Importance of direct metal-π coupling in electronic transport through conjugated single-molecule junctions

Article abstract of DOI:10.1021/ja308626m

We study the effects of molecular structure on the electronic transport and mechanical stability of single-molecule junctions formed with Au point contacts. Two types of linear conjugated molecular wires are compared: those functionalized with methylsulfi

Full text of DOI:10.1021/ja308626m

Article
pubs.acs.org/JACS
Importance of Direct Metal−π Coupling in Electronic Transport
Through Conjugated Single-Molecule Junctions
Jeffrey S. Meisner,†,⊥ Seokhoon Ahn,†,⊥ Sriharsha V. Aradhya,‡ Markrete Krikorian,†,∥
Radha Parameswaran,§ Michael Steigerwald,*,† Latha Venkataraman,*,‡ and Colin Nuckolls*,†
†Department of Chemistry, Columbia University, New York, New York 10027, United States
‡Department of Applied Physics and Applied Mathematics, Columbia University, New York, New York 10027, United States
§Departments of Chemistry and Physics, Barnard College, New York, New York 10027, United States
S
* Supporting Information
ABSTRACT: We study the effects of molecular structure on the electronic
transport and mechanical stability of single-molecule junctions formed with Au
point contacts. Two types of linear conjugated molecular wires are compared:
those functionalized with methylsulfide or amine aurophilic groups at (1) both or
(2) only one of its phenyl termini. Using scanning tunneling and atomic force
microscope break-junction techniques, the conductance of mono- and
difunctionalized molecular wires and its dependence on junction elongation
and rupture forces were studied. Charge transport through monofunctionalized
wires is observed when the molecular bridge is coupled through a S−Au donor−
acceptor bond on one end and a relatively weak Au−π interaction on the other
end. For monofunctionalized molecular wires, junctions can be mechanically
stabilized by installing a second aurophilic group at the meta position that, however, does not in itself contribute to a new
conduction pathway. These results reveal the important interplay between electronic coupling through metal−π interactions and
quantum mechanical effects introduced by chemical substitution on the conjugated system. This study affords a strategy to
deterministically tune the electrical and mechanical properties through molecular wires.
linker, is removed leaving only a single para linker, the
conductance decreases even further. We show that the
conductance of our molecular wires that have at least one
para linker decays exponentially with increasing oligomeric
length and that they have step lengths corresponding to their
molecular length. Both of these results indicate that we are
probing the conductance of single-molecule junctions, as
opposed to junctions formed by overlapping or interdigitated
molecular dyads. That is, these measurements allow us to
conclude that monofunctionalized stilbene molecules do not
readily form junctions where molecules conduct via inter-
molecular carrier transfer (i.e., π−π-stacking interactions).5
INTRODUCTION
■
This study describes the mechanism of conduction through
asymmetric molecular junctions containing conjugated mole-
cules having only one electrode-binding “linker” group.1 Linker
groups are aurophilic functional groups that bind the molecule
between Au electrodes, such as thiols (−SH), primary amines
(−NH2), and methylsulfides (−SMe).2 Typical molecules
employed in single-molecule electronics are conjugated or
short aliphatic molecules that are functionalized at each end
with linker groups. Here, conjugated olefins of varying lengths
are end-functionalized with methylsulfide and amine linkers at
one or both terminal phenyl rings. To explore quantum
mechanical effects3 we vary the position of these linkers
between the meta or para positions. The conductance and
rupture forces of single-molecule junctions formed from these
molecules are measured using the break-junction (BJ)
technique with a scanning tunneling microscope (STM)2a
and an atomic force microscope (AFM).4 We find that for
measurable conductivity to occur at least one of the rings must
have a linker para to the olefin providing strong electronic
coupling to the electrode. We measure both the highest
conductivity and the narrowest distribution of conductance for
olefins with two para linkers. When one of these para linkers is
replaced with a meta linker, the conductance decreases by
almost an order of magnitude, due to a reduction in the Au−
molecule−Au coupling. If this mechanical contact, the meta
EXPERIMENTAL METHODS
■
Syntheses. For this study, we synthesized three different
vinylogous series of methylsulfide-functionalized trans-α,ω-diphenyl-
oligoenes as well as 3-(methylthio)stilbene. Each series ranges in
length from the stilbene (n = 1) to the triene (n = 3) and is displayed
in Figure 1A and 1B. A convenient shorthand is used to name these
compounds (PPn, PMn, Pn, and M1), which includes the linker
substitution (P = para, M = meta) and the length of the oligomer (n).
