Modular Commercial Plant Thesis

The chemical plant of tomorrow and the future : process-
design intensification at different production scales

Citation for published version (APA):
Vural - Gursel, I. (2015). The chemical plant of tomorrow and the future : process-design intensification at
different production scales. Technische Universiteit Eindhoven.
Document status and date:

Published: 23/11/2015
Document Version:
Publisher’s PDF, also known as Version of Record (includes final page, issue and volume numbers)
Please check the document version of this publication:

• A submitted manuscript is the version of the article upon submission and before peer-review. There can be
important differences between the submitted version and the official published version of record. People
interested in the research are advised to contact the author for the final version of the publication, or visit the
DOI to the publisher's website.
• The final author version and the galley proof are versions of the publication after peer review.
• The final published version features the final layout of the paper including the volume, issue and page
numbers.
Link to publication
General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners
and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
• Users may download and print one copy of any publication from the public portal for the purpose of private study or research.
• You may not further distribute the material or use it for any profit-making activity or commercial gain
• You may freely distribute the URL identifying the publication in the public portal.
If the publication is distributed under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license above, please
follow below link for the End User Agreement:
www.tue.nl/taverne
Take down policy

If you believe that this document breaches copyright please contact us at:
openaccess@tue.nl
providing details and we will investigate your claim.
Download date: 25. Feb. 2022

The Chemical Plant of Tomorrow and
the Future
Process-Design Intensification at Different Production Scales
PROEFSCHRIFT
ter verkrijging van de graad van doctor

aan de Technische Universiteit Eindhoven,
op gezag van de rector magnificus prof.dr.ir. F.P.T. Baaijens,
voor een commissie aangewezen door het College voor Promoties,
in het openbaar te verdedigen op
maandag 23 november 2015 om 14:00 uur
door
Iris Vural Gürsel
geboren te Eskisehir, Turkije

Dit proefschrift is goedgekeurd door de promotoren en de samenstelling van
de promotiecommissie is als volgt:
voorzitter: prof.dr.ir. J.C. Schouten

1e promotor: Prof.Dr. V. Hessel
copromotor(en): Dr. Q. Wang
Dr. T. Noël
leden: Prof.Dr.-Ing. N. Kockmann (TU Dortmund)
prof.dr.ir. A.I. Stankiewicz (TU Delft)
Dr. S. Kuhn (KU Leuven)
prof.dr.ir. M. van Sint Annaland
Het onderzoek of ontwerp dat in dit proefschrift wordt beschreven is

uitgevoerd in overeenstemming met de TU/e Gedragscode
Wetenschapsbeoefening.
In process intensification,
the journey is the reward.
Prof. Dr. Frerich J. Keil
This research is funded by the European Research Council (ERC) advanced
grant on “Novel Process Windows – Boosted Micro Process Technology”
under grant agreement number 267443.
Chemical Plant of Tomorrow and the Future

Iris Vural Gürsel
Eindhoven University of Technology, 2015
A catalogue record is available from the Eindhoven University of

Technology Library
ISBN: 978-90-386-3957-4
Printed by Ipskamp Drukkers
Cover design by Rachel van Esschoten, DivingDuck Design
Table of Contents
Summary ix
1. Introduction 1
1.1. Novel Process Windows 1
1.2. Flow Separation and Coupling with Flow Reaction 5
1.2.1. Flow separation units 5
1.2.2. Coiled flow inverter 8
1.2.3. Multi-step flow synthesis 9
1.3. Process Intensification at Plant Level 11
1.3.1. Modular, compact plants 11
1.3.2. End-to-end process integration 14
1.4. Scope and Outline of the Thesis 15
References 17
2. Potential of Intensification through Flow Processing and Use of Modular 25
Plants to Reduce Development Time
2.1. Introduction 26
2.2. Effect of Intensification Fields and Modular Plants 27
2.2.1. Effect of the transport intensification (1st) field 28
2.2.2. Effect of the chemical intensification (2nd) field 30
2.2.3. Effect of the process-design intensification (3rd) field 31
2.2.4. Effect of the compact, modular plants 32
2.3. Case Scenarios with Stepwise Intensification 33
2.3.1. Fine-chemistry 33
2.3.2. Bulk-chemistry 37
2.4. Conclusion 39
References 40
3. Connected Reaction and Metal Separation in Flow – Drop-in Process 45
Integration at Lab Scale
3.2. Homogeneous Metal Catalyst Separation/Recycle in Flow 48
3.3. Experimental 50
3.3.1. Experimental setup of microseparation unit 50
3.3.2. Experimental setup of one-pot CuAAC click reaction and inline 51
copper scavenging in flow
3.4. Results and Discussion 52
v
3.4.1. Performance characterization of the flow liquid-liquid extraction 52
unit
3.4.2. Continuous copper scavenging using flow liquid-liquid 54
extraction unit
3.4.3. One-pot CuAAC click reaction and inline metal scavenging in 58
flow
3.5. Conclusion 59
3.6. Appendix 60
3.6.1. General chemicals information 60
3.6.2. General analytical information 60
3.6.3. Typical experimental procedure for performance characterization 60
of the flow liquid-liquid extraction unit
3.6.4. Typical experimental procedure for continuous copper 61
scavenging using flow liquid-liquid extraction unit
3.6.5. Typical experimental procedure for recovery of EDTA 61
3.6.6. Typical experimental procedure for one-pot CuAAC click 62
reaction and inline copper scavenging in flow
References 62
4. Scale-up of Flow Extraction Utilizing Coiled Flow Inverter – Drop-In 67
Process Integration at Pilot Scale
4.2. Coiled Flow Inverter 70
4.3. Experimental 71
4.3.1. Model systems used 71
4.3.2. Experimental setup 72
4.3.3. Flow separation devices 73
4.4.1. Liquid – liquid two-phase flow patterns in CFI 74
4.4.2. Extraction performance with variation of aqueous phase volume 77
fraction at fixed total flow rate
4.4.3. Extraction efficiency with variation of flow rate and comparison 78
of CFI with straight tube
4.4.4. Performance of flow separators 82
4.5. Conclusion 89
4.6. Appendix 91
4.6.1. General chemicals information 91
4.6.2. General analytical information 91
4.6.3. Investigation of liquid-liquid two-phase flow patterns 91
4.6.4. Determination of extraction efficiency 91
vi
4.6.5. Comparison with straight tube 92
4.6.6. Measurement of separator performance 92
4.6.7. Calibration curves for GC-FID analysis 93
4.6.8. Thermodynamic equilibrium concentration curves 93
References 94
5. Compact Modular Plants for Fine Chemistry – Modular Process 99
Integration at Pilot Scale
5.1.1. Need for faster market and product launch – “window of 100
opportunity”
5.1.2. Current chemical plant concepts 101
5.1.3. Modular plants, compact and pre-manufactured 101
5.1.4. Micro process technology 104
5.1.5. Cash flow analysis for chemical plants 106
5.2. Methodology 107
5.2.1. Demonstration processes and market applications considered 107
5.2.2. Capital investment estimation 109
5.2.3. Operating cost estimation 111
5.2.4. Cash flow analysis 113
5.3.1. Capital investment estimation of fine chemical 115
5.3.2. Operating cost estimation of fine chemical 116
5.3.3. Cash flow analysis of fine chemical 118
5.3.4. Comparison of chemical product classes – bulk chemical, fine 121
chemical and pharmaceutical
5.4. Conclusion 125
5.5. Appendix 127
References 132
6. Decentralized Biofuel Production – Modular Process Integration at 135
Commercial Scale
6.3.1. Process simulation 137
6.3.2. Mass, carbon and energy balance 140
6.3.3. Cost analysis 140
6.4. Conclusion 143
References 143
vii
7. Bulk Chemistry Process Simplification – End-to-end Process Integration 145
at Commercial Scale
7.2.1. Process model development 149
7.2.2. Energy analysis 149
7.2.3. Cost analysis 150
7.2.4. Environmental analysis – LCA methodology 150
7.3.1. Process simulation 151
7.3.2. Energy requirements 154
7.3.3. Equipment cost estimation 155
7.3.4. Life cycle assessment 157
7.4. Conclusion 161
References 162
8. Bulk Chemistry Heat Integration – End-to-end Process Integration at 165
Commercial Scale
8.3.1. Data extraction 170
8.3.2. Defining energy targets 170
8.3.3. Initial heat exchanger network 171
8.3.4. Designing heat exchanger network with pinch analysis 171
8.4. Design Considerations – Compact Heat Exchangers 175
8.4.1. Commercial microchannel-based heat exchanger examples 175
8.4.2. The benefits of compact heat exchangers 176
8.4.3. Heat exchanger network design change implications 178
8.5. Conclusion 180
8.6. Appendix 182
References 183
9. Conclusions and Recommendations 185
9.1. Conclusions 185
9.2. Recommendations 190
References 193
List of Publications 195
Acknowledgements 199
About the Author 202
viii
Summary
Chemical Plant of Tomorrow and the Future – Process-Design Intensification at
Different Production Scales
Vision: Europe’s chemical industry is under pressure to find sustainable and cost-efficient
solutions to improve its competitiveness in Europe and worldwide, The European Union
with its public-private partnerships in the FP7 and SPIRE/H’2020 programs has started top-
down to bring to reality the vision of the “Factory of the Future/Future Chemical
Manufacturing” which considers development of compact, modular plants with embedded
process intensified equipment. A major driver came from the existing modularization
assembly technology. Additionally, a step change improvement can be realized with
process intensification. Micro process technology is one of the key enablers and it is
implemented within the modular plants. Meanwhile, several of such compact plants are in
pilot operation and one is even used for production in industry. Aim of this research is to
complement such step-changes in chemical production by a bottom-up approach and look
for chances of intensification of plants.
Moreover, a holistic and systemic approach should be governed which gives focus to the
chemical process as a whole, which does allow quantum leaps through unifying
contradictions to a new window of opportunity (as the TRIZ method does for systematic
development of innovative patent ideas). With a new field of process-design intensification
under Novel Process Windows this can be achieved. It heads for integrated and simplified
smart-scale flow process design. It is beyond the existing transport intensification field that
considerably improves mass and heat transfer and increasingly explored chemical
intensification field that uses highly intensified, unusual and typically harsh process
conditions to boost micro-processing. Process-design intensification can make use of
transport and chemical intensification when transferring these to full-process level and thus
making the impact holistic. The use of these three intensification fields of flow processing
together with the use of modular plants can enable reduction in number of apparatus in the
plant, a reduction in apparatus size, and a higher predictability in the scale-out of the
apparatus. All these have potential for (much) decreased overall development time. This is
discussed in Chapter 2 with generic impact analysis.
One possibility to bring step-change improvement in chemical production such as
pharmaceutical production is to enable switch from batch to all flow synthesis. It was seen
that further development of flow separation units is required to achieve multi-step synthesis
in flow. The coupling of microseparation units with microreactors is not rare anymore, but
still far from being commonly applied. In this regard, Chapter 3 describes the development
ix
of a flow liquid-liquid extraction unit that enables metal separation in flow and its coupling
with a metal catalyzed flow reactor.
Connected Reaction and Metal Separation in Flow – Drop-in process integration at lab
scale (Ch 3): Increasing usage of metal catalyzed reactions calls for efficient removal of
metal traces from the product. Especially for pharmaceutical substances this purification is
vital to meet stringent requirements. Therefore, a continuous metal separation unit was
developed and optimized which enables liquid-liquid extraction based on slug flow using
chelating agents and phase separation with a porous fluoropolymer membrane. It was
coupled to the copper-catalyzed azide-alkyne cycloaddition reaction studied in one-pot
eliminating the need to isolate and handle potentially explosive azide. The triazole product
was attained in flow with yield of up to 92% and the level of purity requirement of 15 ppm
was met in one stage of extraction.
The vision is to achieve such multi-step synthesis in flow on pilot scale. This calls for scale-
up of separation units. This study is rare in micro flow literature and community and thus
conducted here as explained in Chapter 4 for a liquid –liquid extraction unit.
Scale-up of Flow Extraction Utilizing Coiled Flow Inverter – Drop-in process integration at
pilot scale (Ch 4): A milli-scale flow extraction setup was developed using coiled flow
inverter (CFI, 3.2 mm ID, 210 cm length) and a phase separator. Two separator concepts
were investigated: membrane-based and wettability-based. The performance of the setup
was tested for two model systems that differ in terms of interfacial tension. For the first
system flow rate up to 7.2 l/h was achieved. It is important to point out that slug flow
regime that is limited to microfluidic systems was maintained even at such pilot scale (58
m3/y). This is the highest flow rate reported in literature for slug flow and also for phase
separation using capillary force for immediate separation of phases. For both systems
extraction efficiency close to thermodynamic equilibrium was attained. Compared to a
straight tube, using CFI showed 20 % higher performance at high flow rates due to
intensified secondary flow and inversions.
Another possibility to bring step change improvement in chemical production is the use of
modular plants. They are seen as key to faster market introduction of a new product (50%
idea) by shortening time-to-market resulting in higher cash flows. In Chapter 5, detailed
cost calculations are done to quantify the benefits of modular plants and their synergy with
micro process technology.
Compact Modular Plants for Fine Chemistry – Modular process integration at pilot scale
(Ch 5): The economic benefits of modular plants and their combination with process
intensified smart-scaled equipment was studied at the example of the EcoTrainer of Evonik
Industries. A detailed cash-flow analysis was carried out for the fine chemical of 2,4-
dihydroxy benzoic acid at 200 t/y capacity. It revealed that the new modular plant
technology has a faster payback and a higher earning as compared to conventional
x
technology; in particular when serving high-priced markets and the combination with micro
process technology is advantageous.
Modular compact plants can also enable decentralized production, this is studied in Chapter
6 with biofuel production taken here as application case. It involves an initial techno-
economic analysis to analyze the feasibility of a novel process for biofuel production.
Decentralized Biofuel Production – Modular process integration at commercial scale (Ch
6): Biomass being widely distributed and its composition differing from location and
season, requires biofuel production to be done locally in small-scale. Decentralized
production in modular compact plants was taken to benefit from flexibility, scalability and
standardization enabling savings in engineering and construction expenditure. Process
integration and intensification was achieved by coupling of reforming of biogas and bio oil
in a microchannel plate heat exchanger. Complex separation steps were eliminated by
production via methanol. Cost calculations showed that these benefits can outweigh the
higher investment cost of decentralized plants. The calculation of cost of production of
gasoline revealed that this biofuel production process can be economically competitive with
petroleum-based and Fischer-Tropsh process.
Finally, a step change improvement can be realized by bringing direct chemistries on a full-
process level. Direct chemistries have the potential to reduce the cascades of separation
units. The demonstration example taken here is the direct adipic acid synthesis which is an
important intermediate for the nylon industry. Chapter 7 describes the investigation of the
process simplification potential of this direct route by cost, energy and environmental
analysis. It also includes comparison with the commercial two-step route to benchmark the
flow process against conventional technology.
Bulk Chemistry Process Simplification – End-to-end process integration at commercial
scale (Ch 7): The direct adipic acid synthesis route is through the oxidation of cyclohexene
with hydrogen peroxide. It offers an innovative alternative to the commercial route that has
two-steps. The use of microreactor was considered to safely handle hydrogen peroxide and
to fasten reaction rate. To propose the flow diagram and conditions for this direct route, the
reaction characteristics were considered together with the downstream equipment of the
two-step commercial route. Process simulation was done and a comparison with two-step
commercial route revealed profound simplification due to requirement of much less process
units. This resulted in total purchase cost of equipment to be halved although a more
advanced flow reactor was used. Through simplification or elimination of energy intensive
separation units, energy requirement was reduced significantly. Life cycle assessment
showed that for a number of impact categories the direct process is greener; yet there are
also categories for which the conventional route is more environmentally sustainable.
In Chapter 8 an energy efficient process is searched with the application of heat integration
for the novel direct micro-flow route of adipic acid synthesis introduced in the previous
xi
chapter. Also, the importance of heat exchanger selection is highlighted with information
regarding compact (including microchannel-based) heat exchangers.
Bulk Chemistry Heat Integration – End-to-end process integration at commercial scale (Ch
8): To design an energy efficient heat exchanger network for this direct route pinch analysis
was employed. Aspen Energy Analyzer was used to design the improved heat exchanger
network. Compared with the initial heat exchanger network where energy requirements are
supplied with utility streams, the improved heat exchanger network designed enabled up to
70 % saving in utility cost which enabled to pay back the extra capital cost requirement in
less than one year. Further, it was shown that with the use of compact heat exchangers, due
to their higher heat transfer coefficient, reduced space requirement and increased safety
compared to shell and tube heat exchangers, it is possible to achieve operating cost as well
as capital cost reduction which can add to the benefits achieved by heat integration
Overall, this research provides insight at different production scales what chemical plant of
tomorrow and the future can look like. The main conclusions of this thesis are summarized
and recommendations are provided in Chapter 9.
xii
CHAPTER 1
Introduction
1.1. Novel Process Windows
The chemical industry is under pressure due to strong competition with emerging markets
(especially in the Near East and Asia), increasing product requirements demanded by the
customer and increasing environmental and social legislations.1–3 Therefore, there is a
strong demand for sustainable and cost-efficient solutions. Often this cannot be solved
anymore by process optimization since many processes are performed almost always at
utmost efficiency. Entirely rethinking of the chemical process and its process development
is required.4,5 A step change improvement can be achieved by use of new apparatus and
new approaches in processes. That is found in an entire new way of conceptual thinking in
chemistry and chemical engineering, which manifests in process intensification,6–11 green
chemistry and green engineering12–16. Process intensification presents a set of radically
innovative principles in process and equipment design, which can bring significant benefits
in terms of process and chain efficiency, capital and operating expenses, quality, wastes,
process safety, and more as defined by the European Roadmap for Process Intensification.5
Green chemistry and green engineering approaches bring sustainable solutions in a different
way. Green chemistry promotes new chemical synthesis strategies through use of
renewable and other ‘green’ (air as oxidant, visible light for energy) resources, gives
emphasis on safety, and reduction of waste by yield and selectivity improvements. Green
engineering deals with the design of novel equipment, new production methods, or new
operating modes. 17
Novel Process Windows (NPW) can serve as new technological concept to all three
strategic approaches (Figure 1.1).17–22 It considers the use of high temperature, high
pressure, high concentration (solvent-free), new chemical transformations, explosive
conditions and process simplification and integration to boost synthetic chemistry on both
the laboratory and production scale.22 These could hardly be exploited so far since it would
be beyond the capability of conventional equipment. NPW relate to creating new kinds of
intensification as follows with three kinds of process intensification fields. Those are
transport, chemical and process-design intensification. Latter two constitute NPW.
Chemical intensification is an essential part of green chemistry whereas process-design
intensification is a strong pillar for green engineering with the NPW combining both for
sustainable chemical process development.
1
1. Introduction
NOVEL
PROCESS
WINDOWS
Figure 1.1. Schematic representation of Novel Process Windows (adapted from Hessel et
al.22 )
The American Chemical Society Green Chemistry Institute (ACS GCI) Pharmaceutical
Round Table is a multi-industrial platform with the companies Boehringer, Pfizer, Eli Lilly,
Glaxo-SmithKline, DSM, Johnson & Johnson, AstraZeneca, and Merck (US).22 They
developed a list of ten key green engineering research areas: 1)continuous processing, 2)
bioprocesses, 3)separation and reaction technology, 4)solvent selection, 5)process
intensification, 6)life-cycle analysis, 7)integration of chemistry and engineering, 8)scale-up,
9)process energy intensity, and 10)mass and energy integration.23 Micro process technology
is a major force in the switch of processes from batch to flow and thus is given the top 1
priority. The study of microreactors falls as well under the key areas of 3 and 5. Key areas 2
and 4 refer to chemical intensification and 6 to 10 refer to process-design intensification.
Micro process technology allows processing far from state of the art. It enables to reach
conditions which are not accessible, or which are at least unscalable with conventional
equipment.17 Microreactors are not the only modern devices to achieve process
intensification. Other notable examples are spinning-disk, membrane and microwave
/ultrasound/plasma reactors and reactive separation, hybrid separation units.8–11,24–28
Nevertheless, microreactors are considered to be the most potent and typical example.22
Therefore the concept of NPW is considered with micro process technology here.
2
1.1. Novel Process Windows
The transport intensification field is given by the miniaturization of flow dimensions which
considerably improves mass and heat transfer.29,30 The reduction or elimination of mass
transfer hindrance increases (effective) reaction rates; the ability to heat up and cool down
fast keeps residence times as short as chemically needed and thus improves productivity;
the latter is also given by the ability to operate even highly exothermic reactions with large
heat release due to superior cooling (impacting selectivity as well by lowering side
reactions).19,31,32 All of this allows to reach high space-time yields and gives thus rise to
compact reactors and plants.33,34 In addition, transport intensification and the use of short
residence times, almost at the kinetic limit (intrinsic kinetics), allow to lower the share of
parallel and follow-up reactions, respectively, enhancing selectivity.35 Furthermore the
small level of hold-up in microreactors reduces the damage potential of toxic or explosive
reagents. The improved safety in microreactors may even be the door-opener to further
scale-up development otherwise prohibited.36 The impact of efficient mixing and heat
transfer, short residence time accompanied by the improved safety within microchannels
has already been demonstrated by numerous successful research groups.37–49 After enabling
this development from a batch protocol into a continuous flow setup in the laboratory,
scale-up to achieve commercial production was also studied.50–57
The chemical intensification field refers to using highly intensified and typically harsh
process conditions to exploit faster kinetics.21,22 Massive increases in temperatures,
pressures and/or concentration can speed up reactions by orders of magnitude and reaction
times shrink from hours to minutes and even to seconds.19 Meanwhile there are numerous
examples for the high temperature flow processing (most often conducted in the
superheated regime) also in combination with high pressure.58–65 The mechanical stability
of microstructured devices allows one to run processes in a high pressure regime in safety.20
Apart from facilitating superheated conditions, high pressure affects the reaction through
the influence on the reaction equilibrium and rate constants and reaction medium via the
effect on its physical properties.66–70 There are also some examples for the concentration
effect.71–75 Solvent-less and -free operations profit from increased concentration of the
reactants and the reduced solvent load reduces waste and simplifies/eliminates of process
steps for the removal of solvents. In addition, the use of alternative solvents (ionic liquids,
supercritical fluids, fluorous fluids) allow to tailor the reaction environment in terms of
diffusivity, separation capability and dissolution power.76–79 Besides speeding up of
chemistry, a new chemical transformation based on reduced or combined elemental steps is
also considered as chemical intensification (also as this has large chance to change the
process design, see below). This includes direct syntheses80–82 and multi-step syntheses in
“one-flow” (in analogy to “one-pot”)83–85. Speeding-up chemistry is a means to make
chemistries happen which are otherwise to slow. Changing chemistry is finally the key to
sustainability, for efficient use of materials and reduction of waste.
3
1. Introduction
The process-design intensification field is a third means and is based on process

simplification and integration. While these are common engineering principles also for
conventional processes, simplification in flow can reach much further, due to the power of
the chemical intensification mentioned above. Similarly, process integration can come to
new horizon in flow, as the miniaturization can lead to entire new level of compactness, in
the way microelectronics have impressively demonstrated.
Yet, there is a (current) drawback about this which can hamper implementation. Flow
processes with micro- or other smart-scale reactors are different in design than conventional
reactor systems. Micro flow process design is currently not covered and supported through
the existing education / training and through its plant- and process design software such as
Aspen for the following reasons86,87 :
 No standard modules exist – the (by orders-of-magnitude) intensified components
often do not match to each other and to existing process units and control structure.
 As a result of the aforementioned, no additive (plug & play-type) design from units
into a process-chain exists – rather design is interactive: each flow unit development
has again the potential of changing the whole plant design.
 Flow process design is reaction dominant, whereas today’s plants are utility- and
separation dominated – with totally different cost, energy and process control
structure.
 Integration of process units has to be considered as a new step in between the process
unit and plant level in process development into production.
In view of the last statement, in a more advanced, but not so distant step, due to the
compactness of the basic elements within the units and high speed of processing, integrated
reactors will partly replace consecutive arrangement of otherwise separated units – which is
process integration. Already today heating and reaction or exothermic and endothermic
reactions are combined in microreactors.88–90 This is already largely given for the field of
fuel processing utilizing microreactors comprising energy and mass flow integration.91 In
fine chemistry, this is not so prominent so far. A logical step is to combine separation and
reaction, and actually that has been started with the use of membranes.92–95 Another
eminent integration issue is to incorporate functionalities such as process control and
analysis which can enable optimum operation of the system.96,97
In continuation of the latter, as discussed above new chemical transformations have large
chance to change the process design. A direct (one-step) synthesis can largely eliminate
separation equipment needed for otherwise consecutively done reactions – which is process
simplification.86,87,98 If not eliminated, there is a chance for replacing separation units by
simpler and more inexpensive units. Other chances for process simplification are provided
by elimination of heating/cooling units through achieving room temperature processing at
same productivity, purification units due to a decrease in the levels of by-product to traces
through increased selectivity and recycle loops through higher reaction yields.18,99
4
1.2. Flow Separation and Coupling with Flow Reaction

1.2.1. Flow separation units
The majority of microchemical synthesis applications were so far restricted to one or few
reaction steps with the separation done offline.33,44,46 Therefore, further development of
microseparation units is needed. From about 2011 onwards, there is an increasing number
of developed microseparation devices for chemical purposes.100 Still most of these devices
have been characterized as stand-alone tools, i.e. without connection to a flow reaction
device. These devices were mainly liquid-liquid extraction units. Recently, several
distillation 101–104, absorption 105–107 and chromatography 108–110 units were also described.
Since the early investigations in micro process technology, microextraction has received a
considerable amount of attention. With miniaturization, a high surface area to volume ratio
leads to enhanced mixing efficiency and mass transfer between the two phases.35 Several
microextraction devices are found differing according to how immiscible phases contact
each other. With the use of different mixing elements and process conditions (e.g. velocity
of the phases, velocity ratio of the phases, physical properties of the phases) different flow
patterns are observed.111 Generally four flow patterns can be observed in microcapillaries,
i.e. the parallel flow, the slug flow, the droplet flow, and the dispersed flow (Figure 1.2).
Figure 1.2. Liquid-liquid flow patterns in microchannels (adapted from Holbach and
Kockmann112); a) parallel flow; b) slug flow; c) droplet flow; d) dispersed flow.
The flow pattern used has a large effect on the mixing and separation performance.113 Slug
flow was shown to have better mass transfer compared to parallel flow.114 This is due to
5
1. Introduction
slug flow providing convection through internal circulations as well as interfacial diffusion
between two adjacent slugs.111 Mass transfer in dispersed flow was seen to be the most
effective due to larger interfacial area.115 Several micromixer types have been evaluated for
this purpose such as interdigital116–118 and split-and-recombine119,120 type micromixers.
They can achieve even higher specific interfaces than slug flow does by virtue of the small
droplets. However, the phase separation performance should be taken into consideration
when evaluating the whole extraction process. Emulsions are often very stable and long
time may be required for the coalescence of produced droplets.115 On the other hand, phase
separation of slug flow can be much easily achieved in short times. Therefore, slug flow has
found wide practical use in flow applications.
There has been extensive research on slug flow especially within studies of identifying
liquid-liquid two-phase flow regimes. The mass transfer performance using different types
of microchannels for various non-reacting and reacting systems were investigated and
pressure drop measurements were made.121–128 Slug flow was also applied for liquid-liquid
extraction applications.111,114,115,129–131 In some of these applications phase separation was
also achieved in flow. Some studies were focused on the development of phase separation
units for separating the aqueous and organic phases of the slug flow.132–135 Effective liquid-
liquid slug flow separation was achieved with PTFE membrane or microdevice that
selectively lets the organic phase pass through.129,132,133,135 Also, it was achieved with
wettability-based separation units with branched hydrophobic and hydrophilic
outlets.111,125,134 Review of most prominent literature on slug flow is given in Table 1.1 with
information besides others the maximum flow rate achieved with slug flow regime. So far,
slug flow has only been realized on a laboratory scale. Highest flow rate of 4.8 L/h was
presented by Cervera-Padrell et al.135 for the application of phase separation of toluene and
water. This research enables operation up to 7.2 L/h in slug flow regime as explained in
detail in Chapter 4.
Table 1.1. Review of studies on liquid-liquid slug flow reported in literature
Reference Microchannel Maximum System Separation Application

Flow Rate
Dessimoz T 12 ml/h Hexane- - Flow
et al. d = 400 μm trichloroacetic regime,
H
(2008) 121 glass chip acid- mass
water+NaOH transfer
Burns and Cross 18 ml/h Kerosene-acetic - Mass
Ramshaw d = 380 μm acid- transfer
H
(2001) 122 glass chip water+NaOH
Dummann Y 36 ml/h Aromatic- - Reaction
et al. d = 0.5-1 mm nitration acid
t
(2003) 123 capillary tube
6
Sarkar et Cross 72 ml/h Water-succinic - Flow regime

al. (2012) d = 272 μm acid-n-butanol
124 H
serpentine
channel
glass chip
Ghaini et Y 160 ml/h N-butyl Wettability Mass
al. (2010) d = 0.5-1 mm formate- based transfer
125 t
capillary tube water+NaOH separator
Kashid et T-square: T-square: Water-acetone- - Flow
al. (2011) d = 400 μm 180 ml/h toluene regime,
126 H
T-trapezoidal: T-trap.: mass
d = 400 μm 180 ml/h transfer
H Y-rect.:
Y-rectangular: 120 ml/h
d = 269 μm Concentric:
H
Concentric: 300 ml/h
d = 1600 μm Caterpillar:
t
Caterpillar: 420 ml/h
d = 150 μm
H
Cherlo et T 200 ml/h Kerosene-water - Flow regime
al. (2010) d = 0.54,1mm
127 H
Kashid Y 200 ml/h Water- - Flow

and Agar d =0.25 - cyclohexane regime,
t
(2007) 128 1mm pressure
capillary tube drop
analysis
Kralj et al. d = 255 μm 3 ml/h DMF-DCM- PTFE Extraction
H
(2007) 129 serpentine water membrane
channel DMF-DEE-
silicon chip water
Fries et al. T 18 ml/h Toluene-water- Capillary Extraction
(2008) 114 d = 225 μm vanillin force based
H
PDMS chip separator
Castell et T 24 ml/h Chloroform- PTFE Phase
al. (2008) d = 654 μm water device separation
132 H
PTFE chip
Jovanovic Y 36 ml/h Toluene-2- - Flow
et al. d = 0.25 mm butanol-water regime,
t
(2012) 130 capillary tube mass
transfer,
extraction
Okubo et T 300 ml/h Dodecane- - Extraction
al. (2008) d = 0.5–1 mm phenol-water
115 t
capillary tube
7
1. Introduction
Kashid et Y 120 ml/h Kerosene-acetic Wettability Extraction

al. d = 0.5–1 mm acid-water based
t
(2007) 111 capillary tube Water-iodine- separator
kerosene
Water-succinic
acid-n-butanol
Adamo et T 960 ml/h Hexane-water PTFE Phase
al. (2013) d = 1.6 mm Ethyl acetate- membrane separation
133 t
capillary tube water
Kashid et 6 parallel 1200 ml/h Water- - Flow
al. (2010) Y cyclohexane regime,
131
d = 0.5 mm Kerosene-acetic mass
t
capillary tube acid-water transfer,
extraction
Gaakeer Y 3000 ml/h Water-n- Wettability Phase
et al. d = 1.5 mm heptane based separation
t
(2012) 134 capillary tube separator
Cervera- T 2400 ml/h THF-water PTFE Phase
Padrell et d = 1.6 mm (THF- Toluene-water membrane separation
t
al. (2012) capillary tube water)
135
4800 ml/h
(toluene-
water)
Present Cross 3000 ml/h Butyl acetate- PTFE Extraction
work d = 3.2 mm (butyl water-acetone membrane
t
capillary tube acetate- Toluene-water- /
Coiled flow water- acetone Wettability
inverter acetone) based
7200 ml/h separator
(toluene-
water-
acetone)
1.2.2. Coiled flow inverter

Liquid-liquid extraction consists typically of two process steps, namely mixing and phase
separation. Effective phase separation can be achieved, for example, with PTFE membrane
or wettability based separator units as discussed above. Mixing and contacting of the two
liquid phases, as outlined in Section 1.2.1., has been achieved typically inside a straight
capillary tube (or just slightly bended for reasons of compactness). Using the latter
deliberately and tightly, however, enhanced mixing performance can be achieved with
curved channels that enable generation of vortices.136 Because of the action of centrifugal
force that acts perpendicular to the flow direction of the fluid, a secondary flow develops in
curved channels that is called Dean vortices.137 Dean vortices enhance the radial mixing in
8
case of single-phase flow and thus narrower residence time distribution compared to
straight capillaries can be achieved.138,139
Mixing can be further enhanced by complete flow inversions by inducing 90° bends at
equal intervals along the length of the tube.139 Coiled flow inverter (CFI) is based on this
principle. The direction of the centrifugal forces inducing the formation of secondary flow
profile in helically coiled tube is changed with the 90° bends. Therefore, the secondary flow
profile is developed after the 90° bend in a different plane that is perpendicular to the
previous plane (see Figure 1.3). As a result, the radial mixing can be further enhanced in
CFI. Hence, a narrow residence time distribution close to an ideal plug flow reactor can be
achieved even at laminar flow regimes.140,141
Figure 1.3. CFI structure and cross-sectional velocity contours (adapted from Singh et al.142
and Mandal et al.143)
1.2.3. Multi-step flow synthesis

The fine chemical synthesis typically involves multiple reactions and intermediate
separations. To achieve such synthesis in continuous flow several strategies have been
used. One strategy was telescoping reaction sequences in which one molecule automatically
converts to the next alike a domino cascade .144–146 In many cases however this is not
9
1. Introduction
feasible, e.g., due to impurities formed affecting the following reactions. A second strategy
involves the use of multi-step flow synthesis with individual reactors connected in a row
with the flow separation in between the reaction steps achieved by solid supported reagents
packed into columns in flow. They allow scavenging of unreacted reagents enabling the
product to be collected at the outlet without requirement of further purification. Different
types of such columns can be connected in series allowing multiple reactions to take
place.147–153 However, the necessity to periodically replace the cartridges interrupts the
continuous flow process. Also, they are not found suitable for applications at large scale.154
The final strategy is to implement micro flow separation units together with microreactors
to allow purification in continuous flow. This coupling of different microfluidic unit
operations in continuous flow has great potential for waste and energy minimization. It is
also very beneficial for reactions involving generation of hazardous intermediate products.
However, as discussed above although a great progress has been made to develop the
required flow separation units still more research is required especially to achieve operation
at high throughput. A very few commercial microextraction devices are now available, e.g.
the flow liquid-liquid extraction FLLEX module from Syrris, UK155 and liquid-liquid
separators from Zaiput Flow Technologies156. The coupling of microseparation units with
microreactors is not rare anymore, but still far from being commonly applied. So far, there
are only few demonstrations 42,157–159.
Figure 1.4. Schematic of the microfluidic setup for the multistep synthesis of carbamates157
A good example to illustrate the sequence of functions in a practical multi-step flow system
is provided by a microfluidic reactor network with two reactors and a microfluidic
extraction unit in between was realized for the synthesis of biaryls starting from substituted
phenols.159 Impurities generated in the first reaction are poisonous for the catalyst system of
the second reaction. The intermediate separation step enabled the extraction of the
10
1.3. Process Intensification at Plant Level
impurities with aqueous solvent and the aqueous phase was separated and removed from
the system with phase separation. In another example, a continuous-flow, multistep
microchemical synthesis of carbamates starting from aqueous azide and organic acid
chloride by using the Curtius rearrangement reaction was described (See Figure 1.4).157 It
involved three reaction steps and two separation steps. The first separator enabled
separation of organic and aqueous phases after the first reactor. The second separator was
used to remove the nitrogen gas generated in the second reactor from the liquid stream. As
a final example, a multistep Heck synthesis was achieved with two reaction steps and two
separation steps in between.158 The intermediate was removed by microfluidic extraction
after first reaction. Next, solvent switch was enabled by a microfluidic distillation.

1.3.1. Modular, compact plants
Modular plants were seen as main enablers to reduce development time as discussed at the
48th Tutzing-Symposium with the “Die 50%-Idee – Vom Produkt zur Produktionsanlage in
der halben Zeit” (translated: The 50% idea – from product to production facility in half the
time).160,161 This shortening of the time-to-market can help increase Net Present Value of an
investment and reduce investment risks (see Figure 1.5). This engineering concept of
modularization is increasingly discussed in literature for chemical production.162–167
Figure 1.5. 50%-Idea - from product to production facility in half the time168
Modularization is already established in equipment and instrument manufacturing
industries such as electronics (TV, computer), automotive and the airline industry.169,170
However, modularization is, so far, rarely applied to the production of chemicals.171
11
1. Introduction
Demonstration projects hereto were undertaken in European Union’s FP7 Call for Process
Intensification.172 Such theoretical and experimental first steps are in line with the holistic
ambition and vision of the “Factory of the Future” and “Flexible Future Production
Strategies”, i.e., to combine micro process technology with modular plants.173 This full-
chain ambition of European funding policy is expressed as the realization of “new,
intensified process and plant concepts for speeding up market penetration, for enhancing
the product life-cycle and improving sustainable production”.173 Five large-scale projects
on Future Factories with many of the major industrial players are: F3 Factory, COPIRIDE,
POLYCAT, PILLS, and SYNFLOW.174–178 These projects target basically all the different
chemical industries, including bulk-chemical, fine-chemical and pharmaceutical
manufacturing, as well as applications in polymerization, cosmetics, and renewable
resources. These five Future Factory EU projects supplement each other in focusing on
different aspects of the whole value added chain – starting from new tailored catalyst
development over new reactors towards new types of processing and finally approaching
the plant and systemic process evaluation level. This is esteemed to give the first building
blocks for the sustainable chemical production platforms of the future. ACHEMA 2012 has
put modular plants 1st in the list of top five trends in the process industry.179
These plants are not yet built fully modularly but are built as so called skid mounted units.
The advantages to such manufacturing concept have already been demonstrated for the
construction of pilot plants (such as the ones built by company Zeton B.V.). Within the F3
Factory project seven industrial case studies are demonstrated. The developed process
equipment containers and respective process equipment assemblies are integrated with the
backbone infrastructure services at INVITE facility (see Figure 1.6). This backbone facility
supplies utilities like electricity, heat, cooling water and service gases.
Figure 1.6. Bayer’s Process Equipment Container Unit installed at INVITE (from F3
Factory newsletter174)
12
Among the most developed Future Factory approaches which has approach commercial
level is a compact container-platform plant concept of Evonik Industries, the Evotrainer
(meanwhile renamed as the EcoTrainer), which is designed as multifunctional and all-
purpose plant platform180–182 (see Figure 1.7). The concept orients particularly on the
manufacture of specialty products with tonnage which are at present typically made as
campaign products in conventional multi-product plants. It provides a mobile chemical
development platform which can be upgraded to a production plant of a capacity of 200
metric tons per year.180
Figure 1.7. The Evotrainer: a modern, compact, low-investment plant platform (Courtesy
of Evonik Industries, meanwhile renamed as the EcoTrainer)
The EcoTrainer provides a modern, compact and highly efficient process- and plant concept
covering the whole manufacturing chain to the chemical product in one system – including
reactant, product, by-product supply and storage, reaction and purification.182 This
EcoTrainer concept links research with process development, production and plant
engineering. It is ideally suited for apparatus for process intensification – in particular milli-
and micro process technological apparatus.182
The EcoTrainer is divided into several cabinets for logistics, reaction/purification and
instrumentation, and control.181 Its infrastructure supplies the raw materials, product and
by-product logistics and the utility requirements Every EcoTrainer is tailored to the reaction
that will take place with the processing equipment being adapted to the specific chemical
process.180 It can house a complete chemical plant or just a single module, such as a reactor
or a processing step. It has high flexibility concerning diverse chemical syntheses and
correspondingly diverse products.
13
1. Introduction
1.3.2. End-to-end process integration

In literature and community related to microreactors and other intensified smart-scaled
equipment, it was common to focus on the design of one equipment and testing its
processing – the drop-in (retrofit) concept. However, with process-design intensification we
focus our attention towards the complete chemical process, therefore it is regarded as end-
to-end process integration. It considers the comprehensive evaluation of new, novel
production routes or processing approaches in an early development stage. It involves cost,
energy and environmental analysis based on process simulation. Since it is concerned with
the whole system and considers the interdependence of its parts it is holistic in approach.
Such research is becoming more and more important as evidenced by the increasing
number of holistic process papers in all chemical engineering literature (see Figure 1.8).
The term holistic covers comprehensive, full-chain, multiparameter (criteria), system-
engineering analysis or similar complexity-driven analysis. A paper from the DOW
company entitled Holistic Route Selection is in accordance to this vision.183
Figure 1.8. The number of holistic papers published22

Ex-ante analysis has already been done to aid in process development by enabling
comparison between chemical synthesis routes, different process conditions, raw material
alternatives, etc. for the flow reaction under consideration. Life cycle assessment (LCA)184
is a common method employed to evaluate the environmental impact and it is mostly used
with cost assessment tools. Such evaluation approaches are helpful to provide an early-bird
view, the viability of the intensified units with NPW conditions in terms cost efficiency and
sustainability. Several examples of this analysis are found in the literature for the use of
high temperature and pressure conditions and the use of alternative solvents such as ionic
liquids and supercritical carbon dioxide.17,29,64,185
The process-design intensification through flow processing is new in literature and this
research gives first introduction to it. It has been started to be investigated for the first time
by the author on one reaction example of direct adipic acid synthesis. Apart from the
14
1.4. Scope and Outline of the Thesis
reactor all the downstream processing units is considered in the evaluation. This is exactly
where innovation through the new flow process design has to set in. There is another PhD
project of the ERC Advanced Grant Novel Process Windows (Hessel) supporting it (by
Minjing Shang) with microreactor development to achieve this direct synthesis in flow.82,186
This research gave light to further end-to-end studies including European Union large scale
projects BIOGO and MAPSYN.187,188 Accordingly, now new two PhD projects are
dedicated to such comprehensive analysis under these projects in Hessel’s group.
1.4. Scope and Outline of the Thesis

Process intensification spans to all scales existing in a chemical process, from molecular to
meso- and macroscale. Molecular scale considers the molecules such as molecular
collisions and molecular diffusion. Maximization of the effectiveness of intra- and
intermolecular events is given as a primary principle for process intensification as it can
change the inherent kinetics of chemical reactions. It is the least documented scale so far in
literature. Still there are numerous examples showing the effect of intensification on this
scale as highlighted by Stankiewicz.9 At the macroscale, process units such as reactors are
considered for process intensification which is highly researched and the examples are ever
increasing. As mentioned before microreactors are considered to be the most potent and
typical example for process intensification, as outlined by Hessel. Such reaction-oriented
research in flow chemistry is also studied by other research groups as given in Section 1.1.
On a process level, this resembles the drop-in concept with the microreactor replacing a
conventional reactor.
In this research focus is given more on the process side. There are mainly three parts of
integrated drop-in, modular and end-to-end process integration (Figure 1.9). The first is
called ‘Integrated Drop-in’ and it refers to coupling of microreactor with microseparator in
flow. As described in Section 1.2.3 such multi-step synthesis in flow has very few
demonstrations. Therefore, it is taken in this research and studied in Chapter 3. It describes
connecting metal catalyzed microreaction and metal separation enabled with
microextraction unit in flow to achieve the product separated from the catalyst. The vision
is to achieve multi-step synthesis in flow on pilot scale. The scale-up of microreaction units
has been demonstrated. However, there is very limited research on microseparation units
and the study of their scale-up is very rare. Thus it is conducted here in Chapter 4 for a
flow extraction unit utilizing coiled flow inverter and a phase separator.
The second part of research considers the use of the new compact, modular container
technology and it is referred to as ‘Modular’ process integration. There are known benefits
of such modular production platform such as saving in the design and construction time and
expenditure. An analysis was lacking in the literature to quantify these benefits in terms of
impact on costs. Therefore, in Chapter 5 a detailed discounted cash flow analysis is done
15
1. Introduction
to estimate how these advantages of modular plants are reflected in the Net Present Value.
Furthermore, the combined benefits of process intensification enabled by smart-scaled
process-intensified reaction equipment and of modular compact plants was analyzed with
six scenarios for a fine chemical production. The modular plants are also beneficial because
of their flexibility and scalability. Therefore, they are ideally suited for decentralized
production. This is evaluated with application case of decentralized biofuel production in
Chapter 6. It involves an initial techno-economic analysis to analyze the feasibility of a
novel process for biofuel production.
The third part of research is on comprehensive assessment of the flow process as a whole. It
provides an ‘End-to-end’ vision of intensified process design. It is performed for the
reaction example of direct adipic acid synthesis. A step change improvement can be
realized by bringing such direct chemistries on a full-process level. The process
simplification potential of the direct adipic acid synthesis route is investigated in Chapter 7
by cost, energy and environmental analysis. In Chapter 8 an energy efficient process is
searched with the application of heat integration for the novel direct micro-flow route of
adipic acid synthesis introduced in the previous chapter.
The use of the three intensification fields introduced in Section 1.1 together with the use of
modular plants can enable reduction in number of apparatus, reduction in apparatus size,
and higher predictability in the scale-out of the apparatus. All these have potential for
(much) decreased overall development time which is discussed in Chapter 2.
Figure 1.9. A schematic representation of the chapters in the thesis
16
References
References
1. High-Level Group on the Competitiveness of the European Chemicals Industry,
Final Report: European Chemical Industry, Enabler of a Sustainable Future,
European Commission, Luxembourg, 2009.
2. P. Harnick, The Future of the European Chemical Industry, Publication no.
1001503, KPMG International, Brussels, 2010.
3. K. Griesar, EU vs China: Challenges and Opportunities for European Companies.
The Future of the Chemical Industry, ACS Symposium Series, Vol. 1026, Chapter
3, 47-49, Washington DC, 2009.
4. J. C. Charpentier, Chemical Engineering Journal, 2007, 134, 84–92.
5. European Roadmap on Process Intensification, Creative Energy, SenterNovem,
The Hague, The Netherlands, 2007.
6. A. I. Stankiewicz and J. A. Moulijn, Chemical Engineering Progress, 2000, 96, 22–34.
7. J. R. Burns, J. N. Jamil, and C. Ramshaw, Chemical Engineering Science, 2000, 55,
2401–2415.
8. A. Stankiewicz, Chemical Engineering Research and Design, 2006, 84, 511–521.
9. T. Van Gerven and A. Stankiewicz, Industrial & Engineering Chemistry Research,
2009, 48, 2465–2474.
10. P. Oxley, C. Brechtelsbauer, F. Ricard, N. Lewis, and C. Ramshaw, Industrial &
Engineering Chemistry Research, 2000, 39, 2175–2182.
11. A. Stankiewicz, Chemical Engineering and Processing: Process Intensification,
2003, 42, 137–144.
12. P. T. Anastas and J. Warner, Green Chemistry: Theory and Practice, Oxford
University Press, Oxford, 1998.
13. C. Ramshaw, Green Chemistry, 1999, 1, G15.
14. S. J. Haswell and P. Watts, Green Chemistry, 2003, 5, 240–249.
15. B. P. Mason, K. E. Price, J. L. Steinbacher, A. R. Bogdan, and D. T. McQuade,
Chemical reviews, 2007, 107, 2300–18.
16. D. J. C. Constable, A. D. Curzons, and V. L. Cunningham, Green Chemistry, 2002,
4, 521–527.
17. V. Hessel, D. Kralisch, and U. Krtschil, Energy & Environmental Science, 2008, 1,
467–478.
18. V. Hessel, Chemical Engineering & Technology, 2009, 32, 1655–1681.
19. V. Hessel, B. Cortese, and M. H. J. M. de Croon, Chemical Engineering Science,
2011, 66, 1426–1448.
20. S. Borukhova and V. Hessel, in Process Intensification for Green Chemistry, eds.
K. Boodhoo and A. Harvey, John Wiley & Sons, Ltd, Chichester, UK, 2013.
21. T. Illg, P. Löb, and V. Hessel, Bioorganic and Medicinal Chemistry, 2010, 18,
3707–3719.
22. V. Hessel, D. Kralisch, N. Kockmann, T. Noël, and Q. Wang, ChemSusChem,
2013, 6, 746–89.
17
1. Introduction
23. C. Jimenez-Gonzalez, C. S. Ponder, Q. B. Broxterman, and J. B. Manley, Organic

