Carbonyl compounds in gas and particle phases of mainstream cigarette smoke

https://doi.org/10.1016/j.scitotenv.2011.07.065Get rights and content

Abstract

Carbonyl compounds (carbonyls) are important constituents of cigarette smoke and some are toxic and may be carcinogenic or mutagenic to humans. In this study carbonyl emissions in the gas and particle phases of mainstream cigarette smoke were assessed by GC–MS with pentafluorophenyl hydrazine (PFPH) derivatization. Seven brands of cigarettes and one brand of cigar common in the UK market and having differing nicotine, tar and carbon monoxide yields were investigated. Sixteen carbonyl components were identified in gaseous emissions and twenty in the particle phase. In the gaseous emissions, acetaldehyde presented as the predominant species, followed by formaldehyde, 2-propenal, and pentanal. In the particulate emissions, 1-hydroxy-2-propanone was the most abundant followed by formaldehyde, benzaldehyde, and 2,5-dimethylbenzaldehyde. Significant differences were found in carbonyl emissions among the brands of cigarettes. The gaseous carbonyl emissions varied in the range of 216–405 μg cigarette−1 (μg cig−1) and the particulate carbonyl emissions varied in the range of 23–127 μg cig−1. Positive correlations were found between the total emission of carbonyls, tar yield and carbon monoxide yield. Similar gas/particle (G/P) partitioning ratios of carbonyls were found among all cigarettes, which implies that G/P partitions of carbonyls in smoke mainly depend on the physical properties of the carbonyls. The gaseous carbonyl emissions were enhanced by 40% to 130% when some of the water, accounting for 8–12% of cigarettes in mass, was removed from the tobacco. Non-filtered cigarettes showed significantly higher carbonyl emissions compared to their filtered equivalents. Carbonyl particulate accounted for 11–19% by mass of total particulate matter from tobacco smoke. The cigar generated 806 μg cig−1 gaseous and 141 μg cig−1 particulate carbonyls, which is 2–4 times greater than the cigarettes.

Highlights

► Carbonyl emission factors in both gas (16 species) and particle (20 species) phases from tobacco smoke were investigated. ► Sensitive GC/MS method with pentafluorophenyl hydrazine (PFPH) derivatisation was used in this study. ► Formaldehyde, acetaldehyde, 1-hydroxy-2-propanone were found in as dominant carbonyls. ► Positive correlation was found between total emission of carbonyls, tar yield and carbon monoxide yield.

Introduction

Tobacco smoke contains thousands of chemical compounds, some of which are directly hazardous and potentially carcinogenic or mutagenic to humans (IARC, 2004). Cigarette smoking is associated with cancer, birth defects and heart disease, and is one of the leading preventable causes of premature death globally (IARC, 2004). The chemical composition of tobacco smoke has long been a subject of interest to the health and analytical science communities. Hazardous volatile organic compounds (VOCs), carcinogenic polycyclic aromatic hydrocarbons (PAHs), nicotine and some carcinogenic nitrosamines specific to tobacco have been examined extensively in the last two decades (Baek and Jenkins, 2004, Charles et al., 2007, Charles et al., 2008, Ding et al., 2006, Gmeiner et al., 1997, Harrison et al., 1996, Kayali and Rubiobarroso, 1995, Lee, 1995, Polzin et al., 2007, Slezakova et al., 2009, Xie et al., 2003, Zanieri et al., 2007). Although several carbonyl species have been identified in cigarette smoke (without quantification) they have not been studied fully until now. Ten carbonyl compounds have been classified as probably or possibly carcinogenic to humans by the International Agency for Research on Cancer (IARC, 2004) including formaldehyde, acetaldehyde, 2-propenal, 3-buten-2-one, some of which have been found in tobacco smoke. Carbonyls from tobacco smoke have also been identified as an important source of carbonyls to indoor air, where they may irritate the respiratory system of non-smokers and induce asthma in children and have even been linked with leukemia (Feng et al., 2004, Marchand et al., 2006, Pang and Mu, 2007). Carbonyls in tobacco smoke can react with some low-density lipoproteins on the skin, promoting the modification of these proteins and contributing to the development of smoking-related diseases (Freeman et al., 2005). Accurate quantitative assessments of carbonyls in cigarette smoke are therefore essential to help estimate human and environmental exposure. Systematic measurements of carbonyls in cigarette smoke are scarce even though this information is of the utmost importance to the assessment of the risks of carbonyls on human health.

