- Email: [email protected]
Fermentation of wet-exploded corn stover for the production of volatile fatty acids
Accepted Manuscript Fer mentation of wet-exploded cor n stover for the pr oduction of volatile fatty acids Nanditha Murali, Sebastian Fernandez, Birgitte Kiaer Ahring PII: DOI: Reference:
S0960-8524(16)31663-7 http://dx.doi.org/10.1016/j.biortech.2016.12.012 BITE 17388
To appear in:
Bioresource Technology
Received Date: Revised Date: Accepted Date:
21 October 2016 1 December 2016 2 December 2016
Please cite this article as: Murali, N., Fernandez, S., Kiaer Ahring, B., Fer mentation of wet-exploded cor n stover for the pr oduction of volatile fatty acids, Bioresource Technology (2016), doi: http://dx.doi.org/10.1016/ j.biortech.2016.12.012
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
2
Fermentation of wet-exploded corn stover for the production of volatile fatty acids
3
Nanditha Murali, Sebastian Fernandez, and Birgitte Kiaer Ahring*
4
Bioproducts, Sciences and Engineering Laboratory, Washington State University, Tri-Cities, Richland, WA-99354
1
5 6 7 8 9 10 11
Abstract
12
Volatile fatty acids (VFA) have been used as platform molecules for production of biofuels and
13
bioproducts. In the current study, we examine the VFA production from wet-exploded corn
14
stover through anaerobic fermentation using rumen bacteria. The total VFA yield (acetic acid
15
equivalents) was found to increase from 22.8g/L at 2.5% total solids (TS) to 40.8g/L at 5%TS. It
16
was found that the acetic acid concentration increased from 10g/L to 22g/L at 2.5% and 5%TS,
17
respectively. An increased propionic acid production was seen between day 10 and 20 at 5%TS.
*
Corresponding author
Dr. Birgitte K. Ahring Bioproducts Sciences and Engineering Laboratory Washington State University, Tri-cities 2710 Crimson Way Richland, WA 99354 Tel.: 01-(509)-372-7682 Fax: 01-(509)-372-7690. E-mail address: [email protected] 1
18
Valeric acid (4g/L) was produced at 5%TS and not at 2.5%TS. Composition analysis showed
19
that 50% of the carbohydrates were converted to VFA at 5%TS and 33% at 2.5%TS. Our results
20
show that rumen fermentation of lignocellulosic biomass after wet explosion can produce high
21
concentrations of VFA without addition of external enzymes of importance for the process
22
economics of lignocellulosic biorefineries.
23 24
Key Words: Acetic acid, Volatile Fatty Acids, Anaerobic Fermentation, Corn Stover, Rumen.
25 26
1. Introduction
27
Volatile Fatty Acids (VFA), primarily short-chain carboxylic acids such as acetic, propionic and
28
butyric acids can be produced from petrochemical sources and further form biomass feedstocks.
29
VFA has been suggested as valuable platform molecules for the production of biofuels and
30
bioproducts (Chang et al. 2010). Bio-based VFA are now produced from anaerobic fermentation
31
of organic materials such as waste products or agricultural residues (Solomon et al. 2007). Use of
32
lignocellulosic biomass through biochemical conversion to produce biofuels and bio-products
33
from biomass have been shown to be burdened by the need of enzyme addition leading to high
34
operational cost (Giampietro et al. 1997).
35 36
Volatile fatty acids are the main intermediates in the production of methane from organic
37
materials (Ahring et al. 1995) and have been used as a control parameter to assess effectiveness
38
of anaerobic digestion (Pind et al. 2003). Under high loading of a bioreactor, a higher VFA
39
production will occur and methane production can be reduced as VFA production usually occurs
40
fast within 1 – 3 days, as opposed to 7– 15 days for the production of methane, leading to an 2
41
initial drop in pH (Pham et al. 2012). Studies have shown that these VFA can further be
42
harvested and used for production of biofuels such as ethanol, propanol or butanol by catalysis
43
(Holtzapple and Granada, 2009).
44 45
The market for acetic acid is growing at ca. 5% a year and estimated to 16155 metric ton by
46
2020. Acetic acid is used as building block for production of various chemicals, such as vinyl
47
acetate monomer (main use), purified terephthalic acid, acetate esters, and acetic anhydride of
48
importance for the synthetic fiber and textile industry as well as for ink and pesticides (Mordor
49
intelligence Inc, 2016). Also other acids such as propionic and butyric has growing markets for
50
many applications (Grandview Research Inc, 2015). Producing sustainable VFA from biomass
51
materials can, therefore, be of major importance in the future.
52 53
A wide variety of biomass feedstocks, predominantly agricultural residues, have been used for
54
the production for VFA (Kadam and McMillan, 2003). In the United States, corn is the most
55
widely grown crop (Kadam and McMillan, 2003) and it is estimated that approximately 153
56
million dry tons of corn stover is available as biomass feedstock every year (Walsh et al. 2000).
57
The main advantages of using lignocellulosic biomass for VFA production, is that the biomass
58
has a lower cost, is generally abundantly available and is not in competition with food,
59
renewable, and reduces net CO2 emissions (Berndes et al. 2001). In raw biomass, cellulose fibers
60
(about 35-45% of the biomass) are encased in a hemicellulose (20-25%) and lignin (15-20%)
61
matrix via carbon – carbon linkages (Himmel et al. 2007; Agbor et al. 2011). This matrix has to
62
be broken down for making major parts of the cellulose accessible for fermentation. The native
63
form of biomass is recalcitrant to enzymatic hydrolysis with cellulase enzymes and requires 3
64
pretreatment to remove the hemicellulose-lignin linkages holding the biomass together (Lynd et
65
al. 2002). Conventional pretreatment methods with acids, or alkali or heat (Lynd et al. 2008)
66
introduces harsh chemicals, which could be detrimental for microbial growth required for
67
fermentation. Wet Explosion (WEx) pretreatment, on the other hand, uses oxygen and water at
68
elevated temperature and pressure and the process is terminated by ‘exploding’ the biomass into
69
a flash tank (Biswas et al. 2015). In addition to not using any harsh chemicals, which is both
70
detrimental to microbes and further will be in the fermentation broth., WEx also produces lower
71
amounts of inhibitory products like furfural and hydroxymethylfurfural (Njoku et al. 2013). WEx
72
also opens up the crystalline structure of cellulose (Rana et al. 2012), thus making cellulose more
73
accessible to microbial enzymes produced during fermentation (Varga et al. 2003; Ahring and
74
Munck, 2009).
75 76
Zeachem has suggested to use Mororella thermoaceticum for producing acetic acid as platform
77
molecule for bio-products and biofuel production using cellulosic sugars as substrate (Verser and
78
Eggeman, 2005). However, this fermentation needs sterility to operate along with the need of
79
costly nutrient additions (Verser and Eggeman, 2005). When compared to such monocultures,
80
robust mixed cultures can result in improved process economics since it does not require
81
sterilization and is more resistant to inhibitory chemicals present in pretreated biomass (such as
82
lignin compounds) (Agler et al., 2011). Ruminants like cattle are known to produce VFA via
83
anaerobic fermentation, with acetic, propionic and butyric acids being the predominant products
84
(Dijkstra, 1994). Previous studies have shown that stable mixed microbial culture such as rumen
85
have the ability to produce necessary enzymes for degradation of carbohydrate polymers into
86
monomers as well as for VFA production from both C5 and C6 sugars (Hess et al. 2011; Boaro 4
87
et al. 2014). However, few studies have successfully tested rumen fermentation of lignocellulosic
88
biomass like corn stover into VFA. The primary goal of the current study is to optimize VFA
89
production in bioreactors operated with pretreated lignocellulosic biomass (corn stover) at
90
different TS concentrations (2.5 and 5%) using inoculum from the rumen grown under
91
mesophilic conditions.
92 93
Pretreated corn stover is used as carbon source in our experiments and extra nutrients are
94
provided in the form of corn steep liquor (CSL) (Amartey and Jeffries, 1994). CSL is a
95
byproduct of the corn milling process where shelled corn is first steeped to form corn steep
96
water, which is then evaporated to form corn steep liquor. On average, CSL contains 5% lactic
97
acid and 0.1% VFA (Kerr and berlin, 1933) and it also contains amino acids (Hull et al. 1996),
98
which could aid in the growth of the microbes. In our study, the concentration of CSL is initially
99
evaluated to assess its effectiveness for growing our bacterial consortia on pretreated corn stover
100
with CSL as sole further nutrient.
101 102 103
2. Materials and methods:
104
2.1. Inoculum: Rumen fluid was collected at a mobile slaughter facility in Richland, WA. The
105
animal weighed 1200 pounds and was fed 15 pounds of grain (molasses, corn wheat and barley)
106
and alfalfa per day. The samples were collected in bottles, which were filled to capacity, to
107
minimize oxygen in the bottle and transported to the laboratory on ice. Upon arrival in the lab,
108
the samples were transferred in a sterile bench to sterile bottles, degassed under 80% N2: 20%
109
CO2 for 30 minutes, sealed and stored at –20°C until further use. The nitrogen carbon dioxide 5
110
gas mix was obtained from Oxarc® Inc. (Pasco, WA, USA). As inoculum, 5% rumen fluid was
111
added to each bioreactor.
112 113
2.2. Substrate: Corn stover was obtained from Iowa State University, Ames, IA and milled to 2-
114
mm size particles using Retsch SM 200 cutting mill (Retsch® Inc. PA, USA). The corn stover
115
was pretreated using Wet Explosion Pretreatment (WEx) method at 10% solids loading as
116
previously described (Biswas et al. 2014) at 190°C for 30 min with oxygen loading of 7.5%.
117
These conditions were selected for maximum accessibility of the cellulose to cellulolytic
118
enzymes produced during rumen fermentation (data not shown).
