&delta&13C Run Setup Overview (from Jay Brandes, Skiddaway):
Samples for &delta&13C of DIC are often treated with mercuric chloride (HgCl2). Mercuric chloride is extremely toxic! HgCl2 can be absorbed throught the skin. Always wear gloves and eyewear when in contact with DIC samples. Furthermore, be concerned about any potential residues that can end up on the countertops, sinks, secondary containtiners (etc..), and dispose of the used vials with a mind that they can contain fatal concentrations of poison and should not come into contact with humans.
1. Run at 300 microliters/min carrier, 50 ul/min acid, no oxidant. Different flow rates result in slightly different uncorrected isotope numbers, so be consistent. Faster flow rates result in narrower but smaller peaks and slightly degraded standard deviations.
2. Use standard acid concentration, no silver nitrate. More acid does not help.
3. I use the surveyor autosampler and set it up to do 4 replicate injections, every 200 seconds. 25 microliter loop. You could inject less if you wish but I don't see the point in changing it out. I typically take the mean of the last three injections, with the caveat that you should monitor peak areas. The first peak usually is different from the others because of carryover and sample wash issues. An unexpectedly small peak area should be tossed out from the average. You can also use the manual injector, it is more tedious but if you want to run small samples (and you can run 1 uL samples! ) this is the way to go.
4. I have seen no issues with corrosion. I make sure the autosampler is well flushed/washed before and after each run.
5. prefilter your DIC samples through a 0.2 or better syringe filter, taking care to rinse said filter and to fill vial with no turbulence or bubbles.
6. Biggest problem so far has been the autosampler vials. Do not use standard teflon/silicone septa, they allow CO2 degassing and you do see a difference over the course of a few hours. You will want to run standards every 5-10 samples anyway to correct for drift. I use butyl rubber septa but the autosampler has this bad tendency to push the septa into the vial instead of going through it. If it does push the septum down then the first peak will be smaller and garbage isotopically. If you have the ability to cool the autosampler tray I think this will help, but I haven't tried it.
7. I run at 80oC on the ox furnace. I get slightly better results this way.
8. I use water with 1 drop conc. H3PO4 added prior to degassing as carrier. I routinely get 30 mv m/z 44 background levels in the Isolink this way. Degas both carrier and acid as you would for organic samples. I also degas the autosampler wash fluid and use the same 1 drop acid in it.
9. Other issue is standards, best approach is to standardize a bunch of seawater (artificial or otherwise) using the gasbench method and then use this as a secondary standard.