Process Research and Development, 2011, 15, 912–917.
24. C. Ramshaw and S. Cook, Chemical Engineer, 2005, 774, 42–44.
25. M. Meeuwse, J. Van Der Schaaf, and J. C. Schouten, Industrial and Engineering
Chemistry Research, 2010, 49, 1605–1610.
26. F. Visscher, J. van der Schaaf, M. H. J. M. de Croon, and J. C. Schouten, Chemical
Engineering Journal, 2012, 185-186, 267–273.
27. C. Buchaly, P. Kreis, and A. Górak, Chemical Engineering and Processing:
Process Intensification, 2007, 46, 790–799.
28. A. Górak and A. Stankiewicz, Annual Review of Chemical and Biomolecular
Engineering, 2011, 2, 431–451.
29. D. Kralisch, I. Streckmann, D. Ott, U. Krtschil, E. Santacesaria, M. Di Serio, V.
Russo, L. De Carlo, W. Linhart, E. Christian, B. Cortese, M. H. J. M. de Croon, and
V. Hessel, ChemSusChem, 2012, 5, 300–11.
30. J. C. Charpentier, Chemical Engineering and Technology, 2005, 28, 255–258.
31. W. Ehrfeld, V. Hessel, and H. Lowe, Microreactos: New Technology for Modern
Chemistry, Wiley-VCH, Weinheim, 2000.
32. K. Geyer, J. D. C. Codée, and P. H. Seeberger, Chemistry - A European Journal,
2006, 12, 8434–8442.
33. V. Hessel and H. Löwe, Chemical Engineering & Technology, 2003, 26, 531–544.
34. V. Hessel, S. Hardt, and H. Lowe, Chemical Micro Process Engineering:
Fundamentals, Modelling and Reactions, Wiley-VCH, Weinheim, 2004.
35. V. Hessel, A. Renken, J. C. Schouten, and J. Yoshida, Micro Process Engineering
a Comprehensive Handbook, Wiley-VCH, Weinheim, 2009.
36. X. Zhang, S. Stefanick, and F. J. Villani, Organic Process Research and
Development, 2004, 8, 455–460.
37. T. Noël and V. Hessel, ChemSusChem, 2013, 6, 405–407.
38. T. Noël and S. L. Buchwald, Chemical Society reviews, 2011, 40, 5010–29.
39. R. L. Hartman, J. P. McMullen, and K. F. Jensen, Angewandte Chemie
International Edition, 2011, 50, 7502–7519.
40. J. Wegner, S. Ceylan, and A. Kirschning, Advanced Synthesis and Catalysis, 2012,
354, 17–57.
41. C. G. Frost and L. Mutton, Green Chemistry, 2010, 12, 1687–1703.
42. R. L. Hartman and K. F. Jensen, Lab on a chip, 2009, 9, 2495–507.
43. S. V Ley and I. R. Baxendale, Chimia, 2008, 62, 162–168.
44. C. Wiles and P. Watts, European Journal of Organic Chemistry, 2008, 2008, 1655–
1671.
45. A. Kirschning, W. Solodenko, and K. Mennecke, Chem. Eur. J., 2006, 12, 5972–90.
46. K. Jähnisch, V. Hessel, H. Löwe, and M. Baerns, Angewandte Chemie
47. K. F. Jensen, Chemical Engineering Science, 2001, 56, 293–303.
18
References
48. C. Wiles and P. Watts, Micro Reaction Technology in Organic Synthesis, CRC
Press, Taylor & Francis Group, LLC, Boca Raton, 2011.
49. T. Wirth, Ed., Microreactors in Organic Synthesis and Catalysis, Wiley-VCH,
Weinheim, 2008.
50. A. Tonkovich, D. Kuhlmann, A. Rogers, J. McDaniel, S. Fitzgerald, R. Arora, and
T. Yuschak, Chemical Engineering Research and Design, 2005, 83, 634–639.
51. E. V. Rebrov, I. Z. Ismagilov, R. P. Ekatpure, M. H. J. M. de Croon, and J. C.
Schouten, AIChE Journal, 2007, 53, 28–38.
52. M. Freemantle, Chemical & Engineering News Archive, 2003, 81, 36–37.
53. N. Kockmann, M. Gottsponer, B. Zimmermann, and D. M. Roberge, Chemistry - A
European Journal, 2008, 14, 7470–7.
54. N. Kockmann, M. Gottsponer, and D. M. Roberge, Chemical Engineering Journal,
2011, 167, 718–726.
55. N. Kockmann and D. M. Roberge, Chemical Engineering & Technology, 2009, 32,
1682–1694.
56. N. Kockmann and D. M. Roberge, Chemical Engineering and Processing: Process
Intensification, 2011, 50, 1017–1026.
57. P. Barthe, C. Guermeur, O. Lobet, M. Moreno, P. Woehl, D. M. Roberge, N. Bieler,
and B. Zimmermann, Chemical Engineering & Technology, 2008, 31, 1146–1154.
58. T. Razzaq and C. O. Kappe, Chemistry, an Asian journal, 2010, 5, 1274–89.
59. T. Razzaq, T. N. Glasnov, and C. O. Kappe, Chemical Engineering & Technology,
2009, 32, 1702–1716.
60. T. Razzaq, T. N. Glasnov, and C. O. Kappe, European Journal of Organic
Chemistry, 2009, 2009, 1321–1325.
61. M. Damm, T. N. Glasnov, and C. O. Kappe, Organic Process Research &
Development, 2010, 14, 215–224.
62. C. O. Kappe and E. Van der Eycken, Chemical Society reviews, 2010, 39, 1280–1290.
63. V. Hessel, C. Hofmann, P. Lob, J. Lohndorf, H. Lowe, and A. Ziogas, Organic
Process Research & Development, 2005, 9, 479–489.
64. F. Benaskar, V. Hessel, U. Kxtschil, P. Lob, and A. Stark, Organic Process
Research and Development, 2009, 13, 970–982.
65. A. C. Varas, T. Noël, Q. Wang, and V. Hessel, ChemSusChem, 2012, 5, 1703–7.
66. F. Trachsel, C. Hutter, and P. von Rohr, Chemical Engineering Journal, 2008, 135,
S309–S316.
67. K. A. Swiss and R. A. Firestone, J. Phys. Chem. A, 2000, 104, 3057–3063.
68. R. M. Tiggelaar, F. Benito-Lopez, D. C. Hermes, H. Rathgen, R. J. M. Egberink, F.
G. Mugele, D. N. Reinhoudt, van den A. Berg, W. Verboom, and H. J. G. E.
Gardeniers, Chemical Engineering Journal, 2007, 131, 163–170.
69. F. Benito-López, R. J. M. Egberink, D. N. Reinhoudt, and W. Verboom,
Tetrahedron, 2008, 64, 10023–10040.
70. S. Borukhova, A. D. Seeger, T. Noël, Q. Wang, M. Busch, and V. Hessel,
ChemSusChem, 2015, 8, 504–12.
19
1. Introduction
71. P. Loeb, V. Hessel, H. Klefenz, H. Lowe, and K. Mazanek, Lett. Org. Chem., 2005,
2, 767–779.
72. S. Taghavi-Moghadam, A. Kleemann, and G. Golbig, Organic Process Research &
Development, 2001, 5, 652–658.
73. H. Löwe, V. Hessel, P. Löb, and S. Hubbard, Organic Process Research &
Development, 2006, 10, 1144–1152.
74. D. J. C. Constable, C. Jimenez-Gonzalez, and R. K. Henderson, Organic Process
75. S. Borukhova, T. Noël, B. Metten, E. de Vos, and V. Hessel, ChemSusChem, 2013,
6, 2220–2225.
76. F. Trachsel, B. Tidona, S. Desportes, and P. Rudolf von Rohr, The Journal of
Supercritical Fluids, 2009, 48, 146–153.
77. X. Han and M. Poliakoff, Chemical Society reviews, 2012, 41, 1428–36.
78. C. Yan, J. Fraga-Dubreuil, E. Garcia-Verdugo, P. A. Hamley, M. Poliakoff, I.
Pearson, and A. S. Coote, Green Chem., 2008, 10, 98–103.
79. S. C. Stouten, Q. Wang, T. Noël, and V. Hessel, Tetrahedron Letters, 2013, 54,
2194–2198.
80. J. F. Ng, Y. Nie, G. K. Chuah, and S. Jaenicke, Journal of Catalysis, 2010, 269,
302–308.
81. K. Jähnisch, M. Baerns, V. Hessel, W. Ehrfeld, V. Haverkamp, H. Löwe, C. Wille,
and A. Guber, Journal of Fluorine Chemistry, 2000, 105, 117–128.
82. M. Shang, T. Noël, Q. Wang, and V. Hessel, Chemical Engineering & Technology,
2013, 36, 1001–1009.
83. S. J. Broadwater, S. L. Roth, K. E. Price, M. Kobaslija, and D. T. McQuade,
Organic & biomolecular chemistry, 2005, 3, 2899–2906.
84. A. M. Hyde and S. L. Buchwald, Angewandte Chemie - International Edition,
2008, 47, 177–180.
85. J. Andersen, S. Bolvig, and X. Liang, Synlett, 2005, 19, 2941–2947.
86. V. Hessel, I. Vural Gürsel, Q. Wang, T. Noël, and J. Lang, Chemical Engineering
& Technology, 2012, 35, 1184–1204.
87. V. Hessel, I. Vural Gürsel, Q. Wang, T. Noël, and J. Lang, Chemie Ingenieur
Technik, 2012, 84, 660–684.
88. M. S. Mettler, G. D. Stefanidis, and D. G. Vlachos, Industrial & Engineering
89. S. R. Deshmukh and D. G. Vlachos, Industrial & Engineering Chemistry Research,
2005, 44, 4982–4992.
90. K. Shah and R. Besser, Chemical Engineering Journal, 2008, 135, S46–S56.
91. C. Cremers, A. Pelz, U. Stimming, K. Haas-Santo, O. Görke, P. Pfeifer, and K.
Schubert, Fuel Cells, 2007, 7, 91–98.
92. K. L. Yeung, X. Zhang, W. N. Lau, and R. Martin-Aranda, Catalysis Today, 2005,
110, 26–37.
20
References
93. S. V. Karnik, M. K. Hatalis, and M. V. Kothare, Journal of

Microelectromechanical Systems, 2003, 12, 93–100.
94. T. Cui, J. Fang, A. Zheng, F. Jones, and A. Reppond, Sensors and Actuators B:
Chemical, 2000, 71, 228–231.
95. W. N. Lau, K. L. Yeung, and R. Martin-Aranda, Microporous and Mesoporous
Materials, 2008, 115, 156–163.
96. D. J. Quiram, K. F. Jensen, M. A. Schmidt, P. L. Mills, J. F. Ryley, M. D. Wetzel,
and D. J. Kraus, Industrial & Engineering Chemistry Research, 2007, 46, 8292–8305.
97. J. P. McMullen and K. F. Jensen, Annual review of analytical chemistry (Palo Alto,
Calif.), 2010, 3, 19–42.
98. I. Vural Gürsel, Q. Wang, T. Noël, and V. Hessel, Chemical Engineering
Transactions, 2012, 29, 565–570.
99. V. Hessel, C. Hofmann, H. Löwe, A. Meudt, S. Scherer, F. Schönfeld, and B.
Werner, Organic Process Research & Development, 2004, 8, 511–523.
100. E. Y. Kenig, Y. Su, A. Lautenschleger, P. Chasanis, and M. Grünewald, Separation
and Purification Technology, 2013, 120, 245–264.
101. A. Sundberg, P. Uusi-Kyyny, and V. Alopaeus, Chemical Engineering Research
and Design, 2009, 87, 705–710.
102. A. Ziogas, V. Cominos, G. Kolb, H.-J. Kost, B. Werner, and V. Hessel, Chemical
Engineering & Technology, 2012, 35, 58–71.
103. R. L. Hartman, H. R. Sahoo, B. C. Yen, and K. F. Jensen, Lab on a Chip, 2009, 9,
1843–1849.
104. K. F. Lam, E. Cao, E. Sorensen, and A. Gavriilidis, Lab on a chip, 2011, 11, 1311–
1317.
105. H. Niu, L. Pan, H. Su, and S. Wang, Industrial and Engineering Chemistry
Research, 2009, 48, 8629–8634.
106. C. Ye, G. Chen, and Q. Yuan, Chinese Journal of Chemical Engineering, 2012, 20,
111–119.
107. N. N. Gao, J. X. Wang, L. Shao, and J. F. Chen, Industrial and Engineering
108. N. T. Tran, I. Ayed, A. Pallandre, and M. Taverna, Electrophoresis, 2010, 31, 147–73.
109. C. T. Culbertson, S. C. Jacobson, and J. M. Ramsey, Analytical Chemistry, 2000,
72, 5814–5819.
110. A. B. Jemere, D. Martinez, M. Finot, and D. J. Harrison, Electrophoresis, 2009, 30,
4237–4244.
111. M. N. Kashid, Y. M. Harshe, and D. W. Agar, Industrial Engineering Chemistry
Research, 2007, 46, 8420–8430.
112. A. Holbach and N. Kockmann, Green Processing and Synthesis, 2013, 2, 157–167.
113. N. Assmann, A. Ładosz, and P. Rudolf von Rohr, Chemical Engineering &
Technology, 2013, 36, 921–936.
114. D. M. Fries, T. Voitl, and P. R. von Rohr, Chemical Engineering and Technology,
2008, 31, 1182–1187.
21
1. Introduction
115. Y. Okubo, T. Maki, N. Aoki, T. Hong Khoo, Y. Ohmukai, and K. Mae, Chemical
Engineering Science, 2008, 63, 4070–4077.
116. H. Pennemann, S. Hardt, V. Hessel, P. Löb, and F. Weise, Chemical Engineering &
Technology, 2005, 28, 501–508.
117. P. Löb, H. Pennemann, V. Hessel, and Y. Men, Chemical Engineering Science,
2006, 61, 2959–2967.
118. K. Benz, K. P. Jackel, K. J. Regenauer, J. Schiewe, K. Drese, W. Ehrfeld, V.
Hessel, and H. Lowe, Chemical Engineering & Technology, 2001, 24, 11–17.
119. K. Mae, T. Maki, I. Hasegawa, U. Eto, Y. Mizutani, and N. Honda, Chemical
Engineering Journal, 2004, 101, 31–38.
120. F. Schönfeld, V. Hessel, and C. Hofmann, Lab on a chip, 2004, 4, 65–69.
121. A. L. Dessimoz, L. Cavin, A. Renken, and L. Kiwi-Minsker, Chemical Engineering
Science, 2008, 63, 4035–4044.
122. J. R. Burns and C. Ramshaw, Lab on a chip, 2001, 1, 10–5.
123. G. Dummann, U. Quittmann, L. Gröschel, D. W. Agar, O. Wörz, and K.
Morgenschweis, Catalysis Today, 2003, 79-80, 433–439.
124. P. S. Sarkar, K. K. Singh, K. T. Shenoy, A. Sinha, H. Rao, and S. K. Ghosh,
Industrial and Engineering Chemistry Research, 2012, 51, 5056–5066.
125. A. Ghaini, M. N. Kashid, and D. W. Agar, Chemical Engineering and Processing:
126. M. N. Kashid, A. Renken, and L. Kiwi-Minsker, Industrial & Engineering
127. S. K. R. Cherlo, S. Kariveti, and S. Pushpavanam, Industrial & Engineering
128. M. N. Kashid and D. W. Agar, Chemical Engineering Journal, 2007, 131, 1–13.
129. J. G. Kralj, H. R. Sahoo, and K. F. Jensen, Lab on a chip, 2007, 7, 256–263.
130. J. Jovanović, E. V. Rebrov, T. A. Nijhuis, M. T. Kreutzer, V. Hessel, and J. C.
Schouten, Industrial and Engineering Chemistry Research, 2012, 51, 1015–1026.
131. M. N. Kashid, A. Gupta, A. Renken, and L. Kiwi-Minsker, Chemical Engineering
Journal, 2010, 158, 233–240.
132. O. K. Castell, C. J. Allender, and D. A. Barrow, Lab on a chip, 2009, 9, 388–96.
133. A. Adamo, P. L. Heider, N. Weeranoppanant, and K. F. Jensen, Industrial &
134. W. A. Gaakeer, M. H. J. M. de Croon, J. van der Schaaf, and J. C. Schouten,
Chemical Engineering Journal, 2012, 207, 440–444.
135. A. E. Cervera-Padrell, S. T. Morthensen, D. J. Lewandowski, T. Skovby, S. Kiil, and
K. V. Gernaey, Organic Process Research and Development, 2012, 16, 888–900.
136. F. Schönfeld and S. Hardt, AIChE Journal, 2004, 50, 771–778.
137. A. P. Sudarsan and V. M. Ugaz, Lab on a chip, 2006, 6, 74–82.
138. E. B. Nauman, Chemical Engineering Science, 1977, 32, 287–293.
139. A. K. Saxena and K. D. P. Nigam, AIChE Journal, 1984, 30, 363–368.
22
References
140. S. Klutz, S. K. Kurt, M. Lobedann, and N. Kockmann, Chemical Engineering

Research and Design, 2015, 95, 22–33.
141. S. K. Kurt, M. G. Gelhausen, and N. Kockmann, Chemical Engineering &
Technology, 2015, 38, 1122–1130.
142. J. Singh, N. Kockmann, and K. D. P. Nigam, Chemical Engineering and
Processing: Process Intensification, 2014, 86, 78–89.
143. M. M. Mandal, P. Aggarwal, and K. D. P. Nigam, Industrial & Engineering
144. D. Webb and T. F. Jamison, Chemical Science, 2010, 1, 675.
145. W. Shu, L. Pellegatti, M. A. Oberli, and S. L. Buchwald, Angewandte Chemie -
146. A. R. Bogdan, S. L. Poe, D. C. Kubis, S. J. Broadwater, and D. T. McQuade,
Angewandte Chemie - International Edition, 2009, 48, 8547–8550.
147. S. V. Ley, I. R. Baxendale, G. Brusotti, M. Caldarelli, A. Massi, and M. Nesi, Il
Farmaco, 2002, 57, 321–330.
148. R. I. Storer, T. Takemoto, P. S. Jackson, D. S. Brown, I. R. Baxendale, and S. V
Ley, Chemistry - A European Journal, 2004, 10, 2529–47.
149. I. R. Baxendale and S. V Ley, Ernst Schering Foundation symposium proceedings,
2006, 151–85.
150. I. R. Baxendale, S. C. Schou, J. Sedelmeier, and S. V Ley, Chemistry - A European
Journal, 2010, 16, 89–94.
151. I. R. Baxendale, S. V Ley, A. C. Mansfield, and C. D. Smith, Angewandte Chemie
152. L. J. Martin, A. L. Marzinzik, S. V Ley, and I. R. Baxendale, Organic letters, 2011,
13, 320–3.
153. I. R. Baxendale, J. Deeley, C. M. Griffiths-Jones, S. V Ley, S. Saaby, and G. K.
Tranmer, Chemical communications (Cambridge, England), 2006, 2566–8.
154. S. G. Newman and K. F. Jensen, Green Chemistry, 2013, 15, 1456–1472.
155. Syrris, http://syrris.com/flow-products/africa-modules/fllex-flow-liquid-liquid-
extrator.
156. Zaiput Flow Technologies, https://www.zaiput.com/liquid-liquid-separators.
157. H. R. Sahoo, J. G. Kralj, and K. F. Jensen, Angewandte Chemie (International ed.
in English), 2007, 46, 5704–5708.
158. R. L. Hartman, J. R. Naber, S. L. Buchwald, and K. F. Jensen, Angewandte Chemie
159. T. Noël, S. Kuhn, A. J. Musacchio, K. F. Jensen, and S. L. Buchwald, Angewandte
Chemie International Edition, 2011, 50, 5943–5946.
160. G. Schembecker and T. Bott, Chemie Ingenieur Technik, 2009, 81, 1094–1095.
161. J. Kussi and G. Schembecker, Chemie Ingenieur Technik, 2010, 82, 2031–2031.
162. J. Lang, F. Stenger, and R. Schütte, Chemie Ingenieur Technik, 2012, 84, 883–884.
163. J. Rottke, F. Grote, H. Fröhlich, D. Köster, and J. Strube, Chemie Ingenieur
Technik, 2012, 84, 885–891.
23
1. Introduction
164. T. Seifert, S. Sievers, C. Bramsiepe, and G. Schembecker, Chemical Engineering

and Processing: Process Intensification, 2012, 52, 140–150.
165. C. Bramsiepe, S. Sievers, T. Seifert, G. D. Stefanidis, D. G. Vlachos, H. Schnitzer,
B. Muster, C. Brunner, J. P. M. Sanders, M. E. Bruins, and G. Schembecker,
Chemical Engineering and Processing: Process Intensification, 2012, 51, 32–52.
166. A. Brodhagen, M. Grünewald, M. Kleiner, and S. Lier, Chemie Ingenieur Technik,
2012, 84, 624–632.
167. Ł. Hady and G. Wozny, Chemie Ingenieur Technik, 2012, 84, 597–614.
168. Chemie Ingenieur Technik, 2012, 84, 561–561.
169. T. Skjott-Larsen and P. B. Schary, Managing the Global Supply Chain, Vol. 3,
Copenhagen Business School Press, Copenhagen, 2007.
170. H.-P. Wiendahl, J. Reichardt, and P. Nyhuis, Handbuch Fabrikplanung, Carl
Hanser Verlag, Munchen, 2009.
171. S. Shah, Chemical Engineering Progress, 2007, 103, 36–41.
172. NMP Call, Nanosciences, Nanotechnologies, Materials and new Production
Technologies Calls: FP7-NMP-2009-LARGE-3, 2009.
173. European Commission, Research – Industrial Technologies, Factories of the future
(ww-ec.europa.eu/research/industrial_technologies/lists/factories-of-the-
future_en.html), 2010.
174. The F3 Factory Project, www.f3factory.com.
175. The CoPIRIDE Project, www.copiride.eu.
176. The POLYCAT Project, www.polycat-fp7.eu.
177. The PILLS Project, www.fp7pills.eu.
178. The SYNCLEAN Project, www.eranetbioenergy.net.
179. ACHEMA Daily, June 18, Breaking News, pg 8, 2012.
180. A. Geipel-Kern, Copiride Project: A Platform for Tomorrow’s Production
Solutions, http://www.process-
worldwide.com/engineering_construction/articles/305327/, Process, 03.02.2011.
181. J. Lang, F. Stenger, and H. Richert, Evonik Elements, 2011, 12–17.
182. V. Hessel, J. Lang, and D. Kralisch, CHEManager Europe, 2010, 5-6, 7.
183. R. B. Leng, M. V. M. Emonds, C. T. Hamilton, and J. W. Ringer, Organic Process
Research & Development, 2012, 16, 415–424.
184. ISO 14040, Envioronmental Management-life Cycle Assessment-Principles and
Framework, 2006.
185. U. Krtschil, V. Hessel, P. Löb, D. Reinhard, S. Hübschmann, and D. Kralisch,
186. M. Shang, T. Noël, Q. Wang, Y. Su, K. Miyabayashi, V. Hessel, and S. Hasebe,
187. The BIOGO Project, http://www.biogo.eu/.
188. The MAPSYN Project, http://www.mapsyn.eu/.
24
CHAPTER 2
Potential of Intensification through Flow
Processing and Use of Modular Plants to
Reduce Development Time
This chapter is based on :
Hessel, V., Vural Gursel, I., Wang, Qi, Noël, T. & Lang, J. (2012). Potential analysis of
smart flow processing and micro process technology for fastening process development -
Use of chemistry and process design as intensification fields. Chemical Engineering &
Technology, 35(7), 1184-1204.
Hessel, V., Vural Gursel, I., Wang, Q., Noël, T. & Lang, J. (2012). Potenzialanalyse von
Milli- und Mikroprozesstechniken für die Verkürzung von Prozessentwicklungszeiten :
Chemie und Prozessdesign als Intensivierungsfelder. Chemie Ingenieur Technik, 84(5), 660-
684.
Abstract:
Flow processes with microstructured reactors allow paradigm changes in process
development and thus can enable a faster development to the final production plant. So far,
this was mainly achieved via transport intensification – improved mixing and heat transfer
which increase productivity and possibly improve selectivity. A newer idea is chemical
intensification through deliberate use of harsh chemistries at unusual (high) pressure,
temperature, concentration, and reaction environment which again increases productivity. A
very new idea is the process design intensification – reaction-intensified flow processes
need less separation expenditure and the small unit size together with the high degree in
functionality gives large potential for system integration. The modular nature of the small
flow units allow an easy implementation to modern modular plant environments (Future
Factories) which enables to perform all the testing cycles (lab, pilot, production) in one
plant environment. The final result can be a reduced number of apparatus in the plant, a
reduced apparatus size, and a higher predictability in the scale-out of the apparatus. All
these have potential for (much) decreased overall development time.
25
2. Potential of Intensification through Flow Processing
2.1. Introduction
The pull from the market and the push from technology have initiated a race, in which the
lead time of bringing innovation into plant commissioning and final product has become a
crucial role. Therefore a dedicated team of professionals from university and industry
discussed at the 48th Tutzing-Symposium “Die 50%-Idee – Vom Produkt zur
Produktionsanlage in der halben Zeit” (translated: The 50% idea – from product to
production facility in half the time).1,2
Chemical process development is divided into different phases. It starts in the laboratory
exploring the chemical synthesis where a suited chemical protocol is established;
sometimes and especially in pharmacy using fast, automated screening methods. The
process concept is further developed and validated using modeling techniques and lab and
pilot plant operation.3,4 Process design is then elaborated through basic engineering and
detailed engineering studies when design data for each component is determined. Having
purchased the components, the construction and commissioning of the production plant is
initiated. Since in traditional engineering the phases need to be done in sequence, with each
phase needing the input from the former phase, the only option to reduce the time is to
reduce the time of one or more phases – here standardization through pre-manufactured
modules assembled into a modular and highly functional plant environment is a major
innovation approach.
Micro process technology has initiated a rethinking of processes from batch to flow and
even has entered the stage of commercial valorization, as e.g. demonstrated by giving
continuous processing Top-1 priority in the 10 Green Engineering prime measures stated by
the multi-industrial ACS GCI Pharmaceutical Roundtable.5 Still, microstructured reactors
are a new technology for chemical production. Experience is just based on some reports
about companies which approached or entered production with such flow reactors in the
last years; the total number of reports being estimated to be around 50 or so in the year
2011.6–10 Some deeper insight in cost details was mainly given in industrial presentations
and as well in a few papers both from industry and academia.11–14 Quite detailed
sustainability view has been given from the life-cycle assessment and cost analyses.15–18
However, the reduction of development time – ‘time-to-market’ – albeit also often
mentioned as motivation for flow process development, has so far seldom investigated in
depth or even quantified. This lack in knowledge is mainly due to the fact that development
time cannot be modeled. And also different from the other investigations, no experimental
data are available which can be used as input for further analysis. There is the need to
consider the full-chain process (with all or most components utilizing micro process
technology). This is at present quite complicated, since such kind of processes do simply
not exist. In almost all currently known cases only the reactor within the total process chain
(with all its reaction and separation units) was changed to microreactor. This is sufficient to
26
2.2. Effect of Intensification Fields and Modular Plants
judge on costs, sustainability, and safety; but the full potential in speeding up process
development can just vaguely be seen at this time.
Despite these deficiencies in data availability/modeling, we approached as first group the
eminent topic of speeding up process development in a qualitative theoretical impact
analysis as follows.
What is precisely clear at this point of time is which key steps are at hand to speed up the
reaction and finally process development, starting from intensification of the chemistry
itself up to using complete compact plant architectures.19–21 Thus, this chapter aims at
showing in a generic (‘impact-wise’) manner the respective chances of a flow process
chemistry development under the frame of a holistic plant development. The generic keys
considered follow the analysis of intensification given in chapter 1. The first is based on
utilizing - much more than in the past and now in deliberate manner - the so-called
chemical and process-design intensification fields in addition to the well-known transport
intensification field of microstructured reactors.21 The second tries to connect the flow
developments with the just recently undertaken developments on intensified and modular
plants, and here most notably those using container environments. Microreactors are not the
only modern devices to achieve process intensification however they are considered to be
the most potent one. This is why this reactor tool is considered in this study.
The effect of the intensification measures and use of modular plants will be given and,
wherever possible, will be quantified. This will give the industrial expert an idea on the
reduction in process development, although this cannot be straightforwardly quantified.
However, additional development tasks for flow processes such as the exploration of new
reaction pathways and of new, harsh process conditions may shoot down advantages gained
by faster development through these key technologies. Some of these tasks are related to
technology ripeness, and accordingly the technologies considered were classified as ready-
for-use and near-future-use. Increased process development time may also be given by the
less characterized performance of large-scale microstructured reactors and of
microseparators even at the laboratory stage. While it is thus pretty clear that the step-
change potential in process intensification can result in a significant reduction in
development time, this is so far alone a theoretical potential. The full release of this into a
practical potential suffers from the above mentioned unknowns; being pretty normal for a
young technology.

Following the arguments given in the introduction about that an overall judgment on the
speed-up of process development through micro process technology and flow chemistry is
quite difficult, it is here aimed to give answers at a level reduced in complexity, i.e. on the
27
apparatus level; assuming that such effects are at least beneficial, additive, and at best
synergetic on the higher process level. The distribution into individual impacts allows
further in the last section of this chapter to make a scenario-wise analysis which approaches
in intensification ensure already today a good practice in fastened process development,
which are better used in near future, and which do not save time in process development at
all (but might give superior costing and sustainability).
The impact of the three intensifications fields can be:
 Reduced size of apparatus

 Reduced number of apparatus
 Ease of scalability of the apparatus
All of these three measures certainly and inevitably should lead to a (much) decreased
overall development time.
2.2.1. Effect of the transport intensification (1st) field

Transport intensification can make possible reduction in reactor volume and number when
switching from batch to flow processing. This is due to improved mass and heat transfer
through the miniaturization of flow dimensions. Microreactors enable high surface to
volume ratios, short diffusion paths and well defined flow profile.22 Process parameters can
be controlled easily.19 Therefore, reaction rates are increased, residence times are decreased
and side reactions are suppressed.21 This results in increased yield and selectivity. This also
allows high space-time yields to be reached enabling compact apparatus.19 Reaction times
can be shortened from hours to seconds with the use of much smaller reactor volumes.23
Even for large-scale applications, the volumes of the reactors do not exceed a few liters,
whereas the industrial stirred vessels used are a factor of 1000 larger. Also, with the high
yield provided a single microreactor can replace series of batch reactors .
Transport intensification can also enable reduction in the number of apparatus to be used
and investigated in process development. This can be achieved through the elimination of
heating/cooling units through achieving room temperature processing at same productivity,
purification units due to a decrease in the levels of by-product to traces through increased
selectivity and recycle loops through higher reaction yields.19
Furthermore, the small level of holdup in microreactors reduces the damage potential of
toxic or explosive reagents. Also, fast and highly exothermic reactions can be handled
safely preventing hot spots.24 The improved safety in microreactors is essential for further
development when moving from laboratory to production scale of formerly prohibited
reactions in batch processing.
Less clear is the potential of a batch-to-flow exchange of more than one apparatus,
particularly concerning separation units, up to a process scheme composed of flow units
28
only. While after about 15 years development there is at hand a palette of process
components in microreaction technology (reactors, mixers and heat exchangers.), the
development of robust and preparative flow separation components is still largely missing.
Actually, this is not due to missing flow concepts and components for separation. A number
of microfluidic separation concepts were developed, but most often in the frame of lab-
chip’s analytical and biochemical developments. Since these devices were developed for a
completely different flow range than usual for chemical applications, it is rather difficult to
couple these to microreactors. From about 2011 onwards, there is a slowly, but steadily
increasing number of developed microseparation devices which are developed dedicatedly
for chemical purposes. Initially, these devices were mainly liquid-liquid extraction units.25
Recently, several micro distillation-/rectification, absorption and chromatography units
were also described.26 Still most of these devices have been characterized as stand-alone
tools, i.e. without connection to a flow reaction device. So far there are only few
demonstrations of coupling of microseparation units with microreactors.27–29
For the reasons given above, Lonza in 2008 preferred the step-wise integration of flow
reactors rather than of reactors to separators (see Figure 2.1).30 Yet, as mentioned this is
pharma’s preferred drop-in strategy and the mission of this thesis is to go beyond. Lonza
has now, also an adapted view to fully integrated flow processes with microreactors and
other units in flow.31
Figure 2.1. Lonza prediction of future development of micro process engineering for
commercial pharmaceutical production in 2008
29
2.2.2. Effect of the chemical intensification (2nd) field

Enhancing chemistry is named chemical intensification and this is one of the two
fundamental pillars of the new process chemistry approach termed Novel Process Windows
(NPW).15,19–22 Chemical intensification stands for the use of highly intensified, unusual, and
typically harsh process conditions to boost microprocess technology. Chemical
intensification follows the constitutional elements of every reaction protocol, that is, to
define the optimal reaction path/synthetic strategy; to determine the required reagents; and
to boost the reactivity by virtue of temperature, pressure, and concentration changes. It has
resulted for many reactions speed-up by a factor of 100 or even 1000. This necessarily will
facilitate for given productivity replacing several large volume batch reactors with a
microreactor enabling both reduction in reactor volume and number. It also suggests that
for the required higher productivity the need of numbering-up can be reduced significantly.
By using intensification approached it has been demonstrated by companies, most notably
Lonza, that one (or a few) microreactor is sufficient to achieve industrial fine chemical
production.32–35This significantly reduces the work required in scale-up.
High temperature and pressure flow processing is a widely applicable approach as initiated
by the group of Kappe.36–38 In laboratory equipment, for example, glass flask, autoclaves,
etc. the reaction temperature is mostly maintained by using, for example, a water bath, oil
bath, electrical heating, liquid nitrogen and dry ice mixtures in various alcohols.
Microstructured reactors now offer improved heat transfer capabilities and therewith better
temperature control during the chemical process. Also they allow fast heating-up and
cooling-down steps. Therefore microreactors are ideally fit to exploit high temperatures to
speed up the reaction rate. The reaction times are reduced from typically (many) hours to
minutes and sometimes even to seconds.
When reaction temperatures above the boiling point of a solvent used (superheated
processing) are targeted, it is necessary to increase the system pressure to avoid boiling and
to keep the reaction mixture in a liquid state. The mechanical stability of microstructured
devices allows one to run processes in a high pressure regime in safety. Usually
pressurization is assured by either HPLC or high pressure syringe pumps and a back
pressure regulator after the microreactor. Both laboratory and technical batch processing is
typically limited by this condition as autoclave operation is complex, requires larger
volumes, and needs longer reaction times when compared to microreactors. There are
numerous examples available demonstrating such transformation and the associated
massive speed up in reactivity under superheated conditions in microreactors.36–43
Apart from facilitating superheated conditions, high pressure affects the reaction through
the influence on the reaction equilibrium and rate constants and reaction medium via the
effect on its physical properties. In case where occupied volume by the product after the
reaction is lower than given for the two separate reactants, the overall volume change is
30
negative and application of high pressure will have an acceleration effect. This has been
demonstrated for cycloaddition and Claisen rearrangement reactions.37 Increased viscosity
of the liquid at high pressure was also found to accelerate chemical processes.44 Due to
increase in solubility of gases with the with the increase in pressure, gas-liquid reactions
can be performed with high concentrations of dissolved gases increasing reaction rates.45,46
The use of solvents in chemical synthesis is in most cases necessary to dissolve reactants
and to adjust the reaction volumes to the standard processing space of common glass flasks.
Solvents have to be removed later (stripping), which is energy intensive. Solvents cannot be
recycled in all cases so the generated waste may be enormous and deteriorates the eco-
efficiency of the process. Large volume of solvents also slow down reactions by creating
larger diffusion paths and reducing molecular collision.19 Depending on the reaction
mechanism, an increase in concentration (reduction of solvent) can speed up reactions. A
solvent-free operation might even result in a volume reduction which is so extreme that the
remaining reaction volume is too small to be produced efficiently with batch technology.
Smart-scaled flow apparatus are for this limitation an enabling solution.47–49
In addition, the tailoring of the solvent conditions by a much wider use of temperature and
pressure (affecting solubility, viscosity, dielectric constant, etc.) can enhance reactivity and
selectivity of such species. This can be even strengthened when using alternative solvents,
such as ionic liquids or supercritical fluids which can speed-up reactions be orders of
magnitude.36,50–52 Their low energy requirement, low toxicity and flammability and easy
recovery make them considered as green.53 Ability to have superheated processing with
microreactors enable switch from harmful to green solvents that were not possible to be
used at higher temperatures with batch technology.37
2.2.3. Effect of the process-design intensification (3rd) field

The second NPW pillar is termed process-design intensification and is based on process
simplification and integration. Transport and chemical intensification can have an impact
on the whole process (holistic). These new possibilities can lead to a radical rethinking of
the whole process design. The most evident effect on a full process level is that due to the
intensified operation the higher yield and selectivity achieved with lower solvent loads can
lead to reduction in separation requirements and equipment sizes. Simplification of the
process can also be achieved with the ability to turn former catalytic reaction into non-
catalytic. Due to intensified operation with microstructured reactors catalyst use could be
omitted simplifying process.54,55
In a more advanced, but not so distant step, due to the compactness of the basic elements
within the units and high speed of processing, integrated reactors will partly replace
consecutive arrangement of otherwise separated units. Already today heating and reaction
31
or exothermic and endothermic reactions are combined in microreactors.56–58 This is already

largely given for the field of fuel processing utilizing microreactors comprising energy and
mass flow integration.59 In fine chemistry, this is not so prominent so far. A logical step is
to combine separation and reaction, and actually that has been started with the use of
membranes.60–63 Another eminent integration issue is to incorporate functionalities such as
process control and analysis which can enable optimum operation of the system.64,65
Equipment number and size reduction has a direct impact on the plant footprint, this makes
scale-up much simplified reducing process development time. A comprehensive analysis of
current possibilities in flow process integration is given in both references.21,66
On top of that, a step change may be realized via exploring new reaction designs
(pathways). A one-step synthesis can largely simplify process via elimination of separation
equipment needed for otherwise consecutively done reactions. If not eliminated, there is a
chance for replacing separation units by simpler and less expensive units. This leads to
significant reduction in separation equipment and size. This can change the ‘plant footprint’
from being separation-dominated (as given by the numbers and sizes of respective
equipment) to reaction-dominated. Engineering, procurement, and construction phases can
be performed significantly faster with much less number of equipment required.
Consequently, it can have a major impact on process development time, however the time
spent on development of completely new chemistry needs to be considered as well.
The combination of process steps can have a time dimension, when fast, simultaneous
operation of several steps occurs within one process (such as multiple reactant dosing) or
when several reactions are combined in one process through the flow-chemistry adaptation
of a multistep one-pot synthesis.67 It may also have a spatial dimension through the
integration of function and system, i.e., combined reaction and separation or combined
reaction and heat exchange.
2.2.4. Effect of the compact, modular plants

The vision is to have transfer from batch to flow chemistry, favorably by use of micro-
process technology, intensify the operation with the use of NPW and to develop modular,
compact, fast-responsive and flexible production facilities which the intensified equipment
will be embedded.
There are two benefits of such modularization. By premanufacturing of standardized base
units and assembling them in a lego-type manner, fabrication/assembly time is shortened
and costs are saved. Thus, this reduces the development time to production. A second
benefit arises when in one tailored plant environment, all the development stages, lab, pilot
and production, can be conducted.68 This gives a compact joint concept for utility, process
control and safety. Several recent funding top themes approach these two issues this such as
the Factory of the Future which is the central objective within EU’s Flexible Future
32
2.3. Case Scenarios with Stepwise Intensification
Production Strategies.69,70 The full-chain ambition is the realization of ‘‘new, intensified

process and plant concepts for speeding up market penetration, for enhancing the product
life-cycle and improving sustainable production’’.70 Among other large-scale EU projects,
especially the projects CoPIRIDE71,72 and F3 Factory73–75 were intended to develop modular
production units for the chemical industry and to develop new intensified continuous
processes utilizing new/enhanced reactor technologies such as microreactors for integration
in these plants. Within F3 Factory project, several Process Equipment Containers were
successfully installed and integrated with the backbone infrastructure services in the
INVITE facility by the industrial project partners including Evonik, BASF and Bayer.76–78
Within CoPIRIDE project, modular container unit was demonstrated with the EcoTrainer
developed by Evonik Industries.68,72 EcoTrainer is designed as an all-purpose plant
platform. It includes in one unit, reactant, product, by-product supply and storage, reaction
and purification.79 The design of the container meets fire prevention standards. Emergency
exits are provided, and there is a containment basin in compliance with water management
regulations.79 This allows the realization of highly efficient solutions for warranty on-the-
job and environmental safety standards, for process control, cross compliance, and
maintenance. Due to the entire modular plant concept comprising all utilities – far beyond
the reaction & processor parts – it covers the whole development cycle.
The modular plants are seen as main enablers to cut in half the time it takes to get through
the development process from the product idea to the final production process (50%
idea).1,2 With its standardized infrastructure, utility and process control it supports also the
standardization of modules. It is efficient in terms of construction and mounting. Main
advantage is a short assembly time and drastic cost reduction in terms of planning and
construction. The continuous processing with smaller apparatus leads to a decrease in plant
footprint. It also allows fast plant start-up and shut-down. For many of the arguments given
above, the innovation cycle can be shortened. But besides these technical arguments,
modular plants provide also financial flexibility and in this way speed-up of management
procedures and decisions needed in the process development. The lower risk in
development allows to lower interest rates to finance plant installation and to intensify cash
flow at specific higher plant investment.68 This results in an earlier return-on-investment
and a larger flexibility to react to changes in volatile markets.

2.3.1. Fine chemistry
In the following, the main case scenarios relevant for flow process development are
considered which can be derived from the process intensification fields given above. Their
relative time-scales are judged by an impact analysis based on expert opinions taken from
33
the latter. This leads to a prognosis of shortenings, all done as a theoretical intrinsic
potential of the enabling technologies, ignoring specific effects of different reactions and
their bottlenecks. For separation, a near-future scenario was considered rather than the
present state of the art due to the analysis presented above regarding the development of
flow separators. This is seen to better reflect the intrinsic technology status.
Analysis - of such generic nature and being superficial to specific data of individual
processes - naturally cannot be taken at all to forecast real time-scale, i.e. how many years
such process development might take. What it can provide, however, is to give
recommendation for priority in using enabling technologies, as far as speed-up of process
development is considered; see Figure 2.2. It can give an idea where substantial and where
just mediocre savings in process time are expected. It can oppose benefits in costs and
sustainability to those of process development time.
2.3.1.1 Predictions based on theoretical impact analysis
Figure 2.2. Generic impact analysis of intensification fields on the process development
time in fine chemistry. The different cases and their total and partial time-scales are meant
to present trends, rather than a concrete time scale.
The current case to be considered for a flow process is with only reactor being changed
from batch to flow, while all other equipment remains as batch. This flow scheme is now
widely used in fine chemistry, has been taken over by industry and transferred even to
production.
34
Case (1) Batch processing needs a certain development time in laboratory and all other
steps (pilot, production) then consume a longer amount of time; here summarized as scale-
out. As typical time 6 years was assumed.
Case (2) Microreactors, being commercially available, allow in this way for many
applications “automatically” transport intensification (mixing, heating/cooling, safety) (1st
field). While certainly for other applications, such as multiphase systems, more elaborate
process development is needed. Transport intensification can save process development
time mainly for reasons of apparatus size decrease. As discussed above it can also enable
reduction in the number of apparatus and can make possible scale-up of formerly prohibited
reactions due to improved safety in microreactors. Higher predictability in the scale-out of
the apparatus is achieved with microreactors since the performance achieved at lab scale
can be maintained at higher scale resulting in decrease in the scale-out time.
Case (2) essentially has been studied in literature and there was no need for this thesis to
reinvestigate.
Case (3) Quite straightforward, although more elaborate is to gain chemical intensification
(2nd field). Therefore the development in lab can be longer but it can save even more
process development time in scale-out. As explained above chemical intensification can
speed-up reactions tremendously leading to significant reduction of apparatus size. Since
they suffice to handle higher productivity, some scale-out steps can be eliminated as well.
In this manner, it is not unrealistic to expect development times of 3 years under good
conditions.
Case (3) essentially has been studied and realized in the ERC Grant framing this thesis. See
works of Shang, Shahbazali, Borukhova, Stouten and and more.80–84
Case (4) The use of one plant environment during all development stages such as provided
by container plants will reduce the scale-out time. In conventional plant approach the
development starts in lab to test the feasibility and the reaction conditions. Large
production campaigns can only be performed in another production environment with
larger units in shift-work organization. With the modular container plants all development
stages can be completed in one environment resulting in simplification of scale-out.
Container plants can save time also by standardization of modules enabling saving in
planning and construction time and expenditure. This enables a faster time-to-market.
Case (4) essentially has been studied in this thesis; see chapter 5.
Case (5) In near future, flow separation units might sometimes be added leading to a full
(integrated) flow process. For some applications, a separator-specific intensification (1st
field) might be beneficial also for the overall process. This adds development time. This
effect will probably not be reduced much in the next decade or so. Thus, at least for the
35
time being, longer development times need to be considered. Still, such undertaking is
highly valuable, as it will lead to cost improvements and better sustainability.
Case (5) essentially has been studied in this thesis; see chapter 3.
When really integrating reaction and separation, the additional separation intensification
automatically sets in place and the additional efforts are a must, leading to the same
conclusions as given above of longer development time. The development of new reaction
designs leading to new process designs (3rd field) has large potential, but also needs much
additional chemistry and catalyst development. Thus, although this takes large promises in
the direction of saving costs and providing a more sustainable chemistry; it is not so suited
for time reduction in process development for fine-chemical production.
2.3.1.2 Companies’ strategies for fine-chemical process development in flow
Lonza, Visp/Switzerland has given most insight in a chemical company’s approach and
success in process development for continuous flow production. Lonza has designed and
tested a series of microstructured devices in continuous-flow plants, and performed lab
studies of pharmaceutical reactions with successful transfer to commercial production.
Summing up all reports from Lonza, at least 10 reactions were transferred to production
based on year 2011 analysis, but probably now in 2015 there are more.32–35,85,86
Lonza’s microstructrured reactors are designed for scale-out and, due to smart increase in
dimensions, the favorable single channel approach might be used throughout the whole
process development avoiding parallelization from lab development to pilot-scale
production. Four microreactor classes were described of increasing channel sizes, with
glass as material for laboratory and steel for piloting and production.32
DSM in Austria and NicOx, based in France, developed a pilot-scale process based on
numbering-up multiple parallel microreactors to make an arthritis drug. The drug,
naproxcinod, is the nitroxybutyl-substituted form of naproxen, the well-known nonsteroidal
anti-inflammatory drug. This compound containing a nitro group is difficult to make, since
nitration reactions are strongly exothermic. Together with Corning, Inc. a microreactor
system was designed that combines three process steps - the nitration reaction,
neutralization, and workup. Dr. Hartmann from DSM quoted the achieved decrease in
process development time: “Along with the safety and processing advantages, NicOx
appreciated how quickly large amounts of material could be made available … the
development efforts and investment are much less."7 The pilot-scale process of DSM with a
throughput of hundreds of kilograms was achieved after about six months work. The whole
time from feasibility studies to large-scale production took only about 18 months.7
36
2.3.2. Bulk chemistry

2.3.2.1 Predictions based on theoretical impact analysis
For bulk-chemical processes, the impact and time demands are different from the
conclusions given for fine chemistry; see Figure 2.3. Process design with all its chances for
mass- and energy integration is here the key. Thus, the idea for developing a flow process
in bulk chemical industry must be to forecast process simplification based on the
opportunities provided by microreaction technology. So, the first step is to chemically
establish a new reaction route which allows to simplify process design. Subsequently, the
next step is to intensify this new (direct) process utilizing microreactors with the 2nd
chemical intensification field. Finally, a full-flow process needs incorporation of
microseparators and there might be additional momentum through the 1st transport
intensification field. The key to process time reduction here is to achieve process
simplification in the first step which reduces the number of equipment that needs to be
developed (finally at large scale) in flow. Similar to Figure 2.2, this can be made in
stepwise fashion as follows.
Figure 2.3. Generic impact analysis of intensification fields on the process development
time in bulk chemistry. The different cases and their total and partial time-scales are meant
to present trends, rather than a concrete time scale.
37
Case (1) Development times in bulk chemistry, based on batch and other conventional
(including classical continuous) processing, are much longer than for fine chemistry. As
typical time 10-12 years was assumed. This is not so much due to development time in
laboratory, but caused by the more complex scale-out (pilot, production) requiring long
times for planning, engineering and construction.
Case (2) Microreactors for bulk chemistry are commercially not available, yet even their
prototyping has hardly been demonstrated, although some challenging developments were
undergone in this direction. Thus, different from fine chemistry, transport intensification
(mixing, heating/cooling, safety, 1st field) cannot “automatically” be utilized now. It rather
requires further development, especially for heat integration and catalyst immobilization.
Rather than using “pure” microtechnology, the solution probably is given by mesoscaled or
flat-plate reactors with catalysts as foams or fixed-beds, providing continuous processing.
This has compromises to transport intensification to allow productivity high enough for the
large scales. It is still expected that with such solutions at hand the scale-out can be reduced
due to the replication or representativeness of the process units from lab to production.
Especially with the implementation of chemical intensification (2nd field) further benefits in
ease of scalability can be achieved with significant reduction in apparatus size.
Case (2) has not been studied and realized in this thesis. Yet, we tackle with predicting
operation and costs of such unknown large-scale microstructured reactors; see chapter 7.
Case (3) Since separation units comprise the majority of the process units in bulk
chemistry, the addition of flow separation units with their changes for replication and ease
of scalability should have an additional, actually stronger impact on decreasing scale-out
time as compared for flow transfer of only the reaction unit. Therefore, although
development of flow separators constitutes an additional development time, in overall the
development time can be reduced.
Case (3) has not been studied and realized in this thesis. Yet, our separation concept as
described in chapter 4 is valid for fine-chemical capacities.
Case (4) In near future, the development of new reaction routes might allow to propose
simplification of the process-design which is the process-design intensification field (3rd
field). This adds a development time for the new reaction design. However due to apparatus
size as well as number decrease plant footprint is much reduced. Therefore, savings are
expected as scale-out is much simpler and in sum this may lead to a shorter overall
development time.
Case (4) has been studied in this thesis; for the first time in literature. See chapter 7.
Case (5) In future, process integration of flow reaction and separation can be achieved
under the process-design intensification field (3rd field). This adds a development time for
the new integrated reaction-separation design. The scale-out of such solutions, however,
38
2.4. Conclusion
may not be possible just by strict replication and this will add development times so that the
overall benefit in time savings is considered to be minor.
Case (5) has not been studied and realized in this thesis. It is a very new concept and
recently being investigated in literature with membranes. 60–63
As a net result, the longer development times in bulk chemistry offer larger possibilities for
savings than given in fine chemistry; on the other side, the state of the art in development of
respective (micro-/milli-/meso-) structured flow reactors is much lower as for fine-chemical
flow reactors and this adds additional development challenges and times. Thus,
opportunities and challenges largely come to balance and savings in development time are
predicted to be similar as for fine chemistry. However, cases (4) and (5) make also evident
that when releasing the true flow potential for bulk chemistry – which is process
simplification and integration – the main argument is probably not saving in development
time, but is savings in costs and gains in sustainability (which is not shown in this article).
In the examples that use microstructured equipment for applications at a scale of the upper
limit of fine chemistry and bridging to bulk chemistry, they were inserted into the existing
large scale plant (retrofit). They were motivated by safety, heat-transfer limitations and
requirement of throughput increase.87
2.4. Conclusion
After having seen almost two decades of flow-related developments of micromixers,
microreactors, micro heat exchangers, and recently microseparators, the time is ripe for the
next round of the implementation of microreactor and flow techniques. This is devoted
towards bringing all of the above-mentioned devices to industrial productivity and
embedding them in a modern Future Factory type plant environment with an overarching
process and plant approach.
Intensified flow processing and micro process technology can make process development
somewhat simpler and save development time. One main reason is the higher predictability
in the scale-out of the apparatus achieved with microreactors since the performance
achieved at lab scale can be maintained at higher scale resulting in decrease in the scale-out
time. This allows also to develop modular reactor and separator concepts which fit well to
modular plant concepts (such as containers), developed also currently. With the modular
container plants all development stages can be completed in one environment resulting in
simplification of scale-out. Container plants can save time also by standardization of
modules enabling saving in planning and construction time and expenditure.
Another main reason is that microstructured reactors and similar flow and process
intensified equipment (milli/meso) enable transport and chemical intensification, e.g. open
39
the door to harsher process chemistries. This allows to safely process high throughput in
smaller equipment. A final main reason, but much less exploited so far as compared to the
other two mentioned, is the chance for radical changes in process design when using a flow
intensified plant called process-design intensification. The focus is here much on the
reaction part and this can largely reduce the number and size of separation equipment. Such
process simplification is at best not only achieved by the higher selectivity of flow
reactions, but by an additional radical change in reaction design, e.g. by newly developed
direct syntheses or multi-step cascaded reactions in one flow. Such far-fetching
development adds development time, with chance for pay back due to reduced plant
footprint. Process integration, reaction with separation, may reduce somewhat the impact
on time savings in process development, but undoubtedly offers large chances for cost
saving and sustainability.
Finally, with the present microreactor and flow technology, it seems that time savings in
process development are likely for fine-chemical applications up to the border of bulk-
chemical processes; but that for true bulk-chemical applications, the cost and sustainability
arguments are the main motivation for present investigations here.
References
3. G. H. Vogel, Process Development: From the Initial Idea to the Chemical
Production Plant, Wiley-VCH, Weinheim, Germany, 2006.
4. J. Harmsen and J. B. Powell, Sustainable Development in the Process Industries:
Cases and Impact, John Wiley & Sons, Hoboken, NJ, 2010.
6. P. L. Short, Chemical & Engineering News, 2006, 84, 38.
7. A. M. Thayer, Chemical and Engineering News, 2009, 87, 17–19.
8. M. McCoy, Chemical and Engineering News, 2010, 88, 10.
9. P. L. Short, Chemical and Engineering News, 2008, 86, 37–38.
11. D. Schmalz, M. Häberl, N. Oldenburg, M. Grund, H. Muntermann, and U. Kunz,
Chemie Ingenieur Technik, 2005, 77, 859–866.
12. D. M. Roberge, L. Ducry, N. Bieler, P. Cretton, and B. Zimmermann, Chemical
Engineering and Technology, 2005, 28, 318–323.
13. U. Krtschil, V. Hessel, D. Kralisch, G. Kreisel, M. Küpper, and R. Schenk,
CHIMIA International Journal for Chemistry, 2006, 60, 611–617.
40
References
14. E. Klemm, H. Döring, A. Geisselmann, and S. Schirrmeister, Chemical

