Alpha-tocopherol problem

[Garlarg]

Hi all!

I need to test alpha-tocopherol assay (UDU/CU) in one of our product and I have a problem.
There are five charges of alpha-tocopherol samples.
I inject the standard at the start for five times and once after every 10 sample injectios (UDU/CU). I calculate the system precision from the standard areas, but the areas are increasing consecutively.
For example:

Area: RSD:
1196.52
1234.09
1249.63
1259.31
1263.93 2.2
1279.50 2.3
1306.67 2.8
1371.54 4.1
1417.73 5.4
1442.56 6.3

I should have to keep the RSD under 2.0.

The method is gradient (ACN:H2O) 25 min.
The column is Lichrospher 100 RP-8 5um, temp.: 32C.

I tried to change the column, tried to play with the temperature and tried on another instruments too. Nothing helped. There are two other compunds in the sample so I can't change the solvent conditions.

Please help.
Thanks.

[aceto_81]

your areas are increasing, so this must be carryover, or your sample diluent is evaporating.

Did you try: -inject a blank solution after a standard?
-inject the same solution from different vials, so that
the septa is unpierced for every injection

gl

Ace

[Garlarg]

your areas are increasing, so this must be carryover, or your sample diluent is evaporating.

Did you try: -inject a blank solution after a standard?
-inject the same solution from different vials, so that
the septa is unpierced for every injection

gl

Ace

- I tried to inject blank but nothing special, there is no carryover
- the first five injection in my list are from one vial (where the RSD begin).
- the eluent evaporation couldn't be the problem because the sample areas are proper.

This is the chromatogram. The blue is the first, the red is the last injection:


Same but magnified:

[aceto_81]

Maybe I'm wrong, but it looks like the second large peak is increasing, but the third one is decreasing?

Maybe a sample degradation problem?

Ace

[LCFlo]

Hi Garlarg,

some time ago, I had a similar problem with bad precision of a Tocopherol determination in a semisolid pharmaceutical preparation (I used a LiChrospher RP-18 column).

Standard solutions: many injections possible with good precision
Sample solutions: decrease of area from second injection and bad recovery

I searched with Google for reasons and on hit page 35 or so, I found a link to an Austrian University. There, they described that due to the use of Tocopherole as an antioxidant, it is easily oxidized and analytics might be difficult. The problem shall be solved by addition of ascorbic acid to the sample solution. Thus, I tried to add pure ascorbic acid to the sample before adding the solvent. Then, I mixed, filtered and performed the chromatography: good recovery / good precision.

Maybe this might also be a way for you.

Florian

[Garlarg]

Thank you LCFlo, it's a good idea. I will try it.
Maybe do you have some reference/link to this method?

Thanks.

[LCFlo]

Dear Garlarg,

I just tried to google it again, but I couldn't find the article. But for your information: I use the column already described with a mobile phase consisting of 90% AcN / 10% H2O, Flow = 1.5 mL/min, lambda = 290 nm.

Regards

Florian

[Garlarg]

I tried the ascorbic acid yesterday, but the areas now are not incresing but decreasing the same way.
Sorry but I can't change the method because of the ather two compounds.
I'll try with less ascorbic acid maybe I can counterbalance the area incresing/decreasing.

[HW Mueller]

Why would one use a reducing agent if the problem is not due to a loss of sample via oxidation?
I also wonder why aceto´s comment on the other peak is ignored (though I see the second peak to decrease, the third increase).
The suspicion here is that if you can explain this you have your solution.

(Incidentally, the redox properties of ascorbic are possibly concentration and purity dependent here).

[Garlarg]

Why would one use a reducing agent if the problem is not due to a loss of sample via oxidation?

Because I don't know the end-product of the oxidation. What if the oxidized compund increase the original signal?

I also wonder why aceto´s comment on the other peak is ignored (though I see the second peak to decrease, the third increase).
The suspicion here is that if you can explain this you have your solution.

Believe me the decrease of the 2. (desogestrel) compund is insignificant compared to the tocoferol. At the end the tocoferol area RSD~6.4% and the desolgestrel RSD~0.1%.

(Incidentally, the redox properties of ascorbic are possibly concentration and purity dependent here).

Yes it seems and I'm stucked here.

[HW Mueller]

Garlarg, you may take a look at
http://www.cyberlipid.org/vite/vite0001.htm

This confirms what my first reaction was: the oxidized product should not be likely to coelute with the tocopherol. Also at what wavelength do you operate?

On the RSD: Does this shown chromatogram represent an "outlier" concerning the desogestrel?

Did you do a blank inj. after the red chromatogram?

[aceto_81]

...(though I see the second peak to decrease, the third increase)...

ooops, my fault....

Ace