How to choose the correct GC column – Part 1

A month or so ago I gave a short training session to a few beginners on how to go about choosing a GC column. After all, there are thousands of Restek column catalog numbers to choose from (not to mention numerous custom catalog numbers), so if an analyst doesn’t have a method which specifies which column to use, how can one simplify this selection process?

Well, I explained how I work through this process and since the feedback was positive, I thought I would share it with our blog readers. Just like my posts on GC Capillary Column Killers, I decided to make this a four-part series so I could focus each post on just one or two topics. Because an entire book could be written on column selection, I will approach each post based upon generalizations, and not any specific analysis. As a result, I will not discuss application-specific columns, as these are self-explanatory, nor will I discuss any specific method which lists one or more recommended columns. I am going to approach this topic as if no published method is available (i.e. method development).

Now that everyone has an idea what this series is about, let’s start at the beginning of the column selection process. I’m sure everyone has their own way to narrow down column choices for a particular analysis; let me share mine with you. I’ll begin by listing four general categories that I use to help simplify column selection.

 

My four categories of narrowing down column selection:

1. I look at the boiling points of the compounds of interest.

2. I consider the polarity of the compounds (and polarity of the sample/standard solvent, if applicable).

3. I consider compound concentrations, column capacities, and detector used.

4. I look at any other details which may help narrow down the choice to just one, or a handful, of column catalog numbers.

I use the four categories listed above to help narrow down possible column choices into one of the four categories listed below.

 

My four categories of GC columns:

A. Gas analysis columns (solid phase/adsorption).

B. Volatile columns (liquid phase/ partition with β values* less than approximately 50).

C. Semi-volatile columns (liquid phase/ partition with β values * greater than approximately 50).

D. High-temperature columns (liquid phase/ partition high-temperature fused-silica and metal (MXT) polysiloxane phase columns).

*β value = phase ratio = r/2df (r=internal radius of tubing; df = phase film thickness).

 

 

Part 1 of this series will discuss how to use compound boiling points to narrow down column choices.

 

Boiling point

In order to select a correct GC column, one must know the boiling points of the compound(s) of interest, and of all other compounds in the sample(s). After all, gas chromatography is all about the volatility of compounds, and if they aren’t volatile, you will never be able to analyze them using a GC.

More specifically, I use published boiling points to narrow down column selection. The reason I specify published is because in a GC system, the temperature at which compounds will actually vaporize (boil) will be less than the published value (typically found listed in MSDS’s). Supposedly (I am no chemical engineer) this is because in a closed, inert, and turbulent system compounds will boil at lower temperatures than at atmospheric pressure (which was used to determine the published boiling point temperatures). In addition, because we can look up published boiling points (bp’s) for most compounds, those are what I use. I have summarized the relationship between published boiling points and column selection below:

All compounds which have bp’s less than approximately 50°C - use a column designed for gas analysis (solid phase/adsorption).

One or multiple compounds with bp’s between approximately 50°C and 220°C – use a Volatiles column (liquid phase/ partition).

One or multiple compounds with bp’s between approximately 100°C and 450°C – use a Semi-Volatiles column (liquid phase/ partition).

All compounds with bp’s greater than approximately 400°C – use a high temperature column (liquid phase/ partition).

As you can see, there is overlap, which is fine as long as you remember that in order to chromatograph compounds, the temperature needs to be high enough so that none of the compounds are allowed to condense anywhere in the system, including the column. In other words, you need to use a column with a high enough maximum temperature that all compounds can elute off of it. For example, if you need to analyze both volatile and semi-volatile compounds in the same analytical run, you will need to use a semi-volatiles column.

So, if you know only the boiling points of compounds, is this enough information to make an accurate column selection? Sometimes the answer is “yes”, sometimes the answer is “no”. When the answer is “no”, hopefully using this post, in conjunction with the next three posts, will narrow down column selection even more.