1 s2.0 S0306261917304750 Main

Applied Energy 199 (2017) 4568
Contents lists available at ScienceDirect
Applied Energy
journal homepage: www.elsevier.com/locate/apenergy
A review of salt hydrates for seasonal heat storage in domestic

applications
P.A.J. Donkers a, L.C. Sgtoglu a, H.P. Huinink a,, H.R. Fischer b, O.C.G. Adan a,b
a
Technical University Eindhoven, Den Dolech 2, 5600 MB Eindhoven, The Netherlands
b
TNO, De Rondom 1, 5612 AP Eindhoven, The Netherlands
h i g h l i g h t s
We report an evaluation of potential hydrate reactions for heat storage application.

Thermodynamic data of almost 600 hydrate reactions are collected.
A shortlist of 25 TCM hydrate reactions is identified based on thermodynamic data.
Salt hydrates as seasonal heat storage is not realistic for large scale implementation.
a r t i c l e i n f o a b s t r a c t
Article history: A literature review is performed in order to collect and analyse the thermodynamic data of an utmost
Received 6 February 2017 number of salt hydrate reactions (i.e., 563 reactions are reviewed). These data allow us to evaluate the
Received in revised form 18 April 2017 theoretical possibilities and limitations of salt hydrates as thermochemical materials (TCMs) for seasonal
Accepted 26 April 2017
heat storage in the built environment (1 GJ/m3 on system level). Two filters are used for evaluation. Filter
Available online 9 May 2017
1 including three criteria i.e., an ideal hydration reaction with a capacity larger than 2 GJ/m3, a hydration
temperature of 65 C (suitable for domestic hot water) or higher, a dehydration temperature below
Keywords:
100 C to profit as much as possible from the solar heat that can be harvested. Only four of the studied
Hydrates
Database
hydrates fit with these demands. For selecting a larger number of hydrates, a second filter is introduced
TCM with less demanding constraints. It is expected that modifications on heat storage system level are
Solid sorption needed to reach an acceptable system performance with the hydrates selected through filter 2 (hydration
Review reaction with a capacity larger than 1.3 GJ/m3, a hydration temperature of 50 C or higher, a dehydration
Long-term heat storage temperature below 120 C). Based on this filter, a shortlist of 25 TCM hydrate reactions are identified,
including the four of filter 1. The shortlist is analyzed by considering price, chemical stability, reaction
kinetics and safety for domestic environment. Based on this additional analysis with the used constraints,
K2CO3 is determined to be the most promising candidate for open or closed systems, but has a low energy
density. Based on the review of 563 hydrate reactions, we concluded that no ideal salt exists for seasonal
heat storage under the considered boundary conditions. With the current concept of seasonal heat stor-
age, including closed and open systems, whereby only one dehydration cycle per year is performed under
a system energy density of 1 GJ/m3, it is not realistic for large scale implementation to use pure salt
hydrates as heat storage material. By adjusting the constraints, such as multiple cycles per year or higher
water vapor pressures, salt hydrates can still be used as TCMs. It should be mentioned that MgSO47H2O,
MgSO46H2O and CaCl26H2O are not listed in our shortlist of 25 TCMs, although these hydrates are com-
monly suggested in the literature as promising TCM for seasonal heat storage. The present study on pT-
characteristics shows, however, that these salts are not fitting the demands of such a heat storage system.
2017 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).
Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
1.1. Heat storage. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Corresponding author.
E-mail address: h.p.huinink@tue.nl (H.P. Huinink).
http://dx.doi.org/10.1016/j.apenergy.2017.04.080
0306-2619/ 2017 The Author(s). Published by Elsevier Ltd.
This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
46 P.A.J. Donkers et al. / Applied Energy 199 (2017) 4568
1.2. Aim. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
2. System description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3. Collection of data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
4. Selection procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
5. Database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
6. Evaluation of top 25 reactions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
6.1. Energy density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
6.2. Volume variation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
6.3. Hydration temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
6.4. Dehydration temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
6.5. Melting point . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
6.6. Deliquescence vapor pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
6.7. Price . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
6.8. Safety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
6.9. Chemical stability of the complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
6.10. Hydration/dehydration kinetics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
6.11. Short list . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
7. Conclusion and outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
Appendix A. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Appendix B. K2CO3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
Appendix C. Supplementary material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
1. Introduction The amount of reactive gas L inside salt complex MX is called the
loading of the salt. The formation reaction of MXnL is exothermic,
1.1. Heat storage i.e. it produces energy what can be used when for heating purposes.
The enthalpy of this formation reaction is Dr Hm!n
P P
Currently society is moving from carbon-based to more renew- reactant DHi products DHi < 0. The reverse decomposition reaction
able energy sources in order to become less dependent on fossil of MXnL is endothermic, Dr Hm!n DHn!m thus costs energy. This
fuels. A dominant part of the energy consumption of European res- happens during summer heat storage periods. The equilibrium reac-
idential sector (around 70% of the total consumption [1]) is related tion in Eq. (1) implies that the maximum loading of a salt MX at a
to domestic space heating and hot tap water generation. A cleaner temperature T is determined by the vapor pressure of L(g).
sourcing of this part of the energy sector will have a large impact
on the carbon production. For generation of carbon-free heat,
1.2. Aim
new energy production techniques must be implemented, the
majority of which are based on capturing solar radiation. However,
During the past decade, many researchers have investigated
solar radiation fluctuates on different time scales, i.e., hourly, daily
TCMs as heat storage materials. The first generation of salt
and seasonally The power generated fluctuates, resulting in a vari-
hydrates based on TCMs have already been developed, varying
able and unpredictable supply of heat. For matching heat demand
from labscale [911] to field demonstrations [1214]. A complete
and supply, heat storage systems that account for the timescale of
overview of the systems constructed in the last decade is given
radiation fluctuations are required.
by Scapino et al. [6]. A great body of research is also available on
In this article, the focus is on seasonal storage in the built envi-
high potential salts for temperature storage below 100 C, such
ronment in the future, requiring a storage capacity of about 7
as MgSO4 [1519], MgCl2 [1922], SrBr2 [7,23], Na2S [12,24] and
12 GJ in a typical West European dwelling based on the passive
CaCl2 [22,25,26] which have been studied in detail. Storage of heat
house standard (http://www.passivehouse.com/). This storage
for temperature applications between 100 and 300 C already
capacity is based on an average dwelling in the Netherlands with
shows some promising results with salts based on CaO/Ca(OH)2
a floor area of 120 m2, with the passive house standard of
[27] and CaC2O4/CaC2O4H2O [28]. Furthermore, some reviews
15 kWh/m2 for newly built houses and 28 kWh/m2 in renovated
have been published on TCMs [5,2934], that use the energy den-
houses [2,3]. A promising heat storage concept is based on a ther-
sity as selection criterion, with one exception focusing on applied
mochemical reaction, which was suggested by Goldstein [4] in the
working conditions [5] during hydration/dehydration. In the latter
sixties and gained interest in the last decade [57]. The solid mate-
study three salts were selected: MgSO4, LaCl3 and SrBr2 based on
rials involved in these reactions are called thermochemical materi-
dehydration below 105 C and rehydration at 20 mbar vapor pres-
als (TCMs). Key advantages with respect to techniques like sensible
sure at 25 C, which corresponds to the saturated vapor pressure in
heat storage and phase change materials (PCM) include nearly
equilibrium with a water reservoir at 17 C. However, the missing
loss-free storage period and high energy density. In general, a
parameter for selection in this review is the generated temperature
gas-solid equilibrium reaction can be represented by:
T h during the hydration reaction, since this temperature is the
MX nLs
MX mLs n m Lg; 1 highest output temperature the heat battery can deliver.
For introduction of TCMs on the market, it is important that
wherein MXnL(s) is a solid salt complex consisting of a salt MXmL TCMs are able to match the demands of the customers. As a first
(s) and (n-m) mol of reactive gas L. In the current literature reactive indication it is therefore necessary to determine if TCMs can theo-
gas L is considered to be H2O, NH3 or CH3OH. As the targeted heat retically match such demands. In the present work, we attempted
storage system should be used in residential areas, NH3 and CH3OH to analyze and extend the search for pressure-temperature (pT)
are not considered because of currently strict Dutch safety regula- data on the basis of demanded working conditions of a TCM reactor
tion [8]. As a result, H2O is considered a reactive gas in this article. in the built environment, i.e. a system that can store 10 GJ, deliver
P.A.J. Donkers et al. / Applied Energy 199 (2017) 4568 47
hot tap water at 65 C and can be charged in summer with the help externally released/supplied to the system dependent on TCM
of solar panels. The goal is to select TCMs which fulfill the temper- dehydration/hydration.
ature demand and energy density of the consumer, with the help of The working conditions of TCM systems are determined by the
available pT data, and analyse the plausibility of using these TCMs. phase diagram of the TCM in question. A phase diagram indicates
Firstly, the process of data collection will be summarized, secondly the conditions under which a certain TCM undergoes hydration
the selection criteria are explained. Based on these criteria a short or dehydration. In the left bottom corner of Fig. 1, a schematic
list of the most promising salts will be generated and discussed in phase diagram of reaction from Eq. (1) is shown. Applying a condi-
view of the target application. tion (combination of water vapor pressure and temperature) below
the solid line results in a hydrate MXmH2O. For hydrating this
material, a condition should be applied above the solid line. In a
2. System description system, the temperature of the TCM can be varied with the help
of a heat exchanger. Vapor pressure inside the TCM heat storage
Since a heat storage of approximate 10 GJ stored heat is needed system is determined by the temperature of the water compart-
to overcome seasonal fluctuations [35], a system level energy den- ment. Consequently the applied temperature is directly related to
sity of 1 GJ/m3 is considered in this paper, meaning that 10 m3 of the vapor pressure according to the equilibrium line between liq-
reactor should be placed in housing. The system energy density uid water and water vapor (the dotted line). The conditions of
(considering the entire installation, including TCM material, piping, hydration and dehydration can be found by combining both equi-
valves, control devices) is based on the energy density of 5 times librium lines in one diagram.
the energy density of a water tank by a DT of 50 C. Ten cubic In the case of hydration (producing heat), the initial material in
meters of reactor is probably still an overestimation of the avail- the TCM reactor is MXmH2O and the temperature in the system is
able space for such a system as the space is limited i.e. apartments. T w1 , meaning that the vapor pressure inside the system is equal to
In Fig. 1 the concept of a heat storage system with help of a TCM ph . The applied condition around the TCM in that case is above the
is schematically given. On the top, the reactor system is shown, equilibrium line between MXmH2O and MXnH2O. Consequently
where two compartments are drawn, one filled with a TCM and MXmH2O hydrates into MXnH2O. During this reaction the tem-
the other with water, in between these compartments a valve is perature of the TCM reactor will increase as the hydration reaction
located. is an exothermic reaction. As long as the temperature of the TCM
For heat storage, two main concepts are considered, closed and reactor is below T h the hydration reaction will continue. At tem-
open systems [7]. In the case of a closed system both compart- perature T h , both phases (MXmH2O and MXnH2O) can exist as
ments are part of the system and all water necessary for the hydra- that is the equilibrium temperature corresponding to vapor pres-
tion/dehydration reactions is stored within the system. In the case sure ph . The vapor pressure in the system can only remain constant
of an open system, the water is not stored in the system itself, but as the temperature of the water in the system is kept constant. This
Water
TCM source
Open system
Closed system
Winter discharging Summer charging
MXmH2O (s) + (n-m)H2O (g) MXnH2O(s) ; (n>m) MXnH2O(s) MXmH2O (s) + (n-m)H2O (g); (n>m)
log (pwater vapor)
log (pwater vapor)
pd
ph
Tw1 Th Tw2 Td
Temperature Temperature
Fig. 1. Top: A schematic overview of a TCM reactor. Bottom: the reactions involved during discharging and charging of the reactor are given.
means in the case of TCM hydration, the evaporation heat of the are all thermodynamic quantities at p0 = 1 bar and T 0 = 298.15 K.
water should be compensated by heating the water vessel, other- Note that the reaction enthalpy is defined for a certain reaction
wise the temperature of the water vessel decreases and the vapor from an initial (m) to a final (n) hydration state (n > m). For one salt
pressure in the reactor will decrease. With help of a heat exchanger hydrate, different enthalpies of reactions and thus reaction condi-
the reactor can release heat in a controlled way. tions may exist depending on the considered reactions.
Note: we assume that both the temperature of the TCM reactor The uncertainty of the calculated TCM temperature by a given
and water vessel can vary independently from each other. vapor pressure is calculated with help of a 95% interval. Hereby
In the case of dehydration (regenerating the TCM) a certain it is assumed that the error in T; p0 and R are negligible in Eq.
temperature is applied to the TCM T d with a heat exchanger and (2). In the error calculations based on enthalpy/entropy data, the
the reactor is filled with MXmH2O. In that case, the vapor pressure error in enthalpy and entropy is taken from the literature source
applied to the system by the TCM is equal to pd . As long as this or if the source does not mention the error the maximum error is
vapor pressure is higher than the vapor pressure of the water ves- assumed to be five times the unity of the most significant digit
sel, the TCM will dehydrate. Since that will cost energy, the TCM i.e. in case the enthalpy is given by 40.3 kJ/mol, the error is
should be heated in order to maintain the same vapor pressure. assumed to be 0.5 kJ/mol.
The added water in the gas phase will condensate in the water ves- An example of the error calculation is given in Fig. 2. The pT-
sel, which will increase the temperature in the water vessel and data measured by Polyachenok et al. [39] is plotted, including
the equilibrium vapor pressure in the heat storage system. These the calculated data with the help of Eq. (2) and the enthalpy/
losses can be compensated by cooling the water vessel. entropy values of Glasser [38]. Both data sources fit with each
In this case, it is assumed that cooling and heating of the water other. By small variations in equilibrium vapor pressure observed
vessel has the same source: a ground pump. In that case the tem- for CuCl2 based on the calculation with Eq. (2), the temperature
perature in the vessel T w2 will be higher in the summer than in the error of the condenser/evaporator of 14 C is found in the temper-
winter T w1 as a ground pump delivers a higher temperature in the ature range of 0100 C to match with a certain hydration/dehy-
summer. dration temperature. In the case the hydration/dehydration
temperature is known, the error in pressure is on the order of
3. Collection of data 10%. This means that the calculated condenser temperature can
have an error on the order of 5 C.
The thermodynamic data on salts are comprised of the follow- With help of the collected data, various parameters are calcu-
ing parameters: pT-data, crystal densities of the considered lated which are necessary to know for selecting an appropriate
hydrates, reaction enthalpies, entropies and melting points. A total hydrate reaction for the foreseen application. The energy density
of 262 salts (563 reactions) were considered (see Table 1), of which of a system is calculated for an open system (no water storage
the majority of the data originates from The International Critical included in the system) according to the following equation:
Tables [36] and Glasser [37]. In case pT data were not available,
jDH0m!n j n m
they were deduced from the reaction energies [38]. E=Vopen qn ; 3
The basic thermodynamic equation for equilibrium between a
Mn
condensed phase (solid or liquid) and the vapor phase of a pure wherein Mn [kg/mol] is the molar mass of the highest hydrate and
substance, under conditions of low pressure, is used for this inves- qn [kg/m3] the crystal density of the highest hydrate. In case a com-
tigation [38]: bined transition of two or more reactions (k) is considered, the
energy density can be calculated according to:
p DG0m!n DH0m!n DS0m!n
ln ; 2
p0 RT RT R Rik 0
i1 jDH mi !ni j ni mi
E=Vopen qn : 4
DG0m!n DH0m!n T DS0m!n ; Mn
!,
X X For the closed system (water is stored inside the system), the vol-
DH0m!n DH0i DH0i n m; ume of water molecules involved in the reaction is considered
products reactants
!, and the energy density is also calculated by:
X X
DS0m!n DS0i DS0i n m; 1
E=Vclosed E=V open ; 5
products reactants 1 n m Mqnn Mq w
w
wherein R is the gas constant, R = 8.31451 JK1mol1, T [K] the wherein M w [kg/mol] is the molar mass of the liquid water and qw
temperature, p [bar] the vapor pressure, and p0 the reference pres- [kg/m3] the density of liquid water. The energy density of both open
sure. The enthalpy DH0m!n , entropy DS0m!n and Gibbs energy DG0m!n and closed systems depend strongly on the accuracy of the crystal
density and the reaction enthalpy and can be calculated similarly
as for the T TCM . Although extreme accuracy is taken into account,
Table 1
variations of 2% on the energy densities can be feasible.
Overview of the data inside the database. In case the output temperature of the TCM reaction is not suffi-
cient for the targeted heating application, a second heating step is
Considered:
Salts 262
possible. On the basis of the first reaction, water is heated with DT,
Reactions 563 generating a vapor pressure necessary for inducing the second
Based on:
reaction wherein the TCM produces the demanded temperature.
pT-data 169 In that case, the energy density of the system with a double hydra-
Enthalpy 310 tion step E=V openII will drop for an open system according to:
Combined transitions 83
1
Parameters: E=VopenII E=Vopen n o. ;
E=V 397 1 C p DT jDMHwe j n mqn M
Mn
w
E=Vopen
Hydration temperature 563
6
Water
1 Water vapor
CuCl22H2O-CuCl2
Thermodynamic data
Boundary of 100% interval
thermodynamic data
pT-data*
0.1
pwater vapor (bar)

