Biohydrogen production from wheat straw hydrolysate using Caldicellulosiruptor saccharolyticus followed by biogas production in a two-step uncoupled process

.


Introduction
Foreseeing the advancements in the energy infrastructure and end-user technologies, the growth in world energy consumption can be expected to slow down [1].However, the supply of fossil fuels is expected to hit rock-bottom in coming decades [1].Moreover, un-restrained usage of fossil fuels has contributed to growing concern over global warming.Hence, it is more than evident that the world needs alternative, renewable energy sources which should also be environmental friendly.
Of late, agricultural residues are increasingly being considered as a potential source of renewable biomass.Estimations of agricultural residues are about 10 10 tons/year globally, corresponding to 4.7 Â 10 10 GJ of energy (about 9% of the global energy consumption in 2008 [1]), and about twothirds consists of cereal residues [2].Wheat straw is a lignocellulosic biomass, consisting of 35e40% cellulose, 20e30% hemicelluloses and 8e15% lignin [3].These sugars can potentially be used in microbial fermentations to produce biofuels, such as, bioethanol, biogas and hydrogen.So far, however, bioethanol production from lignocellulosic biomass has not been successful enough due to a variety of techno-economic challenges [4e6].Alternatively, studies have shown efficient production of hydrogen (H 2 ) from wheat straw hydrolysate (WSH) by dark fermentation (DF) [7e9].H 2 is widely considered as a fuel of the future due to its properties of rapid burning speed, no emissions of greenhouse gases, higher energy density, low minimum ignition energy and a very high research octane number [10e13].Currently, H 2 is mainly produced by reforming fossil fuels making it a non-renewable and noncarbon neutral, which is in contrast to what DF of agricultural residues has to offer.The thermophilic Caldicellulosiruptor saccharolyticus possesses the ability of producing H 2 via DF at yields near the theoretical maximum of 4 mol H 2 /mol of hexose consumed [14].In addition, C. saccharolyticus can naturally ferment a wide range of poly-, oligo-and mono-saccharides including sugars present in lignocellulosic hydrolysate [15].Moreover, the absence of 'carbon catabolite repression' enables it to co-ferment glucose, xylose and arabinose among other sugars [16].
During the DF, the highest theoretical maximum yield of H 2 can be obtained only when acetate is the major by-product [17].The latter, contains as much as 67% of the total energy present in the substrate.This energy can be retrieved in the form of H 2 by either photo-biological process or microbial electrolysis, which are both, however, still under development [18].Alternatively, the effluent from DF can be transferred to an anaerobic digester, wherein acetate can be converted to CH 4 by acetoclastic methanogenesis, which is a reliable and an industrially established process [3,18].Various studies of combined H 2 and CH 4 production in a two-step process have been reported in recent years [9,19].Furthermore, H 2 and CH 4 together can give a mixture termed hythane, which has superior combustion properties compared to CH 4 alone [20].
So far, DF has been carried out largely in a continuously stirred tank reactor (CSTR), in which sparging is needed to actively remove hydrogen to keep the hydrogen partial pressure ðp H2 Þ to a minimum [21,22].Nitrogen is usually used for sparging at lab-scale, as it is a cheap and inert gas.However, separation of N 2 from H 2 is tedious and thus not exploitable at industrial scale.As an alternative, CO 2 is relatively easier to separate from H 2 , but has a detrimental effect on growth of C. saccharolyticus [23].Finally, the CH 4 produced in the anaerobic digestion (AD) can, in principle, be used as sparging gas in the DF, producing hythane, after removal of CO 2 .
The ability of C. saccharolyticus to ferment wheat straw was observed previously [7].However, since the experiments were performed on raw wheat straw, they were continued for long duration (about 45 days [7]), which makes it economically unfeasible.On the other hand, various pretreatment methods can generate by-products which may inhibit microbial growth [24,25].Hence, in this study, we demonstrate the fermentability of pre-treated wheat straw by C. saccharolyticus and its ability to sustain growth in the presence of CH 4 .We also demonstrate the feasibility of the two-step process, wherein, the wheat straw hydrolysate (WSH) is fermented to produce H 2 in a CSTR by C. saccharolyticus and the effluent produced is converted to CH 4 by methanogens in a UASB reactor.During this study, the reactors performing DF and AD were uncoupled.Ideally, however, both the reactors should be coupled together as described previously [26].

