Identifying diversity in nanoscale electrical break junctions

Article abstract of DOI:10.1021/ja103327f

The realization of molecular-scale electronic devices will require the development of novel strategies for controlling electrical properties of metal|molecule|metal junctions, down to the single molecule level. Here, we show that it is possible to exert c

Full text of DOI:10.1021/ja103327f

Published on Web 06/10/2010
Identifying Diversity in Nanoscale Electrical Break Junctions
Santiago Mart´ın,†,| Iain Grace,‡ Martin R. Bryce,*,§ Changsheng Wang,§
Rukkiat Jitchati,§,⊥ Andrei S. Batsanov,§ Simon J. Higgins,† Colin J. Lambert,*,‡ and
Richard J. Nichols*,†
Centre for Nanoscale Science and Department of Chemistry, UniVersity of LiVerpool,
LiVerpool L69 7ZD, U.K., Department of Physics, Lancaster UniVersity,
Lancaster LA1 4YB, U.K., Department of Chemistry and Centre for Molecular and Nanoscale
Electronics, Durham UniVersity, Durham DH1 3LE, U.K., Departamento de Qu´ımica
Orga´nica-Qu´ımica F´ısica, Facultad de Ciencias, UniVersidad de Zaragoza, 50009 Zaragoza, Spain,
and AdVanced Organic Materials and DeVices Laboratory, Department of Chemistry, Faculty of
Science, Ubon Ratchathani UniVersity, Warinchumrap, Ubon Ratchathani 34190, Thailand
Received April 20, 2010; E-mail: r.j.nichols@liverpool.ac.uk; m.r.bryce@durham.ac.uk;
c.lambert@lancaster.ac.uk
Abstract: The realization of molecular-scale electronic devices will require the development of novel
strategies for controlling electrical properties of metal|molecule|metal junctions, down to the single
molecule level. Here, we show that it is possible to exert chemical control over the formation of
metal|molecule...molecule|metal junctions in which the molecules interact by π-stacking. The tip of an
STM is used to form one contact, and the substrate the other; the molecules are conjugated
oligophenyleneethynylenes (OPEs). Supramolecular π-π interactions allow current to flow through
the junction, but not if bulky tert-butyl substituents on the phenyl rings prevent such interactions. For
the first time, we find evidence that π-stacked junctions can form even for OPEs with two thiol contacts.
Furthermore, we find evidence for metal|molecule|metal junctions involving oligophenyleneethynylene
monothiols, in which the second contact must be formed by the interaction of the π-electrons of the
terminal phenyl ring with the metal surface.
graphically defined nanogaps,3,14 or chemical synthesis of
nanoscale structures bridged by target molecules.15 The mol-
ecules investigated have increased in sophistication lately, to
encompass “longer” molecular bridges of organic or metal-
organic oligomers,6,16-21 supramolecular interactions such as
double-stranded DNA,22-26 noncovalent bonding of individual
Introduction
The feat of trapping single molecules within contact junctions
can be achieved through a number of techniques which utilize
scanning probe microscopes,1-6 break junctions,7-13 litho-
† Department of Chemistry, University of Liverpool.
‡ Department of Physics, Lancaster University.
§ Department of Chemistry, Durham University.
| Departamento de Quimica Organica-Quimica Fisica, Zaragoza University.
⊥ Department of Chemistry, Ubon Ratchathani University.
(1) Xu, B. Q.; Tao, N. J. J. Science 2003, 301, 1221–1223.
(2) Cui, X. D.; Primak, A.; Zarate, X.; Tomfohr, J.; Sankey, O. F.; Moore,
A. L.; Moore, T. A.; Gust, D.; Harris, G.; Lindsay, S. M. Science
2001, 294, 571–574.
(12) Huber, R.; Gonzalez, M. T.; Wu, S.; Langer, M.; Grunder, S.; Horhoiu,
V.; Mayor, M.; Bryce, M. R.; Wang, C. S.; Jitchati, R.; Schonenberger,
C.; Calame, M. J. Am. Chem. Soc. 2008, 130, 1080–1084.
(13) Kergueris, C.; Bourgoin, J. P.; Palacin, S.; Esteve, D.; Urbina, C.;
Magoga, M.; Joachim, C. Phys. ReV. B 1999, 59, 12505–12513.
(14) Osorio, E. A.; Bjornholm, T.; Lehn, J. M.; Ruben, M.; van der Zant,
H. S. J. J. Phys.: Condens. Matter 2008, 20, 374121.
(15) Champagne, A. R.; Pasupathy, A. N.; Ralph, D. C. Nano Lett. 2005,
5, 305–308.
(3) Kubatkin, S.; Danilov, A.; Hjort, M.; Cornil, J.; Bredas, J. L.; Stuhr-
Hansen, N.; Hedegard, P.; Bjornholm, T. Nature 2003, 425, 698–
701.
(4) Haiss, W.; van Zalinge, H.; Higgins, S. J.; Bethell, D.; Hobenreich,
H.; Schiffrin, D. J.; Nichols, R. J. J. Am. Chem. Soc. 2003, 125, 15294–
15295.
(16) Sedghi, G.; Sawada, K.; Esdaile, L. J.; Hoffmann, M.; Anderson, H. L.;
Bethell, D.; Haiss, W.; Higgins, S. J.; Nichols, R. J. J. Am. Chem.
