Problems with getting a stable PFPD signal

[big boss 88]

Good day all,

I have a Varian 3800 with PFPD, I have set all parameters for the sulfur in gas analyses as per ASTM 6228, the problem is that I am trying to get the instrument to stabilize before the run and while watching the PFPD signal it is never stable,

So the PFPD signal is fluctuating before the run jumping from -2 to 2 and I remember it was very table and fluctuating in the range 0.005 up or down,

When I start the instrument anyway, the run is very noisy with noise up to 1000,

The square root is off,

[muGC]

What does your emission look like in WinPulse?

From the little information given, my initial guess is that your combustor is dirty or broken/chipped and the CH* C2* and OH* emissions are large and spreading into the sulfur region.

[big boss 88]

My software is Galaxies and the PFPD settings are as follows:
Heater: 250 degrees
Square root mode off
Tube voltage 600 V
Trigger level 200 mV
Delay 6.0 ms
Width 20.0 ms
Gain factor 2
Air 1: 17
H2: 13
Air 2: 10
The detector is actually pulsing and I can hear it.
The column is CP-Sil 5 CB for sulfur, 30m, 0.32mm, 4um
The flow rate is 2 ml/min.

[muGC]

Just because you are pulsing doesn't mean that you are able to achieve tick-tock mode. Can you post a screenshot of your emission?

[big boss 88]

Excuse my little knowledge of PFPD, do you need me to post a chromatogram? Is this what you meant by posting a screen of the emission?

[muGC]

No, the emission is located in your WinPulse program that controls the PFPD. For an example please refer to page 19 in chapter 3 of your OI PFPD manual.

[big boss 88]

Actually my PFPD is from Varian so I referred to Varian manual and they mentioned getting the emission by connecting a chart recorder which is not available in my lab. Also, they mentioned something about adjusting the tick tock mode by playing with the flow rates and watching the chart.

The only way to get some recording is by running blank run and watch the Galaxies software and here I am posting a screenshot from my system.

Please see the below screenshot:

http://postimg.org/image/49me5hy5b/

[muGC]

Ah Varian, someone else on the forums will have to help you then. Apparently the OI PFPD is different as it has its own software for looking at the emission. Looking at the emission is helpful for obtaining tick tock mode and using the PFPD. However it is not completely necessary. You can adjust the flows and listen for the tick tock mode. I was suggesting that your PFPD combustor tube may be dirty or broken and this is easily seen when looking at the emission.

The screenshot you posted is just your GC run baseline, Galaxies will not show you the emission to my knowledge.

[Peter Apps]

I have a Varian PFPD on a 450GC. It works very well with the default settings given in the operator's manual (on CD). There is a different diameter combustion tube for the different modes.

What has changed (maintenance, column, gas cylinder) between when you had a stable baseline and now ?

Peter

[big boss 88]

I opened the detector and located the combustor which was dirty as you expected with white deposits.

Is there any cleaning procedure for that?

I tried with methanol and sonication then I tried nitric acid with heating. Both were not sucesful to remove all deposits.

[muGC]

Not sure about the older Varian model but the OI 5380 PFPD combustor tube cannot be cleaned according to the manual. You must obtain a new one.

Also, keep in mind that the combustor can be dirty but the "dirt" may not visible to the eyes. Oils from your fingers will contaminate the combustor, make sure you use gloves and treat the new combustor with care.

[big boss 88]

Peter and muGC,

Thank you for your time to read and answer my questions,

I did place an order to buy a new combustor which usually take 2-3 months, meanwhile, I have cleaned the combustor as much as I could,

Now, the baseline is very stable and clean,

I ran a sample with high H2S concentration 5% in natural gas and got a good signal but as intense as I expected,

I ran a standard with 100 ppm H2S and could not see anything even when I closed the split nothing appeared,

It looks that I have a sensitivity issue hear since the PFPD should be very sensitive to detect 1 ppm of H2S,

Any thoughts on improving the sensitivity of the detector of optimize it to work better,

[muGC]

Since you are using D6228, you are injecting with a gas sampling valve. Correct?

Sensitivity problems do not always mean the detector is bad. It may not be sensitivity but an injection problem. Are you using passivated (e.g. Sulfinert) tubing, fittings, gas sampling loop to inject your sample? It is possible you could be losing sulfur before it reaches the GC. Additionally, the gas sampling loop, gas sampling valve, or the gas sampling valve rotor could be partially plugged limiting the amount of sample being injected.

If the detector is the problem, here are some potential problems with your PFPD:
Combustor is dirty (likely your problem).
Excess or improper flows through the combustor.
Your PMT voltage could be too low (although 600V is fairly high for the OI PFPD).
Gate is set incorrectly (6-26 ms should be fine though).

[big boss 88]

You are right muGC I am using a gas sampling valve and the loop size is 0.5 mL; however, it should be 1 mL as per ASTM D 6228,

The valve is from VICI, it is a two position 6-ports injection valve as per the below website:

http://www.vici.com/support/app/app11j.php

The valve settings are 0.01 min to inject and after 1 min to close,

For the tubing, I am not suing the recommended PEF tubing as they are not available and instead I am using he normal tubing and I tried to minimize the lengths as much as possible,

I will try to open the valve and to check the rotor condition,

[Peter Apps]

When you opened the detector to check the combustor, did you isolate the photomultiplier from the power supply ?

Peter