UPLC Column problems

[koen_shimadzu]

This are just compromises you have to make.
As you can pump and inject almost everything what is liquid with conventional HPLC, it is a different story with a tuned UPLC system.

[H]

Regarding backflushing your columns: I also read that you cannot backflush Acquity columns however, after talking with the engineers they told me that it is now actually ok to do this. The reason for the change in attitude apparently is because when these columns first came out they had a wider opening at the entry to the column compared with that at the end of the column and as a result backflushing could result in the column packing material being washed out of the column. Now I am told that this has changed and the entry and exit are of similar proportions and therefore this is no longer an issue. Good Luck!



Interesting what you posted about your pressure increase experiences. Sounds familiar however ours in the past have not stabilised but just kept on going - until the inevitable Overpressure. One thing we have noticed is that if we lower, to say 0.05ml/min to conserve mobile phase when not running or stop the flow, the pressure build up on the next injection is immediate. One idea we are therefore trialing is to keep flow constantly going through the system in an effort to minimise any pressure build up - Will keep you all posted!

[AA]

Just a couple of comments from someone who has seen it all.

1-The inlet and outlet frits of the acquity columns is 0.2 um. This was not always the cas, at the beginning, the out let frit was 2.0 um, so backflushing at that time was bad. So can you backflush, yes, will it help., usually not, the columns are packed to run most effiecently in the direction indicated on the column. In some of the4 experiments that I have tried with backflushing, you get a reduction of backpressure when you turn the column back around, but it never comes back to the starting point.

2- There are 2 tools that can help, the first is the inlet filter assembly, its been out for some time and is just what is says it is a pre column filter, I may be mistaken, but I belive the filter frit in it is 0.2 uM, and is replaceable. Gaurd columns have just been introduced, I saw they a Pittcon last week, I dont know any details about size (ie length), materials, ask your waters rep.

3- The most commom culprit in situations like the one described (that is increasing backpressure with each run untill the column/system finally overpressures) is bacterial contamination from the buffer side of the system. The solutions are just what you might expect, filter your buffer (0.45 um membranes are good enough), make sure your glassware is clean (not contaminated with microbes from the lab next door who you share the dish washer with). Do not leave buffers sitting on top of the system for dayas and days (life will find a way).

4- Other things that I have seen cause similar problems
- material in sample precipitating in the vial when place in cooled sample chambers
-material coming off of cheap nylon membrane filters and making the buffer more contaminated that it started out.
- glass particles shedding from poorly maintained glass frit filtering apperatus.
- silicate materials from dishwashers that are not working (rinsing) properly
- Organic mobile phases causing degradation of parafilm used to seal mobile phase bottles (this one kills check valves too), use tin foil instead (its cheaper too)
- silicate material coming from cheap borosicate glass bottles when using basic buffers (use high quality pyrex bottles), make sure they can handle basic buffers (I have gone to PTFE bottles for basic buffers).
- bad water (this is a tricky one), if you have control of the water used in the lab, invest in the best quality point of use system you can afford. Stay away (far away) from bottled "HPLC grade" water, it is a common contamination problem, very common.

There are probably a few others, thats all I could think of at the moment.

[H]

Just an update concerning our over pressure problems - we seem to have cracked it We looked into the columns that developer1974 recomended earlier in this thread. We decided to trial the Phenomenex Luna HST C18(2) 2.5µm 50x2.0mm as a substitute for our poor performing Waters Acquity BEH C18 1.7µm column.

A few test runs with the Luna proved very positive! The chromatography was near identical to what we had achieved with the Waters column. Backpressure was down to around 4,000psi pmax and we have sucessfully run >1000 injections with no pressure increases whatsoever.

We have spent months in talks with Waters trialing every recommendation they gave us, and a few ideas of our own also.

I would suggest anyone having similar problems with their systems trial these columns on their UPLC and save themselves months of problems.

They are cheaper too!

[Uwe Neue]

You will not get the same level of performance from a 2.5 micron particle and a 1.7 micron particle. This is exactly the same as a comparison of a 5 micron and a 3.5 micron particle. However, if the larger particle serves you well, be happy...

[developer1974]

Hi H, Just to clarify, I did not try the Luna columns (2.5um), I tried the Supelco Asecentis Express (2.7um, with 1.7um fused core). I am getting some excellent results, almost GC type peaks, but with v. low backpressure. I will have a look at the Luna myself, and I recomend you have a look at the Ascentis Express.

[H]

Hi developer1974, I looked at the link that you posted and read all about the Supelco Ascentis Express columns. They sounded good so I ordered one straight away - unfortunately we are still waiting for it.

When I mentioned the Supelco column to my boss he passed on a filer that he had recieved for the Phenomenex equivalent column. I ordered one of those also to try as we have had lots of experience with standard Luna columns in our lab and have always been happy with the quality of chromatograohy we get from them, it arrived the next day

We have had nearly 2000 injections out of the Luna HST on our Acquity system and we are so far very happy with the results we are getting. We will no doubt still try the Supelco column when it eventually arrives. The Ascentis is actually slightly cheaper but we have a good relationship with Phenomenex so may swing slightly more towards them.

