double peak methanol; residual solvent

[centraing]

Hello,

My english is not verry good, so sorry for this .
i'am having a problem with the residual solvent testing by Ph. Eur.
When i inject a standaard class 2A in water, the methanol peak appears as a double peak and is verry broad. Can anyone tel how i can make it better.

thanks already !

ing

GC-kolom: RTX-624,30m, 0.32mm, 1.8um
flow:2ml/min (35cm/sec)
splitflow:1:5
Liner: 0.25ml special voor HS
injectionvolume: 1ml
injectionspeed:10ml/min

[chromatographer1]

You have too large a sample volume into too small an injection liner.

Increase your liner volume and decrease your sample size. increase split ratio. increase carrier flow rate.

good luck,

Rodney George

[krickos]

You have too large a sample volume into too small an injection liner.

Increase your liner volume and decrease your sample size. increase split ratio. increase carrier flow rate.

good luck,

Rodney George

Either that or it is a 1ml gas sample which is standard for Ph Eur/USP residual solvent analysis with headspace injection. In that case the injection speed seems to slow, takes 6s to feed the gas sample into the injector which should be tooo slow, try to increase injector speed.

Also would be nice with a more complete instrument setting, GC oven profile etc and and concentration of standard (µg/ml and how many ml in vial) in order to help better.

[chromatographer1]

I thought they meant 1µL not 1mL for a test solution injected directly by liquid syringe.

But are they injecting from a gas syringe at 10mL/min of a 1 mL volume over six seconds then?

If this is the case, then perhaps they are getting too much water vapor in the sample from heating the headspace vial at too high a temperature. The water overloading the column can then cause the malformation of the methanol peak.

They should lower the temperature of the HS oven. In my research in HS back in the 1990s I used a short piece of 0.53mm ID bonded PEG capillary column ahead of the OV-1301 column to compensate for the water vapor plug. I kept my HS oven to 80°C. This was quite effective and was used to separate ethanol from ether in the elution order. See Jun 1997 Analytical Chemistry p2003

Rodney George

[centraing]

Hello,

Thank you already for your advice!

I'll give you more GC settings to compleet your advice:
GC:
oven: 40degree for 20min - 10degree/min to 220degree - 220degree for 20min.
detector: 260 degree
injector: 140 degree
flow: 2ml/min
split: 1:5
liner:0.25ml special for headspace

Headspace:
oven: 80degree
needle: 85degree
vial: 20ml (with 5ml water and 1ml standaard class 2A (0.15mg MeOH/1 mlwater))
incubation: 60min
injection volume: 1ml
injection speed: 10ml/min

Our serviceman sie that the injectionspeed must be 80% of the total gasflow on the GC (= 2ml/min + 10ml/min split), thus thats why the injetionspeed is 10ml/min. Is this a (well)reasoned advise?

At first I had a 0.53mm*3um kolom, but the resolution between acetonitril and dichloromethane was <1.0 it has to be >1.0. There was no problem then with the methanol peak. Now I'm using thus a 0.32mm*1.8um with the same settings for GC and HS. The resolution is fine, but now the peak for MeOH-EtOH-IPA is really bad. All the other peaks are fine. So the problem sit at front of the chromatogram (the polair components).

Can you tell me how much I can change the Ph. Eur. description for the GC and HS setting? What is allowed?
I'll hope that this is enough information about my GC-analysis for now.

Ing

[aldehyde]

1 uL of water will expand to about 1000 uL of vapor, so if your liner is only 0.25 mL and your sample is like 10 mL of vapor its not going to happen. The above post is right on, you need to inject less. Get a liner with a larger internal volume and increase split flow so the sample gets out of the inlet faster.

[centraing]

I understand your advise, but i'm already injecting 1ml vapor because the injection techniek is headspace. Information on the internet (for the detemanation of residual solvent USP467) gives just a liner of 1mm (0.25ml) special for headspace. This confusses me !!

ing

[haiedc]

As I understand your analysis is headspace so 1ml is not too much, but I used to use only 250ul loop in headspace.
I think you try to increase the split ratio to 1:10 or 1:20 to see the effect. Probably it's a case of overloading.

