Document KRK77d5gYzzwmxy9Gdk2K3VJK

, 1812 A M . chem.1881. 53,1612-1620 Chemical Class Separation and Characterization of Organic Compounds in Synthetic Fuels Douglas W. Later, Milton L. Lee,' Kelth D. Bartle,' Robert C. Kong, and Danlel L. Vadlaros Oepertment of Chemistry, mrn Young lhivm&, Row, uteh 84602 I k separatkn method k described for the Mentlfkatlon of organlc compounds in g n t b t k fuel ploductr. Prdractiona- solvent fractionation scheme to analyze coal liquefaction producta before and after hydmtreatment. They reported the tlon of crude synfuel matedab into dlscrete chemical dames occurrence of neutral polycyclic aromatic hydrocarbons a d was perf- by adsorption cohnnn chromatography using sulfur,oxygen,and nitrogen heterocyclic aromaticcompounds m a l quantitk d neutral ahmhrum oxide and rllkk acld. Subsequenl hlgh-resdutlon separatkn of brdMduai components was achieved by using capillary cokmn gas chromatography, and rpedtlc compound types were determined by gas chromatographic retentlon data and combhd gas chr+ matagraphy-msrr spectrometry. The prkrclpal chomkal classes Investigated in a sokent-refhd coal llqukl were aC In another separation scheme developed for solvent-refind coal (12),nine fractions were isolated from silica gel on &e basis of chemical functionality. Schiller and Mathiason (13) reported the development of a chromatographic procedure using neutral alumina to fractionate a wide variety of coal. derived solids and liquids for analysis by maas spectrometry and other spectral methods. The principal compound types investigated were saturated hydrocarbons, aromatic hydro- lphatk hydromrbom, pdycydlc aromatic hybocarbonr, lycycllc aromatk sultw heterocycles, nHrogm pdycyclk carbons,benzofurans, ethers, nitrogen compounds, and hydroxyl compounds. A combination of ion exchange, coordi. aromatic compounds, and hydroxyl pdycyck aromatic hy- nation, and adsorption chromatography was employed by drocarbons. The nltrogen-containlng aromatk compounds Jewell et al. (14) for the separation of petroleum distillah. were further separated into secondary nitrogen polycyclk This method has also been used to separate coal-derived aromatk heterocycles, amlno polycyck aromatic hydro- products into five fractions: acids, bases, neutral nitrogen carbons, and tertiary nitrogenpdycydlc aromatic heterocy- compounds, saturated hydrocarbons, and aromatic hydro- cles to f8diitate thek Identifkatlon. carbons. This method is commonly known as the S a technique and has been widely used. Several problems exist with the previously described sep Current emphasis on the development of alternate energy sources has stimulated production of synthetic fuels derived from oil shale, tar sands, and coal. Although technology for aration methods. First, most schemes are complicated, time-consuming,tedious,and require highly trained personnel to execute each step with precision and accuracy. Long periods producing liquid and solid fuels from these feed stocks has of time are required for pretreatment, cleaning, preparation, been available since the early 1900s (I),the chemical char- and activation of resins and adsorbent packings. Many acterizationof these produds has recently received increased attention. There are several reasons for determining specific methods use expensive materials, and large volumes of highpurity solvents are necessary for each chromatographic or compound types in synfuels. The identification and elimi- partitioning step. Solvent extractionand partitioning of polar nation of specific toxic and carcinogenic compounds would types of compounds can cause formation of tars and stable reduce the environmental and occupational health hazards associated with the production and combustion of these emulsions (15)which contribute to the loes of componente and inefficient separations. Finally, many separation schemes are materials (2-4). Certain heteroatom species are known to decrease the efficiency of catalytic processes and contribute hindered by incomplete chemical class separation and