119 120
2.2.1. Chemicals: 5N Sodium hydroxide (Sigma Aldrich, Saint louis, MO, USA) was used to
121
maintain the pH of the fermenters and 6.5. 4mM sulfuric acid (Sigma Aldrich, Saint louis, MO,
122
USA) was used to both prepare the HPLC samples and as an eluent for HPLC. Corn steep liquor
123
(2% concentration) was from Sigma Aldrich, Saint louis, MO, USA.
124 125 126
2.3. Fermentation:
127
2.3.2. Assessing the effect of corn steep liquor in batch VFA fermentation – Initial batch
128
fermentations were done in 50mL serum vials to assess the effect of corn steep liquor at different
129
solids loading (2.5% and 5% TS) using rumen inoculum. In two sets of batch experiments, wet-
130
exploded corn stover supplemented with 2% (w/w) corn steep liquor (Lawford and Rousseau,
131
1997) was used as the substrate. To the other two sets of batch experiments, only corn stover
6
132
without the corn steep liquor was used as substrate. 5% rumen fluid was used as inoculum for all
133
the fermentations.
134 135
2.3.2. Effect of substrate concentration on VFA fermentation – Two reactors (3L Applikon®
136
ezControl Autoclavable bioreactor (Applikon Biotechnology B.V, Netherlands) were set up with
137
pretreated corn stover – one with 2.5% TS and one with 5% TS (Fig. 1) to compare the effect of
138
substrate concentration on rumen fermentation. The bioreactors were initially sterilized by
139
autoclavation and filled with 800ml of pretreated corn stover. 40ml of rumen inoculum was
140
aseptically added to the bioreactors while the reactors were gassed using 80:20 wt% N2/CO2
141
gaseous mixture. 5M sodium hydroxide solution was used for maintaining pH during anaerobic
142
fermentation. The pH of the bioreactor was initially adjusted to 6.5 using 5M sodium hydroxide
143
solution and temperature was set to 37°C with a stirrer speed at 150 rpm. The fermentation was
144
operated semi-continuously with replacement of 100ml of fresh media every day and an overall
145
hydraulic retention time of 6 days.
146 147 148 149
2.4. Analyses:
150
2.4.1. Measurement of VFA Using HPLC – The liquid samples from the fermentation were
151
centrifuged at 10,000 rpm for 10 minutes and the supernatant and analyzed using high
152
performance liquid chromatography (HPLC) analysis after filtration through a 0.2 micron filter
153
and dilution (3 times) using 4mM sulfuric acid. Analysis were done using an UltiMate® 3000
154
HPLC system (Dionex, Sunnyvale, CA) with an Aminex® 87H Column 250 X 4.6mm (Bio – 7
155
Rad, Hercules, CA) and a Shodex RI – 101 refractive index detector. Sulfuric acid (4mM) in
156
water was used as the eluent, flowing through the 87H column at a constant flow rate of
157
0.6mL/min in a constant temperature oven at 60°C. The total time for analysis of the
158
fermentation sample was 68 minutes.
159 160
2.4.2. Gas Analyses using Gas Analyzer – Random head space gas measurements were done
161
using the Universal Gas Analyzer, UGA Series (Stanford Research Systems, Sunnyvale, CA).
162 163
2.4.2. Calculations: The major acids produced in the fermentation are acetic, propionic and
164
butyric acids. These were converted to acetic acid equivalents to report total VFA. The
165
calculations are as follows:
166
= + ∗ ∗ ! . " # + $%& ∗ ∗ "". #
.
!
.
167
Where, molar mass of acetic acid = 60.05; molar mass of propionic acid = 74.08 and molar mass
168
of butyric acid = 88.11. The molar mass of the acids is then multiplied by the number of grams
169
of oxygen required to completely breakdown the acids to carbon dioxide to yield total VFA in
170
acetic acid equivalents.
171 172
2.4.2. Feedstock characterization – Total solids (TS) and volatile solids (VS) were measured
173
using the method described by Sluiter et al. (2008). Crucibles containing liquid effluent was
174
placed in a drying oven at 105 ± 3°C overnight for 20 hours. The dried samples were allowed to
175
cool to room temperature before weighing them. The samples were placed back in the oven and
176
dried to constant weight. After weighing for total solids (TS), ash content was measured, based
177
on the method described by Wychen and Laurens (Wychen and Laurens, 2013), by heating the 8
178
crucible at 575 ± 25°C in a muffle furnace for 6 hours. The burnt samples were cooled to room
179
temperature before being weighed.
180 181
2.4.3. Composition analysis – For the determination of total carbohydrates and lignin, the
182
fermentation feed and effluent were air dried at room temperature thoroughly and the solid
183
fraction was used to analyze total carbohydrates and lignin (both soluble and insoluble) content
184
using the method described by Sluiter et al. (2011). Total carbohydrates (cellulose and
185
hemicellulose) and insoluble lignin was measured after hydrolysis, while soluble lignin was
186
analyzed using UV spectrophotometer (Jenway 6405 UV/Visible NJ, USA) at 320nm (Sluiter et
187
al. 2011). All measurements were done in triplicates.
188 189
3. Results and Discussion
190
3.1. Effect of corn steep liquor on batch rumen fermentation of wet exploded corn stover
191
The total VFA production and individual VFA’s obtained from batch anaerobic rumen
192
fermentation of wet exploded corn stover with and without corn steep liquor was assessed (data
193
not shown). It was seen that the batch experiments with corn steep liquor produced almost 30%
194
more VFA than the experiments without CSL while only insignificant concentrations of VFA
195
was present in 2% CSL (Liggett and Koffler, 1948). We also found that lactic acid present in
196
the pretreated corn stover supplemented with corn steep liquor was only 0.24g/L and hence
197
would not significantly impact VFA yields. Since these preliminary experiments were done
198
in serum vials in batch mode, continuous monitoring of pH was not possible. In the absence of
199
corn steep liquor, there was no significant difference between total VFA production at 2.5% and
200
5% TS. However, through addition of 2% CSL, the total VFA yields at 5% TS was found to give 9
201
almost 1.4 times higher than that at 2.5% TS. This could indicate that additional nutrient is
202
needed to maximize VFA production at increased solids loading.
203 204
3.2. Semi-continuous fermentation - Comparison of total VFA yield
205
The VFA concentration expressed as total VFA produced during fermentation of the rumen
206
culture at different solids loadings are shown in Fig. 4. A study by Pham and co-workers,
207
showed that total VFA production of 5 to 7g/L during fermentation of 30g/L microalgae (Pham
208
et al. 2012). However, they found that when the substrate concentration was increased by 50%,
209
the VFA yield increased to between 10 to 15g/L. We further found that the total VFA yield was
210
22.8g/L (or 1.79 g/g VS) with 2.5% TS and 40.8g/L (or 2.11 g/g VS) at 5% solids loading. Our
211
studies found that acetic acid was the major VFA produced followed by propionic acid and
212
butyric acid (Fig. 2). In the reactor with 2.5% TS, acetic acid concentration increased to 13g/L
213
(0.94g/g VS), while propionic acid concentration increased to 10g/L (0.78g/g VS) after 10 days.
214
However, in the reactor with 5% TS, acetic acid concentration increased to 22g/L (1.14 g/g VS),
215
while maximum propionic acid concentration was 10g/L (0.52 g/g VS). We also found that
216
valeric acid concentration increased to 4g/L (0.21g/g VS) in the reactor with 5% TS (Fig 2) and
217
that valeric acid was only produced in this reactor. Other studies using rumen bacteria under
218
thermophilic conditions (60°C) showed far lower concentrations (417mg/L of VFA) from
219
cellulose (Nissila et al. 2011). Studies published using pure cellulose and starch substrates for
220
rumen fermentation showed VFA yields ranging from 0.417g/L to as high as 4.563g/L (Nissila et
221
al. 2011). Another study using modified lime pretreatment for pretreatment of lignocellulosic
222
biomass resulted in a maximum VFA production of 5.32g/L which accounted for around 0.5g
223
VFA/g biomass added (Kim et al. 2013). In comparison with existing literature on production of 10
224
VFA, it can be clearly seen that significantly higher VFA concentrations were obtained in our
225
current study on pretreated biomass material.
226 227
Studies using monocultures such as Moorella thermoacetica have shown acetic yields of 17-
228
18g/L from pretreated and enzymatically hydrolyzed lignocellulosic biomass hydrolysate and the
229
main factor that adversely affected acetic acid yields was the inefficient consumption of
230
arabinose, mannose and galactose (Ehsanipour et al. 2016). Apart from lower VFA yields when
231
compared to our current study, pure cultures also need sterile conditions and further requires
232
sugars for their conversion while our study is done non-sterile and without hydrolyzing the
233
pretreated biomass material. Previous proteomic and metabolomic studies on anaerobic rumen
234
fermentation done in our laboratory indicated increased ß-glucosidase activity and the activity of
235
Clostridia and Fibirobacter clusters during fermentation of avicel (Boaro et al. 2014). This study
236
shows that unlike pure cultures, an anaerobic cellulose fermenting culture established with
237
rumen fluid as inoculum will have both cellulase producing microbes along with acidogens
238
producing VFA.
239 240
3.3. Comparison of VFA production as function of solids loading in reactors
241
As shown in Figure 3, the total VFA production expressed in acetic acid equivalents increased as
242
a function of time for the first 21 days and was initially higher at lower solids loading. After
243
fermentation for around 10 days, higher VFA production was seen with 5% TS while there was
244
no significant difference in the VFA production at 2.5% and 5% TS between 5 and 10 days.
245
After 10 days it seems like the bioreactors were stabilizing (or acclimating) to the wet-exploded
246
corn stover. This can be attributed to the sudden increase in the VFA productivity after the 10th 11
247
day from 0.07 to 0.19 g/L/h at 2.5% TS and 0.19 to 0.77 g/L/h at 5% solids TS. At the end of the
248
experiment (day 62) a stable microbial productivity of 0.24 g/L/h and 0.40 g/L/h was maintained
249
in the bioreactors with 2.5% and 5% TS, respectively.
