UHPLC – A Small Step for Man, or a Giant Leap for Mankind? - Stuart Jones

In the mid 1990’s, HPLC was being described as a mature product. Most of the development was considered to have been done, and soon it would become a catalogue product like a hotplate stirrer. How far from
reality that has turned out to be. HPLC is developing faster now than at any time since it was first introduced. Just think of the changes with mass spec, nano HPLC, high temperature HPLC, column developments, and
by no means least, UHPLC.

At its outset in the early 1970’s, HPLC used columns as long as 1 meter filled with 40-50um particles, and with injections made with a syringe through a septum directly into the packing bed at the top of the column. Then
came microparticulate silicas, and 50cm followed by 25cm columns were introduced, packed with 10um materials. Separations were much better, but UV detectors were still spectrophotometers with a flow cell.
Remember the old Cecil with a Waters M6000 pump? It was a ‘new day in HPLC’.

The issue with 10um packings was that they needed to be slurry-packed at high pressure, and the back pressure was reliant on a really narrow particle size distribution. However by the late 1970’s we had 5um packings using irregular silica, with good efficiencies and a range of good bonded phases (microbondapak, partisil etc). By about 1980 we had spherical particles, offering a lower back pressure and which were much easier to pack. Suddenly a whole range of column packing companies arrived - some are still with us today. Thanks to them, the price of columns reduced from about £250 to nearer £100. At this stage, HPLC looked very similar to the way it is today. Columns 15 or 25cm, flow rate 1ml/min, temperature ambient – 50oC, back pressure around 2000psi, with UV detection as the method of choice for most users.

The Need for Speed
We always need to analyse samples more quickly. Yesterday is never soon enough. Increasing regulation means more and more testing. Bringing a new drug to market requires thousands of tests on thousands of products and their metabolites. Clinicians want clinical results straightaway if possible. And we now have more work to do, with less people to do it.

The initial approach in the early 1980’s was to use 3um packing materials. Smaller particle sizes give more efficient separations, and hence it was possible to use shorter columns and hence get shorter run times. As a rough guide, a time saving of about 50% was possible. Surprisingly though, 3um columns did not receive the rapturous response that might have been expected. One eminent scientist from the MRC who retired last year once told me that his run time had changed from 30 minutes to 12 minutes, but his sample preparation time had increased from 12 minutes to 30 minutes. His column cost had increased by 30%, and the column was much more sensitive to blocking up, and hence both his samples and mobile phase now needed filtering.

Another approach has been to use monolithic silicas. This is essentially a rod of highly porous silica, inserted into the column tube. Its high porosity allows the use of higher flow rates, which offers the faster analysis times. This is primarily the domain of Merck, and a number of applications are available from them. The speed improvement comes at a price, and columns are slightly more than double the price of conventional columns. Again, although they offer a speed increase, this technology has also not received the high market uptake that might have been expected.

High temperature is another option. It is still under development, but water has been found to act like methanol as an eluent at 200oC. The viscosity is much lower, and the problems mainly include finding columns,
which are stable under hot water conditions, and getting the temperature down before the flow cell to avoid bubble formation. Cambridge Scientific is developing an FID detector for HPLC to overcome the need for this.

A number of suppliers are now offering fused core particles, otherwise known as core-shell particles or superficially porous particles. The centre of the particle is non porous and coated with a thin layer of highly porous
material, giving virtually no band broadening and hence offering a highly efficient separation using a short column.

The final approach has been to use sub 2um particles, which has opened the door for uHPLC.