This article was downloaded by: [Purdue University] On: 13 April 2015, At: 16:28 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Food Additives & Contaminants Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tfac19
Polycyclic aromatic hydrocarbons in olive oils on the Italian market a
a
a
Edoardo Menichini , Adriana Bocca , Franco Merli , a
Daniela Ianni & Fabio Monfredini
a
a
Istituto Superiore di Sanità , viale Regina Elena 299, Roma, 00161, Italy Published online: 10 Jan 2009.
To cite this article: Edoardo Menichini , Adriana Bocca , Franco Merli , Daniela Ianni & Fabio Monfredini (1991) Polycyclic aromatic hydrocarbons in olive oils on the Italian market , Food Additives & Contaminants, 8:3, 363-369, DOI: 10.1080/02652039109373985 To link to this article: http://dx.doi.org/10.1080/02652039109373985
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sublicensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
FOOD ADDITIVES AND CONTAMINANTS, 1991, VOL. 8, NO. 3 , 3 6 3 - 3 6 9
Polycyclic aromatic hydrocarbons in olive oils on the Italian market * EDOARDO MENICHINI, ADRIANA BOCCA, FRANCO MERLI, DANIELA IANNI † and FABIO MONFREDINI Istituto Superiore di Sanità, viale Regina Elena 299, 00161, Roma, Italy
Downloaded by [Purdue University] at 16:28 13 April 2015
(Received 15 November 1990; revised 25 February 1991; accepted 8 March 1991) The six olive oils and seven virgin olive oils which are most consumed in Italy were analysed for 28 polycyclic aromatic hydrocarbons (PAHs). The aim was to evaluate whether a carcinogenic hazard for the general population can derive from the dietary intake of this food, which is consumed particularly highly in the Mediterranean area. The analytical method involved extraction by liquid-liquid partition, nitration on silica gel, clean-up by thin-layer chromatography on silica gel, and analysis by high-resolution gas chromatography with a flame ionization detector. The 3- and 4-ring PAHs which are most abundant in the environment were found in all samples, at individual levels up to ca. 40 µg/kg (for phenanthrene); no important difference was observed between olive oils and virgin olive oils. PAHs which are most suspected of being carcinogenic for humans were not detected (limit of detection, ca. 3 µg/kg). The average yearly intake of the detected PAHs through this food was estimated at ca. 0.56 mg per capita. Keywords: polycyclic aromatic hydrocarbons, olive oil, virgin olive oil
Introduction
The occurrence of polycyclic aromatic hydrocarbons (PAHs) in foods, as a result of sorption from a contaminated environment or from food preparation methods, is a known and widely reviewed matter (Howard and Fazio 1980, Santodonato etal., 1981, Fazio and Howard 1983, IARC 1983). It is also of primary health interest, owing to the carcinogenic activity of many PAHs (IARC 1983, 1987). More specifically, other reviews have dealt with the PAH presence in olive oils (Tiscornia et al. 1982) and in various vegetable oils (Mariani and Fedeli 1984). Marketed olive oils, apparently not influenced by direct sources of PAH emissions, have been analysed previously, although in most investigations as a single sample. Individual PAHs have generally been detected at /tg/kg levels in refined oil (Howard etal. 1966, Biernoth and Rost 1968, Fabian 1968, Joe etal. 1979, Hopia etal. 1986, Welling and Kaandorp 1986, Stijve and Hischenhuber 1987, Speer etal. 1990), the most commonly detected PAHs being AN, BaA, BghiP, CHR, FA, IP, PHE, PY (for abbreviations, see table 1). Higher
* Note. The results of this work were presented in part at the 1 ° Congresso Nazionale di Chimica degli Alimenti (1st National Congress on Food Chemistry) (Messina and Giardini-Naxos, Italy, 9-13 October 1990). † Guest of the Istituto Superiore di Sanita. 0265-203X/9I $3.00 © 1991 Taylor & Francis Ltd.
364
E. Menichini et al. Table 1. PAHs in olive oils and virgin olive oils.3
Abbrev.
Olive oils Fluorene Phenanthrene Anthracene Fluoranthene Pyrene BaFL 11 H-Benzo [a] fluorene CHR + TRI Chrysene + triphenylene ?
