Bioresource Technology 201 (2016)182-190 Contents lists available at ScienceDirect BIORESOURCE TECHNOLOGY Bioresource Technology ELSEVIER journal homepage:www.elsevier.com/locate/biortech Effect of thermal,acid,alkaline and alkaline-peroxide pretreatments on CrossMark the biochemical methane potential and kinetics of the anaerobic digestion of wheat straw and sugarcane bagasse Silvia Bolado-Rodriguez*,Cristina Toquero,Judit Martin-Juarez,Rodolfo Travaini, Pedro Antonio Garcia-Encina Department of Chemical Engineering and Environmental Technology.University of Valladolid.Calle Doctor Mergelina s/n,47011 Valladolid,Spain HIGHLIGHTS Highest methane productions were obtained from thermally pretreated whole slurries. Furfural and 5-HMF released in acid pretreatment inhibited methane production. High phenolic compounds release required a microorganism's acclimation period. Lignin degradation provided the highest hydrolysis rates when inhibition was defeated. A novel kinetic model is proposed combining hydrolysis and microorganisms inhibition ARTICLE INFO ABSTRACT Article history: The effect of thermal,acid,alkaline and alkaline-peroxide pretreatments on the methane produced by the Received 23 September 2015 anaerobic digestion of wheat straw(WS)and sugarcane bagasse(SCB)was studied,using whole slurry Received in revised form 17 November 2015 and solid fraction.All the pretreatments released formic and acetic acids and phenolic compounds,while Accepted 18 November 2015 Available online 28 November 2015 5-hydroxymetilfurfural(HMF)and furfural were generated only by acid pretreatment.A remarkable inhi- bition was found in most of the whole slurry experiments,except in thermal pretreatment which improved methane production compared to the raw materials(29%for WS and 11%for SCB).The alkaline Keywords: pretreatment increased biodegradability (around 30%)and methane production rate of the solid fraction Biogas Pretreatment of both pretreated substrates.Methane production results were fitted using first order or modified Inhibition Gompertz equations,or a novel model combining both equations.The model parameters provided infor- Lignocellulosic material mation about substrate availability,controlling step and inhibitory effect of compounds generated by Biodegradability each pretreatment. 2015 Elsevier Ltd.All rights reserved. 1.Introduction Among renewable sources,the lignocellulosic biomass is one of the most attractive alternatives for bioenergy production (second Bioenergy production from renewable sources is becoming generation technology)since it is available in high quantities crucial in order to address the growing demand for energy and at a low cost (Badshah et al.,2012).This study focuses on and the need to reduce greenhouse gas emissions,owing to the bioenergy production from two of the major agricultural ligno- unavoidable depletion of fossil fuel reserves and the environmental cellulosic residues:wheat straw (WS)and sugarcane bagasse consequences of global warming (Karagoz et al.,2012). (SCB).Wheat straw represents the largest fraction of agricultural waste in many countries,including Spain.Most of this wheat straw is commonly used for mulching or as fodder and the rest is burnt or left unused.For this reason,its use for biofuel production is grow- Abbreviations:BMP.biochemical methane potential:NP.normalized production ing worldwide (Menon and Rao,2012).Sugarcane bagasse is an of methane:SCB.sugarcane bagasse:TS,total solids:TKN,total Kjeldahl nitrogen: VS.volatile solids:WS,wheat straw. abundant lignocellulosic residue produced in many tropical coun- Corresponding author.Tel.:+34983 423958. tries,such as Brazil,India and Colombia.This bagasse is commonly E-mail address:silvia@iq uvaes (S.Bolado-Rodriguez) used for generating electricity by combustion,as animal feedstock, http://dx.doiorg/10.1016/j.biortech.2015.11.047 0960-8524/2015 Elsevier Ltd.All rights reserved
Effect of thermal, acid, alkaline and alkaline-peroxide pretreatments on the biochemical methane potential and kinetics of the anaerobic digestion of wheat straw and sugarcane bagasse Silvia Bolado-Rodríguez ⇑ , Cristina Toquero, Judit Martín-Juárez, Rodolfo Travaini, Pedro Antonio García-Encina Department of Chemical Engineering and Environmental Technology, University of Valladolid, Calle Doctor Mergelina s/n, 47011 Valladolid, Spain highlights Highest methane productions were obtained from thermally pretreated whole slurries. Furfural and 5-HMF released in acid pretreatment inhibited methane production. High phenolic compounds release required a microorganism’s acclimation period. Lignin degradation provided the highest hydrolysis rates when inhibition was defeated. A novel kinetic model is proposed combining hydrolysis and microorganisms inhibition. article info Article history: Received 23 September 2015 Received in revised form 17 November 2015 Accepted 18 November 2015 Available online 28 November 2015 Keywords: Biogas Pretreatment Inhibition Lignocellulosic material Biodegradability abstract The effect of thermal, acid, alkaline and alkaline-peroxide pretreatments on the methane produced by the anaerobic digestion of wheat straw (WS) and sugarcane bagasse (SCB) was studied, using whole slurry and solid fraction. All the pretreatments released formic and acetic acids and phenolic compounds, while 5-hydroxymetilfurfural (HMF) and furfural were generated only by acid pretreatment. A remarkable inhibition was found in most of the whole slurry experiments, except in thermal pretreatment which improved methane production compared to the raw materials (29% for WS and 11% for SCB). The alkaline pretreatment increased biodegradability (around 30%) and methane production rate of the solid fraction of both pretreated substrates. Methane production results were fitted using first order or modified Gompertz equations, or a novel model combining both equations. The model parameters provided information about substrate availability, controlling step and inhibitory effect of compounds generated by each pretreatment. 2015 Elsevier Ltd. All rights reserved. 1. Introduction Bioenergy production from renewable sources is becoming crucial in order to address the growing demand for energy and the need to reduce greenhouse gas emissions, owing to the unavoidable depletion of fossil fuel reserves and the environmental consequences of global warming (Karagöz et al., 2012). Among renewable sources, the lignocellulosic biomass is one of the most attractive alternatives for bioenergy production (second generation technology) since it is available in high quantities and at a low cost (Badshah et al., 2012). This study focuses on bioenergy production from two of the major agricultural lignocellulosic residues: wheat straw (WS) and sugarcane bagasse (SCB). Wheat straw represents the largest fraction of agricultural waste in many countries, including Spain. Most of this wheat straw is commonly used for mulching or as fodder and the rest is burnt or left unused. For this reason, its use for biofuel production is growing worldwide (Menon and Rao, 2012). Sugarcane bagasse is an abundant lignocellulosic residue produced in many tropical countries, such as Brazil, India and Colombia. This bagasse is commonly used for generating electricity by combustion, as animal feedstock, http://dx.doi.org/10.1016/j.biortech.2015.11.047 0960-8524/ 2015 Elsevier Ltd. All rights reserved. Abbreviations: BMP, biochemical methane potential; NP, normalized production of methane; SCB, sugarcane bagasse; TS, total solids; TKN, total Kjeldahl nitrogen; VS, volatile solids; WS, wheat straw. ⇑ Corresponding author. Tel.: +34 983 423 958. E-mail address: silvia@iq.uva.es (S. Bolado-Rodríguez). Bioresource Technology 201 (2016) 182–190 Contents lists available at ScienceDirect Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
