GC-MS Solutions Enhancing Food and Environmental Testing

GC/MS solutions for environmental and food testing Ensure a Healthier World from the Ground Up Application Compendium Introduction 3 Environmental Analysis of PFAS and Environmental Contaminants in Soil 4 and Oat Plants with GC/Q-TOF Analysis of Phthalate with Hydrogen Carrier Gas 18 Automated Sample Prep Using Agilent PAL3 RTC 28 for EPA 8270E SVOC Analysis by GC/TQ Novel Column Chemistry Raises the Bar on Sensitivity 37 and Data Accuracy in SVOCs Analysis Accurate Mass Library for PFAS Analysis in Environmental Samples 49 Food Enhanced Longevity and Revolutionized Robustness for 61 Sensitive Detection of 190 Pesticides Over 800 Injections Brewing Excellence: Quantitating Over 200 Pesticides in Black Tea 76 with Steady Performance and Maximized Uptime by GC/MS/MS Essential Oils Analysis Using GC/MS with Hydrogen 101 and the Agilent Hydrolnert Source Ethylene Oxide and 2-Chloroethanol in Spices and Oilseeds 126 Using QuECHERS GC/MS/MS Analysis of Aldehydes in Beer by Agilent PAL 3 Autosampler 133 and 5977C GC/MSD 2 Table of Contents Agilent 5977C GC/MSD Agilent 7000E GC/TQ Agilent 7010D GC/TQ Agilent 7250 GC/Q-TOF 3 Ensuring the safety and quality of our environment and food supply is a significant and evolving challenge. Analytical testing laboratories are under constant pressure to meet the latest regulations for contaminants such as pesticides, PFAS, phthalates, and SVOCs. Precise, reliable, and efficient analytical workflows are paramount to maintaining the safety and quality of our food and environment. Agilent GC/MS systems deliver high-quality data, maximize productivity, and minimize downtime. Built-in instrument intelligence promotes ease-of-use and robust performance, while sustainability features work to lower your laboratory’s environmental footprint. The application notes in this compendium demonstrate these advanced GC/MS solutions for the analysis of water, soil, air, food, and consumer products. See how to meet today's—and prepare for tomorrow’s—environmental and food testing challenges. Meet the ever-evolving analytical challenges for environmental and food testing with Agilent GC/MS solutions Application Note Environmental Authors Luann Wong, Gabrielle Black, and Thomas Young Department of Civil and Environmental Engineering, University of California, Davis Sofia Nieto, Matthew Giardina, Matthew Curtis, and Tarun Anumol Agilent Technologies, Inc. Abstract Soil is one of the major environmental repositories of per- and polyfluoroalkyl substances (PFAS)1 , and the presence of PFAS in soil can potentially lead to ground water and food contamination. Current PFAS methods typically only cover 40 to 80 PFAS and vastly underestimate their presence in many environmental samples based on mass balance studies.2,3 Further, liquid chromatography/mass spectrometry (LC/MS) has significant limitations with respect to the analysis of some of the volatile classes of PFAS, which is where gas chromatography/mass spectrometry (GC/MS) should be considered as an important complementary technique. This study describes different approaches for the extraction and analysis of PFAS in soil and plants using an Agilent 7250 gas chromatography/quadrupole time-of-flight mass spectrometer (GC/Q-TOF). PFAS and other environmental contaminants were detected using a target screening methodology based on an accurate mass personal compound database and library (PCDL) of these pollutants. A broader range of contaminants was also identified in the soil and plant samples, including polychlorinated biphenyls (PCBs), polybrominated diphenyl ethers (PBDEs), and flame retardants, using a nontargeted screening and an extensive unit mass NIST23 library. Analysis of PFAS and Other Environmental Contaminants in Soil and Oat Plants Using High-Resolution GC/Q-TOF 4 Return to Table of Contents 2 Introduction PFAS are persistent synthetic organic pollutants with a potential to bioaccumulate.4 The list of PFAS substances curated by the Environmental Protection Agency (EPA) currently includes nearly 8,000 PFAS chemicals based on structure5 ranging from volatile PFAS, such as ubiquitous fluorotelomer alcohols (FTOHs), to long-chain PFAS, including the most commonly detected perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS). While long-chain PFAS are being phased out of production, the manufacturing of shorter chain length PFAS is increasing due to the assumption that more volatile PFAS are less toxic. These shorter chain length PFAS such as 6:2 FTOH are more difficult to detect with LC/MS using established methods, and recent studies have indicated that they are equally toxic.6,7 Soil is a significant reservoir of PFAS as well as of many other persistent environmental contaminants, and thus can contribute to contamination of ground water, atmosphere, and biota. Therefore, to better understand the source and transport of these contaminants, both soil and plant extracts have been analyzed using the Agilent 7250 GC/Q-TOF. To maximize the sensitivity of PFAS detection, a target screening approach based on a PFAS accurate mass library was used. The PFAS PCDL used in this study included over 150 electron ionization (EI) PFAS spectra along with retention times (RTs) and retention indices (RIs), and is described elsewhere.8 In addition to PFAS, many persistent pollutants were identified in both soil and plants, where both target and nontarget screening workflows were used. These pollutants included pesticides, polyaromatic hydrocarbons (PAHs), PCBs, PBDEs, and flame retardants. Experimental Sample collection Soil and oat plants were sampled from two fields in California (F1 and F2) that have historically received biosolids. The soil samples were collected prior to the application of biosolids (labeled PreA for preapplication). A certified USDA organic (Org) field was also sampled prior to treating the subplots with compost (Comp) and compost and lime (C&L). The compost was collected as well. The soil was also sampled at harvest time (Hvst). Plants were collected in the same regions as the soil samples. Sample preparation The soil and plant samples were either extracted with methylene chloride (DCM) for liquid injections or subjected to headspace solid-phase microextraction (HS-SPME). For DCM extraction, 2 g of soil was weighed into a 50 mL glass centrifuge tube and 5 mL of DCM was added. The 50 mL centrifuge tubes containing the samples were vortexed using a Heidolph Multi Reax Vibrating Test Tube Shaker at speed 5 for 5 minutes and centrifuged at 3,000 rpm for 5 minutes. Approximately 0.5 mL of supernatant extract was transferred into 2 mL autosampler vials. Whole plant samples that included stems, leaves, seeds, and seed pods were cut into 2 to 5 mm sections. Then, 2 g of plant samples were extracted with methylene chloride the same way as the soil. Method blank samples for each set were also generated. For HS-SPME, the soil (2 g) and finely chopped plant material (1 g) were transferred into a 20 mL headspace vial, and either 2 or 3 mL of DI water were added, respectively. SPME conditions The HS-SPME was performed using an Agilent PAL 3 CTC autosampler. Four different fibers were tested (Agilent 100 µm PDMS, 95 µm CWR/PDMS, 65 µm DVB/PDMS, and 80 µm DVB/CWR/PDMS, part number 5191-5878), and the SPME conditions were optimized. The fiber conditioning was carried out at 300 °C for 5 minutes. The samples were equilibrated for 10 minutes, and the SPME fiber was inserted into the vial headspace. Extraction was carried out at 50 °C for 35 minutes at 300 rpm (programmed for 10 seconds on, 2 seconds off cycle), with the desorption into the GC inlet at 250 °C for 7 minutes. The GC injection port was equipped with the 0.75 mm id liner for SPME analysis and the resistant to wear Merlin Microseal septa. 5 3 Data acquisition The GC/MS analysis was performed using an Agilent 7250 GC/Q-TOF system. All the data were acquired in full spectrum acquisition mode. Two different GC columns were used to acquire the data. The DB-624 is a midpolar GC column and provided the best retention and separation for GC-amenable PFAS compounds. This column was used for PFAS screening using the PFAS PCDL. The nonpolar DB-5ms column was also used to take advantage of RI information available for all the compounds with EI spectra in the NIST23 library. The data acquisition parameters are described in Table 1. Table 1. Data acquisition parameters. Agilent DB-5ms Agilent DB-624 MS Agilent 7250 GC/Q-TOF GC Agilent 8890 GC Inlet Multimode inlet, Agilent Ultra Inert 4-mm liner single taper with wool Inlet Temperature 70 °C for 0.01 min, 300 °C/min to 250 °C Injection Volume 1 µL Column Agilent J&W DB-5ms Ultra Inert (UI), 30 m × 0.25 mm, 0.25 µm Agilent DB-624 Ultra Inert, 30 m × 0.25 mm, 1.4 µm Oven Temperature Program 35 °C for 2 min, 7 °C/min to 210 °C, 20 °C/min to 300 °C, 4 min hold 30 °C for 2 min, 3 °C/min to 75 °C, 2 °C/min to 110 °C, 10 °C/min to 210 °C, 20 °C/min to 240 °C, 2 min hold Column Flow 1.2 mL/min constant flow 1 mL/min constant flow Carrier Gas Helium Transfer Line Temperature 250 °C Quadrupole Temperature 150 °C Source Temperature 200 °C Electron Energy 70 eV Emission Current Variable by time segment, 0.01 to 5 µA Spectral Acquisition Rate 5 Hz Mass Range (Tune) 50 to 1,200 m/z Data processing The nontargeted workflow was performed in Agilent MassHunter Unknowns Analysis software (version 12.1) and involved the SureMass chromatographic deconvolution and the NIST23 EI library search. RIs and accurate mass information were used to confirm the compound identification. The suspect screening was performed using the GC/Q-TOF Screener tool of MassHunter Quantitative Analysis software (version 12.1) and accurate mass libraries for pesticides and PFAS. Results and discussion Characteristic EI fragmentation of PFAS One of the approaches that could be beneficial for PFAS screening in complex matrices is a suspect screening approach since it allows for high sensitivity and specificity of detection. When using a high-resolution accurate mass GC/MS, this approach can be greatly facilitated by using accurate mass libraries to screen for a large number of target compounds that could, in theory, be unlimited. Thus, the accurate mass GC/MS PCDL for over 100 volatile and semivolatile PFAS compounds that was previously created6 was used in this work for PFAS screening in soil and plant samples. The PFAS compound classes in the PCDL included perfluoroalkyl iodides (PFAIs), fluorotelomer iodides (FTIs), fluorotelomer alcohols (FTOHs), fluorotelomer olefins (FTOs), fluorotelomer acrylates (FTACs), fluorotelomer methacrylates (FTMACs), fluorotelomer carboxylic acids (FTCA), fluorotelomer unsaturated carboxylic acids (FTUCA), perfluoroalkane sulfonamides (FASA), among others, many of which are uniquely amenable to GC/MS analysis. The electron ionization (EI) mode was chosen for the PFAS PCDL as a more universal technique compared to chemical ionization. EI covers a broader range of PFAS compound classes and allows users to easily screen for other contaminants in the same data file. While many PFAS compounds can highly fragment in EI, most of them nevertheless have specific fragment ions that could be selected by the GC/Q-TOF suspect screening algorithm or manually, as target or qualifier ions. Some of the most typical and specific fragments for the different PFAS compound classes are shown in Table 2. 6 4 Characteristic Fragments Neutral Loss (m/z) PFAS Class; % of Base Ion, as Maximum Observed for a Given PFAS Class FTOH PFAI FTI FTAC FTMAC FTO PFAL FASA [M]+ 0 – 40 100 30 90 – – – [M-I]+ 126.9045 – 100 – – – – – – [M-H2 O-HF]+ 38.0168 100 – – – – – – – [M-CHO-F]+ 48.011 – – – – – – 90 – [M-H2 O-F-HF-C2 H2 ]+ 83.0308 80 – – – – – – – [M-H2 O-2F]+ 56.0074 70 – – – – – – – [M-C2 F5 ]+ 118.992 – – 50 – – – – – [M-HF-I]+ 146.9107 – – 50 – – – – – [M-H2 O-CF3 ]+ 87.0058 30 – – – – – – – [M-H2 O-2F-CF3 ]+ 126.0026 30 – – – – – – – [M-F]+ 18.9984 6 – – 10 5 5 – 1 [M-CHO-2F]+ 66.9995 – – – – – – 25 [M-SO2 -CH3 ]+ 78.9854 – – – – – – – 25 [M-H2 O-CF2 ]+ 68.0074 25 – – – – – – – [M-HF]+ 20.0062 20 – – – – – – – [M-2F-CF3 ]+ 106.992 – – – – – 20 – – [M-H]+ 1.0078 15 – 1 – – – – – [M-CH3 ]+ 15.0235 – – – – 10 – – 5 [M-H-HF]+ 21.0141 15 – – – – – – – [M-F-2HF]+ 59.0109 15 – – – – – – – [M-H2 O-F]+ 37.009 15 – – – – – – – [M-CF3 -HF]+ 89.0014 – – 10 – – 5 – – [M-F-HF]+ 39.0046 – – – – – 10 – – [M-NH2 SO2 ]+ 79.9806 – – – – – – – 10 [M-C2 H3 -2F]+ 65.0203 – – – – – 10 – – [M-CHO]+ 29.0027 – – – – – – 5 – [M-SO2 -H]+ 64.9697 – – – – – – – 5 [M-SO2 -F]+ 82.9603 – – – – – – – 5 [M-SO2 -CF3 -HF]+ 152.9633 – – – – – – – 5 Table 2. Characteristic fragments of volatile PFAS in EI. 7 5 Selection of SPME fiber for soil analysis Four different SPME fibers were evaluated for the ability to extract volatile compounds (including PFAS) from soil: PDMS, CWR/PDMS, DVB/PDMS, and DVB/CWR/PDMS. The test was performed using soil (2 g) sampled from the same location, mixed with 2 mL of water, and run under the same SPME conditions. Total ion chromatograms (TIC) generated by each fiber tested are shown in Figure 1. Both DVB/PDMS and DVB/CWR/PDMS fibers produced a significant number of peaks and showed the ability to extract a wide range of compounds. The number of the identifiable peaks was also evaluated for each of the fibers (Table 3). DVB/PDMS and DVB/CWR/PDMS had a comparable number of library hits that was slightly higher for DVB/CWR/PDMS. Therefore, it was selected for further analysis. Table 3. SPME fibers performance. The number of components generated by the SureMass deconvolution algorithm as well as number of NIST23 library hits (Library Match Score cutoff 75) are shown. Fiber Type Number of Components Number of Hits PDMS 422 228 CWR-PDMS 514 419 DVB-PDMS 687 560 DVB-CWR-PDMS 683 570 0 2.5 5.0 7.5 0 2 4 6 0 2 4 0 2 4 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 PDMS 100 µm CWR-PDMS 95 µm DVB-PDMS 65 µm DVB-CWR-PDMS 80 µm ×107 ×107 ×107 ×107 Acquisition time (min) Counts Counts Counts Counts Figure 1. SPME fiber performance on soil samples. 8 6 Detection of volatile PFAS in soil and plant samples using accurate mass PFAS library For PFAS detection using the accurate mass PFAS PCDL for GC/Q-TOF, the midpolar DB-624 GC column (for more detail, see Table 1) was used. Both targeted and nontargeted methodologies that involved the GC/Q-TOF Screener and the Unknowns Analysis software, respectively, were used to identify PFAS in soil and plant samples. An advantage of using the nontargeted analysis is that both accurate mass libraries, as well as large comprehensive public libraries such as NIST, could be used to screen for the contaminants simultaneously without reinjection. There are several benefits of the targeted, PCDL-based suspect screening approach, which is performed entirely in MassHunter Quantitative Analysis software and has been described in detail previously.9,10 Some of the main advantages of this approach include high sensitivity and a high degree of flexibility and automation of the data analysis method setup and results validation. The validation requires minimal human involvement prior to the reporting. Together this provides a user with a highly effective and time-saving tool for targeted analysis. A few PFAS compounds were detected when analyzing the data from SPME. An example of a compound identified in a few soil and plant samples using the GC/Q-TOF Screener is shown in Figure 2. This compound is a volatile 6:2 fluorotelomer alcohol, frequently detected in environmental matrices. Due to the trace amount of this compound, it was not found in a nontargeted approach. Figure 2. 6:2 FTOH detected in soil using SPME and PFAS PCDL-based screening approach. The mirror plot at the top shows the deconvoluted compound spectrum versus the spectrum from the PFAS PCDL. The mirror plot at bottom displays only target and qualifier ions. Sample: soil, F1 9 7 When performing nontarget analysis, the chromatographic deconvolution uses the SureMass algorithm, which is specifically optimized for high-resolution EI data to ensure high speed, sensitivity, and integrity of spectral extraction. While the suspect screening approach provided the highest degree of sensitivity and was able to detect more PFAS compounds, one of the more abundant PFAS was detected in both approaches (Figures 3A and 3B). All the PFAS compounds identified in soil and plant samples were detected by matching the PFAS PCDL. Figure 3. PFAS (ethyl perfluorobutyl ether) identified in DCM soil extracts using PFAS PCDL in both target screening (A) and an untargeted, deconvolution-based approach in the Unknowns Analysis software (B). A Sample: soil, PreA F1 B Sample: soil, PreA F1 10 8 Overall, the HS-SPME approach for PFAS extraction worked best and provided a higher number of identified volatile PFAS compounds. The amounts of PFAS detected (summarized in Table 4) were estimated based on the standard injections where the actual concentrations in soil and plant samples have not been determined. Identification of other contaminants in soil and oat plants The identification of the additional contaminants in soil and plant samples was performed for both DCM extracts and SPME. However, while SPME allowed better detection of volatile compounds, a higher number of the environmental contaminants that included PCBs, PBDEs, pesticides, PAHs, and flame retardants, were detected in DCM extracts and will thus be the focus of the further discussion. The separation was carried out using the DB-5ms UI column to be able to use the RI values from the extensive NIST23 library, and thus increase confidence in compound identification by using the RI penalty function for library hits. This column phase is also compatible with the GC/Q-TOF Pesticide PCDL, which could be considered for screening GC/Q-TOF accurate mass EI data for pesticides and PAHs. After a quick prescreening, the identified pollutants were grouped by contaminant classes and approached separately. PCBs and PBDEs were identified in the nontargeted approach using the Unknowns Analysis and NIST23 library. To eliminate false positives based on the accurate mass EI data while searching a unit mass library, the Unknowns Analysis ExactMass tool was used. This tool is described in further detail in Figure 4A, which shows an example of one of the BDEs detected in a soil extract. There were 20 different PCBs and PBDEs detected in the soil extracts (Figure 4B). The only oat plant extract where this group of contaminants (BDE-47) was detected was grown in field F2. Note that PCBs and PBDEs were not detected by SPME due to their high boiling point. Another prominent group of contaminants identified in soil extracts was pesticides, and the GC/Q-TOF Screener workflow with Pesticide PCDL was used for quick and streamlined detection of these pollutants. The original version of the Pesticide PCDL, which is RT-based, was supplemented with RIs to be able to use the PCDL in the screener workflow together with the data acquired using a different chromatographic method. The GC/Q-TOF Screener method was set up in agreement with the SANTE guidelines. However, RT windows were expanded to allow for an additional RT error introduced using a different chromatographic method. Table 4. PFAS detected in soil and plants by HS-SPME using the accurate mass PFAS PCDL and suspect screening approach. The estimated amounts (in pg on column) are shown. Compound RT Quantifier Ion Soil Samples Plant Samples F1 PreA F1 Hvst F2 Hvst C&L Hvst Compost Hvst Organic Hvst Organic Compost F1 F2 Compost C&L Org Ethyl Perfluorobutyl Ether 4.4 218.9851 150.2 – – – – – – – – – – – 6:1 Fluorotelomer Alcohol 20.94 130.9915 – 2 – – – – – 2.2 – – – – 6:2 Fluorotelomer Alcohol 23.59 296.0054 – 7.5 0.3 – – – 6.9 2.5 – – – – N-Methylperfluorooctanesulfonamide 43.1 93.9957 0.3 3.4 0.9 2.1 0.4 1.2 0.9 0.2 – – – – 11 9 Figure 4. PCBs and PBDEs in soil DCM extracts using NIST23. (A) An example of BDE detected in soil sample collected from F1 at the time of harvesting. ExactMass table (bottom-left panel) shows how well the accurate mass fragment ions matched the unit mass library hit and thus provides the additional confirmation of compound ID. The most selective and abundant ions are highlighted in the mirror plot when m/z corresponds to the library hit formula. The arrow in the component chromatogram points to the identified component EICs. (B) Bar graph showing responses of PCB and PBDE for all the soil samples where they have been identified. A Sample: soil, F1 B BDE-99 0 50,000 100,000 150,000 200,000 250,000 300,000 Area BDE 154 BDE 99 BDE 100 PCB 194 BDE 47 PCB 170 PBDE 71 PCB 171 PCB 183 PCB 128 PCB 177 PCB 153 PCB 146 PCB 123 PCB 149 PCB 151 PCB 122 PCB 125 PCB 98 PCB 11 F1 (PreA) F1 (Hvst) F2 (Hvst) Compost Compost (Hvst) CL (Hvst) Org (Hvst) 12 10 Since the DCM extraction method was not optimized specifically for pesticide recovery from plant matrices, only soil samples were processed. In total, over 50 pesticides were detected in soil extracts (Figure 5 and Table 5). A large number of pesticides were detected in compost and compost-treated soils, and a few pesticides were also identified in organic soil extracts. Another interesting observation was that the insecticides fipronil sulfide and fipronil sulfone were consistently found in the same soil samples. Also, conazole fungicides such as propiconazoles, myclobutanil and difenconazole were mostly detected in compost and compost-treated soils. Table 5. Pesticides detected in soil extract using the accurate mass Pesticide PCDL and suspect screening approach. Compound Name RT RT Delta* Library Match Score F1 PreA F1 Hvst F2 Hvst C&L Hvst Compost Hvst Org Hvst Org Compost 1,2,4-Trichlorobenzene 14.62 0.17 98.1 x Diuron Metabolite 16.56 0.28 99.9 x x x x x x 1,2,3,5-Tetrachlorobenzene 16.98 0.32 90.7 x 2,4,6-TCP/2,4,6-Trichlorophenol 17.44 0.25 99.3 x x x Trace Nicotine 17.56 0.05 97.9 x Lufenuron 18.71 0.22 99.7 x x 3,4-DCA/3,4-Dichloroaniline 18.88 0.21 99.9 x x x Pentachlorobenzene 20.42 0.35 99.4 x x x DEET/Diethyltoluamide 21.46 0.22 82.1 Trace Trace Trace x x Trace x 2,3,4,5-Tetrachloroanisole 22.74 0.32 99.8 x x Bromoxynil 23.14 0.09 99.9 Trace x x HCB/Hexachlorobenzene 23.52 0.36 99.7 x x x Trace x x x Dichloran (Dicloran) 23.84 0.16 97.8 x x x Swep (MCC) 24.27 0.10 85.6 x x x PCP/Pentachlorophenol 24.27 0.24 99.7 Trace x x Figure 5. GC/Q-TOF Screener window of MassHunter Quantitative Analysis software when screening for pesticides in soil extracts using the Pesticides PCDL. Sample: soil, F1 13 11 Compound Name RT RT Delta* Library Match Score F1 PreA F1 Hvst F2 Hvst C&L Hvst Compost Hvst Org Hvst Org Compost Pyrimethanil 25 0.10 82.5 Trace x Chlordene 25.1 0.03 98.5 x Pentachloroaniline 25.74 0.24 99.7 x x x x x Dithiopyr 26.85 0.40 93.7 x x x x x x Anthraquinone 27.59 0.05 99.7 x x x x x x x 4,4'-Dichlorobenzophenone 27.86 0.02 80.3 x x Fipronil Sulfide 28.13 0.43 99.7 x Trace x x Cyprodinil 28.27 0.05 99.1 Trace x x Diuron 28.28 0.31 78.1 Trace x Trace Fluopyram 28.49 0.18 93 x x x Chlorbenside 28.99 0.21 92.2 x x x Chlordane-trans (γ-Chlordan) 28.82 0.03 99.7 x x x x x x x Triclosan 28.85 0.03 99.3 x x x x x Chlordane-cis (α-Chlordan) 29.06 0.04 99.9 x x x x x x x Nonachlor-trans 29.11 0.07 99.9 x x x x x x x Flutolanil 29.23 0.11 85.2 x x Fludioxonil 29.27 0.18 99.6 x Dieldrin 29.5 0.05 84.6 x Trace Trace p,p'-DDE 29.42 0.04 99.2 x x x x x x x Oxadiazon 29.43 0.14 97.7 x x x o,p'-DDD (Mitotane) 29.52 0.07 99.9 x x x Trace x x Fipronil Sulfone 29.34 0.33 98.6 Trace x Trace x x Myclobutanil 29.48 0.11 98.8 x x p,p'-DDD 30.03 0.02 99.5 x x x x x x Trace Nonachlor-cis 30.03 0.04 99.9 x x x Trace Trace x Carfentrazone-ethyl 30.32 0.13 98.8 x Bromoxynil octanoate 30.39 0.06 88.4 x x Trace Propiconazole I 30.43 0.11 96.4 x x Trace x Chloridazon (PAC) 30.44 0.06 91.5 x Propiconazole II 30.5 0.04 96 x x Trace x Tebuconazole 30.7 0.03 92.7 x x x Chlorbenside Sulfone 30.7 0.39 89.6 Trace Trace x x x Bifenthrin 31.08 0.09 98.6 Trace x Trace x x x x cis-Permethrin 32.19 0.00 99.1 x x x Trace x x x trans-Permethrin 32.28 0.01 81 x Trace Trace Difenconazole II 34.09 0.01 88.9 Trace x * RT delta is recalculated from RI. Trace indicates that the Library Match Factor is below 75. 14 12 PAHs are included into the Pesticide PCDL and were screened together with pesticides in a single workflow. PAHs were mostly detected in soil extracts. However, phenanthrene and fluoranthene were also identified in most plant samples. This was not unexpected considering that phenanthrene and fluoranthene were the most abundant PAHs detected in the soil. A prominent group of contaminants identified in soil and oat plants was flame retardants. The accurate mass spectra for most of these compounds are already included in the Pesticide PCDL. For a more comprehensive coverage of this group of pollutants, a couple of flame retardants identified in the Unknowns Analysis with NIST23 library, missing in the Pesticide PCDL, were added to the Quant method directly from the Unknowns Analysis software, and thus were screened together with the rest of the targets. Remarkably similar responses were observed between soil and plant extracts for this group of pollutants (Figure 7). Among the most abundant flame retardants identified in soil and plant samples were tributyl phosphate, tris(2-chloropropyl) phosphate, and tris(3-chloropropyl) phosphate, the phosphorus flame retardants of frequent use. 0 500,000 1,000,000 1,500,000 2,000,000 2,500,000 Area 0 5,000,000 10,000,000 15,000,000 20,000,000 25,000,000 30,000,000 35,000,000 Benzo[ghi]perylene Dibenz[a,h]anthracene Benzo[a]pyrene Benzo[k]fluoranthene Benz[a]anthracene Fluoranthene Anthracene Phenanthrene Fluorene Acenaphthene Acenaphthylene F1 (PreA) F1 (Hvst) F2 (Hvst) Compost Compost Naphthalene (Hvst) CL (Hvst) Org (Hvst) Figure 6. PAHs detected in soil DCM extracts using the accurate mass Pesticide PCDL and suspect screening approach. The bar graph shows the PAH peak area. 15 13 Conclusion PFAS analysis in environmental matrices is a challenging undertaking. In this application note, different approaches for PFAS extraction from soil and plants as well as downstream workflows of data processing have been discussed. The most efficient and sensitive approach for volatile PFAS analysis in soil and plants suggested here is HS-SPME combined with the suspect screening based on the PFAS accurate mass library and high-resolution accurate mass GC/Q-TOF. In addition, both soil and plant extracts were screened for other contaminants, and various pollutants including PCBs, PBDEs, PAHs, pesticides, and flame retardants were identified using targeted, nontargeted, and combined methodologies. Figure 7. Flame retardants detected in soil and plant DCM extracts using a combined screening approach that included both accurate mass PCDL as well as NIST23 library. The bar graph shows flame retardant response in soil (S) and plant (P) extracts. S S P S P S S P S P S P 0 100,000 200,000 300,000 400,000 500,000 600,000 700,000 Area 2-Ethylhexyl diphenylphosphate (octicizer) TPP/Triphenyl phosphate Bis(3-chloro-1-propyl)(1-chloro-2-propyl)phosphate Tris(3-chloropropyl)phosphate TCPP/Tris(2-chloropropyl) phosphate Tri(2-chloroethyl) phosphate TBP/Tributylphosphate 2,4,6-Tribromophenol 2,4,6-Tribromoanisole F1 (PreA) F1 (Hvst) F2 (Hvst) Compost Compost (Hvst) CL (Hvst) Org (Hvst) 16 www.agilent.com DE45375629 This information is subject to change without notice. © Agilent Technologies, Inc. 2024 Printed in the USA, May 6, 2024 5994-7351EN References 1. Brusseau, M. L.; Anderson, R. H.; Guo, B. PFAS Concentrations in Soils: Background Levels versus Contaminated Sites. Sci. Total Environ. 2020, Oct 20; 740, 140017. DOI: 10.1016/j.scitotenv.2020.140017 2. Lin, H.; Taniyasu, S.; Yamazaki, E.; Wu, N.; Lam, P. K. S.; Eun, H.; Yamashita, N. Fluorine Mass Balance Analysis and Per- and Polyfluoroalkyl Substances in the Atmosphere. J. Hazard. Mater. 2022, Apr 28; 435, 129025. DOI: 0.1016/j.jhazmat.2022.129025 3. Spaan, K.; Van Noordenburg, C.; Plassman, M.; Schultes, L.; Shaw, S. D.; Berge,r M.; Heide-Jørgensen, M. P.; Rosing-Asvid, A.; Granquist, S.; Dietz, R.; et al. Fluorine Mass Balance and Suspect Screening in Marine Mammals from the Northern Hemisphere. Environ. Sci. Technol. 2020, Mar 12, 54(7), 4046–4058. DOI: 10.1021/ acs.est.9b06773 4. Schildroth, S.; Rodgers, K. M.; Stynar, M.; McCord, J.; Poma, G.; Covaci, A.; Dodson, R. E. Per-and Polyfluoroalkyl Substances (PFAS) and Persistent Chemical Mixtures in Dust from U.S. Colleges. Environ. Res. 2021, Apr 15, 206, 112530. DOI: 10.1016/j.envres.2021.112530 5. Williams, A. J; Gaines, L. G. T.; Grulke, C. M.; Lowe, C. N.; Sinclair, G. F. B.; Samano, V.; Thillainadarajah, I.; Meyer, B.; Patlewicz, G.; et al. Assembly and Curation of Lists of Per- and Polyfluoroalkyl Substances (PFAS) to Support Environmental Science Research. Front. Environ. Sci. 2022, Apr 5, 10, 1–13. DOI: 10.3389/fenvs.2022.850019 6. Sunderland, E. M.; HuHu, X. C.; Dassuncao, C.; Tokranov, A. K.; Wagner, C. C.; Allen, J. G. A Review of the Pathways of Human Exposure to Poly- and Perfluoroalkyl Substances (PFASs) and Present Understanding of Health Effects. J. Expo. Sci. Environ. Epidemiol. 2019 Mar 29, (2), 131–147. DOI: 10.1038/s41370-018-0094-1 7. Rice, P. A.; HuAungst, J.; Cooper, J.; Bandele, O.; Kabadi, S. V. A Comparative Analysis of the Toxicological Databases for 6:2 Fluorotelomer Alcohol (6:2 FTOH) and Perfluorohexanoic Acid (PFHxA). Food Chem. Toxicol. 2020 Apr, 138, 111210. DOI: 10.1016/j.fct.2020.111210 8. Wong, L.; Black, G.; Young, T.; Nieto, S. Accurate Mass Library for PFAS Analysis in Environmental Samples and Workflow for Identification of Pollutants in Drinking Water Using GC/Q-TOF. Agilent Technologies application note, publication number 5994-6966EN, 2023. 9. Van Gansbeke, W.; Albertsdóttir, A. D.; Polet, M.; Van Eenoo, P.; Nieto, S. Introducing Semi-Automated GC/Q-TOF Screening with the AssayMAP Bravo Sample Prep Platform for Antidoping Control. Agilent Technologies application note, publication number 5994-6702EN, 2023. Application Note Environmental Authors Dinh Loc Dao, Minh Trung Tran, Xuan Dai Phan, Quoc Trung Pham, Thi Linh Nguyen, and Tuan Dat Ho IndoChina Center of Excellence and Red Star Vietnam Company Limited – CMS. Ho Chi Minh City, Vietnam Boonraksa Srisawang Agilent Technologies, Inc. Singapore, Singapore Abstract Gas chromatography/mass spectrometry (GC/MS) plays a pivotal role in the analysis of phthalates in consumer products. In light of the recent helium supply challenges, the adoption of hydrogen (H2 ) as a carrier gas has gained prominence. However, using hydrogen as carrier gas in many GC/MS analyses may cause hydrogenation or dechlorination problems in the ion sources that use helium gas as carrier gas. The Agilent GC 8890, coupled with the Agilent 5977C GC/MSD and using the Agilent HydroInert source with hydrogen as the carrier gas with the backflush technique, provides an effective solution to these challenges. It not only meets International Electrotechnical Commission (IEC) requirements but also successfully addresses the issue of instability when dealing with complex sample matrices. Analysis of Phthalate with Hydrogen Carrier Gas Using the Agilent HydroInert source on a challenging cable sample in gas chromatography/mass spectrometry 18 Return to Table of Contents 2 Introduction Phthalates are a group of chemical compounds that are synthesized from phthalic acid. They are widely used as plasticizers in many products, including children's toys and electrical and electronic equipment. Unfortunately, phthalates can leach out of plastic materials and enter the environment. This exposure can have adverse effects on human health. The prevalence of phthalates in various products and the associated health risks have led to the implementation of regulations governing their use. Relevant policies include the European Union’s updated directive under the Restriction of Hazardous Substances (RoHS) framework, designated as 2015/863/EU1 , and its 2007 Registration, Evaluation, Authorization, and Restriction of Chemicals (REACH) framework.2 Traditionally, helium has been the preferred carrier gas for GC/MS analysis. However, due to recurring helium shortages and rising costs, hydrogen is emerging as a viable alternative. Hydrogen is more cost-effective and efficient than helium, making it a promising prospect for phthalate analyses. This application note centers on the analysis of phthalates using a GC/MS system in Selected Ion Monitoring (SIM) mode with hydrogen serving as the GC carrier gas. When transitioning to hydrogen for GC/MS analysis, several critical factors must be considered. Hydrogen, being a chemically reactive gas, can induce chemical reactions within various components of the GC/MS system, including the inlet, column, and sometimes the mass spectrometer Electron Ionization (MS EI) source. These reactions can alter the mass spectra of peaks in the total ion chromatogram (TIC), potentially leading to the misidentification of compounds in the analysis results. The Agilent HydroInert source was developed to mitigate these concerns related to the MS source. The HydroInert source, when used with hydrogen as the carrier gas, preserves mass spectral accuracy and allows users to continue using existing helium-based mass spectral libraries and quantitative methods. This application note uses the solvent extraction method, which is in accordance with the analytical methodology outlined in IEC standard 62321-8:2017.3 The presence of residual sample matrix components in this method can contaminate essential column components, such as the inlet, column, and ionization source. This contamination can adversely affect the accuracy and reliability of the instrument during analysis. The self-cleaning capability of the hydrogen gas ionization source, combined with the backflush technique, helps address issues related to sample background effects, reduces maintenance frequency, and ensures precision and accuracy throughout the analysis process. Experimental Reagents and chemicals – Tetrahydrofuran (THF) 99.9% was purchased from Sigma-Aldrich. – HPLC-grade acetonitrile (ACN) was obtained from Supelco. – Benzyl benzoate (BB) 98%, benzyl butyl phthalate (BBP) 98%, dibutyl phthalate (DBP) 98%, di-isobutyl phthalate (DIBP) 98%, di-n-pentyl phthalate (DPENP) 2,500 mg/L, di-n-hexyl phthalate (DHEXP), dicyclohexyl phthalate (DCHP) > 98%, di(2-ethylhexyl) phthalate (DEHP) > 98%, dinoctyl phthalate (DNOP) > 98%, di-isononyl phthalate (DINP) > 98%, and di-isodecyl phthalate (DIDP) were obtained from Sigma-Aldrich. Standards solution and standards preparation Phthalate ester standard solutions were prepared by diluting a mixture of phthalate ester standards at a concentration of 100 mg/L in a mixture of THF:ACN (1:1) to create a series of standards with concentrations of 0.2, 0.5, 1, 2, and 5 µg/mL, respectively. Also, each of the standard solutions was supplemented with an internal standard, BB, at a concentration of 1 µg/mL. Sample preparation 1. Weigh 300 mg of the sample and transfer it to a 40 mL vial. Record the mass to the nearest 0.1 mg. 2. Add 10 mL of THF and 30 µL of internal standard BB (1,000 µg/mL) to the vial. 3. Seal the vial tightly and place it in an ultrasonic bath. Sonicate for 30 to 60 minutes until the sample dissolves. 4. After the sample has completely dissolved, let the vial cool to room temperature. 5. Carefully add 20 mL of ACN drop by drop into the vial to precipitate the sample matrix. 6. Let the resulting solution stand at room temperature for 30 minutes. 7. Allow the polymer to settle or filter the mixture through a 0.45 µm polytetrafluoroethylene membrane. 