Mutagenicity assessment of aerosols in emissions from wood combustion in Portugal

Environmental Pollution 166 (2012) 172e181

Contents lists available at SciVerse ScienceDirect

Environmental Pollution journal homepage: www.elsevier.com/locate/envpol

Mutagenicity assessment of aerosols in emissions from wood combustion in Portugal B. Vu a, b, C.A. Alves a, b, C. Gonçalves a, b, C. Pio a, b, F. Gonçalves b, c, R. Pereira b, d, * a

Departamento de Ambiente da Universidade de Aveiro, Campus de Santiago, 3810-193 Aveiro, Portugal CESAM e Centro de Estudos do Ambiente e do Mar, Campus de Santiago, 3810-193 Aveiro, Portugal c Departamento de Biologia da Universidade de Aveiro, Campus de Santiago, 3810-193 Aveiro, Portugal d Departamento de Biologia da Faculdade de Ciências da Universidade do Porto, Rua do Campo Alegre, 4169-007 Porto, Portugal b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 1 October 2011 Received in revised form 14 February 2012 Accepted 1 March 2012

Polycyclic aromatic hydrocarbon (PAH) extracts of fine particles (PM2.5) collected from combustion of seven wood species and briquettes were tested for mutagenic activities using Ames test with Salmonella typhimurium TA98 and TA100. The woods were Pinus pinaster (maritime pine), Eucalyptus globulus (eucalypt), Quercus suber (cork oak), Acacia longifolia (golden wattle), Quercus faginea (Portuguese oak), Olea europea (olive), and Quercus ilex rotundifolia (Holm oak). Burning experiments were done using woodstove and fireplace, hot start and cold start conditions. A mutagenic response was recorded for all species except golden wattle, maritime pine, and briquettes. The mutagenic extracts were not correlated with high emission factors of carcinogenic PAHs. These extracts were obtained both from two burning appliances and start-up conditions. However, fireplace seemed to favour the occurrence of mutagenic emissions. The negative result recorded for golden wattle was interesting, in an ecological point of view, since after confirmation, this invasive species, can be recommended for domestic use. Ó 2012 Elsevier Ltd. All rights reserved.

Keywords: Wood smoke Mutagenicity Polycyclic aromatic hydrocarbon Ames test

1. Introduction Residential biofuel combustion for cooking and heating has posed several environmental problems and health risks (Bølling et al., 2009; Zielinska and Samburova, 2011). Burning of wood and other biofuels for domestic application is one of the major sources of regional and local air pollution, principally regarding the production and release of particulate matter and organic compounds (Caseiro et al., 2009; Fine et al., 2004a,b; Gaeggeler et al., 2008; Gonçalves et al., 2010). Biomass combustion is one of the major sources of aerosol in United States (Holden et al., 2011) and in Europe during the winter (Gelencsér et al., 2007; Puxbaum et al., 2007). In Portugal, it was estimated that this source accounts, on average, for 18% of the atmospheric aerosol mass (Borrego et al., 2010). The particulate matter borne from domestic combustion is dominated by submicron particles, which, through inhalation, enter and reach the upper and lower regions of the respiratory tract and induce human health problems. The health implications by fine aerosol from domestic wood combustion may be more serious than has previously been thought (Bølling et al., * Corresponding author. E-mail addresses: [email protected], (R. Pereira).

[email protected],

[email protected]

0269-7491/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.envpol.2012.03.005

2009; Boman et al., 2003). On the other hand, particulate matter interferes with the radiative properties of the atmosphere, supporting or offsetting the greenhouse gas effects depending on their size and their organic and inorganic content (Aunan et al., 2009; Danny and Kaufmann, 2002). In addition to climate change and air quality models, source apportionment methodologies applied to ambient data require detailed biomass burning emission factors. Due to the lack of specific data, biomass burning profiles obtained for the United States (e.g. Fine et al., 2001, 2002, 2004a,b), Scandinavian countries (Johansson et al., 2004), and mid-European Alpine regions (Schmidl et al., 2008) have been applied in source apportionment studies carried out in Southern European countries. However, the emission of pollutants from domestic combustion is a function of many variables, among others, type and quality of wood, burning appliance and combustion conditions (e.g. temperature and oxygen supply) (e.g. Johansson et al., 2004). To overcome the lack of emission factors for the combustion of major biofuels in Portugal, a series of source tests were conducted to establish the chemical composition of particulate matter that has an aerodynamic diameter smaller than 2.5 mm (PM2.5) emitted from the burning in woodstove and fireplace devices (Alves et al., 2011; Gonçalves et al., 2011). The organic chemicals associated with airborne particles, especially polycyclic aromatic hydrocarbons (PAHs) from extractable

B. Vu et al. / Environmental Pollution 166 (2012) 172e181

organic matter (EOM) have been widely investigated in studies exploring the mutagenic and potentially carcinogenic activity of ambient particulate matter (Atkinson and Arey, 1994; Baek et al., 1991; IARC, 1998). Many research using source apportionment techniques have pointed out that wood smoke and mobile source emissions contribute to a substantial amount of airborne extractable organics and mutagenicity of ambient air (Atkinson and Arey, 1994; Baek et al., 1991; Bari et al., 2010; Claxton et al., 2004; Wang et al., 2010). Studies show that PAHs and their derivatives generated from domestic cooking and domestic heating are mutagenic agents, with mutagenic potential varying due to combustion appliances, fuels and extraction methods (Claxton et al., 2004; van Houdt et al., 1986; Mukherji et al., 2002; Oanh et al., 2002). The Ames assay, a plate incorporation test which is based on genetically engineered microorganisms, offers a quick method for mutagenicity assessment and has been used widely as a screening tool for detecting mutagens and possible carcinogens (Maron and Ames, 1983; Mortelmans and Zeiger, 2000). Two bacterial strains of Salmonella typhimurium e strain TA98 and strain TA100 e are used to determine frameshift mutations and base e pair substitution mutations, respectively. Metabolic activation by the S9 mix, consisting of a rat liver microsomal fraction, obtained after induction of drug metabolising enzymes with an exposure to Aroclor1254 (Mortelmans and Zeiger, 2000), is introduced to the test for the purpose of detecting the toxicity of metabolic products of the testing chemicals (Maron and Ames, 1983; Mortelmans and Zeiger, 2000). Taking into account that emission profiles of domestic burning in Portugal have not been created before, there are no studies about the toxic effects of emitted PAHs from domestic combustion in typical appliances. In order to characterise and evaluate the residential biofuel combustion in Portuguese context, the study was designed to determine the mutagenicity of PM2.5 associated with combustion of eight different biofuels under two burning conditions in two household burning appliances (fireplace and woodstove) commonly used in Southern Europe. In a tier approach, the Ames assay performed in this study has used strains TA98 and TA100 in the absence and presence of metabolic activation as recommended in Mortelmans and Zeiger (2000). The two strains were commonly used in other comparative studies because the testing results are comparable and reproducible (Claxton et al., 2004). In addition, the pre-incubation assay, distinguished from other Ames test protocols by the exposure of the tester strain for a short period of time in a small volume containing the test agent with buffer or S9 mix prior to plating on glucose minimal agar, is believed to enhance the reaction between mutagenic metabolites and the tester strain (Mortelmans and Zeiger, 2000). The indirect e acting mutagenicity of each set of PAHs has been assayed by the Ames test using S. typhimurium TA98 and TA100 with S9 mix. The direct-acting mutagenicity of each set of PAHs has been assayed by the Ames test using the same strains, but without in vitro metabolic activation. Such study can determine which fuel is least harmful, which combustion stage and which burning device poses the least mutagenic risk. 2. Material and methods 2.1. Biomass fuel selection According to the Portuguese Forest Inventory (2005), the top seven most prevalent tree species in Portugal are Pinus pinaster (maritime pine), Eucalyptus globulus (eucalypt), Quercus suber (cork oak), Acacia longifolia (golden wattle), Quercus faginea (Portuguese oak), Olea europea (olive), and Quercus ilex rotundifolia (Holm oak). These tree species were selected for this study. In addition, biomass briquettes commonly used in domestic heating nowadays, especially in urban areas, made of wastes from forest cleaning activities and/or from local wood processing industries, were also included in this study.

