On the basis of these experimental

On the basis of these experimental results, the chemical transformation in CSP is qualitatively described as follows (Figure 4). First, the high stressed grain-to-grain contacts are formed by the applied uniaxial pressures throughout the compact after the particle rearrangement. To mitigate the high local pressure, BaTiO3 at contact interfaces is dissolved under alkali-based CSP into soluble species, such as [Ti(OH)x4 – x] and Ba2+ (or BaOH+), where the reaction is enhanced more than in non-contact areas. As the concentration of the soluble ions increase, diffusion from contacts to non-contacts (low-stressed pores) is enhanced by the chemical potential difference between those areas. The pressure relaxation by the dissolution and following diffusion results in an increase in the contact area, leading to the Sr-poor grain boundary growth, as clearly shown in Figure 3B. When the concentration of the dissolved ions approaches a critical value (supersaturation) at the low stressed pore surfaces, the ions including Sr2+ precipitate as the Ba1-xSrxTiO3 based on the thermodynamic stability,24 where the precipitation reaction might be influenced by the concentration of dissolved ions, basicity, and temperature.35 In this way, the pressure solution at intergranular phase enhances the kinetics of three successive steps: dissolution, diffusion, and precipitation. If the chemical reaction is controlled by the local thermodynamic equilibria at the interfaces rather than by the global equilibrium in the whole sintering flux, the interface-coupled dissolution-precipitation reactions will be likely to dominate the dynamic chemical reaction during the pressure solution creep in the same way, consistent with the proposed mechanism in the geological field.36,37 Consequently, the pressure-driven re-equilibration of oxides at solid/flux interfaces enables the densification even at low processing temperature, leading to inhomogeneous distribution of unreacted and newly precipitated grains. In conventional thermally-driven sintering processes (T/Tm of > 0.5), the understanding of elemental mechanisms allowed us to design nano and microstructures with desired functions, thanks to precise determination of optimum conditions such as the thermal cycle,38 sintering atmosphere,39 diameter of starting powders,40 ball milling time,41 and co-doping.42 Similarly, the experimental verification of pressure-driven low temperature process (T/Tm < 0.2), as shown in this study, would be beneficial to not only the fundamental understanding of CSP mechanism but also the ongoing CSP research for the fabrication of novel doped and heterogeneous ceramics with the desired function.