GC choice & infrastructure requirements

[AlexW]

Hello,

My company is planning to get a GC for general chemical analysis (reaction monitoring etc), and I'm trying to find out what sort of machine would suit us and what we need other than the instrument itself. My only experience with GCs is as a routine user, so some of these questions are probably rather basic...

We're looking at getting a second-hand GC-FID (MS would be ideal, but rather too expensive). My questions are:
1) Do you have any particular recommendations for a (cheap!) basic GC model? I'm getting mixed messages about the HP 5890 - on the plus side, it's a standard and reliable, but no longer supported so breakages might be rather terminal. Obviously I don't want to be responsible for something which dies after a year or two.
2) What additional infrastructure is required? I know gases (air and carrier, with oxygen/moisture traps, plus hydrogen which will have to be housed outside) and a dedicated power supply, but is there anything else? I assume ventilation is not essential since there's nothing nasty in the exhaust from the FID?
3) Does GC choice restrict which columns can be used, or are they all a standard fitting?

Thanks,
Alex

[skunked_once]

Here is the first set of opinions on your questions.

1-Yes the 5890 is no longer supported but it is a very reliable instrument and there are several companies that still provide aftermarket support. If you can find someone in your area that offers such service, a 5890 could be a workable option. On the other hand, the newer 6890 models are starting to show up on the used instrument lots so prices should start to go down.

2-Be sure to use a good grade of pure helium for carrier gas with the appropriate filters as you mentioned.

I assume ventilation is not essential since there's nothing nasty in the exhaust from the FID?

I wouldn't be too sure about that. What is in your samples? Are there possible noxious or corrosive products from the FID combustion process? The instrument should be in a well ventilated room. A room with stable temperatures and a lack of strong drafts is best.

3-You should be aware that the inlet/outlet on the instrument is designed for the type of columns you plan to use. You can't use a packed column on a capillary GC inlet for instance.

Post here for further advice.

[aldehyde]

Yeah I wouldn't recommend anything below a 6890 for a new project like this. A 6890 in good condition has many years of service in it, parts are plentiful, and has many options of current software (usually controlled by chemstation or empower).

Are you monitoring reaction conditions in the gas phase, or will you be doing liquid injections?

What kind of compound is it, you mention FID. If you want flexibility you should go with a 6890, if you plan on doing something that would only require one inlet/detector combination a 6850 is cheap and doesn't take up much space. It can be expanded to use headspace and mass spec so its still pretty versatile. They're still pretty easily serviced for a small instrument.

[AlexW]

Thanks for the replies.

We'll be using it for liquid injections

What kind of compound is it, you mention FID.

We'll be using it for a variety of compounds - mostly hydrocarbon / alcohols, and some silanes. I'm only familiar with FID - I thought the other detector types were more specific, e.g. NPD for P and N, ECD for compounds containing heavy atoms?

You should be aware that the inlet/outlet on the instrument is designed for the type of columns you plan to use

I don't think we're likely to be using packed columns, so would a split/splitless type injector be best?

For the ventilation, we won't be using anything particularly horrible. The nastiest is likely to be some thiols which will certainly make themselves known if they escape! The room already has extraction, which could be connected up, but it would be nice if this could be avoided. I guess we'll see whether it's needed when in operation, it could then be fitted.

There are people who will service 5890s round us, so that's definitely a possibility. Hopefully we'll find a complete system so won't have to worry about getting mismatched parts - I've seen 6890s (no workstation) for around £4000, so I'll push for that option...

Thanks again!

[skunked_once]

Alex,

It looks like the FID is a good choice for your applications and yes a split/splitless injector is the way to go. The decision to find a used 6890 is also a wise choice. Good luck with your GC setup.

[larkl]

Be careful, as most of the sample will be exiting the split vent if you run in split mode. If the materials are noxious or toxic, I'd route both the septum purge and the split vent lines to a hood or other appropriate venting location.

[JTM]

I'd say go with the 6890s or an SRI box.

I really like working with SRI, as well.

[AICMM]

AlexW,

I am going to swim upstream here and enthusiastically recommend a 5890. Tons of parts available (at least in the US), very solid chassis. The main part that dies is the heater assembly and you can buy these from Restek. Everything else, I can list two or three companies that can provide parts. Everyone one I work with is still using them.

This recommendation is especially true if you don't need to get really exotic about your chromatography. If you don't require EPC or Dean's switching (for example) it will do everything you ask it to.

The bigger question is how do you need a GC configured. I would put some thought into that before sitting down to purchase.

Best regards,

AICMM

[AlexW]

Thanks for all the advice. In the end it looks likely that we'll go for a 5890 on price grounds. It turns out that the main cost is likely to be software - refurb 5890s often come with computer & chemstation or similar, which saves a lot compared to buying new.

I'll keep my eyes peeled for a 6890 though...

Time to do some reading on what sort of configuration would be best for our chemistry then!

[Consumer Products Guy]

We're a "big" company with 4 GCMS and 4 GC units; three of the GC units are 5890, all computerized.

Yes, they're out of service by Agilent, but there are zillions out there, still running great. My recommendation is to be sure that it contains the HPIB card already, and hopefully includes HPIB/GPIB card for the computer, like a 82341 or 82350, as the earlier 82335 board only fits in older model computers.

As to software: Agilent made available to us free Y2K software in late 1999, for ChemStation A.06.03, which we still use on a few units (with NT4, Windows 95, Windows 2000).