Anal Bioanal Chem (2003) 375 : 381–388 DOI 10.1007/s00216-002-1698-8

O R I G I N A L PA P E R

Dawit Z. Bezabeh · Holly A. Bamford · Michele M. Schantz · Stephen A. Wise

Determination of nitrated polycyclic aromatic hydrocarbons in diesel particulate-related standard reference materials by using gas chromatography/mass spectrometry with negative ion chemical ionization Received: 8 July 2002 / Revised: 31 October 2002 / Accepted: 13 November 2002 / Published online: 16 January 2003 © Springer-Verlag 2003

Abstract Gas chromatography/mass spectrometry (GC/ MS) with negative ion chemical ionization (NICI) detection was utilized for quantitative determination of nitrated polycyclic aromatic hydrocarbons (nitro-PAHs) in diesel particulate-related standard reference materials (SRMs). Prior to GC/MS analysis, isolation of the nitro-PAHs from the complex diesel particulate extract was accomplished using solid phase extraction (SPE) and normal-phase liquid chromatographic (LC) fractionation using an amino/cyano stationary phase. Concentrations of eight to ten mononitroPAHs and three dinitropyrenes were determined in three diesel particulate-related SRMs: SRM 1650a Diesel Particulate Matter, SRM 1975 Diesel Particulate Extract, and SRM 2975 Diesel Particulate Matter (Industrial Forklift). The results from GC/MS NICI using two different columns (5% phenyl methylpolysiloxane and 50% phenyl methylpolysiloxane) were compared to each other and to results from two other laboratories for selected nitro-PAHs. 1-Nitropyrene was the most abundant nitro-PAHs in each of the diesel particulate SRMs (19.8±1.1 µg g–1 particle in SRM 1650a and 33.1±0.6 µg g–1 particle in SRM 2975). Three dinitropyrene isomers were measured in SRM 1975 at 0.5–1.4 µg g–1 extract and in SRM 2975 at 1–3 µg g–1 particle. Keywords Diesel particulate matter · Gas chromatography/mass spectrometry · Nitrated polycyclic aromatic hydrocarbons · Normal-phase liquid chromatography · Standard reference materials (SRMs)

D. Z. Bezabeh · H. A. Bamford · M. M. Schantz · S. A. Wise (✉) Analytical Chemistry Division, National Institute of Standards and Technology, MD 20899-8392, Gaithersburg, USA e-mail: [email protected] Present address: D. Z. Bezabeh Bureau of Alcohol, Tobacco, and Firearms, National Laboratory Center, 1401 Research Boulevard, Rockville, MD 20850, USA

Introduction Nitrated polycyclic aromatic hydrocarbons (nitro-PAHs) are of particular interest to the environmental analytical community because of the extraordinary mutagenic and carcinogenic activities displayed by some members of this class of compounds [1, 2, 3, 4, 5, 6]. Nitro-PAHs in the environment are formed primarily as a result of reactions between PAHs and nitrogen oxides and/or nitric acid, all of which are commonly found in combustion effluents [7, 8, 9]. Diesel engines, in particular, have been a major source of nitro-PAHs [9,10]. Typically nitro-PAHs are found at low µg g–1 concentration levels in complex environmental matrices; therefore, analytical methods used for the measurement of nitro-PAHs must exhibit both high selectivity and sensitivity. Several analytical techniques have been employed for the determination of nitro-PAHs. Measurements of nitroPAHs have been reported using gas chromatography (GC) with various detection methods (electron capture detection [11], nitrogen selective detection [12], reductive electrochemical detection [13], negative and positive ion chemical ionization mass spectrometry [6, 14], and high-resolution mass spectrometry [15]) and using high-performance liquid chromatography with chemiluminescence [9, 14], ultraviolet detection [14, 16], and fluorescence detection [13, 17]. A selective and sensitive mass spectrometric (MS) detection method, which minimizes or eliminates fragmentation, would be an ideal analytical choice for the determination of nitro-PAHs. The nitro-functional group is an electron-withdrawing group and has a large cross-section for electron capture; hence, nitro-PAHs are expected to form negative ions more readily than positive ions. Bezabeh et al. [18] have demonstrated the propensity of nitro-PAHs to readily form negative ions in the presence of more concentrated background. In the negative ion chemical ionization (NICI) mode, methane is used as a reagent gas and highly electronegative compounds such as nitro-PAHs are readily ionized by resonance capture of the thermal electrons. The captured low-energy electrons are then expected