8. Inject on GC/MS. 19 3 Instrumentation The experimental system used in this study was configured (Figure 1) to address and reduce potential challenges arising from the use of hydrogen as the carrier gas and the complexity of the sample matrix in phthalates analysis. The system comprised an 8890 GC configured with an Agilent 7693A automatic liquid sampler (G4513A) and a split/splitless inlet coupled to a 5977C gas chromatography/mass selective detector (GC/MSD) configured with a HydroInert source. Tables 1 and 2 show the GC and MS conditions for phthalate analysis and the SIM ion parameters used for data acquisition. Parameter Value Agilent 8890 GC SSL Inlet 250 °C, Mode: split (5:1) Liner Inlet liner, Ultra Inert, split, low pressure drop, glass wool (p/n 5190-2295) Injection Volume 1 µL Column 0.7 mL/min, constant flow mode (H2  ) Agilent DB-5ms Ultra Inert, 30 m × 0.18 mm, 0.18 µm (p/n 121-5522UI) Restrictor 1.1 mL/min, constant flow mode (H2  ) Purge flow: 3 mL/min Agilent DB-5ms Ultra Inert 5 m × 0.15 mm, 0.15 µm (p/n 165-6626) Oven Program 110 °C (0.5 min) 30.5 °C/min to 280 °C (0.5 min) 30.5 °C/min to 320 °C (2 min) Agilent 8890 GC Backflush Parameters Inlet Pressure (Backflushing) 2 psi Backflush Pressure 20 psi Void Volumes 5.1715 Backflush Time 4 min Agilent 5977C GC/MSD Source Temperature 280 °C MS Quadrupole 150 °C MS Transfer Line 280 °C Acquisition Type SIM mode Gain Factor 1.0 Table 1. GC and MS conditions for phthalate analysis. Compound RT (min) Quantifier Ion (m/z) Qualifier Ions (m/z) BB 5.133 194 105, 212 DIBP 5.395 149 150, 167 DPP 5.724 150 149, 223 DPENP 6.347 149 150, 237 BBP 7.002 149 150, 251 DBP 7.071 149 91, 150 DCHP 7.632 149 150, 167 DEHP 7.666 149 150, 167 DNOP 8.23 149 150, 249 DINP 8.554 293 127, 149 DIDP 8.898 307 149, 167 Table 2. SIM ion parameters used for data acquisition (same with helium as carrier gas). Figure 1. System configuration. S/SL Inlet = split/splitless inlet, PUU = purged ultimate union. Liquid injector HydroInert source witn 9 mm extractor lens EI source Agilent 8890 GC Agilent 5977C GC/MSD PSD (H2 ) S/SL inlet (H2 ) PUU Column Restrictor 20 4 Results and discussion Chromatographic performance IEC 62321-8 describes a 17-minute GC method for the determination of phthalates in polymers by GC/MS. When using hydrogen as the carrier gas, an elution profile similar to previous studies conducted with helium was observed. The combination of hydrogen as the carrier gas and a smaller diameter column resulted in a faster run time compared to helium. The SIM chromatogram of the 1 µg/mL standard solution containing the phthalate is shown in Figure 2. Good chromatographic peak shapes and signal-to-noise ratio (S/N) were obtained for 10 phthalates (excluding the internal standard) in 15 minutes (including backflushing time). Linearity and range The linearity and range of the analysis were determined by constructing solvent and standard addition calibration curves for all phthalate compounds within the concentration range of 0.2 to 5 mg/L. Internal standard (BB) concentration was 1 mg/L. The linearity was confirmed by observing R2 values exceeding 0.996. Figure 3 displays representative calibration plots for four phthalate compounds. Also, the bias percentage for all phthalate compounds at the lowest concentration standard fell within the acceptable range of 80 to 120% of the actual concentration (Table 3). Chemicals Type Origin Weight R2 %Bias STD 1 BBP Linear Include 1/x 0.99921 106.4 DIBP Linear Include 1/x 0.99849 104.9 DBP Linear Include 1/x 0.99841 111.3 DCHP Linear Include 1/x 0.99889 112.1 DEHP Linear Include 1/x 0.99902 108.0 DIDP Linear Include 1/x 0.99912 996.9 DPENP Linear Include 1/x 0.99931 104.9 DNOP Linear Include 1/x 0.99802 115.8 DINP Linear Include 1/x 0.99935 100.1 DPP Linear Include 1/x 0.99920 106.3 Table 3. Linear least-squares regression weighted calibration data for 10 phthalates. Figure 2. SIM TIC of 10 phthalates at 1 mg/L and BB as internal standard (ISTD) at 1 mg/L. 0 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25 2.50 2.75 3.00 Acquisition time (min) 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0 10.5 Counts ×105 1 2 2 3 3 21 5 Relative concentratio n 0 0.4 0.8 1.2 1.6 2 2. 4 2. 8 3. 2 3. 6 4 4. 4 4. 8 5. 2 Relative responses 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 y = 3.705611x – 0.319318 R2 = 0.9992 R = 0.9998 Relative concentratio n 0 0.4 0.8 1.2 1.6 2.0 2. 4 2. 8 3. 2 3. 6 4. 0 4. 4 4. 8 5. 2 Relative responses 012345678 y = 1.571538x – 0.164829 R2 = 0.9984 R = 0.9997 Relative concentratio n 0 0.4 0.8 1.2 1.6 2.0 2. 4 2. 8 3. 2 3. 6 4. 0 4. 4 4. 8 5. 2 Relative responses 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 y = 2.374077x – 0.235474 R2 = 0.9990 R = 0.9997 Relative concentratio n Relative responses ×1 0 1 ×1 0 1 ×1 0 1 BBP DBP DEHP DIBP0 0. 5 1. 0 1. 5 2. 0 2. 5 3. 0 3. 5 4. 0 4. 5 5. 0 0 0. 2 0. 4 0. 6 0. 8 1. 0 1. 2 1. 4 1. 6 1. 8 y = 3.736018x – 0.332544 R2 = 0.99849070 Figure 3. Calibration curves of four phthalates in ROSH 3 from 0.2 to 5 mg/L. 22 6 Method detection limit (MDL) The MDL was estimated based on the standard deviation (SD) of the analysis results of nine spiked polymer samples on two different days at 50 mg/kg. Note that phthalates were not detected in the original polymer samples. The MDLs for phthalate in the polymer matrix ranged from 1.80 to 3.74 mg/kg, as illustrated in Figure 4. For all 10 phthalate compounds, the calculated relative standard deviations (RSDs) were below 20%. These MDL values meet the low detection limits required for phthalate analysis in electrotechnical products. Recovery, repeatability, and reproducibility The recovery, repeatability, and reproducibility were evaluated based on the results of the analysis of 14 spiked samples on two different days. The samples were spiked at concentrations of 50 and 500 mg/kg. The average values of recovery obtained for the spiked samples at different concentrations ranged from 89.6 to 101.1%, as shown in Figure 5. Recovery between 70 and 120% is considered satisfactory based on the limits specified in IEC 62321-8. In addition, the %RSD of the recovery values calculated from 14 spiked samples on three days for each concentration was less than 20%, which satisfies the requirements of IEC 62321-8. Figure 4. MDLs of phthalates in polymer samples. MDL (mg/kg) RSD% 0.5 1.0 0 1.5 2.0 2.5 3.0 3.5 4.0 0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0% 3.5% 4.0% 3.65 2.15 3.37 3.57 2.71 2.19 3.36 3.07 3.74 1.80 3.52% 2.07% 3.23% 3.36% 2.60% 1.93% 2.97% 2.97% 3.27% 1.72% BBP DIBP DBP DCHP DEHP DIDP DINP DPENP DNOP DPP 23 7 Figure 5. (A) Recovery% at 50 mg/kg of phthalate in polymer samples. (B) Recovery% at 500 mg/kg of phthalate in polymer samples. Recovery% RSD% Recovery% RSD% 0% 1% 2% 3% 4% 80% 85% 90% 95% 100% 105% 0% 1% 2% 3% 4% 5% 80% 85% 90% 95% 100% 105% 89.8% 90.6% 92.2% 90.2% 98.7% 98.4% 89.6% 99.3% 90.6% 3.52% 2.07% 3.23% 3.36% 2.60% 1.93% 2.97% 2.97% 3.27% 1.72% BBP DIBP DBP DCHP DEHP DIDP DINP DPENP DNOP DPP 91.7% 89.3% 92.3% 93.0% 92.3% 96.3% 101.1% 90.3% 92.7% 89.6% 4.0% 1.1% 2.4% 2.9% 3.2% 2.2% 2.9% 2.1% 2.6% 2.8% BBP DIBP DBP DCHP DEHP DIDP DINP DPENP DNOP DPP 90.3% A B 24 8 Analysis of various cable samples All MDLs and limits of quantification (LOQs ≈ 3MDLs) were below the lowest calibration point for all compounds. Therefore, the reporting limits for this study were defined as any value greater than or equal to the lowest calibration point for the respective compound. Three real-world cable samples were assessed for the presence of phthalate. The chromatogram and result of three cable samples are presented in Figure 6 and Table 4. Compound Cable 1 Cable 2 Cable 3 DIBP Detected Detected Not detected DPP Detected Detected Not detected DPENP Not detected Not detected Not detected BBP Not detected Not detected Not detected DBP Not detected Not detected Not detected DCHP Not detected Not detected Detected DEHP Significant level detected Significant level detected Not detected DNOP Not detected Not detected Significant level detected DINP Significant level detected Detected Not detected DIDP Significant level detected Detected Not detected Not detected: below the calibration range tested Detected: within the calibration range Significant level detected: above the calibration range Table 4. Phthalate result analysis of three cable samples. Figure 6. TIC SIM of phthalates in three cable samples. 2 3 3 3 1 2 3 1 2 2 3 1 2 2 3 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4 5 6 7 8 9 10 8.573 7.642 0 0.5 1.0 1.5 2.0 2.5 4 5 6 7 8 9 10 7.676 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Acquisition time (min) 4 5 6 7 8 9 10 8.336 Counts Counts Counts ×107 ×107 ×107 25 9 Method robustness in cable matrix When using helium gas for phthalate analysis on cable samples, a common issue is a significant reduction on compound responses, which cause high %RSD for certain types of cable matrices. The method robustness when using Agilent HydroInert source with hydrogen carrier gas, in combination with the backflush technique, was assessed through continuous analysis of standards and three different cable sample matrices. The %RSD values for the response of 1 mg/L standard points of phthalates analyzed throughout the analysis of the three cable sample matrices are presented in Table 5. Sample N (Standard Injection) RSD Area DIBP DPP DPENP BBP DBP DCHP DEHP DNOP DINP DIDP Cable 1 (80 injections) 24.000 7.00 6.62 5.74 5.68 6.35 5.85 5.68 5.96 5.67 6.35 Cable 2 (30 injections) 24.000 6.11 6.16 6.08 5.83 5.72 6.30 5.96 5.66 2.41 3.44 Cable 3 (30 injections) 15.000 4.16 3.75 3.73 4.11 3.89 3.98 3.64 3.95 1.96 2.08 Table 5. The %RSD values for the response at 1 mg/L in the method robustness test. The %RSD of the response, calculated from the 1 mg/L standard without any maintenance (inlet, ion source), remained below 7% for all compounds across the three cable sample matrices. Also, the %RSD was consistently below 20% for all compounds, showcasing outstanding quantitation stability even when continuously subjected to a complex extract (Figure 7). 26 www.agilent.com DE27764049 This information is subject to change without notice. © Agilent Technologies, Inc. 2024 Printed in the USA, March 18, 2024 5994-7215EN Conclusion This application note demonstrates the configuration of the Agilent 8890 GC coupled with the Agilent 5977C, using an Agilent HydroInert source with hydrogen as the carrier gas in combination with the backflush technique. In addition to fulfilling the requirements of the International Electrotechnical Commission 62321 standard for the analysis of phthalates in electrical products, it also highlights the method's robustness and addresses the issue of instability when analyzing samples with complex components. This is achieved through the self-cleaning properties of hydrogen gas within the ion source and the use of the backflush technique to remove components that are difficult to evaporate. References 1. European Union (EU) Commission Delegated Directive 2015/863 of 31 March 2015 Amending Annex II to Directive 2011/65/EU of the European Parliament and of the Council Concerning the List of Restricted Substances, C/2015/2067, OJ L 137, 4.6.2015, p. 10–12, http://data. europa.eu/eli/dir_del/2015/863/oj 2. European Union Regulation No. 1907/2006. Registration, Evaluation, Authorisation, and Restriction of Chemicals (REACH), Establishing a European Chemicals Agency. Annex XIV. https://eur-lex.europa.eu/legal-content/EN/TX T/?uri=CELEX:02006R1907-20231201 3. International Electrotechnical Commission (IEC) 62321-8, Determination of Certain Substances in Electrotechnical Products—Part 8: Phthalates in Polymers by Gas Chromatography/Mass Spectrometry (GC-MS), Gas Chromatography/Mass Spectrometry Using a Pyrolyzer/ Thermal Desorption Accessory (Py/TD-GC-MS) Int. Electrotech. Commission 2017, pp. 1–70. Figure 7. The response values at 1 mg/L in the method robustness test. 0 1 2 3 4 5 6 7 2468 10 12 14 16 18 20 22 Cable 2 ×104 DIDP DIDP DIDP DINP DINP DINP DNOP DNOP DNOP DEHP DEHP DEHP DCHP DCHP DCHP DBP DBP BBP BBP BBP DPENP DPENP DPENP DPP DPP DBP DPP DIPB DIPB DIPB BB BB BB 0 2 4 6 8 10 12 14 246 8 10 12 14 16 18 20 22 24 Cable 1 ×105 0 2 4 6 8 10 12 0246 8 10 12 14 16 Cable 3 ×104 Application Note Environmental Authors Gwen Lim Sin Yee, CTC Analytics AG, Zwingen, Switzerland Aimei Zou, Agilent Technologies, Inc. Abstract Growing demand for SVOC analysis at laboratories reveals the analysis bottlenecks, e.g., a lack of practical sample preparation experience, low sample throughput, and high consumption of chemical solvents. Online automated sample preparation is gaining attention as a solution to address these laboratory challenges. An automated workflow solution for quantitation of SVOCs in water samples, combining calibration, sample preparation, and detection was developed on Agilent gas chromatography/triple quadrupole mass spectrometer (GC/TQ) using the PAL3 robotic tool change (RTC) system in this study. Automated Sample Preparation Using the PAL3 RTC System for EPA 8270E Semivolatile Organic Analysis by GC/TQ 28 Return to Table of Contents 2 Introduction Semivolatile organic compounds (SVOCs) analysis is widely implemented at analytical laboratories. The SW846 Compendium EPA 8270 method provides guidelines on conditions and quality control (QC) checks to ensure successful analysis of SVOCs using gas chromatography/ mass spectrometer (GC/MS). Furthermore, the EPA 8270E method, which was released in 2018, includes GC/MS with an MS/MS detector. The MS/MS selectivity ensures better lower limit of quantitation (LLOQ) thus delivers reliable analysis results1 . Growing demand for SVOC analysis at laboratories reveals the analysis bottlenecks, e.g., a lack of practical sample preparation experience, low sample throughput, and high consumption of chemical solvents. Online automated sample preparation is gaining attention as a solution to address these laboratory challenges. This application note presents a proofof-concept on a novel automated sample preparation workflow modified based on EPA 3510C method² (that is a procedure for isolating organic compounds from aqueous samples), followed by Agilent gas chromatography/triple quadrupole mass spectrometry (GC/TQ) analysis according to EPA 8270E. The novel sample preparation workflow involved liquid-liquid extraction (LLE) of water samples using dichloromethane (DCM) at two specific pH conditions. The combination of these two extracts was then analyzed by the GC/TQ. Surface water samples were prepared automatically and online by the PAL3 robotic tool change (RTC) system prior to GC/TQ analysis. Likewise, working calibration standards, method blank samples, and matrix-spiked QC samples were also prepared automatically by PAL3. Then, 100 analytes of SOVCs were tested and evaluated in this study. Experimental Instrumentation An Agilent 7000 series triple quadrupole mass spectrometer was coupled to an Agilent 8890 GC with a back split/splitless inlet (SSL) and splitless inlet liner (part number 5190-2293, 900 μL, single taper, wool, Ultra). The ion source was equipped with a 9 mm diameter drawout lens (part number G3870-20449). The system was autotuned using the etune algorithm embedded in the Agilent MassHunter software version 10.1. Table 1 contains the conditions and operating parameters for both GC and MS. A PAL3 Series II RTC system (Figure 1) was used as a liquid handling platform for the calibration/sample preparation and injection onto the GC/TQ system in the study. The PAL3 system was equipped with a vortex mixer and various tray holders and racks (for 2 mL, 10/20-mL vials). Various liquid syringe tools were used fitting different volumes of PTFE coated smart syringe. Solvent module and fast wash module were also configured with the PAL3 system. GC Conditions Injection Volume 2.0 μL Column Agilent J&W DB-UI8270D, 30 m x 250 μm x 0.25 μm (part number 122-9732) Inlet Temperature 250 °C Injection Mode Pulsed splitless Carrier Gas Helium, constant flow, 1.2 mL/min Transfer Line Temperature 320 °C Oven Program 40 °C hold for 0.5 minutes 25 °C/minute to 260 °C, hold for 9.3 minutes 5 °C/minute to 280 °C, hold for 13.3 minutes 25 °C/minute to 320 °C, hold for 18.9 minutes MS Parameters Acquisition Mode dMRM Ion Source Temperature 320 °C Quadrupole Temperature 150 °C Ionization EI mode EMV Mode Gain factor (10) Solvent Delay 1.5 minutes Cycles Per Second 15 Table 1. Agilent 8890 GC and 7000 series instrument parameters. The integrated PAL3-GC/TQ system was controlled by Agilent MassHunter Workstation GC/MS Data Acquisition 10.1, offering an easy user experience with a single software system. The operating window is captured in Figure 2. Figure 1. PAL3 RTC system on an Agilent 7000 series GC/TQ. 29 3 Calibration preparation by the PAL3 system A total of 10 calibration levels were automatically prepared by the PAL3 system for the experimental work in this study. The automated procedure is illustrated in Figure 3. A stock standard solution containing 100 analytes based on the EPA 8270E target list was manually prepared at the concentration of 300 µg/mL (ppm) in DCM. Agilent semi-volatiles internal standard (part number ISM-563-1, 2000 ppm) was manually diluted by DCM at 40 ppm as ISTD. Both stock standard solution and ISTD were loaded onto the predefined vial positions according to the method parameters. As indicated in Figure 3, the 3 intermediate standard solutions were prepared from the stock standard solution by the PAL3 system, and then were used to prepare the 10 levels of working calibration standards from 0.01 to 20 ppm. Lastly, 5 μL of ISTD were then spiked into each vial of calibration standards at a final concentration of 2 ppm. Sample preparation by the PAL3 system Surface water was used as a sample to test out the performance of automated sample preparation by the PAL3 system. According to EPA 3510C, manual liquid-liquid extraction (LLE) using a separatory funnel is defined for aqueous samples, which involves large sample size and high chemical/reagents consumption. Based on the LLE described in EPA 3510C, an automated workflow was modified and developed on the PAL3 system in this study. The automated sample preparation workflow is shown in Figure 4. 1 g of NaCl was manually weighed into a 20 mL vial followed by adding 15 mL of the water sample. The vial was capped securely and placed on the sample rack (PAL R60 rack for 10/20-mL vial). The rest of the sample preparation workflow steps were then done by the PAL3 system. The analytes from the water sample were enriched 10-fold during the workflow. Figure 2. The integrated PAL3-GC/TQ system operating window by MassHunter software 10.1. PAL Parameters Figure 3. Automated preparation for calibration standards by the PAL3 system. Stock STD 300 ppm Manual Preparation Automation by PAL3 ISTD 40 ppm Intermediate STD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 Cal 1 0.01 ppm Working Calibration Standards Cal 2 0.02 ppm Cal 3 0.05 ppm Cal 4 0.1 ppm Cal 5 0.2 ppm Cal 6 0.5 ppm Cal 7 1 ppm Cal 8 5 ppm Cal 9 10 ppm Cal 10 20 ppm Figure 4. Automated sample preparation by the PAL3 system. Weigh 1 g of NaCI into vial Manual Preparation Automation by PAL3 Add 15 mL of sample Load sample vial on PAL3 Vortex mixing Vortex mixing Vortex mixing Spike 5 µL of ISTD Vortex mixing Inject 2 µL onto GC Add 50 µL of 6M NaOH Spike target analyte standard (or surrogate standard) Add 1.5 mL of DCM Add 100 µL of 50% HSO LLE Vortex extraction Phase separation 3 minutes LLE Vortex extraction Phase separation 3 minutes Transfer 50 µL lower layer extract to 250 µL vial Transfer 50 µL lower layer extract to 250 µL vial 30 4 Online analysis sequence As illustrated in Figure 5, a batch of online analysis sequence includes working calibration standards, method blank (MB), which is unspiked matrix blank, and matrix-spiked QC samples. First, 10 points of calibration standards were prepared by the PAL3 and subsequently analyzed via GC/TQ. Next, MB was prepared by the PAL3 and immediately injected into GC/TQ for quantitative analysis. In the meantime, PAL3 moved forward to the next sample preparation while GC/TQ was continually working on the analysis of MB. As a result, the integrated PAL3-GC/TQ system allowed sample preparation and sample analysis to proceed in a parallel mode. Thus, the overall lab productivity was increased through automation and eliminating waiting time between runs. Results and discussion Compound identification The acquisition method including multiple reaction monitoring (MRM) transitions, collision energy, and retention time (RT) used for this study was based on the existing well-developed method from a previous application note3 . Figure 6 shows a representative MRM chromatogram of the 100 analytes at 5 μg/mL (Cal 8) prepared by the PAL3 system. The symmetric sharp peaks demonstrate the efficient chromatographic separation of targets within the retention time window. Figure 5. Online analysis sequence on the integrated PAL3-GC/TQ system. PAL3 RTC System Preparing Intermediate STD and Calibration Standards Preparing Method Blank (MB) Preparing Matrix-Spiked QC-1 Preparing Matrix-Spiked QC-2 Preparing Matrix-Spiked QC-3 GC/TQ System Analyzing Calibration Standards Analyzing Method Blank (MB) Analyzing Matrix-Spiked QC-1 Analyzing Matrix-Spiked QC-2 Analyzing Matrix-Spiked QC-3 Analysis Timeline Figure 6. Representative MRM chromatogram for 100 analytes at 5 µg/mL (Cal 8) and ISTDs at 2 µg/mL in DCM prepared by the PAL3 system. x10 Chromatogram of 100 analytes at 5 µg/mL with ISTDs at 2 µg/mL in DCM 1 0.95 0.9 0.85 0.8 0.75 07 0.65 0.6 0.55 0.5 0.45 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 0 3 4 5 6 7 8 9 10 Counts vs. Acquisition Time (min) 11 12 13 14 15 16 17 31 5 Initial calibration performance Initial calibration (ICAL) performance was evaluated in terms of linearity, response factor (RF), and accuracy of calibration standards. The results are summarized in Table 2. The overall working range of the method for all analytes was determined to be 0.01 to 20 μg/mL, while some data points for certain compounds may be deleted at the low and high ends of the calibration range to meet the method performance criteria based on EPA 8270E. In this study, 96% of compounds achieved R>0.995 (LR mode) with minimum 5 points and 97% of compounds met the accuracy requirement for each calibration level. The %RSD of RF is within 20% for all analytes, demonstrating the excellent performance done by the integrated PAL3-GC/TQ system for automated calibration preparation and acquisition analysis. The ISTD was also assessed to determine if the method sensitivity and stability was maintained throughout the whole process. 5 μL of ISTD mixture was added to each calibration level and matrix-spiked QC to reach the final concentration of 2 μg/mL. The absolute RT change for ISTDs was within the regulatory recommendation of ≤ 30 secs. The response of all ISTDs in the individual standard was obtained within 70 to 150% of average response throughout the final calibration range, meeting the EPA performance criteria1 . Method sensitivity based on LLOQ The method sensitivity was evaluated based on the LLOQ in this work. The lowest point in the ICAL is defined as LLOQ that met the performance criteria including linearity, RF, and accuracy¹. The summary of the LLOQ for all analytes is listed in Table 2. The LLOQ of 100 analytes was distributed across 0.01 to 0.5 μg/mL as shown in Figure 7. Overall, 39 out of 100 compounds obtained LLOQ ≤ 0.02 μg/mL, demonstrating the excellent sensitivity of the method developed on the PAL3- GC/TQ. Method blanks Method blanks (MBs) must be carried out through all stages of sample preparation and analyzed for the compounds of interest as a safeguard against lab contamination caused from the sample, the reagents used, and the preparation workflow. In this study, duplicate MBs were prepared by the PAL3 system following the same method script except for the addition of analytes/surrogates. Target concentration for all compounds in MBs was obtained less than 50% of the LLOQ, although positive presence was observed for certain compounds, demonstrating that lab contamination was controlled to the desired level. Compound Name Quantifier Transition Linearity (LR Model) RT (min) RF RSD of RF LLOQ (μg/mL) Recovery (%) RSD of Recovery (n=3) 1,2,4-Trichlorobenzene 179.9 -> 109.0 0.9998 5.64 1.04 0.7% 0.01 115 4% 1,2-Dichlorobenzene 146.0 -> 111.0 0.9999 4.69 1.12 0.5% 0.01 109 5% 1,3-Dichlorobenzene 146.0 -> 111.0 0.9997 4.50 1.12 0.2% 0.01 105 5% 1,3-Dinitrobenzene 168.0 -> 75.0 0.9993 7.16 0.07 6.6% 0.2 102 12% 1,4-Dichlorobenzene 146.0 -> 111.0 0.9998 4.56 1.09 0.5% 0.01 105 8% 1,4-Dinitrobenzene 168.0 -> 75.0 0.9987 7.09 0.04 3.0% 0.2 114 4% 1-Bromo-2-nitrobenzene 156.9 -> 75.9 0.9995 6.56 0.20 0.8% 0.01 122 9% 1-Chloronaphthalene 162.0 -> 127.1 0.9988 6.88 1.45 2.7% 0.02 129 6% 1-Methylnaphthalene 142.0 -> 114.9 0.9998 6.48 1.59 0.7% 0.01 117 3% 1-Naphthylamine 143.1 -> 115.1 0.9951 7.70 0.29 4.3% 0.2 51 30% 2,2'-oxybis[1-chloropropane] 121.0 -> 77.0 0.9998 4.77 0.05 1.7% 0.02 108 16% 2,3,4,6-Tetrachlorophenol 230.0 -> 165.9 0.9980 7.72 0.08 7.5% 0.5 64 3% 2,3,5,6-Tetrachlorophenol 230.0 -> 165.9 0.9988 7.68 0.07 0.6% 0.5 60 4% 2,4,5-Trichlorophenol 195.8 -> 97.0 0.9995 6.70 0.35 10.2% 0.5 63 7% 2,4,6-Trichlorophenol 195.8 -> 97.0 0.9985 6.66 0.46 1.8% 0.05 64 6% 2,4-Dichlorophenol 162.0 -> 63.0 0.9996 5.57 0.98 2.0% 0.01 59 11% 2,4-Dimethylphenol 107.1 -> 77.1 1.0000 5.36 0.99 1.6% 0.01 67 14% 2,4-Dinitrophenol 184.0 -> 79.0 0.9931 7.50 0.01 3.6% 0.5 53 2% 2,4-Dinitrotoluene 165.0 -> 63.0 0.9998 7.59 0.09 3.8% 0.5 100 9% 2,6-Dinitrotoluene 165.0 -> 63.0 0.9990 7.19 0.11 1.4% 0.2 121 9% Table 2. Analytical performance summary for analytes. 32 6 Compound Name Quantifier Transition Linearity (LR Model) RT (min) RF RSD of RF LLOQ (μg/mL) Recovery (%) RSD of Recovery (n=3) 2-Acetylaminofluorene 222.9 -> 181.1 0.9966 11.82 0.07 6.7% 0.5 119 5% 2-Chloronaphthalene 162.0 -> 126.9 0.9992 6.86 2.48 1.8% 0.05 118 5% 2-Chlorophenol 128.0 -> 64.0 0.9998 4.37 0.35 2.0% 0.01 54 3% 2-methyl-4,6-dinitrophenol 198.0 -> 121.0 0.9949 7.99 0.03 3.6% 0.5 53 7% 2-Methylnaphthalene 142.0 -> 141.0 0.9984 6.39 2.81 0.2% 0.01 117 4% 2-Nitroaniline 138.0 -> 92.0 0.9988 6.96 0.14 1.4% 0.5 104 9% 2-Nitrophenol 138.9 -> 81.0 0.9987 5.34 0.25 1.2% 0.01 61 9% 2-Picoline 93.1 -> 66.0 0.9997 3.20 0.28 2.3% 0.01 40 5% 3-Methylcholanthrene 268.1 -> 252.1 0.9997 15.61 0.81 1.2% 0.5 101 7% 4,4'-DDD 234.8 -> 164.9 0.9986 11.04 1.46 1.5% 0.1 123 6% 4,4'-DDE 245.8 -> 176.0 0.9984 10.56 1.15 1.3% 0.1 116 6% 4,4'-DDT 234.8 -> 164.9 0.9985 11.51 0.89 1.1% 0.05 99 8% 4-Aminobiphenyl 168.1 -> 167.1 0.9986 8.68 0.21 8.6% 0.5 59 11% 4-bromophenyl phenyl ether 248.0 -> 141.0 0.9994 8.44 0.52 1.9% 0.1 112 6% 4-chloro-3-methylphenol 107.0 -> 77.0 0.9998 6.23 0.61 1.1% 0.05 55 13% 4-Chloroaniline 127.0 -> 65.0 0.9991 5.77 0.48 2% 0.02 19 14% 4-Chlorophenyl phenyl ether 141.1 -> 115.1 0.9968 7.94 0.45 0.7% 0.02 123 2% 4-Nitroaniline 138.0 -> 108.1 0.9996 7.97 0.14 6.4% 0.5 76 8% 7,12-Dimethylbenz[a]anthracene 256.1 -> 241.1 0.9998 14.45 1.51 1.5% 0.1 121 5% Acenaphthene 152.9 -> 77.0 0.9997 7.44 0.17 0.6% 0.01 113 6% Acenaphthylene 151.9 -> 102.0 0.9998 7.27 0.17 0.3% 0.01 114 8% Aldrin 262.7 -> 192.6 0.9997 9.69 0.15 1.4% 0.01 105 5% Aniline 93.0 -> 66.0 0.9999 4.27 0.68 0.9% 0.01 23 15% Anthracene 177.9 -> 152.0 0.9958 8.94 0.93 4.5% 0.2 118 5% Azobenzene 77.0 -> 51.0 0.9975 8.10 1.42 2.3% 0.2 119 7% Benz[a]anthracene 228.1 -> 226.1 1.0000 12.36 1.60 1.2% 0.5 122 7% Benzo[a]pyrene 252.1 -> 250.1 0.9998 15.04 1.79 3.5% 0.1 113 5% Benzo[b]fluoranthene 252.1 -> 250.1 0.9997 14.46 2.20 1.4% 0.5 120 4% Benzo[g,h,i]perylene 276.1 -> 274.1 0.9998 17.39 1.67 3.8% 0.5 112 5% Benzo[k]fluoranthene 252.1 -> 250.1 0.9979 14.47 1.80 0.1% 0.5 114 5% Benzyl alcohol 108.0 -> 79.0 0.9999 4.65 0.50 0.8% 0.02 59 22% BHC-alpha 180.8 -> 144.9 0.9995 8.44 0.50 2.3% 0.02 109 7% BHC-beta 180.8 -> 144.9 0.9994 8.65 0.38 1.7% 0.1 117 6% BHC-delta 218.8 -> 182.8 0.9936 8.97 0.39 1.5% 0.1 106 6% BHC-gamma 218.8 -> 182.9 0.9991 8.74 0.35 3.0% 0.1 112 8% bis(2-Chloroethoxy)methane 93.0 -> 63.0 0.9998 5.46 1.75 1.2% 0.01 112 10% bis(2-Chloroethyl)ether 93.1 -> 63.0 0.9999 4.31 0.86 0.5% 0.01 101 19% Bis(2-ethylhexyl) phthalate 149.0 -> 65.0 0.9997 12.44 1.40 1.7% 0.02 124 5% Butyl benzyl phthalate 149.0 -> 65.0 1.0000 11.38 0.91 2.4% 0.05 128 6% Chrysene 226.1 -> 224.1 0.9985 12.38 0.73 7.8% 0.1 111 8% Dibenz[a,h]anthracene 278.1 -> 276.1 0.9995 16.88 0.86 4.6% 0.5 104 6% Dibenzofuran 167.9 -> 139.1 0.9959 7.61 1.53 0.2% 0.5 116 7% Dieldrin 262.9 -> 193.0 1.0000 10.70 0.14 1.9% 0.05 113 6% 33 7 Compound Name Quantifier Transition Linearity (LR Model) RT (min) RF RSD of RF LLOQ (μg/mL) Recovery (%) RSD of Recovery (n=3) Diethyl phthalate 149.0 -> 65.0 0.9997 7.83 1.03 0.5% 0.1 114 8% Dimethyl phthalate 163.0 -> 77.0 0.9999 7.13 0.96 0.4% 0.1 105 11% Di-n-butyl phthalate 149.0 -> 65.0 0.9923 9.46 3.15 1.5% 0.1 125 5% Di-n-octyl phthalate 149.0 -> 65.0 0.9993 13.90 1.94 2.7% 0.1 133 5% Diphenylamine 167.0 -> 166.2 0.9985 8.06 0.80 3.1% 0.2 115 8% Endosulfan I 241.0 -> 206.0 0.9997 10.41 0.08 2.0% 0.1 121 8% Endosulfan II 240.7 -> 205.9 0.9999 11.05 0.05 1.0% 0.05 115 6% Endosulfan sulfate 271.6 -> 236.7 0.9963 11.52 0.21 1.6% 0.02 113 8% Endrin 262.7 -> 190.5 0.9995 10.94 0.03 0.7% 0.05 94 12% Ethyl methanesulfonate 109.0 -> 78.9 0.9999 3.91 0.26 1.7% 0.01 79 5% Fluoranthene 200.9 -> 199.9 0.9996 10.17 0.62 2.3% 0.1 121 7% Fluorene 166.0 -> 165.1 0.9954 7.95 1.80 0.8% 0.1 121 6% Heptachlor 273.6 -> 238.7 0.9999 9.37 0.19 0.7% 0.02 104 5% Heptachlor epoxide 352.7 -> 216.7 0.9996 10.03 0.03 1.8% 0.05 112 7% Hexachlorobenzene 283.7 -> 213.8 0.9997 8.49 0.51 2.5% 0.1 110 5% Hexachlorobutadiene 224.7 -> 189.9 0.9998 5.83 1.26 0.9% 0.01 112 1% Hexachlorocyclopentadiene 236.7 -> 143.0 0.9987 6.53 0.12 0.5% 0.05 87 4% Hexachloroethane 200.9 -> 165.9 1.0000 4.99 0.88 1.0% 0.01 105 3% Isophorone 82.0 -> 54.0 1.0000 5.27 0.82 1.4% 0.01 112 11% Methoxychlor 226.9 -> 211.9 0.9995 12.24 0.23 0.5% 0.05 103 8% Methyl methanesulfonate 80.0 -> 64.9 0.9999 3.91 0.05 2.0% 0.01 80 5% Naphthalene 128.1 -> 102.1 0.9998 5.72 1.37 0.6% 0.01 113 5% N-Nitro-o-toluidine 152.0 -> 106.0 0.9993 7.96 0.10 6.9% 0.5 71 10% N-Nitrosodiethylamine 102.0 -> 85.0 0.9999 3.70 0.07 2.8% 0.01 97 5% N-Nitrosodi-n-butylamine 84.1 -> 56.0 0.9998 6.08 0.14 0.2% 0.02 113 12% N-Nitrosodi-n-propylamine 113.1 -> 71.0 0.9998 4.88 0.05 2.1% 0.02 109 12% N-Nitrosomethylethylamine 88.0 -> 42.0 0.9999 3.25 0.11 2.3% 0.01 66 5% N-Nitrosomorpholine 116.0 -> 86.0 0.9999 4.90 0.10 2.1% 0.05 55 13% N-Nitrosopiperidine 114.0 -> 84.1 0.9998 5.19 0.14 2.8% 0.02 106 7% N-Nitrosopyrrolidine 100.1 -> 55.1 0.9996 4.87 0.07 0.7% 0.05 69 13% p-Dimethylaminoazobenzene 225.1 -> 120.1 0.9995 10.82 0.26 2.2% 0.5 144 6% Pentachloronitrobenzene 248.8 -> 213.8 0.9997 8.70 0.17 1.2% 0.1 105 6% Phenanthrene 177.9 -> 152.0 0.9985 8.91 1.32 2.6% 0.2 119 3% Phenol 94.0 -> 66.1 0.9996 4.21 0.50 0.5% 0.01 19 18% Pronamide 173.0 -> 145.0 0.9976 8.73 1.07 1.2% 0.5 116 6% Pyrene 201.1 -> 200.0 0.9997 10.45 0.84 1.2% 0.2 119 7% Thionazin 143.0 -> 79.0 0.9997 7.91 0.13 2.8% 0.1 118 8% 1,4-Dichlorobenzene-d4 (ISTD) 149.9 -> 114.9 N.A. 4.55 N.A. N.A. N.A. N.A. N.A. Acenaphthene-d10 (ISTD) 161.9 -> 159.9 N.A. 7.40 N.A. N.A. N.A. N.A. N.A. Chrysene-d12 (ISTD) 240.0 -> 235.9 N.A. 12.36 N.A. N.A. N.A. N.A. N.A. Naphthalene-d8 (ISTD) 135.9 -> 107.9 N.A. 5.70 N.A. N.A. N.A. N.A. N.A. Perylene-d12 (ISTD) 263.9 -> 259.9 N.A. 15.04 N.A. N.A. N.A. N.A. N.A. Phenanthrene-d10 (ISTD) 187.9 -> 160.0 N.A. 8.88 N.A. N.A. N.A. N.A. N.A. 34 8 27 (0.01 ppm) 12 (0.02 ppm) 13 (0.05 ppm) 19 (0.1 ppm) 9 (0.2 ppm) 20 (0.5 ppm) Figure 7. LLOQ distribution of 100 compounds. Matrix-spiked QC recovery Three technical replicates of matrix-spiked QC (n=3, 2 μg/mL in the final extract) were prepared by the PAL3 system in order to evaluate the reproducibility and robustness of the automated sample preparation. Each QC was analyzed by GC/TQ in duplicates account for the homogeneity of the QC solution and the repeatability of spiked recovery. The recovery values and %RSD are summarized in Table 2. Overall, 96% of compounds met recovery 50 to150%, and 98% of compounds obtained RSD of recovery ≤20% as shown in Figure 8A and 8B, respectively. The obtained results indicate that this automated protocol developed on the PAL3-GC/TQ is suitable, offering good reproducibility and robustness for SVOC analysis according to EPA 8270E. 0% 5% 10% 15% 20% 25% 30% 35% 40% 0 20 40 60 80 100 RSD of Recovery/% Target 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 0 20 40 60 80 100 Matrix spiked QC Recovery/% Target 8A 8B Figure 8. Matrix-spiked QC recovery (A) at 2 μg/mL in the final extract and %RSD of recovery (B). A B 35 Conclusion An automated workflow solution for quantitation of SVOCs in water samples, combining calibration/sample preparation and detection was developed on Agilent gas chromatography/ triple quadrupole mass spectrometer (GC/TQ) using the PAL3 robotic tool change (RTC) system in this study. The analytical performance parameters were evaluated based on EPA 8270E, meeting acceptance criteria for more than 90 out of 100 compounds. The PAL3 system provides various tools and modules enabling the automated preparation of calibration standards and samples to meet diverse customer needs, resulting in less manual work for the user. Agilent 7000 series triple quadrupole mass spectrometer coupled to 8890 GC offers excellent selectivity and sensitivity to target analytes. This newly developed automated workflow on the integrated PAL3-GC/TQ system offers an easy to use and more environmentally friendly solution for users by reducing chemicals/standards consumption as well as waste. This automated solution will enhance lab productivity and reduce costs significantly. References 1. Method EPA 8270E: Semivolatile Organic Compounds by Gas Chromatography/Mass Spectroscopy, Revision 6, June 2018. 2. EPA Method 3510C: Separatory Funnel Liquid-Liquid Extraction 3. A Fast Method for EPA 8270 in MRM Mode Using the 7000 Series Triple Quadrupole GC/MS, Agilent Technologies application note, 5994-0691EN, 2019 www.agilent.com DE14313123 This information is subject to change without notice. © Agilent Technologies, Inc. 2024 Published in the USA, February 1, 2024 5994-7138EN Application Note Environmental Authors Vanessa Abercrombie, Frans Biermans, Anastasia Andrianova, Joel Ferrer, and Ashlee Gerardi Agilent Technologies, Inc. Abstract As ionization sources continue to advance, lowering limits of detection and increasing confidence in analyte identification, column technologies can also be used in conjunction to push the practical limits of sensitivity and data accuracy. Increases in the detection of analyte response also result in unwanted increases in the detection of background noise. Gas chromatography/mass spectrometry (GC/MS) column technology that lowers interfering column bleed ions and elevated bleed baselines, maintains peak shape for active compounds, and can withstand aggressive thermal cycling can greatly enhance the performance and productivity of GC/MS methods. This application note examines how column attributes like bleed, inertness, and thermal stability can further benefit the sensitivity and accuracy of an MS. This study illustrates achievable data parameters, like sensitivity limits at trace levels, retention time consistencies, and data accuracy for active semivolatile organic compounds (SVOCs), when the Agilent 7010D triple quadrupole GC/MS (GC/TQ) system is used. Novel Column Chemistry Raises the Bar on Sensitivity and Data Accuracy in the Analysis of Semivolatile Organic Compounds Return to Table of Contents 37 2 Introduction Governmental regulatory authorities have established method and performance criteria for GC/MS measurement of SVOCs that are identified as pollutants in environmental and industrial matrices. The United States Environmental Protection Agency (U.S. EPA) method 8270, for example, contains a list of over 200 compounds, some of which can be susceptible to unwanted chemical activity in the instrument flow path, resulting in data quality degradation. If the performance criteria of method 8270 are not met, system maintenance is often needed, such as liner replacement followed by column trimming or replacement, resulting in unplanned instrument downtime . Monitoring of the DFTPP tuning standard, which contains 4,4'-DDT, pentachlorophenol, and benzidine, validates the suitability of the flow path and monitors when maintenance should be performed. The breakdown of 4,4'-DDT to 4,4'-DDE, and 4,4'-DDD, as well as the tailing factors of benzidine and pentachlorophenol, tests the flow path inertness, indicating the activity of susceptible acidic and basic analytes. GC columns contribute the largest surface area in the sample flow path and, therefore, are a critical factor in controlling interferences in the analytical path. Agilent Ultra Inert (UI) GC liners, along with an inert GC column phase, can improve the robustness of SVOCs analyses.1 Stationary phases used in GC/MS analysis of SVOCs are typically comprised of liquid polymers with a polysiloxane backbone. When heat is applied to the column during routine use, the terminal end of the stationary phase polymer can bend back and attack itself; this is called "backbiting." Ring structures, which are thermodynamically stable, are liberated from the stationary phase, increasing background noise and raising the baseline; this can be problematic for low signal-to-noise (S/N) analytes. Peak integration can become less repeatable, lowering quantitation accuracy. In addition, the increase in freed ring structures—which fragment in the ion source—and analytes can cause spectral interference in extracted mass spectra and decrease the qualitative score of a library spectral hit. The Agilent J&W HP-5Q and DB-5Q GC columns have an increased thermal stability at upper temperature limits, allowing for less spectral interference, lower levels of column bleed, and better data quality, especially for heavier analytes that may suffer from issues with lower S/N.2 The new Agilent high-efficiency ion source (HES) 2.0, as seen in Figures 1 and 2, is equipped with a novel dipolar RF lens that redirects carrier gas and low mass ions by > 95%. The deflected ions land on adjacent lenses and are pumped out before entering the mass analyzer, providing reduced noise and extended instrument robustness while maintaining sensitivity. A ramped RF amplitude versus mass is implemented to avoid spectrum tilt. The reduction of noise allows for a further increase of sensitivity to attogram-level detection limits. Built-in intelligence features such as SWARM autotune and early maintenance feedback further enhance instrument performance and diagnostic capabilities. The DB-5Q GC column and the HES 2.0 work in concert to increase the robustness of difficult analyses such as that of SVOCs.3,4 Figure 1. Front view of the Agilent HES 2.0. Figure 2. Side view of the Agilent HES 2.0. 