Open Access LCMS systems: pitfalls and remedies for the “perfect storm “ - PART 1
A modern day Medicinal Chemistry Department is heavily dependent on having reliable and robust LCMS systems generating high quality data needed to support both ongoing reaction monitoring and final product QC. Although current LCMS technology is generally highly reliable such systems are frequently subjected to a certain level of “abuse” from users with variable knowledge and training levels.
The single most common issue in my experience is that of overloading of the system with over-concentrated samples, most usually samples taken from ongoing reactions. Over time this leads to a build-up of contamination in the systems and manifests as a poor “wavy” baseline with integrated peaks in the blank chromatogram. While this may be less of a problem for reaction monitoring, it will certainly present a problem for the chemist running a final compound QC, having to establish an adequate level of purity. And what of course is a typical knee-jerk reaction of the Chemist? Understandably it may be to increase the concentration of his sample in an effort to overwhelm the effect of the carry-over or system peaks on his analysis, thereby adding to the overall load on the system and contributing further to the problem. The hardest hit chemists will be those working on chemotypes with poor chromophores to whom the background peaks will be disproportionately significant. But, of course, upping the concentration of your sample will be counter-productive to showing it in its best (true) light if the linear range of the detector is exceeded. Whereas a target peak height of, say, 700-1500mAU should be sought as this will give the best data, once it exceeds that upper level impurities in the sample will be over-estimated as they are still increasing linearly with sample load but the product is not. As a result, the apparent %purity of the product is under-estimated.
Beware also the effects of residual catalysts in your samples converting the separation column into a “catalytic converter” capable of on-column degradation of subsequent samples again leading to low apparent purity.
Another factor to consider is the cost of repeat analysis as sample concentrations are driven ever higher. Queues lengthen and dissatisfaction grows! In a worst case scenario it is even possible for a chemist to resort to prep LC in an effort to remove impurities from his compound that aren't really there!
Does this cascade of errors sound at all familiar? What can be done to prevent it?
In Part 2 of this article I will present a number of strategies to reduce or eliminate the effects I have discussed here and to help ensure trouble free running of open access LCMS systems that contribute to a productive Chemistry Department.