Bull. Spec. CORESTA Congress, Lisbon, 2000, p. 177, S1

Determination of nicotine in tobacco: collaborative study

DAVIS R.E.; BENNETT C.B.; THOMSEN H.V.; FRANKE J.E.; JOHNSTON K.S.
Philip Morris USA, Research Center, Richmond, VA, USA
An international collaborative study was performed to compare several analytical methods for the determination of nicotine in tobacco that are in current use around the world. Five nicotine methods were evaluated and compared, specifically methyl-t-butyl ether (MTBE) extraction with capillary-column gas chromatography (GC), n-hexane extraction with capillary-column GC, n-hexane extraction with packed-column GC, methanol/ammonia extraction with capillary-column GC, and aqueous extraction with continuous flow analyzer (CFA) colorimetry. A total of 37 laboratories participated in the study, with between 9 and 18 laboratories submitting data per nicotine method. Repeatability, reproducibility, and mean nicotine statistics were calculated and compared for each method. Results for reproducibility (%) and mean nicotine difference (%, relative to the mean of the three capillary-GC methods), respectively, for each method are as follows: MTBE method (2.5%, -1.40%), hexane-capillary (4.5%, +0.06%), methanol/ammonia (3.7%, +1.34%), CFA (4.4%, +4.08%), and hexane-packed (5.8%, +4.14%). Pair-wise group comparison tests with simultaneous 95% confidence intervals were used to compare the sample nicotine values between any two given methods. Eight of the ten pair-wise comparisons were statistically different at 95% confidence, the two statistically indistinguishable pair-wise comparisons being CFA vs. hexane-packed and hexane-capillary vs. methanol/ammonia. The results of this collaborative study will be useful toward the goal of standardizing on a reference method for nicotine analysis in tobacco and tobacco products.