ran?“

an. 3.?!

‘9:
5.2.3!
.5

dc...

(an . , . . . r: U\ 09.54.“.
, .éfin Lu”. 4.

_:;.fi

marmwfi

1 ‘
i. .. 9

‘ «a
S \ \u-uu’
.{

.
l

5.5.x; i:
L ,isfi. .
"inruar: ...
(I:

ass-Fun. an“.

4.
“an.“
{.i 931. 4": -.
[gay-hubs
1.!)
33.11%

:II
‘

‘ 3:93

1!!! 2.339.)... ‘

it VPSILKJ‘ stil
9639“”...-

c. {L3 \7 .

l . . x. . J“: A

a {39.5}!
.1: ii.($Ir$.
‘31:!) iii...
ail

\
«.1411;
5:3... .
:‘351134. 1f
I-xxgux _ . \s
(I I534:|0:v$:)9’.l| A...
1.1.1:...1!‘ X
It .3 3.13% 9.
3‘.€!rl.)i
!.£IO..
.7. .I 2.13
n (I
a

 

 

k
‘5) 221.
ill-O ‘
A .
Exitiéfixiifiki ..i.u
i». 331; . é
. 1.3 .1379 ‘1! .‘x‘yulll
s. L,.‘.\..§.flnhla ”.1 . t
‘ A 1 fl. . IOI‘ ‘.

«AK...»

..§l\
14‘ ail

TH‘EblS

#—

'DED v‘t CV:
20/) tut mud,
: ' ' . :-
Mtcnzgan State

. I .-
1 U. 9.2?“
la; «3 cu: {31:},

This is to certify that the
dissertation entitled

A NEW PARADIGM FOR CELLULOSIC ETHANOL
PRODUCTION FROM AMMONIA FIBER EXPANSION
(AFEX)-PRETREATED BIOMASS

presented by
Ming Woei Lau

has been accepted towards fulfillment
of the requirements for the

Doctoral degree in Chemical Engineering

W7, M

Major Professor’s Signature

”ML 8, ’Lalo

 

 

Date

MSU is an Affirmative Action/Equal Opportunity Employer

|---.-.-—.-.--_ — 4

PLACE IN RETURN Box to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

We ”II
x *- 1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K:IProj/Acc8Pres/ClRC/DateDue.indd

 

 

A NEW PARADIGM FOR CELLULOSIC ETHANOL PRODUCTION
FROM AMMONIA FIBER EXPANSION (AFEX)-
PRETREATED BIOMASS
By

Ming Woei Lau

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

2010

ABSTRACT

A NEW PARADIGM FOR CELLULOSIC ETHANOL PRODUCTION FROM
AMMONIA FIBER EXPANSION (AFEX)-PRETREATED CORN STOVER

By

Ming Woei Lau

Lignocellulosic ethanol can provide an environmentally-friendly alternative to
petroleum-based fuels. By utilizing locally-produced feedstocks as carbon sources for
transportation fuel, ethanol can substantially alter the energy profile of transportation
fuels, reducing geopolitical implications and dependence on foreign energy supplies.
Obtaining fermentable sugars in a cost-effective fashion is the central barrier to realizing
the commercial potential of lignocellulosic ethanol. The costs associated with
hydrolysate conditioning, nutrients, and enzymes, which were projected at 45% of the

total processing cost, must be reduced to improve the overall economics.

Feedstock pretreatment has pervasive impacts on downstream processes,
particularly on fermentation. On a comparable basis, Ammonia Fiber Expansion (AFEX)-

pretreated biomass is significantly more fermentable than that from dilute acid

pretreatment. This study confirmed that fermentation of AFEX-pretreated corn stover
using Saccharomyces cerevisiae eliminated the requirement for washing of pretreated
biomass, detoxification and nutrient supplementation. Overall ethanol yield at 191 g/kg
(64 gal/ton) corn stover was achieved at a final titer of 40 g/L. Fermentations were
completed within 72 hr in a high-cell-density fermentation. Fermentations can be
conducted with similar effectiveness with the recycled cells, for at least another three

generations without the addition of fresh cells.

The proposed integrated cellulosic ethanol production utilizes APEX-pretreated
biomass as the exclusive source of carbon, nitrogen and nutrients for ethanol
fermentation and cellulase production. The carbon (sugar) source is divided between
ethanol and enzyme production at a weight ratio of 4:1. Maximum ethanol yield under
this carbon partition scheme is 267 g EtOH/kg CS or 90 gal/ton. About 60% of the
extractable nitrogen and nutrients from AFEX-pretreated corn stover is projected to be
used for enzyme production. This in-house production would be self-sustained if 12% of
the bioavailable nitrogen source was assimilated for saccharolytic enzyme production

and to provide optimal set of activities for biomass deconstruction.

To my parents, Kai Karn Lau and Geok Lan Tey.
Their unconditional love and support made this work possible

ACKNOWLEDGEMENTS

I thank all those who counseled, encouraged, and supported me throughout this
endeavor of my Ph.D. career. I would like to specially thank my mentor, Prof. Bruce E.
Dale, for providing guidance and support in my academic and professional
developments. I am grateful to many other faculty members who offer invaluable advice
and resources to my project; they include my Ph.D. committee members, Prof. Lonnie
0. Ingram from University of Florida and Prof. Lee R. Lynd from Dartmouth College. I
also deeply appreciate all the members of Biomass Conversion Research Laboratory
(BRCL) for encouragement and support. Among them, I particularly appreciative of Dr.
Venkatesh Balan, Christa Gunawan and Ming lie Jin from the fermentation group that
worked with me on multiple projects. Thanks to Bryan D. Bals and Rebecca J. Garlock for

providing editing and peer-review assistance.

TABLE OF CONTENTS

 

 

 

 

LIST OF TABLES ........................................................................................... - - - VIII
LIST OF FIGURES ............................................................................ -- X
ABBREVIATIONS ......................................................................... - ...... XIV
CHAPTER I BACKGROUND AND GENERAL INTRODUCTION 1
WHAT IS LIGNOCELLULOSIC BIOMASS? .............................................................................. 1
WHY CORN STOVER ............................................................................................................ 2
WHY LIGNOCELLULOSIC ETHANOL? ................................................................................... 2
LIGNOCELLULOSIC BIOMASS PROCESSING ......................................................................... 3
Feedstock pretreatment — overcoming the recalcitrance of plant materials ................ 3
Bio-mediated processes — cellulase production, enzymatic hydrolysis and
fermentation .................................................................................................................. 5
Ethanologenic strains for cellulosic ethanol production ............................................... 6
Conventional approach in cellulosic ethanol production .............................................. 9
RESEARCH SCOPE AND OBJECTIVES .................................................................................. 11

CHAPTER II PRETREATMENT COMPARISON BETWEEN AFEX AND DILUTE

 

ACID PRETREATMENT - 13
INTRODUCTION ................................................................................................................. 13
MATERIALS AND METHOD ................................................................................................ 14
RESULTS ............................................................................................................................. 23
DISCUSSION ....................................................................................................................... 31
CONCLUSIONS ................................................................................................................... 35

CHAPTER III STRAIN COMPARISON FOR CELLULOSIC ETHANOL

 

FERMENTATION -- _ ..... - ..... 36
INTRODUCTION ................................................................................................................. 36
MATERIALS AND METHODS .............................................................................................. 37
RESULTS ............................................................................................................................. 40
DISCUSSION ....................................................................................................................... 49
CONCLUSIONS ................................................................................................................... 54

vi

CHAPTER IV CELLULOSIC ETHANOL PRODUCTION USING AFEX AND
SACCHAROMYCES CEREVISIAE 424A(LNH-ST) AS PLATFORM TECHNOLOGY ....56

INTRODUCTION ................................................................................................................. 56
MATERIALS AND METHODS .............................................................................................. 57
RESULTS ............................................................................................................................. 65
DISCUSSION ....................................................................................................................... 77
CONCLUSIONS ................................................................................................................... 80

CHAPTER V A NEW PARADIGM FOR CELLULOSIC ETHANOL PRODUCTION:
UTILIZATION OF AFEX- PRETREATED CORN STOVER AS SELF-SUSTAINED

 

CARBON AND NITROGEN SOURCES 82
INTRODUCTION ................................................................................................................. 82
MATERIALS AND METHODS .............................................................................................. 83
RESULTS ............................................................................................................................. 89
DISCUSSION ..................................................................................................................... 101
CONCLUSIONS ................................................................................................................. 104
CHAPTER VI GENERAL CONCLUSIONS AND RESEARCH OUTLOOK ................ 105
GENERAL CONCLUSIONS ................................................................................................. 105
RESEARCH OUTLOOK ...................................................................................................... 109

APPENDIX A EXPLORATION OF CONSOLIDATED BIOPROCESSING (CBP) ON
AF EX-PRETREATED CORN STOVER USING THERMOANAEROBACTERIUM

 

SACCHAROLYTICUM ALKZ 112
INTRODUCTION ............................................................................................................... 112
MATERIALS AND METHODS ............................................................................................ 113
RESULTS AND DISCUSSION ............................................................................................. 119
CONCLUSIONS ................................................................................................................. 127

APPENDIX B ETHANOLIC FERMENTATION OF HYDROLYSATES FROM
AMMONIA FIBER EXPANSION (AFEX) PRETREATED CORN STOVER AND

 

DISTILLERS GRAIN 129
INTRODUCTION ............................................................................................................... 129
MATERIALS AND METHODS ............................................................................................ 133
RESULTS ........................................................................................................................... 141
DISCUSSION ..................................................................................................................... 147
CONCLUSIONS ................................................................................................................. 155
REFERENCE .......... . -_ ........... 157

 

 

 

vii

Table 1

Table 2

Table 3

Table 4

Table 5

Table 6

Table 7

Table 8

Table 9

LIST OF TABLES

Typical biomass constituents in the energy crops ................................................ 2

Summary for the features and genetic engineering approaches for selected
cellulosic ethanol producer .................................................................................. 8

Seed culture media recipe for the three ethanologenic strains ......................... 39
Results summary for fermentation using E. coli K011, Z. mobilis AX101 and S.
cerevisiae 424A(LNH-ST) in 2% w/w corn steep liquor (CSL) or yeast extract
peptone (YEP) ..................................................................................................... 43

Qualitative summary of the relative fermentation performance of Z. mobilis
AX101, S. cerevisiae 424A(LNH-ST) and E. coli K011 .......................................... 46

Summary of parameters for four experiments examining the effects of pH,
temperature, initial cell density, and washing and nutrient supplementation on
xylose fermentation of corn stover hydrolysate ................................................ 61
Experimental details on the construction of mass balance ................................ 64

Summary of fermentation parameters for complex media and hydrolysates 70

Output sugars analysis from enzymatic hydrolysis of 6.0% cellulose loading
AFEX-CS (after normalization) ............................................................................ 76

Table 10 Future projections for ethanol yield and titer based on 6% cellulose loading of

AF EX-CS ............................................................................................................... 79

Table 11 Amino acid concentration of AFEX-water extract (16.7% Solids Loading) and

AFEX corn stover enzymatic hydrolysate (18.0% solids loading) ....................... 96

Table 12 Trace elements and vitamins in AFEX-water extract (16.7% solids loading) and

AFEX corn stover enzymatic hydrolysate (18.0% solids loading) ....................... 97

viii

Table 13 Total projected carbon and nitrogen entering enzyme production ............... 100
Table 14 MTC Media for T. saccharolyticum ALK2 Growth ........................................... 116
Table 15 Concentration of nitrogenous compounds in commercial enzymes .............. 123

Table 16 Summary of fermentation parameters by ethanologenic strains E. coli K011
and ML01 .......................................................................................................... 144

LIST OF FIGURES

Figure 1 Schematic illustration of biological-mediated events in various biomass
processing strategies ............................................................................................ 5

Figure 2 Conventional cellulosic ethanol production approach adopted by the National
Renewable Energy Laboratory (NREL) .................................................................. 9

Figure 3 Operating cost distribution for the conventional cellulosic ethanol production
projected by NREL .............................................................................................. 11

Figure 4 Schematic describing the preparation of AFEX-pretreated corn stover water
extract ................................................................................................................. 17

Figure 5 Mass balance comparison between AFEX and dilute acid pretreatment on corn
stover (A) AFEX pretreatment and (B) dilute acid pretreatment. ...................... 24

Figure 6 Cell density after 24 hr of fermentation in YNB-supplemented water extracts
from dilute acid pretreatment at different solids loading (5%, 7.5%, 30%[Sund])
and AFEX pretreatment. (A) 424A(LNH-ST) and (B) K011 .................................. 26

Figure 7 Sugar consumption after 24 hr of fermentation in YNB-supplemented water
extracts from dilute acid pretreatment at different solids loading (5%, 7.5%,
30%[Sund]) and AFEX pretreatment in water extract at 7.5% solids loading
equivalent (A, C) and 15.0% solids loading equivalent (B, D). ........................... 27

Figure 8 Time course of fermentation of water extract from 7.5% solids loading
equivalent of AFEX and dilute acid pretreated corn stover. .............................. 28

Figure 9 Enzymatic hydrolysis yield on the water-insoluble AFEX-pretreated corn stover
and dilute acid pretreated corn stover at 5.1% glucan loading, pH 4.8 and 50°C.
............................................................................................................................ 29

Figure 10 Mass balance and energy content for non-carbohydrate insoluble solids after
pretreatment, washing and enzymatic hydrolysis. ............................................ 30

Figure 11 Fermentation using E. coli K011, Z. mobilis AX101 and S. cerevisiae 424A(LNH-
ST) in 2% w/v corn steep liquor (CSL) with (A) glucose as sole carbon source, (B)
glucose and xylose mixture at ratio 7:3 and (C) xylose as sole carbon source. . 42

Figure 12 By-products concentration during (A) glucose fermentation at 72 hr, (8) co-
fermentation at 168 hr and (C) xylose fermentation at 168 hr, in CSL media
using E. coli K011, Z. mobilis AX101 and S. cerevisiae 424A(LNH-ST). .............. 47

Figure 13 The effect of water-soluble compounds from AFEX-treated corn stover on (A)
cell growth (B) percent glucose consumption and (C) percent xylose
consumption after 24 hr fermentation. ............................................................. 48

Figure 14 Fermentation using E. coli K011 on YEP media supplemented with
commercial enzymes, 15% solids loading equivalent of AFEX-pretreated corn
stover water extract or a combination of commercial enzymes and the water
extract ................................................................................................................. 49

Figure 15 Fermentation using (A) Z. mobilis AX101, (8)5. cerevisiae 424A(LNH-ST), (C) E.
coli K011, in enzymatic hydrolysate from 6.0% glucan loading of AFEX-
pretreated corn stover. ...................................................................................... 51

Figure 16 Effect of (A, 8) pH, (C) temperature and (0) initial 00 on fermentation using
hydrolysate from enzymatically-digested AFEX-treated CS using Saccharomyces
cerevisiae 424A(LNH-ST) ..................................................................................... 66

Figure 17 Fermentation of (A) complex media; (B) CS-Hydrolysate I; (C) CS-Hydrolysate ll
using S. cerevisiae 424A(LNH-ST). ...................................................................... 68

Figure 18 Effect of washing and nutrient supplementation on xylose consumption in the
fermentation of hydrolysates from enzymatically-digested AFEX-treated CS... 69

Figure 19 Cell density profile of the fermentation using 424A(LNH-ST) ......................... 71
Figure 20 Effect of AFEX-CS water extract concentration on cell growth, metabolic yield
and specific xylose consumption rate under (A) nutrient-rich and (B) nutrient-
Iimiting conditions. ............................................................................................. 73
Figure 21 Corn stover to ethanol mass balance analysis. ................................................ 75
Figure 22 AFEX-corn stover hydrolysate fermentation. (A) fermentation at high cell

density using cell recycle and (B) xylose consumption over four generation
without fresh cell addition. ................................................................................ 91

xi

Figure 23 Sugar yield after enzymatic hydrolysis using fermentation broth from
Trichoderma reesei RUT-C30. ............................................................................. 92

Figure 24 The integrated biomass processing and bioconversion platform using AFEX-
pretreated corn stover as exclusive carbon, nitrogen and nutrient source. ..... 98

Figure 25 Proposed carbon and nitrogen flow for core and complementary units. (A)
flow diagram for the proposed unit operations, (B) carbon and nitrogen flow
for each stream .................................................................................................. 99

Figure 26 Total projected carbon and nitrogen entering enzyme production .............. 100

Figure 27 Fermentation using S. cerevisiae 424A(LNH-ST) in (A) laboratory media, (8)
18% w/w AFEX-CS liquid hydrolysate (solid-free) and (C) 18% w/w AFEX-CS
hydrolysate (slurry, SSCF) ................................................................................. 111

Figure 28 Batch fermentation of AFEX-CS hydrolysate at 3.6% solids loading equivalent
using T. saccharolyticum ALK2.(A) fermentation profile and (B) initial and final
sugar composition. ........................................................................................... 120

Figure 29 Fedbatch fermentation of AFEX-pretreated corn stover using T.
saccharolyticum ALK2. (A) initial (if no fermentation occurs) and final sugar
concentration with rich and lean nutrient supplementation, (B) fermentation
using rich nutrient supplementation and (C) fermentation using lean nutrient
supplementation. ............................................................................................. 122

Figure 30 Sugar composition of enzymatic hydrolysate from AFEX-pretreated corn
stover utilizing Accelerase at 5.6 (EH1) and 3.8 (EH2) mg protein/g biomass and
a combination of four complex enzymes (Spezyme CP, Novozyme 188,
Multifect Xylanase and Multifect Pectinase; EH3). .......................................... 125

Figure 31 Sugar consumption pattern for (A) glucose and (B) xylose fermentation in
water extract of AFEX-pretreated corn stover at varying solids loading
equivalent ......................................................................................................... 127

Figure 32 (A) Fermentation of Hydrolysate A (6% cellulose loading CS-hydrolysate) by E.
coli K011, (B)Fermentation of Hydrolysate A by E. coli ML01,(C) Fermentation
of Hydrolysate B (7:10 dilution of Hydrolysate A) by E. coli ML01, and (0) Time
course of cell density during fermentation in Hydrolysate A and B ................ 146

Figure 33 (A) Time course of cell density during fermentation in Hydrolysate C (3%
cellulose loading DDGS hydrolysate) and D (DDGS hydrolysate with sugars
addition), (B) Fermentation of Hydrolysate C by E. coli ML01, (C) Fermentation
of Hydrolysate D by E. coli ML01. .................................................................... 149

xii

Figure 34 (A) Time course of cell density during fermentation in Hydrolysate E (Mixture
of CS and DDGS hydrolysates at ratio 7:3 ), (B) Fermentation of Hydrolysate E
by E. coli K011, (C) Fermentation of Hydrolysate E by E. coli ML01. ............... 150

Figure 35 (A) Time course of cell density, (B) Glucose consumption profile in
fermentation of corn stover hydrolysates with or without nutrient
supplementation. ............................................................................................. 153

Figure 36 DDGS to ethanol analysis. DDGS was AFEX-pretreated and enzymatically

hydrolyzed using Spezyme CP and Novozyme 188 for three days before
fermentation using adapted strain E. coli ML01 .............................................. 156

xiii

AFEX
ATCC
CBP
CSL
CS
Conc
(NC
AASU
NREL
(DD
SLE
SHF
SSCF
SSF
XyI
YEP

YNB

ABBREVIATIONS

Ammonia Fiber Expansion

American Type Culture Collection

Consolidated Bioprocessing

Corn Steep Liquor

Corn Stover

Concentration

Glucose

Michigan State University

National Renewable Energy Lab

Optical Density

Solids Loading Equivalent

Separate Hydrolysis and Fermentation
Simultaneous Saccharification and Co-Fermentation
Simultaneous Saccharification and Fermentation
Xylose

Yeast Extract and Peptone

Yeast Nitrogen Base

xiv

CHAPTER I BACKGROUND AND GENERAL INTRODUCTION

WHAT IS LIGNOCELLULOSIC BIOMASS?

Lignocellulosic biomass, a mix of complex organic polymers, consists of cellulose,
hemicellulose, lignin, and other extractive (phenolics and proteins) and non-extractive
components. Cellulose is a linear polymer of B-1,4-D-glucopyranose units with a highly
ordered and crystallized structure [1]. Hemicellulose, a branched carbohydrate, consists
of several types of sugars including xylose, glucose, arabinose, galactose and mannose.
Xylose is the predominant sugar in hemicellulose for most herbaceous biomass. Lignin is
a highly cross-linked phenylpropylene polymer which is covalently bonded to
hemicellulose [2]. It provides structural integrity to plant materials and protects plant
carbohydrates from hydrolytic activity. Typical composition of lignocellulosic biomass is

as listed in Table 1.

Due to the abundance and biorenewability of lignocellulosic materials, the ability to
utilize these materials as the carbon sources would dramatically reduce the cost for raw

materials and the carbon footprints of the chemicals produced.

Table 1 Typical biomass constituents in the energy crops

 

 

 

 

 

 

 

 

 

 

 

 

 

Cellulose Xylan Lignin Protein Reference

Corn Stover 37.5 22.4 17.6 3.0 [3]
Switchgrass 32.2 22.3 17.3 3.0 [4]
Reed 26.5 16.3 14.8 4.5 [4]
Canarygrass

Bagasse 44 22 22 2 [5]
Wheat Straw 31 20 25 - [6]
Poplar 47.4 16.3 31.9 - [7]

 

Unit: Percentage of the total dry biomass

WHY CORN STOVER

Corn stover, an agricultural residual from the existing farming operation, can be

harvested at 80-100 million dry tonnes per year in a sustainable fashion [8]. Corn stover

is also the most abundant existing agricultural residue in the United States and its

production is concentrated in the Midwestern region [8]. Effective conversion of this

lignocellulosic feedstock to ethanol could double the total ethanol production from the

corn industry. Due to the similarity of herbaceous biomass in term of composition,

understanding on how to process corn stover will also be applicable to other important

energy crops such as switchgrass.

WHY LIGNOCELLULOSIC ETHANOL?

Ethanol, in the search for a sustainable source for transportation fuel, has emerged as

the leading alternative to replace petroleum. Currently, ethanol is produced from edible

crops such as corn and sugarcane. However, apart from competing with food supply,

these feedstocks may not have the potential to substantially alter the energy profile in

 

the transportation sector due to the scale of production required [9]. Lignocellulosic
feedstocks, the non-edible part of plant materials, are abundant carbohydrate sources
and available at a competitive cost [10]. Thus, the capability to utilize plant materials for

ethanol production could substantially reduce global dependence on fossil fuels.

LIGNOCELLULOSIC BIOMASS PROCESSING

Feedstock pretreatment - overcoming the recalcitrance of plant materials
Lignocellulosic biomass is inherently recalcitrant primarily due to the protection of
cellulose by lignin, high crystallinity of cellulose, low accessible surface area and
cellulose sheathing by hemicellulose [11, 12]. Thermochemical pretreatments are
commonly used to increase susceptibility of plant carbohydrate for enzymatic hydrolysis
[2]. Several features have been identified to be important for feedstock pretreatment,

including the following:

0 Effective in creating reactive fiber

0 High hemicellulose recovery

0 Low levels of inhibitor production

0 Low requirement for particle size reduction
0 Low cost for construction materials

Various chemicals such as sulfuric acid, sulfur dioxide, ammonia, sodium hydroxide and
lime have been investigated as the potential catalyst for pretreatment. The effects
commonly found in thermochemical pretreatments include increase in accessible

surface area, decrystallization of cellulose, removal of hemicellulose and lignin removal

3

[13]. Among them, dilute acid pretreatment (using sulfuric acid as catalyst) and
ammonia fiber expansion (AFEX) are two important feedstock pretreatment methods

due to their ability to generate highly reactive biomass [14].

Dilute acid pretreatment has been extensively investigated for the past two decades.
Acid hydrolyzes hemicellulose and other extractives in the plant materials, leaving
cellulose in the remaining water—insoluble solids. Although effective at breaking down
hemicellulose, acid-catalyzed reactions on these complex materials have significant
drawbacks; approximately 10-30% of the hemicellulose is degraded to non-sugar
byproducts [15]. These degradation products from sugars and lignin are oftentimes
inhibitory toward saccharolytic enzymes and fermentation [16—19]. The primary
degradation products found in the hydrolysate are furfural, 4-hydroxymethylfurfural,
and acetic acid [17]. Detoxification through calcium hydroxide treatment is incorporated
with cellulosic ethanol production using dilute acid as the feedstock pretreatment

platform [20, 21].

Ammonia fiber expansion (AFEX) uses concentrated ammonia at temperatures from 90
to 150°C and explosive decompression to disrupt the lignocellulosic structure. AFEX is a
dry-to-dry process due to low water use and the volatility of anhydrous ammonia at
normal conditions [22]. This facilitates ethanol production at higher concentration,
reducing water use and energy input during distillation. Furthermore, AFEX-pretreated
feedstocks are highly fermentable presumably due to reduced production of inhibitory

compounds and preservation of nutrient content during the pretreatment [2].

Bio-mediated processes — cellulase production, enzymatic hydrolysis and fermentation
Cellulose utilization and ethanol fermentation are the two important processes in the
bio-mediated conversion for cellulosic ethanol production. Cellulose utilization involves
the breakdown of the carbohydrates to fermenting sugars and the sugars are converted

to ethanol through microbial catalysts during fermentation.

These processes are envisioned to be conducted in one of three configurations i.e. (i)
separate hydrolysis and fermentation (SHF), (ii) simultaneous saccharification and (co)-
fermentation (SSF/SSCF) and (C) consolidating bioprocessing (CBP) [2]. These

configurations are described in Figure 1.

 

Biologically-mediated
events

SHF SSF SSCF CBP

 

Cellulase Production

Cellulose Hydrolysis

Hexoses Fermentation

 

Pentoses Fermentation

 

 

 

 

Figure 1 Schematic illustration of biological-mediated events in various biomass
processing strategies

Note: Modified from [2]. SHF: Separate hydrolysis and fermentation, SSF: Simultaneous
saccharification and fermentation, SSCF: Simultaneous saccharification and
cofermentation, CBP: Consolidated bioprocessing.

SHF allows hydrolysis and fermentation to be conducted in their respective optimal
conditions. The optimal pH and temperature for celluloytic enzymes fall in the ranges of
pH4-5 and 50-60°C, respectively. However, conventional ethanologenic strains are
mesophiles which ferment between 25 and 35°C. The significant challenge of SHF
compared to SSF is overcoming sugar inhibition[23]. Cellulases and hemicellulases are
reported to be inhibited by the simpler forms of sugars (end-product inhibition) which
reduce fermentable sugar yield [24]. The effect of sugar inhibition is particularly

noticeable when the solid loading is greater than 10%.

Both SHF and SSCF require addition of exogenous cellulase. Due to the complex nature
of lignocellulosic biomass, a wide spectrum of enzymatic activity includes
endoglucanase, exoglucanase, B-glucosidase, endoxylanase, exoxylanase, B-xylosidase
and a-arabinofuranosidase are required for biomass deconstruction[25]. The cost of
these enzymes remains the primary challenge to the overall economics of cellulosic
ethanol production. Hence, development of strains that are able to effectively saccharify
plant carbohydrates and convert fermentable sugars to ethanol in the configuration of
consolidated bioprocessing (CBP) has been pursued. Mature technology of CBP is

projected to reduce the cost of biological processing by 78% compared to SSCF [26].

Ethanologenic strains for cellulosic ethanol production
The ethanol fermentation using lignocellulosic biomass must achieve high overall yield,
titer and productivity. Relative to corn ethanol industry, co-fermentation of hexoses and

pentoses in a plant hydrolysate is essential in cellulosic ethanol production.

6

Lignocellulosic hydrolysate that contains degradation compounds from pretreatment
exposes the fermenting strain to potential inhibitory chemicals that can interfere with
cellular activities. Hence, an ethanologenic strain that is relevant for commercial

production must have the following features [27]:

o Ethanol yield higher than 90% based on total plant carbohydrate

0 Final ethanol concentration greater than 40 g/L

0 Overall productivity greater than 1.0 g/L/hr, respectively

0 Robust growth in lignocellulosic hydrolysate

0 Simple nutrient requirement for growth
Besides traditional brewer’s yeast (Saccharomyces cerevisiae), there are several known
ethanologenic strains that have advantageous features as cellulosic ethanol producers.
Zymomonas mobilis produces ethanol from hexoses at higher metabolic yield,
concentration (>100 g/L) and specific rate than S. cerevisiae due to its ability to utilize
glucose anaerobically through the Enter-Doudoroff pathway [27]. Escherichia coli K12
utilizes both hexoses and pentoses to produce ethanol. However, the native abilities of
the ethanologens are limited particularly regarding overall ethanol yield. Two general
metabolic engineering strategies were exploited for strain development, (i) expand
fermentable sugar of ethanologens to include pentose sugars and (ii) improve metabolic
ethanol yield by minimizing byproduct formation. The summary of the genetic

engineering approach of three promising ethanologens is presented in Table 2.

Table 2 Summary for the features and genetic engineering approaches for selected
cellulosic ethanol producer

 

E. coli K011

Z. mobilis AX101

S. cerevisiae
424A(LNl-l-ST)

 

Advantageous
Features (of
Native Strain)

Ability to utilize
hexoses and
pentoses

Ability to produce
ethanol at near
theoretical
maximum

-- Ability to produce
ethanol from
hexose sugars at
near theoretical
maximum

-- The value of yeast
cells as coproduct
of fermentation

 

Disadvantageous
Features (of
Native Strain)

-- Produce ethanol at
25% of the maximum
yield

-- Relative low
ethanol tolerance

-- Inability to utilize
pentose sugar such
as xylose and
arabinose

-- Inability to utilize
pentose sugar such
as xylose and
arabinose

 

 

 

 

 

 

 

Metabolic -- Genes encoding -- Genes encoding -- Gene cloning
Engineering for pyruvate for xylose multiple copies of
Approach decarboxylase and isomerase, xylose reductase,
alcohol xylulokinase, xylose
dehydrogenase || arabinose ‘ dehydrogenase and
from Zymomonas isomerase, xylulokinase into
mobilis was ribulokinase, the yeast’s
heterologously ribulase, 5- chromosome
expressed in E. coli phosphate 4
K12 epimerase,
-- The succinic acid transaldolase and
pathway was transketolase from
disrupted through E. coli were cloned
gene knockout and expressed in
Zymomonas mobilis
Improvements Ethanol is produced Both xylose and Xylose can be
at near theoretical arabinose can be utilized for ethanol
maximum, 4 fold utilized for ethanol production at 90%
increase from the production of the theoretical
native capability maximum
Reference [28] [29] [30-32]

 

 

Conventional approach in cellulosic ethanol production

Biomass processing and conversion technologies centered on dilute acid pretreatment
involve (i) cellulose solids and hemicellulose hydrolysate separation; (ii) detoxification
unit of the hemicelluloses hydrolysate and (iii) nutrient supplementation to improve

general fermentability of the acid-pretreated materials as shown in Figure 2[33].

Enzymes
Nutrients Yeast

 
     
 

 

" '..‘~~.' -‘-
‘lr'l .I,‘ '
.fI ~-.'

Dilute Acid ' 5‘ Stream If];

  
 

:' 3*M.‘7“Ifi . ..
-."-. a»
, , Ethanol
F“

 

 

 

 

 

   
 
   

 

 
    

Mmetment Separation 3‘ :
"I ‘ . .1 ' [Pentosei Ethanol
Detoxification -

, _ _ 3 -"?ii*:‘r.i;1‘j"'._:f?

Nutrients Yeast

Figure 2 Conventional cellulosic ethanol production approach adopted by the National
Renewable Energy Laboratory (NREL)

 

 

In an industrial setting, countercurrent washing has been proposed to separate the
solids and liquid streams. This separation method introduces a greater water use [34].
The pretreatment must also be conducted at a greater solids loading to achieve a given

final titer due to the introduction of the additional water wash stream.

Degradation products from acid pretreatment such as furfural, 4-hydroxymethyfurfural
and acetic acid are shown to be inhibitory toward microorganisms. Removal of these
inhibitors through physical, chemical and biological interactions has been extensively
investigated [17, 35, 36]. Detoxification through calcium hydroxide (overliming) or
activated carbon was reported to be effective [20, 37]. However, these detoxification
methods often remove sugars and reduce the nutrient content of the plant hydrolysate.

Detoxification, if applied, is estimated to comprise 20% of the total processing cost [38].

Nutrient supplementation is generally regarded as the integral part of cellulosic ethanol
production. Expensive complex nutrients such as yeast extract are not relevant as a
supplement for ethanol production. Corn steep liquor (CSL), a co-product from corn
ethanol industry, is regarded as a relatively economical source of nutrients for cellulosic
ethanol fermentation [39]. The cost of commercial nutrients is expected to be at 5-8
cents/gal ethanol [39]. The cost of cellulase, detoxification through ammonium
hydroxide, and nutrient supplementation using corn steep liquor consists of 45% of the
total operating cost (Figure 3) from the economic projection by the National Renewable

Energy Lab [40].

10

 

Electricity

 

 

—- Lf_1 . «I

Waste Disposal

 

 

 

 

 

 

 

Other Raw Materials

._._.___.___..v, ‘

 

Ammonia

 

 

 

Sulfuric Acid

Cellulase

 

_._..._._ .._ .._..— .‘

CSL (Nutrients)

 

Operating Cost Percentage
(cents/gal EtOH) (%)

Feedstock 90 57.29%
15.6 9.93%
Cellulase 35 22.28%
Sulfuric Acid 4.1 2.61%
T Ammonia 20.7 13.18%
Other Raw 2.8 1.78%
21 134%
Electricity -13.2 -8.40%

 

 

 

 

 

 

 

 

 

Feedstock

 

 

 

 

 

Figure 3 Operating cost distribution for the conventional cellulosic ethanol production

projected by NREL

17 r j T I I I I I

20

40 60 80

Operating Cost (cents/gal EtOH)

Note: Modified from[40]. Ethanol yield = 72.6 gal/ton; Assumed feedstock cost =

USD65.3/dry ton.

RESEARCH SCOPE AND OBJECTIVES

My work addresses applied and fundamental understanding on overall lignocellulosic

processing from an integrated perspective

0 Comparison of feedstock pretreatments and ethanologenic microorganisms for

lignocellulose bioprocessing

0 Investigation of cellulosic ethanol production using AFEX as pretreatment and

S. cerevisiae 424A(LNH-ST) as ethanologenic strain

0 Comprehensive nutrient content analysis of plant materials hydrolysate

Ultimately, this research seeks to demonstrate a novel and straightforward paradigm for
cellulosic ethanol production centered on AFEX pretreatment to reduce the overall

processing cost.

