Development of colorimetric sensor array for discrimination of herbal medicine

Abstract

Today, traditional systems of medicines (such as herbal distillates) become important resources for providing healthcare benefits. The ability to discriminate among closely similar herbal products is critical to ensure their efficacy. This article proposes a pattern-based recognition approach for the rapid discrimination of herbal distillates using a low-cost and sensitive colorimetric sensor array composed of 25 indicators. The color changes of the sensor exposed to the vapor of the herbal distillates can be monitored easily with an ordinary flatbed scanner. The digital representation of the array response was analyzed with hierarchical clustering analysis (HCA) and principal component analysis (PCA). Using new variable selection strategy, 6 indicators among the 25 employed indicators were selected as discriminant elements of the array. So, a complete discrimination (with 100% accuracy) of 46 herbal distillates was achieved. The proposed sensor represented a better resolution when analytes were placed in an oven at 85 °C for 45 min. This colorimetric sensor array demonstrates excellent potential for quality assurance/control applications of herbal distillates.