PEER-REVIEWED ARTICLE bioresources.

com

WATER RETENTION VALUE MEASUREMENTS OF CELLULOSIC

MATERIALS USING A CENTRIFUGE TECHNIQUE
Qingzheng Cheng,a * Jingxin Wang,a Joseph F. McNeel,a and Peter M. Jacobson a

A centrifugal method has been modified and applied to the assessment

of water retention value (WRV) in cellulosic materials. Microcrystalline
cellulose (MCC), small particles/fibrils isolated from MCC using high-
pressure homogenizer, and pulp fibers saturated in water were
centrifuged at different speeds and times with filter paper and/or a
membrane acting as the filter in the WRV measurement setup. As
centrifugal speed, time, and filter pore-size increased, lower WRVs were
obtained. Smaller MCC particles/fibrils retained more water than the as-
received MCC and pulp fibers. The results are useful for WRV
measurements of cellulosic materials, especially for microfibrillated
cellulose and small cellulosic fibrils.

Keywords: Biomass; Cellulose; Water retention value; Centrifuge; Microfibrillation

Contact information: a: Division of Forestry and Natural Resources, West Virginia University, P. O. Box
6125, Morgantown, WV 26506-6125; *Corresponding author:q.cheng@mail.wvu.edu

INTRODUCTION

Cellulose, isolated from biomass, especially woody biomass, is renewable,

biodegradable, and the most abundant natural biopolymer. There are about 100 billion
metric tons of cellulose produced naturally every year in the world, while the total
biomass is about 280 billion metric tons. Natural cellulosic fibers are synthesized mainly
in plants such as grasses, reeds, stalks, and woody vegetation (Lima and Borsali 2004).
For example, there are about 2.41 million dry tons of wood residues produced per year as
a result of manufacturing processes in West Virginia, which can be used for bio-material
and biofuel productions (Wang et al. 2006). The applications of cellulosic materials, such
as pulp fibers, microfibrillated cellulose (MFC), microcrystalline cellulose (MCC), and
fibril aggregate include paper, foods, cosmetics, paints, nonwoven textiles, oil field
services, medicine, and polymer reinforcements (Turbak et al. 1983; Yano and Nakahara
2004; Zimmermann et al. 2004; Cheng et al. 2007; Cheng et al. 2009).
Several approaches have been explored to characterize cellulosic materials,
including physical and chemical procedures. Cellulose is a high-molecular-weight linear
homopolymer, consisting of repeating β-D-glucopyranosyl units joined by (1-4)
glycosidic linkages in a variety of arrangements. Due to interaction of its hydroxyl groups
with water molecules and the supramolecular structure of different cellulose with
different available inner surface (e.g. less ordered amorphous regions, void structure)
where sorption processes take place, cellulose has a high hygroscopicity. The interaction
between cellulose and water plays a significant role during the chemical and physical
process and isolation of cellulose (Klemm et al. 1998). Water retention value (WRV) is a
useful reference to evaluate the performance of cellulosic materials relative to moisture

Cheng et al. (2010). “Cellulose Water Retention Value” BioResources 5(3), 1945-1954. 1945
PEER-REVIEWED ARTICLE bioresources.com
behavior. The TAPPI UM 256 is a standard test method for pulp WRV to measure the
performance of pulps relative to dewatering behavior on paper machine (TAPPI 1981).
Water retention value is a measure of water retained by a material after centrifuging
under standard conditions. The WRV of many materials has been investigated, especially
cellulosic samples, such as pulp fibers (Stana-Kleinschek et al. 2002; Hubbe and
Heitmann 2007), microfibrillated cellulose (MFC) (Turbak et al. 1983; Herrick et al.
1983), microcrystalline cellulose (MCC) (Tomer and Newton 1999), and cellulosic fibril
aggregates (Cheng et al. 2007; Wang and Cheng 2009). WRV can be used to indirectly
evaluate the degree of fibrillation of the fibers for the MFC and cellulosic fibril
aggregates isolations.
The WRV of cellulosic materials can be measured under different conditions. For
pulp fibers, the standard conditions from TAPPI used for WRV measurement include a
representative specimen with 1400 g/m2 on an oven-dry fiber basis, a relative centrifugal
force (RCF) of 900 G, and a centrifuge duration time of 30 min (TAPPI 1981). A similar
European standard DIN 53 814 for assessing the water retention capacity of cellulosic
fibers is based on a determination of the quantity of water that fibers can absorb and
retain under strictly controlled conditions (fiber soaking for 2 hours and centrifuging for
20 min with 3000 rpm) (Stana-Kleinschek et al. 2002). For smaller cellulosic materials
(e.g. MFC, MCC, cellulosic fibril aggregates), however, there is no standard WVR
measurement method. Glass filtering crucibles or metal cups with fine wire screens are
not suitable for very fine cellulosic particles/fibrils. Furthermore, the RCFs and
centrifugal times vary in different references, such as 2000 G and 15 min for pulp fibers
(Rom et al. 2007), 3600 RPM and 10 min for MFC (Herrick et al. 1983), 900 G and 30
min for cellulosic fibril aggregates (Cheng et al. 2007; Wang and Cheng 2009).
Factors that may influence WRV measurements include sample weights,
centrifugal time and force, pore size of filters used in the measurement setup, and
cellulosic particle/fibril sizes. The objectives of this study were to investigate the effects
of these factors on WRV of cellulosic materials, to determine an appropriate setup for the
WRV measurements of smaller cellulosic particles/fibrils.

