Performing Dynamic Analysis Using Benchtop XRD Analyzers

In this interview, AZoM talks to Chris Shaffer, Business Development Manager, North America, for Thermo Scientific XRF, XRD and OES products, about x-ray diffraction and the importance of sample preparation.

What is x-ray diffraction and how is it used?

X-ray diffraction is a characterization technique used to look at the crystalline structure of a material. With it, we investigate periodic arrangement in a sample at the atomic scale by striking it with an x-ray source and measuring the diffracted intensity.

This is a non-destructive technique, and so whatever you put into instrument comes out the exact same way. Therefore you can test the sample in multiple ways without worrying about the influence of the x-rays.

How does x-ray diffraction work?

Depending on the spacing of the sample’s crystalline structures, we get a pattern which measures planes perpendicular to the vertical axis. Depending on the sample intrinsic properties and selected sample preparation we could face one of three scenarios.

The first is random powder orientation, which is ideal. That means crystallites are distributed in the volume equally in all direction. The second is single crystal, where by definition a single diffraction plane is in one specific orientation. The third is preferred orientation, where it is not completely perfectly aligned, but one face of the crystalline structure is in a predominant alignment to the x-ray source and detector, it is a case half-way between perfect random orientation and perfect single alignment.

We are looking at a periodic arrangement in long order, which is going to give us strong intensities. Again, we have the atomic coordinates, and depending on which way we are striking that sample, we get a specific diffraction plane. This is indicated by the Miller indices. The HKL indices indicate what plane we are diffracting off.

Each plane will give us a different response, and a different angle for the peak on the scan. What we want is long order diffraction. We get this occurring multiple times. The more times this occurs, the bigger the peak we see.

We can look at each of those planes of diffraction and relate that to a peak position. This is almost like a fingerprint, or a thumbprint to indicate what we are looking at. We are looking at the location and proportion of the peaks. That is later compared with a reference database that we use to identify the sample.

You can have what is known as preferred orientation, where one plane of diffraction happens to align better than the other planes, meaning that you get an exaggerated intensity off of that one plane versus the rest. This is non-ideal, but there are mathematical ways to mitigate this during data processing.

How is x-ray diffraction performed?

Normally we can do sample rotation that will help homogenize the analysis. Alternatively, we can do transmission, where we shoot through the sample. You can even use capillary holders, where you are spinning and obtaining all the possible orientations. In these, you are eliminating any kind of packing issue.

With powder samples, we want to be very careful regarding how hard we press the sample into the aluminum cups. If you are pressing very hard, you are going to have a more preferred orientation. Normally, you just use a microscope slide and press very gently, so that it becomes even. That eliminates some of this preferred orientation. Some materials do no exhibit this behavior at all and others where crystals have needle or platelet shapes where it is stronger.

Are there any complications when using x-ray diffraction?

There are inherent problems with certain crystalline structures. In x-ray diffraction, we can also see something called amorphous content. An amorphous phase has no long-term arrangement in its atomic structure. Hence it only shows a broad diffraction peak.

If we get something that does not repeat, we get a huge amount of micro diffraction peaks that are going to make this big hump in the spectra. That is an amorphous halo. Most of the time you are going to get partial crystallinity if you have amorphous content. For example, if you have something like glass.

Thermo Scientific™ ARL™ EQUINOX 100 X-Ray Diffractometer

Glass is melted silica powder. It is completely amorphous, and so you are going to get this huge hump in the spectra. Ideally, we want something that is crystalline, so that it has this long arrangement that we can get good diffraction peaks out of. More commonly we have a portion of amorphous content, with a portion of crystallinity.

It is very important to know about this amorphous content as it has important implications in several fields, for instance Pharma where it correlates with the bioavailability. The more amorphous content you have in pharma, the better bioavailability and ability to absorb into your bloodstream. The more crystalline it is, the less bioavailable it is. In different areas, it is very important to get a certain percentage of crystallinity versus amorphous.

For regular powder diffraction of minerals and inorganic samples, you do not want amorphous. We run a sample and identify what we see in the pattern. We mark the peak positions, or have the software automatically do it, and then we use a database that has a spectral library to identify what we have collected.