[krickos]

Probably it's a case of overloading.

Agreed, most of the settings are pretty much standard, the volatiles inlet I use have a volyme of like 32-37µl compared to a 0,250ml still works fine.

0,15mg/ml water can cause 2 things:

A. Overload as stated fact is that 5 time 0,15mg gives 7500µg intotal in vial (which is pretty high), due to 1. using water as main solvent, " headspace oven temp way above methanol you get most methanol in gasphase. 2. Reducing column dimesion reduces your sample load capacity.

B. Rodney mention the water vapour issue, though myself I have only encountered it above 80°C in HS oven. (then again most method I have developed have been based on DMF as solvent).

So as suggestion EP/USP have a set of different typical HS parameters, as resulution is an issue with the 0,53mm id column, pick the setting with 60°C for headspace oven and see if that works better (less solvents in vapour) or validate your own method.
You do not seem to have much lean way otherwise as flow 2ml/min ~30cm/s and column oven temp is 40°C with column that is pretty much optimized.


Ohh one last thing, EP if recalling right mentions that sample/standard preparation might have to be adjusted compared to their default methods.
USP do not have that added at least yet.

[chromatographer1]

Yes, that is a lot of methanol for a HS vial. Keep it under 100µg per analyte per vial.

How to trouble shoot the problem?

Inject 10 or 100 times less than you usually do. Does the problem go away?

If yes then fix the problem's cause. If no, then fix the column.

Easy enough, no?

best wishes,

Rodney George

[centraing]

Thanks for the advice!
The solution in the headspace vial is only 0.15mg MeOH in 1ml water. I see now that I wrote it wrong. So after incubation ther wil be 1500ug MeOH in the vial, right? Is this also to much for injection?

Last friday I thit som tests:
- I set the injectionspeed to 40ml/min. The MeOH peak is now one peak, but is still broad. I also think that the area of the MeOH is much smaller now. Is this because of to fast injection en than the liner wil run over?

- When I change the liner to a 3mm liner, the MeOH peak is again a double peak.

So I think that the faster injection is the best option, but I'am afraid that the wil liner run over! Can anyone tell me of I'am right?

Greeting Ing

[carl.nott]

I don't know if your GC has Electronic Pressure Control (EPC), but if it does (assuming it does) you could use a pulsed split or pulsed splitless injection.

Clarifying because of English: The 'pulsed' means that the pressure in the injection port is increased from, say, 5psi to 30psi immediately after the sample is injected. This helps compress solvent peaks.

[jwhite]

I've seen this on one of my FID instruments. Clipping the injection end of the column has always corrected my split methanol peak.

[centraing]

Hello everyone

I think I had found the problem wath cause most of the wrong chromatografie.
When I first injected the classe 2A, the chormatogram displays:
- Methanol peak is not a nice gauches form (begins with a small bump, than a large peak and at the and a second small bump)
- resolution between ACN and Dichloromethane = 1.30

Because there was memory effect at the end of the chromatogram, I had cleaned the liner and injection syringe with Aceton and hexaan, after this I dryed the liner by 105 degrees. Then I injected the classe 2A mix again and the methanol peak had form the double peaks.

Today I had seen that the resolution between ACN and Dichloormethaan was 0.8. I can only think the problem is the liner, right. Must a liner always be deactiveted? If yes, can anyone tell me how to deactivate the 1mm liner?

Greeting Ing
from Holland

[Peter Apps]

Hello Ing

The bumps you see may not be methanol - they may be contaminants. Try washing the syringe with methanol and nothing else, and then purging it very thoroughly with clean gas at elevated temperature. Then inject from the standard headspace.

This might get rid of the bumps, but it will not solve the problem of the peak being too wide. As Rodney says, you need to increase the split ratio, or decrease the injected volume. Under the conditions that you have at present you are simply volume overloading the column.

Peter