TPH Measurements
The Advantage of Using GC/MS
Robert Haddad, Ph.D.
ENTRIX, Inc.
John MacMurphey
ZymaX envirotechnology, inc
Abstract
This study provides analytical data supporting the use of GC/MS to quantify TPH in environmental samples. Analysis of these data shows that there is no statistical difference between TPH values quantified using GC/MS and those derived from conventional TPH methods using GC/FID. Further, the GC/MS analysis used in this study is readily adaptable to most environmental laboratories currently performing the volatile and semi-volatile EPA analyses (i.e., 8260 and 8270, respectively; EPA SW 846).
The strategic implications of this result are: (1) additional information (e.g., PAHs) often required when conducting a risk based assessment can be derived from pre-existing site assessment data, thereby decreasing the cost and time required to obtain the additional samples and analytical information; and (2) unique TPH distributions can be critically evaluated using mass spectral analysis in order to ascertain the nature and potential source of the compounds present in the TPH mixture. The later point can be crucial when specific sources and their distributions are unknown. The benefits of having MS data available on a routine basis seem to increase as a function of the size and the public profile of the project.
Introduction
As more attention is given to managing environmental liability using risk-based approaches (e.g., ASTM-RBCA framework), it is clear that there is a need to quantify more than just total petroleum hydrocarbons (TPH). For example, most risk-based approaches for petroleum contaminated sites require a quantitative understanding of the levels and distribution of the polycyclic aromatic hydrocarbon (PAH) concentrations present in site media as well as the TPH concentrations. Current practice is to use chromatographic methods with flame ionization detection (GC/FID; e.g., EPA 8015m, AK102, etc.) to screen for TPH and then use this information to decide which samples need additional higher resolution analyses (e.g., gas chromatography/mass spectrometry [GC/MS]; EPA 8260, EPA 8270, etc.). Depending on the site and the site's history, this can mean that the sample extract is re-analyzed using GC/MS methods or, for older sites, that the site has to be re-sampled and the re-acquired site media re-submitted to the laboratory for additional analyses.
Additionally, results of EPA 8015 analyses can and do produce hydrocarbon distributions that are noted as being "unlike the calibration standard". Because the current analytical methods do not commonly involve an extract clean-up step (e.g., silica gel), these results can be due to many different sources and may not be related to petroleum hydrocarbons at all. In these cases, GC/FID results are of limited use in determining the specific nature of the organic material with the unusual hydrocarbon distribution.
In order to decrease the analytical costs for larger sites and to increase our clients' future options, we have developed a method certified by the California Department of Health which allows us to quantify TPH using GC/MS analysis. Further, the availability of the full scan GC/MS data allows one to perform a mass spectral analysis and to evaluate unusual hydrocarbon distributions in terms of unknown and tentatively identified compounds. The availability of the GC/MS data coupled with a competent mass spectral analysis can often provide enough information to discern the nature of the organic material in the sample.
Despite higher initial analytical costs, the benefits of this approach are discussed in terms of case studies in which closure of a site using a risk-based approach was undertaken several years following an extensive site assessment. The ability to re-analyze the GC/MS data collected during the initial assessment, provide and support substantive chemical arguments for the presence of naturally occurring crude oil and biogenic hydrocarbons, and provide PAH concentrations saved the client considerable time and expense in needless re-sampling, extraction, and analytical analysis.
Theoretical Concerns
Flame Ionization Detectors
The difference in using FID versus MS as a means of quantification is not a trivial issue. The basic principle of the FID is quite simple (see Grob, 1995). Column effluent (carrier gas and organic compounds) exiting the gas chromatograph is burned in a small hydrogen flame. Ionic fragments and free electrons formed in the combustion are collected via an electrostatic field surrounding the flame, and an electronic current proportional to the amount of sample entering the flame is created. Because the FID responds only to oxidizable carbon atoms, fully oxidized carbon compounds (e.g., CO2) and non-oxidizable compounds (e.g., polysulfides or S8) are not detected. Furthermore, the fact that the compounds are detected following the flame ionization means that while detector response is proportional to the amount of oxidizable material that was ionized, it is not impacted by the chemical nature of the compound itself. Thus, while the FID response can vary as a function of the amount of material being ionized in the flame, it does not vary significantly as a function of the chemistry of the compound; a C6 alkane (hexane) will exhibit an FID detector response similar to that of a C6 aromatic (benzene). Additionally, FID response has been shown to be linear over nearly 6 orders of magnitude differences in compound concentration. For these and other reasons (high sensitivity, great reliability, etc.) the FID is considered universal detector for oxidizable carbon compounds.
