Structural Effects on Electrical Conduction of Conjugated Molecules Studied by Scanning Tunneling Microscopy

Article abstract of DOI:10.1021/jp0018450

We have studied electrical conduction of conjugated molecules with phenyl rings embedded into alkanethiol self-assembled monolayers (SAMs), to investigate the molecular structural effect on the electrical conduction. Scanning tunneling microscope (STM) images of this surface revealed that the conjugated molecules with phenyl rings adsorbed mainly on defects and domain boundaries of the pre-assembled alkanethiol (nonanethiol C9) SAM and formed conjugated domains. In the case of conjugated molecules with one or three methylene groups between the sulfur and phenyl rings, the measured height of the conjugated molecular domains depended on their lateral sizes, while a strong dependence was not observed in the case of conjugated molecules without a methylene group. By analyzing size dependence on the height of the conjugated molecular domain, we could evaluate the electronic conductivity of the molecular domains. As a result of the analysis, to increase the vertical conduction of the molecular domains, one methylene group was found to be necessary between the sulfur and aromatic phenyl rings. Local barrier heights on the conjugated molecular domains in all the cases were larger than on the C9 SAM surface, suggesting that the increase in the vertical conductivitity is not likely to be due to the lowering of the local barrier height, but can be attributed to the conjugated molecular adsorption. X-ray photoelectron spectra (XPS) and ultraviolet light photoelectron spectra (UPS) revealed that the carrier density among conjugated molecular SAMs does not depend on the number of methylene groups between the sulfur and phenyl rings, suggesting that the higher vertical conduction of conjugated molecules with one methylene group can probably be attributed to higher transfer probability of carriers during the STM measurements.

Full text of DOI:10.1021/jp0018450

11680
J. Phys. Chem. B 2000, 104, 11680-11688
Structural Effects on Electrical Conduction of Conjugated Molecules Studied by Scanning
Tunneling Microscopy
Takao Ishida,*,†,‡ Wataru Mizutani,†, Nami Choi,§ Uichi Akiba,|| Masamichi Fujihira,|| and
Hiroshi Tokumoto†
Joint Research Center for Atom Technology (JRCAT), National Institute for AdVanced Interdisciplinary
Research (NAIR), 1-1-4 Higashi, Tsukuba, Ibaraki, 305-8562, Japan, PRESTO-Japan Science and Technology
Corporation (JST), 4-1-8 Honcho, Kawaguchi, Saitama 332-0012, Japan, JRCAT-Angstrom Technology
Partnership (ATP), 1-1-4 Higashi, Tsukuba, Ibaraki, 305-0046, Japan, and Department of Biomolecular
Engineering, Faculty of Bioengineering and Biotechnology, Tokyo Institute of Technology, 4259 Nagatsuta,
Midori-ku, Yokohama 226-8501, Japan
ReceiVed: May 18, 2000; In Final Form: August 25, 2000
We have studied electrical conduction of conjugated molecules with phenyl rings embedded into alkanethiol
self-assembled monolayers (SAMs), to investigate the molecular structural effect on the electrical conduction.
Scanning tunneling microscope (STM) images of this surface revealed that the conjugated molecules with
phenyl rings adsorbed mainly on defects and domain boundaries of the pre-assembled alkanethiol (nonanethiol
C9) SAM and formed conjugated domains. In the case of conjugated molecules with one or three methylene
groups between the sulfur and phenyl rings, the measured height of the conjugated molecular domains depended
on their lateral sizes, while a strong dependence was not observed in the case of conjugated molecules without
a methylene group. By analyzing size dependence on the height of the conjugated molecular domain, we
could evaluate the electronic conductivity of the molecular domains. As a result of the analysis, to increase
the vertical conduction of the molecular domains, one methylene group was found to be necessary between
the sulfur and aromatic phenyl rings. Local barrier heights on the conjugated molecular domains in all the
cases were larger than on the C9 SAM surface, suggesting that the increase in the vertical conductivitity is
not likely to be due to the lowering of the local barrier height, but can be attributed to the conjugated molecular
adsorption. X-ray photoelectron spectra (XPS) and ultraviolet light photoelectron spectra (UPS) revealed that
the carrier density among conjugated molecular SAMs does not depend on the number of methylene groups
between the sulfur and phenyl rings, suggesting that the higher vertical conduction of conjugated molecules
with one methylene group can probably be attributed to higher transfer probability of carriers during the
STM measurements.
1. Introduction
techniques4-9 or the mechanical break junction technique.10 For
example, Bumm et al. estimated the conductance of single
conjugated molecules embedded in an insulative SAM film
using scanning tunneling microscopy (STM).4 They measured
the height of the molecules adsorbed on a metal surface with
the molecular axis almost vertical to the surface. Interestingly,
the measured conductivities for conjugated molecules using
SPM4-9 are larger than those of undoped conductive organic
materials;11 the measured conductivities are order of 10-3 S/cm,
which are of the order larger than those of undoped conductive
polymers (10-7-10-10 S/cm 11). This is one of the unexplained
areas concerning molecular conduction as measured with SPM
techniques. Moreover, detailed experimental data regarding the
effect of molecular structure on the electrical conduction are
not yet available for measurements made with SPM techniques,
even though there exist many kinds of effective conjugated
groups, like phenylene or thiophene. For example, while the
effect of molecular length (the number of phenyl rings) on the
conductivity is important in understanding the electrical proper-
ties of conductive molecules, there has been no systematic
experimental study suing STM, to evaluate the effect of the
number of phenyl rings. The effect of the contact between the
molecule and the metal electrode on the electrical conduction
is another important problem which must be solved before the
realization of molecular devices. Since the insertion of an
insulating methylene group between the metal and aromatic
As the trend toward smaller devices continues, the use of
individual molecules to fabricate device components becomes
more and more attractive. Organic molecules are inherently
nanoscale in size and highly uniform in nature.1 Furthermore,
organic molecules can be synthesized with unique properties
that could be used to promote their self-assembly with one
another and to specific surfaces, and to perform functions that
can provide memory and logic operations. For these reasons,
the field of molecular electronics has generated considerable
interest in recent years.