As examples, PP2 denotes para−para′-dithiomethyl-diphenylbuta-
diene and PM3 denotes para-meta′-dithiomethyl-diphenylhexatriene.
Both difunctionalized (PPn and PMn) and monofunctionalized (Pn
Received: August 30, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja308626m | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Molecular conductance step lengths9 and the slopes of the
conductance features (βS) in the 2D histograms were determined by
fitting the average conductance profile; both procedures are outlined
in detail in the Supporting Information.
Simultaneous conductance and force measurements were obtained
using a custom-built conducting AFM setup.4a Single-molecule
junctions were formed between a gold-coated commercial AFM
cantilever (NanoAndMore, Inc.) and a gold-on-mica substrate.
Conductance is measured across the tip/sample junction at a bias of
75 mV. The force is measured simultaneously by monitoring the
deflection of a laser focused on the back of the cantilever. AFM-BJ
measurements are carried out on PP1, PM1, and P1. In each case, 2D
force-displacement histograms are constructed from measurements on
over 8000 junctions.
Theoretical Methods. Density functional theory (DFT) elec-
tronic structure calculations examined shape and energy differences
among the most relevant molecular orbitals (MOs) for molecular
conductance. All calculations were preformed using Jaguar10 with the
B3LYP hybrid functional and the 6-31G** basis sets. The molecular
geometries were fully optimized. The final geometries, total energies,
and MO energies for each molecule in this study are given in the
Supporting Information.
RESULTS AND DISCUSSION
■
We begin by comparing the conductance and rupture forces of
molecular junctions of four stilbene derivatives having either
one or two methylsulfide linkers in either the meta or para
positions (PP1, PM1, P1, and M1). Their 1D logarithmically
binned conductance histograms are obtained from STM-BJ
measurements and are compared in Figure 2A. The histogram
for PP1 shows a clear molecular conductance peak, indicating
that reproducible single-molecule junctions are formed
throughout thousands of measurements. The sharpest peak at
10−3 G0 is characteristic of a conjugated stilbene having two
para linkers, as has been shown before.3d,11 A characteristic
conductance signature (peak in the histogram) appears for P1
even though it contains only one linker. The peak conductance
for P1 is almost 2 orders of magnitude lower than that for PP1;
therefore, electronic coupling across the P1 junction is weaker
than with PP1. From the width of the peak, we conclude that
the junction conductance varies significantly more than in the
case of PP1, which has two linkers. We quantify the peak width
for all junctions in Table 1 using the half-width at half-
maximum. We next compare P1 with M1, which also has only
one linker, but now at the meta-position, we see no peak in this
histogram indicating that junctions with a conductance above
our instrument noise (∼10−7 G0) are not formed with M1. This
trend is expected since meta linkers do not provide strong
electronic coupling into the π-system.3a Nonetheless, they do
provide a mechanical link as we have previously shown.3d
Indeed, we see with PM1 that the additional mechanical
stability provided by the meta linker yields both a higher
conductance and a narrower distribution compared with P1.
Conductance values for each compound are given in Table 1.