173

2.2. Sampling method The burning tests were performed at the combustion facilities of the Department of Environment, University of Aveiro, Portugal. The facility structure and operation conditions were described in detail elsewhere (Calvo et al., 2011; Fernandes, 2009; Tarelho et al., 2011). Two types of residential biomass combustion appliances were selected for the source tests: i) a cast iron woodstove (SolzaimaÒ, model Sahara) with handheld control of combustion air (primary air underbed feed), and ii) a traditional Portuguese brick open fireplace with no control of combustion air. Both of them were operated manually in batch mode. Two experimental protocols were performed for each appliance and wood type: “cold start” and “hot start” to evaluate the influence of the temperature and fuel ignition process on the combustion fuel emission profile. The cold start experiments included the start-up phase, i.e. the combustion began with the appliance at ambient temperature. On the other hand, the hot start experiments were initiated by loading the combustion chamber with a batch of fuel, which was already at a temperature of around 100
C and with the presence of a small amount of burning char from a batch of fuel burned previously. During each combustion experiment, which lasted between 45 and 90 min, about 5.9 and 6.2 kg of wood was burned, using three consecutive batches of around 2 kg each. This procedure aimed at mimicking traditional fires in residential combustion appliances, since as the load of fuel burns down, a new batch of one or more logs is added to the bed of charcoal. Each combustion experiment lasted between 45 min and 90 min. The duration of the combustion experiments depended on the densities of the different biofuels. Softwoods (pine) are less dense and more resinous, burning faster than hardwoods (other wood types). The slight differences in starting temperatures of the “hot” combustion tests are also a result of distinct biofuel properties, with heating values for softwood ranging from 18.40 to 20.52 MJ kg1 and for hardwood ranging from 17.38 to 23.05 MJ kg1 (Telmo and Lousada, 2011). The wood was burnt as split logs of 30e50 cm in length and around 10 cm in diameter. Fires were ignited with pinecones and small kindling pieces cut from the same wood being burnt. The wood was purchased from a local seller and, in order to carry out the combustion tests in a reproducible manner, logs of approximately the same size were chosen. However, as it occurs in reality, it is impossible to replicate different burning cycles with exactly the same amount of wood and logs with the same absolute size. These small variations do not significantly affect the emissions, as it is argued in the discussion section. PM2.5 was collected in a dilution tunnel coupled to the woodstove or fireplace chimneys. A dilution tunnel was employed because it stimulates the rapid cooling and dilution that occurs as exhaust mixes with the atmosphere (Lipsky and Robinson, 2006). The dilution tunnel consisted of a cylindrical tube with 0.20 m internal diameter and 11 m length. The sampling of PM2.5 was made at a dilution ratio of 25:1. PM2.5 were collected using an Echo sampling head connected to a TECORA sampler (model 2.004.01, Italy) operating at a flow of 2.3 m3 h1, onto quartz fibre filters (47 mm diameter) located 10 m downstream the dilution tunnel entrance. 2.3. Polycyclic aromatic hydrocarbons extraction method The quartz fibre filters were previously fired at 500
C to eliminate organic contaminants and weighed before and after sampling to determine the mass of collected PM2.5. The weighing was performed by a microbalance (Sartorius M5P) after 24 h equilibrium in a room with controlled temperature and humidity. The mass of the PM2.5 collected on each filter was determined by triplicate weightings. If the replicate measurements disagreed from the original by more than 5%, filters were returned to the handling containers, placed in the conditioning environment, and weighted again after some time. Elemental carbon (EC) and organic carbon (OC) in quartz fibre filters were analysed by a thermal e optical transmission technique following a short multie step temperature protocol, first in inert N2 and then in an oxidising atmosphere (N2/ O2) (Alves et al., 2011). Half of the area of each quartz filter was sequentially extracted with dichloromethane and methanol (Fisher Scientific). After filtration, the extracts of each PM2.5 sample were vacuum concentrated and dried under a gentle nitrogen stream. The total organic extract was divided into five fractions by flash chromatography with silica gel and a variety of solvents of increasing polarity. The following solvents were used to elute the different compound classes: (1) 22.50 mL n-hexane (fraction 1, aliphatics); (2) 22.50 mL tolueneen-hexane (8.40:14.10) [fraction 2, polycyclic aromatic hydrocarbons (PAHs)]; (3) 22.50 mL n-hexane-dichloromethane (11.25:11.25) (fraction 3, carbonyl compounds); (4) 30.00 mL ethyl acetateen-hexane (12.00:18.00) (fraction 4, n-alkanols, sterols and other hydroxyl compounds); and (5) 30.00 mL solution of pure formic acid in methanol (4%, v/v) (fraction 5, acids and sugars). After separation, each fraction was vacuum concentrated and evaporated by ultra pure nitrogen stream. Details on the methodology for the extraction of organic compounds have been reported elsewhere (Alves et al., 2011). The fractionated extracts were then analysed by gas chromatography e mass spectrometry (GC model 6890, quadrupole MSD 5973 from HewlettePackard and GC Trace Ultra, quadrupole DSQ II from Thermo Scientific). Extracts of PAHs were coinjected with a mixture of deuterated internal standards: 1,4-dichlorobenzene-D4, naphtalene-D8, acenaphthene-D10, phenanthrene-D10, chrysene-D12 and

174

B. Vu et al. / Environmental Pollution 166 (2012) 172e181

perylene-D12. Calibration was based on the EPA610 mix (acenaphthene, acenaphthylene, anthracene, benzo[a]anthracene, benzo[a]pyrene, benzo[b]fluoranthene, benzo[k]fluoranthene, benzo[ghi]perylene, chrysene, dibenzo[a,h]anthracene, fluoranthene, fluorene, indeno[1,2,3-cd]pyrene, naphthalene, phenanthrene and pyrene), but also included some other aromatics (benzo[j]fluoranthene, benzo[e] pyrene, perylene, p-terphenyl and retene), which were found to be representative compounds after injection of the organic extracts in the scan mode (Gonçalves et al., 2011). To obtain a screening of all individual compounds present in each PAH extract, the first injection was in the full scan mode. The second injection of each extract used the much more sensitive and selective SIM mode to look for specific PAHs that were suspected to be present. Although other aromatic forms (e.g. Cl-PAHs and nitro-PAHs) have not been identified in the full scan mode, their presence in the PAH extracts, at levels below the detection limit or co-eluted with other chromatographic peaks, cannot be ruled out. Thus, a very small mass fraction of PAHs may have not been quantified. The PM2.5 particle emission factor was determined by particle mass and mass of fuel burnt in dry basis, taking into account the dilution applied to the exhaust. The PAH emission factors were calculated as a mass fraction of total OC, which, in turn, was expressed as a mass fraction of total PM2.5 emitted. After separated from other organic fractions, PAH extracts obtained from each combustion experiment were dissolved in 1.6 mL of DMSO. Dilutions of these extracts were tested in the Ames assays.

average number of revertant colonies in the plates was two times greater than those recorded in the negative control plates (Mortelmans and Zeiger, 2000). According to OECD guidelines (1997), this last criterion is sufficient to assume mutagenic potential, even without a doseeeffect relationship. When suspicions arouse about the potential toxicity of some extracts, one-way analysis of variance (ANOVA) followed by a Dunnet test were performed to check for a significant decrease in the number of revertant colonies on plates in comparison with the negative control. A level of significance of a ¼ 0.05 was used to reject the null hypothesis. The statistical analysis was performed using the SPSS 17.0 software for Windows. Regression analyses, using the same software were performed, when an apparent dose-related increase was recorded, to check for week mutagen extracts. Principal component analysis (PCA) was performed to reveal patterns among extracts, based on all the variables assessed, using burning devices (fireplace, woodstove) and combustion conditions (hot start, cold start) as covariables (Quinn and Keough, 2002). First and in order to test for a liner distribution in extracts data, an assumption of the PCA analysis (Quinn and Keough, 2002), a detrended correspondence analysis (DCA) was performed. Since ordination axes length recorded was lower than 1SD, an indication of linearity, the PCA analysis was performed.

3. Results Table 1 presents the emission factors of PM2.5 and PAHs from combustion of each wood species by woodstove and fireplace, during the cold start combustion and subsequently with hot combustion chamber. Twenty one PAHs have been detected in the wood smoke, including 13 carcinogenic/mutagenic PAHs originated from wood combustion characterised by IARC (1998) (benzo[a]fluoranthene, benzo[b]fluoranthene, benzo[k]fluoranthene, benzo[i]fluoranthene, benzo[e]pyrene, benzo[ghi]perylene, chrysene, dibenzo [ah]anthracene, indeno[1,2,3-cd]pyrene, perylene and pyrene). The majority of wood smoke samples contain 17-20 PAHs. Wood smoke from hot combustion of maritime pine by woodstove contains the least number of PAHs, only eight. Emission factors of 13 carcinogenic/mutagenic PAHs (mg kg1 wood burned, dry basis) in wood smoke are also included in Table 1. Eucalypt, Holm oak, olive, Portuguese oak and cork oak have the highest emissions of those PAHs. Table 1 also shows the emission factors of fluoranthane and benzo[a]pyrene (BaP) (mg kg1 wood burned, dry basis) obtained in each combustion experiment, as those PAHs account for the majority of mutagenicity of PAHs emitted by wood combustion (Gustafson et al., 2008). Combustion of olive and Holm oak releases the largest quantity of these two PAHs. According to the PCA analysis (Fig. 1), the first component was mainly correlated with carcinogenic PAHs and with BaP and fluoranthene emission factors. This component was also responsible by joining almost all the extracts in the negative part setting them apart from extracts obtained from Holm oak, olive, eucalyptus and

2.4. Mutagenicity testing method The total PAH extracts from each combustion experiment were tested for mutagenicity performing the Ames assay with S. typhimurium TA98 and TA100 strains with and without metabolic activation by the S9 fraction (Trinova BiochemÒ GmbH, Giessen, Germany) in order to assess the indirect and direct-acting, frameshift and base-pair substitution mutagens. Each sample was tested in four to five doses in range of mg L1, decreasing in two e fold magnitude, with three replicates per dose. The test was performed in plates containing two distinct layers of agar: 25 mL of the bottom agar (glucose minimal agar) to provide support media and 2 mL of top agar to deliver the extract concentration, the S9 mix (in assays with S9) or buffer (in assays without S9) and the tester strain to the bottom agar. All reagents were prepared according to Maron and Ames (1983) and Mortelmans and Zeiger (2000). The tester strain was pre-incubated overnight in Oxoid nutrient Broth number 2 (Oxoid, England). The rat liver microsomal fractions for metabolic activation S9 was obtained in lyophilised form, purchased from Trinova BiochemÒ GmbH. In each test, a negative control consisting of DMSO solvent blank (50 mL/plate) was included to determine the spontaneous revertants. In addition, a positive control consisting of known mutagens was used to confirm the reversion properties and specificities of each strain, activity of the S9 mix and other components presented in the assay. For experiments without S9, 2 e nitrofluorene (10 mg/plate) and sodium azide (10 mg/ plate) were used in positive controls for test with S. typhimurium TA98 and TA100, respectively. For assays with the inclusion of S9, 2-aminoanthracene (10 mg/plate) was used in positive control for both bacterial strains, TA98 and TA100. The plates were incubated at 37
C, for 48 h, and subsequently, the number of revertant colonies formed in each plate was counted. 2.5. Statistical analysis For the mutagenicity tests, results were considered positive, either when a doserelated increase in the number of revertant colonies was observed or when the

Table 1 Emission factors of PM2.5, total PAHs, 13 carcinogenic/mutagenic PAHs, and fluoranthene and benzo[a]pyrene for domestic combustion. Wood species

Briquettes Cork oak Eucalypt Golden wattle Holm oak Maritime pine Olive Portuguese oak a b c d

PM2.5 emission factor (g kg1 wood burned, dry basis)a

PAH emission factor (mg kg1 wood burned, dry basis)b

13 carcinogenic/mutagenic PAH emission factor (mg kg1 wood burned, dry basis)c

Fluoranthene and benzo[a] pyrene emission factor (mg kg1 wood burned, dry basis)d

