Optimization of microwave-hydrogen peroxide pretreatment of cellulose

Abstract

A hydrogen peroxide (H2O2) solution was adapted for microwave pretreatment of microcrystalline cellulose, which can be further used for heavy metal adsorption. The H2O2 concentration, temperature, and retention time were the key factors affecting the microwave/hydrogen peroxide pretreatment process. A Box-Benhken design (BBD) with response surface methodology (RSM) was employed to design and optimize the microwave-hydrogen peroxide pretreatment process (H2O2 pretreatment) of cellulose. After the H2O2 pretreatment, the crystallinity of cellulose decreased by 20% and the degree of polymerization (DP) decreased by up to 30%. The optimal conditions obtained by BBD were a H2O2 concentration of 8.37%, a temperature of 90 °C, and a retention time 5.33 min. Under these conditions, a minimum DP of 91.74 was achieved. The results indicated that all three of the factors notably affected the reduction of cellulose polymerization degree and pronounced interactions existed among the response variables. The predictive model developed was able to optimize the pretreatment process for the reduction of cellulose polymerization degree, which could improve the cellulose modification reactivity.

Full Article

Optimization of Microwave-Hydrogen Peroxide Pretreatment of Cellulose

Jingjing Su,^a,b Hongxiang Zhu,^a,b,* Lijun Wang,^a,b Xinliang Liu,^a,b Shuangxi Nie, ^a,b and Jianhua Xiong ^c,d

A hydrogen peroxide (H₂O₂) solution was adapted for microwave pretreatment of microcrystalline cellulose, which can be further used for heavy metal adsorption. The H₂O₂concentration, temperature, and retention time were the key factors affecting the microwave/hydrogen peroxide pretreatment process. A Box-Benhken design (BBD) with response surface methodology (RSM) was employed to design and optimize the microwave-hydrogen peroxide pretreatment process (H₂O₂ pretreatment) of cellulose. After the H₂O₂pretreatment, the crystallinity of cellulose decreased by 20% and the degree of polymerization (DP) decreased by up to 30%. The optimal conditions obtained by BBD were a H₂O₂concentration of 8.37%, a temperature of 90 °C, and a retention time 5.33 min. Under these conditions, a minimum DP of 91.74 was achieved. The results indicated that all three of the factors notably affected the reduction of cellulose polymerization degree and pronounced interactions existed among the response variables. The predictive model developed was able to optimize the pretreatment process for the reduction of cellulose polymerization degree, which could improve the cellulose modification reactivity.

Keywords: Hydrogen peroxide; Microwave; Response surface methodology (RSM); Cellulose pretreatment

Contact information: a: College of Light Industry and Food Engineering, Guangxi University, Nanning 530004, China; b: Guangxi Key Laboratory of Clean Pulp & Papermaking and Pollution Control, Nanning 530004, China; c: School of Environment, Guangxi University, Nanning 530004, China; d: Department of Paper and Bioprocess Engineering, State University of New York – College of Environmental Science and Forestry, Syracuse, NY, 13210; *Corresponding author: zhx@gxu.edu.cn

INTRODUCTION

Because of increasing concerns about energy consumption and environmental protection, cellulose is becoming a more commonly used renewable resource material. However, considerable hydrogen bonding in the intramolecular and intramolecular domains of the cellulose molecule and the complexity of the cellulose morphology and aggregate structures decrease its solubility with reagents; thus, cellulose exhibits low reactivity and poor uniformity during chemical reactions. A reduction in the crystallinity of cellulose must be achieved by various methods to improve the availability and accessibility of cellulose to reagents. There are several methods involving the physical and chemical pretreatment of cellulose (Zhao et al. 2006; Moharram and Mahmoud 2008; Peng et al. 2013). The most popular pretreatment incorporates physical and chemical methods, including microwave alkali treatment and ultrasonic alkali treatment (Khajavi et al. 2013; Peng et al. 2013; Ni et al. 2014, 2015). However, these pretreatment methods require large quantities of alkali and a long processing duration of 12 to 16 h (Gurgel et al. 2008a,b; Hokkanen et al. 2013). Presently, research is focusing on the development of a pretreatment method that combines a reduction in crystallinity with an improvement in process efficiency and environmental protection.