Mass Spectrometers
In marked contrast to this, detection using a mass spectrometer can be significantly affected by both the amount and the chemical nature of the compound (see McCloskey, 1990). This is due, in part to the ionization potential of the specific compound. It is this attribute which makes the MS such a useful tool in characterizing specific compounds.
Because of the significant differences in detector response as a function of the chemistry, quantification using mass spectrometry is usually done only when there is a specific standard for the compound of interest. Using GC/MS to quantify mixtures can produce inaccurate results. For example, consider the detection and bulk quantification of a benzene and hexane mixture. Each compound has 6 carbons, however, the aromaticity of the benzene molecule will cause a greater response from the MS relative to the non-aromatic (aliphatic) hexane molecules. Using a calibration curve based on a 50% mixture one can determine a response factor for the whole mixture. In reality, the whole mixture response factor calculated during the calibration is a sum of the detector response factors for benzene and for hexane. The next step would be to apply this whole mixture response factor to an unknown mixture and then quantify the concentration of this unknown mixture.
This approach is a potential source of concern in using the MS to quantify the mixture. If the unknown mixture has the same percentages of benzene and hexane as is in the standard, then the MS quantification will be accurate. If, however, the unknown has a benzene percentage > 50%, then the greater response to aromatics seen in the MS can theoretically give rise to an overestimation of the whole mixture concentration. Simply put, the MS calibration will assume a 50% mixture and will "see" the enhanced response (due to the higher benzene content) as meaning a larger concentration of the entire 50% mixture. As noted above, this problem is avoided with the FID due to the relatively small difference between the benzene and hexane response factors.
The magnitude of this problem is currently unknown for environmental samples containing petroleum contamination. However, due to the potential benefits of analyzing TPH mixtures using GC/MS, a study was designed to test the hypothesis that the TPH values derived from GC/MS analysis would be significantly different from those derived from GC/FID analysis of the same sample.
Analytical Methods and Study Results
Method
The soil samples in this study were extracted by EPA method 3550. This method uses sonication to extract petroleum hydrocarbons from solid samples. The extracts were analyzed by both GC/MS and GC/FID.
Both methods ran a five point calibration curve using diesel #2 as a standard. The average response factor for the five calibration standards was calculated following the procedures in EPA method 8270. Samples were quantified against the calibration curve using the equations presented in EPA method 8270. The area used in the calculation for both samples and standard was the total area present. This procedure was followed for both GC/FID and GC/MS. For GC/MS the area used for calculations was from the total ion current (TIC), which was generated in the full scan acquisition mode.
Results were reported as TPH in mg/kg.
Study Results
Soil samples from several environmental sites containing a variety of petroleum contaminants (e.g., diesel, crude oil, motor oil, etc.) were used in this study. The samples were extracted as described above and aliquots of the extracts were analyzed using both GC/FID and GC/MS. Analysis of the Quality Assurance/Quality Control data (internal standards, surrogate recoveries, matrix spikes, matrix spike duplicates) all documented that the data were valid. The GC/FID and GC/MS TPH nC10-nC40 results for the 28 samples which had TPH detections greater than 10 mg/kg (laboratory practical detection limit) are presented in Table 1. Also presented are the differences between the paired measurements and an indication (in terms of percent of the GC/FID result) of how different the GC/MS and the GC/FID results are. This last column in Table 1 shows that the GC/MS results range from between -15.7% to +10.4% of the paired GC/FID results. A log-normal transformation of the data showed the distribution to be normal. Using a paired two-tail t-test, no statistical difference (at P<0.05) was observed between the GC/FID and the GC/MS data sets.
