X-ray Fluorescence: Simplified Instrument Operation Brings Opportunities in Industry, and a Need for Education

Today, many X-ray fluorescence (XRF) instruments, particularly handheld units, are cheaper and easier to use than ever. As a result, industrial use of the technique is on the rise, often involving workers who are not spectroscopists or even scientists. Maggi Loubser, group chief chemist at PPC Cement in Johannesburg, South Africa, recently talked to Spectroscopy about how XRF is being used in industry today, and the resulting need for more XRF training.

Spectroscopy: How do you see XRF being used in industry today?

Loubser: One of the most important recent developments in XRF spectroscopy is the fact that energy dispersive XRF spectroscopy is becoming almost as good as wavelength dispersive XRF. Not only are these systems cheaper, but the instruments are so much smaller, with fewer auxiliary systems needed to run them, and they’re becoming more and more accurate. With these systems, you put XRF spectroscopy within the means of people who could never afford a spectrometer before. We see that with mining companies in Africa. Companies that always had to send their samples away for analysis can now afford to have a desktop energy dispersive system and actually do their analysis at the mine. Instead of waiting days or weeks for results, they can get immediate results, and can manage their operations much better. With fast turnaround times on analysis, if something is wrong, they can immediately take corrective action.

This is true in my industry, too — cement production — where we mainly use XRF for mix design and quality control. We literally have results within five minutes of a sample being taken in the plant, and we’ve got automated systems that report results directly in a feedback loop. If something major is wrong, you can stop the process immediately and prevent damage down the line. When you can make informed decisions quickly like that, it’s better for the entire business.

Spectroscopy: Has the way your industry uses XRF changed in recent years?

Loubser: Definitely. In the past, we mainly used it for quality control (QC), looking at major elemental analysis in raw materials, in mix designs, and for process control down to the final product certification. But in the last few years, environmental legislation has become more stringent, and now we have to do trace element analysis as well. Our work has basically doubled in the last three to four years as a result.

Previously, such trace analysis was done with inductively coupled plasma (ICP), but I have worked for the last two and a half years on a system where we have validated XRF for this work. It’s not the ultratrace analysis you might do for research purposes. We basically want to know if there is less than the legal limit of a heavy metal in our materials, or do baseline studies. XRF is very effective in these ranges and relatively cheap compared to ICP, and of course, the man-hour input is a lot less with XRF.

Another expansion in XRF use stems from our use of alternative fuels. Cement kilns traditionally have been coal burning, but more and more, we are now using alternatives like rubber tires or other polymers that get shredded and burned as fuel. It’s a dual opportunity. On the one hand, you get a cheaper fuel than coal, and on the other hand, you get rid of environmentally hazardous waste that you don’t want in a landfill. Because once the rubber is burned, the remaining inorganic ash is then used in cement. The cement captures whatever nasty elements are in there, and it can’t leach out. So it’s safer than having these materials in a landfill.

But that means we need much tighter controls of what is in the raw materials and the final cement. You also have to make sure that none of the ingredients would have a detrimental effect on the performance of the cement, either. For this kind of work, we’ve been doing more and more XRF.

Spectroscopy: Do you use only XRF, or do you also use x-ray diffraction (XRD)?

Loubser: We see more and more use of XRD as complementary technique to XRF. For many years, XRD was an academic tool and not used much in product or process analysis. But as user-friendly software has been developed for XRD, it is being incorporated into industrial applications more and more. We now use XRF analysis in conjunction with XRD analysis in many of our plants. That way, in addition to looking at the elements in the process, we can look at the phases that are formed. The combined data give you the full picture. That combination of XRF and XRD is also used in other industries here in South Africa, such as heavy minerals and iron ore.

Spectroscopy: Are you using XRF in any automated systems?

Loubser: Everything is becoming more and more automated in large industries, and XRF can be useful there. Because XRF can be almost a push-button technique, once it is set up correctly, you can have systems where a sample can be taken out of a process with an autosampler. The sample comes out of a chute on a robot that prepares it automatically and puts it on the XRF instrument, and the results go through a feedback loop on a laboratory-information management system (LIMS). The minute something is out of limits, you get an alarm. You can almost have hands-free XRF. We’ve got a plant that is almost totally automated and are looking into implementing this in other plants in the future.

This is also wonderful because the quantity of data collected enables you to do data mining. It’s very easy to plot trends. You can also tie in the analytical data with the production QC data of how things were running in the plant at that time.

Spectroscopy: In these industrial settings, there are probably few spectroscopists on staff. Does that matter?

Loubser: I have found that in South Africa, like in many other parts of the world, there has been an extreme loss of spectroscopy skills over the last decade. To a large extent, this is because modern spectrometers have become a black box. People can push a button and get results without understanding what they are doing. So a large focus of mine now is teaching, such as at the short course at the University of Western Ontario that Charlie Wu runs. We want to give spectroscopists the basic knowledge in the techniques that we were all brought up with, because there is less and less of that happening at universities. And we can’t get away from the black box techniques; as a matter of fact, they are becoming more and more prevalent. One of the biggest growing markets is the handheld spectrometers. I jokingly call the people who use handheld devices “cowboy spectroscopists,” because they aim and shoot. And because it’s so easy, because you can aim and shoot, people with no chemical background often use these handheld spectrometers.

Sometimes these users challenge the data that come from a good laboratory, because they don’t understand the limitations of the technique, or use it for applications it was not set up for. They get results from a handheld spectrometer and think those results should always be the same as those you get from a high power wavelength dispersive laboratory spectrometer. They take whatever appears on the screen as the absolute truth. And in that lies the danger. The energy dispersive desktop XRF units and the handhelds are probably the biggest current trend in development, and incredibly exciting, but there is an associated danger. Because they are so easy to handle, many people don’t care to find out what they are doing and use the data blindly.

Spectroscopy: Do you think those users are interested in learning?

Loubser: Oh yes. In this year’s courses at Western Ontario, we’ve had the largest class ever, with students ranging from their mid 20s to their mid 50s. Some of them have been working in spectroscopy laboratories for quite a few years, but they’ve never learned the background, so it’s quite an opportunity for them to understand why things happen. We find that the more people start to understand, the more they hunger for knowledge.