J Ind Microbiol Biotechnol (2008) 35:331–341

DOI 10.1007/s10295-008-0322-0

ORIGINAL PAPER

Flexible bioreWnery for producing fermentation sugars,

lignin and pulp from corn stover
Kiran L. Kadam · Chim Y. Chin · Lawrence W. Brown

Received: 10 October 2007 / Accepted: 21 January 2008 / Published online: 14 February 2008
© Society for Industrial Microbiology 2008

Abstract A new bioreWning process is presented that adhesive production is discussed as well as its use as
embodies green processing and sustainable development. In cement and feed binder. As a baseline application the hemi-
the spirit of a true bioreWnery, the objective is to convert cellulosic sugars captured in the hydrolyzate liquor can be
agricultural residues and other biomass feedstocks into used to produce ethanol, but potential utilization of xylose
value-added products such as fuel ethanol, dissolving pulp, for xylitol fermentation is also feasible. Markets and values
and lignin for resin production. The continuous biomass of these applications are juxtaposed with market penetra-
fractionation process yields a liquid stream rich in hemicel- tion and saturation.
lulosic sugars, a lignin-rich liquid stream, and a solid cellu-
lose stream. This paper generally discusses potential Keywords BioreWnery · Cellulosic ethanol ·
applications of the three streams and speciWcally provides Low-molecular weight lignin · Pulping · Dissolving pulp
results on the evaluation of the cellulose stream from corn
stover as a source of fermentation sugars and specialty
pulp. Enzymatic hydrolysis of this relatively pure cellulose Introduction
stream requires signiWcantly lower enzyme loadings
because of minimal enzyme deactivation from nonspeciWc Fuel ethanol is in ascendance recently. The U.S. is planning
binding to lignin. A correlation was shown to exist between to replace its gasoline consumption by 20% over the next
lignin removal eYciency and enzymatic digestibility. The 10 years with alternative fuels [8]. The current fuel ethanol
cellulose produced was also demonstrated to be a suitable market in the U.S. is about 180 billion gal/year. This pro-
replacement for hardwood pulp, especially in the top ply of vides an opportunity for cellulosic ethanol since corn-based
a linerboard. Also, the relatively pure nature of the cellu- ethanol can supply only about 12.8–17.8 billion gal/year
lose renders it suitable as raw material for making dissolv- based on NCGA (National Corn Growers Association,
ing pulp. This pulping approach has signiWcantly smaller ChesterWeld, MO, USA) estimates [39].
environmental footprint compared to the industry-standard In this context, a new bioreWning process is presented
kraft process because no sulfur- or chlorine-containing that embodies green processing and sustainable develop-
compounds are used. Although this option needs some min- ment. In the spirit of a true bioreWnery, the objective is to
imal post-processing, it produces a higher value commodity convert agricultural residues and other biomass feed-
than ethanol and, unlike ethanol, does not need extensive stocks into value-added products such as fuel ethanol, dis-
processing such as hydrolysis or fermentation. Potential use solving pulp, and lignin for resin production. Successful
of low-molecular weight lignin as a raw material for wood commercialization of this technology would result in a
sustainable green process with positive environmental
impacts such as reduction in emissions of greenhouse
gases and criteria pollutants. Although data speciWc for
K. L. Kadam (&) · C. Y. Chin · L. W. Brown
corn stover are discussed, the proposed bioreWnery sce-
PureVision Technology, Inc., 511 McKinley St.,
Ft Lupton, CO 80621, USA nario is generically applicable to other biomass feed-
e-mail: Kiran@PureVisionTechnology.com stocks.

123
332 J Ind Microbiol Biotechnol (2008) 35:331–341

Alternative approaches to developing a bioreWnery 6. Leads to Xexible bioreWneries that are capable of pro-
ducing a wide range of petroleum-substitute prod-
There are several technologies for developing a bioreWnery, ucts—not limited to biofuels—and shown to be
and they fall into two major approaches: thermochemical economically feasible based on preliminary analysis.
and biochemical. This discussion is limited to the latter. In
this arena, concentrated acid hydrolysis, two-stage dilute
acid hydrolysis, and enzymatic hydrolysis are the major Process description
options for producing fermentation sugars for conversion
into transportation fuels. Concentrated acid hydrolysis is The basic process schematic for biomass fractionation is
well understood; however, the economic upside compared shown in Fig. 1. A screw feeder meters feedstock into the
to its current performance is limited. Although the enzy- extruder (a 27 mm twin-screw extruder from Entek Extrud-
matic approach has to bear the highest current costs due to ers, Lebanon, OR, USA). The feedstock is wetted with
undeveloped enzyme markets, it can leapfrog the more water to assist plug formation. Varying screw pitch along
established acid-based routes due to its untapped long-term the length facilitates the creation of two dynamic plugs in
potential for improved economics [30]. This is attributable the Wrst stage, and high pressure pumps deliver liquids into
to advantages of the enzymatic route such as superior the reactor against operating pressures. Pretreatment occurs
yields, minimal byproduct formation, low energy require- in the Wrst stage via autohydrolysis (although the system is
ments, and milder operating conditions [15, 25, 26, 31, 54]. designed to allow acid addition as well). The sugars formed
Enzymatic hydrolysis of cellulosics requires a pretreat- travel countercurrently and are transported quickly—
ment step, which in the case of PureVision is a fraction- depending on the liquid Xow rate—toward the exit, and
ation step. Some of the chief alternatives include: dilute degradation reactions are hindered as a result. Furthermore,
acid, SO2 impregnation followed by steam explosion, and solid/liquid (S/L) separation is accomplished in situ and at
organosolv. Compared to dilute acid or SO2 impregnation process temperature. A progressive cavity pump is used to
followed by steam explosion, the fractionation process dis- discharge the Wrst stage liquor while maintaining prevailing
cussed here diVers in terms of producing a value-added lig- pressure in the reaction chamber; it is used in an unconven-
nin product rather than burning it as fuel. This approach is tional way in that the pressure is progressively reduced
truer to the spirit of a bioreWnery. NonspeciWc binding of rather than raised. No heat loss occurs because the hot
cellulases to lignin is a major factor during enzymatic dewatered pretreated solids (»50%) advance to the second
hydrolysis and, unlike these approaches, biomass fraction- stage without Xashing as is the case when S/L separation is
ation results in a low-lignin substrate. Also, instead of a decoupled from the hydrolyzer as with a plug-Xow reactor.
batch or plug Xow reactor used in these alternative pretreat- Cocurrent deligniWcation occurs in the second stage with
ments, the fractionation process uses countercurrent mode sodium hydroxide as a catalyst. Two alternating valves dis-
in pretreating the biomass. Although the organosolv pro- charge the reacted slurry while maintaining prevailing pres-
cess also produces a value-added lignin product, the frac- sure. The exiting slurry is subjected to hot S/L separation,
tionation process uses much shorter residence times for yielding a liquid stream rich in lignin and a solid cellulose
deligniWcation and also produces a discrete hemicellulosic product.
sugar stream. The following summarizes distinguishing
features of the PureVision fractionation technology:
1. Ability to fractionate biomass into its three major com-
Biomass 2nd Stage
ponents. Feed
Water Liquor
Discharge
2. Continuous countercurrent pretreatment of biomass. Motor
Water Acid Alkali
Valve
3. Production of low-lignin cellulose (2–4% Klason lig- S/L
Separation
nin on dry weight basis) that requires less enzyme, Extruder Barrel
1st stage
resulting in decreased cost of fermentation sugars and liquor Cellulose
Product
ethanol.
1st Stage: 2nd Stage
4. Production of a puriWed, sulfur-free, low-molecular Feedstock
Wetting
Countercurrent Cocurrent
Hot Water/Acid Alkali
weight (MW) lignin, which can be used as a biobased Extraction Extraction

