?

Taray,

Rock Matrix effects - ALL THE TIME! What a mess. The major use of pXRF (portable XRF) is for classifying metals and their grades. That’s real easy, only a couple of elements involved and usually well separated in energy and homogeneously distributed. Now take a rock with lots of, and I mean lots of, elements mixed in a jumbled fashion some near the surface some deeper down some as a host material some as an inclusion, some metallic, some oxides, some silicates etc and you have a mess trying to get a good analysis. A quantitative analysis vs a qualitative analysis are two different animals ?. Barring the interferences of over lapping element energies it’s pretty straight forward to determine qualitative elemental makeup. But when you want to get the how much question answered that’s where the problems come in.

While you may just shoot a rock to get an idea of what and how much, it’s really is not telling much other than what the beam is pointing at which is probably a lot of different materials and then when you take a second shot a couple of cm away it looks different. So you want a small beam size and a camera showing what you’re actually shooting at. To get around that you take a bulk sample crush it to a powder, blend it and put it into a XRF cup with a thin mylar or prolene window film and run that. It helps but it’s not quite that easy, so while it somewhat homogenizes the sample it still has matrix effects that need to be corrected for quantization. You do that by calibrating the sample matrix to a lab standard matrix of similar composition and run a real calibration curve of the elements of interest. If you’re running? grade control at a mine you’ll go to that effort, if you’re prospecting you’ll shoot lots of rocks looking for anomalous hits then bulk sample, crush and run a cup in the lab to make sure it wasn’t some small grain getting your excitement up.

When it really looks good you’ll send the sample off for an ICP/MS analysis and hope for the best.? Matrix effects are grain size, homogeneity, energy and material dependent. Low energy fluorescence is stopped short deeper in and fluorescence itself can fluoresce other elements so it’s a question of how much makes it in and out to the detector. The losses and additions are hard to take care of quantitatively. Gamma spec calibrations are easy as the source is, well, the source while with XRF you put in a source flux and need to calculate and calibrate all the matrix additions and losses and interactions and all at low energy.?

That paper is pretty good and shows the extents a lab needs to address to get reliable and correct results.

I don’t go to that extent nor do I need to as I’m looking for anomalies. I do have a 15 ton pellet press but I only use that for the laser induced breakdown spectroscopy where the laser pulse will blow powdered material all over the place unless pressed into a pellet.

Dud

?

?

Hi guys

Recently I came across an article stating that the largest error in XRF comes from sampling errors.

Flat metals or alloys are the best?

I see crushed? ??and plain??rocks being compared here..

Powdered(unpressed) or irregular non homogenous rocks cause errors.

M.S. Shackley

Elemental Interference During XRF Analysis

While the measurement of fluorescent peaks in XRF seems straightforward, there are a number of interference issues that must be accounted for in all analyses. Many natural rocks consist of several different minerals of highly variable composition and structure. Even natural glasses that are amorphous with no crystalline structure are mixtures of a wide variety of chemical elements. This variable composition causes rocks to affect the behaviour of photons in highly complex ways. These effects on light translate directly to complexities in interpreting the fluorescence radiation that is detected in the XRF spectrometer. The complexities are collectively known as matrix effects which can be subdivided into overlap effects and mass absorption effects.?

Sharing a xrf prep file?

Dude ..Have you seen any difference between these rock sampling?

Taray

?

?

Thanks?

Taray

?