250 251
3.4 Comparison of biomass composition before and after fermentation
252
Compositional analysis was done on the wet exploded corn stover used as fermenter feed and the
253
effluent obtained at end of the experiments (62 days) and the data is shown in Table 1. It can be
254
seen from Table 1 that the feed had approximately 51wt% carbohydrates and that approximately
255
50% of the carbohydrates was converted to VFA from anaerobic fermentation at 5% TS while
256
only around 33% of the carbohydrates was converted at 2.5% TS. It can be seen from Table 1
257
that, at 2.5% TS that the hemicellulose degradation was considerably lower (16.3wt% in feed
258
reduced to 12.4wt% in effluent) when compared to that at 5% TS (14.6wt% in feed to 5.4wt% in
259
effluent). A similar trend was also seen for cellulose conversion between 2.5% TS (36.8wt% in
260
feed to 22.8wt% in effluent) and 5% TS (36.1wt% in feed to 19.6wt% in effluent) but the
261
difference in values between the two solid loadings was found to be lower for cellulose
262
compared to hemicellulose conversion. This can be attributed to a greater stability of the rumen
263
bacteria towards hemicellulose and cellulosic conversion to volatile fatty acids at higher solids
264
loading. The efficient conversion of hemicellulose along with cellulose by the rumen microflora
265
at 5% TS is significant and shows great potential for further optimization of the anaerobic
266
fermentation to higher loadings to further maximize VFA production.
267 268
Assuming inefficient lignin utilization by the rumen bacteria during anaerobic fermentation,
269
lignin-carbohydrate ratio can be used to quantify effectiveness of carbohydrate utilization 12
270
towards VFA production. It can be seen from Table 1 that the lignin-carbohydrate ratio increased
271
from 0.88 to 2.68 at 5% TS while it increased from 0.82 to 1.75 at 2.5% TS confirming
272
significant biomass sugar utilization by rumen bacteria during anaerobic fermentation to produce
273
VFA. However, in the contrary, higher TS loading (5%) showed a lower carbohydrate-to-VFA
274
conversion of 3.17 g VFA/g carbohydrate when compared to 2.5% TS (5.09 g VFA/g
275
carbohydrate). This trend can also be clearly seen by comparing the acetic acid concentrations,
276
given in Table 1, between feed and effluent. At 2.5% solids loading, 10.06 g/L of acetic acid is
277
produced after consumption of 17.9wt% carbohydrates resulting in 2.25 g acetic acid produced/g
278
carbohydrate consumed while at 5% solids loading, 18.43 g/L of acetic acid is produced after
279
consumption of 25.7wt% carbohydrates resulting in only 1.43 g acetic acid produced/g
280
carbohydrate. While not directly indicated in Table 1, there was negligible amount of C5 or C6
281
sugars in the liquid effluent at both TS. The only difference between the fermentation conditions
282
at both TS is the lignin content which is equal to 1.09 g lignin/L (acid soluble + insoluble lignin
283
given in Table 1) at 2.5% TS and 2.23 g lignin/L at 5% TS and further studies to optimize this
284
process can significantly enhance effective biomass conversion at higher TS loading to obtain
285
higher carbohydrate conversion to selectivity produce VFA. The lower acetic acid production at
286
5% TS loading, could be a result of lignin being converted to methane. Previous studies have
287
reported that wet exploded lignin can be converted to methane (Ahring et al. 2015) and after
288
pretreatment, 44% lignin was converted to methane using anaerobic digestion. The study also
289
showed that pretreated lignin produced 4.5 times higher methane than the non-pretreated lignin
290
(Ahring et al. 2015).
291 292
3.5. Comparison of individual VFA as a function of time and solids loading 13
293
As indicated previously, the major acids produced during rumen fermentation of wet exploded
294
corn stover were acetic, propionic and butyric acids, with trace amounts of formic and valeric
295
acids. The time-based concentration profile at 2.5% and 5% TS for the individual VFA is shown
296
in Fig. 2 (a) & (b), respectively. Liquid fraction analysis showed only trace amounts of C5
297
(xylose) and C6 (glucose) sugars at different sampling times for both 2.5% and 5% TS. This
298
indicates that the sugar oligomers and monomers obtained through wet explosion pretreatment of
299
corn stover was directly be converted to VFA along with the polymers in the material. The
300
culture in the reactors was, therefore, capable of producing the necessary cellulolytic enzymes
301
for getting into the sugar oligomers and polymers.
302 303
It can be seen from Figure 2 that acetic acid was the most predominant VFA produced through
304
anaerobic rumen fermentation of pretreated corn stover. Although issues with the feeding pump
305
caused a decrease in VFA concentration between days 26 and 40 in the reactor fed with 2.5%
306
TS, once the problem was resolved the fermentation stabilized again at the same initial VFA
307
level. This shows that rumen fermentations can be maintained at stable levels for extended
308
periods of time and will have some resilience to perturbations. From Figure 2 it can be seen that
309
increasing amounts of propionic acid was produced at 5% solids loading especially between days
310
10-20. Though studies have indicated that an increase in the propionic acid concentration can
311
slightly affect the rate of anaerobic digestion (AD) and in turn VFA production (Gourdon and
312
Vermande, 1987), it can be seen that the rate of anaerobic fermentation at 5% solids loading was
313
not affected by this propionic acid concentration and overall the VFA concentration was
314
significantly higher compared to the 2.5% TS reactor. Gas analyses shows that increases in the
315
hydrogen concentrations further increased acetic acid production in the rumen fermentation (Fig. 14
316
4). In the reactor with 5% TS slight increases in acetic acid concentration are seen coinciding
317
with increased hydrogen concentration. This might be attributed to activity by homo-acetogens
318
present in the rumen, which might convert the hydrogen together with CO2 to acetic acid (Ni et
319
al. 2011; Wolin et al. 1997) even that the hydrogen concentration generally was maintained at
320
1.5 wt% primarily due to the presence of methanogens in the rumen microflora. The
321
methanogens reduce hydrogen, carbon dioxide and formate to produce methane, which was the
322
major gas produced in both the reactors, as seen in Fig. 4 (Pind et al. 2003). Another significant
323
difference at different solid loadings is the significant amounts of valeric acid produced at 5%
324
TS. This was not found in the samples collected from the 2.5% TS bioreactor. Further studies are
325
currently being done to understand the differences in microbial composition in the two reactors.
326 327
4. Conclusion:
328
In this study, we have established an active and stable culture in two bioreactors using rumen
329
fluid as inoculum. Anaerobic fermentation of wet exploded corn stover at 2.5% and 5% TS
330
produced 22.8g/L and 40.8g/L of total VFA (acetic acid equivalents), respectively. These results
331
show, by far, the highest level of VFA at mesophilic anaerobic fermentation conditions.
332
Compositional analysis showed efficient cellulose conversion and a higher hemicellulose
333
conversion at 5% TS when compared to 2.5%. Further understanding the mechanism of lignin
334
conversion by rumen bacteria after pretreatment can significantly increase VFA yields.
335 336 337 338
References: 1. Agbor, V. B., Cicek, N., Sparling, R., Berlin, A., Levin, D. B., 2011. Biomass pretreatment: Fundamentals towards application. Biotechnol. Adv. 29, 676-685. 15
339
2. Agler, M. T., Wrenn, B. A., Zinder, S. H., Angenent, L. T., 2011. Waste to bioproduct
340
conversion with undefined mixed cultures: The carbohydrate platform. Trends
341
Biotechnol. 70-78.
342
3. Ahring, B. K., Biswas, R., Ahamed, A., Teller, P. J., Uellendahl, H., 2015. Making lignin
343
accessible for anaerobic digestion by wet-explosion pretreatment. Biores. Technol. 175,
344
182-188.
345
4. Ahring, B. K., Munck, J., 2009. Method for treating biomass and organic waste with the
346
purpose of generating desired biologically based products. US Patent No. 0178671 A1.
347
5. Ahring, B. K., Sandberg, M., Angelidaki, I., 1995. Volatile fatty acids as indicators of
348
process imbalance in anaerobic digestors. Appl. Microbiol. Biotechnol. 43, 559-565.
349
6. Amartey, S., Jeffries, T. W., 1994. Comparison of corn steep liquor with other nurients in
350
the fermentation of D-Xylose by Pichia stipitis CBS 6054. Biotechnol. Lett. 6(2), 211-
351
214.
352 353 354
7. Berndes, G., Azar, C., Kaberfer, T., Abrahamson, D., 2001. The feasibility of large-scale lignocellulose based bioenergy production. Biomass Bioenergy. 20(5), 371-383. 8. Biswas, R., Uellendahl, H., Ahring, B. K., 2015. Wet explosion: a universal and efficient
355
pretreatment
356
doi:10.10076/s12155-015-9590-5.
357 358
process
for
lignocellulosic
biorefineries.
Bioenergy
Res.
1-16.
9. Biswas, R., Uellendahl, H., Ahring, B. K., 2014. Wet explosion pretreatment of sugarcane bagasse for enhanced enzymatic hydrolysis, Biomass Bioenergy 61, 104-113.
359
10. Boaro, A. A., Kim, Y. M., Konopka, A. E., Callister, S. J., Ahring, B. K., 2014.
360
Integrated 'omics analysis for stdying the microbial community response to a pH
16
361
perturbation of a cellulose-degrading bioreactor culture. FEMS Microbiol. Ecol. 90, 802-
362
815.
363 364 365 366 367 368 369 370 371 372
11. Chang, H. N., Kim, N. J., Kang, J., Jeong, C. M., 2010. Biomass-derived volatile fatty acid platform for fuels and chemicals. Biotechnol. Bioprocess Eng. 15(1), 1-10. 12. Dijkstra, J., 1994. Production and absorption of volatile fatty acids in the rumen, Livest. Prod. Sci. 39. 61-69. 13. Ehsanipour, M., Suko, A. V., Bura, R., 2016. Fermentation of lignocellulosic sugars to acetic acid by Moorella thermoacetica. J. Ind. Microbiol. 43, 807-816. 14. Giampietro, M., Ulgiati, S., Pimental, D., 1997. Feasibility of large-scale biofuel production. BioSci. 47(9), 587-600. 15. Gourdon, R., Vermande, P., 1987. Effects of propionic acid concentration on anaerobic digestion of pig manure. Biomass 13, 1-12.