No. 1
No. 2
No. 3
No. 4
No. 5
No. 6
Median
nd 12
nd 4 nd 3 2 nd nd
nd 8 nd 5 5 nd nd
nd 11 nd 7 7 nd 7
nd 15
6 10 2? 6
3 41 4 15 14 2 7
10 12 nd 8
nd 11 nd 6 9 nd 6
Virgin olive oils Fluorene Phenanthrene Anthracene Fluoranthene Pyrene BaFL 11 H-Benzo [a] fluorene CHR + TRI Chrysene + triphenylene ?
No. 7
No. 8
No. 9
No. 10
No. 11
No. 12
No. 13
7 42 5 10 10 nd 3
4 38 7 10 7 nd nd
3 27 6 10 9 nd 3
5 36 8 11 10 nd 5
6 38 8 12 11 nd 5
3 32 7 10
3 25 6 7 7 nd nd
FL PHE AN FA PY
FL PHE AN FA PY
Downloaded by [Purdue University] at 16:28 13 April 2015
Concentration (jig/kg)1"
Compound
nd
nd
.
10 nd 3
Median 4
36 7 10 10 nd 3
nd: not detected (for limit of detection, see text). ?: doubtful identification. a The following PAHs were not detected in any sample: 2-methylanthracene; 3,6-dimethylphenanthrene; 11Hbenzo [b] fluorene; benzo[b]naphtho[2,l-d]thiophene; benzo [ghi] fluoranthene; cyclopenta[cd] pyrene (CPP); benz [a) anthracene (BaA); benzo [b] fluoranthene (BbFA); benzo [j] fluoranthene (BjFA); benzo [k] fluoranthene (BkFA); benzo [e] pyrene; benzo [a] pyrene (BaP); perylene; indeno[l,2,3-cd)pyrene (IP); dibenz [a,c] anthracene; dibenz[a,h] anthracene (DBahA); picene (PIC); benzo [ghi] perylene (BghiP); dibenzo [def ,mno]chrysene (known as anthanthrene, ATR); coronene. "Mean values of duplicate measurements.
concentrations, up to 40/ig/kg of BaP and 196/tg/kg of PHE, were found in one sample of refined oil (Morgante 1973) and two samples of virgin olive oils (Ciusa et al. 1970, Morgante 1973). (For a comparative exposition of such determinations, see the review of Menichini and Bocca 1991.) Particularly high concentrations were found recently (Corradetti et al. 1988) in Italian virgin oils deriving from plants exposed to industrial emissions of pitch condensate (BaP and PHE, up to ca. 60 and 3800 /tg/kg, respectively) and, to a lesser extent (PHE, ca. 800 iig/kg), in virgin oils deriving from plants exposed to vehicular exhausts. Following these findings, and also considering the high consumption of this food in Italy as well as in other Mediterranean countries, it appeared important to check whether PAHs present in the environment actually contaminate olive oils found on the market to any serious extent, with reference to those commonly measured in most foods (some /tg/kg per individual PAH; Santodonato et al. 1981); hence, whether these oils pose any special carcinogenic risk for the general population. In this paper, we present the results of an investigation concerning the occurrence of 28 PAHs in both the olive oils and the virgin olive oils most consumed in Italy, at a minimum level of a few /tg/kg. (In European Community countries, 'olive oil' designates a blend of refined and virgin olive oils, besides being a general term.) With some modifications, we used the analytical method previously adopted in the cited study of Corradetti et al. (1988), validated by some of us through a quality control test (Menichini et al. 1991). The general procedure is substantially the same as has been adopted in recent years for such determinations: it involves a
PAHs in olive oils
365
liquid-liquid extraction scheme derived from Natusch and Tomkins (1978), cleanup on silica gel and analysis by high-resolution gas chromatography (hrGC).
Experimental Sampling The olive oils and virgin olive oils which are most consumed in Italy were analysed: six and seven brands, respectively, representing 57% and 50% of the corresponding market (ASSITOL, 1989). The oils (one bottle per each brand) were purchased from shops in Rome during 1989-90. They were stored in the dark pending analysis, which was performed before the product expiry date.
