First-principles Study Of Cation Diffusion Behavior In The Dielectric Ceramics/NiZn Ferrite

This paper addresses the requirements of broadband,miniaturization and high performance of electronic devices in the 5G era,and takes the main materials of the circulator,dielectric ceramics and magnetic ferrite composite co-fired substrate,as the main research object.Uses the first principles and machine learning method to systematically study the cation diffusion problem of dielectric ceramics/Ni Zn ferrite composite substrate,understanding its cation diffusion mechanism,providing theoretical and technical support for the development of highperformance microwave ceramics/gyromagnetic materials.Firstly,the self-diffusion process of cations in three crystals was studied by using density functional theory.The diffusion barrier was used to evaluate the difficulty of cation diffusion.Firstly,the self-diffusion process of cations in three matrix crystals was studied by using density functional theory.The diffusion barrier was used to evaluate the difficulty of cation diffusion.The doping model is used to simulate the process of cation entering the opposite matrix.The best doping site that can evaluate the cation is formed by doping and the interdiffusion process is studied.The diffusion coefficient of the cation at 1500 K is estimated by the diffusion barrier.The diffusion couple model is established to simulate the ion diffusion depth after 3 h at 1500 K.The results show that for MT crystal,the self-diffusion barrier is generally lower than the inter-diffusion barrier.The diffusion barrier of Zn ion is close to its self-diffusion barrier about 1e V,and the diffusion depth is about 8nm at 1500 K for 3 h.For CT crystals,the diffusion barriers of the two diffusion processes are much larger than those in other crystals,and the diffusion behavior of ions is difficult to occur.In the NZF crystal,the diffusion coefficient of Ca ions is the lowest,but due to the high self-diffusion coefficient of Ca ions,the number of Ca ions entering the NZF lattice is also small,followed by Mg ions.The diffusion depth of Mg ions is about 100 nm after 3 h at 1500 K.The MT(001)-NZF(111)interface model was established to evaluate the interface stability.The diffusion process of Zn and Mg ions was simulated by this interface model,and the diffusion barrier of the two cations at the interface was calculated.The diffusion barrier of Zn ions is lower and the diffusion is easier.A cation intragranular diffusion model was established using machine learning methods to predict the diffusion activation of cations.500 diffusion-related data were collected from literature and databases as data sets,and 13 atom-related data were used as eigenvalues to evaluate the feature weights.It was found that the melting point Mh and the maximum bond energy Beh in the matrix had a significant weight.Finally,a diffusion model was established using a random forest model to predict the diffusion activation energy of cations.This provides new ideas and methods for us to evaluate ion diffusion.