How do you choose a second column for GC confirmation?

[LocoLobo]

How do you choose a confirmation column in GC? Is there a standard for this? Do you use McReynolds constants or something else? A polarity chart? Where I work, if we use a DB-1 or 5 for the first analysis, we can use a 35, or 17 (or something even more polar) for confirmation.

I understand we need to select a column with different separation characteristics in order for our confirmation to be valid. Somewhere I heard someone mention a points system where different phases were 1 pt(?), MS was 2 pts, etc. Fine, but does someone know how you quantify the difference between columns? Has somebody done that in the past? Can you point me at such a discussion/article?

I would like to create a database of our columns with their separation characteristics. So, how do you do it?

thanks,

Paul

[chromatographer1]

Locolobo,

You must be as crazy as a fox when choosing a confirmation column. (sorry, could not help myself)

The most important factor in choosing a confirmation column is the chemical nature of the analyte you are measuring and of likely impurities which could coelute with it on your primary column. There are relatively few bonded phase capillary columns with different selectivities from which to choose:

methyl silicone
cyano silicone
phenyl silicone
trifluoropropyl silicone
cyano-phenyl silicone
polyethylene glycol
ionic liquids (newly available)

Omitting columns using a gas solid mechanism of separation, determine a secondary column which separates known impurities and your analyte of interest in a different order from the primary column. McReynolds constants may be helpful if you have a copy.

Generally it is wise to combine a non-polar column with a polar column.

As is often the case in discussing chemistry, it helps to get specific rather than general when discussing these kinds of issues.

Look at retention times published for your analytes and remember that the 'same' phase from different vendors may have slight differences in selectivity. Do not depend upon published values as pressure and temperature can affect the separations you seek to confirm.

The good news?

Job security. Do the research and document the reasons for your choice for the authorities who need to understand and approve your choice.

best wishes,

Rod

[LocoLobo]

Thanks Chromatographer1,

To be a little more specific we are doing pesticide residue analyses in a wide variety of samples (foliage, animals, feeds, soils, etc.). Our QA calls for confirmation of positive samples with a second column. I realize "pesticides" isn't very specific either, but we really don't know what tomorrow brings in the door. It could be for Chlorpyrifos in soil (easy) to Spinosads in bees (not so easy).

The manager makes the decision as to which columns we can comare results for. I 'm not disagreeing with his decisions but would like to be able to make these decisions based more on science than just experience. I respect experience and understand it is based on what works but still want to ask how and why.

We do use all the phases you listed above except the ionic liquids. As much as is practicable we will choose a confirmation column based on the polarity charts; ie. non-polar to mid-polar to polar. More and more we use GC/MS and LC/MSMS which opens up other questions as to confirmation.

Hopefully, somebody can point me to some references on the subject. I appreciate your response and will look up the ionic liquids.

Thanks,

Paul

[gpronger]

For pesticides, many manufacturers offer recommended pairing for pesticide confirmation.

For environmental work, (and running pesticides in what sounds like natural products would be fairly related) the problem is that on any given sample when we're running a long list of analytes, the likelihood of having something extract that has the same retention time as our target analytes approaches 100%. Its from this issue that a lot of the USEPA methods require that for GC work we confirm any positive detect with a column of dissimilar phase.

As Rod states, what you need to do is select a column with significant differences in polarity (while still chromatographing the compound). It may be easier to think about this in terms of what wouldn't be a good choice. For instance the common DB-1 is a methylsilicone based stationary phase, and a DB-5 is a 5% phenyl, methylsilicone phase column. Though there is a slight difference in polarity, for all practical purposes the separation between these two columns is very, very similar. Now, USEPA method 507 uses a DB-5 for primary and a DB-1701 for confirming. The DB-1701 is a 14%-Cyanopropyl-phenyl)- methylpolysiloxane phase; which is fairly more polar than the 5% phenyl methyl silicone column. Most column manufacturers have come out with proprietary columns for pesticides which tend to work a lot better than what was originally referenced in the USEPA methods, but because they are proprietary, the manufacturers don't give us much in the way of column information.

In any case, what we're attempting to do is select columns with significantly different stationary phases so that we have a higher level of certainty in our measurements. When we review the manufacturers, the more info we have regarding their polarity the better you can make these determinations.

[chromatographer1]

gpronger's post was excellent.

Most if not most users of pesticide analysis use a 5% phenyl type phase.

OV-1301 or OV-1701 or OV-225 phase could be excellent alternatives for confirmation purposes. PEG or Ionic liquids might not be for that class of analytes. OV-200 or OV-210 could be a good choice as well, or perhaps not.

Look for articles that discuss phase selectivity, and especially look at articles published by Agilent, Supelco, Restek, and other vendors about pesticide analysis.

Good luck, I hope you find the information you seek.

Rod

[Bigbear]

Don't over look the vendor's catalog for lots of information. We always have used J&W ( now Agilent) and there is a section in the Agilent catalog that recommends column pairings for quite a lot of methods.

[gpronger]

Regarding column selection, though tedious, we do have an option with the data typically supplied by the vendors (chromatograms).

If you are patient, you can sort out elution order between columns and look for two where the actual order is changing significantly.

You'll need to double-check temperature limits, etc., on the columns, but otherwise you can use this approach to use columns from different manufacturers.

Greg

[LocoLobo]

Thanks Everybody.

Your answers have been helpful. We mostly use J&W ourselves so that simplifies the catalog/literature search. We have a fairly good selection of columns so that isn't a problem. Time to sit down and do some reading.