Overcoming computational uncertainties to reveal chemical sensitivity in single molecule conduction calculations

Dynamic double lattice of 1-adamantaneselenolate self-assembled monolayers on Au{111}

We report a complex, dynamic double lattice for 1-adamantaneselenolate monolayers on Au{111}. Two lattices coexist, revealing two different binding modes for selenols on gold: molecules at bridge sites have lower conductance than molecules at three-fold hollow sites. The monolayer is dynamic, with molecules switching reversibly between the two site-dependent conductance states. Monolayer dynamics enable adsorbed molecules to reorganize according to the underlying gold electronic structure over long distances, which facilitates emergence of the self-organized rows of dimers. The low-conductance molecules assume a (7 × 7) all-bridge configuration, similar to the analogous 1-adamantanethiolate monolayers on Au{111}. The high-conductance molecules self-organize upon mild annealing into distinctive rows of dimers with long-range order, described by a (6√5 × 6√5)R15° unit cell.

Ghost transmission: How large basis sets can make electron transport calculations worse

The Landauer approach has proven to be an invaluable tool for calculating the electron transport properties of single molecules, especially when combined with a nonequilibrium Green's function approach and Kohn-Sham density functional theory. However, when using large nonorthogonal atom-centered basis sets, such as those common in quantum chemistry, one can find erroneous results if the Landauer approach is applied blindly. In fact, basis sets of triple-zeta quality or higher sometimes result in an artificially high transmission and possibly even qualitatively wrong conclusions regarding chemical trends. In these cases, transport persists when molecular atoms are replaced by basis functions alone ("ghost atoms"). The occurrence of such ghost transmission is correlated with low-energy virtual molecular orbitals of the central subsystem and may be interpreted as a biased and thus inaccurate description of vacuum transmission. An approximate practical correction scheme is to calculate the ghost transmission and subtract it from the full transmission. As a further consequence of this study, it is recommended that sensitive molecules be used for parameter studies, in particular those whose transmission functions show antiresonance features such as benzene-based systems connected to the electrodes in meta positions and other low-conducting systems such as alkanes and silanes.

Exchange and correlation in molecular wire conductance: nonlocality is the key

We study real-time electron dynamics in a molecular junction with a variety of approximations to the electronic structure, toward the ultimate aim of determining what ingredients are crucial for the accurate prediction of charge transport. We begin with real-time, all electron simulations using some common density functionals that differ in how they treat long-range Hartree-Fock exchange. We find that the inclusion or exclusion of nonlocal exchange is the dominant factor determining the transport behavior, with all semilocal contributions having a smaller effect. In order to study nonlocal correlation, we first map our junction onto a simple Pariser-Parr-Pople (PPP) model Hamiltonian. The PPP dynamics are shown to faithfully reproduce the all electron results, and we demonstrate that nonlocal correlation can be readily included in the model space using the generator coordinate method (GCM). Our PPP-GCM simulations suggest that nonlocal correlation has a significant impact on the I-V character that is not captured even qualitatively by any of the common semilocal approximations to exchange and correlation. The implications of our results for transport calculations are discussed.

Stochastic modulation in molecular electronic transport junctions: molecular dynamics coupled with charge transport calculations

The experimental variation in conductance that can be expected through dynamically evolving Au-molecule-Au junctions is approximated using molecular dynamics to model thermal fluctuations and a nonequilibrium Green's function code (Hückel-IV 2.0) to calculate the charge transport. This generates a statistical set of conductance data that can be used to compare directly with experimental results. Experimental measurements on Au-single molecule junctions show a large variation in conductance values between different identically prepared junctions. Our computational results indicate that the Au-Au and the Au-molecule fluctuations provide extensive geometric freedom and an associated broad distribution in calculated conductance values. Our results show agreement with experimental measurements of the low bias voltage conductance and conductance distribution for both thiol-Au and amine-Au linker structures. -

Successful a Priori Modeling of CO Adsorption on Pt(111) Using Periodic Hybrid Density Functional Theory

The adsorption of CO on the surface of metals such as Pt(111) is of great interest owing to the industrial importance of the catalytic oxidation of pollutant CO. To date, reliable high-level calculations of this process have not been possible, a situation often referred to as the "CO/Pt(111) puzzle". Standard generalized-gradient-approximation density functional theory approaches fail to capture key details of the binding, such as the location of the adsorption site, while cluster approaches using alternative methods show some but insufficient improvement. Using a new computational methodology combining hybrid density functionals containing non-local Hartree-Fock exchange with periodic imaging plane-wave-based techniques, we demonstrate that key aspects of the adsorption of CO on Pt(111), including the identification of the absorption site and CO frequency change, can now be adequately modeled. The binding is dominated by both CO dative covalent bonding and metal-to-molecule pi back-bonding, effects requiring realistic alignment of both the molecular HOMO and LUMO orbitals with respect to the metal Fermi energy.

