TPH analysis by GC:valley to valley or baseline

[fractionated]

I will analyze the TPH concentration of diesel oil using GC-FID. As i noticed as the peaks get closer and closer, the peaks start to move farther from the baseline and create a mountain-like and curve space between the early eluting peaks and the heavier peaks. In analysis of TPH, which is a more acceptable method: to include the analysis of the hump or baseline integration or to perform a valley-to-valley integration? If quantification is done by external standard using the same diesel oil which integration is preferred? How about if quantification is performed using an internal standard say hexadecane-d34 and its peak is part of the big hump/did not touch the baseline, how can we perform the quantification.A million thanks for your help.

[Johnny Rod]

The individual peaks are riding on the big hump/peak/baseline shift/whatever so skim them off i.e. valley-to-valley. You'd only use the "original" baseline where you have some peaks (not too many either, 2 or maybe 3) close together of similar heights. A typical rule of thumb is a maximum height ratio of 1:20 for splitting, more than that, skim off.

[fractionated]

thank you Rod. I am sorry that I havent gotten clearly your answer. Do you mean that I only measure the peaks that are riding on the hump and ignore the areas on the hump in the computation? Do you usually use peak height or peak are for quantification? Will valley to valley involve a manual integration that is tedious in Gc analysis of hydrocarbons in complex crude oil or petroleum products?

[Yama001]

TPH analysis normally is intended to measure that 'hump' (UCM - unresolved complex mixture). If you need single peaks riding along that hump, valley to valley is appropriate there, as well as for any internal standard that may be used.

The hump is measured by constructing a flat baseline to grab that area - the manner in which this baseline is drawn is not consistent amoung laboratories, in my experience. A particular difficulty is accounting for column bleed.

Our lab draws a baseline that encompasses the entire chromatogram, drawn at the lowest point possible. An instrument blank that is treated in the same manner is subtracted. This is by no means perfect, as the area from an instrument blank may vary as the GC system degrades. We try to ensure these errors are less than our reporting limit.