HPLC column - 4.6, 3.0, or 2.1 mm

[parkc23]

Hi, I am considering to purchase Agilent Poroshell column with 150mm length. However, not sure about a diameter between 4.6, 3.0, or 2.1 mm. My HPLC can handle back pressure of 600 bar. Do I have to be worried about system dead volume if a a narrow diameter column such as 2.1 mm is chosen? Any suggestions? Thanks.

[Consumer Products Guy]

A modern system should be able to handle the two smaller diameter columns with no issues. In my book, 4.6mm is "last decade" in most cases, so get better results, and save some solvent.

Actually, some 4.6 x 50 columns can save solvent and reduce run times too, depends what your separation and needs really require.

[juddc]

If you have a rig that can handle 600 bar, the smaller diameter columns should be fine. I tend to prefer 3.0mm columns, but that's because I used to do a lot of work with pretty complex matrices and found the wider bore columns to be slightly more rugged.

At the moment, I'm running an ancient LC and find that it's quite possible to generate really nice separations using this type of column in a 4.6-mm diameter, so you should do fine with either the 2.1 or 3.0-mm.

[Andy Alpert]

I hope that the embedded image comes through. It shows the consequences of using a 2.1-mm i.d. column in an HPLC system where the extracolumn dead volume is optimized for the flow rates appropriate for columns with i.d. of 4.6 mm. At 0.2 ml/min, peaks exhibit pronounced frontal tailing. When the flow rate is increased to 0.4 ml/min, the effects of extracolumn dead volume are cut in half and the tailing all but disappears. Conclusion: Just because your HPLC system can handle the backpressure involved doesn't mean that you don't have to worry about other aspects of the setup. Take pains to minimize the extracolumn dead volume: Use shorter connections with lower i.d. and use a smaller detector cell (if you're using absorbance detection). Another consideration is the precision of the flow rate. If your flow rate is 200 ul/min, then at 99% solvent A/ 1% solvent B, solvent B's flow rate is 2 ul/min. Same concern applies for solvent A at the end of the gradient. Many HPLC systems don't deliver solvent with this degree of precision, so the beginning and end of gradients will not be precise. Check the specs of your system before you buy a 2.1-mm i.d. column to use with it. Incidentally, if you scale down the flow rate proportionately, then the backpressure with a 2.1- or 3.0-mm i.d. column will be the same as the backpressure with a 4.6-mm i.d. column.

Andy Alpert
PolyLC Inc.

[juddc]

Andy is correct, of course, though I admit that I have never met an LC that could push north of 8000 psi that was not reasonably well optimized for 2.1mm diameter columns out of the box. I know a Waters Alliance or Agilent 1100 could run those without issue. 8k psi is at or near UPLC range, isn't it?


I can and do run 3.0 mm columns on my 20 year old ghetto component setup without issue, though for columns with ~2.6 uM particles I tend toward 4.6 mm so that I can keep my flow rates up around 1 ml/min and I find that this works very well. For any system designed past about 1997, 2.1mm should not be an issue and 3.0mm will work perfectly well.

With that said, perhaps we should ask the OP for particulars on their instrument and its dead volume before making any recommendations. Mea Culpa!

[Gerhard Kratz]

And beside pumps, old or brand new, don't Forget the flow cell volume in your detector. Decreasing ID of the column should also be followed by an decreasing flow cell volume.

[parkc23]

Thanks everyone for your kind input.
Gerhard, how small the flow cell should be for 2.1 x 150 mm column?

[Gerhard Kratz]

Standard flow cell volume is between 10 to 14µl
Semi micro flow cells are from 2,5 - 3,2 to 5µl
Micro cells from 140nl - 0,9µl to 2µl, depending from which manufacturer is your detector. Around 2µl I would recommend, or a 0,9µl cell. And the tubing should have not more than 0,1mmID and as short as possible.

[paul_b]

A note of caution - if you have an older HPLC e.g. an Alliance the large system dead volumes (approaching 1ml!) will play havoc if you want to run gradients on a small bore column.

However we now routinely run fused core columns on our 1996 HP1100 (high pressure mixing low dead volume) and get great results.