Jump to main content.

RARE - Development of a Systematic Approach to Accurately Measure Trace Levels of VOCs in Soils


A majority of the contaminated sites within the United States contain volatile organic compounds (VOCs), either as industrial chlorinated solvents or petroleum products commonly referred to as BTEX (benzene, toluene, ethyl benzene, and xylene). The present EPA methods for low level VOC analyses (RCRA Methods 5035A and the CLP method), commonly referred to as volatile organic analyses (VOA), are very matrix dependent and therefore can be biased low. The VOCs diffuse into the micropores or are adsorbed onto the soil matrix. This negative method bias can result in false negative situations where significant contamination may go undetected. Also, water soluble compounds are a major problem with the present methods because their low purging efficiencies result in high detection levels.

Risk assessment is a crucial component of the site remediation decision-making process. The present EPA low level methods are not capable of measuring many VOCs at the required levels for risk assessment, EPA Region 3 Risk Based Concentration (RBC) levels, EPA Region 9 Preliminary Remediation Goals (PRG) levels, or the state cleanup levels. Recently at a University of Massachusetts Remediation Conference, a paper was presented on the magnitude of the problem and concluded that resolving this issue is critical for human health and ecological risk assessments. It is critical that EPA develops a method to accurately measure trace VOCs for the CLP Superfund program, RCRA and Brownfields site redevelopments, and the various state programs in Region 1. Some of the sites within New England where risk assessments are problematic include: Woburn, MA, Wells G & H, Otis Air Force Base, and Nyanza, MA. Below is a VOC target list for this project with the compounds identified that have low RBC and PRG quantitation levels.


The objective of this project is to develop an analytical method which uses thermal desorption (200-300 degrees C) to extract and accurately measure VOCs in soil and sediment samples to satisfy the EPA and state regulatory program requirements. Thermal extraction will be examined because the technique is a simpler and more efficient extraction technique than the present EPA purge methods. VOCs shall be detected and quantified by a gas chromatograph/mass spectrometer (GC/MS).

Thermal extraction is also presently used for extracting semivolatile compounds in soil samples by RCRA Method 8275A. This fact encourages investigation of this technique for the quantification of both VOCs and semivolatile compounds. It is anticipated that the semivolatile compounds shall be analyzed by a second GC/MS. This will be the first ever dual GC/MS method.

Progress to Date:


This research project shall be conducted in several phases.

Phase 1. Method Development. The entire sample collection, preservation, and analysis process must be controlled to achieve accurate results. Initially, an appropriate sample collection vessel must be selected. The present VOA septa (Teflon-faced) and caps must be evaluated for their ability to withstand elevated temperatures (~150 degrees C) without degradation and/or artifact formation. Other chromatographic septa and cap materials may need to be tested should problems be encountered. This phase will be conducted by EPA and the results will be provided to the contractor for use in subsequent phases of the program.

          Thermal Extraction: Thermal extraction is the principle for the proposed method. It must be optimized to exhaustively extract VOCs from samples with a wide range of moisture contents while minimizing any thermal degradation of the target compounds. Fortunately, most VOCs are thermally stable as demonstrated by a thermal desorption step at 250 degrees C from an absorbent in the present methods. The Archon autosampler will be initially used as the thermal extraction device.

          A well-characterized soil starting material will be spiked to cover the complete boiling point range for VOCs and semivolatiles as identified in the attachment. Samples shall be prepared with minimal solvent (i.e., methanol) usage. If the minimal solvent usage results in too much GC/MS interference, a new set of spiked soil samples will be prepared following a different method. Samples will be stored at <7 degrees C prior to analysis.

          Analyses will be performed following a modified SW-846 Method 5035A purge and trap without the addition of water. The method will be further modified to have purging temperatures tested at 40, 60, 80, 100, 120, and 140 degrees C.

Phase 2. Interference Checking. Both moisture and organic carbon can interfere with the extraction of VOCs. Three well-characterized soils with varying ranges of organic carbon contents will be spiked using the same VOC standard used in Phase 1. Six sets of spiked soils with the following moisture contents: dry, 10, 20, 30, 40, and 50% w:w will be prepared. Analyses of one set from each soil will be performed at 40, 60, 80, 100, 120, and 140 degrees C purging temperatures.

          During the analysis, moisture interference will be removed using a combination of the VOCARB 3000 trap and the Tekmar Moisture Control Loop present in the Tekmar-Dohrmann 3100 Concentrator. If these techniques are insufficient to remove the moisture interference, other moisture removing techniques will be investigated.

Phase 3. Chromatographic Splitting. As discussed earlier, one of the objectives of this project is to extract VOCs and semivolatile compounds from the soil matrix. The semivolatile compounds must be separated and prevented from entering the GC/MS for VOC analysis. The split gas flow will be transferred to a second GC column that would enter another GC/MS to analyze the semivolatile fraction, creating the first ever dual GC/MS method.

          Initially, once the two GC/MS systems are "connected," spiked soil samples shall be prepared in a similar manner as Phase 1. It should be noted that these samples will be analyzed using the optimized procedure defined in Phase 2; hence, the samples will be analyzed at 1 (and not 6) temperatures. If the splitting technique is successful, then a second full suite of spiked soils samples (for result verification) shall be prepared following the same requirement as in Phase 2.

Phase 4. Method Validation. An initial demonstration of capability (IDC) and method detection limit (MDL) study will be done. Internal quality control limits will be developed by running 25 internal quality control (QC) samples (IS area counts, surrogate recoveries, laboratory and field blanks, LFB, MS, MSD) on various matrices, some having high moisture (>70%) content. Various external QC samples from CLP, NIST, and ERT will be analyzed. A study will be conducted using performance evaluation materials (PEMs) and the results compared to the proposed (i.e., newly developed) method and the EPA low level method (SW-846 Methods 5035A and 8260) for the determination of VOCs in soils. All samples will be prepared in the same VOA vials, caps, and septa as used during Phases 1 and 2.

Phase 5. Sample Preservation. Using the method developed and validated in the earlier phases of this study, sample preservation techniques will be evaluated. Spiked soil samples will be prepared using the same VOC standard used in Phase 2. Sufficient samples shall be prepared to test the sample preservation techniques of storage at <7 degrees C, with the addition of 0.7 g of tribasic sodium phosphate dodecahydrate (Na3PO4·12H20) to each 5 g sample.


Principal Investigators:

Brian Schumacher, Supervisory Physical Scientist
John Zimmerman, Resident Physical Scientist
Mike Hiatt, Chemist

Characterization and Monitoring Home Page
Environmental Sciences | Research & Development
National Exposure Research Laboratory
Send questions or comments to Steve Gardner
Email: gardner.steve@epa.gov

Local Navigation

Jump to main content.