373
16. Grandview Research Inc, Global propionic acid market by application (animal feed,
374
calcium & sodium propionate, cellulose acetate propionate) expected to reach USD 1.53
375
billion,
376
https://www.grandviewresearch.com/press-release/global-propionic-acid-market.
377
accessed: December 2015.
by
2020:
Grandview
Research,
Inc.
December
2015, Last
378
17. Hess, M., Sczyrba, A., Egan, R., Kim, T. W., Chokhawala, H., Schroth, G., Shujun, L.,
379
Clark, D. S., Chen, F., Zhang, T., Mackie, R. I., Pennacchio, L. A., Tringe, S. G., Visel,
380
A., Woyle, T., Wang, Z., Rubin, E. M., 2011. Metagenomic Discovery of Biomass-
381
Degrading Genes and Genomes from Cow Rumen. Science 331(6016), 463-467.
17
382
18. Himmel, M. E., Ding, S. Y., Johnson, D. K., Adney, W. S., Nimlos, M. R., Brady, J. W.,
383
Foust, T. D., 2007. Biomass reacalcitrance - engineering plants and enzymes for biofuels
384
production, Science. 315, 804-807.
385
19. Holtzapple, M. T., Granda, C. B., 2009. Carboxylate platform: The MixAlco process Part
386
I: Comparison of three biomass conversion platforms. Appl. Biochem. Biotechnol. 156,
387
95-106.
388 389 390 391
20. Hull, S. R., Yang, B. Y., Venzke, D., Kulhavy, K., Montgomery, R., 1996. Composition of corn steep water during steeping. J. Agricult. Food Chem. 44, 1857-1863. 21. Kadam, K. L., McMillan, J. D., 2003. Availability of corn stover as a sustainable feedstock for bioethanol production. Bioresour. Technol. 88, 17-25.
392
22. Kerr, R. W., Berlin, H., 1933. Treatment of steep water. USA Patent No. 1918812.
393
23. Kim, N. K., Park, G. W., Kang, J., Kim, Y. C., Chang, H. N., 2013.Volatile fatty acid
394
production from lignocellulosic biomass by lime pretreatment and its applications to
395
industrial biotechnology. Biotechnol. Bioprocess Eng. 18, 1163-1168.
396
24. Lawford, H. G., Rousseau, J. D., 1997. Corn steep liquor as a low cost effective nutrition
397
adjunct in high-performamce Zymomonas ethanol fermentations, Appl. Biochem.
398
Biotechnol. 63, 287-304.
399 400
25. Liggett, R. W., Koffler, H., 1948. Corn steep liquor in microbiology. Bacteriol. Rev. 12(4), 297-311.
401
26. Lynd, L. R., Laser, M. S., Bransby, D., Dale, B. E., Davison, B., Hamilton, R., Himmel,
402
M., Keller, M., McMillan, J. D., Sheehan, J., Wyman, C. E., 2008. How biotechnology
403
can transform biofuels. Nature Biotechnol. 26, 169-172.
18
404
27. Lynd, L. R., Weimer, P. J., Zyl, W. H., Pretorius, I. S., 2002. Microbial cellulose
405
utilization: Fundamentals and biotechnology. Microbiol. Mol. Biol. Rev. 66, 506-577.
406
28. Mordor Intelligence Inc., Global acetic acid market – Growth, trends and forecasts (2015
407
to 2020), September 2016, http://www.mordorintelligence.com/industry-reports/global-
408
acetic-acid-market-industry?gclid=CI_z1Lfxkc8CFUGTfgod4WcODw. Last accessed:
409
September 2016.
410
29. Ni, B. J., Liu, H., Nie, Y. Q., Zeng, R. J., Du, G. C., Chen, J., Yu, H. Q., 2011. Coupling
411
glucose
fermentation
and
homoacetogenesis
for
elevated
acetate
production:
412
Experimental and mathematical approaches. Biotechnol. Bioeng. 108, 345-353.
413
30. Nissila, M. E., Tahti, H. P., Rintala, J. A., Puhakka, J. A., 2011. Thermophilic hydrogen
414
production from cellulose with rumen fluid enrichment cultures: Effects of different heat
415
treatments. Int. J. Hydrogen Energy 36, 1482-1490.
416
31. Njoku, S. I., Uellendahl, H., Ahring, B. K., 2013. Comparing oxidation and dilute acid
417
wet explosion pretreatment of Cocksfoot grass at high dry matter concentration for
418
cellulosic ethanol production. Energy Sci. Eng. 1(2), 89-98.
419
32. Pham, T. N., Nam W. J., Jeon, Y. J., Yoon, H. H., 2012. Volatile fatty acids production
420
from marine microalgae by anaerobic fermentation. Bioresour. Technol. 124, 500-503.
421
33. Pind, P. F., Angelidaki, I., Ahring, B. K., 2003. Dynamics of the anaerobic process:
422 423 424
Effects of volatile fatty acids. Biotechnol. Bioeng. 82(7), 791-801. 34. Rana, D., Rana, V., Ahring, B. K., 2012. Producing high sugar concentrations from loblolly pine using wet explosion pretreatment. Bioresour. Technol. 121, 61-67.
19
425
35. Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., Crocker, D.,
426
2011. Determination of structural carbohydrates and lignin in biomass. Technical Report
427
NREL/TP-510-42618, Colorado: National Renewable Energy Laboratory, pp. 1-15.
428
36. Sluiter, A., Hames, B., Hyman, D., Payne, C., Ruiz, R., Scarlata, C., Sluiter, J.,
429
Templeton, D., Wolfe, J., 2008. Determination of total solids in biomass and total
430
dissolved solids in liquid process samples. Technical Report NREL/TP-510-42641 USA,
431
pp. 1-6.
432 433
37. Solomon, B. D., Barnes, J. R., Halvorsen, K. E., 2007. Grain and cellulosic ethanol: History, economics and energy policy. Biomass Bioenergy. 31, 416-425.
434
38. Varga, E., Schmidt, A. S., Reczey, K., Thomsen, A. B., 2003. Pretreatment of corn stover
435
using wet oxidation to enhance enzymatic digestibility. Appl. Biochem. Biotechnol. 104,
436
37-50.
437 438
39. Verser, D., Eggeman, T., 2005. Process for producing ethanol. USA Patent No. 6,927,048 B2.
439
40. Walsh, M. E., Perlack, R. L., Turhollow, A., Ugarte, D. D. I. T., Becker, D. A., Graham,
440
R. L., Slinsky, S. E., Ray, D. E., 2000. Biomass feedstock in the United States: 1999 state
441
level analysis, Tennessee: Oak Ridge National Laboratory, pp. 1-16.
442
41. Wolin, M. J., Miller, T. L., Stewart, C. S., 1997. Microbe-Microbe interactions, in: P. N.
443
Hobson, C. S. Stewart (Eds.), The Rumen Microbial Ecosystem. Blackie Academic &
444
Professional, London, pp. 467-491.
445 446 447
42. Wychen, S. V., Laurens, L. M., 2013. Determination of total solids and ash in algal biomass. Technical Report NREL/TP-5100-60956 USA, pp. 1-7. Figure Caption List 20
448
Figure 1. Schematic representation of the semi-continuous VFA bioreactor
449 450
Figure 2. Concentrations of individual VFA (g/L) produced from rumen fermentation of wet
451
exploded corn stover at (a) 5% TS; and (b) 2.5% TS. The decrease in VFA’s between day 26
452
and 40 was due to problems with the feed pump.
453 454
Figure 3. Total VFA concentrations in acetic acid equivalents in g/L from rumen fermentation of
455
wet exploded corn stover using 2.5% and 5% TS
456 457
Figure 4. Comparison of gas production and acetic acid concentration in reactors with 2.5% and
458
5% TS from rumen fermentation of wet exploded corn stover: (a) hydrogen and methane
459
concentrations (wt%), where hydrogen concentration (wt%) is in the secondary y axis, as a line
460
graph and methane concentration (wt%) is in the y axis as a scatter plot; (b) acetic acid
461
concentration (g/L)
462 463
Table Caption List
464
Table 1. Compositional analysis of separated solid and liquid fractions of feed and effluent
465
before and after rumen fermentation at day 62 of wet exploded corn stover at 2.5% and 5% solids
466
21
467
22
468
23
469
24
470 471
25
472
Table 1 Compositional analysis of separated solid and liquid fractions of feed and effluent before
473
and after rumen fermentation at day 62 of wet exploded corn stover at 2.5% and 5% solids
474
loading 2.5% TS Feed
5% TS
Effluent
Feed
Effluent
50.7
25.0
Solid Fraction* Total Carbohydrates (%g/g 53.1
35.2
Cellulose (%g/g biomass)
36.8
22.8
36.1
19.6
Hemicellulose (%g/g biomass)
16.3
12.4
14.6
5.4
Soluble Lignin (%g/g biomass)
2.64
4.37
2.91
5.11
Insoluble Lignin (%g/g biomass)
40.8
57.1
41.7
61.8
Lignin-Carbohydrate Ratio
0.82
1.75
0.88
2.68
Acetic acid (g/L)
0.34
10.4
2.47
21.9
Propionic acid (g/L)
0.17
6.88
0.11
7.03
Butyric acid (g/L)
0.11
1.39
0.56
2.80
Valeric acid (g/L)
0.00
0.00
0.00
0.88
biomass)
Liquid Fraction
475
* Solid fraction was obtained after filtration and washing with water.
476 477 478 479
26
480 481
Fermentation of wet-exploded corn stover for the production of volatile fatty acids
482 483
Highlights
484
•
Volatile fatty acids (VFA) production was studied using rumen bacteria.
485
•
Fermentation of pretreated corn stover to VFA.
486
•
VFA yield increased to 40.8g/L at 5% Total Solids (TS) from 22.8g/L at 2.5% TS.
