Having the ability to tune the oxidation state of Cu
nanoparticles is essential for their utility as catalysts. The degree of
oxidation that maximizes product yield and selectivity is known to
vary, depending on the particular reaction. Using first principles
calculations and XANES measurements, we show that for sub-
nanometer sizes in the gas phase, smaller Cu clusters are more
resistant to oxidation. However, this trend is reversed upon
deposition on an alumina support. We are able to explain this result
in terms of strong cluster-support interactions, which differ
significantly for the oxidized and elemental clusters. The stable
cluster phases also feature novel oxygen stoichiometries. Our
results suggest that one can tune the degree of oxidation of Cu
catalysts by optimizing not just their

The self-assembly of supramolecular architectures bound together by noncovalent interactions offers, at present, the most promising route to constructing devices at the nanoscale. However, the ability to pick molecular components that will result in a desired supramolecular structure and functionality remains a challenge. We suggest these goals can be met by using easily computed descriptors that correlate molecular form and formula with supramolecular architecture and energetics. Using a combination of theory and experiment, we show the feasibility of such an approach for a set of molecules comprised of three hosts (carboxylic acid derivatives of phenyleneethynylene) and five guests (naphthalene, phenanthrene, benzo-c-phenanthrene, benzo-ghi-perylene, and coronene), self-assembled on highly oriented pyrolitic graphite. Both scanning tunneling microscopy experiments and density functional theory calculations show that the host–guest combinations display rich structural diversity, assembling in hexagonal, linear, or random glass-like patterns. For certain host–guest combinations, the introduction of the guest triggers structural reorganization, including a disorder-to-order transition. By correlating with computed free energies, we formulate host and guest descriptors. These descriptors can be evaluated at essentially zero computational cost, as they depend only on the geometry and number of chemical motifs of specific types in the isolated molecules. However, they can successfully predict the structures and energetics of the host–guest assemblies, including systems not included in the original database used when determining the form of the descriptors. Structures of the same kind are found to cluster in the descriptor space. This suggests the way forward for the descriptor-based rational design of self-assembled nanostructured systems on surfaces.

Etching of semiconductors by halogens is of vital importance in device manufacture. A greater understanding of the relevant processes at the atomistic level can help determine optimal conditions for etching to be carried out. Supersaturation etching is a seemingly counter-intuitive process where the coverage of the etchant molecules on the surface to be etched is > 1. Here we use density functional theory computations of reaction pathways and barriers to suggest that supersaturation etching of Si(001) by Br2 should be more effective than conventional etching by Br2, as well as both conventional and supersaturation etching by Cl2. Analysis of our results shows that this is due in part to the larger size of bromine atoms, and partly due to Br-Si bonds being weaker than Cl-Si bonds. We also show that for both conventional and supersaturation etching, the barrier for the rate limiting step of desorption of SiX2 units is lower when the halogen X is Br rather than Cl. This contributes to the overall reaction barrier for supersaturation etching being lower for Br2 than Cl2.

We suggest that the reactivity of Au nanocatalysts can be greatly increased by doping the oxide substrate on which they are placed with an electron donor. To demonstrate this, we perform density functional theory calculations on a model system consisting of a 20-atom gold cluster placed on a MgO substrate doped with Al atoms. We show that not only does such substrate doping switch the morphology of the nanoparticles from the three-dimensional tetrahedral form to the two-dimensional planar form, but it also significantly lowers the barrier for oxygen dissociation by an amount proportional to the dopant concentration. At a doping level of 2.78%, the dissociation barrier is reduced by more than half, which corresponds to a speeding up of the oxygen dissociation rate by five orders of magnitude at room temperature. This arises from a lowering in energy of the s and p states of Au. The d states are also lowered in energy, however, this by itself would have tended to reduce reactivity. We propose that a suitable measure of the reactivity of Au nanoparticles is the difference in energy of sp and d states.

