Bioresource Technology 100 (2009) 58905897

Contents lists available at ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

High consistency enzymatic hydrolysis of hardwood substrates

Xiao Zhang a,*, Wenjuan Qin b, Michael G. Paice a, John N. Saddler b
a b

FPInnovations Paprican Division (PAPRICAN), 570 boul. St-Jean, Pointe-Claire, Quebec, Canada H9R 3J9 Faculty of Forestry, University of British Columbia, 2424 Main Mall, Vancovuer, BC, Canada V6T 1Z4

a r t i c l e

i n f o

a b s t r a c t
The feasibility of using a laboratory peg mixer to carry out high consistency enzymatic hydrolysis of lignocellulosic substrates was investigated. Two hardwood substrates, unbleached hardwood pulp (UBHW) and organosolv pretreated poplar (OPP), were used in this study. Hydrolysis of UBHW and OPP at 20% substrate consistency led to a high glucose concentration in the nal hydrolysate. For example, a 48 h enzymatic hydrolysis of OPP resulted in a hydrolysate with 158 g/L of glucose. This is the highest glucose concentration ever obtained from enzymatic hydrolysis of lignocellulosic substrates. Fermentation of UBHW and OPP hydrolysates with high glucose content led to high ethanol concentrations, 50.4 and 63.1 g/L, respectively after fermentation. Our results demonstrate that using common pulping equipment to carry out high consistency hydrolysis can overcome the rheological problems and greatly increase the sugar and ethanol concentrations after the hydrolysis and fermentation. 2009 Elsevier Ltd. All rights reserved.

Article history: Received 16 September 2008 Received in revised form 1 May 2009 Accepted 20 June 2009 Available online 29 July 2009 Keywords: High consistency hydrolysis Lignocellulosic substrates Fermentation Ethanol Rheological problem

1. Introduction Bioconversion of renewable lignocellulosic feedstock, such as wood and agricultural residues, to liquid fuel provides a possible means of reducing greenhouse gasses and alleviating the pressure from fossil fuel shortage (Wyman and Hinman, 1990; Galbe and Zacchi, 2002). A typical lignocellulose-to-ethanol process based on biological routes consists of at least four major steps: partial or complete removal of lignin and hemicellulose by pretreatment, enzymatic hydrolysis of polysaccharides to sugar monomers, fermentation of sugars to ethanol, and recovery of ethanol from the process stream. Despite signicant advances in both scientic understanding and process engineering towards the commercialization of such a bioconversion process, an economically feasible industrial biomass-to-biofuel process is yet to be developed (Sun and Cheng, 2002; Van Wyk, 2001). While most of the process steps need to be further optimized, the enzyme hydrolysis step has been identied as a major techno-economical bottleneck in the entire wood-to-ethanol bioconversion process. Conventional enzymatic hydrolysis of lignocellulosic materials is typically carried out at a substrate consistency below 5% solids content. This results in a sugar concentration below 5% in the hydrolysate and, subsequently, a nal ethanol concentration less than 2% (w/w) after fermentation. In contrast, enzymatic hydrolysis and fermentation of starch-based substrates (e.g. corn) is commonly performed at a substrate loading above 20% of dry matter and over 10% (w/w) nal ethanol concentration is typically obtained after fermentation.
* Corresponding author. E-mail address: [email protected] (X. Zhang). 0960-8524/$ - see front matter 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.biortech.2009.06.082

It is likely that increasing substrate loading during hydrolysis of lignocellulose will lead to increased sugar concentration and higher nal ethanol content after fermentation. This approach will bring economic savings to the bioconversion process, such as reducing capital and operational cost for hydrolysis and fermentation, and minimizing energy consumption during distillation/evaporation and other downstream processes. A previous technoeconomic assessment has suggested that an increase in substrate loading from 5% to 8% (w/w) can reduce the total production cost by nearly 20% (Stenberg et al., 2000; Wingren et al., 2003). A further increase in substrate loading will provide even more signicant cost saving. However, the studies carried out at that time did not achieve an effective hydrolysis at a substrate consistency above 10% using either separate hydrolysis and fermentation (SHF) or simultaneous saccharication and fermentation (SSF) approaches. The study recognized a number of barriers to high consistency hydrolysis including (1) high concentration of brous materials reduces mass transfer rate; (2) high substrate consistency leads to high concentration of inhibitory substances; and (3) high sugar concentration causes severe end-products inhibition effects. Unlike starch-based feedstock, the lignocellulosic substrate is a brous material with a high degree of polymerization (DP). In a water suspension, brous substrates can interact with each other and form bre ocs or, on a larger scale, bre networks. This leads to a considerable increase in the viscosity of the substrate matrix, and creates a so-called rheological problem when the mass transfer rate in the substrate matrix is signicantly hindered. Due to the limited amount of free water present in the matrix, it takes a much longer time to liquefy the matrix and carry out

X. Zhang et al. / Bioresource Technology 100 (2009) 58905897

5891

effective hydrolysis. Rheological problems associated with mixing pulp bre suspensions have long been recognized in pulp and paper manufacturing. Dealing with high substrate consistency is a common practice in wood pulp bleaching. Industrial bleaching equipment is designed to handle pulps at various consistencies, typically up to a maximum of 35%. Medium or high consistency mixing devices can effectively break down bre ocs and networks formed in pulp suspensions above 20% consistency. In this study, we examined the feasibility of using a peg mixer to carry out enzyme hydrolysis of lignocelluloses at high substrate consistency. Two substrates were used, unbleached hardwood pulp (UBHW) and organosolv pretreated poplar (OPP).