Soc. 2008, 130, 8582–8583.
(5) Haiss, W.; Nichols, R. J.; van Zalinge, H.; Higgins, S. J.; Bethell, D.;
Schiffrin, D. J. Phys. Chem. Chem. Phys. 2004, 6, 4330–4337.
(6) Tao, N. J. Nat. Nanotechnol. 2006, 1, 173–181.
(7) Reed, M. A.; Zhou, C.; Muller, C. J.; Burgin, T. P.; Tour, J. M. Science
1997, 278, 252–254.
(17) Wang, C. S.; Batsanov, A. S.; Bryce, M. R.; Martin, S.; Nichols, R. J.;
Higgins, S. J.; Garcia-Suarez, V. M.; Lambert, C. J. J. Am. Chem.
Soc. 2009, 131, 15647–15654.
(18) Wu, S. M.; Gonzalez, M. T.; Huber, R.; Grunder, S.; Mayor, M.;
Schonenberger, C.; Calame, M. Nat. Nanotechnol. 2008, 3, 569–574.
(19) Yamada, R.; Kumazawa, H.; Noutoshi, T.; Tanaka, S.; Tada, H. Nano
Lett. 2008, 8, 1237–1240.
(8) Weber, H. B.; Reichert, J.; Weigend, F.; Ochs, R.; Beckmann, D.;
Mayor, M.; Ahlrichs, R.; von Lohneysen, H. Chem. Phys. 2002, 281,
113–125.
(20) Leary, E.; Van Zalinge, H.; Higgins, S. J.; Nichols, R. J.; de Biani,
F. F.; Leoni, P.; Marchetti, L.; Zanello, P. Phys. Chem. Chem. Phys.
2009, 11, 5198–5202.
(9) Smit, R. H. M.; Noat, Y.; Untiedt, C.; Lang, N. D.; van Hemert, M. C.;
van Ruitenbeek, J. M. Nature 2002, 419, 906–909.
(10) Gonzalez, M. T.; Wu, S. M.; Huber, R.; van der Molen, S. J.;
Schonenberger, C.; Calame, M. Nano Lett. 2006, 6, 2238–2242.
(11) Venkataraman, L.; Klare, J. E.; Nuckolls, C.; Hybertsen, M. S.;
Steigerwald, M. L. Nature 2006, 442, 904–907.
(21) Mayor, M.; von Hanisch, C.; Weber, H. B.; Reichert, J.; Beckmann,
D. Angew. Chem., Int. Ed. 2002, 41, 1183–1186.
(22) Xu, B. Q.; Zhang, P. M.; Li, X. L.; Tao, N. J. Nano Lett. 2004, 4,
1105–1108.
9
10.1021/ja103327f  2010 American Chemical Society
J. AM. CHEM. SOC. 2010, 132, 9157–9164 9157

A R T I C L E S
Mart´ın et al.
Chart 1. Compounds Featured in This Work (1, 2, 4, and 5) and
That of Haiss et al.28 (1) and Wu et al.18 (3 and 6) (Acetyl
Protecting Groups on the Sulfur Atoms Have Been Omitted for
Clarity)
for S-OPE-S molecules. The latter observation demonstrates
that care is needed when assigning conductance histogram
features. To aid in the attribution of conductance histogram
peaks to these different junction types, we have used measure-
ments of the statistical distribution of break-off distances (the
distance at which junctions are pulled apart during repeated STM
retraction experiments). Our measurements have also allowed
us to assign a previously uncharacterized conductance histogram
feature to a new type of single molecule junction configuration
in which the phenyl end of an OPE monothiol (S-OPE) is
directly contacted to gold. DFT and transport computations give
further insights into electronic transmission through these novel
structures.
Experimental and Theoretical Methods
Synthesis. Molecule 1 was available from an earlier study,28
and 2, 4, and 5 were synthesized and characterized as described in
the Supporting Information.
Conductance Measurements. An Agilent STM running Picos-
can 4.19 Software was used for all single molecule conductance
measurements which were performed at room temperature in air.
Molecular adlayers were formed on Au(111) substrates. These
substrates were formed from commercially available gold on glass
samples with a chromium adhesive layer (Arrandee) which were
flame annealed immediately prior to use. Flame annealing involved
heating the gold slide until it took on a slight orange hue. It was
then kept in this state for approximately 30 s-1 min, but care was
taken to ensure that the sample did not overheat. Molecular
adsorption was achieved by immersing the electrode for 30 s in 5
× 10-5 M THF solutions. The sample was then rinsed in ethanol
and gently blown dry in a stream of nitrogen gas. Gold STM tips
were fabricated from 0.25 mm Au wire (99.99%) which was freshly
electrochemically etched for each experiment at +2.4 V in a mixture
of ethanol (50%) and HCl (50%).
Electrical measurements were performed using an STM and the
I(s) method, described in detail previously.4,17,28,29 In brief, this
method involves the repeated formation and cleavage of molecular
bridges generally formed between gold contacts (a Au STM tip
and a Au substrate). In the I(s) technique the electrical conductance
of the junction is measured as the molecule is fully extended in
the gap between STM-tip and substrate as the tip is rapidly retracted.