382

to induce little fragmentation [19]. Accordingly, NICI has been widely utilized as the most selective and sensitive detection mode for the determination of nitro-PAHs by gas chromatography/mass spectrometry (GC/MS) [3, 6, 10, 20, 21, 22]. One of the first natural environmental matrix standard reference materials (SRMs) issued by the National Institute of Standards and Technology (NIST) for the determination of trace organic constituents was SRM 1650 Diesel Particulate Matter [23]. SRMs are certified reference materials issued by NIST. SRM 1650, which was issued in 1985, had certified concentrations for five PAHs and 1-nitropyrene and noncertified concentrations for six additional PAHs, three nitro-PAHs, and 9-fluorenone. At NIST certified concentration values for natural environmental matrix SRMs are typically assigned based on the agreement of results from two or more independent analytical methods [24]. If only one analytical method is used, the results are typically listed as reference values (previously denoted as noncertified values). For the determination of the certified value of 1-nitropyrene in SRM 1650 in 1985 [23], results were obtained from four techniques: GC/MS with electron impact (EI) ionization and NICI, and liquid chromatography (LC) with electrochemical detection and fluorescence detection (after on-line conversion of the nitroPAH to the corresponding amine) [13]. The noncertified values reported in 1985 for the three additional nitro-PAHs were based only on LC–fluorescence measurements [13]. Recently SRM 1650 was reanalyzed for the determination of PAHs to update the original certified and noncertified concentration values reported in 1985 and to assign values for additional PAHs. This material was reissued in 2000 as SRM 1650a Diesel Particulate Matter [25]. The analytical methods and results used for the value assignment of the PAHs in SRM 1650a are described in Poster et al. [26]. The objective of this study was to develop an improved analytical method for quantitative determination of nitroPAHs in complex environmental matrices using GC/MS NICI and to apply this method to the analysis of SRM 1650a and two new diesel particulate-related SRMs, SRM 1975 Diesel Particulate Extract and SRM 2975 Diesel Particulate Matter (Industrial Forklift). A normal-phase LC procedure was developed for isolation of the nitro-PAHs from the complex diesel particulate extract prior to GC/MS analysis on two columns with different selectivity for the separation of nitro-PAHs. Identification and quantification of selected nitro-substituted PAHs in SRM 1650a, SRM 2975, and SRM 1975 are presented in this paper. These results represent the most extensive characterization of nitro-PAHs reported in these three diesel-related SRMs.

Experimental Materials Three diesel particulate-related SRMs were analyzed: SRM 1650a Diesel Particulate Matter [25] (originally issued as SRM 1650 [23]), SRM 1975 Diesel Particulate Extract [27], and SRM 2975 Diesel Particulate Matter (Industrial Forklift) [28]. SRM 1587 Nitrated

PAHs in Methanol was used as a calibration solution and to evaluate the detection sensitivity of the method. All SRMs were obtained from the NIST Standard Reference Material Program (Gaithersburg, MD). The diesel particulate material used to prepare SRM 1650a (i.e., SRM 1650) was originally obtained through the Coordinating Research Council, Inc. (Atlanta, GA). The particulate material was collected from the heat exchanger of a dilution tube facility following 200 engine hours of particulate accumulation. Several direct injection four–cycle diesel engines, operating under a variety of conditions were used to generate this particulate material [23, 25]. SRM 2975 was obtained from the Donaldson Company, Inc. (Minneapolis, MN) where it was collected from a filtering system designed specifically for diesel-powered forklifts [29]. SRM 1975 was prepared by Soxhlet-extracting 5.7 kg of diesel particulate material (SRM 2975) with dichloromethane. Following extraction, the concentrated solution (8 L) was ampouled with each amber glass ampoule containing approximately 1.2 mL of solution. 2-Nitrofluorene (≥99%), 9-nitroanthracene (≥99%), 1-nitropyrene (≥99.9%), 3-nitroflouranthene (>99%), 7-nitrobenz[a]anthracene (≥99%), 6-nitrochyrsene (≥99%), and 6-nitrobenzo[a]pyrene (≥99%) (Midwest Research Institute, Kansas City, Missouri), 2-nitrofluoranthene (>99%) (Chemsyn, Lenexa, KS), and 9-nitrophenanthrene (AccuStandard, New Haven, CT) were used as calibration standards for the mono nitro-PAHs. 1,3-Dinitropyrene (≥99%), 1,6-dinitropyrene (≥99%), and 1,8-dinitropyrene (≥99%) (Chemsyn Science Laboratories) were used as calibration standards for the dinitropyene isomers. Deuterium-labeled nitro-PAHs: 2-nitrofluorene-d9 (≥98%), 9-nitroanthracene-d9 (≥99%), 1-nitropyrene-d9 (≥99%), 3-nitrofluoranthene-d9 (≥99%), 6-nitrochrysene-d11 (≥99%), as well as a deuterium-labeled dinitropyrene mixture: 1,3-dinitropyrene-d7, 1,6-dinitropyrene-d7, and 1,8-dinitropyrene-d7 (Cambridge Isotope Laboratories, Andover, MA) were used as internal standards. HPLCgrade solvents (dichloromethane, hexane, and methanol) were used to prepare the calibration solutions.

