Bioresource Technology 135 (2013) 116–119

Contents lists available at SciVerse ScienceDirect

Bioresource Technology

journal homepage: www.elsevier .com/locate /bior tech

Short Communication

Comparative study of various pretreatment reagents on rice huskand structural changes assessment of the optimized pretreated rice husk

Teck Nam Ang, Gek Cheng Ngoh ⇑, Adeline Seak May ChuaDepartment of Chemical Engineering, Faculty of Engineering, University of Malaya, 50603 Kuala Lumpur, Malaysia

h i g h l i g h t s

" Among the pretreatment reagents, HCl hydrolyzed rice husk (RH) the best." The optimized pretreatment condition is mild compared with other similar studies." The increased pore volume & size of pretreated RH favors fungal fermentation.

a r t i c l e i n f o

Article history:Available online 24 September 2012

Keywords:Response surface methodologyOptimizationRice hullsPretreatmentStructural assessment

0960-8524/$ - see front matter � 2012 Elsevier Ltd. Ahttp://dx.doi.org/10.1016/j.biortech.2012.09.045

⇑ Corresponding author. Tel.: +60 3 7967 5286; faxE-mail address: ngoh@um.edu.my (G.C. Ngoh).

a b s t r a c t

The performance of alkalis (NaOH and Ca(OH)2) and acids (H2SO4, HCl, H3PO4, CH3COOH, and HNO3) inthe pretreatment of rice husk was screened, and a suitable reagent was assessed for subsequent optimi-zation using response surface methodology. From the assessment, HCl that hydrolysed rice husk well wasoptimized with three parameters (HCl loading, pretreatment duration, and temperature) using Box–Behnken Design. The optimized condition (0.5% (w/v) HCl loading, 125 �C, 1.5 h) is relatively mild, andresulted in �22.3 mg TRS/ml hydrolysate. The reduced model developed has good predictability, wherethe predicted and experimental results differ by only 2%. The comprehensive structural characterizationstudies that involved FT-IR, XRD, SEM, and BET surface area determination showed that the pretreatedrice husk consisted mainly of cellulose and lignin. Compared to untreated rice husk, pretreated rice huskpossessed increased pore size and pore volume, which are expected to be beneficial for fungal growthduring fermentation.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Rice husk is a by-product of rice milling industry, and it repre-sents approximately 20% by weight of rough rice (Hashim et al.,1996). In 2008, there were estimated 125 Mt and 137 Mt rice huskgenerated in Asia and worldwide, respectively (FAOSTAT, 2010).Rice husk contains relatively high cellulose content (40–60%), andit is widely available at relatively low cost. Attributing to its recal-citrant nature, direct conversion of untreated rice husk usually re-sults in low product yield. Therefore, pretreatment is necessary topartially disrupt the recalcitrant structure to achieve delignificationof the lignocellulosic biomass. This renders the solid substrate moreaccessible to enzyme or microorganism during bioconversion.

Prevailing rice husk pretreatments reported in literatures in-clude acid (Dagnino et al., 2013), alkaline (Saha and Cotta, 2008;Singh et al., 2011), hydrothermal (Zheng et al., 2007), alkaline per-oxide (Saha and Cotta, 2007), and ionic liquid dissolution pretreat-ments (Ang et al., 2011; Lynam et al., 2012). Among them, acid and

ll rights reserved.

: +60 3 7967 5371.

alkaline pretreatments have been extensively used for pretreatinglignocellulosic biomass (Kaar and Holtzapple, 2000; Saha andCotta, 2008).

To date, no specifically effective reagent in pretreating rice huskhas been reported. To have a greater insight into the pretreatmentof rice husk, alkalis and acids were screened and assessed in thestudy, and the best performing reagent was subjected to subse-quent optimization study. Various structural analyses includingFourier transform–infrared spectroscopy (FT-IR), X-ray diffraction(XRD), scanning electron microscopy (SEM), and Brunauer–Emmett–Teller (BET) surface area determination were conductedto assess the compositional, chemical, and structural changes ofthe biomass that is imparted by the selected pretreatment.

2. Methodology

2.1. Materials

Rice husk was collected from Ng Trading Company, Selangor,Malaysia. The sample was washed and dried at 55 �C before being

T.N. Ang et al. / Bioresource Technology 135 (2013) 116–119 117

milled to approximately 30 mesh size, and was stored in a dry cab-inet prior to use.