Woodstove

Fireplace

Woodstove

Fireplace

Woodstove

Fireplace

Woodstove

Fireplace

Hot start

Cold start

Hot start

Cold start

Hot start

Cold start

Hot start

Cold start

Hot start

Cold start

Hot start

Cold start

Hot start

Cold start

Hot start

Cold start

3.36 12.05 10.12 9.12 4.69 1.66 7.66 16.01

14.69 18.75 19.26 11.55 14.15 5.62 22.29 25.83

18.36 11.99 11.68 0.84 21.71 5.94 20.22 5.85

9.14 28.99 19.06 10.45 14.81 8.11 26.01 27.93

10.18 19.03 13.58 17.99 10.36 17.70 33.31 16.03

99.22 29.11 29.41 30.97 35.28 47.59 70.49 74.51

43.84 22.62 6.11 1.76 78.42 23.91 45.26 14.17

25.88 68.83 31.10 22.40 31.46 81.05 112.24 49.12

1504 2958 3628 4908 3808 2039 5348 1742

2995 3554 12,648 5356 13,969 2642 17,751 3151

836 2337 737 377 37,122 2042 2560 560

1703 11,997 2347 5180 3938 6476 19,213 18,549

289 339 1080 1233 1000 139 1474 226

722 781 2921 1278 3584 630 4189 676

220 672 216 120 9182 521 622 181

466 2127 640 1221 918 485 4487 4025

Gonçalves et al., 2011. PAH emission factor ¼ (total PAHs/OC)*(OC/PM2.5)*(PM2.5 emission factor). 13 carcinogenic/mutagenic PAH emission factor ¼ (13PAHs/OC)*(OC/PM2.5)*(PM2.5 emission factor). Fluoranthene and benzo[a]pyrene emission factor ¼ [(fluoranthene þ BaP)/OC]*(OC/PM2.5)*(PM2.5 emission factor).

B. Vu et al. / Environmental Pollution 166 (2012) 172e181

175

those more correlated with component 1 mainly explained by carcinogenic/mutagenic PAHs (except for samples from Portuguese cork and oak, obtained from the fireplace with cold start) (Fig. 1). The remaining samples that were not determined as mutagen/ weak mutagen are inclusive for two reasons: there are scattered elevated revertant counts and the testing concentration is far below the maximum concentration 5e10 mg/plate suggested by Mortelsman and Zeiger (2000) for pure compounds. Six samples were statistically proven having significant decrease in the number of revertant colonies on plates in comparison with the negative control by one-way ANOVA followed by a Dunnet test (p < 0.05). All of them are in test with metabolic activation. The detailed statistical results of those samples are presented in Table 4. Since these samples have proved to be toxic to one of the S. typhimurium strains, it is not feasible to conclude about their genotoxicity/mutagenicity. 4. Discussion 4.1. Polycyclic aromatic hydrocarbons emissions

Fig. 1. PCA biplot of all the emission extracts based on emission factors of PM2.5, total PAHs, 13 carcinogenic/mutagenic PAHs, and fluoranthene and benzo[a]pyrene. Brbriquettes; Co-cork oak; Eu-eucalypt; Gw-golden wattle; Ho-Holm oak; Mpmaritime pine; O-olive; Po-Portuguese oak; W-woodstove; F-fireplace; HSehot start, CS-cold start. Circles surrounding samples for which a positive result (based on the criteria of an induction factor greater than 2, comparatively to the control) was recorded in the Ames assay without S9.

Portuguese oak, almost all from cold start. However, this analysis did not reveal a clear pattern in the distribution of samples related to burning devices and conditions. The eigenvalues for the first two components were 0.930 and 0.001, and the component of the PCA accounted for 99.9% of the total variability among extracts. Since covariables contributed for explaining an insignificant proportion of the variability, only the analysis performed without the covariables matrix is presented. The extract obtained from Holm oak burned on a fireplace with a hot start was the one more correlated with carcinogenic PAHs emission factors. The number of revertant colonies obtained from the mutagenicity tests of PAH extracts of PM2.5 particles emitted in the combustion of the seven wood species and briquettes by woodstove and fireplace at two burning conditions (hot start and cold start) are presented in Table 2 and Table 3. Fourteen in total mutagenic tests of this study have an apparent dose-related increase in the number of revertant colonies. They were samples obtained from combustion of briquettes, eucalypt, Holm oak, maritime pine, olive, and Portuguese oak. However, only for the olive burning extract obtained in fireplace with hot start conditions and, tested for TA98 without S9, a significant doseeresponse line was obtained (r2 ¼ 0.676; df ¼ 1.8; p ¼ 0.004). This sample was already characterised as mutagenic based on the second criteria presented above. When the two-fold principle was applied, five samples showed positive result in tests with strain TA98 and TA100 in the absence of S9 mixture (Tables 2 and 3) The positive samples in Ames with strain TA98 are Holm oak (woodstove, hot start), cork oak (fireplace, cold start), olive (fireplace, hot start) and Portuguese oak (fireplace, cold start). The only wood species whose combustion released an extract mutagenic in Ames test with TA100 was eucalypt, using woodstove in hot start condition. According to the PCA biplot, all the samples for which a strong positive result was obtained in the Ames test without S9, where not

PM2.5 emissions were higher during the combustion experiments in the fireplace than in the woodstove (Table 1). In general, emission factors of particulate matter, total PAHs, the 13 mutagenic PAHs, and flouranthene and BaP were higher during the experiments with cold start for both appliances. Lower temperatures result in a lower degree of conversion (oxidation) of the biomass (solid and pyrolysis products), thus originating a higher emission of unburned chemical compounds. The particulate matter emissions obtained from the combustion of maritime pine, the only softwood burned, were the lowest among all, whether in the woodstove or in the fireplace. Emissions from the woodstove at higher temperatures included a higher PAH content in terms of mass ratio to OC (Gonçalves et al., 2011) than those of the fireplace, whereas comparable average total PAH emissions were obtained in both combustion appliances during cold start combustion. It is known that PAH amounts increase with the increase of temperatures and oxygen content in supplied air (Lu et al., 2009; and references therein). In the present study, the fireplace was operated at lower temperatures than those of the woodstove, but the O2 content in supplied air was around 20%, while values lower than 15% were provided to the woodstove (Gonçalves et al., 2011). In spite of PAH emissions being dependent on the combustion temperature, their proportions in particulate matter only are remarkably different when comparing modern high-efficient to traditional equipments, such as those used in this study. In fact, PAHs are extensively formed from wood combustion at temperatures above 800
C, but not below 600
C (Alén et al., 1996; Kjällstrand and Petersson, 2001). Besides the combustion temperature, which is directly related to the type of wood-burning appliance, the biofuel type constitutes another factor that may influence the PAH emissions. Softwood (e.g. pine) is usually less dense and more resinous, produces more flames and sometimes more intense heat, but over a shorter period of time, burning faster than hardwood. Thus, softwoods generally produces higher PAH mass fractions in particle emissions than hardwoods. In this study, the PAH emissions from pine combustion were significantly higher than those of the other wood types, especially in emissions from woodstove. Combustion of softwood versus hardwood also generates different emission profiles for specific aromatic compounds. Retene was the dominant PAH found in the softwood smoke, whilst it was present at very small levels in the hardwood combustion emissions. Retene is the fully aromatised thermal alteration product of the resin acids present in conifer woods (Ramdahl, 1983). During air starved conditions at

Wood species

Hot start ng PAHs/plate

Cold start Number of revertants (mean  SD)a TA98

Briquettes

Cork oak

Golden wattle

Holm oak

Maritime pine

Olive

23.5 13.0 14.0 12.5 12.5 12.3 136.3 16.0 17.0 13.7 17.3 16.0 16.7 121.7 12.5 11.7 12.3 15.7 10.0 16.7 121.7 15.0 11.0 9.5 14.5 8.5 12.3 136.3 25.0 23.0 17.3 15.0 18.7 12.0 149.3 14.3 17.3 14.3 20.7 12.7 12.0 149.3 14.0 15.3 14.0 18.3 13.7 12.0 149.3

TA100  2.1(1.91)b  4.2(1.05)  2.8(1.04)  2.1(1.01)  4.9(1.01)  3.5  2.5  1.0(0.96)  2.6(1.02)  5.7(0.82)  7.5(1.04)  1.7(0.96)  4.2  36.3  0.7(0.75)  2.1(0.70)  3.2(0.74)  3.2(0.94)  1.7(0.60)  4.2  36.3  1.4(1.22)  2.8(0.89)  2.1(0.77)  2.1(1.18)  6.4(0.69)  3.5  2.5  0.0(2.08)  3.0(1.92)  3.2(1.44)  2.0(1.25)  3.2(1.56)  5.6  17.8  2.9(1.19)  5.5(1.44)  1.5(1.19)  1.5(1.73)  1.5(1.06)  5.6  17.8  1.0(1.17)  5.5(1.28)  2.6(1.17)  3.2(1.52)  3.1(1.14)  5.6  17.8

364.3 455.3 432.0 504.7 372.7 561.7 UC 147.0 148.3 143.0 144.7 157.0 97.3 568.3 207.5 184.7 196.7 153.3 168.0 97.3 568.3 257.7 332.3 270.7 244.0 378.0 259.0 823.0 263.3 294.3 303.3 317.3 244.7 259.0 823.0 311.1 301.7 344.7 326.3 305.3 259.0 823.0 280.3 280.7 290.0 253.7 304.0 259.0 823.0

     

ng PAHs/plate TA98/S9

162(0.65) 22.3(0.81) 96.6(0.77) 208(0.90) 98.0(0.58) 47.4

 20.7(1.51)  14.4(1.52)  13.1(1.47)  5.7(1.49)  26.6(1.61)  31.9  120.7  54.4(2.13)c  20.6(1.90)  27.2(2.02)  33.1(1.58)  22.0(1.73)  31.9  120.7  34.6(0.99)  86.8(1.28)  13.7(1.05)  25.2(0.94)  64.2(1.46)  48.3  61.5  43.9(1.02)  8.1(1.14)  44.4(1.17)  17.8(1.23)  38.8(0.94)  48.3  61.5  8.0(1.20)  46.9(1.16)  76.2(1.33)  107.0(1.26)  50.8(1.18)  48.3  61.5  29.0(1.08)  5.9(1.08)  40.0(1.12)  26.3(0.98)  39.0(1.17)  48.3  61.5

