Machine Learning Screening of Metal-Ion Battery Electrode Materials.

Rechargeable batteries provide crucial energy storage systems for renewable energy sources, as well as consumer electronics and electrical vehicles. There are a number of important parameters that determine the suitability of electrode materials for battery applications, such as the average voltage and the maximum specific capacity which contribute to the overall energy density. Another important performance criterion for battery electrode materials is their volume change upon charging and discharging, which contributes to determine the cyclability, Coulombic efficiency, and safety of a battery. In this work, we present deep neural network regression machine learning models (ML), trained on data obtained from the Materials Project database, for predicting average voltages and volume change upon charging and discharging of electrode materials for metal-ion batteries. Our models exhibit good performance as measured by the average mean absolute error obtained from a 10-fold cross-validation, as well as on independent test sets. We further assess the robustness of our ML models by investigating their screening potential beyond the training database. We produce Na-ion electrodes by systematically replacing Li-ions in the original database by Na-ions and, then, selecting a set of 22 electrodes that exhibit a good performance in energy density, as well as small volume variations upon charging and discharging, as predicted by the machine learning model. The ML predictions for these materials are then compared to quantum-mechanics based calculations. Our results reaffirm the significant role of machine learning techniques in the exploration of materials for battery applications.

Structural and electronic properties of layered nanoporous organic nanocrystals.

In this work, we present the optimized geometric stacking of several layered nanoporous organic nanocrystals (NONs) and the stacking effect on their electronic structure. Hexagonal layered structures, C12H6-h2D, B6N6H6-h2D and C6N6-h2D are built from aromatic organic molecular units benzene, borazine and 1,3,5-triazine, respectively while oblique structures, C10N2H4-o2D, C8N4H2-o2D, C10P2H4-o2D and C10As2H4-o2D, are built from pyridine, 1,3-diazine, phosphinine and arsinine, respectively. Our density functional theory calculations show stacking energy profiles of NONs that are similar to graphene in both the stand-alone and bulk C12H6-h2D and B6N6H6-h2D structures while the rest of the studied layered materials deviate from the perfect AB stacking. The number of layers as well as the stacking configuration significantly influence the electronic properties of these materials. Indirect to direct band gap crossovers from the bulk to monolayers are observed in all of the NONs except in C6N6-h2D which exhibits a direct band gap in both the monolayer, isolated few-layers, and bulk. Furthermore, it is observed that the electronic nature of C10As2H4-o2D changes from a semiconducting character in the isolated monolayer to a metallic character in the bulk. The porous nature and the stability of these layered NONs combined with the electronic properties observed in this work point at them as valuable materials for potential applications in nanoelectronics and gas separation membranes, as well as deep ultraviolet optoelectronics and laser devices.

Computational Study of Ortho-Substituent Effects on Antioxidant Activities of Phenolic Dendritic Antioxidants.

Antioxidants are an important component of our ability to combat free radicals-an excess of which leads to oxidative stress, which is related to aging and numerous human diseases. Oxidative damage also shortens the shelf-life of foods and other commodities. Understanding the structure-activity relationship of antioxidants and their mechanisms of action is important for designing more potent antioxidants for potential use as therapeutic agents as well as preservatives. We report the first computational study on the electronic effects of ortho-substituents in dendritic tri-phenolic antioxidants, comprising a common phenol moiety and two other phenol units with electron-donating or electron-withdrawing substituents. Among the three proposed antioxidant mechanisms, sequential proton loss electron transfer (SPLET) was found to be the preferred mechanism in methanol for the dendritic antioxidants based on calculations using Gaussian 16. We then computed the total enthalpy values by cumulatively running SPLET for all three rings to estimate electronic effects of substituents on overall antioxidant activity of each dendritic antioxidant and establish their structure-activity relationships. Our results show that the electron-donating o-OCH3 group has a beneficial effect while the electron-withdrawing o-NO2 group has a negative effect on the antioxidant activity of the dendritic antioxidant. The o-Br and o-Cl groups did not show any appreciable effects. These results indicate that electron-donating groups such as o-methoxy are useful for designing potent dendritic antioxidants while the nitro and halogens do not add value to the radical scavenging antioxidant activity. We also found that the half-maximal inhibitory concentration (IC50) values of 2,2-diphenyl-1-picrylhydrazyl (DPPH) better correlate with the second step (electron transfer enthalpy, ETE) than the first step (proton affinity, PA) of the SPLET mechanism, implying that ETE is the better measure for estimating overall radical scavenging antioxidant activities.

