Laserfiche WebLink
18/12/1995 20:44 9163519357 ANNE M FARR PAGE 05 <br /> a 99 <br /> I <br /> toSO <br /> 034 <br /> i <br /> I Fg. 2a, (left)_ Gas chromatographic trace of a gasoline standard (without MTBE) diluted <br /> to 100 Pam per billion. Notice that there are many peaks, none of which can be separately <br /> identified without mass spectrometry. The prominent peaks at 5.75. I0.50 and 16 69 minutes <br /> emergence represent iso-pentane,2-methyl-pentane and toluene respectively. <br /> ' 2b, (right). GC trace of household well water with MTBE as the only detectable <br /> contaminant, with a peak at 5.34 minutes. It was quantified at 106 ugA (ppb) The smaller <br /> peaks to the right are attributed to"column bleed"_ <br /> If MTBE is suspected in the sample, the particular GC peak can be analysed by mass <br /> spectrometry (GCM) as it comes through the column. An example of the resultant peaks is <br /> given in Fig. 3. Such analysis provides a choice of a small number of compounds most likely <br /> to produce the peaks.The chemist then has to do some detective work to pick the compound <br /> most likely m be present given the chemical contcxL Support for the choice comes from a library <br /> € of standards of mass spectrometric patterns. <br /> f The head space GC and GCWS techniques are not optimized for the detection and <br /> quantification of MTBE,so the detection limit Ls in the range of S ppb for GC and 100 ppb for <br /> GUMS. The problem is that MTBE, with its high water solubility and low vapor pressure, <br /> only slighdy partitions into the gas phase of a headspace vial at 70 C. To lower the detection <br /> limit. a purge and trap technique can be used to further concentrate MTBE prior to analysis by <br /> GC,or the sample can be directly injected into the column. <br /> '• el 2 <br />