(pH 4.8) with differing concentrations of the substrates and enzyme dosages described earlier. All the asks were xed in a controlled environment incubator shaker (New Brunswick Scientic Co., Edison, NJ, USA). The enzymatic hydrolyses were carried out at 50 C with a rotating speed of 200 rpm for up to 96 h at various substrate consistencies. The cellulose-to-glucose conversion yield is dened as the glucose amount in the liquid phase product divided by the cellulose content (as glucose) in the substrate. 2.4. Enzymatic hydrolysis in peg mixer A laboratory peg mixer (Diagram 1) manufactured by the Pulp and Paper Research Institute of Canada (Pointe Claire, QC) was used in this study for carrying out high consistency hydrolysis of UBHW and OPP. The peg mixer has a working volume of 9000 l and approximately 800 g (oven dried weight) substrate was used for batch hydrolysis under the same conditions as the shake asks (temperature, pH and enzyme dosage) except that the mixing speed was set at 20 rpm. Prior to the hydrolysis, the substrate, enzyme and buffer were mixed thoroughly in a Hobart mixer (Hobart, North York ON) before they were transferred to the peg mixer. 2.5. Fermentation of the hydrolysate An industrial adapted Saccharomyces cerevisiae strain (T2 yeast provided by Tembec Inc., Tmiscaming, Qubec) was used for fermentation. The yeast cells were inoculated into 250 ml of YEPD medium (Yeast extract 1%, Peptone 2% and glucose 2%), incubated at 30 C in a rotary shaker (200 rpm) for 24 h. The yeast cells were collected by centrifugation at 5000g for 10 min at 4 . The pellet was washed three times with 10 ml phage storage buffer (PSB). Yeast cells from this preparation were then inoculated into 60 ml of pre-hydrolysate or pure glucose solution. The nal cell concentration was 5.5 mg/mL. The pH of the glucose controls and the hydrolysates was adjusted to 6.5 using 50% NaOH prior to the fermentation after the addition of 0.3% yeast extract, 0.5% peptone and (NH4)2HPO4 to a nal concentration of 20 mM. The fermentation experiment was carried out in 125 ml serum bottles containing 60 ml of hydrolysate. Fermentation was carried out in a rotary shaker (New Brunswick Scientic Co., Edison, NJ, USA) at 30 for up to 96 h. 2.6. Sugar, ethanol and inhibitors analysis During the hydrolysis, 0.5 ml aliquots were taken at different reaction times, and immediately ltered through a 0.45 lm mem-

2. Methods 2.1. Substrates Unbleached hardwood kraft pulp (UBHW) was obtained from a Canadian kraft pulp mill. The organosolv pretreated poplar (OPP) was prepared in Papricans pilot plant by cooking poplar wood chips in 50% (w/w) aqueous ethanol solution with 1.25% H2SO4 as catalyst at 170 C for 60 min (Pan et al., 2006). The extractives content of UBHW and OPP was determined by PAPTAC (Pulp and Paper Technical Association of Canada) standard procedures (STANDARD G.13 and G.20) using acetone as a solvent. The total lignin content (acid soluble lignin and acid in-soluble lignin) of UBHW and OPP was measured following PAPTAC standard procedures G.8 and G.9. The ltrate obtained from lignin analysis was collected and used for sugar analysis. The sugar monomers in the ltrate, including arabinose, galactose, glucose, xylose, and mannose, were separated by an anion exchange column (Dionex CarboPacTM PA1) on a Dionex DX-600 Ion Chromatograph system (Dionex, Sunnyvale, CA) equipped with an AS50 autosampler and a GP50 gradient pump. De-ionized water was used as an eluent at a ow rate of 1 mL/min; 1 M NaOH was used to equilibrate the column after elution of sugars. To optimize baseline stability and detector sensitivity, 0.2 M NaOH was added post column. After being ltered through 0.45 lm nylon syringe lters (Chromatographic Specialties Inc.), a 20 ll sample was injected on the column. The sugar monomers were monitored by a ED50 electrochemical detector with parameters set for pulsed amperometric detection. Sugar standards were prepared and analyzed using the same procedure in order to calibrate the instrument before sample analysis. The Chromeleon 6.5 software was used to control the chromatograph system and quantify sugar concentrations. The viscosity of UBHW and OPP was measured following the Scandinavian pulp, paper and board standard testing process SCAN-C15:62. The degree of polymerization of cellulose (DP) present in UBHW and OPP was estimated using the viscosity value of the respective pulp samples. 2.2. Enzymes Celluclast 1.5L (cellulase) and Novozym 188 (b-glucosidase) were obtained from Novozymes North America (Franklinton, NC). The activity of Celluclast is 80 lter paper units per milliliter (FPU/mL). The activity of Novozym 188 is 450 cellobiase units per milliliter (CBU/mL). The enzyme dosage was 20 FPU cellulase supplemented with 80 CBU of b-glucosidase per gram of cellulose in the substrate. 2.3. Enzymatic hydrolysis in shake asks The batch hydrolysis experiments were carried out in 500-ml asks. The reaction solution contained 200 mM acetate buffer

Diagram 1. The inner chamber of the laboratory peg mixer.