The Green's Function Density Functional Tight-Binding (gDFTB) Method for Molecular Electronic Conduction

A review is presented of the nonequilibrium Green's function (NEGF) method "gDFTB" for evaluating elastic and inelastic conduction through single molecules employing the density functional tight-binding (DFTB) electronic structure method. This focuses on the possible advantages that DFTB implementations of NEGF have over conventional methods based on density functional theory, including not only the ability to treat large irregular metal-molecule junctions with high nonequilibrium thermal distributions but perhaps also the ability to treat dispersive forces, bond breakage, and open-shell systems and to avoid large band lineup errors. New results are presented indicating that DFTB provides a useful depiction of simple gold-thiol interactions. Symmetry is implemented in DFTB, and the advantages it brings in terms of large savings of computational resources with significant increase in numerical stability are described. The power of DFTB is then harnessed to allow the use of gDFTB as a real-time tool to discover the nature of the forces that control inelastic charge transport through molecules and the role of molecular symmetry in determining both elastic and inelastic transport. Future directions for the development of the method are discussed.

Benchmarking the performance of density functional theory based Green’s function formalism utilizing different self-energy models in calculating electronic transmission through molecular systems

Electronic transmission through a metal-molecule-metal system is calculated by employing a Green's function formalism in the scattering based scheme. Self-energy models representing the bulk and the potential bias are used to describe electron transport through the molecular system. Different self-energies can be defined by varying the partition between device and bulk regions of the metal-molecule-metal model system. In addition, the self-energies are calculated with different representations of the bulk through its Green's function. In this work, the dependence of the calculated transmission on varying the self-energy subspaces is benchmarked. The calculated transmission is monitored with respect to the different choices defining the self-energy model. In this report, we focus on one-dimensional model systems with electronic structures calculated at the density functional level of theory.

The symmetry of single-molecule conduction

We introduce the conductance point group which defines the symmetry of single-molecule conduction within the nonequilibrium Green's function formalism. It is shown, either rigorously or to within a very good approximation, to correspond to a molecular-conductance point group defined purely in terms of the properties of the conducting molecule. This enables single-molecule conductivity to be described in terms of key qualitative chemical descriptors that are independent of the nature of the molecule-conductor interfaces. We apply this to demonstrate how symmetry controls the conduction through 1,4-benzenedithiol chemisorbed to gold electrodes as an example system, listing also the molecular-conductance point groups for a range of molecules commonly used in molecular electronics research.

Molecular origins of conduction channels observed in shot-noise measurements

Measurements of shot noise from single molecules have indicated the presence of various conduction channels. We present three descriptions of these channels in molecular terms showing that the number of conduction channels is limited by bottlenecks in the molecule and that the channels can be linked to transmission through different junction states. We introduce molecular-conductance orbitals, which allow the transmission to be separated into contributions from individual orbitals and contributions from interference between pairs of orbitals.

Understanding the inelastic electron-tunneling spectra of alkanedithiols on gold

We present results for a simulated inelastic electron-tunneling spectra (IETS) from calculations using the "gDFTB" code. The geometric and electronic structure is obtained from calculations using a local-basis density-functional scheme, and a nonequilibrium Green's function formalism is employed to deal with the transport aspects of the problem. The calculated spectrum of octanedithiol on gold(111) shows good agreement with experimental results and suggests further details in the assignment of such spectra. We show that some low-energy peaks, unassigned in the experimental spectrum, occur in a region where a number of molecular modes are predicted to be active, suggesting that these modes are the cause of the peaks rather than a matrix signal, as previously postulated. The simulations also reveal the qualitative nature of the processes dominating IETS. It is highly sensitive only to the vibrational motions that occur in the regions of the molecule where there is electron density in the low-voltage conduction channel. This result is illustrated with an examination of the predicted variation of IETS with binding site and alkane chain length.