Sample preparation Based on a previous study by Schantz et al. [30] in which high molecular mass PAHs (>276 Da) were found to be extracted more effectively from diesel particulate matter using pressurized fluid extraction (PFE) than conventional Soxhlet extraction, a preliminary investigation of the efficacy of Soxhlet extraction compared to PFE for nitro-PAHs was conducted using SRM 1650a. The results indicated that both extraction techniques yielded comparable results. PFE was chosen for extraction of the particulate materials (four samples each of SRM 1650a and SRM 2975) because of its relatively rapid extraction time and lower solvent consumption compared to Soxhlet extraction. An Accelerated Solvent Extractor (Dionex Corporation, Sunnyvale, CA) was used for the PFE. The following parameters, optimized by Schantz et al. [30] for the determination of PAHs, were used for PFE extraction of SRM 1650a and SRM 2975: pressure at 13.8 MPa (2,000 psi) and temperature at 100 °C for 5 min, 3 static cycles for 5 min each. Dichloromethane-cleaned sodium sulfate was used to fill the cell volume. All sample analytes, calibration solutions, and blanks were extracted using dichloromethane. Three independent calibration solutions containing the mononitro-PAHs were prepared in dichloromethane at different concentrations to yield a three-point calibration curve. Into four separate 22-mL stainless steel cells, subsamples of approximately 70 mg (mass known) of SRM 1650a or SRM 2975 were spiked with approximately 0.5 mL (mass known) of internal standard solution in dichloromethane. The three calibration standards were also spiked with the internal standards in the same manner. For quantitative determination of the dinitropyrene isomers in SRM 2975, different calibration solutions containing three dinitropyrene isomers were prepared using the above steps. All sample, calibration, and blank solutions for SRM 1650a and SRM 2975 were extracted by PFE. The calibration solutions were processed through each step of the extraction and cleanup procedure with the samples in order to serve as both recovery and response factor solutions.

383 For analysis of SRM 1975 Diesel Particulate Extract, four ampoules were randomly selected and used for sample preparation. Into four separate vials, approximately 0.8 mL (mass known) of SRM 1975 solution was spiked with approximately 0.5 mL (mass known) of corresponding internal standard solution. Calibration solutions were spiked with the internal standard using similar steps applied for the calibration standard of SRM 1650a and SRM 2975. Each of the extracts and their corresponding calibration and blank solutions were concentrated to 0.5 mL under a steam of nitrogen with an automated evaporation system, and then subjected to an aminopropyl solid-phase extraction (SPE) cartridge using 40 mL of 20% dichloromethane in hexane to reduce potential interference of polar constituents. Prior to its use, each SPE column was rinsed with 20 mL of dichloromethane. The eluent from the SPE column was collected in a clean (baked at 500 °C for 18 h) glass evaporation vessel and concentrated to 0.5 mL as described above. The concentrated solution was then subjected to normal-phase LC fractionation using a 5 µm, 9.6 mm×25 cm semi-preparative Chromegabond amino/cyano column (ES Industries, West Berlin, NJ) to isolate the nitro-PAHs and dinitropyrenes. Initially three different stationary phases, that is, an amino/cyano phase, Ring Sep (a proprietary phase), and a nitro phase (ES Industries, West Berlin, NJ), were evaluated for isolation of PAHs, mononitro-PAHs, and dinitroPAHs from diesel particulate-related materials. The mobile phase was 20% dichloromethane in hexane at a flow rate of 5 mL min–1. Following the LC fractionation of each sample, the amino/cyano column was back flushed with 60 mL of dichloromethane. The column was then equilibrated with 150 mL of 20% dichloromethane in hexane before proceeding with the next sample fractionation. Mononitro-PAHs and dinitro-PAHs (dinitropyrene isomers) were collected in two separate fractions. Fraction collection for mononitro-PAHs was started immediately prior to the elution of 9-nitroanthracene (at approximately 6 min or 30 mL) until after the elution of 6-nitrobenzo[a]pyrene (approximately 12.5 min or 62.5 mL). The dinitropyrene fraction collection was started after the end of mononitro-PAH fraction collection (12.5 min or 62.5 mL) and ended at approximately 40 min or 200 mL. The fractions were concentrated to 0.5 mL prior to GC/MS analysis. A second set of three subsamples of each SRM was prepared for GC/MS analysis using the same PFE conditions and SPE cleanup, but no normal-phase LC cleanup was used for the SRM 1975 and SRM 2975 samples. The second set of samples was analyzed by GC/MS using a GC column with different selectivity (see details below). Three independent calibration solutions were used to prepare a seven-point calibration curve. Response factors were determined using the response of the nearest calibration point.

