University of Groningen
Impact of derivatization on electron transmission through dithienylethene-based
photoswitches in molecular junctions
Van Dyck, Colin; Geskin, Victor; Kronemeijer, Auke J.; de Leeuw, Dago M.; Cornil, Jerome;
Cornil, Jérôme
Published in:
Physical Chemistry Chemical Physics
DOI:
10.1039/c3cp44132f
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to
cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date:
2013
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Van Dyck, C., Geskin, V., Kronemeijer, A. J., de Leeuw, D. M., Cornil, J., & Cornil, J. (2013). Impact of
derivatization on electron transmission through dithienylethene-based photoswitches in molecular
junctions. Physical Chemistry Chemical Physics, 15(12), 4392-4404. DOI: 10.1039/c3cp44132f
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the
author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the
number of authors shown on this cover page is limited to 10 maximum.
Download date: 17-06-2017
PCCP
View Article Online
PAPER
Downloaded by University of Groningen on 04 March 2013
Published on 30 January 2013 on http://pubs.rsc.org | doi:10.1039/C3CP44132F
Cite this: Phys. Chem. Chem. Phys., 2013,
15, 4392
View Journal | View Issue
Impact of derivatization on electron transmission
through dithienylethene-based photoswitches in
molecular junctions†
Colin Van Dyck,*a Victor Geskin,a Auke J. Kronemeijer,b Dago M. de Leeuwcd and
Jérôme Cornila
We report a combined Non-Equilibrium Green’s Function – Density Functional Theory study of molecular
junctions made of photochromic diarylethenes between gold electrodes. The impact of derivatization of the
molecule on the transmission spectrum is assessed by introducing: (i) substituents on the diarylethene core;
and (ii) linker substituents between the gold surface and the diarylethene. We illustrate that substituents on
Received 20th November 2012,
Accepted 24th January 2013
the core shift considerably the HOMO/LUMO level energies in gas phase but do not change the I–V
DOI: 10.1039/c3cp44132f
the transmission spectrum and interfacial electronic reorganization upon chemisorption. In contrast, the
different linker substituents under study modulate the conductivity of the junction by changing the degree of
www.rsc.org/pccp
orbital hybridization with the metallic electrodes and the degree of orbital polarization.
characteristics of the molecular junctions; this behaviour has been rationalized by establishing links between
A. Introduction
Since the first concept of a molecular rectifier, published by Aviram
and Ratner in the seventies,1 the field of molecular electronics has
kept challenging researchers at both the applied and fundamental
levels.2–5 The last few decades have experienced the development of
experimental setups to measure the electrical conductivity of
devices at the molecular scale and of theoretical tools to predict
I–V characteristics of molecular junctions. This endeavor stems
from the prospects of moving microelectronics to the realm of
nanoelectronics. Another key feature in dealing with molecules is
the ability to tune the I–V characteristics by exploiting the huge
versatility of chemical synthesis.
The idea of introducing molecular photoswitches between
electrodes to generate functional molecular junctions recently
a
Laboratory for Chemistry of Novel Materials, University of Mons,
Place du Parc 20, B-7000 Mons, Belgium. E-mail: [email protected];
Fax: +32 (0)65 37 38 61; Tel: +32 (0)65 37 38 62
b
Optoelectronics Group, Cavendish Laboratory, University of Cambridge,
J. J. Thomson Avenue, Cambridge CB3 0HE, UK. E-mail: [email protected];
Fax: +44 01223 337706; Tel: +44 01223 365262
c
Zernike Institute for Advanced Materials, University of Groningen, Nijenborgh 4,
9747 AG, Groningen, The Netherlands. E-mail: [email protected];
Tel: +31 50 363 8750
d
Philips Research Laboratories, High Tech Campus 4, 5656 AE, Eindhoven,
The Netherlands. E-mail: [email protected]
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
c3cp44132f
4392
Phys. Chem. Chem. Phys., 2013, 15, 4392--4404
Fig. 1 Chemical structures of the dithienylethene derivatives studied. (Top)
photochromic core switching from the closed isomer, with a conjugated pathway
connecting the linkers R 0 , to the open isomer disrupting the conjugation between
the thiophene units. (Bottom) nature of the different substituents added in the
center of the photochromic core (R) and of the thiolated units used at the
terminal positions (R 0 ) to bind the molecule to the gold electrodes.