Methane, the primary constituent of natural gas, binds too weakly to nanostructured carbons to meet the targets set for on-board vehicular storage to be viable. We show, using density functional theory calculations, that replacing graphene by graphene oxide increases the adsorption energy of methane by 50%. This enhancement is sufficient to achieve the optimal binding strength. In order to gain insight into the sources of this increased binding, that could also be used to formulate design principles for novel storage materials, we consider a sequence of model systems, that progressively take us from graphene to graphene oxide. A careful analysis of the various contributions to the weak binding between the methane molecule and the graphene oxide shows that the enhancement has important contributions from London dispersion interactions as well as electrostatic interactions such as Debye interactions, aided by geometric curvature induced primarily by the presence of epoxy groups. 

Wedemonstrate that the deposition of a self-assembled monolayer of alkanethiolates on a 1 nmthick
cobalt ultrathin film grown on Au(111) induces a spin reorientation transition from in-plane to outof-
plane magnetization. Using ab initio calculations, we show that a methanethiolate layer changes
slightly both the magnetocrystalline and shape anisotropy, both effects almost cancelling each other
out for a 1 nmCo film. Finally, the change in hysteresis cycles upon alkanethiolate adsorption could be
assigned to a molecular-induced roughening of the Co layer, as shown by STM. In addition, we
calculate how a methanethiolate layer modifies the spin density of states of the Co layer and we show
that the spin polarization at the Fermi level through the organic layer is reversed as compared to the
uncovered Co. These results give new theoretical and experimental insights for the use of thiol-based
self-assembled monolayers in spintronic devices.

The growth, morphology, and magnetic structure of ultrathin Cr films grown on a Ag(001) substrate are studied using low-energy electron diffraction (LEED), angle-resolved photoemission spectroscopy (ARPES), and ab initio density functional theory (DFT) calculations. The presence and temperature dependence of c(2×2) half-order spots in the LEED pattern, for low electron energies, along with the presence of characteristic Cr 3d bands in the ARPES spectra, confirm the existence of antiferromagnetic ordering for the Cr monolayer case. Our experiments are also consistent with the presence of a p(1×1) Ag overlayer on top of the Cr layer, suggesting the existence of a Ag/Cr/Ag(001) sandwich structure at the surface. Our DFT calculations confirm that this is the most favored geometric and magnetic structure of the system. The Cr layer is found to retain a “two-dimensional” character with enhanced Cr 3d magnetic moments, despite being buried below a Ag monolayer, due to the absence of significant hybridization between Cr 3d and Ag 4d electronic states. The coverage dependence of the magnetic ordering indicates a maximum ordering above the expected monolayer coverage, possibly due to intermixing between Ag and Cr atoms in the overlayer.

We have used ab initio density functional theory, incorporating van der Waals corrections, to study twisted
bilayer graphene (TBLG) where Stone-Wales defects or monovacancies are introduced in one of the layers. We compare these results to those for defects in single-layer graphene or Bernal stacked graphene. The energetics of defect formation is not very sensitive to the stacking of the layers or the specific site at which the defect is created, suggesting a weak interlayer coupling. However, signatures of the interlayer coupling are manifested clearly in the electronic band structure. For the “γγ” Stone-Wales defect in TBLG, we observe two Dirac cones that are shifted in both momentum space and energy. This up/down shift in energy results from the combined effect of a charge transfer between the two graphene layers and a chemical interaction between the layers, which mimics the effects of a transverse electric field. Charge density plots show that states near the Dirac points have significant admixture between the two layers. For Stone-Wales defects at other sites in TBLG, this basic structure is modified by the creation of minigaps at energy crossings. For a monovacancy, the Dirac cone of the pristine layer is shifted up in energy by ∼ 0.25 eV due to a combination of the requirement of the equilibration of Fermi energy between the two layers with different numbers of electrons, charge transfer, and chemical interactions. Both kinds of defects increase the density of states at the Fermi level. The monovacancy also results in spin polarization, with magnetic moments on the defect of 1.2–1.8 μB.