raw material for manufacturing a myriad of industrial in situ S/L

separation
and consumer products.
= Dynamic plugs
5. Process versatility to gear the cellulose stream toward
glucose or pulp/cellulose derivatives. Fig. 1 Schematic of the PureVision process

123
J Ind Microbiol Biotechnol (2008) 35:331–341 333

Materials and methods T414om98 Internal tearing resistance of paper

T460om02 Air resistance of paper
Feedstock T494om01 Tensile properties of paper and paperboard
T511om02 Folding endurance of paper
Colorado corn stover was used as feedstock, which has the T566om02 Bending resistance of paper.
following major components (dry wt basis): 35.2% glucan,
19.8% xylan and 16.9% lignin.
Results and discussion
Enzymatic hydrolysis
Typical process performance and characteristics of frac-
NREL’s (National Renewable Energy Laboratory, Golden, tions are given here for background. Operating conditions
derived from preliminary optimization are summarized
CO, USA) Laboratory Analytical Procedure #009 was used
here. The Wrst stage—operated at 210°C in autohydrolysis
for enzymatic sacchariWcation of corn stover solids [5].
mode—yields 60–65% xylose recovery. The second
SacchariWcation was carried out in 10 mL Oak Ridge cen-
stage—with operating conditions of T = 220°C, NaOH
trifuge tubes (polycarbonate, Nalgene) and the insoluble
solids concentration was 5 wt%. The tubes were placed in charge = 0.06 g/g biomass, and residence time = 11 min—
the shaker at about 30° angle to ensure bulk mixing. The achieves ¸90% overall deligniWcation and yields about 85–
90% of incoming glucan (solids fraction). Characteristics of
solids were subjected to carbohydrate and Klason lignin
the three fractions are given below.
analyses [50] prior to enzymatic hydrolysis. Sugar concen-
trations were determined by HPLC [51]. First stage liquor: relatively low in furfural and HMF;
furfural <2 g/L or 1.5% of xylan in the incoming corn
Pulp properties stover feedstock (xylan equivalent basis); HMF <0.1 g/
L or 0.05% of glucan in the incoming corn stover feed-
Corn stover pulp was diluted and mixed to break any Wber stock (glucan equivalent basis)
bundles and dirt lumps. The mixer speed was then reduced Solids stream:cellulose with a residual lignin content of
and grit was removed from the pulp by overXow Xotation of 2–3 wt%; can be geared toward pulp/cellulose deriva-
the clean pulp from the agitated bucket. The clean pulp was tives
dewatered; this step also cleaned the pulp removing Second stage liquor: low-MW lignin; average MW of
remaining dissolved materials. The resultant pulp was used 1,000 versus 50,000 for kraft lignin, which is highly
in making handsheets to assess its potential. condensed
Standard tests performed on the handsheets were gram-
A typical bioreWnery would yield from 100 mt of dry corn
mage, caliper, tensile strength, stretch, tensile energy absorp-
stover: 28–30 mt dry cellulose (wet solid/slurry form), 12–
tion (T.E.A.), burst (Mullen), tear, folds, stiVness and porosity
(Gurley). Unbleached pine and Southern hardwood pulps 15 mt dry lignin (dry solid form), and 12–13 mt dry xylose
from a pulp mill were washed and dewatered. The mill pulps (solution form). Various applications of these fractions are
discussed below.
were then reWned in a laboratory valley beater. The standard
composition is equal parts of pine and hardwood for the top
Enzymatic hydrolysis
ply of a linerboard. The blends substitute the Pure Vision pulp
for various amounts of the hardwood pulp. The following
TAPPI (Technical Association of the Pulp and Paper Industry, Economical bioconversion requires the following: a well-
pretreated substrate, an eYcient cellulase system, and an
Norcross, GA, USA) standard protocols were used:
eVective microorganism. Cellulose hydrolysis is eVected
T200sp01 Laboratory beating of pulp (valley beater via a synergistic action of endoglucanases, exoglucanases,
method) and
-glucosidases. Table 1—adapted and extended from
T205sp02 Forming handsheets for physical test of pulp Esteghlalian et al. [14]—lists enzyme-, substrate- and pro-
T220sp01 Physical testing of pulp handsheets cess-related factors that aVect the eYciency of enzymatic
T227om99 Freeness of pulp (Canadian standard freeness) hydrolysis, which constitutes a major cost for the overall
T402sp03 Standard conditioning and testing atmo- bioconversion process [17]. By minimizing residual lignin
spheres for paper, board, pulp handsheets, and related in the substrate, PureVision’s process addresses the issue of
products irreversible binding of enzymes onto lignin.
T403om02 Bursting strength of paper The cellulose stream from the PureVision process is
T411om97 Thickness of paper, paperboard and com- adaptable to either sugar or pulp applications; the high cel-
bined board lulose content of solids from the two-stage process renders

123
334 J Ind Microbiol Biotechnol (2008) 35:331–341

Table 1 Factors aVecting enzymatic hydrolysis

Enzyme related factors Substrate related factors Process related factors

Reaction heterogeneity Cellulose crystallinity Process conWguration (reconciliation

