Phthalate ester residues--their metabolism and analysis in fish.

During the course of our analysis of fish and other aquatic materials for pesticide residues, unknown components were often detected. Concern about the identity of late eluting gas-liquid chromatographic (GLC) peaks in fish and water extracts led to the identification of phthalic acid esters (PAEs) as environmental contaminants (1). Further study of these contaminants with a GLC-mass spectrometer (GLC-MS) led to their identification as di-2-ethylhexyl phthalate (DEHP) and di-n-butyl phthalate (DBP). Preliminary data from acute toxicity tests indicate that these compounds are relatively low in toxicity. Chronic studies, however, indicate that they may have more subtle effects on reproduction of some aquatic species (2). Residue data obtained from a variety of areas, dietary materials, and samples related to fish rearing operations have established some information on the incidence of PAEs in relation to pesticides and polychlorinated biphenyl contaminants (Table 1). PAE residues in water appeared to be correlated with rivers draining industrialized or heavily populated areas (John Hesse, Michigan Department of Natural Resources, East Lansing, Michigan 1972 personal communication). However, the distribution of PAEs in the aquatic environment remains to be defined. PAEs are commonly used as plasticizers, especially in poly(vinyl chloride) plastics. In

During the course of our analysis of fish and other aquatic materials for pesticide residues, unknown components were often detected. Concern about the identity of late eluting gas-liquid chromatographic (GLC) peaks in fish and water extracts led to the identification of phthalic acid esters (PAEs) as environmental contaminants (1). Further study of these contaminants with a GLC-mass spectrometer (GLC-MS) led to their identification as di-2-ethylhexyl phthalate (DEHP) and di-n-butyl phthalate (DBP). Preliminary data from acute toxicity tests indicate that these compounds are relatively low in toxicity. Chronic studies, however, indicate that they may have more subtle effects on reproduction of some aquatic species (2).
Residue data obtained from a variety of areas, dietary materials, and samples related to fish rearing operations have established some information on the incidence of PAEs in relation to pesticides and polychlorinated biphenyl contaminants (Table 1). PAE residues in water appeared to be correlated with rivers draining industrialized or heavily populated areas (John Hesse, Michigan Department of Natural Resources, East Lansing, Michigan 1972 personal communication). However, the distribution of PAEs in the aquatic environment remains to be defined.
PAEs are commonly used as plasticizers, especially in poly(vinyl chloride) plastics. In addition, DBP and DEHP have been used as an insect repellent and pesticide, respectively (3). Production of PAEs exceeded 800 million pounds in 1969 (4). Despite their widespread use, only limited information exists oli the fate of PAEs in biological systems or the aquatic environment. Water quality criteria have not been adopted for PAE compounds, but such criteria are currently under consideration. Our study of PAEs was divided into four areas to better define the interaction of PAEs with aquatic organism. In vivo and in vitro metabolism studies were conducted to determine the uptake, degradation, and residue composition of PAEs in fish. Further studies were undertaken to develop analytical methods applicable to the GLC determination of residues which may be associated with the presence of PAEs in the aquatic environment. PAE and metabolite identities are also confirmed by GLC-MS and computer processing of spectral information.
In Vivo Metabolism of DEHP PAEs are detected with existing GLC methods for pesticide residues. However, we did not know whether PAEs were degraded into other products as well. Therefore, we exposed channel catfish (Ictalurus punctatus) and fathead minnows (Pimephales promelas) to 14 C-carboxyl-labeled di-2-ethylhexyl phthalate to determine the identity and composition of PAE degradation products.
Identification of the degradation products is a prerequisite for development of analytical methods and further toxicological studies of PAEs.
