Identification and Emission Factors of Molecular Tracers in Organic

Applied Geochemistry 16 (2001) 1513±1544 www.elsevier.com/locate/apgeochem Identi®cation and emission factors of molecular tracers in organic aerosols from biomass burning Part 1. Temperate climate conifers

Daniel R. Oros 1, Bernd R.T. Simoneit * Environmental and Petroleum Geochemistry Group, College of Oceanic and Atmospheric Sciences, Oregon State University, Corvallis, OR 97331, USA

Received 1 March 2000; accepted 1 September 2000 Editorial handling by R. JaeÂ

Abstract Smoke particulate matter from conifers subjected to controlled burning, both under smoldering and ¯aming conditions, was sampled by high volume air ®ltration on precleaned quartz ®ber ®lters. The ®ltered particles were extracted with dichloromethane and the crude extracts were methylated for separation by thin layer chromatography into hydrocarbon, carbonyl, carboxylic acid ester and polar fractions. Then, the total extract and individual fractions were analyzed by gas chromatography and gas chromatography±mass spectrometry. The major organic components directly emitted in smoke particles were straight chain aliphatic compounds from vegetation wax and diterpenoid acids (biomarkers) from resin. The major natural products altered by combustion included derivatives from phenolic (lignin) and monosaccharide (cellulose) biopolymers and oxygenated and aromatic products from diterpenoids. Other biomarkers present as minor components included phytosterols, both the natural and altered products, and unaltered high molecular weight wax esters. Polycyclic aromatic hydrocarbons (PAH) were also present, however, only as minor constituents. Although the concentrations of organic compounds in smoke aerosols are highly variable and dependent on combustion temperature, the biomarkers and their combustion alteration products are source speci®c. The major components are adsorbed or trapped on particulate matter and thus may be utilized as molecular tracers in the atmosphere for determining fuel type and source contributions from biomass burning. # 2001 Elsevier Science Ltd. All rights reserved.

1. Introduction which in¯uence atmospheric chemical, optical and radiative properties through direct (absorption and The application of biomass burning as a method for scattering of solar and terrestrial radiation) and indirect clearing vegetated (forest, grassland, etc.) areas and for (modi®cation of cloud processes) mechanisms (e.g. domestic heating, cooking, etc. signi®cantly increases IPCC, 1990, 1992). Natural (unaltered) and thermally the input of organic aerosol components to the atmo- altered (pyrolysis) derivative compounds from vegeta- sphere. Biomass burning is an important primary source tion released by biomass burning events can be utilized of soot and organic particulate matter in emissions as speci®c indicators for identifying fuel source inputs, transport mechanisms and receptor fate in samples of atmospheric ®ne particulate matter. The aim of this study is to report the organic chemical * Corresponding author. Tel.: +1-541-737-2155; fax: +1- 541-737-2064. composition of smoke particulate matter emitted by E-mail address: [email protected] (B.R.T. Simoneit). ¯aming and smoldering combustion of fuel from con- 1 Current address: College of Pharmacy, 102 Wegner Hall, ifers (gymnosperms) constituting the predominant spe- Washington State University, Pullman, WA 99164, USA. cies of western North America. In general, each

0883-2927/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved. PII: S0883-2927(01)00021-X 1514 D.R. Oros, B.R.T. Simoneit / Applied Geochemistry 16 (2001) 1513±1544 individual plant species emits a ``chemical ®ngerprint'' containing organic constituents (e.g. PAH and oxy- of natural and thermally altered organic constituents PAH) which have mutagenic and genotoxic potential upon burning, which is source speci®c and unique in (e.g. Arcos and Argus, 1975; IARC, 1989). Considering composition. The incomplete thermal combustion of that conifer wood and associated biomass are major organic natural product precursors results in emission solid fuel sources for heating of homes and cooking (e.g. products which still retain structural characteristics of ®replaces, woodstoves), besides wild®res, it is also the precursor (molecular markers). From these products necessary to identify the components of smoke emis- it is possible to determine precursor/product relation- sions in order to make air quality assessments and to ships and reaction pathways. The dominant directly determine human exposure levels to particle bound emitted and thermally altered molecular markers may organic compounds. be used as speci®c tracers for tracking emissions speci®- cally from conifer (gymnosperm) burning. For example, it has been shown that the burning of conifer biomass 3. Experimental methods from temperate regions yields characteristic tracers from diterpenoids as well as phenolics and other oxygenated 3.1. Sampling species from lignin, which are recognizable in urban airsheds (Simoneit and Mazurek, 1982; Hawthorne et Samples were collected from temperate zone forested al., 1992; Rogge et al., 1993b, 1998; Simoneit et al., areas of California and Oregon, USA and Durango, 1993, 1999; Standley and Simoneit, 1994). Emission Mexico away from urban areas and major roads factors have only been determined for a limited number (Table 1). The branches (1±2 cm diameter), needles (dry of conifer smoke samples (Rogge et al., 1998). Thus, and green), with bleed resin and cones of conifers were more information is necessary for modeling biomass collected from various levels in the canopy of each single burn emissions in air basins or air masses. Furthermore, tree species. These mixtures of wood, bark, needles and it is important to know the organic compound compo- cones when burned more closely resemble wild®res sition of smoke emitted by burning of dominant bio- rather than stove or ®replace burning. All vegetation mass species in order to model mass chemical (reactions, samples were placed in paper bags and allowed to dry kinetics) and physical (radiative heat transfer) behavior over a two week period. Weight measurements were of organic aerosols in the atmosphere and to determine taken before and after burning to determine the total the contribution of regional biomass burning to global mass of plant material consumed. Using a controlled climate change. ®re, vegetation samples were burned completely to the embers under both ¯aming and smoldering conditions. The emitted smoke was collected on an organically 2. Background clean quartz ®ber ®lter (annealed at 550C for 3 h; 95% particle size retention >1.0 mm) using a high volume air The varying temperature and aeration conditions sampler located approximately 1.5 m diagonally above during burning determine the molecular alteration and and to the side of the ¯ames in the smoke plume. Emis- transformation of the organic compounds emitted from sions from burning biomass are primarily ®ne (<2.0 biomass fuel. The heat intensity and the duration of mm) particles (e.g. Schauer et al., 1996; Rogge et al., ¯aming and smoldering conditions determine the dis- 1998), thus no provisions were made to remove coarse tributions and ratios of the natural versus altered com- particles during sampling of these burn tests. Smoke was pounds present in conifer smoke. The primary chemical typically sampled for 5 min periods at a suction ¯ow reactions that occur under ¯aming conditions (tem- rate of 40 ft3/min (1.13 m3/min). After sampling, a por- perature >300C) include pyrolysis, bond cleavage, ®s- tion of each ®lter (8.8 cm2) was cut out and set aside for sion, and tarry and volatile product formation organic C and elemental C analysis (Johnson et al., (Sha®zadeh, 1984). Under smoldering conditions (tem- 1981; Birch and Cary, 1996). The collection ®lters were perature <300C, this occurs at the start of the ®re, i.e. then placed in precleaned 300 ml jars with Te¯on lined ®refront and after ¯aming) organic compounds and lids to which 10 ml of chloroform was added. The jars their altered products are released by a steam stripping/ were then stored at 4C until further chemical extraction vaporization eect, with the extent of this process was conducted. dependent on fuel moisture content. The primary chemical reactions that occur under smoldering conditions 3.2. Extraction and fractionation include depolymerization, water elimination, fragmentation, oxidation, and char formation (Sha®zadeh, Each ®lter was extracted using ultrasonic agitation for 1984). 3Â20 min periods using 200 ml of dichloromethane

Biomass smoke and other source emissions (e.g. pet- (CH2Cl2). The solvent extract was ®ltered using a Gel- roleum, coal) introduce airborne ®ne particulate matter man Swinney ®ltration unit containing an annealed D.R. Oros, B.R.T. Simoneit / Applied Geochemistry 16 (2001) 1513±1544 1515

Table 1 Conifer species sampled for biomass burning in this study

Common name Botanical name Collection region

Apache pine Pinus engelmannii Sierra Madre Occidental, Mexico California redwood Sequoia sempervirens Eureka, CA Douglas ®r Pseudotsuga menziesii Willamette Valley, Corvallis, OR Eastern White pine Pinus strobus Willamette Valley, Corvallis, OR Lodgepole pine Pinus contorta North Tumalo Creek, OR Montezuma pine Pinus montezumae Sierra Madre Occidental, Mexico Mountain hemlock Tsuga mertensiana Willamette National Forest, OR Noble ®r Abies procera Willamette Valley, Philomath, OR Paci®c Silver ®r Abies amabilis Willamette Valley, Corvallis, OR Ponderosa pine Pinus ponderosa Willamette Valley, Corvallis, OR Port Orford cedar Chamaecypris lawsonia Willamette Valley, Corvallis, OR Sitka spruce Picea sitchensis Willamette Valley, Corvallis, OR Western White pine Pinus monticola North Tumalo Creek, OR

glass ®ber ®lter for the removal of insoluble particles 3.3. Instrumental analyses (Simoneit and Mazurek, 1982). The ®ltrate was ®rst concentrated by use of a rotary evaporator and then a The total extract and the fractions were analyzed by stream of ®ltered N2 gas. The ®nal volume was adjusted capillary gas chromatography (GC, Hewlett-Packard to exactly 4.0 ml by addition of CH2Cl2. Aliquots were Model 5890A) with a 30 mÂ0.25 mm i.d. fused silica then taken for derivatization. Alkanoic acid and phe- capillary column coated with DB-5 (J&W Scienti®c, ®lm nolic moieties in the extracts were methylated using thickness 0.25 mm) which was temperature programmed diazomethane in diethyl ether prepared from the pre- as follows: hold at 65C for 2 min, ramp to 300Cat6C/ cursor N-methyl-N0-nitro-N-nitrosoguanidine (Pierce min, hold isothermal at 300C for 20 min. All samples Chemical Co.) (Schlenk and Gellerman, 1960). were analyzed by capillary gas chromatography-mass The methylated extracts were separated by pre- spectrometry (GC±MS) using a Hewlett-Packard Model parative thin layer chromatography (TLC) on silica gel 5973 MSD quadrupole mass spectrometer operated in the plates (Analtech, Inc.) with a mobile phase eluent mix- electron impact mode at 70 eV and coupled to a Hewlett- ture of hexane:diethyl ether (9:1) (Simoneit and Packard Model 6890 gas chromatograph. The GC was Mazurek, 1982). This procedure allows for determina- equipped with a 30 mÂ0.25 mm i.d. fused silica capillary tion of chemical information on single molecular groups column coated with DB-5 (J&W Scienti®c, ®lm thickness or functional group series, which may not be detected 0.25 mm) and operated using the same temperature pro- due to coelution in the total extract mixture. The 4 gram as described above, with He as carrier gas. fractions removed from the TLC plates contained the following classes of compounds: (1) n-alkanes, n- 3.4. Compound identi®cation and quantitation alkenes, and saturated and unsaturated cyclic di- and triterpenoid hydrocarbons; (2) n-alkanones, n-alkanals Compound identi®cations are based on comparisons and polycyclic aromatic hydrocarbons; (3) n-alkanoic with authentic standards, GC retention time, literature acids (as methyl esters) and saturated and unsaturated mass spectra and interpretation of mass spectrometric di- and triterpenoid ketones and acids; and (4) n-alka- fragmentation patterns. Quantitation of the homo- nols, terpenols and polar organics. The fourth fraction logous compound series was conducted by comparison and the total extract were converted prior to analysis to of the GC peak area with that of a co-injected known trimethylsilyl derivatives by reaction with N,O-bis-(tri- standard (e.g. perdeuterated tetracosane, n-C24D50). methylsilyl)-tri¯uoroacetamide (BSTFA) plus 1% tri- methylchlorosilane for approximately 3 h at 70C. Compound quantitation for TLC was determined from 4. Results and discussion the volume of methylated total extract that was applied to each TLC plate. Brie¯y, after separation, The major organic components identi®ed in the solu- extraction, ®ltration and concentration of each frac- ble lipid fraction of the conifer smoke samples and their tion (F1-F4), each was then dissolved in a solvent emission factors (mg/kg of conifer fuel burned) are given volume equivalent to the initial TLC plate loading in Table 2. These emission factors are preliminary in volume. nature since only one burning test was conducted for 1516 Table 2 Emission factors (mg/kg of conifer fuel burned) of the major organic constituents in conifer smokea

