By now, most of you are probably familiar with U.S. EPA Method 5035A. For those who don't know, this method contains detailed instructions on how to collect soil samples for volatile organic compound (VOC) analysis. 5035A is heir to an earlier soil-sampling method, which was often called “jar packing” (scoop soil into a 4-oz wide mouth glass jar, and screw the lid on tightly). This method typically resulted in significant losses of VOCs, because by the time you got your sample to the lab, a significant amount of the original VOC concentration had either volatilized or biodegraded.
The “solution” to jar packing was EPA 5035, which was first published in December, 1996. In the first edition of the method, you were required to inset soil into a 40-mL VOA vial containing one of two preservatives: sodium bisulfate or methanol. In July of 2002, EPA 5035 was modified to “5035A.” The latest edition contains an appendix with 65 pages of additional information on how to collect samples. In Section A8.0 (p. 56) of 5035A, details on sample collection using an empty vial are provided.
In 2003, EPA published Contract Lab Protocol (CLP) guidelines for 5035A. The preferred or recommended option for 5035A is to use an empty VOA vial to collect the sample. The vial can then be frozen, followed by laboratory analysis (water purge).
Here's how this works:
- Extrude 5-g of soil into a 40-mL empty VOA vial with a magnetic stir bar. Screw-on the cap tightly.
- Ice the sample to 4°C. Transport the sample to the lab.
- Freeze the sample in the lab to at least -7° C (14-day holding time).
- Analyze the sample by adding deionized water, surrogates and internal standards and purge on a closed-loop purge and trap system (Archon or Tekmar).
According to EPA, the key advantage of this method is that once the vial is sealed, it is never opened again; hence, no volatilization. Furthermore, you don't have to fuss with adding preservatives (e.g., methanol), and it's easy for the lab to freeze arriving samples.
Yes, all of this sounds good. Yes, using an empty VOA vial and freezing is easy to do. No one disputes that. Yet, there is a glitch in this method and that's the point of this article.
So What's the Glitch?
When you extrude soil into an empty VOA vial, there will be significant losses of VOCs, even if you screw the cap on within a few seconds. Think about it—you are inserting soil into an empty container with no preservative. That's the glitch. The problem lies in the field and not in the lab. Yes, once you seal the container, it's never opened again; however, that doesn't remedy the first step that took place in the field. If you insert soil with VOCs into an empty container, there will be losses—plain and simple.
The irony of the current EPA CLP protocol for 5035A is that it appears we've come full circle and essentially returned to jar packing. Thus, if you use option #1 under the current EPA CLP protocol for 5035A, be advised that the first step in the field is a risky venture. Again, we understand and acknowledge that freezing mitigates volatilization. We understand and acknowledge that once you seal the vial, it remains sealed, even during the analysis (water purge). Yet, it is that first step in the field that remains the Achilles heel.
Put to the Test
In 2002, Sorini et al. performed a test to determine if extrusion of soil containing VOCs into an empty VOA vial can result in loss of VOCs from the soil. A series of tests was performed to compare VOC concentrations in spiked soil samples extruded into 40-mL empty VOA vials versus VOC concentrations in spiked soil samples extruded into 40-mL VOA vials containing 5 mL of methanol.
The testing, involved the use of 5-gram soil samples—49 percent sand, 26 percent silt, 24 percent clay, 5.3 percent organic material, and 14 percent moisture. Spiking was performed to give VOC concentrations of about 250 µ g/Kg to about 600 µg/Kg (high level) in the soil samples. The analytes of interest were vinyl chloride, MeCl 2 , MTBE, 1,1-dichloroethane, CDCE, chloroform, benzene, TCE, toluene, PCE, ethyl benzene, and o-xylene. Each sample was spiked with 100 µL of spiking solution that was prepared by adding a methanol solution containing the analytes of interest to gasoline-saturated water. All analyses were performed using EPA Methods 5035 (U.S. EPA 1996a) and 8260B (U.S. EPA 1996b). The following sets of samples were prepared and analyzed to evaluate the loss of VOCs due to extrusion into empty VOA vials:
A) three soil samples in 5-gram En Core samplers: spiked with 100 µL of spiking solution, extruded into VOA vials containing 5 mL of methanol, stored at 4 ± 2°C for 24 hours, analyzed
B) three soil samples in 5-gram En Core samplers: spiked with 100 µL of spiking solution and extruded into empty VOA vials; 5 mL methanol injected through septum of each vial using a 23-gauge needle, stored at 4 ± 2°C for 24 hours, analyzed
C) three 5-gram soil samples in VOA vials: soil surface spiked with 100 µL of spiking solution added through opening of the vial with immediate capping after spiking; 5 mL methanol injected through septum of each vial using a 23-gauge needle, stored at 4 ± 2°C for 24 hours, analyzed
D) 3 VOA vials each containing 5 mL of methanol: 100 µL of spiking solution injected through septum of each vial using a 23-gauge needle, stored at 4 ± 2°C for 24 hours, analyzed
The average concentration of each of the analytes of interest in each of the various soil samples (A, B, and C described above) was calculated and compared to the average concentration of the analyte in the spiked methanol solution (D) by calculating average percent recovery. These values are given in Table 1. As shown in Table 1, the average percent recovery values for the analytes of interest in the soil samples that were spiked directly into the VOA vials and in the soil samples that were spiked and extruded into VOA vials containing methanol are very similar, except for vinyl chloride, which is the most volatile of the analytes tested.
