ABSTRACT
Ordering of ferroelectric polarization1 and its trajectory in response to an electric field2 are essential for the operation of non-volatile memories3, transducers4 and electro-optic devices5. However, for voltage control of capacitance and frequency agility in telecommunication devices, domain walls have long been thought to be a hindrance because they lead to high dielectric loss and hysteresis in the device response to an applied electric field6. To avoid these effects, tunable dielectrics are often operated under piezoelectric resonance conditions, relying on operation well above the ferroelectric Curie temperature7, where tunability is compromised. Therefore, there is an unavoidable trade-off between the requirements of high tunability and low loss in tunable dielectric devices, which leads to severe limitations on their figure of merit. Here we show that domain structure can in fact be exploited to obtain ultralow loss and exceptional frequency selectivity without piezoelectric resonance. We use intrinsically tunable materials with properties that are defined not only by their chemical composition, but also by the proximity and accessibility of thermodynamically predicted strain-induced, ferroelectric domain-wall variants8. The resulting gigahertz microwave tunability and dielectric loss are better than those of the best film devices by one to two orders of magnitude and comparable to those of bulk single crystals. The measured quality factors exceed the theoretically predicted zero-field intrinsic limit owing to domain-wall fluctuations, rather than field-induced piezoelectric oscillations, which are usually associated with resonance. Resonant frequency tuning across the entire L, S and C microwave bands (1-8 gigahertz) is achieved in an individual device-a range about 100 times larger than that of the best intrinsically tunable material. These results point to a rich phase space of possible nanometre-scale domain structures that can be used to surmount current limitations, and demonstrate a promising strategy for obtaining ultrahigh frequency agility and low-loss microwave devices.
ABSTRACT
Calculation of Raman scattering from molecular dynamics (MD) simulations requires accurate modeling of the evolution of the electronic polarizability of the system along its MD trajectory. For large systems, this necessitates the use of atomistic models to represent the dependence of electronic polarizability on atomic coordinates. The bond polarizability model (BPM) is the simplest such model and has been used for modeling the Raman spectra of molecular systems but has not been applied to solid-state systems. Here, we systematically investigate the accuracy and limitations of the BPM parameterized from the density functional theory results for a series of simple molecules, such as CO2, SO2, H2S, H2O, NH3, and CH4; the more complex CH2O, CH3OH, CH3CH2OH, and thiophene molecules; and the BaTiO3 and CsPbBr3 perovskite solids. We find that BPM can reliably reproduce the overall features of the Raman spectra, such as shifts of peak positions. However, with the exception of highly symmetric systems, the assumption of non-interacting bonds limits the quantitative accuracy of the BPM; this assumption also leads to qualitatively inaccurate polarizability evolution and Raman spectra for systems where large deviations from the ground state structure are present.
ABSTRACT
Progress in the application of machine learning (ML) methods to materials design is hindered by the lack of understanding of the reliability of ML predictions, in particular, for the application of ML to small data sets often found in materials science. Using ML prediction for transparent conductor oxide formation energy and band gap, dilute solute diffusion, and perovskite formation energy, band gap, and lattice parameter as examples, we demonstrate that (1) construction of a convex hull in feature space that encloses accurately predicted systems can be used to identify regions in feature space for which ML predictions are highly reliable; (2) analysis of the systems enclosed by the convex hull can be used to extract physical understanding; and (3) materials that satisfy all well-known chemical and physical principles that make a material physically reasonable are likely to be similar and show strong relationships between the properties of interest and the standard features used in ML. We also show that similar to the composition-structure-property relationships, inclusion in the ML training data set of materials from classes with different chemical properties will not be beneficial for the accuracy of ML prediction and that reliable results likely will be obtained by ML model for narrow classes of similar materials even in the case where the ML model will show large errors on the data set consisting of several classes of materials.
