Octocrylene Assay per USP suitability issues

[shaorinor]

Hi all, I am presently having issues with meeting the suitability requirement of a 2% RSD for standard injections.

For those unfamiliar with the assay, the standard and sample prep are straightforward dissolution of octocrylene into acetone. These solutions are then injected into the GC using a split injection.

I'm still fairly new to the GC, but I've always thought obtaining high precision on a run employing a split injection is a nightmare unless you incorporate an internal standard to compensate for the variability of the split. Am I wrong on that one? Unfortunately, the USP doesn't include use of an internal standard, so I'm stuck trying to figure out ways to improve the run within the constraints of USP guidelines in order to pull off the necessary precision.

Anyone have any ideas/recommendations? For reference, the assay is being run on a Perkin Elmer Autosystem GC w/ FID.

Thanks for any help you can give.

[Consumer Products Guy]

Only the USP would detail GC as raw material assay for a compound that 99% of the civilized world assays in the finished product using HPLC.

We actually send our SPF raw materials to an outside lab for such assays, as we didn't want to set up a cGMP-qualified GC, purchase columns, USP reference standards etc. We assay the finished products using HPLC.

OK, your RSD issue: maybe try different liner or liner type, adjust the split ratio, etc. What RSD are you obtaining, like: are you close?

[shaorinor]

Yah, some of these USP assays are a joke. We actually test Octocrylene in finished product via HPLC as well.

I'm using the exact column measurements/phase the USP suggests, though there are probably a few different columns that would match the phase called for by the USP. We've tried reduced the split ratio to make it less of a factor to no avail. The chromatography is actually incredibly ugly even with the USP defined split and it gets uglier with less of a split.

I've gotten it down to about 3% RSD, but it seems to be holding there. With the injections being an hour long, I've had to pull it off the system so we could test other client's samples and get them out of the way.

The assay for Avobenzone per USP is the same way. Dissolve in acetone and run through a split injection on a GC with a 2% RSD. Takes a couple runs, but I'm actually able to get that one to just barely meet suitability.

[DSP007]

Good afternoon.

For a split mode and method of external standard RSD = 2% is not bad. Fundamental error makes no split mode, but the precision of sampling AutoSampler.
Try to set a slow plunger n in the recruitment and fast plunger in the injection, if your AutoSampler allows.
Certainly better to use an internal standard and HPLC, but you seem tied to the method of USP.

Can be recommended to gain a large (10 parallel inputs) matrix injections of each sample or standards, then perform mathematical processing and rejection of "serious missteps" - "a kind of mathematical voodoo" (Student criterion or criteria for extreme points of the sample).

[Peter Apps]

The adjustable (i.e. excluding autosampler engineering etc) parameters that determine the repeatability of a split injection are:

- syringe volume in relation to sample volume - do not go below 1 ul from a 10 ul syringe for example.

- stable sample temperature, organic liquids have quite large coefficients of thermal expansion, if you sample temperature changes their volume and mass/volume concentration changes.

- Injection volume, the smaller the volume the less pressure pulse you get as the sample vaporizes and the more repeatably the pneumatics can control the split.

- pre and post injection dwell and injection speed

- inlet temperature, the hotter the bigger the pressure pulse, but it has to be high enough to vaporise the analyte quickly

- solvent, boiling point and vapour volume interact with inlet temperature to affect pressure pulses

- inlet liner geometry (glass wool, baffles, goosenecks, internal diameter) and deactivation

- split ratio; the higher the split flow the less it is likely to change due to vaporization presure pulses (so reducing the split was the wrong thing to do)

- column dimensions; wider columns can take higher flows which flush the inlet more quickly

- column stationary phase and programme; stationary phase and thermal focussing can sharpen peaks that left the inlet as less than perfect bands.

All this assumes that basic maintenance like septa and liner changes are good.

To get the best repeatability all of this list needs attention. I would start with injection dwell times and speed, and inlet liner geometry.

Peter

[krickos]

Hi Shaorinor

In addition to Peters post I would first like to give you the advice to post a summary of the procedure when asking a question, not all here has acess to USP for details. More details=> quicker help and less questions.

I agree with Peters recommendations, would just like to point out that you do analyze realtively high boilers with polar functions, which typically require a correct chosen liner that preferbly is as inert (decativated) as possible.