SH-GC-MSD Residual ethanol

[MSCHemist]

Thanks for all your advice regarding static headspace which is new to me. I've been reviewing the forum archives especially for Chramatographer1's posts and all the residual ethanol posts and HS parameters.

Here is what I've come up for for my residual ethanol method. I get two types of sample liquid flavors in soybean oil, propylene glycol, triacetin, or Neobee and spray dry encapsulated or pan dry encapsulated. I need to ensure they have minimal ethanol for halaal and my goal is to get down to below 100ppm in the samples for LOQ. I have a Tekmar 7000 deactivated sample pathway (unfortunately except for the cut end of the inlet carrier line) and a 6890N-5975MSD. Liner 1mmID. column Innowax or 1701 30m .25mm .25um film.

Here is what I am thinking. I've read Chromatographer1's advice regarding using a low sample volume in the vial. For diluent I am thinking of Using alpha-pinene. I have a ton available it has good oil and alcohol solubility and some/minimal water solubility with a Bp of 155 deg C. I might need a separate method with water for the encapsulated samples though.

Parameters
-100mg sample weighed into a headspace vial
-300ul of alpha pinene containing standard for standard addition
-9mL vials

Vial pressure 15PSI N2 supplied by 6890 aux EPC
Platen temp 80 deg
Loop temp 165 deg
Tline temp 175 deg
Sample equil 2 min
mix time 8 min
mix power 4 (max recommended for 9ml vial)
post mix equil 1 min
Pres time 0.2min
Pres equil .05min
Loop fil (1 ml) 0.25min
Loop equil 0.05min
inject 1 min

Inlet 200 deg C
3:1 pulsed split 3 ml column 9 ml split 2-3 septa purge so a total of ~15 ml/min transferline for 1 minute
Column flow 1ml/min after 1 minute
40 deg for 5 minutes +3 to 75 deg C +20 to 200 5min.
GC/MS SIM 45&46 m/z
I may have to loose the pulse if ethanol doesn't focus well.

Any comments?

[Peter Apps]

My first thought is that your liquid samples are already in nice high boiling solvents, so why dilute them with pinene ?

Diffusion in the higher MW liquids is slower than in water and they are more viscous so do not mix as well with the sloshing action of the headspacer, so you might need more equilibration plus shaking.

Try it with no vial pressurization - it eliminates a source of variation and the increased temperatures provides enough excess pressure to flush the loop.

The loop fill time might be too short if you are aiming to bleed the loop down to atmospheric pressure.

Peter

[MSCHemist]

I need a carrier to dilute the ethanol and add to the samples. I also thought the viscocity of the soybean oil may be a problem. I read somewhere back that a shorter loop fill is better but not too short. I've been running 0.5min so I think I will go with that.

viewtopic.php?f=2&t=12378
viewtopic.php?f=2&t=14167
http://www.danispa.it/empower/file/AN_1 ... OL_rev.pdf

[Peter Apps]

I need a carrier to dilute the ethanol and add to the samples.

Are you doing calibration by known additions ? How many samples in each matix would you usually do in a batch ?

Peter

[MSCHemist]

Like most things in my lab I usually get sporadic samples. I do a whole lot of different analyses but none very often other than basic Flavor TIC's with SPME and quarterly 3-mcpd checks. I'd get at most 2 or 3 samples at a time in a week so since I'd have to do a full calibration each time anyways why not just do standard addition and save the headache with matrix issues.

One of the main reasons I got the 7000 though was doing standard addition by hand with manual SPME was getting rediculous and consuming entire days. I am playing arround with the water method first.

I am also looking to impliment a rancidity method down the line.
http://waset.org/publications/9997968/s ... d-matrices

[Peter Apps]

I agree that with only 2 or three samples in a batch known additions is the method of choice.

If the samples are big enough I would look at taking 10 g of the sample and spiking with 1 ul = 1 mg of ethanol for your lowest addition, mixing thoroughly then subsampling that to the HS vial, rather than use the pinene. Check weigh the addition on a four or five figure balance and use the actual mass added in the calibration. If you work with mass all the way through you also avoid the nasty business of trying to pipette viscous liquids.

Peter

[MSCHemist]

Just thought I'd update. I had to get rid of the pulsed low split. The 6890 EPC was unable to generate it reliably. I switched to a 0.5 film 30m .32 wax column with a 1.3 ml/min column flow and a 5 to 1 split. I saw about a 30% drop in sensitivity going from 3:1 to 5:1, however I was getting %RSD's of 20%+ using the pulsed split and retention time shifts of almost 0.5 minutes out of 4 min.

I had to use the manual controller for vial pressure because the aux EPC would go over pressure after vial pressure time and have nowhere to bleed so the GC would not go ready. Agilent recommended that I could buy a bleed restrictor that they have for their systems but I opted for the manual controller.

I am still considering the 2ml loop for flavors work as I need maximal sensitivity.

[RogerBardsley]

Hi MSCHemist

I am an Applications Chemist for Teledyne Tekmar. Have you considered adding an internal standard? This might help the high RSD.

[MSCHemist]

Yep I am planning to use acetonitrile as an Internal standard. Isopropanol doesn't resolve well enough on the wax column I am using. Oddly I was doing work with some beer and whiskeys and found that there was endogenous n-propanol (not huge ammounts). It is well known the 6890 EPC has trouble with low splits so that was the source of the high %RSD. The only other issue is the tailing on the Ethanol so baseline.