GC/MS analysis Two sets of GC/MS analyses of the three diesel particulate-related SRMs were obtained. For the first set of samples, duplicate GC/ MS NICI analyses of the mononitro-PAH and dinitropyrene fractions isolated from each sample were performed using a 0.25-mm i.d.×30 m fused silica capillary column containing a 5% phenylsubstituted methylpolysiloxane phase, 0.25-µm film thickness (DB–5MS, J&W Scientific Inc., Folsom, CA). The initial oven temperature was held at 60 °C for 2 min, then increased at 45 °C min–1 to 150 °C and held for 10 min, then increased at 5 °C min–1 to 300 °C and held for 15 min. Helium was used as the carrier gas at a constant flow of 1.2 mL min–1. Methane was used as the reagent gas for NICI. All injections were 1.0 µL on column. The transfer line was maintained at 300 °C. For the second set of GC/MS analyses a column with a 50% phenyl-substituted methylpolysiloxane phase, 0.25-mm i.d.×30 m, 0.25-µm film thickness (DB–17MS, J&W Scientific Inc., Folsom, CA) was used. For the analyses with the 50% phenyl-substituted methylpolysiloxane column, the GC/MS was operated under the same conditions as described above for the 5% phenyl-substituted methylpolysiloxane column. For all of the GC/MS analyses selected ion monitoring (SIM) was used to monitor the ions of interest. The monitored mass-to-

charge (m/z) ions for nitro-PAHs in SRM 1975 and 2975 were: 211 (2-nitrofluorene), 220 (2-nitrofluorene-d9), 223 (9-nitroanthracene and 9-nitrophenanthrene), 232 (9-nitroanthracene-d9), 247 (3-nitrofluoranthene, 2-nitrofluronathene, and 1-nitropyrene), 256 (3-nitrofluoranthene-d9 and 1-nitropyrene-d9), 273 (7-nitrobenz[a]anthracene and 6-nitrochrysene) 284 (6-nitrochrysene-d11), and 297 (6-nitrobenzo[a]pyrene). For the dinitropyrenes in both SRM 1975 and SRM 2975, the following ions were monitored in SIM mode: 292 (1,3-dinitropyrene, 1,6-dinitropyrene, and 1,8-dinitropyrene) and 300 (1,3-dinitropyrene-d8, 1,6-dinitropyrene-d8, and 1,8-dinitropyrene–d8). The corresponding deuterium-labeled nitro-PAH was used as an internal standard for the analytes of interest with the exception of 9-nitrophenathrene for which 9-nitroanthracene-d9 was used and 7-nitrobenz[a]anthracene and 6-nitrobenzo[a]pyrene for which 6-nitrochrysene-d11 was used. The reported values for 1-nitrobenzo[e]pyrene and 3-nitrobenzo[e]pyrene in SRM 1975 and SRM 2975 were obtained based on the response factor for 6-nitrobenzo[a]pyrene.

Results and discussion LC fractionation A normal-phase LC procedure was developed and utilized to isolate mono- and dinitro-PAHs from the extract of the diesel particulate-related SRMs. Three LC columns with different stationary phases (Nitro, RingSep, and Amino/ Cyano) were initially investigated for isolation of nitroPAHs. Of the three columns, the amino/cyano column displayed a distinct group separation for the following three classes of compounds: PAHs, mononitro-PAHs, and dinitro-PAHs (dinitropyrene isomers). The amino/cyano stationary phase has been previously reported for the isolation of nitro-PAHs [31]. The normal-phase LC separation of the PAHs, mononitro-PAHs, and dinitropyrene isomer fractions from SRM 1975 is shown in Fig. 1. Unlike the amino/cyano column, co-elution between PAHs and nitroPAHs was observed on the RingSep and nitro columns. The LC fractionation procedure eliminates or minimizes potential interferences and enhances the sensitivity. Moreover, the sample cleanliness obtained from the isolation of the nitro-PAHs has an added benefit in the GC/MS analysis by maintaining a clean retention gap and increasing column longevity. However, the LC cleanup was not required for the analysis of SRM 1975 and SRM 2975 as indicated by the results of the second set of analyses on the 50% phenyl phase, in part because of the high concentration of nitro-PAHs. However, this normal-phase LC step is crucial in the determination of low concentrations (ng g–1) of nitro-PAHs in air particulate samples. A detailed discussion of the advantages of the LC fractionation procedure for the determination of nitro-PAHs in two air particulate SRMs is presented elsewhere [32].