2
0.01 0
1E-3
0.0026 0.0028 0.0030 0.0032 0.0034
-1
T (K ) -1
Fig. 2. An example of a phase diagram composed from pT-data from experiments [39] and pT-data calculated with help of thermodynamic data and Eq. (2) [38]. The numbers
indicate the water loading in the different regions. Between the water vapor line and the CuCl22H2O-CuCl2 line, the solid dihydrate will deliquescence, but the exact
conditions are unknown.
wherein C p [J/(kgK)] is the heat capacity of liquid water and DHe Table 2
[J/mol] the evaporation enthalpy of water. In a closed system the Thermodynamic criteria for selecting hydrates suitable for seasonal heat storage. E=V
energy density changes to: in this table refers the energy density of the TCM, whereby only the volume of the
TCM is considered. A vapor pressure of 12 and 20 mbar corresponds to respectively a

q Mw water temperature of 10 and 17 C.
E=VclosedII E=VopenII 1 n m n 7
M n qw Filter 1 2
E=V [GJ/m3] >2 >1.3
The volume variation during hydration/dehydration is calcu-
Hydration
lated on the basis of the crystal structure density according to: p [mbar] 12 12
T [C] >65 >50
M n qm
DV 1 100%: 8
qn Mm Dehydration
p [mbar] 20 20
The price is calculated based on price per kg Pkg;i , deducted from T [C] <100 <120
the loading i of the salt hydrate indicated by the supplier and T melting > T dehydration > T dehydration
neglecting the cost of water. The price per MJ P MJ is calculated
according to:
to store 10 GJ of heat, in order to overcome seasonal fluctuations
Mi [35,40], within 10 m3. Therefore, an energy density of 2 GJ/m3 on
PMJ P kg;i 106 : 9
jDH0m!n j material level (without considering water storage, open system
configuration) is targeted as the reactor system (piping, valves,
By referring to the current database it is possible to select an
tubing) and material porosity will decreasing the energy density
appropriate hydrate reaction for any application. As the enthalpy
on system level. Beside volumetric considerations, the TCM should
and entropy are known for all reactions in the database, an equilib-
provide temperatures of domestic hot water (DHW) (T > 65 C) and
rium vapor pressure can be found for each TCM temperature and
space heating (HW) (T > 40 C) within one heating step. This tem-
vice versa with the help of Eq. (2). This database will guide the
perature should be reached with a corresponding vapor pressure of
material and system developers during initial screening of suitable
12 mbar, which is equal to an equilibrium vapor pressure of a
hydration/dehydration reactions for particular applications based
water source at 10 C. This 10 C is based on keeping the tempera-
on crucial demands: output temperature (T hydration ), regeneration
ture of the evaporator constant with help of a borehole based on
temperature (T dehydration ) and energy density E=V.
the ground temperature and the ground temperature at 7 m below
surface is approximately 12 C over the year in the Netherlands
4. Selection procedure [41].
Loading of the heat storage system is foreseen to be performed
An appropriate TCM for seasonal heat storage should meet a with heat from solar panels. Different panels are available for
certain set of thermodynamic conditions. These conditions are domestic applications; flat plate collectors and evacuated tubular
listed in Table 2. We used two different set of thermodynamic con- collectors. Depending on the specific collector design, different
ditions (filters) for selecting suitable candidates. These filters temperatures can be reached, where the power output of the col-
reflect a strict and flexible selection criteria, as explained hereafter. lector depends on the output temperature. In case higher output
In addition, an analysis is made on non-thermodynamic conditions temperatures are required the total collector energy output in a
such as price, chemical stability and safety. year will decrease. In this case a dehydration temperature of
Filter 1 selects a TCM which can fulfill the demands of a partic- 100 C is considered, where a low-cost flat plate reactor can gener-
ular heat storage system. This heat storage system should be able ate 1.8 GJ/m2 per year [42] with given output temperature. During
dehydration explained as before, the condenser temperature is a

significant variable, which will rise during condensation. In order
for the dehydration process to continue, the condenser should be
cooled. A temperature of 17 C inside the water drain is considered
as reasonable, provided there is a borehole or air cooling. As a con-
sequence, the TCM should dehydrate at a temperature below
100 C and with a water vapor pressure of 20 mbar. In considera-
tion of mechanical stability, the melting point should be above
the dehydration temperature.
The second filter is introduced as a compromise between the
strict boundary conditions for an ideal salt hydrate and achievable
boundary conditions acknowledged to available salt hydrates. The
energy density is lowered to a value of 1.3 GJ/m3 on material level,
which results in approximately 1 GJ/m3 in a closed system. The
hydration/dehydration temperatures are increased/decreased
respectively. This filter will definitely impact system level design,
e.g., higher dehydration temperatures and additional heating to
reach hot tap water temperatures.
5. Database
The database consists of 563 entries with complementary ther-

modynamic data from multiple sources. The full list is given in
Appendix A. As mentioned before, different dehydration reactions
may occur with one salt. For that reason, multiple hydrate transi-
tions are considered in this database.
A histogram of the energy densities of the available TCM reac-
tions (only 397 records within our database contain information
about energy densities) is plotted in Fig. 3. As can be seen, the his-
togram peaks around 1 GJ/m3. The number of hydrate reactions
with an energy density above 2 GJ/m3 (filter 1) is only 114.
Analyzing the maximum hydration and minimum dehydration
temperature at the conditions stated in Table 2, two histograms
are generated, plotted in Fig. 4A and B, respectively. The shape of
the histogram plotting the hydration temperature peaks around Fig. 4. A histogram of 361 studied hydration reactions showing the maximum
30 C, implying that most salts have a too low output temperature hydration temperature with a vapor pressure of 12 mbar (A) and the minimum
for generating domestic hot tap water under the applied condi- dehydration temperature with a vapor pressure of 20 mbar (B).
tions. Here we stress that hydration reactions of MgSO4 (monohy-
drate to hepta- or hexahydrate at 24 and 21 C, respectively) and
CaCl2 (anhydrous to hexahydrate at 31 C), often considered as The minimum temperatures necessary for the different dehy-
promising salts and extensively investigated [1519,22,25,26], fail dration reactions are given in Fig. 4B. This figure shows that most
in this respect. Reactions with a TCM temperature below 10 C are salts lose water readily at 40 C to a condenser at 20 mbar of water
suspicious and indicated as such in Appendix A. vapor pressure. If the temperature window of filter 1 is applied on
this dataset, 165 salts fit the hydration conditions and 415 salts fit
the dehydration conditions, respectively. Reactions with a TCM
temperature below the 17 C are suspicious and are indicated as
such in Appendix A.
By applying all conditions of filter 1 at once on the dataset, the
number of possible TCM candidates reduces to four: Na2S, LiCl,
EuCl3 and GdCl3. The last two salts are rare earth metals and cannot
be used on a large scale with reasonable prices. Furthermore GdCl3
is toxic (MSDS safety sheet) and LiCl is strongly corrosive [43] and
expensive [44], which makes both candidates unfavorable for
application. The last material is the most promising of these four
materials. It has a high theoretical energy density and the dehydra-
tion temperature is relatively low. On the other hand, Na2S is cor-
rosive [43,45], reactive [46] with the risk of outgassing of H2S
[47,48]) and Na2S is mentioned as a dual use material by the Aus-
tralia Group [49]. Beside Na2S, the US government considers H2S (a
potential outgas of Na2S) as a high priority chemical threat, as well
as being used as a potential weapon of mass destruction by terror-
ists [48]. Experience with Na2S can be found in several heat storage
Fig. 3. A histogram of the energy density of 361 studied hydration reactions. The
projects like TEPIDIUS [12], SWEAT [24] and MERITS [50,51]. These
reaction energy density is calculated on the basis of the molecular volume of the projects all suffered initially from corrosion and faced variable
highest hydrate in the reaction and the enthalpy change. results by overcoming this issue i.e. by coating all surfaces inside
Fig. 5. A selection of the database, which fits the drafted working conditions of filter 1 and/or 2, where the gray shaded area fits the working conditions of filter 1. The
maximum hydration and the minimum dehydration temperature of the different hydrate couples are plotted against the reaction energy density on material level (open
system). The vapor pressure is equal to 20 mbar and 12 mbar during dehydration and hydration, respectively. In addition the lowest melting temperature of the involved
hydrates within the reaction is plotted.
the reactor [51]. Until now, it is unknown how the performance of porosity of 30% in a closed system has a system energy density
the coatings will be on time scales of years. Also concerns are similar to a TCM material with an energy density of 1.8 GJ/m3
raised regarding the release of the last 1.5 water molecules Na2- and equal porosity in an open system. Practically this means, that
S5H2O, whereby the melting temperature and dehydration tem- in case the 1 GJ/m3 at system level is demanded, the material
perature are equal at the current vapor pressure, thereby limiting should be at least above 1.4 GJ/m3 and 2.0 GJ/m3 for an open and
the loading power [24]. In case no complete dehydration is closed system, respectively. Hereby, we did not even consider vol-
reached, the energy density will drop from 2.79 GJ/m3 to 1.58 GJ/ ume for the reactor itself, piping, insulation, valves, control sys-
m3. In addition, local variation in the temperature during dehydra- tems, etc.
tion can cause melting of the TCM, which challenges the stability of Besides different energy densities on system level between
the performance of the heat storage system. open and closed system, another important difference exists
If filter 2 is considered, a shortlist of 25 hydration reactions between these systems. In a closed system, the pressure is mini-
remains. A summary of thermodynamic conditions is given in mized (low-pressure/vacuum), and as a result, the vapor transport
Fig. 5. The hydration and dehydration temperatures are plotted is extremely fast, however the heat transfer from salt to heat
against the energy density for the hydration reactions, as well as exchanger determines the power output. Conversely, in an open
the known melting temperatures of the hydrates. The initial and system, the water transport is slow, but the air blown through
final hydration states are indicated below the salt labels. Table 3 the system can be used as heat conductive medium. As a result,
gives a more detailed overview of the thermodynamic data and for both systems, different issues should be considered for a suffi-
application considerations of these 25 TCMs. In the next section cient output power [62].
a detailed evaluation of this long list will be performed. At this point it is stressed that not all salt hydrates can be used
in an open system due to unwanted side effects. For example, Na2S
should to be used in a closed system to avoid i.e. the transition of
6. Evaluation of top 25 reactions
Na2S into Na2CO3 as Na2 is reactive with CO2 from the atmosphere
[63]. The best performing reactions solely based on the energy den-
Table 3 lists the hydrate reactions, selected on the basis of filter
sity criterion are Na2S (0.55), GdCl3 (06) and EuCl3 (06) in an
2 working conditions. In this section all factors (thermodynamic
open as well as a closed system.
and non-thermodynamic) are evaluated to arrive at a short-list of
TCMs mostly suitable for seasonal heat storage.
6.2. Volume variation
6.1. Energy density
The relative variation of crystal volume during a hydration/
Heat storage system performance only partly depends on the dehydration reaction is shown in Table 3. This number is calculated
material energy density. The effective system energy density is based on the crystal density of highest and lowest hydrates
dependent on the choice of an open or closed system. involved in the studied reaction (both listed in Table 3). Fig. 7
In Fig. 6, the energy density of a pure TCM is plotted against the shows the crystal volume variation of all considered dehydration
energy density of a reactor at different porosities. In all cases an reactions in the database against the energy density involved in
open system leads to a higher energy density, but the difference this reaction. The figure shows a linear trend of 22% volume varia-
between the open and closed system decreases with increasing tion per GJ, meaning that in general the crystal volume varies more
porosity. For example, based on Fig. 6, a TCM of 3.0 GJ/m3 and a in case the energy density of the hydrate increases. Grains are in
52
Table 3
List of the most suitable 25 hydrate reactions based on the working conditions stated in Table 2 with the chosen parameters. The salts are sorted in descending order based on the energy density of an open system. Energy densities open
system of salts whereby an open system is unsafe as a result of toxic gas formation are given in italic and the energy density of the closed system is used for the ranking. The deliquescence point is based on the vapor pressure whereby
the hydrate will form a deliquescence and can no longer transform to a higher hydrate at 25 C. As a reference, the price is given in euro per kg, which is the price of 1 kg of the stable hydrate under ambient conditions. R: rare earth
metal; Tx: acute toxic category 13; T: acute toxic category, : based on the price of LiCl, : not exactly known, : unknown.
Basis Highest Lowest Number Energy Energy Density Density Volume T hydration (C) T dehydration Lowest Deliquescence Price Price LD50 Other Chemical Based on: Point of
hydrate hydrate of density density highest lowest variation with (C) with melting vapor pressure (euro/ (euro/ (mg/ comments stability pT-data/ concern
hydrates open closed hydrate hydrate (%) pw = 12 mbar pw = 20 mbar point at 25 C kg)# MJ) kg) on health Enthalpy
in system system (g/cm3) (g/cm3) in reaction involved (mbar) [53] (H)
reaction (GJ/m3) (GJ/m3) [52] [52] (C) [source]
GdCl3 6 0 2 2.70 1.56 2.48 4.54 61 90 98 R 102 H [37] Rare earth
EuCl3 6 0 2 2.61 1.52 2.42 4.86 65 89 97 R 3527 H [37] Rare earth
CrCl2 3 0 2 2.11 1.31 2 3.59 61 61 68 1870 Instable H [37] Instable
P.A.J. Donkers et al. / Applied Energy 199 (2017) 4568

