FRACTIONATION AND CHARACTERIZATION OF SOLUBILIZED BIOPOLYMERS
FROM ALKALINE PULPING LIQUORS
By
Ryan Joseph Stoklosa

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Chemical Engineering – Doctor of Philosophy
2014

ABSTRACT
FRACTIONATION AND CHARACTERIZATION OF SOLUBILIZED BIOPOLYMERS
FROM ALKALINE PULPING LIQUORS
By
Ryan Joseph Stoklosa
The conversion of lignocellulosic biomass to renewable liquid fuels or chemicals offers
one approach to decrease dependence on fossil fuels. Attempts at commercializing the
biochemical production of biofuels from lignocellulose has been hindered by the costs associated
with pretreatment, enzyme production, and feedstock transportation. The infrastructure
established by the forest products industry offers a way to decrease initial capital costs, and to
generate new products and revenue from an industry showing declined profits. This work
addresses two general themes of applying alkaline pulping chemistries to: (1) extract, separate,
and recover chemically altered non-cellulosic biopolymers from woody biomass and (2) evaluate
the impact that the removal of non-cellulosic biopolymers have on the yields of monomeric
sugars by enzymatic hydrolysis. These general themes were addressed through three studies.
First, bench scale alkali extractions performed on a variety of hardwoods indicated that
hemicellulose and lignin extractability was dependent on biomass species, which in turn were
related to initial composition and properties of the cell walls. Ethanol precipitation could recover
the hemicellulose fraction and following precipitate bleaching the recovered material yielded 6580% hemicellulose, depending on biomass source. Molecular weight characterization of the
recovered hemicellulose by size-exclusion chromatography (SEC) indicated the presence of
lignin and carbohydrates eluting at the 3-12 kDa range, and the application of novel reducing end
quantification estimated the number average degree of polymerization (DPN) would reach an
asymptotic limit of 25 for switchgrass and 120 for silver birch. The high extraction yields of

hemicellulose identified conditions that could be utilized to remove hemicellulose prior to
chemical pulping or during alkaline pretreatments.
Lignin characterization was performed on fractions generated from a softwood kraft
black liquor to determine the properties of solubilized lignin that resulted in their phasepartitioning behavior during a novel acidification process using CO2. A bulk of the lignin
precipitated between a pH of 11.6 to 10.0, with fractions obtained at higher pH contaminated by
larger quantities of aliphatic extractives and polysaccharides. Lignin fractions were characterized
for functional groups by 1H and 13C NMR, molecular weight by gel-permeation chromatography
(GPC), and lignin monomer generation by analytical pyrolysis-GC/MS. Select structural and
chemical properties of the lignin fractions were correlated based on their linear dependencies to
relate phase-partitioning behavior with respect to pH.
Lastly, alkaline pulping trials were conducted in a pressurized reactor vessel to determine
hemicellulose and lignin dissolution during a soda pulping process and to understand how
pulping conditions may impact hydrolysis yields. Hemicellulose dissolution increased up to
170°C followed by a decrease in the polysaccharide content of the black liquor, an effect of
polysaccharide degradation and re-adsorption to the residual wood fibers. Lignin dissolution
increased throughout the pulping trials, and at the most severe condition (1 hour at 170°C) only
10% of the lignin remained in the woody biomass. The alkaline pulping trials also produced a
digestible substrate that generated over an 80% yield of monomeric glucose.

ACKNOWLEDGEMENTS
I would like to offer a word of thanks to my advisor Dr. David Hodge for all of his
guidance and support during my years completing my doctoral research at Michigan State
University. Also, a word of thanks goes to my Ph.D. committee for their advice and suggestions
towards the development of my research. To all my colleagues in Dr. Hodge’s lab, I want to
offer my gratitude for the work we have collaborated on over the years, and all the thoughtful
discussions that occurred when analyzing experimental results. Lastly, special thanks to former
lab manager Kristine Van Winkle for her advice and expertise by diagnosing problems and
troubleshooting the analytical equipment located in the lab.
Additionally, I would also like to offer a word of thanks to the people who graciously
donated feedstocks for this research: Curt Lindström at Smurfit Kappa AB (Piteå, Sweden),
Todd Smith at the Devereaux Sawmill (Pewamo, MI) and Dr. Raymond Miller (MSU) for
providing wood chips and Dr. Pascal Kamden (MSU) for use of his Wiley Mill.
Special thanks would also like to be given to Dr. Mark Thies and his graduate student
Julian Velez from Clemson University for the opportunity to collaborate on an industrial related
research project.

iv

TABLE OF CONTENTS
LIST OF TABLES .................................................................................................................vii
LIST OF FIGURES ...............................................................................................................viii
KEY TO ABBREVIATIONS ................................................................................................xii
CHAPTER 1. NON-CELLULOSIC CELL WALL BIOPOLYMERS AS A SOURCE FOR
CONVERSION TO FUELS AND CHEMICALS ................................................................1
1.1 Introduction ..............................................................................................1
1.2 Hemicellulose Composition in Woody Biomass Feedstocks ..................9
1.3 Existing Liquor Streams ..........................................................................14
1.3.1 Chemi-Mechanical Pulping Glucomannan Streams .................14
1.3.2 Polysaccharides from Viscose Pulping .....................................16
1.3.3 Sulfite Pulping: Lignin and Hemicellulose ...............................19
1.3.4 Lignin Recovery by Acidification ............................................21
1.4 Pre-Extractions of Hemicellulose Prior to Pulping ..................................23
1.4.1 Hot Water Pre-Extractions ........................................................23
1.4.2 Alkali Pre-Extractions ...............................................................25
1.4.3 Renewable Products from an IFBR ..........................................27
1.5 Further Investigating the Direct Fractionation of Black Liquor ..............29
CHAPTER 2. EXTRACTION, RECOVERY, AND CHARACTERIZATION OF
HARDWOODAND GRASS HEMICELLULOSES FOR INTEGRATION INTO
BIOREFINING PROCESSES ...............................................................................................32
2.1 Introduction ..............................................................................................32
2.2 Materials and Methods .............................................................................37
2.2.1 Materials and Biomass Composition ........................................37
2.2.2 NaOH Extraction and Analysis of Extracts ..............................38
2.2.3 Characterization of Biopolymers ..............................................39
2.2.4 Precipitate Bleaching ................................................................41
2.3 Results and Discussion ............................................................................41
2.3.1 Biomass Composition ...............................................................41
2.3.2 Extraction and Recovery ...........................................................42
2.3.3 Chemical and Physical Characterization ..................................50
2.4 Conclusions ..............................................................................................56
APPENDIX ....................................................................................................59
CHAPTER 3. CORRELATING LIGNIN STRUCTURAL FEATURES TO PHASE
PARTITIONING BEHAVIOR IN A NOVEAL AQUEOUS FRACTIONATION OF
SOFTWOOD KRAFT BLACK LIQUOR ............................................................................62
3.1 Introduction ..............................................................................................62
3.2 Materials and Methods .............................................................................65
3.2.1 Materials ...................................................................................65
v

3.2.2 Lignin Acetylation ....................................................................67
3.2.3 Proton and Quantitative 13C NMR Analysis .............................68
3.2.4 Gel Permeation Chromatography .............................................69
3.2.5 Thermogravimetric Analysis ....................................................69
3.2.6 Analytical Pyrolysis ..................................................................70
3.2.7 Data Correlation ........................................................................71
3.3 Results and Discussion ............................................................................71
3.3.1 Lignin Yields and Non-Lignin Contaminants...........................71
3.3.2 Structural Features of Recovered Lignin Fractions ..................74
3.3.3 Relative Lignin Molecular Weight by Gel Permeation
Chromatography (GPC) .....................................................................76
3.3.4 Thermogravimetric Analysis of Lignin Fractions.....................79
3.3.5 Analytical Pyrolysis ..................................................................81
3.3.6 Property Correlations ................................................................83
3.4 Conclusions ..............................................................................................85
APPENDIX ....................................................................................................87
CHAPTER 4. HEMICELLULOSE DISSOLUTION AND THE INFLUENCE OF LIGNIN
REMOVAL ON ENZYMATIC HYDROLYSIS FOR SODA PULPED HARDWOODS ..90
4.1 Introduction ..............................................................................................90
4.2 Materials and Methods .............................................................................95
4.2.1 Biomass Composition ...............................................................95
4.2.2 Soda Pulping Trials ...................................................................95
4.2.3 Analysis of Black Liquor Fractions ..........................................97
4.2.4 Particle Size Reduction for Enzymatic Hydrolysis ...................98
4.2.5 Enzymatic Hydrolysis ...............................................................99
4.3 Results and Discussion ............................................................................100
4.3.1 Biomass Composition and Yields After Pulping ......................100
4.3.2 Hemicellulose and Klason Lignin Solubilization ....................103
4.3.3 Enzymatic Hydrolysis of Residual Soda Pulped Hardwoods ...108
4.3.4 Initial Hydrolysis Rates.............................................................118
4.4 Conclusions .............................................................................................121
REFERENCES ......................................................................................................................123

vi

LIST OF TABLES

Table 1.1. Brief description of non-cellulosic biopolymer recovery techniques and their
mechanisms ............................................................................................................................7
Table 1.2. Summary of hemicellulose composition in select grass species...........................10
Table 2.1. Composition of biomass as a percentage of the dry weight; N/D = not detected, NQ =
not quantified .........................................................................................................................42
Table 3.1. Liquid-lignin fractions obtained by the unique pH-based fractionation process (Velez
& Thies, 2013) and used for this study ..................................................................................66
Table B.1. Aromatic compound identifications for pyro-GC/MS .........................................89
Table 4.1. Soda pulping trial conditions ................................................................................96

vii

LIST OF FIGURES

Figure 1.1. An example of the kraft pulping process with chemical recovery ......................5
Figure 1.2. Hypothetical example of hardwood components dissolved during alkaline pulping
followed by the recovery of hemicellulose and lignin by precipitation.................................7
Figure 1.3. Structural representation of a cellulose chain ......................................................9
Figure 1.4. Hardwood hemicellulose structure with a 4-O-MeGA side group and an α-L-Rhap1,2-linked-α-D-GalpA at the reducing end of the GX chain .................................................11
Figure 1.5. Softwood hemicellulose structure of GGM .........................................................13
Figure 1.6. Softwood hemicellulose structure of GAX .........................................................13
Figure 1.7. Potential process diagram for utilizing soda pulping (or alkali pre-extraction) to
generate biofuels ....................................................................................................................25
Figure 2.1. Total mass extractability for the five biomass sources ........................................43
Figure 2.2. Recoverability of the extracted polymers with a 2:1 (v/v) ethanol precipitation for the
five biomass sources ..............................................................................................................44
Figure 2.3. Hemicellulose extractability by alkali for the five biomass sources ...................45
Figure 2.4. Relationship between hemicellulose extracted to the total amount of alkali extracted
material ..................................................................................................................................47
Figure 2.5. Quantified hemicellulose content of the cell wall polymers recoverable by ethanol
precipitation ...........................................................................................................................48
Figure 2.6. Recovery of plant cell wall polymers from sugar maple by ethanol or pH
precipitation ...........................................................................................................................48
Figure 2.7. Recovery of plant cell wall polymers from aspen by ethanol or pH precipitation
................................................................................................................................................49
Figure 2.8. Recovery of plant cell wall polymers from switchgrass by ethanol or pH precipitation
................................................................................................................................................49
Figure 2.9. Size exclusion chromatography elution profiles for commercial birch glucuronoxylan
with elution times for dextran standards plotted for reference ..............................................51

viii

Figure 2.10. Size exclusion chromatography elution profiles for alkali-extracted, ethanolprecipitated biopolymers from sugar maple with increasing alkalinity and bleaching .........53
Figure 2.11. Apparent number average DP of recovered precipitates estimated using the BCA
reducing end assay .................................................................................................................55
Figure A.1. Estimated degree of polymerization distribution of unsonicated Sigma-Aldrich birch
xylan using Waters Ultrahydrogel 250+500 column set in series .........................................60
Figure A.2. Concentration of reducing ends in solution from Sigma-Aldrich beech and birch
xylan sonicated for up to 20 minutes .....................................................................................60
Figure A.3. UV absorbance at 280 nm of alkali-extracted, ethanol precipitated biopolymers for
aspen following elution by size-exclusion chromatography ..................................................61
Figure A.4. Correlation of quantified ligin-to-hemicellulose ratio to the ratio of the two SEC
peaks (UV peak height to RI peak height) proposed to represent aromatics and polysaccharides
................................................................................................................................................61
Figure 3.1. Schematic of the pH-based fractionation process for separation of liquid-lignin
fractions within narrow (~0.5) pH bands from kraft black liquor via acidification with CO2 at
elevated temperatures and pressures ......................................................................................66
Figure 3.2. Quantified neutral polysaccharide content of fractions F3 through F7 with estimated
weight average polysaccharide content .................................................................................72
Figure 3.3. Quantified aliphatic H to MeO ratio of fractions F1 through F7 as determined by 1H
NMR as a proxy for extractives-to-lignin ratio......................................................................73
Figure 3.4. Phenolic hydroxyl and aliphatic hydroxyl content of lignin fractions by 1H NMR
(mol OH/mol lignin monomer) ..............................................................................................75
Figure 3.5. Decrease in Cγ (mol Cγ/mol lignin monomer) with decreasing separation pH by 13C
NMR ......................................................................................................................................76
Figure 3.6. Estimated lignin molecular weights by GPC (acetylated lignin, THF mobile phase)
................................................................................................................................................77
Figure 3.7. GPC elution profile of nonacetylated lignin fractions in DMF + 0.05 M LiCl mobile
phase, referenced to polystyrene standards ............................................................................78
Figure 3.8. Lignin fraction mass loss by TGA.......................................................................79
Figure 3.9. First derivatives of mass loss curves for lignin fractions ....................................80

ix

Figure 3.10. Relative normalized abundance of aromatic monomers released by pyrolysis of
lignin ......................................................................................................................................81
Figure 3.11. Correlation map for lignin properties of fractions F3-F7 demonstrating strong
correlation between most properties; A-OH = Aliphatic OH, P-OH = Phenolic OH. ...........83
Figure B.1. Representative 1H NMR spectra of acetylated lignin fractions showing identification
of typical structures in lignin as well as aliphatic protons that might be found in softwood
extractives ..............................................................................................................................88
Figure B.2. Representative quantitative 13C NMR spectra of a typical lignin fraction identifying
the Cγ in a β-O-4 bond and a methoxyl group used to determine β-O-4 content...................88
Figure B.3. Methoxyl content for the seven lignin fractions recovered ................................89
Figure 4.1. H-factor relationship for chemical pulping .........................................................95
Figure 4.2. A general temperature profile for a H-factor cook at 1147 .................................97
Figure 4.3. Composition of the three hardwood feedstocks before and after soda pulping, and
after LHW pretreatment for hybrid poplar.............................................................................100
Figure 4.4. Pulp yield, and residual xylan and Klason lignin content as a function of H-factor.
................................................................................................................................................102
Figure 4.5. Solubilization of xylan and Klason lignin during birch soda pulping (H = 1147).
................................................................................................................................................104
Figure 4.6. Solubilization of xylan and Klason lignin during maple soda pulping (H = 1147) ..
................................................................................................................................................105
Figure 4.7. Solubilization of xylan and Klason lignin during poplar soda pulping (H = 1147).
................................................................................................................................................105
Figure 4.8. Glucose enzymatic hydrolysis yields for birch, maple, and poplar after the high Hfactor trial; enzyme loading 20 mg per g of glucan. ..............................................................109
Figure 4.9. Xylose enzymatic hydrolysis yields for birch, maple, and poplar after the high Hfactor trial; enzyme loading 20 mg per g of glucan. ..............................................................109
Figure 4.10. Glucose enzymatic hydrolysis yields for soda pulped poplar at an H-factor of 1147,
525, and 114; enzyme loading of 20 mg per g glucan ...........................................................112
Figure 4.11. Xylose enzymatic hydrolysis yields for soda pulped poplar at an H-factor of 1147,
525, and 114; enzyme loading of 20 mg per g glucan ...........................................................112

x

Figure 4.12. Glucose enzymatic hydrolysis yields for LHW poplar; enzyme loadings of 20 and 5
mg per g glucan. .....................................................................................................................113
Figure 4.13. Xylose enzymatic hydrolysis yields for LHW poplar; enzyme loadings of 20 and 5
mg per g glucan......................................................................................................................114
Figure 4.14. Glucose enzymatic hydrolysis for soda pulped birch (H = 1147) at enzyme loadings
of 20, 10, and 5 mg per g glucan............................................................................................115
Figure 4.15. Xylose enzymatic hydrolysis for soda pulped birch (H = 1147) at enzyme loadings
of 20, 10, and 5 mg per g glucan............................................................................................115
Figure 4.16. Disintegration hydrolysis yields versus knife-milled hydrolysis yields for glucose
................................................................................................................................................117
Figure 4.17. Disintegration hydrolysis yields versus knife-milled hydrolysis yields for xylose
................................................................................................................................................117
Figure 4.18. Initial glucose hydrolysis rates versus residual lignin content ..........................119
Figure 4.19. Initial xylose hydrolysis rates versus residual lignin content ............................119

xi

KEY TO ABBREVIATIONS

4-O-MeGA: 4-O-methyl glucuronic acid
AQ: anthraquinone
Araf: arabinofuranosyl
BCA: bicinchoninic acid
C1, C2, or C3: number of additional carbons on aromatic compounds
CTMP: chemi-thermomechanical pulping
𝐷𝑃! : number average degree of polymerization
G: guaiacol monomer
GAGX: glucuronoarabino(galacto)xylan
Galp: galactopyranosyl
GalpA: galactouronopyranosyl acid
GAX: glucuronoarabinoxylan
GGM: galactoglucomannan
Glcp: glucopyranosyl
GlcpA: glucuronopyranosyl acid
GM: glucomannan
GPC: gel permeation chromatography
GX: glucuronoxylan
H: hydrogen group
HMF: hydroxymethylfurfural
HPLC: high performance liquid chromatography

xii

IFBR: Integrated Forest Bio-Refinery
LCC: lignin-carbohydrate complex
LHW: liquid hot water
Manp: mannopyranosyl
MeO: methoxyl group
MW: molecular weight
PDADMAC: polydiallyldimethylammonium chloride
Ph: phenol monomer
PSAQ: polysulfide-anthraquinone
Py-GC/MS: Pyrolysis gas chromatography-mass spectrometry
Rhap: rhamnopyranosyl
RI: refractive index
S: dimethoxybenzene and dimethoxyphenol monomers
SEC: size exclusion chromatography
SSF: simultaneous saccharification and fermentation
SPORL: sulfite pretreatment to overcome the recalcitrance of lignocellulose
TMP: thermomechanical pulping
UV: ultraviolet
XOS: xyloolgosaccharides
Xylp: xylopyranosyl

xiii

CHAPTER 1. NON-CELLULOSIC CELL WALL BIOPOLYMERS AS A SOURCE FOR
CONVERSION TO FUELS AND CHEMICALS
1.1 Introduction
There exists an overall effort to reduce petroleum consumption predicated upon the rising
costs of the petroleum supply and increased consciousness of the societal and environmental
impacts that accompany petroleum utilization. The supply of petroleum on a global basis has
4

started to level off around 9.0x10 thousand barrels per day making the cost of a barrel of oil
increase to the point where alternative liquid transportation fuels derived from biomass
feedstocks can be competitive (B. Dale, 2008; International Energy Statistics - Petroleum
Production, 2013). Furthermore, the increase in not only liquid transportation fuel consumption,
but also general energy consumption (i.e. electricity, natural gas, coal, etc.) will only rise in the
coming decades as developing nations improve modes of transportation and quality of life for
their citizens (B. Dale, 2008; B. E. Dale & Ong, 2012). Converting biomass feedstocks to
renewable liquid fuels or value-added chemicals is one solution to offset petroleum consumption,
and to temper the release of greenhouse gases (GHG) that originate from outside the biosphere.
In the United States first-generation ethanol production from corn has been successfully
commercialized to decrease overall petroleum consumption by passenger vehicles. To
successfully displace further petroleum consumption the development of second-generation
biofuels and commodity chemicals from cellulosic sources (i.e. lignocellulosic biomass) is
currently under development.
Much of the United States already has dedicated land and infrastructure to utilize
lignocelluloic biomass as a feedstock for biological or catalytic conversion to renewable fuels
and chemcials. A recent combined USDA/DOE study assessed the biomass inventory available
1

from agricultural and forestry required to displace 30% of US petroleum consumption (Perlack
& Stokes, 2011) and estimated that 27% of the biomass inventory would come from woody
biomass. This greater utilization of woody biomass will push the planting of trees from dedicated
forestry areas to more traditional agricultural plots of land. The USDA Forestry Service has
indicated that hybrid poplars (Populus spp.) planted on agricultural land with a short rotation
schedule can produce yields ranging from 37 to 55 dry tons per acre with a 6-year rotation
schedule and 900 trees per acre; however, more general planting stocks can yield 5 to 12 dry tons
per acre per year under current management systems (White, 2010). Additionally, woody based
resources already contribute to a sizable portion of energy usage as reported in 2008 that 2.1 x
15

10

BTUs are consumed from woody based resources; this includes black liquor from pulping

processes (White, 2010). Woody biomass utilization is even more attractive within the United
States based on the Renewable Fuel Standard (RFS) stipulated in the Energy Independence and
Security Act of 2007 that mandates at least 16 billion gallons of cellulosic ethanol must be
produced by 2022. However, the updated RFS places substantial restrictions on the use of woody
biomass to reach these mandates including the requirement that only biomass that was grown on
land that was actively managed as wood plantations in 2007 is eligible for classification as use as
an “advanced biofuel” (EPA, 2013).
Regardless, the research into utilizing woody biomass as a source for renewable fuels and
chemicals should continue. Illustrating this potential in 2012 it was reported that the United
7

States produced 5.66 x 10 metric tons (air dried) of wood pulp (Pulp and Paper Capacities
Survery 2011-2016, 2012). This high capacity of pulp production shows a functioning industry
already developed in the United States with dedicated logistical transportation operations that
can be adapted to develop forest biorefineries. To this end the primary barriers holding forest
2

biorefinery development back include species potential, economically competitive and
sustainable feedstock supply, and appropriate production technologies (Mayfield, Foster, Smith,
Gan, & Fox, 2007). Moreover, the pulp and paper industry is the largest consumer of woody
biomass and has the existing infrastructure for both its collection and transportation, and the
capital-intensive chemical pulp mills for thermochemical fractionation and conversion. Based
upon this it is reasonable to expect that future incremental technology developments for the
chemical and thermochemical conversion of woody biomass can be developed within the
framework of existing chemical pulping technologies and facilities. Moreover, the breakthroughs
in process development associated with pulp mills to generate greater utilization of hemicellulose
and lignin fractions from woody biomass will be helpful to further offset petroleum consumption
on a larger scale.
One of the necessary steps for the production of biologically generated fuels or chemicals
is a restructuring of the pulp and paper industry. Currently, North American pulp mills are in a
period of decreased profitability and consolidation due to a number of economic factors such as
increased competition from newer pulp mills with higher production capacities located outside of
the United States. For the purposes of increasing profitability while also contributing to the
development of renewable fuels and chemicals, the adjustment of current pulp mills to Integrated
Forest Biorefineries (IFBR) is one way to achieve these goals. The IFBR concept allows for
greater utilization of hemicellulose and lignin fractions from woody biomass to develop a new
product portfolio of specialty derived fuels and chemicals from woody biomass to accompany
cellulosic pulp production (Ragauskas et al., 2006; Van Heiningen, 2006). Technological
developments for an IFBR can be quite analogous to current petrochemical refineries. The
existing global infrastructure for energy and the many synthetic products from petrochemical

3

feedstocks is a result of the development of technologies in the last century that allowed for
economical extraction and refining (i.e. fractionation and chemical upgrading) from the raw
materials of crude oil and natural gas. Similarly, biorefining can be considered as the
fractionation and conversion of plant biomass to a suite of products and energy that include, for
example, stationary power, heat, fuels, chemical intermediates, solvents, polymers, and modified
plant fibers such as chemical pulps and cellulose derivatives. In order to develop successful,
economic processes for the biorefining of lignocellulosic biomass, similar developments are
required for the extraction, fractionation, and catalysis of the raw material to fuels and chemicals.
Continuing with this comparison, while the petroleum components consist of miscible liquids
and gases, with separations based on liquid-vapor equilibrium such as distillation and adsorption,
the biomass fractionation systems are different in that the chemical constituents are typically
solutes in an aqueous phase. The properties of hydrocarbons are very well characterized and are
used for process design and, in particular, separations. Similarly, the dissolved plant cell wall
components in an aqueous phase need further characterization and study to determine the
relevant processes that will be implemented to separate and recover the fractions for biological or
chemical conversion to renewable products.
For this reason thermochemical pulp mills offer an opportunity for the enhanced
utilization of hemicellulose polysaccharides and aromatic lignin biopolymers. The two primary
types of pulp mills consist of the kraft process and the sulfite process. Worldwide the kraft
process dominates the pulp production field with more than 80% (Sjostrom, 1993) of global pulp
production performed through this process. This process is preferred for chemical pulping due to
the favorable economics of chemical recovery and the robustness to produce strong pulps from
hardwoods or softwoods. Through the kraft process hemicellulose and lignin components are

4

solubilized into the aqueous alkaline liquid phase by the peeling reaction and nucleophilic
2-

aromatic substitution of sulfur by S from Na2S at elevated temperatures, respectively.
Dissolved hemicellulose can be stabilized from further depolymerization in the alkaline medium
when its aldehyde end groups are oxidized to carboxylic acids, but at the elevated temperature
settings the dissolved polysaccharides are degraded to aldonic acids (Johansson & Samuelson,
1977; Sjostrom, 1993). During the chemical recovery process, which is illustrated in Figure 1.1,
the dissolved plant cell wall fractions of hemicellulose and lignin are concentrated through
successive evaporation followed by combustion to produce heat and power within the pulp mill
itself.

Fresh Feed NaOH and
Na S
2

Wood
Chips

Kraft
Pulping

Delignified
Wood

Pulp for
Papermaking

Black
Liquor

White Liquor
(Recovered
NaOH and Na2S)
Chemical
Recovery

Brownstock
Washer

Evaporators
Green
Liquor
Recovery
Boiler

High
Pressure
Steam

Figure 1.1. An example of the kraft pulping process with chemical recovery.

5

Lignin recovery through the form of black liquor precipitation is produced at a limited scale for
the production of binders, resins, or activated carbons (Vishtal & Kraslawski, 2011), while
hemicellulose utilization, as a value-added product, is noticeably absent. On a more limited scale
the sulfite process produces cellulosic pulp by the addition of the chemical components sulfite or
bisulfite. This process has more variability since it can occur at acid, neutral, or alkaline
conditions. All three conditions can affect the lignin structure to some degree through
sulfonation, cleavage of certain bond linkages (e.g. β-O-4), and lignin softening and
solubilization (Sjostrom, 1993). Due to the larger amounts of sulfur present in the process and
+

2+

2+

the option for different balancing metal cations (e.g. Na , Ca , Mg ), the chemical recovery of
sulfite black liquor is more challenging and presents greater environmental concerns. The
dissolved plant cell wall fractions, however, have been utilized much more than their
counterparts in the kraft process. The recovery of lignosulfonates is practiced globally while
older sulfite mills would produce bioethanol through the conversion of hemicellulose
polysaccharides by the acidic sulfite process that hydrolyzes hemicellulose to monomeric sugars
in the spent black liquor (Gargulak & Lebo, 2000; Lawford & Rousseau, 1993).
Understanding the underlying chemistries behind these pulping processes gives great
incentives to recover hemicellulose and lignin fractions for greater utilization. Figure 1.2 shows a
hypothetical case of plant cell wall fractions that are dissolved from a hardwood species during a
pulping stage and the options for recovery. The presented case in Figure 1.2 only applies to
alkaline processes. In these processes the pH remains relatively constant, but could slightly
-

decrease due to consumption of OH as hemicellulose and lignin are solubilized. Notice that the
precipitation of dissolved plant cell wall biopolymers will not be exclusively pure hemicellulose
or pure lignin. Precipitation techniques (e.g. addition of organic solvent or pH adjustment) will
6

yield fractions consisting of both hemicellulose and lignin fractions due to their crosslinking
associations within native plant cell walls (Himmel et al., 2007).

Insoluble Solids (g/g dry wood)

1.0

Cellulose
Lignin
Glucuronoxylan
Hemicellulose
Glucomannan
Acetate
Extractives
Sugar Degradation
Products
Inorganics (Pulping Chemicals)

Pulp, e.g. Soda-AQ
or Semi-Chemical

0.5

Potential
Lignin
Precipitation

Potential HC
Precipitation

Soluble Solids (g/g dry wood)

0

Black Liquor
Fractionation

0.5

Potential Biopolymer
Recovery Stages

Pulping Process
(Increasing Time and Temp.)

Figure 1.2. Hypothetical example of hardwood components dissolved during alkaline pulping
followed by the recovery of hemicellulose and lignin by precipitation.
Although precipitation to recover non-cellulosic biopolymers from existing liquor streams is
convenient, it is not the only option for recovery nor does it make a process scheme
7

economically advantageous. Some of the techniques for the recovery of dissolved plant cell wall
components are listed in Table 1.1 and these techniques will be expanded upon.
Table 1.1. Brief description of recovery techniques and mechanisms
Method for Recovery
Acidification

Primary Components
Recovered/Mechanism
Lignin – Aggregation by
increasing protonation of
hydroxyl groups.

Precipitation with Organic
Solvents

Hemicellulose –
Polysaccharide insolubility in
excess organic solvents.

Ultrafiltration

Hemicellulose and lignin –
Utilizing molecular weight
cutoff filters to obtain
oligomeric biopolymers.

