|
|
|||
|
|
||||
|
Analytical
Method Development
|
|
|
Bill Katz, President November 2009 Once the choice of a particular analytical technique is made, the next challenge facing the analyst is in either adapting an existing analytical protocol or developing a method from scratch. Both approaches present challenges to the analytical chemist; applying a method from the literature, regardless of how well documented, must be adapted to your specific instrument configuration and test requirements. Having to create a totally new method either on a new system platform or adding a new compound presents it’s own unique challenges moving the analyst into a truly research environment. When developing a new method, the scientist must consider a variety of performance parameters that must be met to have a reliable and reproducible method. Factors including accuracy (how well does your method come to producing the value from an actual reference standard), precision (how well does your method reproduce the value from multiple measurements on the same sample), working curve linearity, and detection limits. There are no correct numbers for any of these; rather, need to be determined by what is regarded as acceptable for your specific analytical situation. For example, accuracy of +/- 10 % may be acceptable if dealing with very low levels, i.e., part per billion, but may be totally unacceptable when detecting much higher concentrations. Any analytical method must produce data that are of use in that particular situation whether it be incoming materials evaluation, R&D projects, manufacturing support, quality control, etc. The analytical chemist must closely interface with his/her counterpart that will be using the data to ensure that the data are not only technically valid but of use in real world applications. An excellent analytical method that is limited to 1 part per million is of little use if the user of the data needs meaningful results down to a part per billion. A critical component to the analytical method, and a part that is often overlooked, is in the preparation of the sample. If the sample matrix is heterogeneous, requires extraction and/or concentration, or any other physical or chemical treatment step prior to the actual analysis are all factors that can effect the ultimate output. Accordingly, as part of the method, the analyst must address such sample preparation issues and decide on a statistically meaningful sample size to minimize any possible sampling induced errors. 2499 Rice St., Suite # 90, Roseville, MN 55113 (651) 493-8832
|