These data and their paired distributions are illustrated in Figure 1 which compares: (1) the theoretical GC/MS TPH values expected if there were a 1:1 relationship between the GC/FID and GC/MS TPH values; (2) the +10% curves for the theoretical 1:1 GC/FID:GC/MS relationship (a measure of the method precision); and (3) the measured GC/MS TPH values. An important point made by this visualization is that there is no apparent bias in the GC/MS TPH results (relative to those from the GC/FID analyses).
Discussion
The data presented above indicate that for soils contaminated with a variety of petroleum compounds/mixtures, the use of GC/MS to quantify TPH is statistically valid. In terms of the discussion regarding the potential problems associated with use of the mass spectrometer to quantify mixtures, these data argue that differences due to these theoretical concerns (at least in this data set) are less than or equal to the actual precision of the GC/FID analysis. Further, Thus, these data indicate that there is no analytical bias as was hypothesized on the basis of MS response to weathered samples. As pointed out in the introduction, this knowledge provides lower cost analytical options for clients beginning to assess larger contaminated sites.
Strategic Implications - Future Data Needs
The strategic implications of this work are at least two fold. First, they provide the project manager with a way to initially gather data that can be used in different ways as the assessment process proceeds. In one study recently conducted, a site had been assessed for hydrocarbon contamination several years ago. At the time, approximately 150 soil samples were analyzed for TPH in the nC10-nC40 range. The TPH concentrations ranged from below detection (<10mg/kg) to >10,000 mg/kg. Following the initial assessment work, the client choose to conduct a human health risk assessment. During the initial data gaps analysis, it was noted that although polycyclic aromatic hydrocarbons (PAH) were chemicals of potential concern (COPCs), these compounds had not been measured during the earlier assessment phase of the work.
Had the assessment TPH work been conducted using the conventional EPA 8015m GC/FID analyses, the project manager would have had to re-mobilize a drill rig and contractors in order to re-obtain samples from areas identified as potential hot spots. The PM would then need to have these samples extracted and analyzed following one of the EPA methods for quantifying PAHs (e.g., EPA 8270). Assuming that only 10 percent of the samples required PAH quantification, this could have required 1-2 additional field drilling days, additional laboratory time, and the associated costs.
Because the assessment was conducted using the GC/MS TPH analyses, the laboratory had the full scan GC/MS data on tape. A list of the samples requiring PAH analysis was developed and sent to the laboratory, the laboratory de-archived the data tapes, and re-analyzed the data for the appropriate PAHs . The client was charged for de-archiving the tape, the re-analysis and production of an EPA 8270 analytical data report. This cost was substantially less than the amount that would have been required for mobilizing a drill rig and complete laboratory analysis of a new soil sample. The PAH data was then directly incorporated into the risk assessment as paired TPH/PAH samples.
Strategic Implications - Evaluation of Unique TPH Distributions
The second strategic implication of this work is that the availability of full scan mass spectral data allows for a mass spectral evaluation of the sample in terms of the specific compounds that are present in the mixture. The GC/MS total ion current (TIC) fingerprint (analogous to the GC trace from a GC/FID analysis) presented in Figure 2 is from a site where a potential for petroleum-derived contamination existed; however, to date, no substantial petroleum contamination had been found in this part of the site. The TPH result from this sample, 170 ug/L, was above the action level of 100 ug/L and was used as basis by the agencies to require additional assessment for petroleum derived hydrocarbons.
Prior to conducting additional field assessment, a mass spectral analysis was performed on the full scan MS data. This analysis concluded that (1) the bulk of the material present in the sample was not attributable to petroleum-derived hydrocarbon contamination; in fact, many of the resolved peaks could be tentatively identified as sulfur and nitrogen-rich mono aromatic compounds. Analysis of samples collected on the same day by the same drilling crew showed similar distributions of nitrogen and sulfur-rich aromatic compounds. Further testing proved that these compounds were derived from drill rig contamination during the field assessment. This conclusion was accepted by the agencies following their review of the mass spectral analysis and the client was able to close part of the site to further assessment activity.
