Home      Faculty      Research      Fast Dynamics Laboratory      Contact us

Project 2: Nanoscale Spin-Transport Systems

2.1. Magnetic Nanojunctions

This part of the project involves close collaboration between Tsymbal, who has expertise in the field of theoretical condensed matter physics, and Doudin, a junior faculty specialist in the synthesis of nanoscale materials by electrochemical techniques. Recently we used the impurity-assisted tunneling model to explain the properties of the electrodeposited ultra-small MTJs, which demonstrate remarkable TMR behavior (see Figure below). Calculations of TMR performed for disordered MTJs using the Landauer-Büttiker theory, showed that the energy of the impurity state modifies drastically the TMR value, allowing a possible sign reversal.

Magnetoresistance of Ni/NiO/Co junctions of 0.01 µm2 area, curves measured at 4.2K, showing (a) the highest magnitude of TMR observed, corresponding to a TMR ratio of 0.17 (40% using the standard definition of TMR), (b) a small magnitude TMR with a scale magnified by a factor 5, and (c) the largest negative magnitude of TMR, corresponding to a TMR ratio of –0.11 (–25% using the standard definition of TMR).

Comparison between the theory and experiments is shown by the statistical distributions in the Figure below. The broad variation of the TMR values indicates that the specifics of atomic arrangement in magnetic nanojunctions have a considerable impact on spin-dependent transport.

Measured (a) and calculated (b) distribution of magnetoresistance values in Ni/NiO/Co nanojunctions.

Our plan is to synthesize magnetic nanojunctions, of a controlled size between a few atoms and 10 nm. A two-step process will be used. Two electrodes separated by a 50 nm to 500 nm gap will be patterned by standard optical lithography and subsequently by focused ion beam (FIB) milling. In a second step, electrochemical deposition will take place over this pattern, with in-situ and real-time visualization, as well as in-situ electrical characterization. This process will also be performed in the presence of a sweeping magnetic field, allowing investigations of the magnetoresistive properties of the synthesized nanojunctions during their growth. The Figure below shows our actual capabilities for patterning. The electrodeposition is performed very slowly on the two needles separated by a small gap, filling the gap until an electrical contact appears. The resistance between two electrodes is monitored by AC measurements.

Scanning electron microscopy picture of two patterned Au electrodes. Left: after FIB milling, with less than 100 nm separation. Right: After subsequent electrodeposition of Ni, with a separation smaller than 10 nm.

This set-up offers a number of unique possibilities:

· the surface states can be controlled to avoid native oxides, which can strongly influence point-contact studies on magnetic materials.

· contacts are made in a “soft” manner, avoiding stress on samples, a major cause of unreliable mechanical point contacts.

· electrodeposition is a reversible process; we can therefore open or close the atomic-sized contact by controlling the applied potential.

Characterization of the magnetic properties are performed in-situ using Kerr microscopy techniques, with a resolution (500 nm) that allows us to determine if a domain wall configuration exists in the neighborhood of the constriction.
The theoretical work aimes at developing a new method for calculating spin-dependent conductance, applying this method to various magnetic nanojunctions, understanding the most important mechanisms causing magnetoresistive phenomena, and suggesting new structures and materials for the associated experiments. Calculations of the electronic structure of magnetic nanojunctions will be based on the tight-binding linear-muffin-tin-orbital recursion method and the conductance will be modeled within the Landauer-Büttiker theory. This method is a powerful first-principles tool for studying magnetotransport across arbitrary non-periodic systems consisting of up to 1000 atoms, including non-collinear magnetic structures. In particular, we will study the domain-wall resistance. The domain wall of an atomic-size contact is constrained by the area of the contact. This allows us to model the domain-wall resistance from first-principles. We will study the influence of the size and the shape of the nanojunctions and the effect of impurities on the magnitude of magnetoresistance and address the problem of the spin-dependent conductance quantization in these systems.