I have always used a specification of 20%RSD for the precision portion of a Residual Solvents analysis. Now I am looking for documented support for this specification. Does anyone know where I can get such support. Thanks
By Memerick on Friday, August 8, 2003 - 10:41 am:
What specifications are typical for everyone else when they run a Residual Solvent analysis on GC? Or any GC run?
By Anonymous on Monday, August 11, 2003 - 06:25 am:
In the USP compendium there is a specification of 15% variance. I am not sure if this is RSD or run to run variation. Unfortunately I do not have ready access to the method at this time, but that is where I would start looking.
By Anonymous on Monday, August 11, 2003 - 06:44 am:
I've had no problem meeting 1% RSDs for replicate std injections from a Perkin Elmer HS-40 headspace sampler (no int std). I wouldn't set a spec at that level, but often we are in the 0.5 to 1% RSD range. I'm also very curious as to what RSDs people are setting for HS and direct GC injections without internal stds.
By Anonymous on Monday, August 11, 2003 - 08:36 am:
Split injections in conventional GC have traditionally given 1% RSDs given a Chaney adapter or good autosampler injection. It can be worse but rarely better, depending upon of course the injector system, the solvent, and the analytes involved.
Again depending upon the analyte and its matrix and concentration, while using fixed loop valve (Varian) and timed sample (Perkin Elmer) headspace injections I have gotten 1-3% RSDs.
Going to very low levels (50 ng solvent content per vial) back in 1997 using a Genesis HS (Tecmar 7000) I got less than 5% RSDs from 18 common and USP analytes including Pyridine and Dioxane.
In headspace RSD values depend upon the hardware as much as the analytes and the concentration.
Both PE HS-40 and the Varian Genesis are good instruments. No brand bias was intended if perceived from the above.
By Memerick on Tuesday, August 12, 2003 - 08:23 am:
I appreciate the input. It appears the %RSD we have been getting is right at 10% where as we have an outside lab that is close to 1%. I was really looking for a reference document that would allow an adjustment to a method transfer protocol. The requirement for the transfer was the sample preparations had to have a %RSD of less than 10.
By Anonymous on Wednesday, August 13, 2003 - 06:59 am:
One of the keys to headspace precision is proper and consistent sealing of the vials. This is a real problem, especially on the CTC when using steel seals which require quite a bit of pressure to crimp properly. The new electric crimpers will give a more consistent seal which should help precision. All manual crimpers are not designed the same, and some that will work with a flat top vial do not work well with a crimp top vial.
One place where I worked was having trouble getting precise results from headspace, and when I picked up a vial and tried to check the seal by turning the seal I pulled of the seal and septum. The operator then overcompensated and started over-crimping the vials which also hurt precision. Make sure you are following the manufacturer's recommendations closely.