Downloaded by [Purdue University] at 16:28 13 April 2015
PAHs The 28 PAHs to be determined (table 1) were selected so as to include: 3- to 7-ring PAHs; PAHs most commonly found in the environment (Grimmer 1983, Chap. 2); 'probably' or 'possibly' carcinogenic PAHs (IARC 1987) generally detected in air (Grimmer 1983, sect. 3.5; Menichini 1991); PAHs which can indicate vehicular emissions, particularly from gasoline motors, as a source of contamination (IARC 1983); PAHs detected in previous olive oil analyses, to be used for comparison purposes. Materials With regard to the chemicals, the precautions in limiting sample degradation and the apparatus for concentrating solutions, reference can be made to a previous paper (Menichini et al. 1990). In addition, ATR, CPP and PIC were purchased from LabService (Bologna, Italy). Dimethyl sulphoxide (DMSO) and /i-pentane were UV-spectroscopy grade, the other solvents HPLC grade. Silica gel thin-layer chromatography (TLC) plates (20 cm X 20 cm, 1 mm thick layer) were obtained from Merck (Darmstadt, FRG) and washed prior to use with redistilled-in-glass acetone. As internal standards (ISs), a solution of o-terphenyl (o-T), /j-terphenyl (p-T) and triphenylbenzene (TPB) was used, ca. 20 /*g/ml each in cyclohexane. The purity of all analytical material was checked by blank tests at the beginning, middle and end of analysis of each of the two subsets of samples (i.e. olive and virgin olive oils); a peak with the retention time of PHE was found in blank samples at levels near the limit of detection and was subtracted from the results. Analyses were carried out with a gas chromatograph (C. Erba, mod. Mega 5160) equipped with a FID detector, a capillary column (fused silica SE-54 Nordibond 25 m x 0-32 mm, purchased from C. Erba), and an integrator. Extraction and clean-up Each oil sample, i.e. each brand, was analysed in duplicate. The oil was homogenized by shaking. Two 10'0g aliquots were weighed; one was spiked with 50 PL\ of IS solution (the spiked specimen was unknown to the operator processing the oils). The spiked oil was shaken and both specimens were stored in the dark for 24 h. Then, each specimen was dissolved in 20 ml of /z-pentane and successively partitioned with three 10 ml portions of DMSO. The upper pentane layer was discarded. The DMSO extracts were combined and 60 ml of cold water were added
Downloaded by [Purdue University] at 16:28 13 April 2015
366
E. Menichini et al.
(water temp: ca. 4°C. Caution: considerable heat can be generated when mixing DMSO and water). PAHs were back-extracted by partitioning three times with 50 ml of cyclohexane. The three combined cyclohexane extracts were washed with two 100 ml portions of water and concentrated to ca. 50 ml. The extract was filtered through a 20 cm X 22 mm i.d. glass column containing, from the bottom: a degreased cotton-wool plug, a 1 cm silica gel layer, and a 5 cm anhydrous Na2SC"4 layer (the column was pre-washed with 10 ml of dichloromethane (DCM) and 10 ml of cyclohexane). After sample elution, an additional 10 ml of cyclohexane and two 5 ml portions of DCM were percolated, collected and combined with the filtered extract. The resulting solution was concentrated to ca. 50 /tl and applied to the TLC plate which was developed with 1:1 «-hexane/toluene to a height of 12 cm. The PAH spot (/?F: ca. 0*8) was detected under 254 nm UV light. The silica gel was scraped off, poured into a 2 cm i.d. glass tube equipped with sintered glass disc, and eluted with 4 ml of DCM; the silica gel was then washed with two 1 ml portions of DCM. Eluate and washings were combined, concentrated to ca. 50 /tl, solvent exchanged into cyclohexane, and taken up to ca. 100 /tl. Analysis The operating conditions were as follows. Oven temp: programmed from 90 to 280°C at 8°C/min and then held isothermally; detector temp: 300°C; injection system: cold on-column; injection volume: 1 /tl; carrier gas: helium, 2 ml/min. PAH identification and quantification were performed by comparison with a reference mixture containing all PAHs and the three ISs (each compound, ca. 10/ig/ml). ISs were used for calculating relative retention times and estimating PAH recovery in each oil analysis. For the latter purpose, the ratios between recoveries of PAHs and ISs had been calculated in preliminary tests (see following) and, based on such ratios, PAH measurements were correspondingly corrected. As reported above, ISs were added to only one specimen, so a comparison between the spiked and the unspiked specimen chromatograms made it possible to evaluate which IS peak was free from blank interference and could be used for recovery calculation; when interference was lacking, the IS whose retention time was nearest to those of detected PAHs was used. In the conditions reported above, the estimated limit of detection of PAHs was ca. 1-3 and 3-5 /tg/kg in olive oil and virgin oil samples, respectively. Recovery and repeatability PAH recoveries were determined in triplicate, both before and after analysing the whole set of oils. Ten gram samples of an olive oil ('blank oil') were spiked with 50 fil of the reference mixture (see 'Analysis'). (Blank oil had been analysed in triplicate previously and found to contain a few PAHs at /tg/kg levels which were subtracted from the results.) After 24 h, spiked samples were processed as actual samples. PAHs were recovered in the range 65-85%. Poorer recoveries were only obtained for the lightest PAH (FL, ca. 60%) and for methylated PAH derivatives (MAN and DMP, ca. 50%), as previously reported for similar extraction schemes (Natusch and Tomkins 1978, Menichini et al. 1990). As to ISs, o-T and p-T were recovered in the same range as PAHs, while TPB was recovered at a low 45%. The repeatability of the method was evaluated through the coefficients of variation (CVs) associated to measurements of each PAH performed during
PAHs in olive oils
367
recovery tests. All CVs were in the range of 4-18%, with a mean value of 12%. For o-T, p-T and TPB, the CV was 13, 10 and 6%, respectively.