2.
Materials and methods

Wheat straw hydrolysate
WSH was produced by steam acid pretreatment and enzymatic hydrolysis of wheat straw obtaining an energy content of 11.9 MJ/kg of dry matter (DM) in the WSH.Glucose and xylose were the main sugars and the chemical oxygen demand (COD) was estimated to be 196 g/l.The detailed composition of the hydrolysate has been reported previously [27].The pre-treated hydrolysate was centrifuged for 15 min at 4900 rpm to remove any remaining solid matter.Subsequently, the supernatant is then allowed to pass through a Whatman's no.1 filter paper supported by a nylon membrane to get rid of insoluble particulate matter.The pH of this clarified hydrolysate was adjusted to pH 7 with 12.5 M NaOH.The filtered neutral hydrolysate was sterilized by filtration using disposable Acrocapä (pore size e 0.2 mm) filters and the filtrate was collected in sterile screw cap bottles and stored at À20 C until further use.

Microorganism and culture medium
C. saccharolyticus DSM 8903 was purchased from the Deutsche Sammlung von Mikroorganismen und Zellkulturen (Braunschweig, Germany).A modified DSM 640 medium was used as a base medium for all cultivations throughout this work [23].Routine subcultures and inoculum development were conducted in 250 mL serum bottles containing 50 mL of medium under a N 2 atmosphere.Anoxic solutions of glucose, xylose and arabinose were autoclaved separately and were added to the sterile medium at the required concentration.Filter sterilized WSH was added to a sterile serum bottle and was kept under a N 2 atmosphere.

Experimental set-up and operation
Batch cultures of dark fermentation were carried out at 70 C using 250-mL serum flasks containing 50 mL liquid medium.The preparation of anaerobic flasks was as follows: the modified DSM 640 medium without the carbon source was added to the flasks and thereafter, the flasks were sealed with butyl stoppers and aluminium crimps.Subsequently, the headspace of the flasks was flushed with N 2 unless stated The chemostat cultures were carried out as described previously [22] except for the following modifications.In continuous mode, the reactor was fed with a fresh medium containing (per litre of deionised water) NH 4 Cl 0.9 g, MgCl 2 .7H 2 O 0.4 g, KH 2 PO 4 0.75 g, K 2 HPO 4 1.5 g, Yeast extract 1 g, resazurin 1 mg, trace element solution SL-10 [28] 1 mL and WSH (10% v/v) as a substrate but omitting cysteine-HCl.WSH at 10% v/v contained approximately 11 g/L of total monosaccharide sugars with 23 mg/L of 5-(hydroxymethyl)furfural (HMF) and 114 mg/L of furfural [27].The reactor was sparged with either 100% N 2 or a gas mixture containing N 2 þ CO 2 (60%:40% v/v) at the flow rate of 6 L/h.The steady states were obtained at four different conditions, i.e.Case I, low growth rate (D ¼ 0.05 h À1 ), N 2 sparging; Case II, higher growth rate (D ¼ 0.15 h À1 ), N 2 sparging; Case III, low growth rate (D ¼ 0.05 h À1 ), sparging with a mixture of N 2 (60% v/v) and CO 2 (40% v/v); and Case IV, higher growth rate (D ¼ 0.15 h À1 ), sparging with a mixture of N 2 (60% v/v) and CO 2 (40% v/v).The steady states were determined after at least five volume changes based on the stability of CO 2 and H 2 levels and biomass concentration.The effluent generated from the chemostat was collected, mixed together and stored at 4 C before use in AD.
Batch cultures of AD were performed in triplicates using the effluent from DF.The flasks were incubated at 37 C for 31 days.The experimental procedure and set-up was as described earlier [27,29].Methane production using the effluent of dark fermentation was performed in UASB reactors in duplicate and under mesophilic (37 C) conditions.The active reactor volume was 0.8 L and the up-flow velocity was 0.08 and 0.09 mL/h.The rest of the reactor configuration was as previously described [30].A modified basic anaerobic nutrient solution (BA) was used to supplement the effluent [31], in that, ammonium chloride was substituted with Urea (1 g/L), as the latter is a rich nitrogen source and also a buffering agent.The effluent collected from DF had a pH of 6.6 and a COD of 16.2 g/l before addition of the BA medium.After addition of the BA medium, the pH and the COD changed to 6.9 and 15.3 g/l, respectively (Table 2).Prior to the treatment of the DF effluent, the UASB reactor was continuously fed with the WSH containing about 10 g/l of fermentable sugars.When the feed was switched to DF effluent, the reactors were operated at an OLR of 5.0 g COD/L/day (HRT of 2 days) until they reached stability.Increase in the organic loading rate was performed by decreasing the hydraulic retention time (HRT).The HRT was decreased from 2.5 to 1.5 days and corresponded to an increase in OLR of 6.0e10.5 g COD/L/day.The treatment period was 49 days.