21.9 19.0 17.5 14.0 12.7 22.0 279.7 15.0 12.3 16.7 13.7 25.7 20.0 364.7 20.5 19.5 18.0 14.7 18.0 20.0 364.3 19.0 16.0 17.3 12.3 18.7 15.0 259.7 16.7 19.0 21.0 16.0 13.7 15.0 259.7 18.3 18.7 15.3 17.7 15.7 15.0 259.7 19.0 21.3 16.0 14.7 21.3 15.0 259.7

                                                

TA100/S9 2.9(1.00) 1.4(0.86) 0.7(0.80) 3.5(0.64)* 3.1(0.58)* 1.0 25.8 1.7(0.75) 0.6(0.62) 6.4(0.84) 2.5(0.69) 6.5(1.29) 2.6 11.7 2.1(1.03) 3.5(0.98) 0.0(0.90) 1.5(0.74) 5.6(0.90) 2.6 111.7 2.6(1.27) 1.7(1.07) 3.2(1.15) 0.6(0.82) 3.5(1.25) 1.0 27.3 4.5(1.11) 4.6(1.27) 3.0(1.4) 1.0(1.07) 2.5(0.91) 1.0 27.3 3.2(1.22) 1.2(1.23) 1.2(1.02) 2.1(1.18) 3.2(1.05) 1.0 27.3 3.5(1.27) 7.2(1.42) 1.7(1.07) 3.8(0.98) 2.1(1.42) 1.0 27.3

87.0 96.0 86.7 98.3 89.7 87.7 360.0 96.7 108.7 120.7 93.7 104.3 98.3 405.0 114.3 99.0 102.7 100.3 111.7 98.3 405.0 84.7 74.7 87.0 86.0 96.7 87.7 360.0 71.7 67.3 68.3 71.3 65.3 78.3 331.7 56.7 79.0 72.3 72.3 60.3 78.3 331.7 78.0 64.7 70.0 64.7 63.7 78.3 331.7

 7.0(0.99)  24.4(1.10)  15.2(0.99)  9.1(1.12)  3.1(1.02)  11.6  43.6  4.7(0.98)  11.0(1.11)  14.2(1.23)  18.6(0.95)  10.6(1.06)  10.1  36.3  18.5(1.16)  18.2(1.01)  16.0(1.04)  2.9(1.02)  19.9(1.14)  10.1  36.3  6.1(0.97)  10.0(0.85)  8.5(0.99)  6.6(0.98)  10.4(1.10)  11.6  43.6  16.3(0.91)  6.0(0.86)  5.1(0.87)  4.2(0.91)  4.7(0.83)  4.6  76.1  5.5(0.72)*  18.4(1.01)  2.1(0.92)  4.0(0.92)  4.2(0.77)  4.6  76.1  39.9(1.00)  6.7(0.83)  2.6(0.89)  7.5(0.83)  7.2(0.81)  4.6  76.1

Number of revertants (mean  SD) TA98

1001 500 250 125 63 0 PC 235 117 59 29 15 0 PC 567 283 142 71 35 0 PC 314 157 78 39 20 0 PC 388 194 97 49 24 0 PC 1064 532 266 133 66 0 PC 630 315 158 79 39 0 PC

TA100

12.0 10.0 14.7 14.0 11.3 9.0 213.3 15.5 13.0 11.0 8.0 10.0 12.3 136.3 15.0 9.5 14.0 9.5 13.0 12.3 136.3 16.5 23.0 13.3 14 7.0 12.3 136.3

 2.8(1.33)  0.0(1.11)  8.1(1.66)  3.6(1.56)  5.9(1.26)  5.6  8.7  0.7(1.26)  1.4(1.05)  2.8(0.89)  0.0(0.65)  1.4(0.81)  3.5  2.5  2.8(1.22)  0.7(0.77)  2.8(1.14)  3.5(0.77)  0.0(1.05)  3.5  2.5  2.1(1.33)  4.2(1.86)  2.3(1.13)  1.4(1.14)  1.4(0.57)  3.5  2.5

10.0 10.5 13.5 19.5 12.3 136.3 12.7 15.3 18.0 14.5 18 16.7 121.7 11.7 11.0 11.0 13.5 12.5 12.3 136.3

 5.7(0.81)  0.7(0.85)  3.5(1.09)  3.5(1.58)  3.5  2.5  3.8(0.76)  4.6(0.92)  4.2(1.08)  0.7(0.87)  1.4(1.08)  4.2  36.3  8.3(0.95)  2.8(0.89)  4.2(0.89)  0.7(1.09)  7.1(1.01)  3.5  2.5

121.7  116.7  145.3  156.3  158.3  97.3  568.3  142.5  138.0  137.5  102.5  181.0  119.3  610.7  289.7  277.7  330.3  297.7  268.3  259.0  823.0  323.0  299.0  329.3  269.0  259.7  259.0  823.0  451.5  519.3  537.0  450.7  612.7  561.7  UC 152.7  146.7  145.0  144.3  146.3  97.3  568.3  643.0  577.7  429.3  353.7  504.0  561.7  UC

TA98/S9 31.0(1.25) 17.0(1.20) 38.9(1.49) 28.3(1.61) 19.3(1.63) 31.9 120.7 3.5(1.19) 12.7(1.16) 17.7(1.17) 13.4(0.86) 67.9(1.51) 33.1 12.9 45.5(1.12) 19.5(1.07) 31.5(1.28) 22.4(1.15) 19.6(1.04) 48.3 61.5 14.9(1.25) 24.6(1.15) 19.5(1.27) 33.7(1.04) 62.6(1.00) 48.3 61.5 44.5(0.80) 90.7(0.92) 61.9(0.96) 97.6(0.80) 172.6(1.09) 47.4 10.7(1.57) 1.5(1.51) 8.2(1.49) 18.8(1.48) 30.7(1.50) 31.9 120.7 40.7(1.14) 81.1(1.03) 54.6(0.76) 78.2(0.63) 80.1(0.90) 47.4

18.7 17.3 17.0 16.3 19.0 20.0 364.3 20.7 15.7 20.3 19.3 16.7 22.0 279.7 21.3 17.3 17.3 15.7 15.3 15.0 259.7 21.7 16.7 18.7 21.7 20.0 15.0 259.7 26.5 18.0 15.0 12.3 13.3 22.0 279.7 19.0 18.3 18.3 15.0 18.3 20.0 364.7 26.7 24.0 17.7 19.3 16.5 22.0 279.7

 6.0(0.94)  4.2(0.87)  4.7(0.85)  1.5(0.82)  5.3(0.95)  2.6  111.7  4.6(0.94)  4.7(0.73)  2.1(0.92)  8.5(0.88)  2.3(0.76)  1.0  25.8  1.5(1.42)  4.7(1.15)  5.7(1.15)  4.9(1.05)  0.6(1.02)  1.0  27.3  2.1(1.45)  4.5(1.11)  8.5(1.25)  2.1(1.45)  3.6(1.33)  1.0  27.3  0.7(1.20)  4.2(0.82)  4.4(0.68)*  0.6(0.56)*  4.0(0.60)*  1.0  25.8  1.0(0.95)  4.9(0.92)  8.3(0.92)  3.5(0.75)  5.7(0.92)  2.6  11.7  4.7(1.21)  7.2(1.09)  3.5(0.80)  0.6(0.88)  4.9(0.75)  1.0  25.8

TA100/S9 123.0 123.7 117.3 120.0 134.0 98.3 405.0 105.0 99.0 90.7 112.0 105.3 360.0 87.7 74.0 91.0 77.7 80.7 78.3 87.7 360.0 94.0 47.3 61.3 63.3 63.3 78.3 331.7 88.7 87.0 96.0 81.0 97.5 87.7 360.0 125.0 121.0 106.0 109.0 91.7 98.3 405.0 106.7 99.3 93.0 91.7 96.3 360.0 87.7

 7.2(1.25)  24.2(1.26)  30.0(1.19)  8.9(1.22)  3.0(1.36)  10.1  36.3  13.0(1.20)  16.1(1.13)  8.1(1.03)  2.6(1.28)  10.1(1.20)  43.6  11.6  3.6(0.84)  9.6(1.04)  2.5(0.89)  7.0(0.92)  19.9(0.89)  11.6  43.6  19.8(1.20)  6.8(0.60)*  7.6(0.78)  0.6(0.81)  12.6(0.81)  4.6  76.1  4.5(1.01)  10.8(0.99)  6.6(1.10)  12.3(0.92)  19.1(1.11)  11.6  43.6  22.1(1.27)  2.6(1.23)  7.2(1.08)  7.8(1.11)  10.6(0.93)  10.1  36.3  9.5(1.22)  5.7(1.13)  6.2(1.06)  19.6(1.05)  16.6(1.10)  43.6  11.6

B. Vu et al. / Environmental Pollution 166 (2012) 172e181

Eucalypt

225 113 56 28 14 0 PC 240 120 60 30 15 0 PC 236 118 59 29 14 0 PC 302 151 75 38 19 0 PC 138 69 34 17 9 0 PC 474 237 119 59 30 0 PC 618 309 154 77 39 0 PC

176

Table 2 Mutagenicity of PAH extracts of particles emitted in domestic wood combustion by woodstove, hot start and cold start, to S. typhimurium TA98, TA100 in the absence and presence of S9 mixture.

*Values significantly different from the negative control (p < 0.05), n ¼ 3. a Values are means  Standard Deviation of 2e3 plates. b Numbers in parenthesis are the ratios of test values to negative control (in DMSO) values. c Ratios of test values to negative control equal to or above 2 are marked in bold; 0: negative control in DMSO; PC: positive control; UC: uncountable.