467–478.
16. S. Huebschmann, D. Kralisch, V. Hessel, U. Krtschil, and C. Kompter, Chemical
17. D. Kralisch, I. Streckmann, D. Ott, U. Krtschil, E. Santacesaria, M. Di Serio, V.
Russo, L. De Carlo, W. Linhart, E. Christian, B. Cortese, M. H. J. M. de Croon, and
V. Hessel, ChemSusChem, 2012, 5, 300–11.
18. D. Ott, D. Kralisch, I. Denčić, V. Hessel, Y. Laribi, P. D. Perrichon, C. Berguerand,
L. Kiwi-Minsker, and P. Loeb, ChemSusChem, 2014, 7, 3521–33.
2011, 66, 1426–1448.
2013, 6, 746–89.
22. T. Illg, P. Löb, and V. Hessel, Bioorganic and Medicinal Chemistry, 2010, 18,
3707–3719.
& Technology, 2012, 35, 1184–1204.
24. T. Illg, V. Hessel, P. Löb, and J. C. Schouten, Chemical Engineering Journal,
2011, 167, 504–509.
Technology, 2013, 36, 921–936.
in English), 2007, 46, 5704–5708.
30. D. M. Roberge, B. Zimmermann, F. Rainone, M. Gottsponer, M. Eyholzer, and N.
Kockmann, Organic Process Research & Development, 2008, 12, 905–910.
31. D. M. Roberge, Lonza “Factory of Tomorrow” with Flow Processes &
MicroReactors, Presented at MicroNano Conference ’13, 11-12 December 2013,
Ede, The Netherlands, 2013.
2011, 167, 718–726.
33. N. Kockmann and D. M. Roberge, Chemical Engineering and Processing: Process
41
34. N. Kockmann, M. Gottsponer, B. Zimmermann, and D. M. Roberge, Chemistry - A

European Journal, 2008, 14, 7470–7.
35. N. Kockmann and D. M. Roberge, Chemical Engineering & Technology, 2009, 32,
1682–1694.
Chemistry, 2009, 2009, 1321–1325.
38. T. Razzaq, T. N. Glasnov, and C. O. Kappe, Chemical Engineering & Technology,
2009, 32, 1702–1716.
Development, 2010, 14, 215–224.
40. C. O. Kappe and E. Van der Eycken, Chemical Society reviews, 2010, 39, 1280–
1290.
42. E. R. Murphy, J. R. Martinelli, N. Zaborenko, S. L. Buchwald, and K. F. Jensen,
Angewandte Chemie (International ed. in English), 2007, 46, 1734–7.
43. R. E. Martin, F. Morawitz, C. Kuratli, A. M. Alker, and A. I. Alanine, European
Journal of Organic Chemistry, 2012, 2012, 47–52.
44. K. A. Swiss and R. A. Firestone, J. Phys. Chem. A, 2000, 104, 3057–3063.
45. R. M. Tiggelaar, F. Benito-Lopez, D. C. Hermes, H. Rathgen, R. J. M. Egberink, F.
G. Mugele, D. N. Reinhoudt, van den A. Berg, W. Verboom, and H. J. G. E.
Gardeniers, Chemical Engineering Journal, 2007, 131, 163–170.
46. F. Trachsel, C. Hutter, and P. von Rohr, Chemical Engineering Journal, 2008, 135,
S309–S316.
47. P. Loeb, V. Hessel, H. Klefenz, H. Lowe, and K. Mazanek, Lett. Org. Chem., 2005,
2, 767–779.
48. S. Taghavi-Moghadam, A. Kleemann, and G. Golbig, Organic Process Research &
Development, 2001, 5, 652–658.
49. H. Löwe, V. Hessel, P. Löb, and S. Hubbard, Organic Process Research &
Development, 2006, 10, 1144–1152.
50. F. Trachsel, B. Tidona, S. Desportes, and P. Rudolf von Rohr, The Journal of
Supercritical Fluids, 2009, 48, 146–153.
51. X. Han and M. Poliakoff, Chemical Society reviews, 2012, 41, 1428–36.
52. C. Yan, J. Fraga-Dubreuil, E. Garcia-Verdugo, P. A. Hamley, M. Poliakoff, I.
Pearson, and A. S. Coote, Green Chem., 2008, 10, 98–103.
53. L. Á. de Cienfuegos, R. Robles, D. Miguel, J. Justicia, and J. M. Cuerva,
ChemSusChem, 2011, 4, 1035–48.
54. P. Löb, H. Löwe, and V. Hessel, Journal of Fluorine Chemistry, 2004, 125, 1677–
1694.
55. P. B. Palde and T. F. Jamison, Angewandte Chemie (International ed. in English),
2011, 50, 3525–8.
42
References
56. M. S. Mettler, G. D. Stefanidis, and D. G. Vlachos, Industrial & Engineering

57. S. R. Deshmukh and D. G. Vlachos, Industrial & Engineering Chemistry Research,
2005, 44, 4982–4992.
58. K. Shah and R. Besser, Chemical Engineering Journal, 2008, 135, S46–S56.
59. C. Cremers, A. Pelz, U. Stimming, K. Haas-Santo, O. Görke, P. Pfeifer, and K.
Schubert, Fuel Cells, 2007, 7, 91–98.
60. K. L. Yeung, X. Zhang, W. N. Lau, and R. Martin-Aranda, Catalysis Today, 2005,
110, 26–37.
61. S. V. Karnik, M. K. Hatalis, and M. V. Kothare, Journal of
Microelectromechanical Systems, 2003, 12, 93–100.
62. T. Cui, J. Fang, A. Zheng, F. Jones, and A. Reppond, Sensors and Actuators B:
Chemical, 2000, 71, 228–231.
63. W. N. Lau, K. L. Yeung, and R. Martin-Aranda, Microporous and Mesoporous
Materials, 2008, 115, 156–163.
64. D. J. Quiram, K. F. Jensen, M. A. Schmidt, P. L. Mills, J. F. Ryley, M. D. Wetzel,
and D. J. Kraus, Industrial & Engineering Chemistry Research, 2007, 46, 8292–
8305.
65. J. P. McMullen and K. F. Jensen, Annual review of analytical chemistry (Palo Alto,
Calif.), 2010, 3, 19–42.
66. S. Borukhova and V. Hessel, in Process Intensification for Green Chemistry, eds.
K. Boodhoo and A. Harvey, John Wiley & Sons, Ltd, Chichester, UK, 2013.
67. H. Pennemann, S. Hardt, V. Hessel, P. Löb, and F. Weise, Chemical Engineering &
Technology, 2005, 28, 501–508.
69. NMP Call, Nanosciences, Nanotechnologies, Materials and new Production
Technologies Calls: FP7-NMP-2009-LARGE-3, 2009.
70. European Commission, Research – Industrial Technologies, Factories of the future
(ww-ec.europa.eu/research/industrial_technologies/lists/factories-of-the-
future_en.html), 2010.
71. The CoPIRIDE Project, www.copiride.eu.
73. The F3 Factory Project, www.f3factory.com.
74. M. Henig, Invite and F3 Factory, The European Chemical Factory of the Future Is
Taking Shape at the Invite Research Center in Germany, Process, 27.09.2011,
(http://www.process-worldwide.com/engineering_construction/articles/332787/).
75. A. Geipel-Kern, F3 Project, What Will Tomorrow’s Chemical Plant Look Like?,
Process, 10.07.2010, (http://www.process-
worldwide.com/engineering_construction/articles/300750/).
43
76. F3 Factory Newsletter, Issue 1, August 2010,

(http://f3factory.com/scripts/pages/en/newsevents/F3_Factory_Newsletter_1.pdf).
77. F3 Factory Newsletter, Issue 2, September 2011,
78. F3 Factory Newsletter, Issue 3, August 2012,
2013, 36, 1001–1009.
81. M. Shang, T. Noël, Q. Wang, Y. Su, K. Miyabayashi, V. Hessel, and S. Hasebe,
6, 2220–2225.
ChemSusChem, 2015, 8, 504–12.
84. S. C. Stouten, Q. Wang, T. Noël, and V. Hessel, Tetrahedron Letters, 2013, 54,
2194–2198.
85. D. M. Roberge, N. Bieler, M. Mathier, M. Eyholzer, B. Zimmermann, P. Barthe, C.
Guermeur, O. Lobet, M. Moreno, and P. Woehl, Chemical Engineering and
Technology, 2008, 31, 1155–1161.
86. P. Barthe, C. Guermeur, O. Lobet, M. Moreno, P. Woehl, D. M. Roberge, N.
Bieler, and B. Zimmermann, Chemical Engineering & Technology, 2008, 31,
1146–1154.
44
CHAPTER 3
Connected Reaction and Metal
Separation in Flow – Drop-In Process
Integration at Lab Scale
Vural Gursel, I., Noël, T., Wang, Q. & Hessel, V. (2015). Separation/recycling methods of
homogeneous transition metal catalysts in continuous flow. Green Chemistry, 17, 2012-
2026.
Vural Gursel, I., Aldiansyah, F., Wang, Q., Noël, T. & Hessel, V. (2015). Continuous metal
scavenging and coupling to one-pot copper-catalyzed azide-alkyne cycloaddition click
reaction in flow. Chemical Engineering Journal, 270, 468-475.
Abstract:
Increasing usage of catalytic chemistry calls for efficient removal of metal traces. This
chapter describes the development and optimization of a scavenger-based extraction in flow
to remove metal catalysts. It enables liquid–liquid extraction with slug flow and phase
separation with a porous fluoropolymer membrane. The use of this unit for copper
scavenging of various copper sources from organic solvent was studied. The effects of
scavenger type (EDTA, DTPA, EDDS), concentration and pH were also investigated. Such
analysis allowed to achieve extraction performance as high as 99% at pH of the scavenger
solution adjusted to 9.4 and molar ratio of scavenger to copper of 10. Process integration is
achieved by coupling this unit downstream to a flow reaction using homogeneous metal-
based catalysis, i.e. presenting a continuous uninterrupted metal scavenging unit. The
copper-catalyzed azide-alkyne cycloaddition click reaction is studied in one-pot eliminating
the need to isolate and handle potentially explosive azide. The triazole product is attained in
flow with high yield of up to 92% with 30 min residence time. The level of purity
requirement for pharmaceuticals is met in one stage of extraction.
45
3. Connected Reaction and Metal Separation in Flow
3.1. Introduction
Process intensification through micro-process technology allows to reach processing far
from state of the art through entirely new and innovative equipment and means of
processing in continuous flow mode 1–3. In the past, it was based mainly on transport
intensification that improves mixing and heat exchange 4,5. More recently, chemical
intensification is increasingly explored which uses highly intensified, unusual and typically
harsh process conditions to boost micro-processing (high temperature, high pressure, high
concentration, safety) 6–11. The last two mentioned means are an essential part of Green
Chemistry. Beyond, as third momentum, process-design intensification heads for simplified
and integrated flow process design (Green Engineering) 12. The latter two intensification
fields constitute ‘Novel Process Windows’, a term coined by Hessel in the past and further
developed by our group 1,13–15.
In the microreactor literature and community it is common to focus largely to one
equipment (designed and tested) and its processing. However, with process-design
intensification we focus our attention toward the chemical process as a whole, thereby
rendering a true holistic picture 16,17. The majority of microchemical synthesis applications
were so far restricted to one or few reaction steps with the separation done offline 18–20.
Microreactors are highly developed at scales up to industrial production 21,22. However,
further development of microseparation units is needed. From about 2011 onwards, there is
an increasing number of developed microseparation devices 23. Still most of these devices
have been characterized as stand-alone tools, i.e. without connection to a flow reaction
device. The coupling of microseparation units with microreactors is not rare anymore, but
still far from being commonly applied. So far, there are only few demonstrations 24–27. To
be able to achieve multi-step synthesis in flow and make the flow process chain complete,
further development of microseparators is required.
Since the early investigations in micro process technology, microextraction has received a
considerable amount of attention. Microextraction involves mixing and contacting of the
organic and aqueous phases and consequently achieving the phase separation. Extraction
based on liquid-liquid slug flow in microchannels facilitates convective mass-transfer
between the phases 28. Due to internal circulation flow, high mixing performance is
achieved 4. After extraction the two phases can be easily separated. Effective liquid-liquid
separation can be achieved using membranes 29,30. Separation of the two phases can also be
performed with branched hydrophobic and hydrophilic outlets 31,32.
For extraction of copper and a wide range of metals, chelating agent solutions are effective
33
. These scavenging agents extract metal by forming a metal-chelate complex which can
further be removed from the solution 34. It enables high extraction efficiency of the metal
and is fast, compatible with many solvents and causes no loss of product. It was used
effectively for extraction from spent catalyst 34–36 and also waste waters 37,38. Most used
46
3.1. Introduction
scavenging agents include ethylene diamine tetraacetic acid, EDTA and diethylene triamine
pentaacetic acid, DTPA 39. Due to their high complexation strength they are effective
scavenging agents but are resistant to degradation 36. Biodegradable [S,S]-ethylene diamine
disuccinic acid ([S,S]-EDDS) has been proposed as an alternative scavenging agent and it
has also been found to have high extraction efficiency 35,36,40.
The copper-catalyzed azide-alkyne cycloadditon (CuAAC) was taken as the reaction of
industrial relevance. It is established as one of the most important examples of click
chemistry 41–43. The resulting 1,2,3 triazoles find use in variety of applications from drug
discovery and development to material science and chemical biology 41,42,44–47. Several
approaches to perform CuAAC in continuous flow processing have been reported 9,48–50.
This is motivated by improved mass and heat transfer as well as the safety aspects, azide
being unstable, and scalability of flow processing opposed to conventional batch
procedures. Microwave irradiation was also tested and major reduction in reaction time was
observed 48,51,52. Another advance in this field is generation of the required azides in situ 50–
53
. This one-pot methodology eliminates the need to synthesize azide separately in
continuous flow 54. All these examples used heterogeneous catalysis. Although
heterogeneous catalysis has advantage of being recyclable, the copper species leach into the
solution 48,49. Scavenger resins in cartridges has been used to remove these copper species
from the solution. One example is QuadrapureTM TU that has been successfully used for
CuAAC reactions 48,49. However, the need to periodically replace these cartridges interrupts
the continuous process.
A microseparation unit has been developed to achieve continuous uninterrupted extraction
of metal catalyst in flow. It involves liquid-liquid extraction based on slug flow with the use
of chelating agent and phase separation using porous fluoropolymer membrane. The
coupling of CuAAC reaction with continuous copper extraction was achieved in flow in
previous studies 9. However, three stages were required to reduce the copper content of
triazole product below the limit for APIs of 15 ppm 55. Since this is not optimal, more
extensive analysis of microseparation unit was needed. This is carried out and presented in
this chapter. In this study, first the performance of the microseparation unit is benchmarked
using EFCE-recommended extraction systems. Also its performance is compared with a
similar flow extraction unit described in literature. Then the copper extraction experiments
are performed that enable continuous uninterrupted separation of copper species in solution.
To achieve the optimum extraction performance, the use of different types of copper
catalysts and metal scavengers (EDTA, DTPA and EDDS) are investigated together with
variation of metal scavenger concentration and pH. Based on this analysis and optimization
of parameters, the copper scavenging unit is coupled to the CuAAC click reaction using
homogeneous copper catalyst with the aim to get the triazole product at high yield and
purity in one stage of separation.
47
3.2. Homogeneous Metal Catalyst Separation/Recycle in Flow

Homogeneous catalysis is of great interest for its high activity and selectivity. However,
their separation from the product remains a challenge. Transition metals and especially
platinum group metals are used extensively. They are expensive and there are strict
regulations on the permitted levels of these metals in pharmaceutical products.55 Moreover,
we are running short of these metals. Therefore, their effective separation and recycle is of
great importance. In industry, classical methods of homogeneous catalyst separation
involve thermal or other costly processes. The most common method of distillation requires
high energy and also can result in irreversible deactivation of the catalyst system.56 This has
stimulated researchers to look for entirely new processes as more gentle, low energy
alternatives to separate /recycle homogeneous catalysts in continuous flow. These include
heterogenization, scavenging, use of biphasic conditions and organic solvent nanofiltration.
There are numerous successful demonstrations on laboratory scale and recently also on an
industrial scale. The advantages and disadvantages of them are summarized in Table 3.1.
Table 3.1. Summary of the advantages and disadvantages of the separation/recycling
methods of homogeneous metal catalysts in continuous flow
Separation methods Advantages Disadvantages

Heterogenization ○ Catalyst is immobilized on ○ Lower catalytic activity
supports enabling easy reuse ○ Significant problem of leaching
in the case of the homogeneous
mechanism
○ Catalyst deactivation due to
deposition of impurities
Scavenging ○ High efficiency ○ Need for replacement of
scavenging columns after saturation
○ No loss of product ○ Use of scavenging columns not
○ Especially advantageous for suitable at larger scales
trace amounts of metals
Use of biphasic conditions ○ The phase containing the ○ The chemistry needs to be
catalyst can be separated and compatible with the biphasic
recycled easily reaction conditions
○ Applicability at the industrial ○ Systems with fluorous solvents,
scale ionic liquids and supercritical fluids
have high initial cost
Organic solvent ○ Low energy separation ○ Catalysts need to be molecular
nanofiltration ○ Catalyst retained enabling weight enlarged to achieve high
easy reuse retention performance
○ Applicability at the industrial ○ Concentration polarization can
scale occur for dead-end filtration
48
3.2. Homogeneous Metal Catalyst Separation/Recycle in Flow
A schematic representation of these methods is given in Figure 3.1. There are two ways of
heterogenization of homogeneous catalyst. First is by immobilizing homogeneous catalysts
on solid support. The drawback is the significant leaching and degradation of the catalyst
during its use and recycle leading to loss of metal.57 Because of the high catalyst cost this
has an economic impact and this is the reason this method has no application at a large
scale. However, there are successful examples, and commercialization in future can be
possible.58–61 Second is recently studied immobilization of homogeneous catalyst in a
stationary liquid phase. In this system, the catalyst is dissolved in a thin film of liquid
which is supported within the pores of silica. Ionic liquids have been used for this
purpose.62–64 With the successful examples of the supported ionic liquid systems, there is
potential for their further application, especially since only a small amount of ionic liquid is
required.
Figure 3.1. Overview of separation/recycling methods of homogeneous metal catalysts in

continuous flow
Homogeneous catalyst separation can also be achieved by scavenging which is typically
done utilizing solid-supported scavenging resins packed into columns 65,66. The scavenging
agents extract metal by forming a metal-chelate complex which can further be removed
from the solution 34. They are very efficient in capturing very low amounts of metals in
metals which is attractive for pharmaceutical production on a small scale 67,68. However,
they need to be replaced when saturated interrupting the continuous process. As a kind of
49
extension and as a solution to this problem, scavenging agents in solution has been
investigated in this chapter which holds promise to be applied at a large scale.
Another method is the use of liquid–liquid biphasic conditions. This is achieved with one
phase selectively dissolving the catalyst while the other dissolving the product. After
separation of the phases, the phase containing the catalyst can be easily recycled to the
reactor. There are several commercial applications of aqueous-organic and organic-organic
biphasic conditions for large scale production showing economic and environmental
benefits.69–71 Biphasic systems can also be created with fluorous solvents, ionic liquids and
supercritical carbon dioxide which have a number of successful applications on a lab-scale
due to their unique properties72,73 being utilized for facile catalyst separation.74–79 They have
not found commercial application so far, most likely due to the high cost of the chemicals
or the equipment required. But with efficient processes, they can be recycled almost totally
with a long lifetime.69 Then their initial selling price becomes less significant. With the
number of successful examples of their application increasing, it can be expected that their
advantages will be exploited in industry especially in fine chemical production.
Finally, there is a relatively new method of organic solvent nanofiltration (OSN). It has
been studied in flow for a large range of reactions to retain the homogeneous catalyst in the
reactor.80–82 Also catalysts have been molecular weight enlarged with dendrimers,83,84
polymers85 and polyhedral oligomeric silsesquioxanes86,87 to achieve better retention with
OSN. There are numerous examples of their application in lab. Currently there are also few
examples of its application by industry.56,88–91 The economic potential is seen to be very
promising.92 With the ongoing development, more and better performing commercial
membranes are becoming available. Their industrial applications are thus expected to
increase in the near future.
3.3. Experimental
3.3.1. Experimental setup of microseparation unit
For testing the performance of the microseparation unit for liquid-liquid extraction
experiments are performed with one 5 mL syringe loaded with feed containing solute and
another 5 mL syringe loaded with extraction solvent. For copper scavenging experiments a
5 mL syringe is loaded with mixture of copper catalyst and organic solvent, another 5 mL
syringe is loaded with aqueous EDTA solution. Equipment configuration is shown in
Figure 3.2. Plastic syringes from DB are used. The two syringes are mounted to Chemyx
Fusion 200 Touch Syringe Pump. All the tubing used in this setup is of PFA capillary
tubing (1/16" OD x 500 μm ID). T-mixer used for mixing of the two coming streams from
syringe pump is made of PEEK with 500 μm ID purchased from IDEX Health and Science.
Fluidic connections are made with ¼-28 flat bottomed flangeless fittings (IDEX Health and
50
3.3. Experimental
Science). Contacting of the phases takes place on capillary tubing of 100 cm length. Two
phases are then separated with flow liquid liquid extraction (FLLEX) module by Syrris Ltd.,
UK and are subsequently collected in vials and analyzed. Syrris FLLEX module contains a
porous PTFE membrane (0.22 μm pore size) placed inside a separator chip. The membrane
is hydrophobic and selectively allows organic stream to pass through. To ensure good
separation the pressure across the membrane is controlled via two back pressure regulators.
To achieve pressure difference along the separator it is connected to an air or nitrogen
source of 4 bar. The back pressure is set to 2 bar. Then the cross membrane pressure is
adjusted to achieve good separation of the phases.
Figure 3.2. Experimental setup of microseparation unit used for liquid-liquid extraction and
copper scavenging experiments
Figure 3.3. Experimental setup of one-pot CuAAC click microreactor coupled with copper
scavenging unit in flow
3.3.2. Experimental setup of one-pot CuAAC click reaction and inline copper
scavenging in flow
Equipment configuration of one-pot CuAAC click reaction coupled to copper scavenging is
shown in Figure 3.3. Chemyx Fusion 200 Touch Syringe Pump is used to deliver reagents
from BD plastic syringes to the reactor. A 3 mL syringe with phenylacetylene,
benzylbromide, sodium azide (NaN3) and DMSO is mounted to the syringe pump. Another
3 mL syringe containing the copper catalyst and DMSO is also mounted to the syringe
pump. All tubing used in this setup is made of PFA capillary tubing (1/16" OD x 500 μm
ID). T-mixer used to mix the reagents is made of PEEK with 500 μm ID purchased from
IDEX Health and Science. Fluidic connections are made with ¼-28 flat bottomed flangeless
51
fittings (IDEX Health and Science). Reactor has a volume of 400 μL made of PFA capillary
tubing (200 cm length, 1/16" OD x 500 μm ID). It is submerged in an oil bath with
temperature controlled via thermocouple (IKA ETS-D5) and maintained with a heating
plate (IKA RCT Basic). Backpressure regulator of 40 psi (IDEX Health and Science) is
used. Upon exiting the reactor, solution is quenched with ethyl acetate (EtOAc) and EDTA
which are loaded into two separate 5 mL BD plastic syringes delivered by another Chemyx
Fusion 200 Touch Syringe Pump. With a PEEK cross mixer (IDEX Health and Science)
these two streams are brought together with the reaction outlet mixture. They flow along
PFA capillary tubing (100 cm length, 1/16" OD x 500 μm ID) and then are delivered to
Syrris FLLEX module. The organic and aqueous phases are separated and subsequently
collected in vials and analyzed.
3.4. Results and Discussion

3.4.1. Performance characterization of the flow liquid-liquid extraction unit
Three two-phase model systems are used to evaluate the extraction efficiency of
the flow liquid-liquid extraction unit (Table 3.2). The first one constitutes the
extraction of dimethylformamide (DMF) from dichloromethane (DCM) into water.
A second system is extraction of acetone from water into toluene. Third system is
extraction of succinic acid from water into n-butanol. The latter two model systems
are recommended standard systems of the European Federation of Chemical
Engineering (EFCE). They were applied in the past successfully by us to
characterize flow extraction of dispersed flows (emulsions) made by
micromixers.93 They differ considerably in their interfacial tension where water
(acetone) / toluene system has a high value and water (succinic acid) / n-butanol
system has a low value as seen in Table 3.2.
For the system of water (succinic acid) / n-butanol the phase separator failed to
achieve complete separation of the phases for all the cross membrane pressure
variation (50-300 mbar) made. Due to very low interfacial tension of this system,
an emulsion readily forms and it is difficult to break.
Table 3.2. Model systems used to evaluate performance of the liquid-liquid extraction unit
Model System Feed Solvent Interfacial Tension (mNm-1)

1 DMF in DCM Water 39.1
2 Acetone in Water Toluene 33.0
3 Succinic acid in Water n-Butanol 1.7
52
3.4.1.1. Model system of DMF-DCM-water

For the model system of DMF-DCM-water, experiments are performed with three different
DMF content in DCM of 9.2 mol.%, 11.5 mol.% and 13.7 mol.% and the resulting
concentration of DMF in the extract stream is measured. Values from our experiments
match closely with experimental results achieved with a similar flow extraction unit
exhibiting slug flow in literature29 as seen in Figure 3.4. With their silicon microfabricated
device packaged with the membrane, flow rates of aqueous and organic phases of 25 μL
min-1 each were achieved.29 With this device flow rates of aqueous and organic phases of
500 μL min-1 each is achieved.
For a system with a constant partition coefficient, the extraction yield is constant. By
performing a mass balance, this DMF extraction yield is calculated as 60 %. The partition
coefficient which is the ratio of mol.% of DMF in extract to mol.% of DMF in feed is
nearly constant as evidenced by good fit of experimental points to straight line. The
partition coefficient which is the slope of this line is found as 0.15. These values agree well
with equilibrium data obtained from shake-flask experiments and literature.29 Thus, a single
equilibrium extraction stage is achieved.
Figure 3.4. Extraction of DMF from DCM into water, performance comparison of our
work and literature
3.4.1.2. Model system of acetone-water-toluene
For the EFCE recommended standard system of acetone-water-toluene, a satisfactory
partition coefficient up to 0.5 is achieved, yet still below the thermodynamic partition
coefficient of 0.8. This is expected as the flow rate of the setup is limited to 60 mL h-1 due
to maximum flow rate limit of the phase separator. At flow rates above 800 mL h-1,
53
conditions close to thermodynamic equilibrium were reached for this model system in
previous experiments utilizing micromixers.93 Yet otherwise, a perfect phase separation is
consistently achieved for this system indicating good performance of the membrane.
Feed and solvent are fed at equal rate and alternate slugs are visualized. Slug lengths are
seen to get larger by decreasing the flow rate (Figure 3.5). This observation agrees with the
work of Dessimoz et al. for the toluene water system studied with a T-mixer.94
According to the performance characterization with the three model systems, flow liquid-
liquid extraction unit is seen to have good, applicable performance apart from the systems
that have low interfacial tension.
Figure 3.5. Flow pattern of acetone-water-toluene system for 9 mL h-1 (left) and 30 mL h-1
(right)
3.4.2. Continuous copper scavenging using flow liquid-liquid extraction unit

With this confirmation of the good performance of the liquid-liquid extraction unit, copper
scavenging experiments were performed. The use of different types of copper catalysts and
metal scavengers were investigated together with variation of metal scavenger
concentration and pH.
3.4.2.1. Copper scavenging performance for different types of copper catalysts
Copper scavenging of various copper sources (CuBr(PPh3)3, [Cu(phen)(PPh3)2]NO3, CuI,
CuSO4, Cu(OAc)2) from organic solvent was studied. The copper extraction performance
achieved with the unit for the various copper sources tested is presented in Table 3.3.
Toluene is selected as solvent and it is fed at the same flow rate with the aqueous EDTA.
However, CuI, CuSO4 and Cu(OAC)2 are not miscible in toluene so a mixture of DMF and
toluene is used. In all cases slug flow is observed. Experimental results show excellent
performance of copper extraction at EDTA concentration of 0.016 M and pH of 9.4 and
extraction performance as high as 99% is achieved. Catalyst source of CuBr(PPh3)3 is used
as basis to study the effects of other parameters. In aqueous solution CuII is the more stable
oxidation state. CuI ions in solution can be oxidized into CuII form. The CuII form is the
only species which complexes with EDTA in aqueous solution. The copper-EDTA complex
formed is very stable with a high stability constant of log K 18.8, making this process
practically irreversible.95
54
Table 3.3. Copper scavenging performance for different types of copper sources
Copper Source Solvent Copper Content (ppm) Extraction

Feed Aq. outlet Org. outlet (%)
CuBr(PPh3)3 Toluene 123 120 3 98

[Cu(phen)(PPh3)2]NO3 Toluene 123 122 1 99
CuI DMF/Toluene 123 121 2 98
CuSO4 DMF/Toluene 162 136 26 84
Cu(OAc)2 DMF/Toluene 183 160 23 88
3.4.2.2. Copper scavenging performance with variation of scavenger type and

concentration
For the experiment with CuBr(PPh3)3 in toluene given in Table 3.3, the molar ratio of
scavenger to copper is 10 using 0.016 M EDTA. The effect of changing the scavenger
concentration is also investigated (Table 3.4). For EDTA concentration of 0.0016 M, which
corresponds to an equimolar scavenger and copper concentration, the copper extraction
performance of 91% is achieved. The residual copper content in the organic phase is still
below API limit indicating that lower amounts of scavenger are sufficient to achieve
required performance.
Additionally, the use of other metal scavengers, such as DTPA and [S,S]-EDDS, is studied
for both scavenger concentrations for extraction of CuBr(PPh3)3 in toluene. These results
are also presented in Table 3.4. DTPA shows slightly higher extraction performance than
EDTA at 0.0016 M concentration. This can be explained by its molecular structure (Figure
3.6). EDTA has six coordination sites (four acid and two amine) available to bind the metal
whereas DTPA has eight sites (five acid and three amine). This behavior is also observed
by others.39 [S,S]-EDDS, a structural isomer of EDTA, shows lower performance. This is
because it forms relatively weak complexes with metal.36 But this property enables it to be
easily biodegradable making it attractive to be used.
For studying the effects of other parameters the concentration is set to 0.016 M and
scavenger type of EDTA is used. At this concentration, EDTA and DTPA have similar
extraction performance. EDTA is the most widely used scavenger in literature therefore it is
selected as basis for analysis of other effects.
55
Table 3.4. Copper scavenging performance for different types of copper sources
Metal Scavenger Concentration Copper Content (ppm) Extraction

(M) Feed Aq. outlet Org. outlet (%)
EDTA 0.0160 123 120 3 98

DTPA 0.0160 142 139 3 98
[S,S]-EDDS 0.0160 118 114 4 97
EDTA 0.0016 126 115 11 91
DTPA 0.0016 124 118 6 95
[S,S]-EDDS 0.0016 118 81 37 70
Figure 3.6. Molecular structure of EDTA (left), DTPA (middle) and [S,S]-EDDS (right)
3.4.2.3. Copper scavenging performance with variation of scavenger solution pH
The acidity of the scavenger solution also affects the extraction performance. Therefore, the
effect of pH of scavenger solution is studied in the range 6-11 for the extraction of
CuBr(PPh3)3 in toluene using 0.016 M EDTA. As seen in Figure 3.7, extraction
performance increases with the increase in pH up to pH of about 10 and after that it shows a
decrease. Therefore, the optimum pH range is seen to be between 8.5 and 9.5. For studying
effects other than pH an optimum value of pH is set as 9.4. It is explained in literature that
the increase in pH changes the protonation stage providing availability of metal binding
sites.35
56
Figure 3.7. Copper scavenging performance with variation of scavenger solution pH

3.4.2.4. Recovery and reuse of chelating agent EDTA
Recovery and recycling of EDTA is attractive in terms of cost saving. It is also beneficial
regarding environmental concern of EDTA being not easily degradable. EDTA recovery is
performed by acidification of the aqueous outlet from phase separation using sulfuric acid.
Lowering the pH breaks the metal-chelate complex and EDTA is recovered in the form of
precipitate. About 87 % of EDTA is recovered by filtration. To investigate its reusability it
is redissolved to identical concentration with fresh EDTA experiments. Its copper content is
analyzed and seen to be less than 1 ppm indicating it can be recovered mostly free from
metal. Significant extracting performance is observed with the recycled EDTA of 92 %
(Table 3.5). Overall, recovered EDTA still has high extraction efficiency.
Further recycle was not performed but in literature extraction efficiency of recovered
EDTA was investigated up to four recycles. It was seen that extraction efficiency decreased
after each recycle. After the fourth recycle, approximately 20 % less extraction efficiency
than fresh EDTA was observed and consequently no further recycling was carried out.35
Overall, recycled EDTA was still effective and showed that EDTA can be used effectively
up to four cycles.34
Table 3.5. Copper scavenging performance comparison between fresh and recycled EDTA
Metal Scavenger Concentration Copper Content (ppm) Extraction

(M) Feed Aq. outlet Org. outlet (%)
Fresh EDTA 0.0160 123 120 3 98

Recycled EDTA 0.0160 134 127 11 92
57
3.4.3. One-pot CuAAC click reaction and inline metal scavenging in flow
After demonstrating the high performance of the continuous copper scavenging unit,
reaction experiments were started. The copper-catalyzed cycloaddition of benzylazide and
phenylacetylene is studied. However, instead of producing benzylazide separately and then
reacting it with phenylacetylene a one-pot methodology is adapted.51 Accordingly,
benzylazide is formed in situ from benzylbromide and sodium azide and is directly reacted
with phenylacetylene with the reaction scheme given in Figure 3.8. In this way isolation
and handling of instable azide is eliminated and the process is simplified. It makes the
process also safer by enabling consumption of the azide as soon as it forms. The catalyst of
[Cu(Phen)(PPh3)2]NO3 is selected based on previous work showing this catalyst to give the
highest performance.9 A tubular PFA microreactor of 400 μl volume is used. For reaction
temperature of 170 °C, pressure of 40 psi and catalyst loading of 5 mol.%, a good yield of
66 % is obtained in 10 minutes. By increasing the reaction/residence time, the triazole yield
is improved up to 92 % in 30 minutes. After achieving good performance with the
microreactor we set to explore the efficiency of our copper scavenging unit by connecting it
after the reactor with the setup seen in Figure 3.3. The reactor outlet is mixed with 0.016 M
EDTA and ethyl acetate that is used for quenching. Scavenging takes place along the tube
and then the mixture is passed through the FLLEX module. Complete separation of organic
and aqueous phases is achieved. The copper content in the organic phase is analyzed and
found to be lower than API limit of 15 ppm for three residence time variations studied as
given in Table 3.6.
Figure 3.8. One-pot copper-catalyzed azide alkyne cycloaddition

Table 3.6. Triazole yield and copper scavenging performance with variation of residence
time
Residence time Triazole yield Copper Content (ppm) Extraction

(min) (%) Feed Aq. outlet Org. outlet (%)
10 66 577 574 3 99
20 84 547 533 14 97
30 92 547 538 9 98
58
3.5. Conclusion
3.5. Conclusion
In order to develop a complete chemical processes on a micro-scale or related somewhat
higher continuous scale (milli, meso), the joint development of reaction as well as
separation techniques in microfluidic or other smart-scaled flow devices is required. It is
seen that to achieve this further especially the development of microseparation units and
respective effective flow protocols is needed. We decided to start our investigations on
microseparation and its connection with the copper-catalyzed azide-alkyne cycloadditon
(CuAAC) reaction. A continuous uninterrupted metal scavenging unit is developed based
on liquid-liquid extraction using slug flow. The phase separation is achieved with a porous
PTFE membrane. The unit is tested with model extraction systems and found to give good,
applicable performance apart from the system with low interfacial tension. With such
confirmation, copper scavenging experiments were performed. The use of different types of
copper catalysts and metal scavengers were investigated together with variation of metal
scavenger concentration and pH. For scavenger to copper molar ratio of 10, the scavengers
EDTA, DTPA and EDDS all showed high performance of about 98 %. For molar ratio of 1,
DTPA showed slightly better performance than EDTA whereas EDDS showed lower
performance as explained by their molecular structure. With the lower EDTA
concentration, the residual copper content in organic is found still below API limit
indicating less amount of scavenger is sufficient to achieve required performance. For
studying the effects of other parameters the molar ratio of 10 and scavenger type of EDTA
is used. The scavenger was seen to have high performance for extracting various copper
sources with up to 99 % extraction efficiency. Strong effect of pH of scavenger solution on
extraction performance was seen. The optimum pH range for copper extraction with EDTA
was found to be between 8.5 and 9.5. Recovery of EDTA was also performed. The recycled
EDTA successfully extracted 92 % of copper, only 6 % lower than fresh EDTA. Finally,
the copper scavenging unit was coupled to one-pot CuAAC click reaction and yield of 92 %
was obtained with 30 minutes residence time using 5% mol loading of
[Cu(Phen)(PPh3)2]NO3 as homogeneous catalyst. Residual copper was efficiently removed
in flow and the level of purity required for pharmaceuticals (APIs) was met in one stage of
separation.
In conclusion, a microextraction unit has been developed for metal catalyst scavenging,
widely applicable for modern organic synthesis. Coupling it to chemical reaction in flow
enables high yield and low metal content of the desired product. This opens doors to exploit
process-design intensification at a multi-step synthesis with vision to achieve synthesis on a
pilot scale.
59
3.6. Appendix
3.6.1. General chemicals information
Chemicals were purchased in their highest purity available from Sigma Aldrich.
Scavenging solutions from ethylene diamine tetraacetic acid (EDTA) and diethylene
triamine pentaacetic acid (DTPA) were prepared by mixing with demineralized water
(purified with a Millipore Elix UV-5 machine) and careful addition of ammonium
hydroxide (NH4OH) 5 M drop by drop until the solids disappear and the desired pH is
achieved. [S,S]-ethylene diamine disuccinic acid ([S,S]-EDDS) was purchased as solution
(35% water), it is diluted to desired concentration and NH4OH is added to get the desired
pH.
3.6.2. General analytical information

Analysis for performance characterization experiments are done using GC-FID on a Varian
430-GC with a flame ionization detector using a CP-Sil 8 CB column (length 60 m,
diameter 0.25 mm, film thickness 0.1 μm), with helium as the carrier gas. External
calibration is performed by making solutions of varying concentrations 0, 2, 4, 6, 8, and 10
wt.% of solute in organic phase. Sample from organic outlet is taken and analyzed for
amount of solute.
Analysis for copper scavenging experiment is performed using Inductively Coupled Plasma
– Optical Emission Spectroscopy (ICP-OES) from Sysmex SpectroBlue. 1 mL sample of
aqueous outlet (containing EDTA–copper complex) is diluted to 100 mL using oven dried
volumetric flask, sample is then analyzed using ICPOES with prior calibration. Calibration
is performed by making a calibration curve from 0 to 10 ppm of copper. Calibration
samples were prepared using 1000 mg L-1 TraceCERT standard solution available from
Sigma–Aldrich.
Analysis for one-pot CuAAC click reaction to determine yield of triazole is done using 1H
NMR. Nuclear magnetic resonance spectra were recorded on a Varian Mercury 400
instrument. All 1H NMR are reported in δ units, parts per million (ppm), and were
measured relative to the signal of tetramethylsilane (0 ppm) in deuterated solvent (CDCl3).
Prior to NMR analysis, high purity toluene is added as internal standard for quantification
of yield. The following abbreviations were used to explain the multiplicities: s = singlet, d
= doublet, t = triplet, q = quartet, m = multiplet.
3.6.3. Typical experimental procedure for performance characterization of the

flow liquid-liquid extraction unit
Experimental procedure for model system of dimethylformamide (DMF) - dichloromethane
(DCM) - water for feed of 10 wt. % (11.5 mol.%) DMF in DCM is described. One 5 mL
syringe is loaded with a solution of 10 wt. % DMF in DCM. Another 5 mL syringe is
60
3.6. Appendix
loaded with demineralized water which is the solvent for this system. A single syringe
pump is used to deliver the chemicals in two syringes at flow rate of 30 mL h making
total flow in the unit 60 mL h . They are mixed with a T-mixer and flow along 100 cm
capillary tubing where partitioning takes place. The phases form alternate slugs along the
tube and they are separated passing through the Syrris FLLEX module. Cross membrane
pressure is adjusted to 250 mbar to achieve good separation of the phases. The organic
stream is collected from the top and aqueous stream from the bottom of the module into
vials. Analysis of the organic phase is performed using GC-FID to determine the amount of
solute. The amount of solute in aqueous stream is then determined with mass balance and
subsequently partitioning coefficient is calculated. The experiment is conducted at room
temperature.
3.6.4. Typical experimental procedure for continuous copper scavenging using

flow liquid-liquid extraction unit
Experimental procedure of extraction of copper using CuBr(PPh3)3 as copper source in
toluene using 0.016 M EDTA with 9.4 pH is described. In a volumetric flask EDTA (116.8
mg) is added with 25 mL demineralized water. Solution pH is adjusted adding drop by drop
5 M NH4OH (about 10 drops) until pH of 9.4 enabling EDTA to be fully dissolved. One 5
mL syringe is loaded with this 0.016 M EDTA aqueous solution. Another 5 mL syringe is
loaded with CuBr(PPh3)3 (7.8 mg) dissolved in 5 mL toluene (123 ppm Cu). A single
syringe pump is used to deliver the chemicals in two syringes at flow rate of 30 mL h
making total flow in the unit 60 mL h . They are mixed with a T-mixer and flow along
capillary tubing of 100 cm where chelation between EDTA and copper takes place. The
phases form alternate slugs along the tube and they are separated passing through the Syrris
FLLEX module with the cross membrane pressure adjusted to 250 mbar. The aqueous
phase collected from the bottom contains the copper source removed from the organic
phase and organic phase collected from the top is mostly free from copper. The aqueous
phase is analyzed using ICP-OES to determine the copper amount scavenged. The
experiment is conducted at room temperature.
3.6.5. Typical experimental procedure for recovery of EDTA

To the sample from aqueous outlet of the copper scavenging experiment, sulfuric acid 5M
is added drop by drop until pH of 2 is attained resulting in a slurry which is monitored with
a pH meter. Slurry is then filtered using oven dried Whatman filter paper with known dry
weight. After filtering, filter paper with the captured solids is dried in 70 °C oven for 120
minutes and subsequently weighed to determine percentage recovery. The solid is then
redissolved and analyzed using ICP-OES to determine the copper content of the recovered
EDTA. This recovered EDTA is then reused in copper scavenging experiments by
preparing a solution of 0.016 M.
61
3.6.6. Typical experimental procedure for one-pot CuAAC click reaction and
inline copper scavenging in flow
An oven dried volumetric flask is charged with sodium azide (39 mg, 0.6 mmol dissolved
in DMSO), phenylacetylene (0.45 mmol), benzylbromide (0.45 mmol). It is flushed with
argon and DMSO is added to make the solution volume 3 mL and is loaded in a syringe.
Another 3 mL syringe is loaded with [Cu(phen)(PPh3)2]NO3 (5% mol, 18.6 mg, 0.0225
mmol) dissolved in DMSO. A single syringe pump is used to deliver the chemicals in two
syringes. Flow rate is adjusted for each residence time variation (1.2 mL h each for 10
minutes residence time). They are mixed with a T-mixer and then flowed through the
reactor with a volume of 400 μL. The reactor is submerged in an oil bath temperature
adjusted to 170 °C. A backpressure regulator of 40 psi (2.76 bar) is used. Upon exiting the
reactor the reactor stream is quenched with ethyl acetate and 0.016 M EDTA loaded into
two separate 5 mL syringes. These two streams are delivered by an additional syringe pump
with a flow rate adjusted to 7.2 mL h for 10 minutes residence time experiment. With a
cross mixer these two streams are brought together with the reaction outlet mixture. They
flow along PFA capillary tubing of 100 cm length where chelation between EDTA and
copper takes place. Finally passing through the Syrris FLLEX module organic and aqueous
phases are separated. The aqueous phase sample is analyzed using ICP-OES to determine
the copper amount scavenged. The organic phase sample is washed with demineralized
water and extracted with diethylether four times (4 2 mL). The organic layers after each
extraction are collected. The combined organic is dried over MgSO4 and filtered. Then the
solvent is removed under vacuum to yield triazole as solid. Deuterated solvent of CDCl3
and toluene as internal standard (0.5 mmol) is then added and 1H NMR analysis is done to
determine the yield. 1-benzyl-4-phenyl-1H-1,2,3-triazole: 1H NMR (400MHz, CDCl3): δ =
5.57 (s, 2H), 7.25 – 7.44 (m, 8H), 7.66 (s, 1H), 7.73-7.85 (m, 2H) ppm.
References
2. A. I. Stankiewicz and J. A. Moulijn, Chemical Engineering Progress, 2000, 96, 22–
34.
3. J. C. Charpentier, Chemical Engineering Journal, 2007, 134, 84–92.
Development, 2010, 14, 215–224.
62
References

Chemistry, 2009, 2009, 1321–1325.
6, 2220–2225.
ChemSusChem, 2015, 8, 504–12.
& Technology, 2012, 35, 1184–1204.
2011, 66, 1426–1448.
467–478.
2011, 167, 504–509.
2013, 6, 746–89.
Technik, 2012, 84, 660–684.
20. C. Wiles and P. Watts, European Journal of Organic Chemistry, 2008, 2008, 1655–
1671.
2011, 167, 718–726.
Journal, 2010, 158, 233–240.
in English), 2007, 46, 5704–5708.
63

