Quantitative water determination with TCD

[SWPM1977]

Is there anyone who was a method to determine hydrates or water in API ???? No matter what the matrix might be ???

[chromatographer1]

Riedel-de Haën should be able to help you with any technical questions.

http://www.sigmaaldrich.com/fluka/appli ... ab1_07.pdf

best wishes,

Rod

[chromatographer1]

To determine water content of any matrix with a single method is very unlikely.

I have performed water determinations of solvents using a R polymer packed porous polymer bead column and a TCD. The solvent was diluted with an anhydrous acetonitrile solvent in order to achieve linearity of the water response.

I have not heard of any TCD method to determine water. This customarily is performed by a Karl Fisher titration which is why I directed you to Riedel-de Haën.

best wishes,

Rod

[Consumer Products Guy]

Decades ago we did assay for water in products and raw materials by GC-TCD using a packed column. But that was when KF titrations used pyridine-based reagents (slower, worse endpoint) and KF units were manual to semi-automatic. Like Rod states: I'd sure try Karl Fischer first. Riedel-de Haën and others have a range of solvents and titrants, and Brinkmann/Metrohm titrators (and I'm sure their competitors) have a lot of information to provide.

[SWPM1977]

We do not have packed inlet/column. It has to be achieved with capillary column.

Method will be used for qualifing the detector.

KF is not an option because it is too hygroscopic and results differ upto 1%. Sample prep has to be done under 5% RH. So for GC it could work.

Any ideas ?

[chromatographer1]

I would use a suitable dissolution solvent that dissolves your API and has a measurable and consistent amount of water (almost all solvents do, even hexane).

I would then try to find a set of conditions that allow you to inject directly onto a capillary column (not on-column) using a packed capillary inlet liner.

Use a splitter only if no other way is possible.

I would use a capillary column which separates water from your dissolution solvent and from any residual solvent found in your API.

Sounds easy, huh? Well, it ain't.

I would try to stay with non-silicone phase capillary columns if possible.

You should expect to replace liners every few injections so deposited API in your liner will not affect recovery of your water.

Good luck,

Rod

[chromatographer1]

I don't understand why KF won't work. It is done with many hydroscopic materials quite successfully. You may need the titrator to be placed in a dry room of less than 5% RH. I have done this personally 15 years ago for a large pharmeceutical company and found the accuracy and low levels of measurement to be excellent.

When you say 1% variation, is that 1% of absolute quanity of water, for example, 4.5% plus or minus 0.045% (which is quite excellent) or 1% of the bulk composition of your API, 4.5% plus or minus 1% (which is not so good).

just curious,

Rod

[AICMM]

SWPM1977,

Thick film CP-Sil1 or SP-Sil 5 60 meters long gives a nice (but not perfect, tails a bit) peak for water around 400ppm. Work was done with a split injection with a high pressure sampling valve into a standard Agilent split injection port. Column was isothermal, detector was helium ionization detector.

Sample was a monomer that eluted after the water. Problem is compounded by the poor flash of the water out of the monomer but that appears to be on the way to being worked out.

Work is, as yet, unpublished because the company won't release it yet. Method is being developed to replace KF since the personnel exposure to the monomer will go way down with GC over KF. Sorry I cannot say more except to say that, depending upon conditions, yes it is do-able.

Best regards.

[jlorier]

Hi,
I've been running water content in most organic solvents for years using a Poraplot-Q column and a TCD detector. We use splitless injection, and get as low as about 10 ppm of water without a problem.