Sample Preparation Problem in Residual Solvents Analysis

[Terry]

Dear all,

I am developing an External Standard test method for the determination of residual solvents in an API.

The sample preparation procedure is as follows:
Transfer about 25 mg of the sample, accurately weighed, to a 20-ml headspace vial, add 5.0 ml of water, apply the stopper, mix.

Vigorous shaking of the headspace vial with hand can not get the sample dissolved at ambient temperature. But the sample is dissolved, when the vial is incubated at 80 ℃ for 30 min in Agilent G1888 HS sampler.

I remember the USP <467> requires that the article under test be dissolved in diluent. I have no idea whether the proposed sample preparation above meet the USP requirement. I hope someone can help clarify this.

Many thanks in advance.

Terry

[krickos]

Hi Terry

First a sidenote: 25mg sample sounds a bit low would expect rather 250mg.

In any case. Both USP and Ph Eur have general methods that are very similar for water souleble and non water souleble substances.

If water does not work you can always try DMF instead, I also think USP is about to add DMSO as an alternative.

In any case the problem can occur in all the alternatives in worst case and the problem is not totally uncommon.
The most import thing is that API is dissolved when heating in the headspace oven as you already concluded.

So yes it would be okey but it would be prudent to check different extraction times around 30min to verify that you have reached the equilibrum.

Cheers

[Terry]

Hi Krickos,

Thank you for your prompt reply.

The residual solvents to be determined are Methanol (MeOH) and Isopropyl Alchol (IPA).

I have tried other diluents including DMSO, DMF and DMI, each of them showed significant interference for either MeOH or IPA. So water as diluent is my last resort.

As far as the amount of sample are concerned, 250mg/5mL means the concentration of MeOH and IPA in the aqueous External Standard Solution are 150 μg/mL and 250 μg/mL respectively. I have tried six replicate injection of this solution, the RSD for the peak area of MeOH and IPA are close to 10%. Someone on this board told me the variability is due to the high concentration. Then I diluted the solution 10-fold, the RSD is much better. So the amount of sample is 25 mg.

I would draw a equilibration curve to get the optimal incubation time in the coming days.

Regards

P.S. I have not seen you on this board since you summer holiday, how was it?

[krickos]

Hi again Terry

Thanks the vacation in Greece was excellent, 3year old keep asking us to go back.

September was hectic, been in India for a week visiting colleuges and an external API suplier.

Start to get back on track again

[mbicking]

Try to get a copy of the USP method for comparison. But here are some of the details.

Dissolve about 250 mg of sample in 25 mL water. Add 5 mL of this solution to a headspace vial, along with 1 mL of diluent (water). The sample does not have to dissolve. Any solvents present will be released at high temperature.

Prepare your standards at a concentration that equals the concentration limit in the standard. Add 1 mL to a vial with 5 mL of water.

[Terry]

Hi Micking

Thank you for your insight.

'Dissolve about 250 mg of sample in 25 mL of water. Add 5 mL of this solution to a headspace vial...'

In my opinion, these two bold words mean that the sample have to dissolve. If not, what they get is suspention, not solution.

By the way, the test method I am hitting is an External Standard method, not the same as Standard Addition method in the USP. So the concentration may be different.

Best wishes

Terry

[mbicking]

My mistake. You are correct. Change the wording to "Add" 250 mg to 25 mL water and measure 5 mL of this "suspension."

Procedures A and B in the USP method are "limit tests" based on an external standard comparison. Procedure C for quantification is a standard addition method, although the calculation method is unusual.

[vliu]

I remember reading the RSD limits to be not more than 15%, cannot remember if that was the USP or EP method, so I guess your 10% RSD is good enough for the analysis, no need to dilute.