FT-NIR characterization with chemometric analyses to differentiate goldenseal from common adulterants

Area of Science:

• Herbal Medicine
• Analytical Chemistry
• Spectroscopy

Background:

• Goldenseal (Hydrastis canadensis L.) is a popular herb facing supply challenges and adulteration.
• Market value exceeded $32 million in the US in 2014.
• Wild goldenseal is listed in the Convention on International Trade in Endangered Species, increasing adulteration risks.

Purpose of the Study:

• To evaluate Fourier transform near-infrared spectroscopy (FT-NIR) coupled with chemometric models for detecting goldenseal adulteration.
• To compare the effectiveness of partial least squares (PLS) regression, soft independent modeling of class analogy (SIMCA), and moving window principal component analysis (MW-PCA).
• To identify the best chemometric approach for detecting specific adulterants in goldenseal.

Main Methods:

• Utilized FT-NIR spectroscopy for nondestructive analysis.
• Applied three chemometric models: PLS, SIMCA, and MW-PCA.
• Tested models computationally on goldenseal mixed with adulterants (yellow dock, yellow root, coptis, Oregon grape) at varying concentrations (2% to 95%).

Main Results:

• All models differentiated pure goldenseal from adulterants.
• PLS model was ineffective for detecting adulteration.
• SIMCA model detected yellow root and Oregon grape at 10% levels.
• MW-PCA model detected yellow dock at ≥15% and coptis at ≥5%.

Conclusions:

• NIR spectroscopy combined with chemometrics offers a rapid detection method for goldenseal adulteration.
• The choice of chemometric model is crucial and depends on the specific adulterant and desired sensitivity/specificity.
• This approach is valuable for industry and investigators to ensure marketplace authenticity.