Diglyceride and Triglyceride Inconsistent Chromatography

[II5tickmanII]

Hi, i currently analyse Fats and Oils for Diglycerides and Triglycerides and recently have been experiencing inconsistent chromatography in the area where the DGs are coming off. Its driving me crazy!

For the large part chromatography is typical, flat baseline, sharp peaks, system suit within tolerances etc etc.

But, occasionally (say once every 20-30 injections) an injection will show sharp peaks which start small get larger and then fade away as the baseline rises with the oven program. The peaks are evenly spaced start when the oven is at approx 250C and fade up to when the oven is at approx 340C. It happens regardless of sample type (blank/std/sample) and is not related to contamination in vials as i cannot reproduce the "ghost spikes". Sometimes the peaks are small (5-10mV) and occasionally they are larger (15-20mV)

This happens on 2 different systems with 2 different set ups but always around the DG area. I use a HP5890 Series II and a YL6100 both with autosampler.

DG Method details DB-5MS 15M x 0.25mm x 0.25um, 2.0ml/min, 25:1 split,
Inj/det temp are 360C Oven program 200C start 10C/min Ramp to 340C Hold for 16mins

TG Method details DB-17HT 30M x 0.25mm x 0.15um, 2.0ml/min, 20:1 split, Inj/Det temp at 370C, Oven program 100C start 50C/min Ramp to 350C Hold for 24mins

I have to the best of my knowledge ruled out as much as possible however do not have much experience with these methods and would appreciate any thoughts or advice regarding this. I think the initial ramps especially for the TG method are quite aggressive but i do not see how they would produce spikes. Thanks

[Consumer Products Guy]

We would first make trimethylsilyl derivatives of the diglycerides.

[seeviewonder]

We would first make trimethylsilyl derivatives of the diglycerides.

Do you use any internal standards? Also, are you sure that there is enough silating agent present? I work in the biodiesel industry and have had problems like this before while running free and total glycerin tests on finished biodiesel.

[II5tickmanII]

Thanks for the responses, from what I have read the Fats and Oils industry have some similarities to the Biodiesel Industry in terms of DG and TG methods even If the C chains of interest differ.

The TG and DG methods are separate and each method looks specifically for either the TGs or the DGs. Do the system set ups and programs look typical? I have recently joined the company and "inherited" the methods which haven't been looked at in detail in some time sadly. There us no internal standard for the TGs. The DGs use Triacontane as an ISTD

The DG method uses TMSI as the silylating reagent and has a long burn off period to remove the TGs from carryover into subsequent injections as there is no sample cleanup to isolate the MGs and DGs. In a typical injection the DGs can be split into the 1,2- and 1,3- isomers and further more into the saturated and unsaturated types. 100ul of TMSI is added to a mixture of 400ul Pyridine and 100ul ISTD (In Iso-Octane) making a final solution of 600ul total

[seeviewonder]

We do free glycerin, MG, DG, and TG all in the same run. We use a DB-5HT column with a cool-on injection. The methods are similar, only we start at 50C and it ramps through a program and finally holds at 380C for ten minutes at the end. We use a flow rate of 3 mL/s of helium as the carrier gas.
If you are looking for a good int. std. for DG and TG I would recommend a tricaprin std. from Supelco (Prod. # 44897-U). Hopefully this can help.

[Peter Apps]

My best guess is that the contaminant peaks are siloxanes - coming from the column or from deactivated galss wool in the inlet. Why they only appear every now and then is a puzzle - could some of your samples contain traces of water ?

Peter

[II5tickmanII]

Hmm, its unlikely that they would contain water as we perform KF on the Cocoa Butter samples and they are never above 0.1%, if water is present at all.

My initial thought was Siloxanes from the column/liner wool or potentially small fragments of septa (?) but as you have stated Peter why they are not consistent or reproducible is beyond me. Preventative maintenance is performed on the system such as changing liners, septa etc periodically. I do see it less frequently now but the fact they appear at all bothers me. I dont like unexpected suprises of the bad sort in my chromatograms haha!