overlap of compound typesinto adjacent fractions. Improvement of to the instability of liquid fuels during storage (5,6). Synthetic fuels are also a valuable source of chemical materials for the manufacture of pharmaceuticals, pesticides, herbicides, and existing separation technology is necessary if faster analysie and better evaluation of the numerous types of synfuel products are to be achieved. dyes (3, 7). Because of the complex nature of synthetic fuel mater&, In this paper, the development of a two-step separation method using neutral alumina and silicic acid adsorption fractionation according to chemical class is usually required chromatography is described. This procedure minimizes the before identification of individual componentscan be achieved. problems previously discussed and is rapid, reproducible, Traditionally, separation schemes have been developed by using solvent partitioning methods and column chromatog- economical, and efficient. Crude solid and liquid synfuel products can be fractionated into seven chemical classes by raphy. The organic analytical group at the National Bureau this method. This separation scheme can be completed in lesa of Standards has isolated neutral oils, acids, and bases using a solvent extraction method with subsequent high-perform- than 6 h and uses less than 500 mL of solvent and approxi- mately 12 g of adsorbent packing. The applicability of thie ance liquid chromatographic (HPLC) and gas chromato- fractionationmethod is demonstrated by the characterization graphic analysis to separate and identify individual organic of a solvent-refined coal (SRC) liquid heavy distillate. compounds in synfuels (8). The analyticalgroup at Oak Ridge National Laboratory has developed a fractionation scheme using an acid-base extraction followed by solid-liquid chromatography on Sephadex LH-20,silicic acid, and basic alumina (1,9,IO). They reported the separation of aliphatic and aromatic hydrocarbons aa well as nitrogen heterocyclic and polar aromatic compounds. Wilson et aL (11) used a similar EXPERIMENTAL SECTION An SRC II heavy distillate sample (boding point range '2604d "C) from the Fort Lewis, WA, pilot plant (using West V w coal fromthe Pittsburgh seam) WBB provided by Pacific N o r t h d Laboratoryfor analysis his mate^ is of pilot plant origin should not n d y be considered representative of c o d scale production. A schematicdiagram of the separation method is shown in Figure 1. First, neutral aluminum oxide (Brockmaa ANALYTICAL CHEMISTRY. VOL. 53, NO. 11, SEPTEMBER 1981 1013 TSAMPLE NEUTRAL' AWINA A II II Wne BenZwp Mom Elhand Telrahydroluan Emno1 1111 --0 1. Chemical class separation scheme for synthetic fuel prod- &: neubalpolVcyclicarornatk (PAC), nitrogen polvCycrC iromatlc compounds (N-PAC), hydroxyl polycyclic aromatic hydro- -arbons (HPAH). polycyclic aromatic oxygen heterocycles (PAOH), 3o)ycyck aromatic sulhr heterocycles (PASH), secondary nitrogen m U c heterocydes (2'PANH). amho polvcyclc a r m & Wocarbons (APAH), and tertiary nitrogen polycyclic aromatic het- VOCyCles (3O-PANl-I). diphatic hydrocarbons; A-2, neutral polycyclic aromatic com- ~ u n d a(PAC); A-3, nitrogen polycyclic aromatic compounds N-PAC); and A-4, hydroxy polycyclic aromatic hydrocarbons HPAH). [Theacronyms used in thispaper are assignedaccordmp `1)the format deked by M e et aL (16).]Approximately 0.1-0.3 < of crude sample was d w l v e d in a few milliliters of chloroform and adsorbed onto 3 g of neutral alumina. The solvent was removed from the alumina by vigorously stirring the mixture mder a gentle stream of dry nitrogen gas. The alumina with W p l e was then packed on top of an 11-mm i.d. column which u y contained 6g of neutral alumina The sample was then eluted with the following chromatographicgrade solvents fraction A-1,20 mL of hexane; fraction A-2,50 mL of benzene; fraction A470 mLof Chloroform(containing0.75% ethanol preservative); M o nA-4,50mL of 10%ethanol in tetrahydrofuran (containing no inhibifora). pdycyclicaromaticoxygen heterocycles(PAOH) and polycyclic -tic sulfur heterocycles(PASH)are eluted with the polycydic -tic hydrocarbons (PAH) in fractionA-2. Becawthe PASH such low concentration in coal-derived materials, the alucolumn procedure wa8 scaled up (100g of neutral alumina) 10g of coal liquid (preadsorbed onto 50 g of neutral alumina) used to obtain a neutral PAC fraction. The PASH were then *Paratad from the PAH and PAOH by a procedure previously ':Ported by Willey et al. (17). Briefly, the PASH were oxidized mth H a 2to their correspondingsulfones, the sulfoneswere then *parated from the other unoxidized neutral PAC on silica gel fractionA-2a), reduced with LiAlH, to the original PASH and -Purified on silica gel (fraction A-2b). ssveral dietinctclasses of nitmgen heterocyclic compoundswere P m n t in the initial N-PAC, fraction A-3. Consequently, a was developed to subfractionate the N-PAC into three additional classes of compounds using silicic acid adsorption "-hromatogrphy.The alumina N-PAC fraction was adsorbed onto 'd g ofailicicacid (MaUinckrodt No.2&17,100-meshpowder, used *F i v e d from supplier) and the solvent removed by stirring muture under dry nitrogen gas. A 22-mm i.d. column was with a hexane slurry of 2 g of silicic acid. The silicicacid the adsorbed sample was then packed on top of the 2-g and eluted to give the following fractions and compound hfraction S-1, secondary nitrogen polycyclic aromatic (2O-PANH)with 50 mL of 1:l (wv) hexane:benzene; fraction S-2, enriched amino polycyclic aromatic hydrocarbone (APAH)with 30 mL of benzene; fraction S-3, tertiary nitrogen polycyclic aromatic heterocycles (3O-PANH)with 50 mL of 1:l (v:v) benzeneanhydrous ethyl ether. Each of the alumina and silicic acid fractions was concentrated to approximately 2 mL on a rotary evaporator, taken to dryness under nitrogen gas, weighed, and dissolved in an appropriate volume of benzene for capillary column gas chromatographic analysis. A Hewlett-Packard 5880 gas chromatographequipped with a fused-silica capillary column (20m x 0.30 mm i.d.) coated with SE52 methylphenylsiiconestationary phase (film thickness 0.25 am) was used to obtain chromatograms of each fraction. The oven was held at 50 OC for 2 min during injection and then temperature programmed at 3 OC/min (alumina fractions)or 4 OC/min (PASH and silicic acid fractions) to a final temperature of 260 "C and held for 5 min. A flame ionization detector (FID) was generally used,but nitrogen and sulfur heterocycle fractions were also analyzed by using a nitrogen-selective detector (NPD) and a sulfur-selectivedetector (FPD). A Perkin-Elmer Sigma 2 gas chromatograph equipped with these detectors was used. Identificationof specificcompoundswas accomplished by usmg gas chromatographicretention data and combined gas chroma- tography-mass spectrometry (GC-MS).Gas chromatographic conditions were similiar to those already described. A Hewlett-Packard 5982A quadrupolemass spectrometerwas operated in the electton impact mode (70eV electron energy). Spectra were acquired and processed with a Hewlett-Packard 5934A data system. Infrared spectrometry of the HPAH fraction was performed by evaporatingthe solvent from the sample and placing an aliquot between two salt plates. The sample was then analyzed by scanning from 4000 to 600 cm-' on a Beckman ACCULAB 2 infrared spectrometer. RESULTS AND DISCUSSION The separation of oxygen and nitrogen heterocyclic com- pounds in petroleum produds by adsorption chromatography on alumina has been extensively studied by Snyder and Buell (18). Similarly,Schiller (13.19)used alumina chromatography to develop a rapid, simple separation method for heterocyclic compounds in synthetic fuels. The alumina step reported in this paper is a modified version of the method developed by Schiller. It was found in this study that column chromatography on neutral alumina minimized sample loas due to irreversible adsorption, eluted a wide range of solutes rapidly, and provided reproducible class separation according to the functionality of the heteroatom present in the polycyclic aromatic