We have performed ab initio density functional theory calculations, incorporating London dispersion
corrections, to study the absorption of molecular hydrogen on zigzag graphene nanoribbons whose edges have been functionalized by OH, NH2, COOH, NO2, or H2PO3. We find that hydrogen molecules always preferentially bind at or near the functionalized edge, and display induced dipole moments. Binding is generally enhanced by the presence of polar functional groups. The largest gains are observed for groups with oxygen lone pairs that can facilitate local charge reorganization, with the biggest single enhancement in adsorption energy found for “strong functionalization” by H2PO3 (115 meV/H2 versus 52 meV/H2 on bare graphene). We show that for binding on the “outer edge” near the functional group, the presence of the group can introduce appreciable contributions from Debye interactions and higher-order multipole electrostatic terms, in addition to the dominant London dispersion interactions. For those functional groups that contain the OH moiety, the adsorption energy is linearly proportional to the number of lone pairs on oxygen atoms. Mixed functionalization with two different functional groups on a graphene edge can also have a synergistic effect, particularly when electron-donating and electron-withdrawing groups are combined. For binding on the “inner edge” somewhat farther from the functional group, most of the binding again arises from London interactions; however, there is also significant charge redistribution in the π manifold, which directly reflects the electron donating or withdrawing capacity of the functional group. Our results offer insight into the specific origins of weak binding of gas molecules on graphene, and suggest that edge functionalization could perhaps be used in combination with other strategies to increase the uptake of hydrogen in graphene. They also have relevance for the storage of hydrogen in porous carbon materials, such as activated carbons

Using van derWaals-corrected density functional theory calculations, we explore the possibility of engineering the local structure and morphology of high surface-area graphene-derived materials to improve the uptake of methane and carbon dioxide for gas storage and sensing. We test the sensitivity of the gas adsorption energy to the introduction of native point defects, curvature and the application of strain. The binding energy at topological point defect sites is inversely correlated with the number of missing carbon atoms, causing Stone-Wales defects to show the largest enhancement with respect to pristine graphene (~ 20%). Improvements of similar magnitude are observed at concavely curved surfaces in buckled graphene sheets under compressive strain, whereas tensile strain tends to weaken gas binding. Trends for CO2 and CH4 are similar, though CO2 binding is generally stronger by ~ 4–5 kJ mol-1. However, the differential between the adsorption of CO2 and CH4 is much higher on folded graphene sheets and at concave curvatures; this could possibly be leveraged for CH4/CO2 flow separation and gas-selective sensors.

Molecular spintronics seeks to unite the advantages of using organic molecules as nanoelectronic
components, with the benefits of using spin as an additional degree of freedom. For technological
applications, an important quantity is the molecular magnetoresistance. In this work, we show that
this parameter is very sensitive to the contact geometry. To demonstrate this, we perform ab initio
calculations, combining the non-equilibrium Green’s function method with density functional theory,
on a dithienylethene molecule placed between spin-polarized nickel leads of varying geometries. We
find that, in general, the magnetoresistance is significantly higher when the contact is made to sharp
tips than to flat surfaces. Interestingly, this holds true for both resonant and tunneling conduction
regimes, i.e., when the molecule is in its “closed” and “open” conformations, respectively. We find
that changing the lead geometry can increase the magnetoresistance by up to a factor of ∼5. We
also introduce a simple model that, despite requiring minimal computational time, can recapture our
ab initio results for the behavior of magnetoresistance as a function of bias voltage. This model re-
quires as its input only the density of states on the anchoring atoms, at zero bias voltage. We also find
that the non-resonant conductance in the open conformation of the molecule is significantly impacted
by the lead geometry. As a result, the ratio of the current in the closed and open conformations can
also be tuned by varying the geometry of the leads, and increased by ∼400%
 

We have studied the co-adsorption of Br2 and H2 on Si(001), and obtained co-adsorption energies and the surface phase diagram as a function of the chemical potential and pressure of the two gases. To do this, we have used density functional theory calculations in combination with ab initio atomistic thermodynamics. Over large ranges of bromine and hydrogen chemical potentials, the favored configuration is found to be either one with only Br atoms adsorbed on the surface, at full coverage, in a (3 × 2) pattern, or a fully H-covered surface in a (2 × 1) structure. However, we also find regions of the phase diagram where there are configurations with either only Br atoms, or Br and H atoms, arranged in a two-atom-wide checkerboard pattern with a (4 × 2) surface unit cell. Most interestingly, we find that by co-adsorbing with H2, we bring this pattern into a region of the phase diagram corresponding to pressures that are significantly higher than those where it is observed with Br2 alone. We also find small regions of the phase diagram with several other interesting patterns.