emerged.6,7 Prototypical molecules are diarylethene derivatives
that can adopt two different structures, i.e. a closed and open
isomer, interconverting under the action of light at different
wavelengths (see Fig. 1). The two isomers have comparable
sizes but differ in the length of their conjugation paths (longer
for the closed isomer), which promote different electronic gaps and
transmission properties, and hence electrical conductivities. Moreover, their thermal stability makes them attractive candidates in the
This journal is
c
the Owner Societies 2013
View Article Online
Downloaded by University of Groningen on 04 March 2013
Published on 30 January 2013 on http://pubs.rsc.org | doi:10.1039/C3CP44132F
Paper
PCCP
field of molecular electronics.8–10 Recent experimental studies
based on dithienylethenes chemisorbed either on gold or
carbon nanotube electrodes have demonstrated a switching
effect and its reproducibility.11–19 However, further improvements
in the reversibility of the photochromic effect in the junction and in
the current ratios between closed and open isomers are required.20
This motivates theoretical studies aiming at a molecular description
of the switching effect and at the definition of the key parameters
governing the current ratios. Theory further allows for a fundamental understanding of I–V characteristics on well-defined molecular architectures in contrast to many experimental measurements
lacking details about the exact contact geometry or the number of
molecules in the junction.
Several theoretical papers have previously studied dithienylethene-based molecular junctions, most often with gold electrodes,
addressing in particular the origin of conductivity switching and
the ratio between the closed and open isomers, as well as the
reversibility of the photoswitching associated with quenching
processes by the electrodes. All these transport studies are based
on the coherent Landauer’s picture of conductivity (described by
transmission probabilities)21 and the Non-Equilibrium Green’s
Function (NEGF) method5,22–26 combined with Density Functional
Theory (DFT) or a tight-binding approximation.
Earlier studies27–31 were either limited to equilibrium (zero
bias, no I–V curves) or to a bias applied not self-consistently,
that is, the electronic structure of the entire junction is not
re-optimized for every applied voltage. However, they pinpointed
the principal differences between the two forms in the junction:
closer alignment to the Fermi level of the electrodes and a higher
delocalization of the HOMO level in the closed isomer thus
yielding a better conductivity. Maximum current ratios around
22 have been reported for the unsubstituted dithienylethene core
with the sulfur atom of the thiol group anchored on a top
position over the Au(111) surface; an increase up to 57 has been
obtained by attaching laterally a thienylthiol unit.27,28
More recently, fully self-consistent studies with gold,32–34
silver and nickel35 metal electrodes appeared, leading
to improved I–V curves. Generally confirming the previous
findings, they showed in particular that the polarization of
the orbitals by the electric field is a key ingredient to be also
taken into account.33,34 This polarization effect has been
demonstrated by examining the shape of the molecular orbitals
of isolated molecules under the influence of an electric field
(without coupling to open electrodes)33 and via a Projected
Density Of States (PDOS) analysis that does not allow for the
visualization of the orbitals within the junctions.34 It has been
suggested that the orbitals of the open isomer are more
polarizable under bias than those of the closed form, leading
to a decrease in the transmission probability of the former with
increasing bias.
