The analytical utility of thermally desorbed polydimethylsilicone membranes for in-vivo sampling of volatile organic compounds in and on human skin

A thermally-desorbed polydimethylsilicone (PDMS) membrane approach with analysis by gas chromatography-mass spectrometry has been developed and characterised, to enable the VOC arising in, and on skin, from glandular secretions, exogenous materials, products of perfusion from blood, and microbiological metabolites to be sampled in a single procedure. In-vitro studies using a series of volatile fatty acid standards indicated that the recovery efficiency of the technique increased with decreasing volatility; for example, the recovery of hexanoic acid was 3.3 times greater than that for 2-methylpropanoic acid. The relative standard deviation of the methodology decreased with decreasing volatility; RSD = 19% for 2-methylpropanoic acid and RSD = 7% for hexanoic acid. Sampled-mass vs. response relationships were modelled satisfactorily using linear regression analysis with regression coefficients in the range 0.95 to 0.998. In-vivo reproducibility was assessed though the analysis of the responses of 1-dodecane, 3,7-dimethyloct-1-ene, 2-propenoic acid, 2-ethylhexyl ester, 2-ethylhexan-1-ol, butanoic, 2-ethylhexylester, and junipen (1,4-methanoazulene, decahydro-4,8,8-trimethyl-9-methylene-); six compounds selected at random retention times from a GC-MS chromatographic VOC profile of human skin containing several hundred resolved and partially resolved compounds. Five samples were obtained simultaneously from the forearm of a healthy male participant. The in-vivo sample masses were estimated to be in the range 50 pg to 100 ng per sample with observed RSD falling between 15% and 32%; in line with a Horwitz trend. Increasing the sample time from 5 min to 120 min generally resulted in an enrichment of the VOC recovered, and for many VOC substantial increases in sensitivity (x7) were observed over this time range as the PDMS sampling-patch approached equilibrium with the underlying skin. Nevertheless, more volatile components, 2,4,6-trimethylcarbazole for instance, were observed to be lost from the analysis with increasing sample time, in a manner analogous with breakthrough behaviour in adsorbent traps. Finally, a 10 day storage study at 4 degrees C suggested that micro-biological factors were significant in their effect on sample stability. Significant changes (up to x8) were observed in the masses of compounds recovered post storage. These studies confirmed that polydimethylsilicone membrane sampling patches of human skin provide rich and analytical useful data. It is important to note that care in experimental design is needed to avoid sampling artefacts being introduced through sampling selectivity, and/or, sample instability where samples are stored for longer than 24 h at 4 degrees C or higher.