Volatile and Semivolatile Organics by Gas Chromatography in the Presence of Surfactant:

Static Headspace and Capillary Column Technique

1.0 Scope and Application

The following is tile analytical method to be used for liquid samples (from Hill AFB) shipped in 12 ml vials in the presence of surfactants. The GC/Static Headspace (HS) technique has proven to be comparable to the widely used GC/Purge and Trap method (using "dynamic" purge and trap technique) for the analysis of VOC's in soil and water (see attachment 1).

2.0 Summary of Method

Upon receipt of samples, the vials are stored at 4⁰C until analysis. Since GC analysis for organics depends upon the surfactant concentration, the surfactant concentration in the samples will first be quantified by either HPLC or L/L extraction. This data will be used for adjusting all solutions to the same final surfactant concentration. A 2 ml aliquot of the adjusted solution is then transferred to a 22 ml sample vial for GC/HS analysis. The results will be calculated relative to either an external (preliminary study) or internal standard.

3.0 Interferences

Raw GC data from all blanks, samples, and spikes must be evaluated for interferences. If possible these interferences must be eliminated. Common interferences for this method are listed below.

1) Contamination by carryover can occur, similar to GC/MS analysis, as documented in the Phase I Work Plan (Appendix H-4). In order to minimize carryover, the pipets used for transferring tile solution will be rinsed with non-polar (acetone) and polar (ultrapure water) solutions before and after each transfer.

2) Other interferences could occur due to high surfactant background concentrations in the samples (e.g., impurities such as ethylene oxide or alcohol in commercial surfactants). Background interferences from the surfactant solution which may create overlapping peaks should be carefully evaluated and substracted. Whenever an unusual sample is encountered, it should be followed by the analysis of pure surfactant to check for possible contamination.

3) The surfactant type and concentration will also affect the sweeping efficiency of the HS system. For example, higher surfactant concentrations will tend to reduce the sweeping efficiency due to strong organic-micelle interactions. Therefore, the calibration for organics should specify the type and concentration of surfactant used. In general, the samples should be adjusted to the same surfactant concentration before GC analysis to reduce errors.

• If any excess phase exists (e.g., NAPL, soil), the detection precision will likely

decrease. If an excess phase is visually observed, the sample should be filtered with a 0.2 mm Acrodisc filter (Gelman cat # LC13 PVDF) before analysis.

4.0 Apparatus and Materials

Gas Chromatograph- Shimadzu 17A with split/splitless injection

Automatic Sampler- Tekmar 7000 Headspace Autosampler and Tekmar 7050 Carrousel (for

controlling the sample sequencing of up to 50 vials).

Data Station- Varian Star Chromatography Workstation

Column- J&W Scientific DB-5, 30 meter, 0.53 mm I.D., 1.5 mm film thickness

Bottles- Glass with Teflon-lined crimp tops.

5.0 Reagents

Water- SYBRON-Barnstead double deionized water

Organic Analytes- 97% purity or greater

Surfactants- The highest purity obtained from the manufacturers

Internal Standard- Perchloroethylene

6.0 Sample Containers, Collection, Preservation, and Handling:

6.1 Sample Containers

Aqueous samples will be contained in 12-ml glass sample vials (Fisher Catalog N 03-338-29C) with Teflon-faced septa caps. The glass vials or the caps are not reused.

6.2 Sample Collection

Each sample vial will be completely filled with aqueous sample, such that no headspace exists, and then capped. The glass vials are not opened until the time of sub-sampling or analysis.

6.3 Transportation and Storage

Similar to the procedure described in the Phase I work plan (Appendex H1)

7.0 Procedure

7.1 Sample Preparation.

In order to account for the effects of surfactants, the samples are split and the surfactant concentration is determined. Sample solutions will be adjusted to the same surfactant concentration by adding the necessary amount of surfactant to the dilute sample. A 2.0 ml aliquot is removed from the adjusted sample vial and put into a 22 ml Tekinar 7000 Autosampler vial with a Teflon-lined septum and crimp cap.

7.2 Operating Conditions for the Shimaz:du 17A GC:

Injection port- 25șC

Oven Temp Initial Time- 1 minute

Oven Temp Initial Value- 40șC

Oven Temp Program Rate- 8șC/min

Oven Temp Final Time- 0 minute

Oven Temp Final Value- 23OșC

Run Time- 30 minutes

7.3 Operating Conditions for the Tekinar 7000 Headsapce Autosampler with Tekinar 7050 Carrousel

Platen Temp- 60 șC

Sample Equil. Time-15 minutes

Mix Time- 2 minutes

Mix Power (0-10)- 5

Stabilize Time- 0.3 minute

Cryo Module Cooldown

Time- 5 minutes at 0 ⁰C (by Liquid N₂)

Pressure Time- 0.5 minute

Pressure Equl. Time- 0.5 minute

Sample Loop- 5 ml

Sample Loop Temp- 200 șC

Loop Time- 0.75 minute

Loop Equl. Time- 0.05 minute

Injection Time- 1 minute

Cryo Injection Time- 2 minute at 200 șC

Line Temp- 120 șC

Injection per Vial- 5 w/multipuncture (can up to 9)

Sample Option- Concentrated

Sample Pre-equl.