(soluble enzyme vs. insoluble substrate) of optimal conditions for cellulases
vs. ethanologen)
Irreversible binding of enzymes onto lignin Cellulose degree of polymerization Operating parameters: pH, temperature,
duration
Gradual loss of synergism in cellulase mixture Feedstock particle size Enzyme loading
Substrate dependence of synergism and Lignin barrier Solids level
binding (speciWcity) of enzyme (content and distribution)
components
End-product inhibition Substrate’s available surface area Bulk mixing, diVusion limitations
(pore volume)
Thermal inactivation of enzyme Cell wall thickness (coarseness) Batch versus continuous mode
Substrate composition Reactor design

them a good candidate for making dissolving pulp, which 50

Total glucan hydrolysis, % theoretical

can further be converted into cellulose derivatives. Sugar
production will be discussed Wrst. The corn stover solids 40
24 h
R2 = 0.9
from the two-stage process were subjected “as-is” to enzy-
matic hydrolysis using Genencor’s Spezyme. The NREL 30
base case assumes an enzyme loading of 12 FPU/g cellu- 12 h
R2 = 0.9
lose [1]; an enzyme loading of 5 FPU/g cellulose was used 20
in this work. A typical hydrolysis proWle is depicted in
Fig. 2. 10
Figure 3 shows a direct eVect of lignin removal on enzy-
matic digestibility. Conversely, an inverse relationship is 0
found between lignin content of solids and enzymatic 70 75 80 85 90 95 100

digestibility (data not shown). This is attributed to nonspe- Lignin removal, % theoretical

ciWc binding of cellulases to lignin. It can also be inter- Fig. 3 EVect of lignin removal eYciency on enzymatic digestibility
preted to mean that for lower lignin content of solids we
can use proportionately lower enzyme loading to achieve
Predicted enzyme loading to achieve same glucan

14
the same hydrolysis performance. If we assume that we can
12
extrapolate based on the design point with lowest lignin
hydrolysis, FPU/g cellulose

removal eYciency, we get the plot of Fig. 4. 10

Other investigators have shown net enzyme consump-

8
tion with lignin-free cellulose to be reduced by as much as a
6
90
Total glucan hydrolysis, % theoretical

80 4
Untested design space Tested design space

70
2
60
0
50 30 40 50 60 70 80 90 100
Lignin removal, % theoretical
40

30 Fig. 4 Predicted enzyme loading to achieve same glucan hydrolysis

20

10 factor of Wve compared to that needed to hydrolyze biomass

0
after conventional pretreatments that leave a signiWcant
0 24 48 72 96 120 144 168 fraction of residual lignin [35]. The predicted reductions in
Elapsed time, h enzyme consumption are much smaller in our work. Cellu-
Fig. 2 Total glucan hydrolysis proWle for “as is” solids containing 4% lose digestibility in Xow through reactors has also been
lignin at enzyme loading of 5 FPU/g cellulose shown to be related to xylan removal as well as lignin

123
J Ind Microbiol Biotechnol (2008) 35:331–341 335

removal [58]; in our work xylan removal was constant. A ethanol concentrations. The PureVision substrate, besides
caveat should be oVered that extrapolation in the untested needing less enzyme, achieves 50% higher ethanol concen-
design space is probably not entirely reliable. However, tration compared to a typically pretreated substrate.
some speculations can be put forward albeit with caution.
Using pretreated corn stover containing 28% lignin (tradi- Pulp and paper production
tional dilute-acid pretreatment with lignin removal
eYciency of 10–20% theoretical), about 80% hydrolysis Corn stover pulp, selected properties of which are listed in
was obtained by Kadam et al. [27] in 7 days with 12 FPU/g Table 2, was blended with other pulps to test it as a partial
cellulose. In comparison, a similar hydrolysis performance substitute for hardwood pulp. Unbleached pine and South-
is possible with the PureVision substrate containing 4% lig- ern hardwood pulps were utilized in conjunction with corn
nin (lignin removal eYciency of ca. 90% theoretical) at stover pulp. The corn stover pulp substitutes for various
5 FPU/g cellulose (Fig. 2). Although the predicted enzyme amounts of the hardwood pulp in the blends, with pine pulp
loading based on Fig. 4 is >12 FPU/g cellulose, these remaining at 50%. The standard composition is equal parts
observations at least qualitatively agree with Fig. 4. Also, of pine and hardwood, i.e., 50% each for the top ply of a
the former substrate was washed whereas the latter was linerboard, which is a target application.
used “as is.” Hence, an unwashed substrate may need an Corn stover pulp substituted hardwood pulp at 10 and
enzyme loading closer to that predicted. 30% levels, the hardwood pulp portion was reduced accord-
ingly with softwood pulp staying constant at 50%. TAPPI
Advantages of PureVision substrate standard handsheets of the pure corn stover pulp and pulp
blends were made for physical testing. Figure 6 shows that
As the solids substrate from PureVision’s fractionation pro- corn stover pulp may be blended at 10–20% level without
cess is very high in cellulose (85–90%), 10% solids release too much degradation of pulp properties. Corn stover pulp
enough sugars to yield about 5% ethanol during SSF/HHF reduced the strength of the sheet as the proportion was
(SSF: simultaneous sacchariWcation and fermentation; increased. Bulk, porosity and Taber stiVness showed slight
HHF: hybrid hydrolysis and fermentation), which is consid- increase showing that the substitution of corn pulp pro-
ered as a minimum ethanol concentration in terms of rea- duced open structure sheet than the standard softwood/
sonable distillation energy. Obviously, increasing solids hardwood blend. Caliper and porosity increased indicating
loadings achieved through a fed-batch mode will yield that the corn pulp produced a more bulky, open sheet than
higher sugar and hence, higher ethanol concentration the standard hardwood. A 10% substitution of hardwood
thereby reducing distillation energy. It should be possible pulp by the corn stover pulp does not show much change in
to achieve Wnal eVective solids concentration of 25% in a most of the properties investigated.
fed-batch mode. Figure 5 illustrates the beneWt of high glu-
can content of pretreated solids on predicted theoretical eth- Environmental burdens of pulping
anol concentration in wt% (taking into account CO2 loss)
assuming SSF/HHF with 25% solids (Wnal eVective con- Environmental burdens of kraft pulping process are signiW-
centration). It should be emphasized that these are merely cant (Table 3). Extremely malodorous emissions of reduced
theoretical (based on 0.51 g ethanol/g sugar) and not actual sulfur compounds—measured as total reduced sulfur (TRS)
16
Theoretical ethanol concentration, wt%