487
•
Valeric acid production was seen only at 5% TS but not at 2.5% TS.
488
•
Composition analysis showed 33% carbohydrates conversion at 2.5% and 50% at 5% TS.
489 490
27
S0960-8524(16)31663-7 http://dx.doi.org/10.1016/j.biortech.2016.12.012 BITE 17388
To appear in:
Bioresource Technology
Received Date: Revised Date: Accepted Date:
21 October 2016 1 December 2016 2 December 2016
Please cite this article as: Murali, N., Fernandez, S., Kiaer Ahring, B., Fer mentation of wet-exploded cor n stover for the pr oduction of volatile fatty acids, Bioresource Technology (2016), doi: http://dx.doi.org/10.1016/ j.biortech.2016.12.012
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
2
Fermentation of wet-exploded corn stover for the production of volatile fatty acids
3
Nanditha Murali, Sebastian Fernandez, and Birgitte Kiaer Ahring*
4
Bioproducts, Sciences and Engineering Laboratory, Washington State University, Tri-Cities, Richland, WA-99354
1
5 6 7 8 9 10 11
Abstract
12
Volatile fatty acids (VFA) have been used as platform molecules for production of biofuels and
13
bioproducts. In the current study, we examine the VFA production from wet-exploded corn
14
stover through anaerobic fermentation using rumen bacteria. The total VFA yield (acetic acid
15
equivalents) was found to increase from 22.8g/L at 2.5% total solids (TS) to 40.8g/L at 5%TS. It
16
was found that the acetic acid concentration increased from 10g/L to 22g/L at 2.5% and 5%TS,
17
respectively. An increased propionic acid production was seen between day 10 and 20 at 5%TS.
*
Corresponding author
Dr. Birgitte K. Ahring Bioproducts Sciences and Engineering Laboratory Washington State University, Tri-cities 2710 Crimson Way Richland, WA 99354 Tel.: 01-(509)-372-7682 Fax: 01-(509)-372-7690. E-mail address: [email protected] 1
18
Valeric acid (4g/L) was produced at 5%TS and not at 2.5%TS. Composition analysis showed
19
that 50% of the carbohydrates were converted to VFA at 5%TS and 33% at 2.5%TS. Our results
20
show that rumen fermentation of lignocellulosic biomass after wet explosion can produce high
21
concentrations of VFA without addition of external enzymes of importance for the process
22
economics of lignocellulosic biorefineries.
23 24
Key Words: Acetic acid, Volatile Fatty Acids, Anaerobic Fermentation, Corn Stover, Rumen.
25 26
1. Introduction
27
Volatile Fatty Acids (VFA), primarily short-chain carboxylic acids such as acetic, propionic and
28
butyric acids can be produced from petrochemical sources and further form biomass feedstocks.
29
VFA has been suggested as valuable platform molecules for the production of biofuels and
30
bioproducts (Chang et al. 2010). Bio-based VFA are now produced from anaerobic fermentation
31
of organic materials such as waste products or agricultural residues (Solomon et al. 2007). Use of
32
lignocellulosic biomass through biochemical conversion to produce biofuels and bio-products
33
from biomass have been shown to be burdened by the need of enzyme addition leading to high
34
operational cost (Giampietro et al. 1997).
35 36
Volatile fatty acids are the main intermediates in the production of methane from organic
37
materials (Ahring et al. 1995) and have been used as a control parameter to assess effectiveness
38
of anaerobic digestion (Pind et al. 2003). Under high loading of a bioreactor, a higher VFA
39
production will occur and methane production can be reduced as VFA production usually occurs
40
fast within 1 – 3 days, as opposed to 7– 15 days for the production of methane, leading to an 2
41
initial drop in pH (Pham et al. 2012). Studies have shown that these VFA can further be
42
harvested and used for production of biofuels such as ethanol, propanol or butanol by catalysis
43
(Holtzapple and Granada, 2009).
44 45
The market for acetic acid is growing at ca. 5% a year and estimated to 16155 metric ton by
46
2020. Acetic acid is used as building block for production of various chemicals, such as vinyl
47
acetate monomer (main use), purified terephthalic acid, acetate esters, and acetic anhydride of
48
importance for the synthetic fiber and textile industry as well as for ink and pesticides (Mordor
49
intelligence Inc, 2016). Also other acids such as propionic and butyric has growing markets for
50
many applications (Grandview Research Inc, 2015). Producing sustainable VFA from biomass
51
materials can, therefore, be of major importance in the future.
52 53
A wide variety of biomass feedstocks, predominantly agricultural residues, have been used for
54
the production for VFA (Kadam and McMillan, 2003). In the United States, corn is the most
55
widely grown crop (Kadam and McMillan, 2003) and it is estimated that approximately 153
56
million dry tons of corn stover is available as biomass feedstock every year (Walsh et al. 2000).
57
The main advantages of using lignocellulosic biomass for VFA production, is that the biomass
58
has a lower cost, is generally abundantly available and is not in competition with food,
59
renewable, and reduces net CO2 emissions (Berndes et al. 2001). In raw biomass, cellulose fibers
60
(about 35-45% of the biomass) are encased in a hemicellulose (20-25%) and lignin (15-20%)
61
matrix via carbon – carbon linkages (Himmel et al. 2007; Agbor et al. 2011). This matrix has to
62
be broken down for making major parts of the cellulose accessible for fermentation. The native
63
form of biomass is recalcitrant to enzymatic hydrolysis with cellulase enzymes and requires 3
64
pretreatment to remove the hemicellulose-lignin linkages holding the biomass together (Lynd et
65
al. 2002). Conventional pretreatment methods with acids, or alkali or heat (Lynd et al. 2008)
66
introduces harsh chemicals, which could be detrimental for microbial growth required for
67
fermentation. Wet Explosion (WEx) pretreatment, on the other hand, uses oxygen and water at
68
elevated temperature and pressure and the process is terminated by ‘exploding’ the biomass into
69
a flash tank (Biswas et al. 2015). In addition to not using any harsh chemicals, which is both
70
detrimental to microbes and further will be in the fermentation broth., WEx also produces lower
71
amounts of inhibitory products like furfural and hydroxymethylfurfural (Njoku et al. 2013). WEx
72
also opens up the crystalline structure of cellulose (Rana et al. 2012), thus making cellulose more
73
accessible to microbial enzymes produced during fermentation (Varga et al. 2003; Ahring and
74
Munck, 2009).
75 76
Zeachem has suggested to use Mororella thermoaceticum for producing acetic acid as platform
77
molecule for bio-products and biofuel production using cellulosic sugars as substrate (Verser and
78
Eggeman, 2005). However, this fermentation needs sterility to operate along with the need of
79
costly nutrient additions (Verser and Eggeman, 2005). When compared to such monocultures,
80
robust mixed cultures can result in improved process economics since it does not require
81
sterilization and is more resistant to inhibitory chemicals present in pretreated biomass (such as
82
lignin compounds) (Agler et al., 2011). Ruminants like cattle are known to produce VFA via
83
anaerobic fermentation, with acetic, propionic and butyric acids being the predominant products
84
(Dijkstra, 1994). Previous studies have shown that stable mixed microbial culture such as rumen
85
have the ability to produce necessary enzymes for degradation of carbohydrate polymers into
86
monomers as well as for VFA production from both C5 and C6 sugars (Hess et al. 2011; Boaro 4
87
et al. 2014). However, few studies have successfully tested rumen fermentation of lignocellulosic
88
biomass like corn stover into VFA. The primary goal of the current study is to optimize VFA
89
production in bioreactors operated with pretreated lignocellulosic biomass (corn stover) at
90
different TS concentrations (2.5 and 5%) using inoculum from the rumen grown under
91
mesophilic conditions.
92 93
Pretreated corn stover is used as carbon source in our experiments and extra nutrients are
94
provided in the form of corn steep liquor (CSL) (Amartey and Jeffries, 1994). CSL is a
95
byproduct of the corn milling process where shelled corn is first steeped to form corn steep
96
water, which is then evaporated to form corn steep liquor. On average, CSL contains 5% lactic
97
acid and 0.1% VFA (Kerr and berlin, 1933) and it also contains amino acids (Hull et al. 1996),
98
which could aid in the growth of the microbes. In our study, the concentration of CSL is initially
99
evaluated to assess its effectiveness for growing our bacterial consortia on pretreated corn stover
100
with CSL as sole further nutrient.
101 102 103
2. Materials and methods:
104
2.1. Inoculum: Rumen fluid was collected at a mobile slaughter facility in Richland, WA. The
105
animal weighed 1200 pounds and was fed 15 pounds of grain (molasses, corn wheat and barley)
106
and alfalfa per day. The samples were collected in bottles, which were filled to capacity, to
107
minimize oxygen in the bottle and transported to the laboratory on ice. Upon arrival in the lab,
108
the samples were transferred in a sterile bench to sterile bottles, degassed under 80% N2: 20%
109
CO2 for 30 minutes, sealed and stored at –20°C until further use. The nitrogen carbon dioxide 5
110
gas mix was obtained from Oxarc® Inc. (Pasco, WA, USA). As inoculum, 5% rumen fluid was
111
added to each bioreactor.
112 113
2.2. Substrate: Corn stover was obtained from Iowa State University, Ames, IA and milled to 2-
114
mm size particles using Retsch SM 200 cutting mill (Retsch® Inc. PA, USA). The corn stover
115
was pretreated using Wet Explosion Pretreatment (WEx) method at 10% solids loading as
116
previously described (Biswas et al. 2014) at 190°C for 30 min with oxygen loading of 7.5%.
117
These conditions were selected for maximum accessibility of the cellulose to cellulolytic
118
enzymes produced during rumen fermentation (data not shown).
119 120
2.2.1. Chemicals: 5N Sodium hydroxide (Sigma Aldrich, Saint louis, MO, USA) was used to
121
maintain the pH of the fermenters and 6.5. 4mM sulfuric acid (Sigma Aldrich, Saint louis, MO,
122
USA) was used to both prepare the HPLC samples and as an eluent for HPLC. Corn steep liquor
123
(2% concentration) was from Sigma Aldrich, Saint louis, MO, USA.
