UPLC

[Uwe Neue]

I did not want to chime in on the subject, but Chris's last post makes it inevitable.

I argued myself exactly as he did, when I killed my first 3 micron column (in maybe 1984 or so) within the first 24 hours of just standard use. The application was trivial, not even real samples, just standards!!! The column was bought from a competitor, before my company had 3 micron columns. My somewhat premature conclusion was that 3 micron particles won't go anywhere. I was wrong! As we all know, 3 micron columns are today in routine use. Properly cared for, they last forever (I have data to prove this).

The exact same thing has already happened with the sub-2-micron-particle columns packed for UPLC. Initially, within the use stages of early adapters and adopters, the column lifetime was often short. But significant improvements in column design and manufacturing have eliminated these initial difficulties already.

At this point in time, the issue is quite different. If it took you 15 minutes to do an analysis with a classical particle size and column, and you spent roughly 5 weeks on this before the column died, then your column lifetime was 5 weeks for 1000 analyses. With UPLC, you can do this analysis in 3 minutes, and the same number of analyses (1000) is done in one week (without overnight runs). If your column lasts only for a week now, I don't think that you can blame the column, nor do you have any reason to complain in general.

At the same time, there is always room for further improvements...

[A.Nonymous]

Chris,

indeed it's sample preparation related.
Never heard about the 10% rule, but it seems logical.
Although we use an inline filter (even for HPLC) 0.2µm, the columns (inlet frit?) get plugged.

And as the waters service engineers say:
If you have 1,000 injections on a 5µm column, don't expect 10,000 on an Acquity!

I think the sample preparation is gaining importancy, so the time you win with UPLC maybe (more or less) consumed by sample preparation steps.

This is a balance which is different for any application, and no other but you can make this up.

[Chris Pohl]

Uwe,

I agree that many factors play into how many injections one can make before a column is plugged including: the stationary phase particle size, the particle size distribution, the column length, the mobile phase linear velocity, the particle size of sample contaminants, the size distribution of these sample contaminants, the concentration of the sample contaminants, the injection volume, whether or not the sample is filtered prior to injection and the type of filter used, the rigidity of contaminating sample particulate, whether or not centrifugal sample prep is used, to name just a few. Nonetheless, it is inescapable that more care must be applied to prevent column damage as the particle size diminishes. This is the reason why columns based on 5 µm media are still very popular. The level of care needed to prevent contamination is significantly less, all other parameters being equal. This does not mean that small particle size media isn't useful. But it does mean that there is a price to pay in terms of cost per sample and ease of use. Whether these costs are warranted depend very much on the analytical situation.

A.Nonymous

I agree that if you get 1000 injections with a 5 µm column you shouldn't expect 10,000 with 1.7 µm media but this doesn't go far enough. You shouldn't expect 1000 injections either.

[sksrinivas]

Hello All

Been quietly following this particular thread for a long time. Can't resist the temptation to throw my hat in.

I live in Bangalore, where Waters India has its HQ, by the way.

Several of my clients have been asking me "Do or don't we purchase a UPLC?".

This question also arose during a recent training program I conducted here.

And to all of them, my response is the same "Why do you think you can't live without a UPLC?."

The answer to that question depends on what the user is actually looking for, and what he/she is willing to pay for it.

If run-time is a major worry, then I recommend that they first take a close look at their existing methods and see if they can improve on them - mobile phase changes, better column, flow rate, etc..

More often than not, their problems lie in the method - not the technology.

What if the method is OK and higher sample throughput is the concern? In this case, I usually tell them to consider buying more HPLC systems, or an autoinjector, or even consider a column switching system for their existing HPLC.

From what I know of UPLC prices, I think one can get two HPLC's for the price of one UPLC. (as far as I know).

Further, the price one pays for adoption of new technology is not just the initial capital cost. It's the price of changeover as well - in terms of operator time, revalidation of SOP's, spares, re-configuring the lab, etc...

There is another important aspect to consider - is adoption of the new technology a regulatory requirement?