Research, 2007, 46, 8420–8430.
33. J. Yu and D. Klarup, Water, Air, & Soil Pollution, 1994, 75, 205–225.
34. S. Goel, K. K. Pant, and K. D. P. Nigam, Journal of Hazardous Materials, 2009,
171, 253–261.
35. G. Chauhan, K. K. Pant, and K. D. P. Nigam, Industrial & Engineering Chemistry
Research, 2013, 52, 16724–16736.
36. G. Chauhan, K. K. Pant, and K. D. P. Nigam, Industrial Engineering Chemistry
Research, 2012, 51, 10354–10363.
37. D. Kołodyńska, Chemical Engineering Journal, 2010, 165, 835–845.
38. L. M. Gando-Ferreira, I. S. Romão, and M. J. Quina, Chemical Engineering
Journal, 2011, 172, 277–286.
39. S. Goel and A. Gautam, Hydrometallurgy, 2010, 101, 120–125.
40. M. Kowalczyk, Z. Hubicki, and D. Kołodyńska, Chemical Engineering Journal,
2013, 221, 512–521.
41. P. Wu and V. V Fokin, Aldrichimica Acta, 2007, 40, 7–17.
42. M. Meldal and C. W. Tornøe, Chemical reviews, 2008, 108, 2952–3015.
43. S. Hassan and T. J. J. Müller, Advanced Synthesis & Catalysis, 2015, 357, 617–
666.
44. J. E. Moses and A. D. Moorhouse, Chemical Society reviews, 2007, 36, 1249–1262.
45. H. C. Kolb and K. B. Sharpless, Drug Discovery Today, 2003, 8, 1128–1137.
46. V. D. Bock, H. Hiemstra, and J. H. Van Maarseveen, European Journal of Organic
Chemistry, 2006, 2006, 51–68.
1290.
48. M. Fuchs, W. Goessler, C. Pilger, and O. Kappe, Adv. Synth. Catal., 2010, 352,
323–328.
49. C. D. Smith, I. R. Baxendale, S. Lanners, J. J. Hayward, S. C. Smith, and S. V Ley,
Organic & biomolecular chemistry, 2007, 5, 1559–1561.
50. A. R. Bogdan and N. W. Sach, Advanced Synthesis & Catalysis, 2009, 351, 849–
854.
51. P. Appukkuttan, W. Dehaen, V. V Fokin, and E. Van der Eycken, Organic Letters,
2004, 6, 4223–4225.
52. A. D. Moorhouse and J. E. Moses, Synlett, 2008, 14, 2089–92.
53. J. Andersen, S. Bolvig, and X. Liang, Synlett, 2005, 19, 2941–2947.
54. M. M. E. Delville, P. J. Nieuwland, P. Janssen, K. Koch, J. C. M. van Hest, and F.
P. J. T. Rutjes, Chemical Engineering Journal, 2011, 167, 556–559.
55. European Medicines Agency, Note for Guidance on Specification Limits for
Residues of Metal Catalysts, Doc. Ref. CPMP/SWP/QWP/4446/00, 2002.
64
References
56. R. Franke, M. Rudek, and G. Baumgarten, Evonik Science Newsletter, 2010,

elements30, 6–11.
57. S. G. Newman and K. F. Jensen, Green Chemistry, 2013, 15, 1456–1472.
58. A. J. Sandee, D. G. Petra, J. N. Reek, P. C. Kamer, and P. W. van Leeuwen, Chem.
Eur. J., 2001, 7, 1202–8.
59. G. W. Lamb, F. A. Al Badran, J. M. J. Williams, and S. T. Kolaczkowski,
Chemical Engineering Research and Design, 2010, 88, 1533–1540.
60. D. Cantillo and C. O. Kappe, ChemCatChem, 2014, 6, 3286–3305.
61. A. Kirschning, W. Solodenko, and K. Mennecke, Chem. Eur. J., 2006, 12, 5972–
90.
62. S. C. Stouten, T. Noël, Q. Wang, and V. Hessel, Chemical Engineering and
63. U. Hintermair, Z. Gong, A. Serbanovic, M. J. Muldoon, C. C. Santini, and D. J.
Cole-Hamilton, Dalton transactions, 2010, 39, 8501–10.
64. R. Duque, E. Öchsner, H. Clavier, F. Caijo, S. P. Nolan, M. Mauduit, and D. J.
Cole-Hamilton, Green Chemistry, 2011, 13, 1187–1195.
65. M. Baumann, I. R. Baxendale, and S. V Ley, Molecular diversity, 2011, 15, 613–
30.
66. S. V. Ley, I. R. Baxendale, R. N. Bream, P. S. Jackson, A. G. Leach, D. A.
Longbottom, M. Nesi, J. S. Scott, R. I. Storer, and S. J. Taylor, Journal of the
Chemical Society, Perkin Transactions 1, 2000, 3815–4195.
67. P. H. Seeberger, Nature chemistry, 2009, 1, 258–60.
68. C. J. Welch, J. Albaneze-Walker, W. R. Leonard, M. Biba, J. DaSilva, D.
Henderson, B. Laing, D. J. Mathre, S. Spencer, X. Bu, and T. Wang, Organic
69. D. J. Cole-Hamilton and R. P. Tooze, Eds., Catalyst Separation, Recovery and
Recycling, Springer Netherlands, Dordrecht, The Netherlands, 2006, vol. 30.
70. B. Cornils, Organic Process Research & Development, 1998, 2, 121–127.
71. J. Mol, Journal of Molecular Catalysis A: Chemical, 2004, 213, 39–45.
72. K. Mikami, Ed., Green Reaction Media in Organic Synthesis, John Wiley & Sons,
2008.
73. F. M. Kerton and R. Marriott, Alternative Solvents for Green Chemistry, Royal
Society of Chemistry, Cambridge, UK, 2013.
74. A. B. Theberge, G. Whyte, M. Frenzel, L. M. Fidalgo, R. C. R. Wootton, and W. T.
S. Huck, Chemical communications, 2009, 6225–7.
75. J. F. B. Hall, X. Han, M. Poliakoff, R. A. Bourne, and M. W. George, Chemical
communications, 2012, 48, 3073–5.
76. S. Liu, T. Fukuyama, M. Sato, and I. Ryu, Organic Process Research &
Development, 2004, 8, 477–481.
77. P. B. Webb, T. E. Kunene, and D. J. Cole-Hamilton, Green Chemistry, 2005, 7,
373–379.
78. P. Li, J. S. Moore , and K. F. Jensen, ChemCatChem, 2013, 5, 1729–1733.
65
79. S. Sharma, K. C. Basavaraju, A. K. Singh, and D.-P. Kim, Organic letters, 2014,
16, 3974–7.
80. H. P. Dijkstra, G. P. M. van Klink, and G. van Koten, Accounts of Chemical
Research, 2002, 35, 798–810.
81. M. Janssen, C. Müller, and D. Vogt, Green Chemistry, 2011, 13, 2247–2257.
82. L. Peeva, J. Arbour, and A. Livingston, Organic Process Research & Development,
2013, 17, 967–975.
83. N. Brinkmann, D. Giebel, G. Lohmer, M. T. Reetz, and U. Kragl, Journal of
Catalysis, 1999, 183, 163–168.
84. H. P. Dijkstra, N. Ronde, G. P. M. van Klink, D. Vogt, and G. van Koten,
Advanced Synthesis & Catalysis, 2003, 345, 364–369.
85. J. Fang, R. Jana, J. A. Tunge, and B. Subramaniam, Applied Catalysis A: General,
2011, 393, 294–301.
86. A. Kajetanowicz, J. Czaban, G. R. Krishnan, M. Malińska, K. Woźniak, H.
Siddique, L. G. Peeva, A. G. Livingston, and K. Grela, ChemSusChem, 2013, 6,
182–92.
87. M. Janssen, J. Wilting, C. Müller, and D. Vogt, Angewandte Chemie (International
ed. in English), 2010, 49, 7738–41.
88. M. Priske, K.-D. Wiese, A. Drews, M. Kraume, and G. Baumgarten, Journal of
Membrane Science, 2010, 360, 77–83.
89. D. Ormerod, B. Bongers, W. Porto-Carrero, S. Giegas, G. Vijt, N. Lefevre, D.
Lauwers, W. Brusten, and A. Buekenhoudt, RSC Advances, 2013, 3, 21501–21510.
90. VITO, Integration of separation technology in (bio) chemical processes,
https://vito.be/en/integration-separation-technology-bio-chemical-processes.
91. SP Process Development, New recycling methodology for homogeneous noble
metal catalysts, http://www.sp.se/sv/units/sppd/affarsomraden2/Sidor/default.aspx.
92. J. Balster and A. Boam, Organic Solvent Nanofiltration (OSN), Evonik presentation
at the EMS Summer School,12 July 2012, Nancy.
93. K. Benz, K. P. Jackel, K. J. Regenauer, J. Schiewe, K. Drese, W. Ehrfeld, V.
Hessel, and H. Lowe, Chemical Engineering & Technology, 2001, 24, 11–17.
Science, 2008, 63, 4035–4044.
95. A. E. Martell and R. M. Smith, Critical Stability Constants, Plenum Press, New
York, 1974.
66
CHAPTER 4
Scale-up of Flow Extraction Utilizing
Coiled Flow Inverter – Drop-In Process
Integration at Pilot Scale
Vural Gursel, I., Kurt, S.K., Aalders, J., Wang, Q., Noël, T., Nigam, K.D.P., Kockmann, N.
& Hessel, V. (2015) Utilization of milli-scale coiled flow inverter in combination with
phase separator for continuous flow liquid-liquid extraction processes. Chemical
Engineering Journal, 283, 855-868.
Kurt, S.K., Vural Gursel, I., Hessel, V., Nigam, K.D.P. & Kockmann, N. (2015) Liquid-
liquid extraction system with microstructured coiled flow inverter and other capillary
setups for single-stage extraction applications. Chemical Engineering Journal, 284, 764-
777.
Abstract:
Process-design intensification situated under the umbrella of Novel Process Windows heads
for process integration with the coupling of flow devices and here most development is
needed for flow separators. The vision is to achieve multi-step synthesis in flow on pilot
scale. This calls for scale-up of separation units. This study is rare in micro flow literature
and community and thus conducted here. The coiled flow inverter (CFI) was considered as
the right tool to work at higher flow rates. We report hereby for the first time, to best of our
knowledge, the use of the CFI for immiscible liquid-liquid mixing and its application for
liquid-liquid extraction. A milli-scale CFI with tube internal diameter of 3.2 mm and tube
length of 210 cm was used where partitioning takes place in slug flow. Phase separation
was achieved with PTFE membrane flow separator or slit shaped flow separator composed
of glass and Teflon rectangular capillaries. The microextraction unit was tested for
European Federation of Chemical Engineering recommended model extraction systems of
toluene-water-acetone and n-butyl acetate-water-acetone. The effects of variation of total
flow rate at equal volume fraction as well as variation of aqueous volume fraction at
67
4. Scale-up of Flow Extraction Utilizing CFI
constant total flow rate on extraction performance were investigated. High extraction
efficiency close to thermodynamic equilibrium was achieved for both systems. The CFI
setup was also compared with straight tube and CFI showed about 20 % higher extraction
performance with the increase in flow rate. In straight tube, parallel flow developed at
higher flow rates whereas the CFI kept operating in the slug flow regime. Membrane
separator could be used up to 20 ml/min, however pure organic phase was attained
throughout. With slit shaped separator, for the butyl acetate-water-acetone system, 20 %
breakthrough at both outlets occurred at a flow rate of 32 ml/min. For the model system of
toluene-water-acetone flow rate up to 120 ml/min (7.2 l/h) was achieved at same
breakthrough level indicating the influence of interfacial tension on the operating window
of separator. It is important to point out that slug flow regime that is limited to microfluidic
systems was maintained even at such pilot scale (58 m3/y). It was seen that higher flow rate
is achieved in this study than the ones reported for slug flow in literature.
4.1. Introduction
The new and innovative equipment available with micro-process technology opens up new
opportunities of processing that are far from conventional practices 1–4. This new concept
was called Novel Process Windows 5–7. It has two fundamental pillars that are chemical
intensification and process-design intensification. Chemical intensification makes use of
highly intensified, unusual and typically harsh process conditions to boost micro-processing
8–17
. Process-design intensification is based on process simplification and integration 18–20. It
takes into consideration the chemical process as a whole, therefore is holistic in approach
21–23
. The majority of research in micro-process technology was focused on the
development of process units to be utilized in reaction, mixing and heat exchange 24–29.
Application of these units on large scale has also been studied 30–35. However, the
development of flow separation units is still largely limited. They are highly desired and
required to enable downstream unit operations in flow. Therefore, to realize a whole
production process in flow, further development of flow separators is asked for that can be
coupled to other processing units. This coupling is not rare anymore, but far from being
commonly applied 36–40. Furthermore, for pilot-scale application, flow separation units that
can handle high throughput are required. This was rarely studied 41, thus development of a
flow separation unit on a moderately large scale is conducted here.
There is an increasing number of flow separators developed dedicatedly for chemical
purposes 42. Liquid-liquid extraction has gained the most interest 43 but recently several
distillation 44–47, absorption 48–50 and chromatography 51–53 units were also described.
Liquid-liquid extraction is very important for chemical industry and it especially finds wide
application in pharmaceutical production 41. Additionally, high surface-to-volume ratios in
microchannels is especially advantageous for effective liquid-liquid contacting 43. Liquid-
liquid extraction has two parts of mixing and phase separation. At micro- and milliscales
68
4.1. Introduction
surface tension dominates over gravitational and inertial forces 41. Thus, surface tension is
used as the driving force for separation 54.
With the use of different mixing elements and process conditions different flow patterns are
observed 55. The flow pattern used has a large effect on the mixing and separation
performance 43. Slug flow was shown to have better mass transfer compared to parallel flow
56
. This is due to slug flow providing convection through internal circulations as well as
interfacial diffusion between two adjacent slugs 55. Mass transfer in dispersed flow was
seen to be the most effective due to larger interfacial area 57. However, the phase separation
performance should be taken into consideration when evaluating the whole extraction
process. Emulsions are very stable and long time is required for the coalescence of
produced droplets 57. On the other hand, phase separation of slug flow can be much easily
achieved in short times. Effective liquid-liquid slug flow separation was achieved with
PTFE membrane or microdevice that selectively lets the organic phase pass through
41,54,58,59
. Also, it was achieved with wettability-based separation units with branched
hydrophobic and hydrophilic outlets 55,60,61. Therefore, slug flow can be regarded as suited
best for practical use in flow applications. So far, slug flow has only been realized on a
laboratory scale. Highest flow rate of 4.8 L/h was presented by Cervera-Padrell et al.41 for
the application of phase separation of toluene and water.
Liquid-liquid extraction based on slug flow was achieved in previous studies with a setup
including a T-mixer, straight tube where partitioning takes place and a PTFE membrane
phase separator 39. Maximum flow rate attainable was 1 ml/min. Its coupling to the copper-
catalyzed azide-alkyne cycloaddition (click chemistry) reaction was achieved in flow. The
vision now is to achieve such multi-step synthesis in flow on pilot scale. This calls for
scale-up of flow separation unit and this is carried out here. So far, no application examples
of extraction using microeffects in higher-throughput are reported. For performing the
mixing, the coiled flow inverter (CFI) introduced by Saxena and Nigam62 was selected to
enable efficient mixing at higher flow rates. For performing the phase separation two
different units were tested. The first is a PTFE membrane separator developed by IMM
which is the Institut fuer Mikrotechnik Mainz GmbH (meanwhile renamed as Fraunhofer-
ICT Mainz) and the second is a slit shaped flow separator composed of glass and Teflon
rectangular capillaries developed by Gaakeer et al.60 Extracting efficiency of the setup was
tested for the European Federation of Chemical Engineering (EFCE) recommended model
extraction systems of toluene-water-acetone and n-butyl acetate-water-acetone 63. Flow
pattern was characterized using high-speed photographic system. Effects of variation of
flow rate and aqueous phase fraction on extraction efficiency were investigated. The
extraction efficiency of the setup was compared with a straight tube setup. Lastly, the
performances of the two separators were evaluated.
69
4.2. Coiled Flow Inverter

Numerous mixing strategies have been proposed to enhance mixing performance in
microfluidic systems. Curved channels offer a simple design that enable generation of
vortices.64 Formation of vortices remarkably contribute to convective mixing and, thus, the
axial dispersion can be decreased.65–68 Because of the action of centrifugal force that acts
perpendicular to the flow direction of the fluid, a secondary flow develops in curved
channels.69 The magnitude of this effect is characterized by the dimensionless Dean
number.70 At higher Dean numbers higher mixing performance is attained.71 Dean vortices
enhance the radial mixing in case of single-phase flow and thus narrower residence time
distribution compared to straight capillaries can be achieved.62,72
Mixing is further enhanced by arranging the curved channels in a way that chaotic
advection is created. It was achieved by complete flow inversions by inducing 90° bends at
equal intervals along the length of the tube.62 Coiled flow inverter (CFI) is based on this
principle. The direction of the centrifugal forces inducing the formation of secondary flow
profile in helically coiled tube is changed with the 90°-bends. Therefore, the secondary
flow profile is developed after the 90° bend in a different plane that is perpendicular to the
previous plane (Figure 4.1). As a result, the radial mixing can be further enhanced in CFI in
comparison to helically coiled tube. Hence, a narrow RTD close to an ideal plug flow
reactor can be achieved even at laminar flow regimes.73,74
Figure 4.1. CFI structure and cross-sectional velocity contours and streamlines at the inlet,
before and after 90° bend, where is the inner tube diameter and is the coil diameter
(adapted from Kurt et al.75).
70
4.3. Experimental
CFI was shown to be very effective as a heat exchanger 76–80 and as an inline mixer of two
miscible liquids 81,82 Furthermore, it was used for chemical reaction 83 and to study fluid
flow in two-phase gas-liquid systems 84–87. CFI was tested already for pilot plant scale with
tube internal diameter of 10 mm allowing for a liquid flow rate up to 2000 l/h 77.
These studies show that the high mixing efficiency of CFI for single phase and two phase
gas-liquid systems is very well understood. The use of CFI for immiscible liquid-liquid
mixing is open for research. The study of two-phase flow characteristic of CFI for the
application of liquid-liquid extraction is to be discussed below with a milli-scale CFI for
pilot scale. It has also been investigated in micro-scale with CFI of 1 mm inner diameter75.
The complete system was able to be operated with slug flow pattern for total volumetric
flow rates and volumetric flow ratios (aq/org) in the range of 1-8 ml/min and 0.5-2.0,
respectively. Extraction efficiency results revealed that the microstructured coiled flow
inverter provides better extraction efficiency up to 20 % compared to straight capillaries at
constant contact times. For a constant length of tube, increasing slug flow velocities slightly
increased the extraction efficiency of CFI due to increase in the intensity of Dean vortices.
The results show that the Dean vortices inside the helically coiled tubular devices offers
enhanced mixing than the internal circulations that occur in straight capillaries in case of
slug flow pattern. Volumetric liquid-liquid mass transfer coefficient values were found 1.5
to 2 folds greater compared to straight capillaries. Therefore the microstructured CFI was
found very promising for extraction purposes and CFIs are also compact in design which is
a further advantage compared to straight capillaries.
4.3. Experimental
4.3.1. Model systems used
Liquid-liquid extraction of two systems were considered: acetone from its aqueous solution
into toluene (toluene-water-acetone) and acetone from its aqueous solution into butyl
acetate (butyl acetate-water-acetone). These are two of the standard test systems
investigated by European Federation of Chemical Engineering (EFCE) 63. They differ in
terms of interfacial tension. In both systems 10 wt. % of acetone in water was used as
aqueous feed. The interfacial tension of the systems at this acetone concentration was
calculated with the correlation given in the EFCE book 63. The models systems are given in
Table 4.1.
Table 4.1. Model systems used
Model systems Interfacial tension [mN/m]

1) Toluene / Water + Acetone (10 wt.%) 17,8
2) n-Butyl Acetate / Water + Acetone (10 wt.%) 8,6
71
4.3.2. Experimental setup

The experimental setup used is shown in Figure 4.2. For the visualization of the flow
methylene blue was added to the aqueous phase. Aqueous feed was loaded into two plastic
syringes and these two syringes were mounted to one Chemyx Fusion 200 Touch syringe
pump. Organic feed was loaded into one plastic syringe and it was mounted on to another
Chemyx Fusion 200 Touch syringe pump. The syringes were connected to 1/8" OD, 1.55
mm ID PFA capillary tubing. The liquids were introduced to Swagelok stainless steel 1/8"
tube OD union cross mixing element. The aqueous streams enter from the sides and organic
stream enters from the middle as shown in Figure 4.2. CFI made with PFA tube (1/4" OD,
3.2 mm ID), which is explained in detail in the next paragraph, was attached downstream of
the union cross. At the end of this mixing and contacting stage, a phase separator was
connected. The separated phases from the two outlets of the phase separator were collected
in sampling bottles. Experiments were carried out in room temperature.
The CFI was made by coiling 1/4" OD PFA tube with tube internal diameter ( ) of 3.2 mm
around a 1" OD stainless steel cylindrical bar that is welded to make a square shape. This
results in coil diameter (shown as in Figure 4.1) of 32 mm. This is the distance from
center to center of the tube so it is found as the sum of cylindrical bar diameter and tube
outside diameter.
(a)
(b)
Figure 4.2. (a) Schematic diagram and (b) Photograph of experimental setup
72
4.3. Experimental
The corresponding curvature ratio ( ) is then equal to 10. With the square shape,
90° bends are introduced and it enables the direction of the centrifugal forces to change 3
times resulting in more efficient mixing. There is equal tube length before and after the
bends. Including the straight tube parts at the inlet and outlet of CFI, the mixing contacting
part has a total tube length of 210 cm.
4.3.3. Flow separation devices

For performing the phase separation two different units were tested. The first is a PTFE
membrane separator developed by IMM which is the Institut fuer Mikrotechnik Mainz
GmbH (meanwhile renamed as Fraunhofer-ICT Mainz) and the second is a slit shaped flow
separator composed of glass and Teflon rectangular capillaries developed by Gaakeer et al.
60
. The pictures of the separators are provided in Figure 4.3.
Figure 4.3. Flow separation devices, left: membrane separator, middle: slit shaped
separator and right: inside view of the bars within slit shaped separator
The IMM membrane separator is composed of microstructured plates with the membrane
sandwiched in between. Each of the two plates have 10 channels and channels have
dimensions of length x width x depth = 38 mm x 500 μm x 280 μm. The membrane used is
Pall Zefluor 0.5 μm pore PTFE membrane. It was cut to 44 mm x 14 mm from a 47 mm
round disk. The membrane allows permeation of the organic phase while aqueous phase is
retained. The device is sealed with allen screws. 1/4" OD PFA tube coming from the CFI was
directly connected to the separator inlet. The outlet tubes used were each 30 cm PFA tubes of
1/16" OD, 500 μm ID. The setup with the membrane separator is given in Figure 4.4.
The slit shaped separator consists of a slit between two glass bars and a slit between two
Teflon bars assembled in stainless steel housing. Stainless steel spacers are used to set the
desired slit height. In this study slit height (H as shown in Figure 4.3, right) of 0.4 mm was
used. Separation is achieved based on the preferential wetting properties of the liquids on
solid material. Organic phase has affinity toward the hydrophobic Teflon capillary whereas
aqueous phase has affinity to flow through the hydrophilic glass capillary. Details about this
device can be found in Gaakeer et al.88 From the CFI the tube was reduced to 1/16" OD,
500 μm ID as it enters the slit shaped separator. The side outlets of the separator are open to
73
atmosphere. To drain the separator two PFA tubes of 1/16" OD, 500 μm ID were connected
at the bottom outlets but there was no flow at these outlets during operation of the
experiments. The setup with the slit shaped separator is shown in Figure 4.2 b.
Figure 4.4. Photograph of the experimental setup with the membrane separator

4.4.1. Liquid – liquid two-phase flow patterns in CFI
To study the two-phase liquid-liquid patterns a frame of 10 ml/min organic flow rate by 10
ml/min aqueous flow rate was chosen. This is the same frame used in the papers by Kashid
et al.89,90 In other studies of flow regime, the flow rates used were lower. The aqueous and
organic phases were fed at all combinations of 2, 4, 6, 8, 10 ml/min of each phase.
Snapshots for each condition were taken. Not to make the figure overcrowded, pictures of
all conditions are not shown but a selection of them is provided. These are seen in Figure
4.5 for the toluene-water system. For the butyl acetate-water almost identical pictures were
obtained. On all pictures a bar showing the diameter of the tube is placed on one slug to
give an indication of size. The experiments show that the water is the dispersed phase and
forms convex shaped slugs while the organic is the continuous phase and has a concave
shape. Slug flow regime is defined with the slug lengths longer than the diameter of the
channel.91 This was observed almost in all points of the map. Only at very low aqueous to
organic ratios of 2 to 8 and 2 to 10 the slug sizes are slightly smaller than the diameter.
These points can be regarded to be in drop flow regime.91 For all the experiments no
coalescence or break-up of the slugs were observed.
It was observed that with this system the total flow rate doesn’t have much influence on the
slug size. This was also observed by Jovanovic et al.92 However, in some studies it was
seen that the slug lengths decrease with the increase in flow rate.91,93 This was explained to
occur due to rapid penetration of one phase into other segregating the stream into more
74
slugs. The reason why this was not observed by us could be because of the larger tube
diameter used in this study, this effect could be seen at much higher flow rates.
Figure 4.5. Flow pattern map of toluene-water system

The slug size was observed to be highly dependent on the ratio of the aqueous and organic
flow rates. It was seen that the aqueous slug length decreases as the organic flow rate
increases at constant aqueous flow rate. This is to be expected since with the decrease in
relative flow rate of the aqueous phase its volume fraction in the channel would decrease.
Similarly it was observed that the aqueous slug length increases as the aqueous flow rate
increases at constant organic flow rate.
Kashid and Agar91 observed that by using larger diameter capillaries the area in the flow
map where uniform slug flow is observed expanded. When 0.5 mm ID tube was used the
maximum flow rate for both liquids was 1.2 ml/min. This was increased to 1.7 ml/min each
when 0.75 mm and 1 mm ID tubes were used. This was the flow rate limit in their study.
Observing slug flow regime in a wider flow rate range using 3.2 mm ID tube in present
study is in accordance with their finding.
In their more recent study, Kashid et al.90 investigated using six parallel capillaries as scale-
up of their previous system. They observed that the slug flow regime expanded six fold
enabling it to be achieved at a flow rate of 10 ml/min for both liquids. They concluded that
this scale-up enabled the flow pattern can be preserved while total throughput is increased.
In this present study, it is also observed that slug flow regime can be preserved while higher
flow rates are achieved. Several authors studied mapping of the flow regimes in terms of
75
dimensionless numbers to define quantitatively the influence of different forces. Dessimoz

et al.93 made flow transition maps based on mean Capillary ( ) and Reynolds ( )
numbers. Zhao et al.94 proposed to the Weber ( ) number for the mapping of the flow
patterns to express whether interfacial tension or inertia is the dominant force.
Several authors plotted their maps as vs. .92,94,95 For the maximum flow rates
used in Figure 4.5 of 10 ml/min of each phase the calculation of these dimensionless
numbers were made for toluene-water (T-W) and butyl acetate-water (B-W) systems. This
is given in Table 4.2. Please note that acetone was not used in flow pattern investigation
experiments and therefore is not included also in these calculations. For interfacial tension
34 and 14 mN/m are used for toluene-water and butyl acetate-water, respectively.
Table 4.2. Reynolds, Capillary and Weber numbers of the organic and aqueous phases for
the two systems
, ,
System
ml/min ml/min
T-W 10 10 103 75 3.4 ×10-4 5.5 ×10-4 3.5 ×10-2 4.1 ×10-2
B-W 10 10 85 75 1.0 ×10-3 1.3×10-3 8.9 ×10-2 1.0 ×10-1
Dessimoz et al.93 calculated the average value of the two phases for the dimensionless
numbers and . They plotted vs. and identified a transition line
between parallel and slug flow regime. The equation of this curve can be calculated from
the figure and it shows that to be in slug flow regime the ratio of to should
be below about 8.3 10-5 . Making this calculation using Table 4.2 it is seen that the ratio is
much below this value confirming the slug flow regime observed in our studies. It was also
calculated for the different composition of aqueous and organic phases. It was seen that the
ratio is higher when the aqueous to organic ratio is higher and also higher for the butyl
acetate-water system. However, for all the conditions the ratio of to was
much below this threshold value.
Jovanovic et al.92 observed slug flow at values below 10-2. The values calculated for
our system are below this value as well. The Weber number values for both phases being
below 1 indicate that interfacial tension dominates compared to inertial forces producing
slugs. It is explained that the interfacial tension tends to reduce the interfacial area, while
inertia and viscous forces act to extent and drag the interface downstream.94 Surface
stresses resist break-up and keeps the aqueous phase in a slug structure.
76
4.4.2. Extraction performance with variation of aqueous phase volume

fraction at fixed total flow rate
The aqueous volume fraction is the aqueous flow rate divided by the total flow rate.
Extraction performances of the two systems were investigated with variation of aqueous
phase volume fraction from 0.1 to 0.9, while keeping the total flow rate constant. For
toluene-water-acetone system given in Figure 4.6 total flow rate was kept at 20 ml/min.
And for butyl acetate-water-acetone system given in Figure 4.7 total flow rate was kept at
10 ml/min.
It was observed that for toluene-water-acetone system (Fig. 4.6), the extraction efficiency
decreases with the increase in the aqueous volume fraction. This is because by increasing
the aqueous volume fraction the amount of toluene that can extract acetone from water
decreases.
Toluene-water-acetone
system
Figure 4.6. Extraction efficiency as a function of the aqueous volume fraction for the
toluene-water-acetone system, total flow rate: 20ml/min
Butyl acetate-water-acetone
system
Figure 4.7. Extraction efficiency as a function of the aqueous volume fraction for the butyl
acetate-water-acetone system, total flow rate: 10ml/min
77
For butyl acetate-water-acetone system (Fig. 4.7), it was observed that the aqueous volume
fraction has less influence with the extraction efficiency staying at 94 % ± 3. It is observed
that the extraction efficiency has a maximum. This is because mass transfer is influenced by
internal circulation flow in both water and butyl acetate slugs. With the increase in aqueous
volume fraction, the length of water slugs increases therefore mass transfer in these slugs
decreases. However, the length of butyl acetate slugs decreases enabling enhanced mass
transfer. Due to these opposing effects a maximum is seen. This was also observed by
Okubo et al.57
4.4.3. Extraction efficiency with variation of flow rate and comparison of CFI
with straight tube
The extraction efficiency was studied as a function of the total flow rate with aqueous
volume fraction kept at 0.5 so equal flow rate of aqueous and organic streams were fed. In
order to evaluate the extraction efficiency achieved using CFI, its performance was
compared with a simple straight tube. Experiments were performed by replacing only the
CFI part with a straight tube of same diameter and length. The results are presented in
Figure 4.8 and 4.9 for toluene-water-acetone system and butyl acetate-water-acetone
system, respectively.
For the toluene-water-acetone system, study was done up to a maximum of 120 ml/min (7.2
l/h) which was the maximum limit of the pump. It can be seen from Figure 4.8 for both
systems extraction efficiency increases with increase in flow rate up to about 30 ml/min.
This is because the intensity of internal circulations within the slugs increases enabling
better mass transfer. The difference in extraction efficiency between CFI and straight tube
is about 8 % at this part. After that, for the straight tube the extraction efficiency starts to
drop. This is due to the decrease in the contact time. For straight tube this effect
overshadows the convection through internal circulations. This was also observed by
others.55,90 The difference in extraction efficiency between CFI and straight tube than
expands to about 20 %.
As mentioned earlier, secondary flow is generated due to centrifugal forces inside coiled
tubes. Dean number is a measure of the magnitude of the secondary flow. It represents the
balance between inertial, centrifugal and viscous forces. By rearrangement it is calculated
as:
(4.1)
where is the curvature ratio described previously and is 10 for this system. CFI exploits
this effective utilization of secondary forces to enhance mixing. It also employs multiple
flow inversions that create chaotic advection leading to higher radial mixing. The better
performance of the CFI results from this. It was observed that even at much higher flow
78
rates high extraction efficiency was maintained. This is because with the increase in flow
rate, Reynolds number increases and thus also Dean number increases intensifying
secondary flow (Table 4.3 and 4.4).
For the butyl acetate-water-acetone system shown in Figure 4.9, study was done up to 50
ml/min (3 l/h) which was the limit for the flow separator due to lower interfacial tension of
this system. Similar observation is made with the previous system.
Figure 4.8. Extraction efficiency as a function of the total flow rate comparing CFI with
straight tube for the toluene-water-acetone system with schematically drawn flow patterns
Figure 4.9. Extraction efficiency as a function of the total flow rate comparing CFI with
straight tube for the butyl acetate-water-acetone system with schematically drawn flow
patterns
79
For the straight tube systems, the extraction efficiency increases up to about 8 ml/min as
seen in Figure 4.9 where the difference with the CFI is on average about 10%. After that its
extraction efficiency drops due to decrease in contact time and the difference with CFI
increases to about 21% in average. For the CFI, high extraction performance was sustained
due to Dean vortices and especially with flow inversions that create chaotic advection.
It was observed for both systems that the extraction efficiency of CFI dropped slightly with
the increase in flow rate further signifying a small influence of the decrease in contact time.
From Figure 4.8 and 4.9 this drop is not very visible due to the scale of the graphs.
Therefore more close up figures are provided in Figures 4.10 and 4.11.
Toluene-water-acetone
system
Figure 4.10. Extraction efficiency as a function of the total flow rate for the toluene-water-
acetone system
Butyl acetate-water-acetone
system
Figure 4.11. Extraction efficiency as a function of the total flow rate for the butyl acetate-
water-acetone system
80
For the toluene-water-acetone system, the extraction efficiency increases steadily up to 96

% at 100 ml/min and then it slightly drops to 93 % at 120 ml/min which was the flow rate
limit of the pumps. For the butyl acetate-water-acetone system, the extraction efficiency
increases and reaches 100 % and then slightly drops and plateaus at about 96 %.
While carrying out experiments it was observed that although CFI operated in slug flow
regime, with the increase in flow rate parallel flow occurred in the straight tube. This
change in the flow regime is also shown schematically in Figures 4.8 and 4.9. This is due to
viscous and inertia forces having more influence with the increase in the flow rate. To see
this, dimensionless numbers are calculated for the two systems considering equilibrium
distribution of acetone for determining the physical properties of the aqueous and organic
phases. Reynolds ( ), Capillary ( ) and Weber ( ) numbers of the organic and
aqueous phases are given in Table 4.3 for toluene-water-acetone system and Table 4.4 for
butyl acetate-water-acetone system at the selected flow rates in the range studied for the
systems.
Also to show the difference between the CFI and straight tube, and the change with the
flow rate Dean ( ) number is calculated using Eq. 4.1 and presented in the Tables 4.3 and
4.4. For that, a two phase Reynolds ( ) is calculated which is also given in tables using
mixture properties for density and viscosity and considering total flow rate of the system
for velocity.
Capillary number shows the ratio of viscous force to interfacial tension force and it is
directly proportional to velocity. With the increase in the flow rate and therefore velocity,
viscous forces increases and produces parallel flow since they extend and drag the interface
downstream. Weber number gives the ratio of inertia force to interfacial tension and it is
proportional to square of velocity. Therefore, with the increase in flow rate it increases
rapidly and inertia forces become more prominent causing the change in the flow regime to
parallel flow in straight tube. This was also observed by others working with straight
capillaries.89,92–94
Table 4.3. Two-phase Reynolds and Dean numbers together with Reynolds, Capillary and
Weber number of the organic and aqueous phases for toluene-water-acetone system
,
ml/min
10 79 25 53 32 3.2 ×10-4 6.1 ×10-4 1.7 ×10-2 2.0 ×10-2
30 237 75 160 97 9.6 ×10-4 1.8 ×10-3 1.5 ×10-1 1.8 ×10-1
60 474 150 320 194 1.9 ×10-3 3.6 ×10-3 6.1 ×10-1 7.0 ×10-1
120 948 300 639 387 3.8 ×10-3 7.3 ×10-3 2.4 2.8
81
Table 4.4. Two-phase Reynolds and Dean numbers together with Reynolds, Capillary and
Weber number of the organic and aqueous phases for butyl acetate-water-acetone system
,
ml/min
10 73 23 45 32 8.1 ×10-4 1.3 ×10-3 3.6 ×10-2 4.1 ×10-2
30 220 70 134 95 2.4 ×10-3 3.9 ×10-3 3.2 ×10-1 3.7 ×10-1
50 366 116 224 158 4.0 ×10-3 6.5 ×10-3 0.9 1.0
For the toluene-water-acetone system, Weber numbers for both phases get above 1 around
80 ml/min. Indeed after around this flow rate we observed that the slug flow became
somewhat irregular showing still slugs but not of equal length as depicted in Figure 4.8.
However, overall the CFI system continued operation in slug flow regime in the same
conditions where the straight tube changed to parallel flow at higher flow rates. This
difference is most likely due to the fact that there are secondary flow effect and flow
inversions in CFI whereas in straight tube these are not present. In straight tube systems it
was seen that slug flow lends itself to stable parallel flow at higher flow rates 94. In CFI,
with the increase in flow rate Dean number increases (Tables 4.3 and 4.4) enabling increase
in angular velocity and intensification of vortices. We believe that centrifugal forces and
angular velocity present in CFI acts to overcome the downstream dragging effect of stream
wise viscous and inertia forces enabling to keep the slug structure.
4.4.4. Performance of flow separators

It is important to achieve separation of the phases immediately following extraction and
contacting of the phases. Gravity based separation was employed commonly to achieve this
separation. However, this separation is generally slow and adds to the contacting time and
affects the extraction performance. Therefore, it is necessary to adapt a separation method
that enables immediate separation of the phases. At micro- and milliscale surface tension
dominates over gravitational forces. Thus, capillary forces can be used effectively as the
driving force for separation of immiscible liquids. There are generally two methods used
for this: wettability based separation and membrane based separation. In this study, we use
example units of both types.
The wettability based separator of slit shaped flow separator developed by Gaakeer et al.60
has Teflon and glass bars with rectangular channels. The hydraulic pressure drop at each
side of the separator can be calculated using Hagen-Poiseuille equation for pressure drop:
12
∆ (4.2)
82
where is the viscosity, is the volumetric flow rate of the phase through that outlet, ,
and are length, width and height of the slit which are 5 mm, 10 mm and 0.4 mm
respectively. Viscosities of both phases are calculated based on the distribution at
thermodynamic equilibrium of acetone at equal flow rate of the organic and aqueous feeds.
To enable proper operation of the separator, capillary pressure ∆ must be higher than the
pressure difference across the interface. The ∆ for the slit shaped separator can be
predicted using the Young-Laplace equation:
2 2
∆ (4.3)
Where is the interfacial tension. Gaakeer et al.60 indicated that the capillary pressure must
be at least twice the hydraulic pressure in the opposite slit to prevent breakthrough.
Hydraulic pressure in the Teflon slit, hydraulic pressure in the glass slit and capillary
pressure with respect to flow rate is given in Figure 4.12 together with half of capillary
pressure as an indicative threshold for breakthrough for the two systems studied.
Figure 4.12. Hydraulic and capillary pressure with respect to flow rate for slit shaped
separator with slit height of 0.4 mm
When the hydraulic pressure of water in the glass slit exceeds the capillary pressure
threshold it is expected to observe water breakthrough at the Teflon slit. In a similar way,
when the hydraulic pressure of organic in the Teflon slit exceeds the capillary pressure
threshold it is expected to observe organic breakthrough at the glass slit. It can be expected
from Figure 4.12 that water breakthrough at the Teflon slit will occur at a lower flow rate
than organic breakthrough at the glass slit. Also, it can be seen that for the lower interfacial
tension system of butyl acetate-water-acetone the breakthrough at both sides occur at a
lower flow rate that the other system. For this reason good separation performance can be
maintained at much lower flow rates that the toluene-water-acetone system.
83
The membrane based separator was developed by IMM and it houses a PTFE membrane
with a pore size of 0.5 μm and dimensions of 44 mm x 14 mm. For membrane based
separators to function properly there are two conditions. First is to prevent the aqueous
phase to cross the membrane and exit with the permeate phase which is the organic phase.
For this, the pressure difference across the membrane should be less than the capillary
pressure, ∆ . This breakthrough pressure can be estimated with Young-Laplace equation
where membrane pores are taken to be an array of cylinders:
2
∆ (4.4)
The second condition is that the pressure drop at the retentate (aqueous) side of the
membrane should be above the sum of pressure drop at the permeate (organic) side of the
membrane and the pressure difference across membrane for the desired flow rate (∆ )
which can be approximated using Hagen-Poiseuille equation as:
8
∆ (4.5)
where is the viscosity of organic phase, is the membrane thickness, is the flow rate of
organic phase, is the number of pores and is the pore size. This is the pressure
difference required to enable all the organic phase to flow through the membrane.
4.4.4.1. Toluene-water-acetone system
From Figure 4.12, it can be expected that, for the toluene-water-acetone system, the
hydraulic pressure of water in the glass slit will exceed the capillary pressure threshold for a
flow rate of about 26 ml/min. After this flow rate it is expected to observe water
breakthrough at the Teflon slit. The same can be expected for the toluene to breakthrough at
the glass slit above about 50 ml/min.
As can be seen from Figure 4.13 left side, the breakthrough of water at the Teflon outlet is
observed from about 25 ml/min as expected. For the limit of this study of 120 ml/min it
reaches about 20 % breakthrough. The breakthrough of toluene at the glass slit is observed
from 32 ml/min but stays below 5 % up to 50 ml/min. Above 50 ml/min it increases up to
20 % at 120 ml/min.
When the membrane separator was used (Figure 4.13, right), it was observed that some of
the organic phase was retained in the aqueous phase. This resulted in about 30 % of organic
phase to be lost at the aqueous side. This can be because some of the organic phase is
entrained in the aqueous phase. It can also be because no differential pressure was applied
there was not enough pressure difference to force all the organic to pass through the
membrane. The operation was limited to flow rate of 20 ml/min because pumps failed to
pump the liquids forward above this flow rate. This could also be because no pressure
difference mechanism was applied. Cervera-Padrell et al.41 employed back pressure
84
regulator in the aqueous outlet to generate high fluidic resistance but below the capillary
pressure to achieve perfect separation. Castell et al.59 created pressure difference between
the two outlets by connecting the aqueous outlet to a sealed vessel where its outlet pressure
is modulated. In this study, initial attempts were made to improve the operation by
generating pressure difference by changing the length and diameter of the outlet tubes at
each side and by using back pressure regulators. But the presence of many factors studied
in this chapter: different systems with different interfacial tension, different flow rates and
different aqueous volume fractions, it was found cumbersome to employ a different element
to create the required pressure difference for each condition. Development of a separator
with an integrated pressure controller like the device of Adamo et al.58 that can handle the
high flow rate required in this study could be the solution to this problem. However, this
was found outside the scope for this study. Nevertheless, with the membrane separator at
permeate side always pure organic phase was attained (Figure 4.13, right bottom) meaning
the PTFE membrane did its job successfully to only let the organic phase to pass through.
This shows that capillary pressure is not exceeded since breakthrough of the aqueous phase
is not taking place at the flow range studied. If it is required to have pure organic phase,
than membrane has similar performance with the slit separator which has pure organic
phase until 26 ml/min but there is significant loss of some organic phase at the aqueous side
which could be overcome using an integrated pressure control system.
Figure 4.13. Separator performance as a function of total flow rate for the toluene-water-
acetone system – volume fraction of organic phase and aqueous phase in two outlets, left
slit shaped separator: top glass outlet and bottom Teflon outlet, right membrane separator:
top retentate outlet and bottom permeate outlet
85
The variation of aqueous volume fraction on the separation performance was also
investigated. The result for the toluene-water-acetone system is given in Figure 4.14. When
the slit shaped separator was used (Fig. 4.14, left), by increasing the aqueous flow rate
hydraulic pressure in the glass slit increases. Accordingly at aqueous volume fractions
above 0.6, water breakthrough in the Teflon slit was observed. This is also due to the fact
that while the aqueous flow rate increases, organic flow rate decreases. When there is
significant difference in the flow rates the inertial flow forces overcome the wetting forces
and the high flow rate liquid of aqueous exits from the low flow rate liquid of organic’s
side. This was also observed by Kashid et al.55 When membrane separator was used (Fig.
4.14, right), no breakthrough of the aqueous phase through the membrane occurred.
However, at low aqueous volume fraction when the organic flow rate is much higher than
the aqueous flow rate, it is observed that there is high amount of organic at the retentate
side. This is because of the low amount of fluidic resistance at the aqueous side at the low
flow rate; the organic phase with high flow rate can also exit from this side. With the
increase in the aqueous volume fraction this is seen to decrease. If pure organic phase is
required at high aqueous volume fractions membrane separator could be preferred.
Figure 4.14. Separator performance as a function of aqueous volume fraction for the
toluene-water-acetone system – volume fraction of organic phase and aqueous phase in two
outlets, left slit shaped separator: top glass outlet and bottom Teflon outlet at total flow rate:
20 ml/min, right membrane separator: top retentate outlet and bottom permeate outlet at
total flow rate: 10 ml/min.
86
4.4.4.2. Butyl acetate-water-acetone system

The interfacial tension has a major influence on the operation of separators using capillary
forces. As can be seen from Figure 4.12, the operating window of the separator reduced
significantly with the lowering of the interfacial tension from the toluene-water-acetone
system to the butyl acetate-water-acetone system. It was also observed by Kralj et al.54 that
by lowering the interfacial tension the maximum operating flow rate decreases
significantly.
From Figure 4.12, it is seen that the hydraulic pressure of water in the glass slit will exceed
the capillary pressure threshold for a flow rate of about 12 ml/min. After this flow rate it is
expected to observe water breakthrough at the Teflon slit. The same can be expected for the
toluene to breakthrough at the glass slit above about 20 ml/min. From experiments it was
observed that (Figure 4.15, left) at both outlets breakthrough occurred above about 16
ml/min and increased rapidly up to 40 % at 50 ml/min. The experiments were stopped at
this flow rate due to high breakthrough. Breakthrough of 20 % occurred at about 32 ml/min
which is much lower than 120 ml/min attained for the toluene-water-acetone system
indicating the influence of the interfacial tension.
When the membrane separator was used (Figure 4.15, right), similar observation with the
toluene-water-acetone was made. It was seen that some of the organic phase was retained in
the aqueous phase resulting about 40 % of organic to be lost at the aqueous side. At the
permeate side pure organic phase was attained throughout. If it is required to have pure
organic phase, than membrane has slightly better performance than the slit separator which
has pure organic phase until 16 ml/min but there is significant loss of some organic phase at
the aqueous side which could be overcome using an integrated pressure control system.
The result for the butyl acetate-water-acetone system with variation of aqueous volume
fraction with the total flow rate fixed at 10 ml/min is given in Figure 4.16. When the slit
shaped separator was used (Fig. 4.16, left), similar observation with the toluene-water-
acetone system is made but here the effect is seen to be larger. This is because the capillary
pressure is lower for this system due to lower interfacial tension system. At aqueous
volume fractions above 0.5, water breakthrough in the Teflon slit was observed and it
increased up to 20% where maximum 10% breakthrough was observed in the first system.
When membrane separator was used (Fig. 4.16, right), again no breakthrough of the
aqueous phase through the membrane occurred and at retentate side high amount of organic
left with the aqueous phase at low aqueous volume fractions as also observed for the
toluene-water-acetone system. With the increase in aqueous volume fraction although the
percent loss of organic at the aqueous side is almost the same, the amount that is lost
decreases with the decrease in the total organic feed. Since also the aqueous amount
increases this results in the decrease in the organic percentage at the retentate side seen in
Fig. 4.16, right.
87
Figure 4.15. Separator performance as a function of total flow rate for the butyl acetate-
water-acetone system – volume fraction of organic phase and aqueous phase in two outlets,
left slit shaped separator: top glass outlet and bottom Teflon outlet, right membrane
separator: top retentate outlet and bottom permeate outlet
Figure 4.16. Separator performance as a function of aqueous volume fraction for the butyl
acetate-water-acetone system – volume fraction of organic phase and aqueous phase in two
outlets, left slit shaped separator: top glass outlet and bottom Teflon outlet, right membrane
separator: top retentate outlet and bottom permeate outlet, total flow rate: 10 ml/min
88
4.5. Conclusion
4.5. Conclusion
In this chapter a liquid-liquid extraction setup developed using milli-scale CFI (3.2 mm ID,
210 cm length) and phase separator was described. The extraction performance of the setup
was tested for the two widely used and EFCE recommended model systems of toluene-
water-acetone and n-butyl acetate-water-acetone. We report for the first, to best of our
knowledge, the application of CFI for immiscible liquid-liquid mixing. Also, the study of
flow separators that can handle flow rates suitable for industrial scale continuous
pharmaceutical production is very rare and this was successfully achieved here. For the
toluene-water-acetone a flow rate up to 120 ml/min was studied which corresponds to 7.2
l/h and 58 m3/y. It is important to point out that the slug flow regime that is limited to
microfluidic systems is maintained even at such pilot scale. The literature review on slug
flow reveals that this is the highest flow rate achieved so far. There are several
microreactors developed that enables contacting of liquid phases at high flow rate.
However, there is very limited number of examples that achieve the phase separation part
of liquid-liquid extraction at these flow rates. Most of these examples involve the use of
settlers where phase separation is achieved using macro force of gravity and it is slow. In
this study, after contacting of the phases the phase separation is achieved immediately using
micro effect of capillary force at high throughput. Two separator concepts utilizing
capillary forces were investigated that are PTFE membrane separator and slit shaped
separator composed of glass and Teflon outlets enabling separation based on wettability.
Flow pattern of two-phase toluene-water and butyl acetate-water systems characterized
with high-speed photographic system revealed the dependence of the slug length depends
on the ratio of organic and aqueous flow rates. The dimensionless numbers calculated in the
frame of 10 ml/min organic flow rate by 10 ml/min aqueous flow rate confirm the presence
of slug flow regime with the interfacial tension dominating over inertia and viscous forces.
Extraction performances with variation of aqueous volume fraction at constant total flow
rate and with variation of total flow rate at equal volume fraction were investigated. For
both systems close to thermodynamic extraction efficiency was achieved (96% for first,
100% for second). The extraction efficiency of the setup was compared with a straight tube
setup. Straight tube efficiency was seen to drop at higher flow rates due to mass transfer
limitation with the reduced contact time. For CFI it was observed that even at much higher
flow rates high extraction efficiency was maintained due to intensified secondary flow and
flow inversions achieved. The difference in extraction efficiency between CFI and straight
tube was seen to be about 20 % at higher flow rates. It was observed that although CFI
operated in slug flow regime throughout, parallel flow occurred in the straight tube with the
increase in flow rate due to viscous and inertia forces becoming more prominent. It is
believed that the centrifugal forces and angular velocity present in CFI acts to overcome
these stream wise forces.
89
Lastly, the performances of the two separators were evaluated. Slit shaped flow separator
performed well with the breakthrough occurring at flow rate expected where hydraulic
pressure exceeds the capillary pressure threshold. Breakthrough of 20 % occurred at about
120 ml/min for the toluene-water-acetone system whereas it was observed at 32 ml/min for
the butyl acetate-water-acetone system indicating the influence of the interfacial tension on
the operating window of the separators. With the variation of the aqueous volume fraction it
was observed that at high aqueous volume fraction water breakthrough in the Teflon outlet
occurs due to increase in hydraulic pressure of the water in the glass slit. The membrane
separator flow rate was limited to 20 ml/min. It was seen that some of the organic phase
was retained in the aqueous phase while at the permeate side pure organic phase was
attained throughout. If pure organic phase is required, then the two separators performed
similarly with the slit shaped separator enabling pure organic phase until 26 ml/min for
toluene-water-acetone system and 16 ml/min for butyl acetate-water-acetone system.
However, there is significant loss of organic phase at the aqueous side which could be
overcome by a separator with an integrated pressure controller. With the variation of the
aqueous volume fraction it was observed that no breakthrough of the aqueous phase
through the membrane occurred but at retentate side at low aqueous volume fractions high
amount of organic left at the retentate side due to low amount of fluidic resistance at this
side with the low flow rate of aqueous phase. With increase in the aqueous volume fraction
this was seen to be reduced due to decrease in the flow rate of the organic phase.
Acknowledgements
I like to thank Dr. Patrick Loeb from Fraunhofer-ICT Mainz (formerly Institut fuer
Mikrotechnik Mainz GmbH) for providing the PTFE membrane separator used in the
studies and Wim Gaakeer and his supervisors Dr.ir. John van der Schaaf and Prof.dr.ir. Jaap
Schouten for the slit shaped flow separator.
90
4.6. Appendix
4.6. Appendix
4.6.1. General chemicals information
Chemicals were purchased in their highest purity available from Sigma Aldrich. Methylene
blue stock solution was made by dissolving 0.498 gram methylene blue in 100 ml
demineralized water (purified with a Millipore Elix UV-5 machine) resulting in 0.016 M
solution. Acetone solutions were made by dissolving 24.32 gram acetone in a 250 ml
volumetric flask with demineralized water.
4.6.2. General analytical information

Quantitative analysis of the acetone transferred to organic phase was done using GC-FID
on a Varian 430-GC with a flame ionization detector using a CP-Sil 8 CB column (length
60 m, diameter 0.25 mm, film thickness 0.1 μm), with helium as the carrier gas. Sample
from organic outlet was taken and analyzed for amount of solute. External calibration was
performed by making solutions of varying concentrations 0, 2, 4, 6, 8, and 10 wt. % of
acetone in butyl acetate. For toluene system calibration was performed with concentrations
of 0, 1, 3.5, 5.5, 6.5, 10 and 13.7 wt. % of acetone. (Appendix 4.6.7)
4.6.3. Investigation of liquid-liquid two-phase flow patterns

To study the flow patterns observed high speed imaging system was used. For this
visualization experiments acetone was not used. The aqueous phase of water was colored
with the methylene blue solution by adding 0.5 ml of the prepared solution to 250 ml of
demineralized water. Syringes were filled with the pure liquid phases and placed in separate
syringe pumps. The aqueous and organic phases were fed at both equal and unequal flow
rate combination in the range of 1-10 ml/min of each phase. As soon as uniform slug flow
regime was achieved, flow was recorded by MotionPro X3 high speed camera. The camera
was placed at a set distance from the setup to take picture of a straight section of tubing. It
was adjusted to capture approximately 3 slugs in an exposure. Two lamps were placed as
light source. Several snapshots were taken for each set of flow conditions. Image analysis
was done with MotionPro software. Pictures are presented to give an indication of the size
of the slugs at that condition. The length of the tube diameter is placed on the pictures as a
measure bar to get an estimation of the size and to be able to make comparison of slug
lengths with other flow conditions. The absolute lengths of slugs were not calculated.
4.6.4. Determination of extraction efficiency

Experimental procedure
Two 20 ml syringes were willed with 10 wt. % acetone solution and mounted on one
syringe pump. One 50 ml syringe was filled with organic feed (either toluene or butyl-
acetate) and mounted on to another syringe pump. Experiments were conducted both equal
91
and unequal flow ratios of the phases. To study the effect of flow rate, the flow ratio was
kept equal to one and flow rate was increased equally. To study the effect of aqueous
volume fraction, aqueous and organic phases were fed at different flow rate combinations
with the total flow rate kept equal to 20 ml/min for toluene system and at 10 ml/min for the
butyl acetate system. The aqueous volume fraction is defined as the aqueous flow rate
divided by the total flow rate. After all the system was filled with uniform slugs, one more
residence time was waited before a sample was taken. At least three samples from organic
outlet were taken from each set of experiments and analyzed using GC-FID to determine
amount of acetone. The mean value was recorded.
Calculation of extraction efficiency
Extraction efficiency is the ratio of the amount of solute transferred to the maximum
amount transferable defined by the thermodynamic equilibrium. To calculate extraction
efficiency ( ) Eq. 4.6 was used:
, ,
100 (4.6)
, ,
where , , , , and , are the inlet, outlet and equilibrium concentration of

acetone in the organic phase, respectively. The inlet concentration of acetone in the organic
phase is zero; all acetone enters with the aqueous phase. The outlet concentration of acetone
in the organic phase is determined using GC-FID and the calibration curves as described in
the Appendix. The equilibrium concentration was determined with shake flask experiments
at different aqueous volume fractions and straight line fitting was done (Appendix 4.6.8).
At equal flow rate of the phases (aqueous volume fraction of 0.5) and acetone concentration
of 10 wt.% the thermodynamic partition coefficient is given in literature as 0.88 for toluene
system and 0.98 for butyl acetate system.63 The experimentally determined thermodynamic
partition coefficients are 0.87 for toluene system and 0.98 for butyl acetate system. This
indicates the validity of the experimental method used.
4.6.5. Comparison with straight tube

