Instruction for the Experiment: Mechanically Controllable
Break Junctions
Xiang Dong, Maristella Coppola, Volodymyr Handziuk, Dirk Mayer
Peter Grünberg Institut, PGI-8, Bioelektronik, Forschungszentrum Juelich
1. INTRODUCTION
In the last five decades scientists have tried to study the electrical properties of single molecules
in order to use them as promising components of electronic devices. Several techniques have
been used and developed to solve this task. Among these, Mechanically Controllable Break
Junction (MCBJ) method has been defined as a fundamental technique [1] that enables us to
measure the electrical properties of different molecular species, such as molecular insulators,
wires, diodes, and molecular switches [2]. In this approach, a molecule is wired, via covalent
bond, between two metal electrodes. The distance between the electrodes can be tuned
precisely, down to picometer resolution, to match the length of a single molecule.
The purpose of this experiment is to realize a metal-molecule-metal junction and to investigate
its own electrical properties.
More specifically, there are three main goals:
1. Determine the value of quantum conductance G0
2. Determine the single molecular conductance
3. Determine the I-V characteristics of the single molecule junction
2. PRINCIPLE
JUNCTION
OF
THE
MECHANICALLY
CONTROLLABLE
BREAK
Our investigation will be performed using two different setups: MCBJ_enviromental and
MCBJ_vacuum, that allow us to conduct experiments in air and vacuum respectively.
Figure 1. Top View schematic of nanofabricated electrodes
Samples, needed for these measurements will be fabricated by scientists from Institute of
Bioelectronics (PGI-8). To obtain electrodes with dimensions in the nanometer scale it is
necessary to use electron beam lithography. Furthermore it is possible to realize free-standing
junction by means of reactive ion etching (RIE).
The samples will be mounted into a homemade three-point bending apparatus, see
Figure 2. This name derives from the fact that the bending of the substrate is performed by
applying force in three points: the push rod in the middle and two counter supports on the sides.
The push rod is moved along the Z direction by a piezoelectric motor, and therefore exerts force
at the center of the sample. This process causes an elongation of the electrodes and narrowing
the constrictions until single-point contact is realized and finally they break.
In order to make the bending process “mechanically” feasible, it is necessary to use flexible
substrate, hence spring steel is the material of choice. Consequently to make it “controllable”
there is a geometrical key factor to take into account: the attenuation factor. The latter is defined
2
as a  x / z  6ut / L of 3.3  107 where x is the gap distance change, and z is the
push rod displacement, which implies that we are, in principle able to control the distance between
the electrodes with sub-picometer accuracy; u is the width of the suspend bridge; t is the
thickness of the steel substrate; L is the length between the two outer support points.
Figure 2. The setup of MCBJ
Figure 3. Principle of the MCBJ
3. EXPERIMENT
3.1 Setting the experiment
Before the electrical measurement can be performed, one should get familiar with the
measurement setup configuration.
Each group of students would have to:
1. mount the sample in three points bending setup
2. Contact it to a measurement circuit (probe station)
3. Get familiar with the software of the Keithley Multichannel Parameter Analyzer. This
device is operated as voltage source and measures the current of the break junction as
function of time and bias voltage. You will get an instruction and record IV responses of
model devices
Figure 4.Probe station
Figure 5. Software interface and the hardware
of Keithley for electric measurement. From
above to bottom is Keithley interface test
environment, Keithley 595 Quasistatic CV
meter, Keithley 590 CV analyzer respectively
4. You will learn how to operate the piezo actuator, see figure 6 and figure 7. The piezo
actuator PIE-755 is responsible for the movement of the push rod in z direction and is
controlled via a computer interface. You can set defined starting and end positions for the
movement of the push rod. Furthermore, the moving velocity can be adjusted.
Figure 6. Piezo actuator PI E-755
Figure7. Interface of PI Mikromove for PIE-755
3.2 Data acquisition
The task will be to bend the nanofabricated gold electrodes until a gap is formed between them.
Concurrently the conductance of the system will be measured. By bending the electrodes will be
stretched, the constrictions will be narrowed, the gold atoms will rearrange until single-point
contact Figure 8. At this point, corresponding conductance values of one or multiples of
G0 (G0  2e2 / h) can be obtained due to quantization of the electron transport through the
junction. After breaking the junction, the electron transport is controlled by tunneling processes.
The distance between the electrodes for both opening and closing directions can be controlled by
bending or relaxing the substrate, respectively. The gap between the electrodes can be adjusted
with picometer accuracy.
Once a nanogap has been realized it is necesary to break and close repeatedly a gold wire
junction and to record several tens of breaking curves. From these individual breaking traces you
should calculate a conductance histogram and determine the value of G0 for a defined bias.