Hydrogen peroxide (H₂O₂) is a type of chemical oxidant that is referred to as a green oxidant (Wang et al. 2009; Sheldon 2015). Some resistant materials can be oxidized using oxidation products containing only water and oxygen, which are compounds that result in minimal environmental damage (Menova and Cibulka 2012; Hartman et al. 2015). The standard oxidation reduction potential of H₂O₂ is 1.8 V, which increases to 2.8 V after treatment with additional reagents (potassium permanganate, hypochlorite, and chlorine dioxide), producing the hydroxyl radical, ·OH. The hydroxyl radical has a strong oxidizing potential. Microwave pretreatment offers an alternative method to traditional heat-treatment, with the advantages of heating rapidly and minimal energy losses (Li and Xu 2013; Tyagi and Lo 2013; Ni et al. 2014). Similarly, microwave irradiation (with H₂O₂) is characteristically well-distributed, efficient, and highly penetrable (Hou et al. 2008; Hashem et al. 2014). Accordingly, this method is recommended for improving sludge disintegration (Wang et al. 2009), the hydrolysis of cellulose (Ni et al. 2015), and the pretreatment of cotton fabrics (Hashem et al. 2014). However, microwave irradiation (with H₂O₂), as a pretreatment of cellulose to obtain a modified material, has yet to be fully investigated.

Response surface methodology (RSM) is a useful tool based on statistical analysis that can be used to construct models and evaluate their influence on multiple factors and their interactions (Ya-Wei et al. 2015). Therefore, RSM has been employed in many studies as a powerful tool to explore the interactions among multiple factors. Predictive models, which are the most representative, have been used to analyze and optimize the operation parameters in many fields, including chemical (Adalarasan et al. 2015), medical (Chojnicka-Paszun and de Jongh 2014), and energy (Antonopoulou et al. 2012; Chen et al. 2012). This process could result in the development of desirable responses and reduce the number of experiments required. In comparison with the central composite design, the Box-Behnken design requires relatively few experiments and offers higher efficiency. The Box-Behnken design requires not all influencing factors to be at a high level at the same time and ensures that all the experimental conditions are within a safe operating area (Ragonese et al. 2002; Kehoe and Stokes 2011).

The present study investigated a pretreatment of cellulose using H₂O₂ and microwave irradiation. Response surface methodology was employed to determine the effect of the H₂O₂ concentration, temperature, and retention time on cellulose and to estimate the interactions among these factors. The RSM model contained 17 response variables to construct the optimal pretreatment conditions.

EXPERIMENTAL

Materials

Microcrystalline cellulose (Aladdin Inc., Shanghai, China), hydrogen peroxide (Chengdu Kelon Chemical Reagent Factory, Chengdu, China), and copper ethylene diamine solution (Paper Research Institute of China) were obtained for use in the study. A microwave accelerated reaction system (MARS6) was obtained from CEM Corp., North Carolina, USA.

Methods

To explore the optimal value of the influencing factors, a single-factor experiment was employed to determine the initial scope (Table 1). The Box-Benhken design (BBD) method was used to explore the importance of temperature, H₂O₂ concentration, and retention time on the cellulose polymerization degree. In theory, the decreasing trend of DP and crystallinity corresponding to the change of the pretreatment condition was consistent, and the decreasing trend of DP was more obvious. Therefore, the DP was selected as the response value. In comparison with the central composite design method, the BBD method requires fewer experimental parameters, which results in increased efficiency (de Almeida Borges et al. 2013). The other important reason for choosing BBD was that BBD avoids “the corner points”, meaning experiments were avoided for all factors at extreme low or high levels, which ensures the operation safety. A three-level, three-factorial Box-Benhken design (BBD) was programmed using Design-Expert 8.0.6 software (Statease Inc., Minneapolis, USA). The H₂O₂ concentration (X₁), temperature (X₂), and retention time (X₃) were adjusted over a range of 6% to 12%, 80 to 100 °C, and 0 to 10 min, respectively, with corresponding optimal values of 9% (X₁), 90 °C (X₂), and 5 min (X₃). The X_i represented the actual values of the three variables (X_1, X₂, and X₃), according to the experimental design (Table 2). The parameters were standardized according to Eq. 1 (Li and Xu 2013),

 i = 1, 2, 3 (1)

where X_i is the coded value of the independent variable, X_i is the actual value of the independent variable, X₀ is the actual value of the independent variable at the center point, and ∆X_i is the step change of X_i corresponding to a unit variation of the dimensionless value.

