Ni dosing in galvanic bath with ICP-OES. Non repetable results

Hello all,

Our lab has acquired a brand new ICP-OES 5800, mainly to run weekly analysis on our galvanic baths.
The method development was going well, but we have an issue with our Nickel bath, as the ICP consistently gives results below what's expected.

Previously, this analysis was done with an AA spectroscope, with results around 75.0g/L Ni.
Now, the ICP hardly gives any results above 71g/L Ni.

Worst, those results are not repetable if the same solutions are analyzed twice.
For example I had 2 dilutions from the same bath, E1 and E2. dilution 1/2000 in water with Ni standard 50/25/10/5 ppm.

They were tested in a first sequence with Intelligent rinse set on moderate, and a second sequence on thorough.
I wanted to do a screening of the main emission rays to check if some would give better results so I selected about 30 rays.
The table below is quite representative of what I'm getting every time - it can be worse.

  Ni 230.299 nm 
Intelligent Rinse Moderate Thorough
Blank (ppm) 0.29 0.3
E1 (g/L) 70.97 73.86
E2 (g/L) 71.27 71.99
QC (10ppm) 9.95 10.03

As you can see, even if it's the same samples that have been run twice, there is a big discrepancy between E1 and E2.
Our first results (70.97 and 71.27 g/L) are way below what our AA gives us (78g/L).

This problem does not occurs when other Ni containing baths are tested, so I was wondering if an interference not listed in the ICP Expert tables could occur.
This bath contains HBO3, does it matter? Can Boron smother the signal and give the flickering results?

If anyone has any idea, please let me know.

Best regards,


  • Hi

    I will suggest to get onsite application support from your local agilent as instrument is new and in warranty...

  • Charles, what is your rinse solution? are you using an internal standard? have you verified with multiple wavelengths of Ni?

  • Hello,
    I use HNO3 1% as a rinse solution.

    Yes, as it was recommended to me, I use Ar as an Internal Standard, specfically the 420.067 nm emission ray.
    Indeed, I checked if this issue happened on other wavelength and this is the case
    In fact, The results gotten at 230.299 are among the higer ones.

    After posting my previous message, I finally realized that the odd results for the blank solution (0.3ppm) was due to a mistake on my part.
    There's been a slight contamination, futher essay with a really clean blank solution showed no presence of Ni - no memory effect.

  • Hi Charles,

    Please be aware that Ar as internal standard doesn't correct for completely for physical interferences. For example, if the nebulizer efficiency drops, due to viscosity, then it's not corrected (completely) by the Ar internal standard. it's good to consider an online internal standard element like Y, Sc, In etc.

    I agree that a application support could help you. I cost something upfront but speeds up the investigation and validation process dramatically. 


Was this helpful?