2.2. Assessment of pretreatment reagents

Screening was conducted on two alkalis and five acids, namelysodium hydroxide (NaOH, Merck), calcium hydroxide (Ca(OH)2,Sigma–Aldrich), sulphuric acid (H2SO4, Fisher Scientific), hydro-chloric acid (HCl, Merck), phosphoric acid (H3PO4, Ajax Chemicals),acetic acid (CH3COOH, Riedel–de Haen) and nitric acid (HNO3,Scharlau Chemie). With reference to the literatures (Chang et al.,1997; Saha and Cotta, 2008), the screening were carried out atthe following conditions: (i) concentration of reagent, 0.5% (w/v);(ii) rice husk loading, 10.0% (w/v); (iii) water loading, 10.0 ml/g;(iv) pretreatment temperature, 100 ± 1 �C; (v) pretreatment dura-tion, 2 h. Total reducing sugars (TRS) content was employed asan indirect pretreatment indicator, and its content in hydrolysatewas determined by DNS method (Miller, 1959). The main actionof acids and alkalis is to dissolve hemicellulose, to some extent cel-lulose, and lignin. Thus, TRS released from the hydrolysis of hemi-cellulose/cellulose can suitably reflect the extent of structuredisruption in rice husk. Pretreatment reagent that releases thehighest TRS in hydrolysate was selected for subsequent optimiza-tion studies. To investigate the compositional changes impartedby the pretreatment reagents, cellulose, hemicellulose, lignin, andash contents of the pretreated rice husk samples were character-ized by using the Association Official Analytical Chemists (AOAC,(2005)) official methods.

2.3. Optimization of pretreatment

The pretreatment reagent selected in Section 2.2 was employedin the optimization of rice husk pretreatment. Three parametersknown to have effect on pretreatment, namely reagent loading(X1), heating duration (X2) and heating temperature (X3), wereinvestigated. To determine the low and high levels of the chosenparameters, preliminary tests were conducted at: (i) reagent load-ing, 0–4% (w/v); (ii) heating duration, 1–6 h; (iii) heating tempera-ture, 60–140 �C. The tests were performed by one-factor-at-a-timeapproach, and TRS detected in hydrolysate was measured as re-sponse for optimization of pretreatment.

The range of each parameter determined from the preliminarytests was applied in the optimization study using Box–Behnkenexperimental design (BBD). TRS detected in the hydrolysate wasdetermined as response (Y). The maximum release of TRS in hydro-lysate was determined using response surface methodology (RSM),and the regression analysis of optimization data was performedwith the aid of Design-Expert Version 6.0.6 (Stat-Ease Inc.,Minneapolis).

2.4. Analytical techniques

The FT-IR spectra of the rice husk samples between 600 and4000 cm�1 at 4 cm�1 nominal resolution were recorded at roomtemperature with a FT-IR/FT–FIR spectrometer (Perkin Elmer,Spectrum 400, USA). The spectra were presented in relative trans-mittance percentage (%) of wave number (cm�1) and the back-ground was recorded with empty cell.

The crystallinity of the rice husk samples was examined by XRDmeasurement performed with a D8 Advanced X-ray diffractometer(Bruker AXS, USA) using Cu Ka monochromatized radiation at40 kV and 40 mA at ambient temperature. The samples werescanned and the intensities were recorded in 2h range from 10�to 80� with a step size of 0.02�. The crystallinity index (CrI) ofthe rice husk samples was calculated by using equation as reportedby Parikh et al. (2007).

The structural changes of the pretreated rice husk were as-sessed with scanning electron microscope Quanta™ 200 FESEM(FEI, USA) operated at 2–5 kV accelerating voltage under lowvacuum.

The surface area and average pore size of rice husk sampleswere determined by nitrogen adsorption isotherm at 77 K using ahigh-performance six-sample surface area and pore size analyzerAutosorb�-6B (Quantachrome, Florida, USA). The nitrogen adsorp-tion–desorption isotherm was operated at relative pressure P/P0 of0.3, where P is the system pressure and P0 is the initial pressure at1 bar.