Using first principles calculations, we have studied the dielectric properties of crystalline α- and β-phase silicon germanium nitrides and silicon carbon nitrides,

(A = Si, B = Ge or C,

). In silicon germanium nitrides, both the high-frequency and static dielectric constants increase monotonically with increasing germanium concentration, providing a straightforward way to tune the dielectric constant of these materials. In the case of silicon carbon nitrides, the high-frequency dielectric constant increases monotonically with increasing carbon concentration, but a more complex trend is observed for the static dielectric constant, which can be understood in terms of competition between changes in the unit-cell volume and the average oscillator strength. The computed static dielectric constants of C

N

, Si

N

, and Ge

N

are 7.13, 7.69, and 9.74, respectively.

With a view towards optimizing gas storage and separation in crystalline and disordered nanoporous
carbon-based materials, we use ab initio density functional theory calculations to explore the effect
of chemical functionalization on gas binding to exposed edges within model carbon nanostructures.
We test the geometry, energetics, and charge distribution of in-plane and out-of-plane binding of
CO2 and CH4 to model zigzag graphene nanoribbons edge-functionalized with COOH, OH, NH2,
H2PO3, NO2, and CH3. Although different choices for the exchange-correlation functional lead to a
spread of values for the binding energy, trends across the functional groups are largely preserved for
each choice, as are the final orientations of the adsorbed gas molecules. We find binding of CO2 to
exceed that of CH4 by roughly a factor of two. However, the two gases follow very similar trends with
changes in the attached functional group, despite different molecular symmetries. Our results indicate
that the presence of NH2, H2PO3, NO2 , and COOH functional groups can significantly enhance
gas binding, making the edges potentially viable binding sites in materials with high concentrations
of edge carbons. To first order, in-plane binding strength correlates with the larger permanent and
induced dipole moments on these groups. Implications for tailoring carbon structures for increased
gas uptake and improved CO2/CH4 selectivity are discussed

First-principles electronic structure calculations are presented on a
variety of Au compounds and speciesencompassing a wide range of formal
oxidation states, coordination geometries, and chemical environmentsin order
to understand the potentially systematic behavior in the nature and energetics of d
states that are implicated in catalytic activity. In particular, we monitor the
position of the d-band center, which has been suggested to signal catalytic activity
for reactions such as CO oxidation. We find a surprising absence of any kind of
correlation between the formal oxidation state of Au and the position of the dband
center. Instead, we find that the center of the d band displays a nearly linear
dependence on the degree of its filling, and this is a general relationship for Au
irrespective of the chemistry or geometry of the particular Au compound. Across
the compounds examined we find that even small calculated changes in the d-band filling result in a relatively large effect on the
position of the d-band center. The results presented here have some important implications for the question of the catalytic
activity of Au and indicate that the formal oxidation state is not a determining factor

We have studied electronic structure of Fe-deposited Au(111) by performing ab initio density functional

theory calculations. We find that the magnetic moment on the deposited Fe layer is enhanced as compared to that in bulk iron. We observe a large number of new states on the Fe-deposited surface — one of which is in the majority spin channel having similar dispersion to that on the clean surface, and others in the minority spin channel. The effective mass of electrons in surface states near the Fermi level increases on Fe deposition. The electronic properties are found to be insensitive to the stacking of near-surface layers. We need to use very thick slabs in our calculations to avoid splitting of surface states due to spurious interactions between the two surfaces of the slab. Using the local density of states profiles for different surface states, we conclude that in scanning tunneling microscope experiments one can detect two of the surface states — one in the majority channel below the Fermi level, and another in the minority channel appearing just above the Fermi energy. We compare our results to those from scanning tunneling spectroscopy experiments.

