Can anyone help me with the quantitation of low levels (~0.02%) of benzalkonium chloride [BAK] (as preseervative) in an aqueous injectable suspension containing 2.5% lecithin (an emulsifier for the active drug substance)?

Both substances have very similar character and reliable separation of BAK has proved elusive.

BAK is a mixture of homologous quaternary amines (alkylbenzyldimethylammonium chlorides) containing alkyl groups of various lengths, the primary two being C-12 and C-14.

Lecithin is a lipid mixture of phosphotidyl cholines (also quaternary amines) containing alkyl groups of various lengths (mostly C-16 and C-18). Unlike BAK, it supposedly has some of both acid and base character (zwitter ion).

Solubilities are very similar but literature states lecithin is soluble in ether while BAK is not.

I am able to clean my sample of the majority of lecithin by salting it out using NaCl in acetonitrile. A two-phase system results with the lipids forming a lower layer, and BAK concentrated in the upper ACN layer.

I am obtaining excellent chromatography for the two BAK homologue peaks (C-12 & C-14), using a cyano column with mobile phase composed of buffer and acetonitrile and detection at 214nm. However, I have been unable to achieve adaquate separation of BAK from interferences arising from small amounts of still-remaining lecithin moieties in my sample, no matter what mobile phase component ratio or buffer pH I try.

I have attempted additional pre-analysis clean-ups (e.g. extraction, SAX SPE), but none have proved useful so far.

Can anyone suggest an alternative or a further sample clean-up to remove trace interfering lipids that still remain after salting-out?

Any input would be greatly appreciated.

You might be able to do the separation on a cation exchanger or even straight silica (for SPE sample preparation). You should be able to find conditions, where the quaternary amine is retained, while the phosphatidyl cholines are not. You may need to break the hydrophobic interaction of the phosphatidyl cholines with some organic, while the ionic interactions of the quarternary amines should keep them nicely retained.
Sorry, no direct experience, just should - should - should ....

I may be confused, but is SAX SPE not a strong anion exchange medium? Is the analyte of interest BAK, not a quaternary amine and therefore a weak cation? It seems to me that a cation SPE would retain the BAK while other components pass through. BAK could then be flushed off with an appropriate high concentration cation eluent and collected. Volumes would have to be measured for both the sample sol'n passed through the SPE containing the BAK and the eluent in which the isolated BAK is collected. It might also be possible to concentrate the BAK in the SPE device before removing and collecting it. I also think ion chromatography would be best for analysis, if available. A gradient methane sulfonic acid eluent with suppressed conductivity detection would be my first thought.
Try contacting Ken Conroe at Waters for information concerning the SPE suggestion, KEN_CONROE@WATERS.COM.
Try contacting Dionex for information concerning suppressed conductivity chromatography, WWW.WATERS.COM
My personal e-mail is HJAMIESON99@HOME.COM. I can be contacted by e-mail at work by by selecting my name on this message.

I've had success using an ion-pairing agent (alkyl sulfonate) and low pH when separating benzalkonium chlorides via reversed phase LC. You may have to experiment a bit with chain length, column (C8, C18, phenyl, etc...), and organic modifiers, but you should (there's that word again) be able to find suitable conditions with this approach.