Final Report Summary - RLUCIM (Resilient large unit cell inorganic materials)
Inorganic materials are important in technologies that are essential to daily life in advanced societies, such as storage of information in computers, the generation of power and the production of chemicals for both basic (clothing, food) and advanced (aerospace, cars) applications. The world is facing increasing pressure on available resources as population grows and non-renewable elements are depleted. There is thus a need for an efficient way to make new inorganic materials with the properties that society needs – for example, a new material based on readily available elements may be needed to replace an existing one based on rare elements, and will need to have similar properties. This project combines theoretical and experimental work to allow the efficient identification of new materials with complicated compositions and structures, and to evaluate the properties of these new materials. The focus is on oxide materials because they have a wide range of applications.
The project developed a method to predict stable structures for complex functional materials by using building units that are stable themselves (because they have chemically sensible structures) while sufficiently simple to allow almost all possible structures to be built from them. By combining these building units computationally and evaluating the stability of the resulting structures, we are able
to identify the most likely structure for a particular composition of matter of an oxide material. This method has been demonstrated in the discovery of a new material which has a good performance as a cathode (one of the two electrodes) of a solid oxide fuel cell. The method does not aim to predict all possible structures, but to assist the experimental scientist in identifying new functional materials in complicated systems with multiple component elements and large structures. The first new material identified by this approach exemplifies the RLUCIM concept – it has a large unit cell containing six different elements in nine distinct environments in the crystal structure, and this structure is resilient both to the introduction of multiple properties (oxide ion conduction, electronic conduction and catalysis of the oxygen reduction reaction) and to the real operating conditions of a solid oxide fuel cell. We have applied this method to materials with different structures. As well as predicting entirely new structures, it is important to be able to make controlled and predictable changes to known materials in order to modify their properties. We have used calculations to predict the outcome of substantial changes to known materials which we then verified experimentally, thus increasing the lithium ion conductivity of a known material and identifying a new family of candidate solid lithium electrolytes. New solid lithium electrolytes are needed to improve the safety of lithium batteries.
The described work above was performed on bulk materials, but thin films are important for a variety of applications, such as information storage. We have combined computation with crystal chemical knowledge to grow materials with structures and compositions that are not stable in the bulk by the assembly of pre-identified building units. The assembly of heterostructures, in which distinct materials are combined into a thin film, is often assumed to occur without large rearrangements of their individual structures. We identified how atomic rearrangements are required at the interfaces between units to allow the growth of structures that would appear to be highly strained and thus unstable without these rearrangements, thereby enabling the creation of new properties at these interfaces. By combining computational prediction with layer-by-layer growth of unit cells of magnetic but non-piezoelectric materials, we produced a piezoelectric magnet at room temperature – this is a rare example of emergent behaviour in hard condensed matter (it is an important and well-known phenomenon in soft materials), where the assembly displays properties not shown by the constituent units. This approach uses symmetry to produce two properties in the same material that normally do not co-exist, and removed the usual chemical restrictions on relationships between properties and composition.
Computation, thin film growth and bulk synthesis have been used to demonstrate how to predictably access the resilient large unit cell materials targeted over a wide range of length scales. A working level protocol combining these components to discover new complex functional materials has been demonstrated.
The project developed a method to predict stable structures for complex functional materials by using building units that are stable themselves (because they have chemically sensible structures) while sufficiently simple to allow almost all possible structures to be built from them. By combining these building units computationally and evaluating the stability of the resulting structures, we are able
to identify the most likely structure for a particular composition of matter of an oxide material. This method has been demonstrated in the discovery of a new material which has a good performance as a cathode (one of the two electrodes) of a solid oxide fuel cell. The method does not aim to predict all possible structures, but to assist the experimental scientist in identifying new functional materials in complicated systems with multiple component elements and large structures. The first new material identified by this approach exemplifies the RLUCIM concept – it has a large unit cell containing six different elements in nine distinct environments in the crystal structure, and this structure is resilient both to the introduction of multiple properties (oxide ion conduction, electronic conduction and catalysis of the oxygen reduction reaction) and to the real operating conditions of a solid oxide fuel cell. We have applied this method to materials with different structures. As well as predicting entirely new structures, it is important to be able to make controlled and predictable changes to known materials in order to modify their properties. We have used calculations to predict the outcome of substantial changes to known materials which we then verified experimentally, thus increasing the lithium ion conductivity of a known material and identifying a new family of candidate solid lithium electrolytes. New solid lithium electrolytes are needed to improve the safety of lithium batteries.
The described work above was performed on bulk materials, but thin films are important for a variety of applications, such as information storage. We have combined computation with crystal chemical knowledge to grow materials with structures and compositions that are not stable in the bulk by the assembly of pre-identified building units. The assembly of heterostructures, in which distinct materials are combined into a thin film, is often assumed to occur without large rearrangements of their individual structures. We identified how atomic rearrangements are required at the interfaces between units to allow the growth of structures that would appear to be highly strained and thus unstable without these rearrangements, thereby enabling the creation of new properties at these interfaces. By combining computational prediction with layer-by-layer growth of unit cells of magnetic but non-piezoelectric materials, we produced a piezoelectric magnet at room temperature – this is a rare example of emergent behaviour in hard condensed matter (it is an important and well-known phenomenon in soft materials), where the assembly displays properties not shown by the constituent units. This approach uses symmetry to produce two properties in the same material that normally do not co-exist, and removed the usual chemical restrictions on relationships between properties and composition.
Computation, thin film growth and bulk synthesis have been used to demonstrate how to predictably access the resilient large unit cell materials targeted over a wide range of length scales. A working level protocol combining these components to discover new complex functional materials has been demonstrated.