Why are my flash column chromatography peaks splitting?

Split peaks? Multiple peaks? Are they really a problem and what causes the issue?

In this post I will discuss what split peaks are and what to do to fix the problem.

Seeing more than a single peak for each compound in your flash chromatography?  If you are then you may have a defective column and yes, this is a problem. If only one peak is splitting, then the issue is method/chemistry related, a future blog topic.

But what causes the defect and can you fix it?  The main causes of split peaks are a flash column with poor bed integrity or a partially blocked flow path (inlet filter/frit).  By poor bed integrity I mean that there is either a void space at the top of the cartridge, void somewhere in the packed bed, or there is channeling occurring.

A flash cartridge with a void space at the top (a void space in indicated by the lack of media just under the top frit or filter) can cause eddy effects and sample dilution upon injection. Increased elution bands are typically seen and sometimes, due to different migration rates, the bands elute as doublets or even multiples, Figure 1.

Figure 1. A comparison of normal chromatography (top) and chromatography showing multiple split peaks (bottom). Chromatograms are of the same sample with normalized elution scales.

To rectify this problem you can add more of the same media to the cartridge if you can remove the top of the cartridge, Figure 2.  This is a feature most commercially available flash cartridges do not have.  In those cases you need to replace the cartridge.

Figure 2. Biotage® SNAP cartridges designed with removable cap provide the option to fill a void space with fresh media.

If you use flash cartridges with removable caps, e.g. Biotage® SNAP cartridges, you can remove the cap, the top filter (frit), and add more of the same media.  After replacing the frit and the cap, it is ready for use. You can also clean the frit and reuse it if you suspect the frit is dirty.

Column channeling is a different issue that occurs when the media density in the cartridge is lower at the walls then in the middle of the column.  This can be due to poor packing techniques, poor media particle size homogeneity (smaller particles work their way to the column bottom over time), different solvation/wetting rates during equilibration/use, or plastic swelling (especially with normal-phase solvents).  With this scenario, some of the sample will elute more quickly than the others since it by-passes much of the media; this is the scenario in Figure 1.

Commercially available flash columns are not typically prone to channeling or voiding, though these issues can occur.  Because cartridge packing is likely automated by your vendor (we use automation to pack our cartridges at Biotage), cartridge performance is typically consistent.  Automation, however, is only part of the solution.

A column or cartridge is only as good as the media with which it is packed. Media with a broad particle distribution will have a greater chance of voiding or channeling than media with a narrow particle range.  This does not mean that media with a large median particle diameter will void or channel but that if the particle size distribution is large there may be an issue.

Most flash media used today has a nominal median particle size of 50 µm though smaller median diameter particles (25/30 µm) are gaining in popularity. Depending on your cartridge vendor, they may publish the media physical property data.

Most vendors today do use media with a particle distribution between 40 and 63 µm.  This is a relatively tight distribution and, if packed properly, should not void or channel.  If the media you use has a broader particle range, then smaller particles can settle to the bottom of the column potentially creating a split in the column bed. When this happens the larger particles can aggregate, if you will, and create a bridge, due to increased particle to particle friction. This void may be invisible to the eye and the only way it is noticed is when the chromatography is giving you two (or more) peaks for the price of one, not necessarily a bargain.

If you notice splitting peaks (all of the peaks, not just one or two) you can try to remove the void by physically tapping the cartridge walls.  This may disturb the particle bridge and remove the void but may introduce other issues such as peak tailing or peak fronting, depending on how well the particles resettle.

So, yes, it is possible to fix the split peak issue if you use cartridges with removable caps but it may just be easier to replace the cartridge.

What do you do with poorly-performing cartridges?

For information on flash method development and optimization, click on this link and register.

Click the link below for access to a survey regarding this blog; I’d love to see your thoughts. http://www.surveygizmo.com/s3/2716481/Flash-Blog-Survey


Published by

Bob Bickler

Technical Specialist, Biotage

2 thoughts on “Why are my flash column chromatography peaks splitting?”

  1. Years ago when I first started to use large, pre-packed flash columns, I used to worry that they might not have an effective distribution frit. The one on our old Prochrom DAC column had an eight stage mechanical device and was extremely efficient though bulky. I have stopped worrying, it seems they do have a distribution frit, a single stage device which is adequate for flash peaks. When I liquid load a highly coloured sample I can see how the band is distributed across the top of the column. If the sample is viscous then often part of the frit is relatively impervious and I see uneven distribution of the sample. It looks like a problem likely to cause tailing but with flash, the peaks are so broad that it simply disappears.
    With dry loading, the distribution is perfect, probably because the sample going through the distribution frit is non viscous.

    1. Hi Derek,

      Good input. I too remember dynamic axial compression type columns and their ability to eliminate voiding, maximize bed integrity, and sample flow distribution.


Leave a Reply

Your email address will not be published. Required fields are marked *