Dr. René de Kloe, Applications Specialist, EDAX
All our senses are aimed at observation. We feel, see, hear, smell and taste things to experience the world around us. We are relying on our senses to make many of our day-to-day decisions and choices. And especially in the upcoming holiday season, shops and companies in the business of selling things cleverly use shiny advertisements, brochures, fragrances, and unbeatable product descriptions to entice us to select their wares. All the time hoping that we will succumb to our senses that focus on the superficial appearance of products before thinking things through.
We must be very careful not to let this very successful marketing strategy subconsciously guide us when analyzing materials as well. During our work as microscopists we are continuously selecting samples, cutting and preparing them to expose a feature of interest, and then choosing the analytical tool and actual analysis area. How sure can we be that we really get representative and objective information?
As a geologist, I was taught to take your distance from a rock outcrop and look it over before going into any detail, knowing that the context of your observations is crucial in your interpretation. Then I would go in close to look, feel, and yes sometimes actually taste the rock in order to try to identify what I was actually looking at and how the overall structure fit in the geological setting of the area.
Observing this distance is crucial for your understanding of structures, but in some cases, you cannot get out far enough to see the bigger picture and then you must make do with what you can see.
Perhaps an extreme example is what I did for my PhD research. I have studied the occurrence and distribution of nm-scale films of amorphous material along grain boundaries in experimentally deformed rocks that originate deep inside the Earth. In total I may have characterized a few cubic microns of material but based on that I tried to draw conclusions on the effects of these melt layers on the movements of entire continents!
In microanalysis, we are suffering from the same problem. Microscopy inherently means that you cannot look at the wider picture and when you are looking at extremely small-scale features, their size combined with a practical image resolution may limit the observable surface even further. And one of the most difficult questions you then must ask yourself before starting an analysis is, if the analysis area is representative. And that can be a really tricky question. How objective are we all when browsing the sample surface to find a spot to collect the data? Don’t we all tend to preferentially pick an area that looks promising? I am not so sure that that would always be the most representative region.
It is not that long ago that the acquisition limits in EDS and EBSD were caused by the detector technology. For EDS mapping, we were quite happy if you could collect your data with 50,000 input counts per second and a 50% dead time. This meant that when you were collecting a 512 x 400 pixel map where you wanted to have, say 1000 X-ray counts per pixel, it would take you a few hours. And after that time someone else would be hovering behind you, eager to use the microscope. This seriously limited the sample area that could be analyzed and as a researcher you needed to think carefully about your analysis strategy to get representative information.
The area that can be analyzed has changed dramatically with the introduction of the latest EDS detector technology. These detectors are capable of processing more than two million input counts and get maximum throughputs of 850,000 counts per second. You can now get the same area analysis in a matter of minutes, which allows you to analyse more samples or simply more areas on your sample. Alternatively, you can choose to get a wider view and collect large area mosaic maps to minimise the risk of unintended preferential area selection and get more representative data.
A similar dramatic improvement has occurred in EBSD technology. When I started as EBSD application specialist at EDAX in 2001, my first EBSD demo system could collect at least two points per second when it was not raining and the moon was in the right quarter (or perhaps more realistically, if I was really lucky to have a good sample with strong patterns). The map below was one of my first maps that I collected when getting to know the system and I still use it today as an example to show different typical EBSD mapping features, such as grain boundaries, subgrain boundaries, twins, and slip planes. This map contains “only” 124,405 points but took an 8.5-hour overnight scan to collect.
The same map today would take less than half a minute to collect with a Velocity™ EBSD detector. Or when you would like to take a little wider view you can combine beam and stage movements to collect a 2.5 million point scan of an entire sample in about 15 minutes.
These technological improvements allow you to be more efficient with your time and collect the same data much faster. But alternatively, it can effectively open our eyes and allow us to investigate much larger areas to see the bigger picture. Just be careful when you look at things from a bit further away, sometimes at the end of the day it may seem that these things start looking back at you!