Analytical methods
For dark fermentation, gas in the headspace of the serum flasks and the CSTR was analysed for CO 2 and H 2 by gas chromatography, using a dual channel Micro-GC (CP-4900; Varian gas chromatography, Middelburg, The Netherlands), as previously described [28].The results were analysed with a Galaxie Chromatography workstation (v 1.9.3.2).The optical density of the culture was measured at 620 nm using a U-1000 spectrophotometer (Hitachi, Tokyo, Japan).The cell-free culture medium was used as a blank while measuring the optical density of the cultures.The cell dry weight was determined as previously described [32].The metabolites, sugars, 5-(hydroxymethyl)furfural and furfural in DF were analysed by HPLC (Waters, Milford, MA, USA) as described previously [22].The samples collected during anaerobic digestion were analyzed for pH, COD, NH 4 þ eN, partial and total alkalinity, volatile fatty acids, gas volume and composition.Methods of sample collection and analysis for the methane potential batch test and UASB reactor were as previously described [27].The volume of methane and hydrogen were corrected for using the standard conditions (0 C, 1 atm).

Calculations
The volumetric H 2 productivity (mM/h) was calculated using the ideal gas law and the H 2 and CO 2 concentrations in the headspace of the serum flasks or CSTR.In case of the CSTR, the calculations were based on the flow rate of the effluent gas and the accompanying partial pressures of H 2 and CO 2 .In case of serum flasks, the product gas was allowed to accumulate in the headspace, which is the basis for the calculation.

Fermentability of wheat straw hydrolysate in DF
Media containing 10% or lower levels of WSH showed comparable biomass and H 2 yields, (Fig. 1(A)).Even though, the differences observed were insignificant, yet a decreasing trend can be observed in maximum obtainable H 2 productivities with increasing WSH concentration (Fig. 1(A)).Hardly any or no significant growth and H 2 accumulation was observed in the flasks containing 20% WSH (data not shown).Interestingly, H 2 accumulation and cell growth appears to be enhanced in WSH compared to a medium with only pure sugars (Figs.1(A) and 2).For obvious reasons, 10% v/v of WSH was added in a growth medium used in further experiments.

3.2.
Growth of C. saccharolyticus in presence of methane H 2 productivities and biomass yield seemed to be unaffected by CH 4 (Figs.1(B) and 2).Interestingly, the flasks containing 100% CH 4 in the headspace appeared to have slightly higher H 2 yields compared to those containing 100% N 2 in the headspace (Figs.1(B) and 2).Yet again, the flasks containing WSH showed relatively better biomass formation and H 2 accumulation at a higher maximum growth rate than those containing only pure sugars (Figs.1(B) and 2).All batch experiments displayed coconsumption of glucose, xylose and arabinose.However, xylose was the most preferred substrate regardless of the growth conditions (Fig. 2).
Although, CH 4 is slightly beneficial; for safety reasons, N 2 was used in all following experiments instead, as both do not affect the performance of C. saccharolyticus negatively.Thus, the gas mixture of N 2 þ CO 2 was assumed to mimic the nonupgraded flue gas (CH 4 þ CO 2 ) from the AD (Case III and IV).Similarly, cultures sparged with N 2 were assumed to be the same as if sparged with CH 4 (Case I and II ).