124 125 126
2.3. Fermentation:
127
2.3.2. Assessing the effect of corn steep liquor in batch VFA fermentation – Initial batch
128
fermentations were done in 50mL serum vials to assess the effect of corn steep liquor at different
129
solids loading (2.5% and 5% TS) using rumen inoculum. In two sets of batch experiments, wet-
130
exploded corn stover supplemented with 2% (w/w) corn steep liquor (Lawford and Rousseau,
131
1997) was used as the substrate. To the other two sets of batch experiments, only corn stover
6
132
without the corn steep liquor was used as substrate. 5% rumen fluid was used as inoculum for all
133
the fermentations.
134 135
2.3.2. Effect of substrate concentration on VFA fermentation – Two reactors (3L Applikon®
136
ezControl Autoclavable bioreactor (Applikon Biotechnology B.V, Netherlands) were set up with
137
pretreated corn stover – one with 2.5% TS and one with 5% TS (Fig. 1) to compare the effect of
138
substrate concentration on rumen fermentation. The bioreactors were initially sterilized by
139
autoclavation and filled with 800ml of pretreated corn stover. 40ml of rumen inoculum was
140
aseptically added to the bioreactors while the reactors were gassed using 80:20 wt% N2/CO2
141
gaseous mixture. 5M sodium hydroxide solution was used for maintaining pH during anaerobic
142
fermentation. The pH of the bioreactor was initially adjusted to 6.5 using 5M sodium hydroxide
143
solution and temperature was set to 37°C with a stirrer speed at 150 rpm. The fermentation was
144
operated semi-continuously with replacement of 100ml of fresh media every day and an overall
145
hydraulic retention time of 6 days.
146 147 148 149
2.4. Analyses:
150
2.4.1. Measurement of VFA Using HPLC – The liquid samples from the fermentation were
151
centrifuged at 10,000 rpm for 10 minutes and the supernatant and analyzed using high
152
performance liquid chromatography (HPLC) analysis after filtration through a 0.2 micron filter
153
and dilution (3 times) using 4mM sulfuric acid. Analysis were done using an UltiMate® 3000
154
HPLC system (Dionex, Sunnyvale, CA) with an Aminex® 87H Column 250 X 4.6mm (Bio – 7
155
Rad, Hercules, CA) and a Shodex RI – 101 refractive index detector. Sulfuric acid (4mM) in
156
water was used as the eluent, flowing through the 87H column at a constant flow rate of
157
0.6mL/min in a constant temperature oven at 60°C. The total time for analysis of the
158
fermentation sample was 68 minutes.
159 160
2.4.2. Gas Analyses using Gas Analyzer – Random head space gas measurements were done
161
using the Universal Gas Analyzer, UGA Series (Stanford Research Systems, Sunnyvale, CA).
162 163
2.4.2. Calculations: The major acids produced in the fermentation are acetic, propionic and
164
butyric acids. These were converted to acetic acid equivalents to report total VFA. The
165
calculations are as follows:
166
= + ∗ ∗ ! . " # + $%& ∗ ∗ "". #
.
!
.
167
Where, molar mass of acetic acid = 60.05; molar mass of propionic acid = 74.08 and molar mass
168
of butyric acid = 88.11. The molar mass of the acids is then multiplied by the number of grams
169
of oxygen required to completely breakdown the acids to carbon dioxide to yield total VFA in
170
acetic acid equivalents.
171 172
2.4.2. Feedstock characterization – Total solids (TS) and volatile solids (VS) were measured
173
using the method described by Sluiter et al. (2008). Crucibles containing liquid effluent was
174
placed in a drying oven at 105 ± 3°C overnight for 20 hours. The dried samples were allowed to
175
cool to room temperature before weighing them. The samples were placed back in the oven and
176
dried to constant weight. After weighing for total solids (TS), ash content was measured, based
177
on the method described by Wychen and Laurens (Wychen and Laurens, 2013), by heating the 8
178
crucible at 575 ± 25°C in a muffle furnace for 6 hours. The burnt samples were cooled to room
179
temperature before being weighed.
180 181
2.4.3. Composition analysis – For the determination of total carbohydrates and lignin, the
182
fermentation feed and effluent were air dried at room temperature thoroughly and the solid
183
fraction was used to analyze total carbohydrates and lignin (both soluble and insoluble) content
184
using the method described by Sluiter et al. (2011). Total carbohydrates (cellulose and
185
hemicellulose) and insoluble lignin was measured after hydrolysis, while soluble lignin was
186
analyzed using UV spectrophotometer (Jenway 6405 UV/Visible NJ, USA) at 320nm (Sluiter et
187
al. 2011). All measurements were done in triplicates.
188 189
3. Results and Discussion
190
3.1. Effect of corn steep liquor on batch rumen fermentation of wet exploded corn stover
191
The total VFA production and individual VFA’s obtained from batch anaerobic rumen
192
fermentation of wet exploded corn stover with and without corn steep liquor was assessed (data
193
not shown). It was seen that the batch experiments with corn steep liquor produced almost 30%
194
more VFA than the experiments without CSL while only insignificant concentrations of VFA
195
was present in 2% CSL (Liggett and Koffler, 1948). We also found that lactic acid present in
196
the pretreated corn stover supplemented with corn steep liquor was only 0.24g/L and hence
197
would not significantly impact VFA yields. Since these preliminary experiments were done
198
in serum vials in batch mode, continuous monitoring of pH was not possible. In the absence of
199
corn steep liquor, there was no significant difference between total VFA production at 2.5% and
200
5% TS. However, through addition of 2% CSL, the total VFA yields at 5% TS was found to give 9
201
almost 1.4 times higher than that at 2.5% TS. This could indicate that additional nutrient is
202
needed to maximize VFA production at increased solids loading.
203 204
3.2. Semi-continuous fermentation - Comparison of total VFA yield
205
The VFA concentration expressed as total VFA produced during fermentation of the rumen
206
culture at different solids loadings are shown in Fig. 4. A study by Pham and co-workers,
207
showed that total VFA production of 5 to 7g/L during fermentation of 30g/L microalgae (Pham
208
et al. 2012). However, they found that when the substrate concentration was increased by 50%,
209
the VFA yield increased to between 10 to 15g/L. We further found that the total VFA yield was
210
22.8g/L (or 1.79 g/g VS) with 2.5% TS and 40.8g/L (or 2.11 g/g VS) at 5% solids loading. Our
211
studies found that acetic acid was the major VFA produced followed by propionic acid and
212
butyric acid (Fig. 2). In the reactor with 2.5% TS, acetic acid concentration increased to 13g/L
213
(0.94g/g VS), while propionic acid concentration increased to 10g/L (0.78g/g VS) after 10 days.
214
However, in the reactor with 5% TS, acetic acid concentration increased to 22g/L (1.14 g/g VS),
215
while maximum propionic acid concentration was 10g/L (0.52 g/g VS). We also found that
216
valeric acid concentration increased to 4g/L (0.21g/g VS) in the reactor with 5% TS (Fig 2) and
217
that valeric acid was only produced in this reactor. Other studies using rumen bacteria under
218
thermophilic conditions (60°C) showed far lower concentrations (417mg/L of VFA) from
219
cellulose (Nissila et al. 2011). Studies published using pure cellulose and starch substrates for
220
rumen fermentation showed VFA yields ranging from 0.417g/L to as high as 4.563g/L (Nissila et
221
al. 2011). Another study using modified lime pretreatment for pretreatment of lignocellulosic
222
biomass resulted in a maximum VFA production of 5.32g/L which accounted for around 0.5g
223
VFA/g biomass added (Kim et al. 2013). In comparison with existing literature on production of 10
224
VFA, it can be clearly seen that significantly higher VFA concentrations were obtained in our
225
current study on pretreated biomass material.
226 227
Studies using monocultures such as Moorella thermoacetica have shown acetic yields of 17-
228
18g/L from pretreated and enzymatically hydrolyzed lignocellulosic biomass hydrolysate and the
229
main factor that adversely affected acetic acid yields was the inefficient consumption of
230
arabinose, mannose and galactose (Ehsanipour et al. 2016). Apart from lower VFA yields when
231
compared to our current study, pure cultures also need sterile conditions and further requires
232
sugars for their conversion while our study is done non-sterile and without hydrolyzing the
233
pretreated biomass material. Previous proteomic and metabolomic studies on anaerobic rumen
234
fermentation done in our laboratory indicated increased ß-glucosidase activity and the activity of
235
Clostridia and Fibirobacter clusters during fermentation of avicel (Boaro et al. 2014). This study
236
shows that unlike pure cultures, an anaerobic cellulose fermenting culture established with
237
rumen fluid as inoculum will have both cellulase producing microbes along with acidogens
238
producing VFA.
239 240
3.3. Comparison of VFA production as function of solids loading in reactors
241
As shown in Figure 3, the total VFA production expressed in acetic acid equivalents increased as
242
a function of time for the first 21 days and was initially higher at lower solids loading. After
243
fermentation for around 10 days, higher VFA production was seen with 5% TS while there was
244
no significant difference in the VFA production at 2.5% and 5% TS between 5 and 10 days.
245
After 10 days it seems like the bioreactors were stabilizing (or acclimating) to the wet-exploded
246
corn stover. This can be attributed to the sudden increase in the VFA productivity after the 10th 11
247
day from 0.07 to 0.19 g/L/h at 2.5% TS and 0.19 to 0.77 g/L/h at 5% solids TS. At the end of the
248
experiment (day 62) a stable microbial productivity of 0.24 g/L/h and 0.40 g/L/h was maintained
249
in the bioreactors with 2.5% and 5% TS, respectively.