EXPERIMENTAL

Materials
Two cellulose sources were used as raw materials: microcrystalline cellulose
(MCC, Avicel PH-101, supplied by FMC BioPolymer) and pulp fiber (supplied by
Kimberly-Clark Worldwide, Inc.). The size of MCC air jet particles in wt % was ≤ 1.00
of +60 mesh and ≤ 35 of +200 mesh. The loose bulk density was 0.26-0.31 g/cc, and the
degree of polymerization of MCC was ≤ 350. The pulp fibers were about 30 µm in width
and 2 to 5 mm in length. Small cellulosic particles and fibrils were isolated from MCC by
using a high-pressure homogenizer (NanoDeBee, Bee International Inc., South Easton,
MA, USA) with 138 MPa pressure and 5 and 20 passes. Whatman filter paper #3
(particle retention: 6 µm), #42 (particle retention: 2.5 µm), and membrane (pore size: 0.2
μm, Whatman Nuclepore) were used to detect the influence of different pore sizes of
filters on WRVs. The scanning electron microscope images (SEM, Hitachi S-4700) of

Cheng et al. (2010). “Cellulose Water Retention Value” BioResources 5(3), 1945-1954. 1946
PEER-REVIEWED ARTICLE bioresources.com
pulp fiber, MCC, and homogenizer treated cellulosic particles/fibrils are shown in Fig. 1.
The dimension and structure varied differently among the three tested materials. The pulp
fibers (~30 µm in width and 2~5 mm in length) were much bigger than the MCC (~50
µm) and treated MCC particles/fibrils (most < 10 µm).

Fig. 1. SEM images of pulp fibers (a), MCC (b), and treated MCC using a homogenizer with 138
MPa pressure and 20 passes (c)

Methods
The water retention value measurement is based on subjecting a sample of
aqueous slurry or water-saturated pulp to a standard centrifugal force with standard
conditions of time, temperature, and apparatus. WRV is defined as the percent ratio of
water contained in the sample after centrifuging at a certain force and time relative to the
dry weight of the sample. A 25-mm diameter stainless steel cap with diameter of 1 mm
holes (D) and a filter (E) were used to hold the wet sample (G) (Fig. 2). Two or four caps
were filled with wet sample of designated weight and placed inside the carriers. After
centrifuging (AccuSpin 400, Fisher Scientific) for a designated time and a designated
relative centrifugal force (RCF) at room temperature, the samples were removed and
placed onto aluminum foil, the “wet weights” were measured, and then the samples were
oven dried at approximately 103 °C until they reached constant mass (dry weights). The
WRV was calculated from Equation 1 (Cheng et al. 2007; Wang and Cheng 2009),

WWet − WDry
WRV = × 100% (1)
WWet

where, WWet is the weight of the sample after centrifuging and WDry is the absolute dry
weight of the sample.

Relative centrifugal force (RCF) is given in multiples of the earth gravity G. It is a

dimensionless number that allows one to compare the efficiency of separation or
sedimentation of diverse instruments because it is independent of the instrument used.
RCF is only related to radius and speed of the centrifugation (Equation 2) (Fisher
Scientific 2003).

