绿竹甲酸预处理脱木素的响应面法优化

Optimizing Lignin Removal of Green Bamboo during Formic Acid Pretreatment by Response Surface Methodology

中文关键词:  甲酸预处理  响应面法优化  木素溶出率  纤维素酶水解  葡萄糖得率

Key Words:formic acid pretreatment  response surface methodology  lignin removal rate  cellulose enzymatic hydrolysis  glucose yield

中文摘要:

      利用Design-Expert统计软件对影响绿竹甲酸预处理过程中木素溶出率的4个自变量(甲酸含量、液比、最高温度和保温时间)进行四因素五水平的中心复合设计,建立全因素二项式方程(模型Ⅰ)。以参数的显著性为判定依据,对全因素二项式方程进行简化处理,得到模型Ⅱ。将这2个模型用于木素溶出率预测，尽管2个模型的变异系数相近,且预测值与实测值的相关系数均高于0.9,但模型Ⅱ的信噪比比模型Ⅰ高近20%,表明简化模型不仅可较好地预测绿竹甲酸预处理过程中的木素溶出率,且稳定性较高。木素溶出率与固体基质中纤维素酶水解转化率之间为指数关系；但随木素溶出率的增加,纤维素溶出率也随之增加,最终酶水解葡萄糖得率呈先升高后降低的变化趋势；为获得更多的葡萄糖,较优的木素溶出率约为50%。

Abstract:

      Central composite design (four-factor and five-level)of four independent variables influencing lignin removal rate(formic acid charge X1, the ratio of liquor to solid X2, pretreatment maximum temperature X3 and the pretreatment time at maximum temperature X4) during formic acid pretreatment of green bamboo was conducted by Design-expert software. Based on this design, a binomial equation with full factors (Model Ⅰ) was developed to predict lignin removal rate. By removing the parameters without influence on lignin removal, the amount of parameters in prediction model was reduced from 14 to 5 based on the significance of single and multi parameters. As a result, a simplified binomial equation was obtained (Model Ⅱ). The coefficient variations (C.V.) of these two prediction models were similar. In the two models, the square of coefficients (R2) of both measured lignin removal rate and prediction value were higher than 0.9. But the ratio of signal-to-noise (SNR) of the simplified equation was 20% higher, which indicating the effectiveness and stability of the simplified model for predicting lignin removal during formic acid pretreatment of green bamboo. The exponential relationship between lignin removal (RL) and cellulose enzymatic conversion (CEC) could be expressed as CEC=96.7+93.0/exp(0.04RL), where R2 was about 0.998. With the increase of RL, the cellulose removal (RC) also increased, but enzymatic hydrolysis glucose yield (YG) increased initially then reduced. The optimal RL was approximately 50%.