New imaging method for material analysis

The method is able to take sharp, high-resolution images without damaging the material by overexposing it. To do this, the scientists developed an algorithm that can recognise patterns in underexposed images. In the scientific journal Nature, they describe the method of Coherent Correlation Imaging (CCI) and present results for samples from thin magnetic layers.

The world in its smallest dimensions is full of movement and in constant flux. Even in solid, virtually unchangeable materials, such fluctuations can result in unusual properties, as in the case of high-temperature superconductors, for example. The fluctuations in phases are particularly strong when a material changes its state, for example from solid to liquid during melting. However, science also investigates changes in the state of a material from non-conductive to conductive, non-magnetic to magnetic or changes in the crystal structure. Many of these processes are used for technical purposes or also play a role in living organisms.

Problem of sample destruction by high-energy radiation
However, it is difficult to observe these processes precisely or even to make a film of the fluctuation patterns. The problem is that the fluctuations can be very fast and take place on the scale of nanometres - a millionth of a millimetre. Even high-resolution X-ray and electron microscopes could not record this fast, random movement. In fact, the problem is even of a fundamental nature, as the example of a photograph makes clear: A minimum amount of lighting is required for every sharp image of an object. If you want to enlarge the object, i.e. zoom in, you have to increase the lighting. Even more light is required if the snapshot is to be taken with a very short exposure time in order to freeze the movement in the image at a certain point in time. Increasingly better spatial and temporal resolution eventually leads to the point where a microscopic object has to be illuminated so strongly that it is altered or even destroyed by the high-energy radiation („illumination“). This is precisely the point that science has reached in recent years: Snapshots taken with free-electron lasers, the most intense X-ray sources available today, inevitably led to the destruction of the sample under investigation.

An algorithm helps to analyse poorly exposed images
An international research team has now found a solution to this dilemma. The starting point was the realisation that the fluctuation patterns in the materials are often not so random. If you only look at a very small area of the sample, you will notice that certain spatial patterns occur again and again. However, it is not possible to predict when and how often which pattern will appear.

The scientists developed a new method for non-destructive imaging, which they call Coherent Correlation Imaging (CCI): To create a video, they continue to take many snapshots of the sample one after the other. In doing so, they reduce the radiation to such an extent that the sample remains intact. However, this means that the fluctuation pattern in the sample can no longer be recognised in a single recording. However, the images still contain enough information to distinguish them from each other and categorise them into groups. To do this, the team first had to develop a new algorithm that analyses the correlations between the images - hence the name of the method. The images in each group are very similar and therefore very likely to originate from a specific fluctuation pattern. A clear picture of the sample only emerges when all the images in a group are analysed together. The scientists can now rewind the film and assign a sharp image of the state of the sample at that point in time to each image.

Example: Fluctuations of magnetic domains filmed
The researchers used this new method to analyse an interesting problem from the world of magnetism. They looked at microscopically small patterns that occur in very thin ferromagnetic layers. These layers form domes (areas with different directions of magnetisation). Similar magnetic layers are used in today's hard drives to encode the data as bits „0“ and „1“ on the hard drive using the different domains. Until now, it was believed that these patterns were very stable. But is this really the case?

To find out, the team investigated just such a thin magnetic layer at one of the most modern X-ray radiation sources, the National Synchrotron Light Source II on Long Island near New York, using the newly developed CCI method. In fact, they found that the patterns do not change at room temperature. However, if the temperature is raised only slightly to 37°C, the areas begin to move back and forth in leaps and bounds and displace each other. The scientists observed this „dance of the domes“ for several hours and then created a kind of „map“ showing the preferred location of the boundaries between the domes. With this map and the film of the movements, it is now possible to better understand the magnetic interactions in the materials and use them for future applications in novel computer architectures.

Outlook
The scientists' next goal is to use the new imaging method at free-electron lasers such as the European XFEL in Hamburg to gain insights into even faster processes on the smallest length scales. They are convinced that their method will help to better understand the role of fluctuations and stochastic processes in the properties of modern materials and thus also discover new ways of utilising them in a targeted manner.

Article from "LABO" dated 20 January 2023