Abstract: To investigate the effects of different de-enzyming treatments on the quality of Polygonum multiflorum Thunb. and screen the optimal primary processing method for this medicinal herb, fresh Polygonum multiflorum Thunb. samples were subjected to different de-enzyming processes and a non-de-enzyming control treatment. Phenotypic characteristics of processed samples were collected via direct observation and colorimeter measurement. High performance liquid chromatography (HPLC) was adopted to establish the fingerprint of Polygonum multiflorum Thunb. and determine the contents of four components: 2,3,5,4′-tetrahydroxystilbene-2-O-â-D-glucoside, emodin-8-O-β-D-glucoside, physcion-8-O-β-D-glucoside and emodin. Multivariate statistical methods including principal component analysis (PCA), cluster analysis (CA) and entropy weight-technique for order preference by similarity to an ideal solution (EW-TOPSIS) were used for sample classification and comprehensive quality evaluation.The results showed that except for the high-temperature de-enzyming group (HG105), all other de-enzyming treatments deepened the surface color, increased the hardness of Polygonum multiflorum Thunb. slices, and significantly elevated the chroma value of sample powders. The change intensity followed the order of steam de-enzyming (SGsq, YGsq, HGsq) > boiling de-enzyming (SZsq) > microwave de-enzyming (WBsq). Among the 10 batches of Polygonum multiflorum Thunb. samples under different processing treatments, the content of 2,3,5,4′-tetrahydroxystilbene-2-O-β-D-glucoside ranged from 3.793% to 6.907%, emodin-8-O-β-D-glucoside from 0.198% to 0.343%, physcion-8-O-β-D-glucoside from 0.055% to 0.123%, and emodin from 0.007% to 0.192%. The sample treated with 50 ℃ drying de-enzyming (HG50sq) showed the highest contents of 2,3,5,4′-tetrahydroxystilbene-2-O-β-D-glucoside and emodin-8-O-β-D-glucoside. A total of 16 common peaks were matched in the 10 batches of samples, with all similarity values greater than 0.900. Both PCA and CA divided the 10 batches of samples into two identical categories; EW-TOPSIS analysis indicated that WBsq and HG50sq samples possessed the best quality, while high-temperature drying de-enzyming (HG105) showed the poorest quality. The HG50sq treatment was conducive to the overall quality improvement of Polygonum multiflorum Thunb.

Key words: Polygonum multiflorum Thunb., de-enzyming treatment, fingerprint, EW-TOPSIS, quality evaluation