Carbohydrate Polymers 335 (2024) 122077

Contents lists available at ScienceDirect

Carbohydrate Polymers
journal homepage: www.elsevier.com/locate/carbpol

Size-structure-property relationship of wood particles in aqueous and dry

insulative foams
Elizabeth Dobrzanski a, b, Elisa S. Ferreira b, c, d, Praphulla Tiwary e, Prashant Agrawal e,
Richard Chen e, Emily D. Cranston a, b, c, *
a
Department of Chemical and Biological Engineering, University of British Columbia, Vancouver, BC V6T 1Z4, Canada
b
Bioproducts Institute, University of British Columbia, Vancouver, BC V6T 1Z4, Canada
c
Department of Wood Science, University of British Columbia, Vancouver, BC V6T 1Z4, Canada
d
Brazilian Nanotechnology National Laboratory (LNNano), Brazilian Center for Research in Energy and Materials (CNPEM), Campinas, SP, Brazil
e
Plantee Bioplastics Inc., Vancouver, Canada

A R T I C L E I N F O A B S T R A C T

Keywords: Three size-fractionated samples of pine beetle-killed wood particles were used to prepare lightweight insulative
Wood foam foams. The foams were produced by foam-forming an aqueous slurry containing wood particles (125–1000 μm),
Aqueous stability a polymer binder, and surfactant, followed by oven drying. The effect of wood particle size on the aqueous foam
Forest residue
stability, structure, and performance of insulative foams was investigated. While all aqueous foams were highly
Thermal insulation
Sustainable construction material
stable, aqueous foam stability increased with decreasing particle size. For dry foams, the cell size distribution was
similar for all particle sizes as it was primarily controlled by the surfactant; differences occurred within the cell
wall structure. A size-structure-property relationship was identified using x-ray micro-computed tomography
where smaller particles produced lighter cell wall frameworks, leading to lower densities and decreased thermal
conductivity and compressive strength. Larger particles produced denser cell wall frameworks that were more
resistant to deformation, although all dry foams had sufficient mechanical properties for use as insulation panels.
Thermal conductivity for all wood particle size-fractionated samples was <0.047 W m− 1 K− 1 making the foams
similar to expanded polystyrene/polyurethane and supporting their use as thermal insulation in buildings.

1. Introduction Resources Canada funded research to replace petrochemical-based

insulative foams with bio-based foam materials (Natural Resources
Lightweight materials, specifically foams, are excellent thermal in­ Canada, 2020). Replacement of traditional insulation with bio-based
sulators. Conventional foams used as insulation for buildings include foams would also assist in achieving building sustainability certifica­
expanded polystyrene (EPS), polyurethane (PU), and minerals wools, tion like LEED or BREEAM (Schiavoni et al., 2016).
which can reduce energy consumption by up to 64 % in the summer and We previously showed that lightweight, high bio-based content
37 % in the winter (Llantoy et al., 2020). However, their use is linked to wood foams are an environmentally friendly alternative to
environmental and health concerns as manufacturing creates significant petrochemical-based foams used as building insulation (Ferreira et al.,
greenhouse gases, lethal gases are produced during fires, and they 2022). These foams have thermal and mechanical properties that are
become landfill waste at the end of their life (Schiavoni et al., 2016). competitive with conventional rigid insulative foams and are resistant to
Replacing EPS, PU, mineral wool, and other petrochemical-based foams humidity yet still recyclable. More specifically, the foams are made from
with bio-based foams has gained widespread interest as organizations wood particles that are foamed in water with surfactant and a water-
worldwide aim to reduce these harms. In Europe, the Circular Bio-based soluble polymer binder and then oven dried (Praphulla et al., 2022).
Europe Joint Undertaking is funding a series of projects to replace The promising thermal properties are attributed to the low density of the
petrochemical-based products in the construction industry (Circular Bio- resulting foam and the hierarchical pore structure consisting of open
based Europe Joint Undertaking, 2023), while in Canada, Natural cells templated by air bubbles and closed cells from the natural

* Corresponding author at: Department of Chemical and Biological Engineering, University of British Columbia, Vancouver, BC V6T 1Z4, Canada.
E-mail addresses: edobrz@student.ubc.ca (E. Dobrzanski), elisa.ferreira@lnnano.cnpem.br (E.S. Ferreira), praphulla@plantee.ca (P. Tiwary), prashant@plantee.
ca (P. Agrawal), richard@plantee.ca (R. Chen), emily.cranston@ubc.ca (E.D. Cranston).

https://doi.org/10.1016/j.carbpol.2024.122077
Received 19 December 2023; Received in revised form 11 March 2024; Accepted 18 March 2024
Available online 19 March 2024
0144-8617/© 2024 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-
nc/4.0/).
E. Dobrzanski et al. Carbohydrate Polymers 335 (2024) 122077

