Relative response factors calculation - salts and free acids

[labcat]

Hallo everybody.
Recently I had a discussion with a collegue about the right way to calculate Relative Response Factors (multiplication factor to be applied to the amounts calculated by area normalisation procedure) for related compounds, for an HPLC procedure.
The analyte is the salt of an acid, and sometimes we also analyse the free acid.
We have the reference compounds for the related compounds, some of them are also salts, others are free acids. I corrected all the salts weighings of the prepared solutions to the free acids values, based on the fact that the species that I analyse by HPLC (what I "see" at the UV detector) are free acids. According to me if I consider some of the compounds as salts and some as free acids I'll introduce a bias; in fact related compounds in the "true" analyte (not the one spiked for validation purposes) are of course, all salts as the main analyte, therefore if I calculate some of them as free acid I'll overestimate.
I'll apply the same RRFs also to the free acid analysis.

The other opinion is that I should use uncorrected weighings, regardless if the weighed compound is an acid or a salt.

What do you think about that?

[HW Mueller]

One uses moles.
Incidentally, how do you know that the UV sees "free acids", and what is a "free acid" here?

[labcat]

I realise that I have to give some more details.
The analyte is an active pharmaceutical ingredient.
As I said, the analysis method is an area normalisation procedure; final results are given as percent by weigh.
Related compounds references to be used for the solns preparation (to calculate the RRF) are given to us in some cases as salts, in other cases as free acids.
When I received them as salts, I corrected their weighing to the free acids, by multiplyng for their molecular weight ratios.

Almost all of them has the same "molecular skeleton" of the main compound, and same carboxylic groups; those which have not the same molekular skeleton still are carboxylic acids.
The analyte can be both a free acid or a salt, so I expect that when in a real sample the impurities are also present in the same form, but at the pH of the mobile phase (close to neutrality) all of them are likely to be almost completely dissociated (known pKa are at least 1.5 units below mobile phase pH). This is why I deem it more correct to consider all of the compounds as "acids", as when in the mobile phase environment I can reasonably suppose that they are present not as salts of the original cation.

If I use the moles to obtain the RRF values, then it seems to me that I would have to correct back the found results from percent by moles to percent by weigh, correct?

I know that all these questions may seem trivial, therefore I posted them as "student question", but ultimately it is the accuracy of the release analysis which will be affected!

[HW Mueller]

Right.
(And remember that some of us are swallowing pills).

[labcat]

Many thanks!