Growth of C. saccharolyticus on WSH in controlled bioreactors
In chemostats, four different experimental conditions were employed (using the growth rate and sparging gas composition as variables, Cases I to IV), with a medium containing 10% WSH as carbon source.Out of the four conditions studied, a low growth rate (D ¼ 0.05 h À1 ) and sparging the reactor with N 2 resulted in the highest H 2 yield and best of substrate conversions (Table 1).The substrate conversion efficiency decreased increasing growth rate and when CO 2 was present in the sparging gas.Surprisingly, at a higher growth rate (D ¼ 0.15 h À1 ), the culture sparged with N 2 þ CO 2 displayed a higher H 2 yield and higher specific H 2 production rate than the one sparged with N 2 (Table 1).Also, the highest lactate yield per mole of hexose was observed in the latter case compared to the other conditions.However, the average volumetric H 2 productivity was about 40% higher in the reactors sparged with N 2 only (Table 1, 5.1 L H 2 /L/day) than the reactors sparged with N 2 þ CO 2 (Table 1, 2.9 L H 2 /L/day).The overall conversion of substrate in the dark fermentation was found to be in the range of 19e88% (Table 1).Regardless of the growth conditions the culture was able to reduce the potential growth inhibitors (5-(hydroxymethyl)furfural and furfural) present in the WSH (Table 1).Cultures sparged with N 2 þ CO 2 displayed higher medium osmolalities than their counterparts performed with N 2 sparging (Table 1).Similarly, low amounts of biomass were obtained in chemostats sparged with N 2 þ CO 2 which were accompanied by higher amounts of residual sugars and consequently lower conversions.The specific consumption rate for xylose was significantly higher than that for glucose in the cultures sparged with N 2 þ CO 2 (Case III and IV, Table 1), whereas the opposite was true for the cultures sparged with N 2 (Case I and II, Table 1).Carbon and redox recovery was significantly higher than 100% in all the cases studied (Table 1).

3.4.
Production of methane from the effluent collected from DF During anaerobic digestion of the collected DF effluent, an increase in the organic loading rate from 6.0 to 10.5 g COD/L/ day resulted in an increase in methane productivity (Table 2).Further increase in the organic loading rate to 15.4 g COD/L/ day (1.0 day HRT) resulted in an increased methane production rate, i.e. 3.95 L/L/day, after 6 days of treatment time (data not shown).At a stable organic loading rate of 10.5 g COD/L/ day (equivalent to 1.5 days HRT) a maximum methane production rate of 2.64 L/L/day (Table 2) was observed.The methane yield ranged from 0.28 to 0.26 L/g COD independent of the OLR and the methane content in biogas was about 60% (Table 2).
Stable operational conditions prevailed throughout the entire treatment period.The pH remained stable at around 7.50 for all applied OLRs.The effluent of the UASB reactor   contained low concentrations of COD (<1 g/L) and volatile fatty acids (<0.1 g/L).Furthermore, the COD of the medium fed to the UASB reactor was reduced by approx.95% after the treatment.Addition of modified anaerobic medium resulted in a need of a high reactor buffer capacity, which was reflected in the partial alkalinity that ranged from 5.4 to 5.8 g/L.The concentration of the buffer species NH 4 þ eN, in the reactor varied from 0.66 to 0.74 g/L as a consequence of urea mineralization (Table 2).