250 251
3.4 Comparison of biomass composition before and after fermentation
252
Compositional analysis was done on the wet exploded corn stover used as fermenter feed and the
253
effluent obtained at end of the experiments (62 days) and the data is shown in Table 1. It can be
254
seen from Table 1 that the feed had approximately 51wt% carbohydrates and that approximately
255
50% of the carbohydrates was converted to VFA from anaerobic fermentation at 5% TS while
256
only around 33% of the carbohydrates was converted at 2.5% TS. It can be seen from Table 1
257
that, at 2.5% TS that the hemicellulose degradation was considerably lower (16.3wt% in feed
258
reduced to 12.4wt% in effluent) when compared to that at 5% TS (14.6wt% in feed to 5.4wt% in
259
effluent). A similar trend was also seen for cellulose conversion between 2.5% TS (36.8wt% in
260
feed to 22.8wt% in effluent) and 5% TS (36.1wt% in feed to 19.6wt% in effluent) but the
261
difference in values between the two solid loadings was found to be lower for cellulose
262
compared to hemicellulose conversion. This can be attributed to a greater stability of the rumen
263
bacteria towards hemicellulose and cellulosic conversion to volatile fatty acids at higher solids
264
loading. The efficient conversion of hemicellulose along with cellulose by the rumen microflora
265
at 5% TS is significant and shows great potential for further optimization of the anaerobic
266
fermentation to higher loadings to further maximize VFA production.
267 268
Assuming inefficient lignin utilization by the rumen bacteria during anaerobic fermentation,
269
lignin-carbohydrate ratio can be used to quantify effectiveness of carbohydrate utilization 12
270
towards VFA production. It can be seen from Table 1 that the lignin-carbohydrate ratio increased
271
from 0.88 to 2.68 at 5% TS while it increased from 0.82 to 1.75 at 2.5% TS confirming
272
significant biomass sugar utilization by rumen bacteria during anaerobic fermentation to produce
273
VFA. However, in the contrary, higher TS loading (5%) showed a lower carbohydrate-to-VFA
274
conversion of 3.17 g VFA/g carbohydrate when compared to 2.5% TS (5.09 g VFA/g
275
carbohydrate). This trend can also be clearly seen by comparing the acetic acid concentrations,
276
given in Table 1, between feed and effluent. At 2.5% solids loading, 10.06 g/L of acetic acid is
277
produced after consumption of 17.9wt% carbohydrates resulting in 2.25 g acetic acid produced/g
278
carbohydrate consumed while at 5% solids loading, 18.43 g/L of acetic acid is produced after
279
consumption of 25.7wt% carbohydrates resulting in only 1.43 g acetic acid produced/g
280
carbohydrate. While not directly indicated in Table 1, there was negligible amount of C5 or C6
281
sugars in the liquid effluent at both TS. The only difference between the fermentation conditions
282
at both TS is the lignin content which is equal to 1.09 g lignin/L (acid soluble + insoluble lignin
283
given in Table 1) at 2.5% TS and 2.23 g lignin/L at 5% TS and further studies to optimize this
284
process can significantly enhance effective biomass conversion at higher TS loading to obtain
285
higher carbohydrate conversion to selectivity produce VFA. The lower acetic acid production at
286
5% TS loading, could be a result of lignin being converted to methane. Previous studies have
287
reported that wet exploded lignin can be converted to methane (Ahring et al. 2015) and after
288
pretreatment, 44% lignin was converted to methane using anaerobic digestion. The study also
289
showed that pretreated lignin produced 4.5 times higher methane than the non-pretreated lignin
290
(Ahring et al. 2015).
291 292
3.5. Comparison of individual VFA as a function of time and solids loading 13
293
As indicated previously, the major acids produced during rumen fermentation of wet exploded
294
corn stover were acetic, propionic and butyric acids, with trace amounts of formic and valeric
295
acids. The time-based concentration profile at 2.5% and 5% TS for the individual VFA is shown
296
in Fig. 2 (a) & (b), respectively. Liquid fraction analysis showed only trace amounts of C5
297
(xylose) and C6 (glucose) sugars at different sampling times for both 2.5% and 5% TS. This
298
indicates that the sugar oligomers and monomers obtained through wet explosion pretreatment of
299
corn stover was directly be converted to VFA along with the polymers in the material. The
300
culture in the reactors was, therefore, capable of producing the necessary cellulolytic enzymes
301
for getting into the sugar oligomers and polymers.
302 303
It can be seen from Figure 2 that acetic acid was the most predominant VFA produced through
304
anaerobic rumen fermentation of pretreated corn stover. Although issues with the feeding pump
305
caused a decrease in VFA concentration between days 26 and 40 in the reactor fed with 2.5%
306
TS, once the problem was resolved the fermentation stabilized again at the same initial VFA
307
level. This shows that rumen fermentations can be maintained at stable levels for extended
308
periods of time and will have some resilience to perturbations. From Figure 2 it can be seen that
309
increasing amounts of propionic acid was produced at 5% solids loading especially between days
310
10-20. Though studies have indicated that an increase in the propionic acid concentration can
311
slightly affect the rate of anaerobic digestion (AD) and in turn VFA production (Gourdon and
312
Vermande, 1987), it can be seen that the rate of anaerobic fermentation at 5% solids loading was
313
not affected by this propionic acid concentration and overall the VFA concentration was
314
significantly higher compared to the 2.5% TS reactor. Gas analyses shows that increases in the
315
hydrogen concentrations further increased acetic acid production in the rumen fermentation (Fig. 14
316
4). In the reactor with 5% TS slight increases in acetic acid concentration are seen coinciding
317
with increased hydrogen concentration. This might be attributed to activity by homo-acetogens
318
present in the rumen, which might convert the hydrogen together with CO2 to acetic acid (Ni et
319
al. 2011; Wolin et al. 1997) even that the hydrogen concentration generally was maintained at
320
1.5 wt% primarily due to the presence of methanogens in the rumen microflora. The
321
methanogens reduce hydrogen, carbon dioxide and formate to produce methane, which was the
322
major gas produced in both the reactors, as seen in Fig. 4 (Pind et al. 2003). Another significant
323
difference at different solid loadings is the significant amounts of valeric acid produced at 5%
324
TS. This was not found in the samples collected from the 2.5% TS bioreactor. Further studies are
325
currently being done to understand the differences in microbial composition in the two reactors.
326 327
4. Conclusion:
328
In this study, we have established an active and stable culture in two bioreactors using rumen
329
fluid as inoculum. Anaerobic fermentation of wet exploded corn stover at 2.5% and 5% TS
330
produced 22.8g/L and 40.8g/L of total VFA (acetic acid equivalents), respectively. These results
331
show, by far, the highest level of VFA at mesophilic anaerobic fermentation conditions.
332
Compositional analysis showed efficient cellulose conversion and a higher hemicellulose
333
conversion at 5% TS when compared to 2.5%. Further understanding the mechanism of lignin
334
conversion by rumen bacteria after pretreatment can significantly increase VFA yields.
335 336 337 338
References: 1. Agbor, V. B., Cicek, N., Sparling, R., Berlin, A., Levin, D. B., 2011. Biomass pretreatment: Fundamentals towards application. Biotechnol. Adv. 29, 676-685. 15
339
2. Agler, M. T., Wrenn, B. A., Zinder, S. H., Angenent, L. T., 2011. Waste to bioproduct
340
conversion with undefined mixed cultures: The carbohydrate platform. Trends
341
Biotechnol. 70-78.
342
3. Ahring, B. K., Biswas, R., Ahamed, A., Teller, P. J., Uellendahl, H., 2015. Making lignin
343
accessible for anaerobic digestion by wet-explosion pretreatment. Biores. Technol. 175,
344
182-188.
345
4. Ahring, B. K., Munck, J., 2009. Method for treating biomass and organic waste with the
346
purpose of generating desired biologically based products. US Patent No. 0178671 A1.
347
5. Ahring, B. K., Sandberg, M., Angelidaki, I., 1995. Volatile fatty acids as indicators of
348
process imbalance in anaerobic digestors. Appl. Microbiol. Biotechnol. 43, 559-565.
349
6. Amartey, S., Jeffries, T. W., 1994. Comparison of corn steep liquor with other nurients in
350
the fermentation of D-Xylose by Pichia stipitis CBS 6054. Biotechnol. Lett. 6(2), 211-
351
214.
352 353 354
7. Berndes, G., Azar, C., Kaberfer, T., Abrahamson, D., 2001. The feasibility of large-scale lignocellulose based bioenergy production. Biomass Bioenergy. 20(5), 371-383. 8. Biswas, R., Uellendahl, H., Ahring, B. K., 2015. Wet explosion: a universal and efficient
355
pretreatment
356
doi:10.10076/s12155-015-9590-5.
357 358
process
for
lignocellulosic
biorefineries.
Bioenergy
Res.
1-16.
9. Biswas, R., Uellendahl, H., Ahring, B. K., 2014. Wet explosion pretreatment of sugarcane bagasse for enhanced enzymatic hydrolysis, Biomass Bioenergy 61, 104-113.
359
10. Boaro, A. A., Kim, Y. M., Konopka, A. E., Callister, S. J., Ahring, B. K., 2014.
360
Integrated 'omics analysis for stdying the microbial community response to a pH
16
361
perturbation of a cellulose-degrading bioreactor culture. FEMS Microbiol. Ecol. 90, 802-
362
815.
363 364 365 366 367 368 369 370 371 372
11. Chang, H. N., Kim, N. J., Kang, J., Jeong, C. M., 2010. Biomass-derived volatile fatty acid platform for fuels and chemicals. Biotechnol. Bioprocess Eng. 15(1), 1-10. 12. Dijkstra, J., 1994. Production and absorption of volatile fatty acids in the rumen, Livest. Prod. Sci. 39. 61-69. 13. Ehsanipour, M., Suko, A. V., Bura, R., 2016. Fermentation of lignocellulosic sugars to acetic acid by Moorella thermoacetica. J. Ind. Microbiol. 43, 807-816. 14. Giampietro, M., Ulgiati, S., Pimental, D., 1997. Feasibility of large-scale biofuel production. BioSci. 47(9), 587-600. 15. Gourdon, R., Vermande, P., 1987. Effects of propionic acid concentration on anaerobic digestion of pig manure. Biomass 13, 1-12.