38 3 Experimental The Agilent 8000 Series semivolatiles standard (part number SVM-8270-1)—a representative mixture of semivolatile acids, bases, and neutrals—was prepared in dichloromethane (DCM) for calibration standards to be analyzed at 10 to 1,000 pg on column. A semivolatile internal standard mix (part number CRM48902) was procured from Sigma-Aldrich (Saint Louis, MO, U.S.). The tuning standard, containing a mixture of benzidine, pentachlorophenol, 4,4'-diphenyltrichloroethane (4,4'-DDT), and decafluorodiphenyltrichloroethane (DFTPP) at 25 µg/mL, was used to obtain MS calibration and tuning settings. A composite mixture of soils extracted with DCM prepared for method 8270 analysis, which is a representative matrix residue that is typically encountered in the lab, was procured from Pace Analytical (Mt. Juliet, TN, U.S.). An Agilent 8890 GC coupled with an Agilent 5977B GC/MSD and Inert Extractor source, as well as an 8890 GC coupled with a 7010D triple quadrupole GC/MS (GC/TQ) system, upgraded with the HES 2.0, were used for the analysis. Parameter Value Agilent 8890 GC Inlet 300 °C, Pulsed splitless mode Injection Volume 0.5 mL Inlet Liner Agilent UI inlet liner, split, low pressure drop (p/n 5190-2295) Injection Pulse Pressure 30 psi until 0.6 min Purge Flow to Split Vent 50 mL/min at 0.6 min Septum Purge Flow 3 mL/min Oven 40 °C (0.5 min), ramp 10 °C/min to 100 °C, ramp 25 °C/min to 260 °C, ramp 5 °C/min to 280 °C, 15 °C/min to 320 °C (5 min), 10 °C/min to 330 °C (10 min), 10 °C/min to 340 °C (10 min) Column Carrier Gas Helium, 1.3 mL/min, constant flow Column – Agilent J&W DB-5Q, 30 m × 0.25 mm, 0.25 µm (p/n 122-5532Q) – Agilent J&W DB-5ms UI, 30 m × 0.25 mm, 0.25 µm (p/n 122-5532UI) – 5ms-Type column X, 30 m × 0.25 mm, 0.25 µm – 5ms-Type column Y, 30 m × 0.25 mm, 0.25 µm Inlet Connection Split/splitless inlet Outlet Connection MSD Table 1. GC parameters for the Agilent 8890 GC. Parameter Value Agilent 5977B GC/MSD Source Agilent Inert Extractor source Mode Scan (35 to 500 amu) Solvent Delay 2.5 min Source Temperature 300 °C Quadrupole Temperature 175 °C Gain 1.0 Table 2. MS parameters for the Agilent 5977B GC/MSD. Parameter Value Agilent 7010D GC/TQ Source Agilent HES Mode Dynamic multiple reaction monitoring (dMRM)/scan Solvent Delay 2.5 min Source Temperature 300 °C Quadrupole Temperature 175 °C Gain 1.0 Table 3. MS parameters for the Agilent 7010D GC/TQ. 39 4 Peak SVM-8270-1 Quantitative Qualitative RT Compound Precursor Ion Product Ion Dwell CE Precursor Ion Product Ion Dwell CE 1 3.043 NDMA 74 44 56.07 6 74 42 56.07 14 2 6.411 Phenol 94 66.1 14.17 15 94 65.1 14.17 20 3 6.670 Chlorophenol-2 128 64 9.22 15 128 63 9.22 30 4 6.940 1, 3-Dichlorobenzene 146 111 8.33 15 146 75 8.33 30 4.5 6.960 1, 4-Dichlorobenzene-d4 150 115 9.31 15 150 78 9.31 3 5 7.074 1, 4-Dichlorobenzene 146 111 13.21 15 146 75 13.21 30 6 7.313 1, 2-Dichlorobenzene 146 111 9.63 15 146 75 9.63 30 7 7.700 Nitrosodi-n-propylamine N- 113.1 71 7.14 10 101 70 7.14 0 8 7.450 Methylphenol-2 (Cresol o-) 108 107 7.96 15 107 77 7.96 15 9 7.500 bis(2-Chloro-1-methylethyl)ether 121 77 7.68 5 121 49 7.68 30 10 7.700 Methylphenol-4 (Cresol p-) 108 107.1 8.36 15 107 77.1 8.36 15 11 7.800 Hexachloroethane 200.9 165.9 7.19 15 118.9 83.9 7.19 35 12 7.900 Nitrobenzene 123 77 8.28 10 77 51 8.28 15 13 8.250 Isophorone 138 82 10.46 5 82 54 10.46 5 14 8.370 Nitrophenol, 2- 138.9 81 11 15 109 81 11 10 15 8.450 Dimethylphenol 2,4- (2, 4-xylenol) 122.1 107 12.24 10 107.1 77.1 12.24 15 16 8.600 bis(2-Chloroethoxy)methane 95 65 10.37 5 93 63 10.37 5 17 8.700 Dichlorophenol, 2,4- 163.9 63 9.65 30 162 63 9.65 30 18 8.800 Trichlorobenzene, 1,2,4- 179.9 145 10.32 15 179.9 109 10.32 30 18.5 8.805 Naphthalene-d8 136.1 108.1 9.59 20 136.1 84.1 9.59 25 19 8.900 Naphthalene 128.1 102.1 12.38 20 128.1 78.1 12.38 20 20 8.980 Chloroaniline,4- 127 92 13.18 15 127 65 13.18 20 21 9.070 Hexachlorobutadiene 226.9 191.9 28.71 15 224.8 189.9 28.71 1 22 9.570 Phenol 4-chloro-3-methyl- 142 107 34.24 15 107 77 34.24 15 23 9.750 Methylnaphthalene, 2- 142.1 141.1 22.34 15 141.1 115.1 22.34 15 24 9.940 Hexachlorocyclopentadiene 237 143 17.85 20 237 119 17.85 20 25 10.060 Trichlorophenol, 2,4,5- 197.9 97 16.66 25 195.9 97 16.66 25 26 10.100 Trichlorophenol, 2,4 6- 198 97 17.45 30 196 97 17.45 30 27 10.300 Chloronaphthalene, 2- 162 127.1 21.04 20 162 77 21.04 35 28 10.400 Nitroaniline, 2- 138 92 25.89 15 138 65 25.89 25 29 10.600 Dimethyl phthalate 163 92 23.89 30 163 77 23.89 20 30 10.670 Dinitrotoluene, 2,6- 165 90.1 20.02 15 165 63 20.02 25 31 10.740 Acenaphthylene 152.1 102.1 14.15 30 151.1 77 14.15 25 32 10.840 Nitroaniline, 3- 138 92 11.18 15 138 80 11.18 5 32.5 10.826 Acenaphthene-d10 164.1 162.1 10.43 15 162.1 160.1 10.43 20 33 10.910 Acenaphthene 154.1 127 10.43 40 153.1 77 10.43 45 34 10.950 Phenol, 2,4-dinitro- 184 107 12.76 25 184 79 12.76 25 35 11.010 Nitrophenol, 4- 138.9 109 12.71 5 109 81 12.71 10 36 11.120 Dibenzofuran 168.1 139.1 13.47 25 139.1 63 13.47 35 37 11.080 Dinitrotoluene, 2,4- 165 119 12.55 5 165 63 12.55 45 38 11.340 Diethyl phthalate 149 93 12.42 15 149 65 12.42 20 39 11.460 Fluorene 166.1 165.1 10.14 15 165.1 163.1 10.14 35 40 11.470 Chlorophenyl phenyl ether, 4- 204 77 10.34 30 141.1 115.1 10.34 20 Table 4. Quantitative/qualitative transitions (dMRM based) for Agilent 7010D GC/TQ acquisition parameters. 40 5 Peak SVM-8270-1 Quantitative Qualitative RT Compound Precursor Ion Product Ion Dwell CE Precursor Ion Product Ion Dwell CE 41 11.480 Nitroaniline, 4- 138 108.1 10.1 5 108 80 10.1 15 42 11.510 DNOC (2-methyl-4 6-dinitrophenol) 198 167.9 13.44 5 198 121 13.44 10 43 11.630 Azobenzene 105 77.1 11.37 5 77 51 11.37 15 44 11.970 4-Bromophenyl phenyl ether 250 141 22.75 20 248 141 22.75 20 45 12.020 Hexachlorobenzene 283.8 213.9 28.52 30 248.9 214 28.52 15 46 12.220 Pentachlorophenol 265.9 167 30.92 25 165 130 30.92 25 47 12.430 Phenanthrene 178.1 152.1 22.36 25 176.1 150.1 22.36 25 47.5 12.360 Phenanthrene-d10 188.3 160.2 18.93 20 188.3 158.2 18.93 35 48 12.490 Anthracene 178.1 152.1 18.72 25 178.1 151.1 18.72 30 49 12.650 Carbazole 167 139 33.75 45 167 89 33.75 60 50 13.000 Di-n-butyl phthalate 149 121 74.98 15 149 65 74.98 25 51 13.690 Fluoranthene 202.1 152.1 26.52 30 201.1 200.1 26.52 15 52 13.980 Pyrene 202.1 151 21.27 45 201.1 200 21.27 15 53 14.900 Butyl benzyl phthalate 149 65 55.86 25 91 65 55.86 15 54 15.860 Benz[a]anthracene 228.1 226.1 23.12 30 226.1 224.1 23.12 35 54.5 15.842 Chrysene-d12 240.2 236.2 16.17 3 236.1 232.1 16.17 40 55 15.930 Chrysene 226.1 224.1 23.59 40 113.1 112.1 23.59 10 56 16.000 bis(2-Ethylhexyl) phthalate 167 149 23.09 5 149 65 23.09 25 57 17.450 Di-n-octyl phthalate 149 93 27.51 20 149 65 27.51 25 58 18.050 Benzo[b]fluoranthene 252.1 250.1 18.69 35 126 113.1 18.69 10 59 18.150 Benzo[k]fluoranthene 252.1 250.1 18.56 30 126.1 113.1 18.56 10 60 18.700 Benzo[a]pyrene 252.1 250.1 21.83 35 125 124.1 21.83 10 60.5 18.754 Perylene-d12 264.2 260.1 16.16 35 260.1 256.1 16.16 40 61 20.580 Indeno[1,2,3-cd]pyrene 276.1 274.1 30.66 40 137 136 30.66 15 62 20.660 Dibenz[a,h]anthracene 278.1 276.1 24.37 35 276.1 274.1 24.37 35 63 21.080 Benzo[g,h,i]perylene 276.1 274.1 45.33 45 138 137 45.33 1 Results and discussion Reduced bleed stabilizes GC/MS baselines According to EPA method 8270, the GC/MS system must meet the required performance criteria to confirm suitability before samples can be analyzed. The system suitability results, along with the chromatographic resolution criteria of closely eluting structural isomer pairs, have been described previously for an Agilent UI glass wool liner and UI 5ms-type column.1 The chromatographic performance of the J&W DB-5Q GC column was tested for the suitability of method 8270 by GC/MS, and a representative chromatogram is demonstrated in Figure 3. When comparing the J&W DB-5Q to a conventional 5ms-type GC column, a significant decrease in column bleed stabilized the chromatographic baseline, which is ideal when analyzing low concentrations of compounds at high temperatures (Figure 4). 41 6 Figure 3. A representative chromatogram of 8,270 compounds analyzed on an Agilent J&W DB-5Q GC column and collected using an Agilent 5977B GC/MSD. Acquisition time (min) 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 0 20 40 60 80 100 10 30 50 70 90 Counts (%) Figure 4. A standard of 50 pg on column of 8,270 compounds, analyzed on an Agilent J&W DB-5Q (blue) and the 5ms-type column Y (red), and collected on an Agilent 5977B GC/MSD. 17 18 19 20 21 22 23 24 1 3 5 7 9 11 2 4 6 8 10 Acquisition time (min) Counts (%) 42 7 Reduced column activity improves peak symmetry of problematic analytes In EPA method 8270, column activity is observed as increased peak tailing, resulting in S/N loss. This is commonly observed with the more active compounds in the analyte panel. In Figure 5, the peak symmetry and S/N ratio for 2,4-dinitrophenol, a problematic analyte, was compared at 250 pg on column, analyzed on both a DB-5Q column and a conventional 5ms-type column. The column activity of the 5ms-type column X resulted in a loss of peak symmetry that significantly decreased analyte sensitivity. Also, the peak shape of another problematic compound, 2-methyl4,6-dinitrophenol, was compared through analysis on two conventional 5ms-type columns (marked as X and Y) and the DB-5Q, as shown in Figure 6. The inertness of the DB-5Q allowed for better S/N, leading to better sensitivity for this difficult phenolic compound. Lastly, Figure 7 demonstrates the comparison of pentachlorophenol on the DB-5Q and a conventional 5ms-type column. Again, when analyzed at the same concentration, the tailing factor of pentachlorophenol increased, leading to less S/N response. When working with difficult analytes, such as those in EPA method 8270, inert column chemistries, such as those observed on the DB-5Q, will optimize sensitivity. Figure 5. Integrated peak 250 pg on column 2,4-dinitrophenol, analyzed on the 5ms-type column X and an Agilent J&W DB-5Q GC column. Symmetry: 2.3 S/N: 715.2 Acquisition time (min) 10.7 10.8 10.9 11.0 11.1 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Agilent J&W DB-5Q Acquisition time (min) 10.6 10.7 10.8 10.9 11.0 0 1 2 3 4 5 6 7 8 9 5ms-Type column X Symmetry: 4.47 S/N: 8.83 ×102 ×103 Counts (%) Counts (%) A B 43 8 Figure 6. Integrated peak 250 pg on-column 2-methyl-4,6-dinitrophenol, analyzed on the 5ms-type columns X and Y, and an Agilent J&W DB-5Q GC column. ×104 Acquisition time (min) 11.3 11.4 11.5 11.6 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 Symmetry: 1.7 S/N: 310.5 Agilent J&W DB-5Q C Counts ×103 Acquisition time (min) 11.2 11.3 11.4 11.5 11.6 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 Symmetry: 6.9 S/N: 8.3 5ms-Type column Y B Counts ×103 Acquisition time (min) 11.2 11.3 11.4 11.5 11.6 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Symmetry: 2.4 5ms-Type column X S/N: 212.9 2-Methyl-4,6-dinitrophenol A Counts Figure 7. Integrated peak 250 pg on-column pentachlorophenol, analyzed the 5ms-type column Y and an Agilent J&W DB-5Q GC column. Pentachlorophenol Acquisition time (min) 12.0 12.1 12.2 12.3 12.4 0 0.5 1.0 1.5 2.0 2.5 3.0 Symmetry: 2.1 S/N: 245.0 Agilent J&W DB-5Q Acquisition time (min) 11.9 12.0 12.1 12.2 12.3 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 Symmetry: 3.2 S/N: 82.8 5ms-Type column Y ×104 ×104 A Counts Counts B 44 9 Improved durability leads to consistent chromatography To stress the robustness of the GC system in a real-world application, a heavy soil matrix was diluted in DCM and analyzed over multiple thermal cycles. DFTPP tuning standard was analyzed every five matrix injections, using %DDT breakdown as an indicator to replace the inlet liner. The inlet liner and septum were replaced every 20 matrix injections, as they failed the method criteria after %DDT breakdown. The peak shapes of pentachlorophenol and benzidine were used as indicators of increased column activity resulting from matrix accumulation and thermal degradation. Figure 8 demonstrates that after 200 matrix injections, the retention times and peak shapes for all test compounds were consistent with minimal data quality reduction. The DB-5Q column is durable to withstand routine, high-throughput cycling shape, even when working with difficult matrices such as soil extracts. Figure 8. DFTPP tune mix initially (black) and after 200 matrix injections (blue) on an Agilent J&W DB-5Q column. Acquisition time (min) 8 9 10 11 12 13 14 15 16 1 2 3 4 Tune mix after 200 matrix injections Tune mix initially Peak Compound 1 Pentacholophenol 2 DFTPP 3 Benzidine 4 DDT 45 10 Optimal sensitivity and acquisition flexibility The improved sensitivity of the HES 2.0 ion source, with the triple off-axis detector configuration, allows for fast MRM speeds. Data can now be acquired in dMRM and scan modes simultaneously. This improvement allows for targeted and untargeted analyses at the same time. The dMRM is useful in setting up MRM methods, as once the retention time is inputted into Agilent MassHunter acquisition software, the dwell time is calculated automatically. While this streamlines the method setup process, if retention times shift from matrix accumulation or thermal instability, it can be helpful to collect scan data and dMRM in the same acquisition method. In Figures 9 and 10, a 10 pg on-column standard was analyzed using dMRM/scan collection mode. Figure 9 demonstrates the extracted scan chromatogram, which is zoomed-in on later-eluting compounds to display their S/N ratios. With the combination of the improved HES 2.0 ion source, along with the improved thermal stability of the DB-5Q column, it is possible to simultaneously acquire selective methods, such as dMRM and scan mode. Selectivity matching eases column adoption A standard EPA method 8270 analysis was conducted at 1,000 pg on-column, using the same instrumentation and method conditions, on a DB-5ms UI and a DB-5Q column. The similar selectivity allows for upgrading analytical methods without the need for more development, as seen in Figure 11. Also, with the same selectivity, there is no need to update retention times, which makes the DB-5Q compatible with existing retention time locking libraries. Figure 9. A standard of SVOCs analyzed at 10 pg on column, collected by dMRM/scan mode with the extracted scan data. 20.5 20.6 20.7 20.8 20.9 21.0 21.1 21.2 m/z = 276 m/z = 278 m/z = 279 m/z = 274 61 62 63 Acquisition time (min) Acquisition time (min) 345678 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 10 30 50 70 90 0 20 40 60 80 10 pg On-column of SVM-8270-1, collected in scan mode Counts (%) Peak Compound S/N 61 Indeno(1,2,3-cd)pyrene 14.53 62 Dibenz[a,h]anthracene 11.03 63 Benzo[g,h,i]perylene 25.72 46 11 Figure 11. The Agilent J&W DB-5Q has similar selectivity to an Agilent J&W DB-5ms UI, as demonstrated in the analysis of 8,270 compounds. Agilent DB-5ms UI Agilent DB-5Q Figure 10. A standard of SVOCs analyzed at 10 pg on column, collected by dMRM/scan mode with the extracted MRMs. Acquisition time (min) 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 0 10 20 30 40 50 60 70 80 90 100 Counts (%) 61 62 63 Acquisition time (min) 20.3 20.4 20.5 20.6 20.7 20.8 20.9 21.0 21.1 21.2 Peak Compound S/N Ratio 61 Indeno(1,2,3-cd)pyrene 157.8 62 Dibenz[a,h]anthracene 81.9 63 Benzo[g,h,i]perylene 208.5 47 www.agilent.com DE-000135 This information is subject to change without notice. © Agilent Technologies, Inc. 2024 Printed in the USA, August 14, 2024 5994-7686EN Conclusion This application note demonstrates that the Agilent J&W DB-5Q GC column can exceed the performance requirements of EPA method 8270. Agilent Ultra Inert chemistry across the sample flow path will maintain peak symmetry of problematic analytes, leading to improved limits of detection and accurate integration. Ultralow-bleed chemistry stabilizes baselines and reduces interfering bleed ions. High-temperature stability allows for the repeated temperature cycling needed for high throughput methods, even when analyzing heavy, complex soil matrices. The matching selectivity of the J&W DB-5Q compared to the Agilent J&W DB-5ms UI allows for seamless adoption, including compatibility with existing retention time locking libraries. The analytical performance of the DB-5Q coupled with the upgraded Agilent HES 2.0 allows for optimal sensitivity, as well as the ability to perform targeted and nontargeted analyses in tandem. References 1. Smith Henry, A. Comparison of Fritted and Wool Liners for Analysis of Semivolatile Organic Compounds by Gas Chromatography/Mass Spectrometry, Agilent Technologies application note, publication number 5994-2179EN, 2020. 2. How Does Bleed Impact GC/MS Data and How Can It Be Controlled? Agilent Technologies technical overview, publication number 5994-7586EN, 2024. 3. Andrianova, A.; Zhao, L. Brewing Excellence: Quantitating Over 200 Pesticides in Black Tea with Steady Performance and Maximized Uptime by GC/MS/MS, Agilent Technologies application note, publication number 5994-7436EN, 2024. 4. Reaser, B. C. Enhanced Longevity and Revolutionized Robustness for the Sensitive Detection of 190 Pesticides over 800 Injections, Agilent Technologies application note, publication number 5994-7385EN, 2024. Application Note Environmental Authors Luann Wong, Gabrielle Black, and Thomas Young Department of Civil and Environmental Engineering, University of California, Davis, CA, U.S. Sofia Nieto Agilent Technologies, Inc. Abstract Development of accurate mass libraries in environmental applications is key in expanding the scope of monitored compounds and allowing for target/suspect detection with high confidence. It also provides the opportunity to use a targeted data analysis approach that offers higher sensitivity and flexibility compared to nontarget screening. This application note describes the development and use of an accurate mass personal compound database and library (PCDL) of per- and polyfluoroalkyl substances (PFAS) for the Agilent 7250 GC/Q-TOF and demonstrates how the PCDL can be applied in both target as well as nontarget screening approaches using environmental samples, such as drinking water extracts. This study also demonstrates the benefits of using the high-resolution accurate mass GC/Q-TOF in nontarget screening using NIST23 and third-party libraries for identifying a substantial number of other contaminants of industrial origin in drinking water. Accurate Mass Library for PFAS Analysis in Environmental Samples and Workflow for Identification of Pollutants in Drinking Water Using GC/Q-TOF Return to Table of Contents 49 2 Introduction PFAS are emerging contaminants of increasing concern due to their environmental persistence, toxicity, and capability of bioaccumulation. There are currently thought to be over 6,000 PFAS that have been commercially produced1 , and recent studies have shown that many emerging PFAS detected in the environment can be volatile or semivolatile in nature2–4. Therefore, many analytical techniques are necessary for PFAS detection. Gas chromatography/mass spectrometry (GC/MS) is typically used for detecting volatile and semivolatile nonpolar PFAS compounds. In this study, the 7250 GC/Q-TOF system was used to take advantage of its high-resolution for detecting compounds with mass defects that are different from that of complex environmental matrixes. To ensure the most sensitive and reliable detection of PFAS, an accurate mass library that includes over 150 electron ionization (EI) PFAS spectra and contains both retention times (RTs) and retention indices (RIs) was created. The PFAS PCDL was further tested using both target and nontarget approaches when analyzing the drinking water extracts. In addition, to fully benefit from the GC/Q-TOF accurate mass capability combined with full-spectrum acquisition, enabling nontarget detection, NIST23 and the third-party library MassBank of North America (MassBank.us5 ) were also used to identify other contaminants in drinking water, with the false positives being effectively removed based on accurate mass information. Thus, many pollutants were identified in drinking water, including disinfection by-products (DBPs), industrial chemicals originated from personal care products, pharmaceuticals, as well as pesticide residues. Experimental Sample preparation The drinking water samples were collected at two different locations in California, U.S. and represented two different water source categories: a small surface water (Weaverville) and a mixed surface and ground water (Irvine). Water samples (2.4 L) were extracted on a multimode solid phase extraction (SPE) using HLB, WAX, WCS, and Isoelut ENV sorbents, and eluted with 5% methyl tert-butyl ether (MTBE) in methanol (MeOH), dichloromethane (DCM), 0.5% NH4 OH in 1:1 ethyl acetate (EtAc):MeOH, and 1.7% formic acid in 1:1 EtAc:MeOH. The combined extracts were concentrated, solvent exchanged to EtAc, and diluted tenfold. Table 1. Data acquisition parameters. GC and MS Conditions Agilent DB-5ms Agilent DB-624 MS Agilent 7250 GC/Q-TOF GC Agilent 8890 GC Inlet Agilent multimode inlet, Ultra Inert 4 mm liner, single taper with wool Inlet Temperature 70 °C for 0.01 min; 300 °C/min to 250 °C Injection Volume 1 µL Column Agilent J&W DB-5ms Ultra Inert (UI), 30 m × 0.25 mm, 0.25 µm Agilent DB-624 Ultra Inert, 30 m × 0.25 mm, 1.4 µm Oven Temperature Program 35 °C for 2 min; 7 °C/min to 210 °C, 20 °C/min to 300 °C, 4 min hold 30 °C for 2 min; 3 °C/min to 75 °C, 2 °C/min to 110 °C, 10 °C/min to 210 °C, 20 °C/min to 240 °C, 2 min hold Column Flow 1.2 mL/min constant flow 1 mL/min constant flow Carrier Gas Helium Transfer Line Temperature 250 °C Quadrupole Temperature 150 °C Source Temperature 200 °C Electron Energy 70 eV Emission Current Variable by time segment, 0.01 to 5 µA Spectral Acquisition Rate 5 Hz Mass Range (Tune) 50 to 1,200 m/z Data acquisition and data processing GC/MS analysis was performed using an Agilent 8890 GC coupled to an Agilent 7250 GC/Q-TOF using the data acquisition parameters described in Table 1. PFAS accurate mass spectra of GC-amenable compounds were acquired from individual PFAS standards. The chromatographic deconvolution and library search were performed in Agilent MassHunter Unknowns Analysis software, version 11.1. Accurate mass electron ionization (EI) fragments were converted to the theoretical m/z using Agilent MassHunter Qualitative Analysis software, version 10.0, and the spectra were exported into the accurate mass Agilent Personal Compound Database and Library (PCDL) Manager, version 8.0. The Agilent GC/Q-TOF Pesticide PCDL, PFAS PCDL, NIST23, as well as MassBank.us were used to perform initial compound identification. Prior to performing the library search with MassBank.us, the spectra, along with metadata information from this database, were exported in the PCDL format using Agilent ChemVista software, version 1.0, as described elsewhere.6–7 RIs and accurate mass information were used to confirm the compound identification. Statistical analysis was performed in Agilent Mass Profiler Professional (MPP), version 15.1. 50 3 Results and discussion Accurate mass library for PFAS To create an accurate mass GC/MS PCDL, spectra were collected for over 100 volatile and semivolatile PFAS compounds. Accurate mass fragment ions were automatically annotated with formulas based on accurate mass information and isotope ratios using MassHunter Qualitative Analysis software (Figure 1). The fragment formula annotations were verified, corrected when necessary, and automatically converted to the theoretical m/z. Figure 1. (A) Extracted ion chromatogram (EIC) of the molecular ion and fragment formula annotation of spectrum for one of the PFAS compounds in Agilent MassHunter Qualitative Analysis software. (B) The PFAS PCDL contains EI spectra as well as the metadata, including molecular structure and database identifiers. 0 1 2 3 19.718 17.0 17.5 18.0 18.5 19.0 19.5 20.0 20.5 21.0 21.5 22.0 22.5 23.0 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 107.0303 [C4 H5 F2 O]+ 68.9947 [C F3]+ 130.9914 [C3 F5]+ 376.0138 168.9879 [C9 H5 F13 O]+ [C3 F7]+ 245.0003 [C6 H2 F9]+ 195.0034 [C5 H2 F7]+ 289.0065 [C8 H3 F10]+ 337.0078 [C9 H4 F11 O]+ 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 3-(Perfluorohexyl)-1,2-epoxypropane ×106 A B Acquisition time (min) Counts ×106 Counts Mass-to-charge (m/z) 51 4 The PFAS compound classes include perfluoroalkyl iodides (PFAIs), fluorotelomer iodides (FTIs), fluorotelomer alcohols (FTOHs), fluorotelomer olefins (FTOs), fluorotelomer acrylates (FTACs), fluorotelomer methacrylates (FTMACs), fluorotelomer carboxylic acids (FTCA), fluorotelomer unsaturated carboxylic acids (FTUCA), perfluoroalkane sulfonamides (FASA), and more (Figure 2). To acquire spectra for the PFAS PCDL, the mid-polar DB-624 GC column (30 m × 0.25 mm, 1.4 µm) was used to ensure the best retention and separation of the challenging volatile PFAS. In addition to the RTs, RIs for the mid-polar column phase were also calculated for all compounds. Inclusion of RIs provides the flexibility of the GC method when using the PFAS PCDL as soon as GC column phase stays the same. The remaining metadata, including compound structures and database identifiers, were added using PCDL Manager. Compound overlap of the volatile and semivolatile PFAS classes between the accurate mass PFAS PCDL and NIST23 library is shown in Table 2. Table 2. Compound overlap between the PFAS PCDL and NIST23 library. Percent Unique to PCDL Total Number All 53 158 PFCA 55 29 FTO 50 6 PFAI and FTI 17 6 FTCA and FTUCA 67 9 FTAC and FTMAC 25 8 FTOH 40 15 FASA 8 12 A significant number of the PFAS compounds (over 50%) were found to be present uniquely in the PFAS PCDL. In particular, the spectra of many per- and polyfluorinated carboxylic acids, fluorotelomer olefins, and fluorotelomer alcohols were found to be unique to the PCDL, thus highlighting the value of the accurate mass PFAS library in PFAS research. Figure 2. Examples of different PFAS compound classes from the PFAS PCDL. +EI MS1 QTOF FV=70 60 80 100 120 140 160 180 200 220 240 260 280 300 320 0 20 40 60 80 100 +EI MS1 QTOF FV=70 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 0 20 40 60 80 100 +EI MS1 QTOF FV=70 30 50 70 90 110 130 150 170 190 210 230 250 270 290 0 20 40 60 80 100 +EI MS1 QTOF FV=70 40 80 120 160 200 240 280 320 360 400 440 480 520 0 20 40 60 80 100 55.01784 100.00 99.04405 37.80 168.98828 21.40 3.92 426.99982 1.41 1.30 3.15 230.98508 1.04 +EI MS1 QTOF FV=70 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400 Abundance 0 20 40 60 80 100 +EI MS1 QTOF FV=70 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 Abundance 0 20 40 60 80 100 +EI MS1 QTOF FV=70 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 Abundance 0 20 40 60 80 100 +EI MS1 QTOF FV=70 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 Abundance Abundance Abundance Abundance Abundance 0 20 40 60 80 100 218.98508 100.00 68.99466 59.25 126.90392 32.40 29.32 176.90073 20.56 99.99306 11.20 3.55 Nonafluoro-1-iodobutane (PFBI) 1,1,1,2,2,3,3,4,4-Nonafluoro-6-iodohexane (6:2 FTI) 8:2 Fluorotelomer acrylate (8:2 FTAC) Perfluoroheptanoic acid (PFHpA) Methyl perfluorohexanoate 6:2 Fluorotelomer alcohol (6:2 FTOH) Perfluorooct-1-ene Mass-to-charge (m/z) Mass-to-charge (m/z) Mass-to-charge (m/z) Mass-to-charge (m/z) Mass-to-charge (m/z) Mass-to-charge (m/z) Mass-to-charge (m/z) Mass-to-charge (m/z) 238.89754 345.88956 312.98813 380.97549 518.01691 68.99466 100.00 130.99147 50.23 99.99306 21.27 180.98828 16.76 281.00186 15.55 218.98508 6.61 246.98000 1.85 308.99680 1.51 146.98639 1.08 68.99466 100.00 118.99147 45.56 318.97870 23.32 168.98828 22.49 230.98508 51.00408 21.05 2.30 280.98187 1.02 69.03349 100.00 268.03287 89.62 95.01031 8.74 51.00408 8.37 118.99147 8.33 253.00940 5.26 204.03682 1.48 149.04086 1.45 380.97549 68.99466 100.00 84.54 130.99147 74.31 180.98828 31.56 92.99466 22.98 230.98508 11.52 280.98187 5.32 330.97870 4.05 95.01031 100.00 127.01654 48.66 296.00537 51.00408 21.59 10.86 168.98828 6.72 230.98508 5.98 363.00491 5.70 373.92084 100.00 227.01015 40.57 140.91957 30.78 68.99466 29.23 247.01639 0.74 354.92245 0.06 285.92725 0.05 2,2,3,3,4,4,4-Heptafluorobutyl methacrylate (3:1 FTMAC) 52 5 PFAS in drinking water extracts For PFAS detection, the extracts of drinking water were separated on a DB-624 column and analyzed using the 7250 GC/Q-TOF. To be able to detect early-eluting volatile PFAS, the emission current was set up by segment, as shown in Table 1, thus excluding the solvent peak from the detection. Both target and nontarget approaches were evaluated using the PFAS PCDL. When performing the nontarget analysis, the chromatographic deconvolution was carried out in the MassHunter Unknowns Analysis software using a SureMass algorithm, which is optimized for complex, high-resolution EI data. The PFAS PCDL then was used to search the deconvoluted spectra with RT matching. One of the PFAS—a transformation product of the perfluorocarboxylic acid— was identified in drinking water extract using this approach (Figure 3A). An additional benefit of nontarget screening is the use of multiple libraries (including libraries containing unit mass spectra) that can all be searched simultaneously. Figure 3. Example of PFAS (methyl perfluorooctanoate) identified in drinking water samples using PFAS PCDL in a nontarget approach using (A) Agilent MassHunter Unknowns Analysis software, and (B) a target GC/Q-TOF screening approach. A B 53 6 One of the advantages of the target approach based on the GC/Q-TOF Screener tool of the MassHunter Quantitative Analysis software (described in detail previously8 ) and PCDL is that all the parameters could be set up individually for every compound in the method. This approach allows for a substantial flexibility when performing the screening method optimization at the data processing level, enabling the highest degree of sensitivity and specificity. Another significant benefit of this approach is that it saves the time usually spent reviewing the results. The GC/Q-TOF Screener algorithm validates quantifier and qualifier ions based on outliers, and for most compounds, either confirms or rejects their presence automatically. Only a few compounds remain highlighted to indicate that a manual review might be necessary for confirmation of compound identity. The same PFAS compound—methyl perfluorooctanoate—that was identified in drinking water extracts using a nontarget approach was also detected using the GC/Q-TOF Screener with library match scores (LMS) of > 99 (Figure 3B). It has previously been reported that perfluoroalkyl carboxylic acids can be converted to corresponding methyl esters in the presence of methanol9 , thus plausibly explaining the presence of the methyl ester of PFOA in drinking water extracts. Identification of other contaminants in drinking water samples To screen for additional contaminants in drinking water samples in a nontarget manner, the GC/Q-TOF Pesticide PCDL, NIST23 library, and MassBank.us were used. The choice of the DB-5ms UI column enabled RI matching while searching the NIST23 library, thus enhancing the confidence in compound identification. The ExactMass tool in MassHunter Unknowns Analysis software was used to eliminate the false positives based on the accurate mass information and molecular formula of the hit. This is particularly practical when using unit mass libraries such as NIST23 (Figure 4A) and MassBank.us. Over 100 contaminants were identified and confirmed using accurate mass information (Figure 4A and 4B, and Tables 3 and 4) from the sample without re-injection. Among the identified contaminants, one of the significant groups was disinfection by-products, formed when chlorine and bromine interact with organic matter. These compounds included halomethanes and haloacetic acids, which are the most common disinfection by-products. Other prominent groups of contaminants included compounds originating from industrial processes (such as those used in cleaning products and manufacturing of plastics, dyes, and pharmaceuticals), PAHs and their derivatives, as well as pesticides. Approximately 400 hits per sample with LMS > 70 were detected using MassBank.us, including over 20 contaminants—mostly DBPs and PAHs. Since this library does not contain RI information for most of the compounds, many hits could potentially be false positives. One such example is shown in Figure 5. The hit from MassBank.us was 1-bromooctane with a high LMS of 88.3 and an RI of 1,134, according to NIST23. This would make a difference of over 400 RI units between the hit and the compound in question, indicating that the ID is likely incorrect. The difference between the compound RI and the NIST23 hit was only 10 RI units (Figure 5B). Note that the hits from both NIST23 and MassBank.us perfectly match the accurate mass information displayed in the ExactMass tables (Figure 5). 