Subject(s)
Machine Learning , Reproducibility of ResultsABSTRACT
Despite more than 50 years of investigation, it is still unclear how the underlying structure of relaxor ferroelectrics gives rise to their defining properties, such as ultrahigh piezoelectric coefficients, high permittivity over a broad temperature range, diffuse phase transitions, strong frequency dependence in dielectric response, and phonon anomalies. The model of polar nanoregions inside a non-polar matrix has been widely used to describe the structure of relaxor ferroelectrics. However, the lack of precise knowledge about the shapes, growth and dipole patterns of polar nanoregions has led to the characterization of relaxors as "hopeless messes", and no predictive model for relaxor behaviour is currently available. Here we use molecular dynamics simulations of the prototypical Pb(Mg1/3,Nb2/3)O3-PbTiO3 relaxor material to examine its structure and the spatial and temporal polarization correlations. Our simulations show that the unusual properties of relaxors stem from the presence of a multi-domain state with extremely small domain sizes (2-10 nanometres), and no non-polar matrix, owing to the local dynamics. We find that polar structures in the multi-domain state in relaxors are analogous to those of the slush state of water. The multi-domain structure of relaxors that is revealed by our molecular dynamics simulations is consistent with recent experimental diffuse scattering results and indicates that relaxors have a high density of low-angle domain walls. This insight explains the recently discovered classes of relaxors that cannot be described by the polar nanoregion model, and provides guidance for the design and synthesis of new relaxor materials.
ABSTRACT
Carbons are ubiquitous electrocatalytic supports for various energy-related transformations, especially in fuel cells. Doped carbons such as Fe-N-C materials are particularly active towards the oxidation of hydrazine, an alternative fuel and hydrogen carrier. However, there is little discussion of the electrocatalytic role of the most abundant component - the carbon matrix - towards the hydrazine oxidation reaction (HzOR). We present a systematic investigation of undoped graphitic carbons towards the HzOR in alkaline electrolyte. Using highly oriented pyrolytic graphite electrodes, as well as graphite powders enriched in either basal planes or edge defects, we demonstrate that edge defects are the most active catalytic sites during hydrazine oxidation electrocatalysis. Theoretical DFT calculations support and explain the mechanism of HzOR on carbon edges, identifying unsaturated graphene armchair defects as the most likely active sites. Finally, these findings explain the 'double peak' voltammetric feature observed on many doped carbons during the HzOR.
ABSTRACT
The existence of domain walls, which separate regions of different polarization, can influence the dielectric, piezoelectric, pyroelectric and electronic properties of ferroelectric materials. In particular, domain-wall motion is crucial for polarization switching, which is characterized by the hysteresis loop that is a signature feature of ferroelectric materials. Experimentally, the observed dynamics of polarization switching and domain-wall motion are usually explained as the behaviour of an elastic interface pinned by a random potential that is generated by defects, which appear to be strongly sample-dependent and affected by various elastic, microstructural and other extrinsic effects. Theoretically, connecting the zero-kelvin, first-principles-based, microscopic quantities of a sample with finite-temperature, macroscopic properties such as the coercive field is critical for material design and device performance; and the lack of such a connection has prevented the use of techniques based on ab initio calculations for high-throughput computational materials discovery. Here we use molecular dynamics simulations of 90° domain walls (separating domains with orthogonal polarization directions) in the ferroelectric material PbTiO3 to provide microscopic insights that enable the construction of a simple, universal, nucleation-and-growth-based analytical model that quantifies the dynamics of many types of domain walls in various ferroelectrics. We then predict the temperature and frequency dependence of hysteresis loops and coercive fields at finite temperatures from first principles. We find that, even in the absence of defects, the intrinsic temperature and field dependence of the domain-wall velocity can be described with a nonlinear creep-like region and a depinning-like region. Our model enables quantitative estimation of coercive fields, which agree well with experimental results for ceramics and thin films. This agreement between model and experiment suggests that, despite the complexity of ferroelectric materials, typical ferroelectric switching is largely governed by a simple, universal mechanism of intrinsic domain-wall motion, providing an efficient framework for predicting and optimizing the properties of ferroelectric materials.
