Research

We do research nanoelectronics in the cryolab and with molecules.

Nanoelectronics with molecules

Principle of a mechanically controllable break junction. While bending the flexible substrate using a push-rod, a metallic constriction can be broken to form a pair of atomic sized contacts. The top SEM picture shows a free-standing Au constriction on top of a polyimide-coated stainless steel substrate.

At the ultimate limit of control, nanometer scale devices should be fabricated with atomic precision. Reaching this level of precision is extremely challenging and in many cases impossible. Molecules however can be chemically synthesized with atomic precision and in large quantities. To some level, molecules or also self-assembled metallic and semiconducting clusters can be seen as the smallest building blocks where structure and electronic as well as electro-optical functions can be engineered. Instead of preparing solid-state devices by continuously reducing their size, we therefore combine here micro- and nano-fabrication techniques with chemical bonding and self-assembly to build molecular devices. In such devices, the major electrodes and contact lines are continuous, metallic or semiconducting materials. The nanometer-scale electronically active units are then chemically bonded or adsorbed within the electrode. To enable the formation of molecular junctions, the short organic compounds under investigation need therefore to feature chemical anchoring groups able to bind the contact electrodes. The fabrication and study of molecular devices is highly interdisciplinary and mixes physics, chemistry and engineering. Controlling the positioning, reactivity and stability of molecules is more easily achieved in liquids and at room temperature than in high vacuum at cryogenic temperatures. We therefore perform most of our experiments at room temperature, either in vacuum or in a controlled atmosphere as well as in liquid cells. For testing specific transport properties of the devices, we can also work in vacuum at temperatures ranging between 80K and 400K.

The electric circuit for 4-terminal EM is shown on the top. The voltage drop U over the junction is maintained constant and equal to a preset reference value Uref by the feedback system. The bottom pictures show a scheme of a four-terminal device and a SEM micrograph of an actual device.

The first challenge in the fabrication of a molecular device consists in preparing proper electrodes. This is still today a challenging task and different routes can be followed. Our workhorse is a mechanically controllable break junction (MCBJ) setup allowing to break open and close nanometer scale metallic wires patterned on a flexible substrate. The breaking open of the wire permits the formation of atomic contacts as illustrated in the figure. The anchoring of molecules can then be attempted many times by operating the break junction in a liquid environment containing the molecules to study. A possible alternative to mechanically opening gaps in electrodes consists in doing it electrically. The technique is based on electromigration: the displacement of metallic ions due to a large momentum transfer by electrons at large current densities. To prevent an avalanche breakdown of the metallic bridge during such a process, the experimental trick illustrated in the figure can be played. By performing the electromigration in a 4-terminal geometry with a control electronics, the thinning of the bridge can be controlled and a thermal runaway prevented. Another attractive route to prepare molecular junctions, but this time at a larger scale, makes use of nanometer-scale metallic particles. We synthesize Au nanoparticles a few nm in diameter and coat them with monothiolated alkanes linkers. A self-assembly step followed by a transfer of the assembled particles array on a substrate results the SEM micrograph shown in the figure. The particles form a well-ordered array into which conjugated molecules with particular properties such as electrochemical or optical switching can then be inserted. The particle arrays represent a powerful platform for building networks of molecular junctions and testing their functionality.

To study the fundamental properties of molecular junctions, we mostly work with MCBJ. The junctions can be formed by operating the setup in a liquid cell containing molecules able to bridge the gap. We developed dedicated electronics to characterize the molecular junctions over more than six orders of magnitude in conductance. Using this approach, we can investigate the basic mechanisms controlling the transport properties of molecular junctions. For instance, most of the single-molecule devices use thiols as binding groups to Au electrodes. Other groups, such as amines for example, have been used as well but results are so far still not complete, as many different combinations are possible. It is for instance interesting to form molecular junctions with different binding moieties on the opposite sides in order to form asymmetric junctions. Different binding chemistries will result in different mechanical and electronic coupling. We even recently could show that molecules having only one synthesized binding terminal group can form molecular bridges via pi-staking (S. Wu et al., 2008). There is today a qualitative understanding of electrical transport in molecular junctions for low-bias voltages. In this limit, the conductance through a short oligomer is determined by through-molecule tunneling, which depends exponentially on the length of the molecule and hence on the electrode separation. This exponential dependence has been demonstrated with different techniques by different groups and can be considered an established fact. The physical parameter is the decay constant β, which is determined by an effective barrier height. This height on its own depends on the alignment of the orbital (the HOMO and LUMO) relative to the Fermi energy of the metal electrode. The problem is however more complex as it involves in addition to through molecule tunneling other parameters, which have to be obtained in a self-consistent manner. This is the screening and potential drop along the molecule [38], a possible charge transfer and corresponding reorganization energy, the broadening of electron levels and vibronic and polaronic effects on transport. We therefore strive to gain deeper insight by adding control knobs on the molecular junctions. A working direction consists in adding an electrochemical gate to modulate the transport of the junction by shifting its electronic levels.

SEM micrograph of Au contacts (large pads) deposited on top a stamped line of Au nanoparticles (darker rectangle, in the back of the pads). The zoom is a TEM image showing the inner structure of a printed array of nanoparticles: a regular hexagonal packing can be seen. The interparticle distance is about 2nm.

SEM picture of seven Si nanowires and leads, all etched into the silicon device layer of a Silicon-on-Insulator wafer. The metallic contacts are not visible. The inset shows a single nanowire still capped with its SiO2 layer acting as an etch-mask.

We are also interested in more applied nanoelectronic devices. Field-effect transistors (FETs) made from semiconducting nanowires (NWs) have a great potential as electronic bio-chemical sensors if they can be integrated as an array in a CMOS-compatible architecture together with microfluidic channels and interfacing electronics. In addition to the expected high sensitivity and superior signal quality due to on-chip correlation analysis in time and space, such NW sensors could be mass manufactured at reasonable costs and readily integrated into electronic diagnostic devices to facilitate bed-site diagnostics and personalized medicine. Typical nanowires are shown in the figure. These wires were etched directly into the silicon device layer of SOI (Silicon-on-insulator) wafers. A great advantage of nanowire FETs is their small size which make them ideal candidate for future implanted sensing devices. To target specific biomarkers, the nanowires need to be selectively functionalized with the proper receptors. Here also, an interdisciplinary approach merging the competences from physicists, chemists and biologists is important.

Go to previous page.