12

CHAPTER II PRETREATMENT COMPARISON BETWEEN AFEX AND DILUTE

ACID PRETREATMENT

INTRODUCTION

Feedstock pretreatment in cellulosic ethanol production plays an integral role in
lignocellulosic biomass processing due to the inherent recalcitrance of plant material
[1]. The selection of pretreatment method has far-reaching impacts on the overall
process, including feedstock handling, biological conversions, and downstream

processing [41].

Among pretreatments, dilute acid pretreatment and AFEX are regarded as promising
candidate for large scale cellulosic biofuels production. High sugar recovery can be
achieved by both pretreatments for corn stover from a previous comparative study[42].
However, a comprehensive comparison of pretreatments concerning their impacts on
fermentation is needed. In this chapter, the impacts of these two pretreatments from a
system-wide perspective are examined. | evaluate the interactions of dilute acid

pretreatment and AFEX with enzyme requirements, hydrolysate fermentability and

13

lignin preservation. The microbial platform used for the pretreatment comparison

involves Saccharomyces cerevisiae 424A(LNH-ST) and Escherichia coli K011.

MATERIALS AND METHOD

Corn stover (CS)

Corn stover was supplied by the National Renewable Energy Laboratory (NREL, Golden,
Colorado). It was milled and passed through a 4 mm screen. The moisture content was
approximately 7% (total weight basis). The milled corn stover was kept at 4°C for long
term storage. This corn stover contains 34.1% cellulose, 20.4% xylan, 3.3% arabinan and

2.3% protein on a dry weight basis.

Dilute acid pretreated corn stover from pilot-scale continuous (Sund) reactor at NREL
This dilute acid pretreatment has been carried out as described previously [15].
Pretreatment was conducted at 190°C. The solids and sulfuric acid loading of the
pretreatment were reported as 30% (w/w) and 0.048 g/g dry corn stover, respectively.

The whole slurry from the reactor was used in this study.

Pretreatment

Ammonia fiber expansion (AFEX)

14

The AFEX pretreatment was performed in a 2.0 L pressure vessel (Parr Instruments,
Moline, IL). The reactor was equipped with thermocouples and a pressure sensor. AFEX
on CS was conducted at 62.5% solids loading. The reactor was preheated to 100-110°C
and prewetted corn stover (150 g dry CS + 90 g distilled water) was loaded into the
vessel. The lid was bolted shut. Anhydrous ammonia (150 g) was preheated in a 500 mL
stainless steel cylinder (Parker Instrumentation, Jacksonville, AL) until its pressure
reached 4.48 MPa (650 psi). Heated ammonia was then transferred into the reactor to
initiate the reaction. The initial and final temperatures of the pretreatment were
130iS°C and 110:5°C, respectively. The reactor pressure was released after 15 min
through an exhaust value. AFEX-pretreated CS was then air-dried in a fume hood

overnight.

Bench-scale dilute acid pretreatment

The dilute acid pretreatment was performed with a 1.0 L Parr reactor made of Hastelloy
C (Parr Instruments, Moline, IL) equipped with a thermocouple (Extech Instruments,
Waltham, MA) and a helical impeller (3.5 in.) on a two-piece shaft. The impeller was
driven by a variable speed DC motor assembly (Parr Instruments, Moline, IL). Corn
stover was presoaked in 1.0% w/v dilute sulfuric acid solution at 5.0% and 7.5% solids
(w/w) overnight. Total weight of the pretreatment mixture was 800 g. The presoaked
slurry was transferred into the reactor which was then sealed and fitted to the impeller
driver motor. The impeller speed was set at 150 rpm. The reactor was heated rapidly to

and maintained at 14012°C in about 2 min and maintained at this temperature in a sand

15

W

3F

Er

TE:

bath for 40 min. At the end of the reaction time, the reactor was cooled to below 50°C.
The diluted acid pretreated corn stover slurry was filtered through Whatman no. 1 filter

paper. Details on the apparatus and experimental procedure are as described [43]

Fermentation on water extract of soluble compounds from pretreated CS

Water extract preparation

Four water extract of pretreated corn stover was prepared for fermentation studies, (i)
AFEX-CS pretreated at 62.5% solids loading, (ii) dilute-acid-CS pretreated at 5.0% solids
loading in the benchscale reactor, (iii) dilute-acid-CS pretreated at 7.5% solids loading in
the benchscale reactor and (iv) dilute—acid-CS pretreated at 30.0% solids loading in

continuous pilot reactor (Sund)

AFEX-pretreated corn stover was washed with distilled water at a ratio of 1 g dry CS to 5
mL of water to produce a water extract (20% solids loading equivalent). In each batch of
washing, distilled water was preheated to 60—70°C and added to 100 g (dry weight
equivalent) of AFEX-CS. The water content of the wetted AFEX-CS was reduced by using
an in-house manufactured press. The washing was conducted in three cycles, i.e. water-
extract from a previous cycle of washing was used for the next cycle of washing (Figure
4). In the final cycle of washing, the moisture content of the washed AFEX-CS was

reduced to 77i3%. The AFEX-CS water extract was used for the fermentation.

16

AFEX-CS 1

Distilled
Water

Water
Preheating to
60-70°C ‘
500 mL
Yflirlled Water

Mixing

. Wm” .. I z. 1,, 1‘
. Washing and (
Press at 80kPa 1.
(Washed Solids

~350 mL:
Wash ‘
Stream I

   
  
 

   

Stream
Recycle 52

   

,1. 7,, 1, 7.

i20% Solids Loading]

Equivalent Wash I
Stream
L_4

Figure 4 Schematic describing the preparation of AFEX-pretreated corn stover water

Hemicellulose hydrolysate from the dilute acid pretreatment stage was used as the

water extract. Hemicellulose hydrolysates prepared from 5% and 7.5% solids loading

during bench-scale dilute acid pretreatment were concentrated to 20% solids loading

equivalent (1 g input CS in 5 mL liquid) through rotary evaporation under vacuum at

17

 

 

75°C.

achie

VVherI

respec

ForCS
nfl.oil
wasrni
sohdslr
avaHabl
lnsolubl

benchsc

Seedcul
Seedtul
pmpmec
DEptone
3” Initial.

KOILSOC

75°C. Xylose concentration was used as the indicator for the concentration factor

achieved during the evaporation (Eq. 1).

X l
(Solids Loading)f =-[D(—y']]ix(Solids Loading)i ---- Eq. 1

I
Where [Xyl], fand i denote for concentration of xylose, final and initial condition,

respectively.

For CS from the Sund reactor, distilled water was added so that the mixture contained 5
mL of liquid to 0.51 g of dry water-insoluble pretreated corn stover. The diluted slurry
was mixed by rigorous shaking and centrifuged at 6000 xg. The supernatant was as 20%
solids loading equivalent. No mass balance around Sund pretreatment was made
available, therefore it was assumed that percentage of input CS remaining as water-
insoluble-solids after the pretreatment in the Sund reactor is the same as that of

benchscale dilute acid pretreatment, i.e. 51%.

Seed culture preparation

Seed cultures of E. coli K011, S. cerevisiae 424A(LNH-ST) and Z. mobilis AX101 were
prepared in 100 mL of complex media YEP_GX (5 g/L bacto yeast extract + 10 g/L bacto
peptone + 30 g/ L glucose + 20 g/L xylose) by inoculating frozen (-80°C) culture stock at
an initial cell density of 0.1 unit OD600nm. Culture temperature and period of E. coli

K011, Saccharomyces cerevisiae 424A(LN H-ST) and Zymomonas mobilis AX101 are 37°C,

18

 

18hr; 3
anaero

runate

Ftnnen
Fennen
eancts
tornsto
lose.CA
ammonII
xwosein
leSpchv
sohdloac

cOntamm

EathlveH
to aid Stir
cuhureym
mICIOPIaII
A” Openinl
PIOduted,

MEWS/'09 4

18hr; 30°C, 18hr; 30°C, 42hr, respectively. The cultures were conducted under largely
anaerobic conditions and mixed at 150 rpm agitation. The grown cells were used to

initiate fermentations.

Fermentation procedure--microplate fermentation

Fermentations of E. coli K011, S. cerevisiae 424A(LNH-ST) and Z. mobilis AX101 in water
extracts of 7.5% and 15.0% solids-loading-equivalent of the three types of pretreated
corn stover were conducted in 24-well cell culture microplates (BD Falcon #353047, San
Jose, CA). The media were supplemented with water extract, yeast base nitrogen with
ammonium sulfate (YNB, MP Biomedicals, Lot Number: 4027512-119914), glucose and
xylose in appropriate buffer (SOmM) at final concentrations of 16.7 g/L, 9 g/L and 35 g/L,
respectively. Distilled water was added to dilute the water extracts to 7.5% and 15.0%
solid loading equivalent. Chloramphenicol (50 mg/L) was added to reduce risk of

contamination.

Each well contained 2.0 mL media and was added with a glass bead (6 mm in diameter)
to aid stirring. Seed cultures were prepared as described above and the microplate cell
culture was initiated at 00(600nm) of 0.5. The microplate was sealed and fixed on the
microplate clamp system (Applikon Inc, Springfield, IL) in an incubator shaker (150 rpm).
An opening (about 1mm diameter) was made on the seal to vent carbon dioxide
produced. Initial pH for E. coli K011 was at 7.0 and at 5.5 for Z. mobilis AX101 and S.

cerevisiae 424A(LNH-ST). Incubation temperature was the same as seed culture

19

 

 

 

cond
424A
sped
HPLC

in the

ibnne
Ferme
extract
extract
10 EM
wasadc
conduct
was pH.

dESIgn31

Enzymat
Enzyma)
To Diepa
Washed I“
Water. FOr

conditions. The fermentations were conducted for a designated period (K011,
424A(LNH-ST): 24hr; AX101: 48hr). Cell density was measured using a
spectrophotometer at 600nm. Sugars and fermentation products were analyzed using a
HPLC system with a Biorad Aminex HPX-87H colume as described [3]. Error bars shown

in the results are standard deviations oftriplicates.

Fermentation procedure--shake flask fermentation

Fermentations of K011 and 424A(LNH-ST) were further conducted in shake flask. Water
extract from AFEX and dilute acid pretreatment were supplemented with 1 g/L yeast
extract and 2 g/L peptone with appropriate buffer. Sugar levels were adjusted to about
10 g/L glucose and 50 g/L xylose. Final solids loading equivalent was at 7.5%. lnoculum
was added to achieve an initial cell density of 0.1 OD600nm. Fermentation was
conducted at the strains’ respective optimal pH and temperature. K011 fermentation
was pH-adjusted every 24hr using 6M KOH. Fermentation samples were taken at

designated period throughout the 120hr culture.

Enzymatic hydrolysis

Enzymatic hydrolysis of water-insoluble solids of the pretreated CS

To prepare water-insoluble materials, pretreated CS from both pretreatments was
washed with distilled water at a ratio 1 dry g (input C5 to pretreatment) to 50 mL of
water. For bench-scale dilute acid pretreated CS, the designated amount of distilled

water was poured into a filter system with Whatman filter paper (No 4) under vaccum.

20

 

The solid
pretreate
for 24 hr
Ahereac
system. 1

content C

The wate
nhnures
ceHMaseI
Novozynt
Multifect
Spefirum
and Mult
188 was I
COndqu
OligomEr

IIIpHCate

The solids remaining on filter paper were dried under vaccum at 60°C. For AFEX-
pretreated CS, the washing was achieved in two stages; (i) Incubation at 250rpm, 50°C
for 24 hr at 5% solids loading equivalent and (ii) two cycles of centrifugation at 6000 x g.
After each cycle of centrifugation, the supernatant was decanted through the filter
system. The total weight of water-insoluble solids was measured and the carbohydrate

content of the solids was analyzed using NREL protocol LAP-002.

The water-insoluble materials were enzymatically-hydrolyzed using either (i) cellulase
mixtures and (ii) cellulase + hemicellulase mixtures at pH 4.8, 50°C for 144 hr. The
cellulase mixture consisted of Spezyme CP [86.7 mL/kg CS; 15 FPU/g cellulose] and
Novozyme 188 [87.5 mL/kg CS; 64 pNPGU/g cellulose]. The hemicellulase mixture was
Multifect Xylanase [12.7 mL/kg CS] and Multifect Pectinase [12.7 mL/kg CS]. The
spectrum of activities for the commercial enzymes was as reported [44]. The Spezyme
and Multifect enzymes were obtained from Genencor Inc. (Palo Alto, CA) and Novozyme
188 was purchased from Sigma-Aldrich Co. (St. Louis, MO). Enzymatic hydrolysis was
conducted at 5.1% glucan loading. Glucose and xylose in both monomeric and
oligomeric forms were measured. Error bars shown are standard deviations of

triplicates.

21

Mass b;
Carboh
Afterp
were VI
and dn
were n
of the I
oligom
hemice
anhydr

around

Klason
The d”.
Billy-m£
maREr
CODCEn
and60

maIIET I

Mass balance construction

Carbohydrate mass balance around pretreatment

After pretreatment, the pretreated solids from both AFEX and dilute acid pretreatment
were washed with water at a ratio of 1 g input biomass to 50 mL of water. Total mass
and dry matter content (%) of the input and output materials around the pretreatments
were recorded. Volumes of the water extracts were recorded. Glucan and xylan content
of the dry matters were analyzed using NREL protocol LAP-002. Both monomeric and
oligomeric (LAP-014) sugars of the water extract from AFEX-pretreated CS and
hemicellulose hydrolysate from dilute acid pretreated CS were analyzed. Total
anhydrous equivalent of glucose and xylose were calculated for input and output

around both pretreatments. Percent carbohydrate conserved was calculated using Eq. 2.

(Anhydrous Equivalentof Sugar)output

 

SugarPreservation(%)= x100%--- Eq. 2

(AnhydrousEquivalentof Sugar)mput

Klason lignin mass balance around pretreatment

The dry matter mass of the input and output materials around pretreatment and
enzymatic hydrolysis were recorded. The total percent of Klason lignin in the dry
matters before and after pretreatment was analyzed using NREL LAP-002. The final acid
concentration, temperature and residence time for the assay was 4% sulfuric acid, 121°C
and 60 min, respectively. The Klason lignin was calculated by multiplying the total dry

matter with the percentage of Klason lignin.

22

Residual solids analysis and heat value estimation

After enzymatic hydrolysis at 5.1% glucan loading, unhydrolyzed solids were separated
by centrifugation, washed twice using distilled water, and dried under vacuum at 55°C.
Total dry weight was recorded. Glucan and xylan in the unhydrolyzed solids was
analyzed using NREL LAP-002. Residual non-carbohydrate solids and their heating value
were estimated is described in Eq. 3 and Eq. 4:

Residual Solids (g) = Total Unhydrolyzed Dry Solids (g)— Carbohydrate in the Solids--Eq. 3

 

Energy content (kJ) = 0.90 x Residual solids (g) x 25.4 kJ/g Eq.4
This is done by assuming 90% of the total residual solids is lignin and the rest of 10% has

negligible heat value. The heat value of lignin used (25.4 kJ/g) was as reported [45].

RESULTS

Sugar and lignin preservation during AFEX and dilute acid pretreatment

AFEX pretreatment on CS at 62.5% solids loading preserved all its carbohydrates. Nearly
10% of the AFEX-pretreated CS carbohydrate was water-soluble, of which two-thirds
was monomeric or oligomeric xylose (Figure 5A). However, 13% of the xylose sugar was
degraded in the dilute acid pretreatment at 5% solids loading. About half of the total
input solids was solubilized in the acid solution. While 59% of the total remaining solids
after dilute acid pretreatment is glucan, its xylan content was reduced to about 3%
(Figure 58). Forty-two percent of the total output sugars from dilute acid pretreatment
was water-soluble, predominantly in monomeric form. The concentration of the total

sugars in the acid liquid stream was 14 g/L.

23

£80m zuofima .cozmbcmocOO «0%: Eat Row ”2388 52m: $3209: of Lou— cozficoO >88 of...”

E8: 88.0: 58:

@3202. um. .EmEummbma of op 22.32: 39: :o comma mEnmo. 8:9. 80;.“ am So 62:8 mm; 2958sz Eon Go: 86%
58 638me Co 9.283 2628.832 @528. 3:8 Xom new 9°03 pm 35:28 825 EmEamwbmi Bed. 325 new me<

#585865 Eon

33:6 EV new “cm—58:98 xm:< A3 86% 58 :6 2me8me Bow 33.6 6cm xE< c3253 acmtquoo 8:23 $.22 m 23mm

3x 96 u so #8 z
8.6 m5 m do “no
_>x 22 m m.oam.o~
2G 22 m 33:.
Emmbm cmm>>

7

 

 

 

 

 

 

 

 

\ /\ 9
ma. m 8.93.: 5m. 394.3
. - . I I. 0 . - .
323+: 1 S 1 S H 333% 2
86343.: m % W. m % m 86 5.96.?
h m- 5 WI m. S 3 I
8 32 8525 an... m... w. aw m V 8 3.83
. I . a
3 98 S I $ 8 I m. M
K . \K k
2% 3m EmEummme

 

22 325 a

2
.5 2.0 ”no“?
go 9.0 3.3.3
3x 22 8.93.8
so 22 8.88.8
Emmbm cmm>>

 

 

v

 

 

 

 

 

 

\II/\ /
Lm. mmdweda .mn Lm. mmdnvdfi
3333.: m .9 M m. 7. 3x 3.93.8
Isomeflmmm W. M w W. mm N 5308.3
0 as? .8532, w. m. N m. S B 83.02
3.33: outflow WW.
5 W
[1.ka k
ONI ..m
a EoEummme
Xmm< H.4

 

24

Regarding Klason lignin content (at assay condition: 4% sulfuric acid, 121°C, 60 min),
AFEX completely preserved the lignin. In the dilute acid pretreatment, 12% of the Klason

lignin was not preserved in the remaining solids (Figure 5).

Fermentability of water extracts of the pretreated corn stovers

Categorically, the water extract from AFEX-treated CS exhibited significantly higher
fermentability with regards to cell growth, and glucose and xylose consumption in both
S. cerevisiae 424A(LNH—ST) and E. coli K011. Remarkably, the water extract from dilute
acid pretreatment inhibited the growth of K011 severely in both 7.5% and 15% solids
loading (Figure 6) over the tested fermentation period (24hr) (Figure 6). Comparing
different dilute acid pretreatment approaches, the water extract of pretreated corn
stover from the Sund reactor was more inhibitory than the benchscale, low solid loading
pretreatments. The cell density of 424A(LNH-ST) in the Sund-CS water extract was about

half of that of acid pretreated CS at laboratory bench-scale.

In contrast, all tested AFEX-CS water extracts were high fermentable. In essence no
inhibitory effect on cell growth was observed. Fermentations of AFEX-treated material
performed similarly to that of yeast base nutrient (YNB, 13.7 g/L). in the case for K011
fermentation, xylose consumption in AFEX-CS water extract (7.5% solids loading
equivalent) was two-fold higher than that of YNB. Complete glucose (8-10g/L)
fermentation was achieved regardless of pretreatments by the S. cerevisiae (Figure

7A,B). While better xylose fermentation was achieved in K011 than 424A(LNH-ST) in

25

£de new xmfi mm; “.38“ muombxm .325 of *o u:o_m>_:cm mEumo.
mE_om .Ecoomoo *o :5. ad *o 3?.ch :3 35:. 53> mE:_o> mezzo; ..E o.~ am 3233:: =m>>é~ E nopoaucoo mm; cozmucmctmn.

:9. E 25 5-:23313 .Emeaeaa xm:< Em :23§om £3 £3 $58. 8:8
$285. um “5:53me Eon 32% So: mtmbxm boom; uB:mEm.aa:m-mz> E cozmpcmEcfi *o E «N Eta Bacon :3 m meswm

cm>3m Eou “.38me E0: Emmbm nmm>>

  
  

 

 

 

ox. 2 / 0/ J. ./.
J. o 0 b o O b
o« Ox. 9 no a! 90 ob
// b 00 . J. 9 .0 Av .(J. 9
44% 0 £3; 9y co s :2
Av + fo or o + ,w of of
M S
T m é.
1
u m
_ " nix
_ _ p
u n N am ..M.
m
m _ 0
. M 9
M m m
. m .m m
H u -w
$0.2” um.” *mfi um i Rod." an" Xmfi um
i Emobm gnu; “Samba 52$ . Emobm 525 “£353 525
..u 3
_ m m _ <
pp —--_PFLl_-ph+p— »-»._-b.—-_>

 

 

 

 

 

 

 

 

 

-—.—.— m

26

532638. r<m NHmm ncm ._\m Hum um m. 32? new 38:3
33:: .Ecoomoo u_o «E: 2. 96 >5ch :8 33:: £35 mE:_o> macro; ..E o.~ um $23.2... zogém E coyoancoo 8.5 5.033583”.

.8 .3 22933 9.63 3:9. gofl .25 G
5 “533:3 9552 $59. .Xms um Hombxm .325 E “5623wa VG”? can :ncam§0m .Xmfi :me wc_omo_ £59. “5.8:“.
pm “5.53.3... Eon 33:. 80.. $8.88 .325 no~co8m_aa:m-mz> .._ 53358.8 .o E «N .mtm cozaEzmcoo .mmsm 5 83m:
.965 Sou 238me So: Emmbm 52$

.26 64 826 826 64 926 926

        

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

027 $Nw4 WOCAJWJ 0 . 9.00. P#& 6 {am HQ 6’ {Mwwd fCCQW/ C . {an Fifi av {om VG 0’
. to
. m
Tod U
. m x
. ON w M.
. d O
. cm m. we
U
I ow \I
. %
:9. SHE . om (
. . . [L 539'
w I 00 D 1 low U
] l x O .._ 1 C
j . . g: 3
. w
, n - ON . . ON
. _ m D
. _ . ow . . ow Awu m
1+ 0
. _ . . . I. S
_ O a
. _ -8 . :5 u
u . . %
. -ow 1 I om l\
igl. r tgf [L I
8am . . _ _ . §<

 

 

 

 

 

 

27

AFEX-CS water extract, the opposite trend was observed in dilute-acid-CS water extract
due to the inhibitory nature of the water extract and the strain robustness toward the
inhibitors. The time course of K011 and 424A(LNH-ST) fermentations in both water
extracts also showed similar trends as the microplate fermentations regarding strain
robustness of 424A(LNH-ST) and better xylose fermentation of K011 in AFEX-CS water
extract (Figure 8). At the low initial cell density tested, K011 could not grow in the water

extract from the acid-pretreated corn stover over the fermentation period.

 

 

 

 

 

 

 

 

50‘w d 50‘\\FM—‘ ‘

1 ‘ 'lV. ‘

:3 40d , - E 4oq ' », -

U 30‘ - U 30‘ . ~

8 ‘ * S ‘ v -XylAFEX ‘ 4

B 20‘ v d k: 20‘ +XylAcid ‘

% 101W § 10; T‘GchFEX

3, <' i 3, \ +Gchcid .
o .‘g-T,a,=, wit: 0 afivr‘. .4i.-..:

0 20 4o 60 30 100 120 o 20 4o 60 80 100 120

Time (hr) Time (hr)

Figure 8 Time course of fermentation of water extract from 7.5% solids loading
equivalent of AFEX and dilute acid pretreated corn stover.

Fermentation was initiated with 0.1 unit ODGOOnm and the water extracts with
supplemented with 1 g/L yeast extract and 2 g/L peptone.

28

Enzymatic hydrolysis of washed pretreated solids

Enzymatic hydrolysis at 5.1% glucan loading of washed solids from both pretreatments
achieved similar total glucose yields at 82% when the cellulase-only mixture was used.
However, AFEX achieved 6% higher glucose yield when both cellulase and hemicellulase
was added (the difference was within the margin of error) (Figure 9). The added
hemicellulase mixture also improved xylose yield in AFEX improved from 83% to 91%. In
contrast, the hemicellulase mixture does not affect sugar yields in dilute acid pretreated
solids. This is probably due to the low xylan content (about 3%) in the solids. The
proportion of glucose and xylose oligomers to the total sugars in the hydrolysates is
about 12.5% and 25.0%, respectively (Figure 9).

E Oligomeric
- Monomeric
, 100 I

 

 

80—

  
  
 

Ch
0
1

 

Glucose Yield (%)
Xylose Yield (%)
J;
O

N
O
1

 

 

\.\C\ “O “O “0 \AC C\ “O “O
[*de pYEtguAC\dgloeNa pflw “0“” pS'E‘I‘S‘LC chip” dd“
0““ \N

Pretreatment Type

Figure 9 Enzymatic hydrolysis yield on the water-insoluble AFEX-pretreated corn stover
and dilute acid pretreated corn stover at 5.1% glucan loading, pH 4.8 and 50°C.

Note: Xylan content in water-insoluble dilute acid pretreated corn stover is very low
(3%), HC: Hemicellulases

29

Energy content of non-carbohydrate residual solids

Non-carbohydrate residual solids from AFEX and dilute acid pretreatment and hydrolysis
are 19.1 g and 15.8 g per 100 g of untreated corn stover, respectively. Based on the
calculation method listed, biomass processing technology based on AFEX pretreatment
is able to generate 737 kJ/kg more energy from the residual solids than that of dilute
acid pretreatment. About 23.2% (AFEX) and 19.3% (Dilute Acid) of the heating value in

the untreated corn stover remained in the non-carbohydrate residual solids (Figure 10).

 

 

 

 

 

 

 

 

 

 

 

 

A: AFEX
B: Dilute Acid Pretreatment
K \r \ r 19.111.3g a
‘5 § 2" Residual Solids
09 on 5c”, A 76.51 1.7 ,2 us - ‘5: 437.0130“;
A:100.0gCS E 5;: go 3 as?) 3 b
4.. 4: o _ Washed AFEXCS g m 3 3 Energy Content
‘ m in W ‘O > 7, _ ;
c o
8:100.0g cs g m 33 8 51010-18 E‘ %‘ 3 8 1531023 ,
a.» 3 21 -' Washed Acid CS L” 5 3 5. Residual Solids
E. g 0 363.3149 u
b
\ /\ J K / Energy Content

Figure 10 Mass balance and energy content for non-carbohydrate insoluble solids after
pretreatment, washing and enzymatic hydrolysis.

3 Residual solids = (Recorded total dry solids left unhydrolyzed)-(Dry glucan and xylan in
the solids)

b Energy content = 0.9 x Residual solids (g) x 25.4 kJ/g. This is done by assuming 90% of
the total residual solids is lignin and the rest of 10% has negligible heat value. The heat
value of lignin used (25.4 kJ/g) was as reported

30

DISCUSSION

Dilute acid pretreatment reduces maximum possible product yield by 10%

Overall yield has the greatest impact on the profitability or viability of a commercial
process. To make cellulosic ethanol a cost-effective fuel, efforts to increase ethanol yield
per unit mass of biomass (corn stover) at a given product titer deserve the highest
priority. In this regard, AFEX preserves all carbohydrates while effectively increasing the
susceptibility of the pretreated corn stover to hydrolytic enzymes. Unlike AFEX, acid-
catalyzed pretreatment hydrolyze hemicellulose almost completely. Unfortunately,
monomeric pentoses are further degraded to byproducts such as furfural under acid
treatment conditions. In this report, about 13% of xylan was lost through chemical
degradation. However, a greater degree of degradation (20-30%) was reported at a
higher solids loading of dilute acid pretreatment [15]. This reduces the maximum
product yield by 10%. In other words, using dilute acid as the catalyst for feedstock
pretreatment would require a 10% increase in raw material cost to achieve the same
yield compared to AFEX assuming that both technologies achieve 100% bioconversion of
carbohydrate. In any mature chemical process for commodities, raw material is the
dominant factor in the processing costs [46]. Hence, selection of a pretreatment that

highly preserves plant carbohydrates is critical for long term success in this industry.

31

Pretreatment dictates the fermentability of pretreated biomass

Apart from the preservation of carbohydrate, an ideal pretreatment reduces the
generation of inhibitory degradation compounds. AFEX—pretreated CS is highly
fermentable using both bacteria and yeast. In certain cases, the soluble fraction of AFEX-
pretreated CS has been shown to be beneficial to microbial growth [47]. In contrast,
corn stover hydrolysate from dilute acid pretreatment is significantly more inhibitory
and difficult to be fermented. The nitrogenous (amides and amines) reaction products
formed during ammonia—lignocellulose reactions are generally non-inhibitory toward
microbial growth. These degradation products would otherwise be organic (aliphatic
and phenolic) acids in acid catalyzed reactions[35]. Fermentation at higher initial cell
density, nutrient supplementation and/or detoxification are likely needed to alleviate or
overcome their inhibitory effects of acid pretreatment [17, 29] judging from the
extensive investigations that on the related subjects centered on dilute acid pretreated

materials.

The pretreated biomass, which inherently possesses various nutrients, can serve as the
carbon and nitrogen sources for bioconversion. If there is no interference from
inhibitory degradation compounds on fermentation, cellulosic ethanol fermentation can
be conducted in an approach similar to the corn ethanol industry (where hydrolysate

conditioning and high initial cell density are not required).

32

Pretreatment determines feasible biomass processing configurations

Due to the nature of pretreatment, particularly with respect to the degree of
hemicellulose solubilization, inhibitor generation and nutrient preservation, different
biomass processing strategies that maximize the advantages of each pretreatment
should be exploited. Dilute acid pretreatment effectively hydrolyze hemicellulose,
eliminating the need for hemicellulases during enzymatic hydrolysis. Nevertheless, the
hemicellulase stream is inhibitory toward enzymes and microorganisms. Therefore,
separation of solids and the hemicellulose stream as previously proposed [48] is
essential to minimize the adverse effects of the inhibitors from the bioconversion of the
remaining solids. However, important technical issues need to be solved in a cost-
effective fashion, including (i) separation of solids and liquid with low fresh water use
and (ii) effective fermentation of the hemicellulose stream at high sugar concentration

without significant conditioning.

AFEX-centered biomass processing can be performed in a straightforward manner
where the pretreated biomass (cellulose and hemicellulose) can be converted to
ethanol after enzymatic hydrolysis and fermentation without washing or stream
separation [47]. In this report, washing was done in AFEX-pretreated corn stover to
establish a basis for comparison. Due to high fermentability of AFEX-pretreated biomass,
washing, nutrient supplementation and high initial cell density are not required during
fermentation stage[47]. In comparison to dilute acid pretreatment, a relatively large

portion of oligomeric xylose is present in AFEX hydrolysate. Exploitation of

33

hemicellulase-secreting strains such as Thermoanaerobacterium saccharolyticum to
biologically process AFEX-pretreated materials could address this issue without added

cost of hemicellulase[49].

AFEX enhances co-product generation and diversity

Reduction in greenhouse gas emissions by cellulosic ethanol E85 relative to petroleum
gasoline is projected to be 68-102% and it is largely due to the heating value of residual
solids (primarily lignin) to generate steam or electricity as a co-product [50, 51]. Our
results indicate that AFEX-centered cellulosic technology is expected to have about 17%
more available energy from the insoluble lignin residue compared to dilute acid. This
also implies that the selection of pretreatment directly affects the magnitude of
environmental benefits brought about by a cellulosic ethanol plant beyond the direct
impact of the pretreatment process. Nevertheless, a definitive conclusion on the impact
of different pretreatments on various environmental benefits can only be made after

careful life cycle analysis based on these experimental data.

Lignin removal is a function of severity in terms of acid concentration, temperature and
residence time [52], and part of the solubilized lignin can be recovered [53]. Hence,
optimization of acid pretreatment condition that takes into account of the economics of
recoverable lignin can be important. However, the recovery process will inevitably
increase the processing cost relative to a production where lignin is preserved in the

solid residue and not washed out.

34

Furthermore, a pretreatment that does not produce inhibitors that are detrimental to
living cells could well determine the product diversity of the prospective full scale
cellulosic biorefinery. AFEX has been shown to improve the rumen digestibility of
lignocellulosic materials [54], and thus providing a valuable co-product (animal feed)
from the biorefinery [55] through unlocking plant carbohydrate for animal production.
This approach can create a scenario where food and fuel production are cooperative

rather than competitive.

CONCLUSIONS

Ammonia fiber expansion (AFEX), a dry-to-dry pretreatment process, completely
preserves Klason lignin and carbohydrate. In comparison, 13% of xylan was degraded to
byproduct and 12% of the Klason lignin was not preserved in the dilute acid pretreated
corn stover. Categorically, streams resulting from AFEX-CS displayed significantly better
fermentability than those from dilute acid. While dilute acid pretreatment eliminates
the need for hemicellulolytic enzymes for hydrolysis, AFEX-centered cellulosic
technology simplifies production steps, reduces the requirement for nutrient
supplementation, increases the diversity of co-products and potentially enhances the

environmental benefits beyond the direct impact of the pretreatment processes.

35

CHAPTER III STRAIN COMPARISON FOR CELLULOSIC ETHANOL

FERMENTATION

INTRODUCTION

Development of microbial platforms has been extensively pursued to achieve cost-
competitive ethanol yield, titer and productivity [56, 57]. Among the ethanologenic
strains, Saccharomyces cerevisiae [29, 58], Zymomonas mobilis [31] and Escherichia coli
[28, 59] have been widely investigated and developed for cellulosic ethanol production.
An economically-attractive cellulosic technology requires the strain to achieve ethanol

yield, titer and rate higher than 90%, 40 g/L (5.1%v/v), 1.0 g/L/hr, respectively [27].

Despite the wealth of publications on strain development, efforts to compare their
performance are often hampered by the variations in experimental conditions such as
sugar type and concentration, media nutrient levels, initial cell density, feedstock
pretreatment selection and detoxification (if applied)[60-62]. In this work, a common
platform to obtain comprehensive fermentation parameters using S. cerevisiae

424A(LNH-ST), Z. mobilis AX101, and E .co/i K011 as the fermenting strains is
36

e sta b Iished. In addition, the effect of the water-soluble substances (mainly
Dret reatment-mediated reaction compounds) from AFEX pretreated corn stover on the
grow-th and fermentation of these three strains is investigated. Fermentation of

er‘Z\/r11atic hydrolysate from AFEX-pretreated corn stover at high solids loading was also

exam ined.