54 7 Figure 4. Examples of the contaminants identified in drinking water extracts using (A) NIST23 and (B) Agilent GC/Q-TOF Pesticide PCDL. The ExactMass tool (outlined in orange) helped to provide additional confirmation of unit mass library hits based on accurate mass. Compound ions are highlighted in the mirror plot when m/z corresponds to the library hit formula. A B 55 8 Table 3. Contaminants identified in drinking water using the NIST23 library with LMS > 75. *The cases where delta RI was calculated considering predicted RIs rather than experimental (experimental not available) are denoted by an asterisk. Some of the prominent disinfection by-products are highlighted in red. RT Compound Name Match Score Formula RI Difference 4.79 Bromodichloromethane 95.4 CHBrCl2 –56 4.81 Chloral 78.8 C2 HCl3 O –9 4.91 Dichloroacetonitrile 86.4 C2 HCl2 N –76 4.95 Chloromethylmethyl sulfide 94 C2 H5 ClS –59* 5.11 Dimethyl disulfide 98.4 C2 H6 S2 –35 5.35 Methyldiallylamine 85.7 C7 H13N –50* 5.47 Bromoacetonitrile 82.8 C2 H2 BrN 1 5.95 Dibromochloromethane 95.5 CHBr2 Cl –25 6.01 Tetrachloroethylene 96.5 C2 Cl4 –12 6.04 1,1-Dimethyl-3-chloropropanol 88.4 C5 H11ClO 7 6.34 Bromoacetone 87.8 C2 HBrClN –3 6.59 Dichloroacetic acid methyl ester 89.2 C3 H4 Cl2 O2 –7* 7.67 Tribromomethane 98.2 CHBr3 –10 8.24 Methyl bromo(chloro)acetate 77.4 C3 H4 BrClO2 –3 8.31 Dibromoacetonitrile 86.6 C2 HBr2 N –15 10.63 2,2-Dichloroacetamide 83.5 C2 H3 Cl2 NO –4* 10.73 1,2-Dichlorbenzene 98.5 C6 H4 Cl2 9 14.12 Naphthalene 81.9 C10H8 –4 15.45 Caprolactam 89.6 C6 H11NO 3 16.43 2-Methylnaphthalene 89.3 C11H10 –1 16.64 Phthalic anhydride 92.5 C8 H4 O3 5 16.98 Benzamide 82.8 C7 H7 NO 18 18.05 Biphenyl 83.2 C12H10 –1 18.18 Benzeneacetamide 84.3 C8 H9 NO 13 19.27 Dimethyl phthalate 75.1 C10H10O4 8 19.96 Acenaphthene 91.2 C12H10 –4 20.18 4-Methylbiphenyl 79.4 C13H12 –4 20.28 2,4-Di-tert-butylphenol 90.2 C14H22O 10 20.56 Dibenzofuran 92.4 C12H8 O –5 21.23 1-Bromododecane 75.7 C12H25Br –10 21.45 Diethyltoluamide (DEET) 78.1 C12H17NO 10 21.69 Diethyl phthalate 96 C12H14O4 8 21.71 Fluorene 75.3 C13H10 –4 22.01 2-(Methylmercapto)benzothiazole 77.8 C8 H7 NS2 2 22.43 Benzophenone 94.8 C13H10O 4 22.60 Tributyl phosphate 93 C12H27O4 P 7 23.51 Hexachlorobenzene 97.2 C6 Cl6 9 RT Compound Name Match Score Formula RI Difference 24.19 9H-Fluoren-9-one 97.1 C13H8 O 8 24.26 9H-Fluoren-9-ol 81.5 C13H10O 9* 24.90 Anthracene 94.4 C14H10 0 24.91 Tris(2-chloroisopropyl)phosphate 82.2 C9 H18Cl3 O4 P 27 25.02 Benzo[h]quinoline 88.2 C13H9 N –2 25.53 2,4-Diphenyl-4-methyl-2(E)-pentene 76.5 C18H20 8 25.55 Benzo[f]quinoline 91.1 C13H9 N –1 25.73 Carbazole 76.8 C12H9 N –4 25.96 Di-sec-butyl phthalate 90.8 C16H22O4 –2 26.09 3,3-Diphenyl-2-propenenitrile 82.7 C15H11N 18* 26.27 3-Methyldibenzothiophene 80.8 C13H10S –5 26.66 3-Methylphenanthrene 84 C15H12 1 27.00 2-Methylanthracene 88.5 C15H12 –27 27.31 Dibutyl phthalate 92.4 C16H22O4 9 27.59 9,10-Anthracenedione 93.2 C14H8 O2 –27 28.21 Octachlorostyrene 88.6 C8 Cl8 –7 28.36 Cyclic octaatomic sulfur 93.1 S8 –18 28.51 Drometrizole 82.3 C13H11N3 O –5 28.53 Fluoranthene 97.8 C16H10 –12 28.54 Phenindione 79 C16H10 –28* 28.91 Dibenzothiophene sulfoxide 87.2 C12H8 OS –41* 28.94 Pyrene 89.3 C16H10 –25 29.02 1-Azapyrene 78.4 C15H9 N 2 29.34 Bisphenol A 84.1 C15H16O2 26* 29.37 2-Amino-9-fluorenone 83.3 C13H9 NO 2* 29.69 Bis(4-chlorophenyl) sulfone 77 C12H8 Cl2 O2 S –1 29.87 2,2'-Methylene-bis-(4-methyl-6-tbutylphenol) 87.5 C23H32O2 2 30.79 Benzo[b]naphtho[1,2-d]thiophene 77.2 C16H10S 13 30.90 7H-Benz[de]anthracen-7-one 89.2 C17H10O 85 30.94 Benzo[b]naphtho[2,1-d]thiophene 81.3 C16H10S –16 31.04 Phthalic acid, di(2-propylpentyl) ester 94.8 C24H38O4 –5 31.38 Bis[3,4-dichlorophenyl]sulfone 82.3 C12H6 Cl4 O2 S 3* 31.48 Bumetrizole 78.6 C17H18ClN3 O 57 31.51 Benz(a)anthracene-7,12-dione 76.1 C18H10O2 –48* 31.82 Bis(2-ethylhexyl) isophthalate 84.7 C24H38O4 –35 32.37 Decachlorobiphenyl 94.3 C12Cl10 –81 56 9 Figure 5. Compound misidentified by MassBank.us due to lack of the RI information. (A) MassBank.us hit. (B) NIST23 hit. A B Table 4. Additional contaminants identified in drinking water using the Agilent GC/Q-TOF Pesticide PCDL. RT Compound Name Match Factor Formula 6.17 2-Picoline 96.7 C6 H7 N 6.90 Methanesulfonate-methyl 79.6 C2 H6 O3 S 8.17 PPD/p-Phenylenediamine 80.0 C6 H8 N2 8.40 o-Toluidine 82.3 C7 H9 N 8.93 Thanite 83.9 C13H19NO2 S 9.17 Benzaldehyde 98.5 C7 H6 O 9.52 Phenol 89.6 C6 H6 O 11.46 Acetophenone 94.3 C8 H8 O RT Compound Name Match Factor Formula 11.99 2,4,5-Trimethylaniline 82.7 C9 H13N 12.90 2-Nitrophenol 77.1 C6 H5 NO3 22.29 DPA/Diphenylamine (DFA) 84.7 C12H11N 22.44 Isoxadifen 93.3 C16H13NO3 24.64 Benzylbenzoate 83.0 C14H12O2 25.96 DIBP/Diisobutyl phthalate 86.1 C16H22O4 27.00 1-Methylphenanthrene 85.3 C15H12 57 10 Additionally, an interesting case was observed for a compound with an RI of 1,858, whereby MassBank.us likely provided correct ID (with the LMS of 85.8) while NIST23 did not (Figure 6), proving the value of including third-party libraries in a compound identification workflow. The compound’s most likely ID is thioxanthene, with a NIST23 experimental RI of 1,977, and an AI-predicted RI of 1,876. The experimental RI for this compound, used for the NIST23 library search, provided a significant RI delta of 124 RI units. However, the experimental RI for this compound was only based on one data point, and thus may not be accurate. The AI-predicted NIST23 RI generated a significantly smaller RI delta (18 RI units). Due to the large RI difference between the compound RI and the NIST23 experimental RI of thioxanthene, another NIST23 hit, 1-methyldibenzothiophene with the lower LMS of 81.1 was chosen (Figure 6B). Among the contaminants with the highest response, which were identified in drinking water extracts using all three libraries, were mostly PAHs, DBPs, and phthalates (Figure 7). Individual samples of drinking water from the same group represented different households. The contaminants in drinking water extracts were detected at a wide range of concentrations, estimated to be from low- and sub-ppb levels (for pesticides) to hundreds of ppb (in the case of PAHs), suggesting that an extended dynamic range might be desirable for this application. Statistical analysis was performed in the MPP software, where the differences between Weaverville and Irvine water sources (n = 5 per group) were evaluated and displayed on a volcano plot (Figure 8). The volcano plot displays fold change versus statistical significance, and is used to quickly detect differences between the two groups. Compounds that were present in higher concentrations in Irvine water compared to Weaverville are colored in red and shown in the upper-right quadrant. Compounds that were found at higher concentrations in Weaverville water extracts compared to Irvine are colored in blue and displayed in the upper-left quadrant. Most contaminants occurred at higher levels in drinking water from Irvine (a more densely populated urban area) compared to Weaverville. Figure 6. Compound likely misidentified by NIST23 due experimental RI information available for only one data point. (A) MassBank.us hit. (B) NIST23 hit. A B 58 11 Figure 8. Comparison of water sourced in Irvine versus Weaverville on volcano plot, showing log2 of fold change (FC) versus –log10 of p-value. Figure 7. High-level contaminants identified in the drinking water of different households (n = 5 for each group) from Irvine (IR) and Weaverville (WV). Peak area Chlorodibromomethane 9H-Fluoren-9-one Anthracene Dibutyl phthalate Fluoranthene Tributyl acetylcitrate 0 5 10 15 20 25 ×105 IR WV 59 www.agilent.com DE50093873 This information is subject to change without notice. © Agilent Technologies, Inc. 2023 Printed in the USA, December 5, 2023 5994-6966EN Conclusion Accurate mass libraries of environmental contaminants (such as PFAS) broaden the scope of suspects screened in environmental samples and increase confidence in the identification of pollutants. The accurate mass library described in this application note, containing over 150 PFAS EI spectra, including several emerging volatile PFAS, enabled identification of PFAS in drinking water samples using both nontarget and target workflows. Additional contaminants were identified in drinking water from two different source categories, including disinfection by-products, PAHs, pesticides, and other industrial contaminants. Two drinking water sources were compared, and a higher number of contaminants were identified in the water extracts from Irvine compared to Weaverville. Acknowledgement The authors would like to thank the U.S. Environmental Protection Agency (EPA) for preparing the PFAS standards and providing them to Agilent Technologies through a material transfer agreement. References 1. Poly- and Perfluoroalkyl Substances (PFAS) Overview and Current Activities. https://www.loudounwater.org/ residential-customers/facts-about-pfas 2. Hammer, J.; Endo, S. Volatility and Nonspecific van der Waals Interaction Properties of Per- and Polyfluoroalkyl Substances (PFAS): Evaluation Using Hexadecane/Air Partition Coefficients. Environ. Sci. Technol. 2022 Nov 15, 56(22), 15737–15745. DOI: 10.1021/acs.est.2c05804 3. Liu, X. Understanding Semi-volatile Organic Compounds (SVOCs) in Indoor Dust. Indoor Built Environ. 2022 Jan 10, 31(2), 291–298. DOI: 10.1177/1420326x211070859 4. Del Vento, S.; Halsall, C.; Gioia, R; Jones, K; Dachs, J. Volatile Per- and Polyfluoroalkyl Compounds in the Remote Atmosphere of the Western Antarctic Peninsula: an Indirect Source of Perfluoroalkyl Acids to Antarctic Waters? Atmos. Pollut. Res. 2012, 3(4), 450–455. DOI: https://doi.org/10.5094/APR.2012.051 5. Wohlgemuth, G.; Mehta, S. S.; Mejia, R. F.; Neumann, S.; Pedrosa, D.; Pluskal, T.; Schymanski, E. L.; Willighagen, E. L.; Wilson, M.; Wishart, D. S.; et al. SPLASH, a Hashed Identifier for Mass Spectra. Nat. Biotechnol. 2016, 34, 1099–1101. https://massbank.us/ 6. Agilent ChemVista Library Manager. Agilent Technologies technical overview, publication number 5994-5924EN, 2023. 7. Valdiviez, L.; Fiehn, O.; Nieto, S. Differences in Metabolic Profiles of Individuals with Heart Failure Using High-Resolution GC/Q-TOF. Agilent Technologies application note, publication number 5994-6858EN, 2023. 8. Van Gansbeke, W.; Albertsdóttir, A. D.; Polet, M.; Van Eenoo, P.; Nieto, S. Introducing Semi-Automated GC/Q-TOF Screening with the AssayMAP Bravo Sample Prep Platform for Antidoping Control. Agilent Technologies application note, publication number 5994-6702EN, 2023. 9. Hanari, N.; Itoh, N.; Ishikawa, K.; Yarita, T.; Numata, M. Variation in Concentration of Perfluorooctanoic Acid in Methanol Solutions During Storage. Chemosphere 2014 Jan, 94, 116–20. DOI: 10.1016/j. chemosphere.2013.09.040 Application Note Environmental Author Brooke C. Reaser Agilent Technologies, Inc. Abstract Multiresidue pesticide analysis has become one of the most difficult but important analytical challenges for those using gas chromatography and mass spectrometry. The Agilent 8890 gas chromatograph (GC) coupled with the Agilent 7010 triple quadrupole mass spectrometer (GC/TQ) with a high efficiency source 2.0 (HES 2.0) upgrade is an analytically accurate, robust, and reproducible instrument for multiresidue pesticide analysis of complex samples. The analysis of 190 pesticides in a spinach extract was conducted using an Agilent QuEChERS extraction kit across 800 injections. Only GC inlet maintenance was required over the duration of the injections. The instrument configuration that enabled robust performance included a multimode inlet, a mid-column backflushing configuration, and the HES 2.0. No degradation of the analytical method, sensitivity, or instrument performance occurred, allowing for the high-throughput, accurate, robust, and sensitive detection of pesticides in spinach. Enhanced Longevity and Revolutionized Robustness for the Sensitive Detection of 190 Pesticides over 800 Injections Return to Table of Contents 61 2 Introduction Multiresidue pesticide analysis remains an important analytical challenge for food safety.1-7 Pesticides used to improve crop yield can end up in the final product, raising concerns for consumer safety. As a result, multiple governing bodies worldwide have published requirements for the maximum legal residue limit (MRL) or tolerance for pesticides allowed in a product. However, the number of pesticides used in food products continues to grow as novel chemicals are introduced, which in turn increases the complexity of the multiresidue pesticide analysis. For especially difficult matrices, QuEChERS, which stands for quick, easy, cheap, effective, rugged, and safe, has become widely accepted as a sample preparation technique for multiresidue pesticide analysis.1 Agilent QuEChERS extraction kits provide prepackaged dispersive and extraction products, extraction salts, and ceramic homogenizers in easy-to-use kits. The kits are ready-made for various methods, including methods of the Association of Official Agricultural Chemists (AOAC)2 and European Standard (EN).3 QuEChERS extracts of food commodities can be analyzed by either GC or high performance liquid chromatography (HPLC) combined with a mass spectrometer (MS) or tandem mass spectrometers (MS/MS).4,5 Depending on the extent of sample cleanup and the food commodity being analyzed, QuEChERS extracts can cause contamination of the instrument, resulting in poor data quality.6 This contamination can exhibit as loss of sensitivity, retention time shifting, poor peak shape, and more. Regular maintenance of the instrument, including GC inlet maintenance, GC column trimming, and ion source cleaning, is required to ensure the robustness and accuracy of the method results. Backflushing is one of the key practices in which GC/MS/MS analyses of complex matrices can be improved.7 Backflushing refers to the reversing of flows in the capillary column so that unwanted matrix components are flushed out of the GC split vent instead of proceeding to the detector. Backflushing can provide improved method robustness and help minimize the required maintenance of the mass spectrometer. Agilent has introduced the new HES 2.0 ion source as part of the 7010D triple quadrupole mass spectrometer (TQ), and it is also available as an upgrade to 7010A/B/C GC/TQ instruments. The HES 2.0 ion source provides improved system robustness, allowing the analysis of hundreds of injections of pesticides in food matrices with only GC maintenance and ion source cleaning necessary. The HES 2.0 delivers the same unparalleled analytical sensitivity for ultratrace-level analysis as the original HES. Multiresidue pesticide analysis in spinach extract was carried out using a 7010B GC/TQ upgraded with the HES 2.0. Matrix-matched standards were used to analyze and quantify over 400 injections of baby spinach extract spiked with 50 ppb of multiresidue standards, demonstrating both method and instrument robustness. The multimode inlet (MMI) and backflushing between two 15 m columns allowed for minimal downtime for GC inlet maintenance. Sensitivity and quantitative accuracy were maintained without any maintenance performed on the mass spectrometer. Experimental GC/TQ analysis An 8890 GC with a 7010B TQ system upgraded with the HES 2.0 was used for analysis. The instrument and method were configured as outlined in a previous application note7 , as shown in Figure 1. Agilent 7010 with HES 2.0 PSD (helium) Agilent 8890 GC Liquid Injector Multimode inlet (helium) HES 2.0 Agilent HP-5ms UI 15 m, 0.25 × 0.25 Figure 1. The Agilent 8890/7010B GC/TQ system upgraded with the HES 2.0 and system configuration. 62 3 The GC was equipped with an Agilent 7650A automatic liquid sampler (ALS) and 50-position tray. The GC used an MMI to achieve a temperature-programmed splitless injection. Mid-column backflush was carried out using an Agilent Purged Ultimate Union (PUU) installed between two identical 15 m columns; the 8890 GC pneumatic switching device (PSD) module allowed for fewer occurrences of regular maintenance. The method parameters are listed in Table 1. Table 1. Agilent 8890 GC and Agilent 7010B upgraded with the HES 2.0 ion source conditions for pesticide analysis. GC Instrument Agilent 8890 with Fast Oven, Auto Injector and Tray Inlet Multimode Inlet (MMI) Mode Splitless Purge Flow to Split Vent 15 mL/min at 0.75 min Septum Purge Flow 3 mL/min Septum Purge Flow Mode Switched Injection Volume 1.0 µL Injection Type Standard L1 Air Gap 0.1 µL Gas Saver Off Inlet Temperature 60 °C for 0.1 min, then to 280 °C at 600 °C/min Postrun Inlet Temperature 310 °C Postrun Total Flow 25 mL/min Carrier Gas Helium Inlet Liner Agilent Ultra Inert 2 mm dimpled liner (p/n 5190‑2297) Oven Initial Oven Temperature 60 °C Initial Oven Hold 1 min Ramp Rate 1 40 °C/min Final Temperature 1 170 °C Final Hold 1 0 min Ramp Rate 2 10 °C/min Final Temperature 2 310 °C Final Hold 2 3 min Total Run Time 20.75 min Postrun Time (Backflushing) 1.5 min Equilibration Time 3 min Column 1 Type Agilent HP‑5ms UI (p/n 19091S‑431UI) Length 15 m Diameter 0.25 mm Film Thickness 0.25 µm Control Mode Constant flow Flow 1.00 mL/min Inlet Connection Multimode inlet (MMI) Outlet Connection PSD (PUU) PSD Purge Flow 5 mL/min Postrun Flow (Backflushing) –7.873 mL/min Column 2 Type Agilent HP‑5ms UI (p/n 19091S‑431UI) Length 15 m Diameter 0.25 mm Film Thickness 0.25 µm Control Mode Constant flow Flow 1.200 mL/min Inlet Connection PSD (PUU) Outlet Connection MSD Postrun Flow (Backflushing) 8.202 mL/min MSD Model Agilent 7010B Source Agilent HES 2.0 Vacuum Pump Performance turbo Tune File Atunes.eihs.tune.xml Solvent Delay 3 min Quad Temp (MS1 and MS2) 150 °C Source Temperature 280 °C Mode dMRM or Scan He Quench Gas 4 mL/min N2 Collision Gas 1.5 mL/min MRM Statistics Total MRMs (dMRM mode) 552 Minimum Dwell Time (ms) 2.63 Minimum Cycle Time (ms) 82.42 Maximum Concurrent MRMs 48 EM Voltage Gain Mode 10 63 4 The Agilent Pesticide and Environmental Pollutant (P&EP) database (P&EP 4, part number G9250AA) was used to easily and rapidly create the dynamic multiple reaction monitoring (dMRM) method. This method, which enabled the analysis of 190 pesticides with a total of 552 MRMs, resulted in a maximum of 48 concurrent MRMs, as shown in Figure 2. QuEChERS sample preparation The sample preparation procedure is summarized in Figure 3. A bag of frozen organic baby spinach was homogenized using a spice grinder. Then, eight replicates were prepared. For each replicate, 15 g of the homogenized spinach were weighed into a 50 mL test tube. Then, two of the replicates were designated as samples, while the remaining six were designated for pooled matrix-matched standards. Fifteen microliters of the internal standard mixture (part number 5190-0502) diluted to 50 ng/µL was added to the two spinach samples. To all eight replicates, 15 mL of 1% acetic acid in acetonitrile was added and the mixture was vortexed until well mixed. To each Figure 2. Concurrent MRMs versus retention time. Retention time (min) Concurrent MRMs 0 4 8 12 16 20 24 28 32 36 40 44 48 52 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Figure 3. Sample preparation workflow. QuEChERS extraction kit Agilent QuEChERS dispersion kit 15 g of spinach + 15 mL of ACN with 1% AA, vortex Take 8 mL supernatant Shake and centrifuge Shake and centrifuge Dilute, spike, and prepare in-vial Sample analysis on GC/TQ 64 5 50 mL tube, the salt packet and two homogenizers from the QuEChERS kit (part number 5982-5755) were added for extraction. These were shaken for 1 minute, then centrifuged at 4,000 rpm with a maximum radius of 17.4 cm for 5 minutes. For the samples and the pooled standards, 8 mL of the supernatant was transferred to a tube with salt for dispersion using the QuEChERS kit (part number 5982-5058). These were shaken for 30 seconds then centrifuged as before for 5 minutes. The supernatant for each of the samples was removed and placed in an amber glass vial. For the pooled standards, the supernatant of all replicates was removed and mixed in a large amber jar for standard preparation. Standard preparation The multiresidue matrix-matched standards were created from the FDA analytical reference standards kit (part number PSM-101). Mixes A, B, C, D, E, L, M, N, O, and P from the PSM-101 pesticides mix were combined to create a 10 ppm stock standard of 190 pesticide residues. The stock solution was then diluted in acetonitrile down to the following nominal concentrations: 1,000, 100, 10, and 1 ppb. In a GC vial, the standards were combined with the spinach extract. The internal standard mixture of parathion-d10 and alpha-BHC-d6 , and acetonitrile were combined until the nominal concentration of the internal standard was 50 ppb and the nominal concentrations of the matrix-matched standards in a total volume of 1,500 µL were as follows: 0.1, 0.5, 1, 5, 10, 50, 100, 253.3, 500, and 1,000 ppb. Extra vials of the 50-ppb matrix-matched standard were made for quantification to test robustness. The samples were diluted by a factor of three in the vials to match the matrix concentration in the standards. This 3x dilution factor was determined by analyzing the matrix alone in full scan mode as described in a previous application note7 to ensure that the instrument was not overloaded or saturated by the matrix. Vials with 250 µL glass inserts were used with 100 µL of each standard or sample in-vial for GC analysis. Sequence Each sequence included 102 injections of spinach extract, either as a sample or matrix-matched standard. Additional injections of blank acetonitrile were used to evaluate system cleanliness by ensuring no analyte carryover or additional background contamination. A representative MRM chromatogram of the 50-ppb matrix-matched standards in spinach extract is shown in Figure 4A with a zoomed in portion shown in Figure 4B. The sequence included: – Matrix-matched calibration curve (10 pts) – Two spinach samples – Matrix-matched calibration curve (10 pts) – Matrix-matched standards (50 ppb) × 60 times – Matrix-matched calibration curve (10 pts) × 2 times, each standard in duplicate Figure 4. A representative chromatogram (A) and a zoomed in portion of the chromatogram (B). ×108 A Acquisition time (min) Counts 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 ×107 B Acquisition time (min) Counts 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 65 6 To achieve more than 400 injections of the 50 ppb matrix-matched standard, seven sequences were run for a total of 819 injections, of which 714 were spinach matrix injections. After each sequence, the GC inlet liner and septum were changed, and the 5 µL syringe was changed as necessary. Also, the GC vials were refreshed with samples and standards that had been stored in the freezer. No additional instrument maintenance was performed. Results and discussion Of the 190 pesticide residues analyzed by GC/TQ, 114 were selected for further study based on their analytical response and performance. These 114 residues were chosen because their calibration curves were either linear or quadratic throughout the seven sequences. These curves did not require extensive analyst intervention, such as manual integration, and had calibration curves that encompassed the 50 ppb point so that the 50 ppb matrix-matched standards for robustness could be calculated as samples. Any points on the calibration curve that had signal-to-noise (S/N) < 3, were interfered with by contaminants, or had accuracy greater than or equal to ± 25% were excluded (≥ ± 25%). A table summarizing these residues can be found at the end of the application note in Table 2. Many of the pesticide residues could accurately be quantified down to 0.5 or 0.1 ppb while maintaining S/N > 10 and quantification accuracy of < 25%. For example, Figures 5A, 5B, and 5C show the 0.1 ppb peak, corresponding qualifiers, and calibration curve of DCPA, respectively. The data are defined well by a quadratic calibration curve over the calibration range 0.1 to 1,000 ppb through four orders of magnitude. DCPA has many MRLs as defined by the US FDA down to 50 ppb in various fruits and vegetables. Figure 5. Integrated peak (A) and qualifiers (B) of the 0.1 ppb peak of DCPA as well as the full calibration curve (C). ×103 ×102 A B ×10 C 1 Acquisition time (min) Counts + MRM CID at 25.0 (298.9 & 221.0) 10.26 10.28 10.30 10.32 10.34 10.36 10.38 10.40 10.42 10.44 Acquisition time (min) 10.26 10.28 10.30 10.32 10.34 10.36 10.38 10.40 10.42 10.44 -0.2 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 10.325 min -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 298.9 & 221.0 300.9 & 223.0 331.8 & 300.9 Ratio = 89.2 (95.1%) Ratio = 42.0 (97.0%) DCPA Relative concentration 0 1 2 3 4 5 6 7 8 9 10 -0.2 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 y = -0.013350x2 + 2.835571x + 0.001611 R2 = 0.99899365 R = 0.99969055 Weight: 1/x Relative abundance (%) Relative responses 66 7 Figures 6A and 6B show the chlorpyrifos 0.5 ppb peak and qualifiers (respectively), while Figure 6C shows the corresponding calibration curve. The curve is linear through 3.5 orders of magnitude and chlorpyrifos has MRLs in various food commodities, the lowest of which is 0.01 ppm in food items such as egg, fig, grape, and apple. Figure 6. Integrated peak (A), qualifiers (B) of the 0.5 ppb peak of chlorpyrifos as well as the full calibration curve (C). ×104 ×102 A B ×10 C 1 Acquisition time (min) Counts Relative concentration Relative abundance (%) Relative responses -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 196.9 & 169.0 198.9 & 171.0 313.8 & 257.8 Ratio = 103.0 (108.4%) Ratio = 43.1 (98.7%) + MRM CID at 15.0 (196.9 & 169.0) 10.16 10.18 10.20 10.22 10.24 10.26 10.28 10.30 10.32 Acquisition time (min) 10.16 10.18 10.20 10.22 10.24 10.26 10.28 10.30 10.32 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 10.220 min Chlorpyrifos 0 1 2 3 4 5 6 7 8 9 10 -0.2 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 y = 2.182005x + 0.002352 R2 = 0.99501738 R = 0.99950640 Weight: 1/x2 67 8 In terms of spinach MRLs, Figures 7, 8, and 9 provide three examples. Figure 7 shows the results for bifenthrin, which is quadratic through 3.5 orders of magnitude down to 0.5 ppb and has an MRL of 0.2 ppm in spinach. Diazinon, linear through 3.5 orders of magnitude down to 0.5 ppb, is shown in Figure 8, with an MRL in spinach of 0.7 ppm. Figure 7. Integrated peak (A), qualifiers (B) of the 0.5 ppb peak of bifenthrin as well as the full calibration curve (C). ×104 ×102 A B ×10 C 2 Acquisition time (min) Counts Relative concentration Relative abundance (%) Relative responses Bifenthrin 0 1 2 3 4 5 6 7 8 9 10 -0.2 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 y = 0.459107x2 + 18.988674x + 0.012908 R2 = 0.99465789 R = 0.99640909 Weight: 1/x2 + MRM CID at 25.0 (181.2 & 165.2) 14.14 14.16 14.18 14.20 14.22 14.24 14.26 14.28 14.30 14.32 Acquisition time (min) 14.14 14.16 14.18 14.20 14.22 14.24 14.26 14.28 14.30 14.32 2 3 4 5 6 7 8 14.202 min -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 181.2 & 165.2 181.2 & 166.2 166.2 & 165.2 Ratio = 87.4 (109.7%) Ratio = 52.5 (107.2%) 68 9 Figure 8. Integrated peak (A), qualifiers (B) of the 0.5 ppb peak for diazinon as well as the full calibration curve (C). ×103 ×102 A B C Acquisition time (min) Counts Relative concentration Relative abundance (%) Relative responses Diazinon 0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 y = 0.638177x – 3.640200E-004 R2 = 0.98721937 R = 0.99294761 Weight: 1/x2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 137.1 & 84.0 199.1 & 93.0 Ratio = 64.7 (96.5%) + MRM CID at 10.0 (137.1 & 84.0) 8.44 8.46 8.48 8.50 8.52 8.54 8.56 8.58 8.60 8.62 Acquisition time (min) 8.44 8.46 8.48 8.50 8.52 8.54 8.56 8.58 8.60 8.62 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 8.523 min 69 10 Lastly, boscalid is shown in Figure 9 with an MRL of 1 ppb; it is easily defined by a quadratic curve with four orders of magnitude down to 0.1 ppb. In the spinach sample, all three residues fell below the limit of quantification (LOQ) of the analytical method, and below their respective spinach MRLs. In the spinach samples, both boscalid and bifenthrin fell below the limit of detection (LOD) as no peak was detected for either residue. However, a small amount of diazinon was detected that was not present in the solvent blank. However, the area of the peak fell well below the LOQ and had an S/N very close to 3, and therefore approached or fell below the LOD. ×103 ×102 ×102 A B C Acquisition time (min) Counts Relative concentration Relative abundance (%) Relative responses -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 140.0 & 112.0 140.0 & 76.0 111.9 & 76.0 Ratio = 87.0 (104.4%) Ratio = 32.9 (99.3%) + MRM CID at 10.0 (140.0 & 112.0) 16.84 16.86 16.88 16.90 16.92 16.94 16.96 16.98 17.00 17.02 Acquisition time (min) 16.84 16.86 16.88 16.90 16.92 16.94 16.96 16.98 17.00 17.02 0 1 2 3 4 5 6 7 Boscalid 16.909 min 0 1 2 3 4 5 6 7 8 9 10 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 y = 0.635140x2 + 8.900401x + 0.002796 R2 = 0.97434891 R = 0.99281390 Weight: 1/x2 Figure 9. Integrated peak (A), qualifiers(B) of 0.1 ppb peak of boscalid as well as the full calibration curve (C). 70 11 The innovative HES 2.0 source enabled enhanced response stability for the analyzed pesticides over 400 replicate injections of the matrix-matched calibration standard at 50 ppb analyzed within a sequence with over 800 total injections. Figure 10 shows the 400 replicates of the 50 ppb matrix-matched standard, calculated as samples from the corresponding calibration curves across six sequences. The X-axis on the top corresponds to the number of injections of just the 50 ppb matrix-matched standard for robustness. The lower X-axis corresponds to the total injection number of spinach QuEChERS matrix, and therefore excludes blank injections. The concentration in ppb versus injection number plot shows how robust and accurate the analysis was. The results showed all but one point falling within ± 20% error, the usual % error given by GC/TQ data, and most points well within ± 10% of the actual. The %RSD for 400 replicates of all 114 residues are summarized in Table 2, with nearly 80% of the 114 residues having %RSD < 10%, and only five having %RSD above 20%. The robustness of the analysis and the instrument is clearly shown, with only inlet maintenance required between sequences and no further maintenance of the instrument needed. Figure 10. Calculated concentration of pesticides in 50 ppb matrix-matched standard over the course of 714 injections of spinach QuEChERS extract. 0 50 100 150 200 250 300 350 400 30 35 40 45 50 55 60 65 70 23 73 165 257 349 441 533 583 675 Adjusted 50 ppb spiked standard injection number ppb QuEChERS injection number Parathion BHC-alpha (benzene hexachloride) Pentachlorobenzonitrile Pentachlorothioanisole Fenthion 71 12 Table 2. Summary of 114 residues, including calibration range, calibration curve fit, and %RSD over the 400 injections of 50 ppb matrix-matched standard injections for robustness. Those results with %RSD > 20% are highlighted in red. Name Retention Time (minutes) Min Cal (ppb) Max Cal (ppb) Curve Fit Curve Weight %RSD Ethiolate 4.609 0.5 1,000 Linear 1/x2 7.6 Dichlorvos 4.826 0.5 1,000 Quadratic 1/x 10.4 Nicotine 5.424 0.5 1,000 Linear 1/x2 11.9 Biphenyl 5.615 1 1,000 Linear 1/x2 6.2 2‑Phenylphenol 6.457 5 1,000 Quadratic 1/x 7.5 Pentachlorobenzene 6.566 0.1 1,000 Linear 1/x2 4.8 Tecnazene 7.118 0.1 1,000 Linear 1/x2 3.9 Diphenylamine 7.186 0.1 1,000 Quadratic 1/x2 6.5 Ethoprophos 7.238 5 1,000 Quadratic 1/x 12.3 2,3,5,6‑Tetrachloroaniline 7.297 0.5 1,000 Linear 1/x2 4.9 Chlorpropham 7.325 0.5 1,000 Quadratic 1/x 7.5 Trifluralin 7.462 1 1,000 Quadratic 1/x2 6.6 Benfluralin 7.496 0.1 1,000 Quadratic 1/x2 7.2 BHC‑alpha (Benzene Hexachloride) 7.881 1 1,000 Linear 1/x2 1.0 2,6‑Diisopropylnaphthalene 8.02 5 1,000 Quadratic 1/x 5.5 Hexachlorobenzene 8.024 0.5 1,000 Quadratic 1/x 4.5 Ethoxyquin 8.034 0.1 1,000 Quadratic 1/x2 25.3 Dichloran 8.04 1 250 Linear 1/x2 7.7 Simazine 8.043 5 1,000 Linear 1/x2 7.0 Pentachloroanisole 8.073 0.1 1,000 Linear 1/x2 3.3 Atrazine 8.124 1 1,000 Linear 1/x2 5.5 Beta‑BHC 8.278 1 1,000 Linear 1/x2 4.9 Terbuthylazine 8.363 1 1,000 Linear 1/x2 14.6 BHC‑gamma (Lindane, Gamma HCH) 8.398 1 1,000 Linear 1/x2 4.5 Pentachloronitrobenzene 8.478 0.5 1,000 Quadratic 1/x2 33.7 Pentachlorobenzonitrile 8.515 0.1 1,000 Quadratic 1/x2 3.1 Diazinon 8.526 0.5 1,000 Linear 1/x2 5.8 Pyrimethanil 8.53 0.1 1,000 Linear 1/x2 5.0 BHC‑delta 8.763 1 1,000 Quadratic 1/x 11.4 Triallate 8.817 1 1,000 Quadratic 1/x2 4.4 Iprobenfos 8.942 0.5 1,000 Quadratic 1/x2 17.3 Pirimicarb 8.976 1 1,000 Linear 1/x2 8.9 Pentachloroaniline 9.178 0.1 1,000 Linear 1/x2 4.6 Propanil 9.193 0.5 1,000 Quadratic 1/x2 11.4 Metribuzin 9.256 0.5 1,000 Quadratic 1/x2 5.8 Dimethachlor 9.255 0.5 1,000 Linear 1/x2 6.0 Vinclozolin 9.372 0.1 1,000 Linear 1/x2 9.5 Chlorpyrifos‑methyl 9.404 0.1 1,000 Quadratic 1/x2 8.4 Parathion‑methyl 9.403 5 1,000 Quadratic 1/x 9.3 Ametryn 9.495 0.5 1,000 Quadratic 1/x2 6.5 Tolclofos‑methyl 9.496 5 1,000 Linear 1/x2 5.8 Prometryn 9.541 0.1 1,000 Quadratic 1/x2 6.9 Pirimiphos‑methyl 9.85 0.5 1,000 Quadratic 1/x2 8.9 Fenitrothion 9.855 0.1 1,000 Quadratic 1/x2 10.2 72 13 Name Retention Time (minutes) Min Cal (ppb) Max Cal (ppb) Curve Fit Curve Weight %RSD Ethofumesate 9.877 0.5 1,000 Quadratic 1/x2 6.5 Malathion 9.995 10 1,000 Quadratic 1/x 20.0 Pentachlorothioanisole 10.032 0.1 1,000 Linear 1/x2 3.5 Metolachlor 10.166 5 1,000 Linear 1/x2 6.2 Fenthion 10.187 0.5 1,000 Linear 1/x2 2.4 Chlorpyrifos 10.224 0.5 1,000 Linear 1/x2 4.5 Parathion 10.242 1 1,000 Linear 1/x2 2.7 Triadimefon 10.275 0.5 1,000 Quadratic 1/x 11.6 Tetraconazole 10.319 0.5 1,000 Quadratic 1/x2 7.5 DCPA (Dacthal, Chlorthal‑dimethyl) 10.328 0.1 1,000 Quadratic 1/x 7.3 Isocarbophos 10.346 0.1 1,000 Linear 1/x2 4.7 Butralin 10.496 0.5 1,000 Linear 1/x2 5.7 Cyprodinil 10.672 0.5 1,000 Linear 1/x2 7.8 MGK‑264 10.709 0.1 1,000 Linear 1/x2 7.3 Pendimethalin 10.797 0.5 1,000 Linear 1/x2 4.6 Penconazole 10.826 0.5 1,000 Quadratic 1/x2 9.8 Heptachlor Exo‑epoxide 10.904 10 1,000 Quadratic 1/x 14.1 Fipronil 10.915 0.5 1,000 Quadratic 1/x 10.7 Triadimenol 11.008 0.5 1,000 Quadratic 1/x2 7.1 Quinalphos 11.01 1 1,000 Linear 1/x2 3.3 Chlordane‑trans 11.326 100 1,000 Quadratic 1/x 13.6 DDE‑o,p' 11.363 0.1 500 Quadratic 1/x2 11.8 Mepanipyrim 11.458 0.5 1,000 Linear 1/x2 7.1 Flutriafol 11.596 0.5 1,000 Quadratic 1/x 6.6 Flutolanil 11.664 0.5 1,000 Quadratic 1/x2 4.6 Napropamide 11.697 0.5 1,000 Linear 1/x2 8.5 Hexaconazole 11.73 0.5 1,000 Linear 1/x2 8.6 Isoprothiolane 11.776 0.1 1,000 Linear 1/x2 4.8 Prothiofos 11.782 0.5 1,000 Linear 1/x2 5.2 Fludioxonil 11.819 0.5 1,000 Quadratic 1/x2 6.9 DEF 11.871 0.5 1,000 Linear 1/x2 7.4 DDE‑p,p' 11.91 0.1 500 Quadratic 1/x2 9.8 Oxyfluorfen 11.988 0.5 1,000 Linear 1/x2 5.2 Myclobutanil 12.013 0.5 1,000 Quadratic 1/x2 8.2 Buprofezin 12.066 1 1,000 Quadratic 1/x2 6.2 Bupirimate 12.089 0.5 1,000 Linear 1/x2 6.7 Kresoxim‑methyl 12.093 0.5 1,000 Quadratic 1/x 5.9 Chlorfenapyr 12.326 0.5 1,000 Quadratic 1/x2 4.8 Endrin 12.425 5 1,000 Quadratic 1/x 10.3 Ethion 12.718 5 1,000 Linear 1/x2 7.5 Benalaxyl 13.167 0.5 1,000 Quadratic 1/x2 7.7 Trifloxystrobin 13.223 0.5 1,000 Linear 1/x2 5.3 Quinoxyfen 13.222 0.1 1,000 Linear 1/x2 7.7 Endosulfan Sulfate 13.328 1 1,000 Quadratic 1/x2 13.2 Tebuconazole 13.565 0.5 1,000 Quadratic 1/x2 6.6 Nuarimol 13.595 0.5 1,000 Quadratic 1/x2 6.4 Triphenyl Phosphate 13.659 0.5 1,000 Quadratic 1/x2 6.1 73 14 Name Retention Time (minutes) Min Cal (ppb) Max Cal (ppb) Curve Fit Curve Weight %RSD Piperonyl butoxide 13.662 0.5 1,000 Quadratic 1/x2 5.9 Epoxiconazole 13.876 0.5 1,000 Quadratic 1/x2 8.5 Spiromesifen 14.014 1 1,000 Quadratic 1/x 15.5 Tetramethrin I 14.207 0.5 1,000 Quadratic 1/x 7.4 Bifenthrin 14.179 0.5 1,000 Quadratic 1/x2 4.4 EPN 14.226 10 1,000 Quadratic 1/x2 5.8 Bromopropylate 14.221 0.5 1,000 Quadratic 1/x2 5.2 Etoxazole 14.375 0.5 1,000 Quadratic 1/x2 6.3 Tebufenpyrad 14.398 0.5 1,000 Quadratic 1/x2 6.9 Fenamidone 14.449 0.5 1,000 Quadratic 1/x2 7.4 Tetradifon 14.72 5 1,000 Quadratic 1/x2 9.2 Metrafenone 15.648 0.5 1,000 Quadratic 1/x2 6.4 Bitertanol I 15.857 0.5 1,000 Quadratic 1/x2 9.7 Spirodiclofen 15.976 5 1,000 Quadratic 1/x2 49.3 Pyridaben 16.081 0.5 1,000 Quadratic 1/x2 5.8 Cyfluthrin I 16.484 5 1,000 Quadratic 1/x2 22.2 Fenbuconazole 16.535 5 1,000 Quadratic 1/x2 9.5 Cypermethrin I 16.8 5 1,000 Quadratic 1/x 23.8 Boscalid 16.909 0.1 1,000 Quadratic 1/x2 8.4 Ethofenprox 17.096 0.5 1,000 Quadratic 1/x2 5.9 Difenoconazole I 18.148 1 1,000 Quadratic 1/x 12.4 Azoxystrobin 18.787 10 1,000 Quadratic 1/x2 11.2 Dimethomorph I 19.175 5 1,000 Quadratic 1/x 15.3 74 www.agilent.com DE34182072 This information is subject to change without notice. © Agilent Technologies, Inc. 2024 Printed in the USA, May 3, 2024 5994‑7385EN Conclusion The analytical performance of the Agilent 7010 Series triple quadrupole mass spectrometer (GC/TQ) upgraded with the HES 2.0 electron ionization (EI) source was demonstrated for multiresidue pesticide analysis. The system demonstrates analytical sensitivity as the same or better than the original HES as well as excellent accuracy and robustness. The 7010 GC/TQ with HES 2.0, coupled with an Agilent 8890 GC with an MMI inlet and 15 m × 15 m mid-column backflush configuration minimizes instrument downtime by allowing for inlet maintenance without requiring the cooling of the heated zones. With no impact to the analytical method or degradation of the instrument performance, this application demonstrates the ability of the instrument to provide robust and reliable analytical results, including for challenging matrices. References 1. Anastassiades, M.; Lehotay, S. J.; Stajnbaher, D.; Schenck, F. J. Fast and Easy Multiresidue Method Employing MeCN Extraction/Partitioning and “Dispersive Solid-Phase Extraction” for the Determination of Pesticide Residues in Produce. J. AOAC Int. 2003, 86, 412− 431. 2. Pesticide Residues in Foods by MeCN Extraction and Partitioning with Magnesium Sulfate. Official Methods of Analysis of AOAC International; AOAC International: Gaithersburg, MD, 2007; Method 2007.1. 3. Lehotay, S. J. QuEChERS Sample Preparation Approach for Mass Spectrometric Analysis of Pesticide Residues in Foods. Methods Mol. Biol. 2011, 747, 65–91. doi: 10.1007/978-1-61779-136-9_4. PMID: 21643905. 4. Alder, L.; Greulich, K.; Kempe, G.; Vieth, B. Residue Analysis of 500 High Priority Pesticides: Better by GC-MS or LC–MS/ MS? Mass Spectrom. Rev. 2006, 25, 838–865. 