Mid-infrared Spectroscopy Monitoring Of Bio-ethanol Fermentation Process:determination Of Ethanol And Glucose

Zymomonas mobilis, which is a native bacterium for ethanol, has a high tolerance for glucose (about 40 g/L) and ethanol (about 9.5 g/L). So it has a high yield of ethanol. From other researchers'work, we find that it has a special way of glycometabolism and energy metabolism. Entner-Doudoroff is a major way for Z. mobilis metabolism, which is a unique way, its intermedial metabolites link with TCA. Infrared spectrum (IR) is a quick and precise analytical technique, is used bio-analysis widely especially for turbid system. We use this technique in Z. mobilis's bio-ethanol fermentation to detect the content of bio-ethanol. ATR cell solve the problem of infrared spectrum used in quantitative analysis, it makes absorbance power of spectrum have nothing to do with the samples. So it makes to come true the quantitative analysis of ethanol, without the influence of sample content.Summarize the research of former laboratory work, in our article the work focal point is put in simultaneously quantitative analysis glucose and ethanol, without the interference of water, bacterium and some major media in the broth. Scan these substances of infrared spectrum for all bands from 4000 cm^-1 to 650 cm^-1, in order to make sure the wave number in the quantitative analysis. From these works we make 1200 cm^-1 to 954 cm^-1 as the finally analysis range. On these grounds we establish mathematics split model, to simulate the peak position of glucose and ethanol. At last we split the absorbance of glucose and ethanol successfully with fitting rate above 95%. The initial IR spectrum in the broth is used in predicting the content of ethanol and glucose in the bio-ethanol fermentation, without a simple porization approach, derivative spectroscopy, matrix simulation, etc. In the study, original spectrum of the broth and the spectra of pure components have been adopted to content analysis. In this study, we do the validation based on the IR mathematics split model. And Gas chromatography and glucose oxidase method is chosen as reference. Finally, the standard curve of glucose and ethanol are: the peak area (1200 cm?1to 954 cm^-1) = 0.037871+1.955 [Glucose], Absorbance (1045 cm^-1) = 0.0059549+0.0063644 [Ethanol]. The calibration linear range (infrared spectrum) is from 16-100 g/L (glucose) and 6.32-79 g/L (ethanol), respectively, and coefficient correlation (R²) are 0.9999. The standard errors of correction are 0.2561 g/L (glucose) and 0.6487 g/L (ethanol), the standard errors of validation are 0.358 g/L (glucose) and 1.0357 g/L (ethanol). The standard errors of prediction are 2.688 g/L (glucose) and 1.659 g/L (ethanol) for the synthetic samples, respectively. For the real fermentation, they are 12.367 g/L (glucose) and 1.9612 g/L (ethanol), respectively. The standard errors of prediction is 12.367 g/L for glucose in the real fermentation, the value is much bigger than synthetic samples. So we compare the results of infrared spectrum and glucose oxidase method. We choose 6 group sample to analyze these two methods (weighting method as the standard validation method), the glucose oxidase method equation describing the simple calibration line is Absorbance = -0.010169+0.20992 [Glucose], 0.2-4 g/L, R²=0.9999. The standard errors of correction is 0.01458 g/L, The standard errors of validation is 0.02119 g/L. The standard error of prediction is 5.366 g/L. However, the result from infrared spectrum were R²=0.9999, the standard error of correction is 0.2561 g/L, The standard errors of validation is 0.3580 g/L. The standard error of prediction is 4.411 g/L. From these dates we find the glucose quantitive analysis is more accurate by infrared spectrum than glucose oxidase method in the broth.