I. Homologous series n-Alkanes (Natural products) n-Tetradecane C H 198 224 A

14 30 1513±1544 (2001) 16 Geochemistry Applied / Simoneit B.R.T. Oros, D.R. n-Pentadecane C15H32 212 1033 699 104 A n-Hexadecane C16H34 226 67 586 2305 29 156 196 120 934 778 723 A n-Heptadecane C17H36 240 112 56 4368 13 86 113 26 143 90 88 192 A n-Octadecane C18H38 254 134 59 7997 71 74 32 156 67 244 196 157 472 A n-Nonadecane C19H40 268 251 314 7653 108 383 40 384 94 699 180 509 184 1496 A n-Eicosane C20H42 282 4397 927 18391 134 775 43 446 77 1753 1477 354 497 1408 A n-Heneicosane C21H44 296 688 516 11558 338 720 77 982 315 1422 204 1050 8624 1934 A n-Docosane C22H46 310 861 641 12705 277 660 114 743 185 1045 385 795 798 710 A n-Tricosane C23H48 324 787 680 3853 271 787 403 1306 201 700 577 866 1310 718 A n-Tetracosane C24H50 338 417 1116 1706 432 1140 82 1421 63 335 386 844 529 846 A n-Pentacosane C25H52 352 287 488 408 185 303 277 636 93 238 331 634 476 655 A n-Hexacosane C26H54 366 363 246 650 97 159 77 32 209 64 273 427 155 611 A n-Heptacosane C27H56 380 409 277 1856 212 137 167 412 927 1780 127 945 746 196 A n-Octacosane C28H58 394 144 279 1751 97 707 52 94 307 399 110 393 176 81 A n-Nonacosane C29H60 408 265 477 2713 239 110 138 266 441 833 81 1253 560 116 A n-Triacontane C30H62 422 870 1157 1882 108 669 49 586 393 179 377 706 1327 A n-Hentriacontane C31H64 436 3713 3350 391 1307 72 37 571 483 2866 262 111 A n-Dotriacontane C32H66 450 48 35 44 646 739 310 A n-Tritriacontane C33H68 464 49 118 1629 13 9157 A n-Tetratriacontane C34H70 478 45 465 A Total alkanes 13906 11168 80830 3156 10502 1679 7224 4808 11390 4310 22467 16371 11595 CPI 0.9 1.2 0.7 1.5 1.4 2.4 1.3 1.2 1.2 0.5 3.4 3.1 0.9

Cmax 20 31 20 24 33 23 24 27 27 20 33 21 21 n-Alkenes (Alteration products) n-Tridec-1-ene C13H26 182 445 S n-Tetradec-1-ene C14H28 196 1473 638 556 S n-Pentadec-1-ene C15H30 210 1880 S n-Hexadec-1-ene C16H32 224 782 508 4951 148 100 713 129 413 S n-Heptadec-1-ene C17H34 238 893 61 8648 121 38 122 90 392 359 S n-Octadec-1-ene C18H36 252 482 385 17981 338 33 476 135 1053 1454 510 555 A n-Nonadec-1-ene C19H38 266 671 417 24158 607 35 840 206 1337 83 1066 190 796 S

(continued on next page) Table 2 (continued)

Compound name Composition M.W. Apache California Douglas Eastern Lodgepole Montezuma Mountain Noble Paci®c Ponderosa Port Sitka Western ID pine redwood ®r White pine pine hemlock ®r Silver pine Orford spruce White basisb pine ®r cedar pine n-Eicos-1-ene C20H40 280 923 142 25723 36 1139 39 1111 255 1753 208 453 A n-Heneicos-1-ene C21H42 294 692 870 73067 675 38 1169 192 1474 438 1047 443 32945 S n-Docos-1-ene C22H44 308 1583 1459 8450 1972 114 1667 709 5534 1010 1617 1342 2431 S n-Tricos-1-ene C23H46 322 614 426 24109 263 31 44 134 517 258 727 668 494 S n-Tetracos-1-ene C H 336 1566 484 3166 192 746 332 1244 262 1313 605 599 516 958 S 24 48 1513±1544 (2001) 16 Geochemistry Applied / Simoneit B.R.T. Oros, D.R. n-Pentacos-1-ene C25H50 350 112 469 408 163 35 201 197 357 902 499 2216 S n-Hexacos-1-ene C26H52 364 363 1547 89 79 97 513 134 611 S n-Heptacos-1-ene C27H54 378 409 703 128 203 479 196 S n-Octacos-1-ene C28H56 392 144 210 455 81 S n-Nonacos-1-ene C29H28 406 134 S Total alkenes 9234 5220 196709 229 6599 696 7256 2179 14541 2602 9857 5440 41697 CPI 0.6 0.8 2.1 nd 0.4 0.3 0.6 0.6 0.4 0.4 0.9 0.5 7.3

Cmax 22 22 21 24 22 24 22 22 22 22 22 22 21 Isoprenoids (Alteration product)

Neophytadiene C28 û56 278 452 704 I Carboxylic acids (Natural products) n-Heptanoic acid C7H14 O2 130 293 S n-Octanoic acid C8H16 O2 144 2904 717 2264 1690 2167 3313 538 473 S n-Nonanoic acid C9H18 O2 158 2931 1248 1092 3278 1620 2754 4115 511 S n-Decanoic acid C10H20 O2 172 3028 947 209 2725 232 1775 1505 3205 586 S n-Undecanoic acid C11H22 O2 186 1579 1060 763 1317 395 889 1251 74 1778 1445 S n-Dodecanoic acid C12H24 O2 200 25856 14213 1337 19445 12521 16194 4985 345 5897 1578 4186 S n-Tridecanoic acid C13H26 O2 214 1986 1980 1846 1296 418 778 1379 157 715 614 S n-Tetradecanoic acid C14H28 O2 228 31530 26817 3283 29728 10462 11498 602 71 8346 748 16240 8859 5252 S n-Pentadecanoic acid C15H30 O2 242 9331 868 10110 6362 1757 3945 1384 66 7480 1095 52940 2890 1785 S n-Hexadecanoic acid C16H32 O2 256 102702 41515 12040 51891 27103 19701 1773 699 39672 4503 57036 39278 32165 A n-Heptadecanoic acid C17H34 O2 270 7222 3033 5605 9465 1219 2631 4297 115 3007 800 5411 2491 1821 S n-Octadecanoic acid C18H36 O2 284 26167 6940 5098 48436 7284 7659 4991 316 17814 2728 8303 12766 18372 A n-Nonadecanoic acid C19H38 O2 298 21778 7750 727 20680 6623 827 137 6086 3005 S n-Eicosanoic acid C20H40 O2 312 83678 16842 13656 170310 14261 26293 8479 387 23511 10161 18324 17892 29838 S n-Heneicosanoic acid C21H42 O2 326 3831 1688 2579 22901 846 4693 285 81 12799 1390 3209 30749 841 S n-Docosanoic acid C22H44 O2 340 30243 28652 37705 65029 13468 17720 34397 895 41689 13004 43635 45229 12015 S n-Tricosanoic acid C23H46 O2 354 8257 3664 2640 14190 1339 4302 290 110 5884 3465 5921 4484 1968 S n-Tetracosanoic acid C24H48 O2 368 20002 12056 22822 31058 7294 18283 28638 401 25663 10639 33639 26668 1822 S n-Pentacosanoic acid C25H50 O2 382 1053 782 3405 372 1367 523 227 1533 2511 1403 407 S 1517

(continued on next page) 1518 Table 2 (continued)

Compound name Composition M.W. Apache California Douglas Eastern Lodgepole Montezuma Mountain Noble Paci®c Ponderosa Port Sitka Western ID pine redwood ®r White pine pine hemlock ®r Silver pine Orford spruce White basisb pine ®r cedar pine n-Hexacosanoic acid C26H52 O2 396 1863 1801 1845 2772 696 4226 4362 307 5135 1856 4572 2066 1089 S n-Heptacosanoic acid C27H54 O2 410 441 331 218 161 430 516 341 S n-Octacosanoic acid C28H56 O2 424 1846 430 664 1286 82 172 2596 35 4371 760 S n-Nonacosanoic acid C29H58 O2 438 834 121 170 1126 3254 S n-Triacontanoic acid C H O 452 1851 1404 3109 793 4725 527 S 30 60 2 1513±1544 (2001) 16 Geochemistry Applied / Simoneit B.R.T. Oros, D.R. n-Hentriacontanoic acid C31H62 O2 466 726 331 3472 S n-Dotriacontanoic acid C32H64 O2 480 5818 488 9410 729 1608 552 S n-Tritriacontanoic acid C33H66 O2 494 1012 S n-Tetratriacontanoic acid C34H68 O2 508 2830 S n-Octadecadienoic acid C18H32 O2 280 3517 17116 186 399 S n-Octadecenoic acid C18H34 O2 282 45835 385 17981 111978 35673 476 1053 1454 510 555 S a,o-Nonanedioic acid C9H16 O4 188 2677 421 S a,o-Undecanedioic acid C11H20 O4 216 331 4073 S a,o-Tridecanedioic acid C13H24 O4 244 6526 464 149 S a,o-Tetradecanedioic acid C14H26 O4 258 197 S a,o-Pentadecenedioic acid C15H26 O4 270 727 9775 S a,o-Pentadecanedioic acid C15H28 O4 272 1287 209 S a,o-Hexadecenedioic acid C16H28 O4 284 683 S a,o-Hexadecanedioic acid C16H30 O4 286 276 7970 5665 1141 A a,o-Heptadecanedioic acid C17H32 O4 300 2865 278 S a,o-Octadecenedioic acid C18H32 O4 312 2469 3231 S a,o-Octadecanedioic acid C18H34 O4 314 4902 1273 1565 S a,o-Nonadecanedioic acid C19H36 O4 328 3732 352 1790 S a,o-Eicosanedioic acid C20H38 O4 342 2460 152 573 1172 211 S a,o-Heneicosanedioic acid C21H40 O4 356 1388 176 S a,o-Docosanedioic acid C22H42 O4 370 321 378 67 S a,o-Tricosenedioic acid C23H42 O4 382 1366 S a,o-Tricosanedioic acid C23H44 O4 384 11897 S a,o-Pentacosenedioic acid C25H46 O4 410 628 S a,o-Pentacosanedioic acid C25H48 O4 412 985 419 S a,o-Heptacosenedioic acid C27H50 O4 438 136 S a,o-Heptacosanedioic acid C27H52 O4 440 4627 9305 S a,o-Nonacosanedioic acid C29H56 O4 468 1200 9305 S 7-Phenylheptanoic acid C13H18 O2 206 1589 9305 I 8-Phenyloctanoic acid C14H20 O2 220 1287 9305 I o-Methoxystearic acid C19H36 O3 312 6994 I o-Methoxy-10- C19H38 O4 330 1279 12129 I hydroxystearic acid

(continued on next page) Table 2 (continued)

11,18-Dimethoxy-10- C20H40 O5 360 5858 I hydroxystearic acid

Total carboxylic acids 437248 189783 179300 633032 156680 153988 95598 6169 253023 51279 281412 198678 116825 CPI 6.1 7.9 4.1 5.4 15.6 5.0 11.0 3.4 4.0 6.4 2.5 3.6 14.2 ..Oo,BRT ioet/ApidGohmsr 6(01 1513±1544 (2001) 16 Geochemistry Applied / Simoneit B.R.T. Oros, D.R. Cmax 16 16 22 20 16 20 22 22 22 22 16 22 16 n-Alkanones (Natural products) n-Hexadecan-2-one C16H32 O 240 59 2530 S n-Heptadecan-2-one C17H34 O 254 24 2187 S n-Octadecan-2-one C18H36 O 268 1920 194 202 1474 S n-Nonadecan-2-one C19H38 O 282 371 818 225 1031 S n-Eicosan-2-one C20H40 O 296 88 630 113 1326 S n-Heneicosan-2-one C21H42 O 310 961 386 1396 183 3212 S n-Docosan-2-one C22H44 O 324 1000 628 113 1396 S n-Tricosan-2-one C23H46 O 338 5660 298 484 160 210 232 1675 2950 S n-Tetracosan-2-one C24H48 O 352 14793 297 94 177 105 117 715 1605 S n-Pentacosan-2-one C25H50 O 366 1090 193 74 160 358 444 842 606 805 S n-Hexacosan-2-one C26H52 O 380 149 28 710 105 1735 S n-Heptacosan-2-one C27H54 O 394 387 54 164 174 825 416 5334 S n-Octacosan-2-one C28H56 O 408 267 108 928 S n-Nonacosan-2-one C29H58 O 422 319 134 238 383 3457 S n-Triacontan-2-one C30H60 O 436 847 S n-Hentriacontan-2-one C31H62 O 450 97 206 328 2991 S n-Dotriacontan-2-one C32H64 O 464 401 S n-Tritriacontan-2-one C33H66 O 478 234 S 6,10,14-Trimethyl-2- C18H36 O 268 473 2837 S pentadecan-2-one n-Nonacosan-10-one C29H58 O 422 777 73 868 625 S Total Alkanones 24712 3040 3705 2774 1419 5314 2908 2685 32969 3755 CPI 0.4 3.2 nd nd 1.0 3.5 1.3 4.2 2.8 nd 1.7 nd nd