Table 1. Treatment Levels

Table 2. Matrix of the Experimental Design

Each response variable Y was assessed as a function of three, first-order effects (X₁, X₂, and X₃), three interaction effects (X₁*X₂, X₁*X₃, and X₂*X₃), and three, second-order effects (X₁², X₂², and X₃²), and can be described according to Eq. 2,

where Y is the predicted response, b₀ is the intercept coefficient, b_i is the linear term, b_ii is the squared effects term, and b_ij is the interaction term.

Data obtained from BBD for the optimization of the pretreatment conditions were used to generate regression coefficients for the second-order multiple regression models. Analysis of variance (ANOVA) was performed using Design-Expert 8.0.6 software. The coefficient of variation and the statistical significance were used to assess the quality of the fit to the polynomial model equation. Regression coefficient significance was assessed using F- and t-test parameters. The optimum values of the selected variables were obtained by analyzing the contour plot of the response surface and then solving for the regression equation.

A microwave reactor from the microwave accelerated reaction system was used for the cellulose pretreatment process. At full power, the microwave reactor delivered approximately 1800 W of microwave energy at a frequency of 2450 MHz and was controlled by a microcomputer that monitored the operations. According to the experimental design, the pretreatment was performed in a digestion tank with H₂O₂ solution (solid-liquid ratio 1:20) and microcrystalline cellulose at 80 to 90 °C for 0 to 15 min. The microwave reactor was digitally programmed to control the experimental conditions, which included a detection system to measure real-time temperature and power. After the reaction was complete, the mixture was cooled, centrifuged, and dried to a constant weight in an oven. Lastly, the microcrystalline cellulose (MCC) was treated with alkali (Hokkanen et al. 2013).

The degree of polymerization (DP) of the cellulose was determined according to the GB/T1548 (1989) (Halidan Manat et al. 2006) testing standard. The specific method to dissolve the cellulose in copper (Cu) ethylene diamine solvent involved mixing fully to obtain a certain concentration solution (0.5 g/L). The viscosity was determined using a Nordic standard viscometer (Beijing, China), and then the DP was calculated by Eq. 3,

where [η] is the intrinsic viscosity.

The crystallinity index (CrI) was determined by the diffracted intensity of Cu radiation (1.54 Å), using an X-ray diffractometer (X-ray double crystal powder diffractometer, Smart Lab, RIGAKU, Japan). The operating voltage and current were set at 40 kV and 30 mA, respectively. The diffracted intensity was measured at a scan rate of 8°/min for 2 h ranging from 10° to 50°, and the step size was 0.012°. Before the tests, the samples were dried to a constant weight at 50 °C. The CrI is an indicator of the amount of crystalline versus amorphous structures, calculated from the diffracted intensity data. This method assumes a two-phase structure (crystalline amorphous) and a line between the intensity minima to obtain an arbitrary background to the diffraction trace, thus separating an arbitrary crystalline phase from an arbitrary amorphous phase (Ju et al. 2015). The CrI was calculated using crystallinity analysis software MDI Jade 5(Materials Data Ltd, USA), according to Eq. 4 (Focher et al. 2001),

where I_c is the intensity of the diffraction from the crystalline region and I_B is the intensity of the diffraction of the amorphous area.

RESULTS AND DISCUSSION

Quadratic Models for All Response Variables

The model showed that the DP was a function of X₁, X₂, and X₃. The 17 variables were assessed using ANOVA second-order equations that incorporated the interactions between the three variables and the coefficients for each variable and fitted them to quadratic models. The ANOVA F-value described the significance of individual factors and their interactions, whereas the P-value illustrates the significance of the coefficient. The F-value of 1054 was large enough to signify that the model was significant. There was a 0.01% chance that the highly significant F-value could have occurred from noise. The original model from the Design-Expert 8 software, including the interaction terms, X₁*X₂, X₁*X₃, and X₂*X₃, indicated that the model terms were significant. In this case X₁, X₂, X₃, X₁*X₂, X₁*X₃, X₂*X₃, X₁², X₂², X₃² resulted in significance. Model parameters with a P-value less than 0.0500 indicate model terms are significant. Values greater than 0.1000 indicate the model terms are not significant. If there are many insignificant model terms (not counting those required to support hierarchy), model reduction may improve your model. The ANOVA results are shown in Table 3.

A ‘lack-of-fit F-value’ of 0.67 signified that the model was not significant, which was related to the error rate. According to the model, there was a 61% chance that the F-value size occurred from noise.

