S.Bolado-Rodriguez et aL/Bioresource Technology 201 (2016)182-190 189 provided worse results than the untreated materials in terms of medium.The total methane potential for this experiment both methane potential and hydrolysis rate coefficient.Ferreira (Bo Boz)reached the theoretical methane yield for SCB.The same et al.(2013)used a first order equation to model the methane pro- theoretical methane potential value was achieved using the solid duction from untreated and steam exploded WS,demonstrating fraction of alkaline pretreated SCB,applying the first order model. that hydrolysis was the limiting step.For steam exploded WS,at The removal of liquid pretreatment,as for WS reduced inhibition 150C and 15 min,fitting parameters very similar to those from identified hydrolysis as the limiting step.Data fitted to a first order our thermal pretreatment,they obtained a methane potential of kinetic resulted in a very high methane potential but with a low 194.6 mL CH4/g VS and a hydrolysis coefficient of 0.085 d- hydrolysis kinetic coefficient. Very low values of methane production potential were obtained A strong inhibitory effect was found for all the alkaline peroxide from the solid fraction of both acid pretreated materials.The pretreated materials,probably generated by the degradation com- degradation compounds retained in the solid fraction probably pounds but also by some remaining peroxide.The Gompertz equa- inhibited the growth of most of the microorganisms.However, tion was used to fit the cumulative methane production from WS the small quantity of methane production obtained fitted ade- pretreated whole slurry with a methane potential very similar to quately to a first order kinetic,limited by the hydrolysis step. that of the untreated material.After a lag period of 17 days,the The microorganism species that were able to grow,biodegraded inhibition was completely defeated,reaching the highest degrada- the material with hydrolysis coefficients in a similar range to those tion rates.The 30 days period was not enough for microorganism for untreated materials. acclimation in the SCB whole slurry experiment. Alkaline pretreatment provided different results for both sub- The high inhibition of alkaline peroxide pretreatment also strates,probably attributable to the aforementioned differences affected the solid fraction experiments.Cumulative methane pro- in lignin structure and composition in tested materials.Basic pre- duction from both materials was fitted to a combination of the first treatment attacks mainly the lignin,and a higher concentration order kinetic and the Gompertz model,according to Eq.(3).Almost of phenolic compounds was found in SCB pretreatment liquid half of the total methane potential belongs to each kinetic model. (Table 2). Kinetic hydrolysis coefficients of first order reactions are the high- Methane production was limited by microorganism growth in est of all studied pretreatments,but the microorganisms require a the anaerobic degradation of WS pretreated whole slurry.The long lag period (19 days)for both substrates.Once the microorgan- Gompertz equation provided the best fit of these experimental isms are acclimated to the medium,very high biodegradation rates results,with a methane production potential higher than that of are also obtained from the Gompertz model fit. the untreated straw and a short lag period (around 4 days).How- The effect of inhibitory compounds released during pretreat- ever,the final methane production after 30 days of the experiment ment was evidenced for the four pretreatments studied for both was the same as that of the untreated WS.Therefore,alkaline pre- substrates but with different intensity.Experiments with low treatment increased the straw potential biodegradability by lignin inhibition fit to a first order kinetic model,but the hydrolysis rate solubilization,but this effect was counteracted by the reduction on coefficient decreases.Experiments with medium inhibition fit to hydrolysis rate by inhibition.Nevertheless,Reilly et al.(2015). the Gompertz model,since they are controlled by the microorgan- working with a whole slurry of Ca(OH)pretreated wheat straw ism growth.In these cases,after the lag period,when microorgan- (48 h,20C,7.4%(w/w)).obtained high anaerobic degradation isms are acclimated,a high degradation kinetic is usually achieved. rates.Their experiments achieved 202 mL-CH4/g VS after only Some combined models have been found,where a fraction of the 5 days of digestion. volatile solids is biodegraded by microorganisms able to grow in Anaerobic degradation of the alkaline pretreated solid fraction the inhibitory environment (first order kinetic)and the degrada- was fitted by a first order kinetic.Most of the inhibitory com- tion of another fraction of volatile solids requires microorganisms pounds were removed with the pretreatment liquid.The methane acclimation(Gompertz kinetic).Additional research work is neces- potential was close to that of the whole slurry pretreated material. sary in order to discover the methane production working in con- however lignin removal without inhibition remarkably increased tinuous bioreactors with acclimated microorganisms. the hydrolysis rate coefficient,obtaining the best methane produc- tion rate results.Liu et al.