96-RP128.03 EPA/600/A-96/071 Interpretation and Data Quality Evaluation of Hourly Measurement of Target VOCs by AutoGC/MS at the New Hendersonville, TN, Southern Oxidants Study Site, June 17-27, 1995 E. Hunter Daughtrey, Jr., Jeffrey R. Adams, Christopher R. Fortune, Keith G. Kronmiller, Karen D. Oliver ManTecb Environmental Technology, Inc. P.O. Box 12313 Research Triangle Park, NC 27709 William A. McClenny National Exposure Research Laboratory U.S. Environmental Protection Agency 79 Alexander Drive Research Triangle Park, NC 27711 ------- 96-RP128.03 INTRODUCTION The North American Research Strategy for Tropospheric Ozone (NARSTO) program has sponsored the development and evaluation of monitoring methods for volatile organic compounds (VOCs) that are associated with the formation of tropospheric ozone. Target VOC lists have been developed that are designed to be reflective of the VOCs of concern in ozone formation. These lists are evolving as understanding of the processes develops.1*2 Our samples were analyzed for a target list that included 57 ozone precursor compounds, the 41-compound TO-14 list, and 10 selected aldehydes. Two field autoGC systems were deployed at the New Hendersonville, TN, site which was one of the intensive measurement sites for the 1995 Southern Oxidants Study (SOS). This site, approximately 15 mi northeast of Nashville, was selected to be a downwind site, although as data in a companion paper demonstrated,3 this was rarely the case for the June 17-27 period that we made measurements. In this paper, quality assurance issues related to the calibration, sampling integrity, and artifacts are discussed, and quantitative measures of data quality are described. Canister and sorbent tube samples were also taken concurrently with some of the hourly measurements, and the concurrent measurements are compared in this paper. EXPERIMENTAL Two autoGC/MS systems for monitoring VOCs in air were field tested. Both systems were housed in an environmentally controlled trailer and required no cryogenic liquids for sample collection or chromatographic separation. Trailer The outside dimensions of the mobile laboratory are 8.5 ft wide by 25.5 ft long, including the tongue, and the inside dimensions are 8 ft by 20 ft. The laboratory is furnished with two instrument benches, which measure 10 and 11 ft long, two gas bays that will accommodate a total of 8 cylinders, a computer desk, a storage closet, a small refrigerator, and storage shelves and drawers. Two heating/air conditioning units (20 A, 120 V) and two ceiling ventilation fans are available for temperature control. The laboratory is equipped with two 50-A 240-V electrical boxes and a total of 64 outlets. A telescoping meteorological tower which came with the trailer malfunctioned and was replaced with a demountable tower. Air was taken from the 30-m-high sampling manifold (erected by Georgia Institute of Technology) by heated transfer line into a distribution manifold that was inside the monitoring trailer. XonTech/Saturn System One autoGC system includes a prototype sample preconcentrator designed by XonTech, Inc. (Van Nuys, CA), that is based on the Stirling refrigeration cycle. This preconcentrator is interfaced to a Varian Saturn II GC/Ion Trap mass spectrometer (Walnut Creek, CA), The preconcentration system consists of two multisorbent traps (Tenax-GR/Carbotrap B/Carbosieve S-HJ) for sample collection at ambient temperatures. Once the air sample has been concentrated on the trap, the sample is desorbed onto a second (unpacked) trap which has been cooled to -165 °C by the Stirling closed- cycle cooler. The sample is effectively focused on this second trap so that upon desorption at 100 °C to the GC column, the analytes are well separated without using subambient GC oven temperature programming. After separation on the Varian Star 3400 GC column, the analytes are detected with the Saturn ion trap detector. The system has been evaluated with the TO-14 2 ------- 96-RP128.03 