Agilent 1200 (RRS), Waters UPLC, or ThermoFisher Fast System

[Euro08Soccer]

Here in Germany the most common LC for an ABI Mass Spec come from Shimadzu at the moment.

The facts that convinced us:
The autosampler shows reproducibility (typically we find 0.1% RSD @ 1ul at cycle times of 1min). The sales guy showed me an application where they did UV runs at 25 seconds cycle time. But peaks become too narrow for satisfying MRMs then.

Best thing is: the system is upgradable. 30kE for the frontend, runs at pressures up to 300 bar or so. If we decide, that we need 650 bar, we can upgrade the system for less than 10kE. And it plugs directly into Analyst. No hassle with drivers developed by the lc vendor.

The Drawbacks at other systems, that I know of:
Waters injector has a lot more carryover, and less precise injection. Limitation of flow. At the moment we can go up to 2x10ml /min. Should we install the higher pressure option, this decreases to 2x5.

Thermo: correct me if I am wrong: their Pump seems to be a low pressure side mixing system. Kicked out of our considerations immediately.

Agilent: Only one dampener. If you want to go for low gradient delay you have to use an undampened high pressure gradient flow.

[Hfranz]

But peaks become too narrow for satisfying MRMs then.

For my understanding: What do you mean with MRM?
Do you know what the pressure upgrade kit includes?

Agilent: Only one dampener. If you want to go for low gradient delay you have to use an undampened high pressure gradient flow.

Well, for high-pressure analysis, Agilent takes the pulse damper out of the flow path...

[Euro08Soccer]

MRM means Multi Reaction Monitoring for quant purposes. As you have to select a precursor ion, allow time for fragmentation and then mass analysis again, peaks should not become too narrow.

What the Shimadzu upgrade kit includes: I'm not sure anymore. I received that information at analytica trade fair. I remember, that their name for the solution was UFLC XR (a label on the components, printed in orange)

HFranz: Whats the pressure limit og the Dionex software for Shimadzu systems? Does the upgraded system work with it?

[Stryder08]

MRM means Multi Reaction Monitoring for quant purposes. As you have to select a precursor ion, allow time for fragmentation and then mass analysis again, peaks should not become too narrow.

What the Shimadzu upgrade kit includes: I'm not sure anymore. I received that information at analytica trade fair. I remember, that their name for the solution was UFLC XR (a label on the components, printed in orange)

HFranz: Whats the pressure limit og the Dionex software for Shimadzu systems? Does the upgraded system work with it?

Sharp, narrow peaks are good! Who told you that narrow peaks were a bad thing?

[Hfranz]

HFranz: Whats the pressure limit og the Dionex software for Shimadzu systems? Does the upgraded system work with it?

Chromeleon Software does not introduce any pressure limit. This is purly caused by the instrument limitations.
For the upgrade: I guess they use a different firmware at least for the pump. The compatibility with the driver would then need testing.

[mbicking]

A more general question you should consider is, "do I really need 'UPLC' for my separation?"

All of these systems should work well within their stated specifications, but do you really need that much separating ability? If you have a very complex sample (100 - 150 components) and you need a good separation of all/most of those components, then you probably would need a long column packed with very small (1.8 um) particles. These columns require very high pressures.

But if you only have a few components, you can get a very good separation in just a minute or two using a shorter column packed with 3 um particles. This can be done on any good quality HPLC. You don't need to spend extra money on a high pressure system.

[Euro08Soccer]

Sharp, narrow peaks are good! Who told you that narrow peaks were a bad thing?

peaks should only be so narrow that you obtain an acceptable number of data points. When you have a peak width of 1second, this is not possile for more complex MS experiments.

This can be done on any good quality HPLC. You don't need to spend extra money on a high pressure system.

