The Column - December 2008 - (Page 10)

Incognito The Column www.thecolumn.eu.com December 2008 Jeffery Coolidge/Getty Images Perfecting Method Specfications Following last month’s Incognito in which I highlighted a decline in the quality of gas chromatography method specifications, this month’s column concentrates on how we might properly specify HPLC methods. I see literally hundreds of HPLC method specifications a year and while most of them are fine as general “guidelines” for performing a chromatographic separation, some fall short of the standards required for the quality control analysis of drug product potency, for example. I always try to frame the quality of a specification in terms of my ability to transfer the method to another laboratory with a different set of equipment, reagents and a different operator; a reflection of the “intermediate precision” and “robustness” often measured during method validation. If I have enough information to perform this transfer without recourse to the originator then I consider the method to be well specified. Whether the method is robust, accurate or precise is not known at this stage of course! So, my ideal HPLC method specification might include… method robustness). After all we are defining a method specification — not a method absolute! Methods using the 0.1% TFA approach in which analytical robustness is increased as all acids and bases are fully protonated (and the eluent pH is well away from the pKa of all/most analytes), may not have such a critical reliance on pH accuracy. However, in some methods a pH change of just 0.01 pH unit may have a profound effect on the resolution of a critical peak pair. I feel it’s only fair to let my colleagues know this! It has been suggested that the “allowable” variation in the pH of a validated HPLC method should be between 0.1 and 0.2 pH units.1,2 I will follow up with an article on allowable “tweaks” in the coming months! I always specify the required grade of solvents as, for example, HPLC water purity is more critical for gradient HPLC or certain grades of acetonitrile contain a relatively high concentration of sodium ions and are less suitable for LC–MS experiments because of adduct formation in electrospray mode. Sample Preparation Sample diluent: 10% acetonitrile (aq.) The final sample diluent may well be specified within the sample preparation section of the method specification. However, having the diluent specified in the chromatographic conditions helps when troubleshooting peak shape issues. One can immediately judge whether these issues arise from a mismatch in the eluotropic strength of the sample diluent and the initial gradient eluent composition. Pump Type: Binary pump Flow-rate: 1.0 mL/min Compressibility A: 0.46 104 (atm 1) Compressibility B: 0.96 104 (atm 1) Flow-rate gradient: 10 mL/min/min Maximum pressure: 400 bar Minimum pressure: 5 bar For the gradient profile see Table 1. Specifying the pump type may seem a limiting factor to some, however, unless the method has been validated on both binary and quaternary systems, the differences in the high- and low-pressure mixing characteristics (respectively), of these systems may make critical differences to the separations obtained. The compressibility of the solvents used should be specified and entered into the data acquisition software to obtain optimum flow-rate precision and to prevent solvent channel cross-talk and back flow in high pressure mixing systems that can affect the reproducibility of gradient formation. Some modern instruments are able to automatically apply Eluent 0.1% TFA (aq.) pH 2.1 ( pH 0.05) (Water 2 mAU @ 254 nm) Eluent B: Acetonitrile ( 99.9% grade) It’s unusual for a specification to give a range for the eluent pH. However, I am a firm believer in letting my colleagues know what the method limitations are, especially for parameters that have been identified as critical during method validation (e.g., using a Plackett-Burman statistical analysis approach to identify critical parameters when assessing Author: Incognito E-mail: incognito@thecolumn.eu.com Eluent A: 10 http://www.thecolumn.eu.com

Table of Contents for the Digital Edition of The Column - December 2008

The Column - December 2008
Contents
Do Wrinkle Creams Really Work?
News
Incognito
Market Trends and Analysis
Tips and Tricks: GPC/SEC
UHPLC for the Analysis of Paracetamol
Q&A

The Column - December 2008