The DP was predicted based on the results obtained from the simulation conducted and Eq. 2. The model in Table 3 was statistically significant. The sequence of influence of the three factors on the degree of polymerization was X₁ > X₂ > X₃. Figure 1 illustrates that the error between the actual value and predicted value was minimal, and the actual value and predicted value on both sides was evenly dispersed diagonally, suggesting that the equation well-described the model. In addition, the regression coefficient was high (R² = 0.99), indicating that the model successfully described the correlation relationship between the factors and the DP (Shukla et al. 2014). Therefore, a quadratic model was chosen in accordance with the experimental results.

Usually, if the coefficient of variation (CV) of a model is less than 10%, then the model has good repeatability. For the DP, the coefficient of variation was 0.41%, indicating that the model exhibited an excellent ability to predict the observed data. The adequate precision value (Table 3) was employed as a measure of the signal-to-noise ratio. The precision value compares a range of predicted values at the design points to the average prediction error. A precision value of 88.497 indicated that an adequate signal was produced (a ratio > 4 is considered desirable) (Ahmadi et al. 2005). Therefore, this model can be used to navigate the design space. The coefficients of the quadratic model were calculated using least-squared means, multiple linear regression analysis, and the model’s goodness-of-fit, and were evaluated according to the R² coefficient. The value of square of the correlation coefficient (R²) was 0.99, shows that the quadratic model described the experimental result well. Because of the high R² coefficient, the resulting quadratic model can be considered suitable for describing the relationship among H₂O₂ concentration (X₁), temperature (X₂), and retention time (X₃). The validity of BBD was checked by predicted values versus actual values and Normal Plot of Residuals (Fig. 1)

According to Eq. 2 and Table 3, Eq. 5 and Eq. 6 were produced. Eq. 5 was the final equation in terms of Coded Factors. Eq. 6 was the final equation in terms of actual factors.

(5)

(6)

Equation 6 represents the actual factors in their original units. According to Eq. 3, the highest coefficient value resulted from the relationship between H₂O₂ concentration and DP, indicating that DP was dependent upon the H₂O₂ concentration. Moreover, the values of b₁ were positive, suggesting that the value of DP increased with increasing H₂O₂ concentration, while the opposite was true for X₂ and X₃. The value of b₁₃ was greater than b₁₂,₂₃, indicating that the effects of the interaction between X₁ and X₃ was more pronounced. This was supported by the F-values obtained when the interaction between the variables was considered (Table 3).

Table 3. ANOVA of the Surface Quadratic Model of the Degree of Polymerization

Fig. 1. Predicted values versus actual values for the polymerization degree (DP) (a); Normal Plot of Residuals for the polymerization degree (DP) (b).

Pretreatment Condition Optimization

The response surface plot and the contour map of DP are shown in Fig. 2. The DP of cellulose was dependent on X₁, X₂, and X₃. In comparison to the density of the contour plot in Fig. 2, the H₂O₂ concentration and microwave processing temperature exhibited the greatest influence on the degree of polymerization. This result was in agreement with the ANOVA results and the regression equation: the F-value of X₁*X₃ was greater than X₂*X₃. According to Eq. 5, the linear programming function was used to obtain the best pretreatment conditions, which included an H₂O₂ concentration of 8.37%, a temperature of 90 °C, and a retention time of 5.32 min. The minimum value of DP, obtained from these experimental conditions, was 91.74.

A)

B)

C)

Fig. 2. Three-dimensional response surface plots for the degree of polymerization:

A) H₂O₂ concentration and temperature; B) H₂O₂ concentration and retention time; and C) temperature and retention time

Cellulose Crystallinity

Crystallinity values pre- and post-treatment are listed in Fig. 3 and Table 4. Cellulose peaks, including cellulose I and cellulose II, appeared at different peak positions. The peaks of cellulose I were located at approximately 22.6°, 16.2°, and 22.6°. The peak at 16.2° was very close to the peak at 14.8°, forming a relatively wide diffraction peak. Meanwhile, the peaks of cellulose II were located at approximately 12°, 19.9°, and 21.7° (Nam et al. 2016).

