Plate Count in UHPLC/RRLC

[Vlad Orlovsky]

What is typical plate count per meter for UHPLC/RRLC columns. I know it is greatly depends on particle size, compound, system optimization, etc. I am interested to see the average plate count for something like toluene (or any other standard) for 1.7, 1.9 and 2.2 um columns. I looked at few articles and the biggest number I saw was 200K plates/meter for 1.7 um particles. We tried to pack 1.9 um and got about 140K (3.2x50 mm column, showed 6800-6900 plates), but I was hoping for 300K, based on the fact that we pack our 5 um material with 85-120K plates per meter

[Uwe Neue]

Plates per meter? Why not plates per mile?

We pack to between 280 000 and 300 000 (at its best) for 1.7 micron particles. Since you are still green at this, maybe you could be happy with 210 000 plates for your 1.9 micron particles.

[Vlad Orlovsky]

Thanks, Uwe.

I found this article (don't know how accurate it is):
http://www.phenomenex.com/cms400min/lit ... epaper.pdf

According to this reference even theoretical plates for 1.7 um is 260K/meter and for 1.9 um it is 240K/meter, with practical values being at 160K and 130K. We are already there with our 1.9 um silica after 3 attempts to pack with new material (I guees not so "green" after all). If the increase is only 30-50% and pressure goes up 5 times why such a hassle to go from 100K (typical for 5 um) to 160K?? Isn't it better to look for selectivity, gradients, different columns?

[Uwe Neue]

To cite a marketing paper by Phenomenex as a good source for such information appears a bit strange. Maybe you can ask Bryan Evans about the maxium plate count that is achievable with a particular particle size, and how they measure it at his place. Also, the equation in my book was chosen to be rather conservative.
This discussion is rather futile without a solid understanding of all the details that go into the bragging rights of plates per meter.

The other rather fundamental thing about small particles is not the fact that one gets more plates per meter, but rather the shorter time to reach real plates. You can get the same plate count 9 times faster with 1.7 micron particles than with 5 micron particles. THAT is the big deal about small particles. And that is true no matter how you measure the plates.

[Vlad Orlovsky]

Uwe,

I did a Google search and Phenomenex reference came on the first page. I reviewed bunch of articles and posted only one. Supelco references are showing 220-260K plates for 1.7 materials (including Waters silicas). I am simply trying to establish a goal for myself, in terms of plate count, because the first thing you check when you pack column is plate count and symmetry. I can pack any material with efficiency close to theoretical after 5-7 attempts with any material. In three attempts our plate count went from 70K to 140K with symmetry close to 1. So I am trying to see if it is possible to get 250-300K with 1.9 particles or we need to look at HPLC system, column hardware, etc. That's all, I am not questioning quality of Waters columns and was looking for references. Unfortunatle I don't have your book, but will gladly accept it as a birthday present (my BD was 3 days ago)

[bisnettrj2]

Vlad, I'm going to throw one thing in, then get out of the way. I read on the Waters website (High Strength Silica Column Care and Use Manual) that the systems that they perform their column QC on are modified so heavily in favor of removing any instrumental contribution to the plate counts that they are "no longer commercially viable", so you may have to tweak your LC to get rid of any extra-column volume to get the best measurement of your column packing procedure. Oh, and if Uwe is giving his book away for birthdays, mine was two weeks ago.

[Uwe Neue]

Congratulations to your birthdays! I wish you high resolution...

I am sorry, I can't give away my book. For one thing, the information content is so high, I would loose my edge at this forum if everybody would have a copy of it and would know its content. Also, it is a loosing proposition for me. The author of a science book only gets 3% to 5% of the sales price, so I would need to sell a ton of books to simply break even. However, you can talk to your Waters sales rep. Vlad, if you buy enough columns from Waters, you might get a free book from your rep.
Or, you can do it like Phenomenex did: you can buy a copy of my book and increase the credibility of your marketing brochures. But please, be more specific than Phenomenex was about the information they took from my book...

[bisnettrj2]

Well, thanks anyway, Uwe. As a belated gift, maybe you can elaborate on how Waters modifies their systems to perform their isocratic column tests? Maybe Vlad can learn something useful to achieve column measurements that are on par with Waters...

Also (perhaps I should post a new thread), but why do any instrument manufacturers design systems to allow any extra-column band-spreading? Why aren't UV detectors placed in-series with the column compartment with some sort of ZDV-union from the column to the detector cell? Isn't extra-column volume kind of unnecessary with modern columns and instrumentation, especially with the advent of sub-2 micron columns? I'm sure there is a viable, economical reason for current system designs, but it still doesn't make me happy. Sorry, my 1/2 cent, go on with your discussion....

[Bryan Evans]

Hi Vlad -

Imtakt measures column efficiency using semi-micro HPLC.
The link below has data for Cadenza CD-C18:

http://www.imtaktusa.com/site_media/fil ... TI064E.pdf

Values will be dependent (among other things) upon: solute, k', eluent, ect.

If your company is known for 5um, why not manufacturer 3um particles?
The performance is much higher than 5um - and system pressure is conducive to HPLC.