Method for toluene determination

[alemaggot]

Hi guys,

Today I developed a method for toluene determination in my drug powder.
The instrument parameter are:

Head space: Tequilibrium: 80°C, Tneedle: 85°C; Ttransferline: 120°C. tequilibirum: 15 min. tinjection:0.03 min. Static head space mode.

GC: Colum zebron 624 30m x 0,32 mm x 1,80 um. Toven: 70 °C x 5 min increase 20°C min to 150°C. Tinjector: 210°C. Tdetector: 220°C. Carrier: 21PSI. Split: 5:1.

Standard preparatiron: Weight toluene in flask with DSMO, diluite it in other flask with water to the right concentration.
Vial of standard solution: 5 ml of STD solution
Sample vial: weight powder directly in vial and add the same quantitative of DMSO is present in STD solution and diluite to 5ml with water.

I motivate my choice:
- Diluite with water and not with DMSO because toluene is very poorly soluble in water and, in this mode, the excration by head space system is plus simply and effective. (with dilution in only DMSO the peak respound is little).
- 5 ml of solution in vial because K costant of toluene is very low. Then I work with high sample volume.

However I want ask you some things:
- Toluene boiling point is 110°C. If I work at 80°C in head space oven can I excract all of it from my powder? For this compound I know that the peak area respound dosen't change highly with temperature increase.The best choice for this analyte is to work with high liquid phase in HS vials.

- I discussed with my QA. I support that in the calculation for the standard concentration the volume of STD solution that I put into head space vials dosen't count. If I know that in my powder, for example, I weight 100mg of it, and that the limit in ppm of my residual solvent is 1000ppm I know that the final concentration of my standard solution is 0,1 mg/ml. This value is indipendent from volume in head space vials. If I put 100 ul or 10 ml in head space vial I've the same concentration. The only that change is component respound. Directly dipendent from his K value.

- My QA support, instead, that you must calculate, in head space analysis, standard concentration from the volume that you put into head space vials. Or if I put 1 ml of STD solution I must work with 0,1 mg/ml concentration. If I put 0,5 ml I must work with 0,2 mg/ml concentration ecc ecc

Who is right in your opinion?

If you have same idea to improve my method efficency, let me know please

Thank you very much!

[chromatographer1]

From your post I believe you are doing a limit test for toluene in your drug, 1000 ppm to be exact.

Thus, if you use a 100mg sample weight then 100 micrograms of the 100 mg is toluene at the limit.

1000 ppm of 100 mg = 100 mcg

If you have the drug that contains 0 mcg of toluene then you can weigh out 100mg of this material and can add 100 mcg of toluene to make your limit standard which can be directly compared to a 100mg drug sample that contains an unknown amount of toluene. The amount of liquid you use to dissolve and dilute has no bearing on a direct comparison as long as both samples are dissolved/diluted with the same amount of DMSO/water.

Of course, given the solubilities of toluene in DMSO, and in water, you should add the measured amount of toluene by preparing a known concentration (w/v) of toluene in DMSO and then adding a measured amount of DMSO to add the toluene to the HS vial. For example if your DMSO solution contains 1mcg/mcL (1g/L) then 100 mcL of the solution will deliver 100 mcg of toluene to your vial.

To perform a std addition test then additional solutions of toluene in DMSO must be prepared: perhaps 2g/L and 0.5g/L. Then when adding 100mcL of each DMSO solution will add 200 mcg and 50 mcg of toluene.

If you do not perform a std addition test then you are able only to do a limit test, and any claim of measurement of the toluene will be incorrect, as you cannot claim accuracy unless you have demonstrated and documented the linearity of recovery in your research.

There is nothing improper in adding water to your dissolution-equilibration solvent as long as the drug dissolves completely at the equilibration temperature and remains in solution an adequate time to achieve a reproducible partition of the toluene into the headspace of the vial.

Your procedure of diluting DMSO containing toluene with water in a flask is not a good practice as losses of toluene can occur during this process. It is better to add the DMSO containing Toluene to your HS vial and then immediately adding the required amount of water to the vial.

Unless you have the drug in a dry powder that contains no trapped toluene you cannot perform a limit test as you will not have a comparison reference std preparation to compare against your new drug sample.

best wishes,

Rod

[alemaggot]

Thanks Rod for your reply.