We confirm the mechanical enhancement provided by the
meta linker by measuring force in AFM-BJs of PP1, PM1, and
P1. The rupture forces are measured by pulling the junction
apart until it breaks and (to a first-order approximation)
establish an upper limit on the strength of the weakest Au−
molecule interaction. Stilbenes with two methylsulfide linkers
(regardless of their position on the phenyl rings) are stabilized
in the junction and give rupture forces of 0.5 nN, as is the case
with PP13d and PM1 (Figure 2C). In contrast, we found that
the rupture force for P1 is smaller than 0.3 nN (twice the
instrumental noise). Our rupture force results for each stilbene
Figure 1. Chemical structures of trans-stilbene and all-trans-oligoene
molecular wires: (A) difunctionalized series, having only para-
positioned linker groups (PPn) or a mixture of para and meta linkers
(PMn); (B) monofunctionalized wires Pn and M1 contain either one
meta or one para linker. (C) Stilbene derivatives end-functionalized
with primary and tertiary amines.
and M1) molecules were synthesized through Wittig6 and Horner−
Wadsworth−Emmons reactions (see the Supporting Information for
synthetic details and characterization).7
A set of three additional stilbene derivatives bearing amine linkers
were synthesized to deconvolute the electronic and mechanical
contributions in molecular junctions: (E)-3,4′-diaminostilbene
(PM1A) and (E)-4-amino-stilbene (P1A) as analogous compounds
to the conducting (E)-(methylthio)stilbenes above, as well as (E)-3-
dimethylamino-4′-aminostilbene (PM1TA), where A = amine and TA
= tertiary amine.
STM and AFM-BJ Measurements. STM-BJ measurements2a,b
were performed in dilute solutions (1
0.1 mM in 1,2,4-
trichlorobenzene) of molecular wires using a gold-on-mica substrate
and a gold STM tip (cut Au wire, 0.25 mm diameter, 99.998%, Alfa
Aesar). Gold atomic point contacts were repeatedly formed and
broken in the solution of molecules under a 500 mV voltage applied to
the junction with a 100 kΩ resistor in series. As the point contacts are
broken, one or a few molecules may bind to bridge the gap between
the broken Au contact, thereby forming molecular junctions. The
electrodes are then pulled farther apart until the junction is broken.
Conductance (current/voltage) is measured as a function of piezo
displacement yielding individual conductance traces. In doing so, our
STM-BJ method does not take consecutive measurements on a single
junction. Instead, at the end of each measurement the junction is
destroyed. Then, before forming a new junction, the tip and substrate
are smashed together and pulled apart, forming fresh electrodes and
thus a new junction. For each molecule, over 5000 traces are collected
and analyzed by creating one-dimensional (1D) conductance and two-
dimensional (2D) conductance-displacement histograms2d,8 that reveal
statistically relevant information on junction conductance, as well its
evolution under junction elongation.
B
dx.doi.org/10.1021/ja308626m | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Table 1. Tabulation of Conductance Parameters from STM-
BJ Measurements
a
Molecular length is taken from DFT optimized structures.
Molecules do not form conductive junctions. Length refers to the
b
c
through-space distance between terminal linker group heteroatoms.
d
Length refers to the through-space distance from the para-linker
e
group heteroatom to the most distant carbon atom. Conductance
peak widths are determined using the half-width at half-maximum on
the high-conductance side of the peak, since low-conductance half-
maxima are sometimes lost in the experimental noise.
result of a strengthened Au−π interaction, where the tunneling
pathway is coupling directly into the π-space of the second ring
while the meta linker secures that end to the electrode surface.
Similar metal−π (Pt or Ag) interactions have been used to
rationalize the conductance mechanism through symmetric
metal−benzene−metal junctions at low temperature,12 as well
as molecular junctions of C60 and stacked oligomers of
paracyclophane using Au electrodes at room temperature.13
To understand the conduction mechanism in greater detail, it
is necessary to verify that conduction occurs through the
molecular backbone of a single molecule. Therefore, we
measured the length dependence of conductance through a
series of oligomers PPn, PMn, and Pn, with n = 1, 2, and 3.
The peak conductance of each oligomer is given in Table 1. In
Figure 2B, we plot the conductance histogram peaks against the
molecular length for each series. The para−para bound series,
PPn, is especially useful as a control group representing typical
linear α,ω-diphenyl-oligoenes.14 The decay constant (β), which
describes the exponential decrease of conductance with
Figure 2. (A) 1D conductance histograms constructed using
logarithmic bins for junctions of PP1 (red), PM1 (blue), P1
(black), and M1 (dashed). M1 does not form conductive junctions.