Current steps are seen in the retraction process which are taken to
be characteristic of the cleavage of Au|molecule(s)|Au electrical
junctions.4 The height of these current steps is repeatedly measured,
and the resulting current step heights are then plotted in a
conductance histogram. In this work we have performed I(s) scans
from the position defined by the set-point values of tunneling current
(I0) and tunneling voltage (Ut) to a distance of +2.5 nm with a
scan rate of 25 nm s-1. Since the histogram peaks span a relatively
wide range, to optimize observation we have recorded data using
both low and high current preamplifiers (see Results and Discussion
and Supporting Information). The voltage to length conversion
factor of the STM was calibrated using images of Au(111)
monatomic steps (0.235 nm height) while the low and high current
preamplifiers were calibrated using resistors.
base pairs across metallic gaps,27 and the interaction of aromatic
molecules via π-stacking.18 Understanding and controlling
intermolecular interactions is an important step toward using
individual molecules or supramolecular assemblies as building
blocks for electronic devices. Wu et al. recently made the
landmark discovery of the formation of electrical junctions
through π-stacking of adjacent molecules in mechanically
formed break junctions.18 They reported conductance data for
pairs of oligophenyleneethynylene (OPE) molecular “rods”
interacting through π-π interactions to bridge nanoscale gaps
between gold electrical contacts. In OPE molecules with thiol
groups at both ends (S-OPE-S), the molecular bridges featured
Au-S chemisorption at both contacts with a single clear
conductance histogram peak attributed to single Au|S-OPE-S|Au
junctions. Using analogous molecules with a single thiol
terminus (S-OPE) led to replacement of this histogram peak by
another peak of considerably lower conductance.18 This was
attributed to a pair of S-OPE molecules, adsorbed at the
adjacent gold contacts and interacting by π-stacking, resulting
in a correspondingly longer bridge length and lower conduc-
tance. Furthermore, they showed that formation of these bridges
was efficient for OPEs with three phenylene groups but
markedly less defined for OPEs with two phenylene groups.18
This demonstrates that a sufficient footprint of overlap is
necessary for effective π-stacking.
In this manuscript, we address the formation of junctions
using molecules (Chart 1) in which the ability to π-stack is
chemically controlled through the use of bulky side groups that
prevent efficient intermolecular overlap. This approach aids in
the assignment of histogram features to π-stacking. We further
demonstrate the competitive formation of both π-stacked
junctions and “conventional” single Au|S-OPE-S|Au junctions
Conductance histograms were constructed from the current-
distance curves by adding all the current data from ∼450 I(s) curves
showing discernible plateaus. Typically, in a data collection series
for a given target molecule more than 5500 I(s) scans were recorded.
Since only a fraction of the I(s) scans resulted in Au|molecule|Au
electrical junction formation, curves not symptomatic of junction
formation, including those with an exponential decay characteristic
(23) van Zalinge, H.; Schiffrin, D. J.; Bates, A. D.; Haiss, W.; Ulstrup, J.;
Nichols, R. J. ChemPhysChem 2006, 7, 94–98.
(24) Cohen, H.; Nogues, C.; Naaman, R.; Porath, D. Proc. Natl. Acad.
Sci. U.S.A. 2005, 102, 11589–11593.
(25) Cohen, H.; Nogues, C.; Ullien, D.; Daube, S.; Naaman, R.; Porath,
D. Faraday Discuss. 2006, 131, 367–376.
(26) Ullien, D.; Cohen, H.; Porath, D. Nanotechnology 2007, 18, 424015.
(27) Chang, S. A.; He, J.; Lin, L. S.; Zhang, P. M.; Liang, F.; Young, M.;
Huang, S.; Lindsay, S. Nanotechnology 2009, 20, 185102.
(28) Haiss, W.; Wang, C. S.; Grace, I.; Batsanov, A. S.; Schiffrin, D. J.;
Higgins, S. J.; Bryce, M. R.; Lambert, C. J.; Nichols, R. J. Nat. Mater.
2006, 5, 995–1002.
(29) Haiss, W.; Martin, S.; Scullion, L. E.; Bouffier, L.; Higgins, S. J.;
Nichols, R. J. Phys. Chem. Chem. Phys. 2009, 11, 10831–10838.
9
9158 J. AM. CHEM. SOC. VOL. 132, NO. 26, 2010

Diversity in Nanoscale Electrical Break Junctions
A R T I C L E S
of tunneling, were rejected from histogram analysis. The probability
of forming molecular junctions is lower with the I(s) technique
than the in situ BJ technique, since the former avoids metallic
contact between the tip and substrate; a comparative study had
indicated that avoiding tip-surface metallic contact promotes
observation of the lower conductance group.30 Conductance
histograms are presented in units of the conductance quantum G0
) 2e2/h ) 77.4 µS and errors are evaluated by fitting histogram
peaks with Gaussian curves.
In this study, the determination of “break-off distance” is
important in the assignment of histogram peaks.17,28–30 “Break-
off distance” refers to the estimated separation at which the
molecular junction cleaves during an I(s) retraction experiment.
Cleavage is either at the Au-molecule contact or between the
contact surface atoms themselves (Au-Au bonds).31,32 By calibrat-
ing the tip-to-sample distance an assessment of the break-off
distance can be made which can help to distinguish π-stacked
junctions from other junction configurations. Full details are given
in the Supporting Information.