3. Results and discussion

3.1. Assessment of pretreatment reagent

Based on the results of screening for suitable pretreatment re-agent, each alkali and acid had pretreated rice husk to a varying ex-tent. Low amount of TRS was detected in the hydrolysates of NaOHand Ca(OH)2 pretreatments. This is because the alkalis only frac-tionally hydrolysed hemicellulose and cellulose (Weil et al.,1994), but mainly delignified rice husk (Brannvall, 2004). Generally,findings from the assessment show that acids were better pretreat-ment reagents than alkalis (Fig. A1). The highest TRS was detectedin the hydrolysate of HCl pretreatment (15.0 ± 0.6 mg/ml), followedby HNO3 (12.2 ± 0.1 mg/ml) and H2SO4 (7.2 ± 0.3 mg/ml). Pretreat-ment with CH3COOH, H3PO4, NaOH, and Ca(OH)2 produced lessthan 1 mg TRS/ml hydrolysate, which is similar to pretreatmentusing only water.

The characterization of rice husk samples showed that all acid-pretreated rice husk had reduced hemicellulose content (Fig. 1),which explains the acids main role were hydrolysing the amor-phous hemicellulose in the substrate (Orozco et al., 2007). Besides,cellulose was also partially hydrolysed during the acid pretreat-ments of rice husk (Weil et al., 1994), particularly pretreatmentswith strong acids, such as HCl, HNO3 and H2SO4. The availabilityof more reactive protons disrupt hydrogen bonding of cellulosechain prior to hydrolysis resulting in higher TRS yield (Orozcoet al., 2007).

Among the reagents, HCl hydrolysed rice husk and released thehighest amount of TRS during pretreatment. This signifies theeffectiveness of HCl pretreatment in disrupting rice husk structure,while retaining significant amount of cellulose (�60%) in pre-treated rice husk. Thus, HCl pretreatment was further optimized.

3.2. Optimization of pretreatment

The preliminary pretreatment study shows that TRS yield in-creased sharply in the first 2 h of pretreatment and reached plateauafter 3 h (Fig. A2a). In HCl loading study, TRS yield increased shar-ply with HCl loading range between 0.25 and 0.75% (w/v) as moreprotons are available for the hydrolysis of rice husk (Fig. A2b).However, little increment in TRS yield was observed with HCl load-ing higher than 1.0% (w/v) signifies that the rate of hydrolysis islimited by the surface area of rice husk available for reaction. Fur-thermore, the TRS yield was found to increase proportionally withthe pretreatment temperature (Fig. A2c).

The low and high levels of HCl loading (X1), pretreatment dura-tion (X2), and temperature (X3) were determined in the preliminarystudies. The design matrix of BBD including the response (Y) is gi-ven in Table 1, where Y is the TRS detected in the hydrolysate. Fromthe runs, the highest TRS (23.9 mg/ml) was obtained with pretreat-ment conditions at 0.75% (w/v) HCl loading, 120 �C for 2 h. Theleast TRS (10.9 mg/ml) was detected when pretreatment was con-ducted at 0.25% (w/v) HCl loading, 100 �C for 2 h. Only minute

Fig. 1. Chemical composition of rice husk after acid and alkaline pretreatments (based on same initial sample weight).

118 T.N. Ang et al. / Bioresource Technology 135 (2013) 116–119

amount of furfural (�5% of TRS) was detected in the hydrolysates,whereas 5-(hydroxymethyl)furfural was not detected.

The Analysis of Variance (ANOVA) of the reduced quadraticmodel confirmed the significance and goodness of fit of the model(Table 2). The reduced model expressed in actual terms is given inEq. (1). The reduced model explained 95.2% of the variability in theoptimization pretreatment of rice husk.

Y ¼ �156:32þ 49:74X1 þ 9:89X2 þ 2:36X3 � 7:39X21

� 0:008X23 � 0:31X1X3 � 0:08X2X3 ð1Þ

Table 1Design matrix of BBD and response.

