•SEI layer on the Sn anode of a Na-ion battery consists of Na2O and Na2CO3.•Ionic conductivity of the SEI layer is very low close to the discharged state.•Ionic conductivity of Na2O is much larger than that of Na2CO3.Structure, stability, and ionic conductivity of the SEI layer on Sn anodes in Na-ion batteries (NIBs) are studied using experimental and theoretical methods. Raman spectra show that the SEI layer consists of Na2O and Na2CO3, the latter becoming more dominant close to the discharged state (at 0.3 V). According to our theoretical phase diagrams, Na2O can be stable in the whole voltage range of charge/discharge (from 0.0 V to 1.90 V), but Na2CO3 can decompose under carbon and/or oxygen poor conditions, leading to the formation of Na2O. These findings are in agreement with our experimental cyclic voltammetry and Raman spectra as function of voltage. Both compounds of the SEI layer have very low ionic conductivity close to the discharge state (0.2–0.3 V), but the ionic conductivity of Na2O is much larger than that of Na2CO3 for a wide range of voltages from 0.4 V to the charge state (∼1.5 V). This work suggests that engineered artificial SEI with Na2O or naturally formed SEI in a carbon and/or oxygen poor environment can improve the conductivity of the SEI layer in NIBs.

Influence of metal oxide (MO) supports on nanoparticle (NP) catalysts is still under investigation. Theoretical studies demonstrate that active defect sites on the surface of a MO support can affect the structure and activity of metal clusters. In the present work, we show that even defect-free surfaces of MOs can cause considerable restructuring and accumulation of interfacial charges on Pt NPs of size 1 nm (Pt55). Independent of the type of MO support, we find that supported Pt55 behaves like a conductor since the binding energy of a test adsorbate on top of it is similar to that on an intact Pt55. However, adsorption energy at binding sites close to the perimeter of the nanoparticle/support interface can vary by 1.8 eV depending on the distance between the adsorbate and surface cations (possibility of forming ionic bonds) as well as the amount and sign of charges (ionization energy) of interfacial Pt atoms.

Highlights•DRIFTS measurements show three bands for CO adsorbed on Pt nanoparticles depending on the applied flow and pretreatment conditions.•DFT calculations indicate that the IR bands are due to a combined size and site effect.•For fully covered small nanoparticles the IR bands are attributed to all binding sites.•For fully covered larger nanoparticles the dominant contribution is related to {111} facets but the other bands are still site independent.Extensive research has been devoted to the assignment of IR bands of CO adsorbed on Pt nanoparticles, which are widely used in heterogeneous and electrocatalysis (e.g. fuel cells). In contrast to single crystal studies, the assignment of CO adsorption to the nanoparticle structure is still controversial. Here we present a case study where we assign CO adsorption bands to the structure of Platinum nanoparticles with a given size distribution. Using a special diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) cell allows to achieve high quality data under in-situ conditions. Temperature dependent CO adsorption spectra are resolved into three bands which depend on the applied flow and pretreatment conditions. Our calculations using Density Functional Theory (DFT) can mimic the experimental findings and link these bands to the particle structure. By explicitly calculating the IR spectra of CO/Pt nanoparticles of different sizes we show that the IR bands are due to a combined size and site effect. For fully covered small nanoparticles the IR bands are attributed to all binding sites. For larger nanoparticles the dominant contribution is related to {111} facets but the other bands are still site independent. Here we provide a tool to assign CO adsorption bands on Platinum nanoparticles with a given size distribution. This can be related to the structure–acitvity relationship which is required for a tailored catalyst design.Graphical abstract

•Adsorption of Li2Sx on graphene is dominated by physisorption.•Vacancy defective graphene catches one S atom from Li2Sx, leading to the formation of a Li2Sx−1 molecule.•Li2Sx−1 molecules adsorb weakly on the created S-doped graphene.•SW and vacancy defect sites can not improve the ability of graphene to catch lithium polysulfides.Adsorption of Li2Sx on pristine and defective (Stone-Wales (SW) and vacancy) graphene is studied using density functional theory. Results show that the interaction between Li2Sx and graphene is dominated by dispersion interaction (physisorption), which depends on the size of molecule as well as the existence and type of defect sites on graphene. We find that single Li2Sx molecules interact only slightly stronger to the SW sites than to the defect-free sites, but they interact very strongly with single-vacant defects. In the later cases, the vacant site catches one S atom from the Li2Sx molecule, leading to the formation of a Li2Sx−1 molecule, which adsorbs weakly on the created S-doped graphene. This study suggests that defect sites can not improve the ability of graphene to catch lithium polysulfides in Li-S batteries.