Cr2+
LiCl 1 0 2 2.08 1.36 1.76 2.09 41 66 72 99 3.5 [54] 37 35.53 1629 pT [36] Price
LiBr 1 0 2 2.01 1.37 2.67 3.51 37 103 110 2 [54] 37 55 1800 H [37] Price
FeCl2 2 0 2 1.93 1.26 2.39 3.24 43 53 59 120 10 895 Instable H [37] Instable
Fe2+
CsF 1 0 2 1.79 1.20 4.67 84 91 >10 N/A H [37] Price
Ca(ClO4)2 4 0 2 1.75 1.17 2.12 2.8 33 92 100 4500 Explosive H [37] Safety
CuCl2 2 0 2 1.74 1.13 2.51 3.39 42 53 59 >150 21 [55] 3 4.33 584 T H [37] pT Price
[39]

Na2S 5 0.5 3 2.79 1.58 1.58 1.86 60 66 82 82 >11 [56] 0.65 0.51 208 H2S pT [56] Safety/Instable
formation
RbF 1 0 2 1.57 1.10 2.9 3.56 30 84 91 >10 N/A H [37] Price
CrCl2 2 0 2 1.57 1.07 2 3.59 57 62 69 1870 Instable H [37] Instable
Cr2+
CaCl2 2 0 3 1.54 1.06 1.84 2.15 35 63 111 176 7.8 [57] 0.29 0.52 1940 pT [57] Deliquescence
and higher
hydrates
Mg(NO3)2 6 2 2 1.53 1.04 1.67 2.03 41 61 68 89 16 5440 Loss of N2 H [37] Instable
LiNO2 1 0 2 1.51 1.07 1.63 94 102 13 [55] 37 39 N/A Loss of N2 H [37] Price
Mg(NO3)2 2 0 2 1.51 1.08 2.03 2.36 31 105 113 16 [54] 5440 Loss of N2 H [37] Instable
LiI 3 1 2 1.49 1.02 2.38 3.18 40 97 95 5.5 [54] 37 59 6500 H [37] Price
LaCl3 7 3 2 1.48 1.03 2.23 2.58 30 66 73 91 R 2370 pT [58] Rare earth
KAl(SO4)2 3 0 2 1.39 1.01 2.2 2.78 35 57 63 92 6986 pT [36] Kinetics [59,5]
MnI2 4 0 2 1.39 0.90 2.88 5.02 53 59 63 80 N/A Tx H [37] Safety
VOSO4 3 1 2 1.35 0.98 2.28 2.5 24 73 80 >10 N/A T H [37] Price/Safety
K2CO3 1.5 0 2 1.30 0.96 2.18 2.33 22 59 65 >150 C 14 [54] 1 1.67 1870 H[37] pT
[36]
MgCl2 6 2 3 1.93 1.24 1.56 1.89 47 61 104 117 10 [54] 0.18 0.14 3800 HCl pT [60,61] Instable
formation
Na2S 5 2 2 1.77 1.17 1.58 1.69 37 66 73 82 >11 [56] 0.65 0.83 208 H2S pT [56] Safety/Instable
formation
Na2S 2 0.5 2 1.60 1.14 1.69 1.86 38 75 82 82 >11 [56] 0.65 1.51 208 H2S pT [56] Safety/Instable
formation
A final note on void spaces within grains should be made

regarding reaction kinetics: cracking may open internal pathways
in grains for water, which improves water transport and thus heat
generation [66]. Thus, a compromise between heat conductivity,
water transport and energy density is inevitable.
The reactions in this list with the least volume variations are
K2CO3 (01.5), VOSO4 (13) and Mg(NO3) (02).
6.3. Hydration temperature
The hydration temperature indicates the maximum tempera-

ture that can be generated in a certain reaction at the given vapor
pressure of 12 mbar. This is important in view of the application, as
the output temperature of the TCM will determine if a salt is only
feasible for domestic heating or also for hot tap water as well. In
case the TCM is insufficient for generating the required tempera-
ture, additional after-heating is necessary. This can be achieved
with electrical heating, but post-heating with a second TCM reac-
Fig. 6. The TCM energy density plotted against the reactor energy density. Different tor is an option as well. This can be accomplished by generation
porosities for open and closed systems are considered. of higher vapor pressure at the evaporator.
The temperatures in Table 3 are calculated for a system with a
water vapor pressure of 12 mbar. In case the vapor pressure
generally dehydrated heterogeneously over the grain and so a increases, i.e. the evaporator temperature increases, the maximum
grain will not shrink in a regular fashion. As a consequence, void reaction temperature will increase as well. So, if part of the heat
spaces will be formed inside the initial grain as locally pieces of generated with the first reactor is added to the evaporator, a higher
the grain contract. The new formed voids will decreases the reaction temperature can be reached, being of interest to reactions
mechanically stability of the original grain. This will results in pul- with output temperatures between 50 and 65 C. The drawback of
verization as the grains are multiple times loaded/unloaded [27]. this approach is the reduction of the effective energy density (see
The pulverization of the grains changes the permeability of the Eqs. (6) and (7)). For example, in case of a 2-step system on CuCl2,
bed and thus the performance of the entire reactor also changes wherein the temperature of the evaporator is increased 13 C to
[14]. In addition, swelling and shrinkage of the grains may induce reach an output temperature of 65 C, the energy density drops
stresses within the heat exchanger. from 1.74 GJ/m3 to 1.21 GJ/m3 and from 1.13 GJ/m3 to 0.72 GJ/m3
Another implication of volume changes is related to the heat for an open and closed system, respectively. This drop in energy
conductivity of the materials. As for many hydrates the thermal density is significant, but if only part of the volume of TCM is used
conductivity is on the order of 0.5 W m1 K1 [64] and that of air for domestic hot water, where the drop in energy density of the
is on the order of 0.024 W m1 K1, where the empty space in entire system is less significant.
between the grains will act as an insulator. Based on the effective The reactions with the highest hydration temperatures are Mg
medium theory, the heat conductivity of a grain with 20% or 40% (NO3)2 (02), LiBr (01) and LiI (13).
void space within the grain will be reduced by 40% and 70%,
respectively [65]. In order to minimize the void space, it is favor- 6.4. Dehydration temperature
able to choose a material with a small volume variation. However,
such smaller volume variation is accompanied with a lower energy The dehydration temperature is the minimum dehydration
density (see in Fig. 7). temperature at which the TCM can dehydrate to the desired load-
ing with a given water vapor pressure of 20 mbar, based on the
condenser temperature in the summer. The temperature inside
the TCM can be accomplished in several ways e.g., solar collectors,
electrical heating, waste heat. The output temperatures of solar
collectors strongly depend on the strength of the solar radiation.
This radiation intensity fluctuates by strength, depending on the
hour of the day, cloudiness and season. Generally a lower dehydra-
tion temperature results in a reduction of solar collector surface for
the production of the desired temperature and power output [42].
Ideally, the dehydration temperature should be kept as low as pos-
sible, to be able to charge the battery multiple times a year. Besides
the appropriate salt selection, this also depends on the condenser
temperature. In practice, the condenser temperature is the temper-
ature of the water reservoir in summer, i.e. 17 C in Western Eur-
ope climate.
Based on the dehydration temperature, FeCl2 (02), CuCl2 (02),
KAl(SO4)2 (312) and MnI2 (04) are the best reactions, i.e. display
the lowest dehydration temperature.
6.5. Melting point