structure. Compounds which containnitrogen heteroatomsof different functionality can be separated on silicic acid by varying the eluent polarity. Silicic acid has previously been used to separate indoles and carbazoles from other N-PAC in tobacco smoke (20)and synfuels (9). The silicic acid procedure reported in thispaper was developed to complementthe alumina method by providing an e q d y simple and rapid method for subsequent chemical class separation of the nitrogen-containing polycyclic aromatic compounds. Using a small quantity of adsorbent (2.5 g) reduces the contact time of the sample components with the silica structure and decreases degradation, artifact formation, and sample loss n o r m d y associated with chromatography on silica adsorbenta. Numerous compounds are present in each chemical class, even after extensive prefractionation, and further high-res* lution separation is required prior to the identification of specific components. The application of capillary column GC-MS to resolve and determine compounds in complex mixtures of PAC has been well established (21). This simple alumina-silica fractionation scheme combined with capillary column GC-MS provides a rapid, effective, and reproducible method for the detailed chemical characterizationof synthetic fuels. Additionally,thistwo-step separation method can d y be scaled up to produce larger quantities of sample which ia 1014 ANALYTICAL CHEMISTRY, VOL. 53. NO. 11, SEPTEMBER 1981 --_ -2. ~ ~ ~ C h m m e ~ o f a p h e t i C ~ f r e c t b PneaAkn-un1ba.srefertocanpoudsbtsdhTabbI!!. Con were 20 m X 0.30 mm fuwd4ca caplllarv cokwm coated wlth SE-52, temperetwo proqemnedfrm 40 OC to 250 OC at 3 OC/min. ,EEYPERATUIEICI TIME l m n l , 0 50 10 00 20 3a 140 , 200 50 $0 250 7b Flgwr 3. Capillary column gas chromatogram of PAH fraction A-2. Peak numbers refer to mmpomds lsted In Table 111. Conditions in Figwe 2. sometimes requiredfor the concentration of trace components and for bioassay to determine mutagenicity or carcinogenicity. The quantitative data obtained by using thisprocedure are presented in Tables I and II. A total recovery of greater than 90% was obtained by this method for the separation of an SRC IIheavy dmtillate into chemical classes. A chroniatogram of each of the fractions is shown in Figures 2-9. Tables ID-V list the compound types identified in each fraction. Aliphatic Hydrocarbons. Due to the nonpolar nature of the aliphatic hydrocarbons, this class of compounds elutes immediately from alumina with hexane. The gas chromatographicretentiontime of standard octadecaoe, n-Cle,was used to determine n-Cl8 in the aliphatic fraction. Mass spectral data supplemented the identification of n-C18and were used to determine the other constituentsin the homologous series of straight-chain hydrocarbons. The nonlinearity of the base line in the n-C16-n-Cpregion (see Figure 2) is caused by the numerous unresolved components (related olefins and branched isomere of the straighkhain parent hydrocarbons). The relativelylow mncentrationof the aliphatic hydrocarbone in thiscualliquid illustratesthe highlyaromaticnature of thia product. The structural features of saturates in coal liquids are similar to those of some petroleums and both can be T&!e 1. Weight Percent of Neutral Alumina Fractior from an SRC I1 Heavy Distillate Coal Liquid fraction compound type fractiona w A-1 aliphatic hydrocarbons A-2 neutral PAC A-3 N-PAC A-4 HPAH 5.5 56.6 20.0 -14.0 total 96.1 a Average values for three different assays of SRC I1 heavy distillate. Table 11. Weight Percent of Silicic Acid Subtractions h m an SRC I1 Heavy Distillate Coal Liquid wt 9% 0 wt % heavy fraction compound type of N-PAC distillatt S-1 2"-PANH 5-2 enrichedAPAH 5-3 3"-PANH 59.6 8.5 -14.9 11.9 1.7 -3.0 total 83.0 16.6 -AllSEICi I 50 100 rio 200 2x) T*tfnnl r 0 lo 20 Jo 40 so F W 4. CapHlary column gas chromatagram of PASH fraction A-2b. Peak numbers refer to compounds listed in Table IV. conditions were 20 m X 0.30 mm fused-sib capihry column coated with SE-52, temperature programmed from 40 OC to 250 OC at 4 O C l m l n . ----I -TuIErci TDilE(nkl I , 0 do rb do x) L 'b1. CaPRary cdum gas chromatopm of W A C fractkm A-3. 