HPLC is mandatory in every pharmacopoeia. Had it not been a mandatory requirement, it would interesting to see how many pharmaceutical companies would buy an HPLC system.

This is why new technologies like CE, monoliths, and the like, never really took off in India.

CE was first launched in India by Beckman with great fanfare, sometime in the early '90's. Sixteen years later, the number of CE systems in India can be counted on the fingers of one's hand.

Monolith columns - same story in India. Are monoliths "better" than packed columns? Debatable, arguable, too early to tell. Will they replace packed columns? Not yet. Are monoliths a regulatory requirement? No.

So goodbye Mr Monolith Vendor. Nothing personal, but sell me something else.

Ditto with UPLC / nanoHPLC / call it what you will.

The technology is new, it's exciting, and it comes to us from some fine companies. It's something we'd all like to have.

BUT - Will your lab/research/product development collapse without it?

Can you afford it? Can you afford to maintain it? Do you really think savings on solvents and run-times will offset the other costs of owning it?

Is it a regulatory requirement? Will you lose your USFDA/GLP certifications without it? Will your quality auditors send you a warning letter, or are they happy with what you already have?

I'd say these are the questions you need to ask yourself, take a hard look at the answers, and then decide.

Personally, I'd love to own the latest, the newest, the trendiest analytical systems that money can buy. But... there's always a but.




Just offering my opinion on the subject. No prejudice to anyone.



Warm regards to all.

SKS

[Rafael Chust]

Would you mind that I copy and translate your text to present to my customers and colleagues?

I fully agree with everything you said!

Regulatory affairs in general are a PITA for vendors, and if they want a new technology to be spreaded in small markets like Portugal, they should make an effort in the larger markets (say US, UK and Germany) to make their option as "the one"!

Otherwise, all these new technology it will just be a toy for rich labs!

[sksrinivas]

Hello Rafael:

Thank you for your comments. You can copy and translate my text, no problem.

Warm rgds,

SKS

[AA]

sksrinivas


If we all had an attitude like that we would still all be using chart recorders and cutting and weighing out the peaks.

AA

[Uwe Neue]

I may need to comment on several items on Srinivas post, but let me first of all address the issue of regulatory compliance, since this is important.
From the standpoint of the basic technology, UPLC and HPLC are the same thing. If you can get a chromatographic separation with an HPLC method, you can get the same separation with an UPLC method. The only difference is that UPLC is faster.

To drive this point further: the US Pharmacopeia is in the process of changing the definition of the particle size range in their relevant L categories of columns down to 1.7 micron. The way implementations work, they will first be published in the Pharmacopeial Forum, and then they will be implemented in the USP book. I have not checked if it has been published in the Pharmacopeial Forum. If it has not been done, it is in the works. The plan is to implement these changes in the next issue of the US Pharmacopeia, which is due to be published in Feburary next year.

So there is not an impediment to UPLC from the regulatory standpoint.

[Uwe Neue]

I just checked: the use of 1.7 micron particles has been incorporated in the Nov/Dec 2005 issue of the Pharmacopeial Forum under the subject of the proposed revision of the Chromatographic Reagents.

So it is on its way towards implementation, and there is no reason dismiss UPLC because of the erroneous idea that it might not be compliant with the regulatory bodies.

[Uwe Neue]

Now that we clarified any issues about regulatory concerns, let us go back to the current status of the technology.

By now, UPLC is not brandnew any more. It has been around for a while, and while this was happening, the technology matured already. A second generation of detectors is available, and many small changes have been made that have improved the reliability of the technique. It has grown out of its baby shoes by now. Many people are buying it: I walk by the packaging line every morning, and these things are growing like mushrooms and leave the plant quickly. So I suppose that somebody is buying these things...

Those that are aware of the technology may have also noticed that Agilent has just introduced their own version of a UPLC instrument with an increased pressure capability and with a strong focus on their version of the smaller particles. Note that they are also not continuing their older 1100 technology. Now the two largest players in the HPLC field are both offering aspects of the higher-pressure technology. From my standpoint this means that the technology of higher pressure is here to stay.

In addition, the argument for the technology is rather simple: same as before, but now either better or faster or both.