14
Table 2 Selected properties of corn stover pulp

12 Parameters Values

10 Fiber properties
8 Length (mm) 0.77
Width (
m) 38.0
6
Coarseness (mg/m) 0.11
4
Curl index (%) 23.2
2
Typical pretreated PureVision Weighted distribution of Wbers (%)
substrate substrate
0 0–0.2 mm 15.4
50 60 70 80 90 100
0.2–0.5 mm 27.0
Glucan content of pretreated solids, wt%
0.5–1.0 mm 34.2
Fig. 5 Predicted eVect of glucan content of pretreated solids on theo- 1.0–3.0 mm 18.9
retical ethanol concentration (taking into account the loss of CO2 3.0–7.0 mm 4.5
mass)—25% solids during SSF/HHF

123
336 J Ind Microbiol Biotechnol (2008) 35:331–341

10 100 Furthermore, separation of hemicellulose before pulping

is also beneWcial to downstream operations. Removal of 4-
8 2
80 O-methyl--D-glucuronic acid groups, which exist in the
Tear index (mN·m /g)
xylan structure as part of the arabinoglucuronoxylan in
6 Breaking length (km)
60 softwood and glucuronoxylan in hardwood, occurs during
Tensile index (Nm/g) alkaline pulping of wood. This leads to the formation of 2-
4 40 O-(4-deoxy-
-L-threo-hex-4-enopyranosyluronic acid)-D-
2
Burst index (kPa·m /g) xylitol (hexenuronic acid-D-xylitol); this structure is com-
2 20
monly referred to as hexenuronic acid and abbreviated as
3
Bulk density (g/cm )
HexA [7, 32, 53]. HexA reacts with numerous bleaching
0 0
agents, such as ozone, chlorine dioxide or peracids [3, 6,
0 20 40 60 80 100 55], and thus consumes these expensive and environmen-
Corn stover pulp in blended product, wt% tally malevolent chemicals. HexA is responsible for as
Fig. 6 Properties of blends made with pulps from the two-stage frac- much as half of the bleach demand for hardwood kraft
tionation process pulps. Ragauskas and Perine [43] recommend a prehydroly-
sis step of hot-water extraction prior to pulping since it
Table 3 Emissions from pulp and paper manufacturing improves deligniWcation during pulp bleaching. They claim
Parameter Maximum recommended value that this is one of the two fundamental approaches that
would transform pulp mills into forest bioreWneries, the
Air emissions
other being the use of black liquor as a source of chemicals,
Particulate matter 100 mg/Nm3 (for recovery furnace) followed by gasiWcation of liquor residuals. Hence, prehy-
Hydrogen sulWde 15 mg/Nm3 (for lime kilns) drolysis is expected to be equally beneWcial as practiced in
Total sulfur emitted the PureVision process, and the pulp thus produced will
SulWte mills 1.5 kg/t ADP require less bleaching chemicals. Also, this fractionation
Kraft and other mills 1.0 kg/t ADP approach recovers lignin as an intermediate that can be
Nitrogen oxides 2.0 kg/t ADP used in making chemicals such as adhesives.
Liquid eZuents
COD 15 kg/t ADP Short Wber pulps
AOX 2 kg/t ADP
Total P 0.05 kg/t ADP Although it is considered necessary in North America to
Total N 0.4 kg/t ADP use long Wbers, i.e., with a Wber length of 2–4 mm, short
World-Bank [57] Wber pulps are routinely used elsewhere in the world. In the
ADP air dried pulp, COD chemical oxygen demand, AOX adsorbable manufacture of printing and writing papers long Wbers have
organic halides a tendency to Xocculate and hence, it is necessary to shorten
the Wbers by beating. With short Wber pulps only a low-
including hydrogen sulWde, methyl mercaptan, dimethyl intensity beating is needed to attain the necessary sheet
sulWde, and dimethyl disulWde—occur during kraft pulping. strength. Concomitantly, short Wbers contribute to smoother
Typical emission rates, in kg/mt of air dried pulp (ADP), surface and higher opacity. Although some printing papers
are: TRS 0.3–3, particulate matter 75–150, sulfur oxides are manufactured from 100% hardwood pulp, about 20% of
0.5–30, nitrogen oxides 1–3, and volatile organic com- long Wber pulp is necessary to enhance the wet strength of
pounds (VOCs) 15 (from black liquor oxidation). Liquid the paper web when passing through the press section of the
eZuents release chlorinated organic compounds (which paper machine, especially the faster machines. For low-
may include dioxins, furans, and others, collectively grade wrapping kraft papers, the furnish can be made up of
referred to as adsorbable organic halides or AOX) at a rate mostly short Wber pulp, with a small amount of long Wber
of 0–4 kg/mt of ADP. Potential environmental pollution pulps. Corrugating board is made of one or several layers of
abatement for the kraft process is expected to be primarily corrugating medium covered on one or both sides by kraft
capture and control because there are few if any methods liner or chip-board. However, the best corrugating medium
available for eliminating pollution at the source. SulWte product is manufactured from a neutral sulWte semichemi-
mills emit 50% higher TRS than kraft mills. Green pulp cal short Wber pulp. Thus, for the most important grades of
production via the PureVision process will eliminate all of papers, short Wber pulps can be used.
the TRS emissions, a major environmental beneWt. Moneti- Average Wber lengths for various hardwoods range from
zation of these advantages can oVer a signiWcant edge to the 0.65 to 1.85 mm [21, 42]. Corn stover Wbers from the Pure-
pulping process proposed here. Vision process are then similar in length to eucalyptus

