GC Baseline drift and noise

[Chris Jones]

Hi there,

I am running a Shimadzu GC17A with He as a carrier and FID. I'm typically analysing essential oils from Indian sandalwood. Temp program 80 initial, ramped to 180 degrees at 4 per min, then hold for the remainder of the run (total time 45 min). Total flow is about 55 ml/min.

Up until a few months back it was running beautifully. Now, after someone else installed their column and conditions, I'm getting diabolical chromatography. This is described by a rising baseline with temperature, and a very slow leveling out of the baseline after a run. The rise is about 1500 uV; when it was running fine I was getting a rise of 100 uV. Also, the baseline is really noisy; peaks and troughs of 60-70 uV, compared to 5 uV under normal good conditions.

I have tightened the column and fittings - no change.
I have taken the column out of the detector and sealed it with a ferrule and steel wire (as per instructions) and re-run the temperature program - same baseline rise and noise ...
Ahha! must be the FID. Replaced the FID with a spare...same results. Replaced the multiplier...same results!!
hmmm...
OKAY, whole new GC. I had the technician bring in a brand new GC 2010 which we new would work. Installed column, set up method and waited... SAME BASELINE RISE and shocking noise.
So, all we are left with must be the gases entering the FID. The hydrogen has not been changed, but the air cyllender has been changed. I will check this in the morning (18 hours time).
HAS ANYONE HAD THIS PROBLEM BEFORE?? Please help!!

Thanks,
Chris Jones.

[chromatographer1]

Now fix it.

Your gases appear quite contaminated and have probably contaminated the tubing. I do not know the source of your gases but you should use replacement cylinders of gases you know are good and do a bit of cleaning and/or replacement of the supply lines until you locate the problem.

Most likely it is bad helium from a contaminated cylinder since an oven program seems to increase the problem. I assume you are using capillary columns with a helium makeup gas. It could be hydrogen or air as well if the heat from the oven is releasing trapped contamination at the detector lines.

Your traps may be saturated (if you are using traps). If the carrier is the problem your entire inlet system is contaminated and must be made pristine. First however, find out which gas(es) are causing the problem.

Good luck. (I hate it when these things happen) But remember to clean lines thoroughly or you will never find the source of the problem.

Rod

[Chris Jones]

Hi Rod, and thanks for the reply. I know for a fact that it isn't the helium, as I let the column flow into the oven while the FID inlet was completely sealed. Once the oven was on it's way up to 180 degrees, the baseline rises. Gremlins I knew it!

I'm going home for the night anyway!

...One last thing - When the oven was sitting idle at 80 degrees, and I let the baseline stablise, something very peculular happens. While the immediate noise is around 60-70 uV, over a period of 20 mins it performs a full drop and rise, not unlike a sine wave with an amplitude of about 300 uV. I kept it at 80 and let it do this over a 3 hour period. Sure enough, a perfect sinewave baseline occurs!! Wavelength about 20 mins. Bizzare eh?
CHRIS

[GoldLeader]

Chris-
I recently experienced a problem exactly as yours where the detector gave a sinusoidal signal over about a 15 min cycle. For me it turned out to be that the zero air generator I use had gone bad (oven in the generator cycling on/off every 15 min...). I replaced this with a cylinder and have had a teady baseline since.
Good Luck,
-Jeff

[Chris Jones]

Ahh, Thanks Jeff.

I am still confused as to how the sinewave baseline can occur when I'm using bottled gases. I am arranging to have the hydrogen cylinder exchanged, and hopefully clean out the line. However, if the H was contaminated, wouldn't the baseline rise to a fairly noisy point than stop rising (assuming the cylinder is a homogenous mixture of H and contaminants)?

Cheers,
CHRIS

[chromatographer1]

You probably have heavy hydrocarbons producing this baseline. Itn't it wonderful to see C80s and C100s being separated by carbon number?

You got a bad contaminated bottle of gas.

It may have taken a while after installation for the grease to get to the detector through your system, but you have grease all through your pneumatics and if you don't clean and/or replace, you will never again get a good baseline.

Sometimes traps are a good idea as well as replacement on a yearly basis.

Good luck,

Rod

[Les]

It's not impossible to have a bad regulator on the air tank. Pressure pulses may look like sin wave if the diaphram is sticking. Most of the time I get noisy signals from FID is usually caused by partial blockage of the jet.