Ann. occup. Hyg., Vol. 35, No. 6, pp. 603-611, 1991. Primed in Great Britain.

0003-4878/91 S3.00 + 0.00 Pcrgamon Press pic V- 1991 British Occupational Hygiene Society.

VALIDATION OF A HIGH-PERFORMANCE LIQUID CHROMATOGRAPHY/FLUORESCENCE DETECTION METHOD FOR THE SIMULTANEOUS QUANTIFICATION OF FIFTEEN POLYCYCLIC AROMATIC HYDROCARBONS

(Received 4 October 1990 and in final form 16 July 1991) Abstract—A high-performance liquid chromatography/fluorescence method using multiple wavelength shift for simultaneous quantification of different PAH compounds was developed. The new method was superior to the methods of DONG and GREENBERG [/. Liquid Chromatogr. 11,1887-1905 (1988)] and WISE et ai. ^Polycyclic arotnat. Hydrocarb. (in press)] with respect to sensitivity of detection of the majority of 15 PAH compounds, and in particular of naphthalene, acenaphthene, fluorene and benzo(b)fluoranthene. The method of validation analysis employed showed that the new method is in statistical balance meaning that no systematic errors, and only small unsystematic errors, could be demonstrated. Furthermore, the method had a good reproducibility and a high sensitivity.

INTRODUCTION POLYCYCLIC aromatic hydrocarbons (PAH), several of which have both mutagenic and carcinogenic properties, are often present as components of organic aerosols. Studies in laboratory animal models have demonstrated that some of the volatile polycyclic aromatic hydrocarbons, for example pyrene, potentiate the effect of the carcinogenic PAH compound benzo(a)pyrene (IARC, 1983). Furthermore, exposure to airborne PAH is considered a major environmental and occupational health problem and an increased incidence of lung cancer has been demonstrated in workers exposed to high levels of airborne PAH (BJORSETH and BECHER, 1986). During the last decade several new methods for identification and quantification of airborne PAH after separation by high-performance liquid chromatography (HPLC) have been published. The National Institute of Safety and Health (NIOSH, 1985) has recommended a standard method to quantify 16 different PAH compounds, DONG and GREENBERG (1988) described a method using a two wavelength shift to quantify different PAH compounds after separation on HPLC, and recently WISE et al. (1991) introduced multiple wavelength shift to quantify several of the 16 PAH compounds mentioned above. However, the concentrations of different airborne PAH compounds often differ widely, and these methods cannot be used for single-stage quantification of different PAH compounds in air samples from, for example, smokehouses. Whenever analytical results have large economic and political consequences (for example in the evaluation of health risk of new pharmaceutical compounds or of occupational exposure) the analytical method employed needs to be validated statistically with 603

Downloaded from http://annhyg.oxfordjournals.org/ at McMaster University Library on January 5, 2015

ASE MARIE HANSEN,* INGE LISE BRINK OLSEN,! E. HOLST* and OTTO MELCHIOR POULSEN* *Danish National Institute of Occupational Health. Lerso Parkalle 105, DK-2100 (. openhagen, Denmark; and f Briiel & Kjaer A/S, Njerum Hovedgade 18, 2850 Naerum, Denmark

604

A. M. HANSEN et al.

respect to accuracy, precision, limit of detection and limit of quantification (LONG and WINEFORDNER, 1983; AARONS et al., 1987).

The present study describes the validation of a new method for the simultaneous detection of 15 of the 16 reference PAH compounds. This is based on fluorescence detection in combination with multiple changes both in excitation and emission wavelengths. One of the 16 reference PAH compounds (acenaphthylene) cannot be excited, hence this compound was excluded from the study. The results of the new method are compared with those of DONG and GREENBERG (1988) and of WISE et al. (1991).