123
J Ind Microbiol Biotechnol (2008) 35:331–341 337

Wbers which have Wber length of 0.65 mm. Eucalyptus is pulp exhibiting both a very low xylan content of about
widely used in Brazil and S. Africa. Eucalyptus Wber length ·2.6 wt% and a very low mannan content of about
is small; it is actually the shortest hardwood Wber available. ·1.5 wt%. In their process, these low xylan and mannan
Hence, the corn stover Wbers may be used in selected paper levels could not be achieved with treatments of only an
applications. alkali extraction stage, only a xylanase treatment stage, or
only two stages of a xylanase treatment stage and an alkali
Dissolving pulp extraction stage. In an alternative process for producing dis-
solving pulps, most of the hemicellulose is removed in a
Cellulose is a polymer with a molecular formula prehydrolysis step followed by sulfate pulping to high cel-
(C6H10O5)n. Dissolving pulp is a high-purity cellulose pulp lulose pulps containing small amounts of hemicellulose.
with low levels of hemicellulose and lignin; typical dissolv- This pulp is then extracted with NaOH (50–150 g/L) at a
ing pulp speciWcations are tabulated in Table 4. These pulps temperature of 10–60°C and washed to remove any residual
are treated to form soluble reactive carbohydrate chains that hemicellulose from the pulp. The pulp exhibits reduced
are then extruded into Wbers or Wlms. This relatively pure swelling tendencies, improved mercerizing properties, and
cellulose is well suited as a raw material for various cellu- reduced energy needs in the process [29].
lose-based products such as staple Wbers, Wlms, cellulose Cellulose pulp from the PureVision fractionation process
ethers (carboxymethyl, ethyl, methyl), cellulose esters (ace- is well suited for making dissolving pulp because it is
tates, propionates, butyrates) and regenerated celluloses already high enough in cellulose and low in xylan and lig-
such as viscose and microcrystalline cellulose. nin, as is shown in Table 4. A simple alkaline treatment
would be suYcient to produce dissolving pulp.
Dissolving pulp production
Fiber separation
Dissolving pulp was produced until World War II exclu-
sively from puriWed cotton linters. For lower purity applica- The low replacement levels of corn stover pulp for hard-
tions such as for viscose, it was produced via the acid wood pulp stems from its short Wber length, i.e., 0.77 mm.
sulWte process using somewhat higher temperature and However, based on Wber length distribution, about 15–20%
acidity in concert with longer cooking times to remove of the Wber can be used directly for paper applications. Sub-
more of the hemicellulose. The customary kraft process is stitution of hardwood pulp with these longer Wbers is
not proWcient in hemicelluloses removal, and residual expected at much higher levels, i.e., up to 50% or total
pentosans interfere with the chemical conversion of cellu- replacement of hardwood pulp in the blends. Using the
lose to viscose, cellulose ethers or acetates. The hemicellu- remainder of shorter stover Wbers thus separated for cellu-
loses can only be eVectively solubilized when exposing lose derivatives is a prudent approach.
wood chips to acid hydrolysis prior to alkaline pulping. Fiber separation segregates pulp Wbers into diVerent
Currently, dissolving pulp is produced via the acid sulWte fractions based on their speciWc physical properties such as
and prehydrolysis kraft processes. The acid sulWte process Wber length or Xexibility. The Wbers with particular attri-
is the most common, but suVers from a disadvantage of butes can then be gainfully directed to the most appropriate
pulps with a broad MW distribution of cellulose. process and product. For example, separated long Wbers can
A method for making dissolving pulp from cellulosic be used in a high-value reinforcing pulp, the shorter Wbers
Wber is disclosed in the patent literature [19, 24]. The Wber being sent to lower-end applications [16]. The board indus-
is treated in a three-stage sequence: (1) alkali extraction try utilizes Wber separation to selectively improve the facets
stage, (2) xylanase treatment stage, and (3) additional alkali of various layers [13, 28]. Also, Wber fractionation and sub-
extraction stage. Placing the xylanase treatment stage sequent processing of recycled Wbers production have been
between two alkali extraction stages yields a dissolving demonstrated to improve strength and brightness [37]. In
industrial pressure screens such as the Messmer Fiber Clas-
siWer, the Wber is circulated at a high uniform velocity
Table 4 Corn stover pulp versus typical dissolving pulp speciWcations across the screen thereby compelling the Wbers to orient
(dry wt%)
themselves parallel to the wire mesh. Achieved Wber orien-
Component SpeciWcation Corn stover pulp tation is not disturbed because water Xows through the
-Cellulose ¸94 89.0
screen at a low enough velocity, and the design also pre-
Pentosans ·3.0 3.9
vents Wbers from jamming the screen.
Average Wber length is a fundamental pulp property that
Extractives ·0.2 2.9
has been shown, other aspects remaining the same, to relate
Ash ·0.5 2.4
to the strength properties of paper. It may be beneWcial to

123
338 J Ind Microbiol Biotechnol (2008) 35:331–341

separate long Wbers from the PureVision process and use The high abundance of lignin as a waste product in pulp
them in a paper application. The remaining shorter Wbers mills has attracted attention as a raw material alternative to
can be used for dissolving pulp. nonrenewable petroleum-derived chemicals in the produc-
For North America, it is probably necessary to separate tion of wood adhesives. Lignin substitution into phenol
out the long Wbers for paper applications. Such fraction- formaldehyde (PF) resins is generally limited to <20–30%
ation also removes Wnes from the furnish and relinquishes it because cure times increases with the amount of lignin.
to the short Wber stream. As mentioned above, fractionation Bagasse-derived lignin has been used as an adhesive for
is an essential part of producing multilayer paperboard. The particleboard manufacture [20]. Hume et al. [22, 23]
long Wber component is mostly responsible for strength, describe a polyvinyl alcohol and lignin sulfonate containing
and the short Wber components contribute to the smooth- adhesive possessing suYcient adhesion, wherein the lignin
ness and opacity of the sheet. In corrugated containers, the sulfonate to polyvinyl alcohol ratio can vary from 1 to 8.
short fraction is used as the corrugated medium, while the PF resins are an important adhesive employed in the pro-
stronger long fraction is used for the liner [4]. Thus, Wber duction of wood-based panels for exterior use, which
separation yields two Wber streams that are more valuable requires superior water resistance. They are produced via
individually than the original feed stream by itself. This the reaction between phenol and formaldehyde; this reac-
approach is applicable to Wbers from the PureVision pro- tion is catalyzed by alkali and yields a thermosetting poly-
cess. mer called a resole. Other phenolic compounds such as
resorcinol can also react with formaldehyde in a similar
Lignin way to provide polymers with similar structure and proper-
ties. However, phenol is a toxic substance of petrochemical
MW distribution of hydrolyzed lignin from second stage origin, and the recent jump in crude oil prices makes a case
liquor is shown in Fig. 7. Using an average phenylpropane for developing alternative resin feedstock. Lignin, a benign
unit MW of 180–200, the peak at around 1,000 can be inter- renewable resource, can be used as a raw material to pro-
preted as a pentamer. A majority of the lignin has a MW of duce environmentally friendly industrial products such as
<10,000; this is in contrast with an average MW of 50,000 wood adhesives and sealers and also reduce demand on fos-
seen for kraft lignin. A lower MW signiWes more reactivity sil fuels, a nonrenewable resource. Also, phenol prices have
thereby facilitating further conversion to value-added prod- doubled in recent years to about $0.8 per lb due to rising
ucts. crude prices.
Lignin can be methylolated by reaction for 5 h at 55°C
Lignin-based adhesives with formaldehyde (formaldehyde/lignin ratio of 0.38 by
weight) in the presence of sodium hydroxide (NaOH/lignin
A variety of wood adhesive types are currently available for ratio of 0.2 by weight); this construct can be used in place
wood utilization, and there is a large market for wood adhe- of phenol [12]. Use of a methylolated softwood ammonium
sives. North American adhesive resin requirements for lignosulfonate as partial substitute of phenol in resol resins
wood composites were 1.8 million t [45]. Resin adhesives manufacture has been reported [2]. Sellers and coworkers
are a vital part of glued wood composites and constitute have done extensive work in successfully using lignins as
32% of the total manufacturing cost of the item being glued an extender in PF resin adhesives [40, 41, 44, 46–48, 56].
and marketed. Hence, lignin from the PureVision process can be similarly
utilized to make PF resins.
1,200
Xylose utilization
1,000