Cmax 24 21 nd nd 22 24 21 27 25 nd 27 23 nd n-Alkanols (Natural products) n-Octadecanol C18H38 O 270 333 A n-Nonadecanol C19H40 O 284 415 A n-Eicosanol C20H42 O 298 634 A n-Heneicosanol C21H44 O 312 316 A 1519 (continued on next page) 1520 Table 2 (continued)

Cmax 24 24 22 22 22 24 24 22 22 22 29 22 22 Wax esters (Natural products)

Nonyl dodecanoate C21 H42O2 326 357 S Decyl dodecanoate C22 H44O2 340 1232 S Undecyl C23 H46O2 354 937 S dodecanoate

Dodecadienyl C24 H44O2 364 2074 S dodecanoate

Dodecyl C24 H48O2 368 851 415 S dodecanoate

Tridecyl C25 H50O2 382 1012 1509 365 S dodecanoate

Tetradecadienyl C26 H48O2 392 1749 S dodecanoate

Tetradecenyl C26 H50O2 394 5314 S dodecanoate

Tetradecyl C26 H52O2 396 4607 S dodecanoate

Pentadecyl C27 H54O2 410 1412 S dodecanoate

Dodecenyl Tetradecanoate C26 H50O2 394 3350 S Dodecyl tetradecanoate C26 H52O2 396 615 3435 924 S Tridecyl tetradecanoate C27 H54O2 410 488 S Hexadecenyl dodecanoate C28 H54O2 422 3250 S Hexadecyl dodecanoate C28 H56O2 424 1333 S (continued on next page) Table 2 (continued)

Tetradecyl tetradecanoate C28 H56 O2 424 1513 5243 1026 724 S Octadecyl tetradecanoate C32 H64 O2 480 604 S Hexadecyl hexadecanoate C32 H64 O2 480 645 S Heneicosanyl dodecanoate C33 H66 O2 494 3164 S

Total wax esters 12439 10373 14234 4876 3083 365 2136 645 1513±1544 (2001) 16 Geochemistry Applied / Simoneit B.R.T. Oros, D.R.

II. Biomarkers

Sesquiterpenoids (C15) (Natural products)

5-Hydroxycalamanene C15 H22O 218 14606 I a-Calacorene C15 H24O 200 1446 I Aromadendrol C15 H24O 220 694 I b-Oplopenone C15 H24O 220 16900 I 6-Deoxygeigerin C15 H20 O3 248 16836 I (Alteration products)

Cadalene C15H18 198 921 1263 I Calamenene C15H22 202 1317 3460 I

Diterpenoids (C20 ) (Natural products)

Dehydroabietane C20H30 270 7971 2714 607 2872 228 160 2899 1023 A Hibaene C20H32 272 183 28482 I Isophyllocladene C20H32 272 645 I Isopimaradiene C20H32 272 3309 I Laurene C20H32 272 372 468 I 5b-Podocarpa-8,11,13- C18 H24 O2 272 9362 2805 I trien-16-oic acid

Rimuene C20H32 272 947 4798 789 I Manoyl oxide C20 H34O 290 488 3467 14048 I Totarol C20 H30O 286 57197 I Abietol C20 H32O 288 4390 3773 I Abietic acid C20 H30 O2 302 17488 24338 8524 13178 2885 26141 13127 6456 3651 A iso-Pimaric acid C20 H30 O2 302 124568 3387 1649 59988 7983 6143 1388 7627 3916 12399 19284 A Palustric acid C20 H30 O2 302 18956 6044 7407 8018 2400 16365 872 4673 A Pimaric acid C20 H30 O2 302 17284 22722 7637 6116 2990 A Sandaracopimaric acid C20 H30 O2 302 2550 4266 4807 695 3955 5712 A Daniellic acid C20 H28 O3 316 26314 113 3313 I Polyaltic acid C20 H28 O3 316 10773 I

Copalic acid C20 H34 O2 318 274 I 1521 (continued on next page) 1522 Table 2 (continued)

Agatholic acid C20H34 O3 322 596 3331 9790 Pinifolic acid C22H36 O4 364 267 402 1561 I (Alteration products)

Bisnorsimonellite C17H20 224 2637 I ..Oo,BRT ioet/ApidGohmsr 6(01 1513±1544 (2001) 16 Geochemistry Applied / Simoneit B.R.T. Oros, D.R. Retene C18H18 234 25003 274 6510 25339 4165 25043 573 963 3842 7797 2634 34717 A Dihydroretene C18H20 236 2105 2554 6024 90 931 1376 5050 I Tetrahydroretene C18H22 238 5044 5808 1500 9891 1674 2493 16133 I Methylretene C19H20 248 194 801 I 7-Isopropyl-1,2,3,4- C18H20 O 252 1209 I tetrahydrophenanthrene- 1-aldehyde

18-Norabieta-4,6,8,11,13- C19H24 252 2077 1768 I pentaene

Simonellite C19H24 252 2333 2268 446 4737 802 392 584 865 2411 I Dihydrosimonellite C19H26 254 548 6620 I 18-Norabieta- C19H26 254 14379 1358 7731 I 2,8,11,13-tetraene

18-Norabieta- C19H26 254 1761 194 231 I 3,8,11,13-tetraene

18-Norabieta- C19H26 254 4397 4485 1477 14267 I 4,8,11,13-tetraene

18-Norabieta-4(19), C19H26 254 4000 I 8,11,13-tetraene

19-Norabieta-4(18), C19H26 254 1403 1276 1280 291 12614 I 8,11,13-tetraene

19- or 18-Norabieta- C19H26 254 554 I 6,8,11,13-tetraene

Dehydroabietin C19H28 256 11338 3307 804 3540 1264 684 1053 27770 A 18-Norabieta- C19H28 256 1329 6551 1823 574 5644 263 2857 317 9986 I 8,11,13-triene

Deisopropyldehydro- C17H22 O2 258 1807 7516 1179 1076 5073 I abietic acid

7-Oxo-18-norabieta- C19H26 O 270 I 8,11,13-triene

7-Oxo-19-norabieta- C19H26 O 270 385 I 8,11,13-triene

16,17-Bisnordehydro- C18H24 O2 272 902 1543 460 I abietic acid

(continued on next page) Table 2 (continued)

Pimara-8(9),15-diene C20H32 272 125 4474 I Abieta-8,11,13,15- C20 H26O 282 645 I tetraen-18-al

3-Oxo-12-hydroxy- C19 H22 O2 282 7011 I simonellite ..Oo,BRT ioet/ApidGohmsr 6(01 1513±1544 (2001) 16 Geochemistry Applied / Simoneit B.R.T. Oros, D.R. Abieta-8,11,13- C20 H28O 284 563 I triene-7-one

Dehydroabietal C20 H28O 284 6908 5813 984 1741 13596 8956 A 6,7-Dehydroferruginol C20 H28O 284 1564 19396 I Dehydroabietol C20 H30O 286 1097 I 3-Oxo-16,17- C18 H22 O3 286 121 I bisnordehydroabietic acid

7-Oxo-16,17- C18 H22 O3 286 6761 I bisnordehydroabietic acid

13-Oxopodocarp- C18 H26 O3 290 674 1419 3850 I 8(14)-en-18-oic acid

1-Methyl-7-isopropyl- C19 H36 O2 296 79153 1371 2510 15162 786 186 12799 7873 30749 155 I 1,2,3,4-tetrahydrophenanthrene-1-carboxylic acid

Abieta-6,8,11,13,15- C20 H24 O2 296 9521 1764 457 4012 I pentaen-18-oic acid

13-Isopropenyl-5a- C20 H26 O2 296 1497 I podocarpa-6,8,11,13- tetraen-16-oic acid

13-Isopropyl-5a- C20 H26 O2 298 1474 911 2360 587 5386 731 774 8002 435 I podocarpa-6,8,11,13- tetraen-16-oic acid

Abieta-6,8,11,13- C20 H26 O2 298 6298 39070 4802 5398 9527 8777 A tetraen-18-oic acid

Abieta-8,11,13,15- C20 H26 O2 298 18874 6937 39844 3173 9465 3121 3275 609 2892 A tetraen-18-oic acid

Abieta-7,13,15- C20 H28 O2 300 974 I trien-18-oic acid

Abieta-8,11,13- C20 H28 O2 300 6552 1544 17892 I trien-18-oic acid

Dehydroabietic acid C20 H28 O2 300 137919 4054 14891 57798 12764 28689 5476 6732 24103 19508 972 26019 24384 A 8,11,15-Isopimaratrien- C20 H28 O2 300 23723 188 I 18-oic acid 1523

(continued on next page) 1524 Table 2 (continued)

13-Isopropyl-5a- C20 H28O2 300 1474 5235 195 617 523 7265 102 I podocarpa-8,11,13- trien-16-oic acid

6,8,15-Pimaratrien- C20 H28O2 300 41741 I 18-oic acid ..Oo,BRT ioet/ApidGohmsr 6(01 1513±1544 (2001) 16 Geochemistry Applied / Simoneit B.R.T. Oros, D.R. 13a-Abieta-7,9(11)- C20 H30O2 302 12820 I dien-18-oic acid

Abieta-8,13(15)- C20 H30O2 302 23335 6937 29000 4304 10808 2270 5301 1014 I dien-18-oic acid

3- or 7-Hydroxyabietal C20 H30O2 302 49582 I 8,15-Isopimaradien- C20 H30O2 302 69603 87918 4646 32772 16903 I 18-oic acid

8,15-Pimaradien- C20 H30O2 302 84439 15110 29055 10158 27103 496 4646 54743 11821 162206 A 18-oic acid

10a(H)-9,10- C20 H30O2 302 682 5637 6767 I Secodehydroabietic acid

10b(H)-9,10- C20 H30O2 302 690 8755 17972 7518 1003 5639 7382 5572 I Secodehydroabietic acid

16-Norisocopalan-15-oic acid C20 H34O2 306 3620 I 7-Oxoabieta-5,8,11,13- C20 H24O3 312 27716 13080 1473 339 85 6176 1705 I tetraen-18-oic acid

3- or 7-Oxoabieta-8,11,13,15- C20 H24O3 312 11966 2037 384 I tetraen-18-oic acid

3- or 7-Oxoabieta-8,11,13- C20 H24O3 314 3322 2456 2142 4312 I trien-18-oic acid

3- or 7-Hydroxyabieta- C20 H24O3 314 18318 I 8,11,13,15-tetraen- 18-oic acid

Methyl dehydroabietate C21 H30O2 314 36110 1287 6623 12667 25716 I 3-Oxodehydroabietic acid C20 H26O3 314 2161 2750 I 7-Oxodehydroabietic acid C20 H26O3 314 24301 980 10335 1223 6639 457 696 11824 3180 1114 A 12- or 14-Hydroxy- C20 H28O3 316 2905 696 I dehydroabietic acid

15-Hydroxydehydroabietic C20 H28O3 316 1145 1489 A acid 0 Methyl-13-(2 -oxopropyl)- C20 H28O3 316 3478 I podocarpa-8,11,13- trien-15-oic acid

3-Oxoabietic acid C20 H28O3 316 2695 I (continued on next page) Table 2 (continued)

3- or 7-Oxopimaric acid C20 H28O3 316 I Agatholic acid C20 H34O3 322 596 3331 9790 I Acetyldihydroabietol C22 H36O2 332 1928 I Propylabieta-8,11,13,15- C23 H32O2 340 382 I tetraenoate ..Oo,BRT ioet/ApidGohmsr 6(01 1513±1544 (2001) 16 Geochemistry Applied / Simoneit B.R.T. Oros, D.R. 3-Oxopinifolic acid C20 H30O5 350 2060 298 I 3- or 7-Acetoxyabietic acid C22 H32O4 360 7730 409 I Dimethyl dihydroagathate C22 H36O4 364 596 88 I Succinyl7- C24 H32O5 400 923 144 1004 I oxodehydroabietol