ABSTRACT
High-temperature order-disorder phase transitions play an important role in determining the structure and physical and chemical properties of non-stoichiometric transition metal carbides. Due to the large number of possible carbon vacancy arrangements, it is difficult to study these systems with first-principles calculations. Here, we construct a simple atomistic potential capable of accurately reproducing the energetics of the carbon vacancy arrangements in cubic Mo2C and Ti2C obtained from density functional theory calculations. We show that this potential can be applied to correctly predict the transition temperatures between the ordered and disordered states in Monte Carlo simulations on large supercells and reveal the extend of local order in the disordered phases of Mo2C and Ti2C that show interesting physical and chemical properties. We find that even the high-temperature disordered phase exhibit a relatively high degree of local order as indicated by the relatively small change in the root mean square number of C atom neighbours of Mo/Ti compared to the ordered phase (from 3.0 to 3.1-3.2). This atomistic potential enables the study of how the structure of these carbides can be tuned through the synthesis temperature to control the properties of carbide materials that are related to the degree of disorder in the system such as catalytic activity and electrical conductivity and play an important role in applications of these carbides. Fundamentally, the successful modelling of these carbides suggests that despite the presence of metallic, covalent and ionic interactions, bonding in carbides can be modelled by simple and physically intuitive interatomic potentials.
ABSTRACT
Fe-N-C electrocatalysts hold a great promise for Pt-free energy conversion, driving the electrocatalysis of oxygen reduction and evolution, oxidation of nitrogen fuels, and reduction of N2, CO2, and NOx. Nevertheless, the catalytic role of iron carbide, a component of nearly every pyrolytic Fe-N-C material, is at the focus of a heated controversy. We now resolve the debate by examining a broad range of Fe3C sites, spanning across many typical size distributions and carbon environments. Removing Fe3C selectively by a non-oxidizing acid reveals its inactivity towards two representative reactions in alkaline media, oxygen reduction and hydrazine oxidation. The activity is assigned to other pre-existing sites, most probably Fe-Nx. DFT calculations prove that the Fe3C surface binds O and N intermediates too strongly to be catalytic. By settling the argument on the catalytic role of Fe3C in alkaline electrocatalysis, we hope to spur innovation in this critical field.
ABSTRACT
Metallocorroles and metalloporphyrins (M-N-C) are some of the best alternative molecular catalysts for the replacement of the expensive platinum-group metals (PGM) in oxygen reduction reaction (ORR) catalysis in polymer electrolyte membrane (PEM) fuel cells. To date, Co-based corroles have shown the best performance, but still suffer from the poor stability and the toxicity of the Co metal. Mn-N-C are more stable than Co-N-C, and are also less reactive towards peroxide formation. In this work, using first-principles density functional theory calculations, we study the improvement of the Mn-based corrole ORR activity by exploiting hydrogen bonding with substituent groups to modify the adsorption energies of the ORR intermediates and obtain higher onset potential (Vonset) values. We found that by using phenyl acetic acid as a substituent, Vonset increased from 0.54 V for the unsubstituted corrole to â¼0.9 V which is competitive with the Vonset of the Co-based corroles. Our results suggest that hydrogen bonding with substituent groups should be considered in the analysis and design of the reactivity of active sites in non-PGM ORR catalysts.
ABSTRACT
Ferroelectrics have recently attracted attention as a candidate class of materials for use in photovoltaic devices, and for the coupling of light absorption with other functional properties. In these materials, the strong inversion symmetry breaking that is due to spontaneous electric polarization promotes the desirable separation of photo-excited carriers and allows voltages higher than the bandgap, which may enable efficiencies beyond the maximum possible in a conventional p-n junction solar cell. Ferroelectric oxides are also stable in a wide range of mechanical, chemical and thermal conditions and can be fabricated using low-cost methods such as sol-gel thin-film deposition and sputtering. Recent work has shown how a decrease in ferroelectric layer thickness and judicious engineering of domain structures and ferroelectric-electrode interfaces can greatly increase the current harvested from ferroelectric absorber materials, increasing the power conversion efficiency from about 10(-4) to about 0.5 per cent. Further improvements in photovoltaic efficiency have been inhibited by the wide bandgaps (2.7-4 electronvolts) of ferroelectric oxides, which allow the use of only 8-20 per cent of the solar spectrum. Here we describe a family of single-phase solid oxide solutions made from low-cost and non-toxic elements using conventional solid-state methods: [KNbO3]1 - x[BaNi1/2Nb1/2O3 - δ]x (KBNNO). These oxides exhibit both ferroelectricity and a wide variation of direct bandgaps in the range 1.1-3.8 electronvolts. In particular, the x = 0.1 composition is polar at room temperature, has a direct bandgap of 1.39 electronvolts and has a photocurrent density approximately 50 times larger than that of the classic ferroelectric (Pb,La)(Zr,Ti)O3 material. The ability of KBNNO to absorb three to six times more solar energy than the current ferroelectric materials suggests a route to viable ferroelectric semiconductor-based cells for solar energy conversion and other applications.