In the baseline scenario, the xylose-rich Wrst stage liquor is

Normalized intensity

800
expected to be fermented to ethanol. This can be achieved
600 in separate pentose fermentation or in SSCF (simultaneous
sacchariWcation and cofermentation) mode. However, it can
400
also be used to make specialty products; it should be
200
emphasized that these will be small markets and a single
large reWnery may saturate the market. One such possibility
0 is xylitol. The Wrst stage liquor from PureVision bioreWnery
100 1,000 10,000 100,000 1,000,000
is expected to contain about 50 g/L xylose and a concentra-
Apparent molecular weight
tion of 150 g/L is usually needed for industrial xylitol fer-
Fig. 7 Molecular weight distribution of hydrolyzed lignin mentation. China is a major producer of xylitol, and xylose

123
J Ind Microbiol Biotechnol (2008) 35:331–341 339

is typically concentrated by triple eVect evaporation prior to estimated lignin market is 1.8 million mt/year, and a few
conversion to xylitol. Alternatively, reverse osmosis (RO) PureVision bioreWneries can be built without saturating
can be used to concentrate the feed 15% xylose. RO is con- these markets. Furthermore, other even larger volume
sidered preferable over evaporation for concentrating sug- applications are possible in the future such as an antioxi-
ars as it prevents carmelization and saves energy. Data on dant additive for hot-mix asphalt (HMA) [18, 34, 60].
concentrating sugars such as glucose, maltose, lactose, and HMA is used to build new roads, but its most prevalent
xylose using RO are available in the literature [33, 59]. application is in patches and repairs. About 500 million t of
NanoWltration has also been used for concentrating xylose HMA are produced every year [60], and future bioreWneries
liquor in the production of xylitol [36]. Studies are under- can exploit these yet untapped markets.
way to ferment the Wrst stage liquor using Debaryomyces The value of xylitol market is currently $340 million
hansenii, which has been shown to be an eYcient xylitol with applications in mouthwashes and toothpastes, chewing
fermenter [10, 11, 52]. gums as well as in foods for special dietary uses. With xyli-
The Wrst stage liquor contains pentoses in oligomeric tol prices currently ranging from $4–5 per kg, this repre-
form. Xylo-oligosaccharides (XOs), also called xylo-oligo- sents <100,000 t/year. Global xylitol consumption was
mers, are low-digestible sugars and utilized by most BiWdo- 43,000 t in 2005, the U.S. and Western Europe accounting
bacterium species. They are sold as functional food for 30 and 37% of the total xylitol consumption, respec-
additive, mostly in Asian markets, and selectively promote tively [9]. Hence, xylitol is a niche market from a bioreWn-
the proliferation of biWdobacteria in human intestine. ery standpoint. The market price of prebiotic XOs is about
Hence, XOs can be classiWed as prebiotics. Prebiotics are $15 per kg [38]; however, this is deWnitely a niche applica-
deWned as “nondigestible food ingredients that are selec- tion.
tively metabolized by colonic bacteria that have the capac-
ity to improve health.” They are distinct from most dietary
Wbers like pectin, celluloses, xylan, which are not selec- Conclusion
tively metabolized in the gut. The desired degree of poly-
merization for XOs as prebiotics is 2–7, which can be A new bioreWning process is presented that—in the spirit of
achieved via controlled enzymatic hydrolysis by an endo- a true bioreWnery—converts corn stover and other biomass
xylanase. feedstocks into value-added products such as fuel ethanol,
dissolving pulp, and lignin for resin production. The contin-
Markets and values uous biomass fractionation process yields a liquid stream
rich in hemicellulosic sugars, a lignin-rich liquid stream
Although a detailed economic analysis is required to assess and a solid cellulose stream. Enzymatic hydrolysis of this
process feasibility, which is a focus of another publication relatively pure cellulose stream requires signiWcantly lower
in print, it is instructive to present the value-added nature of enzyme loadings because of minimal enzyme deactivation
products that are possible. As mentioned above, the U.S. is from nonspeciWc binding to lignin. A correlation was
planning to replace its gasoline consumption by 20% over shown to exist between lignin removal eYciency and enzy-
the next 10 years with alternative fuels [8]. Given the cur- matic digestibility. The cellulose produced was demon-
rent fuel ethanol market in the U.S. of ca. 180 billion gal/ strated to be a suitable replacement for hardwood pulp,
year and the limited capacity of corn-based ethanol, the especially in target application of the top ply of a liner-
market for cellulosic ethanol is predicted to be substantial. board. Also, the relatively pure nature of the cellulose ren-
Fuel ethanol prices generally follow gasoline prices. Hence, ders it suitable as raw material for making dissolving pulp.
at current relatively high crude oil prices, ethanol produc- This pulping approach has signiWcantly smaller environ-
tion via the bioreWning process presented here should even- mental footprint compared to the industry-standard kraft
tually be economically feasible. process because no sulfur- or chlorine-containing com-
World production of dissolving pulp in 2003 was 3.7 pounds are used. Along with use as cement and feed bind-
million t [49]. Although the market is much smaller than ers, low-MW lignin can potentially be used in wood
kraft pulp, dissolving pulp commands $4,000 per t versus adhesive production. As a baseline application, the hemi-
about $500 per t for the latter. Wood adhesive resin con- cellulosic sugars captured in the hydrolyzate liquor can be
sumption was 2 million t in 2001 with market prices of used to produce ethanol, but potential utilization of xylose
approximately $0.46 per kg. Phenolic resin market was for xylitol fermentation is also feasible. Although data spe-
about 500,000 t/year, which is a small market considering ciWc for corn stover are presented, the proposed bioreWnery
the scale of future bioreWneries. However, other potential scenario is generically applicable to other biomass feed-
larger-volume lignin markets are as concrete binder ($275 stocks. Successful commercialization of this technology
per mt) and as feed binder ($385–465 per mt). The current would result in a sustainable green process with positive

