Chromatography Forum

Hello all,

I have been working on a method that has just given very odd results. I have plasma from rats that were dosed with 50mg of mesalamine (5-aminosalicylic acid), which I have extracted with acetonitrile.

My system seems to check out just fine, I have a gradient of .4% phosphate buffer at pH=2.7 and methanol, on a C18 Waters HSS column, and I am analyzing at 254 and 320 nm. My standards have roughly twice the absorbance at the upper wavelength for the analyte and it's degradant.

The problem is, when I inject the extracted rat plasma samples, my absorbance values at the upper wavelength seem to drop to levels that are too far below my 1ug/mL standard to quantitate. The peaks seen in the lower wavelength seem to be unaffected, and end up being larger than the upper wavelength values.

My retention times are very consistent between the two readings, and I am getting really good separation, so I don't think that there is co-elution. I have a few system suitability/standard reinjections, and those don't seem to be affected. So, I am at a loss as to what would cause this. I think it's something in the plasma that is killing my absorbance, but I am not sure what that would be or how to get rid of it.

Any ideas?

I have a quite long experience of analysing 5-ASA, but not in plasma though.

5-ASA has very little retention on C18, and we use cation exchange for this analysis. What k' value do you get for your 5-ASA peak in your system? Having very little retention makes it probable that you have interferences showing up at 254 nm (increasing the absorbance).

A more far-fetched theory is that the 5-ASA peak elutes partly in the sample solvent which may have a different pH that the standard (?). The UV-spectra of 5-ASA at pH 2.7 and pH 7.4 are very different.

Thank you, Mattias.

I haven't calculated the k' value, per se, but I am getting peak splitting on very small peaks when I increase the injection volume in my samples. I see the same behavior in my standards when they are much more concentrated. Clearly, there is something co-eluting with my sample that I can't see. My standards are in 25/75 water/ACN and my samples are extracted in 100% ACN. I hoped those were similar enough that it wouldn't cause an issue. I may try to add a bit of phosphate buffer to my extracted samples so that I can see if it affects the readings.

I did see that several white papers talked about trouble retaining the analyte on C-18 columns, and my retention time is very close after the injection front. However, it is consistent, I am not seeing overlapping with any other peaks, and I am able to separate out related compounds. I think if my peaks were much larger, I would have to re-evaluate that part of the method.

This was intended to be a fast and "almost accurate" method, but those results have just stymied me!

I think you have no choice but to increase the retention of your 5-ASA peak, since you will get interferences from all the unretained components of your sample. The high ACN content of your sample is also problematic.

You can try to add some ion-pair reagent to your mobile phase (10-20 mM of SDS can work). I would also suggest to dilute your samples with this mobile phase as much as you can.

We do this analysis on a Primesep A column, which only requires some TFA in H2O/ACN (but then you need to buy a new column).

I had been using a much slower flow rate, 0.3 mLs per minute, which did increase the retention time. However, I think that might also increase the retention times of all the similar compounds, so I don't end up changing anything in the end. I haven't really found a good method for increasing the retention time of the 5-ASA itself, aside from using 98% MPA at the beginning of my gradient. Any more organic mobile phase, and my peaks start to come out in my dead volume.

I am not very familiar with ion-pairing, though. I have seen SDS used before, but I was hoping that I had achieved some of that behavior with the high phosphate concentration in my mobile phase buffer. Is phosphate not strong enough for the same effect? This probably leads into my next question, why would the high ACN in my samples be a problem? Is that a solubility issue?

Again, thanks for all your help!

What are dimensions of your column? You might have peak splitting, if your retention is too short, and you have a mismatch between your diluent and mobile phase (including any ions).

Changing flow rates is not doing anything because your void moves too

I would suggest considering alretative approach - mixed-mode columns. Mixed-mode will give you much better retention and better tolerance to mismatched diluent/mobile phase.

The phosphate ions you have in your mobile phase will not create any ion-pair effect. The high ACN content of your sample will cause your peaks to be eluted by the solvent in your sample - causing broad ugly peaks. The sample solvent should ideally be equal or weaker than your initial mobile phase. That is not easy to acheive if your sample is dissolved in 100% ACN!

To get good retention of 5-ASA (k' > 2) in this situation you must either use a column with cation-exchange properties or use ion-pair reagents. In the end I think you will save money by going directly to a mixed-mode column (like Primesep A or Primesep 100).

This should be an ideal candidate for a HILIC separation. You would get good retention having the sample in acetonitrile would give you peak compression on the column.
Unfortunatley we do not have any application note for amino salisylic acid but the separation would be simmilar to this application on ascorbic acid and dehydroxy ascorbate.

http://www.sequant.com/files/documents/ ... c_Acid.pdf

If you have any questions regarding HILIC please do not hesitate to contact us at
support@mercksequant.com
or
chromatography@merck.de

I tend to forget HILIC, but Bintang is absolutely correct. With your sample solvent, HILIC should be perfect.