That depends on what applications you are interested in. In general, XRF manufacturers invest more in selected applications and therefore stronger in certain applications. You can see this by the major make up of their customers. Some are stronger in RoHS, cement, mining, handheld etc.

I think you ask the question too broadly as the short answer is that most if not all XRF instruments can analyze geological materials. I think you should think about what kind of elemental range, precision and lower limits of detection you need for your work. Secondarily you should also think about if you want a laboratory based system, portable or handheld unit. All instruments and for that matter vendors come with trade offs. In terms of vendors you should also think about what level of application and service support you want/need. Your location also is a factory as not every vendor has the same level of support in every country. In short I urge you to think carefully about what you need before approaching a vendor as it is not a question of which XRF is best, it is a question of which XRF and vendor is best for you!

Portable analyzers usually come with a lower maximum current setting than benchtop units.

Are you looking for bench-top (laboratory) based XRF or handheld (portable) XRF? Also, what type of samples/calibration standards will you be using?

I agree with the others, you pose a very difficult question! I consider XRF to be one of the widest scientific techniques availabe when you consider all the variations available. Maybe a good start would be to get a copy of this simple but very useful book http://fluxana.com/company/book-x-ray-fluorescence-analysis

Guidelines for XRF analysis:  They have written about all of their knowledge concerning method development for EDXRF and WDXRF. Especially the commodity chapters dealing with the creation of application-dependent measurement programs for WDXRF instruments should be noted. There are recommendations for selecting element lines and the associated background positions. Application-dependent scans show possible line overlaps and influences due to selection of the analysing crystal and collimator. All important parameters are addressed and their influences explained for each application. Wherever possible, comparisons to EDXRF are also made. In my opinion, this book is a must for every developer of XRF methods.

A good starting point, from a technical point of view, is the range of energies required. If you are looking for a specific set of XRF energies then the choice of accelerating voltage and/or filter material will need to be chosen based on that range.

Handheld (portable) x-ray systems are likely to offer a fixed kVp setting and changing filtration so as to tune the excitation energy, based on the user's needs.

Benchtop (lab-based) systems offer a range of kVp settings - full freedom is available here and, additionally, custom filters can be used. There is a greater degree of freedom and stronger anode current - though, sometimes this is not necessary.

These are some advantages and disadvantages of handheld XRF:

1. Does not have vacuum or gas purge chamber. Therefore, their detection range normally range from Al to U. On the other hand, bench top units normally has detection range from Na to U. Some higher end bench top are be able to detect from C to U. (Shimadzu EDX-8000)
2. Handheld would be advantageous when you wanted results onsite and when you do not want to cut samples to smaller size to fit into the XRF chamber.
3. If you have lots of sample, a bench top with an auto sampler can save you time where you do not need to wait and reload one sample at a time. This is not available in a handheld unit.
4. With a lower powered X-Ray tube, if your interest are some very heavy rare earth, Handheld may not be strong enough to excite sample and needs to use secondary lines.

If you are interested in an XRF for research you should look at the Bruker Tracer III-SD. This is the only XRF instrument made in a handhled form that has advantages over everything else out. You can create your own filters with the Tracer III. All others come standard with a yellow, blue, green, and red filter. You can create your own custom calibrations based on your own materials. You can also set your own current and voltage instead of the manufacturer setting it for you. Finally, the best for last, they are the only handheld on the planet with a vacuum allowing for lower levels of detection. The icing, you can purge helium through it and get even lower! Find out for yourself. Oh and here are just a few institutions with multiple units for research. I can give you a list of the best of 1000 universities and museums in 40 countries that have Tracers. I can send you several publications done using this instrument for research.

1. Gallery of Art Washington DC USA (3 Tracers)
2. Institution USA and Panama (8 Tracers)
3. Museum Los Angles USA (3 Tracers)
4. Museum of Art New York City USA (3 Tracers)
5. British Museum London England (1 Tracer)
6. Historic Scotland Edinburg Scotland (1 Tracer)

It would be nice to to see more informative posts and not sales pitch. Could you explain what blue, green, yellow and red filters have to do with XRF instrument? Do you mean optical camera? While vacuum chamber looks like nice (even wouldn't work for all samples) feature, how does helium lets get even lower, could you explain?

I apologize to everyone for making this sound like a sales pitch. I do not work for Bruker directly. I was trying to help. Now to answer your question. There are specific filters that are used in a filter wheel, whether automatic or manual, that allow certain energies on the spectrum to be better targeted allowing for better flourescence of particular elements. For example, the yellow filter might contain 1 mil of Titanium and 12 mils of aluminum and could be better used to excite lets say Fe to Rh. With respect to vacuum, people are running core samples for oil and gas and are getting very low levels in the lighter elements such as Mg, Al, Si, P, S, and K. I see no reason this could not work for most applications. I have personally used this on non-uniform samples such as soil, mining, and core. I would think this could work just as well on more uniform sample as well. You would have to ensure a proper window is used as not to break and suck contamination into the instrument. In addition, using the helium, people have now been able to see Na as well as the other light elements to even lower levels. In using helium you remove the window and purge helium through the instrument which further helps to remove air from the detector and clean the surface of the material you are testing. Without sounding like sales, I have seen this in action and it is currently being used today primarily by folks in oil and gas core analysis. I have not seen this used in other applications yet. Of course, all that I speak of here I have only seen in handheld XRF. I have not seen this in action in a labratory benchtop XRF yet. Not sure if that is available anywhere or not. Hope this helps. Again my apologies for my previous postings. I am new to posting on here and did not intend to upset anyone by trying to sell something. Thank you.

This may help too:

http://www.911metallurgist.com/blog/compare-handheld-xrf-analyzer-brands

I also need to purchase a handheld XRF analyzer within 6 weeks, and still have a lot of researching to do.

The application is for PMI of incoming material in a manufacturing environment (products constructed primarily of sheet metal). We buy four alloys of stainless, two grades of aluminum, and also Galvalume and Galvanize steels. We take in an average of thirty coils, four times per week. I intend to develop a program where we'll put this tool in the hands of operations team member, and minimize potential for errors in logging data (wifi may be the ultimate feature). 

In this environment/application, I am looking for something low cost / high durability (I anticipate the tool will be dropped), and preferably an established program to provide a working backup tool in times we need to have ours serviced, perhaps as part of an optional warranty. 

So far I've found the tools to be as low as $17K (Spectra) up to $35K (Niton). Olympus seems to have a program in place to provide backup. Have yet to look into Bruker and Oxford. 

If you have any suggestions or comments about your experiences with any of these tools for similar application, your input would be greatly appreciated, thanks!

Hi James, I get great service/cost from https://www.911metallurgist.com/profile/DH/