honeycomb structure of the wood cell walls (Ferreira et al., 2022). average milled wood particle size and aqueous stability, agreeing with
In the quest for truly sustainable construction materials, using forest our hypothesis that the resulting dry foams will also have decreased
residues as the biomass source is particularly promising. Forest residues thermal conductivity. However, for industry adoption, a holistic
are materials generated by the forestry industry that have little to no approach must be considered. From a cost and sustainability perspec­
economic value (Kizha & Han, 2016). They include materials like bark, tive, the structure and processing relationships elucidated here will help
dead trees, off-cuts, and small-diameter trees, and are often used for low- optimize energy consumption for wood foam manufacturing. For
value products. Mountain pine beetle-killed wood is one example of a example, if similar properties are achieved for different particle sizes,
forest residue local to the authors. British Columbia, Canada, experi­ then the largest size – i.e., the lowest energy – may be used for further
enced a mountain pine beetle epidemic that peaked from 2004 to 2005 studies and scale up (Ou & Cai, 2020). Concomitant to this, aqueous
and devastated the province’s lodgepole pine forests (Corbett et al., foams that are highly stable but can be quickly dewatered and dried
2016). The beetles killed over 723 million cubic metres, or 20 % of the under oven or ambient conditions may be favourable due to lower en­
province’s forests, transforming viable timber into undesirable material. ergy use (Chen et al., 2022).
The total impact of the mountain pine beetle epidemic is projected to
affect British Columbia’s economy by a loss of $57.37 billion in GDP 2. Methods & materials
over a 45-year period, from 2009 to 2054 (Corbett et al., 2016).
Mountain pine beetle-killed wood is undesirable because its macro 2.1. Materials
properties are degraded and unreliable compared to wood from a
healthy, living tree. The wood is known to undergo significant Pine beetle-killed wood (lodgepole pine, Pinus contorta) was pro­
morphological and chemical changes during infestation and death (Woo vided by Plantee Bioplastics Inc. and was sourced from Prince George,
et al., 2005), leading to a high number of defects (Loeffler & Anderson, British Columbia. Polyvinyl alcohol (PVA) (87–90 % hydrolyzed, mo­
2018). During processing, the wood splits, cracks, and produces more lecular weight 30,000–70,000 g mol− 1) and sodium dodecyl sulfate
fines, all of which decrease its value. (SDS) (purity ≥98.5 %) were purchased from Sigma Aldrich and used as
Milling the mountain pine beetle-killed wood is a method to received. Slurries were prepared with Millipore Milli-Q distilled deion­
circumvent the unreliable macro properties of the trees. Size reduction is ized water (18.2 MΩ cm resistivity).
a common first step in transforming waste wood into value-added ma­
terials (Ihnat et al., 2020). The still-intact microstructure of the wood is 2.2. Wood milling
retained and may be re-used in a composite-like material when pressed
into products like particleboard or fibreboard (Schubert et al., 2023). The first pass of milling pine beetle-killed wood was performed using
However, the properties of milled wood and how they influence light­ a hammer mill (10HMBL, GlenMills, Clifton, USA) to reduce the wood
weight solid materials like foams have not been studied. Past work has chip size to under 1 cm. The second pass was performed using a knife
focused on lightweight lignocellulosic materials consisting of pulp fibres mill (SM100 Comfort, Retsch, Haan, Germany) until the particles could
(Harb et al., 2023; Wu et al., 2022), microfibrillated cellulose or nano­ pass through a 1 mm sieve.
cellulose (Jiang et al., 2021; Ren et al., 2023), or a combination of both
(Liu et al., 2018; Yin et al., 2023). To produce these materials, a rela­ 2.3. Size separation of milled wood
tively high amount of chemical or mechanical refining is required to
liberate cellulose from wood and their methods often use high energy Three wood particle size-fractionated samples were obtained by
drying processes to circumvent capillary forces and avoid material loading the tap-sieve shaker (RX-29 RO-TAP, W.S. Tyler, Mentor, USA)
collapse (Ferreira et al., 2021). Generally, milled wood has been rele­ with material from the knife mill and using a single desired mesh size for
gated to minor roles as filler in petrochemical-based composites (Kumar sieving. The size-fractionated samples were obtained by using their
et al., 2011) or concrete (Mangi et al., 2019). corresponding sieve size and running the tap-sieve shaker until at least
This study investigates the size-structure-property relationship of 50 g of particles were obtained. Each size-fractionated sample only in­
size-fractionated samples of milled wood particles from mountain pine cludes particles that passed through the corresponding sieve, i.e., < 300
beetle-killed wood to elucidate the importance of particle size in μm contains all the particles that passed through the 300 μm sieve and
aqueous and dry foams. Our process is water-based; aqueous foams are none that remained in it (Table 1). Continuous particle size distributions
known to be excellent carriers of particles, such as fibres, to create (e.g., 300 μm – fines) were intentionally chosen instead of bracketed
lightweight, bulky structures (Hjelt et al., 2020). Adding particles can sizes (e.g., 300 μm – 125 μm) to assess whether more energy input was
affect the surface tension, bubble size, rheology, and stability of the required to reduce overall particle size to significantly improve insu­
aqueous foam phase which in turn affects the drying process and final lative properties, and to investigate the effect of including larger parti­
foam performance. The stability of the aqueous foam is generally cles with smaller ones. The inclusion of small particles, down to fines,
improved by increasing mixing speed, increasing surfactant concentra­ has been highlighted previously as overlooked yet critical for final ma­
tion, introducing particles or fibres, and increasing particle surface area terial properties (Hyll, 2015; Odabas et al., 2016). We believe that size
(Hjelt et al., 2020). fractionation of wood into small particle size brackets is not industrially
We hypothesize that if we decrease the average size of milled wood feasible nor necessary for foam production. To obtain a frequency dis­
particles, we obtain more stable aqueous foams that retain more air in tribution graph of wood particles in each size-fractionated sample,
their structure, resulting in dry foams that have lower thermal con­ additional sieves were added for better resolution. A known mass of at
ductivity. This size-stability trend is fairly well-studied for Pickering least 50 g was added to each 10 min run of the tap-sieve shaker, and the
foams/emulsions that incorporate nano- to microscale particles, but frequency was calculated by measuring the mass remaining in each
high consistency fibre/particle-laden foams with micro- to millimeter sieve.
sized particles are less understood (Hjelt et al., 2020). Here we quantify
aqueous foam stability visually and by using an accelerated ageing 2.4. Aspect ratio, length, and width of wood particles
method based on light transmittance during centrifugation, allowing us
to link particle size and aqueous foam stability and anticipate dry foam Images of milled wood from a mesh size over 125 μm were taken by
cell structure and final foam properties. Aqueous foam stability is also scattering the particles on a backlit glass screen, with care taken to
discussed in the context of particle shape heterogeneity, aspect ratio and ensure no particles were touching for ease of image analysis (Fig. S1).
specific surface area. Images were taken with a phone camera on a tripod with a ruler as a
The results presented here show a positive trend between decreased scale bar and were prepared for analysis in ImageJ (version 1.53u)

2
E. Dobrzanski et al. Carbohydrate Polymers 335 (2024) 122077

Table 1 following the ImageJ User Guide (Rasband, 2012). Particles from a mesh
Size-fractionated wood particle samples: sieve number and size, and particle size size under 125 μm were imaged using an optical microscope (Eclipse
and particle shape information obtained from ImageJ analysis of photographs LV100 POL, Nikon, Japan) and processed equivalent to the photographs
and optical microscopy images. (Fig. S2). The “Analyze Particles” command in ImageJ was used to
Size fraction obtain aspect ratio, with length and width obtained from the aspect
<1000 <300 <125 ratio’s fitted ellipse major and minor parameters (Table 1). Aspect ratio
μm μm μm frequency distributions were obtained from over 10,000 particles using
TAPPI sieve number 18 50 120
the Histogram “Data Analysis” tool in Microsoft Excel. Particle mea­
(ASTM, E11-22) surements from smaller size-fractionated samples were re-used for the
Sieve opening size (μm) 1000 300 125 larger size-fractionated samples, as they are all continuous distributions
Image Average particle aspect 3.0 3.4 2.6 from the smallest particle size to their respective largest size.
analysis ratio*
Average particle length 910 520 65
(μm) 2.5. Calculated specific surface area
Maximum particle length 3010 2400 900
(μm) Specific surface area (SSA, in cm2 g− 1) for the milled wood was
Minimum particle length 1.8 1.8 1.8 approximated by fractionating a <1000 μm sample into 13 distinct sieve
(μm)
Average particle width 380 150 22
sizes (Fig. 1(b)) and multiplying the individual opening size of each
(μm) sieve by the weighted aspect ratio distribution. Weassumed the shape to
Maximum particle width 1600 810 220 be a rectangular prism devoid of pores and surface roughness (Bekhta &
(μm) Hiziroglu, 2002). The average SSA was calculated from the weighted
Minimum particle width 1.5 1.5 1.5
SSA for each aspect ratio. Eq. (1) for SSA was followed, where Sp is the
(μm)
surface area of the particle, Vp is the volume of the particle, and ρw is the
*
The standard deviation for all average aspect ratio values was ca. 2. As the density of lodgepole pine which was taken as 0.41 g cm− 3 (Trent et al.,
samples were all polydisperse, the maximum and minimum length and width
2006).
(calculated from the aspect ratio) are provided to represent the breadth of the
distribution instead of the standard deviations (which are large).