Table 4. Crystallinity (CrI) of Untreated and Microwave-Hydrogen Peroxide-Pretreated Cellulose

Fig. 3. X-ray diffraction patterns of A) untreated and B) microwave-hydrogen peroxide-pretreated cellulose

The characteristic peaks of the untreated microcrystalline cellulose were located at 2θ = 14.7°, 16.8°, 22.8°, and 34.7°. The diffractogram revealed a relatively ordered structure, with a narrow peak at 22.8° and a diffuse peak between 14.8° and 16.2°. The sharper diffraction peak at 22.8° indicates a region of higher crystallinity. Linear cellulose molecules are linked laterally by hydrogen bonds to form linear bundles, giving rise to a crystalline structure. Individual fibrillary units consist of long periods of ordered crystalline regions interrupted by disordered amorphous regions. The pretreatment of cellulose causes a rearrangement of the crystal packing of chains from native cellulose I (chains aligned in parallel) to cellulose II (anti-parallel arrangement). This change is irreversible and is normally accompanied by a decrease in crystallinity. The microwave-hydrogen peroxide pretreatment loosened the hydrogen bonding within the fiber bundles, causing the microfibers to move apart. After the pretreatment process, the crystallinity of cellulose decreased 20%, resulting from the activation of cellulose and the increase in accessibility of the fiber.

Degree of Polymerization of Cellulose

The results obtained in this study are listed in Table 5. The degree of polymerization for cellulose was determined by the Cu ethylene diamine solution viscosity method after the pretreatment process of cellulose. After processing, the degree of polymerization for cellulose declined, especially for the optimal value group (Runs 2, 8, 11, 12, and 17), by up to 30% (degree of polymerization for untreated microcrystalline cellulose was 132.6), which was similar to the effect of alkali pretreatment (NaOH 20 wt. %) with ultrasound for 12 to 16 h. The decrease of cellulose DP was mainly due to the oxidation of hydroxyl free radical. Hydroxyl free radical produced by the decomposition of hydrogen peroxide could cut off the long chain of cellulose and then reduce the DP. Low concentration (H₂O₂) and temperature were of no advantage to the reaction. High concentration would produce excess H₂O₂, and high temperature can achieve the response quickly though not completely. In a microwave frequency alternating electromagnetic field, the chance of cellulose hydrogen bond fracture was increased with the increasing of collision frequency between the molecules. However, with the extension of reaction time, a lot of heat would be generated within the molecules as a result of microwave heating. Then the hydrogen bonds which have been broken could form again, and thus the results backfire. Therefore, identifying the tendency of the decrease of DP and optimizing of experimental conditions were necessary.

Table 5. Degree of Polymerization (DP) of Pretreated Cellulose

CONCLUSIONS

1. A quadratic model was established, according to RSM, based on the best fit to the experimental data. The model yielded a high coefficient of determination, indicating that the model was capable of predicting an optimized treatment process.
2. The effects of the three factors on the degree of polymerization for microwave-hydrogen peroxide pretreated cellulose were X₁ > X₂ > X₃, and the optimal conditions obtained by BBD were an H₂O₂ concentration of 8.37%, a temperature of 90 °C, and a retention time of 5.33 min. The crystallinity and DP values decreased by 20% and up to 30%, respectively.
3. In comparison with the alkali pretreatment, the microwave-hydrogen peroxide pretreatment obtained a similar effect. The optimal conditions required a shorter treatment time (5 min) and a lower reagent dosage (8.37%).

ACKNOWLEDGMENTS

This project was sponsored by the Guangxi Youth Natural Science Fund (GXNSFBA053025), the National High Technology Research and Development Program (“863″Program) of China (2009AA06A416), the National Natural Science Foundation of China (51108261), the Guangxi Science and Technology Research Program (14251009), and the Natural Science Foundation of Guangxi (2013GXNSFFA019005).

REFERENCES CITED

Adalarasan, R., Santhanakumar, M., and Rajmohan, M. (2015). “Application of grey Taguchi-based response surface methodology (GT-RSM) for optimizing the plasma arc cutting parameters of 304L stainless steel,” The International Journal of Advanced Manufacturing Technology 78(5-8), 1161-1170. DOI: 10.1007/s00170-014-6744-0

Ahmadi, M., Vahabzadeh, F., Bonakdarpour, B., Mofarrah, E., and Mehranian, M. (2005). “Application of the central composite design and response surface methodology to the advanced treatment of olive oil processing wastewater using Fenton’s peroxidation,” Journal of Hazardous Materials 123(1-3), 187-195. DOI: 10.1016/j.jhazmat.2005.03.042