Theoretical Methods. The electronic and transport properties
of the molecules in Chart 1 contacted between gold leads were
calculated with the ab initio transport code SMEAGOL.33 This
method uses the Hamiltonian provided by the density functional
theory code SIESTA34 to obtain the zero bias electron transmission
coefficients. To begin with, the relaxed geometry of the molecule
was calculated; here we used a double-ꢀ polarized basis set, an
energy cutoff of 150 Ry to define the real space grid, and the local
density approximation35 to calculate exchange and correlation
energy. The molecular coordinates were relaxed until all forces were
<0.02 eV /Å. For the conductance calculations, the molecules were
extended between Au(111) leads; nine atoms per layer were used,
and six layers of gold were enough to converge the transmission.
To model the asymmetry of the STM junction, a flat surface was
used for the bottom electrode contact and a pyramid of gold atoms
for the top electrode (e.g., Figure 3), with the terminal thiol binding
to the tip atom. For the junctions containing two molecules, we
calculated the optimal “stacking” geometry by minimizing the
ground state energy by varying the overlap length X and the
separation distance D (Figure 5; more details of the theoretical
approach can be found in the Supporting Information).
Figure 1. Conductance histograms for compounds 4, 1, 5, 2 recorded at
low set-point current (5 nA, left panel), using a low current amplifier, and
high set-point current (60 nA, right panel), using a high current amplifier,
respectively. Conductance data are presented in units of the conductance
quantum G0 ) 2e2/h ) 77.4 µS. Ut ) 0.6 V.
In this study, we analyze conductance data for compounds
1, 2, 4, and 5. The molecules have been chosen to aid the
assignment of multiple peaks occurring in conductance histo-
grams to differing junction configurations. The monothiolated
derivatives (4 and 5) are chosen to exclude the formation of
Au|S-OPE-S|Au single molecule junctions, while the sterically
hindered derivatives (2 and 5) are chosen to impede π-stacking.
This, together with statistical analysis of the distance at which
junctions cleave, aids in the assignment of the measured
conductance histograms. In what follows, we use the I(s) method
for conductance determination.4,17,28,29 In this method, a Au
STM tip is brought close to a Au(111) surface controlled by
the set-point current I0, on which is adsorbed a low (submono-
layer) coverage of the target molecules. The feedback loop is
switched off and the tip is withdrawn while the tunneling current
is measured. The current-vertical distance I(s) curves are then
analyzed statistically via histogram plots to determine the
electrical properties of the molecular junctions. This method
has the advantage that it is possible, by careful calibration, to
measure simultaneously both the tunneling current I and the
tip height at which the junction breaks down (either by Au-S
or by Au-Au bond cleavage) which we label sbreak-off (see the
Method section and the Supporting Information for full details).
Figure 1 shows the conductance histograms for compounds
1, 2, 4, and 5. The left panel shows data recorded with the low
current amplifier and low set-point current (I0) values (I0 ) 5
nA and Ut ) 0.6 V). The right panel shows data recorded with
a high current (lower sensitivity) amplifier at higher set-point
currents (I0 ) 60 nA and Ut ) 0.6 V). The choice of low or
Results and Discussion
Molecules employed in, or of relevance to, this study are
shown in Chart 1. Compound 1 is the “reference” compound
for our current investigation. It featured previously in a
publication by Haiss et al.28 and was characterized as having a
conductance of 2 nS when untilted in a contact junction with
gold leads. Compound 2 has bulky tert-butyl substituents,
designed to prevent π-stacking. Compound 3 is analogous to 1
and was studied by Wu et al.18 Compounds 4-6 have a single
thiol contact. In the seminal study by Wu et al.,18 it was found
that molecules with two contacting groups such as 3 formed
single molecule junctions with chemisorption at both gold
contacts, whereas monothiols such as 6 formed junctions in
which the metal contacts were bridged by two π-stacked
molecules.
(30) Haiss, W.; Martin, S.; Leary, E.; van Zalinge, H.; Higgins, S. J.;
Bouffier, L.; Nichols, R. J. J. Phys. Chem. C 2009, 113, 5823–5833.
(31) Xu, B. Q.; Xiao, X. Y.; Tao, N. J. J. Am. Chem. Soc. 2003, 125,
16164–16165.
(32) Rubio-Bollinger, G.; Bahn, S. R.; Agrait, N.; Jacobsen, K. W.; Vieira,
S. Phys. ReV. Lett. 2001, 87, 026101.
(33) Soler, J. M.; Artacho, E.; Gale, J. D.; Garcia, A.; Junquera, J.; Ordejon,
P.; Sanchez-Portal, D. J. Phys.: Condens. Matter 2002, 14, 2745–
2779.
(34) Rocha, A. R.; Garcia-Suarez, V. M.; Bailey, S.; Lambert, C.; Ferrer,
J.; Sanvito, S. Phys. ReV. B 2006, 73, 085414.
(35) Perdew, J. P.; Zunger, A. Phys. ReV. B 1981, 23, 5048–5079.
