Periodic Reporting for period 3 - MatMech (Live Tapings of Material Formation: Unravelling formation mechanisms in materials chemistry through Multimodal X-ray total scattering studies)
Berichtszeitraum: 2022-02-01 bis 2023-07-31
The development of new functional materials for e.g. batteries or catalysis relies on our understanding of the relation between material structure, properties and synthesis. While the intense focus on ‘materials by design’ have made it possible to predict the properties of many materials given an atomic arrangement, knowing how to synthesize it is a completely different story. Material synthesis methods are to a large degree established by extensive parameter studies based on trial-and-error experiments, which slows down the development of novel materials. Specifically, our knowledge of particle nucleation is lacking, as even non-classical views on nucleation such as the concept of pre-nucleation clusters do not apply an atomistic view of the nucleation process. In the project MatMech, we apply new methods in X-ray total scattering and Pair Distribution Function analysis to follow nucleation processes to establish the framework needed for predictive material synthesis. One of the large challenges in studying nucleation is the limited characterization methods that can give structural information on materials without long-range order. In the project, we use time-resolved X-ray total scattering, which gives new possibilities for following structural changes in a synthesis, all the way from a solution over amorphous intermediates to crystalline materials. However, the analysis of the large amount of X-ray scattering data is an enormous bottleneck in such studies, and the data may not always provide enough information to result in a unique structure solution. We are therefore at the same time developing a new multimodal approach for scattering data analysis including methods from data science and Machine Learningc
In this first part of the project, we have made achievements in in situ synchrotron studies of material formation, as well as in data analysis of X-ray total scattering. We have developed setups for in situ X-ray total scattering and Small Angle X-ray scattering experiments, and we have used these for new studies of metal oxide formation. Specifically, for molybdate and tungsten oxides, we have revealed the influence of precursor polyoxometalate structures on the final material during synthesis, which has made it possible to predict the outcome of a synthesis based on the synthesis parameters and starting material.
We have also developed new methods for Pair Distribution Function analysis, taking advantage of structure mining and Machine Learning methods. With the continued development of these methods, it will be possible to overcome a huge bottleneck in structural analysis, namely identifying a model for further data analysis. We expect this to have an impact not only in Pair Distribution Function analysis, but in scattering data analysis in general.
The project progress has been heavily affected by the COVID-19 pandemic, as most of our scheduled synchrotron beamtimes have been cancelled or postponed, which delays progress in all work packages. We expect to be able to do the planned experiments when travel restrictions ease but currently, the content of the work packages have been adjusted to be feasible under the current restraints. We have therefore focused more on development of data analysis methods than originally planned.
We have also developed new methods for Pair Distribution Function analysis, taking advantage of structure mining and Machine Learning methods. With the continued development of these methods, it will be possible to overcome a huge bottleneck in structural analysis, namely identifying a model for further data analysis. We expect this to have an impact not only in Pair Distribution Function analysis, but in scattering data analysis in general.
The project progress has been heavily affected by the COVID-19 pandemic, as most of our scheduled synchrotron beamtimes have been cancelled or postponed, which delays progress in all work packages. We expect to be able to do the planned experiments when travel restrictions ease but currently, the content of the work packages have been adjusted to be feasible under the current restraints. We have therefore focused more on development of data analysis methods than originally planned.
As described above, we have obtained several results which give completely new insight into material formation. For example, we have established that the ionic cluster present at the moment of material nucleation directly influences the material that forms in a synthesis.
In the remaining part of the project, where synchrotron travel and in situ experiments will again be possible, we expect to do extensive studies of the formation of especially mixed metal oxides. While we have established that the ionic clusters present in solution affect the formation of simple binary oxides, we now want to move on to establish how doping or incorporation of a second and third metal in the structure takes place. With more complex experiments again becoming possible, we will furthermore be able to combine our total scattering experiments with more spectroscopy experiments.
We will also continue our work using data science methods for analysis of scattering data. In a collaboration with researchers from Department of Chemistry University of Copenhagen, we have recently started using e.g. interpretive Machine Learning in the analysis of X-ray scattering data.
In the remaining part of the project, where synchrotron travel and in situ experiments will again be possible, we expect to do extensive studies of the formation of especially mixed metal oxides. While we have established that the ionic clusters present in solution affect the formation of simple binary oxides, we now want to move on to establish how doping or incorporation of a second and third metal in the structure takes place. With more complex experiments again becoming possible, we will furthermore be able to combine our total scattering experiments with more spectroscopy experiments.
We will also continue our work using data science methods for analysis of scattering data. In a collaboration with researchers from Department of Chemistry University of Copenhagen, we have recently started using e.g. interpretive Machine Learning in the analysis of X-ray scattering data.