Histograms were generated from over 5000 traces without data
selection, using 100 bins/decade. (B) Plot of molecular length vs
conductance for para−para series (PPn; red) and para−meta series
(PMn; blue). Exponential decay constants were found to be β = 0.23
0.03 and 0.27
0.08 Å−1, respectively. The effective contact
resistance for PPn is 520 kΩ and for PMn is 1820 kΩ. Conductance
for Pn (black) is represented with the mean peak conductance (black
circles) within a distribution of the inner 50th percentile (black boxes)
and the inner 80th percentile (bars). Due to junction-to-junction
variations, larger samples sizes were used for Pn (over 30 000
individual conductance traces). (C) The 2D force histogram of PM1
junctions compiled from ∼8000 traces. The statistically averaged force
profile (overlaid blue trace) shows an abrupt drop at zero-
displacement of 0.5 nN corresponding to the average force required
to rupture the junction. (Inset) Average force profiles from
measurements of PP1 and PM1 junctions, which yield a rupture
force of 0.5 nN for both.
increasing molecular length is found to be β = 0.23
0.02
Å−1, in good agreement with published results.15 For the PMn
series, we also find an exponentially decreasing conductance
with a decay constant of β = 0.27 0.08 Å−1, which is similar to
that of PPn. This indicates that conduction is through the π-
system of a single molecule and not by any other mechanism,
such as π−π-stacking between dimers.5 When measuring
conductance of the Pn series, we found that conductance
peak values shifted within half of an order of magnitude over
the course of the experiment (thousands of consecutive traces).
The experimentally observed variation in conductance is
illustrated by the error bars in Figure 2B (see the Supporting
Information for details of the variability analysis). The
are overlaid in Figure 2C (inset). Consequently, the interaction
between the Au electrode and unsubstituted phenyl ring does
not contribute much to the mechanical stability of the junction
at room temperature. Since linkers in the meta position behave
as typical mechanical contacts (albeit without providing
electrical coupling), we postulate that the conductance
modulation found between the PMn and Pn series is the
C
dx.doi.org/10.1021/ja308626m | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
variability increased with molecular length (P1 < P2 < P3),
making it impossible to determine a valid decay constant.16
Outside of the monosubstituted Pn series, the variation during
and between different experimental runs (changing tip and
substrate) caused insignificant variations in the most frequently
measured conductance. This is typically the result obtained
with methylsulfide linkers.
To further understand junction evolution, we show how the
conductance changes during elongation. By constructing 2D
conductance-displacement histograms, we can determine the
maximal junction length, or step length, as well as trends in the
molecule−electrode coupling. Figure 3A displays the 2D
histograms17 of the stilbenes junctions, PP1, PM1, and P1
(all others are given in the Supporting Information). Their
molecular conductance features extend about 0.6 nm. We note
that they are ∼0.7 nm shorter than the molecular lengths,
which is common in STM-BJs due to snapback relaxations at
the Au electrodes.18 We quantify the step length for each
molecule as detailed in the Supporting Information and
summarize them in Table 1. For each oligomeric series, the
step length increases linearly with molecular backbone length
(Figure 3D).2c The fact that all three molecular systems scale
linearly with molecular wire length, and with similar slopes,
indicates that in each case the molecular junctions are formed
with only a single molecule. The step lengths of Pn are slightly
longer (0.02−0.04 nm) than those of PPn. We attribute the
difference to changes in the linker group2c (para-SMe vs Au−π)
and not the formation of π−π-stacked dimers since significantly
larger step lengths would then be expected.
The 2D histograms also show that the slope of the
conductance-versus-displacement curves changes significantly
between each series in the order, PPn > PMn > Pn. The white
lines overlaid in Figure 3A−C represent the statistically
averaged decreasing conductance, whose slopes are tabulated
in Table 1 for all our molecules. Steeper slopes correspond to
junctions whose conductance decreases with elongation due to
weaker metal−molecule coupling. For PMn and Pn series,
where conduction occurs through π−metal overlap, the
conductance should depend sensitively on junction geometry,
and specifically, conductance should decrease as the area of
overlap between the electrode and the molecular backbone is
decreased with increasing elongation.