MATERIALS AND METHODS

AFEX-pretreated corn stover (AFEX-CS)

Corn stover (CS) was obtained from NREL (Golden, Colorado), milled and passed through
a 4 mm screen. The untreated corn stover consisted of 33.2% cellulose, 22.4% xylan,
3.3% arabinan and 2.3% protein on a dry weight basis. The pretreatment conditions
were as follows: temperature at 110-130 °C; catalyst loading at 1.0 g anhydrous
ammonia to 1.0 g dry corn stover ratio; water loading at 0.6g water to 1.0 g dry corn
stover; 15 min retention time. Each pretreatment batch contained 150 g corn stover on
a dry weight basis. The AFEX apparatus, pretreatment conditions and experiment

procedures were as reported on page 14.

Microbial strains
MetabolicalIy-engineered ethanologens used in this investigation are Saccharomyces
cerevisiae 424A(LNH—ST), Zymomonas mobilis AX101 and Escherichia coli K011. Strains

Saccharomyces cerevisiae 424A(LNH-ST), Zymomonas mobilis AX101 were provided by

37

 

P u rd ue University and the National Renewable Energy Laboratory (NREL)respectively.
Stra i n E. coli K011 was obtained from American Type Culture Collection (ATCC) with
desi gnated number 55124. Genetic modification and reported fermentation

performance were previously reported [28, 31, 63, 64].

COI‘n steep liquor (CSL)
Fern'iGoldTM Corn Steep Liquor (Lot: 154-07) from Cargill, Inc (Minneapolis, MN) was
Used as the nitrogen source for fermentation. Technical information from Cargill, Inc

it1dicated that FermGoldTM CSL contained 48.0-52.0% dissolved solids and 19.5-23.5%
total protein. To prepare 20%w/v CSL, 200 g of FermGoldTM CSL was diluted to total
volume of 1.0 liter with distilled water after pH was adjusted to 7.0 with regent grade
KOH. The insoluble solids were separated from the liquid by centrifugation at 5,000 x g
for 30 min. The 20%w/v CSL was sterile-filtered (0.22pm) and used for media

preparation.

AFEX-CS water extract preparation
AFEX-pretreated corn stover was washed with distilled water at a ratio of 1 g dry CS to 5
mL of water to produce a water extract at 20% solids loading equivalent as described on

page 16. The AFEX-CS water extract was used for the fermentation studies.

38

AF EX—CS enzymatic hydrolysate 6% glucan loading (18% solids loading)

Enzym atic hydrolysate from AFEX—CS was hydrolyzed using both cellulase and

hem i cellulase commercial mixtures. The cellulase mixture consisted of Spezyme CP

[86-7 mL/kg CS; 15 FPU/g cellulose] and Novozym 188 [43.7 mL/kg CS; 32 pNPGU/g

cellulose]. The hemicellulase mixture was Multifect Xylanase [12.7 mL/kg CS] and

MU\tifect Pectinase [8.9 mL/kg CS]. Enzymatic hydrolysis was performed for 96 hr at pH

4-8. 50°C and 250 rpm agitation. Other details were as described previously [47]

Seed culture preparation

The frozen (-80°C) glycerol stock was transferred to 100 mL liquid media (nitrogen

Source, 50 g/L total sugar, appropriate buffer and antibiotics) in a 250 mL unbaffled

flask. The cells were grown overnight under largely anaerobic conditions at their

respective temperatures and initial pH, 150 rpm agitation. Details of culture

temperature, initial pH, antibiotics, sugar levels and nitrogen source are as listed in

Table 3

Table 3 Seed culture media recipe for the three ethanologenic strains

 

 

 

 

 

 

 

 

 

 

Strain Temp Buffer/ pH Antibiotics Sugars Conc Nitrogen Source
(’0

K011 37 0.1 M 50 mg/L 50 g/L for Glucose- 2.0% w/v CSL for
MOPS/ Chloramph only and fermentation in
7.0 enicol fermentation; CSL fermentation;

AX101 30 0.05 M 30 mg/L 30 g/L + 20 g/L for 5.0 g/L yeast
Phosphate Ampicillin co fermentation extract+ 10.0 g/L
/5.5 and xylose-only peptone for water

424A- 30 0.05 M 50 mg/ L extract and AFEX

(LNH- Phosphate Ampicillin hydrolysate

ST) / 5.5 fermentation

 

39

 

W ate r extract fermentation
F erm e ntation was conducted on E. coli K011 using 15.0% solids-loadings-equivalent of
AF EX—WS with or without addition of commercial enzymes at loadings described in
previ 0 us section [47]. YEP (5 g/L yeast extract, 10 g/L peptone), 50 g/L glucose and 25
g/L Xylose were added into the media mixtures. Fermentation was conducted at 37°C,
pH 7 .0, in a 125 mL shake flask with a 50 mL working volume. Initial cell density was at

US CD (600nm). Error bars shown in the results are standard deviations of duplicates.

HPLC analysis
The concentrations of glucose, xylose, ethanol, acetate, formate, lactate, glycerol and
Xylitol in the fermentation and culture experiments were analyzed using HPLC with a
Biorad Aminex HPX-87H column (Hercules, CA). The column temperature was
maintained at 60°C and the mobile phase (5 mM H2504) was kept at 0.6 mL/min flow

rate. The HPLC system used was as reported [3].

RESULTS

Fermentations using CSL as nutrients supplement

Fermentations using 2% w/v CSL as nitrogen source indicated that these three strains
effectively produce ethanol from glucose or a mixture of glucose and xylose. During
glucose fermentation, the fermentations were completed within 72 hr (Figure 11A) and

ethanol was produced at concentrations higher than 40 g/L. In particular, 5. cereivisiae

40

424A( LNH-ST) has the highest rate of glucose utilization at 4.16 g/L/hr (Figure 11A).
However, an increase in xylose concentration correlated with a decrease of the overall
fer-m entation rate. Overall sugar consumption rates compared between the glucose and
XV'Ose fermentation were closest for E. coli K011 followed by Z. mobilis AX101 and then
5- cereivisiae 424A(LNH-ST) (Figure 11,Table 4). Remarkably, xylose fermentation in 5.
Ce reivisiae 424A(LNH-ST) achieved only 37.9% of xylose consumption after 168hr.
Nevertheless, xylose fermentation by S. cereivisiae 424A(LNH-ST) was completed when
Using YEP as the nutrients supplement. Specific ethanol productivities of fermentations
USing the bacteria (AX101 and K011) as the fermenting strain were at least twice as
great as those for S. cereivisiae 424A(LNH-ST), regardless the type of carbon source

(Table 4).

Metabolic ethanol yield and byproducts profiles

For 2. mobilis AX101 and 5. cereivisiae 424A(LNH-ST), metabolic ethanol yield appeared
to decrease in complete fermentation of xylose-containing CSL media (Table 4). Carbon
source (glucose or xylose) did not significantly affect the metabolic yield in E. coli K011
fermentation. This trend is also reflected through the profile of targeted byproducts.
The total concentrations of the targeted net-byproducts formation in xylose-containing
fermentation increased for Z. mobilis AX101 and S. cereivisiae 424A(LNH-ST) compared
to glucose-only fermentation, but were essentially unchanged for E. coli K011 (Figure

12). In xylose-containing fermentation (both xylose only and cofermentation) by AX101

41

.Eizdéme 8:388
.m 8... 892 888 .N 22 3 0.8 n8 .29. :8 m .8 M8 0.2 8 8:228 em; _.a 2; 838858 .5888 :5. no
3 E2933 >5ch :8 am 883:: new :oEucoo ofioemmcm >392 Eve: 52958 80.8: 9: E 863:8 mm; cozmacmcto”.

.858 c098 28 mm 80.? UV cam mm 032 am 835.: 803x cam 882w Am: .858 concmo 28 8 882m 2V
5:5 :Q 58: 83... 58 <3 gm 5 5-12:3? 8538 .m 88 832 88:. .N .28. =8 .m mew: 8:35.28 S 28:

 

 

 

               

 

 

 

 

 

 

152:: :5 we: :5 we:
2: o8. 8 8 o 02 oi 8 8 o o2 9: 8 8 8 8 o
— I. p p p p p . o . — b — . P p O . .‘p . . . . . . — . O
. m. . C
x m. D
-o~ Al I - -8 n
w m. m
. a co . fl %
3 m 3
row W S I row 0
3 u
S O S
v n u r 1 n
w m w
-8 m. - 38+ -8 w - -8 w
w 832], m m.
a :Oxlll m. N
U
-8 V - -8 I - -8 fl
f - - h -
88 U - -8H m - 08 <
_ p p p P . p p L . . p — — p —

 

 

 

 

 

 

 

 

 

42

00980 908 <0 0N + 823 88> 208 <0 02 am; a
Hm. 002000.. 00 802, 833300020 000 3.03 2.00.39: 30.3.8 2 00508 mi 0
50070 cOEmEmE.£-8 new _>x “Emu-0 0033:0800 0.0 ”Emu 03200.8 05 .8 swam-0:5 m

 

 

 

 

 

 

 

 

 

 

 

 

 

 

43

 

 

 

 

 

 

 

 

 

20.8 ~08 20.8 0.8 083x
00.820 80+ :2- <2 0.8 002 5.0203? 002 \. 00>
~00“ 00.08 ma“ 0.0“
00.08020 000+ ~m.0- <2 0.00 0.0m 30-120330
00.0w H00“ N0“ 20“
8.0830 ~<.0+ v04- <2 0.00 memo 832
20.8 ~08 0.8 0.8 083x
00030.0 ~50+ :2- <2 200 0.03 :00. 002 >00
H00“ 2.0.0“ 0008 0.0“ 0.0“
8.820 20+ 80. 0:. 0.8 «.8 502058
~0.0sH 00.08 0.020“ 0.0“ «10.8
8.001.000 550+ m~.0- 320.. 0.00 mém H0300. 803x
~00“ mod“ 00.0“ 23 0.08 0m + 08020
00.08000 ~50+ wmd- 0H.~- 200 N00 :00. + 00 )8
mod“ 00“ 0.08
00.08000 00.0 <2 02¢- N00 003 05-1203»?
mad“ H08 0.0“
8.0304” 00.0 <2 <00- ~.mm 0.03 833
00.08 00“ 001A 880.0
0008004 0.00 <2 nmd- 0.50 0.03 :00. 00a >00
.E 8.0 2;...

.. .._»... u ESE £070 28.0 23.0 $00 .30 .83.
333262.. :05 _>x 0.0 23> coznfiamcou SB. >553
:05 050000 a 22)}. 3.283095 o_._uoE:_o> 0:332). 5030 5930 0:8 355:2

305 2.3000 «oat-xv “80> .o :00. 8:0:
. . . < 0.0 . _ a
w 030.33% m. new 833 $.38: N 39. .2 .0 0:0 0038:0850 .8 30888 3.3.02 c 0.03
N a 5.12%: .

02 8%.. .2.

XV‘ito I is the primary byproduct and it contributed about 70% of the total measured
by!) roducts. In fermentation by 5. cereivisiae 424A(LNH-ST), net productions of 67 and
81 m g glycerol/g total consumed sugar(s) were observed in glucose and co-
ferTNentation, respectively (Figure 12A & B). In addition, xylitol production during co-
ferMentation contributed to the lower metabolic yield observed compared to glucose-
only fermentation. Organic acids were identified as the predominant group of

b\IDI‘oducts from fermentation by E. coli K011 (Figure 12). Although total concentrations

5

of the targeted byproducts were at 6.0-6.S g/L regardless of carbon source, the
byproduct profile varied substantially. While acetate formation increased from 21.4
(glucose-only) to 47.5 mg/g consumed sugar (xylose-only), lactate production
diminished during xylose-only fermentation. Of all fermentations, glucose fermentation
by Z. mobilis AX101 achieved the highest metabolic yield and lowest targeted byproduct

formation.

Fermentation using AFEX-CS water extract

AFEX-CS water extract was used to provide a representative compound profile found in
the pretreated biomass without the involvement of enzymatic hydrolysis. Fermentations
by these three strains exhibited similar patterns; in that moderate levels of AFEX-CS
water extract improved cell growth but the degree of improvement decreased as the
strength of the water extract increased. However, a greater cell density was achieved in
most of the water extract-containing fermentations relative to the control (YEP with no

water extract) (Figure 13A).

44

Th e rate of xylose fermentation correlated well with the cell growth pattern (Figure 13).
E. coli K011 consumed xylose completely at the highest rate (close to 20 g/L/hr) at 5%
$0|id 5 loading equivalent of water extract. However, the rate decreases substantially as
the Solids loading increased. In the highest tested solids loading, 5. cereivisiae
424A( LNH-ST) has the greatest xylose consumption rate (12.8 g/L/hr) followed by E. coli
K011 and Z. mobilis AX101. Although able to ferment at the highest specific rate (g/hr/g
tens), Z. mobilis AX101 consumed both sugars at the lowest volumetric rate. Z. mobilis
AX101 also appears to have the lowest tolerance toward water-soluble compounds in
AF EX-CS. The cell density of Z. mobilis AX101 at 24 hr decreased by 66% when the solids
‘Oading was increased from 5% to 15% (Figure 13A). This decrease for both S. cerevisiae
424A(LNH-ST) and E. coli K011 was 26%. The effect of AFEX-CS reaction compounds at
15% solids loading on glucose fermentation was practically negligible for E. coli K011
(Figure 14). However, these compounds are shown to be rather inhibitory toward xylose
fermentation (Figure 14). The xylose consumption rate within 96hr in water extract

fermentation was five times lower than that of the control experiment.

Fermentation using AFEX-CS Hydrolysate (18% Solids Loading) and AFEX-CS Water
Extract

All tested strains were able to grow and completely consume glucose on the AFEX-CS
hydrolysate without washing of the pretreated biomass, nutrient supplementation or
detoxification (Figure 15). Similar to co-fermentation in CSL (Figure 118, Table 5), xylose

fermentation is considerably slower than glucose fermentation. In the hydrolysate

45

 

@Mentation, xylose fermentation from the tested bacteria (AX101 and K011) was very

900 r; less than 20% of the total xylose was consumed (Figure 15A, C; Table 5). Hence,

X\I\c>se fermentation became the bottleneck for yield, concentration and rate for the

bacteria. However, nearly complete xylose consumption was achieved in S. cereivisiae

424A(LNH-ST) fermentation at a metabolic yield of 0.47 gram ethanol per gram

consumed sugars.

Table 5 Qualitative summary of the relative fermentation performance of Z. mobilis

AX101, S. cerevisiae 424A(LNH-ST) and E. coli K011

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Fermentation Parameters AX101 424A(LNH-ST) K011
Glucose In Corn Steep Very Fast Very Fast Fast
Consumption Liquor
In Lignocellulosic Average Very Fast Average
Hydrolysate
Xylose In CSL Co- Average Very Slow Fast
Consumption Fermentation
In Lignocellulosic Very Slow Average Very Slow
Hydrolysate
Nutrient Glucose-only Low Low Low
Requirement Co-fermentation Low Low Low
Xylose-only Average High Low
Growth Robustness Average Very High High
Metabolic Yield Very High High High
46

 

 

 

2.0220303 00.20.3030 .0 0:0 882 0.0.52: .N .200. :8 .0 007.: £000.. ._mu 5 .._; 00H 00 0038:0803

03.? UV 0:0 E 00H “0 00330085060 Am: 0; N0 00 0039:0850 @8020 A3 0550 0030500008 303020400 NH 050:.

£900
<30 0002 :9.

 

 

 

 

 

 

 

 

A9

(1/3) uonenuaouog) SanpOJd

52$.
<49 H002 :9.

 

 

 

 

 

Ila

(1/3) uonenuaouog SthpOJd

sea.
<09 802 :00.

 

 

 

 

(“I/3) uonenuaouog 51:)an Jd_,(8

47

 

 

 

 
 

0:3000 in 0.0 0:0 000.020 um00> <0 m.~ £05 002082003. 00.5
000:. 80.005800 0:0 202000000. 6500000 00 0:: 0.0 0:0 <0 0N .._\m 0m 0.02, 50:00 :00 0:0 03.? 08020 00 5305:0050
_0_::_ 05. 000.030 0_no.00:0 200.0. .00:: 0829 0:38? .7: 0.~ 00 0:20 0:02, «N :_ 006:0:8 0.0.5 00:08:00.0 0:0

0203:9500 E 0.0. .0000 :o0.0E:m:8 033x 0:00.00 UV 0:0
:o_u0:.:m:oo 000020 0:00.00 Am: 5.50.0 :00 As :0 .0300. :50 00000.38“? Eot 00::00E8 0328-80.02, 00 0.00000 0,: ma 050:

 
   
 

£0.00 250029.050 £0.00 250200.050 mEmbm 0_:0mo_o:0£m
<09. 839. Sex <0: 352 :9. m N H
I r 0
d
d a
0 - ON 0
3 a 8
0 w 4
1 .
X D
.M i n
w 00 m m
a m )
O 0 O
U u u
S
n 1 00 m w
w (
d w
u. . m
m w.
W -00 \o/w . $0: I.
. i .0... I .
so I a
i 00H . 0:202:00 05003 00:3 .
_ . p . p U - . _ h p m — . _ . NH <

 

 

 

 

 

 

 

 

 

 

 

 

l'l'l'l'l'l'l'l'l

Glc Xyl ‘
+-~ I - Enzyme Only

SO-l

  
   
 
  

 

 

 

j 404 —°—+~— Water Extract Only ‘
30 + A -~~ Enzyme + Water Extract *
C _ _
Q 30
in!
to ..... 4
:2 \rs\-- ‘
c - “‘ “- “~01-.. -
8 2 0 §\\‘\-<- }‘ ‘-‘—--‘
1: “‘+\ .....
.. ~~i
U 10 -i -
0 . , . , . . .” . =13!“‘:*3' . , f .
0 20 40 60 80 100 120 140 160 180 '
Time (hr)

Figu re 14 Fermentation using E. coli K011 on YEP media supplemented with
commercial enzymes, 15% solids loading equivalent of AFEX-pretreated corn

stover water extract or a combination of commercial enzymes and the water
extract.

.Fe rmentation was conducted at 37 °C, pH 7.0 (adjusted during fermentation) and was
' n itiated at 0.5 ODGOOnm.

DISCUSSION

Rationale behind the platform for comparison
I h t his chapter, I first compared glucose, xylose and co-fermentation in the CSL and
f0 ' I owed by co-fermentations on AFEX-CS enzymatic hydrolysate to elucidate its effects
0 T‘- rnicrobial growth pattern and xylose utilization. Fermentations using CSL reveal

1: Q
"I‘nentation performance of respective ethanologens without the interference from

49

 

 

 

reaction products from the pretreated biomass. CSL has also been regarded as a
eco n omical nitrogen source in large scale fermentation [39]. Lignocellulosic hydrolysate
fro m AFEX-CS without washing, detoxification and supplementation provided the actual
su ga r media for cellulosic ethanol production. This investigation platform would enable

us to evaluate the strains based on their intrinsic fermentation ability and robustness for

in d u strial applications.

Benchmarking of the non-cellulase—secreting ethanologenic strains on CSL media

The three tested strains were able to produce ethanol with a metabolic yield between

82.4—93.2% of theoretical maximum in both glucose and co-fermentation at

COncentrations of 40 g/ L or higher, at a rate over 0.72 g/L/hr (0-48hr). These parameters

a re comparable to those projected to be necessary for a viable cellulosic ethanol

in d ustry [27]. The growth robustness of E. coli K011 and S. cerevisiae 424A(LNH-ST)

5 h 0 u Id be sufficient to withstand the potential inhibitory effect of AFEX-CS degradation
C0 m pounds at very high solids loading (greater than 25%). However, in the case for
A F EX-hydrolysate, robustness of the strain can be readily increased through cell

a d a ptation, proven for both 2. mobilis AX101 (Figure 15A) and K011 [3]

H igher hexose to pentose ratio improves overall fermentation efficiency and

e .
chomIcs

m ong all tested strains, regardless of media used, fermentations wuth higher glucose to

X
y ‘ Qse ratios yielded better results in term of ethanol yield, concentration and rate;

50

[Ir _1

 

 

these are the three most important parameters dictating the overall economics of

ce l l u losic ethanol production. Pentose-only fermentation, even in a naturally-occurring
xylose-metabolizing strain such as E. coli, has proven to be more difficult than hexose
fe rm entation. One proven cause is the lack of precursors to synthesize products derived
from 2-ketoglutarate[65]. In the heterologous pentose metabolic system, further
Com plicating issues are associated with pentose transport [66] and redox balance [57,
58] must be resolved. The increase of hexose (glucose and mannose) to pentose ratio in

biomass could be achieved through plant genetic engineering [67].

 

 

 

 

 

 

 

 

 

 

 

 

 

60'1-1 .0 SvarrvIv 60,......
A 4 ( B J +---- Glucose ‘ C
504‘ ~ 50- —O—---O--Xylose - 50: -
__l - i 4 4' +--*--Ethanol-
\ 0‘
00404 . a \ 4 4o- -
V l.
C b I
.9 30. ' —-“" - . 30: -
‘0' .
‘0 ‘ .
& |’ . ‘-
E 20- x’ ‘ “~ . 4 20- -
2 l ..
o 10" ’0‘ \.‘ -l .. 10.. _,
U I]. -‘
O ' ' TL' I :77 '_ 0 ' I ' I ' I ' O i I Y I -' I '-
0 4O 80 120 160 0 40 80 120 160 0 40 80 120 160
Time (hr) Time lhrl Time (hr)

F i gure 15 Fermentation using (A) Z. mobilis AX101, (B) S. cerevisiae 424A(LNH-ST), (C) E.
coli K011, in enzymatic hydrolysate from 6.0% glucan loading of AFEX-
pretreated corn stover.

I:
de :- Inentation was conducted under largely anaerobic condition and initiated at cell
6 e hsity equivalent to 0.5 unit OD600nm Temperature and pH were controlled at 37°C,
6‘ a for K011, and 30°C, 5.5 for AX101 and 424A(LNH-ST). Solid lines: Seed culture in YEP;

Q“':ted lines: Seed culture in 3% glucan loading of AFEX-CS hydrolysate.

51

I“. _ “II

 

 

 

Degradation products from pretreated biomass complicates xylose fermentation

Xyl 0 se fermentation in lignocellulosic hydrolysate is substantially more challenging
rel ative to co-fermentation in CSL. Xylose consumption in the bacteria fermentations

we 5 considerably weaker than in S. cerevisiae 424A(LNH-ST) fermentation. The selective
inhibition, presumably from reaction-degradation products from pretreated biomass, on

xylose fermentation is not well understood and deserves extensive investigation.

ln most pretreatments of biomass at acidic pH such as dilute acid and steam explosion, a
he|"I"Iicellulose hydrolysate stream is produced and separated from the solids stream for
fermentation [43]. Inhibition due to an acidic pH pretreatment was reported to be
considerably more extensive than inhibition due to AFEX pretreatment (2].
Fermentation of a hemicellulose hydrolysate can be very difficult; detoxification,
n Utrient supplementation and/or high initial cell density may be required to increase its
fe rmentability. Improved fundamental understanding of the inhibitory mechanism

Se l ectively targets xylose fermentation, presumably by these degradation products,

C0 u Id help alleviate this crucial process bottleneck.

Advantages of the individual fermenting strains

8 ‘ cZerevisiae 424A(LNH-ST) exhibited the best glucose fermentation rate and highest
QVQ rall fermentation yield in the presence of AFEX-CS degradation compounds with

a Q ertable metabolic ethanol yield. The bacterial pathways were shown to be effective,

a
Q "I ieving higher ethanol production per unit cell biomass and metabolic yields

52

 

   

 

w

 

(com pared to the yeast). In particular, E. coli K011 showed robust growth and co-

fe rm entation in AFEX-CS water extract (Figure 13). This is contrary to the general
p e rception that E. coli has low growth robustness [27]. Fermentation pH (near-neutral
p H for K011) might be one of the key influences, as organic acids are less inhibitory in
dissociated form [68]. E. coli K011 was also able to produce and tolerate ethanol
concentrations at least up to 45 g/L (Figure 11,Table 4). Although the maximum ethanol
tolerance of E. coli might not be comparable to Z. mobilis or S. cerevisiae [27], the ability

to produce and tolerate ethanol between 40-70 g/L is probably sufficient for commercial

cellulosic ethanol production.

Compatibility between strain selection and biomass processing strategy
Th e selection of an ethanologenic strain for cellulosic ethanol production is highly
d ependent on the process design, particularly feedstock pretreatment selection and the

CO n figurations of the biologically-mediated processes (whether enzymatic hydrolysis

a n d fermentation are combined).

3- cerevisiae 424A(LNH-ST) is highly robust and able to ferment both glucose and xylose
re a sonably well, even at high solids loading, but likely requires more nutrients for
effective pentose-only fermentation (Table 5). E. coli K011 is able to tolerate a relatively
h i g ‘1 concentration of AFEX-CS degradation compounds and produces ethanol at a high

’7'". Q‘tabolic yield and rate. Unfortunately, xylose utilization in degradation compound-

QQ
htaining media (high solids loading) is severely affected. Z. mobilis AX101 appears to

53

m _

 

 

be an excellent ethanologenic strain for media with low levels of inhibitors due to its

su p erior metabolic yield and glucose fermentation rate (Table 5). Evidently, selection of
a p ret reatment with low inhibitory compound generation is the crucial factor,

p a rticularly for the Z. mobilis AX101 microbial platform.

The importance of a low-inhibitor-generating pretreatment

f

The feedstock pretreatment chemistries and temperatures dictate the profile (type and
concentration) of degradation compounds produced. AFEX produces substantially less

inhibitory compounds compared to pretreatments at acidic pH [2]. Compatibility

bElivveen pretreatment, fermenting strain and configurations of bio-mediated processes
may well the key for the viability/profitability of this industry. Reducing the production

levels of inhibitory degradation compounds would inevitably widen the range of suitable

eth anologenic strains and ultimately improve overall economics.

CO NCLUSIONS

Th e tested ethanologens are able produce ethanol from a CSL-supplemented co-

fe rI’Tientation at a metabolic yield, final concentration and rate greater than 0.42 g/g

Co "I sumed sugars, 40 g/L and 0.7 g/L/hr (0-48hr), respectively. Xylose-only fermentation

of t he tested ethanologenic bacteria is 5-8 times faster than S. cerevisiae 424A(LNH-ST)

in
t he CSL fermentation.

54

 

 

   

All tested strains grow and co-ferment sugars at 15% solids loading equivalent of AFEX-
pretreated corn stover water extract. However, both E. coli K011 and S. cerevisiae
424A(LNH-ST) exhibit higher growth robustness than AX101. In 18% solids loading
lignocellulosic hydrolysate from AFEX pretreatment, complete glucose fermentations
can be achieved at a rate greater than 0.77 g/L/hr. In particular, the glucose
consumption by S. cereivisiae 424A(LNH-ST) is higher than 3.1 g/L/hr. Furthermore, S.
cereivisiae 424A(LNH-ST) consumed xylose in lignocellulosic hydrolysate at the greatest
extent and rate, leading to its decided advantage in overall ethanol yield and titer over

other tested strains.

55

CHAPTER IV CELLULOSIC ETHANOL PRODUCTION USING AFEX AND

SACCHAROMYCES CEREVISIAE 424A(LNH-ST) AS PLATFORM TECHNOLOGY

INTRODUCTION

Ammonia fiber expansion (AFEX) has been shown (in Chapter II) to be an effective
pretreatment method for generating highly fermentable substract [3]. Saccharomyces
cerevisiae 424A(LNH-ST) produced ethanol at the highest overall yield and rate
compared to other tested ethanologens in Chapter III [57]. In this chapter, I seek to take
advantage of the promising features of AFEX pretreatment and S. cerevisiae 424A(LNH-
ST) to formulate and demonstrate an industrially-relevant strategy for fermentation

using lignocellulosic biomass.

The requirements for: i) high starting cell density, ii) hydrolysate conditioning and iii)
nutrient supplementation to conduct lignocellulosic fermentation are evaluated. I also
construct a comprehensive mass balance based on conversion of carbohydrates in
untreated biomass to ethanol to elucidate the current status and the bottlenecks of the

technology. To better understand the phenomenon of slow xylose utilization in

56

hydrolysate, the interactions between degradation products, xylose metabolism, cell

growth and media nutrient content are described.

MATERIALS AND METHODS

Corn stover (CS)

The composition of the corn stover was as described on page 14

AFEX pretreatment

The AFEX pretreatment was conducted as described on page 14

AFEX-corn stover (AFEX-CS) water extract

Washing of CS was conducted by spraying distilled water on AFEX-treated CS at a ratio
of 1 g dry CS to 5 mL of water (20% solids loading) as described on page 16. Washed CS
solids were then enzymatically-hydrolyzed. This water extract was used for

fermentation studies.

Enzymatic hydrolysis
The AFEX-treated corn stover was enzymatically-hydrolyzed by commercial enzymes
mixtures. The cellulase mixture consisted of Spezyme CP [86.7mL/kg CS; 15 FPU/g

cellulose] and Novozyme 188 [43.7 mL/kg CS; 32 pNPGU/g cellulose]. The hemicellulase

S7

mixture was Multifect Xylanase [12.7 mL/kg CS] and Multifect Pectinase [8.9 mL/kg CS].
The spectrum of activities for the commercial enzymes were as reported [44]. The
Spezyme and Multifect enzymes were obtained from Genencor Inc (Palo Alto, CA) and
Novozyme 188 was purchased from Sigma-Aldrich Co (St. Louis, MO). These enzyme
mixtures, unless otherwise stated, were used in all hydrolysis experiments. The cellulose
loading for the hydrolysis was kept at 6.0% by weight (60 g of cellulose per 1 kg of
hydrolysis mixture) which corresponds to 17.6% and 16.2% solids loading, in unwashed
and washed AFEX-CS respectively. The reaction was carried out for 96 hr at pH 4.8 (0.05
M phosphate buffer), 50°C and 250 rpm agitation. Each hydrolysis was conducted in a
1.0 L baffled flask with 500 g total saccharification mixture. Chloramphenicol (Cm) was
added to a final concentration of 50 mg/L to minimize the risk of contamination. After
96 hr, the hydrolysis mixture was centrifuged twice at 5,000xg for 30 min to separate
the liquid solution from unhydrolyzed solids. The supernatant (not sterilized) was used

for fermentation.

Microorganism and seed culture preparation

Xylose-fermenting Saccharomyces cerevisiae 424A(LNH-ST) was obtained from Dr Nancy
W. Y. Ho from Purdue University through MSU Material Transfer Agreement (MTA06-
119). Details of the metabolic engineering of this yeast have been reported [64, 69]. To
prepare seed culture, the strain was grown on YEP (5 g/ L yeast extract, 10 g/L peptone,
20 g/L glucose) plate for one to two days at 30°C. The cells were transferred to liquid

YEP media supplemented with 50 g/L glucose in an unbaffled flask. The 424A(LNH-ST)

58

seed was grown overnight at 30°C, 150 rpm agitation, under a largely anaerobic
conditions. The cell density of a typical 424A(LNH-ST) overnight seed culture reached 15

units absorbance at 600nm.

Fermentation

Fermentations were conducted at a working volume of 70 mL in a 250 ml unbaffled flask
at 30°C, initial pH5.5, 150 rpm agitation. A designated volume of seed culture was
centrifuged (15,000xg) for 5 min and the yeast cell pellet was resuspended into the
media to initiate fermentation. The flasks were capped with rubber stoppers pierced
with a needle to vent carbon dioxide formed during fermentation. Samples were taken
during the course of fermentation. Cell density was measured using a
spectrophotometer at 600nm (DU Series 700 UV/Vis, Beckman Coulter). One unit of
absorbance is approximately equal to 0.55 g (dry-wt-cell)/ L. Glucose, xylose, glycerol

and ethanol profiles were measured using HPLC as previously described [3].

Complex media and hydrolysates

Three different fermentation media were prepared: (i) complex media (YEP + 70 g/L
glucose+ 40 g/L xylose); (ii) hydrolysate I: AFEX-CS hydrolysate at 6.0% glucan loading;
(iii) Hydrolysate ll: AFEX-CS hydrolysate at 6.0% glucan loading with supplemental sugars
(10 g/L glucose + 12 g/L xylose), which is the monomeric sugar equivalent of the
oligomeric sugar content in the hydrolysate. The initial pH of the hydrolysates was

adjusted to 5.5 using reagent grade KOH. The starting cell density of fermentation was

59

at 2.0 (OD600nm) which is approximately 1.1 g dry-cell-wt./L. Hydrolysates were
fermented without prior detoxification or nutrient supplementation. Volumetric
productivities of glucose, xylose and ethanol were calculated from their concentration
gradients over the first 6, 12, 24 hr of fermentation, respectively. Specific productivities
(g/L/hr/g cell) were calculated by dividing volumetric productivities by the respective
values for dry cell mass. Metabolic ethanol yield was estimated based on total

consumed glucose and xylose. The theoretical maximum yield was 0.51 g EtOH/g sugar.

Effect of pH, temperature, initial cell density, nutrient supplementation and washing
on xylose fermentation in hydrolysate

Several fermentations were carried out to investigate the effect of (i) pH, (ii)
temperature, (iii) initial cell density, (iv) nutrient supplementation and (v) minimal
washing on xylose utilization in the cofermentation of hydrolysate. Enzymatic hydrolysis
was carried out with the cellulase mixture only. Initial glucose and xylose concentrations
in the hydrolysate were 55:2 g/L and 23:1 g/L, respectively. Other experimental details

were as listed in Table 6.

60

 

Table 6 Summary of parameters for four experiments examining the effects of pH,
temperature, initial cell density, and washing and nutrient supplementation on

xylose fermentation of corn stover hydrolysate

 

 

 

 

 

 

 

 

 

 

 

 

 

Experiment Effect of Effect of Effect of
Effect of pH Tem era ture Initial Cell Washing and
Parameter p Density Nutrient Suppl.
. . 3.5, 4.5,
Initial pH 5.5, 6.5 5.5 5.5 5.5
Temfcrfture 30 25,30,35,37 30 30
Initial Cell
Density 0.5 0.5 1256’ $120 0.5
(OD600 nm) ' ’ '
Yes
Washing on C5 No No No (Minimal
Washing)
Yes
Nutrient (10 g/L Yeast
Supplementation NO NO NO Extract, 20 g/ L
Peptone)
Relevant Figure Figure 16A & B Figure 16C Figure 160 Figure 18

 

Water extract (With or without YEP supplementation)

The water extract was used to investigate the effects of soluble compounds on AFEX-CS

on fermentation. Each fermentation medium contains the water extract, 10 g/ L glucose,

40 g/L xylose and 50 mg/L Chloramphenicol. YEP (5 g/L yeast extract and 10 g/L

peptone), was added to create nutrient-rich conditions. The final concentrations of the

water extract in the fermentation media ranged from 0 to 16% solids-loading-

equivalent. The reaction was carried out at 15 mL working volume in 20 mL screw-

capped vial. Three glass beads (1 cm in diameter) were added into each vial to aid

61

 

 

stirring. Cell density and specific xylose consumption over the first 24 hr of fermentation

were estimated.