S.Bolado-Rodriguez et aL/Bioresource Technology 201 (2016)182-190 183 or as fuel in boilers that produce low-pressure steam.However,the pretreatment(Rabelo et al.,2011:Talebnia et al.,2010).The pro- surplus that remains leads to environmental and storage problems cess is usually carried out at mild temperatures.using hydrogen (Costa et al.,2014:Travaini et al.,2013). peroxide(H2O2)and NaOH,leading to a lesser formation of inhibi- Three types of energy can be produced from these lignocellu- tors than in other processes. losic wastes through thermochemical or biochemical processing: The determination of the kinetic of the anaerobic digestion pro- liquid fuels such as bioethanol,gaseous fuels such as biogas,and vides important information about the effect of the inhibitory com- electricity by combustion (Menon and Rao,2012). pounds generated by the pretreatment on the biodegradability. Biogas,composed mainly of methane and carbon dioxide,is and to determine if the hydrolysis is the limiting step.There are considered a clean and renewable form of energy.It has the advan- several models of the kinetic analysis of biogas production process; tage of being easy to implement for consumers,and easy to it all depends on the types of substrate used for anaerobic digestion produce on a local level,such as small-scale farms (Taherdanak and the controlling step. and Zilouei,2014).Biogas can be produced through the anaerobic The Gompertz model is well known among the available models digestion of many types of wastes,and is considered one of the for the kinetic behavior of the anaerobic digestion process consid- most efficient technologies,since high energy recovery and ering inhibition.The Gompertz equation is used to estimate the environmental benefits can be achieved (Ferreira et al.,2013). kinetic parameters;biogas yield potential,duration of the lag Nevertheless,the biodegradability of biomass residues is lim- phase,and maximum biogas production rate (Krishania et al.. ited by its lignocellulosic structure.Therefore,efficient pretreat- 2013).However,when the hydrolysis reaction is the rate limiting ment digestion could accelerate the hydrolysis and improve the step of the overall process,as in the anaerobic degradation of some biogas production(Sambusiti et al.,2013).However,the realization lignocellulosic substrates,the first order model is commonly used of a pretreatment frequently produces degradation compounds to estimate the extent of the reaction,and the hydrolysis constant. that can act as inhibitors:organic acids(acetic,formic and levuli- Both parameters can be used in a global model of the anaerobic nic),furan derivatives [furfural and 5-hydroxymethylfurfural digestion process (such as ADM1)to predict the performance of (5-HMF)]and phenol compounds,affecting overall cell physiology anaerobic digesters (Ferreira et al.,2013). and often decreasing viability and productivity (Chandel et al.. The present study aims to establish the influence of four pre- 2011). treatments (thermal autoclaving.dilute HCI autoclaving.dilute Different pretreatment methods have been studied,depending NaOH autoclaving and alkaline peroxide)in the production of on the characteristics of each lignocellulosic feedstock (Karagoz biogas from sugarcane bagasse and wheat straw,and to study et al.,2012),including biological,chemical,physical processes,or the kinetics of anaerobic digestion in order to determine the a combination of them.Among them,this work focuses on thermal, influence of inhibitory compounds present in both the liquid phase dilute acid,dilute alkaline and oxidative pretreatments. and the solid phase. The thermal pretreatments are considered eco-friendly,green processing technologies.Energy recovery from biomass for fuel is excellent,often with values as high as 80%(Chandra et al., 2.Methods 2012a).Thermal pretreatments have been applied to improve the anaerobic digestibility of different agriculture substrates such as 2.1.Materials wheat straw,sorghum forage and sugarcane bagasse (Costa et al.. 2014;Sambusiti et al.,2013).The non-addition of chemicals avoids Two lignocellulosic substrates were used in this study:WS,pro- the corrosion problems,and decreases the formation of toxic vided by the Castilla Leon Institute of Technological Agriculture compounds.Other advantages include the lower requirement of from Valladolid(Spain).and SCB(surplus after milling in a sugar/ chemicals for the neutralization of the hydrolysates produced, ethanol factory).donated by Usina Vale,City of Onda Verde-SP and the smaller amount of waste produced in comparison to other (Brazil).Wheat straw and sugarcane bagasse were washed for processes(Ferreira et al..2013). particulate material removal,dried in a ventilated oven at 42C Acid pretreatment is widely applied due to its low cost and high and ground in an agricultural crusher to a size of 3-5 mm.Both efficiency to hydrolyze hemicellulose into monomeric sugars with- substrates were kept in an oven at 45C until they reached a out dissolving lignin(Ferreira et al.,2013).However,this pretreat- constant weight prior to compositional analysis and different ment is corrosive and generates high concentrations of toxic pretreatments.The chemical composition of both substrates is compounds,making it necessary to recover the acids in order to presented in Table 1. make the process economically feasible (Talebnia et al.,2010). The main substrates studied for this pretreatment are wheat straw, 2.2.Pretreatments sorghum forage and sugarcane bagasse (Sambusiti et al.,2013). and different acids such as sulphuric,hydrochloric,phosphoric, Four different pretreatments were applied to both substrates in maleic,peracetic or nitric acids have been investigated (Badshah this study:thermal autoclaving(A).dilute HCI autoclaving (B). et al.,2012:Chandel et al.,2011:Costa et al.,2014:Krishania etal,2013). The alkaline pretreatment is typically used in lignocellulosic Table 1 Composition of raw materials. materials with a high lignin content,such as wheat straw and sug- arcane bagasse(Rabelo et al,2011:Taherdanak and Zilouei,2014). Parameter Wheat straw Sugarcane bagasse Alkaline pretreatments performed with bases such as sodium, Total solids(g TS/kg) 916.24±121 91922±084 potassium,calcium and ammonium hydroxides are effective in Volatile solids (g vS/kg) 818.83±1.52 907.96±1.10 modifying the structure and solubilizing the lignin.In addition. N-TKN (g N/kg) 4.85±0.09 2.51±0.02 the alkaline pretreatment reduces the degree of inhibition in TCOD(g Oz/kg) 1150.40±4.99 1188.85±2.43 Cellulose (w/w) 35.19±029 46.21±0.10 methane fermentation and provides a lower cost of production Hemicellulose w/w)" 22.15±021 20.86±0.05 (Ferreira et al.,2013:Krishania et al.,2013). Total lignin w/w)" 22.09±0.80 22.67±0.04 The use of an oxidizing compound in combination with an alka- Acid insoluble lignin w/w) 18.17±021 19.53±0.03 line pretreatment is becoming more common in order to improve Ash w/w) 7.49±029 1.19±0.10 the digestibility of crop residues,compared with an alkaline Dry basis calculated composition