For the realization of molecular devices, new conductive
organic materials, e.g., carbon nanotubes2 or conjugated mo-
lecular wires3 have been discovered. Since these materials are
very attractive, many interesting studies have appeared concern-
ing the electron transfer along the molecular axis theoretically
and experimentally using scanning probe microscope (SPM)
* Corresponding Author. Present address: Mechanical Engineering
Laboratory, 1-2 Namiki, Tsukuba, Ibaraki 305-8564, Japan. E-mail:
tishida@mel.go.jp.
† JRCAT-NAIR.
‡ PRESTO- JST.
§ JRCAT-ATP.
|| Tokyo Institute of Technology.
Permanent Address: Electrotechnical Laboratory, Tsukuba, Ibaraki 305-
8568, Japan.
10.1021/jp0018450 CCC: $19.00 © 2000 American Chemical Society
Published on Web 11/17/2000

Electrical Conduction of Conjugated Molecules
J. Phys. Chem. B, Vol. 104, No. 49, 2000 11681
phenyl ring is expected to be useful for understanding the
metal-molecule contact, changing the number of methylene
groups is effective for the investigation of the contact effect. In
this case, we can easily imagine that the molecular conduction
with many methylene spacers should be lower than the
conjugated molecule without methylene groups.
CHART 1: Molecular Structures of Conjugated
Molecules Used in This Study
Meanwhile, phase-separation for mixed SAMs has been
investigated by SPM from the viewpoint of nanoscale pat-
terning.12-21 Recently, we used conjugated molecules, e.g.,
phenylene oligomers to form a phase-separated surface,18-21 and
found that various sizes of domains of these conjugated
molecules implanted into the insulate alkanethiol SAMs. We
further evaluated both the vertical and lateral conductivities of
conjugated molecular domains using a conducting disk model,
where the intermolecular interaction may increase the electrical
conduction.18,19 The investigation of nanometer molecular scale
domains is important when the conjugated molecule is to be
used as one of the interconnection units of molecular device.
From this point of view, we believe that our data concerning
the size and height can provide a useful information for future
molecular device fabrication. In addition to the above significant
experimental studies, some interesting theoretical calculations
were performed concerning the conduction of conjugated
molecules present between two metal electrorodes.22,23
In the present study, we extended the above experiment to
several kinds of phenylene oligomers with thiol groups in order
to understand molecular structural effects on the electrical
conduction. We used several kinds of conjugated molecules with
two or three aromatic rings to evaluate the effect of the number
of aromatic rings. Since it was expected that the presence of a
methylene group affects both the molecular arrangements24 and
electrical conduction, we used conjugated molecules having one
or three methylene groups between the sulfur and the aromatic
rings, or no methylene group to investigate the metal-molecular
contact effect. We evaluated the electrical conduction of
conjugated molecules by analyzing the dependence of the
measured height of the conjugated molecular domains on their
lateral sizes. We present our data of local barrier height images
taken simultaneously with STM images, to discuss the reason
for the increase in the vertical conductivity of conjugated
domains. Finally, we display the photoelectron spectra to discuss
the relationship between the carrier density among the mono-
layer and electrical conduction data.
chloride as solvents (TP3, TP0, BP0, and BP1). After being
taken out of the solution, the Au substrates were rinsed with
pure solvent to remove physisorbed multilayers. To insert the
conjugated molecules, the Au substrate with pre-assembled
SAMs, were immersed into 0.1 mM solution for 1-12 h.
STM Measurements. STM images were obtained using a
Seiko Instruments SPA340 unit in air, or a homebuilt UHV-
STM system with a typical tunneling current of 20 pA and a
tip biases of 0.85-1.2 V, corresponding to a tunneling resistance
of 40-60 GΩ. The local barrier height (LBH) image was taken
simultaneously with the STM topography.13,19 The LBH images
were measured by applying a small sinusoidal voltage to the
z-axis piezoactuator to give the gap s a small modulation ds
and then measuring the corresponding variation of the tunneling
current with a lock-in amplifier. The LBH can be calculated
from the following equation:28
φA(eV) ) 0.952[d(ln I)/ds(Å)]2
(1)
where I is the tunneling current and s is the tip-sample
separation. To obtain a LBH image simultaneously with a STM
image, we measured the d(ln I) at each point, in addition to
measuring the tip height. In our experiment the modulation
frequency was set at 4.0 kHz, which was higher than the cutoff
frequency of the feedback loop (less than 1.0 kHz), but lower
than the response cut off the current amplifier of the STM. The
amplitude of the modulation was such that the corresponding
ds was 0.03 nm, whose value is smaller than the gap distance
of about 0.2 nm.
XPS and UPS Measurements. High-resolution XPS spectra
were recorded using a VG Scientific Inc. ESCALAB 220iXL
system with a monochromatic Al kR X-ray source (1486.6 eV).
The binding energy was calibrated using the Au (4f7/2) peak
energy (84.0 eV) as an energy standard. The X-ray power, the
pass energy of the analyzer, and take-off angle of the photo-
electrons were set at 180 W, 20 eV, and 90°, respectively. XPS
peaks were fitted using the spectra processing program in the
XPS system.21 In the case of UPS measurements, He I line (21.2
eV) was used as a light source. The UV light power and take-
off angle were 150 W and 20°, respectively.