5. Chamkasem, N.; Ollis, L. W.; Harmon, T.; Lee, S.; Mercer, G. Analysis of 136 Pesticides in Avocado Using a Modified QuEChERS Method with LC-MS/MS and GC-MS/MS. J. Agric. Food Chem. 2013, 61(10), 2315–2329. DOI: 10.1021/jf304191c 6. Lehotay, S. J.; Han, L.; Sapozhnikova, Y. Automated Mini-Column Solid-Phase Extraction Cleanup for High-Throughput Analysis of Chemical Contaminants in Foods by Low-Pressure Gas Chromatography-Tandem Mass Spectrometry. Chromatographia 2016, 79(17), 1113–1130. DOI: 10.1007/s10337-016-3116-y. 7. Andrianova, A.; Zhao, L. Five Keys to Unlock Maximum Performance in the Analysis of Over 200 Pesticides in Challenging Food Matrices by GC/MS/MS. Agilent Technologies application note, publication number 5994-4965EN, 2022. Application Note Food Authors Anastasia A. Andrianova and Limian Zhao Agilent Technologies, Inc. Abstract This application note presents results for the sensitive and robust quantitation of 246 pesticides in black tea extract with the Agilent 7010D Triple Quadrupole Mass Spectrometer (GC/TQ) featuring a second-generation High Efficiency Source 2.0 (HES 2.0), that addresses the challenges posed by residual pesticide analysis in complex matrices. By optimizing sample preparation and using state-of-the-art GC/MS hardware, including ion source technology and midcolumn backflushing, excellent calibration performance and sensitivity at low-ppb levels were achieved. The method demonstrated exceptional ruggedness and robustness over 800 consecutive injections of a black tea extract spiked with pesticides at 2 ppb, with high precision and low RSDs, ensuring prolonged instrument uptime and maximum throughput. The demonstrated limits of quantitation (LOQs) were as low as 0.01 ppb for over a third of evaluated compounds, and the calibration range spanned up to five orders of magnitude while meeting SANTE 11312/2021 guidelines. This application note highlights intelligent GC/TQ features, such as early maintenance feedback and an instrument health status dashboard, maintaining confidence in the results for a high-throughput analysis. The updated data acquisition platform provides enhanced user experience, including a new implementation of the retention time locking functionality. Brewing Excellence: Quantitating Over 200 Pesticides in Black Tea with Steady Performance and Maximized Uptime by GC/MS/MS 76 Return to Table of Contents 2 Introduction Tea is among the most common nonalcoholic beverages consumed worldwide. Like many foods, tea cultivation relies heavily on pesticide application to combat pests, leading to concerns over pesticide residues intensifying.1 Assessing pesticide levels in tea is essential for evaluating safety, and is required by many regulatory bodies, including the European Commission and the US Environmental Protection Agency.2,3 A complete workflow for pesticide testing in tea includes sample extraction via QuEChERS, followed by extract cleanup, and subsequent testing with liquid and gas chromatography coupled with triple quadrupole mass spectrometry (LC/TQ and GC/TQ).4 Workflow performance should enable sufficient method sensitivity, calibration range, pesticide recovery from the extraction, and precision. Sensitivity requirements are set based on the maximum residue levels (MRLs), which are the highest levels of pesticide residue that are legally allowed in or on food or feed when pesticides are applied correctly. The ability to calibrate over a wide dynamic range allows the varying MRLs for individual compounds monitored in the commodity, which can vary from 10 ppb to 100 ppm. When a specific pesticide lacks an established MRL, a default limit of 10 ppb is commonly applied. Efficiency of extraction and cleanup are characterized in terms of recovery of matrix spikes, and precision is expressed in terms of relative standard deviation (RSD) of repeat analyses. This application note presents a complete GC/TQ workflow solution for the accurate and reliable analysis of 246 volatile and semivolatile pesticides in black tea. Excellent analytical performance of the workflow was achieved through a combination of cutting-edge technology and optimized methodology that included: – Sample preparation using QuEChERS extraction, followed by EMR mixed-mode pass-through cleanup using Agilent Captiva EMR–GPD cartridges – Agilent 8890 GC hardware and GC supplies – Novel electron ionization (EI) source technology with HES 2.0 – Built-in GC/TQ MS intelligence and new software functionality for method setup, maintenance, and system health evaluation The presented workflow allowed for quantitating 246 pesticide residues in black tea with LOQs as low as 0.01 ppb for 34% of the targets, at or below 0.1 ppb for 74% of compounds, and below 2 ppb for 96%. Matrix-matched calibrations demonstrated excellent accuracy over a wide dynamic range, spanning up to five orders of magnitude over 0.01 to 1,000 ppb in the complex black tea extract. Method ruggedness was demonstrated through maintaining measurement accuracy with good precision (RSDs < 20% for 176 compounds) for black tea extract spiked at 2 ppb sequentially analyzed over 800 runs spanning 17 days of continuous analysis. The new HES 2.0 ion source is equipped with a novel dipolar radiofrequency (RF) lens that redirects the carrier gas ions and, as a result, enables improved system robustness and maximizes uptime while maintaining unparalleled analytical sensitivity. Experimental GC/TQ analysis The 8890 GC and 7010D GC/TQ systems (Figure 1A) were used and configured to achieve the best sensitivity, maintain a wide calibration range, and provide the most rugged method performance. The GC was configured with the Agilent 7693A automatic liquid sampler (ALS) and 150-position tray. The system used a multimode inlet (MMI) operated in temperature-programmed splitless injection mode (also known as cold splitless). The injection parameters were optimized for maximizing sensitivity while limiting carryover. Midcolumn backflush capability was provided by the Agilent Purged Ultimate Union (PUU) installed between two identical 15 m columns, and the Agilent 8890 pneumatic switching device (PSD) module (Figure 1B). The instrument operating parameters are listed in Table 1. Data were acquired in dynamic MRM (dMRM) mode, which provides the capability for large multi-analyte assays and the accurate quantitation of narrow peaks by an automatically determined most-efficient dwell time distribution. The dMRM capability enabled successful analysis for a large panel of 246 pesticides, with 749 total MRM transitions with up to 64 concurrent MRMs. Furthermore, dMRM allows the analyst to add and remove additional analytes with ease. The acquisition method was retention time locked to match the retention times in the Agilent MassHunter Pesticides and Environmental Pollutants MRM Database 4.0 (P&EP 4.0)5 , 77 3 Figure 1. The Agilent 8890 GC system with the Agilent 7010D GC/TQ system (A) and system configuration (B). PSD (helium) Agilent 8890 GC Liquid Injector Multimode inlet (helium) Agilent 7010D TQ MS HES 2.0 Agilent HP-5ms UI, 15 m × 0.25 mm, 0.25 µm Agilent HP-5ms UI, 15 m × 0.25 mm, 0.25 µm A B which was used to seamlessly create the MS method. The use of P&EP 4.0 increased the ease and speed of setting up a targeted dMRM method. High method selectivity in the presence of coeluting matrix components was achieved by selecting the best MRM transitions from up to nine transitions available for each compound in the P&EP 4.0 database. Three targets were not available in the P&EP 4.0 database (flonicamid, bioallethrin, and cycloxydim). For these compounds, MRM transitions were developed using Agilent MassHunter Optimizer for GC/TQ, operating in Start from Full Scan mode. The Optimizer is fully integrated into MassHunter Acquisition 13.0 for GC/MS (Figure 2). The acquisition method was retention time locked to the P&EP database with chlorpyrifos-methyl eluting at 9.143 minutes. The retention time locking functionality, integrated in MassHunter Acquisition 13.0 for GC/MS, has an updated user-friendly and intuitive interface (Figure 3). It allows for semi-automated or manual compound selection, provides a choice to use three or five points for retention time locking calibration, and features both a visual and quantitative assessment of the calibration curve fit, while providing the tools to maintain excellent precision of retention times, even after column trimming. Full scan data acquisition mode was used for the initial screening of the matrix extract. This screening was used to evaluate in-source loading, and for monitoring the efficiency of the sample cleanup procedure that followed QuEChERS extraction. Agilent MassHunter Workstation software, including Agilent MassHunter Acquisition 13.0 for GC/MS, MassHunter Quantitative Analysis 12.1, and MassHunter Qualitative Analysis 12.0 packages were used in this work. 78 4 Table 1. Agilent 8890 GC system with Agilent 7010D gas chromatograph and mass spectrometer conditions for pesticide analysis. Parameter Value GC Agilent 8890 with fast oven, auto injector and tray Inlet MMI Mode Cold splitless Purge Flow to Split Vent 60 mL/min at 3 min Septum Purge Flow 3 mL/min Septum Purge Flow Mode Switched Injection Volume 1.0 µL Injection Type Reversed 2-layer (L2, L1) L1 Airgap 0.2 µL L2 Volume (ISTD) 0.2 μL L2 Airgap 0.2 μL Gas Saver On at 30 mL/min after 5 min Inlet Temperature 60 °C for 0.1 min, then to 280 °C at 600 °C/min, hold for 5 min, then to 325 °C at 600 °C/min Postrun Inlet Temperature 310 °C Postrun Total Flow 25 mL/min Carrier Gas Helium Inlet Liner Agilent Ultra Inert 2 mm dimpled liner (p/n 5190-2297) Oven Oven Program 60 °C for 1 min; 40 °C/min to 170 °C; Hold 0 min; 10 °C /min to 310 °C; Hold 2.25 min Total Run Time 20 min Postrun Time 1.5 min Equilibration time 0.5 min Column 1 Type Agilent HP-5ms UI, 15 m × 0.25 mm, 0.25 µm (p/n 19091S-431UI-KEY) Control Mode Constant flow Flow 1.0 mL/min (then retention time locked) Inlet Connection MMI Outlet Connection PSD (PUU) PSD Purge Flow 5 mL/min Postrun Flow (Backflushing) –7.873 Parameter Value Column 2 Type Agilent HP-5ms UI, 15 m × 0.25 mm, 0.25 µm (p/n 19091S-431UI-KEY) Control Mode Constant flow Flow 1.2 mL/min (then retention time locked) Inlet Connection PSD (PUU) Outlet Connection MSD Postrun Flow (Backflushing) 8.202 MSD Model Agilent 7010D Source HES 2.0 Vacuum Pump Performance turbo Tune File atunes.eihs2.jtune.xml Solvent Delay 3.75 min Quad Temperature (MS1 and MS2) 150 °C Source Temperature 280 °C Mode dMRM or Scan He Quench Gas 2.25 mL/min N2 Collision Gas 1.5 mL/min MRM Statistics Total MRMs (dMRM mode) 749 Minimum Dwell Time 5.42 ms Minimum Cycle Time 85.01 ms Maximum Concurrent MRMs 64 EM Voltage Gain Mode 10 Scan Parameters Scan Type MS1 Scan Scan Range 45 to 450 m/z Scan Time (ms) 220 Step Size 0.1 amu Threshold 0 EM Voltage Gain Mode 1 79 5 Figure 2. Agilent MassHunter Optimizer software for GC/TQ, used for the automated development of MRM transitions. 80 6 Figure 3. New Agilent retention time locking software in Agilent MassHunter Acquisition 13.0 for GC/MS. 81 7 Sample preparation Black tea powder was obtained from a local grocery store. Black tea powder (2 g) was extracted with a modified QuEChERS extraction using acetonitrile (ACN) with 2% formic acid and EN extraction salt. The crude tea extract then was mixed with 2% of acidic buffer. The sample mixture was cleaned by EMR mixed-mode pass-through cleanup using Agilent Captiva EMR–GPD 6 mL. The sample eluent was dried with anhydrous MgSO4 to remove water residue completely before GC/MS/MS analysis. The sample preparation procedure flowchart is shown in Figure 4, and details will be discussed in a separate application note. The entire sample preparation procedure resulted in a 5x dilution factor. Pesticide standards Agilent GC pesticide standards 1 through 12 (part numbers PSM-100-A through -L) and Agilent GC/LC pesticide standards 1, 2, and 3 (part numbers PSM-100-AA, PSM-100-AB, PSM-100-AC) were used for preparing matrix matched calibration standards. A combination of the 15 used standards yielded a mix of 246 pesticides commonly regulated by the FDA, USDA, and other global governmental agencies. Figure 4. QuEChERS sample preparation and cleanup method for black tea. 2 g of black tea Agilent QuEChERs EN extraction kit Mechanical shaker Centrifuge – Dry sample hydration and equilibration – Addition of extraction solvent Sample cleanup Sample extraction Black tea extract on Agilent Captiva EMR–GPD Sample eluent drying with MgSO4 Sample analysis on the Agilent 7010D GC/TQ Crude extract 2 g of – Premixing with 2% acidic buffer – Loading 2.5 to 3 mL on cartridge 82 8 Matrix-matched calibration Calibration performance was evaluated using a series of matrix matched calibration standards, ranging from 0.01 to 1,000 ppb, including 0.01, 0.05, 0.1, 0.5, 1, 2, 5, 10, 50, 100, 200, 500, and 1,000 ppb. The standard parathion-d10 (Agilent QuEChERS IS standard number 6, part number PPS-610-1) was used as the internal standard for quantitation of the target pesticides. It was added at 0.2 μL through reversed sandwich injection with the ALS to a final concentration of 10 ppb in the injected sample. An appropriate calibration function, either linear or quadratic, guided by the lower value of the relative standard error (RSE) was used. A weighting factor of 1/x allowed for maintaining accuracy across the entire calibration range. The deviation of the back-calculated concentrations of the calibration standards from the true concentrations, using the calibration curve in the relevant region, did not exceed ± 20%. The concentrations expressed in ppb (w:v) correspond to the pesticide concentration in the injected sample. The QuEChERS sample preparation procedure described in the Sample preparation section resulted in a dilution factor of 5. Hence, the concentrations measured in the injected samples were five times lower than the corresponding concentration in the black tea sample, expressed in µg/kg. Analyte protectants were not used in this work. The preliminary investigation showed that analyte protectants did not have a response enhancement effect on most of the compounds when analyzed in the rich and complex black tea matrix. It is of note that analyte protectants often significantly enhance target analyte response and stability as is described in-depth in the peer-reviewed literature.6 Recovery evaluation Sample preparation efficiency was evaluated by performing recovery studies. The surrogate black tea matrix was spiked at two different levels, 10 and 50 ppb, with six replicates at each level. The samples were extracted and cleaned up. Assuming 100% recovery, pesticide concentrations in the prespiked samples were expected to be 2 and 10 ppb due to a 5x dilution rate. A blank black tea extract was postspiked with the pesticide standard to achieve final concentrations of 2 and 10 ppb. The prespiked and postspiked samples were analyzed, and the response areas were compared. The recovery was measured as the ratio of the pesticide peak area in the prespiked sample to the area in the postspiked samples. Results and discussion As contemporary GC/MS technology continues to advance, so do the expectations for high sample throughput, intuitive user-friendly system setup and configuration, and streamlined maintenance. The demands for enhanced analytical performance are driven by the evolving regulations in pesticide residue analysis and food safety. Several best practices to achieve the best GC/TQ performance in pesticide residue analyses were described in Agilent application note 5994-4965EN.7 This work presents a complete workflow for analyzing 246 pesticides in black tea, while implementing the previously described best practices and offering further method and technology enhancements. The innovative HES 2.0 yielded enhanced GC/TQ performance stability, as evidenced by precise results at the low concentration of 2 ppb over 800 consecutive injections of complex black tea extract. The technology and method enhancements that enable unparalleled GC/TQ performance while ensuring stable and reliable results in a high-throughput setting are outlined in this application note and grouped into four categories: sample preparation, GC instrumentation and supplies, MS electron ionization technology advancements, and instrument intelligence and software functionality. Effective matrix cleanup Sample preparation is a key component of performing successful pesticide analysis. Performing analysis of samples prepared by QuEChERS extraction, particularly when analyzing complex pigmented samples such as black tea without adequate cleanup, can lead to increased system maintenance. The parts of the system that are affected without adequate sample cleanup include liner replacement, GC column trimming, and inlet and MS source cleaning. As a result, throughput is decreased. Further, the presence of large amounts of matrix can affect the accuracy of results, often most pronounced with difficult to analyze pesticides. The EMR mixed-mode pass-through cleanup using Captiva EMR with Carbon S cartridges is a simplified procedure that demonstrates an improvement on both sample matrix removal, and overall recovery and reproducibility of targets. As shown in Figure 5, the abundance of the TIC signal in full scan data acquisition mode was noticeably reduced for black tea extract after cleanup when comparing the crude extracts before cleanup. 83 9 Performing matrix screening in full scan data acquisition mode, as shown in Figure 5, facilitates the evaluation of in-source matrix loading, as discussed in 5994-4965EN.7 Every MS source has a limitation on the amount of material present in the source, at any point in time, to maintain optimal performance. Quantitation accuracy of the analysis can be significantly compromised if the source is overloaded with matrix. Hence, it is essential to analyze the matrix in full scan mode to evaluate the TIC and maintain optimal GC/TQ performance. For best performance with the HES 2.0 source, it is recommended to have a TIC full scan abundance below 7 × 107 counts when analyzing with an EM gain set to 1. As shown in Figure 5, black tea extract is complex, featuring abundant matrix components. Cleaning up the extract is key to lowering the matrix background that leads to adequate in-source loading, enhancing selectivity and sensitivity, widening the dynamic range, and allowing for less frequent system maintenance, increasing productive uptime. GC instrumentation and supplies Midcolumn backflushing: The Agilent 8890 GC provides an easy-to-use midcolumn backflushing functionality that results in increased sample throughput through shorter analysis times and less frequent column maintenance. Midcolumn backflush allows for the elution of the high boiling point matrix components from the column in a shorter time and without eluting high-boiling matrix into the MS. Midcolumn backflushing is a technique in which the carrier gas flow is reversed after the last analyte has exited the column and all MS data are collected. The oven is then held at the final temperature in postrun mode, with the reversed carrier gas flow through the first column. The high boilers are eluted back out the head of the column and into the split vent trap. The ability to reverse the flow is provided by the PUU. The PUU is a tee that is inserted, in this case, between two identical 15 m columns. During the analysis, a small makeup flow of carrier gas from the 8890 PSD is used to sweep the connection. During backflushing, the Figure 5. Scan TIC of black tea extract. The green trace corresponds to matrix sample with Agilent Captiva-EMR cleanup, and the red trace corresponds to matrix sample without cleanup. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3.0 3.1 3.2 3.3 3.4 3.5 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 15 min ×108 Acquisition time (min) Counts Extract before cleanup After cleanup 84 10 makeup flow from the PSD is raised to a much higher value, sweeping high boilers backward out of the first column while simultaneously providing forward flow in the second column. For the configuration in this application, the backflushing time was 1.5 minutes. More details about using PSD for backflushing in the Agilent 8890 GC system can be found in Agilent application note 5994-0550EN.8 Figure 6 illustrates the effectiveness of the backflush technique in reducing cycle time without carryover of the black tea matrix. The cycle time was reduced by 50% and the columns did not have to be exposed to the higher bake-out temperatures for an extended time. Using backflushing, excess column bleed and heavy residues are not introduced into the MSD, thereby reducing ion source contamination. 1 1 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 It took an additional 20 minutes with the columns heated to 310 °C to remove these high boilers Run stopped at 20 minutes and backflushed for 1.5 minutes The following blank has no carryover peaks Run stopped at 20 minutes, no backflushing, no bakeout The following blank has carryover peaks Siloxanes from the septum Siloxanes from the septum Acquisition time (min) Figure 6. TIC Scan chromatograms of black tea extract, followed by analysis of an instrument blank with: column bake-out, with backflush, and without backflush or bake-out. 85 11 The backflush setup process has been simplified with the introduction of new tools that allow for making a capillary flow technology (CFT) connection with ease. These tools include the gold-plated flexible metal ferrules (part number G2855-28501) and the GC column installation preswaging tool for Flexible Metal ferrules into Capillary Flow Technology devices (part number G3440-80227)—shown in Figure 7. Figure 7. Flexible metal ferrules (part number G2855-28501) (A) and the GC column installation preswaging tool for Flexible Metal ferrules into Capillary Flow Technology devices (part number G3440-80227) (B). A B Additionally, MassHunter Acquisition 13.0 for GC/MS provides intuitive guides for backflushing setup and review. Figure 8 shows the backflush overview tab in the GC Method Editor in MassHunter Acquisition 13.0 for GC/MS. Figure 8. Backflush summary in Agilent MassHunter Acquisition 13.0 for GC/MS. 86 12 GC injection optimization: Efficiently volatilizing the sample in the GC inlet is an essential component of successful GC/MS analysis. Various sample introduction techniques are aimed at preserving thermally labile and active compounds. In this work, cold splitless and solvent vent injection modes were evaluated. As shown on the left in Figure 9, the use of solvent vent mode for analyzing black tea extract resulted in very large caffeine carryover into the subsequent analyses. To reduce caffeine carryover, cold splitless injection mode was used (Figure 9, at right). Increase of the splitless purge time to 3 minutes resulted in enhanced method sensitivity without deteriorating the chromatographic peak shape for the targets. 0 1 2 3 4 5 0 1 2 3 4 5 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 0 1 2 3 4 5 0 1 2 3 4 5 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Black tea in solvent vent mode Following ACN blank in solvent vent mode Black tea in cold splitless mode Following ACN blank in cold splitless mode 1 2 3 4 8.6 8.7 8.8 8.9 9.0 9.1 Five sequential blanks Caffeine Caffeine Clean blank Same scale ×108 ×108 ×108 Same scale ×108 ×108 Acquisition time (min) Counts Counts Counts Counts Figure 9. Injection optimization for analyzing black tea: cold splitless (on the right) reduces carryover of caffeine compared to solvent vent mode (on the left). 87 13 HES 2.0: Novel electron ionization (EI) source technology Equipped with the novel HES 2.0 EI source, the 7010D demonstrated sensitivity that allows for ultra-trace level detection when analyzing pesticides. The new HES 2.0 ion source is equipped with a novel dipolar RF lens that redirects the carrier gas ions and, as a result, enables improved system robustness and unparalleled analytical sensitivity. Figure 10 shows MRM chromatograms for selected pesticides at 0.01 ppb in black tea extract. The overlaid chromatograms show repeatability over seven replicate injections, and the response RSD% as a measure of precision. Appendix Table 1 shows the LOQs for all analyzed pesticides. LOQs as low as 0.01 ppb were observed for 34% of the targets, at or below 0.1 ppb for 74% of compounds, and below 2 ppb for 96%. The number of compounds expressed in percent, with their respective LOQs, are plotted in Figure 11. Figure 10. MRM chromatograms with seven replicate injections for the selected pesticides at the LOQ of 0.01 ppb in black tea extract and their calibration curves. Bromophos-ethyl Concentration (ng/mL) 0 200 400 600 800 1,000 Concentration (ng/mL) 0 200 400 600 800 1,000 Concentration (ng/mL) 0 200 400 600 800 1,000 Concentration (ng/mL) 0 200 400 600 800 1,000 Concentration (ng/mL) 0 200 400 600 800 1,000 -0.2 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 y = 23,817.677338x + 874.434909 R2 = 0.99951780 R = 0.99991699 RSE = 9.1 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 11.3 11.4 11.5 11.6 Chlorfenson RSD = 9% 0.01 ppb N = 7 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 10.9 11.0 11.1 11.2 11.3 Bromophos ethyl RSD = 8% 0.01 ppb N = 7 0.01 to 1,000 ppb 0.01 to 1,000 ppb 0.01 to 1,000 ppb 0.01 to 1,000 ppb 0.01 to 1,000 ppb Chlorfenson -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 y = 103,879.642468x + 3,894.601308 R2 = 0.99936894 R = 0.99990218 RSE = 10.2 DDE-o,p' 0 1 2 3 4 5 6 7 8 y = 83,323.889419x + 6,145.523095 R2 = 0.99949459 R = 0.99992324 RSE = 6.4 DDE-p,p' 0 1 2 3 4 5 6 y = 62,964.732203x + 6,196.477029 R2 = 0.99949744 R = 0.99993069 RSE = 6.8 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 11.0 11.2 11.4 11.6 11.8 DDE -o,p' RSD = 5% 0.01 ppb DDE -p,p' RSD = 6% 0.01 ppb N = 7 Chloropropylate 0 1 2 3 4 5 6 7 y = 72,085.499690x + 3,773.703370 R2 = 0.99957751 R = 0.99991508 RSE = 7.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 12.05 12.15 12.25 12.35 Chloropropylate RSD = 12% 0.01 ppb N = 7 ×102 ×103 ×103 ×107 ×107 ×107 ×107 ×103 ×108 Acquisition time (min) Acquisition time (min) Acquisition time (min) Acquisition time (min) Counts Counts Counts Counts Responses Responses Responses Responses Responses 88 14 Figure 11. Percentage of compounds with their respective LOQ levels (in ppb) in black tea extract. 0% 20% 40% 60% 80% 100% 120% 0.01 0.05 0.1 0.5 1 2 10 Percent of compounds (out of 246) with LOQ at or belo w LOQ (in ppb) Figure 10 also shows the matrix-matched calibration performance in black tea extract with excellent linearity maintained over five orders of magnitude, ranging from 0.01 to 1,000 ppb. All calibration curves were inspected, and if needed, were trimmed to comply with SANTE 11312/2021 guidelines.2 Appendix Table 1 provides information on calibration ranges and quality of calibration fit for all compounds. The R2 correlation coefficient for all targets was > 0.99. The RSE was used as an additional criterion for demonstrating the calibration curve quality. The RSE provides an improved criterion for evaluation of calibration curves, as it is consistent for evaluation of all curve fitting types.9 In this work, the calibration curves for all compounds had RSE values below 20. For the compounds for which quadratic calibration fit was used, a linear calibration curve fit can be used instead by narrowing the calibration range. For example, oxyfluorfen could be calibrated over five orders of magnitude, from 0.01 to 1,000 ppb using quadratic calibration fit with R2 = 0.9995 and RSE = 14. Alternatively, a linear calibration fit could be applied over a calibration range of 0.01 to 500 ppb with R2 = 0.9960 and RSE = 26. The selection of the calibration curve fit was guided by the lower RSE value. Some pesticides are known to present a particular challenge for analysis. As stated in the EURL Analytical Observation Report10, captan and folpet are analytically among the most challenging pesticides due to their nonamenability to LC/TQ, and their tendency to degrade both in solution as well as in the GC inlet. Figure 12 demonstrates that captan and folpet could be quantitated with great precision at LOQs as low as 2 and 0.5 ppb, respectively. Freshly diluted standard, acidified sample extracts, and optimized injection conditions with cold splitless injection were key to achieving high recoveries and precision in analyzing captan and folpet. Deltamethrin, a synthetic pyrethroid, elutes at the end of the chromatographic run and is also known to be difficult for GC/MS analysis.11 As shown in Figure 12, deltamethrin could be reliably quantitated down to 0.5 ppb using the developed method. Other compounds shown in Figure 12 include organochlorine pesticides, aldrin, dieldrin, and endrin, and the two most widely used multipurpose pyrethroids, cypermethrin and cyfluthrin, quantitated with great precision and excellent linearity over a wide dynamic range. 89 15 Figure 12. MRM chromatograms with eight replicate injections for the selected challenging pesticides. Included are LOQ levels and their calibration curves. 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 9.8 9.9 10.0 Aldrin RSD = 8% 0.5 ppb N = 8 Aldrin 0 1 2 3 4 5 6 7 y = 6,985.401438x + 203.727553 R2 = 0.99952950 R = 0.99993060 RSE = 8.8 0.4 0.5 0.6 0.7 0.8 0.9 1.0 10.6 10.7 10.8 10.9 11.0 Folpet RSD = 7% 0.5 ppb Folpet 0 50 100 150 200 0 1 2 3 4 5 6 y = 2,692.336742x + 830.709066 R2 = 0.99889601 R = 0.99995851 RSE = 16.4 0.5 to 200 ppb 1 to 500 ppb 0.05 to 1,000 ppb 0.5 to 1,000 ppb 0.1 to 500 ppb 0.1 to 1,000 ppb 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 11.6 11.8 12.0 12.2 Endrin RSD = 9% 0.5 ppb Dieldrin RSD = 12% 0.5 ppb 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 10.55 10.65 10.75 10.85 10.95 Captan RSD = 10% 2 ppb N = 8 N = 8 Endrin 0 1 2 3 4 5 6 7 y = 7,758.571223x + 792.052413 R2 = 0.99958608 R = 0.99993430 RSE = 5.5 Captan -1 0 1 2 3 4 5 6 7 y = 1,198.868510x + 256.553484 R2 = 0.99734455 R = 0.99927622 RSE = 18.9 0.50 0.54 0.58 0.62 0.66 0.70 0.74 0.78 0.82 0.86 0.90 0.94 0.98 1.02 18.0 18.1 18.2 18.3 18.4 Deltamethrin RSD = 10% 0.5 ppb N = 8 Deltamethrin 0 100 200 300 400 500 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 y = 1,873.813407x + 335.177249 R2 = 0.99627763 R = 0.99887520 RSE = 17.2 1.0 1.4 1.8 2.2 2.6 3.0 3.4 3.8 4.2 4.6 5.0 16.1 16.3 16.5 16.7 Cyfluthrins RSD = 6% 1 ppb Cypermethrins RSD = 5% 1 ppb N = 8 Cypermethrin I 0 200 400 600 800 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 y = 12,792.103487x + 12,912.393135 R2 = 0.99849640 R = 0.99974393 RSE = 16.6 Concentration (ng/mL) 0 200 400 600 800 Concentration (ng/mL) 0 200 400 600 800 1,000 1,000 1,000 Concentration (ng/mL) Concentration (ng/mL) Concentration (ng/mL) 0 100 200 300 400 500 Concentration (ng/mL) ×103 ×103 ×103 ×105 ×105 ×106 ×103 ×103 ×103 ×106 ×106 ×107 Acquisition time (min) Acquisition time (min) Acquisition time (min) Acquisition time (min) Acquisition time (min) Acquisition time (min) Counts Counts Counts Counts Counts Counts Responses Responses Responses Responses Responses Responses 90 16 Recovery and precision To validate the complete workflow solution and ensure that enhanced matrix cleanup did not have a negative impact on pesticide recovery, a study aimed at recovery and precision evaluation was performed. Two concentrations were selected for the study (10 and 50 ng/g) in dry black tea, resulting in 2 and 10 ppb in the final extract due to the 5x dilution factor. Figure 13 shows target results at 10 and 50 ng/g in black tea, demonstrating the acceptable recoveries achieved for the majority of pesticides, even for the common problematic pesticides such as planar and labile. Longevity and maximized throughput with confidence The ruggedness of the analysis was demonstrated by analyzing a challenging black tea extract spiked with pesticides at 2 ppb. The area of the analyte response was monitored over 800 consecutive injections. Analyte response, normalized by the internal standards (ISTD), remained consistent over 800 injections that spanned over 400 hours of continuous running with RSDs < 20% for 176 compounds. Figure 14 shows the response for 60 compounds, normalized by the ISTD and by the average response for each analyte. 0% 20% 40% 60% 80% 100% 120% 140% 160% 180% Dichlorvos Dichlorobenzonitrile, 2,6- Biphenyl Mevinphos, EChlormephos Propham Pebulate Etridiazole Nitrapyrin cis-1,2,3,6-Tetrahydrophthalimide Methacrifos Chloroneb Crimidine 2-Phenylphenol Isoprocarb I Pentachlorobenzene Heptenophos DEET Chlorfenprop-methyl Omethoate Thionazin Flonicamid Propachlor Ethoprophos Cycloate Chlorpropham Ethalfluralin DMSA Trifluralin Benfluralin Monocrotophos Dicrotofos Sulfotep Bromoxynil Promecarb Cadusafos Phorate BHC-alpha (benzene hexachloride) Desmedipham Hexachlorobenzene Dichloran Dimethoate Pentachloroanisole Propazine BHC-beta DMST (Tolylfluanid metabolite) Propetamphos Profluralin BHC-gamma (Lindane, gamma HCH) Cyanophos Terbufos Pentachloronitrobenzene Fonofos Diazinon Pyrimethanil Fluchloralin Phosphamidon I Dinitramine Tefluthrin Paraoxon-methyl BHC-delta Isazofos Etrimfos Triallate Chlorothalonil Iprobenfos Formothion Bromocyclen Pentachloroaniline Desmetryn Dichlofenthion Propanil 2,4,4'-Trichlorobiphenyl (BZ #28) Malaoxon Vinclozolin Transfluthrin Parathion-methyl Chlorpyrifos-methyl Cymiazole Tolclofos-methyl Alachlor Fuberidazole Heptachlor Prometryn Paraoxon Ronnel Prosulfocarb Octachlorodipropyl ether Pirimiphos-methyl 2,2',5,5'-Tetrachlorobiphenyl (BZ #52) Fenitrothion Methiocarb Dipropetryn Malathion Ioxynil Dichlofluanid Parathion-d10 Metolachlor Phorate sulfone Aldrin Anthraquinone Chlorpyrifos Parathion Flufenacet Nitrothal-isopropyl DCPA (Dacthal, Chlorthal-dimethyl) Isocarbophos Chlorthion Isobenzan Trichloronat Fenson Bromophos Pirimiphos-ethyl Fosthiazate I Isopropalin Cyprodinil Isofenphos-methyl Isodrin Pendimethalin Terbufos sulfone Chlozolinate Heptachlor exo-epoxide Esbiothrin Bioallethrin Chlordane-oxy Tolylfluanid Isofenphos Mecarbam Chlorfenvinphos Heptachlor endo-epoxide Fipronil Captan Quinalphos Phenthoate Dinobuton Procymidone Folpet Chlorbenside Methidathion Bromophos-ethyl Chlordane-trans DDE-o,p' 2,2',4,5,5'-Pentachlorobiphenyl (BZ #101) Tetrachlorvinphos Chlordane-cis Endosulfan I (alpha isomer) Ditalimfos Picoxystrobin Flutriafol Fenamiphos Nonachlor, transChlorfenson Iodofenphos Prothiofos Isoprothiolane Flubenzimine Profenofos DDE-p,p' Dieldrin Oxyfluorfen Myclobutanil DDD-o,p' Methoprotryne Azaconazole Dibromobenzophenone, 4,4'- Isoxathion Binapacryl Nitrofen Ethylan Chlorfenapyr Endrin Carbophenothion-Methyl Chloropropylate 2,3',4,4',5-Pentachlorobiphenyl (BZ #118) Endosulfan II (beta isomer) Fensulfothion Flamprop-isopropyl DDD-p,p' Aclonifen DDT-o,p' Ethion Chlorthiophos Tetrasul 2,2',4,4',5,5'-Hexachlorobiphenyl (BZ #153) Sulprofos Triazophos Famphur Carbophenothion Methoxychlor olefin Cyanofenphos Edifenphos DDT-p,p' Endosulfan sulfate 2,2',3,4,4',5'-Hexachlorobiphenyl (BZ #138) Diclofop-methyl Diflufenican Propargite Piperonyl butoxide Captafol Nitralin Mefenpyr-diethyl Benzoylprop-ethyl Iprodione Spiromesifen Tetramethrin I Pyridaphenthion Endrin ketone Dimoxystrobin Phosmet Bifenthrin Bromopropylate EPN Picolinafen Dicofol, p, p'- Fenpropathrin 2,2',3,4,4',5,5'-Heptachlorobiphenyl (BZ #180) Phenothrin I Tetradifon Furathiocarb Phosalone Azinphos-methyl Leptophos Cyhalothrin (Lambda) Cyhalofop-butyl Tralkoxydim Mirex Acrinathrin Pyrazophos Azinphos-ethyl Cycloxydim (Focus) Permethrin, (1R)-cisPermethrin, (1R)-transCoumaphos Dioxathion Butafenacil Cyfluthrin I Cypermethrin I Halfenprox Flucythrinate I Ethofenprox Silafluofen Fenvalerate I Fluvalinate-tau I Deltamethrin 10 ppb in tea → 2 ppb in the extract 50 ppb in tea → 10 ppb in the extract Figure 13. Pesticide recoveries in black tea at 10 and 50 ppb shown for all 244 pesticides. Figure 14. Stability of the peak area for pesticides spiked at 2 ppb into black tea extract, normalized by the ISTD and the average response, over 800 consecutive injections with the Agilent 8890 GC and 7010D GC/TQ systems. 