9
J. AM. CHEM. SOC. VOL. 132, NO. 26, 2010 9159

A R T I C L E S
Mart´ın et al.
high current amplifiers enables the observation of either lower
or higher conductance peaks. In particular, the choice of low
current amplifier and low set-point currents effectively restricts
observation to events (current jumps in I(s) curves) at the low
conductance range (<4 × 10-5 G0). This choice of low current
amplifier and low set-point current could be viewed, by way of
an analogy, as low pass filtering of the data, since it restricts
observations to low-current events at high sensitivity. The high
current amplifier used at higher set-point currents, on the other
hand, can be viewed as high pass removal of the low current
events that are “lost” in the background appearing near the
ordinate. By using this combination of low and high current
amplifiers a greater range of conductance histogram peaks can
be measured with high sensitivity.
Guide lines are provided in Figure 1 to highlight common
features between histograms. Starting with the low current
amplifier data, the compounds without the sterically blocking
tert-butyl substituents (1 and 4) feature a peak labeled π-1 which
is attributed to the formation of Au|molecule...molecule|Au
junctions in which the molecules π-stack (vide infra). The
histogram of 4 shows a shoulder on this peak (labeled π-2;
although π-2 is not clearly apparent for 1 where its appearance
may be masked by an adjacent higher conductance and more
intense peak marked A). Note that these π -peaks do not appear
for compounds 2 and 5, which have the sterically blocking tert-
butyl substituents. This leads to the conclusion that the π-peaks
do indeed arise from π-stacking, in agreement with the previous
recognition of π-stacking in OPE derivatives by Wu et al.18
However, a glance at our overall conductance histogram data
shows marked differences as well, in that we observe for certain
derivatives the resolution of multiple distinctive conductance
peaks.
In the low current amplifier data for 1 (Figure 1, left panel),
a peak marked A at higher conductance values than the π-peak
is seen. This peak is not apparent for compounds 4 and 5, the
monothiol analogues, but it is also present for the other dithiol
analogue 2. This leads us to attribute the A peak to Au|S-
OPE-S|Au single molecule junctions with chemisorbed Au-S
contacts at both ends of the molecular bridge. We show later
in the text that the break-off distance for these A features is
consistent with the expected length of a Au|S-OPE-S|Au
junction, while the π-peaks feature considerably longer break-
off lengths. Histograms for 5 (monothiol and tert-butyl groups)
feature neither A nor π-peaks, while histograms for 2 feature
only A peaks.
We now turn to the high current amplifier data (Figure 1,
right panel). Two histogram peaks are observed in this range,
and these are marked with the dashed guidelines as “T” and
“B”, respectively. Peak B appears only for the dithiolated
analogues (1 and 2) and is assigned, like the A peaks, to
Au|S-OPE-S|Au single molecule junctions, albeit with a higher
conductance than the A peaks. Molecules 4 and 5 will not be
able to form such Au|S-OPE-S|Au junctions, since they are
monothiols, and this is consistent with the absence of both A
and B peaks for these molecules. It has been previously shown
that Au|S-backbone-S|Au junctions exhibit different conduc-
tance groups which have been related to different adsorption
geometries/conformations of the sulfur chemisorbed at the gold
contacts.5,30,36-38 It appears likely that a similar phenomenon
Figure 2. Break-off distance histograms for 4 (top) and 1 (bottom), for
the T and π-1 peaks (compound 4), and for the A, B and π-1 peaks
(compound 1).
is occurring here for the OPE derivatives. However, what is
surprising is the appearance of an additional peak labeled “T”
for the two monothiol compounds (4 and 5). Break-off distance
data (vide infra: Figure 2) show this peak to be consistent with
junctions of a single molecule length (rather than π-stacked
junctions). However, since both 4 and 5 are monothiols, we
infer that this peak arises from the end phenyl group making
direct physical contact with a gold lead and forming one end
of the electrical junction. Such junctions have not been identified
clearly at the single molecule level before, but contacts of this
nature could be formed in larger area junctions in which
monothiols are sandwiched between a pair of closely spaced
metal contacts.39-43 The direct interaction of benzene with
platinum contacts in an STM configuration has also been
reported by Kiguchi et al., with the formation of Pt|benzene|Pt
junctions (benzene sandwiched “face-on” between two Pt leads)
with high conductance.44
In the discussion above, break-off distance determinations
have been referred to as an aid to peak assignment. This
important information is not available from the MCBJ experi-
ments of Wu et al.18 and is an added advantage of using the
I(s) technique to study these systems. These are now discussed
in more detail (see Supporting Information to obtain more
information about the determination of sbreak-off histograms).
Figure 2 shows break-off distance histograms for 4 that feature
the π and T group conductance peaks. The π-peak is character-
ized by a break-off distance of 2.9 ( 0.3 nm, which is
considerably longer than a fully extended junction containing
a single 4 molecule (2.19 nm), while the T-peak has a much
(38) Zhou, J. F.; Chen, F.; Xu, B. Q. J. Am. Chem. Soc. 2009, 131, 10439–
10446.
(39) Cui, X. D.; Zarate, X.; Tomfohr, J.; Sankey, O. F.; Primak, A.; Moore,
A. L.; Moore, T. A.; Gust, D.; Harris, G.; Lindsay, S. M. Nanotech-
nology 2002, 13, 5–14.
(40) Beebe, J. M.; Kim, B.; Frisbie, C. D.; Kushmerick, J. G. ACS Nano
2008, 2, 827–832.
