INTEGRATED BIOLOGICAL PROCESSES FOR CONVERSION OF
TM
AFEX PRETREATED BIOMASS TO ETHANOL

By
Mingjie Jin

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
CHEMICAL ENGINEERING
2012

ABSTRACT
INTEGRATED BIOLOGICAL PROCESSES FOR CONVERSION OF
TM
AFEX PRETREATED BIOMASS TO ETHANOL
By
Mingjie Jin
Ethanol production from lignocellulosic biomass has gained much momentum due to its
benefits to energy security, reduction of green house gas emission, as well as both environmental
and social sustainability. The technology for lignocellulosic ethanol production, however, is not
yet fully commercialized. The major issues impeding the cellulosic ethanol production in the
biochemical route include the high enzyme loadings needed, long enzymatic hydrolysis time,
slow xylose fermentation and low ethanol productivity, which result in a high production cost.
TM

Ammonia Fiber Expansion (AFEX

) is a leading alkaline pretreatment. It provides a

biomass substrate with high enzymatic digestibility and high fermentation potential. Previous
TM

studies on ethanol production from AFEX

pretreated biomass focused on a separate

enzymatic hydrolysis and fermentation process (SHF). However, integrated biological processes,
such as simultaneous saccharification and co-fermentation (SSCF) and consolidated
bioprocessing (CBP), are strongly believed to have lower cost compared to SHF. This
TM

dissertation work studies the integrated biological processes performed on AFEX

pretreated

biomass and resolves the aforementioned issues for biochemical production of cellulosic ethanol.

The slow xylose fermentation issue in hydrolysate by Saccharomyces cerevisiae 424A
(LNH-ST) was quantitatively studied. The xylose fermentation inhibition was not only from
degradation products but also from ethanol and metabolites generated during glucose
fermentation (Chapter II). Based on such understanding, a two-step SSCF process was developed,
in which xylan was hydrolyzed and fermented ahead of glucan/glucose. As a result, xylose
fermentation was greatly improved (Chapter III). Through the study of conventional SSCF on
TM

AFEX

treated corn stover (CS), it was found that pre-hydrolysis to generate a certain amount

of glucose was crucial for achieving a good result (Chapter IV). A high solids loading process
can save much cost for ethanol production. However, ethanol yield decreased with increasing
solids loading. From an economic point of view, 6% (w/w) glucan loading was the optimal solids
TM

loading during SSCF of AFEX

CS (Chapter V). For improvement of productivity, a

continuous SSCF process using multi-stage continuous stirred tank reactors (CSTRs) was
developed based on the kinetic studies of the two reactions in SSCF (enzymatic hydrolysis and
fermentation)(Chapter VI). Based on the fundamental understandings of the cellulosic ethanol
production, a novel industrially-relevant integrated biological process was developed. This
process shortened the biological processing time (including enzymatic hydrolysis and
fermentation) from 11 days to around 2 days, reduced enzyme loading by more than 1/3 and
enhanced ethanol productivity by 2-3 times (Chapter VII). Consolidated bioprocessing (CBP)
eliminates the enzyme cost and is believed to be the ultimate low-cost industrial configuration
for cellulosic ethanol production. CBP studies using Clostridium phytofermentans on
TM

AFEX

CS at both low and high solids loadings showed promising results (Chapter VIII & IX).

AFEX is a trademark of MBI, International of Lansing, Michigan.

This work is dedicated to my parents, Jianlin Jin and Meiying Yu, without
whose unconditional love and support this effort could not have been
possible.

iv

ACKNOWLEDGMENTS

I would like to thank my distinguished committee members: Professors Bruce E. Dale
(Chair), Venkatesh Balan, Gemma Reguera, Robert M. Worden and Robert P. Hausinger, for
their support, encouragement and valuable suggestions. I would like to specially thank my
advisor Dr. Dale and Dr. Balan for providing detailed guidance and support during my Ph.D.
study period. I am grateful to my QB program advisor Dr. Reguera and Professor Lee R. Lynd
from Dartmouth College for providing anaerobic fermentation training and resources.
I am indebted to all the Biomass Conversion Research Laboratory (BCRL) members for
their encouragement, insightful comments and support. I would like to particularly thank Christa
Gunawan for working with me on many projects and Bryan Bals for support on the technoeconomic analyses. I would also like to thank Ming Woei Lau, Dahai Gao, Leonardo da Costa
Sousa, Rebecca Garlock, James Humpula, Pete Donald, Mary Ann Vigil, Nirmal Uppugundla,
Nick Posawatz and Derek Marshall for their support.
My thanks go to all the people ever helped and encouraged me. Thank the Great Lakes
Bioenergy Research Center (GLBRC). I thank Xiongjun Shao (Dartmouth College) and Jenna
Young for their help on the anaerobic fermentation training.

v

TABLE OF CONTENTS

LIST OF TABLES ................................................................................................................... xi
LIST OF FIGURES ................................................................................................................ xii
ABBREVIATIONS ................................................................................................................ xvi
CHAPTER I BACKGROUND AND GENERAL INTRODUCTION .................................... 1
1.1 Cellulosic ethanol production ..................................................................................... 1
1.1.1 Background ...................................................................................................... 1
1.1.2 Cellulosic ethanol production........................................................................... 2
1.1.3 Biomass pretreatment ....................................................................................... 3
TM
1.1.4 Ammonia fiber expansion (AFEX ) pretreatment ........................................ 5
1.1.5 Process configurations based on four biological steps ..................................... 7
TM
1.2 Ethanol production from AFEX
pretreated biomass ........................................... 15
TM
1.2.1 High digestibility and high fermentability of AFEX
pretreated biomass . 15
TM
1.2.2 Problems of ethanol production from AFEX
pretreated biomass ............. 22
1.3 Research objectives ................................................................................................... 25
CHAPTER II QUANTITATIVELY UNDERSTANDING REDUCED XYLOSE
TM
TREATED CORN STOVER
FERMENTATION PERFORMANCE IN AFEX
HYDROLYSATE .................................................................................................................... 26
Abstract ........................................................................................................................... 26
2.1 Introduction ............................................................................................................... 27
2.2 Materials and Methods .............................................................................................. 28
TM
2.2.1. AFEX treated corn stover ......................................................................... 28
TM
2.2.2 AFEX -CS hydrolysate preparation ............................................................ 28
TM
2.2.3 AFEX -CS water extract preparation ......................................................... 29
2.2.4 Microorganisms and seed culture preparation................................................ 29
2.2.5 Fermentations in YEP medium or hydrolysates ............................................. 30
2.2.6 Xylose conversion experiments using non-growing cells .............................. 30
2.2.7 HPLC analyses ............................................................................................... 32
2.2.8 Measurement of viable cells density .............................................................. 32
vi

2.3 Results and discussion .............................................................................................. 32
TM
2.3.1 Slow xylose fermentation in AFEX -CS hydrolysate................................. 32
2.3.2 Effect of degradation products on specific xylose consumption rate ............. 34
2.3.3 Effect of ethanol on specific xylose consumption rate................................... 35
2.3.4 Effect of culture medium on specific xylose consumption rate ..................... 38
2.3.5 Effects of degradation products on cell growth and nutrients addition on
xylose consumption ................................................................................................. 40
2.3.6 Understanding the xylose consumption problem during hydrolysate
fermentation by S. cerevisiae 424A......................................................................... 42
2.4 Conclusions ............................................................................................................... 44
CHAPTER III SIMULTANEOUS SACCHARIFICATION AND CO-FERMENTATION OF
TM
PRETREATED SWITCHGRASS.......................................................................... 45
AFEX
Abstract ........................................................................................................................... 45
3.1 Introduction ............................................................................................................... 46
3.2 Materials and Methods ............................................................................................. 47
3.2.1 Switchgrass ..................................................................................................... 47
TM
Pretreatment ................................................................................... 47
3.2.2 AFEX
3.2.3. Enzymatic Hydrolysis (EH) .......................................................................... 47
3.2.4 Microorganism and seed culture preparation ................................................. 48
3.2.5 Fermentation of hydrolyzate .......................................................................... 48
TM
3.2.6 SSCF of AFEX -SG.................................................................................... 49
TM
3.2.7 Preparation of washed AFEX -treated SG.................................................. 50
3.2.8 Mass balance .................................................................................................. 50
3.3 Results and discussion .............................................................................................. 51
3.3.1 Traditional SSCF vs. two-step SSCF (un-optimized) .................................... 51
3.3.2 Optimization of two-step SSCF conditions .................................................... 52
TM
3.3.3 Effects of EH residue and washed AFEX -SG on hydrolyzate fermentation
................................................................................................................................. 57
3.3.4 Mass balance comparison between SHF and two-step SSCF ........................ 59
3.3.5 The limiting factors during SSCF .................................................................. 60
3.4 Conclusions ............................................................................................................... 61
CHAPTER IV SIMULTANEOUS SACCHARIFICATION AND CO-FERMENTATION OF
TM
AFEX
PRETREATED CORN STOVER .......................................................................... 65
Abstract ........................................................................................................................... 65
4.1 Introduction ............................................................................................................... 66
4.2 Materials and Methods .............................................................................................. 67
TM
4.2.1 AFEX treated corn stover .......................................................................... 67
4.2.2 Microorganism and seed culture preparation ................................................. 67
TM
4.2.3 Separate hydrolysis and fermentation (SHF) of AFEX -CS ...................... 68
4.2.4 Simultaneous Saccharification and Co-fermentation ..................................... 68
vii

TM

4.2.5 Preparation of washed AFEX -CS .............................................................. 69
4.2.6 Mass balance .................................................................................................. 69
4.2.7 Measurement of viable cell density ................................................................ 69
4.3 Results and Discussion.............................................................................................. 69
4.3.1 Traditional SSCF process vs. Two-step SSCF process .................................. 69
4.3.2 Effect of unhydrolyzed solids on xylose fermentation ................................... 72
4.3.3 Effect of pre-hydrolysis time on SSCF performance ..................................... 73
4.3.4 Effect of solids and liquids from different pre-hydrolysis on xylose
fermentation and yeast cell viability ....................................................................... 74
4.3.5 Glucose concentration after pre-hydrolysis is the key for xylose fermentation
during SSCF ............................................................................................................ 79
4.3.6 Mass balance comparison of different process............................................... 80
4.4 Conclusions ............................................................................................................... 84
CHAPTER V EFFECT OF SOLIDS LOADING ON SSCF ............................................... 85
Abstract ........................................................................................................................... 85
5.1 Introduction ............................................................................................................... 86
5.2 Materials and Methods .............................................................................................. 87
TM
pretreated corn stover ..................................................................... 87
5.2.1 AFEX
5.2.2 Enzymatic hydrolysis ..................................................................................... 87
5.2.3 Microorganism and seed culture preparation ................................................. 88
5.2.4 Fermentation of enzymatic hydrolysate (SHF) .............................................. 88
5.2.5 Simultaneous saccharification and co-fermentation (SSCF) ......................... 88
5.2.6 Techno-economic analyses ............................................................................. 89
5.3 Results and Discussion.............................................................................................. 90
5.3.1 SSCF performances at different solids loading .............................................. 90
5.3.2 Fermentation performances of solids-free hydrolysate at different solids
loading ..................................................................................................................... 92
5.3.3 SSCF performances with YEP supplementation ............................................ 94
5.3.4 Effect of solids loading on monomeric sugar conversion and ethanol yield.. 96
5.3.5 Techno-economic analyses ............................................................................. 98
TM

CHAPTER VI CONTINUOUS SSCF OF AFEX
PRETREATED CORN STOVER .... 101
Abstract ......................................................................................................................... 101
6.1 Introduction ............................................................................................................. 102
6.2 Materials and Methods ............................................................................................ 103
TM
6.2.1 AFEX
pretreated corn stover ................................................................... 103
6.2.2 Enzymatic Hydrolysis .................................................................................. 104
6.2.3 Microorganisms and seed culture preparation.............................................. 104
6.2.4 Hydrolysate fermentation (SHF) and batch SSCF ....................................... 104
6.2.5 Continuous SSCF ......................................................................................... 105
6.2.6 Measurement of viable cells density ............................................................ 107
6.2.7 Sugar, fermentation products analyses and mass balance ............................ 107
6.3 Results and Discussion............................................................................................ 107
viii

6.3.1 Enzymatic hydrolysis comparison of SHF and SSCF .................................. 107
6.3.2 Fermentation comparison between SHF and SSCF ..................................... 109
6.3.3 Fermentation kinetics of batch SSCF ............................................................112
6.3.4 Continuous SSCF ..........................................................................................113
6.3.5 Comparison of continuous SSCF with batch SHF and batch SSCF .............117
6.3.6 Discussion on the continuous SSCF process design .................................... 120
6.4 Conclusions ............................................................................................................. 123
CHAPTER VII NOVEL INTEGRATED BIOLOGICAL PROCESSES FEATURING HIGH
ETHANOL PRODUCTIVITY, ENZYME RECYCLING AND YEAST CELLS REUSE . 124
Abstract ......................................................................................................................... 124
7.1 Introduction ............................................................................................................. 125
7.2 Materials and Methods ............................................................................................ 129
TM
pretreated corn stover ................................................................... 129
7.2.1 AFEX
7.2.2 Microorganisms and seed culture preparation.............................................. 129
7.2.3 BCRL SHF and SSCF processes .................................................................. 129
7.2.4 Conventional SHF and SSCF processes....................................................... 131
7.2.5 Measurement of viable cell density .............................................................. 131
7.2.6 HPLC analyses and mass balance ................................................................ 132
7.3 Results and discussion ............................................................................................ 132
7.3.1 Effect of solids loading................................................................................. 132
7.3.2 Optimization of fermentation conditions ..................................................... 136
7.3.3 Optimization of enzyme loading for cycle 2 of BCRL SHF process ........... 136
7.3.4 BCRL SHF process results ........................................................................... 137
7.3.5 BCRL SSCF process results ......................................................................... 140
7.3.6 Comparisons of process ethanol productivity and ethanol yield of different
process ................................................................................................................... 141
7.4 Conclusions ............................................................................................................. 149
TM

PRETREATED CORN STOVER AT LOW SOLIDS
CHAPTER VIII CBP OF AFEX
LOADING ............................................................................................................................ 150
Abstract ......................................................................................................................... 150
8.1 Introduction ............................................................................................................. 151
8.2 Materials and Methods ............................................................................................ 152
TM
8.2.1 AFEX pretreated corn stover ................................................................... 152
8.2.2 Particle size reduction .................................................................................. 152
8.2.3 CBP microbe and seed culture preparation .................................................. 152
8.2.4 CBP fermentation ......................................................................................... 153
8.2.5 Simultaneous Saccharification and Co-fermentation ................................... 153
8.2.6 Measurement of glucan and xylan conversions ........................................... 154
TM
8.2.7 Preparation of washed AFEX -CS ............................................................ 154
8.3 Results and discussion ............................................................................................ 155
8.3.1 Effect of temperature on CBP performance ................................................. 155
8.3.2 Effect of inoculation size on CBP performance ........................................... 159
ix

8.3.3 Effect of nutrients and pH on CBP performance.......................................... 160
TM
TM
8.3.4 Comparison of CBP on washed AFEX -CS and unwashed AFEX -CS 162
8.3.5 Effect of particle size on CBP performance ................................................. 162
8.3.6 Comparison between CBP by C. phytofermentans and SSCF by S. cerevisiae
424A ...................................................................................................................... 164
8.4 Conclusion .............................................................................................................. 165
TM

PRETREATED CORN STOVER AT HIGH SOLIDS
CHAPTER IX CBP OF AFEX
LOADING ............................................................................................................................ 167
Abstract ......................................................................................................................... 167
9.1 Introduction ............................................................................................................. 168
9.2 Materials and Methods ............................................................................................ 169
TM
9.2.1 AFEX
pretreated Corn Stover ................................................................. 169
9.2.2 Particle size reduction .................................................................................. 169
TM
9.2.3 Preparation of washed AFEX -CS……………………………………....169
9.2.4 CBP microbe seed cultures preparation ....................................................... 169
TM
9.2.5 Water extract of AFEX -CS preparation ................................................... 169
9.2.6 CBP fermentation ......................................................................................... 169
9.2.7 Measurement of glucan and xylan conversions ........................................... 170
9.3 Results and discussion ............................................................................................ 170
9.3.1 CBP at 4% glucan loading with and without nutrients supplementation ..... 170
9.3.2 Effect of solids loading on CBP performance .............................................. 171
9.3.3 Effects of acetate and ethanol on CBP performance and sugar consumption
............................................................................................................................... 173
9.3.4 Effect of degradation products on fermentation ........................................... 176
9.4 Conclusions ............................................................................................................. 179
CHAPTER X PERSPECTIVES ........................................................................................... 180
10.1 Techno-economic analyses of various processes .................................................. 180
10.2 Perspectives on the BCRL integrated biological processes improvement ............ 182
10.2.1 Ethanol yield enhancement ........................................................................ 182
10.2.2 Thermo-tolerant xylose-fermenting ethanologen ....................................... 183
10.2.3 Improvement of ethanol yield for the BCRL SHF process ........................ 184
10.2.4 Improvement of enzyme adsorption ........................................................... 184
10.2.5 Alleviation of solids accumulation ............................................................. 184
10.3 Perspectives on the solids inhibitory effect on fermentation ................................ 185
10.4 Perspectives on the CBP process .......................................................................... 186
APPENDIX MODIFIED NREL 2011 TECHNO-ECONOMIC MODEL PARAMETERS 189
REFERENCES ..................................................................................................................... 200

x

LIST OF TABLES

Table 1 Comparison of SHF/SHcF, SSF/SSCF and CBP .......................................................... 14
TM

Table 2 Major degradation products (µg analyte/g substrate) in AFEX
TM

Table 3 Nutrients in 6% glucan loading AFEX

-CS ........................... 22

treated corn stover hydrolysate. ................... 23
TM

Table 4 Fermentation performance (SHF) on AFEX

treated biomass .................................... 24

Table 5 Xylose conversion experiment design for investigating the effects of water extract,
a
ethanol, cell culture medium and fermentation metabolites on xylose consumption ................ 31
Table 6 Summary of parameters for the experiments examining the effects of the fraction of total
cellulases loaded and temperature during step 1, pH and initial OD on two-step SSCF
performance .................................................................................................................................. 51
TM

a

Table 7 Summary of SHF and two-step SSCF experiments on 6% glucan loading AFEX -SG
....................................................................................................................................................... 64
a

Table 8 Concentration and composition of pre-hydrolyzed solids ............................................. 75
Table 9 Byproduct profiles of continuous SSCF at different flow rates..................................... 115
Table 10 Summary of experimental parameters for the effects of temperature, inoculation size,
TM
a
nutrients, and pH on CBP of AFEX -CS by C.phytofermentans. ......................................... 155
Table 11 Effects of temperature, inoculation size and nutrients on lactate, acetate, formate, and
TM
a
ethanol production during CBP of AFEX -CS by C.phytofermentans. ................................ 157
Table 12 Effect of pH on lactate, acetate, formate, and ethanol production during CBP of AFEXa
CS by C. phyfermentans. ........................................................................................................... 158
TM

Table 13 Summary of AFEX

-CS water extract (WE) effect on cellobiose fermentation. .... 179

xi

LIST OF FIGURES

Figure 1 Fuel ethanol production in the U.S. ................................................................................ 2
Figure 2 Biochemical process for cellulosic ethanol production .................................................... 3
TM

Figure 3 AFEX

pretreatment. .................................................................................................... 5

Figure 4 Process configurations (adaptated from (Lynd, 1996a)). ................................................. 7
Figure 5 Xylose metabolic pathway in S. cerevisiae 424A (LNH-ST). ......................................... 8
TM

Figure 6 Mass balance of SHF process based on AFEX
pretreatment (adapted from (Lau &
Dale, 2009)). ................................................................................................................................. 17
Figure 7 Fermentation performances of S. cerevisiae 424A (a,b) and E. coli KO11 (c,d) in 6%
TM
(w/w) glucan loading AFEX -CS hydrolysate (a,c) and YEP medium (b,d). .......................... 33
TM

pretreated corn stover on specific xylose
Figure 8 Effect of water extract of AFEX
consumption rates of S. cerevisiae 424A (a) and E. coli KO11 (b). ............................................. 36
Figure 9 Effect of ethanol on specific xylose consumption rates of S. cerevisiae 424A (a) and E.
coli KO11 (b). ............................................................................................................................... 37
Figure 10 Effect of culture medium (YEP and hydrolysate) and culture time (24h and 48h) on
specific xylose consumption rates of S. cerevisiae 424A (a) and E. coli KO11 (b). .................... 39
Figure11 Cell growth and xylose consumption of S. cerevisiae 424A (a) and E. coli KO11(b)
TM
during fermentations in YEP medium, AFEX -CS hydrolysate and hydrolysate with YEP. ... 41
TM

Figure 12 Understanding the xylose fermentation problem during AFEX -CS hydrolysate
fermentation by S. cerevisiae 424A. ............................................................................................. 42
Figure 13 Schematic diagram of traditional SSCF (A) and two-step SSCF (B). ........................ 50
Figure 14 Traditional SSCF (A) vs. two-step SSCF (B). ............................................................. 51
xii

Figure 15 Effects of various factors on xylose fermentation and final ethanol concentration
during two-step SSCF. .................................................................................................................. 54
Figure 16 Effect of adding 2% (w/w) Tween 80 on xylose consumption, ethanol production (a),
and glucose fermentation (b) during two-step SSCF. ................................................................... 56
TM

-SG on fermentation performance in 6%

Figure 17 Effects of EH residue and washed AFEX
TM

glucan loading AFEX

-SG hydrolyzate. ................................................................................... 58

Figure 18 Mass balance analysis for SHF (A), two-step SSCF (B1), and two-step SSCF with 2%
Tween 80 (B2). ............................................................................................................................. 62
TM

Figure 19 SSCF of AFEX

-CS by applying two processes. ..................................................... 69

Figure 20 Comparison of xylose fermentation (a) and cell viability (b) during SHF and
traditional SSCF. ........................................................................................................................... 70
Figure 21 Effect of pre-hydrolysis time on xylose fermentation (a) and cell viability (b) during
traditional SSCF. ........................................................................................................................... 73
Figure 22 Effect of pre-hydrolyzed solids on xylose fermentation (a) and cell viability (b). ...... 76
Figure 23 Xylose consumption (a) and cell growth (b) during fermentations of hydrolysate from
6 h pre-hydrolysis and 24 h pre-hydrolysis................................................................................... 77
Figure 24 Effect of glucose supplementation on xylose consumption (a) and cell viability (b)
during 6 h pre-hydrolysis SSCF.................................................................................................... 79
TM

Figure 25 SSCF of AFEX

-CS using Accellerase enzymes and 6 h pre-hydrolysis. ............... 80

Figure 26 Mass balance for the SHF (a), SSCF with 24 h pre-hydrolysis (b), and SSCF with 6 h
TM
pre-hydrolysis using Accellerase enzymes (c) on AFEX -CS. ................................................. 82
Figure 27 Effect of ethanol concentration on glucose (a) and xylose (b) release during enzymatic
hydrolysis at different temperature. .............................................................................................. 84
TM

Figure 28 Glucose (a), xylose (b) and ethanol (c) profiles during SSCF of unwashed AFEX CS at different glucan loading (GL). ............................................................................................ 90
TM

Figure 29 Xylose and ethanol profiles during fermentations of 5% glucan loading AFEX
TM

-CS

TM

hydrolysate, 7% glucan loading AFEX -CS hydrolysate and 9% glucan loading AFEX -CS.
....................................................................................................................................................... 92
Figure 30 Xylose, ethanol and OD600 profiles during fermentations of 5% glucan loading
TM

AFEX

TM

-CS hydrolysate, 7% glucan loading AFEX
xiii

-CS hydrolysate and 9% glucan loading

TM

AFEX -CS, with glucose and xylose concentrations all adjusted to 74.5 and 43.0 g/L,
respectively. .................................................................................................................................. 94
Figure 31 Comparison of final xylose and ethanol concentrations during SSCF of unwashed
TM
TM
TM
AFEX -CS, unwashed AFEX -CS with supplementation of YEP and washed AFEX -CS
with supplementation of YEP at different glucan loading. ........................................................... 95
Figure 32 Effect of solids loading on monomeric sugar conversion (a) and ethanol yield (b)..... 97
Figure 33 Effect of SSCF solids loading on minimum ethanol selling price (MESP). ................ 99
Figure 34 Continuous SSCF process diagram ............................................................................ 105
Figure 35 Time courses for polymeric (Poly-), oligomeric (Oligo-) and monomeric (Mono-)
glucose (a) /xylose (b) during enzymatic hydrolysis of batch SSCF and SHF........................... 108
Figure 36 Fermentation comparison of batch SSCF and SHF. ................................................... 110
Figure 37 Consumed glucose/xylose, ethanol production, glucose/xylose consumption rate, and
ethanol production rate during batch SSCF. ............................................................................... 112
Figure 38 Comparison of three flow rates (12, 24, and 48 ml/h) on glucose (a) and xylose (b)
consumption, ethanol production (c), viable cell density (d), glucose-to-ethanol (e) and xyloseto-ethanol (f) conversions during continuous SSCF. .................................................................. 116
Figure 39 Process ethanol productivity comparisons of different flow rates during continuous
SSCF (a), and productivity comparisons of SHF, batch SSCF and continuous SSCF (b). ........ 117
Figure 40 Mass balance of continuous SSCF, batch SSCF(A) and SHF(B). ............................. 118
Figure 41 Levenspiel plots of –Fglc.o/rglc against Xglc (a) and –Fxyl.o/rxyl against Xxyl (b). .... 122
Figure 42 Flow chart of the Biomass Conversion Research Laboratory (BCRL) SHF process (a)
and SSCF process (b). ................................................................................................................. 128
Figure 43 Tank setup and volume change for BCRL SHF and SSCF processes. ...................... 130
Figure 44 Effect of solids loading on 24 h enzymatic hydrolysis, the following 24 h SSCF or
hydrolysate fermentation (a), viable cell density and cell viability after 24 h SSCF (b)............ 133
Figure 45 Effect of initial OD (a), pH (b) and temperature (c) on xylose consumption, ethanol
production and cell viability (d) during the first cycle 24 h SSCF at 7% (w/w) glucan loading.134
Figure 46 Optimization of enzyme loading for cycle 2 during the BCRL SHF process. ........... 135
Figure 47 BCRL SHF process results. ....................................................................................... 138

xiv

Figure 48 Effect of enzyme loading profile on ethanol production (a), sugar to ethanol
conversions (b), and viable cell density (c) during BCRL SSCF process. ................................. 142
Figure 49 Mass balance of BCRL SHF and SSCF processes .................................................... 144
Figure 50 Enzyme activities of proteins in the removed hydrolysate after enzymatic hydrolysis
during fast SHF process. ............................................................................................................. 146
Figure 51 Effect of enzymatic hydrolysis residual solids concentration on fermentation in
hydrolysate. ................................................................................................................................. 147
Figure 52 Comparisons of process ethanol productivity (a) and ethanol yield (b) of different
processes. .................................................................................................................................... 148
Figure 53 Effects of temperature (a), inoculation size (b), nutrient addition (c) and pH (d) on
TM
glucan and xylan conversion during CBP of AFEX -CS by C.phytofermentans. .................. 156
Figure 54 Comparison of sugar conversions (a) and product generation (b) during CBP on
TM
TM
washed AFEX -CS and un-washed AFEX -CS.................................................................. 161
TM

Figure 55 Effect of particle size of AFEX -CS on sugar conversion and product generation
during CBP by C. phytofermentans. ........................................................................................... 163
Figure 56 Comparison of sugar conversions achieved by CBP using C. phytofermentans and
TM
SSCF using S. cerevisiae 424A on AFEX -CS. ..................................................................... 164
Figure 57 Mass balance comparison between CBP by C. phytofermentans and SSCF by S.
TM
cerevisiae 424A on AFEX -CS .............................................................................................. 166
TM

Figure 58 CBP of 4% glucan loading AFEX -CS with (w/) and without (w/o) nutrients
supplementation. ......................................................................................................................... 172
Figure 59 Effect of solids loading on sugar conversions (a) and products production (b) during
CBP. ............................................................................................................................................ 174
Figure 60 Effect of ethanol and acetate on sugar conversions of CBP. ...................................... 175
Figure 61 Effect of ethanol and acetate on glucose and xylose consumption. .......................... 176
TM

Figure 62 Effect of AFEX -CS water extract (WE) on cellobiose consumption (a), cell growth
(b), ethanol (c) and acetate (d) production during cellobiose fermentation by C.phytofermentans.
..................................................................................................................................................... 178
Figure 63 Techno-economic analyses of various processes. ...................................................... 181
Figure 64 Rearranged BCRL SHF process for solving accumulated solids problem................. 185
xv

ABBREVIATIONS

AFEX

Ammonia Fiber Expansion

BCRL

Biomass Conversion Research Laboratory

CBP

Consolidated bioprocessing

CFU

Colony forming unit

CS

Corn Stover

CSL

Corn steep liquor

CSTR

Continuous stirred tank reactor

EH

Enzymatic hydrolysis

EtOH

Ethanol

Glc

Glucose

LCC

Lignin-carbohydrates complex

MESP

Minimum ethanol selling price

SG

Switchgrass

SHF/SHcF

Separate hydrolysis and fermentation/co-fermentation

SSCF

Simultaneous saccharification and co-fermentation

SSF

Simultaneous saccharification and fermentation

WE

Water extract

XD

Xylitol dehydrogenase
xvi

XK

Xylulokinase

XR

Xylose reductase

Xyl

Xylose

YE

Yeast extract

YEP/YP

Yeast extract and peptone medium

xvii

CHAPTER I BACKGROUND AND GENERAL INTRODUCTION
1.1 Cellulosic ethanol production
1.1.1 Background
Development of renewable transportation fuels (biofuels) has gained much momentum due
to the increasing energy demand, limited reserves of petroleum, required energy independence as
well as environmental concerns (Farrell et al., 2006; Greene & Council, 2004; Hallac &
Ragauskas, 2011; Wyman, 2008; Wyman, 2007). Currently, the US is consuming 20 million
barrels of crude oil per day with over 70% going to transportation (Gray et al., 2006; Shaw et al.,
2008) and the transportation area almost totally relies on petroleum (Wyman, 2007). The largescale use of fossil fuels is recognized as a major factor resulting in the green house effect
(Houghton et al., 2001). Since the use of fossil fuels is not sustainable, fossil fuels utilization as
the main world’s energy is greatly threatened (Yu et al., 2003). In such an atmosphere,
substitutes to fossil fuels for transportation are badly needed. Bioethanol (CH3-CH2-OH), which
is renewable and produces little green house gas when consumed (Hansen et al., 2005), is
thought to be a promising liquid biofuel to replace a large amount of petroleum in transportation.
Fuel ethanol production in the US keeps increasing from year to year (Fig. 1). Today, over
13 billion gallons of ethanol are produced per year by the US (www. http://www.eia.gov/) with
corn as the major substrate. However, corn ethanol is not sufficient to make a big impact on the
transportation fuel use (Yang & Wyman, 2008). Besides, fuel production using corn, which is a
food, is controversial. The US government’s goal is to replace 30% of the gasoline consumption
by biofuels by 2030 (Perlack et al., 2005), which requires around 60 billion gallons ethanol.
Lignocellulosic biomass, such as corn stover (CS), swithgrass (SG), agricultural and forestry
residues, which amounts to 1.3 billion tons (corresponding to about 80 billion gallons ethanol)
1

per year produced by the US (Perlack et al., 2005), can offer sufficient ethanol volume to achieve
the government’s goal and make a big impact on the petroleum use in the transportation area.
Corn ethanol production is a mature technology with glucose from enzymatic liquefaction and
saccharification of starch fermented to ethanol by Saccharomyces yeasts (Gray et al., 2006).
Cellulosic ethanol technology is more complicated and not yet fully commercialized.
14000
Fuel ethanol

Million Gallons

12000
10000
8000
6000
4000
2000
0
1980

1985

1990

1995
2000
Year

2005

2010

Figure 1 Fuel ethanol production in the U.S.
1.1.2 Cellulosic ethanol production
Cellulosic ethanol refers to the ethanol produced from lignocellulosic biomass.
Lignocellulosic biomass typically contains 36-61% of cellulose and 13-39% of hemicelluloses
(Olsson & Hahn-Hägerdal, 1996), which are the sugars for ethanol production. These sugars are
buried in a complex carbohydrate-lignin network (Chundawat et al., 2011a), which is highly
recalcitrant to enzymes hydrolysis and thus renders pretreatment necessary for the biochemical
route of cellulosic ethanol production (Yang & Wyman, 2008). Biochemical cellulosic ethanol
production process is composed of 7 steps (Fig. 2):
 Lignocellulosic biomass (feedstock) production and harvest.
 Pretreatment to disrupt the structure of plant cell wall making substrates more digestible

2

to enzymes (Yang & Wyman, 2008).
 Enzyme production. To hydrolyze cellulose to glucose, at least three types of cellulases
are needed: endoglucanase (1,4-β-D-glucan glucanohydrolase; EC 3.2.1.4), exoglucanase
(1,4-β-D-glucan cellobiohydrolase; EC 3.2.1.91) and β-glucosidase (cellobiase or β-Dglucoside glucohydrolase; EC 3.2.1.21) (Kabel et al., 2006). Xylanases are required for
hydrolyzing hemicelluloses (Zhong et al., 2009).
 Enzymatic Hydrolysis (EH) of pretreated feedstock to fermentable sugars.
 Fermentation of hexose to ethanol. The most widely used genus for ethanol production is
Saccharomyces yeasts. The reaction can be expressed as (Hamelinck et al., 2005):
C6H12O6

2C2H5OH + 2CO2

 Fermentation of pentose to produce ethanol.
3C5H10O5

5C2H5OH + 5 CO2

The theoretical yield is 0.51 g ethanol / g sugar (Hamelinck et al., 2005).
 Ethanol Recovery.

Feedstock
harvest & size
reduction

Pretreatment

Four Biological Steps
C6 sugars
Fermentation
Enzymatic
Hydrolysis
C5 sugars
Fermentation
Enzyme
Production

Ethanol
Recovery

Figure 2 Biochemical process for cellulosic ethanol production
1.1.3 Biomass pretreatment
Linocellulosic biomass is composed of three major polymers: cellulose, hemicelluloses
and lignin. Cellulose is linear and comprises β-D-glucopyranose moieties linked by β-(1,4)
3

glycosidic bonds (Somerville et al., 2004). Cellulose chains bundled together form microfibrils
which are further packed to create cellulose fibers (Chundawat et al., 2011a). Hemicelluloses are
heterogeneous polymers consisting of various sugars including xylose, arabinose, galactose,
glucose and mannose (Saha, 2003). Hemicelluloses act like a “coat” to cellulose (Agbor et al.,
2011). Lignin glues the components of lignocellulosic biomass together by forming lignincarbohydrates complex (LCC) linkages and thus makes the lignocellulosic biomass recalcitrant
to microbial/enzymatic attack. Three major mono-lignols for forming lignin are p-coumaryl
alcohol (H lignin), coniferyl alcohol (G lignin) and sinapyl alcohol (S lignin) (Vanholme et al.,
2010). Due to these plant cell wall properties, the major factors affecting enzymatic hydrolysis of
lignocellulosic biomass include high crystallinity of cellulose, low accessible surface area for
enzymes, high lignin content and high degree of polymerization. The role of pretreatment is to
modify the properties of the lignocellulosic biomass for high digestibility by enzymes (e.g.
decrystallize cellulose, increase accessible surface area, remove lignin, de-polymerization, alter
lignin structure and remove hemicelluloses).
Many pretreatment technologies have been invented, such as physical pretreatment
comminution and biological pretreatment using white-rot fungi (Agbor et al., 2011). However,
the most efficient and cost-effective ones belong to physicochemical pretreatment such as dilute
acid, steam explosion, liquid hot water, lime, ammonia recycle percolation (ARP), soaking
TM

aqueous ammonia (SAA) and ammonia fiber expansion (AFEX

) (Poth et al., 2011).

Depending on the choice of pretreatment the composition and properties of the substrate may
change and thereby affect the digestibility and fermentability during the downstream processes
(Yang & Wyman, 2008). During physicochemical pretreatment, it is almost impossible to avoid
the formation of degradation products, which are inhibitory to enzymatic hydrolysis and

4

fermentation. The degradation products could derive from glucose such as 5-hydroxymethyl-2furaldehyde (HMF), or from xylose such as furfural, or from lignin such as phenolics (Almeida
et al., 2007). Degradation compounds are also formed by cleaving LCC linkages (e.g. acetic acid)
(Almeida et al., 2007). Ideal pretreatment technology should not only provide a biomass
substrate with high enzyme digestibility and high fermentability but also have low capital and
operating costs. Eggeman et al. (Eggeman & Elander, 2005) techno-economically compared the
TM

existing leading pretreatments and concluded that dilute acid and AFEX

are the two most

cost-effective pretreatments with a potential minimum ethanol selling price (MESP) of $1.35 and
$1.41 per gallon, respectively.
TM

1.1.4 Ammonia fiber expansion (AFEX
TM

AFEX

) pretreatment

is an alkaline thermo-chemical pretreatment using ammonia as the catalyst (Balan

et al., 2010; Morales-Rodriguez et al., 2011). As shown in Fig. 3, untreated biomass with certain
TM

moisture content and ammonia are fed into the preheated AFEX

Gaseous
Ammonia

Liquid
Ammonia

Biomass

reactor, in which

Reaction

Expansion

Pretreated
Biomass

Temperature: 40-180 oC
Ammonia-to-biomass ratio: 0.3-2
Water-to-biomass ratio: 0.1-2
Pretreatment time: 5-60 min

TM

Figure 3 AFEX
pretreatment.
This figure is in color. For interpretation of the references to color in this and all other figures,
the reader is referred to the electronic version of this dissertation.

5

pretreatment reactions occur under certain temperature and pressure conditions. After 5-60 min,
the pressure is released and ammonia is recycled. The pretreated biomass is typically darker than
the untreated biomass.
TM

AFEX

pretreatment is a “dry to dry” process with no liquid stream. Two major reactions
TM

occur during AFEX

are ammonolysis and hydrolysis, which cleave linkages in carbohydrate-

lignin network. Some ammonia soluble lignin is extracted and relocated on the surface of the
biomass (Chundawat et al., 2011c). Pores in the biomass are formed. Lignocellulosic ultra and
TM

macro structures are altered (Chundawat et al., 2011c). After AFEX

pretreatment, cellulose is

decrystallized, accessible surface area is increased, and hemicellulose is partially depolymerized.
TM

Cellulose and hemicellulose conversions for most AFEX

treated biomass during enzymatic

hydrolysis achieved greater than 90% (Yang & Wyman, 2008). Degradation products formed
TM

during AFEX

pretreatment showed mild inhibitory effect on yeast fermentation (Lau & Dale,
TM

2009). The level of organic acids like degradation products in AFEX

pretreated corn stover is

much lower than in diluted acid pretreated corn stover (Chundawat et al., 2010), probably due to
TM

the ammonolysis reaction during AFEX

converting the compounds into amide rather than
TM

acid forms. As a result, the acetamide to acetic acid ratio in AFEX
TM

around 5:1 (Chundawat et al., 2010). Moreover, AFEX
plant biomass for yeast fermentation (Lau et al., 2012).

6

pretreated corn stover is

preserves the nutrients naturally in

1.1.5 Process configurations based on four biological steps
SHF

SHcF

SSF

SSCF

CBP

Enzymes
production
Enzymatic
hydrolysis
C6 sugar
fermentation
C5 sugar
fermentation

Figure 4 Process configurations (adaptated from (Lynd, 1996a)).
SHF: separate hydrolysis and fermentation; SHcF: separate hydrolysis and co-fermentation;
SSF: simultaneous saccharification and fermentation; SSCF: simultaneous saccharification and
co-fermentation; CBP: consolidated bioprocessing.

Biological events (Fig. 2) involved in the processing of harvested plant biomass for ethanol
production include saccharolytic enzyme production, enzymatic hydrolysis, hexose fermentation
and pentose fermentation (Lynd, 1996). Depending on how the four biological events are
integrated, the process configuration can be separate hydrolysis and fermentation/cofermentation (SHF/SHcF), simultaneous saccharification and (co)fermentation (SSF/SSCF), or
consolidated bioprocessing (CBP) (Fig. 4) (Lynd, 1996). In the configuration of SHF/SHcF, the
four biological events are conducted separately (co-fermentation performs hexose and pentose
fermentation together). In the configuration of SSF/SSCF, enzymatic hydrolysis and
fermentation are conducted in the same bioreactor. CBP carries out the four biological events
together in a single bioreactor. This dissertation work involves SHcF, SSCF and CBP. But I
would like to use term SHF to refer to SHcF since SHcF belongs to the category of SHF.
1.1.5.1 Co-fermentation of glucose and xylose

7

Xylose in the form of xylan/hemicelluloses is the second most abundant sugar in the
lignocellulosic biomass (Girio et al., 2010). From an economic-viability point of view, it is
critical to convert xylan/xylose as well as cellulose/glucose to ethanol (Weber et al., 2010).
During the past decades, numerous microorganisms, such as Escherichia coli (Olsson et al.,
1995), Zymomonas mobilis (Zhang et al., 1995), and Schefferosomyces (Pichia) stipitis (Jeffries
et al., 2007) have been genetically modified to accomplish this goal. E.coli can natively consume
xylose as well as arabinose (Weber et al., 2010). Genetically modified strain E.coli KO11
possesses pyruvate decarboxylase and alcohol dehydrogenase genes (PET operon) from Z.
mobilis, which makes it a homoethanolic fermentation strain (Ingram et al., 1999). Z. mobilis
AX101 with the seven xylose and arabinose metabolism genes integrated into the genome was
also able to ferment pentose to ethanol with high ethanol yield (Mohagheghi et al., 2002). S.

NADP+ (NAD+) control is the key for ethanol production using S. stipitis

Xylitol
NAD+
NADH

Xylulose

Xylulose-5-phosphate

Xylose metabolic pathway in 424A

Xylose isomerase

Xylose
Xylitol
Xylulokinase
Reductase Dehydrogenase

Xylose
NADPH (NADH) stipitis is also a native xylose consumer. Dissolved oxygen

(Skoog & Hahn-Hagerdal, 1990).
Saccharomyces cerevisiae, the oldest ethanol producer used
by humans has also been engineered to ferment xylose
(Olofsson et al., 2008b; Sedlak & Ho, 2004). For instance, S.
cerevisiae 424A (LNH-ST) with its chromosome integrated
with multiple copies of xylose reductase (XR) and xylitol
dehydrogenase (XD) genes from S. stipitis and endogenous
xylulokinase gene (XK) was able to ferment xylose with high

Pentose Phosphate Pathway

Figure 5 Xylose metabolic
pathway in S. cerevisiae
424A (LNH-ST).
The figure is in color.

rates (Sedlak & Ho, 2004). However, due to the fact that

8

+

NADPH is the preferable cofactor for XR and NAD is the sole cofactor for XD (Fig. 5), the
redox of those two reactions is imbalanced, which results in xylitol production (Ha et al., 2011).
Mutation of XR and screen of the XR mutants preferring NADH is a way to solve this problem
but this approach reduced the xylose fermentation rate (Bengtsson et al., 2009; Watanabe et al.,
2007). Ha et al. applied both wild XR and XR mutants, which turned out to be a better way to
reduce xylitol production and at the same time keep high xylose consumption rate (Ha et al.,
2011). Strains with the xylose isomerase gene instead of XR and XD genes could also bypass the
redox imbalance problem (Kuyper et al., 2005; Madhavan et al., 2009).
After all those endeavors, the xylose fermentation rate, however, is still much lower
compared to glucose fermentation. One reason for slow xylose consumption is that S. cerevisiae
does not have a specialized xylose transport system. Xylose relies on the hexose transporters to
be transported into the cells (Bertilsson et al., 2008). However, the affinities of hexose
transporters are much lower for xylose when compared to glucose. As a result, xylose
fermentation begins only after glucose is nearly depleted in the fermentation broth. It is
interesting to note that the expression of hexose transporters is regulated by the concentration of
glucose (Bertilsson et al., 2008; Diderich et al., 1999). In order to create the optimum conditions
for xylose uptake, the glucose concentration must be maintained at a low but not zero level
(Bertilsson et al., 2008). Recently, Ha et al. constructed a recombinant xylose-fermenting S.
cerevisiae with a cellodextrin transporter and intracellular β-glucosidase (Ha et al., 2011). The
strain is able to co-ferment cellobiose and xylose and thus bypasses the problem of cofermenting glucose and xylose since the glucan in the lignocellulosic biomass can be degraded to
just cellobiose.

