identification and emission rates of molecular tracers in

Identification and emission rates of molecular tracers in coal smokeparticulate matter

D.R. Orosa,b, B.R.T. Simoneitb,*aEnvironmental Sciences Graduate Program, Cordley Hall, Oregon State University, Corvallis, OR 97331, USA

bEnvironmental and Petroleum Geochemistry Group, College of Oceanic and Atmospheric Sciences, Oregon State University, Corvallis, OR 97331, USA

Received 5 November 1998; received in revised form 9 July 1999

Abstract

The abundances and distributions of organic constituents in coal smoke particulate matter are dependent on thermal combustion tempera-ture, ventilation, burn time, and coal rank (geologic maturity). Important coal rank indicators from smoke include (1) the decreases in CPIs ofn-alkanoic acids, UCM and phenolic compounds with increasing rank, and (2) the increase in the homohopane index�S=�S1 R�� withincreasing rank. Coal smoke emissions may be identified in atmospheric samples by (1) the unresolved to resolved component ratios�U=R�;(2) the distributions and abundances of aromatic molecular markers, specifically picene, alkylated picenes and alkylhydropicenes, (3) the17a(H),21b(H)-hopane to 22R-17a(H),21b(H)-homohopane ratio (range 0.05–0.35), and (4) the presence of other source-specific molecularmarkers.q 2000 Elsevier Science Ltd. All rights reserved.

Keywords: Lignite; Brown coal; Sub-bituminous coal; Bituminous coal; Molecular biomarkers; Hydrocarbons; Burning

1. Introduction

The burning of coal as a fuel significantly increases theinput of particulate organic aerosol components to the atmo-sphere. This source of smoke emissions, as other particleemissions, influences atmospheric chemical, optical andradiative properties through direct (adsorption and scatter-ing of solar and terrestrial radiation) and indirect (modifica-tion of cloud processes) mechanisms [1,2]. The aim of thisstudy is to report the chemical composition of smoke parti-culate matter emitted by flaming and smoldering combus-tion of four coals ranging in maturity from lignite tobituminous rank. This information is important for under-standing the organic component contribution of coal burn-ing emissions to atmospheric chemistry. It complementsexisting work on the characterization of organic emissionsfrom biomass burning [3–10] and from fossil fuels [11,12],and in atmospheric source apportionment studies [13–15].The identification of combustion products may be used as achemical fingerprint for identifying source inputs, transportmechanisms and receptor fate in samples of atmosphericfine particulate matter. In this case, the directly emittedand thermally altered molecular biomarkers may be usedas specific tracers for tracking coal smoke emissions inthe environment.

1.1. Coal composition

During the coalification process, plant remains undergo asequence of physical, biochemical and chemical changes(diagenesis, then catagenesis) which results in a series ofcoals of increasing rank or maturity [16–18]. The coal rankswith the progression of maturity are peat! lignite! sub-bituminous! bituminous! anthracite [19]. The sequenceof chemical reactions associated with the coalificationprocess is (1) dehydration, (2) loss of oxygen containingfunctional groups, (3) alkylation and (4) oligomerization[20]. The degree of completion of chemical reaction withinthe depositional environment determines the structure ofcoals. The general structural characteristics of coal havebeen described [17,19] and include a series of fused benzenerings (aromatics, Ar) with associated functional groups,linked together by ether linkages, sulfur bridges, or by–CH2– and –CH2–CH2– bridges. As maturity increases, thedegree of aromaticity increases and the number of linkagesor bridges decreases. The more reactive oxygen bearingfunctional groups (i.e. –OCH3, –COOH, –OH) alsodecrease in relative abundance, whereas the least reactivefunctional group (i.e. CyO) increases. Nitrogen may also bepresent in reactive forms such as amines, particularly inyounger coals such as lignites [19,21]. As coal maturityincreases, nitrogen forms change into more condensedstructures (i.e. pyridines, quinolines, pyrroles and

Fuel 79 (2000) 515–536

0016-2361/00/$ - see front matterq 2000 Elsevier Science Ltd. All rights reserved.PII: S0016-2361(99)00153-2

www.elsevier.com/locate/fuel

* Corresponding author. Tel.:11-541-737-2155; fax:11-541-737-2064.

carbazoles ([22]). Sulfur occurs in coals as sulfide, disulfideor mercaptan in aliphatic or aromatic structures [23].

1.2. Coal combustion

The varying temperature conditions during burning deter-mine the molecular alteration and transformation of theorganic compounds emitted from coal. The heat intensityand the duration of smoldering and flaming conditionsdetermine the distributions and ratios of the natural versusaltered compounds present in coal smoke. Under flamingconditions (temperature.3008C), the weaker bridges inthe coal structure, particularly the –CH2–CH2– bonds andthe –CH2–O–CH2– ether linkages, are broken first, butnot the Ar–CH2–Ar bridges or the Ar–O–Ar linkages. Thisis immediately followed by the loss of functional groups,which then result in the formation of noncondensible gases(i.e. CO2, H2O, CH4) [19,24]. The primary chemical reac-tions that occur under flaming conditions include pyrolysis,bond cleavage, fission, and tarry and volatile product forma-tion. It is generally accepted that free radical reactionscontrol the pyrolysis chemistry of most organic substances.A detailed description of the free radical reactions occurringin coal pyrolytic processes has been reported [25].

Under smoldering conditions (temperature,3008C),organic compounds and their altered products are releasedby a volatilization/steam stripping effect. The extent of thisprocess is dependent on the coal moisture content, which ishigher in less mature coals. The primary chemical reactionsthat occur under smoldering conditions include depolymer-ization, water elimination, fragmentation, oxidation, charformation and volatilization/steam stripping of moleculartracers.

Some important physical and chemical properties whichgovern the behavior of coal during combustion processesinclude moisture content, specific gravity, reactivity, struc-ture and aromaticity, thermal conductivity and heat capacity[19]. These properties coupled with mass transfer effects(resistances to molecular diffusion or convective transportof reactants and products due to changes in particle size andgas pressure) may result in variations in product yields anddistributions [26].

1.3. Solvent soluble components of coal

Various organic constituents isolated from extracts (bitu-men) of different rank coals have been identified andinclude: homologous series ofn-alkanes,n-alkanoic acids,a,v-alkanedioic acids,v-hydroxyalkanoic acids,n-alkanolsand alkan-2-ones in lignite [27,28]; with additional alkylcy-clohexanes and tetracyclic diterpanes (phyllocladanes) fromvarious German coals [29,30]; diterpenoid acids from coni-fer resin in brown coal [31]; isoprenoids and hopanoic acidsfrom asphaltenes and resins of bituminous coal [32]; azaar-enes, polycyclic aromatics, picenes and hydropicenes fromWashington bituminous and anthracite coals [33]. Numer-ous other coal extracts and laboratory pyrolysis studies have

been reported. These biomarker molecules and their thermalalteration products are released in coal smoke emissions.

Coal smoke and other source emissions (e.g. petroleum)introduce airborne fine particulate matter containing organicconstituents (e.g. polyaromatic hydrocarbon (PAH) andoxy-PAH) which have mutagenic and genotoxic potential[34,35]. Considering that coal is still collected and used as asolid fuel source for indoor heating of homes or cooking(e.g. Hopi Indian Reservation, New Mexico, USA; EasternEurope; Asia) it is also necessary to identify the componentsof smoke emissions in order to make indoor air qualityassessments and to determine human exposure levels toparticle-bound organic compounds.

2. Materials and methods

2.1. Sampling

Four coal samples ranging in rank from lignite to bitumi-nous coal were collected and include lignite (Fortuna Mine;Aachen, Germany); brown coal (Leuna, commercialbriquette, Thu¨ringen, Germany); sub-bituminous (KayentaWepo Formation, Black Mesa, Arizona); and bituminouscoal (Wales, Great Britain). Using a controlled fire, coalsamples were burned under both flaming and smolderingconditions. The emitted smoke was collected on an organi-cally clean quartz fiber filter (annealed at 5508C for 3 h;100% particle size retention.2.0mm) using a high volumeair sampler located approximately 1.5 m diagonally aboveand to the side of the flames in the smoke plume. Smoke wassampled for approximately 10-min periods at a suction flowrate of 40 ft3/min (1.13 m3/min). After sampling, a portionof each filter (8.8 cm2) was cut out and set aside for volatileorganic carbon (VOC) and elemental carbon (EC) analysis[36,37]. The collection filters were then placed inprecleaned 300-ml jars with Teflon lined lids to which10 ml of chloroform was added. The jars were stored at48C until chemical extraction was conducted.

2.2. Extraction and fractionation

Each filter was extracted using ultrasonic agitation forthree 20-min periods using 200 ml of dichloromethane(CH2Cl2, nannograde, glass distilled). The solvent extractwas filtered using a Gelman Swinney filtration unit contain-ing an annealed glass fiber filter for the removal of insolubleparticles [38]. The filtrate was first concentrated by use of arotary evaporator and then a stream of filtered nitrogen gas.The final volume was adjusted to exactly 4.0 ml by additionof CH2Cl2. Aliquots were taken for derivatization. Alkanoicacid and phenolic moieties in the extracts were methylatedusing diazomethane in diethyl ether prepared from theprecursor N-methyl-N0-nitro-N-nitrosoguanidine (PierceChemical Co.) [39].

The methylated extracts were separated by preparativethin layer chromatography (TLC) on silica gel plates

D.R. Oros, B.R.T. Simoneit / Fuel 79 (2000) 515–536516

(Analtech, Inc.) with a mobile phase eluent mixture of hexa-ne:diethyl ether (9:1). This procedure allows the determina-tion of chemical information on single molecular groups orfunctional group series, which may not be detected due tocoelution in the total extract mixture. The four fractionsremoved from the TLC plates contained the followingclasses of compounds: (1)n-alkanes and saturated and unsa-turated cyclic di- and triterpenoid hydrocarbons, (2)n-alka-nones,n-alkanals and polycyclic aromatic hydrocarbons, (3)n-alkanoic acids (as methyl esters) and saturated and unsa-turated di- and triterpenoid ketones, and (4)n-alkanols,terpenols and polar organics. The fourth fraction and atotal extract aliquot were converted before analysis totrimethylsilyl derivatives by reaction withN,O-bis-(trimethylsilyl)trifluoroacetamide (BSTFA) plus 1%trimethylchlorosilane for approximately 3 h at 708C. Aschematic of the sample treatment and separation procedureis given in Fig. 1. The separation procedure follows themethod first used by [38] and includes some modifications.

2.3. Instrumental analyses

The total extracts and separated fractions were analyzed

by capillary gas chromatography (GC, Hewlett–PackardModel 5890A) with a 25 m× 0:20 mm i.d. fused silicacapillary column coated with DB-5 (J&W Scientific, filmthickness 0.25mm) which was temperature programmed asfollows: hold at 658C for 2 min, ramp to 3008C at 68C/min,hold isothermal at 3008C for 20 min. All samples wereanalyzed by capillary gas chromatography–mass spectro-metry (GC–MS) using a Hewlett–Packard Model 5973MSD quadrupole mass spectrometer operated in the elec-tron impact mode at 70 eV and coupled to a Hewlett–Pack-ard Model 6890 gas chromatograph. The GC was equippedwith a 30 m× 0:25 mm i.d. capillary column coated withDB-5 (J&W Scientific, film thickness 0.25mm) andoperated using the same temperature program describedearlier with helium as carrier gas.