2. Experimental Section
Chemicals. We used the following molecules: nonanethiol
(C9, Aldrich), [1,1′:4′,1′′-terphenyl]-4-thiol (TP0), [1,1′:4′,1′′-
terphenyl]-4-methanethiol (TP1), [1,1′:4′,1′′-terphenyl]-4-pro-
panethiol (TP3), 4-biphenylthiol (BP0), and 4-biphenylmethane-
thiol (BP1). The molecular structures are shown in Chart 1.
Synthesis methods of TP0, TP1, BP0, and BP1 molecules have
been described elsewhere.24-26 TP3 molecules were synthesized
as shown in Scheme 1.27
Au Deposition and SAM Formation. An atomically flat Au
(111) surface was epitaxially grown on mica by vacuum
deposition under a base pressure of about 4 × 10-8 Torr. The
mica was preheated at 440 °C for 4 h before deposition. The
deposition rate of Au was kept at 0.2 nm/s. After the deposition,
the substrate was annealed at 480 °C for 30-60 min to obtain
large terraces on the Au surface. The flatness of the terraces of
the Au surface was checked by STM to be atomically flat over
200 nm. For the C9 SAMs, the Au substrates were immersed
into 1 mM ethanol solution for 24 h or more. In the case of
conjugated molecules, we used chloroform (TP1) and methylene
3. Results and Discussion
3.1. STM Images of Mono-component SAMs. First, we
describe STM data of the mono-component SAMs of series of
C9 and conjugated molecules because the molecular arrange-
ment of these conjugated molecules on the Au surface is
important to discuss the electrical conduction of the nanometer

11682 J. Phys. Chem. B, Vol. 104, No. 49, 2000
Ishida et al.
Figure 1. STM images of the SAMs (1); (a) C9 SAM as deposited; (b) magnified image of (a); (c) TP0 SAM after dipping into 0.1 mM TP0
solution for 48 h; (d) TP1 SAM after dipping into 0.1 mM TP1 solution for 48 h; (e) magnified image of (d); (f) TP3 SAM after dipping into 0.1
mM TP3 solution for 48 h; (g) magnified image of (f). In (b), (e), (g), we observed structures with a molecular distance of about 0.5 nm, while this
arrangement was not obtained in the TP0 SAM. STM images of C9 and TP3 SAMs were measured with a typical tunneling current of 20 pA and
a tip bias of 0.85 V. On the other hand, the STM images of TP0 and TP1 SAMs were obtained with a typical tunneling current of 100 pA and a
tip bias of 0.5 V.
SCHEME 1: Synthesis of TP3 Molecule
molecular domains. Figure 1 shows the STM images of the
mono-component SAMs. For the C9 SAM, after the immersion
of C9 solution for 24 h, we could see (x3 × x3) R30°
structures with a molecular distance of 0.5 nm, with domain
boundaries and depressions (Figure 1a,b).29,30 On the other hand,
in the case of TP0, the TP0 molecules are likely to adsorb on

Electrical Conduction of Conjugated Molecules
J. Phys. Chem. B, Vol. 104, No. 49, 2000 11683
the Au surface while retaining the herringbone structure which
is observed for the clean Au (111) surface (Figure 1c).31
Although it has been reported that TP0 molecules arranged with
the same (x3 × x3) R30° structure on a Au surface,25 we could
not observe a clear molecular arrangement with STM in the
magnified image. Similar STM images were taken in the case
of BP0 (data not shown). However, we obtained molecular
images in the cases of TP1 (Figure 1d,e), TP3 (Figure 1f,g)
and BP1 (data not shown). For the conjugated molecular SAMs,
the height difference of the depression is about 0.2 nm, whose
value is almost the same as that of C9 SAM. However, the
density of depressions is higher than that of C9 SAM.
into 0.1 mM TP1 solution for 1 h (Figure 2a), a phase-separated
STM image was observed. For the C9 SAMs, the C9 molecules
form 10-20 nm diameter domains, and these domain boundaries
are clearly seen in the STM image (cf. Figure 1a). After
insertion, protrusions which could be assigned to conjugated
molecular domains surround the C9 domains and have a
diameter of 10-20 nm whose size is almost identical to that of
C9 domains (cf. Figure 2a,c,d). Therefore, we can conclude that
the conjugated molecules in the solution are considered to adsorb
gradually on the Au surface, mainly on the uncovered areas
such as defects and domain boundaries of the pre-assembled
C9 SAM.20
Tao et al. evaluated the structure of these aromatic derivatized
thiols with SPM and made electrochemical measurements and
concluded that the structure and molecular densities of these
SAMs were dependent on the number of the aromatic rings and
substituted groups.24 They proposed that there exist two types
of molecular orientations, when one methylene group is not
present between the sulfur and aromatic rings such as BP0, TP0,
i.e., sp and sp3 configurations, causing molecular disordering
(see inset in Figure 1). The presence of these molecular
orientations was predicted from theoretical calculations.32 On
the other hand, if the molecule has one methylene group between
the sulfur and aromatic rings, these conjugated molecules form
a well-ordered structure, i.e., commensurate (x3 × x3) R30°
structures.
The cross-sectional profiles across the TP1 domains embed-
ded in the C9 SAM (Figure 2b) indicate that the apparent height
difference of the TP1 domains in the STM image is not uniform
and depends on the TP1 domain size. In cases of TP3 and BP1,
the phase-separated STM images were obtained in a similar way
to those of TP1 (Figure 2c,d). We investigated the relationship
between the domain sizes of conjugated molecules and the
observed height differences of phase-separated SAMs. Figure
3 shows the relationship between the domain sizes of the
conjugated molecules and observed height differences. It should
be noted that the apparent height difference of the TP1 domains
in the STM image is not uniform, but changes as shown in
Figure 2b. The smallest protrusion which can be probably
attributed to a single TP1 molecule shows lowest (Figure 3a).