9

Lau et.al (Lau et al., 2010) compared ethanologens E. coli KO11, S. cerevisiae
TM

424A(LNH-ST) and Z. mobilis AX101 in AFEX

pretreated corn stover hydrolysate

fermentation and concluded that S. cerevisiae 424A(LNH-ST) is the best among the tested
strains in terms of xylose fermentation, tolerance to biomass degradation products and ethanol
yield.
1.1.5.2 Separate enzymatic hydrolysis and (co)fermentation (SHF/SHcF)
Currently, the most efficient cellulase enzymes are produced from Trichoderma reesei.
During past a few years, the cost of cellulase enzymes has been reduced by more than tenfold
(Olofsson et al., 2008a). The current cellulase enzymes cost estimated by National Renewable
Energy Laboratory (NREL) is about $4.24/kg protein (Humbird et al., 2011). Nevertheless, the
enzyme expense is still one of the major outlays for cellulosic ethanol production, accounting for
15.7% of the total cost based on the near future technology (Humbird et al., 2011). Enzyme
companies such as Novozymes and Genencor are leading the production of cellulase enzymes.
The optimal conditions using Trichoderma cellulases for enzymatic hydrolysis are 4550 °C, pH 4.8 and high agitation rate (250 rpm in shake flask (Lau & Dale, 2009)) (Taherzadeh
& Karimi, 2007). On the other hand, the optimal conditions for ethanol fermentations using
current ethanologens (S. cerevisiae, S. stipitis, Z. mobilis and E. coli) are around 30-37 °C, pH
5.5-7.0, and low agitation rate (around 150 rpm in shake flask (Lau & Dale, 2009)) (Lau et al.,
2010; Taherzadeh & Karimi, 2007). The major advantage of SHF/SHcF compared to SSF/SSCF
is that enzymatic hydrolysis and fermentation can be carried out at their own optimal conditions
(Taherzadeh & Karimi, 2007). However, enzymes during hydrolysis is easily inhibited by its
end-products (sugars), especially during high solids loading enzymatic hydrolysis (Kristensen et
al., 2009; Philippidis & Smith, 1995), which demands somewhat longer hydrolysis time and high

10

enzyme loading to achieve high sugar conversions. Another problem of this process is the high
risk of contamination during enzymatic hydrolysis due to the long reaction time and high sugar
concentrations (Taherzadeh & Karimi, 2007). Enzymatic hydrolysis is the limiting step for SHF,
which determines the overall ethanol yield (Lau & Dale, 2009).
1.1.5.3 Simultaneous saccharification and (co)fermentation (SSF/SSCF)
With enzymatic hydrolysis and sugar fermentation combined, SSF/SSCF removes sugar
inhibition on enzymatic hydrolysis thus increases the hydrolysis sugar yield and reduces
contamination risk. Moreover, SSF/SSCF reduces the overall reaction time and reactor volume
(Kristensen et al., 2009). However, the optimal conditions for hydrolysis and fermentation are
different. SSF/SSCF sacrifices the optimal conditions for both enzymatic hydrolysis and
fermentation. Typically, for Trichoderma reesei cellulases – Saccharomyces SSF system the
temperature is kept at 37°C as a compromise (Dien et al., 2003b). In addition, SSF/SSCF
introduces a new inhibitor (ethanol) for enzymatic hydrolysis. But the inhibitory effect from
ethanol is much lower compared to cellobiose or glucose (Taherzadeh & Karimi, 2007). SSF has
been widely investigated on the steam and diluted acid pretreated biomass while SSCF studies
were limited (Olofsson et al., 2008a).
1.1.5.4 Consolidated bioprocessing (CBP)
Both SHF/SHcF and SSF/SSCF require expensive enzymes production in a separate
bioreactor, which is a major cost barrier for commercialization of cellulosic ethanol (Banerjee et
al., 2010). CBP which carries out the four biological events together in a single bioreactor
reduces the enzyme production cost and at the same time possesses all the advantages of SSCF.
A strong case has been made that CBP is the ultimate low-cost industrial configuration to
produce cellulosic ethanol (Lynd et al., 2005; Sendich et al., 2008). However, currently there’s

11

no perfect microbe or microbe consortium capable of degrading lignocelluloses, utilizing all the
carbohydrates and at the same time producing ethanol at a high yield as CBP requires (Lynd et
al., 2005).
Intensive research has been carried out focusing on two strategies to develop CBP microbes
(Lynd et al., 2005). The first one is genetic engineering of an excellent ethanol producer for
production of cellulases and hemicellulases. So far, many kinds of endoglucanases,
exoglucanases, β-glucosidase, β-xylanases and β-mannanases, of which most are fungal origin,
have been expressed in S. cerevisiae (Crous et al., 1995; Den Haan et al., 2007; Ilmen et al.,
2011; Reinikainen et al., 1992; Romanos et al., 1991; Vanarsdell et al., 1987). With high initial
cell density, direct fermentation of amorphous cellulose resulting in a yield of 0.45 g ethanol per
g substrate corresponding to 88.5% of the theoretical yield was achieved by co-expression of
three types of cellulase in S. cerevisiae (Fujita et al., 2004). The other strategy is genetic
modification of a cellulolytic microbe, which can efficiently degrade lignocellulosic materials, to
make it also capable of producing ethanol at a high yield. Among cellulolytic microbes, the
thermophilic anaerobic bacterium Clostridium thermocellum received the most attention
(Demain et al., 2005). C. thermocellum can degrade cellulose efficiently through a cellulase
complex system called “cellulosome” whose specific activity is much higher compared to
Trichoderma cellulases (Carere et al., 2008). C. thermocellum can also produce ethanol by
utilizing sugars from cellulose degradation but with production of a large amount of by-products:
acetate and lactate (Desvaux, 2006). Xylanases are included in the enzymes secreted by C.
thermocellum which can break down hemicelluloses (Morag et al., 1990). However, C.
thermocellum cannot utilize pentose. So co-culture of C. thermocellum with other thermophilic
microbes which can consume pentose is a good option and received considerable investigation

12

(Demain et al., 2005). Ethanol yield reached as high as 1.8 mol per mol of anhydroglucose on
MN300 cellulose by co-culturing C.thermocellum and Clostridium thermohydrosulfuricum (Ng
et al., 1981). Thermoanaerobacterium saccharolyticum which can ferment xylan and almost all
the soluble sugars from biomass with ethanol, acetate and lactate as the end metabolites is also
considered as a suitable species to co-culture with C. thermocellum (Lynd et al., 2005). Shaw
and co-workers have successfully developed an ethanologic strain T. saccharolyticum ALK2 by
knockout of the genes involved in organic acid formation (acetate kinase, phosphate
acetyltransferase, and L-lactate dehydrogenase) (Shaw et al., 2008). Similar targeted gene
knockouts have also been done on C. thermocellum to eliminate side product synthesis and
improve the ethanol yield (Argyros et al., 2011). By co-culture of the engineered C.
thermocellum with T. saccharolyticum ALK2 on 92 g/L Avicel (pure cellulose), an ethanol titer
of 38 g/L was achieved, which is the highest ethanol titer achieved by thermophilic CBP
microbes to date (Argyros et al., 2011). Nevertheless, the performance of this co-culture system
on real biomass is still unknown. The low tolerance of the two microbes to biomass degradation
products as well as ethanol could be an issue.
The advantages of anaerobic thermophilic microbes over yeast as CBP strains are: 1) they
ferment almost all the sugars in lignocellulosic materials; 2) no air supply is needed since
enzyme production is also anaerobic as well as ethanol fermentation, which reduces costs; 3)
High fermentation temperature accelerates enzymatic hydrolysis, prevents contamination,
facilitates ethanol recovery, and cuts cooling cost. 4) Low cell yield makes most of the sugars
convert to ethanol. 5) Much higher enzyme specific activities require less quantity of enzymes
and in turn less nutrients and sugars. 6) Utilizing hexose and pentose at the same time results in

13

higher pentose consumption rate. 7) Oligosaccharides can also be used, leading to a higher
ethanol yield (Zhang & Lynd, 2005).
The mesophilic anaerobic strain Clostridium phytofermentans whose genome encodes the
highest number of enzymes for degradation of lignocellulosic material among sequenced
clostridial genomes (Weber et al., 2010), is also a promising CBP microbe. It secretes individual
enzymes (non-complexed) like fungi and can degrade both cellulose and hemicelluloses to
fermentable sugars. In addition, C. phytofermentans can consume almost all the sugars present in
lignocellulosic biomass and produce ethanol and acetate as the major products (Warnick et al.,
2002; Weber et al., 2010). It is named Q-Microbe by Qteros Company. The major disadvantage
compared to C. thermocellum is that the enzymes cannot function at their optimal/high
temperature because of the low thermo-tolerance of the microbe.
The summary of the three biological processes is listed in Table 1. CBP is the most
promising process but currently no perfect CBP microbe has been reported in the literature.
SHF/SHcF and SSF/SSCF are the current feasible industrial processes. So far, there is no
systematic comparison study of the three processes on any pretreated biomass.
Table 1 Comparison of SHF/SHcF, SSF/SSCF and CBP
1

SHF/SHcF
Sugar inhibition during enzymatic
hydrolysis

SSF/SSCF

CBP

High

Low

Low

CBP

2

Low

CBP

3

Low

MediumNone
None
None
High
Contamination risk
High
Low
Low
Low
Lowest
Overall processing time
Long
Medium
N/A
N/A
Short
No. of reactors
4-3
2-3
1
1
1
Overall cost
High
Medium
Low
Low
Low
1
2
Footnote: CBP use engineered S. cerevisiae; CBP use mesophilic anaerobic microbes such as
Enzyme loading

High

3

C. phytofermentans; CBP use thermophilic anaerobic microbes such as C. thermocellum and T.
saccharolyticum

14

TM

1.2 Ethanol production from AFEX

pretreated biomass
TM

Before this dissertation work, the studies of cellulosic ethanol production from AFEX

pretreated biomass were almost all based on the SHF configuration. The critical properties of one
pretreated biomass for cellulosic ethanol production include enzyme digestibility and
TM

fermentability. AFEX

pretreated biomass is both high digestible and high fermentable, which

make it a suitable substrate for cellulosic ethanol production.
TM

1.2.1 High digestibility and high fermentability of AFEX

pretreated biomass
TM

High sugar conversions have been achieved on various AFEX

pretreated feedstocks at

high solids loading (6% (w/w) glucan loading) using commercial enzymes, which showed the
TM

high digestibility of AFEX

pretreated biomass (Bals et al., 2010; Krishnan et al., 2010; Lau &
TM

Dale, 2009). For instance, enzymatic hydrolysis of AFEX

treated corn stover at 6% (w/w)

glucan loading reached glucan and xylan conversions as high as 86.7% and 85.0%, respectively
(Fig. 6). However, 19% of the output sugars were oligomers, which cannot be fermented to
ethanol by S. cerevisiae. Moreover, it took 96 h to reach such high sugar conversions, which is
too long for an industrial process.
Fermentability of a pretreated biomass is largely determined by two factors: 1) toxicity of
TM

degradation products; 2) nutrients level. AFEX

pretreatment generates low-toxicity

degradation products and preserves nutrients naturally in the plant biomass. The residual
ammonia (unrecovered) in the pretreated biomass also serves as nutrients (nitrogen source)
TM

during fermentation. Hence, AFEX

pretreated biomass is well known having high

15

TM

fermentability. The degradation products and nutrients profiles in the AFEX

pretreated corn
TM

stover have been quantified, which are shown in Table 2 and Table 3, respectively. AFEX

pretreated corn stover contains much lower concentrations of toxic compounds (such as acetic
acid) compared to diluted acid pretreatment but has higher concentrations of low-toxic amides,
TM

which is the result of ammonolysis reaction during AFEX
TM

toxicity. AFEX

and is the major reason for low

pretreated corn stover hydrolysate has a total amino acid concentration and

ammonia concentration of as high as 1231±44 mg/L and 750±50 mg/L, respectively, with excess
trace elements for yeast fermentation, which renders nutrients supplementation unnecessary
during fermentation (Lau et al., 2012; Lau & Dale, 2009). Due to its high fermentability, almost
all the tested xylose-fermenting ethanologens such as S. cerevisiae 424A(LNH-ST), E. coli
TM

KO11, Z. mobilis AX101 and P. stipitis FPL-061 grew well in AFEX

pretreated biomass

hydrolysate without detoxification and nutrients supplementation (Lau et al., 2010; Li et al.,
2010; Shao et al., 2010).
TM

Lau & Dale (Lau & Dale, 2009) conducted a mass balance study based on AFEX

pretreatment using commercial enzymes for enzymatic hydrolysis and S. cerevisiae 424A(LNHST) for fermentation (Fig. 6). It was found that little sugar was lost during pretreatment (high
sugar recovery), which guaranteed potential high ethanol yield. Both sugar yield during
enzymatic hydrolysis and ethanol yield during fermentation were high. Without washing,
detoxification and nutrients supplementation, 191.5 g ethanol was produced from 1 kg untreated
corn stover with the final ethanol concentration of 40 g/L. The ethanol metabolic yield during
hydrolysate fermentation was 92.9%, which was higher than that (83.6%) for fermentation in

16

- Spezyme CP: 87.9ml (7.7g Protein)
- Novozyme 188: 43.6ml (6.5g Protein)
- Multifect Xylanase: 18.4ml (0.8g Protein)
- Multifect Pectinase: 8.0ml (0.7g Protein)

Untreated Corn
stover: 1Kg
(Dry Weight)
- Glucose: 385.8g
- Xylose: 231.8g

TM

- Yeast innoculum
(Dry Weight):
4.2g

AFEX treated Corn
stover: 1Kg
(Dry Weight)

ENZYMATIC
HYDROLYSIS

- Glucose: 380.4 g
- Xylose: 235.3 g

Time: 96h
Temp.: 50°C

AFEX

FERMENTATION - Ethanol: 191.5g (40g/L)
- Glucose: 286.1g
Time: 168h
- Xylose: 128.6g
Temp.: 30°C
- Gluco-oligomers: 43.8g
- Xylo-Oligomers: 71.5g

Unhydrolyzed solids
- Glucose: 55.8g
- Xylose: 31.7g

- Xylose: 11.0g
- Gluco-oligomers: 43.6g
- Xylo-Oligomers: 72.4g
Conversion:
Glucan: 86.7%
Xylan: 85.0%
Ethanol Metabolic yield: 92.9%

TM

Figure 6 Mass balance of SHF process based on AFEX
pretreatment (adapted from (Lau & Dale, 2009)).
Enzymatic hydrolysis was carried out at 6%(w/w) glucan loading using commercial enzymes. Glucose and xylose co-fermentation
was conducted using S. cerevisiae 424A(LNH-ST).

YEP medium (yeast extract (10 g/L) and peptone (20 g/L) medium). Degradation products were attributed to this increment (Lau &
TM

Dale, 2009). Similar high ethanol metabolic yield using S. cerevisiae 424A(LNH-ST) was also found on other AFEX

pretreated

biomass (Table 4). This was probably due to the fact that some ATP being used for resistance of degradation products such as weak
organic acids rather than for cell growth (Bellissimi et al., 2009), which in turn reduced carbon flux into cell production and hence
increased carbon flux into ethanol production.

17

TM

Table 2 Major degradation products (µg analyte/g substrate) in AFEX
Adapted from (Chundawat et al., 2010).

Acetic acid
Levulinic acid
Furfural/HMF
Acetamide/phenolic amides
Pyrazine/imidazole derivatives
Syringaldehyde
Phenolic acids
Total soluble sugars

-CS

TM

Untreated corn
stover
1610
171
72
3
196
35,927

Dilute-acid treated
corn stover
34770
3649
23640
149
3151
318,300

AFEX
treated
corn stover
4610
24
645
39801
945
11
1183
81,930
TM

1.2.2 Problems of ethanol production from AFEX

pretreated biomass

To make the cellulosic ethanol production process cost-effective, it is essential for
fermentation technology to meet the following criteria: ethanol titer > 40 g/L, ethanol metabolic
-1 -1

yield > 90% and productivity > 1.0 g·L ·h (Dien et al., 2003a). While requirements for
TM

ethanol titer and ethanol metabolic yield have been achieved on some AFEX

pretreated

-1 -1

biomass, the ethanol productivity was far away from 1.0 g·L ·h (Table 4).
The major cause for low ethanol productivity is slow xylose fermentation. Typically, in 6%
TM

(w/w) glucan loading AFEX

hydrolysate, glucose is completely consumed in 18-24 h.

However, it requires more than 96 h to consume around 28 g/L xylose. As a result, the ethanol
-1 -1

productivity can be as high as 4.8 g·L ·h during the glucose fermentation period (Table 4), but
the overall average productivity is low. Xylose fermentation in corn steep liquor (CSL) or YEP,
however, is much faster compared to hydrolysate fermentation (Lau et al., 2010). The reduced
xylose consumption rate in hydrolysate fermentation was attributed to both cell growth inhibition
and xylose metabolism inhibition by degradation products (Lau & Dale, 2009). Xylose
22

TM

fermentation in AFEX

hydrolysates is also affected by nutrient levels in the feedstocks. For

instance, the nutrient level in sugar cane leaf is higher compared to sugar cane bagasse. As a
TM

Table 3 Nutrients in 6% glucan loading AFEX
Adapted from (Lau et al., 2012).

treated corn stover hydrolysate.

Amino
acids
(mg/L)
Asp

Free

Total

8.4±1.7

75.9±1.7

Mg (mg/L)

168.42±3.24

Glu

0.0±2.4

133.8±2.4

Ca (mg/L)

242.87±7.72

Ser

16.8±3.8

104.2±3.8

Mn (mg/L)

2.32±0.53

Gly

5.2±5.8

127.2±5.8

Co (µg/L)

11.3±3.8

His

4.5±2.3

34.3±2.3

Ni (µg/L)

13.5±5.3

Thr

17.6±4.6

98.9±4.6

Cu (µg/L)

116.2±9.3

Arg

17.1±3.2

55.0±3.2

Zn (µg/L)

505.7±51.3

Ala

11.6±2.9

110.2±2.9

Mo(µg/L)

15.9±0.6

Pro

30.4±2.3

108.7±2.3

Fe (µg/L)

296.4±74.5

Tyr

30.0±2.5

28.6±2.5

Val

9.9±2.2

68.8±2.2

Met

2.6±1.9

19.4±1.9

Pantothenic
Acid
Pyridoxine

Ile

7.6±2.2

55.4±2.2

Nicotinic Acid

10.87±1.38

Leu

0.0±3.8

93.6±3.8

Biotin

~0.05

Lys

18.4±1.3

25.7±1.3

Thiamine

~0.66

Phe

15.7±3.8

91.6±3.8

Total AA

195.8±28.3

1231±43.8

Trace elements

Vitamins (µM)

4+

NH
(mg/L)

750±50

23

1.50±0.12
1.26±0.18

TM

Table 4 Fermentation performance (SHF) on AFEX

treated biomass

Average
EtOH
EtOH
EtOH
metabolic titer
produc.
yield
(g/L)
-1 -1
(g·L ·h )
Saccharomyces cerevisiae 424A(LNH-ST)
Corn
1.6
0.24
92.9%
40.0
Stover
TM

AFEX
pretreated
biomass

Max. EtOH
produc.
-1 -1
(g·L ·h )

Xyl.
cons.
(%)

Final
xyl.
conc.
(g/L)

82.2

5.0

Ref.

(Lau & Dale,
2009)
(Krishnan et
al., 2010)
(Krishnan et
al., 2010)
(Zhong et al.,
2009)
(Lu et al.,
2009)
(Li et al.,
2010)
(Shao et al.,
2010)
(Shao et al.,
2010)

Cane leaf

2.4

0.51

91.6%

36.4

87.0

3.8

Sugarcane
Bagasse

3.0

0.28

91.6%

33.7

62.7

12.3

Rice straw

1.7

0.26

95.3%

37.0

69.2

7.4

Poplar

4.8

0.74

93.0%

35.5

78.7

3.5

0.32

86.2%

30.9

57.0

13.0

0.39

93.2%

28.4

72.0

2.1

0.41

89.2%

29.8

79.5

1.6

0.8

0.25

71.7%

29.7

92.4

1.8

(Zhong et al.,
2009)

0.52

0.23

67.7%

27.6

88.4

2.8

(Zhong et al.,
2009)

0.22

96.5%

32.0

31.0

20.0

(Lau et al.,
2010)

0.22

N/A

31.0

10.3

26.0

(Lau et al.,
2010)

Forage
N/A
sorghum
Corn
1.1
silage
Whole
1.2
corn plant
P. stipitis FPL-061
Rice straw
P. stipitis DX-26
Rice straw

Z. mobilis AX101
Corn
0.42
Stover
E. Coli KO11
Corn
0.46
Stover

TM

result, S. cerevisiae 424A(LNH-ST) grew to an OD600 of 14 in AFEX

pretreated sugar cane

leaf hydrolysate, while it grew to an OD600 of 9 in bagasse hydrolysate, which in turn affected
xylose consumption (87% in 72 h vs. 62.7% in 120 h; table 4) (Krishnan et al., 2010).

24

Moreover, if calculation is based on enzymatic hydrolysis time plus fermentation time
(process ethanol productivity), the ethanol productivity value for the current SHF process is
much lower since at least 96 h is required for enzymatic hydrolysis reaction. Besides, high
enzyme loading is required during high solids loading enzymatic hydrolysis, which is a barrier
for commercialization of cellulosic ethanol due to the high cost of cellulase enzymes.
1.3 Research objectives
The integrated biological processes (SSCF&CBP) are strongly believed to have lower cost
TM

compared to the SHF process. But so far, they have not been studied on AFEX
TM

biomass. Moreover, the current SHF processing of AFEX

pretreated

biomass has the following

problems: 1) relatively long enzymatic hydrolysis time; 2) high enzyme loading; 3) slow xylose
fermentation and low ethanol productivity. This dissertation work’s goal is to: 1) study the
TM

integrated biological processes on AFEX

pretreated biomass with the results compared to

SHF process; 2) solve the aforementioned problems in the current process.

25

CHAPTER II QUANTITATIVELY UNDERSTANDING REDUCED
TM

XYLOSE FERMENTATION PERFORMANCE IN AFEX

TREATED

CORN STOVER HYDROLYSATE

Abstract
Reduced xylose fermentation performance has been an issue during fermentation of
TM

AFEX

hydrolysate using Saccharomyces cerevisiae 424A (LNH-ST) or Escherichia coli

KO11. To better understand why fermentation performance is reduced, the effects of compounds
present in the fermentation broth on xylose consumption were quantitatively studied. The
compounds include biomass degradation products, ethanol and fermentation metabolites. The
xylose consumption capability of E. coli KO11 was almost totally inhibited by the presence of
both degradation products and ethanol. On the other hand, for S. cerevisiae 424A, 89% reduction
of xylose consumption rate was found during hydrolysate fermentation. Degradation products,
ethanol and fermentation metabolites were responsible for 32%, 24% and 33% of such reduction,
respectively. Those results suggest that to further improve the xylose fermentation in hydrolysate,
strains should be selected not only for degradation products tolerance but also for ethanol and
fermentation metabolites tolerance.

26

2.1 Introduction
Saccharomyces cerevisiae 424A (LNH-ST) is a genetically engineered xylose-fermenting
strain. Its chromosome is integrated with multiple copies of xylose reductase (XR) and xylitol
dehydrogenase (XD) genes from Schefferosomyces stipitis and an endogenous xylulokinase gene
(XR) (Ho et al., 1999). Escherichia coli KO11 possesses pyruvate decarboxylase and alcohol
dehydrogenase genes (PET operon) from Zymomonas mobilis, which makes it a homoethanolic
fermentation strain (Ingram et al., 1999). Both strains were able to consume xylose quickly in
corn steep liquor or YEP (yeast extract and peptone) medium (Lau et al., 2010). Nevertheless, in
TM

6% (w/w) glucan loading AFEX

treated corn stover hydrolysate, the xylose consumption was

slow for both of the strains.
It is well known that glucose is always fermented first during co-fermentation of glucose
and xylose by S. cerevisiae or E. coli (Weber et al., 2010). S. cerevisiae does not have
specialized xylose transporters (Bertilsson et al., 2008). It depends on hexose transporters to
transport xylose. Due to the much lower affinity of hexose transporters for xylose, glucose is
always transported and hence fermented first (Bertilsson et al., 2008). Xylose consumption in E.
coli is typically repressed due to carbon catabolite repression in the presence of glucose which
reduces the expression of both xylose transporters and key enzymes for xylose metabolism
(Görke & Stülke, 2008; Ren et al., 2009). Therefore, xylose consumption does not start until
TM

glucose is nearly depleted. Thus, during fermentation of high solids loading AFEX

hydrolysate there are always fermentation metabolites and 20-40 g/L ethanol (mostly generated
from glucose fermentation) as well as degradation products present in fermentation broth when
xylose is being consumed (Lau et al., 2010). Degradation products generated from pretreatment
27

were considered to cause poor xylose fermentation (Lau & Dale, 2009; Lau et al., 2010).
However, the effects of ethanol and fermentation metabolites on xylose consumption have not
been quantified, especially in combination with degradation products.
The present study investigates and quantifies the effects of degradation products, ethanol,
TM

fermentation metabolites and nutrients, on xylose consumption in AFEX

corn stover

hydrolysate using S. cerevisiae 424A or E. coli KO11.
2.2 Materials and Methods
TM

2.2.1. AFEX

TM

AFEX

treated corn stover
TM

pretreated corn stover (AFEX

-CS) was supplied by the Biomass Conversion

Research Laboratory (BCRL), Michigan State University, and contained glucan and xylan
TM

contents of 34.1% and 20.4%, respectively. The AFEX

pretreatment procedure has been

described in Chapter I. Pretreatment conditions included: ammonia to biomass loading 1.0 g/g
o

dry biomass, water loading 0.6 g/g dry biomass, temperature 140 C and residence time 15 min.
TM

Unless otherwise stated, AFEX

treated corn stover was used for enzymatic hydrolysis and

fermentation experiments with no washing, conditioning, nutrient supplementation or
detoxification.
TM

-CS hydrolysate preparation

2.2.2 AFEX

TM

AFEX

-CS hydrolysate was prepared by using a commercial enzyme mixture including

Spezyme CP (Genencor Inc, USA) 22.4 mg protein/g glucan (15 FPU/g glucan), Novozyme 188
(Sigma-Aldrich, USA) 38.4 mg protein/g glucan (64 pNPGU/g glucan), Multifect xylanase 2.6
mg protein/ g glucan and Multifect pectinase (Genencor Inc, USA) 4.7 mg protein/ g glucan.
28

Enzymatic hydrolysis was carried out at 6% (w/w) glucan loading (corresponding to 18% (w/w)
o

solid loading) in a 2.0 L baffled flask with 500 g total mixture at pH 4.8, 50 C, and 250 rpm. 50
mg/L chloramphenicol (Cm) was used to avoid microbial contamination. After 96 h hydrolysis,
hydrolysate was centrifuged at 14,000 rpm for 30 min. Supernatant was sterile filtrated through a
0.22 µm filter and kept in a sterile bottle for fermentation use.
TM

2.2.3 AFEX

-CS water extract preparation
TM

Water extract of AFEX
o

-CS was prepared by spraying 1L of preheated distilled water
TM

(around 50 C) on 240 g of AFEX

-CS and then pressing the wetted biomass to a moisture

content of 55 ± 3%. The water extract obtained was sprayed on the biomass again and then the
biomass was pressed again. The extraction process was finished by 3 times of spraying and
pressing. The final water extract obtained was equivalent to a concentration of 8% (w/w) glucan
loading. Sugars contained in the water extract included 0.8 g/L glucose, 1.5 g/L xylose, 2.9 g/L
oligomeric glucose and 8.8 g/L oligomeric xylose. The water extract (concentration: 8% (w/w)
glucan loading) was then sterile filtrated and diluted to different concentrations (2%, 4% and 6%
(w/w) glucan loading) for this study. The water extract was used as a surrogate to represent
TM

soluble degradation products in the AFEX

-CS hydrolysate. Xylose and ethanol were

supplemented in the water extract for xylose conversion experiments.
2.2.4 Microorganisms and seed culture preparation
The genetically engineered xylose-fermenting strain S. cerevisiae 424A(LNH-ST) (Ho et
al., 1999) was obtained from Purdue University. E. coli KO11 was purchased from American
Type Culture Collection (ATCC) with a designated number of 55124. To facilitate comparisons
between the two strains, seed cultures of both strains were prepared on YEP medium (10 g/L
29

yeast extract and 20 g/L tryptone) with 50 g/L glucose in a 250 ml Erlenmeyer flask with
working volume 100 ml. MOPS (pH 7.0) at a concentration of 0.05 M was used for the seed
culture of E. coli KO11. Frozen glycerol stocks were used for inoculation. The cultures of S.
o

o

cerevisiae 424A and E. coli KO11were grown at 30 C and 37 C, respectively and 150 rpm
under micro-aerophilic conditions for 24 h.
2.2.5 Fermentations in YEP medium or hydrolysates
Fermentations on YEP medium (yeast extract 5 g/L; tryptone 10 g/L) or hydrolysate were
conducted at 150 rpm in 125 ml Erlenmeyer flasks with a working volume of 50 ml. 0.05 M
phosphate buffer and 0.05 M MOPS buffer were used for S. cerevisiae 424A and E. coli KO11
fermentations to maintain pH at 5.5 and 7.0, respectively. During E. coli KO11 fermentations,
pH was adjusted twice per day to maintain the set point. Temperatures for S. cerevisiae 424A
o

o

and E. coli KO11 fermentations were 30 C and 37 C, respectively (Lau et al., 2010). To
maintain largely anaerobic conditions and release carbon dioxide, rubber stoppers with a needle
piercing them were used to cap the flasks. S. cerevisiae 424A and E. coli KO11 fermentations
were initiated with an OD600 of 2.0 and 0.5, respectively.
2.2.6 Xylose conversion experiments using non-growing cells
After 24 or 48 h fermentations on YEP medium or hydrolysate, 10 ml fermentation broth
was harvested and centrifuged at 4400 rpm for 10 min. Supernatant was discarded. Cell pellets
were washed using 0.9% NaCl solution. The cell pellets obtained were re-suspended in various
solutions with a total volume of 10 ml. The various solutions included water extract at different
concentrations and water extract with different concentrations of ethanol (Table 5). Either 0.05
M phosphate buffer (pH 5.5, for S. cerevisiae 424A) or 0.05 M MOPS buffer (pH 7.0, for E. coli

30

KO11) was used for pH control. The 10 ml cell and other component mixtures were transferred
o

to a 15 ml screw-capped vial and incubated at 150 rpm and 30 C (for S. cerevisiae 424A) or
o

37 C (for E. coli KO11) for 3 hours. Samples were taken for xylose concentration analysis by
HPLC both before and after the incubation. Specific xylose consumption rate was calculated by
dividing the amount of consumed xylose by the length of reaction time (3 h) and by the density
of the cells.
Experiments were conducted in triplicates and the averages and standard deviations were
given in the figures. Minitab15 Statistical Software (2006 Minitab Inc, Pennsylvania, USA) was
used to perform t-test for the determination of statistical significance. All the values of specific
xylose consumption rate reduction and cell biomass production reduction shown in the figures 8,
Table 5 Xylose conversion experiment design for investigating the effects of water extract,
a
ethanol, cell culture medium and fermentation metabolites on xylose consumption
Effect of
water extract

Parameter
Water extract conc.

b

Ethanol conc. (g/L)
Cells from
CHT for S.
c
cerevisiae
CHT for E. coli
Relevant data

Effect of
ethanol

Effect of culture medium

Effect of ferm.
metabolites

0, 2%, 4%,
6%
0
Hyl.Ferm.

6%

6%

0

0, 20,30,40
Hyl.Ferm.

30
Hyl.Ferm. & YEP Ferm.

35
YEP Ferm.

24 h

24 h

24 h & 48 h

24 h

48 h
Fig. 8

48 h
Fig. 9

24 h & 48 h
Fig. 10

N/A
Fig. 10

a

The experiments were conducted using non-growing cells in a 15 ml screw-capped vial with
10 ml solution. The solution with xylose 30 g/L contained water extract and ethanol at different
concentrations as shown in the table. Either 0.05 M phosphate buffer (pH 5.5, for S. cerevisiae
424A) or 0.05M MOPS buffer (pH 7.0, for E. coli KO11) was used for pH control. The 15 ml
o
o
vial was incubated at 150 rpm and 30 C (for S. cerevisiae 424A) or 37 C (for E. coli KO11) for
3 h.
b
Water extract concentrations were expressed as glucan loading equivalent.
c

Cell harvest time (CHT) from hydrolysate fermentation or YEP fermentation for xylose
conversion experiments.
31

9, 10 &12 are statistically significant (p<0.05).
2.2.7 HPLC analyses
Concentrations of glucose, xylose and ethanol were determined using the Shimadzu HPLC
system equipped with a refractive index detector and a BioRad Aminex HPX-87H column. The
mobile phase was 5 mM aqueous sulfuric acid solution flowing at a rate of 0.6 mL/min. The
o

column temperature was maintained at 60 C.
2.2.8 Measurement of viable cells density
Viable cell density was measured in colony forming unit (CFU) per ml. Samples were
diluted and 20 µL of each diluted sample was plated on YPD agar medium (5 g/L yeast extract,
10 g/L tryptone, 25 g/L glucose, 20 g/L agar and 50 mg/L chloramphenicol). Dilution rate was
varied to make sure the number of colony on a single plate was between 20 to 100. During
dilution, the sample solutions were vigorously vortexed to prevent cell clumping or adhering to
o

the solid biomass. Single colonies were counted after the plate was incubated for 24 h at 30 C
and viable cell density was calculated accordingly.
2.3 Results and discussion
TM

2.3.1 Slow xylose fermentation in AFEX

-CS hydrolysate

TM

AFEX

-CS hydrolysate fermentations and the fermentations in YEP medium with

similar concentrations of glucose and xylose by S. cerevisiae 424A and E. coli KO11 are shown
in Figure 7. Both strains fermented xylose rapidly in YEP medium with approximate 30 g/L of
xylose completely consumed in 48 and 96 h, respectively (Fig. 7b&d). The xylose fermentation
performance in hydrolysate, however, was reduced significantly (Fig. 7a&c). S. cerevisiae 424A
and E. coli KO11 consumed around 21 and 5 g/L xylose in 168 h, respectively, thereby reducing

32

424A-Hydrolysate

Glucose
Ethanol
OD600

50
40

b)

Xylose
CFU

4.0x107
0 20 40 60 80 100 120 140 160 180
Time (h)

50

Xylose
OD600

40
30
20
10
0

2.0x108

40

1.6x108

30

1.2x108

20

8.0x107

10
0

4.0x107
0

d)

KO11-Hydrolysate

Glucose
Ethanol

2.4x108

50

Concentration(g/L) or OD600

60

424A-YEP

60

CFU/ml

8.0x107

10

Concentration(g/L) or OD600

2.0x108

1.2x108

20

c)

2.4x10

1.6x108

30

0

8

Concentration(g/L) or OD600

60

CFU/ml

Concentration(g/L) or OD600

a)

0 20 40 60 80 100 120 140 160 180
Time (h)

60

10

20
30
Time (h)

40

50

KO11-YEP

50
40
30
20
10
0

0 10 20 30 40 50 60 70 80 90 100
Time (h)

TM

Figure 7 Fermentation performances of S. cerevisiae 424A (a,b) and E. coli KO11 (c,d) in 6% (w/w) glucan loading AFEX -CS
hydrolysate (a,c) and YEP medium (b,d).
0.05 M phosphate buffer and 0.05 M MOPS buffer were used for S. cerevisiae 424A and E. coli KO11 fermentations to maintain
pH at 5.5 and 7.0, respectively. YEP medium contained 5 g/L yeast extract and 10 g/L tryptone.
33

both ethanol yield and productivity. To understand the reduced xylose fermentation
performances in hydrolysate, we investigated the effects of possible inhibitors such as
degradation products, ethanol and the growth medium on specific xylose consumption rate. The
inhibitory effect of degradation products on cell growth was also quantified since the xylose
consumption rate was determined by both specific xylose consumption rate and cell density.
(Xylose consumption rate equals to cell density multiply by specific xylose consumption rate.)
2.3.2 Effect of degradation products on specific xylose consumption rate
Degradation products (Table 2; profile reported by Chundawat et al (Chundawat et al.,
TM

2010)) were represented using the AFEX

-CS water extract prepared as described above. The

TM

AFEX

-CS water extract should have a very similar degradation products profile compared to

TM

AFEX

-CS hydrolysate. The effect of water extracts at different concentrations was

investigated by performing xylose conversion experiments using non-growing cells from
hydrolysate fermentations. The non-growing cells for S. cerevisiae 424A and E.coli KO11 were
harvested at 24 h and 48 h, respectively. At this time, glucose was depleted and cell growths had
reached the stationary phase (Fig. 7a&c). During the xylose conversion experiment, no
significant cell growth was observed.
With no inhibitors in the solution, the specific xylose consumption rates for S. cerevisiae
-1

-1

424A and E. coli KO11 reached as high as 0.408 and 0.435 g·h ·g cell , respectively (Fig. 8).
-1

However, in real hydrolysate fermentations these values were reduced to 0.056 and 0.017 g·h ·g
-1

cell , respectively. About 86% and 96% reductions on specific xylose consumption rate were,

34

therefore, observed for S. cerevisiae 424A and E. coli KO11 fermentations in 6% (w/w) glucan
loading hdyrolysates, respectively.
Water extract showed no significant impact on the specific xylose consumption rate of S.
cerevisiae 424A until the concentration reached as high as 6% glucan loading equivalent, at
which point a 13% reduction in specific xylose consumption rate was observed (Fig. 8a). The
robustness of S. cerevisiae 424A was also highly rated by Lau et al. (2010). The specific xylose
consumption rate of E. coli KO11, however, was severely affected by water extract. Up to 76%
reduction was observed in water extract at the concentration of 6% (w/w) glucan loading
equivalent (Fig. 8b).
2.3.3 Effect of ethanol on specific xylose consumption rate
During fermentation of 6% (w/w) glucan loading hydrolysate, there is always 20-40 g/L
ethanol (mostly produced from glucose) present in fermentation broth during the xylose
consumption period (Fig. 7). Therefore, the effect of 0, 20, 30, and 40 g/L ethanol on specific
xylose consumption rate was investigated in 6% (w/w) glucan loading equivalent water extract
by applying xylose conversion experiments. With more than 20 g/L ethanol in water extract, the
specific xylose consumption rate of S. cerevisiae 424A was reduced from 0.355 to around 0.230
-1

-1

g·h ·g cell (Fig. 9a). This is a 31% reduction when compared to the control (without any
inhibitor). No significant difference of specific xylose consumption rate was found under these
conditions for different concentrations of ethanol in the range of 20-40 g/L.

35

Specific xylose consumption rate
(g/h/g cell)

a)

0.45
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00

13%
reduction
86%
reduction
Hyl. ferm.(24h)

0

Specific xylose consumption rate
(g/h/g cell)

b)

0.45
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00

1

2 3 4 5 6 7 8
Water extract conc.(%)
(glucan loading equivalent)

9

76%
reduction
96%
reduction

Hyl. ferm.(48h)
0

1

2 3 4 5 6 7 8
Water extract conc.(%)
(glucan loading equivalent)

9

TM

Figure 8 Effect of water extract of AFEX
pretreated corn stover on specific xylose
consumption rates of S. cerevisiae 424A (a) and E. coli KO11 (b).
Xylose conversion experiments were conducted in a 15 ml screw-capped vial and incubated at
o
o
150 rpm and 30 C (for S. cerevisiae 424A) or 37 C (for E. coli KO11) for 3 h. Non-growing
cells were harvested from hydrolysate fermentations at 24 h (for S. cerevisiae 424A) or 48 h (for
E. coli KO11). The solutions used for these experiments contained different concentrations of
water extract and buffer for pH control.
For S. cerevisiae 424A, the specific xylose consumption rate of the cells (from
hydrolysate fermentation) in xylose conversion solution (containing 6% glucan loading
equivalent water extract, 30 g/L ethanol and 30 g/L xylose) is higher than the value in a real

36

hydrolysate fermentation. Most of the components and their concentrations in the solution and in
the fermentation broth were similar. Two major factors could be responsible for this observation.
One is the additional degradation products, such as acetate that might be released during
enzymatic hydrolysis. But S. cerevisiae 424A had high degradation products tolerance (Fig.8a),
which renders this explanation less likely. The other factor was the inhibitory metabolites

Specific xylose consumption rate
(g/h/g cell)

a)
0.45
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00

With no inhibitors

13%
reduction

With 6% WE

31%
reduction
86%
reduction

42%
reduction

Hyl. ferm.(24h)
0

10

20

30

40

EtOH (g/L)

Specific xylose consumption rate
(g/h/g cell)

b)
0.45
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00

With no inhibitors
76%
reduction

96%
reduction

With 6% WE
24%
reduction
0

10

Hyl. ferm.(48h)

20

30

40

EtOH (g/L)
Figure 9 Effect of ethanol on specific xylose consumption rates of S. cerevisiae 424A (a) and E.
coli KO11 (b).
Results were obtained from xylose conversion experiments in 6% (w/w) glucan loading water
extract (WE). Non-growing cells harvested from hydrolysate fermentation were used.
37

produced by yeast fermentation while cells were in a dynamic process of growing and dying. In
order to explore the second possibility, we harvested the 24 h non-growing yeast cells cultured in
YEP medium and tested the specific xylose consumption rate of these cells in the solution
containing 35 g/L ethanol, 30 g/L xylose and 0.05 M phosphate buffer (pH 5.5). This solution is
similar to the YEP fermentation broth at 24h except that there were no fermentation metabolites
in this solution. The specific xylose consumption rate was around two times higher than the one
in the YEP fermentation, which indicates that the fermentation metabolites reduced the specific
xylose consumption rate in YEP medium by about half (Fig. 10a).
The xylose consumption of E. coli KO11 was totally inhibited by ethanol concentrations
greater than 20 g/L ethanol in 6% (w/w) glucan loading equivalent water extract (Fig. 9b).
However, during real hydrolysate fermentations, E. coli KO11 at 48 h (ethanol: 31.1 g/L) still
-1

-1

consumed xylose at a rate of 0.017 g· h ·g cell . Such differences might be caused by the
activity loss during cell sample preparation including centrifuge, washing, and cell re-suspension.
2.3.4 Effect of culture medium on specific xylose consumption rate
The xylose consumption ability of cells cultured in YEP medium and in hydrolysate was
compared by performing xylose conversion experiment in the solution containing 6% glucan
loading equivalent water extract, 30 g/L ethanol and 30 g/L xylose (Fig. 10). For both S.
cerevisiae 424A and E. coli KO11, cells from YEP medium showed higher specific xylose
consumption rates compared to the cells from hydrolysate. Particular nutrient(s) in YEP medium
probably helped induce the expression of sugar transporters or enzymes involved in xylose
metabolism, which in turn enhanced the xylose-fermenting ability or improved tolerance to
various inhibitors. We also observed that the yeast cells harvested at 48 h showed slightly higher
xylose consumption rates compared to the cells harvested at 24 h.

38

Specific xylose consumption rate
(g/h/g cell)

a)

Specific xylose consumption rate
(g/h/g cell)

b)

0.45
Cells fromYEP

0.40

Cells from hydrolysate

0.35
0.30
Metabolites
51% Reduction

0.25
0.20
0.15
0.10
0.05
0.00

24h cells 48h cells 24h cells Real ferm.(24h)
(solution A) (solution A) (solution B)

0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00

24h cells

48h cells Real ferm.(48h)

Figure 10 Effect of culture medium (YEP and hydrolysate) and culture time (24h and 48h) on
specific xylose consumption rates of S. cerevisiae 424A (a) and E. coli KO11 (b).
Results were obtained from xylose conversion experiments in solution A: 6% glucan loading
equivalent water extract, 30 g/L ethanol, 30 g/L xylose and 0.05M phosphate buffer (pH 5.5);
and solution B: 35 g/L ethanol, 30 g/L xylose and 0.05M phosphate buffer (pH 5.5) for S.
cerevisiae 424A and in solution A’ (6% glucan loading equivalent water extract, 30 g/L ethanol,
30 g/L xylose and 0.05M MOPS buffer (pH 7.0) for E. coli KO11.

39

2.3.5 Effects of degradation products on cell growth and nutrients addition on xylose
consumption
Since it is very difficult to separate degradation products from nutrients, and both are
TM

present in AFEX

-CS, the effect of degradation products on cell growth was investigated by

comparing maximum cell density in YEP medium (plus pure sugars) to the maximum cell
density in the hydrolysate with supplementation of YEP. Both glucose and xylose concentrations
were the same. Under such conditions, nutrients in both media were in excess and were assumed
not to affect performance. A 22% reduction of yeast cell biomass production and 11% reduction
of E. coli cell biomass production were observed in hydrolysate supplemented with YEP when
compared to the fermentation in YEP medium supplemented with sugars (Fig. 11). Such mild
TM

inhibitory effects of degradation products on cell growth was probably due to the AFEX

pretreatment forming much less acids compared to dilute acid pretreatment (Chundawat et al.,
TM

2010). The ammonolysis reaction occurring during AFEX

pretreatment converted a large

portion of compounds into amides instead of acids. For instance, the ratio of acetamide over
TM

acetic acid in AFEX
TM

AFEX

pretreated corn stover was 5.4 and the acetic acid concentration in

pretreated corn stover was 7.5 times lower compared to the diluted acid pretreated

sample (Chundawat et al., 2010).
Surprisingly, the 24 h specific xylose consumption rate of S. cerevisiae 424A in
hydrolysate with YEP was very close to that without YEP. YEP can act as an electron acceptor
like furfural to alleviate the imbalance of redox caused by xylose fermentation in anaerobic
conditions (Van Zyl et al., 1989; Wahlbom & Hahn-Hägerdal, 2002). Besides, as mentioned
above, enzyme systems involved in xylose metabolism might also be enhanced in the presence of
40

YEP. However, apparently, YEP did not help to increase the specific xylose consumption rate in
the hydrolysate, although it enhanced yeast cell density and overall xylose consumption when
compared to the hydrolysate without YEP (Fig. 11a). For E.coli, the xylose consumption rate in
hydrolysate with YEP was very close to that for hydrolysate without YEP although the maximal
cell density was higher (Fig. 11b). This was probably due to the extremely low specific xylose
consumption rate of E. coli in the presence of both degradation products and ethanol as shown
above.
a)

25

15
10
5
0

b)

25

OD.hyl+YEP

Xylose (g/L)

15

10
8
6

10

OD.hyl

5
0

Xyl.YEP
0

OD600

OD.YEP

20

OD600

Xylose (g/L)

20

0

20
18
16
OD.hyl+YEP
14
OD.hyl
12
10
8
Xyl. Hyl 6
4
Xyl. Hyl +YEP
2
Xyl.YEP
0
10
20
30
40
50
60
Time (h)
14
Xyl. Hyl +YEP
Xyl. Hyl
12
OD.YEP

4
2

0
10 20 30 40 50 60 70 80 90 100
Time (h)

Figure11 Cell growth and xylose consumption of S. cerevisiae 424A (a) and E. coli KO11(b)
TM
during fermentations in YEP medium, AFEX -CS hydrolysate and hydrolysate with YEP.
The initial glucose and xylose concentrations were approximately 55 and 25 g/L, respectively.
41

2.3.6 Understanding the xylose consumption problem during hydrolysate fermentation by S.
cerevisiae 424A
The overall extent of xylose consumption is determined by both cell density and the
specific xylose consumption rate. During hydrolysate fermentation by S. cerevisiae 424A, we
typically observed cell growth only during glucose fermentation. Cells were in a stationary or
dying phase during xylose fermentation (Fig. 7a). Shortage of ATP could be one major cause for
the inability of S. cerevisiae 424A to grow on xylose anaerobically (Kuyper et al., 2004). Due to
the inhibitory effect of degradation products on cell growth, 22% less cell biomass was produced
during glucose fermentation, which resulted in 22% fewer cells for xylose
consumption/fermentation (Fig.11a&Fig.12). After glucose fermentation, the composition of
broth changed with the release of ethanol and fermentation metabolites, which caused 31% and
42% reduction of specific xylose consumption rate. Degradation products were responsible for a
13% reduction in the specific xylose consumption rate. Therefore, compared to the ideal
conditions without any inhibitors present (maximal cell growth and maximal specific xylose

Dedegradation Normalized: 100% reduction on overall xylose consumption
products
Degradation products: 36%
Ethanol: 27%
Fermentation metabolites: 37%
Degradation
22% reduction
products
Ferm. metabolites
Cell growth
Ethanol
on Glc.
13%
42%
31%
reduction reduction reduction
6% improvement
of ethanol
metabolic yield

Specific xylose
consumption rate
(g/h/g cell)

TM

Figure 12 Understanding the xylose fermentation problem during AFEX
fermentation by S. cerevisiae 424A.