2.4. Compound identification and quantitation

Compound identifications are based on comparisons withauthentic standards, GC retention time, literature mass spec-tra and interpretation of mass spectrometric fragmentationpatterns. Quantitation of the compounds was conducted bycomparison of GC peak area with that of a coinjected known

D.R. Oros, B.R.T. Simoneit / Fuel 79 (2000) 515–536 517

Fig. 1. Schematic of the sample treatment, extraction and separation procedure.


Table 1Emission rates (mg/kg of coal burned) and percent abundances (relative to major compound) of the predominant organic constituents in coal smoke samples(identification criteria: A� matches with authentic standard; I� interpreted from mass spectrum fragmentation pattern; S� interpolated from homologousseries fragmentation pattern; nd� not determined.)

Compound name Composition MW Coal smoke

Lignite Brown Sub-bituminous Bituminous ID*basis

(mg/kg) (rel. %) (mg/kg) (rel. %) (mg/kg) (rel. %) (mg/kg) (rel. %)

I. Homologous seriesn-Alkanesn-Tetradecane C14H30 198 0 0 4 0.4 0 0 0 0 An-Pentadecane C15H32 212 0 0 11 1 0 0 27 3.2 An-Hexadecane C16H34 226 42 0.8 17 1.5 27 1.5 43 5.1 An-Heptadecane C17H36 240 79 1.5 25 2.2 96 5.4 58 6.9 An-Octadecane C18H38 254 148 2.8 31 .28 303 17.1 64 7.6 An-Nonadecane C19H40 268 227 4.3 37 3.4 467 26.3 94 11.1 An-Eicosane C20H42 282 308 5.8 43 3.9 617 34.8 132 15.7 An-Heneicosane C21H44 296 364 6.9 60 5.4 658 37.1 151 17.9 An-Docosane C22H46 310 453 8.6 74 6.6 770 43.4 154 18.2 An-Tricosane C23H48 324 571 10.8 93 8.4 905 51 162 19.3 An-Tetracosane C24H50 338 591 11.2 86 7.8 1217 68.6 171 20.3 An-Pentacosane C25H52 352 610 11.6 98 8.8 1340 75.6 140 16.6 An-Hexacosane C26H54 366 531 10.1 89 8 1294 73 120 14.3 An-Heptacosane C27H56 380 587 11.1 117 10.6 1386 78.2 127 15.1 An-Octacosane C28H58 394 576 10.9 84 7.6 1159 65.4 92 10.9 An-Nonacosane C29H60 408 710 13.5 152 13.7 1035 58.4 67 8 An-Tricontane C30H62 422 382 7.2 165 14.9 872 49.2 47 5.6 An-Hentriacontane C31H64 436 669 12.7 97 8.8 636 35.9 41 4.9 An-Dotriacontane C32H66 450 164 3.1 57 5.1 572 32.3 42 5 An-Tritriacontane C33H68 464 297 5.6 19 1.7 162 9.1 0 0 An-Tetratriacontane C34H70 478 113 2.1 0 0 0 0 0 0 ACPI 1.2 1.1 1 1Cmax 29 30 27 24n-Alkenesn-Tetradecene C14H28 196 0 0 4 0.3 0 0 13 1.6 Sn-Pentadecane C15H30 210 0 0 10 0.9 0 0 21 2.5 Sn-Hexadecene C16H32 224 133 2.5 16 1.4 0 0 51 6 Sn-Heptadecene C17H34 238 99 1.9 20 1.8 87 4.9 52 6.1 Sn-Octadecene C18H36 252 208 3.9 23 2 149 8.4 51 6 An-Nonadecene C19H38 266 233 4.4 26 2.3 298 16.8 69 8.1 Sn-Eicosene C20H40 280 340 6.4 43 3.9 291 16.4 90 10.7 An-Heneicosene C21H42 294 474 9 40 3.6 370 20.9 88 10.4 Sn-Docosene C22H44 308 501 9.5 59 5.4 520 29.3 58 6.9 Sn-Tricosene C23H46 322 760 14.4 47 4.3 382 21.6 52 6.2 Sn-Tetracosene C24H48 336 597 11.3 98 8.9 0 0 46 5.4 Sn-Pentacosene C25H50 350 578 11 49 4.4 0 0 40 4.8 Sn-Hexacosene C26H52 364 833 15.8 1107 100 0 0 31 3.7 Sn-Heptacosene C27H54 378 1010 19.2 0 0 0 0 0 0 Sn-Octacosene C28H56 392 1049 19.9 0 0 0 0 0 0 SCPI 0.9 0.1 1.2 0.9Cmax 28:1 26:1 22:1 20:1n-Alkan-2-onesn-Tridecan-2-one C13H26O 198 0 0 6 0.6 8 0.5 33 3.9 Sn-Tetradecan-2-one C14H28O 212 0 0 7 0.6 26 1.4 44 5.2 Sn-Pentadecan-2-one C15H30O 226 33 0.6 10 0.9 25 1.4 46 5.4 Sn-Hexadecan-2-one C16H32O 240 94 1.8 12 1 26 1.5 41 4.9 Sn-Heptadecan-2-one C17H34O 254 31 0.6 19 1.7 59 3.3 50 5.9 Sn-Octadecan-2-one C18H36O 268 46 0.9 15 1.3 56 3.2 133 15.8 Sn-Nonadecan-2-one C19H38O 282 25 0.5 23 2 46 2.6 68 8.1 Sn-Eicosan-2-one C20H40O 296 65 1.2 23 2.1 41 2.3 109 12.9 Sn-Heneicosan-2-one C21H42O 310 50 1 33 3 48 2.7 68 8 Sn-Docosan-2-one C22H44O 324 30 0.6 29 2.6 30 1.7 110 13 Sn-Tricosan-2-one C23H46O 338 33 0.6 63 5.7 108 6.1 136 16.1 Sn-Tetracosan-2-one C24H48O 352 39 0.7 44 3.9 49 2.7 74 8.8 S


Table 1 (continued)




n-Pentacosan-2-one C25H50O 366 86 1.6 101 9.2 79 4.5 29 3.5 Sn-Hexacosan-2-one C26H52O 380 61 1.2 63 5.7 48 2.7 68 8.1 Sn-Heptacosan-2-one C27H54O 394 190 3.6 104 9.4 112 6.3 114 13.5 Sn-Octacosan-2-one C28H56O 408 139 2.6 79 7.1 73 4.1 52 6.1 Sn-Nonacosan-2-one C29H58O 422 230 4.4 150 13.6 157 8.9 45 5.3 Sn-Triacontan-2-one C30H60O 436 149 2.8 56 5 87 4.9 21 2.5 Sn-Hentriacontan-2-one C31H62O 450 210 4 94 8.5 89 5 20 2.3 Sn-Dotriacontan-2-one C32H64O 464 93 1.8 16 1.4 9 0.5 0 0 Sn-Tritriacontan-2-one C33H66O 478 0 0 52 4.7 9 0.5 0 0 S6,8,10-Trimethylpentadecan-2-one C18H36O 268 0 0 24 2.2 0 0 0 0 SCPI 1.2 1.9 1.7 0.9Cmax 29 29 29 23n-Alkanalsn-Tridecanal C13H26O 198 0 0 8 0.7 0 0 0 0 Sn-Tetradecanal C14H28O 212 17 0.3 13 1.2 0 0 0 0 Sn-Pentadecanal C15H30O 226 33 0.6 15 1.4 0 0 0 0 Sn-Hexadecanal C16H32O 240 27 0.5 16 1.4 0 0 0 0 Sn-Heptadecanal C17H34O 254 15 0.3 19 1.7 0 0 0 0 Sn-Octadecanal C18H36O 268 24 0.5 21 1.9 0 0 0 0 Sn-Nonadecanal C19H38O 282 36 0.7 20 1.8 0 0 0 0 Sn-Eicosanal C20H40O 269 14 0.3 26 2.4 0 0 0 0 Sn-Heneicosanal C21H42O 310 14 0.3 21 1.9 0 0 0 0 Sn-Docosanal C22H44O 324 14 0.3 40 3.6 0 0 0 0 Sn-Tricosanal C23H46O 338 7 0.1 21 1.9 0 0 0 0 Sn-Tetracosanal C24H48O 352 14 0.3 57 5.1 0 0 0 0 Sn-Pentacosanal C25H50O 366 14 0.3 13 1.2 0 0 0 0 Sn-Hexacosanal C26H52O 380 101 1.9 32 2.9 0 0 0 0 Sn-Heptacosanal C27H54O 394 14 0.3 13 1.2 0 0 0 0 Sn-Octacosanal C28H56O 408 66 1.3 26 2.4 0 0 0 0 Sn-Nonacosanal C29H58O 422 14 0.3 34 3.1 0 0 0 0 Sn-Triacontanal C30H60O 436 127 2.4 11 1 0 0 0 0 SCPI 2.8 1.5 nd ndCmax 30 24 0 0n-Alkanoic acidsn-Octanoic acid C8H16O2 144 0 0 0 0 13 0.7 16 1.9 Sn-Nonanoic acid C9H18O2 158 0 0 0 0 14 0.8 127 15 Sn-Decanoic acid C10H20O2 172 0 0 0 0 36 2 14 1.6 Sn-Undecanoic acid C11H22O2 186 0 0 10 0.9 45 2.5 11 1.4 Sn-Dodecanoic acid C12H24O2 200 48 0.9 21 1.9 75 4.2 39 4.6 Sn-Tridecanoic acid C13H26O2 214 34 0.7 12 1.1 18 1 40 4.7 Sn-Tetradecanoic acid C14H28O2 228 63 1.2 25 2.2 131 7.4 234 27.7 Sn-Pentadecanoic acid C15H30O2 242 79 1.5 20 1.8 126 7.1 131 15.6 Sn-Hexadecanoic acid C16H32O2 256 135 2.6 30 2.7 160 9 107 12.6 An-Heptadecanoic acid C17H34O2 270 24 0.5 27 2.5 91 5.1 216 25.6 Sn-Octadecanoic acid C18H36O2 284 106 2 29 2.6 77 4.3 226 26.8 Sn-Nonadecanoic acid C19H38O2 298 66 1.3 29 2.6 74 4.2 226 26.9 Sn-Eicosanoic acid C20H40O2 312 118 2.2 38 3.5 45 2.6 174 20.6 Sn-Heneicosanoic acid C21H42O2 326 105 2 49 4.4 30 1.7 62 7.4 Sn-Docosanoic acid C22H44O2 340 175 3.3 69 6.2 55 3.1 94 11.2 Sn-Tricosanoic acid C23H46O2 354 114 2.2 34 3 169 9.5 0 0 Sn-Tetracosanoic acid C24H48O2 368 271 5.1 319 28.8 93 5.2 0 0 Sn-Pentacosanoic acid C25H50O2 382 245 4.7 68 6.1 51 2.9 0 0 Sn-Hexacosanoic acid C26H52O2 396 326 6.2 463 41.8 167 9.4 0 0 Sn-Heptacosanoic acid C27H54O2 410 616 11.7 83 7.5 72 4 0 0 Sn-Octacosanoic acid C28H56O2 424 5270 100 147 13.3 386 21.8 0 0 Sn-Nonacosanoic acid C29H58O2 438 1007 19.1 60 5.4 144 8.1 0 0 Sn-Triacontanoic acid C30H60O2 452 3727 70.7 120 10.8 280 15.8 0 0 Sn-Hentriacontanoic acid C31H62O2 466 511 9.7 14 1.2 158 8.9 0 0 Sn-Dotriacontanoic acid C32H64O2 480 1923 36.5 28 2.5 0 0 0 0 S