100 200 300 400 500 600 700 BHC-alpha (benzene hexachloride) 11% Coumaphos 15% Fenitrothion 11% Chlorpyrifos 11% Ethalfluralin 10% Esbiothrin 9% Propazine 10% Tefluthrin 10% 2,2',4,4',5,5'-Hexachlorobiphenyl (BZ #153) 9% Aldrin 12% Myclobutanil 11% Chlordane-oxy 11% BHC-beta 10% DDD-o,p' 10% Fenpropathrin 11% Parathion 7% DMSA 10% Bioallethrin 9% Diazinon 10% Iprobenfos 9% 2,2',3,4,4',5'-Hexachlorobiphenyl (BZ #138) 9% Dieldrin 10% Nonachlor, trans- 10% Chlorfenvinphos 11% Bromophos 10% DDT-p,p' 9% Fonofos 12% Fenson 9% Trifluralin DCPA (Dacthal, Chlorthal-dimethyl) 11% Pyrimethanil 10% 2,2',5,5'-Tetrachlorobiphenyl (BZ #52) 10% 2,2',3,4,4',5,5'-Heptachlorobiphenyl (BZ #180) 9% Endrin 15% Chlorfenson 10% Heptachlor endo-epoxide 10% Prothiofos 10% DDE-o,p' 8% Piperonyl butoxide 11% EPN 11% Benfluralin 10% Metolachlor 9% Fluchloralin 8% 2,2',4,5,5'-Pentachlorobiphenyl (BZ #101) 9% Bifenthrin 11% Endrin ketone 11% Chlordane-cis 10% Heptachlor exo-epoxide 10% Chlorthiophos 11% DDE-p,p' 9% Sulfotep 12% Pentachloronitrobenzene 11% Profluralin 9% Pirimiphos-ethyl 10% Dinitramine 11% 2,3',4,4',5-Pentachlorobiphenyl (BZ #118) 9% Chloropropylate 8% Nitrofen 10% Chlordane-trans 10% Deltamethrin 25% Column trim Replaced syringe 0% 20% 40% 60% 80% 100% 120% 140% 160% 180% 200% 91 17 The graph shows that analyte responses were stable and within 80 to 120% throughout the entire study that lasted over 17 days of continuous analysis. The RSDs for each of the target responses are shown in the legend in Figure 14, with most of them below 12%. The absolute responses in terms of peak areas also remained consistent throughout the longevity study. For example, the RSDs on the peak areas over 800 injections for early-eluting BHC-beta, mid-range eluting fenson, and late-eluting coumaphos were 9%, 10%, and 16%, respectively. The maintenance performed during the robustness testing involved septum and liner replacement every 100 injections. With the midcolumn backflush configuration and the use of the temperature-programmed MMI inlet, inlet liner and septum replacement could be performed in under four minutes, providing a productivity boost to the workflow. Two inches of the GC column head were trimmed after 500 injections. The use of the backflushing allowed for a substantial increase in the number of injections before column head maintenance was needed when analyzing a complex black tea extract. Similar to GC inlet maintenance, column trimming could be performed efficiently in a short amount of time (5 to 10 minutes) and did not require MS cool-down and venting due to the midcolumn configuration coupled with the temperature-programmed MMI. The autoinjector syringe was replaced after 600 injections in the longevity study, as noted in Figure 14, with a total of 1,000 injections made with the syringe. The decision to replace the syringe was driven by the decreased measurement precision resulting from a higher variability in the target and ISTD responses. The replacement procedure was performed as guided by the user manual for the 7693A ALS.12 The precision of measurements was restored after the syringe replacement. As a result, the graph in Figure 14 shows an increased response variability between 500 and 600 injections. This effect was particularly pronounced for deltamethrin, which can present a challenge for achieving good precision at low concentrations. Washing the syringe needle support foot is an additional autoinjector maintenance procedure to consider when analyzing challenging samples to minimize carryover and ensure precision. It is of note that there was no need to perform GC inlet or MS source cleaning during the entire study, which spanned over 1,000 injections, including the calibration assessment and precision and recovery studies. The exceptional method ruggedness shown in this work was achieved by: – Following the key practices to successful pesticide analysis outlined in this application note and in another application note7 – Performing effective sample preparation and cleanup – Using the state-of-the art GC and MS technology in the 8890 and 7010D GC/TQ systems GC/TQ intelligence and new software functionality The health and status of the GC/TQ system was continuously monitored though the longevity study by using the Early Maintenance Feedback functionality in MassHunter Acquisition 13.0 for GC/MS. Figure 15A shows a screenshot of the MS health status, featuring the electron multiplier (EM) voltage at last tune, filament age, pump maintenance schedule, and time since the source was cleaned. The dashboard allows for daily tracking for essential maintenance procedures and reminds the user to perform maintenance on time. In addition to the dashboard view, the built-in intelligent functionality in the 7010D GC/TQ system allows for plotting tune-related parameters over time to track the EI source health and performance. The plot for the EM voltage is shown in Figure 15B. The routine tune check procedures, which could be built into the sequence table via keywords, are helpful in guiding when the EM gain curve needs to be updated. This procedure allows for adjusting the EM voltage to maintain a stable response while not altering tune parameters and ion ratio, maintaining MS tune and method calibration validity. 92 18 A B Consider maintenance EMV monitored over a five-month duration of experiment preparation and execution Figure 15. The early maintenance dashboard for GC/TQ (A) and the EM voltage plot for the tune history (B) shown in Agilent MassHunter Acquisition 13.0 for GC/MS. 93 19 Conclusion This application note presents a workflow solution for analyzing pesticides in black tea with the new 7010D GC/TQ, allowing for the quantitation of 246 pesticide residues at trace levels with LOQs as low as 0.01 ppb for 34% of the targets, at or below 0.1 ppb for 74% of compounds, and below 2 ppb for 96%. Matrix-matched calibration allowed for excellent accuracy over a wide dynamic range, spanning up to five orders of magnitude over the 0.01 to 1,000 ppb range in a complex black tea extract. Method ruggedness was demonstrated through maintaining measurement accuracy with good precision (RSDs < 20% for 176 compounds) for black tea extract spiked at 2 ppb sequentially analyzed over 800 runs and spanning over 17 days of continuous analysis. The key components for a robust workflow included a combination of efficient sample preparation and cleanup, the Agilent 8890 GC hardware, functionality, and GC supplies, the novel EI source technology with HES 2.0, and lastly, the built-in GC/TQ intelligence and new software functionality. References 1. Mehri, A.; Taleb, R.; Elaridi, J.; Hassan, H. F. Analytical Methods Used to Determine Pesticide Residues in Tea: A Systematic Review. Appl. Food Res. 2022, 2(1), 100131. 2. Analytical Quality Control and Method Validation Procedures for Pesticide Residues Analysis in Food and Feed. SANTE 11312/2021, 2021. 3. Tolerances and Exemptions for Pesticide Chemical Residues in Food. Title 40 U.S. Code of Federal Regulations, US EPA. 4. Lozano, A.; Rajski, L.; Belmonte-Valles N.; Uclés, A.; Uclés, S.; Mezcua, M.; Fernández-Alba, A. Pesticide Analysis in Teas and Chamomile by Liquid Chromatography and Gas Chromatography Tandem Mass Spectrometry Using a Modified QuEChERS Method: Validation and Pilot Survey in Real Samples. J. Chrom. A 2012, 1268, 109–122. 5. The Agilent MassHunter Pesticide and Environmental Pollutants MRM Database (P&EP 4.0). G9250AA. https:// www.agilent.com/en/product/gas-chromatographymass-spectrometry-gc-ms/gc-ms-application-solutions/ pesticides-environmental-pollutants-4-0-mrm-database 6. Maštovská, K.; Lehotay, S. J.; Anastassiades, M. Combination of Analyte Protectants to Overcome Matrix Effects in Routine GC Analysis of Pesticide Residues in Food Matrixes. Anal. Chem. 2005, 77, 8129–8137 7. Five Keys to Unlock Maximum Performance in the Analysis of Over 200 Pesticides in Challenging Food Matrices by GC/MS/MS. Agilent Technologies application note 5994-4965EN, 2022. 8. Using the PSD for Backflushing on the Agilent 8890 GC System. Agilent Technologies application note, publication number 5994 0550EN, 2018. 9. Burrows, R. Parr, R. Evaluating the Goodness of Instrument Calibration for Chromatography Procedures. LCGC Supplements Special Issues 2020 11-01-20, 38(11), 35–38. 10. EURL-SRM – Analytical Observation Report. Quantification of Residues of Folpet and Captan in QuEChERS Extracts Version 3.1 (last update: 06.04.17). 11. Kim, L.; Baek, S.; Son, K.; Kim, E.; Noh, H. H.; Kim, D.; Oh, M.; Moon, B.; Ro, J.-H. Optimization of a Simplified and Effective Analytical Method of Pesticide Residues in Mealworms (Tenebrio molitor Larvae) Combined with GC–MS/MS and LC–MS/MS. Mol. 2020, 25(15), 3518. 12. Agilent 7693A Automatic Liquid Sampler. Installation, Operation and Maintenance. Agilent Technologies, 2023. 94 20 Name RT Transition Calibration Range (ppb) CF CF R2 Relative Standard Error Methamidophos 4.520 141.0 & 64.0 0.1 – 1,000 Linear 0.9994 7.5 Dichlorvos 4.643 184.9 & 93.0 0.05 – 1,000 Linear 0.9988 11.2 Dichlorobenzonitrile, 2,6- 5.210 171.0 & 100.0 0.01 – 1,000 Linear 0.9990 10.4 Biphenyl 5.390 154.1 & 153.1 0.5 – 1,000 Quadratic 0.9991 11.8 Mevinphos, E- 5.578 127.0 & 94.9 0.5 – 1,000 Linear 0.9994 7.9 Acephate 5.679 136.0 & 94.0 5 – 1,000 Quadratic 0.9990 8.1 Chlormephos 5.687 153.9 & 121.1 0.5 – 1,000 Linear 0.9979 5.0 Propham 5.740 178.9 & 137.1 1 – 1,000 Linear 0.9986 9.6 Pebulate 5.774 128.0 & 57.1 0.5 – 1,000 Linear 0.9983 6.6 Etridiazole 5.798 213.1 & 185.0 0.1 – 500 Linear 0.9986 10.7 Nitrapyrin 5.804 194.0 & 158.0 0.05 – 1,000 Quadratic 0.9994 8.7 cis-1,2,3,6-Tetrahydrophthalimide 5.956 151.1 & 80.0 0.1 – 1,000 Linear 0.9994 6.1 Methacrifos 6.027 207.9 & 180.1 0.05 – 1,000 Quadratic 0.9996 11.4 Chloroneb 6.110 191.0 & 113.0 0.01 – 1,000 Linear 0.9995 8.2 Crimidine 6.212 170.9 & 142.1 0.1 – 1,000 Linear 0.9994 10.4 2-Phenylphenol 6.213 169.1 & 115.1 0.5 – 1,000 Linear 0.9995 5.6 Isoprocarb I 6.295 136.0 & 121.1 0.5 – 1,000 Linear 0.9994 9.3 Pentachlorobenzene 6.311 251.9 & 217.0 0.01 – 1,000 Linear 0.9984 11.4 Heptenophos 6.585 124.0 & 89.0 0.01 – 500 Linear 0.9990 12.5 DEET 6.600 191.0 & 190.0 0.5 – 1,000 Linear 0.9981 11.5 Chlorfenprop-methyl 6.696 165.0 & 102.0 0.01 – 1,000 Linear 0.9989 14.0 Omethoate 6.773 110.0 & 47.0 0.1 – 1,000 Quadratic 0.9997 8.8 Thionazin 6.781 143.0 & 79.0 0.1 – 1,000 Quadratic 0.9998 9.5 Flonicamid 6.859 173.9 & 68.9 0.01 – 1,000 Linear 0.9994 11.5 Propachlor 6.865 176.1 & 57.1 0.05 – 1,000 Quadratic 0.9997 8.5 Ethoprophos 6.996 157.9 & 97.0 0.1 – 1,000 Quadratic 0.9997 8.2 Cycloate 7.017 154.1 & 83.1 0.05 – 1,000 Linear 0.9991 16.2 Chlorpropham 7.080 171.0 & 127.1 0.05 – 500 Linear 0.9993 13.7 Ethalfluralin 7.109 275.9 & 202.1 0.05 – 1,000 Quadratic 0.9997 11.4 DMSA 7.169 200.0 & 108.0 2 – 1,000 Quadratic 0.9963 17.5 Trifluralin 7.217 306.1 & 264.0 0.05 – 1,000 Quadratic 0.9997 11.8 Benfluralin 7.251 292.0 & 264.0 0.1 – 1,000 Quadratic 0.9996 12.0 Monocrotophos 7.258 192.0 & 127.0 0.1 – 1,000 Quadratic 0.9998 7.4 Dicrotofos 7.264 193.0 & 127.1 0.5 – 1,000 Quadratic 0.9998 10.1 Sulfotep 7.349 321.8 & 201.9 0.05 – 1,000 Quadratic 0.9997 10.4 Bromoxynil 7.395 276.8 & 88.0 0.05 – 1,000 Quadratic 0.9997 7.1 Promecarb 7.399 135.1 & 115.1 2 – 1,000 Linear 0.9967 16.1 Cadusafos 7.405 158.8 & 97.0 0.01 – 1,000 Quadratic 0.9997 13.1 Phorate 7.475 121.0 & 47.0 0.5 – 1,000 Linear 0.9983 11.1 BHC-alpha (Benzene Hexachloride) 7.609 218.9 & 183.0 0.01 – 1,000 Quadratic 0.9998 9.9 Desmedipham 7.690 181.0 & 122.0 2 – 1,000 Linear 0.9985 11.8 Hexachlorobenzene 7.741 283.8 & 213.9 0.01 – 1,000 Quadratic 0.9996 10.7 Dichloran 7.771 160.1 & 124.1 0.01 – 1,000 Linear 0.9996 6.7 Dimethoate 7.781 87.0 & 46.0 0.01 – 1,000 Linear 0.9997 11.8 Pentachloroanisole 7.797 279.9 & 236.8 0.05 – 1,000 Linear 0.9993 8.0 Appendix Table 1. Calibration performance for 246 pesticides in black tea using the Agilent 7010D GC/TQ equipped with the Agilent High-Efficiency Source (HES) 2.0. 95 21 Name RT Transition Calibration Range (ppb) CF CF R2 Relative Standard Error Propazine 7.933 229.1 & 58.1 0.01 – 1,000 Linear 0.9988 13.5 BHC-beta 8.010 218.9 & 183.1 0.01 – 1,000 Linear 0.9994 12.2 DMST (Tolylfluanid Metabolite) 8.032 214.0 & 106.0 2 – 1,000 Quadratic 0.9954 17.7 Propetamphos 8.079 138.0 & 64.0 0.1 – 1,000 Linear 0.9988 6.9 Profluralin 8.087 318.1 & 199.1 0.05 – 1,000 Quadratic 0.9997 12.7 BHC-gamma (Lindane, Gamma HCH) 8.119 216.9 & 181.0 0.01 – 1,000 Linear 0.9981 15.6 Cyanophos 8.135 242.9 & 109.0 0.05 – 1,000 Linear 0.9993 12.8 Terbufos 8.137 230.9 & 129.0 0.1 – 1,000 Linear 0.9994 12.5 Pentachloronitrobenzene 8.195 141.9 & 106.9 0.01 – 1,000 Quadratic 0.9996 9.4 Fonofos 8.223 246.1 & 109.0 0.01 – 500 Linear 0.9981 11.9 Diazinon 8.264 137.1 & 84.0 0.05 – 1,000 Quadratic 0.9997 17.0 Pyrimethanil 8.269 198.0 & 118.1 0.01 – 500 Linear 0.9990 12.6 Fluchloralin 8.299 325.8 & 62.9 0.05 – 500 Quadratic 0.9995 13.5 Phosphamidon I 8.339 127.0 & 95.0 0.5 – 500 Linear 0.9981 17.4 Dinitramine 8.382 260.7 & 241.0 0.05 – 1,000 Quadratic 0.9997 11.6 Tefluthrin 8.400 177.1 & 127.1 0.01 – 1,000 Linear 0.9986 16.0 Paraoxon-methyl 8.411 229.9 & 106.1 0.05 – 500 Linear 0.9947 18.3 BHC-delta 8.489 219.0 & 183.1 0.5 – 1,000 Quadratic 0.9998 16.5 Isazofos 8.504 256.9 & 162.0 0.01 – 1,000 Quadratic 0.9997 13.8 Etrimfos 8.523 292.0 & 153.1 0.01 – 500 Linear 0.9985 16.7 Triallate 8.540 268.0 & 184.1 0.05 – 1,000 Linear 0.9995 10.8 Chlorothalonil 8.568 265.9 & 168.0 0.1 – 500 Quadratic 0.9971 10.7 Iprobenfos 8.673 203.9 & 91.0 0.01 – 1,000 Linear 0.9986 13.9 Formothion 8.763 124.9 & 47.0 0.1 – 1,000 Quadratic 0.9997 16.0 Bromocyclen 8.764 271.8 & 236.9 0.1 – 1,000 Linear 0.9996 8.6 Pentachloroaniline 8.897 158.0 & 123.0 0.5 – 1,000 Linear 0.9986 14.7 Desmetryn 8.916 213.0 & 58.1 0.05 – 1,000 Linear 0.9972 11.7 Dichlofenthion 8.961 279.0 & 223.0 0.01 – 500 Linear 0.9975 11.4 Propanil 8.980 161.0 & 99.0 0.01 – 1,000 Linear 0.9983 16.8 2,4,4'-Trichlorobiphenyl (BZ #28) 9.030 256.0 & 186.0 0.05 – 1,000 Linear 0.9981 11.7 Malaoxon 9.103 126.9 & 99.0 2 – 1,000 Quadratic 0.9989 15.0 Vinclozolin 9.128 187.0 & 124.0 0.05 – 1,000 Linear 0.9962 15.8 Transfluthrin 9.129 163.1 & 143.1 0.1 – 1,000 Linear 0.9972 14.3 Parathion-methyl 9.151 262.9 & 109.0 0.5 – 500 Quadratic 0.9991 18.2 Chlorpyrifos-methyl 9.151 288.0 & 93.0 0.05 – 1,000 Quadratic 0.9995 12.0 Cymiazole 9.213 218.0 & 144.1 2 – 1,000 Linear 0.9986 12.3 Tolclofos-methyl 9.242 267.0 & 93.0 0.5 – 1,000 Linear 0.9987 12.8 Alachlor 9.280 237.0 & 160.1 1 – 1,000 Linear 0.9969 12.2 Fuberidazole 9.306 184.0 & 156.2 5 – 500 Linear 0.9921 19.5 Heptachlor 9.330 271.7 & 236.9 0.1 – 200 Linear 0.9990 15.4 Prometryn 9.339 241.0 & 58.2 5 – 200 Quadratic 0.9967 19.4 Paraoxon 9.383 148.9 & 119.0 50 – 1,000 Quadratic 0.9982 11.3 Ronnel 9.411 286.9 & 272.0 1 – 1,000 Linear 0.9989 11.7 Prosulfocarb 9.424 251.0 & 128.2 0.1 – 1,000 Quadratic 0.9995 14.9 Octachlorodipropyl Ether 9.431 129.9 & 94.9 0.5 – 1,000 Quadratic 0.9996 10.0 Pirimiphos-methyl 9.610 290.0 & 125.0 0.01 – 1,000 Quadratic 0.9996 9.3 96 22 Name RT Transition Calibration Range (ppb) CF CF R2 Relative Standard Error 2,2',5,5'-Tetrachlorobiphenyl (BZ #52) 9.617 289.9 & 219.9 0.01 – 1,000 Linear 0.9997 11.9 Fenitrothion 9.622 125.1 & 47.0 0.01 – 1,000 Linear 0.9994 11.3 Methiocarb 9.628 168.0 & 109.1 2 – 1,000 Quadratic 0.9998 11.3 Dipropetryn 9.748 255.1 & 222.1 0.01 – 500 Linear 0.9982 13.5 Malathion 9.759 172.9 & 99.0 0.01 – 1,000 Quadratic 0.9997 12.1 Ioxynil 9.780 370.8 & 117.0 0.05 – 1,000 Linear 0.9982 12.2 Dichlofluanid 9.785 167.0 & 97.0 1 – 500 Quadratic 0.9991 13.8 Metolachlor 9.913 238.0 & 162.2 0.01 – 1,000 Linear 0.9993 12.5 Phorate Sulfone 9.914 199.0 & 97.0 0.1 – 500 Linear 0.9975 11.7 Aldrin 9.932 254.9 & 220.0 0.05 – 1,000 Linear 0.9995 8.8 Anthraquinone 9.941 208.0 & 152.2 0.05 – 1,000 Linear 0.9990 19.1 Chlorpyrifos 9.968 313.8 & 257.8 0.05 – 1,000 Linear 0.9994 11.7 Parathion 9.983 291.0 & 109.0 0.01 – 1,000 Linear 0.9985 15.2 Flufenacet 10.004 151.0 & 95.0 0.5 – 1,000 Linear 0.9994 11.9 Nitrothal-isopropyl 10.057 254.0 & 212.0 0.5 – 500 Linear 0.9967 16.3 DCPA (Dacthal, Chlorthal-dimethyl) 10.068 298.9 & 221.0 0.01 – 1,000 Linear 0.9993 14.1 Isocarbophos 10.114 136.0 & 69.0 0.5 – 1,000 Linear 0.9985 13.1 Chlorthion 10.156 125.1 & 47.1 0.05 – 1,000 Quadratic 0.9995 14.7 Isobenzan 10.186 274.7 & 240.0 0.1 – 1,000 Linear 0.9992 11.2 Trichloronat 10.199 296.8 & 268.9 0.01 – 1,000 Linear 0.9987 14.5 Fenson 10.210 141.0 & 77.1 0.01 – 1,000 Linear 0.9996 6.7 Bromophos 10.294 330.9 & 315.9 0.01 – 1,000 Linear 0.9986 15.8 Pirimiphos-ethyl 10.294 318.1 & 166.1 0.05 – 1,000 Linear 0.9977 14.8 Fosthiazate I 10.299 195.0 & 103.0 0.05 – 1,000 Quadratic 0.9997 12.1 Isopropalin 10.350 280.1 & 238.1 0.01 – 500 Linear 0.9975 16.4 Cyprodinil 10.413 225.2 & 224.3 0.05 – 1,000 Linear 0.9990 14.4 Isofenphos-methyl 10.420 199.0 & 121.0 0.01 – 1,000 Quadratic 0.9995 9.9 Isodrin 10.442 193.0 & 123.0 0.05 – 1,000 Linear 0.9985 10.9 Pendimethalin 10.522 251.8 & 162.2 0.05 – 1,000 Quadratic 0.9995 11.5 Terbufos Sulfone 10.573 264.0 & 199.0 0.05 – 1,000 Quadratic 0.9996 8.7 Chlozolinate 10.586 186.0 & 109.0 0.1 – 1,000 Quadratic 0.9995 8.6 Heptachlor Exo-epoxide 10.616 354.8 & 264.9 0.01 – 1,000 Linear 0.9993 15.3 Esbiothrin 10.622 123.0 & 93.0 2 – 1,000 Linear 0.9975 15.2 Bioallethrin 10.629 123.0 & 81.0 2 – 1,000 Linear 0.9924 12.2 Chlordane-oxy 10.629 114.9 & 51.1 0.5 – 1,000 Linear 0.9996 3.8 Tolylfluanid 10.639 238.0 & 137.0 0.05 – 1,000 Quadratic 0.9996 14.6 Isofenphos 10.674 212.9 & 121.1 0.01 – 1,000 Quadratic 0.9994 10.7 Mecarbam 10.675 130.9 & 86.0 0.5 – 1,000 Quadratic 0.9992 13.0 Chlorfenvinphos 10.679 266.9 & 159.0 0.01 – 200 Linear 0.9986 11.0 Heptachlor Endo-epoxide 10.683 135.0 & 99.0 0.5 – 1,000 Linear 0.9993 6.6 Fipronil 10.698 366.8 & 212.8 0.05 – 1,000 Quadratic 0.9997 8.8 Captan 10.738 149.0 & 70.0 1 – 500 Linear 0.9973 18.9 Quinalphos 10.738 298.0 & 156.0 0.01 – 500 Quadratic 0.9992 17.6 Phenthoate 10.741 274.0 & 125.0 0.01 – 1,000 Quadratic 0.9988 8.1 Dinobuton 10.743 211.0 & 163.0 0.5 – 1,000 Quadratic 0.9999 9.0 Procymidone 10.850 282.8 & 96.0 0.01 – 1,000 Linear 0.9996 8.3 97 23 Name RT Transition Calibration Range (ppb) CF CF R2 Relative Standard Error Folpet 10.851 259.8 & 130.1 0.5 – 200 Linear 0.9989 16.4 Chlorbenside 10.904 125.0 & 89.0 0.01 – 1,000 Linear 0.9993 8.2 Methidathion 11.007 125.0 & 47.0 0.5 – 1,000 Linear 0.9993 16.2 Bromophos-ethyl 11.022 358.7 & 302.8 0.01 – 1,000 Linear 0.9995 9.1 Chlordane-trans 11.024 271.7 & 236.9 0.05 – 1,000 Linear 0.9995 8.4 DDE-o,p' 11.073 246.0 & 176.2 0.01 – 1,000 Linear 0.9995 6.4 2,2',4,5,5'-Pentachlorobiphenyl (BZ #101) 11.111 325.9 & 255.9 0.01 – 1,000 Linear 0.9995 7.7 Tetrachlorvinphos 11.166 329.0 & 108.9 0.01 – 1,000 Linear 0.9972 16.9 Chlordane-cis 11.288 372.8 & 265.9 0.01 – 1,000 Linear 0.9994 15.4 Endosulfan I (Alpha Isomer) 11.290 194.9 & 160.0 0.1 – 1,000 Linear 0.9993 14.8 Ditalimfos 11.299 242.9 & 148.1 0.05 – 500 Linear 0.9977 14.8 Picoxystrobin 11.307 145.0 & 102.1 0.05 – 1,000 Linear 0.9994 14.1 Flutriafol 11.335 123.1 & 75.1 0.05 – 1,000 Linear 0.9995 11.9 Fenamiphos 11.360 303.0 & 154.0 0.1 – 500 Linear 0.9959 17.1 Nonachlor, trans- 11.369 406.8 & 299.8 0.05 – 1,000 Linear 0.9993 11.9 Chlorfenson 11.374 175.0 & 111.0 0.01 – 1,000 Linear 0.9994 10.2 Iodofenphos 11.466 376.8 & 361.8 0.05 – 1,000 Quadratic 0.9998 12.4 Prothiofos 11.488 308.9 & 238.9 0.05 – 500 Linear 0.9975 16.8 Isoprothiolane 11.498 162.1 & 85.0 0.01 – 1,000 Linear 0.9986 13.4 Flubenzimine 11.538 186.0 & 69.0 2 – 500 Quadratic 0.9974 10.5 Profenofos 11.544 207.9 & 63.0 0.1 – 200 Linear 0.9984 12.4 DDE-p,p' 11.613 246.1 & 176.2 0.01 – 1,000 Linear 0.9995 6.8 Dieldrin 11.713 262.9 & 193.0 0.5 – 1,000 Linear 0.9990 13.8 Oxyfluorfen 11.721 252.0 & 146.0 0.01 – 1,000 Quadratic 0.9995 14.0 Myclobutanil 11.750 179.0 & 125.1 0.01 – 1,000 Quadratic 0.9995 12.6 DDD-o,p' 11.783 235.0 & 165.1 0.01 – 1,000 Linear 0.9993 12.7 Methoprotryne 11.788 256.0 & 212.1 0.01 – 1,000 Quadratic 0.9996 12.6 Azaconazole 11.865 217.0 & 173.1 0.01 – 1,000 Linear 0.9994 11.4 Dibromobenzophenone, 4,4'- 11.913 340.0 & 183.0 0.05 – 1,000 Quadratic 0.9997 7.6 Isoxathion 11.941 313.0 & 177.0 0.05 – 1,000 Quadratic 0.9993 14.9 Binapacryl 12.003 100.0 & 82.0 5 – 1,000 Quadratic 0.9988 18.9 Nitrofen 12.011 282.9 & 253.0 0.01 – 500 Linear 0.9963 14.5 Ethylan 12.041 223.1 & 193.1 0.05 – 500 Linear 0.9976 14.7 Chlorfenapyr 12.051 328.0 & 247.0 0.5 – 1,000 Linear 0.9983 14.8 Endrin 12.108 262.8 & 193.0 0.5 – 1,000 Linear 0.9996 5.5 Carbophenothion-methyl 12.167 125.0 & 47.0 0.1 – 1,000 Linear 0.9986 19.2 Chloropropylate 12.187 139.0 & 75.0 0.01 – 1,000 Linear 0.9996 7.2 2,3',4,4',5-Pentachlorobiphenyl (BZ #118) 12.222 325.9 & 255.9 0.01 – 1,000 Linear 0.9996 7.6 Endosulfan II (Beta Isomer) 12.274 206.9 & 172.0 0.1 – 1,000 Quadratic 0.9989 13.0 Fensulfothion 12.284 293.0 & 97.0 0.01 – 1,000 Linear 0.9978 14.8 Flamprop-isopropyl 12.305 276.0 & 105.1 0.05 – 1,000 Linear 0.9991 10.4 DDD-p,p' 12.369 237.0 & 165.1 0.01 – 1,000 Linear 0.9993 13.7 Aclonifen 12.397 264.1 & 194.2 0.1 – 1,000 Quadratic 0.9995 10.3 DDT-o,p' 12.430 235.0 & 199.1 0.01 – 1,000 Linear 0.9974 19.4 Ethion 12.431 231.0 & 129.0 0.01 – 1,000 Linear 0.9962 19.9 98 24 Name RT Transition Calibration Range (ppb) CF CF R2 Relative Standard Error Chlorthiophos 12.484 268.9 & 205.1 0.05 – 1,000 Linear 0.9983 13.9 Tetrasul 12.572 321.7 & 252.0 0.01 – 1,000 Linear 0.9987 11.5 2,2',4,4',5,5'-Hexachlorobiphenyl (BZ #153) 12.610 359.9 & 289.9 0.01 – 1,000 Linear 0.9992 10.9 Sulprofos 12.650 322.0 & 156.0 0.01 – 500 Linear 0.9975 12.5 Triazophos 12.662 161.2 & 134.2 1 – 500 Quadratic 0.9995 15.0 Famphur 12.810 218.0 & 109.0 2 – 1,000 Linear 0.9982 14.6 Carbophenothion 12.826 342.0 & 157.0 0.05 – 500 Linear 0.9974 14.5 Methoxychlor Olefin 12.837 308.0 & 238.0 0.01 – 1,000 Linear 0.9984 13.5 Cyanofenphos 12.906 169.0 & 77.1 0.1 – 1,000 Linear 0.9988 10.7 Edifenphos 12.940 309.9 & 172.9 0.5 – 1,000 Quadratic 0.9998 13.3 DDT-p,p' 13.027 235.0 & 165.2 0.01 – 1,000 Linear 0.9976 19.9 Endosulfan Sulfate 13.032 271.9 & 237.0 0.5 – 1,000 Quadratic 0.9997 18.2 2,2',3,4,4',5'-Hexachlorobiphenyl (BZ #138) 13.118 359.9 & 289.9 0.01 – 1,000 Linear 0.9996 6.5 Diclofop-methyl 13.284 339.9 & 252.9 0.01 – 500 Linear 0.9973 13.1 Diflufenican 13.310 266.0 & 246.1 0.01 – 1,000 Linear 0.9971 18.7 Propargite 13.327 231.0 & 135.0 0.5 – 1,000 Linear 0.9993 13.7 Piperonyl Butoxide 13.380 176.1 & 103.1 0.1 – 1,000 Quadratic 0.9995 9.0 Captafol 13.440 310.8 & 78.8 10 – 1,000 Quadratic 0.9987 19.3 Nitralin 13.551 315.9 & 274.0 0.1 – 500 Quadratic 0.9993 13.9 Mefenpyr-diethyl 13.608 253.0 & 189.0 0.01 – 500 Quadratic 0.9989 15.8 Benzoylprop-ethyl 13.699 292.0 & 105.0 0.1 – 1,000 Linear 0.9991 11.8 Iprodione 13.721 313.8 & 55.9 0.1 – 500 Linear 0.9984 12.4 Spiromesifen 13.722 272.0 & 254.2 0.1 – 500 Quadratic 0.9994 15.7 Tetramethrin I 13.814 164.0 & 77.1 5 – 1,000 Quadratic 0.9990 17.1 Pyridaphenthion 13.822 340.0 & 199.0 0.05 – 1,000 Quadratic 0.9996 12.9 Endrin Ketone 13.876 316.9 & 101.0 0.01 – 500 Quadratic 0.9993 13.6 Dimoxystrobin 13.880 205.0 & 58.0 0.1 – 1,000 Quadratic 0.9996 8.4 Phosmet 13.917 160.0 & 77.1 2 – 1,000 Quadratic 0.9993 10.6 Bifenthrin 13.922 181.2 & 165.2 0.1 – 500 Quadratic 0.9990 13.8 Bromopropylate 13.928 338.8 & 182.9 0.05 – 1,000 Quadratic 0.9993 10.7 EPN 13.935 169.0 & 77.1 0.05 – 1,000 Quadratic 0.9991 13.3 Picolinafen 13.958 376.0 & 238.1 0.01 – 200 Linear 0.9978 17.1 Bifenazate 13.975 168.1 & 61.9 10 – 1,000 Quadratic 0.9990 6.6 Dicofol, p, p'- 13.976 183.9 & 141.2 1 – 1,000 Linear 0.9972 18.8 Fenpropathrin 14.056 265.0 & 89.0 0.01 – 1,000 Quadratic 0.9996 13.5 2,2',3,4,4',5,5'-Heptachlorobiphenyl (BZ #180) 14.299 393.8 & 323.8 0.01 – 1,000 Linear 0.9992 12.2 Phenothrin I 14.399 122.9 & 81.1 0.1 – 1,000 Linear 0.9987 11.5 Tetradifon 14.424 158.9 & 111.0 0.01 – 1,000 Linear 0.9992 7.4 Furathiocarb 14.437 163.1 & 135.1 2 – 1,000 Linear 0.9992 6.1 Phosalone 14.590 182.0 & 75.0 0.05 – 500 Linear 0.9958 19.2 Azinphos-methyl 14.626 160.0 & 77.0 2 – 1,000 Quadratic 0.9993 11.4 Leptophos 14.638 171.0 & 51.0 0.05 – 1,000 Quadratic 0.9995 12.4 Cyhalothrin (Lambda) 14.698 181.1 & 152.1 5 – 1,000 Quadratic 0.9979 12.9 Cyhalofop-butyl 14.703 357.1 & 229.1 0.01 – 500 Linear 0.9958 16.1 99 www.agilent.com DE28615044 This information is subject to change without notice. © Agilent Technologies, Inc. 2024 Printed in the USA, May 16, 2024 5994-7436EN Name RT Transition Calibration Range (ppb) CF CF R2 Relative Standard Error Tralkoxydim 14.830 137.0 & 57.0 0.05 – 1,000 Linear 0.9990 7.9 Mirex 14.865 271.8 & 236.8 0.01 – 1,000 Linear 0.9994 13.5 Acrinathrin 15.045 247.0 & 68.0 1 – 1,000 Quadratic 0.9996 12.8 Pyrazophos 15.144 221.0 & 149.0 0.01 – 1,000 Quadratic 0.9994 14.1 Azinphos-ethyl 15.228 160.0 & 77.1 0.5 – 1,000 Quadratic 0.9997 12.4 Cycloxydim (Focus) 15.500 178.0 & 80.9 0.1 – 1,000 Quadratic 0.9997 8.0 Permethrin, (1R)-cis- 15.622 163.0 & 91.0 2 – 1,000 Quadratic 0.9978 18.4 Permethrin, (1R)-trans- 15.744 163.0 & 127.0 0.01 – 1,000 Linear 0.9990 12.0 Coumaphos 15.880 361.9 & 109.0 0.05 – 500 Linear 0.9972 16.0 Dioxathion 15.963 271.0 & 96.9 0.1 – 500 Linear 0.9969 19.8 Butafenacil 15.988 331.0 & 180.0 0.01 – 500 Linear 0.9973 14.4 Cyfluthrin I 16.202 163.0 & 127.0 0.5 – 1,000 Linear 0.9980 15.8 Cypermethrin I 16.510 163.0 & 127.0 0.1 – 1,000 Linear 0.9985 16.6 Halfenprox 16.565 262.9 & 169.0 0.05 – 1,000 Quadratic 0.9994 11.9 Flucythrinate I 16.725 156.9 & 107.1 0.01 – 1,000 Linear 0.9992 13.7 Ethofenprox 16.798 163.0 & 107.1 0.1 – 1,000 Linear 0.9989 14.3 Silafluofen 16.944 286.0 & 207.0 0.1 – 1,000 Quadratic 0.9995 9.7 Fenvalerate I 17.428 167.0 & 125.1 0.05 – 1,000 Linear 0.9988 12.6 Fluvalinate-tau I 17.601 250.0 & 200.0 0.1 – 1,000 Quadratic 0.9996 15.2 Deltamethrin 18.152 252.9 & 174.0 0.1 – 500 Linear 0.9963 17.2 Application Note Food & Beverage Testing Author Bruce D. Quimby Agilent Technologies, Inc. Abstract Due to ongoing concerns with the price and availability of helium (He), many laboratories are looking for alternative carrier gases for their gas chromatography/mass spectrometry (GC/MS) methods. This application note describes the conversion of a typical GC/MS method for the qualitative analysis of flavor and fragrance compounds in essential oils from helium to hydrogen (H2  ). The Agilent 8890 GC coupled with the Agilent 5977C GC/MSD system were used with hydrogen carrier gas and a new source that has been optimized for hydrogen operation—the Agilent HydroInert source. Unlike most conventional electron ionization (EI) sources, the HydroInert source provides excellent mass spectral fidelity for flavor compounds when using hydrogen. To further increase confidence in compound identification, deconvoluted mass spectra and linear retention indexes (RI) from Agilent MassHunter Unknowns Analysis software were searched against the NIST23 mass spectral library. Using the Agilent Method Translator tool, a column and chromatographic conditions for hydrogen were chosen that allowed reduction of the analysis time by a factor of 2.5 compared to the typical helium method. By proper selection of instrument configuration and operating conditions, the system with hydrogen carrier gas can generate results comparable to those with helium, but with significantly reduced run time. Qualitative Analysis of Essential Oils Using GC/MS with Hydrogen Carrier Gas and the Agilent HydroInert Source Return to Table of Contents 101 2 Introduction Essential oils are widely used as a source of flavors and fragrances in both food and nonfood consumer products. Quality control analysis of essential oils has long been challenging due to the hundreds of terpenes and terpenoid compounds that can be present in the oils. To address this, methods using high-resolution capillary GC combined with MS have typically been employed. Searching the acquired mass spectra against libraries of flavor and fragrance compounds can be performed for identification, but is usually insufficient because many compounds, especially isomers, give similar spectra. For this reason, RIs are often used as a complement to spectral matching for more dependable identifications. The measured RI of an unknown is used in conjunction with the results of the spectral library search to determine the best candidate for identification. This application note describes the conversion of a typical GC/MS method for the qualitative analysis of flavor and fragrance compounds in essential oils from helium to hydrogen. The two methods are then applied to the analysis of two common essential oils—orange oil, of Brazilian origin, and neroli oil—for comparison. The hydrogen method was further evaluated using different EI source components to determine the optimal source for maintaining spectral fidelity. The conversion of a method from helium to hydrogen carrier gas requires consideration of the chromatographic parameters such as column choice, column flow, amount injected, and temperature program rates.1 In addition, it is important to consider the MS parameters of column flow rate and MS source configuration. All these considerations and parameters are included in the Agilent EI GC/MS Instrument Helium to Hydrogen Carrier Gas Conversion User Guide. 1 When determining the chromatographic parameters for the hydrogen method, it is desirable to have a method that has: – Column dimensions that result in a high enough inlet pressure for accurate flow control – Similar chromatographic resolution to the original helium method – Maintenance of the same analyte elution order as the original helium method – Use of a column flow near the optimum for the column diameter used – Use of a column flow near the optimum for MS source sensitivity The Agilent Method Translator tool2-4 is a calculator designed to greatly simplify this process. It was used in this application note. For the MS parameters, the column flow rate should be kept within or near the optimum range of approximately 0.8 to 1.4 mL/min to maximize the MS response. Another important consideration is the choice of MS EI source hardware. The concern is that some analytes undergo reactions with hydrogen in the source, changing their ion ratios and spectra, thus reducing their library match scores (LMS), and possibly resulting in misidentification. With the Agilent inert extractor source used in the Agilent 5977 series GC/MSD systems, this effect has been reduced in the past using an extractor lens with a larger diameter, such as 9 mm. However, this only partially addressed the problem, as many compounds such as nitro compounds and some terpenes and terpenoids still exhibited reactions. For this reason, Agilent developed the HydroInert source5,6, which greatly reduces or eliminates these reactions. To evaluate the effectiveness of the HydroInert source, the qualitative analysis of the two oils was carried out with the helium method using the standard 3 mm inert extractor source and the hydrogen method using the standard HydroInert source, which is equipped by default with the 9 mm lens. In addition, the oils were analyzed with the hydrogen method and a conventional inert extractor source using both the 3 and 9 mm extractor lenses for comparison. The next consideration is how to process the data files to obtain the mass spectrum of each oil component and search it against a library. In the past, this was done largely by obtaining the apex or average spectrum over the peak then subtracting a baseline spectrum taken next to the peak. The resulting spectrum was then searched against the spectral library. While this process is effective for handling a few peaks that were relatively well resolved, it becomes overwhelming with large numbers of peaks and/or overlapping peaks. Fortunately, there is now a powerful solution for mass spectral identification called Agilent MassHunter Unknowns Analysis (MHUA), which is part of the Agilent MassHunter Quantitative Analysis software suite. MHUA uses spectral deconvolution to extract clean analyte spectra from the complex overlapping peaks. The deconvolution and library search processes are automated and take approximately 1 to 8 minutes per data file depending on the file size, library size, computer hardware, and so on. The result is cleaner spectra than with the previous approach, which therefore results in higher LMS and greater confidence in peak identifications.7 102 3 A second major feature of MHUA is the ability to calculate the RI for each peak. If the searched library has the appropriate reference RI values associated with spectral entries, the measured RI value for an unknown spectrum can be used to filter the spectral search results. This is especially important in the identification of essential oil components because of spectral similarities. RI values were used in this application note for this reason. There are multiple mass spectral libraries available with RIs for flavor and fragrance compounds. For example, the Adams library8 has been used widely for many years for this analysis. Recently, NIST released a newer version of their mass spectral library (NIST23), which has numerous enhancements. Among them are the incorporation of the entire Adams library and the expansion of semistandard, nonpolar RI entries to cover all EI spectra. "Semistandard, nonpolar" refers to phases such as HP-5, DB-5, HP-5ms, and other 5% phenylmethyl silicone phases.9 The new RI values are either experimental values, if available, or artificial intelligence (AI)-generated values. Note that the AI-generated values have better accuracy than the previous "estimated" values. The new semistandard, nonpolar values are of specific interest here because this type of stationary phase is commonly used in essential oil analysis with GC/MS. Also, these are the NIST23 values currently usable with the RI function in MHUA. When the RI function of MHUA is used, the experimental semistandard, nonpolar RI values are used if available; if not, the AI-generated values are used. Therefore, NIST23 is the library used here. Experimental Column selection For the reference helium method, the column and conditions chosen are like those frequently used in the past.8,10 A 30 m × 0.25 mm id, 0.25 μm Agilent J&W HP-5ms Ultra Inert (UI) column (part number 19091S-433UI) was used with an oven temperature program from 60 to 240 °C at 3 °C/min. Although some of the older methods10 used constant pressure control mode for column flow, constant flow mode is far better in terms of MS performance. A constant column flow rate of 1.0 mL/min of helium was used. For the hydrogen method, a 20 m × 0.18 mm id, 0.18 μm Agilent J&W HP-5ms UI column (part number 19091S-577UI) was chosen. This column makes an excellent choice for several reasons: – The column dimensions provide an inlet pressure with hydrogen that is high enough for accurate flow control. – The column provides similar or better chromatographic resolution compared to the helium method. – The phase ratio is the same, helping maintain the same analyte elution order as the original method. – A column flow of hydrogen near the optimum for both chromatographic separation and MS source sensitivity can be used. Method translation The Method Translator tool is included as part of the Agilent MassHunter acquisition software or can be downloaded for standalone use from Agilent.com: https://www.agilent. com/en/support/gas-chromatography/gccalculators. The free download includes the Method Translator, Vapor Volume Calculator, Pressure Flow Calculator, and Solvent Vent Calculator tools, which are all useful when developing GC methods. After installation, the Method Translator is opened from an icon on the desktop. The opened Method Translator is shown in Figure 1. First, the chromatographic parameters of the original helium method are entered in the left column labeled Original Method Parameters. The carrier gas type (He), column dimensions, column outlet pressure, and oven temperature program ramp should be entered first. The column outlet flow is then entered. As shown in Figure 1, the other values such as phase ratio, inlet pressure, and so on, are calculated automatically. Next, the carrier gas type (H2  ), column dimensions, and column outlet pressure for the hydrogen method are entered in the right side of the calculator under the Calculated Method Parameters column. After entry, the calculated hydrogen parameters are displayed. In the upper-left corner, the calculated speed gain is shown as 2.5877, meaning that the predicted retention times (RTs) with the hydrogen method would be a factor of approximately 2.6 smaller than the helium method. The calculated oven 103 4 ramp rate for hydrogen would be 7.763 °C/min. Note that it would be easier to have the oven ramp rate and the speed gain closer to 7.5 and 2.5, respectively. This can be done by selecting Speed gain and entering 2.5 into the field. The parameters are recalculated, resulting in the desired parameters. Figure 1 shows the results. Note that the calculated flow for the hydrogen method shown in Figure 1 is 0.84004 mL/min. Before using the method, retention time locking (RTL) was used to make the RT of n-pentadecane in the hydrogen method precisely 2.5 times faster than that with the helium method. This made comparison of RTs easier. The resulting flow for the hydrogen method after RTL was 0.958 mL/min. For library searching, this step is not necessary as the RI calibration accounts for differences in flow. Figure 1. Agilent Method Translator tool, used to determine method parameters for conversion of the helium method to hydrogen. 