3s In Figue 2. 150 200 250 UI &J do h Peak numbers refer to mmpwnds listed In Table 111. condtknswdlo PAC fraction eluted from the alumina column. Benzene was ehoeen over toluene aa the eluent becaw it ia a slightly more ~ I aerolvent and reducedthe overlap of the polycyclic furans PAOH) intothe chloroform fraction. The PAH (see Figue 4 are the chemical clam of highest concentration in this SRC I heavy dietillateand generally are mqjor constituents in most a-derived fuel producta. The literature in this area ie too Xbmive to cover here but hae recently been reviewed (23, 1). The high degree of alkylation in thie PAH fraction is of uticular interest. A rough correlation between degree of blation of PAHstructures in energy-related materiala and UtageniciQ haa been reportedby Grieet and co-workers (25). rtenaive alkylation ia observed in all fractions of this SRC liquid, and in some inetancesthe alkylated speciea are more 'Udant than the parent PAC. Although the PASH, Figure 4, comprise less than 1% of e totalheavy distillate, the detailed characterization of the Vur-containing compounds in synthetic fuel products is Portant in light of recent evidence that specific sulfur terocyclee have been shown to poeee%s signifcant mutagenic ti* (~627)F.urthermore,the r e m d of sulfurfrom fuels desirable to prevent the formation of noxious sulfur gaaes l h g combustion and to prevent catalyst poisoning in coal conversion reactors. The identification of specific sulfurcontaining speciesin synthetic fuels cau provide insightsinto better methods of sulfur removal. N-PAC. The elution of the N-PAC from neutral alumina with chloroform could be visually obeerved by the migration of a dark band down the column. Analysis by gas chromatography using a nitrogen-selective NPD verified that essen- tially all of the componenta in thie fraction were nitrogen- containingcompounds. Figure 5 is an FID chromatogram of, the N-PAC fraction. Primary (APAH), secondary (2O-PANH), and tertiary (3O-PANH) nitrogen heterocycles (see Figwea 6-8) occur collectively in the alumina N-PAC fraction. Additional chemical Cleas sepamtionwas essentialfor threereamna Firet, the difficulty of identifying individual constituents in this fraction was increased by the fact that more than one compoundeluted from the capillary column at the same time. For example, tetrahydrocarbazole coeluted with two b e n m quinoline isomers. This was also the case with many of the alkylated c a r h l m alkylated ben7.oquinOline8, and threering APAH. Second, different concentration levela between chemical classes ohcured g r o u p of compounds which are important to the overall chemical behavior of the sample. . -I,,. I m '1 li 1616 ANALYTICAL CHEMISTRY. VOL. 53, NO. 11. SEPTEMBER 1981 -Table 111. Compounds Identified in the Alumina Chromatographic Fractions of an SRC I1 Heavy Distillate Coal Liquid peak no. mol wt compound peak no. mol wt compound 1 226 2 240 3 254 4 268 5 282 6 296 7 310 Aliphatic Hydrocarbon Fraction A-1 (Figure 2) 8 324 9 338 10 352 11 366 12 380 13 394 14 408 PAC Fraction A-2 (Figure 3) 1 142 C,-naphthalene 14 196 C,-biphenyl/C,-dibenzothiophene 156 C,-naphthalene 1 5 204 1-phenyInaphthalene 170 C,-naphthalene 16 192 C,-phenanthrene 154 biphenyl 17 204 2-phenylnaphthalene 154 acenaphthene 18 206 C,-phenanthrene 168 dibenzofuran 1 9 202 fluoranthene 2 166 fluorene 20 202 PYrene 3 168 C,-biphenyl/C,-acenaph thene 21 not identified 4 5 168 184/180 CC,,--nbaipphhtehnayleln/eC/,C-a,c-felnuaoprehntheene 22 216 23 216 benzo[a]fluorene benzo[ b ]fluorene 6 182 C,-dibenzofuran 24 216 C,-pyrene/C,-fluoranthene 7 182 C,-biphenyl 25 230 C,-pyrene/C,-fluoranthene 8 1801182 C,-fluorene/C,-biphenyl 26 228 benz[a ]anthracene 9 1981194 C,-naphthalene/C,-fluorene 21 228 chrysene 10 184 dibenzothiophene 28 242 C,-benz [ a ]anthracene/C,-chrysene 11 182 tetrah ydrophenanthrene 29 252 benzofluoran thene 12 196 C,-biphenyl 30 252 benzopyrene 13 178 phenanthrene N-PAC Fraction A-3 (Figure 5) 1 131 C,-indole 19 193/207 C,-benzoquinoline/C,-benzoquinoline 2 157 3 145 4 159 C,-quinoline C,-indole C, -indole 20 181 21 193 22 207 C,-carbazole C,-benzoquinoline C,-benzoquinoline . 