Run Parameter Response

X1 (%, w/v) X2 (h) X3 (�C) Y (mg/ml)

1 0.25 1 120 18.32 0.25 3 120 21.63 1.25 1 120 21.54 1.25 3 120 22.35 0.25 2 100 10.96 0.25 2 140 23.67 1.25 2 100 19.08 1.25 2 140 19.59 0.75 1 100 16.110 0.75 1 140 23.511 0.75 3 100 19.112 0.75 3 140 20.213 0.75 2 120 23.914 0.75 2 120 23.315 0.75 2 120 23.116 0.75 2 120 23.117 0.75 2 120 22.2

The optimization study shows that HCl loading and pretreat-ment temperature have greater effect compared to the pretreat-ment duration, and both these influencing parameters interactedin the pretreatment of rice husk as shown in Fig. 2. Higher temper-ature is needed to achieve the maximal TRS yield when lower HClloading is used, and vice versa. The TRS yield showed a quadraticdependence on the pretreatment temperature with high effect ofterm X3

2.Numerical optimization was conducted for the maximization of

TRS release, and the optimum pretreatment conditions were 0.5%(w/v) HCl loading, 125 �C for 1.5 h with a predicted TRS yield of22.8 mg/ml. Verification of the optimized pretreatment conditionconfirmed the TRS released at 22.3 ± 0.3 mg/ml, which is a mere2% discrepancy from the predicted value that further suggeststhe accuracy of the reduced model.

Table 2Analysis of variance (ANOVA) of quadratic model for optimization of rice huskpretreatment.

Source Sum ofsquare

Degree offreedom

Meansquare

F-value P-value

Model 173.04 7 24.72 25.44 <0.0001X1 7.88 1 7.88 8.11 0.0192X2 1.85 1 1.85 1.91 0.2007X3 59.46 1 59.46 61.19 <0.0001X1

2 14.43 1 14.43 14.85 0.0039X3

2 39.43 1 39.43 40.58 0.0001X1X3 37.33 1 37.33 38.42 0.0002X2X3 9.83 1 9.83 10.11 0.0112Residual 8.75 9 0.97Lack of fit 7.28 3 1.46 3.96 0.1034Pure error 1.47 4 0.37Corrected total 181.79 16

Fig. 2. Response surface plot of HCl loading and pretreatment temperature on TRSyield (pretreatment duration = 2 h).

T.N. Ang et al. / Bioresource Technology 135 (2013) 116–119 119

3.3. Characterization of pretreated rice husk

FT-IR investigation revealed that the pretreated rice husk exhib-its enhanced cellulosic content as shown by the intensified bandsat 1046 and 2924 cm�1 (Fig. A4). Besides, cellulose portion of thepretreated rice husk has more disordered structure, and ligninwas also observed in the HCl pretreated rice husk. These findingsare complemented by chemical composition analysis of the ricehusk samples. The pretreated rice husk consists of mainly cellulose(65.1%, w/w) and lignin (26.7%, w/w) which can be seen from theincreased intensity of relative transmittance of cellulose- and lig-nin-related bands in the FT-IR spectrum, whereas hemicellulosethat is easily hydrolysed was not detected.

Pretreated rice husk was slightly more crystalline than the un-treated rice husk after HCl pretreatment. The increased in the CrIfrom 51.5 of the untreated rice husk to 56.3 of the pretreated ricehusk is due to the removal of amorphous hemicellulose and cellu-lose from rice husk by HCl during pretreatment. The evidence forthe removal is confirmed by the SEM micrographs, whereby at1500� magnification, pretreated rice husk exhibited a smoothersurface than the untreated rice husk (Fig. A5).

Furthermore, the BET surface area analysis demonstrated thatthe specific surface area of the pretreated rice husk was signifi-cantly reduced from 21.2 to 9.6 m2/g, which is due to the hydroly-sis of amorphous hemicellulose/cellulose in rice husk and mergingof smaller pores into larger pores. The total pore volume and poresize of the rice husk after pretreatment had increased from1.76 � 10�2 to 1.81 � 10�2 cc/g and 33.2 to 75.2 Å, respectively.The pretreated rice husk with higher pore volume and bigger poresize is of advantageous for fungal growth during fermentation (Hsuet al., 2010), in spite of its slightly higher CrI and lower specific sur-face area compared to the untreated rice husk. Nevertheless, theunderlying relationship of structural features and biomass digest-ibility of the pretreated substrate depends on the size of enzymeused in the enzymatic hydrolysis, and the accessibility of microbesto substrate during fermentation.

4. Conclusion

HCl has been found to be more superior to the other investi-gated reagents, and its subsequent HCl pretreatment optimizationgives relatively mild optimal condition involving low HCl loading,low pretreatment temperature and short duration. The reducedmodel exhibits 2% discrepancy between the predicted and experi-mental values that suggests its accurate predictability. Structuralcharacterization analyses showed that the pretreated rice husk,which consisted mainly of cellulose and lignin, had increased inpore size and pore volume, which is favorable for the attachmentand growth of fungus during fermentation.