Corn stover to ethanol analysis
To construct a corn stover to ethanol mass balance analysis, sugars (glucose and xylose
in monomeric, oligomeric and polymeric forms) and ethanol content were measured

before and after each process, i.e. pretreatment, enzymatic hydrolysis and

Ln-

fermentation.
Around pretreatment
The total weight of dry corn stover before and after pretreatment was recorded.
Glucan and xylan contents (as a percentage) of untreated and AFEX-treated corn
stover were estimated using the NREL protocol (LAP-002).
Around enzymatic hydrolysis
A known amount of pretreated corn stover was added into the system. Total sugars
solubilized in the liquid phase or remaining in unhydrolyzed solids were determined.
Soluble oligomeric sugars in the liquid stream were estimated according to the NREL
protocol (LAP-014).
Preparation of washed unhydrolyzed corn stover dry matter
Prior to the analysis, unhydrolyzed corn stover was washed twice using distilled
water at a ratio of 1 g wet unhydrolyzed corn stover to 5 mL of water. The mixture
was centrifuged for 30 min at 6,000 xg to separate the washed pellet from the

supernatant. The volumes of water extracts were recorded, and the concentration

62

of glucose and xylose were estimated. After washing, a representative sample of
unhydrolyzed corn stover was dried under vacuum at 60°C (Vacufuge, Eppendorf)
overnight. The glucan and xylan contents in this dry matter were estimated using
the NREL protocol (LAP—002). The remaining solids were dried in an oven at 110°C
and the total dry weight of the washed solids (from two different dryers) was
recorded.

Sugars and ethanol estimation in liquid streams

The liquid streams after saccharification were analyzed to determine ethanol,
monomeric glucose and monomeric xylose concentrations. To estimate the
concentration of both oligomeric and monomeric sugars in the stream, acid
hydrolysis (H2504) was conducted to hydrolyze sugars from oligomers to monomers.
Experimental details were as described in the NREL protocol (LAP-014).

Mass closure analysis:

Mass closure (as a percentage) was calculated by dividing total sugars (glucose or
xylose) exiting the unit by total input sugars. Total output sugars were then
normalized to 100%.

Around fermentation

The amount of input sugars into the fermentation unit was known. Total ethanol
produced and the remaining sugars (both monomeric and oligomeric) after
fermentation were estimated. The process yield around fermentation was calculated

using Eq.5.

63

 

 

v

p
[ii

tar

\
‘i‘x

"“J‘ilé

 

[EtOH)f - [EtOH)i

- Yie|d(%)=
[Gk+thx051

 

x 100% --------- Eq 5

Where [EtOH] is concentration of ethanol, [Glc+Xyl] is total concentration of glucose
and xylose, f and i denote for final and initial conditions, respectively. Data collected
from the analyses were adjusted to the basis of 1.000 kg of dry untreated CS.
Detailed experiment procedure on mass balance construction was as described in

Table 7.

Table 7 Experimental details on the construction of mass balance

 

Unit Operation Recorded Data Analysis Data

Glucan and xylan contents (in
percentage) of untreated and
AFEX-treated corn stover were
estimated using NREL protocol
(LAP-002)

Glucan and xylan contents (in
percentage) using NREL protocol
(LAP-002)

 

AFEX Total dry weight of corn
Pretreatment stover before and after AFEX

 

Total weight of AFEX-treated
CS entering the system

 

Total volume of liquid
Enzymatic hydrolysate

Hydrolysis Total volume of water extract
of unhydrolyzed solids

Concentration of both oligomeric
and monomeric sugars in the
stream (NREL LAP-014).

 

 

Glucan and xylan contents (in
percentage) using NREL protocol
(LAP-002)

Concentration of both oligomeric
and monomeric sugars in the

Total volume of hydrolysate stream before and after
Fermentationa entering the system fermentation (NREL LAP-014). Final
ethanol concentration and cell
density (absorbance at 600nm) was
measured.

3 Volume change during fermentation is assumed negligible

Total dry weight of washed
unhydrolyzed solids

 

 

 

 

 

 

64

 

RESULTS

Optimization of fermentation conditions

Optimal conditions for fermentation of AFEX-CS-hydrolysate are pH 5.5 and 30°C
(among the conditions tested) with respect to overall fermentation yield and rate. While
fermentation at pH 6.5 exhibited the highest xylose consumption rate, it had the lowest
metabolic yield (79.5%) among tested pH values with the highest level of glycerol (5.0
g/L) formation. Judging from the overall ethanol yield, pH 5.5 was determined to be the
optimal pH (Figure 16A &B). Regarding temperature optimization, xylose utilization in
fermentations at 35°C and 37°C essentially ceased after 72 hr; 10.9 g/L and 16.1 g/L of
xylose were left unconsumed. Xylose fermentation at 25°C was slower than at 30°C

(Figure 16C).

The initial rate of xylose consumption is directly correlated to the initial cell density
(Figure 160). However, the difference between the extent of xylose consumption (after
144 hr) was rather small (<3 g/L). Final ethanol yield was almost independent of initial
cell density. In order to be more industrially-relevant, relatively low cell densities (51.1 g
dry-cell-wt./L) were used. With the exception of hydrolysate fermentation at pH 3.5,
glucose fermentations were completed within 18 hr (Glucose consumption profiles are

not shown in Figure 16).

65

 

 

 

 

 

 

 

 

 

 

A25 ' I V I V I r I ' I V 35 ' I v I v I V I V I fl
M 1 BBQ; , .
20. . ‘ l ‘ ‘ . _ . ‘
3 u . . _. 4 j 25“ ‘ : I,
d /
s - a /
‘J ‘ A i ‘c’ 4 /
C 10. ‘ i O 15 ° / ‘
8 + pH3.5 . U //—I— pH3.5-
—>~ _._ pH4.5 ‘ g 10 / ”—.‘pH45‘
>< 5‘ A ‘ ... A HS
pH5.5 u; 5 A p .
0 ‘r pH6.5 0* 0; _ _ _/ pH6.S‘
0 20 4To 6'0 8'0 100 120 0 20 40 60 80 100 120
25 f I ' I V I V I ' DZS v r fi f 1

 

20‘

 

 

 

 

 

 

 

A “k A +oo=0.s

—J ', V v V v —’ . 00:6-0

\ 15- ’. ‘\_ -l \

£9 . q‘ \\ .92 A 00:12.0

U ‘ ‘ 7 U .

C J \a ' ‘d c * 00:18.0

8 10 +25 (deg)C“‘» \“ 8

— l+~30(deg)c "- ‘ _ . \‘

>~ _‘ 7’“ O.‘ >‘ _ ‘ _- ..

x 5 **f 35 (deslc x 5 I “"7““:
0‘ V 37 (deg)C ‘ 0‘ ‘
0 ' 3'0 ' 60 ' 90 15.0 150 0 3'0 ' 6'0 ' 9'0 ' 1&0 ' 150

Time (hr) Time (hr)

Figure 16 Effect of (A, B) pH, (C) temperature and (D) initial OD on fermentation using
hydrolysate from enzymatically-digested AFEX-treated CS using Saccharomyces

cerevisiae 424A(LNH-ST).

(Glucose consumption profiles are not shown in this graph). Note: Ethanol production
from 0 to 25 g/L shown in Panel B is largely due to glucose fermentation

Laboratory media vs AFEX-CS hydrolysate

Fermentations using 424A(LNH-ST) in complex media (YEP) and AFEX-CS-hydrolysates
derived from saccharification at 6% glucan loading were compared side-by-side under
identical conditions (Figure 17). These hydrolysates were fermented without
conditioning (no washing, nutrient supplementation or detoxification) at 1.1 g dry-cell-

wt./ L starting cell density. The difference between Hydrolysate l and II is that an

66

additional 10 g/L glucose and 12 g/L xylose were supplemented into Hydrolysate II to
examine the scenario in which soluble oligomeric glucose and xylose in the hydrolysate

were also utilized.

Strain 424A(LNH-ST) was able to grow well in both complex media and AFEX-CS-
hydrolysates and achieved cell densities greater than 6.0 g/L within 12hr of
fermentation (Table 8; Figure 19). In complex media, 70 g/L glucose and 40 g/L xylose
were completely consumed within 48 hr (Figure 17A). The volumetric glucose
consumption rate was 7.3 g/L/hr; 0—6 hr, which was roughly seven-fold higher than the
xylose consumption rate. For fermentations in AFEX-CS-hydrolysate, volumetric xylose
consumption rates were an order of magnitude lower than for glucose. This indicates
that xylose utilization was more susceptible to inhibition. Surprisingly, specific glucose
consumption rates in hydrolysate achieved 10.6 g/L/hr/g cell; 0-6 hr, substantially
higher than that in complex media (Table 8). Fermentations in the hydrolysates achieved
higher metabolic yields with lower xylitol formation (0.46 g ethanol/g consumed sugars;

0.3 g/L xylitol) than complex media (0.43 g ethanol/g consumed sugars; 3.2 g/L xylitol).

67

 

 

 

 

 

 

 

 

 

 

 

 

I .
A 60- 3 + -
50- - ~ «
4 l ‘ <
..l 40. . _

E < +GIC 4 1 j
LC) 30' +Xy| ‘ J a
o q -—+—EtOH . ‘ ‘
U 20 < - -
loj -l - l

O V IL' I f I ' 0 I v! I ' I ‘ I ' I 0 | v- 1 v I v I v I '

O 20 40 60 80 0 40 80 120 160 0 40 80 120 160

Time (hr) Time (hr) Time (hr)

Figure 17 Fermentation of (A) complex media; (B) CS-Hydrolysate l; (C) CS-Hydrolysate II
using S. cerevisiae 424A(LNH-ST).

Fermentations were initiated with 1.1 g (dry-wt.)/L of 424A(LNH-ST) inoculum, carried
out at 30°C, pH5.5 and 150rpm under largely anaerobic condition. The hydrolysates
were neither detoxified nor externally nutrient-supplemented.

Washing removes both degradation products and biomass nutrients

The impacts of washing AFEX-pretreated CS (prior to enzymatic hydrolysis) on xylose
utilization and the requirements for nutrient supplementation were investigated.
Washing of AFEX-CS does not improve xylose utilization unless additional nutrients are
provided. Xylose fermentation for unsupplemented washed-CS hydrolysate had the
lowest rate (0.12 g/L/hr, 0-96hr) followed by unsupplemented unwashed-CS hydrolysate
(Figure 18). This observation is attributed to the loss of nutrients through washing.
Besides removing degradation products [70], the water extract of AFEX-CS contains
residual ammonia from pretreated materials and other biomass components that are
important nutrient sources for fermentation (Table 11, Table 12 in Chapter V). With YEP

supplementation, washed-CS hydrolysate had a better xylose fermentation (0.20 g/L/hr,

68

0—96hr) than unwashed-CS. Evidently, even under nutrient-rich conditions, the presence

of degradation products affected xylose utilization.

 

24 I I I I ' l ' ' ' I
. \ —-I— Unwashed_5uppl
+ Unwashed w/o Suppl _

+ Washed_Suppl
+ Washed w/o Suppl

N
O
1

5.;
m
1
1

Concentration (g/L)
9° . F3

A
l
1

 

 

 

O ' I ' I ' I ' I T I
O 20 4O 6O 80 100

Time (hr)

Figure 18 Effect of washing and nutrient supplementation on xylose consumption in the
fermentation of hydrolysates from enzymatically-digested AFEX-treated CS.

(Glucose consumption profiles are not shown in this graph)

The effects of soluble substances from AFEX-CS on xylose fermentation

The relationship between xylose fermentation with the levels of degradation products
and nutrient content in fermentation media was further investigated using the water
extract from AFEX-CS. In nutrient—rich conditions, cell growth decreased as the
concentration of AFEX-CS water extract increased. However, the level of specific xylose
consumption remained virtually constant at 0.25 g/L/hr/g cell; 0-24hr (Figure 20A). This

suggested that xylose utilization was lowered due to the reduced cell mass production

69

..\..>>-__mo->.n m mmd o. .33 2038.68.an m. .558 am 8:36QO .o “E: «:0 5053585. ..o E em .3? 3.3% :3 .
..mmsm 3m Hmd mm 383...... 83 22> 10m. 30.5.85 .03.? new 382m 38:38 .33 co woman 3.23.3 22> 2.832)....
3.02638. ..r. «N new NH6 3...» o... .06 3.26:8... 3...

6.3293; 2.. .._ Emacou .mmnm o..mEom_.o on“ .o Em.m>.:cm .mmsm 25895:. of m. 5...; .30.? .\m NH .25 882m ..\m 3 .0
Emma 5580.33 5.? 000m £33 an ....mm .8 msnmo. cmoEm 9% am umummm_u->_.mozmc.>~:m. mu toummbkmrz 80.. 3820.3... u
5.8 .wéza um .58 .8 mcimo. 583% $0 um nofimmfi$=83m§>~c$ mu nmummbkmmd. :8: 3320.3... 0
5:93. 303. .R 3 + Umbxm Hmmm> 9.6mm. <m m a
3\.ts-..wo->.n m H.H
um mm; Bacon ..8 .mEc. £83.38 050.698 2092 .09.: cozmgmm EEOmH new m9... 0. om Hm 3639.8 0.03 2033585. a

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Gd“ m6“ H6“ 06 wdw odh odn v.0“ mdfi
. i . . . 83> o. > -
H 0.2.. m ohm odm mm SH. 3:. wd 0.0. mm. Q? m mm O we a: _ n I 8 O
No m.ww m.~m mé mH- ms- wd m6- m.m- ode me msm U. 3320.3..-3 m
ad... mdw H.o.+. de mdh o6“ Hdu Nd“ Nd“
. H . . .
N o m m o.mm o.mm mé m.~- We- md H.H- ms- one o mm o mm a A...» <
.._?s s... .x. . :9. . 3x . 0.0 . :om . 3x . sol .._}. .3. so
53 23.223 :33 2.8 .Eul.
33> .. 23> a 3.3.2
_ 3.2.2. . 3.26:3... 3.33:3... :05 0:8 .33
$30... H.822
:3 053% u_.uoE:_o> .2... ..qucos.

 

 

 

 

 

 

 

8.8283,. new 3.8... .6388 .8 330528 5.3.585. we EmEEsm m 03......

7O

 

>

 

Cell Dry Weight (g)

 
  
  

+YEP

 

—0— Hydrolysate I

+ Hydrolysate ll ‘

 

'1'0‘2'01310'4'0's'otsTo'7'o'80

Time (hr)

 

 

w

.o .o 3" l" l"
m m 0 N b
1 4 1 1 .

0.44
J
02*

Cell Density (g/L)

 

.0
o

T I ' l ' T ' l ' l

+ Control

- —0— Enzyme Solution
: + 5% Water Extract

 

 

Figure 19 Cell density profile of the fermentation using 424A(LNH-ST)

Time (hr)

(A) Cell density profile of the fermentations reported in Figure 17

(B) Fermentation was conducted in 100 mL total volume, at 30 °C, pHS.5 and 150rpm,
the initial cell density was at 0.055 g (dry-wt)/L. Control: SOmM phosphate buffer (PB),
Enzyme Solution: 50 mM PB+ 0.43 mL Spezyme CP+0.22 mL Novozyme 188+0.064 mL
Multifect Xylanase+0.045 mL Multifect Pectinase. 5% Water Extract: 50mM PB in final
water extract concentration at 5% solids-loading equivalent. Glucose and xylose (both at

25 g/L) was supplemented to each fermentation.

71

in the presence of degradation products, while xylose metabolism in cells was
apparently not affected. In nutrient-limiting conditions, where the AFEX-CS water
extract was the sole source of nutrient, cell growth is proportional to the concentration
of the water extract. The relationship between the specific xylose consumption rate and
the water extract concentration exhibited a left-skewed curve with the highest rate
(0.15 g/L/hr/g cell, 0-24hr) at 4% solids loading water extract (Figure 208). These results
imply that, in nutrient-limiting conditions, (i) a low level of degradation products
resulting from AFEX pretreatment stimulated xylose metabolism within cells, but the
rate of xylose utilization was limited by cell growth (nutrient availability) and (ii) at high
water extract concentration (28% solids loading), inhibition of xylose metabolism was
the primary factor affecting xylose utilization. Under both conditions, biomass
degradation products resulting from the pretreatment (soluble products in the water
extract) were shown to increase metabolic ethanol yield (Figure 20A & B). Fermentation
using a water extract at 16% solids-loading-equivalent as the sole nutrient source
(Figure 203) achieved comparable cell growth to the YEP-supplementation fermentation

(Figure 20A).

Current status and bottlenecks of the technology

Current technology using corn stover as feedstock, AFEX as the pretreatment
technology and S. cerevisiae 424A(LNH-ST) as the ethanologenic strain in separate
hydrolysis and fermentation (SHF) with a starting cell density at 1.1 g dry-cell-wt./L, was

able to achieve 191.5 g EtOH/kg untreated CS (Figure 21), 60.8% of the theoretical

72

flall
111..

n» A
V'Iuk

maximum yield, at an ethanol concentration of 40.0 g/L (5.1 v/v%) without the need for

washing, detoxification and nutrient supplementation.

 

J

10 ' I ' T ' T ' I l l ' 0'6

0 30

g

m

. A H‘

‘6’-
( O“ ‘J‘5 m
8" g :30
A '<‘ O
< i- 030 :7:
00 B; x
V m‘ E
> m r‘)
L" 8— us 9
W I S
0:) a?) 3
Q Q- magi
_ 3 C
— m4 3
a) 3 1
U m— 43,5 ‘7‘
c n
on E
2* 3

V.

VH DUO

 

 

 

 

..a

O

'l

O

03
l

 

J

l

. I Metabolic Yield“ \'
6 . [3:3 Cell Density » _ - 0'5
+ Spec Xyl Cinsm i ‘ .

g1

' i 0.4

1

Cell Density (g/L)

(“03 8/1u/1/8)ulsu03|Ax junadg

 

(swans wsuog 8 Mg}; 8) ppm anoqelaw

 

 

 

.0
w

o 2 4 8 12 16
Wash Stream Conc (% Solids Loading Equivalent)

Figure 20 Effect of AFEX-CS water extract concentration on cell growth, metabolic yield
and specific xylose consumption rate under (A) nutrient-rich and (B) nutrient-
limiting conditions.

Fermentations were initiated with 0.3 g (dry-wt.)/L of 424A(LNH-ST) inoculum, carried
out at 30°C and 150rpm. Data points presented were at 24 hr.

73

(ll

‘\ .

U4
5‘
LU

During AFEX pretreatment, there was no sugar loss, total material mass was increased
by 1-2%, probably due to the ammonia binding onto the biomass [71]. The mass balance
around the enzymatic hydrolysis step achieved 98.9% and 107.1% closure for glucose
and xylose, respectively. In enzymatic hydrolysis at 17.6% solids loading, 85.8% of the
total input sugars (glucose and xylose) were hydrolyzed and solubilized, of which, 78.2%
was hydrolyzed to their monomers (Figure 21; Table 9). About two-thirds (62.0%) of the
total oligomeric sugars were xylose. Fermentation using 424A(LNH-ST) effectively
converted monomeric glucose and xylose with 88.5% ethanol yield. As expected,

oligomeric sugars were not utilized by 424A(LNH-ST).

The three process parameters having the strongest influence on ethanol production
economics are yield, titer and rate. Enzymatic hydrolysis at high solids loading has been
identified as the primary bottleneck affecting overall yield and titer for the cellulose-to-
ethanol bioconversion. One-third of the total output sugars were oligomers or polymers
which could not be utilized by 424A(LNH-ST). To improve overall ethanol productivity,

efforts should focus on increasing the xylose consumption rate during fermentation.

74

 

.0_.0Eom..o “9.0 “0:050:22 ”0.). .03.; ”.>x 08020 ”0.0 8.02.2.9: 0320?. 0... 2... 0080.30 0.0
m0 5 8:00:00 30.02.0050 .8063. Av.0..0m 2S5 $0.5 8 0:032:00. 9:08. :83.» $0.0 8 08030.80 83 09.0.02. 0.02.35
5.23 0.:0mo.oc050 05 mm A 5-12.73? 00330.00 .0. “Em $205.00. EmEHmmbwa 0.080.008 8 x92. .8 0.0me mm? 0.7.205 0;...

8.9.0.5 00:22. 80:. .0550 o. .0300. 500 HN 0.3mm

_>x m edflam 22.8
0:0 m H.303 08290.55

           

3x .06 m 0.84.? 3x .36 u 3.3.:
0.0 .95 m @393 0.0 .95 m 0.98.?
_>x .22 m 0.30.: E. .22 m miemd
2.00 umm0> m wdeKN 0.0 .05. w deHdwN

           

        

, 3x 0 o.mnm.mm~
20 m 33.8.“.

.3. w 0.03am
20 m 38.0.».

    

            

       

.035 500 .
9. mHodHoSH _

.9on Eou
0. 80.9884

     

Em 9. a.
:9.
m 0.03::

 

3820.3:
230:

E2300... 00mm.» 08.02.00
33 EB m Nd $855

75

038.32 .oz “<2

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. I . . I . H
00H H 2+0 20 <2 00H 0 m+w Hmm <2 00H .NHHmem <2 .30..
N: ~.30.B <2 Ema c.0053 <2 3: H.memm <z 0050.035
00.0 0.000.0mm 0.~H~.00H 000 0.03.02 ~.0Hm.0v m.mw 0,000.03 0.~H<.$ .2033 02820.03.
Nwfi mafimfifi fifiwmdm 0.0m 0.3%: 130.3 v.2 0.000.? 0.30.0 .0Eom..o 050..
NS 0035‘? 0.30.00 0mm. 03.0.02 532K N4; 0.03.03 fimfllwm .0Eo:o.>. . .
omhwwwou m .\u 00%”.8 u ..\m 00MWW00 0 <0
00.0.; + 0303.0 033x 0303.0

 

 

 

 

.co.um~..mE.oc .0000. mu-xm.u.< 0508. 03.2.00 $0.0 .0 $9.20.... 0.008520 80.. 0.02000 0.00:0. 39:0 0 033

76

DISCUSSION

Significance of the process integration

Despite abundant published research on ethanol production from lignocellulosic
materials, comprehensive system-wide studies with industrially-relevant performance
metrics are still lacking. Washing and detoxification steps, though often incorporated to
improve the fermentability of the hydrolysate, may well be prohibitively expensive [38].
The straightforward cellulosic ethanol technology reported here, which converts sugars
from lignocellulosic materials to ethanol without washing, detoxification and nutrient
supplementation, is significant in the search for a highly competitive cellulosic ethanol
production strategy. A final ethanol titer of 40 g/L, a benchmark concentration for

commercial cellulosic ethanol production, was achieved.

Using AFEX as the pretreatment and S. cerevisiae 424A(LNH-ST) as the fermenting strain
were the key innovations responsible for achieving this progress. Hydrolysate from
lignocellulosic biomass is generally regarded to be nutrient-deficient [72]. However, our
results strongly contradict this perception. The perceived nutrient deficiency is likely
due to pretreatment at high temperature and acidic pH followed by washing which
degrade and/or remove nutrients. In contrast, nutrients provided through AFEX-
pretreated corn stover were sufficient to support robust yeast growth. The
saccharolytic enzyme preparation provides relatively little nutrient value (Figure 19).

Cellulosic ethanol technology therefore need not depend on commercial nutrient

77

supplements such as yeast extract or corn steep liquor. Yeast cells can also be produced
as a valuable co-product using this platform. A recent publication using E. coli K011 to
ferment enzymatic hydrolysate from AFEX-treated corn stover has also confirmed the

general fermentability of corn stover [3].

Important steps in improving yield and titer I .1.

However, further improvements in overall yield and titer are required to make this

 

technology more commercially-attractive. Results presented herein suggest that efforts
to increase the availability of fermentable sugars at high concentrations deserve the
highest priority. This can be done through (i) optimization of the enzyme formulation
for effective saccharification of both cellulose and hemicellulose; (ii) development of
ethanologenic strains which consume a wider range of substrates (both mono- and
oligosaccharides) and (iii) integration of enzymatic hydrolysis and fermentation in the
configurations of simultaneous saccharification and co-fermentation (SSCF) or
consolidated bioprocessing (CBP) to alleviate sugar inhibition at high solids loading and
reduce enzyme dosage. In this report, enzymatic hydrolysis and fermentation have
been conducted separately in order to facilitate understanding of the individual
processes. Nevertheless, simultaneous saccharification and fermentation (SSF) was
shown to be beneficial compared to separate hydrolysis and fermentation (SHF) [73], in
spite of a lower reaction temperature that was adopted to allow fermentation at the
expense of greater efficiency during enzymatic hydrolysis. The full potential of these

configurations will be better realized if both hydrolysis and fermentation share common

78

optimal conditions. Developing an oligosaccharide-utilizing strain will be particularly

advantageous for AFEX-centered cellulose technology as a sizeable portion of sugars in

the hydrolysate are oligomers. As a near-future projection, the overall yield could be

increased to 251.4 g EtOH/kg untreated CS with final titer of 52.5 g/L or 6.7%(v/v)

(Figure 168) if complete utilization of soluble sugars is achieved (Table 10).

Table 10 Future projections for ethanol yield and titer based on 6% cellulose loading of

 

 

 

 

 

 

 

 

 

AFEX-CS
Areas of Conc Conc
Label Improvement g/kg gall ton Lg! L) (%v/v)
current CT N.A. 191.46 64.10 40.00 5.07
Technology
CT + Complete
NFT1 Monomeric Xylose 196.69 65.86 41.09 5.21
Utilization
NFT1 + Half of Soluble
Near-Future NFT2 Oligosaccharides 224.04 75.01 46.81 5.93
Technology Utilization
NFT1 + Complete
NFT3 . somb'e . 251.39 84.17 52.52 6.66
Oligosaccharides
Utilization
NFT3 + Half of Sugars
MT1 Utilization in 272.15 91.12 56.86 7.21
Mature Unhydrolyzed Solids
Technology NFT3 + Complete
MT2 Sugars Utilization in 292.92 98.07 61.20 7.76
Unhydrolyzed Solids
Theoretical No Yield Loss in Any 314.96 105.46 65.80 8.34
Maxnmum Stage

 

 

 

 

 

 

 

 

 

79

Interaction between degradation products and xylose fermentation

Xylose fermentation was more susceptible than glucose fermentation to inhibition by

the degradation products from AFEX pretreatment, resulting in prolonged fermentation.

However, as our results indicate, the effects of AFEX degradation products on

fermentation are complex. Generally, AFEX degradation products increase metabolic

yield by reducing the formation of fermentation by-products. Certain degradation .._,

compounds have been postulated to act as electron acceptors to provide redox balance

 

in xylose metabolism [58, 74]. An equally important benefit of the degradation products
from AFEX-pretreated biomass is that they increase metabolism of sugars which
translates into higher specific ethanol production rates. Degradation compounds such as
organic acids and 4-hydroxybenzaldehye have been shown to stimulate fermentation
when present at moderate levels [16, 75]. Nonetheless, they inhibit the cells from
propagating to a density warranted by available nutrients. In a commercial setting, a
two-fold increase from the existing xylose consumption rate is likely required. Instead of
supplementing with excess nutrients or detoxifying hydrolysates, improving xylose
fermentation through cell recycle to increase cell density might be a more cost-effective

solution.

CONCLUSIONS
Current technology using corn stover as feedstock, AFEX as the pretreatment

technology and Saccharomyces cerevisiae 424A(LNH-ST) as the ethanologenic strain in

80

separate hydrolysis and fermentation (SHF) was able to achieve 191.5 g EtOH/kg
untreated corn stover, at an ethanol concentration of 40.0 g/L (5.1 v/v%) without
washing of pretreated biomass, detoxification or nutrient supplementation. Enzymatic
hydrolysis at high solids loading was identified as the primary bottleneck affecting
overall ethanol yield and titer. Degradation compounds in AFEX-pretreated biomass
were shown to increase metabolic yield and specific ethanol production while
decreasing the cell biomass generation. Nutrients inherently present in corn stover and
those resulting from biomass processing are sufficient to support microbial growth
during fermentation. This platform offers the potential to improve the economics of
cellulosic ethanol production by reducing the costs associated with raw materials,

process water and capital equipment.

81

 

 

CHAPTER V A NEW PARADIGM FOR CELLULOSIC ETHANOL

PRODUCTION: UTILIZATION OF AFEX- PRETREATED CORN STOVER AS

SELF-SUSTAINED CARBON AND NITROGEN SOURCES

INTRODUCTION

Besides carbohydrates, plant materials contain other extractives including nutrients
[76]. Although present at relatively low levels compared to carbohydrate and lignin,
plant nutrients can be vital for a cost-effective lignocellulose bioconversion. Further
understanding of nutrient characteristics of plant hydrolysate is required to facilitate

rational approaches for fermentation optimization.

In this chapter, the nutrient content of enzymatic hydrolysate from AFEX pretreated
corn stover at high solids loading, identifying potential excess or limiting nutrients is
investigated. I seek to demonstrate biomass processing and bioconversion strategies
which feature (i) utilization of AFEX—pretreated corn stover as the self-sustained carbon
and nutrient source, (ii) complete and rapid fermentation of AFEX-CS hydrolysate at high

solids loading and (iii) conceptual proof of the feasibility to utilize nutrients inherent

82

from AFEX-pretreated corn stover to support saccharolytic enzyme production for the

hydrolysis of lignocellulosic biomass.

MATERIALS AND METHODS

Corn Stover (CS)

The composition of the corn stover was as described on page 14

AFEX pretreatment

The AFEX pretreatment was conducted as described on page 14

Enzymatic hydrolysis at 18% solids loading

AFEX-CS was hydrolyzed using both cellulase and hemicellulase commercial mixtures.
The cellulase mixture consisted of Accelerase 1000. The hemicellulase mixture was
Multifect Xylanase and Multifect Pectinase. Enzymatic hydrolysis at total mass of 2.0 kg
(18.0% solids loading) was performed for 96 hr at pH 4.8, 50°C and 1200 rpm agitation in

an bioreactor (Biobundle 2L, Applikon Inc).

Corn stover and enzyme feeding strategies
To achieve proper liquefaction and stirring throughout the enzymatic hydrolysis. Both

AFEX-CS and commercial enzymes were fed in batches. AFEX-CS was fed in 5 batches

83

(6.0%, 3.0%, 3.0%, 3.0%, 3.0% solids loading) with respective intervals of 2, 1, 1, 1.5, 1.5
hr between feeding. Regarding enzyme feeding, two-thirds of the total enzymes was
added during the first 8 hr ofthe feeding (1/3, at 0 hr; 1/6, at 4hr; 1/6, at 8hr). The
remaining one-thirds of the enzymes were fed continuously to the reactor for 36 hr
(1/6, 8-24 hr; 1/6, 24-48hr). A total protein loading of 6.7 mg protein/g biomass was
used. The enzyme loading in volume is 106.1 mL Accelerase 1000, 12.4 mL Multifect

Xylanase and 12.4 mL Multifect Pectinase per 1.0 kg of dry CS.

Water extraction of AFEX-pretreated corn stover
AFEX-pretreated corn stover was washed with distilled water at a ratio of 1 g dry CS to 5
mL of water to produce a water extract at 20% solids loading equivalent as described on

page 16.

Nutrient analysis

The nutrient content in (i) AFEX-CS enzymatic hydrolysate at 18% solids loading, (ii)
enzyme solution (used during the enzymatic hydrolysis) and (iii) water extract of AFEX-
CS was analyzed. The nutrient of interest includes ammonium ion, protein, amino acids,

trace elements, and vitamins. The method used for each analysis is described as follows.

Ammonia
Free ammonia in the solutions was analyzed through an enzymatic assay from R-

biopharm AG (Cat no: 11112732035, Darmstadt, Germany). The solution was diluted to

84

 

an appropriate level for assay detection. The level of reduction of NADH, which
indicates the concentration of ammonia in the solution, was measured at absorbance
wavelength of 340 nm using a spectrophotometer. A standard ammonia solution
(control experiment) was tested to ensure the accuracy of the results. Other
experimental details and enzymatic chemistry explanation can be found in the

manufacturer’s instruction manuals.

Amino acids
The analyses for amino acid concentrations were conducted in MSU Macromolecular
Structure Facility through a High Performance Liquid Chromatography (HPLC) system
equipped with 3 Nova Pak C18 (3.9mm><150mm; Waters). Operational details of the
system were as described [77]. The full range of amino acid was analyzed.
Free Amino Acids
About 500 uL of respective solutions was filtered (Millipore Centricon), 20uL of the
filtered elute was derivatized with AccQ Tag (Waters), 10% of the total derivatized
sample was injected into the HPLC system.
Protein Amino Acids
The three solutions were dried under vacuum (SpeedVac, Savant) and hydrolyzed
with GM HCl at vapor phase at 100°C for 24 hrs. The hydrolyzed dry samples were
solubilized in 100 uL of 20mM HCl and IOuL of the mixture was derivatized with
AccQTag (Waters). 10% of the derivatized mixture was injected into 3 Nova Pak C18

(3.9mmXI50mm; Waters).

85

Trace elements

Trace elements were measured by an inductively-coupied-plasma mass spectrometry
(lCP-MS) in the MSU Department of Geological Sciences.

One hundred milligrams of sample were digested in 5 mL of 70% HN03 (Optima, Fisher
Scientific, Pittsburgh, PA) and placed in an ultrasonic bath for 1 hr to ensure the
materials was uniformly distributed in solution. The sample was placed on a hot plate at
70°C for 36 hrs. Then, 5 mL of H202 were added and the solutions were allowed to be
evaporated. Nitric acid (2%, SmL) was used to dissolve the residue at 70°C for 2 hr. This
solution was diluted in distilled water by a factor of 200 and run in the ICP-MS for full

mass scan analyses.

