Kinesis Solutions Techtips

We are often asked for advice on selecting the ’right’ column, and among the characteristics that we ask people to consider are low bleed, a high maximum temperature limit, column inertness, and resistance to chemical degradation (primarily by oxygen, water, and extreme sample pH).

However, what customers usually have in mind is the ’right’ column for getting the ’right’ separation.

In this regard another question often crops up, which may be summarised as: "Should I be using a polar or a non-polar column?" Answering this second question is much easier and usually helps in providing an answer to the first.

Polar columns separate compounds on the basis of their polarity, whereas non-polar columns separate compounds purely on differences in boiling point. This is fine if there is at least 2°C difference in boiling point between all the compounds in your sample. If there isn’t, then it will be very difficult to separate them using a non-polar phase, even with a very long column. Examples of this can be seen with comparison analyses of xylene isomers (Figure 1) and FAMES (Figure 2), where you need to use much more polar phases to separate compounds according to differences in polarity, rather than relying on differences in boiling point. A guide to the polarity of various phase types is shown in Figure 3

Figure 1

Figure 2

Figure 3

Figure 4

When problems with co-elutions are encountered with a 5% phenyl column, or when a different elution order is required for confirmation of compound identities, mid-polarity columns, such as the 35% & 50% phenyl, can be used. These still have relatively high maximum temperature limits, so they can often be used in the same GC as a 5% phenyl column, and allows each sample to be injected only once and split onto the 2 separate columns.

So, in order to choose the right column for your application you need to know something about all the compounds you are trying to separate. Remember - this includes any contaminants that are likely to be in your real world samples, not just the compounds that are in the standard that you may be working with initially. Do they have boiling points that are different enough for them to be separated on a non-polar column? Or are the boiling points very similar? If this is the case, then are there differences in polarity that you exploit, using a more polar column?