Tyxyapplication Research On Near Infrared Spectroscopy In Production Technology Of 2,3,5-Trimethyl Hydroquinone

2,3,5-trimethyl hydroquinone (TMHQ) is an important intermediate of vitamin E. With the wide application of vitamin E in recent years, the demand of TMHQ increases year by year. There are many methods to synthesis TMHQ, one of the most advanced routes is 2,3,6-trimethyl phenol oxidation-reduction method. This method has less reaction steps and mild condition, is a hotspot of current process research. The whole production process of TMHQ including oxidation, extraction, neutralization, washing, enrichment, reduction, centrifugal filtration, vacuum drying and so on. Many process parameters during production can affect the yield and quality of final product. However, some key parameters in the process of production controlled only by experienced workers manually. The traditional detection methods need tedious pre-treatment, such as high performance liquid chromatography (HPLC), the testing result lags behind the production process seriously, so it’s hard to find problems in time. Many drawbacks have been exposed in the traditional mode. The labile process parameters will lead to an unstable product quality and differences between batches. Therefore, a fast and efficient method to monitor process parameters has great application value.Near infrared spectroscopy (NIRS) analysis technology is a kind of process analysis tool. Because of advantages of rapid and non-destructive, it has been widely used in pharmaceutical and chemical industry.This article focus on extraction, hydrogenation reduction reaction and vacuum drying processes. Analysis models for these three technological links were established by using NIRS and chemometrics.1 Optimized the purification method of crude TMBQ, and made improvement for water vapor distillation apparatus. Investigated reduction effect of four kinds of reductants to TMBQ, including sodium dithionite, palladium/ hydrogen, hydrazine hydrate, sodium borohydride, eventually chose palladium/ hydrogen as the best, and optimized the process parameters.2 This study selected the extraction process and established PLS quantitative model for TMBQ with NIRS and optimized the model with SPA, iPLS and correlation coefficient method. Eventually, the best model established with 4385.33cm-1 -5152.86 cm-1,5928.11 cm-1 -6309.94 cm-1 selected by iPLS produced the values of Rp=0.996 and RMSEP=0.1350. The method established is expected to reach rapid detection of TMBQ in extraction process.3 First, established a detection method of TMBQ and TMHQ by high performance liquid chromatography(HPLC) and obtained primary data. Second, collected spectrum of reaction liquid by NIR. Then associated spectral data with primary data by PLS algorithm and established quantitative analysis model of reactant and product respectively. Third, optimized the model by adopting a variety of algorithm. We built two models that realized fast process monitoring of hydrogenation reduction reaction and these models could judge abnormal conditions during reaction.4 We use a method to configure a certain gradient of water content in TMHQ solid powder. The diffuse reflectance spectra of samples were associated with water content in TMHQ using PLS algorithm. Investigated various pretreatment methods and wavelength selection methods to optimized the model.