Conclusions
The study presented above provides evidence that use of GC/MS to quantify TPH in environmental samples (at least in soil samples) provides results that are statistically similar to those derived from conventional TPH methods using GC/FID. The analysis is readily adaptable to most environmental laboratories currently performing the volatile and semi-volatile EPA analyses (i.e., 8260 and 8270, respectively).
The strategic implications of this result are: (1) additional information (e.g., PAHs) often required when conducting a risk based assessment can be derived from pre-existing assessment data, thereby decreasing the cost and time required to obtain the additional samples and analytical information; and (2) unique TPH distributions can be critically evaluated using mass spectral analysis in order to ascertain the nature and potential source of the compounds present in the TPH mixture. The benefits of having MS data available on a routine basis seem to increase as a function of the size and the public profile of the project.
References
Environmental Protection Agency, SW-846, 1995
Grob, R.L. (1995) Modern practice of gas chromatography, 3rd Edition. Wiley, New York, 888 pp.
McCloskey, J.A. (1990) Mass Spectrometry. Academic Press, San Diego, 960 pp.
Robert Haddad is a Senior Consultant with ENTRIX, Inc. where he applies his technical expertise in petroleum geochemistry to evaluating and allocating environmental liability. He holds a B.S. in geology and a Ph.D. in organic geochemistry. ENTRIX, Inc., 590 Ygnacio Valley Rd., Suite 200, Walnut Creek, CA 94596, Voice (510) 935-9920, FAX (510) 935-5368, E-MAIL rhaddad@entrix.com
John MacMurphey is Laboratory Director and partner of ZymaX envirotechnology, inc., an analytical laboratory specializing in organic analysis by GC/MS and metals by ICP/MS. He holds a B.S. in chemistry. ZymaX envirotechnology, inc., 71 Zaca Lane, Suite 110, San Luis Obispo, CA 93401, Voice (805) 544 4696, FAX (805) 544 8226, E-MAIL john@ZymaXusa.com
Robert Haddad, Ph.D.
ENTRIX, Inc.
John MacMurphey
ZymaX envirotechnology, inc
Abstract
This study provides analytical data supporting the use of GC/MS to quantify TPH in environmental samples. Analysis of these data shows that there is no statistical difference between TPH values quantified using GC/MS and those derived from conventional TPH methods using GC/FID. Further, the GC/MS analysis used in this study is readily adaptable to most environmental laboratories currently performing the volatile and semi-volatile EPA analyses (i.e., 8260 and 8270, respectively; EPA SW 846).
The strategic implications of this result are: (1) additional information (e.g., PAHs) often required when conducting a risk based assessment can be derived from pre-existing site assessment data, thereby decreasing the cost and time required to obtain the additional samples and analytical information; and (2) unique TPH distributions can be critically evaluated using mass spectral analysis in order to ascertain the nature and potential source of the compounds present in the TPH mixture. The later point can be crucial when specific sources and their distributions are unknown. The benefits of having MS data available on a routine basis seem to increase as a function of the size and the public profile of the project.
Introduction
As more attention is given to managing environmental liability using risk-based approaches (e.g., ASTM-RBCA framework), it is clear that there is a need to quantify more than just total petroleum hydrocarbons (TPH). For example, most risk-based approaches for petroleum contaminated sites require a quantitative understanding of the levels and distribution of the polycyclic aromatic hydrocarbon (PAH) concentrations present in site media as well as the TPH concentrations. Current practice is to use chromatographic methods with flame ionization detection (GC/FID; e.g., EPA 8015m, AK102, etc.) to screen for TPH and then use this information to decide which samples need additional higher resolution analyses (e.g., gas chromatography/mass spectrometry [GC/MS]; EPA 8260, EPA 8270, etc.). Depending on the site and the site's history, this can mean that the sample extract is re-analyzed using GC/MS methods or, for older sites, that the site has to be re-sampled and the re-acquired site media re-submitted to the laboratory for additional analyses.
Additionally, results of EPA 8015 analyses can and do produce hydrocarbon distributions that are noted as being "unlike the calibration standard". Because the current analytical methods do not commonly involve an extract clean-up step (e.g., silica gel), these results can be due to many different sources and may not be related to petroleum hydrocarbons at all. In these cases, GC/FID results are of limited use in determining the specific nature of the organic material with the unusual hydrocarbon distribution.
In order to decrease the analytical costs for larger sites and to increase our clients' future options, we have developed a method certified by the California Department of Health which allows us to quantify TPH using GC/MS analysis. Further, the availability of the full scan GC/MS data allows one to perform a mass spectral analysis and to evaluate unusual hydrocarbon distributions in terms of unknown and tentatively identified compounds. The availability of the GC/MS data coupled with a competent mass spectral analysis can often provide enough information to discern the nature of the organic material in the sample.
Despite higher initial analytical costs, the benefits of this approach are discussed in terms of case studies in which closure of a site using a risk-based approach was undertaken several years following an extensive site assessment. The ability to re-analyze the GC/MS data collected during the initial assessment, provide and support substantive chemical arguments for the presence of naturally occurring crude oil and biogenic hydrocarbons, and provide PAH concentrations saved the client considerable time and expense in needless re-sampling, extraction, and analytical analysis.
Theoretical Concerns
Flame Ionization Detectors
The difference in using FID versus MS as a means of quantification is not a trivial issue. The basic principle of the FID is quite simple (see Grob, 1995). Column effluent (carrier gas and organic compounds) exiting the gas chromatograph is burned in a small hydrogen flame. Ionic fragments and free electrons formed in the combustion are collected via an electrostatic field surrounding the flame, and an electronic current proportional to the amount of sample entering the flame is created. Because the FID responds only to oxidizable carbon atoms, fully oxidized carbon compounds (e.g., CO2) and non-oxidizable compounds (e.g., polysulfides or S8) are not detected. Furthermore, the fact that the compounds are detected following the flame ionization means that while detector response is proportional to the amount of oxidizable material that was ionized, it is not impacted by the chemical nature of the compound itself. Thus, while the FID response can vary as a function of the amount of material being ionized in the flame, it does not vary significantly as a function of the chemistry of the compound; a C6 alkane (hexane) will exhibit an FID detector response similar to that of a C6 aromatic (benzene). Additionally, FID response has been shown to be linear over nearly 6 orders of magnitude differences in compound concentration. For these and other reasons (high sensitivity, great reliability, etc.) the FID is considered universal detector for oxidizable carbon compounds.
Mass Spectrometers
In marked contrast to this, detection using a mass spectrometer can be significantly affected by both the amount and the chemical nature of the compound (see McCloskey, 1990). This is due, in part to the ionization potential of the specific compound. It is this attribute which makes the MS such a useful tool in characterizing specific compounds.
Because of the significant differences in detector response as a function of the chemistry, quantification using mass spectrometry is usually done only when there is a specific standard for the compound of interest. Using GC/MS to quantify mixtures can produce inaccurate results. For example, consider the detection and bulk quantification of a benzene and hexane mixture. Each compound has 6 carbons, however, the aromaticity of the benzene molecule will cause a greater response from the MS relative to the non-aromatic (aliphatic) hexane molecules. Using a calibration curve based on a 50% mixture one can determine a response factor for the whole mixture. In reality, the whole mixture response factor calculated during the calibration is a sum of the detector response factors for benzene and for hexane. The next step would be to apply this whole mixture response factor to an unknown mixture and then quantify the concentration of this unknown mixture.
This approach is a potential source of concern in using the MS to quantify the mixture. If the unknown mixture has the same percentages of benzene and hexane as is in the standard, then the MS quantification will be accurate. If, however, the unknown has a benzene percentage > 50%, then the greater response to aromatics seen in the MS can theoretically give rise to an overestimation of the whole mixture concentration. Simply put, the MS calibration will assume a 50% mixture and will "see" the enhanced response (due to the higher benzene content) as meaning a larger concentration of the entire 50% mixture. As noted above, this problem is avoided with the FID due to the relatively small difference between the benzene and hexane response factors.
The magnitude of this problem is currently unknown for environmental samples containing petroleum contamination. However, due to the potential benefits of analyzing TPH mixtures using GC/MS, a study was designed to test the hypothesis that the TPH values derived from GC/MS analysis would be significantly different from those derived from GC/FID analysis of the same sample.
Analytical Methods and Study Results
Method
The soil samples in this study were extracted by EPA method 3550. This method uses sonication to extract petroleum hydrocarbons from solid samples. The extracts were analyzed by both GC/MS and GC/FID.
Both methods ran a five point calibration curve using diesel #2 as a standard. The average response factor for the five calibration standards was calculated following the procedures in EPA method 8270. Samples were quantified against the calibration curve using the equations presented in EPA method 8270. The area used in the calculation for both samples and standard was the total area present. This procedure was followed for both GC/FID and GC/MS. For GC/MS the area used for calculations was from the total ion current (TIC), which was generated in the full scan acquisition mode.
Results were reported as TPH in mg/kg.
Study Results
Soil samples from several environmental sites containing a variety of petroleum contaminants (e.g., diesel, crude oil, motor oil, etc.) were used in this study. The samples were extracted as described above and aliquots of the extracts were analyzed using both GC/FID and GC/MS. Analysis of the Quality Assurance/Quality Control data (internal standards, surrogate recoveries, matrix spikes, matrix spike duplicates) all documented that the data were valid. The GC/FID and GC/MS TPH nC10-nC40 results for the 28 samples which had TPH detections greater than 10 mg/kg (laboratory practical detection limit) are presented in Table 1. Also presented are the differences between the paired measurements and an indication (in terms of percent of the GC/FID result) of how different the GC/MS and the GC/FID results are. This last column in Table 1 shows that the GC/MS results range from between -15.7% to +10.4% of the paired GC/FID results. A log-normal transformation of the data showed the distribution to be normal. Using a paired two-tail t-test, no statistical difference (at P<0.05) was observed between the GC/FID and the GC/MS data sets.
These data and their paired distributions are illustrated in Figure 1 which compares: (1) the theoretical GC/MS TPH values expected if there were a 1:1 relationship between the GC/FID and GC/MS TPH values; (2) the +10% curves for the theoretical 1:1 GC/FID:GC/MS relationship (a measure of the method precision); and (3) the measured GC/MS TPH values. An important point made by this visualization is that there is no apparent bias in the GC/MS TPH results (relative to those from the GC/FID analyses).
Discussion
The data presented above indicate that for soils contaminated with a variety of petroleum compounds/mixtures, the use of GC/MS to quantify TPH is statistically valid. In terms of the discussion regarding the potential problems associated with use of the mass spectrometer to quantify mixtures, these data argue that differences due to these theoretical concerns (at least in this data set) are less than or equal to the actual precision of the GC/FID analysis. Further, Thus, these data indicate that there is no analytical bias as was hypothesized on the basis of MS response to weathered samples. As pointed out in the introduction, this knowledge provides lower cost analytical options for clients beginning to assess larger contaminated sites.
Strategic Implications - Future Data Needs
The strategic implications of this work are at least two fold. First, they provide the project manager with a way to initially gather data that can be used in different ways as the assessment process proceeds. In one study recently conducted, a site had been assessed for hydrocarbon contamination several years ago. At the time, approximately 150 soil samples were analyzed for TPH in the nC10-nC40 range. The TPH concentrations ranged from below detection (<10mg/kg) to >10,000 mg/kg. Following the initial assessment work, the client choose to conduct a human health risk assessment. During the initial data gaps analysis, it was noted that although polycyclic aromatic hydrocarbons (PAH) were chemicals of potential concern (COPCs), these compounds had not been measured during the earlier assessment phase of the work.
