Sunscreen drug product assay

[jeputnam]

I am trying to develop a method to assay oxybenzone, octocrylene, avobenzone, octyl salicylate and homosalate - they are all present in one product. I have developed a couple decent methods (Gemini C18 5um 86/14 MeOH/pH5 buffer and Prodigy ODS3 5um MeOH/pH2.1 gradient), but they each failed validation. I'm having two problems with resolution: (1) avobenzone splits into two peaks - one large and one small, the small one co-elutes with oxybenzone; and (2) when I try to alter conditions to get better resolution between avobenzone, cis-homosalate, octyl salicylate, and trans-homosalate the peaks get broader and continue to over lap. Any suggestions? I've done some literature searches, but I haven't been able to find anything out there with these 5 analytes.

[Consumer Products Guy]

I'm going to assume that your samples are dilute enough, and you're injecting 5ul or less, as sunscreens are great UV absorbers and are present in high concentrations. I don't believe we've ever done a mixture of all five of these, but I can share that we've used C-18 and mobile phases containg THF. For proprietary reasons I can't provide more details on this. It's not a new reference, but you may want to review J. DiNunzio and R. Gadde, Journal of Chromatography 519 (1990), pp. 117-124

[Bruce Hamilton]

If your peaks are splitting, it may be due to a mismatch between your initial mobile phase and the sample solvent.

Try dissolving the samples in the initial mobile phase, and perhaps adding a third component to your mobile phase and sample solvent, conveniently suggested by CPG.

Please keep having fun,

Bruce Hamilton

[jeputnam]

I appreciate the advice, but the homosalate is always present as 2 isomers. Also, the mobile phase is methanol/buffer and the extraction is in methanol.

[Bryan Evans]

Try reducing your injection volume (as CPG discussed) and see if that
helps with peak splitting. The peak distortion can occur if the solvent
strength of your sample diluent is much stronger than your mobile phase.

There was an interesting poster done on this topic at ASMS. The
author injected 5uL, 10uL, and 50uL of 100% acn into various ODS
phases. Certain ODS phases handle this better than others.

If your interested - shoot me an email and I can forward it on to you.

[jeputnam]

Maybe I didn't use the right terminology to describe what is happening. What is happening is not a single peak splitting into two peaks, but an analyte become two species - homosalate becomes cis and trans isomers, and an unknown isomer splits off the main avobenzone peak. This is not a chromatographic problem - these species are really there. It just means that I have seven peaks to resolve rather than five.

[Bruce Hamilton]

I assume you have performed a literaure search but, if not, some of the following might offer some useful guidance. a more specific targetted search could find more relevant publications.

Bruce Hamilton

Assay of common sunscreen agents in suncare products by high-performance liquid chromatography on a cyanopropyl-bonded silica column.
Silvia Simeoni, Rosanna Tursilli, Anna Bianchi and Santo Scalia,
Journal of Pharmaceutical and Biomedical Analysis
Volume 38, Issue 2, 15 June 2005, Pages 250-255

Determination of sixteen UV filters in suncare formulations by high-performance liquid chromatography
D.J. Schakel, , D. Kalsbeek and K. Boer
Journal of Chromatography A
Volume 1049, Issues 1-2, 17 September 2004, Pages 127-130

High-performance liquid chromatographic assay for common sunscreen agents: application to in vivo assessment of skin penetration and systemic absorption in human volunteers.
V. P. Sarveiya, A. Hall, and H. A. Benson
Journal of Chromatography (2004) B, vol. 803, no.2, pp. 225-231

[Noser222]

Maybe I didn't use the right terminology to describe what is happening. What is happening is not a single peak splitting into two peaks, but an analyte become two species - homosalate becomes cis and trans isomers, and an unknown isomer splits off the main avobenzone peak. This is not a chromatographic problem - these species are really there. It just means that I have seven peaks to resolve rather than five.

Just a quick thought, but could increasing temperature speed up the conversion so that you get one peak?

[Uwe Neue]

Homosalate has two steric centers. One at the oxygen that attaches the cyclohexane ring to the rest of the structure, and one at the methyl group at the cylohexane ring. This gives two chemically distinct species that can be separated by RP.
Avobenzone does not have such a thing. Either the second peak of avobenzone is a degradation product, or it is an injection artifact. I would think that the 1,3-ketone is rather sensitive - to light, pH etc., and sample preparation could be critical. This could explain the degradation. It is also possible that artificial peaks are created from injecting the sample from a sample solvent that has a much higher elution strength than the mobile phase. This would cause peak distortion for the early peaks in a gradient, but no more for later peaks.
I suggest that yoy tell us the composition of the sample solution.

[frododawn]

I am having the same problem trying to separate Cis and Trans Homosalate and Octisalate. All of the methods act as if there isn't a problem here. One of the isomers of Homosalate runs at the same time as Octisalate. Does anyone know how to separate these?

[Consumer Products Guy]

I am having the same problem trying to separate Cis and Trans Homosalate and Octisalate. All of the methods act as if there isn't a problem here. One of the isomers of Homosalate runs at the same time as Octisalate. Does anyone know how to separate these?

Yes, homosalate isomers will elute as two peaks. The following chromatogram shows one way we do this, our cGMP-validated procedure. I didn't have this technology a few years ago when this topic was first discussed.

Isocratic on RP column. I can't post the mobile phase or column type, proprietary, sorry.

If you can't read the identifications, last peak is trans-homosalate, preceded by octisalate, that preceded by cis-homosalate, preceded by avobenzone and first octocrylene. Very early is our solvent peak.

[paulw]

I performed sunscreen assays a number of years ago. We would do octinoxate, octisalate, homosalate, avobenzone, oxybenzone and octocrylene all in the same run with many of them in the same product. If memory serves we used a SB-C18, 4.6x250mm, with a mobile phase of 85MeOH/15H20/1Acetic Acid/~1.5g triethylamine HCl. The run time was about 26 minutes for all separation.

I'm not saying the method is perfect, but I'm saying it worked. If someone got a little sloppy with making the mobile phase, then we could potentially lose resolution between some of the peaks.

[frododawn]

Awesome!! That helps alot. I appreciate it!!