On-line GC - How to setup

[ChemistSA1]

Hi,

I appreciate you taking the time to read this.

I have a 1L high-pressure batch reactor. I'm interested in hooking it up to one of my GC's for on-line analysis of liquid samples.

I'll need some sort of time control and pneumatic actuator, because I want the whole system to be automated.

I reckon I'll need a filter between the GC injection valve and my reactor, and some sort of presusre control built into it too. In fact, I'll probably need a computer for this sort of system.

Have any of you guys done anything like this by yourselves before? I'm not an analytical chemist, I'm chemical engineer.

Cheers guys

[aidnai]

Automation and online sampling are a fairly common undertaking. I have worked with and set up a couple such systems myself.

What make/model is your GC? (also what software do you use to control it?) Is your GC already set up to do the analysis? (i mean appropriate column, detector, etc). Do you have a loop sampler already, or is it just an injection port? Do you have compressed air or nitrogen to run pneumatic valves?

We need to figure out how much work you have in front of you, and what hardware you are missing...

[ChemistSA1]

Automation and online sampling are a fairly common undertaking. I have worked with and set up a couple such systems myself.

What make/model is your GC? (also what software do you use to control it?) Is your GC already set up to do the analysis? (i mean appropriate column, detector, etc). Do you have a loop sampler already, or is it just an injection port? Do you have compressed air or nitrogen to run pneumatic valves?

We need to figure out how much work you have in front of you, and what hardware you are missing...

Thanks for the fast reply!

It's a Clarus 500 from Perkin Elmer. The software is called TotalChrom™.

Yes, the GC is already setup to do my analysis off-line(I have the appropriate column, and a FID detector). I have two injection ports and an autosampler. I have access to both nitrogen and compressed air, aswell as helium, hydrogen etc. I also have a spare PC I can use if needed.

Cheers

[aidnai]

Ok next step -- how do you plan to transfer sample from the sample point to the GC?

If you transfer liquid, depending on the distance and line size, your hold up may be quite large. If you need to use smaller amounts of sample, or if you need quick GC response to changes in the sample, you may want to consider vaporizing your sample and transferring it to the GC in a heated sample line.

A system that I am setting up at this very moment uses a heated sample line to carry condensables in the vapor phase to the GC. We have a 5 meter line and low liquid flow rates (~0.15mL/min) so this adaptation was pretty much forced... I am modifying the GC injection zone (basically adding another oven to the GC ^_^) so I can keep the injection valve, sample loop, and transfer lines warm with no cold spots. We have used a do-it-yourself approach, using thermocouples and heat tape/wire from cole-parmer and using Solo controllers from Automation Direct; there may be a pre-built solution available as well.

Once you get the sample to the GC, you need a pneumatic actuated valve. For a standard set up (no column switching or backflushing) you want a six-port injection valve -- check out VICIs selection here. The valve will use solenoid to control the nitrogen/air and turn the valve; the solenoid is controlled by a relay usually in the GC but possibly controlled by the computer or other hardware.

The valve can replace one of your injector ports or be used in series with one of them. There are advantages either way.

I don't have any experience with the clarus 500, but I have used turbochrom (totalchrom's ancestor) plus external digitizer boxes to do automation of valves, relays, method start, column switching, etc. Hopefully someone else can jump in and give you more specific advice.

Good luck,
aidan

[ChemistSA1]

Ok next step -- how do you plan to transfer sample from the sample point to the GC?

If you transfer liquid, depending on the distance and line size, your hold up may be quite large. If you need to use smaller amounts of sample, or if you need quick GC response to changes in the sample, you may want to consider vaporizing your sample and transferring it to the GC in a heated sample line.

A system that I am setting up at this very moment uses a heated sample line to carry condensables in the vapor phase to the GC. We have a 5 meter line and low liquid flow rates (~0.15mL/min) so this adaptation was pretty much forced... I am modifying the GC injection zone (basically adding another oven to the GC ^_^) so I can keep the injection valve, sample loop, and transfer lines warm with no cold spots. We have used a do-it-yourself approach, using thermocouples and heat tape/wire from cole-parmer and using Solo controllers from Automation Direct; there may be a pre-built solution available as well.

Once you get the sample to the GC, you need a pneumatic actuated valve. For a standard set up (no column switching or backflushing) you want a six-port injection valve -- check out VICIs selection here. The valve will use solenoid to control the nitrogen/air and turn the valve; the solenoid is controlled by a relay usually in the GC but possibly controlled by the computer or other hardware.

The valve can replace one of your injector ports or be used in series with one of them. There are advantages either way.

I don't have any experience with the clarus 500, but I have used turbochrom (totalchrom's ancestor) plus external digitizer boxes to do automation of valves, relays, method start, column switching, etc. Hopefully someone else can jump in and give you more specific advice.

Good luck,
aidan

Hello Aidan,

Firstly, thank you.

Secondly, I have no choice but to use a line between 0.5 and 1 metre in length due to the location of the reactor & GC.

The hold up is not really much of an issue to me. I am essentially analysing something similar to a mixture of benzene and cyclohexane at 30 minute intervals over 3-7 days. There are high boiling point polyaromatic impurities present in the feed, and I need to keep an eye on these in the GC (highest boiling point compound has a B.P of 250 deg.c at atmospheric pressure)

But there is an issue which has just sprung into my mind, namely that the reactor is operating under 20-30 bar of pressure, I guess I'm going to have to put in some kind of pressure reduction line into my setup?

EDIT: I just realized something else, I'm going to have to put a some sort of filter into my system, because I don't want fine catalyst particles (that are created in my reaction from catalyst pellets through mechanical attrition) getting into my GC!!!

Thanks,

Christina

[AICMM]

Christina,

Yes, you want to filter the sample stream, Swagelok makes several really good in-line filters that you can use. I would suggest two sizes, something like 7 micron and 2-micron.

The fact that your system operates at pressure should really help you quite a bit. It might give you a force to push sample from the reactor to the GC. The issue I can think of is your sample in-line returning to reactor and not actually having any pressure drop to move the sample from reactor to instrument and back to reactor. Having said that, I would suggest you use a 4 port liquid sampling valve to make your injection since you are moving liquid samples around (sounds like you are at least.) Using that type of valve, you will only be shooting something on the order of 2 uL or so.

The next question is whether everything will vaporize. Normally what do is connect the 4 port valve to to an injector (splitless/split for example) and this will help vaporize everything. In addition, the drop in pressure will help but this is still something you need to consider.

Unless your batch reactor is running hot, I would be wary of running my transfer lines really hot. You do not necessarily want to vaporize your sample anywhere but in the injection port.

Finally, for safety reasons, consider pressure build ups if you have shut-off valves in your final system and make sure you put some pressure relief systems in to deal with the possibility of vaporizing blocked samples.

Best regards,

AICMM