Acknowledgement

The authors thank University of Malaya for the financial sup-port (UMRG RG006/09AET and PS161/2010A). The authors alsothank Ng Trading Company, Selangor, Malaysia for providing therice husk sample.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.biortech.2012.09.045.

References

Ang, T.N., Yoon, L.W., Lee, K.M., Ngoh, G.C., Chua, A.S.M., Lee, M.G., 2011. Effiency ofionic liquids in the dissolution of rice husk. BioResources 6 (4), 4790–4800.

AOAC, 2005. Official Methods of Analysis of the Association of Official AnalyticalChemists. The Association Official Analytical Chemists, Washington.

Brannvall, E., 2004. Pulping technology. In: Ek, M., Gellerstedt, G., Henriksson, G.(Eds.), Pulp and Paper Chemistry and Technology, vol. 2. Walter de GruyterGmbH & Co, Berlin.

Chang, V.S., Burr, B., Holtzapple, M.T., 1997. Lime pretreatment of switchgrass. Appl.Biochem. Biotechnol. 63–65, 3–19.

Dagnino, E.P., Chamorro, E.R., Romano, S.D., Felissia, F.E., Area, M.C., 2013.Optimization of the acid pretreatment of rice hulls to obtain fermentablesugars for bioethanol production. Ind. Crop. Prod. 42, 363–368.

FAOSTAT, 2010. Food and Agricultural Commodities Production, vol. 2010, Food andAgriculture Organization of the United Nations.

Hashim, A.B., Aminuddin, H., Siva, K.B., 1996. Nutrient content in rice husk ash ofsome Malaysian rice varieties. Pertanika J. Trop. Agric. Sci. 19 (1), 77–80.

Hsu, T.C., Guo, G.L., Chen, W.H., Hwang, W.S., 2010. Effect of dilute acidpretreatment of rice straw on structural properties and enzymatic hydrolysis.Bioresource Technol. 101, 4907–4913.

Kaar, W.E., Holtzapple, M.T., 2000. Using lime pretreatment to facilitate the enzymichydrolysis of corn stover. Biomass Bioenerg. 18, 189–199.

Lynam, J.G., Reza, M.T., Vasquez, V.R., Coronella, C.J., 2012. Pretreatment of rice hullsby ionic liquid dissolution. Bioresource Technol. 114, 629–636.

Miller, G.L., 1959. Use of dinitrosalicylic acid reagent for determination of reducingsugar. Anal. Chem. 31, 426–428.

Orozco, A., Ahmad, M., Rooney, D., Walker, G., 2007. Dilute acid hydrolysis ofcellulose and cellulosic bio-waste using a microwave reactor system. ProcessSafety and Environmental Protection, Trans. I. Chem. E. Part B 85 (B5), 446–449.

Parikh, D.V., Thibodeaux, D.P., Condon, B., 2007. X-ray crystallinity of bleached andcrosslinked cottons. Textil. Res. J. 77 (8), 612–616.

Saha, B.C., Cotta, M.A., 2007. Enzymatic saccharification and fermentation ofalkaline peroxide pretreated rice hulls to ethanol. Enzyme Microb. Tech. 41,528–532.

Saha, B.C., Cotta, M.A., 2008. Lime pretreatment, enzymatic saccharification andfermentation of rice hulls to ethanol. Biomass Bioenerg. 32, 971–977.

Singh, A., Tuteja, S., Singh, N., Bishnoi, N.R., 2011. Enhanced saccharification of ricestraw and hull by microwave-alkali pretreatment and lignocellulolytic enzymeproduction. Bioresource Technol. 102, 1773–1782.

Weil, J., Westgate, P., Kohlmann, K., Ladisch, M.R., 1994. Cellulose pretreatments oflignocellulosic substrates. Enzyme Microb. Tech. 16, 1002–1004.

Zheng, G.J., Zhou, Y.J., Zhang, J., Cheng, K.K., Zhao, X.B., Zhang, T., Liu, D.H., 2007.Pretreatment of rice hulls for cellulase production by solid substratefermentation. J. Wood. Chem. Tech. 27, 65–71.