Direct Utilization of Liquor
Streams

Hemicellulose – Direct
conversion of monomeric or
oligomeric hemicellulose after
autohydrolysis or sulfite
treatment.

References
(Gidh, Decker, Vinzant,
Himmel, & Williford, 2006;
Guerra et al., 2007; Ragnar,
Lindgren, & Nilvebrant, 2000;
Sarkanen, Teller, Hall, &
McCarthy, 1981)
(Al-Dajani & Tschirner, 2008;
Bian, Peng, Peng, Xu, & Sun,
2010; Stoklosa & Hodge,
2012)
(Ibn Yaich, Edlund, &
Albertsson, 2012; Krawczyk,
Oinonen, & Jönsson, 2013;
Saadatmand, Edlund,
Albertsson, Danielsson, &
Dahlman, 2012)
(Franco et al., 2012;
Helmerius, von Walter, Rova,
Berglund, & Hodge, 2010; G.
S. Wang, Pan, Zhu, Gleisner,
& Rockwood, 2009)

Furthermore, processes to purify recovered streams of hemicellulose or lignin will be needed to
generate more pure fractions with high yields. Some of this recovery is already practiced, but to
offset further petroleum consumption within the United States a greater commitment to their
recovery and conversion to liquid fuels or value-added chemicals should be practiced. Cellulosic
pulp is still a desired commodity for paper production that can be sold at $1017 per metric ton
for bleached softwood kraft pulp (FOEX Indexes, LTD., 18th March 2014). Converting the
cellulosic portion of woody biomass to biobased fuels or chemicals can be accomplished.
However, in a pulp mill setting it is more likely to focus on the non-cellulosic fractions to
8

increase product output and revenue. This can be accomplished with the introduction of new
upstream processes to extract and recover the non-cellulosic biopolymers, or new recovery
processes after a traditional chemical pulping stage. The choice and availability of feedstocks
(i.e. hardwood or softwood) will determine extraction and recovery before or after a pulping
stage and is best exemplified based on the hemicellulose composition for different feedstocks.

1.2 Hemicellulose Composition in Woody Biomass Feedstocks
In contrast to the cellulose portion within plant cell walls, the composition of
hemicellulose can vary extensively between monocots, hardwoods, and softwoods. Cellulose is a
homopolysaccharide of β-1-4-linked glucopyranosyl (Glcp) units that form microfibrils within
the plant cell wall, partly due to hydroxyl groups emanating from one chain and forming
hydrogen bonds with other chains in proximity. Figure 1.3 gives a structural representation of
cellulose.
OH HOH 2C

OH HOH 2C

1

HO
O 4

β O4
O

HOH 2C

HO

1

O

HO

β O4

βO 4

O

1

O
O

HO

OH HOH 2C

1
OH

Figure 1.3. Structural representation of a cellulose chain.
Unlike cellulose, hemicelluloses are amorphous heterpolysaccharides that vary in composition
and structure with respect to species. The dry mass percentage of hemicellulose in the plant cell
wall can vary from 20 to 30%, which represents a significant amount of material that can be
isolated and recovered for conversion processes. When considering grasses select species are
being investigated as potential bioenergy crops for conversion to liquid transportation fuels

9

(David & Ragauskas, 2010; Lau, Gunawan, & Dale, 2009; Xu, Cheng, Sharma-Shivappa, &
Burns, 2010). The Poaceae family, or true grasses, include such varieties of monocots as
bamboo, maize, and barley. Essentially, all members of this family contain a form of xylan as the
dominant polysaccharide chain (Hampton, Haworth, & Hirst, 1929). Differences amongst
hemicellulose distribution in grasses start with side chain substitution on the polysaccharide
backbone. These differences are summarized in Table 1.2 by listing the primary polysaccharide
backbone and side-chain substitutions for each species.
Table 1.2. Summary of hemicellulose composition in select grass species.
Species Type

Primary Polysaccharide
Backbone

Side-Chain
Substitutions

Reference

Bamboo
(Arundinaria
japonica)

Heteroxylans

L-arabinofuranose
(Araf), Ogalactopyranosyl
(Galp), and acetyl

(Wilkie & Woo,
1977)

Corn Stover (Zea
mays)

Glucuronoarabino(galacto)xylan
(GAGX)

Araf, acetyl
residues, and 4-Omethyl glucuronic
acid (4-O-MeGA)
residues

(Carpita &
Whittern, 1986;
Naran, Black,
Decker, & Azadi,
2009)

Barley (Hordeum
vulgare L.)

Glucuronoarabinoxylans (GAX)
with linear β-glucans

Araf, and acetyl

(Izydorczyk &
MacGregor, 2000)

An important distinguishing feature of the cell walls from this family include the presence of
ferulic acid in the primary cell wall (Carpita, 1996). This aromatic acid forms feruloylated
oligosaccharides that crosslink to Araf side chains located on the xylan chain (Saulnier,
Vigouroux, & Thibault, 1995). This is notable due to the ester linkage of ferulic acid to the
hemicellulose polysaccharide that can be broken by mild pretreatment conditions and has been
correlated to show improved digestibility of the polysaccharides by enzymes (Li et al., 2012).

10

These covalently linked structures are important for plant physiological development, but
become a hindrance in the deconstruction of biomass for producing biofuels or value-added
chemicals. Taken together, grass bioenergy crops have great potential for the development of
second generation bioethanol production (Pauly & Keegstra, 2008); however, in this context of
non-cellulosic biopolymer recovery in an IFBR setting they will not be discussed because of the
characteristic property of non-wood feedstocks containing higher amounts of silicon that make
downstream processing in a chemical pulping process difficult (Khiari, Mhenni, Belgacem, &
Mauret, 2010; Madakadze, Radiotis, Li, Goel, & Smith, 1999).
Compared to grasses, the composition of hemicelluloses in woody lignocellulosic
biomass is considerably different. Hardwoods primarily contain glucuronoxylan (GX) that
contains a backbone of β-1,4-linked xylopyranosyl (Xylp) units as shown in Figure 1.4.
OH

O

O
4
RO

O
HO
H 3CO

OH

β
1

HO
O

4

OR

1

O

β

COOH
O

OH HO

O4
RO

O

α O3

OH 1

O
H 3C

OH

2

OH
O

HO

αO

1

OH
O

1
α

O
CO2H

Figure 1.4. Hardwood hemicellulose structure with a 4-O-MeGA side group and an α-L-Rhap-12-linked-α-D-GalpA at the reducing end of the GX chain.

Substitutions occur frequently along the polysaccharide backbone with either acetyl groups at the
C-2 or C-3 position corresponding to 7 acetyl groups per 10 xylose units, or 4-O-MeGA at the C2 location corresponding to 1 residue for every 10 xylose residues (Sjostrom, 1993; Whistler &
Richards, 1970). The presence of 4-O-MeGA units and the lack of Araf side chains distinguishes
11

hardwood xylans from grass xylans. Previous studies have shown that native GX isolated from
various hardwood species have number average degrees of polymerization (𝐷𝑃! ) ranging in
value from 150 to 215 (Goring & Timell, 1960; Koshijima, Timell, & Zinbo, 1965; LeBel,
Goring, & Timell, 1963; Timell, 1960). A distinguishing feature for hardwood GX is their ability
to become resistant against degradation during a traditional chemical pulping process utilizing
alkaline chemistry. At the reducing end of the GX chain a sequence of uniquely linked
carbohydrates can stop the alkali peeling reaction. This sequence, shown in Figure 1.3, starts at
the reducing end with a xylose unit connected by an α-1,4-linked galacturonic acid (GalpA)
residue, which in turn is connected by an α-1,2-linked rhamnopyranosyl (Rhap) unit, ending with
an α-1,3-linked xylose unit to continue the GX chain (Johansson & Samuelson, 1977; Sjostrom,
1993). The initial reducing end unit of the GX chain is degraded by β-elimination to a
saccharinic acid group, which will lead to the exposure of new reducing ends (Whistler &
Richards, 1970). The combination of the Rhap and GalpA unit can hinder the peeling reaction
from progressing further down the GX chain, followed ultimately with a substitution occurring
with the 4-O-MeGA to stabilize the chain against further degradation. For this reason hardwoods
can be viewed as an attractive feedstock to obtain extended utilization of the hemicellulose
portion in pulp mills utilizing soda or kraft pulping processes (Van Heiningen, 2006).
Alternatively, softwood hemicellulose not only has different compositions, but also
physical properties when compared to hardwood hemicellulose. The primary hemicellulose
polysaccharides in softwoods are glucomannans (GM) or galactoglucomannan (GGM). These
polysaccharides contain β-1,4-linked glucopyranosyl (Glcp) and mannopyranosyl (Manp) units
with side chains of α-1,6-linked galactopyranosyl (Galp) units in GGM (Ebringerova, 2006). A
partial substitution of acetyl groups at the C-2 and C-3 position still occur for GGM and GM but
12

at a lower quantity of only one acetyl group per 3-4 hexose units (Sjostrom, 1993). GAX are also
present in softwoods at low quantities (5-10% by weight) and are similar to GX from hardwoods
through having C-2 substitutions with 4-O-MeGA in a ratio of 2 residues per 10 xylose units, but
are dissimilar by having side chains of α-L-Araf at a ratio of 1.3 residues per 10 xylose units
(Sjostrom, 1993). Figures 1.5 and 1.6 show the different structures for GGM and GAX.
HOH 2C
O

O 4
HO

RO

βO 4

OH
HO

OH

1

OR
O

1 HOH 2C

βO 4

βO 4
O
OR

OH2C

1

RO
O

1

RO

βO

HOH 2C

RO

1
α

OH
O
CH2OH

Figure 1.5. Softwood hemicellulose structure of GGM.
OH
O

O4
HO

HO

HO
H 3CO

OH
O
COOH

O

1
1
α

O

βO 4

β O4

O

OH

1

βO 4

O
OH

O

1

O

1

βO

1
α

OH

CH2OH

HO

OH

Figure 1.6. Softwood hemicellulose structure of GAX.
Previous extractions of GGM and GAX from softwoods by different chemical pulping
techniques yielded characteristic 𝐷𝑃! of 90-102 and 89-120, respectively (Jacobs & Dahlman,
2001). By having lower quantities of GAX and lacking the unique Rhap to GalpA sequence at
13

the polysaccharides reducing end, dissolved hemicellulose from softwoods have greater potential
towards degradation when exposed to alkali. Consequently, the isolation and recovery of
hemicellulose from hardwoods and softwoods as a source for conversion to liquid transportation
fuels or value-added chemicals can be developed by the use of different processes. This process
integration could include alkali pre-extractions for hardwoods (i.e. a less severe soda pulping
process) while softwoods would best be utilized in an autohydrolysis or acid sulfite process
(Phillips, Jameel, & Chang, 2013; Saadatmand et al., 2012; Stoklosa & Hodge, 2012; Van
Heiningen, 2006; Yoon, Tunc, & Van Heiningen, 2011; Zhu et al., 2010).

1.3 Existing Liquor Streams
To improve hemicellulose and lignin utilization from woody biomass, existing black
liquor streams produced by the pulping industry may play an important role by recovering these
plant cell wall fractions. This section discusses current operations performed in the pulping
industry along with the type of hemicellulose or lignin that can be recovered, how they are
separated from liquor streams, and their potential conversion to fuels or chemicals.

1.3.1 Chemi-Mechanical Pulping Glucomannan Streams
Although not as prevalent as kraft or sulfite pulping, chemi-thermomechanical pulping
(CTMP) or thermomechanical pulping (TMP) is utilized to produce pulps with higher overall
yields and higher lignin content. The pulping conditions are less severe in terms of pH, but still
offer the potential to recover hemicellulose oligomers that are dissolved during the process.
Liberated hemicellulose fractions that contain both sugar monomers and oligomers can be
collected from process water used during CTMP or TMP. In the case of using a dilute sulfuric
14

acid pre-extraction on a hardwood feedstock the hemicellulose fraction in the plant cell wall is
hydrolyzed to monomers while the residual wood chips are mechanically refined to produce a
pulp that still maintains industrially important strength properties (W. Liu, Hou, Mao, Yuan, &
Li, 2012; W. Liu, Yuan, Mao, Hou, & Li, 2012). The hydrolyzed sugars can serve as a feedstock
for ethanol fermentation. Alternatively, hemicellulose oligomers have applications as a source
for barrier films with low oxygen permeability in the packaging industry. Dissolved oligomers
within process water fractions after TMP can be concentrated through the use of ultrafiltration or
microfiltration (A. Andersson, Persson, Zacchi, Stalbrand, & Jonsson, 2007; Persson, Krawczyk,
Nordin, & Jonsson, 2010). GGM from softwoods have been studied as a potential source for the
barrier film production. A CTMP process applied to spruce produced soluble polysaccharides
that passed a 10 kDa membrane cutoff recovered 90% of the hemicellulose fraction with a
sufficient lignin content to promote oxidation by laccase enzymes for barrier film production
(Krawczyk et al., 2013). Comparatively, an economic analysis on spruce hemicellulose
recovered from TMP process water fractions at 30 g/L and 80% purity yielded a net product of 4
metric tonnes produced daily at €670 per metric tonne, which is less expensive than the
petrochemical ethylene vinyl alcohol (Persson, Nordin, Zacchi, & Jönsson, 2007). Aside from
CTMP or TMP, woody biomass prehydrolysis in water prior to pulping can be performed
producing liquid side streams of hydrolyzed hemicellulose. Both GGM from softwood
hemicellulose and GX from hardwood hemicellulose recovered after prehydrolysis have been
evaluated for barrier film production (Ibn Yaich et al., 2012; Saadatmand et al., 2012) by
incorporating additives such as carboxymethylcellulose or microfibrillated cellulose for barrier
films derived from hemicellulose to show that tensile strength and oxygen permeability
properties meet the performance requirements associated with current barrier film production.

15

From a process point of view current CTMP and TMP mills would have to add separation
and recovery stages for process water fractions. Hemicellulose oligomers could be separated
through an ultrafiltration step whereby the polysaccharides oligomers are recovered for use as
barrier films and the concentrated monomeric sugar could be collected for a fermentation
process. Although lignin removal by CTMP or TMP is low, these processes are known to
deacetylate biomass, which has implications for the fractions solubilized and the residual solid
fractions. The noteworthy trends include diffusion limited release of large hemicellulose
polymers (i.e. high 𝐷𝑃! ), which can be increased after the refining stage, and GGM insolubility
after deacetylation (Konn, Pranovich, & Holmbom, 2006). Furthermore this release of acetyl
groups as acetic acid can inhibit downstream fermentation indicating that an additional
detoxification process should be added before fermentation (Sun, 2009). Although the
production of a liquid fuel can be achieved through CTMP or TMP dissolved hemicellulose, the
isolation and recovery of higher molecular weight hemicellulose may provide better utilization as
a source for barrier film production.

1.3.2 Polysaccharides from Viscose Pulp:
Viscose pulping is a process capable of producing derivatized forms of cellulose for the
production of semi-synthetic fibers, but has further applications due to the removal of
hemicellulose and lignin components. Traditionally, fibers such as rayon or cellophane can be
produced from the viscose process which incorporates steeping woody biomass in 18% NaOH
solution followed by taking the residual alkali solids and treating them with carbon disulfide to
produce a semi-synthetic fiber called cellulose xanthate (Sjostrom, 1993). The high alkali
concentration used allows a large fraction of hemicellulose to be dissolved, which can then be

16

separated and recovered for a source of oligosaccharides. High purity xylan from the alkali
extract can be recovered through dialysis followed by precipitation with sulfuric acid to yield
oligosaccharides (Griebl, Lange, Weber, Milacher, & Sixta, 2006). Griebl et al. (2006)
characterized xylan oligosaccharides (XOS) by hydrothermolysis to assess the stability of xylan
as a renewable material and found that the soluble fraction of xylan after hydrothemrolysis
contained neutral XOS, acidic XOS, and degradation products due to the high temperature
pretreatment (Griebl et al., 2006). Additionally, XOS comprised the largest fraction of dissolved
plant cell wall material from an autohydrolysis study performed on southern mixed hardwoods at
temperatures exceeding 150°C and indicated that up to 70% of the total hemicellulose content
dissolved at the highest severity (Tunc & van Heiningen, 2008). Commercially, acid extractions
are feasibly performed in pulp mills producing viscose pulps, usually as an acid sulfite stage.
Examples of this include the Dömsjö mill in Örnsköldsvik, Sweden where the hemicellulose
sugars are fermented to ethanol and the Lenzing mill in Lenzing, Austria where the xylose and
acetic acid are recovered and furfural is produced from the xylose. Furthermore, succinic acid
production by genetically engineered strains of Escherichia coli was successful at the laboratory
scale in utilizing all hemicellulosic sugars originating from spruce after a two-stage acid
pretreatment (Hodge, Andersson, Berglund, & Rova, 2009). Succinic acid production is not for a
biologically derived fuel, but instead as a commodity chemical with applications in food
ingredients, pharmaceutical additives, diesel fuel additives, and as a 4-carbon building block for
higher value polymers (Song & Lee, 2006; Werpy, Frye, & Holladay, 2006; Zeikus, Jain, &
Elankovan, 1999). As with many other commodity chemicals succinic acid is currently produced
from petrochemically derived butane through maleic anhydride (Zeikus et al., 1999). This

17

example of a biologically derived commodity chemical shows one other instance of how
petroleum dependence can be decreased through biomass conversion.
Another possibility for recovering hemicellulose fractions from viscose pulps is through
the use of sequential extractions with alkali or the use of specific hemicellulases to remove
hemicellulose polysaccharides. Bleached hardwood kraft pulp contain appreciable amounts of
GX that can be removed in high purity through extractions with alkali or nitren (i.e. a metal
complex of nickel(II)-hydroxide and tris(2-aminoethyl)amine); both gave similar xylan yields
but alkali xylans exhibit a lower DPN with higher substituent values than nitren xylans (Janzon,
Saake, & Puls, 2008). The combination of alkali extractions with a treatment of xylanases can
produce a range of xylans from polymeric to XOS. Hakala et al. (2013) found that xylan from
hardwood kraft pulp could be separated into two fractions that included one with a high yield of
original pulp xylan (60%) and one with a high molecular weight (40,000 Da) but with a lower
yield (~7%), and that xylan with essentially unaltered molecular weight plus XOS could be
obtained after an alkaline extraction (Hakala, Liitia, & Suurnakki, 2013). This also has been
shown previously with cold alkali extractions on bleached hardwood kraft pulp and recycled
fibers before xylanase treatment to enhance the xylan removal (Jackson, Heitmann, & Joyce,
1998). Viscose pulps, consequently, have the potential to allow an advantage in utilizing the
entire plant cell wall fractions since the three main biopolymers of cellulose, hemicellulose, and
lignin are separated completely. Through aqueous or biological based separation procedures,
followed by recovery of oligomeric or polymeric material if needed, the hemicellulose
polysaccharides can be obtained as a renewable substrate for the conversion to other fuels or
chemicals.

18

1.3.3 Sulfite pulping: Lignin and hemicelluloses
Although sulfite pulping is not practiced as much as kraft pulping, the residual black
liquor from sulfite pulping processes or similar processes can still contribute to renewable fuel or
chemical production. During sulfite pulping lignin is sulfonated under acidic and basic
conditions leading to solubilization. Also, the acidic chemistry allows for hemicellulose
hydrolysis to lower molecular weight oligomeric and monomeric sugars. Ethanol was once
produced at a larger volume scale from sulfite black liquors by fermenting the hexose sugars of
softwood GGM. Due to the gradual phase-out of sulfite mills in favor of the kraft process, only a
few sulfite mills remain that continue to practice ethanol fermentation of hemicellulose sugars
such as the Tembec pulp mill in Temiskaming, Quebec (Williamson, 2013). Even with the
decline of sulfite pulping, research continues for producing bioethanol from spent sulfite liquors.
A recombinant strain of Escherichia coli produced variable results for ethanol production, from
70% to 92% conversion efficiency of sugars for hardwood spent sulfite liquors (Lawford &
Rousseau, 1993). Recombinant yeasts from Candida shehatae, Pichia stipitis, and
Saccharomyces cerevisiae were able to ferment spent sulfite liquors originating from softwood
feedstocks with ethanol concentrations ranging from 13.0 to 18.8 g/L, and yields corresponding
to 0.34 to 0.45 g/g (ethanol/sugar) (Olsson & HahnHagerdal, 1996). While acidic sulfite pulping
is more conducive for obtaining a stream of hydrolyzed sugars, alkaline sulfite pulping can be
viewed as a pretreatment by removing lignin from the wood cell wall while maintaining a
stabilized polysaccharide chain in the wood chips. Franco et al. (2012) explored alkaline sulfite
pulping followed by disk refiner to produce pulps amenable to fermentation by commercial
Saccharomyces cerevisiae with a maximum ethanol concentration of 22.0 g/L (Franco et al.,
2012).

19

Additionally, a new subset of research has looked into a process known as sulfite
pretreatment to overcome the recalcitrance of lignocellulose (SPORL) (Shuai et al., 2010; G. S.
Wang et al., 2009; Zhu, Pan, Wang, & Gleisner, 2009). This process differs from traditional
sulfite pulping by operating at different chemical conditions and preparing a feedstock that is
readily digestible to monomeric sugars for the production of bioethanol (G. S. Wang et al.,
2009). Softwoods have been the common feedstock analyzed because of their perceived
recalcitrance by having higher lignin content, and possessing a larger fraction of hexose sugars in
the hemicellulose fraction comprised of GGM. High glucan conversion during enzymatic
hydrolysis can be achieved only after 48 hours for most of the conditions analyzed; however,
some of these high conversions occur with high loadings of sulfuric acid where corrosion can be
an issue within the reactor vessel. Though the acid loadings are high, it serves the purpose of
maximizing hemicellulose removal and simultaneously hydrolyzing the polysaccharides in the
spent liquor to monomeric sugars for fermentation to bioethanol. After pretreatment SPORL
residual wood chips are passed through a disk refiner for size reduction followed by enzymatic
hydrolysis to yield a counterpart sugar stream for fermentation. Consequently, fermentation can
be performed on almost all of the polysaccharides from the woody biomass minus the losses
from conversion to furan degradation components. These furans are formed under acidic
conditions by dehydration reactions that form furfural from pentose sugars and HMF from
hexose sugars. For SPORL pretreatment the spent liquor is conditioned for the removal of these
products either by rotary evaporation or adsorption to a polymeric adsorbent (e.g. Amberlite
XAD-4) (Lan, Gleisner, Zhu, Dien, & Hector, 2013; Zhu et al., 2010). Previous fermentation
trials have been completed with engineered strains of S. cerevisiae to utilize hexose and pentose
sugars for the production of ethanol. A combined fed-batch simultaneous saccharification and

20

fermentation (SSF) procedure on lodgepole pine achieved a commercially feasible ethanol
concentration of 47.4 g/L, corresponding to 285 L/tonne of wood, while independent spent liquor
fermentation separate from SSF on pretreated lodgepole pine could produce 276 L/tonne wood,
or 72% theoretical yield (Lan et al., 2013; Zhu et al., 2010). With this process being analogous to
sulfite pulping, further recovery of lignosulfonates after pretreatment can diversify the product
portfolio for increased revenue since lignins can be valuable co-products obtained from woody
biomass meeting range of product applications including adsorbents, epoxy resins, adhesives,
and pellets for combined heat and power generation (Holladay, Bozell, White, & Johnson, 2007;
Lora & Glasser, 2002).

1.3.4 Lignin Recovery by acidification
Tangential to biologically derived liquid fuels, lignin recovery from aqueous black
liquors can potentially integrate with existing kraft or soda pulping mills to generate renewable
co-products and decrease recovery boiler loads. During alkali pulping a sizable fraction of the
lignin from the plant cell wall is dissolved and can be recovered from the aqueous phase by
lowering the pH through acidification that will protonate lignin moieties containing ionizable
groups to increase aggregation (Gidh et al., 2006; Guerra et al., 2007; Ragnar et al., 2000;
Sarkanen et al., 1981). Besides direct addition of acid it is also possible to decrease the pH of
black liquor by utilizing CO2. Nagy et al. (2010) characterized precipitated lignin and residual
lignin remaining in the filtrate after pH reduction and found that lignin still dissolved in the
filtrate had more carboxylic and phenolic groups while the precipitated lignin had more
methoxyl content (Nagy, Kosa, Theliander, & Ragauskas, 2010). Alternatively, Velez and Thies
(2013) described obtaining lignin from a softwood kraft black liquor by sparging CO2 into the

21

bulk aqueous phase at elevated temperature and pressure (115°C and 6.2 bar) where the
precipitated lignin emerges from the black liquor as a highly solvated suspension that is termed
“liquid-lignin” (Velez & Thies, 2013). Softening point analysis performed on these “liquidlignin” fractions showed that the solvated fractions were all below 115°C, but the dried fractions
tested up to 375°C for their softening point (Velez & Thies, 2013). This “liquid-lignin” offers the
advantage of not requiring extra separation processes, such as microfiltration or nanofiltration,
that would be needed to recover soluble lignin (Jönsson, Nordin, & Wallberg, 2008; Mänttäri,
Van der Bruggen, & Nyström, 2013). Some process design and costing has been evaluated for
the acidification of black liquor by CO2. The equipment was specified to recover only 9.5% of
the lignin residing as a soluble component in the black liquor and 21% of the recovered lignin
would be sold off for specialty chemical production and the rest used as a fuel within a pulp mill
(Loutfi, Blackwell, & Uloth, 1991).
Today there are small subsets of companies that do practice the recovery of lignin on an
industrial scale for renewable materials applications. The “LignoBoost” process utilizes the
precipitation of lignin by the addition of CO2 followed by filtering. After filtering the lignin is redispersed in water and acidified further which aids in obtaining higher lignin yields with lower
ash and carbohydrate content, and lowers the amount of acidic washing water that is needed
(Tomani, 2010). The company Borregaard based in Sarpsborg, Norway also utilizes lignin
recovery for the production of a range of chemicals including industrial binders, emulsion
stabilizers, and even vanillin. Processes and companies such as the ones discussed show the
potential that lignin can have in terms of producing value-added chemicals alongside the liquid
fuels derived from hemicellulosic components.

22

1.4 Pre-Extraction of Hemicellulose Prior to Pulping
1.4.1 Hot Water Pre-Extractions
Existing pulping liquor streams can be utilized for the recovery of dissolved
hemicellulose and lignin for use as conversion to liquid fuels or higher value chemicals.
Alternatively, the implementation of new processes into pulp mills to pre-extract desired cell
wall components will enhance the utilization of hemicellulose and lignin fractions while
providing product diversification. These processes could be based around extractions with hot
water (i.e. autohydrolysis) or alkali prior to pulping. Most pulp mills have on-site water
treatment that can allow autohydrolysis processes to make use of closed water loops in the mill.
Processes utilizing hot water at elevated temperatures and pressure fall right in line for the
production of specialty fiber applications, but the additional removal of hemicellulose
components as a soluble side stream gives rise to further applications for the development of
renewable products. Specific studies have focused on utilizing autohydrolysis for the production
of a soluble sugar stream through the hydrolysis of hemicellulose polysaccharides followed by a
chemical pulping stage to obtain cellulosic fibers. Helmerius et al. (2010) looked at preextraction of Betula pendula by autohydrolysis or varying alkali loadings to assess the impact of
hemicellulose removal, conversion to value-added chemicals, and downstream pulping on the
residual wood chips after extraction (Helmerius et al., 2010). The alkali extractions produced
lower recoveries of hemicellulose than compared with the water extraction, most likely due to
the degradation of hemicellulose polysaccharides to saccharinic acids, but the alkali extracted
wood chips produced favorable kraft pulp yields and strength properties when compared to a
control kraft cook (Helmerius et al., 2010). Although water extractions at elevated temperatures
do not necessarily connect well with kraft pulping processes due to pH fluctuations, it was found

23

that a hydrolysable liquor was produced by the autohydrolysis of Betula pendula which could
produce succinic acid by fermentation (Helmerius et al., 2010). The main adverse consequence
of autohydrolysis processes on woody biomass is the decline in pulp yield and quality due to the
depolymerization occurring to a small portion of the cellulose microfibrils (Helmerius et al.,
2010; Mendes, Rocha, Sousa, & Carvalho, 2011).
Apart from unwanted cellulose microfibril hydrolysis and pulp yield decreases, the
hemicellulose fraction removed by autohydrolysis can degrade at the elevated temperatures (e.g.
180-240°C) used in the process through dehydration reactions that form furfural (Borrega,
Nieminen, & Sixta, 2011). Although these side reactions are unwanted, the autohydrolysis of
woody biomass can be implemented in such a way to make them desirable. For example it was
found that a hemicellulose content below 3% after autohydrolysis decreased the pulp yield after a
soda/anthraquinone (AQ) pulping stage, but the yield of cellulose could be improved by
supplementing the pulping stage with sodium borohydride (Borrega, Tolonen, Bardot, Testova,
& Sixta, 2012). The degradation of xylose to furfural can also be exploited since furfural is a
commodity chemical that is typically produced by the acid hydrolysis of plant substrates
(Mamman et al., 2008). While the hydrolysis liquor after autohydrolysis contains many different
dissolved organic components (i.e.hemicellulose, lignin, furfural), detoxifying can help to
remove inhibitory fermentation components by using activated carbon (AC), oxidized AC, or
polydiallyldimethylammonium chloride (PDADMAC) to absorb lignin and furfural inhibitors at
83.3% and 100%, respectively, in high quantities (X. Liu, Fatehi, & Ni, 2012). Ultimately, if
some decreases in pulp yield and quality can be tolerated within a certain range the
autohydrolysis of woody biomass prior to chemical pulping offers many advantages in terms of

24

producing different liquid fuels or chemicals; especially if refiner energy savings can be
correlated to hemicellulose removal (Kenealy, Houtman, Laplaza, Jeffries, & Horn, 2007).

