A Coherent Approach to 3D X-ray microscopy

High-resolution 3D X-ray imaging is essential for many fields of science. It lets us study the intricate inner works of everything from integrated circuits to bone and microfossils. In a recently published paper, the authors describe the first ptychographic X-ray computed tomography experiment performed at the NanoMAX beamline. A porous nickel inverse opal structure was successfully imaged in three dimensions with a resolution of 37.3 nm. The Swedish/American team of researchers suggest improvements of which several have meanwhile been implemented at the NanoMAX beamline.

Ptychographic imaging is a versatile method that provides images of a sample without the use of an objective lens. It can supply structural information with high resolution on samples from very different science areas (see the bottom of the page for further reading about the many possibilities).

Ptychography is a scanning method using a pencil beam, as opposed to full-field imaging using a larger beam. The sample is scanned through the X-ray beam and scattering patterns are recorded. A mathematical algorithm is then used to retrieve the image from the diffraction patterns.

For ptychographic imaging to work, it requires the X-ray beam to be highly coherent. X-rays are electromagnetic waves with crests and throughs. If we imagine the X-ray source to be like a stone thrown in a pond, the waves emitted are all in phase for a fully coherent beam. No rings cross each other anywhere. Then a coherent source means that the wave hits the sample the same way more or less in every point and the scattering patterns together can be used to create the full image.

“MAX IV is the first of the 4th generation synchrotron radiation sources which offer a larger coherent fraction of the beam which is essential for ptychographic imaging. At present, pretty much the full coherent flux can be used for ptychographic experiments in the forward direction,” says Maik Kahnt, a postdoc at the NanoMAX beamline.

However, only scanning a sample at one specific angle results in a two-dimensional projection of the sample. To construct a three-dimensional image the sample needs to be rotated slightly, scanned again, rotated a bit more and so on. The sample imaged in the first ptychographic experiment at NanoMAX was rotated in 181 steps. Each projection was scanned with 81 lines, generating 169 diffraction patterns from each line. That’s a huge amount of data and a lot of computing power needed to do the reconstructions.

“Among the biggest challenges of the experiment were the preparation of the micron-sized free-standing samples as well as handling the huge amount of raw data recorded during the experiment,” says Kahnt.

The paper is thought to be a bit of instruction and inspiration for future users who want to do ptychographic imaging at NanoMAX.

“In the article we identify the limiting factors at the time of the experiment – only using part of the coherent flux due to limitations imposed by the detector, having to virtually realign the sample between angular movements, the scanning overhead due to the way the detectors are read and air scattering as the detector was not mounted inside a flight tube under vacuum yet,” explains Kahnt. “Most of these problems have been solved in the meantime. The beamline is now equipped with a new detector that allows us to use the full coherent flux offered by the 3GeV ring of MAX IV, in 90% of the experimental settings. The detector is mounted inside a flight tube under vacuum, which significantly increases the quality of the recorded diffraction patterns. There are also new centring stages allowing easier alignment of the sample to the centre of rotation. The next steps are to figure out what is limiting the achievable resolution now – most likely, the sample stability and the overall beam stability – and improve these even further.”

The team is working on several fronts to make ptychographic imaging even more accessible for users of NanoMAX.

“We will work on making the type of experiments presented in this study a standard option at NanoMAX. It means that in the future, it will be easier to set up and perform the experiments for users. For example, I created various pieces of software to run the data-heavy and computing heavy reconstructions spread over the nodes of the MAX IV cluster, as soon as the data is recorded,” concludes Kahnt.



Maik Kahnt, Simone Sala, Ulf Johansson, Alexander Björling, Zhimin Jiang, Sebastian Kalbfleisch, Filip Lenrick, James H. Pikul and Karina Thånell, First ptychographic X-ray computed tomography experiment on the NanoMAX beamline, J Appl Crystallogr 53 (6), DOI: 10.1107/S160057672001211X


Other examples of ptychographic imaging


About the sample

The sample was an inverse opal made of nickel. An opal – you might have seen them as gemstones – is what is called a photonic crystal. A photonic crystal has an internal periodic structure such that it will only reflect very specific colours of visible light in specific directions. This gives the iridescent effect of butterfly wings and opals. Now, to make inverse opal the researchers carefully stacked tiny plastic beads in a structure similar to that in a natural opal where the tiny spheres are silica. Then they covered the whole structure in nickel, dissolved the plastic beads and were left with the same type of structure but with holes instead of beads. And the result is also a photonic crystal but it will reflect other colours. It is the perfect test sample as the nickel structures formed between the beads are nanometer-sized and easy to find.