5892

X. Zhang et al. / Bioresource Technology 100 (2009) 58905897

brane lter. The glucose concentration in the resulting ltrate was then determined by the above-mentioned ion chromatograph method. All values are averages obtained from experiments performed in duplicate. Aliquots of 0.5 ml were taken periodically from the fermentation broth to determine the ethanol and glucose concentration. Samples were rst centrifuged to remove the yeast cells and then ltered through a 0.45 lm membrane lter (Millipore, Bedford, MA) The ethanol concentration was determined by gas chromatography as previously described (Robinson et al., 2002) using a HP 6890 series GC system. The glucose concentration was measured by the same HPLC method as described in the previous section. 2.7. Data analysis and number of sample replicates All the results reported for batch hydrolysis and fermentation are the mean values of at least six replicates based on two batches of repeated experiments under the same conditions. Three samples were taken for each time point. The mean and standard deviation were calculated by Origin version 7SR2 (OriginLab Corp., Northampton, MA). 3. Results 3.1. Chemical composition of unbleached hardwood pulp (UBHW) and organosolv pretreated poplar (OPP) As shown in Table 1, UBHW has a cellulose content of approximately 80% with 19.6% of xylan. The pulp contains a small amount of lignin with low extractives content. The UBHW represents an ideal pretreated wood substrate as it contains a minimum content of lignin and other contaminants. To test whether the same high consistency hydrolysis approach can be applied to other substrates, an organosolv pretreated poplar (OPP) was prepared using pretreatment conditions described previously (Pan et al., 2006). The chemical composition of the organosolv pretreated poplar (OPP) is also shown in Table 1. The OPP contains approximately 87% cellulose with little xylan. The lignin content of OPP is slightly higher than that of UBHW. The most signicant difference between the UBHW and OPP is the signicant amount of acetone extractives detected in OPP. Most of these extractives are likely derived from low molecular weight phenolic compounds which are soluble in acetone. Previous work (Pan et al., 2006) suggested a higher lignin content in OPP substrate than that found in the current study. However, their lignin analysis was performed on samples without

solvent extraction where these phenolic compounds can be precipitated during acid hydrolysis and contribute to the total lignin content. 3.2. Hydrolysis of UBHW at 2% and 5% consistency in shake asks The hydrolysability of UBHW was rst determined in shake asks at 2% and 5% (w/w) consistencies. As shown in Fig. 1A, hydrolysis at 2% substrate consistency for 48 h resulted in a glucose concentration of about 17 g/L, while hydrolysis at 5% substrate consistency produced approximately 40 g/L glucose in the nal hydrolysate. When the percentage of cellulose-to-glucose conversion was determined (Fig. 1B), it was found that most of the cellulose present in 2% UBHW substrate was converted to glucose within 24 h of incubation with 20 FPU/g and 80 CBU/g of enzyme loading. Increasing substrate consistency to 5% led to a slightly lower cellulose-to-glucose conversion rate, approximately 95% after 48 h. This is probably due to end-product inhibition effects from the glucose and cellobiose produced (Xiao et al., 2004). 3.3. High consistency hydrolysis of unbleached hardwood pulp (UBHW) We next examined the possibility of hydrolyzing UBHW at different substrate consistencies, from 2% to 20% at 3% intervals (Ta-

A
Glucose concentration (g/L)

45 40 35 30 25 20 15

5% substrate consistency

2% substrate consistency
10 5 0 0 10 20 30 40 50

Hydrolysis time (hours)

B 120
Cellulose conversion (%)
Table 1 The chemical composition of unbleached hardwood pulp (UBHW) and organosolv pretreated poplar (OPP). Component Unbleached hardwood pulp (UBHW) Acetone extractives Cellulose (as glucan) Cellulose (as glucose) Xylan Lignin Acid soluble Acid in-soluble Organosolv pretreated poplar (OPP) Acetone extractives Cellulose (as glucan) Cellulose (as glucose) Xylan Lignin Acid soluble Acid in-soluble (Weight percentage %) 0.15 0.02 79.1 0.4 84.3 0.4 19.6 0.4 0.63 0.04 1.06 0.05 8.17 0.06 86.5 0.4 92.3 0.4 1.46 0.03 0.34 0.01 2.08 0.01

2% substrate consistency

100

80

5% substrate consistency

60

40

20

0 0 10 20 30 40 50

Hydrolysis time (hours)

Fig. 1. Enzymatic hydrolysis of UBHW (unbleached hardwood pulp) at 2% and 5% substrate consistencies in shake asks, based on (A) glucose concentration formed and (B) percent cellulose conversion.

