New directions in high temperature dielectrics: unlocking performance of doped tungsten bronze oxides through mechanistic understanding

点击翻译按钮获取中文摘要

New higher temperature, high-voltage multilayer ceramic capacitors (MLCCs) are required to advance power electronics - an important technology in the energy transition to net zero CO2 emissions by 2050. Wide-bandgap semiconductor technologies for power electronic equipment already provide active components that can operate at 250-300C, allowing reductions in heatsink size and equipment weight. However due to the high switching speeds of wide-bandgap devices, passive and active components must be in close proximity, demanding high temperature operation of the (passive) capacitors. In addition to applications in renewable energy distribution, there are demands for higher temperature capacitors in transport electrification where electronic equipment needs to operate at high ambient temperatures. Unfortunately existing Class II capacitors, which are all based on the perovskite crystal structure, can only operate to 125-175 C. Global research into new higher temperature capacitor materials over the past decade has failed to produce any dielectric material suitable for mass market MLCCs, now manufactured using inexpensive nickel metal internal electrodes. The obstacle has been the presence of bismuth or lead oxide in the ceramic formulation. This would cause the dielectric materials and electrodes to degrade in the high temperature, chemically reducing atmospheres used to manufacture modern MLCCs. In a shift of research direction, we have recently obtained proof-of-principle that a new type of dielectric based on the tungsten bronze crystal structure offers uniformly high permittivity (>1300 +/- 15%) over the requisite -55 to 300 C temperature range. The material is based on strontium sodium niobate (SNN) co-doped with only 1-2.5 at.% calcium, yttrium and zirconium. Although promising, the dielectric properties fall short of the exceptional performance levels required of a next generation capacitor material. For example, dielectric losses (currently 4%) exceed industrial specifications (2.5%). Unlocking the true potential of the new tungsten bronze approach is severely hindered by a lack of knowledge as to underpinning mechanisms. For example, why low levels of dopants create extremely diffuse twin temperature-dependent dielectric anomalies. In preliminary work we have demonstrated that composition-structure-property relationships for existing temperature stable dielectrics based on titanate perovskites do not apply to this new type of high-temperature dielectric. We propose to unlock the true potential of tungsten bronzes by application of new scientific understanding to overcome existing limitations. We will discover how to raise permittivity and reduce dielectric losses in doped SNN ceramics across the challenging temperature range -55 to 300 C by studying how structure (crystal, nano, micro, defect) is modified by specific dopant systems, using a combination of: electron, neutron and synchrotron diffraction; atom column resolution electron microscopy; electrochemical impedance spectroscopy. First principles simulations will also assist us in interpreting experimental findings and developing structure-property models. From this framework of understanding, new compositions will be designed. Final materials selection criteria will include a range of other dielectric parameters, including dielectric breakdown strength and energy storage density. Our capacitor industry partner KEMET will help evaluate materials and conduct highly accelerated lifetime testing. The best material will be demonstrated within the project in a wide-bandgap switching cell with integrated high-voltage DC-link MLCCs. Alongside direct engagement with established company collaborations, wider benefits will be maximized by developing new activities with industry. This will be achieved in part using the resources of the University of Leeds's Research and Innovation Service, and the new Innovation and Enterprise Centre, Nexus.