Some studies (Vuarnoz, 2015; Williams, 2015) have pointed to chemical binding between TSNA and the tobacco matrix that are not extracted using standard extraction procedures. This potential negative bias may cause inconsistencies in ‘mass balance’ determinations of HTP products. This study used a systematic approach to optimize an accelerated solvent extraction technique to differentiate the ‘free’ and ‘bound’ TSNA available in the tobacco product.

Accelerated solvent extraction was performed using a Dionex ASE 350 using the same extraction solutions and instrumental parameters as the standard procedure for TSNA determinations. However, the extraction process was sequentially optimized modifying the temperature, sample size, static time and the number of extraction cycles. To investigate the potential of artifact formation, l-ascorbic acid was added to the extraction solution.

The optimized conditions to extract both the ‘bound’ and ‘free’ TSNA were determined to be: 200 °C, for a 0.5 g sample size and a 15-minute static or cycle time, with a single cycle being adequate. The ‘free’ TSNA were extracted at 45 °C with all other parameters remaining the same, producing results consistent with those determined using the standard extraction procedures.

Detailed results of the optimization process will be presented. The impact of ASE on NNN, NAT and NAB was found to be negligible. However, the estimated ‘free’ NNK fraction was calculated to be anywhere from 16 % to 41.7 % of the total NNK content determined by ASE.