Machine Learning the Voltage of Electrode Materials in Metal-Ion Batteries.

Machine-learning (ML) techniques have rapidly found applications in many domains of materials chemistry and physics where large data sets are available. Aiming to accelerate the discovery of materials for battery applications, in this work, we develop a tool ( http://se.cmich.edu/batteries ) based on ML models to predict voltages of electrode materials for metal-ion batteries. To this end, we use deep neural network, support vector machine, and kernel ridge regression as ML algorithms in combination with data taken from the Materials Project database, as well as feature vectors from properties of chemical compounds and elemental properties of their constituents. We show that our ML models have predictive capabilities for different reference test sets and, as an example, we utilize them to generate a voltage profile diagram and compare it to density functional theory calculations. In addition, using our models, we propose nearly 5000 candidate electrode materials for Na- and K-ion batteries. We also make available a web-accessible tool that, within a minute, can be used to estimate the voltage of any bulk electrode material for a number of metal ions. These results show that ML is a promising alternative for computationally demanding calculations as a first screening tool of novel materials for battery applications.

Tunable Broadband Nanocarbon Transparent Conductor by Electrochemical Intercalation.

Optical transparent and electrical conducting materials with broadband transmission are important for many applications in optoelectronic, telecommunications, and military devices. However, studies of broadband transparent conductors and their controlled modulation are scarce. In this study, we report that reversible transmittance modulation has been achieved with sandwiched nanocarbon thin films (containing carbon nanotubes (CNTs) and reduced graphene oxide (rGO)) via electrochemical alkali-ion intercalation/deintercalation. The transmittance modulation covers a broad range from the visible (450 nm) to the infrared (5 µm), which can be achieved only by rGO rather than pristine graphene films. The large broadband transmittance modulation is understood with DFT calculations, which suggest a decrease in interband transitions in the visible range as well as a reduced reflection in the IR range upon intercalation. We find that a larger interlayer distance in few-layer rGO results in a significant increase in transparency in the infrared region of the spectrum, in agreement with experimental results. Furthermore, a reduced plasma frequency in rGO compared to few-layer graphene is also important to understand the experimental results for broadband transparency in rGO. The broadband transmittance modulation of the CNT/rGO/CNT systems can potentially lead to electrochromic and thermal camouflage applications.

Potassium Ion Batteries with Graphitic Materials.

Graphite intercalation compounds (GICs) have attracted tremendous attention due to their exceptional properties that can be finely tuned by controlling the intercalation species and concentrations. Here, we report for the first time that potassium (K) ions can electrochemically intercalate into graphitic materials, such as graphite and reduced graphene oxide (RGO) at ambient temperature and pressure. Our experiments reveal that graphite can deliver a reversible capacity of 207 mAh/g. Combining experiments with ab initio calculations, we propose a three-step staging process during the intercalation of K ions into graphite: C → KC24 (Stage III) → KC16 (Stage II) → KC8 (Stage I). Moreover, we find that K ions can also intercalate into RGO film with even higher reversible capacity (222 mAh/g). We also show that K ions intercalation can effectively increase the optical transparence of the RGO film from 29.0% to 84.3%. First-principles calculations suggest that this trend is attributed to a decreased absorbance produced by K ions intercalation. Our results open opportunities for novel nonaqueous K-ion based electrochemical battery technologies and optical applications.

Hexagonal BC3: A Robust Electrode Material for Li, Na, and K Ion Batteries.

We have investigated the stability, maximum intercalation capacity, and voltage profile of alkali metal intercalated hexagonal BC3 (MxBC3), for 0 < x ≤ 2 and M = Li, Na, and K. Our calculations, based on dispersion-corrected density functional theory, show that these intercalation compounds are stable with respect to BC3 and their bulk metal counterparts. Moreover, we found that among all MxBC3 considered, the maximum stable capacity corresponds to an x value of 1.5, 1, and 1.5 for Li, Na, and K, respectively. These values are associated with large gravimetric capacities of 572 mA h/g for Na and 858 mA h/g for Li and K. Importantly, we show that metal intercalated hexagonal BC3 has the advantage of a small open-circuit voltage variation of approximately 0.49, 0.12, and 0.16 V for Li, Na, and K, respectively. Our results suggest that BC3 can become a robust alternative to graphitic electrodes in metal ion batteries, thus encouraging further experimental work.