GC/MS NICI method development GC/MS with negative ion chemical ionization (NICI) was utilized for quantitative determination of selected nitroPAHs in the diesel particulate-related SRMs. NICI has been demonstrated to selectively ionize any electronega-

384 Fig. 1 Normal-phase LC analysis of SRM 1975 (Diesel Particulate Extract) on an amino/cyano column. Elution times of the PAH, mononitroPAH, and dinitro-PAH fractions are indicated

tive groups including nitro-PAHs in a complex environmental extract. Hence, in addition to minimizing interference by developing SPE and LC fractionation procedures prior to the GC/MS analysis, employing NICI detection further minimizes any interferences that may elute with the nitro-PAH fraction. The EI and NICI modes were examined for sensitive ionization of nitro-PAHs in SRM 1587 Nitrated PAHs in Methanol in the preliminary stage of this study. Comparison of EI and NICI peak areas for nitro-PAHs revealed that the NICI mode yielded two orders of magnitude greater sensitivity than the EI mode. Characterization of the nitro-PAHs in diesel particulaterelated SRMs was commenced after the analytical method was developed and validated using the certified concentrations of seven nitro-PAHs in SRM 1587. The primary difference in the two sets of GC/MS analyses reported in this paper is the use of two different GC stationary phases for the separation of the nitro-PAHs. The first set of results in this study was obtained using the 5% phenyl methylpolysiloxane column that has typically been used for PAH and nitro-PAH measurements. Recently, the 50% phenyl methylpolysiloxane has been shown to provide better selectivity for the separation of several critical pairs of PAH isomers [26, 33], and therefore this stationary phase was evaluated for the separation of nitro-PAH isomers, particularly the 2-nitrofluoranthene and the 3-nitrofluoranthene, which coelute on the 5% phenyl methylpolysiloxane phase. A detailed comparison and discussion of the 50% phenyl and 5% phenyl methylpolysiloxane phases for the separation of nitro-PAH isomers is described elsewhere [32]. For SRM 1975 and SRM 1650a, results for selected nitro-PAHs were obtained from two other laboratories for comparison (Battelle and Environment Canada). Both laboratories used a GC/MS NICI method with a 5% phenyl

methylsiloxane column. Environment Canada used GC with high-resolution mass spectrometry (HRMS) with NICI [15]. Analysis of SRM 1650a Diesel Particulate Matter The GC/MS NICI method was applied to the analysis of SRM 1650a Diesel Particulate Matter for the determination of the mononitro-PAHs. The quantitative results for selected mononitro-PAHs in SRM 1650a obtained using both the 5% and 50% phenyl methylpolysiloxane columns are listed in Table 1, and the results are compared with certified and noncertified values in the original NIST Certificate of Analysis for SRM 1650 [23] and with results reported by Chiu and Miles [15]. 1-Nitropyrene is the only nitro-PAH for which a certified value was provided in the original SRM 1650 Certificate of Analysis. The measured values for 1-nitropyrene in the present study (19.8±1.1 µg g–1 and 18.3±0.3 µg g–1) are in good agreement with the original certified value (19±2 µg g–1) and the value reported by Chiu and Miles (18.9±1.4 µg g–1) [15]. Scheepers [34] recently determined 1-nitropyrene in SRM 1650a and reported a concentration of 18.8±1.1 µg g–1 (n=3). These recent results indicate that the concentration of 1-nitropyrene in SRM 1650 has been stable for the past 17 years since the original certification measurements. The concentration for 6-nitrobenzo[a]pyrene for the two NIST methods is also in good agreement with the noncertified value reported in the original Certificate of Analysis and the results of Chiu and Miles [15]. However, the current results for the other two nitro-PAHs reported in the original SRM 1650 Certificate of Analysis as noncertified values, 2-nitrofluorene and 7-nitrobenz[a]anthracene, are not in agreement with the original measurements in 1985 or with the measurements reported by Chiu and Miles [15]. For 2-nitrofluorene both

385 Table 1 Concentrations of selected nitro-PAHs in SRM 1650a Diesel Particulate Matter

2–Nitrofluorene 9–Nitroanthracene 9–Nitrophenanthrene 3–Nitrofluoranthene 2–Nitrofluoranthene 1–Nitropyrene 7–Nitrobenz[a]anthracene 6–Nitrobenzo[a]pyrene

GC/MS NICIa (5% phenyl) (µg g–1)

GC/MS NICIb (50% phenyl) (µg g–1)

NIST Certificate of Analysisc (µg g–1)

Chiu and Milesd (5% phenyl) (µg g–1)