Additional bars indicate the shortest and longest dimensions of
the noncircular domains. Up to a domain size of 3-6 nm, the
TP1 domain height increases and saturates at around 0.6-0.7
nm. In the case of TP3 implanted into C9 SAM, the domain
height saturates at around 0.5-0.6 nm (Figure 3b). For the BP1
embedded into C9 SAM, the saturated height value decreased
to be at about 0.3 nm (Figure 3c).
Poirier reported that Au surface atoms are forced out of the
surface layer by relaxation of the compressed herringbone
structure, and then the depressions as shown in Figure 1 are
formed during the n-alkanethiol adsorption.33 Poirier described
in the same paper, however, that such a feature did not appear
in the case of aromatic thiols with shorter chains.33 In addition,
Hara et al. reported that the adsorption of 4-mercaptopyridine
with an aromatic ring induced the periodic herringbone structure
without making depressions.34 Thus, the remaining of the
herringbone structure for TP0 adsorption is likely to be a
characteristic feature in the case of conjugated molecules without
a methylene group and should be related to the two molecular
orientations in these kinds of molecular species. One possible
explanation of the difference is as follows. Conjugated molecules
such as TP0 without a methylene group, adsorbed along the
corrugations on the Au reconstructed surface at the initial stage
of SAM growth, with two kinds of molecular orientation.
However, since the bond between the sulfur and conjugated
aromatic ring is expected to be rigid and needs a large energy
to change the molecular direction, these molecules could not
form commensurate (x3 × x3) R30° structures. On the other
hand, in the case of conjugated molecules with some methylene
groups, (x3 × x3) R30° structures appear after immersion,
because the methylene groups may provide some flexibility to
arrange the molecular directions.
The size dependence of the conductive molecular domain
height indicates that the vertical conductance of the domains
increases as the number of molecules in the domain increases.18-20
To discuss the electrical conduction, here we show a schematic
drawing of conjugated molecules embedded into C9 SAMs
(Figure 4). The lengths of TP and BP units are expected to be
about 1.4 and 1.0 nm, respectively.25 When one methylene group
is located between sulfur and TP or BP unit, the molecular
length of 0.12 nm is increased.35 Thus, the molecular lengths
of TP1, TP3, and BP1 are expected to be 1.55, 1.85, and 1.15
nm, respectively. Since the tilt angle of these molecules is
considered to be about 20° and these molecules are expected
to adsorb onto an Au surface at sp3 configuration from our
previous XPS data about TP1,18 the thickness of TP1, TP3, and
BP1 SAMs are estimated to be 1.46, 1.73, and 1.08 nm,
respectively. From the XPS data the thickness of C9 SAM is
1.01 nm. Thus, the expected height differences of TP1 + C9,
TP3 + C9, and BP1 + C9 systems are 0.45, 0.72, and 0.07
nm, respectively. We made measurements with a tapping mode
atomic force microscope (AFM) and obtained height differences
between the C9 and these conjugated molecules. These values
of TP1 + C9 and TP3 + C9 are 0.40 nm and 0.80 nm,
respectively. On the other hand, we could not observe specific
topography higher region on the AFM image in the case of BP1
+ C9 SAM. The AFM data agrees with the estimated height
difference from the XPS data. In the case of TP1 + C9 SAM,
if we determine the tunneling gap between C9 and tip to G0
(nm), the tunneling gaps at the single molecule and larger
molecular domains are estimated to be G0 - 0.24 nm and G0
+ 0.18 nm. In the same manner, for the TP3 + C9 and BP1 +
Also, the increase in the depressions of the TP1 and TP3
SAMs compared with C9 SAM could be attributed to the surface
layer relaxation. Perhaps the conjugated molecules with one or
more methylene groups, like TP1 or TP3, generate stronger
stress during adsorption than C9, while TP0 cannot give such
a stress to Au surface.
3.2. STM Images and Observed Height Dependence on
Domain Sizes of Conjugated Molecules (1): TP1, TP3, and
BP1. Here, we describe the STM images of conjugated
molecules implanted into C9 SAMs and the apparent height
dependence on the conjugated molecular domain size.18-20
Figure 2 shows a series of STM images and cross-sectional
profiles of C9 + conjugated molecular SAMs. After dipping

11684 J. Phys. Chem. B, Vol. 104, No. 49, 2000
Ishida et al.
Figure 2. STM images of the SAMs (2); (a) TP1 in C9 SAM after dipping into 0.1 mM TP chloroform for 1 h; (b) the cross-sectional profiles
across of the TP1 domains embedded in C9 SAM (c) TP3 in C9 SAM after dipping into 0.1 mM TP3 methylene chloride solution for 10 h; (d) BP1
in C9 SAM after dipping into 0.1 mM BP1 methylene chloride solution for 10 h. In these binary SAMs, all the STM images were taken with a
typical tunneling current of 20 pA and a tip bias of between 0.85 and 1.3 V with UHV-STM.
Figure 3. (a) Relationship between the domain sizes of TP1 in C9 SAM and observed height differences: (b) same plot of TP3 in C9 SAM; (c)
same plot of BP1 in C9 SAM.
C9 SAMs, these values are assumed to be G0 - 0.52 nm, G0 -
0.22 nm (TP3), G0 + 0.05, and G0 + 0.30 nm (BP1).
the STM image are more than 0.2 nm. In the present study, in
the case of BP1 + C9 SAM, a similar tendency was observed.
We consider these data to be the strong evidence that size
dependence is mainly due to electronic condition. These data
suggested that the ordering effect on the size dependence is
smaller than that of the conduction as expected, when the
conjugated molecules can form ordered structure.