Total Diterpenoids 733880 46367 148413 619352 116951 259767 18981 23538 133145 238205 127543 193502 433517

Triterpenoids (Natural products)

24,25-Dinorlupa-1,3,5(10)- C28 H42 378 3491 1060 1718 2659 I triene

3a-Methoxyfriedelene C31H52 O 440 1060 I 3b-Methoxyfriedelene C31H52 O 440 1676 I 3a-Ethoxyfriedelene C32H56 O 456 2506 I 3b-Ethoxyfriedelene C32H56 O 456 743 I Steroids (Natural products)

Campesterol C28H48 O 400 180 3 84 148 98 95 8 173 184 I Stigmasterol C29H48 O 412 85 I b-Sitosterol C29H50 O 414 2253 421 568 2310 849 1284 282 1876 1282 141 1407 1300 66 I (Alteration products)

5-Pregnene C21 H34 286 3163 3633 I 7-Pregnene C21 H34 286 4907 4251 I 24-Ethyl-19-norcholesta- C28 H38 374 651 433 205 465 518 359 I 1,3,5(10),6,8,14-hexaene

24-Ethyl-19-norcholesta- C28 H40 376 367 1693 311 295 377 561 119 488 657 I 1,3,5(10),6,8-pentaene

24-Ethyl-19-norcholesta- C28 H42 378 3713 891 2870 I 1,3,5(10),8-tetraene

24-Methyl-19-norcholesta- C28 H44 380 56 I 1,3,5(10)-triene 1525

(continued on next page) 1526 Table 2 (continued)

24-Ethyl-14a(H)- C28H44 380 1228 I 1(10-6)-abeo-19- norcholesta-5,7,9-triene

24-Ethyl-14b(H)- C28H44 380 1566 1157 I 1(10-6)-abeo-19- ..Oo,BRT ioet/ApidGohmsr 6(01 1513±1544 (2001) 16 Geochemistry Applied / Simoneit B.R.T. Oros, D.R. norcholesta-5,7,9-triene

24-Ethyl-1-methyl- C29H44 392 511 I 19-norcholesta-5,7,9,14- tetraene

24-Ethyl-1-methyl- C29H46 394 1908 431 I 19-norcholesta-5,7,9- triene

24-Ethyl-4-methyl- C29H46 394 145 327 425 537 500 140 1126 I 19-norcholesta-1,3,5(10)- triene

24-Ethyl-4-methyl- C29H46 394 547 550 306 486 I 19-norcholesta-1,3,5(10)- triene (isomer)

24-Ethyl-4-methyl- C29H46 394 1908 2152 1006 I 19-norcholesta-1,3,5(10)- triene (isomer)

24-Ethyl-14b(H)- C29H46 394 1590 1276 I 1(10-6)-abeo-cholesta- 5,7,9-triene

24-Ethylcholesta-2,4-diene C29H46 396 870 I 24-Ethylcholesta-4,22-diene C29H46 396 3721 I Stigmasta-3,5-diene C29H48 396 2114 298 785 1562 1757 1879 122 782 I 5a(H)-24-Ethylcholest-2-ene C29H48 398 221 I Stigmast-4-ene C29H50 398 371 1413 379 722 864 I Stigmast-5-ene C29H50 398 187 1319 578 670 I Stigmasta-3,5- C29 H46O 410 576 268 258 I dien-7-one

Stigmasta-4,6- C29 H46O 410 1283 I dien-3-one

Stigmast-4-en-3-one C29 H48O 412 496 268 I Total Steroids 17315 1271 8014 6660 12097 1379 2617 10061 4414 8671 6237 4133 6792

III. Polycyclic aromatic hydrocarbons (PAH) (Alteration products) (continued on next page) Table 2 (continued)

Phenanthrene C14H10 178 2073 639 1479 4197 1477 7687 7861 396 7475 743 3284 715 2001 A Anthracene C14H10 178 207 64 315 606 714 2917 56 90 1330 140 647 183 A 4(H)-Cyclopenta[def] C15H10 190 512 88 2430 336 428 A phenanthrene 9-Methylanthracene C H 192 83 47 96 167 57 557 326 76 1095 35 122 19 67 A 15 12 1513±1544 (2001) 16 Geochemistry Applied / Simoneit B.R.T. Oros, D.R. 1-Methylphenanthrene C15H12 192 1714 113 672 3490 739 7920 1895 232 2458 951 421 346 2864 A 2-Methylphenanthrene C15H12 192 968 488 480 1503 568 3506 3440 165 2929 352 547 124 1362 A 3-Methylphenanthrene C15H12 192 308 314 259 761 205 1897 1027 71 6697 381 261 41 339 A 9-Methylphenanthrene C15H12 192 103 744 396 297 164 1450 3930 83 3285 124 498 41 270 A C1-Anthracenes/ C15H12 192 3176 1705 1903 6219 1733 15331 10619 627 16463 1843 1848 570 4902 A phenanthrenes

Fluoranthene C16H10 202 1728 1535 6467 6633 669 6742 1194 1750 3240 4493 A Pyrene C16H10 202 1071 865 1009 1480 1861 3191 521 4318 6370 426 3394 A 2-Phenylnaphthalene C16H12 204 1148 932 A C2-Anthracenes/ C16H14 206 5507 1568 2247 6168 1084 7608 194 365 2318 2176 16710 A phenanthrenes

11(H)-Benzo[a]¯uorene C17H12 216 355 26104 922 1188 A C1-Pyrenes C17H12 216 493 1300 2524 381 7963 A C3-Anthracenes/ C17H16 220 9829 1946 8486 1681 5252 215 744 3033 1301 11397 S phenanthrenes

Benzo[ghi]¯uoranthene C18H10 226 305 639 1919 A Cyclopenta[cd]pyrene C18H10 226 1979 429 1059 1919 S Benz[a]anthracene C18H12 228 127 A Chrysene/triphenylene C18H12 228 179 531 581 204 251 1724 566 1210 A C1-Chrysenes C19H14 242 263 48 A Benzo[b/k]¯uoranthene C20H12 252 582 364 A Benzo[a]pyrene C20H12 252 376 38 253 578 1045 2321 A Benzo[e]pyrene C20H12 252 803 402 A Perylene C20H12 252 426 A Anthanthrene C22H12 276 290 A Benzo[ghi]perylene C22H12 276 81 190 A Indeno[1,2,3-cd]pyrene C22H12 276 499 515 A Total PAH 25039 10663 14193 32406 42239 63094 44561 5063 62461 19918 9378 8998 69961

IV. Phenols (lignin pyrolysis) (Natural products)

Catechol C6H6O2 110 38264 5182 9586 20885 6372 10898 4101 5252 32152 1381 27840 13169 2047 A Resorcinol C6H6O2 110 4317 8291 A

Cinnamic acid C9H8O2 148 560 354 A 1527 (continued on next page) 1528 Table 2 (continued)

Compound name Composition M.W. Apache California Douglas Eastern Lodgepole Montezuma Mountain Noble Paci®c Ponderosa Port Sitka Western ID pine redwood ®r White pine pine hemlock ®r Silver pine Orford spruce White basisb pine ®r cedar pine p-Coumaric acid C9H8O3 164 7218 430 5585 A Eugenol C10H12 O2 164 17034 30522 973 17285 2294 13252 A (Alteration products)

Hydroquinone C6H6O2 110 4769 5163 25729 7540 15711 I ..Oo,BRT ioet/ApidGohmsr 6(01 1513±1544 (2001) 16 Geochemistry Applied / Simoneit B.R.T. Oros, D.R. Benzoic acid C7H6O2 122 4206 698 354 I Dihydroxytoluene C7H8O2 124 17570 S Guaiacol C7H8O2 124 324 4674 I 4-Hydroxyphenylethanol C8H10 O2 138 2043 I Dihydrocinnamol C9H12 O2 152 7615 I 3-Hydroxybenzoic acid C8H8O3 152 1614 2405 I 4-Hydroxybenzoic acid C8H8O3 152 2618 4345 5807 4425 3559 23206 40 2968 I Vanillin C8H8O3 152 9897 1559 3239 11498 2546 9641 405 7015 424 7196 2802 279 I Vanillyl alcohol C8H10 O3 154 224 I Acetovanillone C9H10 O3 166 1220 3441 2251 6718 5411 511 5587 292 I Vanillic acid C8H8O4 168 6707 1126 4307 8351 3609 5706 1783 7833 790 5125 2194 477 I Pyrogallol C9H14 O3 170 19044 1246 2455 17808 4192 4296 3872 12883 5861 2472 48 I Guaiacylpropenal C10H10 O3 178 12253 3207 1098 1174 7106 1063 8550 7413 I Coniferyl alcohol C10H12 O3 180 330 8797 I Guaiacylacetone C10H12 O3 180 10064 2007 3860 10881 10885 888 10137 521 8108 262 I 4-Methoxy-3- C9H10 O4 182 275 I hydroxybenzoic acid

Homovanillyl alcohol C10H14 O3 182 1266 6644 34634 9791 12886 861 14053 370 10376 9700 177 A 3-Vanillylpropanol C10H14 O3 182 12139 15990 13364 2840 I 3-Methylcatechol C10H16 O2 196 2839 7236 4151 14674 2624 13810 17024 I 4-Methylcatechol C10H16 O2 196 22952 2961 2621 1359 1250 I Homovanillic acid C10H12 O4 196 7083 999 4053 8087 3514 4641 1139 6400 1067 3870 2621 951 A 0 3,3 -Dimethoxy- C16H16 O4 272 11728 1854 3644 17183 4664 11996 559 2636 6331 532 6899 7394 I 4,40-dihydroxystilbene

Divanillyl C16H18 O4 274 18874 601 1331 4679 465 I 1,2-Divanillylethane C18H22 O4 302 133 I Tetrahydro-3,4- C20H24 O5 344 7597 1191 2546 7400 5227 3474 575 1376 3951 308 4384 3378 419 I divanillylfuran

Pinoresinol C20H22 O6 358 I Total phenolics 199678 32006 62049 231919 66285 134097 20852 54710 179475 8258 124070 54577 4952

V. Monosaccharides (cellulose pyrolysis)

(Alteration products)

Galactosan C6H10 O5 162 14537 5384 3707 21005 3058 12148 107 4958 13385 214 8850 1804 229 A (continued on next page) Table 2 (continued)

Compound name Composition M.W. Apache California Douglas Eastern Lodgepole Montezuma Mountain Noble Paci®c Ponderosa Port Sitka Western ID pine redwood ®r White pine pine hemlock ®r Silver pine Orford spruce White basisb pine ®r cedar pine ..Oo,BRT ioet/ApidGohmsr 6(01 1513±1544 (2001) 16 Geochemistry Applied / Simoneit B.R.T. Oros, D.R.