compound mixture, a polar VOC mixture, and the ozone precursor mixture. Excellent results have been achieved with regard to trapping efficiency and linearity of response.4 Perkin Elmer ATD 400/Q-Mass 910 System The second benchtop GC/MS system to be evaluated was made available to EPA by the Perkin Elmer corporation (Norwalk, CT) and it includes an ATD 400 Auto Thermal Desorption system, an Autosystem gas chromatograph, and a Q-Mass 910 mass spectrometer. The ATD 400 is the sample concentration and desorption device for the system, and it incorporates a multisorbent packed cold trap that is cooled by a Peltier cooler. The ATD 400 is connected to the Autosystem GC through a heated transfer line that is connected within the GC oven to a high-capacity, fused-silica capillary column (50 m x 0.32 mm x 5.0 #tm dimethyl polysiloxane). The GC column is connected to the Q-Mass 910 quadrupole mass selective detector through a direct interface. The overall design of this system makes it feasible to collect and analyze VOCs without the requirement for a cryogen, such as liquid nitrogen. This system was evaluated in two ways to determine its performance in VOC sample analysis using a sorbent tube sampling method. 1. Perkin Elmer sorbent tube collection—sorbent tube used to collect sample from the manifold or directly in an outdoor setting is desorbed on ATD 400 by using the cold trap for concentration 2. Summa-polished canister collection—sample transferred from canister onto a sorbent tube for subsequent analysis using ATD 400 cold trap for concentration RESULTS AND DISCUSSION Quality Assurance Project Han Before we went to the field, a Quality Assurance Project Plan was written and approved by appropriate ManTech and EPA QA and management personnel. In the plan, we set out our objectives for the study, schedules, responsibilities of each staff member, and our overall study design, meeting all criteria for a Class HI Project.5 The approximate overall goal for each quantitative QA objective is given in the following tabulation. Specific target objectives were set for each compound, based on the previous performance history of each compound on each analytical system. XonTech ATD 400/Q-Mass QA Objectives Saturn Manifold Canisters Sorbent Tube Method Detection Limit 0.1 ppbv 0.5 ppbv 0.5 ppbv 0.5 ppbv Relative Percent Difference 10% 25% 25% 25% Internal Audit Accuracy 20% 30% 30% N/A Completeness 75% 75% 75% 75% Target Compounds The target lists of compounds for which we calibrated the GC/MS systems are given in Table 1. We made a concerted effort to coordinate with other NARSTO awl SOS researchers to seek common target lists as much as we could. Although we understood that the standard target list of the Photochemical Assessment Monitoring Station (PAMS) program was undergoing reevaluation, we used the Alphagaz Ozone Precursor 57-compound standard, as we had it available and all systems had been characterized with this standard. We calibrated for the TO-14 target list because of our long history of using this standard and because we could use it as a benchmark for quick field evaluation of system performance. Because of the expressed interest in oxygenated compounds, we 3 ------- 96-RP128.03 added an aldehyde/ketone standard to our target list. It was prepared by injecting 100 fxL of dilute aqueous solution of selected aldehydes and ketones into an evacuated canister and was used as a semiquantitative calibration standard. Storage stability was only marginally evaluated because of the press of time in preparing for the study. The limited evaluation showed that the very volatile and the heavier compounds both demonstrated low recoveries. Further work is required to develop methods for the preparation of adequate quantitative standards for these compounds. Audit Standard Analysis We conducted a blind analysis on both GC/MS systems. The audit canister was shipped to the field site by the NERL Quality Assurance Branch. The results of these analyses