Validation of the use of an automatic diluter in HPLC analyses by comparison with traditional methods

In an analysis protocol, dilution is nowadays the last step that is never ( or seldom) fully automated. Yet special appliances called diluters are comm ercially available. In spite of the potential advantages in time and solven t saving, analysts are still reluctant to use these diluters, mainly becaus e of the insufficient proof that they are able to give accurate and precise results. Validation by gravimetry did not represent an irrevocable argumen t since it could not account for memory effect. That is the reason why we c hose to carry out our validation of the dilution step in HPLC analyses with an automatic diluter through a comparison with classical methods using fla sks and pipettes or burettes. A parent solution made of concentrated Spiram ycin was diluted by the various methods. The control analysis was operated using HPLC with UV detection. The experimental design involved two operator s and results obtained were processed using statistical tools such as ANOVA and regression. Conclusions were that in any case the diluter was at least as effective as traditional methods. Observed variability introduced by di lution was 0.30% RSD with diluter, and 0.45% RSD with classical methods. In addition, as was shown by the results of the lack of fit test, diluter acc uracy is quite compatible with linearity studies in HPLC. Consequently, it is possible and even advantageous to use a diluter instead of manual method s.