For comparison of the performance of the CFI with the straight tube, experiments were
performed replacing the CFI with a tube of same diameter and length. Due to the long
length of the tube required, to be able to fit in the fume hood, the tube was placed as a one
large circle. It is assumed that there will be negligible secondary flow effects with such a
large coil. The same experimental procedure as CFI setup was followed for determination
of extraction efficiency.
4.6.6. Measurement of separator performance

After all the system was filled with uniform slugs and two residence times has passed over
the separator, the collector bottles were removed and measuring cylinders of 50 ml were
92
4.6. Appendix
placed at the outlets. This way startup effects were excluded. Visual inspection of the
measuring cylinders was done. Since the aqueous phase was colored it was easy to evaluate
the separation performance and measure the amount of one phase leaving at the other
phase. In case of incomplete separation, volume fraction of the two phases at the two
outlets was determined by volumetric measurement.
4.6.7. Calibration curves for GC-FID analysis

To determine the calibration curve for the two systems, known amounts of acetone were
dissolved in the organic phase and put into the GC-FID. The area of the acetone peak was
then divided by the area of the organic phase peak, this is plotted on the x-axis in Figure
4.17. The acetone concentration is plotted on the y-axis. Appropriate trendlines were fitted
to these points using Excel. The goodness of fit of both curves are shown with R2 and are
close or equal to one. This shows that the calibration curves are appropriate to determine
the acetone concentration of organic samples from experiments.
Figure 4.17. Calibration curves for GC-FID
4.6.8. Thermodynamic equilibrium concentration curves

To determine the thermodynamic equilibrium concentration experimentally shake flask
method was used. Aqueous phase of 10 wt. % acetone in water was combined with the
organic phase at different volumetric fractions of the aqueous phase. After equilibrium was
reached sample was taken from the organic phase and analyzed with GC-FID. With the
calibration curve explained in the previous section the equilibrium concentration of acetone
in the organic phase is determined. This is plotted with respect to the aqueous volume
fractions they correspond to. Straight line fitting was done to these points using Excel. The
goodness of fit of both curves are shown with R2 and are close to one. This shows that these
93
curves are appropriate to determine the equilibrium concentration of acetone in organic at

different aqueous volume fractions studied.
Figure 4.18. Equilibrium concentration curves
References
2013, 6, 746–89.
3. T. Noël and V. Hessel, in New Trends in Cross-Coupling: Theory and Applications,
ed. T. Colacot, Royal Society of Chemistry, London, 2014, pp. 610–644.
2011, 167, 504–509.
467–478.
2011, 66, 1426–1448.
Chemistry, 2009, 2009, 1321–1325.
Development, 2010, 14, 215–224.
1290.
94
References

6, 2220–2225.
ChemSusChem, 2015, 8, 504–12.
16. S. C. Stouten, T. Noël, Q. Wang, and V. Hessel, Chemical Engineering and
17. Y. Su, N. J. W. Straathof, V. Hessel, and T. Noël, Chem. Eur. J., 2014, 20, 10562–
89.
Transactions, 2012, 29, 565–570.
& Technology, 2012, 35, 1184–1204.
Technik, 2012, 84, 660–684.
21. I. Vural Gürsel, Q. Wang, T. Noël, V. Hessel, and J. T. Tinge, Industrial
Transactions, 2013, 35, 775–780.
23. Q. Wang, I. Vural Gürsel, M. Shang, and V. Hessel, Chemical Engineering
Journal, 2013, 234, 300–311.
26. C. Wiles and P. Watts, Chemical communications, 2011, 47, 6512–35.
28. V. Hessel, H. Löwe, and F. Schönfeld, Chemical Engineering Science, 2005, 60,
2479–2501.
29. K. F. Jensen, Chemical Engineering Science, 2001, 56, 293–303.
30. S. Braune, P. Pöchlauer, R. Reintjens, S. Steinhofer, M. Winter, O. Lobet, R.
Guidat, P. Woehl, and C. Guermeur, Chimica Oggi, 2009, 27, 26–29.
31. X. Ye, M. D. Johnson, T. Diao, M. H. Yates, and S. S. Stahl, Green Chemistry,
2010, 12, 1180–1186.
32. D. M. Roberge, N. Bieler, M. Mathier, M. Eyholzer, B. Zimmermann, P. Barthe, C.
Guermeur, O. Lobet, M. Moreno, and P. Woehl, Chemical Engineering and
Technology, 2008, 31, 1155–1161.
33. N. Zaborenko, M. W. Bedore, T. F. Jamison, and K. F. Jensen, Organic Process
34. P. Styring and A. I. R. Parracho, Beilstein Journal of Organic Chemistry, 2009, 5.
2011, 167, 718–726.
95

in English), 2007, 46, 5704–5708.
39. I. Vural Gürsel, F. Aldiansyah, Q. Wang, T. Noël, and V. Hessel, Chemical
Engineering Journal, 2015.
41. A. E. Cervera-Padrell, S. T. Morthensen, D. J. Lewandowski, T. Skovby, S. Kiil,
and K. V. Gernaey, Organic Process Research and Development, 2012, 16, 888–
900.
Technology, 2013, 36, 921–936.
44. A. Sundberg, P. Uusi-Kyyny, and V. Alopaeus, Chemical Engineering Research
and Design, 2009, 87, 705–710.
45. A. Ziogas, V. Cominos, G. Kolb, H.-J. Kost, B. Werner, and V. Hessel, Chemical
46. R. L. Hartman, H. R. Sahoo, B. C. Yen, and K. F. Jensen, Lab on a Chip, 2009, 9,
1843–1849.
47. K. F. Lam, E. Cao, E. Sorensen, and A. Gavriilidis, Lab on a chip, 2011, 11, 1311–
1317.
48. H. Niu, L. Pan, H. Su, and S. Wang, Industrial and Engineering Chemistry
Research, 2009, 48, 8629–8634.
49. C. Ye, G. Chen, and Q. Yuan, Chinese Journal of Chemical Engineering, 2012, 20,
111–119.
50. N. N. Gao, J. X. Wang, L. Shao, and J. F. Chen, Industrial and Engineering
51. N. T. Tran, I. Ayed, A. Pallandre, and M. Taverna, Electrophoresis, 2010, 31, 147–
73.
52. C. T. Culbertson, S. C. Jacobson, and J. M. Ramsey, Analytical Chemistry, 2000,
72, 5814–5819.
53. A. B. Jemere, D. Martinez, M. Finot, and D. J. Harrison, Electrophoresis, 2009, 30,
4237–4244.
Research, 2007, 46, 8420–8430.
56. D. M. Fries, T. Voitl, and P. R. von Rohr, Chemical Engineering and Technology,
2008, 31, 1182–1187.
96
References
57. Y. Okubo, T. Maki, N. Aoki, T. Hong Khoo, Y. Ohmukai, and K. Mae, Chemical
Engineering Science, 2008, 63, 4070–4077.
59. O. K. Castell, C. J. Allender, and D. A. Barrow, Lab on a chip, 2009, 9, 388–96.
61. A. Ghaini, M. N. Kashid, and D. W. Agar, Chemical Engineering and Processing:
62. A. K. Saxena and K. D. P. Nigam, AIChE Journal, 1984, 30, 363–368.
63. T. Misek, R. Berger, and J. Schroter, Standard Test Systems for Liquid Extraction,
EFCE Publication Series No. 46, Rugby, England, 1985.
64. F. Schönfeld and S. Hardt, AIChE Journal, 2004, 50, 771–778.
65. V. Kumar, M. Aggarwal, and K. D. P. Nigam, Chemical Engineering Science,
2006, 61, 5742–5753.
66. L. A. M. Janssen, Chemical Engineering Science, 1976, 31, 215–218.
67. J. H. van den Berg and R. S. Deelder, Chemical Engineering Science, 1979, 34,
1345–1347.
68. J. A. Koutsky and R. J. Adler, The Canadian Journal of Chemical Engineering,
1964, 42, 239–246.
69. A. P. Sudarsan and V. M. Ugaz, Lab on a chip, 2006, 6, 74–82.
70. W. R. Dean, Proceedings of the Royal Society A: Mathematical, Physical and
Engineering Sciences, 1928, 121, 402–420.
71. F. Jiang, K. S. Drese, S. Hardt, M. Küpper, and F. Schönfeld, AIChE Journal, 2004,
50, 2297–2305.
72. E. B. Nauman, Chemical Engineering Science, 1977, 32, 287–293.
73. S. Klutz, S. K. Kurt, M. Lobedann, and N. Kockmann, Chemical Engineering
Research and Design, 2015, 95, 22–33.
74. S. K. Kurt, M. G. Gelhausen, and N. Kockmann, Chemical Engineering &
Technology, 2015, 38, 1122–1130.
75. S. K. Kurt, K. D. P. Nigam, and N. Kockmann, Proceedings of the ASME-ICNMM
2015, 2015.
76. J. Singh, N. Kockmann, and K. D. P. Nigam, Chemical Engineering and
77. V. Kumar, M. Mridha, A. K. Gupta, and K. D. P. Nigam, Chemical Engineering
Science, 2007, 62, 2386–2396.
78. M. M. Mandal, V. Kumar, and K. D. P. Nigam, Chemical Engineering Science,
2010, 65, 999–1007.
79. V. Kumar and K. D. P. Nigam, International Journal of Heat and Mass Transfer,
2005, 48, 4811–4828.
80. J. Singh, N. Choudhary, and K. D. P. Nigam, The Canadian Journal of Chemical
Engineering, 2014, 92, 2185–2201.
97
81. M. M. Mandal, P. Aggarwal, and K. D. P. Nigam, Industrial & Engineering

82. M. Mridha and K. D. P. Nigam, Chemical Engineering Science, 2008, 63, 1724–
1732.
83. M. M. Mandal, C. Serra, Y. Hoarau, and K. D. P. Nigam, Microfluidics and
Nanofluidics, 2010, 10, 415–423.
84. J. Singh, V. Verma, and K. D. P. Nigam, Industrial & Engineering Chemistry
Research, 2013, 52, 207–221.
85. S. Vashisth and K. D. P. Nigam, Industrial & Engineering Chemistry Research,
2008, 47, 3630–3638.
86. S. Vashisth and K. D. P. Nigam, Industrial & Engineering Chemistry Research,
2007, 46, 5043–5050.
87. S. Vashisth and K. D. P. Nigam, Chemical Engineering and Processing: Process
Chemical Engineering Journal, 2012, 207-208, 440–444.
89. M. N. Kashid, A. Renken, and L. Kiwi-Minsker, Industrial & Engineering
Journal, 2010, 158, 233–240.
92. J. Jovanović, E. V. Rebrov, T. A. Nijhuis, M. T. Kreutzer, V. Hessel, and J. C.
Schouten, Industrial and Engineering Chemistry Research, 2012, 51, 1015–1026.
Science, 2008, 63, 4035–4044.
94. Y. Zhao, G. Chen, and Q. Yuan, AIChE Journal, 2006, 52, 4052–4060.
95. S. K. R. Cherlo, S. Kariveti, and S. Pushpavanam, Industrial & Engineering
98
CHAPTER 5
Compact Modular Plants for Fine
Chemistry – Modular Process Integration
at Pilot Scale
Vural Gursel, I., Hessel, V., Wang, Q., Noël, T. & Lang, J. (2012). Window of opportunity :
potential of increase in profitability using modular compact plants and micro-reactor based
flow processing. Green Processing and Synthesis, 1(4), 315-336.
Abstract:
Recently much focus has been given to a new type of chemical production plant, with the
aim of a much faster time-to-market (“50% idea”) and better cash-flow revenue. The main
enabling technology is to have the plants pre-manufactured and assembled by a modular
construction and to use innovative, smart-scale processing and apparatus technology, in
order to achieve a compact overall plant footprint. Focal points in such technology are on
the one hand, flow processing, with micro process technology as a cutting-edge
cornerstone, and on the other hand, the container framework. Yet, other process-intensified
technologies are suitable as well such as milli-flow or upgraded batch technologies. Finally,
process robustness and short-time applicability make the decision. In this chapter, a
CAPEX and OPEX analysis of the new plant technology is done, at the example of the
Evotrainer production platform (meanwhile renamed as the EcoTrainer). This platform is
pre-manufactured in serial and provides all the utilities needed around the reactor and e.g.
separator to be tailored and inserted. Since the fine-chemical application comprises the
envisaged target market of the Evotrainer infrastructure and also since for this application
all relevant data were at hand, the cost analysis investigations started with and were
99
5. Compact Modular Plants for Fine Chemistry
centered around this chemical process. The same procedure was applied as model-based
scenarios for market applications in bulk chemistry and pharmacy for comparison. It is
shown, in many facets, that the Evotrainer infrastructure based plants indeed have a faster
payback and higher earning as compared to conventional technology; particularly when
serving high-priced markets. Further, the combination with novel chemical routes or novel
processing (Novel Process Windows) is advantageous. Micro process technology is one of
the key enablers and was taken here, since the dataset of such technology was readily
available to due to past research efforts and there is some general belief in the combination
to the so-called “Future Factories”. Yet, it stands also for any other process intensification
technology which can achieve the same performance level and which is able to satisfy the
needs of a producing industry.
5.1. Introduction
5.1.1. Need for faster market and product launch – “window of opportunity”
Europe’s chemical industry is under pressure in view of new emerging markets and
production capabilities in Asia and the Near East.1 Lower costs of production, lower taxes,
different approval, environmental and safety procedures, and increasingly skilled personnel
in Asia allow the chemical industry to grow at a high rate and to increase the global high
market share.2 To improve competitiveness of the chemical industry in Europe and
worldwide, new innovative production and plant technologies are needed. Besides, markets
are continuously diversifying, as products are increasingly designed for special customer
needs.3 New products have to be developed in even shorter time periods and thus the time
to market has to be shortened correspondingly.4 The current plant concepts used in the
chemical industry cannot necessarily satisfy all of these market conditions to the full extent
anymore.
To reduce risk over the whole time line from the product idea to the product launch requires
a reduction in the capital investment.5 A critical issue is the numerical ramp up scenario of
the product quantity sold. Firstly, intelligent handmade experiments in the lab help to
reduce the technical risk of product development and an early-bird provision of test samples
helps to make a timely contract with the future customer. Especially in volatile markets, it
is highly important to enter at the right time in the “window of opportunity” which is
facilitated when engineering is done at an early stage. A window of opportunity denotes a
short period of time during which an opportunity must be acted on or missed. This
established term in medicine and social sciences, is increasingly used also in the economy
for volatile markets, to describe the right time to approach a customer. It may be too early,
e.g., if the manufacturing of the product with the targeted high quality and functionality
cannot be proven (documented probes on a pilot/production level) or if adherence to a
100
5.1. Introduction
production schedule is not secured. It may be too late when the competitor is already on the
market, the client is in contracts with other parties, and product prices decline.6,7
5.1.2. Current chemical plant concepts

Bulk chemical industry builds high capacity plants as a consequence of economies of scale.
They are optimized for the efficient production of a single product.8 The plant is built from
individually designed equipment in a customized way. Because of their size and the degree
of optimization required, design and engineering takes a long time and costs are higher.9
Also, procurement of the equipment and construction of the plant is lengthy and
expensive.10 These all lead to a long time-to-market. Fine chemical production is carried
out batch-wise in multi-purpose plants and is flexible regarding production rate and
different products produced.8 Yet, it is nonetheless inefficient in terms of energy and raw
material consumption, as the equipment is not optimized for the requirements of the
products and recycling and heat recovery is hard to implement.11
When the demand for customized, high-value products in low annual capacities increases
and a much shorter market entry is required, these existing plant concepts are not
necessarily optimal anymore.
5.1.3. Modular plants, compact and pre-manufactured

Modular compact plants, being partly or fully preconfigured in their modules, parts and
interfaces, constitute an innovative approach to address this situation.12 They are seen as
main enablers to cut in half the time it takes to get through the development process from
the product idea to the final production process. This was discussed at the 48th Tutzing-
Symposium “Die 50%-Idee – Vom Produkt zur Produktionsanlage in der halben Zeit”
(translated: The 50% idea – from product to production facility in half the time).4,13 With
having entirely new concepts and apparatus at hand (process intensification), investigations
of the advantages of modular design and its economic potential, have gained new
momentum and intensified recently. Bramsiepe and Schembecker12 pointed out the
advantage of modular process design in terms of planning process that enables laboratory to
production with small effort. Hady and Wozny14 presented a modular concept with the aim
of showing engineering and equipment reuse, creating saving in terms of knowhow and
cost. Brodhagen et al.15 showed that with a shorter time-to-market achieved, modularization
is more economical than conventional batch technology. Rottke et al.16 described an
approach to configure plant design using modules, making it easier to determine the cost of
different layouts.
Standardization through pre-manufactured modules assembled into highly functional plant
environment facilitates a desired lead time reduction to respond to market changes in an
efficient manner. It was pointed out by Bramsiepe et al.9 that with a modular plant concept,
elimination of specialized field installation and customized design is possible by
101
standardization, leading to savings in construction and design time and expenditure.

Through the reduction of the planning and construction period, the Net Present Value
(NPV) of a modular plant was 30 % higher compared to multiproduct batch plant.11 For the
calculation of the capital investment for a modular plant, Shah17 suggested that the factors
which are used to estimate based on total equipment cost should be reduced, but the degree
of reduction was not specified.
Evonik Industries have developed a modular, mobile, compact platform that supports an
individual chemical process with a highly functionalized infrastructure – the Evotrainer
(meanwhile renamed as the EcoTrainer, see Figure 5.1).18,19 It triggers a far fetching
modular plant approach, with a holistic view comprising far beyond the reaction parts. The
concept orients particularly on the manufacture of specialty products with tonnage, which
are at present typically made as campaign products in conventional multi-product plants.20
It can be upgraded to a production plant of a capacity of some 100 tons per year. It takes
into account the whole process development to make easier the conversion from lab to
production by including raw material and product logistics, separation parts, utilities,
process control and environmental and safety measures.21
Figure 5.1. The EcoTrainer of Evonik Industries19

Standardized infrastructure, utility and process control supports the standardization of
modules.9 The compatibility of the modules offers the opportunity to rearrange the same
design to meet the needs of different processes. Design and engineering time is reduced
significantly through pre-manufactured modules. Standardization reduces the need for
customized equipment, piping and structural design.17 Thus, efforts for detailed equipment
102
5.1. Introduction
design will be reduced, as well as detailed piping and instrumentation design. The
engineering job can be focused to find the optimal configuration of modules for the specific
process.11 Since the engineering task is significantly easier, there is faster transfer from lab
to production.
Modular plants are efficient in terms of construction and mounting. Construction time can
be significantly reduced by the opportunity of preassembling modules in a workshop where
tools and machinery to build will be already available. With the use of modular components
it is possible to minimize additional field installation. The different modules will only be
plugged together on the production site. Accordingly, the costs for scaffolding, construction
tools and rentals and personnel located at the construction site can be minimized. A further
advantage of the concept is that construction will not be affected by the weather conditions.
Expensive site inspections can be omitted, as testing can be partly done on the
preassembled plant in the workshop.9 The sum of these benefits helps to reduce the
investment cost and time-to-market.
Additional benefits can be gained in terms of operating costs as well. Due to modularity,
only the interaction between the modules needs to be managed, which lowers the labor
requirement. Accordingly, the cost of operating labor and its supervision will be reduced.
Also, owing to standardization, the maintenance of modules will be easier and no
requirement for special parts makes it less costly.
5.1.3.1. Project workflow compression by modular plants
To be able to bring a new product to market faster than other products produces an
opportunity to seize high margins in the early phase of a product launch. Because of the
capacity of money to earn further money, the financial gain earned earlier is worth more
than the same amount earned later.22
In Figure 5.2 the difference between classical project workflow is shown in comparison to
the proposed workflow based on modular plants. In a classical project work flow after
initial selection of the process to be used, the process flow diagram is prepared. The
flowsheet shows the arrangement of the equipment selected, operating conditions and
stream flow rates and compositions.23 The flowsheet is formed by the material and energy
balance calculations which are carried out using process simulation. It is used as a basis for
the design of instrumentation and control, equipment and piping and structure. Piping and
instrument diagrams are made and equipment specification sheets are prepared via detailed
design. Using these sheets, the procurement of equipment and materials is done.
The field construction and equipment installation is done and after field inspection and
testing, the plant is started up. The workflow based on modular components deviates
considerably from classical project workflow. Some major steps are eliminated or
extremely simplified. From the process selection, different configurations of the modules
103
are considered and a modular assembly plan is made.17 With the use of process simulation
software, different modular processes are compared and optimal configuration is
identified.11 Then, the step from process flow diagram to final piping and instrument
diagram is prepared with minimum effort, due to standard information regarding each
module. The modules are constructed in workshops and field installation includes only
plugging the modules together.9 Since the checks of the plant are made before delivery,
after installation on site it can be started up. These differences in workflows indicate the
saving in engineering and construction time and corresponding cost reduction.
Figure 5.2. Project workflow – Classical workflow vs. workflow based on modular
components. It is assumed that the modular components are “off-the-shelf” or benefit from
serial manufacture of the modular plant. Otherwise, there would also be work steps as
process flow sheeting etc. similar to classical project description.
Figure 5.2 naturally presents a generic, yet ideal picture. Real-case engineering can be more
complex and need to take into account additional time-consuming engineering activities to
comply with company or governmental specific regulations. It has to be further considered
that the modular project is re-using equipment from earlier projects, for which equipment
and process and instrumentation diagram (P&ID) design is required, as well as
procurement. Thus, an ultimate comparison needs to mirror a first-time built classical
project with a modular project where a large part of initial engineering has already been
done.
5.1.4. Micro process technology

5.1.4.1. Cost saving potential
Micro process technology provides a potential for improvements to chemical processes
regarding economic sustainability. An economic advantage principally can occur from a
104
5.1. Introduction
reduction in capital expenditure (CAPEX) through lower investment and from a reduction
in operating costs (OPEX) by introducing this technology. Merck Company together with
the Technical University Clausthal presented a four staged potential analysis for the
microreaction technology as theoretical, technical, material and economical potential.24
Cost saving from using a microreactor was discussed for the example reaction of nitration
to 3-methyl-4-amido-5-nitro benzoic acid. Roberge et al.25 presented a cost analysis of
continuous vs. batch production for large scale pharmaceutical production. For a large-scale
unit, the CAPEX for a microreactor system can be as high as, or even higher than that of
the batch process. Regarding OPEX, the raw material costs account for 30–80 %. Higher
product yield and quality attained represents the main cost saving potential for
microreactors. This also has a direct influence on labor and waste treatment costs.
Krtschil et al. and Azurchem Company made an economic evaluation of the 4-
cyanophenylboronic acid formation process.26 Besides the assessment of the existing
microchemical process, capacity increase case scenarios were described. Five- and 10-fold
increases in capacity were considered. The plant size per given production rate can be
decreased with the order-of-magnitude change in productivity, enabling a reduction of the
overall costs to 25 % of the investigated microchemical process with the 10-fold capacity
increase.26 It was also noted that the equipment cost having a low share should have a
minor impact on the decision to go for the novel technology.
Hessel et al.27 performed a cost analysis for the Kolbe-Schmitt synthesis of 2,4-dihydroxy
benzoic acid. The base case considered is microreactor production at 4.4 tons/year, and the
final product cost is calculated as approximately 91 €/kg (based on a reaction time of about
4 s). A microreactor production under the same conditions and with the same productivity,
but being operated at a reaction time of 2 h under reflux conditions, would run into a final
product cost of approximately 17,350 €/kg, which is fully out of economic range. For a
similar throughput, the operating cost for a 20 L batch-reactor is approximately 107 €/kg.
This economy of scale can be seen in relation: a 1 L batch, which matches roughly the
space requirements of the microreactor, including tubings (same scale of equipment),
accounts to a product cost of approximately 985 €/kg. This one order of magnitude increase
in product costs also reflects the economy of scale of chemical industry – in negative
manner. For the microstructured reactor with a 10-fold throughput, the operating cost is
reduced to approximately 57 €/kg which reflects the economy of numbering up of flow
chemistry.27,28
5.1.4.2. Plant embedment
It is an aim of this chapter to analyze for the first time the combined benefits of process
intensification enabled by smart-scaled process-intensified reaction equipment and of
modular compact plants. The first can be upgraded and integrated batch technology or,
following a prime industrial trend29, continuous processing, favorably with small-scale flow
105
equipment, which may be milli-reactors or, when needed, microstructured reactors. This
chapter focuses in a methodological way on (virtual) microreactor applications; yet they
stand for the use of any similar high-performing reactor technology, and finally, the most
robust will be chosen if the performance is similar. The innovative microstructured reactors
have so far been implemented in a more conventional plant environment and tailored to the
needs of these reactors and process intensification. Thus, so far, it remains unclear which
potential dedicated process control, utility, and safety systems may add. Here, it is aimed at
a financial answer; the processing advantages need to be determined experimentally
through demonstration projects.
5.1.5. Cash flow analysis for chemical plants

The NPV is a central tool in discounted cash flow analysis and is a standard method for
using the time value of money to appraise long-term projects.30 It is used to find the sum of
future cash flows discounted back to its present value. Thus, the NPV is sensitive to the
discount rate used. The discount rate is the minimum acceptable return that must be
achieved by an investment.30 It is adjusted to compensate for the degree of risk in an
investment with the rate increasing with the risk level. Since the modular plant concept
lowers the risk in process development, this allows lower discount rates, which depends on
risk. This in turn enables a higher NPV to be achieved.
Modular compact plants can be quickly built and applied at any location with approval for
chemical manufacture and needed installations, facilities, and piping to take advantage of
economical potentials. Producing directly at a customer’s site can be attractive regarding
adapting the product to varying requirements quickly; also if the product is hazardous to
transport.11 Depending on the process, if raw material consumption is high, production
close to the source can be advantageous in terms of transportation costs and availability of
resources. Another argument for distributive manufacturing is if the process has high
energy demands and is moved to a site where energy is reliably available at a low price.
Production in close proximity to an existing plant might also determine the plant location
decision. In such Verbund fashion, compact plants can be used to produce derivatives from
the product or upgrade a small volume of products.17 They can also be used for new product
production whose market chances are unknown.
A detailed calculation of capital investment and operating cost for a modular plant, to our
best knowledge, has not been made so far in open literature quantifying the benefits of
modular design. In particular, this has not been done considering the step-change character
of process intensification and smart-scale technologies such as microstructured reactors,
with the aim of checking if they add to the advantage of modular compact plants.
We would like to express clearly at this point that naturally for a very new technology such
as the Evotrainer plant infrastructure, a number of assumptions had to be made. These are
106
5.2. Methodology
listed clearly throughout the text. Wherever possible, in the context of an industrial
development, a rationalization of the assumption is made. Clearly those parameters, such as
the cost reduction by pre-manufacture or the risk-related lowering of the rent, have a
decisive impact on the outcome of the study. These are to be proven by industrial
demonstration activities. For all these reasons, we have assigned this study as an
investigation towards model-based scenarios (including virtual microreactor applications)
making clear that the generic outcome is the main ambition in the sense of a parametric
sensitivity and a proof of principle for a new economic model in chemical manufacturing.
5.2. Methodology
5.2.1. Demonstration processes and market applications considered
The chemical manufacturing in the compact and premanufactured plant infrastructure was
analyzed for three application fields – bulk chemistry, fine chemistry, and pharmacy. These
markets have very different production volumes, product prices, added value, plant
technologies and more8, as summarized in Table 5.1. It was clear and expected that the bulk
chemical manufacturing in the novel plant infrastructure is economically not efficient. Yet,
this case serves as a comparison to complete the picture given here.
Table 5.1. Difference in characteristics of bulk chemicals vs. fine chemicals
Characteristics Bulk chemical Fine chemical Pharmaceutical

4 6 2 4
Volume (tons/yr) 10 – 10 10 -10 10-103
Price ($/kg) < 10 > 10 > 100
Added value Low High Very high
Processing Continuous Batch-wise Batch-wise
Plants Dedicated Multi-purpose Multi-purpose
Flexibility Low High High
Safety and environmental efforts Relatively low High Relatively high
The fine chemical considered here is 2,4-dihydroxybenzoic acid, which was chosen due to
extensive past experience with experimental microreactor work and cost analysis for the
respective synthesis, i.e., all relevant cost data (materials, equipment, performance, etc.)
were at hand. In the following, the published background is given; cost information is part
of the introduction given above. In this study investigations were centered around this fine
chemical process. As bulk-chemical, adipic acid manufacturing was considered, for which
extended results are at hand. Also, experimental research on the adipic acid synthesis in
flow is performed in our group. The pharmaceutical product considered is naproxen, a
commercial drug. Reasons for this selection were the general relevance of this product on
107
the pharma market, the fact that patent protection is not given anymore (market prices
orient more on function/production rather than on development costs) and, last but not least,
that a derivative of this product (naproxcinod) has been made at large scale in flow with
microreactors.31 More details concerning bulk chemistry and pharmacy application are
given in results section.
2,4-Dihydroxybenzoic acid is used as an intermediate in the manufacture of dyes,
pharmaceuticals and of additives in photography, and cosmetics.32 It is produced via
conversion of resorcinol with aqueous potassium carbonate solution at 100 °C and applying
a carbon dioxide pressure of 4.5 bar. A by-product of 2,6-dihydroxybenzoic acid (γ-
resorcylic acid) is formed and can be separated by recrystallization.32
Scheme 5.1. Reaction scheme of the Kolbe-Schmitt reaction giving 2,4 –dihydroxybenzoic
acid
The reaction of resorcinol to 2,4-dihydroxybenzoic acid in aqueous potassium bicarbonate
solution at high temperatures and high pressure using a microreactor rig was studied
previously.33 The reaction is shown in Scheme 5.1. The achieved yield was up to 45 %,
which compares closely to the batch operation carried out for at least 2 h (only reaction
time considered here); much higher yields are difficult to reach due to the reversible nature
of the reaction. Microreactor superheated processing at 200 °C and 40 bar in comparison to
the batch operation carried out at 100 °C and 1 bar, enabled a reduction of reaction time by
a factor of 450 at comparable yields. Cost analysis of the microreactor and batch reactor
operation of the realized process was done previously at a production rate of 0.55 kg/h with
a five-tube reactor.27,28 This corresponds to a production rate of 4.4 tons/year assuming an
8000 h/year operation. Also, the production rate with a 10-fold higher throughput (44
tons/year) equivalent to microstructured reactors operating in parallel was considered. It
was found in other studies that new approaches such as microwave heating (MW) and using
ionic liquids (IL) can increase the yields compared to the conventionally heated process
from about 40 % up to 60 % and allow operation at a lower reaction temperature.34 This
improvement in yield can have a profound effect in terms of cost, since synthesis of high-
value product is made from expensive raw materials. However, the investment for
microwave heating and the high cost of ionic liquid material (due to missing recycling at
this point of time) presents economic burden for these processes.
108
5.2. Methodology
In this study, a capacity of 200 tons/year is selected according to the production capacity of
the compact and pre-manufactured plant infrastructure. The process data of raw material
and utility requirements are based on the previous experimental data of the microreactor
and batch reactor operation at a yield of 45 % for both. Besides these two cases, a virtual
case of 60 % yield with the same microreactor operation was considered to take into
account the potential of achieving high yields with the use of microreactors, which has been
demonstrated for so many reactions35 (yet the Kolbe-Schmitt reaction here is rather an
exception, if not considering the IL- and MW-processing).
5.2.2. Capital investment estimation

The capital needed to supply the required manufacturing and plant facilities is called the
fixed capital investment (FCI), while that necessary to start the plant up and operate to the
point when income is earned is termed the working capital (WC) with the sum making the
total capital investment (TCI).30 WC is recovered at the end of the plant life. FCI is
subdivided into manufacturing FCI, also known as direct cost, and non-manufacturing FCI,
also known as indirect cost. Direct costs include purchased equipment, purchased
equipment installation, instrumentation and controls, piping, electrical systems, buildings
and service facilities.30 Indirect costs include engineering and supervision costs covering
the design of the plant and purchasing, procurement and construction supervision as well as
contingency allowance to cover for unforeseen circumstances.30
Capital cost estimation is done based on an estimate of the total purchase cost of the major
equipment items required for the process and the other costs being estimated as factors of
the equipment cost. This method is called the factorial estimate method. After the
preparation of the material and energy balances for the process and preparation of the
flowsheets, sizing of major equipment items is done. Then an estimation of the purchased
cost of equipment is made. In literature log-log plots or correlations for the main types of
process equipment based on cost data obtained from vendors are provided30,36 that are used
here for estimation.
For the detailed factorial estimate method used here, the factors allow the total capital
investment to be calculated based on information of the purchased equipment cost. These
factors are given in literature and are determined for typical chemical plants of three
general process types – solid, solid-liquid and liquid processing. Table 5.2 gives the ratio
factors that are used here for estimating FCI based on delivered equipment cost. These
factors are based on work by Peters et al.30, where yard improvements, legal expenses and
contractor’s fee are neglected in this context. The delivered equipment cost is calculated
with a delivery allowance of 10 % of purchased equipment cost.
The compact and pre-manufactured plant infrastructure of the Evotrainer includes
instrumentation and controls, electrical systems, building and service facilities in its
109
structure. The associated costs of these in calculating the direct costs with the detailed
factorial estimate method are already incurred in the Evotrainer infrastructure. Therefore,
the factors used to calculate these in this scenario are eliminated and the corresponding total
cost is replaced by the efforts about the Evotrainer infrastructure. Due to the compact
structure of the plant and the standardization of units as discussed previously in Section
5.1.3, the associated costs for the purchased equipment installation and piping are adjusted
to lower than for a conventional plant, as an estimate by 35 %. Engineering and supervision
costs included in the indirect costs are also expected to be significantly lower than for a
conventional plant, due to the elimination of customized equipment, piping and structural
design as discussed previously in Section 5.1.3. Also, the contingency allowance can be
reduced significantly due to the lower risk of design errors. The factors in their estimation
are decreased accordingly as an estimate by 50 %.
Table 5.2. Fixed capital investment determination based on detailed factorial estimate
method (as given by Peters et al.30).
Solid Solid-fluid Fluid

processing processing processing
plant plant plant
Direct costs
Delivered equipment cost (E) 1.00 1.00 1.00
Purchased equipment installation 0.45 0.39 0.47
Piping 0.16 0.31 0.68
Instrumentation & controls 0.18 0.26 0.36
Electrical systems 0.10 0.10 0.11
Buildings 0.25 0.29 0.18
Service facilities 0.40 0.55 0.70
Total Direct Plant Cost 2.54 2.90 3.50
Indirect Costs
Engineering and supervision 0.33 0.32 0.33
Contingency 0.35 0.37 0.44
Total Indirect Plant Cost 0.68 0.69 0.77
Fixed Capital Investment 3.22 3.59 4.27
The corresponding fixed capital investment calculation for the Evotrainer based scenario
can be done according to Table 5.3. This table is based on factors used for a solid-fluid
processing plant by Peters et al. given in Table 5.2. Since the Evotrainer is a pre-
manufactured plant infrastructure, part of the costs for instrumentation and controls, electric
systems, buildings, and service facilities are constant and here a fixed value was assumed
which shall not be disclosed here due to confidentiality. The working capital is then
estimated as 15 % of the TCI.30 With the knowledge of the fixed capital investment as
explained above, the TCI is found.
110
5.2. Methodology
The fixed cost, mentioned in Table 5.3 and considered for further calculations, stands for
the current Evotrainer infrastructure (third generation) with given demands and flexibility
on scale and possibly on application and fits to the production targets considered (200 t/a).
Yet, future generations of such a modular plant with a much different scale and project task,
may have other fixed costs. That would naturally change the outcome of the calculations.
Table 5.3. Fixed capital investment determination based on detailed factorial estimate
method for the Evotrainer scenario
Solid-fluid
processing
plant
Direct costs
Delivered equipment cost (E) 1.00
Purchased equipment installation 0.25
Piping 0.20
Instrumentation & controls
Electrical systems
Buildings Evotrainer
Service facilities
Total Direct Plant Cost
Indirect Costs
Engineering and supervision 0.16
Contingency 0.185
Total Indirect Plant Cost
Fixed Capital Investment
5.2.3. Operating cost estimation

Operating costs include direct costs and general expenses. Direct costs, also called as
manufacturing costs, refer to the costs of producing the product at the production site. In
addition to these costs are general expenses for contributing to corporate functions of
research and development and selling of the product.30 The direct costs are divided into
variable and fixed operating costs. Variable costs include raw material, utilities and
catalysts and solvents.36 Fixed costs include operating labor, supervision, maintenance,
plant overheads, laboratory charges, depreciation, taxes and insurance. General expenses
include administrative, distribution and marketing and research and development costs.30
The variable cost items are determined based on mass and energy balances and operating
labor is determined based on the production rate and the type of processing plant. The other
items are estimated based on the known operating cost items and fixed capital investment.
This factorial method of estimation of operating cost can be seen in Table 5.4. These factors
used for estimation are based on average values given by Peters et al.30 and Sinnott.36
111
Table 5.4. Operating cost determination based on factorial estimates (as given by Peters et
al.30 and Sinnott36)
Operating cost items Factor Basis

Variable Costs
Raw materials - Required input
Utilities - Required input
Catalysts and solvents - Required input
Fixed Costs
Operating labour - Required input
Operating supervision 0.15 of operating labor
Maintenance and repairs 0.05 of FCI
Plant overheads 0.50 of operating labor
Laboratory charges 0.15 of operating labor
Local taxes 0.02 of FCI
Insurance 0.01 of FCI
Depreciation / Capital charges - Calculated separately
Direct Production Cost - Variable + Fixed Costs
General Expenses 0.20 Direct Production Cost
Total Product Cost -
The Evotrainer infrastructure enables the reduction of operating costs due to its modularity
and compact structure in terms of cost of operating labor and due to standardization of units
in terms of cost of maintenance, as discussed previously in Section 5.1.3. The factors in
their estimation are decreased accordingly as an estimate by 20 %, which is a rough-cut
assumption taken from the very first experience. Since some other cost items, including
operating supervision, plant overheads and laboratory charges, are related to operating labor
as explained above, their corresponding costs are reduced accordingly as well.
Table 5.5. Annual depreciation rates with modified accelerated cost recovery system
(MACRS) depreciation method 5-year recovery period
Recovery year Depreciation rate, %

1 20.00
2 32.00
3 19.20
4 11.52
5 11.52
6 5.76
Depreciation is calculated separately because it is not the same each year. A depreciation
method termed modified accelerated cost recovery system (MACRS) with a 5-year
112
5.2. Methodology
recovery period and half-yearly convention is selected, which is typically preferred for
chemical plants.30 Table 5.5 shows the annual depreciation rates for this method.
5.2.4. Cash flow analysis

A cumulative cash flow diagram allows visualization of the forecasted cumulative new cash
flow over the life of a project. It gives an idea of the resources required and timing of the
earnings. The cash flow is found by adding depreciation to net profit after taxes and
subtracting the TCI:
Net profit = (sales-total operating costs (including depreciation)) (1-φ) (5.1)
Cash flow = net profit depreciation total capital investment (5.2)

where φ is income tax rate.
For doing cash flow analysis, the schedule of the project is decided. Generally, for a
chemical plant, the construction takes place heavily in the 1st year and in the 2nd year the
remaining construction is completed.36 Accordingly, it is taken here that 30 % of the FCI is
spent in the 1st year and 70 % of the FCI is spent in the 2nd year. Then, in the 3rd year,
working capital is allocated and the plant starts production. Production at full design
capacity does not usually occur in the 3rd year, so that production at 50 % capacity is taken
for this year. This gives 50 % of revenue in this year. In terms of the operating cost, 50 %
of the variable cost (VC) but 100 % of fixed costs (FC) occur in this year. From the 4th
year, production at full capacity takes place until the end of project, with 100 % of annual
revenue and 100 % of operating costs. For this evaluation 8 years of operation is considered
with the end of the project at the 10th year, due to 2 years of construction. At the end of the
project working capital is recovered by the sale of materials and supplies. The cash flow
schedule is given in Table 5.6.
Due to the reduction of the planning and construction period, it was assumed that the
construction of the Evotrainer plant can be achieved in 1 year. So, 100 % of FCI is spent in
1st year and in the 2nd year, working capital is allocated and the plant starts production.
Production at full design capacity does not usually occur in the 2nd year, so 50 % of VC and
50 % of revenue is taken but 100 % of FC occur this year. From the 3rd year, production at
full capacity takes place. The end of the project is at the 9th year, with 8 years of operation
selected here. The cash flow schedule is given in Table 5.7.
Table 5.6. Cash flow schedule for the conventional plant
Year FCI, % WC, % VC % FC, % Revenue, %

1 30 0 0 0 0
2 70 0 0 0 0
3 0 100 50 100 50
4-10 0 0 100 100 100
113
Table 5.7. Cash flow schedule for the Evotrainer plant
Year FCI, % WC, % VC % FC, % Revenue, %

1 100 0 0 0 0
2 0 100 50 100 50
3-9 0 0 100 100 100
The cash flows show the value in the year in which they occur, so they stand for the future
worth of the projected industrial manufacture. Since cash flows occur in different years
throughout the project, it is necessary to convert them to equivalent values. This is done by
discounting future cash flows to a particular point in time. In this way, the time value of
money is taken into consideration. The time value of money is related to the capacity of
money to earn money. The money earned in any year can be reinvested as soon as it is
available and can start to earn a return. NPV calculation is made here to assess the
profitability of a projected industrial manufacture. The net cash flow in each year of the
project is brought to its present value at the start of the projected industrial manufacture by
discounting it at a chosen discount rate. The NPV is then found as the sum of the present
values of the future cash flows30 :
(5.3)
1
where is cash flow in year , is project life in years and is discount rate.
The value of NPV is strongly dependent on the discount rate chosen. It will have a lower
value if a higher discount rate is selected. The appropriate discount rate to use is determined
from the minimum acceptable rate of return. It is the rate of earning that must be achieved
in order for it to be acceptable to the investor, given its risk and opportunity cost of
forgoing other projects.30 The discount rate is determined based on the rate from safe
investment such as on government bonds (8 %) and adjusted to account for uncertainties
associated with a new project. With the increase in risk for the project, the risk premium
added on top of the risk-free rate increases.

Since the fine-chemical application comprises the envisaged target market of the Evotrainer
infrastructure and also since for this application all relevant data were at hand, the cost
analysis investigations started with and were centered around this chemical process. From
the earnings made and common data on the compact and premanufactured plant Evotrainer
infrastructure gathered, the same procedure was applied to the bulk-chemical (adipic acid)
and pharmaceutical (naproxen) applications for comparison.
114
5.3.1. Capital investment estimation of fine chemical

The process equipment for the 2,4-dihydroxybenzoic acid manufacture used included
pumps for feeding the reactor at the required pressure, a mixer for mixing the reactants,
heat exchangers for heating the reactor feed and cooling the reactor effluent and a reactor
unit supplied with heating.33 The choice of apparatus refers to flow equipment in the case of
microreactor operation and to batch equipment in the case of batch operation. For the
microreactor operation, the reaction conditions are 200 °C and 40 bar, whereas for the batch
operation, 100 °C and 1 bar were used. Besides the reaction part downstream processes are
considered as well. After cooling to about 60 °C, water is added to the suspension until
complete dissolution of the solid is achieved. Then, the resulting mixture is acidified with
hydrochloric acid at room temperature. Then, by cooling to about 15 °C, crystals are
formed which are then filtered and dried. The product is attained with about 99 % purity.
For the downstream processes, the same batch equipment was used for both operation types
(batch/flow) consisting of a dissolving tank, crystallizer, filter and dryer.
In addition to the normal flow process with the standard yield, a virtual flow process with a
higher (60 %) yield, termed “micro 60 %”, was investigated as well. This refers to the
proven capability of microreactors to increase selectivity, which actually was also found for
the Kolbe-Schmitt synthesis, yet with use of microwaves (which was not considered as
equipment here).34
Thus, equipment cost estimates are made for three processes (batch, micro, micro 60 %).
The equipment for the reaction part is based on the costs of lab set-ups determined
previously for 4.4 tons/year production.27 For the batch operation, scaling up of the batch
reactor and apparatus is done and for the flow operation, the microreactors are numbered
up. The downstream equipment is simulated with Aspen and their respective costs are
determined using the literature.36 The microreactor operation has a higher cost than the
batch operation, due to the higher cost of flow equipment in comparison to batch
equipment. However, the difference is not so high (about 25 %), mainly because the
downstream equipment used is the same. The equipment cost for the microreactor operation
with 60 % yield is estimated to be 10 % lower than the microreactor operation with 45 %
yield, since a lower amount of raw material goes through to achieve the same amount of
product, making it possible to use smaller equipment.
Applying the methodology described in Section 5.2.2, the detailed factorial estimate
method based on the total purchased equipment cost, the FCI for the three processes (batch,
micro, micro 60 %) is determined for the conventional plant and the Evotrainer plant. With
the three process types taking place in two plant types, a corresponding six cases are
considered in this evaluation.
115
The cost of the Evotrainer infrastructure, comprising the costs of instrumentation and
controls, electrical systems, building and service facilities, were estimated. The sum of
these costs calculated separately with factorial estimates for the conventional plant is
slightly higher than that of the cost of the Evotrainer infrastructure. The other direct costs of
installation and piping are considerably lower for the Evotrainer plant regarding the factors
used, as explained in Section 5.2.2.
Also, the indirect costs of engineering and supervision and contingency allowance are
lower due to lower factors used. This makes the capital investment for the Evotrainer plant
around 15 % lower than that for the conventional plant. Figure 5.3 shows the FCI estimated
for the six cases showing the direct and indirect plant costs separately.
It should be noted here that batch operation in the conventional plant has a very close
capital investment with the microreactor operation in the Evotrainer plant. This shows the
opportunity of using Evotrainer infrastructure for applying micro process technology to be
competitive, with conventional operation in terms of capital investment which cannot be
achieved normally due to the higher cost of more advanced flow reactors.
Figure 5.3. Fixed capital investment estimated for the six cases considered
5.3.2. Operating cost estimation of fine chemical

The operating cost estimation is started with estimation of raw material, utility and
operating labor costs. The raw material requirement is derived from the experimental data
for the 45 % yield for the micro and batch operations and for the 60 % yield it is estimated
by stoichiometry. A five- to six-fold excess of potassium bicarbonate to resorcinol is used
conventionally. However, for the microreactor operation, this could be reduced to three-
fold excess, enabling saving in terms of raw material cost. The microreactor operation with
116
60 % yield has a considerably lower raw material cost, due to lower raw material
requirement. The raw material costs are based on wholesale price information.37 The
operating labor is estimated based on rule of thumb of solid-fluid plants as 10 employee-
hours per ton of product.30 For the microreactor operation 1.5 times lower labor of 6.6
employee-hours per ton of product is taken, partly accounting for a faster product
manufacture by estimation. An average common labor rate of 34 €/employee-hour is taken
based on the Engineering News Record.38 This corresponds to one to two operators, which
is a reasonable estimate for such a small plant of 200 tons/year production.
The other operating costs are determined by factors of the known values as described in
Section 5.2.3. Figure 5.4 shows the operating costs estimated for the six cases split into the
variable costs, fixed costs and general expenses.
Since the Kolbe-Schmitt synthesis presents the case of a synthesis of high-value product
from relatively expensive raw materials, the raw material cost dominates the operating
costs. The raw material costs constitute around 66 % of the operating costs and the total
variable costs sum up to around 80 %. Accordingly, the share of other operating cost items
is low. The use of an Evotrainer infrastructure affects the operating labor and maintenance.
However, since these cost categories have a low share in the overall operating cost, the
effect of the Evotrainer infrastructure is seen as very small for this process example. For
process examples with cheaper raw material the effect of the Evotrainer plant would be
larger. It is seen that compared to batch, lower operating cost values can be achieved with
the microreactor. Also, the increase in yield in the microreactor operation with 60 % yield
case enables a profound reduction of the operating costs, due to the raw material cost
reduction, which has a big effect on the overall operating cost.
Figure 5.4. Operating cost estimated for the six cases considered
117
5.3.3. Cash flow analysis of fine chemical

In Table 5.8, the cash flow analysis is given for the case of batch process in a conventional
plant. The schedule of the project is as described in Section 5.2.4 and Table 5.6 for the
conventional plant. In the 3rd year, the working capital is allocated and the plant starts
production at 50 % capacity. After the 3rd year, 100 % of revenue is earned and 100 % of
operating cost (OC) is spent. With the eight years of operation selected the end of the
project is at the 10th year. At the end of the project, the working capital is released and
taken as a positive increment to the cash flow.
The revenue, that is the income earned from the sales of the product, is found by sales price
58 €/kg37 multiplied by the amount of 200,000 kg of product produced. The annual
depreciation charge is calculated with rates of MACRS depreciation method given in Table
5.5 for the recovery period. Gross profit is calculated as the difference between the revenue
and the operating cost, including depreciation. The net profit is calculated as gross profit
after taxes with Eq. 5.1. Then, the cash flow (CF) is found using Eq. 5.2. The cumulative
cash is calculated to see the cash position over the life cycle of the project. The present
value (PV) of the cash flow in each year is then determined by multiplying with PV factor
for the discount rate selected which is calculated as 1 . The NPV up to year is the
cumulative sum of all the present values of cash flows up to that year as given in Eq. 5.3.
For the case of batch process in a conventional plant with the discount rate of 15 % the
NPV is found as 9076 103 €. Cash flow analysis tables for the other cases are given in the
appendix.
Table 5.8. Cash flow analysis for the case of batch process in a conventional plant (all cost
values in the unit of 103 €)
End of Year 1 2 3 4 5 6 7 8 9 10
FCI -377 -880
WC -222 222
Fraction of Capacity 0.5 1 1 1 1 1 1 1
Annual Revenue 5800 11600 11600 11600 11600 11600 11600 11600
Annual OC w/o depr. -3892 -6482 -6482 -6482 -6482 -6482 -6482 -6482
Annual Depreciation -251 -402 -241 -145 -145 -72
Annual Gross Profit 1657 4716 4877 4973 4973 5046 5118 5118
Annual Net Profit 1077 3065 3170 3233 3233 3280 3327 3327
Annual CF -377 -880 1106 3467 3411 3377 3377 3352 3327 3548
Cumulative CF -377 -1257 -150 3317 6728 10106 13483 16835 20162 23710
PV factor (15 %) 0.87 0.76 0.66 0.57 0.50 0.43 0.38 0.33 0.28 0.25
PV of annual CF -328 -668 730 1976 1706 1452 1283 1106 931 887
PV cumulative CF -328 -996 -266 1710 3416 4868 6151 7258 8189 9076
118
The important consideration here is the selection of the discount rate as explained
previously in Section 5.2.4, which is the minimum acceptable rate of return suitable for
each case. For a low level of risk in investment, a lower rate can be selected. Since with the
Evotrainer plants there is a lower risk in bringing the product from idea to production, a
lower discount rate can be used. That is evident from Figure 5.2, with its fewer respective
steps in development and the more straight forward combined development from laboratory
to production in this pre-manufactured facility, mentioned in the corresponding text.
As micro process technology is fairly new technology, it can be considered to have a higher
risk than the established process technology of batch production. Accordingly, a higher rate
is selected for micro processes. The discount rates selected for the different cases based on
the level of risk estimation are given in Table 5.9. The known conventional batch process
rate of return was taken as 15 % and by estimation reduced by 3 % when being operated in
a container. The risk increase by use of micro process technology was considered with a 5
% higher rate.
Table 5.9. Minimum acceptable rate of return (discount rate, ) selected for the cases
Case Discount rate, %

