1.8um silica on conventional LC

[Paulhurleyuk]

Whilst there is lots of theoretical chromatograms with amazing sped floating about from various Ultra High Pressure LC system vendors, I haven't seen many genuine 'in the wild' applications that are that impressive.

As Koen, Tom, Hassel and Bryan are trying to say is that it is very likely that you don't need the extra extra plate count from UPLC, and you could use sub 3 and sub 2 micron particles in a standard HPLC system at a fraction of the cost.

Obviously vendors are keen to sell their ultra high pressure LC systems, but in reality, there's a limited number of applications that can't be done in an extra 30 seconds on conventional equipment. Other avenues you might want to look at are parallel LC where you can do twice the work in the same time span (try here.)

Just my $0.02....

Paul.

[Uwe Neue]

The point of the discussion was the choice of particle size. You will always gain in either performance or run time by reducing the particle size. This is true for a 10 minute separation on a 3.5 micron particle that becomes a 5 minute separation on a 1.7 micron particle. It was also true in the past, when a 30 minute separation on a 10 micron particle became a 15 minute separation on a 5 micron particle. Where you need to be depends on your goals.

I am not advocating trying to do 10 sec separations. I have only demonstrated that it can be done, and how one does something like this, since Yazawa challenged me.

On how do do things with the tools and the equipment that you have, you may want to read some or all of the publications.

[jzt]

As the users, we currently have no need for < 1 min HPLC analysis. We had some debates within our department, some analyst even said “since we run most of our analysis overnight anyway, it doesn’t matter if the sequence is done one hour after I left or by 7 am the next morningâ€

[tom jupille]

I'll inject a comment here based on Koen's earlier post (all of these analogies are based on the North American auto market).

Think of a 250 mm / 10 μm column as a VW Beetle (I'm not talking about the new Beetle, I'm talking about the old Beetle: 60 hp, air cooled engine). Primitive, easy to fix, tolerant of bad roads, amateur maintenance and low-octane gas. Great technology in its day, still quite a few around (at least here in California), but I wouldn't buy a new one today.

Think of a 150 mm / 5 μm column as a Toyota Camry or Honda Accord. Reasonably fast, reasonably comfortable family sedan with automatic transmission and air conditioning. Neither fast nor sexy, but it will run on the highway in comfort, is reasonably economical, uses regular gas, even your teenage kids can drive it safely, and you can get it serviced by almost anyone.

Think of a 100 mm / 3 μm column as a Porsche 911. It's fast, it's sexy, it will spend more time in the shop, uses premium gas, and requires a skilled driver -- and you want to be very careful about potholes, curbs, and steep driveways.

Think of a 50 mm / sub-2 μm column as a Ferrari. It's very fast, very sexy, requires premium gas, and only a trained driver on a closed course can exploit its full potential.

I don't see too much to argue with in the statement that a small column packed with smaller particles will generally provide faster separations than a column packed with larger particles -- if the users of the method have the hardware and the skills to cope with it.

If I'm running on a race track, I want the Ferrari. If I'm driving from San Francisco to Salt Lake City, I want the Porsche. If I'm sending my kid off to college, I want him to have the Camry.

[Uwe Neue]

Tom,

I like your analogy - with a modification. The 10 micron 25 cm column is a horse or a horse-drawn buggy, and a 50 mm sub-2 micron column is a BMW driving on the left lane on a German autobahn. If you drive in the left lane and see a BMW in your rear-view mirror, you get out of the way...

[Koen Hollebekkers]

Haha, Tom
I really like your explanation of it.

Uwe, what I also see on the German autobahn is that this BMW frequently is landed in the crash barrier. Mainly because of speeding……….

But I agree with you both, the advantages are clear.

[anupama]

we were using UPLC/MS & we transferred our method which was on LC/MS .We saw difference S/N ratio was less & we can quantify lesser amout easily.

[sfe-co2]

I don't see where this back and forth argument is taking us. The gist of the matter is simple. UPLC represents a technological advancement (I don't have one ) over conventional HPLC. If the technolohy is suited to your separation objectives, and you have the money to acquire it (now, I'm really jealous), then good luck to you.....go for it.

The 1.7 um particle size has a very flat Van Deempter curve over a wide range of linear velocities (compare that to 10 um particles). Very nice. Uwe, what's the reduced plate height on one of these columns, at 40 degrees Celsius? About 3?

Oh yeah, I'm not associated with any manufacturers in any way. And I drive a Toyota Camry....

[sfe-co2]

No really, no joke.....I really drive a Toyota Camry

[Uwe Neue]

We QC at room temperature. Typical results for the reduced plate height are about 2.5.

[sfe-co2]

Thanks Uwe. Appreciate the response, even if it was just for personal curiosity

[BradVM]

Sorry I HAVE to disagree that lower particle size means a shorter run time. If you refuse to optimize other parameters in your separation simultaneously, then yes, a smaller particle size should yield a faster separation. But once you take a multivariant optimization, yout notice a minimum particle size requirment for a particular separation.

Dr. Chester (from P&G I believe) is working on these types of optimizations. I think he's also an editor for some journal.


T. L. Chester, Anal. Chem., 1997, 69, 165A-169A

T. L. Chester, J. Chromatogr. A, 2003, 1016, 181-193

T. L. Chester and S. O. Teremi, J. Chromatogr. A,2005, 1096, 16-27

Brad VanMiddlesworth

[tom jupille]

Brad, you have to read my entire statement:

a small column packed with smaller particles will generally provide faster separations than a column packed with larger particles -- if the users of the method have the hardware and the skills to cope with it.

(italics added)

You're right: in the real world, pressure constrains what you can do with small particles and things like extra-column volume, detector response time, and data system sampling rate constrain what you can do with small columns. That was the starting point of this thread .

Someone correct me if I'm wrong (it happens ), but my recollection is that pressure was one of the constraints that Tom Chester applied in his optimizations; it's been a while since I looked at his stuff, and I don't recall whether he explicitly incorporated extra-column volume into his model.

[BradVM]

I'll have to look over the paper in more detail. What I remember now was a rather small extracolumn volume (0.001) and some reasonable pressure limit, 6000 MPa sounds familiar.

You're definately right about the italics though. When I read about Dr. Chesters work it was as if he had told me the world was round. I was quite impressed that for a particular separation he was suggesting that 3 um dp would be faster than a sub-2.

Still trying to get over the intimidation factor.