Since the conductance of the PPn series is systematically
higher than the corresponding PMn series, and since meta-
substitution appears to be nonconductive, our data suggest that
conduction in PMn junctions occurs through the combination
of a Au−S interaction in the para-substituted ring and a direct
Au−π interaction in the meta-substituted ring. To verify this
hypothesis, and to distinguish the role of the linker at the meta
position from its electronic influence into the π-system, we
employed a series of amine-terminated stilbenes where, in
contrast to the methylsulfide stilbenes, we are able to retain
electronic effects while disrupting the mechanical coupling at
the meta-linker (PM1A and PM1TA from Figure 1C).
Molecules terminated with primary amines (RNH2) bind to
the Au electrodes, forming molecular junctions in the STM-BJ
setup. However, when primary amines are methylated to form
tertiary amines (RNMe2), they do not bind to Au.2b In this way,
the linker group’s electron-donating contributions to the MOs
are preserved while arresting additional mechanical stabiliza-
tion.
Conductance histograms of PM1A, P1A, and PM1TA
junctions are compared in Figure 4. The conductance peak
for the para−meta bound junction (PM1A) is almost an order
of magnitude higher than that of the para−π bound junction
(P1A), similar to what was observed for the methylsulfide
analogues (PM1 vs P1). However, when the meta-NH2 of
PM1A is replaced by meta-NMe2, the conductance of the
resulting PM1TA drops by an order of magnitude, overlapping
the conductance of monofunctionalized P1A. With these direct
structural comparisons, we conclude that the electrical pathway
in the para−meta bound series is through a Au−π interaction
and that the meta linker serves as a mechanical stabilizer that
enhances the electronic coupling between the terminal phenyl
group and the Au electrode.
Theoretical Analysis. To gain insight into the electronic
structure of our molecular junctions, DFT calculations were
carried out to examine the shape of MOs. The methylsulfide-
functionalized stilbenes (PP1, PM1, and P1) are considered.
The most relevant MOs are those nearest to the Fermi level of
the electrodes. Since these junctions are generally highest-
Figure 3. 2D conductance-displacement histograms are generated
from over 5000 individual measurements on (A) PP1, (B) PM1, and
(C) P1 single-molecule junctions. Colored contour represents the
number of counts × 1000. Step lengths are marked by black arrows,
representing the 90th percentile of conductance, at 0.57, 0.56, and 0.61
nm, respectively. (D) Plot of the step lengths vs molecular wire lengths
of PPn (red), PMn (blue), and Pn (black) series, with slopes of 0.52,
0.62, and 0.52, respectively. Molecular wire lengths were determined
by DFT calculations.
D
dx.doi.org/10.1021/ja308626m | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 4. Log-binned conductance histograms of amine-functionalized
stilbene molecular wires: para−meta PM1A (blue), para−π P1A
(green), and methylated para−meta PM1TA (black) were generated
from 5000 individual conductance traces.
occupied molecular orbital (HOMO) conducting,19 we
examine the HOMO, HOMO-1, and HOMO-2 of PP1 and
PM1, which contain the p−π orbitals of sulfur and the olefin
(CC), as well as the HOMO and HOMO-1 for compounds
P1. Since other MOs are much lower in energy, their
contributions become negligible.
First, we inspect the results of the para−para case (PP1),
which are shown in Figure 5A. We see that the orbitals are
strongly coupled across the molecular backbone and include
both sulfur lone pairs (SLP). In contrast, the two SLP's in the
para−meta compound (PM1; Figure 5B) are decoupled. In
fact, the energy cost to mix these orbitals and establish a
conduction pathway (through each SLP) is roughly equal to the
energetic cost (∼0.85 eV) required to disrupt that in the
previous case (PP1). In light of this, it would be reasonable to
expect PM1 to not conduct at all. However, upon deeper
inspection we find density located on the phenyl ring opposite
the para linker and purpose this as an alternative conduction
pathway. This secondary path is also predicted to exist in PP1;
however, since it is expected to be more resistive, it may not be
experimentally distinguishable in PP1. In the special case of
PM1, the secondary path may be the only available
conductance pathway.
