A method to improve the performance of wide dye concentration measurement based on the multi‐pathlength spectrophotometry

Accurate measurement of dye concentration in a dyebath is a basic step for dyeing process control and dyebath reuse. The dyebath usually covers a wide concentration range. In this article, a method based on multi-pathlength spectroscopy was proposed to improve the test performance for wide dye concentration. Partial least squares regression (PLSR) models were constructed on the combined pathlength spectra and evaluation metrics coefficient of determination ( R 2 ) and root mean square error (RMSE) were adopted for performance comparison. Results of the single-component experiment at different pathlengths (10, 5, 3, 2 and 1 mm) showed that for CI Acid Violet 54 and CI Acid Blue 113 the introduction of spectra at the 5 mm pathlength yielded greater performance. Further additions of lower pathlengths spectra were not as favourable as that of the 5 mm pathlength spectra for the models based on 10 mm spectra. While the upper concentration limits for three dyestuffs were raised to 0.50 g/L and 5 and 10 mm pathlengths were selected for further analysis, results demonstrated that the introduction of spectra at the 5 mm pathlength generally promoted the test performance for all three dyestuffs. Based on the results of single-component experiment, 10 and 5 mm pathlengths were eventually decided as the detection pathlengths for three-component solutions, and the concentration ranges of CI Acid Yellow 117, CI Acid Violet 54 and CI Acid Blue 113 were 0.03–0.48, 0.03–0.48 and 0.02–0.40 g/L, respectively. Results of three-component experiments suggested that evaluation metrics R 2 for the three dyestuffs were raised from 0.409, 0.409 and 0.980 to 0.957, 0.992 and 0.998, respectively, by the multi-pathlength spectroscopy technique.