by
juddc » Wed May 05, 2010 12:14 am
HI Bruce
Your argument is entirely rational, your points are well made, and worthy of consideration. An IR may indeed be the best way to go in the long run, but it doesn't address the question he asked. He asked how he might go about assaying for dimethicone with an LC.
If this is to become a QC release assay, then the IR method may be necessary. If it's a one off analysis, then GPC might be the easiest way for him to go and to do that he wouldn't need any specialized system beyond the appropriate column(s) (in my opinion). It's his decision. Developing an assay, looking at batch to batch variation of his raw material, then looking at a placebo should give an indication of whether he might have a hope at a reasonable method.
I'd guess that 2-5 days fooling with GPC columns and MP's, then another day looking at various batches of raw material & a placebo and he'll know whether he'll need to buy an IR or not. In my experience, mgmt would look much more kindly upon such a capital expenditure after you show them the results of a few experiments demonstrating that you cannot or should not do the assay with existing equipment.
The NF is lovely, but it's anything but definitive. For instance, the NF specifies GC/FID for analysis of a number of sunscreens, but that doesn't mean that LC isn't an entirely reasonable method to use. I'd argue that LC may actually be superior given that the NF methods are for drug substances and the matrix of a sunscreen product is significantly more complex than a single component - and there's usually lots of water around. GC's hate water, last I checked.
In a pinch, I've used an LC to assay for some things that folks here might consider crazy. Two recent examples come to mind:
Fe2+ (phenanthroline complex, then RP with PDA detection -I saw single digit ppb levels easily). I would have much rather used an AA or ICP, but don't have one and it would have taken longer to send samples out. The matrix was relatively simple, it was a one time deal, and a known addition spike gave good recovery, so it was done.
FD&C Dye lakes (metal chelated dyes). I'd be surprised if you find a single HPLC method for these in the literature, but it can be done and it works quite well. It took me a couple of weeks to get squared away.
The hammer can be used pretty creatively and I'm don't think one should limit one's creativity because it falls outside what the NF specifies. As long as the important variables are accounted for, there's no reason not to give it a shot. If it doesn't work, it could buy him an IR after all.