Batch, Conventional 15
Batch, Evotrainer 12
Micro, Conventional 20
Micro, Evotrainer 17
Figure 5.5. Cumulative cash flow diagram for the six cases considered
119
The diagram of cumulative cash flow over the course of the project life for the six cases is
given in Figure 5.5. It is seen that in comparison to the conventional plant, the Evotrainer
plant enables a higher cumulative cash flow for each process case. Also, the capital
investment is repaid quicker with the reduction of the construction period. The microreactor
operation with 60 % yield has the lowest operating cost and highest cash flows are attained
accordingly. Due to synthesis of the high value product, the reduction in operating costs has
a significant effect. This makes the investment cost difference a minor influence.
Accordingly, the batch operation has the lowest cash flow, because of its higher operating
cost and cannot profit from its lowest capital investment.
Cumulative cash flow diagrams do not take into account the time value of money.
Accordingly, NPV calculation is made. As explained before, with a higher discount rate,
lower NPV values can be achieved for the same investment consideration. The resulting
NPV for the six cases using discount rates given in Table 5.9 is given in Figure 5.6.
The reduction of risk achieved with the compact and pre-manufactured Evotrainer plant
enables higher NPV values to be achieved for each process considered. In comparison to
the conventional plant for the reaction studied, with the use of the Evotrainer infrastructure,
an increase in NPV of around 40 % is obtained for each process type. Regarding the
process type, although the microreactor processes have a higher discount rate, a still higher
NPV than with the batch process can be achieved due to lower operating cost.
Figure 5.6. Net Present Value for the six cases considered at different discount rates
120
Figure 5.7. Net Present Value for the six cases considered at same discount rate of 15 %
To eliminate the effect of discount rates selected, the NPV calculation is also done by using
a 15 % rate for all cases. Accordingly, the discount rate used for batch, Evotrainer case is
increased, batch, conventional case is kept the same and the other cases are decreased
compared to the cases given in Figure 5.6. The resulting NPV for the six cases is given in
Figure 5.7. Due to the changed discount rates the NPV of batch, Evotrainer case decreased
compared to Figure 5.6.
In Figure 5.7, the same discount rate is considered and accordingly the same risk level is
taken. That is why the high difference in NPV seen in Figure 5.6 achieved with the
Evotrainer infrastructure owing to lower risk, is not seen in Figure 5.7. However, using the
same risk level, the Evotrainer plant still enables higher NPV values. In comparison to the
conventional plant the Evotrainer infrastructure gives around 18 % higher NPV at same risk
level. This difference is due to a lower operating cost achieved with the Evotrainer plant
and the reduction in the construction period that enables higher financial gain to be
achieved since the money earned earlier is worth more than that earned later. Considering
the process type, the microreactor process is seen to have a higher NPV, due to the
profound effect of the reduction in the operating cost achieved.
5.3.4. Comparison of chemical product classes – bulk chemical, fine chemical

and pharmaceutical
As bulk-chemical, adipic acid manufacturing was considered, for which extended results
are at hand. Recently, a one-step oxidation of cyclohexene to adipic acid by 30% hydrogen
peroxide (H2O2) has been reported.39 We investigated the process design intensification
121
through the microreactors and flow processing of this one-step adipic acid synthesis.
Further analysis of this can be found in Chapters 7 and 8. For this route, the use of
microreactor technology is selected to overcome limits in interfacial transfer, to safely
handle hydrogen peroxide, to explore new, harsher process chemistries, and to test for
better selectivity at much reduced reaction times (transport and chemical intensification
fields). Current commercial production processes for adipic acid are carried out in two
steps: the first step involves the production of so-called KA oil (a mixture of
cyclohexanone, the ketone or K component, and cyclohexanol, the alcohol or A
component). The second stage involves the oxidation of the KA oil to adipic acid, with an
excess of strong nitric acid. The reaction scheme of the two routes are given in Scheme 5.2.
Scheme 5.2. Reaction scheme of commercial 2-step and one-step adipic acid synthesis
Here, four cases are considered at a capacity of 200 tons/year according to the production
capacity of the Evotrainer infrastructure: two-step commercial process in conventional or
Evotrainer plant and 1-step micro process in conventional or Evotrainer plant. The
investment, operating cost and revenue that are used in making cash flow analysis for these
cases are based on the previous study of the two-step and one-step operations through
process simulation.40 The investment cost calculation is done with the same factorial
estimate method as used above, based on the total purchase equipment cost estimated for
the two-step process and the one-step micro process. Due to a more compact plant design
achieved with the one-step micro process, a significant reduction in capital cost occurs. A
higher impact of using the Evotrainer infrastructure is seen for the two-step process in the
reduction of the capital cost.
The variable operating cost calculation is then done based on the mass and energy balance
that enables determination of the raw material and utility costs. The operating labor is
estimated as 10 employee-hours per ton of product for commercial and 6.6 employee-hours
per ton of product for the microreactor operation. The other operating costs are determined
by factors of the known values, carrying out the same methodology as before. The revenue
is calculated using the wholesale price of 3000 €/ton.41 The cash flow and thereafter the
NPV calculations are done as explained for the previous example. Cash flow analysis tables
for these cases are given in the appendix. Again, the same discount rates given in Table 5.9
122
are used for analysis where batch representing the commercial operation in this bulk-
chemical process example.
The total operating cost is greater than the revenue for this process example, showing that it
is not worth investing at this very low production rate. Accordingly, minus NPV values are
attained showing that the margin is not enough to recover the investment (see Figure 5.8).
Actually, a production capacity of 130,000–450,000 tons/year defines the commercial range
seen for adipic acid synthesis.36 At a higher production rate, the effect of the fixed operating
costs decreases, enabling higher margins to be achieved.
Figure 5.8. Net Present Value for adipic acid synthesis (bulk-chemical case)
The pharmaceutical product considered is naproxen. It is a nonsteroidal anti-inflammatory
drug. It is used to relieve pain from various conditions such as headaches, muscle aches,
tendonitis, dental pain, and menstrual cramps. It also reduces pain, swelling, and joint
stiffness caused by arthritis, bursitis, and gout attacks.42 It was introduced to the market in
1976 and the patent expired in December of 1993. The Syntex manufacturing process used
today has the starting material 2-bromo-6-methoxynaphtelene (BMN), which is converted
to a Grignard reagent which is then coupled with a salt of bromopropionic acid. This way,
the corresponding d,l-acid is produced with a yield > 90 %. This is then efficiently resolved
using N-alkylglucamine (> 95 %).43 The reaction scheme is given in Scheme 5.3.
Scheme 5.3. Reaction scheme of naproxen synthesis

The investment cost of this process is determined based on the equipment required for the
Grignard reaction step and the resolution step, with a racemization-recycling loop. The
123
equipment includes a reactor and related parts, a mixer for mixing the resolving agent, a
heater, a filter for separation of the insoluble salt and a separator for recovery and recycle of
the resolving agent.
Regarding the operating cost, again the major share refers to the raw material cost due to
the production of a high value product use of relatively expensive raw materials. There are
various alternatives to produce naproxen, and production with the starting material BMN
enables the lowest operating cost, with BMN being the least expensive.43 Due to the labor
intensive resolution step, the operating labor is estimated as 20 employee-hours per ton of
product. The other operating costs are determined by factors of the known values carrying
out the same methodology as before. The revenue is based on a wholesale price of
naproxen at 25 kg of approximately $10,000.44 Here, two cases are considered at a capacity
of 200 tons/year according to the production capacity of the Evotrainer infrastructure:
naproxen synthesis with micro in conventional or Evotrainer plant. Cash flow analysis
tables for these cases are given in the appendix. Since micro use is considered 20 %
discount rate is taken for the conventional plant and 17 % discount rate is used for the
Evotrainer plant as in previous cases.
The selection of naproxen as an example is motivated by a well-published industrial
application example of micro process technology. In 2009, the development was started to
get a new drug, naproxcinod to treat patients with osteoarthritis, to market; it reached Phase
III studies.31 Naproxcinod is a naproxen derivative which contains a nitrate group as a
substituent. Compounds containing a nitrate group are difficult to make, as nitration
reactions must be handled carefully due to producing products that can violently
decompose. Highly diluted, biphasic conditions and specialized safety equipment are
necessary to realize this type of processing in a classical batch manufacturing process.
Beyond this safety issue, the selectivity in nitration reactions and work-up are to be
considered to permit extraction and neutralization of the nitrated product. A microreactor
system enabled this eco-efficient production at a few hundred tons capacity of naproxcinod
per year and combined three main steps: the nitration reaction, neutralization and work-
up.31 Thus, it would have been of interest to evaluate the use of microreactors for the
synthesis of naproxcinod as a further step from the naproxen process described above.
However, due to lack of data about this process, naproxen has been studied here.
The resulting NPV calculation for the pharmaceutical cases is given in Figure 5.9 together
with fine-chemical micro process cases and adipic acid 1-step micro process cases for
comparison.
It is evident that the operation of a much more costly product, such as naproxen, has a very
positive effect on the cash-flow. Using a compact and pre-manufactured plant
infrastructure, already after 1 year, a substantial payback and earning is achieved, while a
similar trend is observed with conventional plant technology only after 2 years. The overall
124
5.4. Conclusion
earnings over the years with pharma products are much higher as compared to the two other
application cases. Thus, a GMP-type upgrading of the Evotrainer infrastructure technology
provides a promising business case.
Figure 5.9. Net Present Value comparison of the three chemical product classes
5.4. Conclusion
A detailed calculation of capital investment and operating cost for a modular plant is made
for the first time, quantifying the benefits of modular design. For the evaluation, Evonik’s
new compact and pre-manufactured production platform, termed the Evotrainer (meanwhile
renamed as the EcoTrainer), was considered. Standardization through pre-manufactured
modules assembled into a highly functional plant environment, facilitates savings in the
design and construction time and expenditure. Additional benefits can be gained in terms of
the operating costs as well, regarding operating labor and maintenance. A detailed factorial
estimate method, which is used in making an economic analysis for conventional plants, is
employed to make an analysis for the plant based on the Evotrainer infrastructure.
Regarding the advantages gained with the Evotrainer plant, these factors are either replaced
with the cost of the Evotrainer infrastructure or reduced, since the latter already includes
instrumentation and controls, electrical systems, and building and service facilities in its
structure. The study also considers the step-change character of process intensification and
smart-scale technologies, such as microstructured reactors, together with modular compact
125
plants; the latter also stand for any other modern robust high-performance equipment to
hand in the chemical industry. As stated in the introduction, the main generic aim of this
study was to provide parametric sensitivity and a proof of principle for a new economic
model in chemical manufacturing using model based scenarios (including virtual
microreactor applications). Therefore, several assumptions were made that are stated in the
text.
The chemical manufacturing in the compact plant infrastructure (e.g., framed in a
container-like module) was analyzed for three virtual microreactor applications. These
model-based scenarios for market applications in bulk chemistry, fine chemistry and
pharmacy, take into consideration all possible markets regarding production volumes,
product prices etc. The fine chemical considered was 2,4-dihydroxybenzoic acid, which
was chosen due to extensive past experience with experimental microreactor work. CAPEX
and OPEX calculations and a following cash flow analysis were carried out. In comparison
to the conventional plant, with the use of the Evotrainer infrastructure based plant, an
increase in NPV of around 40 % was obtained; using a lower discount rate for the
Evotrainer plant, it has a lower risk in bringing the product from idea to production.
However, with the use of the same risk level, the Evotrainer plant still enables higher NPV
values (e.g., 18 % for the fine chemical case considered at 200 t/a and the same discount
rate of 15%). This difference is due to the lower operating cost achieved with the
Evotrainer plant and the reduction in the construction period. Considering the process type,
microreactor (or generally intensified) processes were seen to have a higher NPV than that
of the batch processes, due to the profound effect of the reduction in operating cost
achieved. It was seen that the Evotrainer infrastructure and micro process technology can
be synergistic in costs.
The study carried out for the bulk chemical (adipic acid) did not give favorable results as
expected, indicating that bulk chemical manufacturing in the Evotrainer based plant is
economically generally not efficient. In the pharma case (naproxen), a high positive effect
on the cash-flow was seen due to higher value-added making it attractive to be produced in
the Evotrainer based plant.
126
5.5. Appendix: Cash flow analysis tables

Fine-chemical:
Table 5.10. Cash flow analysis for the case of batch process in a Evotrainer plant – fine-
chemical (all cost values in the unit of 103 €)
End of Year 1 2 3 4 5 6 7 8 9
FCI -1128
WC -199 199
Annual Revenue 5800 11600 11600 11600 11600 11600 11600 11600
Annual OC w/o depr. -3837 -6427 -6427 -6427 -6427 -6427 -6427 -6427
Annual CF -1128 1156 3489 3438 3408 3408 3385 3362 3561
Cumulative CF -1128 28 3516 6954 10362 13770 17155 20518 24079
PV factor (12 %) 0.89 0.80 0.71 0.64 0.57 0.51 0.45 0.40 0.36
PV of annual CF -1007 921 2483 2185 1934 1727 1531 1358 1284
PV cumulative CF -1007 -86 2397 4582 6516 8242 9774 11132 12416
Table 5.11. Cash flow analysis for the case of micro process in a conventional plant – fine-
End of Year 1 2 3 4 5 6 7 8 9 10
FCI -485 -1131
WC -285 285
Annual Revenue 5800 11600 11600 11600 11600 11600 11600 11600
Annual OC w/o depr. -2853 -4710 -4710 -4710 -4710 -4710 -4710 -4710
Annual CF -485 -1131 1744 4659 4587 4544 4544 4511 4478 4763
Cumulative CF -485 -1616 128 4788 9375 13918 18462 22972 27451 32214
PV factor (20 %) 0.83 0.69 0.58 0.48 0.40 0.33 0.28 0.23 0.19 0.16
PV of annual CF -404 -785 1009 2247 1843 1522 1268 1049 868 769
PV cumulative CF -404 -1189 -180 2067 3910 5432 6700 7749 8617 9386
127
Table 5.12. Cash flow analysis for the case of micro process in a Evotrainer plant – fine-
End of Year 1 2 3 4 5 6 7 8 9
FCI -1308
WC -231 231
Annual Revenue 5800 11600 11600 11600 11600 11600 11600 11600
Annual OC w/o depr. -2788 -4646 -4646 -4646 -4646 -4646 -4646 -4646
Annual CF -1308 1819 4667 4608 4573 4573 4547 4520 4751
Cumulative CF -1308 511 5178 9786 14359 18932 23479 27999 32750
PV factor (17 %) 0.85 0.73 0.62 0.53 0.46 0.39 0.33 0.28 0.24
PV of annual CF -1118 1328 2914 2459 2086 1783 1515 1287 1156
PV cumulative CF -1118 211 3125 5584 7670 9452 10967 12255 13411
Table 5.13. Cash flow analysis for the case of micro 60 % process in a conventional plant –
fine-chemical (all cost values in the unit of 103 €)
End of Year 1 2 3 4 5 6 7 8 9 10
FCI -442 -1030
WC -260 260
Annual Revenue 5800 11600 11600 11600 11600 11600 11600 11600
Annual OC w/o depr. -2198 -3598 -3598 -3598 -3598 -3598 -3598 -3598
Annual CF -442 -1030 2185 5366 5300 5261 5261 5231 5201 5461
Cumulative CF -442 -1472 713 6079 11379 16640 21901 27132 32333 37794
PV factor (20 %) 0.83 0.69 0.58 0.48 0.40 0.33 0.28 0.23 0.19 0.16
PV of annual CF -368 -716 1264 2588 2130 1762 1468 1217 1008 882
PV cumulative CF -368 -1083 181 2769 4899 6660 8129 9345 10353 11235
128
Table 5.14. Cash flow analysis for the case of micro 60 % process in a Evotrainer plant –
fine-chemical (all cost values in the unit of 103 €)
End of Year 1 2 3 4 5 6 7 8 9
FCI -1236
WC -218 218
Annual Revenue 5800 11600 11600 11600 11600 11600 11600 11600
Annual OC w/o depr. -2141 -3541 -3541 -3541 -3541 -3541 -3541 -3541
Annual CF -1236 2247 5377 5321 5288 5288 5263 5238 5456
Cumulative CF -1236 1011 6387 11709 16997 22285 27548 32787 38243
PV factor (17 %) 0.85 0.73 0.62 0.53 0.46 0.39 0.33 0.28 0.24
PV of annual CF -1056 1641 3357 2840 2412 2062 1754 1492 1328
PV cumulative CF -1056 585 3942 6782 9194 11255 13009 14501 15829
Bulk-chemical :
Table 5.15. Cash flow analysis for the case of 2-step commercial process in a conventional
plant – bulk-chemical (all cost values in the unit of 103 €)
End of Year 1 2 3 4 5 6 7 8 9 10
FCI -646 -1508
WC -380 380
Annual Revenue 300 600 600 600 600 600 600 600
Annual OC w/o depr. -524 -647 -647 -647 -647 -647 -647 -647
Annual Gross Profit -655 -736 -461 -295 -295 -171 -47 -47
Annual Net Profit -655 -736 -461 -295 -295 -171 -47 -47
Annual CF -646 -1508 -604 -47 -47 -47 -47 -47 -47 333
Cumulative CF -646 -2154 -2758 -2806 -2853 -2900 -2947 -2994 -3041 -2709
PV factor (15 %) 0.87 0.76 0.66 0.57 0.50 0.43 0.38 0.33 0.28 0.25
PV of annual CF -562 -1140 -397 -27 -23 -20 -18 -15 -13 82
PV cumulative CF -562 -1702 -2099 -2126 -2150 -2170 -2188 -2203 -2217 -2135
129
Table 5.16. Cash flow analysis for the case of 2-step commercial process in a Evotrainer
plant – bulk-chemical (all cost values in the unit of 103 €)
End of Year 1 2 3 4 5 6 7 8 9
FCI -1577
WC -278 278
Annual Revenue 300 600 600 600 600 600 600 600
Annual OC w/o depr. -421 -544 -544 -544 -544 -544 -544 -544
Annual Gross Profit -436 -448 -247 -125 -125 -35 56.23 56.23
Annual Net Profit -436 -448 -247 -125 -125 -35 36.55 36.55
Annual CF -1577 -399 56.23 56.23 56.23 56.23 56.23 36.55 314.55
Cumulative CF -1577 -1976 -1920 -1864 -1807 -1751 -1695 -1658 -1344
PV factor (12 %) 0.89 0.80 0.71 0.64 0.57 0.51 0.45 0.40 0.36
PV of annual CF -1408 -318 40.03 35.74 31.91 28.49 25.44 14.76 113.43
PV cumulative CF -1408 -1726 -1686 -1650 -1619 -1590 -1565 -1550 -1436
Table 5.17. Cash flow analysis for the case of 1-step micro process in a conventional plant
– bulk-chemical (all cost values in the unit of 103 €)
End of Year 1 2 3 4 5 6 7 8 9 10
FCI -422 -985
WC -248 248
Annual Revenue 300 600 600 600 600 600 600 600
Annual OC w/o depr. -208 -730 -730 -730 -730 -730 -730 -730
Annual CF -422 -985 -156 -130 -130 -130 -130 -130 -130 118
Cumulative CF -422 -1407 -1563 -1694 -1824 -1954 -2085 -2215 -2345 -2228
PV factor (20 %) 0.83 0.69 0.58 0.48 0.40 0.33 0.28 0.23 0.19 0.16
PV of annual CF -352 -684 -90 -63 -52 -44 -36 -30 -25 19
PV cumulative CF -352 -1036 -1126 -1189 -1241 -1285 -1321 -1352 -1377 -1358
130
Table 5.18. Cash flow analysis for the case of 1-step micro process in a Evotrainer plant –
bulk-chemical (all cost values in the unit of 103 €)
End of Year 1 2 3 4 5 6 7 8 9
FCI -1204
WC -212 212
Annual Revenue 300 600 600 600 600 600 600 600
Annual OC w/o depr. -208 -677 -677 -677 -677 -677 -677 -677
Annual CF -1204 -120 -77 -77 -77 -77 -77 -77 135
Cumulative CF -1204 -1324 -1401 -1478 -1555 -1631 -1708 -1785 -1650
PV factor (17 %) 0.85 0.73 0.62 0.53 0.46 0.39 0.33 0.28 0.24
PV of annual CF -1029 -88 -48 -41 -35 -30 -26 -22 33
PV cumulative CF -1029 -1116 -1164 -1206 -1241 -1271 -1296 -1318 -1285
Pharmaceutical :
Table 5.19. Cash flow analysis for the case of naproxen process in a conventional plant –
pharmaceutical (all cost values in the unit of 103 €)
End of Year 1 2 3 4 5 6 7 8 9 10
FCI -371 -865
WC -218 218
Fraction of Capacity 0,5 1 1 1 1 1 1 1
Annual Revenue 32300 64600 64600 64600 64600 64600 64600 64600
Annual OC w/o depr. -15889 -26849 -26849 -26849 -26849 -26849 -26849 -26849
Annual CF -371 -865 10536 24677 24621 24588 24588 24563 24538 24756
Cumulative CF -371 -1235 9301 33978 58599 83187 107775 132339 156877 181633
PV factor (20 %) 0,83 0,69 0,58 0,48 0,40 0,33 0,28 0,23 0,19 0,16
PV of annual CF -309 -600 6097 11900 9895 8235 6862 5713 4756 3998
PV cumulative CF -309 -909 5188 17088 26983 35218 42080 47793 52548 56546
131
Table 5.20. Cash flow analysis for the case of naproxen process in a Evotrainer plant –
pharmaceutical (all cost values in the unit of 103 €)
End of Year 1 2 3 4 5 6 7 8 9
FCI -1050
WC -185 185
Fraction of Capacity 0,5 1,0 1 1 1 1 1 1
Annual Revenue 32300 64600 64600 64600 64600 64600 64600 64600
Annual OC w/o depr. -15785 -26745 -26745 -26745 -26745 -26745 -26745 -26745
Annual CF -1050 10623 24723 24676 24648 24648 24627 24606 24791
Cumulative CF -1050 9573 34296 58973 83621 108269 132896 157501 182292
PV factor (17 %) 0,85 0,73 0,62 0,53 0,46 0,39 0,33 0,28 0,24
PV of annual CF -897 7760 15437 13169 11242 9609 8206 7007 6034
PV cumulative CF -897 6863 22299 35468 46710 56319 64524 71532 77566
References
1. Eurofound, European Monitoring Centre on Change (EMCC),
http://www.eurofound.europa.eu/emcc/.
2. Cefic, Facts & Figures 2011, The European chemical industry in a worldwide
perspective, http://www.cefic.org/Global/ Facts-and-figures-images/Graphs
2011/FF2011-chapters-PDF/ Cefic_FF Rapport 2011.pdf.
3. European Commission – Communication on an Industrial Competitiveness Policy
for the European Union (EU), http://
ec.europa.eu/enterprise/sectors/chemicals/documents/ reach/archives/white-
paper/background/communication/ summary_en.htm.
5. G. H. Vogel, Process Development: From the Initial Idea to the Chemical
Production Plant, Wiley-VCH, Weinheim, Germany, 2006.
6. D. N. Sull, Made in China: What Western Managers Can Learn from Trailblazing
Chinese Entrepreneurs, Harvard Business School Press, Boston, 2005.
7. P. H. Thomas, Windows of Opportunity: 21 Steps to Successful Selling, Key Porter
Books, Toronto, 1984.
8. A. Cybulski, J. A. Moilijn, M. M. Sharma, and R. A. Sheldon, Fine Chemicals
Manufacture: Technology and Engineering, Elsevier Science B.V., Amsterdam,
2001.
132
References
9. C. Bramsiepe, S. Sievers, T. Seifert, G. D. Stefanidis, D. G. Vlachos, H. Schnitzer,

B. Muster, C. Brunner, J. P. M. Sanders, M. E. Bruins, and G. Schembecker,
Chemical Engineering and Processing: Process Intensification, 2012, 51, 32–52.
10. T. Bott and G. Schembecker, Jahrestreffen PAAT-Fachgemeinschaft 2009,
Weinheim, http://www.processnet.
org/processnet_media/Veranstaltungen/48Tutzing/
die+50prozent_idee+vortrag+bott_schembecker.pdf.
11. T. Seifert, S. Sievers, C. Bramsiepe, and G. Schembecker, Chemical Engineering
and Processing: Process Intensification, 2012, 52, 140–150.
12. C. Bramsiepe and G. Schembecker, Chemie Ingenieur Technik, 2012, 84, 581–587.
14. Ł. Hady and G. Wozny, Chemie Ingenieur Technik, 2012, 84, 597–614.
15. A. Brodhagen, M. Grünewald, M. Kleiner, and S. Lier, Chemie Ingenieur Technik,
2012, 84, 624–632.
16. J. Rottke, F. Grote, H. Fröhlich, D. Köster, and J. Strube, Chemie Ingenieur
Technik, 2012, 84, 885–891.
17. S. Shah, Chemical Engineering Progress, 2007, 103, 36–41.
19. Chemistry on Wheels, Evonik Industries,
http://corporate.evonik.com/en/products/product-stories/pages/chemistry-on-
wheels.aspx.
22. Wikipedia, Time value of money, http://en.wikipedia.org/wiki/
Time_value_of_money.
23. R. Sinnott and G. Towler, Chemical Engineering Design, Elsevier Ltd, Oxford, 5th
edn., 2008.
24. D. Schmalz, M. Häberl, N. Oldenburg, M. Grund, H. Muntermann, and U. Kunz,
Chemie Ingenieur Technik, 2005, 77, 859–866.
25. D. M. Roberge, L. Ducry, N. Bieler, P. Cretton, and B. Zimmermann, Chemical
26. U. Krtschil, V. Hessel, D. Kralisch, G. Kreisel, M. Küpper, and R. Schenk,
CHIMIA International Journal for Chemistry, 2006, 60, 611–617.
27. V. Hessel, C. Hofmann, P. Lob, J. Lohndorf, and H. Lowe, Chem. Eng. Trans.,
2005, 6, 49–54.
28. V. Hessel, U. Krtschil, P. Lob, H. Lowe, D. Kralisch, G. Kreisel, M. L. Kupper,
and R. Schenk, Proceedings of the 2006 AIChE Annual Meeting, San Francisco,
CA, November 12–17, 2006.
133
30. M. S. Peters, K. Timmerhaus, and R. West, Plant Design and Economics for
Chemical Engineers, McGraw-Hill, Columbus, 2004.
32. E. Ritzer and R. Sundermann, in Ullmann’s Encyclopedia of Industrial Chemistry,
Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, Germany, 2000.
34. F. Benaskar, V. Hessel, U. Kxtschil, P. Lob, and A. Stark, Organic Process
36. R. Sinnott, Coulson and Richardson’s Chemical Engineering, Volume 6: Chemical
Engineering Design, Elsevier B.V., Oxford, 2005.
37. ICIS Indicative chemical prices, http://www.icis.com/ chemicals/channel-info-
chemicals-a-z/.
38. Engineering News Record (ENR), Economics – current costs,
http://enr.construction.com/economics/current_costs/.
39. K. Sato, M. Aoki, and R. Noyori, Science, 1998, 281, 1646–1647.
Transactions, 2012, 29, 565–570.
41. Sigma Aldrich, http://www.sigmaaldrich.com.
42. WebMD, http://www.webmd.com.
43. P. J. Harrington and E. Lodewijk, Organic Process Research & Development,
1997, 1, 72–76.
44. Voigt Global Distribution, Fine chemicals, http://www.vgdusa.
com/chemicals_n.html.
134
CHAPTER 6
Decentralized Biofuel Production –
Modular Process Integration at
Commercial Scale
Vural Gursel, I., Wang, Q., Noël, T., Hessel, V., Kolb, G. & van Veen, A. (2014). Heat-
integrated novel process of liquid fuel production from bio resources – Process simulation
and costing study. Chemical Engineering Transactions, 39, 931-936.
Abstract:
Innovative ways of providing novel sustainable solutions to reduce the emissions and
consumption of exhaustible resources is called for. Production of liquid fuels from
bioresources satisfies this call and it is aimed to be achieved with BIOGO project. Proposed
process involves coupled reforming of biogas and bio oil followed by methanol production
which is then converted to gasoline. This novel route through methanol eliminates
challenging separation steps of Fischer-Tropsh process. Decentralized production in
modularized small scale plants with process intensified equipment is considered.
Preliminary techno-economic analysis is carried out here to analyze the feasibility of the
process. Process simulation reveals that carbon efficiency of 56 % and energy efficiency of
42 % can be achieved. Cost calculations show that with the total cost of production of
gasoline at € 0.5/L; it is economically competitive with petroleum-based production and
more economical than Fischer-Tropsh process. Comparison of 35 decentralized small
modular plants with one centralized large scale plant reveals that the reduced construction
time and operating cost of decentralized plants can outweigh the higher investment cost.
135
6. Decentralized Biofuel Production
6.1. Introduction
Concerns about the depletion of fossil fuel reserves, increased energy demand, society’s
pressure for clean and sustainable production make use of bioresources as energy source an
attractive solution. Although, recently there has been increasing research on renewable
energy sources, still around 90 % of the world energy needs are met with non-renewable
sources.1
Bioenergy is the largest source of renewable energy and can provide heat, electricity as well
as transport fuels. International Energy Agency foresees that biofuels contribution to total
transport fuel could increase from 2 % today to 27 % by 2050.1 The scenario suggests that
this will be possible by advanced biofuel technologies that are not yet commercially
available. These advanced (second generation) biofuels manufactured from various types of
biomass, which are not competitive to food production, can substitute current fossil derived
gasoline and diesel fuels.2 The two main catalytic processes for biomass to liquid fuel
production are Fischer-Tropsch (FT) and Methanol-to-hydrocarbons processes. FT is an
established process already applied on large scale worldwide in coal and natural gas based
plants.2 There is one commercial methanol to gasoline plant using methanol produced from
natural gas.3 Methanol to olefins process is also developed. A further development is the
Mobil olefins to gasoline/diesel process. They are at demonstration phase.3
BIOGO, a new project funded by the European commission, proposes to combine bio oil
and biogas sources to produce syngas which is then converted to methanol. The methanol
produced as intermediate is finally transformed into liquid fuels with a sharper product
distribution than the ones from the FT process. The novelty of this route comes also from
coupled syngas generation. Decentralized production in modular plant environment
utilizing flow reactors is considered. In this chapter, results from the preliminary study done
to analyze feasibility of the proposed novel process is presented with process simulation
and cost analysis.
6.2. Methodology
Important problems associated with synthesis from biomass are; it is widely distributed, and
its composition and amount differs with location and season. Due to its large water content
its transportation is costly. Also large stocks of biomass needs to be stored near the
production plant because of its low energy density. Therefore, decentralized fuel production
is found preferable in this project. Moreover, production in modular plants is taken to
benefit from feed flexibility, scalability and faster time-to-market.4 They are formed by pre-
manufactured modules assembled into highly functional plant environment that facilitates
piping, utility, control and safety requirements in its infrastructure.5
For ease of transportation and storage, biomass is usually densified by torrefaction or
pyrolysis.6 With fast pyrolysis biomass is converted to liquid bio oil.7 Bio oil is up to 10
136
times the energy density of biomass and it has diverse chemical composition, typically with
low hydrogen (7 %) and high oxygen (45 %) content.8 Another bio source of energy is
biogas. Biogas is produced by decomposition of organic matter mainly of plant and animal
wastes or landfills. It is primarily methane and carbon dioxide along with other trace gases.9
For liquid fuel production, these bioresources need to be converted to syngas. The two main
syngas production processes are partial oxidation and steam reforming. Conventionally
steam reforming occurs in catalytic fixed bed reactors placed in direct-fired furnaces.10
Long contact times are required due to heat transfer limitations. In this project coupling of
exothermic partial oxidation with endothermic steam reforming in a microchannel plate
heat exchanger reactor is considered.11 This reactor design provides intensive mixing and
enhanced mass and heat transfer that allows very short contact times.12
Fischer-Tropsch (FT) synthesis from syngas irrespective of operating conditions cannot
produce uniquely diesel and gasoline. Therefore, FT product upgrading is required that
involves challenging separation steps of many side products. Methanol produced from
syngas can either be converted to gasoline or olefins.3 Product upgrading is simpler. There
is significantly less gaseous products and light saturates and also no heavy products that
require cracking. BIOGO project is formed accordingly for decentralized liquid fuel
production from bioresources of biogas and bio oil through methanol synthesis in
modularized small scale plants with process intensified equipment including microreactors.
To analyse the feasibility of the proposed project, a preliminary techno-economic analysis
is performed. First, the process scheme is developed and a mass and energy balance is
done. This is carried out with Aspen PlusTM modeling. Production capacity of 117.6 t/d
gasoline is selected according to the production capacity of a small scale plant.13 In this
way the viability of the proposed process scheme can be shown. Second, cost analysis is
done by first estimating capital and operating costs. Accordingly cost of production of
gasoline is calculated for rate of return of 10 % and 20-year plant life. This can assist to
compare this route with values from FT synthesis and conventional fossil based production.
Also, the economics of decentralized small scale production is assessed in comparison with
large scale plant with a discounted cash flow analysis.

6.3.1. Process simulation
The preliminary study to assess the BIOGO project is started with a process simulation
done with Aspen PlusTM. Biomass handling and drying, pyrolysis, biogas cleaning, air
separation and wastewater treatment are not included in the process simulation but are
taken into account in cost calculations. The process flow diagram of generation of liquid
fuel from biogas and bio oil is given in Figure 6.1.
137
138
Figure 6.1. Process flow diagram of liquid fuel production from biogas and bio oil
Biogas is fed together with bio oil from fast pyrolysis to the reformer in 4:1 mole ratio.
High feed temperature is required to achieve a high yield of syngas and to inhibit coke
formation. The reformer temperature is set to 900 °C and pressure to 45 bar to achieve high
methane conversion. Coupled reforming of the two streams enables high hydrogen content
of the biogas to be utilized by bio oil. The rapid and intimate heat transfer eliminates coke
formation. Pure oxygen is input to the reformer to provide the heat required for the
reforming reactions. Its amount is adjusted to limit formation of carbon dioxide (CO2)
instead of carbon monoxide (CO). Steam is manipulated to achieve a hydrogen to carbon
monoxide (H2/CO) ratio of two. Therefore, further reverse water-gas-shift reaction is not
required. The reformer outlet gas is separated from water, carbon dioxide and methane.
Water can be recycled after treatment to the reformer. The methane left after reforming is
sent to a combustion unit with oxygen to recover energy.
The syngas composed of H2 and CO is fed to methanol synthesis. This is an equilibrium
reaction with two moles of H2 and one mole of CO converted to one mole of methanol. The
reactor temperature and pressure are set to 250 °C and 45 bar respectively. Recycle is
employed to maximize yield of methanol. Because of water-gas-shift reaction, some CO is
converted to CO2 which needs to be taken into consideration. The syngas mixture is
typically adjusted to contain 4 – 8 mole % CO2 for maximum selectivity and a
stoichiometric ratio defined as H2 CO2 ⁄ CO CO2 of two is preferred for methanol
synthesis.2 The recycle is adjusted to get this stoichiometric ratio equal to two and the
combined syngas with recycle contains 6 mole % CO2. 10 % of the recycle syngas stream is
purged to prevent accumulation. This purge stream is used as fuel gas and fed to
combustion unit.
The methanol product has a composition of 2.8 mole % water and 97.2 mole % methanol. It
is fed to the methanol to gasoline reactor. The inlet conditions are adjusted to 350 °C and
20 bar.2 The methanol is converted to 56 wt. % water and 44 wt. % hydrocarbons. 100 %
methanol conversion is assumed. The hydrocarbon product composition is modelled based
on typical composition in literature.3 Hydrocarbon composition is 1.5 wt. % light gas, 5 wt.
% C3, 11.5 wt. % C4, and 82 wt. % C5+. The crude hydrocarbons are separated into
finished liquid fuel products of which 86 wt. % is gasoline, 14 wt. % is LPG and the
balance is fuel gas.
Using methanol as intermediate ensures a high conversion to liquid fuels. The light gas
formation is much lower than FT synthesis. These light gas streams would need to be
recompressed and recycled leading to extra energy requirement and loss of sellable product.
A process route via methanol therefore avoids challenging gas separations and recycle
which are required for conventional FT route. Thus, a more compact process scheme is
achieved.
139
6.3.2. Mass, carbon and energy balance

Mass, carbon and energy balance from simulation is presented in Table 6.1. The inputs to
the process simulation are biogas, bio oil and oxygen while the outputs are gasoline, LPG,
CO2 and water. Although there is also water inlet to the system, the net balance of water is
written on the table indicating outlet water streams are more than enough to satisfy the inlet
water by recycle. 100 % balance is attained. In the carbon balance carbon is input to the
process by biogas and bio oil and exits the process via gasoline, LPG and CO2. The carbon
conversion efficiency to liquid fuels is calculated as 55.8 % which is slightly higher than
reported in previous biomass to liquid studies.13 This is mainly due to reduced separation
requirement of the product in this process enhancing carbon utilization. 100 % of carbon is
accounted for. Energy balance is given in Table 6.1 on a LHV basis. The energy efficiency
of the process to liquid fuels is calculated by dividing the total energy output of gasoline
and diesel by the total energy input to the process from biogas and bio oil. This energy
efficiency is calculated as 41.8 %.
Table 6.1. Overall mass, carbon and energy balance
Mass Balance Carbon Balance Energy Balance

t/d kmol/h MW
Inputs
Biogas 357.3 484.7 108.1
Bio oil 162.5 230.0 61.6
Oxygen 453.0 0.0 0.0
Outputs
Gasoline 117.6 343.3 61.0
LPG 19.2 55.3 10.1
CO2 501.5 316.1 0.0
Water 334.5 0.0 0.0
6.3.3. Cost analysis

The sections of the plant include biomass pyrolysis, coupled reforming, methanol synthesis,
methanol to gasoline (MTG) conversion process. Combustion unit is used to recover heat.
Oxygen is produced on site with air separation unit. Water is recycled inside the unit with
water treatment. Water treatment unit also includes steam turbine to generate electricity
from waste heat in the process. Therefore, system is self-sufficient without external utility
requirement. The direct costs of the units are calculated using estimates from several
literature sources. Biomass pre-processing, fast pyrolysis and combustion unit costs are
estimated using a scaling factor of 0.7 to values from Wright et al.14 The other unit costs are
estimated using cost parameters given in Baliban et al.15,16 Chemical engineering plant cost
140
index is used to bring the unit costs to Q4 2013 (567.3). For conversion to euro from dollars
1.33 exchange rate is taken. The total direct cost includes cost of equipment, installation,
instrumentation, piping and service facilities. The indirect costs of engineering and
construction expenses and contingency are expected to be lower with modular plants since
engineering and construction time and expenditure is much reduced with standardization
and pre-manufacturing. Accordingly, indirect costs are estimated to be 50 % lower than
conventional plants. The fixed capital investment is calculated as sum of total direct cost
plus the indirect cost as given in Table 6.2. This is converted to annual basis to be
incorporated into cost of production. Annual capital charge is calculated by multiplying the
investment with a factor that depends on rate of return and plant life. For the selected rate
of return of 10 % and plant life of 20 years the factor used is 0.117.17 The other
contributions to total cost of production come from raw materials, operating & maintenance
(O&M) cost and LPG. LPG is sold as by-product so indicated as negative value. The cost
parameters used in calculation are also given in Table 6.2.
Table 6.2. Breakdown of investment cost (left), total cost of production and cost
parameters (right)
MM€ MM€/yr €/L

Biomass pre-processing 3 Biomass 5 0.10
Biomass fast pyrolysis 5 Biogas 11 0.21
Combined reforming 20 O&M 3 0.07
Syngas separation 7 Capital charge 10 0.20
Methanol synthesis 5 LPG -4 -0.07
MTG process 7 Total Cost of
LPG - gasoline separation 3 Production 26 0.50
Combustion 13
Air separation 9
Wastewater treatment 2 Item Cost parameters
Total Direct Cost 75 Biomass € 62.0per dry ton
Indirect Cost 12 Biogas € 0.12per Nm3
Fixed Capital Investment 87 LPG € 0.32per litre
The total cost of production is calculated both in MM€/yr and in €/L using gasoline lower
heating value of 32 MJ/L. The latter value represents the price of gasoline from which this
process becomes economically competitive. The current gasoline price is 0.56 €/L (given
regular gasoline spot price of 2.8 $/gal18) is higher than 0.50 €/L estimated in this study
indicating this process can be competitive with petroleum-based process. It should be noted
that the estimations are sensitive to the cost parameters selected in the study. Baliban et al.
141
found a similar value for MTG process (0.51 €/L gasoline) and a higher value for FT
process (0.59 €/L gasoline equivalent) at same production capacity.13 The FT process cost
of production is higher mainly due to higher investment cost required for this process with
higher separation demand.
Higher capacities up to 50 times the capacity used in this study (5,880 t/d gasoline) was
also studied. The investment and fixed operating cost information of this large scale plant is
taken from Baliban et al.13 to make comparison of decentralized plants with a centralized
large scale plant. Due to seasonality large scale plants can’t run at full capacity. 70 %
capacity usage is assumed in this study, so one large plant of 5,880 t/d gasoline total
capacity is compared with 35 plants of 117.6 t/d gasoline capacity each. Large scale plants
collect biomass typically from 80 km radius so the associated transportation cost is added to
the biomass cost giving 75 €/t. One year construction time is taken for small-plants and
three year for the large plant. The plant-life is taken as 20 years. The resulting net present
value (NPV) calculation is given in Figure 6.2.
Figure 6.2. Net Present Value of decentralized small-scale plants and centralized large
plant
The investment cost is higher for decentralized plants because of economies of scale.
However, due to lower construction time, financial gain is earned earlier and due to time
value of money this gives higher NPV. Also, operating cost is lower due to no added
transportation cost for biomass. It is seen that these positive effects can outweigh the
negative effect of higher investment requirement.
142
6.4. Conclusion
6.4. Conclusion
In the proposed novel production plant process integration (reaction coupling), process
simplification (elimination of complex separation steps), modular processing and process
intensification (microreactor utilization) is achieved. Syngas can be generated with coupled
reforming at a stoichiometric ratio suitable to directly feed to methanol synthesis while
simultaneously avoiding coke and minimizing methane formation. Neither challenging gas
separation nor extensive light gas recycle loops are required in contrast to Fischer-Tropsch
based routes. Therefore a compact process design is attained. Carbon efficiency to liquid
fuels of 56 % and energy efficiency to liquid fuels of 42 % is achieved. Cost calculation
reveals that with the price parameters used this proposed novel process is economically
competitive with petroleum-based process and is more economical than FT process. Capital
cost reduction is possible due to lower requirement of product separation and upgrading,
also with lowered indirect costs with modular processing. Transportation which has
associated costs and emissions is reduced by decentralized production. Discounted cash
flow analysis show that decentralized plants can be preferable to large scale plants because
they can be built in shorter time with faster time-to-market. Also due to seasonality and
presence of variety of bioresources feed flexibility and capacity adaptation is important and
can be achieved with these plants Liquid fuels produced from bioresources cause lower
greenhouse gas emissions since carbon dioxide is captured by the feedstock and it offsets
carbon dioxide emission from burning fuel. The introduction of environmental regulations
such as carbon tax will improve the economics of biomass to liquid processes and can
enable the shift from fossil-based processes. This chapter gives an ex-ante analysis before
start of the BIOGO project activities and more coming insight can lead to a differentiated
picture.
Acknowledgements
Funding by the Advanced European Research Council Grant “Novel Process Windows –
Boosted Micro Process” under grant agreement number 267443 and funding by FP 7 EU
project BIOGO under grant agreement number 604296 are kindly acknowledged.
References
1. IEA, Technology Roadmap: Biofuels for Transport, OECD/IEA, Paris, 2011.
2. M. Iglesias Gonzalez, B. Kraushaar-Czarnetzki, and G. Schaub, Biomass
Conversion and Biorefinery, 2011, 1, 229–243.
3. S. Tabak and S. Yurchak, Catalysis Today, 1990, 6, 307–327.
4. I. Vural Gürsel, V. Hessel, Q. Wang, T. Noël, and J. Lang, Green Processing and
Synthesis, 2012, 1, 315–336.
& Technology, 2012, 35, 1184–1204.
143
6. E. Peduzzi, G. Boissonnet, G. Haarlemmer, and F. Marechal, Chemical

Engineering Transactions, 2013, 35, 553–558.
7. M. A. G. Figueiredo and F. L. Mendes, Chemical Engineering Transactions, 2012,
29, 691–696.
8. Q. Zhang, J. Chang, T. Wang, and Y. Xu, Energy Conversion and Management,
2007, 48, 87–92.
9. P. Pagliai and R. Di Felice, Chemical Engineering Transactions, 2012, 29, 871–
876.
10. J. R. Rostrup-Nielsen, in Catalysis Science and Technology, ed. J. R. A. a. M.
Boudart, Springer, Berlin, 1984, vol. 5, pp. 1–117.
11. G. Kolb, T. Braune, J. Schurer, and D. Tiemann, Chemie Ingenieur Technik, 2013,
85, 1619–1623.
2013, 6, 746–89.
13. R. C. Baliban, J. A. Elia, and C. A. Floudas, Energy & Environmental Science,
2013, 6, 267–287.
14. M. M. Wright, J. A. Satrio, R. C. Brown, D. E. Daugaard, and D. D. Hsu, Techno-
Economic Analysis of Biomass Fast Pyrolysis to Transportation Fuels, NREL/TP-
6A20-46586, 2010.
15. R. C. Baliban, J. A. Elia, and C. A. Floudas, Computers & Chemical Engineering,
2011, 35, 1647–1690.
16. R. C. Baliban, J. A. Elia, V. Weekman, and C. A. Floudas, Computers & Chemical
Engineering, 2012, 47, 29–56.
edn., 2008.
18. Energy & Oil Proces - Bloomberg, www.bloomberg.com/energy.
144
CHAPTER 7
Bulk Chemistry Process Simplification –
End-to-end Process Integration at
Commercial Scale
Vural Gursel, I., Wang, Q., Noël, T. & Hessel, V. (2012). Process-design intensification :
direct synthesis of adipic acid in flow. Chemical Engineering Transactions, 29, 565-570.
Vural Gursel, I., Wang, Q., Noël, T., Hessel, V. & Tinge, J.T. (2013). Improving energy
efficiency of process of direct adipic acid synthesis in flow using pinch analysis. Industrial
and Engineering Chemistry Research, 52(23), 7827-7835.
Wang, Q., Vural Gursel, I., Shang, M. & Hessel, V. (2013). Life cycle assessment for the
direct synthesis of adipic acid in microreactors and benchmarking to commercial process.
Chemical Engineering Journal, 234, 300-311.
Abstract:
Synergy of micro dimensions and harsh operating conditions results in a three-step
intensification within a reaction system: transport, chemical and process-design
intensification. Transport intensification that improves mass and heat transfer has been
vastly demonstrated. Emerging chemical intensification uses highly intensified, unusual and
typically harsh process conditions to boost micro-processing (high-T, high-p, high-c,
safety). Process-design intensification is a new field in micro processing and it heads for
integrated and simplified smart-scaled (micro-meso) flow process design in a holistic
picture. The latter two constitute Novel Process Windows. As a demonstration example,
direct adipic acid synthesis is considered which is an important intermediate for the nylon
industry. It provides an innovative alternative to the conventional process currently used.
The conventional process occurs in two steps from oxidation of cyclohexane by air
followed by nitration oxidation. The direct route is from cyclohexene and uses hydrogen
peroxide as oxidant. This results in higher adipic acid yield and simplified process.
Drawbacks of the direct synthesis are long reaction time and increased safety issues, which
145
7. Bulk Chemistry Process Simplification
can be overcome by using microreactors. The reaction rate is increased by the largely
improved mass transfer and the use of higher temperature. Based on laboratory flow
experiments a full-chain process simulation was made for the direct route using Aspen Plus.
This was used to make cost, energy and environmental analysis of this new process.
Moreover, comparison with the two-step process was made to benchmark the flow process
against conventional technology. Profound simplification of the process scheme of direct
process leads to fewer unit operations enabling reduction of investment cost, although the
microreactor cost is higher than those of conventional reactors. Through simplification or
elimination of energy intensive separation units energy requirement can be reduced
significantly. Life cycle assessment (LCA) shows that for a number of impact categories
the direct process is greener; yet there are also categories for which the conventional route
is more environmentally sustainable.
7.1. Introduction
Microreaction technology has attracted much attention in last 20 years because of its high
capability in mass transfer and heat transfer, as well as its process window widening, e.g. to
allow safe operation even in explosive regimes or with toxic or hazardous reactants or
intermediates.1–3 Nearly all microreactor studies are on the synthetic/reaction engineering
level; the process design aspect, especially for a larger productivity scale (>1000 t/y) is
missing in scientific literature. In such context, this chapter studies how microreaction
technology donates to the end-to-end process integration for a bulk chemical process.
Process-design intensification through microreaction technology is new in literature and it
is studied for the first time at the example of direct adipic acid synthesis.4,5 Process-design
intensification benefits from interdisciplinary integration of chemistry, micro process
technology and process engineering. A step change improvement can be realized via
bringing new chemical transformations such as direct synthesis on a full-process level.
Such direct chemistries have the potential to reduce the cascades of separation units
considerably. A decreased number of apparatus combined with a decreased size results in a
smaller plant foot-print. The new flow-based process design then can have an entirely new
cost, energy and sustainability structure.
Adipic acid is an important precursor for the production of nylon 6,6.6 According to the
market report7, the global synthetic adipic acid market was worth USD 4,898.5 million in
2010 and is expected to reach USD 8,063.1 million in 2018. The conventional way to
produce adipic acid comprises the two-step catalyzed air / nitric acid oxidation of
cyclohexane (see Figure 7.1, top).8 The first step involves the production of so-called KA
oil (a mixture of cyclohexanone, the ketone or K component, and cyclohexanol, the alcohol
or A component). In general, KA oil is produced by air oxidation of cyclohexane. In order
to ensure the high selectivity to KA oil of over 75 % in the first air-oxidation step, it is
146
7.1. Introduction
necessary to keep the conversion of cyclohexane very low, to 4-6%.9 To cope with that,
capital and energy intensive distillation equipment is needed to separate and recycle
unconverted cyclohexane from the product stream. The second stage involves oxidation of
the KA oil with nitric acid where adipic acid is obtained in 92-96 % yield.10 The high
concentration (50-60 wt. %) and large consumption of nitric acid causes serious problems,
since nitric acid is highly corrosive.10 Moreover, nitric acid oxidation produces significant
quantities of nitrous oxide (N2O) and NOx (NO, NO2, and higher oxides) gases.11
Accordingly abatement technologies are adapted to reduce these emissions to the
environment. Generally thermal decomposition is used where high amount of energy is
required.12 Despite the efficient recovery still large amounts are emitted. N2O emission
from this process, which is considered to cause global warming and ozone depletion13,
corresponds to 5 to 8 % of the total amount released by man worldwide.14
Figure 7.1. Reaction scheme of commercial two-step route and direct route for adipic acid
synthesis
Accordingly, an ideal process would proceed at higher conversion and without the use of
nitric acid. The solution can be a selective direct process with an oxidant that causes no
emission to be removed. A direct route by the one-step all-air oxidation of cyclohexane is
researched for this purpose15–18, yet this is so far not in commercial use.
In virtually the same manner and now using another oxidant, a direct route was realized
through the oxidation of cyclohexene with hydrogen peroxide in batch or membrane
reactors19–21 (see Figure 7.1, bottom). This direct route is process simplified (one-step) and
has higher yield as compared to the conventional route. Also water is produced as by-
product instead of nitrous gases in conventional process. Sato et al. described the direct
oxidation of cyclohexene to adipic acid by 30% hydrogen peroxide (H2O2) with Na2WO4
and [CH3(n-C8H17)3N]HSO4 as phase-transfer catalyst.19 This aqueous, organic biphasic
reaction was seen to enable high yield (93 %) in batch experiments under conditions that
are entirely free of organic solvents and halides.19,22 Jiang et al. studied alternative phase
147
transfer catalysts due to [CH3(n-C8H17)3N]HSO4 being expensive and tedious to synthesize.