Fig. 7. The volume variation of dehydration reactions plotted against the energy
density of the TCM. The linear fit indicates that on average each stored GJ results in During dehydration of pure hydrates, the melting point deter-
a 22% volume variation of the hydrate. mines the maximum dehydration temperature. If the temperature
exceeds the melting point, the grains will clog together, which tenance and revision, operation, dismantlement and accidents.
results in a change in porosity of the reactor, affecting the power Safety impact refers to health effects or environmental damage
in- and output. Because of this, CaCl2 hexahydrate and tetrahydrate that may be produced by a chemical.
cannot be used as TCM above 40 C as they have melting points of
30 and 45 C, respectively. As a consequence the transition of CaCl2 1. The LD50-values (i.e. Lethal Dosis for 50% of subjects) of the
(02) should be handled with care as during hydration at lower considered salts is listed in Table 3, in which we categorize an
temperatures, the higher hydrates (hexahydrate and tetrahydrate) LD50 below 25 mg/kg as highly toxic; between 25 and 200 as
can be formed. If this is the case, clogging of the grains is still pos- toxic and between 200 and 2000 as harmful [72]. Based on this
sible in case the grains are heated above the melting points, which classification, GdCl3 and NiCl2 are toxic, and Na2S is harmful,
makes it challenging to work with CaCl2 as a TCM. but a borderline case. Note that from the point of view of the
As the power during dehydration is strongly affected by the dif- customer, the LD50 value is merely an indication of health risks,
ference in minimum dehydration temperature and applied tem- as the customer will not directly interact with the materials.
perature [6769], reactions such as CuCl2 (02), K2CO3 (01.5) 2. Special attention is devoted to the following hydrates that are
and FeCl2 (02) are more favorable. explicitly mentioned to be toxic or acute toxic in the MSDS
safety sheets: MnI2, VOSO4 and CuCl2, although this is not
6.6. Deliquescence vapor pressure shown based on the LD50 values.
3. Possible side-reactions due to outgassing or catalytic effect of
Deliquescence is a key parameter for grain stability both in view salts should be addressed case-specifically. Examples are H2S
of hydration at low temperatures and TCM storage. In case of del- and HCl-formation in the case of Na2S and MgCl2, respectively.
iquescence, the grains can clog to each other, changing the local Ca(ClO4)2 is a strong oxidizing agent, which is of major concern
porosity of the TCM. This affects the power which can be deliv- in fires and relevant for working conditions.
ered/absorbed by the reactor as the vapor transport will be
hampered. 6.9. Chemical stability of the complexes
The known deliquescence vapor pressures at 25 C are listed in
Table 3, which is only an indication. For specific applications a dif- A heat battery based on TCMs should reasonably have a service
ferent temperature should be selected. As can be seen, some life of about 1520 years, preferably without replacing the TCM
hydrates will tend to deliquescence in case the hydration vapor material. A stable material performance is desired, i.e. the TCM
pressure is applied when the TCM is still cold (T < 25 C). For exam- should not decompose or transform into another material. Based
ple MgCl2 and CaCl2 can deliquescence during the initial hydration on this criterion, CrCl2 an FeCl2 are eliminated from Table 3, since
process as the TCM is still cold and the vapor pressure of 12 mbar is Cr2+ and Fe2+ are prone to oxidation in a humid environment [73].
applied, which is unwanted. A different oxidation state will affect the pT characteristics, thus
Different strategies for anticipating deliquescence effects can be being unwanted.
used. For example a preheating step can avoid deliquescence or A second point of concern is the decomposition by outgassing.
start at a lower condenser temperature. Furthermore, deliques- For example, MgCl2 is known to produce HCl gas at high tempera-
cence as well as melting in a TCM reactor might be accepted in case tures above 140 C [60], consequently reducing the amount of TCM
a stabilization technique is used to avoid clogging, such as matrix in the reactor. Recent studies indicate that at dehydration temper-
impregnation [70] or microencapsulation by polymers [71]. A con- atures below 100 C during cyclic hydration/dehydration HCl gas is
stant grain structure is desired, to secure a steady heating/dehy- produced [74]. This reduces the amount of TCM and the HCl gas
dration rate of the reactor. may induce corrosion and increases the pressure inside a closed
Based on the deliquescence criteria, CuCl2 (02), Mg(NO3)2 (0 system.
2;26) and LiNO2 (01) are the most suitable reactions. Although outgassing of Na2S is not studied as thoroughly as
MgCl2, it is mentioned by several scientists that Na2S can emit
6.7. Price H2S [47,63], which is strongly corrosive. Another hydrate in the list
that may decompose under the considered working conditions is
The price of the material is a crucial boundary condition affect- Mg(NO3)2 [75].
ing the economic feasibility of any heat storage system. For that For all these hydrates, the stability of the material should be
reason, rare earth metals, like EuCl3 and GdCl3, are thus not consid- considered in view of the service life duration. The rate of chemical
ered. The prices mentioned in this article are based on industrial decomposition will determine whether a hydrate should be
scale produced materials where for the real application bulk mate- rejected or not. Current data on decomposition rates are insuffi-
rial will be used as the base material. We tried to estimate the cient to exclude the above mentioned hydrates, although these
prices of the material as accurately as possible. Hereby, we cooper- salts are unfavorable as a first choice from this perspective.
ated with a distributor and manufacturer within the chemical and
food market [44] to come up with realistic market prices. The price 6.10. Hydration/dehydration kinetics
is given in euro per kg, as a reference, which is the price of 1 kg of
the stable hydrate under ambient conditions. The effective price, The reaction kinetics of a TCM have a strong effect on the power
expressed in euro per joule, determines the actual price per energy. in- and output of a heat battery. Generally increasing the DT
Based on price as an indicator, MgCl2 is currently the most between the applied temperature and the minimum dehydration
promising candidate, with reservation of bulk price fluctuations temperature will increase the dehydration rate [76]. This also holds
for the listed salts in the future. Other good candidates are Na2S for increasing the vapor pressure: increasing the difference
(0.55; 25) and CaCl2 (02). between the applied vapor pressure and the minimum hydration
vapor pressure will increase the hydration rate [77].
6.8. Safety For most hydrates, multiple dehydration data can be found in
literature and in most cases the dehydration data deviate from
Safety is an important aspect in the salt selection procedure, as each other as i.e. the grain size [69], temperature profile [78], sam-
the TCM will be applied in the domestic environment involving all ple holder [79] strongly affect the observed characteristics. In gen-
stages of the heat storage system life cycle, from installation, main- eral, for each material a single study can be conducted to
summarize all available data whereby the history of the grains is As the goal of this review is to find a TCM for domestic applica-
considered. In addition, a similar study can be performed on hydration over a period of 2030 years, the list additionally analyzes
tion experiments although the amount of available literature is other critical parameters: safety, chemical stability, kinetics and
limited, as controlling the vapor pressure during hydration in a price. It turns out that almost each material in this top 25 candi-
TGA/DSC is complicated. date list has its own challenges or foreseen problems. One of the
As we did not perform a study on the hydration/dehydration largest problems is the price: eleven selected materials are disre-
kinetics for each material, it is difficult to distinguish based on garded after price investigations.
kinetics. One exception we made, based on our own lab experi- Finally, a single candidate K2CO3 remains from the list of 25
ments and literature [5,59], KAl(SO4)2 shows very slow (i.e. too candidate salt hydrates under the boundary conditions considered.
slow) hydration reaction kinetics, which we see as problematic K2CO3 has an energy density on material level that is still 6 times
for our application. larger than water with a DT of 50 C and has no heat loss during
storage. By taking into account that the heat storage system also
consumes volume, the energy density will drop significantly. Rela-
6.11. Short list tively small challenges on system level are expected concerning
maintenance, as K2CO3 is not strongly corrosive, no higher hydrates
The final column of Table 3 summarizes the critical points of are known and it has no known unsafe side-reactions. As it does not
concern i.e. potential showstopper for using salt hydrates to store reveal any decomposition reactions, an open system is an option
heat under the considered boundary conditions. For the majority with K2CO3. MgCl2 can be considered as a candidate as well pro-
of the listed salts, concerns arise regarding the material price, sta- vided a solution for HCl-outgassing and deliquescence is found.
bility of the TCM or safety of the system. K2CO3 is the only remain- The 1.3 GJ/m3 energy density of K2CO3 makes it rarely unlikely
ing candidate without points of concern, despite the fact it is the that 10 GJ of heat will be stored in domestic applications, as this
lowest-ranked salt in our shortlist in terms of energy density. Note means that the material consumes almost 8 m3, without consider-
that K2CO3 can operate in an open system without safety or stabil- ing porosity, volume of reactor vessel and heat exchangers. As a
ity issues, contrary to high potential candidates like Na2S and consequence, a heat storage system on the basis of K2CO3 with a
MgCl2. The difference in energy density will be strongly reduced single hydration/dehydration cycle a year will be unlikely based
in that case. In addition, K2CO3 performs sufficiently well consider- on the storage volume. In a multicyclic application, however, the
ing dehydration temperature, melting point, deliquescence point amount of stored heat decreases per cycle, likewise decreasing
and price. As K2CO3 is a relatively new salt in view of thermal heat the amount of material, storage volume and price of the system.
storage, a short overview of the literature is given in Appendix B. Moreover, an increased number of cycles per year appears more
Other salts with potential are MgCl2, assuming HCl production feasible when the dehydration temperature is lower. At lower
can be minimized and deliquescence can be avoided or CuCl2 in dehydration temperatures, a solar system or waste heat can more
case of multiple hydration/dehydration cycles a year. easily deliver the required temperature.
A lower dehydration temperature however, means in general a
lower hydration (i.e. output) temperature as well. This may be
7. Conclusion and outlook anticipated by increasing the evaporator temperature or by serial
application of TCM systems. K2CO3 has a relatively low dehydra-
An extensive review of 563 hydrate reactions is performed, tion temperature, which may facilitate multiple cycles per year.
resulting in a database, wherein the thermodynamic data of these To further assess its potential as a heat storage material additional
reactions are summarized (see Appendix A). With help of the cur- TGA/DSC experiments were performed, which will be published
rent database it is possible to select an appropriate hydrate reac- subsequently as full papers.
tion for any application. For example an industry application will Based on an extensive review of 563 hydrate reactions, no ideal
have different working temperatures, number of cycles and other salt exists for seasonal heat storage under the boundary conditions
safety regulations than applications in households. As the enthalpy considered. With the current concept of seasonal heat storage,
and entropy data of 563 hydrate reactions are listed in this data- including concepts of closed and open systems, whereby only
base, for each reaction an equilibrium vapor pressure can be found one dehydration cycle per year is performed and a system energy
for a given temperature and vice versa. This database will help the density of 1 GJ/m3, it is not realistic for large scale implementation
material and system developers for performing initial screening on to rely on pure salt hydrates as heat storage materials. Although
suitable hydration/dehydration reactions for particular applica- the present view of seasonal heat storage seems unprofitable, mul-
tions based on crucial demands: output temperature (T hydration ), tiple usages per year, utilization in peak shaving [80] or storing
regeneration temperature (T dehydration ) and energy density E=V. waste heat [81] are all promising prospects. In conclusion: chang-
In this study the database is used to evaluate the theoretical ing the concepts behind seasonal heat storage is in our view neces-
potentials and limitations of salt hydrates as thermochemical sary to overcome energy density and price issues, which will
materials (TCM) for heat storage in the built environment under provide TCM with an opportunity to be a valid technique for heat
boundary conditions relevant to (seasonal) use in domestic envi- storage in the future.
ronments. A list of 25 candidate salts is composed from this data-
base based on four criteria, i.e. energy density on material level Acknowledgments
above 1.3 GJ/m3, hydration temperature above 50 C, dehydration
temperature below 120 C and a melting point above the dehydra- We like to thank Henry van der Meer of Caldic for his input on
tion temperature. market prices of promising hydrates in the top 25 list. Part of the
Considering these conditions, commonly suggested salt work in this article is performed within NL Topsector multi-
hydrates like CaCl2 (hexa- and tetrahydrate) and MgSO4 (hepta- annual programme CCO (compact heat storage and conversion)
and hexahydrate) did not fit the demands for seasonal heat storage and the CREATE project in the European Unions Horizon 2020
(for domestic heating and hot tap water using the hydration reac- research and innovation programme under grant agreement No
tion). According to the database, it is impossible to reach the 680450. This paper only reflects the authors view and the Com-
required temperatures during hydration with a reasonable energy mission is not responsible for any use that may be made of the
storage density. information it contains.
Appendix A porosity; DH is the enthalpy of the reaction; DS the entropy of the

reaction; T hyd the maximum hydration temperature by 12 mbar
Although care is taken to minimize errors in the following table, water vapor pressure and T deh the minimum dehydration temper-
some mistakes are found during detailed analysis of the table. The ature by 20 mbar water vapor pressure. The used type of thermo-
data which did not fit lab experiments or seems unrealistic are still dynamic data and source is given in column source and in case
included, but indicated with an asterisk . As not all materials could of the pT data the minimum and maximum temperature of the
be studied in detail, some mistakes will still be unmarked in the used pT data is given in columns T min and T max . For the combined
table. transitions, DH and DS are not given, but can be calculated with
Salt gives the basis on the hydrate in the reaction; H is the high- help of the data of the involved hydrates, which are also given in
est loading in the reaction; L the lowest loading in the reaction; this table. The error of T deh is on the order of 13 C, the error in
E=V is the energy density of the reaction in an open system without E=V is on the order of 2%.
Salt H L E=V DH DS T hy T deh Source T min T max Comment

(GJ/m3) (kJ/mol) (J/(molK) (C) (C) (K) (K)
AgF 4 2 51.6 139 22 28 H [37]
AgF 4 1 22 60 Com
AgF 4 0 22 38 Com
AgF 2 1 56.2 136 53 60 H [37]

AgF 2 0 53 38 Com
AgF 1 0 59.2 158 33 38 H [37]

Al2(SO4)3 6 0 3.19 102.3 150 293 306 H [37]

Al2O3 3 1 51.9 156 2 7 H [37]

Al2O3 3 0 2 133 Com
Al2O3 1 0 81.6 169 126 133 H [37]

Al4C3 6 0 170.9 141 760 785 H [37]
AlCl3 6 0 3.78 89.4 154 204 215 H [37]
AlF3 3 0.5 3.00 78.4 152 145 154 H [37]

As2O5 4 0 50.3 150 2 6 H [37]
AuCl3 2 0 56.5 158 20 25 H [37]
Ba(ClO4)2 3 0 3.30 146.6 138 613 634 H [37]

Ba(IO3)2 1 0 0.25 27.0 115 94 90 H [37]

Ba(N3)2 1 0 0.55 44.3 126 1 7 H [37]
Ba(OH)2 8 7 0.39 43 ? Com
Ba(OH)2 8 3 1.91 58.5 149 43 49 H [37]
Ba(OH)2 8 1.3 2.55 43 ? Com

Ba(OH)2 8 1 2.66 43 39 Com
Ba(OH)2 7 3 1.61 31 49 Com
Ba(OH)2 7 1.3 2.28 31 ? Com
Ba(OH)2 7 1 2.40 58.0 150 31 37 pT [36] 291 338
Ba(OH)2 3 1.3 1.24 33 ? Com

Ba(OH)2 3 1 1.47 57.5 152 33 39 H [37]
Ba(OH)2 1.3 1 0.35 59.0 160 32 38 pT [36] 285 360
Ba(OH)2 1 0 1.08 56.1 127 71 79 pT [36] 350 373
Ba(OH)2 1 0 1.20 60.6 145 61 67 H [37]
Ba(OH)2 0 1 116.0 92 630 658 pT [36] 903 1263 L:BaO

BaAl2O4 1 0 83.7 149 183 193 H [37]

BaBr2 2 1 0.58 50.0 135 19 25 pT [36] 293 333
BaBr2 2 1 0.66 57.3 147 40 46 H [37]
BaBr2 2 1 0.70 61.0 160 38 44 pT [82] 298 378
BaBr2 2 0 1.37 38 42 com

BaBr2 1 0 0.53 40.0 100 21 29 pT [36] 313 348
BaBr2 1 0 0.76 57.8 151 36 42 H [37]
BaBr2 1 0 0.76 58.0 177 20 23 pT [82] 383 398
BaCl2 6 2 43.5 115 14 21 pT [36] 278 283
BaCl2 6 1 14 39 Com
BaCl2 6 0 14 55 Com
Appendix A (continued)


BaCl2 2 1 0.50 39.3 153 41 39 H [37]
BaCl2 2 1 0.67 53.0 140 32 38 pT [36] 273 323
BaCl2 2 1 0.75 59.0 156 33 39 pT [82] 298 373
BaCl2 2 0 1.50 32 55109 Com

BaCl2 1 0 0.88 60.0 100 49 55 pT [36] 298 373
BaCl2 1 0 0.99 67.8 145 101 109 H [37]
BaCl2 1 0 1.17 80.0 190 77 82 pT [82] 373 398
BaI2 6 0 2.21 40.4 100 23 31 pT [36] 293 333
BaI2 2 1 0.55 55.2 147 29 35 H [37]

BaI2 2 0 1.33 66.3 148 29 156 Com
BaI2 1 0 0.86 147 156 H [37]
BeSO4 3 2 0.60 55.7 155 20 26 H [37]
BeSO4 2 1 0.99 70.1 147 109 117 H [37]
BeSO4 2 0 1.97 109 110 Com

BeSO4 1 0 1.27 68.4 146 102 110 H [37]
Ca(ClO4)2 4 0 1.75 64.1 139 92 100 H [37]
Ca(H2PO4)2 1 0 53.2 119 69 78 H [37]

Ca(IO3)2 6 1 1.41 46.4 153 17 14 H [37]

Ca(IO3)2 6 1 1.41 46.4 153 17 14 H [37]
Ca(IO3)2 1 0 0.90 89.9 147 226 237 H [37]
Ca(NO3)2 4 3 0.42 52.5 133 37 44 H [37]
Ca(NO3)2 4 3 0.45 56.0 140 40 47 pT [36] 283 313
Ca(NO3)2 4 3 0.49 61.0 161 34 40 pT [83] 293 315
Ca(NO3)2 4 2 0.86 53.0 135 38 44 pT [83] 293 312

Ca(NO3)2 4 0 1.71 53.0 130 43 50 pT [83] 293 303
Ca(NO3)2 3 2 0.48 55.4 139 43 50 H [37]
Ca(NO3)2 3 2 0.49 55.8 140 43 50 pT [83] 293 324
Ca(NO3)2 3 2 0.52 60.0 140 46 53 pT [36] 288 313
Ca(NO3)2 2 0 1.28 59.5 151 44 51 pT [83] 293 324

Ca(NO3)2 2 0 1.29 60.2 151 48 54 H [37]
Ca(OH)2 0 1 2.68 88.7 107 349 365 pT [36] 574 830 L:CaO
CaBr2 6 0 2.67 61.9 142 74 81 H [37]
CaCl2 6 4a 0.86 54.8 142 35 41 pT [57] 273 301
CaCl2 6 4b 0.98 62.6 166 36 42 pT [57] 273 301

CaCl2 6 4 1.03 66.1 182 32 36 H [37]
CaCl2 6 2 2.06 32 49 Com
CaCl2 6 1 2.58 32 71 Com
CaCl2 6 0 3.10 32 71 Com
CaCl2 4a 2 1.21 60.4 155 43 49 pT [57] 273 316

CaCl2 4 2 1.22 60.8 153 48 54 H [37]
CaCl2 4 1 1.82 48 54 Com
CaCl2 4 0 2.46 48 112 Com
CaCl2 2 1 0.63 50.6 114 63 71 pT [57] 273 327
CaCl2 2 1 0.64 51.2 145 63 71 H [37]