or as fuel in boilers that produce low-pressure steam. However, the surplus that remains leads to environmental and storage problems (Costa et al., 2014; Travaini et al., 2013). Three types of energy can be produced from these lignocellulosic wastes through thermochemical or biochemical processing: liquid fuels such as bioethanol, gaseous fuels such as biogas, and electricity by combustion (Menon and Rao, 2012). Biogas, composed mainly of methane and carbon dioxide, is considered a clean and renewable form of energy. It has the advantage of being easy to implement for consumers, and easy to produce on a local level, such as small-scale farms (Taherdanak and Zilouei, 2014). Biogas can be produced through the anaerobic digestion of many types of wastes, and is considered one of the most efficient technologies, since high energy recovery and environmental benefits can be achieved (Ferreira et al., 2013). Nevertheless, the biodegradability of biomass residues is limited by its lignocellulosic structure. Therefore, efficient pretreatment digestion could accelerate the hydrolysis and improve the biogas production (Sambusiti et al., 2013). However, the realization of a pretreatment frequently produces degradation compounds that can act as inhibitors: organic acids (acetic, formic and levulinic), furan derivatives [furfural and 5-hydroxymethylfurfural (5-HMF)] and phenol compounds, affecting overall cell physiology and often decreasing viability and productivity (Chandel et al., 2011). Different pretreatment methods have been studied, depending on the characteristics of each lignocellulosic feedstock (Karagöz et al., 2012), including biological, chemical, physical processes, or a combination of them. Among them, this work focuses on thermal, dilute acid, dilute alkaline and oxidative pretreatments. The thermal pretreatments are considered eco-friendly, green processing technologies. Energy recovery from biomass for fuel is excellent, often with values as high as 80% (Chandra et al., 2012a). Thermal pretreatments have been applied to improve the anaerobic digestibility of different agriculture substrates such as wheat straw, sorghum forage and sugarcane bagasse (Costa et al., 2014; Sambusiti et al., 2013). The non-addition of chemicals avoids the corrosion problems, and decreases the formation of toxic compounds. Other advantages include the lower requirement of chemicals for the neutralization of the hydrolysates produced, and the smaller amount of waste produced in comparison to other processes (Ferreira et al., 2013). Acid pretreatment is widely applied due to its low cost and high efficiency to hydrolyze hemicellulose into monomeric sugars without dissolving lignin (Ferreira et al., 2013). However, this pretreatment is corrosive and generates high concentrations of toxic compounds, making it necessary to recover the acids in order to make the process economically feasible (Talebnia et al., 2010). The main substrates studied for this pretreatment are wheat straw, sorghum forage and sugarcane bagasse (Sambusiti et al., 2013), and different acids such as sulphuric, hydrochloric, phosphoric, maleic, peracetic or nitric acids have been investigated (Badshah et al., 2012; Chandel et al., 2011; Costa et al., 2014; Krishania et al., 2013). The alkaline pretreatment is typically used in lignocellulosic materials with a high lignin content, such as wheat straw and sugarcane bagasse (Rabelo et al., 2011; Taherdanak and Zilouei, 2014). Alkaline pretreatments performed with bases such as sodium, potassium, calcium and ammonium hydroxides are effective in modifying the structure and solubilizing the lignin. In addition, the alkaline pretreatment reduces the degree of inhibition in methane fermentation and provides a lower cost of production (Ferreira et al., 2013; Krishania et al., 2013). The use of an oxidizing compound in combination with an alkaline pretreatment is becoming more common in order to improve the digestibility of crop residues, compared with an alkaline pretreatment (Rabelo et al., 2011; Talebnia et al., 2010). The process is usually carried out at mild temperatures, using hydrogen peroxide (H2O2) and NaOH, leading to a lesser formation of inhibitors than in other processes. The determination of the kinetic of the anaerobic digestion provides important information about the effect of the inhibitory compounds generated by the pretreatment on the biodegradability, and to determine if the hydrolysis is the limiting step. There are several models of the kinetic analysis of biogas production process; it all depends on the types of substrate used for anaerobic digestion and the controlling step. The Gompertz model is well known among the available models for the kinetic behavior of the anaerobic digestion process considering inhibition. The Gompertz equation is used to estimate the kinetic parameters; biogas yield potential, duration of the lag phase, and maximum biogas production rate (Krishania et al., 2013). However, when the hydrolysis reaction is the rate limiting step of the overall process, as in the anaerobic degradation of some lignocellulosic substrates, the first order model is commonly used to estimate the extent of the reaction, and the hydrolysis constant. Both parameters can be used in a global model of the anaerobic digestion process (such as ADM1) to predict the performance of anaerobic digesters (Ferreira et al., 2013). The present study aims to establish the influence of four pretreatments (thermal autoclaving, dilute HCl autoclaving, dilute NaOH autoclaving and alkaline peroxide) in the production of biogas from sugarcane bagasse and wheat straw, and to study the kinetics of anaerobic digestion in order to determine the influence of inhibitory compounds present in both the liquid phase and the solid phase. 2. Methods 2.1. Materials Two lignocellulosic substrates were used in this study: WS, provided by the Castilla & León Institute of Technological Agriculture from Valladolid (Spain), and SCB (surplus after milling in a sugar/ ethanol factory), donated by Usina Vale, City of Onda Verde-SP (Brazil). Wheat straw and sugarcane bagasse were washed for particulate material removal, dried in a ventilated oven at 42 C and ground in an agricultural crusher to a size of 3–5 mm. Both substrates were kept in an oven at 45 C until they reached a constant weight prior to compositional analysis and different pretreatments. The chemical composition of both substrates is presented in Table 1. 2.2. Pretreatments Four different pretreatments were applied to both substrates in this study: thermal autoclaving (A), dilute HCl autoclaving (B), Table 1 Composition of raw materials. Parameter Wheat straw Sugarcane bagasse Total solids (g TS/kg) 916.24 ± 1.21 919.22 ± 0.84 Volatile solids (g VS/kg) 818.83 ± 1.52 907.96 ± 1.10 N-TKN (g N/kg)* 4.85 ± 0.09 2.51 ± 0.02 TCOD (g O2/kg)* 1150.40 ± 4.99 1188.85 ± 2.43 Cellulose (% w/w)* 35.19 ± 0.29 46.21 ± 0.10 Hemicellulose (% w/w)* 22.15 ± 0.21 20.86 ± 0.05 Total lignin (% w/w)* 22.09 ± 0.80 22.67 ± 0.04 Acid insoluble lignin (% w/w)* 18.17 ± 0.21 19.53 ± 0.03 Ash (% w/w)* 7.49 ± 0.29 1.19 ± 0.10 * Dry basis calculated composition. S. Bolado-Rodríguez et al. / Bioresource Technology 201 (2016) 182–190 183