42

-CS hydrolysate

-1 -1

consumption rate), the overall xylose consumption rate (g·L ·h ) was reduced by 89% during
hydrolysate fermentation, among which degradation products, ethanol and fermentation
metabolites accounted for 32%, 24%, and 33%, respectively.
Under ideal conditions without any inhibition, S. cerevisiae 424A is able to grow to a
dry cell weight of 6.5 g/L in hydrolysate and reaches a maximum specific xylose consumption
-1

-1

TM

rate of 0.408 g· h ·g cell . Under such conditions, xylose in AFEX

-CS hydrolysate

(typically 28-30 g/L) can be completely consumed in 10.6-11.3 hours with a xylose consumption
-1

-1

rate of 2.65 g· h ·L . In real hydrolysate fermentations, however, the dry cell weight and 24 h
-1

-1

specific xylose consumption rate decreased to 5.1 g/L and 0.056 g· h ·g cell , respectively.
Based on those values, the 28 g/L xylose in hydrolysate can be theoretically consumed in 98 h.
Nevertheless, the xylose consumption rate was continually decreasing during fermentation due to
the decreasing viable cell density and xylose concentration (Fig. 7a). Therefore, it took 150 h
(excluding the glucose fermentation time of 18 h) to reduce xylose concentration to around 5 g/L.
It was noteworthy that there was a short but higher specific xylose consumption rate period
during late glucose fermentation at around 10-18 h. This was due to the fact that a low
concentration of glucose could help induce sugar transporters that have higher affinity for xylose
-1

(Bertilsson et al., 2008). The specific xylose consumption rate reached about 0.15 g· h ·g cell

-1

at that time.
With YEP supplementation, cell density was enhanced to 6.7 g/L in hydrolysate (Fig. 11a)
and 28 g/L xylose was reduced to around 5 g/L in 30 h (excluding glucose fermentation time 18
h; Fig. 11a). Nutrients affected cell growth and hence affected xylose fermentation. Different
feedstocks have different levels of nutrients and hence exhibit different xylose fermentation
43

TM

performance. For instance, in 6% (w/w) glucan loading of AFEX

treated sugar cane leaf

hydrolysate, the S. cerevisiae 424A cell density reached 6.9 g/L (dry weight basis) and xylose
TM

consumption reached 87.0% in a 72 h fermentation. However, cell density in AFEX

sugar

cane bagasse hydrolysate was 4.4 g/L and xylose consumption of 62.7% was achieved in 120 h
(Krishnan et al., 2010). Fermentation of mixed biomass hydrolysates containing more nutrients
could be a potential way to enhance xylose fermentation.
2.4 Conclusions
TM

S. cerevisiae 424A and E. coli KO11 both have the potential to finish AFEX

-CS

hydrolysate xylose fermentation in 24h. However, due to its poor tolerance to degradation
products and ethanol, E. coli KO11 lost most of its ability to consume xylose during hydrolysate
fermentation. The xylose consumption rate of S. cerevisiae 424A in hydrolysate was reduced to a
great extent (89%) due to the presence of not only degradation products but also ethanol and
fermentation metabolites. Degradation products reduced yeast cell biomass production by 22%
and the specific xylose consumption rate by 13%. Ethanol and fermentation metabolites reduced
the specific xylose consumption rate by 31% and 42%, respectively.

44

CHAPTER III SIMULTANEOUS SACCHARIFICATION AND COTM

FERMENTATION OF AFEX

PRETREATED SWITCHGRASS

Abstract
It is well known that simultaneous saccharification and co-fermentation (SSCF) reduces
cellulosic ethanol production cost compared to separate hydrolysis and fermentation (SHF).
TM

However, the traditional SSCF process of converting Ammonia Fiber Expansion (AFEX

)

pretreated switchgrass to ethanol using commercial enzymes and Saccharomyces cerevisiae
424A(LNH-ST) gave reduced ethanol yield due to lower xylose consumption. To overcome this
problem a two-step SSCF process was developed, in which xylan was hydrolyzed and fermented
first followed by the hydrolysis and fermentation of glucan. Important parameters such as
temperature, cellulases loading during xylan hydrolysis and fermentation, initial OD600 for
inoculation of S. cerevisiae 424A(LNH-ST), and pH were studied for best performance.
Compared with traditional SSCF, the two-step SSCF showed higher xylose consumption and
higher ethanol yield. The sugar conversion was also enhanced from 70% by enzymatic
hydrolysis to 82% by two-step SSCF. One important finding is that the residue from enzymatic
hydrolysis of switchgrass plays a significant role in reducing xylose consumption and ethanol
metabolic yield during SSCF.

45

3.1 Introduction
Switchgrass (Panicum virgatum), a C4 perennial grass, has been identified as one of the
most promising biomass feedstocks to produce cellulosic ethanol (Walsh et al., 2003). It has
been estimated that 171 million tons of switchgrass (SG) can be economically produced in the
US (Walsh et al., 2003) and that 94% life cycle greenhouse gas emission reductions are possible
using ethanol produced from SG compared to gasoline (Schmer et al., 2008).
Simultaneous saccharification and co-fermentation (SSCF) carries out enzymatic
hydrolysis, xylose and glucose fermentation in a single bioreactor. It has several advantages
such as: lower cost, shorter process time, reduced contamination risk and less inhibitory effects
during enzymatic hydrolysis compared to separate hydrolysis and fermentation (SHF)
TM

(Taherzadeh & Karimi, 2007). By applying both AFEX

and SSCF on corn stover, the

minimum ethanol selling price (MESP) was estimated to be as low as $1.03/gallon based on the
future technology (Sendich et al., 2008). In the model it was assumed that both monomeric
TM

glucose and xylose from AFEX

treated biomass were completely fermented to ethanol.

However, relatively low xylose consumption (< 40%) was found when SSCF was performed on
steam-pretreated biomass at 9% water insoluble solids (WIS) by using xylose-fermenting strain S.
cerevisiae TMB3400 (Olofsson et al., 2008b). By applying pre-fermentation of xylose, the
xylose utilization was enhanced to 77% in fed-batch SSF at 10% WIS loading (Bertilsson et al.,
2009).
Previous results in Chapter II have shown that S. cerevisiae 424A (LNH-ST) consumed
xylose slowly compared to glucose, especially in hydrolysates. Degradation products, ethanol
and fermentation metabolites are the major inhibitors for xylose consumption. To alleviate
ethanol and fermentation metabolite inhibition, performing xylose fermentation ahead of glucose
46

TM

fermentation might help. In the present study, a two-step SSCF process on AFEX

-treated

switchgrass (using commercial enzymes and S. cerevisiae 424A) was developed. Two-step SSCF
gives higher ethanol yield with improved xylose consumption compared to the traditional SSCF
TM

process on AFEX

TM

-SG. Process improvements include hydrolyzing AFEX

-SG with

hemicellulases (xylanase and pectinase) first to release xylose which was then fermented by S.
cerevisiae 424A, followed by adding cellulases to hydrolyze glucan to glucose and continue the
fermentation. Several process parameters for the two-step SSCF were optimized and results were
compared with both traditional SSCF and SHF process.
3.2 Materials and Methods
3.2.1 Switchgrass
The species SG used in this study was Cave-in-Rock grown at Michigan State University
(East Lansing, MI, USA) and harvested in mid October (Bals et al., 2010). The glucan, xylan and
lignin contents were estimated using the standard method published by National Renewable
Energy Laboratory (NREL) (Sluiter et al., 2004) and have been reported in our previous
publication (Bals et al., 2010) as 33.6%, 25.3%, and 16.7%, respectively.
TM

3.2.2 AFEX

TM

AFEX

Pretreatment

pretreatment conditions used for this study include: ammonia to biomass loading
o

2.0 g/g dry biomass, water loading 0.5 g/g dry biomass, temperature 140 C and residence time
30 minutes.
3.2.3. Enzymatic Hydrolysis (EH)
A commercial enzyme mixture was used for enzymatic hydrolysis. The mixture was
composed of Spezyme CP (Genencor Inc, USA; protein 88 mg/mL) 89.3 ml/kg SG, Novozyme
47

188 (Sigma-Aldrich, USA; protein 150 mg/ml) 89.9 ml/kg SG, Multifect xylanase (protein 35
mg/mL) 31.5 ml/kg SG and Multifect pectinase (Genencor Inc, USA; protein 90 mg/mL) 22.2
ml/kg SG. The protein concentrations were estimated by Dairyone (Ithaka, NY) using Kjeldahl’s
method. These enzymes activities have been reported by Dien et al. (Dien et al., 2008). The
glucan loading (Zhong et al., 2009) used in this study was 6% by weight. The reaction was
carried out in a 2.0 L baffled flask with 500 g total mixture at pH 4.8 (50 mM phosphate buffer),
o

50 C, and 250 rpm for 96 h. A final concentration of 50 mg/L chloramphenicol (Cm) was used
to avoid microbial contamination. Biomass and enzymes were fed in two batches at a 3 h time
interval. Hydrolyzate was harvested by centrifuging at 14,000 rpm for 30 min and sterile filtering
the supernatant through a 0.22 μm filter.
3.2.4 Microorganism and seed culture preparation
Xylose-fermenting S. cerevisiae 424A(LNH-ST) (Ho et al., 1999) obtained from Prof.
Nancy W. Y. Ho, Purdue University was used for this study. Seed culture was prepared on YEP
medium (5 g/L yeast extract, 10 g/L peptone and 50 g/L glucose) by inoculating a frozen
glycerol stock into a 250 ml flask with a working volume of 100 ml. The initial OD of the seed
o

culture was approximately at 0.1. The strain was grown at 30 C and 150 rpm under
microaerophilic conditions (Zhong et al., 2009) for 20 h. The final optical density at 600 nm
(OD600) and glucose concentration of the seed culture were around 10 and below 1 g/L,
respectively.
3.2.5 Fermentation of hydrolyzate
Fermentations of hydrolyzate were carried out in 250 ml Erlenmeyer flasks with a working
o

volume of 100 ml at pH 5.5, 30 C and 150 rpm. The flasks were capped with rubber stoppers to

48

maintain microaerophilic conditions. Yeast cell pellets used for inoculation were obtained by
centrifuging the seed culture at 4000 rpm for 8 min. The initial OD for fermentation of
hydrolyzate was 2.0. Glucose, xylose and ethanol concentrations were analyzed by HPLC with a
Biorad Aminex HPX-87H column described previously in Chapter II section 2.2.7.
Ethanol metabolic yield was calculated based on the theoretical ethanol yield from available
glucose and xylose, which is 0.51 g ethanol/g sugar.
TM

3.2.6 SSCF of AFEX

-SG

SSCF experiments were conducted in 250 ml baffled flasks with 100 g total mixture at 180
rpm in a shaking incubator (Innova, New Brunswick, NJ). Enzyme loading and glucan loading
were the same as for enzymatic hydrolysis. For traditional SSCF (Fig.13A), all the enzymes were
o

added at the beginning during pre-hydrolysis at pH 4.8, 50 C and 250 rpm for 8 h. Then pH,
o

temperature, and agitation rate were adjusted to 5.2, 30 C, and 180 rpm, respectively. Yeast
cells were inoculated at an initial OD of 4 to perform SSCF for another 144 h. For two-step
SSCF (Fig.13B), xylanase and pectinase were added at first to perform pre-hydrolysis for 8 h.
Then yeast cells and a small percentage of Spezyme CP and Novozyme 188 were added to the
hydrolyzate to carry out SSCF (step 1) for 60 h. During step 1, most of the xylan was hydrolyzed
to xylose and then fermented by yeast cells. Little glucan was hydrolyzed and fermented during
this step. After 60 h, the rest of Spezyme CP and Novozyme 188 enzymes were fed to release
glucose from glucan, which was subsequently fermented to ethanol by the yeast (SSCF step 2).
Unless otherwise stated, fermentations were performed with no washing of pretreated
biomass, detoxification or nutrient supplementation.

49

TM

3.2.7 Preparation of washed AFEX

-treated SG

TM

AFEX

TM

dry AFEX

-treated SG wash experiments were conducted using 50 ml distilled water per g

treated SG. The pretreated biomass was washed 5 times to make sure all the

soluble degradation products were removed. The washed biomass was used the same day.
3.2.8 Mass balance
Mass balance calculation was based on the analyses of ethanol, monomeric, oligomeric, and
polymeric sugars before and after each process (enzymatic hydrolysis and fermentation for SHF;
fermentation for SSCF) as reported earlier (Lau & Dale, 2009).

Figure 13 Schematic diagram of traditional SSCF (A) and two-step SSCF (B).

50

Table 6 Summary of parameters for the experiments examining the effects of the fraction of total
cellulases loaded and temperature during step 1, pH and initial OD on two-step SSCF
performance

Parameter

Effect of
percentage of total
cellulases loading

Effect of
temperature

Effect of initial
OD600

Effect of pH

Percentage of total
cellulases loading

0%,2%,4%,6%,8%

4%

4%

4%

30
5.2
4
Fig. 15A1, A2

30, 33, 35
5.2
4
Fig. 15B1,BA2

35
5.2
1, 4, 10, 16
Fig. 15C1, C2

35
4.8, 5.2, 5.5
4
Fig. 15D1, D2

Temperature (°C)
pH
Initial OD600
Relevant figure

Concentration (g/L)

A. Traditional SSCF

30

Xylose

25
20
Ethanol

15
10

Glucose

5
0

0

Concentration (g/L)

35

35

B. Two-step SSCF (un-optimized)
Step 2
Step 1
Xylose
25
30
20
15

Ethanol

Glucose

10
5
0

20 40 60 80 100 120 140 160
Time (h)

0

20 40 60 80 100 120 140 160
Time (h)

Figure 14 Traditional SSCF (A) vs. two-step SSCF (B).
o
Both experiments were carried out at 6% glucan loading, initial OD 4, pH 5.2, 30 C, and 180
rpm. For this experiment of two-step SSCF, no cellulases were added during step 1.

3.3 Results and discussion
3.3.1 Traditional SSCF vs. two-step SSCF (un-optimized)
Traditional and two-step SSCF experiments were conducted at identical conditions at 6.0%
TM

(w/w) glucan loading (equivalent to 17.9% (w/w) total solids loading) of AFEX
51

-treated SG.

In this two-step SSCF experiment, no cellulases were added during step 1. Results are compared
side-by-side in Fig. 14. For the traditional SSCF process, after 8 h pre-hydrolysis using
hemicellulase and cellulase, glucose and xylose concentrations reached 30.7 and 25.5 g/L,
respectively. After inoculation of the yeast, glucose concentration was quickly reduced to 1.3 g/L
in 14 h and then maintained at a low level (< 4 g/L) during the whole fermentation time.
However, xylose concentration remained at nearly the same level (about 25 g/L) during the
entire process. A relatively low final ethanol concentration (22 g/L) was achieved after 144 h
fermentation. For the two-step SSCF process (un-optimized), glucose and xylose concentrations
were 8.1 and 25.1 g/L, respectively, after 8 h pre-hydrolysis. During step 1 of SSCF (0-60 h),
glucose and xylose concentrations were reduced to 0.6 and 18.7 g/L, respectively with 10.6 g/L
ethanol produced. In the step 2 (after cellulase addition after 60 h), xylose concentration
remained the same at about 19 g/L and glucose concentration increased to 15 g/L in 36 h.
However, glucose concentration was reduced to 0.9 g/L at 128 h and ethanol concentration
reached the maximum of 26.3 g/L. These results clearly demonstrate that the two-step SSCF
process results in higher ethanol yield by enhanced xylose consumption and is superior to the
traditional SSCF process.
3.3.2 Optimization of two-step SSCF conditions
Conditions optimized during the two-step SSCF process include: the fraction of the total
cellulases loaded and temperature during step 1 of SSCF, initial OD, and pH (Table 6). Different
cellulases fractions loaded during step 1 resulted in different xylose consumption rates (Fig.
15A1) (assuming that xylan hydrolysis was not affected by feeding cellulases at low levels).
Among the tested values, 4% of total cellulases led to the fastest xylose consumption, the lowest
final xylose concentration (18.5 g/L), and highest ethanol yield (27.3 g/L, Fig. 15A2). Regarding

52

o

o

temperature optimization, xylose was consumed much faster at both 33 and 35 C than at 30 C
(Fig. 15B1) even without considering the higher xylose release rate at higher temperature
(Taherzadeh & Karimi, 2007). The final ethanol and xylose concentrations were higher and
o

lower, respectively when step 1 of SSCF was conducted at either 33 or 35 C (Fig. 15B2).
Higher initial OD resulted in higher xylose consumption rates, lower final xylose concentration
(6.27 g/L at OD 16), and higher ethanol yield (36.1 g/L at OD 16, Fig. 15C1, C2). Judging from
final xylose and ethanol concentrations, pH 5.5 was the optimum for two-step SSCF among the
tested values (Fig. 15 D2). Glucose concentrations (data are not shown) in all tested conditions
were below 2 g/L after 192 h. Considering the above analysis and the cost of cellulosic ethanol
production, the following conditions were chosen as the optimal ones to conduct two-step SSCF:
o

4% of the total cellulases loading and 35 C during step 1, initial OD 4, and pH 5.5. Under the
optimum conditions, 32.1 g/L final ethanol concentration was achieved with a final xylose
concentration of 11.2 g/L (Fig. 16).

53

22
0%
2%
4%
6%
8%

18
16

30 40
Time (h)

50

20
15
10
5
0

20

5

15
0% 2% 4% 6%8% 0% 2%4% 6% 8%
Ethanol 6%-Percentage of total cellulases

0

10

20

50

OD=1
OD=10

OD=4
OD=16

15
10
5
10

20

30 40
Time (h)

50

30
25

10

20

5

15
30

33 35 30 33
Temperature ( oC)

35

Xylose30 Xylose33 Xylose35 Ethanol30 Ethanol33 Ethanol35

30

60

10
--

40

C2

Xylose
Ethanol

25

35

20

30

15

25

10

20

5

15

0

0

35

15

60

20

Xylose
Ethanol

20

1

4

10 16 1 4
Initial OD600

10 16

10

Xylose1 Xylose4 Xylose10Xylose16Ethanol1 Ethanol4Ethanol10 thanol 16
E

Figure 15 Effects of various factors on xylose fermentation and final ethanol concentration
during two-step SSCF.
Fraction of total amount of cellulases fed during step 1 (A1), Initial OD (B1), Temperature
during step 1 (C1), pH (D1) and the corresponding SSCF results after 192 h (A2, B2, C2, and
D2).

54

Ethanol concentration (g/L)

30 40
Time (h)

10
40

B2

0

25 C1

0

25
10

25

30 C
33 oC
35 oC

35
30

15

60
o

25 B1

Xylose
Ethanol

20

0

20

40

A2

Ethanol concentration (g/L)

Xylose concentration (g/L)

10

Xylose concentration (g/L)

20

0

Xylose concentration (g/L)

Xylose concentration (g/L)

24

25

Xylose concentration (g/L)

Xylose concentration (g/L)

A1

Ethanol concentration (g/L)

26

Xylose concentration (g/L)

Xylose concentration (g/L)

20

25

pH4.8
pH5.2
pH5.5

15
10
5
0

0

10

20

30 40
Time (h)

50

40

D2

Xylose
Ethanol

20

35
30

15

25
10
5
0

60

20
15
4.8

5.2

5.5 4.8 5.2
pH

5.5

Xylose4.8 Xylose5.2 Xylose5.5 Ethanol4.8 Ethanol5.2 Ethanol5.5

10

Ethanol concentration (g/L)

Figure 15 (cont’d)
25 D1

--

Poor xylose fermentation is a key bottleneck in traditional SSCF processes for cellulosic
ethanol production. This is probably due to the large amount of glucose released during the
process, which inhibited xylose uptake by hexose transporters (Bertilsson et al., 2008).
In addition, xylose consumption was also inhibited by ethanol and fermentation metabolites
released during glucose fermentation (Chapter II). By performing the two-step SSCF under
optimal conditions, xylose fermentation was greatly improved. This process avoided large
amount of ethanol and fermentation metabolites inhibition on xylose fermentation. Glucose
concentration was kept at a low but not zero level during step 1 of SSCF by supplementing with
a low concentration of cellulases, which could not only help induce the sugar transporters
expression used for xylose uptake (Bertilsson et al., 2008), but also help overcome the inhibition
from degradation products (Bellissimi et al., 2009). At the beginning of fermentation, there was a
low oxygen concentration in the broth. This was also important to xylose fermentation since
oxygen could help to alleviate the redox imbalance caused by xylose metabolism (Ruohonen et
al., 2006). Presence of nutrients at the beginning of the two-step SSCF process could be another
beneficial factor which probably stimulated the yeast growth and hence stimulated the xylose
consumption.

55

a)

Concentration (g/L)

45

EtOH. SSCF
EtOH. SSCF with tween
Xyl. SSCF
Xyl. SSCF with tween

40
35
30
25
20
15
10
5
0

Glucose concentration (g/L)

b)

0

40

80
120
Time (h)

160

200

20
SSCF
SSCF with tween

18
16
14
12
10
8
6
4
2
0

0

40

80
120
Time (h)

160

200

Figure 16 Effect of adding 2% (w/w) Tween 80 on xylose consumption, ethanol production (a),
and glucose fermentation (b) during two-step SSCF.
o
The experiments were conducted at optimized conditions: initial OD 4, pH 5.5, 35 C and 4% of
total cellulases loading during step 1 of SSCF.

In all tested SSCF experiments, little xylose was consumed after 48 h even with high
xylose concentration and low glucose concentration in the broth. This was also found by the
SSCF studies with TMB3400 (Bertilsson et al., 2009; Olofsson et al., 2008b) and was mostly

56

attributed to the decreased cell viability (Rudolf et al., 2005) and the inactivation of transport
systems.
TM

3.3.3 Effects of EH residue and washed AFEX

-SG on hydrolyzate fermentation

To investigate the effects of solids content during SSCF on xylose fermentation and ethanol
TM

metabolic yield, EH residue and washed AFEX

-SG biomass were added to the hydrolyzate

derived from 6% glucan loading enzymatic hydrolysis (Fig. 17). Both glucose and xylose were
fermented well in the absence of solid residues (Fig. 17A). Glucose (45.5 g/L) was consumed
completely in 18 h and xylose was reduced from 31.8 g/L to 1.9 g/L in 96 h. The final ethanol
concentration reached 34.6 g/L with an ethanol metabolic yield of 89.7%. With the solid EH
residue (75 g/L, equivalent to the same concentration in final SSCF broth), however, only 11.8
g/L xylose was consumed during 144 h fermentation (Fig. 17B). The ethanol metabolic yield was
reduced to 82% (Fig. 17D). The xylose consumption rate and ethanol metabolic yield were also
TM

-SG) or 114 g/L

lowered by adding 29 g/L (equivalent to 1% glucan loading AFEX
TM

(equivalent to 4% glucan loading AFEX

TM

-SG) washed AFEX

-SG (Fig. 17C, D). These

results clearly suggest that the presence of solid biomass (especially the EH residue) during
SSCF affected both xylose consumption and ethanol metabolic yield. The mechanisms by which
EH residue affects the xylose metabolism are unclear at present.
In an attempt to alleviate these effects, 2% (w/w) of Tween 80 was added during the
two-step SSCF process (Fig. 16). With Tween 80 addition, glucose concentration was enhanced
to 10.2 g/L after pre-hydrolysis. Adding surfactants is known to improve sugar yield during
enzymatic hydrolysis (Alkasrawi et al., 2003; Kim & Chun, 2004; Kim et al., 2006; Tu et al.,

57

2009). This improvement is attributed to reduction of unproductive binding to the lignin
(Eriksson et al., 2002a). The final sugar conversion, however, was slightly decreased when

Concentration (g/L)

40
30
20
10
0

0

40

Xylose Concentration (g/L)

50 B

Glucose
Xylose
Ethanol

20

40
60
Time (h)

80

C

30
20
10

AFEX-SG 0 g/L
AFEX-SG 29g/L
AFEX-SG 114g/L
EH residue

0
-10

0

20

40

60

40
30
20
10
0

100

80

0

92

Ethanol Metabolic Yield (%)

Concentration (g/L)

50 A

20

40

60 80 100 120 140
Time (h)

D

90
88
86
84
82
80

100

AFEX-SGwith 1% glucan AFEX4% glucan AFEXwith EH residue -AFEX-SG AFEX-SG SG residue
SG-Hyl
SG-Hyl
SG-Hyl with SG
SG-Hyl EH
0g/L
29 g/L 114 g/L

Time (h)
TM

Figure 17 Effects of EH residue and washed AFEX

-SG on fermentation performance in 6%

TM

glucan loading AFEX -SG hydrolyzate.
A, fermentation of hydrolyzate without solids; B, fermentation of hydrolyzate with EH residue; C,
TM
effects of washed AFEX -SG and EH residue on xylose fermentation in hydrolyzate; D, effects
TM

of washed AFEX

-SG and EH residue on ethanol metabolic yield. Fermentations carried out at
o

initial OD 2, pH 5.5, 30 C and 150 rpm.

compared with the two-step SSCF without Tween 80 (Fig. 18B1, B2). This was probably
because of the higher ethanol concentration present in the fermentation broth due to the addition
of Tween 80. Ethanol is an important factor limiting enzymatic hydrolysis during SSCF

58

(Taherzadeh & Karimi, 2007). The maximum glucose accumulation after feeding cellulases at 60
h was reduced from 14.3 to 5.1 g/L (Fig. 16b), which indicated that cellular activity was
improved. The xylose consumption rate and ethanol metabolic yield were also enhanced with the
addition of Tween 80 (Fig. 16a, Fig. 18B2). Similar behavior was also found during SSF by the
addition of Tween 20 (Alkasrawi et al., 2003). Since Tween 80 did not show any influence
during solid-free fermentation (Tu et al., 2009), it probably reduced the binding of yeast cells to
the solid biomass during SSCF, which inhibited the cell viability.
3.3.4 Mass balance comparison between SHF and two-step SSCF
To demonstrate the current status and bottlenecks of the technology, mass balances were
TM

done for both SHF and two-step SSCF of AFEX

-SG (with and without 2% Tween 80
TM

addition) (Fig. 18, Table 7). The SHF process yielded 178.4 g ethanol from 1 kg AFEX

-SG

with glucan and xylan conversions of 66.5% and 74.7%, respectively (Fig. 18A). Ethanol
metabolic yield was 89.7%. SHF of Cave-in-Rock switchgrass (cut in October) showed low
sugar conversion during enzymatic hydrolysis, which was also found in an earlier report (Bals et
al., 2010). However, ethanol metabolic yield was high during fermentation of the solid-free
hydrolyzate. Enzymatic hydrolysis has been identified as the bottleneck for SHF (Zhong et al.,
2009). Two-step SSCF with initial OD 4 enhanced glucan and xylan conversions to 80.3% and
84.3%, respectively (Fig. 18B1). These enhancements were probably due to the removal of
enzymatic hydrolysis products (sugars) by the yeast as soon as they were produced, thus
preventing sugar inhibition (Xiao et al., 2004). The glucan conversion enhancement was also due
to the hydrolysis and fermentation of xylan before cellulases were added, which reduced the
inhibition of xylan and xylose oligomers to cellulases (Kumar & Wyman, 2009b). However, the
ethanol metabolic yield was reduced to 72.7% and the final ethanol yield was 165.3 g per kg
59

TM

AFEX

-SG. About 55.0 g monomeric xylose remained unutilized in the fermentation broth.
TM

Based on the total sugars in the AFEX

-SG, 76.8% of glucose and 51.0% of xylose were

consumed by yeast; 3.5% of glucose and 33.2% of xylose remained in the fermentation broth;
19.7% of glucose and 15.7% of xylose were unhydrolyzed and left in the residue. By adding 2%
of Tween 80, ethanol metabolic yield was improved to 77.7% (Fig. 18B2) and the final ethanol
TM

yield was enhanced to 183.5 g per kg AFEX

-SG. However, 40.7 g monomeric xylose was

still un-utilized.
Xylose consumption and the final ethanol yield (but not the ethanol metabolic yield) of twostep SSCF were improved by increasing the initial OD (Table 7). However, this may not be a
cost-efficient approach unless the yeast cells can be reused. Though Tween 80 addition improved
both xylose consumption and ethanol metabolic yield, this is probably too expensive for
cellulosic ethanol production. A cheap surfactant with the same effect of Tween 80 would be an
TM

important advance. About 191 g ethanol per kg of washed AFEX

-SG (all the degradation

products removed) was achieved by SSCF when supplementing YEP (Table 7). Overall, 91%
xylose consumption and 77.2% ethanol metabolic yield were realized. However, sugar
conversion was reduced to 77% probably because of the high ethanol concentration which is
known to inhibit enzymes (Taherzadeh & Karimi, 2007). Washing and nutrient supplementation
TM

will also increase the production cost of cellulosic ethanol from AFEX

treated biomass.

3.3.5 The limiting factors during SSCF
Reduced xylose consumption and ethanol metabolic yield are two key issues for SSCF,
TM

which affect the overall ethanol yield and titer. The presence of solid AFEX

60

treated biomass,

especially the EH residue (mostly insoluble lignin) was partly responsible for reduced
performance versus solid-free fermentation. Another factor which may affect xylose
consumption and ethanol metabolic yield during SSCF is the interaction between solid biomass
and degradation products produced during AFEX. Degradation products caused lower yeast cell
production and thereby increased ethanol metabolic yield but decreased xylose consumption
during hydrolyzate fermentation (Chapter II). Based on this assumption, cell production should
TM

be increased when SSCF is conducted on washed AFEX

-SG (degradation products removed)

with supplementation of YEP, which in turn increase xylose consumption and decrease ethanol
metabolic yield. Though the results showed enhanced xylose consumption, the ethanol metabolic
yield was also increased (Table 7). High cell density (Table 7; SSCF with initial OD 10) only
benefited xylose consumption with the presence of degradation products. Therefore, it seems that
the removal of degradation products helped increase the ethanol metabolic yield. This implies the
interaction between solid biomass and degradation products also played a role in the reduction of
ethanol metabolic yield since degradation products alone do not reduce ethanol metabolic yield.
3.4 Conclusions
The two-step strategy was an efficient way to improve xylose consumption and ethanol
yield during SSCF. The optimized conditions for two-step SSCF were found to be: 4% of total
o

amount of cellulases loading and 35 C during step 1, initial OD 4, and pH 5.5. Compared to the
traditional SSCF (at 6% glucan loading), the two-step SSCF reduced the final xylose
concentration from 26.4 g/L to 11.2 g/L and increased the final ethanol concentration from 22.0
g/L to 32.1g/L. 80.3% of glucan conversion and 84.3% of xylan conversion were achieved with
ethanol metabolic yield 72.7%. Enzymatic hydrolysis residue played a significant role in the low
xylose consumption and ethanol metabolic yield.
61

A) SHF
- Spezyme CP: 89.3 ml (7.9 g Protein)
- Novozyme 188: 89.9 ml (13.5 g Protein)
- Multifect Xylanase: 31.5 ml (1.1 g Protein)
- Multifect Pectinase: 22.2 ml (2.0 g Protein)

AFEX treated
Switchgrass: 1 Kg
(Dry Weight)
- Glucose: 388.9 g
- Xylose: 287.5 g

-Yeast innoculum
- Yeast inoculum
(Dry Weight):
5.0 g

ENZYMATIC
HYDROLYSIS
Time: 96 h
Temp.: 50°C

FERMENTATION - Ethanol: 178.4 g (34.6 g/L)
- Glucose: 235.0 g (45.5 g/L)
- Xylose: 165.0 g (31.8 g/L)
- Gluco-oligomers: 23.6 g (4.7 g/L)
- Xylo-oligomers: 50.1 g (9.8 g/L)

Time: 96 h
Temp.: 30°C

- Glucose: 130.4 g
- Xylose: 72.6 g

- Xylose: 9.9g (1.9 g/L)
- Gluco-oligomers: 23.6g (4.7 g/L)
- Xylo-Oligomers: 50.1g (9.8 g/L)
Conversion:
Glucan: 66.5%
Xylan: 74.7%
Ethanol Metabolic yield: 89.7%

Figure 18 Mass balance analysis for SHF (A), two-step SSCF (B1), and two-step SSCF with 2% Tween 80 (B2).
o
For SHF, enzymatic hydrolysis was carried out at 6% glucan loading, pH 4.8, 50 C, 250 rpm for 96 h. Fermentation of
o

hydrolyzate was conducted at initial OD 2, pH 5.5, 30 C, 150 rpm for 96 h. For two-step SSCF, experiments were performed at 35
o

C and 4% of total cellulases loading during step 1 of SSCF, initial OD 4, pH 5.5, and 180 rpm.

62

Figure 18 (cont’d)

B1) Two-step SSCF
- Spezyme CP: 89.3 ml (7.9 g Protein)
- Novozyme 188: 89.9 ml (13.5 g Protein)
- Multifect Xylanase: 31.5 ml (1.1 g Protein)
- Multifect Pectinase: 22.2 ml (2.0 g Protein)
- Yeast inoculum
(Dry Weight):
11.2 g
AFEX treated
Switchgrass: 1Kg
(Dry Weight)

SSCF
192 h

Ethanol: 165.3g
(32.1g/L)
- Glucose: 2.4 g (0.5 g/L)
- Xylose: 55.0 g (11.2 g/L)
- Glc-oligomers: 11.2 g (2.2 g/L)
- Xyl-Oligomers: 40.5 g (8.2 g/L)

- Glucose: 388.9 g
- Xylose: 287.5 g

Conversion:
Glucan: 80.3%
-Glucose: 76.5 g Xylan: 84.3%
-Xylose: 45.2 g
Ethanol Metabolic yield: 72.7%

B2) Two-step SSCF+2% Tween 80
- Spezyme CP: 89.3 ml (7.9 g Protein)
- Novozyme 188: 89.9 ml (13.5 g Protein)
- Multifect Xylanase: 31.5 ml (1.1 g Protein)
- Multifect Pectinase: 22.2 ml (2.0 g Protein)
- Yeast inoculum
(Dry Weight):
11.2 g
AFEX treated
Switchgrass: 1 Kg
(Dry Weight)

SSCF
192 h

Ethanol: 183.5 g
(35.1 g/L)
- Xylose: 40.7g (8.3 g/L)
- Glc-oligomers: 5.2g (0.8 g/L)
- Xyl-Oligomers: 43.4g (8.8 g/L)

- Glucose: 388.9 g
- Xylose: 287.5 g

-Glucose: 76.1 g
-Xylose: 47.9 g
63

Conversion:
Glucan: 80.4%
Xylan: 83.4%
Ethanol Metabolic yield: 77.7%

TM

Table 7 Summary of SHF and two-step SSCF experiments on 6% glucan loading AFEX
Sugar
b
conversion
(%)
70

SHF
SSCF-OD 4
SSCF-OD10
e

SSCF-tween
SSCF-washed
f
SG

Xylose
Final ethanol
c
consumption concentration
(g/L)
(%)
94
34.6

a

-SG

Ethanol
d
metabolic yield
(%)
89.7

Ethanol yield
(g/kg AFEX
SG)
178

82
85
82

73
84
79

32.1
35.0
35.1

72.7
71.7
77.7

165
180
184

77

91

36.4

77.2

191
o

a, For SHF, enzymatic hydrolysis was carried out at 6% glucan loading, pH 4.8, 50 C, 250 rpm
o

for 96 h. Fermentation of hydrolyzate was conducted at initial OD 2, pH 5.5, 30 C, 150 rpm for
o

96 h. For two-step SSCF, experiments were performed at 35 C and 4% of total cellulases
loading during step 1 of SSCF, pH 5.5, and 180 rpm.
b, Related to both oligmeric and monomeric sugars (glucose and xylose)
c, Based on available monomeric xylose
d, Based on the theoretical yield
e, two-step SSCF with initial OD 4 and 2% (w/w) Tween 80
f, two-step SSCF of washed AFEX-SG with initial OD 4 and supplementation of YEP.

64

CHAPTER IV SIMULTANEOUS SACCHARIFICATION AND COTM

FERMENTATION OF AFEX

PRETREATED CORN STOVER

Abstract
Xylose consumption by Saccharomyces cerevisiae 424A(LNH-ST) during simultaneous
TM

saccharification and co-fermentation (SSCF) of AFEX

pretreated switchgrass was inhibited

by unhydrolyzed solids. Such inhibitory effects were not found in unhydrolyzed solids from
TM

AFEX

TM

pretreated corn stover (AFEX

-CS). However, the xylose consumption was still

unsatisfactory during 6 h pre-hydrolysis SSCF. By extending the pre-hydrolysis time to 24 h or
longer, the xylose consumption was improved significantly. In order to better understand the
reasons for such improvement, the hydrolysate slurries after 6 h pre-hydrolysis and 24 h prehydrolysis were studied and compared. We found that the glucose concentration after prehydrolysis was the critical factor that determined cell viability and hence xylose consumption
o

during SSCF. Low temperature (30 C) and ethanol inhibition were shown to be the factors
limiting hydrolysis rate and hence productivity during SSCF.

65

4.1 Introduction
Corn stover is an agricultural residue abundantly available in the US, which makes it a
suitable feedstock for biofuel production (Perlack et al., 2005). Regarding biofuel production
from corn stover, much effort has focused on environmental impacts (Mann et al., 2002), energy
analysis (Luo et al., 2009), and technology development (Lau & Dale, 2009). Lack of
commercial conversion technologies is currently a bottleneck.
TM

Due to the merits of high digestibility and high fermentability of AFEX

pretreated

biomass, a yield of 191.5 g ethanol per kg corn stover was realized without detoxification or
TM

nutrient supplementation after AFEX

pretreatment (Lau & Dale, 2009). However, these

experiments were done by separate hydrolysis and fermentation (SHF) which has issues of
higher capital cost, longer processing time, higher contamination risk and sugar inhibition for
enzymatic hydrolysis compared to simultaneous saccharification and co-fermentation (SSCF)
(Taherzadeh & Karimi, 2007). While SSCF is generally superior to SHF, there are still several
concerns about this process such as the optimum temperature discrepancies between
saccharolytic enzymes and fermentation microbes as well as ethanol inhibition on enzymes
o

(Taherzadeh & Karimi, 2007). S. cerevisiae 424A(LNH-ST) works best at 30 C (Lau & Dale,
o

2009) while, the Trichoderma enzymes have an optimum temperature of around 50 C
TM

(Taherzadeh & Karimi, 2007). Therefore, the SSCF of AFEX
TM

(AFEX

pretreated switchgrass

o

-SG) was performed at 30-35 C (Chapter III). Nevertheless, the sugar conversions

during SSCF were still higher compared to SHF. That study also identified the inhibitory effect
of unhydrolyzed biomass solids on xylose fermentation, which caused poor xylose consumption.

66

By applying two-step SSCF, which mainly hydrolyzes xylan and ferments xylose prior to
glucose, xylose consumption was greatly improved. Xylose fermentation before glucose has the
advantages of alleviating redox imbalance and promoting xylose consumption due to the
presence of oxygen and many nutrients at the beginning as well as avoiding ethanol inhibition on
xylose consumption. The key for the two-step SSCF is to keep the glucose release rate at a low
level during the xylose fermentation period so that it does not affect xylose uptake by S.
cerevisiae (Bertilsson et al., 2008).
In order to improve our understanding of the SSCF process, the performance of two-step
TM

SSCF and traditional SSCF were further compared on AFEX
TM

(AFEX

pretreated corn stover

TM

-CS). The effect of unhydrolyzed AFEX

-CS solids and the causes of the
TM

unsatisfactory xylose consumption during SSCF of AFEX

-CS were studied. The effect of

ethanol and temperature on enzymatic hydrolysis was also investigated.
4.2 Materials and Methods
TM

4.2.1 AFEX

treated corn stover

Please see Chapter II section 2.2.1.
4.2.2 Microorganism and seed culture preparation
Seed cultures of S. cerevisiae 424A(LNH-ST) were prepared on YEP medium (10 g/L yeast
extract and 20 g/L tryptone) with 75 g/L glucose and 25 g/L xylose in a 250 ml Erlenmeyer flask
with a working volume of 100 ml. Frozen glycerol stock was used for inoculation and the initial
o

seed culture OD600 was 0.1. The culture was incubated at 30 C and 150 rpm under microaerophilic conditions for 24 h.

67

TM

4.2.3 Separate hydrolysis and fermentation (SHF) of AFEX

-CS

TM

Enzymatic hydrolysis (EH) of AFEX

-CS was performed in a 250 ml baffled flask with

100 g total mixture. The EH procedure is shown in Chapter II section 2.2.2. After 96 h
hydrolysis, the hydrolysate was harvested by centrifuge at 14,000 rpm for 30 min to remove
unhydrolyzed solids and was then sterile filtered through a 0.22 µm filter. The filtered
o

hydrolysate was stored at 4 C in a sterile bottle for fermentation use. For study of the
unhydrolyzed solids effect on fermentation, the hydrolysate was used as is without centrifuge
and filtration (unfiltered hydrolysate).
Fermentation of hydrolysate procedure is the same as described in Chapter III section 3.2.5.
4.2.4 Simultaneous Saccharification and Co-fermentation
SSCF was conducted using the same enzymes and biomass loading as used for the SHF
experiments. SSCF was also conducted in a 250 ml baffled flask with a total mixture weight of
100 g. For the traditional SSCF process, all the enzymes were added at the beginning and preo

hydrolyzed the biomass at 50 C, pH 4.8 and 250 rpm. After 6 h pre-hydrolysis, the temperature,
o

pH and agitation rate were changed to 30 C, 5.5 and 180 rpm, respectively, and yeast cells were
inoculated at an OD600 of 2.0 to initiate the SSCF. For two-step SSCF, xylanase, pectinase and
25% of the Spezyme CP enzymes were added at the beginning for 6 h pre-hydrolysis. The other
75% of the Spezyme CP and all of Novozyme 188 enzymes were added 48 h after yeast
inoculation. Glucose, xylose and ethanol concentrations were analyzed using HPLC with a
Biorad Aminex HPX-87H column as described previously in Chapter II section 2.2.7.

68

TM

4.2.5 Preparation of washed AFEX

-CS

The procedure used is the same as described in Chapter III section 3.2.7.
4.2.6 Mass balance
Please see Chapter III section 3.2.8.
4.2.7 Measurement of viable cell density
Please see Chapter II section 2.2.8.
4.3 Results and Discussion
4.3.1 Traditional SSCF process vs. Two-step SSCF process
The two-step SSCF process with xylose fermentation ahead of glucose fermentation
resulted in better xylose consumption and higher ethanol yields compared to traditional SSCF
TM

when processing AFEX

-SG. Here the performance of traditional SSCF and two-step SSCF on

TM

AFEX

-CS was compared (Fig. 19). For the traditional process, xylose in the fermentation

Concentration (g/L)

broth remained at a nearly constant concentration (around 11 g/L) during the whole process time.
45
Glc. (1)
Xyl. (1)
EtOH (1)
40
Glc. (2)
Xyl. (2)
EtOH (2)
35

30
25
20
15
10
5
0

0

24

48

72 96 120 144 168
Time (h)
TM
Figure 19 SSCF of AFEX -CS by applying two processes.
Process 1 (Traditional SSCF): 6h pre-hydrolysis using all of the enzymes followed by
inoculation of yeast; Process 2 (two-step SSCF): 6 h pre-hydrolysis using xylanase, pectinase
and 1/4 Spezyme CP followed by inoculation of yeast, the rest of Spezyme CP and Novozyme
188 was added at 48 h.
69

This indicates that the xylose consumption rate by the yeast was the same as the xylose release
rate by the enzymes. At the end of the process, a concentration of 10.6 g/L xylose was left in the
broth unutilized. The two-step SSCF process did not improve xylose fermentation. About 12.5
g/L xylose remained in the broth at the end of the process. Ethanol produced during the two
processes was 36.3 and 32.3 g/L, respectively.
a)
Filtered Hyl.
UnFiltered Hyl. (Solids: 90 g/L)
Hyl.+ washed AFEXTM-CS (140 g/L)
SSCF
SSCF of washed biomass +YEP

25

Xylose (g/L)

20
15
10
5
0

b)

0

24

48

72 96
Time (h)

120 144 168

1.8x108
1.6x108

CFU/ml

1.4x108
1.2x108
1.0x108
8.0x107
6.0x107
4.0x107
2.0x107
0.0
0

24

48

72 96 120 144 168
Time (h)
Figure 20 Comparison of xylose fermentation (a) and cell viability (b) during SHF and
traditional SSCF.
SHF fermentations were carried out in filtered hydrolysate (without solids), unfiltered
hydrolysate (with unhydrolyzed solids), and filtered hydrolysate with 140 g/L washed
TM
TM
AFEX -CS. 6 h pre-hydrolysis SSCF was performed on unwashed AFEX -CS without
TM

nutrients supplementation and washed AFEX
g/L & yeast extract 5 g/L).

-CS with supplementation of YEP (tryptone: 10

70

Different biomass properties/compositions might have caused different results for twoTM

step SSCF on AFEX

TM

-SG and AFEX

TM

-CS. AFEX

-SG contains 25.3% xylan and was

easily solubilized by using only xylanase and pectinase (released around 25 g/L xylose after preTM

hydrolysis). On the other hand, AFEX

-CS contains 20.4% xylan and was found to be very

difficult to solubilize using only xylanase and pectinase even with a 10% higher dosage. Because
of this solubilization issue, there was almost no free water and little xylose available for
fermentation after pre-hydrolysis. Therefore, we tried adding part of Spezyme CP enzyme, which
has low β-glucosidase activity, to help solubilize the biomass during the pre-hydrolysis. Several
loadings were tried. The one reported here with 25% of the total Spezyme CP enzymes added at
the start gave the best xylose consumption results. We found it difficult to balance solubilization
of the biomass with minimization of the glucose release so as not to influence xylose
fermentation. Apparently, 25% of the total Spezyme CP enzymes gave a too high glucose release
rate and negatively affected xylose fermentation. The effect of cellulases loading on xylose
consumption was reported previously in Chapter III. There are two potential ways to apply twoTM

step SSCF on AFEX

-CS and give good results: (i) by increasing the amount of xylanase and

pectinase use for pre-hydrolysis; (ii) by using pure endo-glucanase to help solubilize biomass.
However, both methods will increase the cost of the process since enzymes are still expensive for
cellulosic ethanol production (Humbird et al., 2011). Since the two-step SSCF method did not
TM

-CS, we subsequently chose to work on

help improve the xylose fermentation on AFEX

traditional SSCF to better understand the poor xylose fermentation issue.