Downloaded by [Purdue University] at 16:28 13 April 2015
Results and discussion
Table 1 shows the results of PAH determinations in oils. The repeatability of the duplicate measurements (not individually reported) of each oil was very good: for any pair of analysed specimens and any PAH, the difference between the two measurements was at the most 3 /tg/kg, and typically only 1 /tg/kg. A lower limit of detection, by approximately one order of magnitude, could be achieved (though beyond the aim of this work) if samples were analysed by high-performance liquid chromatography (HPLC) with a fluorimetric detector. However, we used hrGC since it allows a better separation of the 28 PAHs which we planned to determine, according to criteria reported in 'PAHs' section. The chromatogram of a virgin oil sample is shown in figure 1. PAHs which were detected are commonly found in the general environment, and particularly in the atmosphere; they are all 3- or 4-ring compounds. Their carcinogenicity classification (IARC 1987) is of 'inadequate evidence' in experimental animals (except for CHR: 'limited evidence'; but its identification in this survey is doubtful); among the 28 PAHs, none of those 'probably' or 'possibly' carcinogenic to humans (BaA, BbFA, BjFA, BkFA, BaP, DBahA, IP) was detected. Median concentrations of detected PAHs were
Food Additives & Contaminants Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tfac19
Polycyclic aromatic hydrocarbons in olive oils on the Italian market a
a
a
Edoardo Menichini , Adriana Bocca , Franco Merli , a
Daniela Ianni & Fabio Monfredini
a
a
Istituto Superiore di Sanità , viale Regina Elena 299, Roma, 00161, Italy Published online: 10 Jan 2009.
To cite this article: Edoardo Menichini , Adriana Bocca , Franco Merli , Daniela Ianni & Fabio Monfredini (1991) Polycyclic aromatic hydrocarbons in olive oils on the Italian market , Food Additives & Contaminants, 8:3, 363-369, DOI: 10.1080/02652039109373985 To link to this article: http://dx.doi.org/10.1080/02652039109373985
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sublicensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
FOOD ADDITIVES AND CONTAMINANTS, 1991, VOL. 8, NO. 3 , 3 6 3 - 3 6 9
Polycyclic aromatic hydrocarbons in olive oils on the Italian market * EDOARDO MENICHINI, ADRIANA BOCCA, FRANCO MERLI, DANIELA IANNI † and FABIO MONFREDINI Istituto Superiore di Sanità, viale Regina Elena 299, 00161, Roma, Italy
Downloaded by [Purdue University] at 16:28 13 April 2015
(Received 15 November 1990; revised 25 February 1991; accepted 8 March 1991) The six olive oils and seven virgin olive oils which are most consumed in Italy were analysed for 28 polycyclic aromatic hydrocarbons (PAHs). The aim was to evaluate whether a carcinogenic hazard for the general population can derive from the dietary intake of this food, which is consumed particularly highly in the Mediterranean area. The analytical method involved extraction by liquid-liquid partition, nitration on silica gel, clean-up by thin-layer chromatography on silica gel, and analysis by high-resolution gas chromatography with a flame ionization detector. The 3- and 4-ring PAHs which are most abundant in the environment were found in all samples, at individual levels up to ca. 40 µg/kg (for phenanthrene); no important difference was observed between olive oils and virgin olive oils. PAHs which are most suspected of being carcinogenic for humans were not detected (limit of detection, ca. 3 µg/kg). The average yearly intake of the detected PAHs through this food was estimated at ca. 0.56 mg per capita. Keywords: polycyclic aromatic hydrocarbons, olive oil, virgin olive oil