184 S.Bolado-Rodriguez et aL/Bioresource Technology 201 (2016)182-190 dilute NaOH autoclaving(C),and alkaline peroxide(D).The opera- methane yields are expressed as the volume of methane under tional conditions were the same as those selected by Toquero and standard conditions,i.e.0C and 1 atm for gases,as the Interna- Bolado (2014),according to the optimal experimental settings tional Union of Pure Applied Chemistry(IUPAC)defines,per gram previously reported for each pretreatment (Akhtar et al.,2001: of VS in substrates fed into the assay(N mL CH4/g VS). Cao et al.,2012:Karagoz et al.,2012;Sun and Cheng.2005).In Theoretical methane yields,calculated from the characteriza- autoclave pretreatments A,B,and C,milled and dried WS or SCB tion performed for both substrates,were as follows:449 mL were slurried for 5 min with distilled water,1.5%w/w HCl solution. CH4/g VS for WS;and 420 mL CH4/g VS for SCB.These values are and 1%w/w NaOH solution,respectively,in a 500 mL screw cap consistent with those calculated by Ferreira et al.(2013)for WS bottle with a solid:liquid ratio of 1:10 w/w,and then autoclaved (444 mL CH4/g VS)and Badshah et al.(2012)for SCB(415 mL CH4/ at 121 C for 60 min.In alkaline peroxide pretreatment(D).milled g VS).considering the stoichiometric conversion of the organic and dry WS and SCB were slurried for 5 min with 5%w/w H2O2,in a matter. solid:liquid ratio of 1:20,the pH was then adjusted to 11.5 with 2 M NaOH and the mixture was placed in a rotatory shaker at 2.4.Analytical methods 50C and 120 rpm for 60 min. After pretreatment,and once cooled down to room tempera- Total solids (TS).volatile solids and total Kjeldahl nitrogen ture,the slurry obtained from each pretreatment was recovered. (TKN)were measured following the procedures given in Standard and the residual solid was separated by vacuum filtration till max- Methods for Examination of Water and Wastewater (APHA imum liquid removal and dried in a ventilated oven at 45 C for 2005).Total chemical oxygen demand(TCOD)was determined 48 h.Liquid fractions from every pretreatments were stored in a according to the standard method UNE 77004:2002 based on the refrigeration chamber for biogas production and compositional dichromate method only in the initial raw material (Ferreira analysis.Solid fraction,or whole slurry were used as the substrates et al.,2013). in the subsequent step of anaerobic digestion.All experiments The analytical methods of the National Renewable Energy Lab- were conducted in triplicate and the results were averaged. oratory (NREL)were followed to determine substrate composition in terms of ash,lignin,cellulose (as glucose),and hemicellulose (as 2.3.Anaerobic biodegradability xylose)(Sluiter et al..2012).High performance liquid chromatogra- phy (HPLC)was used to measure glucose,xylose,formic acid,acetic Biochemical methane potential(BMP)tests were carried out to acid,HMF and furfural,using a Bio-Rad HPX-87P column at 80C study the biodegradability of raw and pretreated substrates in with MilliQ water as the mobile phase for sugars and a Bio-Rad duplicate following the protocol of Angelidaki et al.(2009).Batch HPX-87H column at 50C with 0.005 M H2SO4 as the mobile phase mode assays were performed under mesophilic conditions in for acids,both at 0.6 mL/min.A Waters 2414 refractive index was borosilicate glass bottles of 2 L volume (260 mm height,160 mm used as detector (Travaini et al.,2013).The total content of pheno- diameter and a 40 mm bottleneck).The effluent from a pilot- lic compounds in the samples was determined by the Folin-Ciocal- scale mesophilic anaerobic digester processing mixed sludge from teu method (Singleton and Rossi,1965)with gallic acid as the a municipal wastewater treatment plant,with a volatile solids(VS) calibration standard.The biogas composition (CO2.H2S,O2,N2. concentration of 14.0t1.5 g VS/kg was used as inoculum for tests. CH4)was measured by gas chromatography using a varian Two series of experiments (test 1 and test 2)were performed in CP-177 3800 GC-TCD equipped with a CP-Molsieve 5A and a CP- order to determine the influence of the pretreatment and the Pora BOND Q columns,using helium as the carrier gas.All analyses inhibitory effect of compounds present in the liquid phase:(1) were performed in duplicate and all chromatographic standards using the whole slurry (solid and liquid fractions):and (2)using were of analytical grade,and MilliQ Ultrapure water was used. only the solid fraction. Raw substrates,whole slurries or solid fractions from pretreat- ment batches were adjusted to 10%w/w soil content in all the 2.5.Determination of kinetic parameters experiments,using either the pretreatment liquid (assays with whole slurry)or distilled water(assays with raw substrates and The cumulative methane production data from the experiments solid fractions from pretreatments).NaOH or HCI were added,if was fitted either to a first order model (Eq.(1))or to the modified necessary,in order to pre-neutralize the samples up to values of Gompertz equation (Eq.(2))(Lay et al.,1996)or to a combination pH 8 for alkaline samples or pH 5.5 for acid samples.The sludge of both.The first one was applied successfully in many reports on was added in a ratio substrate/inoculum around 0.5 g VS/g VS,in anaerobic biodegradability tests when the hydrolysis reaction was order to obtain a ratio 1 of hydrolysable material in substrate the rate-limiting step of the global process(Ferreira et al..2013). g VS inoculum,considering that sugars (hydrolysable material) The modified Gompertz model described the cumulative methane make up half of the substrate (Ferreira et al.,2014).The pH,after production in batch assays when an inhibitory behavior was adding the activated sludge,was always between 6.5 and 7.The observed,assuming that the methane production was a function working volume was approximately 400 mL in order to have of bacterial growth.Moreover,the model parameters were calcu- enough headspace for gas production.A control test without lated by minimizing the least square difference between observed substrate was also conducted,aiming to check the methanogenic and predicted values. activity of the inoculum. B=Bo·[1-exp(-kH·t)] (1) Before starting the test,the bottles were closed with rubber septa and aluminium crimps.Helium gas was circulated inside the gas chamber for 5 min,and the test started after releasing a-Bexpf-exp+ (2) the pressure.The bottles were placed horizontally in a rotary desk with constant mixing under mesophilic conditions in a thermo- In these equations,B represents the cumulative methane static room (35.10.3C). production (mL CH4/g VS)and t is the length of the assay(d).These Biogas production in the headspace of each bottle was mea- models estimate the methane production potential Bo(mL CH4/ sured periodically by a manual pressure transmitter (PN5007, g VS,related to the substrate biodegradability).the hydrolysis range 0-1 bar,IFM Electronics)over a period of 30 days.Biogas coefficient kH (d-1).the maximum biogas production rate Rm composition was determined by gas chromatography.Specific (mL CH4/g VS-d),and the lag time (d)
dilute NaOH autoclaving (C), and alkaline peroxide (D). The operational conditions were the same as those selected by Toquero and Bolado (2014), according to the optimal experimental settings previously reported for each pretreatment (Akhtar et al., 2001; Cao et al., 2012; Karagöz et al., 2012; Sun and Cheng, 2005). In autoclave pretreatments A, B, and C, milled and dried WS or SCB were slurried for 5 min with distilled water, 1.5% w/w HCl solution, and 1% w/w NaOH solution, respectively, in a 500 mL screw cap bottle with a solid:liquid ratio of 1:10 w/w, and then autoclaved at 121 C for 60 min. In alkaline peroxide pretreatment (D), milled and dry WS and SCB were slurried for 5 min with 5% w/w H2O2, in a solid:liquid ratio of 1:20, the pH was then adjusted to 11.5 with 2 M NaOH and the mixture was placed in a rotatory shaker at 50 C and 120 rpm for 60 min. After pretreatment, and once cooled down to room temperature, the slurry obtained from each pretreatment was recovered, and the residual solid was separated by vacuum filtration till maximum liquid removal and dried in a ventilated oven at 45 C for 48 h. Liquid fractions from every pretreatments were stored in a refrigeration chamber for biogas production and compositional analysis. Solid fraction, or whole slurry were used as the substrates in the subsequent step of anaerobic digestion. All experiments were conducted in triplicate and the results were averaged. 