General questions

[Roxanne42]

Hi there,

I am very new at GC but the facility I am managing does have an analytical side, which I am also in charge of. I've been trying to optimize a test separation and I've read an awful lot online and in a few books but I needed some clarifications on a few points...

(1) I am using an Agilent 6850 with FID. I was injecting a compound with a bp of 265 C, dissolved in MeOH (I'm aware I didn't pick the right solvent). I set the detector and inlet T at 285 and the oven at 60 (isothermal). I'm using a 30m, 0.32 column with helium as my carrier gas. I set the inlet pressure to 8 psi (according to a table), split injection of 500:1 (fairly concentrated sample) and 1 uL injection.

According to Restek's backflash calculator, this would lead to excessive backflash. I have read in some places that backflash does not apply to split injections but I also have read that it does... Should I be concerned?

(2) I found that setting a higher pressure (25 psi, like was suggested bv previous users) led to peak tailing (although I had to use a lower split ratio of 100:1 due to the higher pressure). How should one go about adjusting pressure and flow rate for optimization?

(3) Pulse pressure related to injections - contains the sample expansion and transfers the analytes to the column faster - is this something I should consider doing, especially when using MeOH as solvent?

(4) What three top solvents are generally the best to use for dissolving samples (once solubility is assured)? Do you use a high solvent grade for GC-FID?


Thank you so much for your help,

Roxanne.

[Peter Apps]

Hi Roxanne

With the very high split ratios that you are using, "backflash is irrelevant. 500:1 is very high, what is the concentration of the sample, and have you calculated how much of the analyte goes onto the column ?

Why are you running isothermal at 60 degC ?. How long does the peak take to come out ?

To optimise flow through the column (and therefore inlet pressure) inject some hydrocarbon gas (from a cigarette lighter or LPG). Measure the time that it takes to elute (in other words the retention time for its peak) and calculate the mean velocity of the carrier gas as column length divided by lighter gas retention time. For helium the optimum carrier gas velocity is 35 cm/s, for hydrogen 50 cm/s. 8 psi should be about right for the an inlet pressure to give you about the right flow velocity, but with a very high split ratio you might need to ncrease the pressure slightly. Unless there is something odd going on 25 psi is way too high. Also you should not need to change the split ratio when you change pressure (but to keep the ratio the same you will have to change the split flow because the volume of gas going into the column changes with pressure).

Until you have sorted out a straightforward split injection do not be tempted to play with pulsed pressure and opther fancy options - they just give a new set of interactive parameters that have to be optimised.

The most popular solvents (which might not necessarily mean that they are the best) are hexane, dichloromethane and ?ethyl acetate. As long as it is inert, volatile and not water you can use almost anything.

Good luck

Peter

[Roxanne42]

Hi there,

With the very high split ratios that you are using, "backflash is irrelevant. 500:1 is very high, what is the concentration of the sample, and have you calculated how much of the analyte goes onto the column ?

Actually, the concentration was unknown (and after the first injection, higher than I had thought). I agree the split ratio is a little high and that the sample should definitely be diluted. We were told by the Agilent rep that the column should see a maximum of 10^-4 M and that one should aim for peak heights between 30-300 pA...

At what point is backflash a concern for split injections?

Why are you running isothermal at 60 degC ?. How long does the peak take to come out ?

It comes out at about 2.4 minutes. I do isothermal since we really wish to minimize runtimes due to the high-throughput nature of the facility (will analyze 96 samples overnight). I can do the run at 40, but it takes a little longer.

For helium the optimum carrier gas velocity is 35 cm/s, for hydrogen 50 cm/s. 8 psi should be about right for the an inlet pressure to give you about the right flow velocity, but with a very high split ratio you might need to ncrease the pressure slightly. Unless there is something odd going on 25 psi is way too high. Also you should not need to change the split ratio when you change pressure (but to keep the ratio the same you will have to change the split flow because the volume of gas going into the column changes with pressure).

The average velocity is automatically calculated in my software so I just need to entire desire flow with split ratio. So from what I understand, I set my split ratio and then set the average velocity to be around 35 cm/s, which will lead to a given pressure (I believe the split flow adjusts itself accordingly). Does the column flow rate have to be between certain values for good results? Does it have a lower/upper limit? And when given the choice to run in constant pressure mode or constant flow is one preferred over the other?

As for 25 psi being too high; our current instruments were obviously setup by someone else. When the GC is not in use (overnight/week-ends), it is set to a method called "sleep mode", where pretty much everything is OFF, except for the carrier gas, which is set to 25.00 psi (giving a column flow of 7.7 mL/min and a total flow of 90.5 mL/min). Should I change these parameters and if so, what would be a good value? It seems awful high to me just for preserving column integrity...


Thank you so much for all your help,

Roxanne.

[Peter Apps]

Hi Roxanne

the easy part first: having 7.7 ml going down the colum and 85 ml or so through the split while the GC is on standby is a waste of gas. The GC has a gas saver (on the inlet and column flow control windows) which kicks in at a specified time after an injection - use that to limit the standby flow.

My personal opinion is that with any sensible split ratio (i.e. 10:1 and above) you do not need to worry about solvent vapour volume.

How much of an analyte you can put onto a column depends on the match between the analyte and the mobile phase - usually it is between 10 ng and 100 ng, sometimes up to 1 microgram. 10 minus 4 M gives you more than 1 mg which is way to high (and with a 500:1 split the sample would have to be nearly all analyte).

The retention time is very short - not even double the dead time. What is the MW of the analyte (or better yet what is it !?), can you give even a rough estimate of the concentration ?. What do you have to separate it from (if anything) ?. What is the stationary phase in the column - if it is non-polar (say methyl silicone or 5% phenyl silicone) I would guess that the peak that you are looking at might be the methanol solvent

You are right to set the flow on the software, but Chemstation sometimes changes values on the inlet screen when you edit on the column screen and vice versa, so always go back to check that the settings are what you want.

If the gas flows are a little bit out you do not see a dramatic change in chromatographic performance, but the closer you are to the optimums, the better the separation. To optimise speed you can increase flow rate (never run at lower than optimum flows) - try putting the helium flow up to 50 cm/s.

Regards

Peter