Extraction-Several fish were combined and ground in an Osterizer blender with Na2SO4 (1:4, v/w). Each sample was then packed into a glass column (1-2 cm ID) and  7000  60  DDT-30, dieldrin-60  20  190  20  170  20  140  30  400  30  200   80   20 Chlordane-720 extracted with 1% H3PO4 in acetone (5). The ratio of extraction solvent to sample size was 10:1 (v/w). Extraction efficiency for removal of radioactivity from the fish tissue was greater than 98% as determined by scintillation counting of tissue and the extracted activity. After dilution of the extract to a known volume (10 ml/g tissue), an aliquot corresponding to 2 g of fish tissue was neutralized with concentrated NH40H. The neutralized solution was evaporated to 2-3 ml at 1000C using a hot plate. The sample was transferred to a centrifuge tube with the use of a 1:1 (v/v) mixture of methanol and acetone containing 0.1% acetic acid. The sample was centrifuged to remove insoluble material, i.e., (NH4 )3PO4, and the solution transferred into a conical centrifuge tube. The centrifuged material was resuspended in the methanol:acetone:acetic acid solution and recentrifuged. The tube was washed again to remove all of the radioactivity, and the supernatants combined and concentrated to 0.2 ml.
Chromatographic Characterization-The concentrated solution was applied in a 15 cm streak, 1.5 cm from the bottom of a Mallinckrodt Chrom AR-1000 chromatographic sheet (20 x 20 cm). Combined standard compounds were spotted at a point 1 cm from the vertical edge of the sheet. The sheet was suspended from a stainless steel rack with spring clips and placed in an equilibrated glass chromatographic tank and developed with petroleum ether:diethyl ether:acetic acid (80:20:5, v/v/v). The solvent front was allowed to migrate 19 cm (ca. 15 min). The sheet was then removed, dried, and examined under a shortwave ultraviolet lamp. The bands corresponding to phthalate esters and metabolites were located and marked (Fig. 1).
The radioactivity in each band, except for. the conjugate band at the origin, was determined by shredding that section of the sheet in a liquid scintillation vial. Scintillation cocktail (20 ml of Beckman Fluoralloy in toluene) was added, each vial counted, and corrected for quench. Quench was minimal, and counting efficiency was 85-95%.
The band containing conjugates was shredded into a 25 ml centrifuge tube containing 15 ml of 0.5N acetate buffer and 1 ml of Ketodase and incubated overnight at 370C. This suspension was acidified with 2 ml H3 PO4, cooled, and extracted three times with 10-ml portions of diethyl ether (Caution! Cool the centrifuge tube after addition of H3PO4). The combined diethyl ether extracts were dried with Na2 SO4 in an Erlenmeyer flask, transferred to a conical centrifuge tube, and evaporated to 0.2 ml. This solution was streaked in a 2.5 cm long band on a Chrom AR sheet. Authentic standards were spotted and the sheet developed as previously described. The origin and the bands corresponding to each standard were examined for radioactivity.
Derivatization-Material from the channel catfish extracts with an Rf corresponding to the phthalic acid standard was recovered from the chromatosheet fiber by extraction three times at pH 1.5 with 10 ml diethyl ether, dried with anhydrous Na2SO4, divided into equal aliquots, and evaporated to 1 ml. Then, one aliquot was reacted with ethereal diazomethane, and the esterified material and unreacted aliquot were evaporated to 0.1 ml. Esterified and unreacted aliquots, and a dimethyl phthalate (DMP) standard were spotted on an Eastman 6060 Chromatogram silicia gel sheet which was developed with petroleum ether:diethyl ether:acetic acid (80:20:2, v/v/v).
Chromatographic spots corresponding to DMP and phthalic acid were recovered and transferred to a scintillation vial. The radioactivity in the diazotized sample was recovered in the spot corresponding to DMP. A similar derivatization study was made with the material corresponding to mono-2ethylhexyl phthalate that had been separated from the channel catfish extract. Separation of the methylated 2-ethylhexyl phthalate and mono-2-ethylhexyl phthalate was not complete with the solvent system used previously. Thus, a GLC examination to determine the identity of the isolated methylated monoester was undertaken. The GLC elution of the methyl ester corresponded to that of methyl 2-ethylhexyl phthalate.