Scalable holey graphene synthesis and dense electrode fabrication toward high-performance ultracapacitors.

Graphene has attracted a lot of attention for ultracapacitor electrodes because of its high electrical conductivity, high surface area, and superb chemical stability. However, poor volumetric capacitive performance of typical graphene-based electrodes has hindered their practical applications because of the extremely low density. Herein we report a scalable synthesis method of holey graphene (h-Graphene) in a single step without using any catalysts or special chemicals. The film made of the as-synthesized h-Graphene exhibited relatively strong mechanical strength, 2D hole morphology, high density, and facile processability. This scalable one-step synthesis method for h-Graphene is time-efficient, cost-efficient, environmentally friendly, and generally applicable to other two-dimensional materials. The ultracapacitor electrodes based on the h-Graphene show a remarkably improved volumetric capacitance with about 700% increase compared to that of regular graphene electrodes. Modeling on individual h-Graphene was carried out to understand the excellent processability and improved ultracapacitor performance.

Accurate Surface Chemistry beyond the Generalized Gradient Approximation: Illustrations for Graphene Adatoms.

Simulations of surface chemistry often use density functional theory with generalized gradient approximations (GGAs) for the exchange-correlation functional. GGAs have well-known limitations for gas-phase chemistry, including underestimated reaction barriers, and are largely superseded by meta-GGAs and hybrids. Our simulations of O and Li adatoms on graphene add to a growing body of evidence that GGAs have similar limitations on surfaces and that meta-GGAs and screened hybrids are computationally feasible for such systems. Meta-GGAs and screened hybrids systematically improve accuracy, just as they do for gas-phase chemistry, motivating their continued exploration in surface chemistry.

Lithium Adsorption on Graphene: From Isolated Adatoms to Metallic Sheets.

We have studied Li adsorption on graphene for Li concentrations ranging from about 1% to 50% by means of density functional theory calculations. At low adsorbant densities, we observe a strong ionic interaction characterized by a substantial charge transfer from the adatoms to the substrate. In this low concentration regime, the electronic density around the Li adatoms is well localized and does not contribute to the electronic behavior in the vicinity of the Fermi level. For larger concentrations, we observe the formation of a chemically bound Li layer characterized by a stronger binding energy as well as a significant density of states above the Fermi level coming from both graphene and the two-dimensional Li sheet.

Accurate prediction of the electronic properties of low-dimensional graphene derivatives using a screened hybrid density functional.

Over the last several years, low-dimensional graphene derivatives, such as carbon nanotubes and graphene nanoribbons, have played a central role in the pursuit of a plausible carbon-based nanotechnology. Their electronic properties can be either metallic or semiconducting depending purely on morphology, but predicting their electronic behavior has proven challenging. The combination of experimental efforts with modeling of these nanometer-scale structures has been instrumental in gaining insight into their physical and chemical properties and the processes involved at these scales. Particularly, approximations based on density functional theory have emerged as a successful computational tool for predicting the electronic structure of these materials. In this Account, we review our efforts in modeling graphitic nanostructures from first principles with hybrid density functionals, namely the Heyd-Scuseria-Ernzerhof (HSE) screened exchange hybrid and the hybrid meta-generalized functional of Tao, Perdew, Staroverov, and Scuseria (TPSSh). These functionals provide a powerful tool for quantitatively studying structure-property relations and the effects of external perturbations such as chemical substitutions, electric and magnetic fields, and mechanical deformations on the electronic and magnetic properties of these low-dimensional carbon materials. We show how HSE and TPSSh successfully predict the electronic properties of these materials, providing a good description of their band structure and density of states, their work function, and their magnetic ordering in the cases in which magnetism arises. Moreover, these approximations are capable of successfully predicting optical transitions (first and higher order) in both metallic and semiconducting single-walled carbon nanotubes of various chiralities and diameters with impressive accuracy. This versatility includes the correct prediction of the trigonal warping splitting in metallic nanotubes. The results predicted by HSE and TPSSh provide excellent agreement with existing photoluminescence and Rayleigh scattering spectroscopy experiments and Green's function-based methods for carbon nanotubes. This same methodology was utilized to predict the properties of other carbon nanomaterials, such as graphene nanoribbons. Graphene nanoribbons may be viewed as unrolled (and passivated) carbon nanotubes. However, the emergence of edges has a crucial impact on the electronic properties of graphene nanoribbons. Our calculations have shown that armchair nanoribbons are predicted to be nonmagnetic semiconductors with a band gap that oscillates with their width. In contrast, zigzag graphene nanoribbons are semiconducting with an electronic ground state that exhibits spin polarization localized at the edges of the carbon nanoribbon. The spatial symmetry of these magnetic states in graphene nanoribbons can give rise to a half-metallic behavior when a transverse external electric field is applied. Our work shows that these properties are enhanced upon different types of oxidation of the edges. We also discuss the properties of rectangular graphene flakes, which present spin polarization localized at the zigzag edges.