Mannosan C6H10O5 162 11773 818 3787 21019 3212 14619 341 5100 12217 584 6964 3645 744 A Levoglucosan C6H10O5 162 40878 11256 16189 57468 14791 27977 2750 19452 40707 3381 43522 10448 422 A Total monosaccharides 67188 17457 23683 99492 21061 54745 3198 29510 66309 4179 59335 15897 1396

VI. Unknowns Total unknowns 117441 56639 429 321648 1916 13219 68163 48043 212328 2559 35223

VII. Miscellaneous Unresolved complex 581006 110939 307576 1129924 270713 710099 85892 305 685106 68573 688853 344263 113813 mixture (mg/kg) Unresolved components: 0.6 0.8 1.2 1.0 1.1 1.2 0.8 1.4 1.1 0.7 0.9 0.9 0.8 resolved components (U:R) Organic carbon per 46908 3040 3030 35104 5194 16069 2599 2663 21393 1495 31957 11466 2395 amount burned (mg/kg) Elemental carbon 1663 159 178 451 188 284 59 163 640 158 468 224 915 (mg/kg) Organic carbon/elemental 28 19 17 78 28 57 44 16 33 9 68 51 3 carbon (OC/EC) Methylphenanthrenes/ 1.5 2.6 1.2 1.4 1.1 1.9 1.3 1.4 2.1 2.4 0.5 0.8 2.4 phenanthrene (MP:P)

a These data are also available from the corresponding author expressed as mg/g organic carbon (OC) emitted. b Identi®cation criteria: blank space indicates not present or below detection limit; nd=not determined; A=matches with authentic standard; S=interpolated from homologous series fragmentation pattern; I=interpreted from mass spectrum fragmentation pattern. All compositions and molecular weights are as the compounds occur in smoke (i.e. underivatized). CPI for n-alkanes and n- alkenes: [CPI=ÆC13ÀC35 /ÆC12 ÀC34] from Mazurek and Simoneit (1984); CPI for n-alkanoic acids, n-alkanols and n-alkanones: [CPI=ÆC12 ÀC34/ÆC13 ÀC35 ] from Mazurek and Simoneit (1984). Organic carbon in this case represents the solvent extractable matter. 1529 1530 D.R. Oros, B.R.T. Simoneit / Applied Geochemistry 16 (2001) 1513±1544 each conifer species included in this study. The dis- even C number predominance (CPI range from 0.3 to tributions of the molecular classes include the following: 7.3, average=1.2), and Cmax varying from 21 to 24 and homologous series of aliphatic compounds (n-alkanes, 22 predominant in 69% of all samples (Table 2, also n-alkenes, n-alkanoic acids and wax esters); polycyclic Figs. 1b, 2b and 3b). Alkenes are not major components aromatic hydrocarbons (PAH); monosaccharides from in plant waxes and their origin has been inferred to be cellulose; phenolics from lignin; and steroid and terpe- from biomass fuel (Abas et al., 1995). The n-alkenes are noid (mainly diterpenoid) biomarkers. The distributions formed primarily by the thermal dehydration of n-alka- and abundances of the conifer smoke constituents are nols (which show even C number predominance: strongly dependent on combustion conditions (e.g., Mazurek and Simoneit, 1984) and to a minor degree smoldering versus ¯aming, duration). Thus, the values from the n-alkanes by oxidation during incomplete reported here should not be used as absolute but as combustion (Abas et al., 1995). The distributions of n- relative chemical ®ngerprints for these sources. The alkenes showing even C number predominances and biomarkers are source speci®c and may be used as con- Cmax at 24, coupled with the low abundances of n-alka- ®rming tracers for transport and fate studies of conifer nols with Cmax at 22 or 24 (Table 2), further supports an smoke emissions in the environment. origin from n-alkanols for these molecules.

4.1. Homologous compound series 4.1.3. n-Alkanoic acids

The n-alkanoic acids range from C7 to C34,showa Examples of the typical GC±MS TIC (total ion cur- strong even to odd C number predominance (CPI range rent) traces for the total extract and TLC fractions of from 2.5 to 15.6, average 6.9), and Cmax at 16, 20 or 22 conifer smoke samples (Douglas Fir, Mountain Hem- (Table 2, also Figs. 1c±4c). These compounds, which are lock, Ponderosa Pine, and Sitka Spruce) are given in basic units of plant fats, oils and phospholipids, are iden- Figs. 1±4. The TIC traces of the total extracts of the ti®ed here as a major molecular class for all conifer smoke smoke samples show the distributions and relative samples. There are also minor contributions from unsa- abundances of the major organic constituents, while the turated fatty acids, both oleic (C18:1) and linoleic (C18:2). TIC traces of the TLC fractions F1 to F4 show the distributions and abundances of the aliphatics, aromatics 4.1.4. a,o-Alkanedioic acids and molecular biomarkers separated according to func- Series of a,o-alkanedioic acids are present and range tional group and polarity properties. The TLC separa- from C9 to C29 (Table 2, also Fig. 1e). The most com- tion procedure was conducted on all smoke samples in mon a,o-alkanedioic acid in conifer smoke is C20 (pre- order to best identify a source speci®c chemical ®nger- sent in 38% of all samples). The photo-oxidation print that is representative of conifer smoke emissions. product (Stephanou and Stratigakis, 1993) of C18:1 and Thus, the discussion will focus on the identity and dis- C18:2 alkenoic acids, a,o-nonanedioic acid, is present in tributions (C number range and maxima, Cmax, and C only a single sample. The a,o-alkanedioic acids have preference indices, CPI; Mazurek and Simoneit, 1984) been identi®ed from a variety of sources and in the of the major aliphatic homologs and biomarkers. environment (Simoneit, 1989; Rogge et al., 1993a; Hil- demann et al., 1994; Abas et al., 1995; Gogou et al., 4.1.1. n-Alkanes 1996). High molecular weight a,o-alkanedioic acids

The distribution of n-alkanes in conifer smoke (C10±C24) have been identi®ed in rural aerosol particles (Table 2, also see examples in Figs. 2b and 4b) ranges in and their source may be oxidation products of o-

C chain length from C14 to C34 and shows an odd to hydroxy alkanoic acids from vegetation polyester bio- even C number predominance (CPI range from 0.5 to polymer (Simoneit and Mazurek, 1982). The identi®ca- 3.4, average=1.5). This distribution suggests an n- tion here of all acids con®rms a source contribution alkane contribution from epicuticular waxes (Oros et from the burning of biomass. al., 1999). Vascular plants synthesize epicuticular waxes containing odd carbon number n-alkanes usually in the 4.1.5. n-Alkanones

C25±C33 range with C29 or C31 as dominant homologs The straight chain ketones as n-alkan-2-ones range which often contribute up to 90% of all parans found from C16 to C33 (Table 2) and show an odd to even C in plant waxes (Kolattukudy, 1970). The Cmax for the n- number predominance (CPI range from 0.4 to 4.2, alkanes are diverse and vary from 20 to 33. The n- average=2.3). The Cmax ranged from 21 to 27. The n- alkane distributions con®rm an input from epicuticular alkan-2-ones are mainly derived from the partial com- wax sources (Cmax5C27 present in 31% of all samples). bustion of aliphatic precursors (Simoneit, 1978).

4.1.2. n-Alkenes 4.1.6. n-Alkanols The n-alkenes are primarily terminal ole®ns (i.e. alk- Homologous series of n-alkanols with even to odd C

1-enes). They range from C13 to C28, with an odd to number predominances are present in conifer smoke D.R. Oros, B.R.T. Simoneit / Applied Geochemistry 16 (2001) 1513±1544 1531

Fig. 1. GC±MS total ion current traces of Douglas Fir smoke particulate matter: (a) total extract showing the major organic components, (b) F1 fraction showing n-alkenes and steroid biomarkers, (c) F2 fraction showing n-alkanoic acids,(d) F3 fraction showing n-alkanoic acids and diterpenoid biomarkers, and (e) F4 fraction showing n-alkanol and n-alkanedioic acids (numbers refer to C chain length of n-alkanes, A=n-alkanoic acid, OH=n-alkanol, U=unknown, UCM=unresolved complex mixture).

(CPI=4.1 and 5.7). The n-alkanols ranged from C18 to In contrast to the primary alcohols, the free second- C30 with Cmax at 22 and 24 (Table 2, also Figs. 1e±3e). ary alcohol n-nonacosan-10-ol is present as a major The n-alkanols from C20 to C30 are predominantly of an component in all of the conifer smoke samples (Table 2, epicuticular wax origin. Figs. 1a, 2a, 3e and 4a). This compound has been pre- 1532 D.R. Oros, B.R.T. Simoneit / Applied Geochemistry 16 (2001) 1513±1544

Fig. 2. GC±MS total ion current traces of Mountain Hemlock smoke particulate matter: (a) total extract showing the major organic components, (b) F1 fraction showing n-alkanes, n-alkenes, and steroid biomarkers, (c) F2 fraction showing n-alkanoic acids, PAH, and diterpenoid biomarkers, (d) F3 fraction showing n-alkanoic acids and diterpenoid biomarkers, (e) F4 fraction showing n-alkanoic acids and n-alkanols (abbreviations as in Fig. 1 and IS=internal standard). D.R. Oros, B.R.T. Simoneit / Applied Geochemistry 16 (2001) 1513±1544 1533

Fig. 3. GC±MS total ion current traces of Ponderosa Pine smoke particulate matter: (a) total extract showing the major organic components, (b) F1 fraction showing n-alkanoic acids, PAH and diterpenoid biomarkers, (c) F2 fraction showing n-alkanoic acids and diterpenoid biomarkers, (d) F3 fraction diterpenoid biomarkers, and (e) F4 fraction showing n-alkanoic acids, n-alkanols, and diterpenoid biomarkers (abbreviations as in Fig. 1). 1534 D.R. Oros, B.R.T. Simoneit / Applied Geochemistry 16 (2001) 1513±1544

Fig. 4. GC±MS total ion current traces of Sitka Spruce smoke particulate matter: (a) total extract showing the major organic components, (b) F1 fraction showing n-alkanes, n-alkenes, and diterpenoid and steroid biomarkers, (c) F2 fraction showing n-alkanoic acids, PAH, and diterpenoid biomarkers, (d) F3 fraction showing n-alkanoic acids and diterpenoid biomarkers, (e) F4 fraction showing n-alkanoic acids and diterpenoid biomarkers (abbreviations as in Fig. 1). D.R. Oros, B.R.T. Simoneit / Applied Geochemistry 16 (2001) 1513±1544 1535 viously identi®ed as a major component in epicuticular Simoneit, 1997). Both dehydroabietic acid and retene waxes from gymnosperm species (Tulloch, 1976, 1987; have been proposed as candidate molecular tracer com- Schulten et al., 1986). pounds for coniferous wood combustion (Ramdahl, 1983; Simoneit et al., 1993; Standley and Simoneit, 1994). 4.2. Molecular biomarkers The product-precursor relationship for the diterpenoids in conifer smoke may follow an alteration path- Molecular biomarkers (i.e. biomarkers) are organic way which commences with the dehydrogenation of compounds of biological origin that show little or no abietic acid to dehydroabietic acid with subsequent change in chemical structure from their parent organic decarboxylation to dehydroabietin and full aromatiza- molecule (i.e. natural product) found in living organ- tion to retene (Simoneit, 1998) (Fig. 5). Dehy- isms. Such molecules are characterized by their restric- droabietane, which is also present, may dehydrogenate ted occurrence, source speci®city, molecular stability to simonellite and then proceed to retene (Standley and and suitable concentration for analytical detection Simoneit, 1994). The other resin acids initially rearrange (Mazurek and Simoneit, 1984). The major biomarkers to the abietane skeleton before dehydrogenation to identi®ed in the conifer smoke samples include diterpe- dehydroabietic acid. They also eliminate ethylene and noids, monosaccharide derivatives from cellulose, dehydrogenate to bisnordehydroabietic acid. methoxyphenols from lignin, sterols and wax esters, The ratios of the total natural and altered compounds including their thermal alteration products. It has been that contain the abietane skeleton to the total natural shown that these high molecular weight compounds are and altered compounds that contain the pimarane ske- directly volatilized into smoke by an injection mechan- leton (A/P, abietane skeletons/pimarane skeletons) ism similar to steam volatilization/stripping. Subsequent range from 0.9 to 8.3 (average=3.3) (Table 2). The A/P condensation onto or entrapment into preexisting par- ratios are distinct for each conifer smoke sample as is ticulate matter when the smoke plume is diluted and shown in Fig. 6. Thus, they may be useful indicators of cooled provides the means for their incorporation into source speci®c burn emissions. the atmospheric aerosol phase (Simoneit et al., 1993). 4.2.2. Monosaccharide derivatives 4.2.1. Diterpenoids Cellulose and hemicellulose biopolymers which are The major biomarkers present in conifer smoke are mainly responsible for structural strength compose the diterpenoids and their thermal alteration products approximately 40±50% and 20±30% dry weight of (Table 2, also Figs. 1a±4a). Diterpenoids are important wood, respectively (Sergejewa, 1959; Petterson, 1984). A biomarker constituents of many higher plants, especially cellulose molecule is a long-chain, linear polymer made of conifers, in their resins (Erdtman et al., 1968; Rier et up of 7000±12,000 d-glucose monomers, while a hemi- al., 1969; Zinkel and Clarke, 1985; Zinkel et al., 1985; cellulose molecule is a 100±200 sugar monomers poly- Simoneit, 1986, 1998; Zinkel and Magee, 1987; Lorbeer saccharide mixture of glucose, mannose, galactose, and Zelman, 1988; Barrero et al., 1991; Simoneit et al., xylose, arabinose, 4-O-methylglucuronic acid and 1993, 1999; Mazurek and Simoneit, 1997). Many soft- galacturonic acid (Sergejewa, 1959; Parham and Gray, wood species are proli®c resin producers and have well 1984). It is the burning of wood at temperatures established systems of horizontal and vertical ducts ®l- >300C which gives rise to the source speci®c molecular led with resin in the wood (Parham and Gray, 1984). tracers, i.e. mainly 1,6-anhydro-b-d-glucopyranose, also The predominant biomarkers identi®ed in conifer called levoglucosan (Table 2, also Figs. 1a±4a). Levo- smoke have the abietane and pimarane skeletons which glucosan has been previously reported in biomass burn- are the major diterpenoids produced by gymnosperms in ing and atmospheric particles (Hornig et al., 1985; the northern hemisphere (Thomas, 1970). The most com- Locker, 1988; Simoneit et al., 1999). Levoglucosan is the mon diterpenoid natural products present in the smoke predominant organic component in smoke from Mon- samples are iso-pimaric acid with lesser amounts of tezuma Pine and is detectable in the smoke samples pimaric acid, sandaracopimaric acid and abietic acid from all conifers. The low levels of levoglucosan repor- (Table 2). The major thermal alteration (oxidation) prod- ted here in conifer smoke samples are due to extraction ucts are 8,15-pimaradien-18-oic acid, dehydroabietic acid, using only CH2Cl2 (typically 10Â greater using a polar 1-methyl-7-isopropyl-1,2,3,4-tetrahydrophenanthrene-1- solvent). Levoglucosan is emitted at such high con- carboxylic acid followed by lesser amounts of retene and centrations that it is detectable in aerosol particulate 7-oxodehydroabietic acid. Dehydroabietic acid is the matter at considerable distances from the combustion major organic component in smoke from Apache Pine sources (Simoneit et al., 1999). but occurs in all samples analyzed (Table 2). Dehy- droabietic acid has been regarded both as a partially 4.2.3. Methoxyphenols altered atmospheric oxidation product and a pyrolysis Lignin biopolymer comprises approximately 20±30% product of resin acids (Simoneit, 1986; Mazurek and of the dry weight of wood (Sergejewa, 1959; Petterson, 1536 D.R. Oros, B.R.T. Simoneit / Applied Geochemistry 16 (2001) 1513±1544