(2015)used the Gompertz model to fit cumulative methane production from the solid fraction of WS pre- 4.Conclusions treated with KOH.They obtained weak inhibition with low lag periods (1-2 days)for KOH loadings in the range from 2%to 50% Thermal pretreatment provided the lowest solubilization of w/w,for pretreatments at 20C during 48 h. sugars,VS and inhibitory compounds and the methane production The fit of cumulative methane production from alkaline pre- from whole slurry improved the raw material potential,with treated whole slurry SCB required a combination of both models. hydrolysis being the controlling step.Acid solubilized high concen- As observed for the solid fraction of acid pretreated SCB,only a trations of sugars,generating furfural and HMF with a strong inhi- small fraction of material was available for anaerobic degradation bition effect.Alkali solubilized lignin,generating acetic acid and from the starting time,according to a first order kinetic.Most of phenolic compounds;BMP experiments required an acclimation the microorganisms growth was inhibited,probably due to the period for whole slurry but obtained the highest methane produc- high acetic acid and phenolic compound concentration (Heiske tion from solid fractions.Alkaline-peroxide pretreatment caused et al..2013:Krishania et al..2013)and a lag period of 13 days higher inhibition in SCB than in WS,but high methane production was required.So.the Eq.(3)was used to fit the experimental rates were attained for solid fractions of both substrates after the results: acclimation period. B-Bo1·[1-exp(-kHt】+Bo2 Acknowledgements exp-exp me(-t)+1 (3) Bo2 This work was supported by the research unit UIC 071 of the regional government"Junta de Castilla y Leon-JCyL",Spain.Rodolfo In this case,Bo is the methane potential from microorganisms Travaini is also grateful to "Conselho Nacional de Desenvolvimento resistant to inhibitory compounds and Bo is the methane potential Cientifico e Tecnologico-CNPq",Brazil(238059/2012-0)for provid- from microorganisms able to acclimate in the pretreatment ing his Doctorate Scholarship.provided worse results than the untreated materials in terms of both methane potential and hydrolysis rate coefficient. Ferreira et al. (2013) used a first order equation to model the methane pro￾duction from untreated and steam exploded WS, demonstrating that hydrolysis was the limiting step. For steam exploded WS, at 150 C and 15 min, fitting parameters very similar to those from our thermal pretreatment, they obtained a methane potential of 194.6 mL CH4/g VS and a hydrolysis coefficient of 0.085 d1 . Very low values of methane production potential were obtained from the solid fraction of both acid pretreated materials. The degradation compounds retained in the solid fraction probably inhibited the growth of most of the microorganisms. However, the small quantity of methane production obtained fitted ade￾quately to a first order kinetic, limited by the hydrolysis step. The microorganism species that were able to grow, biodegraded the material with hydrolysis coefficients in a similar range to those for untreated materials. Alkaline pretreatment provided different results for both sub￾strates, probably attributable to the aforementioned differences in lignin structure and composition in tested materials. Basic pre￾treatment attacks mainly the lignin, and a higher concentration of phenolic compounds was found in SCB pretreatment liquid (Table 2). Methane production was limited by microorganism growth in the anaerobic degradation of WS pretreated whole slurry. The Gompertz equation provided the best fit of these experimental results, with a methane production potential higher than that of the untreated straw and a short lag period (around 4 days). How￾ever, the final methane production after 30 days of the experiment was the same as that of the untreated WS. Therefore, alkaline pre￾treatment increased the straw potential biodegradability by lignin solubilization, but this effect was counteracted by the reduction on hydrolysis rate by inhibition. Nevertheless, Reilly et al. (2015), working with a whole slurry of Ca(OH)2 pretreated wheat straw (48 h, 20 C, 7.4% (w/w)), obtained high anaerobic degradation rates. Their experiments achieved 202 mL-CH4/g VS after only 5 days of digestion. Anaerobic degradation of the alkaline pretreated solid fraction was fitted by a first order kinetic. Most of the inhibitory com￾pounds were removed with the pretreatment liquid. The methane potential was close to that of the whole slurry pretreated material, however lignin removal without inhibition remarkably increased the