Edge effects on the characteristics of li diffusion in graphene.

We study the adsorption and diffusion of Li atoms on the surface of planar graphenes by means of density functional theory. When the dimensionality of graphene is reduced to a quasi-one-dimension, armchair and zigzag edges appear. We show that the presence of these edges affects not only the reactivity of the carbon material toward the adsorption of Li adatoms but also their diffusion properties. These properties strongly depend on the specific morphology of the edges. Our results indicate that Li adatoms will diffuse toward the edges while Li diffusion channels appear along the ribbon axis. For most of the diffusion paths studied here, energy barriers are lower than those in graphene. This effect is significantly more pronounced toward the edges, where energy barriers can be up to 0.15 eV smaller than those in in graphene, producing an increase of up to 2 orders of magnitude in the diffusion coefficient at room temperature. Our results indicate that electrodes fabricated with these materials should increase the power of Li-ion batteries.

Enhanced electrochemical lithium storage by graphene nanoribbons.

Herein, we report the electrochemical Li intake capacity of carbonaceous one-dimensional graphene nanoribbons (GNRs) obtained by unzipping pristine multiwalled carbon nanotubes (MWCNTs). We have found that nanotubes with diameters of approximately 50 nm present a smaller reversible capacity than conventional mesocarbon microbead (MCMB) powder. Reduced GNRs improve the capacity only marginally over the MCMB reference but present a lower Coulombic efficiency as well as a higher capacity loss per cycle. Oxidized GNRs (ox-GNRs) outperform all of the other materials studied here in terms of energy density. They present a first charge capacity of approximately 1400 mA h g(-1) with a low Coulombic efficiency for the first cycle (approximately 53%). The reversible capacity of ox-GNRs is in the range of 800 mA h g(-1), with a capacity loss per cycle of approximately 3% for early cycles and a decreasing loss rate for subsequent cycles.

Electronic properties of the biphenylene sheet and its one-dimensional derivatives.

We have studied the electronic properties and relative stability of the biphenylene sheet composed of alternating eight-, six- and four-carbon rings and its one-dimensional derivatives including ribbons and tubes of different widths and morphologies by means of density functional theory calculations. The two-dimensional sheet presents a metallic character that is also present in the planar strips with zigzag-type edges. Armchair-edged strips develop a band gap that decreases monotonically with the ribbon width. The narrowest armchair strip considered here (0.62 nm wide) presents a large band gap of 1.71 eV, while the 2.14 nm wide armchair strip exhibits a band gap of 0.08 eV. We have also found that tubes made by rolling these ribbons in a seamlessly manner are all metallic, independent of their chirality. However, while the calculated energy landscape suggests that planar strips present a relative stability comparable to that of C(60), in the tubular form, they present a more pronounced metastable nature with a Gibbs free energy of at least 0.2 eV per carbon higher than in C(60).

Magnetic exchange couplings from noncollinear spin density functional perturbation theory.

We propose a method for the evaluation of magnetic exchange couplings based on noncollinear spin density functional calculations. The method employs the second derivative of the total Kohn-Sham energy of a single reference state, in contrast to approximations based on Kohn-Sham total energy differences. The advantage of our approach is twofold: It provides a physically motivated picture of the transition from a low-spin to a high-spin state, and it utilizes a perturbation scheme for the evaluation of magnetic exchange couplings. The latter simplifies the way these parameters are predicted using first principles: It avoids the nontrivial search for different spin states that needs to be carried out in energy difference methods, and it opens the possibility of "black-boxifying" the extraction of exchange couplings from density functional theory calculations. We present proof of concept calculations of magnetic exchange couplings in the H-He-H model system and in an oxovanadium bimetallic complex where the results can be intuitively rationalized.