(a)

500 µm

(b) 40 (c) 50
< 125 µm < 125 µm
< 300 µm < 300 µm
30 40
< 1000 µm < 1000 µm
Frequency (%)

Frequency (%)

30
20
20
10
10

0 0
0 200 400 600 800 1000 1 2 3 4 5 6 7 8 9 10
Sieve Opening Size (µm) Aspect Ratio
Fig. 1. (a) Optical microscopy image of the <1000 μm size-fractionated sample dispersed in water showing heterogeneity in wood particle size and shape. (b)
Frequency distribution by mass of wood particles in each size-fractionated sample. (c) Weighted aspect ratio of wood particles in each size-fractionated sample.
Average aspect ratio is found in Table 1.

3
E. Dobrzanski et al. Carbohydrate Polymers 335 (2024) 122077

( )
Sp 1 tests.
SSA = (1)
Vp ρ w
2.11. X-ray micro-computed tomography (μCT)
2.6. Foam preparation
2D image slices of the wood foams were obtained using a Xraia 520
Slurries were prepared by solubilizing 4 g of PVA and 0.048 g of SDS Versa μCT apparatus (Zeiss, Oberkochen, Germany) operating at 40 kV.
in 69 g of water and then adding 12 g of size-fractionated milled wood Images were collected at 10 μm resolution with an iKon-L camera
particles (Ferreira et al., 2022). Aqueous foams were prepared from the equipped with a 2048 × 2048 sensor (Andor, Belfast, Northern Ireland).
slurries using a homogenizer (Ultra Turrax T25, IKA, Staufen, Germany) Image processing occurred in ImageJ using the MorphoLibJ plugin for
at 9000 RPM until the foam volume remained constant, approximately morphological segmentation, and cell size was measured using Feret’s
10 min. Preparation took place in a laboratory setting with a thermostat diameter. For each size-fractionated sample, over 6000 measurements of
set to 21 ◦ C and uncontrolled relative humidity, ranging from 25 to 70 % cell size were obtained and expressed as a frequency distribution. Den­
RH. The aqueous foams were poured into silicone baking molds (3.5 × sity of the foams was approximating using ImageJ (version 1.53u) by
3.5 × 1.5 cm) to fill, but not overflow, each mold. Drying took place converting a 500-slice stack of images to 8-bit, binary images, and
overnight in an oven (Isotemp 100 L Oven Gravity, Thermo Fisher Sci­ counting the number of black and white pixels using the “Histogram”
entific, Langenselbold, Germany) at 60 ◦ C. Average mass of the dried command. Black pixels represented voids within the foam structure and
foam samples was 2.2 ± 0.2 g and average volume was 18.4 ± 1.7 cm3. white pixels represented the solid foam.

2.7. Aqueous foam stability 3. Results and discussion

Stability of the aqueous foams was visually assessed by pouring 50 3.1. Milled wood particles are heterogeneous in size and shape
mL of foam into a 50 mL graduated cylinder and recording the change in
foam and drained liquid height over a period of 2 h until the foams The milled wood particles used in our insulative foams were het­
appeared to reach equilibrium. Stability was further assessed using a erogeneous in size and shape (Fig. 1a) regardless of size fraction. The
LUMiSizer Dispersion Analyzer (LUM GmbH, Berlin, Germany): aqueous wood particle dimensions and size distributions were dictated by the
foam was added immediately after foam preparation to four LUM 10 mm sieve size used to fractionate the knife-milled wood (Fig. 1b). As shown
polyamid rectangular synthetic cell cuvettes (LUM GmbH, Berlin, Ger­ in Table 1, the smallest average dimension was 22 μm while the largest
many) to a 23 mm fill height. All samples were centrifuged for over 30 was 910 μm (with individual values spanning from 1.5 to 3010 μm).
min with the following conditions: 25 ◦ C, 4000 RPM, and a transmitted Aspect ratios ranged from 1 to 10 for all size-fractionated samples
near-infrared wavelength of 870 nm. Integral transmission data was (Fig. 1c). Breakage appeared to occur along the direction of the wood
obtained directly from the LUMiSizer SEPView version 3.4.0, focusing grain, exposing tracheids in some wood particles (Fig. S3). While this
on the portion of the curve that represented supernatant development. variability makes image analysis a challenge because most methods rely
Spatio-temporal plots were obtained by running the raw LUMiSizer data on consistent shapes or colour values, we believe it contributes to
through a Python script kindly provided by Drs. Bapat and Segets (Bapat structural stability within our dry foams. The relationship between
& Segets, 2020). particle morphology and packing has been widely studied for its appli­
cations in civil engineering and agriculture: as particles become more
2.8. Thermal conductivity heterogeneous in size, packing density increases due to an increased
number of contact points (Conzelmann et al., 2020; Sohn & Moreland,
The thermal conductivity of dried foams was determined using a 1968; Wiacek & Molenda, 2014). Similar to these previously studied
transient thermal conductivity instrument equipped with a Modified systems, we expect that smaller wood particles will pack between larger
Transient Plane Source (TCi, C-Therm, Fredericton, NB, Canada) sensor. wood particles, creating a geometric interlocking effect.
Tests were carried out on six samples per size fraction with average Three continuous size-fractionated samples of milled wood were
dimensions of 3.4 × 3.5 × 1.5 mm. To ensure complete contact with the investigated (<125, <300, and <1000 μm) to elucidate the relationship
sensor, a 500 g weight was placed on top of each tested sample. between wood particle size and resulting foam structure and properties.
Size fractionation also acts as an analogue for milling energy input:
2.9. Calculated R-value although our samples have all been taken from the same source and have
been milled for the same amount of time, a smaller size-fractionated
R-value relates the thermal conductivity of a material to its thickness sample can represent further milling time/energy to create a consis­
and is commonly used in the construction industry. It is shown as Eq. (2) tently smaller average size distribution. Samples were size fractionated
and is calculated by dividing thickness t by thermal conductivity k and using a tap-sieve shaker (Table 1) where the 300 and 125 μm sieve sizes
multiplying by 5.678 to convert metric (m2 K W− 1) to imperial (h ft2 ◦ F were chosen based on the bimodal distribution of the <1000 μm size
Btu− 1) (Vosper & Wiersma, 1988). fraction taken directly from the knife mill, which had peaks at both 300
(t) and 125 μm (Fig. 1b, black circles). Sieving decreased the size and
R − value = × 5.678 (2) narrowed the original broad distribution of wood particle sizes (Fig. 1b).
k Small particles (~100 μm) were present in all samples, with an increase
Higher R-values indicate that the material is a better insulator, i.e., in frequency from 10 to 40 % as sieve size decreased. Additionally, the
less material is required to insulate a building. shape of the size distribution curves changed after sieving, exhibiting a
narrowing bimodal distribution from the <1000 to the <300 μm size
2.10. Compression testing fraction and a single peak for the <125 μm size fraction. The bimodal
distribution may represent the size difference between milled bark and
The compressive strength at 10 % compression was determined using wood, as seen previously (Rezaei & Sokhansanj, 2018).
a universal testing machine (model 5969, Instron, Norwood, MA, USA), The aspect ratios and aspect ratio distributions (Table 1 and Fig. 1c)
using compression platen fixtures and following test conditions from confirm the polydisperse nature of the milled wood particles observed in
ASTM D1621. The rate of compression was 1.3 mm min− 1 (ASTM D695). Fig. 1a, even after sieving. All size-fractionated samples had a similar
Compression values were averaged from the same six samples for each aspect ratio distribution centered around 3, with the smallest <125 μm
size-fractionated wood particle sample used in the thermal conductivity sample displaying a significantly larger peak at 2. The large peak at