1.4.2 Alkali Pre-Extractions
Alternatively, the pre-extraction of woody biomass with alkali presents a much more
favorable option for integration with current alkali based chemical pulping processes. Figure 1.7
presents a potential process that utilizes a soda pulping process (only alkali) for the production of
biologically or catalytically derived fuels. Alkali pre-extractions will not only extract
hemicellulose polysaccharides but will also pull out more lignin than autohydrolysis processes,
and allow current chemical pulping processes to decrease the alkali loading in the actual pulping
stage and facilitate improved recovery boiler operation by sending a black liquor stream with
lower solids content. Alkali extractions have been executed under low temperature conditions to
assess feasibility and at high temperature conditions to prove industrial relevance.

Soda
Pulping

Wood
Chips
Regenerated
Alkali

Chemical
Recovery

Delignified Wood

Enzymatic
Hydrolysis
Sulfur-free Black
Liquor

Sugars for
Biochemically
Derived
Biofuels

Gasification
Sulfur-free Syngas
Fuel
Synthesis

Thermochemically
Derived Biofuels

Figure 1.7. Potential process diagram for utilizing soda pulping (or alkali pre-extraction) to
generate biofuels.

25

Low temperature alkali extractions have been performed previously at room temperature and up
to 90°C with a multitude of woody biomass feedstocks. Giant bamboo of the species Bambusa
balcooa cv. Roxburgh had 13.6% of its initial xylan content extracted after exposure to 1 M
NaOH at 90°C for 4 hours that resulted in improvements for pulp yields after kraft pulping, no
reduction in pulp viscosity, and only a slight decrease in pulp strength properties (Vena, Brienzo,
del Prado García-Aparicio, Görgens, & Rypstra). Comparatively, Al-Dajani and Tschirner
(2008) investigated mild alkaline extractions at low temperatures (50-90°C) for Populus
tremuloides (aspen) to find that for one metric ton of wood chips supplied up to 40-50 kg of total
wood components can be dissolved by alkali and showed further that kraft pulping on the
extracted wood chips needed lower chemical loadings and shorter cooking time (Al-Dajani &
Tschirner, 2008). A disadvantage that stands out from this study was a 10% decrease in the tear
index and a relatively low yield of extracted hemicellulose; however, the benefits of the alkali
extraction prior to pulping still show promise since the overall pulp properties are very
comparable to control kraft cooks (Al-Dajani & Tschirner, 2008). While these low temperature
studies show promise for hemicellulose extraction and recovery from aqueous alkali liquors,
higher temperatures studies are a better indicator for hemicellulose utilization at industrial
process parameters.
At higher extraction temperatures (125-160°C) with alkali (NaOH or green liquor) it is
possible to remove 5-40% of the total wood components depending on temperature and alkali
charge. Yoon et al. (2011) found that a 3% green liquor extraction with 0.05% AQ showed
improved delignification rates, higher yields, higher tear resistance, lower refiner response, and a
2.5-3% lower effective alkali consumption when compared to traditional kraft cooks (Yoon et
al., 2011). Apart from batch alkali pre-extractions, it is also possible to obtain higher value from

26

hemicellulose in woody biomass by different extraction processes or post-pulping extractions. A
modified twin-screw extruder was analyzed for extracting hemicellulose from poplar found that
90% of the initial hemicellulose could be extracted at much lower liquid-to-solid loadings when
compared to batch studies and hemicellulose solubilization could be correlated to the specific
mechanical energy that was absorbed by the process material (N'diaye, Rigal, Larocque, &
Vidal, 1996). Also, it is possible to utilize post-alkali extractions after kraft pulping to extract
and later recover hemicellulose. Post-alkali extractions were performed at room temperature to
remove hemicellulose from birch kraft pulp as a means to produce xylan films for use in the
packaging industry. Alekhina et al. (2013) showed that 60% of the xylan present in the pulp
could be extracted by alkali at room temperature with the extracted xylan being an essentially
pure polysaccharide with no acetyl side units and with a very low content of 4-O-MeGA
(Alekhina, Mikkonen, Alén, Tenkanen, & Sixta, 2013). A stable, self-supporting pure xylan film
could not be formed, but through carboxymethylation it was shown that with an increase in the
degree of substitution the film elongation at break improved along with the reduction in oxygen
permeability while water vapor permeability increased; thus showing the potential that xylan has
as a biodegradable film for packaging in the food industry (Alekhina et al., 2013). As discussed
above hemicellulose recovery and utilization can be varied and diverse. The end usage of
hemicellulose will ultimately be determined by the economics associated with industrial
chemical facilities and the value of the end product desired.

1.4.3 Renewable Products from an IFBR
To determine end products derived from hemicellulose, researchers have looked at
economic studies revolving around coupling pre-extraction processes with current pulp mills in

27

an IFBR setting. The proposed IFBR model to convert pulp mills has been analyzed mostly from
the perspective of producing bioethanol for liquid fuel utilization. An economic model of an
IFBR was evaluated on three scenarios: traditional kraft pulping, hemicellulose pre-extraction
prior to pulping for ethanol production, and hemicellulose pre-extraction prior to pulping
coupled with the conversion of short cellulosic fibers to ethanol (Huang, Lin, Ramaswamy, &
Tschirner, 2009; Huang, Ramaswamy, Al-Dajani, & Tschirner, 2010). This study found that with
a fixed feed rate of 2000 Mg wood per day a pulp-mill IFBR utilizing alkali pre-extraction, the
hemicellulose and short cellulosic fibers converted to ethanol produced 10.04 MM gal (0.038
3

3

MM m ) of ethanol per year with a minimum-selling price of $1.86/gal ($491/m ) (Huang et al.,
2010). The economic simulation found with hemicellulose pre-extraction the pulping capacity
could be increased by 22%, but both pulp-mill IFBR simulations found that less steam would be
produced in the recovery boiler due to a lower solids content in the black liquor; however, this
could be offset by a 13% fuel usage reduction in the lime kiln from a lower loading of chemical
inorganics (Huang et al., 2010).
Aside from ethanol production, value-added chemicals also have industrialization
potential in a biorefinery. A cultivar of larch, Larix sibirica Lebed, was analyzed for use in a
biorefinery utilizing hot water extractions to dissolve hemicellulose at optimized conditions
(140-150°C for 60-90 minutes) followed by acid hydrolysis and fermentation or SSF with
Bacillus coagulans MXL-9 for the production of lactic acid (Horhammer, Walton, & van
Heiningen, 2011). Fermentation after acid hydrolysis produced favorable results by converting
36 g/L of monomeric sugars to 28 g/L of lactic acid (78% yield), but SSF did not produce
optimal results, which was attributed to a lack of enzymes that were specific to the larch species
and not to degradation products (e.g. furfural, acetate, etc.) since they constituted less than 1% of
28

the hot water extract (Horhammer et al., 2011). After hot water extraction kraft pulp yields were
lower by 4-5%, but this yield could be improved through polysulfide-anthraquinone (PSAQ)
pulping to produce a good papermaking pulp with similar yields (Horhammer et al., 2011).

1.5 Further Investigating the Direct Fractionation of Black Liquors
The highest volume of liquor streams generated from pulping processes, and therefore
containing the most dissolved hemicellulose and lignin, originate in kraft and sulfite pulping
mills. Although recovery of dissolved non-cellulosic plant cell wall fractions are possible from
these streams, negative side effects still remain. At the temperatures and chemical loadings
traditionally used in those pulping processes it is expected that a sizeable portion of the
hemicellulose would be degraded to saccharinic acids, ultimately giving low yields of
hemicellulose. Lignin would be dissolved in high quantities, but most of the lignin recovered
would be sulfonated. A market does exist for sulfonated lignin to be used as a binding agent,
however, if other materials are more desirable the lignin needs to be in a more pure form. For
example, lignin can be used as a precursor to generate carbon fibers as long as it is free from
contaminants, such as inorganics, that can contribute to undesirable broad melting point
temperatures (Compere, Griffith, Leitten, & Pickel, 2005). Additionally, the sulfur compounds in
the cooking chemicals for these pulping processes still present environmental concerns for
downstream processing. Black liquor gasification presents a thermal catalytic conversion route
for lignin to produce drop-in fuels, but residual sulfur components diminish the potential
benefits. After gasification H2S is formed which will require the use of a scrubber to remove the
component from the stack gas, thus increasing capital cost. Also, any residual sulfur remaining in
the syngas will decrease the yield of drop-in fuel produced since sulfur is known to contaminate
29

catalysts. Kraft and sulfite pulping processes should not be required to remove sulfurous cooking
chemicals from their processes, instead, the options to extract, recover, and separate noncellulosic biopolymers from wood biomass can be achieved from alkali pre-extractions prior to
pulping, or an alkali only (i.e. soda) pulping process.
The research presented in this dissertation investigated how alkali pre-extractions, kraft
pulping, and soda pulping liquors can be fractionated to recover chemically altered noncellulosic biopolymer fractions that have potential to be used in biological or catalytic
conversion processes. In chapter 2 alkali pre-extractions were performed on a variety of
hardwoods and a single species of grass to determine the ability of alkali to preferentially extract
hemicellulose or lignin, followed by the recovery and characterization of the altered
biopolymers. The lone monocot analyzed and a species of birch produced the highest yields of
hemicellulose recovered, thus indicating high extractability in alkali prior to pulping. For chapter
3 a detailed characterization was carried out on seven fractions of lignin obtained from softwood
kraft black liquor at discrete pH levels. This research explored how each fraction was similar or
different in terms of characteristics such as contaminating polysaccharides, functional group
content, molecular weight, thermal degradation, and lignin monomers generated by analytical
pyrolysis. Correlations were made on several properties to understand how the pH at
fractionation can produce distinct fractions of lignin to be used for select conversion processes.
Lastly, chapter 4 investigates how a soda pulping process influences hemicellulose and lignin
dissolution during the process, and producing a digestible cellulosic substrate after pulping.
Dissolved hemicellulose and lignin were quantified at specific time points during the heat up and
cook phase of the soda pulping process to determine the mass quantities obtained in the aqueous
solution. After pulping the residual wood chips were enzymatically hydrolyzed to determine
30

yields of soluble sugars. The goal here was to identify if soda pulping is an effective process to
generate soluble sugars from cellulose as opposed to generating pulp for papermaking. Taken
altogether the research evaluated in this dissertation attempts to understand more thoroughly how
altered cellulosic and non-cellulosic plant cell wall components affected by alkali treatment
processes can contribute to the advancement of renewable fuels, chemicals, and materials.

31

CHAPTER 2. EXTRACTION, RECOVERY, AND CHARACTERIZATION OF
HARDWOODAND GRASS HEMICELLULOSES FOR INTEGRATION INTO
BIOREFINING PROCESSES

This work has been published as original research on 3rd August 2012, in Industrial and
Engineering Chemistry Research, Volume 51, and Pages 11045 – 11053.

2.1 Introduction
Hemicelluloses are the second most abundant class of polysaccharides after cellulose in
the biosphere, comprising up to 40% by mass of plant cell walls. Xylans are the predominant
hemicelluloses in the secondary cell walls of angiosperms and represent an immense,
underutilized resource. Xylans have many potential commercial applications with increasing
recent interest in their use for the production of green chemicals (C. Andersson, Hodge,
Berglund, & Rova, 2007), renewable biofuels (Girio et al., 2010), and green polymers (Hansen &
Plackett, 2008), among others. However, the current utilization of xylans is very limited due to
the challenges of economic extraction, recovery, and conversion to these higher value products.
One approach for improving the economic utilization of xylans is through integration with other
processes that fractionate and convert plant cell wall biopolymers to fuels and chemicals in a
“biorefinery” approach. Biorefining approaches utilizing xylan can be envisioned to be as large
as current pulp mill capacity utilizing woody based feedstocks to extract xylan prior to chemical
pulping (Helmerius et al., 2010; Ragauskas et al., 2006; Van Heiningen, 2006) or with a
cellulosic biofuels process utilizing chemical pretreatment and enzymatic hydrolysis of
polysaccharides (Wyman et al., 2005).

32

Angiosperm xylans consist of a β-1,4-linked Xylp backbone with the possibility for
substitutions along this backbone that may include O-acetyl subunits at the C-2 and/or C-3
position of the Xylp backbone (Carpita & Whittern, 1986; Rosell & Svensson, 1975; Teleman,
Tenkanen, Jacobs, & Dahlman, 2002; Timell, 1960), and Araf subunits attached with α-1,2
and/or α-1,3 linkages (Ebringerova, 2006). As a proposed cross-linking glycan, xylans are
known to form strong associations with cellulose microfibrils or to each other and their ability to
associate with other glycans is thought to be through H-bonding and van der Waals forces
between unsubstituted regions along the backbone (Kohnke, Ostlund, & Brelid, 2011; Linder,
Bergman, Bodin, & Gatenholm, 2003). Glucuronoarabinoxylans (GAXs), along with mixedlinkage β-glucans, are known to be the dominant cross-linking glycans in the primary cell walls
of cereals while glucuronoxylans (GXs) are the principal hemicelluloses in the secondary walls
of all angiosperms. The glucomannans (GMs) and galactoglucomannans (GGMs) consist of a
backbone of β-1,4-linked Glcp and Manp units with α-1,6-linked Galp subunits present in GGMs
(Ebringerova, 2006). GGMs comprise only a small fraction in the cell walls of angiosperms, but
are the most abundant hemicellulose in gymnosperms.
A defining characteristic of the hemicelluloses is that these are soluble in alkali and can
be extracted from the cell wall with minimal polymer depolymerization (Carpita, 1983) apart
from deacetylation due to saponification of acetyl esters. This improved solubility is presumably
due to deprotonation of hydroxyl groups on hemicelluloses at high pH. As a consequence of this,
alkaline processes specifically for extraction of xylans have been a topic of ongoing research for
woody dicots (Al-Dajani & Tschirner, 2008; Karlsson, Roubroeks, Glasser, & Gatenholm, 2006;
N'diaye et al., 1996) and agricultural residues such as corn stover (Cheng, Zhan, Fu, & Lucia,
2010). As mentioned previously, integration with chemical pulping or cellulosic biofuels

33

processes in a biorefinery approach, in particular using alkaline chemistries for the extraction,
fractionation, or pretreatment, opens up new possibilities for utilizing all the cell wall fractions in
an integrated manner to generate higher value fuels and chemicals.
The kraft process has developed as the dominant chemical pulping process due to the
favorable economics of chemical recovery and the robustness of the process to produce strong
pulps from a wide range of feedstocks, with over 80% of the chemical pulp produced globally
utilizing the this process (Sjostrom, 1993). During any alkaline pulping (i.e. kraft, soda,
soda/AQ, etc.) the lignin is significantly depolymerized and solubilized while a large proportion
of the hemicellulose is degraded to carboxylic acids and ultimately concentrated and burned in a
recovery boiler. Due to this, hemicellulose extraction prior to pulping has been proposed for
integration into kraft pulp mills (Helmerius et al., 2010; Van Heiningen, 2006; Yoon, van
Heiningen, & Krishnagopalan, 2008) using methods that include, among others, hot water
pretreatment, dilute acid hydrolysis, neutral sulfite, and mild alkaline extraction steps.
Commercially, acid sulfite or steam hemicellulose hydrolysis is feasibly performed in pulp mills
producing viscose (dissolving) pulps using sulfite or kraft pulping processes, respectively.
However, integrating a hemicellulose pre-extraction step into a kraft process can be challenging
for applications where fiber strength is a requirement. Solubilized hemicellulose re-deposited on
fibers provides one component of the strength characteristics of the resulting pulps (Molin &
Teder, 2002) and this has recently been shown through hot water xylan extraction that
significantly decrease the strength of the resulting kraft fibers cooked to similar Kappa numbers
and refined to equivalent freeness values relative to alkaline extractions (Helmerius et al., 2010).
Additionally, acid extractions prior to alkali pulping can result in incomplete impregnation of
alkali into chips resulting in uneven cooks and increased alkali requirements. For this reason we

34

propose that alkaline hemicellulose extractions can be performed prior to pulping in a manner
that result in high hemicellulose yields without significant degradation and integrates well with
existing alkaline chemical pulping.
Apart from alkaline chemical pulping, another area of interest for the effects of alkaline
treatment on the biopolymers contained in plant cell walls is chemical pretreatments for
cellulosic biofuels processes utilizing a chemical pretreatment coupled to an enzymatic
depolymerization of the polysaccharides to generate fermentable sugars. A wide range of
reaction chemistries are possible and these pretreatments generally share common features that
may involve chemically modifying, depolymerizing, solubilizing, and/or physically
redistributing a portion of the hemicelluloses and lignins in the cell wall. For acidic
pretreatments (Wyman et al., 2005) where much of the recent research and nascent
commercialization attempts have been focused, the hemicelluloses are hydrolyzed to low MW
oligomers, monomers, or sugar dehydration products and when oligomers are present, the
deacetylation is often incomplete (unpublished observations). Alkaline pretreatments such as
NaOH (Sills & Gossett, 2011; Xu et al., 2010), Ca(OH)2 (Chang, Kaar, Burr, & Holtzapple,
2001), and anhydrous (Wyman et al., 2005) or aqueous ammonia (T. H. Kim, Kim, Sunwoo, &
Lee, 2003) or alkaline oxidative pretreatments such as Na2CO3-O2 (Klinke, Ahring, Schmidt, &
Thomsen, 2002), Ca(OH)2-O2 (Chang, Nagwani, Kim, & Holtzapple, 2001), and NaOH-H2O2
(Banerjee, Car, Scott-Craig, Hodge, & Walton, 2011) impact the hemicellulose (and lignin)
fraction in a significantly different ways than acidic pretreatments.
There is still a need to better understand the fundamental behavior of plant cell wall
biopolymers during alkaline biorefining processes and how their properties are affected by the

35

biorefining processing conditions. Laboratory analytical methods for hemicellulose isolation
involve, for example, alkali dissolution followed by alcohol precipitation (2-propanol, methanol,
or ethanol) with oxidative bleaching of contaminating lignin in the precipitate (Glasser, Kaar,
Jain, & Sealey, 2000; Puls, Schroder, Stein, Janzon, & Saake, 2006), while existing approaches
to lignin recovery from alkaline black liquors are based on lowering of the pH with an acidifying
reagent such as CO2, acetic acid, or H2SO4 (Ohman, 2006) to decrease the solubility of lignin by
increasing aggregation. Sugar separation may be desirable for the for the purpose of removing
sugars from the liquors which may be ultimately combusted and recovered for reuse in the
process.
This work proposes to better characterize the alkaline extractability, solubility, and
properties of the biopolymers (particularly the hemicelluloses) solubilized by alkaline conditions
that may ultimately integrate this extraction into either hardwood alkaline pulping, or grass or
hardwood alkaline pretreatments to convert polysaccharides through a biochemical platform.
Within this scope, this work has three specific aims. These are to: (1) determine the extractability
of cell wall polymers from diverse hardwoods and a representative grass by increasing alkali, (2)
determine the recoverability of these cell wall polymers by solubility-based separations, and (3)
quantify the recovered biopolymer fractions in terms of composition and molecular weight
distributions. In support of the third aim, analytical methods are adapted and compared for
estimating molecular weight distributions of alkali-extracted hemicelluloses.

36

2.2 Materials and Methods
2.2.1 Materials and Biomass Composition
Four species of hardwoods were tested along with one cultivar of switchgrass (Panicum
virgatum, cv. Cave-In-Rock) obtained from Dr. Farzaneh Teymouri (Michigan Biotechnology
Institute, Lansing, MI). The hardwood feedstocks were Acer saccharum (sugar maple) obtained
from Todd Smith (Devereux Sawmill, Pewamo, MI), Betula pendula (silver birch) obtained from
Curt Lindström (Smurfit-Kappa Kraftliner AB, Piteå, Sweden) and hybrid Populus spp.
including Populus tremula x tremuloides (hybrid aspen) and Populus nigra x maximowiczii cv.
NM6 (hybrid poplar) obtained from Dr. Raymond Miller (Michigan State University Extension).
Debarked wood chips were dried to approximately 5% moisture, milled (Wiley Mini-Mill,
Thomas Scientific) to pass a 2 mm screen, and stored sealed at 4°C until use.
Biomass composition was determined according to NREL/TP 510-42618 (Sluiter et al., 2008)
except that the procedure was scaled down by a factor of three and that HPLC quantification of
monosaccharides was performed with a Bio-Rad Aminex HPX-87H column using 5 mM H2SO4
as the mobile phase and RI detection. Ash and extractives were quantified according to
NREL/TP 510-42622 (Sluiter, Hames, et al., 2005) and NREL/TP 510-42619 (Sluiter, Ruiz,
Scarlata, Sluiter, & Templeton, 2005), respectively. Commercial GXs (Sigma-Aldrich #X4252
and X0502) which are derived from alkali-extracted, bleached Fagus sylvatica) (European
beech) and Betula pendula (silver birch) were utilized as model xylans for testing SEC and the
reducing end assay.

37

2.2.2 NaOH Extraction and Analysis of Extracts
Alkaline extractions were performed using, 1 g of biomass (dry basis) at NaOH
concentrations ranging from 0 g/L to 75 g/L at a total liquid volume of 20 mL (L:W of 20) in 50
mL disposable centrifuge tubes. These were incubated in a water bath at 85˚C for one hour with
this temperature chosen as a compromise between potential polysaccharide degradation versus
improved extraction. Immediately after extraction the centrifuge tubes were cooled in ice water
for 10 minutes followed by centrifugation (Eppendorf 5804R) at 4,500 x g for 10 minutes. The
liquid extract was decanted and saved for further analysis. The insoluble solids were resuspended
in water and washed by repeated centrifugation and decanting until neutral. The samples were
then transferred to pre-dried aluminum dishes, followed by drying overnight at 105˚C to
determine the total amount of material removed by the extraction.
Analysis of the liquid phase consisted of two parts. First the procedure NREL/TP 51042623 was implemented to assess the amount of dissolved polysaccharide material in the
alkaline extract (Sluiter et al., 2006). A small sample of extract was withdrawn and placed into a
pressure tube followed by the addition of 72% (w/w) H2SO4 to yield a 4% H2SO4 solution for
hydrolysis. The calculation provided by the procedure to yield a 4% acid solution was designed
for extracts from a dilute acid pretreatment of biomass. This calculation was modified slightly to
account for the high concentration of OH- in solution. After the hydrolysis procedure was
complete the liquid hydrolysate was separated from precipitated insoluble solids by
centrifugation. The hydrolysate was syringed filtered and analyzed for sugars by HPLC as
outlined in section 2.2.1. The precipitated solids were washed until neutral, dried overnight at
105°C, and weighed to determine Klason lignin content. The second analysis performed on the
alkaline liquor extract was a 2:1 (v:v) ethanol precipitation for the recovery of the dissolved
38

biopolymers. Two parts ethanol were added to one part of sample in a centrifuge tube and
allowed to precipitate for two hours at -4°C. The precipitate that was formed was washed with
95% (v/v) ethanol by repeated cycles of centrifugation at 2,885 x g for 10 minutes followed by
decanting. Gravimetric analysis was performed on the precipitate after washing to account for
the total amount of polymeric material that could be recovered from a given extraction.
Acid precipitation was performed on the liquid phase samples after NaOH extraction with
a 1 M H2SO4 solution for pH adjustment. Samples of the liquid extract were titrated to either a
pH of 2 or 6. After titrating the samples were centrifuged at 2,885 x g for 10 minutes followed by
decanting the supernatant. The precipitated material was centrifuged washed twice with 15 mL
of 5 mM H2SO4 followed by two washes with 15 mL of ethanol. These solvents were selected
for washing instead of water so as to not lose any material due to dissolution. After washing
samples were transferred to pre-dried and weighed aluminum weigh dishes and dried overnight
at 105°C. Gravimetric analysis followed to determine the amount of material recovered by pH
adjustment.

2.2.3 Characterization of Biopolymers
After washing, the precipitate was dried in an incubator at 65°C to drive off most of the
remaining ethanol without complete drying since the solubility properties were found to be
significantly reduced upon complete drying. The solids content of the dried precipitates were
determined and samples were suspended in a known volume of water and heated in a water bath
at 50°C until re-solubilized (approximately 3 hours).
NREL/TP 510-42623 was used to quantify the amount of polysaccharides associated with
the recovered biopolymer precipitates (Sluiter et al., 2006). The pH of the solution was measured
39

so the proper amount of 72% (w/w) H2SO4 could be added to yield a 4% (w/w) H2SO4 solution
for hydrolysis. After completion of the hydrolysis procedure samples were allowed to cool down,
and then centrifuged to remove any insoluble material that could have precipitated out of
solution upon addition of the acid. The liquid hydrolysate was decanted and syringed filtered
prior to analysis by HPLC. If any insoluble material was present the solid material was washed
and dried for gravimetric analysis.
Size exclusion chromatography (SEC) was performed using a Waters (Milford, MA)
Ultrahydrogel 250 column with elution in an aqueous mobile phase containing 0.1 M NaNO3
and 0.01 M NaOH with detection by RI and UV at 280 nm. Prior to analysis the re-solubilized
samples were mixed in a 1:1 with 0.2 M NaNO3 and 0.02 M NaOH. The column was calibrated
with dextran standards (Sigma-Aldrich) consisting of molecular weights of 1,000, 5,000, 12,000,
25,000, 50,000, and 80,000 g/mol. For the final comparisons concerning the size of the polymers
the molecular weights of the dextran standards were converted to degree of polymerization (DP)
using the conversion factor of 132 g/mol monomer subunit for a xylan polymer.
The number average degree of polymerization (DPN) of the recovered material was
determined using a bicinchoninic acid (BCA) assay, which was performed according to Zhang
and Lynd (Zhang & Lynd, 2005) only using xylose in the range of 5-60 µM as the standard. The
monomer concentration of xylose was determined by the same 4% acid hydrolysis method
discussed when determining the composition of the liquid phase extract and precipitate,
including correction for xylose dehydration. The ratio of the xylose monomer concentration and
reducing end concentration give an estimate of the apparent DPN of the recovered
polysaccharides assuming no interference by other polymers.
40

2.2.4 Precipitate Bleaching
For the bleaching procedure, 10 g/L of the precipitate was suspended in a solution of 2%
(w/v) H2O2 and 1.0 mM NaOH solution. This was incubated for one day at room temperature,
after which ethanol was added in a 2:1 ratio the flask was placed in a freezer (-4°C) for 2 hours to
precipitate the bleached polymers. The bleached material was washed with ethanol by
centrifugation at 2,885 x g for five minutes during each cycle. After washing the bleached
material was dried in the similar manner stated above for the original precipitated material. After
drying the bleached material was resuspended in water and the previously described analyses
were performed for composition by 4% acid hydrolysis, apparent 𝐷𝑃! by BCA, and molecular
weight distribution by SEC.

2.3 Results and Discussion
2.3.1 Biomass Composition
The original composition of the biomass used in this work is presented in Table 2.1.
Standard deviations were lower than 5% after triplicate measurements for the sugar and lignin
content. Arabinan was not detected (N/D). This table shows that the glucan fraction
(overwhelmingly cellulose) represents the major component by weight and is relatively unaltered
by the extraction. A small extractable fraction of this glucan may derive from starch, sucrose,
mixed-linkage β-glucans, xyloglucans, or glucomannans. The hemicellulose fraction is
characterized as arabinan and xylan + mannan + galactan since the analytical method does
distinguish between these sugars. The majority of the hemicellulose sugars should be xylose
since it is well established that the xylans are the predominant hemicelluloses in woody dicots

41

and grasses. The term “hemicellulose” will be used throughout this text as up to 10% of this
fraction may include GMs and GGMs.
Table 2.1. Composition of biomass as a percentage of the dry weight; N/D = not detected, NQ =
not quantified.

Sugar
Maple
Silver
Birch
Hybrid
Aspen
Hybrid
Poplar
Switchgrass

Glucan

Xyl/Man/Gal

Ara

Klason
Lignin

42.4

20.7

N/D

24.6

3.4

39.3

25.3

N/D

16.6

31.9

16.9

1.3

42.3

18.3

35.0

26.6

Acetyl Ash

Extractives

NQ

0.5

3.5

4.9

4.6

0.2

2.8

11.2

26.0

4.8

2.9

12.0

4.2

N/D

23.8

3.0

1.7

3.5

7.5

2.0

16.7

2.5

4.3

5.3

7.6

The birch and switchgrass had the highest fraction of hemicellulose sugars, while the birch and
aspen had the most acetylated xylans. The maple and aspen had the highest fraction of Klason
lignin in their cell walls and the switchgrass was particularly high in ash, perhaps due to the
incorporation of silica into the cell walls of grasses rather than calcium in the case of dicots
(Carpita, 1996). The birch and aspen had more than 14% total extractives plus “not quantified”
material that would likely be extracted by alkali. Comparatively switchgrass contains a large
fraction of “not quantified” material, which includes uronic acids originating in pectin or as
glucuronsyl substitutions on the xylan and possibly protein.

2.3.2 Extraction and Recovery
After the series of NaOH extractions were performed, the alkali extraction was quantified
by total material lost from the cell walls (Figure 2.1) and the total recoverable plant cell wall
polymers were quantified by the mass recovered by alcohol precipitation (Figure 2.2). Data are
42

not presented on an extractives-free basis and error bars indicate data range from duplicate
extractions. The data plotted in Figure 2.1 show that increasing NaOH loadings on the biomass
had the obvious effect of increasing the amount of material solubilized from the cell walls and
that the most material is removed from switchgrass and aspen while maple is the most resistant
to alkali extraction. More material is removed from the aspen and switchgrass, which may be a
consequence of the higher extractive content in the biomass, 12% and 5.3%, respectively, along
with the contribution from the unquantified fraction.