Results and Discussion
Exposure of fish to 14C-labeled DEHP enabled us to separate biological degradation products by using TLC chromatosheets. The radioactivity was distributed among four products which were then identified by TLC or GLC, chromatography after enzymatic hydrolysis, and/or derivatization, to be mono-2-ethylhexyl phthalate, a monoester glucuronide, and phthalic acid and a phthalic acid glucuronide. The glucosides, probably glucuronides, remained near the origin during TLC. Hydrolysis of these materials with Ketodase, followed by TLC, showed that they were formed from phthalic acid and mono-2-ethylhexyl phthalate. The monoester aglycone represented 98% of the activity found as glucuronides. The presence of appreciable amounts of mono-2-ethylhexyl phthalate in the fish suggests that this metabolite is refractive to further degradation. The relative composition of separated metabolites in the channel catfish extracts were determined by scintillation counting (Table 2).
An additional study was made to characterize the composition of the material extracted with diethyl ether at pH 1.5 and to determine whether the glucuronides could be separated from unconjugated metabolites without prior chromatographic separation. The radioactive material remaining in the aqueous phase after extraction of an unhydrolyzed fish extract at pH 1.5 was determined. After hydrolysis of the ether extract, chromatographic examination revealed a much higher ratio of phthalic acid conjugate than in the initial sample extract. The monoester glucuronide was found to be partially extracted with diethyl ether at pH 1.5. Thus, chromatographic resolution of the glucuronide prior to hydrolysis was necessary to determine the ratio of the glucuronides of the monoester and phthalic acid. The composition of the residues differed from those determined in the channel catfish ( Table 3). The ratio (Tables 2 and 3). of mono-2-ethylhexyl phthalate to DEHP was lower in the fathead minnows than channel catfish, as were the relative levels of conjugates. The significance of these differences cannot be assessed at this time because the concentration of DEHP in water, and the exposures were different. However, the monoester residues in fish represent an appreciable portion of the total DEHP residues that are accumulated. Therefore, the monoesters should be considered in analyses for phthalate ester residues and toxicological studies should be designed to assess the impact of phthalate esters on fish and other aquatic forms.
Environmental Health Perspectives 162 Table 3. Composition of residues in fathead minnows continuously exposed to C14di-2-ethylhexyl phthalate (DEHP In Vitro Metabolism of Phthalic Acid Esters While PAE residues have been detected in fish from many locations (1), metabolism of phthalates has not been reported. Metabolism of some xenobiotics by fish liver preparations has been studied (6)(7)(8)(9). In most reports to date, the compounds studied were organophosphate pesticides. Generally metabolism of xenobiotics by fish takes place in hepatic microsomes. Therefore, we examined the metabolism of 14C-di-2ethylhexyl phthalate and 14C-di-n-butyl phthalate in vitro by microsomal preparations from channel catfish liver.

Experimental Procedures
Carboxyl-1 4C-di-2-ethylhexyl phthalate (specific activity 1.64 mCi/mmole) and carboxyl-14C-di-n-butyl phthalate (specific activity 1.53 mCi/mmole) were purchased from Mallinckrodt. An authentic standard of mono-2-ethylhexyl phthalate was prepared in this laboratory and characterized with TLC, GLC, and GLC-MS. Standards of DEHP and DBP were obtained from Monsanto Company and Chem. Services, Inc.
Male channel catfish (300-600 g) were obtained from the Fish Farming Experimental Station, Bureau of Sport Fisheries and Wildlife, Stuttgart, Arkansas. The spinal cord of each fish was severed, and the liver removed. Hepatic microsomal preparations were obtained by differential centrifugation (1 ).
