Dionex Acclaim E2 column issue - scaling from HPLC to U-HPLC

[bisnettrj2]

This is a topic I wish Uwe was still around for....

I have been running a column from Dionex for years - the Acclaim E2 250x4.6mm 5 micron. I get reproducible separations from column lot-to-lot at 52:48 H2O:MeOH at 1.0 mL/min and 30 degrees C. I recently purchased a 150x3.0mm 3 micron Acclaim E2 column and expected to be able to directly transfer my conditions to this column - 0.71 mL/min, 25.52 uL injection volume, same temp and mobile phase constituents (according to Thermo's hplctransfer.com and my own independent calculations from this document http://www.iptonline.com/articles/publi ... -print.pdf). However, I can't get a decent separation unless I go to 45:55 H2O:MeOH, or change my temperature parameters to 26 C and 51:49 H2O:MeOH.

Below is the comparison between the exact same column on the two systems I'm using (there is a difference in the injection volume -40uL for the 1290, 100 for the 1100 - and the wavelength recorded - 254 nm with no reference versus 254 nm with 360,100 reference)



And here is the comparison between the new 3 micron column and the existing 5 micron column:



I'm losing resolution somewhere between the last three peaks, as can be seen in the comparison between the upper and lower chromatograms.

Also, this seems to be a 'known' issue with these columns, as evidenced by the following:

http://www.dionex.com/en-us/webdocs/417 ... -Jul09.pdf

On page 14, an example chromatogram is shown using 51:49 H2O:MeOH at 0.9 mL/min and 26 C, with good separation, which just doesn't compute if you're trying to directly transfer a method from the 5 micron to 3 micron phase.

My question is this: What could be the reasons behind this discrepancy? Is this evidence of poor stationary phase reproducibility with reduction of particle size? Is there something else I should consider when looking at scaling phases from HPLC to UPLC columns? Is there just something I'm missing here?

[tom jupille]

Peaks 13 & 14 on the 5 micron have merged on the 3 micron, so you're definitely seeing a selectivity difference (as opposed to a simple difference in plate counts). Like you, I suspect a difference in the stationary phase chemistry. It's hard to imagine a pressure-related shift in equilibria, but you could check by easily enough by slowing down the flow rate to get the same back pressure (about 0.43 mL/min ??).

Your past experience with the 5-micron suggests that their lot-to-lot reproducibility is pretty good. It may be that this is simply an anomalous batch. If it were my problem, I'd contact Dionex and try to get a new 3-micron column from a different batch. If it matches the one you have, then it indicates that they are making the 3-micron material differently.

[bisnettrj2]

Thank you Tom. I did execute a run around 0.43mL/min (not posted, same results, 14 instead of 15 peaks, co-elution at the end of the run). I'll contact Dionex on Monday and inquire. The pressure related to the 3-micron run was around 380 bar, so I too doubt and pressure-related selectivity differences.

The only other thing I could think of was frictional heating during the run, which might explain why 26 degrees C instead of 30 C got me the separation I desired? However, this is probably not the case here due to the relatively low pressure. Is there a way to determine whether frictional heating is an issue in a U-HPLC separation?

[tom jupille]

With a 3-micron packing, you're really not into "UHPLC" territory (don't get me started on that term!); 3-micron materials have been commercially available since the mid-80's. I don't think you would see this big a selectivity shift from pressure until you got up into the 10kpsi region.

Ditto for frictional heating; that's a function of the amount of energy you have to put in to get flow and the surface / volume ratio of the column. Again, with 3-micron material in a 3-mm column, I can't see that it would be an issue. And, if it were, I'd expect to see most of the peaks shift to lower k'.

All in all, I'd vote for chemistry.

[bisnettrj2]

Tom:

Thank you for the reply. However, I think the question still stands - how can I tell if frictional heating is affecting a high-pressure separation (I won't use the UPLC/U-HPLC terms, as they irk you.. ).

[tom jupille]

You really can't. That's what has muddied the argument about pressure-related selectivity shifts (is it really the result of increased pressure or is it heating?). You could get some evidence by looking at k' values (for *all* the peaks, not just the last 4). If dropping the temperature restores all the peaks to their earlier k', that's a pretty good argument for the temperature effect.

A quick-and-dirty way to do it would be to scale-expand the faster chromatograms until the solvent fronts matched on all three chromatograms (5-micron, 3-micron high temp, and 3-micron low temp). If it's a temperature effect, all the peaks in the first and third chromatograms should line up, with the peaks

All of that said, in this case the drop in particle size is partially compensated by the decrease in column length, and the narrower bore sheds heat more effectively. It's not "evidence", but my gut feel agrees with yours that your conditions aren't different enough to account for the selectivity change.

[bisnettrj2]

Thanks for the input Tom. I'll post back if I get anywhere with Dionex, and if it turns out to be a bad lot or the 'norm' with the 3 micron phase.

[XL]

Bisnettrj2,

Please ask your sales contact (Alex?) to contact me. I will arrange for sending you a replacement column. I also appreciate that you send your current column to me for investigation.

Xiaodong Liu
1228 Titan Way
Sunnyvale, CA 94086

Thanks!

[bisnettrj2]

XL/Xiadong:

I contacted Alex via email this afternoon regarding this issue. I'll try to send off the column tomorrow. Thanks for the reply, hopefully we can get this figured out.

-Robert

[bisnettrj2]

XL:

Hopefully Alex informed you that my new 150x3.0mm 3-micron column performed as expected, and I was able to reproduce the 8330 testing chromatogram. Did you have a chance to test my column and determine if there was anything wrong with it?

Also, is Dionex planning on introducing a sub-2-micron version of this phase?

[XL]

Robert,

I am glad that the replacement column works well for you. Please feel free to contact me (xiaodong.liu@dionex.com) and Alex if you have any further questions regarding this and other Acclaim columns. We are in the process of investigating the previolus column. I will let you know the result in a separate email.

Regarding the sub-2-um version of Acclaim E2 column, it should be quite straightforward to transfer the column chemistry from a larger particle size to a smaller one. Now the question is: how much will the users benefit from it? Sub-2 micron particle columns are more prone to column clogging due to "dirty" samples compared to columns packed with larger particles. This requires some serious effects on sample preparation, and the use of in-line filter and/or precolumns. While the thorughput for each separation is higher, the time for sample preparation could be longer. To be honest, I am not sure about the "net" benefit of the sub-2um version of this specialty column. But I will surely contact you if we need beta test sites for testing such column.

Thanks for the interests, feedback and comments!

[bisnettrj2]

XL:

The obvious benefit to a sub-2 column of this type is that the EPA method demands a positive confirmation of all tentative analyte identifications above the detection limit on a secondary column that has an alternative selectivity to the primary column. I use the E2 as a dual primary (against a C18), and if both runs were faster, I could turn over more samples, have higher throughput, yada yada. But please, if Dionex wants a beta tester for a sub-2 E2, I have two Agilent 1290s I can test them on.

[XL]

Robert,

I see what you mean. Do you mean you are using or planning to use UHPLC C18 column, and are in need of a UHPLC type Acclaim E2 column to increase the sample throughput? I will be in touch on your request.

[bisnettrj2]

Both. If Dionex wants to beta-test a sub-2 E2, I'll volunteer, but only if they can provide a guard column/cartridge system to protect the analytical column. Too many sub-2's I look at have no guard system available, which might be fine if you're running clean pharmaceuticals, but is not fine when you're running soil extracts from some old firing range the Army decided it didn't want anymore, like I do.