5 198 C,-diphenyl ether 23 195 C,-carbazole 6 185 C,-quinoline 24 207 C,-benzoquinoline 7 183 tetrahy drobenzoquinoline 25 195 C,-carbazole 8 178 phenanthrene 26 195 C,-carbazole 9 179 benzoquinoline 27 209 C,-carbazole 10 179 benzoquinoline 25 203 azafluoranthene 11 197 C,-tetrahydrobenzoquinoline 29 203 azapyrene 12 171/179 tetrahydrocarbazole/benzoquinoline 30 223 C4carbazole 13 197 C,-tetrahydrobenzoquinoline 31 2171237 C,-azapyrene/C,-azafluoranthene/C,-carbazole 14 167 carbazole 32 231 C,-azapyrene/C,-azafluoranthene 15 193 C , -benzoquinoline 33 221 tetrahydrobenzocarbazole 16 185 C,-tetrahydrocarbazole 34 229 naph thoquinoline 17 193 C,-benzoquinoline 35 217 benzocarbazole iii 181/193 C,carbazole/C,-benzoquinoline 36 231 C ,-benzocarbazole HPAH Fraction A-4 (Figure 9) 1 148 C,-indanol 17 198 C,-hydroxybiphenyl 2 170 3 162 hydroxy biphenyl C,-indanol - 18 19 182 182 hydroxy fluorene hydroxy fluorene 4 184 C,-hydroxy biphenyl 20 198 C,-hydrox ybiphenyl 5 184 C,-hydroxybiphenyl 21 152 hydroxyfluorene 6 176 7 198 C,-indanol C,-hy droxy biphenyl 22 198 C,-hydroxy biphenyl 23 2261196 C,-h ydroxybiphenyllc ,-hydroxyfluorene 8 170 9 170 hydroxy biphenyl hydroxy biphenyl 24 25 212 212/196 C,-h ydroxybiphenyl C,-hydroxy biphenyllc ,-hydroxyfluorene 10 190 C,-indanOl ;26 220 naphthylphenol 11 212 C,-hydroxy biphenyl , 2 7 2261210 C,-hydroxybiphenyl/C,-hydroxy fluorene 1 2 1841204 C,-hydroxy biphenyl/C,-indanol 28 234 C,-naphthylphenol 13 184 C,-hydroxy biphenyl 29 2401224 C,-hydroxybiphenyl/C,-hydroxy fluorene 14 184 C,-hydroxy biphenyl 30 220 naphthylphenol 1 5 212 16 182 C,-hy droxybiphenyl hy &oxy fluorene 31 220 32 234 nca.p-nhatDhyhlthpYhelDnohlenol Although the APAH were lower in concentration than either the secondary or tertiary PANH (see Table 11), it has been suggested that the APAH are largely responsible for the biological activity of synfuel products (9,28). Third, detailed assignment of structure wa.s c.om* p."licated by the presence of anthracenes have molecular weights of 193 amu md 9 fragment similarly by mass spectrometry making exad structural identification nearly impoeeible. Therefore,it WH neceseery to separatethe nitrogemcontainingcornpouwh ..-.subfractions. 7,.I - c. .r I L ~ : - ,nnl 9 I I i i' I 'i I 'i I I 19 I- TUPERAlWIEIC' 50 loo 150 100 25L) TlYEImrl , 0 10 20 a 4a - 50 6. CapMary column gas chromatogam of 2'PANH fraction S 1 . Peak numbers refer to compounds listed in T a w V. conditions were W h FbWe 4. c i ,rapER*rrrirci so 100 do ZOO 2y, 0'W I n h 1 m a0 30 4 0 so c(su,7. CapiJlary colwnnges chromatogam of enriched APAH fractkm 5 2 . Peak numbers refer to compounds kted in Table V. ChdtkUM *has in Figure 4. T*le Iv. CompoundsIdentified in the PASH Subfraction A-2b (Figure 4) of an SRC I1 Heavy Distillate Coal Liquid Peakno. mol wt compound peakno. molwt compound 1 166 fluorene 2 180 C,-fluorene 9 202 fluoranthene 10 226 C,-dibenzothiophene 3 188 tetrahydrodibenzothiophene 11 208 phenanthro[4,5-bcd]thiophene 4 184 dibenzothiophene 5 178 phenanthrene 12 13 202 222 pyrene C,-phenantho[ 4,ti-bcd ]thiophene 6 178 anthracene 14 234 benzonaphthothiophene 7 198 C,-dibenzothiophene 15 248 C,-benzonaphthothiophene -tion and identification of several pyridines and anilines; !'umber of two- and three-ring N-PAC have also been dentifinidsynthetic fuel products (7,23,19,29-33).Rel- fewer studies have been reported on the isolation and ion of compounds containing a larger number of one- and two-ring N-PAC systems are of general interest because they are present in higher concentration levels, are more volatile, some are commercially useful, and a few are biologically active. However, it is believed that isomers of increasing ring number generally display a higher degree of mutagenicity and carcinogenicity. For example, carbazole