Cheng et al. (2010). “Cellulose Water Retention Value” BioResources 5(3), 1945-1954. 1947
PEER-REVIEWED ARTICLE bioresources.com
2
⎛ n ⎞
RCF = 11.18 × ⎜ ⎟ ×r (2)
⎝ 1000 ⎠

where, r is the radius of centrifugation in cm from the centrifuge center to sample mass
center and n is the speed in rpm.

Fig. 2. Scheme of water retention value measurement setup for centrifuge, A: Centrifuge bucket,
B: Adapter, C: Spacer, D: Stainless steel cap with holes, E: Filter paper/Membrane, F: Tissue for
water absorption, G: Sample

Untreated MCC was used to investigate how sample weight, centrifugal time,
centrifugal force, and filters with different pore sizes influence the WRVs of cellulosic
materials. Samples of certain weight were immersed in distilled water for 2-3 days to
make sure that the material was fully saturated before centrifuging. For the sample weight
test, five weight levels of each specimen were used: 0.25, 0.5, 0.75, 0.1, and 1.25 grams
for MCC, and 0.06, 0.12, 0.18, 0.24, and 0.3 grams for treated MCC using a NanoDeBee
homogenizer at 138 MPa and 5 passes (MCC were saturated in distilled water for 2-3
days before homogenization). Three RCF levels: 900, 1350, and 1800 G, and four
centrifugal time levels: 5, 10, 20, and 30 min. were used. The Whatman filter paper #3
(particle retention: 6 µm), #42 (particle retention: 2.5 µm), and membrane with 0.2-μm
pore-size (Whatman Nuclepore) were used for examining the effects of filters of different
pore sizes on WRV measurements. The pulp fiber and treated MCC using a homogenizer
at 138 MPa and 20 passes were also used to detect the effects of cellulosic particle/fibril
sizes on the WRV measurements. Six to eight replicates of each test were conducted to
examine the variation of the WRVs. Single parameter test of analysis of variance
(ANOVA) was used to examine the influence significance of the measurement conditions
on WRVs (α=0.05).

Cheng et al. (2010). “Cellulose Water Retention Value” BioResources 5(3), 1945-1954. 1948
PEER-REVIEWED ARTICLE bioresources.com

RESULTS AND DISCUSSION

Effects of Specimen Weight

Figure 3 shows the WRVs of MCC (Fig. 3a) and treated MCC (Fig. 3b) at varied
specimen masses. Only the effect of lower specimen weight (between 0.25 g to 0.5 g) on
WRVs of MCC was significant (p=0.0058) (Fig. 3a). Although there was about 12%
reduction in WRVs when specimen weight changed from 0.25 g to 0.5 g, the difference
of WRV change was not dramatic, as shown by the error bar overlap. There were no
significant differences among the WRVs of other different weight specimens (from 0.5 g
to 1.25 g) (p values from 0.06 to 0.91). The higher standard deviation for lower weight
samples may be attributed to the scale and drying processes. The WRV is independent of
the surface area and hence that the centrifuge is effectively removing surface water in
some weight range for large cellulosic materials, such as filter paper fibers (Kato and
Cameron 1999). For MCC, a higher specimen weight could be better due to smaller
influences and lower standard deviations. A sample weight of 0.5 grams was used for
most of the following experiments. For treated MCC (138 MPa and 5 passes using a
homogenizer), the results showed that all the sample weights (from 0.06 g to 0.3 g) did
not significantly influence the WRVs (p values from 0.05 to 0.88) (Fig. 3b), which
indicates that WRV is independent to the sample weights. However, it is notable that
higher standard deviations for smaller mass samples were mainly attributable to the lower
masses. More care should be taken during the sample weight measurements for small
samples.

