An automated GC/MS system for the analysis of volatile and semi-volatile organic compounds in water

This paper describes a GC/MS system capable of performing Volatile Organic Analysis on liquids, solids, and air. When combined with a syringe auto-injector, the system is completely automated for both volatile and semi-volatile analyses. An OI Analytical Model 4551 Vial Multisampler and an OI Analytical DPM-16 Multisampler are interfaced and then connected to an OI Analytical Model 4560 Sample Concentrator, an HP Model 5971 MSD, an HP Model 7673 Auto-Injector, and an HP Model 5890 Series II GC to form a multi-tasking GC/MS system. This system is shown to allow greater versatility in the laboratory.


Introduction
For today's laboratory needs, multi-tasking equipment is increasingly necessary, particularly for gas chromatography/mass spectrometry (GC/MS) systems. Equipping a GC/MS system to perform a single analysis is costly but necessary when protocol requires a mass spectrometer detector (MSD) for positive analyte identification. Not only is a suitable GC/MS system needed (these have been available for some time), but also the ability to create a multi-tasking system. It is necessary to interface the required apparatus to perform a variety of analyses.
Until recently, USEPA Method 524.2 or 624 requirements have limited the GC/MS system by requiring that the GC injector interface to a purge-and-trap sample concentrator. This has become less of a limitation with the appearance of certain interfaces (such as the Low-Dead-Volume Injector) on the market, which allow more utility ofthe system. However, the weaknesses ofthese systems are related more to sample introduction than to GC. There is a need to analyse a variety of matrices (for example water, soil, extractables, air) by sample concentration.
Using the GC/MS system for more than volatiles-only analyses is also desirable (for example, for base neutral acids (BNA) analysis). This requires using a syringe autosampler, combined with a sample concentrator, creating several multisampling capabilities.
The GC/MS system described in this paper demonstrates such capabilities. It incorporates the use of a capillary column (interfaced directly to the MS), which separates a wide range of analytes and has a large temperature range. The inlet is a standard split-splitless (S/SL) injection port. More importantly, the laboratory can perform a wide range of volatile and semi-volatile organic analyses using this new multi-tasking system. Module (ECM) (OI Analytical) supplied a constant gas flow to the GC system. In all states, except the sample concentrator's 'DESORB' state, the carrier gas supply flowed from the mass flow controller through the six-port switching valve in the Model 4560 into the inlet. At this juncture, the gas flow was controlled by adjusting the column head pressure control to allow 1.0 ml/min (at 30C column temperature) of the total flow into the column. The remainder of the flow was split out of the injector (the injection purge valve was ON) and the septum purge vent was then turned OFF.
The advantage of this flow control scheme is that the total gas flow, supplied by the ECM, desorbs the trap when placed in line with the injector (see figure 3). The higher gas flow needed to efficiently desorb analytes from the trap is provided and the requirement for low gas flow to the MSD is also met. By adjusting the total gas flow, the range, or sensitivity, of the analysis can easily be adjusted.        figure 5. For this analysis, a 5 ml nickel sample loop was used. The sample is transferred under pressure from the sample vial to overfill the loop volume; the sample is then transferred from the loop through a Standard Injection Module (SIM). As the sample begins to pass through the SIM valve (see figure 6), the module rotates the valve to place a 10 ml gas-tight syringe filled with internal standard solution in line with the valve rotor's internal volume. The SIM advances the syringe plunger, filling the volume of the valve (nominal 10 gl); the rotor rotates back in line with the sample which then sweeps the internal standard solution to the sparge vessel. A second module can inject a surrogate standard solution (Spiker) based on the same valve logic.  This column was more than adequate for the separation of these analytes, even showing separation of all the xylene isomers. However, to obtain good dichlorodifluoromethane (Freon 12) resolution, a subambient oven temperature was used (see figure 8). An initial temperature of-80C was used in this analysis, although higher temperatures are used when Freon 12 is not a compound of interest. Table 5 shows that excellent retention time (Rt) stability was achieved over the entire range of analytes.
Results and discussion  Table 7 shows the SIM reproducibility during the three-week test period of this analysis with a mixture ot the internal standard and surrogates. This standard was added to every sample analysed at an injection interval of 1. Reproducibility is very good for the internal standard and the p-Bromofluorobenzene surrogate. The percentage RSD for the 1,2-Dichlorobenzene-d4 was 16. This high value was first perceived as a hardware problem until further investigation showed good reproducibility for the other surrogates, particularly for 1,4-Dichlorobenzene-d,.    Also, the relatively poor percentage RSD value was a result of low responses and varied with the presence or absence of the hydrogenated isomer, 1,2-Dichlorobenzene. A possible explanation is de-deuteration due to hydrogen exchange either in the MSD ion source, or during the gas phase of the analysis. This occurrence will be investigated further in future analyses. Table 8 shows the reproducibility of a SIM and a Spiker, a second SIM module that has been set to inject a standard solution at an alternating interval between 0-99 (interval 2 for this analysis). The Spiker option can be used to add surrogates to a sample or blank, or to hold a second volume of the internal standard solution. The Spiker option can then be sequenced by the Model 4560 microprocessor to 5 ppb Internal      With a 5 ml sample loop, it is possible to analyse multiple replicates for a 40 ml VOA vial. By adjusting the loop fill time to a minimum and adjusting the needle depth, all but the final ml of water can be sampled. Table 9 shows the reproducibility of two replicates taken across the calibration range.
The usefulness of any quantitative analysis automatic sampling system is limited by the amount of carryover Methylene chloride @ 1.4 gg/1 Naphthalene @ 1.2 gg/1 Tetrachloroethene @ 0.9 gg/1 all others < 0.6 gg/1 experienced when a high level sample is encountered. A multi-tasking GC/MS system is useful to both laboratories operating on a limited capital budget, and labora-