USP Residual Solvent

[Meerkat]

Hey Forum,
I running this res solv on xylazine per USP vers. 37 (current USP).
Using Class 1 Res Solvent Mixture A.
We observing continual co-elution of "Benzene" and "1,2 Dichloroethane".
Interestingly enough, when I prep a comparison std to the precise concentrations of "Benzene" and "1,2 DCE" made up of lab reagent Benz and 12 DCE, observed that the peak areas are approximate to what would have been if they were separated. So, this shows that the peaks are indeed co-eluting and the area counts are additive.
But how to separate them?
I used: DB624, 30 m x .53 x 3.0 film,
Inlet = 140 C, Flow velocity = 35 cm/sec, split = 1:5.
Oven = 40C, hold 10mins, ramp 10C/min to 240, hold 20 mins (prolly over-kill but hey it's verbatim per the USP meth.)
Detector= 240 C.

chrom picture forthcoming.
Any ideas on how to separate these two "bosom buddies"?

[Consumer Products Guy]

If the "total" is less than the limit, can you report that way?

[AaronAIT]

I achieve pretty ok separation with a 30x0.32x1.8 J&W DB-624. It's not complete separation by any means, but you can make out 2 distinct peaks. When I used to use a column similar to yours, they eluted as a single peak again.

The USP is aware that this co-elution is a pretty common issue - do you require it for system suitability purposes, or are you analysing your API for levels of both these solvents?

[dlbenach]

We use a Restek 1301, 30m x 0.53mm x 3um and get good resolution. To get better resolution, USP allows you to increase the column length by 70% and decrease the ID by a specified amount, I forget what that percentage is. Have you looked at dead volume in your system or a cold spot in the transfer line (assuming your using a loop system)?