Fig. 3. WRVs with standard deviations of MCC with centrifugal force of 900 G and time of 5 min
(a) and treated MCC with centrifugal force of 900 G and time of 20 min (b) at varied specimen
masses

Cheng et al. (2010). “Cellulose Water Retention Value” BioResources 5(3), 1945-1954. 1949
PEER-REVIEWED ARTICLE bioresources.com
Effects of Centrifugal Time and Force
Generally, WRVs of MCC decreased with the increase of centrifugal time and
RCFs (Fig. 4). For RCF of 900 G, the WRV reduction trend was somewhat linear, while
there were no significant differences between 20 min and 30 min for both RCFs of 1350
G (p=0.121) and 1800 G (p=0.087), respectively, which indicates that no more water
could be forced out from materials under these forces after 20 min. At higher centrifugal
speeds (higher RCFs), due to higher pressure forced to the materials, it caused lower
WRVs than those of lower RCFs. Similar results showed that increasing centrifugal
speed can decrease the measured WRVs of the kenaf core material because higher forces
can force out more water from the material than low forces (Lips et al. 2009).
120

5 Minutes 10 Minutes
100 20 Minutes 30 Minutes

80
WRV (%)

60

40

20

0
900 1350 1800

Centrifuge Force (G)

Fig. 4. Water retention values of MCC (0.5 g) with varied centrifugal times and varied relative
centrifugal forces

Effects of Filters with Different Pore Sizes

The pore size of the filter paper/membrane was found to dramatically influence
the MCC’s WRVs for all four centrifugal times tested (Fig. 5). The bigger pore-size
filters are used, the lower WRVs are possible to obtain. For example, at the centrifugal
time of 5 min, the WRV from filter paper #3 (particle retention 6 µm) was 24% and 40%
lower than those from filter paper #42 (particle retention: 2.5 µm) and membrane (0.2 μm
pore-size), respectively. Larger pore-size resulted in lower WRVs because water is easily
forced out through bigger pore-size filters, and cellulosic fine particles may not block the
filter pores easily during the centrifuging process. This indicates that choosing filters of
appropriate size is important for the setup of WRV measurements. When filter paper #3
was used, there was no significant difference between the WRVs of centrifugal times of 5
and 10 min (p=0.601). Approximately a 5% decrease was exhibited gradually from 10
min to 20 min, and then to 30 min. However, for the membrane, centrifugal time
influenced the WRVs drastically due to the small pores that made the water passing the
filter much slower. Meanwhile, it is worthy of note that the materials of filter paper
(cellulose fiber) and the membrane (polycarbonate) were different, which may affect the
WRV measurements because they have different hydrophilic properties.

Cheng et al. (2010). “Cellulose Water Retention Value” BioResources 5(3), 1945-1954. 1950
PEER-REVIEWED ARTICLE bioresources.com
160

Membrane # 42 Filter # 3 Filter

140

120
WRV (%)

100

80

60

40

20

0
5 10 20 30

Centrifuge Time (Minutes)

Fig. 5. Water retention values of MCC (0.5 g) with different filter paper/membranes at varied
centrifugal times (relative centrifugal forces: 900 G).

Effects of Cellulosic Particle/Fibril Sizes

The WRVs of pulp fiber samples and homogenizer treated MCC particles/fibrils
(138 MPa and 20 passes) were derived at varied centrifugal time, force, and filter
paper/membrane (Tables 1 and 2). As shown in Fig. 4 for the MCC, the WRVs of pulp
fiber decreased with the increasing of centrifugal times and RCFs (Table 1). Also, the
filters with different pore sizes were found to drastically influence the pulp fiber’s WRVs
as well. When the centrifugal time is 5 min, the WRV using membrane was 40% higher
than that of filter paper #42, and 43% higher than that of filter paper #3.

Table 1. WRV of Pulp Fiber Samples with Varied Centrifugal Time, Force, and
Filter Paper/Membrane (mass: 0.5 g)
Centrifugal Time (Min.) Centrifugal Force (G) Filter WRV ± STD* (%)
30 900 #3 142.3 ± 3.8
30 1350 #3 104.6 ± 6.1
30 1800 #3 90.7 ± 0.8
20 1800 #3 90.5 ± 4.1
10 1800 #3 97.4 ± 3.1
5 1800 #3 105.7 ± 2.2
5 1800 #42 112.3 ± 1.1
5 1800 Membrane 185.8 ± 95.4
* STD: standard deviation.

Cheng et al. (2010). “Cellulose Water Retention Value” BioResources 5(3), 1945-1954. 1951
PEER-REVIEWED ARTICLE bioresources.com
Table 2. WRVs of Homogenizer Treated MCC (HPH: 138 MPa & 20 passes)
Particles/Fibrils Samples with Varied Centrifugal Time, Force, and Filter
Paper/Membrane (mass: ~ 0.08 g)
Centrifugal Time (Min.) Centrifugal Force (G) Filter WRV ± STD* (%)
30 900 #3 523.0 ± 25.2
30 1350 #3 452.9 ± 54.4
30 1800 #3 360.0 ± 33.0
30 1800 #42 380.0 ± 22.9
30 1800 Membrane 548.1 ± 58.3
20 1800 Membrane 767.1 ± 187.2
10 1800 Membrane ---**
* STD: standard deviation; ** Not centrifuged completely, and with some water on the top of the
samples.

