Avaliação de equipamento espectrômetro no infravermelho próximo no processo de classificação de tabaco.

Abstract

The use of tools that allow strict control over product quality and have low operating costs can be the difference for an industry to remain the leader in its sector. Aiming at improving measurements of moisture and nicotine in processes involving the sector of receiving tobacco, the company Souza Cruz S.A. acquired the process analyzer TM710e, a near infrared spectrophotometer (NIR) with noninvasive measurement, as these variables are closely related to the classification of tobacco and directly influence the amount to be paid to the supplier. The methodology used to plan and define the validation parameters was the Guideline on Validation of Analytical Methods defined by INMETRO. Through this guidance, the results of TM710e can be compared with the results of classical measurements used in quality control in the organization, as well as verify their deviations. Through the tests to identify the selectivity and robustness of TM710e, variables were identified such as the inappropriate response time and place of installation of the equipment, the absorption spectrum of the rope and the different colors that tobacco that contributed to the appearance of uncertainty and deviations in the measurements. The place of installation of the equipment was modified during this research as well as a reduction in the response time, facts that, in compliance with the periodic calibrations, contributed to significant reductions in the coefficient of variation and increased the accuracy of the equipment, therefore enabling optimization. From the experiments and statistical analysis to determine the accuracy of the analytical methods involved and compare them to the performance, it was found that the characteristics of the equipment are compatible with the requirements of the method under study. In order to reduce the deviations and improve the robustness of the equipment, the variation due to different colors of tobacco could be optimized by calibration for each type of tobacco and by automating the exchange of parameterization by a sophisticated color sensor, or even by the treatment of images. Thus, the self- adjustment of the Trim could also be performed. It would be possible to prevent the absorption of the spectrum of the rope; however, as the reading of this material is possible and the diversion generated by such is low, it is understood that the optimization of other variables would be enough to improve the performance of equipment. Given the above, the purpose of evaluating the near infrared spectrophotometer equipment TM710e through its monitoring, analysis of purchasing processes and classification of tobacco, quantifying the levels of moisture and nicotine and exploratory data analysis in order to investigate the variables that affect the performance equipment was completed. The results obtained with precise testes could serve as the basis for the company to define the acceptance criteria and to implement this process control through near infrared spectroscopy in all tobacco classification units. Keywords: near infrared spectroscopy (NIR), process control, tobacco grading, moisture, nicotine.