By Anonymous on Thursday, May 15, 2003 - 03:31 am:
When checking linearity of the methods for API and its impurities should I use the product with different amounts of analyte prepared as the sample will be in the future, or can I just use standard solutions of diffrent concentrations?
This may be a simple question but I have limited experience, and all papers I have found so far don’t state that clearly, so I hope you will find some time to answer it.
Best regards
By readski on Thursday, May 15, 2003 - 04:04 am:
Use standard solutions to determine linearity... then when analyzing samples make sure you are within the linear response range.
By Anonymous on Thursday, May 15, 2003 - 04:40 am:
A Basic follow-up question- Is a serial dilution better for linearity determination than independent weighings of standard to determine linearity? What about for the standard curve in the actual analysis?
By readski on Thursday, May 15, 2003 - 07:50 am:
A serial dilution should be more linear (assuming that ones dilution skills are better than quantitative transfer skills). Some may argue this point. I am not sure what you mean by standard curve in actual analysis. Do you want to use a multi level calibration? This would be determined by the software that you are using.
By Anonymous on Monday, May 19, 2003 - 07:54 am:
Ask yourself: If the preparation of your stock solution was wrong and your dilutions are o.k. (fine linearity !) how will you notice your fault?
By Anonymous on Monday, May 19, 2003 - 08:49 am:
Make the stock solution twice!
By readski on Tuesday, May 20, 2003 - 06:31 am:
I guess at some point one must trust their ability to make a standard. Errors will happen. Unless this is the very first time making a standard one may have previous data to compare (peak areas, etc).
If one makes multiprep standards what do you do if the linearity is off? reprep them all? How do you know which prep may be bad?
By Daren on Wednesday, May 21, 2003 - 05:05 am:
My experience is that your are looking at two different parts of the method, linearity and then quantification. For checking linearity, serial dilutions from one stock will be best to demonstrate linearity of the analyte detection (at this point it doesn't matter if your stock prep isn't exact since you are not quantitating anything). If this a high level validation (phase III or registration) I would repeat this step two more times.
For quantifying samples it is recomended to use two or three separate standard stock weighings within one curve or as single points forced through zero.