No Arabic abstract
{it Ab initio} computational methods for electronic transport in nanoscaled systems are an invaluable tool for the design of quantum devices. We have developed a flexible and efficient algorithm for evaluating $I$-$V$ characteristics of atomic junctions, which integrates the non-equilibrium Greens function method with density functional theory. This is currently implemented in the package {it Smeagol}. The heart of {it Smeagol} is our novel scheme for constructing the surface Greens functions describing the current/voltage probes. It consists of a direct summation of both open and closed scattering channels together with a regularization procedure of the Hamiltonian, and provides great improvements over standard recursive methods. In particular it allows us to tackle material systems with complicated electronic structures, such as magnetic transition metals. Here we present a detailed description of {it Smeagol} together with an extensive range of applications relevant for the two burgeoning fields of spin and molecular-electronics.
We have investigated the electronic properties of a C_60 molecule in between carbon nanotube leads. This problem has been tackled within a quantum chemical treatment utilizing a density functional theory-based LCAO approach combined with the Landauer formalism. Owing to low-dimensionality, electron transport is very sensitive to the strength and geometry of interfacial bonds. Molecular contact between interfacial atoms and electrodes gives rise to a complex conductance dependence on the electron energy exhibiting spectral features of both the molecule and electrodes. These are attributed to the electronic structure of the C_60 molecule and to the local density of states of the leads, respectively.
Molecular electronic devices are the upmost destiny of the miniaturization trend of electronic components. Although not yet reproducible on large scale, molecular devices are since recently subject of intense studies both experimentally and theoretically, which agree in pointing out the extreme sensitivity of such devices on the nature and quality of the contacts. This chapter intends to provide a general theoretical framework for modelling electronic transport at the molecular scale by describing the implementation of a hybrid method based on Green function theory and density functional algorithms. In order to show the presence of contact-dependent features in the molecular conductance, we discuss three archetypal molecular devices, which are intended to focus on the importance of the different sub-parts of a molecular two-terminal setup.
Electrically manipulating the quantum properties of nano-objects, such as atoms or molecules, is typically done using scanning tunnelling microscopes and lateral junctions. The resulting nanotransport path is well established in these model devices. Societal applications require transposing this knowledge to nano-objects embedded within vertical solid-state junctions, which can advantageously harness spintronics to address these quantum properties thanks to ferromagnetic electrodes and high-quality interfaces. The challenge here is to ascertain the devices effective, buried nanotransport path, and to electrically involve these nano-objects in this path by shrinking the device area from the macro- to the nano-scale while maintaining high structural/chemical quality across the heterostructure. Weve developed a low-tech, resist- and solvent-free technological process that can craft nanopillar devices from entire in-situ grown heterostructures, and use it to study magnetotransport between two Fe and Co ferromagnetic electrodes across a functional magnetic CoPc molecular layer. We observe how spin-flip transport across CoPc molecular spin chains promotes a specific magnetoresistance effect, and alters the nanojunctions magnetism through spintronic anisotropy. In the process, we identify three magnetic units along the effective nanotransport path thanks to a macrospin model of magnetotransport. Our work elegantly connects the until now loosely associated concepts of spin-flip spectroscopy, magnetic exchange bias and magnetotransport due to molecular spin chains, within a solid-state device. We notably measure a 5.9meV energy threshold for magnetic decoupling between the Fe layers buried atoms and those in contact with the CoPc layer forming the so-called spinterface. This provides a first insight into the experimental energetics of this promising low-power information encoding unit.
We have developed a controlled and highly reproducible method of making nanometer-spaced electrodes using electromigration in ambient lab conditions. This advance will make feasible single molecule measurements of macromolecules with tertiary and quaternary structures that do not survive the liquid-helium temperatures at which electromigration is typically performed. A second advance is that it yields gaps of desired tunnelling resistance, as opposed to the random formation at liquid-helium temperatures. Nanogap formation occurs through three regimes: First it evolves through a bulk-neck regime where electromigration is triggered at constant temperature, then to a few-atom regime characterized by conductance quantum plateaus and jumps, and finally to a tunnelling regime across the nanogap once the conductance falls below the conductance quantum.
We present a study of the effects of inelastic scattering on the transport properties of various nanoscale devices, namely H$_2$ molecules sandwiched between Pt contacts, and a spin-valve made by an organic molecule attached to model half-metal ferromagnetic current/voltage probes. In both cases we use a tight-binding Su-Schrieffer-Heeger Hamiltonian and the inelastic effects are treated with a multi-channel method, including Pauli exclusion principle. In the case of the H$_2$ molecule, we find that inelastic backscattering is responsible for the drop of the differential conductance at biases larger than the excitation energy of the lower of the molecular phonon modes. In the case of the spin-valve, we investigate the different spin-currents and the magnetoresistance as a function of the position of the Fermi level with respect to the spin-polarized band edges. In general inelastic scattering reduces the spin-polarization of the current and consequently the magnetoresistance.