Single Molecule transistors

Motivation

This project deals with the fabrication of electronic devices from single molecules. Using nanofabrication techniques we make electrical contact to single molecules. We then study the transport properties of these devices at low temperatures (below 4 K).

Project Summary

We are interested in making electronic devices from single molecules. These devices show interesting physical properties due to their small size. Work has been done previously on such devices by other research groups. A few examples of such research are the fabrication of transistors from single Carbon-60 molecules, carbon nanotubes and conjugated organic molecules.

Project details

The type of devices we make. A small molecule (1-2 nm long) is attached to two gold electrodes which are made on an oxidized silicon surface

In order to make devices like the one represented above, we have to first create two gold electrodes that are 1-2 nm apart. The technique we use was first used by the McEuen group to make devices. We start by making a thin gold wire (dimensions of 200 nm * 10 nm * 30 nm). These wires are made at the Cornell Nanofabrication Facility using standard lithographic techniques.

An SEM micrograph of a 20 nm - wide gold wire made at the CNF using electron beam lithography.

We then apply some voltage across the ends of the wire. The wire is like a tiny fuse - at some voltage and current it will break. if this procedure is done right, the break in the wire can be as small as 0.5 - 2 nm (see Park et. al.'s paper on the subject).  


A TEM image of a gold wire broken by applying voltage across its ends. The break can be as small as 0.5 nm. Such gaps are difficult to image. Here we have chosen a wire with a large break of 7 nm.

Now we are ready to stick stuff between the electrodes. Currently we are studying molecules made in Hector Abruna's group in the chemistry department at Cornell.
Representations of the molecules studied. At the center of each of the molecules is a single metallic atom (cobalt,manganese or iron) shown in dark blue. It is linked to two sets of 3 nitrogen atoms that are shown in blue. Each of the sets of 3 nitrogen atoms is a part of a terpyridinyl group (like three benzene rings stuck together). At the ends of the molecule are thiol groups (shown in yellow) which bond very strongly to gold surfaces. The atoms in black are carbon. The two molecules are synthesized with different lengths by inserting a 5 carbon alkyl chain in one of the molecules (top).


These molecules have some interesting electrochemical properties (see the Abruna group page for some electrochemistry details).
We perform the solid-state equivalent of the electrochemistry experiment. By wiring up these molecules to two leads, we can send electrons onto the central atom by changing the voltage between the two leads. Current thus flows by electrons hopping on and off a single metallic atom.
In order to be able to stabilize this small molecule between the gold contacts, we have to cool it down to low temperatures. We use three different cryostats to measure these devices. For the coarsest measurements we use a dipstick that sits in a liquid helium dewar. This gets the temperature down to 4.2 K. We use a pumped helium-4 cryostat for better measurements (temperature down to 1.5 K). The temperature of the sample can be controlled between 1.5 K and room temperature in this cryostat. Lowering the temperature gives us better energy resolution on our data. For the best resolution we use a dilution refrigerator (temperature down to ~ 0.05 K). We can control the temperature of the sample between 0.05 K and ~ 1 K in this cryostat.
The main feature of the transport properties of devices made with the longer molecule is that they exhibit "coulomb blockade". This "single electron tunneling" has been studied in many previous experiments in our group. The feature of this experiment is that the piece of metal is exactly one atom.


Coulomb blockade in the longer molecule. The different curves are for different gate voltages.

We also see excited quantum states that contribute to the transport through these molecules. We believe that these states are the vibrational modes of the molecule. Connie Chang is currently simulating the vibrational modes of these molecules.


Differential conductance of the device as a function of gate and source voltages. The light areas are higher conductance. See Mandar's paper if you have trouble understanding this figure.

The shorter molecule is coupled much better to the electrodes. As a consequence of the strong coupling between the leads and the cobalt atom, we see the Kondo effect in these short-molecule devices.

Differential conductance of the device as a function of bias voltage. The peak at zero bias is due to the Kondo effect. Bare gold wires (inset) do not show this.

The Kondo effect has many interesting consequences, such as the splitting of the conductance peak in a high magnetic field, shown below.


Differential conductance as a function of source bias and applied magnetic field. Bright areas are high conductance. The zero field peak splits into two peaks at high field.

We are currently studying the Kondo effect in these molecules in greater detail. We are also studying other interesting quantum dots with this technique.

People involved

The nanofabrication and measurements are done by Abhay Pasupathy and Jiwoong Park. We work with Dan Ralph and Paul McEuen. The molecules are synthesized by Jonas Goldsmith who works with Hector Abruna. Connie Chang, who works with Jim Sethna, is simulating the normal modes of these molecules.