March 28, 2024, 08:10:30 AM
Forum Rules: Read This Before Posting


Topic: Interference in mercury analysis by cold vapor generation that depresses signal  (Read 4455 times)

0 Members and 1 Guest are viewing this topic.

Offline javhert

  • Regular Member
  • ***
  • Posts: 18
  • Mole Snacks: +1/-0
Hi everyone.

We received a sample from a proficiency testing scheme for analysis of several analytes, mercury among them. Although the sample was synthetic, its matrix was made to resemble a industrial effluent.

We proceeded with the usual mercury analysis by Atomic Absorption with Cold Vapor generation (CVAA), doing the usual KMNO4 sample digestion abd further reduction of hydroxilamine. Our local regulatory method mandates use of sodium borohydride instead of stannous chloride. When adding it to the sample before reading absorbance, a black precipitate was formed. Result was quite low; however, what was notticed and very weird is that any further samples that were processed gave negative results or quite low recoveries when using the quality control solutions, even when using another vessel for borohydride addition. The entire system had to be washed in order to restore the correct readings.

Haven't found a quick answer for it. My initial guess is that the sample was high on sulfate or sulfur compounts so that sulfide ions were formed, which could generate the mercury sulfide precipitate that doesn't volatilize towards the quartz tube. That however doesn't explain why the following samples could be affected. I suppose there could be deposition of a certain substance in inner layers of the tube that conducts the gas to the quartz tube but here I can only speculate.

Sample was reported to contain 0,1 mg / L when our reading was of about 0,01 mg / L. Notorious metals we found in the sample where copper and iron at the 5 mg / L level and Zinc at the 10 mg / L level.

Anyone has had an issue like this when analyzing mercury? Is there a reactive able to supress this intereference?

Offline javhert

  • Regular Member
  • ***
  • Posts: 18
  • Mole Snacks: +1/-0
Well a little bit of investigation gave this hint. From the EPA analysis at ultratrace levels of mercury:

Gold-coated cloth filter: Soak 2 m2 of cotton gauze in 500 mL of 2% gold chloride solution at pH 7. In a hood, add 100 mL of 30% NH2OHHCl solution, and homogenize into the cloth with gloved hands. The material will turn black as colloidal gold is precipitated. Allow the mixture to set for several hours, then rinse with copious amounts of deionized water. Squeeze-dry the rinsed cloth, and spread flat on newspapers to air-dry. When dry, fold and place over the intake prefilter of the laminar flow
hood.
CAUTION: Great care should be taken to avoid contaminating the laboratory with gold dust.
This could cause interferences with the analysis if gold becomes incorporated into the samples
or equipment. The gilding procedure should be done in a remote laboratory if at all possible.


One possible explanation could be that the sample for testing had gold chloride on it which when reduced formed the colloidal gold precipitate. Particles of it could easily be deposited on the conduct and do a fair job at adsorbing the mercury vapor that would otherwise go onto the quartz tube.

Although it seems a little fancy and expensive, looks plausible for a proficiency test if you feel like giving the labs quite a rough time. I gotta say well played.

Any other theory is welcome, so far I'll be working with this premise. Although it is very unlikely that we ever find a real sample with dissolved gold in it, let's see if there is a way to get rid of it just in case we enter another test. If anything it would look like a mental exercise.

Offline NaCHOargaez

  • Very New Member
  • *
  • Posts: 1
  • Mole Snacks: +0/-0
Hello good afternoon. I am Ignacio. I am an analyst in a laboratory and recently I received a test to determine several metals by atomic absorption among those mercury. I also had those same problems but however I did not know how to solve or eliminate that interference. Could you tell me if you were able to solve it? I would appreciate your response in advance.

Sponsored Links