Combi-PAL headspace injection again

[Karen01]

I know when using a loop headspace sampler ones does not need a septum purge (or just a very low non 0 flow)...

What about when using a syringe headspace Sampler like the Combi-PAL? I would think it would be the same situation... but I may be wrong. (BTW the GC is a 7890)

Any suggestions on getting smaller bands fro the early peaks (the first one is too broad) when the higher boiler peaks are already smaller than I would like?

Given that from what I've read here, the sample injection rate should be no more than about 1/2 the flow rate of carrier into the inlet (and that is where I'm at so I guess I can't inject faster), and that I can't drop the GC oven temp at injection any more (would be too close to ambient) or raise the HS oven much, any suggestions?

Thanks,
- Karen

[Peter Apps]

Hi Karen

The function of the septum purge is to stop muck from the septum from diffusing down into the body of the inlet, and then getting onto the column to make interfering peaks and ugly baseline humps - so it needs to be on all the time.

Is the extra width of the early peak actually affecting resolution and/or integration ? If not then it is not really a problem.

Please give us your operating conditions - how do you expect us to troubleshoot without information ?

Peter

[Karen01]

The function of the septum purge is to stop muck from the septum from diffusing down into the body of the inlet, and then getting onto the column to make interfering peaks and ugly baseline humps - so it needs to be on all the time.

With a loop headspace autosampler on a 5890 with transfer line needle through the septum, I had the septum purge outlet caped off for years with no baseline issues. I did it because it was just diluting the sample.... When doing headspace the inlet is typically much cooler than when doing a liquid injection.

Is the extra width of the early peak actually affecting resolution and/or integration ?

Yes it is.

Please give us your operating conditions - how do you expect us to troubleshoot without information ?

I posed the question generally on purpose because I hoped for a discussion of principles.

In any case
column: FFAP 30m, 0.32 mm, 1 um (to hold up low boilers)
Oven 45C (room can get warm so I don't want to go lower) for 0.5 min then 10C/min to 200
Carrier: He @ 2.5 mL/Min Splitless
Inj Vol: 350 uL @ 20 mL/s

First peak (at peak) off at about 2.5 min (not at work at the moment)

Thanks,
- Karen

[Peter Apps]

The function of the septum purge is to stop muck from the septum from diffusing down into the body of the inlet, and then getting onto the column to make interfering peaks and ugly baseline humps - so it needs to be on all the time.

With a loop headspace autosampler on a 5890 with transfer line needle through the septum, I had the septum purge outlet caped off for years with no baseline issues. I did it because it was just diluting the sample how can carrier gas going out throrugh the septum purge dilute the sample ?. Even if you have the headspacer plumbed to supply all the gas to the inlet, any of the sample going out to the septum purge simply increases the effective split ratio. The only way to dilute the sample is to add extra gas to it..... When doing headspace the inlet is typically much cooler than when doing a liquid injection good point, another possible issue is that once the transfer line is through the septum it stays there, whereas with liquid or gas injections there are repeated penetrations.

Is the extra width of the early peak actually affecting resolution and/or integration ?

Yes it is.

Please give us your operating conditions - how do you expect us to troubleshoot without information ?

I posed the question generally on purpose because I hoped for a discussion of principles. OK, general principals you shall have. The 1 ml default sample loops of valve and loop samplers are a hold over from the days of packed columns. To work effectively with capillary columns they need split injections, and that means having an inlet connection with extra gas going in and out, and issues of poorly swept volumes, adsorption and gas mixing. To be compatible with capillary GC the samples need loop volumes of 20 - 100 ul, with a zero dead volume connection between the transfer line and the column (or better yet with the column connected direct to the six-port valve). With either loop or syringe injections and with splitless transfer to columns up to and including 0.32 mm diameter through a conventional split-splitless inlet, the gas flow rate through the inlet is too slow to effectively flush the internal volume quickly enough to ensure narrow peaks for compounds that are not stationary phase focussed - especially if the inlet has a "splitless" liner which has a larger internal volume in order to accommodate the vapour cloud from a hot injection of volatile solvent. With syringe injections there is the additional problem that the inlet pressure causes back-flow into the syringe as the needle penetrates the septum, with subsequent issues of incomplete mixing and incomplete transfer of the original sample volume into the inlet. The position of the tip of the needle or transfer line relative to the top of the column can impact what fraction of the headspace sample gets into the column, and what fraction flows away down the split outlet or diffuses into contact with active metal surfaces.