Introduction
The occurrence of polycyclic aromatic hydrocarbons (PAHs) in foods, as a result of sorption from a contaminated environment or from food preparation methods, is a known and widely reviewed matter (Howard and Fazio 1980, Santodonato etal., 1981, Fazio and Howard 1983, IARC 1983). It is also of primary health interest, owing to the carcinogenic activity of many PAHs (IARC 1983, 1987). More specifically, other reviews have dealt with the PAH presence in olive oils (Tiscornia et al. 1982) and in various vegetable oils (Mariani and Fedeli 1984). Marketed olive oils, apparently not influenced by direct sources of PAH emissions, have been analysed previously, although in most investigations as a single sample. Individual PAHs have generally been detected at /tg/kg levels in refined oil (Howard etal. 1966, Biernoth and Rost 1968, Fabian 1968, Joe etal. 1979, Hopia etal. 1986, Welling and Kaandorp 1986, Stijve and Hischenhuber 1987, Speer etal. 1990), the most commonly detected PAHs being AN, BaA, BghiP, CHR, FA, IP, PHE, PY (for abbreviations, see table 1). Higher
* Note. The results of this work were presented in part at the 1 ° Congresso Nazionale di Chimica degli Alimenti (1st National Congress on Food Chemistry) (Messina and Giardini-Naxos, Italy, 9-13 October 1990). † Guest of the Istituto Superiore di Sanita. 0265-203X/9I $3.00 © 1991 Taylor & Francis Ltd.
364
E. Menichini et al. Table 1. PAHs in olive oils and virgin olive oils.3
Abbrev.
Olive oils Fluorene Phenanthrene Anthracene Fluoranthene Pyrene BaFL 11 H-Benzo [a] fluorene CHR + TRI Chrysene + triphenylene ?
No. 1
No. 2
No. 3
No. 4
No. 5
No. 6
Median
nd 12
nd 4 nd 3 2 nd nd
nd 8 nd 5 5 nd nd
nd 11 nd 7 7 nd 7
nd 15
6 10 2? 6
3 41 4 15 14 2 7
10 12 nd 8
nd 11 nd 6 9 nd 6
Virgin olive oils Fluorene Phenanthrene Anthracene Fluoranthene Pyrene BaFL 11 H-Benzo [a] fluorene CHR + TRI Chrysene + triphenylene ?
No. 7
No. 8
No. 9
No. 10
No. 11
No. 12
No. 13
7 42 5 10 10 nd 3
4 38 7 10 7 nd nd
3 27 6 10 9 nd 3
5 36 8 11 10 nd 5
6 38 8 12 11 nd 5
3 32 7 10
3 25 6 7 7 nd nd
FL PHE AN FA PY
FL PHE AN FA PY
Downloaded by [Purdue University] at 16:28 13 April 2015
Concentration (jig/kg)1"
Compound
nd
nd
.
10 nd 3
Median 4
36 7 10 10 nd 3
nd: not detected (for limit of detection, see text). ?: doubtful identification. a The following PAHs were not detected in any sample: 2-methylanthracene; 3,6-dimethylphenanthrene; 11Hbenzo [b] fluorene; benzo[b]naphtho[2,l-d]thiophene; benzo [ghi] fluoranthene; cyclopenta[cd] pyrene (CPP); benz [a) anthracene (BaA); benzo [b] fluoranthene (BbFA); benzo [j] fluoranthene (BjFA); benzo [k] fluoranthene (BkFA); benzo [e] pyrene; benzo [a] pyrene (BaP); perylene; indeno[l,2,3-cd)pyrene (IP); dibenz [a,c] anthracene; dibenz[a,h] anthracene (DBahA); picene (PIC); benzo [ghi] perylene (BghiP); dibenzo [def ,mno]chrysene (known as anthanthrene, ATR); coronene. "Mean values of duplicate measurements.