Overall energy output
On average, about 50% of the energy in wheat straw has been retrieved across all the scenarios of the hythane process.The energy output from DF was highest for Case I and lowest for Case IV.Although, the composition of effluent generated during different Cases of DF was different, due to the mixing of all the effluent together before its treatment, a scenario-specific energy output could not be determined for AD.Hence, a maximum energy output observed during AD was assumed to be true in all the scenarios of hythane (Table 3), which was significantly higher than the energy output from any of the DF Cases (Table 3).About 85% of the overall energy present in straw is contained in the sugars, of which 60% (average of all hythane scenarios, Table 3) has been successfully retrieved in the form of H 2 and CH 4 in the present hythane process.

Dark fermentation
In this study, C. saccharolyticus was successfully cultured on WSH, provided that the concentration of WSH is less than 20% (v/v).C. saccharolyticus has been seen previously to grow efficiently on hydrolysates of wheat straw and Miscanthus, juices of sweet sorghum and sugar beet as well as on raw feedstocks, such as, maize leaves, Silphium trifoliatum leaves, potato peels, carrot pulp and paper sludge [34e39].C. saccharolyticus has been observed to sustain growth in a medium containing up to 2 g/L of common growth inhibitors found in WSH, viz., 5-  i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 8 ( 2 0 1 3 ) 9 1 2 1 e9 1 3 0 (hydroxymethyl)furfural and/or furfural [34].However, the concentrations of these inhibitors in the WSH used in this study were far below 2 g/L [27].On the other hand, the osmolality of the medium containing 20% WSH was found to be about 0.26 Osmol/kg of H 2 O, which is well above the critical osmolality, i.e. 0.22 Osmol/kg of H 2 O, reported for substantial growth inhibition in a growing culture of C. saccharolyticus [23].
Hence, the inability of C. saccharolyticus to initiate growth on higher concentrated WSH is related to its limited osmotolerance.
The results herein revealed that C. saccharolyticus is as unaffected by CH 4 as by N 2 .To our knowledge no information is available in the literature about the ability of thermophiles like C. saccharolyticus to grow in the presence of CH 4 .Performance on WSH (10% v/v) was slightly better than on artificial medium, which might be due to the presence of marginal amounts of soluble proteins and amino acids in WSH [8,9,27].No obvious explanation could be found for the observed slight beneficiary effect of the presence of CH 4 compared to N 2 (Figs.1(B) and 2).Nevertheless, it strongly suggests that sparging with upgraded CH 4 can be an appropriate alternative.However, to obtain purified CH 4 , CO 2 should be removed from the flue gas of the AD reactor, which will incur significant additional costs.To reduce these costs, the DF reactor can be sparged with the non-upgraded flue gas of the AD reactor i.e. mixture of CH 4 and CO 2 .In addition, C. saccharolyticus can sustain growth in non-sparging conditions in the reactor [22], which opens an opportunity to alleviate the costs of sparging.However, H 2 yields obtained in the absence of sparging are much lower due to formation of more undesirable byproducts such as lactic acid, which is also not a preferred substrate for acetoclastic methanogenesis in AD [40,41].Hence, absence of sparging in the DF reactor can affect both DF and AD.A thorough techno-economic evaluation of the entire process may conclude the best applicable alternative.
The maximum overall H 2 productivities observed in the hythane scenario (Case I, Table 1) is at least five times higher than the average H 2 productivity reported by Kongjan et al. [9].Moreover, the productivities observed in all the Cases in this study are comparable to previously reported values for C. saccharolyticus, ranging from 2.3 to 9.7 L of H 2 /L/day, the highest of which was achieved when hydrolysed potato steam peels were used as a substrate [14,34e38].The observation of significantly lower H 2 yield in Case II may have been due to overflow metabolism, i.e. high glycolytic flux causing a metabolic shift at the pyruvate node to lactate formation.Overall, the combination of low biomass, volumetric H 2 productivity and sugar conversion efficiency of cultures sparged with N 2 þ CO 2 clearly illustrate the dramatic effect of CO 2 in the sparging gas (Case III and IV, Table 1).A previous investigation on the effect of sparging with CO 2 in C. saccharolyticus cultures [23], revealed that the inherent formation of bicarbonate increased the osmotic potential to critical levels.As a consequence, extensive cell lysis occurs in the culture resulting in higher protein and DNA concentration in the culture broth [23].Nevertheless, this nutrient-rich lysate might benefit the growth of the remaining cells, therefore displaying higher specific H 2 production rates observed in cultures sparged with CO 2 (Case III and IV, Table 1).Alternatively, the observation of CO 2 stimulating growth of C. saccharolyticus on xylose [42] might have improved specific H 2 productivity in Case III and IV.
None of the Cases studied showed complete consumption of sugars which could indicate a limitation of an essential nutrient.It can be argued that it might be sulphur.Firstly, phosphoric acid (H 3 PO 4 ), instead of sulphuric acid (H 2 SO 4 ), was used in the mild acid pretreatment of wheat straw used in this study, thus eliminating a potential sulphur source from the medium [27].Secondly, the influents of all DF cases were supplemented with yeast extract as the only sulphur source.With a minimal concentration of 1 g/L it may not have provided adequate amounts of sulfur.Finally, wheat straw itself contains very negligible amounts of sulfur [43].However, further experiments are needed to explore this hypothesis as they were out of the scope of this study.
The higher carbon and electron (redox) recovery observed in all the cases may have been due to traces of non-hydrolyzed disaccharides and/or oligosaccharides in WSH.This also may have resulted in a possible overestimation of H 2 yields in the respective cases.