Vitamins

Five vitamins important for industrial fermentations were analyzed using a LC/MS/MS
(Quattro Micro, Waters) using a Water Symmetry C-18 column. The mobile phase was
run at 0.3 mL/min with a gradient of 1 mM perfluoroheptanoic acid and acetonitrile.
Mass spectra was acquired for 6 min using electrospray ionization in positive ion mode.
The capillary voltage, extractor voltage and RF lens voltage was set at 3.17 W, 4.00 V
and 0.3 V, respectively. The source temperature and desolvation temperature were at
110°C and 350°C. The desolvation gas flow was set at 400L/hr. Collision energies and
source cone potentials were optimized for each transition using Waters QuanOptimize

software. Data was acquired with MassLynx 4.0 and processed with QuanLynx software.

86

Solids-liquid separation

Remaining solids and liquid hydrolysate after the enzymatic hydrolysis were separated
(i) by centrifugation and (ii) an in-house manufactured press using miracloth
(Calbiochem, San Diego, CA) as the filter to retain moist unhydrolyzed solids. The
moisture content (total weight basis) of the remaining solids after press is 54il%. The

solids-free liquid hydrolysate was used for fermentation.

Trichoderma reesei fermentation

Seed culture

Trichoderma reesei RUT-C30 (provided by Dr. Jonathan Walton from MSU Plant Biology
Department) was used to investigate the feasibility of using constituents from AFEX-
pretreated corn stover as the source of carbon and nitrogen for protein production.
Seed cultures were prepared in YEPG solution (10 g/L yeast extract + 5 g/L peptone + 50
g/L glucose) by inoculating T. reesei cells grown on agar plate into the media and was

grown in 30°C,150 rpm for 48 hr.

Substrate/Nutrients for the fermentation

Fermentation was conducted in a 500 mL stirred-tank fermentor (Qplus, Sartorius,
Germany) with a working volume of 300 mL at 28°C. The fermentation media was a
mixture of 10% solids loading equivalent (SLE) AFEX-CS water extract + 1.8% SLE of
AFEX-CS enzyme hydrolysate. The dissolved oxygen (DO) level was controlled at 25%

saturation and the stirring was at 400-600rpm. The DO was controlled in cascade by

87

varying air supplementation first and followed by changing stirring speed. About 30mL

of sample was taken every 24 hr. The fermentation was stopped after 5 days.

Ethanol fermentation

Seed culture

Enzymatic hydrolysate with 6.0% SLE was diluted at a factor of 3:10 and used as the
seed culture media. No other sugar or nutrient supplements were added. Frozen
glycerol stock of Saccharomyces cerevisiae 424A(LNH-ST) was inoculated into the 6.0%
SLE enzymatic hydrolysate at initial cell density corresponding to OD600nm of 0.1. The
seed culture was grown at 30°C for 18 hr and used as the inoculum for ethanol

fermentation.

Fermentation and cell recycle

Fermentation was conducted in a 250 mL Erlenmeyer flask with 70 mL working volume.
The fermentation temperature was kept at 30°C through a water bath regulated by a
heater recirculator. A six-spot magnetic stirrer plate was placed underneath the water
bath to drive the magnetic bar which was placed in the flask. Fermentation was initiated
by inoculating the seed culture at a cell density corresponding to OD600nm of 0.5. Cell
increase equivalent to OD600nm of 20 was continuously fed into the fermentation
broth in 2 intervals (10 units per each interval, 16-24hr and 40-48hr). Fermentation was
allowed for 72hr. Cell density, sugar and ethanol concentration were measured as

described [47].

88

The feasibility of recycling cells without addition of fresh cells was tested for 4
generations of fermentation. The yeast cells generated from previous fermentation
were used to initiate fermentation at cell density corresponding to OD600nm of 0.5, and
the remaining cells were cultured in 30 mL of 18% solids loading AFEX-CS hydrolysate for
2 hr in 30°C, 150rpm (termed as cell preincubation unit). The concentrated cell-
hydrolysate mixture was then stored in an ice-water bath and continuously fed into the

hydrolysate in 2 batches (16-24hr and 40—48hr) from the cell preincubation unit.

Mass balance construction and HPLC analyses
Mass balance and HPLC analyses of glucose, xylose, xylitol, glycerol and ethanol were

conducted as described in the Chapter IV

RESULTS

Core bioprocessing and bioconversions

Enzymatic hydrolysis

Enzymatic hydrolysis using 6.7 mg protein/g biomass was able to produce sugar stream
at 110 g/L which accounted for 85% of the total carbohydrate entering the overall
process. About one-seventh of the soluble sugars was oligomers. Among them, 70% is
xylooligomers. In essence, the same sugar yield was achieved compared to the previous
chapter where Spezyme CP and Novozyme 188 were added instead of Accelerase 1000.

Protein concentration of each enzyme is as listed in Table 15 in Appendix A

89

Fermentation

The first generation of AFEX-hydrolysate fermentation essentially finished after 72 hr
with 58 g/L glucose was completely consumed within 16 hrs and xylose was utilized at a
rate of 0.37 g/L/hr; 0-60hr. All tested generations of fermentation achieved the same
sugar consumption rate at 0.36:0.01 g/L/hr without addition of fresh cells. Metabolic
ethanol yield of 0.45—0.48 g/g consumed sugar was achieved throughout these
fermentations (Figure 22). This affirmed that the high recyclability of yeast cells in AFEX-
CS hydrolysate fermentation. The final cell densities of each generation of fermentation

were 29, 32, 39 and 39 unit absorbance at CD 600nm, respectively.

Complementary unit operations

Seed culture production

Low level of fresh inoculation is expected to be necessary in an industrial production. .
In our study, enzymatic hydrolysate at 6.0% solids loading equivalent was able to
generate cell densities at 7 unit OD600nm within 18 hr. Assuming that fresh cell at 0.5
unit OD600nm is need in each generation of fermentation, an inoculum volume at 7%

that of the ethanol fermentor is sufficient to support this demand.

90

 

 

00.0000 ..00 0.6.0.. 00000.2, 00.00.0000 .00. .0>0 00.00.000.000
000.? .0. 000 0.0>00. ..00 00.00 3.0000 ..00 ..m... 00 00:00:00.0. .<. 00.000.00.00. 000020.02. .0>0...” 0.00-xm.u.< - 0.00...

0.0000 00.00

 

 

 

 

+ =00 ummm>
.2. 2::
00200804000080
. . , o \u I I I I)
_ 0.0: . 0.0:
.1 x . 00:00:00.2; Kin. 00.02.000.05
. ..Am . =00 . . :00
m r I II I \ u
3 . .
i O . .
w . u
. m l4-) 0
h . . n
r _ 00.00.008.00. _ .u 00:00:08.0“.
_ _
f I I I I \

 

 

 

          

   

0.5..
00.002.000.05
:00

 

 

 

00.00.000.00

 

 

I
.0800 E58

00 25 om
.5288. :00 so...

I
:5 @6080
00 25 0.0
0.0.0000. ..00 30..

91

00:00:05.0“.

 

 

0.002905
0000.05
30000.0.
0.00.00 000m <

 

 

20 I ' I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3 i ,
R 15 - [:1 OligAra -
Z lllllllllllllll OIingl

.g E OligGlc

E Ara

.._, 10 - ‘

g 0\\\\\\\\ Xyl

g M Glc

O

U

a. 5 - 4
ca

00

3 1

m ..

0 Illlllllllllllllllllllllll!!llllllllllllllllllll|||ll|
Control RUT—C30 Broth

Hydrolysis Media

Figure 23 Sugar yield after enzymatic hydrolysis using fermentation broth from
Trichoderma reesei RUT—C30.

Control: Hydrolysis in water; RUT-C30 broth: Hydrolysis in fermentatio broth of
Trichoderma reesei RUT-C30. Enzymatic hydrolysis was conducted for 96hr, at pH 4.8,
50°C.

Note: Olig: Oligomeric, Mo: Monomeric, Ara: Arabinose, Glc: Glucose, Xyl: Xylose.

Proof of concept for in-house enzyme production

Enzyme production from Trichoderma reseei RUT-C30 was conducted using
carbohydrate and nitrogen sources from AFEX-CS water extract (10% SLE) and enzymatic
ll‘lyd rolysate (1.8% SLE). The fermentation broth after 4 days of culture was used to
C0nduct enzymatic hydrolysis on washed AFEX-CS at 5% solids loading. The result

COnfirmed that half of the total sugars in AFEX—CS was hydrolyzed to soluble sugars after

92

72 hr. About 75% of the total xylan was solubilized. Among the soluble sugars, 36% was

oligomers of which 70% is xylooligomers (Figure 23)

Ammonia and protein

AFEX-CS hydrolysate contained 0.80 g/L of ammonium ions, resulting from residual
ammonia from pretreatment, this free ammonium ion accounted for 0.1% of the total
input ammonia for AFEX pretreatment. After enzymatic hydrolysis, total protein in
enzymatic hydrolysate was 0.88 g/L, of which more than one-thirds (0.37 g/L) was in

free amino acids (Table 11).

This protein originated from enzyme addition (for saccharification) and the inherent
nutrient content from plant biomass. Total added enzyme protein corresponded to 1.8
g/L, if all protein were to remain in liquid phase. Evidently most of the enzyme protein
was bound to biomass during enzymatic hydrolysis and removed alongside with

unhydrolyzed solids. This conclusion is consistent with the previous report [78].

Minerals
Ten important trace elements for fermentation were analyzed through an ICP-MS. The
Concentration of these ten minerals exceeded required levels for yeast fermentation
Suggested by Walker [79]. Plant biomass was the predominant source for trace
elements in the AFEX-CS hydrolysate (Table 12). Magnesium, a macroelement which is

usUally limited in most industrial fermentations [80], is present at sufficient level (99.6

93

 

 

mg/L) in AFEX-CS hydrolysate. However, the concentration of calcium exceeds the
reference value by 4 orders of magnitude; potentially undermining the metabolic

reaction catalyzed by magnesium presumably due to interference of calcium on

magnesium uptake and cellular utilization [81].

Vitamins

 

Panthothenic acid, pyridoxine and nicotinic acid are available at the respective
concentration of 3.0001, 2.1000 and 26800.2 uM, respectively, which are considered
of appropriate levels for industrial fermentations (Table 12).The respective
concentrations of thiamine and biotin were below 1 uM. Supplementation is needed if
the fermenting strains have strict growth requirements for these vitamins. Nevertheless,

these vitamin concentrations are comparable to what is commonly present in wort used

in the brewery industry [82].

Proposed scheme for integrated cellulosic ethanol production

To fully utilize the nitrogenous compounds generation during AFEX-pretreatment, the
proposed scheme features (i) core biomass processing and conversion, (i) cellulase
production, (iii) seed culture maintenance based on AFEX-pretreated biomass as the
exclusive source for carbon, nitrogen and nutrients (Figure 24). After enzymatic
hYclrolysis, solids and liquid was separated. The moist residual with 50% moisture

Corltent was washed with water (1 g wet solids to 1 mL water). The water extract, solid-

free diluted hydrolysate produced after moisture content of the washed solids is

94

 

reduced again to 50%, is used for seed culture maintenance. The remaining washed
solids were directed to enzyme production. Ethanol fermentation was conducted on
solid-free enzymatic hydrolysis and yeast cell can be separated from the broth for

recycle or as purge stream before ethanol distillation step (Figure 24).

One-fifth of the total input carbohydrate is projected to channel for enzyme production
and 78% was used for ethanol production (Figure 25). Most of the carbon source (71%)
provided to enzyme production is from unhydrolyzed solids which contain residual
cellulose and hemicellulose. The maximum overall ethanol yield using the proposed
carbon partition is 267g EtOH/kg CS or 90 gal/ton. About 60% of the accounted
nitrogenous sources (ammonium, amino acid and acetamide) are directed to enzyme
production (Table 13) and 40% for ethanol fermentation. Protein bound on
unhydrolyzed solid residues is assumed non-bioavailable. Both carbon and nitrogen
sources used for seed culture maintenance is less than 2% of the bioavailable nitrogen
source (Figure 25). If 20% of the total nitrogen source is assimilated for enzyme

production, enzymatic hydrolysis at 10.6 mg/g biomass can be conducted.

95

Table 11 Amino acid concentration of AFEX-water extract (16.7% Solids Loading) and
AFEX corn stover enzymatic hydrolysate (18.0% solids loading)

 

AFEX-Water Extract

AFEX-Hydrolysate

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Components (mg/ L) (mg/ L)
Free I Total Free I Total

NH4+ 750050 800250

Asp 8.7203 48320.2 7.720.7 54520.1
Glu 6920.4 73.4203 7220.5 76020.2
Ser 24220.0 50321.9 43.9221 74020.6
Gly 5320.0 72020.0 13220.8 88.8-20.9
His 3.1203 15.0201 6720.4 24.8202
Thr 14.3205 39120.4 40.0227 75.0203
Arg 9.0209 26620.0 34.0215 37.8203
Ala 30.721.6 78.7-20.1 32621.7 83.6-20.4
Pro 43.0253 72324.7 51.1224 75.6-23.8
Tyr 16326.0 30120.1 42821.6 23520.3
Val 7521.3 44.0202 19.1-21.4 56320.3
Met 3.2213 9.4207 5.9212 11320.1
lIe 8220.4 26220.0 13.8-20.8 41.924]
Leu 7.8200 50.6202 19.8209 66820.4
Lys 1320.0 10320.0 1220.2 15.0204
Phe 5720.1 50.9202 29.6212 72.1205

Total 195.72145 6972231 36862173 87722128

 

 

 

 

 

 

96

 

78. 05833 2: :8 m:_0> 8:280: 05 00 now: 002, .5035 20265 :8 now: to; :_ 0:002: 5305:0028 583$ 3303 9: 0
.33 «com 09:03 >0 umummmmam 00 0:25 3:95.00 80.; :8 m:_m> 8:288 05. 0

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

00.0-30 000,. 00.0.. 0582,:
000-000 000,. 00.0,. 0:20 m
00.00-0000 $0000.00 00.30000 .21 05. 00582 m
03-000 00003.0 00.0003 0562;: m.
3.0-00.0 00.00000 20000.0 22 200505000
0000 00.00000 8000: 09.. 0“. w
000.0 0000.000 00.00000 50003222 as.
000 0000000 00000: 050 00 1
0000.0 00.00000 00.00000 .0008 8 m
00.0 0000000 0000.000 <08 .322 _z m
0000.0 00.00000 00.00000 :38 8 m
00 00.00000 0000000 30000005. 0.2 m.
~00 00,300.00 000000.20 520.8 8 ..s.
00 00.000000 00.300000 528000.). 0.2
00 0000000: 000000000 82330: V.
a... «20> 38223: 625 2.3.03.5"? 0.5
3:283. in“?

 

 

 

 

 

 

.0208. 22.8 00.0:
300293: 0:00:55 .0530 So... VG“? 0:0 3:602 02.00 000.0: Umbxm 5003-503 E 22:33 new 35:33 89:. NH ~33

 

 

.cozficmctfi ncm $3.93; 33835 toaasm B 203.23% 83.3

8% tea cozuscoa mE>Ncm .m._ mmmmmuoa 335E358 ucm Axon E .8592»: HE: 95.333 28 m 8955 :08an m7:

.QUBSOm “£253: Dem :wth—E

509.8 933.88 mm .3on 58 35833-5”? was: 8.853 5332.805 new 95893 32:03 BEESE 2t 3 95m:

238$ __mo

_. l
_
_

cozmfiaow
10m. £054.00

 

 

E5 1.»
cozmnaocfikcl

=mo

 

98

Em;

 

T

 

 

829565“.

 

 

 

 

 

 

   

22.983 m_m> o. > :0. m.
n_:c_._-mu__omv cam—”hmcI ..Wom “Xm—
wwoi . m. H >>

 

 

Eoo

 

 

 

9330 J;

9mm

$.06»:

8:
-28

 

 

 

328m

 

 

2.8 895
fl 8:25 #252 88395. gas

 

8mm E HECEmmE; H 9512.”. H

 

 

.magmom
22.8

 

802% £000 .2 30: cmmobE new 0088 Ev
$203230 :03 00300.0 05 .8 :5.me 30: c3 .32.: EmucmequEou 000 0.00 .8 30: :0m0Ec 000 000.00 08000:. mm 05m:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

$3 $3 momma 03% 3 $3. @323
. 835 830 830 595 .38
08.0 03m “.03 m 08.0 E 3% m0.329....
08.0 20.0 H.030 83 03 083 coma E 23. osefisfiea
Sod Sod Sod 220 08.0 830 08.0 3 5285.5
Bod 30.0 93 SS. 983 ~83 08.0 E maes8<
mu m .0 m~ m .0 ON 39; .0>0 Lw>o ..m>o
r8293: >0“. m>0c MW... M0: -308 50 um 50$ 505
.mmam .3.“ n 3 .3“; D Umbxm “WNW? u 3mm“: 3:300:08
.8025 . .2 . s. .325 u g 2, van: .0 a
n Emmbm o Emmbm m 802% v Emmbm m 6085 N 80.2% H Emmbm m
wco_m..m>cooo_m 8895
A'Il Tim hm~m>> N me< 523% Eo
:8 0cm 053000505 0.00 32 o
m >
w
_
.203
0000:2502 00:25 cozosuoi H +
05:30 000m :83me mEficw H V
5.03% magmom W <

mu=ow

99

s‘ :—~.~ fiflw .s xsymks.‘

 

 

 

 

 

 

 

 

. 00.0 00:0 03200
0.2000. 00 0 00.0 .00: 0 0329.... 50:0 =38
mmd 0_0< 00_E<\0_30.0
€000.02. mwd Hmd 82000.8(
0~.~ 00.0.3094
300033 .00 0803
«00.02000 00.00300... 0000000890
03.030035: 00.500
3000.03 .000... 000300 _000._.

 

 

 

 

00300020 00.>~00 m0_.300 0000.00 000 000.00 03003.0 .33 ma 0.0:

3. .3005. am 0.:mE 00 0000.. 0300.500 0_ 00:00 0320.02.00 0_ 30.03093 0

0000:8000 0200.20. .00 030=0>003 000 000.0000 0. 000 00.0000 05 0_ 03000000 000 0_ 0000.03 0320.030: 00 003 530.0 0
«000300.030 xm.“_< w0_.:0 00>.000.0 002, 530.0 ”00_00E:00< 0

03003000030; 0_ 020.3000 _33 05 00 $00 ”00.000502 0

.00. 05.. 0328.8 m

30.000088 .0>000 0.00 .3 030.:
.30000 00.0 0.0.0 00300. 0.02000 05 3 m0_0.0000 0300.500 0. 00.00 00.80 000 030.0 .00_ 05.000.80 00 00003000000 05 “302

0.000 mN 0.330

100

 

 

DISCUSSION

A new paradigm for cellulosic ethanol production

Consolidated bioprocessing (CBP) has been widely regarded and developed as the most
economical approach to produce cellulosic biofuels. Fundamentally, this advanced
technology relies on the microbial platform to conduct complex multi-parametric
reactions which include (i) enzyme (cellulase and hemicellulase) production, (ii)
enzymatic hydrolysis of biomass (cellulose and hemicellulose) and (iii) bioconversion of
both hexoses and pentoses in one reactor. I herein propose a new paradigm where
enzyme production, enzymatic hydrolysis and fermentation are conducted in separate
reactors and controlled at respective optimal conditions. Nevertheless, ethanologenic
CBP strain can play an pivotal role in this processing approach by increasing final ethanol
yield and titer through digesting oligomer formed in the enzymatic hydrolysate, a major
bottleneck in the overall process [47]. This paradigm also features integrated biomass
processing and bioconversion strategies solely based on AFEX-pretreated corn stover as
the exclusive source for carbon, nitrogen and nutrients and established ethanologenic

strain Saccharomyces cerevisiae for fermentation.

The role of ammonia in pretreatment, enzyme production and ethanol fermentation
Ammonia, introduced during AFEX pretreatment, contributes to three major goals (i)
overcoming recalcitrance of biomass (releasing plant carbohydrate), (ii) producing
inhibitory degradation compounds at low levels, and (iii) preserving and enriching the

nitrogenous content of the pretreated biomass. About 1.5 g ammonia was chemically-

101

bonded in each 100 g of dry corn stover. In the proposed scheme, the ammonium ion
and acetamide (a reaction product between acetyl group and ammonia) are the main
nitrogen sources for enzyme production. This effectively creates a cost-sharing scenario

between pretreatment, enzyme production, and nutrient supplementation.

Xylose fermentation and yeast cells

Xylose consumption has been identified as the primary rate-limiting factor for overall
production in Chapter IV [47]. Through increasing cell density during fermentation,
xylose fermentation was essentially completed within 72hr (Figure 22). More
importantly, fermentation using recycled yeast cells was as effective as using fresh cells
at least up to 4 generations, therefore substantially reducing the need for seed
inoculum after the first fermentation was initiated. Due to the net increase of cell
density in each generation of fermentation, a purge stream of yeast cell as a co-product
is expected. Unlike in SSF where cells and solids are in a mixture, yeast cells in the solid-

free fermentation broth can be readily separated by sedimentation.

Nutrient is sufficient for ethanol fermentation

Nutrient content, in the form of ammonium ion, protein, trace elements and vitamins,
inherently from AFEX-pretreated corn stover was sufficient to support microbial growth
for industrial fermentation. Analyzed nutrient content is considered adequate compared
to reference value suggested for yeast fermentation [81, 82]. Ethanol fermentation can

exclusively rely on AFEX-pretreated corn stover as the source for carbon, nitrogen and

102

other nutrient requirements. Fermentation using corn steep liquor as the commercial
nutrient supplementation was projected to cost 5-8 cent/gal, a significant cost
particularly in biofuels production where processing cost is expected to be at 1.0

USD/gal to be commercially-attractive.

Potential self-dependence on enzyme production

Saccharolytic enzymes include a wide range of cellulolytic and hemicellulolytic enzymes
are identified as the major part of the processing cost of cellulosic ethanol production. A
recent study estimated cellulase cost at 32.2 cents/gal EtOH, a dominant operating cost
in the ethanol production [84]. Nevertheless, the actual cost of these enzymes is still
uncertain, and would largely depend on the traditional supply and demand dynamics.
Hence, the ability to produce enzymes in-house using water extract of AFEX-treated
corn stover and residual unhydrolyzed biomass solids is significant in the search for a
bioprocessing strategy where biorefinery can be independent from external enzyme
sources. Assuming that 20% of the bioavailable nitrogen-equivalent compounds was
converted to saccharolytic enzymes optimal for biomass deconstruction, enzymatic
hydrolysis at 10.6 mg enzyme/g biomass can be conducted. This enzyme loading is 3.9

mg/g higher than what is utilized in enzymatic hydrolysis in this chapter.

103

CONCLUSIONS

Cellulosic ethanol production based on AFEX-pretreated corn stover can depend on the
biomass as the sole source for carbon, nitrogen and nutrients source for both enzyme
production and ethanol production. In this proposed scheme, about 80% of the carbon
source is used for ethanol production and 20% is channeled for complementary unit
operations i.e. enzyme production and seed inoculum maintenance. Assuming that 20%
of the extractable nitrogenous source could be converted to saccharolytic enzymes,
enzymatic hydrolysis at 10.6 mg/g dry corn stover enzyme loading can be conducted.
Close to 80% of the total xylose can be fermented within 72 hr through high cell density
fermentation. The fermentation using yeast cells were as effective as that of fresh cells
at least up to 4 generations, effectively reducing the need for fresh cells to conduce
fermentation. The nutrient content analysis affirmed that enzymatic hydrolysate from
18% solids loading of AFEX-pretreated corn stover contains sufficient nitrogen, trace

elements and vitamins for yeast fermentation.

104

CHAPTER VI GENERAL CONCLUSIONS AND RESEARCH OUTLOOK

GENERAL CONCLUSIONS

Cellulosic ethanol provides an environmentally-friendly alternative to petroleum-based
fuels. By utilizing locally-produced feedstocks as carbon sources for transportation fuel,
ethanol can create self-dependence on energy supply and reduce geopolitical
implications. However, major technological challenges to produce ethanol in a cost-
effective manner must be overcome to realize the commercial potential of cellulosic

ethanol.

The levels of overall ethanol yield, concentration and rate indicate the maturity of the
technology. The cost associated with hydrolysate conditioning, nutrients, and enzymes
have also been the major barriers for its commercialization. Our study directly addresses
these issues by investigating and developing AFEX-pretreated corn stover as the model
substrate to provide understanding and practical solutions that result in cost reductions

in those areas.

105

 

To achieve this goal, I focused to identify (i) a pretreatment that generates fermentable
pretreated biomass (ii) an ethanologenic strain that produces ethanol at the highest
yield, titer and rate and (iii) the major bottlenecks for the overall process. Overcoming
biomass recalcitrance without generating inhibitors can eliminate the cost associated
with hydrolysate conditioning in terms of detoxification and nutrient supplementation.
The nature of pretreatment chemistry dictates the inhibitory effect of the pretreated
materials. It was found that degradation compounds generated from AFEX pretreatment
is substantially more benign to the growth of ethanologenic E. coli and S. cerevisiae in
comparison to dilute acid pretreatment and this effectively eliminates the need for
detoxification, which is projected to cost 20% of the total processing cost. Lignin, an
important biomass constituent for co-product generation, is highly preserved in AFEX-
pretreated corn stover. It is expected that 13% more energy can be harvested in an

AFEX-centered process relative to that of dilute acid pretreatment.

The ability of an ethanologenic strain to utilize all sugars derived from lignocellulosic
biomass and convert them to ethanol is the key process that affects overall process
economics. In this report, E. coli K011, Z. mobilis AX101 and S. cerevisiae 424A(LNH-ST)
were tested on AFEX-CS hydrolysate from high solids loading hydrolysate without
detoxification and commercial nutrient supplementation. These three ethanologens
were able to grow and consume glucose completely within 72hr. However, the extent
and rate of xylose consumption become the yield- and rate-determining factor in

fermentation. Fermentation using S. cerevisiae 424A(LNH-ST) yielded the highest level of

106

afiano

wasob:

fiatZO?
asbmer
ST) was

mmmer

To fUl’thl

Ethanol:

L1

(7')

i' est 6C
theOver;
hasbeen
72hr The

‘ r
‘9 mEHta

I}.
~gnOCEHL
WkogEHO

539mm y“

 

ethanol in all tested strains. Nearly 20-fold increase in cell density after fermentation
was observed and this confirmed the feasibility of producing yeast cell as the co-product
from lignocellulosic fermentation.

From a system-wide perspective, 191 g ethanol can be produced from 1.0 kg (64
gal/ton) of corn stover at a titer of 40 g/L. Enzymatic hydrolysis at high solids loading
have been identified as the primary bottleneck that affects the overall process. I found
that 20% of the output sugar from hydrolysis unit was oligomers and 15% remained in
polymeric and water-insoluble form. Xylose fermentation using S. cerevisiae 424A(LNH-
ST) was prolonged. Overall fermentation rate must increase by 2-3 fold to suit

commercial production.

To further improve the overall yield, a consolidated bioprocessing (CBP)-enabling
ethanologen Thermoanaerobacterium saccharolyticum ALK2 has shown to be able to
digest 60-70% of the total oligomeric sugars in the enzymatic hydrolysate. This brought
the overall product yield close to 77 gal/ton (Appendix A). High cell density fermentation
has been demonstrated as a viable approach to achieve complete fermentation within
72hr. The ability to recycle the cells at least up to 4 generations while achieving similar

fermentation efficacies without fresh cell addition was also confirmed.

Lignocellulosic hydrolysate from AFEX-pretreated corn stover contains 4-5 g/ L of
nitrogenous source in the form of ammonium, amino acids and acetamide. While no

attempt was made in our study to understand the bioavailability of trace elements, it

107

wase
fenne
atsuf
fenne
lprop
Straig
unns)
sourc
Idaxk
gal/IC
stove:
Drodc
sohibi
asshn

CODdL

was expected that essential minerals are in excess relative to the requirement of most
fermentation microorganisms. Regarding vitamins, although most vitamins are present
at sufficient levels, biotin and thiamine might need to be supplemented if the utilized
fermentation strain has the stringent growth requirements of these two vitamins.

I proposed an integrated biomass processing and conversion approach that features (i)
straightforward pretreatment, enzymatic hydrolysis and ethanol fermentation (core
units), (ii) in-house cellulase production and (iii) seed inoculum unit. Carbon (sugar)
source was divided between core and complimentary units at weight ratio of 4:1.
Maximum ethanol yield under this carbon partition scheme is 267 g EtOH/kg CS or 90
gal/ton. About 60% of the extractable nitrogen and nutrients from AFEX-pretreated corn
stover was projected to be used for cellulase production. As a proof of concept, enzyme
produced from Trichoderma reseei RUT-C30 from the proposed scheme was shown to
solubilize 47% of the total polymeric plant sugars. If 20% of the available nitrogen is
assimilated for enzyme production, enzymatic hydrolysis at 10.6 mg/g biomass can be

conducted.

This new paradigm of cellulosic ethanol production utilized AFEX-pretreated corn stover
as the exclusive source for carbon, nitrogen and nutrients to ethanol and enzyme
fermentation within the integrated biorefinery without detoxification and commercial
nutrient supplementation. The potential cost reduction is substantial as 45% of the total

operating cost is comprised of raw materials associated these unit processes.

108

SO

by
P“

0P

Ce;
Sac
use
en:
sec
fen
Dro
Evic

Sho

The

Dret

RESEARCH OUTLOOK

It was concluded from our study that 14% of the total sugar remained in unhydrolyzed
solids and 15% as soluble oligomers in Chapter IV. To improve overall yield based on the
proposed biomass processing scheme, (i) effective sugar solubilization during enzymatic
hydrolysis and (ii) utilization of the whole spectrum of soluble sugar for ethanol
production are required. Specific to achieving these two primary goals, I suggest that
optimization of cellulase production and incorporation of a CBP—enabling (cellulase-

secreting) ethanologen to address both challenges.

Cellulase production optimization

Saccharolytic microorganisms such as Trichoderma sp. and Aspergi/Ius sp. are widely
used for industrial cellulase production due to their ability to secrete a wide range of
enzymes for biomass deconstruction. Nevertheless, the levels and types of enzyme
secreted are highly dependent on strain type, induction method, carbon source and
fermentation condition [85]. To maximize sugar solubilization, high secretion level to
produce an optimal enzyme mixture from the saccharolytic microorganism is needed.
Evidently, detailed optimization of cellulase production with regards to those factors

should be pursued.

The ability to utilize acetamide, the most abundant nitrogenous compound in AFEX-

pretreated biomass, as the nitrogen source expanded the bioavailable nitrogen source

109

als
Th
dLJ

bic

Inc
Cor
prc
enz
par
The
inte
Stre
abil
fea:
Opt.
dev

by 73% in AFEX-centered process. Native saccharolytic strains such as Aspergi/Ius
nidulans secretes acetamidase [86]. Expression of amdS gene in a heterologous host was
also reported to enable growth using acetamide as the sole nitrogen source [87].
Through assimilation of acetamide as nitrogen source and effective enzyme production
during fermentation, cellulase can be produced in excess as a potential coproduct in the

biorefinery.

Incorporating CBP microorganisms on AFEX-pretreated biomass processing platform
Consolidated bioprocessing is generally regarded as the most economical biomass
processing approach due to its projected ability to eliminate the need for exogenous
enzyme and sugar inhibition on cellulolytic enzymes. However, the fermentation
parameters from reported CBP has not met the levels required for industrial production.
Therefore, to achieve an effective CBP-center cellulosic ethanol production, several key
interactions must be further investigated, they include (i) the growth robustness of the
strain with respect to ethanol and degradation compounds from pretreatment, (ii) the
ability and extent of oligomer uptake, (iii) the enzyme profile of the strain and (iv) the
feasibility of exploiting directed evolution to engineer enzyme secretory profile toward
optimal hydrolysis of AFEX-pretreated biomass. Most CBP-enabling ethanologens
developed are anaerobic bacteria and natively saccharolytic. It is expected that the

enzyme profile could be directed through the available carbon sources.

110

Fundar
Mofip
fenner
longte
achiev
ceHssr
ferniel
theob
IdentH
negafi
ofthe

fernne

Conc (g/L)

FigUre

Fundamental understanding on cell-solids interactions

Most proposed biomass processing configurations involve direct contact between
fermenting cells and plant biomass. The cell-solids interaction was found to reduce the
long term viability of yeast cells; therefore lower ethanol yield and titer (Figure 26) were
achieved in this processing approach. Interactions between insoluble solid residues and
cells such as cell adsorption and shear stress can potentially reduce cell viability and the
fermentation efficiency in slurries (solid-liquid mixture). They are the likely causes for
the observed substantial reduction in xylose utilization in slurry—state fermentation.
Identification, understanding and alleviation of the solid-cell interactions which exert
negative impacts on overall fermentation deserve further extensive investigation as it is
of the greatest significance during the process of integrating enzymatic hydrolysis and

fermentation in the configuration SSCF or CBP.

 

 

 

 

 

 

 

 

 

 

 

 

 

70’6‘.,.r.I 70Cr11. :
60 - - 60- -
50 - 4 50— _
’3 . .
\
£0 40 — - 40- -
g 1 d ‘
o 30- — _
U «I a
20- - -
10- — ..
O ' IL' I ' f' 0 I ' l ' I V l
0 20 40 60 80 0 40 80 120 0 40 80 120

Time (hr) Time (hr) Time (hr)

Figure 26 Fermentation using S. cerevisiae 424A(LNH-ST) in (A) laboratory media, (B)
18% w/w AFEX-CS liquid hydrolysate (solid-free) and (C) 18% w/w AFEX-CS
hydrolysate (slurry, SSCF)

111

AP
(C!

III

"UT

The
sea
aral

INST

COU
Con
DIOI
Df0(
(figs
robL
Enzy
conC

hYdr

APPENDIX A EXPLORATION OF CONSOLIDATED BIOPROCESSING
(CBP) ON AFEX-PRETREATED CORN STOVER USING

THERMOANAEROBACTERIUM SACCHAROL YTIC UM ALK2

INTRODUCTION

Thermoanaerobacterium saccharolyticum is a native saccharolytic microorganism which
secretes various hemicellulolytic enzymes including endoxylanase, B-xylosidase,
arabinofuranosidase and acetyl esterase for xylan digestion [49, 88, 89]. The
metabolically-engineered strain ALK2 ferments ethanol close to thereotical maximum
[49]. A biomass processing strategy centered on AFEX and T. saccharolyticum ALK2
could potentially simplify cellulosic ethanol production in the configuration of
Consolidated Bioprocessing (CBP) which is projected to be the most economical
production strategy [46]. This research serves as the early efforts to develop CBP
processing based on AFEX-pretreated materials. I evaluate (i) the level of oligomer
digestion on enzymatic hydrolysate from AFEX-pretreated materials, (ii) the growth
robustness of ALK2 in water extract of AFEX-pretreated corn stover and (iii) levels of
enzyme loading required to generate enzymatic hydrolysate containing high
concentration of oligomeric sugars. Fermentation parameters of ALK2 on lignocellulosic

hydrolysate will also be investigated.