We report on a comparative density functional theory (DFT) study of ethylene carbonate (EC) decomposition on unlithiated, Li-covered, and fully-lithiated Sn surfaces as well as a pure Li surface. Results show that EC molecule does not dissociate on unlithiated Sn modelled by the clean β–Sn(100) surface, but rather adsorbs intactly via dispersion interaction. However, EC molecule spontaneously dissociates over Li-covered β–Sn(100), fully-lithiated Li17Sn4(001), and Li(100) surfaces. The dissociated EC molecule (C2H4+CO3) is energetically much more favorable than the intact EC molecule on these surfaces. We find that the dissociative adsorption of EC molecule is stronger on Li/β–Sn(100) and Li17Sn4(001) than on Li(100) in spite of larger electron transfer in the last case. This result indicates that besides electron transfer, chemical composition also plays an important role on dissociation of EC molecules. A detached CO3−2 anion catches two Li atoms from the Li-rich surfaces and forms a Li2CO3 molecule, which is the building block of a main component of the solid electrolyte interphase (SEI) layer on Sn anodes according to the experimental observations.

Structure and activity of nanoparticles of hexagonal close-packed (hcp) metals are studied using first-principles calculations. Results show that, in contact with a nitrogen environment, high-index {132} facets are formed on hcp metal nanoparticles. Nitrogen molecules dissociate easily at kink sites on these high-index facets (activation barriers of <0.2 eV). Analysis of the site blocking effect and adsorption energies on {132} facets explains the order of activity of hcp metals for ammonia synthesis: Re < Os < Ru. Our results indicate that the high activity of hcp metals for ammonia synthesis is due to the N-induced formation of {132} facets with high activity for the dissociation of nitrogen molecules. However, quite different behavior for adsorption of dissociated N atoms leads to distinctive activity of hcp metals.

The lithiation process in TiO2-B as a function of Li content x is studied using density functional theory. Considering a variety of possible pathways, we find that stepwise insertion of Li in TiO2-B layers, followed by insertion of Li in O layers, is kinetically the most favorable scenario. Diffusion coefficient (D) varies only by 2 orders of magnitude with x for Li contents up to x = 0.75 (3.03 × 10–10 cm2/s ≤ D ≤ 7.94 × 10–8 cm2/s), but it becomes almost 6 and 26 orders of magnitude smaller for 0.75 < x ≤ 1.0 and 1.00 < x < 1.25, respectively. To make a direct comparison to experimental measurements, the maximum diffusion length was estimated for different values of C-rate. It is found that, for any reasonable value of C-rate between 0.1C and 10C, bulk TiO2-B cannot attain capacities larger than 251 mAh g–1. However, larger capacities up to 335 mAh g–1 can be obtained by nanosized TiO2-B. In addition, we find that a capacity higher than 335 mAh g–1 cannot be achieved. Our results suggest that the small discharge capacity of bulk TiO2-B is due to the large increase in the energy barrier of Li diffusion for Li contents above x = 0.75.

•We study lithiation of Si from a kinetic and thermodynamic point of view.•Anisotropic lithiation of Si is driven by an anisotropy of mobilities.•The anisotropy of mobilities does not originate from orientation-dependent kinetics of Li.•Anisotropy of mobilities is a consequence of orientation-dependent interface energies.•The present work provides insights into lithiation of Li-alloying anode materials.Silicon has the highest known theoretical capacity (∼4140 mAhg−1) to store lithium. Among different forms of Si, Si nanowires (SiNWs) are the most promising candidates for the next-generation lithium-ion batteries. Lithiation of SiNWs is a complex process, which is not very well understood. Here, we present density functional theory calculations on Li incorporation in SiNWs using surface and interface geometries. Our results show that, initially, Li intercalation proceeds through Si(110) facets, leading to the formation of an amorphous Li2Si shell. For interfaces between the lithiated (amorphous Li2Si) shell and unlithiated (pristine c-Si) core region we find that the Li intercalation barriers are independent of the actual interface orientation, while interface energies show an orientation dependence similar to surface energies. In particular, a-Li2Si/c-Si(111) is most favorable while a-Li2Si/c-Si(110) is least favorable. Since high-energy interfaces typically show a higher mobility than low-energy interfaces, the experimentally observed anisotropic swelling of SiNWs can be understood on the basis of interface energetics.

Mechanism of Li diffusion at the LiCoO2(104) surface and in bulk LiCoO2 is studied using density functional theory calculations. We find that there is almost no barrier for the diffusion of Li between the two topmost surface layers. The results show that Li intercalation occurs by the diffusion of Li ions from the first layer to the divacancy of Li sites created by removal of two neighboring Li ions in the first and second layer. However, Li deintercalation occurs by the diffusion of Li ions from the second layer to the missing row of topmost Li sites. The energy barrier for the process of intercalation/deintercalation of Li between the second and third surface layers is also lower than that in the bulk. This finding indicates that nanosized LiCoO2 with a large surface area/volume ratio is a promising cathode material for fast charging/discharging Li-ion batteries.

