Varian versus Perkin Elmer

[Sandrin]

My department is buying GC with headspace for residual solvents analysis, and we have to chose between Perkin Elmer CLARUS 500 GC/PPC with TurboMatrix HeadSpace sampler 40 and Varian 430-GC s combi pall/LIQ&HEADSPACE 2mL&10/20mL TRAY.
Please comment what is your experience with these two producers.
Software - Pharma GLP?
Headspace verus combi pall?
Many thanks!
Sandrin

[MHD]

We have a PE headspace sampler in our lab, it worked well!
Why not try Agilent?

[Sandrin]

The offer was much expensive, is it so much better?

[Jana P]

In our laboratory we have a Perkin Elmer and an Agilent system.

The Agilent system has given us many headaches and has been very difficult for us to use routinely and have confidence in the results.

Our Perkin Elmer system has been running well for a number of years now and i would definitely be looking at buying a Perkin Elmer if a system ever needed replacing or upgrading.

good luck

[Peter Apps]

The CombiPal has the advantage that it can be upgraded to do a wide range of sample handling - at the minimum the configuration that you are thinking of can do liquid injections as well as headspace. I have tested the Gerstel clone of the CombiPal for headpsace, and its repeatability is poor - rsd of approx 2% under strict repeatability conditions, mainly because the temperature control of the sample oven is poor. This was similar to an off-the-shelf Agilent G1888.

I have never used Perkin Elemer instruments.

As ever, the best advice is to give samples to the suppliers and tell them to run them, then compare results.

You should also look at how good the technical support is in your part of the world.

Peter

[unhpian]

We have a Perkin Elmer HS 110 headspace sampler, and have had some problems with it. For some unknown reason the interior of the sampling body rusted (PE tech couldn't find a reason) and had to be replaced. We also have an unsolved problem with cDCE dissapearing, despite numerous visits from the PE "experts".

I've never used the Varian HS sampler, so I can't compare them directly - but the PE HS 110/40 is not without it's own problems.

[larkl]

I've used three different CombiPals. Two of them ran routinely with no problems. The third would break the needle almost daily when it would inject. Numerous service visits later and no fix. We forced Varian to replace it with a Tekmar HT3, as we've had very good results with the Tekmar equipment. This has run quite well. Varian buys the CombiPal and didn't seem to be able to do much to fix it when it wasn't right. Having said that, the other two CP continue to run well with no problems

[chromatographer1]

I have used a Tekmar 7000 (aka Varian Genesis) and a PE HS40 for several years.

Both were reliable and performed well within their design limits. Both were capable of research and routine residual solvent work.

I found the Tekmar easier to service but its design is less complicated and older than the PE unit.

Either choice would be a safe bet, but get the inert package on the Tekmar for better results.

I have no experience with the Agilent unit. I am wary of any HS autosampler that uses a syringe for sampling, that is my issue, not necessarily a problem technically as far as I know.

best wishes,

Rodney George
consultant

[GaryR]

We have three Agilent 6890s with PE headspace samplers. We recently replaced our PE GCs with the Agilent units as we had serious issues with GLP compliance with the PE software at the time. (PE Hardware was great.)
The main advantage to the Agilent GCs was that they are controllable using Millenium software, PEs are not. The pain of validating another software package just for a couple of GCs was just not worth it. (we run >50 Waters HPLCs using Millenium). We also had some issues at the time as to the level of 21CFR part 11 compliance for PE software. (This was a couple of years back, not sure what they're like now)
Can't comment about the Varian GCs as I've not used one since we decommissioned our 3300GC ( what a workhorse! )

[Peter Apps]

Just a comment on the bent needle problem raised by larkl - I had a similar problem with a Gerstel, but only whe I used butyl rubber septa a low (close to room temp) sample temperatures - the vial septa became stiff and grippy enough to bend the needle a little bit as it went into the vial, and then the bend in the needle made it hit the edge of the injector nut and commit autosampler suicide. It was fine with silicone septa, and with butyl septa at higher tmperatures.

Peter

[mcbart]

why not consider Shimadzu 2010 GC + HTA HT200H headspace autosampler?

see application note:
http://www.phenomenex.com/lib/TN_63660808.pdf

or here (pag.14):
http://www.shimadzu.co.uk/Upload/file/news200901e.pdf

[aldehyde]

Varian was purchased by Agilent recently, I don't think anyone is sure yet what is going happen as far as maintenance on existing Varian equipment and what the direction of new hardware would be.

I work on a lot of Agilent headspaces (both the 1888 and the 7694) and I'm not a huge fan--but hear me out. They are good instruments but in the wrong hands they tend to have serious issues.