For the para−π bound molecule (P1; Figure 5) the ring
systems and SLP are strongly coupled to one another. Due to
the lack of a second linker, a single conductance pathway exists
in P1, which closely resembles that of the secondary path found
in PM1, leading to conducting junctions, despite lacking in the
mechanical stability provided by a second linker. This MO
interpretation helps to identify relationships linking the intrinsic
properties of the bridging molecule with the conductance
mechanism and overall device performance. Further studies
that include the mapping of the density of states and couplings
of the Au electrodes are beyond the scope of this study. Our
theoretical results agree with our experimental findings and give
weight to a dominant secondary pathway in PM1 that is similar
to that found in P1.
Figure 5. DFT calculated isosurfaces of MOs relevant in single-
molecule conductance. Geometries of PP1, PM1, and P1 were
optimized and MOs calculated at the B3LYP/6-31** level of theory in
the gas phase. HOMO energies were normalized to 0 eV for easy
comparison. Contour values set at 0.75.
of a single molecule rather than though π−π-stacked dimers.
The conductance mechanism differs between the para−para
bound (PPn) and para−π bound molecules (Pn, P1A, and
P1TA). Since the Pn series is functionalized with only one
linker, a mechanically weak, yet electronically coupled, Au−π
interaction completes the molecular circuit. The limited
strength of this new interaction causes significant variability
in junction-to-junction conductance. However, installing a
second linker at the meta position stabilizes this interaction
by securing the terminal phenyl group to the electrode surface,
such as for the para−meta compounds (PMn and PM1A).
Meta linkers result in quantum electronic interference effects
that suppress the conductance, but the position of the linkers
(meta vs para) does not significantly alter their mechanical
attachment to the Au electrodes. By exploiting this attribute we
strengthen a direct molecule−metal interaction that dominates
in the charge transport of monofunctionalized molecular
junctions, in turn enabling us to quantify the charge transport
properties arising from Au−π interactions. This strategy may
prove useful in future molecular-scale device architectures for
the positioning of molecules onto electrode surfaces while
mediating their electronic couplings, such as is needed for
advancement in single-molecule rectification.20
CONCLUSION
■
The mechanism of conductance through monofunctionalized
single-molecule wires was determined by measuring the
electrical and mechanical properties in single-molecule
junctions. By using a combination of rational molecular design,
STM- and AFM-BJ techniques, we showed that mono- and
difunctionalized molecular wires conduct through the backbone
E
dx.doi.org/10.1021/ja308626m | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(8) Kamenetska, M.; Koentopp, M.; Whalley, A.; Park, Y. S.;
Steigerwald, M.; Nuckolls, C.; Hybertsen, M.; Venkataraman, L. Phys.
Rev. Lett. 2009, 102, 126803.
(9) Parameswaran, R.; Widawsky, J. R.; Vazquez, H.; Park, Y. S.;
Boardman, B. M.; Nuckolls, C.; Steigerwald, M. L.; Hybertsen, M. S.;
Venkataraman, L. J. Phys. Chem. Lett. 2010, 1, 2114.
(10) Jaguar, 7.8 ed.; Schrodinger, L.L.C.: New York, NY, 2011.
(11) Widawsky, J. R.; Kamenetska, M.; Klare, J.; Nuckolls, C.;
Steigerwald, M. L.; Hybertsen, M. S.; Venkataraman, L. Nanotechnology
2009, 20, 434009.
(12) (a) Kiguchi, M.; Tal, O.; Wohlthat, S.; Pauly, F.; Krieger, M.;
Djukic, D.; Cuevas, J. C.; van Ruitenbeek, J. M. Phys. Rev. Lett. 2008,
101, 046801. (b) Kiguchi, M.; Miura, S.; Takahashi, T.; Hara, K.;
Sawamura, M.; Murakoshi, K. J. Phys. Chem. C 2008, 112, 13349.