X. Zhang et al. / Bioresource Technology 100 (2009) 58905897

5893

ble 2). It was observed that increasing pulp consistency resulted in a decrease in the amount of free water in the substrate matrix. The higher the initial consistency, the longer it takes to liquefy the substrate matrix (Table 2). In shake asks at 20% substrate consistency, the UBHW only liquees after 40 h incubation in the presence of cellulase enzymes. These experiments demonstrate that the shake-ask method is not suitable for evaluating high consistency hydrolysis of lignocellulosic feedstock. A peg mixer (Diagram 1), which is commonly used for oxygen delignication, was found to be capable of providing effective mixing of UBHW at high consistency. As shown in Table 2, it only took 1 h to liquefy UBHW at 20% consistency in the peg mixer compared to 40 h in a shake ask at the same enzyme loading. The hydrolysability of UBHW at 20% substrate consistency in the peg mixer was then evaluated. It was anticipated that a high substrate loading will raise the cellobiose concentration in the hydrolysate which will in turn elevate the end-product inhibition effects on cellulase enzymes (cellobiohydrolases and endoglucanases). Therefore, we chose a mixture of cellulase and b-glucosidase at an enzyme loading of 20 FPU of cellulase with 80 CBU of b-glucosidase per gram of cellulose. As shown in Fig. 2A, a signicant increase in glucose concentration was obtained during hydrolysis of UBHW at 20% consistency. The glucose content reached 144 g/L after 96 h of incubation. This is the highest glucose concentration ever reported from batch hydrolysis of a lignocellulose substrate. Hydrolysis of UBHW at 2% consistency was also carried out in the peg mixer (Fig. 2) to compare with the results obtained from shake ask experiments (Fig. 1). Similar hydrolysis proles (Fig. 2) were obtained and both attained 100% cellulose-to-glucose conversion after 24 h of incubation. However, at 20% substrate consistency, the cellulose-to-glucose conversion rate is only about 84% after 96 h of hydrolysis. Extending the hydrolysis to longer time resulted in little increase in glucose concentration (data not shown). Typically, cellulase hydrolysis of cellulose follows a two-phase curve, with an initial logarithmic phase and a subsequent asymptotic phase (Ramos et al., 1993). A number of factors contribute to the slow conversion rate in the later hydrolysis phase. Among these factors, end-products such as cellobiose and glucose were shown to play a major role in hindering hydrolysis (Tengborg et al., 2001). It is anticipated that the end-products inhibition effect will become severe at high substrate loading. A previous study has demonstrated that the presence of 100 g/L glucose in the hydrolysate can reduce the efciency of cellulase hydrolysis by 80% (Xiao et al., 2004). The lower conversion rate at 20% consistency compared to 2% is mainly due to the inhibition effects from the high glucose content in the hydrolysate (Xiao et al., 2004). 3.4. High consistency hydrolysis of organosolv pretreated poplar (OPP) The OPP was hydrolyzed at both 2% and 20% substrate consistencies under the same conditions as applied to UBHW in the peg mixer. As shown in Fig. 3, the OPP demonstrates a high hydrolysability at 2 % substrate consistency. The substrate released 16.8 g/L of glucose after 12 h enzyme hydrolysis (Fig. 3A) which represents 91% of the available cellulose (as glucose) in the OPP (Fig. 3B). A complete conversion of all the cellulose-to-glucose was obtained after 60 h of enzymatic hydrolysis. Hydrolysis of OPP at 20% substrate consistency yielded a signicantly higher glu-

A
Glucose concentration g/L)

160

20% substrate consistency

140 120 100 80 60 40 20 0 0 20 40 60 80 100

2% substrate consistency

Hydrolysis time (hours)

B 120
100

2% substrate consistency

Cellulose conversion (%)

80

60

20% substrate consistency

40

20

0 0 20 40 60 80 100

Hydrolysis time (hours)

Fig. 2. Enzymatic hydrolysis of UBHW (unbleached hardwood pulp) at 2% and 20% substrate consistencies in a peg mixer, based on (A) glucose concentration formed and (B) percent cellulose conversion.

cose concentration. The glucose content reached 158 g/L in the hydrolysate after 48 h of enzyme hydrolysis which is even higher than that obtained from UBHW. The amount of released glucose after 48 h of hydrolysis corresponded to a cellulose-to-glucose conversion of about 85%. There is little increase in sugar concentration beyond 48 h hydrolysis of OPP at 20% substrate consistency. High consistency hydrolysis of OPP at lower enzyme loadings, 3 and 10 FPU/g, was also tested (Fig. 4). It is apparent that decreasing enzyme loading led to reduced sugar concentration and lower cellulose-to-glucose conversion rate. However, a 50% reduction in enzyme dosage only reduced nal glucose concentration by about 21% (from 158.2 to 124.5 g/L). Even at 3 FPU/g of enzyme loading, it is still possible to obtain a glucose concentration above 80 g/L (Fig. 4A) with 43% cellulose-to-glucose conversion yield (Fig. 4B) after 96 h of hydrolysis. The hydrolysability of OPP at 30% substrate consistency was also determined (Fig. 5). As shown in Fig. 5, increasing OPP consistency from 20% to 30% resulted in a large increase in glucose con-

