I don't have any experience with Horiba, but Bruker has always been very willing to work with users (particularly academic users) on applications like this. It may even be possible to arrange a demonstration or loan through a local distributor of either manufacturer.

I've never heard any of my customers talk about changes in their geological samples during analysis by XRF or XRD. Sample preparation is something to consider though. Fusion and ignition will certainly be irreversible and dissolution or glycolation may present challenges as well.

Which Horriba instrument are you talking about? XGT line is more suited for materials/failure analysis area (I did evaluate one of them once), even though it can give you element map of sample. I am not sure resolution is good enough, though - it is quite a bit worse than that of SEM (1 micron diameter spot). I also never liked combined machines (matter of taste I guess). Mesa-500W looks like any other mini-bench-top EDXRF machine, I think it could do the job, but for that you will need to destroy (crush/mill) your samples - as for any other "conventional" XRF. If you really need chemical mapping, I would look for SEM with ED or WD sensor, in that case you would need only some carbon spattering on your samples in order to make mapping. It would be a bit more expensive, but in most cases mapping will be much better.

That is why i asked about sample destruction, since making pressed pellets and fused glass discs will result in obvious sample destruction.

The Horiba instrument that was advertised in the Elements Journal from February 2014 is this one XGT-7200 X-ray Analytical Microscope. It was unclear to me if samples will have to go trough carbon coating, step that i will gladly avoid because that interference seems unbearable for the sample owner. I will perform diffuse reflectance spectroscopy which is non destructive, but the fact that deriving quantifiable measures from a set of elelments is a hard chemometric task for reflectance alone, wanted to know if anyone around here knows of another way of measuring the major and trace elements that I mentioned above.

Laser Induced Breakdown Spectroscopy is also being tested. But previous experiences indicate problems when measuring Si and Ti.

XTG as any XRF machine doesn't require carbon or any other coating (You don't need grounding, as with electrons). I have few concerns, however - spatial resolution (check if 1 microns spot, giving you resolution of 1 micron +/-, is enough), quantification precision (I wouldn't expect much, even in full vacuum mode, since here calibration is a problem), ability of software to map elements or phases (some SEM do that, I don't know if you need it). You can also try Helmut Fischer XRF machines (upper range is pretty good), even though I don't think they have vacuum and/or X-Ray optics. You mentioned diffuse reflectance - did you consider using FTIR microscopy and/or Raman microscopy? I've seen both actively used with geological samples.

The 1µm spots are perfect, since they are way more small than the diffuse reflectance spot that I am measuring on nature and will allow me to measure the chemistry of the unknown materials. The idea is that we can complement field measurements with lab tests. Problem is that some of the samples that I should measure quantitatively are sacred to the owners, so no destructive analyses could be performed on them. Hence, crushing/milling and carbon coating of unaltered samples (along SEM, EPMA) are out of question.

The other problem with FTIR microscopy and Raman microscopy was the cutting of samples. Sample owners consider the cutting of rock sections as sample destruction as well. Thats why I wanted to know how degraded samples would be from these µXRF spectrometers. Thank you very much for your ideas and insights!

Which system will be appropriate for you depends... Either the mapping XRF or an "environmental" SEM (with soft vacuum, so no coating is required) would work - and there are variants on these instruments too. I understand you don't want to damage the samples in any way. So, the main questions I see are;

Would your sample fit inside the instrument of choice? The largest sample stage in these instruments is around A4 size - not that you can analyse across all of that but you have to fit the sample in. So the range of motion in the "x-y" axes is important. The mapping XRF *typically* has a larger sample stage than SEM - exceptions exist, although they are more difficult to find.

Can the instrument cope with the sample roughness? Once the sample is in, can the optics allow you to analyse the areas you want? So the range of motion in the "z" axis is important.

What analytical size and pattern do you want? These vary from sub-micron in SEM to ~30 micron in the XRF depending on the instrument. You can analyse spots, lines or areas - what do you want for for your research?

What elements do you want to quantify? (remember your samples and the elements of interest in the future will be different). Large sample XRF spectrometers may have different tubes (but typically Rh or Mo) and tube emission overlaps might affect your data. Having said that, the elements you identify in your original post should be OK with Rh or Mo tubes. But mainly, XRF are more sensitive for row 4 elements whereas SEM are more sensitive for Row 3 elements and of course can quantify into Row 2.

A large uncertainty in my understanding is the software that drives quantification. Of course it would always be possible to build your own calibrations, but in terms of inbuilt calibrations to quantify rough samples in variable vacuums - that might be difficult to achieve good accuracy. But also, for the best quantification, you may have problems with analysis of small volumes in your certified reference materials as well. Might be best to check those by mapping the chemistry of a small area (say, 1 cm^2) first before relying on the analytical data from a small volume "spot" analysis. These would be questions for your local supplier or lab that has these instruments.

It may be that one instrument alone won't satisfy your non-destructive elemental analysis needs and you may need to combine data from several instruments. Good luck with your work!

As stated above, there are many points to consider when measuring bulk minerals: I.E. Particle size, Mineralogical Matrix Effects, Analysis Volume, Sample Prep Methods, XRF software modeling and system configuration (portable, bench or high power floor systems).

A 4KW floor system with mapping will address all known and unknown sample matrices but cost may be prohibitive. Most bench systems are 50-200W and may suffice for 10+ppm LLD sample requirements. Bench systems will always have limits in primary filters, crystals, sample handling and power.

WDXRF analysts always consider LLD, element composition and matrix effects when selecting a tool for analysis. Good luck with your research.