71

4.3.2 Effect of unhydrolyzed solids on xylose fermentation
It was found that the inhibitory effect of unhydrolyzed solids was the major cause for poor
TM

xylose fermentation during SSCF of AFEX
TM

solids of AFEX

-SG. Therefore, we investigated the unhydrolyzed

TM

-CS and AFEX

-CS biomass itself on xylose fermentation in hydrolysate

TM

-SG (Fig. 20). The hydrolysate containing 52.3 g/L glucose

the same way as we did for AFEX

and 25.5 g/L xylose was obtained by 96 h enzymatic hydrolysis. Surprisingly, no inhibitory
effect of those solids on xylose fermentation (Fig. 20a) and cell viability (Fig. 20b) was observed
except for a longer lag phase for the unfiltered hydrolysate fermentation (with unhydrolyzed
solids 90 g/L). Xylose was consumed to levels below 5 g/L in 168 h, which was a typical rate for
TM

AFEX

-CS SHF fermentation using S. cerevisiae 424A(LNH-ST) (Lau & Dale, 2009).

Comparing to SHF fermentations, SSCF showed lower cell viability (Fig. 20b) and poor xylose
fermentation (Fig. 20a). It appears that the xylose fermentation was strongly correlated with cell
TM

-CS with

viability. To test this hypothesis, we performed SSCF on washed AFEX

supplementation of YEP. Using this approach, we found the cell viability was enhanced to a
similar level as in SHF (Fig. 20b) and the xylose consumption was also improved to
approximately the SHF level with the final xylose concentration around 3 g/L (Fig. 20a). The
TM

correlation between cell viability and xylose fermentation was also observed on AFEX
TM

The unhydrolyzed solids of AFEX

-SG.

-SG severely reduced cell viability (data not shown), which
TM

was likely the cause for poor xylose fermentation in the presence of AFEX

72

-SG solids.

TM

Since the poor xylose fermentation during SSCF of AFEX

-CS was probably not due to

the solids inhibition and xylose consumption during fermentation of unfiltered hydrolysate was
good, which could be considered as an SSCF process with 96h pre-hydrolysis period, we decided
it would be worthwhile to investigate the effect of pre-hydrolysis time.
4.3.3 Effect of pre-hydrolysis time on SSCF performance
a)

30
6h pre-hydrolysis
24h pre-hydrolysis
48h pre-hydrolysis
96h pre-hydrolysis

Xylose (g/L)

25
20
15
10
5
0

0

24

48

b)

72 96 120 144 168
Time (h)

1.4x108

CFU/ml

1.2x108
1.0x108
8.0x107
6.0x107
4.0x107
2.0x107
0.0
0

24

48

72 96 120 144 168
Time (h)

Figure 21 Effect of pre-hydrolysis time on xylose fermentation (a) and cell viability (b) during
traditional SSCF.
The glucose concentrations after 6, 24, 48, and 96 h pre-hydrolysis were 23.5, 46.0, 50.2 and
52.3 g/L, respectively.

73

Four pre-hydrolysis times were tested (6, 24, 48, and 96 h) during SSCF (Fig. 21). As
discussed above, the xylose consumption during 6 h pre-hydrolysis SSCF was not as good as
during SHF experiments. However, as the pre-hydrolysis time was extended to 24, 48 or 96 h,
the xylose consumption was improved and achieved the same level as SHF (Fig. 21a). The yeast
cell viabilities were also enhanced (Fig. 21b), which is a possible reason for improved xylose
consumption. The correlation between cell viability and xylose consumption was further
validated by these results.
The initial purpose of pre-hydrolysis was to solubilize the biomass quickly at a high
temperature so that a liquid, more homogenous environment could be formed for yeast
fermentation. Though 6 h pre-hydrolysis solubilized the biomass well, the xylose fermentation
during SSCF was poor. Hence, in addition to solubilizing the biomass, it is important to create a
better environment in which to initiate yeast cell growth and thereby xylose consumption. Prehydrolysis time directly influences the concentration and composition of unhydrolyzed solids as
well as the composition of liquid hydrolysate. To understand why a longer pre-hydrolysis time is
superior, we studied the effects of both the solids portion of the medium and the liquid portion on
xylose fermentation and yeast cell viability.
4.3.4 Effect of solids and liquids from different pre-hydrolysis on xylose fermentation and
yeast cell viability
After pre-hydrolysis for different time periods, the unhydrolyzed solids were harvested by
centrifugation and washed using water and 1M sodium chloride to remove both degradation
products and enzymes adsorbed on the solids. Removal of enzymes avoided further sugar release
and changes in solids composition. The compositions of solids obtained were determined using
the NREL protocol (Sluiter et al., 2004) and are shown in Table 8. The effects of such solids on

74

xylose fermentation and yeast cell growth were investigated in hydrolysate produced in 96 h (Fig.
22). The solids concentrations were the same as after pre-hydrolysis during SSCF and ranged
from 90 to 135 g/L (Table 8). No significant differences of xylose consumption (Fig. 22a) and
cell viability (Fig. 22b) were observed between fermentations with 6 h pre-hydrolyzed solids and
fermentations with longer time pre-hydrolyzed solids. All the xylose consumption rates and cell
viabilities were as good as achieved during SHF. Therefore, the solids portion of the media is
probably not the factor causing xylose consumption difference among the SSCFs using different
pre-hydrolysis time.

Table 8 Concentration and composition of pre-hydrolyzed solids

a
c

Solids composition
Acid
Glucan
Arabinan
Xylan(%)
insoluble
(%)
(%)
lignin (%)
0
195
34.1
20.4
4.2
11.0
6
135±5.0 27.0±0.1 15.5±0.4
1.9±0.0
27.9±0.6
24
117±2.4 22.7±1.3 13.9±0.6
1.8±0.0
33.4±1.4
48
102±6.8 21.9±1.2 13.0±0.6
1.8±0.0
34.4±1.5
96
90±3.3
20.0±1.2 12.3±0.9
1.8±0.0
37.0±1.9
a
o
Pre-hydrolysis was carried out at 50 C, 250 rpm, pH 4.8
Prehydrolysis
time (h)

b
c

Solids
conc.
b
(g/L)

Ash(%)
6.1
7.2±0.3
8.5±0.1
8.8±0.4
8.7±0.2

Unhydrolyzed solids, dry weight basis
Composition analysis was performed on unhydrolyzed solids using the NREL protocol

75

a)

30
6h pre-hydrolysis
24h pre-hydrolysis
48h pre-hydrolysis
96h pre-hydrolysis

Xylose (g/L)

25
20
15
10
5
0

0

20 40 60 80 100 120 140 160 180
Time (h)

b)

1.4x108
1.2x108

CFU/ml

1.0x108
8.0x107
6.0x107
4.0x107

2.0x107
0.0
0

20 40 60 80 100 120 140 160 180
Time (h)

Figure 22 Effect of pre-hydrolyzed solids on xylose fermentation (a) and cell viability (b).
Solids were harvested after hydrolyzing at different time and washed using water, 1M NaCl,
and water for 3 times. The washed solids were then washed again using 96h hydrolysate and
added into 96 h hydrolysate for fermentation study. The glucose concentration was 50 g/L.
The solids concentrations were the same as SSCF. Solids concentration and solids
compositions were shown in Table 8.

76

a)

24

6h hyl.
6h hyl.+Sugar adj.
6h hyl.+heat-treated+Sugar adj.
24h hyl.
24h hyl.+heat-treated

Xylose (g/L)

20
16
12
8
4
0
b)

0

10

20

30 40 50
Time (h)

60

70

80

10

20

30 40 50
Time (h)

60

70

80

14
12

OD600

10
8
6
4
2
0

0

Figure 23 Xylose consumption (a) and cell growth (b) during fermentations of hydrolysate from 6
h pre-hydrolysis and 24 h pre-hydrolysis.
Liquid hydrolysate was harvested after 6 h or 24 h hydrolysis by centrifuge and sterile filtration.
Fermentations were performed on both heat-treated and untreated hydrolysate. To fair
comparison, glucose and xylose in 6h hydrolysate was adjusted to the same concentrations as 24
h hydrolysate. The one with no sugar adjustment was also investigated. Oligomeric glucose and
xylose in 6h hydrolysate, 6 h heat-treated hydrolysate, 24 h hydrolysate, and 24 h heat-treated
hydrolysate were 4.0 and 17.1 g/L, 1.7 and 14.8 g/L, 3.6 and 15.2 g/L, and 5.3 and 16.2 g/L,
o
respectively. Heat treatment was conducted at 105 C for 10 min.
The liquid portion after 6 h pre-hydrolysis (6 h hydrolysate) and 24 h pre-hydrolysis
(24 h hydrolysate) were also obtained by centrifugation. Fermentations were performed under 5

77

different sets of conditions to investigate the effect of the liquid portion on xylose fermentation
and yeast cell growth (Fig. 23): (i) 6 h hydrolysate, (ii) 6 h hydrolysate with glucose and xylose
adjusted to the same concentrations as 24 h hydrolysate, (iii) heat-treated 6h hydrolysate with
glucose and xylose adjusted, (iv) 24 h hydrolysate and (v) heat-treated 24 h hydrolsyate. Heat
treatment was used to denature the enzymes in the hydrolysate. No significant differences were
observed for xylose consumption (Fig. 23a) and cell growth (Fig. 23b) between the 6 h
hydrolysate fermentations and 24 h hydrolysate fermentations with the same sugar
concentrations (no matter if heat-treated or not). The 6h hydrolysate fermentation (without
adjusting sugar concentrations) showed lower cell density and hence slower xylose consumption
but still consumed xylose to a similar low level as did the 24 h hydrolysate fermentation (Fig. 23).
In summary, the solids portion and the liquid portion prepared under different prehydrolysis regimes showed no significant effect on xylose fermentation and cell viability. The
improved cell viability and better xylose consumption for longer pre-hydrolysis SSCFs
compared to 6h pre-hydrolysis SSCF are likely due to the higher glucose concentrations
achieved after pre-hydrolysis. The glucose concentrations after 6, 24, 48, and 96h pre-hydrolysis
were 23.5, 46.0, 50.2 and 52.3 g/L, respectively. Assuming the yeast cell biomass yield per g
glucose was fixed, higher glucose concentrations would produce more yeast cells. During SSCF,
glucose was still being released, but the release rate was probably too slow to satisfy yeast cell
o

growth demand, especially at a low temperature (30 C) (Fig. 27a). Besides, S. cerevisiae
424A(LNH-ST) can hardly grow on xylose anaerobically might due to the deficiency of ATP
during xylose metabolism (Kuyper et al., 2004). Therefore, the cell growth was only seen in the
first 24 h (Fig. 21b), which was mostly the glucose fermentation period (Fig. 19). After 24 h, the
cell viabilities began to decrease (Fig. 21b). It seems that the first 24 h cell growth determines

78

the overall cell viability and hence xylose consumption and that the glucose concentration after
pre-hydrolysis determines the first 24 h cell growth. To test this hypothesis, we conducted the
following investigation.
4.3.5 Glucose concentration after pre-hydrolysis is the key for xylose fermentation during
SSCF
a)

25
6h pre-hydrolysis
24h pre-hydrolysis
6h pre-hydrolysis+Glc.

Xylose (g/L)

20
15
10
5
0
b)

0

24

48

72 96 120 144 168
Time (h)

1.6x108
1.4x108
CFU/ml

1.2x108
1.0x108
8.0x107
6.0x107
4.0x107
2.0x107
0.0
0

24

48

72 96 120 144 168
Time (h)
Figure 24 Effect of glucose supplementation on xylose consumption (a) and cell viability (b)
during 6 h pre-hydrolysis SSCF.
Glucose was supplemented after 6 h pre-hydrolysis to the similar concentration as 24 h prehydrolysis (around 46.0 g/L).

79

We supplemented glucose after 6 h pre-hydrolysis to 46.0 g/L (the same concentration as
24 h pre-hydrolysis) and conducted the SSCF study. The cell viability was enhanced to a similar
level as 24 h pre-hydrolysis SSCF (Fig. 24b) and the xylose consumption was also improved to a
similar level (Fig. 24a). This observation supports the hypothesis that the low glucose
concentration after 6 h pre-hydrolysis caused low viable cell density and thereby poor xylose
consumption during SSCF.
To further test this hypothesis, we substituted the Spezyme CP enzymes with Accellerase
enzymes which had higher activities and performed 6h pre-hydrolysis SSCF (Fig. 25). Glucose
concentration reached approximately 32.3 g/L after 6 h pre-hydrolysis, higher than that achieved
using Spezyme CP enzymes (23.5 g/L). Once again, the cell viability and xylose consumption
were better than those for Spezyme CP. As a result, after 168 h, 39.9 g/L ethanol was produced
with 3.2 g/L xylose left in the broth.
4.3.6 Mass balance comparison of different process
For a better understanding of the system, mass balance studies were performed on the

Concentration (g/L)

CFU/ml

processes of SHF, 24 h pre-hydrolysis SSCF using Spezyme CP enzymes and 6 h pre-hydrolysis
45
1.4x108
40
1.2x108
35
Glucose
1.0x108
30
Xylose
Ethanol
25
8.0x107
CFU/ml
20
6.0x107
15

10

4.0x107

5

2.0x107

0

0

24

48

72 96 120 144 168
Time (h)
TM
Figure 25 SSCF of AFEX -CS using Accellerase enzymes and 6 h pre-hydrolysis.
Accellerase 1500, Accellerase XY, and Multifect pectinase with the protein loadings of 24, 6,
and 6 mg/g glucan, respectively were used for this experiment.
80

SSCF using Accellerase enzymes. The sugar conversions and ethanol yield of SHF process (Fig.
26a) were lower than those observed in a previous study (Lau & Dale, 2009), likely due to
enzyme activity loss during storage in a refrigerator for almost 3 years. Compared to SHF, 24 h
pre-hydrolysis SSCF increased sugar conversions and hence increased ethanol yield from 168.0
TM

to 188.9 g ethanol per kg AFEX

-CS (Fig. 26b). 6 h pre-hydrolysis SSCF using Accellerase
TM

enzymes gave a higher ethanol yield (193.2 g ethanol per kg AFEX

-CS) (Fig. 26c).

Oligomeric sugar loss, especially oligomeric xylose loss, was an issue for all the processes.
Compared to SHF, SSCFs resulted in lower oligomeric glucose content in the fermentation broth.
However, the oligomeric xylose content was found to be similar in both processes.
Another feature of these processes is low ethanol productivity. The process ethanol
productivities for SHF, 24 h pre-hydrolysis SSCF using Spezyme CP enzymes and 6 h pre-1 -1

hydrolysis SSCF using Accellerase enzymes were 0.13, 0.20, and 0.23 g·L ·h , respectively.
SSCF shortened the overall process time by combining enzymatic hydrolysis and fermentation
o

and hence had a higher productivity than SHF. However, the low processing temperature (30 C)
and the presence of ethanol limited the hydrolysis rates (Fig. 27) during SSCF and hence limited
the further improvement of productivity. Thermo-tolerant strains and ethanol-tolerant enzymes
are required to fully exploit the merits of the SSCF process. It is interesting to note that with the
increase of ethanol concentration the hydrolysis rates dropped most rapidly for hydrolyses
o

conducted at 50 C, and the optimum temperature for enzymatic hydrolysis in the presence of
o

ethanol was found to be 45 C (Fig. 27).

81

a)

SHF
- Spezyme CP: 84.2 ml (7.4 g Protein)
- Novozyme 188: 85.3 ml (12.8 g Protein)
- Multifect Xylanase: 24.7 ml (0.86 g Protein)
- Multifect Pectinase: 17.4 ml (1.56 g Protein)

AFEX treated
corn stover: 1Kg
(Dry Weight)

Enzymatic
hydrolysis
96h

- Yeast inoculum
(Dry Weight):
5.3 g

- Glucose: 250.7 g (52.3 g/L)
- Xylose: 122.3 g (25.5 g/L)
- Glc-Oligomers: 38.7 g (7.8 g/L)
- Xyl-Oligomers: 85.4 g (17.3 g/L)

- Glucose: 378.7 g
- Xylose: 266.9 g

Solid residue
-Glucose: 89.3 g
-Xylose: 59.2 g

Conversion:
Glucan: 76.4%
Xylan: 77.8%
Ethanol Metabolic yield: 93.1%

Ferm.
168h

Ethanol: 168.0 g
(35.0 g/L)

- Xylose: 19.2 g (4.0 g/L)
- Glc-oligomers: 38.7 (7.8 g/L)
- Xyl-Oligomers: 85.4 g (17.3 g/L)

Figure 26 Mass balance for the SHF (a), SSCF with 24 h pre-hydrolysis (b), and SSCF with 6 h pre-hydrolysis using Accellerase
TM
enzymes (c) on AFEX -CS.

82

Figure 26 (cont’d)
b)
SSCF (24h pre-hydrolysis)
- Spezyme CP: 84.2 ml (7.4 g Protein)
- Novozyme 188: 85.3 ml (12.8 g Protein)
- Multifect Xylanase: 24.7 ml (0.86 g Protein)
- Multifect Pectinase: 17.4 ml (1.56 g Protein)
- Yeast innoculum
(Dry Weight):
5.3 g
AFEX treated
corn stover: 1Kg
(Dry Weight)

SSCF
24+168h

Ethanol: 188.9 g
(37.9 g/L)
- Xylose: 23.3 g (4.5 g/L)
- Glc-oligomers: 20.2 g (4.3 g/L)
- Xyl-Oligomers: 86.4 g (17.6 g/L)

- Glucose: 378.7 g
- Xylose: 266.9 g

Conversion:
Glucan: 85.2%
Solid residue
Xylan: 82.3%
-Glucose: 56.1 g
Ethanol Metabolic yield: 90.3%
-Xylose: 47.3 g

c)

SSCF (6h pre-hydrolysis)
- Accellerase 1500: 122.1 ml (8.18 g Protein)
- Accellerase XY: 70.6 ml (2.05 g Protein)
- Multifect Pectinase: 23.0 ml (2.05 g Protein)
- Yeast inoculum
(Dry Weight):
5.3 g
AFEX treated
corn stover: 1Kg
(Dry Weight)

SSCF
6+168h

- Glucose: 378.7 g
- Xylose: 266.9 g

Ethanol: 193.2 g
(39.9 g/L)
- Xylose: 16.3 g (3.2 g/L)
- Glc-oligomers: 18.9 g (4.0 g/L)
- Xyl-Oligomers: 82.1 g (16.8 g/L)

Solid residue
-Glucose: 53.2 g
-Xylose:44.7 g

83

Conversion:
Glucan: 86.5%
Xylan: 85.9%
Ethanol Metabolic yield: 88.0%

10.0

a)

30oC

Glucose (g/L)

9.5

35oC
45oC

40oC
50oC

9.0
8.5
8.0
7.5
7.0
6.5
6.0
5.5

b)

0

5

10

5.5

15 20 25 30
Ethanol (g/L)

30oC

35oC
45oC

Xylose (g/L)

5.0

35

40

45

40oC
50oC

4.5
4.0
3.5
3.0
2.5

0

5

10

15 20 25 30 35 40 45
Ethanol (g/L)
Figure 27 Effect of ethanol concentration on glucose (a) and xylose (b) release during
enzymatic hydrolysis at different temperature.
Enzymatic hydrolysis was carried out at pH 4.8, 200 rpm and 1% glucan loading for 24 h.
Spezyme CP, Novozyme 188, Mutifect Xylanase, and Mutlfict Pectinase loadings were 22.4,
38.4, 2.6 and 4.7 mg/g glucan, respectively.
4.4 Conclusions
In S. cerevisiae 424A(LNH-ST) fermentations, xylose consumption has a strong
correlation with yeast cell viability. Glucose concentration after pre-hydrolysis determines the
TM

cell viability and hence xylose fermentation during SSCF of AFEX
TM

AFEX

-CS. Unhydrolyzed

-CS solids has no inhibitory effect on cell growth and xylose fermentation. We found

o

that 45 C is the ideal temperature for SSCF.
84

CHAPTER V EFFECT OF SOLIDS LOADING ON SSCF

Abstract
A high solid loading is desired in an industrial relevant process. However, the sugar conversions
typically decrease with increasing solids loading during enzymatic hydrolysis and SSF. So far,
TM

no one has investigated solids loading effect on SSCF process, especially SSCF on AFEX

TM

pretreated biomass. In the present study, the solids loading effect during SSCF of AFEX

pretreated corn stover was investigated. Techno-economic analyses were done to show the solids
loading effect on the economics of cellulosic ethanol production. A decreased sugar conversion
with increasing solids loading was also observed during SSCF. The increased degradation
products and xylose accumulation with increasing solids loading were proven not to be the major
factors causing such phenomenon but they did affect the overall economics of cellulosic ethanol
production. 6% (w/w) glucan loading was shown as the optimal solids loading for SSCF on
TM

AFEX

corn stover.

85

5.1 Introduction
Processing lignocellulosic biomass at high solids loading is desired in a biorefinery. A
high solids loading process can potentially generate a high concentration of sugars and in turn a
high titer of products (such as ethanol), which facilitates products recovery (lower distillation
cost for ethanol production). In addition, high solids loading typically means less reactor volume
required, less energy for heating and cooling as well as less waste water (Kristensen et al., 2009).
Thus, both capital cost and operating cost can be reduced when a high solids loading process is
applied. However, sugar conversion during enzymatic hydrolysis normally linearly decreases
when solids loading increases (Cara et al., 2007; Hodge et al., 2008; Kristensen et al., 2009).
Numerous factors could be responsible for such phenomenon including increased degradation
products inhibition (Panagiotou & Olsson, 2007), increased lignin concentration (Pan, 2008),
enzyme non-productive adsorption to lignin (Kristensen et al., 2009), decreased water
concentration (Kristensen et al., 2009), mass transfer limitations (Hodge et al., 2008), and
increased end-product inhibition (Kristensen et al., 2009). Kristensen et al. (Kristensen et al.,
2009) concluded that the major cause for decreased sugar conversion at high solids loading was
decreased enzymes adsorption to biomass, which was inhibited by increased end-products
(sugars) concentration. SSF/SSCF removes the enzymatic hydrolysis products (sugars) and thus
reduces the sugar inhibition on enzymes. However, studies showed that sugar conversion still
decreases with increasing solids loading during SSF (Jørgensen et al., 2007; Mohagheghi et al.,
1992) but not as much as for SHF (Jørgensen et al., 2007). The inhibition from increased ethanol
concentration might be a major reason. However, the accumulated xylose during SSF could be
another factor limiting enzyme performances during SSF of lignocellulosic biomass since xylose
was not consumed by the SSF microbe. To the author’s knowledge, the effect of solids loading

86

on SSCF performance has not been studied. SSCF removes not only glucose but also xylose and
thereby might be able to reduce the effect of high solids loading.
In the present study, the goal is to study the effect of high solids loading on enzymatic
TM

hydrolysis and ethanol yield during SSCF of AFEX

-CS and to figure out which solids loading

is the most economic one based on current technologies.
5.2 Materials and Methods
TM

5.2.1 AFEX

pretreated corn stover

Untreated corn stover (GLBRC corn stover) with glucan content and xylan content of 36.2%
TM

and 22.5%, respectively, was AFEX

pretreated in a 5 gallon reactor in Michigan

TM

Biotechnology Institute (MBI). AFEX

pretreatment conditions were: ammonia to biomass
o

loading 1.0 g/g dry biomass, water loading 0.6 g/g dry biomass, temperature 100 C and
residence time 15 minutes.
5.2.2 Enzymatic hydrolysis
Enzymatic hydrolysis was carried out at 5%, 7% and 9% (w/w) glucan loading in a 250 ml
baffled flask with a biomass-enzyme-water mixture of 100 g. Enzymes applied included
Accellerase 1500, Accellerase XY and Multifect pectinase (Genencor Inc., USA) in protein
loadings of 20, 5 and 5 mg/g glucan, respectively. Hydrolysis was conducted in an incubator at
o

pH 4.8, temperature 50 C and 250 rpm for 7 days. After 7 days hydrolysis, the glucose
concentrations for 5%, 7% and 9% glucan loading were 48.7, 64.1 and 74.5 g/L, respectively.
The xylose concentrations were 25.7, 35.2, and 41.5 g/L, respectively.

87

5.2.3 Microorganism and seed culture preparation
S. cerevisiae 424A(LNH-ST) was used for fermentation of glucose and xylose in this study.
Seed culture preparation has been described in Chapter IV section 4.2.2.
5.2.4 Fermentation of enzymatic hydrolysate (SHF)
The fermentation procedure has been described in Chapter III section 3.2.5. Fermentations
were conducted on both sugar concentration unadjusted hydrolysates and sugar concentration
adjusted hydrolysates. The sugar concentration adjusted hydrolysates were prepared by adding
glucose and xylose to the hydrolysates to reach the same concentrations of glucose and xylose
(74.5 and 43.0 g/L, respectively) in different hydrolysates.
5.2.5 Simultaneous saccharification and co-fermentation (SSCF)
SSCF experiments were performed in the same baffled flask and the same enzyme loading
as enzymatic hydrolysis. Solids loading investigated included 4%, 5%, 6%, 7% and 9% (w/w)
glucan loading. The 6 h pre-hydrolysis SSCF procedure described in the Chapter IV section 4.2.4
TM

-CS,

was used for this study. SSCF experiments were conducted on unwashed AFEX
TM

unwashed AFEX

-CS with supplementation of YEP (yeast extract 5 g/L and tryptone 10 g/L),

as well as washed AFEX

TM

TM

-CS with supplementation of YEP. The washed AFEX

-CS was

prepared as described in Chapter III section 3.2.7.
Monomeric sugar conversions during SSCF were calculated based on the produced ethanol
and remaining sugars in the fermentation broth with the assumption of 90% ethanol metabolic
yield for fermentation.

88

5.2.6 Techno-economic analyses
Techno-economic analyses were performed based on a modified NREL 2011 model
(Humbird et al., 2011). The modified model was developed by a colleague Dr. Bryan Bals. The
model strips out biomass handling and pretreatment, and calculates capital, raw material, and
energy costs as well as total ethanol and electricity revenue based on inputs regarding biomass
composition, enzymatic hydrolysis and fermentation performance. The cellulase enzymes
production area was also removed in the model and replaced with a fixed cost of enzyme (on a
$/kg protein basis). In the modified model, AFEX

TM

pretreatment is assumed to be carried out in

the local biomass processing depots (LBPDs). The costs of biomass pre-processing including
pretreatment are assumed to be $50/tonne untreated biomass (Bals & Dale, 2012) with an
ammonia use of 22 g/kg biomass. The calculations for all the costs and revenues are based on the
NREL model and adjusted proportionally for differences in the composition, enzymatic
hydrolysis and fermentation inputs. Different areas of the biorefinery were adjusted based on
different inputs. For example, the costs of purifying the ethanol via molecular sieving were
adjusted based on the difference in ethanol produced via NREL's assumptions and the user's
assumptions, whereas the costs associated with fermentation were based on the differences in
solid loading and total residence time. The cost adjustments were performed linearly for energy
and raw material (for example, a 5% increase in ethanol production relative to NREL resulted in
a 5% increase in energy use in the molecular sieve), whereas the cost of capital was adjusted
based on the average sizing factor of equipment in that area (generally around 0.6). For example,
a 5% increase in ethanol production would increase the capital cost of the molecular sieving by
1.05^0.6= 3% increase. All financial assumptions (lifetime of the plant, interest rate on loan,
taxes, depreciation method, proportion of funding by equity, working capital, etc), all design

89

assumptions (equipment used, power requirements of equipment, design of wastewater treatment
and boiler/turbogenerator, etc), and all assumptions around the efficiency of the generator are
identical to the NREL model.
5.3 Results and Discussion
5.3.1 SSCF performances at different solids loading
4%, 5%, 6%, 7% and 9% glucan loading SSCF were performed using unwashed
TM

AFEX

-CS. The results were compared in Fig. 28. Time 0 was set for the beginning of SSCF

after 6 h pre-hydrolysis. Ethanol concentration increased with increasing solids loading but not
proportionally. Glucose conversion to ethanol by fermentation was all rapid even at 9% glucan
loading.

a)

35
4% GL
5% GL
6% GL
7% GL
9% GL

Glucose (g/L)

30
25
20
15
10
5
0

0

24

48

72 96
Time (h)

120 144 168

Figure 28 Glucose (a), xylose (b) and ethanol (c) profiles during SSCF of unwashed
TM
AFEX -CS at different glucan loading (GL).

90

Figure 28 (cont’d)

Ethanol (g/L)

b)

50
45
40
35
30
25
20
15
10
5
0

0

24

48

c)

72 96
Time (h)

120 144 168

72 96
Time (h)

120 144 168

35

Xylose (g/L)

30
25
20
15
10
5
0

0

24

48

91

However, the xylose consumption was reduced with increasing solids loading (Fig. 28b).
At 9% glucan loading, there was 26.1 g/L xylose remaining in the fermentation broth at the end
of fermentation, which was a huge sugar waste for a biorefinery. The reduced xylose
fermentation performance with increasing solids loading is probably a synergistic effect of
degradation products, ethanol and fermentation metabolites as already shown in Chapter II.
Solids concentration might be another factor decreased the overall xylose consumption. Higher
solids loading correspond to higher solids concentration during SSCF, which influence the
viscosity and shear stress (Pimenova & Hanley, 2004). However, it is illustrated in Chapter VII
Fig. 51 that solids concentration (even up to 200 g/L) has little effect on S. cerevisiae 424A
fermentation performance including xylose consumption. To further prove the small effect of
solids during SSCF, fermentations of solids-free hydrolysate were conducted.
5.3.2 Fermentation performances of solids-free hydrolysate at different solids loading
45

Xylose & EtOH (g/L)

40
35
30
25
20
15
10
5
0
0

24

48

72

96

Time (h)
Xylose.5%
EtOH.5%

Xylose.7%
EtOH.7%

Xylose.9%
EtOH.9%

TM

Figure 29 Xylose and ethanol profiles during fermentations of 5% glucan loading AFEX
TM

hydrolysate, 7% glucan loading AFEX

-CS

TM

-CS hydrolysate and 9% glucan loading AFEX

92

-CS.

This was an SHF experiment. Enzymatic hydrolyses were conducted at 5%, 7% and 9%
glucan loading for 7 days. The obtained hydrolysates with solids removed by centrifugation were
used for fermentation. Initial glucose concentrations for 5%, 7% and 9% glucan loading
hydrolysates were 48.7, 64.1 and 74.5 g/L, respectively. These glucose were all fermented to
ethanol in 24 h. The xylose consumption trend is similar to the one in SSCF (Fig. 29): the higher
solids loading the poorer xylose consumption, which proved that the solids during SSCF were
not the main factor limiting xylose fermentation. At the end, around 34.7, 39.1 and 42.0 g/L
ethanol, respectively, were produced from 5%, 7% and 9% glucan loading hydrolysate.
To illustrate the effect of increased degradation products concentration on xylose
fermentation performance, sugars in 5%, 7% and 9% glucan loading hydrolysate were all
adjusted to the same levels before fermentation was carried out (Fig. 30). High solids loading
TM

means more degradation products as well as more nutrients since AFEX

-CS contains both

ingredients. In this fermentation at the same sugar concentrations, cell density (OD) reached the
same level for all three glucan loading hydrolysates (Fig. 30). The benefits of more nutrients on
cell growth at higher solids loading were probably offset by the inhibition of more degradation
products. After 96 h fermentation, around 18.2, 12.8 and 8.0 g/L xylose were consumed in the
5%, 7% and 9% glucan loading hydrolysates, respectively. The increased concentration of
degradation products substantially increased the inhibition on xylose fermentation probably
through the synergistic effect with ethanol and fermentation metabolites.

93

55
36

50

32

40

28

35

24

30

20

25

OD600

Xylose & EtOH (g/L)

45

16

20

12

15

8

10

4

5

0

0
0

24
Xylose.5%
EtOH.5%
OD600.5%

48
72
96
Time (h)
Xylose.9%
Xylose.7%
EtOH.9%
EtOH.7%
OD600.9%
OD600.7%

Figure 30 Xylose, ethanol and OD600 profiles during fermentations of 5% glucan loading
TM

AFEX

TM

-CS hydrolysate, 7% glucan loading AFEX

-CS hydrolysate and 9% glucan loading

TM

AFEX -CS, with glucose and xylose concentrations all adjusted to 74.5 and 43.0 g/L,
respectively.

5.3.3 SSCF performances with YEP supplementation
One goal of this study is to see if the sugar conversions decrease with increasing solids
loading during SSCF with both glucose and xylose removed by fermentation. However, with the
fact that xylose removal was not satisfactory at high solids loadings, the SSCF results on
TM

unwashed AFEX

-CS might not strong enough to see the effect of xylose removal. Therefore,
TM

two other strategies were applied. One approach used SSCF of unwashed AFEX

-CS with

supplementation of YEP. Supplementation of YEP could help enhance yeast cell growth and
thereby improve xylose consumption as shown in Chapter II Fig. 11. The other method involved
TM

SSCF of washed AFEX

-CS with supplementation of YEP. The removal of degradation

94

products could definitely help improve xylose consumption. Experimental results are shown in
Fig. 31.
With the supplementation of YEP, xylose consumption was substantially improved during
TM

SSCF on unwashed AFEX

-CS (Fig. 31a), especially at solids loadings higher than 6% glucan

a) 30
Remaining xylose (g/L)

Unwashed AFEXTM-CS
25

Unwashed AFEXTM-CS + YEP

20

Washed AFEXTM-CS + YEP

15
10
5
0

b)

70

4%

5%
6%
7%
Glucan loading

Unwashed AFEXTM-CS

60
Ethanol (g/L)

9%

Unwashed AFEXTM-CS+YEP

50

Washed AFEXTM-CS+YEP

40
30
20
10
0

4%

5%
6%
7%
9%
Glucan Loading
Figure 31 Comparison of final xylose and ethanol concentrations during SSCF of unwashed
TM
TM
TM
AFEX -CS, unwashed AFEX -CS with supplementation of YEP and washed AFEX -CS
with supplementation of YEP at different glucan loading.

loading. For instance, the final (168 h) xylose concentration at 9% glucan loading SSCF was
reduced from 26.1 g/L without YEP to 8.9 g/L with YEP. Due to the improvement of xylose
95

fermentation, the ethanol yield was enhanced (Fig. 31b). For 9% glucan loading SSCF, ethanol
titer was enhanced from 45.4 g/L without YEP to 51.9 g/L with YEP. The xylose removal effect
TM

was even better when SSCF was performed on washed AFEX

-CS with supplementation of

YEP. In this situation, the final xylose concentrations for different solids loadings were all
reduced to around 1 g/L (Fig. 31a). As a result, an ethanol titer as high as 60.3 g/L was achieved
at 9% glucan loading (Fig. 31b). The successful removal of both glucose and xylose during
SSCF guaranteed the following analyses of the solids effect on sugar conversions and ethanol
yield during SSCF.
5.3.4 Effect of solids loading on monomeric sugar conversion and ethanol yield
In the SSCF experiments, a decreased monomeric sugar conversion with increasing solids
loading was still observed (Fig. 32a). The sugar conversion decreased with increasing glucan
TM

loading at a similar rate on unwashed AFEX

TM

-CS and on unwashed AFEX

-CS with YEP

(slope: -3.3 vs. -3.8). As already shown above, xylose accumulation in fermentation broth
increased with increasing glucan loading during SSCF on unwashed biomass. Substantial
reduction of xylose accumulation by supplementing YEP during SSCF for better xylose
fermentation did not help on the increase of sugar conversion. This indicates that some
determining factors might “mask” the effect of xylose inhibition on sugar conversion. With no
degradation products and no accumulation of xylose, the sugar conversion was still decreasing
with increase of glucan loading during SSCF of washed biomass. The slope for this decrease is 2.7 (Fig. 32a), which indicates that by eliminating the degradation products effect the decrease
rate was slightly slowed down. Therefore, neither accumulated xylose or degradation products
was not the determining factor causing the decrease of sugar conversion with increasing solids
loading.
96

Monomeric sugar conversion (%)

a)
100
90

y = -2.7445x + 101.34
R² = 0.8646

80
70

y = -3.8339x + 99.176
R² = 0.9416

y = -3.3286x + 95.392
R² = 0.9908

60
Unwashed biomass
Unwashed biomass + YEP
Washed biomass + YEP

50
40
3

4

5

6

7

8

9

10

11

Glucan loading (%)

b)

EtOH yield (g/g AFEXTM-CS)

0.35
0.30

y = -0.0145x + 0.3636
R² = 0.9307

0.25
0.20

y = -0.0148x + 0.2923
R² = 0.963

y = -0.0175x + 0.2949
R² = 0.9956

0.15
0.10

Unwashed biomass
Unwashed biomass + YEP
Washed biomass + YEP

0.05
0.00
3

4

5

6

7

8

9

10

11

Glucan loading (%)

Figure 32 Effect of solids loading on monomeric sugar conversion (a) and ethanol yield (b).
The monomeric sugar conversion was calculated by dividing released total glucose and xylose
by total glucose and xylose in the initial biomass.

97

Since it has been concluded that end-products (sugars) are the major cause for the
decreased sugar conversion at high solids loading during enzymatic hydrolysis (Kristensen et al.,
2009) and the sugar conversion still decreased with increasing solids loading during SSCF, it is
most likely that ethanol is the major cause of sugar conversion reduction during SSCF. The
impact of ethanol on enzymatic hydrolysis has been shown in Chapter IV Fig. 27. Experiments
similar to SSCF but without production of ethanol (e.g. perform SSCF at aerobic conditions) will
be an interesting study to further test this. Oligomeric sugars, especially oligomeric xylose,
increase in concentration with increasing solids loading and are another inhibitor for enzymatic
hydrolysis (Qing et al., 2010). Studies in Chapter III (Fig. 18) and Chapter VI (Fig. 26) showed
that SSCF did not reduce the production of oligomeric xylose much compared to SHF. Thus,
oligomeric xylose is an unavoidable inhibitor in high solids loading processes based on current
pretreatment and enzymes system.
5.3.5 Techno-economic analyses
For determining the optimal solids loading of SSCF for lignocellulosic ethanol production,
techno-economic analyses were performed using a modified NREL model. Minimum ethanol
TM

selling price as a function of glucan loading is shown in Fig. 33. On unwashed AFEX

-CS, 5%

glucan loading displayed the lowest MESP, which was around $2.18/gal. However, with the
improvement of xylose fermentation by supplementation of YEP, the lowest MESP shifted to 6%
glucan loading, which is around $2.02/gal. This indicates that xylose fermentation (besides sugar
conversion) is also a determining factor in terms of production economics. A better xylosefermenting strain or a better xylose-fermenting process can easily reduce the current MESP. On
TM

washed AFEX

-CS, 5%, 6%, 7% and 9% glucan loading showed almost the same lowest

MESP, which is around $1.85/gal. Removal of degradation products reduced the rate of sugar
98

conversion decrease with solids loading and in turn made high solids loading process more
profitable. However, the costs for washing and YEP addition, which are expensive, were not
TM

included. The techno-economic analyses on unwashed AFEX

-CS with supplementation of

TM

YEP and on washed AFEX

-CS were more about to show the benefits of xylose fermentation

improvement and reduction of degradation products inhibitors through pretreatment
improvement rather than showing the real economics for these two scenarios.

Unwashed AFEXTM-CS

2.8

MESP ($/gal.)

Unwashed AFEXTM-CS + YEP
Washed AFEXTM-CS +YEP

2.6
2.4
2.2
2.0
1.8
4

5

6
7
8
Glucan loading (%)

9

Figure 33 Effect of SSCF solids loading on minimum ethanol selling price (MESP).
TM
TM
SSCF was performed on unwashed AFEX -CS, unwashed AFEX -CS with
TM

supplementation of YEP (yeast extract 5 g/L and tryptone 10 g/L), and washed AFEX -CS
with supplementation of YEP. Techno-economic analyses were conducted with the assumptions
of ethanol metabolic yield 90%, no oligomeric sugars generation as well as no difference
between glucan/glucose and xylan/xylose in terms of monomeric sugar conversion and
conversion to ethanol. The costs for YEP addition and washing were not accounted.

This techno-economic model applied does not consider the threshold of ethanol
concentration that makes distillation economic, which is 40 g/L (Dien et al., 2003a). When
considering this factor, 6% (w/w) glucan loading should be the optimal solids loading for SSCF
99

TM

on unwashed AFEX

-CS since 6% is the minimum solids loading that made 40 g/L ethanol

titer possible (Fig. 31). In addition, the techno-economic analyses were done with the
assumptions of no oligomeric sugars generated and no difference between glucan/glucose and
xylan/xylose in terms of monomeric sugar conversion and conversion to ethanol, which made the
calculated MESPs lower than real values. But this did not affect the economic comparison of
different solids loadings.

100

TM

CHAPTER VI CONTINUOUS SSCF OF AFEX

PRETREATED CORN

STOVER

Abstract
High productivity processes are critical for commercial production of cellulosic ethanol. One
high productivity process - continuous hydrolysis and fermentation - has been applied in the
sugarcane and corn ethanol industries. However, little research related to this process has been
conducted on cellulosic ethanol production. In the present study, the kinetics of both batch SHF
(separate hydrolysis and co-fermentation) and SSCF (simultaneous saccharification and coTM

fermentation) of AFEX

TM

(Ammonia Fiber Expansion) pretreated corn stover (AFEX

-CS)

were studied and compared. Subsequently, a SSCF process was designed to evaluate continuous
TM

hydrolysis and fermentation performance on AFEX

-CS in a series of continuous stirred tank

reactors (CSTRs). Based on similar sugar to ethanol conversions (around 80% glucose-to-ethanol
conversion and 47% xylose-to-ethanol conversion), the overall process ethanol productivity for
continuous SSCF was 2.3 and 1.8 fold higher than batch SHF and SSCF, respectively. Slow
xylose fermentation and high concentrations of xylose oligomers were the major factors limiting
further enhancement of productivity.

101

6.1 Introduction
Batch SHF (i.e. performing enzymatic hydrolysis and fermentation separately) has been
TM

widely applied for ethanol production from AFEX

treated biomass using commercial

enzymes and a xylose-fermenting yeast strain Saccharomyces cerevisiae 424A (LNH-ST)
(Krishnan et al., 2010; Lau & Dale, 2009; Li et al., 2010; Shao et al., 2010; Zhong et al., 2009).
The maximum volumetric ethanol productivities during the glucose fermentation period in 6%
-1 -1

(w/w) glucan loading hydrolysates were all above 1.0 g·L ·h . However, the average
-1 -1

fermentation productivities were all below 1.0 g·L ·h , which does not meet industrial
production requirements (Dien et al., 2003a). If the ethanol productivity is calculated based on
not only fermentation time but also on enzymatic hydrolysis time, the productivity values are
even lower. SSCF, in which enzymatic hydrolysis and fermentation are performed in a single
bioreactor, is a potential way to reduce the overall reaction time and hence enhance productivity
(Taherzadeh & Karimi, 2007). However, the differences between optimum temperature and pH
between enzymatic hydrolysis (EH) and yeast fermentation decrease this advantage of SSCF as
shown in Chapter IV. Increasing solids loading and hence increasing ethanol titer is one way to
enhance productivity. However, with increasing solid loading the sugar conversions and xylose
fermentation efficiencies decrease as shown in Chapter V. The most economic solids loading
based on current enzymes and microbes is 6% (w/w) glucan loading as discussed in Chapter V.
Another way to improve ethanol productivity is by applying continuous hydrolysis and
fermentation (Brethauer & Wyman, 2010).
The continuous stirred tank reactor (CSTR) system, especially with more than two reactors,
provides a way to save the total reactor volume and hence enhance the volumetric productivity

102

(Brethauer & Wyman, 2010). With automation, CSTR reduces operations such as filling and
cleaning of the reactors, as done during batch fermentation, and thus reduces the overall
production cost. Moreover, the inocula required for CSTR seeding are negligible if the reactions
proceed for a long time (autocatalytic system). Continuous hydrolysis and fermentation has been
applied successfully in both industrial sugarcane ethanol production (Zanin et al., 2000) and corn
ethanol production in China (Bai et al., 2008; Brethauer & Wyman, 2010). The CSTR system
has also been studied on fermentations of pretreated biomass hydrolysate (such as enzymatic
TM

hydrolysate of diluted acid pretreated spruce wood) but not on AFEX

hydrolysate (Brethauer

& Wyman, 2010). In addition, to the author’ knowledge, a continuous system has not yet been
applied on SSCF of pretreated biomass.
In the present study, the ethanol productivity was enhanced by applying multi-stage
TM

CSTRs during fermentation of AFEX

-CS. Batch SHF and SSCF processes were studied and

compared first. To compare SHF and SSCF processes, ethanol productivity was calculated based
on enzymatic hydrolysis time plus fermentation time (process ethanol productivity). Since it
gave higher ethanol productivity, further investigation on SSCF was performed in a continuous
mode. Three flow rates for continuous process were investigated. Based on the mass balance,
batch SHF, SSCF and continuous SSCF were comprehensively compared.

6.2 Materials and Methods
TM

6.2.1 AFEX

pretreated corn stover

The pretreated corn stover used in this study was the same as used in Chapter II (see section
TM

2.2.1 for details). The AFEX

-CS (particle size 4 mm) was dried in the hood, until the

103

moisture content was below 10%, milled to pass a 0.5 mm sieve, and then used as is with no
washing, detoxification or nutrient supplementation for enzymatic hydrolysis or fermentation.
6.2.2 Enzymatic Hydrolysis
TM

Enzymatic hydrolysis (EH) of AFEX

-CS at 6% (w/w) glucan loading was conducted in a

500 ml fermentor (Biostat, Sartorius) with 400 g total mixture of biomass, enzymes and water.
Enzymes used for the hydrolysis include Accellerase 1500, Accellerase XY, and Multifect
pectinase (Genencor Inc, USA) with protein loadings of 24, 6, and 6 mg/g glucan, respectively.
o

Temperature, pH and agitation speed were controlled at 50 C, 4.8, and 800 rpm, respectively.
50 mg/L chloramphenicol (Cm) was used during hydrolysis to avoid microbial contamination.
Samples (5 ml) were taken during hydrolysis using a 10 ml syringe for sugar analysis. After 168
h hydrolysis, pH was adjusted to 5.5 and the hydrolysate was used for SHF fermentations at 220
o

rpm and 30 C without removing the unhydrolyzed solids.
6.2.3 Microorganisms and seed culture preparation
S.cerevisiae 424A was used for this study. For details, please see Chapter IV section 4.2.2.
6.2.4 Hydrolysate fermentation (SHF) and batch SSCF
Batch SHF/ SSCF fermentations were also carried out in 500 ml fermentors (Biostat,
Sartorius) with 400 g total mixture. For SHF fermentation, hydrolysate (168 h hydrolysis) was
inoculated with yeast cell pellets (obtained by centrifuging the seed culture) at an initial OD600
of 2.0. The fermentation was carried out for 168 h. For batch SSCF, the same glucan and enzyme
loadings which were employed in enzymatic hydrolysis were used. 24 h pre-hydrolysis of
TM

AFEX

o

-CS was performed first at 50 C, pH 4.8, and 800 rpm. Then temperature, pH, and
o

agitation speed were changed to 30 C, 5.5 and 220 rpm, respectively, with yeast cell pellets
104

inoculated at an initial OD600 of 2.0. After inoculation, fermentation was started and proceeded
for another 168 h. Samples (5 ml) were taken during SSCF using a 10 ml syringe for
sugar/ethanol analysis (mass balance). Triplicates of the experiments were performed. Averages
and standard deviations of the experimental results are shown in the figures and table.
6.2.5 Continuous SSCF
AFEX-CS,
Water,
Enzymes

Reactor A
(EH)

Yeast cells
Initial OD=2
(inoculate once)

Reactor B
(SSCF)

Hyl.

