Story of 2 Active Substances in a Dosage Form

[syx]

Short story, we have 2 active substance, just called A and B, in a tablet.
They have different character each other:
The elder, A, 2 mg per tablet, is a basic compound. The retention in reverse-phase column is affected greatly by pH.
Its brother, B, 0.25 mg per tablet, is unionized substance and not affected by pH. B’s retention is more influenced by percentage of organic solvent in mobile phase.

Once upon a time, we wanted to develop dissolution method for the tablets. 500 ml of water is chosen to be the dissolution media. Due to very low concentration, we need to inject 100 ul of sample solution.

Then the problem came, repeated injection from one vial gave highly variable for A though the %RSD of B was less than 1.0. We tried to use higher buffer concentration (from 0.025M to 0.050M), the problem was still occurred… I do not know why. Could someone please explain it…

Note: We use manual injector with 100 ul loop.

[HW Mueller]

What happens if you adjust the pH of the sample to that of the mobile phase?

[syx]

I have tried to change pH value from 6.5 to 7.5. In this condition (percentage of organic solvent is not different) A has longer retention time than B. We still got the same problem.
Today I used the same condition but I slow down the flow rate… the problem is decrease (but %RSD still over 2.0%). When I noticed the chromatogram, the peak heights of troubled series are different just for peak A. Peak B has small variability (%RSD < 1.0) though it has lower area than peak A.

Note: Capacity factor of peak A in unmodified mobile phase is about 1.9.

[unmgvar]

it might be totally off,
but how do you perform your integration?
do you take care to have different integration parameters for both compounds?
many times this is also a reason for bad RSD since the integration on the smaller peak could be very badlly perform.
we have had that problem with new people in our lab when they needed to calculate tablets of SMX-TMP for the first time.

[HW Mueller]

You never bothered about the pH of thre sample?

[syx]

It is dissolution sample in water as the media. Should I dilute it in mobile phase? If I should, I will get very low concentration...

I have checked the integration, it is ok. The problem is not in the base, but in the height. A's height has great variability, though B's in the same solution has constant height and, off course, constant area.

The similar HPLC condition (with less buffer capacity, different injection volume and diluent) has been used to determine A and B in the tablets. It has no problem.

[HW Mueller]

Just adjust the pH of your sample to match the mobile phase and see what happens.

[tom_mizukami]

Since your B is very repeatable I don't think there is anything physically wrong with your system. Perhaps you're basic A material is interacting with silanols in your vials and you are actually injecting different amounts of A depending on the temperature or time.

Does you data trend with time, have you tried polypropylene vials and volumetric ware?

[syx]

Does you data trend with time, have you tried polypropylene vials and volumetric ware?

No and not yet...

[syx]

Here is data of both substances in different concentration. The data of each same level is taken from the same solution.
Lv|********A*******|********B********|
* | Conc | Area | %RSD | Conc | Area | %RSD |
1 | 1.64 | 92618 | 2.25 | 0.207 | 38490 | 0.72 |
2 | 1.99 |114147 | 9.19 | 0.252 | 45615 | 0.24 |
3 | 2.42 |140478 | 4.38 | 0.307 | 56120 | 0.25 |
4 | 3.30 |181450 | 4.34 | 0.398 | 72278 | 0.39 |
5 | 4.00 |213290 | 3.23 | 0.502 | 90270 | 0.10 |
6 | 4.82 |257931 | 5.07 | 0.606 | 109840 | 0.20 |

Note: Conc in ppm.
Suggested concentration from 100% dissolution of active substances are 4 ppm and 0.5 ppm for A and B respectively. (Each tablet contains 2 mg A dan 0.25 mg B; dissolution medium: water, 500 ml).

[HW Mueller]

Now I don´t follow this at all anymore. You inject different amounts of the same solution (the "conc" is a misnomer?)? If so, you probably have a variable ion as well as a pH incompatibility which will effect only the amine.

[syx]

Mr. Muller, I am sorry if I make a misunderstanding here. I mean I made 6 solutions, each contains substance A and B, in different concentration (can be defined as linearity test). I injected them in the same volume: 100 uL, in replication.
You are right that A has amine.

[HW Mueller]

No new suggestion, it just might be wise to put emphasis on integration parameters, integration method (area vs hight), and solvent incompatibilities. If the hights are variable, as you state, the chromatography is not optimal nor robust (for the amine).

[ivanvins]

What kind of column you are using? Basic compunds require usually highly endcapped or base deactivated columns to give reasonably symmetrical peaks. On badly or nonendcapped columns you can observe peak tailing, sorption, saturation effects and other well known problems, resulting in the observed behaviour.

Ivan Vins

[syx]

Mr. Muller, you’re right (again) about ionic strength in sample solution. I mixed sample solution in dissolution medium and 0.1M phosphate buffer pH 7.5 (1:1) and then injected it. The area variability of A is decreased below 2.0%!! Thank you.
It means I should work in low concentration…

I will compare the result I have got with the result of lowered buffer concentration of mobile phase (from 0.050M to 0.025M). The buffer for mixture of sample solution is also reduced to 0.05M.

We use column as you said, Mr. Ivanvins. Assymetry (5%) of A is less than 2.0. Thank you.