concentrations, up to 40/ig/kg of BaP and 196/tg/kg of PHE, were found in one sample of refined oil (Morgante 1973) and two samples of virgin olive oils (Ciusa et al. 1970, Morgante 1973). (For a comparative exposition of such determinations, see the review of Menichini and Bocca 1991.) Particularly high concentrations were found recently (Corradetti et al. 1988) in Italian virgin oils deriving from plants exposed to industrial emissions of pitch condensate (BaP and PHE, up to ca. 60 and 3800 /tg/kg, respectively) and, to a lesser extent (PHE, ca. 800 iig/kg), in virgin oils deriving from plants exposed to vehicular exhausts. Following these findings, and also considering the high consumption of this food in Italy as well as in other Mediterranean countries, it appeared important to check whether PAHs present in the environment actually contaminate olive oils found on the market to any serious extent, with reference to those commonly measured in most foods (some /tg/kg per individual PAH; Santodonato et al. 1981); hence, whether these oils pose any special carcinogenic risk for the general population. In this paper, we present the results of an investigation concerning the occurrence of 28 PAHs in both the olive oils and the virgin olive oils most consumed in Italy, at a minimum level of a few /tg/kg. (In European Community countries, 'olive oil' designates a blend of refined and virgin olive oils, besides being a general term.) With some modifications, we used the analytical method previously adopted in the cited study of Corradetti et al. (1988), validated by some of us through a quality control test (Menichini et al. 1991). The general procedure is substantially the same as has been adopted in recent years for such determinations: it involves a
PAHs in olive oils
365
liquid-liquid extraction scheme derived from Natusch and Tomkins (1978), cleanup on silica gel and analysis by high-resolution gas chromatography (hrGC).
Experimental Sampling The olive oils and virgin olive oils which are most consumed in Italy were analysed: six and seven brands, respectively, representing 57% and 50% of the corresponding market (ASSITOL, 1989). The oils (one bottle per each brand) were purchased from shops in Rome during 1989-90. They were stored in the dark pending analysis, which was performed before the product expiry date.
Downloaded by [Purdue University] at 16:28 13 April 2015
PAHs The 28 PAHs to be determined (table 1) were selected so as to include: 3- to 7-ring PAHs; PAHs most commonly found in the environment (Grimmer 1983, Chap. 2); 'probably' or 'possibly' carcinogenic PAHs (IARC 1987) generally detected in air (Grimmer 1983, sect. 3.5; Menichini 1991); PAHs which can indicate vehicular emissions, particularly from gasoline motors, as a source of contamination (IARC 1983); PAHs detected in previous olive oil analyses, to be used for comparison purposes. Materials With regard to the chemicals, the precautions in limiting sample degradation and the apparatus for concentrating solutions, reference can be made to a previous paper (Menichini et al. 1990). In addition, ATR, CPP and PIC were purchased from LabService (Bologna, Italy). Dimethyl sulphoxide (DMSO) and /i-pentane were UV-spectroscopy grade, the other solvents HPLC grade. Silica gel thin-layer chromatography (TLC) plates (20 cm X 20 cm, 1 mm thick layer) were obtained from Merck (Darmstadt, FRG) and washed prior to use with redistilled-in-glass acetone. As internal standards (ISs), a solution of o-terphenyl (o-T), /j-terphenyl (p-T) and triphenylbenzene (TPB) was used, ca. 20 /*g/ml each in cyclohexane. The purity of all analytical material was checked by blank tests at the beginning, middle and end of analysis of each of the two subsets of samples (i.e. olive and virgin olive oils); a peak with the retention time of PHE was found in blank samples at levels near the limit of detection and was subtracted from the results. Analyses were carried out with a gas chromatograph (C. Erba, mod. Mega 5160) equipped with a FID detector, a capillary column (fused silica SE-54 Nordibond 25 m x 0-32 mm, purchased from C. Erba), and an integrator. Extraction and clean-up Each oil sample, i.e. each brand, was analysed in duplicate. The oil was homogenized by shaking. Two 10'0g aliquots were weighed; one was spiked with 50 PL\ of IS solution (the spiked specimen was unknown to the operator processing the oils). The spiked oil was shaken and both specimens were stored in the dark for 24 h. Then, each specimen was dissolved in 20 ml of /z-pentane and successively partitioned with three 10 ml portions of DMSO. The upper pentane layer was discarded. The DMSO extracts were combined and 60 ml of cold water were added