112

hflAl

Corr
Corr
Coka
appr
tern

2.3%;

Corr
Ivhk
pref
Iflas
ng
Dies
reac
l30:
thro

0V6!

MATERIALS AND METHODS

Corn stover and ammonia fiber expansion (AFEX)

Corn stover was provided by the National Renewable Energy Laboratory (NREL, Golden,
Colorado). It was milled and passed through a 4 mm screen. The moisture content was
approximately 7% (total weight basis). The milled corn stover was kept at 4°C for long
term storage. This corn stover contains 33.2% cellulose, 22.4% xylan, 3.3% arabinan and

2.3% protein on a dry weight basis.

Corn stover was AFEX-pretreated in a 2.0L pressure vessel (Parr Instruments, Moline, IL.)
which was equipped with a thermocouple and a pressure sensor. The reactor was
preheated to 100-110°c and prewetted corn stover (150 g dry CS + 90 g distilled water)
was loaded into the reactor. The reactor was bolted shut and vacuum. A separate 500
mL stainless steel cylinder (Parker Instrumentation, Jacksonville, AL) was heated until its
pressure reached 4.48 MPa (650 psi) and the heated ammonia was transferred to the
reactor containing biomass. The initial and final temperatures of the pretreatment were
130iS°C and 11015°C, respectively. The reactor pressure was released after 15 min
through an exhaust valve. The AFEX-pretreated CS was air-dried in a fume hood

overnight.

113

)1:-

ce

fer

En;

C0!”

int

Wer

ofd.

Preparation of enzymatic hydrolysate (at 18.1% Solids Loading)

AFEX-CS was enzymatically-hydrolyzed using cellulase and hemicellulase mixtures at pH
4.8, 50°C for 96hr. The solids loading of the enzymatic hydrolysis was at 18.1% w/w
(corresponding to 6.0% cellulose loading). The cellulase mixture consisted of Spezyme
CF [86.7 mL/kg CS; 15 FPU/g cellulose] and NovozymeTM 188 [87.5 mL/kg CS; 64
pNPGU/g cellulose]. The hemicellulase mixture was Multifect Xylanase [12.7 mL/kg CS]
and Multifect Pectinase [8.9 mL/kg CS]. The spectrum of activities for the enzymes was
previously reported [44]. The Spezyme and Multifect enzymes were obtained from
Genencor Inc. (Palo Alto, CA) and Novozyme 188 was purchased from Sigma-Aldrich Co.
(St. Louis, MO). After 96 hr of hydrolysis, liquid hydrolysate was separated from solids by
centrifugation at 6,000xg for 30 min. Liquid hydrolysate was used for the following

fermentation studies. Kanamycin at 30 mg/ L was used to reduce risk of contamination.

Enzymatic hydrolysis using Accelerase 1000 (Genencor Inc., Palo Alto, CA) as sole
complex enzyme at respective volume loading of 106 and 70 uL/g dry CS was conducted
in the similar condition. The corresponding protein loadings are 5.6 and 3.8 mg/ g dry
CS, respectively. Oligomeric glucose, xylose and arabinose of all enzymatic hydrolysates
were measured using NREL protocol LAP 014. Error bars shown are standard deviations

of duplicates.

114

 

Det

Pro

188

ana

cen

adc

pre

W8

pre

fol

pre

Ul'l

Ni‘

NI

CC

St

Determination of protein concentration in complex enzymes

Protein concentration of commercial enzymes Accelerase 1000, Spezyme CP, Novozyme
188, Multifect Xylanase, Multifect Pectinase were determined through nitrogen content
analyses of protein precipitate of respective enzymes. Each complex enzyme was
centrifuged (13,000 x g) for 5 min, 0.20 mL of clear supernatant of the enzyme was
added with 0.25 mL 100% w/v trichloroacetic acid (TCA) and 0.80 mL distilled water to
precipitate protein in the enzyme solution. After 5 min of incubation at 4°C, the mixture
was centrifuged at 13,000 x g for 5 min and the supernatant was decanted. The
precipitate was washed with 1.0 mL cold (4°C) acetone twice, each washing was
followed by centrifugation and the removal of acetone water extract. Washed protein
precipitate was removed to a crucible (a sample holder for nitrogen analyzer) and dried

under vacuum.

Nitrogen content within the precipitate was determined using a Skalar Primacs SN Total
Nitrogen Analyzer (Breda, The Netherlands). The principle behind the nitrogen analysis
is based on Dumas method using EDTA as the standards. Nitrogen content was
converted to protein content by multiplying a factor of 6.25. Errors represented are

standard deviation of duplicate experiments.

Thermoanaerobacterium saccharolyticum ALK2 seed culture preparation
MTC media (Table 14) supplemented with 10 g/L of glucose and 10 g/L xylose was

prepared and 50 mL of this media was sterile-filtered before transferring into a sterile

115

200 mL serum bottle. The bottle was intermittently purged with nitrogen (filtered
through 0.22pm pore) and vacuumed for 5 cycles to create anaerobic condition. Each
cycle lasted approximately for 40 sec. The frozen glycerol stock (-80°C) of ALK2 was
anaerobically inoculated into the media. The bottle was incubated in a 55°C shaker
(Innova 4080, New Brunswick Scientific) at 180 rpm agitation for about 18 hr. The grown
cells were used as the inoculum for fermentation studies. To achieve inoculation
without adding nutrients from seed culture process, grown cells were centrifuged and
supernatant (fermentation broth) was removed carefully under anaerobic condition.

Sterile distilled water (at volume equals to removed broth) was added to resuspend the

pellet cells. Appropriate volume of inoculum was taken for inoculation.

Table 14 MTC Media for T. saccharolyticum ALK2 Growth

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Final Conc.
Yeast extract 10 g/ L
Solution A Tryptone 5 g/ L
MES (buffer) 10 g/L
Citric acid potassium salt 2.00 g/L
Citric acid monohydrate 1.25 g/L
Solution B Sodium sulfate (Na2504) 1.00 g/L
Potassium dihydrogen phosphate (KH2P04) 1.00 g/L
Sodium bicarbonate (NaHCOg) 2.50 g/L
. Urea 5.00 g/ L
S°'"t'°" C Ammonium chloride (NH4CI) 1.50 g/L
Magnesium chloride hexahydrate
(MgCl2.6H20) 1'00 g/L
Solution D Calcium chloride dehydrate (CaCl2.2H20) 0.20 g/L
Iron (II) chloride tetrahydrate (FeCl2.4H20) 0.10 g/L
L-cysteine hydrochloride monohydrate 1.00 g/L
Pyridoxamine dihydrochloride 0.020 g/L
, P-Aminobenzoic acid 0.004 g/L
S°'"t'°" E D-Biotin 0.002 g/L
Vitamin B12 0.002 g/L

 

116

 

Batcl
Liqui
pH-a
final
Tabl
filte

and

byt
carr

DEF

Fec
Fec
NJ)
ext
Sys
of
We
ml
fe:

SU

Batch fermentation of enzymatic hydrolysate at low solids loading equivalent

Liquid hydrolysate from 18% solids loading saccharification of AFEX-CS was diluted and
pH-adjusted to 6.2 using KOH. MTC media was supplemented to the hydrolysate at the
final concentration listed in

Table 14. The final dilution factor was at 1:5. The diluted hydrolysate (50 mL) was sterile
filtered and transferred into a 250 mL sterile serum bottle. The bottle was vacuumed
and purged with nitrogen as described in the previous section. Inoculation was achieved
by transferring 5.0 mL of seed culture into the diluted hydrolysate. Fermentation was
carried out in a 55°C shaker at 180 rpm agitation. Samples were taken at designated
period. Glucose, xylose and arabinose in monomeric and oligomeric forms were

analyzed through HPLC and NREL Protocol LAP-014.

Fed-batch fermentation of enzymatic hydrolysate at high solids loading equivalent
Fedbatch fermentation was conducted in a custom-made fermentor (NDS Technologies,
NJ) equipped with a pH probe. Temperature of the fermentor was controlled by an
external water bath recirculator. Feeding and pH were controlled by Sartorius Aplus
system (Goettingen, Germany). Initial volume of the reactor is 120 mL which consisted
of 20 mL enzymatic hydrolysate at 18% solids loading, nutrient supplement and distilled
water (for dilution). For rich media fermentation, 1.0 g yeast extract, 0.5 g peptone, 10
mL of concentrated stock for solution B, C, D and E was added; while for lean media
fermentation, 0.2 g yeast extract, 0.1 g tryptone and stock for solution D was used as

supplement. The fermentation media was pH-adjusted to 6.2 with KOH and sparged

117

with nitrogen for about 10 min to create anaerobic condition. Seed culture (10mL) was
inoculated to initiate fermentation. Undiluted 18% solids loading enzymatic hydrolysate
at pH 6.2 (rich media: supplemented with 10 g/L yeast extract and 5 g/L peptone, lean
media: no supplementation), was used as the feed. Feeding started 4 hr after
inoculation at the rate of 4.0 mL/hr until 180 mL of feed volume was added into the
fermentor. Samples were taken at the designated period. Glucose, xylose, arabinose (in
monomeric form) and ethanol were analyzed using HPLC. Oligomeric sugars were

analyzed through acid hydrolysis based on NREL Protocol LAP-014.

Fermentation using water extract of AFEX-pretreated corn stover (inhibitory testing)
AFEX-CS water extract at 20% solids loading equivalent was produced as previously
described [47]. In essence, AFEX-pretreated corn stover was washed with distilled
water at a ratio of 1 g dry C5 to 5 mL of water and the water extract was removed from
the biomass through an in-house manufactured press. The water extract contained 3.7
g/L anhydrous-equivalent glucose (1.7 g/L monomer + 2.0 g/L oligomer) and 8.3 g/L
anhydrous-equivalent xylose (1.7 g/L monomer + 6.9 g/L oligomer). The water extract
was diluted to final solids-loading-equivalents (SLE) of 3.0, 6.0, 9.0, 12.0 and 15.0%. Each
fermentation media contained 7.8 g/ L anhydrous-equivalent glucose and 14.6 g/L
anhydrous-equivalent xylose for carbon source; 5.0 g/L yeast extract and 2.5 g/L
tryptone for nutrient supplementation; and 100 mM phosphate buffer at pH 6.0. Seed
culture at 1.5 mL was inoculated to initiate fermentation. Total volume of the each

fermentation is 15.0 mL. Sample was taken after designated periods. Sugar and ethanol

118

was analyzed in a HPLC system. Errors reported were the standard deviations of

duplicate experiments.

HPLC analysis

Sugars and ethanol in this work was analyzed by a HPLC system (Waters 2695) equipped
with an autosampler unit. The column was Biorad Aminex HPX-87H column, and is
maintained at 60°C. The mobile phase used was SmM H2504. Other details about the

analysis were as reported previously [47].

RESULTS AND DISCUSSION

Batch fermentation

Fermentation using Themoanerobacterium saccharolyticum ALK2 in undetoxified AFEX—
CS at 3.6 solids loading equivalent achieved complete fermentation of monomeric
glucose, xylose and arabinose within 24 hrs (Figure 27A). In addition, over 50% of the
oligomeric sugars were consumed (Figure 27B). In essence, all consumed sugar was
converted to ethanol (0.50 g EtOH/g consumed sugars). Due to unconverted oligomeric
sugars, the process yield is 0.44 g EtOH/g initial sugars. Xylooligomers consisted of the
largest unconsumed sugars (64%) (Figure 27). Arabinose-containing oligomers were not
further enzymatically-hydrolyzed indicating that T. saccharolyticum lacks of related

enzymes.

119

Hemicellulose, highly complex and heterogenous in nature, comprises of linear
homoxylan, arabinoxylan, glucuronoxylan and glucuronoarabinoxylan. Hence, the
lacking of complete enzyme system for hydrolysis of the xylo- and arabino-oligomers
were not surprising. Examination of the structure of the residual oligomers could help
identify the enzyme activity which is lacking in the T. saccharolyticum enzyme secretory

system.

ALK2, a genetically-engineered homoethanologenic strain, produced ethanol at
theoretical maximum and practically eliminated organic acid production during
fermentation. These results are consistent with the previous fermentation report [90]

based on synthetic media.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A 12 . ' I ' I ' I ' I ' I ' I ' I ' B
- LT
\
E 32
3 ‘ .5
.5 _ E
E 1.5.
g u
8 . i E.
3
0 ' I ' I ' I 1 I '-l ' I ' T-' m
0 10 20 30 40 50 60 70 80
Time (hr) Time (hr)

Figure 27 Batch fermentation of AFEX-CS hydrolysate at 3.6% solids loading equivalent
using T. saccharolyticum ALK2.(A) fermentation profile and (B) initial and final
sugar composition.

lnoculum size used to initiate fermentation was 10% (by volume) of fermentation
volume. Fermentation was carried out in 55°C, 180rpm under strictly anaerobic
condition. Olig: Oligomeric, Mo: Monomeric, A: Arabinose, G: Glucose, X: Xylose.

120

Fedbatch fermentation

In rich nutrient-supplemented fermentation, close to 90% of the total sugars in the
hydrolysate was consumed, metabolic yield of 0.45 g EtOH/ g consumed sugars was
achieved (Figure 28A). Fermentation was completed within 64 hr after inoculation; 15 hr
after feeding was concluded. Total oligomeric sugar utilization has achieved 60%. This
result (the extent of oligomer consumption) is in agreement with the results from batch
fermentation at low solids loading equivalent liquid hydrolysate. I demonstrated that
ALK2 is able to grow and produce ethanol to 30 g/L at 0.45 g/L/hr (0-64hr) from the
hydrolysate containing degradation compounds equivalent to 11.7% solids loading of

AFEX-CS (Figure 28B).

In lean-nutrient-supplemented fermentation, the levels of both total sugar consumption
and metabolic yield were 80% of those in rich-nutrient-supplemented fermentations.
Ethanol production apparently ceased 40 hr after inoculation. This is likely due to (i) the
lack of important nutrients for the fermentation and/or (ii) T. saccharolyticum cells were

not fully adapted to the lignocellulosic hydrolysate environment (Figure 28C).

121

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

[In
A B ' ‘ C T'V‘ l l I I
A 35- - 35- -
< . .
£0 E 30- - E 30- -
0 . . .
2 3 25- . £9 25- —
+5 0 . . 0 .
b .g 20- — .9 20- -
J—I
S e - ~ 2 -
g E 15- +MOG - E 15- -
a) d . a) u
8 g 10- ' MOX _ g 10- _
0 o - +Et0H . o . .
no 5- — U s- _
:3 . . .
U)
— 0 . , . 7 .j-fi T: 0 . , . . . . . , .
0 88(Rich)88(Lean) o 20 40 60 80 100 o 20 4o 60 80 100
Time (hr) Time (hr) Time (hr)

Figure 28 Fedbatch fermentation of AFEX-pretreated corn stover using T.
saccharolyticum ALK2. (A) initial (if no fermentation occurs) and final sugar
concentration with rich and lean nutrient supplementation, (B) fermentation
using rich nutrient supplementation and (C) fermentation using lean nutrient
supplementation.

Note: Olig: Oligomeric, Mo: Monomeric, A: Arabinose, G: Glucose, X: Xylose.

The inability to utilize oligomers for fermentation was identified as one of the main
primary bottlenecks for cellulosic ethanol production. Hence, the results presented
herein suggested that integration of CBP microorganisms like ALK2 into AFEX-centered
biomass processing technology can increase overall ethanol yield from 64.1 gal/ton;
191.4 g EtOH/kg (as previously reported [47]) to 76.8 gal/ton; 229.5 g EtOH/kg; 270 g
EtOH/kg. However, ALK2 fermentation was supplemented with rich nutrients in this
work; further adaptation of the strain to lignocellulosic hydrolysate environment might
able to mitigate this requirement. Final ethanol was also lower than the previous report

[47] due to fedbatch strategy to keep sugar levels low throughout the fermentation.

122

Table 15 Concentration of nitrogenous compounds in commercial enzymes

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

With TCA Protein Without TCA Protein Ammonium
Precipitation Precipitation (Whole Content/Total

(Protein Precipitate) Enzyme Solution) Nitrogen

Total Protein Total Protein git-2:1:

Nitrogen Equalent N'troge" Equ'va'ent Solution)

mg/mL

Acclegggase 8.5001 53.1006 14.0006 106.6045 0.64/0.53

Spezyme CP 13.4011 83.5068 17.1007 87603.5 0.14/012

Novozyme 188 10.6001 66.3007 34.1020 21290123 8.32/6.85

MU't'feCt 5.0003 31.0017 7.2010 45.3058 0.61/0.50
Xylanase

MU'F'feCt 8.3000 51.9002 12.2002 75901.3 028/023
Pectinase

 

The nitrogen content of enzyme solution, its protein precipitate and ammonium ion was

analyzed to show the protein concentration and other non-protein nitrogen source such

as ammonium in the commercial enzyme complexes (Table 15). Protein content in all

the tested enzymes ranged from 31-84 g/ L. The protein nitrogen to total nitrogen ratio

was between 60—80% for all tested enzymes with exception of Novozyme 188, in which

about 70% of the total nitrogen is from non-protein sources. Approximately 30% of the

total non-protein nitrogen of Novozyme 188 comprises of ammonium ions, 3

commonly-used nitrogen supplement in commercial fermentation. To accurately

quantify protein content of a complex media, protein precipitation followed by nitrogen

content analysis has been demonstrated to be effective to avoid interference due to

soluble impurities in the enzyme solution.

123

 

Er

Er

at

ut
37

c0
mc
(El
inc

0f

Thi
ree
loa
Acc
ran
Wh.‘
ALK
frOr
50a

leo

Enzymatic hydrolysis at different protein loadings

Enzymatic hydrolysis at 18.1% solids loading using a complex of commercial enzymes
(Spezyme CP, Novozyme 188, Multifect Xylanase and Multifect Pectinase) yielded sugars
at 110 g/L using protein loading at 11.1 mg protein/g biomass. Enzyme hydrolysis
utilizing Accelerase only achieve a hydrolysis with sugar concentration at 102 g/L and 96
g/L; at respective protein loading of 5.6 and 3.8 mg/g biomass (Figure 29). Protein
concentration of respective enzymes were analyzed and listed in Table 15. Oligomers to
monomers ratio increased from 33% (EH3 in hydrolysis uses 4 complex enzymes) to 80%
(EH2, in hydrolysis uses 3.8 mg/g protein loading). Total oligomers concentration also
increased from 27 g/L (EH3) to 43 g/L (EH2). Xylooligomers are the largest (60%) group

of oligomeric sugar presented across the hydrolysates (Figure 29).

This suggests that carbohydrates in AFEX-pretreated corn stover can be hydrolyzed
readily to a mixture of oligomers and monomer at high solids loading and low enzyme
loading (3.8-5.6 mg protein/g biomass). Assuming that 70% of the total protein in
Accelerase is cellulolytic enzymes, cellulase loading responsible for these hydrolyses
ranged from 2.7-3.9 mg cellulase/g biomass. This level of enzyme loading is close to
what utilized in the existing corn ethanol industry [91].

ALK2 have been shown to hydrolyze oligomers present in the enzymatic hydrolysate
from AFEX-pretreated corn stover. Further integrating a CBP ethanologen such as T.
saccharolyticum ALK2 that secrete various cellulolytic enzymes particularly for

xylooligomer hydrolysis can be highly advantageous due to reduction in requirement

124

and all:

enzym

Flgu

Not

and alleviation of sugar inhibition through simultaneous sugar consumption and

enzymatic hydrolysis.

" OligA |]I|]]]]]]]]]]]|I[) MoA E OligX
-:. Mo)( W OligG' W MoG

 

120 .0 .“

 

100*

80_ _— — _

  
 

   

Sugar Concentration (g/L)

 

 

EH2 EH3
Enzylme(s) Used for Hydrolysis
Figure 29 Sugar composition of enzymatic hydrolysate from AFEX-pretreated corn
stover utilizing Accelerase at 5.6 (EH1) and 3.8 (EH2) mg protein/g biomass and
a combination of four complex enzymes (Spezyme CP, Novozyme 188,
Multifect Xylanase and Multifect Pectinase; EH3).

Note: Olig: Oligomeric, Mo: Monomeric, A: Arabinose, G: Glucose, X: Xylose.

125

Wate
Ferm
glucc
bet“
wate
degr
Nev.
30).

lign

COl'l

The
the
hy«
a S
of
5U

UT

 

Water extract fermentation

Fermentation media consists of AFEX-CS water extract supplemented with additional
glucose and xylose. The total sugars are equal in all the fermentations, but the ratios
between oligomeric and monomeric sugars are different because most of the sugars in
water extract of AFEX-CS was oligomers. At low solids loading (3-9% SLE, low levels of
degradation compounds), glucose was consumed at a higher rate than xylose,
Nevertheless, xylose was consumed faster than glucose at SLE of 12% and 15% (Figure
30). This feature is distinctively different from most reported co-fermentation of
lignocellulosic fermentation where xylose is inhibited when fermentation was

conducted in high SLE [3, 47].

The primary challenge in utilizing ALK2 in industrial cellulosic ethanol production is on
the growth robustness of this strain in sub-optimal growth media such as lignocellulosic
hydrolysate. Fermentation at media containing 15% of SLE was prolonged (Figure 30). At
a similar SLE and nutrient level, S. cerevisiae 424A(LNH-ST) was able to consume 30 g/L
of glucose completely within 24 hr [47]. Provided that overall growth robustness can be
substantially improved through adaptation as demonstrated [3], by exploiting the
unique feature of ALK2 on carbon source preference and CBP ability, T. saccharolyticum
ALK2 could provide a highly economical processing approach for cellulosic ethanol

production.

126

H;

CO
FEr
SOC
Dl’O
Abo
eng,‘
Xylo.
high

 

 

 

 

 

 

 

 

 

 
 

 

A7 J 812 l l T fj l '

:T " :10 ‘ 7

ED 39 . —-—3%

8 ‘ U ‘) +591

0 g 8 \ o _

3 _ g; . +9%

§ § 6- +12% -

5 - ,3 - ——<—15%

.2 ‘ '2 4- -

L d O)

“é . g '

8 - c 2- -

O O

2 5 *
r l

 

 

   

40 80 120 160 0 40 80 120 160
Time (hr) Time (hr)

Figure 30 Sugar consumption pattern for (A) glucose and (B) xylose fermentation in

water extract of AFEX—pretreated corn stover at varying solids loading
equivalent

CONCLUSIONS

Fermentation of enzymatic hydrolysate from AFEX-pretreated corn stover using T.
saccharolyticum ALK2 consumed 90% of the total sugars (monomers and oligomers) to
produce ethanol close to the theoretical maximum with productivity of 0.62 g/L/hr.
About 60% of the oligomers were metabolized by T. saccharolyticum ALK2, a genetically-
engineered ethanologen which is capable of secreting hemi-lcellulolytic enzymes.
Xylooligomers were the largest group of unconsumed sugars. Hydrolysates containing
high concentrations of oligomeric sugars can be generated from AFEX pretreated corn
stover at low enzyme loading (3.8-5.6 mg protein/g corn stover). ALK2 has unique

127

carbon source preference; xylose was favored relative to glucose in a media containing
high levels of degradation compounds. However, higher growth robustness at lower
nutrient levels is likely required for a commercial production. Utilizing T.
saccharolyticum ALK2 as the fermentation platform could increase the overall yield to

76 gal/ton due to Oligosaccharides utilization.

128

APPE

AMN

DlSTl

INTRI

DEpe
serio
Ethai

PTOrT

Cellu
EStin

aChH

 

APPENDIX B ETHANOLIC FERMENTATION OF HYDROLYSATES FROM
AMMONIA FIBER EXPANSION (AFEX) PRETREATED CORN STOVER AND

DISTILLERS GRAIN

INTRODUCTION

Dependence on petroleum as the nearly exclusive source of transportation fuel poses a
serious threat to our economy, environment and energy security [9]. Recently, cellulosic
ethanol has gained wide attention from both governmental and private sectors as a
promising alternative fuel option due to the abundance of cellulosic materials [92].
Amongst cellulosic materials, corn stover (CS) is considered an important feedstock for
cellulosic ethanol production. The National Renewable Energy Laboratory (NREL)
estimated that a harvest of 80-100 million dry tonnes per year of corn stover can be

achieved in a sustainable fashion [8].

129

Unfortunately, cellulosic substrates are perceived to be nutrient-deficient. Hence, extra
costs due to external nutrient supplement are deemed unavoidable [72]. Unlike most
cellulosics, corn distiller’s dried grains with solubles (DDGS) is a nutrient-rich feedstock
derived from the corn-based ethanol industry [93]. DDGS is used for animal feed due to
its high nutrient content [94]. Nonetheless, the potential to utilize DDGS for nutrient
supplementation in cellulosic ethanol production has apparently not been recognized.
Instead, rich laboratory media were supplemented to conduct efficient fermentations
on hydrolysates from corn stover treated by various pretreatment technologies [15, 29,
95, 96]. In addition, cellulosic hydrolysates contain degradation products which could
inhibit both enzymatic hydrolysis and fermentation [35, 97]. Fermentation at high solids
loading and recycle of liquid streams in industry will probably lead to further build-up of
inhibitors. Thus, various methods have been investigated to remove the toxins [20, 36,
98, 99]. However, the cost of detoxification can be as high as 22% of the total ethanol
production cost [38]. To avoid detoxification, efforts have been made to increase
through genetic engineering the tolerance toward inhibitors of ethanologens [100]. The
technique of selective evolution had been exploited to increase the ethanol tolerance
limit of Escherichia coli K011 without altering the genetic map [59]. It is expected that a
similar technique can be utilized to enhance ethanologenic strains in terms of

robustness and fermentation performance.

130

Escherichia coli K011 is one of the very first ethanologens engineered for fermentation
of sugars from cellulosic materials. E. coli B, the parental strain of E. coli K011, has the
natural ability to utilize several pentoses and hexoses found in cellulosic hydrolysates.
The integration of pyruvate decarboxylase (pdc) and alcohol dehydrogenase II (ath)
from Zymomonas mobilis into E. coli B resulted in high ethanol yield [28]. However, its
ethanol tolerance and growth robustness are relatively low compared to other

ethanologens [27].

Ammonia Fiber Expansion (AFEX), recognized as one of the leading pretreatment
technologies [2], offers several advantages including reduced production of inhibitory
compounds and nutrient addition due to residual ammonia [22]. Unlike other
pretreatments such as dilute acid pretreatment and steam explosion, AFEX is a dry-to-
dry process (no liquid stream produced after pretreatment) with essentially complete
solids recovery. I believe that the AFEX pretreatment could make possible a new
paradigm of cellulosic ethanol fermentation: a paradigm in which neither extensive
detoxification nor significant nutrient supplementation is required. This paper
represents an early step toward realizing this new paradigm. Therefore, hydrolysis and
fermentation on AFEX-treated feedstocks under the stringent baseline conditions (low
initial cell density, without washing the pretreated materials, detoxifying or
supplementing the hydrolysate with nutrients) are investigated. The key questions

investigated here are: i) the feasibility of ethanol fermentation on AFEX-treated com

131

 

stover hydrolysates at industrially-relevant concentrations under the baseline
conditions, ii) the value of nutrient-rich feedstocks such as DDGS in fermentation and iii)
the usefulness of selective evolution in adapting ethanologenic strains for fermentation.
Engineered ethanologen E. coli K011 is used to conduct these investigations in Separate
Hydrolysis and Fermentation (SHF). In some ways this represents a ”worst case” for

hydrolysis and fermentation of AFEX-treated CS and DDGS.

132

Ml

of

re-

pri

Bu

cel

etl

AF

AF

Dre

inc

MATERIALS AND METHODS

Materials

Corn stover (CS), premilled and passed through a 10 mm screen, was a kind gift from the
National Renewable Energy Laboratory (NREL, Golden, Colorado). The moisture content
of the untreated CS was about 7% (total weight basis). Feedstock analysis by NREL
revealed an estimated composition of 34.1% cellulose, 22.8% xylan, 4.2% arabinan, 2.3%
protein in the corn stover. The DDGS, obtained from Big River Resources (West
Burlington, IA), was dried to roughly 11.5% (total weight basis). According to analysis
performed at the Purdue University, the content of the DDGS consists of 16.006.6%
cellulose, 8.2:3.3% xylan and 5.213.3% arabinan, 5.2% starch and 26.4 % protein. The
ethanologenic strain E. coli K011 was obtained from American Type Culture Collection

(ATCC) with designated number 55124.

AFEX pretreatment

AFEX pretreatment was conducted on CS and DDGS under the optimal conditions as
previously described [101, 102]. For CS, pretreatment temperature was 90°C with
moisture content of prewetted CS at 60% (dry weight basis) and ammonia loading of 1.0
kg anhydrous ammonia/1.0 kg dry corn stover. The pretreatment of DDGS was at 70 °C
with moisture content of DDGS at 13.0% (dry weight basis) and ammonia loading of 0.8
kg anhydrous ammonia/1.0 kg dry DDGS. For both feedstocks, after holding the vessel at

the target temperatures for five minutes, pressure was rapidly released to accomplish

133

theta
[102'

ah-d

Enzy
The
LAP-
CA)
ceH

50°

Hig
The
gra
con
AF
de
ad
39

en

En

the expansion. AFEX-apparatus description and operational details were as reported
[102]. After the pretreatment, the materials were left under a fume hood overnight to

air-dry the treated feedstock.

Enzymatic hydrolysis

The enzymatic hydrolysis of AFEX-treated CS and DDGS was carried out based upon the
LAP-009 protocol from NREL. Commercial enzymes, Spezyme CP (Genencor, Palo Alto,
CA) and B—glucosidase (Novozyme 188, Bagsvaerd, Denmark), were loaded at 15 FPU/g
cellulose and 64 pNPGU/g cellulose respectively. The reaction was conducted at pH 4.8,

50°C and 200 rpm shaking.

High solids loading enzymatic hydrolysis

The solids loading for each biomass sample was approximately 18% by weight (180
grams of dry solids per liter of hydrolysis mixture). Due to the difference in cellulose
contents, the corresponding cellulose loadings were 6% for CS and 3% for DDGS. The
AFEX-treated feedstocks were enzymatically hydrolyzed without any prior washing or
detoxification. For C5, the feedstock and the corresponding enzyme loadings were
added fed-batch wise with 9% solids loading each for two consecutive days. To be
specific, in an initial saccharification mixture of 448 g, 44 g of dry CS and corresponding
enzyme solutions were added to 396 mL of buffer solution (0.06 M citrate buffer, pH4.8)
in a 1L unbaffled Erlenmeyer flask. After 24 hr, another 44 g of dry CS and corresponding

enzymes solutions were added into the reaction flask to complete the fed-batch

134

addition with a final saccharification mixture of 500 g. Chloramphenicol (Cm) was added
to a final concentration of 50 mg/L to prevent microbial growth. The total enzymatic
reaction time was 144 hr. The hydrolysis of DDGS was conducted in single batch for 72
hr. The final hydrolysates were centrifuged and the supernatants were sterile-filtered

for further fermentation studies.

1% cellulose loading enzymatic hydrolysis

The 1% cellulose loading hydrolyses were conducted to compare the effect of solids
loading on sugar yields for both AFEX-treated CS and DDGS. The experiments were
conducted in a total 15 g saccharification mixture (AFEX-treated CS or DDGS in buffered
enzyme solution) in screw-capped vials with a total volume of 25 mL. Both feedstocks
were added in a single-batch, the reaction times for CS and DDGS were six days and

three days, respectively.

Mass balance for enzymatic hydrolysis

After enzymatic hydrolysis, reaction mixtures were centrifuged to separate the pellet
from the supernatant. Only glucose and xylose were determined during the analysis
because they are the predominant monomeric sugars present in CS-hydrolysate which
are also fermentable by E. coli K011. To obtain accurate sugar yield estimates,
monomeric sugars content in supernatants, pellets and reaction flasks (due to
irretrievable hydrolysate left in the flasks) were taken into account in the mass balance

calculation. The pellets formed after centrifugation were washed with distilled water at

135

fixe

W35

W35

mo

COl'l

anc

sug

not

Cul

All:

VOIl

sha

car

SEIi

KO:

Seri

8r0

Unc

fixed proportions (1 g wet pellet in 10 mL water). The sugar concentrations of the
washed streams and the moisture content of the washed pellets were analyzed. The
washed solids dry matter was calculated based on the moisture content. To estimate
monomeric sugars retained in the pellets, it was assumed that 1) the monomeric sugar
concentration in washed stream was equal to that of the solution contained in the pellet
and 2) the density ofthe solution in the pellets was 1.0 g/mL. Theoretical possible
sugars are calculated based on cellulose and hemicellulose i.e. starch from DDGS was

not included in the calculation.

Culture and fermentation

All cultures and fermentations, if not stated otherwise, were carried out at a working
volume of 100 mL solution in 250 mL unbaffled Erlenmeyer flask at 37°C, 100 rpm
shaking. The flasks were capped with rubber stoppers pierced with a needle to vent

carbon dioxide.

Selective evolution

K011 was selectively evolved in two stages, first on solid medium consisting of
hydrolysate resulting from 3% cellulose loading of CS-hydrolysis mixture and followed by
serial culture in liquid hydrolysates. The ethanologenic strain, E. coli K011, was first
grown overnight at 35°C in Luria-Bertani (LB) agar plate supplemented with 2% xylose

under 50 mg/L Chloramphenicol (Cm) selective pressure. The raised colonies were

136

tra

Crr

cel

lnit
37°
wa:
int
suc.
cell
3.5

wer
Ioac

inoc

K01
The

Strai

fliecj

OVEH

transferred to solid medium (CS-hydrolysate) and grown successively at two alternate

Cm concentrations (50 mg/L and 600 mg/L) for three cycles.