Table 1 (continued)




n-Tritriacontanoic acid C33H66O2 494 121 2.3 0 0 0 0 0 0 Sn-Tetratriacontanoic acid C34H68O2 508 425 8.1 0 0 0 0 0 0 SCPI 4.3 3.2 1.4 0.9Cmax 28 26 28 14n-Alkenoic acidsn-Decenoic acid C10H18O2 170 0 0 0 0 6 0.3 0 0 Sn-Undecenoic acid C11H20O2 184 0 0 0 0 12 0.7 0 0 Sn-Tridecenoic acid C13H24O2 212 14 0.3 0 0 23 1.3 0 0 Sn-Tetradecenoic acid C14H26O2 226 55 1 0 0 41 2.3 0 0 Sn-Pentadecenoic acid C15H28O2 240 32 0.6 0 0 31 1.8 0 0 Sn-Hexadecenoic acid C16H30O2 254 132 2.5 0 0 61 3.4 0 0 Sn-Heptadecenoic acid C17H32O2 268 32 0.6 0 0 41 2.3 0 0 Sn-Octadecenoic acid C18H34O2 282 34 0.6 0 0 156 8.8 42 5 Sn-Nonadecenoic acid C19H36O2 296 14 0.3 0 0 83 4.7 0 0 Sn-Eicosenoic acid C20H38O2 310 115 2.2 0 0 19 1.1 0 0 Sn-Heneicosenoic acid C21H40O2 324 14 0.3 0 0 17 1 0 0 Sn-Docosenoic acid C22H42O2 338 87 1.7 0 0 15 0.8 0 0 Sn-Tricosenoic acid C23H44O2 352 0 0 0 0 29 1.6 0 0 SCPI 4 nd 1.2 ndCmax 16:1 nd 18:1 18:1n-Alkanolsn-Octadecanol C18H38O 270 537 10.2 0 0 0 0 0 0 Sn-Nonadecanol C19H40O 284 0 0 0 0 0 0 0 0 Sn-Eicosanol C20H42O 298 313 5.9 8 0.7 0 0 0 0 Sn-Heneicosanol C21H44O 312 0 0 0 0 0 0 0 0 Sn-Docosanol C22H46O 326 447 8.5 33 3 0 0 0 0 An-Tricosanol C23H48O 340 173 3.3 4 0.4 0 0 0 0 Sn-Tetracosanol C24H50O 354 609 11.6 102 9.2 0 0 0 0 Sn-Pentacosanol C25H52O 368 402 7.6 7 0.7 0 0 0 0 Sn-Hexacosanol C26H54O 382 754 14.3 83 7.5 0 0 0 0 Sn-Heptacosanol C27H56O 396 514 9.8 7 0.6 0 0 0 0 Sn-Octacosanol C28H58O 410 1276 24.2 72 6.5 0 0 0 0 Sn-Nonacosanol C29H60O 424 426 8.1 2 0.2 0 0 0 0 Sn-Triacontanol C30H62O 438 1363 25.9 28 2.5 0 0 0 0 Sn-Hentriacontanol C31H64O 452 411 7.8 0 0 0 0 0 0 Sn-Dotriacontanol C32H66O 466 960 18.2 6 0.6 0 0 0 0 SCPI 3.2 16.1 nd ndCmax 30 24 nd ndn-Alkylbenzenesn-Heptylbenzene C13H20 176 0 0 0 0 0 0 1 0.1 Sn-Octylbenzene C14H22 190 0 0 0 0 0 0 5 0.6 Sn-Nonylbenzene C15H24 204 0 0 0 0 23 1.3 11 1.4 Sn-Decylbenzene C16H26 218 80 1.5 7 0.6 18 1 19 2.3 Sn-Undecylbenzene C17H28 232 37 0.7 6 0.6 69 3.9 24 2.8 Sn-Dodecylbenzene C18H30 246 120 2.3 16 1.4 165 9.3 22 2.7 Sn-Tridecylbenzene C19H32 260 142 2.7 16 1.4 163 9.2 29 3.4 Sn-Tetradecylbenzene C20H34 274 156 3 15 1.3 166 9.3 30 3.5 Sn-Pentadecylbenzene C21H36 288 129 2.4 6 0.6 163 9.2 26 3 Sn-Hexadecylbenzene C22H38 302 149 2.8 14 1.3 146 8.3 23 2.8 Sn-Heptadecylbenzene C23H40 316 122 2.3 6 0.5 160 9 21 2.5 Sn-Octadecylbenzene C24H42 330 111 2.1 6 0.5 297 16.8 17 2.1 Sn-Nonadecylbenzene C25H44 344 155 2.9 6 0.6 179 10.1 18 2.1 Sn-Eicosylbenzene C26H46 358 172 3.3 15 1.3 234 13.2 16 1.9 Sn-Heneicosylbenzene C27H48 372 241 4.6 6 0.6 152 8.6 12 1.4 Sn-Docosylbenzene C28H50 386 288 5.5 6 0.5 137 7.7 11 1.3 Sn-Tricosylbenzene C29H52 400 298 5.7 17 1.5 114 6.4 5 0.6 Sn-Tetracosylbenzene C30H54 414 443 8.4 17 1.5 156 8.8 3 0.3 Sn-Pentacosylbenzene C31H56 428 263 5 8 0.7 88 5 2 0.2 Sn-Hexacosylbenzene C32H58 442 217 4.1 0 0 0 0 0 0 S


Table 1 (continued)




CPI 1.3 1.3 1.2 1Cmax 30 30 30 20n-Alkylnitrilesn-Tetracosanyl nitrile C24H47N 349 0 0 6 0.6 0 0 0 0 Sn-Pentacosanyl nitrile C25H49N 363 0 0 2 0.2 0 0 0 0 Sn-Hexacosanyl nitrile C26H51N 377 0 0 12 1.1 0 0 0 0 Sn-Heptacosanyl nitrile C27H53N 391 0 0 0 0 0 0 0 0 Sn-Octacosanyl nitrile C28H55N 405 0 0 12 1.1 0 0 0 0 Sn-Nonacosanyl nitrile C29H57N 419 0 0 0 0 0 0 0 0 Sn-Triacontanyl nitrile C30H59N 433 0 0 31 2.8 0 0 0 0 SCPI nd 27.2 nd ndCmax nd 30 nd ndII. BiomarkersHopanoids22,29,30-Trisnorneohop-13(18)-ene C27H44 368 356 6.8 0 0 0 0 0 0 I22,29,30-Trisnorhop-17(21)-ene C27H44 368 325 6.2 12 1.1 150 8.4 17 2 I17a(H)-22,29,30-Trisnorhopane C27H46 370 195 3.7 10 0.9 430 24.3 67 8 I17b(H)-22,29,30-Trisnorhopane C27H46 370 442 8.4 15 1.4 492 27.7 23 2.8 I17a(H),21b(H)-29-Norhopane C29H48 398 56 1.1 19 1.7 656 37 85 10.1 I17b(H),21a(H)-29-Norhopane C29H48 398 566 10.7 7 0.6 729 41.1 43 5.1 I17b(H),21b(H)-29-Norhopane C29H48 398 0 0 6 0.5 236 13.3 0 0 IHop-17(21)-ene C30H48 410 1263 24 10 0.9 0 0 1 0.1 I17a(H),21b(H)-Hopane C30H50 412 97 1.8 10 0.9 335 18.9 45 5.3 I17b(H),21a(H)-Hopane C30H50 412 1111 21.1 6 0.5 473 26.7 42 5 I17b(H),21b(H)-Hopane C30H50 412 0 0 0 0 43 2.4 0 0 I17b(H),21b(H)-Homohopene C31H52 424 49 0.9 0 0 0 0 0 0 I22R-17a(H),21b(H)-Homohopane C31H54 426 840 15.9 12 1 160 9 17 2 I22S-17a(H),21b(H)-Homohopane C31H54 426 49 0.9 1 0.1 40 2.3 10 1.1 I17b(H),21a(H)-Homohopane C31H54 426 195 3.7 5 0.5 170 1.6 19 2.2 I22R-17b(H),21b(H)-Homohopane C31H54 426 253 4.8 3 0.3 129 7.3 0 0 I17a(H),21b(H)-Bishomohop-31-ene C32H54 438 233 4.4 0 0 0 0 0 0 I17b(H),21a(H)-Bishomohop-31-ene C32H54 438 122 2.3 0 0 0 0 0 0 I22R-17b(H),21a(H)-Bishomohop-31-ene

C32H54 438 75 1.4 0 0 0 0 0 0 I

22R-17a(H),21b(H)-Bishomohopane C32H56 440 0 0 2 0.2 111 6.3 16 1.9 I22S-17a(H),21b(H)-Bishomohopane C32H56 440 0 0 0 0 22 1.3 11 1.2 I17b(H),21a(H)-Bishomohopane C32H56 440 147 2.8 2 0.2 84 4.7 0 0 I17b(H),21b(H)-Bishomohopane C32H56 440 62 1.2 0 0 0 0 0 0 I17b(H),21b(H)-Trishomohop-32-ene C33H56 452 77 1.5 0 0 0 0 0 0 I22R-17a(H),21b(H)-Trishomohopane C33H58 454 0 0 0 0 0 0 3 0.3 I22S-17a(H),21b(H)-Trishomohopane C33H58 454 0 0 0 0 0 0 1 0.2 IHomohopane index, C31 [S/(S1 R)] 0.05 0.08 0.20 0.3717a(H),21b(H)-Hopane/22R-17a(H),21b(H)-Homohopane

