EN13130-7:2004 Glycols Migration in Aqueous Simulants

[mbdixon]

Hi I have been trying to calibrate my HP6890GC with FID detector to run the aqueous standards as described in the EN13130-7:2004 Materials and articles in contact with foodstuffs. Plastics substances subject to limitation. Determination of monoethylene glycol and diethylene glycol in food simulants

Inlet Conditions : Splitless, Temp 250°C, Purge Time 1.00, purge Flow 50.0, saver Flow 20.0, Saver Time 5.00 - Drilled Uniliner Direct injection (hole at top)
Injection 1µl manual
Column ZB- Wax Plus (30m x 0.32 mm ID x 0.25µm thickness)
Oven 80 (2mins) to 160°C @ 10°C/min, hold for 2.00min, then 50°C/min to 250 for 2 minutes
FID Detector 250°C (H2 25, Air 350, Make Up N2 29.0)
Carrier Gas Nitrogen @ 3.0ml/min

The MEG peak is very broad (but symetrical), whilst my DEG/Butan1,4 Diol (IS)peaks are OK but not fully resolved. (see attached chromatogram) Using a 50/50 mix of IPA and water as the matrix did not make any difference to the peak shape quality.

Having done a calibration the DEG has a correlation factor of 0.9996 of which I'm more than happy with, but the MEG is just 0.97. Will I just have to live with that or can anyone suggest anything to improve it?

Any suggestions??

7501518570_6eca4a440a_n by Mark Dixon, on Flickr

[Yama001]

Perhaps this is a case where a retention gap will help?

I find that the 6890 inlet often does not behave *reasonably* for lower boilers and often resort to a split or a pulsed split at low split flows - perhaps 1:1 to 5:1. It almost seems the inlet in splitless mode (which I would prefer here on a 5890) insists on flashing well back into the inlet pumbing.

I have some colleages that like to do a splitless with the valve opening at 0 minutes - another odd approach and one that also seems to be favored due to (accidental?) less interactions in the inlet.

[Peter Apps]

Hi Mark

Those peaks are really ugly ! Before we get too hung up on the inlet, what is the history of the column, and do you have any performance checks on it ? With a similar Restek wax phase in megabore I can get diol peaks 2 s wide with split, splitless or Uniliner direct injections.

Peter

[mbdixon]

Hi Peter,

The column is quite new, a couple of months old. The only perforamnce checks is the CoA that came with it from Phenomenex, and all looks in order. I did also use it to run some derivatised carboxylic acids on it and got very sharp peaks. (Hydrocinnamic acid ~2s). I haven't run any performace checks.

Some things I'm stuck with regards the system (at least in the short term):
Nitrogen as a carrier gas. It is also used on other machines as part as standard methods. To fit a new line for hydrogen or helium isn't going to happen!
Manual injection with the S/SL injector - The company will not see any justification in spending a couple of grand on an autosampler.
Myself - I'm an anlytical jack of all trade type, not a chromatography specialist and I'm a one man band in the lab here, so if it sometimes appears I'm missing something obvious please bear with me!

Yama,

A guard column is something I've considered but I want to try and solve it with what I've got.

[Yama001]

I dug up some stuff I did a few years back for ethylene glycol on an older pre EPC 5890.

RTX-Wax 30 meter x 0.25 mm ID x 0.25um film. I likely set a head pressure (Helium) at 15 to 25 psi. IIRC, the flow was likely 1 to 3 cc/min (constant pressure on the old 5890, though).

2 min at 45 deg, 10 /min to 180, hold 2 min.

Inlet 250, Splitless, 1 uL, valve on at 0.8 min, split flow 50 cc/min. Gooseneck liner, I can't recall if wool was used.

I had a sharp peak at maybe 8 min - calibrated 1 to 100 ug/mL. The sample was aqueous, mix 50/50 with methanol; use methanol to rinse the syringe.

It seems like you should be able to get a decent peak; though I am never surprised at what I see from a 6890 onwards in splitless mode / low boilers. The column is easily degraded, aggressive clipping (a meter of more) was useful. We would do these once a year; the column always seemed shot at first, but would come back. I recall this and other direct aqueous injections jobs were always a problem, with many false starts and swapping amoung various wax columns and operators.

I did use an autosampler, though we started with manual injection. Solvent /Air flush, and keep the syringe in the inlet for a few seconds before withdrawing.

I think the nitrogen carrier should work, just somewhat broader peaks.

[mbdixon]

Cheers Yama,

I have my guard column/retention gap on order, so I'll pick this up again when I return from a short break middle of next week.

Cheers for your help chaps, and have a good weekend.

[Peter Apps]

Hi Mark
Manual injections and nitrogen carrier gas will not (on their own) cause such extreme peak broadening. Can you take a couple of steps back and make up the glycols in methanol and inject some of that ?, have the column initial temp at about 80C so that no methanol will condense in the column. If that does not give sharp peaks then try some hydrocarbons or something else very tractable, dissolved in a non-polar solvent. If the peaks are still wide then there is something wrong with the column, or some muck stuck somewhere in the inlet or the detector.

Peter

[mbdixon]

glycols by Mark Dixon, on Flickr

Chaps,

I think I've got it about as best as I'm going to. In the end the settings I've used are:

Inlet Conditions : Splitless, Temp 250°C, Purge Time 1.00, purge Flow 50.0, saver Flow 20.0, Saver Time 5.00 - Drilled Uniliner Direct injection (hole at top)
Injection 1µl manual
Column ZB- Wax Plus (30m x 0.32 mm ID x 0.25µm thickness)
Pre-Column 10m
Oven 80 (2mins) to 140°C @ 10°C/min, hold for 1.00min, the 160°C @ 10°C/min, hold for 2.00min then 50°C/min to 250 for 2 minutes
FID Detector 250°C (H2 25, Air 350, Make Up N2 29.0)
Carrier Gas Nitrogen @ 3.0ml/min

Thanks for eveyones help. Now a similar sort of thing with terephthalic acid to sort out.....