(41) Wold, D. J.; Frisbie, C. D. J. Am. Chem. Soc. 2001, 123, 5549–5556.
(42) Ishida, T.; Mizutani, W.; Akiba, U.; Umemura, K.; Inoue, A.; Choi,
N.; Fujihira, M.; Tokumoto, H. J. Phys. Chem. B 1999, 103, 1686–
1690.
(36) Li, C.; Pobelov, I.; Wandlowski, T.; Bagrets, A.; Arnold, A.; Evers,
F. J. Am. Chem. Soc. 2008, 130, 318–326.
(43) Kushmerick, J. G.; Whitaker, C. M.; Pollack, S. K.; Schull, T. L.;
Shashidhar, R. Nanotechnology 2004, 15, S489-S493.
(44) Kiguchi, M.; Murakoshi, K. Thin Solid Films 2009, 518, 466–469.
(37) Li, X. L.; He, J.; Hihath, J.; Xu, B. Q.; Lindsay, S. M.; Tao, N. J.
J. Am. Chem. Soc. 2006, 128, 2135–2141.
9
9160 J. AM. CHEM. SOC. VOL. 132, NO. 26, 2010

Diversity in Nanoscale Electrical Break Junctions
A R T I C L E S
Table 1. Tabulated Experimental and Theoretical Data for Compounds 4, 1, 2, and 5a
Column 1
Column 2
Column 3
Column 4
Column 5
Column 6
Column 7
Column 8
Molecular
length
/nm
(theoretical
gold...gold
separation)
Adjusted
Theoretical
Conductance/
Experimental
Conductance
Unadjusted
Theoretical
Conductance/
10-5 G0
Adjusted
Theoretical
Conductance/
10-5 G0
Expt
Conductance/
10-5 G0
Sbreak off/
nm
Molecule
Peaks
4
π-1
π-2
T
π-1
A
B
A
B
0.61 ( 0.11
1.19 ( 0.20
4.86 ( 1.63
0.57 ( 0.14
2.02 ( 0.18
10.68 ( 2.25
1.98 ( 0.42
10.56 ( 2.82
s
2.92 ( 0.26
2.65 ( 0.18
2.10 ( 0.12
2.82 ( 0.10
2.26 ( 0.09
2.02 ( 0.14
2.18 ( 0.14
1.91 ( 0.16
s
2.84
s
3.31
s
45.7
3.96
18
119.0
20.7
126.0
0.003
66.4
0.37
s
0.6
s
2.19
2.82
2.51
2.27
2.51
2.26
(3.62)
2.20
13.5
0.41
2.41
6.12
2.3
6.45
0.005
32.4
2.8
0.7
1.2
0.6
1.2
0.6
s
1
2
5
π-1
T
3.79 ( 0.12
1.94 ( 0.18
8.5
a Column 2: conductance peak label (see text for a description). Column 3: conductance measured from experimental histograms. Column 4:
experimentally measured break-off distance from calibrated break-off distance measurements. Column 5: the theoretical Au...Au contact separation for
the molecule(s) extended in the junction for this configuration. Column 6: The unadjusted theoretically computed conductance. Column 7: the adjusted
theoretically computed conductance (see text). Column 8: The ratio of column 7 (adjusted theoretical values) to column 3 (experimental values).
Figure 3. Models of the considered junction configurations for the A peak of compound 1 (a), the T peak of compound 4 (b), and the π-1 peak of
compound 4 (c).
shorter break-off distance of 2.1 ( 0.1 nm, consistent with a
single molecule length junction. Errors are based on the full
width at half height of Gaussian curves fitted to the histogram
distribution of break-off distances. Break-off distance histograms
for 1 are also shown in Figure 2 (bottom panel). Similar to 4,
the π-peak is characterized by a “long” break-off distance of
2.8 ( 0.1 nm. The A and B peaks, however, occur at
considerably shorter break-off distances of 2.3 ( 0.1 and 2.0
( 0.1 nm, respectively, consistent with single molecule junc-
tions. The difference in break-off distances between the A and
B peaks may arise from different surface binding sites and thiol
headgroup(s) sitting in step or defect sites to give a closer
Au...Au effective contact spacing for the B group. Such a
phenomenon has been noted before and discussed in detail, for
both thiol and pyridyl contacting groups.17,30 Table 1 sum-
marizes the experimental data and peak assignments.
To test the experimental assignments and to provide further
insight into junction configurations involving π-stacking or
direct contact of the phenyl group and a gold contact, theoretical
calculations on these systems have been performed using a
combination of DFT and nonequilibrium Green’s functions
formalisms. Figure 3 shows illustrative configurations considered
for 1 (Figure 3a) and 4 (Figure 3b) between gold contacts. In
addition to the configuration shown in Figure 3a, we have also
considered configurations in which the thiol binds at one contact
to a higher coordination “step”-type site (see Supporting
Information). Figure 3b shows 4 binding to the upper contact
through a thiol and to the lower contact through a phenyl ring
in a face-on configuration to a gold adatom, representative of a
high coordination site. We have also considered phenyl group
coordination to the surface end-on through the terminal hydrogen
(Figure 6, right). Figure 3c shows an optimized aromatic
stacking configuration for 4.