0.1 0.8 2.1 2.6

Steroids4-Methyl-24-ethyl-19-norcholesta-1,3,5(10)-triene

C29H46 394 58 1.1 0 0 0 0 0 0 I

b-Sitoster-2-ene C29H50 398 230 4.4 0 0 217 12.2 0 0 I5a(H),14a(H)-24-Ethylcholestane C29H52 400 0 0 0 0 0 0 4 0.5 ITerpenoidsRetene C18H18 234 101 1.9 0 0 282 15.9 0 0 IDihydroretene C18H20 236 82 1.6 0 0 0 0 0 0 ISimonellite C19H24 252 91 1.7 0 0 213 12 0 0 IDehydroabietin C19H28 256 57 1.1 0 0 0 0 0 0 I10-Nor-6,12-dioxoabieta-1,5(10),7,9(11),13-pentaene

C19H20O 264 335 6.4 0 0 0 0 0 0 I

12-Hydroxysimonellite C19H24O 268 294 8.6 0 0 0 0 0 0 I19-Nor-7-oxoferruginol C19H24O 268 110 2.1 0 0 0 0 0 0 IDehydroabietane C20H30 270 278 5.3 0 0 0 0 0 0 I


Table 1 (continued)




3-Oxo-12-hydroxysimonellite C19H22O2 282 1132 21.5 0 0 0 0 0 0 IAbieta-8,11,13-trien-7-one C20H28O 284 69 1.3 0 0 0 0 0 0 I6,7-Dehydroferruginol C20H28O 284 1471 27.9 0 0 0 0 0 0 IFerruginol C20H30O 286 422 8 0 0 0 0 0 0 I24,25-Dinorursana-1,3,5(10),12-tetraene

C28H40 376 0 0 0 0 246 13.9 0 0 I

24,25-Dinoroleana-1,3,5(10),12-tetraene

C28H40 376 1577 29.9 0 0 100 5.7 0 0 I

24,25-Dinorlupa-1,3,5(10),22(29)-tetraene

C28H40 376 0 0 0 0 99 5.6 0 0 I

24,25-Dinorlupa-1,3,5(10)-triene C28H42 378 0 0 0 0 99 5.6 0 0 IA-Neoursa-3(5),12-diene C30H48 408 648 12.3 0 0 0 0 0 0 ILupa-2,22-diene C30H48 408 0 0 3 0.3 0 0 0 0 IOleana-2,12-diene C30H48 408 921 17.5 0 0 0 0 0 0 IUrsana-2,12-diene C30H48 408 275 5.2 0 0 0 0 0 0 IOlean-12-ene C30H50 410 1161 22 0 0 0 0 0 0 IAllobetul-2-ene C30H48O 424 0 0 117 10.6 0 0 0 0 Ia-Amyrone C30H48O 424 186 3.5 0 0 0 0 0 0 Ib-Amyrone C30H48O 424 823 15.6 0 0 0 0 0 0 IFriedelin C30H50O 426 3481 66.1 50 4.5 0 0 0 0 IAromatic biomarkers3,3,7-Trimethyl-1,2,3,4-tetrahydrochrysene

C21H22 274 67 1.3 0 0 0 0 0 0 I

3,4,7-Trimethyl-1,2,3,4-tetrahydrochrysene

C21H22 274 0 0 9 0.8 0 0 0 0 I

Picene C22H14 278 0 0 0 0 0 0 94 11.1 IC1-Picenes C23H16 292 0 0 0 0 0 0 41 4.9 IC2-Picenes C24H18 306 0 0 0 0 76 4.3 11 1.3 I2,9-(1,8)-Dimethylpicene C24H18 306 31 0.6 12 1.1 93 3.2 0 0 I1,2-Dimethyl-1,2,3,4-tetrahydropicene C24H22 310 262 5 18 1.6 0 0 0 0 I2,2-Dimethyl-1,2,3,4-tetrahydropicene C24H22 310 0 0 0 0 95 5.2 0 0 I1,2,9-(1,7,8)-Trimethylpicene C25H20 320 102 1.9 4 0.4 189 10.7 0 0 I2-Ethyl-9-methyl-1,2-dihydropicene C25H22 320 306 5.8 0 0 0 0 0 0 I1,4-Dimethyl-9-methyl-1,2-dihydropiecene

C25H22 322 77 1.5 0 0 0 0 0 0 I

1,2-(50-Isopropylcyclopenteno)-7-methylchrysene

C25H24 324 0 0 37 3.3 0 0 0 0 I

1,2,9-Trimethyl-1,2,3,4-tetrahydropicene

C25H24 324 357 6.8 44 4 39 2.2 0 0 I


C25H24 324 339 6.4 47 4.3 674 38 0 0 I

3,10-Dimethyl-3,4-(300-isopropylcyclopenteno)-1,2-dihydrochrysene

C26H26 340 461 8.7 0 0 377 21.2 0 0 I

1,2,4a,9-Tetramethyl-1,1a,2,3,4,4a-hexahydropicene

C26H28 340 334 6.3 0 0 456 25.7 0 0 I

2,2,4a,9-Tetramethyl-1,1a,2,3,4,4a-hexahydropicene

C26H28 340 0 0 0 0 187 10.6 0 0 I

3,10-Dimethyl-3,4-(30-isopropylcyclopenteno)-1,2,3,4-tetrahydrochrysene

C26H30 342 443 8.4 0 0 414 23.3 0 0 I

1-4,4a,9-Tetramethyl-1,1a,2,3,4,4a,5,6-octahydropicene

C26H30 342 0 0 58 5.2 687 38.8 0 0 I

1,2,4a,9-Tetramethyl-1,1a,2,3,4,4a,5,6-octahydropicene

C26H30 342 180 3.4 0 0 472 26.6 0 0 I

2,2,4a,9-Tetramethyl-1,1a,2,3,4,4a,5,6-octahydropicene

C26H30 342 0 0 19 1.7 257 14.5 0 0 I

9-Methyl-3,4-indeno-(30-isopropylcyclopenteno)picene

C33H28 424 0 0 226 20.4 0 0 0 0 I

standard (e.g. perdeuterated tetracosane, C24D50). Emissionrates were determined from weight measurements of samplebefore and after burning.

3. Results and discussion

The organic compounds identified in the solvent

soluble fraction of coal smoke particles are given inTable 1 with the emission rates and percent abundances(relative to the maximum compound concentration). Thedistributions of the molecular classes include thefollowing: homologous series ofn-alkanes, n-alkenes,n-alkan-2-ones,n-alkanals,n-alkanoic acids,n-alkenoicacids, n-alkanols, n-alkylbenzenes andn-alkylnitriles;polycyclic aromatic hydrocarbons (PAH); phenolics;


Table 1 (continued)




III. Polycyclic aromatic hydrocarbonsTotal C3-Naphthalenes C13H14 170 0 0 142 12.8 75 4.2 61 7.2 IPhenanthrene C14H10 178 129 2.5 31 4.8 152 8.6 239 28.3 AAnthracene C14H10 178 0 0 6 0.6 85 4.8 105 12.4 A9-Methylphenanthrene C15H12 192 23 0.4 14 1.2 123 6.9 122 14.5 A1-Methylphenanthrene C15H12 192 116 2.2 26 2.4 176 9.9 125 14.9 A2-Methylphenanthrene C15H12 192 35 0.7 11 1 128 7.8 108 12.8 A3-Methylphenanthrene C15H12 192 40 0.8 26 2.4 172 9.8 92 10.9 A9-Methylphenanthrene C15H12 192 32 0.6 26 2.4 291 16.4 246 29.2 ATotal C1-Anthracenes/phenanthrenes C15H12 192 245 4.6 26 2.4 902 50.9 694 82.3 AFluoranthene C16H10 202 40 0.8 0 0 323 18.2 614 72.8 APyrene C16H10 202 7 0.1 29 2.6 355 20 561 66.6 ATotal C2-Anthracenes/phenanthrenes C16H14 206 174 3.3 75 6.7 1773 100 843 100 A7(H)-Benzol[c]fluorene C17H12 216 0 0 0 0 0 0 761 90.3 I11(H)-Benzol[b]fluorene C17H12 216 0 0 0 0 482 27.2 182 21.6 ITotal C1-Pyrenes/fluoranthenes C17H12 216 128 2.4 37 3.3 783 44.2 610 72.4 ITotal C3-Anthracenes/phenanthrenes C17H16 220 196 3.7 23 2 554 31.2 97 11.5 SBenzo[ghi]fluoranthene C18H10 226 0 0 0 0 58 3.3 0 0 ACyclopental[cd]pyrene C18H10 226 0 0 0 0 187 10.6 224 26.6 SBenz[a]anthracene C18H12 228 0 0 17 1.5 0 0 70 8.3 AChrysene/triphenylene C18H12 228 14 0.3 18 1.6 104 5.9 323 38.4 ATotal C2-Pyrenes C18H14 230 0 0 0 0 178 10.1 0 0 I1,2-Diphenylbenzene C18H14 230 0 0 0 0 0 0 237 28.1 ITotal C1-Chrysenes C19H14 242 0 0 18 1.6 878 49.5 372 44.1 IBenzo[k]fluoranthene C20H12 252 0 0 0 0 46 2.6 273 32.4 ABenzo[a]pyrene C20H12 252 0 0 0 0 53 3 194 23.1 ABenzo[e]pyrene C20H12 252 0 0 0 0 51 2.9 253 30.1 APerylene C20H12 252 0 0 0 0 158 8.9 720 85.4 Atotal C2-Chrysenes C20H16 256 0 0 26 2.4 367 20.7 0 0 IAnthanthrene C22H12 276 0 0 0 0 88 5 253 30.1 ABenzo[ghi]perylene C22H12 276 0 0 0 0 78 4.4 291 34.5 AIndeno[cd]pyrene C22H12 276 0 0 0 0 0 0 158 18.7 ACoronene C24H12 300 0 0 0 0 0 0 74 8.8 AIV. Phenols (methoxy)Catechol C6H6O2 110 304 5.8 34 3 31 1.7 16 1.8 I1,3-Dihydroxybenzene (resorcinol) C6H6O2 110 105 2 0 0 0 0 15 1.8 I1,5-Bisguaiacylpentane-1,5-dione C19H20O6 344 587 11.1 318 28.8 23 1.3 0 0 I2-Guaiacyl-5-(20-guaiacylethyl)-tetrahydrofuran

C20H24O5 344 421 8 268 24.2 21 1.2 0 0 I

Divanillyl C16H18O4 274 271 5.1 0 0 0 0 0 0 I1,2-Divanillylethane C18H22O4 302 306 5.8 0 0 0 0 0 0 IPinoresinol C20H22O6 358 336 6.4 83 7.5 0 0 0 0 IV. MiscellaneousUCM (g/kg) 96 73 40 24 IU:R 3.2 2.9 3.3 3.3Elemental carbon (g/kg) 0.169 0.205 0.474 1.378VOC/Elemental Carbon 88 87 11 6

and hopanoid, steroid and terpenoid molecular biomarkers.The VOC and EC data for the coal smoke samples are alsogiven in Table 1. The distributions and abundances of the coalsmoke constituents are strongly dependent on coal rank and oncombustion temperature (smoldering versus flaming condi-tions) and duration. Thus, the values reported here shouldnot be used as absolute but as relative chemical fingerprintsfor these sources. The compounds are first discussed in classesand then summarized as the major markers in the totalmixtures.