Table 2 The inuence of substrate consistencies on liquefaction time during hydrolysis of UBHW (unbleached hardwood pulp) in shake asks and peg mixer. Hydrolysis in shake asks Substrate consistency (%) Liquefaction time (h) 2 0 5 0 8 2.5 11 6 14 12 17 28 20 40 Hydrolysis in peg mixer 2 0 20 1

5894

X. Zhang et al. / Bioresource Technology 100 (2009) 58905897

A 180
160

20% substrate consistency

A
Glucose concentration (g/L)

180 160 140

20 FPU/g

Glucose concentration (g/L)

140 120 100 80 60 40

10 FPU/g
120 100 80 60 40 20 0

3 FPU/g

2% substrate consistency
20 0 0 20 40 60 80 100

20

40

60

80

100

Hydrolysis time (hours)

Hydrolysis time (hours)

B 120
Cellulose conversion (%)
100

B
2% substrate consistency
Cellulose conversion (%)

100

20 FPU/g
80

10 FPU/g
60

80

60

20% substrate consistency

3 FPU/g
40

40

20

20

0 0 20 40 60 80 100

0 0 20 40 60 80 100

Hydrolysis time (hours)

Fig. 3. Hydrolysis of organosolv pretreated poplar (OPP) at 2% and 20% substrate consistency in a peg mixer, based on (A) glucose concentration formed and (B) percent cellulose conversion.

Hydrolysis time (hours)

Fig. 4. Hydrolysis of organosolv pretreated poplar (OPP) at 20% substrate consistency in a peg mixer using 3, 10 and 20 FPU/g of cellulase loadings. (A) Glucose concentration formed and (B) percent cellulose conversion.

centration during the entire hydrolysis course with a nal glucose concentration of 21% at 96 h of hydrolysis. However, the celluloseto-glucose conversion yield is lower than that obtained at 20%. It should be mentioned that the initial OPP substrate has a solid content slightly below 30% (27%). Therefore, this substrate has to be dewatered by pressing prior to enzyme hydrolysis. Although opportunities exist to further increase high sugar concentration at higher substrate consistencies, 2030% is likely to be the practical substrate consistency range to carry out high consistency hydrolysis of lignocellulosic substrates, as most of the substrate would have solids content in this range after the pretreatment stage. 3.5. Fermentability of the hydrolysate obtained from high consistency hydrolysis of UBHW and OPP The hydrolysates obtained from hydrolysis of UBHW and OPP at 20% consistency represent the highest glucose concentrations that have been obtained from batch enzymatic hydrolysis of lignocelluloses. There has been little information on the fermentability of realistic hydrolysate with such high sugar concentrations. It can also be expected that high substrate loading may lead to an increased amount of potential inhibitors in the hydrolysates. Therefore, it is critical to determine how well yeast will ferment these hydrolysates. The hydrolysates obtained from 48 h hydrolysis of

240 220

Glucose concentration (g/L)

30 % substrate consistency

200 180 160 140 120 100 80 60 40 20 0 0 20 40 60 80 100

20% substrate consistency

Hydrolysis time (hours)

Fig. 5. Comparison of glucose concentrations obtained from hydrolysis of OPP at 20% and 30% substrate consistencies using 20 FPU/g of cellulose loading.

UBHW and OPP substrates at 20% consistency were collected and used for the subsequent fermentation experiments. Two glucose solutions were prepared as controls at the sugar concentrations present in UBHW and OPP hydrolysates. The initial glucose concen-

X. Zhang et al. / Bioresource Technology 100 (2009) 58905897

5895

tration of UBHW hydrolysate prior to fermentation was about 112 g/L and the control pure glucose solution was 110 g/L. The fermentation experiment was carried out for 96 h; the glucose reduction and ethanol production were determined during the fermentation. The yeast showed a high fermentability on pure glucose solution. Nearly all the sugars were metabolized after 12 h fermentation (Fig. 6). The ethanol production reached approximately 44 g/L at this time and started to level off (Fig. 7). The nal ethanol concentration (after 96 h) in the fermentation broth was 48.4 g/L which is about 86% of the theoretical glucose-to-ethanol conversion yield (based on a theoretical yield of 0.51 g of ethanol per gram of glucose). The yeast was also able to effectively utilize glucose in UBHW hydrolysate to produce a signicant amount of ethanol. Compared to glucose control, there was an initial lag phase observed in the glucose reduction and ethanol production during UBHW hydrolysate fermentation. The depletion of glucose occurred after 36 h of fermentation with an ethanol production of 46 g/L at the same time. The nal ethanol concentration (after 96 h) was 50.4 g/L which is 88% of the theoretical yield. The fermentation of OPP hydrolysate was then tested under the same conditions and compared to a control medium containing 150 g/L of pure glucose. The initial glucose concentration in OPP hydrolysate was about 149 g/L. The yeast again demonstrated a
120

good capability to ferment concentrated glucose solution. Almost all the glucose was used up within the initial 12 h of the fermentation with an ethanol production of nearly 60 g/L (Figs. 8 and 9). A higher nal ethanol concentration (after 96 h), 62.3 g/L, was obtained compared to the previous glucose control (Fig. 9). However the conversion yield was lower, 81% vs. 86% of the theoretical yield. Again, an initial lag phase was observed during OPP hydrolysate fermentation compared to the control medium with similar glucose content. The maximum ethanol production was achieved after 24 h of fermentation. The nal ethanol concentration (after 96 h) from fermenting OPP hydrolysate was 63.1 g/L which is equivalent to 83% of the theoretical yield.