123
340 J Ind Microbiol Biotechnol (2008) 35:331–341

environmental impacts such as reduction in emissions of 18. GuVey FD, Robertson RE, Hettenhaus JR (2005) Lignin as an anti-
greenhouse gases and criteria pollutants. oxidant in petroleum asphalt. In: The 27th symposium on biotech-
nology for fuels and chemicals, Denver, CO
19. Heitmann JA, Joyce TW, Jackson LS (2001) Method for making
Acknowledgments We would like to thank the U.S. Department of dissolving pulp from paper products containing hardwood. United
Energy (DOE) for Wnancial support, David Templeton and Dr. David States patent 6,254,722
Johnson of NREL (National Renewable Energy Laboratory, Golden, 20. Hemingway RW, Conner AH (1989) Adhesives from renewable
CO, USA) for help in corn stover compositional and lignin MW distri- resources. Oxford University Press, Oxford
bution analyses, and Prof. Gopal Krishnagopalan of Auburn University 21. Horn RA (1978) Morphology of pulp Wber from hardwoods and
for pulp testing. inXuence on paper strength Research Paper FPL 312. Forest Prod-
ucts Laboratory
22. Hume RM, LaBrash RA, Vander Giessen MJ (1986a) Bonding
method employing an adhesive which contains in an aqueous base
References an eVective amount of a polyvinyl alcohol and a lignin sulfonate.
United States patent 4,608,111
1. Aden A, Ruth M, Ibsen K, Jechura J, Neeves K, Sheehan J, Wal- 23. Hume RM, LaBrash RA, Vander Giessen MJ (1986b) Polyvinyl
lace R, Montague L, Slayton A, Lukas J (2002) Lignocellulosic alcohol and lignin sulfonate containing adhesive. United States
biomass to ethanol process design and economics utilizing co-cur- patent 4,564,649
rent dilute acid prehydrolysis and enzymatic hydrolysis for corn 24. Jackson LS, Joyce TW, Heitmann JA (2001) Method for making
stover NREL/TP-510-32438 dissolving pulp from paper products containing hardwood Wbers.
2. Alonso MV, Oliet M, Rodríguez F, Astarloa G, Echeverría JM United States patent 6,254,722
(2004) Use of a methylolated softwood ammonium lignosulfonate 25. Kadam KL (2005) Bioethanol production technologies: econom-
as partial substitute of phenol in resol resins manufacture. J Appl ics, environmental impact, and policy issues. In: Lal B, Reddy
Polym Sci 94(2):643–650 MRVP (eds) Wealth from waste: trends and technologies. TERI
3. Bergnor-Gidnert E, Tomani PE, Dahlman O (1998) InXuence on Press, New Delhi, pp 445–470
pulp quality of conditions during the removal of hexenuronic ac- 26. Kadam KL, Camobreco VJ, Glazebrook BE, Forrest LH, Jacobson
ids. Nordic Pulp Paper Res J 13(4):310–316 WA, Simeroth DC, Blackburn WJ, Nehoda KC (1999) Environ-
4. Bliss T (1987) Pulp fractionation can beneWt multilayer paper- mental life cycle implications of fuel oxygenates production from
board operations. Pulp Paper 61(2):104–107 california biomass, NREL/TP-580-25688 NREL/TP-580-25688.
5. Brown L, Torget R (1996) Enzymatic sacchariWcation of lignocel- National Renewable Energy Laboratory
lulosic biomass, LAP-009 (laboratory analytical procedure). Na- 27. Kadam KL, Rydholm EC, McMillan JD (2004) Development and
tional Renewable Energy Laboratory, Golden validation of a kinetic model for enzymatic sacchariWcation of lig-
6. Buchert J, Teleman A, Harjunpää V, Viikari L, Vuorinen T (1995) nocellulosic biomass. Biotechnol Prog 20:698–705
EVect of cooking and bleaching on the structure of xylan in con- 28. Kohrs M (1992) The applications for fractionation technology. Pa-
ventional pine kraft pulp. Tappi J 78(11):125–130 per Technol 33(3):10–12
7. Buchert J, Teleman A, HV., T. M., L. Viikari, and T. Vuorinen 29. Lindberg EAS, Ljungqvist K-J (1982) Method in the production of
(1994) Presented at the 3rd European workshop on Lignocellulo- dissolving pulp. United States patent 4,361,463
sics and pulp (EWLP). In: Proceedings of the 3rd European work- 30. Lynd L, Cushman J, Nichols R, Wyman C (1991) Fuel ethanol
shop on Lignocellulosics and pulp (EWLP), Stockholm, Sweden from cellulosic biomass. Science 251:1318–1323
8. Bush GW (2007) Twenty in ten: strengthening energy security and 31. Mandels M, Hontz L, Nystrom J (1974) Enzymatic hydrolysis of
addressing climate change. http://www.whitehouse.gov/infocus/ waste cellulose. Biotechnol Bioeng 16:1471–1493
energy/ 32. Maréchal A (1993) Acid extraction of the alkaline wood pulps
9. Business-Communications-Company (2000) Xylitol consumption (kraft or soda/AQ) before or during bleaching reason and opportu-
by geographic area through 2005. The Global Market for Polyols nity. J Wood Chem Technol 13(2):261–281
by Product, Report C230, chap 4, Table 89 33. Matsuura T, Sourirajan S (1971) Reverse osmosis separation of
10. Carvalheiro F, Duarte L, Lopes S, Parajo J, Pereira H, Girio F some organic solutes in aqueous solution using porous cellulose
(2006) Supplementation requirements of brewery’s spent grain acetate membranes. Ind Eng Chem Process Des Develop
hydrolysate for biomass and xylitol production by Debaryomyces 10(1):102–108
hansenii CCMI 941. J Ind Microbiol Biotechnol 33(8):646–654 34. McCready N, Williams C (2007) Utilization of agriculturally de-
11. Carvalheiro F, Duarte L, Medeiros R, Girio F (2004) Optimization rived lignin as an antioxidant in asphalt binder. In: 4th world
of Brewery’s spent grain dilute-acid hydrolysis for the production congress on industrial biotechnology and bioprocessing, Orlan-
of pentose-rich culture media. Appl Biochem Biotechnol 113– do, FL
116:1059–1072 35. Moniruzzaman M, Dale BE, Hespell RB, Bothast RJ (1997) Enzy-
12. Çetin NS, Özmen N (2003) Studies on lignin-based adhesives for matic hydrolysis of high moisture corn Wber pretreated by AFEX
particleboard panels. Turk J Agric For 27:183–189 and recovery and recycle of the enzyme complex. Appl Biochem
13. Cox MT (1989) Presented at the TAPPI pulping conference Biotechnol 67:113–126
14. Esteghlalian A, Srivastava V, Gilkes N, Gregg D, Saddler J (2000) 36. Murthy GS, Sridhar S, Shyam Sunder M, Shankaraiah B, Rama-
An overview of factors inXuencing the enzymatic hydrolysis of krishna M (2005) Concentration of xylose reaction liquor by nano-
lignocellulosic feedstocks. ACS Symp Ser 769:100–111 Wltration for the production of xylitol sugar alcohol. Sep Purif
15. Ghose TK, Ghosh P (1978) Bioconversion of cellulosic sub- Technol 44(3):205–211
stances. J Appl Chem Biotechnol 28:309–320 37. Musselmann W, Menges W (1982) Concept and function of a
16. Gooding RW, Olson JA (2001) Fractionation in a Bauer–McNett wastepaper Wbre fractionation system and practical operating
classiWer. J Pulp Paper Sci 27(12):423–428 experience. Woch Papierfab 11/12:368–379
17. Gregg D, Boussaid A, Saddler J (1998) Techno-economic evalua- 38. Nakakuki T (2002) Present status and future of functional oligo-
tions of a generic wood-to-ethanol process: eVect of increased cel- saccharide development in Japan. Pure Appl Chem 74(7):1245–
lulose yields and enzyme recycle. Bioresour Technol 63(1):7–12 1251