Fig. 5. The thermal alteration products of diterpenoid biomarker precursors observed in conifer smoke samples (only major compounds are shown).

1984). The lignin biopolymers are derived from p-cou- 1992; Edye and Richards, 1991; Simoneit et al., 1993; maryl, coniferyl and sinapyl alcohols and contain Mazurek and Simoneit, 1997). mainly anisyl, vanillyl and syringyl nuclei (Simoneit et The phenolics in conifer smoke are composed mainly al., 1993). Gymnosperm lignin is enriched in the con- of lignin pyrolysis products, lignans and dimers of sub- iferyl alcohol precursor and on burning produces pri- stituted phenols. The predominant phenolic biomarkers marily vanillyl moieties. Burning (pyrolysis) of wood in conifer smoke include catechol, pyrogallol, vanillin, injects these lignin nuclei into smoke as breakdown homovanillic acid, vanillic acid, homovanillyl alcohol products such as acid, aldehyde, ketone and alkyl deri- and acetovanillone (Table 2, also Figs. 2a±4a). The vatives of the methoxyphenols (Hawthorne et al., 1988, phenol substitution (i.e. 3-methoxy-4-hydroxy) pattern D.R. Oros, B.R.T. Simoneit / Applied Geochemistry 16 (2001) 1513±1544 1537

Fig. 6. The ratios of the total of natural and altered compounds with the abietane skeleton to the total of natural and altered compounds with the pimarane skeleton (A/P) for each conifer smoke sample. is consistent with an origin from gymnosperm (soft- al., 1988; Simoneit, et al., 1993). The lignin phenols, wood) (Simoneit et al., 1993). The phenolic compound lignans and secondary dimers have mainly coniferyl guaiacylacetone is also present. Guaiacyl derivatives are alcohol type phenolic structures, thus they may be uti- potential biomarker tracers for both hard and soft- lized as biomarker tracers for conifer combustion emis- woods (Hawthorne et al., 1988). A major lignan of sions. conifer smoke is tetrahydro-3,4-divanillylfuran (Table 2). Lignans have been described previously as tracers for 4.2.4. Steroids distinguishing between coniferous and deciduous wood The sterols, generally comprised of the C28 and C29 smoke emissions (Simoneit et al., 1993). Secondary phytosterol compounds, are constituents of plant lipid products as dimers of substituted phenols are present membranes and waxes (Goad, 1977; Goodwin, 1980). and include divanillyl and 1,2-divanillylethane. Both are The sterol biomarkers are present in all conifer smoke derived from coniferyl alcohol type precursors and have samples (Table 2, also Figs. 1a and 2a). The natural been previously identi®ed in wood smoke (Hawthorne et product b-sitosterol is the most common sterol in con- 1538 D.R. Oros, B.R.T. Simoneit / Applied Geochemistry 16 (2001) 1513±1544 ifer smoke immediately followed by campesterol and aromatization products are also found, mainly as 24- less so by stigmasterol, also the natural products. Sev- ethyl-4-methyl-19-norcholesta-1,3,5(10)-triene and its eral C29 thermal alteration products from the sterol structural isomers. The thermal alteration products of precursor stigmasterol are present and include stigmast- sterol precursors are summarized in Fig. 7 and can be 3,5-diene, stigmast-5-ene and stigmast-4-ene. Various used as general indicators for burning of higher plant

Fig. 7. The thermal alteration products of phytosterol precursors observed in conifer smoke samples (only major compounds are shown). D.R. Oros, B.R.T. Simoneit / Applied Geochemistry 16 (2001) 1513±1544 1539 lipids (Simoneit, 1989; Simoneit et al., 1993). Pregnenes, Simoneit, 1998; Oros and Simoneit, 2000). They are thus the unsaturated C21 thermal cracking products from not exclusive markers for biomass combustion. sterols by loss of the side chain, are present in Lodge- The ratio of methylphenanthrenes to phenanthrene pole and Ponderosa Pine smoke samples and include the (MP:P) has been previously used as an indicator of isomers with double bonds at the C-5 and C-7 positions anthropogenic in¯uences in the environment: 0.5 for (Fig. 7). Overall, the phytosterols and their alteration atmospheric fallout (Takada et al., 1991), 0.5±1.0 for products are present only as minor constituents in con- combustion sources (Prahl and Carpenter, 1983), 1.0 for ifer smoke. street and urban dusts (Takada et al., 1990, 1991), 2.0± 6.0 for fossil fuel (Prahl and Carpenter, 1983), and 4.0 4.2.5. Wax esters for crankcase oil (Pruel and Quinn, 1988). The range of Wax esters have been previously reported in conifer the MP:P ratio determined for the conifer smoke sam- cuticular waxes (Tulloch, 1987; SuÈ mmchen et al., 1995; ples is 0.5±2.6 (average=1.6), which is proposed here as Oros et al., 1999). These compounds form crystalline a potential indicator for conifer burning emissions. zones in the cuticles of needles and leaves which act as transport barriers to diminish the loss of water (Rie- 4.4. Unresolved complex mixture derer and Schneider, 1990). Here, in the smoke extracts, the wax esters range mainly from C21 to C33 (total C An unresolved complex mixture (UCM) of structu- number of compounds) and have exclusively saturated rally complex isomers and homologs of branched and fatty acid and alcohol moieties (Table 2). The major cyclic hydrocarbon compounds (Eglinton et al., 1975) homolog and dominant Cmax is 26 in 7 of the samples eluting between C14 and C34 alkanes is present as a where wax esters are present. Alkanoic acid moieties major organic component of all conifer smoke total range from C12 to C16 and n-alkanols from C9 to C21, extracts (Figs. 1a±4a). The UCM, which has been thor- with common combinations of acid to alcohol moieties oughly examined in petroleum sources, is comprised of of C12 to C13,C14 to C13 and C14 to C14 predominating. compounds which are relatively inert to microbial The combinations of the acid and alcohol moieties vary degradation (Gough and Rowland, 1990; Killops and considerably from species to species, thus these com- Al-Juboori, 1990). The ratio of UCMto resolved compounds may be useful source indicators for conifer spe- ponents (U:R) has been used as a parameter for the cies in smoke samples. The presence of very high indication of petroleum contribution to aerosol particle molecular weight wax esters (>C40) would require con- samples (Mazurek and Simoneit, 1984). The U:R ratios ®rmation by high temperature GC or high temperature for conifer smoke samples were quanti®ed from the GC±MS (Elias et al., 1997, 1998). total extract in order to determine if this parameter is useful for distinguishing between conifer and fossil fuel 4.3. Polycyclic aromatic hydrocarbons (petroleum and coal) derived combustion source emissions (Table 2). The conifer smoke U:R ratios range All biomass ®res are pyrolysis processes causing the from 0.6 to 1.4 (average=1.0). The close similarity in formation of polynuclear aromatic hydrocarbons U:R ratios suggests that this parameter is conservative (PAH) from (a) the high temperature thermal alteration for these smoke emissions. Several U:R ratios have been of natural product precursors in the source organic determined from more mature fossil fuel derived com- matter and (b) the recombination of molecular frag- bustion emission sources which include the following: ments in the smoke (Simoneit, 1998). The identi®cations lignite coal=3.2 and bituminous coal=3.3 (Oros and and abundances of over 30 PAH compounds present in Simoneit, 2000); catalyst-equipped automobile engine the conifer smoke samples are given in Table 2. The exhaust=5.5 and heavy-duty diesel truck engine major PAH are phenanthrene, anthracene, C1-phenan- exhaust=9.3 (Rogge et al., 1993a). Thus, the lower U:R threnes/anthracenes (since anthracene is a minor PAH, ratio of conifer smoke shows that this parameter is use- the alkylanthracenes are expected to be negligible, based ful for distinguishing between conifer biomass burning on compound elucidation for other combustion sam- and fossil fuel derived combustion source emissions. ples, Simoneit, 1998), ¯uoranthene and pyrene followed Ultimately, the U:R ratio may be used as an indicator by lesser amounts of C2- and C3-phenanthrenes, C1- for identifying atmospheric transport trajectories from pyrenes, 11(H)-benzo[a]¯uorene and chrysene (Figs. 2c regional biomass burning and fossil fuel combustion and 3b). Certain PAH that exhibit mutagenic and gen- emission containing air parcels. This is especially useful otoxic potential such as benz[a]anthracene, benzo[a]- for determining the contributions of organic matter pyrene and cyclopenta[c,d]pyrene (Arcos and Argus, derived from rural versus urban emission sources. 1975; IARC, 1989), are also present, however only as The U:R ratios for aerosol samples collected over the minor constituents. The PAH identi®ed here are also western United States for rural (0.2±4.0), mixed (1.4± emitted by internal combustion engines, coal burning, 3.4), and urban (0.9±25.0) areas have been reported and other anthropogenic sources (Rogge et al., 1993a; (Mazurek and Simoneit, 1984). Generally, urban aero- 1540 D.R. Oros, B.R.T. Simoneit / Applied Geochemistry 16 (2001) 1513±1544 sols were shown to contain the largest component of Table 3 petroleum-derived compounds, while rural and mixed Major compound groups identi®ed in conifer smoke rural/urban environments showed variable contribu- Major compound Total emission factor (%) tions of anthropogenic pollutants. The average U:R group (mg/kg)a Abundanceb ratio for conifer smoke (1.0) suggests that the UCMof rural aerosol particles from the western United States Diterpenoids 3093Æ225 32.3 consists mainly of recent (immature) organic matter Carboxylic acids 2753Æ167 28.6 derived from conifer and perhaps other biomass com- Methoxyphenols 1172Æ76 12.2 bustion source emissions, such as grass smoke released Monosaccharide 463Æ31 4.8 from agricultural ®eld burning, and less pronounced derivatives Polycyclic aromatic 408Æ23 4.2 fossil fuel combustion emissions. hydrocarbons Alkenes 302Æ53 3.1 4.5. Organic and elemental carbon Alkanes 199Æ21 2.1 Alkanols 123Æ9 1.3 The concentrations of organic C (OC) and elemental Steroids 90Æ5 0.9 C (EC, i.e. black soot) in the dierent conifer smoke Alkanones 83Æ10 0.9 samples are given in Table 2. The organic to elemental C Wax esters 48Æ5 0.5 ratios (OC/EC) show a range from 3 to 78 (aver- Unknowns 877Æ98 9.1 age=35). The OC/EC ratios for the conifer smoke a Sum and standard deviation for each compound group samples are elevated in comparison to ambient air sam- listed in Table 2. ples collected from rural sites (Crater Lake, OR=12.4, b % Abundance relative to sum of total emission factors. Carus, OR=6.5, and Sauvie, OR=4.1) and urban areas (Los Angeles=1.6, New York=1.4, Santiago, Chile=1.7, China=1.5) (Didyk et al., 2000). The low (needles, cones and branches of 1±2 cm thickness) which OC/EC ratios for mainly urban and suburban areas were relatively immature in plant structural develop- indicate a strong in¯uence from petroleum derived ment. combustion emissions. The conifer smoke OC/EC ratios are much less than that measured from an ambient air 4.7. Major and unique biomarker tracers sample collected at a remote area (South Atlantic Ocean=160, Didyk et al., 2000) where in¯uence from The major biomarker compounds identi®ed for con- both petroleum combustion and biomass burning emis- ifers to be applied as potential tracers in smoke and in sions is negligible. This distribution indicates that the the atmosphere are given in Table 4. These are the OC/EC ratio measured for conifer smoke may be useful diterpenoid natural products (iso-pimaric acid, pimaric in distinguishing this source from petroleum derived acid, abietic acid and sandaracopimaric acid) and their combustion emissions such as those found in rural and dominant combustion alteration products (8,15-pimar- urban areas and from natural emissions (background) adien-18-oic acid, dehydroabietic acid, 1-methyl-7-iso- found in remote areas. propyl-1,2,3,4-tetrahydrophenanthrene-1-carboxylic acid, retene and 7-oxodehydroabietic acid). The major 4.6. Major compound groups biomarker derived from combustion of cellulose biopolymer is levoglucosan, which has been previously pro- The total emission factors (mg/kg) and % abundances posed as a tracer for cellulose burning. Galactosan and of the major compound groups identi®ed in conifer mannosan are also detectable as secondary cellulose smoke are given in Table 3. Of the total resolved com- derivatives. The major tracers from lignin combustion ponents the major compound groups are the diterpe- are methoxyphenolic compounds, including vanillin, noids (32%) from bleed resins, carboxylic acids (29%) vanillic acid, homovanillic acid, homovanillyl alcohol from internal lipids and methoxyphenols (12.2%) and acetovanillone, typical of the predominant coniferyl derived from lignin. Other compound groups such as alcohol content of the precursor biopolymer. Com- steroid biomarkers and aliphatic homologous series are pound series such as n-alkanes, n-alkenes, n-alkanoic present at low abundances (<5%). Although wood is acids, n-alkanones, n-alkanols, PAH, phytosterols, composed mostly of cellulose (40±50% of dry weight of anhydrosaccharides (e.g. levoglucosan), and UCMare wood, d.w.w.), with lesser amounts of hemicelluloses not source speci®c, because they are generally found in (20±30% of d.w.w.), and lignin (20±30% of d.w.w.) all biomass combustion emissions (Simoneit, 1984, (Sergejewa, 1959; Petterson, 1984), the % abundance of 1989; Abas et al., 1995; Simoneit et al., 1999). However, the monosaccharide derivatives (4.8%) from cellulose some of these compound series are indicative of species alteration is less than the methoxyphenols. This obser- speci®c biomass burning, when coupled with the directly vation is likely due to burning of the selected plant parts emitted and thermally altered biomarker compounds. D.R. Oros, B.R.T. Simoneit / Applied Geochemistry 16 (2001) 1513±1544 1541