Figure 2.1. Total mass extractability for the five biomass sources.
Specifically for switchgrass, it is well-established that grass cell wall xylans and lignins are
particularly susceptible to extraction by alkali with more than 50% typically extractable (Hartley,
1983) while higher temperature alkali pretreatments solubilize a comparable portion of the cell
wall (Xu et al., 2010). Factors that contribute to this are that alkali-labile ferulate-mediated ester
cross-links between lignin and GAX are a feature of the secondary cell walls of commelinid
monocots (Iiyama, Lam, & Stone, 1990).

43

Figure 2.2. Recoverability of the extracted polymers with a 2:1 (v:v) ethanol precipitation for the
five biomass sources.
Furthermore, grass lignins have significantly higher free phenolic contents than dicot lignins
which gives them a higher solubility in alkali (Lapierre, Jouin, & Monties, 1989). Contributing
factors to cell wall polymer extractability of the woody dicots tested include differences in the
extractives abundance, secondary cell wall thickness, and vascular structure density between
aspens and poplars that have bulk densities 50% lower than maples and birches (Simpson &
TenWolde, 2007), although these are not measured for this work.
The data plotted in Figure 2.2 show the recovery profiles for the selected biomass
feedstocks. Because the alcohol-precipitated material is presumed to represent primarily
solubilized cell wall polymers, this is used to set an upper limit on polymer recoverability. On a
mass basis, switchgrass generates the largest amount of recoverable material, while the
hardwood samples produce less, albeit, with a similar trend amongst the four. On a percentage
basis sugar maple gives close to full recovery of the material removed by NaOH extraction
44

indicating that the material extracted is primarily polymeric. The other feedstocks perform lower
on an overall amount recovered basis with switchgrass recoveries of around 80%, birch 67%,
poplar 60%, and aspen yielding the lowest recovery of extracted material at 37.5%. Differences
in the recovery can be attributed to differences in the abundance of low MW organics solubilized
during alkali extraction (e.g. low DP polymers, extractives, acetic acid).
The total amount of the hemicelluloses solubilized by NaOH is plotted in Figure 2.3,
which shows that increasing alkali concentration increases hemicellulose extraction.

Figure 2.3. Hemicellulose extractability by alkali for the five biomass sources.
The amount of hemicellulose extracted from sugar maple, hybrid poplar, and aspen show very
similar trends, while the hemicellulose extracted from birch and switchgrass are significantly
higher. This is potentially due to combination of factors including the higher content of overall
hemicellulose in these two plants (Table 2.1), the lower lignin content which may make for a
more accessibly secondary cell wall, and the aforementioned properties of grass cell wall
polymers that increase their alkali extractability. Once again the switchgrass outperforms the
other feedstocks by removing up to 60% of the hemicellulose present based on the composition
45

in Table 2.1. Interestingly the birch sample produces the second greatest fraction of
hemicellulose removed. As previously discussed the aspen feedstock produced the most material
solubilized from the hardwood samples as shown in Figure 2.1, however, this does not
specifically indicate that hemicellulose components are extracted in the greatest quantity from
aspen. The larger component of extractives in the aspen sample, 12% as shown in Table 2.1, can
presumably explain this factor.
Overall, important properties that may affect plant cell wall polysaccharides in solution
could be proposed to include (1) the polymer size, (2) both the degree of substitution of the
polymer backbone and the properties (e.g. ionizable groups) of the substituent (hydroxyl, acetyl,
glucuronosyl, or arabinosyl), (3) the properties of the solvent (including the pH), and (4) the
potential for aggregation with itself or with other polymers. It is known that gradient extraction
at increasing alkali concentrations (Carpita, 1983; Carpita & Whittern, 1986; Peng et al., 2010)
or gradient recovery with solvents of at different strengths or with different properties (Bian et
al., 2010) can yield hemicelluloses with different properties. For example, graded alkali
extraction of maize coleoptiles (Carpita & Whittern, 1986) identified 3 GAX fractions that
differed in the level of substitution and composition, with more substituted GAX going into
solution first, and mixed-linkage β-glucans going into solution at higher alkali (2-3 M KOH) due
to closer association with cellulose microfibrils. Successive extractions with alkali on isolated
holocellulose have shown decreases in molecular weight starting at concentration levels of 1M
KOH (Peng et al., 2010).
Data from Figures 2.1, 2.2, and 2.3 are replotted in Figures 2.4 and 2.5 to show the
amount of hemicellulose extracted versus the overall amount of material removed (Figure 2.4) or
the yield of hemicellulose in the precipitated polymers versus NaOH loading (Figure 2.5). The

46

offset in Figure 2.4 is presumably due to differences in amount of “extractives” that are likely to
be easily solubilized during alkali extraction. In Figure 2.5 the hemicellulose content of the
precipitated polymers shows a comparatively high fraction, around 80%, present in the birch
precipitate while the other feedstocks all have a similarly lower composition.

Figure 2.4. Relationship between hemicellulose extracted to the total amount of alkali extracted
material.
A higher fraction of hemicellulose can be obtained after a bleaching of the maple precipitate as
indicated in Figure 2.5. Accordingly this shows that although the precipitates are not comprised
of pure polysaccharides, they can undergo further processing to obtain a relatively pure
polysaccharide component.
Figures 2.6, 2.7, and 2.8 compare the recovery of plant cell wall polymers as a function
of extraction condition for the sugar maple, aspen, and switchgrass, repspectively, based on
precipitation with 2:1 (v:v) ethanol, acidification to pH 6, or acidification to pH 2. The obvious
trend is that alcohol precipitation yields the highest recovery of material, followed by
precipitation at pH 6 and pH 2. Another clear observation from this plot is that the maple extracts
(Figure 2.6) were significantly more soluble at acidic pH than extracts from the other biomass
47

types which may be due to differences in the degree of substitution by glucuronic acid, although
this was not quantified.

Figure 2.5. Quantified hemicellulose content of the cell wall polymers recoverable by ethanol
precipitation.

Figure 2.6. Recovery of plant cell wall polymers from sugar maple by ethanol or pH
precipitation.
48

Figure 2.7. Recovery of plant cell wall polymers from aspen by ethanol or pH precipitation.

Figure 2.8. Recovery of plant cell wall polymers from switchgrass by ethanol or pH
precipitation.

49

2.3.3 Chemical and Physical Characterization
Subsequent polymer size characterization of the extracted and recovered biopolymers
was performed using size exclusion chromatography (SEC) in addition to a method based on
quantification of polysaccharide reducing ends. The use of SEC presents challenges for
quantifying the molar mass distributions of mixtures of non-cellulosic heteropolysaccharides
from the cell wall. Qualitative analysis of MW distributions by SEC requires the implicit
assumptions that all polymers will yield a signal that is proportional to the polymer mass, that the
dextran calibration standards translate to hemicellulose polymer sizes, that size effects are the
only contribution to the separation, and that the polymers do not interact. Illustrating this, Saake
et al. (Saake, Kruse, & Puls, 2001) determined that calibration of SEC separations with pullulan
standards gives a higher apparent molar masses of xylans than when an in-line viscometer is
used. The negative charges on glucuronosyl substitutions may be strongly repelled by the anionic
stationary phase (hydroxylated methacrylate) in the column resulting in these polyanionic
polymers eluting first. However, it has been shown that 0.1 M NaNO3 used in the mobile phase
is sufficient to overcome these effects in aqueous SEC of hemicelluloses (Ponder & Richards,
1997a).
Summaries of some of the SEC results are presented in the following figures. Figure 2.9
shows an SEC elution profile for an alkali-extracted, bleached birch GX (Sigma-Aldrich X0502)
with the elution times for dextran MW standards plotted for reference. Without sonication a
broad polymer distribution was apparent between approximately 11-14 minutes elution time
which corresponds to a range of degrees of polymerization of between 70-1000 assuming
primarily β-1,4 linked Xylp subunits (132 g/mol monomer) and an additional large peak eluted at

50

approximately 10 minutes representing all the polymers above the exclusion limit for the column
(which is approximately 100 kDa according to the manufacturer).

Figure 2.9. Size exclusion chromatography elution profiles for commercial birch glucuronoxylan
with elution times for dextran standards plotted for reference.

When multiple columns were run in series (Waters Ultrahydrogel 500 and 250), increasing the
exclusion limit to approximately 500 kDa, a second large peak with an apparent DP of greater
than 1000 was still present as shown in the Appendix on page 60, Figure A.1. This peak is
proposed to represent aggregates of hemicellulose, which are known to form tight hydrogen
bonding networks or junction zones between deacetylated chains which are especially resistant to
decrystallization (Ebringerova, Hromadkova, Burchard, Dolega, & Vorwerg, 1994). Saake et al.
(Saake et al., 2001) found that KOH extraction of hardwood GX resulted in SEC chromatograms
with a similar bimodal distribution with the higher MW fraction possibly corresponding to
aggregates of deacetylated polymers in an aqueous mobile phase. They also found that DMSO
51

extraction of xylan (which does not deacetylate the polymer) did not result in the high MW peak.
Although DMSO may help in reducing aggregation, the chemical was not used as a solvent for
SEC analysis because the column set chosen could only handle up to a 20% composition (v:v) of
an organic solvent. In the work by Saake et al. (Saake et al., 2001) an organic phase composition
of 90% (v:v) was needed to disrupt the peak of high molecular weight components while a 20%
(v:v) organic solvent composition did not suffice. Sonication of samples prior to analysis was
found to disrupt these aggregates somewhat and allow for better separation (Figure 9); although,
it is anticipated that controlling aggregation may remain a problem and inflating the apparent
molecular weight. Moreover, Figure A.2 in the Appendix on page 60 shows the sonication of
beech and birch GX followed by the reducing end analysis indicating that the concentration of
reducing ends did not increase with sonication time; thus, sonication did not contribute to
depolymerization and instead only to aggregate disruption.
The elution behavior of ethanol-precipitated polymers from the maple extracted with
varying levels of alkali is presented in Figure 2.10. The chromatogram is bounded on either side
by a low MW peak eluting at approximately 15.5 minutes representing salts, solvent, or other
material below the permeation limit and a high MW peak eluting at 10 minutes representing
polymer aggregates above the exclusion limit. Between these two bounds there are two distinct
peaks present. The peak at 14 minutes can be removed by H2O2 bleaching (Figure 2.10) and is
the only peak giving a strong UV absorbance at 280 nm as shown in the Appendix on page 61 ,
Figure A.3. This peak decreases with respect to the peak at 11.5 minutes with increasing alkali
extraction.

52

Figure 2.10. Size exclusion chromatography elution profiles of alkali-extracted, ethanol
precipitated biopolymers from sugar maple with increasing alkalinity and bleaching.
In the Appendix on page 61, Figure A.4, the lignin to hemicellulose ratio of the precipitates
estimated from the composition analysis correlated very well to the ratio of the peak heights of
the peak at 14 minutes to the peak at 11.5 minutes.
Bleaching was found to increase the fraction of neutral sugars in recovered precipitates
from 20-50% (w/w) to 60-80% for a range of biomass types and extraction conditions (Figure
2.5). All unbleached extracts (aspen, birch, poplar, and switchgrass) gave the same elution
behavior as the maple with two major peaks at the same times representing presumably
hemicellulose or hemicellulose aggregates and lignin or LCCs and showing the trend that the
polysaccharide peak increases relative to the aromatic peak with increasing alkalinity during the
extraction.
In order to provide an additional independent estimate of the recovered polysacchide
DPN, an assay was adapted to estimate the number of free anomeric carbons on a polysaccharide
chain as a function of extraction condition and biomass source. Analytical methods that quantify
53

the anomeric or “reducing ends” of the polysaccharide chain have been used for estimating the
DPN of hardwood glucuronoxylans (Johansson & Samuelson, 1977; Sturgeon, 1973), cellulose
(Kongruang, Han, Breton, & Penner, 2004; Zhang & Lynd, 2005), cellulose derivatives
(Vlasenko, Ryan, Shoemaker, & Shoemaker, 1998), and pectins (Barford, Magidman, Phillips, &
Fishman, 1986; Pitifer, McLellan, & Vanburen, 1994), among others. Since the (4-O-Me)-GlcAp
subunits are known to be substituted onto the polymer backbone in GX and GAX as an α-1,2
glycosidic bond (Teleman et al., 2002) and the Araf subunits are attached to GAX and AX with
α-1,2 and/or α-1,3 linkages (Ebringerova, 2006), there should only be one unique anomeric
carbon per polymer chain in the xylans. The results for the DPN estimated with this method are
plotted in Figure 2.11 with all of the recovered cell wall biopolymer precipitates showing similar
profiles where the DPN increased until a threshold value was reached. This increase in the
apparent degree of polymerization with increasing alkali could be interpreted as increasing
extraction of higher MW hemicelluloses due to improved solubility at increasing pH.
A subsequent alkaline-H2O2 bleaching stage was performed on the maple precipitates in
order to decrease possible interference in the assay from non-polysaccharides in the recovered
biopolymer precipitates. This bleaching was found to significantly increase the measured DPN
(Figure 2.11) from a range of 50-60 for unbleached maple to 120-150 for the higher NaOH
extraction conditions. Commercial alkali-extracted and bleached GXs from beech (SigmaAldrich X4252) and birch GXs (Sigma-Aldrich X0502) were also analyzed this method and gave
DPN values of 156 and 123, respectively.

54

Figure 2.11. Apparent number average DP of recovered precipitates estimated using the BCA
reducing end assay.

DPN values of alkali-extracted hardwood GX in the range of 100-150 have been determined by
SEC MALDI-MS (Jacobs & Dahlman, 2001), by terminal reducing end analysis (Sturgeon,
1973), by a number of end group analysis methods (Johansson & Samuelson, 1977), viscometry
and sedimentation (LeBel et al., 1963) in the range of 180-215 using viscometry for birch
(Goring & Timell, 1960) and aspen (Koshijima et al., 1965) indicating that this method may
provide a more accurate estimate for actual polymer size.
Possible reasons that the bleaching step increases the apparent DPN include any one or
combination of proposed mechanisms. One mechanism may be due to interference non-glycans
in the precipitates. Since the BCA assay is based on quantifying the stoichiometric reduction of
2+

Cu

+

to Cu by the carbonyl group in a terminal aldose on a polysaccharide, polymers like

proteins (which are also known to participate in this reduction) and aldehydes present in lignin
55

may also contribute to this reduction. Other Cu-reducing compounds would have the net effect of
increasing the apparent number of polymer chains quantified and decrease the quantified value
for the DPN. If this is the case, then the observed trend in Figure 2.11, which shows the apparent
DPN increasing with increasing alkali, may be due to the increasing hemicellulose-to-lignin ratio
whereby the impact of non-polysaccharides would be less at higher alkali extraction conditions.
Another possibility is that the bleaching stage results in destruction of the reducing capacity of
the polysaccharide end group due to modification by the oxidant (Potthast, Rosenau, & Kosma,
2006) or by alkali (Johansson & Samuelson, 1977) which would also have the net effect of
decreasing the quantifiable reducing ends and increasing the apparent DPN.

2.4 Conclusions
Overall, the quantified amount of lignin and hemicellulose extractability was dependent
on the biomass source, and that increasing alkali solubilized an increasing fraction of the
hemicellulose. The alcohol precipitation recovered most of the extracted hemicellulose, while
acidification removed significantly less and performed particularly poorly for the maple.
Precipitates for most of the biomass sources and extraction conditions contained 30-50% (w/w)
hemicellulose, which was increased to 65-80% with alkaline-H2O2 bleaching. A notable
exception was the birch extracts, which contained more than 70% hemicellulose without
bleaching.
Size exclusion chromatography characterization of the recovered biopolymers indicated a
low-MW lignin or lignin carbohydrate complex fraction estimated in the range of 3-12 kDa as
calibrated by dextran molar mass standards. This peak could be removed by alkaline-H2O2
56

bleaching. The SEC separation of the polysaccharides exhibited a bimodal distribution with one
peak in the DP range of 100-1000 and a very high MW peak at estimated DP values of greater
than 1000 representing xylan aggregates. This aggregate peak could be removed by sonication.
The results from the reducing end assay estimated apparent DPN values that were significantly
lower than these obtained by SEC and showed increasing DPN values with increasing alkali that
approached asymptotic limits ranging from about 25 for the switchgrass to 120 for the birch. It
was found that bleaching these precipitates increased the asymptotic DPN value from
approximately 80 for the maple to 160 indicating potential interference due to other nonpolysaccharides or due to oxidation of the anomeric end of the polysaccharide.
The significance of these results is that it was demonstrated the extractability and
recoverability of alkali-solubilized hemicelulloses from plant cell walls which may have
important implications for identifying high-yielding conditions for hemicellulose extraction prior
to pulping or total sugar recovery for alkaline pretreatments. One important finding was the
identification that the hemicelluloses in silver birch (Betula pendula), which is the most
important commercial hardwood in Scandinavia and Russia for papermaking, were exceptionally
well suited for extraction and yielded high recoveries and precipitates that contained a low
content of contaminating lignin. It was affirmed that monocot hemicelluloses (and lignins) are
highly susceptible to alkaline extraction due to the unique properties and features of their cell
walls, their lignins, and potentially differences in cell wall polymer crosslinking. Additionally,
this work demonstrated the potential for quantifying the aggregation of alkali-extracted
hemicelluloses. This understanding of aggregation should be useful for identifying conditions
that encourage polymer self-association and could be important for developing solubility-based

57

recovery processes to separate polysaccharides (and lignins) from alkali pulping or pretreatment
liquors.

58

APPENDIX

59

APPENDIX

Column Set
Permeation Limit
Column Set
Exclusion Limit

Xylan Aggregates

Figure A.1. Estimated degree of polymerization distribution of unsonicated Sigma-Aldrich birch
xylan using Waters Ultrahydrogel 250+500 column set in series.

Figure A.2. Concentration of reducing ends in solution from Sigma-Aldrich beech and birch
xylan sonicated for up to 20 minutes.

60

Figure A.3. UV absorbance at 280 nm of alkali-extracted, ethanol precipitated biopolymers for
aspen following elution by size exclusion chromatography.

Figure A.4. Correlation of quantified lignin-to-hemicellulose ratio to the ratio of the two SEC
peaks (UV Peak Height to RI Peak Height) proposed to represent aromatics and polysaccharides.

61

CHAPTER 3. CORRELATING LIGNIN STRUCTURAL FEATURES TO PHASE
PARTITIONING BEHAVIOR IN A NOVEAL AQUEOUS FRACTIONATION OF
SOFTWOOD KRAFT BLACK LIQUOR

This work has been published as original research on 16th August 2013, in Green Chemistry,
Volume 15, and Pages 2904 – 2912.

3.1 Introduction
It is generally accepted that the sustainable production of heat, power, fuels, and
chemicals is imperative for the long-term economic stability and overall human welfare (B. E.
Dale & Ong, 2012). The production of fuels and chemicals from renewable feedstocks such as
plant biomass is one such route for achieving this goal of long-term sustainability. One such
8

renewable feedstock, kraft lignin, is currently produced at the scale of over 6.3x10 kg of kraft
lignin annually as a co-product of the chemical pulping of wood (Vishtal & Kraslawski, 2011)
with the majority of this lignin burned in recovery boilers for the production of process heat and
power. Lignin offers the potential to be a renewable feedstock for solid and liquid fuels,
chemicals, and materials, and while the recovery of lignosulfonates for products from sulfite
pulping is widely practiced, the recovery of lignin from kraft black liquors is practiced only on a
limited scale. Currently, products from polymeric lignin in kraft black liquors are limited to the
DMS, DMSO, DMSO2 family of solvents (Calvo-Flores & Dobado, 2010), lignin precipitated by
acidification for use as a solid fuel (Loutfi et al., 1991; Nagy et al., 2010; Tomani, 2010), and
lignin as an emulsifier to improve the dispersion of fillers in industrial emulsions such as cement
and asphalt (Lora & Glasser, 2002; Wei, Yang, Yang, & Wang, 2012). Additional proposed
62

products from kraft lignin include synthesis gas for the production of biofuels (Francis, Shin,
Omori, Amidon, & Blain, 2006), phenolic resins, epoxy resins, and adhesives for wood
composites (Stewart, 2008), sequestrants, adsorbents, aromatic monomers (Holladay et al., 2007;
Lora & Glasser, 2002), and low-cost lignin-synthetic polymer blend fibers to produce carbon
fibers (Suhas, Carrott, & Riberio Carrott, 2007). In addition to lignins from sulfite and kraft
lignins, lignin streams will become available from the deployment of cellulosic biofuels
technologies. With their diverse properties, lignins may have value for applications other than
heat and power, including materials, fuels, and chemicals.
As a consequence of the random nature of polymerization to form lignin, the
characterization of polymeric lignin is challenging, as individual lignin polymers may exhibit a
diverse distribution of properties, structures, and chemical functionalities, with quantified
properties representing average bulk values. Although recent work has been directed at
characterizing individual polymers (Morreel et al., 2010), solvent and pH-based fractionation of
lignins, by differences in these properties, is often applied in order to characterize the abundance
and distribution of structural features in fractionated subsets of lignin polymers (Villaverde, Li,
Ek, Ligero, & de Vega, 2009). Besides fractionation for the purpose of characterization,
industrially it may also be desirable to generate lignin fractions that are more suitable for a given
application, with properties within a desired tolerance. Important lignin properties for polymer
applications include the phenolic hydroxyl content in non-syringyl monomers for phenolic resins
(Froass, Ragauskas, & Jiang, 1996; Robert, Bardet, Gellerstedt, & Lindfors, 1984), thermal
properties such as a narrow melting-point range for carbon-fiber applications (H. Chum et al.,
1989; Elder & Soltes, 1981; Monteil-Rivera, Phuong, Ye, Halasz, & Hawari, 2013), and aliphatic
hydroxyl groups and even phenolic hydroxyl groups that under certain reaction conditions have

63

been identified as carbamylation sites to produce polyurethanes from lignin (Yelle, Ralph, &
Frihart, 2011).
Alkali-solubilized lignins are known to undergo increasing aggregation with decreasing
pH due to the increasing protonation of ionizable groups in lignin (Gidh et al., 2006; Guerra et
al., 2007; Ragnar et al., 2000; Sarkanen et al., 1981). This phenomenon has been exploited by a
number of approaches to lignin recovery from alkaline black liquors, in which the pH is
decreased with an acidifying reagent such as CO2, acetic acid, or H2SO4 (Loutfi et al., 1991;
Ohman, 2006) in order to effect a precipitation of lignin from solution. The resultant lignin
precipitate is then recovered by filtration (Jönsson et al., 2008; Mänttäri et al., 2013). This
acidification-filtration approach is currently being performed on a limited commercial scale to
generate both conventional lignins from the kraft black liquors of hardwoods and softwoods
(Nagy et al., 2010; Tomani, 2010), as well as sulfur-free lignins from the black liquors of sodapulped flax (Lora & Glasser, 2002) and sugar cane bagasse (Nowakowski, Bridgwater, Elliott,
Meier, & de Wild, 2010). Besides precipitation, ultrafiltration and nanofiltration have also been
investigated as a strategy for the recovery of soluble lignin directly from the black liquor
(Keyoumu et al., 2004).
Researchers have recently developed and investigated a novel black liquor lignin
recovery technology whereby the black liquors are acidified with CO2 under elevated
temperatures and pressures and a portion of the lignin phase-separates from the bulk black liquor
solution as a suspension of solvated lignin, referred to as “liquid lignin”, which can be separated
without filtration (Lake & Blackburn, 2011; Velez & Thies, 2013). This lignin recovery
technology is proposed to have a number of positive process attributes including improved
energy economy and processability (Lake & Blackburn, 2011). To facilitate the characterization
64

of this liquid lignin phase, in previous work Velez and Thies (Velez & Thies, 2013) acidified an
alkaline kraft black liquor with CO2 as described above, but for their investigation the
acidification was done incrementally in pH units of ~0.5, such that the fraction of the liquid
lignin that separated out over a given pH range was isolated and quantified. The water content
and softening point of each fraction in its “hydrated” state at process operating conditions was
previously determined, as was the ash content of each fraction. However, no structural properties
of the fractions isolated from the liquid-lignin phase, such as molecular weight and the frequency
of ionizable groups on the lignin polymer such as aromatic hydroxyl (free phenolic), were
determined. This, then, was the subject of this study. In particular, the goal of this work was to
(1) characterize a number of the molecular properties for narrow pH-based fractions of softwood
lignin and (2) to relate the properties of these liquid-lignin fractions to their pH-driven, phasepartitioning behavior.

3.2 Material and Methods
3.2.1 Materials
As described above and reported in detail elsewhere (Velez & Thies, 2013), “liquid
lignin” fractions were phase-separated from an industrial softwood alkaline black liquor via a
unique pH-based fractionation technique, using CO2 as the acidifying reagent at elevated
temperatures and pressures (115°C and 6.2 bar). Upon acidification, these lignin fractions phaseseparated and were collected as either a continuous “liquid lignin” phase or a hydrated lignin
precipitate containing between 30-50% water by weight. Information about each of these
fractions, including the pH at the end of each fraction collection and the mass percentage of each
fraction relative to the total lignin recovery, is given in Table 3.1.
65

Table 3.1. Liquid-lignin fractions obtained by the unique pH-based fractionation process
(Velez & Thies, 2013) and used for this study.

Liquid-Lignin
Fraction
F1
F2
F3
F4
F5
F6
F7
a

a

Mass Percentage of Lignin
Recovered
Per Fraction (%)
Cumulative Recovery
(%)
0.8
0.8
0.6
1.4
3.4
4.8
31.8
36.6
36.2
72.8
22.6
95.4
4.6
100

Final pH Obtained Per
Fraction
12.8
12.1
11.6
11.1
10.6
10.0
9.5

Softwood Kraft black liquor with pH of 13.6 served as feed to the acidification process

These fractions were subsequently further acidified with 1N H2SO4 to a pH of 2.5 as suggested
by Compere and Griffith (Compere & Griffith, 2009). A schematic of this procedure is given in
Figure 3.1.

Figure 3.1. Schematic of the pH-based fractionation process for separation of liquid-lignin
fractions within narrow (~0.5) pH bands from kraft black liquor via acidification with CO2 at
elevated temperatures and pressures.
66

Note that fresh softwood kraft black liquor was used as the feed only for the first fractionation.
The partially spent black liquor from the first fractionation (i.e., the un-precipitated top phase
designated as Spent Black Liquor 1) served as the feed to the second fractionation, Spent Black
Liquor 2 from the second fractionation served as the feed to the third fractionation, etc. This
procedure was repeated to yield fractions at approximately every 0.5 pH increments, with early
fractions collected over a wider pH band due to the lower recovery yields. After vacuum-filtering
the fractions from the spent acid solution, the samples were air-dried and their ash content was
determined using NREL/TP-510-42622 (Sluiter, Hames, et al., 2005). Klason lignin and neutral
polysaccharide content were quantified by NREL/TP-510-42618 (Sluiter et al., 2008), and the
uronic acid content was assayed enzymatically with a D-Glucronic/D-Galacturonic Kit (Cat. No.
K-URONIC, Megazyme Intl. Ireland, Bray, IE).

3.2.2 Lignin acetylation
1

The lignin fractions were acetylated for analysis by H NMR and Gel Permeation
Chromatography (GPC) according to the acetylation procedure used by Gosselink et al.
(Gosselink et al., 2004), except that ethanol was used in place of methanol. First, approximately
400 mg of a lignin fraction was suspended in 8 mL of a 1:1 (v/v) solution of acetic anhydride and
pyridine. For lignin fractions F1 and F2 (see Table 3.1), only 80 mg of material was suspended in
1.5 mL of the acetylating mixture due to the limited availability of sample. The acetylation was
allowed to proceed for 24 hours at room temperature in a shaking incubator at 180 rpm. Then 30
mL of ethanol was added to each sample, followed by brief stirring and incubation for an
additional thirty minutes. Solvents were then removed by drying under nitrogen. This ethanol
washing and nitrogen-drying procedure was repeated three times to ensure a complete removal
67

of the acetic acid and pyridine. The remaining material was dissolved in 25 mL of CHCl3 and
then washed twice with 250 mL of deionized water using a separatory funnel. The CHCl3
solution was then added slowly to 75 mL of diethyl ether to precipitate out the acetylated lignin
fraction.

3.2.3 Proton and quantitative

13

C NMR Analysis

1

Quantitative H NMR was performed with an Agilent 500/54 Premium Shielded 500
MHz NMR equipped with a Varian 7600 Autosampler. A sample of lignin was dissolved at a
concentration of 44 mg/mL in deuterated chloroform containing pentaflurorbenzaldehyde (PFB)
as an internal standard and a trace amount of tetratmethylsilane (TMS) as a reference peak.
Samples were run using a 30° pulse with a 4 s pulse delay and a total of 800 scans. Integration
was performed using VnmrJ v. 3.2 software (Agilent Technologies, Santa Clara, CA) using the
integration regions for protons in acetylated aliphatic hydroxyl, acetylated phenolic hydroxyl,
and methoxyl groups with peak assignments according to Ralph et al (Ralph, Ralph, & Landucci,
2004). Figure B.1 in the Appendix on page 88 gives a reference to the peaks identified.
Quantitative

13

C NMR was performed on fractions F4-F7 in DMSO-d6 at a sample

concentration of 300 mg/mL using a 30° pulse angle with a recycle delay time of 2 s, an
acquisition time of 1.36 s, and an observed pulse of 3.4 µs using a minimum of 5000 scans. The
analyses included a decoupling mode to reduce the Nuclear Overhauser Enhancement (NOE).
Integration was done on the regions showing methoxyl functionality (δ = 54 – 57 ppm) and Cγ in
the β-O-4 structure (δ = 59 – 61 ppm), with the relative abundance of a β-O-4 taken as the ratio
of this peak to the methoxyl content based on the peak assignments of Ralph et al (Ralph et al.,
68

2004). Figure B.2 in the Appendix on page 88 gives a representative spectrum used to quantify
β-O-4 content. Other structures that contained a hydroxylated Cγ were presumed to have
chemical shifts far enough downfield (δ	
  = 61.8 - 64.4 ppm) from the Cγ in a β-O-4 (δ	
  = 60.0
ppm) that they would not overlap.

