why is the semi quantitative result incorrect?


1. We recently ran standard and sample in Semi-Quantitative mode by ICP-MS, but the results were not what we expected. see the below picture.

For Li, expected results: concentration of STD2.0 should double the concentration of STD1.0 because we diluted 2 ml of stock standard into STD2.0, while only 1.0 ml of stock was diluted in STD1.0

However, ICP-MS obtained the results without big difference between them. 

2. in data processing, we set STD1.0 as SQStd, one of the blank as SQBlk, and one of pre-run which had internal standard as SQISTD. 

The standard solution concentration was set as the following: Li 25ng/mL, Cr 110ng/mL, As 1.5 ng/mL and 150 ng/mL. Internal standard as auto add (Sc, Ge, In, Te and Bi)

3. What we found counts for each element were changed after the data processing. 

Originally, for Li, counts of STD2.0 is almost double STD 1.0.see below picture 

However, the counts for Li were totally changed and became not proportional.  

for Li counts for STD2.0 is 4429.3; counts for STD1.0 is 3428.98 , blank is 142.86

4. my questions are 

 why did the counts become non-proportional after standard correction? the response of counts/concentration should be the same for the same element. 

do we usually use a tuning solution ( concentration is 10ppb) as a standard solution for semi-quantitative? since mass hunter wrote this statement? 

Can we use other elements as standdard? how do we choose a concentration of standard and how to choose elements of standard for semi-quantitative analysis? 

Thank you

Best regards, 

Di Xu

  • In SQ everything prior to the SQStd is not corrected and everything after is corrected. So any standards prior to the SQStd will have nonsense values. You could try changing the process order and dragging the SQStd above the other standards and reprocessing to see if that helps some. Also, SQ does not use the counts from the full quant table (table where you enter your integration times).

    Instead, it performs a 2 second scan of the entire mass range. Some accuracy is lost, and it can have 30-50% error. The more elements you have in your SQStd, the higher accuracy your results will be in SQ but some elements will still come out poor simply due to high background and poor interference removal on a single quad system (namely Cl, Si, S, etc). But those results can still be used for qualitative studies. 

  • Hi Chelsea, 

    Thank you for answering our questions. 

    I have one question about your answer. do you mean no matter how much integration time I set for each element, SQ only spends total 2 seconds scanning all of the elements ( not 2 seconds for each)? 

    and one more question, Does SQ use 6 points in the peak pattern I set in SQ mode, or SQ has another setting for this parameter? 

    I will try one more time next week. we will try to improve accuracy by adding more elements as you suggested. Let's see the results. 

    Thank you 

    best regards, 

    Di Xu

  • Hello Di Xu, 

    That is correct, no matter how much integration your full quant method uses SQ does a 2sec scan total for every element not 2sec each. SQ uses a 1 point peak pattern. You can actually view the spectra in data analysis, just look for the spectra icon where your internal standard graph shows up, click it and then change where it says tune mode to Quick Scan. 

Was this helpful?