Unless exceptions are noted, the incubation conditions for these in vitro studies were as follows: 0.5 ,umole of NADPH, 0.03-0.06 ,umole of 14C-carboxyl phthalic acid ester, and microsomal preparation equivalent to 200 mg of liver. Incubation was conducted in 2.0 ml of 0.05M Tris-HCI buffer (pH 7.4), with shaking in a water bath at 25°C. Incubation was usually for 2 hr. Following incubation the pH of the system was adjusted to <pH 2 with concentrated HCl. PAEs and metabolites were then extracted from the incubation medium by use of diethyl ether.
PAEs and certain metabolites were identified by using TLC, GLC, or GLC-MS. PAEs and their metabolites were separated by TLC on 0.2 mm silicia gel plates containing a fluorescent indicator. DEHP and DBP were cochromatographed with authentic standards. TLC plates were developed three times with petroleum ether:diethyl ether:acetic acid (77:20:3). 14C-Phthalic acid esters and their metabolites were visualized using autoradiography. Kodak x-ray plates were exposed to thin-layer chromatograms. Radioactive areas on the TLC plates were scraped, placed in a disposable pipet, and extracted with diethyl ether or diethyl ether containing 5% acetic acid. Some radioactive areas were scraped directly into scintillation vials for quantitation. Derivatives of some PAE metabolites were prepared by reacting samples extracted from TLC scrapings with excess diazomethane to form methyl esters.
Radioactive PAEs and metabolites were quantitated by using liquid scintillation spectrometry.

Results and Discussion
At least four metabolites of DBP were detected by autoradiography (Fig. 2). Three of these metabolites (numbered 0, 1, and 2) have not been identified, and analogous metabolites were not found in the in vivo study with DEHP. The fourth metabolite has been identified in the in vivo and in vitro experiments as either mono-2-ethylhexyl or mono-n-butyl phthalate, depending on the parent compound.
The formation of the three as yet unidentified metabolites of either PAE was inhibited by carbon monoxide ( shown. Formation of the monoester metabolite of either di-2-ethylhexyl or di-n-butyl phthalate was unaffected by carbon monoxide or a lack of oxygen. Metabolism of DBP and DEHP was affected similarly by air, nitrogen, or carbon monoxide. As with the effects of air, nitrogen or carbon monoxide, NADPH also had no effect on formation of the monoester metabolites of either phthalic acid diester (Table  5). It is obvious, however, that NADPH is required for the formation of metabolites, 0, 1, and 2.
Enzymatic degradation of DBP was more complete than that of DEHP (Table 6), 97% degradation of DBP and only 6% of DEHP occurring within 2 hrs. The total amount of metabolism, however, was found to vary considerably with hepatic microsomes prepared from different channel catfish.  The length of incubation time has a definite effect on the metabolism of DBP (Fig. 3). Formation of the monoester metabolite (mono-n-butyl phthalate) appears to be relatively independent of time, at least for the initial 60 min of incubation. A slight increase in formation of mono-n-butyl phthalate was noted during the final 60 min of incubation but the maximum amount of metabolite 0 was formed in 120 min. Maximum amounts of metabolites 1 and 2 were present after 15 min incubation, and further incubation only decreased the amounts of these metabolites.
An additional study was undertaken to clarify whether the unidentified metabolites 0, 1, and 2 were derived from the monoester metabolite or parent PAE. Reincubation of monoester metabolites of DEHP and DBP with a microsomal preparation resulted in no further metabolism of the monoesters.
Results of the inhibition and cofactor experiments indicate that at least two separate and distinct microsomal enzyme systems degrade PAEs. One enzyme system, responsible for the formation of metabolites 0, 1, and 2, is carbon monoxide-sensitive and requires NADPH. The other enzyme, responsible for the formation of monoester metab- olites, is not carbon monoxide sensitive and does not require NADPH. These observations indicate that the monoester is a terminal metabolite in this system and that all metabolites derive from the parent phthalic acid diester. While insufficient evidence is available to identify definitely the enzyme systems present, it does seem likely that metabolites 0, 1, and 2 are formed by a mixed function oxidase while the monoester metabolites may be formed by an esterase.