In serial culture, the concentration of antibiotic Cm was maintained at 100 mg/L. The
cells on solid medium were inoculated into hydrolysate to an initial ODSSOnm of 0.1.
Initially, the hydrolysates were kept at neutral pH. Each culture was grown for 24 hr at
37°C before successive transfer to a more concentrated hydrolysate. No pH adjustment
was made during these cultures. The corresponding cellulose loading of the hydrolysate
in the first liquid culture was 3.5% and this was followed by 0.5% increment in each
successive transfer until reaching 6%. Culture transfer to the hydrolysate with higher
cellulose loading was conducted only when the culture achieved an 00550.1,11 greater than
3.5 within 24 hr. Mutants adapted to the CS hydrolysate at each stage of serial culture
were stored in 30% glycerol at -70°C. The mutant strain isolated from 6% cellulose-
loading—CS-hydrolysate was termed ML01. This preculture of ML01 was used as the

inoculum for the designated fermentation experiments described below.

K011 seed preparation

The K011 strain was grown on LB plate (2% xylose) under Cm selective pressure. The
strain was grown successively under alternate Cm concentrations i.e. 50 mg/L and 600
mg/L at 35°C. This cycle was repeated three times before the cells were inoculated to LB
media (5% xylose, 100 mg/L Cm) to an initial ODSSOnm of 0.1. The K011 seed grown

overnight was used as the inoculum for the designated fermentations.

137

Fermentation media preparation

Hydrolysates from enzymatic saccharification of 6% cellulose-loading-CS (Hydrolysate A)
and 3% cellulose-loading-DDGS (Hydrolysate C) were the principal media used in the
fermentation. Other fermentation media were formulated based on these two
hydrolysates. They are listed as follows:

0 Hydrolysate B: A mixture of Hydrolysate A (CS-hydrolysate) and distilled water at
7:3 volume ratio.

a Hydrolysate D: A mixture of Hydrolysate C (DDGS-hydrolysate) and externally
added reagent grade glucose (9 g/L) and xylose (22 g/L). These sugars were
added into Hydrolysate C to increase their concentrations to levels comparable
with Hydrolysate A.

o Hydrolysate E: A mixture of Hydrolysate A and Hydrolysate C at 7:3 volume ratio.
Hydrolysate E was used to elucidate the effect of DDGS—hydrolysate on
fermentation with Hydrolysate B serving as the control experiment to determine

the effect of dilution of the Hydrolysate A on fermentation.

Antibiotic Cm was added at a concentration of 50 mg/ L. Total Cm concentration

accumulated from both enzymatic hydrolysis and fermentation steps was 100 mg/L. All

hydrolysates were pH-adjusted using NaOH and sterile-filtered.

138

Ethanol fermentation

Fermentation was carried out under semi-sterile conditions. The initial ODSSOnm of
fermentation was predetermined as 0.05 (approximately 16 mg dry-weight/L).
Designated inocula were centrifuged and the cell pellets were resuspended into the
fermentation medium. Growth was at 37°C with 100 rpm shaking. NaOH (5 M) was
manually added to adjust the pH to neutral every twelve hours. Before and after each
pH measurement and adjustment, the pH probe was thoroughly rinsed sequentially in
four test tubes filled with sterile distilled water and dried with sterile wipes. Samples
(1.0 mL) were taken manually at defined periods. The volume changes due to pH
adjustment and sampling were recorded. No detoxification step and no external
nutrient addition were used. Every fermentation experiment was conducted in

duplicate.

Compositional analysis

The glucose and xylose concentrations in hydrolysates during the course of
fermentation were determined by HPLC with a Biorad Aminex HPX-87P column
(Hercules, California). The HPLC system was equipped with a Waters 410 Differential
Refractometer (Milford, Massachusetts), HPLC pump (Waters 515) and aided by Waters
717plus autosampler. The temperature of the column was maintained at 85°C and the
eluant (water) flow rate was 0.6 mL/min. Glucose and ethanol concentrations were

measured using a YSI 2700 Select Biochemistry Analyzer (Yellow Springs, Ohio) to

139

ei
pa

CE

Mc
ma
MX
em
in p
Wag

confirm the accuracy of the HPLC analysis. Operational details were as described in the

user manual.

Ethanol yield and volumetric ethanol productivity calculations

In this report, ethanol yields were expressed as grams of ethanol produced per gram of
sugars consumed to assess the efficiency of the fermenting strains in converting the
sugars to ethanol. The theoretical yield of ethanol for all consumed monomeric sugars
was 0.51 g EtOH/g sugar. Volumetric ethanol productivity was calculated based on
ethanol produced at the time available glucose was completely consumed. This
parameter along with cell density data were used to evaluate the fermentability of

cellulosic hydrolysates for ethanol production.

Moisture content analysis and cell density measurement

Moisture content analysis was performed to determine the dry weight of solid
materials. In this report, moisture content was measured by an A 81 D Moisture analyzer
MX-50 (Milpitas, California) to determine the weight of feedstocks for hydrolysis
experiments. In addition, this analysis was carried out to determine the liquid retained
in pellet for calculation of monomeric sugar yields. The optical density of the cell culture
was measured using a Turner SP-890 UV Visible Scanning Spectrophotometer (Dubuque,

Iowa). The cell density of fermentation broth was measured at a wavelength of 550nm

140

 

(1 cm light path) after the sample was taken and cooled to 4°C. Readings were corrected

with the dilution factors as necessary.

RESULTS

AFEX pretreatment and enzymatic hydrolysis

AFEX did not significantly change the physical appearance of treated materials except
that they become slightly darker. The moisture contents of the air—dried AFEX-CS and
DDGS were 10.5% and 6% (total weight basis) respectively. CS was added in two batches
to ensure proper liquefaction and mixing of the hydrolysate. After addition of each CS
batch, it was observed that the slurry-like mixture liquefied within two hours. Thus,
mixing was not a limiting factor for enzymatic hydrolysis at 18% solids loading of AFEX-
treated corn stover. After six days of reaction, a hydrolysate containing 52.0 g/L of
glucose and 22.2 g/L of xylose were obtained. The monomeric sugar yields were 65.0%
for glucose and 42.1% for xylose. In 1% cellulose loading saccharification, glucose and
xylose yields for CS after six days were 95% and 70% respectively. Thus, sugars
obtainable from AFEX-pretreated C5 are limited more by enzyme properties than by
inherent reactivity of the AFEX-treated material.

DDGS hydrolysis at 18% solids loading was carried out in a single batch. After three days
of reaction, the resulting DDGS hydrolysate contained 40.0 g/ L of glucose,

corresponding to 97% glucose yield. Xylose yields were minimal (<1.0 g/L). The cellulase

141

 

enzyme complex apparently lacks an effective xylanase for DDGS xylans while this
xylanase is somewhat effective in hydrolyzing AFEX-treated CS. In 1% cellulose loading
saccharification, the glucose yield after 3 days was 105%. Thus essentially, complete

cellulose conversions were achieved in both hydrolyses.

Selective evolution of K011

A serial culture of K011 on solid medium followed by liquid medium was successfully
utilized to fully adapt the ethanologen to the AFEX-treated-CS-hydrolysate environment
while maintaining its efficient ethanol production metabolic pathway. K011 was
transferred to the solid medium (3% cellulose loading CS hydrolysate) with 50 mg/L Cm.
The cell growth was relatively poor but several colonies were able to thrive. These
colonies were transferred to similar solid medium with 600 mg/L Cm. Cell growth was
still poor. However, with each subsequent transfer cycle, the cell growth significantly
improved. After three cycles of culture, the cells from the solid medium were inoculated
into CS-hydrolysate. The cell density (ODSSOnm) after overnight culture at 37°C was 4.2.
Each of the successive transfer cultures in hydrolysates at higher cellulose loadings was

able to obtain similar densities (ODSSOnm between 3.8 and 4.2).

Fermentation of corn stover hydrolysate
The fermentation performance of K011 and adapted mutant ML01 were compared in
Hydrolysate A. Both fermentations were carried out successfully with approximate

optical density (550 nm) of 4.0 at stationary phase. Glucose (52.0 g/L) was completely

142

 

sh

SU

consumed within 72 hr of fermentation using ML01, resulting in 24.9 g/L ethanol
concentration, for an ethanol yield of 93.1% based on total consumed sugars (Figure 31
A, B). A prolonged lag phase of about 24 hr was observed in K011 fermentation while no
observable lag phase was seen for ML01 (Figure 31 D). Ethanol productivity doubled for
ML01 compared to K011 based fermentation (Table 16). Surprisingly, xylose was not
consumed by either strain during the course of fermentation of Hydrolysate A.
However, 58.2% of the xylose in Hydrolysate B was utilized by ML01 (Figure 31C).
Control experiments at comparable sugar concentrations (50 g/L glucose, 20 g/L xylose)
using Luria-Bertani (LB) media (10 g tryptone, 5 g yeast extract, 10 g NaCl in 1.0 L)
showed similar cell densities as that of Hyd rolysate A with complete utilization of both

sugars (results not shown).

143

 

Hod). Uta HHOx .QOU m. mEmtm Ecomoscmim >9 wcmuoEmch 2025:0658 0‘0 \CmEESM ma .2th

..mmzm E8328 3103.“. m Hmd mm; mcmmzm 50.. co“. .0828 “—o Em... .833005 .30.? new 382m 383:8 co comma 22:
823.6 .8 “.3888 mm; 80.328388 .0823 0

358.398 38.32. 05 *0 80.833. Panama 05 33> c3588 mcotm
~38..an “oz ”.<.z .cozmbcmucou “ucou

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.3.0 3.3.: 8.238 0.03.8 H.308 H32 ms 96.
E u new < 3322?»:
w.m-o.m N.HHHo.mN Nodwmmd wduvém mdflméN HHOV. .o 335.). w
. . . i . . . . . . i . u 332962..
c wé m H o+w mm mo oumo H .V 3K mm m o+w Nm H35. + mcmmam o
. . . . . . . i . . i . $8.88. 30.2.3
m w m N < 2 Ho oHoN H N o+m NoH v o+m mH H35. *3 33282:. moon u
mun
. . . i . . i . . i . . i . 0.32 am .33., 82.5.8
m 3m v N mH+m Hm No o+3 c N N+m mm m o+w mH H35. new < 3329?»: m
H6 335.).
méNé H.Hfimd Hodugd H.HHNMm vNHQmN 3...). 3:68. 322.8 £63
v.v-m.m HdHHd HodeNd odHNNm odnméN HHOV. 33292:. 8 <
.230.
AEMWMMwO. 80.2.8388 3.28388“. NW...” SE. aucou 3.830. .33
320.5% 303x 838.38.). a Embm
um 37.ch :3 3269:. .0825 332062..

 

 

 

 

 

 

H35. new HHOV. :8 .m 3.23. u.:omo.ocm£m >9 £33822. 803358.66. u.o 33:83 3 2an

144

Fermentation of DDGS hydrolysate

Protein content in DDGS-hydrolysate was substantially higher than CS-hydrolysate and
hence higher fermentability was expected in DDGS-hydrolysate. As illustrated in Figure
32, the fermentation using ML01 yielded high cell density (ODSSOnm between 8 and 9) for
Hydrolysates C and D. However, the cell density of hydrolysate with supplemental
sugars was consistently lower than that of DDGS hydrolysate (Figure 32 A). The cell
densities of both hydrolysates were more than two-fold that of the control experiment
in LB media. Glucose in both hydrolysates was consumed within 24 hr with ethanol
productivity exceeding 1.0 g/L/hr . Nearly 80% of the xylose in Hydrolysate D was
utilized after 48 hr of fermentation. Ethanol yields based on consumed sugars in

Hydrolysates C and D were 103% and 97% respectively (Table 16).

Fermentation of mixed hydrolysate

K011 and ML01 were tested for fermentation using mixed hydrolysates of CS and DDGS
at volume ratio of 7:3 (Hydrolysate E). Both strains obtained similar cell densities with
0055mm, between 5 and 6 at the stationary phase (Figure 33 A). However, ML01
exhibited better performance in terms of the rate of glucose consumption and ethanol
productivity (Table 16). Glucose was completely consumed within 30 hr by ML01
compared to 40 hr when using K011 (Figure 33 B, C). The ethanol productivity of K011
and ML01 were 0.59 and 0.63 g/L/hr, respectively. Ethanol yields close to 100% of
theoretical yield were achieved. Xylose consumption was slightly improved to 10-20% of

total xylose compared to CS-hydrolysate fermentation (<10% total xylose consumed). As

145

 

for the control experiment, Hydrolysate B (70% Hydrolysate A and 30% distilled water)
fermented by ML01 exhibited a lower ethanol productivity (0.44 g/L/hr) and cell density

(4.3-4.5) at stationary phase (Table 16, Figure 31C).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A 50» B 50'
+Glc

:7 4o_ +XV' . a 40.
E —o— EtOH 39
C C
9 30~ . 9 30-
H H
(O CO
5 20. K: 35 20
c ‘ c ’
OJ 0)
U U
S 10» . S 10.
u u

0 . A . . . ; 0 A L . . -- . .1

O 20 4O 60 80 100 120 140 0 20 40 50 80 100 120 140

Fermentation Progress (hr) Fermentation Progress (hr)

40 . . - . - . 5

C D
E
:3 30* g
B m
z m
o o
S I
‘8 {-3
‘1’ c
g 10' a)
o O
U =
8
0 . . . - . _. - _ 0 A L . . .
0 20 40 60 80 100 120 140 0 20 40 60 80 100
Fermentation Progress (hr) Fermentation Progress (hr)

Figure 31 (A) Fermentation of Hydrolysate A (6% cellulose loading CS-hydrolysate) by E.
coli K011, (B)Fermentation of Hydrolysate A by E. coli ML01,(C) Fermentation
of Hydrolysate B (7:10 dilution of Hydrolysate A) by E. coli ML01, and (D) Time
course of cell density during fermentation in Hydrolysate A and B

The starting cell density of the fermentations was set at 0.05 (ODSSOnm) and the data
presented are averages of the duplicate experiments.

146

Feedstock-to-ethanol mass balance

In AFEX-pretreatment, a solid-to-solid operation, no loss of mass was expected and
none was observed. The residual ammonia on treated—materials was negligible (<0.5%
of the total treated material). Under listed enzyme loading and solids loading, 100 g of
dry CS yielded 24.6 g of glucose and 11.5 g xylose. About 37.0 g of the dry solid residual
material remained after the hydrolysis. Fermentation with K011 and ML01 produced
12.0 g ethanol with similar yield (93%). The final ethanol concentrations were about 25

g/L. However, the xylose consumption was below 10% in both cases.

For DDGS, 17.2 g of glucose was produced from 100 g dry DDGS with hydrolysis of 3%
cellulose-loading. Half of the initial dry matter content of the DDGS was solubilized into
the hydrolysate (Figure 35). The ethanol yield in both DDGS fermentations was at the
theoretical limit with ethanol concentrations of 19.6 g/L and 32.8 g/L for Hydrolysate C
and D, respectively. Glucose and xylose utilizations in Hydrolysate D were essentially

complete.

DISCUSSION

High solids loading enzymatic hydrolysis
As expected, one of the significant limiting steps in converting C5 to ethanol production

was enzymatic hydrolysis using the selected enzymes and loadings. The sugar yields

147

decreased 30% when the cellulose loadings were increased from 1% to 6%. At higher
solids loadings, the concentrations of end products (sugars) and degradation products
increase. Monomeric sugars, particularly glucose, dramatically inhibit the cellulase and
B-glucosidase activities. As suggested by previous studies, simultaneous saccharification
and fermentation (SSF) can be utilized to reduce end product inhibition [23, 103]. The
inhibitory effect of degradation products released during pretreatment (wet oxidation)
of wheat straw on cellulases and hemicellulases has been reported [97]. Even though
AFEX produced degradation compounds at a reduced level [2], it is not surprising that
similar inhibitory effects can be found in CS-hydrolysate at high solids loading. In
addition, xylose yields were lower than glucose yields in both hydrolyses at 1% and 6%
cellulose loadings. Evidently, an optimal enzyme cocktail which contains sufficient
activities for both cellulose and hemicellulose is needed to improve xylose yield in AFEX-

treated CS.

Similarly, the enzymatic hydrolysis of cellulose in DDGS at 1% cellulose loading achieved
higher yield than that of 6% cellulose loading. However, the difference was statistically
insignificant due to relatively large margin of error in cellulose content determination. A
more precise analysis of composition is required to calculate sugar yield accurately for
DDGS. Due to insufficient hemicellulolytic activity of commercial enzymes toward DDGS,

the xylose yield was expected to be minimal, as indeed occurred.

148

 

 

 

 

 

 

 

 

 

 

 

 

 

10 I I I
A . .
8 - .
a a - «
2 8 5 - -
a a
= D 4 ' ‘
a) O
U - 2 ’ —0—Glucose Supplemented ‘
( + No Supplementation
O L 1 . 1 1
0 20 40 60 80
Fermentation Progress (hr)
B so I ' I If ' l
:7 —I—- Ic .
E 40 - + XVI ~
.5 30 - + 30” .
E . .
‘E 20 - .
8 l .
6 1o - _
U I
0 1 _ 1 _. 1 .— 1 .
0 20 40 60 80 100
Fermentation Progress (hr)
C 40 l ' l ' U
:3 ——-—— Glc
E 30 ' _._ EtOH ‘
.5 ' ‘
E 20 - = g .
E s
8 10 - -
C
8 .
0 1 L . _ L _. 1 A
0 20 40 60 80

Figure 32 (A) Time course of cell density during fermentation in Hydrolysate C (3%
cellulose loading DDGS hydrolysate) and D (DDGS hydrolysate with sugars
addition), (8) Fermentation of Hydrolysate C by E. coli ML01, (C) Fermentation

Fermentation Progress (hr)

of Hydrolysate D by E. coli ML01.

The starting cell density of the fermentations was predetermined at 0.05 (ODSSOnm)

and the data presented are averages of the duplicate experiements.

149

>

Cell Density
(OD 550nm)

w

30

10

Concentration (g/ L)

50

O

40

20

10

Concentration (g/L)

 

 

+ K011
+ ML01

 

 

Fermentation Progress (hr)

100

 

20

 

* U 1 1 ' I r 61 fr
+Glc
—v—Xyl

——9—- EtOH

 

4|

1 J — .L

A!‘

4|

1 -

 

 

20

Fermentation Progress (hr)

40 60 80 100 120 140

 

30

 

l l

_ 1 4-1

 

 

20 40 60 80 100 120

Fermentation Progress (hr)

140

Figure 33 (A) Time course of cell density during fermentation in Hydrolysate E (Mixture
of CS and DDGS hydrolysates at ratio 7:3 ), (B) Fermentation of Hydrolysate E
by E. coli K011, (C) Fermentation of Hydrolysate E by E. coli ML01.

The starting cell density of the fermentations was set at 0.05 (ODSSOnm) and the data
presented are averages of the duplicate experiments.

150

Fermentation

Fermentation of corn stover h ydrolysate

The hydrolysates of AFEX-treated CS showed high fermentability. Under the baseline
conditions (low initial cell density, without washing the pretreated materials, detoxifying
or supplementing the hydrolysate with nutrients), AFEX-hydrolysates were benign to the
growth of E. coli, the least robust genetically-engineered ethanologen amongst its
counterparts [27]. Nevertheless, I am fully aware that pH of the hydrolysate is crucial
factor that affects the toxicity of certain compounds, e.g. organic acids. This result
indicated that the nutrient content of the CS-hydrolysate is not as poor as generally

perceived because cell densities similar to culture in LB media were achieved.

Fermentation of DDGS and mixed h ydrolysates

DDGS hydrolysate exhibited very high fermentability. There are two general trends
observed in the fermentations using hydrolysate mixture of CS and DDGS. Compared to
control experiment (Hydrolysate B) which indicating the effect of dilution on the
fermentability of Hydrolysate A, ethanol productivity and cell density at stationary
phase both increased when DDGS-hydrolysate was part of the fermentation mixture.
DDGS improves the fermentability of cellulosic hydrolysates presumably due to high
nutrient content in DDGS. It is noteworthy that 25% of the crude protein and 42% of the
fat from the original DDGS were solubilized in the hydrolysate during saccharification

[101].

151

DDGS as the source for nutrient supplement

DDGS is a nutrient-rich cellulosic feedstock, which can potentially be used as the source
for fermentable sugars and nutrient supplement. The difference in fermentation profiles
of Hydrolysate A and E indicates that DDGS enhance glucose utilization and cell growth
in the fermentation (Figure 34 A & B). Complete glucose utilization was achieved within
48 hr in Hydrolysate E compared to 96 hr in Hydrolysate A. Fermentation in LB-
supplemented CS-hydrolysate (Hydrolysate F) exhibited similar improvements i.e.
increased in cell density and glucose consumption rate (Figure 34 A & B). This further

affirmed the feasibility to replace external nutrient supplements using DDGS.

Xylose utilization

Regardless of the strains used, xylose consumption was low in Hydrolysate A and E. Both
of the hydrolysates produced comparable final ethanol concentrations even though
Hydrolysate E exhibited better fermentability. Interestingly, Hydrolysate B (diluted from
hydrolysate A) produced lower ethanol concentration with more than 60% xylose
consumption. In contrast, xylose consumption was almost complete in DDGS
hydrolysate (Hydrolysate D) with the highest final ethanol concentration amongst tested
hydrolysates. One plausible interpretation of these results is that both the ethanol
concentration and inhibitory compounds present in CS-hydrolysate could reduce xylose
utilization but the levels of these inhibitory compounds were low enough in DDGS
hydrolysate so as not to interfere with fermentation. It is worth pointing out that the

nature of the lignocellulosic content in each feedstock influences the formation of toxic

152

degradation products. In particular, the higher lignin content in CS increases the
potential production of phenolic-based-compounds which are highly inhibitory [35].
Overall, this finding presents a challenge for efficient xylose utilization in co-

fermentation of unconditioned CS—hydrolysate with low initial cell density.

 

>
00

h 03

Cell Densnty
(ODSSOnm)

 

 

 

 

0 __. . . . .
o 20 4o 60 80 100
Fermentation Progress (hr)

 

I f I ' r ' I ‘ I ' I

  
  
  

U1
O

+ CS-I-LB
+ CS+DDGS ‘
+ C5 (no Suppl) ..

A
O

N
O
I V

 

 

   

A l l l l

0 20‘40 6 so 100120 140
Fermentation Progress (hr)

A. l -

 

Glucose Conc (g/L) CD
8

H
O O
. I .

Figure 34 (A) Time course of cell density, (B) Glucose consumption profile in
fermentation of corn stover hydrolysates with or without nutrient
supplementation.

The starting cell density of the fermentations was predetermined at 0.05 (ODSSOnm)
and the data presented are averages of the duplicate experiments.

153

Comparison between K011 and ML01

The mutant strain ML01 exhibited higher ethanol productivity and the difference was
more pronounced in CS-hydrolysate which has lower fermentability. ML01 has a higher
tolerance and adaptability to the CS-hydrolysate environment and thus performed
better than K011. In terms of ethanol yield and xylose utilization, both K011 and ML01
displayed similar performance. However, this technique can be very useful to generate

suitable strains for fermentation at higher solids loadings.

Significance of the results

To our knowledge, this was the first successful ethanolic fermentation of cellulosic
hydrolysates at low initial cell density, without washing, detoxification and external
nutrient supplement. This breakthrough was largely made possible by the nature of the
AFEX pretreatment. AFEX produces relatively low levels of degradation compounds
(Chundawat, personal communication). Thus, the technological challenges to improve
the fermentability of hydrolysates are significantly decreased. Various physical, chemical
and biological detoxification methods have been explored, primarily for acid pretreated
biomass [35, 99, 104, 105]. Nevertheless, detoxification is a significant economic burden
for large-scale production of cellulosic ethanol. According to analysis done at the Lund
University, 22% of the total cost of cellulosic ethanol production was estimated to be

due to detoxification processes [38].

154

The utilization of DDGS, a feedstock with high nutrient and low lignin content, can help
achieve efficient fermentation of cellulosic materials. The fermentation of mixed
hydrolysate sugars from nutrient-poor (CS) and nutrient-rich (DDGS) feedstocks could

provide a viable option for large-scale industrial production of cellulosic ethanol.

CONCLUSIONS

Hydrolysates from both AFEX-pretreated corn stover and DDGS showed high
fermentability with excellent ethanol yield in both K011 and adapted strain ML01.
Nonetheless, incomplete utilization of xylose in CS-hydrolysates was observed. The
technique of selective evolution and the incorporation of nutrient-rich feedstock greatly
enhance fermentation performance. Cellulosic hydrolysates based on AFEX are not as
nutrient-deficient as frequently perceived; the need for external nutrient
supplementation and detoxification can be reduced through the AFEX pretreatment

method and the utilization of nutrient-rich feedstocks such as DDGS.

155

H35. ..8 .m 8.2% 63826 m8? 80.335883 282. 36 3:6. .8
wwH o8>~o>oz 68".. au m8>~mam 9.7.: 632062. 2.8.6855 6.8 63833953 mm; .300 .m.m2m:m .0826 3 .300 mm 33mm

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

m Hm H 28.
66.3.. >5
$32 $0.3
n m H.m 38206.»: m NNH uzwfiowcwwuchnmmg Goo >5 xmu.< moon. to
. .o 80.68086“. ”3.83.0 _ __ m 8H m SH
.105
6.08336...
1333.563. 34 m8>~o>oz 338m:
coasom cox oEoEE<

mu ~83me

156

REFERENCE

Mosier, N., et al., Features of promising technologies for pretreatment of
lignocellulosic biomass. Bioresource Technology, 2005. 96(6): p. 673-686.

Lynd, L.R., Overview and evaluation of fuel ethanol from cellulosic biomass:
Technology, economics, the environment, and policy. Annual Review of Energy
and the Environment, 1996. 21: p. 403-465.

Lau, M.W., B.E. Dale, and V. Balan, Ethano/ic fermentation of hydrolysates from
ammonia fiber expansion (AFEX) treated corn stover and distillers grain without
detoxification and external nutrient supplementation. Biotechnology and
Bioengineering, 2008. 99(3): p. 529-539.

Dien, B.S., et al., Chemical composition and response to dilute-acid pretreatment
and enzymatic saccharification of alfalfa, reed canarygrass, and switchgrass.
Biomass & Bioenergy, 2006. 30(10): p. 880-891.

Chang, V., M. Nagwani, and M. Holtzapple, Lime pretreatment of crop residues
bogasse and wheat straw. Applied Biochemistry and Biotechnology, 1998. 74(3):
p. 135-159.

Kabel, M.A., et al., Effect of pretreatment severity on xylan solubility and
enzymatic breakdown of the remaining cellulose from wheat straw. Bioresource
Technology, 2007. 98(10): p. 2034—2042.

Luo, C., D.L. Brink, and H.W. Blanch, Identification of potential fermentation
inhibitors in conversion of hybrid poplar hydrolyzate to ethanol. Biomass and
Bioenergy, 2002. 22(2): p. 125-138.

Kadam, K.L. and JD. McMillan, Availability of corn stover as a sustainable

feedstock for bioethanol production. Bioresource Technology, 2003. 88(1): p. 17-
25.

157

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

Greene, N., et al., Growing energy: how biofuels can help end America’s oil
dependence. Nature Resources Defense Council, 2004.

Wyman, C.E., Potential synergies and challenges in refining cellulosic biomass to
fuels, chemicals, and power. Biotechnol Prog, 2003. 19(2): p. 254-62.

Hsu, T.-A., Pretreatment of Biomass, in Handbook on Bioethanol, Production and
Utilization, C.E. Wyman, Editor. 1996, Taylor & Francis: Washington, DC.

Wenzl, H.R.J., The Chemical Technology of Wood. 1970, Academic Press. p. 32-
245.

Mosier, N., et al., Features of promising technologies for pretreatment of
lignocellulosic biomass. Bioresource Technology, 2005. 96(6): p. 673-86.

Wyman, C.E., et al., Coordinated development of leading biomass pretreatment
technologies. Bioresource Technology, 2005. 96(18): p. 1959-1966.

Schell, D.J., et al., Dilute-sulfuric acid pretreatment of corn stover in pilot-scale
reactor - Investigation of yields, kinetics, and enzymatic digestibilities of solids.
Applied Biochemistry and Biotechnology, 2003. 105: p. 69—85.

Zaldivar, J., A. Martinez, and LO. Ingram, Effect of selected aldehydes on the
growth and fermentation of ethanologenic Escherichia coli. Biotechnology and
Bioengineering, 1999. 65(1): p. 24-33.

Palmqvist, E. and B. Hahn-Hagerdal, Fermentation of lignocellulosic hydrolysates.
l: inhibition and detoxification. Bioresource Technology, 2000. 74(1): p. 17-24.

Palmqvist, E. and B. Hahn-Hagerdal, Fermentation of lignocellulosic hydrolysates.
II: inhibitors and mechanisms of inhibition. Bioresource Technology, 2000. 74(1):
p. 25-33.

Zaldivar, J. and LO. Ingram, Efiect of organic acids on the growth and
fermentation of ethanologenic Escherichia coli LY01. Biotechnology and
Bioengineering, 1999. 66(4): p. 203-210.

Martinez, A., et al., Effects of Ca(OH)(2) treatments ("overliming") on the
composition and toxicity of bagasse hemicellulose hydrolysates. Biotechnology
and Bioengineering, 2000. 69(5): p. 526-536.

Berson, R.E., et al., Detoxification of actual pretreated corn stover hydrolysate
using activated carbon powder. Appl Biochem Biotechnol, 2005. 121-124: p. 923-
34.

158

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

Teymouri, F., et al., Optimization of the ammonia fiber explosion (AFEX)
treatment parameters for enzymatic hydrolysis of corn stover. Bioresource
Technology, 2005. 96(18): p. 2014-2018.

Takagi, M., et al. A method for production of alcohol direct from cellulose using
cellulose and yeast. in Proceedings of Bioconversion of cellulosic substances into
energy, chemicals and microbial protein. 1977. ”T, New Delhi.

Xiao, Z.Z., et al., Effects of sugar inhibition on cellulases and beta-glucosidase
during enzymatic hydrolysis of softwood substrates. Applied Biochemistry and
Biotechnology, 2004. 113-16: p. 1115-1126.

Saha, B.C., Hemicellulose bioconversion. Journal of Industrial Microbiology and
Biotechnology, 2003. 30(5): p. 279-91.

Lynd, L.R., et al., Consolidated bioprocessing of cellulosic biomass: an update.
Curr Opin Biotechnol, 2005. 16(5): p. 577-83.

Dien, B.S., M.A. Cotta, and T.W. Jeffries, Bacteria engineered for fuel ethanol
production: current status. Applied Microbiology and Biotechnology, 2003. 63(3):
p. 258-266.

Ohta, K., et al., Genetic improvement of Escherichia coli for ethanol production -
chromosomal integration of Zymomonas mobilis genes encoding pyruvate
decarboxylase and alcohol dehydrogenase ll. Applied and Environmental
Microbiology, 1991. 57(4): p. 893-900.

Sedlak, M. and N.W.Y. Ho, Production of ethanol from cellulosic biomass

hydrolysates using genetically engineered Saccharomyces yeast capable of
cofermenting glucose and xylose. Applied Biochemistry and Biotechnology, 2004.
113-16: p. 403-416.

Mohagheghi, A., et al., Cofermentation of glucose, xylose, and arabinose by
genomic DNA-integrated xylose/arabinose fermenting strain of Zymomonas
mobilis AX101. Appl Biochem Biotechnol, 2002. 98-100: p. 885-98.

Zhang, M., et al., Metabolic Engineering of a Pentose Metabolism Pathway in
Ethanologenic Zymomonas mobilis. Science, 1995. 267(5195): p. 240-243.

Mohagheghi, A., et al., Performance of a newly developed integrant of

Zymomonas mobilis for ethanol production on corn stover hydrolysate.
Biotechnology Letters, 2004. 26(4): p. 321-325.

159

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

Aden, A., et al., Lignocellulosic Biomass to Ethanol Process Design and Economics
Utilizing Co-Current Dilute Acid Prehydrolysis and Enzymatic Hydrolysis for Corn
Stover, N.R.E. Lab, Editor. 2002: Golden, CO.

Lee, Y.Y., P. Iyer, and R.W. Torget, Dilute-Acid Hydrolysis of Lignocellulosic
Biomass, in Advances in Biochemical Engineering, T. Scheper, Editor. 1999,
Springer-Verlag: New York, NY.

Klinke, H.B., A.B. Thomsen, and BK. Ahring, Inhibition of ethanol-producing yeast
and bacteria by degradation products produced during pre-treatment of
biomass. Applied Microbiology and Biotechnology, 2004. 66(1): p. 10-26.

Nilvebrant, N.O., et al., Detoxification of lignocellulose hydrolysates with ion-
exchange resins. Applied Biochemistry and Biotechnology, 2001. 91-3: p. 35-49.

Berson, R.E., et al., Detoxification of Actual Pretreated Corn Stover Hydrolysate
Using Activated Carbon Powder, in Twenty-Sixth Symposium on Biotechnology
for Fuels and Chemicals. 2005. p. 923-934.

VonSivers, M., et al., Cost analysis of ethanol production from willow using
recombinant Escherichia coli. Biotechnology Progress, 1994. 10(5): p. 555-560.

Lawford, H.G. and JD. Rousseau, Corn steep liquor as a cost-efiective nutrition
adjunct in high-performance Zymomonas ethanol fermentations. Applied
Biochemistry and Biotechnology, 1997. 63-5: p. 287-304.

Humbird, D. and A. Aden, Biochemical Production of Ethanol from Corn Stover:
2008 State of Technology Model. 2009: Golden, Colorado.

Yang, B. and CE. Wyman, Pretreatment: the key to unlocking low-cost cellulosic
ethanol. Biofuels, Bioproducts and Biorefining, 2008. 2(1): p. 26-40.

Wyman, C.E., et al., Comparative sugar recovery data from laboratory scale
application of leading pretreatment technologies to corn stover. Bioresource
Technology, 2005. 96(18): p. 2026-2032.

Lloyd, TA. and CE. Wyman, Combined sugar yields for dilute sulfuric acid
pretreatment of corn stoverfollowed by enzymatic hydrolysis of the remaining
solids. Bioresource Technology, 2005. 96(18): p. 1967-1977.

Dien, 85., et al., Enzyme characterization for hydrolysis of AFEX and liquid hot-

water pretreated distillers' grains and their conversion to ethanol. Bioresource
Technology, 2008. 99(12): p. 5216-5225.