The method is obviously of in-house design but i am contemplating moving to ISO 29822:2009

[Peter Apps]

Septum fragments is a good call - if they were small enough they might get baked free enough of siloxanes on the run in which they dropped into the inlet that you do not see them on the subsequent runs. What kind of point do you have on the syringe needle, and what kind of septum are you using ?

Peter

[II5tickmanII]

The needle is a typical 10ul/5ul plunger type for GC autosamplers, the actual needle is approx 40mm in depth and the tip is of the Bevel type (standard not sheathed) essentially a 20 degree slice across the cross section of the needle tip

The septa is a an SRG Auto-Sep T septa which i did not forsee any problems with. It does not state the thermal stability of them but i assumed as they are suitable for a PTV they would be ok? They are installed correctly with the inert "disc" side placed down into the holder.

The Liner is a split precision RESTEK #20762 liner with deactivated inner and wool. It is not a focus liner but a single inner diameter at both ends with exception of the slight flute in the middle whereby the glass wool is held in place.

I am by no means a GC expert and as such not in greybeard territory but i do have considerable years excperience with a GC-FID and havent really ever seen this kind of problem. Maybe im missing the obvious.

[Peter Apps]

[quote="II5tickmanII"]The needle is a typical 10ul/5ul plunger type for GC autosamplers, the actual needle is approx 40mm in depth and the tip is of the Bevel type (standard not sheathed) essentially a 20 degree slice across the cross section of the needle tip
quote]

When I build the perfect universe it will have no bevelled needles in it - they are far harsher on septa than the cone- or dome-tipped needles. I would change it if I were you.

Peter

[jdezeeuw]

Hi
sorry I got into this discussion a bit late.

what we sometimes see is that phase-bleed products are getting focused into concrete peaks when fast cooling down is done. This is because top and bottom of GC columns cannot cool EXACTLY at the same rate and at a lower temperature we get focusing. Some GC's are very bad and some are very good.

Can you cool-down with a negative temp program, say -10C/min, and THEN do a next analysis. This way we allow bleed products to "smooth" out without forming discrete peaks.

It may help. If its helps it still is a bleed problem. you may use a lower bleed column like a Rxi-5 or 1HT series, or even consider metal columns.. the deactivations used on metal columns seem to produce a stabilizing effect making non-polar phases bleed 2-4 times less.

[II5tickmanII]

Hi
sorry I got into this discussion a bit late.

what we sometimes see is that phase-bleed products are getting focused into concrete peaks when fast cooling down is done. This is because top and bottom of GC columns cannot cool EXACTLY at the same rate and at a lower temperature we get focusing. Some GC's are very bad and some are very good.

Can you cool-down with a negative temp program, say -10C/min, and THEN do a next analysis. This way we allow bleed products to "smooth" out without forming discrete peaks.

It may help. If its helps it still is a bleed problem. you may use a lower bleed column like a Rxi-5 or 1HT series, or even consider metal columns.. the deactivations used on metal columns seem to produce a stabilizing effect making non-polar phases bleed 2-4 times less.

Thanks for the suggestions both.

Changing the needle will be a good start as i have been told by other sources that the bevel type needles tend to scrape and "pick" at the septum over time potentially leading to the issues discussed above.

Also, regarding the column its resolving power is a little greater than what we strictly require in our lab. For instance resolving 1,2- and 1,3- isomers is ok, further splitting into sat and unsat is not needed as long as total DGs can be measured.

I will see if i can put a negative ramp on at the end though i dont think the software allows it without a strange pulsating effect in the fan which makes it sound like an unplanned maintenance call waiting to happen. as for purchsing another column i would like to but i think i would have to weigh up the cost vs payoff. Other than those are there any other DG GC capillary columns people would recommend through experience.

Hopefully i will be able to resolve this by the end of the next week. Thanks again for the suggestions they are much appreciated. Will let you know if they lead to success.

[jdezeeuw]

Hi
I like to look a bit more in detail. Can you sent me a C'gram please? For me Chromatograms tell me more then 1000 words.
I have been doing glycerides for a long time and last 5 years it became very actual with the biofuels.

jaap.dezeeuw@restek.com

thanks
jaap