Table 4 Some conifer smoke samples contain unique bio- Major molecular tracers identi®ed in conifer smoke markers which may be useful as species speci®c tracers. For example, Mountain Hemlock smoke contains traces Major biomarker Total emission (%) tracer factor (mg/kg)a Abundanceb of 3a-methoxyfriedelene, 3b-methoxyfriedelene, 3a- ethoxyfriedelene and 3b-ethoxyfriedelene (tentative MS Diterpenoids interpretation). These triterpenoid natural products are 8,15-Pimaradien-18-oic acid 400Æ47 4.2 not present in other conifer smoke samples, thus may be Dehydroabietic acid 363Æ36 3.8 iso-Pimaric acid 248Æ35 2.6 unique tracers in smoke from this species. Port Orford 1-Methyl-7-isopropyl-1,2,3,4- 151Æ22 1.6 Cedar smoke shows the sesquiterpenoid natural pro- tetrahydrophenanthrene- ducts 5-hydroxycalamenene, a-calacorene, aromaden- 1-carboxylic acid drol, b-oplopenone and 6-deoxygeigerin along with the Retene 137Æ12 1.4 7-Oxodehydroabietic acid 61Æ7 0.6 diterpenoids hibaene, totarol and 6,7-dehydroferruginol Pimaric acid 57Æ7 0.6 as unique tracers. Montezuma Pine smoke contains the Sandaracopimaric acid 22Æ2 0.2 diterpenoid natural products laurene, rimuene and pini- Methoxyphenols folic acid as unique markers. Laurene and rimuene are Catechol 177Æ12 1.8 also present at lower concentrations in Sitka Spruce Homovanillyl alcohol 101Æ10 1.0 smoke. California Redwood smoke has daniellic acid, Pyrogallol 74Æ7 0.8 Vanillin 57Æ4 0.6 polyaltic acid and 6,7-dehydroferruginol as unique Vanillic acid 48Æ3 0.5 diterpenoid tracers. Apache Pine smoke shows agatholic Homovanillic acid 44Æ3 0.5 acid with a lesser amount of pinifolic acid as unique Tetrahydro-3,4-divanillylfuran 42Æ2 0.4 diterpenoid natural product tracers. 15-Hydro- Acetovanillone 25Æ2 0.3 xydehydroabietic acid is found in Ponderosa Pine and Monosaccharide derivatives Western White Pine smoke. Copalic acid is a unique but Levoglucosan 289Æ18 3.0 Galactosan 89Æ7 0.9 minor diterpenoid tracer in Western White Pine smoke. Mannosan 84Æ6 0.9 The examples of speci®c biomarkers may be useful as Steroids indicative tracers of species speci®c biomass burning. b-Sitosterol 14Æ1 0.1 The relative abundances (%) of key biomarkers from Campesterol 1Æ0.01 0.01 conifer smoke may be used to distinguish fuel type. a Sum and standard deviation for individual compounds listed in Table 5 shows 6 key biomarkers derived mostly from Table 2. internal plant components (levoglucosan, dehy- b % Abundance relative to total resolved organic components. droabietic acid and catechol) and epicuticular wax lipids

Table 5 Abundances (% relative to maximum) of key biomarkers for identifying fuel type

Sample Heptacosane Palmitic Docosanol Levoglucosan Dehydroabietic Acid Catechol Acid

Apache 0.3 74 2 36 100 33 pine California 1 100 0.3 81 10 37 redwood Douglas ®r 4 25 20 100 62 59 Eastern White 0.3 75 18 100 15 36 pine Lodgepole 0.2 38 8 63 100 27 pine Montezuma 1 59 7 100 4 39 pine Mountain 3 13 6 59 100 88 hemlock Noble ®r 10 7 4 100 53 27 Paci®c silver ®r 4 81 11 100 49 79 Ponderosa pine 0.1 3 3 7 100 3 Port Orford 2 100 2 92 2 36 cedar Sitka spruce 1 31 1 10 100 12 Western White pine 1 100 9 4 5 19 1542 D.R. Oros, B.R.T. Simoneit / Applied Geochemistry 16 (2001) 1513±1544

(heptacosane, palmitic acid and docosanol). The dis- those found in rural and urban areas, and from natural tributions of these compounds relative to one another emissions found in remote areas. are dierent and represent the unique chemical and physical characteristics between conifer species. The relative abundances of key biomarkers and homologous Acknowledgements series compounds reported here can collectively be used as speci®c tracers for assessing and tracking emissions This work was funded in part by the US Environ- from burning of conifer fuels. mental Protection Agency (Grant CR 823990-01-0). We thank Dr. W.F. Rogge and an anonymous reviewer for their useful suggestions to improve this manuscript. 5. Conclusions

This work reports the lipid and molecular biomarker References components in smoke from burning of predominant conifers of western North America. The data is of use Abas, M.R., Simoneit, B.R.T., Elias, V., Cabral, J.A., Cardoso, for assessing direct organic composition signatures for J.N., 1995. Composition of higher molecular weight organic particle emissions from conifer fuels during biomass matter in smoke aerosol from biomass combustion in Ama- burning. The abundance order for the major molecular zonia. Chemosphere 30, 995±1015. classes in conifer smoke samples was identi®ed as Arcos, J.C., Argus, M.G., 1975. Chemical Induction of Cancer. the following: UCM>diterpenoids>carboxylic acids> Structural Basis and Biological Mechanisms, Vol. IIA. Aca- methoxyphenols>anhydrosaccharides>PAH>alkenes demic Press, New York. Barrero, A.F., Sanchez, J.F., Alvarez-Manzaneda, R.E.J., >alkanes>alkanols>steroids>alkanones>wax esters. MunÄ oz Dorado, M., Haidour, A., 1991. Endoperoxide Variations in this molecular group order exist among diterpenoids and other constituents from Abies marocana. the conifer species burned, however, these are only Phytochemistry 30, 593±597. minor and usually occur between close ranking mole- Birch, M.E., Cary, R.A., 1996. Elemental carbon-based cular groups. method for monitoring occupational exposures to particulate Although the concentrations of organic compounds diesel exhaust. Aerosol Science and Technology 25, 221±241. in smoke aerosols are highly variable and dependent on Didyk, B.M., Simoneit, B.R.T., Pezoa, A.L., Riveros, L.M., combustion temperature, aeration, and moisture con- Flores, A.A., 2000. Urban aerosol of Santiago, Chile: tent of the source fuel, the biomarkers and their com- organic content and molecular characterization. Atmo- bustion alteration products are source speci®c. The spheric Environment 34, 1167±1179. Edye, L.A., Richards, G.N., 1991. Analysis of condensates major biomarkers identi®ed in the smoke samples are from wood smoke: components derived from poly- useful as tracers for distinguishing the conifer burning saccharides and lignins. Environmental Science and Tech- component in atmospheric aerosol source attributions. nology 25, 1133±1137. The relative abundances of key biomarkers in conifer Eglinton, G., Maxwell, J.R., Philp, R.P., 1975. Organic geo- smoke may also be used to distinguish fuel type. chemistry of sediments from contemporary aquatic environ- The range of the MP:P ratio determined for the con- ments. In: Tissot, B., Bienner, F. (Eds.), Advances in Organic ifer smoke samples is 0.5±2.6 (average=1.6), which may Geochemistry 1973. Editions Technip, Paris, pp. 941±961. be useful for distinguishing conifer burning emissions Elias, V.O., Simoneit, B.R.T., Pereira, A.S., Cardoso, J.N., from fossil fuel and other combustion emissions, and 1997. Mass spectra of triterpenyl alkanoates, novel natural from street and urban dusts. The MP:P range is pro- products. Journal of Mass Spectrometry 32, 1356±1361. Elias, V.O., Simoneit, B.R.T., Pereira, A.S., Cardoso, J.N., posed here as a potential indicator for identifying con- 1998. High temperature gas chromatography with a glass ifer burning emissions in atmospheric and other capillary column for the analysis of high molecular weight environmental samples. Additionally, the average U:R tracers in smoke samples from biomass burning. Journal of ratio for conifer burning emissions (1.0) may also be of High Resolution Gas Chromatography 21, 87±93. utility for distinguishing this source from fossil fuel- Erdtman, H., Kimland, B., Norin, T., Daniels, P.J.L., 1968. derived combustion emissions (e.g., coal combustion, The constituents of the ``Pocket Resin'' from Douglas Fir gasoline and diesel engine exhaust). The ratio also sug- Pseudotsuga menziesii (Mirb.) Franco. Acta Chemica Scan- gests that conifer and other biomass burning emissions dinavica 22, 938±942. are signi®cant contributors of particle bound immature Goad, L.J., 1977. The biosynthesis of plant sterols. In: Tevini, organic matter (UCM) present in the atmosphere of M., Lichtenthaler, H.K. (Eds.), Lipids and Lipid Polymers in Higher Plants. Springer-Verlag, New York, pp. 146±168. rural areas of the western United States. The OC/EC Gogou, A., Stratigakis, N., Kanakidou, M., Stephanou, E.G., ratios measured for conifer smoke samples range from 3 1996. Organic aerosols in Eastern Mediterranean: compo- to 78 (average=35). The distribution indicates that the nents source reconciliation by using molecular markers and OC/EC ratios may be useful to distinguish this source atmospheric back trajectories. Organic Geochemistry 25, 79± from petroleum derived combustion emissions, such as 96. D.R. Oros, B.R.T. Simoneit / Applied Geochemistry 16 (2001) 1513±1544 1543