Deng et al. reported the use of peroxytungstate-organic complex catalyst for the direct
oxidation of cyclohexene to adipic acid without any phase transfer catalyst.20
It has to be noted that the terminology “direct synthesis” is only valid in the strict sense at a
lab-scale. On an industrial level, cyclohexane and cyclohexene are produced from the
hydrogenation of benzene comprising an additional step as shown in Figure 7.1.
This is a highly exothermic reaction and, in co-amplification with the much accelerated
decomposition of H2O223, heat management is especially demanding at high process
temperatures and high production capacity. For these demands for intensified mass/heat
transfer and safety operation, batch reactors are not optimal; instead another type of reactor
which can safely handle such highly exothermic and severe reaction conditions is required.
Given that profile, one largely preferred solution is the use of microreactors.24
Recently, in our group we showed, for the first time, the feasibility of carrying out the
direct oxidation of cyclohexene with H2O2 to produce adipic acid in continuous micro-
flow.25 The reaction rate is significantly increased in the micro-flow consisting of a milli
packed bed reactor with micro interstices between the beads forming a microchannel
network. The yield improved from upto 50% by increasing the reaction temperature from
90°C to 100°C. But with a further increase of temperature from 100°C to 110°C, the yield
decreased. Decomposition of H2O2 at elevated temperature lowers the concentration of
H2O2 which decreases reaction rate and yield.
The yield of 50% is satisfactory, but not so high to simply give a straightforwardly intuitive
and unequivocal argument to commercialize such process, while the H2O2 decomposition
adds further problem. For such situation, holistic guidance in terms of economic, energy
and environmental analysis are needed in order to determine the process feasibility and
what needs to be done to optimize it. Accordingly, in this study process simulation of the
direct route and two-step conventional route was carried out. Comparison of the two routes
were made based on cost of equipment and energy requirement calculation and
environmental impact determination by means of life cycle assessment (LCA).
One of the important goal is to facilitate finding the right path from the starting point of the
process development curve towards a new full-chain continuous flow process design, by
implementing the needs of the latter from the first moment of investigations. Holistic
process route selection is becoming more and more important, as largely increasing number
of holistic process papers in the scientific engineering literature over the last 8 years
shows.3 We see such investigation also as aid for the industrial management and it has
accordingly entered the decision making of bulk-chemical industry.26
148
7.2. Methodology
7.2. Methodology
7.2.1. Process model development
Since, the direct route is not yet commercially applied for adipic acid synthesis, a flow
diagram and the process conditions were not available – this is exactly where innovation
through the new flow process design has to set in. The reaction characteristics of this
process based on laboratory experiments were considered together with the downstream
equipment of the nitric acid oxidation process (second step of commercial route) to propose
the full-chain process for the direct route.
In the direct oxidation process, cyclohexene, hydrogen peroxide together with sodium
tungstate and phase transfer catalyst are fed to the oxidation reactor. The biphasic reaction
takes place at 90°C and 1 bar. Cyclohexene is converted to water soluble adipic acid. The
reactor effluent goes through a separator to remove the organic phase containing unreacted
cyclohexene and by-products. The resulting aqueous stream is treated analogous to reactor
effluent stream in the nitric acid oxidation process to recover adipic acid. First it is fed to a
crude crystallizer operated at 50°C. The crystals are removed by a filter. And the mother
liquor is sent to concentrating still to remove some of the produced water. The remaining
stream goes through a purge crystallizer operated at 30°C to recover any remaining adipic
acid. The purity of the crude adipic acid is high (99%) therefore recrystallization with water
done for nitric acid oxidation process is not required.27 The combined crystals are dried and
the final product is attained.
Process simulations were done using Aspen PlusTM for the commercial route (two-step air
and nitric acid oxidation of cyclohexane) and for the direct synthesis route following the
proposed process. The commercial route was simulated with data from the industrial
production process from relevant patents and literature.6,8–10,28–32 The direct route is
simulated with data on the reaction based on laboratory flow experiments and for the
downstream equipment the properties proposed above is used. Conversion of cyclohexene
of 40 % based on primary result and 50 % based on current result were taken in simulation
of the direct route. Since the experiments are ongoing, it is likely that it can be improved
further. Therefore, also conversion of 98 % was simulated which reflects the best
performance achieved for the direct route in literature in batch experiments. The capacity
selected is 400 kton/year based on average adipic acid production plants.33
7.2.2. Energy analysis

The power, heating and cooling requirements of the processes were determined from the
process simulations that were carried out for both routes. Heat integration that is typically
employed in industrial plants is not considered here. Another chapter is devoted to heat
integration study for the direct route (Chapter 8). In this chapter the aim is to get an idea of
149
the energy requirements of the direct route and compare it with the requirements in
conventional process.
7.2.3. Cost analysis

For further comparison of the routes cost analysis was carried out. This was done so by
comparing the total purchased cost of equipment for the two routes based on the sizing
information from the process simulations done. Equipment costs for major equipment were
estimated using Aspen Process Economic Analyzer34, log-log plots of equipment costs from
books35–37 and internet38. Chemical engineering plant cost index was used to bring the unit
costs to Q4 2013 (567.3). For conversion to euro from dollars 1.33 exchange rate was
taken.
7.2.4. Environmental analysis – LCA methodology

Environmental sustainable development needs methods and tools to measure and compare
the environmental impacts as a result of the production processes. A preferred tool is LCA
which is a method for evaluating the environmental impacts of a material, product, process,
system or service throughout its entire life cycle holistically, i.e. from ‘cradle to factory
gate’, ‘cradle to grave’ or even from ‘cradle to cradle’. The LCA work in this study was
conducted with regard to standards of ISO 1404039 and its related guidelines40. The impact
categories considered in this study are based on CML2001 method41.
Figure 7.2. System boundary of a) direct synthesis, b) conventional synthesis (the dotted
lines indicate system boundaries, full process equipment such as distillation not shown for
reason of graphical simplicity)
150
LCA as considered in this study refers to a “cradle to factory gate” analysis, which means
starting from raw materials and end up with final products. The boundaries for the direct
micro-flow route synthesis and two-step conventional synthesis are shown in Figure 7.2.
The inventory data were adopted from the Ecoinvent v2.2 database incorporated in
Umberto 5.6. The functional unit is “1 kg adipic acid”. The by-products produced in the
process were treated as waste, so no allocation was needed in this study. Only the material
flow and energy flow were considered while the transportation of raw materials was not
considered in the study. The mass and energy flows are collected from process simulation
of the two routes for inventory analysis.
Life cycle impact assessment is a means of examining and interpreting the inventory data
from an environmental perspective. The impact categories42 considered here is based on the
CML2001 method41 incorporated in Umberto 5.6. The impact categories in the present
study are: acidification potential, average European (AP), global warm potential: climate
change in 20 years (GWP 20a), eutrophication potential, average European (EP), freshwater
aquatic ecotoxicity in 20 years (FAETP 20a), human toxicity in 20 years (HTP 20a), marine
aquatic ecotoxicity in 20 years (MAETP 20a), photochemical oxidant creation potential
(POCP), depletion of abiotic resources, terrestrial ecotoxicity in 20 years (TAETP 20a).
Since the inventory data for cyclohexene is not available in database, a sub process to
produce cyclohexene by the partial hydrogenation of benzene is built.43 In this study, an
ideal conversion with 100 % and selectivity with 85 % are assumed in order to simplify the
LCA analysis. In the real process, however, to achieve high selectivity low conversion is
required. So the environmental profile does not include the energy consumed by separation
and recycle of unreacted benzene. Under such simplification of the sub-process, the
proportion of cyclohexene contribution to the impact categories will be lower than that in
real case.

7.3.1. Process simulation
As explained above process simulations were done using Aspen PlusTM for the
conventional route (two-step air and nitric acid oxidation of cyclohexane) based on the
industrial production process and for the direct route following the proposed process. The
process flow diagram of the conventional synthesis is given in Figure 7.3 and process flow
diagram of the direct synthesis is given in Figure 7.4 for 50% conversion of cyclohexene
case. As can be seen from the process flowsheets of the two routes, the direct route leads to
a more compact plant design. With the reduction in step in synthesis there is requirement of
fewer process units: only one reactor is needed and energy intensive separation units are
either eliminated or simplified.
151
152
Figure 7.3. Process flowsheet of Aspen Plus simulation of conventional synthesis

Figure 7.4. Process flowsheet of Aspen Plus simulation of direct synthesis at 50 % conversion
153
In the conventional route there is requirement of 5 reactors (air oxidation, decomposition,

nitric acid oxidation, N2O and NOx decomposition reactors) whereas in the direct route
there is requirement of only one which is the oxidation reactor. The purity of the crude
adipic acid is found high in direct synthesis27 therefore recrystallization is not required
omitting redissolving and refined crystallization units. The low cyclohexane conversion in
the conventional route demands high recycle ratio, whereas in the direct flow route
conversions per pass are much higher. In the conventional route also NOx and N2O
treatment is required. With water being the only by-product in the direct route waste
treatment is simplified as well.
7.3.2. Energy requirements

The power, heating and cooling requirements of the conventional route and direct route are
determined based on the process simulation. Also the effect of change in conversion for the
direct route on the energy requirements is analyzed by making simulations at conversion
levels of 40%, 50% and 98%. These are given in Table 7.1. Power requirement is the total
requirement of compressors and pumps. Cooling requirement is the total duty of the heat
exchangers that enable cooling. Heating requirement is the total duty of the heat exchangers
that enable heating.
Table 7.1. Power, cooling and heating requirements of conventional and direct route at
different conversion levels
Conventional Direct Route Direct Route Direct Route

Route 40% 50% 98%
Power requirement, kW 4,546 0 0 0
Cooling requirement, kW 29,258 42,189 32,299 13,281
Heating requirement, kW 69,277 60,105 48,859 26,773
Total, kW 103,081 102,294 81,158 40,054
With the elimination of compressors and working at atmospheric pressure the power
requirement for the direct route is 0. In real life for the transfer of liquids also power is
required but simulation only considers the power requirement when higher pressure
operation is involved requiring pressure differential.
For the conventional route, there is requirement of large heat input to reheat the large
amount of recycled cyclohexane to the reaction temperature due to low conversion by pass.
For the direct route, there is requirement of 4.4 moles of H2O2 per mole of cyclohexene.
Also H2O2 concentration of <60% is required because the use, storage and transportation of
higher concentration of H2O2 are not desirable for safety reasons.19 Therefore together with
the reactants and products large volume of water needs to be heated and cooled in the
process. There are large heating and cooling requirements for this reason. It is seen that for
154
the 40% conversion case, there is close total energy requirement with the conventional
route although there are much less process units involved. However, with the increase in
conversion the energy requirements are very much reduced. With the 50% conversion
achieved currently and with the ongoing studies for further improvement, it can be said that
direct route is more energy efficient than the conventional route.
7.3.3. Equipment cost estimation

A list of major equipment and their rough cost estimates for the two routes are provided in
Table 7.2. The costs for the direct route is representative of 40-50 % conversion. This table
depicts the difference of the two routes due to the simplification of the process design
achieved with direct route. In the list for the conventional route there are 22 major
equipment whereas in the list given for the direct route there are 10. It is also shows that the
capital investment can be reduced even with the use of more expensive microreactor
technology. The total purchase cost of equipment was seen to be cut approximately in half.
Table 7.2. List of major equipment and their cost estimates for the conventional route and
the direct route
Conventional Route Direct Route

List of Equipment Cost Estimates, M€ List of Equipment Cost Estimates, M€
Pump 1.00 Pumps 0.80
Compressors 7.50 Oxidation reactor 8.00
Oxidation reactors 1.75 Phase separator 0.75
Decomposition reactor 1.00 Distillation 1.00
Neutralization vessel 0.80 Crude crystallizer 1.75
Distillation 1.50 1st Centrifuge/Filter 1.60
Pump 0.40 Concentrating still 1.50
Oxidation reactor 1.50 Purge crystallizer 1.50
nd
Bleacher 0.80 2 Centrifuge/Filter 1.25
Compressors 6.30 Dryer 0.75
DeN2O reactor 1.50
DeNOx reactor 1.50
Crude crystallizer 2.25
1st Centrifuge/Filter 1.95
Concentrating still 1.50
Purge crystallizer 1.75
2nd Centrifuge/Filter 1.60
Distillation 1.75
Redissolving tank 0.45
Refined crystallizer 1.50
3rd Centrifuge/Filter 1.25
Dryer 0.75
Total Purchase Cost 40.30 Total Purchase Cost 18.90
155
In the conventional route there is requirement of costly compressors to compress the air to
be fed into the oxidation reactor and also for the transfer of nitrous gases to treatment. It has
been seen that elimination of the compressor in the direct route with the use of liquid
hydrogen peroxide as oxidant has a big impact in terms of costs.
Also, the reduction in number of steps in synthesis leads to lower number of downstream
units (separators). Correspondingly the total cost of equipment for separators has decreased
in comparison to the two-step route. Figure 7.5, left shows graphically the breakdown of the
costs of equipment. In this figure, the effect of elimination of compressors and the
reduction of downstream equipment in the direct route can be clearly seen.
It is seen that the total cost for reactors is similar for the two routes with microreactor being
used in the direct route. Therefore, application of microreactor technology with process-
design intensification is justified. Although flow reactors have higher cost than
conventional reactors applying them to a simplified process design does not result in higher
capital expenditure.
From Figure 7.5, right the share of costs can be seen within process equipment groups of
auxiliary equipment (pumps and compressors), separators (crystallizers, centrifuges,
distillation columns, separator vessels and dryers) and reactors. It is seen that for the
conventional route auxiliary equipment and separators have higher cost share (about 40 %)
than reactors (about 20 %). In the direct route with the elimination of compressors auxiliary
equipment have a low share. With the reduction in cost of separation equipment the share
of reactors is seen to be larger for the direct route (about 40 %) showing shift in plant
footprint toward more reaction dominated.
Figure 7.5. Total purchase cost of equipment (left) and share of equipment costs (right)
comparison of 2-step conventional and direct route
The analysis solidified the dream of pioneers in continuous small-scale plant design,
Benson and Rinard, who predicted that separation will be much reduced due to advances in
flow reactor performance. Rinard made detailed analysis of a micro-reactor plant concept
156
which he termed mini-plant production. Important criteria in this concept are JIT (just-in-
time) production, zero holdup, inherent safety, modularity and the KISS (keep it simple,
stupid principle) principle.44,45 Benson and Ponton proposed the concept of distributed
small-plant manufacturing referring to small scale, intrinsically safe and highly automated
plants46
7.3.4. Life cycle assessment

The environmental profile for the direct route is detailed for impacts that were mentioned
above in 7.2.4. (see Figure 7.6). Even at a glance it is obvious that the production of H2O2
dominates most of these impact categories. This is because H2O2 production is an energy
and pollution intensive process. 95% of all H2O2 is produced on an industrial scale by the
Anthraquinone oxidation (AO) process, which was developed by BASF. It involves the
sequential hydrogenation and oxidation of an Anthraquinone precursor dissolved in a
mixture of organic solvents followed by liquid–liquid extraction to recover H2O2.47 The
high energy input and generated waste has negative effect on environmental and production
costs.48 Also, the transport, storage, and handling of bulk H2O2 involve hazards. Moreover,
H2 as raw material for producing H2O2 is also made in an energy intensive process, mostly
by the steam reforming of methane to generate H2. It is seen that the H2O2 impact is
therefore much dominant (>50%) for all toxicity-related impacts (FAETP 20a, HTP 20a,
MAETP 20a and TAETP 20a).
Figure 7.6. Environmental profile of the direct synthesis 50% conversion (functional unit:
1 kg adipic acid) 1. AP 2. GWP 20a 3. EP 4. FAETP 20a 5. HTP 20a 6. MAETP 20a 7.
POCP 8. Depletion of abiotic resources 9. TAETP 20a. The line drawn separates raw
material- from process technology contributions.
The second most dominant factor for most impact categories is cyclohexene. Cyclohexene
has much lower impact on all toxicity-related impacts. It is relatively strong for the other
157
impacts. Cyclohexene and H2O2 are the raw materials, the other contributions can be
considered as process technology related. The strong raw material impact – clearly visible
by the line drawn in Figure 7.6 – is dominant at all impact categories.
For the two-step conventional synthesis, the dominating contribution for most of the impact
categories also comes from raw materials (see Figure 7.7). The nitrogen oxides is the
dominating parameter for AP and EP, while the waste treatment is the dominating
parameter for POCP.
Figure 7.7. Environmental profile of the two-step conventional synthesis (functional unit: 1
kg adipic acid) 1. AP 2. GWP 20a 3. EP 4. FAETP 20a 5. HTP 20a 6. MAETP 20a 7.
POCP 8. Depletion of abiotic resources 9. TAETP 20a. The line drawn separates raw
material- from process technology contributions.
So far shares of material- and process-related contributions have been compared in a
relative manner, normalized to 100%. In addition, a quantitative benchmarking between
process options will give insight about the total amount for each impact category. Besides
the 40% and 50% conversion levels taken from experimental results for the direct route, a
respectively higher conversion (98%) was also taken into consideration reflecting batch
conversion of direct route.
The comparison between the direct route (DR) and two-step conventional route (CR) shows
lower GWP for the direct route (see Figure 7.8). Because the recycling of unreacted
reactants is considered in the LCA, the GWP caused by the raw material consumption is
almost the same which is around 6 kg CO2-eq/kg adipic acid with different conversion of
cyclohexene from 40% to 98%. The difference in GWP performances of two routes
becomes even larger when including nitrous oxide emission in the total GWP sum-up
(denoted as CR2 in Figure 7.8). The GWP of emission 1 kg N2O equals to emission 270 kg
CO2 according to the eco-efficiency analysis by BASF.49 Then, the GWP for conventional
route increases from 6.7 to 16.7 kg CO2-eq/kg adipic acid.
158
Although from the GWP point of view the direct synthesis is better compared to the two-
step synthesis, it is also worse for some other impact categories. The human toxicity
potential (HTP) of the direct synthesis is found to be higher compared to the two-step route
(see Figure 7.9). In the direct route, the contribution of H2O2 to HTP is above 90%. In the
two-step conventional route, cyclohexane is the main contributor to HTP, while the nitric
acid has only small share. The total HTP of the conventional route is smaller than the direct
route. Improved conversion does not help here as explained in the discussion of GWP as
the unconverted raw material is always recycled.
Figure 7.8. GWP for direct route with different cyclohexene conversion (40%, 50% and
98%) and also for the two-step conventional route. DR: direct route, CR1: conventional route
without considering the emission of N2O, CR2: conventional route considering N2O emission
Figure 7.9. HTP for direct route with different cyclohexene conversion (40%, 50% and 98%)
and also for the two-step conventional route. DR: direct route, CR1: conventional route
159
Figures 7.8 and 7.9 show that H2O2 is not “green” as claimed by chemists22 because of its
comparable GWP contribution as cyclohexene and high contribution in HTP. It is green in
an atom economy and waste context (only water emission), but not in a more holistic view
“from cradle to factory gate”.
In order to compare the direct micro-flow and the two-step conventional routes holistically,
radar figures with normalized quantitative information are an appropriate presentation
approach (see Figure 7.10). The whole LCA sometimes is in (strong) favour for the direct
micro-flow process, sometimes for the two-step conventional process. It is a typical multi-
criteria decision. It is already seen from the findings given above; yet is spotlighted in
Figure 7.10 giving results from 9 impact categories. The direct route shows much better
results for the impact categories of AP, EP and POCP. The two-step conventional route has
visible better results in the impact categories of FAETP, HTP and MAETP.
It is also seen that even radical process improvements (such as increasing conversion up to
98%) cannot change the overall environmental profile and advantage/disadvantage of a
certain process. Exceptions are given when both the direct and conventional process have
similar data on one impact. Then the smart changes possible by process intensification in
direct route can have decisive character.
Figure 7.10. Comparison of all LCA impact categories for the direct route at different
cyclohexene conversion (40%, 50% and 98%) and the two-step conventional route. 1. AP 2.
GWP 20a 3. EP 4. FAETP 20a 5. HTP 20a 6. MAETP 20a 7. POCP 8. Depletion of abiotic
resources 9. TAETP 20a.
160
7.4. Conclusion
7.4. Conclusion
Adipic acid can be obtained by the direct synthesis from cyclohexene using H2O2 as
oxidant with chances for simplified process as compared to the two-step conventional
process used commercially in industry. Direct synthesis in batch, however, suffers from
long reaction time and safety concerns. That can be overcome by applying a micro-flow
reactor for which the yield was recently increased to 50% by processing at 100 °C in 20
minutes. For the direct micro-flow process still the yield needs to be much further
optimized and scale-up remains a not-tackled issue so far. Nonetheless, it is the ambition of
this chapter to answer in an early stage of process development (ex-ante), if such new direct
micro-flow process is feasible through economic, energy and environmental analysis.
Process-design intensification through microreaction technology is new in literature and it
is studied for the first time at the example of direct adipic acid synthesis.
In order to make comparison, process simulations were done using Aspen PlusTM for the
conventional route (two-step air and nitric acid oxidation of cyclohexane) based on the
industrial production process and for the direct route following the proposed process and
data on the reaction based on laboratory flow experiments. It is demonstrated that the direct
flow process leads to more compact plant design owing to the need of much less process
units (one reactor, less number of downstream units). Energy requirements of the two
routes are determined based on the process simulation. Direct route at different cyclohexene
conversion levels (40%, 50% and 98%) is considered. It is seen that due to low cyclohexane
conversion and requirement of compressors, the conventional route has high energy
requirement. Direct route at 50% conversion, which reflects result of current flow
experiments, has 20% lower energy requirement than the conventional route.
The capital investment is then much lower for the direct route as seen from the cost of
equipment calculations owing to the elimination of compressors and the reduction of
downstream equipment. The cost of the more advanced flow reactor is higher but it is
compensated by requiring only one in comparison with 5 reactors in conventional route.
With the cost of separation equipment being lower, the share of reactor costs is seen to be
larger for the direct route. Process-design intensification which is seen to enable reduction
in investment costs can thus lower the main decision barrier in the management of chemical
industry towards new technologies.
With LCA analysis it was seen that the production of H2O2 dominates most of the impacts
in direct route. It was found that the direct route is better in GWP, especially evident when
N2O-impact on GWP is considered. It also shows much better results for the impact
categories of AP, EP and POCP. Yet, the new proposed route has worse profile in HTP and
some other impact categories (FAETP, MAETP) compared to the conventional route. Thus,
the overall environmental picture is complex and there is not one better and greener
process.
161
This direct route will be economically and environmentally attractive when low cost H2O2
production methods become available. When considered from cradle to process gate, the
Anthraquinone process for H2O2 production is an energy and pollution intensive process.
Only by replacing it through an environmental sustainable process, which can be the direct
H2O2 synthesis out of the elements, the direct route will become “green” on all levels i.e.
creating a double-direct route to adipic acid. Such an in situ or on-site synthesis using H2
and O2 in industry can also reduce the cost of H2O2 production, making its use
economically favorable. Also changes in legislation and taxation policy regarding N2O
emission can be effective for the switch to this direct process.
References
2011, 66, 1426–1448.
2013, 6, 746–89.
& Technology, 2012, 35, 1184–1204.
Technik, 2012, 84, 660–684.
6. M. T. Musser, in Ullmann’s Encyclopedia of Industrial Chemistry, Wiley-VCH,
Weinheim, 2000.
7. Synthetic and Bio-Based Adipic Acid Market – Global Industry Analysis, Size,
Share, Growth and Forecast, 2012 - 2018. Transparency Market Research,
http://www.transparencymarketresearch.com/synthetic-and-bio-based-adipic-
acid.html.
8. A. Castellan, J. C. J. Bart, and S. Cavallaro, Catalysis Today, 1991, 9, 237–254.
9. M. T. Musser, in Ullmann’s Encyclopedia of Industrial Chemistry, Wiley-VCH
Verlag GmbH & Co. KGaA, Weinheim, Germany, 2000.
10. J. P. Oppenheim and G. L. Dickerson, in Kirk-Othmer Encyclopedia of Chemical
Technology, Wiley, New York, 2003, vol. 1, pp. 553–582.
11. Background Report AP-42 Section 6.2 Adipic Acid Production, U.S. Environmental
Protection Agency, http://www.epa.gov/ttnchie1/ap42/ch06/bgdocs/b06s02.pdf.
12. H. Mainhardt, N2O Emissions From Adipic Acid and Nitric Acid Production, Good
Practice Guidance and Uncertainty Management in National Greenhouse Gas
Inventories, The Intergovernmental Panel on Climate Change, http://www.ipcc-
nggip.iges.or.jp/public/gp/bgp/3_2_Adipic_Acid_Nitric_Acid_Production.pdf.
13. S. A. Montzka, E. J. Dlugokencky, and J. H. Butler, Nature, 2011, 476, 43–50.
14. C. Bolm, O. Beckmann, and O. A. G. Dabard, Angewandte Chemie International
Edition, 1999, 38, 907–909.
162
References
15. S.-S. Lin and H.-S. Weng, Applied Catalysis A: General, 1993, 105, 289–308.
16. S. Liu, Z. Liu, and S. Kawi, Korean Journal of Chemical Engineering, 1998, 15,
510–515.
17. D. Bonnet, T. Ireland, E. Fache, and J.-P. Simonato, Green Chemistry, 2006, 8,
556.
18. G. N. Kulsrestha, U. Shankar, J. S. Sharma, and J. Singh, Journal of Chemical
Technology and Biotechnology, 1991, 50, 57–65.
19. K. Sato, M. Aoki, and R. Noyori, Science, 1998, 281, 1646–1647.
20. Y. Deng, Z. Ma, K. Wang, and J. Chen, Green Chemistry, 1999, 1, 275–276.
21. M. G. Buonomenna, G. Golemme, M. P. De Santo, and E. Drioli, Organic Process
22. R. Noyori, M. Aoki, and K. Sato, Chemical Communications, 2003, 1977.
23. G. Goor, J. Glenneberg, and S. Jacobi, in Ullmann’s Encyclopedia of Industrial
Chemistry, Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, Germany, 2000.
2013, 36, 1001–1009.
26. R. B. Leng, M. V. M. Emonds, C. T. Hamilton, and J. W. Ringer, Organic Process
27. H. Jiang, H. Gong, Z. H. G. Yang, X. T. Zhang, and Z. L. Sun, Reaction Kinetics
and Catalysis Letters, 2002, 75, 315–321.
28. W. O. Bryan, Process for Preparing Cycloalkanols and Cycloalkanones, U. S.
Patent, US4238415, 1980.
29. U. N. Besmar, J. B. Lyon, F. J. Miller, and M. T. Musser, Preparation of
Cyclohexanone and Cyclohexanol, U. S. Patent, US4720592, 1988.
30. L. Fodor and B. C. Sutradhar, Process for Oxidation of Cyclohexane, U.S. Patent,
US6703529 B1, 2004.
31. F. Bende, K. Pohl, and H. Vollinger, Adipic Acid Recovery from Nitric Acid
Oxidation, U. S. Patent, US3476805 A, 1969.
32. O. A. Sampson, Preparation of dicarboxylic acids by nitric acid oxidation, U. S.
Patent, US3359308 A, 1967.
edn., 2008.
34. AspenTech, Aspen Process Economic Analyzer,
http://www.aspentech.com/products/aspen-icarus-process-evaluator.aspx.
35. R. Sinnott, Coulson and Richardson’s Chemical Engineering, Volume 6: Chemical
Engineering Design, Elsevier B.V., Oxford, 2005.
36. M. S. Peters, K. Timmerhaus, and R. West, Plant Design and Economics for
Chemical Engineers, McGraw-Hill, Columbus, 2004.
37. J. R. Couper, W. R. Penney, J. R. Fair, and S. M. Walas, in Chemical Process
Equipment, Elsevier, 1990, pp. 663–669.
163
38. Matches’ Process Equipment Cost Estimates,

http://matche.com/equipcost/EquipmentIndex.html.
39. ISO 14040, Envioronmental Management-life Cycle Assessment-Principles and
Framework, 2006.
40. European Commission – Joint Research Centre – Institute for Environment and
Sustainability: International Reference Life Cycle Data System (ILCD) Handbook –
General guide for Life Cycle Assessment – Detailed guidance, Publications Office
of the European Union, Luxembourg, 2010.
41. J. B. Guinée, M. Gorree, R. Heijungs, G. Huppes, R. Kleijn, A. Koning, L. Oers, A.
W. Sleeswijk, S. Suh, H. A. U. Haes, H. Bruijn, R. Duin, and M. A. J. Huijbregts,
Life Cycle Assessment, An Operational Guide to the ISO Standards, Final report,
http://media.leidenuniv.nl/legacy/new-dutch-lca-guide-part-1.pdf, 2001.
42. L. M. Tufvesson, P. Tufvesson, J. M. Woodley, and P. Börjesson, The International
Journal of Life Cycle Assessment, 2012, 18, 431–444.
43. H. Nagahara, M. Ono, M. Konishi, and Y. Fukuoka, Applied Surface Science, 1997,
121-122, 448–451.
44. I. H. Rinard, in Proceedings of the 2nd International Conference on Microreaction
Technology, IMRET 2, New Orleans, USA, 1998, pp. 299–312.
45. N. Saha and I. H. Rinard, in Proceedings of the 4th International Conference on
Microreaction Technology, IMRET 4, Atlanta, USA, 2000, pp. 327–333.
46. R. S. Benson and J. W. Ponton, Chemical Engineering Research and Design, 1993,
71a, 160–168.
47. J. M. Campos-Martin, G. Blanco-Brieva, and J. L. G. Fierro, Angewandte Chemie
(International ed. in English), 2006, 45, 6962–84.
48. G. Centi and S. Perathoner, Catalysis Today, 2009, 143, 145–150.
49. P. Saling, A. Kicherer, B. Dittrich-Krämer, R. Wittlinger, W. Zombik, I. Schmidt,
W. Schrott, and S. Schmidt, The International Journal of Life Cycle Assessment,
2002, 7, 203–218.
164
CHAPTER 8
Bulk Chemistry Heat Integration – End-
to-end Process Integration at Commercial
Scale
Vural Gursel, I., Wang, Q., Noël, T., Hessel, V. & Tinge, J.T. (2013). Improving energy
efficiency of process of direct adipic acid synthesis in flow using pinch analysis. Industrial
and Engineering Chemistry Research, 52(23), 7827-7835.
Vural Gursel, I., Wang, Q., Noël, T. & Hessel, V. (2013). Implementation of heat integration
for efficient process design of direct adipic synthesis in flow. Chemical Engineering
Transactions, 35, 775-780.
Abstract:
An energy efficient process is searched for the novel direct micro-flow route of adipic acid
synthesis introduced in the previous chapter. For systematically and holistically analyzing
the process and in particular its heat integration, pinch analysis is employed. Three
conversion cases of 40%, 50% and 98% is considered. With the use of the Aspen Energy
Analyzer the available energy recovery potential is determined and a new heat exchanger
network is designed with minimum total annual cost. Compared with the initial heat
exchanger network where energy requirements are supplied with utility streams, the
improved heat exchanger network designed enables 50-70 % saving in operating cost which
enables to pay back the extra capital cost requirement in less than one year in all cases. The
benefit of heat integration is seen to be reduced with increased conversion due to the
reduced heat load in the process. The heat exchanger selection is also very important design
consideration. The pinch analysis is restricted to shell and tube heat exchangers, however
the utilization of compact heat exchangers (including microchannel-based) which have
proven advantages in thermal effectiveness, safety and reduced size can enable further
165
8. Bulk Chemistry Heat Integration
benefits in terms of total plant cost and plant complexity. Also, the opportunity to have
lower operable temperature approach gives the possibility to lower the utility requirement
by 20%. As a consequence, for operation of flow processes using micro-meso reactors at
large scale, the utility/energy support and heat exchanger selection should be taken into
consideration. Such holistic thinking has not been detailed and justified so far for micro- or
similar smart-scale flow reactors, to our best knowledge.
8.1. Introduction
Process-design intensification is a new field in micro processing and it considers a
completely new process design that has a new cost, energy and sustainability structure.1–5
The investigation is started with process simplification at reaction level, which is given by
replacing multi-step by direct synthesis. The demonstration example is the direct oxidation
of cyclohexene with hydrogen peroxide for adipic acid synthesis. In previous studies it has
been shown that the direct route enables reduction of number of process units considerably
which results in the total purchase cost of equipment to be cut approximately in half and
energy requirements to be reduced compared with the two-step commercial route.3 Also life
cycle assessment revealed that direct route can not be considered totally green due to
energy and pollution intensive process of H2O2 production.4
Heat integration is typically employed in industrial plants to enable heat recovery by
matching hot and cold process streams. This chapter focuses on its application for the direct
route. In the microreactor literature it is common to consider one unit and intensify its
operation. However, we consider the chemical process as a whole in a holistic context.
Pinch analysis is taken here therefore to evaluate the entire process.6 It enables heat
integration by maximization of process-to-process heat recovery considering the whole heat
exchanger network.
In the process simulation a basic design approach is used where the heating and cooling
requirements are satisfied by the use of external utility streams such as LP steam and
cooling water respectively. By employing heat integration, reduction of utility loads and
consequently operating costs can be achieved. However, energy recovery always involves
additional heat exchangers and hence higher capital cost. Therefore, an optimization is
required so that total cost is minimized. The design of the aimed heat exchanger network is
not easy considering the fact that most processes involve a large number of process and
utility streams. Pinch analysis presents a powerful technique in solving these energy saving
problems.7 Initially developed by Linnhoff 8,9, this can be applied now with speed and
efficiency with the use of available software programs such as Aspen Energy Analyzer.10
In the present work, the effort of process-design intensification is combined with pinch
analysis to enable energy efficient simplified flow process design on production scale for
166
8.2. Methodology
the direct synthesis of adipic acid. Further, the importance of heat exchanger selection is
highlighted with information regarding compact (including microchannel-based) heat
exchangers. Software programs are restricted to conventional shell and tube heat
exchangers due to lack of readily available data on compact heat exchangers. This is
because of their currently limited industrial application. However, it is shown that with the
use of compact heat exchangers, due to their higher heat transfer coefficient, reduced space
requirement and increased safety compared to shell and tube heat exchangers, it is possible
to achieve operating cost as well as capital cost reduction which can add to the benefits
achieved by heat integration.11,12
8.2. Methodology
For the application of pinch analysis to the direct flow oxidation process a stepwise
procedure is followed (Figure 8.1).
Figure 8.1. Steps of pinch analysis

The software program of Aspen Energy Analyzer is employed to carry out these steps of
pinch analysis. The first step is the identification of hot, cold and utility streams in the
process and extraction of thermal data that are supply temperature, target temperature and
heat duty from process heat and material balance. The Aspen PlusTM simulation of the
direct flow oxidation process is used to extract the thermal data and they are given as input
to the Aspen Energy Analyzer. Heating and cooling utilities must also be specified. For this
process LP (low pressure) steam is hot enough to satisfy the heating load of the unit and
cooling water is cold enough to satisfy the cooling load of the unit. Accordingly, the
utilities selected for this process are LP steam and cooling water. Aspen Energy Analyzer
includes a default utility database of utilities typically available on the process site. The
cooling water is specified with 20°C inlet and 25°C outlet temperature. The LP steam is
specified with 125 °C inlet and 124°C outlet temperature.
In order to make economical evaluation some cost data are also required, these are given in
Table 8.1. Utility cost and base operating time are needed to calculate operating cost saving
167
and heat exchanger cost is needed to estimate the investment cost. For estimating the
installed cost of a shell and tube heat exchanger of carbon steel material the equation given
in Table 8.1 is used where A is exchanger area in m2.
Table 8.1. Cost data for pinch analysis
Utility cost* €/GJ

LP Steam 11.24
Cooling Water 0.28
Operating time 8000 h/year
Heat Exchanger Cost €
0.81
Equation** 37614 + 916 A
* Utility cost modified from Turton et al.13 based on current prices.
** Coefficients updated from Hall et al.14 (A ≤ 550m2) based on Nelson-Farrar cost index
for heat exchangers15, 1 € = 1.25 $.
Temperatures of hot and cold streams at any point in the heat exchanger must always have
a minimum temperature difference which is defined as ΔTmin. The next step is the selection
of this ΔTmin. It is related to the thermal effectiveness of the heat exchanger used. Rule-of-
thumb ΔTmin values have been developed from experience for various types of processes for
shell and tube heat exchangers. High temperature processes (over 300°C) use ΔTmin of over
20°C, while low temperature processes (around 0°C) use ΔTmin of under 5°C. Processes at
medium temperature use ΔTmin of typically 10°C16. Accordingly, in this study a ΔTmin value
of 10°C is selected.
Figure 8.2. Combined composite curves on a temperature vs. enthalpy plot
168
8.2. Methodology
With the knowledge of thermal data and ΔTmin composite curves are constructed with the
use of Aspen Energy Analyzer. Composite curves consist of hot composite curve and cold
composite curve placed together on a single temperature – enthalpy plot7 (see Figure 8.2).
The hot composite curve represents the sum of all the heat sources (hot streams) within the
process whereas the cold composite curve represents the sum of all the heat sinks (cold
streams) within the process.6 These curves are separated by ΔTmin and where it is observed
is referred to as pinch.17 Composite curves provide a counter-current picture of heat transfer
and can be used to define energy targets for the process. Heat recovery is possible in the
region where the hot composite overlaps with the cold composite curve and this region
indicates maximum heat integration potential.18 The remaining heating and cooling needs
are supplied by the utilities. The enthalpy difference at the upper end of the curves indicates
minimum hot utility target. Similarly, the enthalpy difference at the lower end of the curves
indicates minimum cold utility target.
The pinch principles are central to the procedure for designing the heat exchanger network.
The main focus is to minimize the total annual cost by taking into account the energy
targets (minimum hot and cold utility) of the process. Aspen Energy Analyzer is used for
performing optimal heat exchanger network design. It recommends several near optimum
heat exchanger network designs with its embedded optimization algorithm.10,19 The
resulting heat exchanger network designs are presented graphically with a grid diagram
which shows the coupling of hot and cold streams, so the process-to-process heat exchange.
The generated design options are compared based on minimum total annual cost and the
one that gives the minimum is selected. For calculating total annual cost, the capital cost is
multiplied by an annualization factor in order to set it on the same basis as the operating
cost. Then the addition of the operating cost and the annualized capital cost yields the total
annual cost (Equation 8.1). The annualization factor is calculated with Equation 8.220:
Total annual cost (TAC) Operating cost Capital cost Annualization factor (8.1)
1
Annualization factor (8.2)
1 1
where is interest rate and is the plant life. For the selected plant life of 10 years and
interest rate of 10 % the annualization factor calculated is 0.162745.
In the following section, the pinch analysis performed with Aspen Energy Analyzer for the
direct flow oxidation process will be described following the steps described above.
The effect of change in conversion for the direct route on the energy requirements is
analyzed by making simulations at conversion levels of 40%, 50% and 98%. Conversion of
cyclohexene of 40 % is based on primary result from laboratory flow experiment and 50 %
is based on the current result achieved. Since the experiments are ongoing, it is likely that it
can be improved further. Therefore, also conversion of 98 % was simulated which reflects
169
the best performance achieved for the direct route in literature in batch experiments and
represents the superficial best performance.

8.3.1. Data extraction
The thermal data of supply and target temperature of hot and cold streams as well as heat
duty for the direct flow oxidation process is extracted from the process simulation and is
given in Table 8.2 for the 40% conversion case. Heating and cooling utilities selected were
LP steam and cooling water respectively.
Table 8.2. Thermal data of hot and cold streams of the direct flow oxidation process
Stream Stream Stream Supply Target Heat Duty

Number Name Type Temp (°C) Temp (°C) (kW)
1 Cyclohexene Cold 20 90 2,705
2 30% H2O2 Cold 20 90 35,719
3 Cryst. 1 feed Hot 90 50 21,450
4 Mother liquor Cold 50 85 17,688
5 Cryst. 2 feed Hot 85 30 18,339
6 Mixer feed Cold 30 50 108
7 Dryer feed Cold 50 110 3,884
8 Product Hot 110 30 2,400
8.3.2. Defining energy targets
Figure 8.3. Combined composite curves of direct flow oxidation process
170
The hot and cold composite curves are constructed with the input thermal data and ΔTmin
value of 10°C. In Figure 8.3 this is given for the 40% conversion case. It is seen that the
minimum cold utility target is 0 and the minimum hot utility target is calculated using
Aspen Energy Analyzer as 17,916 kW for the 40% conversion case. For the other two
conversion cases the minimum cold utility target is also 0. The hot utility target for the 50%
conversion case is 16,560 kW and for the 98% conversion case it is 13,492 kW. The heat
exchanger network designs to be generated will be based on these targets for minimum
energy consumption.
8.3.3. Initial heat exchanger network

The initial network for direct flow oxidation process has all the heating and cooling
requirements of process streams satisfied by utility streams. The heat exchanger network
based on this basic design can be conveniently revealed by a grid diagram (see Figure 8.4,
top) that shows how individual process streams and utility streams of LP steam and cooling
water are matched with heat exchangers.
Hot process streams flow direction is from left to right and cold process streams run from
right to left. A red marking of a heat exchanger indicates a heater with the use of hot utility
and a blue marking of a heat exchanger indicates a cooler with the use of cold utility. If
process-to-process heat exchanger was available it would have been shown with a grey
colour. It is seen that this heat exchanger network can be improved with process-to-process
heat exchange.
8.3.4. Designing heat exchanger network with pinch analysis

Aspen Energy Analyzer is used to generate heat exchanger network (HEN) design options
to satisfy most closely the energy targets. It has an embedded optimization algorithm that
minimizes the total annual cost subject to these constraints. Table 8.3 gives the total heat
exchanger area, capital investment, hot utility duty, cold utility duty and annual operating
cost results for the generated best two design options for the three conversion cases.
Table 8.3. Generated best two HEN design options of direct flow oxidation process for
three conversion cases
Design options Area (m2) Capital cost Hot utility Cold utility Operating cost
(106 €) (kW) (kW) (106 €/y)
40 % Design 1 43,515 13.02 18,516 600 6.00
40 % Design 2 48,124 14.26 17,916 0 5.80
50 % Design 1 30,084 9.12 16,560 0 5.36
50 % Design 2 29,315 8.89 16,851 291 5.46
98 % Design 1 9,400 3.24 13,492 0 4.37
98 % Design 2 8,642 3.01 14,150 658 4.59
171
Figure 8.4. Grid diagram of initial heat exchanger network (top) and improved heat
exchanger network (bottom) for the direct flow oxidation process (40% conversion case)
It is seen that for example for 40 % conversion case, design 2 enables satisfying minimum
energy targets but at a higher capital cost. In order to be able to select the best design option
for each conversion case, total annual cost (TAC) calculation needs to be made. It is found
using Eq. 8.1 and 8.2 explained above. The calculated TAC of the design options for the
three conversion cases is given in Table 8.4. Design 1 is found to be the best design option
with minimum TAC for all three conversion cases. The HEN of design 1 for 40%
conversion case is presented with a grid diagram in Figure 8.4, bottom. For this design
there are fourteen heat exchangers. Eight of the heat exchangers achieve process-to-process
heat recovery which indicates the improved energy efficiency of the direct flow oxidation
process. Of the remaining six, five is supplied with hot utility of LP steam and one is
172
supplied with cold utility of cooling water. The improved HEN of direct flow oxidation
process for the other conversion cases of 50% and 98% are given in the appendix.
Table 8.4. TAC of best two HEN design options of direct flow oxidation process for three
conversion cases
40 % 40 % 50 % 50 % 98 % 98%
Design 1 Design 2 Design 1 Design 2 Design 1 Design 2
TAC, 106 €/y 8.117 8.120 6.845 6.904 4.894 5.077
The generated improved HEN results are compared with the results of the initial network
without heat integration as seen in Table 8.5. In order to achieve process-to-process heat
exchange more heat exchangers are required in the improved HEN, resulting in increased
capital cost. However, the hot and cold utility requirements are very much reduced
indicating a reduced operating cost. The operating cost for 40 %, 50 % and 98 %
conversion is calculated to be reduced by 70 %, 68 % and 50 % respectively. With the
increase in conversion the heat load in the system is reduced creating less room for
improvement. Accordingly, the benefit achieved with heat integration is seen to be reduced.
Table 8.5. Comparison of improved HEN with initial HEN for three conversion cases
Area (m2) Capital cost Hot utility Cold utility Operating cost
(106 €) (kW) (kW) (106 €/y)
40 % Initial HEN 11,836 3.73 60,105 42,189 19.80
40 % New HEN 43,515 13.02 18,516 600 6.00
50 % Initial HEN 9,257 2.99 48,859 32,299 16.08
50 % New HEN 30,084 9.12 16,560 0 5.36
98 % Initial HEN 4,278 1.60 26,773 13,281 8.77
98 % New HEN 9,400 3.24 13,492 0 4.37
Pay-back time can be calculated to see how long it takes for the operating cost saving to
pay back the extra capital cost requirement. This is given in Table 8.6. It is seen that with
the improvement of the HEN major saving in operating cost is achieved which enables to
pay back the extra capital cost requirement in less than one year (4.5 – 8 month) in all
cases. These results indicate that that heat integration is a useful technique for efficient
design of processes as it results in more sustainable and economical condition.
Table 8.6. Pay-back time for three conversion cases
Conversion Capital cost Operating cost saving Pay-back time

Case (106 €) (106 €/y) (months)
40 % 13.02 13.80 11.3
50 % 9.12 10.72 10.2
98 % 3.24 4.41 8.8
173
To estimate the cost of plant or inside battery limit (ISBL) investment, besides cost of
equipment, the cost of equipment delivery, installation, piping, instrumentation, electrical
and civil work should also be taken into consideration.21 The total purchase cost of
equipment for the direct route representative of 40-50% conversion was estimated in
Chapter 7 as 18.9 M€. The total delivered cost of equipment (E) is found to be 25 M€
adding 15% delivery cost to the purchased equipment cost and 15% contingency on
estimation. ISBL can be estimated with factorial estimation where typical factors for the
other cost contributors based on delivered cost of equipment are present in literature and are
given in Table 8.7. The sum of these factors is 2.2 for a fluids-solids process type21 making
the ISBL cost 3.2 times the total delivered cost of equipment. The ISBL cost calculated in
this way is found to be approximately 80 M€.
The fixed capital investment can also be estimated with a factorial estimation which is the
sum of ISBL cost, offsite cost, engineering cost and contingency. Offsite (OS) cost can be
calculated as 40% of ISBL cost.21 Engineering cost then can be calculated as 25% and
contingency as 10 % of ISBL plus offsite cost.21 The fixed capital investment is then found
as 1.89 times the ISBL cost. The capital investment calculated in this way amounts to
approximately 150 M€ (see Table 8.7). It can be seen that the annual operating cost saving
achieved with pinch analysis of 13.8 M€ for 40% conversion case can enable to pay out the
plant cost (ISBL) of 80 M€ in approximately 6 years.
Table 8.7. Plant cost estimation
Fraction Cost, M€
Total delivered cost of equipment(E) 1.0 of E 25.0
Purchased equipment installation 0.5 of E 12.5
Piping 0.6 of E 15.0
Instrumentation & controls 0.3 of E 7.5
Electrical systems 0.2 of E 5.0
Civil work 0.3 of E 7.5
Structures and buildings 0.2 of E 5.0
Lagging and paint 0.1 of E 2.5
ISBL Investment 3.2 of E 80.0
Offsite cost (OS) 0.4 of ISBL 32.0
Engineering cost 0.25 of (ISBL+OS) 28.0
Contingency charges 0.1 of (ISBL+OS) 11.2
Fixed Capital Investment 1.89 of ISBL 151.2
174
8.4. Design Considerations – Compact Heat Exchangers

Pinch analysis presents a novel approach that integrates the design of heat exchange system
to the design of the core process. It is seen in section 8.3 that with its application major
reduction in utility consumption and correspondingly operating cost saving can be
achieved. Another important consideration in process design is heat exchanger selection.
However, current design procedures are mainly based on the use of shell and tube heat
exchangers. This is because software programs handle only this heat exchanger type17. This
is mainly due to lack of readily available data on compact heat exchangers because of their
currently limited industrial application.
Numerous articles have been published regarding the advantages of compact heat
exchangers which are characterized by having a high area density; the ratio of heat transfer
surface to heat exchanger volume.11,12 A typical shell and tube exchanger has an area
density of around 100 m2/m3 on one fluid side.12 A compact heat exchanger has an area
density over 700 m2/m3 in gas side of a gas-liquid heat exchanger and over 400 m2/m3 for
liquid-liquid heat exchanger.12 Micro heat exchangers being the most compact have an area
density greater than 10,000 m2/m3.11
8.4.1. Commercial microchannel-based heat exchanger examples

The microchannel-based heat exchanger technology enables enhanced heat transfer realized
from large area density and heat transfer coefficients that are much higher than typically
seen in conventional heat exchangers.22 This feature allows the microchannel heat
exchangers to be compact and lightweight. By arranging microchannel in pure counter
current flow pattern, high thermal effectiveness of 99 % is possible.23 In contrast to 6-8°C
or even 10°C approach of conventional shell and tube heat exchangers, a less-than 1°C
temperature approach is available which can enable significant operating cost saving.24
IMM develops prototype micro heat exchangers suitable for pilot or small scale production
plants. Laser-welded (WT-series) and brazed (HX-series) type micro heat exchangers are
available.25 They are developed for liquid/liquid, gas/liquid or gas/gas applications.
Counter- or co-current flow regimes are possible. The core elements are chemically etched
microstructured plates in many cases, but rolled or embossed devices are on their way to
reduce the fabrication cost.25 As the name implies, they are joined by high precision laser-
welding or brazing technology forming a stack of aligned plates. They enable enhanced
heat transfer with their high inner specific surface. The brazed series are suitable for
applications at high temperature (up to 800°C) and pressure (up to 500 bar).25 When coated
with catalyst, the heat exchangers can also be used as chemical reactors.
Printed circuit heat exchanger (PCHE) is an established micro- and millichannel heat
exchanger technology commercially manufactured by Heatric Company.26 The core
matrices are formed by chemically milling microchannels (under 200 μm) or millichannels
175
(up to 3 mm) into flat metal plates and then diffusion bonding the plates together into a
single block.23 The complete heat exchanger is composed by welding together as many of
these blocks as the thermal duty required. Diffusion bonding is a solid state joining process
which gives parent metal strength and ductility throughout the entire exchanger23. No
gaskets or braze materials are required for the assembly making the potential for leakage
and corrosion reduced to minimum.26 Because of the compactness provided by its design
they are up to 85% smaller and lighter than an equivalent shell and tube heat exchanger.26
Obviously, this reduction will cut the material and installation cost noticeably. Flexibility is
another feature of PCHE. The standard construction material is stainless steel but it can be
constructed out of a range of materials to meet process requirements. In this way they have
very large temperature capability from cryogenic (-200°C) to 900°C.11 With its structure it
can operate at pressures of more than 600 bar.26 They are also able to incorporate multiple
process streams into a single unit reducing exchanger and piping complexity simplifying
process control.26 They can also incorporate additional unit operations such as chemical
reaction, mass transfer and mixing.26
8.4.2. The benefits of compact heat exchangers