Analysis of Phthalate Esters and Metabolic Products
in Fish by GLC The following discussion on GLC analytical techniques presents procedures we have developed in our study on PAEs and some of the problems which have been encountered. Analysis of PAEs and their degradation products in biological tissues has not, as yet, been reduced to routine analyses. Some progress toward this goal was achieved with the identification of the predominant metabolic products, the monoalkyl esters and corresponding glucuronides. These more polar products cannot be directly analyzed by GLC. GLC analyses are possible, however, if the glucuronides are chemically or enzymatically cleaved and the resulting acid group(s) derivatized to form methyl alkyl phthalates.

Extraction Procedures
The most satisfactory procedure for extraction of the phthalates and metabolic products was achieved with a variation of the column extraction procedure reported by Hesselberg and Johnson (5).
Extraction procedure-The first step is to prepare a homogenous sample by blending tissue and anhydrous Na2SO4 in a 1:4 (w/w) ratio. Then an amount of sample-Na2SO4 mixture equivalent to 20 g of sample is transferred to the extraction column and prepared as described in the literature (5).
The sample is then extracted by percolation with 200 ml of 1% (v/v) H3P04 in acetone and the eluate collected in a casserole dish. The column eluate, is evaporated to approxi-mately 25 ml. The extract is then transferred to a 500 ml separatory funnel with two 5 ml acetone rinses and two 5 ml diethyl ether rinses. To this is added 200 ml of distilled water and the pH of the solution adjusted to 10.0 with 5N NaOH. Cleanup of sample extract-Cleanup of the sample will depend upon the objectives of the analysis. The diesters are not appreciably soluble in water or base and will be removed from the basic solution along with the coextracted lipid by extraction with diethyl ether. The monoesters and phthalic acid can be separated by extraction at pH 1.5 along with some of the monoester conjugates. After removal of the unconjugated material, the monoesters and phthalic acid conjugates are enzymatically cleaved by using To determine the unconjugated phthalate monoester and phthalic acid, proceed with the following step 3. To isolate total monoesters and phthalic acid proceed with step 5, omitting steps 3 and 4. (3) The pH is adjusted to 1.5 with H3PO4 and extraction is performed twice with 50 ml of diethyl ether; (4) the extracts from step 3 are combined, evaporated to 2-4 ml, transferred to a 10 ml calibrated culture tube, and the monoesters and phthalic acid derivatized. solution is transferred to a 250-ml Erlenmeyer flask, warmed gently to remove the diethyl ether, and allowed to cool to room temperature. (7) A 5-ml portion of Ketodase is then added and incubated overnight at 370 C. (8) The solution is transferred to a separatory funnel and the pH adjusted to 1.5 with H3PO4. (9) The extracts are combined and evaporated to 2-4 ml; the monoester and phthalic acid are transferred to a calibrated 10-ml culture tube with diethyl ether and derivatized.

Derivatization
Monoesters and phthalic acid are derivatized effectively by reaction with excess diazomethane (CH2 N2) in diethyl ether. This reagent forms methyl esters rapidly and quantitatively. Diazomethane is readily prepared from Diazald (11). Ethereal diazomethane is added to the solution obtained in the extraction procedure [step (4) or (10)] until the yellow color persists.
An aliquot is diluted or concentrated if required for cleanup or analysis.
Cleanup of sample extracts containing PAEs requires removal of coextracted fish lipids and two approaches may be utilized. One approach is to employ conventional cleanup by using solvent partition as described in the Official Methods of Analysis of the AOAC, 11th edition, 1970, section 29.011, combined with chromatography on Florisil (section 20.014) (12). We determined that DEHP has a p value in hexaneacetonitrile of 0.65 and, thus, its recovery suffers from the same difficulties as p,p'-DDE, while DBP has a p value of 0.15.