are given in Table 2. The QA laboratory analyzed the canister, using a Nutech preconcentrator with Nafion dryer. The canister had been filled from a high-pressure cylinder containing one of the series of audit standards used in the PAMS program. As a measure of precision, we computed the percent relative difference between replicate analyses on each system. Excellent internal precision was found for all measurements. After all results were validated and reported, the field analysis results were compared to the QA lab results. Both the XonTech Saturn and Perkin Elmer results were within acceptable limits and better than the QA lab analysis results, except for the instance where the Perkin Elmer missed the identification of 2,3-dimethylbutane. GC Retention Time Stability- Retention time stability is a critical issue in measuring ozone precursor compounds, as similarity in mass spectra of the prevalent hydrocarbon compounds may cause inaccurate identification of the VOCs. Operating two heat-generating GC/MS systems in a trailer in the summer raises the question of GC retention time drift with variations in the trailer temperature. We suffered periodic air conditioner freeze-up during our 10-day study. This resulted in only one delayed start of the GC, as the trailer ambient temperature was near the GC start temperature. We were particularly concerned about retention time stability in the operation of our system, because we employ absolute rather than relative retention times in our identification. Despite these worries, the retention time stability of both systems was excellent, as illustrated in Figure 1. Data Capture Hourly measurements were taken continuously from June 17 through 27, with the only time lost for daily calibration checks and two power interruptions during the course of the study; samples were taken for 191 hours out of a scheduled 200. This represents 95% data capture for the XonTech/Saturn system. Hie Perkin Elmer system was not operated in a continuous mode, so calculation of completeness does not easily apply. The Perkin Elmer did demonstrate a greater sensitivity to humidity as samples taken during rain events and in the early morning showed depressed response because of the condensation of liquid water in the sampling tube and the previously demonstrated limited pumping capacity of the Q-Mass mass spectrometer. Potential Manifold and Transfer Line Artifacts Prior to the study, concern was raised by study participants about the length of the sampling manifold (30 m) and the successful transfer from the manifold into our sampling van. This concern was raised because of the anticipated "stickiness" of some of the heavier (Cg-CJ0) aldehydes. A cross-linked FEP/PTFE Teflon-coated manifold was especially constructed by University Research Glass (Carrboro, NC) for the Georgia Tech researchers. We procured a 40-ft heated PFA Teflon transfer line (Unique Products, Inc., Hazel Park, MI) and tested it prior to installation. The line 4 ------- 96-RP128.03 was purged with humidified scientific-grade air and the output analyzed; do artifacts were seen above background. We also challenged the line with standard gas from our calibration manifold; no compounds showed diminished response from passing through the transfer line. Additional testing for artifacts has continued since June. CONCLUSIONS Trailer-based near-real-time measurements of VOCs related to ozone formation can be made with high and measurable data quality. Reasonable accuracy and excellent precision were demonstrated. Greater than 95% data capture was achieved for the 10-day study. ACKNOWLEDGEMENTS Although the research described in this paper has been funded wholly or in part by the United States Environmental Protection Agency through Contracts 68-DCW3106 and 68-D5-0049 to ManTech Environmental Technology, Inc., it has not been subjected to Agency review and therefore does not necessarily reflect the views of the Agency, and no official endorsement should be inferred. Mention of trade names or commercial products does not constitute endorsement or recommendation for use. REFERENCES 1. E.C. Apel, J.G. Calvert, R. Zika, M.O. Rodgers, V.P. Aneja, J.F. Meagher, and W.A. Lonneman, "Hydrocarbon measurements during the 1992 Southern Oxidants Study Atlanta Intensive: Protocol and quality assurance," J. Air Waste Manage. Assoc.. 