hydrolysis rate coefficient, obtaining the best methane produc￾tion rate results. Liu et al. (2015) used the Gompertz model to fit cumulative methane production from the solid fraction of WS pre￾treated with KOH. They obtained weak inhibition with low lag periods (1–2 days) for KOH loadings in the range from 2% to 50% w/w, for pretreatments at 20 C during 48 h. The fit of cumulative methane production from alkaline pre￾treated whole slurry SCB required a combination of both models. As observed for the solid fraction of acid pretreated SCB, only a small fraction of material was available for anaerobic degradation from the starting time, according to a first order kinetic. Most of the microorganisms growth was inhibited, probably due to the high acetic acid and phenolic compound concentration (Heiske et al., 2013; Krishania et al., 2013) and a lag period of 13 days was required. So, the Eq. (3) was used to fit the experimental results: B ¼ B01 ½1  expðkH tÞ þ B02 exp  exp Rm e B02 ðk  tÞ þ 1  
ð3Þ In this case, B01 is the methane potential from microorganisms resistant to inhibitory compounds and B02 is the methane potential from microorganisms able to acclimate in the pretreatment medium. The total methane potential for this experiment (B01 + B02) reached the theoretical methane yield for SCB. The same theoretical methane potential value was achieved using the solid fraction of alkaline pretreated SCB, applying the first order model. The removal of liquid pretreatment, as for WS reduced inhibition, identified hydrolysis as the limiting step. Data fitted to a first order kinetic resulted in a very high methane potential but with a low hydrolysis kinetic coefficient. A strong inhibitory effect was found for all the alkaline peroxide pretreated materials, probably generated by the degradation com￾pounds but also by some remaining peroxide. The Gompertz equa￾tion was used to fit the cumulative methane production from WS pretreated whole slurry with a methane potential very similar to that of the untreated material. After a lag period of 17 days, the inhibition was completely defeated, reaching the highest degrada￾tion rates. The 30 days period was not enough for microorganism acclimation in the SCB whole slurry experiment. The high inhibition of alkaline peroxide pretreatment also affected the solid fraction experiments. Cumulative methane pro￾duction from both materials was fitted to a combination of the first order kinetic and the Gompertz model, according to Eq. (3). Almost half of the total methane potential belongs to each kinetic model. Kinetic hydrolysis coefficients of first order reactions are the high￾est of all studied pretreatments, but the microorganisms require a long lag period (19 days) for both substrates. Once the microorgan￾isms are acclimated to the medium, very high biodegradation rates are also obtained from the Gompertz model fit. The effect of inhibitory compounds released during pretreat￾ment was evidenced for the four pretreatments studied for both substrates but with different intensity. Experiments with low inhibition fit to a first order kinetic model, but the hydrolysis rate coefficient decreases. Experiments with medium inhibition fit to the Gompertz model, since they are controlled by the microorgan￾ism growth. In these cases, after the lag period, when microorgan￾isms are acclimated, a high degradation kinetic is usually achieved. Some combined models have been found, where a fraction of the volatile solids is biodegraded by microorganisms able to grow in the inhibitory environment (first order kinetic) and the degrada￾tion of another fraction of volatile solids requires microorganisms acclimation (Gompertz kinetic). Additional research work is neces￾sary in order to discover the methane production working in con￾tinuous bioreactors with acclimated microorganisms. 4. Conclusions Thermal pretreatment provided the lowest solubilization of sugars, VS and inhibitory compounds and the methane production from whole slurry improved the raw material potential, with hydrolysis being the controlling step. Acid solubilized high concen￾trations of sugars, generating furfural and HMF with a strong inhi￾bition effect. Alkali solubilized lignin, generating acetic acid and phenolic compounds; BMP experiments required an acclimation period for whole slurry but obtained the highest methane produc￾tion from solid fractions. Alkaline-peroxide pretreatment caused higher inhibition in SCB than in WS, but high methane production rates were attained for solid fractions of both substrates after the acclimation period. Acknowledgements This work was supported by the research unit UIC 071 of the regional government ‘‘Junta de Castilla y León – JCyL”, Spain. Rodolfo Travaini is also grateful to ‘‘Conselho Nacional de Desenvolvimento Científico e Tecnológico – CNPq”, Brazil (238059/2012-0) for provid￾ing his Doctorate Scholarship. S. Bolado-Rodríguez et al. / Bioresource Technology 201 (2016) 182–190 189