Portuguese oak

234 117 59 29 15 0 PC

23.5 13.0 14.0 12.5 12.5 12.3 136.3

      

2.1(1.91) 4.2(1.05) 2.8(1.14) 2.1(1.01) 4.9(1.01) 3.5 2.5

323.7 309.3 278.7 287.7 310.3 259.0 823.0

      

72.5(1.25) 20.3(1.19) 17.4(1.08) 12.7(1.11) 39.3(1.20) 48.3 61.5

15.3 16.7 13.7 15.7 20.7 15.0 259.7

 1.5(1.02)  4.5(1.11)  0.6(0.91)  1.2(1.05)  0.6(1.38)  1.0  27.3

100.3 89.0 70.3 85.7 78.0 87.7 360.0

 10.8(1.14)  6.1(1.02)  5.5(0.80)  21.1(0.98)  2.6(0.89)  11.6  43.6

685 343 171 86 43 0 PC

16.3 10.0 16.0 12.7 13.3 16.7 121.7

 3.8(0.98)  5.6(0.60)  7.0(0.96)  4.5(0.76)  2.9(0.80)  4.2  36.3

169.7 159.3 171.0 156.0 167.7 97.3 568.3

 10.0(1.74)  3.5(1.64)  1.7(1.76)  24.5(1.60)  20.3(1.72)  31.9  120.7

21.5 17.0 18.3 15.0 22.0 20.0 364.7

 4.9(1.08)  2.6(0.85)  2.5(0.92)  3.6(0.75)  1.0(1.1)  2.6  11.7

132.0  128.0  117.7  130.0  138.3  98.3  405.0 

9.5(1.34) 37.6(1.30) 16.0(1.20) 9.6(1.32) 25.1(1.41) 10.1 36.3

B. Vu et al. / Environmental Pollution 166 (2012) 172e181

177

high burn rates, the total PAH emissions increased considerably. It is a rather well-known fact that overloading the appliance, especially with dry and small wood logs, can give too high heat outputs resulting in air starved combustion conditions (Pettersson et al., 2011). It should be noted that, during the combustion experiments of this study, the continuous monitoring of the combustion process was carried out by measuring the composition of flue gases (O2, CO2 and CO) at the exit of the chimney (Tarelho et al., 2011). The loading of batches of fuel, with approximately the same sizes and avoiding overcharged, were done in a reproducible manner between fireplace and woodstove and between different wood types. It was observed that, independently from the wood type, the general CO and CO2 concentration profiles were very similar. In terms of flue gas composition, the main difference between the two burning appliances was observed for the O2 concentration profiles. Regardless of wood type, during the initial 15 min of the combustion in the woodstove, a decrease from 21% to values around 15% in the O2 concentration was registered. Afterwards, the O2 concentrations increased gradually. The decrease in the O2 concentration in the first period of combustion in the fireplace was not as marked as in the woodstove, since much higher volumes of air enter the combustion chamber. Thus, the more pronounced oxygen-starving conditions observed during the first period of each combustion cycle in the woodstove contribute to higher PAH emissions in comparison with those from the fireplace. In, summary, variations in PAH emissions are mainly due to distinct structural and operatory characteristics of the two appliances, which in turn affect the combustion conditions, rather than to the wood type or the small variations between each experimental burning test (starting combustion temperature, log size and weight). 4.2. Mutagenic activities of different wood species Five in total eight wood species used in this study had emissions with mutagenic activity in Ames assay with S. typhymirium TA98 and/or TA100. They were cork oak, eucalypt, Holm oak, olive, and Portuguese oak. The opposite was recorded only for briquettes, Golden wattle, and maritime pine. All of them were mutagenic in the absence of metabolic agent. Direct-acting mutagenicity in extracts of combustion products of wood has been linked to the presence of polar fraction such as nitroarenes (Claxton et al., 2004); however it is not applicable in all the cases, such as coconut smoke (Bell and Kamens, 1990). Four of the five direct-acting mutagenic samples originated a positive response by TA98, suggesting a shift mutation mechanism in the induction of mutagenicity (Mortelmans and Zeiger, 2000). The remaining sample was mutagenic to TA100, which implies the base-pair substitution mutation mechanism. For all of the mutagenic samples, when S9 was introduced to the test, the mutagenic activity disappeared. This suggests that these samples contained direct-acting base-pair and frameshift mutagens that loose their mutagenicity after being metabolised by enzymes from the S9 liver fraction. Similar trend was observed in the study of Oanh et al. (2002) where the mutagenicity of organic extract from wood combustion decreased when S9 was added. Other samples induced mutagenicity, if had, in one of four possible mechanisms: direct/indirect-acting frameshift/base-pair substitution, either restrained by or reinforced by metabolic activation. Studies have been carried out to compare mutagenicity of different wood species. McCrillis et al. (1992) concluded that emissions from burning pine were more mutagenic than those from burning oak. This is in contrast with the results from this study, where combustion emissions of Holm oak were more potent

Wood species

Hot start ng PAHs/plate

Briquettes

Cork oak

Golden wattle

Holm oak

Maritime pine

Olive

Cold start Number of revertants (mean  SD)a

ng PAHs/plate

TA98

TA100

TA98/S9

TA100/S9

16.0  5.2(0.96)b 19.0  2.6(1.14) 14.3  3.5(0.86) 13.7  3.5(0.82) 14.7  2.9(0.88) 16.7  4.2 121.7  36.3 12.7  4.0(1.41) 8.7  3.2(0.97) 10.0  1.7(1.11) 8.3  2.5(0.92) 10.3  3.8(1.14) 9.0  5.6 213.3  8.7 16.0  0.0(1.30) 15.5  2.1(1.26) 14.5  7.8(1.18) 17.0  7.1(1.38) 12.5  0.7(1.01) 12.3  3.5 136.3  2.5 11.0  4.6(1.22) 10.7  3.5(1.19) 8.7  5.7(0.97) 10.0  5.6(1.11) 12.0  3.6(1.33) 9.0  5.6 213.3  8.7 14.3  1.2(1.19) 18.3  2.9(1.53) 13.7  1.5(1.14) 14.0  6.6(1.17) 22.3  7.1(1.86) 12.0  5.6 149.3  17.8 21.7  4.0(1.81) 21.7  1.2(1.81) 20.5  3.5(1.71) 18.0  10(1.50) 12.0  2.8(1.00) 12.0  5.6 149.3  17.8 27.0  1.4(2.19)c 19.0  2.8(1.54) 19.5  3.5(1.58) 20.0  2.8(1.62) 9.5  0.7(0.77) 12.3  3.5 136.3  2.5

171.3  4.5(1.76) 178.0  24.6(1.83) 178.3  26.7(1.83) 175.7  14.0(1.81) 156.7  6.7(1.61) 97.3  31.9 568.3  120.7 126.0  12.7(1.06) 101.0  11.3(0.85) 115.5  3.5(0.97) 119.5  6.4 (1.00) 111.5  19.1(0.93) 119.3  33.1 610.7  12.9 129.3  1.2(1.33) 145.0  22.9(1.49) 146.7  21.4(1.51) 168.3  34.0(1.73) 152.7  11.2(1.57) 97.3  31.9 568.3  120.7 103.5  19.1(0.87) 107.0  0.0(0.90) 110.5  6.4(0.93) 459.7  0.0(0.95) 96.5  2.1(0.81) 119.3  33.1 610.7  12.9 302.3  49.7(1.17) 368.3  118(1.42) 271.7  16.1(1.05) 281.3  45.0(1.09) 266.3  28.5(1.03) 259.0  48.3 823.0  61.5 313.3  24.5(1.21) 303.3  27.3(1.17) 332.3  50.0(1.28) 279.7  32.1(1.08) 338.0  51.9(1.31) 259.0  48.3 823.0  61.5 159.3  10.7(1.64) 180.0  4.4(1.85) 152.7  28.0(1.57) 152.7  35.9(1.57) 193.3  4.0(1.99) 97.3  31.9 568.3  120.7

20.5  0.7(1.34) 14.7  3.8(0.98) 18.0  5.3(1.2) 17.0  1.0(1.13) 17.0  1.0(1.13) 15.0  1.0 259.7  27.3 26.0  2.8(1.18) 17.7  2.9(0.80) 15.7  4.5(0.71)* 21.3  2.9(0.97) 21.5  0.7(0.98) 22.0  1.0 279.7  25.8 26.5  3.5(1.77) 19.0  4.2(1.27) 26.5  13(1.77) 14.0  3.0(0.70) 15.0  7.2(0.75) 20.0  2.6 364.7  11.7 22.0  3.6(1.00) 24.3  4.9(1.10) 25.7  5.9(1.17) 23.3  2.9(1.06) 19.0  2.6(0.86) 22.0  1.0 279.7  25.8 14.7  4.6(0.98) 19.7  1.5(1.31) 16.0  2.6(1.07) 16.0  1.7(1.07) 15.0  2.6(1.00) 15.0  1.0 259.7  27.3 18.0  2.0(1.20) 16.3  5.9(1.08) 20.7  6.4(1.38) 16.7  4.6(1.11) 16.7  3.8(1.11) 15.0  1.0 259.7  27.3 19.0  1.7(0.95) 25.3  5.5(1.27) 22.7  5.0(1.14) 13.5  2.1(0.68) 13.0  2.8(0.65) 20.0  2.6 364.7  11.7

123.0  2.6(1.25) 115.7  27.0(1.18) 119.7  6.1(1.22) 128.3  11.9(1.31) 105.3  5.5(1.07) 98.3  10.1 405.0  36.3 113.5  9.2(1.29) 103.7  17(1.18) 128.0  20(1.46) 94.0  5.6(1.07) 125.7  23(1.43) 87.7  11.6 360.0  43.6 109.7  8.4(1.12) 110.0  20.7(1.12) 106.7  15.0(1.08) 105.6  16.0(1.07) 120.7  2.1(1.23) 98.3  10.1 405.0  36.3 103.7  5.1(1.18) 77.5  3.5(0.88) 90.0  8.7(1.03) 83.7  12.3(0.95) 94.7  11.9(1.08) 87.7  11.6 360.0  43.6 70.0  8.7(0.89) 69.0  14.7(0.88) 64.3  11.0(0.82) 47.7  15.0(0.61) 57.0  7.0(0.73) 78.3  4.6 331.7  76.1 83.7  15.9(1.07) 63.0  15.5(0.80) 69.7  5.5(0.89) 56.0  8.7(0.71) 87.7  23.1(1.12) 78.3  4.6 331.7  76.1 91.7  3.5(0.93) 95.3  11.6(0.97) 90.0  4.6(0.92) 90.0  15.4(0.92) 97.7  17.6(0.99) 98.3  10.1 405.0  36.3