45(7): 521 (1995). 2. A. Bernardo-Bricker, C. Farmer, P. Milne, D. Riemer, R. Zika, and C. Stoneking, "Validation of speciated nonmethane hydrocarbon compound data collected during the 1992 Atlanta Intensive as part of the Southern Oxidants Study (SOS), "J. Air Waste Manage. Assoc.. 45(8): 591 (1995). 3. W.A. McClenny, E.H. Daughtrey, Jr., K.G, Kronmiller, J.R. Adams, and K.D. Oliver, "Hourly measurement of VOCs by AutoGC at the New Hendersonville, TN, Southern Oxidants Study site, June 17-27, 1995," Paper A1004, 1996 A&WMA Annual Meeting, Nashville, TN. 4. K.D. Oliver, J.R. Adams, E.H. Daughtrey, Jr., W.A. McClenny, M.J. Yoong, M.A. Pardee, E.B. Almasi, and N.A. Kirshen, "Technique for monitoring toxic VOCs in air: Sorbent preconcentration, closed cycle cooler ciyofocusing, and GC-MS analysis," Environ. Sci. Technol.. in press. 5. U.S. Environmental Protection Agency, Preparing Perfect Project Plans: A Pocket Guide for the Preparation of Quality Assurance Project Plans. EPA/600/9-89/087, Available from Guy F. Simes, Quality Assurance Manager, U.S. Environmental Protection Agency, Risk Reduction Engineering Laboratory, Cincinnati, OH, 1989. 5 ------- 96-RP128.03 D.A. Brymer, L.D. Ogle, C.J. Jones, and D.L. Lewis, "Viability of using Summa polished canister for the collection and storage of parts per billion by volume level volatile organics, Environ. Sci. Technol.. 30(1): 188 (1996). ------- Table 1. Target Compounds for Nashville 1995 Summer Study. Ozone Precursors (Alphagaz Standard) Isobutane I -Butene n-Butane frans-2-Butene rij-2-Butene 3-Methyl-l-butenc Isopentane 1 -Pentene n-Pentane Isoprene mrnj-2-Pentene ris-2-Pentene 2-Methyl-2-butene 2.2-Dimethylbutane Cyclopcntenc 4-Methyl-l-pentene 2.3-DimethyIbutane Cyclopentane 2-Methylpentane 3-Methylpentane 2-Me(hyI-1 -pentene n-Hexane franj-2-Hexene a'j-2-Hcxcne 2-4-Dimethylpentant Methylcyclopentane Benzene Cyclohexane 2-Methylhexane 2,3-Dimethylpentane 3-Methylhexane 2,2,4-Trimethylpentane n-Heptane Methylcyclohexane 2,3,4-T rimethylpentane Toluene 2-Methylheptane 3-Methylheptane n-Octane Ethylbenzene m,p-Xyfene Styrene ©-Xylene n-Nonane Isopropylbenzene cr-Pinene n-Propylbenzcne 1,3,5-T rimethylbenzene n-Decane /3-Pinene 1,2,4-Trimethylbenzene Undecane TO-14 Standard (Alphagaz) D ich lorod i fl uoromel h ane Chloromelhane 1,2-Dichloro-1,1,2,2-tetrafluoroethane Chloroethene Bromomethane Chloroelhane Trichlorofluoromethane 1,1-Dichloroethene Dichloromethane 3-ChIoropropene 1,1,2-Trichloro-1,2,2-trifluoroethane 1.1-Dichloroe»hane as-1,2-DichIoroethene Trichloromethane 1.2-Dichloroethane 1,1,1 -Trichloroethane Benzene Carbon tetrachloride 1,2-Dichloropropane Trichloroethene cis-1,3-DichIoropropene trans-1,3-Dichloropropene 1,1,2-Trichloroethane Toluene 1,2-Dibromoethane Tetrachloroethene Chlorobenzene Ethylbenzene m,p-Xylenc Styrene 1,1,2,2-Tetrachloroethane o-Xylene 4-Eihyltoluene 1,3,5-Trimethylbenzene 1,2,4-Trimethylbenzene Benzyl Chloride m-Dichlorobenzene p-Dichlorobenzene o-Dichlorobenzene 1,2,4-Trichlorobenzene Hexachlorobutadiene "Aldehyde Standard" Methacrolein Methyl vinyl ketone Butanal 1,1,1 -Trichloroethane 2-Pentanone Pentana! Trichloroethene Hexanal Tetrachloroethene Heptanal Benzaldehyde Octanal Nonanal Decanal 8 »—¦ 00 © U1 ------- Table 2. Precision and Accuracy of Analysis of Field Audit Canister 01628.