Our data indicated that the single molecular resistances of
these molecules are decreased in the order of TP3 > TP1 >
C9 g BP1. However, the domain resistances of these molecules
In some cases, e.g., TP1 + C9 SAM, where the structural
height difference is expected to be more than 0.4 nm, the
observed height difference seems to be explained by the domain-
size-dependent ordering of the molecules. In our previous paper,
we investigated TP1 + C12 SAMs where the structural height
difference is negligibly small as well, to check this effect.18 In
this case, the estimated thicknesses of the TP1 and C12 SAMs
are almost identical, while the observed height differences in

Electrical Conduction of Conjugated Molecules
J. Phys. Chem. B, Vol. 104, No. 49, 2000 11685
groups is 0 or 1.36 We will discuss the reason on the basis of
the XPS and UPS data later.
In our previous studies, we estimated the single molecular
resistances or molecular conduction by assuming that this
junction behaves ohmically and the gap between the STM tip
and molecule linearly change.18-20 However, the junction STM
tip/molecules/Au was considered to be a tunneling junction.22
In the case of tunneling junction, the quantitative estimation
became very complicated and the numbers of the parameters
(e.g., work function) increased. Moreover, more discussions are
still needed to decide the transfer mechanism through the
junction STM tip/molecules/Au. Therefore, we do not estimate
molecular resistance values in this study. In our previous paper,
we also calculated the lateral conduction of TP1 molecules by
assuming a resistor network model from the size dependence
of height difference.18,19 The lateral conduction of nanometer
molecular domains reflect the increase in the vertical conduction,
i.e., the maximum vertical conduction becomes a higher value
when the lateral conduction is large. Since the physical
description of the size dependence of height difference is not
clear and many interpretations are possible, we do not estimate
the lateral conduction values either in this paper. We will discuss
the physical description of the increase in the vertical conduction
of the molecular domain later based on the LBH data.
3.3. STM Images and Observed Height Dependence on
Domain Sizes of Conjugated Molecules (2): TP0 and BP0.
In cases of binary SAMs of TP0 + C9 and BP0 + C9, the size
dependence of the height difference plots are different from
those as shown previously. Figure 5 shows the relationships
between the domain sizes of conjugated molecules and observed
height differences of TP0 + C9 (Figure 5a) and BP0 + C9
SAMs (Figure 5b). In these SAMs, the apparent heights at
around single molecule (less than 1 nm) were in the range 0.2-
0.5 nm. In the case of TP0 + C9, with the increase in the domain
size, the apparent height difference decreased and saturated at
around 0.3 nm (cf. Figure 5a). A similar tendency was observed
in the case of BP0 + C9 SAM (Figure 5b). As described before,
these conjugated SAMs without a methylene group between the
sulfur and the aromatic rings, tend to have two molecular
directions due to the existence of the two sulfur-Au bond
orientation.24 We consider that our data on the apparent height
might reflect the above structural change. Even in the conjugated
molecular domain at an exact size larger than 5 nm, it is
expected that these two kinds of molecular orientation can cause
molecular disordering. Figure 5c,d shows schematic drawings
of TP0 + C9 (Figure 5c) and BP0 + C9 (Figure 5d) SAMs.
By assuming sp conformation, BP0 and TP0 molecules are
expected to have thicknesses of more than 1.0 and 1.4 nm,
respectively. Tapping mode AFM images showed 0.3 nm height
difference in the case of TP0 + C9, while no clear difference
was observed in the BP0 + C9 SAM, supporting the above
structural assumption.
Figure 4. Schematic drawing of conjugated molecules embedded into
C9 SAMs. (a) TP1 in C9; (b) TP3 in C9; (c) BP1 in C9.
are expected to be in the order TP3 > C9 > TP1 > BP1. Both
the single molecular and domain resistance were decreased by
increasing the number of phenyl rings, suggesting that there
may exist a sum law in the conduction of conjugated molecule.
For example, Samanta et al. calculated that the resistance is
expected to increase with the number of phenyl rings.22
These data may suggest that the domain resistance of
conjugated molecules without any methylene group between
the aromatic rings and the sulfur, did not increase at larger size,
if we assumed the domain height difference was nearly constant.
In Section 3.1, ordered structures were not observed in these
kinds of molecular SAMs without a methylene group. The height
diffence data is likely to be due to the fact that the molecular
disordering prevents the lateral connection of molecules.
Both the single and domain resistances of TP3 are expected
to be higher than those of TP1, suggesting that the increase in
the number of methylene groups between the sulfur and phenyl
rings increases the molecular resistance. Possible reasons for
the higher resistance of TP3 than those of TP1 are as follows:
(a) the decrease in the carrier transport probability between the
metal surface and molecules by the increase in the number of
methylene groups; (b) there may exist a specific doping effect
where large amounts of carrier at the metal surface penetrate
into the conjugated molecules, when the number of methylene
3.4. LBH Measurements of Conjugated Molecules. The
origin of conductivity change can be considered as follows:(i)
increase in the density of states with the domain size. This
phenomena is reported by Zeppenfeld et al., in the case of

11686 J. Phys. Chem. B, Vol. 104, No. 49, 2000
Ishida et al.
Figure 5. Relationship between the domain sizes and observed height differences of (a) TP0 in C9; (b) BP0 in C9. Schematic drawing of TP0 (c)
and BP0 (d) embedded into C9 SAMs.
interface, and then caused the work function change.38-40
Perhaps the strength of the dipole moment can be slightly
TABLE 1: Relative LBH Values of Conjugated Molecules
Embedded in the C9 SAMs. All the LBH Values Were
Taken with a Typical Tunneling Current of 20 PA and a Tip
controlled by the insertion of a methylene group between sulfur
Bias of between 0.85 and 1.3 V with UHV-STM
and conjugated phenyl rings.39 Therefore, the change of LBH
values are likely to be reflected to both the conduction and dipole
moment.
conjugated molecules
relative LBH values (eV)
TP0
0.61
TP1
0.24
TP3
0.11
BP0
0.84
BP1
0.24
submonolayer metal deposition;37 (ii) lowering of the local
barrier height due to conjugated molecule adsorption. In this
case, the local barrier height on the conjugated molecules would
be lower than that on the C9 surface. Direct observation of the
local barrier height with the STM technique is possible,12,19
while it is difficult to observe the increase in the density of
states by STM. To obtain information on the density of states
at valence band directly, photoelectron spectroscopy is suitable.
