• State Suggested Answer
  • Replies 13 replies
  • Answers 1 answer
  • Subscribers 65 subscribers
  • Views 2221 views
  • Users 0 members are here
Related

Agilent Knowledge Portal

Find maintenance, best practice, and troubleshooting articles.

Agilent Knowledge Portal

ICP-MS 7700/7900 105 Pd has a bad recovery

Daning

Hello,

Recently there is one tough problem surrounding me: 105 Pd has a bad recovery after calibration standards. 

HW setting, Tuning, P/A, and R^2 are all fine. The annual PM was done shortly. No Y or Cu in standards to interfere on mass 105. Nothing wrong with the Pd standard and the In internal standard. Two acid rinse after each CalStd and Sample. Everything seems perfect. However, when re-testing the same CalStd as Sample type (as seen in the pic), 105, 106, 108, and 110 Pd lost their CPS together, which caused the conc. dropped as well. 102 and 104 are better but not ideal because Ru is mixed with Pd in our customized standard. Even though the higher the conc, the better the recovery, the same solution gives two results. This issue only happens to Pd for both 7700 and 7900 we have. Other elements are fine. Has anyone encountered such issue before? Is this a software calculation issue?

Parents Reply Children
  • 0 in reply to TRIUMF

    Hi TRIUMF,

    Replicates:3, Peak Pattern: 3 points, Sweeps/Replicate: 100. This setting works well for other element. It is good that you mentioned the RSD. The RSD was so bad for whole Pd families in calibration blank and 1 ppb standard but returned normal (under 5%) in higher levels:

  • 0 in reply to Daning

    1. high RSD is normal for cal blank and cal 1, seeing their conc is so low. Your RSD is good seeing all other stds have <5% RSD

    2. You didn't use internal std when measuring blank samples of 2% HCL, resulting in 0.1% or 0% recovery, so after the correction of ISTD, your corrected CPS of 2% HCL will be superhigh (corr'ed CPS = CPS/% recovery). No surprise that the conc of 2% HCL is so high.

    You either not select ISTD for calculation, or add ISTD to blank samples. 

    3. I think the masshunter I am using is calculating this way - 

    an unknown is calculated by only the stds right before it. So if an unknown was tested right after cal blank and cal 1, the unknown will be cal'ed by the slope of only cal blank and cal 1.

    After this unknown, you continue to test cal 2 and cal 3 followed by a 2nd unknown, then the 2nd unknown will be cal'ed by the slope of only cal 2 and cal 3.

    But after re-arrange the order and group all cal stds together, the unknown can be cal'ed by all the stds slope.

    That's my experience, not sure if it is true.

  • 0 in reply to TRIUMF

    Those 2% HCl sequences are acid rinse without IS. We need at least 5-points calibration curve based on our SOP. You are correct that calibration standards must be established first and then unknown samples added with IS can be measured in the mode of fully quantitation. I always run all of cali stds and then samples. The problem towards Pd is a myth that CPS dropped almost half in samples. 

  • 0 in reply to Daning

    I don't know either. looking forward to seeing answers from experts.

    Did you use preemptive rinsing? If so, is the number of seconds set too high? can you un-check the box?

Was this helpful?
Privacy Statement
Terms of Use
Contact Us
Site Help