CaCl2 2 0 1.54 63 112 Com
CaCl2 1a 0 1.02 59.0 121 103 112 pT [57] 273 365
CaCl2 1b 0 1.27 73.0 160 102 109 pT [57] 273 363

CaCl2 1 0 1.28 73.8 140 147 156 H [37]
CaC2O4 1 0 1.06 69.5 134 134 144 H [28]

CaHPO4 2 0 1.66 61.6 150 57 64 H [37]
CaSeO4 2 0 56.8 128 72 80 H [37]
CaSiF6 2 0 75.1 149 133 142 H [37]

CaSO3 2 0.5 1.17 52.4 147 15 20 H [37]
CaSO3 2 0 1.64 15 68 Com
(continued on next page)


CaSO3 0.5 0 0.57 62.8 149 65 72 H [37]

CaSO4 2 0.5 1.07 53.6 143 26 32 pT [36] 290 375
CaSO4 2 0.5 1.12 55.4 146 32 37 H [37]
CaSO4 2 0 1.25 46.8 127 14 21 pT [36] L: soluble
CaSO4 2 0 1.29 48.4 129 20 26 pT [36] 288 345 L: natural
CaSO4 0.5 0 0.60 65.0 141 94 102 pT [36] 373 435

CaTeO3 1 0 0.70 70.7 147 113 121 H [37]

Cd(ClO4)2 6 4 0.23 20.0 21 66 98 pT [36] 368 598
Cd(NO3)2 4 2 0.91 57.0 142 47 53 H [37]
Cd(NO3)2 2 0 1.20 57.5 143 48 54 H [37]
CdBr2 4 0 1.82 52.1 145 16 22 H [37]

CdBr2 4 0 1.86 53.3 145 21 27 pT [36] 293 333
CdCl2 2.5 0 1.61 53.4 148 19 24 H [37]
CdCl2 1 0 0.85 53.3 134 40 47 H [37]
CdSO4 2.7 1 1.26 51.5 144 15 20 H [37]
CdSO4 2.7 1 1.30 53.1 147 17 23 pT [36] 298 313

CdSO4 2.7 0 2.25 17 65 Com
CdSO4 1 0 1.08 64.6 158 59 65 H [37]
CdSO4 1 0 3.16 190.0 410 152 155 pT [36] 438 448
Ce(SO4)2 5 0 13.42 384.1 132 2200 2261 H [37]
CeCl3 7 0 2.57 61.0 150 54 60 H [37]

CePO4 2 0 1.57 50.5 151 2 6 H [37]
CH3NH2Al(SO4)2 12 0 2.93 65.8 177 36 41 pT [36] 313 349
CH3NH2Cr(SO4)2 12 0 3.33 74.9 204 39 43 pT [36] 308 352
CH3NH2Fe(SO4)2 12 0 4.14 93.0 273 27 31 pT [36] 296 315
Co(ClO4)2 6 0 1.06 22.0 27 101 118 pT [36] 360 493

Co(NO3)2 6 0 2.16 56.0 139 46 53 H [37]
CoBr2 6 0 2.61 58.9 149 45 51 H [37]
CoCl2 6 2 1.79 55.5 148 29 35 pT [36] 293 325
CoCl2 6 2 1.84 57.0 153 29 35 H [37]
CoCl2 6 1 2.33 29 50 Com

CoCl2 6 0 2.82 29 105 Com
CoCl2 2 1 0.91 60.3 154 44 50 H [37]
CoCl2 2 0 1.83 44 105 com
CoCl2 1 0 1.03 61.1 129 96 105 H [37]
CoSO4 7 6 0.36 52.6 146 16 22 pT [84] 293 316

CoSO4 7 6 0.37 54.2 150 20 25 H [37]
CoSO4 7 0 2.76 20 47 com
CoSO4 6 1 1.64 42.8 113 13 20 pT [84] 288 339
CoSO4 6 0 2.64 57.5 147 41 47 H [37]
Cr2(SO4)3 14 8 1.25 57.7 149 39 45 H [37]

Cr2(SO4)3 14 0 4.81 39 292 com
Cr2(SO4)3 8 0 3.46 99.6 149 279 292 H [37]
CrCl2 4 3 0.57 58.4 150 41 47 H [37]
CrCl2 4 2 1.18 41 68 com
CrCl2 4 0 2.40 41 69 com

CrCl2 3 2 0.71 62.4 150 61 68 H [37]
CrCl2 3 0 2.11 61 69 com
CrCl2 2 0 1.57 62.2 149 62 69 H [37]
Cs2Co(SO4)2 6 2 6.08 348.0 1 26 44 pT [36] 10 302
Cs2CO3 3.5 1.5 1.06 64.5 154 65 72 H [37]

CsAl(SO4)2 12 0 2.31 54.8 150 23 28 H [37]
CsAl(SO4)2 12 0 2.78 65.8 170 46 51 pT [36] 323 378
CsCd(SO4)2 6 2 0.48 34.0 90 30 33 pT [36] 299 338

CsCo(SO4)2 12 0 3.71 122.5 364 33 36 pT [36] 300 312
CsCr(SeO4)2 12 0 3.08 73.0 196 43 48 pT [36] 323 368

CsCr(SO4)2 12 0 1.67 39.6 103 14 22 pT [36] 287 376
CsF 1 0 1.80 65.9 148 84 91 H [37]
CsFe(SO4)2 6 4 0.56 64.0 175 31 36 pT [36] 313 363
CsMg(SO4)2 6 2 0.80 45.6 120 19 26 pT [36] 303 341
CsNi(SO4)2 6 2 0.94 53.7 132 45 52 pT [36] 308 355

CsTi(SO4)2 12 0 1.70 56.1 152 24 30 pT [36] 298 368
CsV(SO4)2 12 0 2.01 66.5 178 38 43 pT [36] 323 363
CsZn(SO4)2 6 2 1.11 63.7 166 41 47 pT [36] 308 344

Cu(ClO4)2 8 4 0.29 13.0 1 111 136 pT [36] 355 472
Cu(IO3)2 1 0 60.2 154 43 49 H [37]

Cu(NO3)2 6 0 58.0 152 36 41 H [37]

Cu3(PO4)2 3 0 46.2 20 618 649 H [37]
CuBr2 4 0 2.09 54.8 148 26 31 H [37]
CuCl2 2 0 1.75 58.0 143 50 56 pT [39] 300 415
CuCl2 2 0 1.74 59.1 145 53 59 H [37]

CuF2 2 0 3.04 71.3 152 105 113 H [37]

CuHPO4 1 0 0.63 35.2 24 563 595 H [37]
CuSeO3 2 0 1.30 44.1 111 26 34 H [37]
CuSO4 5 3 1.01 55.3 148 28 34 pT [82] 298 369
CuSO4 5 3 1.03 55.9 149 30 35 H [37]

CuSO4 5 1 2.05 30 40 Com
CuSO4 5 0 2.56 30 151 Com
CuSO4 3 1 1.42 56.6 149 34 40 pT [82] 298 390
CuSO4 3 0 2.30 35 151 Com
CuSO4 1 0 0.93 49.4 81 140 151 pT [36] 303 348

DyCl3 6 0 2.87 70.5 143 121 129 H [37]
DyPO4 2 0 107.5 151 322 335 H [37]
ErCl3 6 0 2.94 71.5 147 117 126 H [37]
ErPO4 2 0 99.2 151 271 283 H [37]

Eu(IO3)3 2 0 23.2 168 2466 2570 H [37]
EuCl3 6 0 2.61 65.8 145 89 97 H [37]

EuPO4 2 0 1.45 49.0 153 7 3 H [37]
FeBr2 2 1 0.52 50.0 120 71 79 pT [36] 383 403
FeCl2 4 2 1.11 56.1 149 31 36 H [37]
FeCl2 4 0 2.22 31 59 Com

FeCl2 2 1 0.91 62.0 144 73 80 pT [36] 363 398
FeCl2 2 0 1.93 65.8 165 53 59 H [37]
FeCl3 6 0 2.51 62.2 154 53 60 H [37]
FeI2 6 3.5 43.4 101 43 52 pT [36] 363 393
FePO4 6 0 2.17 53.8 150 18 23 H [37]

FeSO4 7 4 1.07 52.0 142 19 25 pT [85] 294 321
FeSO4 7 4 1.10 53.8 147 22 28 H [37]
FeSO4 7 1 2.20 19 26 Com
FeSO4 4 1 1.64 53.4 147 20 26 H [37]
FeSO4 4 0 2.29 20 112 Com

FeSO4 1 0 1.24 71.1 152 104 112 H [37]
GdCl3 6 0 2.70 67.4 149 90 98 H [37]
GdPO4 2 0 111.7 151 348 362 H [37]
H2C2O4 2 0 1.38 52.8 131 42 49 pT [36] 248 343
H2O 50.0 141 11 17 H [37] solidgas

H2SeO3 1 0 1.24 53.3 129 49 56 pT [36] 294 384
H2SO4 6.5 4 47.0 120 28 35 H [37]


H2SO4 4 3 49.0 120 41 48 H [37]
H2SO4 3 2 51.5 120 56 64 H [37]
H2SO4 2 1 57.7 124 87 95 H [37]

H2SO4 1 0 71.8 134 150 160 H [37]
H3PO4 0.5 0 62.8 152 60 66 H [37]
HIO3 1 0 1.97 75.0 201 42 47 pT [36] 293 373
HoCl3 6 0 70.2 148 108 116 H [37]
HoPO4 2 0 109.6 151 335 349 H [37]

K2C2O4 2 0 1.21 51.0 124 44 51 pT [36] 323 373
K2C4H4O6 0.5 0 53.0 150 21 23 pT [36] 287 313
K2CO(SO4)2 6 2 61.0 167 35 39 pT [36] 300 349
K2CO3 2 0 1.56 62.5 151 62 69 pT [36] 363 393 dihydrate not observed
K2CO3 1.5 0.5 0.84 63.6 155 59 65 H [37]

K2CO3 1.5 0 1.30 59 65 Com Direct transition
K2CO3 0.5 0 0.52 70.2 157 89 97 H [37]
K2CuCl4 2 0 59.7 152 44 50 H [37]
K2Fe(SO4)2 6 4 0.48 44.0 118 12 19 pT [36] 298 345
K2Mg(SO4)2 6 4 0.44 43.3 116 12 19 pT [36] 300 344

K2Mn(SO4)2 6 2 0.83 43.9 117 13 20 pT [36] 328 372
K2Ni(SO4)2 6 2 0.95 46.0 112 38 45 pT [36] 307 370

K2SO3 1 0 45.6 147 16 12 H [37]
K2Zn(SO4)2 6 2 1.12 55.2 146 30 36 pT [36] 289 337

K3PO4 7 0 49.9 148 2 7 H [37]
K4[Fe(CN)6] 3 0 0.69 52.8 142 24 30 pT [36] 288 353
K4Fe(CN)6 3 0 0.66 50.2 112 66 74 H [37]
K4P2O7 3 0 1.05 60.7 147 58 64 H [37]
KAl(SO4)2 12 3 1.87 56.0 147 33 39 H [37]

KAl(SO4)2 12 0 2.68 60.3 160 7 8 pT [36] 293 353
KAl(SO4)2 3 0 1.31 61.8 152 55 61 H [37]
KAl(SO4)2 3 0 1.39 65.9 163 57 63 pT [36] 303 388
KCr(SO4)2 12 6 52.4 141 22 28 pT [36] 283 352
KF 2 0 2.45 57.0 144 43 49 H [37]
KOH 2 1 1.10 56.7 147 37 43 H [37]

KOH 2 0 2.19 37 203 Com
KOH 1 0.8 0.12 17.9 8 117 141 pT [36] 301 315
KOH 1 0 2.30 87.1 153 193 203 H [37]
LaCl3 7 3 1.48 61.7 145 66 73 H [38,58]
LaCl3 7 1 2.31 66 114 Com

LaCl3 7 0 2.41 48 158 Com
LaCl3 3 1 1.19 68.9 145 106 114 H [38,58]
LaCl3 3 0 1.81 106 158 Com
LaCl3 1 0 0.83 72.5 136 148 158 H [37]
LaPO4 2 0 3.69 113.8 140 395 411 H [37]

Li2SO4 1 0 0.90 55.1 136 48 55 pT [36] 333 381
Li2SO4 1 0 0.92 57.2 139 53 60 H [37]

LiBr 2 1 1.02 56.7 185 2 1 H [37]
LiBr 2 1 1.21 67.5 171 52 57 H [36]
LiBr 2 0 2.03 52 110 Com

LiBr 1 0 1.77 69.5 153 94 101 H [37]
LiBr 1 0 2.01 78.9 173 103 110 H [36]
LiCl 3 2 0.58 44.6 117 17 24 pT [36] 238 331
LiCl 3 1 1.41 17 37 Com
LiCl 3 0 2.22 17 72 Com

LiCl 2 1 1.22 64.0 173 32 37 pT [36] 273 338

LiCl 2 0 2.43 17 72 Com
LiCl 1 0 1.83 62.9 145 73 80 H [37]
LiCl 1 0 2.08 71.3 173 66 72 pT [36] 303 375
LiClO4 3 1 1.44 60.9 146 61 67 H [37]

LiClO4 3 0 2.16 61 59 Com
LiClO4 1 0 1.13 64.0 160 53 59 H [37]
LiI 3 2 0.75 58.9 123 97 105 H [37]
LiI 3 2 0.90 71.2 168 75 81 pT [36] 344 385
LiI 3 1 1.49 75 95 Com

LiI 3 0 2.24 75 151 Com
LiI 2 1 0.90 58.9 128 85 93 H [37]
LiI 2 1 1.09 71.1 161 88 95 pT [36] 350 401
LiI 2 0 1.80 88 151 Com
LiI 1 0.5 0.90 86.2 174 136 143 pT [36] 403 451
LiI 1 0 1.64 78.3 153 142 151 H [37]
LiNO2 1 0 1.51 65.9 143 94 102 H [37]
LiNO3 3 0 2.13 55.6 149 28 34 H [37]
LiOH 1 0 64.3 161 52 58 H [37] l:Li2O
LiOH 0 1 106.7 91 563 589 pT [36] 793 1197 l:Li2O