Again, the WRVs of homogenizer-treated MCC particles/fibrils also decreased

along with the increase of centrifugal time and forces (Table 2). The filters with different
pore sizes influenced the WRVs of the smaller cellulosic particles/fibrils much more than
the big particles/fibers. For example, when the membrane was used, the free water could
not be easily forced out through the membrane filter completely at lower centrifugal time
(10 min), so that water on the top of the samples could be left and observed after
centrifuging. The possible reasons for water remaining in the samples include; that the
small cellulosic fibrils with sizes under 1 µm may completely block the filter pores
and/or a very dense thin film just above the filter may be formed and may also block the
water passing the filter pores during the centrifugation. Several mechanisms may be used
to explain this phenomenon, such as sealing mechanism (Hubbe and Heitmann 2007).
Furthermore, much higher standard deviations were obtained when the membrane,
shorter centrifugal time, and/or lower centrifugal force were used, especially for treated
MCC with larger passes (e.g. 20 passes) (Table 2), while the standard deviations were
much smaller for the treated with fewer passes (e.g. 5 passes) (Fig. 3b), which included
less very fine cellulosic particles/fibrils. All of these results reveal that higher centrifugal
time and/or higher force should be used for smaller cellulosic particles/fibrils when a
very small pore-size membrane is used as the filter in the WRV measurement setup. For
example, for the MCC treated with homogenizer with 138 MPa and 20 passes, a
centrifugal force of 1800 G or higher and time of 30 min. or longer should be used (Table
2).
It is apparent that the WRVs varied depending on materials such as pulp fiber
(Table 1), MCC (Figs. 3 and 4), and homogenizer-treated cellulosic particles/fibrils
(Table 2). The WRV of cellulosic particles/fibrils was about six times of that of MCC.
These cellulosic materials showed different WRVs because the WRV depends on surface
properties of fibers/particles, and water content of cellulosic fibers/particles contained on
surface and bulk water (Maloney et al. 1999; Forsstrom et al. 2005). The homogenizer-
treated cellulosic particles/fibrils, also called MFC, were much smaller than the original
MCC material and highly microfibrillated. The cellulosic particles/fibrils or MFCs had
higher surface area and volume than big size pulp fibers and MCCs. So MFCs could

Cheng et al. (2010). “Cellulose Water Retention Value” BioResources 5(3), 1945-1954. 1952
PEER-REVIEWED ARTICLE bioresources.com
retain higher WRVs because of the fibril and microfibril surface and volumetric
phenomena (Herrick et al., 1983). The smaller the particles/fibrils isolated from raw
materials, the higher the WRVs obtained.

CONCLUSIONS

1. A modified centrifuge setup with stainless steel cap and filters can be used to measure
water retention values (WRVs) of cellulosic materials, especially for small cellulosic
particles/fibrils or microfibrillated cellulose.
2. Lower specimen weight may result in higher standard deviations of WRVs.
3. Lower WRVs of cellulosic materials could be obtained through using higher
centrifugal speed, longer time, and larger pore size of filters.
4. Smaller MCC particles/fibrils can retain more water than untreated MCC and pulp
fibers.
5. Filter membranes with very small pore sizes are not suitable for small cellulosic
particles/fibrils or MFC for the WRV measurement setup if the centrifugal force is
not high enough and/or the centrifugal time is not long enough. Higher forces such as
more than 1800 G for more than 20 min are recommended for these cellulosic
materials.
6. Same filter or membrane should be used for WRV measurement in this setup.

ACKNOWLEDGMENTS

This study was supported in part by the USDA Wood Utilization Research
Special Program at West Virginia University.