3.2.4 Gel permeation chromatography
Gel permeation chromatography (GPC) was performed on an Agilent 1100 series HPLC
equipped with a Waters Styragel HR 4 (Milford, MA, USA) column, using either a system
consisting of tetrahydrofuran (THF) at a 0.5 mL/ min flow rate and 40°C for acetylated lignin
samples or a system consisting of dimethylformamide (DMF) and 0.05 M LiCl at a flow rate of
0.5 mL/min and 80°C with detection by UV absorbance at 280 nm. Lignin samples were
dissolved in the mobile phase and syringe-filtered (0.45 µm PTFE membrane, Whatman
Puradisc Syringe Filters, GE Healthcare, Waukesha, WI) prior to analysis. Monodisperse
polystyrene standards (1.0, 10, 50, 100, 200, 600 kDa, Sigma-Aldrich, St. Louis, MO) were run
on each system for reference. Based on the reference elution times, number and weight average
molecular weights of the lignin fractions were obtained per the procedure given by Malawer and
Senak (Malawer & Senak, 2004).

3.2.5 Thermogravimetric analysis
Thermogravimetric analysis (CH-8603 model, Mettler-Toledo, Schwerzenbach, CH) was
employed to determine the mass loss during slow pyrolysis using nitrogen as the purge gas with
a cell and furnace flow rate of 20 mL/min and a temperature profile of 10°C/min. For each
sample, the furnace was initially purged with nitrogen for 10 min; then the sample was dried for
69

20 minutes at 105°C. Masses were normalized to the weight remaining after complete water
volatilization at 115°C. The normalized mass loss data was plotted as differential thermogravimetric curves from 115°C – 800°C.

3.2.6 Analytical pyrolysis
Analytical pyrolysis was performed as reported previously (Li et al., 2012) using a
microscale pyrolysis unit (CDS Pyroprobe 5250, CDS Analytical, Inc., Oxford, PA) interfaced to
a Shimadzu QP-5050A gas chromatograph/mass spectrometer (GC/MS; Shimadzu Corp,
Columbia, MD). Approximately 1 mg of lignin sample was packed between quartz wool in a
quartz tube with a filler rod. Three replicates of each sample were run. The pyroprobe was set at
a filament temperature of 650°C at a heating rate of 1000°C/s and a hold time of 20 s. The GC
was equipped with a Restek Rtx-1701 column (Restek, Bellefonte, PA), 60 m x 0.25 mm with a
0.25 µm film thickness. The column gas (helium) flow was 1 cm/s with a split ratio of 1:100 so
as to not overwhelm the mass spectrometer. The GC oven temperature program began with a 1min hold at 40°C followed by a heating rate of 8°C/min to 270°C. The injector and detector
temperatures were set to 280°C. The mass spectra were recorded in the electron ionization mode
for m/z from 33 to 400. Identification of compounds was performed by comparing the mass
spectra of the peaks with standard spectra of other compounds from the NIST library to obtain
the most probable matches. Pure compounds (Sigma-Aldrich Co., St Louis, MO) were then used
to confirm the identities of certain peaks based on matching of retention times and mass spectra.

70

3.2.7 Data correlation
Correlation coefficients for selected properties of lignin fractions F3 through F7 were
determined in MATLAB (MathWorks, Natick, MA), using a hierarchical ordering of correlation
coefficients according to the shortest Euclidian distance between clusters. Incomplete data for βO-4 content were handled by determining the correlations only between fractions F4-F7.

3.3 Results and Discussion
3.3.1 Lignin yields and non-lignin contaminants
Physical properties of the seven fractions isolated by our unique pH-based fractionation
process, using the methods briefly described in sections 3.1 and 3.2, and in detail elsewhere
(Velez & Thies, 2013) are given in Table 3.1. The pH at which the collection of a given fraction
was terminated is shown, and the mass percent of total lignin in each fraction is given on an ashfree basis. Results indicate that the majority of the lignin phase-separates between a pH of 10.0
and 11.6, corresponding to fractions F4, F5, and F6, which comprise more than 90% of the
precipitated material.
Non-lignin contaminants such as inorganics, polysaccharides, and extractives (e.g., resin
and fatty acids, sterols/terpenoids, and flavonoids/tannins) are detrimental in some lignin
applications (Compere & Griffith, 2009); for example, for carbon-fiber applications they can
contribute to an undesirably broad melting point (Compere et al., 2005). During black liquor
acidification, it is known that proteins, polysaccharides, and some extractives can precipitate out
at a higher pH than the bulk of the lignin (Compere & Griffith, 2009; Dai, Jameel, & Chang,
2006). Thus, this pH-based fractionation process could result in high levels of contaminants in
the first fractions to phase-separate. However, all samples were found to be comprised primarily

71

1

of lignin as determined by H NMR, which quantified comparable methoxyl contents (between
3.0 and 6.4 mmol/g lignin, see Appendix page 89, Figure B.3) for all samples. The neutral
polysaccharide content of fractions F3 through F7, along with the estimated bulk polysaccharide
content, are presented in Figure 3.2.

Figure 3.2. Quantified neutral polysaccharide content of fractions F3 through F7 with estimated
weight average polysaccharide content.
Sugar composition analysis was not performed on fractions F1 and F2 because of the limited
quantities of sample that were available. The weight-averaged content of neutral polysaccharides
in the precipitated lignin samples, expressed as a percentage of the total lignin precipitate from
the black liquor, was 2.91% by weight and consisted primarily of the polysaccharides contained
in GGMs and arabinoxylans. Fractions F3 and F4 are more highly enriched in polysaccharides,
particularly in glucan, relative to the overall, weight-averaged polysaccharide content, while
fractions F5 through F7 contain significantly less polysaccharides than the bulk. The differences
in glucan content between the fractions indicates that the glucomannans are more likely to
72

partition into the liquid-lignin phase at a higher pH than the xylans. The content of uronic acids
in the fractions was below the detection limit, although this may potentially be due to their
conversion to hexenuronic acids.
1

Using H NMR, the abundance of the aliphatic extractives (e.g., resin and fatty acids,
sterols/terpenoids) relative to lignin was estimated (Figure 3.3) based on integration of the
aliphatic -CH2- and -CH3 peaks (δ = 0.7-1.4 ppm).

F2	
  

F5	
  

F7	
  

F4	
  

F1	
  

F3	
  

F6	
  

Figure 3.3. Quantified aliphatic H to MeO ratio of fractions F1 through F7 as determined by 1H
NMR as a proxy for extractives-to-lignin ratio.
These are presumed to originate primarily from extractives, as polysaccharides and the aliphatic
side regions of lignin should not yield signals in this region with the exception of a terminal
methyl group. This integration is normalized to the molar methoxyl content based on integration
of the methoxyl peak (δ = 3.5-4.0 ppm), which is taken to represent a single lignin monomer
because (1) the overwhelming majority of the lignin is presumed to be comprised of guaiacyl

73

monomers containing a single methoxyl group and (2) demethoxylation is not considered to be
significant during kraft delignification (Pinto, Evtuguin, Neto, Silvestre, & Amado, 2002).
Assuming that terpenes and fatty acids have a mass-to-proton ratio of approximately 10 g/mol H,
while a guaiacyl monomer in lignin will have a molar mass of approximately 180 g/mol
monomer, the same data can be presented as an approximation on a mass ratio basis (see
secondary y-axis on Figure 3.3). These results show two notable trends: (1) fractions F1 and F2
contain two to three times the extractives than the other fractions (approximately 0.12 g aliphatic
extractives per g lignin versus 0.04-0.07 for the other fractions), and (2) the aliphatic extractivesto-lignin ratio in fractions F3 to F7 shows a slight increase with decreasing pH of fractionation.

3.3.2 Structural features of recovered lignin fractions
1

H NMR also quantified the relative abundance of aromatic and aliphatic hydroxyls in

acetylated lignin fractions (Figure 3.4). These results are presented on a mole hydroxyl per mole
methoxyl basis, which can be considered as a proxy for a per mole lignin monomer basis as
defined previously. Each of these sources of hydroxyl groups shows a clear trend for fractions F3
through F7, where the bulk of the material separates and contains the least non-lignin
contaminants. The trend for aliphatic hydroxyl content exhibits minimal variation between the
fractions, with an average value of 1.16 mol of OH mole per monomer, while the phenolic
hydroxyl content shows a nearly 50% increase with decreasing pH from 0.47 to 0.69 mol of OH
per monomer. This is a significant result, as it demonstrates that the aliphatic hydroxyl content of
a lignin species does not contribute (either directly or indirectly) to its tendency to phase-separate
from solution during pH-based fraction.

74

1

Figure 3.4. Phenolic hydroxyl and aliphatic hydroxyl content of lignin fractions by H NMR
(mol OH/mol lignin monomer).
Conversely, phenolic hydroxyl content is a strong predictor of the tendency of a lignin to phaseseparate from solution as a function of pH, either directly (i.e., by pH-dependent deprotonation)
or indirectly (e.g., as it relates to another property such as molecular weight, as phenolic
hydroxyl content increases with decreasing lignin molecular weight).
The relative content of C in a β-aryl ether with respect to the methoxyl content, again as
γ

a proxy for a mole of lignin monomer, was determined by quantitative

13

C NMR (Figure 3.5).

This result shows that the carbon content present in this configuration undergoes about a 40%
decrease with decreasing pH of fractionation, which may be a consequence of a decrease in the
overall molecular weight of the lignin.

75

F5	
  

F4	
  

F6	
  

F7	
  

Figure 3.5. Decrease in Cγ (mol Cγ/mol lignin monomer) with decreasing separation pH by

13

C

NMR.
1

Interestingly, the total molar aliphatic hydroxyl content as determined by H NMR (Figure 3.3)
is about an order of magnitude greater than the molar β-O-4 content as determined by

13

C NMR

(Figure 3.5). This can be interpreted as meaning the aliphatic hydroxyl groups are predominantly
contained on other lignin structures (e.g., terminal aliphatic hydroxyl or one of at least four other
linkages containing a hydroxylated Cγ).
3.3.3 Relative lignin molecular weight distributions by gel permeation chromatography (GPC)
Lignin self-aggregation is significant in aprotic solvents such as THF (Crestini, Melone,
Sette, & Saladino, 2011), necessitating the derivatization of hydroxyl groups, for example, by
acetylation before GPC analysis can be undertaken. Using acetylated lignin fractions and a
mobile phase of THF, estimates of number and weight average molecular weights were obtained
by GPC (Figure 3.6).
76

F4	
  

F3	
  

F1	
  

F5	
  
F6	
  

F2	
  

F7	
  

Figure 3.6. Estimated lignin molecular weights by GPC (acetylated lignin, THF mobile phase).

Column retention volumes were calibrated using monodisperse polystyrene standards. The use of
GPC as a tool for quantitatively determining lignin molecular weights is problematic in that
reported results and are strongly dependent on the solvent properties (e.g., pH, ionic strength,
etc.), lignin properties (e.g., the tendency for lignin to self-associate during analysis and interact
with the stationary phase in the column), and the detection method (Cathala, Saake, Faix, &
Monties, 2003; Contreras, Gaspar, Guerra, Lucia, & Argyropoulos, 2008; D. Dong, 1993;
Glasser, Dave, & Frazier, 1993; Gosselink et al., 2004; Guerra et al., 2007; Harton et al., 2012).
These results show that with decreasing pH, the molecular weight of the precipitated liquidlignin fractions decreases. Thus, the lowest MW lignin species are the last to phase-separate from
solution during pH-based fractionation.
The use of strong polar solvents such as DMF in conjunction with electrolytes such as
LiBr or LiCl can yield effective solubilization of underivatized lignin, as well as effective
77

separation which can be size-fractionated by GPC (Baumberger et al., 2007). However, it has
been noted that polystyrene elutes more slowly than comparably-sized lignin polymers in this
system, resulting in unusually high molecular weight estimates for lignin when using polystyrene
for the reference standards (H. L. Chum, Johnson, Tucker, & Himmel, 1987). This behavior was
confirmed by our GPC results for underivatized lignin in DMF + 0.05 M LiCl (Figure 3.7). It can
be observed that fractions F1 through F4 have a high molecular weight tail, while fractions F5
through F7 are less polydisperse. Comparing the lignin elution profiles, a decrease in molecular
weight is apparent with decreasing pH of fractionation for fractions F3 through F7 (i.e., the
retention times are shifted to higher elution volumes).

Figure 3.7. GPC elution profiles of nonacetylated lignin fractions in DMF + 0.05 M LiCl mobile
phase, referenced to polystyrene standards.

78

These results provide a second independent validation of the trends (but not the absolute values)
for the molecular weights obtained using the THF system (Figure 3.6) of decreasing molecular
weight with decreasing fractionation pH.

3.3.4 Thermogravimetric analysis of lignin fractions
Results from the thermogravimetric analysis (TGA) of fractions F3-F7 are presented in
Figure 3.8; TGA for the entire liquid-lignin phase (still recovered by acidification with CO2, but
not incrementally so as to produce fractions) is also included as a reference.

Figure 3.8. Lignin fraction mass loss by TGA.

The mass-loss curves (Figure 3.8) are normalized to include only the mass remaining after
complete water volatilization from the samples has occurred at 105-115°C. These results show
comparable mass volatilization profiles for all fractionated samples, with the mass fraction of the
original sample remaining being 0.421–0.445 when the temperature reaches 800°C. The
79

derivative curves for the rate of volatilization (-dm/dT) are presented in Figure 3.9 and exhibit
three distinct peaks, representing the loss of water (50-110°C), the loss of polysaccharides (200300°C), and loss of the bulk lignin (300-500°C) (Garcia-Perez, Chaala, Yang, & Roy, 2001).

Figure 3.9. First derivatives of mass loss curves for lignin fractions.

Minor differences in these three peaks were observed among the lignin fractions (F3–F7). The
water peak is associated with the sorbed water, and the thermal behavior of this water (i.e., the
heat of desorption of water in the heating range of 80-130°C) has been correlated to the phenolic
hydroxyl and carboxyl content of the lignin (Vasile, Popescu, Stoleriu, & Gosselink, 2006). The
peak associated with aliphatic hydroxyl groups, corresponding primarily to polysaccharides
(Garcia-Perez et al., 2001; Mousavioun & Doherty, 2010), is most prominent for fractions F3
and F4 (which both show a maximum rather than an inflection point as in the other samples) and
is in agreement with the composition results of Figure 3.2, which show the highest
80

polysaccharide content for these fractions. The mass-loss peak associated with the bulk of the
lignin is relatively unchanged between lignin fractions with the exception of fraction F5, which
exhibits a lower onset temperature. As would be expected, the derivative curve for the
unfractionated lignin bounds all of the derivative curves for the lignin fractions (Figure 3.9).

3.3.5 Analytical pyrolysis
The distribution and relative abundance of aromatic monomers generated from the lignin
fractions by analytical pyrolysis are presented in Figure 3.10. Aromatic compounds are grouped
as dimethoxybenzene or dimethoxyphenol (S), guaiacol (G), or phenol (Ph) monomers
containing one (C1), two (C2), or three (C3) additional carbons, with abundances being defined
as a percentage of the total summed peak areas. Identification of the aromatic compounds used in
this analysis is provided in the Appendix on page 89, Table B.1.

S	
  

G	
  

H	
  

Figure 3.10. Relative normalized abundance of aromatic monomers released by pyrolysis of
lignin.
81

As the calculated abundances do not consider differences in response factors for each compound,
this approach is an approximation of the relative abundance of these sources of aromatic
monomers. However, this is the method that is generally applied by researchers for S/G ratio
estimation (Sykes et al., 2009). It should be noted that the dimethoxylated aromatics are not
likely to originate from syringyl lignins, as most gymnosperms lack the enzymes required for
sinapyl alcohol biosynthesis (Weng & Chapple, 2010). Instead, these dimethoxylated aromatics
are most likely the result of chemical rearrangement during pyrolysis, or they may represent a
single carbon remaining from a residual alkyl-aryl ether bond.
As shown in Figure 3.10, all fractions exhibit a comparable distribution of aromatic
products, albeit with notable exceptions. Interestingly, while the dimethoxylated lignins
comprise a small fraction of total aromatics (with a mass-averaged bulk composition of 6%),
they encompass a wide range of abundances among the fractions (3.6-14.5%). In particular, these
are most enriched in the fractions recovered at the more alkaline pH values and especially
fraction F3. The unmethoxylated phenolics comprise 11% of the mass-averaged bulk
composition but have a range of between 7 and 34% of the total aromatics across all fractions.
Again, fraction F3 shows substantially more Ph+C2 (primarily consisting of p-xylenol) relative to
the other fractions. Considering that fraction F3 represents only 3.1% of the mass of the lignin,
the overall contribution of this difference is relatively minor. Nevertheless, this result provides
evidence that the lignins are more likely to form unmethoxylated phenolics upon pyrolysis (e.g.,
non-guaiacyl lignins) that phase-separate out at higher pHs, and that the guaiacyl-only monomers
preferentially separate out at the lower pH values.

82

3.3.6 Property correlations
Many of the quantifiable properties of the lignins recovered in the pH range where the
majority of the mass is recovered (i.e., fractions F3-F7) correlate to each other and to the pH at
which the fractions were recovered. With the use of a correlation map (Figure 3.11), these linear
dependencies are presented simultaneously for fractions F3-F7 and show exceptionally strong
correlation, with |R| > 0.95 between a number of the lignin properties, as discussed below.

Mn

0.8

Mw

0.6
0.4

pH

0.2
0

β-O-4

-0.2

A-OH

-0.4

P-OH

-0.6

Mass
Recovery

-0.8
Mn

Mw

pH

β-O-4 A-OH P- OH

Correlation	
  Coefficient

1

-1
Mass
Recovery

Figure 3.11. Correlation map for lignin properties of fractions F3-F7 demonstrating strong
correlations between most properties; A-OH = Aliphatic OH, P-OH = Phenolic OH.
The first trend of importance observed is the relationship between the phenolic hydroxyl
content and the cumulative mass recovery, which can serve as a proxy for the alkali solubility. A
high alkali solubility means that the species in a given lignin fraction are highly soluble in the
83

caustic black liquor and tend to stay in solution, and not separate out as a separate lignin phase,
even as the pH is lowered via CO2 acidification. The phenolic hydroxyl content is known to
contribute to improved alkali solubility (Froass, Ragauskas, & Jiang, 1998; Lapierre et al., 1989),
which is one reason for lignin solubilization and removal from the cell wall during alkaline
pulping (D. J. Dong & Fricke, 1995). For example, the range of phenolic hydroxyl content in the
kraft lignins in the present work (from 0.47 to 0.69 mol per monomer, see Figure 3.4) is
significantly higher than the phenolic hydroxyl content of relatively unmodified wood lignins
(typically less than 0.1 mol/lignin monomer) (Chen, 1992). This positive relationship between
alkali solubility and free phenolic content has been observed in the literature, for example, for
the acidification of steam-exploded aspen lignin (H. L. Chum, Ratcliff, Schroeder, & Sopher,
1984) and for the CO2-based precipitation of softwood kraft lignins (Nagy et al., 2010). For the
latter case, the phenolic hydroxyl content (as well as carboxyl hydroxyl and aliphatic hydroxyl
content) as determined by

31

P NMR of phosphitylated lignin correlates with alkali solubility,

that is, with a tendency for the lignin fraction to remain in solution with decreasing pH. Dong et
al. (D. J. Dong, Fricke, Moudgil, & Johnson, 1996) identified a strong pH-dependence in the zeta
potential of kraft lignins, indicating that ionizable groups in lignin may contribute significantly to
their aggregation or repulsion. Finally, we note that the acid dissociation constant of phenolic
hydroxyl groups (pKa = 10.5) is well within the range of our pH-based fractionation experiments
(Guerra et al., 2007; Ragnar et al., 2000); thus, it represents an important property with respect to
the phase partitioning of liquid lignin.
GPC-determined lignin molecular weights exhibited a strong negative correlation with
the phenolic hydroxyl content and a strong positive correlation with the β-O-4 content (Figure

84

3.11). Both of these results are logical, considering that β-aryl ethers are the primary target of
kraft delignification and that the scission of a β-aryl ether results in the generation of a “free”
phenolic hydroxyl group, which confers alkali solubility to these lignin fragments and results in a
decrease in the molecular weight. This relationship has been quantified in the literature, where
the free phenolic content has been found to be strongly inversely correlated to the β-O-4 content
in kraft delignified softwoods (Koda, Gaspar, Yu, & Argyropoulos, 2005) and to the molecular
weight in diverse lignins modified by fungal laccases (van de Pas et al., 2011).

3.4 Conclusions
Liquid lignin fractions solvated in water were recovered from softwood kraft black liquor
using a novel aqueous fractionation process involving sequential CO2 acidification. The bulk of
the material was found to phase-separate between a pH of 11.6 and 10.0. The fractions were
characterized for a range of structural features. Based on the aromatic methoxyl content of these
fractions, it was concluded that the majority of the material by mass in all recovered fractions is
lignin. However, the fractions recovered at the highest pHs were enriched in aliphatic extractives
and polysaccharides relative to the other lignin fractions. Additionally, these higher-pH lignin
fractions were found by pyrolysis GC/MS to yield the highest fraction of non-guaiacyl lignins,
with the dimethoxylated aromatics potentially corresponding to higher aryl ether content. The
aliphatic hydroxyl content was not discernibly different between all of the various fractions,
while DGA/DTA curves showed differences only in the polysaccharide mass-loss peaks.
However, these solvated liquid-lignin fractions exhibited several notable trends in their
structural properties that can be related to their phase-partitioning behavior with respect to pH.
1

Specifically, the phenolic hydroxyl content of a lignin fraction as determined by H NMR was
85

found to exhibit a strong negative linear function with respect to the pH at which that fraction
phase-separated. In other words, lignin fractions with a lower phenolic hydroxyl content were
found to phase-separate from solution first, that is, at the highest pH values. The phenolic
hydroxyl content also had a strong negative correlation to the estimated molecular weights of the
lignin fractions as determined by GPC, but a strong positive correlation to the estimated β-O-4
content as determined by quantitative

13

C NMR. Thus, fractions with the highest molecular

weights and β-O-4 content separated out to form a liquid-lignin phase at a higher pH. Taken
together, this study provides encouraging evidence that our unique pH-based fractionation
approach can be employed to generate kraft lignins that are enriched or depleted in various
structural properties relative to the bulk lignin.

86

APPENDIX

87

APPENDIX

1

Figure B.1. Representative H NMR spectra of acetylated lignin fractions showing identification
of typical structures in lignin as well as aliphatic protons that might be found in softwood
extractives.

Figure B.2. Quantitative

13

C NMR spectra of a typical lignin fraction identifying the Cγ in a β-

O-4 bond and a methoxyl group used to determine lignin β-O-4 content.
88

F3
F2

F5
F4

F6

F1

F7

Figure B.3. Methoxyl content for the seven lignin fractions recovered.
Table B.1. Aromatic compound identification for pyro-GC/MS
Classification
S+C3
S+C2
S+C2
S+C1
S+C1
G+C3
G+C3
G+C3
G+C3
G+C3
G+C2
G+C2
G+C2
G+C1
G+C1
G+C1
G+C1
Guaiacol
Ph+C2
Ph+C2
Ph+C1
Ph+C1
Phenol

Compound
Methoxyeugenol
4-Ethyl-2,6-Dimethoxyphenol
3,4-Dimethoxystryene
2,3-Dimethoxytoluene
2,5-Dimethoxytoluene
Dihydromethyleugenol
Propiovanillione
Guaicylacetone
p-Propylguaiacol
Isoeugenol
Acetovanillone
4-Vinylguaiacol
4-Ethylguaiacol
2-Methoxy-4-Methylphenol
Vanillin
2-Methoxy-6-Methylphenol
3-Methoxy-5-Methylphenol
Guaiacol
3-Ethylphenol
p-Xylenol
p-Cresol
o-Cresol
Phenol
89

Res. Time
(Minutes)
25.9
23.1
20.9
17.6
18.4
23.6
24.7
24.2
20.1
21.0
23.5
19.7
18.9
17.3
22.3
16.8
17.3
15.5
18.1
17.4
16.6
15.9
15.0

Mass Peak (m/z)
194
137
164
152
137
151
151
137
137
164
151
135
137
138
151
123
123
124
107
122
107
108
94

CHAPTER 4. HEMICELLULOSE DISSOLUTION AND THE INFLUENCE OF LIGNIN
REMOVAL ON ENZYMATIC HYDROLYSIS FOR SODA PULPED HARDWOODS
4.1 Introduction
Lignocellulosic biomass feedstocks are targeted as one option to offset petroleum
consumption and to offer a sustainable approach to the production of liquid transportation fuels
and value-added chemicals. The complex structure within lignocellulosic biomass necessitates
the use of a pretreatment process to break plant cell wall recalcitrance for conversion to fuels or
chemicals (Himmel et al., 2007). A range of pretreatment variations have been researched to
understand how the structural components within cell walls can be deconstructed under different
conditions such as pH, temperature, and chemical loading (Mosier et al., 2005). Hemicellulose,
an amorphous polysaccharide, is most significantly affected by pretreatments utilizing liquid hot
water (LHW) (Y. Kim, Mosier, & Ladisch, 2009), dilute acid (Torget, Werdene, Himmel, &
Grohmann, 1990), or high temperature alkaline pretreatments (Oka et al., 2013). Both LHW and
dilute acid pretreatments dissolve hemicellulose into oligomeric fractions followed by hydrolysis
to monomeric constituents (Marzialetti et al., 2008; VanWalsum et al., 1996). Conversely, a high
temperature alkaline pretreatment, a process that is similar to chemical pulping, functions by
removing hemicellulose through the peeling reaction and suspending the polysaccharide as a
solute in the aqueous phase (Hamilton, 1962). The LHW and dilute acid pretreatments do not
significantly affect the lignin structure aside from redistributing some of the aromatic polymer
around the plant cell wall (Sannigrahi, Ragauskas, & Miller, 2008). Pretreatments that strongly
affect lignin structures incorporate alkali, organic solvents, or oxidative chemicals for lignin
removal (Chang, Nagwani, et al., 2001; Sarkanen et al., 1981; Venica, Chen, & Gratzl, 2008).
Pretreatments of this type can either directly dissolve lignin fragments or remove the fragments
90

by breaking specific linkages between the lignin structures itself or the hemicellulose
polysaccharide. Any of the disruptions made to the hemicellulose or lignin structure can result in
the improvement for the conversion of cellulose to a biologically renewable liquid transportation
fuel (Hendriks & Zeeman, 2009). Much of the impediment to commercializing the production of
these types of fuels from cellulose lies in constructing capital-intensive pretreatment processes
(Yang & Wyman, 2008). One way to overcome this issue is to look at industries that already
have a well-developed infrastructure for the chemical processing of woody biomass feedstocks.
In North America, the pulp and paper industry is uniquely positioned to assist with the
conversion of woody lignocellulose by already possessing a well-defined infrastructure of capital
equipment. On a global scale kraft pulping is the chemical process of choice for obtaining
cellulosic fibers for papermaking due to process economics that give sufficient recovery of
cooking chemicals and excess power generation allowing kraft pulp mills to meet their own net
power demands. During this process hemicellulose is solubilized by the peeling reaction and
eventually degraded to isosaccharinic acids (Sjostrom, 1993). Lignin is primarily removed by the
addition of Na2S where the nucleophilic aromatic substitution occurs by sulfur at elevated
temperatures. The obtained pulp is enriched in cellulosic fibers that are used in papermaking,
while the dissolved non-cellulosic biopolymers are sent to a recovery boiler where they are
burned for power generation. Current research and industrial implementation is now trending
towards recovering these biopolymers by fractionating these components from the black liquor
so they can be used as a source to generate renewable fuels or chemicals (Phillips et al., 2013).
This idea of utilizing the non-cellulosic portion of the plant cell wall for the generation of fuels
or chemicals is not new. In previous decades sulfite pulping was a more predominant form of
chemical pulping. When an acidic sulfite cook was utilized the hemicellulose portion within the

91

woody biomass could be hydrolyzed to monomers in the spent liquor followed by fermentation
to produce ethanol (Lawford & Rousseau, 1993). Alternatively, lignin can be recovered by
techniques such as ultrafiltration (Jönsson et al., 2008) or pH adjustment by sparging CO2 into
the aqueous black liquor to precipitate lignin (Nagy et al., 2010; Velez & Thies, 2013), which
has already investigated for commercialization (Lake & Blackburn, 2011; Tomani, 2010).
Depending on the fractionation process utilized the lignin recovered can contain variations in it’s
composition and structure. As an example the CO2 precipitation of lignin can contain variations
in sugar content and functional groups that are correlated with the pH of fractionation that was
achieved (Stoklosa et al., 2013). Although the current pulping industry heavily relies on
sulfurous compounds in chemical pulping, a new direction should be initiated by utilizing soda
pulping when it comes to producing renewable products from woody biomass. In this way
sulfurous compounds can be removed from the processing of woody biomass that produces
benefits both environmentally and for the recovery of lignin in a sulfur-free form that has greater
product applications (Lora & Glasser, 2002).
Integrated Forest Biorefineries (IFBR) are envisioned to mostly utilize the non-cellulosic
plant cell wall fractions for the production of fuels and chemicals by implementing an alkali preextraction (or pretreatment), while the cellulosic fraction can be sent to an actual pulping process
to procure the pulp fiber needed for papermaking (Ragauskas et al., 2006; Van Heiningen, 2006).
The economics associated with the transition to an IFBR must be favorable to an operational
pulp mill for this to occur. A process simulation and economic analysis was performed on an
IFBR working within three scenarios: traditional kraft pulping, hemicellulose pre-extraction
prior to pulping for ethanol production, and hemicellulose pre-extraction prior to pulping
coupled with the conversion of short cellulosic fibers to ethanol (Huang et al., 2009; Huang et
92

al., 2010). By converting both the extracted hemicellulose and short cellulosic fibers to ethanol a
total of 10.04 MM gallons (0.038 MM m3) of ethanol could be produced at $1.86 per gallon
($491 per m3), which simultaneously increased pulping capacity by 22% but the lower solid
contents in the black liquor would produce less steam than a traditional kraft mill (Huang et al.,
2010). However, this lower production of steam could be offset from a 13% reduction in fuel
usage by the lime kiln due to a lower loading of chemical inorganics (Huang et al., 2010).
Furthermore, ethanol is not the only other fuel or value-added chemical that has the option of
being produced in an IFBR. Hot water pre-extraction has been evaluated on both larch and birch
prior to chemical pulping for the purpose of extracting hemicelluloses followed by the
fermentation hemicellulosic sugars to lactic acid (Walton, Bischoff, van Heiningen, & van
Walsum, 2010) or succinic acid (Helmerius et al., 2010), respectively, by different strains of
bacteria. Although high titers of lactic acid or succinic acid were achieved, the hot water
extraction had the additional effect of decreasing pulp yield due to unwanted hydrolysis of
cellulosic microfibrils (Helmerius et al., 2010); but the pulp yields could be improved by adding
polysulfide-anthraquinone (PSAQ) to produce a good papermaking pulp with similar yields
(Horhammer et al., 2011). To ensure that an IFBR can be economical with traditional pulp mills
the inclusion of black liquor gasification is envisioned to boost the attractiveness for this type of
integrated processing. With black liquor gasification it is possible to recover the initial cooking
chemicals in the form of green liquor, increase power generation by converting organics to
synthesis gas that can be sent to a gas turbine or used in other conversion processes to produce
liquid hydrocarbons, and to completely segregate environmentally hazardous sulfur compounds
in a condensed form, especially at pressurized gasification at temperatures greater than 800°C
(Backman, Frederick, & Hupa, 1993; Lindstrom, Jameel, Naithani, Kirkman, & Renard, 2007;

93

Salmenoja, 1993). The main cautionary note about black liquor gasification would be to consider
proper operational parameters to decrease the amount of H2S that can be formed during the
process. Production of H2S would require additional wet scrubbers that would also increase
IFBR capital cost.
One understanding that is missing from IFBR evaluation is the effect that pulping
processes have on obtaining a more readily digestible woody biomass substrate. The research
discussed here focuses on the effect of soda pulping on three different hardwood species and the
implications for the three major cell wall fractions as a substrate source for renewable fuel or
chemical production. Soda pulping trials were conducted on three hardwood species in a
laboratory scale digester at varying temperature and time parameters. During the heat-up phase
and constant temperature cooking phase the solubilization of xylan and lignin was investigated to
explore the potential for xylan recovery from the black liquor while attempting to balance the
xylan retention against delignification for yields of monomeric sugars after enzymatic
hydrolysis. Woody biomass delignification evaluated at different severity cooks was compared
with initial hydrolysis rates to understand how lignin affects yield of sugars during the first hours
of enzymatic hydrolysis. Also evaluated was the effect on hydrolysis yields due to different
methods of comminution and different loadings of enzymes. Overall this work presents an option
for deconstructing the entire plant cell wall contents from woody biomass in a pulping process to
produce renewable fuels or chemicals.

