• Data Fusion 3: Automat

XRF Scanning and Why We Hate Handhelds

X-Ray Florescence (XRF) has amazing capabilities, and offers a great way to get precise measurements of elements that are heavier than Aluminum (in air) or heavier than Sodium (when using a near vacuum environment or displacing air with a light gas). However, there are a few things that make XRF difficult to use:


1. Typically, handheld instruments are slow.


2. Point data is hard to properly compare to volumetric data, such as assay results.


3. Many companies take the black box’s data as absolute truth. While the XRF measures intensity in counts per second, the black box converts data to ppm. It’s common for engineers to get excited when data or numbers are generated, as they think the data is infallible and end up blindly trusting the box. Problems arise when the manufacturer updates or improves the black box, rendering the models outdated and essentially broken, making a proper comparison very difficult.


Life is too short to spend time hovering over a handheld, taking a single measurement and transferring each one into a software program. The process of XRF measurement begs for automation. Also, does it sound like fun to stand next to an x-ray source for hours on end? It sure doesn't seem like a good idea to us.


So, how do you improve XRF measurement and automate it for the best possible results? Here are the nine steps we recommend:


1. Use fast and precise sensors.


2. Work in counts and not in ppm until you have sufficient data to be confident in the results.

3. Clear the X-ray's / electron's path:


a. Use the thinnest possible sensor windows. (throw out the beryllium)

b. Remove the safety plastic.

c. Flood with helium or use vacuums if you want to measure Na and Mg.


4. Collimate and filter the X-ray to improve the signal for energy levels you want to measure.

5. Use lasers to perfect the distance from rough samples. Follow along with a precise

robotic head that keeps the geometry consistent.


6. Use spectrometers that are fast enough to minimize interference between element energies.


7. Control voltage and throw out the batteries.

8. Pick the best X-ray anode for the problem.


9. Filter the X-ray to improve data quality, if needed.

After you do all of that, what do you get? You get this creation:


Custom XRF Instrument Built for GeologicAI Robots

It may not be pretty, but we love it! Why? Because after four years of data collection, the information Enersoft has assembled is some of the most reproducible and reliable data that we have ever seen.


The robots that run these heads are entirely automated. First, they laser scan the sample surface. Next, the robots run the XRF head at a precise distance from the sample. The scanning is faster than you might think, because the sensors are generations more advanced than what is currently used in commercial handhelds. Enersoft's sensors pick up reasonable quality data in just a tenth of a second. Below are some geological samples containing different key elements. The lines compare different scan times to show what can be achieved in 0.1 seconds. Of course, you can always run the process more slowly, depending on the desired data.


What a Fast Instrument Can Do With Rock Samples

It is exciting to see how much these instruments have improved in the past few years. Enersoft is happy to share some sample data or to help facilitate the use of an XRF instrument on your project. Please give us a call.


Thank you for reading our blog posts! Please subscribe if you love core scanning!




© 2020 by Enersoft Inc. 

  • LinkedIn