Chromatography of DBP and DEHP on Florisil is analogous to that of dieldrin. The PAEs are recovered in the 15% eluate from the Florisil column (13). An examination of the 6% diethyl ether-petroleum ether eluate from a Florisil column should be made for DEHP, as this material may partially elute in the 6% fraction if the Florisil is too deactivated. Dieldrin may behave in an analagous manner.
An alternate and preferred approach to separation of lipids and PAEs is the substitution of gel permeation chromatography (GPC) for the partition step in the cleanup Environmental Health Perspectives procedure (13). A modification of the collection volume is required as DEHP elutes somewhat earlier than the chlorinated insecticides. Collection of the GPC effluent is begun after 70 ml of cyclohexane is eluted instead of the recommended 100 ml. Recovery of DBP was quantitative (published procedure) when collection was initiated after 100 ml of elution, but only 55-60% of the DEHP was recovered.

Gas-Liquid Chromatographic Analyses
Di-n-butyl, di-2-ethylhexyl, methyl nbutyl, and methyl 2-ethylhexyl phthalate are readily detected by using the conventional electron capture cell and a lightly loaded GLC column (Fig. 4). Corning GLC-110 glass bead support coated with OV-7 or OV-11 is very satisfactory as GLC column   packings for separation of PAEs (13). Confirmation of the identity of the various esters is readily achieved from their GLC retention time combined with their mass spectrum. The GLC retention time of the PAEs does not pose a significant interference problem in pesticide analyses ( Table 7). The GLC retention times relative to aldrin on an OV-7 column were 1.24 and 14.9 for DBP and DEHP, respectively. The retention time of methoxychlor relative to aldrin (OV-7) was 11.9. The OV-7 GLC column temperature was 1550C and the nitrogen carrier flow was 20 ml/min. The electron capture detector area response relative to an equal amount of DDT was 0.12 for DBP and 0.09 for DEHP.
Various components of laboratory equipment, such as blender seals, disposable plastic gloves and plastic tubing, were found to be sources of PAE contamination in residue samples. Reagent-grade anhydrous Na2 SO4 also contained some DBP and prewashing of the Na2SO4 with petroleum ether was required before its use in tissue extractions.
A factor to be considered in the derivatization of phthalic acid is the volatily of (DMP). Quantitative analysis of DMP is difficult under GLC conditions which elute DEHP in a reasonable time due to the short retention time of the dimethyl ester. Derivatization of phthalic acid to a less volatile derivative can be achieved by esterification with 2-chloroethanol and BC13. The di-2-chloroethyl ester is well suited for chromatographic analysis under GLC conditions similar to those required for chlorinated insecticides (Fig. 4). Use of the 2-chloroethyl derivative requires that a separate aliquot of the recovered metabolic products be derivatized. If chemical hydrolysis was employed to recover phthalic acid from all of the esters and metabolic products in the sample, the 2-chloroethyl diester would be a good choice of derivative.

GLC-MS Identification of Phthalate Esters and Metabolites
The spectra of nonsymmetrical methyl alkyl esters we examined contained frag- (3) (2) Characteristics of the electron impact mass spectra of phthalate esters permit ready identification of the esters as a group of compounds. The structure is indicated by the presence of a moderately intense ion corresponding to (M-R+2H)+ in the spectra of symetrically substituted esters and an ion, R+ or (R-H)+ corresponding to the alcohol moiety (14). For the higher esters, a very intense ion is observed at m/e+ of 149 which corresponds to the protonated phthalic anhydride ion: Cp 0 N? C~~..,O+ H (1) However, no parent ion is generally observed with the higher alkyl esters by using electron impact mass spectrometry (15). Field ionization mass spectrometry was applied to the study of mixtures of phthalate esters and with this technique molecular ions are observed for mixed esters (16). in addition to the R+ or (R-H)+ ion. Dimethyl phthalate does not yield a significant 149 m/e+ ion fragment, and a molecular ion peak is observed (m/e+ 194). The base peak in the spectrum of dimethyl phthalate is 163 m/e+ (15). Combination of GLC with mass spectrometry (GLC-MS) simplifies interpretation of the spectrum as the occurrence of mixtures of compounds is greatly reduced. Computer reduction of the spectra from GLC-MS scans enables one to analyze environmental samples in more detail than is possible with manual data reduction.