2.3. Anaerobic biodegradability Biochemical methane potential (BMP) tests were carried out to study the biodegradability of raw and pretreated substrates in duplicate following the protocol of Angelidaki et al. (2009). Batch mode assays were performed under mesophilic conditions in borosilicate glass bottles of 2 L volume (260 mm height, 160 mm diameter and a 40 mm bottleneck). The effluent from a pilotscale mesophilic anaerobic digester processing mixed sludge from a municipal wastewater treatment plant, with a volatile solids (VS) concentration of 14.0 ± 1.5 g VS/kg was used as inoculum for tests. Two series of experiments (test 1 and test 2) were performed in order to determine the influence of the pretreatment and the inhibitory effect of compounds present in the liquid phase: (1) using the whole slurry (solid and liquid fractions); and (2) using only the solid fraction. Raw substrates, whole slurries or solid fractions from pretreatment batches were adjusted to 10% w/w soil content in all the experiments, using either the pretreatment liquid (assays with whole slurry) or distilled water (assays with raw substrates and solid fractions from pretreatments). NaOH or HCl were added, if necessary, in order to pre-neutralize the samples up to values of pH 8 for alkaline samples or pH 5.5 for acid samples. The sludge was added in a ratio substrate/inoculum around 0.5 g VS/g VS, in order to obtain a ratio 1 of hydrolysable material in substrate/ g VS inoculum, considering that sugars (hydrolysable material) make up half of the substrate (Ferreira et al., 2014). The pH, after adding the activated sludge, was always between 6.5 and 7. The working volume was approximately 400 mL, in order to have enough headspace for gas production. A control test without substrate was also conducted, aiming to check the methanogenic activity of the inoculum. Before starting the test, the bottles were closed with rubber septa and aluminium crimps. Helium gas was circulated inside the gas chamber for 5 min, and the test started after releasing the pressure. The bottles were placed horizontally in a rotary desk with constant mixing under mesophilic conditions in a thermostatic room (35.1 ± 0.3 C). Biogas production in the headspace of each bottle was measured periodically by a manual pressure transmitter (PN5007, range 0–1 bar, IFM Electronics) over a period of 30 days. Biogas composition was determined by gas chromatography. Specific methane yields are expressed as the volume of methane under standard conditions, i.e. 0 C and 1 atm for gases, as the International Union of Pure Applied Chemistry (IUPAC) defines, per gram of VS in substrates fed into the assay (N mL CH4/g VS). Theoretical methane yields, calculated from the characterization performed for both substrates, were as follows: 449 mL CH4/g VS for WS; and 420 mL CH4/g VS for SCB. These values are consistent with those calculated by Ferreira et al. (2013) for WS (444 mL CH4/g VS) and Badshah et al. (2012) for SCB (415 mL CH4/ g VS), considering the stoichiometric conversion of the organic matter. 2.4. Analytical methods Total solids (TS), volatile solids and total Kjeldahl nitrogen (TKN) were measured following the procedures given in Standard Methods for Examination of Water and Wastewater (APHA, 2005). Total chemical oxygen demand (TCOD) was determined according to the standard method UNE 77004:2002 based on the dichromate method only in the initial raw material (Ferreira et al., 2013). The analytical methods of the National Renewable Energy Laboratory (NREL) were followed to determine substrate composition in terms of ash, lignin, cellulose (as glucose), and hemicellulose (as xylose) (Sluiter et al., 2012). High performance liquid chromatography (HPLC) was used to measure glucose, xylose, formic acid, acetic acid, HMF and furfural, using a Bio-Rad HPX-87P column at 80 C with MilliQ water as the mobile phase for sugars and a Bio-Rad HPX-87H column at 50 C with 0.005 M H2SO4 as the mobile phase for acids, both at 0.6 mL/min. A Waters 2414 refractive index was used as detector (Travaini et al., 2013). The total content of phenolic compounds in the samples was determined by the Folin–Ciocalteu method (Singleton and Rossi, 1965) with gallic acid as the calibration standard. The biogas composition (CO2, H2S, O2, N2, CH4) was measured by gas chromatography using a Varian CP-177 3800 GC-TCD equipped with a CP-Molsieve 5A and a CPPora BOND Q columns, using helium as the carrier gas. All analyses were performed in duplicate and all chromatographic standards were of analytical grade, and MilliQ Ultrapure water was used. 2.5. Determination of kinetic parameters The cumulative methane production data from the experiments was fitted either to a first order model (Eq. (1)) or to the modified Gompertz equation (Eq. (2)) (Lay et al., 1996) or to a combination of both. The first one was applied successfully in many reports on anaerobic biodegradability tests when the hydrolysis reaction was the rate-limiting step of the global process (Ferreira et al., 2013). The modified Gompertz model described the cumulative methane production in batch assays when an inhibitory behavior was observed, assuming that the methane production was a function of bacterial growth. Moreover, the model parameters were calculated by minimizing the least square difference between observed and predicted values. B ¼ B0 ½1 expðkH tÞ ð1Þ B ¼ B0 exp exp Rm e B0 ðk tÞ þ 1 ð2Þ In these equations, B represents the cumulative methane production (mL CH4/g VS) and t is the length of the assay (d). These models estimate the methane production potential B0 (mL CH4/ g VS, related to the substrate biodegradability), the hydrolysis coefficient kH (d1 ), the maximum biogas production rate Rm (mL CH4/g VSd), and the lag time k(d). 184 S. Bolado-Rodríguez et al. / Bioresource Technology 201 (2016) 182–190
S.Bolado-Rodriguez et aL/Bioresource Technology 201(2016)182-190 185 3.Results and discussion et al..2011).However,Sambusiti et al.(2013)reported the release of furfural (0.2-0.4 g/100g VS raw material)in the thermal pre- 3.1.Composition of pretreatment liquids treatment of WS at 100C and 160C.Costa et al.(2014)obtained considerably higher sugar release (1.94 g/100g raw material)and The presence and concentration of sugars,volatile solids and some HMF production from the thermal pretreatment of SCB degradation compounds released into the liquid during pretreat- working with a higher temperature (150C)but lower reaction ments can significantly affect the anaerobic digestion of pretreated time(30 min).Their experiments of SCB acid pretreatment,work- biomass.Table 2 shows these values for percentages of sugars ing at 120C and 40 min with 0.63 M HCI.provided appreciably (glucose plus xylose),volatile solids and degradation compounds lower degradation compounds (0.001 g of 5-HMF and 0.8 g of formed and released into the liquid during the four pretreatments furfural by 100 g of raw material). for both substrates Rabelo et al.