The structure of tin (Sn) nanoparticles as function of size and temperature has been studied using density functional theory and thermodynamic considerations. It is known that bulk Sn undergoes a transition from the semiconducting α-phase to the metallic β-phase at temperatures higher than 13.2 °C under atmospheric pressure. Here we show that, independent of temperature, Sn nanoparticles smaller than 8 nm diameter always crystallize in the β-phase structure in thermodynamic equilibrium, and up to a size of 40 nm of the Sn nanoparticles this metallic phase is stable at all reasonable ambient temperatures (≳−40 °C). The transition to the metallic phase is caused by nanoscale stabilization due to the lower surface energies of the β phase. This study suggests that the atomic structure and conductivity of nanostructured Sn anodes can change dramatically with size and temperature. This finding has implication for understanding the performance of future Li-based batteries since Sn nanostructures are considered as one of the most promising anode materials, but the mechanism of nanoscale stabilization might be used as a design strategy for other materials.

The Li–S battery is the most promising candidate for future electric vehicles. The study of stabilities and conductivities of Li2S nanoparticles, which are used as pre-lithiated cathode materials, is crucial for the development of Li–S batteries. Here, we investigate the atomic and electronic structures as well as stabilities of Li2S surfaces and nanoparticles using density functional theory (DFT) and classical electrostatic models. We show that Li2S nanoparticles have octahedral shape and consist of only (111) facets. At low concentrations of Li, the surfaces of nanoparticles are metalized non-polar Li2S(111) surfaces. The metalization is found to be due to the depletion of valence bands of surface states. However, for higher concentrations of Li, the nanoparticle faces are insulator non-polar Li2S(111) surfaces. This study suggests that Li2S nanoparticles with (111) surfaces are very promising cathode materials for Li–S batteries.

We report results from new experiments on C/Re(\(11\bar{2}1\)) to identify threshold conditions for morphological instability of Re(\(11\bar{2}1\)). We have found that adsorption of carbon from 0.35 to 0.85 ML (0.3–6.0 L exposure of C2H2) at T ≥ 800 K leads to faceting of Re(\(11\bar{2}1\)) with formation of three-sided pyramids. Using density functional theory we have investigated binding sites and binding energies of C on planar and faceted Re surfaces as well as generated a surface phase diagram of C/Re to obtain an atomistic understanding of C-induced pyramidal faceting of Re(\(11\bar{2}1\)). The calculations reveal that at low to intermediate coverage, C atoms prefer binding at four-fold sites on the Re surfaces and formation of three-sided pyramids is thermodynamically favored.

Structure and activity of nanoparticles of hexagonal close-packed (hcp) metals are studied using first-principles calculations. Results show that, in contact with a nitrogen environment, high-index {132} facets are formed on hcp metal nanoparticles. Nitrogen molecules dissociate easily at kink sites on these high-index facets (activation barriers of <0.2 eV). Analysis of the site blocking effect and adsorption energies on {132} facets explains the order of activity of hcp metals for ammonia synthesis: Re < Os < Ru. Our results indicate that the high activity of hcp metals for ammonia synthesis is due to the N-induced formation of {132} facets with high activity for the dissociation of nitrogen molecules. However, quite different behavior for adsorption of dissociated N atoms leads to distinctive activity of hcp metals.

The structure of tin (Sn) nanoparticles as function of size and temperature has been studied using density functional theory and thermodynamic considerations. It is known that bulk Sn undergoes a transition from the semiconducting α-phase to the metallic β-phase at temperatures higher than 13.2 °C under atmospheric pressure. Here we show that, independent of temperature, Sn nanoparticles smaller than 8 nm diameter always crystallize in the β-phase structure in thermodynamic equilibrium, and up to a size of 40 nm of the Sn nanoparticles this metallic phase is stable at all reasonable ambient temperatures (≳−40 °C). The transition to the metallic phase is caused by nanoscale stabilization due to the lower surface energies of the β phase. This study suggests that the atomic structure and conductivity of nanostructured Sn anodes can change dramatically with size and temperature. This finding has implication for understanding the performance of future Li-based batteries since Sn nanostructures are considered as one of the most promising anode materials, but the mechanism of nanoscale stabilization might be used as a design strategy for other materials.

Mechanism of Li diffusion at the LiCoO2(104) surface and in bulk LiCoO2 is studied using density functional theory calculations. We find that there is almost no barrier for the diffusion of Li between the two topmost surface layers. The results show that Li intercalation occurs by the diffusion of Li ions from the first layer to the divacancy of Li sites created by removal of two neighboring Li ions in the first and second layer. However, Li deintercalation occurs by the diffusion of Li ions from the second layer to the missing row of topmost Li sites. The energy barrier for the process of intercalation/deintercalation of Li between the second and third surface layers is also lower than that in the bulk. This finding indicates that nanosized LiCoO2 with a large surface area/volume ratio is a promising cathode material for fast charging/discharging Li-ion batteries.