I must create a quantification method, not limit test. I'll use external standard method.

From your reply I've understand that the final concentration of my standard solution depend from the volume of it that I put into the HS vial. Then if I've a solution 1 mg/ml as concentration and I put 1 ml of it into HS vial I've 1 mg of toluene into Vial. Instead if I've put 0,5 ml of it solution into HS vial I've 0,5 mg of toluene into vial. It's correct?

But... If I put 1 ml of standard (diluted in pure DMSO) in one vial and 1 ml of DMSO into other vial and I add 5 ml of water into both vials I change the concentration? Yes. But since this two volume are egual, I don't must change my standard concentration to have 1 mg of toluene into vial. It's correct?
I ask this because I think that is a good idea to add water. In fiirst time I increase the sample volume, and in second time I decrease toluene solubility.

Thank you!

[Consumer Products Guy]

We had a raw amterial that trapped in toluene. We dissolved it all up, then injected by capillary GC at like 40 C.

[chromatographer1]

If you wish to use an external std headspace method to quanitatively measure toluene you will have daunting matrix issues to overcome and I don't think you see the problems in doing it the way you are proceeding. Your error may be minor or it may be considerable.

I won't argue with you if you insist but I can only hope you will change your mind and do a direct injection method with an internal std, or decide to use headspace methodology and use a standard addition method to measure the toluene.

If you persist in following the path you have set out I hope you avoid the troubles I foresee you will have.

GOOD LUCK to you, because you will need it.

Rod

[alemaggot]

Forums are made to discuss togheter. I'm happy for your reply. Also Their are in conflict with my operation

However. You say that if I weigh powder in sample vial and add some solvent is not the same that to use only STD solution in STD vial. Yes. I know that is incorrect. Then I create a "STD powder" only for GC method. I purifiy it from solvent with the use of multiple head space extraction. After I weigh it in STD vial to reproduce the same matrix condition of my sample matrix vial. What you think about it?

Today I've create a problem. My powder is soluble in DMSO but in pratically insoluble in water. If I dissolve it in pure DMSO and after I add water it become turbid. Not good. But I can't use only DMSO for my STD preparation. To extract toluene form DMSO is very difficult.

Idea?

Another question. I've try to put 5ml of STD solution that contatin 0.089 mg/ml (890 ppm for 100 mg of sample) of toluene in water ( 890 mg --> 100 with DMSO. withdraw 1 ml of it--> 100 ml with H2O) into vial and I've found a great solvent respound. But if I use 5 ml of it my standard concentration must decrease, right? From 5 time. It must become 0.0178 mg/ml, right? At this concentration my peak area il low. How can I operato to increase my peak respound?

Thank you very much!

[chromatographer1]

" However. You say that if I weigh powder in sample vial and add some solvent is not the same that to use only STD solution in STD vial. Yes. I know that is incorrect."

I am glad you do.

" Then I create a "STD powder" only for GC method. I purifiy it from solvent with the use of multiple head space extraction. After I weigh it in STD vial to reproduce the same matrix condition of my sample matrix vial. What you think about it?"

If you can make the powder without water or DMSO present in the powder, that would be excellent!



As I have stated so repeatedly on this forum, many HS users try to use too much solvent and sample in their attempt to measure residual solvents.

What is the maximum dissolution concentration of your powder at your equilibration temperature?

For example, will 10mg dissolve into 100 microliters of DMSO at 80C ?

I was able to measure 1 ppm of toluene using 5mg of sample with DMSO as the dissolution solvent (no water or salts) in a 6mL HS vial. At 100mg/mL concentration I only had to use 50 microliters of DMSO to dissolve my 5mg sample. A 10mg sample would require 100 micoliters. Heat this at 80C for 10 minutes and you should have an excellent response for toluene at 100 ppm (1mcg of toluene)

As a point of information, your sample does not have to be dissolved at room temperature to do headspace but at the temperature you use to equilibrate the solution. If your solution is turbid at room temperature check it at your 80C equilibration temp. Is the solution clear?