N.D.e N.D. N.D. N.D. N.D. 19.8±1.1 1.22±0.19 1.69±0.24

0.0462±0.0026 6.08±0.19 0.510±0.010 0.065±0.007 0.201±0.012 18.3±0.3 0.995±0.068 1.44±0.05

(0.27) N.D. N.D. N.D. N.D. 19±2 (2.8) (1.6)

<0.01 6.02 (0.34) 0.369 (0.022) 0.332 (0.025)f N.D. 18.9 (1.4) 0.445 (0.083) 1.38 (0.09)

aMean values of duplicate measurements by GC/MS NICI on four samples. The associated uncertainties represent 95% confidence limits. Nitro–PAH fraction isolated by normal-phase LC prior to GC/MS NICI bMean value of single measurements by GC/MS NICI on three samples. The associated uncertainties represent 95% confidence limits. No normal-phase LC isolation of nitro-PAHs prior to GC/MS NICI cConcentrations listed on the Certificate of Analysis for SRM 1650 issued in 1985; reissued in 2000 as SRM 1650a with update values

for some analytes. The numbers in parenthesis are noncertified or information values dNitro-PAH values reported by Chiu and Miles using GC/HRMS NICI [17]; results are reported as mean±standard deviation (n=3) e N.D. not determined fThe concentration value for 3-nitrofluoranthene includes 2-nitrofluoranthene

Table 2 Concentrations of selected mononitro- and dinitro-PAHs in SRM 1975 and SRM 2975

9–Nitroanthracene 9–Nitrophenanthrene 3–Nitrofluoranthene 2–Nitrofluoranthene 1–Nitropyrene 7–Nitrobenz[a]anthracene 6–Nitrochrysene 6–Nitrobenzo[a]pyrene 1–Nitrobenzo[e]pyreneg 3–Nitrobenzo[e]pyreneg 1,3–Dinitropyrene 1,6–Dinitropyrene 1,8–Dinitropyrene

SRM 1975

SRM 2975

GC/MS NICIa,b GC/MS NICIb,c Battelleb,d Chiu and Milesb,e (5% phenyl) (50% phenyl) (5% phenyl) (5% phenyl) (µg g–1) (µg g–1) (µg g–1) (µg g–1)

GC/MS NICIa,b GC/MS NICIb,c (5% phenyl) (50% phenyl) (µg g–1 particle) (µg g–1 particle)

1.36±0.03 0.266±0.009 1.47±0.01h 16.4±0.1 1.62±0.03 0.782±0.007 0.641±0.006 0.83±0.05 2.2±0.2 0.603±0.011 1.39±0.04 1.55±0.02

1.28±0.02 0.205±0.005 1.62±0.02 0.094±0.004 16.1±0.6 1.96±0.07 0.900±0.015 0.514±0.024 0.670±0.017 1.83±0.07 0.538±0.039 0.934±0.014 1.38±0.04

1.17±0.05 N.D.f 1.23±0.04h

1.19±0.02 0.175±0.006 1.34±0.034h

3.37±0.19 0.767±0.063 3.32±0.10h

17.2±0.4 3.34±0.08 0.71±0.02 0.9±0.2 0.8±0.2 2.8±0.2 0.8±0.2 1.1±0.2 1.0±0.3

16.6±0.34 1.99±0.12 0.873±0.13 0.456±0.012 N.D. N.D. 0.581±0.093 0.640±0.112 0.495±0.143

33.1±0.6 3.03±0.12 1.38±0.07 1.19±0.17 1.72±0.16 7.1±1.4 0.961±0.030 1.95±0.06 2.62±0.16

2.93±0.06 0.454±0.020 4.30±0.033 0.250±0.011 39.6±1.7 5.31±0.38 2.37±0.07 1.65±0.04 1.79±0.05 6.86±0.21 1.15±0.06 2.54±0.22 3.58±0.17

aValues

based on GC/MS NICI measurements at NIST using a 5% phenyl methylpolysiloxane column. The reported value is the mean of duplicate measurements on four samples. Nitro-PAH fraction isolated by normal-phase LC prior to GC/MS NICI bThe associated uncertainties represent a 95% confidence limits cValues based on GC/MS NICI measurements at NIST using a 50% phenyl methylpolysiloxane column. The reported value is the mean of duplicate measurements on three samples. No normal-phase LC isolation of nitro-PAHs prior to GC/MS NICI

dValues

based on GC/MS measurements at Battelle. The reported value is the mean of five measurements eReported values by Chiu and Miles. The reported value is the mean of three measurements using GC/HRMS NICI f N.D. not determined gMean concentration value calculated using a response factor for 6-nitrobenzo[a]pyrene hThe concentration value for 3-nitrofluoranthene includes 2-nitrofluoranthene