3.1. Homologous compound series

The GC traces of the coal smoke samples are

provided in Figs. 2–5. The total extract shows thedistributions and abundances of all major organicconstituents, whereas TLC fractions 1–4 show thedistributions of homologous aliphatic series, PAH andmolecular biomarkers separated according to functionalgroup and polarity properties. This separation aids thecompound identifications in the total mixtures. In order todefine source-specific chemical fingerprints for coal smokeemissions, the discussion will focus on the identity anddistributions (carbon number range and maxima, and carbonpreference indices) [40] of aliphatic homologues andbiomarkers. Where possible, comparisons will be made todistinguish differences between mature (bituminous) versusless mature (lignite) coals.


Fig. 2. GC–MS total in current traces of lignite coal smoke particulate matter: (a) total extract showing the major organic components; (b) F1 fraction showingn-alkanes,n-alkenes, and C31 hopane biomarker, a� 3; 10-dimethyl-3,4-(30-isopropylcyclopenteno)-1,2,3,4-tetrahydrochrysene; (c) F2 fraction showing PAHandn-alkanoic acids; (d) F3 fraction showingn-alkanones and triterpenoid biomarkers; and (e) F4 fraction showingn-alkanols (numbers refer to carbon chainlength ofn-alkanes, A� y-alkanoic acid, OH� n-alkanol, K� n-alkanone, UCM� unresolved complex mixture, IS� internal standard).

3.1.1. n-Alkanes and n-alkenesThe overall distribution ofn-alkanes in coal smoke ranges

in carbon chain length from C14 to C34 and indicates that theoriginal organic matter is derived from higher plant sources.The odd-to-even carbon number predominance.C25

(CPI� 0:9–1:2) and C29 maximum for lignite suggeststhat the volatilizable alkanes are in part epicuticular plantwax. Vascular plants synthesize epicuticular waxes contain-ing odd carbon numbern-alkanes usually in the C25–C33

range with C29 or C31 as dominant homologues whichoften contribute up to 90% of all paraffins found in plantwaxes [41,42]. Both sub-bituminous and bituminous smokesamples have an even-to-odd carbon number predominance(CPI� 0:9) with Cmax at C27 and C24, respectively. The even

carbon numbern-alkanes can be formed by maturationprocesses (diagenesis and catagenesis) or during burningand their presence fits with the higher ranks of these coals.

n-Alkenes have even carbon number predominances(CPI� 0:1–0:9) and range from C14 to C28 with Cmax forall coals of C20 or greater as even carbon numbers.n-Alkenes are formed primarily by the thermal dehydrationof n-alkanols and to a lesser degree from then-alkanes byoxidation during incomplete combustion ([3]). In the case ofcoal burning, they can also derive from bound or free alkylmoieties. The carbon number range forn-alkenes compareswell with the distribution ofn-alkanols (range from C18 toC32 with even carbon number distributions and Cmax at C24

for brown coal and C30 for lignite, respectively) which


Fig. 3. GC–MS total ion current traces of brown coal smoke particulate matter: (a) total extract showing the major organic components; (b) F1 fraction showingn-alkanes andn-alkenes; (c) F2 fraction showingn-alkanoic acids and PAH; (d) F3 fraction showingn-alkanones andn-alkanals; and (e) F4 fraction showingn-alkanols (abbreviations as in Fig. 2 and al� n-alkanal).

supports the dehydration pathway. However, the presence ofn-alkenes in smoke from sub-bituminous and bituminouscoals, which do not containn-alkanols, indicates a deriva-tion from the bulk matrix by cracking of alkyl containingmoieties (e.g.n-alkylaromatics).

3.1.2. n-Alkan-2-onesThe straight chain ketones asn-alkan-2-ones range from

C9 to C33 and show an odd-to-even carbon number predo-minance (CPI range� 0:9–1:9). Lignite through sub-bitu-minous coal smoke samples each had Cmax at C29, whereasbituminous coal had a Cmax at C23. An example of such afraction is shown in Fig. 3(d) for brown coal smoke.n-Alkan-2-ones have been described to have a dual origin. Thefirst is from partial combustion processes and secondarily

from microbial alteration of plant wax lipids in situ [43] orin soils [44]. Their introduction into coal smoke is frompartial combustion of aliphatic precursors.

3.1.3. n-Alkanalsn-Alkanals (aldehydes) range from C13 to C30 and show an

even-to-odd carbon number predominance (CPI range�2–3) with Cmax at C24 or C30. They are only emitted insmoke from the immature coals.n-Alkanals .C25 insediments with an even-to-odd carbon number predomi-nance were interpreted to be derived from vegetativewax [41,45]. Here they may be intermediate thermaloxidation products from alkane or alkanol precursorsbound to the coal matrix.


Fig. 4. GC–MS total ion current traces of sub-bituminous coal smoke particulate matter: (a) total extract showing the major organic components; (b) F1fraction showingn-alkanes; (c) F2 fraction showingn-alkanoic acids and PAH; (d) F3 fraction showingn-alkanoic acids, PAH and C2-picene biomarker; and(e) F4 fraction showing polar compounds (low level) (abbreviations as in Fig. 2).

3.1.4. n-Alkanoic and n-alkenoic acidsThe n-alkanoic acids emitted in smoke from all coals

range from C8 to C34, show a Cmax at C24 or C26 forimmature coals and at C16 for mature coals, and haveeven to odd carbon number predominances(CPI range� 0:9–4:3). These compounds are mainlyderived from precursor higher plant sources and are identi-fied here as a major molecular class for all coal smokesamples. A strong trend is observed for then-alkanoicacid CPI values which decrease with increasing maturity(rank).

The unsaturatedn-alkenoic acids show an even to oddcarbon number predominance (CPI� 1:2 and 4:0) andrange from C10:1 to C23:1 with Cmax at C16:1 and C18:1. Theyoccur mainly in smoke from the lignite and brown coal. The

C16:1 and C18:1 n-alkenoic acids have been used as indicatorsof recent biogenesis in aerosols e.g. [45]. Their presence incoal smoke shows that these compounds are highly stable incoal over geological times, perhaps due to sequesteringwithin the coal matrix where they are protected from chemi-cal degradation.

3.1.5. n-AlkanolsHomologous series ofn-alkanols with even-to-odd

carbon number predominances were identified in smokefrom lignite (CPI� 3:2) and brown coal (CPI� 16:1) butnot from sub-bituminous or bituminous coals. Then-alka-nols ranged from C18 to C32 with Cmax at C24 for brown coaland C30 for lignite. Then-alkanols from C26 to C30 are predo-minantly of a vascular plant wax origin and may be


Fig. 5. GC–MS total in current traces of bituminous coal smoke particulate matter: (a) total extract showing the major organic components; (b) F1 fractionshowingn-alkanes and C29 hopane biomarker; (c) F2 fraction showing PAH; (d) F3 fraction showing PAH; (e) F4 fraction showingn-alkanoic acids and polarcompounds (low level) (abbreviations as in Fig. 2).


Table 2Emissions of source-specific biomarkers identified for smoke samples from individual coal types (all identifications were made from mass spectrometry).Given as emission rates (mg/kg of coal burned).

Compound name Molecular formula Molecular mass Coal smoke

Lignite (mg/kg) Brown (mg/kg) Sub-bituminous (mg/kg) Bituminous (mg/kg)

Aromatic biomarkers3,4,7-Trimethyl-1,2,3,4-tetrahydrochrysene

C21H22 274 9

Picene C21H14 278 94C1-Picenes C23H16 292 41C2-Picenes C24H18 306 76 112,2-Dimethyl-1,2,3,4-tetrahydropicene

C24H22 310 95

1,2,9-(1,7,8)-Trimethylpicene C25H20 320 102 4 1892-Ethyl-9-methyl-1,2-dihydropicene

C25H22 322 306

1,2-(50-Isopropylcyclopenteno)-7-methylchrysene

C25H24 324 37


C25H24 324 357 44 39


C25H24 324 339 47 674

Hopanoids22,29,30-Trisnorneohop-13(18)-ene

C27H44 368 356

22,29,30-Trisnorhop-17(21)-ene C27H44 368 325 12 150 1717a(H)-22,29,30-Trisnorhopane C27H46 370 195 10 430 6717b(H)-22,29,30-Trisnorhopane C27H46 370 442 15 492 2317a(H),21b(H)-29-Norhopane C29H48 398 56 19 656 8517b(H),21a(H)-29-Norhopane C29H48 398 566 7 729 4317a(H),21b(H)-Hopane C30H50 412 97 10 335 4517b(H),21a(H)-Hopane C30H48 412 1111 6 473 4217b(H),21b(H)-Hopane C30H48 412 4322R-17a(H),21b(H)-Homohopane

C31H54 426 840 12 160 17

22S-17a(H),21b(H)-Homohopane

C31H54 426 49 1 40 10

17b(H),21a(H)-Homohopane C31H54 426 195 5 170 1917a(H),21b(H)-Bishomohop-31-ene

C32H54 438 233

17b(H),21a(H)-Bishomohop-31-ene

C32H54 438 122

22R-17b(H)-21a(H)-Bishomohop-31-ene

C32H54 438 75

17b(H),21b(H)-Bishomohopane C32H56 440 6217b(H),21b(H)-Trishomohop-32-ene

C33H56 452 77

PhenolicsCatechol C6H6O2 110 304 34 31 16Divanillyl C16H18O4 274 2711,2-Divanillylethane C18H22O4 302 306TerpenoidsDehydroabietane C20H30 270 278Abieta-8,11,13-trien-7-one C20H28O 284 69Ferruginol C20H30O 286 42224,25-Dinorursana-1,3,5(10),12-tetraene

C28H40 376 246

24,25-Dinorlupa-1,3,5(10),22(29)-tetraene

C28H40 376 99

24,25-Dinorlupa-1,3,5(10)-triene

C28H42 378 99

4-Methyl-24-ethyl-19-norcholesta-1,3,5(10)-triene

C29H46 394 58


Table 2 (continued)

Compound name Molecular formula Molecular mass Coal smoke

Lignite (mg/kg) Brown (mg/kg) Sub-bituminous (mg/kg) Bituminous (mg/kg)

A-Neoursa-3(5),12-diene C30H48 408 648Oleana-2,12-diene C30H48 408 921Usana-2,12-diene C30H48 408 275Olean-12-ene C30H50 410 1161Allobetul-2-ene C30H48O 424 117a-Amyrone C30H48O 424 186b-Amyrone C30H48O 424 823U:R 3.2 2.9 3.3 3.3Homohopane index, C31

�S=�S1 R��0.05 0.08 0.20 0.37

17a(H),21b(H)-Hopane/22R-17a(H),21b(H)-Homohopane

0.1 0.8 2.1 2.6

Fig. 6. GC–MS key ion fragmentograms�m=z� 191� of hopanoid biomarkers in coal smoke particulate matter: (a) Lignite coal smoke; (b) Brown coal smoke;(c) Sub-bituminous coal smoke; and (d) Bituminous coal smoke (numbers refer to compound carbon number, Tt� triterpenoid,R and S� configuration atcarbon-22 in hopanes. C31, a � 17a�H�,21b(H) configuration,ba � 17b�H�,21a(H); maturity parameters:bb � immature,ba � moderately mature, anda � fully mature).

entrapped as such in the coals or bound to the organicmatrix.