(13) (a) Schneebeli, S.; Kamenetska, M.; Cheng, Z.; Skouta, R.;
Friesner, R. A.; Venkataraman, L.; Breslow, R. J. Am. Chem. Soc. 2011,
133, 2136. (b) Park, H.; Park, J.; Lim, A.; Anderson, E.; Alivisatos, A.;
McEuen, P. Nature 2000, 407, 57. (c) Schull, G.; Frederiksen, T.;
Arnau, A.; Sanchez-Portal, D.; Berndt, R. Nature Nanotechnol. 2011, 6,
23. (d) Neel, N.; Kroeger, J.; Limot, L.; Palotas, K.; Hofer, W. A.;
Berndt, R. Phys. Rev. Lett. 2007, 98, 016801.
ASSOCIATED CONTENT
* Supporting Information
Synthetic details and characterization data for all compounds;
details of experimental setup and analysis methods; theoretical
methods and tabulated results. This material is available free of
charge via the Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
mls2064@columbia.edu; lv2117@columbia.edu; cn37@
columbia.edu
■
Present Address
∥Department of Chemistry, Massachusetts Institute of Tech-
nology, Cambridge, MA.
Author Contributions
⊥These authors contributed equally.
Notes
The authors declare no competing financial interest.
(14) PPn, PMn, and Pn represent the first oligoene series having no
main-chain functionalizations to be studied in STM-BJs or other
single-molecule methods. Other oligoene systems studied are based on
carotenoid or α,ω-diphenyl-μ,ν-dicyano structures, which are deco-
rated with methyl or cyano groups along the conjugated backbone
(see: ref 15).
(15) (a) Meisner, J. S.; Kamenetska, M.; Krikorian, M.; Steigerwald,
M. L.; Venkataraman, L.; Nuckolls, C. Nano Lett. 2011, 11, 1575.
(b) He, J.; Chen, F.; Li, J.; Sankey, O. F.; Terazono, Y.; Herrero, C.;
Gust, D.; Moore, T. A.; Moore, A. L.; Lindsay, S. M. J. Am. Chem. Soc.
2005, 127, 1384.
(16) Large variability in the junction conductance of the Pn series
results in a range in the estimation of the decay constant (0 > β > 0.46
Å−1).
(17) 2D histograms are generated using an automated algorithm with
the added requirement that a G0 break is clearly identifiable in the
trace (more than 80% of the traces that start with a conductance
greater than 1 G0 and successfully break satisfy this requirement). In
2D histograms the conductance is binned logarithmically with 200 bins
per decade in conductance (y-axis), while displacement is binned
linearly (x-axis).
(18) Immediately after the formation of gold point contact
electrodes, surface adatoms quickly reorganize, thus widening the
distance between the electrodes. We refer to this as the “snapback”
distance. This was first discovered during measurements at 4 K:
(a) Yanson, A. I.; Bollinger, G. R.; van den Brom, H. E.; Agrait, N.; van
Ruitenbeek, J. M. Nature 1998, 395, 783. For room temperature:
(b) 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.
(19) (a) Venkataraman, L.; Park, Y. S.; Whalley, A. C.; Nuckolls, C.;
Hybertsen, M. S.; Steigerwald, M. L. Nano Lett. 2007, 7, 502. (b) Ma,
G. H.; Sun, L. L.; Zhang, R. X.; Shen, Z. Y.; Hou, S. M. Chem. Phys.
2010, 375, 67. (c) Li, Z.; Kosov, D. S. Phys. Rev. B 2007, 76, 035415.
(d) Quek, S. Y.; Venkataraman, L.; Choi, H. J.; Loule, S. G.;
Hybertsen, M. S.; Neaton, J. B. Nano Lett. 2007, 7, 3477.
(20) Aviram, A.; Ratner, M. A. Chem. Phys. Lett. 1974, 29, 277.