Antonopoulou, M., Papadopoulos, V., and Konstantinou, I. (2012). “Photocatalytic oxidation of treated municipal wastewaters for the removal of phenolic compounds: Optimization and modeling using response surface methodology (RSM) and artificial neural networks (ANNs),” Journal of Chemical Technology and Biotechnology 87(10), 1385-1395. DOI: 10.1002/jctb.3755

Chen, G., Li, J., Song, C., Suo, Y., and You, J. (2012). “A sensitive and efficient method for simultaneous trace detection and identification of triterpene acids and its application to pharmacokinetic study,” Talanta 98, 101-111. DOI: 10.1016/j.talanta.2012.06.053

Chojnicka-Paszun, A., and de Jongh, H. H. J. (2014). “Friction properties of oral surface analogs and their interaction with polysaccharide/MCC particle dispersions,” Food Research International 62, 1020-1028. DOI: 10.1016/j.foodres.2014.05.028

de Almeida Borges, V. R., Ribeiro, A. F., Anselmo, C. D. S., Cabral, L. M., and de Sousa, V. P. (2013). “Development of a high performance liquid chromatography method for quantification of isomers beta-caryophyllene and alpha-humulene in copaiba oleoresin using the Box-Behnken design,” Journal of Chromatography B 940, 35-41. DOI: 10.1016/j.jchromb.2013.09.024

Focher, B., Palma, M. T., Canetti, M., Torri, G., Cosentino, C., and Gastaldi, G. (2001). “Structural differences between non-wood plant celluloses: Evidence from solid state NMR, vibrational spectroscopy and X-ray diffractometry,” Industrial Crops & Products 13(3), 193-208. DOI: 10.1016/S0926-6690(00)00077-7

Gurgel, L. V., Junior, O. K., Gil, R. P., and Gil, L. F. (2008a). “Adsorption of Cu(II), Cd(II), and Pb(II) from aqueous single metal solutions by cellulose and mercerized cellulose chemically modified with succinic anhydride,” Bioresource Technology 99(8), 3077-3083. DOI: 10.1016/j.biortech.2007.05.072

Gurgel, L. V. A., Freitas, R. P. D., and Gil, L. F. (2008b). “Adsorption of Cu(II), Cd(II), and Pb(II) from aqueous single metal solutions by sugarcane bagasse and mercerized sugarcane bagasse chemically modified with succinic anhydride,” Carbohydrate Polymers 74(4), 922-929. DOI: 10.1016/j.carbpol.2008.05.023

Halidan, M., Nurmmamat., and Wumanjiang, E. (2006). “Determination of polymerization degree of plant cellulose by viscosimetry,” China Synthetic Fiber Industry 29(1), 40-42.

Hartman, T., Sturala, J., and Cibulka, R. (2015). “Two-phase oxidations with aqueous hydrogen peroxide catalyzed by amphiphilic pyridinium and diazinium salts,” Advanced Synthesis & Catalysis 357(16-17), 3573-3586. DOI: 10.1002/adsc.201500687

Hashem, M., Abou Taleb, M., El-Shall, F. N., and Haggag, K. (2014). “New prospects in pretreatment of cotton fabrics using microwave heating,” Carbohydrate Polymers 103, 385-391. DOI: 10.1016/j.carbpol.2013.11.064

Hokkanen, S., Repo, E., and Sillanpää, M. (2013). “Removal of heavy metals from aqueous solutions by succinic anhydride modified mercerized nanocellulose,” Chemical Engineering Journal 223, 40-47. DOI: 10.1016/j.cej.2013.02.054

Hou, A., Wang, X., and Wu, L. (2008). “Effect of microwave irradiation on the physical properties and morphological structures of cotton cellulose,” Carbohydrate Polymers 74(4), 934-937. DOI: 10.1016/j.carbpol.2008.05.011

Ju, X., Bowden, M., Brown, E. E., and Zhang, X. (2015). “An improved X-ray diffraction method for cellulose crystallinity measurement,” Carbohydrate Polymers 123, 476-481. DOI: 10.1016/j.carbpol.2014.12.071

Kehoe, S., and Stokes, J. (2011). “Box-Behnken design of experiments investigation of hydroxyapatite synthesis for orthopedic applications,” Journal of Materials Engineering & Performance 20(2), 306-316. DOI: 10.1007/s11665-010-9671-8

Khajavi, R., Atlasi, A., and Yazdanshenas, M. -E. (2013). “Alkali treatment of cotton yarns with ultrasonic bath,” Textile Research Journal 83(8), 827-835. DOI: 10.1177/0040517512467077