Broth

Reactor C
(SSCF)

220rpm,30oC
pH5.5

220rpm,30oC
pH5.5

250ml

800rpm,50oC
pH4.8

250ml

Broth

Reactor D
(SSCF)
220rpm,30oC
pH5.5

Broth

Reactor E
(SSCF)

Broth

220rpm,30oC
pH5.5
400ml

400ml

Figure 34 Continuous SSCF process diagram

Continuous SSCF was performed in a bio-reactor train (CSTR) as shown in Fig. 34. Five
bioreactors (Biostat, Sartorius) were connected in series. Reactor A was used for enzymatic
o

hydrolysis with temperature, pH, and agitation speed controlled at 50 C, 4.8 and 800 rpm,
o

respectively. Reactors B, C, D, E were for SSCF with conditions controlled at 30 C, pH 5.5 and
TM

220 rpm. AFEX

-CS, water and enzymes should keep flowing into Reactor A and the resulting

hydrolysate was pumped to Reactor B. The fermentation broth from Reactor B was pumped to
Reactor C, and then, D and E as shown in Fig. 34. Flow rates from one reactor to another were

105

the same. The volume of Reactor A (VA) depends on the flow rate and pre-hydrolysis time
(residence time). In this study, the pre-hydrolysis time was fixed at 24 h. Therefore, theoretically,
VA is equal to flow rate multiplied by 24 (Table 9). However, due to equipment limitations in the
TM

laboratory, we were unable to continuously feed solid AFEX

-CS into Reactor A and did not

have appropriate bioreactors to accommodate VA. Instead, we performed enzymatic hydrolysis
(400 g total) in another 500 ml bio-reactor (Biostat, Sartorius) for 24 h and then transferred the
contents to Reactor A and made sure the reactor was never empty. Therefore, the pre-hydrolysis
time for biomass in Reactor A varies from 24 h to 54 h depending on the flow rate. Between 24 h
and 54 h during the enzymatic hydrolysis period, sugar concentrations were relatively constant,
with glucose and xylose concentrations between 52-58 g/L and 31-35g/L, respectively (Fig. 35,
Fig. 38a, & Fig. 38b). Thus pseudo-steady state operation was achieved with this setup. Our
criterion for pseudo-steady state was that glucose and xylose concentrations in each SSCF
reactor varied less than 2 g/L in 28 h. The working volumes for Reactor B, C, D, and E were
250, 250, 400, and 400 ml, respectively.
TM

To initiate the continuous SSCF, 24 h pre-hydrolysis of AFEX

-CS was conducted in

Reactor B followed by inoculation of yeast cells (initial OD=2) and fermenting the mixture for
24 h. At the end of this fermentation period, hydrolysate was continuously pumped from Reactor
A to Reactor B and fermentation broth began to be continuously pumped from Reactor B to
Reactor C. Once Reactor C was filled, the fermentation broth in Reactor C started being pumped
to D. When all of the reactors reached the designated volume, the broth was continuously
pumped out from Reactor E. Samples were taken from time to time to verify pseudo-steady state
by measuring sugar/ethanol concentrations and for mass balance calculations.
106

6.2.6 Measurement of viable cells density
Please see Chapter II section 2.2.8.
6.2.7 Sugar, fermentation products analyses and mass balance
Glucose, xylose, ethanol, xylitol, glycerol and acetate concentrations were analyzed by
HPLC using Biorad Aminex HPX-87H column as described previously Chapter II section 2.2.7.
For mass balance details, please see Chapter III section 3.2.8.
6.3 Results and Discussion
6.3.1 Enzymatic hydrolysis comparison of SHF and SSCF
Two reactions existing in both SHF and SSCF processes are enzymatic hydrolysis and
fermentation, which determine fermentable sugar yields and conversion of fermentable sugars to
ethanol, respectively. Enzymatic hydrolysis was found to be the limiting step during SHF
process affecting the final ethanol yield (Lau & Dale, 2009). The performance of this reaction in
both processes was compared in Fig. 35. For the SHF process, enzymatic hydrolysis and
fermentation were performed separately. Therefore, enzymatic hydrolysis was conducted at its
o

optimal conditions (50 C, pH 4.8 and 800 rpm). For the SSCF process, the enzymatic hydrolysis
reaction was carried out at the same optimal conditions during the first 24 h (pre-hydrolysis). But
o

after 24 h, the conditions were changed to favor fermentation (30 C, pH 5.5 and 220 rpm).
Apparently, during the first 24 h the sugar release patterns were the same for SHF and for
SSCF. Glucan and xylan were hydrolyzed quickly to monomeric/oligomeric glucose and xylose,
respectively. Monomeric glucose and xylose concentrations reached 51.5 g/L and 32 g/L,
respectively with 5.9 g/L oligomeric glucose and 14.4 g/L oligomeric xylose produced. The
corresponding monomeric glucose and xylose, oligomeric glucose and xylose conversions were

107

64.5%, 57.0%, 7.4% and 25.6%, respectively. After 24 h, the hydrolysis rates in both processes
were decreased with SHF showing larger hydrolysis rates compared to SSCF. For SSCF, we

Concentration (g/L)

70

a.

60
50
40
30
20
10
0

0

24

48

72 96 120 144 168 192
EH time (h)
Poly-Glc-SHF
Poly-Glc-SSCF
Oligo-Glc-SHF
Oligo-Glc-SSCF
Mono-Glc-SHF
Mono-Glc-SSCF

24

48

45
b.

Concentration (g/L)

40
35
30
25
20
15
10
5
0

0

72 96 120 144 168 192
EH time (h)
Poly-Xyl-SHF
Poly-Xyl-SSCF
Oligo-Xyl-SHF
Oligo-Xyl-SSCF
Mono-Xyl-SHF
Mono-Xyl-SSCF
Figure 35 Time courses for polymeric (Poly-), oligomeric (Oligo-) and monomeric (Mono-)
glucose (a) /xylose (b) during enzymatic hydrolysis of batch SSCF and SHF.
o
For SSCF, prehydrolysis was performed for the first 24 h at 800 rpm, pH 4.8 and 50 C
o
followed by a change to 220 rpm, pH 5.5, and 30 C with yeast cells inoculated. Monomeric
sugars released during SSCF were calculated based on total input sugars and remaining
polymeric and oligomeric sugars.
108

observed about a 20 h lag phase after yeast inoculation (Fig. 36). During the lag phase, the
enzymatic hydrolysis reaction seemed to stop with almost no additional sugars released (Fig. 35).
The enzyme activities could have been fully inhibited by high concentrations of glucose and
o

xylose at low temperature (30 C). After the lag phase, with glucose consumed by yeast
fermentation, the monomeric glucose release continued to increase to 63.1 g/L at 192 h with the
oligomeric glucose decreased to 3.2 g/L. Lower oligomeric glucose concentration was found
during SSCF when compared to SHF. This could be a benefit of glucose removal by the yeast
fermentation and thereby removing the end-product inhibition of β-glucosidase. The released
monomeric xylose during SSCF after inoculation was increased slightly to 33.4 g/L (from 32 g/L)
with oligomeric xylose also increased to 15.6 g/L. Slow xylose removal by yeast fermentation,
and the presence of ethanol could be possible reasons why little xylose is released during SSCF
o

at low temperature (30 C) (Chapter IV, Fig. 27). Overall, for a given EH time, SHF showed
higher monomeric and total sugar conversions compared to SSCF. Thermo-tolerant strains,
which can ferment at high temperatures, closer to the optimum temperature of EH, could
improve EH performance during SSCF (Chapter IV, Fig. 27).
6.3.2 Fermentation comparison between SHF and SSCF
During the fermentation period, glucose was consumed first and rapidly (Fig. 36a) for
both SHF and SSCF. Xylose fermentation was much slower. After 168 h fermentation, SHF
consumed 64.7 g/L glucose and 31.2 g/L xylose with the remaining broth xylose concentration of
6.5 g/L. SSCF consumed 62.5 g/L glucose and 29.6 g/L xylose with the remaining xylose
concentration of 3.8 g/L. The same ethanol concentration (38.8 g/L) was produced. The ethanol
metabolic yields for SHF and SSCF were 78.0% and 81.3%, respectively (Fig. 40), which were

109

Concentration (g/L)

70
60
50

a.

Glc-SHF
Xyl-SHF
EtOH-SHF

Glc-SSCF
Xyl-SSCF
EtOH-SSCF

40
30
20
10
0

0 20 40 60 80 100 120 140 160 180

Fermentation time (h)

CFU/ml

Concentration (g/L)

8

1.2x10
16 b.
14
9.0x107
12
6.0x107
10
3.0x107
8
0.0
6
-3.0x107
4
-6.0x107
2
0
-9.0x107
0 20 40 60 80 100 120 140 160 180

Fermentation time (h)

Cumulative process
ethanol productivity (g/L/h)

Xylitol-SHF
Glycerol-SHF
Acetate-SHF
Xylitol-SSCF
Glycerol-SSCF
Acetate-SSCF
CFU-SHF
CFU-SSCF
0.50
c.
BatchSSCF
0.45
SHF(168h EH)
0.40
SHF(72h EH)
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00
0 20 40 60 80 100 120 140 160 180

Time (h)
Figure 36 Fermentation comparison of batch SSCF and SHF.
Only the fermentation period is shown in this figure. The productivity of SHF (72h EH) was
calculated based on the assumption that the fermentation kinetics were exactly the same as SHF
with 168h EH.
110

lower than the previous study on SHF (Lau & Dale, 2009). The reduced ethanol metabolic yields
in this study might be due to the presence of fine solids during fermentation in both SHF and
SSCF processes. The byproducts produced during SHF and SSCF process were xylitol (1.1 and
2.4 g/L) and glycerol (8.1 and 7.2 g/L) respectively (Fig. 36b). Glycerol production could help
reoxidize NADH produced during xylose fermentation in anaerobic conditions (Medina et al.,
2010) and hence alleviate redox imbalance and facilitate xylose metabolism, but would also
reduce ethanol metabolic yield. Acetate in the fermentation broth was mostly generated from
plant biomass by pretreatment and EH. During yeast fermentation, acetate concentrations were
slightly reduced (Fig. 36b).
Overall, based on 168 h EH, SHF showed much lower process ethanol productivity
compared to SSCF after the first 24 h of fermentation during which period SSCF was in the lag
phase (Fig. 36c). To compare SHF and SSCF more fairly, we also calculated the process ethanol
productivity of SHF based on 72 h EH with the assumption that the fermentation kinetics were
exactly the same as the 168 h EH’s. However, the process ethanol productivity of SHF was still
lower compared to SSCF. Productivity comparison was based on the similar sugar-to-ethanol
conversions. Glucose-to-ethanol conversion, xylose-to-ethanol conversion and ethanol
concentration were 78.3%, 46.9% and 36.0 g/L, respectively for SHF (72 h EH + 96 h ferm.) and
80.3%, 48.2% and 36.9 g/L, respectively for SSCF (24 h prehydrolysis + 120 h SSCF),
-1 -1

respectively. The process ethanol productivities for SHF and SSCF were 0.20 and 0.25 g·L ·h ,
respectively. Although the EH during SHF yielded more sugars at a given time compared to
SSCF, SSCF was more efficient in producing ethanol. Therefore, SSCF was chosen as the mode
to conduct continuous fermentation for enhancing the overall productivity.

111

70

4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
-0.5

60
50
40
30
20
10
0

0

24

Glc., Xyl. cons. rate and
EtOH prod. rate (g/L/h)

EtOH, consumed glc.and xyl. (g/L)

6.3.3 Fermentation kinetics of batch SSCF

48 72 96 120 144 168 192
Fermentation time (h)
Glc.Cons. Rate
Consumed Glc.
Xyl.Cons.Rate
Consumed Xyl.
EtOH Prod. Rate
EtOH
Figure 37 Consumed glucose/xylose, ethanol production, glucose/xylose consumption rate, and
ethanol production rate during batch SSCF.
Those rates can be used to estimate dilution rate for continuous SSCF.

Glucose/xylose consumption, ethanol production and their respective rates are shown in
Fig. 37. The maximum glucose consumption rate, xylose consumption rate and ethanol
-1 -1

production rate were 3.72, 0.44 and 2.71 g·L ·h , respectively. At such rates, the glucose,
xylose and ethanol concentrations were 28.36, 27.17 and 10.54 g/L, respectively (Fig. 36a).
Therefore, if the design of continuous SSCF aims to maximize glucose consumption rate, the
-1

dilution rate (D) should be approximately 0.13 h (the ratio of maximum glucose consumption
rate over the corresponding glucose concentration(Govindaswamy & Vane, 2010)). If the
-1

maximum xylose consumption rate is desired, the dilution rate should be around 0.02 h . To
-1

maximize the ethanol production rate, the dilution rate is 0.26 h . To avoid yeast cell washout,
the dilution rate should be controlled below the maximum specific cell growth rate 0.30 h
112

-1

(calculated from cell growth kinetics, Fig. 36b). Since the xylose consumption rate is much
lower compared to the glucose consumption rate, a multi-stage continuous fermentation applying
a larger working volume (reduced dilution rate and longer residence time) for xylose
fermentation reactors could achieve both high glucose consumption rates and high xylose
consumption rates (Govindaswamy & Vane, 2010). Due to the limitations of our laboratory
reactors, we set the volumes of SSCF reactors (B, C, D, E) at 250, 250, 400 and 400 ml,
respectively (Fig. 34).
6.3.4 Continuous SSCF
Continuous SSCF was conducted according to the diagram in Fig. 34. The flow rates
tested were 12, 24, and 48 ml/h with the corresponding dilution rates shown in Table 9. At a flow
rate of 12 ml/h when pseudo-steady state was reached, glucose consumption was almost
complete in the first SSCF reactor Reactor B (Fig. 38a) with the glucose consumption and
glucose-to-ethanol conversion at 56.5 g/L (Table 9) and 73.6% (Fig. 38e), respectively. The
overall glucose consumption and glucose-to-ethanol conversion were 60.9 g/L and 79.2%,
respectively. The cumulative xylose consumptions in Reactors B, C, D, and E were 11.3, 20.8,
25.0 and 27.5 g/L, respectively (Table 9) with the corresponding xylose-to-ethanol conversions
of 20.9%, 38.4%, 46.2% and 50.8%, respectively (Fig. 38f). When the flow rate was raised to 24
or 48 ml/h, the residence time was reduced with less glucose/xylose consumed and converted to
ethanol (Fig. 38a, b, c, e, f, Table 9). The overall glucose and xylose consumptions for a flow
rate of 24 ml/h was 58.5 and 25.5 g/L, respectively with glucose and xylose-to-ethanol
conversions of 77.0% and 47.5%, respectively. The overall glucose and xylose consumption for
48 ml/h were 55.3 and 17.7 g/L, respectively, with glucose and xylose-to-ethanol conversions of
75.6% and 33.3%, respectively. The final ethanol concentrations for the flow rates 12, 24, and 48

113

ml/h were 38.0, 36.5, and 33.8 g/L, respectively. Therefore, the overall trend was that sugar
consumption and ethanol production decreased as the flow rate increased. However, higher final
process ethanol productivity was observed at higher flow rates (Fig. 39a). The productivity
-1 -1

values for the flow rates 12, 24, and 48 ml/h were 0.28, 0.46 and 0.65 g·L ·h , respectively.
This means higher flow rates converted sugars more efficiently to ethanol but the overall sugar
conversions were reduced. For cost-efficient production of cellulosic ethanol, those two factors
(sugar conversion and ethanol productivity) must be balanced to reach an optimum economic
point.
Xylose fermentation rate was the major factor limiting ethanol productivity. For instance,
at a flow rate 12 ml/h, it required 3 reactors (250 ml + 400 ml + 400 ml) to consume 27.5 g/L
xylose while one reactor (250 ml) was sufficient to consume 56.5 g/L glucose (Fig 5a, b, Table
9). Slow xylose consumption is a major concern for lignocellulosic fermentation (Dien et al.,
2003a; Hahn-Hagerdal et al., 2007). Recently, engineered S. cerevisiae which can simultaneous
co-ferment cellobiose and xylose with the glucose repression of xylose fermentation minimized
(Ha et al., 2011), and naturally xylose-fermenting thermophilic bacterium with knockout of
organic acid genes (Shaw et al., 2008) showed interesting features in xylose fermentation. The
presence of considerable concentrations of oligomeric sugars is another factor affecting overall
sugar to ethanol conversion which in turn affects ethanol titer and ethanol productivity. For
instance, at a flow rate of 24 ml/h, the overall xylan-to-xylose (monomers and oligomers)
conversion reached a high of 86.9%. However, monomeric xylose conversion was merely 58.5%
with 28.4% of oligomeric xylose conversion (16.5 g/L; Fig. 40, Table 9), which resulted in
relatively low xylose-to-ethanol conversion (47.5%) since S. cerevisiae 424A was unable to
consume oligomeric sugars.

114

Table 9 Byproduct profiles of continuous SSCF at different flow rates
Consumed
glucose (g/L)
12 ml/h, VA=288 ml

-1

Reactor A(D=0.04h )
-1

Reactor B(D=0.05h )
-1

Reactor C(D=0.05h )
-1

Reactor D(D=0.03h )
-1

Reactor E(D=0.03h )
24 ml/h, VA=576 ml

-1

Reactor A(D=0.04h )
-1

Reactor B(D=0.10h )
-1

Reactor C(D=0.10h )
-1

Reactor D(D=0.06h )
-1

Reactor E(D=0.06h )
48 ml/h, VA=1152 ml

-1

Reactor A (D=0.04h )
-1

Reactor B (D=0.19h )
-1

Reactor C (D=0.19h )
-1

Reactor D (D=0.12h )
-1

Reactor E (D=0.12h )

Consumed
xylose (g/L)

Oligo-glc.
(g/L)

Oligo-xyl.
(g/L)

Xylitol
(g/L)

Glycerol
(g/L)

Acetate
(g/L)

0

0

6.19±0.43

15.13±1.13

0.00±0.00

0.56±0.02

1.86±0.02

56.54±0.92

11.32±0.33

5.50±0.08

16.40±0.29

0.38±0.02

5.45±0.11

1.47±0.01

56.69±0.94

20.80±0.77

4.89±0.15

16.38±0.23

0.68±0.01

5.59±0.03

1.40±0.06

57.52±1.22

25.00±1.31

4.56±0.34

16.58±0.37

1.04±0.06

6.38±0.28

1.35±0.01

60.58±1.81

27.49±1.41

4.18±0.41

16.33±0.42

1.10±0.02

6.63±0.17

1.40±0.02

0

0

6.96±0.44

16.14±0.92

0.00±0.00

0.53±0.02

1.85±0.05

46.42±2.11

4.55±0.84

6.51±0.14

17.07±0.12

0.00±0.00

3.63±0.19

1.55±0.05

51.11±0.56

10.12±0.66

5.88±0.10

17.56±0.08

0.37±0.03

5.15±0.21

1.43±0.05

54.09±2.23

19.75±1.03

5.35±0.06

16.99±0.13

0.77±0.01

6.38±0.17

1.41±0.03

58.47±1.49

25.51±0.94

4.87±0.07

16.52±0.05

0.93±0.01

6.97±0.21

1.44±0.03

0

0

6.90±0.25

15.96±0.81

0.00±0.00

0.52±0.01

1.83±0.04

31.06±2.46

2.02±0.42

6.87±0.09

16.49±0.02

0.00±0.00

2.07±0.26

1.63±0.08

50.58±0.20

4.99±0.18

6.37±0.04

17.03±0.40

0.05±0.11

4.00±0.21

1.44±0.07

55.92±0.18

13.65±0.64

5.89±0.02

17.15±0.40

0.56±0.03

5.45±0.15

1.36±0.04

55.34±0.47

17.67±0.07

5.64±0.04

17.05±0.03

0.74±0.01

5.96±0.14

1.33±0.02

115

35 b.

50

30

Xylose (g/L)

Glucose (g/L)

60 a.

40
30
20
10

15
10
0

12

24

Flow rate (ml/h)

ReactorA
ReactorD

ReactorB
ReactorE

ReactorC

1.4x108

24

Flow rate (ml/h)

ReactorA
ReactorD
d.

ReactorB
ReactorE

48
ReactorC

1.2x108
1.0x108
8.0x107
6.0x107
4.0x107
2.0x107
0.0
12

24

Flow rate (ml/h)

12

24

Flow rate (ml/h)

ReactorA
ReactorD

ReactorB
ReactorE

24

Flow rate (ml/h)

ReactorA
ReactorD

ReactorC

(Glc -to-EtOH)

100
90 e.
80
70
60
50
40
30
20
10
0

ReactorB
ReactorE

Conversion (%)

ReactorA
ReactorD

12

48

48
ReactorC

100
90 f.
80
70
60
50
40
30
20
10
0

ReactorB
ReactorE

48
ReactorC

(Xyl -to-EtOH)

40
c.
35
30
25
20
15
10
5
0

12

48

CFU/ml

Ethanol (g/L)

20

5

0

Conversion (%)

25

12

24

Flow rate (ml/h)

ReactorA
ReactorD

ReactorB
ReactorE

48
ReactorC

Figure 38 Comparison of three flow rates (12, 24, and 48 ml/h) on glucose (a) and xylose (b)
consumption, ethanol production (c), viable cell density (d), glucose-to-ethanol (e) and xyloseto-ethanol (f) conversions during continuous SSCF.
The glucose-to-ethanol and xylose-to-ethanol conversions were calculated based on total glucose
TM
and total xylose in AFEX -CS.
116

Process EtOH productivity (g/L/h)

Cumulative process
EtOH productivity (g/L/h)

0.8
a.
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
12

48

24

Flow rate (ml/h)
ReactorB
ReactorD

ReactorC
ReactorE

0.5
b.
0.4
0.3
0.2
0.1
0.0
SHF

Batch
SSCF

Continuous
SSCF

Figure 39 Process ethanol productivity comparisons of different flow rates during continuous
SSCF (a), and productivity comparisons of SHF, batch SSCF and continuous SSCF (b).
In (b), process ethanol productivities were compared based on similar glucose-to-ethanol and
xylose-to-ethanol conversions. For SHF, the calculation was based on 72 h enzymatic hydrolysis
and 96 h fermentation assuming the fermentation kinetics are the same as fermentation of 168 h
hydrolysate. The glucose-to-ethanol conversion, xylose-to-ethanol conversion and ethanol
concentration were 78.3%, 46.9% and 36.0 g/L, respectively. For batch SSCF, the calculation
was based on 24 h prehydrolysis and 120 h SSCF. The glucose-to-ethanol conversion, xylose-toethanol conversion and ethanol concentration were 80.3%, 48.2% and 36.9 g/L, respectively. For
continuous SSCF, the calculation was based on flow rate 24 ml/h. The glucose-to-ethanol
conversion, xylose-to-ethanol conversion and ethanol concentration were 77.0%, 47.5% and
36.5 g/L, respectively.
6.3.5 Comparison of continuous SSCF with batch SHF and batch SSCF
Mass balances were completed for comprehensive comparison of those processes (Fig.
40). The highest sugar conversions were accomplished by SHF (168 h EH+168 h fermentation)
followed by batch SSCF (24 h pre-hydrolysis + 168 h SSCF). The highest ethanol yield (195.1 g
TM

ethanol/kg AFEX

-CS) was achieved by batch SSCF (24 h prehydrolysis + 168 h SSCF)
TM

followed by SHF (168 h EH+168 h fermentation) 194.1 g ethanol/kg AFEX

117

-CS and

A)

Yeast cells: 0.24g

Continuous SSCF
(Flow rate 12 ml/h)

Yeast cells: 0.24g
Enzyme protein loading:
Accellerase 1500: 8.184g
Accellerase XY: 2.046g
Mutifect Pectinase: 2.046g

Continuous SSCF
(Flow rate 24 ml/h)

1.0 kg AFEX-CS
-Glcose: 378.9 g
-Xylose: 265.8 g
Yeast cells: 0.24g

Yeast cells: 4.91g

Yeast cells: 4.91g

Continuous SSCF
(Flow rate 48 ml/h)

Batch SSCF
(24h prehydrolysis
+120h SSCF)

Batch SSCF
(24h prehydrolysis
+168h SSCF)

Remaining sugars and EH conversion
-Mono-Glc: 3.9 g; 80.2%
-Mono-Xyl: 21.8 g; 59.0%
-Oligo-Glc: 19.6 g; 5.2%
-Oligo-Xyl: 74.6 g; 28.1%
-Poly-Glc: 55.3 g
-Poly-Xyl: 34.45 g
-Ethanol: 187.5 g
-Ethanol metabolic yield: 84.5%
-Process productivity: 0.28 g/L/h
-Mono-Glc: 2.2 g; 77.0%
-Mono-Xyl: 29.1 g; 58.5%
-Oligo-Glc: 29.6 g; 7.8%
-Oligo-Xyl: 75.5 g; 28.4%
-Poly-Glc: 60.9 g
-Poly-Xyl: 34.8 g
-Ethanol: 177.6 g
-Ethanol metabolic yield: 84.4%
-Process productivity: 0.46 g/L/h
-Mono-Glc: 0.2 g; 75.7%
-Mono-Xyl: 64.5 g; 57.5%
-Oligo-Glc: 33.7 g; 8.9%
-Oligo-Xyl: 80.8 g; 30.4%
-Poly-Glc: 58.4 g
-Poly-Xyl: 32.2 g
-Ethanol: 173.4 g
-Ethanol metabolic yield: 90.7%
-Process productivity: 0.65 g/L/h
-Mono-Glc: 3.4 g; 80.9%
-Mono-Xyl: 32.5 g; 61.3%
-Oligo-Glc: 22.4 g; 5.9%
-Oligo-Xyl: 75.5 g; 28.4%
-Poly-Glc: 50.1 g
-Poly-Xyl: 27.2 g
-Ethanol: 184.7 g
-Ethanol metabolic yield: 83.5%
-Process productivity: 0.25 g/L/h
-Mono-Glc: 3.4 g; 85.5%
-Mono-Xyl: 19.2 g; 63.5%
-Oligo-Glc: 16.3 g; 4.3%
-Oligo-Xyl: 73.4 g; 27.6%
-Poly-Glc: 38.4 g
-Poly-Xyl: 23.7 g
-Ethanol: 195.1 g
-Ethanol metabolic yield: 81.3%
-Process productivity: 0.20 g/L/h

Figure 40 Mass balance of continuous SSCF, batch SSCF(A) and SHF(B).
For SHF (72 h EH+96h fermentation), the fermentation kinetics were assumed to be the same
as those of 168 h hydrolysate fermentation.
118

Figure 40 (cont’d)

B)
1.0 kg AFEX-CS
-Glcose: 378.9 g
-Xylose: 265.8 g

Enzyme protein loading:
Accellerase 1500: 8.184g
Accellerase XY: 2.046g
Mutifect Pectinase: 2.046g

Enzymatic
Hydrolysis
(72h)

Yeast cells:
4.91g

Enzymatic
Hydrolysis
(168h)
-Mono-Glc: 328.9 g; 86.8%
-Mono-Xyl: 191.9 g; 72.2%
-Oligo-Glc: 29.6 g; 7.8%
-Oligo-Xyl: 61.7 g; 23.2%
-Poly-Glc: 20.7g
-Poly-Xyl: 12.3 g

-Mono-Glc: 299.0 g; 78.9%
-Mono-Xyl: 175.4 g; 66.0%
-Oligo-Glc: 33.0 g; 8.7%
-Oligo-Xyl: 69.4 g; 26.1%
-Poly-Glc: 46.8 g
-Poly-Xyl: 21.1g

Fermentation
(96h)

Fermentation
(168h)

-Mono-Glc: 2.2 g; 78.9%
-Mono-Xyl: 50.7 g; 66.0%
-Oligo-Glc: 33.0 g; 8.7%
-Oligo-Xyl: 69.4 g; 26.1%
-Poly-Glc: 46.8 g
-Poly-Xyl: 21.1g
-Ethanol: 163.6 g
-Ethanol metabolic yield: 76.1%
-Process productivity: 0.20 g/L/h

Yeast cells:
4.91g

-Mono-Glc: 0.0 g; 86.8%
-Mono-Xyl: 32.9 g; 72.2%
-Oligo-Glc: 29.6 g; 7.8%
-Oligo-Xyl: 61.7 g; 23.2%
-Poly-Glc: 20.7g
-Poly-Xyl: 12.3 g
-Ethanol: 194.1 g
-Ethanol metabolic yield: 78.0%
-Process productivity: 0.16 g/L/h

TM

continuous SSCF (flow rate 12 ml/h) 187.5 g ethanol/kg AFEX

-CS. However, the
-1 -1

corresponding process ethanol productivities were low at 0.20, 0.16, and 0.28 g·L ·h ,
119

respectively. The highest process ethanol productivities were found in continuous SSCF with
-1 -1

flow rates of 48 and 24 ml/h, and were 0.65 and 0.46 g·L ·h , respectively. To more fairly
compare the three processes , the process ethanol productivity was compared based on similar
sugar to ethanol conversions (glucose and xylose-to-ethanol conversions and ethanol
concentration were 78.3%, 46.9% and 36.0 g/L, respectively for SHF (72 h EH +96 h ferm.);
80.3%, 48.2% and 36.9 g/L, respectively for batch SSCF (24 h pre-hydrolysis + 120 h SSCF);
77.0%, 47.5% and 36.5 g/L, respectively for continuous SSCF with 24 ml/h flow rate) in Fig.
39b. Under such a comparison, continuous SSCF also showed the highest process ethanol
-1 -1

-1 -1

productivity 0.46 g·L ·h , followed by batch SSCF 0.25 g·L ·h (Fig. 39b) and SHF 0.20
-1 -1

g·L ·h . This means that to achieve the same sugar-to-ethanol conversion, continuous SSCF is
a more efficient process. Ethanol productivity increased by around 2 times, therefore the capital
cost for enzymatic hydrolysis and fermentation was reduced by around half. Besides, yeast
inoculation for continuous SSCF is required only once and only in one bioreactor, which would
also reduce the processing cost. For batch SSCF and SHF, however, yeast inoculation is needed
TM

for every batch. For example, the yeast cell inoculation required is 0.24 g per kg AFEX

-CS or

TM

even per ton AFEX

-CS based on continuous SSCF. However, for batch SSCF or SHF, 4.91 g
TM

yeast inoculum is required per 1 kg AFEX

-CS (Fig. 39).

6.3.6 Discussion on the continuous SSCF process design

𝑉=−

𝐹 𝑖𝑜

𝑋𝑖

The bioreactors configuration design in the continuous SSCF process was based on the
CSTR design equation:

𝑟𝑖

(V is the reactor volume (L); Fio is the mass flow in

rate (g/h); i could be glucose or xylose; ri is the conversion rate (g/L/h); Xi is the sugar to ethanol
120

conversion. Fio equals the volumetric flow rate (L/h) multiplied by the initial sugar concentration.
Levenspiel plots (-Fio/ri ~ Xi) were used for reactor sizing (Fogler, 1999). The correlations
between -Fio/ri and Xi were calculated based on experimental data. When the volumetric flow
rate is 12 ml/h, the plots are shown in Fig. 41. The shade areas are the designed reactor volumes.
The designated glucose conversion for the first SSCF reactor (Reactor B) is around 68%, at
which point the curve has a value close to minimum (Fig. 41a), corresponding to a reactor

𝑉=∫ −
0
𝑋

𝐹 𝑖𝑜

𝑑𝑋 𝑖 ). A series of CSTRs function like a plug flow reactor. Thus, three

volume of 250 ml. After that point, a plug flow reactor (PFR) would be most efficient (PFR:
𝑟𝑖

reactors (C, D, E) were used for achieving high volume efficiency. The obtained experiments
results were similar to the design expectations. For instance, the Reactor B achieved 73.6%
glucose-to ethanol conversion and Reactor E reached 79.2% (design expectation is 68% and 79%,
respectively). The xylose-to-ethanol conversions for Reactor B and Reactor E were 20.9% and
50.8%, respectively (design expectation is around 16% and 49%, respectively).
An SSCF model, which could describe the reactions better, might be helpful in further
optimization of the continuous SSCF process. However, even if the process is further optimized
a large reaction volume is still required due to the slow xylose fermentation (Fig. 41b) and a high
enzyme loading is needed, which make the economics of this process not that promising based
on current enzymes and fermenting strain. However, this process is definitely superior to batch
processes. Better enzymes and fermenting strains are needed for better reactions for this process.

121

a) 25
Reactor E volume
400 ml

-FGlc.o/rglc

20

15
Reactor D volume
400 ml
10

5

0
0.0

Reactor C volume
250 ml
Reactor B volume
250 ml

0.1

0.2

0.3
0.4
0.5
0.6
0.7
Glc-to-EtOH conversion (Xglc)

0.8

0.9

b)
10
9

-FXyl.o/rxyl

8

Reactor E volume
400 ml

7
6
5
4
3

Reactor D volume
400 ml
Reactor C volume
250 ml
Reactor B volume
250 ml

2
1
0.0

0.1
0.2
0.3
0.4
Xyl-to-EtOH conversion (Xxyl)

0.5

0.6

Figure 41 Levenspiel plots of –Fglc.o/rglc against Xglc (a) and –Fxyl.o/rxyl against Xxyl (b).
F is the mass flow in rate (g/h); r is the sugar to ethanol conversion rate (g/L/h); X is the sugar
to ethanol conversion.

122

6.4 Conclusions
Enzymatic hydrolysis performed better in the SHF process when compared to SSCF based
on xylose-fermenting strain S. cerevisiae 424A(LNH-ST) and commercial enzymes (including
Accellerase 1500, Accellerase XY and Multifect pectinase). However, SSCF was more efficient
for ethanol production. Based on similar sugar to ethanol conversions (around 80% glucose-toethanol conversion and 47% xylose-to-ethanol conversion), the process ethanol productivity of
continuous SSCF was 2.3 and 1.8 fold higher than that of SHF and batch SSCF, respectively.
Increasing flow rates enhanced ethanol productivity, but decreased overall sugar conversions of
continuous SSCF.

123

CHAPTER VII NOVEL INTEGRATED BIOLOGICAL PROCESSES
FEATURING HIGH ETHANOL PRODUCTIVITY, ENZYME
RECYCLING AND YEAST CELLS REUSE

Abstract
High enzyme loading requirements, slow xylose fermentation and low ethanol productivity
are three of the major issues impeding commercial biochemical production of cellulosic ethanol.
A novel integrated biological process was developed in this study to overcome these problems.
Enzymatic hydrolysis was performed for only 24 h to avoid the slow rate period which begins at
about that time. Unhydrolyzed recalcitrant solids with adsorbed enzymes were recycled to the
subsequent cycles. By this approach, easily digestible biomass was processed first and
recalcitrant biomass was given enough residence time to get hydrolyzed during subsequent
processing steps. Fermentation was conducted using a high yeast inoculation level and was also
completed in 24 h. The yeast cells were then recycled. With this novel processing approach, the
enzyme loading was reduced from 36 to 22.3 and 25.8 mg protein per gram glucan, respectively,
for separate hydrolysis and fermentation (SHF) and for simultaneous saccharification and coTM

fermentation (SSCF) on AFEX

pretreated corn stover. The process ethanol productivity was

enhanced by 2 to 3 fold due to both fast enzymatic hydrolysis and fast fermentation.

124

7.1 Introduction
Cellulosic ethanol production from lignocellulosic biomass has gained considerable
momentum due to both environmental and social sustainability benefits (Farrell et al., 2006;
Greene & Council, 2004; Wyman, 2008). However, the technology is not yet fully
commercialized. There are major issues impeding the cellulosic ethanol production using the
sugar platform, including high enzyme loadings needed, slow xylose fermentation and low
ethanol productivity.
The enzyme loadings typically used for high solids loading enzymatic hydrolysis were
around 30 to 36 mg protein per gram glucan (previous chapters). According to the National
Renewable Energy Laboratory’s (NREL, Golden Colorado) 2011 report, the enzyme cost
accounts for 15.7% of the total cost even if the enzyme loading used is 20 mg per gram glucan
(Humbird et al., 2011). The high enzyme loadings required for the process to obtain high sugar
yields are largely due to the reduced rate of enzymatic hydrolysis with increased conversion/time
(Yang et al., 2006) (Chapter VI, Fig. 35). Cellulose and hemicelluloses in plant cell wall exist in
complex structures within a recalcitrant lignin matrix, and are difficult to hydrolyze (Chundawat
et al., 2011a). Pretreatment disrupts such structures and improves the digestibility of
lignocellulosic biomass (Yang & Wyman, 2008). After pretreatment, much of the heterogenous
biomass is easily digestible while there is a considerable amount that is still recalcitrant (Kumar
& Wyman, 2009a; Zhang et al., 1999; Zhou et al., 2010). Therefore, high rates are typically seen
at the beginning of the enzymatic hydrolysis due to the easily digestible part being hydrolyzed
before the recalcitrant part. Concerning the hydrolysis of the recalcitrant parts of biomass, there
are several factors that likely further reduce the hydrolysis rates, including product inhibition

125

(Eriksson et al., 2002b), enzyme inactivation (Gusakov & Sinitsyn, 1992) and thermal instability
of cellulases (Eriksson et al., 2002b). Reduced hydrolysis rates not only lead to high enzyme
loadings but also result in long enzymatic hydrolysis times. Based on the current separate
hydrolysis and fermentation (SHF) process, it takes 72~168 h to achieve high sugar conversions
at high solids loading and another 96-168 h are required to ferment those sugars (mostly glucose
and xylose) to ethanol since the xylose fermentation is slow. Thus overall biological processing
times are well over 5 days, leading to higher capital and operating costs.
Xylose can be co-fermented with glucose to ethanol by genetically modified strains such as
Escherichia coli KO11 (Olsson et al., 1995), Zymomonas mobilis AX101 (Zhang et al., 1995)
and Saccharomyces cerevisiae 424A(LNH-ST) (Sedlak & Ho, 2004). Lau et al. compared the
TM

performance of those strains in AFEX

(ammonia fiber expansion) pretreated hydrolysate and
TM

concluded that S. cerevisiae 424A(LNH-ST) was a good strain for AFEX

hydrolysate

fermentation (Lau et al., 2010). However, S. cerevisiae 424A(LNH-ST) fermented xylose slowly
TM

in AFEX

hydrolysates. Slow xylose fermentation by this yeast is caused by redox imbalance

(Ha et al., 2011), no specialized xylose transporter (Bertilsson et al., 2008), degradation products
inhibition, ethanol inhibition as well as fermentation metabolites inhibition (Chapter II).
Degradation products mostly inhibit the yeast cell growth during fermentation and hence reduce
the overall xylose fermentation rate. Due to the slow xylose consumption, the fermentation
ethanol productivity (ethanol produced per unit of time) was much lower than the industrial
-1 -1

criterion of about 1.0 g·L ·h

(Dien et al., 2003a). If the ethanol productivity is calculated

based on enzymatic hydrolysis time plus fermentation time (process ethanol productivity), the
productivity of current SHF process would be even lower due to the long enzymatic hydrolysis

126

time. By integrating enzymatic hydrolysis and fermentation, simultaneous saccharification and
co-fermentation (SSCF), especially continuous SSCF, reduces the time required for enzymatic
hydrolysis and hence improves productivity (Chapter III,V&VI). However, the high enzyme
loading required is still a major concern.
Here a novel integrated biological process was invented (BCRL process; BCRL: Biomass
Conversion Research Laboratory) to reduce the enzyme loading, enhance the xylose
fermentation and increase the ethanol productivity. This process can be performed either in a
TM

SHF mode or in a SSCF mode (Fig. 42) and was tested on AFEX

pretreated corn stover

TM

(AFEX

-CS) using commercial enzymes and S.cerevisiae 424A(LNH-ST). In this process, we

hydrolyzed the easily digestible biomass first and avoided the slow-down period by performing
enzymatic hydrolysis only for 24 h. The remaining unhydrolysed solids with some enzymes
adsorbed (Azevedo et al., 2002; Gao et al., 2011; Tu et al., 2007) were recycled to the next cycle
for further hydrolysis. By this approach, part of the enzymes was recycled and thereby the
required enzyme loading was reduced. To achieve fast xylose fermentation, we used a high
inoculum level (OD600=20 instead of 0.5~2) and completed fermentation in 24 h. The high
inoculum level fermentation bypassed the degradation products inhibition on yeast cell growth
and hence improved xylose fermentation. The yeast cells were subsequently recycled. It took 24
h for enzymatic hydrolysis and another 24 h for fermentation to complete one cycle. Then the
liquid part of the hydrolysis/fermentation mixture was harvested and subjected to distillation.
The unhydrolyzed solids (recalcitrant part) were repeatedly transferred to the next
hydrolysis/fermentation cycle to allow more hydrolysis time. After five cycles of

127

hydrolysis/fermentation, the accumulated recalcitrant residual solids were further hydrolyzed in
the last step. Since some enzymes were recycled by this approach and part of the inhibitory
Yeast cells

a)

Cycle 1.
AFEXTM-CS

Enzymatic
hydrolysis
(250rpm,50oC,
pH 4.8,24h)

Water
Enzymes
Cycle 2.
AFEXTM-CS

Fermentation
(iOD=20,150rpm,32oC
pH 5.5, 24h)

Reduced
enzyme loading
Cycle 3.

Enzymatic
hydrolysis
(250rpm,50oC,
pH 4.8,24h)

Centrifuge
/ filtration

Supernatant
(hydrolysate)

Fermentation
(150rpm,32oC
pH5.5, 24h)

Yeast cells

n cycles later (n=5 for the present study)
Solid Residue (Enzymes)
Enzymatic
hydrolysis
(250rpm,50oC,
pH 4.8,24h)

Centrifuge
/ filtration

Ethanol

Yeast cells

Solid Residue (Enzymes)

Cycle 4.
Cycle 5.

Ethanol

Yeast cells

Solid Residue (Enzymes)

Water

Last step
Water

Supernatant
Centrifuge (hydrolysate)
/ filtration

Supernatant
(hydrolysate)

Fermentation
(150rpm, 32oC
pH5.5, 24h)

Ethanol

Solid residue

b)
Water

Enzymes

Yeast cells

Cycle 1.
AFEXTM-CS

Enzymatic
hydrolysis
(250rpm,50oC,
pH 4.8,24h)
Water

Enzymes

Cycle 2.
AFEXTM-CS

Cycle 3.

Enzymatic
hydrolysis
(250rpm,50oC,
pH 4.8,24h)

Centrifuge
/ filtration

Supernatant
(Ethanol)

Solid Residue (Enzymes & yeast)
Supernatant
SSCF
Centrifuge (Ethanol)
(180rpm,32oC,pH5.5
/ filtration
24h)
Solid Residue (Enzymes & yeast)

Cycle 4.
Cycle 5.

Last step

SSCF
(iOD=20,180rpm,
32oC,pH5.5,24h)

n cycles later (n=5 for the present study)
Solid Residue (Enzymes & yeast)
Water

SSCF
(180rpm,32oC,pH5.5
72h)

Centrifuge
/ filtration

Supernatant
(Ethanol)

Solid
residue

Figure 42 Flow chart of the Biomass Conversion Research Laboratory (BCRL) SHF process (a)
and SSCF process (b).
The figure is in color.
128

degradation products were removed with liquid during the five cycles, the hydrolysis of the
residual solids was both more rapid and more complete (higher sugar conversions). No enzymes
were supplemented during the last step of enzymatic hydrolysis.
7.2 Materials and Methods
TM

7.2.1 AFEX

pretreated corn stover

Please see Chapter II section 2.2.1
7.2.2 Microorganisms and seed culture preparation
Please see Chapter IV section 4.2.2.
7.2.3 BCRL SHF and SSCF processes
The BCRL SHF process performed enzymatic hydrolysis for 24 h, followed by
centrifugation at 5300 rpm for 20 min (Fig. 42a). The resulting unhydrolyzed solids were
recycled to the next enzymatic hydrolysis tank and thereby the working volume increased from
cycle to cycle (Fig. 43). The supernatant (hydrolysate) was used for ethanol fermentation at 150
o

rpm, 32 C and pH 5.5 with an initial OD600 of 20 for 24 h. After fermentation, the yeast cells
were collected by centrifugation at 4400 rpm for 8 min and used for the next fermentation
inoculation. As shown in Fig. 42a, solid biomass residues and yeast cells were continually
recycled to the next round of enzymatic hydrolysis and fermentation, respectively. For enzymatic
hydrolysis in the last step, 5 ml phosphate buffer and 55 ml water were added without any new
enzymes or fresh biomass. Enzymatic hydrolysis for cycles 1, 2, and 3 was conducted in a 250
ml baffled flask while cycles 4, 5 and the last step were performed in a 500 ml baffled flask.
The BCRL SSCF process was performed using enzymatic hydrolysis for 24 h followed by
o

changing conditions to pH 5.5, 32 C, and 180 rpm, and inoculating yeast cells (initial OD = 20)

129

(Fig. 42b). After 24 h of SSCF, centrifugation was carried out at 5300 rpm for 20 min. The
supernatant was ready for distillation. The solids were recycled and transferred to the next cycle
SSCF tank and thereby the working volume increased from cycle to cycle (Fig. 43). After 5
cycles, 5 ml phosphate buffer and 55 ml water were added to the solids with no enzymes and
o

fresh biomass supplemented for the last step. SSCF were conducted at 32 C, pH 5.5 and 180
rpm for 72 h.
Mass flow every 24 h

BCRL SHF
Enzymes
Biomass,
Water
EH

Liquid
Hyl.

