I agree it seems to be less “cool” to do XRF compare to XRD. But the contribution to fundamental understanding of the matter cannot be compare. I’m a XRF analyst but I must admit that crystallography can provide a lot more information about matter than XRF. XRF provides information about elemental composition and layer thickness when crystallography determines the atomic and molecular structure of a crystal, the size of atoms, the lengths and types of chemical bonds and confirmed many theoretical models of chemical bonding. In my view XRD help more to understand matter but XRF contributes to the quality of our daily life by ensuring that the products we used are safe, meet their requirements and ensure the quality of many different types of products of the daily life of every human beings. Or even all life on earth!

I agree 100% with what you say. I have been working with XRF (using and selling) for the last 35 years and believe in the technique passionately. What I was trying to say perhaps not eloquently enough, was XRD gets all the “sexy application” attention whilst XRF gets on with doing all the tasks you have listed, but quietly in the background. I’m sure all the Diffractionists out there will take issue with this.

I think that XRF also suffers because outside of Geology it is a technique not often taught anymore in universities, particularly in the US. Many people don't ever encounter it unless they work in an industry that uses it, while XRD is commonly used by materials researchers in a large number of universities. I think for XRF to get more visibility as highly powerful analytical technique we need to get it into the universities worldwide. I think the vendors can do a lot here especially with the handheld and benchtop systems. Universities are often pressed for money, especially those focused on teaching. Vendors often have depreciated or obsolete equipment in their application labs. Imagine if they packaged some of these systems with a little pre-package course material and started donating them to chemistry, material science and geology departments. If you learn the technique in school you are more likely to see applications for it when you get a job. I think these would be market development dollars well spent for the benefit of the industry as a whole. Some food for thought.

I have been working as a "common" user and "professional" application scientist for many years now and never I have had one day being the same as the other. Meaning, there can be a lot of variation and many "exciting" problems to resolved. Of course, when one works in the cement or oil industry, apart from doing some Initial calibrations, there is not much excitement, but those users are probably not interested much in XRF at all. That's fine by me, but in a community like this one I would expect some more respect for XRF, definitely compared to XRD.

I cannot fully agree with selling (or even donatng as I somewhere read) XRF spectrometers to certain communities. We, as vendors, already now have to let our pants down all too often in order to get a sale. Just compare the price of a system today and the price of a system 25 years ago. You will see that prices have actually decreased and that's in absolute values. What can we still offer more? Radioisotopes, that's past tense indeed, Michael. Unless you like calibrating a lot. I STICK TO XRF!

XRF is not taught enough in universities and vendors could promote XRF and themselves in one go. In my opinion, in addition to the relevant inputs that XRF can provide for so many kind of research projects; XRF can offer great benefits to universities for teaching purpose. The fundamentals and the complexity of XRF make it a very interesting topic.

To master XRF, a broad variety of knowledge as well as a wide range of fields of study are needed. Quantum physics and optics of course but chemistry, electronics, mathematics, statistics and all specific fields related to the sample type (geology, metallurgy, petrochemicals, catalysts, environmental) are involved in XRF. Even sample preparation and I have probably skipped some others. So I do see a place for XRF in all university department of science!

I think we are all agreed that XRF is a great technique. I think however, that the very term “XRF” is our problem. To those with only a slight understanding of the technique (i.e. its basic principles), “XRF is XRF is XRF”. However, to those of us who have a deeper understanding of the technique, the term XRF hides a huge diversity of sub-techniques such as Hi-Power WDXRF, Low and Hi-resolution EDXRF, Direct or Indirect (Polarised) Excitation EDXRF, Hand-Held, Micro-Focus, T-XRF, Monochromatic WDXRF. The list goes on. I don’t sell XRF instruments anymore. My company focusses on the sample preparation side of the technique, but I bet I am not the only XRF salesman who was told by a prospect “oh we tried XRF once, but it didn’t work”. Meaning he was sold the wrong kind of XRF sub-technique.

Maybe I didn’t understand your comment in an appropriate way but as Malcolm said XRF is more a diversity of sub-techniques than AN analytical technique.
I do not have your experience but I’m following your path. I have also been working in industrial lab, developing applications, and as application specialist within a XRF instrument manufacturer for many years and, like you, I always enjoy working in this so exciting field that is XRF. If my previous comment reveals a lack of respect for XRF, please receive my sincere apologize because that was not my intention at all. I’m not upset by your comment but I just wanted to clarify the situation and to make sure that you are not upset by mine as well. Viva XRF !

There are many XRF types of instrumentation and they are often marketed and sold in many different ways. Some companies focus on making XRF a commodity and are trying to push for the volume of sales and some focus on helping the customer understand the value proposition of XRF and helping the customer make the best choice to address their analytical problem, even when another technique might make more sense. The former approach can lead to less than optimal approach to sales, i.e. customers being convinced to buy a product that won't do what they need and then ultimately having a bad impression of the technique. I am not involved in the X-ray business anymore, but I my humble observation is that vendors spend too much time fighting each other than focusing on growing the market to new application areas. This will ultimately lead to market erosion, with companies having trouble investing in new XRF technology and product improvements and ultimately having the ability to support their customers. There are ample opportunities to grow the XRF market and a wide range of new application frontiers and industries to expand the technique into. Perhaps their should be a vendor neutral organization to support awareness of XRF and its capabilities.