ABSTRACT
Switchable polarization makes ferroelectrics a critical component in memories, actuators and electro-optic devices, and potential candidates for nanoelectronics. Although many studies of ferroelectric switching have been undertaken, much remains to be understood about switching in complex domain structures and in devices. In this work, a combination of thin-film epitaxy, macro- and nanoscale property and switching characterization, and molecular dynamics simulations are used to elucidate the nature of switching in PbZr(0.2)Ti(0.8)O3 thin films. Differences are demonstrated between (001)-/(101)- and (111)-oriented films, with the latter exhibiting complex, nanotwinned ferroelectric domain structures with high densities of 90° domain walls and considerably broadened switching characteristics. Molecular dynamics simulations predict both 180° (for (001)-/(101)-oriented films) and 90° multi-step switching (for (111)-oriented films) and these processes are subsequently observed in stroboscopic piezoresponse force microscopy. These results have implications for our understanding of ferroelectric switching and offer opportunities to change domain reversal speed.
Subject(s)
Calcium Compounds , Oxides , Drug Stability , Organic Chemicals , Spectrum Analysis, Raman , TitaniumABSTRACT
We use first-principles density functional theory within the local density approximation to ascertain the ground state structure of real and theoretical compounds with the formula ABS3 (A = K, Rb, Cs, Ca, Sr, Ba, Tl, Sn, Pb, and Bi; and B = Sc, Y, Ti, Zr, V, and Nb) under the constraint that B must have a d(0) electronic configuration. Our findings indicate that none of these AB combinations prefer a perovskite ground state with corner-sharing BS6 octahedra, but that they prefer phases with either edge- or face-sharing motifs. Further, a simple two-dimensional structure field map created from A and B ionic radii provides a neat demarcation between combinations preferring face-sharing versus edge-sharing phases for most of these combinations. We then show that by modifying the common Goldschmidt tolerance factor with a multiplicative term based on the electronegativity difference between A and S, the demarcation between predicted edge-sharing and face-sharing ground state phases is enhanced. We also demonstrate that, by calculating the free energy contribution of phonons, some of these compounds may assume multiple phases as synthesis temperatures are altered, or as ambient temperatures rise or fall.
ABSTRACT
The development of durable platinum-group-metal-free oxygen reduction reaction (ORR) catalysts is a key research direction for enabling the wide use of fuel cells. Here, we use a combination of experimental measurements and density functional theory calculations to study the activity and durability of seven iron-based metallophthalocyanine (MPc) ORR catalysts that differ only in the identity of the substituent groups on the MPcs. While the MPcs show similar ORR activity, their durabilities as measured by the current decay half-life differ greatly. We find that the energy difference between the hydrogenated intermediate structure and the final demetalated structure (ΔEdemetalation) of the MPcs is linearly related to the degradation reaction barrier energy. Comparison to the degradation data for the previously studied metallocorrole systems suggested that ΔEdemetalation also serves as a descriptor for the corrole systems and that the high availability of protons at the active site due to the COOH group of the o-corrole decreases the durability.
ABSTRACT
Relaxor ferroelectrics have been a focus of intense attention due to their anomalous properties, and understanding the structure and dynamics of relaxors has been one of the long-standing challenges in solid-state physics. We investigate the local structure and dynamics in 75%PbMg(1/3)Nb(2/3)O(3)-25%PbTiO(3) using molecular dynamics simulations and the dynamic pair distribution function technique. We show that relaxor transitions can be described by local order parameters. The relaxor phase is characterized by the presence of highly anisotropic correlations between the local cation displacements that resemble the hydrogen bond network in water. This contradicts the current model of polar nanoregion inside a nonpolar matrix. We therefore suggest a new model of a homogeneous random network of anisotropically coupled dipoles.