[Rafael Chust]

I think we are seeing the same stuff under different points of view.

I am sympathetic with Srinivas stand-point, as we are talking about countries with low budget to afford new systems and new technology development.

I live in the US for a couple of years and know the local reality. But a Waters column (let's say Symmetry C18 4.6 x 150 mm) in Portugal is sold at roughly 700 € + 21% VAT, meaning over $1000 cost (through a distributor...!

If there is no "regulation" demanding the use a Waters column, believe that everyone would buy a Lichrospher column for 1/3 the price from VWR local subsidiary...

I think when me and Srinivas talk about regulation is not concerning technique, but if it is mandatory or not... If it is, them it would become a costy commodity, otherwise, it is a luxury stuff!

Back in 1985, Waters had 7 (yes, it is the correct number) HPLCs in use in Portugal. In 1987, they were roughly 40... And I know today labs with 8 to 10 Waters HPLC working, besides the Agilent and others...

I have no record of any UPLC working in Portugal - besides a demo unit at the distributor premises. Considering the cost, I think it will take a little bit to have them around here!

[Uwe Neue]

Distributors have different price stuctures. I do not know what the price of a UPLC is in Portugal.

[juddc]

Having just this week completed the purchase of an HPLC, I can say that the difference in price between HPLC and UPLC was not high on the list of deciding factors in the purchase. I could very well have gotten a UPLC in for not too much more than I spent on the HPLC I purchased.

The reason I went with the HPLC is that I need flexibility both in separation mode and detection more than I need either high sensitivity or throughput. It's as simple as that. I need to be able to run my conductivity detector, RI, PDA, fluorescence detector, ELSD, and electrochemical detector and at the moment, I have no good reason to toss them all into the trash. Also, if I buy one UPLC, I'd need to buy 2 because there'd be no point in my developing methods for QC that they couldn't use.

Instead, I purchased a high quality HPLC with a relatively sophisticated software setup (Empower w/ PDA, SysSuit, GPC + AMDS) where I'll still be able to use my old detectors to good effect and do tons of good chromatography for at least the next 5-10 years without worry.

If you need the efficiency of UPLC, you'll be well served by buying one. The price difference isn't that huge. Heck, if your bottom line depends upon throughput, UPLC is probably cheaper in the long run.

IMHO, as they stand HPLC and UPLC are slightly different tools with different strengths and weaknesses. As UPLC becomes more flexible, HPLC may fade, but it will be a while.

I think the old hot-rodder's adage applies: Speed costs money... How fast you wanna go?

[sksrinivas]

Hello All

Thought I'd add my 2p worth - again. At the risk of stoking this UPLC/HPLC issue further !!

Recently I attended a seminar on HPTLC, hosted by Chromline, Indian dealers for Desaga. And someone in the audience asked a question that is analogous to the HPLC/UPLC debate.

The question... is an "old" technique like TLC still relevant today? After all, how sensitive can an HPTLC densitometer be? Nowhere as good as PDA. How many theoretical plates can one get from a TLC plate anyway? 20,000 tops?

And the price of an HPTLC system is more than a comparable HPLC system. Isn't it about time we dumped this medieval chromatographic method?

The answer from the speaker ... Labs are buying far more HPTLC systems today than a decade ago. During the 70’s and 80's, when HPLC was being touted as the new-age "replacement" for TLC, HPTLC vendors were a worried lot. Today, they are hard pressed to keep up with demand!

And he’s dead right.

The reasons are many. HPTLC scores over HPLC, in many analyses. Sample throughputs are much higher. One can apply 25-30 samples to one single TLC plate. Solvent consumption is a few ml. Costs-per-sample are laughable. Sample prep requirements are not as critical as in HPLC. I wouldn't hesitate to apply an extremely dirty sample to an HPTLC plate.

No expensive columns, no injector seals, no check valves, no flow cells, no degassers, no guard columns to change. No fancy training. Cost of ownership – negligible, compared to HPLC.

LOD's are comparable, precision comparable, sensitivities comparable to HPLC. Regulatory compliance is no issue, and that includes 21CFR-11.