However, in the case of phase-separated SAMs, photoelectron
spectroscopy is not available because the spatial resolution of
photoelectron spectroscopy is not so high (about µm order).
Therefore, we measured LBH of these conjugated molecules
embedded into C9 SAMs to understand the origin of the
conductivity change of conjugated molecule on the domain size.
Figure 6 shows the STM and LBH images taken simulta-
neously. For all the cases, the LBH on the conjugated molecular
domains are larger than that on the C9 surface. The estimated
LBH value of the C9 surface is 1.7 ( 0.9 eV. The estimated
LBH values of conjugated molecules are higher than that of
the C9 region, as listed in Table 1. There is a tendency that the
LBH value decreased with the number or presence of methylene
groups between the sulfur and phenyl rings, and our estimated
electrical conduction of conjugated molecules depended on the
number of methylene groups as shown before. Thus, this
tendency should be related to the molecular electrical conduc-
tion. It has been considered that the adsorbed organic molecules
on a metal surface formed a dipole layer at the metal/molecule
It should be noted that the LBH values were not dependent
on the domain size, while the height difference changed with
the domain size in the case of conjugated molecules such as
TP1.18 These data suggest that the increase in the vertical
conductivity of molecular domains is not due to the lowering
of barrier height, but likely to be due to the increase in the
density of states.
3.5. XPS and UPS Data on the Mono-component TP1 and
TP3 SAMs. Our data on the vertical conductivity,18-20 as well
as other groups5-10 measured with SPM, exhibited higher
conductivity than an undoped conductive polymer (10-7-10-10
S/cm 11), even at zero bias, as described in the Introduction.
One possible origin is likely to be the doping effect from the
Au surface to the conjugated molecules. If such a doping
occurred, the carrier density (e.g., electron, hole, etc.) among
the monolayer should depend on the conductivity.
To observe the change of carrier density, XPS or UPS
techniques are considered to be appropriate. For conjugated
molecules such as benzene, C60, it is well-known that the C(1s)
π-π* shake-up satellite peaks appear at a higher binding energy
region of major peak at around 284-285 eV.41-43 Since the
appearance of these satellite peaks is related to the carrier density
at the valence band, we can expect that the XPS data would
depend on the molecular conduction. For example, Onoe et al.43
reported that the C(1s) satellite peak intensity decreased with
the dimerization of C60 molecules due to the decrease in the
carrier density.

Electrical Conduction of Conjugated Molecules
J. Phys. Chem. B, Vol. 104, No. 49, 2000 11687
Figure 6. Schematic drawing of the reason for the increase in the vertical conduction (a) increase in the density of states; (b) decrease in the barrier
height. STM topograpical image (c) and local barrier height image (d) of TP in C9 SAM. All the STM images were taken with a typical tunneling
current of 20 pA and a tip bias of 0.85-1.3 V with UHV-STM.
If the increased vertical conduction, with the decrease in the
number of methylene group was due to the doping effect from
the Au substrate, the shape and intensity of both the satellite
peak and valence band should be different.43 However, since
the intensity of satellite peaks and valence bands on the TP1
and TP3 SAMs are identical, the carrier density among the
monolayer is expected to be almost the same in both the SAMs.
Therefore, we could not confirm the doping effect on the
molecular conductivity in these SAMs and the decrease in the
vertical conduction of TP3 might simply be attributed to the
Figure 7. C(1s) XPS and UPS spectra of (a) TP3 C(1s) region; (b)
TP1 C(1s) region; (c) TP3 valence band; (d) TP1 valence band region.
decrease in the transfer probability of carriers, by inserting the
methylene groups between the sulfur and phenyl rings. More-
over, the bulk conductivities of conductive polymers includes
We measured mono-component TP1 and TP3 SAMs to
inter-intra hopping and tunneling effects, which decrease with
observe information on carrier densities. At the C(1s) region
the conductivity. However, our XPS and UPS data necessarily
between 280 and 300 eV (Figure 7a,b), weak and broad
mean denying the doping effect, because the experimental
structures were detected in the cases of TP3 (Figure 7a) and
conditions of photoelectron spectroscopy are different from STM
TP1 (Figure 7b). We magnified satellite peaks, the shape of
measurements. To confirm the doping effect, further detailed
satellite peaks of TP1 and TP3 are almost identical. The ratios
study is essential.
of the C(1s) satellite/Au(4f) of the TP1 and TP3 SAMs are also
New and interesting physical phenomena of organic mol-
the same values and about 0.08. The peak intensities in both
ecules may be found to control and determine molecular
the SAMs are weaker than those as reported previously.41-43
conduction. For example, recently, large negative differential
Our data suggest that the amount of carrier among the monolayer
resistance was observed in the SAM of conjugated molecules
is very small.
containing nitroamine redox center.45,46 Such new physical
It is also possible to observe valence band structure with XPS.
phenomena should be helpful in developing molecular devices
However, the effect of Au surface on the spectra cannot be
in the future.
ignored in the case of XPS measurements of organic monolayer
less than 2 nm, because the escape depth of photoelectron from
4. Conclusions
the Au substrate is longer than 2 nm.44 Thus, we measured UPS
spectra of the valence band region instead of XPS. The UPS
spectra between -5 and 20 eV in TP3 (Figure 7c) and TP1
(Figure 7d) are also identical. The Fermi level at both the UPS
spectra shifted about 1.7 eV, due to the dipole layer formation
by the adsorption of organic molecules.40
We have measured electrical conduction and LBH of
conjugated molecules embedded into C9 SAMs by STM to
investigate the structure effect on the electrical conduction. We
further measured XPS and UPS for some molecular SAMs, to
understand the origin of higher electrical conduction of these

11688 J. Phys. Chem. B, Vol. 104, No. 49, 2000
Ishida et al.