4. Discussion High consistency hydrolysis and fermentation (20% and above) is a common practice in current starch-based bioethanol production. However, the characteristics of lignocellulosic substrates are different from starch-based feedstock. At low consistencies (<4%), the brous materials are suspended in abundant free water which makes the suspensions easy to be mixed and transferred. However, once the substrate consistency increases to 8%, a greater degree of

180 160

Glucose concentration (g/L)

Glucose concentration (g/L)

100

140 120 100 80 60 40 20 0

Pure glucose OPP hydrolysate

80

Pure glucose UBHW hydrolysate

60

40

20

0 0 20 40 60 80 100

10

20

30

40

50

60

70

80

Fermentation time (hours)

Fig. 6. The decrease in glucose concentration during fermentation of UBHW hydrolysate to ethanol by Saccharomyces cerevisiae.

Fermentation time (hours)

Fig. 8. The decrease in glucose concentration during fermentation of OPP (organosolv pretreated poplar) hydrolysate.

60

70

Ethanol concentration (g/L)

50

60

40

Ethanol concentration (g/L)

50 40 30 20 10 0 0 20 40 60 80 100

30

Pure glucose UBHW hydrolysate

Pure glucose OPP hydrolysate

20

10

0 0 20 40 60 80 100

Fermentation time (hours)

Fig. 7. The production of ethanol during Saccharomyces cerevisiae fermentation of UBHW (unbleached hardwood pulp) hydrolysate.

Fermentation time (hours)

Fig. 9. The production of ethanol during Saccharomyces cerevisiae fermentation of OPP (organosolv pretreated poplar) hydrolysate.

5896

X. Zhang et al. / Bioresource Technology 100 (2009) 58905897

brebre interactions occur, and this leads to a substantial increase in the strength of the bre network. As a result, the character of the suspension changes from one mass of bres in water to wet bre aggregates surrounded by gas (Duff and Titchener, 1975). This creates the so-called rheological problem during mixing and signicantly reduces the amount of free water available for hydrolysis. As shown in Table 2, a signicant increase in the liquefaction time was observed at higher substrate consistency in shake asks. It is apparent that effective mixing to break down the bre network is a critical step to achieving high consistency hydrolysis of lignocellulosic substrates. The peg mixer is conventional equipment employed in the pulp and paper industry to provide effective mixing of medium-consistency (815%) pulp. The results have shown that this equipment can greatly improve the liquefaction rate of UBHW substrate in the presence of cellulase. Using a peg mixer for high consistency hydrolysis resolved the technical issue related to mixing. During the preparation of this manuscript, two studies dealing with the topic of high consistency hydrolysis were published in the literature. Jorgensen and colleagues (Jorgensen et al., 2007) employed an in-house chamber to carry out liquefaction of wheat straw at 40% consistency. After liquefaction, the straw slurry was subjected to subsequent saccharication and fermentation in either SHF or SSF conguration. Forty percent substrate consistency is apparently the highest solid loading that has been attempted so far. However, in a practical bre processing industry (e.g. pulp and paper industry), a pulp consistency between 20% and 25% is typically encountered. Therefore, we chose to use 20% substrate consistency to examine hydrolysability of wood substrate in a peg mixer. As shown in Diagram 1, it appears that the peg mixer can provide a similar mixing effect as the equipment designed by the Danish group; both apply a rotating shaft to break down bre oc and disintegrate bre networks. Although in our study a lower substrate consistency was used, signicantly higher glucose concentrations were obtained from hydrolysis of UBHW and OPP, at 144 g/L (or 144 g/kg), 158 g/L (or 158 g/kg) respectively after 96 h at 20 FPU of cellulase loading. The density is estimated at 1.06 kg/L based on measurement of a pure glucose solution of 150 g/L. While low enzyme loadings, 3 and 10 FPU/g of cellulose, still yield 80.5 and 124.5 g/L of glucose in the nal hydrolysates. In Jorgensens study, 86 and 76 g/kg glucose were produced by hydrolyzing wheat straw at 40% and 20% substrate consistency, respectively. The enzyme loading used in Jorgensens study (7 FPU per gram of dry matter) is equivalent to 13.5 FPU per gram of cellulose as the substrate has a cellulose content of 52%. A high b-glucosidase dosage was supplied (ratio of 5:1 between CBU and FPU) in Jorgensens study, while in our study, a CBU:FPU ratio of 4:1 was applied. In another recent study, Cara et al. (2007) carried out enzymatic hydrolysis of pretreated olive tree biomass at a substrate consistency of up to 30%. The study reported the production of 73 g/L of glucose after 72 h hydrolysis of delignied LHW (liquid hot water)-pretreated olive tree biomass. Caras study employed 15 FPU cellulase with 15 CBU of b-glucosidase per gram of substrate on delignied LHW-pretreated olive tree biomass, which has a cellulose content of 56.7%. It is acknowledged that the substrates used in these two studies are different from UBHW and OPP which contain higher cellulose content. Comparing UBHW and OPP, it was surprising to nd that OPP demonstrated a better hydrolysability at high consistency than UBHW. The initial high cellulose content likely contributed to the high glucose concentration observed during OPP hydrolysis. Although, a similar cellulose-to-glucose yield (84% vs. 85%) was obtained after hydrolyzing UBHW and OPP at high consistency for 96 h, OPP demonstrated a higher initial reaction rate during hydrolysis. The initial velocity (Vi) calculated based on the reaction rate obtained from the rst hour of hydrolysis of OPP, is 0.204 g/g/h