ABSTRACT
The motion of domain walls is critical to many applications involving ferroelectric materials, such as fast high-density non-volatile random access memory. In memories of this sort, storing a data bit means increasing the size of one polar region at the expense of another, and hence the movement of a domain wall separating these regions. Experimental measurements of domain growth rates in the well-established ferroelectrics PbTiO3 and BaTiO3 have been performed, but the development of new materials has been hampered by a lack of microscopic understanding of how domain walls move. Despite some success in interpreting domain-wall motion in terms of classical nucleation and growth models, these models were formulated without insight from first-principles-based calculations, and they portray a picture of a large, triangular nucleus that leads to unrealistically large depolarization and nucleation energies. Here we use atomistic molecular dynamics and coarse-grained Monte Carlo simulations to analyse these processes, and demonstrate that the prevailing models are incorrect. Our multi-scale simulations reproduce experimental domain growth rates in PbTiO3 and reveal small, square critical nuclei with a diffuse interface. A simple analytic model is also proposed, relating bulk polarization and gradient energies to wall nucleation and growth, and thus rationalizing all experimental rate measurements in PbTiO3 and BaTiO3.
ABSTRACT
We show that light drives large-amplitude structural changes in thin films of the prototypical ferroelectric PbTiO3 via direct coupling to its intrinsic photovoltaic response. Using time-resolved x-ray scattering to visualize atomic displacements on femtosecond time scales, photoinduced changes in the unit-cell tetragonality are observed. These are driven by the motion of photogenerated free charges within the ferroelectric and can be simply explained by a model including both shift and screening currents, associated with the displacement of electrons first antiparallel to and then parallel to the ferroelectric polarization direction.
ABSTRACT
Recent efforts to develop durable high-performance platinum-group metal (PGM)-free oxygen reduction reaction (ORR) electrocatalysts have focused on Fe- and Co-based molecular and pyrolyzed catalysts. While Mn-based catalysts have advantages of lower toxicity and higher durability, their activity has been generally poor. Nevertheless, several examples of high-performance Mn-based catalysts have been reported. Thus, it is necessary to understand why Mn-based materials much more rarely show high catalytic ORR performance and to determine the factors that can lead to the achievement of such high performance in these rare cases. We have studied the effects of the changes in the macrocycle structure, axial ligand, distance between the active sites, interactions with the dopant N atoms and the presence of an extended carbon network on the ORR catalysis of various Mn-, Fe-, and Co-based systems through the comparison of the adsorption energies of the ORR intermediates. We find that the sensitivity to the local environment changes is the largest for Mn and is the smallest for Co, with Fe between Mn and Co. Our results showed that the strong binding of OH by Mn and the strong sensitivity of the Mn to the modification of its environment necessitate a precise combination of local environment changes to achieve a high onset potential (Vonset ) in Mn-based catalysts. By contrast, the weaker binding of OH by Fe and Co and their weaker sensitivity to local environment changes lead to a wide variety of local environments with favorable catalytic activity (Vonset >0.7â V) for Co- and Fe-based systems. This explains the scarcity of reported Mn-based pyrolyzed catalysts and suggests that precise material synthesis and engineering of the active site can achieve high-performance Mn-based ORR electrocatalysts with high activity and durability.
Subject(s)
Carbon , Platinum , Carbon/chemistry , Catalysis , Ligands , Oxygen/chemistryABSTRACT
Prediction of properties from composition is a fundamental goal of materials science that is particularly relevant for ferroelectric perovskite oxide solid solutions where compositional variation is a primary tool for material design. Design of ferroelectric oxide solid solutions has been guided by heuristics and first-principles and Landau-Ginzburg-Devonshire theoretical methods that become increasingly difficult to apply in ternary, quaternary, and quintary solid solutions. To address this problem, a multilevel model is developed for the prediction of the ferroelectric-to-paraelectric transition temperature (Tc ), coercive field (Ec ), and polarization (P) of PbTiO3 -derived ferroelectric solid solutions from composition. The characteristics of the materials at different length scales, starting at the level of the electronic structure and chemical bonding of the constituent ions and ending at the level of collective behavior, are analytically related by using ferroelectric domain walls and cationic off-center displacements as the key links between the different levels of the model. The obtained composition-structure-property relationships provide a unified quantitatively predictive theory for understanding PbTiO3 -derived solid solutions. Such a multilevel analytical modeling approach is likely to be generally applicable to different classes of ferroelectric perovskite oxides and to other functional properties, and to materials and properties beyond the field of ferroelectrics.