GC Method for determination of MeOH, isopropanol and DMF

[alemaggot]

Hi guys,

In this time I'm tring to create a new GC method for the determination of the content of residual solvent in my active drug.

The method is create for headspace. I use DMSO such as solvent. Then DMF has a high boiling point and high DMSO solubilty. I'm using the FET technique. I weight 200 mg of my powder in 22ml headspace vial and I add 10ul of standard solution. I heat the vial at 190 °C to boiling DMSO. In this mode I must not to excract DMF from solution, because all of my solvent are in gaseous phase.
In this mode I've a great DMF peak height, then good area value.

But there is a problem. To work with 10ul automatic pipette is not very good. I've got a relevant RSD % (about 3-4 % for 5 STD injection).

Have you some ideaa to reduce my RSD%?

Thank you so much!

[Peter Apps]

2 possibilities; either weigh the actual addition of standard and make a correction for the actual weight added, or dilute the standard in DMSO at a ratio corresponding to the ratio of 10 ul to the volume of DMSO for each sample. Then you add the standard and the DMSO together in one operation.

Peter

[alemaggot]

Hi Peter,

I'm sorry. But I can't understand what you've write. My english comprension is not perfect!
I write to you my actual standard preparation. I weight X g of MeOH, X g of Isopropanol and X g of DMF in a 25 ml volumetric flask ( All three solvent are weight in the same flask). I fill up to volume with DMSO. After I dispense 10 ul of this solution in n vials with automatic pipette and I inject.

[Peter Apps]

Is the 10 ul of standard the only DMSO that you add to the vial ?

Peter

[alemaggot]

Yes. The DMSO solution, that contain all 3 solvent!

[Peter Apps]

On a 5-figure balance; weigh the vial and cap together. Add the standard. Put on the cap. Weigh again. Calculate the actual weight of the standard that you added. Calculate the actual volume that you added form the density of DMSO. Use that actual volume divided by 10 ul as a correction factor in your calculations of analyte amount.

Peter

[chromatographer1]

You should improve your RSD by using a 10-25 microliter syringe and a Chaney adapter.

I would still weigh as Peter suggested.

best wishes,

Rod

ps: One question, does your 200mg of sample melt during this process? And do you keep your sample valve and transfer line at a high temperature also?

[alemaggot]

I don't have balance with 0,01 mg of resolution. I've only balance with 0,1 mg. Then I can't applicate this method...

About temperature. My analyte has a melting point aobut of 300°C. I use my head space system with 190°C of sample heat, 195°C of needle temperature and 200°C of transfer line temperature. My headspace manufactures instruction report that it can work with over 210°C of transferline tempereture and over 200°C of needle temperature!

[chromatographer1]

The syringe with chaney adapter should be more accurate than a pipette.

If your sample does not melt, if it stays a solid, your assay for solvents may not be accurate.

You have to have the sample liquified (whether in solution or melted) for a Henry's Law headspace method to be valid.

Any government regulatory reviewer will call you on this point.

best wishes,

Rod

[Peter Apps]

I don't have balance with 0,01 mg of resolution. I've only balance with 0,1 mg. Then I can't applicate this method...

Not having access to a proper analytical balance (5 figures) must surely make it very difficult to make up calibration standards etc - what kind of quantities do you usually weigh ?

Peter

[RhysB]

Just as a word of warning, I have found that one of the degradation products of DMSO (I believe it is DMS, but not 100% sure) has a residence time remarkably similar to methanol. Caused me quite a few headaches about a year ago.

[chromatographer1]

I think you are correct. Pesty little analyte it is too.

Rod

[alemaggot]

Just as a word of warning, I have found that one of the degradation products of DMSO (I believe it is DMS, but not 100% sure) has a residence time remarkably similar to methanol. Caused me quite a few headaches about a year ago.

Hi Rhys,

Thank for your observation. But In a blank course I don't found extra peaks. The only peaks that I find are the septa release and DMSO peak. Then I think that I'll not have degradation problem in future. It's right?

Thank you!

[alemaggot]

I don't have balance with 0,01 mg of resolution. I've only balance with 0,1 mg. Then I can't applicate this method...

Not having access to a proper analytical balance (5 figures) must surely make it very difficult to make up calibration standards etc - what kind of quantities do you usually weigh ?

Peter

I know it Peter. I know that is a problem too. But I've it. And with it I must work.
I usually perform weight about from 200mg to 10 g.

[alemaggot]

The syringe with chaney adapter should be more accurate than a pipette.

If your sample does not melt, if it stays a solid, your assay for solvents may not be accurate.

You have to have the sample liquified (whether in solution or melted) for a Henry's Law headspace method to be valid.

Any government regulatory reviewer will call you on this point.

best wishes,

Rod

Hi Rod,

I know that many company work with solid product, without dilution.
If the boiling point, for example, of methanol is about 65°C and I heat my sample to 190°C, why methanol doesn't "leave" all from my product?

Thank you so much