Vapor intrusion (VI) of groundwater contaminants like trichloroethene (TCE) has become an issue of increasing concern over the past decade necessitating development of methods to appropriately evaluate it. Determination of TCE concentrations in indoor air can be a crucial part of VI assessments. The conventional and most commonly used EPA TO-15 method (gas chromatography/mass spectrometry [GC/MS]) is limited primarily by protracted turnaround times, multiple house visits, and per-sample cost.  Near-real-time on-site analysis can address these concerns and identify potential interfering indoor sources as well. A commercially available portable GC/MS provides a near-real time analysis, but has high capital costs, requires external carrier gas, and can have significant instrument downtime for costly maintenance. The overall project objective was to evaluate the efficacy of a micro-gas chromatograph (µGC) prototype for detection of low-level TCE concentrations in indoor-air vapor intrusion applications as a potential cost-effective alternative.

The µGC prototype, developed by the University of Michigan, consists of a conventional sampling front-end module and a novel micro-analysis module. The front-end sampling module concentrates VOCs in the TCE vapor pressure range. A microfabricated µfocuser injects the sample onto the separation µcolumns with independent temperature control; scrubbed air is used as the carrier gas. The µdetector consists of an array of 4 different chemiresistor sensors, thus providing compound-specific response patterns. Both modules are controlled by customized software. Laboratory µGC studies showed that TCE detection limits in the low- and sub-ppb range could be obtained and that the µGC was applicable to analysis of other VOCs.

A field demonstration was conducted in the vicinity of Hill Air Force Base (AFB), UT; primarily in a house with known TCE VI. Concurrent reference samples were analyzed by TO-15. Field calibration detection limits were similar to those in the laboratory. TCE levels were varied by creating a negative indoor air pressure relative to sub-slab. Comparison with concurrent TO-15 samples showed that the µGC prototype TCE accuracy was good above its TCE mitigation action level (2.3 ppb; at time of field demonstration), but less accurate below 1 ppb due to interfering VOCs. Long-term results showed that response stability was adequate and could be improved with µdetector temperature control. Temporal and spatial studies were conducted. Temporal TCE variations were effectively tracked by the µGC; including a 48-h unattended, automated run. Spatial studies showed concentration gradients indicating VI entry and an emplaced indoor TCE source. These studies illustrate the efficacy of the µGC prototype in real-world VI applications.

A primary implementation issue is that the µGC is not currently commercially available. Future work is needed to further reduce its size, improve ease of use, improve robustness, incorporate remote communications, and implement hardware and software refinements to improve accuracy. Using cost estimates, a commercial µGC for VI applications is anticipated to be more cost-effective than the traditional TO-15 approach.

This study stands as the first of its kind, where µGC instrumentation was shown capable of sustained, reliable, automated measurements of a trace-level component (TCE) in a complex VOC mixture under field conditions. µGC technology holds great promise for environmental monitoring problems (e.g., VI) where speciated low-concentration measurements are required.