4
E. Dobrzanski et al. Carbohydrate Polymers 335 (2024) 122077

aspect ratio 2 may be a consequence of milling unrefined wood: 2020). High-stable aqueous foams are characterized by minimal volume
although the path of fracture will tend to align with wood cell walls, loss and dimensional changes over time. The longer the aqueous foam is
resulting in elongated particles, during milling the particles are able to retain its volume, the lighter the final dry foam will be, and also
randomly oriented and smaller particles which spend more time in the enables the use of lower energy drying methods such as oven drying. The
mill will become shorter regardless of cell wall orientation (Karinkanta high solids content of our foams also means that less water is required in
et al., 2018). Overall, the milled wood particles are in between being our foam forming process, leading to relatively low drying times.
square/spherical (i.e., aspect ratio 1) and high aspect ratio fibre-like We observed that minor liquid drainage occurred immediately after
particles, such as softwood pulp fibres that generally has an aspect foaming and was the primary instability mechanism in our aqueous
ratio over 40 (Sanchez-Salvador et al., 2022). This heterogeneity is a foams (Fig. 3 and Table 2) – this is similar to previous work on pulp fibre
unique feature of the wood particles that can potentially impact the foams (Koponen et al., 2020). Most of the liquid drained within the first
properties of aqueous and died foams, in contrast to high aspect ratio 30 min and stabilized after 60 min (Fig. 3a). A small effect from wood
pulp fibres or nanocelluloses, which have been much more studied for particle size was observed: less liquid drained from the aqueous foam
foam-forming applications (Ferreira et al., 2021; Nechita & Năstac, with smaller wood particles. This is likely due to the increased specific
2022; Sun et al., 2021). surface area (SSA) from the smaller particles providing resistance to
gravity-induced flow, a relationship that has been observed in both pulp
fibres (Banavath et al., 2011; Seth, 1990) and natural fibres (Nguyen &
3.2. Milled wood greatly increases aqueous foam stability with a small Indraratna, 2019). The calculated SSA of the wood particles (Table 2)
particle size effect corroborates this observation, indicating a method to tune drainage in
aqueous foams, as discussed later in this section.
Our insulative foams are produced by foaming a slurry of wood Aqueous foams without wood particles were highly unstable as
particles, surfactant (SDS) and dissolved polymer (PVA) in water fol­ visualized in Fig. 3a and quantified in Fig. 3b, where all three size-
lowed by oven drying; the final lightweight structure is templated fractionated wood samples increased aqueous foam stability by over
initially by air bubbles in the aqueous foam and the wood particles 90 %. In contrast to pulp fibre foams where the fibres are flexible and
themselves (Figs. 2 and S4). Optical microscopy shows close packing of conform to the bubble shape (Koponen et al., 2018; Lehmonen et al.,
wood particles and bubbles (Fig. 2a), which is a result of both the high 2020), the wood particles may be stabilizing the foam by physically
consistency of our aqueous foam (17 %, or 12 g of dry wood in 69 mL of restricting bubble movement (Fig. 2). While SDS is the main air-water
water) and use of a surfactant. Typical pulp fibre foams have consis­ interface stabilizer (Ferreira et al., 2022) we believe SDS also helps
tencies ranging from 1 to 6 % (Hjelt et al., 2020; A. Koponen et al., make the wood particles and the air bubbles compatible. In combination
2020), with consistencies in recent works including 1.5 % for insulative with PVA, which has a minor surfactant effect (Bhattacharya & Ray,
foams (He et al., 2019); 2.5 % for acoustic foams (Cucharero et al., 2004), these components produce a hybrid water phase around the
2021); and 1.5 % for flame retardant foams (Wu et al., 2022). Low bubbles that increases the energy barrier for bubble deformation and
consistencies are required for these foams because of the hydrophilicity collapse. Overall, all the aqueous foams with wood particles are stable
of pulp fibres, which remove free water from the bulk slurry and means with <5 % liquid drainage, allowing us to use overnight oven drying, or
that proportionally more water is required for foam forming. The even benchtop ambient drying, to obtain lightweight solid foams. The
amphiphilicty of wood is an advantage here as we are able to increase stability also translates to very little shrinkage during drying – volume
the consistency of our foams without affecting our foam forming pro­ loss from aqueous to dry foam was <5 % by volume for all three size-
cess. Combined with the use of a surfactant to reduce surface tension, we fractionated wood samples.
are able to achieve a close-packed structure in our aqueous foams, A faster, quantitative method to measure aqueous foam stability was
resulting high aqueous foam stability. performed using a dispersion analyzer (LUMiSizer), which accelerates
The compatibility of bubbles and wood particles suggests an foam collapse through centrifugation. While the samples are centri­
increased energy requirement for separation leading to high aqueous fuged, light transmittance through the length of the sample is recorded.
foam stability, which is crucial when handling and drying of the pre­ The procedure developed by Bapat and Segets (2020) to visualize time-
cursor wet structure. Instability mechanisms for aqueous foams include resolved settling was used to produce Fig. 4a. We interpret the first front
liquid drainage due to gravity, bubble coarsening, and bubble coales­ to show bubbles rapidly evacuating the foam within the first minute,
cence, which occur at different rates and may be suppressed by intro­ while the second front represents wood particles sliding down the
ducing particles or fibres and by using surface active agents (Hjelt et al.,

(a) (b)

500 µm 100 µm

Fig. 2. Optical microscopy images of aqueous foams with wood particles (<1000 μm): (a) foam templating is achieved by both the wood particles and bubbles, and
close packing of the wood particles is observed. (b) Surface tension of the liquid is low, allowing close contact between the wood particles and the air bubbles and
increasing the energy requirement for typical instability mechanisms to proceed. Optical microscopy images of the <300 μm fraction and <125 μm fraction are
provided in Fig. S4.