The influence of the electrodes Fermi surface on the transport properties of a photoswitching molecule is investigated with state of the art ab initio transport methods. We report results for the conducting properties of the two forms of dithienylethene attached either to Ag or to non-magnetic Ni leads. The I-V curves of the Ag/dithienylethene/Ag device are found to be very similar to those reported previously for Au. In contrast, when Ni is used as electrode material the zero-bias transmission coefficient is profoundly different as a result of the role played by the Ni d bands in the bonding between the molecule and the electrodes. Intriguingly, despite these differences the overall conducting properties depend little on the electrode material. We thus conclude that electron transport in dithienylethene is, for the cases studied, mainly governed by the intrinsic electronic structure of the molecule.

The electronic states of self-organized Fe nanoislands on a Au(111) surface have been investigated

using low-temperature scanning tunneling microscopy and spectroscopy. We show that the local

density of state is dominated by Shockley surface states conned in the nanostructures. Comparing

the experimental dispersion diagram with a free-electron model we derive the eective mass

m=0.39 me and the band onset E0=􀀀420 meV of these states. Ab initio calculations show the

existence of the Shockley surface states in the Fe layer, in agreement with the experiment, and reveal

that they are fully spin polarized.

We present a density functional theory study on the magnetic properties of two-dimensional surface

alloys of the type MxN1 À x (M 1⁄4 Fe, Co and Ni; N 1⁄4 Pt, Au, Ag, Cd and Pb) on Rh(1 1 1) for x 1⁄40.0, 0.25,

   0.33, 0.5, 0.67, 0.75 and 1.0, in two types of geometric arrangements—striped phases or linear-chain

    type, and non-striped phases or mixed checkerboard type. Many pairs among these are bulk-immiscible

   but show mixing on the surface. We find that the trend in the magnetic moment of surface alloys of N

    with a given M follows the number of valence electrons in N: the higher the number of valence

   electrons, the lower the magnetic moment. Overlayer atoms when put on hcp sites show higher

  moment compared to fcc sites. In general, for a given composition x, linear-chain type structures show a

 reduced magnetic moment compared to checkerboard type structures. We find that Pb, when alloyed

  with magnetic elements (Fe, Co and Ni), has a lowering effect on their magnetic moments.

The morphology of small metal clusters can have a big impact on their electronic, magnetic, and chemical properties. This has been shown earlier, for example, for Au

₂₀

clusters on MgO(001), where planar and tetrahedral geometries are possible for the gold atoms. While the planar geometry is more desirable for catalytic applications, it is disfavored in the usual situation. While earlier suggestions that have been made for tilting this balance in favor of the planar isomer are of considerable fundamental interest, they do not easily lend themselves to practical applications. Here, we suggest a conceptually simple but practicable way of achieving the same goal: viz., by doping the MgO substrate with Al atoms. We show, by performing density functional theory calculations, that this stabilizes the planar over the tetrahedral arrangement by an energy difference that is linearly proportional to the dopant concentration and is insensitive to the position of the dopant atom. The charge transferred to the Au cluster also depends monotonically on the doping concentration. This work is of interest for possible applications in the field of gold nanocatalysis.

We have investigated the source of errors in ab initio calculations of thermal properties of the magnetic

metals Fe and Ni and their dependence on the form of the exchange and correlation functional. We used

density-functional theory and density-functional perturbation theory together with the quasiharmonic approxi-

mation to compute the coefficient of thermal expansion, bulk modulus and its pressure derivative, phonon

modes, and Grueneisen parameters of bcc Fe and fcc Ni. In nonmagnetic metals the main source of error in

calculated thermal properties can be attributed to evaluation of properties at incorrect lattice constants, which

in turn may be traced to the choice of exchange and correlation functional. However, for magnetic metals the

properties may be evaluated at both incorrect lattice constant and incorrect magnetic moment. This affects

vibrational properties so that it is no longer true that anharmonic errors are significantly less than errors at

harmonic order.

   Using scanning tunneling microscopy and a diffraction experiment, we have discovered a new ordered

surface alloy made out of two bulk-immiscible components, Fe and Au, deposited on a Ru(0001)

substrate. In such a system, substrate-mediated strain interactions are believed to provide the main

driving force for mixing. However, spin-polarized ab initio calculations show that the most stable

structures are always the ones with the highest magnetic moment per Fe atom and not the ones minimizing

the surface stress, in remarkable agreement with the observations. This opens up novel possibilities for

creating materials with unique properties of relevance to device applications.