(2011)reported high sugar concentrations The pretreatment liquids of both substrates presented low sug- (1.504-2.904 g/L)with lower solid concentrations of sugarcane ars release,with values below 4.04 g/100g raw material in all bagasse(4-8%)in a pretreatment with Ca(OH)2 at 90C over a long cases except for the liquids from the diluted acid pretreatment period (90 h).They also obtained higher sugar concentrations (B).which reached a higher content due to solubilization of hemi- (7.565-21.759 g/L)in an alkaline hydrogen peroxide pretreatment celluloses produced by this type of pretreatment.The experimental at 25C with solids concentrations of between 4%and 15%over pretreatment conditions in this work are identical to those applied 1 h.These high sugar release results may be due to a higher by Toquero and Bolado(2014)for ethanol production from wheat concentration of H202(7.36%(v/v). straw,so the sugar release results are also the same for this material.For most of the pretreatments tested,the sugar release 3.2.Methane production was higher for WS than for SCB.In contrast,the concentration of volatile solids in the pretreatment liquids was greater,in all cases 3.2.1.Test 1:BMP of untreated materials and of pretreated whole for SCB that for WS.The highest volatile solids concentration was slurry found in the liquid phases from the basic pretreatment(C),despite The influence of pretreatments applied to both substrates was the low sugar release in this pretreatment.This high volatile solids studied in terms of methane production,considering three param- release may be related to the remarkable effect of the dilute eters:methane yield,defined as the volume of methane gas alkaline pretreatment disrupting the lignin structure.Liu et al. produced per gram of volatile solid fed;biodegradability,defined (2015)observed a lignin reduction of up to 54.7%in wheat straw as the percentage of the theoretical methane yield determined pretreatment with 50%KOH solution.Costa et al.(2014)extracted for raw substrates;and normalized production of methane (NP). approximately 80%of the lignin content of sugarcane bagasse defined as the ratio between the production of methane per gram using alkaline pretreatment (130C,20 min,NaOH 1 M). of VS from treated and untreated substrates. The main degradation compounds found in the pretreatment Fig.I presents the cumulative methane production curves from liquids were organic acids and phenolics from lignin.Acetic acid test 1,working with whole slurry;the results of WS (a)and SCB(b) and phenolic compounds appear in the liquid phase of all the are shown together for comparison.Results from untreated pretreated samples.HMF and furfural were only detected in hydro- substrates are also presented in this figure. lysates from acid pretreatment samples(B).where the formic acid By comparing the experimental values obtained in the BMP concentration was very low or undetectable.Comparing both tests of raw WS with the theoretical methane yield(449 mL CH4/ substrates,except for in the thermal pretreatment,which pro- g VS).it was determined that 48%of the VS were converted into duced very low concentrations of inhibitors,the production of methane.This result is very close to the 51%obtained with WS degradation compounds was higher for SCB than for WS.The high- by Ferreira et al.(2013).For untreated SCB,the experimental est concentrations of degradation compounds were detected in biodegradability was slightly higher,reaching 52%of the theoreti- samples from the basic pretreatment (C).with 5.80 g/100g of cal value (420 mL CH4/g VS). raw material for WS,and with 8.36 g/100 g of raw material for Concerning the assays with whole slurry of pretreated materi- SCB,respectively.This pretreatment provided the highest releases als,the highest methane yields were obtained for substrates of acetic acid and phenolic compounds,related to the previously pretreated by thermal autoclaving (A).both substrates followed noted high lignin removal effect. comparable behaviors during the digestion process.The anaerobic Sugars and volatile solids release,and degradation compounds biodegradability of WS_A1 increased up to 62%and the variable concentrations obtained in this study are within the range of most NP reached a value of 1.29.Thermal pretreatment also caused an of the results found in the literature (Bustos et al.,2003;Chandel increase in methane production from sample SCB_A1,which Table 2 Sugars (glucose plus xylose).volatile solids and degradation compounds(g/100 g of raw material)released to the liquid during pretreatments of wheat straw and sugarcane bagasse. Sample Degradation compounds Sugars Volatile solids Formic acid Acetic acid HMF Furfural Total phenolics WS_A 1.48±0.01 9.88 0.08±0.01 0.13±0.03 ND ND 0.42±0.01 WS_B 1774±300 27.8 0.06±0.00 0.72±000 0.04±0.00 1.10±0.07 0.80±0.00 WS_C 1.99±0.12 33.42 0.25±0.03 2.60±0.14 ND ND 2.95±0.00 WS_D 4.04±0.07 12.74 0.79±0.01 1.23±0.06 ND ND 0.48±0.01 SCB_A ND 10.43 0.01±0.00 0.08±0.00 ND ND 0.50±0.01 SCB_B 21.42±2.09 30.52 ND 4.11±0.13 0.02±0.00 122±0.11 1.32±0.00 SCB_C 0.98±0.00 37.41 0.12±0.01 4.30±0.24 ND ND 3.94±0.04 SCB_D 1.64±0.11 15.34 1.00±0.05 1.77±0.04 ND ND 1.14±0.00 ND,not detected,under detection limits of the method. Codes:Lignocellulosic material:WS,wheat straw and SCB sugarcane bagasse. Pr therm ving: ite acid autoclaving:C.dilute alkali autoclaving and D.alkaline peroxide
3. Results and discussion 3.1. Composition of pretreatment liquids The presence and concentration of sugars, volatile solids and degradation compounds released into the liquid during pretreatments can significantly affect the anaerobic digestion of pretreated biomass. Table 2 shows these values for percentages of sugars (glucose plus xylose), volatile solids and degradation compounds formed and released into the liquid during the four pretreatments for both substrates. The pretreatment liquids of both substrates presented low sugars release, with values below 4.04 g/100 g raw material in all cases except for the liquids from the diluted acid pretreatment (B), which reached a higher content due to solubilization of hemicelluloses produced by this type of pretreatment. The experimental pretreatment conditions in this work are identical to those applied by Toquero and Bolado (2014) for ethanol production from wheat straw, so the sugar release results are also the same for this material. For most of the pretreatments tested, the sugar release was higher for WS than for SCB. In contrast, the concentration of volatile solids in the pretreatment liquids was greater, in all cases for SCB that for WS. The highest volatile solids concentration was found in the liquid phases from the basic pretreatment (C), despite the low sugar release in this pretreatment. This high volatile solids release may be related to the remarkable effect of the dilute alkaline pretreatment disrupting the lignin structure. Liu et al. (2015) observed a lignin reduction of up to 54.7% in wheat straw pretreatment with 50% KOH solution. Costa et al. (2014) extracted approximately 80% of the lignin content of sugarcane bagasse using alkaline pretreatment (130 C, 20 min, NaOH 1 M). The main degradation compounds found in the pretreatment liquids were organic acids and phenolics from lignin. Acetic acid and phenolic compounds appear in the liquid phase of all the pretreated samples. HMF and furfural were only detected in hydrolysates from acid pretreatment samples (B), where the formic acid concentration was very low or undetectable. Comparing both substrates, except for in the thermal pretreatment, which produced very low concentrations of inhibitors, the production of degradation compounds was higher for SCB than for WS. The highest concentrations of degradation compounds were detected in samples from the basic pretreatment (C), with 5.80 g/100 g of raw material for WS, and with 8.36 g/100 g of raw material for SCB, respectively. This pretreatment provided the highest releases of acetic acid and phenolic compounds, related to the previously noted high lignin removal effect. Sugars and volatile solids release, and degradation compounds concentrations obtained in this study are within the range of most of the results found in the literature (Bustos et al., 2003; Chandel et al., 2011). However, Sambusiti et al. (2013) reported the release of furfural (0.2–0.4 g/100 g VS raw material) in the thermal pretreatment of WS at 100 C and 160 C. Costa et al. (2014) obtained considerably higher sugar release (1.94 g/100 g raw material) and some HMF production from the thermal pretreatment of SCB, working with a higher temperature (150 C) but lower reaction time (30 min). Their experiments of SCB acid pretreatment, working at 120 C and 40 min with 0.63 M HCl, provided appreciably lower degradation compounds (0.001 g of 5-HMF and 0.8 g of furfural by 100 g of raw material). Rabelo et al. (2011) reported high sugar concentrations (1.504–2.904 g/L) with lower solid concentrations of sugarcane bagasse (4–8%) in a pretreatment with Ca(OH)2 at 90 C over a long period (90 h). They also obtained higher sugar concentrations (7.565–21.759 g/L) in an alkaline hydrogen peroxide pretreatment at 25 C with solids concentrations of between 4% and 15% over 1 h. These high sugar release results may be due to a higher concentration of H2O2 (7.36% (v/v). 