3.1.6. n-AlkylbenzenesA homologous series ofn-alkylbenzenes ranging from C13

to C32 was identified in all coal smoke samples. Then-alkyl-benzenes all showed even carbon number predominances(CPI range� 1:0–1:3) and had Cmax at C30 for lignite,brown and sub-bituminous coals, with a Cmax at C20 forbituminous coal. Then-alkyl side chains ranged from C7

to C26 for all samples. These aromatic molecules are likelyderived from the thermal alteration ofn-alkylcyclohexaneprecursors which have been identified in German browncoals [29]. n-Alkylbenzenes have also been reported ascommon constituents of coal pyrolysates and have beenshown to decrease in abundance with increasing coal matur-ity [46]. Such a trend related to maturity was not found forthe few coal smoke samples analyzed here, however, theCPI decreases with higher rank.

3.1.7. n-AlkylnitrilesA series ofn-alkylnitriles ranging from C24 to C30 with a

Cmax at C30 (CPI� 27) was identified only in smoke fromthe brown coal. These compounds have not been reportedpreviously in unburned coal extracts, however, they havebeen found in pyrolysates of kerogens [47], in urban aero-sols [48], and in smoke from charbroiling and meat cookingoperations [49]. We suggest that then-alkylnitriles areformed by pyrolytic processes.

3.2. Aliphatic biomarkers

Molecular biomarkers are organic compounds of a biolo-gical origin that show little or no change in chemical struc-ture from their parent precursor compound found in living

organisms. Such molecules are characterized by theirrestricted occurrence, source specificity, molecular stabilityand suitable concentration for analytical detection[12,40,45,50,51]. The major molecular biomarkers identi-fied in these coal smoke samples include the hopanoidhydrocarbons, steroid hydrocarbons, and terpenoids withtheir thermal alteration products.

3.2.1. Hopanoid hydrocarbonsThe hopanoid hydrocarbon series (C27–C35) is derived

from precursors in the cell membranes of prokaryotes(bacterial source) and cyanobacteria (blue–green algaesource) in sedimentary organic matter over geologicaltime. Hopanoids (C30 mainly) are also known to be presentin certain higher plants such as ferns [52]. The microbialhopanes are thought to derive from diploptene (C30) and aC35 bacteriohopanetetrol precursor which has been isolatedfrom various microorganisms [53,54]. Hopanes in higherplants are derived from squalene via diploptene and consistmainly of the C30 compounds (e.g. hopane, moretane,lupane structures) [13,51].

The distributions and relative abundances of the hopaneseries in the coal smoke samples are given in Table 1 andFig. 6. The hopanes range from C27 to C33 and have the Cmax

at C29 or C31. Lignite smoke, which is the least mature coal,is dominated by 22R-17a(H),21b(H)-homohopane, whereassmoke from the more mature coals have 17a(H),21b(H)-29-norhopane as the most abundant component. It should benoted that lupanes, ursanes and oleananes are not detectable.The identification of these compounds is based primarily ontheir mass spectra and GC retention time in the key ion (m/z:191) fragmentogram [52,13,55].

The stereochemical configurations at the C-17 and C-21positions of hopanes are often used to determine the maturityof a geological sample, where hopanes with configurations


Fig. 7. Bar plot showing the relative abundances (%) and distributions of hopane biomarkers in the coal smoke samples and in exhaust emissions from gasolineand diesel engines.


Tab

le3

Em

issi

ons

ofso

urce

-spe

cific

biom

arke

rsid

entifi

edfo

rco

alsm

oke

(all

iden

tifica

tions

wer

em

ade

from

mas

ssp

ectr

omet

ry)

Com

poun

dna

me

Mol

ecul

arfo

rmul

aM

olec

ular

mas

sC

oals

mok

e

Lign

iteB

row

nS

ub-b

itum

inou

sB

itum

inou

s

(mg/

kg)a

(g/k

gE

C)b

(mg/

kg)a

(g/k

gE

C)b

(mg/

kg)a

(g/k

gE

C)b

(mg/

kg)a

(g/k

gE

C)b

Cat

echo

lC

6H6O

211

030

419

0034

139

3168

1612

C2-

Pic

enes

C24

H18

306

00

00

7616

611

81,

2,9-

(1,7

,8)-

Trim

eth

ylpi

cene

C 25H

2032

010

263

84

1618

941

21,

2,9-

Trim

ethy

l-1,

2,3,

4-te

trah

ydro

pic

ene

C25

H24

324

357

2231

4418

039

85

2,2,

9-T

rimet

hyl-

1,2,

3,4-

tetr

ahyd

rop

icen

eC

25H

2432

433

921

1947

193

674

1471

22,2

9,30

-Tris

norh

op-1

7(21

)-en

eC

27H

4436

832

520

3112

4915

032

717

13

17a

(H)-

22,2

9,30

-T

risno

rhop

ane

C27

H46

370

195

1219

1041

430

938

6750

17b

(H)-

22,2

9,30

-T

risno

rhop

ane

C27

H46

370

442

2763

1562

492

1073

2317

17a

(H),

21b

(H)-

29-N

orho

pane

C 29H

4839

856

350

1978

656

1431

8563

17b

(H),

21a

(H)-

29-N

orho

pane

C 29H

4839

856

635

387

2972

915

9143

3217

a(H

),21b

(H)-

Hop

ane

C30

H50

412

9760

610

4133

573

145

3417

b(H

),21a

(H)-

Hop

ane

C30

H50

412

1111

6944

625

473

1032

4231

22R

-17a

(H),

21b

(H)-

Hom

ohop

ane

C31

H54

426

840

5250

1249

160

349

1713

22S-

17a

(H),

21b

(H)-

Hom

ohop

ane

C31

H54

426

4930

61

440

8710

7

17b

(H),

21a

(H)-

Hom

ohop

ane

C 31H

5442

619

512

195

2117

037

119

14

aG

iven

asem

issi

onra

tes

(mg/

kgof

coal

burn

ed).

bG

iven

asco

ncen

trat

ion

rela

tive

toel

eme

ntal

(bla

ck)

carb

on.

of 17b(H),21b(H) are immature, 17b(H),21a(H) aremoderately mature and 17a(H),21b(H) are fully mature.This maturity parameter stems from the observation thatthe “biological” configuration of 17b(H),21b(H)-22R,imposed on the hopane precursors diploptene and bacterio-hopanetetrol by biochemistry in the living organism, isunstable during diagenesis and catagenesis in sediments,undergoing isomerization to the “geological“ configurations[13,55]. For example, the homohopane precursor has the17b(H),21b(H) stereochemistry and occurs with only theR configuration at the C-22 position. Thus, once it isdeposited into the geological environment, it converts tothe more thermodynamically stable 17a(H),21b(H) con-figuration, with the formation of theR and S epimers atthe C-22 position.

Lignite smoke contains mainly 17b(H),21b(H)-hopaneand 17b(H),21a(H)-moretane configurations, confirmingits immaturity. Brown coal smoke contains both17b(H),21a(H)-moretane and 17a(H),21b(H)-hopaneconfigurations which supports a moderate maturity. Bothsub-bituminous and bituminous coal smoke samples containmainly 17a(H),21b(H)-hopane and minor amounts of17b(H),21a(H)-moretane compounds, confirming thenearly full maturity of these two coals.

In typical petroleum, the extended 17a(H),21b(H)-hopane homologues.C31 have the C-22 epimers at anequilibrium ratio �S=�S1 R�� of 0.6 (homohopane index;[56]). In Fig. 6, the ratio of the 17a(H),21b(H)-homo-hopane R and S epimers is shown to decrease withincreasing coal maturity. The homohopane index�S=�S1R�� for these coal smoke samples varies from 0.05 to0.35 and increases with coal rank (lignite 0.05; browncoal 0.09; sub-bituminous coal 0.20; bituminous coal0.35). Thus this index is useful for identifying differentcoal types.

The distributions and abundances of six selected hopanespresent in coal smoke, and gasoline and diesel enginecombustion exhausts [10] were compared in order to obtaina source-specific hopane signature for coal smoke (Table 2and Fig. 7). The results show that the hopane distributions(% relative to Cmax of hopanes) for gasoline and dieselengine exhausts are very similar (petroleum source);however, both differ from the coal smoke signatures. Theratios of 17a(H),21b(H)-hopane to 22R-17a(H),21b(H)-homohopane determined for all coal smoke samples aregiven in Table 2. The ratios for gasoline (3.7) and diesel(2.5) exhaust were similar [10], and for coal smoke theyranged from 0.1 to 2.6, increasing with higher rank. Theratio can be used to distinguish between coal smoke andgasoline engine emissions, however, the high ratio ofbituminous coal smoke (2.6) does not allow for distinc-tion between this high ranking coal and diesel engineemissions. Thus, another ratio such as the 17a(H),21b(H)-29-norhopane to 22R-17a(H),21b(H)-homoho-pane would distinguish these sources. These ratios mayaid in the apportionment of source emissions from coal

burning versus vehicle exhaust (i.e. petroleum) in theatmospheric environment.

The lupanes, ursanes and oleananes were not present inthese coal smoke samples, which may be due to their non-preservation by initial oxidative sedimentation processesand conversion to aromatic biomarkers. An in-depthdiscussion on the triterpenoids and aromatic biomarkers isincluded later in this text.

3.2.2. Steroid hydrocarbonsSteroids are constituents of plant lipid membranes, and

hydrocarbon derivatives are often identified in geologicalsamples. However, in coals they are usually not prominentbecause of their oxidative loss by diagenetic processesduring coal deposition. Three C29 phytosteroid hydrocar-bons were identified at low levels in these coal smokesamples and includeb-sitoster-2-ene and 4-methyl-24-ethyl-19-norcholesta-1,3,5(10)-triene (both dehydrationproducts of bound phytosterols) and 5a(H),14a(H),17a(H)-24-ethyl-cholestane (a moderately mature steroidhydrocarbon based on itsaaa configuration [55]).

3.2.3. TerpenoidsTerpenoids are ubiquitous components of higher order

vegetation (plant resins, gums and mucilages) and algae.The terpenoids identified in the coal smoke samples includemainly di- and tri-terpenoids with their diagenetic, catage-netic and thermal alteration products. They are also theprecursors of the aromatic biomarkers [40].

The product–precursor relationship for the diterpenoidsidentified in coal smoke may follow an alteration pathwaywhich commences with the dehydration of abietic acid todehydroabietic acid with subsequent decarboxylation todehydroabietin and full aromatization to retene [51].Retene, which is present in these coal smoke samples(except bituminous), is a molecular indicator of gymnos-perm combustion [5]. Dehydroabietane, which is presentin many resins and in the lignite coal smoke, may dehydro-genate to simonellite and then proceed to retene.

