When I ran analysis samples in college we ran a standard solution immediately prior to each element evaluation and then again immediately after to bracket the element under test. No doubt different these days (that was AA) but still I’m sure it’s not a trivial task if you wish to obtain accurate results.
Yes; this. Unfortunately, there is often an discrepancy between academia and the private sector. Private companies have to make money to survive, and at times, they do what they believe is best for them. And "best" might be a compromise between the most accurate method on one hand, and the most profit on the other.
Therein lies my concern, in part. I doubt these labs are purging between every sample. I know for a fact BS does not; I was there in the lab watching them run the tests for the article I wrote many years back. I doubt they've changed their methodology. Back then, they would run 8 samples of lubes, and then run a check-sample on the 9th cycle using pure mineral oil (IIRC). Why 9th? IIRC, the little trays held 8 sample bottles, so they check after running each full tray. Now, I cannot say if this is proper or not, as I've not been able to ever get to "test" my theory.
I also suspect that other popular labs also don't purge or check between samples. Time is money, as is said; they want to maximize profits. It "wastes" a lot of time/money to do so. Don't for one second think that this potential issue is unique to BS; I suspect they all are at risk.
Specifically, my concern is how the ICP machines transfer the fluids prior to the plasma burn. I'm not really concerned about the burn cycle; I think that's very accurate. My issue is that there is some amount of fluid between the sample container (typically a glass vial or similar), and the piping which draws the sample into the machine, then injects that sample into the plasma stream. Unless you "purge" the full tract, there's going to be residual results in the samples; one relative to the next. For BS, the samples run in the ICP are only 1ml in volume! That's a very small sample, and so it's not inconceivable that cross-contamination between samples in the form of residuals is in play.
When samples are run back-to-back, with no purge or check in between, there are two possible outcomes:
- if the samples are similar in nature, what little cross-contamination which exists is unlikely to greatly skew the results. If sample A has 15 ppm of Fe, and sample B has 17ppm of Fe, then any residual in the sample tract is really not going to push the results badly one way or another.
- if the samples are very different, cross contamination can potentially be an issue. If sample A has 38ppm of Al and sample B has 5ppm of Al, then there's likely to be some effect in the results, one relative to the other.
Now, I want to be very clear ... I have no proof of this,
but it would be very easy to prove out with a few gauge R&R tests. But like I said previously, I don't think any of the labs are interested in the results, as they may not be favorable.
