[the inquisitor]

Greetings all. Having searched Mr Google for the last 2 weeks for pointers, I've become stuck and perplexed with Fe determination by AAS. Especially since there is a lot of Mg in my samples and few documents suggest Fe suppression by Mg. Can someone please give advice on a good Fe determination pathway?

The samples:
50 - 200 ppm Fe
25,000 - 30,000 ppm Mg (a lot!)
~3,000 ppm base metals

I understand that the samples should be diluted to dilute the large amounts of dissolved metals. I always perform x10. Fe in x10 sample ~ 20 ppm.

My calibration STDs have been prepared with amounts of Mg that would be similar to that in the analysed x10 samples ( ~2,500 - 3,000 ppm Mg).

No additions of other base metals are made to the Calibration STDs since the Concentration of the base metals in the analysed sample would be < 300 ppm.

2% HNO3 used for Cal STDs. 372 nm line for Fe used. Conc range about 20 - 100 ppm Fe in HNO3.

By putting such a large amount of Mg in the Fe Cal STDs, I'm suppressing the Fe on purpose and Calibrating the AAS with the given Absorbance. The https://en.wikipedia.org/wiki/Beer%E2%80%93Lambert_law should still apply here since I'm unaware of any physical interference to Fe from Mg. This means that when reading samples, so long as the Mg Conc in the samples are similar to the Mg Conc in the Cal STDs, the Fe result should be true. If there is no Mg in a sample and I analyse it using this Calibration curve, then the Fe signal in the sample will NOT be suppressed but I will get a bias high result.

I have attached a graph to try and show what I'm talking about. A sample with an Absorbance of 0.4 with have a Conc of about 50 ppm or 100 ppm depending on which Calibration is used.

Is there anything wrong with my logic here? My other options are Nitrous/Acetylene flame or use a different Fe line.

Ty .

[marquis]

I think you've got the idea and covered the options.

Running nitrous/acetylene can reduce interference, but it usually increases noise.  Sometimes,the noise is increased dramatically. This leads to longer integration times and other problems.

There may be some merit in looking at other lines.  You might find one with no interference from Mg.  This line will probably have lower sensitivity.  That could eliminate some dilution requirements.  But, it will probably take a lot of work to confirm/validate this.

Good luck.

[the inquisitor]

TY Marquis.

We are in the process of checking the AAS results against ICP-OES.

NO2/Acet flame will be explored if the comparisons are bad.

Kind regards.