BCRL SSCF

Enzymes
Biomass,
Liquid
Water
Ferm. broth
Ferm.
EH

Whole
Hyl.

Solids
buildup

Solids
buildup

BCRL SHF-Ferm. (ml)

250
200
150
100
50
0

Volume (ml)

Volume (ml)

BCRL SHF-EH (ml)

1

2

3

4

5

250
200
150
100
50
0

6

1

2

Cycle

2

3
4
Cycle

3
4
Cycle

5

6

BCRL SSCF-SSCF (ml)
Volume (ml)

Volume (ml)

BCRL SSCF-EH (ml)
100
80
60
40
20
0
1

Liquid

SSCF Ferm. broth

5

6

250
200
150
100
50
0
1

2

3
4
Cycle

5

6

Figure 43 Tank setup and volume change for BCRL SHF and SSCF processes.
The last step is denoted as cycle 6. Basically, the volume increase happened in the enzymatic
hydrolysis (EH) tank for the BCRL SHF process and in the SSCF tank for BCRL SSCF process
due to the buildup of solid residues. The last step of fermentation for BCRL SHF process was
carried out in the EH tank after 24 h hydrolysis without solid-liquid separation.
130

For enzymatic hydrolysis (EH) in both processes, the inputs for each cycle including fresh
TM

AFEX

-CS, water, phosphate buffer, and enzymes were 100 g in total. Glucan loadings for

BCRL SHF and SSCF processes were 7.5% and 7% (w/w), respectively, which correspond to
22.0 g and 20.5 g AFEX

TM

-CS biomass in 100 g total mixture, respectively. Enzymes used for

the hydrolysis included Accellerase 1500, Accellerase XY, and Multifect pectinase (Genencor
Inc, USA) with protein loadings of 30, 3, and 3 mg/g glucan, respectively, in the first cycle. The
o

protein loadings varied during cycles 2~5. Enzymatic hydrolysis was performed at 50 C, pH 4.8,
and 250 rpm. Fifty mg/L chloramphenicol (Cm) was used during hydrolysis to avoid microbial
contamination.
7.2.4 Conventional SHF and SSCF processes
o

The conventional SHF process performed enzymatic hydrolysis at 50 C, pH 4.8 and 250
o

rpm for 96 h. The resulting hydrolysate was used for fermentation at 30 C, pH 5.5 and 150 rpm
with an initial OD of 2.0. The conventional SSCF process was carried out by pre-hydrolyzing
o

o

biomass at 50 C, pH 4.8 and 250 rpm for 6 h. The conditions were then changed to 30 C, pH
5.5 and 180 rpm with yeast cells inoculated at an initial OD of 2.0. The enzyme loading for the
conventional processes was the same as for the first cycle of the BCRL processes. All the
experiments were conducted in duplicate with the average and standard deviation shown in
figures.
7.2.5 Measurement of viable cell density
Please see Chapter II section 2.2.8.

131

7.2.6 HPLC analyses and mass balance
For HPLC analyses, please see Chapter II section 2.2.7.
For mass balance details, please see Chapter III section 3.2.8.
7.3 Results and discussion
7.3.1 Effect of solids loading
The effect of solids loading (6%, 7% and 9% glucan loading) was investigated by
performing 24 h enzymatic hydrolysis followed by 24 h hydrolysate fermentation or SSCF (Fig.
44a). One objective was to produce ethanol at an industrially relevant titer of 40 g/L (Dien et al.,
2003a). The ethanol titer of 40 g/L was achieved by a 9% glucan loading for both SSCF and
hydrolysate fermentation. However, the remaining xylose concentration in the fermentation broth
was almost 2 times higher than 7% glucan loading, which means a greater loss of fermentable
sugars if 9% glucan loading is applied since the broth will be removed and subject to distillation
after 24 h fermentation (Fig. 42). Moreover, the cell viability during SSCF for 9% glucan loading
was lower compared to the other solids loadings (Fig. 44b). This implies that the yeast cells were
dying more rapidly at 9% glucan loading and thus fewer viable yeast cells could be recycled for
the subsequent cycles. A glucan loading of 7% produced 38.6 g/L ethanol with 6.0 g/L xylose
remaining in the broth after 24 h hydrolysis and 24 h SSCF. Similar studies using separate
hydrolysis and fermentation (SHF) mode yielded 38.1 g/L ethanol with 4.8 g/L xylose remaining
unutilized. By converting 4 g/L more xylose to ethanol, 40 g/L ethanol could be achieved.
Therefore, the fermentation conditions were optimized.

132

a) 70

Glucose

Xylose

Ethanol

Concentration (g/L)

60
50
40
30
20
10
0

24h 24h 24h 24h 24h 24h 24h 24h 24h
EH SSCF Ferm. EH SSCF Ferm. EH SSCF Ferm.
6%
9%
7%
glucan loading
glucan loading
glucan loading

7x108
8

6x10

0h

Cell viability

24h

105
90

8

5x10

75

8

4x10

60

8

3x10

45

8

2x10

30

1x108

Cell viablility (%)

Viable cell density (CFU/ml)

b)

15

0

6%

7%
Glucan loading

9%

0

Figure 44 Effect of solids loading on 24 h enzymatic hydrolysis, the following 24 h SSCF or
hydrolysate fermentation (a), viable cell density and cell viability after 24 h SSCF (b).
Hydrolysate fermentation (fermentation of the liquid hydrolysate) and SSCF were conducted in
250 ml unbaffled and baffled flasks, respectively, at 30 oC, pH 5.5, OD600=20, and 150 rpm.
Viable cell density was measured in colony forming unit (CFU) per ml. The cell viability was
calculated by using the viable cell density after 24 h SSCF dividing the initial viable cell density.
The figure is in color.

133

Concentration (g/L)

a)

40
35
30

Xyl.OD15
Xyl.OD20
Xyl.OD25

25
20

EtOH.OD15
EtOH.OD20
EtOH.OD25

15
10
5
0

0

5

Concentration (g/L)

b) 40

10
15
20
SSCF time (h)

25

30

35
30

Xyl.pH5.0
Xyl.pH5.5
Xyl.pH6.0

25
20

EtOH.pH5.0
EtOH.pH5.5
EtOH.pH6.0

10
15
20
SSCF time (h)

25

15
10
5
0

0

5

c)

30

Concentration (g/L)

40
35
30

Xyl.30oC
Xyl.32oC
Xyl.35oC

25
20

EtOH.30oC
EtOH.32oC
EtOH.35oC

15
10
5
0

0

5

10
15
20
SSCF time (h)

25

30

Figure 45 Effect of initial OD (a), pH (b) and temperature (c) on xylose consumption, ethanol
production and cell viability (d) during the first cycle 24 h SSCF at 7% (w/w) glucan loading.
The default conditions were OD=20, pH 5.5 and temperature 30 oC. The figure is in color.
134

Figure 45 (cont’d)

Cell viability (%)

d)
100
80
60
40
20
0
15 20 25
OD600

70

Concentration (g/L)

60

5.0 5.5 6.0
pH

Mono-Glc.
Oligo-Glc.

30 32 35
Temp.(oC)

Mono-Xyl.
Oligo-Xyl.

50
40
30
20
10
0
0% 30% 40% 50% 60% 100%
Cycle1Cycle2 Enzy. loading (100%=36mg/g glucan)
24hEH 0h
Cycle2 24hEH

Figure 46 Optimization of enzyme loading for cycle 2 during the BCRL SHF process.
The enzyme loading for cycle 1 is 36 mg/g glucan (100%). Enzymes ratio is fixed at Accellerase
1500 (cellulases): Accellerase XY (xylanase): Mutifect pectinase=10:1:1. The figure is in color.

135

7.3.2 Optimization of fermentation conditions
The initial OD (yeast cell density), pH and temperature were optimized for the process of
24 h enzymatic hydrolysis followed by 24 h SSCF at 7% glucan loading. From the figure (Fig.
45a) it is clear that by increasing the OD more xylose was consumed and more ethanol was
produced. However, the highest OD tested (OD=25) showed lower cell viability (Fig. 45d). The
fact that this yeast strain cannot grow well on xylose anaerobically (probably due to redox
imbalance (Kuyper et al., 2004)) and that the glucose release rate during SSCF might not support
a high enough cell growth rate to balance the cell death rate could explain this result. pH 6.0 was
the best for xylose fermentation (Fig. 45b). Nevertheless, the cell viability was not as good at this
pH as it was at pH 5.5 (Fig. 45d) and it was also far removed from the optimum enzymatic
o

hydrolysis pH 4.8, which would also affect the overall sugar conversion. A temperature of 32 C
produced more than 40 g/L ethanol with reasonable cell viability. The higher ethanol yield might
be due to higher sugar conversions at higher temperature. However, the ethanol metabolic yield
o

is lower at 35 C as is the cell viability (Fig. 45d), which might explain the lower ethanol yield at
o

o

35 C (Fig. 45c). Therefore, an initial OD of 20, pH 5.5 and a temperature of 32 C were chosen
as the optimum conditions for BCRL SSCF. The same conditions were used for the BCRL SHF
process.
7.3.3 Optimization of enzyme loading for cycle 2 of BCRL SHF process
To make sure 40 g/L ethanol was produced, we applied a 7.5% glucan loading for the
BCRL SHF process. Here, unhydrolyzed solids along with adsorbed enzymes were recycled for
subsequent cycles of hydrolysis. By this approach we reduced the enzyme loadings for cycles
2~5. Experiments were carried out to optimize the enzyme loading for cycle 2 of BCRL SHF

136

process to obtain similar sugar concentrations as cycle 1 (Fig. 46). Without adding additional
enzymes, 38.7 g/L glucose and 20.2 g/L xylose were obtained after 24 h enzymatic hydrolysis
during cycle 2, which means the recycled enzymes from cycle 1 were functioning well. About 60%
of the cycle 1’s enzyme loading applied in cycle 2 yielded similar sugar concentrations as were
obtained in cycle 1. Therefore, this enzyme loading (21.6 mg enzyme protein per glucan) was
used for cycle 2, resulting in a 40% enzyme saving. Since the enzymes were being recycled and
were likely accumulating, we used 50% of the cycle 1 enzyme loading (18 mg enzyme protein
per glucan) for cycles 3 to 5.
7.3.4 BCRL SHF process results
By applying the enzyme loadings discussed above and a glucan loading of 7.5%,
consistently high sugar concentrations were produced for cycles 1 to 5 (Fig. 47a). The last step
produced lower sugar concentrations compared to other cycles due to no addition of fresh
enzymes or fresh biomass. It is likely that 24 h was too short a time for the hydrolysis of the final
residual solids, and that these solids were much more difficult to hydrolyze (Zhang et al., 1999).
However, the sugar concentrations obtained were still reasonable. The removal of degradation
products with liquid in previous cycles and lower sugar concentrations present at the beginning
of the last step helped achieve reasonable sugar conversions from the most recalcitrant part of
biomass. Ethanol produced in each cycle was always approximately 40 g/L and the OD was
increasing from cycle to cycle (Fig. 47b), which guaranteed fast fermentation of xylose (Fig.
47c).The ethanol metabolic yield was maintained at around 90% (Fig. 47c).Though this is an
SHF process, the enzymes (mostly β−glucosidase and β−xylosidase) present in the hydrolysate
continued to function. Hence part of the oligmeric sugars was converted to monomeric sugars
during the fermentation process and then fermented to ethanol (Fig. 47c).The overall glucan

137

b)

Concentration (g/L)

70

Mono-Xyl.
Oligo-Xyl.

Mono-Glc.
Oligo-Glc.

60
50
40
30
20
10
0

1

2

4
3
Cycle
c)

Concentration (g/L) & OD600

a)

70
50
30
20
10

5 last step

0
1

2

4
3
Cycle

5 last step

100
90
80
70
60
50
40
30
20
10
0

EtOH metabolic yield (%)

60

Consumed sugar
concentration (g/L)

Mono-Xyl.
Oligo-Xyl.
OD600

40

70
50
40
30
20
10
0

Mono-Glc.
Oligo-Glc.
EtOH

60

1

2

Mono-Glc.
Oligo-Xyl

3
4
5 last step
Cycle
Mono-Xyl.
Oligo-Glc.
EtOH metabolic yield

Figure 47 BCRL SHF process results.
a: Sugar concentrations after 24 h enzymatic hydrolysis; b: fermentation performance; c: sugar
consumption and ethanol metabolic yield during fermentations; d: sugar conversions during
enzymatic hydrolysis, oligomeric to monomeric sugar conversions during fermentation and
overall sugar to ethanol conversions. The enzyme loading for cycle 1 is 36 mg/g glucan (100%).
Enzyme loadings used for cycle 2, 3, 4, 5 were 60%, 50%, 50%, and 50% of the cycle 1
respectively. The average enzyme loading was 22.3 mg/ g glucan. All the conversions were
calculated based on total input glucan and xylan as glucose or xylose equivalents. During
fermentation, some oligomeric glucose/xylose was converted to monomeric glucose/xylose
(Mono-Glc./Mono-Xyl. conversion). The Glc.-EtOH/Xyl.-EtOH conversion was calculated based
on total fermented glucose/xylose. The figure is in color.

138

Figure 47 (cont’d)

Sugar conversion (%)

d) 100

90
80
70
60
50
40
30
20
10
0
Enzymatic hydrolysis
Mono-Glc.
Mono-Xyl.
Total-Glc.
Total-Xyl.

Fermentation
Oligo-Glc.
Oligo-Xyl.
Glc.-EtOH
Xyl.-EtOH

conversion and xylan conversion were 79.4% and 86.3%, respectively for this process (Fig. 47d).
However, the glucose to ethanol and xylose to ethanol conversions were only 68.3% and 37.7%,
respectively. The major sugar loss was due to the oligomeric sugars (Fig. 49a), especially xylose
oligomers (Fig. 47b), which cannot be consumed by yeast strains. Improved enzymes or
engineered yeasts able to consume oligomers would help resolve this problem.
The enzymes used in this study were commercial enzyme cocktails rather than the
optimum defined enzyme mixtures used for AFEX

TM

-CS (Gao et al., 2010) to generate highest

concentrations of monomeric sugars, especially xylose. (Our laboratory does not have the
capacity to produce enough purified enzymes for work such as that reported here.) Large
TM

amounts of xylose oligmers always exist in AFEX

hydrolysates at high solid loadings. In

addition, we kept the enzyme loading ratio (cellulases: xylanases: pectinases) the same for each
139

cycle of the process. However, by analyzing the enzymes in the liquid hydrolysate for each cycle,
we found that more β-glucosidase and xylanase were removed or denatured compared to
cellulases (Fig. 50). This means that more xylanase and β-glucosidase should have been
supplemented during late cycles and therefore the enzyme loading can probably be further
reduced. Optimization of the enzyme cocktail for each cycle of this process needs to be
performed in order to determine the overall potential for enzyme saving using the BCRL process.
7.3.5 BCRL SSCF process results
Since the BCRL SSCF process recycles solids and enzymes to the next cycle of the SSCF
tank, the enzyme loading can be reduced during later cycles. However, as the strain cannot
tolerate high temperatures and the cells were recycled along with the enzymes, the recycled
o

enzymes must work at the fermentation temperature (32 C). Hence, the enzyme reduction
achieved in the BCRL SSCF process is less than it is for the BCRL SHF process. We tested three
enzyme loading profiles (Fig 6): Enzyme loading a: 36 mg/g glucan for each cycle; Enzyme
loading b: 36.0 (100%), 32.4 (90%), 30.6 (85%), 28.8 (80%) and 27.0 (75%) mg/g glucan for
cycles 1~5, respectively, with an average enzyme loading of 30.9 mg/g glucan; Enzyme loading
c: 30.0 (100%), 27.0 (90%), 25.5 (85%), 24.0 (80%) and 22.5 (70%) mg/g glucan for cycles 1~5,
respectively, with an average enzyme loading of 25.8 mg/g glucan. The ratio of enzymes
cocktails applied (cellulases:xylanase:pectinase) was fixed at 10:1:1. The ethanol concentration
for each cycle largely met the criterion of 40 g/L for all three cases (Fig. 48a). Glucose to ethanol
conversions were around 75% to 80% (Fig. 48b). However, the xylose to ethanol conversions
were quite low, around 40% to 50%. Most of the xylose was in oligomeric form and was not
available for fermentation (Fig. 49).

140

One challenge in this process system is the viable cell density. Decreasing viable cell densities
were observed from cycle to cycle (Fig. 48c), which caused a buildup of xylose (xylose loss as
monomers) during later cycles (Fig. 48a & Fig. 49).The solid residue concentrations in cycles 1
to 6 of the BCRL SSCF process were around 88, 125, 153, 170, 185, and 185 g/L, respectively.
One might suspect that high solids concentration could lead to loss of viable cells. However, the
solids had no effect on hydrolysate fermentation (Fig. 51).
7.3.6 Comparisons of process ethanol productivity and ethanol yield of different process
The BCRL SHF and SSCF processes showed much higher process ethanol productivities
compared to conventional SHF or SSCF processes (Fig. 52a) probably due to the time saved (Fig.
52c) by fast hydrolysis and fast fermentation. For the ethanol yield, 7% glucan loading was less
effective than a 6% glucan loading using conventional processes (Fig. 52b). Decreased ethanol
yield with increasing solids loading has been shown and discussed in Chapter V. The BCRL SHF
process at 7.5% glucan loading showed similar ethanol yields compared to conventional
processes at similar glucan loadings but with much lower enzyme loading (22.3 vs. 36 mg
protein/ g glucan). The BCRL SSCF process showed higher or similar ethanol yields (depending
on enzyme loading) at 7% glucan loading compared to conventional fermentation processes at 7%
or 6% glucan loading.
Regarding enzyme reduction, it appears that the BCRL SHF process has greater potential
than the BCRL SSCF process since the recycled enzymes can perform at optimal conditions,
while the recycled enzymes for the BCRL SSCF process cannot perform to their potential due to
the temperature limitations of the strain. If a thermo-tolerant strain is used, the full advantage of
the BCRL SSCF process can perhaps be realized. The enzymes recycling depended on the
enzyme adsorption to the residual solids. But high sugar concentrations in the fast SHF process

141

a)

Cycle 1

Cycle 2

Cycle 3

Cycle 4

182 ml Solids
90 ml Solids 122 mlSolids 152 ml Solids

EtOH conc.
(g/L)
Xylose conc.
(g/L)
EtOH conc.
(g/L)
Xylose conc.
(g/L)
EtOH conc.
(g/L)
Xylose conc.
(g/L)

41.9
5.4
42.3
5.7
40.3
4.6

45.1

42.6
6.3

43.5

40.1
5.5

Last step

210 ml

210 ml

Solids

46.3

47.2

9.2

13.6

44.3

45.2

7.0

9.0

14.2

38.1 Enzyme
10.5 loading b

39.9
8.3

39.6
10.7

37.8
15.0

36.4 Enzyme
9.4 loading c

7.9

40.5 Enzyme
9.7 loading a

Glc.to ethanol conversion
Xyl. to ethanol converion

Sugar to ethanol
conversion (%)

b) 100
90
80
70
60
50
40
30
20
10
0

43.3
5.8

Cycle 5

Enzyme loading Enzyme loading Enzyme loading
a
b
c
Figure 48 Effect of enzyme loading profile on ethanol production (a), sugar to ethanol
conversions (b), and viable cell density (c) during BCRL SSCF process.
Enzyme loading a: 36 mg/g glucan for each cycle; Enzyme loading b: 36.0 (100%), 32.4
(90%), 30.6 (85%), 28.8 (80%) and 27.0 (75%) mg/g glucan for cycle 1~5, respectively with
average enzyme loading 30.9 mg/g glucan; Enzyme loading c: 30.0 (100%), 27.0 (90%), 25.5
(85%), 24.0 (80%) and 22.5 (70%) mg/g glucan for cycle 1 to 5, respectively with average
enzyme loading 25.8 mg/g glucan. The ratio for enzymes (cellulases:xylanase:pectinase) was
fixed at 10:1:1. The figure is in color.

142

Figure 48 (cont’d)
c)

CFU/ml

Enzy.loading a
5.6x108 Cycle1
Enzy.loading b
8
Cycle 2
4.8x10
Enzy.loading c
Cycle 3
Cycle 4
4.0x108
3.2x108
Cycle 5

2.4x108

Last step

1.6x108
8.0x107
0.0

0 24 0 24 0 24 0 24 0 24 0
SSCF time (h)

72

probably inhibited such adsorption (Kristensen et al., 2009) and hence affected enzyme recycling.
The fast SSCF process removed most of the sugars by fermentation but produced ethanol whose
effect on enzyme adsorption is unclear. Therefore, more study is needed in order to effectively
recycle enzymes and specific enzyme activities using this approach. The BCRL SSCF process
achieved higher ethanol yield compared to the BCRL SHF process probably because of the
additional 24 h enzymatic hydrolysis during fermentation even though the hydrolysis was not
performed at its optimal conditions.

143

20.6% glucose
13.7% xylose

a.
100% glucose
100% xylose

BCRL SHF

10.9% glucose
40.7% xylose
0.2% glucose
8.0% xylose

68.3% glucose
37.7% xylose

Loss as
polymers
Loss as
oligomers
Loss as
Monomers

Ethanol
production

12.1% glucose
0% xylose

b.
100% glucose
100% xylose

Loss as
polymers

6.0% glucose

BCRL SSCF
37.2% xylose
Enzyme loading a

Loss as
oligomers

1.7% glucose
16.0% xylose

Loss as
Monomers

80.2% glucose
48.5% xylose
Ethanol
production

c.

12.0% glucose
0.7% xylose
6.7% glucose

BCRL SSCF
39.5% xylose
Enzyme loading b

1.9% glucose
16.6% xylose

79.5% glucose
43.2% xylose

Loss as
polymers
Loss as
oligomers
Loss as
Monomers

Ethanol
production

16.8% glucose
2.8% xylose

d.
100% glucose
100% xylose

7.8% glucose

BCRL SSCF
42.0% xylose
Enzyme loading c

1.4% glucose
16.3% xylose

74.0% glucose
38.9% xylose
Ethanol
production

Figure 49 Mass balance of BCRL SHF and SSCF processes

144

Loss as
polymers
Loss as
oligomers
Loss as
Monomers

Figure 49 (cont’d)
e.
Accellerase 1500 protein: 6.33 g
Accellerase XY protein: 0.63 g
Multifect pectinase protein: 0.63 g

- Yeast innoculum
(Dry Weight):
7.9 g

AFEXTM treated
Corn stover
1Kg
(Dry Weight)

- Ethanol: 164.2 g
BCRL SHF

- Glucose: 378.9 g
- Xylose: 266.9 g

-Glucose: 0.8 g
-Xylose: 21.4 g
- Oligo-glucose: 41.3 g
- Oligo-xylose: 108.6 g

Residual solids:
- Glucose: 78.1 g
- Xylose: 36.6 g

f. Accellerase 1500 protein: 10.2 g
Accellerase XY protein: 1.0 g
Multifect pectinase protein: 1.0 g
AFEXTM treated
Corn stover
1Kg
(Dry Weight)
- Glucose: 378.9 g
- Xylose: 266.9 g

- Yeast innoculum
(Dry Weight):
8.4 g

BCRL SSCF
Enzyme loading a

Residual solids:
- Glucose: 45.8 g

145

- Ethanol: 205.2 g
-Glucose: 6.4 g
-Xylose: 42.7 g
- Oligo-glucose: 22.7 g
- Oligo-xylose: 99.3 g

70

Avicel-glc.
AFEX-CS.glc.
AFEX-CS.xyl.

Conversion (%)

60
50
40
30
20
10
0
1

2

4
3
Cycle

5

Control

Figure 50 Enzyme activities of proteins in the removed hydrolysate after enzymatic hydrolysis
during fast SHF process.
TM
The assay was performed on Avicel and AFEX -CS. Monomeric glucose/xylose conversion is
shown in the figure. The protein concentrations in each cycle hydrolysate were 3.2, 2.7, 2.6, 3.2
and 2.8 mg/ml. The control assumed all the enzymes were removed in the hydrolysate after
cycle 1 (i.e., no enzyme recycle with unhydrolyzed solids stream). The protein concentration for
the control was 5.2 mg/ml. The enzyme activity assay was conducted in micro-plates with
working volume 1.5ml/well, substrate loading 1 mg/well, protein loading 30 µg/well, reaction
time 12 h.

146

a)
40
Concentration (g/L)

35
30
25
Xyl.80
Xyl.150
Xyl.170
Xyl.200

20
15
10

EtOH.80
EtOH.150
EtOH.170
EtOH.200

5
0

0

12

b)

24

36 48 60
Time (h)

2.0x108

72

84

96

80 g/L
150 g/L
170 g/L
200 g/L

CFU/ml

1.6x108
1.2x108
8.0x107
4.0x107
0.0
0

12 24 36 48 60 72 84 96
Time (h)

Figure 51 Effect of enzymatic hydrolysis residual solids concentration on fermentation in
hydrolysate.
Solids concentrations investigated included 80, 150, 170, and 200 g/L. Fermentations were
performed in 250 ml shake flasks with 100 ml working volume at 180 rpm, 32oC, pH 5.5 and
initial OD 2.0. The initial glucose concentration was 43.8±1.4 g/L and was consumed
completely in 24 h for all of the cases. The solids concentrations in cycle 1-6 of the BCRL
SSCF process were around 88, 125, 153, 170, 185, and 185 g/L, respectively.

147

Process EtOH productivity
(g/L/h)

Ethanol yield (g/kg AFEXTM-CS)

a)

b)

0.50
0.45
0.40
0.35
0.30
0.25
0.20
0.15
0.10
0.05
0.00
SHF-6%GL
SHF-7%GL
BCRLSHF-7.5%GL
BCRLSSCFb-7%GL

SSCF-6%GL
SSCF-7%GL
BCRL SSCFa-7%GL
BCRL SSCFc-7%GL

200
180
160
140
120
100
80
60
40
20
0
SHF-6%GL
SHF-7%GL
BCRL SHF-7.5%GL
BCRL SSCFb-7%GL

c)
BCRL SSCF

24h

24h

BCRL SHF

24h

SSCF-6%GL
SSCF-7%GL
BCRL SSCFa-7%GL
BCRL SSCFc-7%GL

24h

SSCF

Enzymatic hydrolysis

168h

6h

SHF

168h

96h

0

Fermentation

50

100

150
Time (h)

200

250

300

Figure 52 Comparisons of process ethanol productivity (a) and ethanol yield (b) of different
processes.
SHF results were based on 96 h enzymatic hydrolysis and 168 h fermentation (c). SSCF results
were based on 6 h pre-hydrolysis and 168 h fermentation. Enzyme loadings for these processes
were 36, 36, 36, 36, 22.3, 36, 30.9, and 25.8 mg/g glucan, respectively. GL: glucan loading. The
figure is in color.
148

High productivity achieved by BCRL processes could potentially reduce the reactor sizes
and hence reduce both the capital costs and operating costs (Humbird et al., 2011). Based on a
rough calculation using the NREL 2011 model (Humbird et al., 2011) and the conventional SHF
TM

process performance on AFEX

-CS, 12 saccharification tanks (250,000 gallon each) and 36

fermentors (1,000,000 gallon each) are required to process corn stover at a scale of 2,205 dry U.S.
ton /day (2,000 metric tonne/day). However, when the BCRL SHF process is applied, the
number of fermentors required is reduced to 12 with the number of saccharification tanks
remaining the same, which means the total reaction volume is reduced by around 62%. The cost
of centrifugation and filtration for the BCRL processes are almost the same as for the
conventional process. Therefore, around 62% of the capital cost in saccharification tanks,
fermentors and their accessories can be saved. Enzymes cost accounts for around 16% of the
total ethanol production cost according to the NREL 2011 model, which is one of the major costs.
The BCRL SHF process saves 38% of the enzymes compared to conventional process and hence
reduces the cost of enzymes by around 38%.
7.4 Conclusions
This novel integrated biological process, the “BCRL Process” saves time by quickly
fermenting xylose, processing the easily digestible biomass first to fully utilize the high
enzymatic hydrolysis rate period, and hydrolyzing the more recalcitrant part of biomass with less
inhibition by degradation products and sugars, and thereby enhances the process ethanol
productivity. Biocatalyst resources were better utilized by recycling yeast cells and enzymes and
thereby reduced the overall processing cost.

149

TM

CHAPTER VIII CBP OF AFEX

PRETREATED CORN STOVER AT

LOW SOLIDS LOADING

Abstract
Consolidated bioprocessing (CBP) is believed to be a potentially cost-efficient and
commercially-viable way to produce cellulosic biofuels. In this study, the performance of the
TM

CBP organism Clostridium phytofermentans (ATCC 700394) on AFEX

treated corn stover

TM

(AFEX

-CS) was evaluated. Fermentation conditions including temperature, inoculation size,

nutrients, and pH were investigated. At optimal conditions with 0.5% (w/w) glucan loading of
TM

AFEX

-CS, C. phytofermentans hydrolyzed 76% of glucan and 88.6% of xylan in 10 days.

These values reached 87% and 102% of those obtained by simultaneous saccharification and cofermentation (SSCF) using commercial enzymes and S. cerevisiae 424A. Ethanol titer for CBP
was found to be 2.8 g/L which was 71.8% of that yielded by SSCF (3.9 g/L). Decomposition
TM

products from AFEX

-CS helped to increase ethanol yield somewhat during CBP. Particle size

played a crucial role in the enhancement of sugar conversion by CBP.

150

8.1 Introduction
Both SHF and SSCF processes require expensive commercial enzymes production in a
separate bioreactor, a major cost barrier for commercialization of cellulosic ethanol (Banerjee et
al., 2010). Consolidated bioprocessing (CBP) (Lynd, 1996) which carries out enzyme production,
enzymatic hydrolysis and fermentation in a single bioreactor reduces the enzyme production cost
and at the same time possesses all the advantages of SSCF. A strong case has been made that
CBP is the ultimate low-cost industrial configuration to produce cellulosic ethanol (Lynd et al.,
2005; Sendich et al., 2008).
Currently, there is no perfect CBP microbe which can degrade lignocellulosic biomass
efficiently and at the same time utilize all the sugars released from biomass to produce mostly
ethanol. Researchers are focusing on either engineering an ethanologen (such as Saccharomyces
cerevisiae) to be able to produce cellulases/ hemicellulases or engineering a lignocellulose
degrader (such as Clostridium thermocellum) to be an efficient ethanol producer (Lynd et al.,
2005). For S. cerevisiae, xylose fermentation will be another challenge and for C. thermocellum,
which cannot consume xylose, co-culture with other bacteria (such as C. thermosaccharolyticum)
(Demain et al., 2005) will be another concern. Clostridium phytofermentans (ATCC 700394),
whose genome encodes the highest number of enzymes for degradation of lignocellulosic
material among sequenced clostridial genomes (Weber et al., 2010), is a promising native
anaerobic CBP microbe. It secretes individual enzymes (non-complexed) like fungi and can
degrade both cellulose and hemicelluloses to fermentable sugars. In addition, C. phytofermentans
can consume almost all the sugars present in lignocellulosic biomass and produce ethanol and
acetate as the major products (Warnick et al., 2002; Weber et al., 2010).

151

To the author’s knowledge, no detailed study has been reported in the literature on the
performance of C. phytofermentans on pretreated biomass. In the present study, C.
TM

phytofermentans was used to conduct CBP on AFEX

TM

treated corn stover (AFEX

-CS) to

produce ethanol. Conditions that might affect CBP performance (including temperature,
inoculation size, nutrients, and pH) were investigated. Effects of biomass washing and biomass
particle size were also studied.
8.2 Materials and Methods
TM

8.2.1 AFEX

pretreated corn stover

Please see Chapter II section 2.2.1.
8.2.2 Particle size reduction
Particle size reduction of AFEX treated corn stover (4 mm) was achieved by using a
centrifugal mill (Model ZM 200, Retsch, Newtown, PA) fitted with various ring sieve
attachments.
8.2.3 CBP microbe and seed culture preparation
C. phytofermentans ATCC 700394 was used for this CBP study. Seed culture was prepared
in a 125 ml serum vial with N2 atmosphere and 50 ml GS-2 cellobiose (5 g/L) medium (Cavedon
et al., 1990) containing (g/L): KH2PO4 1.5, K2HPO4 2.9, urea 2.1, Cysteine-HCl 2.0, MOPS 1.0,
NaCitrate•2H2O 3, yeast extract (YE) 6, MgCl2•6H2O 1, CaCl2•2H2O 0.15, FeSO4•7H2O
o

0.00125, pH 7.0. The culture was incubated at 35 C and 200 rpm in a shaking incubator
(Innova, New Brunswick, NJ).

152

8.2.4 CBP fermentation
CBP experiments were conducted in 125 ml serum vials (except for the experiments at
varying pH) with N2 atmosphere, 50 ml GS-2 medium (without cellobiose) and 0.5% (w/w)
TM

glucan loading AFEX

-CS. Experiments were conducted at 200 rpm. Fermentation

experiments with varying pH was carried out in 500 ml fermentors (Biostat, Sartorius) with
working volume 300 ml and agitation speed of 250 rpm. The fermentation medium was purged
with N2 before inoculation. CBP conditions including temperature, inoculation size, nutrients,
and pH were studied according to Table 10. For nutrient investigation, 6 g/L YE in GS-2
TM

medium were substituted by the nutrients indicated in Table 10. Particle size of AFEX

-CS

used for these studies was 0.5 mm.
Glucose, cellobiose, xylose, ethanol, acetate, lactate and formate concentrations were
analyzed using HPLC with a Biorad Aminex HPX-87H column described previously in Chapter
II section 2.2.7. In all of the CBP fermentation samples tested, no glucose, cellobiose and xylose
were detected.
8.2.5 Simultaneous Saccharification and Co-fermentation
SSCF was used as a control for this study and was conducted by using commercial
enzymes and genetically modified xylose-fermenting yeast S. cerevisiae 424A(LNH-ST) (Ho et
al., 1999). The commercial enzymes mixture was composed of Spezyme CP (Genencor Inc,
USA) 22.4 mg protein/g glucan (15 FPU/g glucan), Novozyme 188 (Sigma-Aldrich, USA) 38.4
mg protein/g glucan (64 pNPGU/g glucan), Multifect xylanase 2.6 mg protein/ g glucan and
Multifect pectinase (Genencor Inc, USA) 4.7 mg protein/ g glucan.

153

o

SSCF was conducted in the same 125 ml serum vials similar to CBP experiments at 30 C,
200 rpm with 50 ml medium containing 5 g/L yeast extract and 10 g/L tryptone. AFEX-CS
glucan loading was 0.5% (w/w). The fermentation medium pH was controlled at 5.2 with 0.05 M
phosphate buffer. The initial OD600 for S. cerevisiae 424A(LNH-ST) was 0.5. SSCF
experiments proceeded for 10 days.
8.2.6 Measurement of glucan and xylan conversions
Unutilized residual biomass samples resulting from CBP or SSCF were obtained by the
following procedures: 1) centrifuge samples at 5300 rpm for 15 min in Falcon centrifuge tubes; 2)
discard supernatant; 3) add water (same amount as supernatant) to wash the pellets; 4) centrifuge
o

again; 5) discard supernatant; 6) dry the solid residue in vacuum oven at 90 C. The glucose and
xylose contents in solid residual biomass were determined by quantitative saccharification (Lu et
al., 2006; Zhang & Lynd, 2003). The glucan/xylan conversions were calculated based on the
initial total glucose/xylose and residual total glucose/xylose.
TM

8.2.7 Preparation of washed AFEX

-CS

Please see Chapter III section 3.2.7.

154

8.3 Results and discussion
8.3.1 Effect of temperature on CBP performance

Table 10 Summary of experimental parameters for the effects of temperature, inoculation size,
TM
a
nutrients, and pH on CBP of AFEX -CS by C.phytofermentans.
Parameter
Temperature
(°C)
Inoculation
size (v/v)
Nutrients

b

c

pH

Relevant data

Effect of
temperature

Effect of
inoculation size

Effect of nutrients

Effect of
pH

25, 30, 35, 40

35

35

35

10%

5%,10%,20%,40%

10%

10%

6 g/L YE

6 g/L YE

7.0
Fig. 53a,Table
11

7.0
Fig. 53b,Table 11

6 g/L YE, 12 g/L
YE,
0.3YP,0.4YP,0.5YP
7.0
Fig. 53c,Table 11

6 g/L YE
6.5,7.0,7.5
Fig. 53d,
Table 12

a

The experiments were conducted in 125 ml serum vials with working volume 50 ml and
agitation at 200 rpm, except for the experiments investigating the effects of pH, which were
carried out in 500 ml fermentors with working volume of 300 ml and agitation at 250 rpm.
b
YE: yeast extract. YP medium: yeast extract 10 g/L and tryptone 20 g/L. The number beside
the YP represents the fraction of YP concentration.
c
Initial pH for serum vial experiments and controlled pH for fermentor experiments.

TM

CBP experiments using AFEX

o

-CS at different temperatures including 25, 30, 35, 40 C

were performed with initial pH 7.0 and inoculation size of 10% (v/v) (Table 10). Glucan and
xylan conversion results after 7 and 10 days are shown in Fig. 53a. At higher temperatures (35
o

th

and 40 C), the two day fermentation results were very close with the 7 day’s glucan
th

conversions of 51% and 20%, xylan conversions of 82% and 77% and 10 day’s glucan
conversions of 52.5% and 21%, xylan conversions of 82.5% and 77.7%, respectively. However,
155

o

at lower temperatures (25 and 30 C), we observed a significant increase in glucan conversions
th

th

Conversion (%)

Conversion (%)

Conversion (%)

Conversion (%)

from 7 day’s 43% and 60% respectively to the 10 day’s 62.4% and 71.6%, respectively.
100
100
b
a
90
90
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
0
40%
5%
10%
20%
40
35
30
25
o
Inoculation (v/v)
Temperature ( C)
Glucan(10d)
Xylan (10d)
Xylan(7d)
Glucan(7d)
Glucan(10d)
Xylan(10d)
100
100
Xylan (pH6.5)
d
c
90
Xylan (pH 7.0)
90
Xylan (pH7.5)
80
80
70
70
60
60
50
50
40
40
30
30
20
Glucan (pH6.5)
10
20
Glucan (pH7.0)
0
6g/LYE 12g/LYE 0.3YP 0.4YP 0.5YP 10
Glucan (pH7.5)
Nutrients
0
0
1
2
3
4
5
6
Glucan(10d)
Xylan(10d)
Time (d)
Figure 53 Effects of temperature (a), inoculation size (b), nutrient addition (c) and pH (d) on
TM
glucan and xylan conversion during CBP of AFEX -CS by C.phytofermentans.

By microscopy we observed few viable cells in fermentation broth after 7 days at higher
o

temperatures (35 or 40 C), while there were still many viable cells after 7 days at lower
o

o

temperatures (25 or 30 C), especially 25 C. It seems the microbe can survive and function

156

longer at lower temperatures, which probably caused the 7 day and 10 day sugar conversion
o

differences at 25 and 30 C. It is worthwhile to mention that the density of viable cells after 10
o

th

days fermentation at 25 C was similarly high when compared to the 7 days fermentation,
which means the sugar conversions might be further enhanced by extending fermentation time at
o

o

25 C. At 30 C, however, most of the cells were dead after 10 days.

Table 11 Effects of temperature, inoculation size and nutrients on lactate, acetate, formate, and
TM
a
ethanol production during CBP of AFEX -CS by C.phytofermentans.

Lactate (g/L)

Formate (g/L)

Acetate (g/L)

EtOH (g/L)

EtOH
/Acetate

o

0.00±0.00

0.86±0.05

2.43±0.19

1.17±0.06

0.48

o

0.00±0.00

0.89±0.01

2.58±0.03

1.39±0.11

0.54

o

0.38±0.08

0.88±0.03

2.38±0.04

0.92±0.07

0.39

40 C

o

0.16±0.12

1.14±0.01

2.33±0.60

0.51±0.13

0.22

5%

0.14±0.06

0.77±0.17

2.23±0.25

0.96±0.01

0.43

10%

0.38±0.08

0.88±0.03

2.38±0.04

0.92±0.07

0.39

20%

0.63±0.00

0.91±0.06

2.33±0.07

0.48±0.05

0.21

40%

0.53±0.03

0.81±0.13

2.39±0.11

0.73±0.09

0.31

6 g/LYE

0.38±0.08

0.88±0.03

2.38±0.04

0.92±0.07

0.39

12 g/LYE

0.42±0.09

0.79±0.05

2.31±0.07

0.87±0.09

0.38

0.3YP

0.20±0.08

1.00±0.02

2.66±0.18

1.03±0.01

0.39

0.4YP

0.22±0.02

1.05±0.09

2.67±0.06

1.30±0.02

0.49

0.5YP

0.35±0.03

1.07±0.08

2.57±0.08

1.17±0.06

0.46

Temperature
/Inoculation
/nutrients
25 C
30 C
35 C

a

th

The concentrations were 10 day results with inoculum concentrations subtracted.

157

In general, we found that xylan conversions were always higher than glucan conversions.
th

th

o

At both the 7 and 10 day, a 30 C fermentation gave the highest glucan and xylan
conversions and hence was chosen as the optimal temperature. Our results contradict reports by
o

Warnick et. al., (2002) that the optimum temperature for C. phytofermentans was 35 to 37 C.
The discrepancy might be caused by differences in carbon source as well as the criteria used.
Warnick et al. used cellobiose as carbon source, which can be consumed directly and quickly by
the strain. Moreover, the criterion for optimum temperature was a high growth rate. In the
TM

present study, however, AFEX

-CS was used as the carbon source, which must be hydrolyzed

before the strain can utilize it. Also, the temperature was selected for better sugar conversion.
Higher cell density with slower growth rate at relatively low temperature phenomena were also
observed by Warnick et al., consistent with the results in this study.
We also observed that ethanol production was affected by temperature (Table 11). Lower
o

th

temperatures (25 and 30 C) resulted in higher ethanol/acetate ratios. The 10 day’s highest
o

ethanol/acetate ratio (0.54) was found at 30 C.

Table 12 Effect of pH on lactate, acetate, formate, and ethanol production during CBP of AFEXa
CS by C. phyfermentans.

pH 6.5
pH 7.0
pH 7.5
a

Lactate
(g/L)
0.18±0.07
0.10±0.03
0.16±0.06

Acetate
(g/L)
2.71±0.08
2.77±0.10
2.74±0.12

Formate
(g/L)
1.03±0.09
1.05±0.05
1.17±0.10

th

Ethanol
(g/L)
1.02±0.11
0.97±0.09
0.71±0.02

Ethanol
/Acetate
0.38
0.35
0.26

The concentrations were 5 day results with inoculum concentrations subtracted.

158

8.3.2 Effect of inoculation size on CBP performance
Inoculation introduced not only microbial cells, cellulases and hemicellulases produced
during seed cultivation but also other metabolites (such as acetate and lactate) which could
inhibit CBP performance. The larger inoculums, the more microbial cells, enzymes as well as
inhibitory metabolites are introduced into CBP systems. Microbial cells and enzymes should
certainly help initiate the CBP process, but some metabolites do not. The optimal inoculation
size is the one best balancing these advantages and disadvantages. It was found that 5% (v/v)
inoculation size led to highest glucan and xylan conversions among the tested values followed by
10% (Fig. 53b). 20% (v/v) inoculation size gave the lowest sugar conversions. 40% (v/v)
inoculation size yielded higher sugar conversions than 20% probably due to the additional
enzymes introduced along with the inoculums from seed culture.
It is well known that CBP microbes tend to produce much less enzymes when grown on
non-cellulosic substrates such as cellobiose (Demain et al., 2005; Lynd et al., 2002). Therefore,
the relatively poor CBP performances caused by large inoculation sizes might also be due to an
inadequate ratio between enzymes and cells in the seed culture (grown on cellobiose) which was
not high enough to give good results at large inoculums levels. If so, the enzymes in the seed
culture were insufficient to hydrolyze biomass rapidly enough to meet the cells’ need for sugars
at the beginning of CBP. As a result, more cells might have died during the early stages of CBP
fermentation when bigger inoculums were used, thereby releasing inhibitory compounds to affect
both enzymatic hydrolysis and cell growth and hence diminish CBP performance.
The ethanol production results (Table 11) showed the same trend as sugar conversions. A 5%
(v/v) inoculation size resulted in the highest ethanol production (0.96 g/L) as well as the highest
ethanol/acetate ratio (0.43).

159

8.3.3 Effect of nutrients and pH on CBP performance
During CBP, both enzyme production and cell growth require nitrogen sources. Apparently,
C. phytofermentans prefers amino acids or peptides (Warnick et al., 2002), which are abundant in
either yeast extract or tryptone. Increasing the concentration of yeast extract did not further
increase either sugar conversion or ethanol production (Fig. 53c and Table 11). When 6 g/L yeast
extract in GS-2 medium was replaced by different fractions of YP medium (yeast extract 10 g/L
and tryptone 20 g/L), enhancements in both sugar conversions and ethanol production were
observed. Addition of 0.4YP resulted in the highest sugar conversions and highest
ethanol/acetate ratio, which were 57.4% of glucan conversion, 83.2% of xylan conversion and
0.49 of ethanol/acetate ratio, respectively. Compared to the control (6 g/L YE), glucan
conversion was increased by 4.9% while ethanol/acetate ratio enhancement was 25.6% of the
control. A higher YP fraction (0.5) resulted in reduced glucan conversion and ethanol/acetate
ratio when compared to 0.4YP. This might be due to the higher initial salt concentration caused
by the higher YP fraction since there are salts in yeast extract and tryptone. As CBP proceeds,
more salts (acetate, lactate and formate) accumulated in the fermentation broth, perhaps
inhibiting cell growth (Lynd et al., 2001) and thereby CBP performance.
Studies of pH effects in fermentors were only able to proceed for 5 days without nitrogen
purging. After 5 days, the fermentation stopped producing gas (mostly CO2 and H2 (Warnick et
al., 2002)) and the pressure of the fermentors became negative, perhaps due to more rapid cell
o

death at 35 C compared to low temperatures (section 8.3.1). Among the tested pH values, the
lower pH resulted in higher ethanol/acetate ratios (Table 12). Low pH stimulating ethanol
production was also found in a mixed culture fermentation (Zhao et al., 2009). Based on the

160

sugar conversions (Fig. 53d) and ethanol/acetate ratio, pH 6.5 was found to be the optimum
among the tested values.
o

At optimal conditions (30 C, 5% (v/v) inoculation size, and 0.4YP), 76% glucan
conversion and 88.6% xylan conversion with 2.8 g/L ethanol (ethanol/acetate ratio 1.08) were

Conversion (%)

100
90 a
80
70
60
50
40
Glucan.washed
30
Xylan.washed
20
Glucan.unwashed
10
Xylan.unwashed
0
0 1 2 3 4 5 6 7 8 9 10 11
Time (d)

3.0

b

Concentration(g/L)

2.5
2.0
1.5
1.0
0.5
0.0
0

1

2

EtOH.washed
EtOH.unwashed
For.washed

3

4

5 6 7 8
Time (d)
Ac.washed
Ac.unwashed
For.unwashed

9

10 11

Lac.washed
Lac.unwashed

Figure 54 Comparison of sugar conversions (a) and product generation (b) during CBP on
TM
TM
washed AFEX -CS and un-washed AFEX -CS.
Abbreviations: EtOH: ethanol; Ac: acetate; Lac: lactate; For: formate.