104 5 MS source hardware The standard inert extractor source with the 3 mm extractor lens is an excellent choice for the helium method when analyzing flavor and fragrance compounds, and was used in this application note. For the hydrogen method, the HydroInert source with the 9 mm extractor lens was used, as it reduces in-source reactions with hydrogen and provides improved peak shape. The hydrogen method was also run with a conventional inert extractor source using both the 3 and 9 mm extractor lenses. These data were compared to that from the HydroInert and helium results to identify components of the oils that were most susceptible to in-source reactions by comparing their spectra and LMS values. Figure 2 shows the system configurations for the helium and hydrogen methods. Chemicals and standards In-house hydrogen with 99.9999% purity specification and low individual specifications on water and oxygen was used as the carrier gas for the hydrogen method. In-house helium with similar specifications was used as the carrier gas for the helium method. Cold-pressed orange oil (Brazil origin) and neroli oil (Morocco origin) were purchased from Sigma-Aldrich (Milwaukee, WI, USA). The oils were diluted to 20:1 (v:v) in ethanol. A custom RI calibration standard consisting of all the n-alkanes from n-C5 to n-C40 plus n-C44 was purchased from Ultra Scientific (now Agilent). All alkanes were at a concentration of 500 ng/µL in n-hexane except n-C13, n-C18, n-C22, n-C28, n-C31, and n-C39, which were at 1,000 ng/µL. The standard was then diluted to 50:1 (v:v) in isooctane. Figure 2. System configurations for the helium and hydrogen methods. Liquid injector S/SL Inlet (hydrogen) HydroInert source with 9 mm extractor lens EI source 20 m × 180 µm, 0.18 µm df Agilent 8890 GC Agilent 8890 GC Liquid injector S/SL inlet (helium) Inert extractor with 3 mm extractor lens Agilent 5977C GC/MSD Agilent 5977C GC/MSD EI source 30 m × 250 µm, 0.25 µm df Agilent HP-5ms UI Agilent HP-5ms UI Helium configuration Hydrogen configuration Table 1. GC and MS conditions for helium and hydrogen methods. Method Parameters Helium Method Hydrogen Method Inlet EPC split/splitless Mode Split 25:1 Column Flow 1.0 mL/min helium 0.958 mL/min hydrogen Injection Volume 1.0 µL Inlet Temperature 250 °C Inlet Liner Agilent universal low pressure drop UI liner with wool (p/n 5190-2295) Column Agilent J&W HP-5ms UI, 30 m × 0.25 mm, 0.25 µm (p/n 19091S-433UI) Agilent J&W HP-5ms UI, 20 m × 0.18 mm, 0.18 µm (p/n 19091S-577UI) Column Temperature Program 60 °C (no hold) 3 °C/min to 240 °C (no hold) 60 °C (no hold) 7.5 °C/min to 240 °C (no hold) Run Time 60 min 24 min MSD Source Agilent inert extractor (3 mm lens) Agilent HydroInert source (9 mm lens) Transfer Line Temperature 300 °C Ion Source Temperature 300 °C Quadrupole Temperature 150 °C EM, Gain Mode 0.1 Mode Scan 40 to 400 m/z TID, A/D Samples TID ON, 8 TID ON, 4 Solvent Delay 2.2 min 0.88 min Tune etune.u 105 6 Results and discussion Retention index calibration The diluted RI calibration standard was run with both methods. The chromatograms are shown in Figure 3. Although the oil sample analysis is finished before n-C26, the standard contains n-alkanes up to n-C40. Therefore, the temperature ramps for both RI calibration methods were extended to 300 °C and held until n-C40 eluted to prevent carryover peaks in subsequent chromatograms. The red arrow in each chromatogram indicates the normal end of run for the oil methods. For determining the RT of the RI calibration compounds, integrating the EIC for m/z 57 is preferred over the total ion chromatogram (TIC), as it has a better signal-to-noise ratio. To use RI values in MHUA, a calibration file needs to be created for each method. The file can be created as a .csv file in Microsoft Excel, or as a text file in Microsoft Windows Notepad. Figure 4 shows the calibration .rtc files created in Notepad. The blue headers are not included in the files; they are included here to indicate the entry format. Each entry consists of the format name, CAS number, RI, and RT, and the associated text file is then saved with a .rtc extension in the filename. The text files are usually saved in either the library directory or the directory containing the data files. Figure 3. EICs at m/z 57 for the RI calibration standard. (A) Helium method; (B) hydrogen method. 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 7.757 15.485 20.532 26.737 1.985 3.036 4.446 6.068 9.432 11.049 12.596 14.075 16.828 18.116 29.390 19.348 21.666 22.754 23.806 24.817 25.792 27.648 28.532 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 19.377 4.804 38.694 7.485 11.051 15.137 23.577 51.260 27.622 31.489 35.177 66.686 73.266 42.049 45.257 48.320 54.087 56.792 59.400 61.918 64.339 68.953 71.138 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70 72 74 13 28 22 18 30 9 14 10 11 12 20 19 17 16 15 29 23 21 25 26 27 24 31 Hydrogen Agilent HP-5ms UI, 20 m × 0.18 mm id, 0.18 µm 240 °C (end of run for samples) 240 °C (end of run for samples) Helium Agilent HP-5ms UI, 30 m × 0.25 mm id, 0.25 µm ×105 A B Acquisition time (min) Acquisition time (min) Counts ×105 Counts 106 7 Brazilian orange oil Figure 5 compares the TICs obtained with the helium and hydrogen methods for the Brazilian orange oil sample. Figure 5A shows the complete time range of elution, and 5B is an expanded view to better compare the peak shapes and chromatographic resolution. As can be seen, by using the Method Translator technique, the relative elution order of peaks is maintained, as is the resolution. Many of the larger peaks in the hydrogen chromatogram exhibited some fronting. This was predicted by the Method Translator tool, which indicated that the hydrogen setup has 36% of the original column capacity of the helium method. However, the chromatographic resolution is still about the same as the helium method. Figure 4. RI calibration text files (.rtc) used in Agilent MassHunter Unknowns Analysis. Helium method Name, CAS, RI, RT Hydrogen method Name, CAS, RI, RT 107 8 Figure 5. (A) Brazilian orange oil with helium and hydrogen methods showing full range of compound elution. (B) Expanded view of earlier elution time region. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Hydrogen Agilent HP-5ms UI, 20 m × 0.18 mm id, 0.18 µm Helium Agilent HP-5ms UI, 30 m × 0.25 mm id, 0.25 µm Brazilian orange oil 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 1 2 3 4 5 6 7 8 9 10 11 12 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Brazilian orange oil, helium method Brazilian orange oil, hydrogen method ×106 ×106 ×106 ×106 A B Acquisition time (min) Counts Counts Counts Counts 108 9 Neroli oil Figure 6 shows the TICs for the neroli oil run with the helium and hydrogen methods. As with the orange oil, the relative elution order of peaks is maintained, as is the resolution. The reduced column capacity of the hydrogen method is again evident in the increased fronting of the large peak at 4.6 minutes in the hydrogen chromatogram. Peak identification with MassHunter Unknowns Analysis The parameters used with MHUA are listed in Table 2. If the library has appropriate RI values for the spectrum entries, RIs can be used as a filter for library hits. The program uses RI if an RI calibration filename is entered in the RT calibration file box. A more detailed description for setting up and running an analysis is shown in the Appendix. Also, an excellent source of information about MHUA is available in a video on the Agilent YouTube channel.7 With the settings listed in Table 2 and a data file analyzed, the program will deconvolute the entire scan file and determine where each detectable peak (component) is. It will then take the deconvoluted (cleaned) spectrum of each component and search it against the library (NIST23). The library entry that best matches the spectrum of the component is checked to see if it exceeds the minimum match factor parameter of 70. If it does, it is next checked to see if the measured RI of the component falls within ± the penalty free range, in this case ± 10 seconds. With the RT mismatch penalty set to Additive and the maximum RT penalty set to 100, if the difference between the measured RT and the library RI (converted to RT) is greater than 10 seconds, the entry is completely discarded. If the difference is less than 10 seconds and the LMS is greater than 70 and higher than the other possible hits, the hit is included in the results table. Figure 6. Neroli oil analyzed with (A) helium and (B) hydrogen methods. 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 Helium Agilent HP-5ms UI, 30 m × 0.25 mm id, 0.25 µm 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Hydrogen Agilent HP-5ms UI, 20 m × 0.18 mm id, 0.18 µm Neroli oil ×106 A Acquisition time (min) Acquisition time (min) Counts ×106 B Counts 109 10 Table 2. Agilent MassHunter Unknowns Analysis method parameters. Parameter Setting RT Window Size Factor 10, 25, 50, 100, 200, 400, 600, 800 Library NIST23.L RT Penalty function Trapezoidal RT Range 10 s Penalty Free Range 10 s RT Mismatch Penalty Additive Maximum RT Penalty 100 Minimum Match Factor 70 Once the entire data file has been processed, the results can be reviewed. Figure 7 shows the results for the Brazilian orange oil with the helium method. Left-clicking on the compound name for one of the listed results displays the deconvoluted component spectrum head to tail with the library spectrum. Figure 7. Agilent MassHunter Unknowns Analysis results for Brazilian orange oil run with the helium method. Identification results TIC chromatogram Component spectrum NIST 23 library spectrum The selected component peak at 16.989 minutes is highlighted in red in the TIC chromatogram, which can be zoomed in for closer inspection. The five most abundant ions are extracted and overlaid in the box to the left of the spectrum display. This is to allow inspection of the peak shapes and apex retention times. If the apex RT or shape of one of the extracted ions is substantially different from the others, this suggests that there might be an interference, which should be considered when interpreting the identification. In practice, reviewing the results consists of going down the list of hits and looking at the Match Factor (Library Match Score), Delta RI, and Base Peak Area. Using the peak at 16.989 as an example, the spectrum has a high-quality match for D-carvone of 98.2 listed, and the head-tail component and library spectra visually match well. The overlay of EICs of the principal ions all have the same shape and apex RT. The delta RI value of 2, which is the difference between the measured RI for the peak and that from the library, is small at 2. Finally, the base peak area and observed peak size in the TIC chromatogram indicate that the response is large enough to produce good-quality spectra. From these observations, the identification of D-carvone is confirmed with high confidence. 110 11 In contrast, the peak at 17.6172, identified as p-mentha-1(7),8(10)-dien-9-ol, has a low LMS of 74.6, a larger delta RI of 4, one of the EICs has a noticeably different apex RT, and the base peak area is small—approximately 190 times smaller than the D-carvone peak. This would be a low-confidence identification. If the reviewer decides it should not be reported, the hit can be removed from the results by right-clicking the name and selecting Delete Components/Hits. If an identification is questioned based on other information, the reviewer can right-click the name in the results table and select Show Alternate Hits. This will display a list of the other spectra in the library that also met the LMS and RI criteria, but with an LMS less than the listed Best Hit. This is useful as sometimes the LMS of the lesser hits is only a fraction of a point smaller. If desired, the reviewer can select one of the alternate hits and set it as the identification. This review process is used to evaluate all the hits. Once completed, the reviewed analysis can be saved and a report can be generated if desired. Evaluating in-source reactions with hydrogen While the inert extractor source with the 3 mm lens is standard for use with helium, it is not the source of choice for use with hydrogen carrier. The metal surfaces inside the source tend to catalyze reactions between hydrogen and some analyte molecules in the source, resulting in peak tailing and spectral changes for some compounds. In the past, substituting the 9 mm extractor lens for the 3 mm lens was used as it reduced the tailing and spectral changes to some degree, but did not eliminate them. For this reason, the HydroInert source was developed. In this section, the spectra of carvone oxide (CAS number 33204-74-9) obtained with hydrogen using the HydroInert source, the inert extractor source with the 3 and 9 mm lenses, and with helium are compared to illustrate the effects of source reactivity. Several other examples are presented in the Appendix. Figure 8 shows the chromatograms and spectra of the carvone oxide peak with the helium and hydrogen methods under the optimized conditions. The library reference spectrum from NIST23 is shown upside down for comparison. With both methods, the deconvoluted spectra have high LMS values of > 95, demonstrating the excellent spectral fidelity provided by the HydroInert source with hydrogen carrier gas. Figure 8. (A) Chromatogram and spectrum of the carvone oxide peak with the helium method. (B) Chromatogram and spectrum of the carvone oxide peak with the hydrogen method and Agilent HydroInert source. 3 mm Inert extractor, He Component RT: 18.4298 40 50 60 70 80 90 100 110 120 130 140 150 160 170 Counts Counts -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 43.0 43.1 123.1 85.1 67.1 95.1 67.0 85.0 109.1 123.0 95.0 55.1 55.0 109.0 137.1 148.0 137.0 148.0 166.0 165.0 3 mm Inert extractor, He LMS = 95.5 9 mm HydroInert H2 LMS = 95.8 Carvone oxide Carvone oxide 16.5 17.0 17.5 18.0 18.5 19.0 19.5 20.0 20.5 21.0 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 16.2812 16.4233 16.9890 16.6057 16.7963 17.4855 17.6172 17.8714 18.1399 18.2331 18.4298 18.7572 18.9074 19.2663 19.5336 20.2660 21.2728 21.0191 9 mm HydroInert, H2 6.6 6.8 7.0 7.2 7.4 7.6 7.8 8.0 8.2 8.4 Counts 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 6.6160 6.8418 8.4586 7.2896 7.5229 7.7307 7.5880 7.3999 7.3203 6.5497 7.0261 8.5487 7.8424 7.1778 8.1341 6.6932 6.7527 Component RT: 7.3999 40 50 60 70 80 90 100 110 120 130 140 150 160 170 Counts -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.2 0 0.4 0.6 0.8 1.0 43.0 43.0 67.1 84.9 123.1 67.0 85.0 95.1 95.0 123.0 55.1 109.1 55.0 109.0 137.1 137.0 147.9 148.0 165.9 166.0 ×105 ×102 ×105 ×102 A B Acquisition time (min) Acquisition time (min) Mass-to-charge (m/z) Mass-to-charge (m/z) 111 12 For comparison, Figure 9 shows the spectra of the carvone oxide peak with hydrogen carrier using the HydroInert source and the inert extractor source with both the 9 and 3 mm extractor lenses. With the 9 mm inert extractor lens shown in the spectrum in Figure 9B, the LMS value is still a respectable 91.2. However, there is clear evidence of some spectral changes. Most notably, the ions 82 and 108 have increased in abundance relative to the rest of the ions in the spectrum. While the degree of spectral fidelity is still useful, it demonstrates that in-source reactions, albeit limited, are occurring. In contrast, the spectrum with the 3 mm inert extractor lens and hydrogen carrier is significantly changed. The spectrum is changed to the extent that the LMS for matching carvone oxide is below the 70 cutoff and thus is not listed, even in the alternate hits list. The search identified the peak as (2,6,6-trimethylbicyclo[3.1.1]heptan-3-yl)methanamine (CAS 61299-72-7), also known as pinane-3-(methylamine). Note that with an LMS value of 84.8 and a close RI match with a delta RI of only –1, this identification looks plausible but is incorrect. Component RT: 7.4187 43.0 82.0 67.0 43.0 123.0 95.0 67.0 85.0 108.0 55.0 95.0 123.0 55.0 109.0 137.0 137.0 151.0 166.0 166.0 9 mm Inert extractor, H2 LMS = 91.2 Component RT: 7.4171 40 50 60 70 80 90 100 110 120 130 140 150 160 170 82.0 95.0 82.0 67.0 108.0 93.0 67.0 43.0 41.0 123.0 107.0 54.0 55.0 135.0 135.0 150.0 121.0 152.0 3 mm Inert extractor, H2 LMS = 84.8 (but wrong compound) Component RT: 7.3999 40 50 60 70 80 90 100 110 120 130 140 150 160 170 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1.0 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1.0 43.0 43.0 67.1 84.9 123.1 67.0 85.0 95.1 95.0 123.0 55.1 109.1 55.0 109.0 137.1 137.0 147.9 148.0 165.9 166.0 9 mm HydroInert, H2 LMS = 95.8 ×102 ×102 A B CCounts Counts -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1.0 ×102 Counts Mass-to-charge (m/z) 40 50 60 70 80 90 100 110 120 130 140 150 160 170 Mass-to-charge (m/z) Mass-to-charge (m/z) Figure 9. (A) Spectrum of the carvone oxide with hydrogen carrier and the Agilent HydroInert source. The reference spectrum is carvone oxide from NIST23. (B) Spectrum with an Agilent inert extractor source and a 9 mm lens. The reference spectrum is carvone oxide from NIST23. (C) Spectrum with an Agilent inert extractor source and a 3 mm lens. The reference spectrum is (2,6,6-trimethylbicyclo[3.1.1]heptan-3-yl)methanamine from NIST23. 112 13 To further investigate the nature of the in-source reaction with the 3 mm inert extractor source, the data file was reanalyzed with MHUA using the same parameters, except not using the RI match criteria. This would list the best hits solely on LMS. If the carvone oxide is reacting with hydrogen in the source to produce a reaction product, the spectral search may reveal what it is. Figure 10 shows the spectrum obtained with (A) the 3 mm lens at the carvone oxide RT compared with (B) that of the best match and (C) the carvone oxide library spectrum. Note that the spectrum obtained with the 3 mm inert extractor source looks very much like a combination of that of carvone oxide and 3-hydroxy-2-methyl-5-(prop-1-en-2-yl) cyclohexanone. Examining the structures in Figure 10, it appears that the epoxide structure of carvone oxide reacts with hydrogen to form the OH group. This example clearly illustrates the perils of using a GC/MS source that allows reactions between hydrogen and analytes, and why the HydroInert source is the best choice when using hydrogen carrier gas. Several other examples are shown in the Appendix. Figure 10. (A) Spectrum obtained with the 3 mm lens at the carvone oxide RT. (B) NIST23 library reference spectrum for the best match when searched without RI filtering, 3-hydroxy-2-methyl-5-(prop-1-en-2-yl)cyclohexanone. (C) NIST23 library reference spectrum for carvone oxide. 30 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 43.0 67.0 85.0 123.0 95.0 55.0 81.0 109.0 71.0 105.0 137.0 148.0 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 43.0 108.0 82.0 67.0 54.0 79.0 95.0 71.0 58.0 122.0 135.0 150.0 NIST23 spectrum of best library match when searched without RI NIST23 spectrum of carvone oxide A B C Counts Counts Counts ×103 ×105 ×103 HO CH3 CH2 H3 C O O CH3 CH2 H3 C O Carvone oxide 3-Hydroxy-2-methyl-5-(prop-1-en-2-yl)cyclohexanone 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 82.0 108.0 67.0 93.0 43.0 123.0 54.0 135.0 150.0 166.0 192.0 222.0 Spectrum at carvone oxide RT 113 14 Analysis results for Brazilian orange and neroli oils Table 3 presents the results of the analysis of the Brazilian orange oil with both the helium method using the 3 mm inert extractor source and the hydrogen method using the 9 mm HydroInert source. Table 4 presents the results for neroli oil with the same methods. The results were reviewed to address questions relevant to converting the method helium to hydrogen carrier gas: – RIs compared to NIST23: The RIs measured with both methods closely matched those in the NIST23 library for most compounds. However, it should be recognized that the RI recognition window used in MassHunter Unknowns Analysis limits the maximum delta RI when listing hits. The second consideration is that if the NIST23 RI value is an AI-predicted value instead of a true experimental value, the errors, and thus the delta RIs, can be larger. – Comparing RIs between helium and hydrogen methods: As seen in the columns listing the difference between the RI measured with helium and that with hydrogen (labeled RI He-RI H2  ), the agreement of the RIs measured with both methods is good. The only exceptions are the earliest peaks and those such as linalool and D-limonene that are chromatographically overloaded, thus shifting their RT. This is one of the benefits of using the Method Translator tool to choose the chromatographic parameters for the hydrogen method, because it maintains the same relative elution order for analytes and RI calibrators between the two methods. – LMS versus NIST23: The deconvolution process in general yields cleaner spectra, which results in improved LMS scores when compared to previous approaches. Looking at the column of LMS scores for the helium results, most of them are > 85. The smaller values can result from smaller responding compounds, overlapping peaks causing spectral interferences, or search results where the analyte is not in the library and an incorrect hit is listed. – Comparing LMS between helium and hydrogen methods: The columns listing the difference between the LMS measured with helium and that with hydrogen (labeled LMS He-LMS H2  ) shows that, in general, there is good agreement between the two methods. The exceptions where the hydrogen method values are significantly lower are either due to lower signal response or spectral interference from overlapping peaks. In general, the signal-to-noise ratios obtained with hydrogen are two to five times less than with helium, and this is reflected in lower LMS scores for the smallest peaks. 114 15 Table 3. Analysis results for the Brazilian orange oil with both the helium and hydrogen methods. (LR = low response; Int = interference.) Compound Name CAS Lib RI Helium Hydrogen with Hydroinert RI HeRI H2 LMS HeLMS H2 RT RI Delta RI LMS RT RI Delta RI LMS Ethane, 1,1-diethoxy- 105-57-7 726 2.445 723 3 98 1.061 732 -6 96 -9 2 Nonane 111-84-2 900 4.807 900 0 95 1.983 900 0 91 0 4 (1R)-2,6,6-Trimethylbicyclo[3.1.1]hept-2-ene 7785-70-8 932 5.688 933 -1 99 2.332 933 -1 97 0 2 1-Heptanol 111-70-6 970 6.566 966 4 98 2.699 968 2 95 -2 2 Bicyclo[3.1.0]hexane, 4-methylene-1-(1-methylethyl)- 3387-41-5 974 6.757 973 1 98 2.759 974 0 98 -1 0 beta-Myrcene 123-35-3 991 7.240 991 0 97 2.953 992 -1 95 -1 2 Octanal 124-13-0 1,003 7.616 1,004 -1 98 3.116 1,006 -3 97 -2 1 3-Carene 13466-78-9 1,011 7.883 1,011 0 98 3.211 1,012 -1 89 -1 9 D-Limonene 5989-27-5 1,031 8.682 1,034 -3 99 3.562 1,037 -6 99 -3 0 1-Methylbicyclo[2.2.1]heptan-exo-2-ol 766-25-6 1,039 8.796 1,037 2 76 3.593 1,039 0 74 -2 2 trans-Sabinene hydrate 17699-16-0 1,070 9.837 1,066 4 74 3.985 1,067 3 77 -1 -3 1-Octanol 111-87-5 1,070 9.919 1,068 2 98 4.016 1,070 0 97 -2 1 cis-Linalool oxide 5989-33-3 1,074 10.039 1,072 2 94 4.063 1,073 1 94 -1 0 trans-Linalool oxide (furanoid) 34995-77-2 1,086 10.630 1,088 -2 95 4.297 1,089 -3 94 -1 0 Benzene, 1-methyl-4-(1-methylethenyl)- 1195-32-0 1,090 10.665 1,089 1 91 4.306 1,090 0 92 -1 -1 Epoxymyrcene,6,7- 29414-55-9 1,090 10.804 1,093 -3 73 4.365 1,094 -4 79 -1 -6 Linalool 78-70-6 1,099 11.051 1,100 -1 97 4.462 1,101 -2 95 -1 1 Nonanal 124-19-6 1,104 11.232 1,104 0 96 4.533 1,105 -1 86 -1 10 LR cis-Pinen-3-ol 1010292-85-2 1,108 11.324 1,107 1 81 4.574 1,108 0 73 -1 8 2-Cyclohexen-1-ol, 1-methyl-4-(1-methylethenyl)-, trans- 7212-40-0 1,123 11.872 1,120 3 95 4.799 1,122 1 96 -2 -1 5-Undecene, 4-methyl- 143185-91-5 1,132 12.217 1,129 3 74 4.925 1,130 2 71 -1 4 7-Oxabicyclo[4.1.0]heptane, 1-methyl-4-(1-methylethenyl)- 1195-92-2 1,133 12.401 1,133 0 86 5.003 1,134 -1 90 -1 -4 cis-p-Mentha-2,8-dien-1-ol 3886-78-0 1,133 12.462 1,135 -2 79 5.042 1,137 -4 89 -2 -10 (+)-(E)-Limonene oxide 6909-30-4 1,139 12.586 1,138 1 97 5.082 1,139 0 98 -1 -1 Cyclohexanol, 1-methyl-4-(1-methylethenyl)-, cis- 7299-41-4 1,144 12.826 1,143 1 95 5.171 1,145 -1 93 -2 1 Cyclohexanol, 1-methyl-4-(1-methylethenyl)-, trans- 7299-40-3 1,161 13.610 1,163 -2 88 5.480 1,164 -3 83 -1 4 Bicyclo[3.3.0]octan-2-one, 7-methylene- 1000151-92-1 1,166 13.822 1,168 -2 84 5.578 1,170 -4 64 -2 19 Int 1-Nonanol 143-08-8 1,173 13.921 1,170 3 98 5.603 1,171 2 96 -1 2 Ethanone, 1-(4-methylphenyl)- 122-00-9 1,183 14.452 1,183 0 93 5.822 1,185 -2 95 -2 -2 p-Mentha-1(7),8-dien-2-ol 35907-10-9 1,186 14.596 1,187 -1 97 5.873 1,188 -2 96 -1 1 alpha-Terpineol 98-55-5 1,189 14.730 1,190 -1 95 5.928 1,191 -2 94 -1 1 Bicyclo[3.1.1]hept-2-ene-2-methanol, 6,6-dimethyl- 515-00-4 1,195 14.966 1,196 -1 84 6.018 1,197 -2 86 -1 -2 Decanal 112-31-2 1,206 15.374 1,206 0 94 6.173 1,206 0 93 0 1 Acetic acid, octyl ester 112-14-1 1,210 15.663 1,212 -2 91 6.287 1,213 -3 79 -1 12 LR, Int Dihydro carveol, iso- 18675-35-9 1,212 15.756 1,215 -3 85 6.334 1,216 -4 78 -1 7 trans-Carveol 1197-07-5 1,217 15.946 1,219 -2 95 6.427 1,221 -4 96 -2 -1 trans-3(10)-Caren-2-ol 1010151-66-5 1,227 16.281 1,227 0 89 6.550 1,229 -2 85 -2 4 2-Cyclohexen-1-ol, 2-methyl-5-(1-methylethenyl)-, cis- 1197-06-4 1,229 16.423 1,230 -1 96 6.616 1,232 -3 97 -2 -1 2,4-Cycloheptadien-1-one, 2,6,6-trimethyl- 503-93-5 1,238 16.606 1,235 3 81 6.693 1,237 1 73 -2 8 Benzaldehyde, 4-(1-methylethyl)- 122-03-2 1,239 16.796 1,239 0 73 6.753 1,241 -2 58 -2 14 LR D-Carvone 2244-16-8 1,246 16.989 1,244 2 98 6.842 1,246 0 98 -2 0 Geraniol 106-24-1 1,255 17.486 1,255 0 72 7.026 1,257 -2 77 -2 -5 p-Mentha-1(7),8(10)-dien-9-ol 29548-13-8 1,262 17.617 1,258 4 75 7.070 1,259 3 73 -1 2 115 16 Compound Name CAS Lib RI Helium Hydrogen with Hydroinert RI HeRI H2 LMS HeLMS H2 RT RI Delta RI LMS RT RI Delta RI LMS 4-Cyclohexylidenebutyraldehyde 937-59-7 1,268 17.871 1,264 4 75 7.178 1,266 2 76 -2 -2 2-Cyclohexen-1-one, 3-methyl-6-(1-methylethenyl)-, (S)- 16750-82-6 1,270 18.140 1,271 -1 91 7.290 1,272 -2 83 -1 7 1-Cyclohexene-1-carboxaldehyde, 4-(1-methylethenyl)- 2111-75-3 1,273 18.233 1,273 0 97 7.320 1,274 -1 84 -1 14 LR Carvone oxide 33204-74-9 1,279 18.430 1,278 1 96 7.400 1,279 0 96 -1 0 Pinocarvyl acetate, cis- 73366-18-4 1,285 18.757 1,285 0 80 7.523 1,286 -1 82 -1 -2 Verbenyl acetate, trans- 1203-21-0 1,291 18.907 1,289 2 79 7.588 1,290 1 80 -1 0 p-Mentha-1,8-dien-7-ol 536-59-4 1,297 19.266 1,297 0 95 7.731 1,298 -1 93 -1 2 2-Propanol, 1-[(1-ethynylcyclohexyl)oxy]- 54644-17-6 1,303 19.534 1,304 -1 77 7.842 1,305 -2 79 -1 -1 1,2-Cyclohexanediol, 1-methyl-4-(1-methylethenyl)- 1946-00-5 1,321 20.266 1,321 0 80 8.134 1,323 -2 84 -2 -4 (1S,4R,5R)-1,3,3-Trimethyl-2-oxabicyclo[2.2.2]octan5-yl acetate 81781-24-0 1,343 21.019 1,339 4 83 8.459 1,342 1 82 -3 1 exo-2-Hydroxycineole acetate 57709-95-2 1,344 21.273 1,345 -1 86 8.549 1,347 -3 89 -2 -2 1-Cyclohexene-1-methanol, 4-(1-methylethenyl)-, formate 29621-55-4 1,356 21.612 1,353 3 87 8.664 1,354 2 90 -1 -4 2-Cyclohexen-1-ol, 2-methyl-5-(1-methylethenyl)-, acetate, (1R-cis)- 7111-29-7 1,358 21.856 1,359 -1 80 8.775 1,361 -3 79 -2 1 Copaene 3856-25-5 1,376 22.546 1,375 1 91 9.024 1,376 0 84 -1 6 2-Methyl-4-(2,6,6-trimethylcyclohex-2-enyl)but-3-en2-ol 56763-65-6 1,406 23.737 1,404 2 83 9.528 1,406 0 85 -2 -2 cis-β-Copaene 18252-44-3 1,432 24.728 1,428 4 93 9.897 1,429 3 89 -1 3 Sesquicineole, 7-epi-1,2-dehydro- 149067-90-3 1,471 26.498 1,472 -1 78 10.618 1,473 -2 79 -1 -1 3-Tetradecen-5-yne, (E)- 74744-44-8 1,488 27.032 1,485 3 72 10.842 1,487 1 73 -2 -1 Valencene 4630-07-3 1,492 27.311 1,492 0 97 10.930 1,493 -1 96 -1 0 Caryophyllene oxide 1139-30-6 1,581 30.784 1,582 -1 73 12.320 1,582 -1 81 0 -7 1,5,9-Cyclododecanetriol 2938-55-8 2,007 45.413 2,005 2 72 18.190 2,006 1 76 -1 -4 3-Eicosyne 61886-66-6 2,032 46.366 2,036 -4 75 18.573 2,037 -5 76 -1 -2 Uvidin C, diacetate 1000501-90-0 2,107 48.516 2,107 0 75 19.432 2,107 0 73 0 2 (9E,11E)-Octadecadienoic acid 544-71-8 2,237 52.406 2,241 -4 74 20.997 2,241 -4 73 0 2 Incensole oxide, acetate 1000513-23-1 2,270 53.383 2,275 -5 72 21.395 2,276 -6 69 -1 3 116 17 Compound Name CAS Lib RI Helium Hydrogen with Hydroinert RI HeRI H2 LMS HeLMS H2 RT RI Delta RI LMS RT RI Delta RI LMS Ethane, 1,1-diethoxy- 105-57-7 726 2.445 723 3 98 1.060 732 -6 97 -9 1 3-Hexen-1-ol 544-12-7 856 3.982 851 5 93 1.685 856 0 80 -5 13 LR 1-Hexanol 111-27-3 868 4.184 863 5 95 1.762 867 1 92 -4 3 Bicyclo[3.1.0]hex-2-ene, 2-methyl-5-(1- methylethyl)- 2867-05-2 929 5.501 926 3 98 2.258 926 3 92 0 6 (1R)-2,6,6-Trimethylbicyclo[3.1.1]hept-2-ene 7785-70-8 932 5.688 933 -1 98 2.334 933 -1 98 0 1 Camphene 79-92-5 952 6.087 948 4 98 2.496 949 3 97 -1 1 Benzaldehyde 100-52-7 962 6.380 959 3 98 2.628 961 1 96 -2 2 Bicyclo[3.1.1]heptane, 6,6-dimethyl-2- methylene-, (1S)- 18172-67-3 978 6.899 978 0 97 2.824 980 -2 98 -2 0 m-Mentha-4,8-diene, (1S,3S)-(+)- 5208-51-5 983 7.047 984 -1 82 2.875 985 -2 77 -1 5 beta-Myrcene 123-35-3 991 7.240 991 0 97 2.952 992 -1 97 -1 0 Cyclohexane, 1-methylene-4-(1-methylethenyl)- 499-97-8 1,004 7.667 1,005 -1 98 3.121 1,006 -2 97 -1 1 cis-Anhydrolinalool oxide 54750-69-5 1,007 7.775 1,008 -1 93 3.166 1,009 -2 79 -1 14 LR, Int 3-Carene 13466-78-9 1,011 7.874 1,011 0 97 3.203 1,012 -1 97 -1 1 1,3-Cyclohexadiene, 1-methyl-4-(1-methylethyl)- 99-86-5 1,017 8.082 1,017 0 97 3.285 1,018 -1 97 -1 0 D-Limonene 5989-27-5 1,031 8.536 1,029 2 99 3.476 1,031 0 99 -2 0 1,3,6-Octatriene, 3,7-dimethyl-, (Z)- 3338-55-4 1,038 8.782 1,036 2 97 3.562 1,037 1 97 -1 0 trans-beta-Ocimene 3779-61-1 1,049 9.174 1,047 2 97 3.724 1,049 0 97 -2 0 gamma-Terpinene 99-85-4 1,060 9.540 1,058 2 99 3.860 1,058 2 99 0 0 trans-Sabinene hydrate 17699-16-0 1,070 9.843 1,066 4 90 3.994 1,068 2 81 -2 9 cis-Linalool oxide 5989-33-3 1,074 10.047 1,072 2 98 4.071 1,073 1 96 -1 2 1,4-Undecadiene, (Z)- 55976-14-2 1,080 10.241 1,077 3 81 4.138 1,078 2 80 -1 1 Cyclohexene, 1-methyl-4-(1-methylethylidene)- 586-62-9 1,088 10.638 1,088 0 98 4.294 1,089 -1 98 -1 0 Linalool 78-70-6 1,099 11.285 1,106 -7 98 4.618 1,111 -12 98 -5 0 Phenylethyl alcohol 60-12-8 1,116 11.611 1,114 2 98 4.725 1,117 -1 96 -3 1 2-Cyclohexen-1-ol, 1-methyl-4-(1-methylethyl)-, cis29803-82-5 1,122 11.938 1,122 0 93 4.837 1,124 -2 93 -2 0 2,4,6-Octatriene, 2,6-dimethyl-, (E,Z)- 7216-56-0 1,131 12.231 1,129 2 99 4.935 1,130 1 99 -1 0 2-Isopropylimidazole 36947-68-9 1,132 12.318 1,131 1 71 4.973 1,132 0 70 -1 1 cis-p-Mentha-2,8-dien-1-ol 3886-78-0 1,133 12.460 1,134 -1 83 5.029 1,136 -3 82 -2 1 Benzyl nitrile 140-29-4 1,144 12.594 1,138 6 98 5.095 1,140 4 94 -2 5 Myroxide 28977-57-3 1,140 12.747 1,142 -2 92 5.144 1,143 -3 71 -1 20 LR, Int Terpinen-4-ol 562-74-3 1,177 14.182 1,177 0 96 5.709 1,178 -1 96 -1 1 Benzenemethanol, alpha,alpha,4-trimethyl- 1197-01-9 1,183 14.515 1,185 -2 91 5.864 1,187 -4 90 -2 1 alpha-Terpineol 98-55-5 1,189 14.790 1,192 -3 99 5.972 1,194 -5 99 -2 0 1,3-Cyclohexadiene-1-carboxaldehyde, 2,6,6-trimethyl- 116-26-7 1,201 15.153 1,200 1 78 6.099 1,202 -1 79 -2 -2 (3E,5E)-2,6-Dimethylocta-3,5,7-trien-2-ol 206115-88-0 1,202 15.481 1,208 -6 88 6.229 1,210 -8 89 -2 -1 Acetic acid, octyl ester 112-14-1 1,210 15.648 1,212 -2 92 6.285 1,213 -3 89 -1 3 Benzofuran, 2-ethenyl- 7522-79-4 1,220 15.994 1,220 0 89 6.428 1,221 -1 86 -1 4 2,6-Octadien-1-ol, 3,7-dimethyl-, (Z)- 106-25-2 1,228 16.337 1,228 0 97 6.577 1,230 -2 98 -2 0 Neral 106-26-3 1,240 16.860 1,241 -1 90 6.776 1,242 -2 88 -1 2 Carvone 99-49-0 1,242 16.966 1,243 -1 87 6.821 1,245 -3 86 -2 2 Linalyl acetate 115-95-7 1,257 17.587 1,258 -1 95 7.079 1,260 -3 92 -2 3 Citral 5392-40-5 1,273 18.124 1,270 3 93 7.283 1,272 1 89 -2 4 Table 4. Analysis results for neroli oil with both the helium and hydrogen methods. (LR = low response; Int = interference.) 117 18 Compound Name CAS Lib RI Helium Hydrogen with Hydroinert RI HeRI H2 LMS HeLMS H2 RT RI Delta RI LMS RT RI Delta RI LMS 2,6-Octadien-1-ol, 3,7-dimethyl-, formate, (Z)- 2142-94-1 1,282 18.575 1,281 1 89 7.452 1,282 0 89 -1 0 Levo-bornyl acetate 5655-61-8 1,285 18.780 1,286 -1 99 7.534 1,287 -2 95 -1 4 Indole 120-72-9 1,294 18.998 1,291 3 99 7.637 1,293 1 99 -2 0 Benzene, (2-nitroethyl)- 6125-24-2 1,302 19.280 1,298 4 89 7.746 1,299 3 86 -1 3 Geranyl formate 105-86-2 1,300 19.483 1,303 -3 91 7.811 1,303 -3 91 0 1 δ-EIemene 20307-84-0 1,338 20.958 1,338 0 86 8.392 1,338 0 96 0 -10 Methyl anthranilate 134-20-3 1,343 21.020 1,339 4 94 8.439 1,341 2 98 -2 -4 alpha-Terpinyl acetate 80-26-2 1,350 21.445 1,349 1 97 8.593 1,350 0 96 -1 2 6-Octen-1-ol, 3,7-dimethyl-, acetate 150-84-5 1,354 21.647 1,354 0 84 8.669 1,354 0 81 0 2 2,6-Octadien-1-ol, 3,7-dimethyl-, acetate, (Z)- 141-12-8 1,364 22.133 1,366 -2 99 8.881 1,367 -3 99 -1 0 Copaene 3856-25-5 1,376 22.543 1,375 1 82 9.026 1,376 0 80 -1 2 Geranyl acetate 105-87-3 1,382 22.966 1,385 -3 98 9.222 1,387 -5 98 -2 0 levo-β-Elemene 515-13-9 1,391 23.236 1,392 -1 96 9.307 1,393 -2 95 -1 1 Benzoic acid, 2-amino-, ethyl ester 87-25-2 1,414 24.091 1,413 1 86 9.662 1,414 0 93 -1 -7 Caryophyllene 87-44-5 1,419 24.340 1,419 0 99 9.751 1,420 -1 99 -1 0 gamma-Elemene 29873-99-2 1,434 24.921 1,433 1 83 9.977 1,434 0 88 -1 -6 Humulene 6753-98-6 1,454 25.707 1,453 1 92 10.293 1,453 1 91 0 1 (E)-beta-Farnesene 18794-84-8 1,457 25.894 1,457 0 96 10.363 1,458 -1 96 -1 0 Alloaromadendrene 25246-27-9 1,461 26.002 1,460 1 92 10.408 1,460 1 84 0 8 gamma-Muurolene 30021-74-0 1,477 26.657 1,476 1 93 10.669 1,476 1 91 0 2 Germacrene D 23986-74-5 1,481 26.824 1,480 1 95 10.736 1,481 0 97 -1 -2 Bicyclogermacrene 24703-35-3 1,496 27.445 1,496 0 98 10.985 1,496 0 97 0 1 alpha-Muurolene 10208-80-7 1,499 27.605 1,500 -1 96 11.048 1,500 -1 93 0 3 alpha-Farnesene 502-61-4 1,508 27.955 1,509 -1 94 11.186 1,509 -1 96 0 -1 γ-Cadinene 39029-41-9 1,513 28.137 1,513 0 97 11.261 1,514 -1 96 -1 1 δ-Cadinene 483-76-1 1,524 28.519 1,523 1 97 11.413 1,524 0 97 -1 1 α-Cadinene 24406-05-1 1,538 29.047 1,537 1 80 11.626 1,537 1 81 0 -2 alpha-Calacorene 21391-99-1 1,542 29.254 1,542 0 94 11.710 1,543 -1 82 -1 13 LR β-Germacrene 15423-57-1 1,557 29.779 1,556 1 96 11.915 1,556 1 87 0 10 Int 1,6,10-Dodecatrien-3-ol, 3,7,11-trimethyl- 7212-44-4 1,564 30.146 1,565 -1 97 12.091 1,567 -3 97 -2 0 Spathulenol 6750-60-3 1,576 30.574 1,576 0 93 12.244 1,577 -1 85 -1 9 Caryophyllene oxide 1139-30-6 1,581 30.788 1,582 -1 92 12.329 1,583 -2 90 -1 2 tau-Cadinol 5937-11-1 1,640 32.985 1,641 -1 96 13.198 1,641 -1 95 0 2 δ-Cadinol 19435-97-3 1,645 33.163 1,646 -1 92 13.276 1,646 -1 88 0 4 alpha-Cadinol 481-34-5 1,653 33.462 1,654 -1 96 13.396 1,654 -1 94 0 2 Naphthalene, 1,6-dimethyl-4-(1-methylethyl)- 483-78-3 1,674 34.205 1,675 -1 92 13.695 1,674 0 78 1 14 LR, Int 8-Heptadecene 2579-04-6 1,677 34.320 1,678 -1 95 13.731 1,677 0 94 1 2 6,10-Dodecadien-1-ol, 3,7,11-trimethyl- 51411-24-6 1,692 34.795 1,691 1 81 13.928 1,690 2 87 1 -6 2,6,10-Dodecatrien-1-ol, 3,7,11-trimethyl-, (Z,E)- 3790-71-4 1,697 35.110 1,700 -3 92 14.058 1,699 -2 89 1 3 trans-Farnesol 106-28-5 1,722 35.968 1,724 -2 97 14.418 1,724 -2 98 0 -1 Farnesol, 2E, 6Z- 3879-60-5 1,742 36.652 1,743 -1 79 14.676 1,743 -1 78 0 1 all-trans-Farnesyl acetate 4128-17-0 1,843 40.112 1,842 1 94 16.057 1,843 0 92 -1 3 Cubitene 66723-19-1 1,878 41.166 1,874 4 82 16.479 1,874 4 78 0 3 m-Camphorene 20016-73-3 1,960 43.728 1,952 8 95 17.508 1,953 7 93 -1 3 p-Camphorene 20016-72-2 1,995 44.815 1,986 9 92 17.942 1,986 9 91 0 1 Hexadecanoic acid, ethyl ester 628-97-7 1,993 45.083 1,995 -2 73 18.054 1,995 -2 75 0 -2 118 19 Conclusion Using the techniques described in this application note, a typical method for the qualitative analysis of essential oils was successfully converted to a method using hydrogen carrier gas. The resulting hydrogen method retains the same chromatographic resolution and relative elution order of the original method, but with a run time 2.5 times shorter. The column capacity with the new method is reduced, being a calculated 36% of the original method, so the amount injected may need adjustment in some cases. The new hydrogen method was then applied to the analysis of two essential oils and found to generate results comparable to those from the helium method. The use of spectral deconvolution and retention index search filtering with Agilent MassHunter Unknowns Analysis gave improved search results that were faster than previous identification methods. The NIST23 library, with expanded content for essential oil components and RIs, allowed identification of a significant portion of the compounds present. The Agilent HydroInert source is a key component in the successful conversion to hydrogen. Without it, in-source reactions were shown to degrade the spectra of some compounds to the point where they were misidentified. References 1. Agilent EI GC/MS Instrument Helium to Hydrogen Carrier Gas Conversion. Agilent Technologies user guide, publication number 5994-2312EN, 2020. 2. Blumberg, L. M. Method Translation in Gas Chromatography. US Patent US6634211B1. 2002. 3. Blumberg, L. M.; Klee, M. S. Method Translation and Retention Time Locking in Partition GC. Anal. Chem. 1998, 70(18), 3828–3839. 4. Agilent GC Calculators and Method Translation Software. Tools available for download from: https://www.agilent. com/en/support/gas-chromatography/gccalculators 5. Agilent Inert Plus GC/MS System with HydroInert Source Applying H2 Carrier Gas to Real-World GC/MS Analyses. Agilent Technologies technical overview, publication number 5994-4889EN, 2022. 6. Godina, L. Flavor and Fragrance GC/MS Analysis with Hydrogen Carrier Gas and the Agilent HydroInert Source, Agilent Technologies application note, publication number 5994-6015EN, 2023. 