LuCl3 6 0 65.5 151 76 83 H [37]
LuPO4 2 0 82.4 151 171 180 H [37]
Mg(ClO4) 6 4 0.74 61.7 127 105 114 pT [86] 273 398
Mg(ClO4) 6 2 1.49 105 133 Com
Mg(ClO4) 6 0 2.45 105 225 Com

Mg(ClO4) 4 2 0.80 61.1 118 123 133 pT [86] 273 398
Mg(ClO4) 4 0 1.61 123 225 Com
Mg(ClO4) 2 0 1.33 77.4 124 213 225 pT [86] 273 473
Mg(NO3)2 6 2 1.53 58.6 139 61 68 H [37]
Mg(NO3)2 6 0 2.29 61 113 Com

Mg(NO3)2 2 0 1.51 68.5 145 105 113 H [37]
Mg(OH)2-MgO 0 1 37.0 87 25 35 pT [36] 308 444 l:MgO
MgBr2 6 1 1.74 50.0 135 19 25 pT [36] 293 333
MgBr2 6 0 3.02 72.5 142 135 144 H [37]
MgBr2 1 0 1.19 69.0 145 109 116 pT [36] 393 438

MgCl2 6 4 0.80 52.0 118 65 72 pT [61] 293 367
MgCl2 6 4 0.85 55.5 130 61 68 pT [87] 305 403
MgCl2 6 4 0.89 58.2 138 61 68 H [37]
MgCl2 6 2 1.93 61 104 Com
MgCl2 6 1 2.48 61 127 Com

MgCl2 6 0 3.12 61 214 Com
MgCl2 4 2 1.25 64.6 140 93 101 pT [61] 293 405
MgCl2 4 2 1.32 67.8 147 96 104 H [37]
MgCl2 4 1 2.01 96 127 Com
MgCl2 4 0 2.82 96 214 Com

MgCl2 2 1 1.01 70.3 143 117 125 H [60]
MgCl2 2 1 1.03 71.3 146 118 127 H [37]
MgCl2 2 0 2.23 118 214 Com
MgCl2 1 0 83.5 141 203 214 H [37]
MgCO3 5 3 53.0 146 20 25 H [37]

MgCO3 3 1 51.9 145 15 21 H [37]
MgSeO3 6 0 58.4 152 38 43 H [37]
MgSeO4 6 4 0.74 52.7 147 17 22 H [37]
MgSeO4 4 1 56.4 147 35 41 H [37]
MgSeO4 1 0 1.52 88.4 84 421 440 H [37]


MgSO3 6 3 55.9 142 41 47 H [37]
MgSO3 3 0 64.1 150 70 77 H [37]
MgSO4 7 6 0.40 59.6 166 24 29 H [38,88]
MgSO4 7 4 1.13 55.6 145 24 29 H [38,88]
MgSO4 7 1 2.27 56.0 141 24 28 H [38,88]

MgSO4 6 5 0.37 49.7 144 6 11 H [38,88]
MgSO4 6 4 0.80 53.4 138 24 30 H [38,88]
MgSO4 6 1 2.08 55.2 137 22 28 H [38,88]
MgSO4 5 4 57.8 145 46 52 H [38,88]
MgSO4 4 2 1.23 58.8 143 55 61 H [37]

MgSO4 2 1 41.6 148 33 29 H [37]
MgSO4 1 0 2.00 107.9 154 315 329 H [37]
MgTeO3 6 5 52.7 147 17 22 H [37]
MgTeO3 5 0 61.0 147 59 66 H [37]
Mn(ClO4)2 6 4 20.0 18 103 127 pT [36] 373 470

MnBr2 6 4 76.5 139 166 176 H [37]
MnBr2 4 1 1.16 41.7 101 32 40 pT [36] 293 373
MnBr2 4 0 2.22 59.7 150 47 53 H [89]
MnBr2 1 0 76.0 163 111 117 pT [36] 413 433
MnCl2 4 2 1.05 51.0 134 27 33 pT [36] 293 353

MnCl2 4 2 1.14 55.7 147 32 37 H [37]
MnCl2 4 1 1.79 32 74 Com
MnCl2 4 0 2.45 32 122 Com
MnCl2 2 1 0.89 63.6 150 68 74 H [37]
MnCl2 2 0 1.78 68 122 Com

MnCl2 1 0 1.10 63.5 128 114 122 H [37]

MnF2 4 50.2 147 5 10 H [37]
MnI2 4 2 0.83 55.2 130 59 67 pT [36] 323 383
MnI2 4 1 1.11 59 63 Com
MnI2 4 0 1.39 59 63 Com

MnI2 2 1 0.38 37.0 77 52 63 pT [36] 339 393
MnI2 1 0 0.53 43.1 95 53 63 pT [36] 393 406
MnSeO4 1 0 78.2 82 357 374 H [37]

MnSO4 7 5 0.74 51.0 149 6 11 H [37]

MnSO4 4 1 1.58 52.0 150 10 14 H [37]
MnSO4 2 1 46.3 124 16 22 pT [36] 282 333
MnSO4 1 0 0.79 45.1 108 40 48 pT [36] 293 333
MnSO4 1 0 1.22 69.5 146 108 116 H [37]
MoO3 1 0 85.6 107 319 334 H [37]
Na0.96Al0.96Si2.04O6 0.5 0 84.5 134 229 240 H [37]

Na2B4O7 10 0 55.5 149 28 33 H [37]
Na2B4O7 5 0 72.8 192 46 51 pT [36] 333 338
Na2C2H4O6 2 0 44.5 119 13 20 pT [36] 282 322
Na2CO3 10 7 0.78 51.0 140 16 22 pT [36] 248 305
Na2CO3 10 7 0.84 54.6 141 35 42 H [37]

Na2CO3 7 1 2.08 51.6 146 13 18 H [37]
Na2CO3 1 0 1.07 58.6 155 33 39 pT [36] 293 371
Na2CO3 1 0 1.08 59.4 156 36 42 H [37]
Na2CrO4 4 0 54.5 143 32 38 H [37]
Na2HAsO4 46.4 120 24 30 pT [36] 273 373

Na2HPO4 12 7 0.86 40.0 99 22 28 pT [36] 248 323
Na2HPO4 12 7 1.15 53.7 149 19 24 H [37]
Na2HPO4 7 2 1.56 49.8 132 22 28 pT [36] 273 372
Na2HPO4 7 2 1.66 53.2 146 21 26 H [37]

Na2HPO4 2 0 57.2 156 26 31 H [37]
Na2S 9 5 1.32 55.3 148 27 33 pT [56] 278 317
Na2S 5 2 1.77 62.9 149 66 73 pT [56] 288 356
Na2S 5 0.5 2.79 66 82 Com
Na2S 2 0.5 1.60 72.0 171 75 82 pT [56] 278 353
Na2S2O3 5 0 1.91 54.0 145 26 32 H [37]

Na2S2O3 5 0 1.95 54.9 144 32 38 pT [36] 288 313

Na2SiO3 9 8 53.1 174 7 4 H [37]
Na2SiO3 8 6 53.8 147 22 28 H [37]
Na2SiO3 6 5 51.9 139 24 30 H [37]
Na2SiO3 5 0 52.6 135 34 41 H [37]

Na2SO3 7 0 52.1 145 16 22 H [37]
Na2SO4 10 7 0.77 52.1 145 16 22 H [37]
Na2SO4 10 0 2.48 50.5 139 15 21 pT [90] 273 306
Na2SO4 7 0 2.21 53.7 151 16 21 H [37]
Na2WO4 2 0 55.3 148 28 34 H [37]

Na3PO4 12 0 55.1 148 27 33 H [37]
Na4P2O7 10 0 54.1 147 24 29 H [37]
NaAl(SO4)2 12 0 2.40 54.9 151 21 27 pT [36] 304 332
NaAl(SO4)2 12 2.32 53.1 H [37]
NaAl(SO4)2 12 2.44 55.9 H [37]

NaBr 2 0 1.67 53.1 145 21 27 H [37]
NaBr 2 0 1.69 53.7 143 27 33 pT [91] 283 324
NaC2H3O2 3 1.5 0.94 58.7 157 30 36 pT [36] 288 323
NaC2H3O2 3 0 30 36 Com
NaC2H3O2 1.5 0 1.22 58.0 152 34 40 pT [36] 288 308

NaClO2 3 0 2.00 55.9 149 30 35 H [37]

NaClO4 1 0 0.71 47.8 147 6 2 H [37]
NaCN 2 0.5 1.30 54.1 148 22 28 H [37]
NaCN 2 0 1.75 22 33 Com
NaCN 0.5 0 0.71 55.2 148 28 33 H [37]

NaCr(SO4)2 12 0 2.30 55.2 146 32 37 pT [36] 324 363
NaHC2O4 3 0 51.8 134 30 37 pT [36] 248 323
NaHSO4 1 0 52.3 134 34 41 H [37]
NaI 2 0 1.47 55.2 139 42 49 pT [36] 295 333
NaI 2 0 1.53 57.3 151 34 39 H [37]

NaIO3 5 1 1.78 51.4 149 8 13 H [37]

NaIO3 5 0 2.23 8 105 com
NaIO3 1 0 1.10 68.4 148 98 105 H [37]
NaKC4H4O6 4 0 56.7 153 26 31 pT [36] 288 313
NaOH 1 0 1.46 50.0 100 93 102 pT [36] 273 323 h: unclear

NaOH 1 0 1.97 65.3 169 45 50 H [37]
Nd2(SO4)3 2 0 83.2 141 201 212 H [37]
NdCl3 6 0 61.7 150 58 64 H [37]
NdPO4 2 0 111.7 136 395 411 H [37]
(NH2OH) Al(SO4)2 12 0 70.0 180 31 36 pT [36] 300 348

(NH2OH) Cr(SO4)2 12 0 70.0 180 31 36 pT [36] 303 345
(NH4)2Co(SO4)2 6 2 0.94 62.0 160 41 47 pT [36] 308 355
(NH4)2Mg(SO4)2 6 4 58.0 152 34 40 pT [36] 302 341
(NH4)2Mn(SO4)2 6 2 47.0 120 19 25 pT [36] 339 363
(NH4)2Zn(SO4)2 6 2 1.22 64.0 170 46 52 pT [36] 304 353

(NH4) V(SO4)2 12 0 2.59 59.0 160 33 39 pT [36] 323 363
NH4(UO2)2F5 4 3 48.5 131 18 24 H [37]
NH4(UO2)2F5 4 0 48.5 131 18 55 H [37]


NH4(UO2)2F5 3 0 17.2 20 31 55 H [37]
NH4Al(SO4)2 3 1 65.7 174 41 46 pT [36] 323 362

NH4Co(SO4)2 12 0 4.15 95.6 293 18 21 pT [36] 288 309
NH4Cr(SO4)2 12 0 2.44 56.3 150 30 36 pT [36] 323 359
NH4Fe(SO4)2 12 0 4.00 90.0 260 28 32 pT [36] 298 314
Ni(ClO4)2 17.9 15 106 132 pT [36] 357 541

Ni(IO3)2 2 0 0.75 41.6 168 33 31 H [37]
Ni(NO3)2 6 58.2 136 64 72 H [37]
NiCl2 6 4 0.72 46.0 119 23 30 pT [36] 288 308
NiCl2 6 4 0.80 51.3 138 22 28 H [37]
NiCl2 6 2 1.73 22 37 Com
NiCl2 6 0 2.70 22 76 Com

NiCl2 4 2 0.81 47.0 119 30 37 pT [36] 288 339
NiCl2 4 2 1.03 59.8 160 33 38 H [37]
NiCl2 4 0 2.09 33 76 Com
NiCl2 2 0 1.97 62.2 145 69 76 H [37]
NiSO4 7 6 0.36 51.7 144 16 21 H [37]

NiSO4 7 4 16 ? Com
NiSO4 7 1 16 45 Com
NiSO4 7 0 16 571 Com
NiSO4 4 1 1.82 57.7 149 39 45 H [37]
NiSO4 4 0 2.31 54.8 143 39 571 H [37]

NiSO4 1 0 0.90 46.3 125 14 20 pT [36] 288 348
NiSO4 1 0 2.73 140.3 144 551 571 H [37]

Pb(CH3CO2)2 3 0 16.0 7 170 195 pT [36] 288 295
PrCl3 7 6 64.4 150 72 79 H [37]
PrCl3 6 0 61.6 151 55 62 H [37]

PtCl4 5 0 62.9 163 43 48 H [37]
PuCl3 6 0 60.5 146 58 65 H [37]

PuO2(OH)2 1 0 34.1 103 31 24 H [37]
Ra(BrO3)2 1 0 60.2 143 62 69 H [37]

Ra(IO3)2 1 0 51.2 154 2 6 H [37]
RaBr2 2 0 59.7 154 41 47 H [37]
RaCl2 2 0 57.7 151 36 41 H [37]
RaI2 0.5 0 70.2 143 119 127 H [37]
Rb2Cd(SO4)2 6 2 44.6 119 14 20 pT [36] 301 341
Rb2Co(SO4)2 6 2 1.16 60.0 156 38 44 pT [36] 301 349

Rb2CO3 3.5 1.5 0.87 51.3 153 3 8 H [37]

Rb2CO3 3.5 1 1.17 3 105 Com

Rb2CO3 3.5 0 1.73 3 81 Com
Rb2CO3 1.5 1 0.42 69.5 151 98 105 H [37]
Rb2CO3 1.5 0 1.23 98 81 Com

Rb2CO3 1 0 0.87 66.5 155 74 81 H [37]
Rb2Mg(SO4)2 6 4 0.51 53.1 141 27 33 pT [36] 302 340
Rb2Mn(SO4)2 6 2 47.2 127 17 23 pT [36] 301 368
Rb2Ni(SO4)2 6 2 49.1 117 48 56 pT [36] 306 388
Rb2Zn(SO4)2 6 2 1.27 65.3 170 44 49 pT [36] 306 355

RbAl(SO4)2 12 0 2.47 57.0 148 37 43 pT [36] 323 370
RbAl(SO4)2 12 0 2.50 57.7 149 39 45 H [37]
RbCr(SO4)2 12 0 56.0 150 29 35 pT [36] 313 363
RbF 1 0 1.57 66.2 149 84 91 H [37]
RbFe(SO4)2 12 0 3.80 87.0 250 31 35 pT [36] 298 329

RbOH 2 1 1.12 62.8 147 69 76 H [37]
RbOH 2 0 69 211 Com

RbOH 1 0 2.08 88.3 153 200 211 H [37]
RbTi(SO4)2 12 0 68.3 180 43 48 pT [36] 313 378
RbV(SO4)2 12 0 57.1 153 30 35 pT [36] 326 364