ACKNOWLEDGMENTS
■
This work has been supported in part by the NSF Career
Award (CHE-07-44185) and the Packard Foundation. This
research was also funded by the National Science Foundation
Center for Chemical Innovation (CCI Phase 1 - Award
Number CHE-09-43957).
REFERENCES
■
(1) (a) Diez-Perez, I.; Hihath, J.; Lee, Y.; Yu, L.; Adamska, L.;
Kozhushner, M. A.; Oleynik, I. I.; Tao, N. Nature Chem. 2009, 1, 635.
(b) Kim, B.; Choi, S. H.; Zhu, X. Y.; Frisbie, C. D. J. Am. Chem. Soc.
2011, 133, 19864. (c) Holmlin, R. E.; Haag, R.; Chabinyc, M. L.;
Ismagilov, R. F.; Cohen, A. E.; Terfort, A.; Rampi, M. A.; Whitesides,
G. M. J. Am. Chem. Soc. 2001, 123, 5075.
(2) (a) Xu, B. Q.; Tao, N. J. Science 2003, 301, 1221.
(b) Venkataraman, L.; Klare, J. E.; Tam, I. W.; Nuckolls, C.;
Hybertsen, M. S.; Steigerwald, M. L. Nano Lett. 2006, 6, 458. (c) Park,
Y. S.; Whalley, A. C.; Kamenetska, M.; Steigerwald, M. L.; Hybertsen,
M. S.; Nuckolls, C.; Venkataraman, L. J. Am. Chem. Soc. 2007, 129,
15768. (d) Martin, C. A.; Ding, D.; Sorensen, J. K.; Bjornholm, T.; van
Ruitenbeek, J. M.; van der Zantt, H. S. J. J. Am. Chem. Soc. 2008, 130,
13198. (e) Hong, W.; Manrique, D. Z.; Moreno-García, P.; Gulcur,
M.; Mishchenko, A.; Lambert, C. J.; Bryce, M. R.; Wandlowski, T. J.
Am. Chem. Soc. 2011, 134, 2292.
(3) (a) Mayor, M.; Weber, H. B.; Reichert, J.; Elbing, M.; von
Hanisch, C.; Beckmann, D.; Fischer, M. Angew. Chem., Int. Ed. 2003,
42, 5834. (b) Solomon, G. C.; Andrews, D. Q.; Hansen, T.;
Goldsmith, R. H.; Wasielewski, M. R.; Van Duyne, R. P.; Ratner, M.
A. J. Chem. Phys. 2008, 129. (c) Ke, S. H.; Yang, W. T.; Baranger, H. U.
Nano Lett. 2008, 8, 3257. (d) Aradhya, S. V.; Meisner, J. S.; Krikorian,
M.; Ahn, S.; Parameswaran, R.; Steigerwald, M. L.; Nuckolls, C.;
Venkataraman, L. Nano Lett. 2012, 12, 1643. (e) Guedon, C. M.;
Valkenier, H.; Markussen, T.; Thygesen, K. S.; Hummelen, J. C.; van
der Molen, S. J. Nat. Nanotechnol. 2012, 7, 305.
(4) (a) Frei, M.; Aradhya, S. V.; Koentopp, M.; Hybertsen, M. S.;
Venkataraman, L. Nano Lett. 2011, 11, 1518. (b) Xu, B. Q.; Xiao, X.
Y.; Tao, N. J. J. Am. Chem. Soc. 2003, 125, 16164.
(5) π−π-staking in STM-BJs has been previously reported; however,
many experimental factors were different from our study, including the
structure of their bridging molecules, longer molecular lengths, and
different solvent system during their STM-BJ measurements. See: Wu,
S. M.; Gonzalez, M. T.; Huber, R.; Grunder, S.; Mayor, M.;
Schonenberger, C.; Calame, M. Nat. Nanotechnol. 2008, 3, 569.
(6) Wittig, G.; Schollkopf, U. Chem. Ber. 1954, 87, 1318.
(7) Wadsworth, W. S. In Org. React.; John Wiley & Sons, Inc.: New
York, 1977; Vol. 25, p 73.
F
dx.doi.org/10.1021/ja308626m | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX