Optimized GC/MS Analysis for PAHs in Challenging Matrices

Abstract

The Agilent 8890 GC combined with an Agilent 5977 Series MSD system was used for the analysis of polycyclic aromatic hydrocarbons (PAHs). By proper selection of instrument configuration and operating conditions, the system provides a robust means of analyzing PAHs in difficult matrices. Midcolumn backflushing, continuous hydrogen source cleaning (JetClean), and use of an alternative drawout lens result in excellent linearity across a calibration range of 1 to 1,000 pg. System precision and robustness are demonstrated with replicate injections of an extract from high organic content soil.

Introduction

PAHs are toxic to aquatic life and are suspected human carcinogens. Because they originate from multiple sources, they are widely distributed as contaminants throughout the world.

PAH Sources:

• Petrogenic: Derived from petroleum inputs associated with fossil fuels
• Pyrogenic: Derived from combustion sources
• Biogenic: Formed from natural biological processes

Given their ubiquitous nature, they are monitored as trace contaminants in many different food products ranging from seafood to edible oils to smoked meats. They are also monitored in the environment including air, water, and soil. PAHs have been analyzed by multiple techniques including HPLC/UV, GC/FID, GC/MS, or GC/MS/MS.

This Application Note focuses on GC/MS in SIM mode. A common calibration range is from 1 to 1,000 pg with an acceptable linearity of R² >0.99. Internal standard (ISTD) area reproducibility is typically specified at ±20 % with calibration standards, and ±30 % with samples.

A number of issues arise with the analysis due to the properties of PAHs. They span wide molecular weight and boiling temperature ranges. Although not considered active or subject to degradation, they are sticky and readily adhere to surfaces. PAHs are subject to desublimation (deposition) and are difficult to vaporize. High temperatures and minimizing surface contact are important. Peak tailing is often seen on the later eluters, resulting in manual integration and extending data review. In some cases, the ISTD response is inconsistent across the calibration range and can lead to problems with linearity of the method.

In addition to the PAH-related challenges, there are often matrix-related problems with the analysis. For example, in food and soil analyses, high boiling matrix contaminants that elute after the analytes can require extended bakeout times to prevent ghost peaks in subsequent runs. The highest boiling contaminants can deposit in the head of the column, requiring more frequent column trimming and adjustment of SIM and data analysis time windows from the resulting retention time shift.

Experimental

This system was configured to minimize the potential problems with the analysis of PAHs in high-matrix samples. The important techniques used were:

• Agilent JetClean: This option on the 5977 Series GC/MSD system provides a low continuous flow of hydrogen (0.33 mL/min) into the source during the analysis. Continuous cleaning of the source with hydrogen has been demonstrated¹⁻³ to significantly improve calibration linearity and precision of response over time for PAH analysis. The need for manual source cleaning, especially with high-matrix samples, is substantially reduced.
• 9 mm extractor lens: The Agilent extractor source provides additional flexibility to meet the specific needs of different analytical challenges. For the analysis of PAHs, a 9 mm extraction lens provides a good choice to minimize the surfaces available for deposition of the PAHs, and contributes, with JetClean, to providing better linearity, precision, and peak shapes.
• Midcolumn backflushing: Backflushing is a technique where the carrier gas flow is reversed after the last analyte has exited the column. After the MS data are collected, the oven is held at the final temperature in post run mode, and the carrier gas flow through the first column is reversed. This reversed flow carries any high boilers that were in the column at the end of data collection out of the head of the column and into the split vent trap. The capability to reverse the flow is provided by the Agilent Purged Ultimate Union (PUU). The PUU is a tee inserted, in this case, between two identical 15 m columns. During the analysis, a small makeup flow of carrier gas from the 8890 pneumatic switching device (PSD) module is used to sweep the connection. During backflushing, the makeup flow from the PSD is raised to a much higher value, sweeping high boilers backwards out of the first of column and forwards from the second. For this configuration, the backflushing time was 1.5 minutes.
• 8890 PSD module: The PSD is an 8890 pneumatics module optimized for backflushing applications. During backflushing, it significantly reduces the flow of helium used compared to previous configurations. The PSD provides for seamless pulsed injections and simpler setup of backflush.

System Configuration

Figure 1 shows the system configuration used. Tables 1 and 2 list the instrument operating parameters. Instrument temperatures must be kept high enough to prevent deposition of the highest boiling PAHs. The inlet and MSD transfer line are maintained at 320 °C. The MS source should be a minimum of 320 °C. Pulsed splitless injections are used to maximize transfer of the PAHs, especially the heavy ones, into the column. The straight bore 4 mm liner with glass wool is a must. The wool transfers heat to the PAHs and blocks the line of sight to the inlet base. If the PAHs condense on the inlet base, they are difficult to vaporize, and sweep back into the column.

Figure 1. System configuration. [Diagram description: Shows the Agilent 8890 GC connected to a 5977 Series GC/MSD. The 8890 GC includes a Liquid Injector and a PSD (Helium) module. The GC column setup involves a 9-mm Extractor lens, an EI Source, and two 15 m DB-EUPAH columns connected via a Purged Ultimate Union (PUU) with a makeup flow from the PSD module. The 5977 Series GC/MSD has JetClean (Hydrogen) connected to its source.]

Table 1. GC and MS conditions for the PAH analysis.

Table 2. SIM ions used for quantifier and qualifiers.

Results and discussion

Initial calibration

Figure 2 shows the SIM TIC of the 100 pg/µL calibration standard. With the parameters chosen, the peak shapes for all PAHs, especially the latest ones, are very good.

The use of the 9 mm lens and continuous hydrogen cleaning often results in a reduced signal-to-noise ratio (S/N), so it is important to check the lowest desired calibration level. As an example, Figure 3 shows the response at the quantifier ion for several of the compounds at the 1 pg level. All analytes at the 1 pg level had sufficient signal for calibration.

Figure 2. SIM TIC of the 100 pg/µL calibration standard. [Description: A chromatogram showing the Total Ion Chromatogram (TIC) in Selected Ion Monitoring (SIM) mode for a 100 pg/µL calibration standard, with numbered peaks corresponding to identified PAHs and ISTDs. The x-axis represents acquisition time in minutes, ranging from approximately 5 to 28 minutes.]

Figure 3. Response at quantifier ion for select compounds in the lowest calibration standard (1 pg). [Description: Three separate, zoomed-in chromatogram plots. The first shows the response for Fluorene (SIM 166) between 7.75 and 8.65 minutes. The second shows responses for Fluoranthene and Pyrene (SIM 202) between 12.3 and 14.7 minutes. The third shows the response for Benzo[ghi]perylene (SIM 276) between 24.8 and 27.5 minutes.]

Stability of response

Table 3 shows the R² values for three ISTD calibrations of the system with seven levels from 1 to 1,000 pg. All analytes show excellent linearity across the entire range.

Figure 4 shows the precision of ISTD peak responses for 60 sequential replicate injections of the 100 pg standard. The RSDs of the ISTD areas were: Naphthalene-d8 (3.3 %), Acenaphthene-d10 (3.2 %), Phenanthrene-d10 (3.4 %), Chrysene-d12 (2.7 %), Perylene-d12 (2.0 %).

Figure 5 shows the calculated concentration for several analytes in the 60 sequential replicate runs of the 100 pg standard. The system exhibits excellent stability of response. The average RSD of the calculated concentrations for all 27 analytes is 1.1 %.

Table 3. R² values of three seven-level ISTD calibrations, 1 to 1,000 pg SIM. [Table content listing compounds, RT, and R² values for three calibration sets.]

Figure 4. ISTD response stability over 60 injections for a 100 pg calibration standard. Areas are normalized to that of the first injection. [Description: A line graph showing the normalized ISTD response (percentage) over 60 injection numbers for five deuterated PAH standards: Naphthalene-d8, Acenaphthene-d10, Phenanthrene-d10, Chrysene-d12, and Perylene-d12. The y-axis ranges from 50% to 120%.]

Figure 5. Stability of calculated concentrations over 60 sequential injections for a 100 pg calibration standard. [Description: A line graph showing the calculated concentration (normalized to 100%) over 60 injection numbers for Naphthalene, Acenaphthylene, Benzo[k]fluoranthene, and Benzo[ghi]perylene. The y-axis ranges from 60% to 120%.]

Stability of response with soil extracts

The soil extract used for the robustness test was deliberately chosen to have a high-matrix content to challenge the system. Figure 6 compares the scan TIC of the extract to that of the 100 pg PAH standard. The soil extract has a very high level of matrix. Note that, for soils with this level of organic content, further sample cleanup should be considered for routine analysis. The sample preparation used was for test purposes only.

To test the robustness of the system, the soil extract was spiked with 100 pg each of the 27 analytes and 500 pg each of the ISTDs. The spiked extract was then injected 60 times. The PAHs were quantitated against the solvent-based calibration curve for each run, and the resulting calculated concentrations were plotted. Figure 7 shows the calculated concentrations for several of the analytes. Naphthalene and benzo[ghi]perylene both show measured concentrations higher than the spiked 100 pg level. These compounds were found to be present in the soil at levels roughly corresponding to the offset in Figure 7. Perylene (not shown) was found at almost 200 pg in the soil.

The average RSD for the calculated concentrations of all 27 analytes was 4.4 %. For 22 of the 27 analytes, the calculated concentration was within 20 % after 60 soil shots, compared to the first injection in the soil. As expected, the heaviest analytes, such as benzo[ghi]perylene, lost response quickest.

Figure 6. Scan TIC of soil extract and PAH 100 pg standard with 500 pg ISTDs, both drawn in the same scale, showing a large amount of material in the extract. [Description: A chromatogram comparing the Total Ion Chromatogram (TIC) of a heavy soil matrix extract against a PAH standard, illustrating the significant matrix interference. The x-axis represents acquisition time in minutes, ranging from approximately 5 to 28 minutes.]

Figure 7. Stability of calculated concentrations over 60 injections of a soil matrix spiked with 100 pg PAH standards and 500 pg ISTDs. [Description: A line graph showing the calculated concentration (normalized to 100% or higher) over 60 injection numbers for Naphthalene, Acenaphthylene, Benzo[k]fluoranthene, and Benzo[ghi]perylene when analyzing a spiked soil matrix. The y-axis ranges from 40% to 140%.]

Post-Maintenance Calibration Check

After the 60 injections of soil extract, inlet maintenance was performed. This consisted of changing the septum, inlet liner, and gold seal, and removing 30 cm from the head of column 1. While the liner and gold seal were out, the inlet was cleaned with cotton swabs saturated with methanol. After maintenance, the 100 ppb calibration standard was run and quantitated using the original calibration curve generated before both of the replicate studies. Table 4 shows the measured concentrations. All analytes were within 7 % of the expected concentration. Table 4 presents the R² values for a full calibration after inlet maintenance. The data in Table 4 demonstrate that the degradation in system performance with the soil is limited to the inlet and column head, as expected.

The source did not require cleaning, as is often the case with matrix levels such as those used here. The use of JetClean and the 9 mm drawout lens greatly reduce the deposits that normally degrade source performance.

Table 4. Calibration check and R² values of 7 level ISTD-calibration 1 to 1,000 pg SIM after the system maintenance. [Table content listing compounds, RT, calculated concentration (%), and R² values after maintenance.]

Conclusions

This system addresses many of the problems encountered with GC/MS PAH analysis. The use of JetClean, the 9 mm drawout lens, higher zone temperatures, and the appropriate liner result in substantial improvements in linearity, peak shape, and system robustness. The greatly reduced need for manual source cleaning provided by JetClean is a welcome productivity improvement for the lab.

For labs analyzing high volumes of samples containing significant matrix interferences, the Agilent 8890/7000D triple quadrupole GC/MS with JetClean and midcolumn backflush offers all the advantages demonstrated here plus the much higher specificity of MS/MS⁴. Use of GC/MS/MS simplifies the data review versus GC/MS by providing much higher selectivity over spectral interferences from the matrix.

References

1. Szelewski, M.; Quimby, B. D. Optimized PAH Analysis Using the Agilent Self-Cleaning Ion Source and Enhanced PAH Analyzer, Agilent Technologies Application Note, publication number 5191-3003EN, 2013.
2. Anderson, K. A.; et al. Modified ion source triple quadrupole mass spectrometer gas chromatograph for polycyclic aromatic hydrocarbons, Journal of Chromatography A 2015, 1419(6), 89-9US.
3. Quimby, B. D.; Prest, H. F.; Szelewski, M. J.; Freed, M. K. In-situ conditioning in mass spectrometer systems, US Patent 8,378,293 Feb 19, 2013.
4. Andrianova, A. A.; Quimby, B. D. Optimized GC/MS/MS Analysis for PAHs in Challenging Matrices, Agilent Technologies Application Note, publication number 5994-0498EN, 2019.