how the internal standard and the blank are applied in the ICP OES


I am brand new user of Agilent 725 ICP-OES. I'm setting up an ICP-OES method for quantification of As, Al, Zn, Fe, Mg, Ca, K, P, Na, K, Ni, V in veg oils and I used Yttrium (Y) as an internal standard in my method. I noticed that internal standard IS stability ratio plot trends downwards. I am wondering what is leading to the IS decreasing from sample to sample. I am using  Agilent ICP Expert II software (version, and I hope someone can help to understand how the software is using the internal standard to generate the elements concentration at the end of the run. Also, I am calibrating the instrument with a multi-element std solution at 2, 5 and 10 ppm concentration including a blank. I am wondering how the internal standard and the blank  are used in the calculation of the calibration curve.  Thanks in advance.

  • Hi,

    I am using ICP-MS.

    In the masshunter software the counts of the IS in your blank are a sort of reference point used to calculate ratios (counts of element (in blank/sample/calstd) divided by counts of IS and then multiplied by 100%. This gives ratios for each element. The ratio's are plotted against the concentration.

    If the IS stability plot deviates a lot from 100% (below 80% or above 120%) then your calblk and calstd's are not correctly matrix matched to your samples.

    Hope this helps 

  • Hi CGGWO,

    Thank you so much for your answer. I am using the ICP-OES with Expert II software and I think the calculation is not as you described, it looks different from masshunter software.

  • From ICP Expert 7.5 (5xxx instruments) help file:

    How Are Internal Standards Applied?

    Internal standardization is applied by taking the intensity of the internal standard measured with the first blank (or standard) and referencing that as 1.0. The intensity of the internal standard in subsequent standards and samples is referenced against the first internal standard measurement and calculated as a ratio. This ratio is then applied to the calculated value of the subsequent standards and samples to give a corrected value. That is, the analyte standards and samples are corrected for any changes which may affect both the internal standard and the analytes.

    Should be the same for Expert II, so FOGs22 was right in his explanation.

  • Hello Andrey,

    Thank you so much for your answer. It's very helpful for me as new user of ICP-OES. Just to make sure so I don't need to correct manually my results as you described, the software has done all the calculation for me, correct?

    Other question please. I am trying to develop a method with the appropriate internal standard for the measurement of phosphorus in vegetable oil (organic matrix). I tried first scandium and then Yttrium. the Sc seems to give higher value compare to Y and the Y ratio keeps falling down during the run, thinking that Y is falling out in the solution (kerosene), I tried stirred it during the run, that improved the Y ratio but still the ratio was 90 % by the 35 analysis. Would it be possible please to provide me with an explanation of why the Y stability  is different from  Sc ?

    Thank you in advance.

  • Sorry, but I don't have any experience with organics, so can't explain it.

  • And yes, ICP Expert does all the calculations, you don't have to do it manually.

  • Again, I am using ICP-MS. In the past I used to measure P in oleochemicals. The samples were directly diluted in NMP (organic solvent). because these solutions contained so much carbon we used a mixture of Ar en O for the plasma gas. If you don't do that the cones became black and partially blocked with carbon really quick. The blockage resulted in less counts the more samples you introduced.

    In the MS I use Ge as internal standard for P determination because of the higher ionisation potential. However that is probably not relevant for ICP-OES.

  • Hello CGGWO,

    Thank you so much for your answer, 

    Indeed, I am using Ar and O2 for the plasma gas. Still I can see some soot formed on the inert tube at the end of the run. Our lab is planning to buy a ICP-MS. Did you notice a large variation in P for the same oelochemical sample when measured with ICP-MS ? would be possible please to know why did you choose the N-Methyl-2-pyrrolidone (NMP) solvent ?

    Thank you so much for you help.

  • Hi FOGs22,

    I have to say I don't know anything about the ICP-OES instrumentation and setup. But if you see carbon in the sample path after the plasma maybe you can increase the O2 a little bit.

    We don't use NMP anymore because we have microwave acid digestion now for all our samples. So carbon is no problem for us anymore and we skipped the introduction of oxygen to the plasma. However, we started using NMP when we bought the ICP-MS first, because NMP is mixable with some water. All our single element standard solution were in watermatrix. So that way we could spike NMP with our standards to setup the calibrations for different elements.

    P determination with NMP was stable. But at that time we were forced to measure it at m/z47 as PO because at m/z 31 there was a significant interference of O2. This also meant that detection limits for P were relatively high. Probably around 2-3 ppm. But this was along time ago. Now all ICP-MS instrument have a collision/reaction cell which eliminates a lot of interferences. Now I measure P at m/z 31.

    Hope this helps!

Was this helpful?