3.2. Methane production 3.2.1. Test 1: BMP of untreated materials and of pretreated whole slurry The influence of pretreatments applied to both substrates was studied in terms of methane production, considering three parameters: methane yield, defined as the volume of methane gas produced per gram of volatile solid fed; biodegradability, defined as the percentage of the theoretical methane yield determined for raw substrates; and normalized production of methane (NP), defined as the ratio between the production of methane per gram of VS from treated and untreated substrates. Fig. 1 presents the cumulative methane production curves from test 1, working with whole slurry; the results of WS (a) and SCB (b) are shown together for comparison. Results from untreated substrates are also presented in this figure. By comparing the experimental values obtained in the BMP tests of raw WS with the theoretical methane yield (449 mL CH4/ g VS), it was determined that 48% of the VS were converted into methane. This result is very close to the 51% obtained with WS by Ferreira et al. (2013). For untreated SCB, the experimental biodegradability was slightly higher, reaching 52% of the theoretical value (420 mL CH4/g VS). Concerning the assays with whole slurry of pretreated materials, the highest methane yields were obtained for substrates pretreated by thermal autoclaving (A), both substrates followed comparable behaviors during the digestion process. The anaerobic biodegradability of WS_A1 increased up to 62% and the variable NP reached a value of 1.29. Thermal pretreatment also caused an increase in methane production from sample SCB_A1, which Table 2 Sugars (glucose plus xylose), volatile solids and degradation compounds (g/100 g of raw material) released to the liquid during pretreatments of wheat straw and sugarcane bagasse. Samplea Degradation compounds Sugars Volatile solids Formic acid Acetic acid HMF Furfural Total phenolics WS_A 1.48 ± 0.01 9.88 0.08 ± 0.01 0.13 ± 0.03 ND ND 0.42 ± 0.01 WS_B 17.74 ± 3.00 27.8 0.06 ± 0.00 0.72 ± 0.00 0.04 ± 0.00 1.10 ± 0.07 0.80 ± 0.00 WS_C 1.99 ± 0.12 33.42 0.25 ± 0.03 2.60 ± 0.14 ND ND 2.95 ± 0.00 WS_D 4.04 ± 0.07 12.74 0.79 ± 0.01 1.23 ± 0.06 ND ND 0.48 ± 0.01 SCB_A ND 10.43 0.01 ± 0.00 0.08 ± 0.00 ND ND 0.50 ± 0.01 SCB_B 21.42 ± 2.09 30.52 ND 4.11 ± 0.13 0.02 ± 0.00 1.22 ± 0.11 1.32 ± 0.00 SCB_C 0.98 ± 0.00 37.41 0.12 ± 0.01 4.30 ± 0.24 ND ND 3.94 ± 0.04 SCB_D 1.64 ± 0.11 15.34 1.00 ± 0.05 1.77 ± 0.04 ND ND 1.14 ± 0.00 ND, not detected, under detection limits of the method. a Codes: Lignocellulosic material: WS, wheat straw and SCB, sugarcane bagasse. Pretreatment: A, thermal autoclaving; B, dilute acid autoclaving; C, dilute alkali autoclaving and D, alkaline peroxide. S. Bolado-Rodríguez et al. / Bioresource Technology 201 (2016) 182–190 185
186 S.Bolado-Rodriguez et aL/Bioresource Technology 201 (2016)182-190 ·Untreated WS·WsAl+ws_Bl·Ws_C1×WsDI Uetreated WS-WS Al_model-----WS Cl model ----WS D1 model furfural concentrations were detected only in samples of this pre 300 (a) treatment.These degradation compounds released to pretreatment liquids could have caused the inactivity or death of the inoculum (Chen et al.,2008:Sezun et al.,2011).In contrast,Costa et al. 250 (2014)obtained lower inhibitory compound concentration (0.33 g/L of furfural and 0.31 g/L of 5-HMF)in the liquid phase of 200 acid pretreated SCB (136C,6.4 min,HCI 0.63 M).achieving a BMP of 122.2 L CH4/kg of substrate.Badshah et al.(2012)obtained a methane yield from acid treated SCB(2%H2SO4.121C,15 min) 150- of 173 L/kg VS,which showed an increase in methane of 18%com- pared to untreated SCB,probably due to the low furfural and no 100- HMF release in these experiments. Cumulative methane production curves from the anaerobic digestion of dilute alkali(C)pretreated materials followed a similar 50- trend for both substrates,with an acclimation period,according to Fig.1.At the end of the test,methane production from sample 0 WS_C1 did not increase in comparison with untreated WS,with 12 13 18 21 24 27 30 33 a NP ratio of 1.00,indicating no effect of this pretreatment in wheat Time (d) straw biodegradability.Chandra et al.(2012b),with an alkaline ·Untresed SCB·SCBA1·SCB8I·SCB CI SCB_D1 -Untreated SCBSCB_Al_model-SCB CI model pretreatment(4%NaOH,37C,120 h),observed a lower methane 300= production of 165.9 L/kg VS.Nevertheless,they reported an (b) increase of 111%in relation to its very low methane production 250 from untreated wheat straw(78.4 L/kg VS).Reilly et al.(2015) obtained a final biomethane potential yield increase from 260 to 313 mL/g VS,but working with very different conditions:0.08 M 200 Ca(OH)2.0.59%(w/v)calculated,pretreatment for 48 h at 20C of 3 mm milled WS particles. However,SCB_C1 showed a slight increase in methane produc- 150- tion compared to raw SCB,with a biodegradability of 55%and NP of 1.05.Similar results were reported by Rabelo et al.(2011)with Ca 100- (OH)2 pretreated SCB(4-8%solid concentrations.90C,90 h) where the methane yield was 180-148 mL/g VS and the biodegrad- 50- ability was 51-42%. The alkali pretreatment did not favour the anaerobic digestion of any of the tested substrates,reducing methane productions dur- 0- ◆ ing the process and resulting in a biodegradability similar or even 0 1215 18 21 24 27 lower than the untreated materials at the end of the experiment. Time(d) The high degradation compound concentrations generated by the high lignin solubilization in alkali pretreatment probably Fig.1.Experimental results and fitting curves of cumulative methane production affected the bacterial growth and the methanogenic activity of from untreated and whole slurry pretreated materials (test 1).(A)Thermal autoclaving:(B)dilute acid autoclaving:(C)dilute alkali autoclaving:(D)alkaline inoculum. peroxide pretreatments.(a)Wheat straw and (b)sugarcane bagasse. Finally,using the complete pretreated material from alkaline peroxide pretreatment(D).WS and SCB were seen to behave com- pletely differently when subjected to anaerobic digestion,as seen in Fig.1.After 17 days of approximately zero methanogenic activ- reached an anaerobic biodegradability of 58%and NP of 1.11.Costa ity,the anaerobic biodegradability of WS_D1 increased up to 51 et al.