SITE INFORMATION AND CORRESPONDENCE

Home Browse Search

Octa6er 2002 • LUSTLine Bulletin 42 Tablol EFFECT OF r r HYDROLYSIS r 800 present in the sample and that the higher concentrations of TBA in the Time of Incubation Percent MTBE undiluted samples were produced by Preserved with (minutes) MTBE(pg/L) TBA(pg/L) Hydrolyzed hydrolysis of MTBE. In 15 undiluted samples,the frac- pH-1 30 195 2HCL o 536 55 57 tion of MTBE that was hydrolyzed 5 so 88 5 343 77% during analysis varied from 22 per- cent to 89 percent, with a median of HCl 0 495 <3 62 percent hydrolyzed.The hydroly- pH=2 30 401 25.2 6.1% sis of MTBE in the unsamdiluted les _. -- — - _60 :; u _ 393. '_ 53.8 _. 130% P 1%trisodmm o 476 3. increased the reported concentration phosphate 30 424 <31 1% of TBA by a factor of four to eight. H>12 60 432 <3 <1% When samples that were diluted 1:10 'The hydrolysis of 1 pg/L of MTBE should yield 0.84 pg/L of TBA. are compared to samples that were diluted 1:100,the extent of hydrolysis in the samples that were diluted 1:10 Table EFFECT OF rHYDROCHLORICr varied from 1 percent to 18 percent with an average of about 9 percent. ON r CONCENTRATION OF These data-quality problems TBA(pg/L) MTBE (pg/L) associated with the hydrolysis of corrected for corrected for Percent MTBE MTBE to TBA illustrate the impor- Sample ID Dilution dilution dilution Hydrolyzed tance of a quality assurance/quality control program. Any significant ML-12-16 none 3,230 89% hydrolysis of MTBE can be detected t10 toss 653 easily if matrix spike samples are included in the analyses. The accu- ML-16-12 none 10.400 83% Aj racy of the analysis is determined by 1:10 2,644 5.8% measuring the concentration of the tto0 2,006 12666 target compound present in a sam- ple,then adding a known concentra- ML-17-12 none 6.170 56% tion of the target compound to a 110 1,493 0.7% replicate sample of the same water(a 00 Aos 73,273 matrix spike)and again determining the concentration of the target com- ML-19-12 none 4,640 _ pound. The concentration in the 1:10 1,309 - _ matrix spike sample should equal the sum of the spiked concentration and 1:100 591 7,216 the original concentration. ML-19-16 none 6,100 _ 68% Therefore, if water samples are 110 1222 7.3% 1 preserved with acid, there is an understandable concern as to 1 1 n0 _ 740 8,551 whether or not any of these data are ML-z3-ts none 719 valid. Unfortunately, the answer to 110 260 2 550 this can only be determined by _ _ reviewing the reports of analytical results from each site of interest.The and shipped to EPA's R.S.Kerr Envi- preservative was diluted tenfold, the things to look for are indications ronmental Research Center for analy- rate of acid hydrolysis of MTBE was of sample-preservation methods, sis using a static headspace sampler reduced tenfold, and the concentra- method operating parameters,qual- (Method 5021). tion of TBA produced from hydroly- ity assurance/quality control results, The water samples were brought sis was reduced. The reported and whether or not confirmatory to 80°C for 30 minutes prior to analy- concentrations in Table 2 are cor- identification of analytes is provided. sis of the headspace by GC/MS. rected for dilution of the sample.The The rate constants published by Replicates of selected groundwater reported concentration of TBA in the O'Reilly et al. (2001) can be used to samples were diluted and then ana- undiluted samples was much higher estimate the stability of MTBE in lyzed. The concentration of TBA than in the diluted samples. water samples. Figure 1 (page 6)pre- reported for a sample was the sum of The last column in Table 2 pre- sents predictions for water samples the concentration of TBA that was sents the fraction of MTBE that was that are preserved at pH=1 and originally present plus the concentra- hydrolyzed during analysis.The frac- pH=2 and stored before analysis at tion of TBA produced from hydroly- tion was calculated by assuming that temperatures of V C, 10°C, and 200 sis of MTBE. the reported concentration of TBA at C.If the samples were refrigerated at For each tenfold dilution, the the highest dilution was the true con- 10'C or lower, less than 5 percent of concentration of acid used as a centration of TBA that was originally ■continued on page 6 5