378 189 95 47 24 0 PC 811 406 203 101 51 0 PC 360 180 90 45 23 0 PC 684 342 171 85 43 0 PC 473 237 119 59 30 0 PC 1124 562 281 140 70 0 PC 1040 520 260 130 65 0 PC

Number of revertants (mean  SD) TA98

TA100

TA98/S9

9.5  0.7(1.06) 10.3  4.0(1.14) 11.0  3.6(1.22) 11.7  1.2(1.3) 12.7  4.6(1.41) 9.0  5.6 213.3  8.7 12.3  0.6(1.37) 13.7  5.5(1.52) 18.7  4.0(2.08) 12.7  3.2(1.41) 13.7  1.5(1.52) 9.0  5.6 213.3  8.7 18.3  1.5(1.10) 17.6  2.6(1.05) 17.3  2.1(1.04) 14.7  5.1(0.88) 15.0  6.1(0.90) 16.7  4.2 121.7  36.3 12.0  2.6(1.33) 10.7  3.2(1.19) 9.7  1.5(1.08) 10.0  2.6(1.11) 10.7  2.1(1.19) 9.0  5.6 213.3  8.7 12.0  0.0(1.33) 15.7  3.5(1.74) 10.3  2.9(1.14) 10.0  2.6(1.11) 12.7  8.1(1.41) 9.0  5.6 213.3  8.7 15.3  3.2(0.92) 19.0  1.0(1.14) 14.7  5.1(0.88) 15.0  5.0(0.90) 16.7  2.1(1.00) 16.7  4.2 121.7  36.3 11.7  4.0(1.3) 11.0  7.0(1.22) 15.7  1.5(1.74) 11.0  0.0(1.22) 15.3  5.0(1.7) 9.0  5.6 213.3  8.7

133.5  3.5(1.12) 121.0  22.6(1.01) 102.0  17.0(0.85) 142.0  31.1(1.19) 142.0  2.8(1.19) 119.3  33.1 610.7  12.9 135.0  0.0(1.13) 142  45.3(1.19) 109.0  8.5(0.91) 114.5  0.7(0.96) 134.5  6.4(1.13) 119.3  33.1 610.7  12.9 345.0  120(1.33) 258.3  24.5(1.00) 295.3  57.7(1.14) 434.3  36.0(1.68) 299.7  61.7(1.16) 259.0  48.3 823.0  61.5 472.3  119(0.84) 475.7  29.9(0.85) 571.0  101(1.02) 498.0  85.5(0.89) 620.3  38.9(1.10) 561.7  47.4 UC 113.0  18.2(1.16) 166.0  42.3(1.71) 117.0  19.9(1.20) 83.0  0.0(0.85) 150  70.7(1.54) 97.3  31.9 568.3  120.7 147.0  20.1(1.51) 161.3  35.1(1.66) 162.0  19.3(1.66) 164.3  18.9(1.69) 157.3  15.4(1.62) 97.3  31.9 568.3  120.7 115.0  7.1(0.96) 109.0  18.4(0.91) 112.5  17.7(0.94) 110.0  2.8(0.92) 104.5  10.6(0.87) 119.3  33.1 610.7  12.9

18.7 18.0 17.0 16.7 15.0 22.0 279.7 21.0 16.0 19.0 12.0

 1.5(0.85)  1.0(0.81)  5.3(0.77)  1.5(0.76)  3.6(0.68)  1.0  25.8  4.2(1.26)  1.4(0.96)  5.7(1.14)  4.1(0.72)

22.0  1.0 279.7  25.8 19.7  3.2(1.31) 19.0  3.6(1.27) 18.7  1.5(1.25) 16.7  4.6(1.11) 15.0  1.0 259.7  27.3 26.7  3.5(1.21) 27.0  4.0(1.23) 20.3  0.6(0.92) 19.0  2.8(0.86) 26.0  1.7(1.18) 22.0  1.0 279.7  25.8 16.5  6.4(0.75) 19.7  4.5(0.90) 21.3  5.0(0.97) 23.3  0.6(1.06) 16.3  5.5(0.74) 22.0  1.0 279.7  25.8 21.7  2.5(1.05) 20.7  2.3(1.04) 14.7  3.1(0.74)* 15.0  2.6(0.75) 19.0  1.0(0.95) 20.0  2.6 364.7  11.7 26.7  4.7(1.21) 24.0  7.2(1.09) 17.7  3.5(0.80) 19.3  0.6(0.87) 16.5  4.3(0.75) 22.0  1.0 279.7  25.8

TA100/S9 105.3  9.3(1.20) 94.5  6.4(1.08) 94.7  15.0(1.08) 117.0  10.5(1.33) 106.0  12.7(1.21) 87.7  11.6 360.0  43.6 108.0  4.6(1.23) 97.7  17.9(1.11) 114.0  11.5(1.30) 106.0  14.1(1.21) 105.0  21.0(1.20) 87.7  11.6 360.0  43.6 72.0  1.0(0.92) 72.7  7.6(0.93) 65.3  8.6(0.83) 63.7  1.5(0.81) 94.0  38.1(1.2) 78.3  4.6 331.7  76.1 122.3  3.5(1.24) 144.0  61.6(1.46) 81.0  6.6(0.82) 100.7  2.1(1.02) 114.0  10.4(1.16) 98.3  10.1 405.0  36.3 105.7  24.8(1.07) 106.3  14.2(1.08) 98.3  14.6(1.00) 109.7  6.4(1.12) 107.3  19.9(1.09) 98.3  10.1 405.0  36.3 127.3  7.1(1.29) 107.3  4.6(1.09) 100.7  8.1(1.02) 105.7  13.3(1.07) 110.0  13.9(1.12) 98.3  10.1 405.0  36.3 98  9.8(1.12) 108.0  20.0(1.23) 126.3  22.2(1.44) 113.0  8.0(1.29) 119.3  22.5(1.36) 87.7  11.6 360.0  43.6

B. Vu et al. / Environmental Pollution 166 (2012) 172e181

Eucalypt

599 300 150 75 37 0 PC 442 221 110 55 28 0 PC 83 42 21 10 5 0 PC 272 136 68 34 17 0 PC 333 166 83 42 21 0 PC 461 231 115 58 29 0 PC 603 302 151 76 38 0 PC

178

Table 3 Mutagenicity of PAH extracts of particles emitted in domestic wood combustion by fireplace, hot start and cold start, to S. typhimurium TA98, TA100 in the absence and presence of S9 mixture.

*Values significantly different from the negative control (p < 0.05), n ¼ 3. a Values are means  Standard Deviation of 2e3 plates. b Numbers in parenthesis are the ratios of test values to negative control (in DMSO) values. c Ratios of test values to negative control equal to or above 2 are marked in bold; 0: negative control in DMSO; PC: positive control; UC: uncountable.

Portuguese oak

528 264 132 66 33 0 PC

12.3 16.3 17.3 13.7 13.0 9.0 213.3

      

2.1(1.37) 4.2(1.81) 4.9(1.92) 4.7(1.52) 2.6(1.44) 5.6 8.7

126.0 101.0 115.5 119.5 98.0 119.3 610.7

      

12.7(1.06) 11.3(0.98) 3.5(0.97) 6.4(1.00) 2.8(0.82) 33.1 12.9

34.0 26.3 30.7 24.0 22.3 22.0 279.7

      

4.2(1.55) 2.1(1.20) 6.4(1.40) 3.5(1.09) 2.1(1.01) 1.0 25.8

102.3 99.3 100.7 98.3 95.0 87.7 360.0

      

15.1(1.17) 21.5(1.13) 17.8(1.15) 16.9(1.12) 9.9(1.08) 11.6 43.6

674 337 168 84 42 0 PC

34.3 25.3 18.3 14.7 20.7 12.0 149.3

      

16.2(2.86) 5.5(2.11) 4.2(1.53) 7.6(1.23) 5.5(1.73) 5.6 17.8

327.7 284.7 300.7 309.3 320.7 259.0 823.0

      

39.3(1.27) 40.0(1.10) 30.4(1.16) 35.9(1.19) 32.2(1.24) 48.3 61.5

19.7  20.7  20.7  16.7  16.3  15.0  259.7 

6.4(1.31) 2.5(1.38) 5.9(1.38) 2.9(1.11) 4.2(1.09) 1.0 27.3

92.3 79.0 81.3 72.3 75.0 78.3 331.7

      

13.6(1.18) 10.5(1.01) 13.0(1.04) 8.4(0.92) 15.1(0.96) 4.6 76.1

B. Vu et al. / Environmental Pollution 166 (2012) 172e181

179

than those of pine. Therefore the conclusion of McCrillis et al. (1992) cannot be generalised for all oak and pine species. Mukherji et al. (2002) did a research comparing mutagenicity of organic matter extracts of PM2.5 from burning of wood (Acacia nilotica), dung cake and briquettes and found out that wood was less mutagenic compared to briquettes. This difference in mutagenicity was not observed in this study, in which briquettes emissions were found out to have comparable mutagenicity as some wood species. The mutagenic potency of EOM of burning other wood species has been established (Asita et al., 1991; Bell and Kamens, 1990; Mukherji et al., 2002; Oanh et al., 2002). Most of them were expressed in mutagenicity emission factors (revertant kg1 wood burnt or revertant g1 extracts). However, the mutagenicity emission factors were not calculated in this study, due to the fact that only PAH fraction was collected and tested in Ames assay. In addition to PAH fraction, the mutagenicity of organic extracts is due to an extensive number of compounds of various chemical classes, including oxygen-containing compounds, nitrogencontaining compound, halogen-containing compounds and so forth (Claxton et al., 2004). Further, there are various factors need to be taken carefully into account when comparing the results of mutagenicity tests between different studies such as the extraction method, bioassay, wood composition, moisture content, burning conditions (burning rate, burning temperature, air supply), stove design, etc (Claxton et al., 2004, 2007). Therefore, the comparison of mutagenic activity of combustion emissions of wood species in this study to other wood species was not carried out. Five wood species showed an average number of revertants on the plates significantly below (p < 0.05) that recorded on the negative control plates, suggesting the toxicity of the samples upon the bacteria strain. They were briquettes, cork oak, golden wattle, Holm oak, and maritime pine. The toxicity restrained the growth of bacteria and might mask the mutagenicity of those samples. Theoretically, repeats of those Ames tests at different range of concentration should be carried out in order to clarify that doubt; however, due to the limit in the sample quantity, those tests were not performed. 4.3. Influence of burning appliance and burning condition on the mutagenicity of emissions Combustion in hot start condition induced more direct-acting mutagenicity. In addition, burning by fireplace produced more mutagenic samples comparing to combustion carried out in woodstove. This is in agreement with the results of van Houdt et al. (1986), who found out that fireplace led to an increase in direct and indirect mutagenicity of biofuel emission in comparison to woodstove. When toxic samples were considered the opposite trend was recorded, the woodstove device was responsible for four of the six toxic extracts. Table 4 Samples with average number of revertants on plates significant below that on the negative control plates. Wood species/Bacteria straina

