Method development for tricky acetone-containing samples on 7900 ICP-MS

Hi TonyGT, I think your variance is matrix related. Could you please inform us of your acetone concentation on the sample vial located in the autosampler? Too much organic highly volatile concentration can indeed affect your plasma conditions. I suppose also you are using the standard nebulizer and the standard torch, delivered with the instrument. Keep also in mind that high organic content can enhance the signal of your low ionizable elements, like As, Se. So I would suggest also a matrix matched calibration. 

Thank you for the reply Efstathios. The final concentration of acetone in autosampler vials is ~0.65%. Matrix-matching with an aliquot of the method blank was one of the first things I tried, and the ISTD signal in the calibrations is more stable than that of the samples. The sample matrix has other organics, but acetone is the only volatile component (I think anyway). Of the other organics, cellulose would be predominant, but I did not expect that alone to lead to instability and high RSD. I estimate the final cellulose concentration to be <10 ppm. Do you think matrix-matching with some sort of cellulose is necessary also? This wouldn't explain the high variance, however, would it?

I am indeed using the standard sample introduction setup and have been keeping a close eye on those components, especially the torch.

Thank you for the reply Efstathios. The final concentration of acetone in autosampler vials is ~0.65%. Matrix-matching with an aliquot of the method blank was one of the first things I tried, and the ISTD signal in the calibrations is more stable than that of the samples. The sample matrix has other organics, but acetone is the only volatile component (I think anyway). Of the other organics, cellulose would be predominant, but I did not expect that alone to lead to instability and high RSD. I estimate the final cellulose concentration to be <10 ppm. Do you think matrix-matching with some sort of cellulose is necessary also? This wouldn't explain the high variance, however, would it?

I am indeed using the standard sample introduction setup and have been keeping a close eye on those components, especially the torch.

Hi again TonyGT, I cannot tell much without seeing the ISTD variance that you are reporting. Your matrix looks really complicated so at the beginning I would say, that this looks like a matrix-effect. Interference like viscosity changes, oveloading the plasma with too much sample, lowering your plasma ionization efficiency with too much elemental content or even the signal enhancement for the not easily ionizable elements like Se, As due to high Carbon content could affect your measurement stability. Try to dilute your samples e.g. 10 times with MilliQ water and then observe your measurement stability if it improves this will explain matrix interference.