GC - PDD issues

[Booth]

Hello all,

I'm new to the forum and I'm looking for some help, hopefully down the line I'll be able to provide some assistance to you guys as well. I've been assigned the task of installing a PDD detector into one of the GC's in my lab. Its our first one, so I don't have many other people to ask for advice. The PDD is a VICI / Valco model PD D3-I-7890 and the GC is an agilent 7890. I have a Restek RT Msieve 5A column (30m, .53mmID, 50um df) and a gcms leak tested valco sampling valve. My helium supply is running through a purifier and it feeds my GC and my detector. I have helium leak tested all of my connections and found nothing to indicate a leak.

The problem I'm having is that when I begin a run, my areas around oxygen and nitrogen are elevated and the peaks are on plateaus in the baseline. After the run ends, I am seeing 2 peaks come out, which I am assuming are oxygen and nitrogen (air)? These are extra peaks (not in my sample) and they always come out after the run has finished regardless of how long or short I make the run.

Has anyone else encountered this issue with a PDD, or does anyone have an idea as to why I'm seeing these plateaus?

My target mix is 1ppm Hydrogen, Nitrogen, Oxygen, Carbon Monoxide, and Methane balanced in Helium.

Many thanks

[CE Instruments]

The two peaks that always occur at the end of the run would point me in the direction of the GSV. Is it manual or air actuated ? I'm guessing air actuated and it returns to the loop fill position at the end of the run ? What happens if you actuate the valve do you get these peaks ? It could just be the disturbance in the gas flow. Have you cheked the valve is plumbed right ?

[AICMM]

My bet is with CE Instruments. You might consider a purged valve head.

Only other thing I can think of is that your EPC for the system is leaking on the internal seals but this should really give you an elevated baseline rather than a peaks. What's your zero at?

Best regards,

AICMM

[Booth]

I've double checked a bunch of things and I'm still having issues. To answer your questions:

Baseline is around 950 eV (VICI/Valco says it shoud be btwn 600 and 2000)

The valve is typically controlled by a nitrogen powered actuator. valve switches from "loop fill" to "inject" for 5 seconds at the beginning of the run and then switches back to "loop fill". I have been performing manual injection without the actuator to make sure that it wasn't causing the peaks. The results look the same when injections are performed manually and with the actuator.

I have also performed a run without an injection at all and there were baseline disturbances that matched my run with an injection (sans target peaks).

When I contacted Valco to select the correct valve for this setup, they told me that a helium purged valve would not be necessary for what I need to do, I'm hoping they are right about that.

My only other thought is that air is somehow getting in through the EPC. The digaram that Valco provides shows a mini gas purifier in between the EPC and the column. I have the provided mini purifier in line between the EPC and the valve port that feeds carrier gas into the system. However there is another gas line coming from the EPC that feeds the inlet (I'm guessing for inlet/septum purge?) This line does not have a mini purifier in line. Could this be the source of the baseline disturbances? Would there be any way to test this theory without buying another mini purifier?

Thanks for your help so far

[chromatographer1]

I suspect you have found the problem.

I would replace the septum and see if that helps.

What is your split ratio and loop size?

If you can use the valve inner loop (2 microliters) as a sample loop and directly connect your column to the valve then you might be able to get by without splitting.

ROd

[AICMM]

Booth,

Based on what you describe, we can sort of eliminate the GSV for the moment. (I am very surprised that Valco said that, at low ppm most of the instruments I work with have purged heads.) The key statement I missed earlier is you said that there is a set of two peaks after the run is over. This inidicates that you are doing something like gas saver at the end which changes flows and allows a burp of air in. Furthermore, you have finished playing with the GSV at that point, based on what you describe.

1) Why not connect GSV direct to column?

2) connect GSV after inlet with a piece of MXT type tubing.

3) Look at your flow program, especially at the end of the run to see what games you play and try taking out said games.

Let us know.

Best regards,

AICMM

[Booth]

Thanks for your help, I disabled the gas saver and tried to take out any unnecessary flow changes. Unfortunately, I was unable to directly connect the MXT tubing directly to the column and complete a run. I was able to plumb the whole thing correctly with the tubing connected to the column directly, but bypassing the inlet prevented the pressure setpoints in the method from being reached. Even with the inlet setpoints set to zero, the column flow was reading zero (not sure if its measured or calculated, but I have a feeling its calculated).

I think the EPC expects to have the inlet in line and can't run without it? If it can, I'm not sure how to set it up.

I think the real problem may be having the inlet in the first place, there's probably a good chance that running a line directly to the column would have solved the problem, and if I was designing a new system, I would not even have an inlet installed in the GC, but since this is a repurposing project, I'm stuck with the inlet.

Does anyone know of a way to bypass the inlet without removing / uninstalling it completely?

[AICMM]

Booth,

Try the second option. Take a piece of tubing from the bottom of the inlet, run it out the oven to the GSV (assuming the GSV is outside the oven) then run another piece of tubing from the GSV into the oven. Hook the column to the second piece of tubing. Keeps the inlet in the configuration but puts the column in line with the valve. Eliminates split injections but at a minimum you can use this to troubleshoot.

best regards,

AICMM

[GasMan]

Booth,

I think that the problem is the auto zero for the inlet. The default is to have this set to on. At the end of the run, the GC will turn the flow off to the inlet, zero the flow sensor, and then turn the flow back on again. To turn it off, try the following. On the GC keyboard click on Options,Calibrate, scroll down to the inlet you are using, and press enter. The first entry is Auto Flow Zero. Set this to OFF

Gasman

[kalylcie]

Verify that the PDD detector is set up correctly. Check the detector parameters, such as lamp voltage, lamp current, and gain settings. buckshot roulette