The Evolution of Capillary Electrophoresis
I have been an enthusiastic advocate for CE since first using the technique in the early 1990’s during my time as a Research Scientist at the old Fisons Pharmaceutical Division soon to be acquired by Astra AB of Sweden. Back in those days we used CE alongside HPLC in a number of areas including: chiral separations using various cyclodextrin additives, free solution CZE for analysis of synthetic nucleotides and other small, organic ions as a far simpler alternative to ion exchange methods, MEKC as a complementary approach to Reverse Phase HPLC and capillary ion analysis with indirect UV detection to determine counter-ion stoichiometry or excess (we didn’t have an Ion Chromatography system).
I can draw your attention to a cross-company exercise that was carried out between 1993 and 1995, that I was a part of, whose goal was to assess the repeatability of CE and to demonstrate the successful transfer of CE methods between independent laboratories. Three exercises were performed and I give the references below so that you can see the results that were achieved using the instrumentation that was available at this time.
And so you can see why, back then, we saw it as a very versatile technique for small molecule analysis with multiple applications. The way that this versatility was largely and rapidly dialled-in through chemistry (the solution chemistry of the running buffers) was very attractive. We even had 3 systems for a modest-sized Department although I suspect funding for instruments was more relaxed in that era. One of the beauties of the technique for a practitioner was its high predictability (in CZE) based on mass to charge ratio. This predictability based on m/z may be even more useful for protein and other large molecule analysis, and the analogy with mass spectrometry is obvious. Any lack of sensitivity could be offset somewhat by the ability to go down to extremely low UV wavelengths possibly as low as 185nm. And of course waste generation is miniscule.
So why has CE failed to maintain a place in drug discovery? The first thing to say is that this question is completely biased to small molecule thinking. There may be a different view to be held in the world of research into alternative therapeutic modalities which I will come back to.
But, in small molecule terms, the big killer was perceived poor reproducibility of both migration time and peak area. This can be traced back to factors such as the source of bulk flow being electro-osmotic and hence dependent on a reproducible capillary internal surface, temperature control issues and injection techniques relying on pressurisation of the sample vial and/or application of a voltage. It seems intuitive that, at least without some effort, these processes are going to be inherently more variable than an electro-mechanical HPLC pump generating flow and a fixed volume injection loop or a metering syringe driven by a stepper motor determining sample injection. In favour of CE, its relative simplicity with few moving parts makes for reliable hardware.
Other factors in CE’s failure to retain its place on the small molecular analytical bench include its scale. If you are supporting Medicinal Chemistry it is one thing to be able to demonstrate a 10% impurity or a good chiral separation but thanks may be somewhat limited if you can’t also offer a realistic solution in terms of a means of purification. Hence, CE would be limited to a place as an analytical/QC tool albeit a valuable one offering orthogonality to LC and a range of applications. There have certainly been times recently when I’ve been struggling away with HILIC when I wished I’d had a system!
The mid-late 1990s saw the advent of combinatorial chemistry and HTS with the drive to higher and higher throughput and hence speed of analysis. This didn’t really play to CE’s strengths as cycle times while reasonable were not as good as LC. For example, flushing protocols to ensure reproducibility bring an inevitable time penalty. MS interfacing seemed to be a bit of a dark art and I’d be keen to hear from anyone as to how that has progressed in the intervening years.
However, I think the most decisive factor was the advent of UPLC heralded by Waters introduction of their Acquity platform in 2004 and associated column technologies. This step-change in LC performance and reproducibility marked a revolution in the power of Liquid Chromatographic separations. The ease of use and sheer generic power of these systems greatly simplified method development making them the clear go-to systems for the non-specialist and specialist user alike for very good reasons. In comparison to LC, there was always a scarcity of scientists with CE experience and knowledge and this lack of expertise meant it was even less likely that CE would be considered even for applications where it would be a better choice. Once that generation of systems were gone, they were gone and even fewer scientists would be familiar with the technique or have the opportunity to become so.
In this way, CE became extinct. Or did it? UPLC might have been the asteroid that wiped out CE for small molecule applications but, as the dinosaurs evolved into birds, CE has not gone away. In the biomolecular field it is alive and kicking. And, given the great and increasing importance of large molecules in the Pharmaceutical Industry, its importance as an analytical tool is surely set to rise accordingly in this area for which it is very well adapted. Scientists working in this field have to be more inclined to select the most appropriate tool for the job as generic approaches are less successful, and CE is most certainly part of their toolkit. Additionally, CE has become incorporated into other dedicated analysers such as for Western Blotting and, of course, its role in DNA analysis is legendary.
Watch this space...
Inter-company cross-validation exercise on capillary electrophoresis. I. Chiral analysis of clenbuterol. J.Chromatogr., 641 (1993) 147-153
An Inter-company cross-validation exercise on capillary electrophoresis testing of dose uniformity of paracetamol content in formulations. Chromatographia, 39 (1994) 180-184
Inter-Company Cross Validation Exercise on Capillary Electrophoresis. Quantitative Determination of Drug Counter-Ion Level. Chromatographia, 40 (1995) 47-50
If any of the issues discussed in this article is of interest to you, please feel free to contact me directly on firstname.lastname@example.org for further information.