The solenoid valves for vent and vial pressurization get hit with sample during each injection so they are a common failure. This is not difficult to troubleshoot or replace on your own, but most users don't know how to determine if their valves are working correctly and don't understand what happens when they are broken (either no sample, decreased sample, non-reproducible peak sizes). I have seen people continue to use a broken headspace without realizing what is going on (for months), blaming bad chromatography and tiny, non-reproducible peaks on the instrument.

The motor groups for the belt and the oven carousel can become mis-aligned which can cause sequences to abort half way through or vials to randomly break/get stuck during loading. This is a little more complicated to fix and should be done by Agilent service people (there is a really nice alignment kit that came out recently that can be used to fix these problems so its not a big deal). However, the alignment problems are usually not caused by the headspace but by people putting it on an uneven surface (say for example, on top of two different tables that have a few cm difference in height and are already bowing slightly because they can't support the weight).

These two issues result in about 80% of the headspace service calls I go on. If you are willing to sit down and learn about the pneumatic path of the headspace, how each part works, and instead of just not doing bad things learn WHY you shouldn't do them then you will probably have a very successful experience.

If you have some horribly put together method that contaminates the system during each run and have no idea how the headspace works it won't matter what instrument you buy, you're going to be pulling your hair out.

[aldehyde]

Also, the Agilent HS syringe is pretty securely mounted, and the opening is on the side of the needle rather than at the end so it is pretty difficult to plug or bend it. If this does happen its not too hard to replace.

[sdegrace]

I can't comment on Varian vs. PE. The Varian LC systems are not good, but I hear the GC systems are good. We have had a terrible experience with the Varian Galaxie software, but I have heard reports from GC users that it is not as bad for controlling GCs. I feel that the manufacturer assertation that the Galaxie software is 21 CFR part 11 compliant to be a gross exaggeration, however, and I feel the design is cumbersome and above all very fragile.

If you are considering the Varian, take a cold, hard look at the software if you are required to purchase Galaxie with it. In particular, consider the impact of the fact that the instance window itself runs the instrument and that there is no way to change the logged in user of the instance window, so that actions taken on a particular run cannot be tracked on a per-user basis, and your run is lost if the instance window must be closed, for example due to one of Galaxie's frequent crashes. If this type of audit trail is important to you, then I feel you should avoid Galaxie.

I do have experience with the Agilent G1888A connected to an Agilent 6890 either via a needle into the split/splitless inlet with manual pressure control, or the manufacturer-recommended splicing of the transfer line into the carrier line of the split/splitless inlet with electronic pressure control. The more I use it, the more I don't like the G1888. I find it very fussy and problem-prone. The manufacturer-recommended configuration is terrible, with vapors passing through bare stainless steel through a cool zone that you have to mitigate yourself.

I think the G1888 would be better if connected to the 6890 via the Agilent volatiles inlet, but overall I am not a huge fan of the Agilent equipment. Of the GCs I have tried the one I liked the best was actually an old Shimadzu, and I would be very favourably disposed towards checking them out for new purchases.

[sdegrace]

I can't comment on Varian vs. PE. The Varian LC systems are not good, but I hear the GC systems are good. We have had a terrible experience with the Varian Galaxie software, but I have heard reports from GC users that it is not as bad for controlling GCs. I feel that the manufacturer assertation that the Galaxie software is 21 CFR part 11 compliant to be a gross exaggeration, however, and I feel the design is cumbersome and above all very fragile.

If you are considering the Varian, take a cold, hard look at the software if you are required to purchase Galaxie with it. In particular, consider the impact of the fact that the instance window itself runs the instrument and that there is no way to change the logged in user of the instance window, so that actions taken on a particular run cannot be tracked on a per-user basis, and your run is lost if the instance window must be closed, for example due to one of Galaxie's frequent crashes. If this type of audit trail is important to you, then I feel you should avoid Galaxie.

I do have experience with the Agilent G1888A connected to an Agilent 6890 either via a needle into the split/splitless inlet with manual pressure control, or the manufacturer-recommended splicing of the transfer line into the carrier line of the split/splitless inlet with electronic pressure control. The more I use it, the more I don't like the G1888. I find it very fussy and problem-prone. The manufacturer-recommended configuration is terrible, with vapors passing through bare stainless steel through a cool zone that you have to mitigate yourself.

I think the G1888 would be better if connected to the 6890 via the Agilent volatiles inlet, but overall I am not a huge fan of the Agilent equipment. Of the GCs I have tried the one I liked the best was actually an old Shimadzu, and I would be very favourably disposed towards checking them out for new purchases.