Figure 8. Single breaking curve showing
quantized conductance of gold.
Figure 9. Conductance histogram of different metals.
3.3 Measure the single molecular conductance
To determine Gm, the conductance of the target molecule, it is necessary to realize a
metal/molecules/metal junction. It is essential that molecule, that one wants to study, has binding
groups, at both ends, able to bridge the two gold electrodes. Target molecules for this experiment
will be Octanedithiol (ODT) and the dipeptide double Cysteine (Cys-Cys), both having thiol group
as terminus, hence strong binding group for the gold. In order to produce a molecular junction,
few droplets of a 1 mM ethanolic solution of a model molecule will be placed onto the sample, at
the junction area.
The whole breaking process of the junction can be monitored by the conductance
measurement of the junction, see for example Figure 8 and 9. Due to the ability of the molecule
to form covalent bonds to the gold electrodes, a metal-molecule-metal junction will form
automatically after the constriction breaks. The two tips will be bridged by only few molecules
since the two tips are atomically sharp. Further elongation of the junction will consecutively result
in a rupture of all molecule-electrode bonds until a single molecule junction remains, see
Figure 10 and 11.
Figure 10. Conductance change in the break process
Figure 11. Metal/molecules/metal junction
Figure 12. Cys-Cys
Gm can be determined from the last plateau. To give consistence to this result it is required to
perform a statistical analysis. Specifically one will need to break and close repeatedly the metalmolecule-metal junction, building and analyzing a conductance histogram, thence to determine
the single molecule conductance for a defined bias. You will investigate one of the following two
model molecules: octanedithiol (CH2)8(SH)2, double-cysteine (C5H10N2O3)(SH)2, see Figure 12.
You should try to estimate single molecule conductance of your molecule from the last plateau
already during the experiment and compare this value with the conductance that you have
determined from the conductance histogram.
3.4 Measure I-V curves of single molecular junction
The single molecule conductance that you have determined from the last plateau can be used as
benchmark to establish a single molecule junction. Therefore, you should elongate a closed
junction until you are close to conductance values of 1G0. At this point you should interrupt the
bending process and wait unit a stable conductance has established. Afterwards you continue
the breaking process stepwise until you observe the value of the single molecule conductance
you have determined before. After you have established a single molecule junction, you will start
recording I-V curves. Therefore you should start at zero bias and increase the bias until you reach
maximum bias (Vmax) values of 0.4V. Afterwards, Vmax should be enlarged to max. 1.1V. Please
record at least 5 I-V curves for each Vmax value. After the I-V response was recorded, you should
try to further reduced the gap size in 0.2 nm / steps (according to the attenuation factor, this
corresponds to a displacement of the push rod by 40 m / step) and measure the I-V
characteristic again. By doing so, you will be able to investigate the influence of gap size on the
I-V characteristic and correspondingly on the tunneling mechanism. To reveal the underlying
tunnel mechanism, you should convert the data from I-V to ln( I / V 2 ) versus 1 / V presentation
and determine the transition voltage.
Figure 12. I-V curves of the single molecule junction with two different gap sizes. Curves A and B were
recorded at different gap sizes. The inset shows the linearly regime of curves A and B within
which indicates that both sets of curves correspond to a single molecule junction.
0.3 V,
References
[1] Xiang, D., Jeong, H., Lee, T. & Mayer, D. Mechanically controllable break junctions for
molecular electronics. Adv. Mater. 25, 4845–4867 (2013)
[2] Zhao, A., Zhang, H. & Hou, J. G. 5 Electron Transport in Single Molecules and Nanostructures
Microsystems and Nanotechnology (2012)
4. SUPPLEMENTARY INFORMATION
The bending beam is realized by using spring steel substrates with dimensions (1.2x5.5x0.2)cm3.
The substrate is insulated by spin coating a layer of polyimide (PI HD-2611, HD microsystem)
about 6µm.
Figure S. Top view SEM picture of the chip (a) and side view of the chips fabrication process which
consists of five main steps 1-5 (b)
This polymer is cured using a ramping program for the temperature: from room temperature up
to 350°C (4°C per min), baking steadily at 350°C for 30 min and again gradually cooling down.
Due to the shape of the substrate and the intrinsic roughness of the spring steel, it is needed to
smoothen the surface before placing the electrodes on top. For this purpose a second layer of PI
is spin coated, at higher speed, on the surface allowing us to have a more homogeneous surface.
Three layers of a positive-tone resist. Polymethylmetacylate (PMMA, ALLRESIST), of increasing
molecular weight, are spin coated in order to realize the nanoelectrodes.