A Workflow for Getting the Most Out of Your Samples and Pushing the Limits of SEM-Based Microanalysis

21^st October 2022 | Author: Dan Haspel

It is interesting how different ways of working seem to dominate amongst different groups of microscope users. In some cases, a certain set of conditions are almost always used, and in others, it is routine to significantly vary how an analysis is performed. There are definitely opportunities for learning from the different approaches that are used so that we can all improve our analyses. One example of this that I was recently involved with - a geological sample on which low magnifications of only a few hundred times or just a couple of thousand times were being used, as the view was that this was all that was needed. Additionally, a uniform accelerating potential of 20 kV was used regardless of the analysis being done, whether it was imaging, energy dispersive x-ray spectrometry (EDS) or electron backscatter diffraction (EBSD), because “that’s what we’ve always done”. In this collection of blogs, I’ll discuss how, with the same system that we use for “standard” analysis, we can push down to the nano-scale. While this will be true for many types of samples, this blog will be showing a geological sample.

First thing’s first…sample preparation

This is probably the most important part of the analysis process, especially for geological samples, as they can be rather tricky to prepare. If this isn’t done correctly, then you will never be able to push the limits of your analysis. In most cases, you will probably end up spending more time preparing the sample than actually analysing it.

In an ideal world, you want to be getting your sample as flat as physically possible. This begins with firstly cutting and mounting your sample, whether you use resin, hot mounting or a glass slide. Once mounted, the sample will go through various stages of polishing. Depending on just how far you want to push things, this could be anything from a simple polish to using increasingly fine grit paper, diamond suspension and finishing off using colloidal silica. However, some minerals don’t seem to like to be manually or vibrationally polished. If this is the case, I would highly recommend broad ion beam polishing (BIB).

All of this will enable your sample to be as flat as possible, which means your micro/nanoanalysis, such as EDS and EBSD, will be able to run at high resolution with improved accuracy. Next, sample needs to be conductive to be analysed in an SEM, and most geological samples being non-conductive. In most such cases, you would want to coat your sample in a thin layer of carbon, with a thickness of ~5-10 nm. This coating is thin enough for EBSD, but also good for zero-tilt EDS. Another option, if your SEM has the capability, could be to use variable pressure rather than coating – however, you won’t be able to get the best resolution when doing so.

Where are you hiding?

Now the sample is mounted, it can be difficult to know where to start. Where in the sample is the mineral that you’re looking for? What features are you trying to find? Most of the time, geological samples can be quite large in SEM terms, and it can be very time-consuming to just navigate around the sample using the SEM manually. So, what better way to navigate a sample than to use AZtec Live Chemical Imaging, which Dr Matt Hiscock wrote about in a previous blog. This is a fantastic part of the software that allows for live chemical imaging, as seen in the video below. The software will generate chemical maps as you navigate around the sample and once you have stopped, the software will begin to generate higher resolution maps. What’s also great, is the Live Trace, which can be seen in the lower right-hand corner. This shows exactly where you have been across your sample. The video shows a user navigating across the fusion crust and then zooming in, enabling finer structures to be identified.

Live Chemical Imaging is extremely useful and quick, and you can easily find points of interest within seconds of starting the acquisition, followed quickly by saving the data from that point. However, it can also be useful to get an overview of the entire sample without manually moving . This allows you to show the entire sample in one image. Alternatively, it allows you to come back to that sample at a later date to re-analyse certain areas in more detail.

For this problem, I would always recommend large area mapping (LAM), which Dr Lucia Spasevski wrote about in previous blog earlier this year. Coupled with EDS, this can be an extremely powerful tool to get an overview of your sample to find what interests you. This can be easily done in our AZtecLive software, where the software can automatically acquire images and maps across an entire sample and montage them all together, to create one single image/map. This single image can be used as a reference image in AZtec for navigation, without the need to touch the SEM controls. It also allows for easy identification of features within the sample, and if you’re a bit of a meteorite fanatic, you will easily notice the fusion crust of this Winchcombe sample. Getting to this point only took ~4 hrs of acquisition, which is much easier and probably quicker than manually navigating around the sample until you find what you’re looking for.

Now we have our single image as a reference to navigate around in AZtec, we can start to interrogate the sample in more detail. However, the LAM may have missed things if the resolution wasn’t good enough.

Found you

Having found your area of interest, whether it’s something special or just something representative, we can now switch to higher magnification analyses. But that’s a story for next time…

I’ll be back to follow up on the second part of this analysis soon – please keep an eye out for my next blog!

About the Author