(gram of glucose produced from 1 g of cellulose in an hour), whereas the Vi from UBHW is 0.146 g/g/h. In order to understand this difference, we further measured the viscosity and determined the DP (degree of polymerization) of cellulose present in both substrates. It was found that OPP cellulose has extremely low viscosity (2.67 mPa s) and DP (207), while UBHW cellulose has a viscosity of 40.3 mPa s and a DP of 1643. OPP was obtained from pulping wood chips at high temperature under acidic conditions. It is likely that this pretreatment signicantly degraded the cellulose macromolecules making them susceptible to cellulase hydrolysis. However, from the results obtained, it was not apparent how the xylan in UBHW and acetone extractives in OPP affected hydrolysability of the substrates at high consistency. The effect of these components on high consistency hydrolysis is currently under investigation. High consistency hydrolysis not only signicantly reduces the capital cost for installation of a hydrolysis vessel; more importantly it also provides a concentrated glucose stream for the subsequent fermentation which will lead to a signicant savings in the distillation cost. A previous study indicated that increasing glucose concentrations from 1.5% to 16% can lead to a sixfold reduction in the distillation cost after fermentation Zacchi and Axelsson (1989)). S. cerevisiae can typically tolerate a high ethanol concentration up to 180 g/L (Lin and Tanaka, 2006). As seen from the results, this industrial adapted S. cerevisiae effectively fermented the two pure glucose controls and assimilated most of the glucose within 12 h. The increase in initial glucose concentration from 110 g/L (pure glucose control for UBHW) to 150 g/L (pure glucose control for OPP) lowered the nal ethanol yield from 0.44 g/g (gram of ethanol per gram of glucose) to 0.42 g/g which is probably due to the inhibition effect from initial glucose concentration (Thatipamala et al., 1992). It was found that substrate (glucose) inhibition starts to have a signicant impact on fermentation yield at glucose concentrations over 150 g/L. The overall product yield is around 0.45 when the substrate concentration is below 150 g/L. However the fermentation yield decreases linearly with in the increase in substrate concentration once the substrate concentration exceeds 150 g/L (Thatipamala et al., 1992). A slower initial rate compared to pure glucose controls was observed during fermentation of lignocellulosic hydrolysates. This is probably due to the initial adjustment of yeast cells to the hydrolysate conditions. The amount of potential inhibitors, including acetic acid, phenolic compounds, furfural and hydroxymethylfurfural were determined. As shown in Table 3, there is an appreciable amount of acetic acid and phenolic compounds present in both hydrolysates with negligible amounts of furfural and hydroxymethylfurfural. However, the nal ethanol yield obtained from the two hydrolysates was not compromised by the presence of these inhibitors. In fact the ethanol concentration obtained from this study is the highest that has been ever reported in the literature from lignocellulose based feedstock. It is also interesting to note that, although the OPP hydrolysate has a higher acetic acid and total phenolic content than UBHW, its exhibits a superior fermentability. For example, the maximum ethanol yield was obtained after 24 h of fermentation of OPP hydrolysate, while it took 36 h to reach to ethanol production peak in UBHW hydrolysate. The organic acids and phenolic compounds are generally considered as inhibitors to yeast. However, the concentration effect of these compounds on yeast fermentation is still under debate. For example, the presence of
Table 3 The amount of potential inhibitory compounds present in UBHW (unbleached hardwood pulp) and OPP (organosolv pretreated poplar) hydrolysates. Concentration Acetic acid (mmol/L) Total phenolic (g/L) UBHW hydrolysate 53.6 2.1 OPP hydrolysate 109.4 5.2

