 So we try to study material imperfections and how they alter the properties and functionalities of materials. The key for this is to understand the atomic structure of defects. So why is this actually important? When you think of technologically irrelevant materials, they are typically all composed of a polycrystalline microstructure. So that means different crystals build up the material. And these crystals are somehow differently oriented to each other. So these different crystals are separated by very, very confined interfaces, so-called grain boundaries. And actually these interfaces, these grain boundaries, which we are interested in. So for example, to give you a perspective in microelectronic devices, these copper interconnects in your smartphone or your computer, these grain boundaries basically in them, they span millions of soccer pitches. And of course they somehow determine the lifetime durability but also safety of these devices. And of course for us it is the key to understand the structure of these interfaces since they dictate how the material performs. For example, early embrittlement, so catastrophic material failure, often occurs at these interfaces. Or for example, the stability of a material, so when the grains coarsen, these interfaces somehow determine the stability of the material but also functional properties. For example, electrical properties are determined by these grain boundaries. For example, in multi-crystalline silicon. And of course for us, the key to basically make use of these interfaces and somehow be able to engineer the material through these grain boundaries, we have to understand their fundamental atomic nature. So the question now is, how do we actually observe the atomic structure of these interfaces? So there's first several problems. You can envision grain boundaries as very confined interfaces. So imagine them as a sheet of paper but actually they are confined to one nanometer. So that means they are 100,000 times thinner than an actual sheet of paper. So there's basically a twofold situation. First we have to develop a method how we can specifically locate such an interface so that we can later on observe it. So basically we take our very thin copper thin film that we have deposited and we use a very small iron beam basically as a sharp razor blade with which we extract specifically a certain part of our sample containing this particular grain boundary. And so the actual method we use then to observe the atomic structure is so-called transmission electron microscope. And in more particularly we use scanning transmission electron microscopy. And what we do here is we basically form a very, very small electron probe. It's only like a tenth of a nanometer, so the size of a few atoms basically. And this beam is formed by high energy electrons. So the electrons almost travel at the speed of light and while we scan our sample, so containing this grain boundary, these electrons interact with the atoms in the sample. And some of these electrons, they get scattered and these scattered electrons, they contain a rich information on our sample. For example, the structure of the sample, how the atoms are bonded or even the chemical composition. To put this in perspective, our microscope, if it would basically increase the resolving power of your eyes, would basically lead to a 700,000 times increase in resolution. So this is how powerful this microscope is. And so now we have this tool with which we can really observe our structure, our material with such a high resolution. And this is of course the key that we understand now how the atoms are arranged with these interfaces so that we really can understand their structure, their properties, and of course ultimately their impact on the material. So of course, when we looked at the atomic structure of these interfaces, what we found is yes, their atomic arrangement was completely disrupted compared to the adjoining crystals, but we found that there is indeed a regular arrangement of the atoms. So basically what you can imagine is that such a grain boundary region is composed as if you would build a lot of Lego blocks. And basically, this is what we call the structural unit. So basically each Lego block would be a structural unit and within this Lego block, the atoms adopt a certain geometric arrangement. What was really puzzling for us that when we looked along this interface in our copper sample that all of a sudden this kind of building blocks, they were abruptly changing. So the structure of the grain boundary was changing abruptly. And this was simply something completely unexpected and we were actually the first ones to be able to observe this with our methodologies. Why is this so interesting? So what we could basically find is that we have two different phases coexisting at this grain boundary. So the grain boundary decomposes into two grain boundary phases. And this is something that was really not expected, especially not expected to occur in an elemental metal such as copper. So basically if you imagine water transformed from the liquid state to the solid state at zero degrees Celsius. So it changes its fate state from the liquid to the solid. So this is basically a similar behavior but now occurring within the interface. This was proposed 50 years ago by Theoretic Concepts. But it has never been observed because it was really too difficult to observe in these very confined spaces. What was also very interesting for us of course when you now think of these two different grain boundary phases is basically sheets of paper and we glue them together. So basically where you have the glue line there's a line defect which separates the two grain boundary phases. And what we actually found is this kind of grain boundary phase junction so this line defect in the interface is sort of dictating how these grain boundary phases transform into each other. So when we treat our material at high temperature, this line defect, this grain boundary phase junction has to move through the sample. But since it's like migration, the motion of this defect is temperature dependent, it somehow stops at an intermediate temperature and freezes in this two phase state so that we can actually observe it in the microscope. So of course one can now ask the question, what do we do with the understanding of the atomic nature of these grain boundary phases? So for me this has like multiple impacts on different levels. So first from the basic science side, of course we as material physicists, we really try to understand the physical nature how it is built up. You want to understand how nature is built up in its most basic form down to the atomic level. And of course this was enabled by our high resolution observations. Beyond that of course, we have to now simply require a paradigm shift in how we treat and see grain boundaries and these interfaces in materials. Because for example when we relate back to embrittlement of a material, so the catastrophic failure along grain boundaries in a material, what is now the impact of these grain boundary phase transformations? Honestly we don't know. How do they now influence the electronic behavior? We still don't know. But it will also give us a tool that we can use in the future to tune the properties of the interface and also with this influence the material properties. For example look back to these copper interconnects in your smartphone. Maybe we can utilize in the future these grain boundary phase transformations. If we just treat them a little differently to enhance the performance and the lifetime of these devices with these different grain boundary phases. Another interesting topic that really fascinated me is that a recent study has shown that these grain boundaries themselves for example exhibit catalytic properties. They are the location where a certain chemical reaction takes place at a higher rate. But also there it's not considered that the interface itself can exist in different states. Can we maybe use grain boundary phase transformations in the phases to further optimize catalytic materials and functional properties in materials? So there's multiple pathways that we are pursuing. On the one hand of course to further advance our understanding of these interfaces but of course the ultimate goal is to really understand them in such a way that we can use them to make a material better. For example real materials contain solutes and impurity elements from the production process or to make the material better. The question is now how do these solutes interact with the different grain boundary phases? This is not known. Can we maybe add a certain element to adjust a certain grain boundary phase and with this a certain property of the interface? This is the things that basically keep us going. Beyond that of course one limitation in the transmission electron microscope is that we only look in projection. So we basically lose the third dimension of our sample and we are basically trying to develop methods so that we can really retrieve every atom in three dimensions within these defects because then we can really build realistic models to make predictions about the interface properties and ultimately our materials. A very intriguing thing about electron microscopy also is that we can basically probe our interfaces under realistic conditions meaning we can heat up the sample as if it would occur in a real life application. So we can study the dynamic behavior and evolution of these grain boundary phases. We can expose them to a strain or we can even put them under gas to see how they dynamically transform. And if we have all these keys at hand then we can really build models to develop new materials to improve the properties of materials so that they are for example less prone to failure, less prone to degradation that they can sustain a better life. It would provide us the ultimate keys to thermodynamically but also kinetically engineer materials.