You are on page 1of 7

The Science of the Total Environment 277 Ž2001.

69 ᎐ 75

Phenolic antioxidants in wood smoke
Jennica Kjallstrand, Goran ¨ ¨ PeterssonU
Department of Chemical En¨ ironmental Science, Chalmers Uni¨ ersity of Technology, Goteborg 412 96, Sweden ¨ Received 15 August 2000; accepted 24 October 2000

Abstract Ten prominent dimethoxyphenols were determined in birchwood smoke from choked and open laboratory burning and in chimney smoke from a tiled stove. The structures of the methoxyphenols are similar to those of the well-known tocopherol and ubiquinol antioxidants. The 2,6-dimethoxyphenols characterising hardwood smoke are stronger antioxidants than the corresponding 2-methoxyphenols present mainly in softwood smoke. The antioxidant activity is highest for the 2,6-dimethoxyphenols with 4-alkenyl and 4-alkyl groups, which constitute 60 ᎐ 70% of the total amount of dimethoxyphenols. Phenolic antioxidants are scavengers of oxygen radicals and should be considered when health hazards of small-scale incomplete biomass burning are estimated. ᮊ 2001 Elsevier Science B.V. All rights reserved.
Keywords: Combustion; Air pollutants; Gas chromatography; Methoxyphenols; Phenoxy radicals; Lipid peroxidation

1. Introduction The impact on health of wood smoke is currently an issue of great concern with respect to small-scale burning of wood as a renewable fuel. Particle-related health hazards, attributed mainly to polycyclic organic compounds, were assessed in a large American study in Boise, Idaho ŽCupitt et

Corresponding author. Tel.: q46-31-7722998; fax: q4631-7722999. E-mail address: ŽG. Petersson ..


al., 1994.. The carcinogens benzene and 1,3butadiene are prominent among the emitted volatile hydrocarbons ŽBarrefors and Petersson, 1995.. An important, but often overlooked fact is that most of the organic compounds emitted from incomplete burning are related in structure to the lignin and carbohydrates of the biomass ŽAlen ´ et al., 1996.. These compounds are therefore very different from the better-known hazardous compounds in combustion emissions from stationary and mobile sources. Analytical data have previously been given for lignin-specific methoxyphenols in smoke from forest biomass burning

0048-9697r01r$ - see front matter ᮊ 2001 Elsevier Science B.V. All rights reserved. PII: S 0 0 4 8 - 9 6 9 7 Ž 0 0 . 0 0 8 6 3 - 9