Burning condition

F

df

Briquettes TA98 Cork oak TA98 Golden wattle TA100 Holm oak TA98 Maritime pine TA100 Maritime pine TA98

Woodstove Hot start Fireplace Hot start Woodstove Cold start Woodstove Cold start Woodstove Hot start Fireplace Cold start

6.992 3.904 7.281 8.137 3.648 4.417

5, 5, 5, 5, 5, 5,

a

All samples were in test with the presence of metabolic agent.

p 10 10 11 10 12 12

0.005 0.032 0.003 0.003 0.031 0.016

180

B. Vu et al. / Environmental Pollution 166 (2012) 172e181

4.4. Correlation between polycyclic aromatic hydrocarbon emissions and mutagenicity In general, there were no evident relations between PAH emissions and the mutagenicity of the samples. In fact, as it was observed by the PCA biplot, only two of the strongest mutagenic extracts were correlated with carcinogenic PAHs. Such fact suggests that other PAHs, besides those analysed, may have been responsible for the mutagenicity of these samples. The most potent mutagens corresponded to samples with moderate emission factors of PM2.5, PAHs, 13 carcinogenic/mutagenic PAHs, fluoranthene and BaP. Those samples had strong direct-acting mutagenicity, which is contrast to the reported indirect mutagenic activities of the 13 carcinogenic/mutagenic PAHs (IARC, 1998). As mentioned above, this suggests again a possibility that the PAH mixture contained other mutagens than 13 identified PAHs, which acting by means of direct way. Besides PAHs, Cl-PAHs and nitro-PAH have also been detected in wood smoke samples, although at much lower levels than the parent compounds, requiring clean-up and analytical methodologies more sophisticated than the conventional techniques (Feilberg, 2000; Kamens et al., 1985; Lee et al., 2005; Ohura et al., 2008). Higher mutagenic activity of these compounds than the parent PAHs has also been verified (e.g. Colmsjöm et al., 1984). Thus, it is possible that the GCeMS method used in our study was not able to detect other aromatic compounds likely present in the organic extract. It should be noted, however, that Cl-PAH are highly susceptible to photodegradation in the atmosphere (Ohura et al., 2008) and that the majority of nitro-PAHs are thought to be formed in the atmosphere after emission from the gas phase reaction of the parent PAHs with OH and NO3 radicals or with O3, leading to more polar compounds, such as nitro-lactones (Atkinson and Arey, 1994). Therefore, the mutagenic potential of these compounds changes once emitted into the atmosphere (Atkinson and Arey, 1994). It has been observed that the PAH fraction generally make up less than 1% of the extracted smoke sample mass, whereas the greatest mass fraction is composed of aliphatics, acids and sugars (Kamens et al., 1985; Gonçalves et al., 2011). In spite of being a minor mass fraction, PAHs contribute to the majority of the total indirect-acting S. typhimurium mutagenicity, while the dominant compounds are not mutagenic (Kamens et al., 1985). As far as we know, no study has investigated the mutagenic activity of PAHs obtained from domestic combustion of biofuel in Mediterranean countries. Many researchers have carried out Ames test for the obtained particles or the extractable organic matter from particles and have attempted to link their mutagenicity and the PAH content (Asita et al., 1991; Mukherji et al., 2002; Oanh et al., 2002). The mass of particles, organic extracts, and PAH content tested in those studies was much higher than in this study. Asita et al. (1991) tested the mutagenicity of smoke condensates obtained from burning of five wood species. Mass of smoke condensates in each Salmonella test was from 0 to 3000 mg/plate, many of which were mutagen in the absence and presence of metabolic agent. TA98 strain was more sensitive to the condensates than the TA100. They attempted to determine the relative contribution of PAHs to the total mutagenicity of the smoke condensates by means of Ames assay with individual PAH, including BaP and pyrene, at dose of 250e4000 ng/plate; however, no mutation was induced. This concentration is considerably higher than PAHs tested in this study (maximum 1124 ng/plate). This suggests the higher mutagenicity of samples tested in this study, even though the mutagenicity of the mixture can be different from the sum of individual test. Mukherji et al. (2002) tested the mutagenic activity of organic extracts of PM2.5 particles emitted by burning of wood, dung cake and biofuel briquettes by a number of traditional

woodstoves. The test doses were 2, 3, and 5 mg extract/plate. The largest contributor to the total mutagenicity of mixture was directacting component (more than 70%). Oanh et al. (2002) tested the mutagenicity of organic extracts of particulate matter, with the mass of PAH in range of 28.7e6160 mg, released from combustion of wood (Pterocarpus indicus), saw dust, and kerosene by different burning appliances. The organic extract from burning of wood by Thailand ceramic woodstove was the most mutagenic one. This extract presented mutagenicity in the absence, as well as in the presence, of S9 mix.

5. Conclusion According to this study, emissions from the woodstove at higher temperatures included a higher PAH content than those of the fireplace, whereas comparable average total PAH emissions were obtained in both combustion appliances during the cold start combustion. A direct mutagenic response (assays without S9) was recorded for different species (except for golden wattle, maritime pine, and briquettes), different burning appliances and burning start conditions. The strong mutagenic extracts were not correlated with higher emission factors of carcinogenic PAHs, suggesting that other PAHs (like Cl-PAHs) not analysed, or synergistic effects between specific PAHS, may have been responsible for their mutagenicity. The Ames test is a useful tool for a first screening of the potential mutagenicity of wood smoke extracts, however when inferences about impacts on human health are to be made, the information provided by this assay should be complemented with other assays, as this one does not mimic the bioavailability of the mutagenic components in the lungs environment. No mutagenic responses have been detected for all the extracts, in the assays with S9 liver fraction. This fraction provides a set of liver enzymes that are common in vertebrates. They can metabolize certain organic compounds to potentially more genotoxic molecules. Therefore the addition of S9 in Ames assay allows extrapolation of the results for vertebrates. Despite no indirect-acting mutagenicity has been recorded, for all the extracts, we are still concerned with consequences on human health. Continuous exposures to these emissions, at which cumulative doses of PM2.5 are expected, especially during the winter, justifies further evaluations.

Acknowledgements This work was funded by the Portuguese Science Foundation (FCT) through the project “Contribution of biomass combustion to air pollutant emission”, PTDC/AMB/65706/2006 (BIOEMI) and by FSE and POPH funds (Programa Ciência 2007). B.N. Vu acknowledges the master grant from European Commission. Special thanks go to Verónica Inês Nogueira for the laboratory assistance. C. Gonçalves thanks the PhD grant SFRH/BD/36540/2007 from FCT.

References Alén, R., Kuoppala, E., Oesch, P., 1996. Formation of the main degradation compound groups and its components during pyrolysis. Journal of Analytical and Applied Pyrolysis 36, 137e148. Alves, C.A., Vicente, A., Monteiro, C., Gonçalves, C., Evtyugina, M., Pio, C., 2011. Emission of trace gases and organic components in smoke particles from a wildfire in a mixed e evergreen forest in Portugal. Science of the Total Environment 409, 1466e1475. Asita, A.O., Matsui, M., Nohmi, T., Matsuoka, A., Hayashi, M., Ishidate Jr., M., Sofuni, T., Koyano, M., Matsushita, H., 1991. Mutagenicity of wood smoke condensates in the Salmonella/microsome assay. Mutation Research 264, 7e14. Atkinson, R., Arey, J., 1994. Atmospheric chemistry of gas e phase polycyclic aromatic hydrocarbons: formation of atmospheric mutagens. Environmental Health Perspectives 102 (Suppl. 4), 117e126.