the current NIST measurements and the Chiu and Miles [15] result are two orders of magnitude lower than the original noncertified value, indicating that the current GC/MS NICI measurements are probably more selective than the original LC–fluorescence result. Although not reported in the original Certificate of Analysis, 2-nitrofluoranthene and 3-nitrofluoranthene were determined using the 50% phenyl methylpolysiloxane

column. When the results for 2-nitro- and 3-nitrofluoranthene are combined, they are in excellent agreement with the result reported by Chiu and Miles [15] using a 5% phenyl methylpolysiloxane phase, where 2-nitrofluoranthene and 3-nitrofluoranthene coelute. Additional nitroPAHs have been measured in SRM 1650a using the 50% phenyl methylpolysiloxane phase and these results are discussed in detail elsewhere [32].

386 Fig. 2 GC/MS analysis of an extract from SRM 1650a (Diesel Particulate Matter) on a 5% phenyl methylpolysiloxane column. Selected ion monitoring of m/z 223, 247, 273, and 297

Analysis of SRM 1975 and SRM 2975 The results for mononitro-PAHs and dinitro-PAHs in SRM 1975 Diesel Particulate Extract and SRM 2975 Diesel Particulate Matter (Industrial Forklift) are summarized in Table 2. Results for SRM 1975 from two other laboratories, Battelle [35] and Environment Canada [15], which analyzed SRM 1975 to assist in the value assignment process, are also included in Table 2 for comparison. For the mononitro-PAHs in SRM 1975 in Table 2, the two NIST results and the results from the other laboratories generally agree within a 30% range with the exception of high results reported by Battelle for 7-nitrobenz[a]anthracene and 6-nitrobenzo[a]pyrene. For 1-nitropyrene, the most abundant nitro-PAH measured in SRM 1975, the NIST GC/MS NICI results obtained using the two columns are in excellent agreement with each other and with the results from the other laboratories (range of results of 6%). The results from NIST and the two laboratories are in good agreement for

the 1,3-dinitropyrene; however, for the other two dinitropyrene isomers, the range of results is within 35%, with the exception of the results reported by Environment Canada, which are 50% to 70% lower than the other three results. For SRM 2975 the concentration of 1-nitropyrene determined in this study using both columns (33.1±0.6 µg g–1 and 39.6±1.7 µg g–1) bracket the concentration reported recently by Scheepers [34] (35.7±2.7 µg g–1, n=3). For the other nitro-PAHs the majority of the results from the two NIST methods agree within 25%. SRM 1650a and SRM 2975 are diesel particulate materials that were collected from different diesel engines; hence, it is worth noting the similarities and differences in relation to the nitro-PAH patterns in both SRMs. Selected ion chromatograms from the analysis of SRMs 1650a and 1975 are depicted in Figs. 2 and 3, respectively. Because SRM 1975 was prepared from particulate matter from SRM 2975, the chromatograms from the GC/MS analysis for the nitro-PAHs in these two SRMs are similar. There-

387 Fig. 3 GC/MS analysis of SRM 1975 (Diesel Particulate Extract) on a 5% phenyl methylpolysiloxane column. Selected ion monitoring of m/z 223, 247, 273, 297, and 292

fore, only the chromatogram from the analysis of SRM 1975 is shown (Fig. 3) and it is representative for both SRM 1975 and SRM 2975. One similarity between SRM 1650a and SRM 2975 is the presence of 1-nitropyrene as the most abundant constituent from the nitro-PAH class of compounds. In general SRM 1975 has more isomers present within each molecular mass group, for example, 3-nitro-

fluoranthene (m/z 247), 6-nitrochrysene (m/z 273), 1-nitrobenz[e]pyrene (m/z 297), and 3-nitrobenz[e]pyrene (m/z 297) are present in greater concentrations relative to other isomers in SRM 1975 compared to SRM 1650a. As shown in Figs. 2 and 3, a number of unknown nitro-PAH isomers were detected (particularly isomers at molecular mass 223, 273, and 297). A more detailed identification and quan-

388

tification of additional nitro-PAH compounds in these diesel particulate-related SRMs and air particulate SRMs are presented elsewhere [32].

Conclusions A method was developed for isolation and quantitative determination of nitro-PAHs in diesel particulate material. Nitro-PAHs were successfully isolated from the solvent extract of diesel particulate matter by utilizing SPE cleanup and normal-phase LC fractionation. The isolated nitro-PAH fractions were analyzed by GC/MS with negative ion chemical ionization. Quantitative values were achieved with good accuracy and precision for selected mono- and dinitro-PAHs in three diesel particulate-related SRMs. These results will be used to assign certified and/or reference values for the concentrations of selected nitro-PAHs in these three diesel particulate-related SRMs. Acknowledgements The authors acknowledge C. Chiu of Environment Canada and M. Nishioka of Battelle Laboratories for their analyses of SRM 1975.

