CALIBRATION ON CHEMSTATION...HOW TO NOT HAVE TO RUN every ru

[johncarrigan]

Hello,

New enough to this. Ihave set up a calibration curve for ELSD on chemstation. Its not linear but is fitted with a quadratic line. THe thing is that every week, I would like to run a single standard or two instead of a whole new curve. I would then like to adjust the curve based on these two samples. Can anyone run thought how this could be done?

Regards,

JC

[nick22]

Typically, ELSD calibration curves use a log-log transform rather than a quadratic. If you use a log-log calibration curve, you could run two samples and generate the calibration curve using those two samples. It will be more prone to error due to fewer points, but if your standards are stable enough, you could use the high and low standards from a full calibration to do this.

[tom jupille]

I wouldn't do that.

ELSD is more complex than UV detection. If the response were *truly* quadratic, then a log-log transform as nick22 suggests would work, but the quadratic fit is only approximate. Aside from the non-linearity, ELSD response is a function of things like the nebulizer gas flow rate and temperature.

All of that said, I would suggest that you do it both ways (i.e. your present quadratic fit and use two well separated points from that for a log-log plot) for a while and compare the results. That will tell the tale.