94

4.2 Materials and Methods
4.2.1 Feedstock Composition
Three chipped hardwood species were analyzed in this study: Acer saccharum (sugar
maple), Betula pendula (silver birch), and Populus nigra X P. maximowiczii cv. NM-6 (hybrid
poplar). Samples were obtained from Curt Lindström of Smurfit-Kappa in Sweden (silver birch),
Dr. Raymond Miller of Michigan State University (hybrid poplar), and Todd Smith of the
Devereaux Saw Mill, Inc. in Pewamo, MI, USA (sugar maple). All samples had moisture
contents around 5-8% and the composition for each feedstock was determined by NREL/TP 51042618 (Sluiter et al., 2008) with slight modifications as discussed in Chapter 2, Section 2.2.1.
Ash and extractive components were quantified according to NREL/TP 510-42622 (Sluiter,
Hames, et al., 2005) and NREL/TP 510-42619 (Sluiter, Ruiz, et al., 2005), respectively.

4.2.2 Soda Pulping Trials
An M/K Systems, Inc. (Peabody, MA, USA) single vessel digester unit with a volume
capacity of 10 L was used for soda pulping trials. A 4:1 (L:kg dry wood) liquid loading was
selected with an 18% loading of NaOH (g NaOH per g dry wood); this corresponds to a NaOH
concentration of 45 g/L. The wood chips were cooked to different H-factor levels. The H-factor
is a single variable representation of the effect temperature and net cooking time have on
delignifcation given by the equation in Figure 4.1 (Sjostrom, 1993).
t

H=

"

∫ exp $# 43.2 −
0

16,113 %
' dt
T &

Figure 4.1. H-factor relationship for chemical pulping

95

The temperature (T) is in Kelvin and time (t) in minutes. Table 4.1 outlines the feedstocks that
were utilized for each cooking trial.
Table 4.1. Soda pulping trial conditions
Combined HFactor (Heat Up
Plus Cook)
114
525
1147

Maximum
Temperature
Temperature (°C) Ramp Rate
(°C/min)
150
0.75 to
135°C, 0.5 to
150°C
150
0.75 to
135°C, 0.5 to
150°C
170
0.75 to
135°C, 0.5 to
170°C

Total Cooking
Time (Hours)

Feedstocks

0.5

Poplar

3

Poplar

1

Birch, Maple,
Poplar

An additional LHW pretreatment was carried out on hybrid poplar in the same digester vessel
with a 4:1 liquid loading at 160°C for 1.5 hours to compare the effectiveness of soda pulping to
LHW pretreatment in terms of generating a digestible substrate by enzymatic hydrolysis.
For the most severe H-factor condition samples of the black liquor were collected during the heat
up phase, and the constant temperature cook phase with the use of an inline condenser. Figure
4.2 shows a general temperature profile over the entire pulping trial marked with red lines
indicating each sampling point during the heat up phase. At the constant temperature phase
samples were collected at 5, 10, 20, 30, 45, and 60 minutes. After system cool down and pressure
reduction the residual black liquor was collected in bulk. Wood chips were subjected to a
thorough wash with deionized water by rinsing the chips in the vessel canister, followed by
soaking the canister containing the wood chips in a 5-gallon container of water. The water was
exchanged up to two times a day until the wash water was clear and the bulk wash liquid pH was
neutral. The chips were then oven dried at 105°C to determine the yield.

96

Figure 4.2. A general temperature profile for a H-factor cook at 1147.

4.2.3 Analysis of Black Liquor Fractions
Each black liquor fraction collected was analyzed for dissolved plant cell wall content to
determine the kinetics of solubilization. Prior to this the pH of each fraction was measured by an
Oakton Instruments (Vernon Hills, IL, USA) pH 510 bench top meter to determine any pH
changes occurring during the process. Samples were then subjected to a 4% acid hydrolysis
according the method NREL LAP TP-510-42623 (Sluiter et al., 2006). Around 4 mL of each
black liquor sample was taken for analysis and diluted to a volume of 8 mL with water. The
black liquor analysis follows the same procedure for alkali extracts outlined in Chapter 2,
Section 2.2.2. The precipitated insoluble solids after hydrolysis were centrifuged and washed at
4,500 RPM for 5 minutes with deionized water in 15 mL conical tubes until the pH was neutral.
The washed solids were transferred to aluminum trays that were previously weighed and placed
97

in an oven to dry at 105°C overnight. Gravimetric analysis of the solids followed to determine
the total amount of precipitated material from the black liquor fractions. This material is
considered to be the amount of Klason lignin removed at each sample point.

4.2.4 Particle Size Reduction for Enzymatic Hydrolysis
Washed and dried wood chips after soda pulping underwent particle size reduction in two
steps. The first step combined the dried wood chips with deionized water at a solids consistency
of 5% (w/w). This mixture of wood chips and water was added to a blender for disintegration. To
produce a consistent disintegrated substrate between each feedstock, operation of the blender
was standardized for the amount of pulses and total time used to disintegrate. For each feedstock
around 20 pulses with the blender were sufficient to disintegrate the wood chips. This material
was then blended at a constant power setting for 20 minutes to produce the final disintegrated
feedstock sample. The disintegrated sample was filtered to remove free water, followed by
drying at 50°C overnight. The second step consisted of using a Thomas Wiley® Mini Mill
(Swedesboro, NJ, USA) and passing the feedstock through a 20 mesh screen. After milling the
samples were collected in zip-top plastic bags for storage. A moisture content analysis was
performed on all the feedstocks after milling with samples added to pre-weighed aluminum trays
and dried in an oven for at least four hours at 105°C. Additionally, samples of birch, maple, and
hybrid poplar (H-factor cooks of 1147 and 525) were left in a disintegrated particle size without
knife milling to test if a preferential enzymatic breakdown occurred between larger and smaller
particle size substrates.

98

4.2.5 Enzymatic Hydrolysis
Milled feedstock were added to 250 mL Erlenmeyer flasks at a 10% (w/v) solids loading
in a citric acid buffer at a concentration of 50 mM and pH around 5.5. This pH was chosen based
upon obtaining higher sugar yields from pretreated biomass (Lan et al., 2013; Z. J. Wang, Lan, &
Zhu, 2013). To prevent the growth of unwanted microorganisms tetracycline was added to the
slurry at a concentration of 40 µg/mL. Cellic® CTec2 and HTec2 from Novozymes (Bagsævard,
Denmark) were the enzyme cocktails utilized in the experiments and added at a CTec to HTec
ratio of 2:1. Protein concentrations were determined according to the Bradford Assay (SigmaAldrich). For all experiments the total enzyme loading was 20 mg enzyme per gram of glucan in
the biomass sample, unless otherwise noted. A piece of Parafilm (Bemis Company, Neenah, WI,
USA) was placed over the top of each flask to prevent evaporation losses. Hydrolysis was carried
out in a shaking incubator at 50°C at 160 RPM. Hydrolysis was allowed to occur over a 5 day
period. During sampling the flasks were removed from the incubator, briefly swirled, and then
400 µL of the hydrolysate was removed and placed in a 1.7 mL microcentrifuge tube, followed
by centrifuging at 13,000 RPM for three minutes. After sampling a new piece of parafilm was
placed on the top of the flasks before being returned to the incubator. After centrifugation the
liquid sample was diluted ten times in water followed by syringe filtering through a 0.22 µm
filter into HPLC vials. Samples were stored in a freezer until HPLC analysis. The conditions for
HPLC were performed according to conditions given in Chapter 2, Section 2.2.1. Glucose and
xylose concentrations were converted to their polymeric form, and yields for each were
calculated as a ratio of the sugar released at the observed time point divided by the theoretical
amount of glucan and xylan supplied based on the composition of the untreated or treated
feedstock.

99

4.3 Results and Discussions
4.3.1 Biomass Composition and Yields After Pulping
The temperature and NaOH loading utilized in this study altered the original hardwood
composition during soda pulping. Figure 4.3 presents the biomass composition after each soda
pulping trial and the lone LHW pretreatment on hybrid poplar. The composition results are
presented on a gram per dry gram original biomass basis. One of the first observations noticeable
is how the cellulose (i.e. glucan) content is relatively unaffected at each condition analyzed.

Figure 4.3. Composition of the three hardwood feedstocks before and after soda pulping, and
after LHW pretreatment for hybrid poplar.

It should be expected that the amorphous portions of the cellulose chain can be removed by
pretreatment (Sannigrahi et al., 2008), however, there is much less of this compared to the
crystalline cellulose that it does not affect the overall cellulose composition. The HPLC method
to quantify biomass composition, the same method utilized in Chapter 2, does not separate the
individual sugar monomers of galactose, mannose, or xylose. For this reason the hemicellulose
100

content is referred to as “Xyl+Man+Gal”. An assumption can be made that the predominant
hemicellulose component is xylan (Glasser et al., 2000) and this term will be used from this point
forward.
In contrast to glucan, the soda pulping and LHW treatments altered the content of xylan
and Klason lignin. The amount of xylan removed was similar for all soda pulping trials. Xylan
content decreases after each treatment with the removed xylan dissolving into the black liquor or
hydrolyzed by hot water. Greater xylan removal occurred during the LHW pretreatment on
poplar, a potentially expected result since LHW pretreatments at high temperatures can readily
hydrolyze the more susceptible hemicellulose fractions due to their lack of hydrogen bonding in
the cell wall (Bobleter, 1994). For the soda pulping trials, acetyl side groups were completely
saponified by alkali and only a small fraction of acetyl remained in the hybrid poplar after LHW.
As Figure 4.3 shows Klason lignin was removed in large quantities from all soda pulping trials
except for the lowest H-factor cook (114) and hardly any lignin removal for the LHW
-

pretreatment. This removal of lignin should be expected as well since OH performs well in
disrupting lignin bonds to dissolve substantial fractions in traditional pulping processes (Venica
et al., 2008). In terms of the overall amount of material solubilized, the most severe pulping trial
(H = 1147) produce similar amounts of material removed for all three feedstocks tested. The
poplar feedstock tested at three different H-factor cooks had similar overall material dissolution
at the 1147 and 525 cooks, but the low H-factor cook at 114 produced less material removed and
more residual solids after cooking. The same trend for Klason lignin in the LHW pretreated
poplar was present again for the overall amount of material removed from the LHW treated
poplar due to the lack of chemical catalysts utilized.

101

When comparing different sets of alkali pulps cooked at different temperatures and times
it is easiest to compare them according to the combined H-factor of the process. The combined
variable H-factor was developed to describe the extent of delignifcation on a woody biomass
feedstock during alkali chemical pulping (Sjostrom, 1993). In Figure 4.4 the content xyland and
Klason lignin are plotted as a function of H-factor, along with the pulp yield. The trends
presented are obvious; with increasing H-factor the pulp yield decreases, while the residual
content of xylan and Klason lignin in the poplar decrease as well.

Figure 4.4. Pulp yield, and residual xylan and Klason lignin content as a function of H-factor.

The high temperatures and chemical loading utilized have an expectation that larger amounts of
xylan can be removed as opposed to lower temperatures and lower NaOH loadings, which have a
-

tendency to lower the pH during the release of organic acids that neutralize OH (Lehto & Alen,
2013). That should not be the case in this work due to the high loading of NaOH utilized.
102

Although the xylan content in the untreated poplar did decrease after soda pulping, the residual
xylan remaining was consistent for the three separate pulping studies. Xylan extraction is known
to be dependent upon both temperature and alkali loading (Testova et al., 2014). The data
presented in Figure 4.4 show that xylan was removed during soda pulping, but the trials had an
overall greater effect on Klason lignin removal. It must be noted that a careful balance must be
obtained when trying to extract desirable plant cell wall components. In this case the greater
amount of xylan remaining in the residual solids is encouraging because it allows the xylan to be
recovered as a monomeric sugar during downstream enzymatic hydrolysis, or to serve as a fiber
adhesive to increase the pulp strength properties (Danielsson & Lindstrom, 2005; Molin &
Teder, 2002). The lower quantity of lignin remaining in the residual solids indicates the
opportunity to recover a majority of the lignin from the aqueous black liquor, and, in terms of
downstream pulp processing, lower chemical loadings can be achieved in a pulp bleaching
process for the purpose of oxidizing the residual lignin.

4.3.2 Hemicellulose and Klason Lignin Solubilization During Pulping
The collected samples of black liquor were analyzed for the amount of dissolved xylan
and Klason lignin in the aqueous phase. Figures 4.5 – 4.7 present the kinetics of solubilization
for both the xylan and Klason lignin portions after the high severity soda pulping trial (H = 1147)
for all three feedstocks. The amount of xylan and Klason lignin was determined after a 4% acid
hydrolysis on the black liquor samples. The dashed black line on each graph indicates the time
point when the cook reached a constant temperature of 170°C. A time of -116 minutes indicates
when the reactor vessel reached a temperature of 100°C, which was 100 minutes into the heat up
phase of the pulp trial. For the solubilization of xylan the trends show a gradual increase in the

103

amount of xylan solubilized from each hardwood species until 170°C. Upon reaching constant
temperature the amount of xylan solubilized decreases for both birch and poplar after 5 minutes
at 170°C, while the amount of xylan solubilized from maple increases 10 minutes into the cook
phase followed by a gradual decrease.

Figure 4.5. Solubilization of xylan and Klason lignin during birch soda pulping (H = 1147).

The decrease in xylan can be attributed to degradation during the cook phase by the xylan
breaking down to form saccharinic acids (Giudici & Park, 1996; Kakola, Alen, Pakkanen,
Matilainen, & Lahti, 2007; Ponder & Richards, 1997b). Xylan re-adsorption to cellulose could
also be responsible for the retention of a substantial fraction of the xylan.

104

Figure 4.6. Solubilization of xylan and Klason lignin during maple soda pulping (H = 1147).

Figure 4.7. Solubilization of xylan and Klason lignin during poplar soda pulping (H = 1147).

105

Alkali solutions at higher temperatures also contribute more towards cleaving the negatively
charged 4-O-MeGA substitutions on the xylan chain that will help to promote xylan aggregation
in the aqueous solution, thus decreasing solubility and increasing xylan adsorption to cellulose
(Linder et al., 2003). Furthermore, xylan present in the surfaces of hardwood kraft fibers have
been shown to contain lower frequencies of uronic acid side groups and a 30% higher average
molecular mass, thus lending more credence for xylan re-adsorption to kraft pulps during the
course of a cook (Dahlman, Jacobs, & Sjoberg, 2003). A combination of degradation and
adsorption would indeed indicate a loss in quantifiable xylan contained in the aqueous black
liquor. Additionally, the amount of xylan that is lost from solution appears to be affected by
species type. More xylan is lost from hybrid poplar as opposed to birch or maple. Xylan is the
dominant hemicellulose polysaccharide in hardwoods, however, GMs can account for 2-5% of
the dry wood weight (Sjostrom, 1993) and it has been shown that deacetylated mannans can
strongly and irreversibly absorb to cellulosic surfaces (Hannuksela, Tenkanen, & Holmbom,
2002). Since GMs were not directly quantified by the HPLC method utilized in this study it is
not known whether the three feedstocks assayed contained different amounts of GM that could
influence the overall amount of hemicellulose lost from the black liquor during pulping.
Conversely, the dissolution of Klason lignin during the soda pulping trials tracks closely
to the total mass loss and increases throughout the duration of the trial as evidenced in Figures
4.5 – 4.7. At the start of the constant temperature phase the amount of Klason lignin dissolved
into the black liquor increases for all three hardwood feedstocks, albeit a lower amount of Klason
lignin is removed from the birch when compared to maple and hybrid poplar. For a typical alkali
pulping process three stages of delignification occur: initial, bulk, and residual (Sjostrom, 1993).
The initial stage of delignification is controlled by diffusion, while bulk delignification is

106

controlled by chemical reactions and occurs at temperatures above 140°C until the residual phase
where there is relatively little lignin remaining in the woody biomass (≈ 10%) (Sjostrom, 1993).
The trends presented in Figures 4.5 – 4.7 show increases in delignification for all three
feedstocks after 40 minutes, corresponding to a temperature of 130°C. When comparing the three
H-factor cooks performed on hybrid poplar in Figure 4.4 the highest H-factor (H = 1147) could
remove up to 84% of the Klason lignin (i.e. a composition yield around 0.031 g final per g
original biomass), which would be expected with the alkali concentration and temperature
utilized. The extent of delignification that can occur in the bulk stage is dependent upon the
molar alkali concentration that is supplied during the initial stage of delignification (Olm &
-

Tistad, 1979). With OH being a strong nucleophile the most notable reactions that occur during
the bulk delignification stage include the cleavage of β-aryl ether bonds in phenolic and nonphenolic units, cleavage of α-aryl ether bonds in phenolic units, and condensation reactions
(Gierer, 1980). Concomitantly, hemicellulose is removed and/or degraded during bulk stage
delignification. In a study conducted on softwood, xylan removal has been shown to be linearly
proportional to delignifcation while polysaccharide peeling was minor, but GM was dissolved
rapidly and 45% degraded by the peeling reaction (Wigell, Brelid, & Theliander, 2007).
However, the data presented in Figures 4.5 – 4.7 indicate that xylan selectivity to lignin removal
is strongly a function of time.

4.3.3 Enzymatic Hydrolysis of Residual Soda Pulped Hardwoods
The insoluble hardwood feedstocks after alkali pulping were investigated by enzymatic
hydrolysis to determine the recovery of monomeric sugars. Investigations into the recovery of
soluble sugars from woody biomass after alkali pulping have not been carried out since chemical
107

pulps have strictly been reserved for the production of paper. However, pretreatments similar to
chemical pulping, such as the Organosolv or SPORL process, have been evaluated as a way to
recover and convert the polysaccharide portion of woody biomass to fuels or chemicals (Araque
et al., 2008; Del Rio, Chandra, & Saddler, 2010; Lan et al., 2013; Pan et al., 2006; Puls & Saake,
2004; G. S. Wang et al., 2009; Z. J. Wang et al., 2013; Zhu et al., 2009). (Lan et al., 2013; G. S.
Wang et al., 2009; Z. J. Wang et al., 2013; Zhu et al., 2009).
Each feedstock after pulping was subjected to enzymatic hydrolysis. The woody biomass
went through a two-stage particle size reduction that was discussed in Section 4.2.4. Figure 4.8
shows the glucose hydrolysis yields for birch, maple, and poplar after the high H-factor cook,
and Figure 4.9 shows the xylose hydrolysis yields for the same feedstocks. The yield of sugar
was calculated based on the concentration of glucan or xylan released into solution at each time
point divided by the amount of glucan or xylan supplied to the overall process (i.e. the initial
pulping process). It is evident that the sugar yields after soda pulping increase significantly when
compared to each untreated feedstock. High yields of glucose (> 80%) were achieved by the
addition of 20 mg of enzyme per g glucan for each feedstock shown in Figure 4.8. Xylose yields
were lower than 40% for each feedstock in Figure 4.9. The low yields account for xylan
extraction during the pulping process, and xylan degradation from the peeling reaction. The
increase in sugar yields for the treated feedstocks can be attributed to lignin removal since it is
understood that the removal of lignin will increase the exposed cellulose surface area and lessen
the loss of enzymes due to non-productive binding (Mansfield, Mooney, & Saddler, 1999). The
yields of soluble sugars after a chemical pulping treatment have not been tested extensively.
Other forms of alkaline pretreatment, such as AFEX™ or lime, have been utilized on woody
biomass to test the efficacy of plant cell wall deconstruction.

108

Figure 4.8. Glucose enzymatic hydrolysis yields for birch, maple, and poplar after the high Hfactor trial; enzyme loading 20 mg per g of glucan.

Figure 4.9. Xylose enzymatic hydrolysis yields for birch, maple, and poplar after the high Hfactor trial; enzyme loading 20 mg per g of glucan.
109

Black locust treated by AFEX™ at 180°C and 1.0 g NH3 per g dry wood, followed by milling,
was shown to greatly improve biomass digestibility at high enzyme loadings by generating close
to 80% yields of monomeric glucose and hemicellulose sugars after three days of hydrolysis
(Garlock, Wong, Balan, & Dale, 2012). However, AFEX™ is a unique alkaline pretreatment
since it does not function by changing the composition of the biomass material but instead opens
up the pore structure of the cell wall and converts the cellulose chain into a more digestible
allomorph (Chundawat, Bellesia, et al., 2011; Chundawat, Donohoe, et al., 2011). Lime
pretreatments on poplar have indicated up to 77% and 74% monomeric sugar conversions for
glucose and xylose, respectively, after three days of enzymatic hydrolysis, but lime pretreatments
also increase the ash content of the biomass due to calcium incorporation (Chang, Nagwani, et
al., 2001). More recently, Oka et al. (2013) determined that alkaline treatment at 150°C with 1 N
and 5 N NaOH solutions on cedar and eucalyptus produced different monomeric conversion
yields after 24 hour hydrolysis. The glucose conversion for cedar reached 10% after 1 N NaOH
treatment and 80% after 5 N NaOH treatment, while eucalyptus could reach 40% conversion
after 1 N NaOH treatment and 70% conversion after 5 N NaOH treatment (Oka et al., 2013). In
this research the NaOH concentration was around 1.1 M and following 24 hours of hydrolysis a
70% or higher yield of glucose could be achieved for birch, maple, and poplar. The yields in
Figure 4.8 are higher than what was achieved in the study by Oka et al. (2013) with 1 N NaOH at
150°C, but differences in yield can be attributed species difference (i.e. cedar is a softwood and
should have a higher lignin content), type of enzymes used, and enzyme loading.
Yields for enzymatic hydrolysis were also compared based on different pulping or
pretreatment conditions, substrate particle size, and enzyme loading. Figure 4.10 shows the
glucose yields for hybrid poplar after the three H-factor trials. The yields of glucose follow the

110

exact same trend in Figure 4.8 up until 24 hour hydrolysis, where a slight increase in the glucose
yield for the 1147 cook. The 525 H-factor cook utilized lower temperature (150°C) but longer
pulping time (3 hours) that could affect the macrostructure of the woody biomass plant cell wall
by redistributing lignin more evenly over the fibers to increase the cellulose surface area to
enzyme exposure. The xylose hydrolysis yields in Figure 4.11 are also improved for the 525 Hfactor cook indicating a better preferential breakdown of the polysaccharides to monomers by the
enzymes. Figures 4.3 and 4.4 indicate that poplar had more lignin removed with the increase in
H-factor severity. Previous work has related the level of delignification to enzymatic hydrolysis
response in hardwoods and softwoods (Mansfield et al., 1999; Schwald, Brownell, & Saddler,
1988; Yu, Jameel, Chang, & Park, 2011) and grasses (Li et al., 2012; Ohgren, Bura, Saddler, &
Zacchi, 2007; Williams & Hodge, 2014). These studies show that a larger quantity of lignin
removal typically improve sugar yields during enzymatic hydrolysis. However, Figures 4.10 and
4.11 indicate that greater sugar yields are obtained from feedstocks with higher residual lignin
content. While this data is slightly confounding it could be an expected result. Previously,
delignification below 5%, along with xylan removal, resulted in overall decreased digestibility
potentially due to aggregation of cellulose microfibrils (Ishizawa et al., 2009).

111

Hydrolysis	
  Time	
  (Hours)

Figure 4.10. Glucose enzymatic hydrolysis yields for soda pulped poplar at an H-factor of 1147,
525, and 114; enzyme loading of 20 mg per g glucan.

Hydrolysis	
  Time	
  (Hours)

Figure 4.11. Xylose enzymatic hydrolysis yields for soda pulped poplar at an H-factor of 1147,
525, and 114; enzyme loading of 20 mg per g glucan.
112

For comparison, LHW pretreated poplar also underwent enzymatic hydrolysis to assess
hydrolysis yield differences when compared to soda pulping. Figure 4.12 presents the glucose
hydrolysis yields for LHW pretreated poplar at two different enzyme loadings.

Figure 4.12. Glucose enzymatic hydrolysis yields for LHW poplar; enzyme loadings of 20 and 5
mg per g glucan.
The LHW pretreatment on poplar starts to level off around a glucan yield of 70% after 4 day
hydrolysis. When compared to the soda pulped poplar the sugar release for LHW treated poplar
is slower based on Figures 4.8 and 4.10 showing a consistent yield of 80% after 1 day hydrolysis.
Xylose yields for LHW pretreated poplar are given in Figure 4.13. The xylose yields steadily
increase for the LHW pretreated poplar over the course of hydrolysis to reach a final conversion
around 30%, but a 40% yield of xylose can be achieved after 24 hour hydrolysis for the soda
pulped poplar as indicated in Figures 4.9 and 4.11. Xylose yields were corrected to account for

113

losses by solubilization during pretreatment, and possible degradation by dehydration reactions;
both losses are common for LHW pretreatments (Garrote, Dominguez, & Parajo, 1999).

Figure 4.13. Xylose enzymatic hydrolysis yields for LHW poplar; enzyme loadings of 20 and 5
mg per g glucan.
The mechanism of LHW pretreatments is the dissolution and hydrolysis of hemicellulose, while
the lignin fraction is only altered structurally and not removed in large quantities (Y. Kim et al.,
2009). The composition chart in Figure 4.3 clearly shows a higher lignin content for LHW
pretreated poplar as opposed to soda pulped poplar at the high H-factor cook. This higher lignin
content probably influences the lower hydrolysis yields obtained for the LHW preatreated
poplar.
Hydrolysis yields were also evaluated by varying the loading of enzymes for hydrolysis
or the substrate particle size. Figure 4.14 and 4.15 presents the hydrolysis yields for glucose and
xylose from soda pulped birch at varying enzyme loadings.
114

Figure 4.14. Glucose enzymatic hydrolysis for soda pulped birch (H = 1147) at enzyme loadings
of 20, 10, and 5 mg per g glucan.