The chromatogram in Fig. Petersson r The Science of the Total En¨ ironment 277 (2001) 69 ᎐ 75 ¨ ŽKjallstrand et al. 1990. 2000. The very high concentrations from choked laboratory burning are partly explained by the enclosure of smoke on choking. 1997. Early on. All these species contain the 4-hydroxy-3. The results in Table 1 compare concentrations and proportions of the ten 2.. In Table 1.. 0. The column was linked to a Varian Saturn 2000 ion trap mass spectrometer. Concentrations were calculated assuming the same response for all the structurally similar dimethoxyphenols ŽKjallstrand et al. Antioxidant properties have also recently been demonstrated for several lignin-related methoxyphenols ŽBarclay et al. Electron impact mass spectra Ž70 eV.70 J. All samples were taken on injector liner glass cartridges Ž0. alkenyl and carbonyl-containing groups of the phenols are remainders of the original C 3 chain in lignin..6-dimethoxyphenols. 1997. because of their crucial importance in living organisms.. as the stationary phase.d. The separations were made on a fused silica open tubular column Ž30 m = 0. Chips and match-size sticks were cut for laboratory burning and coarser lumps for tiled stove burning.. 2000. 1989. and as semi-volatile components of wood smoke particles ŽHawthorne et al. 1 illustrates the separation of the 10 most prominent 2. The same method for choked burning was used in a previous study of smoke from different forest biomass materials ŽKjallstrand et al. The discussion is based on the formation of methoxyphenols from birchwood burning.. Newmark..25 mm i. 2000..d. comparative data are given for the three different types of burning experiments.. 1990. Ogata et al. The antioxidant activities of the phenolic tocopherols and ubiquinols have been extensively studied ŽBarclay et al. Smoke samples were ¨ withdrawn through the bottom hole of the clay pot using a gas-tight syringe. as adsorbent. Season-dried wood or wood with a similar dryness was used. 1997. Niki.5-dimethoxyphenyl Žsyringyl. were recorded.8 mm i. with a cyanopropylphenylsilicone ŽRtx 1701. Phenols possess basic antioxidant properties ŽSimic.. G. 1996.. Chimney smoke from burning in a tiled stove was sampled in the .. with the porous polymer Tenax TA Ž60 ᎐ 80 mesh.. Sy. Tsuda et al.25 ␮ m layer. Helium was used for desorption and as carrier gas..5 l. Laboratory burning was performed in a ceramic pot Ž1.6-dimethoxyphenols for three widely differing burning conditions. Results This study was focussed on the 2. 1994. 1992. 1998. Smoke from choked burning was collected by placing a clay pot upside down over the fire after flame burning about half of the wood.. mrz 35 ᎐ 400. and potential protection of health ŽLeandersson et al.. same way in the chimney pipe outlet.6-dimethoxyphenols in smoke from choked burning of birchwood. which have especially strong antioxidant effects. group. Smoke samples from open burning were taken 50 cm above a similar pot fire with a conventional air pump during the whole period of flame burning. ¨ methoxyphenols were identified as prominent thermal degradation products of wood ŽFaix et al. Several of the statistically significant differences in relative proportions of the 10 phenols between the three 2... which characterises the phenylpropane building blocks of hardwood lignin. The oven temperature was increased 5ЊC miny1 from 40ЊC to 240ЊC. Kjallstrand. They are major components of hardwood smoke and are formed by degradation of hardwood lignin. Our communication relates the antioxidant properties of specific assessed methoxyphenols to the impact of wood smoke on human health.. which are normally not separated in Scandinavian wood fuels. Methods The birchwood burnt originated from the two closely related species Betula ¨ errucosa and Betula pubescens. ¨ 3. 1997. 1997. The concentrations in chimney smoke from stove burning and in smoke from open laboratory burning are similar.. placed in a fume cupboard. Interest in flavonoids and a variety of other plant phenols is rapidly growing because of their antioxidant effects ŽOgata et al.. Thermal desorption at 220ЊC was accomplished in the injector of the gas chromatograph ŽVarian 3800. The alkyl.