Which databases are used to identify peaks in x-ray diffraction spectra?

The most common databases are the ICDD or the ICDS, or a free one called COD. There are many databases available out there. You can take that database, process it and identify your spectrum and what you have. This can be single phase or multi-phase, and using the newer software you can easily quantify how much of each of those phases you have.

That is a way in which we can use x-ray diffraction for quantification and identification. For sample prep in x-ray diffraction, the most common and easiest way is to use a mortar and pestle. If your sample is soft enough, this is a nice way of getting down to a fine particle without destroying the crystallinity of your particle.

The more heightened the processing you have on your particles, the more amorphous content you are going to create because you are destroying the crystalline product. You can use a ball mill for harder materials. This is better compared to something like a ring and puck.

Why is a ball mill better than a ring and puck mill for x-ray diffraction?

A ring and puck mill is designed to destroy the crystalline phase when you are trying to get rid of that component in x-ray fluorescence. It is ideal for x-ray fluorescence, but terrible for x-ray diffraction. The ball mill, which is just based on the impact of shaking and grinding, is ideal for this type of technique.

You can also use what is called a micronizing mill. This can reduce to less than one micron whilst maintaining the crystalline structure. The issue with this mill is that it is expensive and time-consuming to clean, as there are about 50 different little cylindrical spheres, and you have to take it apart and clean each one out between samples. Therefore it is ideal for research but awful for QA/QC.

Which is your preferred method of sample preparation?

With sample prep, we normally just use a microscope slide and put the sample into it. We want to keep the powder even with the outside edge. That gives us our height from the x-ray tube to the sample to the detector. Shifting that height will cause a shift in the peak positions, and so it is very important that we keep that height exact.

Here we have examples of a correctly prepared sample, where you can see it is nice and even, and an overpacked sample. As well as a shift in the peak positioning, we see the background going up and some scatter. This causes a lot of issues because now you are shooting through some of the sample. It is very important that you get good sample prep for x-ray diffraction.

You can do transmission, in which you are basically taking your sample and screwing in two thin polymer windows so that it is x-ray transparent, and instead of doing a reflection off the surface, you are transmitting through the sample. This works very well for organic materials or very small quantities.

You are always going to get better resolution with transmission compared to reflection. The key is that you do not get nearly as much intensity, because you are absorbing some of the x-rays. You can see where you are getting much better resolution between all these peaks, but the intensity is lower and you must count longer.

Another common technique of x-ray diffraction is regular reflection, using zero background, where you are using a specially cut single crystal Silicon piece at a certain angle. You can select the cut in a way that no peak will appear from that Silicon. We can choose different ways you can do very small quantities - micrograms, on Silicon plates, with transmission cups or even capillary.

Can you run a sample without grinding it?

You do not have to grind. However, if you do still have a solid it is going to be primarily a surface analysis, but you can actually put in a scissor jack and lift it up for larger materials. We have had parts of solar cells put in here on a scissor jack, as we did not want to destroy the product.

You may want to run the sample in different orientations so that you are not just seeing one plane of diffraction. You can turn it 90 degrees and run it that way when you are doing a non-prepped sample and cannot rotate or spin it.

How long does it take to perform x-ray diffraction?

How long you measure your sample for depends on how good of a diffractor your sample is. The key is to get a good ratio between your peak and background.

If you have a good diffracting sample, that could be seconds. If you have something that has higher amorphous content or does not diffract so well, then you may have to wait three to four minutes for acquisition.

The other aspect of the time ratio is that you need to know your trace element or your trace phases. If you are only looking for major and minor, then that is going to determine the length of diffraction time. Up to a point, the longer you go the better peak and background you are going to get. Normally, you need about ten thousand counts for a good refinement.

About Chris Shaffer

Chris Shaffer is currently the Business Development Manager, North America, for Thermo Scientific XRF, XRD and OES products. He started working at Thermo Fisher Scientific in 2007 as XRF Applications Specialist before assuming the role of XRF Product Manager in Switzerland for a few years. He also supported our marketing and commercial activities in Latin America and South East Asia. Prior to joining Thermo Fisher Scientific, Chris managed the analytical laboratory at Ferro Corporation as his first job after graduation.