A major advantage of compact heat exchangers is their high thermal effectiveness. The
effectiveness can be defined as the ratio of the actual heat transfer to the maximum possible
heat transfer11. A higher effectiveness implies a closer temperature approach (see Figure
8.5). This gives possibility of further operating cost saving as explained in section 8.4.3.
Figure 8.5. Schematic representation of temperature approach and temperature cross

Most compact heat exchangers allow the application of pure countercurrent flow which
enables better use of temperature driving force.27 It is most valuable when there is a
temperature cross (see Figure 8.5). For shell and tube heat exchangers the solution is to use
multiple shells in series whereas in case of compact this can be achieved with single heat
exchanger.27 This gives possibility of capital cost reduction as explained in section 8.4.3.
176
Another major advantage of compact heat exchangers is their reduced size and weight.11
Figure 8.6 shows for the same heat transfer duty reduction in size and weight compared to
shell and tube heat exchanger. This makes installation easier and less costly. For shell and
tube heat exchangers the installation factor is approximately 3 whereas it ranges from 1.5 to
2 for compact heat exchangers depending on the size of the unit.27 Thus compact heat
exchangers can be cheaper in view of the total installed cost; especially in the case when
the heat exchanger has to be made from expensive material. Also, for a new plant design a
reduction in plant size can be achieved reducing investment cost.
Figure 8.6. Size and weight comparison of a PCHE (front) and a shell and tube heat
exchanger for same duty28
Compact heat exchangers also have improved safety. They have much reduced hold-up
which makes handling hazardous fluids safer.11 The tube rupture hazard in shell and tube
heat exchangers is eliminated.26 They are also less prone to fouling and are able to cope
with rapid temperature changes.29
Shell and tube heat exchangers handle only one hot stream and one cold stream. Some
compact heat exchangers (plate-fin and printed circuit) can enable multi-streaming.11 Figure
8.7 depicts the multi-streaming possibility with a PCHE. This enables additional to reduced
plant complexity, major reduction in heat exchange area and accordingly cost.
Figure 8.7. Multi-streaming possibility enables PCHE to replace three conventional units28
177
Many of these benefits of compact heat exchangers can only be realized if the heat
exchanger selection is handled at the earliest stages of plant design. Especially the saving in
terms of weight and space cannot be seen if the plant layout is already set for shell and tube
heat exchanger. The installed cost should be taken into consideration rather than the
purchase cost since this gives the effect on total cost. Also, the heat exchanger network
design would be different with the improved thermal effectiveness (closer approach
temperature), pure counter flow and multi-streaming. Therefore, in view of total plant cost
the most economic heat recovery can be achieved if the selection of heat exchanger is
undertaken from the beginning considering available heat exchanger technologies. It is
believed that, as the compact heat exchangers become more widely known and applied
there will an increasing demand toward them for general application instead of shell and
tube heat exchangers. Using such innovative tools means considering the effect of process
intensification also on the utility side aiming at decreasing the size of the plant, thereby
rendering a true holistic picture of the intensification.
8.4.3. Heat exchanger network design change implications

With the pure counter current flow achieved with compact heat exchangers major
advantage in terms of capital cost can be attained in case of temperature cross. Whenever
there is a temperature cross (cold stream heated to a temperature greater that the outlet
temperature of the hot stream) the duty is split into multiple shells in series for shell and
tube heat exchanger design.17 For the direct flow oxidation process studied here three heat
exchangers have a high number of shells due to presence of temperature cross for the 40%
conversion case and they make the major part of the capital cost requirement. These heat
exchangers are presented in Table 8.8 for 40% conversion case. Their total cost is found to
contribute to more than 80% of the total capital cost requirement given in Table 8.6.
Table 8.8. Heat exchangers with temperature cross of direct flow oxidation process of new
HEN (40 % conversion case)
Heat Area Capital cost Number Hot T Hot T Cold T Cold T

exchanger (m2) (106 €) of shells in (°C) out (°C) in (°C) out (°C)
1 5514 1.6 12 50 30 20 33
2 10712 3.1 24 85 50 43 66
3 20734 5.9 44 90 58 50 80
With the use of microchannel-based heat exchangers, series of shells used for shell and tube
design can be replaced with single compact heat exchanger. This major reduction in the
heat transfer area results in smaller units to be used in plant which reduces plant
complexity, making operation and control easier and maintenance cheaper. It also enables
capital cost saving by replacing number of units with a single small unit. It is estimated that
in this way the capital cost requirement of heat exchangers can be cut at least by half. This
178
also indicates that the operating cost saving will enable to pay back the extra capital cost in
even shorter time.
It was explained that with the high thermal effectiveness of compact heat exchangers
temperature approach can be reduced. This has the potential to reduce the operating cost by
bringing the composite curves closer to each other increasing the process-to-process heat
integration.
For the direct flow oxidation process studied here there is total overlap of the composite
curves at ΔTmin of 10°C as can be seen in Figure 8.3. In these cases of total overlap the
positive effect of decreasing ΔTmin value on reduced operating cost cannot be seen. In case
of partial overlap this reduction in temperature approach will enable reduction of utility and
correspondingly operating cost. It can therefore be said that the use of microchannel-based
heat exchangers will be advantageous only in this case. This situation is shown for one
process example (from Aspen tutorial guide19) in Figure 8.8.
Figure 8.8. Combined composite curves of Aspen tutorial guide example process with
ΔTmin = 10 °C (top), ΔTmin = 1 °C (bottom)
In top figure the composite curves at ΔTmin of 10°C and at bottom figure the composite
curves at ΔTmin of 1°C is shown. The cold and hot composite curves have partial overlap
and with the reduction of approach temperature they get closer enabling higher process-to-
179
process heat recovery and thus lower utility requirement. In Figure 8.8, cold utility
requirement is shown with a blue two sided arrow and hot utility requirement is shown with
a red two sided arrow. Also, ΔTmin of 10°C is denoted as 1, ΔTmin of 1°C is denoted as 2.
Same denotation is used in Figure 8.9. The reduction in utility requirements are given in
Figure 8.9. Decreasing ΔTmin from 10°C to 1°C enables about 20% reduction in both hot
and cold utility resulting in operating cost saving.
Figure 8.9. Hot utility (top) and cold utility (bottom) requirements vs. ΔTmin for Aspen
tutorial guide example process
8.5. Conclusion
This chapter shows that intensification on the utility side is of primary relevance for process
design intensification of flow processes in micro- or similar smart-scaled reactors at large
capacity, beyond fine chemistry. As such, it probably can constitute a strong pillar in the
overall process-design intensification.
180
8.5. Conclusion
The direct oxidation of cyclohexene with hydrogen peroxide for adipic acid synthesis was
taken as the example for process-design intensification to investigate its potential for
process simplification. It was seen in previous chapter through process simulation that this
direct route leads to a more compact plant design. To design an energy efficient heat
exchanger network for this novel direct flow route in this chapter pinch analysis and the
software Aspen Energy Analyzer is employed. The design is selected from the alternatives
provided by the software by making a total annual cost calculation. By following the
stepwise procedure of such pinch analysis an improved heat exchanger network design is
achieved for the three conversion cases considered: 40% (representing primary result), 50%
(representing current improved result) and 98% (representing superficial best performance).
Compared with the basic design where the heating and cooling requirements are satisfied
by the use of utilities, the improved design enables 50-70 % operating cost saving. It is
calculated that this saving pays back the extra capital requirement for the heat exchangers
in 4.5 – 8 months.
Heat exchanger selection is also a very important design consideration which should be
handled at the earliest stages of plant design so that in view of total plant cost the most
economic heat recovery can be achieved. However, the current software programs are
restricted to conventional shell and tube heat exchangers due to lack of readily available
data on compact heat exchangers. Compact heat exchangers have proven advantages in
terms of thermal effectiveness, safety, size and weight. With their improved heat transfer
that enable operation with approach temperature as low as 1°C, they give the possibility to
reduce the utility requirement by 20%. They can also enable saving in the total plant cost
and enable a reduction in plant complexity with the pure countercurrent flow achieved. For
the case of temperature cross they can achieve the additional capital cost to be halved. As
their application increase in industry and their benefits are most widely recognized it is
believed that there will be an increasing demand towards them.
In conclusion, the use of microprocess or other smart-scaled technology can achieve
considerable capital cost and energy consumption reduction through process-design
intensification lowering the management’s main decision barrier towards new technologies.
Yet, considerable challenges are expected when releasing the theoretical potential into
industrial practice.
181
8.6. Appendix
Figure 8.7. Grid diagram of improved heat exchanger network for the direct flow oxidation
process for 50% conversion case
182
Figure 8.8. Grid diagram of improved heat exchanger network for the direct flow oxidation
process for 98% conversion case
References
& Technology, 2012, 35, 1184–1204.
Technik, 2012, 84, 660–684.
Transactions, 2012, 29, 565–570.
4. Q. Wang, I. Vural Gürsel, M. Shang, and V. Hessel, Chemical Engineering
Journal, 2013, 234, 300–311.
2013, 6, 746–89.
6. A. P. Rossiter, Chemical Engineering Progress, 2010, 106, 33–42.
7. Pinch Technology: Basics for Beginners,
http://www.cheresources.com/content/articles/heat-transfer/pinch-technology-
basics-for-beginners.
8. D. W. Townsend and B. Linnhoff, AIChE Journal, 1983, 29, 742–748.
9. B. Linnhoff and E. Hindmarsh, Chemical Engineering Science, 1983, 38, 745–763.
183
10. AspenTech, Aspen Energy Analyzer, https://www.aspentech.com/products/aspen-

hx-net.aspx.
11. D. A. Reay, Heat Recovery Systems and CHP, 1994, 14, 459–474.
12. Q. Li, G. Flamant, X. Yuan, P. Neveu, and L. Luo, Renewable and Sustainable
Energy Reviews, 2011, 15, 4855–4875.
13. R. Turton, R. C. Bailie, W. B. Whiting, and J. A. Shaeiwitz, Analysis, Synthesis,
and Design of Chemical Processes, Prentice Hall, New Jersey, 3rd edn., 2009.
14. S. G. Hall, S. Ahmad, and R. Smith, Computers & Chemical Engineering, 1990,
14, 319–335.
15. Nelson-Farrar Cost Indexes, Oil and Gas Journal,
http://www.ogj.com/articles/print/volume-112/issue-7/processing/nelson-farrar-
monthly-cost-indexes.html.
16. S.-G. Yoon, J. Lee, and S. Park, Applied Thermal Engineering, 2007, 27, 886–893.
17. G. Polley and R. Mukherjee, in Kirk-Othmer Encyclopedia of Chemical
Technology, 13th edn., 2005, pp. 178–224.
18. G. Kleemann and W. Paul, Chemical Engineering & Technology, 2010, 33, 603–
609.
19. AspenTech, Aspen Energy Analyzer Tutorial Guide, Burlington, MA, 2009.
20. Depreciating Capital Costs, Cost and Financing in WikiEducator,
http://wikieducator.org/Cost_and_Financing_in_Open_Schooling/Capital_Costs/De
preciating_Capital_Costs.
edn., 2008.
22. K. Schubert, J. Brandner, M. Fichtner, G. Linder, U. Schygulla, and A. Wenka,
Microscale Thermophysical Engineering, 2001, 5, 17–39.
23. L. M. Pua and S. O. Rumbold, Industrial Microchannel Devices – Where Are We
Today ?, Presented at First International Conference on Microchannels and
Minichannels, Rochester, New York, 2003.
24. P. Takats and G. Nemeth, MOL Scientific Magazine, 2011, 2, 42–48.
25. IMM Catalogue, http://www.imm-mainz.de/fileadmin/IMM-upload/Flyer-
Katalog_etc/Catalogue09_WT.pdf (accessed: 19 December 2013), .
26. Heatric Brochure, http://www.heatric.com/hres/Heatric%20standard%20brochure
(accessed: 4 August 2015), .
27. C. Haslego and G. Polley, Chemical Engineering Progress2, 2002, 98, 32–37.
28. Heatric, Printed circuit heat exchangers,
http://www.heatric.com/heatric_exchangers_benefits.html.
29. A. Kananeh and J. Peschel, in Heat Exchangers - Basics Design Applications, ed. J.
Mitrovic, InTech, New York, 2012.
184
CHAPTER 9
Conclusions and Recommendations
9.1. Conclusions
Research Highlights in this Œuvre at a Glance:
 Process-design intensification was predicted for the first time by the ERC Advanced Grant
in 2010 and the research presented here is the first manifest of that concept.
 Theoretical potential analysis of process development time reduction for case scenarios in
fine and bulk chemistry was done as a function of the three intensification fields and plant
modularity. Before in literature, only the latter was predicted.
 A continuous metal scavenging unit was developed and coupled to a CuAAC click reac-
tion in flow enabling to reach pharma-grade purity level (ppm) in 1 stage of extraction
 Coupled reaction-separation in flow has very few demonstrations and this was
successfully conducted here; even at the example of a one-pot synthesis.
 Scale-up of liquid-liquid extraction unit was done using milli-scale CFI and a phase
separator, and gave ~20% higher extraction efficiency compared to straight tube setup.
 The use of the CFI for immiscible liquid-liquid mixing and its application for liquid-
liquid extraction was reported for the first time
 The study of flow separators that can handle flow rates suitable for pilot scale
pharmaceutical production is very rare and this was successfully achieved here
 Slug flow regime was maintained at 7.2 l/h (highest reported in literature) and the micro
effect of capillary force was able to be used for phase separation at such high flow rate
 A detailed cash flow analysis for a modular plant was reported for the first time to
quantify the benefits and to make comparison with conventional plants
 The combination of smart-scaled process-intensified equipment with modular compact
plants was analyzed and they were found synergistic in costs
 An initial techno-economic analysis was done to analyze the feasibility of a novel
process for biofuel production considering decentralized production in modular plants;
also here in literature only rare information is given.
 The process simplification potential of a direct adipic acid synthesis route was
investigated by cost, energy and environmental analysis at world-scale capacity
 The latter involved a comprehensive end-to-end Aspen simulation analysis of using
small-scale flow processing at world-scale capacity for the first time.
 Heat integration for the direct adipic acid synthesis was achieved with pinch analysis
and compared to initial design significant utility cost saving is attained
185
9. Conclusions and Recommendations
Figure 9.1. A schematic representation of the three main parts of this thesis
From intensified reactor to intensified process engineering
Chapter 1 outlines the synergy of micro dimensions and harsh operating conditions which
results in a three-step intensification within a reaction system: transport, chemical and
process-design intensification. Transport intensification that improves mass and heat
transfer has been vastly demonstrated. Emerging chemical intensification uses highly
intensified, unusual and typically harsh process conditions to boost micro-processing (high-
T, high-p, high-c, safety). This was predicted first in 2005 by Hessel1 and resulted finally in
the ERC Advanced Grant in which this work was done. Process-design intensification was
in this grant and particular in this thesis explored as a new field in micro processing. It
heads for integrated and simplified smart-scaled (micro-meso) flow process design in a
holistic picture. The latter two constitute Novel Process Windows.
After having seen almost two decades of flow-related developments of micromixers,
microreactors, micro heat exchangers, and recently microseparators, the time was ripe for
the next round of the implementation of microreactor and flow techniques. This is devoted
towards bringing all of the above-mentioned devices together (drop-in & integrate), scaling
these up to industrial productivity and embedding them in a modern Future Factory type
plant environment (modularize) with an overarching process and plant approach. A final,
yet still too less exploited consideration is the chance for radical changes in process design
by process simplification (end-to-end), e.g. by newly developed chemical transformations
such as direct synthesis in flow. These concepts are schematically given in Figure 9.1.
186
9.1. Conclusions
Potential of process development time reduction through flow processing

In Chapter 2, case scenarios are considered with stepwise intensification covering the three
intensification fields of flow processing and the use of modular plants. Such paradigm
changes in process development can enable a faster development to the final production
plant through reduction in the number of apparatus in the plant, reduction in apparatus size,
and higher predictability in the scale-out of the apparatus. Their effect is shown in a generic
(impact-wise) manner giving an idea of their impact on the process development time in
bulk and fine chemistry. The distribution into individual impacts allows to assess which
approaches in intensification ensure already today a good practice in fastened process
development, which are better used in near future, and which do not save time in process
development at all. It is seen that time savings in process development are likely for fine-
chemical applications up to the border of bulk-chemical processes; but that for true bulk-
chemical applications, the cost and sustainability arguments are the main motivation for
investigation.
Connected reaction and metal separation in flow
One possibility to bring step-change improvement in chemical production was seen to be
through enabling multi-step synthesis in flow which is especially attractive to facilitate
switch to continuous processing for fine chemistry/pharmaceutical production. The
essential base unit hereto is coupled reaction-separation. Chapter 3 describes accordingly
the development of a continuous metal separation unit which enables liquid-liquid
extraction based on slug flow using chelating agents and phase separation with a porous
fluoropolymer membrane. Its performance was optimized by investigating the use of
different types of copper catalysts and metal scavengers together with variation of metal
scavenger concentration and pH. It was successfully coupled to the copper-catalyzed azide-
alkyne cycloaddition reaction studied in one-pot (eliminating the need to isolate and handle
potentially explosive azide). The triazole product was attained in flow with yield of up to
92% and the level of purity requirement for pharmaceuticals of 15 ppm was met in one
stage of extraction.
Scale-up of flow extraction unit utilizing CFI
The vision is to achieve such multi-step synthesis in flow on pilot scale. This calls for scale-
up of separation units. This study is rare in micro flow literature and was conducted in
Chapter 4. A liquid-liquid extraction setup was developed using milli-scale CFI (3.2 mm
ID, 210 cm length) and a phase separator. Two separator concepts were investigated: a
PTFE membrane flow separator and a slit shaped flow separator composed of glass and
Teflon rectangular capillaries. The extraction performance of the setup was tested for the
two widely used and EFCE recommended model systems of toluene-water-acetone and n-
butyl acetate-water-acetone. For toluene-water-acetone, a flow rate up to 120 ml/min was
studied which corresponds to 7.2 l/h and 58 m3/y. It is important to point out that the slug
187
flow regime that is considered limited to microfluidics is maintained even at such pilot
scale. For both systems close to thermodynamic extraction efficiency was achieved (96%
for first, 100% for second). Compared to a straight tube, the CFI showed 20 % higher
performance at high flow rates most likely due to the Dean vortices which is intensified by
the flow inversions in the CFI. In straight tube, parallel flow developed at higher flow rates
whereas the CFI kept operating in the slug flow regime.
Compact modular plants
Another possibility to bring step change improvement in chemical production is the use of
compact, modular plants. They are seen as key to faster market introduction of a new
product (50% idea) by shortening time-to-market resulting in higher cash flows. In Chapter
5, a detailed calculation of capital investment and operating cost for a modular plant was
done, at the example of Evotrainer production platform of Evonik Industries, quantifying
the benefits of modular design. The step-change character of process intensification and
smart-scale technologies such as microstructured reactors is considered, in particular to
check if they add to the advantage of modular compact plants. For a fine chemical example,
six scenarios were considered: two plant type (conventional, Evotrainer) and three
processes (batch, micro, virtual micro of high yield). In comparison to the conventional
plant, the use of the Evotrainer infrastructure resulted in significant higher Net Present
Value (NPV) to be achieved with the reduction in construction period. Considering the
process type, microreactor (or generally intensified) processes were seen to have a higher
NPV than that of the batch processes, due to the profound effect of the reduction in
operating cost achieved. It was seen therefore that the modular plants and micro process
technology can be synergistic in costs.
Decentralized production with modular plants
The modular plants are also beneficial because of their flexibility and scalability. Therefore,
they are ideally suited for decentralized production such as biomass conversion processes
(‘Chemistry on Wheels’; Evonik promotion name2). This is because biomass is widely
distributed and its composition and overall amount differs from location and season.
Therefore, decentralized production close to resource is beneficial to save from
transportation and storage costs. Yet, plant modularity at small scale has its price
(seemingly violating the economy of scale) and thus holistic cost calculations, which show
if the total cost is positive or negative, are rare for decentralized biomass plants. Finally the
plant has to meet current gasoline fuel prices, costing enables a check if the goal was
reached.
With such motivation, a holistic cost calculation was done for the application case of
decentralized biofuel production in Chapter 6. It involves an initial techno-economic
analysis to analyze the feasibility of a novel process for biofuel production. Process
integration and intensification was achieved by coupling of reforming of biogas and bio oil
188
9.1. Conclusions
in a microchannel plate heat exchanger. In this process via methanol, neither challenging
gas separation nor extensive light gas recycle loops were required in contrast to Fischer-
Tropsch based routes. Therefore a compact process design was attained with high carbon
and energy efficiency. Cost calculations showed that these benefits can outweigh the higher
investment cost of decentralized plants especially with the use of modular plants that are
flexible to production capacity and that can be built in short time. Cost calculations showed
that with the total cost of production of gasoline at € 0.5/L; this biofuel production process
can be economically competitive with petroleum-based and Fischer-Tropsch process.
Process simplification of a bulk chemical synthesis
Finally, a step change improvement can be realized by bringing direct chemistries on a full-
process level. The demonstration example taken was the direct adipic acid synthesis which
is an important intermediate for the nylon-6,6 industry. Chapter 7 described the
investigation of the process simplification potential of this direct route by cost, energy and
environmental analysis. To propose the flow diagram and conditions for this direct route,
the reaction characteristics were considered together with the downstream equipment of the
two-step commercial route. A reduction in step of synthesis led to a compact plant design
owing to the need of much less process units (one reactor, less number of downstream
units) compared to the commercial route carried out in two-steps. This resulted in the total
purchase cost of equipment to be halved although a more advanced and very costly flow
reactor was used. Process-design intensification which is seen to enable reduction in
investment costs can thus lower the main decision barrier in the management of chemical
industry towards new technologies. The analysis solidified the dream of pioneers in
continuous small-scale plant design, Benson and Rinard, who predicted that separation will
be much reduced due to advances in flow reactor performance.3–5
Through simplification or elimination of energy intensive separation units, energy
requirement was also reduced significantly. Life cycle assessment showed that for a number
of impact categories the direct process shows much better results especially evident when
N2O-impact on GWP is considered; yet there are also categories for which the conventional
route is more environmentally sustainable. The main limitation of the direct route was
found to be the energy and pollution intensive Anthraquinone process for the production of
raw material of H2O2. Since consideration from cradle to gate is made this production
method of the raw material is also affecting the results for the direct route.
Heat integration on the process simplified bulk chemical synthesis
Furthermore, to design an energy efficient process for this route pinch analysis was
employed in Chapter 8. A new heat exchanger network was designed by enabling heat
integration with process-to-process heat exchange. Compared with the initial heat
exchanger network where energy requirements are supplied with utility streams, the
189
improved heat exchanger network enabled up to 70 % saving in utility cost which enabled
to pay back the extra capital cost requirement in less than one year.
Overall impact of this research
Overall, the research provides insight at different production scales what chemical plant of
tomorrow and the future may look like – seen through the goggles of micro-flow chemical
intensification envisioned to intensified process-design scale. The analysis accordingly was
done in the hierarchy of scales as follows. Multi-step synthesis for continuous fine chemical
production is now in lab development. Compact modular plants used to shorten time-to-
market and to achieve decentralized production is now in pilot operation for industrial
application cases. Bulk chemistry process change can enable step change improvement by
bringing a new chemical transformation on a full-process level. However, such exploration
of new reaction pathways and their respective new equipment designs to implement new
process condition will naturally require much time to be realized, more a question of
decades than of years.
9.2. Recommendations
Extension of metal scavenging experiments in flow
In Chapter 3, copper scavenging with continuous liquid-liquid microextraction unit was
described to achieve metal catalyst separation in flow. This study can be extended to test
the metal scavenging performance of the unit for other homogeneous metal catalysts. It can
therefore be applied to other metal catalysed reactions such as palladium. PGM noble
metals have become rare and recycling is a must here. In this sense, the new extraction
concept may be used in applications beyond chemistry such as in mining and metallurgy.
In this study the recovery of the chelating agent EDTA was achieved offline and it was
reused. A further study could be done to develop this scavenger recovery and recycle
system in continuous flow. In this way besides the reaction and separation taking place
together in flow recycle could be achieved as well. Kinetic analysis of metal scavenging has
been performed in literature for batch operation.6–8 A possibility can be to extend this study
to investigate kinetic analysis of metal scavenging for a flow process.
Flow regime investigation
In Chapter 4, a milli-scale coiled flow inverter was used in combination with a phase
separator and liquid-liquid extraction was achieved in slug flow regime. It was observed
that parallel flow regime developed in straight tube with increase in flow rate as also found
in literature. Further analysis would be beneficial to understand why the slug flow regime
can be preserved with CFI at higher flow rates. As a continuation to that, it would be
interesting to find the limit for the CFI where the switch to parallel flow occurs.
190
Accordingly, a flow regime map can be made for the CFI showing at what flow rate and
aqueous volume fraction what type of flow regime is observed. In this study the flow rate of
7.2 l/h was the maximum flow rate achievable with the two syringe pumps used. Therefore,
larger capacity pumps would be required to make this further investigation and one
possibility could be to use peristaltic pumps.
Further scale-up of the flow extraction unit
In this study CFI with internal diameter of 3.2 mm was used. Scale-up of the extraction
system can be studied by using several of these CFIs in parallel or by developing a CFI of a
larger internal diameter. It would be interesting to apply this unit to a fine chemistry or
pharmaceutical production process of industrial importance enabling multi-step synthesis at
high throughput. The extraction is composed of two parts of mixing and phase separation. It
would not be expected to have problems in achieving high flow rates in mixing part which
is achieved with CFI in this case. Main limitation to achieving such high flow rates would
come from the phase separation part. In this study highest flow rate in literature is achieved
using micro effect of capillary force to achieve separation with a slit shaped flow separator
composed of glass and Teflon rectangular capillaries. So, further development of such units
is required to achieve higher throughput. It was not possible to have flow rate above 20
ml/min with the membrane separator. Development of a membrane separator with an
integrated pressure controller like the device of Adamo et al.9 that can handle high flow
rates required could be the solution to this problem. This pressure control apart from higher
flow rates could also extend the applicability of the separator to lower interfacial tension
systems.
Validation of assumptions on modular plants with industry
The main generic aim of Chapter 5 was to provide parametric sensitivity and a proof of
principle for a new economic model in chemical manufacturing of compact modular plants
using model based scenarios (including virtual microreactor applications). Therefore, for
such a very new technology several assumptions had to be made that are stated in the text.
Clearly, the assumption of cost reductions that could be achieved with modular plants need
to be proven by industrial demonstration activities. These are underway in European
Union’s large scale projects on Future Factories.
Extension of the techno-economic study and performing sensitivity analysis
An initial, preliminary techno-economic analysis for a novel process of liquid fuel
production from bioresources was performed in Chapter 6 to analyse the feasibility of the
proposed project. This was carried out before the start of the project activities of the new
project funded by the European commission of BIOGO. Methanol to gasoline converter
product composition is modelled based on typical composition in literature. The
applicability of such composition needs to be verified for biofuel production. The cost of
191
gasoline production is very much dependent on the capital cost estimation and the cost
parameters used in the study. A complete PhD project is now devoted to the comprehensive
assessment of this process therefore a further more detailed work will follow. It is
recommended to perform sensitivity analysis here to investigate the effect of several
parameters such as cost parameters and process conditions on the economics of this
process. Such analysis is performed in the techno-economic studies carried out by the
National Renewable Energy Laboratory (NREL) which can be used as a guideline. 10,11
Validation of the new bulk chemistry process design and additional sensitivity analysis
In Chapter 7 and 8, direct adipic acid synthesis route was studied. Since this direct synthesis
is currently only studied at the reactor level and at lab scale, the downstream equipment and
large scale synthesis was proposed in this study to assess the potential of this route in full-
chain process level holistically. The cost, energy and environmental analysis were done
based on the process simulation made following this proposed process. Therefore, the
results are subject to the validity of the process simulation. This study was conducted in
consultation with Dr. Johan Tinge from DSM. However, a more detailed analysis to reflect
real life application finally needs in-flow from data of industry. Also, a sensitivity analysis
was made in terms of only conversion in this study. It would be good to expand the
sensitivity analysis to investigate the effect of other process parameters.
New hydrogen peroxide production process investigation
The main limitation of this route was found to be the raw material of hydrogen peroxide. It
results in high raw material cost making this route economically infeasible. Its current
production through Anthraquinone process is energy and pollution intensive. This direct
route will be economically and environmentally attractive when low cost H2O2 production
methods become available. H2O2 synthesis from its elements of H2 and O2 directly can be
the solution to reduce the costs. It will also enable the direct route to be considered green on
all levels with the H2O2 production also produced in a green way. Therefore, a further study
can be made to investigate such double direct route to adipic acid.
Process design intensification studies beyond and following this research package
Process-design intensification, as per definition, is a step-change technology. Thus, it is far
from existing practice and largely not implemented in educational curriculum, standard
books, standard software and process-design tools. Moreover, current performances, known
to experts, may be not representative of the full potential. Selection of process data needs so
much care. Validated data are generally lacking. Sometimes extrapolation needs to be
made, which again needs validation. Current process simulation, cost, energy and
environmental analysis given in Chapters 5 through 8 may therefore be far from intrinsic
and theoretical potential. The technology is simply too young and too less documented.
192
Nonetheless and hopefully with the improvements mentioned above, the overall future aim
naturally is to realise a fully-assembled development for one intensified flow process,
meaning to perform all the single objectives given here in the chapters in one process
frame. This thesis could not achieve such total picture for the reasons given above, and
rather provided different selected frames and facets which on their own have still high
innovation.
Thus, the last recommendation is to bring the visions here into reality and to make the Plant
of the Future to the Plant of Tomorrow, i.e. to realise a demonstration example for such
fully-assembled process development. Due to the need to involve production scale, this
needs large-TRL development by public-private partnerships with strong industrial
involvement and lead (TRL = Technology Readiness Level).
References
1. V. Hessel, P. Lob, and H. Lowe, Current Organic Chemistry, 2005, 9, 765–787.
2. Chemistry on Wheels, Evonik Industries,
http://corporate.evonik.com/en/products/product-stories/pages/chemistry-on-
wheels.aspx.
3. I. H. Rinard, in Proceedings of the 2nd International Conference on Microreaction
Technology, IMRET 2, New Orleans, USA, 1998, pp. 299–312.
4. N. Saha and I. H. Rinard, in Proceedings of the 4th International Conference on
Microreaction Technology, IMRET 4, Atlanta, USA, 2000, pp. 327–333.
5. R. S. Benson and J. W. Ponton, Chemical Engineering Research and Design, 1993,
71a, 160–168.
6. G. Chauhan, K. K. Pant, and K. D. P. Nigam, Industrial & Engineering Chemistry
Research, 2013, 52, 16724–16736.
7. G. Chauhan, K. K. Pant, and K. D. P. Nigam, Industrial Engineering Chemistry
Research, 2012, 51, 10354–10363.
8. G. Chauhan, K. K. Pant, and K. D. P. Nigam, Environmental science. Processes &
impacts, 2015, 17, 12–40.
10. M. M. Wright, J. A. Satrio, R. C. Brown, D. E. Daugaard, and D. D. Hsu, Techno-
Economic Analysis of Biomass Fast Pyrolysis to Transportation Fuels, National
Renewable Energy Laboratory, NREL/TP-6A20-46586, 2010.
11. R. M. Swanson, J. A. Satrio, R. C. Brown, A. Platon, and D. D. Hsu, Techno-
Economic Analysis of Biofuels Production Based on Gasification, National
Renewable Energy Laboratory, NREL/TP-6A20-46587, 2010.
193
194
List of Publications
Journal Articles
1. Vural Gursel, I., Kurt, S.K., Aalders, J., Wang, Q., Noël, T., Nigam, K.D.P.,
Kockmann, N. & Hessel, V. (2015) Utilization of milli-scale coiled flow inverter in
combination with phase separator for continuous flow liquid-liquid extraction
processes. Chemical Engineering Journal, 283, 855-868.
2. Kurt, S.K., Vural Gursel, I., Nigam, K.D.P., Hessel, V. & Kockmann, N. (2015).
Liquid-liquid extraction system with microstructured coiled flow inverter and other
capillary setups for single-stage extraction applications, Chemical Engineering Journal,
284, 764-777.
3. Vural Gursel, I., Noël, T., Wang, Q. & Hessel, V. (2015) Separation/recycling methods
of homogeneous transition metal catalysts in continuous flow. Green Chemistry, 17,
2012-2026.
4. Vural Gursel, I., Aldiansyah, F., Wang, Q., Noël, T. & Hessel, V. (2015). Continuous
metal scavenging and coupling to one-pot copper-catalyzed azide-alkyne cycloaddition
click reaction in flow. Chemical Engineering Journal, 270, 468-475.
5. Vural Gursel, I., Wang, Q., Noël, T., Hessel, V., Kolb, G. & van Veen, A. (2014).
Heat-integrated novel process of liquid fuel production from bio resources – Process
simulation and costing study. Chemical Engineering Transactions, 39, 931-936.
6. Wang, Q., Vural Gursel, I., Shang, M. & Hessel, V. (2013). Life cycle assessment for
the direct synthesis of adipic acid in microreactors and benchmarking to commercial
process. Chemical Engineering Journal, 234, 300-311.
7. Vural Gursel, I., Wang, Q., Noël, T. & Hessel, V. (2013). Implementation of heat
integration for efficient process design of direct adipic synthesis in flow. Chemical
Engineering Transactions, 35, 775-780.
8. Vural Gursel, I., Wang, Q., Noël, T., Hessel, V. & Tinge, J.T. (2013). Improving
energy efficiency of process of direct adipic acid synthesis in flow using pinch analysis.
Industrial and Engineering Chemistry Research, 52(23), 7827-7835.
9. Vural Gursel, I., Hessel, V., Wang, Q., Noël, T. & Lang, J. (2012). Window of
opportunity : potential of increase in profitability using modular compact plants and
micro-reactor based flow processing. Green Processing and Synthesis, 1(4), 315-336.
195
10. Vural Gursel, I., Wang, Q., Noël, T. & Hessel, V. (2012). Process-design
intensification : direct synthesis of adipic acid in flow. Chemical Engineering
Transactions, 29, 565-570.
11. Hessel, V., Vural Gursel, I., Wang, Q., Noël, T. & Lang, J. (2012). Potenzialanalyse
von Milli- und Mikroprozesstechniken für die Verkürzung von
Prozessentwicklungszeiten : Chemie und Prozessdesign als Intensivierungsfelder.
Chemie Ingenieur Technik, 84(5), 660-684.
12. Hessel, V., Vural Gursel, I., Wang, Qi, Noël, T. & Lang, J. (2012). Potential analysis
of smart flow processing and micro process technology for fastening process
development - Use of chemistry and process design as intensification fields. Chemical
Engineering & Technology, 35(7), 1184-1204.
Conference Contributions
Oral Presentations
1. Vural Gursel, I. (2015) Milli-scale coiled flow inverter coupled to phase separator for
continuous liquid-liquid extraction applications. Autumn Session of the Dutch Process
Intensification Network (PIN-NL), 18 November 2015, Helmond, The Netherlands.
2. Vural Gursel, I., Aldiansyah, F., Aalders, A., Wang, Q., Noël, T., Hessel, V. & Nigam,
K.D.P. (2014). Continuous metal scavenging and coupling to one-pot click reaction in
flow. NPS-14: 'Fundamentally Innovative', 3-5 November 2014, Utrecht, The
Netherlands.
3. Vural Gursel, I., Aldiansyah, F., Wang, Q., Noël, T. & Hessel, V. (2014). Continuous
metal scavenging in flow coupled to one-pot click reaction. 21st International Congress
of Chemical and Process Engineering CHISA 2014, 23-27 August 2014, Prague, Czech
Republic.
4. Vural Gursel, I., Wang, Q., Noël, T., Lang, J. & Hessel, V. (2014). Window of
opportunity - potential of increase in profitability using modular plants and flow
processing. 21st International Congress of Chemical and Process Engineering CHISA
2014, 23-27 August 2014, Prague, Czech Republic.
5. Vural Gursel, I., Wang, Q., Noël, T., Hessel, V., Kolb, G. & van Veen, A. (2014).
Heat-integrated novel process of liquid fuel production from bio resources – Process
simulation and costing study. 17th Conference on Process Integration, Modelling and
Optimisation for Energy Saving and Pollution Reduction (PRES 2014), 23-27 August
2014, Prague, Czech Republic.
196
6. Vural Gursel, I., Aldiansyah, F., Wang, Q., Noël, T. & Hessel, V. (2014). Connected
reaction and separation in flow for one-pot click reaction. International Conference on
MicroREaction Technology (IMRET13), 23-25 June 2014, Budapest, Hungary.
7. Vural Gursel, I., Wang, Q., Noël, T. & Hessel, V. (2013). Continuous metal
scavenging with a flow liquid-liquid extraction unit. MicroNanoConference'13, 11-12
December 2013, Ede, The Netherlands.
8. Vural Gursel, I., Hessel, V., Noël, T. & Wang, Q. (2013). A new vision in micro
process technology: process-design intensification. Jahrestreffen der Fachgemeinschaft
Process-, Apparate- und Anlagentechnik (PAAT), 18-19 Nov. 2013, Bruchsal,
Germany.
9. Vural Gursel, I., Wang, Q., Noël, T. & Hessel, V. (2013). Implementation of heat
integration for efficient process design of direct adipic synthesis in flow. 16th
Conference on Process Integration, Modelling and Optimisation for Energy Saving and
Pollution Reduction (PRES 2013), 29 September - 2 October 2013, Rhodos, Greece.
10. Vural Gursel, I., Wang, Q., Noël, T., Hessel, V. & Tinge, J.T. (2013). Energy and
operating cost analysis of the smart-scaled flow route directly to adipic acid. 9th
European Conference of Chemical Engineering (ECCE9), 21-25 April 2013, The
Hague, The Netherlands.
Poster Presentations
1. Vural Gursel, I., Wang, Q., Noël, T., Hessel, V., & Nigam, K.D.P. (2015). Milli-scale
continuous extraction with coiled flow inverter connected to phase separator. 10th
European Conference of Chemical Engineering (ECCE10), 27 September - 1 October
2015, Nice, France.
2. Vural Gursel, I., Wang, Q., Noël, T., Kolb, G., Hessel, V., & van Veen, A. (2014).
Process simulation and cost analysis for novel process of liquid fuel production from
bioresources. International Conference on MicroREaction Technology (IMRET13),
23-25 June 2014, Budapest, Hungary.
3. Vural Gursel, I., Hessel, V., Noël, T. & Wang, Q. (2012). Process-design
intensification : direct synthesis of adipic acid in flow. Jahrestreffen der
Fachgemeinschaft Process-, Apparate- und Anlagentechnik (PAAT), 19-20 November
2012, Dortmund, Germany.
4. Vural Gursel, I., Wang, Q., Noël, T. & Hessel, V. (2012). Process-design
intensification : direct synthesis of adipic acid in flow. 15th Conference on Process
Integration, Modelling and Optimisation for Energy Saving and Pollution Reduction
(PRES 2012), 25-29 August 2012, Prague, Czech Republic. (3rd Prize for Best Poster)
197
5. Vural Gursel, I., Wang, Q. & Hessel, V. (2011). Coupling of chemically intensified
microreactors to microseparators. Netherlands Process Technology Meeting (NPS-11),
24-26 October 2011, Arnhem, The Netherlands
Future Publications
Journal Article
Vural Gursel, I., Kockmann, N. & Hessel, V. (2016). Multiphase flow in microstructured
devices - flow separation concepts and their integration into process flow networks.
Chemical Engineering Science, in preparation.
Upon invitation to contribute to Chemical Engineering Science special issue - Multiphase
Microfluidic Engineering
Book
Vural Gursel, I. & Hessel, V. “Modular Compact Intensified Plants”, Wiley-VCH Verlag
GmbH & Co. KGaA, proposal accepted, in preparation.
198
Acknowledgements
In this section I would like to thank deeply the people who have helped and supported me
in my PhD journey and also to those who made it an enjoyable and special ride.
First and foremost I would like to thank my husband Gokalp with all my heart. I know that
without you I would not be able to be in the position that I am today. You have not only
supported my work but also my whole being. You made me laugh when I was sad, calmed
me down when I was upset, nursed me when I was sick, fed me with your delicious food
when I was hungry and encouraged me when I was timid. You have always been there for
me in the 10 years we have been together for better and worse. Thank you for being the
caring, loving, funny, laid back person you are and thank you for being my husband as I
could have not wished for more. I would also like to thank my parents in law Saadet and
Haluk for their gentle heart and support; and for accepting and loving me as I am.
Second I would like to express my sincere gratitude to my dear mother Nazmiye who have
made me the person I am today. Thank you for being the best mom anyone can ask for.
You have always been my rock. You made my life easy so that I could be as successful as I
am now. You have been my personal cleaner, cook, driver, teacher and mentor all at the
same time while your very busy and demanding work schedule. Thank you for loving and
caring for me all the ways that you do. Thank you for still being able to understand and help
me with all my problems. I am so lucky to have you as my mother. I love you so much and
I am sorry for not expressing it enough. I would also like to thank my father for his support
in my success. And lastly I would like to thank my big family; my aunts and cousins for
always being there for me.
My very special thanks go out to my promoter Prof.dr. Volker Hessel. Volker, thank you
for finding me amongst many applications and inviting me to interview for this project. I
appreciate the opportunity you gave me to conduct my PhD research in your group. You
always found the time to give me advice, read and correct my abstracts, presentations,
papers, etc. during your immense workload. I especially want to thank you for your
confidence in me and my capabilities. You have supported and pushed me to grow and
achieve more and therefore form a big part in my accomplishments. You allowed me to
produce many publications and present my work in various conferences. I have learned a
lot from you and will follow these in my academic career.
I would like to thank my supervisor Dr. Qi Wang for the help and support during my
project. Thank you for giving me constructive comments during scientific meetings,
providing feedbacks to improve my papers and recommending new ideas to implement in
my research. Qi, I also would like that you for always welcoming me to ask questions to
you whether for work or personal reasons.
199
I would also like to thank my second supervisor Dr. Timothy Noël. Your supervision was
most valuable in the experimental section of my research. I am grateful for the discussions
that increased the quality of my work. You also improved the writing of my papers with
your corrections and gave advice on how to write high-quality papers. I enjoyed our
personal discussions and I value your recommendations for the future.
I would like to express my gratitude to the European Research Council for providing the
financial support necessary to conduct my research (grant agreement number 267443).
My thanks also goes to the committee members Prof.Dr.-Ing. Norbert Kockmann, prof.dr.ir.
Andrzej Stankiewicz, Dr. Simon Kuhn and prof.dr.ir. Martin van Sint Annaland for
accepting to be in this committee, for taking the time to read my thesis and providing
valuable feedback.
For the collaboration during my research I would like to thank Safa Kutup Kurt and
Prof.Dr.-Ing. Norbert Kockmann from TU Dortmund, Prof. Dr. Krishna Nigam from Indian
Institute of Technology, Prof. Dr. Gunther Kolb, Dr. Johan Tinge from DSM, Dr. Jürgen
Lang from Evonik and Prof. Dr. Andre van Veen from the University of Warwick.
I also had the opportunity to supervise two master students. Ferry and Jasper thank you for
your hard work and providing important input to my thesis. I would like to thank Wim
Gaakeer for allowing me to use his separator and Dr. Patrick Löb from Fraunhofer ICT-
IMM for providing the PTFE membrane separator used in my studies.
I appreciate the support of Carlo, Marlies, Peter and Erik during the experimental work.
Peter thank you for performing the tedious ICP analysis for my research and finding
equipment and supplies based on my requirements. Erik thank you for translating my idea
and rough sketches into computerized design and building up a setup accordingly. You
have the great knowledge of how to connect things together and I appreciate your can do
attitude.
Denise I would also like to thank you for your patience in answering my endless questions
regarding administrative matters and the necessary steps to prepare to my defense. SCR is
very fortunate to have you. I also appreciate the support of our previous secretary Agnes
and current secretary Jose. Jannelies, I will miss our conversations in the corridor and at
your booth in the conferences. I am happy to have met you. I also would like to thank
Martin Timmer for the nice conversation we had on the way you brought me home after a
meeting in Germany.
Stefan my dear officemate, we shared the same office for four years and now it has come to
an end for both of us. Thank you for helping me with my questions about experiments or
life in general in the Netherlands. You are a very nice, giving person and I hope you the
very best in the future. I also wish my new officemates Cecilia and Dario lot of success in
your PhDs.
200
Girls of SFP, Lana, Minjing, Catherine, Elnaz and Smitha and also our former colleagues
Lidia and Ivana, I have enjoyed our little talks. Lana talking to you has always been talking
to one of my friends from Turkey. You are a very strong girl, keep up the good work. Also
to the girls from SCR, thanks for inviting me to the girls dinners. I have enjoyed the food
from around the world and had a great time with the talks and the games. Sorry for not
getting the chance to organize one myself. Dulce thanks for taking the time to answer my
questions about preparation of thesis.
My coursemates Paola, Shohreh and Slavisa, I had a good time with you during the OSPT
courses we had together. I also appreciate your support when my laptop was stolen. Slavisa
thank you for always taking the time to talk to me when we came across in the corridor. It
was nice to share with each other how things are going.
The rest of the people in our SFP group, Carlos, Bhaskar, Hannes, Natan, Nico and
Yuanhai, it was a pleasure having you as my colleagues although we didn’t have the chance
to interact much. Yuanhai thank you for our talks and sharing your experience with me.
Finally I would like to thank the people who have made me feel at home in the Netherlands.
My dear friend Songul and the couple Burcak and Ozgur, you made my life here enjoyable
and enabled me to relax after hard work which was most needed. Thank you for the dinners,
barbecues, stays and all the great time we spent together.
201
About the Author
Iris Vural Gürsel was born on 07.06.1986 in Eskisehir,
Turkey. She finished Koc High School in Istanbul, Turkey in
2004 as class of 2004 Valedictorian with GPA of 5.00/5.00.
She obtained her BSc. degree in Chemical Engineering in
2008 at Bogazici University in Istanbul, Turkey. She
graduated with high honour with GPA 3.86/4.00 ranking 2nd
in the department. She was employed between 2008 and 2011
at Turkish Petroleum Refineries Cooperation (TUPRAS). She
worked at the TUPRAS Izmit Refinery as a Process Engineer
at Technical Services and R&D Department. While working,
she studied MSc in Engineering and Technology
Management at Bogazici University from 2008 to 2010. In 2010, she completed her MSc
with high honour ranked 1st with GPA 4.00/4.00.
In August 2011, she moved to the Netherlands and started her PhD project at Eindhoven
University of Technology in the group of Micro Flow Chemistry & Process Technology.
Her PhD project on Process Simplification and Integration is under the ERC advanced grant
awarded “Novel Process Windows – Boosted Chemistry”. The results of her research are
presented in this dissertation and are currently given as output in 12 peer reviewed journal
papers. She has presented her studies orally or with poster in several national and
international conferences. She was awarded 3rd prize for Best Poster in PRES 2012
conference. She wants to pursue a career in academics and as a next step she is carrying out
postdoctoral research at Utrecht University, Copernicus Institute of Sustainable
Development, group of Energy & Resources to further her career. Her postdoctoral research
is on Sustainability assessment of chemicals and fuels produced from biomass within
CatchBio project.
202

Modular Commercial Plant Thesis

Uploaded by

Copyright:

Available Formats

Modular Commercial Plant Thesis

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Modular Commercial Plant Thesis

Uploaded by

Copyright:

Available Formats

The chemical plant of tomorrow and the future : process-

design intensification at different production scales

Document status and date:

Please check the document version of this publication:

Take down policy

Download date: 25. Feb. 2022

ter verkrijging van de graad van doctor

Iris Vural Gürsel

geboren te Eskisehir, Turkije

voorzitter: prof.dr.ir. J.C. Schouten

Het onderzoek of ontwerp dat in dit proefschrift wordt beschreven is

Chemical Plant of Tomorrow and the Future

A catalogue record is available from the Eindhoven University of

The process-design intensification field is a third means and is based on process

1.2. Flow Separation and Coupling with Flow Reaction

Table 1.1. Review of studies on liquid-liquid slug flow reported in literature

Reference Microchannel Maximum System Separation Application

Sarkar et Cross 72 ml/h Water-succinic - Flow regime

Kashid Y 200 ml/h Water- - Flow

Kashid et Y 120 ml/h Kerosene-acetic Wettability Extraction

1.2.2. Coiled flow inverter

1.2.3. Multi-step flow synthesis

1.3. Process Intensification at Plant Level

1.3.2. End-to-end process integration

Figure 1.8. The number of holistic papers published22

1.4. Scope and Outline of the Thesis

Figure 1.9. A schematic representation of the chapters in the thesis

23. C. Jimenez-Gonzalez, C. S. Ponder, Q. B. Broxterman, and J. B. Manley, Organic

93. S. V. Karnik, M. K. Hatalis, and M. V. Kothare, Journal of

140. S. Klutz, S. K. Kurt, M. Lobedann, and N. Kockmann, Chemical Engineering

164. T. Seifert, S. Sievers, C. Bramsiepe, and G. Schembecker, Chemical Engineering

This chapter is based on :

2.2. Effect of Intensification Fields and Modular Plants

 Reduced size of apparatus

2.2.1. Effect of the transport intensification (1st) field

2.2.2. Effect of the chemical intensification (2nd) field

2.2.3. Effect of the process-design intensification (3rd) field

or exothermic and endothermic reactions are combined in microreactors.56–58 This is already

2.2.4. Effect of the compact, modular plants

Production Strategies.69,70 The full-chain ambition is the realization of ‘‘new, intensified

2.3. Case Scenarios with Stepwise Intensification

2.3.2. Bulk chemistry

14. E. Klemm, H. Döring, A. Geisselmann, and S. Schirrmeister, Chemical

34. N. Kockmann, M. Gottsponer, B. Zimmermann, and D. M. Roberge, Chemistry - A

56. M. S. Mettler, G. D. Stefanidis, and D. G. Vlachos, Industrial & Engineering

76. F3 Factory Newsletter, Issue 1, August 2010,

This chapter is based on :

3.2. Homogeneous Metal Catalyst Separation/Recycle in Flow

Separation methods Advantages Disadvantages

Figure 3.1. Overview of separation/recycling methods of homogeneous metal catalysts in

3.4. Results and Discussion

Model System Feed Solvent Interfacial Tension (mNm-1)

3.4.1.1. Model system of DMF-DCM-water

3.4.2. Continuous copper scavenging using flow liquid-liquid extraction unit

Copper Source Solvent Copper Content (ppm) Extraction

CuBr(PPh3)3 Toluene 123 120 3 98

3.4.2.2. Copper scavenging performance with variation of scavenger type and

Metal Scavenger Concentration Copper Content (ppm) Extraction

EDTA 0.0160 123 120 3 98

Figure 3.7. Copper scavenging performance with variation of scavenger solution pH