Had the assessment TPH work been conducted using the conventional EPA 8015m GC/FID analyses, the project manager would have had to re-mobilize a drill rig and contractors in order to re-obtain samples from areas identified as potential hot spots. The PM would then need to have these samples extracted and analyzed following one of the EPA methods for quantifying PAHs (e.g., EPA 8270). Assuming that only 10 percent of the samples required PAH quantification, this could have required 1-2 additional field drilling days, additional laboratory time, and the associated costs.
Because the assessment was conducted using the GC/MS TPH analyses, the laboratory had the full scan GC/MS data on tape. A list of the samples requiring PAH analysis was developed and sent to the laboratory, the laboratory de-archived the data tapes, and re-analyzed the data for the appropriate PAHs . The client was charged for de-archiving the tape, the re-analysis and production of an EPA 8270 analytical data report. This cost was substantially less than the amount that would have been required for mobilizing a drill rig and complete laboratory analysis of a new soil sample. The PAH data was then directly incorporated into the risk assessment as paired TPH/PAH samples.
Strategic Implications - Evaluation of Unique TPH Distributions
The second strategic implication of this work is that the availability of full scan mass spectral data allows for a mass spectral evaluation of the sample in terms of the specific compounds that are present in the mixture. The GC/MS total ion current (TIC) fingerprint (analogous to the GC trace from a GC/FID analysis) presented in Figure 2 is from a site where a potential for petroleum-derived contamination existed; however, to date, no substantial petroleum contamination had been found in this part of the site. The TPH result from this sample, 170 ug/L, was above the action level of 100 ug/L and was used as basis by the agencies to require additional assessment for petroleum derived hydrocarbons.
Prior to conducting additional field assessment, a mass spectral analysis was performed on the full scan MS data. This analysis concluded that (1) the bulk of the material present in the sample was not attributable to petroleum-derived hydrocarbon contamination; in fact, many of the resolved peaks could be tentatively identified as sulfur and nitrogen-rich mono aromatic compounds. Analysis of samples collected on the same day by the same drilling crew showed similar distributions of nitrogen and sulfur-rich aromatic compounds. Further testing proved that these compounds were derived from drill rig contamination during the field assessment. This conclusion was accepted by the agencies following their review of the mass spectral analysis and the client was able to close part of the site to further assessment activity.
Conclusions
The study presented above provides evidence that use of GC/MS to quantify TPH in environmental samples (at least in soil samples) provides results that are statistically similar to those derived from conventional TPH methods using GC/FID. The analysis is readily adaptable to most environmental laboratories currently performing the volatile and semi-volatile EPA analyses (i.e., 8260 and 8270, respectively).
The strategic implications of this result are: (1) additional information (e.g., PAHs) often required when conducting a risk based assessment can be derived from pre-existing assessment data, thereby decreasing the cost and time required to obtain the additional samples and analytical information; and (2) unique TPH distributions can be critically evaluated using mass spectral analysis in order to ascertain the nature and potential source of the compounds present in the TPH mixture. The benefits of having MS data available on a routine basis seem to increase as a function of the size and the public profile of the project.
References
Environmental Protection Agency, SW-846, 1995
Grob, R.L. (1995) Modern practice of gas chromatography, 3rd Edition. Wiley, New York, 888 pp.
McCloskey, J.A. (1990) Mass Spectrometry. Academic Press, San Diego, 960 pp.
Robert Haddad is a Senior Consultant with ENTRIX, Inc. where he applies his technical expertise in petroleum geochemistry to evaluating and allocating environmental liability. He holds a B.S. in geology and a Ph.D. in organic geochemistry. ENTRIX, Inc., 590 Ygnacio Valley Rd., Suite 200, Walnut Creek, CA 94596, Voice (510) 935-9920, FAX (510) 935-5368, E-MAIL rhaddad@entrix.com
John MacMurphey is Laboratory Director and partner of ZymaX envirotechnology, inc., an analytical laboratory specializing in organic analysis by GC/MS and metals by ICP/MS. He holds a B.S. in chemistry. ZymaX envirotechnology, inc., 71 Zaca Lane, Suite 110, San Luis Obispo, CA 93401, Voice (805) 544 4696, FAX (805) 544 8226, E-MAIL john@ZymaXusa.com