My point exactly.
But it's nice to know, that you can upgrade to the higher pressure tolerance if necessary. So you can compensate for the disadvantage of higher backpressures when increasing length or decreasing particle size.
My favourite particle size is 2.5micron since you can obtain it from numerous vendors. And we all know, that C18 from brand A is not exactly the same as C18 from vendor B (or W )

[Stryder08]

Sharp, narrow peaks are good! Who told you that narrow peaks were a bad thing?

peaks should only be so narrow that you obtain an acceptable number of data points. When you have a peak width of 1second, this is not possile for more complex MS experiments.
- )

You need a very fast MS. Dwell times such as 0.005 seconds. Scan times such as 0.15 seconds.

[Euro08Soccer]

0.15s per data point is not really suitable for fast lc.

@stryder:
In another thread you have stated, that you consider your Shimadzu LC unreliable as a front end. Why is that so? (Which model do you use? what are the precautions you would advise?)

We have made completely opposite experiences. Only negative point in one year: at the beginning there was trouble with the check valves until we changed to another Acetonitrile brand.

Many good points though: For example the rotor seal of the injector does not wear off easily. If I remember correctly, it is specified with 100 000 (!!!) rotations according to the manual. What is that figure for the Acquity? No one from the waters sales team seemed to have that information at hand at analytica trade fair nor can I find it on their web site.

[Elmiko]

I still see a lot of questionable marketing hype in the UPLC and other ultra-high pressure LC claims. There's a lot of empirical evidence that shows that when you go to particle sizes below 2.2um or so, your increases in efficiency drop off rapidly, but back pressure increases dramatically. You can achieve virtually all of the speed improvements UPLC claims to deliver on conventional systems at reasonable pressures (<5000 PSI). Phenomenex has a pretty good white paper discussing their Luna packings on their website. While it may be a marketing tool by the company, it contains some eye opening facts and data which made me think twice about UPLC and similar systems.

[Uwe Neue]

Elmiko:

One has to be really careful with data generated by people who do not have a UPLC instrument nor the capability to make sub-2-micron columns.

A careful look at the Phenomenex data shows that they have been obtained on HPLC instruments. We know that the bandspreading of an HPLC instrument is insufficient to measure the quality of a sub-2-micron column, and on a HPLC instrument one will not be able to see the performance difference between 2.5 micron particles and sub-2-micron particles. Looking at the reduced Knox plots, it is also clear that the sampling rate of the instrument was not set sufficiently high.

To develop VHPLC or UPLC instrumentation and columns was not a trivial effort, and no company in the entire world would have embarked on this endeavour just for blowing some marketing bubbles in the air. But there are several companies that are offering these instruments, and I expect that there is more to come.

So all those skeptics: I recommend to be cautious with data set proclaim that UPLC is a fake.

[Elmiko]

I didn't say fake-just questionable.

[Bryan Evans]

The real benefit to using the new U-HPLC systems is the reduced
extra-column volume and dwell volume. For bioanalytical assays -
I would also look into the autosampler design and see what the instrument
manufacturers are doing to reduce autosampler carryover.

Below are HT applications using Cadenza CD-C18 and Cadenza CW-C18
(both 3um silica particle):

http://www.silvertonesciences.com/files/TI276E.pdf
http://www.silvertonesciences.com/files/TI385E.pdf

Imtakt can pack 250mm & 500 mm (3um) columns - generating
50,000 & 100,000 plates, respectively. The pressures are conducive
to conventional HPLC systems.

[gouki]

Here's my 2 cents on the topic.

I've only experienced agilent 1200 and waters acquity and therefore am unable to comment on the last manufacturer.

Waters

PROs

1) Excellent mixing by pumps (especially when mixing 100% MeCN with Buffer) minimal baseline disturbance observed @210nm
2) Using empower full access to the system is great

CONs

1) Changing and re-characterising sample loops felt like a waste of time
2) Injection carry-over observed
3) Errors observed when using the sample manager system
4) Cannot fit standard HPLC columns (with large ends) easily into the column heater chamber

AGILENT

PROs

1) Injector reproducible and capable from 0.1uL to 100uL
2) No carryover observed
3) Sample tray has a simple layout.

CONs

1) Pump isn't great at mixing (baseline wobble)
Agilent
2) Column compartment has flimsy fittings on the preheater which can break.

After looking at both bits of kit, I decided to go back to standard HPLC.