Table 1 presents a summary of the theoretically computed
and the experimentally determined conductance values. Column
6 shows the unadjusted DFT calculations, while the next column
along shows computations which have been adjusted using a
self-energy and a screening energy correction, following
the procedure of Quek et al.45 It should be noted that, even in
the absence of these corrections, the calculations reproduce the
experimentally observed trend for the different conductance
peaks, although in some instances the theoretical values are
(45) Quek, S. Y.; Kamenetska, M.; Steigerwald, M. L.; Choi, H. J.; Louie,
S. G.; Hybertsen, M. S.; Neaton, J. B.; Venkataraman, L. Nat.
Nanotechnol. 2009, 4, 230–234.
9
J. AM. CHEM. SOC. VOL. 132, NO. 26, 2010 9161

A R T I C L E S
Mart´ın et al.
Figure 4. Transport calculations for the π-stacking configurations of compounds 5 (blue curve and model on left) and 4 (black curve and model on right).
Notice how the bulky tert-butyl groups impede the stacking of 5 (left) and consequently produce a very marked decrease in the transmission at E ) EF. On
the other hand, there is good stacking overlap for compound 4 (right).
Figure 5. Top: the optimized stacking configuration for 4. The energy minimized values for X and D are 1.7 and 0.34 nm, respectively. Bottom: variation
of the junction conductance as a function of parameter X (left) and D (right).
significantly greater than the experimental values. The “semi-
empirical” correction40 improves the quantitative fit. In the
following, we present the ab initio generated transmission curves
that give rise to the “unadjusted” conductance values (Table
1), and finally mention the corrected conductance for which the
quantitative fit is improved.
Figure S7 in the Supporting Information shows transmission
curves for 1 binding through the thiol end groups to single gold
surface atoms at both contacts. The resonance closest to the
Fermi level corresponds to a HOMO feature, indicating that
HOMO mediated hole transport is likely to dominate the
transport behavior. The ab initio conductance at the Fermi level
is (18 × 10-5)G0 which is ∼1 order of magnitude higher than
the experimental value (2 × 10-5)G0. Figure 4 presents
transmission curves for the π-stacking geometry of 4 (black
curve) and 5 (blue curve). The transmission at the Fermi level
is more than 3 orders of magnitude lower for compound 5 which
has the tert-butyl substituents. Figure 5 shows the optimized
(energy minimized) stacking geometries used for 4, which is
characterized by two parameters; “X” represents the total
horizontal distance over which the molecules overlap, while “D”
represents their optimum vertical separation. For 4, we find X
) 1.7 nm and D ) 0.34 nm. Note that, in the optimized
geometry, π-stacking does not involve face-to-face stacking of
the phenyl rings; rather, it involves a staggered configuration
of the rings, with the phenyl ring of one molecule facing an
acetylene linker in the adjacent molecule. In the case of 5 (see
Supporting Information), aromatic stacking involves only one
phenyl ring and a single acetylene linker on either molecule
(X ) 0.8 nm) and the two molecules are further apart (D )
0.40 nm). In the X-ray crystal structure of 5, D ) 0.39 nm
(Figure S4). The marked differences between the D and X
9
9162 J. AM. CHEM. SOC. VOL. 132, NO. 26, 2010

Diversity in Nanoscale Electrical Break Junctions
A R T I C L E S
Figure 6. Transport calculations for the proposed “T” configuration of compound 4 with the phenyl group coordinating with the lower contact face-on
through the end phenyl group to a gold adatom (blue curve and model on left) or end-on to a flat surface (black curve and model on right). Notice how the
favored face-on configuration produces a much higher transmission at E ) EF. We refer to the face-on contact of the phenyl group to the Au adatom as a
“step” position since it is our model for a step or similar defect site on the surface. Other coordination possibilities are considered in the Supporting Information.
parameters for 4 (Figure 5, top) and 5 (see Supporting
Information, Figure S6) are the main cause of the large
differences in the computed conductance values.