The triterpenoids and their thermally altered products aremainly found in lignite smoke and are present in brown coalsmoke only as minor constituents. The triterpenoids are ofthe oleanane, lupane and ursane series and occur as thedienes and partially aromatic hydrocarbon derivatives.These compounds were probably preserved in the coalmatrix as C–O–C bound biomarkers. Further thermaldehydration of these intermediates yields the aromaticbiomarkers.

3.3. Aromatic biomarkers

Aromatic biomarkers are derived from natural products(primarily terpenoids) by diagenetic or catagenetic altera-tion of the molecular precursors (i.e. dehydrogenation anddealkylation). Aromatic biomarkers can be related back toparent precursor compounds in natural lipid fractions (e.g.


plant wax, resin, etc. [57]). Aromatic biomarkers in the coalsmoke samples consist mainly of alkylpicenes and alkylhy-dropicenes, with a minor contribution of alkylhydrochry-senes. The sequence of reactions leading to the formationof the hydropicene and hydrochrysene series has beendescribed previously [51]. Briefly, triterpenoids in the coaldepositional environment are initially oxidized duringdiagenesis to 3-oxo derivatives (e.g.a-amyrin to a-amy-rone) or dehydrated by catagenesis to aD2-olefin (boundtriterpenoids in the coal matrix also yield theD2-triterpenesduring burning). With continued oxic diagenesis andmaturation, theD2-olefins and 3-oxo derivatives undergosuccessive ring aromatizations to yield the alkylhydropiceneseries. The alkylhydrochrysene series is also derived fromtriterpenoids; however, the initial diagenetic oxidation reac-tions of the 3-hydroxy precursor or 3-oxo derivative aredriven by ultraviolet light and produce ring-A cleavageproducts (e.g. roburic acid). The ring-A cleavage productsare then oxidized further with complete loss of the A ring,followed by subsequent aromatization yielding alkylhydro-chrysenes. These aromatic biomarkers are injected as suchor with further dehydrogenation into the smoke during theburning of the coals.

3.4. Polycyclic aromatic hydrocarbons

The emission rates and relative abundances of over 30polycyclic aromatic hydrocarbon (PAH) compounds presentin the coal smoke samples are given in Table 1. PAHs arederived from the high temperature thermal alteration of coalorganic matter. In lignite and brown coal smoke the PAHsare present only as minor constituents, however, in sub-bituminous and bituminous coals the PAHs are highly abun-dant. It should be noted that the alkylpicenes and picene areonly trace components. The major PAHs in sub-bituminousand bituminous coal smoke are alkyl-disubstituted anthra-cenes and phenanthrenes. The high abundances of thesecompounds in the coal smoke reflects the degree of aroma-ticity and Ar–CH2–Ar linkages present in these two maturecoals.

Certain PAHs, such as benz[a]anthracene, benzo[a]pyr-ene and cyclopenta[c,d]pyrene, have mutagenic and geno-toxic potential [34,35]. These PAHs have been identifiedhere and their emission rates indicate that they are signifi-cant in both sub-bituminous and bituminous coal smoke.

3.5. Phenolics

Phenolic compounds are derived from lignin residues inthe coals and are composed mainly of lignin pyrolysisproducts and lignans (dimers of substituted phenols). Cate-chol and resorcinol were identified in all coal smokesamples and are common pyrolysis products from lignin(Table 1). The lignans were identified previously in emis-sions from biomass combustion and have been suggested astracers for distinguishing between coniferous versus decid-uous wood fires [6]. This is the first report of their presence

in coal smoke emissions and thus must be considered inairsheds receiving such emission inputs. The most abundantlignans were 1,5-bisguaiacylpentane-1,5-dione and 2-guaia-cyl-5-(20-guaiacylethyl)-tetrahydrofuran, which are poten-tial tracers for mainly gymnosperm and to a lesser extentangiosperm vegetation [58], both of which have contributedorganic matter to these coals.

The total emission rates were determined for the pheno-lics in all coal smoke samples and range from 31 to2330 mg/kg (lignite at 2330 mg/kg; brown coal at 703 mg/kg; sub-bituminous coal at 75 mg/kg; bituminous coal at31 mg/kg). The phenolic component emission ratesdecrease with increase in coal rank. Therefore, these mole-cules could be used as maturity indicators.

It should be noted that levoglucosan, the thermal break-down product from cellulose, is not detectable in the coalsmoke samples [62]. This confirms the general absence ofcellulose in coals.

3.6. Unresolved complex mixture

An unresolved complex mixture (UCM) of structurallycomplex isomers and homologs of branched and cycliccompounds [59] eluting between C14 and C34 is present asa major organic component of all total extracts and theiraliphatic fractions from the coal smoke samples. TheUCM, which has been thoroughly examined in petroleumsources, is comprised of compounds which are relativelyinert to microbial or thermal degradation [60,61]. TheUCM of coal smoke is unlike that of petroleum because itpartitions into all TLC fractions. This indicates that it is notcomposed entirely of hydrocarbons but is a mixture contain-ing heteroatomic and functional groups. The emission ratesof UCM range from 24 to 96 g/kg (Table 1) and decreasewith increasing coal rank. This parameter is therefore animportant indicator of geological maturity of the fuelwhen assessing coal smoke emissions.

The ratio of UCM to resolved components�U : R� is auseful parameter for the indication of petroleum contribu-tion to aerosol samples [40].U : R ratios for the various coalsmoke samples were determined from the total extract inorder to assess whether this parameter could be used as anindicator for the presence of coal smoke in the atmosphere(Table 1). The respectiveU : R ranges for rural, mixed andurban aerosol samples from the western US are 0.2–4.0,1.4–3.4 and 0.9–25.0 [40]. Generally, urban aerosolscontain the largest ratio of petroleum-derived compounds,whereas rural and mixed rural/urban environments showvariable contributions of these anthropogenic contaminants.TheU : R ratios measured here ranged from 2.9 to 3.3 andare similar for all coal smoke samples. This similarity inU :

R ratio suggests that it may be a general indicator for all coalsmoke emissions. The averageU : R value of 3.1 indicatesthat coal smoke can be identified in urban airsheds that areinfluenced by coal burning and can further be distinguishedfrom the UCM in urban vehicular emission sources.


3.7. Source-specific molecular markers

Dominant source-specific molecular markers were desig-nated for smoke from each coal type (Table 2) and a groupof major markers was selected to represent all coal smokeemissions (Table 3). Lignite smoke (19 markers) may bereadily identified by its dominant C31-hopanes, divanillyland 1,2-divanillylethane lignans, diterpenoids, and dehydro-genated and 3-oxo triterpenoid derivatives. Brown coalsmoke markers (3) include allobetul-2-ene, 3,4,7-trimethyl-1,2,3,4-tetrahydrochrysene and 1,2-(50-isopropyl-cyclopenteno)-7-methylchrysene. Sub-bituminous coalsmoke markers (5) include 17b(H),21b(H)-hopane, C28

triterpenoids and 2,2-dimethyl-1,2,3,4-tetrahydropicene.The bituminous coal smoke markers (2) are picene andmethylpicenes.

The selected molecular markers (17) that are representa-tive of all coal smoke emissions include a series of C27 to C31

hopanes (no C28), phenolics, C2-picenes and C2-hydropi-cenes (Table 3). The molecular markers identified in thedifferent coal smoke samples are source specific and areuseful for confirmation purposes. Combined as such or inratios with other information (e.g. homologous series distri-butions and abundances) they become a powerful tool fordistinguishing atmospheric aerosol source contributionswhen affected by coal burning.

3.8. Volatile and elemental carbon

VOC and EC emission rates for the coal smoke samplesare given in Table 1. The VOC emission rates generallydecreased with increase in coal rank. This is due primarilyto higher levels of reactive oxygen bearing functionalgroups in lower rank coals which are more closely relatedto their biofuel precursors. The EC emission rates increasedwith increase in coal rank. This is due primarily to thegreater percent carbon content resulting from the losses ofhydrogen and oxygen with increased coal maturity. Theratio of VOC/EC decreased with increase in coal rankwith the greatest change occurring between brown coaland sub-bituminous coal smokes. This observation likelyreflects an increased level of catagenetic alteration occur-ring in situ at this stage of the coalification process. TheVOC/EC ratios reported here may be used as indicatorsfor tracking these combustion source emissions in the atmo-spheric environment. However, caution must be exercisedwhen using VOC/EC ratios as numerous sources of carbo-naceous matter (e.g. diesel and non-diesel vehicle exhaust,incinerators, tire debris) contribute to varying extents to theambient background levels of EC [36].

4. Conclusions

Coal burning injects airborne particulate emissions intothe atmosphere containing both natural and thermal altera-tion products from the organic matter of the fuel. The

abundances and distributions of the compounds is depen-dent on combustion temperature, flame aeration, fire dura-tion, and coal rank. The major parameters that may be usedfor identifying coal smoke emissions include (1) the unre-solved to resolved component ratio (U : R range 2.9–3.3),(2) the distributions and abundances of aromatic molecularmarkers, specifically picene, alkylated picenes and alkylhy-dropicene derivatives, (3) the 17a(H),21b(H)-hopane to22R-17a(H),21b(H)-homohopane index (also norhopane/homohopane index), and (4) the source-specific biomarkersthat identify coal smoke in ambient aerosols. Some impor-tant indicators for determining coal rank or maturity of thefuel being burned in smoke samples have been identifiedand include (1) the decrease in CPI ofn-alkanoic acids withincreasing rank, (2) the decrease in UCM emissions withincreasing rank, (3) the decrease in phenolic compoundemissions with increasing rank, and (4) the increase inhomohopane index�S=�S1 R�� with increasing rank.

Acknowledgements

Partial financial support for this work by the US Environ-mental Protection Agency (Grant CR823990-01-0) is grate-fully acknowledged.

References

[1] IPCC, International Panel on Climate Change. In: Houghton JT,Jenkins GJ, Ephraums JJ, editors. Climate change: the IPCC scientificassessment, Cambridge, UK: University Press, 1990.

[2] IPCC, International Panel on Climate Change. In: Houghton JT, Call-ander BA, Varney SK, editors. The supplementary report to the IPCCscientific assessment, Cambridge, UK: University Press, 1992.

[3] Abas MR, Simoneit BRT, Elias V, Cabral JA, Cardoso JN. Composi-tion of higher molecular weight organic matter in smoke aerosol frombiomass combustion in Amazonia. Chemosphere 1995;30:995–1015.

[4] Hawthorne SB, Miller DJ, Barkley RM, Krieger MS. Identification ofmethoxylated phenols as candidate tracers for atmospheric woodsmoke pollution. Environmental Science and Technology1988;22:1191–6.

[5] Ramdahl T. Retene - A molecular marker of wood combustion inambient air. Nature 1983;306:580–2.