(9) Bumm, L. A.; Arnold, J. J.; Dunbar, T. D.; Allara, D. L.; Weiss,
P. S. J. Phys. Chem. 1999, B103, 8122.
(10) Reed, M. A.; Zhou, C.; Muller, C. J.; Burgin, T. P.; Tour, J. M.
Science 1997, 278, 252.
(11) Epstein, A. J. MRS Bulletin 1997, 22, 16.
(12) Stranick, S. J.; Parikh, A. N.; Allara, D. L.; Weiss, P. S. J. Phys.
Chem. 1994, 98, 11136.
(13) Bumm, L. A.; Arnold, J. J.; Charles, L. S.; Dunbar, T. D.; Allara,
D. L.; Weiss, P. S. J. Am. Chem. Soc. 1999, 121, 8017.
(14) Tamada, K.; Hara, M.; Sasabe H.; Knoll, W. Langmuir 1997, 13,
1558.
(15) Ishida, T.; Yamamoto, S.-I.; Mizutani, W.; Motomatsu, M.;
Tokumoto, H.; Hokari, H.; Azehara, H.; Fujihira, M. Langmuir 1997, 13,
3261.
(16) Hobara, D.; Sasaki, T.; Imabayashi, S.; Kakiuchi, T. Langmuir 1999,
15, 5073.
(17) Nishida, N.; Hara, M.; Sasabe, H.; Knoll, W. Jpn J. Appl. Phys.
1997, 36, 2379.
(18) Ishida T.; Mizutani, W.; Akiba, U.; Umemura, K.; Inoue, A.; Choi,
N.; Fujihira, M.; Tokumoto, H. J. Phys. Chem. 1999, B103, 1686.
(19) Mizutani, W.; Ishida, T.; Tokumoto, H.; Jpn. J. Appl. Phys. 1999,
38, 3892.
(20) Ishida, T.; Mizutani, W.; Tokumoto, H.; Choi, N.; Akiba, U.;
Fujihira, M. J. Vac. Sci. Technol. 2000, A18, 1437.
(21) Ishida, T.; Choi, N.; Mizutani, W.; Tokumoto, H.; Kojima, I.;
Azehara, H.; Akiba, U.; Hokari, H.; Fujihira, M. Langmuir 1999, 15, 6799.
(22) Samanta, M. P.; Tian, W.; Datta, S.; Henderson, J. I.; Kubiak, C.
P. Phys. ReV. 1996, B53, R7626.
(23) Magoga, M.; Joachim, C. Phys. ReV. 1999, B59, 16011.
(24) Tao, Y.-T.; Wu, C.-C.; Eu, J.-Y.; Lin, W.-L.; Wu, K.-C.; Chen, C.
Langmuir 1997, 13, 4018.
conjugated molecules. Our experimental data leads to the
following conclusions.
(i) The measured height of the conjugated molecular domains
depended on their lateral sizes, in the case of conjugated
molecules with a methylene group between the sulfur and the
phenyl rings. By analyzing size dependence on the height of
the conjugated molecular domain, we could estimate the
electronic conductivity of the molecular domains. To increase
the vertical conduction of molecular domains, a methylene group
is necessary between the sulfur and aromatic rings. The obtained
single molecular resistances are in the order TP3 > TP1 > C9
g BP1. In the cases of TP0 and BP0 without any methylene
group, strong dependences were not observed, suggesting that
the increase in the vertical conduction of the molecular domain
is not so strong.
(ii) The LBH values on the conjugated molecular domains
in all the cases were larger than that on the C9 surface. The
LBH values were not dependent on the domain size, while the
height difference changed with the domain size in the case of
conjugated molecules such as TP1. These data suggest that the
increase in the vertical conductivity is not due to the lowering
of the local barrier height, but can likely be attributed to the
conjugated molecular adsorption.
(iii) The XPS C(1s) satellite peaks and UPS spectra of TP1
and TP3 are almost identical, indicating that the carrier densities
among these SAMs are almost the same. If the vertical
conduction increase with the decrease in the number of
methylene groups is attributed to the doping effect, then the
shape and intensity of both the satellite peaks and valence bands
should be different. Thus, the reason for the higher electrical
conduction might be due to the higher carrier transfer probability
through the conjugated molecules with one methylene group.
(25) Sabatani, E.; Cohne-Boulakia, J.; Bruening, M.; Rubinstein, I.
Langmuir 1993, 9, 2974.
(26) Himmel, H.-J.; Terfort, A.; Wo¨ll, Ch. J. Am. Chem. Soc. 1998,
120, 12069.
(27) [1,1′:4′,1′′-terphenyl]-4-propanethiol 1H NMR (400 MHz, CDCl3)
δ 1.39 (t, 1H), 1.99 (m, 2H) 2.58 (q, 2H) 2.79 (t, 2H) 7.26-7.67 (m, 13H)
13C NMR δ 24.02 (CH2) 33.59 (CH2) 35.42 (CH2) 127.01-127.46 (benzene
ring) 128.79 (ring), 128.86 (ring), 138.27 (ring) 139.88-140.71 (ring).