adsorption could instead be at a step edge or to a pyramidal
tip; we propose that it is sites of this kind, with face-on
interaction of the phenyl ring and a Au contact, which give rise
to the T peak in the high conductance histogram data. The
conductance values summarized in Table 1 show both experi-
mentally and theoretically that the T group is of higher
conductance than the π-stacking configurations. Calculations
of adsorption energies for candidate phenyl-gold contact
geometries support the likelihood of the step-contact geometry
(for the H end-on to gold configuration we find an adsorption
energy value of 0.302 eV for a distance of 2.0 Å above a gold
atom). By contrast, for a “step contact” (phenyl ring face-on to
the gold adatom contact) a value of 0.86 eV is obtained, where
the phenyl ring is 2.4 Å away from the gold adatom. By
comparison the gold-thiol adsorption energy is larger with a
value of 1.67 eV, which is in reasonable agreement with the
expected binding strength.48 Although there are reported issues
concerning the reliability of DFT computations for the adsorp-
tion energy of benzene on gold,49 the value we estimate for the
phenyl-gold adatom adsorption (0.86 eV) is sufficiently high
compared to kT (0.026 eV) to merit consideration for our room
temperature measurements. We should also add that increasing
the microscopic structure of the contact (e.g., a pyramid of Au
atoms rather than the single gold atom in Figure 6, left) may
for certain configurations enhance the adsorption “footprint” of
the phenyl group and its corresponding adsorption energy. There
are other indications in the literature that phenyl groups
(benzene) can interact strongly enough with gold to exhibit room
temperature surface adsorption, notably the temperature pro-
grammed desorption recorded by Nakazawa and Somorjai of
benzene on gold.50
Figure 5 shows how the conductance (i.e., the transmission
coefficient T(EF)) of 4 varies as the parameters X and D are
changed. The conductance decreases monotonically as D is
increased (Figure 5, lower right). The conductance also increases
as X is increased from 0 to 2 nm (Figure 5, lower left), although
there are peaks and troughs indicating that the theoretically
computed junction conductance is very sensitive to variations
in X at the sub-angstrom level. Lin et al.46 have also theoretically
investigated π-stacked OPE junctions with semi-infinite gold
leads. They concluded that the experimental transport charac-
teristics of the π-stacked OPE junctions18 could be accounted
for with D ) 0.30 nm,46 close to our computed value of 0.34
nm for 4. However, their preferred geometry involved a cofacial
stacking arrangement.46 This seems rather unlikely given what
is generally known about π-stacking.47
Both our experimental and theoretical data agree with the
observation of Wu et al.18 that aromatic coupling is strong
enough to allow junction formation and that these junctions are
sufficiently electronically transmissive to allow observation in
MCBJ or STM configurations. However, our study also shows
that there is a competitive formation of differing junction types
which can be seen and characterized in the STM based I(s)
method, in particular the “T” peak seen for 4 and 5 which has
a short sbreak-off (Figure 2) which we thus assign to the end phenyl
group contacting with gold to form a molecular bridge which
breaks off at the distance consistent with a single molecule
length. We have performed DFT computations to further test
this experimental assignment. Figure 6 presents transmission
curves for 4 with the phenyl group in a “step” contact position
or with the phenyl H end-on to the gold contact (see figure for
illustration). The former geometry gives a transmission at the
Fermi level >1 order of magnitude higher than the latter. This
demonstrates that phenyl group contacts in high coordination
sites can yield highly transmissive systems. These sites are
exemplified by the adatom geometry in Figure 6 (left), but
Conductance values after adjustment using the procedure of
Quek et al.45,51 are presented in Table 1. This shifts the HOMO
resonance further from the Fermi level and reduces the
conductances, giving better agreement with the measured values.
(48) Ulman, A. Chem. ReV. 1996, 96, 1533–1554.
(49) Bilic, A.; Reimers, J. R.; Hush, N. S.; Hoft, R. C.; Ford, M. J. J. Chem.
Theory Comput. 2006, 2, 1093–1105.
(46) Lin, L. L.; Leng, J. C.; Song, X. N.; Li, Z. L.; Luo, Y.; Wang, C. K.
J. Phys. Chem. C 2009, 113, 14474–14477.
(50) Nakazawa, M.; Somorjai, G. A. Appl. Surf. Sci. 1993, 68, 517–537.
(51) Quek, S. Y.; Venkataraman, L.; Choi, H. J.; Loule, S. G.; Hybertsen,
M. S.; Neaton, J. B. Nano Lett. 2007, 7, 3477–3482.
(47) Hunter, C. A.; Sanders, J. K. M. J. Am. Chem. Soc. 1990, 112, 5525–
5534.
9
J. AM. CHEM. SOC. VOL. 132, NO. 26, 2010 9163

A R T I C L E S
Mart´ın et al.
Acknowledgment. This work was supported by EPSRC under
Grants EP/C00678X/1 (Mechanisms of Single Molecule Conduc-
tance) and GR/S84064/01 (Controlled Electron Transport Through
Single Molecules), QinetiQ and the British Department of Trade
and Industry, Royal Society, Northwest Regional Development
Agency, the EC FP7 ITN “FUNMOLS” Project No. 212942, and
the Royal Thai Government (scholarship to R.J.). S.M. acknowl-
edges his Juan de la Cierva position from the Ministerio de Ciencia
e Innovacio´n (Spain). We thank the NWGrid for computing
resources.
The slightly poorer agreement for the phenyl contacts is probably
due to greater uncertainty in the precise geometry of this contact.
Conclusions
We have shown that the ability of π-conjugated molecular
wires to form noncovalent aromatically coupled electrical
junctions can be controlled through the use of chemical
modifications. These chemical modifications can in turn be used
to distinguish the nature of the junctions formed and to identify
differing conductance peaks in histograms, particularly when
used together with measurements of the junction extension at
break-off and supporting DFT computations. The use of
supramolecular interactions, such as aromatic coupling, is a key
long-term goal of molecular electronics. Molecular conductance
measurements using STM break junction techniques can help
to identify such interactions, although they also identify the
diverse range of different junction configurations with widely
different conductance values which can occur for any given
molecular system.
Supporting Information Available: Synthesis and character-
ization of compounds 2, 4, and 5; crystal packing and molecular
stacks in I and 5; conductance traces recorded by the I(s)
technique; details of calculations of optimum stacking geom-
etries for 2 and 4, influence of contact geometries for 1, influence
of side groups for 1 and 2, and a comparison of the uncorrected
and corrected transport calculations. This material is available
free of charge via the Internet at http://pubs.acs.org.
JA103327F
9
9164 J. AM. CHEM. SOC. VOL. 132, NO. 26, 2010