160

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

Pordesimo, L.O., et al., Variation in corn stover composition and energy content
with crop maturity. Biomass and Bioenergy, 2005. 28(4): p. 366-374.

Lynd, L.R., et al., How biotech can transform biofuels. Nature Biotechnology,
2008. 26(2): p. 169-72.

Lau, MW. and B.E. Dale, Cellulosic ethanol production from AFEX-treated corn
stover using Saccharomyces cerevisiae 424A(LNH-ST). Proc Natl Acad Sci U S A,
2009. 106(5): p. 1368-73.

Sticklen, M.B., Plant genetic engineering for biofuel production: towards
affordable cellulosic ethanol. Nat Rev Genet, 2008. 9(6): p. 433-43.

Shaw, A.J., et al., 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, 2008. 105(37): p. 13769-13774.

Wang, M., H. Lee, and J. Molburg, Allocation of energy use in petroleum
refineries to petroleum products - Implications for life-cycle energy use and
emission inventory of petroleum transportation fuels. International Journal of
Life Cycle Assessment, 2004. 9(1): p. 34-44.

Wang, M., C. Saricks, and D. Santini, Effects of fuel ethanol use on fuel-cycle
energy and greenhouse gas emissions, U.S.-Department-of-Energy, Editor. 1999,
Argonne National Laboratory: Chicago.

Yang, B. and CE. Wyman, Effect of xylan and lignin removal by batch and
flowthrough pretreatment on the enzymatic digestibility of corn stover cellulose.
Biotechnology and Bioengineering, 2004. 86(1): p. 88-95.

Gilarranz, M.A., et al., Acid precipitation and purification of wheat straw lignin.
Separation Science and Technology, 1998. 33(9): p. 1359-1377.

Bals, B., et al., Ammonia fiber expansion treatment of eleven different forages:
Improvements to fiber digestibility in vitro. Animal Feed Science and Technology,
2009. [Manuscript Submitted].

Carolan, J.E., S.V. Joshi, and B.E. Dale, Technical and Financial Feasibility Analysis
of Distributed Bi0processing Using Regional Biomass Pre-Processing Centers.

Journal of Agricultural and Industrial Organization, 2007. 5: p. 1-27.

Ingram, L.O., et al., Metabolic engineering of bacteria for ethanol production.
Biotechnology and Bioengineering, 1998. 58(2-3): p. 204-214.

161

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

Jeffries, T.W. and Y.S. Jin, Metabolic engineering for improved fermentation of
pentoses by yeasts. Applied Microbiology and Biotechnology, 2004. 63(5): p. 495-
509.

Karhumaa, K., et al., Comparison of the xylose reductase-xylitol dehydrogenase
and the xylose isomerase pathways for xylose fermentation by recombinant
Saccharomyces cerevisiae. Microbial Cell Factories, 2007. 6: p. -.

Yomano, L.P., S.W. York, and LO. Ingram, Isolation and characterization of
ethanol-tolerant mutants of Escherichia coli K011 for fuel ethanol production.
Journal of Industrial Microbiology & Biotechnology, 1998. 20(2): p. 132-138.

Saha, B.C., et al., Dilute acid pretreatment, enzymatic saccharification and
fermentation of wheat straw to ethanol. Process Biochemistry, 2005. 40(12): p.
3693-3700.

Ohgren, K., et al., Simultaneous saccharification and co-fermentation of glucose
and xylose in steam-pretreated corn stover at high fiber content with
Saccharomyces cerevisiae TMB3400. J Biotechnol, 2006. 126(4): p. 488-98.

Moniruzzaman, M. and LO. Ingram, Ethanol production from dilute acid
hydrolysate of rice hulls using genetically engineered Escherichia coli.
Biotechnology Letters, 1998. 20(10): p. 943-947.

Deanda, K., et al., Development of an arabinose-fermenting Zymomonas mobilis
strain by metabolic pathway engineering. Applied and Environmental
Microbiology, 1996. 62(12): p. 4465-70.

Ho, N.W., et al., Successful design and development of genetically engineered
Saccharomyces yeasts for effective cofermentation of glucose and xylose from
cellulosic biomass to fuel ethanol. Adv Biochem Eng Biotechnol, 1999. 65: p. 163-
92.

Underwood, S.A., et al., Flux through citrate synthase limits the growth of
ethanologenic Escherichia coli K011 during xylose fermentation. Applied and
Environmental Microbiology, 2002. 68(3): p. 1071-81.

Sedlak, M. and NW. Ho, Characterization of the effectiveness of hexose
transporters for transporting xylose during glucose and xylose co-fermentation

by a recombinant Saccharomyces yeast. Yeast, 2004. 21(8): p. 671-84.

Pauly, M. and K. Keegstra, Cell-wall carbohydrates and their modification as a
resource for biofuels. Plant J, 2008. 54(4): p. 559-68.

162

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

Lohmeier-Vogel, E.M., C.R. Sopher, and H. Lee, Intracellular acidification as a
mechanism for the inhibition by acid hydrolysis-derived inhibitors of xylose
fermentation by yeasts. Journal of Industrial Microbiology & Biotechnology,
1998. 20(2): p. 75-81.

Chen, J.H. and CT. Ho, Antioxidant activities of caffeic acid and its related
hydroxycinnamic acid compounds. Journal of Agricultural and Food Chemistry,
1997. 45(7): p. 2374-2378.

Chundawat, S.P.S., B. Venkatesh, and B.E. Dale, Effect of particle size based
separation of milled corn stover on AFEX pretreatment and enzymatic
digestibility. Biotechnology and Bioengineering, 2007. 96(2): p. 219-231.

Solaiman, S.G., G.W. Horn, and EN. Owens, Ammonium Hydroxide Treatment on
Wheat Straw. J. Anim Sci., 1979. 49(3): p. 802-808.

Kadam, K.L. and MM. Newman, Development of a low-cost fermentation
medium for ethanol production from biomass. Applied Microbiology and
Biotechnology, 1997. 47(6): p. 625-629.

Ohgren, K., et al., A comparison between simultaneous saccharification and
fermentation and separate hydrolysis and fermentation using steam-pretreated
corn stover. Process Biochemistry, 2007. 42(5): p. 834-839.

Sonderegger, M., et al., Fermentation performance of engineered and evolved
xylose-fermenting Saccharomyces cerevisiae strains. Biotechnol Bioeng, 2004.
87(1): p. 90-8.

Pampulha, ME. and MC Loureirodias, Combined Effect of Acetic-Acid, Ph and
Ethanol on Intracellular Ph of Fermenting Yeast. Applied Microbiology and
Biotechnology, 1989. 31(5-6): p. 547-550.

Hamelinck, C.N., G. van Hooijdonk, and A.P.C. Faaij, Ethanolfrom lignocellulosic
biomass: techno-economic performance in short-, middle- and long-term.
Biomass & Bioenergy, 2005. 28(4): p. 384-410.

Bergman, T., M. Carlquist, and H. Jornvall, Amino Acid Analysis by High
Performance Liquid Chromatography of Phenylthiocarbamyl Derivatives in
Advanced Methods in Protein Microsequence Analysis, 8. Wittmann-Liebold,
Editor. 1986, Springer Verlag: Berlin. p. 45-55.

Xu, F., et al., Partition of enzymes between the solvent and insoluble substrate
during the hydrolysis of lignocellulose by cellulases. Journal of Molecular
Catalysis B: Enzymatic, 2008. 51(1-2): p. 42-48.

163

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

Walker, G.M., Yeast physiology and biotechnology. 1998, New York, NY: John
Wiley & Sons, Inc. .

Saltokoglu, A. and J.C. Slaughter, The effect of magnesium and calcium on yeast
growth. Journal of the Institute of Brewing, 1983. 89: p. 81-83.

Walker, G.M., Metals in Yeast Fermentation Processes, in Advances in Applied
Microbiology, A.I. Laskin, J.W. Bennet, and GM. Gadd, Editors. 2004, Elsevier
Acadamic Press: New York, NY.

Boulton, C. and D. Quain, Brewing Yeast and Fermentation. 2001, Ames, Iowa:
Iowa State University Press. 638.

Chundawat SPS. , et al., Major cell wall decomposition products produced during
ammonia fiber expansion pretreatment of corn stover. [Manuscript In
Preparation], 2009.

Aden, A., Biochemical Production of Ethanol from Corn Stover: 2007 State of
Technology Model. 2008: Golden, Colorado.

Mach, R.L. and S. Zeilinger, Regulation of gene expression in industrial fungi:
Trichoderma. Applied Microbiology and Biotechnology, 2003. 60(5): p. 515-522.

Kelly, J.M. and M.J. Hynes, Transformation of Aspergillus-Niger by the Amds Gene
of Aspergillus-Nidulans. Embo Journal, 1985. 4(2): p. 475-479.

van den Hondel, C.A., P.J. Punt, and RF. van Gorcom, Production of extracellular
proteins by the filamentous fungus Aspergillus. Antonie Van Leeuwenhoek, 1992.
61(2): p. 153-60.

Lee, Y.E., S.E. Lowe, and JG. Zeikus, Regulation and Characterization of
Xylanolytic Enzymes of Thermoanaerobacterium saccharolyticum BGA-RI. Applied
and Environmental Microbiology, 1993. 59(3): p. 763-771.

Lowe, S.E., M.K. Jain, and JG. Zeikus, Biology, ecology, and biotechnological
applications of anaerobic bacteria adapted to environmental stresses in

temperature, pH, salinity, or substrates. Microbiol Rev, 1993. 57(2): p. 451-509.

Shaw, A.J., et al., Metabolic engineering of a thermophilic bacterium to produce
ethanol at high yield. International Sugar Journal, 2009. 111(1323): p. 164-171.

164

91.

92.

93.

94.

95.

96.

97.

98.

99.

100.

101.

102.

Kwiatkowski, JR, et al., Modeling the process and costs of fuel ethanol
production by the corn dry-grind process. Industrial Crops and Products, 2006.
23(3): p. 288-296.

Schubert, Q, Can biofuels finally take center stage? Nature Biotechnology, 2006.
24(7): p. 777-784.

Spiehs, M.J., M.H. Whitney, and G.C. Shurson, Nutrient databasefor distiller’s
dried grains with solubles produced from new ethanol plants in Minnesota and
South Dakota. Journal of Animal Science, 2002. 80(10): p. 2639-2645.

Ham, G.A., et al., Wet corn distillers by-products compared with dried corn
distillers grains with solubles as a source of protein and energy for ruminants.
Journal of Animal Science, 1994. 72(12): p. 3246-3257.

Mosier, N., et al., Optimization of pH controlled liquid hot water pretreatment of
corn stover. Bioresource Technology, 2005. 96(18): p. 1986-1993.

Kim, T.H., et al., Pretreatment of corn stover by low-liquid ammonia recycle
percolation process. Applied Biochemistry and Biotechnology, 2006. 133(1): p.
41-57.

Panagiotou, G. and L. Olsson, Effect of compounds released during pretreatment
of wheat straw on microbial growth and enzymatic hydrolysis rates.
Biotechnology and Bioengineering, 2006. 96(2): p. 250-258.

Varga, E., et al., High solid simultaneous saccharification and fermentation of wet
oxidized corn stover to ethanol. Biotechnology and Bioengineering, 2004. 88(5):
p. 567-574.

Berson, R.E., et al., Detoxification of actual pretreated corn stover hydrolysate
using activated carbon powder. Applied Biochemistry and Biotechnology, 2005.
124(1-3): p. 923-934.

Petersson, A., et al., A 5-hydroxymethy/fuifural reducing enzyme encoded by the
Saccharomyces cerevisiae ADH6 gene conveys HMF tolerance. Yeast, 2006. 23(6):
p. 455-464.

Bals, B., B. Dale, and V. Balan, Enzymatic hydrolysis of distiller's dry grain and
solubles (DDGS) using ammonia fiber expansion pretreatment. Energy & Fuels,
2006. 20(6): p. 2732-2736.

Teymouri, F., et al., Ammonia fiber explosion treatment of corn stover. Applied
Biochemistry and Biotechnology, 2004. 113-16: p. 951-963.

165

103.

104.

105.

Stenberg, K., et al., Effect of substrate and cellulose concentration on
simultaneous saccharification and fermentation of steam-pretreated softwood
for ethanol production. Biotechnology and Bioengineering, 2000. 68(2): p. 204-
210.

DeMancilha, |.M. and MN. Karim, Evaluation of ion exchange resins for removal
of inhibitory compounds from corn stover hydrolyzate for xylitol fermentation.
Biotechnology Progress, 2003. 19(6): p. 1837 -1841.

Nichols, N.N., et al., Bioabatement to remove inhibitors from biomass-derived
sugar hydrolysates. Applied Biochemistry and Biotechnology, 2005. 121: p. 379-
390.

166

REFERENCE

Mosier, N., et al., Features of promising technologies for pretreatment of
lignocellulosic biomass. Bioresource Technology, 2005. 96(6): p. 673-686.

Lynd, L.R., Overview and evaluation of fuel ethanol from cellulosic biomass:
Technology, economics, the environment, and policy. Annual Review of Energy
and the Environment, 1996. 21: p. 403-465.

Lau, M.W., B.E. Dale, and V. Balan, Ethanolic fermentation of hydrolysates from
ammonia fiber expansion (AFEX) treated corn stover and distillers grain without
detoxification and external nutrient supplementation. Biotechnology and
Bioengineering, 2008. 99(3): p. 529-539.

Dien, B.S., et al., Chemical composition and response to dilute-acid pretreatment
and enzymatic saccharification of alfalfa, reed canarygrass, and switchgrass.
Biomass & Bioenergy, 2006. 30(10): p. 880-891.

Chang, V., M. Nagwani, and M. Holtzapple, Lime pretreatment of crop residues
bogasse and wheat straw. Applied Biochemistry and Biotechnology, 1998. 74(3):
p. 135-159.

Kabel, M.A., et al., Effect of pretreatment severity on xylan solubility and
enzymatic breakdown of the remaining cellulose from wheat straw. Bioresource
Technology, 2007. 98(10): p. 2034-2042.

Luo, C., D.L. Brink, and H.W. Blanch, Identification of potential fermentation
inhibitors in conversion of hybrid poplar hydrolyzate to ethanol. Biomass and
Bioenergy, 2002. 22(2): p. 125-138.

Kadam, K.L. and JD. McMillan, Availability of corn stover as a sustainable

feedstock for bioethanol production. Bioresource Technology, 2003. 88(1): p. 17-
25.

157

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

Greene, N., et al., Growing energy: how biofuels can help end America’s oil
dependence. Nature Resources Defense Council, 2004.

Wyman, C.E., Potential synergies and challenges in refining cellulosic biomass to
fuels, chemicals, and power. Biotechnol Prog, 2003. 19(2): p. 254-62.

Hsu, T.-A., Pretreatment of Biomass, in Handbook on Bioethanol, Production and
Utilization, C.E. Wyman, Editor. 1996, Taylor & Francis: Washington, DC.

Wenzl, H.R.J., The Chemical Technology of Wood. 1970, Academic Press. p. 32-
245.

Mosier, N., et al., Features of promising technologies for pretreatment of
lignocellulosic biomass. Bioresource Technology, 2005. 96(6): p. 673-86.

Wyman, C.E., et al., Coordinated development of leading biomass pretreatment
technologies. Bioresource Technology, 2005. 96(18): p. 1959-1966.

Schell, D.J., et al., Dilute-sulfuric acid pretreatment of corn stover in pilot-scale
reactor - Investigation of yields, kinetics, and enzymatic digestibilities of solids.
Applied Biochemistry and Biotechnology, 2003. 105: p. 69-85.

Zaldivar, J., A. Martinez, and LO. Ingram, Effect of selected aldehydes on the
growth and fermentation of ethanologenic Escherichia coli. Biotechnology and
Bioengineering, 1999. 65(1): p. 24-33.

Palmqvist, E. and B. Hahn-Hagerdal, Fermentation of lignocellulosic hydrolysates.
I: inhibition and detoxification. Bioresource Technology, 2000. 74(1): p. 17-24.

Palmqvist, E. and B. Hahn-Hagerdal, Fermentation of lignocellulosic hydrolysates.
II: inhibitors and mechanisms of inhibition. Bioresource Technology, 2000. 74(1):
p. 25-33.

Zaldivar, J. and LO. Ingram, Effect of organic acids on the growth and
fermentation of ethanologenic Escherichia coli LY01. Biotechnology and
Bioengineering, 1999. 66(4): p. 203-210.

Martinez, A., et al., Effects of Ca(OH)(2) treatments ("overliming") on the
composition and toxicity of bagasse hemicellulose hydrolysates. Biotechnology
and Bioengineering, 2000. 69(5): p. 526-536.

Berson, R.E., et al., Detoxification of actual pretreated corn stover hydrolysate
using activated carbon powder. Appl Biochem Biotechnol, 2005. 121-124: p. 923-
34.

158

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

Teymouri, F., et al., Optimization of the ammonia fiber explosion (AFEX)
treatment parameters for enzymatic hydrolysis of corn stover. Bioresource
Technology, 2005. 96(18): p. 2014-2018.

Takagi, M., et al. A method for production of alcohol direct from cellulose using
cellulose and yeast. in Proceedings of Bioconversion of cellulosic substances into
energy, chemicals and microbial protein. 1977. IIT, New Delhi.

Xiao, Z.Z., et al., Effects of sugar inhibition on cellulases and beta-glucosidase
during enzymatic hydrolysis of softwood substrates. Applied Biochemistry and
Biotechnology, 2004. 113-16: p. 1115-1126.

Saha, B.C., Hemicellulose bioconversion. Journal of Industrial Microbiology and
Biotechnology, 2003. 30(5): p. 279-91.

Lynd, L.R., et al., Consolidated bioprocessing of cellulosic biomass: an update.
Curr Opin Biotechnol, 2005. 16(5): p. 577—83.

Dien, B.S., M.A. Cotta, and T.W. Jeffries, Bacteria engineered for fuel ethanol
production: current status. Applied Microbiology and Biotechnology, 2003. 63(3):
p. 258-266.

Ohta, K., et al., Genetic improvement of Escherichia coli for ethanol production -
chromosomal integration of Zymomonas mobilis genes encoding pyruvate
decarboxylase and alcohol dehydrogenase II. Applied and Environmental
Microbiology, 1991. 57(4): p. 893-900.

Sedlak, M. and N.W.Y. Ho, Production of ethanol from cellulosic biomass

hydrolysates using genetically engineered Saccharomyces yeast capable of
cofermenting glucose and xylose. Applied Biochemistry and Biotechnology, 2004.
113-16: p. 403-416.

Mohagheghi, A., et al., Cofermentation of glucose, xylose, and arabinose by
genomic DNA-integrated xylose/arabinose fermenting strain of Zymomonas
mobilis AX101. Appl Biochem Biotechnol, 2002. 98-100: p. 885-98.

Zhang, M., et al., Metabolic Engineering of a Pentose Metabolism Pathway in
Ethanologenic Zymomonas mobilis. Science, 1995. 267(5195): p. 240-243.

Mohagheghi, A., et al., Performance of a newly developed integrant of

Zymomonas mobilis for ethanol production on corn stover hydrolysate.
Biotechnology Letters, 2004. 26(4): p. 321-325.

159

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

Aden, A., et al., Lignocellulosic Biomass to Ethanol Process Design and Economics
Utilizing Co-Current Dilute Acid Prehydrolysis and Enzymatic Hydrolysis for Corn
Stover, N.R.E. Lab, Editor. 2002: Golden, CO.

Lee, Y.Y., P. Iyer, and R.W. Torget, Dilute-Acid Hydrolysis of Lignocellulosic
Biomass, in Advances in Biochemical Engineering, T. Scheper, Editor. 1999,
Springer-Verlag: New York, NY.

Klinke, H.B., A.B. Thomsen, and BK. Ahring, Inhibition of ethanol-producing yeast
and bacteria by degradation products produced during pre-treatment of
biomass. Applied Microbiology and Biotechnology, 2004. 66(1): p. 10-26.

Nilvebrant, N.O., et al., Detoxification of lignocellulose hydrolysates with ion-
exchange resins. Applied Biochemistry and Biotechnology, 2001. 91-3: p. 35-49.

Berson, R.E., et al., Detoxrfication of Actual Pretreated Corn Stover Hydrolysate
Using Activated Carbon Powder, in Twenty-Sixth Symposium on Biotechnology
for Fuels and Chemicals. 2005. p. 923-934.

VonSivers, M., et al., Cost analysis of ethanol production from willow using
recombinant Escherichia coli. Biotechnology Progress, 1994. 10(5): p. 555-560.

Lawford, H.G. and JD. Rousseau, Corn steep liquor as a cost-effective nutrition
adjunct in high-performance Zymomonas ethanol fermentations. Applied
Biochemistry and Biotechnology, 1997. 63-5: p. 287-304.

Humbird, D. and A. Aden, Biochemical Production of Ethanol from Corn Stover:
2008 State of Technology Model. 2009: Golden, Colorado.

Yang, 8. and CE. Wyman, Pretreatment: the key to unlocking low-cost cellulosic
ethanol. Biofuels, Bioproducts and Biorefining, 2008. 2(1): p. 26-40.

Wyman, C.E., et al., Comparative sugar recovery data from laboratory scale
application of leading pretreatment technologies to corn stover. Bioresource
Technology, 2005. 96(18): p. 2026-2032.

Lloyd, T.A. and CE. Wyman, Combined sugar yields for dilute sulfuric acid
pretreatment of corn stoverfollowed by enzymatic hydrolysis of the remaining
solids. Bioresource Technology, 2005. 96(18): p. 1967-1977.

Dien, B.S., et al., Enzyme characterization for hydrolysis of AFEX and liquid hot-

water pretreated distillers’ grains and their conversion to ethanol. Bioresource
Technology, 2008. 99(12): p. 5216-5225.

160

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

Pordesimo, L.O., et al., Variation in corn stover composition and energy content
with crop maturity. Biomass and Bioenergy, 2005. 28(4): p. 366-374.

Lynd, L.R., et al., How biotech can transform biofuels. Nature Biotechnology,
2008. 26(2): p. 169-72.

Lau, MW. and B.E. Dale, Cellulosic ethanol production from AFEX-treated corn
stover using Saccharomyces cerevisiae 424A(LNH-ST). Proc Natl Acad Sci U S A,
2009. 106(5): p. 1368-73.

Sticklen, M.B., Plant genetic engineering for biofuel production: towards
affordable cellulosic ethanol. Nat Rev Genet, 2008. 9(6): p. 433-43.

Shaw, A.J., et al., 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, 2008. 105(37): p. 13769-13774.

Wang, M., H. Lee, and J. Molburg, Allocation of energy use in petroleum
refineries to petroleum products - Implications for life-cycle energy use and
emission inventory of petroleum transportation fuels. International Journal of
Life Cycle Assessment, 2004. 9(1): p. 34-44.

Wang, M., C. Saricks, and D. Santini, Effects of fuel ethanol use on fuel-cycle
energy and greenhouse gas emissions, U.S.-Department-of-Energy, Editor. 1999,
Argonne National Laboratory: Chicago.

Yang, 8. and CE. Wyman, Effect of xylan and lignin removal by batch and
flowthrough pretreatment on the enzymatic digestibility of corn stover cellulose.
Biotechnology and Bioengineering, 2004. 86(1): p. 88-95.

Gilarranz, M.A., et al., Acid precipitation and purification of wheat straw lignin.
Separation Science and Technology, 1998. 33(9): p. 1359-1377.

Bals, B., et al., Ammonia fiber expansion treatment of eleven different forages:
Improvements to fiber digestibility in vitro. Animal Feed Science and Technology,
2009. [Manuscript Submitted].

Carolan, J.E., S.V. Joshi, and B.E. Dale, Technical and Financial Feasibility Analysis
of Distributed Bioprocessing Using Regional Biomass Pre-Processing Centers.

Journal of Agricultural and Industrial Organization, 2007. 5: p. 1-27.

Ingram, L.O., et al., Metabolic engineering of bacteria for ethanol production.
Biotechnology and Bioengineering, 1998. 58(2-3): p. 204-214.

161

57.

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

Jeffries, T.W. and Y.S. Jin, Metabolic engineering for improved fermentation of
pentoses by yeasts. Applied Microbiology and Biotechnology, 2004. 63(5): p. 495-
509.

Karhumaa, K., et al., Comparison of the xylose reductase-xylitol dehydrogenase
and the xylose isomerase pathways for xylose fermentation by recombinant
Saccharomyces cerevisiae. Microbial Cell Factories, 2007. 6: p. -.

Yomano, L.P., S.W. York, and LO. Ingram, Isolation and characterization of
ethanol-tolerant mutants of Escherichia coli K 011 for fuel ethanol production.
Journal of Industrial Microbiology & Biotechnology, 1998. 20(2): p. 132-138.

Saha, B.C., et al., Dilute acid pretreatment, enzymatic saccharification and
fermentation of wheat straw to ethanol. Process Biochemistry, 2005. 40(12): p.
3693-3700.

Ohgren, K., et al., Simultaneous saccharification and co-fermentation of glucose
and xylose in steam-pretreated corn stover at high fiber content with
Saccharomyces cerevisiae TM83400. J Biotechnol, 2006. 126(4): p. 488-98.

Moniruzzaman, M. and LO. Ingram, Ethanol production from dilute acid
hydrolysate of rice hulls using genetically engineered Escherichia coli.
Biotechnology Letters, 1998. 20(10): p. 943—947.

Deanda, K., et al., Development of an arabinose-fermenting Zymomonas mobilis
strain by metabolic pathway engineering. Applied and Environmental
Microbiology, 1996. 62(12): p. 4465-70.

Ho, N.W., et al., Successful design and development of genetically engineered
Saccharomyces yeasts for effective cofermentation of glucose and xylose from
cellulosic biomass to fuel ethanol. Adv Biochem Eng Biotechnol, 1999. 65: p. 163-
92.

Underwood, S.A., et al., Flux through citrate synthase limits the growth of
ethanologenic Escherichia coli K011 during xylose fermentation. Applied and
Environmental Microbiology, 2002. 68(3): p. 1071-81.

Sedlak, M. and NW. Ho, Characterization of the effectiveness of hexose
transporters for transporting xylose during glucose and xylose co-fermentation

by a recombinant Saccharomyces yeast. Yeast, 2004. 21(8): p. 671-84.

Pauly, M. and K. Keegstra, Cell-wall carbohydrates and their modification as a
resource for biofuels. Plant J, 2008. 54(4): p. 559-68.

162

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

Lohmeier-Vogel, E.M., C.R. Sopher, and H. Lee, Intracellular acidification as a
mechanism for the inhibition by acid hydrolysis-derived inhibitors of xylose
fermentation by yeasts. Journal of Industrial Microbiology & Biotechnology,
1998. 20(2): p. 75-81.

Chen, J.H. and CT. Ho, Antioxidant activities of caffeic acid and its related
hydroxycinnamic acid compounds. Journal of Agricultural and Food Chemistry,
1997. 45(7): p. 2374-2378.

Chundawat, S.P.S., B. Venkatesh, and B.E. Dale, Effect of particle size based
separation of milled corn stover on AFEX pretreatment and enzymatic
digestibility. Biotechnology and Bioengineering, 2007. 96(2): p. 219-231.

Solaiman, S.G., G.W. Horn, and F.N. Owens, Ammonium Hydroxide Treatment on
Wheat Straw. J. Anim Sci., 1979. 49(3): p. 802-808.

Kadam, K.L. and MM. Newman, Development of a low-cost fermentation
medium for ethanol production from biomass. Applied Microbiology and
Biotechnology, 1997. 47(6): p. 625-629.

Ohgren, K., et al., A comparison between simultaneous saccharification and
fermentation and separate hydrolysis and fermentation using steam-pretreated
corn stover. Process Biochemistry, 2007. 42(5): p. 834-839.

Sonderegger, M., et al., Fermentation performance of engineered and evolved
xylose-fermenting Saccharomyces cerevisiae strains. Biotechnol Bioeng, 2004.
87(1): p. 90-8.

Pampulha, ME. and MC. Loureirodias, Combined Effect of Acetic-Acid, Ph and
Ethanol on Intracellular Ph of Fermenting Yeast. Applied Microbiology and
Biotechnology, 1989. 31(5-6): p. 547-550.

Hamelinck, C.N., G. van Hooijdonk, and A.P.C. Faaij, Ethanol from lignocellulosic
biomass: techno-economic performance in short-, middle- and long-term.
Biomass & Bioenergy, 2005. 28(4): p. 384-410.

Bergman, T., M. Carlquist, and H. Jornvall, Amino Acid Analysis by High
Performance Liquid Chromatography of Phenylthiocarbamyl Derivatives in
Advanced Methods in Protein Microsequence Analysis, 8. Wittmann-Liebold,
Editor. 1986, Springer Verlag: Berlin. p. 45-55.

Xu, F., et al., Partition of enzymes between the solvent and insoluble substrate
during the hydrolysis of lignocellulose by cellulases. Journal of Molecular
Catalysis B: Enzymatic, 2008. 51(1-2): p. 42-48.

163

79.

80.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

Walker, G.M., Yeast physiology and biotechnology. 1998, New York, NY: John
Wiley & Sons, Inc. .

Saltokoglu, A. and J.C. Slaughter, The effect of magnesium and calcium on yeast
growth. Journal of the Institute of Brewing, 1983. 89: p. 81-83.

Walker, G.M., Metals in Yeast Fermentation Processes, in Advances in Applied
Microbiology, A.I. Laskin, J.W. Bennet, and GM. Gadd, Editors. 2004, Elsevier
Acadamic Press: New York, NY.

Boulton, C. and D. Quain, Brewing Yeast and Fermentation. 2001, Ames, Iowa:
Iowa State University Press. 638.

Chundawat SPS. , et al., Major cell wall decomposition products produced during
ammonia fiber expansion pretreatment of corn stover. [Manuscript In
Preparation], 2009.

Aden, A., Biochemical Production of Ethanol from Corn Stover: 2007 State of
Technology Model. 2008: Golden, Colorado.

Mach, R.L. and S. Zeilinger, Regulation of gene expression in industrial fungi:
Trichoderma. Applied Microbiology and Biotechnology, 2003. 60(5): p. 515-522.

Kelly, J.M. and M.J. Hynes, Transformation of Aspergillus-Niger by the Amds Gene
of Aspergillus-Nidulans. Embo Journal, 1985. 4(2): p. 475-479.

van den Hondel, C.A., P.J. Punt, and RF. van Gorcom, Production of extracellular
proteins by the filamentous fungus Aspergillus. Antonie Van Leeuwenhoek, 1992.
61(2): p. 153-60.

Lee, Y.E., S.E. Lowe, and JG. Zeikus, Regulation and Characterization of
Xylanolytic Enzymes of Thermoanaerobacterium saccharolyticum BEA-RI. Applied
and Environmental Microbiology, 1993. 59(3): p. 763-771.

Lowe, S.E., M.K. Jain, and JG. Zeikus, Biology, ecology, and biotechnological
applications of anaerobic bacteria adapted to environmental stresses in

temperature, pH, salinity, or substrates. Microbiol Rev, 1993. 57(2): p. 451-509.

Shaw, A.J., et al., Metabolic engineering of a thermophilic bacterium to produce
ethanol at high yield. International Sugar Journal, 2009. 111(1323): p. 164-171.

164

91.

92.

93.

94.

95.

96.

97.

98.

99.

100.

101.

102.

Kwiatkowski, JR, et al., Modeling the process and costs of fuel ethanol
production by the corn dry-grind process. Industrial Crops and Products, 2006.
23(3): p. 288-296.

Schubert, C., Can biofuels finally take center stage? Nature Biotechnology, 2006.
24(7): p. 777-784.

Spiehs, M.J., M.H. Whitney, and G.C. Shurson, Nutrient database for distiller’s
dried grains with solubles produced from new ethanol plants in Minnesota and
South Dakota. Journal of Animal Science, 2002. 80(10): p. 2639-2645.

Ham, G.A., et al., Wet corn distillers by-products compared with dried corn
distillers grains with solubles as a source of protein and energy for ruminants.
Journal of Animal Science, 1994. 72(12): p. 3246-3257.

Mosier, N., et al., Optimization of pH controlled liquid hot water pretreatment of
corn stover. Bioresource Technology, 2005. 96(18): p. 1986-1993.

Kim, T.H., et al., Pretreatment of corn stover by low-liquid ammonia recycle
percolation process. Applied Biochemistry and Biotechnology, 2006. 133(1): p.
41-57.

Panagiotou, G. and L. Olsson, Effect of compounds released during pretreatment
of wheat straw on microbial growth and enzymatic hydrolysis rates.
Biotechnology and Bioengineering, 2006. 96(2): p. 250-258.

Varga, E., et al., High solid simultaneous saccharification and fermentation of wet
oxidized corn stover to ethanol. Biotechnology and Bioengineering, 2004. 88(5):
p. 567-574.

Berson, R.E., et al., Detox1fication of actual pretreated corn stover hydrolysate
using activated carbon powder. Applied Biochemistry and Biotechnology, 2005.
124(1—3): p. 923-934.

Petersson, A., et al., A S-hydroxymethylfurfural reducing enzyme encoded by the
Saccharomyces cerevisiae ADH6 gene conveys HMF tolerance. Yeast, 2006. 23(6):
p. 455-464.

Bals, B., B. Dale, and V. Balan, Enzymatic hydrolysis of distiller's dry grain and
solubles (DDGS) using ammonia fiber expansion pretreatment. Energy & Fuels,
2006. 20(6): p. 2732-2736.

Teymouri, F., et al., Ammonia fiber explosion treatment of corn stover. Applied
Biochemistry and Biotechnology, 2004. 113-16: p. 951-963.

165

103.

104.

105.

Stenberg, K., et al., Effect of substrate and cellulose concentration on
simultaneous saccharification and fermentation of steam-pretreated softwood
for ethanol production. Biotechnology and Bioengineering, 2000. 68(2): p. 204-
210.

DeMancilha, I.M. and MN. Karim, Evaluation of ion exchange resins for removal
of inhibitory compounds from corn stover hydrolyzate for xylitol fermentation.
Biotechnology Progress, 2003. 19(6): p. 1837 -1841.

Nichols, N.N., et al., Bioabatement to remove inhibitors from biomass-derived
sugar hydrolysates. Applied Biochemistry and Biotechnology, 2005. 121: p. 379-
390.

166