ScCl3 71.2 149 111 119 H [37]
SmCl3 6 0 65.6 150 78 85 H [37]
SmPO4 2 0 115.9 145 393 408 H [37]
Sr(BrO3)2 1 0 138.9 147 532 551 H [37]
Sr(IO3)2 6 1 1.97 53.6 156 10 14 H [37]

Sr(IO3)2 6 0 2.36 10 45 Coms
Sr(IO3)2 1 0 0.52 52.0 131 38 45 H [37]
Sr(NO3)2 4 0 51.3 145 12 18 H [37]
Sr(OH)2 8 2 1.63 38.0 101 18 24 pT [36] 288 363
Sr(OH)2 2 1 0.84 41.7 100 35 43 pT [36] 327 373

Sr(OH)2 0 1 2.39 98.0 97 570 575 pT [36] 703 1051 L:SrO
SrBr2 6 1 1.94 56.5 141 45 52 pT [82] 298 360
SrBr2 6 1 1.99 57.9 144 48 54 H [7,38]

SrBr2 6 0 1.77 43.0 96 52 61 pT [36] 293 313
SrBr2 6 0 2.49 48 122 Com

SrBr2 1 0 1.06 72.8 152 114 122 H [37]
SrCl2 6 2 1.58 53.7 143 27 33 pT [82] 298 333
SrCl2 6 2 1.61 54.6 146 28 33 H [37]
SrCl2 6 2 1.87 63.7 175 28 34 pT [36] 248 323
SrCl2 6 1 2.04 28 52 Com

SrCl2 6 0 2.99 28 202 Com
SrCl2 2 1 0.83 58.9 148 46 52 pT [82] 298 403
SrCl2 2 1 0.83 59.4 91 189 202 H [37]
SrCl2 2 0 1.54 55.0 138 45 52 pT [36] 293 373

SrCl2 1 0 1.01 61.1 184 12 15 H [37]
SrTeO3 1 0 64.0 147 75 83 H [37]
TbCl3 6 0 70.2 148 108 116 H [37]
TbPO4 2 0 103.4 151 297 310 H [37]
Tc2O7 1 0 45.8 102 58 67 H [37]

Th(NO3)4 5 4 47.3 148 10 5 H [37]

Th(NO3)4 5 0 10 155 Com
Th(NO3)4 4 0 76.4 147 146 155 H [37]
ThF4 2.5 0 60.6 154 45 51 H [37]
Tl(Al(SO4))2 12 0 68.9 184 41 46 pT [36] 324 362
Tl(Cr(SO4))2 12 0 58.6 157 31 36 pT [36] 306 375

TlOH 0.5 0 52.4 126 53 60 pT [36] 319 413
TmPO4 2 0 103.4 151 297 310 H [37]
U(SO4)2 8 4 53.0 144 22 28 H [37]
U(SO4)2 4 0 51.6 144 15 21 H [37]
UF4 2.5 0 61.1 144 65 72 H [37]

(UO2)3(PO4)2 52.8 15 20 H [37]
(UO2)3(PO4)2 70.1 116 124 H [37]
UO2(NO3)2 6 3 53.9 143 28 34 H [37]
UO2(NO3)2 3 2 59.9 148 52 58 H [37]
UO2(NO3)2 2 1 72.9 147 125 134 H [37]

UO2(NO3)2 1 0 71.2 144 122 131 H [37]
UO2(OH)2 1 0 51.4 135 28 34 H [37]
UO2Br2 3 1 59.2 144 55 62 H [37]
UO2Br2 1 0 76.7 144 155 164 H [37]
UO2Cl2 3 1 60.7 149 54 61 H [37]

UO2Cl2 1 0 74.5 147 135 143 H [37]


UO2F2 3 0 51.8 144 16 22 H [37]
UO2SO4 3.5 3 58.4 164 21 26 H [37]

UO2SO4 3 2.5 47.2 133 8 13 H [37]
UO2SO4 2.5 1 77.0 141 164 174 H [37]

UO3 2 1 47.3 135 5 11 H [37]
UO3 1 0.9 71.4 106 226 239 H [37]
UO3 0.9 0 72.1 156 102 109 H [37]
UOF2 1 0 55.6 147 31 37 H [37]
UOFOH 0.5 0 57.0 145 42 48 H [37]

V2O5 1 0 52.2 147 14 19 H [37]
VOSO4 6 5 0.38 52.0 149 11 16 H [37]
VOSO4 5 3 0.93 56.9 149 35 40 H [37]
VOSO4 3 1 1.35 64.3 149 73 80 H [37]
VOSO4 1 0 1.86 134.7 149 498 517 H [37]

YbCl3 6 0 12.3 147 75 75 H [37]
YbPO4 2 0 113.8 151 361 375 H [37]
YCl3 6 0 2.84 73.6 149 125 133 H [37]

YPO4 2 0 50.7 149 5 9 H [37]

Zn(ClO4)2 6 4 0.13 11.0 4 92 135 pT [36] 373 533
Zn(NO3)2 6 4 0.86 62.1 167 34 39 H [37]
Zn(NO3)2 6 2 34 93 Com
Zn(NO3)2 6 0 34 104 Com
Zn(NO3)2 4 2 0.91 52.6 111 84 93 H [37]
Zn(NO3)2 4 0 84 104 Com

Zn(NO3)2 2 0 1.61 71.5 157 96 104 H [37]
ZnBr2 2 0 1.44 62.3 159 46 52 H [37]
ZnF2 4 0 2.92 55.7 151 26 31 H [37]
ZnSO4 7 6 0.39 57.2 160 19 25 pT [92] 301 309
ZnSO4 7 6 0.40 59.4 164 26 31 H [37]

ZnSO4 6 2 1.65 53.5 146 22 28 H [37]
ZnSO4 6 1 2.00 52.1 141 21 27 pT [36] 291 305
ZnSO4 2 1 0.83 49.6 135 18 24 H [37]
ZnSO4 1 0 1.55 82.6 161 148 157 H [37]
Appendix B. K2CO3
Fig. 8 shows the phase diagram of K2CO3-H2O based on several

