• To participate in the 911Metallurgist Forums, be sure to be Logged-IN
  • Use Add New Topic to ask a New Question/Discussion about Mineralogy or Microscopy .
  • OR Select a Topic that Interests you.
  • Use Add Reply = to Reply/Participate in a Topic/Discussion (most frequent).
    Using Add Reply allows you to Attach Images or PDF files and provide a more complete input.
  • Use Add Comment = to comment on someone else’s Reply in an already active Topic/Discussion.

Semi-Quantitative or Standardless XRF (7 replies)

Maya Rothman
10 months ago
Maya Rothman 10 months ago

What kind of samples do you XRF measure with semi-quantitative (or Standardless) package instead of a conventional calibration? And why? It seems that more and more analysis are made with semi-quantitative (or Standardless) package. Most of the time these applications are already setup by the manufacturer, ready to use right after instrument installation and required low maintenance over time (recalibration or monitoring). So is it always the best method or is it more by convenience?

Helena Russell
10 months ago
Helena Russell 10 months ago

I've found that "standards free" measurements can be misleading, if you don't know how many layers you have in a coating. For instance, if you had a calibration sample of a thin tin coating over nickel, and you calibrated your machine with it, your certainty for tin over nickel would be very good from there on. However if you had a copper component which had a nickel under-plate and tin on top, your software would be confused and the tin coating thickness would be incorrect. The spectrum would give you a clue as copper now would be present, but you wouldn't be able to quantify coating thickness with any accuracy.

Therefore, there are many occasions that you will need some certified samples to re-align the machine, and after a problematic measurement you can assess these sample to make you as the machine operator to think of what may cause inaccuracies in thickness measurement of coatings, or composition of unknown alloys.

I deal with non-ferrous metals, but the idea reads across to other mixed samples, where you may need to know thickness and composition.

There is no definite answer to your question, but you should have some samples of material that you trust to recalibrate yourself as well as the machine.

I've seen a lot of poor XRF measurement in my time, and its always due to lack of calibration and understanding of the fundamental operation of the machines and software used. I hope this sort of helps?

10 months ago
JohnnyD 10 months ago

I think what Maya meant under "semi-quantitative" or "Standard-less" package is Panalytical's Omnian software and it's analogues used for quantitative composition determination on ED and WD XRF machines, and not standard-less (or with only infinite thickness standards) applications in "coating measurement" XRFs (most of them are ED). Speaking of later, I agree with you - one needs to know what he/she is doing;) (I've seen opposite in China few years back, visiting several plating manufacturers, using Helmut Fischer XRF machines, with standard-free applications, who did NOT know what they were doing).

As for former (Omnian, etc.), Maya, I've compared Omnian to conventional silicate calibration of geological samples with fused beads, and it gives decent results. it's good even for preliminary analysis of iron ores (with minute or two additional counting on iron channel). With powder samples it could be worse. I spoke with some users and standard-less is for unknowns only(it has certain stigma), even though it could be replacement of classical calibrations in not a few cases.
I hope, in few years, we will have software able to make users use"standardless" calibrations as first choice, not last.

Maya Rothman
10 months ago
Maya Rothman 10 months ago

 Thank you for your clarification Johnny. That’s right I was thinking about softwares like, Omnian, IQ+, UniQuant, QUANT-EXPRESS, SQX or other types of software using FP (Fundamental Parameter) algorithm or others. These are the one I know but I probably missed some others.

Bob Mathias
10 months ago
Bob Mathias 10 months ago

Apart from many really unknown samples, e.g. scrapings in production processes, dumpings and materials where there are no standards available for, it is certainly very useful to verify the presence of non-certified elements in samples where the composition is supposed to be known for, but in practice is hardly ever complete and/or correct. Then additionally for any sample where the amount or representation form is not in line with the form/amount used for routine analysis. Also to prevent lengthy sample preparation or application set-up in case an indication may be sufficient.

John Koenig
10 months ago
John Koenig 10 months ago

Some years ago I evaluated the use of standard-less analysis of portable XRF for the grade control of the Run of Mine (the ore). For this application standard-less analysis is not the most suitable, and sometimes even using a calibrated equipment the results may not be accurate enough. Some months ago I posted in this group a link to download a paper where I explain why I think using calibrated analysis can be not enough.

In the case of a prospecting project of non-ferrous metals, precius metals, Ta, Nb, REEs, standard-less analysis can be very helpfull to indicate an anomaly content in a sample.

But in any situation it is very important to users have in mind at least the basic concepts to interpret the results properly, what sometimes doesn't happen. I have witnessed some situations where some geologists would desapprove the use of portable XRF in any situation because it is not as accurate as WD XRF (and generally a semi-quantitave or qualitative analyse would be ok). Usually these geologist are very new user and need some technical training to break this barrier.

Victor Bergman
10 months ago
Victor Bergman 10 months ago

I have used UniQuant (standardless/fundamental parameter package) for the determination of elements in dust collected from workplace air. Samples include welding fumes and inhalable dust. It may not necessarily be the best method compared to conventional calibrations however, it has a number of advantages. The main being that for some chemicals suitable dust standards collected on filters may be difficult to make or may not be readily available. Fundamental parameter programs can deal with these types of samples with suitable quality of results for the intended application.

Bill Fraser
10 months ago
Bill Fraser 10 months ago

The fundamental parameters (FP) approach to calibration in XRF is based upon the theoretical relationship between measured x-rays and the concentrations of elements in the sample. It takes into account the instrument parameters (such as tube excitation, geometry, window thickness, primary or secondary filter material and thickness, etc.) as well as the sample composition (spectral and matrix effects). It is generally a good calibration for measuring samples of unknown chemical composition (with the caveat that if there are elements present that are not in the algorithm then this will cause erroneous readings). It may be used successfully for elements in the ppm to 100% level.

Empirical calibrations are created by the user analyzing certified reference materials or well characterized samples to generate a calibration curve for a particular sample set. It is generally suitable only for measuring samples within a narrow calibration range (maybe only 2 or 3 orders of magnitude), and spectral and matrix interferences must be taken into consideration (in some cases they can be ignored if all samples contain the same concentration for example). Analyzing samples with different matrices will give very erroneous results. The biggest problem with empirical calibrations is getting the right number of samples / standards to put into the calibration program. I have seen people use 2 standards and assume calibration is linear, when it is not, and often times reference materials are not available or are cost prohibitive.

Many instruments (laboratory as well as hand held) have several different calibration routines, and the customer should chose the most appropriate for their specific use, based on their particular sample type. Also, these calibrations may be "tweaked" for a specific sample type if the readings are offset. There is no "1 size fits all" in calibration.