It is very confusing at the moment to try and decide what make and model XRF to buy, I have also noticed that the quality and volume of new papers regarding new topics and applications has gone down. I know this happens as techniques become more everyday and seem less exciting but it is a shame.  There is not much common effort and sharing in XRF area. Don't get me wrong, information is exchanged, through groups like ours and XRF Listserver, and there even was some effort to write freeware programs (discontinued CSIRO alphas software comes to mind, and there surely were other projects) but that is so much less than XRD community has (look on freeware diffraction pattern database, for instance).

Papers relevant to most users of XRF have been dropping off in numbers over the last twenty five years. Here's a quick question for you all. When did you last look at a copy of 'X-Ray Spectrometry' ? One thing I have seen happening in the UK over the last couple of years is that there does seem to be an increase in University Chemistry Depts buying benchtop EDXRF to help with the interpretation of XRD data.

Thanks for pushing the conversation forward. I think you have identified two key issues. 

Promoting XRF through scientific achievement and communication of the achievements. Since XRF is not taught as a cutting edge discipline in many universities anymore I think it falls on the vendors who are funding their own research to share the non proprietary aspects of basic development in accessible journals and at conferences. There are a lot of fundamental aspects of X-ray physics and optics where there is research relevant to XRF and that needs to be communicated in a way that relates to the tangible benefits of using XRF to solve problems. Sharing information and promoting the technique should be a key mission for all XRF enthusiasts. 

XRF needs to be a technique where data can easily be shared and integrated with data from other techniques. As Michael pointed out, this is something that the XRD community does well. How about a universal data format for all instrumentation to this easier for users to integrate XRF data from many different types of instruments. The universal data format could easily be adopted by XRF users wanting to refine XRD data. A universal XRF data format makes that a lot easier and increases the usability of XRF.

Interesting ideas were suggested. In addition we need to keep in mind that actually XRF is strongly linked and mainly used in industries; where making money is one of the first objective (if not the only one). Thereby XRF manufacturers function in the same way. From that perspective it is not surprising that manufacturers do not want (or like) to share so much their developments. This is a huge difference compare to XRD which is more used in universities where research and publications are their livelihood. So maybe we should also think in another way to promote XRF like:

What can XRF bring to university? (Here Stephen raised up a nice option) or even more ambitious; how can XRF help to make discoveries and breakthroughs in science?

Academics want tools to help them pursue ground-breaking research, whilst Industry needs to improve profitability i.e. reduce cost of goods by controlling raw materials, or improve quality to maintain or improve their pricing. Therefore XRF is often regarded simply as a (very good) measurement tool. Research scientists on the other hand need more “discovery” tools which is why XRD and techniques like NMR and SEM-EDS are very often found in their departments. Perhaps a way forward for our community is to promote the advancement in the micro-focusing XRF systems and encourage the development of software packages for imaging that the SEM-EDS community enjoys.

It is a very good point, micro-focus machines are nice and very useful in still growing electronics market (or in old, but also big plating market), problem is that they represent quite isolated sub-market of XRF machines. Let me explain - from "Big Guns' only Bruker has decent line of microXRF, and some of them don't even consider "microXRF" as XRF and call it "coating thickness measurement". There is no vacuum, though (no pun intended;)) - there is quite vigorous competition between Helmut Fischer Scientific and Oxford, lots of RoHS applications, and very interesting efforts in coupled techniques made by Horriba Scientific (but no new machines were introduced for some time). Sorry if I don't mention other medium-small players, but it's big players who influence XRF future. I can understand unwillingness to meddle in relatively cheap market (top HFS machine was ~$95K - much lower than WDXRF machines by major manufacturers), or unwillingness to compete with their own SEM/EDS/WDS systems, but that's a fact.

I fully agree with your assessment of the common file format, but you were thinking a level deeper than I was. As you say it is very difficult to be able to normalizes count rates / element between different types of instruments. What I was thinking was a common data transfer format for concentration data. There is not a single standard that all vendors use. This can be a problem especially when interfacing with LIMS systems or trying to collect data from different types of analytical data( Sparc OE, ICP, LOI, XRD etc.). some type of reporting tool built into the software to give users a choice of a few different data export formats would be really helpful. On the standards front, a standard format has been needed for some time. It would make sense for vendors to work with industry partners to define one and make it available in their software. If such a standard format were available it could be used for standards data and final concentration data. It would be in everyone's best interest to do this.

Let’s distinguish two processes: data exportation and data importation. Of course I don’t have experience with all XRF software but I worked and tried some of them and they offer at least few data exportation formats which are easily handle with a LIMS system. Here I speak about the major WD-XRF manufacturers and few ED-XRF manufacturers. So in my opinion, offering different data exportation formats is a must-have from an analytical instrument and I guess that all manufacturers know it as the major parts of their customers want it. So from my experience, data exportation should not be a problem.
Contrariwise I fully agree with both of you that data importation should be implemented as soon as possible (or even quicker)! That’s so painful to have to type all these numbers again and again. And at the moment some manufacturers can offer it, they will have a very strong selling point.