sources. Some discrepancy between the different sources is visible:
the dihydrate of Lescouer [36] is not observed by other researchers
and the corresponding phase equilibrium line is similar to the
sesquihydrated K2CO3. Based on our own measurements with
K2CO3 (supplier Sigma-Aldrich), we expect that the dihydrate is a
misinterpretation of the observed sesquihydrate by Lescouer. In
addition, the transition based on the data of Glasser (the 0.50)
is not observed in our experiments as well. Recent insights point
towards a single 01.5 transition, also observed by Stanish and
Perlmutter [93]. This reduces the dehydration temperature and
complexity of the TCM-reaction. Based on the extrapolated deli-
quescence line in Fig. 8, deliquescence of this material is not
expected under working conditions.
We conclude with kinetic data of K2CO3, reported by Stanish
and Perlmutter [69,94,93,95]. They analyzed the effect of the grain
size, applied vapor pressure and temperature on the reaction rate.
Fig. 8. The phase diagram of K2CO3 in equilibrium with water. The data are a
Also some first cyclic experiments were performed in which they combination of experimental and calculated data from the literature, i.e., Lescouer
indicate a hydration rate influenced by dehydration conditions [36], Glasser [38], Foote [96] and Greenspan [54].
and the number of cycles of the material. Their results indicate that [28] Knoll C, Mller D, Artner W, Welch J, Werner A, Harasek M, et al. Probing cycle
stability and reversibility in thermochemical energy storage CaC2O4H2O as
the composed phase diagram is correct within the expected errors.
perfect match? Appl Energy 2017;187:19.
[29] NTsoukpoe K, Liu H, Le Pierrs N, Luo L. A review on long-term sorption solar
energy storage. Renew Sust Energy Rev 2009;13(9):238596.
Appendix C. Supplementary material [30] Trausel F, Jong de A-J, Cuypers R. A review on the properties of salt hydrates for
thermochemical storage. Energy Procedia 2014;48:44752.
[31] Bales C. Final report of subtask B chemical and sorption storage the
Supplementary data associated with this article can be found, in
overview. Tech rep. Report of IEA Solar Heating and Cooling Programme
the online version, at http://dx.doi.org/10.1016/j.apenergy.2017. Task 32; 2008.
04.080. [32] Essen van V, Gores J, Bleijendaal L, Zondag H, Schuitema R, Helden van W.
Characterization of salt hydrates for compact seasonal thermochemical
storage. In: Proc ASME 3rd int conf energy sustain.
References [33] Yan T, Wang R, Li T, Wang L, Fred I. A review of promising candidate
reactions for chemical heat storage. Renew Sustain Energy Rev
2015;43:1331.
[1] Belad F. Understanding the spectrum of domestic energy consumption:
[34] Deutsch M, Mller D, Aumeyr C, Jordan C, Gierl-Mayer C, Weinberger P, et al.
empirical evidence from France. Energy Policy 2016;92:22033.
Systematic search algorithm for potential thermochemical energy storage
[2] Willems E, Jablonska B, Ruijg G, Krikke T. Energy neutral districts? Key to
systems. Appl Energy 2016;183:11320.
transition towards energy neutral built environment! In: World renew energy
[35] De Jong AJ, Trausel F, Finck C, Van Vliet L, Cuypers R. Thermochemical heat
congr 2011-sustainable cities reg. p. 8.
storage system design issues. Energy Procedia 2014;48:30919.
[3] Jablonska B, Roossien B, Ruijg G, Visser H, Willems E, Bakker E. Energy neutral
[36] Washburn E. International critical tables of numerical data, physics, chemistry
districts in 2050 the Dutch approach. Int J Sustain Build Technol Urban Dev
and technology. 1st Electronic ed. Knovel; 2003.
2012;3(3):1517.
[37] Glasser L. Thermodynamics of inorganic hydration and of humidity control,
[4] Goldstein M. Some physical chemical aspects of heat storage. UN conf new
with an extensive database of salt hydrate pairs. J Chem Eng Data 2014;59
sources energy, Rome, vol. III. p. 4117.
(2):52630.
[5] NTsoukpoe K, Schmidt T, Rammelberg H, Watts B, Ruck W. A systematic multi-
[38] Glasser L, Jenkins H. The thermodynamic solvate difference rule: solvation
step screening of numerous salt hydrates for low temperature thermochemical
parameters and their use in interpretation of the role of bound solvent in
energy storage. Appl Energy 2014;124:116.
condensed-phase solvates. Inorg Chem 2007;46(23):976878.
[6] Scapino L, Zondag HA, Van Bael J, Diriken J, Rindt CC. Sorption heat storage for
[39] Polyachenok O, Dudkina E, Polyachenok L. Thermal stability and
long-term low-temperature applications: a review on the advancements at
thermodynamics of copper(II) chloride dihydrate. J Chem Thermodyn
material and prototype scale. Appl Energy 2017;190:92048.
2009;41(1):749.
[7] Michel B, Neveu P, Mazet N. Comparison of closed and open thermochemical
[40] Pinel P, Cruickshank C, Beausoleil-Morrison I, Wills A. A review of available
processes, for long-term thermal energy storage applications. Energy
methods for seasonal storage of solar thermal energy in residential
2014;72:70216.
applications. Renew Sustain Energy Rev 2011;15(7):334159.
[8] VROM. Tech rep; 2016. <http://www.publicatiereeksgevaarlijkestoffen.nl/>.
[41] Geelen C, Krosse L, Sterrenburg P, Bakker E-J, Sijpheer N. Handboek
[9] Zondag H, Kikkert B, Smeding S, Boer De R, Bakker M. Prototype
Energiepalen. Tech rep; 2003.
thermochemical heat storage with open reactor system. Appl Energy
[42] Trier D. Solar district heating guidelines solar collectors. Tech rep; 2012.
2013;109:3605.
[43] Craig B, Anderson B. Handbook of corrosion data; 1995.
[10] Schmidt M, Szczukowski C, Rokopf C, Linder M, Wrner A. Experimental
[44] Van de Meer (Caldic)H, personal communication.
results of a 10 kW high temperature thermochemical storage reactor based on
[45] Sol A, Mir L, Barreneche C, Martorell I, Cabeza L. Corrosion test of salt
calcium hydroxide. Appl Therm Eng 2014;62(2):5539.
hydrates and vessel metals for thermochemical energy storage. Energy
[11] Gaeini M, Zondag H, Rindt C. Effect of kinetics on the thermal performance of a
Procedia 2014;48:4315.
sorption heat storage reactor. Appl Therm Eng 2016;102:52031.
[46] Warnqvist B. Comments on thermochemical data and fusion temperature for
[12] Brunberg E. The Tepidus system for seasonal heat storage and for cooling. In:
pure sodium sulfide. Thermochim Acta 1980;37:3435.
Proc int semin thermochem heat storage. p. 24760.
[47] Roelands M, Cuypers R, Kruit KD, Oversloot H, de Jong A-J, Duvalois W, et al.
[13] Jong de A-J, Trausel F, Finck C, Van Vliet L, Cuypers R. Thermochemical heat
Preparation & characterization of sodium sulfide hydrates for application in
storage system design issues. Energy Procedia 2014;48:30919.
thermochemical storage systems. Energy Procedia 2015;70:25766.
[14] Michel B, Mazet N, Neveu P. Experimental investigation of an open
[48] Jiang J, Chan A, Ali S, Saha A, Haushalter KJ, Lam W-LM, et al. Hydrogen sulfide-
thermochemical process operating with a hydrate salt for thermal storage of
mechanisms of toxicity and development of an antidote. Sci Rep 2016;6
solar energy: local reactive bed evolution. Appl Energy 2016;180:23444.
(October 2015):20831.
[15] Donkers P, Beckert S, Pel L, Stallmach F, Steiger M, Adan OCG. Water transport
[49] The Australia group (Ed.). Australia group common control list handbook, vol.
in MgSO47H2O during dehydration in view of thermal storage. J Phys Chem C
I; 2016.
2015;119(52):2871120.
[50] de Jong A-J, van Vliet L, Hoegaerts C, Roelands M, Cuypers R. Thermochemical
[16] Posern K, Kaps C. Humidity controlled calorimetric investigation of the
heat storage from reaction storage density to system storage density. Energy
hydration of MgSO4 hydrates. J Therm Anal Calorim 2008;92(3):9059.
Procedia 2016;91:12837.
[17] Posern K, Kaps C. Calorimetric studies of thermochemical heat storage
[51] Jong A-j D, Stevens R, Rentrop C, Hoegaerts C. Coatings for heat storage
materials based on mixtures of MgSO4 and MgCl2. Thermochim Acta
reactors with hygroscopic salts. Energy Procedia 2015;70(0):18292.
2010;502(12):736.
[52] Villars P, Cenzual K, editor. Crystal structure: Datasheet from PAULING FILE
[18] Essen van V, Zondag H, Gores J, Bleijendaal L, Bakker M, Schuitema R, et al.
Multinaries Edition 2012 in SpringerMaterials (http://materials.
Characterization of MgSO4 hydrate for thermochemical seasonal heat storage. J
springer.com). Japan: Springer-Verlag Berlin Heidelberg Material Phases
Sol Energy Eng 2009;131:041014.
Data System (MPDS), Switzerland National Institute for Materials Science
[19] Ferchaud C, Zondag H, Veldhuis J, Boer De R. Study of the reversible water
(NIMS).
vapour sorption process of MgSO47H2O and MgCl26H2O under the conditions
[53] N.L. of Medicine (U.S.) (Ed.). TOXNET: toxicology data network. United States:
of seasonal heat storage. In: Proc 6th Eur therm sci conf. p. 10.
National Institute of Health (U.S.); 2000.
[20] Zondag H, Essen van V, Bakker M. Application of MgCl26H2O for
[54] Greenspan L. Humidity fixed points of binary saturated aqueous solutions. J
thermochemical seasonal solar heat storage. In: 5th int renew energy. p. 8.
Res Natl Bur Stand Phys Chem 1977;81A(1):8996.
[21] Kirsh Y, Yariv S, Shoval S. Kinetic analysis of thermal dehydration and
[55] Rockland L. Saturated salt solutions for static control of relative humidity
hydrolysis of MgCl26H2O by DTA and TG. J Therm Anal 1987;32:393408.
between 5 and 40 C. Anal Chem 1960;32(10):13756.
[22] Rammelberg H, Schmidt T, Ruck W. Hydration and dehydration of salt
[56] Boer Rd, Haije W, Veldhuis J. Determination of structural, thermodynamic and
hydrates and hydroxides for thermal energy storage kinetics and energy
phase properties in the Na2S-H2O system for application in a chemical heat
release. Energy Procedia 2012;30:3629.
pump. Thermochim Acta 2003;395:319.
[23] Michel B, Mazet N, Mauran S, Stitou D, Xu J. Thermochemical process for
[57] Lannung A. Dampfdruckmessungen des Systems Calciumchlorid-Wasser.
seasonal storage of solar energy: characterization and modeling of a high
Zeitschrift fr Anorg und Allg Chemie 1936;228(1):118.
density reactive bed. Energy 2012;47(1):55363.
[58] Sahoo DK, Mishra R, Singh H, Krishnamurthy N. Determination of
[24] Boer R, Haije WG, Veldhuis JBJ, Smeding SF. Solid-sorption cooling with
thermodynamic stability of lanthanum chloride hydrates (LaCl3xH2O) by
integrated thermal storage: the SWEAT prototype. In: Int conf heat powerd
dynamic transpiration method. J Alloys Compd 2014;588:57884.
cycles, Larnaca, Cyprus, no. August.
[59] Marias F, Neveu P, Tanguy G, Papillon P. Thermodynamic analysis and
[25] Molenda M, Stengler J, Linder M, Wrner A. Reversible hydration behavior of
experimental study of solid/gas reactor operating in open mode. Energy
CaCl2 at high H2O partial pressures for thermochemical energy storage.
2014;66:75765.
Thermochim Acta 2013;560:7681.
[60] Kipouros G, Sadoway D. The chemistry and electrochemistry of magnesium
[26] Barreneche C, Fernndez A, Cabeza L, Cuypers R. Thermophysical
production. Adv Molten Salt Chem 1987;6:127209.
characterization and thermal cycling stability of two TCM: CaCl2 and zeolite.
[61] Carling R. Dissociation pressures enthalpies of reaction in MgCl2H2O and
Appl Energy 2015;137:72630.
CaCl2nNH3. J Chem Thermodyn 1981;13:50312.
[27] Criado Y, Alonso M, Abanades J. Kinetics of the CaO/Ca(OH)2 hydration/
[62] Abedin AH, Rosen MA. Closed and open thermochemical energy storage:
dehydration reaction for thermochemical energy storage applications. Ind Eng
energy- and exergy-based comparisons. Energy 2012;41(1):8392.
Chem Res 2014;53:12594601.
[63] Nash B. Infrared reflectance spectra of Na2S with contaminant Na2CO3 effects [80] Perdichizzi A, Barigozzi G, Franchini G, Ravelli S. Peak shaving strategy through
of adsorbed H20 and CO2 and relation to studies of Io. Icarus 1988;368 a solar combined cooling and power system in remote hot climate areas. Appl
(71):3658. Energy 2015;143:15463.
[64] Fopah Lele A, NTsoukpoe K, Osterland T, Kuznik F, Ruck WK. Thermal [81] Miro L, Gasia J, Cabeza L. Thermal energy storage (TES) for industrial waste
conductivity measurement of thermochemical storage materials. Appl Therm heat (IWH) recovery: a review. Appl Energy 2016;179:284301.
Eng 2015;89:91626. [82] Collins E, Menzies A. A comparative method for measuring aqueous vapor and
[65] Landauer R. The electrical resistance of binary metallic mixtures. J Appl Phys dissociation pressures, with some of its applications. J Phys Chem 1936;40
1952;23(7):77984. (3):37997.
[66] Donkers P, Pel L, Adan O. Experimental studies for the cyclability of salt [83] Ewing WW. Calcium nitrate. II. The vapor pressure temperature relations of
hydrates for thermochemical heat storage. J Energy Storage 2016;5:2532. the binary system calcium nitrate water. J Am Chem Soc 1927;49
[67] Fichte P, Flanagan T. Kinetics of dehydration of single crystals of copper (8):196373.
formate tetrahydrate. Trans Faraday Soc 1971;67:146779. [84] Broers P, van Welie G. The system CoSO4-H2O; vapour pressure
[68] Flanagan T. Dehydration studies of lead styphnate monohydrate. Trans measurements from 0 TO 150 C. Recl Trav Chim Pays-Bas 1965;84
Faraday Soc 1959;55:11423. (6):78998.
[69] Stanish M, Perlmutter DD. Kinetics and transport effects in the dehydration of [85] Chou I-M, Seal II R, Hemingway B. Determination of melanterite-rozenite and
crystalline potassium. AIChE J 1983;29(5):80612. chalcanthite-bonattite equilibria by humidity measurements at 0.1 MPa. Am
[70] Aristov Y, Di Marco G, Tokarev M, Parmon V. Selective water sorbents for Mineral 2002;87:10814.
multiple applications, 3. CaCl2 solution confined in micro- and mesoporous [86] Besley LM, Bottomley GA. The water magnesium vapour equilibria over
silica gels: pore size effect on the solidification-melting diagram. React Kinet perchlorate hydrates. J Chem Thermodyn 1969:139.
Catal Lett 1997;61(1):14754. [87] Derby I, Yngve V. The dissociation tensions of certain hydrated chlorides and
[71] Cuypers R, de Jong A, Eversdijk J, Spijker Hvt, Oversloot H, Ingenhut B, et al. the vapor pressures of their saturated solutions. J Am Chem Soc 1916;XXXVIII
Microencapsulation of salts for enhanced thermochemical storage materials. (8):143951.
40th annu meet expo control release soc. Honolulu, Hawaii, USA, vol. 24. p. [88] Steiger M, Linnow K, Juling H, Glker G, El Jarad A, Brggerhoff S, et al.
212. Hydration of MgSO4H2O and generation of stress in porous materials. Cryst
[72] van Leeuwen C, Hermens J. Risk assessment of chemicals: an Growth Des 2008;8(1):33643.
introduction. Kluwer Academic Publishers; 1995. [89] Wagman D, Evans W, Parker V, Schumm R, Halow I, Bailey S, et al. The NBS
[73] The merck index eleventh edition. Chem Eng News Arch 1989; 67 (49). ifc. tables of chemical thermodynamic properties. J Phys Chem Ref Data 1982;11
[74] Ferchaud CJ. Experimental study of salt hydrates for thermochemical seasonal (2):302.
heat storage door. No. April; 2016. [90] Steiger M, Asmussen S. Crystallization of sodium sulfate phases in porous
[75] Mu J, Perlmutter DD. Thermal decomposition of metal nitrates and their materials: The phase diagram Na2SO4-H2O and the generation of stress.
hydrates. Thermochim Acta 1982;56(3):25360. Geochim Cosmochim Acta 2008;72(17):4291306.
[76] Lumpkin J, Perlmutter D. Thermal and water vapor effects on the rate of the [91] Dingemans P. Die dampfspannungen in den dreiphasengleichgewichten des
dehydration reactions of barium chloride. Thermochim Acta systems natriumbromid-wasser. Recl Trav Chim Pays-Bas 1938;57(7):7039.
1995;249:33549. [92] Chou I-M, Seal II R. Determination of goslarite-bianchite equilibria by the
[77] Linnow K, Niermann M, Bonatz D, Posern K, Steiger M. Experimental studies of humidity-buffer technique at 0.1 MPa. Chem Geol 2005;215:51723.
the mechanism and kinetics of hydration reactions. Energy Procedia [93] Stanish M, Perlmutter D. Rate processes in cycling a reversible gas-solid
2014;48:394404. reaction. AIChE J 1984;30(1):5662.
[78] Donkers P, Beckert S, Pel L, Stallmach F, Steiger M, Adan O. Water transport in [94] Stanish M, Perlmutter D. Salt hydrates as absorbents in heat pump cycles. Sol
MgSO47H2O during dehydration in view of thermal storage. J Phys Chem C Energy 1981;26(4):3339.
2015;119(52):2871120. [95] Stanish M, Perlmutter D. Kinetics of hydration-dehydration reactions
[79] Paulik F, Bessenyey-Paulik E, Walther-Paulik K. Differential thermal analysis considered as solid transformations. AlChE J 1984;30(4):55763.
under quasi-isothermal, quasi-isobaric conditions (Q-DTA) Examination using [96] Foote B. Equilibrium in the systems, nickel chloride, cobalt chloride, cupric
transformation-governed heating control and self-generated atmosphere chloride-hydrochloric acid- water. J Am Chem Soc 1923;45(3):6637.
(TGHC-SGA). Thermochim Acta 2003;402:10516.

1 s2.0 S0306261917304750 Main

Uploaded by

Copyright:

Available Formats

1 s2.0 S0306261917304750 Main

Uploaded by

Document Information

Original Description:

Original Title

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

1 s2.0 S0306261917304750 Main

Uploaded by

Copyright:

Available Formats

Applied Energy 199 (2017) 4568

Contents lists available at ScienceDirect

A review of salt hydrates for seasonal heat storage in domestic

We report an evaluation of potential hydrate reactions for heat storage application.

log (pwater vapor)

pwater vapor (bar)

dehydration explained as before, the condenser temperature is a

The database consists of 563 entries with complementary ther-

P.A.J. Donkers et al. / Applied Energy 199 (2017) 4568

A final note on void spaces within grains should be made

6.3. Hydration temperature

The hydration temperature indicates the maximum tempera-

6.5. Melting point

Appendix A porosity; DH is the enthalpy of the reaction; DS the entropy of the

Salt H L E=V DH DS T hy T deh Source T min T max Comment

AgF 1 0 59.2 158 33 38 H [37]

Ba(ClO4)2 3 0 3.30 146.6 138 613 634 H [37]

Ba(OH)2 8 1.3 2.55 43 ? Com

Ba(OH)2 3 1.3 1.24 33 ? Com

Ba(OH)2 0 1 116.0 92 630 658 pT [36] 903 1263 L:BaO

BaBr2 2 0 1.37 38 42 com

Salt H L E=V DH DS T hy T deh Source T min T max Comment

BaCl2 2 0 1.50 32 55109 Com

BaI2 2 1 0.55 55.2 147 29 35 H [37]

BeSO4 2 0 1.97 109 110 Com

Ca(NO3)2 4 2 0.86 53.0 135 38 44 pT [83] 293 312

Ca(NO3)2 2 0 1.28 59.5 151 44 51 pT [83] 293 324

CaCl2 6 4b 0.98 62.6 166 36 42 pT [57] 273 301

CaCl2 4a 2 1.21 60.4 155 43 49 pT [57] 273 316

CaCl2 2 1 0.64 51.2 145 63 71 H [37]

CaC2O4 1 0 1.06 69.5 134 134 144 H [28]

(continued on next page)

Salt H L E=V DH DS T hy T deh Source T min T max Comment

CaSO3 0.5 0 0.57 62.8 149 65 72 H [37]

CaSO4 0.5 0 0.60 65.0 141 94 102 pT [36] 373 435

CdBr2 4 0 1.82 52.1 145 16 22 H [37]

CdSO4 2.7 1 1.30 53.1 147 17 23 pT [36] 298 313

CeCl3 7 0 2.57 61.0 150 54 60 H [37]

Co(ClO4)2 6 0 1.06 22.0 27 101 118 pT [36] 360 493

CoCl2 6 1 2.33 29 50 Com

CoSO4 7 6 0.36 52.6 146 16 22 pT [84] 293 316

Cr2(SO4)3 14 8 1.25 57.7 149 39 45 H [37]

CrCl2 4 0 2.40 41 69 com

Cs2CO3 3.5 1.5 1.06 64.5 154 65 72 H [37]

Salt H L E=V DH DS T hy T deh Source T min T max Comment

CsCr(SeO4)2 12 0 3.08 73.0 196 43 48 pT [36] 323 368

CsNi(SO4)2 6 2 0.94 53.7 132 45 52 pT [36] 308 355

Cu(IO3)2 1 0 60.2 154 43 49 H [37]

CuCl2 2 0 1.74 59.1 145 53 59 H [37]

CuSO4 5 3 1.03 55.9 149 30 35 H [37]

CuSO4 1 0 0.93 49.4 81 140 151 pT [36] 303 348

FeCl2 4 0 2.22 31 59 Com

FePO4 6 0 2.17 53.8 150 18 23 H [37]

FeSO4 4 0 2.29 20 112 Com

H2O 50.0 141 11 17 H [37] solidgas

(continued on next page)

Salt H L E=V DH DS T hy T deh Source T min T max Comment

H2SO4 2 1 57.7 124 87 95 H [37]