Determination of caffeine in black tea leaves by Fourier transform infrared spectrometry using multiple linear regression

Abstract A simple and fast analytical procedure was developed for the determination of caffeine in black tea leaves. The method is based on multiple linear regression treatment of Fourier transform infrared spectrometric data obtained in the wave number range 1800–1300 cm−1 after extraction of caffeine in CHCl3 from tea samples, wetted with an aqueous solution of NH3 (0.1 M). The procedure requires no complex sample preparation. It provided a limit of detection of 0.035 mg/ml, a sampling frequency of 6 h−1 and a coefficient of variation of 0.8% for five independent measurements of a tea sample with 3.68% w/w caffeine content. This procedure provides a drastic reduction in the consuming organic solvent for each sample compare to that of the reference chromatographic determination. The accuracy of technique is evaluated by comparing the obtained results with those of a reference HPLC technique. An average value of 3.60±0.07% w/w was obtained by HPLC for a powdered tea sample which is compromising when is compared to 3.68±0.03% (w/w) obtained by this FTIR technique.