161

obtained in serum vials (initial pH 7.0, final pH 6.1±0.2) after 10 days’ CBP fermentation by
TM

C.phytofermentans (Fig. 54). During degradation of AFEX

-CS there was around 0.2 g/L

acetate released from the biomass itself. If that amount were subtracted from the observed total
acetate, the ethanol/acetate ratio produced by fermentation would be increased to 1.17.
TM

8.3.4 Comparison of CBP on washed AFEX

TM

-CS and unwashed AFEX

-CS

CBP experiments on washed and unwashed AFEX-CS were performed at optimal
conditions in serum vials (initial pH 7.0). The results are compared side by side in Fig. 54.
TM

Higher sugar conversions were observed on washed AFEX

-CS with final glucan and xylan

conversions of 82.6% and 91.4%, respectively. Apparently, removal of decomposition products
TM

present in AFEX

treated corn stover (Chundawat et al., 2010) by water washing helped

improve the CBP performance. Although sugar conversions were improved by washing, the
ethanol yield and ethanol/acetate ratio were reduced from 2.8 g/L and 1.08 to 2.2 g/L and 0.81,
respectively. Enhancement of ethanol metabolic yield by decomposition products was also
TM

observed during yeast fermentation of AFEX

-treated biomass (Lau & Dale, 2009).

8.3.5 Effect of particle size on CBP performance
Particle size is one of the factors limiting enzymatic hydrolysis (Lynd et al., 2002).
Reduction of particle size increases the surface area accessible to enzymes and thereby increases
enzymatic hydrolysis rate (Chundawat et al., 2007; Yeh et al., 2010). Similar results were also
observed in this CBP study (Fig. 55). As the particle size was reduced, the glucan and xylan
conversion rates increased. Among the tested values (4 mm, 0.5 mm, and 0.08 mm), the smallest
particle size (0.08 mm) gave the highest glucan and xylan conversions (83.1% and 89.4%,
respectively) after 10 days’ CBP fermentation. For a particle size of 4 mm, glucan was
162

100
90 a
0.08mm
80
0.5mm
70
4mm
60
50
40
30
20
10
0
0 1 2 3 4 5 6 7 8 9 10 11
Time (d)

Xylan conversion (%)

3.0 c

Acetate concentration (g/L)

EtOH concentration (g/L)

Glucan conversion (%)

100
90 b
80
70
60
50
40
0.08mm
30
0.5mm
4mm
20
10
0
0 1 2 3 4 5 6 7 8 9 10 11
Time (d)

2.5
2.0
1.5
1.0
0.5
0.0

0.08mm
0.5mm
4mm
0 1 2 3 4 5 6 7 8 9 10 11
Time (d)

3.0 d
2.5
2.0
1.5
0.08mm
0.5mm
4mm

1.0
0.5
0.0

0 1 2 3 4 5 6 7 8 9 10 11
Time (d)

TM

Figure 55 Effect of particle size of AFEX
during CBP by C. phytofermentans.

-CS on sugar conversion and product generation

hydrolyzed very slowly during the first 5 days while xylan hydrolysis was still relatively rapid.
After 10 days, the glucan and xylan conversions reached 46.3% and 78.8%, respectively. A
particle size of 0.5 mm showed slower conversion rates compared to 0.08 mm but similar xylan
conversion rates and somewhat lower glucan conversions with nearly the same final ethanol and
acetate concentrations (around 2.8 g/L ethanol and 2.6 g/L acetate).

163

100

10d Conversion (%)

90

Glucan
Xylan

80
70
60
50
40
30
20
10
0

AFEX-CS
(4mm)

AFEX-CS
(0.5mm)

AFEX-CS
(4mm)

AFEX-CS
(0.5mm)

CBP by C.phytofermentans SSCF by S.cerevisiae 424A
Figure 56 Comparison of sugar conversions achieved by CBP using C. phytofermentans
TM
and SSCF using S. cerevisiae 424A on AFEX -CS.

8.3.6 Comparison between CBP by C. phytofermentans and SSCF by S. cerevisiae 424A
A commercial enzyme mixture (containing 15 FPU cellulases) and xylose-fermenting
o

yeast S. cerevisiae 424A were used to perform SSCF at similar conditions (30 C and 200 rpm)
as for CBP. After 10 days’ SSCF, 85.6% glucan conversion and 83.3% xylan conversion were
achieved for a particle size of 4 mm (Fig. 56). Slightly higher conversions (87.4% glucan
conversion and 86.7% xylan conversion) were obtained for the particle size 0.5 mm. Compared
to the SSCF process, CBP using C. phytofermentans reached 54% and 87% of the glucan and
xylan conversions achieved by SSCF at particle size 4 mm. However, at a particle size of 0.5 mm,
CBP was able to reach 87% and 102% of the glucan and xylan conversions achieved by SSCF.
This indicates that CBP has a more stringent requirement for effective sugar release profiles than
SSCF does, probably due to the fact that CBP needs sugars to produce enzymes in addition to

164

sugars for microbial fermentation. Small particle sizes led to faster sugar release rate which
facilitated simultaneous microbial growth and enzyme production, thereby enhancing the sugar
conversions. Although at particle size 0.5 mm CBP sugar conversions were close to those of
SSCF’s it is noteworthy that SSCF reached almost the same conversions in 5 days’ reaction,
twice as fast as CBP.
After 10 days, at a particle size of 0.5 mm, CBP yielded 2.8 g/L ethanol, which is 71.8%
of that yielded by SSCF (3.9 g/L). Potentially, C. phytofermentansis is able to produce more
ethanol than S. cerevisiae from corn stover since it can also ferment galactan/galactose,
arabinan/arabinose, and mannan/mannose (Warnick et al., 2002) present in corn stover. The
lower ethanol concentration obtained in this study is due to byproduct formation, especially
acetate production. Based on the results obtained in this study at low solid loading, 174.1 g
TM

ethanol and 161.6 g acetate are generated from 1 kg AFEX

-CS through CBP using C.

phytofermentans (Fig. 57). The ethanol yield for SSCF using S. cerevisiae 424A is 242.4 g per
TM

kg AFEX

-CS. The yields, especially for CBP will be significantly reduced if the experiments

are conducted at high solid loading due to inhibition by solids and products.
8.4 Conclusion
The optimal conditions for CBP of AFEX-CS by C. phytofermentans ATCC 700394 were
o

found to be: 30 C, 5% (v/v) inoculation size and 0.4YP addition. At these conditions in serum
vials with an initial pH 7.0, 76% of glucan conversion and 88.6% of xylan conversion were
achieved in 10 days with an ethanol yield of 2.8 g/L. Decomposition products removal by
washing improved sugar conversions slightly but also decreased ethanol yield. Higher sugar
yields were observed at smaller particle sizes during CBP.

165

CBP using C.phy
Data collected from 0.5% glucan loading experiments
C.phy

AFEX treated
corn stover: 1Kg
(Dry Weight)

Ethanol: 174.1 g
(2.8 g/L)

CBP
240 h

Acetate: 161.6 g
(2.6 g/L)

- Glucose: 378.7 g
- Xylose: 266.9 g

Conversion:
Glucan: 76.0%
Xylan: 88.6%

-Glucose: 90.9 g
-Xylose: 30.4 g

SSCF using S.cerevisiae 424A
Data collected from 0.5% glucan loading experiments
- Spezyme CP: 7.4 g Protein
- Novozyme 188: 12.8 g Protein
- Multifect Xylanase: 0.86 g Protein
- Multifect Pectinase: 1.56 g Protein

Yeast
cells
AFEX treated
corn stover: 1Kg
(Dry Weight)
- Glucose: 378.7 g
- Xylose: 266.9 g

SSCF
240 h

Ethanol: 242.4 g
(3.9 g/L)
- Glc-oligomers: 8.7 g
- Xyl-oligomers: 36.9 g

-Glucose: 47.7 g
-Xylose: 35.5 g

Conversion:
Glucan: 87.4%
Xylan: 86.7%

Figure 57 Mass balance comparison between CBP by C. phytofermentans and SSCF by S.
TM
cerevisiae 424A on AFEX -CS

166

TM

CHAPTER IX CBP OF AFEX

PRETREATED CORN STOVER AT

HIGH SOLIDS LOADING

Abstract
Consolidated bioprocessing (CBP) using Clostridium phytofermentans (ATCC 700394)
TM

on AFEX

TM

treated corn stover (AFEX

-CS) at low solids loading showed promising results

(Chapter VIII). However, an industrially relevant process requires high solids loading. Therefore,
TM

high solids loading CBP performance on AFEX

-CS was studied. The factors potentially

affecting the performance including solids loading, CBP products acetate and ethanol, and
degradation products resulting from pretreatment were investigated. At 4% (w/w) glucan loading,
TM

C. phytofermentans performed well on AFEX

-CS with no nutrients supplementation and

reached similar sugar conversions as a fermentation with nutrients supplementation. A glucan
conversion of 48.9% and a xylan conversion of 77.9% were achieved after 264 h with 7.0 g/L
ethanol and 8.8 g/L acetate produced. Relatively high concentrations of acetate produced at high
solids loading was found to be the major factor limiting the CBP performance. Degradation
TM

products in AFEX

-CS helped enhance ethanol production.

167

9.1 Introduction
CBP produces enzymes during the fermentation process and eliminates the requirement
of external enzymes addition. Clostridium thermocellum and Clostridium phytofermentans are
two of the promising CBP microbes. C. thermocellum is a thermophilic anaerobic strain and
produces cellulosomes to degrade cellulose (Alexande.Jk, 1968; Lynd et al., 2005; Lynd et al.,
TM

2002; McBee, 1954). This organism solubilized AFEX

TM

pretreated corn stover (AFEX

-CS)

of particle size 0.5 mm with sugar conversions higher than 80% in 4 days at low solids loading
(Shao et al., 2011). However, C. thermocellum cannot ferment xylose in the lignocellulosic
biomass and produced around 0.9 g/L ethanol in that study (data unpublished). C.
phytofermentans is a mesophilic anaerobic strain with its genome encoding the highest number
of cellulases and hemicellulases among the sequenced clostridial genomes (Weber et al., 2010).
It can ferment almost all of the carbohydrates in lignocellulosic biomass (Warnick et al., 2002).
C. phytofermentans also reached around 80% sugar conversion at low solids loading but the
fermentation required 10 days. However, it produced more ethanol (2.8 g/L) than C.
thermocellum did and the ethanol yield of 174.1 g ethanol per kg corn stover is promising.
Nevertheless, the study was conducted at a low solids loading (0.5% glucan loading). Higher
solids loading is preferred for industrial process due to the benefits of lower capital cost, higher
ethanol titers and hence low distillation costs, and less waste water (Kristensen et al., 2009).
Industrially relevant solids loading (high solids loading) should have the potential to give enough
fermentable sugars for ethanol production at a titer of > 40 g/L (Dien et al., 2003a), which
corresponds to at least 4% (w/w) glucan loading for corn stover. To the author’s knowledge there
is very limited information on high solids loading CBP reported in the literature. Here, the high
TM

solids loading CBP performance of C. phytofermentans on AFEX
168

-CS was studied. The

possibility of performing the CBP without nutrients supplementation and the factors potentially
affecting the CBP performance at high solid loadings were investigated.
9.2 Materials and Methods
TM

9.2.1 AFEX

pretreated Corn Stover

Please see Chapter II section 2.2.1.
9.2.2 Particle size reduction
Please see Chapter VIII section 8.2.2.
TM

-CS

9.2.3 Preparation of washed AFEX

Please see Chapter III section 3.2.7.
9.2.4 CBP microbe seed cultures preparation
C. phytofermentans ATCC 700394 was used for this study. For seed cultures preparation
please see Chapter VIII section 8.2.3.
TM

9.2.5 Water extract of AFEX

-CS preparation

Please see Chapter II section 2.2.3.
9.2.6 CBP fermentation
CBP experiments at 2%, 3% and 4% (w/w) glucan loading (corresponding to 5.9%, 8.8%
and 11.7% (w/w) total solids loading, respectively) were carried out in a 500 ml fermentor
(Biostat, Sartorius) with a working volume of 300 ml, agitation speed 400 rpm, pH 6.7, and
o

temperature of 30 C. The fermentation medium was purged with N2 before inoculation. The
TM

inoculation size was 10% (v/v). GS-2 medium (without cellobiose) and washed AFEX

-CS

were used for those experiments except the experiment without nutrient supplementation. CBP
TM

of 4% (w/w) glucan loading AFEX

-CS without nutrients supplementation was started with 4%
169

glucan loading equivalent water extract as the medium and 2% (w/w) glucan loading equivalent
TM

washed AFEX

TM

-CS. Another 2% glucan loading equivalent washed AFEX

-CS was fed

TM

after 96 h. The particle size of AFEX

-CS used was 0.5 mm except for the experiment

investigating different solids loadings, in which a particle size of 4.0 mm was used.
TM

Experiments investigating the effects of AFEX

-CS water extract, acetate, and ethanol

on CBP performance and sugar fermentations were conducted in 125 ml serum vials with N2
atmosphere, 50 ml GS-2 medium (without cellobiose). For CBP, 0.5% (w/w) glucan loading
TM

AFEX

-CS with particle size 0.5 mm was used. For fermentation of pure sugars: 5 g/L

cellobiose or 10 g/L glucose or 10 g/L xylose was used. Experiments were conducted at 200 rpm
o

and 30 C. Glucose, cellobiose, xylose, ethanol, acetate, and lactate concentrations were
analyzed by HPLC with a Biorad Aminex HPX-87H column as described previously in Chapter
II section 2.2.7.
9.2.7 Measurement of glucan and xylan conversions
Please see Chapter VIII section 8.2.6.
9.3 Results and discussion
9.3.1 CBP at 4% glucan loading with and without nutrients supplementation
TM

AFEX

pretreatment is a dry to dry process which conserves nutrients in the plant

biomass and hence there is no nutrients supplementation required for yeast fermentation of
TM

AFEX

TM

treated biomass (Lau & Dale, 2009; Zhong et al., 2009). Here the CBP of AFEX

CS without nutrients supplementation was tried. One experiment with the nutrients
supplementation (GS-2 medium) was used as a control. To avoid high salts concentration that
170

-

potentially inhibits anaerobic microbe fermentation (Lynd et al., 2001), and that could be caused
TM

by the buildup of salts from both GS-2 medium and AFEX

-CS, we used washed biomass (0.5

mm) with soluble compounds removed.
The CBP sugar conversions and products profiles with and without nutrients
supplementation showed a similar trend (Fig. 58). The sugar conversion rates for the nutrients
supplementation experiment were higher with the final conversions also slightly higher than the
experiment with no nutrients supplementation (Fig. 58a). The relatively slower sugar
conversions for the no nutrients supplementation condition could be caused by the inhibition
from the degradation products generated during pretreatment. After 264 h CBP, the runs with
and without nutrients supplementation reached glucan and xylan conversions of 50.9% and
83.1%, and 48.9% and 77.9%, respectively. For both cases, around 7.0 g/L ethanol and 9.0 g/L
TM

acetate (subtracting 1.6 g/L acetate originally present in AFEX

-CS) was produced after 264 h

CBP fermentation (Fig. 58b).
TM

At 4% glucan loading, AFEX

-CS provided sufficient nutrients for the CBP

fermentation by C. phytofermentans. However, the glucan conversions for both cases were low
(around 50%). Sugar conversions did not increase even when the fermentation time was
extended (data not shown). To better understand the CBP performance at high solids loading, we
subsequently investigated several factors that could potentially affect CBP performance on
TM

AFEX

-CS.

9.3.2 Effect of solids loading on CBP performance
It is well known that solids loading plays an important role in the enzymatic hydrolysis of
lignocellulosic biomass (Kristensen et al., 2009). Here we investigated different glucan loadings

171

(2%, 3% and 4%) on the CBP performances. To eliminate the effects of degradation products,
washed biomass with particle size 4 mm was used. The sugar conversions showed similar trends
a)

Glucan(w/o)
Xylan(w/o)
Glucan(w/)
Xylan(w/)

Conversion (%)

100
90
80
70
60
50
40
30
20
10
0

0

48

96

144

192

240

Time (h)

Concentration (g/L)

b)

12
11
10
9
8
7
6
5
4
3
2
1
0

Ethanol (w/o)
Acetate (w/o)
Lactate (w/o)
Ethanol (w/)
Acetate (w/)
Lactate (w/)

0

48

96

144

Time (h)
TM

192

240

Figure 58 CBP of 4% glucan loading AFEX -CS with (w/) and without (w/o) nutrients
supplementation.
The experiments were started with half of the total biomass. The other half was fed at 96 h.
4% glucan loading equivalent water extract was used as medium for no nutrients
supplementation experiment. GS-2 medium was used for nutrients supplementation
experiment. Washed biomass with particle size 0.5 mm was used for those experiments. pH
was controlled at 6.7.

172

as observed for enzymatic hydrolysis (conversion decreased with increase of solids loading) (Fig.
59a). For enzymatic hydrolysis, high sugar (hydrolysis products) concentration limited the final
sugar conversion at high solids loading (Kristensen et al., 2009). However, for CBP we did not
observe sugar accumulation during the process. The potential inhibitors of CBP could be the
accumulated fermentation products such as acetate and ethanol (Fig. 59b), which might inhibit
sugar fermentation, cell growth, enzyme production, and enzymatic hydrolysis. With the addition
of 7.5 FPU/glucan cellulases (Spezyme CP, the optimal pH and temperature are around 4.8 and
o

50 C, respectively) at the beginning of the fermentation, the glucan and xylan conversions of 3%
glucan loading increased from 44.7% and 75.4% to 73.1% and 82.9%, respectively, and the
ethanol and acetate concentrations also increased from 6.2 and 7.6 to 11.2 and 10.9 g/L,
respectively. These data indicate that the inhibition probably acted more on the enzyme
production process than on sugar fermentation.
9.3.3 Effects of acetate and ethanol on CBP performance and sugar consumption
The experiments investigating the effects of acetate and ethanol on CBP performance
TM

were conducted at 0.5% glucan loading in a serum vial using unwashed AFEX

-CS with

particle size 0.5 mm. Different concentrations of acetate and ethanol were investigated (Fig. 60).
Sugar conversions decreased with increased ethanol or acetate concentration and the reduction of
glucan conversion was more pronounced than that of xylan conversion. Compared to ethanol,
acetate showed more severe inhibitory effects on sugar conversions. With the addition of acetate
to 10 g/L, the glucan and xylan conversions dropped from 76.0% and 88.6% to 31.1% and 73.5%,
respectively. A reduction this large was not observed until the ethanol concentration was
increased to 20 g/L, at which point the glucan and xylan conversions dropped to 37.7% and
76.8%, respectively. During high solids loading CBP fermentation, we typically see more and
173

earlier acetate production when compared to ethanol (Fig. 58b). This fact probably played a
major role in the inhibition of sugar conversion during CBP process. The inhibitory effects of
ethanol on glucose and xylose consumptions were similar (Fig. 61a&c). Ethanol concentrations
lower than 10 g/L showed mild inhibition while 20 g/L severely inhibited sugar consumption. At

Glucan
Xylan

Concentration (g/L)

Conversion (%)

100
90
80
70
60
50
40
30
20
10
0

13
12
11
10
9
8
7
6
5
4
3
2
1
0

2%

3%
4%
Glucan loading

3% (7.5FPU)

Ethanol
Acetate

2%

3%
4%
Glucan loading

3% (7.5FPU)

Figure 59 Effect of solids loading on sugar conversions (a) and products production (b) during
CBP.
3% glucan loading CBP performance was also tested with addition of 7.5 FPU cellulase
(Spezyme CP) per glucan. Particle size of the washed biomass used was 4 mm. pH was
controlled at 6.7. Experiments were performed for 264 h.

174

the tested levels (2, 5, and 10 g/L), acetate did not show severe inhibition on glucose
consumption. However, acetate with the concentrations higher than 5 g/L showed severe
inhibition of xylose consumption. The inhibitory effects of 20 g/L ethanol and 10 g/L acetate
were similar on xylose consumption (Fig. 61c&d).

Glucan

Xylan

Conversion (%)

100
90
80
70
60
50
40
30
20
10
0

20
Control 5 10
EtOH (g/L)

2
5
10
Acetate (g/L)

Figure 60 Effect of ethanol and acetate on sugar conversions of CBP.
TM
The experiment was conducted at 0.5% glucan loading using unwashed AFEX -CS with
particle size 0.5 mm for 240 h. Ethanol and acetate at different concentrations were added at
the beginning. There was no addition of ethanol or acetate for the control.

In summary, it is likely that relatively high concentrations of acetate produced during high
solids loading CBP inhibited both enzyme production, especially production of cellulases, and
xylose metabolism which in turn further affected enzymes production, and hence reduced sugar
conversions (Fig. 58).

175

b)

0

10

20

c)

50

60

Acetate(0 g/L)
Acetate(2 g/L)
Acetate(5 g/L)
Acetate(10 g/L)

70

d)
11
10
9
8
7
6
5
4
3
2
1
0
-1
-2

0

10

20

30 40
Time (h)

50

60

50

70

60

Xylose (g/L)

Xylose (g/L)

11
10
9
8
7
6
5
4
3
2
1
0
-1
-2

30 40
Time (h)

11
10
9
8
7
6
5
4
3
2
1
0

Glucose (g/L)

EtOH (0 g/L)
EtOH (5 g/L)
EtOH (10 g/L)
EtOH (20 g/L)

Glucose (g/L)

a)
11
10
9
8
7
6
5
4
3
2
1
0

Ethanol (0 g/L)
Ethanol (5 g/L)
Ethanol (10 g/L)
Ethanol (20 g/L)
0

10

20

30 40 50
Time (h)

60

70

Acetate (0 g/L)
Acetate (2 g/L)
Acetate (5 g/L)
Acetate (10g/L)
0

10

20

30 40
Time (h)

70

Figure 61 Effect of ethanol and acetate on glucose and xylose consumption.
Experiments were conducted using GS-2 medium and glucose or xylose. Ethanol and acetate were
added at the beginning.

9.3.4 Effect of degradation products on fermentation
It has been shown that without nutrients supplementation, C. phytofermentans performed
TM

well on AFEX

TM

-CS. AFEX

-CS contains not only nutrients but also degradation products

resulting from pretreatment which could potentially affect sugar conversions and fermentation
performance during CBP. However, the effect of degradation products on sugar conversions does

176

not seem significant (Fig. 58). Therefore, we only investigated the effect of degradation products
TM

on fermentation performance. The water extract of AFEX

-CS was used to represent the

degradation products and cellobiose was used as the carbon source (Fig. 62 & Table 13). At
high concentration (6% glucan loading), degradation products showed inhibition on the sugar
consumption and cell growth (Fig. 62a&b). However, at lower concentrations (0.5%, 2% and 4%
glucan loading), the inhibitory effect was not substantial (Fig. 62). With low water extract
concentrations, the ethanol metabolic yield and ethanol to acetate ratio were enhanced (Table 13).
For example, with the addition of 2% glucan loading water extract, the ethanol metabolic yield
and ethanol to acetate ratio increased from 48.5% and 1.76 to 67.2% and 2.77, respectively. We
also observed higher ethanol metabolic yield with the addition of 4% glucan loading water
extract.
In Chapter VIII, it has been shown that more ethanol was produced during CBP of
TM

unwashed AFEX

-CS compared to washed material at 0.5% glucan loading. However, CBP at

4% glucan loading without nutrients supplementation produced similar amounts of ethanol as
washed biomass with nutrients supplementation (Fig. 58). During CBP, there are two key types
of reactions taking place inside the cell and which determine the flow of carbon flux. One is cell
growth and enzymes production, which require ATP. Acetate production can provide more ATP
than ethanol production (Zhang & Lynd, 2005). The other reaction is fermentation producing
ethanol and acetate. Apparently, with 4% water extract, more ethanol can be produced during
fermentation since water extract can enhance ethanol metabolic yield. However, with these
amounts of degradation products, more enzyme production might be required to reach similar
sugar conversions as without degradation products since degradation products inhibit enzymatic
hydrolysis (Chundawat et al., 2010). Therefore, it seems likely that more sugars were used for
177

acetate production to generate more ATP for enzyme production, which offset the higher ethanol
metabolic yield of fermentation with the presence of degradation products.
b)

a)

4
3

2.0

OD600

Cellobiose (g/L)

2.5

0% WE
0.5% WE
2% WE
4% WE
6% WE

5

2

0.0

0

10

c)

20
30
Time (h)

40

50

1.4

0% WE
0.5% WE
2% WE
4% WE
6% WE

2.0
1.5
1.0
0.5
0.0

10

50

0% WE
0.5% WE
2% WE
4% WE
6% WE

1.2

Acetate (g/L)

2.5

0

0% WE
0.5% WE
2% WE
4% WE
6% WE
20
30
40
Time (h)

d)

3.0

Ethanol (g/L)

1.0
0.5

1
0

1.5

1.0
0.8
0.6
0.4
0.2

0

10

20
30
Time (h)

40

0.0

50

TM

0

10

20
30
Time (h)

40

50

Figure 62 Effect of AFEX -CS water extract (WE) on cellobiose consumption (a), cell growth
(b), ethanol (c) and acetate (d) production during cellobiose fermentation by C.phytofermentans.
The experiments were conducted in 125 ml serum vials with working volume 50 ml at 30 ºC, 200
rpm with inoculation size 10% (v/v). Water extract concentrations were expressed as glucan
loading equivalent.

178

9.4 Conclusions
TM

Production of cellulosic ethanol from AFEX

pretreated biomass through CBP is promising. However, the strain has to be improved

for less acetic acid production, higher ethanol tolerance as well as fast hydrolysis of the biomass.

TM

Table 13 Summary of AFEX
Water extract
conc.
(glucan loading
equivalent)
0% (t=0 h)
0% (t=48 h)
0.5% (t=0 h)
0.5% (t=48 h)
2% (t=0 h)
2% (t=48 h)
4% (t=0 h)
4% (t=48 h)
6% (t=0 h)
6% (t=48 h)

Glc.
(g/L)
0.00
0.00
0.19
0.10
0.75
0.36
1.51
0.64
2.26
0.00

-CS water extract (WE) effect on cellobiose fermentation.

Cellobiose Xylose
(g/L)
(g/L)
5.00
0.00
0.00
0.00
5.00
0.12
0.00
0.00
5.01
0.49
0.00
0.00
5.02
0.97
0.00
0.00
5.03
1.46
0.23
0.00

Oligoa
Glc
(g/L)
0.00
0.00
0.18
0.15
0.72
0.71
1.45
1.43
2.17
1.13

Oligoa
Xyl
(g/L)
0.00
0.00
0.55
0.32
2.18
1.66
4.37
4.35
6.55
5.51

EtOH
(g/L)
0.12
1.36
0.12
1.82
0.13
2.52
0.14
2.54
0.14
2.08

Acetate
(g/L)
0.08
0.78
0.25
0.98
0.77
1.64
1.47
2.32
2.16
3.14

EtOH
Metabolic
b
yield

Acetate
Metabolic
b
yield

EtOH/
Acetate

48.5%

18.9%

1.76

55.2%

18.2%

2.33

67.2%

18.6%

2.77

63.2%

17.3%

2.80

34.1%

13.1%

1.98

a

Oligomeric sugars were from water extract.

b

Ethanol and acetate metabolic yield were calculated based on consumed sugars. The theoretical yield for ethanol and acetate were
0.51 and 0.667 g per g consumed glucose or xylose.

179

CHAPTER X PERSPECTIVES

This dissertation work studied integrated biological processes (SSCF and CBP) on
TM

AFEX

pretreated biomass for cellulosic ethanol production, investigated fundamental issues

in these processes and based on the gained understandings developed several processes for a
better performance. The developed processes include two-step SSCF (Chapter III), continuous
hydrolysis and fermentation process (continuous SSCF; Chapter VI), and BCRL integrated
biological processes (Chapter VII). The fundamental issues studied include slow xylose
fermentation (Chapter II), factors affecting SSCF (such as SG solids inhibition, CS solids effect,
effects of ethanol and temperature on enzymatic hydrolysis, effect of pre-hydrolysis time and
effect of solids loading), factors affecting CBP performance (such as fermentation conditions,
biomass particle size, degradation products, fermentation products and solids loading). In
particular, the BCRL integrated biological processes achieved rapid hydrolysis and rapid
fermentation, shortened the cellulosic ethanol production time from 11 days to around 2 days,
reduced enzyme loading by more than 1/3 and enhanced productivity by 2~3 times. The
cellulosic ethanol production problems mentioned in Chapter I, such as long enzymatic
hydrolysis time, high enzyme loading, slow xylose fermentation and low ethanol productivity,
were largely resolved.
10.1 Techno-economic analyses of various processes
Techno-economic analyses based on the modified NREL model described in Chapter V
were performed here for overall comparison of different processes. Results are shown in Fig. 63.
The lowest minimum ethanol selling price (MESP) $2.24/gallon goes to the BCRL SSCF process
with the enzyme loading profile b, following by BCRL SSCFs (enzyme loading profile a & c),
180

Minimum Ethanol Selling Price (MESP)
$/gal.

3.1
2.9
2.7
2.5
2.3
2.1
1.9
1.7
BCRL SHF

BCRL SSCF Enzyme
loading c

BCRL SSCF Enzyme
loading b

BCRL SSCF Enzyme
loading a

Continuous SSCF
(48 ml/h)

Continuous SSCF
(24 ml/h)

Continuous SSCF
(12 ml/h)

SSCF (24h+168h)

SSCF (24h+120h)

SHF (168h+168h)

SHF (72h+96h)

1.5

Figure 63 Techno-economic analyses of various processes.
The experimental data for SHF(72 h EH + 96 h Ferm.), SHF (168 h EH+168h Ferm.), SSCF(24 h EH +120 h Ferm.), SSCF(24 h
EH + 168 h Ferm.), and continuous SSCF (flow rates: 12 ml/h, 24 ml/h and 48 ml/h) were from Chapter VI. The experimental
data for BCRL process were from Chapter VII.

181

BCRL SHF process ($2.37/gallon) and continuous SSCF (flow rate 12 ml/h) $2.52/gallon. The
process with the highest MESP $2.97/gallon is the SHF process with enzymatic hydrolysis 72 h
and fermentation 96 h. The batch SSCF (24 h pre-hydrolysis + 168 h fermentation) has almost
the same MESP as the continuous SSCF with flow rate 12 ml/h. This model did not consider the
savings of the continuous process in terms of automation which reduces operations such as
filling and cleaning of the reactors. In addition, the plant lifetime of this model is set for 30 years
instead of 20 years in the 2002 NREL model. This change makes ethanol yield the most
important factor and makes capital cost not as important as in the 2002 model. Therefore, this
dissertation work’s effort on reduction of capital cost by improving ethanol productivity might
be underestimated.
Based on the techno-economic analyses, it seems ethanol yield is now the factor that
limits further reduction of production costs. For instance, among the continuous SSCF processes,
the lowest flow rate (12 ml/h) has the highest ethanol yield and thereby the lowest MESP
although the productivity is lower than for the other two flow rates. The BCRL SHF process has
the highest ethanol productivity and the lowest enzyme loading while it has higher MESP
compared to BCRL SSCF probably due to the lower ethanol yield.
10.2 Perspectives on the BCRL integrated biological processes improvement
10.2.1 Ethanol yield enhancement
The glucose-to-ethanol conversion in the BCRL SSCF process can reach around 80% with
around 12~17% lost as polymers and 6~8% lost as oligomers (Fig. 49). It would be difficult to
convert additional polymeric glucose to ethanol since the conversion is already high and the
remaining unhydrolyzed cellulose is of high recalcitrance. Conversion of oligomeric glucose to
ethanol could increase around 10 g ethanol production from 1 kg corn stover. The xylose-to-

182

ethanol conversion in BCRL SSCF process was around 40%~50%, which is very low, with about
37%~ 42% lost as oligomeric sugars (Fig. 49). The conversion of oligomeric xylose to ethanol is
crucial! It could not only add around 50 g ethanol per kg corn stover to the total yield making the
production of around 260 g ethanol per kg corn stover possible but also could remove the
inhibitory effect of oligomeric xylose on enzymatic hydrolysis of cellulose and oligomeric
glucose. If the xylose-to-ethanol conversion reaches 80%, the MESP for BCRL SSCF process
could be as low as $1.81/ gallon.
The current xylases are probably easily inhibited by high sugar concentrations or high
ethanol concentrations (Fig. 27). Thus, SSCF did not help reduce the oligomeric xylose
production (Fig. 26, 35 & 48). Ethanol tolerant or high sugar concentration tolerant xylanases are
desired for reducing oligomeric xylose in hydrolysate. In addition, a structure study of the
remaining oligomeric xylose would be also important for finding the right xylanase to degrade
them. Obtaining a microbial strain that can consume oligomeric xylose and produce ethanol
would be another approach.
10.2.2 Thermo-tolerant xylose-fermenting ethanologen
o

In Chapter IV Fig. 27, it has been shown that the optimal temperature for SSCF is 45 C
o

and the enzymatic hydrolysis rates are much lower when the temperature is 30-32 C. Using the
current nonthermo-tolerant strain S. cerevisiae 424A for the BCRL SSCF process, the recycled
o

enzymes have to work at a temperature around 30 C, in which temperature the enzymes cannot
function to their best. A thermo-tolerant ethanologen could make the reduction of enzyme
loading for BCRL SSCF process at least as much as BCRL SHF achieved if the SSCF

183

o

temperature can be 45 C. In addition, the ethanol yield or productivity could be further
o

enhanced due to the higher sugar yield at 45 C.
10.2.3 Improvement of ethanol yield for the BCRL SHF process
The reducing cell viability in the BCRL SSCF process is an issue. The BCRL SHF
process does not have this problem. Yeast cells were growing well and were staying at a high
level (OD>20) during the entire process. In addition, based on the current microbial strain, the
BCRL SHF has higher productivity and higher enzyme loading reduction potential compared to
the BCRL SSCF process. The only disadvantage of the BCRL SHF process is the lower ethanol
yield. Further improvement of enzymatic hydrolysis for higher sugar yield by optimization of the
enzymes combination/cocktail will probably make the BCRL SHF process reach the same
ethanol yield as the BCRL SSCF.
10.2.4 Improvement of enzyme adsorption
Enzyme adsorption to biomass is strongly affected by high sugar concentrations. The
effect of ethanol on enzyme adsorption is not clear. Enzyme adsorption is strongly correlated to
the hydrolysis performance. In the BCRL integrated processes, enzyme recycling is dependent
on the adsorption to biomass. The improvement of the adsorption capability of cellulases and
hemicellulases to biomass would not only help improve hydrolysis rate and hence reduce
enzymes loading but also help on the enzymes recycling during the BCRL processes.
10.2.5 Alleviation of solids accumulation
Solids accumulation in the BCRL integrated biological processes is an issue, which
limited the cycles that can be performed. Removal of lignin during pretreatment is a way to
alleviate this problem. Another approach is to reorganize the configuration. For instance, in the
BCRL SHF process, the solids content in the enzymatic hydrolysis tank (Fig. 43) can be
184

maintained to a constant level by directing part of the residual solids to another enzymatic
hydrolysis tank as shown in Fig. 64. This way, the process can be performed with unlimited
cycles and hence probably can further reduce the enzyme loading and enhance the productivity.
The “another enzymatic hydrolysis” tank would replace the “last step” in the previous
configuration. The solids content level in the major enzymatic hydrolysis tank needs to be
investigated.
Cycle 1.

AFEX-CS
Water
Enzymes

Cycle 2.
AFEX-CS
Water
Reduced enzyme
loading
Cycle 3.
Cycle 4.

Enzymatic
hydrolysis

Supernatant goes
to fermentation

(250rpm, 50C,
24h)

Water

Part of Solid Residue
(Enzymes)
Solid Residue (Enzymes)

Enzymatic
hydrolysis

Supernatant goes
to fermentation

(250rpm, 50C,
24h)

Part of Solid Residue
(Enzymes)
Solid Residue (Enzymes)
Solid Residue (Enzymes)

Enzymatic
hydrolysis
(250rpm,
50C)

Solid Residue (Enzymes)

Cycle 5.

Figure 64 Rearranged BCRL SHF process for solving accumulated solids problem.
The figure is in color.

10.3 Perspectives on the solids inhibitory effect on fermentation
This dissertation work found that the enzymatic hydrolysis residue of switchgrass was
highly inhibitory to xylose fermentation by S. cerevisiae 424A while the residue of corn stover
was not. SSCF and CBP processes both have solids together with the microbial cells in the

185

fermentation broth. Therefore, understanding why some feedstock solids have inhibitory effect
on fermentation while others do not is important. Various feedstocks including sugarcane
bagasse, miscanthus, poplar and rice straw have been further investigated (results not shown in
this dissertation). None of those solids showed inhibitory effect. The switchgrass harvested in
different years (2008, 2009, and 2010) were also studied. Only the one harvested in 2008 has
inhibitory effect on both yeast cell growth and xylose fermentation. Currently, the various solids
samples are under characterization in the BioEnergy Science Center (BESC) using various
techniques such as

13

C NMR and

31

P NMR. The characteristics of the solids samples might not

correlate with fermentation inhibition directly. However, they might be able to give some clues,
with what more experiments need to be done to justify hypotheses.
10.4 Perspectives on the CBP process
CBP has much potential in the cellulosic ethanol production. In this dissertation work, the
performances of CBP using C. phytofermentans have been investigated at both low and high
TM

solids loadings. AFEX

pretreated biomass has been proven to be a good substrate for CBP.
TM

The current major bottlenecks for this CBP process on AFEX

pretreated biomass include

slow glucan degradation, production of acetic acid, and low tolerances to acetate and ethanol.
Genetic engineering of this strain for high expression of cellulases and less production of acetic
acid is a possible way to make this CBP process economical for cellulosic ethanol production.
Pretreatment improvement is another approach to achieve faster hydrolysis of biomass through
TM

CBP. For instance, our lab is currently developing an Extractive AFEX

TM

(E-AFEX

) process

for converting cellulose I to cellulose III and at the same time remove part of the lignin in the
biomass. Cellulose III has around 5 times higher enzyme digestibility compared to cellulose I

186

TM

(Chundawat et al., 2011b) which is the typical cellulose type in regular AFEX

Removal of lignin could also reduce the recalcitrance of the pretreated biomass.

187

treated biomass.

APPENDIX

188

APPENDIX MODIFIED NREL 2011 TECHNO-ECONOMIC MODEL
PARAMETERS

The modified NREL model is developed by Dr. Bryan Bals (BCRL, Michigan State University)
based on NREL’s 2011 model (please see Chapter V section 5.2.6). BCRL SSCF process with
enzyme loading b is used here as an example.

Model inputs
Biomass Composition
1000 g/kg untreated
341 g/kg input
234 g/kg input
30 g/kg input
150 g/kg input
50 g/kg input
80 g/kg input
165 g/kg input
Enzymatic Hydrolysis
Solid loading
20.5 % solids input
Cellulase Loading
8.78 mg/g Biomass
Hemicellulase Loading
1.756 mg/g Biomass
High viscosity residence time
28.8 hours
Low viscosity residence time
0 hours
Cellulase Cost
3600 $/Mg
Hemicellulase Cost
4500 $/Mg
Glucan to oligomer conversion
6.7 % glucan
Glucan to monomer conversion
81.4 % glucan
Xylan to oligomer conversion
39.5 % xylan
Xylan to monomer conversion
59.8 % xylan
Arabinan to oligomer conversion
0 % arabinan
Arabinan to monomer conversion
0 % arabinan
Fermentation
Residence Time
28.8 hours
Glucose Conversion
97.7 %
O-Glucose Conversion
0 %
Glucose Metabolic Yield
0.47 g ethanol/g sugar
Xylose Conversion
72.2 %
Fraction of untreated
Glucan
Xylan
Arabinan
Klason Lignon
Ash
Water Extractives
Other

189

O-Xylose Conversion
Xylose Metabolic Yield
Arabinose Conversion
O-Arabinose Conversion
Arabinose Metabolic Yield
Inoculum size
Biomass Input Price
Pre-processing Costs
Biorefinery Size
Electricity selling price
Lignin selling price
Internal Rate of Return
Ethanol selling price
Pretreatment installed cost
Thermal Energy Use
Electricity Use
Ammonia Use
Ammonia Price
Other variable costs

0 %
0.47 g ethanol/g sugar
0 %
0 %
0.45 g ethanol/g sugar
10 % of hydrolysate volume
Others
40 $/Mg
50 $/Mg Untreated
2000 Mg/day
5.72 cents/kW*h
0 $/Mg
10 %
2.5 $/gal
Pretreatment Costs
0 Million $
0 GJ/Mg biomass
0 kWh/Mg biomass
22 g/kg biomass
0 $/Mg
0 $/Mg

Model outputs
MESP
IRR
NPV

MESP
2.254
0.10
0.0000

Ethanol yield (kg/Mg)
195.6
Ethanol yield (gal/ton)
59.41
Ethanol volume (Mmgal/yr)
45.89
Ethanol titer (w/w)%
4.63%
Net electricity (kWh/Mg)
359.6
Economic consideration
Fixed Capital Investment (MM$)
$275.77
TCI/Annual gallon
$6.349
Ethanol Selling Price
$2.25
IRR
10.0%
Feedstock Prod Cost ($/gal)
$0.611
Feedstock Del Cost ($/gal)
$1.374
Enzyme Cost ($/gal)
$0.603
Fixed Costs ($/gal)
$0.148
190

IRR
2.500
1.03
0.0000
195.6
59.41
45.89
4.63%
359.6
$275.77
$6.349
$2.50
102.9%
$0.611
$1.374
$0.603
$0.148

Other Op Costs ($/gal)

$0.072

$0.072

Mass balance
Initial Water
Mglucose
Oglucose
Mxylose
Oxylose
Marabinose
Oarabinose
Pglucose
Pxylose
Parabinose
Solids
Water in solids
Water reacted
Water in hydrolysate
Water added
Water in fermentation
Ethanol
Ethanol conc
Ethanol yield

REFERENCE
FRACTION

3871.429
308.4156
23.48164
159.0136
95.58102
0
0
40.579
1.638
30
271.717
271.717
53.70885
3546.003
487.1429
4033.146
195.581
0.046251
0.247885
65.48441
59.40646
79
0.751981

g/kg solids
g/kg solids
g/kg solids
g/kg solids
g/kg solids
g/kg solids
g/kg solids
g/kg solids
g/kg solids
g/kg solids
g/kg solids
g/kg solids
g/kg solids
g/kg solids

g/kg solids
w/w
L/kg dry BM
gal/Mg dry BM
gal/ton dry BM

Capital Cost
Total water use in hydrolysate
Baseline reference
Fraction of baseline
Size fraction
Total fermentation volume
Reference fermentation
Fraction

3546.0 Lignin present
4093 Reference lignin
Fraction of
0.8663 baseline
1 Lignin energy
4228.7 Reference energy
Fraction of
4479.55 baseline
0.9440

191

271.717
273
0.99530
cal/g entering
1187.3 bm
1167.6
1.0169

WITH BOILER
Process Area 2007 dollors
Area 100: Feedstock handling
Area 200: Pretreatment
Area 200: Conditioning
Area 300: Enzymatic hydrolysis and
fermentation
Area 400: Enzyme production
Area 500: Recovery
Area 600: Wastewater
Area 700: Storage
Area 800: Boiler
Area 900: Utilities
Totals (excl. Area 100)
Warehouse 4.0% of ISBL
Site development 9.0% of ISBL
Additional piping 4.5% of ISBL
Total Direct Costs (TDC)
Prorateable expenses 10.0% of TDC
Field expenses 10.0% of TDC
Home office & construction fee 20.0% of TDC
Project contingency 10.0% of TDC
Other costs (start-up, permits, etc.) 10.0% of
TDC
Total Indirect Costs
Fixed Capital Investment (FCI)
Land
Working capital 5.0% of FCI
Total Capital Investment (TCI)
Lang Factor (FCI/purchased equip cost)b 3.1
TCI per annual gallon

Purchased
14,200,000
19,900,000
1,500,000

Installed
24,200,000
29,900,000
3,000,000

0
0
0

0
0
0

18,500,000
10,700,000
11,100,000
49,300,000
2,800,000
36,500,000
4,000,000
154,500,000

31,200,000
18,300,000
22,300,000
49,400,000
5,000,000
66,000,000
6,900,000
232,100,000
4,200,000
9,400,000
4,700,000
250,400,000
25,000,000
25,000,000
50,100,000
25,000,000

19,758,956
0
20,590,828
47,721,280
4,585,139
65,918,232
6,719,605
165,294,040
1,613,991
3,631,481
1,815,740
172,355,252
17,235,525
17,235,525
34,471,050
17,235,525

19,758,956
0
20,590,828
47,721,280
4,585,139
65,918,232
6,719,605
165,294,040
1,613,991
3,631,481
1,815,740
172,355,252
17,235,525
17,235,525
34,471,050
17,235,525

17,235,525
103,413,151
275,768,403
1,800,000
13,788,420
291,356,823

17,235,525
103,413,151
275,768,403
1,800,000
13,788,420
291,356,823

25,000,000
150,200,000
400,600,000
1,800,000
20,000,000
422,500,000
6.92/gal

104,700,000

232,000,000
0
0
0
232,000,000

6
192

Energy Balance
Ref
1167.6
1008
1740.48

Modified
1187.368 cal/g entering biomass lignin
1008
cal/g entering biomass other
1756.294 cal/g steam