Good luck with your system and thanks for your posts.

H

[Bruce Hamilton]

I would be keen to see the results for the Supelco fused-core column compared to the 2.5um Luna.

The technical white paper for the Luna 2.5 claims the advantage ia that 2.5 um delivers nearly 90% of expected theoretical plates, whereas the sub 2um particles only deliver about 60%, so the performance difference is not as great as predicted using just size. I wonder if instrument limitations contributed?.

The paper is available on the Phenomenex site:-
http://www.phenomenex.com/Phen/EM/Pittc ... epaper.pdf

If anybody wants a copy, but can't access the Phenomenx site for any reason, I"ll email the pdf on request.

Please keep having fun,

Bruce Hamilton

[koen_shimadzu]

Sorry to be rude, but I have the feeling that things are mixed up here.

Like Uwe mentioned, it's never possible to get the same results on a 2.5um column as on a 1.7um column.

GC peaks on an LC system is also enthusiastic. This can only be said based on the plate number comparison.


Why using a 2.5 column on a UPLC system? I don't understand it!

[Wayne Way]

Luna HST 2.5 are totally porous particles (like most everything on the market).

Ascentis Express columns have a solid core and thick porous shell (fused-core). They behave differently (high efficiency) than totally porous particles in regards to the kinetic aspects of chromatography.

[H]

koen, Effeciency was never the problem with our 1.7µm column the problem was that we were unable to run our system as we had constant problems with the system overpressuring.

We have tried every precautionary measure in the book to avoid the overpressure of the system but we are running plant extracts and after months of work trying to resolve these issues we have concluded that the 1.7µm columns just have too small a particle size for our application and are the root of the pressure problems. The Acquity columns are only available in 1.7µm particle size thereforewe have little option but to try columns from other vendors that were recommended to us.

Maybe this approach is not the best for everyone, but for us it works.[/list]

[Bruce Hamilton]

Sorry to be rude, but I have the feeling that things are mixed up here.

Like Uwe mentioned, it's never possible to get the same results on a 2.5um column as on a 1.7um column.

Did you bother reading the white paper I referred to?. Nowhere did I, or the paper, suggest the 2.5 um column would exceed the theoretical plates of a smaller diameter particle. What I quoted was that the theroretical performance gain of sub-2um particles, calculated using particle size, was not achieved.

Some numbers ( just visually estimated from the white paper chart, representing at least 3 runs on 3 different columns )
Size, Theoretical, Actual
1.7um = 265,000, 165,000
1.8um = 252,000, 155,000
1.9um = 240,000, 140,000
2.5um = 170,000, 145,000, Oops, what did you say?
3.0um = 152,000, 115,000

Now, I don't necessarily believe these numbers, hence my comment about the instrument perhaps affecting the difference between theoretical and actual at smaller particle sizes.

But they are interesting for anybody who is encountering problems, such as blockage, with small particle columns, or who want to obtain more plates on a conventional pressure HPLC.

In another post, I mentioned that the 2.7 um fused-core technology particles were claiming about 215,000 N/meter. Sure, the testing conditions would be different to the above, but if the above data is correct, there is a lot of merit in using the 2.5um porous ( eg Phenomenex Luna ) or the 2.7um fused-core ( Mac-Mod Halo, Sulpelco Ascentis Express ) on both normal and high pressure HPLCs.

I haven't tried either of these column technologies out, but they look attractive. If the white paper is incorrect, I'd like to see information that demonstrates and quantitates the errors.

Bruce Hamilton

[koen_shimadzu]

Bruce,

I fully agree with the practical part of the story. There are much more ways for fast LC than sub 2um.
I'm defiantly don't say that 2.5um is rubbish. We yourself promote 2.2um columns and this works really well!!

I'm not the person to judge whether the suggested plate numbers are right or wrong.
But imagine this. What happens if you do this evaluation in a system that is not tuned enough?
The increase efficiency of sub 2um column is destroyed by the dead volume. So it looks that 1.7 material is poor material. But this is misleading.

Nevertheless, I think this topic is the best example that it is up to chromatographers to judge what in practically the best solution for them. And I very appreciate your practical look on this case.

[Uwe Neue]

One should be extraordinarily careful with data sets that demonstrate things that are counter to expectation, especially if the data serve the marketing story of a manufacturer. If the extra-column bandspreading on an instrument is large enough, then one will not see a lot of difference between a 1.7 micron column and a 2.5 micron column, which of course will result in the conclusion that there is not a hell of a lot of a benefit to these small particles.

In addition, there are a lot of additional complications in measuring platecounts on very tiny particles. In principle, plate count measurements become meaningless, if the pressure is high enough.

Bottom line: I would not look at a marketing brochure from a manufacturer who has not yet figured out how to make very small particles for solid answers on the performance benefits of UPLC.

[koen_shimadzu]

I don't want to promote the UPLC technology, but uwe is completely right in this.

In addition, there are a lot of additional complications in measuring platecounts on very tiny particles. In principle, plate count measurements become meaningless, if the pressure is high enough

Do you mean radial temperature gradient Uwe?