Thursday 10 May 2012

Help on: Robustness factorial design for GC

Do you have any problems relating to analytical chemistry for pharmaceuticals or training? Send your questions to the MTS helpdesk using our contact form.

“I am currently validating an analytical method and I want to look at robustness testing by factoral design. I have never done this before and was hope you could point me in the right direction for what kind of set up I could use. It’s a GC method with a temperature ramp, fairly standard in all other aspects.”

"The first step in any robustness study is to select the factors and levels which you are going to investigate. The most common factors investigated for GC methods are the temperatures (injector, detector and column), the flow rate of the carrier gas, and if applicable, the split ratio. Also, if it is a headspace method then some of those method variables may be important for the robustness of the method. The limits typically investigated are approx +/- 10% but really it is up to you to decide how much variation is likely to be experienced in routine use of the method. It may be unlikely that anywhere near 10% variation will be experienced.

The experimental design for the robustness testing could be simply by investigation of each parameter at two extreme levels (above and below that in the method) whilst holding everything else constant, the one factor at a time (OFAT) approach. Alternatively, you have asked about using a factorial design. In this approach the identified factors are investigated simultaneously. This does require an understanding of both how factorial design is applied and how to interpret the necessary statistics that are generated. A commonly used design for analytical methods is the Plackett-Burman design, which can be used for up to 7 factors in 8 runs.

I think that I would recommend that you invest in a suitable design of experiments software package if you decide use a factorial design approach. The training on the package, even if it is just going through the manual, should help you to understand how they are applied and also the limitations of the approach. If you feel that there are only 4 or less factors that are likely to have an effect on the robustness of the method then an OFAT approach may be easier and less time consuming, since the number of experiments will be similar and the time taken to get to grips with using factorial design won’t be necessary."

No comments:

Post a Comment