but are almost negligible at 1000ЊC.8 " 0. Methoxyphenols are the major organic products of lignin pyrolysis up to approximately 800ЊC. arranged in order of estimated decreasing antioxidant strength Burning Firewood Weight Žg.7 12.8 10.6 " 0.5 " 2. A low proportion of glowing ember for the open laboratory burning of small sticks may be related to the observed lower proportions of the carbonyl-containing phenols.3 13.. Kjallstrand.6 8... but also dramatically decreases the proportion of methoxyphenols ŽAlen ´ et al. 1 as small peaks. New devices for residential wood burning at higher temperatures are likely to eliminate emissions of methoxyphenols efficiently.1 " 7.8 " 1. 1998. average " S.5 " 0. Antioxidant properties As illustrated in Fig.3 7.8 21.. Žmg my3 . Proportions Ž%.4 " 0. Time Žmin..9 6.0 " 0.2 5.5 " 1.6¨ dimethoxyphenols because of a different lignin structure.4 " 0. the structural features of the methoxyphenols in wood smoke are similar .7 6. Smoke from average residential burning still contains appreciable proportions of methoxyphenols ŽHawthorne et al. 2.0 " 2.5-dimethoxyphenyl Žsyringyl.4 5. but ¨ those with high retention appear in Fig.4 4. Petersson r The Science of the Total En¨ ironment 277 (2001) 69 ᎐ 75 ¨ 71 Table 1 Antioxidants in smoke from birchwood burning. units in the lignin. G.0 " 1.1 2.1 " 1.9 12.4 19.0 In tiled stove Sticks and lumps 500 In chimney pipe 40 12 6 4.3 " 1.J.5 6..D.9 " 0. SyCH 3 and SyCH 2 CH 3 species. Discussion 4.9 Open in laboratory Match-size sticks 10 Above pot fire Ž50 cm.0 " 1.5 6 146 " 72 22. 1989.9 11.7 14. Smoke from other hardwood species than birch may contain larger proportions of guaiacyl compounds ŽHawthorne et al. Hotter flames for open laboratory and tiled stove burning may explain the higher proportions of the chemically least reactive SyH. compounds. 4. Regardless of the detailed proportions.1..9 " 0.6 12.7 5. Number of samples Total conc.1 " 1.6 15.0 " 1. Softwood smoke contains 2-methoxyphenols in high con- centrations ŽKjallstrand et al.4 12. The three different burning conditions all correspond to incomplete combustion with large emissions of methoxyphenols.6 " 0. 1996.2 " 1.0 " 3.1 " 2. but not 2.6 14. 2000. embracing many of those occurring in practice. SyCHsCHCH3 (E) SyCHsCH2 SyCH2 CHsCH2 SyCH2 CH3 SyCH3 SyCH2 COCH3 SyCHsCHCHO (E) SyH SyCOCH3 SyCHO Choked in laboratory Match-size sticks 2 From covered pot 2 0. The guaiacyl compounds elute earlier than the corresponding syringyl compounds on gas chromatography ŽKjallstrand et al..3 5.3 " 2.3 " 3.7 " 1.8 " 0.9 " 1.9 3.7 " 0.3 15.1 " 0.0 " 1. Sy.. 40 2 6 9. 1989. the results demonstrate that the 10 tabulated di-methoxyphenols are prominent constituents of birchwood smoke from incomplete combustion under very different conditions..2 5.4 " 2.7 5. Increased combustion temperature not only efficiently decreases the emissions of organic compounds.5 burning methods are likely to be due to the differing burning parameters in a complex way.3 5. Total concentrations and percentage proportions of the 10 most prominent 4-hydroxy-3.6 9. Birchwood smoke also contains smaller amounts of 2-methoxyphenols from corresponding 4-hydroxy-3-methoxyphenyl Žguaiacyl. Sampling Volume Žml.

Methyl Žalkyl.. Tocopherols and other phenolic antioxidants react as scavengers of peroxide radicals. The methoxyphenols in wood smoke lack the long hydrocarbon side chains of ␣-tocopherol and human ubiquinol. Structural relationships between wood smoke methoxyphenols. 1997. 1990. 1992. resulting in the formation of oxygen radicals and a decreased antioxidant effect ŽNiki. Kjallstrand. the antioxidants complement the enzymatic defence against oxygen radicals...... The antioxidant properties are due to oxidation of the phenols to comparatively non-reactive phenoxy radicals ŽSimic.. 1997. The stabilising effect of a methoxyl group is stronger than that of a methyl group according to reported redox potentials for phenols with these substituents ŽLind et al. to those of well-known physiological phenolic antioxidants such as ubiquinols and tocopherols. The hardwood phenols with two radical-stabilising methoxyl groups are expected to be more efficient antioxidants than the softwood phenols with only one methoxyl group. 1997. In the living cell. thereby preventing extensive peroxidation of lipids and other essential biomolecules ŽBarclay et al. 2. Using two different in vitro antioxidant activity assays. Petersson r The Science of the Total En¨ ironment 277 (2001) 69 ᎐ 75 ¨ Fig. group in the para position ŽBarclay et al. Niki. G. 1992. 1990. .. and methoxyl Žalkoxyl. illustrating the separation and the proportions of the ten most prominent dimethoxyphenol antioxidants. 1997.72 J. 1. Ubiquinols differ by two hydroxyl groups in para position and are therefore easily autoxidised. groups in ortho and para position stabilise the phenoxy radicals further and strengthen the antioxidant properties. The single phenolic hydroxyl group of the methoxyphenols makes them similar to ␣tocopherol as antioxidants. such a difference has been clearly demonstrated between the two methoxyphenols with a 2-propenyl Žallyl. By analogy Fig. Ogata et al. Total ion chromatogram of a smoke sample from choked birchwood burning.. ubiquinol and ␣-tocopherol. Simic.