5
E. Dobrzanski et al. Carbohydrate Polymers 335 (2024) 122077

Fig. 3. Stability of aqueous foams: (a) visual comparison between our formulation with and without wood particles (<1000 μm). The addition of wood particles
significantly decreases the amount of drained liquid due to gravity over a two-hour period. (b) Percent of liquid drained due to gravity from all size-fractionated
aqueous foams over a two-hour period.

cuvette walls. The change in light transmittance shows a difference from

Table 2
black to white with time (Fig. 4a) and can be plotted as the integral of
Summary of values for liquid drained and its complement, liquid retained. As
transmission (Fig. 4b). A more rapid change in light transmittance with
calculated (approximate) SSA of the milled wood particles increases, liquid
retained within the aqueous foam increases.
time (higher slope) means the foams are more easily collapsed, (less
stable). The results indicate that the foam with the largest particles
Size Change in Liquid Liquid Calculated SSA of
(<1000 μm) was the least stable with 30 % light transmittance in the
fraction foam height drained after retained after wood particles
(%) 2 h (%) 2 h (%) (cm2 g− 1) supernatant phase after 10 min. In contrast, the foam with the smallest
particles (<125 μm) was the most stable, with 12 % light transmittance
No wood 0.0 45.0 55.0 N/A
particles
maintained after 10 min, supporting the observed trend in drainage
<1000 μm 2.8 4.4 95.6 100 measurements due to gravity (Fig. 3b). The aqueous foam stability
<300 μm 2.3 1.5 98.5 400 provided by all three size-fractionated wood samples is high, being
<125 μm 2.2 0.0 100. 1000 equivalent to food-grade emulsions centrifuged at 4000 RPM that
showed a light transmittance ranging from 2.5 to 30 % after 10 min (Niu

Fig. 4. Processed dispersion analyzer (LUMiSizer) data for aqueous foams under centrifugation, i.e., accelerated ageing: (a) spatio-temporal plot (Bapat & Segets,
2020) showing the time-resolved settling behaviour for an aqueous foam made with the <1000 μm wood particle size-fractionated sample. Dashed lines are drawn to
guide the eye to see the development of the first and second fronts. The other size-fractionated samples are shown in Fig. S5. (b) Percent of integral transmission of
light over the area of supernatant development for all three size-fractionated wood particle samples. Steeper slopes correspond to a higher rate of foam collapse and
lower stability.

6
E. Dobrzanski et al. Carbohydrate Polymers 335 (2024) 122077

et al., 2022; Zhang et al., 2021). exceed the requirements outlined in ASTM C578-22 Standard Specifi­
The fast, yet minor, drainage of liquid from our aqueous foams over cation for Rigid, Cellular Polystyrene Thermal Insulation. Therefore, the
the first 30 min suggests some avenues of exploration for quicker and largest particle size fraction may be used to achieve standards for
lower energy drying. The foams can be cast on mesh, or in molds with thermal insulation, but possibilities remain to obtain different material
holes, to dewater before or during drying, decreasing the amount of properties by simply changing the feedstock particle size. A wide range
liquid retained within the mold that would need to be removed by of properties is also encountered for cellulose lightweight materials in
3
evaporation. The dewatered liquid could then be recycled back into the the literature: apparent density may range from 0.01 to 0.49 g cm− and
− 1 − 1
system. Dewatering has been explored previously with fibre-laden thermal conductivity from 0.013 to 0.075 W m K (Ferreira et al.,
foams, though more work could be done to optimize the process in 2021). Our compressive strength at 10 % strain is within the typical
terms of decreasing sample shrinkage and drying time (Keränen et al., range of cellulose lightweight materials of 10–70 kPa. Our wood foams
2023). For example, vacuum-assisted dewatering may improve drying are relatively rigid with small elastic moduli: 5.2 (<125 μm) to 9.8
time but can have a detrimental effect on sample shrinkage (Koponen (<1000 μm) kPa compared to reported values ranging from 20 to
et al., 2020; Lehmonen et al., 2020). The high stability of our aqueous 49,300 kPa. Ultimately, wood particle size can be used to tailor the
foams suggests that gravity- or vacuum-assisted dewatering could be desired foam performance and adapt to cost, ease of manufacturing, and
used. Another route could include the reduction of water in the initial region- and building-specific thermal insulation requirements.
formulation. While the amount of water was held constant for this study, The internal structure of the dry foams is dependent on the aqueous
the formulation could be fine-tuned to reduce the amount of water foam morphology and stability. When dried, the originally spherical
without compromising foamability and final properties. bubbles in the aqueous foam take on a polyhedral shape which is pre­
served by the cell wall structure yet varies slightly depending on the
wood particle size as visualized by μCT (Fig. 6a-f). In the foams prepared
3.3. Structure of the dried foam cell walls dictates thermal and
with <125 μm particles, more air is trapped within cell walls, resulting
mechanical properties
in its lower density and thermal conductivity. The <1000 μm size
fraction shows a synergistic effect between average particle length
To understand how our dry wood foams compare to conventional
(Table 1) and polyhedral cell wall length (Gibson & Ashby, 1997a),
petrochemical-based insulation, the apparent density, thermal conduc­
where only one or two particles make up each cell wall. Since wood
tivity, and compressive strength at 10 % strain were measured (Fig. 5).
particles are rigid and cannot entangle, particle-particle “bonds”
All values fall within a similar range; however, there is a trend with
represent weak points within the structure and are minimized when
respect to particle size. Increasing particle size increased the values in all
using larger particles. This results in a polyhedral cell structure that
measured properties. Essentially, smaller wood particles generated
better resists deformation, with strength also helped by the higher foam
foams with lower apparent densities (from 0.14 to 0.10 g cm− 3), which
density.
resulted in lower thermal conductivity (from 0.047 to 0.042 W m− 1 K− 1)
The average size and size distribution of the mm-scale polyhedral
and lower compressive strength (from 71 to 44 kPa). One-way ANOVA
cells (pores) in dried foams is very similar between all size-fractionated
with a 90 % confidence level revealed statistically significant differences
samples (Fig. 6g) – the average cell size for the <1000 μm sample is 1.1
between size-fractionated samples for density (F(2,15) = 95.7, p < 2.88
mm and for the <300 μm and <125 μm samples is 1.2 mm. This simi­
× 10− 9), thermal conductivity (F(2,15) = 35.1, p < 2.21 × 10− 6), and
larity in physical structure supports our observation about cell wall
compressive stress at 10 % strain (F(2,12) = 4.6, p < 0.03).
structure, as the cells themselves may be regarded as a constant, and
EPS, PU, and mineral wools have a wide range of properties
supports the overall similarities in thermal conductivity and compres­
depending on the manufacturer and product variant, but generally our
sive strength. The polyhedral cells are comparable to those observed in
wood foam properties are similar to these conventional foams (Schia­
our previous work (Ferreira et al., 2022) and are typical of foam struc­
voni et al., 2016). Conversion of thermal conductivity and foam thick­
tures in general (Gibson & Ashby, 1997b). If we regard the cell size as a
ness to R-value per inch (used widely in the construction industry) yield
constant, then the cell wall structures are necessarily less dense for
3.4, 3.3, and 3.1 for <125 μm, <300 μm, and <1000 μm, respectively.
smaller wood particle foams to account for their significantly lower
For comparison, EPS, sprayed PU, and mineral wool have an R-value per
density (Fig. 5a). Quantifying the cell wall density through μCT image
inch around 4, depending on density (Tariku et al., 2023), meaning that
analysis was approximated by converting a stack of 500 image slices to
our foams need to be slightly thicker for equivalent insulative perfor­
black and white images, with white representing the solid foam struc­
mance. Despite statistical differences observed between the final prop­
ture and black representing voids within the foam. The ratio of white to
erties of the three size-fractionated samples, all our foams meet or

Fig. 5. (a) Apparent density, (b) thermal conductivity, and (c) strength at 10 % compression for dried foams produced from the three size-fractionated wood particle
samples (<125, <300, and <1000 μm).