For the VOCs, other than vinyl chloride, these data show losses of approximately 10 percent for both techniques. For the soil spiked in the VOA vials, some loss can be expected prior to capping the vials; and for the samples extruded into methanol, some loss can be expected during the spiking and extrusion steps.
The average percent recovery values shown in Table 1 for the analytes of interest in the soil extruded into the empty VOA vials are much lower than for the other two sets of soil samples, showing a greater VOC loss. These data show that extrusion of soil samples containing high-level concentrations of VOCs into empty VOA vials can result in significantly greater VOC losses from the samples as compared to losses resulting from sample extrusion directly into methanol.
In similar studies, evaluating the use of an empty VOA vial for storage of samples during transportation to the laboratory, the spike was injected onto the soil in the VOA vial, rather than being added to the soil prior to extrusion into the empty VOA vial (Hewitt 1999 and Ricker 1999). As a result, VOC recoveries similar to those listed in the first column of Table 1 were reported, and VOC loss due to extrusion into the empty VOA vials was not observed.
In summary, the authors of this article feel very strongly that you should understand the limitations of the empty vial and freezing method prior to using it. We simply want to convey the scientific facts as we know them.
Is There a Lesson Here?
Deana Crumbling (2002) summed it up best in her paper on data quality. She explains that we have traditionally used analytical methods to define data quality. In other words, if the sample was analyzed within specified limits of precision, then the data must be “good.” Yet we now know that most of the problems with data quality are associated with what happens in the field and not in the lab.
Specifically, the lesson learned is this: How and where a sample gets collected is just as important as how it gets analyzed. In the case of 5035A, the empty vial method is designed to mitigate losses in the lab (close-loop extraction). Yet, what everyone seems to have forgotten is that how the sample gets collected (empty vial) is just as important as what happens in the lab.
So Where Do We Go from Here?
Hopefully EPA will re-evaluate the empty VOA vial option. In the interim, the authors of this article recommend that if you are collecting soil samples for VOC analysis, then you should use one of two techniques:
Once you collect the sample, immediately immerse it in 5-10 mL of laboratory grade methanol; or,
Use some type of hand-operated coring device designed to collect and store soil samples for VOC analysis (e.g. En Core â ) to collect the sample, ice to 4°C and ship to the lab within 48 hrs.
Charles San Juan is Senior Hydrogeologist with the Washington Department of Ecology. He can be reached at (360)407-7191, or firstname.lastname@example.org
Susan Sorini is Lead Scientist with Western Research Institute (WRI). She can be reached at (307)721-2232, or email@example.com.
Table 1. Average Percent Recoveries of VOC a Concentrations from 5-gram Soil b Samples Compared to VOC Concentrations in Spiked Methanol
Spike injected onto soil
| Spiked soil extruded
into MeOH in VOA
| Spiked soil extruded into
empty VOA vial
|83 c (17) d||63(6)||
a Spiked concentrations of the VOCs in the samples ranged from ~250 to ~600 F g/Kg.
b The soil is 49% sand, 26% silt, 24% clay, 5.3% organic material, and 14% moisture.
c Average percent recovery is based on mean concentration values determined for 3 of each of the soil samples and 3 spiked methanol samples.
d The value in parentheses is the percent relative standard deviation of the concentration values in the soil samples. The percent relative standard deviation of the concentration values in the spiked methanol samples ranged from 0.1% to 5%.
- Crumbling, Deana (2002) , In Search of Representativeness: Evolving the Environmental Data Quality Model. Quality Assurance, 9, pp. 179-190.
- Hewitt, A.D., 1999, Storage and Preservation of Soil Samples for Volatile Organic Compound Analysis. CRREL Special Report 99-5, U.S. Army Cold Regions Research and Engineering Laboratory, Hanover, NH.
- Ricker, M.J., 1999, An Easy, Cost-Effective Solution for Sampling Volatile Organic Compounds in Soils. Proceedings of the Fifteenth Annual Waste Testing & Quality Assurance Symposium (WTQA ‘99), Arlington, VA.
- Sorini, S.S., J.F. Schabron, and J.F. Rovani, Jr., 2002, Evaluation of VOC Loss from Soil Samples. Contaminated Soil Sediment & Water, April/May Issue, pp. 39-44.
- U.S. EPA, 1996a, Method 5035: Closed-System Purge-and-Trap and Extraction for Volatile Organics in Soil and Waste Samples. Test Methods for Evaluating Solid Waste: Physical/Chemical Methods (SW-846), Vol 1B.
- U.S. EPA, 1996b, Method 8260B: Volatile Organic Compounds by Gas Chromatography/Mass Spectrometry (GC/MS). Test Methods for Evaluating Solid Waste: Physical/Chemical Methods (SW-846), Vol. 1B.