a) EPA is using the five level calibration mixture recommended in the ASTM draft method for aromatics in gasoline. Pre-made standards for that method can be purchased from at least two vendors. The list of compounds used in the ASTM calibration are generally available from at least one vendor. For further information, contact Carl Scarboro at (313) 668-4209.(10/31/94)

b) Precise results are possible by using a multi-level calibration using the compounds listed in either the ASTM procedure and by using: the response factor for Indan for all uncalibrated aromatics with a molecular mass of 117, the response factor for 1,2,3,5tetramethylbenzene
for all uncalibrated aromatics with a molecular mass of 134, and the response factor for pentylbenzene for all uncalibrated aromatics with a molecular mass of 148. There are undoubtedly some errors in the quantification but they are probably less than 1.0 of the total aromatic number. Leaving out the uncalibrated aromatics would underestimate the total aromatic number. We have no evidence that using a response factor of one is as good as using a response factor of a calibrated compound of a similar mass and structure.(10/31/94)

c)Using the above calibration standards and a reconstructed chromatogram at the masses 117, 134, and 148 of a gasoline like RFA or other high aromatic gasoline, one can identify uncalibrated aromatics with the above molecular weights, assign a single point through zero calibration curve using the curves generated for one of the above three calibrated compounds and produce an adequate calibration table for this procedure.(10/31/94)

d)The 25 vol% only refers to the volume of a single internal standard in the total internal standard mixture; i.e., a compound is specified as 25 vol% of the internal standards. That translates to a much lower number when it is mixed into a sample or standard. Presently, EPA is using 6.0 wgt% ethylbenzene-d10, 2.5 wgt% benzene-d6, and 2.0 naphthalene-d8. This seems to work well, but methylbenzene may be better quantitated with the addition of methylbenzene-d8. We believe other compounds should work as well.(10/31/94)

e)The 2 vol% is the allowable difference for lab duplicates of a sample. That is, duplicate analyses should not differ by more than 2 vol%. In addition, the analysis of the quality control standard or reference material should also not differ from its nominal concentration by 2 vol%. These are fairly wide targets. EPA is currently seeing a precision of 5.0% of point at the 95% confidence interval. That figure translates in a standard or sample containing 25 vol% aromatics as 1.25 vol% for a 95% confidence interval and less than 1.9 vol% at a 99% confidence interval.(10/31/94)

f) If it is necessary to add the unknown aromatics into the total aromatic content, we suggest using the same type of calculation used in the other EPA methods (assume a response factor of 1). This would allow for easy modification of existing software to handle the calculations.

We believe that by following the steps outlined in paragraphs b and c above, and the ASTM procedure one can produce satisfactory results. The use of a response factor of one may or may not produce similar results. We have no data at this time to make an evaluation.(10/31/94)

This question and answer was posted at Consolidated List of Reformulated Gasoline and Anti-Dumping Questions and Answers: July 1, 1994 through November 10, 1997 (PDF)(333 pp, 18.17 MB, EPA420-R-03-009, July 2003, About PDF)