(2014).comparing the effect of hydrothermal,acid and during the last days of the experiment,and the variable NP reached alkaline pretreatments on sugarcane bagasse,also achieved the a value of 1.07,whereas biodegradability of sample SCB_D1 only best values of BMP for hydrothermal pretreatment.They obtained reached the value of 4%with a really low NP ratio of 0.08.The inhi- a methane production value slightly lower than that of our ther- bition in samples of both substrates,with no methane production mal pretreatment (197.5 mL CH4/g substrate)working at 200C. during the first days,probably proceeds from the presence of resid- 10 min. ual chemicals from pretreatment.The concentration of acetic acid The concentration of degradation compounds in liquid from and phenolic compounds is lower than the values obtained for thermal pretreatment was very low,with a production of possible basic pretreatment,where the inhibition was less remarkable. inhibitory compounds of 0.63 g/100g raw material for WS and of The alkaline peroxide pretreatment releases the highest formic 0.69 g/100 g raw material for SCB.Thus,the use of whole slurry acid concentrations and this value is higher for sugarcane bagasse, from pretreatment A was beneficial for the anaerobic digestion of so another possibility could be an inhibitory effect of formic acid two tested substrates,due to the increase in biomass biodegrad- on methane production.Some better results were reported ability and the production of low concentrations of inhibitory by Rabelo et al.(2011)with 36%of biodegradability in alkaline compounds,leading to an enhancement of methane production, peroxide pretreated SCB (4%w/w.7.36%(v/v)H202,pH 11.5, especially for WS. 25C,1h). In the case of substrates pretreated by dilute acid autoclaving In short,with the aim of using a whole slurry of pretreated (B),a clear negative influence in the digestion process was found materials in order to harness pretreatment liquids and to avoid a to the extent that the total inhibition of biogas production was separation step,the anaerobic digestion of samples from thermal observed.As shown in Fig.1,the anaerobic digestion of samples method A provided the highest methane production results.The WS_B1 and SCB_B1 resulted in zero biogas production.HMF and use of chemical pretreatments B,C and D resulted in similar or
reached an anaerobic biodegradability of 58% and NP of 1.11. Costa et al. (2014), comparing the effect of hydrothermal, acid and alkaline pretreatments on sugarcane bagasse, also achieved the best values of BMP for hydrothermal pretreatment. They obtained a methane production value slightly lower than that of our thermal pretreatment (197.5 mL CH4/g substrate) working at 200 C, 10 min. The concentration of degradation compounds in liquid from thermal pretreatment was very low, with a production of possible inhibitory compounds of 0.63 g/100 g raw material for WS and of 0.69 g/100 g raw material for SCB. Thus, the use of whole slurry from pretreatment A was beneficial for the anaerobic digestion of two tested substrates, due to the increase in biomass biodegradability and the production of low concentrations of inhibitory compounds, leading to an enhancement of methane production, especially for WS. In the case of substrates pretreated by dilute acid autoclaving (B), a clear negative influence in the digestion process was found, to the extent that the total inhibition of biogas production was observed. As shown in Fig. 1, the anaerobic digestion of samples WS_B1 and SCB_B1 resulted in zero biogas production. HMF and furfural concentrations were detected only in samples of this pretreatment. These degradation compounds released to pretreatment liquids could have caused the inactivity or death of the inoculum (Chen et al., 2008; Sezˇun et al., 2011). In contrast, Costa et al. (2014) obtained lower inhibitory compound concentration (0.33 g/L of furfural and 0.31 g/L of 5-HMF) in the liquid phase of acid pretreated SCB (136 C, 6.4 min, HCl 0.63 M), achieving a BMP of 122.2 L CH4/kg of substrate. Badshah et al. (2012) obtained a methane yield from acid treated SCB (2% H2SO4, 121 C, 15 min) of 173 L/kg VS, which showed an increase in methane of 18% compared to untreated SCB, probably due to the low furfural and no HMF release in these experiments. Cumulative methane production curves from the anaerobic digestion of dilute alkali (C) pretreated materials followed a similar trend for both substrates, with an acclimation period, according to Fig. 1. At the end of the test, methane production from sample WS_C1 did not increase in comparison with untreated WS, with a NP ratio of 1.00, indicating no effect of this pretreatment in wheat straw biodegradability. Chandra et al. (2012b), with an alkaline pretreatment (4% NaOH, 37 C, 120 h), observed a lower methane production of 165.9 L/kg VS. Nevertheless, they reported an increase of 111% in relation to its very low methane production from untreated wheat straw (78.4 L/kg VS). Reilly et al. (2015) obtained a final biomethane potential yield increase from 260 to 313 mL/g VS, but working with very different conditions: 0.08 M Ca(OH)2, 0.59% (w/v) calculated, pretreatment for 48 h at 20 C of 3 mm milled WS particles. However, SCB_C1 showed a slight increase in methane production compared to raw SCB, with a biodegradability of 55% and NP of 1.05. Similar results were reported by Rabelo et al. (2011) with Ca (OH)2 pretreated SCB (4–8% solid concentrations, 90 C, 90 h) where the methane yield was 180–148 mL/g VS and the biodegradability was 51–42%. The alkali pretreatment did not favour the anaerobic digestion of any of the tested substrates, reducing methane productions during the process and resulting in a biodegradability similar or even lower than the untreated materials at the end of the experiment. The high degradation compound concentrations generated by the high lignin solubilization in alkali pretreatment probably affected the bacterial growth and the methanogenic activity of inoculum. Finally, using the complete pretreated material from alkaline peroxide pretreatment (D), WS and SCB were seen to behave completely differently when subjected to anaerobic digestion, as seen in Fig. 1. After 17 days of approximately zero methanogenic activity, the anaerobic biodegradability of WS_D1 increased up to 51% during the last days of the experiment, and the variable NP reached a value of 1.07, whereas biodegradability of sample SCB_D1 only reached the value of 4% with a really low NP ratio of 0.08. The inhibition in samples of both substrates, with no methane production during the first days, probably proceeds from the presence of residual chemicals from pretreatment. The concentration of acetic acid and phenolic compounds is lower than the values obtained for basic pretreatment, where the inhibition was less remarkable. The alkaline peroxide pretreatment releases the highest formic acid concentrations and this value is higher for sugarcane bagasse, so another possibility could be an inhibitory effect of formic acid on methane production. Some better results were reported by Rabelo et al. (2011) with 36% of biodegradability in alkaline peroxide pretreated SCB (4% w/w, 7.36% (v/v) H2O2, pH 11.5, 25 C, 1 h). In short, with the aim of using a whole slurry of pretreated materials in order to harness pretreatment liquids and to avoid a separation step, the anaerobic digestion of samples from thermal method A provided the highest methane production results. The use of chemical pretreatments B, C and D resulted in similar or Fig. 1. Experimental results and fitting curves of cumulative methane production from untreated and whole slurry pretreated materials (test 1). (A) Thermal autoclaving; (B) dilute acid autoclaving; (C) dilute alkali autoclaving; (D) alkaline peroxide pretreatments. (a) Wheat straw and (b) sugarcane bagasse. 186 S. Bolado-Rodríguez et al. / Bioresource Technology 201 (2016) 182–190