7
E. Dobrzanski et al. Carbohydrate Polymers 335 (2024) 122077

Fig. 6. μCT cross-sectional images from foams prepared with size-fractionated wood particle samples: (a, d) < 125 μm; (b, e) < 300 μm; (c, f) < 1000 μm. (d, e, f)
polyhedral cells illustrated to guide the reader, and (g) resulting cell size frequency distributions.

black pixels then digitally represents the density of each foam. The ratios composite foams from other unrefined rigid biomass, like wood bark and
for <125 μm, <300 μm, and <1000 μm are 0.06, 0.08, and 0.09, agricultural husks, such that material properties can be controlled
respectively, supporting our observation that density decreases as par­ through particle size fractionation.
ticle size decreases. Overall, the thermal and mechanical properties of our wood particle
foams met or exceeded ASTM requirements for cellular polystyrene
4. Conclusion insulative foams and are comparable to commercially available insu­
lative foams. We recommend using the largest size fraction (< 1000 μm)
Aqueous foams prepared from milled wood particles were highly of wood particles tested here for the development of insulative foams,
stable, with low liquid drainage (<5 %) and negligible foam collapse. because it requires the least amount of milling, i.e., the least amount of
The high stability was attributed to the use of surfactants, high solids time, energy, or cost, with minimal thermal and mechanical property
content, and wood particles with heterogeneous size and shape. While loss. Future work will focus on adding functionality to the wood foams,
amphiphilic wood particles with heterogeneous size distribution and including flame retardancy, moisture resistance, and antimicrobial
shape are not generally considered desirable characteristics to work properties, to advance the development of this recyclable and sustain­
with, they contribute to our high aqueous foam stability by allowing able bio-based alternative for petrochemical-based foams.
high solids content and close packing of bubbles and particles (amphi­
philic wood particles) and through interlocking particles which physi­ CRediT authorship contribution statement
cally restrict bubble movement (heterogeneous particle size and shape).
A minor size effect was observed for aqueous foam stability, where Elizabeth Dobrzanski: Writing – review & editing, Writing – orig­
smaller wood particles with higher specific surface area resulted in inal draft, Visualization, Project administration, Methodology, Investi­
higher stability. High aqueous foam stability allowed for slow and low gation, Conceptualization. Elisa S. Ferreira: Writing – review & editing,
energy drying because the foam structure was not compromised over Writing – original draft, Methodology, Investigation, Funding acquisi­
time. There is potential to further decrease drying time and energy by tion, Conceptualization. Praphulla Tiwary: Writing – review & editing,
removing draining liquid during drying through mold design, by adding Project administration, Methodology, Funding acquisition, Conceptu­
holes to the mold or using a mesh, and by implementing a vacuum to alization. Prashant Agrawal: Writing – review & editing, Project
assist liquid drainage. administration, Methodology, Funding acquisition, Conceptualization.
Dry foams containing larger wood particles increased in density Richard Chen: Writing – review & editing, Project administration,
(from 0.10 to 0.14 g cm-3), compression strength (from 43.8 to 70.9 Methodology, Funding acquisition, Conceptualization. Emily D. Cran­
kPa), and thermal conductivity (from 0.042 to 0.047 W m− 1 K− 1). The ston: Writing – review & editing, Writing – original draft, Visualization,
size-property relationships can primarily be explained by the cell wall Supervision, Project administration, Funding acquisition,
structure of the foams. Larger particles produced denser structures Conceptualization.
where only one or two particles made up each cell wall, working syn­
ergistically with the polyhedral cell wall size. Conversely, lighter foams
were obtained with smaller wood particles as they provided higher wet Declaration of competing interest
stability which led to lower density and lower thermal conductivity. The
impact of particle size on the foam density can likely be anticipated in The authors declare the following financial interests/personal re­
lationships which may be considered as potential competing interests:

