I wonder if there are any mass spec jockeys out there who can comment on the relative accuracy of Glow Discharge Mass Spec (GDMS) with ICP Mass Spec (ICPMS) at the ppm level? In other words, if one has data taken from each and compares them side by side, which should one side with? I have results from both analyses on a metal oxide and I’m puzzled as to which I should stand behind.
If you have one clock, you know what time it is. If you have two, you’re never sure.
One lab breathlessly proclaims that GDMS is linear over 9 orders of magnitude, but is subject to 20 to 30 % error owing to a lack of a valid standard (??!@#?). The same fellow says that ICPMS is accurate to 5 % at 100 ppm, but the error is considerably higher at 1 ppm. Good gravy.
No doubt, the answer will contain the words “it depends”. But I wonder what the issues are.
Lamentations on Chemistry 
I will side with GDMS rather than with ICPMS at ppm level since GDMS is direct solid analysis. The strengh of GDMS is the ability to do the analysis of gases which are impossible with ICP due to various technological problems associated with the ICP source itself. The calibration problem associated with GDMS can be prominent since any solid will have the problem for finding a close calibration standard. This is is a start up problem which are solvable for the routine use. ICP OES or ICPMS can solve the problem of developing working standards for GDMS. But GDMS will outperform ICP or ICPMS after the standarisation of method of analysis is carried out.
Someone suggested that it is possible for a solid matrix to have a minor constituent that has concentrated or bloomed on the surface. This material under glow discharge conditions might show a higher level of other components than one with a more homegeneous distribution within the mass. This argument doesn’t seem unreasonable to me. Any thoughts, GDMS people out there??
Yes, GDMS may show some anomoly in results due to blooming. This is the problem of the sample rather than analysis itself.
The solution lies in the depth profiling for trace elemental concentration for trace elements.
Thanks for the input. I figured that my sample (a metal oxide) was to blame. I had a discussion with a GD mass spectroscopist the other day and (I think) he indicated that they are able to extend the sampling process until the mass spectrum stabilizes in some fashion. It could be that he is referring to the depth profiling that you mention.
GDMS is a truly amazing method. Pity that this method is so obscure for most of us out there. I only heard of it recently.