It should be observed that the ␣-carbonyl group also lowers the p K a of the phenols to approximately 7. 1992. The effects are there- .. but may also release counteracting superoxide radicals similarly as ubiquinols ŽBrusick. Petersson r The Science of the Total En¨ ironment 277 (2001) 69 ᎐ 75 ¨ 73 with ubiquinols of differing chain length ŽNiki. which is very different from the extensively studied dietary intake. Phenolate ions are normally stronger antioxidants than the corresponding phenols ŽSimic.2.J.. Methoxyphenols studied in this respect have strong chainbreaking antioxidant activity both in organic solutions ŽBarclay et al.. Barclay et al. The effect is linked to decreased hydrogen bonding for phenols with bulky ortho groups and is therefore observed in aqueous systems. 1997. 1990.. Kjallstrand. 1997.6-dimethoxyphenols in Table 1 can be predicted.. The 2. 1997. 1997.. Newmark. and in aqueous micelle test systems ŽOgata et al. 1997 ... these phenols exist as phenolate ions in blood. SyCH 2 COCH 3 . development of atherosclerosis and initiation of cancer ŽNiki 1997... are expected to have similar activities. Ogata et al. has been demonstrated in different test systems ŽBarclay et al. Niki. 1997. Electron spin trapping studies have demonstrated methoxyphenols to be scavengers of superoxide radicals Ž Ogata et al.. The effect of the conjugated double bond is approximately as large as that of the additional methoxyl group in SyCH 2 CHsCH 2 ŽBarclay et al.6-dimethoxyphenols are likely to counteract propagation of cellular lipid peroxidation more efficiently than antioxidants with similar redox potentials. Another important aspect of antioxidants is their ability to scavenge superoxide radicals.. The additional methoxyl group should also be kinetically favourable compared with corresponding 2methoxyphenols. The electron-withdrawing ␣-carbonyl groups of SyCHO and SyCOCH 3 destabilise the phenoxy radical. A clear difference for GuCHsCHCH 3 . Specific differences in antioxidant activity between the ten 2. this is likely to make the methoxyphenols more easily mobile within and between cell organelles. .. SyCH 2 CHsCH 2 ... was the most efficient scavenger. 1998..4 for the aldehydes SyCHO and GuCHO ŽRagnar et al. The phenols with a CH 2 group adjacent to the aromatic ring ŽSyCH 3 . The only 2. The antioxidative properties of a variety of phenolic compounds in food may counteract basic health problems such as cell ageing. 1999.6-dimethoxyphenol included. which is generally recognised as kinetically favourable for chain-breaking reactions in cellular lipid peroxidation ŽBarclay et al.. The conjugated double bonds of SyCHsCH 2 and SyCHsCHCH 3 stabilise the phenoxy radical. 1996.. We conclude that 2.. 1994. Many of these prevent genotoxic effects by scavenging hydrogen peroxide ŽNakayama. 1997. 1997. Recent interest has focussed on flavonoids and other dietary antioxidants with two or more phenolic hydroxyl groups. Differences in solubility between specific methoxyphenols may cause differences in phase distribution and antioxidant activity. 1990. 4.... The ionic superoxide radical occurs in the aqueous phase. Lind et al. requiring sufficient water solubility of antioxidants in order to be efficient scavengers.. which initiate radical chain reactions. G. 1997. 1997. Evidently this permits efficient trapping of both superoxide radicals and lipid peroxide radicals. 1993.. 1990. 1997. SyCH 2 CH 3 .. Possibly. which are higher than that of SyH ŽBarclay et al. Priyadarsini et al. Health aspects Exposure to wood smoke phenols represents antioxidant intake by breathing.6-dimethoxyphenols have two bulky groups adjacent to the phenolic hydroxyl group. but without bulky ortho groups. resulting in lowered antioxidant activity ŽLind et al. 1997. The solubility of the methoxyphenols is intermediate between those of the lipid-soluble ␣-tocopherol and the water-soluble ascorbic acid.. resulting in especially high antioxidant activity for these prominent smoke components. Superoxide trapping ability essentially parallelled lipid peroxidation inhibition for a range of different phenols. The carbonyl group of SyCHsCHCHO offsets the effect of the conjugated double bond ŽBarclay et al.. Ogata et al. relative to GuCH 2 CHsCH 2 with a non-conjugated double bond. SyCH 2 CHsCH 2 .