Can you increase your equilibration temperature to 85 or 90C ?

best wishes,

Rod

[krickos]

Hi

Slightly of topic but related to Rods low sample technique, there is a risk/flaw assosiated with that that can be worked around. Very low sample weight requires that drug substance (DS) crystals are very homogene with regard to residual solvents which is not always the case or unknown in development so you may get a non repressentive sampling. Some DS are dissolved before manufacturing of drug product so crystal size is less important best case there is some delumping after drying, others are seived or even micronized. Also some solvents goes into crystals others stay on outside or do both.

We have seen that issue with up to 25-50mg samples in few cases i.e. in process controls got within specification, release testing out specification (larger weighing). So the work around is using a higher sample weighing, dissolve and dilute and then add a small sample volume to the HS vial and still be able to do a fast HS analysis.

[chromatographer1]

Quite true. krickos

This is not an unexpected result, and if fact it was a motivation in developing small sample analysis.

Back in the 1990s when I was developing a 'generic' method for HS analysis, there was an occasional history of incomplete or inconsistent drying of pharmaceuticals in many of our plants throughout the world. By being able to measure accurately using small amounts of material, a quick comparison of samples taken from different locations in the drying trays and freeze drying equipment was possible. If of course, the results were identical then great confidence in the homogeneity of the bulk drug was possible.

For example, the residual ethanol content of one drug varied from 1000ppm to over 1% depending upon the location of the tray in the dryer. This determination allowed us to modify the drying procedure until consistent drying was achieved.

If the bulk drug is variable in solvent content the linearity of the standard addition method (three small samples from different locations in the bulk container) will be poor and the results will not pass inspection.

Dissolving a larger sample and placing portions into different vials of the one solution will give you the 'average' solvent content of the portion dissolved, and this can be accurate for solvents even with high volatility if the transfer and dissolution of the sample can be done quickly and there is no use of excessive mixing being required to dissolve the sample. This does avoid errors in single sample testing when a sample of inconsistent HIGH or LOW solvent content is removed from the bulk.

However, a moderate amount of inconsistent solvent content throughout the bulk drug may not be detected if large samples are taken. This can be a source of error (perhaps, less than 0.25 % ?) in other testing results (HPLC, etc) which depend upon accurate weighing of drug content of samples.

By using smaller and smaller portions from the bulk drug sample for HS analysis, smaller and smaller inconsistencies in solvent content can be ascertained.

While some of the drugs I was testing were inexpensive (generic painkillers) some experimental drugs were costing $500,000 per gram. Needless to say I was careful not to spill any when I was weighing out my samples !

In summation, taking more than one sample of a bulk drug is a good practice for release testing to ensure that the bulk is homogeneous. Using smaller samples will allow smaller inconsistencies to be detected.

best wishes,

Rod

[alemaggot]

Today I've try to inject 100ul of STD (in DMSO) solution but I've obtained slow respound too...

Until now the best peak respound that I've obtained was when I've use 1 ml of STD solution with the it's second dilution made in water.

However for my routine analysis I use only 22ml vial. The phase ratio from this vial to your 6ml vial Rod is more different. Can be here the problem?

I can't found the best choice of volume to put into HS vial. My matrix is soluble in DMSO. Then I must use it. Then I must use it in STD preparation too. But toluene is large soluble in DMSO, Then I found problem to extract it from STD solution.

Do you have Other test to try?

Thanks for support

best regards

Alessandro

[chromatographer1]

You have not stated the solubility of your drug in DMSO at 80C.

How much drug will dissolve into 250 microliters of DMSO? 25 mg? 50 mg? 100 mg? 250 mg?

If you add water to make a 1:1 solution of DMSO and water, what is the solubility of the drug at 80C ?

Do you need to add 4mL of water to 1mL of DMSO to get good recovery of the toluene? Can you use less water? How much less?

best wishes,

Rod

[alemaggot]

Hi guys!

Finally I've develop the method. At the end I use 500ul of DMSO for totally dissolve 100mg of sample. Standard solution is prepare in DMSO only. I've adjust the split ratio and injection time to have an acceptable toluene respound.

I am satisfiy

Thanks for your support!

Kind regards

[chromatographer1]

I love a happy ending.





Rod

[alemaggot]

Hi guys

Yersterday I've finished to validate this method. It's very good method. With High accuracy and low Standard deviation.
My LOQ is 10 ppm. Not bad, or not?

Best regards

Ale