8
E. Dobrzanski et al. Carbohydrate Polymers 335 (2024) 122077

Emily D. Cranston reports financial support was provided by Mitacs. He, S., Liu, C., Chi, X., Zhang, Y., Yu, G., Wang, H., Li, B., & Peng, H. (2019). Bio-inspired
lightweight pulp foams with improved mechanical property and flame retardancy
Praphulla Praphulla & Richard Chen has patent #CA 3194371 issued to
via borate cross-linking. Chemical Engineering Journal, 371, 34–42. https://doi.org/
Plantee Bioplastics Inc. If there are other authors, they declare that they 10.1016/j.cej.2019.04.018
have no known competing financial interests or personal relationships Hjelt, T., Ketoja, J. A., Kiiskinen, H., Koponen, A. I., & Pääkkönen, E. (2020). Foam
that could have appeared to influence the work reported in this paper. forming of fiber products: A review. Journal of Dispersion Science and Technology.
https://doi.org/10.1080/01932691.2020.1869035
Hyll, K. (2015). Size and shape characterization of fines and fillers-a review. PHYSICS
Data availability Nordic Pulp & Paper Research Journal, 30(3), 466–487.
Ihnat, V., Lübke, H., Balbercak, J., & Kuňa, V. (2020). Size reduction downcycling of
waste wood. Review. Wood Research, 65(2), 205–220. https://doi.org/10.37763/
Data will be made available on request. wr.1336-4561/65.2.205220
Jiang, S., Zhang, M., Li, M., Zhu, J., Ge, A., Liu, L., & Yu, J. (2021). Cellulose-based
Acknowledgements composite thermal-insulating foams toward eco-friendly, flexible and flame-
retardant. Carbohydrate Polymers, 273. https://doi.org/10.1016/j.
carbpol.2021.118544
The authors thank James Drummond (PPC) for μCT images, Prof. Karinkanta, P., Ämmälä, A., Illikainen, M., & Niinimäki, J. (2018). Fine grinding of wood
Shahabaddine Sokhansanj, Prof. Orlando Rojas, Saeid Soltanian – Overview from wood breakage to applications. In , Vol. 113. Biomass and bioenergy
(pp. 31–44). Elsevier Ltd.. https://doi.org/10.1016/j.biombioe.2018.03.007
(AMPEL) for equipment training and usage, and Prof. Mark Martinez for Keränen, J. T., Jetsu, P., Turpeinen, T., & Koponen, A. I. (2023). Dewatering and
LUMiSizer data visualization algorithm development. Plantee Bio­ structural analysis of foam-formed, lightweight fibrous materials. In , Vol. 18, Issue 1.
plastics acknowledges funding from Natural Resources Canada (NRCan, BioResources (pp. 531–549). North Carolina State University. https://doi.org/
10.15376/biores.18.1.531-549.
2019-F0041-C00001) and the authors acknowledge the Natural Sciences Kizha, A. R., & Han, H. S. (2016). Processing and sorting forest residues: Cost,
and Engineering Research Council of Canada (RGPIN-2018-06818). E. productivity and managerial impacts. Biomass and Bioenergy, 93, 97–106. https://
D. and E. S. F. thanks Mitacs and Plantee Bioplastics (Mitacs Accelerate, doi.org/10.1016/j.biombioe.2016.06.021
Koponen, A., Jäsberg, A., Lappalainen, T., & Kiiskinen, H. (2018). The effect of in-line
IT22973 and IT30768). E. D. C. is grateful for support and recognition foam generation on foam quality and sheet formation in foam forming. Nordic Pulp &
through the University of British Columbia’s President’s Excellence Paper Research Journal, 33(3), 482–495. https://doi.org/10.1515/npprj-2018-3051
Chair initiative, the NSERC E.W.R. Steacie Memorial Fellowship and the Koponen, A., Timofeev, O., Jäsberg, A., & Kiiskinen, H. (2020). Drainage of high-
consistency fiber-laden aqueous foams. Cellulose, 27(16), 9637–9652. https://doi.
Canadian Foundation for Innovation (John R. Evans Leaders Fund) for
org/10.1007/s10570-020-03416-y
equipment. Kumar, V., Tyagi, L., & Sinha, S. (2011). Wood flour-reinforced plastic composites: A
review. Reviews in Chemical Engineering, 27(5–6), 253–264. https://doi.org/10.1515/
REVCE.2011.006/MACHINEREADABLECITATION/RIS
Appendix A. Supplementary data Lehmonen, J., Retulainen, E., Paltakari, J., Kinnunen-Raudaskoski, K., & Koponen, A.
(2020). Dewatering of foam-laid and water-laid structures and the formed web
Supplementary data to this article can be found online at https://doi. properties. In , Vol. 27, Issue 3. Cellulose (pp. 1127–1146). Springer. https://doi.org/
10.1007/s10570-019-02842-x.
org/10.1016/j.carbpol.2024.122077.
Liu, Y., Kong, S., Xiao, H., Bai, C. Y., Lu, P., & Wang, S. F. (2018). Comparative study of
ultra-lightweight pulp foams obtained from various fibers and reinforced by MFC.
References Carbohydrate Polymers, 182, 92–97. https://doi.org/10.1016/j.carbpol.2017.10.078
Llantoy, N., Chàfer, M., & Cabeza, L. F. (2020). A comparative life cycle assessment
(LCA) of different insulation materials for buildings in the continental Mediterranean
Banavath, H. N., Bhardwaj, N. K., & Ray, A. K. (2011). A comparative study of the effect
climate. In , Vol. 225. Energy and buildings. Elsevier Ltd.. https://doi.org/10.1016/j.
of refining on charge of various pulps. Bioresource Technology, 102(6), 4544–4551.
enbuild.2020.110323
https://doi.org/10.1016/J.BIORTECH.2010.12.109
Loeffler, D., & Anderson, N. (2018). Impacts of the mountain pine beetle on sawmill
Bapat, S., & Segets, D. (2020). Sedimentation dynamics of colloidal formulations through
operations, costs, and product values in Montana. In , Vol. 68, Issue 1. Forest Products
direct visualization: Implications for fuel cell catalyst inks. ACS Applied Nano
Journal (pp. 15–24). Forest Products Society. https://doi.org/10.13073/FPJ-D-17-
Materials, 3(8), 7384–7391. https://doi.org/10.1021/acsanm.0c01467
00041.
Bekhta, P., & Hiziroglu, S. (2002). Theoretical approach on specific surface area of wood
Mangi, S. A., Binti Jamaluddin, N., Siddiqui, Z., & Memon, A. (2019). Utilization of
particles. Forest Products Journal, 52, 72–76.
sawdust in concrete masonry blocks: A review. Mehran University Research Journal of
Bhattacharya, A., & Ray, P. (2004). Studies on surface tension of poly (vinyl alcohol):
Engineering and Technology, 38(2), 487–494. https://doi.org/10.3316/
Effect of concentration, temperature, and addition of chaotropic agents. Journal of
INFORMIT.291965731461322
Applied Polymer Science, 93(1), 122–130. https://doi.org/10.1002/app.20436
Natural Resources Canada. (2020, July 5). Plastics challenge: Development of next
Chen, L., Wang, S., Wang, S., Chen, C., Qi, L., Yu, L., … Chen, C. (2022). Scalable
generation bio-based foam insulation. Government of Canada.
production of biodegradable, recyclable, sustainable cellulose-mineral foams via
Nechita, P., & Năstac, S. M. (2022). Overview on foam forming cellulose materials for
coordination interaction assisted ambient drying. ACS Nano, 16(10), 16414–16425.
cushioning packaging applications. Polymers, 14(10), 1963. https://doi.org/
https://doi.org/10.1021/acsnano.2c05635
10.3390/POLYM14101963, 2022, Vol. 14, Page 1963.
Circular Bio-based Europe Joint Undertaking. (2023). Reinvent: Novel products for
Nguyen, T. T., & Indraratna, B. (2019). Micro-CT scanning to examine soil clogging
construction and automotive industries based on bio materials and natural fibres.
behavior of natural fiber drains. Journal of Geotechnical and Geoenvironmental
https://www.cbe.europa.eu/projects/reinvent.
Engineering, 145(9). https://doi.org/10.1061/(asce)gt.1943-5606.0002065
Conzelmann, N. A., Penn, A., Partl, M. N., Clemens, F. J., Poulikakos, L. D., &
Niu, H., Chen, X., Luo, T., Chen, H., & Fu, X. (2022). The interfacial behavior and long-
Müller, C. R. (2020). Link between packing morphology and the distribution of
term stability of emulsions stabilized by gum arabic and sugar beet pectin.
contact forces and stresses in packings of highly nonconvex particles. Physical Review
Carbohydrate Polymers, 291. https://doi.org/10.1016/j.carbpol.2022.119623
E, 102(6). https://doi.org/10.1103/PhysRevE.102.062902
Odabas, N., Henniges, U., Potthast, A., & Rosenau, T. (2016). Cellulosic fines: Properties
Corbett, L. J., Withey, P., Lantz, V. A., & Ochuodho, T. O. (2016). The economic impact
and effects. In , Vol. 83. Progress in materials science (pp. 574–594). Elsevier Ltd..
of the mountain pine beetle infestation in British Columbia: Provincial estimates
https://doi.org/10.1016/j.pmatsci.2016.07.006
from a CGE analysis. Forestry: An International Journal of Forest Research, 89(1),
Ou, L., & Cai, H. (2020). Dynamic life-cycle analysis of fast pyrolysis biorefineries:
100–105. https://doi.org/10.1093/FORESTRY/CPV042
Impacts of feedstock moisture content and particle size. ACS Sustainable Chemistry &
Cucharero, J., Ceccherini, S., Maloney, T., Lokki, T., & Hänninen, T. (2021). Sound
Engineering, 8(16), 6211–6221. https://doi.org/10.1021/acssuschemeng.9b06836
absorption properties of wood-based pulp fibre foams. Cellulose, 28(7), 4267–4279.
Praphulla, P., Chen, R., & Agrawal, P. (2022). FOREST RESIDUE BASED RIGID FOAM
https://doi.org/10.1007/s10570-021-03774-1
PRODUCTS AND PROCESSES OF MANUFACTURE (Patent WO 2022/073126).
Ferreira, E. S., Dobrzanski, E., Tiwary, P., Agrawal, P., Chen, R., & Cranston, E. D. (2022).
Rasband, W. S. (2012). ImageJ user guide. U. S. National Institutes of Health. https://i
Insulative wood materials templated by wet foams. Materials Advances. https://doi.
magej.nih.gov/ij/docs/guide/146-30.html.
org/10.1039/d2ma00852a
Ren, Q., Li, W., Cui, S., Ma, W., Zhu, X., Wu, M., Wang, L., Zheng, W., Semba, T., &
Ferreira, E. S., Rezende, C. A., & Cranston, E. D. (2021). Fundamentals of cellulose
Ohshima, M. (2023). Improved thermal insulation and compressive property of
lightweight materials: Bio-based assemblies with tailored properties. Green
bimodal poly (lactic acid)/cellulose nanocomposite foams. Carbohydrate Polymers,
Chemistry, 23(10), 3542–3568. https://doi.org/10.1039/d1gc00326g
302. https://doi.org/10.1016/j.carbpol.2022.120419
Gibson, L. J., & Ashby, M. F. (1997a). Cellular solids: Structure and properties (2nd ed., pp.
Rezaei, H., & Sokhansanj, S. (2018). Physical and thermal characterization of ground
175–234). The Press Syndicate of the University of Cambridge.
bark and ground wood particles. Renewable Energy, 129, 583–590. https://doi.org/
Gibson, L. J., & Ashby, M. F. (1997b). Cellular solids: Structure and properties (2nd ed., pp.
10.1016/j.renene.2018.06.038
15–51). The Press Syndicate of the University of Cambridge.
Sanchez-Salvador, J. L., Campano, C., Balea, A., Tarrés, Q., Delgado-Aguilar, M.,
Harb, E., Maalouf, C., Bliard, C., Tenpierik, M., Lachi, M., Bogard, F., & Polidori, G.
Mutjé, P., … Negro, C. (2022). Critical comparison of the properties of cellulose
(2023). Thermal performance of starch/beet-pulp composite bricks for building
nanofibers produced from softwood and hardwood through enzymatic, chemical and
insulation at a wall scale. Case Studies in Construction Materials, 18. https://doi.org/
10.1016/j.cscm.2023.e01851

