Smoke taint in wines from bushfire smoke exposure has become a concern for wine producers. Smoke taint compounds are primarily derived from pyrolysis of the lignin component of fuels. This work examined the influence of the lignin composition of pyrolysed vegetation on the types of putative smoke taint compounds that accrue in wines. At veraison, Merlot vines were exposed to smoke generated from five vegetation types with differing lignin composition. Smoke was generated under pyrolysis conditions that simulated bushfire temperature profiles. Lignin and smoke composition of each fuel type along with putative smoke taint compounds in wines were determined. The results showed that, regardless of fuel type, the commonly reported guaiacyl lignin derived smoke taint compounds, guaiacol and 4-methylguaiacol, represented about 20% of the total phenols in wines. Quantitatively, syringyl lignin derived compounds dominated the total phenol pools in both free and bound forms. The contributions of p-hydroxyphenyls were generally similar to the guaiacyl sources. A further unexpected outcome of the study was that pine smoke affected wines had significantly elevated levels of syringols compared to the controls although pine fuel and its smoke emission lacked syringyl products.

This paper describes the establishment of a robust method to determine compound specific δD and δ(13)C values of volatile organic compounds (VOCs) in a standard mixture ranging between C(6) and C(10) and was applied to various complex emission samples, e.g. from biomass combustion and car exhaust. A thermal desorption (TD) unit was linked to a gas chromatography isotope ratio mass spectrometer (GC-irMS) to enable compound specific isotope analysis (CSIA) of gaseous samples. TenaxTA was used as an adsorbent material in stainless steel TD tubes. We determined instrument settings to achieve a minimal water background level for reliable δD analysis and investigated the impact of storage time on δD and δ(13)C values of collected VOCs (176 days and 40 days of storage, respectively). Most of the standard compounds investigated showed standard deviations (SD)<6‰ (δD) when stored for 148 days at 4 °C. However, benzene revealed occasionally D depleted values (21‰ SD) for unknown reasons. δ(13)C analysis demonstrated that storage of 40 days had no effect on VOCs investigated. We also showed that breakthrough (benzene and toluene, 37% and 7%, respectively) had only a negligible effect (0.7‰ and 0.4‰, respectively) on δ(13)C values of VOCs on the sample tube. We established that the sample portion collected at the split flow effluent of the TD unit can be used as a replicate sample for isotope analysis saving valuable sampling time and resources. We also applied TD-GC-irMS to different emission samples (biomass combustion, petrol and diesel car engines exhaust) and for the first time δD values of atmospheric VOCs in the above range are reported. Significant differences in δD of up to 130‰ were observed between VOCs in emissions from petrol car engine exhaust and biomass combustion (Karri tree). However, diesel car emissions showed a high content of highly complex unresolved mixtures thus a baseline separation of VOCs was not achieved for stable hydrogen isotope analysis. The ability to analyse δD by TD-GC-irMS complements the characterisation of atmospheric VOCs and is maybe used for establishing further source(s).

No abstract