B. Vu et al. / Environmental Pollution 166 (2012) 172e181 Aunan, K., Berntsen, T.K., Myhre, G., Rypdal, K., Streets, D.G., Jung-Hun, Woo, Smith, K.R., 2009. Radiative forcing from household fuel burning in Asia. Atmospheric Environment 43, 5674e5681. Baek, S.O., Field, R.A., Goldstone, M.E., Kirk, P.W., Lester, J.N., Perry, R., 1991. A review of atmospheric polycyclic aromatic hydrocarbons: sources, fate and behaviour. Water Air and Soil Pollution 60, 279e300. Bari, M.A., Baumbach, G., Kuch, B., Scheffknecht, G., 2010. Particle-phase concentrations of polycyclic aromatic hydrocarbons in ambient air of rural residential areas in southern Germany. Air Quality and Atmospheric Health 3, 103e116. Bell, D., Kamens, R., 1990. Evaluation of the mutagenicity of combustion particles from several common biomass fuels in the Ames/Salmonella microsome test. Mutation Research Letters 254, 177e183. Bølling, A.K., Pagels, J., Yttri, K.E., Barregard, L., Sallsten, G., Schwarze, P.E., Boman, C., 2009. Health effects of residential wood smoke particles: the importance of combustion conditions and physicochemical particle properties. Particle and Fibre Toxicology 6, 29. doi:10.1186/1743-8977-6-29. Boman, B.C., Forsberg, A.B., Järvholm, B.G., 2003. Adverse health effects from ambient air pollution in relation to residential wood combustion in modern society. Scandinavian Journal of Work, Environment and Health 29, 251e260. Borrego, C., Valente, J., Carvalho, A., Sá, E., Lopes, M., Miranda, A.I., 2010. Contribution of residential wood combustion to PM10 levels in Portugal. Atmospheric Environment 44, 642e651. Calvo, A.I., Tarelho, L.A.C., Duarte, M., Evtyugina, M., Alves, C.A., 2011. Operating conditions of a Portuguese woodstove. In: Proceedings of the 4th International Congress of Energy and Environment Engineering and Management (CIIEM), Merida, Spain. Caseiro, A., Bauer, H., Schmidl, C., Pio, C.A., Puxbaum, H., 2009. Wood burning impact on PM10 in three Austrian regions. Atmospheric Environment 43, 2186e2195. Claxton, L.D., Matthews, P.P., Warren, S.H., 2004. The genotoxicity of ambient outdoor air, a review: Salmonella mutagenicity. Mutation Research 567, 347e399. Claxton, L.D., Woodall Jr., G.M., 2007. A review of the mutagenicity and rodent carcinogenicity of ambient air. Mutation Research 636, 36e94. Colmsjöm, A., Rannug, A., Rannug, U., 1984. Some chloroderivatives of polynuclear aromatic hydrocarbons are potent mutagens in Salmonella typhimurium. Mutation Research 135, 21e29. Danny, L.D.H., Kaufmann, R.K., 2002. Simultaneously constraining climate sensitivity and aerosol radiative forcing. Journal of Climate 15, 2837e2861. Feilberg, A., 2000. Atmospheric Chemistry of Polycyclic Aromatic Compounds with Special Emphasis on Nitro Derivatives. Risø-R-1183(EN). Risø National Laboratory, Roskilde, Denmark. Fernandes, A.P.S., 2009. Emission of PM2.5 and gases from domestic appliances burning. Master’s thesis. University of Aveiro (in Portuguese: http://biblioteca. sinbad.ua.pt/Teses/2010000392). Fine, P.M., Cass, G.R., Simoneit, B.R.T., 2001. Chemical characterization of fine particle emissions from fireplace combustion of woods grown in the northeastern United States. Environmental Science and Technology 35, 2665e2675. Fine, P.M., Cass, G.R., Simoneit, B.R.T., 2002. Chemical characterization of fine particle emissions from fireplace combustion of woods grown in the southern United States. Environmental Science and Technology 36, pp. 1442e1551. Fine, P.M., Cass, G.R., Simoneit, B.R.T., 2004a. Chemical characterization of fine particle emissions from fireplace combustion of woods grown in the Midwestern and western United States. Environmental Engineering Science 21, 387e409. Fine, P.M., Cass, G.R., Simoneit, B.R.T., 2004b. Chemical characterization of fine particle emissions from the woodstove combustion of prevalent United States tree species. Environmental Engineering Science 21, 705e721. Gaeggeler, K., Prevot, A.S.H., Dommen, J., Legreid, G., Reimann, S., Baltensperger, U., 2008. Residential wood burning in an Alpine valley as a source for oxygenated volatile organic compounds, hydrocarbons and organic acids. Atmospheric Environment 42 (35), 8278e8287. Gelencsér, A., May, B., Simpson, D., Sanchez-Ochoa, A., Kasper-Giebl, A., Puxbaum, H., Caseiro, A., Pio, C., Legrand, M., 2007. Source apportionment of PM2.5 organic aerosol over Europe: primary/secondary, natural/anthropogenic, fossil/biogenic origin. Journal of Geophysical Research 112 (D23), D23S04. Gonçalves, C., Alves, C., Evtyugina, M., Mirante, F., Pio, C., Caseiro, A., Schmidl, C., Bauer, H., Carvalho, F., 2010. Characterisation of PM10 emissions from woodstove combustion of common woods grown in Portugal. Atmospheric Environment 44, 4474e4480. Gonçalves, C., Alves, C., Fernandes, A.P., Monteiro, C., Tarelho, L., Evtyugina, M., Pio, C., 2011. Organic compounds in PM2.5 emitted from fireplace and woodstove combustion of typical Portuguese wood species. Atmospheric Environment 45, 4533e4545. Gustafson, P., Ostman, C., Sallsten, G., 2008. Indoor levels of polycyclic aromatic hydrocarbons in homes with or without wood burning for heating. Environmental Science and Technology 42, 5074e5080. Holden, A.S., Sullivan, A.P., Munchak, L.A., Kreidenweis, S.M., Schichtel, B.A., Malm, W.C., Collett Jr., J.L., 2011. Determining contributions of biomass burning and other sources to fine particle contemporary carbon in the western United States. Atmospheric Environment 45, 1986e1993.

181

IARC, 1998. Polynuclear aromatic compounds. Part 1, chemical, environmental and experimental data. In: IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemical to Humans, vol. 32. International Agency for Research on Cancer, Lyon, p. 453. Johansson, L.S., Leckner, B., Gustavsson, L., Cooper, D., Tullin, C., Potter, A., 2004. Emission characteristics of modern and old-type residential boilers fired with wood logs and wood pellets. Atmospheric Environment 38, 4183e4195. Kjällstrand, J., Petersson, G., 2001. Phenols and aromatic hydrocarbons in chimney emissions from traditional and modern residential wood burning. Environmental Technology 22, 391e395. Kamens, R., Bell, D., Dietrich, A., Perry, J., Goodman, R., Claxton, L., Tejada, S., 1985. Mutagenic transformations of dilute wood smoke systems in the presence of ozone and nitrogen dioxide. Analysis of selected high-pressure liquid chromatography fractions from wood smoke particle extracts. Environmental Science and Technology 19, 63e69. Lee, R.G.M., Coleman, P., Jones, J.L., Jones, K.C., Ohmann, R., 2005. Emission factors and importance of PCDD/Fs, PCBs, PCNs, PAHs and PM10 from the domestic burning of coal and wood in the U.K. Environmental Science and Technology 39, 1436e1447. Lipsky, A.M., Robinson, A.L., 2006. Effects of dilution on fine particle mass and partitioning of semivolatile organics in diesel exhaust and wood smoke. Environmental Science and Technology 40, 155e162. Lu, H., Zhu, L., Zhu, N., 2009. Polycyclic aromatic hydrocarbon emission from straw burning and the influence of combustion parameters. Atmospheric Environment 43 (4), 978e983. Maron, D.M., Ames, B.N., 1983. Revised methods for the Salmonella mutagenicity test. Mutation Research 113, 173e215. McCrillis, R.C., Watts, R.R., Warren, S.H., 1992. Effects of operating variables on PAH emissions and mutagenicity of emissions from woodstoves. Journal of the Air and Waste Management Association 42, 691e694. Mortelmans, K., Zeiger, E., 2000. The Ames Salmonella/microsome mutagenicity assay. Mutation Research 455, 29e60. Mukherji, S., Swain, A.K., Venkataraman, C., 2002. Comparative mutagenicity assessment of aerosols in emissions from biofuel combustion. Atmospheric Environment 36, 5627e5635. Oanh, N.T.K., Nghiem, L.H., Phyu, Y.L., 2002. Emission of polycyclic aromatic hydrocarbons, toxicity, and mutagenicity from domestic cooking using sawdust briquettes, wood, and kerosene. Environmental Science and Technology 36, 833e839. OECD, 1997. Test Guideline 471. Bacteria reverse mutation. In: OCED Guidelines for the Testing of Chemicals. Organisation for Economic Co-operation and Development, Paris, France. Ohura, T., Fujima, S., Amagai, T., Shiamagai, M., 2008. Chlorinated polycyclic aromatic hydrocarbons in the atmosphere: seasonal levels, gas-particle partitioning, and origin. Environmental Science and Technology 42, 3296e3302. Pettersson, E., Boman, C., Westerholm, R., Boström, D., Nordin, A., 2011. Stove performance and emission characteristics in residential wood log and pellet combustion, Part 2: wood stove. Energy Fuels 25, 315e323. Portuguese Forest Inventory, 2005. Direcção Geral de Florestas, Lisbon. Puxbaum, H., Caseiro, A., Sanchez-Ochoa, A., Kasper-Giebl, A., Clayes, M., Gelencser, A., Legrand, M., Preunkert, S., Pio, C., 2007. Levoglucosan levels at background sites in Europe for assessing the impact of biomass combustion on the European aerosol background. Journal of Geophysical Research 112 (D23). D23S05/1 e D23S05/11. Quinn, G.P., Keough, M.J., 2002. Experimental Design and Data Analysis for Biologists. Cambridge University Press, Cambridge, UK. Ramdahl, T., 1983. Retene e a molecular marker of wood combustion in ambient air. Nature 306, 580e582. Schmidl, C., Marr, I.L., Caseiro, A., Kotianova, P., Berner, A., Bauer, H., Kasper-Giebl, A., Puxbaum, H., 2008. Chemical characterisation of fine particle emissions from woodstove combustion of common woods growing in mid e European Alpine regions. Atmospheric Environment 42, 126e141. Tarelho L.A.C., Gordaliza A.I.C., Neves D.S.F., Alves C.A., Matos M.A.A., 2011. Characteristics of wood combustion in a Portuguese fireplace and stove. In: Proceedings of the European Biomass Conference and Exhibition, 6e10 June, Berlin, Germany. Telmo, C., Lousada, J., 2011. The explained variation by lignin and extractive contents on higher heating value of wood. Biomass and Bioenergy 35, 1663e1667. van Houdt, J.J., Daenen, C.M.J., Boleij, J.S.M., Alink, G.M., 1986. Contribution of woodstoves and fire places to mutagenic activity of airborne particulate matter inside home. Mutation Research 171, 91e98. Wang, X.Y., Li, Q.B., Luo, Y.M., Lian, Q.D., Xi, M., Ma, J.M., Li, Y., Liu, Y.P., Cheng, C.L., 2010. Characteristics and sources of atmospheric polycyclic aromatic hydrocarbons (PAHs) in Shanghai, China. Environmental Monitoring and Assessment 165, 295e305. Zielinska, B., Samburova, V., 2011. Residential and non-residential biomass combustion: impact on air quality. Encyclopedia of Environmental Health, 819e827.