The photoreversibility critically depends on the nature of the
lateral anchoring group (R 0 in Fig. 1) attached to the photochromic core and on the nature of the electrodes (such as gold,
carbon nanotubes, silver, or nickel). The switching mechanism
is, however, not affected by the anchoring groups or the
electrodes.35–37 For gold electrodes, the irreversibility of the
This journal is
c
the Owner Societies 2013
switching event has been attributed to a strong hybridization of
the molecular orbitals with the gold orbitals.30,31 In the case of
carbon nanotubes37 and graphene38 electrodes, this coupling
has been shown to reduce the excited-state lifetime and hence
the efficiency of the photoswitching.
In a study connecting the molecule to finite gold clusters,30
the width of the transmission peak at the Fermi level has been
correlated to the amount of mixing between the HOMO level
and the gold orbitals, thus supporting the role of the orbital
hybridization in the quenching process.31
In the present study, we report a comprehensive fully
self-consistent analysis of the transport properties of several
dithienylethene derivatives, in their closed and open forms,
inserted between two gold electrodes in well-defined geometries.
Our results further allow us also to speculate on the reversibility of
photoswitching between the two forms in the junction. We analyze
separately and systematically the impact of different substituents
on the ethylene bridge (R) and of different lateral linkers (R0 ) on
the calculated I–V characteristics (see Fig. 1). Six different substitution patterns of the photochromic core have been considered in
order to modulate the gap and the absolute energies of the
electronic levels via inductive/mesomer effects or changes in the
amplitude of the torsion angles for the open isomer.39,40 The three
lateral thiolated linkers involved in our study (a thiophene ring
and a benzene ring with meta and para connections) are those
extensively found in the literature.11,12,15,18,31 The large set of
substituents allows us to compare the evolution of the electronic
properties of the photochromic core in the isolated state and
within the junctions and to assess the nature of the linker which
optimally protects the central responsive unit by reducing the
amount of hybridization with the gold electrodes. A critical
comparison is made with corresponding experimental data
indicating that the switching process is reversible with a metaor para-phenylene linker and irreversible with a thiophene
moiety.11,12,15 Moreover, in order to link the electronic structure
of the compounds in the junctions to the calculated transmission spectra, we visualize the molecular orbitals of the contacted
molecules using the Molecular Projected Self-consistent
Hamiltonian (MPSH) approach. The MPSH approach illustrates
that some orbitals get strongly polarized in a molecular junction under bias and that the orbital shape plays a crucial role in
defining the I–V characteristics. Moreover, we find here that the
degree of polarization depends on the nature of the R 0 linker
substituent. Interestingly, in spite of large variations in the
HOMO energy of the different compounds substituted on their
central photochromic core, the HOMO is aligned similarly with
respect to the Fermi level of the gold electrodes for all compounds, thus pointing to a Fermi level pinning. This teaches us
that designing the alignment of electronic levels in molecular
junctions by the mere consideration of isolated molecular units
could prove highly misleading due to the neglect of the interfacial electronic processes.
The paper is structured as follows: after a brief presentation
of the theoretical methodology in Section B, we discuss the
properties of the isolated molecules and of the junctions at
equilibrium and under bias in Section C. For the sake of clarity,
Phys. Chem. Chem. Phys., 2013, 15, 4392--4404
4393
View Article Online
PCCP
Paper
we distinguish between the impact of R or R 0 substitutions in
each section. Conclusions and perspectives are outlined in
Section D.
Downloaded by University of Groningen on 04 March 2013
Published on 30 January 2013 on http://pubs.rsc.org | doi:10.1039/C3CP44132F
B. Computational methods
We have first optimized the ground-state geometry of each
compound in its two isomeric forms at the DFT level using the
widely used B3LYP hybrid functional41 and a 6-31G** basis set,
as previously done with organic compounds42 especially diarylethenes.40,43 The calculations have been performed with the
Gaussian03 package.44 The electronic structure and transmission
spectrum of the gold-molecule junctions have been calculated using
the Non-Equilibrium Green’s Function formalism based on the DFT
method, as implemented in the ATK2008.10 package.22,25 The
exchange–correlation is treated with the GGA.revPBE functional.45,46
In all junctions, the anchoring geometry was fixed, with the
sulfur atoms anchored on the top position on the gold surface
with an Au–S distance fixed at 2.42 Å, as suggested experimentally
and theoretically,47–49 in order to focus on the impact of the
molecular internal structure on conductivity. More computational
details can be found in the ESI.†
The electronic structure of the molecules within the junction
has been characterized by MPSH analysis,25 which consists in
diagonalizing the Hamiltonian of the scattering region in an
LCAO basis restricted to a chosen set of atoms of the central
region. The MPSH spectrum provides valuable information
about the alignment of the molecular levels with respect to
the Fermi level of the electrodes and helps to rationalize the
nature and intensity of the peaks in the transmission spectra.