Anaerobic digestion of the effluent collected from DF
The maximum methane production rate obtained during anaerobic digestion of the DF effluent collected from a H 2 producing CSTR during this study is significantly higher than a previously reported value (2.1 L CH 4 /L/day) in a similar study where DF effluent was collected from a H 2 producing UASB Scenario: b Since the effluent collected from different Cases of DF was mixed before its treatment in AD, the energy output for the latter was assumed constant in all the scenarios in this study.c A reference case scenario wherein WSH was directly fed to an AD reactor [27].
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 8 ( 2 0 1 3 ) 9 1 2 1 e9 1 3 0 reactor [9].This might be related to the differences in composition of DF effluent, as: i) the DF effluent collected during this study contained mainly acetate whereas, its counterpart in the previous study contained significant amounts of butyrate, propionate and ethanol, along with acetate [9], and ii) acetoclastic methanogens take acetate as a substrate and rely on acetogens for the conversion of butyrate, propionate and ethanol to acetate [40,41].In another study [27], WSH was directly fed to a methanogenic UASB reactor at an OLR of 10.2 g COD/L/day producing methane at a production rate (2.7 L CH 4 / L/day) comparable with the one reported in the present study.So far, sustained organic loading rates up to 15 g COD/L/day have been reported in the treatment of DF effluents in a UASB reactor [3,9,44,45].However, applications of OLRs higher than 15 g COD/L/day were observed to result in accumulation of volatile fatty acids, low COD reductions and low CH 4 yields.In addition, very high OLRs generate vigorous gas production rates, thus inflicting instability to the granular bed and eventually leading to process failure [45].Due to a decrease in methane yield and slight increase in VFA accumulation at higher OLR (10.2 g COD/L/day, Table 2) further increase in OLR was abandoned in this study.
A stable pH within the range of 7e8 has been reported as optimum for acetoclastic methanogenesis [9].Consumption of VFA during AD may have contributed to a pH increase to a suitable range.
Granular anaerobic sludge is known to be more protective for methanogens against inhibitory compounds than liquid granular sludge [46].This could be a reason why batch tests of AD using liquid anaerobic sludge resulted in lower CH 4 yields on DF effluent (w0.22 L CH 4 /g COD) than obtained from effluent treated in the UASB reactor with granular anaerobic sludge (Table 2).