Li, H., and Xu, J. (2013). “Optimization of microwave-assisted calcium chloride pretreatment of corn stover,” Bioresource Technology 127, 112-118. DOI: 10.1016/j.biortech.2012.09.114

Menova, P., and Cibulka, R. (2012). “Insight into the catalytic activity of alloxazinium and isoalloxazinium salts in the oxidations of sulfides and amines with hydrogen peroxide,” Journal of Molecular Catalysis A: Chemical 363-364, 362-370. DOI: 10.1016/j.molcata.2012.07.012

Moharram, M. A., and Mahmoud, O. M. (2008). “FTIR spectroscopic study of the effect of microwave heating on the transformation of cellulose I into cellulose II during mercerization,” Journal of Applied Polymer Science 107(1), 30-36. DOI: 10.1002/app.26748

Nam, S., French, A. D., Condon, B. D., and Concha, M. (2016). “Segal crystallinity index revisited by the simulation of X-ray diffraction patterns of cotton cellulose I beta and cellulose II,” Carbohydrate Polymers 135, 1-9. DOI: 10.1016/j.carbpol.2015.08.035

Ni, J., Na, H., She, Z., Wang, J., Xue, W., and Zhu, J. (2014). “Responsive behavior of regenerated cellulose in hydrolysis under microwave radiation,” Bioresource Technology167, 69-73. DOI: 10.1016/j.biortech.2014.05.066

Ni, J., Teng, N., Chen, H., Wang, J., Zhu, J., and Na, H. (2015). “Hydrolysis behavior of regenerated celluloses with different degree of polymerization under microwave radiation,” Bioresource Technology 191, 229-233. DOI: 10.1016/j.biortech.2015.05.036

Peng, H., Li, H., Luo, H., and Xu, J. (2013). “A novel combined pretreatment of ball milling and microwave irradiation for enhancing enzymatic hydrolysis of microcrystalline cellulose,” Bioresource Technology 130, 81-87. DOI: 10.1016/j.biortech.2012.10.167

Ragonese, R., Macka, M., Hughes, J., and Petocz, P. (2002). “The use of the Box-Behnken experimental design in the optimisation and robustness testing of a capillary electrophoresis method for the analysis of ethambutol hydrochloride in a pharmaceutical formulation,” Journal of Pharmaceutical & Biomedical Analysis 27(6), 995-1007. DOI: 10.1016/S0731-7085(01)00659-8

Sheldon, R. A. (2015). “Recent advances in green catalytic oxidations of alcohols in aqueous media,” Catalysis Today 247, 4-13. DOI: 10.1016/j.cattod.2014.08.024

Shukla, P., Bajwa, U., and Sharma, S. (2014). “Optimization of the level of ingredients for functional dairy beverage using response surface methodology (RSM),” Acta Alimentaria 43(3), 402-411. DOI: 10.1556/AAlim.43.2014.3.6

Tyagi, V. K., and Lo, S.-L. (2013). “Microwave irradiation: A sustainable way for sludge treatment and resource recovery,” Renewable & Sustainable Energy Reviews 18, 288-305. DOI: 10.1016/j.rser.2012.10.032

Wang, Y., Wei, Y., and Liu, J. (2009). “Effect of H₂O₂ dosing strategy on sludge pretreatment by microwave-H₂O₂ advanced oxidation process,” Journal of Hazardous Materials 169(1-3), 680-684. DOI: 10.1016/j.jhazmat.2009.04.001

Ya-Wei, W., Cheng-Min, G., Xiao-Tang, N., Mei-Xue, C., and Yuan-Song, W. (2015). “Multivariate analysis of sludge disintegration by microwave-hydrogen peroxide pretreatment process,” Journal of Hazardous Materials 283, 856-864. DOI: 10.1016/j.jhazmat.2014.10.022

Zhao, H. B., Kwak, J. H., Wang, Y., Franz, J. A., White, J. M., and Holladay, J. E. (2006). “Effects of crystallinity on dilute acid hydrolysis of cellulose by cellulose ball-milling study,” Energy & Fuels 20(2), 807-811. DOI: 10.1021/ef050319a

Article submitted: April 23, 2016; Peer review completed: June 29, 2016; Revisions accepted: July 6, 2016; Published: July18, 2016.

DOI: 10.15376/biores.11.3.7416-7430