(28) Sasaki, M.; Komai, M.; Ozawa, R.; Yamamoto, S. Jpn. J. Appl.
Phys. 1998, 37, 6186.
(29) Poirier, G. E.; Tarlov, M. J. Langmuir 1994, 10, 2853.
(30) Delamarche, E.; Michel, B.; Gerber, C.; Anselmetti, D.; Guntherodt,
H.-J.; Wolf, H.; Ringsdorf, H. Langmuir 1994, 10, 2869.
(31) Wo¨ll, Ch.; Chiang, S.; Wilson, R. J.; Lippel, P. H. Phys. ReV. 1989,
B39, 7988.
(32) Sellers, H.; Ulman, A.; Shindman, Y.; Eilers, J. E. J. Am. Chem.
Soc. 1993, 115, 9389.
Acknowledgment. This work was supported by the New
Energy and Industrial Development Organization (NEDO) of
Japan. We thank Hodogaya Contact Lab. for synthesizing some
useful conjugated molecules. We gratefully acknowledge Drs.
I. Kojima and N. Fukumoto (National Institute of Materials and
Chemical Research, Tsukuba, Japan) for their helpful sugges-
tions and experimental assistance in the XPS measurements.
We also thank Dr. S. P. Jarvis (JRCAT-NAIR) for her help in
English correction.
(33) Poirier, G. E. Langmuir 1997, 13, 2019.
(34) Hara, M.; Sasabe, H.; Knoll, W. Thin Solid Films 1996, 273, 66.
(35) Bain, C. D.; Troughton, E. B.; Tao Y.-T.; Evall, J.; Whitesides, G.
M.; Nuzzo; J. Am. Chem. Soc. 1989, 111, 321.
(36) Nakanishi, S.; Tsukada, M. Jpn. J. Appl. Phys. 1998, 37, 3805.
Nakanishi, S.; Tsukada, M. Jpn. J. Appl. Phys. 1998, 37, L1400. Nakanishi,
S.; Tsukada, M. Surf. Sci. 1999, 438, 305.
(37) Zeppenfeld, P.; Lutz, C. P.; Eigler, D. M. Ultramicroscopy 1992,
42-44, 128.
(38) Campbell, I. H.; Rubin, S.; Zawodzinski, T. A.; Kress, J. D.; Martin,
R. L.; Smith, D. L.; Barashkov, N. N.; Ferraris, J. P. Phys. ReV. 1996, B54,
R14321. Campbell, I. H.; Kress, J. D.; Martin, R. L.; Smith, D. L. Appl.
Phys. Lett. 1997, 71, 3528.
References and Notes
(1) Carter, F. L. Molecular Electric DeVices; Marcel Dekker: New
York, 1982.
(2) Iijima, S. Nature 1991, 354, 56.
(3) Tour, J. M.; Jones, L., II.; Pearson, D. L.; Lamba, J. J.; Burgin, T.
P.; Whitesides, G. M.; Allara, D. L.; Parkh, A. N.; Atre, S. V. J. Am. Chem.
Soc. 1995, 117, 9529.
(39) Zehner, R. W.; Parsons, B. F.; Hsung, R. P.; Sita, L. R. Langmuir
1999, 15, 1121.
(40) Narioka, S.; Ishii, H.; Yoshimura, D.; Sei, M.; Ouchi, Y.; Seki,
K.; Hasegawa, S.; Miyazaki, T.; Harima, Y.; Yamashita, K. Appl. Phys.
Lett. 1995, 67, 1899.
(41) Benning, P. J.; Poirier, D. M.; Ohno, T. R.; Chen, Y.; Jost, M. B.;
Stepniak, F.; Kroll, G. H.; Weaver, J. H.; Fure, J.; Smalley, R. E. Phys.
ReV. 1992, B45, 6899.
(42) Enkvist, Ch.; Lunell, S.; So¨jgren, B.; Bru¨hwiler, P. A.; Svensson,
S. J. Chem. Phys. 1997, 103, 6333.
(4) Bumm, L. A.; Arnold, J. J.; Cygan, M. T.; Dunbar, T. D.; Jones,
L., II.; Allara, D. L.; Tour, J. M.; Weiss, P. S. Science 1996, 271, 1705.
Cygan, M. T.; Dunbar, T. D.; Arnold, J. J.; Bumm, L. A.; Shedlock, N. F.;
Burgin, T. P.; Jones, L., II.; Allara, D. L.; Tour J. M.; Weiss P. S. J. Am.
Chem. Soc. 1998, 120, 2721.
(5) Leatherman, G.; Durantini, E. N.; Gust, D.; Moore, T. A.; Moore,
A. L.; Stone, S.; Zhou, Z.; Lez, P.; Liu, Y. Z.; Lindsay, S. M. J. Phys.
Chem. 1999, B103, 4006.
(43) Onoe, J.; Nakao, A.; Takeuchi, K. Phys. ReV. 1997, B55, 10051.
(44) Bain, C. D.; Whitesides, G. M. J. Phys. Chem. 1989, 93, 1670.
(45) Chen, J.; Reed, M. A.; Rawlett, A. M.; Tour, J. M. Science 1999,
286, 1550.
(46) Seminario, J. M.; Zacarias, A. G.; Tour, J. M. J. Am. Chem. Soc.
2000, 122, 3015.
(6) Kelly, K. F.; Shom, T.-S.; Lee, T. R.; Halas, N. J. J. Phys. Chem.
1999, B103, 8639.
(7) Kergueris, C.; Bourgoin, J. P.; Palacin, S. Nanotechnology 1999,
10, 8.
(8) Salmeron, M.; Neubauer, G.; Folch, A.; Tomitori, M.; Ogletree,
D. F.; Sautet, P. Langmuir 1993, 8, 3600.