Figure 4.15. Xylose enzymatic hydrolysis for soda pulped birch (H = 1147) at enzyme loadings
of 20, 10, and 5 mg per g glucan.
115

The most obvious trend in Figures 4.14 and 4.15 is that hydrolysis yields decrease when lower
enzyme additions are utilized. Excluding pretreatment costs the high cost of enzyme production
for biofuels produced by the biochemical platform is one of the limiting steps that challenge
commercialization (Klein-Marcuschamer, Oleskowicz-Popiel, Simmons, & Blanch, 2012). By
generating high sugar yields at lower enzyme addition it is possible for the production of biofuels
from lignocellulosic biomass to be competitive. The lower enzyme dosages used here show that
a 10 mg enzyme addition per g glucan can achieve very similar glucan and xylan yields when
compared to 20 mg enzyme addition. However, the 5 mg enzyme addition only produces a
glucan yield around 70% and a xylan yield below 30%, thus wasting sugars that can be
recovered at the higher enzyme loading conditions.
Additionally, since a two-stage post-treatment milling was performed on the biomass to
decrease the wood chip particle size, an extra round of enzymatic hydrolysis studies was carried
out on the wood chips that were only “disintegrated” in a laboratory blender to determine
hydrolysis yields. The disintegration step gave the wood chips a more exposed surface and
unlocked a looser fiber structure. Figure 4.16 and 4.17 plot the glucose and xylose hydrolysis
yields obtained for disintegrated wood chips versus the yields from knife-milling. The yields
show each time point for birch and poplar after the high severity trial (H = 1147), and poplar
after the mid-severity trial (H = 525). Initially the yields in Figures 4.16 and 4.17 do not appear
to vary greatly, which would be an unexpected result since a larger particle size have produced
lower hydrolysis yields due to a lower availability of cellulose surface area, or a diffusion limited
step for the enzymes to access the polysaccharides in the substrate (Jorgensen, Kristensen, &
Felby, 2007).

116

Figure 4.16. Disintegration hydrolysis yields versus knife-milled hydrolysis yields for glucose.

Figure 4.17. Disintegration hydrolysis yields versus knife-milled hydrolysis yields for xylose.
117

A paired T-test was used, at a significance level of 0.05, to statistically compare the glucose and
xylose hydrolysis yields obtained for each particle size reduction method. The glucose hydrolysis
yields for soda birch indicated statistically significant differences with a P-value 0f 0.0466, but
the xylose hydrolysis yields produced no statistical differences with a P-value of 0.0619. Both Pvalues are came very close to the cutoff criteria so the effect of particle size on hydrolysis yields
cannot be adequately considered to be important. For the high severity poplar the statistics
indicate that no statistical differences exist for the glucose or xylose yields, both of which
produce P-values greater than 0.05. Alternatively, the mid-severity poplar trial produced
statistical differences for the glucose and xylose hydrolysis yields with P-values less than 0.002.
Overall, differences in particle size did not make significant contributions to the overall
hydrolysis yields. It should be understood that with additional biomass milling a lower quantity
of enzymes can achieve high sugar yields (Garlock et al., 2012). Similar yields of soluble sugars
after enzymatic hydrolysis with lower enzyme loadings and larger particle size shown in Figures
4.14 and 4.15 are encouraging for the aspect of producing a digestible feedstock after a soda
pulping process.

4.3.4 Initial Hydrolysis Rates
The yields of soluble sugars after enzymatic hydrolysis were evaluated further to
understand an initial rate of hydrolysis and the affect of lignin on the initial rate. The initial rates
of hydrolysis were determined based upon a linear regression of the sugar released over the
sample times of 0.5, 1, and 2 hours. Figures 4.18 and 4.19 plot the initial hydrolysis rates for
glucose and xylose, respectively, against the residual lignin content for each treated and
untreated biomass sample.

118

Maple	
  (H	
  =	
  1147)

Birch	
  (H	
  =	
  1147)
Poplar	
  (H	
  =	
  1147)
Poplar	
  (H	
  =	
  525)
Poplar	
  (H	
  =	
  114)
LHW	
  Poplar

Figure 4.18. Initial glucose hydrolysis rates versus residual lignin content.

Maple	
  
(H	
  =	
  1147)

Poplar	
  
(H	
  =	
  1147)

Birch	
  
(H	
  =	
  1147)

Poplar	
  
(H	
  =	
  525)

Poplar	
  
(H	
  =	
  114)

LHW	
  Poplar

Figure 4.19. Initial xylose hydrolysis rates versus residual lignin content.
119

The initial rates for glucose and xylose release from untreated feedstocks are extremely low, an
expected result since less than 10% yields of soluble sugars could be recovered from untreated
feedstocks. When evaluating the initial rates of hydrolysis for birch, maple, and poplar at the
1147 H-factor trial, very little difference exists between both the release of glucose and xylose.
The hybrid poplar treated at three different H-factors shows a steady decrease in the rate of
glucose release, but the rate of xylose release does not appear to change much at all. The
difference in the initial glucose rates for the hybrid poplar substrates is another indication of the
role that lignin plays during enzymatic hydrolysis. Lower H-factor pulps show to contain more
lignin in the residual substrate, as Figures 4.3 and 4.4 indicate, and although high sugar yields are
still obtained with continued hydrolysis, the greater amount of lignin in the residual substrate
could play a role in slowing down the action of the enzymes during the first couple of hours of
hydrolysis. This can be supported from the initial rates achieved on LHW poplar, which are
lower than the soda pulped poplar. Since LHW poplar contained the highest lignin content this is
an extra indication the role lignin has in slowing down the rates of sugar release during the first
house of enzymatic hydrolysis.
Furthermore, Figures 4.18 and 4.19 indicate a correlation between residual lignin content
and the initial hydrolysis rates. Figure 4.19 clearly indicates no change in the rate of xylose
release at short hydrolysis times, while Figure 4.18 shows a steady increase in the glucose
release rate with decreasing lignin content (i.e. higher H-factor). More importantly, Figure 4.18
shows that even the initial glucose hydrolysis rate for LHW poplar falls in line with this trend. It
is also interesting to see that the initial xylose hydrolysis rate in Figure 4.19 is independent of
lignin content. A 95% confidence interval test was performed to see if the initial hydrolysis rates
could be correlated to residual lignin content. The negative slope in Figure 4.18 for glucose did

120

not cross a value of zero for the confidence interval test; thus indicating that a correlation exists
between the initial glucose hydrolysis rates and residual lignin content. The confidence interval
test on the initial xylose hydrolysis rates confirmed that no correlation could be made to the
residual lignin content. Overall, for similarly milled, pretreated materials, lignin content
correlates strongly to the initial glucose hydrolysis yields.

4.4 Conclusions
Soda pulping on three different hardwood species was analyzed to determine
hemicellulose and lignin dissolution during the pulping trial, and the processes ability to produce
a digestible substrate for enzymes to recover monomeric sugars. Composition analysis after soda
pulping indicated high removal amounts for lignin, but the xylan portion of the woody biomass
could be recovered up to 70% of the initial material. For hybrid poplar three different H-factor
pulping trials were evaluated and indicated decreasing yields of lignin with increasing H-factor,
i.e. severity. At the highest H-factor setting around 10% of the lignin remained in the residual
wood chips, while all three H-factor trials could retain around 60-70% of the initial xylan.
During the course of soda pulping, the dissolution of xylan increased up to a point followed by
slight decreases in the amount of xylan dissolved in the aqueous phase. Klason lignin dissolution
continually increased throughout the trial.
The enzymatic hydrolysis of the residual wood chips indicated how soda pulping could
produce a readily digestible substrate. The highest H-factor cooks (H = 1147) produced very
similar glucose and xylose yields over a 5-day hydrolysis interval with final yields around 80%
for glucose and 40% for xylose (after accounting for losses). When comparing different H-factor
cooks on poplar (H = 1147 versus H = 525), the initial yields of sugars up to 24 hour hydrolysis

121

were essentially the same, but a slight increase in the final yields of glucose and xylose, around
90% and 50%, respectively, occurred for the poplar cooked at an H-factor of 525. Also by
decreasing the enzyme loading to 10 mg of enzyme per g of glucan, similar yields of glucose and
xylose were achieved. This is noteworthy since substantial delignification and substantial particle
size reduction may not be necessary. Additionally, a larger substrate particle size did not appear
to hinder the yields of monomeric sugars when compared to knife-milled substrates.
Furthermore, the effect that lignin has on hydrolysis was indicated from the calculation of initial
hydrolysis rates. Lower residual lignin content indicated a higher initial glucose hydrolysis rate
could be achieved. Conversely, the initial xylose hydrolysis rates did not appear to be affected by
the residual lignin content. This study shows how soda pulping can be utilized effectively to
produce a residual black liquor stream that has high contents of hemicellulose and lignin for
downstream recovery in a biorefinery setting, and an altered woody biomass feedstock for
downstream enzymatic hydrolysis to recover monomeric sugars for conversion to liquid fuels or
value-added chemicals.

122

REFERENCES

123

REFERENCES

Al-Dajani, W.W., & Tschirner, U.W. (2008). Pre-extraction of hemicelluloses and subsequent
kraft pulping Part I: alkaline extraction. Tappi Journal, 7(6), 3-8.
Alekhina, M., Mikkonen, K.S., Alén, R., Tenkanen, M., & Sixta, H. (2013). Carboxymethylation
of alkali extracted xylan for preparation of bio-based packaging films. Carbohydrate
Polymers.
Andersson, A., Persson, T., Zacchi, G., Stalbrand, H., & Jonsson, A. S. (2007). Comparison of
diafiltration and size-exclusion chromatography to recover hemicelluloses from process
water from thermomechanical pulping of spruce. Applied Biochemistry and
Biotechnology, 137, 971-983.
Andersson, C., Hodge, D., Berglund, K. A., & Rova, U. (2007). Effect of different carbon
sources on the production of succinic acid using metabolically engineered Escherichia
coli. Biotechnology Progress, 23(2), 381-388.
Araque, E., Parra, C., Freer, J., Contreras, D., Rodriguez, J., Mendonca, R., & Baeza, J. (2008).
Evaluation of organosolv pretreatment for the conversion of Pinus radiata D. Don to
ethanol.
Backman, R., Frederick, W. J., & Hupa, M. (1993). Basic studies on black-liquor pyrolysis and
char gasification. Bioresource Technology, 46(1-2), 153-158.
Banerjee, G., Car, S., Scott-Craig, J. S., Hodge, D. B., & Walton, J. D. (2011). Alkaline peroxide
pretreatment of corn stover: effects of biomass, peroxide, and enzyme loading and
composition on yields of glucose and xylose. Biotechnology for Biofuels, 4, 15.
Barford, R. A., Magidman, P., Phillips, J. G., & Fishman, M. L. (1986). Estimation of degree of
methylation of pectin by pyrolysis-gas chromatography Analytical Chemistry, 58(12),
2576-2578.
Baumberger, S., Abaecherli, A., Fasching, M., Gellerstedt, G., Gosselink, R., Hortling, B., . . . de
Jong, E. (2007). Molar mass determination of lignins by size-exclusion chromatography:
towards standardisation of the method. Holzforschung, 61(4), 459-468.
Bian, J., Peng, F., Peng, P., Xu, F., & Sun, R. C. (2010). Isolation and fractionation of
hemicelluloses by graded ethanol precipitation from Caragana korshinskii. Carbohydrate
Research, 345(6), 802-809.
Bobleter, O. (1994). Hydrothermal degradation of polymers derived from plants. Progress in
Polymer Science, 19(5), 797-841.

124

Borrega, M., Nieminen, K., & Sixta, H. (2011). Degradation kinetics of the main carbohydrates
in birch wood during hot water extraction in a batch reactor at elevated temperatures.
Bioresource Technology, 102(22), 10724-10732.
Borrega, M., Tolonen, L.K., Bardot, F., Testova, L., & Sixta, H. (2012). Potential of hot water
extraction of birch wood to produce high-purity dissolving pulp after alkaline pulping.
Bioresource Technology.
Calvo-Flores, F. G., & Dobado, J. A. (2010). Lignin as Renewable Raw Material. Chemsuschem,
3(11), 1227-1235.
Carpita, N. C. (1983). Hemicellulosic polymers of cell-walls of Zea coleoptiles Plant
Physiology, 72(2), 515-521.
Carpita, N. C. (1996). Structure and biogenesis of the cell walls of grasses. Annual Review of
Plant Physiology and Plant Molecular Biology, 47, 445-476.
Carpita, N. C., & Whittern, D. (1986). A highly substituted glucuronoarabinoxylan from
developing maize coleoptiles Carbohydrate Research, 146(1), 129-140.
Cathala, B., Saake, B., Faix, O., & Monties, B. (2003). Association behaviour of lignins and
lignin model compounds studied by multidetector size-exclusion chromatography.
Journal of Chromatography A, 1020(2), 229-239.
Chang, V. S., Kaar, W. E., Burr, B., & Holtzapple, M. T. (2001). Simultaneous saccharification
and fermentation of lime-treated biomass. Biotechnology Letters, 23(16), 1327-1333.
Chang, V. S., Nagwani, M., Kim, C. H., & Holtzapple, M. T. (2001). Oxidative lime
pretreatment of high-lignin biomass - Poplar wood and newspaper. Applied Biochemistry
and Biotechnology, 94(1), 1-28.
Chen, CL. (1992). Methods in Lignin Chemistry. Heidelberg: Springer.
Cheng, H., Zhan, H., Fu, S., & Lucia, L. A. (2010). Alkali extraction of hemicellulose from
depithed corn stover and effects on soda-aq-pulping. Bioresources, 6(1), 196-206.
Chum, H., Diebold, J., Scahill, J., Johnson, D., Black, S., Schroeder, H., & Kreibich, R. E.
(1989). Biomass pyrolysis oil feedstocks for phenolic adhesives. Acs Symposium Series,
385, 135-151.
Chum, H. L., Johnson, D. K., Tucker, M. P., & Himmel, M. E. (1987). Some aspects of lignin
characterization by high-performance size exclusion chromatography using styrene
divinylbenzene copolymer gels. Holzforschung, 41(2), 97-108.
Chum, H. L., Ratcliff, M., Schroeder, H. A., & Sopher, D. W. (1984). Electrochemistry of
biomass-derived materials. 1. Characterization, fractionation, and reductive electrolysis
125

of ethanol-extracted explosively-depressurized aspen lignin. Journal of Wood Chemistry
and Technology, 4(4), 505-532.
Chundawat, S. P. S., Bellesia, G., Uppugundla, N., Sousa, L. D., Gao, D. H., Cheh, A. M., . . .
Dale, B. E. (2011). Restructuring the Crystalline Cellulose Hydrogen Bond Network
Enhances Its Depolymerization Rate. Journal of the American Chemical Society,
133(29), 11163-11174.
Chundawat, S. P. S., Donohoe, B. S., Sousa, L. D., Elder, T., Agarwal, U. P., Lu, F. C., . . . Dale,
B. E. (2011). Multi-scale visualization and characterization of lignocellulosic plant cell
wall deconstruction during thermochemical pretreatment. Energy & Environmental
Science, 4(3), 973-984.
Compere, A.L., & Griffith, W.L. (2009). Preparation and analysis of biomass lignins. In J. R.
Mielenz (Ed.), Biofuels: Methods and Protocols (Vol. 581, pp. 185-212).
Compere, A.L., Griffith, W.L., Leitten, C.F., & Pickel, J.M. (2005). Evaluation of lignin from
alkaline-pulped hardwood black liquor: ORNL.
Contreras, S., Gaspar, A.R., Guerra, A., Lucia, L.A., & Argyropoulos, D.S. (2008). Propensity of
lignin to associate: light scattering photometry study with native lignins.
Biomacromolecules, 9(12), 3362-3369.
Crestini, C., Melone, F., Sette, M., & Saladino, R. (2011). Milled wood lignin: a linear oligomer.
Biomacromolecules, 12(11), 3928-3935.
Dahlman, O., Jacobs, A., & Sjoberg, J. (2003). Molecular properties of hemicelluloses located in
the surface and inner layers of hardwood and softwood pulps. Cellulose, 10(4), 325-334.
Dai, Q. Z., Jameel, H., & Chang, H. M. (2006). Precipitation of extractives onto kraft pulps
during black liquor recycling in extended delignification process. Journal of Wood
Chemistry and Technology, 26(1), 35-51.
Dale, B. (2008). Biofuels: Thinking clearly about the issues. Journal of Agricultural and Food
Chemistry, 56(11), 3885-3891.
Dale, B. E., & Ong, R. G. (2012). Energy, wealth, and human development: Why and how
biomass pretreatment research must improve. Biotechnology Progress, 28(4), 893-898.
Danielsson, S., & Lindstrom, M. E. (2005). Influence of birch xylan adsorption during kraft
cooking on softwood pulp strength. Nordic Pulp & Paper Research Journal, 20(4), 436441.
David, K., & Ragauskas, A. J. (2010). Switchgrass as an energy crop for biofuel production: A
review of its ligno-cellulosic chemical properties. Energy & Environmental Science, 3(9),
1182-1190.
126

Del Rio, L. F., Chandra, R. P., & Saddler, J. N. (2010). The Effect of Varying Organosolv
Pretreatment Chemicals on the Physicochemical Properties and Cellulolytic Hydrolysis
of Mountain Pine Beetle-Killed Lodgepole Pine. Applied Biochemistry and
Biotechnology, 161(1-8), 1-21.
Dong, D. (1993). Characterization of Kraft Lignin and Investigation of Pulping Effects on Pulp
Yield, Lignin Molecular Mass and Lignin Content of Black Liquor with a Central
Composite Pulping Design. University of Florida, Ann Arbor, MI. (9505752)
Dong, D. J., & Fricke, A. L. (1995). Effects of multiple pulping variables on the molecularweight and molecular-weight distribution of kraft lignin. Journal of Wood Chemistry and
Technology, 15(3), 369-393.
Dong, D. J., Fricke, A. L., Moudgil, B. M., & Johnson, H. (1996). Electrokinetic study of kraft
lignin. Tappi Journal, 79(7), 191-197.
Ebringerova, A. (2006). Structural diversity and application potential of hemicelluloses.
Macromolecular Symposia, 232, 1-12.
Ebringerova, A., Hromadkova, Z., Burchard, W., Dolega, R., & Vorwerg, W. (1994). Solution
properties of water-insoluble rye-bran arabinoxylan Carbohydrate Polymers, 24(3), 161169.
Elder, T.J., & Soltes, E.J. (1981). Pyrolysis of Lignocellulosic Materials. Phenolic Constituents
of A Wood Pyrolytic Oil. Wood and Fiber Science, 12(4), 217-226.
EPA, United States. (2013). Questions and Answers on Changes to the Renewable Fuel Standard
Program (RFS2). Retrieved March 17, 2014, 2014, from
http://www.epa.gov/otaq/fuels/renewablefuels/compliancehelp/rfs2-aq.htm
Francis, R. C., Shin, S. J., Omori, S., Amidon, T. E., & Blain, T. J. (2006). Soda pulping of
hardwoods catalyzed by anthraquinone and methyl substituted anthraquinones. Journal of
Wood Chemistry and Technology, 26(2), 141-152.
Franco, H., Ferraz, A., Milagres, A. M. F., Carvalho, W., Freer, J., Baeza, J., & Mendonca, R. T.
(2012). Alkaline sulfite/anthraquinone pretreatment followed by disk refining of Pinus
radiata and Pinus caribaea wood chips for biochemical ethanol production. Journal of
Chemical Technology and Biotechnology, 87(5), 651-657.
Froass, P. M., Ragauskas, A. J., & Jiang, J. (1996). Chemical structure of residual lignin from
kraft pulp. Journal of Wood Chemistry and Technology, 16(4), 347-365.
Froass, P. M., Ragauskas, A. J., & Jiang, J. E. (1998). NMR studies Part 3: Analysis of lignins
from modern kraft pulping technologies. Holzforschung, 52(4), 385-390.

127

Garcia-Perez, M., Chaala, A., Yang, J., & Roy, C. (2001). Co-pyrolysis of sugarcane bagasse
with petroleum residue. Part I: thermogravimetric analysis. Fuel, 80(9), 1245-1258.
Gargulak, J. D., & Lebo, S. E. (2000). Commercial use of lignin-based materials. In W. G.
Glasser, R. A. Northey & T. P. Schultz (Eds.), Lignin : Historical, Biological, and
Materials Perspectives (Vol. 742, pp. 304-320). Washington: Amer Chemical Soc.
Garlock, R. J., Wong, Y. S., Balan, V., & Dale, B. E. (2012). AFEX Pretreatment and Enzymatic
Conversion of Black Locust (Robinia pseudoacacia L.) to Soluble Sugars. Bioenergy
Research, 5(2), 306-318.
Garrote, G., Dominguez, H., & Parajo, J. C. (1999). Mild autohydrolysis: an environmentally
friendly technology for xylooligosaccharide production from wood. Journal of Chemical
Technology and Biotechnology, 74(11), 1101-1109.
Gidh, A. V., Decker, S. R., Vinzant, T. B., Himmel, M. E., & Williford, C. (2006).
Determination of lignin by size exclusion chromatography using multi angle laser light
scattering. Journal of Chromatography A, 1114(1), 102-110.
Gierer, J. (1980). Chemical aspects of kraft pulping Wood Science and Technology, 14(4), 241266.
Girio, F. M., Fonseca, C., Carvalheiro, F., Duarte, L. C., Marques, S., & Bogel-Lukasik, R.
(2010). Hemicelluloses for fuel ethanol: A review. Bioresource Technology, 101(13),
4775-4800.
Giudici, R., & Park, S. W. (1996). Kinetic model for kraft pulping of hardwood. Industrial &
Engineering Chemistry Research, 35(3), 856-863.
Glasser, W. G., Dave, V., & Frazier, C. E. (1993). Molecular-weight distribution of
(semi)commercial lignin derivatives. Journal of Wood Chemistry and Technology, 13(4),
545-559.
Glasser, W. G., Kaar, W. E., Jain, R. K., & Sealey, J. E. (2000). Isolation options for noncellulosic heteropolysaccharides (HetPS). Cellulose, 7(3), 299-317.
Goring, D. A. I., & Timell, T. E. (1960). Molecular properties of 6 4-O-methylglucuronoxylans
Journal of Physical Chemistry, 64(10), 1426-1430.
Gosselink, R.J.A., Abächerli, A., Semke, H., Malherbe, R., Käuper, P., Nadif, A., & Van Dam,
J.E.G. (2004). Analytical protocols for characterisation of sulphur-free lignin. Industrial
Crops and Products, 19(3), 271-281.
Griebl, A., Lange, T., Weber, H., Milacher, W., & Sixta, H. (2006). Xylo-oligosaccharide (XOS)
formation through hydrothermolysis of xylan derived from viscose process.

128

Guerra, A., Gaspar, A.R., Contreras, S., Lucia, L.A., Crestini, C., & Argyropoulos, D.S. (2007).
On the propensity of lignin to associate: a size exclusion chromatography study with
lignin derivatives isolated from different plant species. Phytochemistry, 68(20), 25702583.
Hakala, T. K., Liitia, T., & Suurnakki, A. (2013). Enzyme-aided alkaline extraction of
oligosaccharides and polymeric xylan from hardwood kraft pulp. Carbohydrate
Polymers, 93(1), 102-108.
Hamilton, J.K. (1962). The behaviour of wood carbohydrates in technical pulping processes.
Pure and Applied Chemistry, 5(1), 197-218.
Hampton, H. A., Haworth, W. N., & Hirst, E. L. (1929). Polysaccharides Part IV The
constitution of xylan. Journal of the Chemical Society, 1739-1753.
Hannuksela, T., Tenkanen, M., & Holmbom, B. (2002). Sorption of dissolved
galactoglucomannans and galactomannans to bleached kraft pulp. Cellulose, 9(3-4), 251261.
Hansen, N. M. L., & Plackett, D. (2008). Sustainable films and coatings from hemicelluloses: A
review. Biomacromolecules, 9(6), 1493-1505.
Hartley, R. D. (1983). Degradation of cell-walls of forages by sequential treatment with sodiumhydroxide and a commercial cellulase preparation Journal of the Science of Food and
Agriculture, 34(1), 29-36.
Harton, S.E., Pingali, S.V., Nunnery, G.A., Baker, D.A., Walker, S.H., Muddiman, D.C., . . .
Langan, P. (2012). Evidence for complex molecular architectures for solvent-extracted
lignins. Acs Macro Letters, 1(5), 568-573.
Helmerius, J., von Walter, J. V., Rova, U., Berglund, K. A., & Hodge, D. B. (2010). Impact of
hemicellulose pre-extraction for bioconversion on birch Kraft pulp properties.
Bioresource Technology, 101(15), 5996-6005.
Hendriks, A., & Zeeman, G. (2009). Pretreatments to enhance the digestibility of lignocellulosic
biomass. Bioresource Technology, 100(1), 10-18.
Himmel, M. E., Ding, S. Y., Johnson, D. K., Adney, W. S., Nimlos, M. R., Brady, J. W., &
Foust, T. D. (2007). Biomass recalcitrance: Engineering plants and enzymes for biofuels
production. Science, 315(5813), 804-807.
Hodge, D. B., Andersson, C., Berglund, K. A., & Rova, U. (2009). Detoxification requirements
for bioconversion of softwood dilute acid hydrolyzates to succinic acid. Enzyme and
Microbial Technology, 44(5), 309-316.

129

Holladay, J.E., Bozell, J.J., White, J.F., & Johnson, D. (2007). Top value-added chemicals from
biomass. Volume II - Results of Screening for Potential Candidates from Biorefinery
Lignin. In PNNL-16983 (Ed.).
Horhammer, H., Walton, S., & van Heiningen, A. (2011). A larch based biorefinery: preextraction and extract fermentation to lactic acid. Holzforschung, 65(4), 491-496.
Huang, H. J., Lin, W. L., Ramaswamy, S., & Tschirner, U. (2009). Process Modeling of
Comprehensive Integrated Forest Biorefinery-An Integrated Approach. Applied
Biochemistry and Biotechnology, 154(1-3), 205-216.
Huang, H. J., Ramaswamy, S., Al-Dajani, W. W., & Tschirner, U. (2010). Process modeling and
analysis of pulp mill-based integrated biorefinery with hemicellulose pre-extraction for
ethanol production: a comparative study. Bioresource Technology, 101(2), 624-631.
Ibn Yaich, A., Edlund, U., & Albertsson, A. C. (2012). Wood Hydrolysate Barriers: Performance
Controlled via Selective Recovery. Biomacromolecules, 13(2), 466-473.
Iiyama, K., Lam, T. B. T., & Stone, B. A. (1990). Phenolic-acid bridges between polysaccharides
and lignin in wheat internodes Phytochemistry, 29(3), 733-737.
International Energy Statistics - Petroleum Production. (2013).
Ishizawa, C. I., Jeoh, T., Adney, W. S., Himmel, M. E., Johnson, D. K., & Davis, M. F. (2009).
Can delignification decrease cellulose digestibility in acid pretreated corn stover?
Cellulose, 16(4), 677-686.
Izydorczyk, M. S., & MacGregor, A. W. (2000). Evidence of intermolecular interactions of betaglucans and arabinoxylans. Carbohydrate Polymers, 41(4), 417-420.
Jackson, L. S., Heitmann, J. A., & Joyce, T. W. (1998). Production of dissolving pulp from
recovered paper using enzymes. Tappi Journal, 81(3), 171-178.
Jacobs, A., & Dahlman, O. (2001). Characterization of the molar masses of hemicelluloses from
wood and pulps employing size exclusion chromatography and matrix-assisted laser
desorption ionization time-of-flight mass spectrometry. Biomacromolecules, 2(3), 894905.
Janzon, R., Saake, B., & Puls, J. (2008). Upgrading of paper-grade pulps to dissolving pulps by
nitren extraction: properties of nitren extracted xylans in comparison to NaOH and KOH
extracted xylans. Cellulose, 15(1), 161-175.
Johansson, M. H., & Samuelson, O. (1977). Reducing end groups in birch xylan and their
alkaline-degradation Wood Science and Technology, 11(4), 251-263.

130

Jönsson, A.S., Nordin, A.K., & Wallberg, O. (2008). Concentration and purification of lignin in
hardwood kraft pulping liquor by ultrafiltration and nanofiltration. Chemical Engineering
Research and Design, 86(11), 1271-1280.
Jorgensen, H., Kristensen, J. B., & Felby, C. (2007). Enzymatic conversion of lignocellulose into
fermentable sugars: challenges and opportunities. Biofuels Bioproducts & BiorefiningBiofpr, 1(2), 119-134.
Kakola, J., Alen, R., Pakkanen, H., Matilainen, R., & Lahti, K. (2007). Quantitative
determination of the main aliphatic carboxylic acids in wood kraft black liquors by highperformance liquid chromatography-mass spectrometry. Journal of Chromatography A,
1139(2), 263-270.
Karlsson, P., Roubroeks, J. P., Glasser, W. G., & Gatenholm, P. (2006). Optimization of the
process conditions for the extraction of heteropolysaccharides from birch (Betula
pendula). In J. J. Bozell & M. K. Patel (Eds.), Feedstocks for the Future: Renewables for
the Production of Chemicals and Materials (Vol. 921, pp. 321-333). Washington: Amer
Chemical Soc.
Kenealy, W.R., Houtman, C.J., Laplaza, J., Jeffries, T.W., & Horn, E.G. (2007). Pretreatments
for converting wood into paper and chemicals. Materials, Chemicals, and Energy From
Forest Biomass (Argyropoulos, DS, ed.). Washington, DC: American Chemical Society,
392-408.
Keyoumu, A., Sjodahl, R., Henriksson, G., Ek, M., Gellerstedt, G., & Lindstrom, M. E. (2004).
Continuous nano- and ultra-filtration of kraft pulping black liquor with ceramic filters - A
method for lowering the load on the recovery boiler while generating valuable sideproducts. Industrial Crops and Products, 20(2), 143-150.
Khiari, R., Mhenni, M. F., Belgacem, M. N., & Mauret, E. (2010). Chemical composition and
pulping of date palm rachis and Posidonia oceanica - A comparison with other wood and
non-wood fibre sources. Bioresource Technology, 101(2), 775-780.
Kim, T. H., Kim, J. S., Sunwoo, C., & Lee, Y. Y. (2003). Pretreatment of corn stover by aqueous
ammonia. Bioresource Technology, 90(1), 39-47.
Kim, Y., Mosier, N. S., & Ladisch, M. R. (2009). Enzymatic Digestion of Liquid Hot Water
Pretreated Hybrid Poplar. Biotechnology Progress, 25(2), 340-348.
Klein-Marcuschamer, D., Oleskowicz-Popiel, P., Simmons, B. A., & Blanch, H. W. (2012). The
challenge of enzyme cost in the production of lignocellulosic biofuels. Biotechnology and
Bioengineering, 109(4), 1083-1087.
Klinke, H. B., Ahring, B. K., Schmidt, A. S., & Thomsen, A. B. (2002). Characterization of
degradation products from alkaline wet oxidation of wheat straw. Bioresource
Technology, 82(1), 15-26.