Kjallstrand J. J Wood Chem Techn 1997. The assessed dimethoxyphenols appear condensed on particles in ambient wood smoke ŽHawthorne et al.. because the respiratory tissues are directly exposed to the wood smoke components on inhalation. Gas chromatographic and mass spectrometric analysis of 36 lignin-related methoxyphenols from uncontrolled combustion of wood.30:1551 ᎐ 1556.:73 ᎐ 90. which trap radicals without release of reactive species. Methoxyphenols from burning of Scandinavian forest plant materials. has been demonstrated to have a chemopreventive effect in a carcinogenicity test ŽTsuda et al. The one-electron reduction potential of 4-substituted phenoxyl radicals in water. Barclay LRC. Shen X. Norris JQ. Chemosphere 2000. Petersson G. 1996. 1994. Locke SJ.9:223 ᎐ 230.. Hawthorne SB. 1996.pyrene ŽWattenberg et al. Oesch P. Suppression of hydroperoxide-induced cytotoxicity by polyphenols.. References Alen ´ R. ˚ Leandersson P. Green tea po¨ AO.824:205 ᎐ 210. J Am Chem Soc 1990. Free Radic Biol Med 1997. Collection and quantitation of methoxylated phenol tracers for atmospheric pollution from residential wood stoves. Cancer Res 1994. Kuoppala E. Vanillin is probably the weakest methoxyphenol antioxidant in wood smoke and stronger effects may be anticipated for all the dimethoxyphenols reported here. Vanillin... Eriksen TE.. Chemical interactions in the lungs between methoxyphenols and polycyclic aromatic compounds may explain part of the lower carcinogenicity of wood smoke particles as compared with particles from vehicle emissions.36:137 ᎐ 148.68:2258 ᎐ 2269.. Environ Health Persp 1994. Vinqvist MR.:75 ᎐ 84. J Anal Appl Pyrolysis 1996. 1994. with respect to the dimethoxyphenol content of the smoke. the emissions of methoxyphenols decrease rapidly and the proportion of polycyclic aromatic hydrocarbons increases rapidly ŽAlen ´ et al. Faix O.74 J.: 235 ᎐ 242.17Ž1 ᎐ 2. For the same amount of extracted material.41:735 ᎐ 741. Kjallstrand J. Lind J.. 1994.54:1991 ᎐ 1993. Brusick D. Krieger MS. Genotoxicity of phenolic antioxidants. Cupitt LT. 4-hydroxy-3-methoxybensaldehyde. Kjallstrand.. In addition to their antioxidative effect. Petersson G. Volatile hydrocarbons from domestic wood burning. Lewtas J. Petersson G. Formation of the main degradation compound groups from wood and its components during pyrolysis. The antioxidant effects of the methoxyphenols in the lungs are likely to explain much of this striking difference.. The antioxidant activities in free radical peroxidation of phospholipid membranes.23:470 ᎐ 475. 1980. Toxicol Ind Health 1993. Inhibition effects of phenols. Dakin KA. the different impact on health of wood smoke particles compared to particles from other sources should be considered when emissions from traditional small-scale wood burning are regulated. Glen WG. Baskin KA. including methoxyphenols. Chemosphere 1995. Faresjo Tagesson C. It is also concluded that hardwood is preferable to softwood for open burning. 1989. The comprehensive Boise campaign estimated the lifetime lung cancer risk due to extractable organic matter from particles ŽCupitt et al. have been observed for the carcinogenic bensoŽa.. The methoxyphenols are predominant thermal degradation products from lignin at temperatures below 800ЊC.112:479 ᎐ 482. Barclay LRC. Fortmann I. Can J Chem 1990. Xi F. G. Petersson r The Science of the Total En¨ ironment 277 (2001) 69 ᎐ 75 ¨ fore different from those of monophenols. Thermal degradation products of wood.23Ž2. Evidently. Ramnas ¨ ¨ O. phenols are also potential trapping agents for electrophilic genotoxic compounds ŽNewmark. Antioxidant properties of phenolic lignin model compounds. Environ Sci Technol 1989. Gas chromatographic separation and mass spectrometric characterisation of monomeric lignin derived products. Nakayama T. lyphenols inhibit oxidant-induced DNA strand breakage in cultured lung cells. . This is of particular interest with respect to lung cancer. the risk was found to be approximately 10 times larger for particles from mobile sources than for particles from residential wood burning. The health effects of the methoxyphenol-rich smoke from open traditional wood burning are therefore expected to be very different from those of the emissions of modern high-temperature stove burning. Miller DJ. J Chromatogr A 1998. Meier D. Holz Roh-Werkstoff 1990. Merenyi ´ G..102Ž4.48:281 ᎐ 285. Barrefors G.. Ramnas ¨ ¨ O. Related polycyclic aromatic compounds are regarded as responsible for the lung cancer effect of urban particulate air pollution ŽCupitt et al. Mathiason MB. At higher temperatures. Exposure and risk from ambient particle-bound pollution in an airshed dominated by residential wood combustion and mobile sources.