10.30804
68.16027
31.63213
64.08026
25.4881
88.41635
0.132001
16.41607
37.69215

0
0
20.25089
0
25.4881
83.46566
0.099262
16.41607
37.69215

kWh/Mg
kW/Mg
kW/Mg
kW/Mg
kW/Mg
kW/Mg
kW/Mg
kW/Mg
kW/Mg

Difference

Area 1
Area 2
Area 3
Area 4
Area 5
Area 6
Area 7
Area 8
Area 9

342.3254 183.4121

158.913244

Heat requirement
1.959
1.784815 GJ/Mg
0.1557
0
2.1147
1.784815
258.4654 218.1458

0.329885

40.3196355 kW*hr/Mg
kW*hr/Mg
199.232879 saved

Economics
FIXED COSTS
Salaries
Benefits
Maintenance
Insurance
Raw Material Costs
Ammonia
Feedstock
Ash disposal
1.5319
A900 Items
A800 Items
A300 Items
2.154
A600 Item
6.821508
Enzymes
39.51

1.9184
1.72656
1.2104935
1.9303788
6.7858323

1.7836
1.60524
1.210494
1.192095
5.791428

MM$/yr
MM$/yr
MM$/yr
MM$/yr

0
63.075001
1.5247006
0.38055
0.01034
1.0166955
0.3891044
27.689928
94.086319

0
63.075
0
0.38055
0
1.016696
0.389104
27.68993
92.55128

MM$/yr
MM$/yr
MM$/yr
MM$/yr
MM$/yr
MM$/yr
MM$/yr
MM$/yr

28.03333

193

Other Revenue
Lignin
Electricity
Ethanol production

Raw
materials
A200
A300

A400

A600
A800

A900
Subtotal
Waste
A800
Subtotal
Byproducts

0
14.417414
45.89366

45.89366 MM gal/yr

kg/hr
104,167
1,981
1,051
1,158

LB/HR
229,688
4,367
2,317
2,554

$/USton
46.8
81.39
406.96
51.55

$/hr
5,374.69
177.73
471.48
65.84

MM$
/yr
45.2
1.49
3.97
0.55

Cents
/gal
74.07
2.45
6.5
0.91

142
44
2,418
164
115
67
16
2,252
<1
895

313
98
5,332
363
254
149
36
4,966
1
1,973

895.32
1,021.93
526.52
51.55
406.96
745.3
275.7
135.65
4,532.17
180.87

140.33
49.96
1,403.60
9.35
51.59
55.34
4.99
336.83
1.23
178.42

1.18
0.42
11.8
0.08
0.43
0.47
0.04
2.83
0.01
1.5

1.94
0.69
19.34
0.13
0.71
0.76
0.07
4.64
0.02
2.46

2
147,140

5
324,443

2,716.10
0.23

7.14
38.11
8,328.49

0.06
0.32
70.36

0.1
0.53
115.3

disposal
Ash Disposal

5,725

12,623

28.86

182.15
182.15

1.53
1.53

2.51
2.51

and

credits
731.57

6.15

10.08

49.13
780.71

0.41
6.57

7,779.08

65.33

0.68
10.76
107.7
6

Feedstock
Sulfuric
Ammonia
CSL
Diammonium
phosphate
Sorbitol
Glucose
CSL
Ammonia
Host nutrients
Sulfur dioxide
Caustic
Boiler
FGD lime
Cooling Tower
chemicals
Makeup water

Grid Electricity 12,797

kW

Area 100 Elec

859

kW

variable

operating

costs

Subtotal
Total

194

$0.0572/k
Wh
$0.0572/k
Wh

Rate of
return

Ethanol
price
2.25391
IRR
0.1
Tax
0.35
TCI
275.768
-2

Year
Ethanol sales
Electricity sales
Fixed costs
Operating costs
Depreciation
Depreciation
Real Depreciation
Real Depreciation
PWFv
Other capital
Loan Payment
Actual Interest

PWFc
Net
Revenue
Losses
Forward
Taxable
Income
Income Tax
Annual Cash Income
-159.165

Equity
Interest

0.4
0.14903
0.08

-1

0

1
90.51
12.62
6.79
85.12
0.14
0.04
2.47
29.99

2
103.44
14.42
6.79
94.09
0.24
0.07
4.76
51.39

3
103.44
14.42
6.79
94.09
0.17
0.07
4.40
36.70

4
103.44
14.42
6.79
94.09
0.12
0.06
4.07
26.21

5
103.44
14.42
6.79
94.09
0.09
0.06
3.77
18.74

6
103.44
14.42
6.79
94.09
0.09
0.05
3.48
18.72

7
103.44
14.42
6.79
94.09
0.09
0.05
3.22
18.74

8
103.44
14.42
6.79
94.09
0.04
0.05
2.98
9.36

8.82
1.80
1.06

66.18
9.00

35.30
13.79
13.24

1.21

1.10

1.00

24.66
13.24
11.42
154.04
0.91

24.66
12.32
12.34
141.70
0.83

24.66
11.34
13.32
128.38
0.75

24.66
10.27
14.39
113.99
0.68

24.66
9.12
15.54
98.45
0.62

24.66
7.88
16.78
81.67
0.56

24.66
6.53
18.12
63.55
0.51

24.66
5.08
19.57
43.97
0.47

-34.48

-51.49

-35.45

-23.57

-14.64

-13.09

-11.51

-0.44

-34.48

-85.97 -121.42 -144.99 -159.63 -172.72 -184.23

14.14

82.70

62.32

-34.48
0.00
-13.44
-12.22
0.00

-85.97 -121.42 -144.99 -159.63 -172.72 -184.23 -184.67
0.00
0.00
0.00
0.00
0.00
0.00
0.00
-7.67
-7.67
-7.67
-7.67
-7.67
-7.67
-7.67
-6.34
-5.76
-5.24
-4.76
-4.33
-3.94
-3.58

195

Year
Ethanol sales
Electricity sales
Fixed costs
Operating costs
Depreciation
Depreciation
Real Depreciation
Real Depreciation
PWFv
Other capital
Loan Payment
Actual Interest

PWFc
Net Revenue
Losses Forward
Taxable Income
Income Tax
Annual Cash
Income

9

10

11

12

13

14

15

103.44
14.42
6.79
94.09

103.44
14.42
6.79
94.09

103.44
14.42
6.79
94.09

103.44
14.42
6.79
94.09

103.44
14.42
6.79
94.09

103.44
14.42
6.79
94.09

0.04
2.94
0.00

0.04
2.94
0.00

0.04
2.94
0.00

0.04
2.94
0.00

0.04
2.94
0.00

0.04
2.94
0.00

24.66
3.52
21.14
22.83
0.42
10.53
184.67
174.14

24.66
1.83
22.83
0.00
0.39
12.22
174.14
161.92

0.35
14.04
161.92
147.88

0.32
14.04
147.88
133.83

0.29
14.04
133.83
119.79

0.26
14.04
119.79
105.75

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

-7.67
-3.25

-7.67
-2.96

16.99
5.95

16.99
5.41

16.99
4.92

16.99
4.47

16.99
4.07

16.99
3.70

16.99
3.36

16.99
3.05

16.99
2.78

16.99
2.52

196

103.44
14.42
6.79
94.09

16
103.4
4
14.42
6.79
94.09

17
103.4
4
14.42
6.79
94.09

18
103.4
4
14.42
6.79
94.09

19
103.4
4
14.42
6.79
94.09

20
103.4
4
14.42
6.79
94.09

0.04
2.94
0.00

0.04
2.94
0.00

0.04
2.94
0.00

0.04
2.94
0.00

0.04
2.94
0.00

0.04
2.94
0.00

0.24
0.22
0.20
0.18
0.16
0.15
14.04 14.04 14.04 14.04 14.04 14.04
105.75 -91.70 -77.66 -63.61 -49.57 -35.52
-91.70 -77.66 -63.61 -49.57 -35.52 -21.48

Year
Ethanol sales
Electricity sales
Fixed costs
Operating costs
Depreciation
Depreciation
Real Depreciation
Real Depreciation
PWFv
Other capital
Loan Payment
Actual Interest

21
103.44
14.42
6.79
94.09

22
103.44
14.42
6.79
94.09

23
103.44
14.42
6.79
94.09

24
103.44
14.42
6.79
94.09

25
103.44
14.42
6.79
94.09

26
103.44
14.42
6.79
94.09

27
103.44
14.42
6.79
94.09

28
103.44
14.42
6.79
94.09

29
103.44
14.42
6.79
94.09

30
103.44
14.42
6.79
94.09

0.02
1.47
0.00

0.00
0.00

0.00
0.00

0.00
0.00

0.00
0.00

0.00
0.00

0.00
0.00

0.00
0.00

0.00
0.00

0.00
0.00

PWFc
Net Revenue
Losses Forward
Taxable Income

0.14
15.51
-21.48
-5.96

0.12
16.99
-5.96
11.02

0.11
16.99
0.00
16.99

0.10
16.99
0.00
16.99

0.09
16.99
0.00
16.99

0.08
16.99
0.00
16.99

0.08
16.99
0.00
16.99

0.07
16.99
0.00
16.99

0.06
16.99
0.00
16.99

0.15
16.99
0.00
16.99

0.00
16.99
2.30

3.86
13.13
1.61

5.94
11.04
1.23

5.94
11.04
1.12

5.94
11.04
1.02

5.94
11.04
0.93

5.94
11.04
0.84

5.94
11.04
0.77

5.94
11.04
0.70

5.94
11.04
1.64

Income Tax
Annual Cash Income

0.00

197

CAPEX
Dependent on solid loading and hyd/fermentation time
A-300
Ethanol Fermentor Agitator
1
$951,337
A-306
Beer Surge Tank Agitator
0.5
$106,105
F-300
Ethanol Fermentor
1 $15,293,882
0.9968 $16,351,324
Dependent on viscous hydrolysate residence time and solid loading
T-310
Saccharification Tank
0.7 $8,006,772
0.7 $8,006,772
Dependent on fermentation OD
A-301
Seed Hold Tank Agitator
0.5
$49,305
A-304
4th Seed Vessel agitator
0.5
$39,262
A-305
5th Seed vessel agitator
0.5
$64,933
F-301
1st Seed fermentor
0.7
$136,630
F-302
2nd Seed fermentor
0.7
$211,288
F-303
3rd Seed fermentor
0.7
$285,582
F-304
4th Seed fermentor
0.7
$708,721
F-305
5th Seed fermentor
0.7 $2,375,827
P-301
Seed Hold transfer pump
0.8
$18,795
P-302
Seed Transfer pump
0.8
$55,698
T-301
Seed Hold tank
0.7
$825,471
$4,771,512
Depends on solid loading only
H-300
Fermentation Cooler
1
$192,524
H-301
Hydrolyzate Cooler
0.7
$214,500
H-310
Fermentor Batch cooler
0.7
$0
P-300
Fermentation Recirc pump
0.8
$109,288
P-306
Beer Transfer pump
0.8
$58,165
P-310
Saccharification Tansfer pump
0.8
$113,746
T-306
Beer Storage tank
0.7 $1,199,170
A-308
Enzyme-Hydrolysate Mixer
0.5
$201,519
$2,088,912

198

0.9968

0.7000
24652.5000
19631.0000
32466.5000
95641.0000
147901.6000
199907.4000
496104.7000
1663078.9000
15036.0000
44558.4000
577829.7000
0.6951
192524.0000
150150.0000
0.0000
87430.4000
46532.0000
90996.8000
839419.0000
100759.5000
0.7218

REFERENCES

199

REFERENCES

Agbor, V.B., Cicek, N., Sparling, R., Berlin, A., Levin, D.B. 2011. Biomass pretreatment:
Fundamentals toward application. Biotechnology Advances, 29(6), 675-685.
Alexande.Jk. 1968. Purification and specificity of cellobiose phosphorylase from Clostridium
thermocellum. Journal of Biological Chemistry, 243(11), 2899-2904.
Alkasrawi, M., Eriksson, T., Borjesson, J., Wingren, A., Galbe, M., Tjerneld, F., Zacchi, G. 2003.
The effect of Tween-20 on simultaneous saccharification and fermentation of softwood to
ethanol. Enzyme and Microbial Technology, 33(1), 71-78.
Almeida, J.R.M., Modig, T., Petersson, A., Hähn-Hägerdal, B., Lidén, G., Gorwa-Grauslund,
M.F. 2007. Increased tolerance and conversion of inhibitors in lignocellulosic
hydrolysates by Saccharomyces cerevisiae. Journal of Chemical Technology &
Biotechnology, 82(4), 340-349.
Argyros, D.A., Tripathi, S.A., Barrett, T.F., Rogers, S.R., Feinberg, L.F., Olson, D.G., Foden,
J.M., Miller, B.B., Lynd, L.R., Hogsett, D.A., Caiazza, N.C. 2011. High Ethanol Titers
from Cellulose by Using Metabolically Engineered Thermophilic, Anaerobic Microbes.
Applied and Environmental Microbiology, 77(23), 8288-8294.
Azevedo, H., Bishop, D., Cavaco-Paulo, A. 2002. Possibilities for recycling cellulases after use
in cotton processing part II: Separation of cellulases from reaction products and released
dyestuffs by ultrafiltration. Applied Biochemistry and Biotechnology, 101(1), 77-91.
Bai, F.W., Anderson, W.A., Moo-Young, M. 2008. Ethanol fermentation technologies from
sugar and starch feedstocks. Biotechnology Advances, 26(1), 89-105.
Balan, V., Bals, B., Chundawat, S.P.S., Marshall, D., Dale, B.E. 2010. Lignocellulosic Biomass
Pretreatment Using AFEX. in: Biofuels, Vol. 581, pp. 61-77.
Bals, B., Rogers, C., Jin, M., Balan, V., Dale, B. 2010. Evaluation of ammonia fibre expansion
(AFEX) pretreatment for enzymatic hydrolysis of switchgrass harvested in different
seasons and locations. Biotechnology for Biofuels, 3, 1.
Bals, B.D., Dale, B.E. 2012. Developing a model for assessing biomass processing technologies
within a local biomass processing depot. Bioresource Technology, 106(0), 161-169.
Banerjee, G., Scott-Craig, J.S., Walton, J.D. 2010. Improving Enzymes for Biomass Conversion:
A Basic Research Perspective. Bioenergy Research, 3(1), 82-92.
Bellissimi, E., van Dijken, J.P., Pronk, J.T., van Maris, A.J.A. 2009. Effects of acetic acid on the
kinetics of xylose fermentation by an engineered, xylose-isomerase-based
Saccharomyces cerevisiae strain. Fems Yeast Research, 9(3), 358-364.
200

Bengtsson, O., Hahn-Hägerdal, B., Gorwa-Grauslund, M.F. 2009. Xylose reductase from Pichia
stipitis with altered coenzyme preference improves ethanolic xylose fermentation by
recombinant Saccharomyces cerevisiae. Biotechnology for Biofuels, 2(1), 1-10.
Bertilsson, M., Andersson, J., Liden, G. 2008. Modeling simultaneous glucose and xylose uptake
in Saccharomyces cerevisiae from kinetics and gene expression of sugar transporters.
Bioprocess and Biosystems Engineering, 31(4), 369-377.
Bertilsson, M., Olofsson, K., Liden, G. 2009. Prefermentation improves xylose utilization in
simultaneous saccharification and co-fermentation of pretreated spruce. Biotechnology
for Biofuels, 2(1), 8.
Brethauer, S., Wyman, C.E. 2010. Review: Continuous hydrolysis and fermentation for
cellulosic ethanol production. Bioresource Technology, 101(13), 4862-4874.
Cara, C., Moya, M., Ballesteros, I., Negro, M.J., González, A., Ruiz, E. 2007. Influence of solid
loading on enzymatic hydrolysis of steam exploded or liquid hot water pretreated olive
tree biomass. Process Biochemistry, 42(6), 1003-1009.
Carere, C.R., Sparling, R., Cicek, N., Levin, D.B. 2008. Third generation biofuels via direct
cellulose fermentation. International Journal of Molecular Sciences, 9(7), 1342-1360.
Cavedon, K., Leschine, S., Canale-Parola, E. 1990. Cellulase system of a free-living, mesophilic
clostridium (strain C7). Journal of bacteriology, 172(8), 4222.
Chundawat, S., Beckham, G., Himmel, M., Dale, B. 2011a. Deconstruction of lignocellulosic
biomass to fuels and chemicals. Annu Rev Chem Biomol Eng, 2, 121-145.
Chundawat, S.P.S., Bellesia, G., Uppugundla, N., Sousa, L.D., Gao, D.H., Cheh, A.M., Agarwal,
U.P., Bianchetti, C.M., Phillips, G.N., Langan, P., Balan, V., Gnanakaran, S., Dale, B.E.
2011b. Restructuring the Crystalline Cellulose Hydrogen Bond Network Enhances Its
Depolymerization Rate. Journal of the American Chemical Society, 133(29), 1116311174.
Chundawat, S.P.S., Donohoe, B.S., Sousa, L.D., Elder, T., Agarwal, U.P., Lu, F.C., Ralph, J.,
Himmel, M.E., Balan, V., Dale, B.E. 2011c. Multi-scale visualization and
characterization of lignocellulosic plant cell wall deconstruction during thermochemical
pretreatment. Energy & Environmental Science, 4(3), 973-984.
Chundawat, S.P.S., Venkatesh, B., Dale, B.E. 2007. Effect of particle size based separation of
milled corn stover on AFEX pretreatment and enzymatic digestibility. Biotechnology and
Bioengineering, 96(2), 219-231.
Chundawat, S.P.S., Vismeh, R., Sharma, L.N., Humpula, J.F., Sousa, L.D., Chambliss, C.K.,
Jones, A.D., Balan, V., Dale, B.E. 2010. Multifaceted characterization of cell wall
decomposition products formed during ammonia fiber expansion (AFEX) and dilute acid
based pretreatments. Bioresource Technology, 101(21), 8429-8438.

201

Crous, J.M., Pretorius, I.S., Vanzyl, W.H. 1995. Cloning and Expression of an AspergillusKawachii Endo-1,4-Beta-Xylanase Gene in Saccharomyces-Cerevisiae. Current Genetics,
28(5), 467-473.
Demain, A.L., Newcomb, M., Wu, J.H.D. 2005. Cellulase, clostridia, and ethanol. Microbiology
and Molecular Biology Reviews, 69(1), 124-154.
Den Haan, R., McBride, J.E., La Grange, D.C., Lynd, L.R., Van Zyl, W.H. 2007. Functional
expression of cellobiohydrolases in Saccharomyces cerevisiae towards one-step
conversion of cellulose to ethanol. Enzyme and Microbial Technology, 40(5), 1291-1299.
Desvaux, M. 2006. Unravelling carbon metabolism in anaerobic cellulolytic bacteria.
Biotechnology Progress, 22(5), 1229-1238.
Diderich, J.A., Schepper, M., van Hoek, P., Luttik, M.A.H., van Dijken, J.P., Pronk, J.T.,
Klaassen, P., Boelens, H.F.M., de Mattos, R.J.T., van Dam, K., Kruckeberg, A.L. 1999.
Glucose uptake kinetics and transcription of HXT genes chemostat cultures of
Saccharomyces cerevisiae. Journal of Biological Chemistry, 274(22), 15350-15359.
Dien, B., Cotta, M., Jeffries, T. 2003a. Bacteria engineered for fuel ethanol production: current
status. Applied Microbiology and Biotechnology, 63(3), 258-266.
Dien, B.S., Cotta, M.A., Jeffries, T.W. 2003b. Bacteria engineered for fuel ethanol production:
current status. Applied Microbiology and Biotechnology, 63(3), 258-266.
Dien, B.S., Ximenes, E.A., O'Bryan, P.J., Moniruzzaman, M., Li, X.L., Balan, V., Dale, B.,
Cotta, M.A. 2008. Enzyme characterization for hydrolysis of AFEX and liquid hot-water
pretreated distillers' grains and their conversion to ethanol. Bioresource Technology,
99(12), 5216-5225.
Eggeman, T., Elander, R.T. 2005. Process and economic analysis of pretreatment technologies.
Bioresource Technology, 96(18), 2019-2025.
Eriksson, T., Borjesson, J., Tjerneld, F. 2002a. Mechanism of surfactant effect in enzymatic
hydrolysis of lignocellulose. Enzyme and Microbial Technology, 31(3), 353-364.
Eriksson, T., Karlsson, J., Tjerneld, F. 2002b. A model explaining declining rate in hydrolysis of
lignocellulose substrates with cellobiohydrolase I (Cel7A) and endoglucanase I (Cel7B)
of Trichoderma reesei. Applied Biochemistry and Biotechnology, 101(1), 41-60.
Farrell, A., Plevin, R., Turner, B., Jones, A., O'hare, M., Kammen, D. 2006. Ethanol can
contribute to energy and environmental goals. Science, 311(5760), 506.
Fogler, H.S. 1999. Elements of chemical reaction engineering. Prentice-Hall International
London.
Fujita, Y., Ito, J., Ueda, M., Fukuda, H., Kondo, A. 2004. Synergistic saccharification, and direct
fermentation to ethanol, of amorphous cellulose by use of an engineered yeast strain
202

codisplaying three types of cellulolytic enzyme. Applied and Environmental
Microbiology, 70(2), 1207-1212.
Gao, D., Chundawat, S.P.S., Krishnan, C., Balan, V., Dale, B.E. 2010. Mixture optimization of
six core glycosyl hydrolases for maximizing saccharification of ammonia fiber expansion
(AFEX) pretreated corn stover. Bioresource Technology, 101(8), 2770-2781.
Gao, D.H., Chundawat, S.P.S., Uppugundla, N., Balan, V., Dale, B.E. 2011. Binding
Characteristics of Trichoderma reesei Cellulases on Untreated, Ammonia Fiber
Expansion (AFEX), and Dilute-Acid Pretreated Lignocellulosic Biomass. Biotechnology
and Bioengineering, 108(8), 1788-1800.
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.
Görke, B., Stülke, J. 2008. Carbon catabolite repression in bacteria: many ways to make the most
out of nutrients. Nature Reviews Microbiology, 6(8), 613-624.
Govindaswamy, S., Vane, L.M. 2010. Multi-stage continuous culture fermentation of glucosexylose mixtures to fuel ethanol using genetically engineered Saccharomyces cerevisiae
424A. Bioresource Technology, 101(4), 1277-1284.
Gray, K.A., Zhao, L.S., Emptage, M. 2006. Bioethanol. Current Opinion in Chemical Biology,
10(2), 141-146.
Greene, N., Council, N.R.D. 2004. Growing energy: How biofuels can help end America's oil
dependence. Natural Resources Defense Council New York, NY.
Gusakov, A.V., Sinitsyn, A.P. 1992. A theoretical-analysis of cellulase product inhibition-effect
of cellulase binding constant, enzyme substrate ratio, and beta-glucosidase activity on the
inhibition pattern. Biotechnology and Bioengineering, 40(6), 663-671.
Ha, S.J., Galazka, J.M., Rin Kim, S., Choi, J.H., Yang, X., Seo, J.H., Louise Glass, N., Cate,
J.H.D., Jin, Y.S. 2011. Engineered Saccharomyces cerevisiae capable of simultaneous
cellobiose and xylose fermentation. Proceedings of the National Academy of Sciences,
108(2), 504.
Hahn-Hagerdal, B., Karhumaa, K., Fonseca, C., Spencer-Martins, I., Gorwa-Grauslund, M.F.
2007. Towards industrial pentose-fermenting yeast strains. Applied Microbiology and
Biotechnology, 74(5), 937-953.
Hallac, B.B., Ragauskas, A.J. 2011. Analyzing cellulose degree of polymerization and its
relevancy to cellulosic ethanol. Biofuels Bioproducts & Biorefining-Biofpr, 5(2), 215-225.
Hamelinck, C.N., Hooijdonk, G.v., Faaij, A.P.C. 2005. Ethanol from lignocellulosic biomass:
techno-economic performance in short-, middle- and long-term. Biomass and Bioenergy,
28(4), 384-410.

203

Hansen, A.C., Zhang, Q., Lyne, P.W.L. 2005. Ethanol-diesel fuel blends - a review. Bioresource
Technology, 96(3), 277-285.
Ho, N., Chen, Z., Brainard, A., Sedlak, M. 1999. Successful Design and Development of
Genetically Engineered Saccharomyces Yeasts for Effective Cofermentation of Glucose
and Xylose from Cellulosic Biomass to Fuel Ethanol. in: Recent Progress in
Bioconversion of Lignocellulosics, pp. 163-192.
Hodge, D.B., Karim, M.N., Schell, D.J., McMillan, J.D. 2008. Soluble and insoluble solids
contributions to high-solids enzymatic hydrolysis of lignocellulose. Bioresource
Technology, 99(18), 8940-8948.
Houghton, J.T., Ding, Y., Griggs, D.J., Noguer, M., Van der Linden, P.J., Dai, X., Maskell, K.,
Johnson, C.A. 2001. Climate Change 2001: the scientific basis. Contribution of Working
Group I to the Third Assessment Report of the Intergovernmental Panel on Climate
Change (IPCC). . Cambridge University Press, UK.
Humbird, D., Davis, R., Tao, L., Kinchin, C., Hsu, D., Aden, A. 2011. Process Design and
Economics for Biochemical Conversion of Lignocellulosic Biomass to Ethanol. NREL
report (NREL/TP-5700-47764).
Ilmen, M., den Haan, R., Brevnova, E., McBride, J., Wiswall, E., Froehlich, A., Koivula, A.,
Voutilainen, S.P., Siika-Aho, M., la Grange, D.C., Thorngren, N., Ahlgren, S., Mellon,
M., Deleault, K., Rajgarhia, V., van Zyl, W.H., Penttila, M. 2011. High level secretion of
cellobiohydrolases by Saccharomyces cerevisiae. Biotechnology for Biofuels, 4.
Ingram, L.O., Aldrich, H.C., Borges, A.C.C., Causey, T.B., Martinez, A., Morales, F., Saleh, A.,
Underwood, S.A., Yomano, L.P., York, S.W., Zaldivar, J., Zhou, S.D. 1999. Enteric
bacterial catalysts for fuel ethanol production. Biotechnology Progress, 15(5), 855-866.
Jeffries, T., Grigoriev, I., Grimwood, J., Laplaza, J., Aerts, A., Salamov, A., Schmutz, J.,
Lindquist, E., Dehal, P., Shapiro, H. 2007. Genome sequence of the lignocellulosebioconverting and xylose-fermenting yeast Pichia stipitis. Nature Biotechnology, 25(3),
319-326.
Jørgensen, H., Vibe‐Pedersen, J., Larsen, J., Felby, C. 2007. Liquefaction of lignocellulose at
high‐solids concentrations. Biotechnology and Bioengineering, 96(5), 862-870.
Kabel, M.A., van der Maarel, M., Klip, G., Voragen, A.G.J., Schols, H.A. 2006. Standard assays
do not predict the efficiency of commercial cellulase preparations towards plant materials.
Biotechnology and Bioengineering, 93(1), 56-63.
Kim, S.B., Chun, J.W. 2004. Enhancement of enzymatic digestibility of recycled newspaper by
addition of surfactant in ammonia-hydrogen peroxide pretreatment. Applied Biochemistry
and Biotechnology, 113, 1023-1031.
Kim, S.B., Kim, H.J., Kim, C.J. 2006. Enhancement of the enzymatic digestibility of waste
newspaper using tween. Applied Biochemistry and Biotechnology, 130(1-3), 486-495.
204

Krishnan, C., Sousa, L.D., Jin, M.J., Chang, L.P., Dale, B.E., Balan, V. 2010. Alkali-Based
AFEX Pretreatment for the Conversion of Sugarcane Bagasse and Cane Leaf residues to
Ethanol. Biotechnology and Bioengineering, 107(3), 441-450.
Kristensen, J.B., Felby, C., Jorgensen, H. 2009. Yield-determining factors in high-solids
enzymatic hydrolysis of lignocellulose. Biotechnology for Biofuels, 2, 11.
Kumar, R., Wyman, C.E. 2009a. Does change in accessibility with conversion depend on both
the substrate and pretreatment technology? Bioresource Technology, 100(18), 4193-4202.
Kumar, R., Wyman, C.E. 2009b. Effect of xylanase supplementation of cellulase on digestion of
corn stover solids prepared by leading pretreatment technologies. Bioresource
Technology, 100(18), 4203-4213.
Kuyper, M., Hartog, M.M.P., Toirkens, M.J., Almering, M.J.H., Winkler, A.A., van Dijken, J.P.,
Pronk, J.T. 2005. Metabolic engineering of a xylose-isomerase-expressing
Saccharomyces cerevisiae strain for rapid anaerobic xylose fermentation. Fems Yeast
Research, 5(4-5), 399-409.
Kuyper, M., Winkler, A.A., Dijken, J.P., Pronk, J.T. 2004. Minimal metabolic engineering of
Saccharomyces cerevisiae for efficient anaerobic xylose fermentation: a proof of
principle. Fems Yeast Research, 4(6), 655-664.
Lau, M.W., Bals, B.D., Chundawat, S.P.S., Jin, M., Gunawan, C., Balan, V., Jones, A.D., E, D.B.
2012. An integrated paradigm for cellulosic biorefineries: utilization of lignocellulosic
biomass as self-sufficient feedstocks for fuel, food precursors and saccharolytic enzyme
production. Energy & Environmental Science, DOI: 10.1039/C2EE03596K.
Lau, M.W., Dale, B.E. 2009. Cellulosic ethanol production from AFEX-treated corn stover using
Saccharomyces cerevisiae 424A(LNH-ST). Proceedings of the National Academy of
Sciences of the United States of America, 106(5), 1368-1373.
Lau, M.W., Gunawan, C., Balan, V., Dale, B.E. 2010. Comparing the fermentation performance
of Escherichia coli KO11, Saccharomyces cerevisiae 424A(LNH-ST) and Zymomonas
mobilis AX101 for cellulosic ethanol production. Biotechnology for Biofuels, 3, 11.
Li, B.Z., Balan, V., Yuan, Y.J., Dale, B.E. 2010. Process optimization to convert forage and
sweet sorghum bagasse to ethanol based on ammonia fiber expansion (AFEX)
pretreatment. Bioresource Technology, 101(4), 1285-1292.
Lu, Y.L., Warner, R., Sedlak, M., Ho, N., Mosier, N.S. 2009. Comparison of Glucose/Xylose
Cofermentation of Poplar Hydrolysates Processed by Different Pretreatment
Technologies. Biotechnology Progress, 25(2), 349-356.
Lu, Y.P., Zhang, Y.H.P., Lynd, L.R. 2006. Enzyme-microbe synergy during cellulose hydrolysis
by Clostridium thermocellum. Proceedings of the National Academy of Sciences of the
United States of America, 103(44), 16165-16169.

205

Luo, L., van der Voet, E., Huppes, G. 2009. An energy analysis of ethanol from cellulosic
feedstock-Corn stover. Renewable & Sustainable Energy Reviews, 13(8), 2003-2011.
Lynd, L.R. 1996. Overview and evaluation of fuel ethanol from cellulosic biomass: technology,
economics, the environment, and policy. Annual review of energy and the environment,
21(1), 403-465.
Lynd, L.R., Baskaran, S., Casten, S. 2001. Salt accumulation resulting from base added for pH
control, and not ethanol, limits growth of Thermoanaerobacterium
thermosaccharolyticum HG-8 at elevated feed xylose concentrations in continuous
culture. Biotechnology Progress, 17(1), 118-125.
Lynd, L.R., van Zyl, W.H., McBride, J.E., Laser, M. 2005. Consolidated bioprocessing of
cellulosic biomass: an update. Current Opinion in Biotechnology, 16(5), 577-583.
Lynd, L.R., Weimer, P.J., van Zyl, W.H., Pretorius, I.S. 2002. Microbial cellulose utilization:
Fundamentals and biotechnology. Microbiology and Molecular Biology Reviews, 66(3),
506-577.
Madhavan, A., Tamalampudi, S., Ushida, K., Kanai, D., Katahira, S., Srivastava, A., Fukuda, H.,
Bisaria, V.S., Kondo, A. 2009. Xylose isomerase from polycentric fungus Orpinomyces:
gene sequencing, cloning, and expression in Saccharomyces cerevisiae for bioconversion
of xylose to ethanol. Applied Microbiology and Biotechnology, 82(6), 1067-1078.
Mann, L., Tolbert, V., Cushman, J. 2002. Potential environmental effects of corn (Zea mays L.)
stover removal with emphasis on soil organic matter and erosion. Agriculture Ecosystems
& Environment, 89(3), 149-166.
McBee, R.H. 1954. The characteristics of clostridium thermocellum. Journal of Bacteriology,
67(4), 505-506.
Medina, V.G., Almering, M.J.H., van Maris, A.J.A., Pronk, J.T. 2010. Elimination of Glycerol
Production in Anaerobic Cultures of a Saccharomyces cerevisiae Strain Engineered To
Use Acetic Acid as an Electron Acceptor. Applied and Environmental Microbiology,
76(1), 190-195.
Mohagheghi, A., Evans, K., Chou, Y.C., Zhang, M. 2002. Cofermentation of glucose, xylose,
and arabinose by genomic DNA-integrated xylose/arabinose fermenting strain of
Zymomonas mobilis AX101. Applied Biochemistry and Biotechnology, 98(1), 885-898.
Mohagheghi, A., Tucker, M., Grohmann, K., Wyman, C. 1992. High solids simultaneous
saccharification and fermentation of pretreated wheat straw to ethanol. Applied
Biochemistry and Biotechnology, 33(2), 67-81.
Morag, E., Bayer, E.A., Lamed, R. 1990. Relationship of cellulosomal and noncellulosomal
xylanases of Clostridium thermocellum to cellulose-degrading enzymes, Vol. 172, pp.
6098-6105.

206

Morales-Rodriguez, R., Gernaey, K.V., Meyer, A.S., Sin, G. 2011. A Mathematical Model for
Simultaneous Saccharification and Co-fermentation (SSCF) of C6 and C5 Sugars.
Chinese Journal of Chemical Engineering, 19(2), 185-191.
Ng, T.K., Benbassat, A., Zeikus, J.G. 1981. Ethanol-Production by Thermophilic Bacteria Fermentation of Cellulosic Substrates by Cocultures of Clostridium-Thermocellum and
Clostridium-Thermohydrosulfuricum. Applied and Environmental Microbiology, 41(6),
1337-1343.
Olofsson, K., Bertilsson, M., Liden, G. 2008a. A short review on SSF - an interesting process
option for ethanol production from lignocellulosic feedstocks. Biotechnology for Biofuels,
1.
Olofsson, K., Rudolf, A., Liden, G. 2008b. Designing simultaneous saccharification and
fermentation for improved xylose conversion by a recombinant strain of Saccharomyces
cerevisiae. Journal of Biotechnology, 134(1-2), 112-120.
Olsson, L., Hahn-Hägerdal, B. 1996. Fermentation of lignocellulosic hydrolysates for ethanol
production. Enzyme and Microbial Technology, 18(5), 312-331.
Olsson, L., Hahn Hägerdal, B., Zacchi, G. 1995. Kinetics of ethanol production by recombinant
Escherichia coli KO11. Biotechnology and Bioengineering, 45(4), 356-365.
Pan, X. 2008. Role of functional groups in lignin inhibition of enzymatic hydrolysis of cellulose
to glucose. Journal of Biobased Materials and Bioenergy, 2(1), 25-32.
Panagiotou, G., Olsson, L. 2007. Effect of compounds released during pretreatment of wheat
straw on microbial growth and enzymatic hydrolysis rates. Biotechnology and
Bioengineering, 96(2), 250-258.
Perlack, R.D., Wright, L.L., Turhollow, A.F., Graham, R.L., Stokes, B.J., Erbach, D.C. 2005.
Biomass as feedstock for a bioenergy and bioproducts industry: the technical feasibility
of a billion-ton annual supply. http://www.osti.gov/bridge.
Philippidis, G.P., Smith, T.K. 1995. Limiting Factors in the Simultaneous Saccharification and
Fermentation Process for Conversion of Cellulosic Biomass to Fuel Ethanol. pp. 117-124.
Pimenova, N.V., Hanley, T.R. 2004. Effect of corn stover concentration on rheological
characteristics. Applied Biochemistry and Biotechnology, 114(1), 347-360.
Poth, S., Monzon, M., Tippkotter, N., Ulber, R. 2011. Lignocellulosic biorefinery: process
integration of hydrolysis and fermentation (SSF process). Holzforschung, 65(5), 633-637.
Qing, Q., Yang, B., Wyman, C.E. 2010. Xylooligomers are strong inhibitors of cellulose
hydrolysis by enzymes. Bioresource Technology, 101(24), 9624-9630.
Reinikainen, T., Ruohonen, L., Nevanen, T., Laaksonen, L., Kraulis, P., Jones, T.A., Knowles,
J.K.C., Teeri, T.T. 1992. Investigation of the Function of Mutated Cellulose-Binding
207

Domains of Trichoderma-Reesei Cellobiohydrolase-I. Proteins-Structure Function and
Genetics, 14(4), 475-482.
Ren, C., Chen, T., Zhang, J., Liang, L., Lin, Z. 2009. An evolved xylose transporter from
Zymomonas mobilis enhances sugar transport in Escherichia coli. Microbial Cell
Factories, 8, 66-66.
Romanos, M.A., Makoff, A.J., Fairweather, N.F., Beesley, K.M., Slater, D.E., Rayment, F.B.,
Payne, M.M., Clare, J.J. 1991. Expression of Tetanus Toxin Fragment-C in Yeast - Gene
Synthesis Is Required to Eliminate Fortuitous Polyadenylation Sites in at-Rich DNA.
Nucleic Acids Research, 19(7), 1461-1467.
Rudolf, A., Alkasrawi, M., Zacchi, G., Liden, G. 2005. A comparison between batch and fedbatch simultaneous saccharification and fermentation of steam pretreated spruce. Enzyme
and Microbial Technology, 37(2), 195-204.
Ruohonen, L., Aristidou, A., Frey, A.D., Penttila, M., Kallio, P.T. 2006. Expression of
Vitreoscilla hemoglobin improves the metabolism of xylose in recombinant yeast
Saccharomyces cerevisiae under low oxygen conditions. Enzyme and Microbial
Technology, 39(1), 6-14.
Saha, B.C. 2003. Hemicellulose bioconversion. Journal of industrial microbiology &
biotechnology, 30(5), 279-291.
Schmer, M.R., Vogel, K.P., Mitchell, R.B., Perrin, R.K. 2008. Net energy of cellulosic ethanol
from switchgrass. Proceedings of the National Academy of Sciences of the United States
of America, 105(2), 464-469.
Sedlak, M., Ho, N.W.Y. 2004. Production of ethanol from cellulosic biomass hydrolysates using
genetically engineered Saccharomyces yeast capable of cofermenting glucose and xylose.
pp. 403-416.
Sendich, E., Laser, M., Kim, S., Alizadeh, H., Laureano-Perez, L., Dale, B., Lynd, L. 2008.
Recent process improvements for the ammonia fiber expansion (AFEX) process and
resulting reductions in minimum ethanol selling price. Bioresource Technology, 99(17),
8429-8435.
Shao, Q., Chundawat, S.P.S., Krishnan, C., Bals, B., Sousa, L.D., Thelen, K.D., Dale, B.E.,
Balan, V. 2010. Enzymatic digestibility and ethanol fermentability of AFEX-treated
starch-rich lignocellulosics such as corn silage and whole corn plant. Biotechnology for
Biofuels, 3.
Shao, X., Jin, M., Guseva, A., Liu, C., Balan, V., Hogsett, D., Dale, B.E., Lynd, L. 2011.
Conversion for Avicel and AFEX pretreated corn stover by Clostridium thermocellum
and simultaneous saccharification and fermentation: Insights into microbial conversion of
pretreated cellulosic biomass. Bioresource Technology, 102(17), 8040-8045.

208

Shaw, A.J., Podkaminer, K.K., Desai, S.G., Bardsley, J.S., Rogers, S.R., Thorne, P.G., Hogsett,
D.A., Lynd, L.R. 2008. Metabolic engineering of a thermophilic bacterium to produce
ethanol at high yield. Proceedings of the National Academy of Sciences of the United
States of America, 105(37), 13769-13774.
Skoog, K., Hahn-Hagerdal, B. 1990. Effect of oxygenation on xylose fermentation by Pichia
stipitis. Applied and Environmental Microbiology, 56(11), 3389.
Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., Crocker, D. 2004.
Determination of structural carbohydrates and lignin in biomass. NREL, Golden, CO.
Somerville, C., Bauer, S., Brininstool, G., Facette, M., Hamann, T., Milne, J., Osborne, E.,
Paredez, A., Persson, S., Raab, T. 2004. Toward a systems approach to understanding
plant cell walls. Science's STKE, 306(5705), 2206.
Taherzadeh, M.J., Karimi, K. 2007. Enzyme-Based Hydrolysis Processes for Ethanol from
Lignocellulosic Materials: a Review. Bioresources, 2(4), 707-738.
Tu, M.B., Chandra, R.P., Saddler, J.N. 2007. Recycling cellulases during the hydrolysis of steam
exploded and ethanol pretreated lodgepole pine. Biotechnology progress, 23(5), 11301137.
Tu, M.B., Zhang, X., Paice, M., McFarlane, P., Saddler, J.N. 2009. Effect of Surfactants on
Separate Hydrolysis Fermentation and Simultaneous Saccharification Fermentation of
Pretreated Lodgepole Pine. Biotechnology progress, 25(4), 1122-1129.
Van Zyl, C., Prior, B.A., Kilian, S.G., Kock, J.L.F. 1989. D-xylose utilization by Saccharomyces
cerevisiae. Microbiology, 135(11), 2791.
Vanarsdell, J.N., Kwok, S., Schweickart, V.L., Ladner, M.B., Gelfand, D.H., Innis, M.A. 1987.
Cloning, Characterization, and Expression in Saccharomyces-Cerevisiae of
Endoglucanase-I from Trichoderma-Reesei. Bio-Technology, 5(1), 60-64.
Vanholme, R., Demedts, B., Morreel, K., Ralph, J., Boerjan, W. 2010. Lignin biosynthesis and
structure. Plant physiology, 153(3), 895-905.
Wahlbom, C.F., Hahn-Hägerdal, B. 2002. Furfural, 5-hydroxymethyl furfural, and acetoin act as
external electron acceptors during anaerobic fermentation of xylose in recombinant
Saccharomyces cerevisiae. Biotechnology and Bioengineering, 78(2), 172-178.
Walsh, M.E., Ugarte, D.G.D., Shapouri, H., Slinsky, S.P. 2003. Bioenergy crop production in the
United States - Potential quantities, land use changes, and economic impacts on the
agricultural sector. Environmental & Resource Economics, 24(4), 313-333.
Warnick, T.A., Methe, B.A., Leschine, S.B. 2002. Clostridium phytofermentans sp nov., a
cellulolytic mesophile from forest soil. International Journal of Systematic and
Evolutionary Microbiology, 52, 1155-1160.

209

Watanabe, S., Abu Saleh, A., Pack, S.P., Annaluru, N., Kodaki, T., Makino, K. 2007. Ethanol
production from xylose by recombinant Saccharomyces cerevisiae expressing proteinengineered NADH-preferring xylose reductase from Pichia stipitis. Microbiology, 153(9),
3044.
Weber, C., Farwick, A., Benisch, F., Brat, D., Dietz, H., Subtil, T., Boles, E. 2010. Trends and
challenges in the microbial production of lignocellulosic bioalcohol fuels. Applied
Microbiology and Biotechnology, 87(4), 1303-1315.
Wyman, C.E. 2008. Cellulosic ethanol: A unique sustainable liquid transportation fuel. Mrs
Bulletin, 33(4), 381-383.
Wyman, C.E. 2007. What is (and is not) vital to advancing cellulosic ethanol. Trends in
Biotechnology, 25(4), 153-157.
Xiao, Z.Z., Zhang, X., Gregg, D.J., Saddler, J.N. 2004. Effects of sugar inhibition on cellulases
and beta-glucosidase during enzymatic hydrolysis of softwood substrates. Applied
Biochemistry and Biotechnology, 113, 1115-1126.
Yang, B., Willies, D.M., Wyman, C.E. 2006. Changes in the enzymatic hydrolysis rate of avicel
cellulose with conversion. Biotechnology and Bioengineering, 94(6), 1122-1128.
Yang, B., Wyman, C.E. 2008. Pretreatment: the key to unlocking low-cost cellulosic ethanol.
Biofuels, Bioproducts and Biorefining, 2(1), 26-40.
Yeh, A.I., Huang, Y.C., Chen, S.H. 2010. Effect of particle size on the rate of enzymatic
hydrolysis of cellulose. Carbohydrate Polymers, 79(1), 192-199.
Yu, J., Corripio, A.B., Harrison, D.P., Copeland, R.J. 2003. Analysis of the sorbent energy
transfer system (SETS) for power generation and CO2 capture. Advances in
Environmental Research, 7(2), 335-345.
Zanin, G.M., Santana, C.C., Bon, E.P.S., Giordano, R.C.L., de Moraes, F.F., Andrietta, S.R.,
Neto, C.C.D.C., Macedo, I.C., Lahr Fo, D., Ramos, L.P. 2000. Brazilian bioethanol
program. Applied Biochemistry and Biotechnology, 84(1), 1147-1161.
Zhang, M., Eddy, C., Deanda, K., Finkelstein, M., Picataggio, S. 1995. Metabolic engineering of
a pentose metabolism pathway in ethanologenic Zymomonas mobilis. Science, 267(5195),
240.
Zhang, S., Wolfgang, D.E., Wilson, D.B. 1999. Substrate heterogeneity causes the nonlinear
kinetics of insoluble cellulose hydrolysis. Biotechnology and Bioengineering, 66(1), 3541.
Zhang, Y.H., Lynd, L.R. 2003. Quantification of cell and cellulase mass concentrations during
anaerobic cellulose fermentation: Development of an enzyme-linked immunosorbent
assay-based method with application to Clostridium thermocellum batch cultures.
Analytical Chemistry, 75(2), 219-227.
210

Zhang, Y.H.P., Lynd, L.R. 2005. Cellulose utilization by Clostridium thermocellum:
Bioenergetics and hydrolysis product assimilation. Proceedings of the National Academy
of Sciences of the United States of America, 102(20), 7321-7325.
Zhao, C.X., O-Thong, S., Karakashev, D., Angelidaki, I., Lu, W.J., Wang, H.T. 2009. High yield
simultaneous hydrogen and ethanol production under extreme-thermophilic (70 degrees
C) mixed culture environment. International Journal of Hydrogen Energy, 34(14), 56575665.
Zhong, C., Lau, M.W., Balan, V., Dale, B.E., Yuan, Y.J. 2009. Optimization of enzymatic
hydrolysis and ethanol fermentation from AFEX-treated rice straw. Applied Microbiology
and Biotechnology, 84(4), 667-676.
Zhou, W., Xu, Y., Schuttler, H.B. 2010. Cellulose Hydrolysis in Evolving Substrate
Morphologies III: Time-Scale Analysis. Biotechnology and Bioengineering, 107(2), 224234.

211