Goodwin, T.W., 1980. The biosynthesis of plant steroids. In: weight terpenoids as indicators of organic emissions from Stumpf, P.W. (Ed.), The Biochemistry of Plants, Vol. 4. terrestrial vegetation. In: Eganhouse, R.P. (Ed.), Molecular Academic Press, New York, pp. 485±507. Markers in Environmental Geochemistry, Amer. Chem. Soc. Gough, M.A., Rowland, S.J., 1990. Characterization of unre- Symp. Series 671. ACS, Washington, DC, pp. 92±108. solved complex mixtures of hydrocarbons in petroleum. Oros, D.R., Simoneit, B.R.T., 2000. Identi®cation and emission Nature 344, 648±650. rates of molecular tracers in coal smoke particulate matter. Hawthorne, S.B., Miller, D.J., Barkley, R.M., Krieger, M.S., Fuel 79, 515±536. 1988. Identi®cation of methoxylated phenols as candidate Oros, D.R., Standley, L.J., Chen, X.J., Simoneit, B.R.T., 1999. tracers for atmospheric wood smoke pollution. Environ- Epicuticular wax compositions of predominant conifers of mental Science and Technology 22, 1191±1196. western North America. Zeitschrift fuÈ r Naturforschung 54c, Hawthorne, S.B., Miller, D.J., Langenfeld, J.J., Krieger, M.S., 17±24. 1992. PM-10 high volume collection and quantitation of Parham, R.A., Gray, R.L., 1984. Formation and structure of semi- and nonvolatile phenols, methoxylated phenols, wood. In: Rowell, R. (Ed.), Chemistry of Solid Wood, Adv. alkanes and polycyclic aromatic hydrocarbons from winter Chem. Series 207. American Chemical Society, Washington, urban air and their relationship to wood smoke emissions. DC, pp. 3±56. Environmental Science and Technology 26, 2251±2262. Petterson, R.C., 1984. The chemical composition of wood. In: Hildemann, L., Klinedinst, D.B., Klouda, G.A., Currie, L.A., Rowell, R. (Ed.), Chemistry of Solid Wood, Adv. Chem. Cass, G.R., 1994. Sources of urban contemporary carbon Series 207. American Chemical Society, Washington, DC, aerosol. Environmental Science and Technology 28, 1565± pp. 57±126. 1576. Prahl, F.G., Carpenter, R., 1983. Polycyclic aromatic hydro- Hornig, J.F., Soderberg, R.H., Barefoot III, A.C., Galasyn, carbons (PAH)-phase associations in Washington Coastal J.F., 1985. Wood smoke analysis: vaporization losses of sediment. Geochimica et Cosmochimica Acta 47, 1013±1023. PAH from ®lters and levoglucosan as a distinctive marker Pruel, R.J., Quinn, J.G., 1988. Accumulation of polycyclic for wood smoke. In: Cooke, M., Dennis, A.J. (Eds.), Poly- aromatic hydrocarbons in crank-case oil. Environmental nuclear Aromatic Hydrocarbons: Mechanisms, Methods, Pollution 49, 89±97. and Metabolism. Batelle Press, Columbus, pp. 561±568. Ramdahl, T., 1983. Retene Ð a molecular marker of wood IARC, 1989. World Health Organization, International Agency combustion in ambient air. Nature 306, 580±582. for Research on Cancer: Lyon, France, IARC Monographs Riederer, M., Schneider, G., 1990. The eect of the environ- 46, 41±155. ment on the permeability and composition of Citrus leaf IPCC, International Panel on Climate Change, 1990. In: cuticles. 2. Composition of soluble cuticular lipids and cor- Houghton, J.T., Jenkins, G.J., Ephraums, J.J. (Eds.), Climate relation with transport properties. Planta 180, 154±165. Change: The IPCC Scienti®c Assessment. University Press, Rier, R., Anderson, A.B., Wong, A., 1969. Terpenoid con- Cambridge, UK. stituents of the pocket resin from coast redwood (Sequoia IPCC, International Panel on Climate Change, 1992. In: sempervirens). Phytochemistry 8, 923±925. Houghton, J.T., Callander, B.A. and Varney, S.K. (Eds.), Rogge, W.F., Hildemann, L.M., Mazurek, M.A., Cass, G.R., Climate Change 1992: The Supplementary Report to the Simoneit, B.R.T., 1993a. Sources of ®ne organic aerosol: 2. IPCC Scienti®c Assessment. University Press, Cambridge, Noncatalyst and catalyst-equipped automobiles and heavy- UK. duty diesel trucks. Environmental Science and Technology Johnson, R.I., Shaw, J.J., Cary, R.A., Huntzicker, J.J., 1981. 27, 636±651. An automated thermal-optical method for the analysis of Rogge, W.F., Mazurek, M.A., Hildemann, L.M., Cass, G.R., carbonaceous aerosol. In: Macias, E.S., Hopke, P.H. (Eds.), Simoneit, B.R.T., 1993b. Quanti®cation of urban organic Atmospheric Aerosol: Source/Air Quality Relationships. aerosols at a molecular level: identi®cation, abundance and Amer. Chem. Soc., Symp. Ser. 167, Washington, D.C., pp. seasonal variation. Atmospheric Environment 27A, 1309± 223±233. 1330. Killops, S.D., Al-Juboori, M.A.H.A., 1990. Characterisation of Rogge, W.F., Hildemann, L.M., Mazurek, M.A., Cass, G.R., the unresolved complex mixture (UCM) in the gas chroma- Simoneit, B.R.T., 1998. Sources of ®ne organic aerosol. 9. tograms of biodegraded petroleums. Organic Geochemistry Pine, oak, and synthetic log combustion in residential ®re- 15, 147±160. places. Environmental Science and Technology 32, 13±22. Kolattukudy, P.E., 1970. Plant waxes. Lipids 5, 259±275. Schauer, J.J., Rogge, W.F., Hildemann, L.M., Mazurek, M.A., Locker, H.B., 1988. The Use of Levoglucosan to Assess the Cass, G.R., Simoneit, B.R.T., 1996. Source apportionment Environmental Impact of Residential Wood-burning on Air of airborne particulate matter using organic compounds as Quality. PhD thesis, Dartmouth College, Hanover, NH. tracers. Atmospheric Environment 30, 3837±3855. Lorbeer, E., Zelman, N., 1988. Investigation of the distribution Schlenk, H., Gellerman, J.L., 1960. Esteri®cation of fatty acids of the non-volatile lipophilic part of rosin in spruce (Picea with diazomethane on a small scale. Analytical Chemistry 32, abies). Holzforschung 42, 241±246. 1412±1414. Mazurek, M.A., Simoneit, B.R.T., 1984. Characterization of Schulten, H.-R., Simmleit, N., Rump, H.H., 1986. Soft ioniza- biogenic and petroleum derived organic matter in aerosols tion mass spectrometry of epicuticular waxes isolated from over remote, rural and urban areas. In: Keith, L.H. (Ed.), coniferous needles. Chemistry and Physics of Lipids 41, 209± Identi®cation and Analysis of Organic Pollutants in Air. Ann 224. Arbor Science/Butterworth Publishers, Boston, pp. 353±370. Sergejewa, A.S., 1959. Chemie des Holzes und der Cellulose. Mazurek, M.A., Simoneit, B.R.T., 1997. Higher molecular Theodor Steinkop Verlag, Dresden. 1544 D.R. Oros, B.R.T. Simoneit / Applied Geochemistry 16 (2001) 1513±1544

Sha®zadeh, F., 1984. The chemistry of pyrolysis and combus- Standley, L.J., Simoneit, B.R.T., 1994. Resin diterpenoids as tion. In: Rowell, R. (Ed.), Chemistry of Solid Wood, Adv. tracers for biomass combustion aerosols. Journal of Atmo- Chem. Series 207. American Chemical Society, Washington, spheric Chemistry 18, 1±15. DC, pp. 489±529. Stephanou, E.G., Stratigakis, N., 1993. Oxocarboxylic and a,o- Simoneit, B.R.T., 1978. The organic chemistry of marine sedi- dicarboxylic acids: photooxidation products of biogenic ments. In: Riley, J.P., Chester, R. (Eds.), Chemical Oceano- unsaturated fatty acids present in urban aerosols. Environ- graphy, 2nd Edition, Chapter 39. Academic Press, New mental Science and Technology 27, 1403±1407. York, pp. 233±2311. SuÈ mmchen, P., MarkstaÈ dter, C., Weinhaus, O., 1995. Compo- Simoneit, B.R.T., 1984. Organic matter of the troposphere. III: sition of the epicuticular wax esters of Picea abies (L.) Karst. Characterization and sources of petroleum and pyrogenic Zeitschrift fuÈ r Naturforschung 50C, 11±14. residues in aerosols over the western United States. Atmo- Takada, H., Onada, T., Harada, M., Ogura, N., 1990. Deter- spheric Environment 18, 51±67. mination of polycyclic aromatic hydrocarbons in urban Simoneit, B.R.T., 1986. Cyclic terpenoids of the geosphere. In: street dusts and their source materials by capillary gas chro- Johns, R.B. (Ed.), Biological Markers in the Sedimentary matography. Environmental Science and Technology 24, Record. Elsevier, Amsterdam, pp. 43±99. 1179±1185. Simoneit, B.R.T., 1989. Organic matter of the troposphere. V: Takada, H., Onada, T., Harada, M., Ogura, N., 1991. Dis- Application of molecular marker analysis to biogenic emis- tributions and sources of polycyclic aromatic hydrocarbons sions into the troposphere for source reconciliations. Journal (PAHs) in street dust from the Tokyo metropolitan area. of Atmospheric Chemistry 8, 251±275. Science of the Total Environment 107, 45±69. Simoneit, B.R.T., 1998. Biomarker PAHs in the environment. Thomas, B.R., 1970. Modern and fossil plant resins. In: Har- In: Neilson, A. (Ed.), The Handbook of Environmental borne, J.B. (Ed.), Phytochemical Phylogeny, Chapter 4.. Chemistry, Vol. 3, Part I. Springer Verlag, Berlin, pp. 175±221. Academic Press, London, pp. 59±79. Simoneit, B.R.T., Mazurek, M.A., 1982. Organic matter of the Tulloch, A.P., 1976. Chemistry of waxes of higher plants. In: troposphere-II. Natural background of biogenic lipid matter Kolattukudy, P.E. (Ed.), Chemistry and Biochemistry of in aerosols over the rural Western United States. Atmo- Natural Waxes. Elsevier, Amsterdam, pp. 235±287. spheric Environment 16, 2139±2159. Tulloch, A.P., 1987. Epicuticular waxes of Abies balsamea and Simoneit, B.R.T., Rogge, W.F., Mazurek, M.A., Standley, Picea glauca: occurrence of long-chain methyl esters. Phy- L.J., Hildemann, L.M., Cass, G.R., 1993. Lignin pyrolysis tochemistry 26, 1041±1043. products, lignans and resin acids as speci®c tracers of plant Zinkel, D.F., Clarke, W.B., 1985. Resin acids of Pinus resinosa classes in emissions from biomass combustion. Environ- needles. Phytochemistry 24, 1267±1271. mental Science and Technology 27, 2533±2541. Zinkel, D.F., Magee, T.V., 1987. Diterpene resin acids from the Simoneit, B.R.T., Schauer, J.J., Nolte, C.G., Oros, D.R., Elias, needle oleoresin of Pinus strobus. Phytochemistry 26, 769± V.O., Fraser, M.P., Rogge, W.F., Cass, G.R., 1999. Levo- 774. glucosan, a tracer for cellulose in biomass burning and Zinkel, D.F., Magee, T.V., Walter, J., 1985. Major resin acids atmospheric particles. Atmospheric Environment 33, 173±182. of Pinus nigra needles. Phytochemistry 24, 1273±1277.