4.3.
Overall energy output and the potential of the process The overall energy yield obtained during this study (average of all hythane scenarios), i.e. approximately 2010 kJ/L of WSH, was about four times higher than the stable overall energy yield reported earlier for a similar study (440 kJ/L of WSH, estimated from Ref. [9]).Thus, in comparison, this study reports a very efficient process with respect to overall energy output.However, in the study performed by Kongjan et al. [9], the total sugar concentration in the culture medium was about twice lower than in this study, which resulted in comparatively lower H 2 and CH 4 yields per litre of WSH and consequently a lower energy yield.
Another study on biohydrogen production from WSH reports an energy yield of 0.96 kJ/g of wheat straw (estimated from Refs.[3,8]) which is two-folds lower than the energy yields obtained in Case I (Table 3) of the DF phase studied herein.In the present study, the overall conversion efficiency for a hythane process i.e. 60% could not match the high conversion efficiency i.e. 71% obtained in a study pertaining to production of biogas using WSH (Table 3 [27],).However, the former will be advantageous, if the aim is to produce hythane.
About 85% of the energy in wheat straw can be retrieved in the form of soluble sugars (Table 3).Although, reasonably high substrate conversion efficiencies can be achieved during DF and AD using the soluble sugars in WSH; the possible losses of sugars during the extensive pre-treatment process can result in much lower overall energy yields (Tables 1e3).Hence, an efficient pre-treatment process is of paramount importance for any hythane-like process.
In the current study, the AD expending about five-folds more process time than DF (1.5 days for AD and 0.28 days for DF), will consequently require reactors with five-folds more volumetric capacity than DF.Reactors with higher volumetric capacity will incur higher capital and operational costs.This can be conveniently avoided simply by operating DF reactors at high HRT (preferably similar to that of AD), which may also aid in achieving higher conversion during DF (Table 1 and 3).
Overall, the process offers a number of benefits with respect to convenience in operation and cost, i) a thermophilic DF process offers less risk of contamination by H 2 -oxidising methanogens in the DF reactor [47], ii) the contaminants can also be kept out of the DF reactor by operating it at relatively higher growth rate [8] and iii) the process can successfully retrieve about 57% of the energy present in wheat straw.More technical details of the process and possible ways of cost reduction have been extensively discussed elsewhere [48].

Conclusions
C. saccharolyticus can efficiently produce H 2 from sugars in WSH.The residual sugars and acids produced can subsequently be converted to CH 4 in a methanogenic UASB reactor.The two-step process gives reasonable conversion efficiencies (about 67% of energy in the sugar fraction of wheat straw), but there remains room for further improvement.Moreover, the performance of C. saccharolyticus is not affected by CH 4 allowing application of this gas for sparging the hydrogenogenic reactor.However, a further extensive techno-economic evaluation is required to determine the best DF set up out of the following scenarios: i) sparging with upgraded CH 4 , ii) sparging with the non-upgraded flue gas from the AD reactor, or iii) no sparging.
An optimized and economically feasible version of this process can potentially complement a bio-refinery, wherein, along with bio-energy other value-added products are also produced from any unutilized parts of renewable agricultural biomass.This study paves a way for further exploration to determine whether a biological hythane process can be a viable alternative for the conversion of lignocellulosic biomass.
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 8 ( 2 0 1 3 ) 9 1 2 1 e9 1 3 0 The energy output for each of the cases was calculated based on lower calorific values (LCV) and the quantity of H 2 or CH 4 produced.The LCV for H 2 and CH 4 are 122 and 50.1 MJ/kg, respectively [33].

Table 2 e
Treatment of dark fermentation effluent in a UASB reactor.
b MPR, methane production rate.

Table 1 e
Results of the continuous fermentations of wheat straw hydrolysate by C. saccharolyticus.Three values for three sugars, i.e. glucose, xylose and arabinose respectively.b ðQ H2 Þ, volumetric hydrogen productivity.c ðq H2 Þ, specific hydrogen productivity.d q sugar , specific sugar consumption rate.e Osmolality was measured in Osmol/kgH 2 O.

Table 3 e
[27]gy output in all scenarios compared with reference scenario.Values for energy contained in wheat straw (19.1 kJ/g) and in its sugar fraction (16.3 kJ/g) were obtained from Kaparaju et al.[3]and Nkemka et al.[27]respectively.