Phthalates Analysis (problems since 2009)

[a.r.spencer]

Back in 2009 we developed an in house method based on the CPSC-CH-C1001-09 method for phthalates analysis utilizing GC-MS. We always seemed to be having a problem with linearity in the range from 0.5 µg/mL to 10 µg/mL. Our calibration curve seemed to be opening up (like the right half of a U).

I located a paper from Shimadzu (http://www.shimadzu.com.br/analitica/ap ... 46E152.pdf) and duplicated it almost exactly. The differences being: (1) I run a Thermo DSQ and not a Simadzu GC/MS-QP2010, but both are quadrapole detectors. (2) Their ion source temperature was 230 °C and my ion source temperature is 250 °C. Once again my calibration curve is opening up. I'm getting correlation coefficients along the lines of 0.97 and 0.98, while the paper is claiming correlation coefficients such as 0.999!

Two years later I'm still using these terrible calibration curves (forcing a linear fit) and still have not determined what is going on.

I was hoping others might be having similar problems with phthalates and have figured out the solution.

Please help!!!

[Peter Apps]

Phthallates are ubiquitous contaminants, it is nearly impossible to get a clean blank and so you lower standards actually contain a bit more phtallate than you think they do, giving an upward deviation of the calibration at the lower end of the curve (if I correctly understand you description of the shape of the line).

Peter