7. MassHunter Unknowns Analysis Video: https://www. youtube.com/watch?v=y_zJkBfnN3g 8. Adams, R. Identification of Essential Oil Components by Gas Chromatography/Mass Spectrometry, version 4, available through Diablo Analytical (Antioch, CA). 9. Sparkman, O. D. NIST 23: The Largest Increases in Compound Coverage for the Tandem and NIST/EPA/ NIH EI Libraries Since NIST Became Curator, Separation Science, July 2023. NIST 23: The Largest Increases in Compound Coverage for the Tandem and NIST/EPA/ NIH EI Libraries Since NIST Became Curator - Separation Science (sepscience.com) 10. David, F.; Scanlan, F.; Sandra, P.; Szelewski, M. Analysis of Essential Oil Compounds Using Retention Time Locked Methods and Retention Time Databases, Agilent Technologies application note, publication number 5988-6530EN, 2002. 119 20 Appendix Setting up MassHunter Unknowns Analysis This section shows how to set up Agilent MassHunter Unknowns Analysis. The parameters shown here were used in this specific application. For other applications, different parameters can be used to optimize the process. Locate the Agilent MassHunter Quantitative Analysis software folder under the Microsoft Windows Start menu and open the MassHunter Unknowns Analysis program. 1. Click File > New Analysis. Navigate to the directory containing your data files. 2. Enter a file name for the analysis. 3. Click File > Add Samples. Select the data files that you want to process. You should see the TIC of your chromatogram appear. 4. Click Method > Edit. Set the parameters as shown in Appendix Figures 1 to 5. Leave these at defaults Leave these at defaults Set these Appendix Figure 1. Setting the parameters for the Peak Detection and Deconvolution tabs in Agilent MassHunter Unknowns Analysis. 120 21 Leave these at defaults Set these values Select NIST23.L Select the .rtc file created in Notepad Appendix Figure 2. Setting the parameters for the Library Search tab in Agilent MassHunter Unknowns Analysis. Leave these at defaults for now Set to 70 Appendix Figure 3. Setting the parameters for the Compound Identification tab in Agilent MassHunter Unknowns Analysis. Leave everything at defaults Appendix Figure 4. Setting the parameters for the Target Match tab in Agilent MassHunter Unknowns Analysis. Leave these at defaults for now Select Apply to Selected Sample, then Close. After closing, go to the Analyze menu and select Analyze Sample. Note: it will take a while to process. Uncheck this Appendix Figure 5. Setting the parameters for the Blank Subtraction tab in Agilent MassHunter Unknowns Analysis. 121 22 Appendix Figure 6. Example result of a deconvoluted/searched data file in Agilent MassHunter Unknowns Analysis. Right-click the title bar and select Add/Remove Columns. Select these columns for now. Library spectrum More examples of in-source reactions eliminated with the HydroInert source In this section, the spectra of several compounds from the Brazilian orange and neroli oils are examined. For each compound, the spectrum obtained using helium and the 3 mm inert extractor source is presented with the spectrum for each compound obtained using hydrogen with the 9 mm HydroInert source, and the inert extractor source with the 9 and 3 mm extractor lenses. As shown in Figures 7 to 11, changes in the hydrogen spectra without the HydroInert source are clear, resulting in reduced LMS values and reduced confidence in identification. 122 23 Appendix Figure 7. Linalool in Brazilian orange oil. Red arrows indicate spectral changes observed with hydrogen carrier and non-HydroInert sources. 3 mm Inert extractor, He LMS = 96.7 9 mm HydroInert, H2 LMS = 95.3 Component RT: 11.0513 40 50 60 70 80 90 100 110 120 130 140 150 160 Counts -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 71.0 93.0 93.0 71.0 43.0 55.0 41.0 55.0 80.0 80.0 121.0 121.0 109.0 136.0 107.0 136.0 150.0 154.0 Component RT: 4.4621 40 50 60 70 80 90 100 110 120 130 140 150 160 170 170 Counts -1.0 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1.0 1.0 71.0 93.1 71.0 93.0 43.0 55.0 41.1 55.1 80.1 80.0 121.1 121.0 105.1 136.0 136.1 107.0 149.9 154.0 166.9 Component RT: 4.4841 40 50 60 70 80 90 100 110 120 130 140 150 160 170 Counts -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 69.1 71.0 41.1 93.1 41.0 93.0 55.0 55.0 79.1 121.1 80.0 121.0 105.1 136.1 107.0 136.0 151.9 9 mm Inert extractor, H2 LMS = 89.5 Component RT: 4.4812 40 50 60 70 80 90 100 110 120 130 140 150 160 170 Counts -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1.0 1.0 71.0 69.1 93.1 93.0 41.1 43.0 55.0 55.1 79.1 121.1 80.0 121.0 105.1 136.2 107.0 136.0 154.0 154.0 3 mm Inert extractor, H2 LMS = 89.7 Brazilian orange: Linalool (CAS 78-70-6) ×102 ×102 ×102 ×102 Mass-to-charge (m/z) Mass-to-charge (m/z) Appendix Figure 8. cis-Carveol in Brazilian orange oil. Red arrows indicate spectral changes observed with hydrogen carrier and non-HydroInert sources. Counts Counts Counts Counts ×102 ×102 ×102 ×102 Mass-to-charge (m/z) Mass-to-charge (m/z) Component RT: 16.4233 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 84.0 84.1 109.1 134.1 91.0 41.0 109.0 55.0 134.0 55.1 69.0 69.0 41.1 93.0 119.0 152.0 152.1 Component RT: 6.6160 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 84.0 84.0 109.0 41.0 91.0 119.0 109.0 134.0 55.0 69.0 134.0 55.0 69.0 41.0 91.0 119.0 152.0 152.0 Component RT: 6.6374 40 50 60 70 80 90 100 110 120 130 140 150 160 170 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0 0 0.2 0.4 0.6 0.8 1.0 84.0 91.1 119.1 134.0 134.1 41.0 79.0 109.0 79.1 55.0 91.0 67.1 109.1 41.1 69.0 55.0 152.0 151.9 Component RT: 6.6350 40 50 60 70 80 90 100 110 120 130 140 150 160 170 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.2 0.4 0.6 0.8 1.0 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0 0 0.2 0.4 0.6 0.8 1.0 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.2 0.4 0.6 0.8 1.0 84.0 119.0 91.0 134.0 134.0 41.0 84.0 79.0 109.0 109.0 55.0 91.0 67.0 69.0 41.0 55.0 152.0 152.0 9 mm Inert extractor, H2 LMS = 88.0 3 mm Inert extractor, H2 LMS = 90.0 Brazilian orange: cis-Carveol (CAS 1197-06-4) 3 mm Inert extractor, He LMS = 95.8 9 mm HydroInert, H2 LMS = 96.7 123 24 Appendix Figure 9. trans-Carveol in Brazilian orange oil. Red arrows indicate spectral changes observed with hydrogen carrier and non-HydroInert sources. Counts Counts Counts Counts ×102 ×102 ×102 ×102 Mass-to-charge (m/z) Mass-to-charge (m/z) Component RT: 15.9464 Component RT: 6.4486 Component RT: 6.4268 Component RT: 6.4461 9 mm Inert extractor, H2 LMS = 84.4 3 mm Inert extractor, H2 LMS = 87.6 Brazilian orange: trans-Carveol (CAS 1197-07-5) 3 mm Inert extractor, He LMS = 94.8 9 mm HydroInert, H2 LMS = 96.0 40 50 60 70 80 90 100 110 120 130 140 150 160 109.0 109.0 84.0 41.0 55.0 84.0 119.0 91.0 69.0 55.0 77.0 134.0 119.0 41.0 69.0 77.0 137.0 152.0 152.0 40 50 60 70 80 90 100 110 120 130 140 150 160 109.0 109.1 84.0 41.0 84.0 55.0 69.0 91.0 119.1 41.1 55.1 69.1 119.0 137.1 137.0 152.0 151.9 40 50 60 70 80 90 100 110 120 130 140 150 160 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1.0 109.0 119.0 91.0 84.0 134.0 41.0 84.0 55.0 91.0 119.0 152.0 67.0 69.0 41.0 55.0 134.0 152.0 40 50 60 70 80 90 100 110 120 130 140 150 160 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1.0 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1.0 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1.0 91.1 109.0 119.1 84.0 109.1 134.1 41.0 55.0 77.1 91.0 119.0 67.1 69.0 152.0 41.1 134.0 55.0 151.9 Appendix Figure 10. p-Cymen-8-ol in neroli oil. Red arrows indicate spectral changes observed with hydrogen carrier and non-HydroInert sources. 3 mm Inert extractor, He LMS = 91.0 9 mm HydroInert, H2 LMS = 89.6 Component RT: 14.5145 40 50 60 70 80 90 100 110 120 130 140 150 160 Counts -1.2 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 Component RT: 5.8637 40 50 60 70 80 90 100 110 120 130 140 150 160 170 40 50 60 70 80 90 100 110 120 130 140 150 160 170 170 Counts -1.0 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1.0 1.0 Component RT: 5.8893 Counts -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 9 mm Inert extractor, H2 LMS = 75.4 Component RT: 5.8907 40 50 60 70 80 90 100 110 120 130 140 150 160 170 Counts -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1.0 1.0 3 mm Inert extractor, H2 LMS = 79.8 Neroli oil: p-Cymen-8-ol (CAS 1197-01-9) ×102 ×102 ×102 ×102 Mass-to-charge (m/z) Mass-to-charge (m/z) 43.0 132.0 117.0 43.0 135.0 91.0 91.0 65.0 65.0 117.0 51.0 77.0 77.0 150.0 105.0 102.0 150.0 43.0 135.0 43.1 117.0 135.0 91.1 91.0 65.1 65.0 117.0 77.0 105.1 150.1 39.0 77.0 51.1 51.0 105.0 150.0 43.0 43.1 134.9 91.1 135.0 116.9 91.0 64.9 77.1 65.0 117.0 50.9 149.9 51.0 77.0 105.0 101.9 150.0 167.9 43.0 117.0 132.0 43.0 91.0 135.0 91.0 65.0 65.0 117.0 51.0 77.0 105.0 77.0 150.0 51.0 105.0 150.0 124 www.agilent.com DE52560836 This information is subject to change without notice. © Agilent Technologies, Inc. 2024 Printed in the USA, January 26, 2024 5994-7058EN Appendix Figure 11. (3E,5E)-2,6-Dimethylocta-3,5,7-trien-2-ol in neroli oil. Red arrows indicate spectral changes observed with hydrogen carrier and non-HydroInert sources. 3 mm Inert extractor, He LMS = 88.1 9 mm HydroInert, H2 LMS = 88.9 40 50 60 70 80 90 100 110 120 130 140 150 160 Counts -1.2 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 40 50 60 70 80 90 100 110 120 130 140 150 160 170 40 50 60 70 80 90 100 110 120 130 140 150 160 170 170 Counts -1.0 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1.0 1.0 Component RT: 6.2480 Component RT: 6.2447 Counts -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 9 mm Inert extractor, H2 LMS = 77.3 40 50 60 70 80 90 100 110 120 130 140 150 160 170 Counts -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1.0 1.0 3 mm Inert extractor, H2 LMS = 75.6 Neroli oil: (3E,5E)-2,6-Dimethylocta-3,5,7-trien-2-ol (CAS 206115-88-0) ×102 ×102 ×102 ×102 Mass-to-charge (m/z) Mass-to-charge (m/z) Component RT: 15.4812 43.0 91.0 91.0 109.0 119.0 43.0 81.0 119.0 79.0 67.0 134.0 55.0 152.0 152.0 134.0 55.0 67.0 Component RT: 6.2287 43.0 91.0 91.0 109.0 79.0 81.0 119.0 43.0 67.0 55.0 152.0 109.0 55.0 67.0 134.0 134.0 152.0 72.0 43.0 93.0 121.0 77.0 91.0 109.0 81.0 119.0 109.0 136.0 67.0 41.0 67.0 55.0 152.0 134.0 55.0 152.0 72.0 43.0 93.0 109.0 121.0 91.0 81.0 77.0 119.0 105.0 136.0 67.0 55.0 152.0 134.0 41.0 53.0 67.0 84.0 152.0 Application Note Food Testing Authors Praveen Arya, Soma Dasgupta, and Vivek Dhyani Agilent Technologies, Inc. Abstract This application note demonstrates the use of an Agilent 8890 GC system coupled with an Agilent 7010 GC/MS triple quadrupole mass spectrometry system to detect and quantify ethylene oxide (EtO) and 2-chloroethanol (2-CE) simultaneously in foodstuffs such as flaxseed, cumin powder, and red chili powder. Sample preparation was done using QuEChERS extraction and a dispersive cleanup process followed by injection into GC/MS/MS through liquid injection. A limit of quantification (LOQ) of 10 ppb for both compounds was achieved in matrix. Average recoveries ranged from 75 to 86% for both compounds. Estimation of Ethylene Oxide and 2-Chloroethanol in Spices and Oilseeds Using QuEChERS Extraction and GC/MS/MS Agilent 8890/7010 triple quadrupole GC/MS system 126 Return to Table of Contents 2 Introduction EtO is used to sterilize foods to eliminate insects and bacteria, such as salmonella. Ethylene chlorohydrin, or 2-CE, is a derivative produced by the reaction of EtO with chlorine ions present in foodstuff. Use of EtO is banned in the European Union (EU) due to its carcinogenic and toxic properties. Previously, methods for analysis of EtO (EtO and 2-CE) have been developed that include acidic conversion of EtO to 2-CE. These methods are time-consuming, labor-intensive, and require large quantities of harmful solvents. Due to the volatile nature of EtO, sample preparation is crucial. In December 2020, the EU Reference Laboratories (EURL) for Residues of Pesticides recommended a single-residue method for the analysis of EtO and 2-CE in sesame seeds using QuEChERS extraction followed by GC/MS/MS analysis. The method adopted in this work demonstrates the use of an automated liquid sampler for sample introduction to the 8890 GC system coupled with a 7010 GC/MS triple quadrupole mass spectrometry system. Parameter Value GC/MS/MS Method Parameters GC Agilent 8890 GC with G4513 autosampler Mass Spectrometer Agilent 7010 triple quadrupole mass spectrometry system Analytical Column Agilent J&W DB-VRX (60 m x 0.25 mm, 1.4 µm) Column Flow Helium: 1.0 mL/min, constant flow Injection Mode Pulsed split (4:1) Injection Volume 2 µL Injection Program Starts at 90 °C (hold for 0.8 min), ramped at 450 °C/min to 250 °C, hold for 10 min Oven Program Starts at 40 °C (hold for 1.0 min), ramped at 10 °C/min to 160 °C, and then at 30 °C/min to 245 °C, hold for 5 min MS Parameters Ionization mode: EI; Ion source temperature: 230 °C; Quadrupole temperature (Q1 and Q2): 150 °C MRM Transitions EtO 44 & 14 (CE:20) 44 & 28 (CE:5) 44 & 29 (CE:5) 2-CE 80 & 31 (CE:5) 80 & 43 (CE:5) 82 & 31 (CE:5) Table 1. GC/TQ parameters. Experimental Standard preparation Due to the high volatility of EtO, its standard solutions were prepared at a low temperature (< 10 °C). As a diluent, acetonitrile was placed in a freezer (at –20 °C) for at least 15 minutes before use. The cold analytical standard solutions were diluted with cold acetonitrile to prepare the stock standard solutions of EtO and 2-CE, which had a concentration of 1 mg/mL each. The stock standard solution was further diluted with acetonitrile to obtain a working standard solution which had a concentration of 10 µg/ml for EtO and 2-CE. All the calibration standards (2.0, 5.0, 10.0, 20.0, 50.0, 100, and 200 ng/mL) were freshly prepared by diluting the stock solution with acetonitrile. Matrix-matched standards of flaxseeds, cumin powder, and red chili powder were prepared by post-spiking the requisite amount of solution to extracts of each matrix. Before analysis, all stock solutions were preserved at a temperature of −20 °C to avoid degradation. Sample preparation Homogenized samples of flaxseed, cumin powder, and red chili powder were processed using the QuEChERS extraction procedure according to the EN 15662 procedure (Figure 1). Approximately 2.00 ± 0 .01 g of each sample was weighed in 50 mL centrifuge tubes. Ten mL of cold water was added to the centrifuge tubes followed by capping and vortex mixing for 1 minute to ensure sample hydration, followed by the addition of 10.0 mL of cold acetonitrile along with two ceramic homogenizers to improve the extraction efficiency. The centrifuge tubes were capped and shaken for 10 minutes. The QuEChERS extraction salt (4 g of MgSO4 , 1 g of NaCl, 1 g of sodium citrate, and 0.5 g of disodium citrate sesquihydrate) was added, and the tubes were shaken for another 3 minutes. The samples were then centrifuged for 5 minutes at 6,000 rpm. The upper acetonitrile layer (6.0 mL) was transferred to QuEChERS Dispersive Kit 15 mL tubes (150 mg of PSA, 150 mg of C18EC , and 900 mg MgSO4 ). The tubes were vortexed for 30 seconds followed by centrifugation at 5,000 rpm for 5 minutes. After centrifugation, the supernatant was transferred into GC vials for analysis. 127 3 Results and discussion Calibration Matrix calibrations were performed for EtO and 2-CE at concentrations of 2, 5, 10, 20, 50, and 100 ng/mL in solvent. Similar calibrations were performed in post-spike matrix extracts for flaxseed, cumin powder, and red chili powder. Excellent values for R2 > 0.99 were obtained. Figures 8 and 9 show linearity of EtO and 2-CE respectively in acetonitrile. Figures 10 and 11 show linearity of EtO and 2-CE respectively in a flaxseed matrix. Figure 1. QuEChERS workflow for the extraction and cleanup of samples. Standard area repeatability A repeatable elution was obtained by injecting a 10 ppb concentration of EtO and 2-CE in a matrix extract. As shown in Table 2, a % RSD data of EtO and 2-CE is calculated from peak areas of 6 replicate injections of 10 ppb matrix standards in flaxseed, cumin powder, and red chili powder. Recovery EtO and 2-CE were spiked in samples of flaxseed, cumin powder, and red chili powder at concentration levels of 20 and 50 ng/g. Acceptable recoveries were obtained by quantification through post-spike matrix-based calibration. Results are highlighted in Table 3. Transfer 6.00 mL of supernatant to a 15 mL dSPE tube. Cleanup Step Agitate for 30 seconds. Centrifuge at 5,000 rpm for 5 minutes. Transfer the clear supernatant to a 2 mL vial for injection. Weigh 2.00 g ± 0.01 g of sample into a 50 mL centrifuge tube. QuEChERS Extraction Step Add water and shake briefly to wet the sample. Add 10.00 mL of ACN and extraction aids and mix for 15 minutes. Add water and shake briefly to wet the sample. Add extraction salts and mix for 3 minutes. Centrifuge at 6,000 rpm for 5 minutes. 128 4 A B EtO 2-CE Figure 3. Overlaid chromatograms from a set of calibration standards from 2 ng/mL to 100 ng/mL; (A) EtO and (B) 2-CE. Figure 2. Chromatograms for (A) EtO at the 10 ng/mL level and (B) 2-CE at the 10 ng/mL level. A B EtO 2-CE 129 5 Figure 4. Quantifier and qualifier peaks of EtO at 10 ng/g matrix standard in flaxseed. Figure 5. Quantifier and qualifier peaks of 2-CE at 10 ng/g matrix standard in flaxseed. Figure 6. Quantifier and qualifier peaks of EtO at 10 ng/g matrix standard in cumin powder. Figure 7. Quantifier and qualifier peaks of 2-CE at 10 ng/g matrix standard in cumin powder. Figure 8. Calibration curve for EtO in acetonitrile (R2 > 0.998). Figure 9. Calibration curve for 2-CE in acetonitrile (R2 > 0.998). Figure 10. Calibration curve for EtO in flaxseed (R2 > 0.999). Figure 11. Calibration curve for 2-CE in flaxseed (R2 > 0.996). 130 6 Number of Injections (Replicates) EtO (Peak Area) 2-CE (Peak Area) Flaxseed Cumin Powder Chili Powder Flaxseed Cumin Powder Chili Powder 10 ng/mL Injection 1 2,266 2,193 3,135 1,065 893 735 Injection 2 2,264 2,205 3,445 1,049 1,035 910 Injection 3 2,545 2,179 3,425 1,129 1,125 880 Injection 4 2,562 2,162 3,440 993 1,048 755 Injection 5 2,392 2,228 3,385 1,236 1,168 845 Injection 6 2,526 2,135 3,335 1,100 979 785 Mean 2,426 2,184 3,361 1,095 1,041 818 SD 138.32 32.76 118 83.04 99.01 70.68 RSD (%) 5.70 1.50 3.51 7.58 9.51 8.64 Table 2. Standard area repeatability for 6 replicates of 10 ppb matrix standard in flaxseed, cumin powder, and red chili powder. Food Matrix EtO 2-CE Spiked Concentration (ng/g) Average Obtained Concentration (ng/g) Recovery (%) Spiked Concentration (ng/g) Average Obtained Concentration (ng/g) Recovery (%) Flaxseed 20 15.12 75.6 20 15.84 79.2 Cumin Powder 50 39.02 78.0 50 43.29 86.6 Chili Powder 50 38.52 77.04 50 42.47 84.9 Table 3. Percent Recovery of EtO and 2-CE in flaxseed and cumin powder spiked at 20 and 50 ng/g levels, respectively. Conclusion An accurate and rugged method was developed which meets the requirements of the EURL for a single-residue method that uses QuEChERS extraction followed by GC/MS/MS analysis for the analysis of EtO and 2-CE in flaxseed, cumin powder, and red chili powder. The LOQ of the method was demonstrated at 10 ng/g for all the tested matrices. Repeatable results are found for 6 successive replicates of matrix-based standards at 10 ng/g concentration levels for EtO and 2-CE. Excellent recoveries were obtained for both EtO and 2-CE in all tested matrices at 20 and 50 ng/g spiked concentration levels. Thus, this study demonstrates the usefulness of the developed method for the routine analysis of food samples for EtO and 2-CE at trace levels. 131 www.agilent.com DE-000364 This information is subject to change without notice. © Agilent Technologies, Inc. 2024 Printed in the USA, September 3, 2024 5994-7720EN References 1. Dunkelberg, H. Carcinogenic Activity of Ethylene Oxide and Its Reaction Products 2-Chloroethanol, 2-Bromoethanol, Ethylene Glycol, And Diethylene Glycol. I. Carcinogenicity of Ethylene Oxide in Comparison with 1,2-Propylene Oxide after Subcutaneous Administration in Mice. Zentralblatt Fur Bakteriologie Mikrobiologie Und Hygiene. 1. Abt. Originale B Hygiene 1981, 174, 383–404. (PubMed) 2. Tateo, F.; Bononi, M. Determination of Ethylene Chlorohydrin as Marker of Spices Fumigation with Ethylene Oxide. Journal of Food Composition and Analysis 2006, 19, 83–87. 3. Regulation (EU) 2015/868 of 26 May 2015 Amending Annexes II, III and V to Regulation (EC) No 396/2005 of the European Parliament and of the Council as Regards Maximum Residue Levels for 2,4,5-T, Barban, Binapacryl, Bromophos-Ethyl, Camphechlor (Toxaphene), Chlorbufam, Chloroxuron, Chlozolinate, Dnoc, Di-Allate, Dinoseb, Dinoterb, Dioxathion, Ethylene Oxide, Fentin Acetate, Fentin Hydroxide, Flucycloxuron, Flucythrinate, Formothion, Mecarbam, Methacrifos, Monolinuron, Phenothrin, Propham, Pyrazophos, Quinalphos, Resmethrin, Tecnazene and Vinclozolin in or on Certain Products. Off. J. Eur. Union L. 2015, 145, 1–71. 4. Regulation (EC) No 149/2008 of 29 January 2008 Amending Regulation (EC) No 396/2005 of the European Parliament and of the Council by Establishing Annexes II, III and IV Setting Maximum Residue Levels for Products covered by Annex I thereto. Off. J. Eur. Union L. 2008, 58, 1–398. 5. EURL-SRM - Analytical Observation Report: Analysis of Ethylene Oxide and Its Metabolite 2-Chloroethanol by the QuOil or the QuEChERS Method and GC-MS/MS. December 2020. Application Note Flavor and Fragrance Authors Yufeng Zhang and Lay Peng Tan Agilent Technologies Inc. Abstract This application note describes a method for quantitation of four aldehydes (hexanal, furfural, phenylacetaldehyde, and trans-2-nonenal) responsible for off-flavors in beer using an Agilent 8890/5977C GC/MSD with an Agilent PAL3 (SPME) autosampler. The method used fully automated, solvent-free extraction and on-fiber derivatization. The aldehyde compounds were derivatized with the derivatization agent O-(2,3,4,5,6-pentafluorobenzyl)hydroxylamine hydrochloride (PFBHA) using the on-fiber derivatization procedure. The derivatization agent was first adsorbed onto an Agilent 65 µm PDMS/DVB fiber. Next, the fiber was inserted into a 20 mL headspace vial containing a 2 mL beer sample with agitation at 60 °C for 30 minutes. Both extraction and derivatization procedures were automatically performed using the PAL3 autosampler. The method demonstrated excellent sensitivity, with detection limits of 0.0009 µg/L for hexanal, 0.52 µg/L for furfural, 0.015 µg/L for phenylacetaldehyde, and 0.003 µg/L for trans-2-nonenal. The limits of the quantification for the four compounds were 0.003, 1.72, 0.05, and 0.01 µg/L, respectively. The four aldehydes were successfully quantified in four beer samples purchased from a supermarket. Good repeatability was demonstrated with RSD < 4.9% based on three replicate injections of the four beer samples for all four aldehydes. Analysis of Aldehydes in Beer by Agilent PAL3 Autosampler and 5977C GC/MSD Return to Table of Contents 133 2 Introduction Aldehydes are a critical group of compounds in beer that significantly influence its flavor and aroma profile. These volatile organic compounds, even at low concentrations, can impact undesirable sensory characteristics such as cardboard, grassy, or stale flavors, which negatively affect the overall quality and consumer acceptance of the product. The presence of aldehydes is often linked to oxidation processes during brewing, packaging, and storage, making them key indicators of beer freshness and stability.1 Hexanal, furfural, phenylacetaldehyde, and trans-2-nonenal are among the most recognized aldehydes responsible for off-flavors, with flavor thresholds of 350 μg/L, 15.157 mg/L, 105 μg/L, and 0.1 μg/L, respectively.2 Monitoring and controlling aldehyde levels in beer are essential for maintaining product consistency and ensuring a high-quality drinking experience. By accurately quantifying aldehydes, brewers can identify potential issues in their production processes, such as oxygen exposure or ingredient degradation, and implement corrective measures to minimize off-flavors. Additionally, understanding the aldehyde profile of beer can aid in the development of new brewing techniques and formulations that enhance flavor stability and shelf life. Solid-phase microextraction (SPME) is a widely used, solvent-free sample preparation technique that operates on the principles of adsorption and desorption. It uses a fiber coated with an extractive phase to concentrate analytes from a sample. Various types of fibers are available, including PDMS, acrylate, carbon WR, DVB, and combinations of these sorbents, to address the differing polarities of analytes. SPME is extensively used in a range of analyses, including environmental characterization, food and flavor analysis, pharmaceutical studies, and forensics investigations. It is well suited for automated sample preparation, resulting in reduced preparation time per sample, minimized risk of human error, and the freeing up of lab personnel from repetitive tasks. This application note demonstrates the use of a PAL3 RTC sampler with SPME tool and 8890/5977C GC/MSD for the analysis of the four aldehyde compounds in beer. The SPME sampling tool is equipped with its own read-and-write chip with preset parameters including the stationary phase and usage tracking. Both sample extraction and derivatization were automated with the PAL3 RTC sampler. Experimental Reagents and samples Aldehyde standards (hexanal, furfural, phenylacetaldehyde, and trans-2-nonenal) and derivatization reagent PFBHA were purchased from Sigma-Aldrich. HPLC-grade methanol was from Merck. The water was Milli-Q Ultrapure. Four different brands of beer were purchased from a local supermarket. Standards preparation A mixed stock solution consisting of 10 µg/mL hexanal, 10 µg/mL phenylacetaldehyde, 1,000 µg/mL furfural, and 1 µg/mL trans-2-nonenal was prepared in methanol. A secondary mixed stock solution consisting of 100 µg/L hexanal, 100 µg/L phenylacetaldehyde, 10 µg/mL furfural, and 10 µg/L trans-2-nonenal was prepared in ultrapure water. Using the secondary mixed stock solution, a series of calibration standards were prepared with a concentration range of 0.05 to 10 µg/L for hexanal, 5 to 1,000 µg/L for furfural, 0.1 to 50 µg/L for phenylacetaldehyde, and 0.025 to 5 µg/L for trans-2-nonenal. Two milliliters of each calibration standard was added to a 20 mL headspace vial and immediately capped for analysis. PFBHA derivatization reagent with a concentration of 60 mg/L was prepared by weighing 30.1 mg of powder PFBHA into a 500 mL volumetric flask and dissolving in ultrapure water up to the 500 mL mark. Ten milliliters of the 60 mg/L PFBHA solution was added to a headspace vial to be used for sample derivatization. Sample preparation Beer samples were stored in a refrigerator at 4 to 6 °C prior to analysis. A 250 mL portion of the beer was poured into a clean plastic bottle and capped. The sample was degassed by hand shaking the capped bottle five times followed by opening the cap to release the carbon dioxide (CO2  ). The degassing of the samples was performed a total of three times. A 2 mL degassed beer sample was transferred to a 20 mL headspace vial, which was immediately capped for analysis. 134 3 PAL3 RTC and GC/MSD parameters The analysis parameters of the PAL3 RTC sampler are shown in Table 1. Table 1. Agilent PAL3 autosampler and GC/MSD parameters for beer analysis. Agilent PAL3 (SPME) Fiber Type Agilent 65 µm PDMS/DVB (p/n 5610-5873) Fiber Conditioning Temperature 250 °C Preconditioning Time 5 min Incubation Time 20 min Incubation Temperature 60 °C Derivatization Time 10 min Agitator Speed 250 rpm Sample Extraction Time 30 min Sample Desorption Time 1 min Post Conditioning Time 5 min Gas Chromatograph Model Agilent 8890 GC GC Column Agilent J&W DB-5ms UI, 30 m × 0.25 mm, 0.25 µm (p/n 122-5532UI) Column Pneumatics Constant flow Carrier Gas Helium Injector Mode Splitless Purge Flow to Split Vent 50 mL/min at 2 min Inlet Temperature 250 °C Injector Liner Agilent Ultra Inert splitless liner (p/n 5190-4047) Flow Rate 1.2 mL/min Oven Temperature Program 60 °C for 2 min 10 °C/min to 140 °C 7 °C/min to 250 °C, hold 3.0 min Equilibration Time 3 min Mass Spectrometer Model Agilent 5977C GC/MSD Ionization Mode EI, 70eV Acquisition Mode Scan Scan Speed N = 2 Scan Range 50 to 520 amu GC Transfer Line Temperature 250 °C Ion Source Temperature 230 °C Quad Temperature 150 °C Results and discussion Compound identification and retention time confirmation The 10 μg/L calibration standard was analyzed in full scan data acquisition mode, and the total ion chromatogram (TIC) is shown in Figure 1. Figure 1. TIC of 10 µg/L aldehydes with Agilent SPME on-fiber PFBHA derivatization. The data file for the 10 µg/L sample was processed using MassHunter Unknowns Analysis software. Automatic deconvolution of the data was performed using Unknowns Analysis to identify the components present exclusively in the sample. From the resulting list of components, the four target aldehydes were identified through library matching against the NIST 23 spectral library, achieving match scores above 80. The retention times (RTs) of the four aldehyde derivatives were identified as 13.299, 13.830, 17.419, and 18.914 minutes, respectively (Figures 2 to 5). 135 4 Figure 2. Hexanal derivative identified with an RT of 13.299 minutes. Figure 3. Furfural derivative identified with an RT of 13.830 minutes. 136 5 Figure 4. Phenylacetaldehyde derivative identified with an RT of 17.419 minutes. Figure 5. trans-2-Nonenal derivative identified with an RT of 18.914 minutes. 137 6 Calibration curves Based on the response of the calibration standard solutions, the calibration curves were plotted for the four aldehyde derivatives. The results are shown in Table 2 and Figures 6 to 9. Table 2. The calibration range and R2 for the four aldehydes. No. Compound Name Calibration Range (µg/L) R2 1 Hexanal 0.05 to 10 0.999 2 Furfural 5 to 1,000 0.998 3 Phenylacetaldehyde 0.1 to 50 0.996 4 trans-2-Nonenal 0.025 to 5 0.998 Figure 6. Calibration curve for hexanal 0.05 to 10 μg/L. Hexanal 0 1 2 3 4 5 6 7 8 9 10 Responses 0 0.2 0.4 0.6 0.8 y = 1,019,452.066254x + 31,382.352344 R2 = 0.9992 R = 0.9997 ×107 Concentration (µg/L) Figure 7. Calibration curve for furfural 5 to 1,000 μg/L. Responses 0 0.2 0.4 0.6 0.8 ×107 Concentration (µg/L) Furfural 0 100 200 300 400 500 600 700 800 900 1,000 y = 10,983.666729x – 17,133.067454 R2 = 0.9980 R = 0.9989 Figure 8. Calibration curve for phenylacetaldehyde 0.1 to 50 μg/L. Responses ×106 Concentration (µg/L) Phenylacetaldehyde 0 5 10 15 20 25 30 35 40 45 50 0 1 2 3 4 5 6 y = 140,635.827993x – 5,702.613289 R2 = 0.9963 R = 0.99970 Figure 9. Calibration curve for trans-2-nonenal 0.025 to 5 μg/L. Responses ×106 Concentration (µg/L) 0 trans-2-Nonenal 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0.5 1.0 1.5 y = 22,317.224704x2 + 233,186.59012 x – 5,014.216502 R2 = 0.9980 R = 0.9979 Type: Quadratic, Origin:Ignore, Weight:1/x 138 7 Quantification results of beer samples Based on the calibration curve setup, quantification was performed for the four aldehydes in the four brands of beer samples that were tested. The quantification results are summarized in Tables 3 to 6. All the aldehydes in the four beer samples were below their respective flavor thresholds. Among the four brands of beer samples that were tested, Brand 4 contained the lowest levels of aldehydes, with 0.45 μg/L hexanal, 6.26 μg/L furfural, 6.64 μg/L phenylacetaldehyde, and 0.031 μg/L trans-2-nonenal. Table 3. Quantification results of hexanal in four brands of beer samples. No. Beer Hexanal Concentration (µg/L) Average Concentration (µg/L) Concentration %RSD (n = 3) 1 Brand 1 0.67 0.67 0.69 0.68 1.7 2 Brand 2 1.53 1.54 1.61 1.56 2.8 3 Brand 3 0.99 1.03 1.04 1.02 2.6 4 Brand 4 0.45 0.46 0.45 0.45 1.3 Table 4. Quantification results of furfural in four brands of beer samples. No. Beer Furfural Concentration (µg/L) Average Concentration (µg/L) Concentration %RSD (n = 3) 1 Brand 1 18.45 18.75 18.45 18.55 0.9 2 Brand 2 49.46 49.39 52.27 50.37 3.3 3 Brand 3 24.81 26.53 25.04 25.46 3.7 4 Brand 4 6.10 6.37 6.30 6.26 2.2 Table 5. Quantification results of phenylacetaldehyde in four brands of beer samples. No. Beer Phenylacetaldehyde Concentration (µg/L) Average Concentration (µg/L) Concentration %RSD (n = 3) 1 Brand 1 11.07 10.41 10.05 10.51 4.9 2 Brand 2 8.36 8.26 8.07 8.23 1.8 3 Brand 3 8.72 8.63 8.92 8.76 1.7 4 Brand 4 6.84 6.59 6.48 6.64 2.8 Table 6. Quantification results of trans-2-nonenal in four brands of beer samples. No. Beer Trans-2-Nonenal Concentration (µg/L) Average Concentration (µg/L) Concentration %RSD (n = 3) 1 Brand 1 0.034 0.035 0.034 0.034 1.7 2 Brand 2 0.061 0.063 0.063 0.062 2.8 3 Brand 3 0.034 0.034 0.035 0.034 1.7 4 Brand 4 0.030 0.031 0.031 0.031 1.3 139 8 With three replicate injections for each beer sample, the concentration %RSDs were calculated to be below 4.9%. The TIC overlays of the three replicate injections are shown in Figures 10 to 13. Figure 10. TIC overlay of the four aldehydes in the Brand 1 beer sample. 1 2 3 4 5 ×105 13.2 13.3 13.4 ×105 13.7 13.8 13.9 14.0 0.2 0.4 0.6 0.8 1.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Hexanal ×106 17.3 17.4 17.5 Phenylacetaldehyde Furfural ×103 18.8 18.9 19.0 19.1 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 trans-2-Nonenal Figure 11. TIC overlay of the four aldehydes in the Brand 2 beer sample. ×106 13.2 13.3 13.4 ×105 13.7 13.8 13.9 14.0 0.2 0.4 0.6 0.8 1.0 1.2 Hexanal ×105 17.3 17.4 17.5 Phenylacetaldehyde Furfural ×103 18.8 18.9 19.0 19.1 trans-2-Nonenal 0.5 1.0 1.5 2.0 2.5 3.0 3.5 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 1 2 3 4 5 6 7 8 140 9 Figure 12. TIC overlay of the four aldehydes in the Brand 3 beer sample. ×105 13.2 13.3 13.4 ×105 13.7 13.8 13.9 14.0 Hexanal ×106 17.3 17.4 17.5 Phenylacetaldehyde Furfural ×103 18.8 18.9 19.0 19.1 trans-2-Nonenal 1 2 3 4 5 6 7 0.2 0.4 0.6 0.8 1.0 1.2 1.4 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Figure 13. TIC overlay of the four aldehydes in the Brand 4 beer sample. ×105 13.2 13.3 13.4 ×105 13.7 13.8 13.9 14.0 Hexanal ×105 17.3 17.4 17.5 17.6 Phenylacetaldehyde Furfural ×103 18.8 18.9 19.0 19.1 trans-2-Nonenal 0.2 0.4 0.6 0.8 1.0 1.2 0.2 0.4 0.6 0.8 1.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 1 2 3 4 5 6 7 141 www.agilent.com/chem/5977c DE58984949 This information is subject to change without notice. © Agilent Technologies, Inc. 2024 Printed in the USA, July 18, 2024 5994-7633EN Determination of detection limits Using the lowest calibration concentration of each compound, the signal-to-noise ratio (S/N) in full scan mode was calculated. The limit of quantification (LOQ) was determined at a S/N of 10, and the limit of detection (LOD) was determined at a S/N of 3. A summary of the LOQ and LOD results is presented in Table 7. Table 7. LOQ and LOD of the aldehydes. Aldehyde S/N LOQ (µg/L) LOD (µg/L) Hexanal (0.05 µg/L) 161 0.003 0.0009 Furfural (5 µg/L) 29 1.72 0.52 Phenylacetaldehyde (0.1 µg/L) 20 0.05 0.015 trans-2-Nonenal (0.025 µg/L) 24 0.01 0.003 Conclusion This application note describes the quantitative analysis of four aldehydes (hexanal, furfural, phenylacetaldehyde, and trans-2-nonenal) responsible for off-flavors in beer using an Agilent 8890/5977C GC/MSD with a PAL3 (SPME) autosampler. This method offers the advantages of full automation, rapid analysis, solvent-free extraction, and on-fiber derivatization. With this automated solution, excellent sensitivity was demonstrated for the detection of hexanal (0.0009 μg/L), furfural (0.52 μg/L), phenylacetaldehyde (0.015 μg/L), and trans-2-nonenal (0.003 μg/L). Four different brands of beer were analyzed, with hexanal detected in the range of 0.45 to 1.56 μg/L, furfural in the range of 6.62 to 50.37 μg/L, phenylacetaldehyde in the range of 6.64 to 10.51 μg/L, and trans-2-nonenal in the range of 0.031 to 0.062 μg/L. Good repeatability was demonstrated with RSD < 4.9% based on three replicate injections of the four beer samples for all four aldehydes. References 1. Aguiar, D.; et al. Assessment of Staling Aldehydes in Lager Beer under Maritime Transport and Storage Conditions. Molecules 2022, 27(3), 600. 2. Moreira, M. T. G.; et al. Aldehyde Accumulation in Aged Alcoholic Beer: Addressing Acetaldehyde Impacts on Upper Aerodigestive Tract Cancer Risks. Int. J. Mol. Sci. 2022, 23(22), 14147. DE-002126 This information is subject to change without notice. © Agilent Technologies, Inc. 2024 Published in the USA, November 19, 2024 5994-7858EN Learn more: www.agilent.com/chem/5977c www.agilent.com/chem/7000e www.agilent.com/chem/7010 Buy online: www.agilent.com/chem/store Find a local Agilent customer center in your country: www.agilent.com/chem/contactus U.S. and Canada 1-800-227-9770 agilent_inquiries@agilent.com Europe info_agilent@agilent.com Asia Pacific inquiry_lsca@agilent.com