X. Zhang et al. / Bioresource Technology 100 (2009) 58905897

5897

100 mmol/L of acetic acid in the media was shown to increase rather than decrease the ethanol yield from approximately 0.41 0.45 g/g by S. cerevisiae (Larsson et al., 1999). Also different types of phenolic compounds could exhibit different effects on S. cerevisiae fermentation (Palmqvist and Hahn-Hagerdal, 2000). As mentioned, the S. cerevisiae strain used in current study has been adapted to spent sulphite liquor which has a high phenolic compounds and acetic acid content. Therefore, it is not surprising that the yeast can generate a high ethanol yield from sugars in the two hydrolysates which have a relatively low phenolic compounds and acetic acid content compared to typical spent sulphite liquor. 5. Conclusions We have demonstrated that a peg mixer, commonly employed in pulping processes, can be used for successful high consistency hydrolysis of lignocellulosic substrates. Hydrolysis of unbleached hardwood pulp (UBHW) and organosolv pretreated poplar (OPP) at 20% substrate consistency led to a high glucose concentration in the hydrolysate. Enzymatic hydrolysis of OPP for 48 h resulted in a hydrolysate with 158 g/L of glucose content. This is the highest glucose concentration that has been obtained from enzymatic hydrolysis of lignocellulosic substrate. Fermentation of UBHW and OPP hydrolysates with high glucose content led to high ethanol concentrations in the nal fermentation broth, much higher than those reported in previous literature. Applying pulping equipment, which is typically designed to handle high and medium-consistency brous materials, for high consistency hydrolysis provides a practical means to overcome the rheological problems encountered in laboratory shake ask experiments. This brings the biomass conversion research a step closer to industrial implementation. The high consistency hydrolysis and fermentation presents a new approach to lignocellulose hydrolysis and fermentation and opens up new windows to examine substrateenzyme interactions. Acknowledgement We would like to thank Novozymes North America Inc. for providing enzymes. Funding was provided by Natural Resources Canada and the National Science and Engineering Research Council of Canada.

References
Cara, C., Moya, M., Allesteros, I., Negro, M.J., Gonzalez, A., Ruiz, E., 2007. Inuence of solid loading on enzymatic hydrolysis of steam exploded or liquid hot water pretreated olive tree biomass. Process Biochemistry 42, 10031009. Duff, G.G., Titchener, A.L., 1975. The disruptive shear stress of pulp networks. Svensk. Papperstidning nr. 13, 474479. Galbe, M., Zacchi, G., 2002. A review of the production of ethanol from softwood. Applied Microbiology and Biotechnology 59, 618628. Jorgensen, H., Vibe-Pedersen, J., Larsen, J., Felby, C., 2007. Liquefaction of lignocellulose at high-solids concentrations. Biotechnology and Bioengineering 96, 862870. Larsson, S., Palmqvist, E., Hahn-Hagerdal, B., Tengborg, C., Stenberg, K., Zacchi, G., Nilvebrant, N.O., 1999. The generation of fermentation inhibitors during dilute acid hydrolysis of softwood. Enzyme and Microbial Technology 24, 151159. Lin, Y., Tanaka, S., 2006. Ethanol fermentation from biomass resources: current state and prospects. Applied Microbiology and Biotechnology 69, 627642. Palmqvist, E., Hahn-Hagerdal, B., 2000. Fermentation of lignocellulosic hydrolysates II: inhibitors and mechanisms of inhibition. Bioresource Technology 74, 2533. Pan, X.J., Gilkes, N., Kadla, J., Pye, K., Saka, S., Gregg, D., Ehara, K., Xie, D., Lam, D., Saddler, J., 2006. Bioconversion of hybrid poplar to ethanol and co-products using an organosolv fractionation process: optimization of process yields. Biotechnology and Bioengineering 94, 851861. Ramos, L.P., Breuil, C., Saddler, J.N., 1993. The use of enzyme recycling and the inuence of sugar accumulation on cellulose hydrolysis by Trichoderma cellulases. Enzyme and Microbial Technology 15, 1925. Robinson, J., Keating, J., Boussaid, A., Manseld, S.D., Saddler, J.N., 2002. The inuence of bark on the fermentation of Douglas-r whitewood prehydrolysates. Applied Microbiology and Biotechnology 59, 443448. Stenberg, K., Bollok, M., Reczey, K., Galbe, M., Zacchi, G., 2000. Effect of substrate and cellulase concentration on simultaneous saccharication and fermentation of steam-pretreated softwood for ethanol production. Biotechnology and Bioengineering 68, 204210. Sun, Y., Cheng, J.Y., 2002. Hydrolysis of lignocellulosic materials for ethanol production: a review. Bioresource Technology 83, 111. Tengborg, C., Galbe, M., Zacchi, G., 2001. Reduced inhibition of enzymatic hydrolysis of steam-pretreated softwood. Enzyme and Microbial Technology 28, 835844. Thatipamala, R., Rohani, S., Hill, G.A., 1992. Effects of high product and substrate inhibitions on the kinetics and biomass and product yields during ethanol batch fermentation. Biotechnology and Bioengineering 40, 289297. Van Wyk, J., 2001. Biotechnology and the utilization of biowaste as a resource for bioproduct development. Trends in Biotechnology 19, 172177. Wingren, A., Galbe, M., Zacchi, G., 2003. Techno-economic evaluation of producing ethanol from softwood: comparison of SSF and SHF and identication of bottlenecks. Biotechnology Progress 19, 11091117. Wyman, C.E., Hinman, N.D., 1990. Ethanol fundamentals of production from renewable feedstocks and use as a transportation fuel. Applied Biochemistry and Biotechnology 24 (5), 735753. Xiao, Z.Z., Zhang, X., Gregg, D.J., Saddler, J.N., 2004. Effects of sugar inhibition on cellulases and beta-glucosidase during enzymatic hydrolysis of softwood substrates. Applied Biochemistry and Biotechnology 113 (16), 11151126. Zacchi, G., Axelsson, A., 1989. Economic evaluation of preconcentration in production of ethanol from dilute sugar solutions. Biotechnology and Bioengineering 34, 223233.