Chemical Forums

Chemistry Forums for Students => Analytical Chemistry Forum => Topic started by: Argon on June 14, 2013, 11:51:43 AM

Title: Question about calibration curve
Post by: Argon on June 14, 2013, 11:51:43 AM

Hello! Ive just started working on a GC and I need to create a calibration curve using an external standart. I've injected 3 different concentrations - 10ppb, 50ppb and 100ppb and I got a nice curve with a correlation coefficient over 0.99. The problem is that the software Im using shows results for the calibration in which there are two fields that annoy me - "specified amount" and "calculated amount". It is clear that the "specified amount" field shows the concentrations that I've entered - 10, 50 and 100. But the calculated amount field shows a significant difference - 6.88 for 10; 42 for 50 and 103 for 100. The last one seems to be ok, but the first two? How can there be such a difference, when the correlation is over 0.99? Please give some advice!  :)

Title: Re: Question about calibration curve
Post by: Arkcon on June 15, 2013, 10:16:59 AM

The software might be working just fine.  Or you method may have another point included, by accident or because the method requires one -- i.e. an internal instrument standard, or maybe forced through zero?  Suggestion:  graph all this by hand, and by hand I mean on Excel or another spreadsheet, and see if these calculated amounts do actually fall on the line.