[Discuss] How to develop method fit for purpose

[catlover]

Hi everyone,

I spent too much time on method development of pesticide residue test....
When problem comes, I tried to change and study every parameters but most of cases, I failed.
I would like to discuss and learn from you all how to develop method in a faster way.
How are you guys find the direction of solution? By experience? By reading?
Thank you if you share.

By the way, sorry for the poor English. =P

[tom jupille]

By experience? By reading?

Yes to both of those.

For experience, you can only get that for yourself .

For reading, my list of recommended books is on our web site: http://www.lcresources.com/resources/resbooks.html

In addition, I'll put in a "plug" for the training courses from Analytical Training Solutions (http://www.analyticaltraining-solutions.org ; by way of "full disclosure", we provide the LC courses for them).

And last, but very much *not* least, keep asking questions here on the Forum!

[James_Ball]

Definitely remember to change only one parameter at a time!

I still fall prey to wanting to change several parameters with each new attempt but one change can counter the benefit of another change and you may never know you were going in the right direction with the one that works. Patience is definitely a virtue in this type of work, and do it methodically too. Work one parameter at a time trying several magnitudes of change to see if results are better or worse before moving to another parameter.

The most difficult thing to do is visualize how multiple parameters interact with each other. It can become as if you are trying to do multidimensional plotting in your head, which is what it actually is for complicated situations. How does parameter A affect parameter B and C and what will the change in B caused by the change in A do to C? Think how Temperature, Pressure and Volume interact, if you can learn to visualize those interactions in your head you will be able to do method development very well.

[HPLCaddict]

One important thing concerning method development: Make sure what you actually want to achieve (in terms of minimum resolution, etc.) *before* starting the development and actually *stop* development once you're sure you've achieved it. I regularily catch myself trying to tickle just a bit more resolution or a bit improvement in peak shape out of the chromatograms...spending hours with these tiny improvements where the initial quick-and-dirty method actually didn't perform that bad .