Waters 2795 vs 2695

[bgiles]

Does anyone have experience using Waters 2795s? I understand that they are "high throughput" systems inted for use with LC-MS. We are having trouble qualifying some refurbished systems we purchased. High %RSD for injections. Wondering if anyone else has had this problem with this model.

thanks

The Dude

[DR]

As far as I know, the plumbing in them is the same. High RSDs typically means either the needle wash is dry/needs priming or the syringe is shot.

[Jason Hendricks]

How high are the RSDs?

Jason

[bgiles]

2-5% using the waters qualification method

[Jason Hendricks]

2 to 5% does seem a little high. In our facility we service both a 2795 and qutie a few 2695s and the autosampler design is different on the 2795 to allow for the high thoroughput so we used to get a higher %RSD on our tests of the 2795. I looked at our data from a few years ago and were getting around 1.25%RSD in partial loop mode (10 uL) and 0.9%RSD in full loop mode (100 uL).

I know that we had quite a few problems in the past with the 2795 and it seemed like most of them related to the autosampler and associated valves.

Without knowing any more it is hard to say what could be going on. Does it pass the compression check and are you sure there are no leaks or air trapped in the syringe?

Good luck
Jason

[bgiles]

Thanks Jason

Not sure the exact details on the testing b/c the systems are off site at our plant. They are sending them to us (RD) because of the issues they had. It may be these paticular systems (they were refurbs) but it seems like there are a lot of refurbished 2795s compared to 2695s out there. Just wondering if this is a system design problem experienced by other users.

The Dude

[Jason Hendricks]

We only had one 2795 compared to 6 2695s so I don't know if it is a trend in the model or not, but we just got ride of our 2795 due to the large number of problems we have had. The Mass Spec group decided that they did not want to spend any more money this time to fix it this time. It is now sitting in storage for spare parts to fix our 2695. I believe a large portion of the problems we had were autosampler related.

[AA]

The 2695 and the 2795 are two very different instruments. Although the solvent delivery is very similar, the autosampler part couldnt be more different. To put it as simply as possible, in th 2695 the needle is part of the flow path which gives very reproducible (low %RSD's) injections. The 2795 is a sip and spit type, that is to say sample is drawn into the needle and then placed into a loop. If everything is not exactly perfect (weak solvent, strong solvent, and sample diluent matching, loop and needle matching, vial type, cap mat type), poor injection reproducibility is the result. The early units were much worse than the later and current units due to some hardware modifications, firmware changes and added calibration routines. If you bought the unit on the so called "grey market" (not refurbished by Waters), it is likely one of those early units and you would be lucky to get less than 2% RSD. A currently produced 2795 should always give less than 1% RSD for area. After all is said and done a 2695 will ALWAYS out perform and 2795 when it comes to injection reproducibility. The 2795 has a lower system volume, better flexability as it can sample from vials or 96/384 well plates. Carryover is often better on the 2795 if you choose the correct wash solvents and wash volumes. As well, injection to injection cycle time is much better on the 2795. Solvent delivery charateristics from both the 2796 and 2695 are identical.

[Newman768]

The injection system of the 2795 seems to be nearly identical with the uplc sample manager system.
From the uplc I know that it is VERY important to degass the weak and the strong needle wash solvent (when using the pressure assisted injection mode). I don't know the 2795 injection system but if the autosampler uses the weak needle wash solvent through the injection cycle for carrying the sample through the tube to the loop this mode is comparible to the so called pressure assisted mode of the uplc system.
To cut the long story short: Use freshly degassed weak and strong needle wash solvents, purge the autosampler and then check if %RSD is as bad as before (hope it's not).
To satisfy my curiosity please give me some response...

[bgiles]

Thanks- I will try this. It might be some time as we are still waiting for our units to arrive. Will let you know once qualification is completed.