Time- 2 minutes

GC Cycle Time- 45 minutes

8.0 Calibration

A calibration curve (five points minimum) must be developed to verify proper operation of the

GC/HS system. Individual calibration curves will be made for each surfactant and different

surfactant concentrations. After fitting the standards, all analytes must meet acceptance criteria.

These criteria include:

• In order to determine if the GC/HS is exhibiting any deteoriated response, the percent relative standard deviation should be less than 20 percent for o-xylene, and 1, 3, 5-trimethylbenzene; less than 25 percent for naphthalene and 0-dichiorobenzene; and less than 35 percent for trichloroethylene and undecane.

• Reproducibility will be monitored through use of a range table or Shewart plot as described in "Handbook for Analytical Quality Control in Water and Wastewater Laboratories" USEPA, June 1972.

The initial calibration curves for total petroleum hydrocarbons are provided (see attachment 2).

As standards are prepared by using the actual analytes, verification of operation and calibration can be easily determined. To help correct for errors due to injection, all analytes are divided by the internal standard (i.e., tetrachloroethylene) and are plotted as area ratios or by simply comparing the peak area in the case of external standard. If any samples have an internal standard area higher than that of the standard, then the average internal standard area from the standards will be used for determining that sample's area ratio. After initial verification, at least three standards should be analyzed after every 10 samples. The standards should include one at or near the quantitation limit, a mid range standard, and a standard at or near the highest concentration of the calibration curve. These analyses are performed to ensure that the calibration does not drift as a function of time. If the calibration checks differ by less than or equal to 20 percent, the initial calibration is assumed to be valid. If this criterion is not met for any one of the calibration checks, corrective actions must be taken.

Surrogate spikes (bromofluorobenzene) will be used to determine the percent recovery or accuracy. The spike is performed by injecting 50 ml of 2000 ppm bromofluorobenzene in methanol into the samples once received in the laboratory.

Since a large number of samples will be collected during the Treatability Studies, only 5 percent of the samples will be analyzed in duplicate and 5 percent of the samples will be blanks.

8.1 Quantitation Limits:

Compound Quantitation Limit (mg/L)

trichloroethylene 50

o-xylene 7

undecane 50

0-dichlorobenzene 5

naphathlene 40

9.0 GC/Headspace Analysis

Once a sample is in the autosampler, the vial will be injected with tetrachloroethylene as an internal standard. In the case of the external standard method, no PCE will be added into the vial. Multiple injections of headspace samples (5 injections at 5 ml each) are made into the GC column where separation and detection occur based on the conditions stated in sections 7.2 to 7.3. A sample chromatogram is given in Attachment 3.

10.0 Data Interpretation

10.1 Qualitative Analyses

Qualitative identification of compounds can be achieved based on retention times. The sample chromatograms are compared with the chromatogram for a mixture of individual components. The relative retention time of sample components should not exceed +/- 0.1 retention units.

10.2 Quantitative Analyses

The quantitation of individual components will be based on the integrated absorbance of that characteristic ion.

The calibration curve is expressed by the following equations

Concentration mg/L = Slope * Peak Area + Constant (External standard)

Concentration mg/L = Parameter 1 * Area Ratio ** Parameter 2. (1nternal standard)

In order to determine the concentration of the sample, the area ratio, which is the area of the target analyte divided by the internal standard, is taken to the power of parameter 2, which should be very close to 1, and multiplied by parameter 1. In the case of external standard, the resulting peak area is compared with the external calibration curve.

11.0 Quality Control

The QC requirements for this project are:

• Analytical equipment will have regularly scheduled maintenance and daily calibration (more frequent when appropriate). In the event that injection of replicate standards results in variations exceeding the criteria levels (as described below), the source of the variation will be identified and appropriate corrective action taken.

2) Before processing any samples, the analyst should demonstrate, through the analysis of a method blank, that interferences from the analytical system, glassware, and reagents are acceptable.

3) A quality control reference sample containing each target analyte is required. The QC reference sample may be prepared from pure standard materials. This reference sample is then injected into the surfactant solution and recovery determined. This should be done at three concentrations; one near the quantitation limit, one mid range, and one at or near the highest concentration used for the calibration curve. For each analyte, the standard deviation of the recovery (mg/L) and the average recovery (mg/L) will be compared with the corresponding acceptance criteria for precision and accuracy

Compound Average Recovery Std. Dev. (%)

trichloroethylene 100.5 24.8

o-xylene 101.2 5.8

1, 3, 5-trimethylbenzene 102.4 5.2

undecane 97.5 4.6

dichlorobenzene 101.1 2.0

naphthalene 98.2 10.5