Downloaded by [Purdue University] at 16:28 13 April 2015
366
E. Menichini et al.
(water temp: ca. 4°C. Caution: considerable heat can be generated when mixing DMSO and water). PAHs were back-extracted by partitioning three times with 50 ml of cyclohexane. The three combined cyclohexane extracts were washed with two 100 ml portions of water and concentrated to ca. 50 ml. The extract was filtered through a 20 cm X 22 mm i.d. glass column containing, from the bottom: a degreased cotton-wool plug, a 1 cm silica gel layer, and a 5 cm anhydrous Na2SC"4 layer (the column was pre-washed with 10 ml of dichloromethane (DCM) and 10 ml of cyclohexane). After sample elution, an additional 10 ml of cyclohexane and two 5 ml portions of DCM were percolated, collected and combined with the filtered extract. The resulting solution was concentrated to ca. 50 /tl and applied to the TLC plate which was developed with 1:1 «-hexane/toluene to a height of 12 cm. The PAH spot (/?F: ca. 0*8) was detected under 254 nm UV light. The silica gel was scraped off, poured into a 2 cm i.d. glass tube equipped with sintered glass disc, and eluted with 4 ml of DCM; the silica gel was then washed with two 1 ml portions of DCM. Eluate and washings were combined, concentrated to ca. 50 /tl, solvent exchanged into cyclohexane, and taken up to ca. 100 /tl. Analysis The operating conditions were as follows. Oven temp: programmed from 90 to 280°C at 8°C/min and then held isothermally; detector temp: 300°C; injection system: cold on-column; injection volume: 1 /tl; carrier gas: helium, 2 ml/min. PAH identification and quantification were performed by comparison with a reference mixture containing all PAHs and the three ISs (each compound, ca. 10/ig/ml). ISs were used for calculating relative retention times and estimating PAH recovery in each oil analysis. For the latter purpose, the ratios between recoveries of PAHs and ISs had been calculated in preliminary tests (see following) and, based on such ratios, PAH measurements were correspondingly corrected. As reported above, ISs were added to only one specimen, so a comparison between the spiked and the unspiked specimen chromatograms made it possible to evaluate which IS peak was free from blank interference and could be used for recovery calculation; when interference was lacking, the IS whose retention time was nearest to those of detected PAHs was used. In the conditions reported above, the estimated limit of detection of PAHs was ca. 1-3 and 3-5 /tg/kg in olive oil and virgin oil samples, respectively. Recovery and repeatability PAH recoveries were determined in triplicate, both before and after analysing the whole set of oils. Ten gram samples of an olive oil ('blank oil') were spiked with 50 fil of the reference mixture (see 'Analysis'). (Blank oil had been analysed in triplicate previously and found to contain a few PAHs at /tg/kg levels which were subtracted from the results.) After 24 h, spiked samples were processed as actual samples. PAHs were recovered in the range 65-85%. Poorer recoveries were only obtained for the lightest PAH (FL, ca. 60%) and for methylated PAH derivatives (MAN and DMP, ca. 50%), as previously reported for similar extraction schemes (Natusch and Tomkins 1978, Menichini et al. 1990). As to ISs, o-T and p-T were recovered in the same range as PAHs, while TPB was recovered at a low 45%. The repeatability of the method was evaluated through the coefficients of variation (CVs) associated to measurements of each PAH performed during
PAHs in olive oils
367
recovery tests. All CVs were in the range of 4-18%, with a mean value of 12%. For o-T, p-T and TPB, the CV was 13, 10 and 6%, respectively.
Downloaded by [Purdue University] at 16:28 13 April 2015
Results and discussion
Table 1 shows the results of PAH determinations in oils. The repeatability of the duplicate measurements (not individually reported) of each oil was very good: for any pair of analysed specimens and any PAH, the difference between the two measurements was at the most 3 /tg/kg, and typically only 1 /tg/kg. A lower limit of detection, by approximately one order of magnitude, could be achieved (though beyond the aim of this work) if samples were analysed by high-performance liquid chromatography (HPLC) with a fluorimetric detector. However, we used hrGC since it allows a better separation of the 28 PAHs which we planned to determine, according to criteria reported in 'PAHs' section. The chromatogram of a virgin oil sample is shown in figure 1. PAHs which were detected are commonly found in the general environment, and particularly in the atmosphere; they are all 3- or 4-ring compounds. Their carcinogenicity classification (IARC 1987) is of 'inadequate evidence' in experimental animals (except for CHR: 'limited evidence'; but its identification in this survey is doubtful); among the 28 PAHs, none of those 'probably' or 'possibly' carcinogenic to humans (BaA, BbFA, BjFA, BkFA, BaP, DBahA, IP) was detected. Median concentrations of detected PAHs were