9
E. Dobrzanski et al. Carbohydrate Polymers 335 (2024) 122077

mechanical processes. International Journal of Biological Macromolecules, 205, Trent, T., Lawrence, V., & Woo, K. (2006). A wood and fibre quality-deterioration model
220–230. https://doi.org/10.1016/j.ijbiomac.2022.02.074 for mountain pine beetle-killed trees by biogeoclimatic subzone. www.pfc.cfs.nrcan.
Schiavoni, S., D’Alessandro, F., Bianchi, F., & Asdrubali, F. (2016). Insulation materials gc.ca.
for the building sector: A review and comparative analysis. Renewable and Sustainable Vosper, F. C., & Wiersma, B. J. (1988). Insulation basics.
Energy Reviews, 62, 988–1011. https://doi.org/10.1016/j.rser.2016.05.045 Wiacek, J., & Molenda, M. (2014). Microstructure and micromechanics of polydisperse
Schubert, M., Panzarasa, G., & Burgert, I. (2023). Sustainability in wood products: A new granular materials: Effect of the shape of particle size distribution. Powder
perspective for handling natural diversity. Chemical Reviews, 123(5), 1889–1924. Technology, 268, 237–243. https://doi.org/10.1016/j.powtec.2014.08.020
https://doi.org/10.1021/ACS.CHEMREV.2C00360/ASSET/IMAGES/LARGE/ Woo, K. L., Watson, P., & Mansfield, S. D. (2005). The effects of mountain pine beetle
CR2C00360_0018.JPEG attack on lodgepole pine wood morphology and chemistry implications for wood and
Seth, R. S. (1990). Fibre quality factors in papermaking — II the importance of fibre fiber quality wood and fiber science. In Wood and fiber science (Vol. 37, Issue 1).
coarseness. MRS Proceedings, 197(1), 143–161. https://doi.org/10.1557/PROC-197- Wu, M., Yu, G., Chen, W., Dong, S., Wang, Y., Liu, C., & Li, B. (2022). A pulp foam with
143/METRICS highly improved physical strength, fire-resistance and antibiosis by incorporation of
Sohn, H. Y., & Moreland, C. (1968). The effect of particle size distribution on packing chitosan and CPAM. Carbohydrate Polymers, 278. https://doi.org/10.1016/j.
density. The Canadian Journal of Chemical Engineering, 46(3), 162–167. https://doi. carbpol.2021.118963
org/10.1002/CJCE.5450460305 Yin, X., Zhang, T., Zhao, T., Wang, K., Xu, Z., & Zhao, Y. (2023). Cellulose-based, flexible
Sun, Y., Chu, Y., Wu, W., & Xiao, H. (2021). Nanocellulose-based lightweight porous polyurethane polyHIPEs with quasi-closed-cell structures and high stability for
materials: A review. In , Vol. 255. Carbohydrate polymers. Elsevier Ltd.. https://doi. thermal insulation. Carbohydrate Polymers, 302. https://doi.org/10.1016/j.
org/10.1016/j.carbpol.2020.117489 carbpol.2022.120385
Tariku, F., Shang, Y., & Molleti, S. (2023). Thermal performance of flat roof insulation Zhang, Y., Lu, Y., Zhang, R., Gao, Y., & Mao, L. (2021). Novel high internal phase
materials: A review of temperature, moisture and aging effects. In Journal of Building emulsions with gelled oil phase: Preparation, characterization and stability
Engineering (Vol. 76)Elsevier Ltd.. https://doi.org/10.1016/j.jobe.2023.107142 evaluation. Food Hydrocolloids, 121. https://doi.org/10.1016/j.
foodhyd.2021.106995

10