is noncarcinogenic,benzo[a]carbezole is moderakb 1818 ANALYTICAL CHEMISTRY, VOL. 53. NO. 11, SEPTEMBER 1981 -9 56' TIy(mn) , 0 lo m 30 a 50 Flgwr 8. Capillary column gas chrometogram of 3'PANH fraction 53. Peak numbers refer to compounds listed in Table V. COndluOns were as in Figure 4. Table V. N-PAC Compounds Identified in the Silicic Acid Subfraction of an SRC I1 Heavy Distillate Coal Liquid - peak no. mol wt compound peak no. mol wt compound 1 145 2 159 3 184 4 173 5 198 6 178 7 171 8 212 9 167 10 167 2"-PANH Fraction S-1 (Figure 6) C,-indole 11 167 carbazole isomer C,-indole 1 2 181 C,carbazole C,-diphenyl ether 13 195 C,-carbazole C,-indole 14 209 C,-carbazole C,-diphenyl ether 1 5 223 C,carbazole phenanthrene 16 237 C,-carbazole tetrahydrocarbazole 17 217 benzocarbazole C,-diphenyl ether 18 217 benzocarbazole carbazole 19 231 C,-benzocarbazole carbazole isomer 20 245 C,-benzocarbazole 1 143 2 169 2 183 4 197 5 181 6 179 7 167 8 193 9 193 1 0 207 Enriched APAH Fraction S-2 (Figure 7) naphthylamine 11 221/207 C,-benzoquinoline/C,-aminophenanthrene aminobiphenyl 1 2 235 C,-benzoquinoline C,-aminobiphenyl 1 3 233 tetrahydronaphthoquinoline C,-amino biphenyl 14 217 aminopyrenelaminofluoranthene aminofluorene 1 5 233 tetrahy dronaphthoquinoline benzo[h]quinolinea 16 229 naphthoquinoline (possibly benz[ c ]acridinea) carbazole C,-benzoquinoline 17 231 18 229 C,-aminopyrene/C,aminofluoranthene naphthoquinoline (possibly benzla lacridinea) aminophenanthrenelaminoanthracene C,-benzoquinoline l290 21 224340 benzan'thione C,-benzanthrone phthalate ester 3"-PANH Fraction 5-2(Figure 8) 1 157 C -quinoline . 8 197 C,-tetrahydroquinoline 2 171 C,-quinoline 9 193 C,-benzoquinoline 3 18511581183- C,-quinolinelunidentifiedpeaks 1 0 2071203 C,-benzquinolinelazafluoranthene 4 199 C,-quinoline 11 221 C,-benzoquinoline 5 183 6 179 tetrahy droquinoline acridinea 1 2 203 13 217 azapyrene C, -azapyrene/C, -azafluoranthene 7 179 phenanthridinealbenzo[f]quinolinea 14 229 naphthoquinoline a Possible structures assigned by retention data comparisons with standard compounds. carcinogenic,and dibenzo[cg]carbazole is a potent carinogen .(34). Therefore,, it. is important -1 to isolate and identify the mostly of two-, three-, and four-ring N-PAC. However, w d has been. ,d.one with. s.ta.ndard N-.PAC, of up t.o fiv.e r i n p @; c TEWERATLRETC I 50 do TI= I- I 0I io 20 30 F@ra 0. Capillary column gas chromatogram of HPAH fraction A-4. as in F w e 2. 1% xy) 25u 40 5a 60 10 Peak numbers refer to compounds listed in Table 111. Conditions were The enriched APAH fraction, Figure 7, is of particular interest. Wilson et al. (28) initially suggested the presence of two-, three-, and four-ring primary aromatic amines in a similar SRC I1heavy distillate. The results presented in this paper confirm the presence of these APAH and also report the Occurrence of several alkylated species of the parent AF'AH. The APAH were not completely separated from the 3O-PANH. Preliminary data suggest that the tertiary nitrogen heterocycles overlapping into the APAH fraction are either alkylated or angular-structured species where the lone pair of electrons on the nitrogen heteroatom is shielded. Therefore, some of the 3O-PANH and the APAH have similar retention characteristics. Many of the 3"-PANH that eluted in this fraction were more than twice as concentrated as the APAH and frequently obscured their presence. Thus, in addition to the retenti n and GC/MS data used, a derivatization method has b$.?nused (35)to ensure positive identification of the APAH listed in Table V. Although there was overlap between these two classes of N-PAC compounds in this fraction, the APAH were sufficiently enriched to verify their PRaence. WAH. The hydroxyl PAH (see Figure 9) were eluted last from neutral alumina with 10% ethanol in tetrahydrofuran. b y HPAH have been shown to be present in synthetic fuel Productsand several methods have been developed for their mhtion and determination (8,29,30,3f$. The major acidic compoundsfound in synfuels are phenol and the