Based on my understanding, the sulfur was largely overestimated by ICP-OES. Usually, there are more than one wavelength you can choose. Can you try other wavelengths, which have less intensities but are far away from arsenic's and manganese's.  

ICP analysis ,you can  determination about 78 elements 

To Dr Wahab:

The CRMs are buttons of various alloys we have made by a 3rd party, each batch has a collated COA from 18 other laboratories. I'm not sure what form the sulfur is in the CRMs. Most of the CRM were completely dissolved by the dilute acid. Two of the high Sb CRMs did have precipitates but we used a modified version of the digestion method using a 5% tartaric acid solution instead of deionized water to prevent the formation of precipitate.

To Dr Liao:

We are currently testing other intensities to try find one where there is minimal interference.

To Dr Wahab

Thank you for your input, it appears me and a couple of our analysts will be troubleshooting this for the next week or so. We are currently doing blank runs to check and for ambient S. Our current wavelength is 181.3 nm, but we maybe playing around with that during our troubleshooting. We have done standard runs over the weekend and we get drifts from the COA value of 7% or less. I apologize if I did not include all necessary information, this is my first time using this particular equipment and also the first time we are using the ICP-OES for something other than environmental testing. Gain thank you for your assistance.