Experimental Procedures
A PDP-12 LDP computer was interfaced to a Perkin-Elmer Model 270-B GLC-MS. The interface was expanded to contain an analogue squaring network with variable gain. The squaring network linearized the relation between the magnetic field intensity and m/e+ ratio. Further, the network resulted in a uniform mass error throughout the mass range (14-550 amu) of approximately ± 0.15 amu. The low resolution software system MASH (as furnished by Digital Equipment Corporation) was used for data Environmental Health Perspectives acquisition and presentation (1 7). Mass fragmentographs were constructed from sequentially acquired spectra using the program MCHROM.
A 3.6 m long x 2 mm ID glass column was filled with 0.3% (w/w) OV-7 coated on 80-100 mesh Coming GLC-110 glass bead support. The column temperature was programmed from 1700C at 5°C/min to 235°C with 18 psig helium as the carrier gas.
Spectra scans were initiated and acquired under computer program control every 8 sec with a scan duration of 4 sec. The scan rate was 3 sec/mass decade over a mass range 14-550, and 70 to 140 sequential scans were acquired during a GLC-MS run. Perfluorokerosine was used for calibration of the mass spectrometer and was introduced into the MS system through a heated variable leak.
After the parent compound or methylated metabolic products (dimethyl phthalate, methyl n-butyl phthalate, methyl 2ethylhexyl phthalate, or di-2-ethylhexyl phthalate) were recovered from TLC and derivatized, the samples were concentrated to 1-5 Mg/Ml. and injected into the GLC-MS apparatus for analysis. Mass searches were then conducted on each set of spectra by using either mass 149 or 163 or both. The spectrum containing the specified base peak was then examined and compared to that of an authentic derivative.

Results and Discussion
A mass fragmentogram was constructed from 140 sequential scans of a mixture of dimethyl, methyl 2-ethylhexyl, and di-2ethylhexyl phthalates by combining the intensities of ion fragments 149 and 163 in each spectrum (Fig. 5). The low apparent concentration of dimethyl phthalate is due to the lack of a significant 149 m/e+ fragment in the spectrum of this compouna ( Table 8). The base peak in the spectrum of methyl 2-ethylhexyl phthalate is 163 (Fig.  6). Note the ion fragment of 112 m/e+ which corresponds to C8H16+, the long chain alkyl portion of the ester. As previously stated, the base peak of the spectrum of di-2-ethylhexyl phthalate contains an intense 149 m/e+ ion (base peak) and an ion fragment corresponding to C8H16+ (112 m/e+) ( Table 9).
Confirmation of the identity of several of the in vitro degradation products of DBP and DEHP was achieved from the spectrum of the methylated monoesters of phthalic acid. The metabolic products were separated by TLC, recovered, and reacted with diazomethane. The reaction products were examined by GLC-MS, and, by using the computer the spectra were examined for masses 149 and 163. Upon location of a GLC peak containing the specified mass(es), the corresponding spectrum of the compound was compared to that of a standard spectrum.
Identification of methyl n-butyl phthalate demonstrated that one ester group was hydrolyzed during in vitro metabolism of DBP (Table 10)  2-ethylhexyl phthalate (Fig. 7), was identified from the in vitro studies with DEHP.
The mass spectra of the methylated metabolic compounds from the in vivo exposures were consistent with those of the in vitro studies. TLC of the in vivo extracts revealed that one spot of lower Rf than phthalic acid remained near or at the origin. This compound has not been identified but could be a hydroxylated derivative of either phthalic acid or the monoester. Acylation of the material remaining at the origin with trifluroacetic anhydride and subsequent reaction with diazomethane may yield a suitable derivative for GLC-MS identification.