The aldehyde content in cigarette smoke was first determined gravimetrically by the reaction of the aldehydes with dimedon to form their derivatives as precipitates (Touey, 1955). Later techniques resolved carbonyl compounds in tobacco smoke using liquid chromatography based on their 2,4-dinitrophenylhydrazones (Houlgate et al., 1989, Shin et al., 2009, Wong et al., 1997). Some information on carbonyls in cigarette smoke had been obtained using gas chromatography in combination with analytical standards (Charles et al., 2007, Lin et al., 2008, Mitschke et al., 2005, Moir et al., 2007). Many analytical methods are limited either in application due to expensive and complicated instrumentation, or sensitivity. Additionally, the above literature generally report on only a small sub-set of carbonyl species, and there is very limited systematic data on a carbonyl as a fully specified chemical class. A further limitation in those studies is that only carbonyls in the gas phase are reported. It is necessary also to study the carbonyls in particle phase since particulate matter accounts for a major component in cigarette smoke.

In this study carbonyl emissions in both the gas and particle phases of cigarette and cigar smoke were determined by a sensitive and versatile analytical GC–MS method based on pentafluorophenyl hydrazine (PFPH) derivatization. The influences of cigarette filter ventilation and moisture content of tobacco were investigated. The correlations between carbonyl yields and nicotine, tar and carbon monoxide yields were examined.

Section snippets

Material

Hexane was purchased from Fisher, UK (HPLC grade). All carbonyl compounds (25 species) were purchased from Sigma-Aldrich company (Gillingham, UK), including formaldehyde (37% in water), acetaldehyde (99.5%), propanal (97%), acetone (99.5%), 2-propenal (95%), 2-butenal (99.5%), butanal (98%), 2-butanone (99%), 2-methyl-2-propenal (95%), 1-penten-3-one (98%), pentanal (99%), 3-methylbutanal (97%), hexanal (98%), 2-furfural (98%), benzaldehyde (99%), 1-hydroxy-2-propanone (90%),

Carbonyls in gas phase

In the gas phase of cigarette smoke 16 species of carbonyls were identified as shown in Fig. 3 and detailed carbonyl emissions for each cigarette type are listed in Table 3. For all cigarettes acetaldehyde was found to be the predominant carbonyl with an emission varying from 76.2 to 187 μg cigarette−1 (μg cig−1), which is consistent with findings reported in the previous studies (Shaughnessy et al., 2001, Shin et al., 2009). For normal untreated cigarettes, formaldehyde, acetone, propanal,

Conclusion

In this study the detailed carbonyl emission factors in both gas and particle phases from cigarette smoke were determined by a sensitive analytical GC–MS method based on PFPH derivatization. 16 carbonyls were identified in the gas phase and 20 species in the particle phase, most of which are known to have negative effects on human health. For instance acetaldehyde and formaldehyde were found to be dominant carbonyls in the gas phase of cigarette (and cigar) smoke. Acetaldehyde emissions varied

Acknowledgments

Pang acknowledges the supports from a BP-Amoco fellowship of the Royal Society and the Chinese National Natural Science Foundation (20807041). Lewis acknowledges support from the Natural Environment Research Council (NERC) and the Engineering and Physical Sciences Research Council (EPSRC) for support of the instrumentation used in this paper. The authors thank Dr. Samuel Edwards for improving the English in this paper.

References (37)

Cited by (83)

View all citing articles on Scopus
View full text