So much so, a customer would be caught in a serious dilemma, if he had to choose between HPLC and HPTLC.

So what's the point? Point is, you can't "replace" one form of chromatography with another.

When diode array was developed, people said UV detectors are finished. Are they?

When tunable UV's were developed, people said fixed wavelength UV's are finished. Are they?

When SFC was developed, people said HPLC is finished. When HPLC was developed, people said HPTLC is finished. When HPTLC was developed, people said TLC is finished. When TLC was developed, people said paper chromatography is finished.

And now it's UPLC. And 1.7 micron. Tomorrow it will be atomic LC, and 1.7 Angstrom, for all you know.

On the backs of big fleas....

Has Whatman stopped selling chromatography paper? Merck stopped selling TLC plates? Stopped selling TLC silica gel? There are 20th century dinosaurs out there who still make their own TLC plates. I'm one of them.

I'm equally at ease with Gilson's 215, Shimadzu's SIL, Waters Empower, Camag's CATS - and I still make my own TLC plates ... just like any other chromatographer, now or in the future.

If you visit a typical, USFDA-approved, Indian chromatography lab, you'll see glass columns, tlc plates, autoinjectors, UV detectors, DAD's, peristaltic pumps, nano LC pumps, chart recorders and CFR-compliant data systems, all merrily rubbing shoulders with each other.

It finally comes down to - what's right for your sample. What's right for your budget. What's right for your work.

Your decision. Your choice.

That’s why I still tell my clients, if your analysis requires nothing more than a fixed wavelength detector, nothing more than an isocratic pump, nothing more than a data integrator – then buy nothing more.

If your analysis requires an LC-MS, with ESI, octopole, multichannel detection, 21CFR software, then buy nothing less.

Enough said, I think?

Warm rgds,

SKS

[H Man]

As a manager of a small lab (as well as the research scientist, secretary, janitor and plumber), fast turn-around-time is very important. We use LC-MS/MS for its high sensitivity, specificity and of course - speed. We can't afford false positive e.g. isobaric compounds that has the same retention time as the analyte of interest (we try to exclude false positive by fragment m/z ratio).

While respecting TLC for its high-throughput rivaling or surpasing HPLC in certain aspects, i don't believe that there is many TLC capable of reaching sensitivity in terms of sub-pg/mL level, as well having the capability in resolving isobaric-structurally-related compounds as aforementioned)...

We just bought a UPLC, with high chromatography resolution, very narrow peaks hence high S/N, coupled with it's high-scan rate UV detector, low-ng/mL quantitation is now a reality - without derivatization.
However, our LC-MS/MS (Ultima) is an older machine, we can still use it for ordinary LC works - running 1 to 4.6mm columns with normal back-pressure i.e. 20s to <400bars without problems. We particularly like the fast injection cycle (fast and efficient injector needle washing cycle). So, FOR NOW anyway, injector related carry-over has not been observed! Very impressed, thus far.

My complains about UPLC?
- Very small column compartment. Limited to 150mm length. It's design appears to exclude certain kind of columns e.g. Many Phenomenex columns i.e. Gemini and Luna are impossible (missed by several mm) to fit into the compartment... this is very suspicious... I find it hard to believe that having designed a complicated piece of UPLC, Waters' engineer failed to design a wider column compartment.
- Wash solvent switch-over takes 20-30mins to complete! Eventhough it is a push of a button, still having to wait for 20-30mins...
- Very expensive Acquity columns!
- (Minor complain) Flow rate limited to 2ml/min.
- Buggy Version 1.2 software, more works needed. Stick to Version 1.1 - largely problem free.

In summary, after using UPLC for a month, we find it hard to go back to ... say Waters 2795XT (which is a piece of robust equipment). The following reasons convince us of the benefit of using it:
The ease of use, fast injection cycle, very low injector carry-over (none so far), able to do ordinary HPLC works, capable of fast analysis and great resolution when used with Acquity columns, great sensitivity (ng/mL) possible with UV detector.

The above is just my personal experience. Hope it helps.

H Man