Headspace with Cold Spot

[mbicking]

I have an Agilent 1888 HS unit connected to a 6890 through the main carrier gas line (not through the septum). The method is for methanol and ethanol in pharma products, using a 60 m x 0.32 mm x 1.0 um Wax column, isothermal at 100 C (not a smart setup, but I did not develop this one!). The problem is signficant tailing for both analytes. This method has been validated with the transfer line going through the septum, but that configuration was in another lab.

The problem is a "cold spot" in the transfer line, just as it comes out of the heated transfer line region, through a fitting, and into the port. The tubing is only about a few cm long, but it is cool to the touch. I can eliminate the problem by using a heat gun on the tubing, but this is obviously not a long-term solution. (It's a nice troubleshooting example, and I can post the chroms if you want to see them!)

This is a GMP environment, so I would like to find a simple solution to heating this region, without resorting to exotic fixes. Heating tape comes to mind, but I was wondering if anyone had something a little more elegant.

[Peter Apps]

Hi Merlin

If you really want elegant, and you have a machine shop available get them to make an aluminium block that fits around the cool tubing, with wells for an inlet heater cartridge and temperature probe. With the probes and cartridge connected as a second inlet you can then control (and monitor and record) the block temperature using the GC and software. Beware that this will probably void the warranty on the GC.

Heater tape with some sort of readout and feedback control would work just as well.

Peter

[AICMM]

mbicking,

I agree wholeheartedly with Peter that the aluminum block is the right way to do this. I have done this myself with additional detector heaters (see discussions about FID heat) and it is very straightforward. Unlike heating tape, you can be assured the heat is controlled via a second injection port heater.

Best regards.

[mbicking]

Thanks Peter. I will look into this.

[Peter Apps]

These cold spots are a frequent topic on the forum - remedial hardware might be a market niche for someone.

Peter

[Ron]

It isn't elegant, but often an effective approach is the wrap the injection port, transfer line, and interface in insulation, then raise the temperature of the injector and transfer line to 150C. This is almost always enough to eliminate the cold spot.

[sdegrace]

The insulation trick did work for me, somewhat - the peaks still had a bit worse shape than before I changed over to the EPC setup with the cold spot, but good insulation helped to mitigate it. I wrapped the cold spot in aluminum foil, then in a layer of glass wool, then in another layer of aluminum foil.

I have to say, I am becoming displeased with the Agilent G1888A. First of all, Agilent specifically recommends the EPC setup with the cold spot passing through bare steel in the injection path. This is what their docs specifically say to do. In fact, they will even recommend to you to leave a long stretch of tubing here (a few inches) so you still have room to put on an autosampler tower on the inlet! I did that (although I did not care about leaving room for an autosampler and did want to minimize the cold area, so I made the tubing very short) and I am living to regret it.

However, with the MPC setup, you have the problem of having to balance two carrier sources, and manually calculate your real split ratios. Plus, I find the manual pressure control on the G1888 to be very poor, at least on our unit. It's very difficult to control the pressure, it tends to be sort of all-or-none, and it is ridiculously difficult to reproduce, requiring very broad ranges of split ratio to be examined for robustness and making sensitivity of the method very problematic to predict (the more flow out the headspace the lower the sensitivity because the higher the real split ratio).

So the choice is between that and a wretched jury rig recommended by Agilent, and/or your own wretched jury rig. I'm more and more coming to the opinion that these are just not good instruments.

Are the PE headspace samplers any better? If some day I have the opportunity to influence a related buying decision, maybe I might want to try and get away from Agilent.

Stephen

[krickos]

Hi Stephen

Maybe I misunderstand you, but I thought Agilent "standard" set up was with the so called volatile inlet in EPC mode?

With the volatile inlet set up the transfer line end up with a "bluish" metal connection where the column nut is fixed to (up through the inlet). The connection is then fitted into the inlet and no cold spots there.
We usually have this volatile inlet on the back injector position so we can run normal split/splitless injection on the front inlet.

//Chris

[sdegrace]

The manual for the G1888A headspace sampler gives two configurations for carrier pressure... the MPC setup uses a special deactivated needle to connect the transfer line much like you would the autosampler tower through the septum. No cold spot in the arrangement, but there are two sources of carrier gas going into the inlet, one controlled by a knob.

The EPC setup, as outlined in the manual, involves splicing the transfer line into the carrier line of the inlet weldment and the cold spot is inherent in this setup. If you talk to an Agilent tech this is exactly how they will tell you to hook up the headspace. So, the cold spot is inherent in how the manufacturer will tell you to connect the HS to the 6890 GC if you follow their docs.

http://www.chem.agilent.com/en-US/Searc ... 2&liid=787
[edit: corrected link to installation manual instead of user manual]

Stephen

[krickos]

Ahh

I see, looks "crazy" compared to how we have it (assume page 49 fig 18 is what you refer to).

Compare that to this set up (page 3):

http://www.chem.agilent.com/Library/Sup ... F03052.pdf

[sdegrace]

No one told me you could purchase an inlet like that. That looks fantastic. Now I want one!!

Stephen

[mbicking]

I have used the older Agilent HS unit, which also had some very bad valves for controlling loop flow and vial pressure. This system had the transfer line going in through the septum.

I opened up the manual valves and then connected the loop and pressure inlets to the Aux outputs on the GC. This allowed me to set the pressures on the keyboard/software. It made the whole system much easier to control. But as noted earlier, the extra gas from the HS goes out the split vent, and increases the split ratio. You need to manually check the loop flow via the split flow to control the split ratio.

I would like to get more info on the volatiles connection. Anyone have a part number or link?

[krickos]

Hi

This pdf gives you an overview of the flow part to the inlet:
http://www.chem.agilent.com/Library/bro ... 6491EN.pdf

this gives you a part break down overview:
http://www.chem.agilent.com/Library/Sup ... a15841.pdf

[sdegrace]

I'd suggest just calling Agilent and getting a quote - I got an informal quote of a little over $5000 Canadian. They recommend you get one of their techs in to do the install, which unless you're very comfortable taking apart your GC may be a good idea.

Stephen

[Bigbear]

I do purge and trap with a 6890 and cut the gas supply line to the injector and plumb the P&T in the loop in series. The desorb ( sample transfer) pressure and time are then controled via the GC.
Are you certain it's a cold spot? I see you used a heat gun, but have you lowered the column flow as 100C is quite hot for those compounds. I do them @ 35C