Magnetic boron nitride nanoribbons with tunable electronic properties.

We present theoretical evidence, based on total-energy first-principles calculations, of the existence of spin-polarized states well localized at and extended along the edges of bare zigzag boron nitride nanoribbons. Our calculations predict that all the magnetic configurations studied in this work are thermally accessible at room temperature and present an energy gap. In particular, we show that the high spin state, with a magnetic moment of 1 muB at each edge atom, presents a rich spectrum of electronic behaviors as it can be controlled by applying an external electric field in order to obtain metallic <--> semiconducting <--> half-metallic transitions.

Accurate solid-state band gaps via screened hybrid electronic structure calculations.

The band energy differences of solids calculated with screened hybrid density functionals, such as the functional of Heyd-Scuseria-Ernzerhof (HSE), reproduce experimental band gaps with a high degree of accuracy. This unexpected result is here rationalized by observing that band energy differences obtained from generalized Kohn-Sham calculations with screened (short-range) Hartree-Fock-type exchange approach the excitation energies obtained via time-dependent density functional calculations with the corresponding unscreened functional. The latter are expected to be the accurate predictions of the experimental optical absorption spectra. While the optimum screening parameter (omega) is system dependent, the HSE standard value of omega=0.11 bohr(-1) represents a reasonable compromise across diverse systems.

Enhanced half-metallicity in edge-oxidized zigzag graphene nanoribbons.

We present a comprehensive theoretical study of the electronic properties and relative stabilities of edge-oxidized zigzag graphene nanoribbons. The oxidation schemes considered include hydroxyl, lactone, ketone, and ether groups. Using screened exchange density functional theory, we show that these oxidized ribbons are more stable than hydrogen-terminated nanoribbons except for the case of the etheric groups. The stable oxidized configurations maintain a spin-polarized ground state with antiferromagnetic ordering localized at the edges, similar to the fully hydrogenated counterparts. More important, edge oxidation is found to lower the onset electric field required to induce half-metallic behavior and extend the overall field range at which the systems remain half-metallic. Once the half-metallic state is reached, further increase of the external electric field intensity produces a rapid decrease in the spin magnetization up to a point where the magnetization is quenched completely. Finally, we find that oxygen-containing edge groups have a minor effect on the energy difference between the antiferromagnetic ground state and the above-lying ferromagnetic state.

Electronic structure and stability of semiconducting graphene nanoribbons.

We present a systematic density functional theory study of the electronic properties, optical spectra, and relative thermodynamic stability of semiconducting graphene nanoribbons. We consider ribbons with different edge nature including bare and hydrogen-terminated ribbons, several crystallographic orientations, and widths up to 3 nm. Our results can be extrapolated to wider ribbons providing a qualitative way of determining the electronic properties of ribbons with widths of practical significance. We predict that in order to produce materials with band gaps similar to Ge or InN, the width of the ribbons must be between 2 and 3 nm. If larger bang gap ribbons are needed (like Si, InP, or GaAs), their width must be reduced to 1-2 nm. According to the extrapolated inverse power law obtained in this work, armchair carbon nanoribbons of widths larger than 8 nm will present a maximum band gap of 0.3 eV, while for ribbons with a width of 80 nm the maximum possible band gap is 0.05 eV. For chiral nanoribbons the band gap oscillations rapidly vanish as a function of the chiral angle indicating that a careful design of their crystallographic nature is an essential ingredient for controlling their electronic properties. Optical excitations show important differences between ribbons with and without hydrogen termination and are found to be sensitive to the carbon nanoribbon width. This should provide a practical way of revealing information on their size and the nature of their edges.

Theoretical nitrogen NMR chemical shifts in octahedral boron nitride cages.

We have calculated the geometrical structure, relative stability, and nitrogen chemical shifts of five boron nitride hollow octahedral cages using density functional theory. Our results show three typical ranges for nitrogen chemical shifts corresponding to each of the nonequivalent magnetic sites of the N atoms. The principal component of the electric field gradient tensor at each 14N site in boron nitride cages is predicted to be much smaller than the corresponding value in borazine, which should reflect in sharper spectral lines and much better resolution.