The monoesters of phthalic acid are formed by fish upon exposure to the diester and should be considered in the analysis of environmental samples. Their toxicity at present, however, remains undetermined. Summary Fish from various locations in the United States have been analyzed for DEHP by GLC. Residue levels ranged from 0.2 to 10.0 ,g/g on a whole fish basis. Higher phthalate ester residues appear to be associated with industrialized areas. Significant residues also occurred in commercially reared channel catfish and dietary contamination was suspected as the source of the phthalate ester. Laboratory contamination from plastics was  of the radioactive residues was examined by TLC. Four metabolic products were separated and identified as: mono-2-ethylhexyl phthalate, the corresponding monoester glucuronide (unknown aglycone), phthalic acid, and a phthalic acid glucuronide. The monoester was the predominant metabolite identified by TLC. DBP was metabolized in vitro sixteen times more rapidly than DEHP by hepatic microsomes from male channel catfish. Formation of monoester metabolites of di-nbutyl and di-2-ethylhexyl phthalates was not inhibited by carbon monoxide and did not require NADPH. Formation of three other unidentified metabolites, however, was inhibited by carbon monoxide and did require NADPH. The phthalic acid monoesters were not further metabolized and appeared to be a terminal metabolite for the in vitro system under study. At least two separate and distinct enzyme systems appeared to metabolize phthalic acid diesters in vitro.
Metabolic products were separated by TLC by using Brinkman 0. the TLC plates, derivatized with diazomethane, and identified by GLC-MS. The spectra of resultant dimethyl phthalate, methyl n-butyl phthalate, and methyl 2-ethylhexyl phthalate were identical to those of the corresponding synthetic esters. Graphical presentation of mass fragments in sequential GLC-MS scans using the PDP-12 LDP computer was extremely useful in examining the metabolites and parent compounds. Routine analysis of phthalate esters and metabolic products required multiple cleanup techniques prior to GLC analysis. Diesters were recovered by conventional residue extraction techniques. Chromatography of these esters of Florisil was analagous to chromatography of dieldrin. Glucuronide conjugates of the phthalic acid monoesters must be enzymatically hydrolyzed and recovered prior to derivatization and GLC analysis.
Recovery of the glucuronides from tissue (20 g) was achieved by column extraction using 200 ml of 1% H3PO4 in acetone. The extract concentrated to 20 ml, was dissolved in 200 ml of water and the pH was adjusted to pH 4.5 and buffered with acetate buffer. The glucuronide was hydrolyzed by incubation with Ketodase (5 ml) for 24-48 hr at 37°C. The pH was lowered to 1.5 with H3PO4, and the monoester and phthalic acid extracted with diethyl ether. After evaporation, the extract was dried with Na2 SO4 (anhydrous) and derivatized with diazomethane prior to GLC analysis.
Derivatization of the PAE monoesters with diazomethane was satisfactory. Dimethyl phthalate, however, has an extremely short GLC retention time, and di-2chloroethyl phthalate was a more suitable derivative for phthalic acid. The electron capture detector was sensitive to the phthalate esters and detector response was approximately 20% of the response for DDE.
Additional investigations have been undertaken to establish which derivatives of the phthalate ester metabolites are suitable for gas chromatographic analysis. The methyl and n-butyl 3-0-trifluoroacetyl phthalates and trimethylsilyl derivative of 3hydroxyphthalatic acid were examined by GC-MS. Only the trimethylsilyl derivative gave satisfactory chromatographic peaks. The mass spectrum of the TMS derivative Environmental Health Perspectives 295:14; 282:6; 398:2. The TMS derivative eluted at 2200 C by use of a 10% SE-30 column programmed from 120 to 2400C at 6.50C/min with ca. 20 ml/min of a He carrier flow.