Chemicals The acetonitrile used was LiChrosolv (Merck, Darmstadt, Germany). Water was obtained from a Milli-Q water purification system (Millipore Waters, Taastrup, Denmark). The PAH standard solution containing naphthalene, acenaphthylene, acenaphthene, fluorene, phenanthrene, anthracene, fluoranthene, pyrene, benz(a)anthracene, chrysene, benzo(b)fluoranthene, benzo(k)fluoranthene, benzo(a)pyrene, dibenz(a,h)anthrancene, benzo(ghi)perylene, and indeno(l,2,3-cd)pyrene was from Supelco (Gland, Switzerland). The PAH reference material, which contained the same 16 PAH compounds but in different proportions, was from the National Bureau of Standards (NBS, 1647a, Gaithersburg, Maryland, U.S.A.). Sample preparation for method evaluation In order to include the recovery of PAH compounds after binding to filters and tubes routinely used for analysis of airborne PAH compounds in the evaluation of the method, a series of dilutions of the Supelco standard solution in 100% acetonitrile (undiluted, 2-, 3-, 5- and 25-fold dilutions) were prepared, and 25 fi\ of each of the five diluted samples was spiked to 37-mm glass fibre filters (Millipore AP 4003-705, Taastrup, Denmark) and styrene-divinylbenzene adsorption tubes (ORBO 43 from Supelco, Gland, Switzerland). PAH components in the samples were extracted ultrasonically (1 h) with 10 ml of acetonitrile in 25 ml Erlenmeyer flasks sealed with laboratory film and placed in an ice bath. Quantification of PAH compounds by HPLC The PAH compounds were separated by reversed-phase HPLC using the following equipment: Pumps model 6000A, a WISP 710B autosampler, a model 720 system controller and a model 730 data module (all from Waters Associates Inc., Milford, U.S.A.). The column (250 x 4.6 mm inner diameter) was a Supelcosil LC-PAH column (5 /im packing) protected with a Waters guard column with Bondapak C 18 /Corasil (37-50/im). The mobile phase was acetonitrile-water (2:3) for the first 5 min, then increasing linearly to 100% acetonitrile for 20 min, and finally 100% acetonitrile for 15 min. The flow rate was 1.0 ml min" 1 , and the injection volume was 25 //I. All HPLC runs were performed at room temperature (2O-22°C). Two different detectors were used: a Waters model 440 absorbance detector (254 nm) was used for quantification of acenaphthylene; and a Perkin-Elmer fluorescence detector model LS-4 (Perkin-Elmer

Downloaded from http://annhyg.oxfordjournals.org/ at McMaster University Library on January 5, 2015

MATERIALS AND METHODS

HPLC method for PAH measurements

605

Corp., Norwalk, U.S.A.) was used to quantify the other 15 PAH compounds using multiple wavelength shifts (Table 1). The elution order of the compounds was confirmed (identified) using a HewlettPackard 1040A diode-array detector with the aid of a Hewlett-Packard HP 85 and a Hewlett-Packard 9121 disc drive (Hewlett-Packard, Waldbronn, Germany). Preparation of calibration plots and method evaluation points

0.5

1.0

1.5

2.0

2.5

3.0

3.5 Jig

FIG. 1. M EP plot of anthracene from a spiked tube. The measured concentrations (K) are plotted against the true concentrations (p).

concentrations of the 15 PAH compounds were used in the MEP samples. At each calibration plot the PAH concentration of the MEP sample was randomly selected from one of thefiveconcentrations employed. This form of method evaluation study eliminates any correlation between MEP samples and the calibration samples, and furthermore makes it possible to recognize systematic and random errors due to changes in for example the equipment employed during the 4-month period of study. Statistics Any chemical method can be characterized by its method evaluation function (M EF). Denoting the result of the chemical analysis Yand the true concentration of the analyte \i the equation of the MEF is:

Downloaded from http://annhyg.oxfordjournals.org/ at McMaster University Library on January 5, 2015

To evaluate the method nine calibration plots for both filter and tube were made over 4 months for each of the 15 PAH compounds using the abovementioned five different dilution levels of the Supelco standard solution. Simultaneously, a method evaluation point (MEP) sample with known concentration of PAH compounds was prepared by diluting in 100% acetonitrile a different batch of the Supelco standard solution or the NBS reference PAH material (dilution between undiluted and 25-fold dilution). The concentration of PAH compounds in the MEP sample was calculated from the calibration plot and the calculated value (Y) was plotted vs the true concentration (JI) of PAH compound in the MEP sample. Consequently, each MEP in the MEP plot (Fig. 1) corresponds to a particular calibration curve. Five different

606

A. M. HANSEN el al.

where aY is the standard deviation of Y given /J. In practice the relative mean square error (RMSE = MSE(/i)//i2) is a more useful parameter. In checking the largest acceptable value of RMSE 1/2 (RMSE1/2(/i)m,%) is selected, and for each value of // RMSE" 2 should be less than R M S E 1 ' 2 ^ ) ^ . At our laboratory RMSE 1/2 (/i) ma!I = 25% is selected for practical reasons. The limit of quantification is then defined as the lowest concentration of/i which will produce a RMSE 1 / 2

fluorescence detection method for the simultaneous quantification of fifteen polycyclic aromatic hydrocarbons.

A high-performance liquid chromatography/fluorescence method using multiple wavelength shift for simultaneous quantification of different PAH compound...
1MB Sizes 0 Downloads 0 Views