alkylphenols. &, HPAH such as hydroxybiphenyls,indanols, naphthols, acenaphthenols, and hydroxyfluorenes have been reported to in these matrices (29). Infrared spectrometry of the HPAH fraction showed the presence of hydroxyl functional groups and indicated there was little if any carbonyl oxygen. ~ h i oabservation agrees well with others who have reported the absence of *bony1 group in coal liquefaction products (29).The major COmPonentsidentified by GC-MS in this fraction were inh o b , hydroxybiphenyls, hydroxyfluorenes,and naphthyl- phenols. Low levels of naphthols were also detected. As with theOther fractions of thiscoal liquid, a hgh alkyl content was ob-ved. The Projected increase in use of synthetic fuels produced coal, shale, and tar sands to meet growing energy de- mands has createda need for faster and more reliable methods Of upis of products and byproducts from these processes. %methodology presented in this paper and applied to the wY6ai of a representativesolvenbrefinedcoalliquid has been to be a rapid, effective, and reproducible means to a - P h e chemical analysis of these materials. Although not *, this separation method has been applied to the analysis of other SRC I1 products, coal tar, SRC recycle solvents, and SRC I process materials. In each case, the compounds in these materials were successfully separated according to chemical functionality as described for the SRC I1 heavy distillate. ACKNOWLEDGMENT The authors thank Robert M. Campbell for his technical assistance. LITERATURE CITED (1) Rubin. I. B.; Ouerin, M. R.; Hardlgee. A. A.; Em,J. L. Evton. R a . 1976. 72. 358-365 (2) Sexton, R. J. Arch Enwiron. Heetm 1910, 7 , 208-231. (3) Dong, M.; schmeltz, I , Lavole, E.; Hoffmann. D. In "Carclqpnesls. vol. 3: Polynuclear Aromatic HyxkocarboM";Jones. P. W.. Fruden- mi.R. I., ~ s .~; h v R~-:n bwyak. 1978; w ga-108. (4) Koppenhal. D. W.; Manahan, S. E. Emton. Sd.TectmoI. 1976. TO, 1104-1107. (5) Drushel. H. V.; Samwmr.A. L. AM/.chem.1966. 3. 19-28. (6) Harhmg. G. K.; &we#. D. M. AMI. Wm. Acta 1902, 26. 514-547. (7) White. C. M.; Schwdgherdt, F. K.; Schultz. J. L. F u d R U c e s S . T&- nol. 1971. 7. 209-215. (8) Hertz. H. S.;Brown, J. M.; Cheskr, S. N.; Quenther, F. R.; L. W.: May, W. E.; Pards, R. M.; Wlse, S. A. Anel. Chem. 1980, 52, w.1650-1657. (9) b,C.+(.; clerk. 8. R.; Gusrh. M. R.; Ma, C. Y.; Rao. T. K. Pep.. Am. Chem. Soc., Uiv. Fud Chem. 1979, 24 (1). 281-291. (10) J.o~nes. A. R.: Querh. M. R.: Clark. 8. R. AMI. Chem. 1977. 49, 1766-1771. . wi)son. B. W.: PetcKson. M. R.; Petw. R. A.; Crmto, J. T. Fud1911. Bo. 289-294. Farcadu, M. Fud 1977. 56 (1). 9-14. schikr, J. E.; Mathlason. D. R. Ami. chem.1977. 49. 1225-1228. bel.D. M.; W h . J. H.; eunpSr. J. W.: plenchw. H.; Latham. 0. R. AMI. chem.1972. 44. 1391-1395. caw, B. R.; ~ oc,. H.; ones. A. R. A&. chem.ser. i s m No. 170, 282-294. Bartle, K. D.; Lee. M. L.; Wlse, S. A. chem.Soc. Rev. 1911. 35. 113-158. m y , C.; Iwao, M.; Castb, R. N.; Lee, M. L. A M I . C h m . 1981. 53, 400-407. Snyder. L. R.; Bud. B. E. AMI. chsm.1068, 40, 1295-1302. SCMller. J. E. AMI. chem.1977. 49. 2292-2294. Snook. M. E.; Amandfib. R. F.; I 4 p s n . H. C.; Chortyk, 0. T. Am/. chem.1978, 50. 68-90. Lee. M. L.; Wright, B. W. J. clwonretoq. Scl. 1980, 18. 345-357. W e . C. M.; Sh&, J. L.; Sharkey. A. G Mtuu(L0ndon) 1977. 268. 620-623. Querln. M. R. In "FWycycUc Hyxkombons and Cancer"; Qdboln. H. V.. TS'O. P. 0. P.. Ms.;Academlc Press: New YON, 1978; Parl I. White, C. M. In "Handbook of Polycyck Aromatk Hy&ocabm"; Bj0rseth. A., Ed.; Wcd Dddtw: New York. h press. @lestt.W. H.; Tomkhs. B. A.; Epbr. J. L.; Rao, T. K. In "Polyw9w Aranetk Hvaouvbars"; Jones. P. w., Lsber. P.. E&.; Ann kba w.Sdence Phlkham Ann Arbor, MI. 1879; pp 395-409. Karchet. W.; Nelen R.; Depaus. R.; van J. In "Roceedlngs of 5th Intemetkmal Symposkm on Pdynudeer Aroma* Hydrocmbod'; Bsttdk C o b b u s Lsboratorks: bhmbus, OH, 1980. Way.R. A.. PadRc Northwest Laboratory. RlcNsnd, WA. 1981, uc PUMIahed work. WUson. B. W.; Pckoy, R. A.; CrWo. J. T. U t . Res. 1980. 79. 193-202.