References 1. Yu WC, Fine DH, Chiu KS, Biemann K (1984) Anal Chem 56: 1158–1162 2. Handa T, Yamauchi T, Ohnishi M, Hisamatsu Y, Ishit T (1983) Environ Internat 9:335–341 3. Bonfanti L, Careri M, Mangia A, Manini P, Maspero M (1996) J Chromatogr A 728:359–369 4. Durant JL, Busby WF, Lafleur AL, Penman BW, Crespi CL (1996) Mutation Res 371:123–157 5. Talaska G, Underwood P, Maier A, Lewtas J, Rothman N, Jaeger M (1996) Environ Health Perspect 104:901–906 6. Schilhabel J, Levsen K (1989) Fresenius Z Anal Chem 333: 800–805 7. Zielinska B, Arey J, Atkinson R, Winer AM (1989) Atmos Environ 23:223–229 8. Atkinson R, Winer AM, Pitts JN (1986) Atmos Environ 20: 331–339 9. Li H, Westerholm R (1994) J Chromatogr A 664:177–182 10. Newton DL, Erickson MD, Tomer KB, Pellizzari ED, Gentry P (1982) Environ Sci Technol 16:206–213 11. Campbell RM, Lee ML (1984) Anal Chem 56:1026–1030 12. Schuetzle D, Jensen TE, Ball JC (1985) Environ Internat 11: 169–181

13. MacCrehan WA, May WE, Yang SD, Benner BA (1988) Anal Chem 60:194–199 14. Robbat A, Corso NP, Doherty PJ, Wolf MH (1986) Anal Chem 58:2078–2084 15. Chiu C, Miles W (1996) Polycyclic Aromat Compd 9:307–314 16. Robbat A, Liu TY (1990) J Chromatogr 513:117–135 17. Liu TY, Robbat A (1991) J Chromatogr 539:1–14 18. Bezabeh DZ, Allen TM, McCauley EM, Jones AD, Kelly PB (1997) J Am Soc Mass Spectrom 8:630–636 19. Tomkins, BA (1985) In: White CM (ed) Nitrated polycyclic aromatic hydrocarbons. Springer, Heidelberg Basel New York, pp 111–114 20. Ramdahl T, Becher G, Bjørseth A (1982) Environ Sci Technol 16:861–865 21. Xu J, Lee FSC (2001) Chemosphere 42:245–250 22. Dimashki M, Harrad S, Harrison RM (2000) Polycyclic Aromat Compd 20:205–223 23. Certificate of analysis, standard reference material (SRM) 1650, diesel particulate matter (1985) National Institute of Standards and Technology (NIST), Gaithersburg, MD 24. May WE, Parris R, Beck C, Fassett J, Greenberg R, Kramer G, Wise S, Gills T, Colbert J, Gettings R, MacDonald B (1999) Definitions of terms and modes used at NIST for value assignment of reference materials for chemical measurements. NIST Special Publication SP260–136, Gaithersburg, MD 25. Certificate of analysis, standard reference material (SRM) 1650a, diesel particulate matter (2000) National Institute of Standards and Technology (NIST), Gaithersburg, MD 26. Poster DL, Lopez de Alda MJ, Schantz MM, Sander LC, Vangel MG, Wise SA (2002) Polycyclic Aromat Compd (in press) 27. Certificate of analysis, standard reference material (SRM) 1975, diesel particulate extract (2000) National Institute of Standards and Technology (NIST), Gaithersburg, MD 28. Certificate of analysis, standard reference material (SRM) 2975, diesel particulate matter (Industrial Forklift) (2000) National Institute of Standards and Technology (NIST), Gaithersburg, MD 29. Wright ME, Klein AD Jr, Stesniak EJ (1991) SAE Technical Paper Series 911852, Warrendale, PA 30. Schantz MM, Nichols JJ, Wise SA (1997) Anal Chem 69: 4210–4219 31. D’Agostino PA, Narine DR, McCarry BE, Quilliam MA (1983) In: Cooke M, Dennis AJ (eds) Polynuclear aromatic hydrocarbons: formation, metabolism and measurement. Battelle Press, Columbus, pp 365–377 32. Bamford HA, Bezabeh DZ, Schantz MM, Wise SA (2003) Chemosphere 50:575–587 33. Poster DL, Lopez de Alda MJ, Wise SA (2000) Polycyclic Aromat Compd 20:79–95 34. Scheepers P (1997) Toxicology Department, Katholieke Universiteit Nijmegen, Nijmegen, The Netherlands, personal communication 35. Nishioka M (1994) Battelle, Columbus, OH, unpublished results