373
16. Grandview Research Inc, Global propionic acid market by application (animal feed,
374
calcium & sodium propionate, cellulose acetate propionate) expected to reach USD 1.53
375
billion,
376
https://www.grandviewresearch.com/press-release/global-propionic-acid-market.
377
accessed: December 2015.
by
2020:
Grandview
Research,
Inc.
December
2015, Last
378
17. Hess, M., Sczyrba, A., Egan, R., Kim, T. W., Chokhawala, H., Schroth, G., Shujun, L.,
379
Clark, D. S., Chen, F., Zhang, T., Mackie, R. I., Pennacchio, L. A., Tringe, S. G., Visel,
380
A., Woyle, T., Wang, Z., Rubin, E. M., 2011. Metagenomic Discovery of Biomass-
381
Degrading Genes and Genomes from Cow Rumen. Science 331(6016), 463-467.
17
382
18. Himmel, M. E., Ding, S. Y., Johnson, D. K., Adney, W. S., Nimlos, M. R., Brady, J. W.,
383
Foust, T. D., 2007. Biomass reacalcitrance - engineering plants and enzymes for biofuels
384
production, Science. 315, 804-807.
385
19. Holtzapple, M. T., Granda, C. B., 2009. Carboxylate platform: The MixAlco process Part
386
I: Comparison of three biomass conversion platforms. Appl. Biochem. Biotechnol. 156,
387
95-106.
388 389 390 391
20. Hull, S. R., Yang, B. Y., Venzke, D., Kulhavy, K., Montgomery, R., 1996. Composition of corn steep water during steeping. J. Agricult. Food Chem. 44, 1857-1863. 21. Kadam, K. L., McMillan, J. D., 2003. Availability of corn stover as a sustainable feedstock for bioethanol production. Bioresour. Technol. 88, 17-25.
392
22. Kerr, R. W., Berlin, H., 1933. Treatment of steep water. USA Patent No. 1918812.
393
23. Kim, N. K., Park, G. W., Kang, J., Kim, Y. C., Chang, H. N., 2013.Volatile fatty acid
394
production from lignocellulosic biomass by lime pretreatment and its applications to
395
industrial biotechnology. Biotechnol. Bioprocess Eng. 18, 1163-1168.
396
24. Lawford, H. G., Rousseau, J. D., 1997. Corn steep liquor as a low cost effective nutrition
397
adjunct in high-performamce Zymomonas ethanol fermentations, Appl. Biochem.
398
Biotechnol. 63, 287-304.
399 400
25. Liggett, R. W., Koffler, H., 1948. Corn steep liquor in microbiology. Bacteriol. Rev. 12(4), 297-311.
401
26. Lynd, L. R., Laser, M. S., Bransby, D., Dale, B. E., Davison, B., Hamilton, R., Himmel,
402
M., Keller, M., McMillan, J. D., Sheehan, J., Wyman, C. E., 2008. How biotechnology
403
can transform biofuels. Nature Biotechnol. 26, 169-172.
18
404
27. Lynd, L. R., Weimer, P. J., Zyl, W. H., Pretorius, I. S., 2002. Microbial cellulose
405
utilization: Fundamentals and biotechnology. Microbiol. Mol. Biol. Rev. 66, 506-577.
406
28. Mordor Intelligence Inc., Global acetic acid market – Growth, trends and forecasts (2015
407
to 2020), September 2016, http://www.mordorintelligence.com/industry-reports/global-
408
acetic-acid-market-industry?gclid=CI_z1Lfxkc8CFUGTfgod4WcODw. Last accessed:
409
September 2016.
410
29. Ni, B. J., Liu, H., Nie, Y. Q., Zeng, R. J., Du, G. C., Chen, J., Yu, H. Q., 2011. Coupling
411
glucose
fermentation
and
homoacetogenesis
for
elevated
acetate
production:
412
Experimental and mathematical approaches. Biotechnol. Bioeng. 108, 345-353.
413
30. Nissila, M. E., Tahti, H. P., Rintala, J. A., Puhakka, J. A., 2011. Thermophilic hydrogen
414
production from cellulose with rumen fluid enrichment cultures: Effects of different heat
415
treatments. Int. J. Hydrogen Energy 36, 1482-1490.
416
31. Njoku, S. I., Uellendahl, H., Ahring, B. K., 2013. Comparing oxidation and dilute acid
417
wet explosion pretreatment of Cocksfoot grass at high dry matter concentration for
418
cellulosic ethanol production. Energy Sci. Eng. 1(2), 89-98.
419
32. Pham, T. N., Nam W. J., Jeon, Y. J., Yoon, H. H., 2012. Volatile fatty acids production
420
from marine microalgae by anaerobic fermentation. Bioresour. Technol. 124, 500-503.
421
33. Pind, P. F., Angelidaki, I., Ahring, B. K., 2003. Dynamics of the anaerobic process:
422 423 424
Effects of volatile fatty acids. Biotechnol. Bioeng. 82(7), 791-801. 34. Rana, D., Rana, V., Ahring, B. K., 2012. Producing high sugar concentrations from loblolly pine using wet explosion pretreatment. Bioresour. Technol. 121, 61-67.
19
425
35. Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., Crocker, D.,
426
2011. Determination of structural carbohydrates and lignin in biomass. Technical Report
427
NREL/TP-510-42618, Colorado: National Renewable Energy Laboratory, pp. 1-15.
428
36. Sluiter, A., Hames, B., Hyman, D., Payne, C., Ruiz, R., Scarlata, C., Sluiter, J.,
429
Templeton, D., Wolfe, J., 2008. Determination of total solids in biomass and total
430
dissolved solids in liquid process samples. Technical Report NREL/TP-510-42641 USA,
431
pp. 1-6.
432 433
37. Solomon, B. D., Barnes, J. R., Halvorsen, K. E., 2007. Grain and cellulosic ethanol: History, economics and energy policy. Biomass Bioenergy. 31, 416-425.
434
38. Varga, E., Schmidt, A. S., Reczey, K., Thomsen, A. B., 2003. Pretreatment of corn stover
435
using wet oxidation to enhance enzymatic digestibility. Appl. Biochem. Biotechnol. 104,
436
37-50.
437 438
39. Verser, D., Eggeman, T., 2005. Process for producing ethanol. USA Patent No. 6,927,048 B2.
439
40. Walsh, M. E., Perlack, R. L., Turhollow, A., Ugarte, D. D. I. T., Becker, D. A., Graham,
440
R. L., Slinsky, S. E., Ray, D. E., 2000. Biomass feedstock in the United States: 1999 state
441
level analysis, Tennessee: Oak Ridge National Laboratory, pp. 1-16.
442
41. Wolin, M. J., Miller, T. L., Stewart, C. S., 1997. Microbe-Microbe interactions, in: P. N.
443
Hobson, C. S. Stewart (Eds.), The Rumen Microbial Ecosystem. Blackie Academic &
444
Professional, London, pp. 467-491.
445 446 447
42. Wychen, S. V., Laurens, L. M., 2013. Determination of total solids and ash in algal biomass. Technical Report NREL/TP-5100-60956 USA, pp. 1-7. Figure Caption List 20
448
Figure 1. Schematic representation of the semi-continuous VFA bioreactor
449 450
Figure 2. Concentrations of individual VFA (g/L) produced from rumen fermentation of wet
451
exploded corn stover at (a) 5% TS; and (b) 2.5% TS. The decrease in VFA’s between day 26
452
and 40 was due to problems with the feed pump.
453 454
Figure 3. Total VFA concentrations in acetic acid equivalents in g/L from rumen fermentation of
455
wet exploded corn stover using 2.5% and 5% TS
456 457
Figure 4. Comparison of gas production and acetic acid concentration in reactors with 2.5% and
458
5% TS from rumen fermentation of wet exploded corn stover: (a) hydrogen and methane
459
concentrations (wt%), where hydrogen concentration (wt%) is in the secondary y axis, as a line
460
graph and methane concentration (wt%) is in the y axis as a scatter plot; (b) acetic acid
461
concentration (g/L)
462 463
Table Caption List
464
Table 1. Compositional analysis of separated solid and liquid fractions of feed and effluent
465
before and after rumen fermentation at day 62 of wet exploded corn stover at 2.5% and 5% solids
466
21
467
22
468
23
469
24
470 471
25
472
Table 1 Compositional analysis of separated solid and liquid fractions of feed and effluent before
473
and after rumen fermentation at day 62 of wet exploded corn stover at 2.5% and 5% solids
474
loading 2.5% TS Feed
5% TS
Effluent
Feed
Effluent
50.7
25.0
Solid Fraction* Total Carbohydrates (%g/g 53.1
35.2
Cellulose (%g/g biomass)
36.8
22.8
36.1
19.6
Hemicellulose (%g/g biomass)
16.3
12.4
14.6
5.4
Soluble Lignin (%g/g biomass)
2.64
4.37
2.91
5.11
Insoluble Lignin (%g/g biomass)
40.8
57.1
41.7
61.8
Lignin-Carbohydrate Ratio
0.82
1.75
0.88
2.68
Acetic acid (g/L)
0.34
10.4
2.47
21.9
Propionic acid (g/L)
0.17
6.88
0.11
7.03
Butyric acid (g/L)
0.11
1.39
0.56
2.80
Valeric acid (g/L)
0.00
0.00
0.00
0.88
biomass)
Liquid Fraction
475
* Solid fraction was obtained after filtration and washing with water.
476 477 478 479
26
480 481
Fermentation of wet-exploded corn stover for the production of volatile fatty acids
482 483
Highlights
484
•
Volatile fatty acids (VFA) production was studied using rumen bacteria.
485
•
Fermentation of pretreated corn stover to VFA.
486
•
VFA yield increased to 40.8g/L at 5% Total Solids (TS) from 22.8g/L at 2.5% TS.
487
•
Valeric acid production was seen only at 5% TS but not at 2.5% TS.
488
•
Composition analysis showed 33% carbohydrates conversion at 2.5% and 50% at 5% TS.
489 490
27