* Analysis Results Replicate Average Compound QA Standard Target Lab Found X/S Found PE Found 1-Butene 8.90 10.50 9.34 5.01 rranj-2-Butene 8.75 10.30 9.94 11.43 3-Methyl-l-butene 11.00 13.10 9.10 10.38 1-Pentene 10.25 11.50 10.75 11.00 Isoprene 9.55 12.00 8.55 9.84 cw-2-Pentenc 9.80 11.50 10,28 12.78 2,2-Dimethylbutane 14.45 17.90 14.85 17.44 4-Methyl-1 -pentene 13.05 14.90 13.77 15.63 2,3-Dimethylbutane 13.85 16.20 14.82 0.00 3-Methylpentane 14.05 16.60 14.94 15.46 it-Hexane 13.85 15.50 14.94 15.91 cw-2-Hexene 13.10 14.70 14.13 15.40 2,4-Dimethylpentane 16.75 18.20 16.52 17.94 Cyclohexane 14.45 16.70 15.96 14.14 2,3-Dimethylpentane 17.40 20.30 18.34 20.05 2,2,4-T rimethylpentane 20.30 23.50 20.84 22.69 Methylcyclohexane 18.80 20.20 19.01 20.35 2,3,4-Trimethylpentane 18.80 21.50 20.12 20.13 2-Methylheptane 19.95 22.80 21.68 23.57 Ethylbenzene 18.65 17.80 17.36 17.34 m.p-Xylene 18.70 17.80 17.60 16.75 o-Xylene 18.15 18.50 18.16 17,97 Isopropylbenzene 19.95 20.50 20.16 20.27 1,3,5-T rimethy lbenzene 18.75 19.20 20.30 18.34 •Abbreviations: X/S * XonTech Saturn, PE = Perkin Elmer, ~~Precision = (a - b) x 2/(a + b) Precision (%)** Accuracy (%)*** X/S PE QA Lab X/S PE 2.68 -41.79 117 104 56 1.11 -1.50 117 113 130 -0.55 -3.32 119 82 94 0.70 -6.00 112 104 107 1.75 -2.59 125 89 103 -0.12 2.03 117 104 130 0.10 -2.16 123 102 120 -1.20 1.77 114 105 119 1.82 ERR 116 107 0 0.00 -1.19 118 106 110 0.60 1.19 111 107 114 0.74 -2.24 112 107 117 0.64 0.47 108 98 107 1.32 0.62 115 110 97 0.76 0.33 116 105 115 0.10 -0.46 115 102 111 -0.09 2.17 107 101 108 1.89 -0.93 114 107 107 1.29 -0.86 114 108 118 2.30 0.27 95 93 92 3.64 0.53 95 94 89 1.98 -0.06 101 100 99 1.56 0.55 102 101 101 2.99 0.09 102 108 97 ~~~Accuracy = Found x 100%/"True\ ------- 96-RP128.03 Oi tlier due to late: start 30.9 31 31.1 Decanal Retention Time, min 31.2 Figure 1. Typical Retention Time Stability for 10-Day Study. 9 ------- J TECHNICAL REPORT DATA 1. REPORT NO. EPA/600/A-96/071 2. 3.REC3 4. TITLE WTO SUBTITLE Interpretation of Data Quality Evaluation of Hourly- Measurement of Target VOCs by AutoGC/MS at the New Henedersonville, TN Southern Oxidants Study Site S. REPORT DATE 6. PERFORMING ORGANIZATION CODE 7. AUTHOR(S) W.A. McClenny; USEPA & E,H. Daughtrey, Jr., J.R. Adams C.R. Fortune, K.G. Kronmiller, and K.D. Oliver; Mantech 8. PERFORMING ORGANIZATION REPORT NO. 9. PERFORMING ORGANIZATION NAME AND ADDRESS Mantech Environmental Research Triangle Park, NC 10. PROGRAM ELEMENT NO. 11. CONTRACT/GRANT NO. 68-DO- 01OS 12. SPONSORING AGENCY NAME AND ADDRESS US Environmental Protection Agency Research Triangle Park, NC 27711 13.TYPE OF REPORT AND PERIOD COVERED publication of proceedings 14 . SPONSORING AGENCY CODS EPA/600/09 15. SUPPLEMENTARY NOTES 16. ABSTRACT The North American Research Strategy for Tropospheric Ozone (NARSTO) program has sponsored the development and evaluation of monitoring methods for volatile organic compounds (VOCs) that are associated with the formation of tropospheric ozone. Target VOC lists have been developed that are designed to be reflective of the VOCs of concern in ozone formation. These lists are evolving as understanding of the processes develop. Our samples were analyzed for a target list that included the 57 ozone precursor compounds, the 41- compound TO-14 list, and 10 selected aldehydes. Two field- deployed autoGC systems were deployed at the New Hendersonville, TN, site which was one of the intensive measurement sites of the 1995 Southern Oxidants study. This site approximately 15 mi northeast of Nashville, was selected to be a downwind site, although, ae data in a companion paper demonstrated, this was rarely the case for the June 17-27 period that we made measurements. In this paper, quality assurance issues related to the calibration, sampling integrity and artifacts are discussed, and quantitative measures of data quality are described. Canister and sorbent tube samples were also taken concurrently with some of the hourly measurements, and the concurrent measurements are compared in this paper. 17. KEY WORDS AND DOCUMENT ANALYSIS a. DESCRIPTORS b.IDENTIFIERS/ OPEN ENDED TERMS c.COSATI 18. DISTRIBUTION STATEMENT RELEASE 19 PUBLIC 19. SECURITY CLASS (This Report) UNCLASSIFIED 21.NO. OF PAGES 20. SECURITY CLASS (This Page) UNCLASSIFIED 22. PRICE ------- |