53:292 ᎐ 298. Jpn J Cancer Res 1994. Holzforchung 1999. Antioxidant activity of magnolol. editorial. Cancer Res 1980. Eriksson T. Reitberger T. G. editor.18:63 ᎐ 70. Plant phenolics as potential cancer prevention agents. Tsuda H. New York: Plenum Press. Hoshi M. Endo T. New York: Pergamon Press.85:1214 ᎐ 1219. Kjallstrand. Rao MNA. 1996:25 ᎐ 34. Urano S. . Niki E. 1992:47 ᎐ 59.pyrene-induced neoplasia. Shimotohno K. ␣-tocopherol and five naturally occurring antioxidants on initiation of hepatocarcinogenesis by 2amino-3-methylimidazow4. honokiol and related phenolic compounds. Coccia JB. Ogata M. JAOCS 1997.:933 ᎐ 941. Chemopreventive effects of ␤-carotene. Molec Aspects Med 1997. Wattenberg LW.5-f xquinoline in the rat. Petersson r The Science of the Total En¨ ironment 277 (2001) 69 ᎐ 75 ¨ Newmark HL. Mechanisms and dynamics of antioxidant action of ubiquinol. Uehara N. Dietary phytochemicals in cancer prevention and treatment.40:2820 ᎐ 2823. Radical formation in 75 ozone reactions with lignin and carbohydrate model compounds. In: Davies KJA. Inhibitory effects of phenolic compounds on benzoŽa.J. Ragnar M. American Institute for Cancer Research. Simic MG. Lam LKT. Physico-chemical properties and antioxidant activities of methoxyphenols.24Ž6. Oxidative damage and repair. Iwahori Y et al. Antioxidant compounds: an overview.74:557 ᎐ 562. Free Radic Biol Med 1998. Guha SN. Priyadarsini KI.