Loss of one homolog of bensalkonium chloride

[Mattias]

Hi,

I have been asked to develop a better method for benzalkonium chloride in our company, and for the chromatography part there is no problem (Primesep).

However, when I inject from the same vial, the area for the C14-homolog decreases for every injection (about 1% per injection). I don't see the same decrease for the C12-homolog. I doubt that this is adsorption to the vial - why would only one homolog be affected? And the areas only decrease after the first injection (i.e. a untouched vial can stand for days in the autosampler). Could BAC form micelles?

The sample is dissolved in water/citrate buffer. Any good ideas of how to solve this?

[Vlad Orlovsky]

Mattias,

Some of the reasons you might see less of C14;
compounds in your sample which have affinity to the stationary phase and to your compound, this might be highly hydrophobic compound or acidic compound with multiple negative charges.
I would also try to inject higher concentrations of BAC and see if it is concentration/amount related. Changing source of vial might give some conclusions too

[Mattias]

I have now tested different sample solvents:

10% acetonitrile - no difference
0.15% TFA - worse (sample had precipitated)
1% NaCl - much better but still a slightly decreasing trend on C14

The salt concentration of the sample seems to be involved at least.

[Mattias]

I did a quick litterature seach on methods for benzalkonium chloride - and all methods I have seen show low recovery of C14 in accuracy studies.

This seems to be a general problem. The CMC for benzalkonium chloride is 5 mM, and the concentration in most products (incl ours) is well below that. If I remember correctly, below CMC, there is an enrichment of the surface active compound at the liquid surface. I guess that the more hydrophobic C14 homolog will be enriched more at the surface?

Could this be solved by adding another surface active substance in the sample? Just asking questions to myself - but any input will be appreciated.

[Vlad Orlovsky]

Mattias,

Your Primesep column has positively charged surface, so BAC is repelled from the stationary phase. If you play with buffer pH and concentration you should be able to increase repulsion and have better recovery, unless your compound absorbs in the system (injector, tubing, etc.). Also changing ACN concentration might improve recovery, although you might have shorter retention. Between amount of ACN (reverse phase) and buffer (cation-exclusion) I am sure you can find conditions were you have a better recovery

[Mattias]

Vlad,

This is something that happens in the vial - and we saw this phenomenon with our old LC method also. With salt in the sample, I can get the recovery up to >98% - which is probably good enough (but I would really like to find 100%!)

[Vlad Orlovsky]

I guess this is happening due to silanols on the surface of the vial. You can probably increase your recovery if you add TFA to the sample, which will supress ionization of silanols on the surface of vial, or add excess of base (sodium hydroxyde, TEA) to cover silanol sites.

[Mattias]

I will try that, thanks for the suggestion!

If it is not the silanols of the vial, but some effect of the surface activity of BAC - I guess I would be optimal to dissolve the sample in a lot of acetonitrile (to make the hydrophobic ends more comfortable in solution). But then I cannot used reversed-phase (Primesep D) anymore (peak distortion).

[XL]

Hi,

There are several ways to improve the recovery of the C14-benzalkonium:

1. Introduce higher organic solvent in the sample diluent (say, 30% or higher MeCN) to make sure the surfactant won't stick to the sample vial, especially when plastic sample vials are used.

2. Introduce higher ionic strength in the sample diluent to mask the silanol group, which I found was more effective than suppressing silanols at low pH.

3. Introduce a more hydrophobic surfactant that doesn't have chromophore (say C18-NMe3+) in the sample vial to mask the silanols. And/or inject high concentration of such surfactant several time to the LC system before the analyte of interest is injected, to mask the places in the flow path that the surfactant can stick to.

I also attach a poster on surfactant analysis for your information. The Acclaim Surfactant column presented in this poster offers a convenient way to separate cationic, anionic, nonionic and amphoteric surfactants. Please pay attention to Figures 2 and 3.

http://www.dionex.com/en-us/webdocs/388 ... actant.pdf

http://www.dionex.com/en-us/columns-acc ... 25924.html

Feel free to contact if you are interested in more details.

Regards,

[Mattias]

To conclude the investigation:

I ended up using 50% EtOH as my sample diluent, and I have no problems with my C14 homolog or foaming of the samples (it is even not possible to get any foam even with vigorous shaking). The method is validated with no problems, and transferred successfully to our QC-lab. Happy ending!