131

Koda, K., Gaspar, A. R., Yu, L., & Argyropoulos, D. S. (2005). Molecular weight-functional
group relations in softwood residual kraft lignins. Holzforschung, 59(6), 612-619.
Kohnke, T., Ostlund, A., & Brelid, H. (2011). Adsorption of Arabinoxylan on Cellulosic
Surfaces: Influence of Degree of Substitution and Substitution Pattern on Adsorption
Characteristics. Biomacromolecules, 12(7), 2633-2641.
Kongruang, S., Han, M. J., Breton, C. I. G., & Penner, M. H. (2004). Quantitative analysis of
cellulose-reducing ends. Applied Biochemistry and Biotechnology, 113, 213-231.
Konn, J., Pranovich, A., & Holmbom, B. (2006). Dissolution of fibre material in alkaline pretreatment and refining of spruce CTMP. Holzforschung, 60(1), 32-39.
Koshijima, T., Timell, T. E., & Zinbo, M. (1965). Number-average molecular weight of native
hardwood xylans Journal of Polymer Science Part C-Polymer Symposium(11PC).
Krawczyk, H, Oinonen, P, & Jönsson, A.S. (2013). Combined membrane filtration and
enzymatic treatment for recovery of high molecular mass hemicelluloses from
chemithermomechanical pulp process water. Chemical Engineering Journal, 225, 292299.
Lake, M.A., & Blackburn, J.C. (2011). United States of America Patent No. US20110294991
A1.
Lan, T. Q., Gleisner, R., Zhu, J. Y., Dien, B. S., & Hector, R. E. (2013). High titer ethanol
production from SPORL-pretreated lodgepole pine by simultaneous enzymatic
saccharification and combined fermentation. Bioresource Technology, 127, 291-297.
Lapierre, C., Jouin, D., & Monties, B. (1989). On the molecular-origin of the alkali solubility of
gramineae lignins Phytochemistry, 28(5), 1401-1403.
Lau, M. W., Gunawan, C., & Dale, B. E. (2009). The impacts of pretreatment on the
fermentability of pretreated lignocellulosic biomass: a comparative evaluation between
ammonia fiber expansion and dilute acid pretreatment. Biotechnology for Biofuels, 2, 11.
Lawford, H. G., & Rousseau, J. D. (1993). Production of ethanol from pulp-mill hardwood and
softwood spent sulfite liquors by genetically-engineered Escherichia coli. Applied
Biochemistry and Biotechnology, 39, 667-685.
LeBel, R. G., Goring, D. A. I., & Timell, T. E. (1963). Solution Properties of Birch Xylan. I.
Measurement of Molecular Weight. Journal of Polymer Science: Part C, 2(1), 19.
Lehto, J., & Alen, R. (2013). Alkaline Pre-treatment of Hardwood Chips Prior to Delignification.
Journal of Wood Chemistry and Technology, 33(2), 77-91.
Li, M. Y., Foster, C., Kelkar, S., Pu, Y. Q., Holmes, D., Ragauskas, A., . . . Hodge, D. B. (2012).
Structural characterization of alkaline hydrogen peroxide pretreated grasses exhibiting
132

diverse lignin phenotypes. Biotechnology for Biofuels, 5(38), (6 June 2012)-(2016 June
2012).
Linder, A., Bergman, R., Bodin, A., & Gatenholm, P. (2003). Mechanism of assembly of xylan
onto cellulose surfaces. Langmuir, 19(12), 5072-5077.
Lindstrom, M., Jameel, H., Naithani, V., Kirkman, A., & Renard, J. (2007). Integrating Black
Liquor Gasification and Pulping: A Review of Current Technology. In D. S.
Argyropoulos (Ed.), Materials, Chemicals, and Energy from Forest Biomass (Vol. 954,
pp. 434-462). Washington: Amer Chemical Soc.
Liu, W., Hou, Q. X., Mao, C. B., Yuan, Z. R., & Li, K. C. (2012). Effect of Hemicellulose Preextraction on the Properties and Bleachability of Aspen (Populus tremuloides)
Chemithermomechanical Pulp. Industrial & Engineering Chemistry Research, 51(34),
11122-11127.
Liu, W., Yuan, Z. R., Mao, C. B., Hou, Q. X., & Li, K. C. (2012). Extracting hemicelluloses
prior to aspen chemi-thermomechanical pulping: Effects of pre-extraction on pulp
properties. Carbohydrate Polymers, 87(1), 322-327.
Liu, X., Fatehi, P., & Ni, Y. (2012). Removal of inhibitors from pre-hydrolysis liquor of kraftbased dissolving pulp production process using adsorption and flocculation processes.
Bioresource Technology, 116, 492-496.
Lora, J. H., & Glasser, W. G. (2002). Recent industrial applications of lignin: A sustainable
alternative to nonrenewable materials. Journal of Polymers and the Environment, 10(12), 39-48.
Loutfi, H, Blackwell, B, & Uloth, V. (1991). Lignin recovery from kraft black liquor:
preliminary process design. Tappi Journal, 74(1), 203-210.
Madakadze, I. C., Radiotis, T., Li, J., Goel, K., & Smith, D. L. (1999). Kraft pulping
characteristics and pulp properties of warm season grasses. Bioresource Technology,
69(1), 75-85.
Malawer, E.G., & Senak, L. (2004). Introduction to Size Exclusion Chromatography. In C. Wu
(Ed.), Handbook of Size Exclusion Chromatography and Related Techniques (Second ed.,
Vol. 91). New York, NY, USA: Marcel Dekker, Inc.
Mamman, A. S., Lee, J. M., Kim, Y. C., Hwang, I. T., Park, N. J., Hwang, Y. K., . . . Hwang, J.
S. (2008). Furfural: Hemicellulose/xylosederived biochemical. Biofuels Bioproducts &
Biorefining-Biofpr, 2(5), 438-454.
Mansfield, S. D., Mooney, C., & Saddler, J. N. (1999). Substrate and enzyme characteristics that
limit cellulose hydrolysis. Biotechnology Progress, 15(5), 804-816.

133

Mänttäri, M., Van der Bruggen, B., & Nyström, M. (2013). Nanofiltration. In S. Ramaswamy,
H.-J. Huang & B. V. Ramarao (Eds.), Separation and Purification Technologies in
Biorefineries (pp. 233-258).
Marzialetti, T., Olarte, M. B. V., Sievers, C., Hoskins, T. J. C., Agrawal, P. K., & Jones, C. W.
(2008). Dilute acid hydrolysis of Loblolly pine: A comprehensive approach. Industrial &
Engineering Chemistry Research, 47(19), 7131-7140.
Mayfield, C. A., Foster, C. D., Smith, C. T., Gan, J. B., & Fox, S. (2007). Opportunities, barriers,
and strategies for forest bioenergy and bio-based product development in the Southern
United States. Biomass & Bioenergy, 31(9), 631-637.
Mendes, C.V.T., Rocha, J.M.S., Sousa, G.D.A., & Carvalho, M.G.V.S. (2011). Extraction of
hemicelluloses prior to kraft cooking: A step for an integrated biorefinery in the pulp
mill. O PAPEL, 72(9), 79-83.
Miao, Z., Grift, T. E., Hansen, A. C., & Ting, K. C. (2011). Energy requirement for comminution
of biomass in relation to particle physical properties. Industrial Crops and Products,
33(2), 504-513.
Molin, U., & Teder, A. (2002). Importance of cellulose/hemicellulose-ratio for pulp strength.
Nordic Pulp & Paper Research Journal, 17(1), 14.
Monteil-Rivera, F., Phuong, M., Ye, M. W., Halasz, A., & Hawari, J. (2013). Isolation and
characterization of herbaceous lignins for applications in biomaterials. Industrial Crops
and Products, 41, 356-364.
Morreel, K., Kim, H., Lu, F. C., Dima, O., Akiyama, T., Vanholme, R., . . . Boerjan, W. (2010).
Mass Spectrometry-Based Fragmentation as an Identification Tool in Lignomics.
Analytical Chemistry, 82(19), 8095-8105.
Mosier, N., Wyman, C., Dale, B., Elander, R., Lee, Y. Y., Holtzapple, M., & Ladisch, M. (2005).
Features of promising technologies for pretreatment of lignocellulosic biomass.
Bioresource Technology, 96(6), 673-686.
Mousavioun, P., & Doherty, W. O. S. (2010). Chemical and thermal properties of fractionated
bagasse soda lignin. Industrial Crops and Products, 31(1), 52-58.
N'diaye, S., Rigal, L., Larocque, P., & Vidal, P. F. (1996). Extraction of hemicelluloses from
poplar, Populus tremuloides, using an extruder-type twin-screw reactor: A feasibility
study. Bioresource Technology, 57(1), 61-67.
Nagy, M., Kosa, M., Theliander, H., & Ragauskas, A.J. (2010). Characterization of CO2
precipitated Kraft lignin to promote its utilization. Green Chemistry, 12(1), 31-34.
Naran, R., Black, S., Decker, S. R., & Azadi, P. (2009). Extraction and characterization of native
heteroxylans from delignified corn stover and aspen. Cellulose, 16(4), 661-675.
134

Nowakowski, D. J., Bridgwater, A. V., Elliott, D. C., Meier, D., & de Wild, P. (2010). Lignin
fast pyrolysis: Results from an international collaboration. Journal of Analytical and
Applied Pyrolysis, 88(1), 53-72.
Ohgren, K., Bura, R., Saddler, J., & Zacchi, G. (2007). Effect of hemicellulose and lignin
removal on enzymatic hydrolysis of steam pretreated corn stover. Bioresource
Technology, 98(13), 2503-2510.
Ohman, F. (2006). Precipitation and separation of lignin from kraft black liquor. (Doctoral),
Chalmers University of Technology.
Oka, D., Kobayashi, K., Isobe, N., Ogawa, Y., Yokoyama, T., Kimura, S., . . . Wada, M. (2013).
Enzymatic hydrolysis of wood with alkaline treatment. Journal of Wood Science, 59(6),
484-488.
Olm, L., & Tistad, G. (1979). Kinetics of the initial-stage of kraft pulping Svensk
Papperstidning-Nordisk Cellulosa, 82(15), 458-464.
Olsson, L., & HahnHagerdal, B. (1996). Fermentation of lignocellulosic hydrolysates for ethanol
production. Enzyme and Microbial Technology, 18(5), 312-331.
Pan, X. J., Gilkes, N., Kadla, J., Pye, K., Saka, S., Gregg, D., . . . Saddler, J. (2006).
Bioconversion of hybrid poplar to ethanol and co-products using an organosolv
fractionation process: Optimization of process yields. Biotechnology and Bioengineering,
94(5), 851-861.
Pauly, M., & Keegstra, K. (2008). Cell‐wall carbohydrates and their modification as a resource
for biofuels. The Plant Journal, 54(4), 559-568.
Peng, F., Ren, J. L., Xu, F., Bian, J., Peng, P., & Sun, R. C. (2010). Fractionation of AlkaliSolubilized Hemicelluloses from Delignified Populus gansuensis: Structure and
Properties. Journal of Agricultural and Food Chemistry, 58(9), 5743-5750.
Perlack, R. D., & Stokes, B. J. (2011). U.S. Billion-Ton Update: Biomass Supply for a
Bioenergy and Bioproducts Industry. Oak Ridge, TN: Oak Ridge National Laboratory.
Persson, T., Krawczyk, H., Nordin, A. K., & Jonsson, A. S. (2010). Fractionation of process
water in thermomechanical pulp mills. Bioresource Technology, 101(11), 3884-3892.
Persson, T., Nordin, A.K., Zacchi, G., & Jönsson, A.S. (2007). Economic evaluation of isolation
of hemicelluloses from process streams from thermomechanical pulping of spruce
Applied Biochemistry and Biotecnology (pp. 741-752): Springer.
Phillips, R. B., Jameel, H., & Chang, H. M. (2013). Integration of pulp and paper technology
with bioethanol production. Biotechnology for Biofuels, 6.
135

Pinto, P. C., Evtuguin, D. V., Neto, C. P., Silvestre, A. J. D., & Amado, F. M. L. (2002).
Behavior of Eucalyptus globulus lignin during kraft pulping. II. Analysis by NMR,
ESI/MS, and GPC. Journal of Wood Chemistry and Technology, 22(2-3), 109-125.
Pitifer, L. A., McLellan, M. R., & Vanburen, J. P. (1994). Analysis of pectin content and degree
of polymerization in orange juice Food Chemistry, 50(1), 29-32.
Ponder, G. R., & Richards, G. N. (1997a). Arabinogalactan from Western larch .1. Effect of
uronic acid groups on size exclusion chromatography. Journal of Carbohydrate
Chemistry, 16(2), 181-193.
Ponder, G. R., & Richards, G. N. (1997b). Arabinogalactan from Western larch, Part III: alkaline
degradation revisited, with novel conclusions on molecular structure. Carbohydrate
Polymers, 34(4), 251-261.
Potthast, A., Rosenau, T., & Kosma, P. (2006). Analysis of oxidized functionalities in cellulose.
Polysaccharides Ii, 205, 1-48.
Pulp and Paper Capacities Survery 2011-2016. (2012). Rome, Italy.
Puls, J., & Saake, B. (2004). Industrially isolated hemicelluloses. In P. Gatenholm & M.
Tenhanen (Eds.), Hemicelluloses: Science and Technology (Vol. 864, pp. 24-37).
Puls, J., Schroder, N., Stein, A., Janzon, R., & Saake, B. (2006). Xylans from oat spelts and birch
kraft pulp. Macromolecular Symposia, 232, 85-92.
Ragauskas, A. J., Nagy, M., Kim, D.H., Eckert, C. A., Hallett, J. P., & Liotta, C. L. (2006). From
wood to fuels: integrating biofuels and pulp production. Industrial biotechnology, 2(1),
55-65.
Ragnar, M., Lindgren, C.T., & Nilvebrant, N.O. (2000). pKa-values of guaiacyl and syringyl
phenols related to lignin. Journal of Wood Chemistry and Technology, 20(3), 277-305.
Ralph, S.A., Ralph, J., & Landucci, L. (2004). NMR Database of Lignin and Cell Wall Model
Compounds. Madison, WI: US Forest Products Laboratory.
Robert, D. R., Bardet, M., Gellerstedt, G., & Lindfors, E. L. (1984). Structural-changes in lignin
during kraft cooking. 3. On the structure of dissolved lignins. Journal of Wood Chemistry
and Technology, 4(3), 239-263.
Rosell, K. G., & Svensson, S. (1975). Studies of distribution of 4-O-methyl-D-glucuronic acid
residues in birch xylan. Carbohydrate Research, 42(2), 297-304.
Saadatmand, S., Edlund, U., Albertsson, A. C., Danielsson, S., & Dahlman, O. (2012).
Prehydrolysis in Softwood Pulping Produces a Valuable Biorefinery Fraction for Material
Utilization. Environmental Science & Technology, 46(15), 8389-8396.
136

Saake, B., Kruse, T., & Puls, J. (2001). Investigation on molar mass, solubility and enzymatic
fragmentation of xylans by multi-detected SEC chromatography. Bioresource
Technology, 80(3), 195-204.
Salmenoja, K. (1993). Black-liquor gasification - theoretical and experimental studies.
Bioresource Technology, 46(1-2), 167-171.
Sannigrahi, P., Ragauskas, A. J., & Miller, S. J. (2008). Effects of Two-Stage Dilute Acid
Pretreatment on the Structure and Composition of Lignin and Cellulose in Loblolly Pine.
Bioenergy Research, 1(3-4), 205-214.
Sarkanen, S., Teller, D.C., Hall, J., & McCarthy, J.L. (1981). Lignin. 18. Associative effects
among organosolv lignin components. Macromolecules, 14(2), 426-434.
Saulnier, L., Vigouroux, J., & Thibault, J. F. (1995). Isolation and partial characterization of
feruloylated oligosaccharides from maize bran. Carbohydrate Research, 272(2), 241-253.
Schwald, W., Brownell, H. H., & Saddler, J. N. (1988). Enzymatic-hydrolysis of steam treated
aspen wood - Influence of partial hemicellulose and lignin removal prior to pretreatment
Journal of Wood Chemistry and Technology, 8(4), 543-560.
Shuai, L., Yang, Q., Zhu, J. Y., Lu, F. C., Weimer, P. J., Ralph, J., & Pan, X. J. (2010).
Comparative study of SPORL and dilute-acid pretreatments of spruce for cellulosic
ethanol production. Bioresource Technology, 101(9), 3106-3114.
Sills, D. L., & Gossett, J. M. (2011). Assessment of commercial hemicellulases for
saccharification of alkaline pretreated perennial biomass. Bioresource Technology,
102(2), 1389-1398.
Simpson, W., & TenWolde, A. (2007). Physical properties and moisture relations of wood. In U.
S. D. o. Agriculture (Ed.), The Encyclopedia of Wood. New York: Skyhorse Publishing,
Inc.
Sjostrom, E. (1993). Wood Chemistry: Fundamentals and Applications. San Diego, CA:
Academic Press.
Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., & Templeton, D. (2005). Determination
of Ash in Biomass NREL Laboratoy Analytical Procedures. Golden, Colorado, United
States: National Renewable Energy Laboratory.
Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., & Templeton, D. (2006). Determination
of Sugars, Byproducts, and Degradation Prodcuts in Liquid Fraction Process Samples
NREL Laboratory Analytical Procedures. Golden, Colorado, United States: National
Renewable Energy Laboratory.

137

Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., & Crocker, D. (2008).
Determination of Structural Carbohydrates and Lignin in Biomass NREL Laboratory
Analytical Procedure. Golden, Colorado, United States: National Renewable Energy
Laboratory.
Sluiter, A., Ruiz, R., Scarlata, C., Sluiter, J., & Templeton, D. (2005). Determination of
Extractives in Biomass NREL Laboratoy Analytical Procedures. Golden, Colorado,
United States: National Renewable Energy Laboratory.
Song, H., & Lee, S.Y. (2006). Production of succinic acid by bacterial fermentation. Enzyme and
Microbial Technology, 39(3), 352-361.
Stewart, D. (2008). Lignin as a base material for materials applications: Chemistry, application
and economics. Industrial Crops and Products, 27(2), 202-207.
Stoklosa, R.J., & Hodge, D.B. (2012). Extraction, Recovery, and Characterization of Hardwood
and Grass Hemicelluloses for Integration into Biorefining Processes. Industrial &
Engineering Chemistry Research, 51(34), 11045-11053.
Stoklosa, R.J., Velez, J., Kelkar, S., Saffron, C.M., Thies, M.C., & Hodge, D.B. (2013).
Correlating lignin structural features to phase partitioning behavior in a novel aqueous
fractionation of softwood Kraft black liquor. Green Chemistry, 15(10), 2904-2912.
Sturgeon, R. J. (1973). Determination of degree of polymerization of xylans Carbohydrate
Research, 30(1), 175-178.
Suhas, Carrott, PJM, & Riberio Carrott, MML. (2007). Lignin - from natural adsorbent to
activated carbon: A review. Bioresource Technology, 98(12), 2301-2312.
Sun, R. C. (2009). Detoxification and separation of lignocellulosic biomass prior to fermentation
for bioethanol production by removal of lignin and hemicelluloses. Bioresources, 4(2),
452-455.
Sykes, R., Yung, M., Novaes, E., Kirst, M., Peter, G., & Davis, M. (2009). High-Throughput
Screening of Plant Cell-Wall Composition Using Pyrolysis Molecular Beam Mass
Spectroscopy. In J. R. Mielenz (Ed.), Biofuels: Methods and Protocols (Vol. 581, pp.
169-183).
Teleman, A., Tenkanen, M., Jacobs, A., & Dahlman, O. (2002). Characterization of O-acetyl-(4O-methylglucurono)xylan isolated from birch and beech. Carbohydrate Research,
337(4), 373-377.
Testova, L., Roselli, A., Costabel, L., Kovasin, K., Tenkanen, M., & Sixta, H. (2014). Combined
Production of Polymeric Birch Xylan and Paper Pulp by Alkaline Pre-extraction
Followed by Alkaline Cooking. Industrial & Engineering Chemistry Research, 53(19),
8302-8310.
138

Timell, T. E. (1960). Isolation and properties of an O-acetyl-4-O-methylglucurono-xyloglycan
from the wood of white birch (Betula papyrifera). Journal of the American Chemical
Society, 82(19), 5211-5215.
Tomani, P. (2010). The lignoboost process. Cellulose Chemistry & Technology, 44(1), 53.
Torget, R., Werdene, P., Himmel, M., & Grohmann, K. (1990). Dilute acid pretreatment of short
rotation woody and herbaceous crops Applied Biochemistry and Biotechnology, 24-5,
115-126.
Tunc, M.S., & van Heiningen, A.R.P. (2008). Hydrothermal dissolution of mixed southern
hardwoods. Holzforschung, 62(5), 539-545.
van de Pas, D., Hickson, A., Donaldson, L., Lloyd-Jones, G., Tamminen, T., Fernyhough, A., &
Mattinen, M.L. (2011). Characterization of fractionated lignins polymerized by fungal
laccases. Bioresources, 6(2), 1105-1121.
Van Heiningen, A. (2006). Converting a kraft pulp mill into an integrated forest biorefinery.
Pulp & Paper-Canada, 107(6), 38-43.
VanWalsum, G. P., Allen, S. G., Spencer, M. J., Laser, M. S., Antal, M. J., & Lynd, L. R.
(1996). Conversion of lignocellulosics pretreated with liquid hot water to ethanol.
Applied Biochemistry and Biotechnology, 57-8, 157-170.
Vasile, C., Popescu, M.C, Stoleriu, A., & Gosselink, R.J.A. (2006). New Tends in Natural and
Synthetic Polymer Science. In C. Vasile & G. Zaikov (Eds.), (pp. 135-163). New York:
Nova Science Publishers, Inc.
Velez, J., & Thies, M.C. (2013). Solvated Liquid-Lignin Fractions from a Kraft Black Liquor.
Bioresource Technology.
Vena, P.F., Brienzo, M., del Prado García-Aparicio, M., Görgens, J.F., & Rypstra, T.
Hemicelluloses extraction from giant bamboo (Bambusa balcooa Roxburgh) prior to kraft
or soda-AQ pulping and its effect on pulp physical properties. Holzforschung, 1-8.
Venica, A. D., Chen, C. L., & Gratzl, J. S. (2008). Soda-AQ delignification of poplar wood. Part
1: Reaction mechanism and pulp properties. Holzforschung, 62(6), 627-636.
Villaverde, J. J., Li, J. B., Ek, M., Ligero, P., & de Vega, A. (2009). Native Lignin Structure of
Miscanthus x giganteus and Its Changes during Acetic and Formic Acid Fractionation.
Journal of Agricultural and Food Chemistry, 57(14), 6262-6270.
Vishtal, A.G., & Kraslawski, A. (2011). Challenges in industrial applications of technical lignins.
Bioresources, 6(3), 3547-3568.

139

Vlasenko, E.Y., Ryan, A.I., Shoemaker, C.F., & Shoemaker, S.P. (1998). The use of capillary
viscometry, reducing end-group analysis, and size exclusion chromatography combined
with multi-angle laser light scattering to characterize endo-1,4-beta-D-glucanases on
carboxymethylcellulose : A comparative evaluation of the three methods. Enzyme and
Microbial Technology, 23(6), 350-359.
Walton, S.L., Bischoff, K.M., van Heiningen, A.R.P., & van Walsum, G.P. (2010). Production of
lactic acid from hemicellulose extracts by Bacillus coagulans MXL-9. Journal of
Industrial Microbiology & Biotechnology, 37(8), 823-830.
Wang, G.S., Pan, X.J., Zhu, J.Y., Gleisner, R., & Rockwood, D. (2009). Sulfite pretreatment to
overcome recalcitrance of lignocellulose (SPORL) for robust enzymatic saccharification
of hardwoods. Biotechnology Progress, 25(4), 1086-1093.
Wang, Z.J., Lan, T.Q., & Zhu, J.Y. (2013). Lignosulfonate and elevated pH can enhance
enzymatic saccharification of lignocelluloses. Biotechnology for Biofuels, 6.
Wei, Z.J., Yang, Y., Yang, R., & Wang, C.Y. (2012). Alkaline lignin extracted from furfural
residues for pH-responsive Pickering emulsions and their recyclable polymerization.
Green Chemistry, 14(11), 3230-3236.
Weng, J.K., & Chapple, C. (2010). The origin and evolution of lignin biosynthesis. New
Phytologist, 187(2), 273-285.
Werpy, T., Frye, J., & Holladay, J. (2006). Succinic acid–a model building block for chemical
production from renewable resources. Biorefineries-Industrial Processes and Products:
Status Quo and Future Directions, 367-379.
Whistler, R.L., & Richards, E L. (1970). Hemicelluloses. In W. Pigman & D. Horton (Eds.), The
Carbohydrates - Chemistry and Biochemistry (Second Edition ed., Vol. 2A, pp. 447469). New York: Academic Press.
White, E.M. (2010). Woody Biomass for Bioenergy and Biofuels in the United States - A Briefing
Paper. Portland, OR.
Wigell, A., Brelid, H., & Theliander, H. (2007). Degradation/dissolution of softwood
hemicellulose during alkaline cooking at different temperatures and alkali concentrations.
Nordic Pulp & Paper Research Journal, 22(4), 488-494.
Wilkie, K.C.B., & Woo, S.L. (1977). Hemicelluloses of bamboos Arundinaria japonica and
Arundinaria anceps. 2. Heteroxylan and hemicellulosic materials from bamboo leaves,
and a reconsideration of general nature of commonly occurring xylans and other
hemicelluloses. Carbohydrate Research, 57(AUG), 145-162.

140

Williams, D. L., & Hodge, D. B. (2014). Impacts of delignification and hot water pretreatment
on the water induced cell wall swelling behavior of grasses and its relation to cellulolytic
enzyme hydrolysis and binding. Cellulose, 21(1), 221-235.
Williamson, M. (2013, April, 2013). Tembec: A Biorefiner Ahead of Its Time. IPW Magazine,
24-27.
Wyman, C.E., Dale, B.E., Elander, R.T., Holtzapple, M., Ladisch, M.R., & Lee, Y.Y. (2005).
Coordinated development of leading biomass pretreatment technologies. Bioresource
Technology, 96(18), 1959-1966.
Xu, J.L., Cheng, J.J., Sharma-Shivappa, R.R., & Burns, J.C. (2010). Sodium hydroxide
pretreatment of switchgrass for ethanol production. Energy & Fuels, 24, 2113-2119.
Yang, B., & Wyman, C.E. (2008). Pretreatment: the key to unlocking low-cost cellulosic
ethanol. Biofuels Bioproducts & Biorefining-Biofpr, 2(1), 26-40.
Yelle, D.J., Ralph, J., & Frihart, C.R. (2011). Delineating pMDI model reactions with loblolly
pine via solution-state NMR spectroscopy. Part 1. Catalyzed reactions with wood models
and wood polymers. Holzforschung, 65(2), 131-143.
Yoon, S.H., Tunc, M.S., & Van Heiningen, A. (2011). Near-neutral pre-extraction of
hemicelluloses and subsequent kraft pulping of southern mixed hardwoods. Tappi
Journal, 10(1), 7-15.
Yoon, S.H., van Heiningen, A., & Krishnagopalan, G.A. (2008). Kraft Pulping Integrated with
mild alkaline pre-extraction of southern mixed hardwoods. Paper presented at the TAPPI
Pulping and Enivonmental Coference, Portland, OR.
Yu, Z. Y., Jameel, H., Chang, H. M., & Park, S. (2011). The effect of delignification of forest
biomass on enzymatic hydrolysis. Bioresource Technology, 102(19), 9083-9089.
Zeikus, J.G., Jain, M.K., & Elankovan, P. (1999). Biotechnology of succinic acid production and
markets for derived industrial products. Applied Microbiology and Biotechnology, 51(5),
545-552.
Zhang, Y.H.P., & Lynd, L.R. (2005). Determination of the number-average degree of
polymerization of cellodextrins and cellulose with application to enzymatic hydrolysis.
Biomacromolecules, 6(3), 1510-1515.
Zhu, J.Y., Pan, X.J., Wang, G.S., & Gleisner, R. (2009). Sulfite pretreatment (SPORL) for robust
enzymatic saccharification of spruce and red pine. Bioresource Technology, 100(8),
2411-2418.
Zhu, J.Y., Zhu, W.Y., Obryan, P., Dien, B.S., Tian, S., Gleisner, R., & Pan, X.J. (2010). Ethanol
production from SPORL-pretreated lodgepole pine: preliminary evaluation of mass

141

balance and process energy efficiency. Applied Microbiology and Biotechnology, 86(5),
1355-1365.

142