Please tell me more about the sample type and the digestion details. Perhaps I can help. Cheers.

The question is very general without giving details, so the answer can only be general too.

1. You must be sure that the measurements are performed correctly and the measurement results differ significantly. Are control solutions used in addition to calibration solutions?

2. It is necessary to be sure that the analyzed samples are physically and chemically homogeneous, and that the test portions (mass taken for digestion) are representative.

3. Is the problem only with trace elements or all?

4. Can the analyzed samples differ from each other so much that a mistake in weighing the test portions must lead to different measurement results?

If the analyzed samples do not differ so much that if they are mixed up, the measurement results will be different, and the test portions are representative, then some preparation errors are likely to be made. Maybe the addition of some reference element to the test portions will help to recognize it?

Such situations usually occur in the laboratory when the performance of analytical activities (preparative, measurement, processing of results) is divided among different people. It is best for the quality of analytical work when all activities with a given sample are performed by the same person, of course with external control using solutions and control samples.

ZJ

1.) sample is non-homogeneous? 2.) you have mis-labeled sample? 3.) you haven't accounted for amount digested? 4.) the nebulizer on your instrument needs attention ------------------> what is significant?

Paul Lechler

The samples that we usually receive are complex samples, often termed as "stabilizers" for PVC. But we also receive samples that simply consist of a single material, such as calcite, magnesium hydroxide, and other metal salts and soaps (metal stearates). Magnesium salts, for example, show greater discrepancies in between repetitions, than say, calcite. With such materials of course, the concentrations we get are usually high, higher than 100 ppm, so I usually dilute these samples, but for mixed matrices such as stabilizers that have not only metals, but also various polyolefins and other organic material, the concentrations usually stay between the calibration curve that we specified (points are 1, 10 and 100). Thanks for the response. Cheers.

Zbigniew Jońca

Apologies, its my first time asking a question here. I'll try my best to provide you with more details to the question.

1- Sadly, the use of QC solutions is not standard practice at our laboratory.

2- We try our best to achieve physical & chemical homogeneity by thorough mixing prior to performing acid digestion. However, it does come to pass that sometimes our solutions do not fully digest inside the microwave, and I'm no expert on microwaves, or their application, but I suspect that its unable to fully direct its power towards all vessels inside microwave, which causes some samples to digest at different levels, even though the solutions look clear most of the time (if there is undigested particulate matter we usually redo the digestion with the addition of HF).

3- We usually do not perform trace element analysis. But I've seen the results having such complications mostly with elements that are expected to have high concentrations in the sample (like Ca in calcite, or Mg in Mg(OH)2)).

4- Maybe, but frankly I do not know how to confirm or deny this.

One thing that struck me is that you mentioned test portions, and whether they're representative. Forgive me my ignorance, I've only been using the ICP for a year or so, but how can we assess representative sample size?

You're also quite right when you say that all steps in the analysis should be performed by the same person to avoid accumulative human error. I will eliminate this at once.

We also add Ytterbium as an internal standard to our samples, which in itself sometimes causes trouble, if, say, by mistake, someone adds a certain quantity of it that deviates from the amount added to the standard solutions. Frankly I don't know how the software calculates the results by using internal standards as a point of reference, but it does happen that significant deviation between sample IS and standard IS concentrations leads to a higher or lower result.

If possible, I'd like your take on the best way to use external control solutions, or QC standards.

Thanks in advance.

Peter Philbrook

1- I would also think of it that way, but we try to assure ourselves of the homogeneity of our samples by thorough mixing in a laboratory blender.

2- It is a common error for sure. But I suppose I wouldn't expect mislabelling to be that frequent.

3- Is there a way to tell how much a sample digested in the acid solution, other than to run it through ICP?

4- Might be, I'll check the nebulizer.

"what is significant"

That part struck me however, I suppose it's a question that I don't know the answer to.