from the beginning
Posted: Tue Nov 28, 2006 7:29 pm
by excalibur30
Is there any basic online source about how the DL determined and how the calculations tak place? For example we have a peak. We also have a calibration curve. How can we determine the amounth of the interest compound.
Thank you so much
Posted: Tue Nov 28, 2006 11:15 pm
by Earl
You have a curve expressed roughly as y=mx+b right. Where y= area and x= concentration. you have an area from your peak (your slope and intercept come from the cal curve) so solve for x. By DL to you mean detection limit? MDL method detection limit. Run your lowest curve point 10 times take the average and find the standard deviation of the calculated value. For example your lowest curve point is 10ppm and you run this 10 times getting answers like 9, 10, 8 and so on. Calculate the standard deviation and multiply the answer by 3.14, this is the MDL. Sorry, I don't know a web site that explains this, but any analytical chemistry text book ought to walk you through this. Microsoft excel can do all the calcs for you.
Posted: Thu Nov 30, 2006 11:14 pm
by KarenR
Most of the USEPA methods also contain a section regarding how to calculate the MDL. These are available on a subsection of the USEPA website (
www.epa.gov) but I don't have the exact link - sorry.
Posted: Fri Dec 01, 2006 2:26 am
by Noser222
There are a couple of other approaches. One is to use the signal-to-noise ratio, if you are able to calculate the baseline noise. A signal-to-noise ratio of 3 is considered the detection limit.
Here's a nice illustration:
http://microfluidics.stanford.edu/snr_f ... rgraph.gif
The other is to calculate the uncertainty of the y-intercept, which Excel can do for you. Multiply this number by 3 and that's the detection limit.