[6] Simoneit BRT, Rogge WF, Mazurek MA, Standley LJ, HildemannLM, Cass GR. Lignin pyrolysis products, lignans, and resin acids asspecific tracers of plant classes in emissions from biomass combus-tion. Environmental Science and Technology 1993;27:2533–41.

[7] Simoneit BRT, Schauer JJ, Nolte CG, Oros DR, Elias VO, Fraser MP,Rogge WF, Cass GR. Levoglucosan, a tracer for cellulose in biomassburning and atmospheric particles. Atmospheric Environment1999;33:173–82.

[8] Standley LJ, Simoneit BRT. Composition of extractable plant wax,resin and thermally matured compounds in smoke particles fromprescribed burns. Environmental Science and Technology1987;21:163–9.

[9] Standley LJ, Simoneit BRT. Resin diterpenoids as tracers for biomasscombustion aerosols. Journal of Atmospheric Chemistry 1994;18:1–15.

[10] Rogge WF, Hildemann LM, Mazurek MA, Cass GR, Simoneit BRT.Sources of fine organic aerosol: 2. Noncatalyst and catalyst-equipped


automobiles and heavy-duty diesel trucks. Environmental Scienceand Technology 1993;27:636–51.

[11] Simoneit BRT. Organic matter of the troposphere. III. Characteriza-tion and sources of petroleum and pyrogenic residues in aerosols overthe western United States. Atmospheric Environment 1984;18:51–67.

[12] Simoneit BRT. Cyclic terpenoids of the geosphere. In: Johns RB,editor. Biological markers in the sedimentary record, Amsterdam:Elsevier, 1986. p. 364.

[13] Gogou A, Stratigakis N, Kanakidou M, Stephanou EG. Organic aero-sols in Eastern Mediterranean: components source reconciliation byusing molecular markers and atmospheric back trajectories. OrganicGeochemistry 1996;25:79–96.

[14] Rogge WF, Hildemann LM, Mazurek MA, Cass GR, Simoneit BRT.Journal of Geophysical Research 1996;101:379–94 also p. 19.

[15] Schauer JJ, Rogge WF, Hildemann LM, Mazurek MA, Cass GR,Simoneit BRT. Source apportionment of airborne particulate matterusing organic compounds as tracers. Atmospheric Environment1996;30:3837–55.

[16] Van Krevelen DW. Coal, typology, chemistry, physics, constitution,Amsterdam: Elsevier, 1961. p. 514.

[17] Murchison D, Westoll TS. Coal and coal-bearing strata, New York:Elsevier, 1968. p. 418.

[18] Tissot BP, Welte DH. Petroleum formation and occurrence, Berlin:Springer, 1984. p. 699.

[19] Tillman DA. Combustion characteristics of lignite and coal: the domi-nant solid fossil fuels, In: The combustion of solid fuels and wastes,London: Academic Press, 1991. p. 378.

[20] Larson J. Coal structure, Chemistry and physics of coal utilization,1980. American Institute of Physics Conference Proceedings, 70.New York: American Institute of Physics, 1981 p. 1–27.

[21] Chen SL, Heap MP, Pershing DW, Martin GB. Fate of coal nitrogenduring combustion. Fuel 1982;61:1218–24.

[22] Vogt RA, Laurendeau NM. Effect of devolatilization on nitric oxideformation from coal nitrogen. Fuel 1978;57:232–5.

[23] Schlosberg RH, editor. Chemistry of coal conversion, New York:Plenum Press, 1985. p. 336.

[24] Solomon PR, Hamblen DG, Carangelo RM, Serio MA, DeshpandeGV. General model of coal devolatilization. Energy and Fuels1988;2:405–22.

[25] Stein SE. Free radicals in coal conversion. In: Schlosberg RH, editor.Chemistry of coal conversion, New York: Plenum Press, 1985. p. 13–44.

[26] Hershkowitz F. Mass transfer effects in coal conversion chemistry. In:Shlosberg RH, editor. Chemistry of coal conversion, New York:Plenum Press, 1985. p. 45–66.

[27] Gonzalez-Vila FJ, Martin F, Cubero F. Composicion quimica de lafraccion cerea del lignito de Puentes de Garcia Rodriguez. Anales deQuimica 1987;83:295–9.

[28] Del Rio JC, Gonzalez-Vila FJ, Martin F. Ceras fosiles en turba ylignito. Composicion quimica y significancia geoquimica. Anales deQuimica 1991;87:217–22.

[29] Schulze T, Michaelis W. Structure and origin of terpenoid hydrocar-bons in some German coals. Organic Geochemistry 1990;16:1051–8.

[30] Wang T-G, Simoneit BRT. Organic geochemistry and coal petrologyof tertiary brown coal in the Zhoujing mine Baise Basin, South China:2. Biomarker assemblage and significance. Fuel 1990;69:12–20.

[31] Li M, Johns RB, Mei B. A study in early diagenesis: Biomarkercomposition of a suite of immature coals and coaly shales. OrganicGeochemistry 1990;16:1067–75.

[32] Jenisch A, Richnow HH, Michaelis W. Chemical structural units ofmacromolecular coal components. Organic Geochemistry 1990;16:917–29.

[33] Barrick RC, Furlong ET, Carpenter R. Hydrocarbon and azaarenemarkers of coal transport to aquatic sediments. EnvironmentalScience and Technology 1984;18:846–54.

[34] Arcos JC, Argus MG. Chemical induction of cancer. Structural basisand biological mechanisms, IIA. New York: Academic Press, 1975.

[35] World Health Organization. IARC Monographs, 46. Lyon, France:International Agency for Research on Cancer, 1989. p. 41–155.

[36] Birch ME, Cary RA. Elemental carbon-based method for monitoringoccupational exposures to particulate diesel exhaust. Aerosol Scienceand Technology 1996;25:221–41.

[37] Johnson RI, Shaw JJ, Cary RA, Huntzicker JJ. An automated thermal-optical method for the analysis of carbonaceous aerosol. Atmosphericaerosol: source/air quality relationsips, ACS Symposium Series, 167.Washington, DC: ACS, 1981. p. 223–33.

[38] Simoneit BRT, Mazurek MA. Organic matter of the troposphere-IINatural background of biogenic lipid matter in aerosols over the ruralWestern United States. Atmospheric Environment 1982;16:2139–59.

[39] Schlenk H, Gellerman JL. Esterification of fatty acids with diazo-methane on a small scale. Analytical Chemistry 1960;32:1412–4.

[40] Mazurek MA, Simoneit BRT. Characterization of biogenic and petro-leum-derived organic matter in aerosols over remote, rural and urbanareas. In: Keith LH, editor. Identification and analysis of organicpollutants in air, Boston, MA: Ann Arbor Science/Butterworth,1984. p. 353–70.

[41] Kolattukudy PE. Plant waxes. Lipids 1970;5:259–75.[42] Kolattukudy PE, editor. Chemistry and biochemistry of natural waxes

Amsterdam: Elsevier, 1976. p. 459.[43] Simoneit BRT. The organic chemistry of marine sediments. In: Riley

JP, Chester R, editors. Chemical oceanography, 7. New York:Academic Press, 1978. p. 233–311 ch. 39.

[44] Morrison RI, Bick W. Long-chain methyl ketones in soils. ChemicalIndustries 1966;596–7.

[45] Simoneit BRT. Organic matter of the troposphere. V: Application ofmolecular marker analysis to biogenic emissions into the tropospherefor source reconciliations. Journal of Atmospheric Chemistry1989;8:251–75.

[46] Gallegos EJ. Alkylbenzenes derived from carotenes in coals by GC/MS. Journal of Chromatographic Science 1981;19:177–82.

[47] Derenne S, Largeau C, Taulelle F. Occurrence of non-hydrolysableamides in the macromolecular constituent of Scenedesmus quadri-cauda cell wall as revealed by15N-NMR: origin of n-alkylnitriles inpyrolysates of ultralaminae-containing kerogens. GeochimicaCosmochimica Acta 1993;57:851–7.

[48] Abas MR, Simoneit BRT. Composition of extractable organic matterof air particles from Malaysia: Initial study. Atmospheric Environ-ment 1996;30:2779–93.

[49] Rogge WF, Hildemann LM, Mazurek MA, Cass GR, Simoneit BRT.Sources of fine organic aerosol: 1. Charbroilers and meat cookingoperations. Environmental Science and Technology 1991;28:1375–88.

[50] Simoneit BRT. Organic matter in eolian dusts over the AtlanticOcean. Marine Chemistry 1977;5:443–64.

[51] Simoneit BRT. Biomarker PAHs in the environment. In: Nielson AH,editor. The handbook of environmental chemistry, Berlin: Springer,1998. pp. 176–221 ch. 5.

[52] Philp RP. 94. Fossil fuel biomarkers—application and spectra. meth-ods in geochemistry and geophysics 23, Amsterdam: Elsevier, 1985.p. 294.

[53] Ourisson G, Albrecht P, Rohmer M. The hopanoids. Palaeochemistryand biochemistry of a group of natural products. Pure and AppliedChemistry 1979;51:709–29.

[54] Ourisson G, Albrecht P, Rohmer M. Predictive microbial biochemis-try—from molecular fossils to procaryotic membranes. Trends inBiological Sciences 1982;236–9.

[55] Peters KE, Moldowan JM. The biomarker guide: interpreting mole-cular fossils in petroleum and ancient sediments, Englewood Cliffs,NJ: Prentice-Hall, 1993. p. 363.

[56] Seifert WK, Moldowan JM. Applications of steranes, terpanes, andmonoaromatics to the maturation, migration, and source of crude oils.Geochimica Cosmochimica Acta 1978;42:77–95.

[57] Johns RB, editor. Biological markers in the sedimentary record,Amsterdam: Elsevier, 1986. p. 364.


[58] Hawthorne SB, Krieger MS, Miller DJ, Mathiason MB. Collectionand quantification of methoxylated phenol tracers for atmosphericpollution from residential wood stoves. Environmental Science andTechnology 1989;23:470–5.

[59] Eglinton G, Maxwell JR, Philip RP. Organic geochemistry of sedi-ments from contemporary aquatic environments. In: Tissot B, BiennerF, editors. Advances in organic geochemistry 1973, Paris: EditionsTechnip, 1975. p. 941–61.

[60] Gough MA, Rowland SJ. Characterization of unresolved complexmixtures of hydrocarbons in petroleum. Nature 1990;344:648–50.

[61] Killops SD, Al-Juboori MAHA. Characterisation of the unresolvedcomplex mixture (UCM) in the gas chromatograms of biodegradedpetroleums. Organic Geochemistry 1990;15:147–60.

[62] Kolattukudy PE, Croteau R, Buckner JS. Biochemistry of plantwaxes. In: Kolattukudy PE, editor. Chemistry and biochemistry ofnatural waxes, Amsterdam: Elsevier, 1976. p. 289–347.


identification and emission rates of molecular tracers in

Documents