Strange drift of retention time on hp5890

[intereco]

hello,

thereis very strange drift of retention time:
- first compound peak shifting more than next ones
- seems difference of drift not linear

see please photo

there is hp5890 series II with EPC, already checked septa and leaks - seems system sealed. Temperature programm mode: 50 C - 8min, after that 10 C/min to 140 C

what is might be? EPC?

thanks!

with best regards,
-vadym

[Peter Apps]

Hi Vadym

Is the carrier set to run at constant pressure or constant flow ?

Poor temperature control is a potential cause - check the oven temperature with an independent thermometer, a recording thermocouple would be easiest.

What compounds are these and what kind of column is it ? Why do you have an 8 minute isothermal hold when all the peaks are so widely separated ? - are you using a thermal desorber or something similar to put the sample onto the column ? What isthe matrix of the samples ?

Peter

[Johnny Rod]

Hmm I think I'd be looking for a leak or perhaps a blocked FID jet slowing th column flow down

[intereco]

Is the carrier set to run at constant pressure or constant flow ?

constant flow

What compounds are these and what kind of column is it ?

Aceton, Ethanol, Toluen, Butyl acetate, Butanol
DB-WAX, 30m, 0,53 ID

Why do you have an 8 minute isothermal hold when all the peaks are so widely separated ? - are you using a thermal desorber or something similar to put the sample onto the column?

manual injection by gas syringe

What isthe matrix of the samples ?

air pollution. transportanton in plastic bag

Peter

Thanks a lot!

With best regards,
-vadym

[intereco]

Hmm I think I'd be looking for a leak or perhaps a blocked FID jet slowing th column flow down

we check for leak but find nothing. Did you mean FID jet too contaminated and it may, even, affect retention time?

with best regards,
-vadym

[Peter Apps]

Hi Vadym

What is the gas flow rate through the column in ml/min ?, and what is the volume of the sample ?

Peter

[intereco]

Hi Vadym

What is the gas flow rate through the column in ml/min ?, and what is the volume of the sample ?

Peter

4.3 ml/min, sample - 1 ml of gas

thank you!

-vadym

[Peter Apps]

Hi Vadym

GC is very simple - only three things can influence retention time: stationary phase, gas flow rate and temperature. Your retention shift is not due to changes in the stationary phase (if it was the compounds of different polarity would be differently affected).

That leaves gas flow and temperature. If you inject an unretained compound, methane for preference, does the retention time change ? If it does you have a gas flow problem. You could also turn off the detector gasses and measure column flow directly at the detector. If the unretained compound always elutes at the same time when the other compounds drift then you have a temperature problem - check the oven temperature with a independent thermometer, not the readout of the GC.

Peter

[intereco]

Thanks Peter Apps for info! First of all, I suspect that it's proportional valve of EPC problem. We already changed one and today I will check it!

with best regards,
-vadym

[Johnny Rod]

In that case you need to manually check the gas flows. I suggested a leak or other flow problem because the earlier peaks are affected more than those at the top of the temperature ramp. Yes a blocked capillary jet can cause longer RTs though with constant flow mode this effect will be lesser as the GC will jack the pressure up to try to overcome the blockage and maintain the same flow. Everything that enters the column head has to come out of the FID jet so a blockage would slow the whole column down.

[Peter Apps]

In that case you need to manually check the gas flows. I suggested a leak or other flow problem because the earlier peaks are affected more than those at the top of the temperature ramp. Yes a blocked capillary jet can cause longer RTs though with constant flow mode this effect will be lesser as the GC will jack the pressure up to try to overcome the blockage and maintain the same flow Sorry Johnny Rod, I have to disagree with you here. The EPC controls flow through the column by controlling pressure in the inlet - it has no way to measure actual column flow and so cannot detect a blockage. The pressure that it uses to drive the flow is calculated from the column dimensions entered to the GC setup and the viscosity and temperature of the carrier gas. You can have all sorts of fun with flow rates if you lie to the GC aobut column dimensions or what carrier gas you are using . Everything that enters the column head has to come out of the FID jet so a blockage would slow the whole column down.

Peter

[Bigbear]

Not to pick nits, but with a FID you can check column flows. Just turn off all gases for the detector and the only thing left is column flow.

[Johnny Rod]

Peter you're right, on a 5890 there is no flow measurement. I've just been tangling with two new Agilents and they do have flow sensors in the inlets. Bigbear that's what he means, not that it's impossible to measure manually. So yes I do think it's a flow/blockage problem. Quickest check is put a new FID jet in and try again.

[Peter Apps]

Hi Johnny Rod

To actually measure the flow into the column would mean putting a flow sensor in the sample stream, or doing it by doppler shift which might be challenge at GC scale and temperatures. Typically there are flow sensors on one or more of the carrier gas inlet line, the split outlet and the septum purge - with the full suite of sensors you can get the column flow by subtracting split plus septum purge from total (unless you have a leak !).

Peter

[intereco]

so, it was problem not with proportional valve but with EPC pressure sensor. As soon as it was replaced everything became fine. But I understand that it's not exclude possible problem with FID jet - I will try replace it soon.

thanks a lot!

with best regards,
-vadym