REFERENCES CITED

Cheng, Q., Wang S., and Rials, T. G. (2009). “Poly(vinyl alcohol) nanocomposites
reinforced with cellulose fibrils isolated by high intensity ultrasonication,”
Composites Part A: Appl. Sci. Manufact. 40, 218-224.
Cheng, Q., Wang, S., Rials T., and Lee, S. H. (2007). “Physical and mechanical
properties of polyvinyl alcohol and polypropylene composite materials reinforced
with fibril aggregates isolated from regenerated cellulose fiber,” Cellulose, 14(6),
593-602.
Fisher Scientific (2003). Centrifuge AccuSpin 400, “Operating Instruction,” p. 31.
Forsström, J., Andreasson, B., and Wågberg, L. (2005). “Influence of pore structure and
water retaining ability of fibers on the strength of papers from unbleached kraft
fibers,” Nord. Pulp Pap. Res. J. 20(2), 176-185.
Herrick, F. W., Casebier, R. L., Hamilton, J. K., and Sandberg, K. R. (1983).
“Microfibrillated cellulose: morphology and accessibility,” J Appl. Polymer Sci.:
Appl. Polymer Symp. 37, 797-813.

Cheng et al. (2010). “Cellulose Water Retention Value” BioResources 5(3), 1945-1954. 1953
PEER-REVIEWED ARTICLE bioresources.com
Hubbe, M. A., and Heitmann, J. A. (2007). “Review of factors affecting the release of
water from cellulosic fibers during paper manufacture,” BioRes. 2(3), 500-533.
Kato, K. L., and Cameron R. E. (1999). “Structure-property relationships in thermally
aged cellulose fibers and paper,” J Applied Polymer Sci. 74, 1465-1477.
Klemm, D., Philipp, B., Heinze, T., Heinze, U. and Wagenknecht, W. (1998).
“Comprehensive Cellulose Chemistry. Fundam. Anal. Methods,” Wiley-VCH,
Weinheim, New York, Chichester, Brisbane, Singapore, Toronto.
Lima, M. M. D. and Borsali, R., 2004. “Rodlike cellulose microcrystals: Structure,
properties, and applications,” Macromolecular Rapid Communications 25(7), 771-
787.
Lips, S. J. J., de Heredia, G. M. I., den Kamp, R., and van Dam, J. E. G., (2009). “Water
absorption characteristics of kenaf core to use as animal bedding material,” Industrial
Crops and Products 29(1), 73-79.
Maloney, T. C., Jaine, J. E., and Paulapuro, H. (1999). “Comments on the measurement
of cell wall water,” Tappi. J. 82(9), 125-127.
Stana-Kleinschek, K., Ribitsch, V., Kreze T., and Fras L. (2002). “Determination of the
adsorption character of cellulose fibres using surface tension and surface charge,”
Mat. Res. Innovat. 6, 13-18.
TAPPI. (1981). TAPPI useful method UM256. “Water retention value (WRV),” TAPPI
Useful Methods, TAPPI Press, Atlanta, USA.
Tomer, G., and Newton, J. M. (1999). “A centrifuge technique for the evaluation of the
extent of water movement in wet powder masses,” Int. J. Pharm. 188, 31-38
Turbak, A. F., Snyder, F. W. and Sandberg, K. R. (1983). “Microfibrillated cellulose,
new cellulose product: Properties, uses, and commercial potential,” J. Appl. Polymer
Sci.: Appl. Polymer Symp. 37, 815-827.
Wang, S., and Cheng, Q. (2009). “A novel method to isolate fibrils from cellulose fibers
by high intensity ultrasonication. Part I: Process optimization,” J. Appl. Polymer Sci.
113: 1270-1275.
Wang, J., Grushecky, S., and McNeel, J. (2006). Biomass Resources, Uses, and
Opportunities in West Virginia, West Virginia University Biomaterials Center,
Division of Forestry and Natural Resources, Morgantown, WV, 103 pp.
Yano, H., and Nakahara, S. (2004). “Bio-composites produced from plant microfiber
bundles with a nanometer unit web-like network,” J. Mat. Sci. 39(5), 1635-1638.
Zimmermann, T., Pohler, E., and Geiger, T. (2004). “Cellulose fibrils for polymer
reinforcement,” Adv. Engin. Mat. 6(9): 754-761.

Article submitted: March 13, 2010; Peer review completed: July 14, 2010; Revised
version accepted: July 29, 2010; Published: July 29, 2010.

Cheng et al. (2010). “Cellulose Water Retention Value” BioResources 5(3), 1945-1954. 1954