In any case
column: FFAP 30m, 0.32 mm, 1 um (to hold up low boilers)
Oven 45C (room can get warm so I don't want to go lower) for 0.5 min then 10C/min to 200
Carrier: He @ 2.5 mL/Min Splitless
Inj Vol: 350 uL @ 20 mL/s 20 ml/s I would think that this is a typo, except that 20 ml/min is still too fast, and 20 ul/s is too slow, giving a minimum starting band width of 17.5 s - so maybe the devil is in the details

First peak (at peak) off at about 2.5 min (not at work at the moment)

Thanks,
- Karen

[WillyOne]

Are we talking about alcohols in beverages? If so can send you some suggestions

[Karen01]

how can carrier gas going out throrugh the septum purge dilute the sample ?. Even if you have the headspacer plumbed to supply all the gas to the inlet, any of the sample going out to the septum purge simply increases the effective split ratio.

That is what I meant.

The 1 ml default sample loops of valve and loop samplers are a hold over from the days of packed columns. To work effectively with capillary columns they need split injections, and that means having an inlet connection with extra gas going in and out, and issues of poorly swept volumes, adsorption and gas mixing. To be compatible with capillary GC the samples need loop volumes of 20 - 100 ul, with a zero dead volume connection between the transfer line and the column (or better yet with the column connected direct to the six-port valve).

For the loop sampler I always used split injections

With either loop or syringe injections and with splitless transfer to columns up to and including 0.32 mm diameter through a conventional split-splitless inlet, the gas flow rate through the inlet is too slow to effectively flush the internal volume quickly enough to ensure narrow peaks for compounds that are not stationary phase focussed - especially if the inlet has a "splitless" liner which has a larger internal volume in order to accommodate the vapour cloud from a hot injection of volatile solvent.

I use 2 mm straight through liner for headspace.

With syringe injections there is the additional problem that the inlet pressure causes back-flow into the syringe as the needle penetrates the septum, with subsequent issues of incomplete mixing and incomplete transfer of the original sample volume into the inlet. The position of the tip of the needle or transfer line relative to the top of the column can impact what fraction of the headspace sample gets into the column, and what fraction flows away down the split outlet or diffuses into contact with active metal surfaces.[/color]

I would think that this is a typo, except that 20 ml/min is still too fast, and 20 ul/s is too slow, giving a minimum starting band width of 17.5 s - so maybe the devil is in the details[/color]

It should have been 20 uL/sec. The advice I've seen here for syringe headspace samplers is to use half the carrier flow rate. Half of 2.5 mL/min = 20.8 ul/sec and that is what I used...

Injecting larger volumes using a split injection to get faster flow does not get me much from what I can see. Is that wrong?


Thanks,
- Karen

[Karen01]

Are we talking about alcohols in beverages? If so can send you some suggestions

I am not dealing with alcoholic beverages but some alcohols and other compounds from aqueous solution, so info on that could be useful!

Thanks,
- Karen

[WillyOne]

Well, that I use is really forzed to achieve a quite good separation of Methanol and Ethyl Acetate and keeping Ethanol under control
Column: 60 m 530µ i.d. 1.0µ thin film Poliethylenglicol (Superwax from Teknokroma
Injection: direct (after have been added Internal Std) volume 0.5 µl ( I'm talking about wine) ( no headspace)
Septum purge closed
Split ratio 4 to minute 5 then 10 ( because Ethanol peak is so big)
Injector temp 150°C
Flow 4.5 ml/min
Oven 45°C for 12 min, then 10°C/min to 110°C wait 5 min and then 20°C/min to 240°C, wait 5 min
Detector FID 260°C.
So I analyze Acetaldehide (poorly retained), Ethyl Acetate, Methanol, 1-propanol, 1-butanol, 2-Butanol, isobutanol and iso amyl alcohol,
Int Istd is Methyl isobuthyl carbinol 1.5 mg/ml in Ethanol
Of course inlet liner wide: 4 mm i.d. packed with glasswool and changed every two weeks SG liners
Hope it helps