123
J Ind Microbiol Biotechnol (2008) 35:331–341 341

39. NCGA (2005) Corn-to-ethanol capacity estimates for 10% fuel 51. Sluiter A, Hames B, Ruiz R, Scarlata C, Sluiter J, Templeton D
ethanol inclusion nationwide by 2015. In: National corn growers (2006) Determination of sugars, byproducts, and degradation
association, ChesterWeld, MO. http://www.ncga.com/ products in liquid fraction process samples. Laboratory Analytical
40. Oh Y, Sellers T, Kim MG, Strickland RC (1994) Evaluation of Procedure, National Renewable Energy Laboratory, Golden
phenol-formaldehyde OSB resins modiWed with lignin residues 52. Tavares J, Duarte L, Amaral-Collaco M, Girio F (2000) The inXu-
from acid-hydrolyzed waste newsprint. For Prod J 44(2):25–29 ence of hexoses addition on the fermentation of d-xylose in De-
41. Olivares M, Aceituno H, Neiman G, Rivera E, Sellers T (1995) baryomyces hansenii under continuous cultivation. Enzyme
Lignin-modiWed phenolic adhesives for bonding radiata pine ply- Microb Technol 26(9–10):743–747
wood. For Prod J 45(1):63–67 53. Teleman A, Harjunpää V, Tenkanen M, Buchert J, Hausalo T,
42. Pawlak JJ, Keller DS (2005) The compressive response of a strat- Drakenberg T, Vuorinen T (1995) Characterisation of 4-deoxy-
-
iWed Wbrous structure. Mech Mater 37:1132–1142 L-threo-hex-4-enopyranosyluronic acid attached to xylan in pine
43. Ragauskas A, Perine L (2007) Presented at the fourth world con- kraft pulp and pulping liquor by 1H and 13C NMR spectroscopy.
gress on industrial biotechnology and bioprocessing. In: 4th world Carbohyd Res 272:55
congress on industrial biotechnology and bioprocessing, Orlando, 54. Van Wyk JPH (2001) Biotechnology and the utilization of bio-
FL waste as a resource for bioproduct development. Trends Biotech-
44. Sellers T (1998) Adhesive industry matching wood composite nol 19:172–177
needs. PanelWorld 39(5):12–15 55. Vuorinen T, Teleman A, Fagerström P, Buchert J, Tenkanen M
45. Sellers T (2000) Presented at the XXI IUFRO world congress. In: (1996) Presented at the international pulp bleaching conference.
XXI IUFRO world congress, Kuala Lumpur, Malaysia In: International pulp bleaching conference, Washington, DC
46. Sellers T, Kim MG, Miller GD, Haupt RA, Strickland RC (1995) 56. Wooten AL, Sellers T, Tahir PM (1988) Reaction of formaldehyde
Comparison of strandboards made with phenol-formaldehyde res- with lignin. For Prod J 38(6):45–46
in and resins modiWed with TVA acid-hydrolysis lignin. For Prod 57. World-Bank (1996) Pollution prevention and abatement: pulp and
J 44(4):63–68 paper mills. World Bank, Environment Department, Technical
47. Sellers T, Lora JH, Okuma M (1994) Organosolv-lignin-modiWed Background Document
phenolics as strandboard binder I. Organosolv lignin and modiWed 58. Yang B, Wyman CE (2004) EVect of xylan and lignin removal by
phenolic resin. Mokuzai Gakkaishi 40(10):1073–1078 batch and Xowthrough pretreatment on the enzymatic digestibility
48. Sellers T, Lora JH, Okuma M (1994) Organosolv-lignin-modiWed of corn stover cellulose. Biotechnol Bioeng 86(1):88–98
phenolics as strandboard binder II. Strandboard manufacture and 59. Yurong L, Bangxiao C, Yiming C, Kangmin L (1987) Concentrat-
properties. Mokuzai Gakkaishi 40(10):1078–1084 ing xylose solution by reverse osmosis with cellulose acetate dry
49. Sixta H (2006) Chemical pulping. In: Sixta H (ed) Handbook of membrane. Desalination 62:193–201
pulp. Wiley-VCH Verlag GmbH & Co., Weinheim 60. Zeman N (2007) ISU evaluates lignin as additive in asphalt manu-
50. Sluiter A (2005) Determination of structural carbohydrates and facturing. Biomass Magazine. http://www.biomassmagazine.com
lignin in biomass. Laboratory Analytical Procedure, National
Renewable Energy Laboratory, Golden

123
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.