The charge transfer occurring upon covalent bonding to the
gold surface has been estimated by computing the x,y planeaveraged density profiles of three subsystems: the total system
(rall), the total system without the molecule (rgold), and the total
system without the two interfacial gold layers on each side of
the molecular unit (rmol).50 The external layers are kept to
ensure the continuity between the scattering region and the
semi-infinite electrodes. We then evaluate the dr profiles using
dr(z) = rall(z) (rmol(z) + rgold(z)). The difference in charge
density arises from the electronic reorganization due to the gold–
molecule–gold interactions. On that basis, we can also build-up the
Rz
interfacial charge transfer by integration: QðzÞ ¼ 0 drðz0 Þdz0 .
C. Results and discussion
a Isolated molecules
We will describe here separately the impact of the core and
lateral substituents on the electronic structure of the photochromic core. For the core substituents, we have introduced fluorine
atoms, cyano and methyl groups on the central double bond as
well as hydrogenated or fluorinated saturated five-carbon rings.
The thiolated linkers introduce an additional thiophene ring or
phenylene ring with a meta/para connection, see Fig. 1.
We focus on the properties of the frontier MOs, as one of them
(HOMO or LUMO, depending on level alignment in the junction)
4394
Phys. Chem. Chem. Phys., 2013, 15, 4392--4404
usually dominates transmission and the chemical potential of the
molecules, defined as the energy at midgap.51 In the junction, the
alignment of the molecular electronic levels with the Fermi level of
the electrodes will be driven by an evolution towards equalization
of the chemical potentials of the two components. However, this
equalization is apparently not necessarily achieved at equilibrium
depending on the actual density of states available.52,53 We will
show that the changes due to derivatization can be considerable
and will discuss the corresponding energy level alignment in the
following sections. The actual position of the frontier electronic
levels of the closed and open isomers in gas phase was speculated
to be a key parameter defining the properties of dithienylethenebased molecular memory devices.54
Core substituents. The DFT calculations show that the backbone of all closed isomers is planar, in agreement with previous
studies.39,40 In contrast, the open isomers exhibit a significant
dihedral angle between the planes of the two thiophene rings,
varying as a function of the nature of the core substituent
(Table 1). In comparison to RQH, the largest distortion is
observed for RQCH3, where the steric hindrance cannot be
accommodated by a strong reduction of the bond angle associated
with the carbon atoms 2, 3, and 4 of the ethylene bridge (see
labelling in Fig. 1), thus translating into a torsion angle increased
by about 10 degrees.
The B3LYP shapes of the HOMO acting as a transport
channel in the junction (plotted with an isovalue = 0.02) and
frontier orbital energies are reported in ESI† (Fig. S2). We
observe that the pattern of the HOMO remains unchanged
upon derivatization of the photochromic core or by changing
the torsion angle (RQCH3). We can directly relate the pattern of
the HOMO level to that of octatetraene and hexatriene for the
closed and open isomers which are displayed in Fig. 2.39 We
also observe that HOMOs do not bear a significant weight on
the substituents. While the substituents do not modify the
shape of the HOMOs, they significantly alter the absolute
energies of the frontier molecular orbitals and the HOMO–
LUMO energy gap (see Table 1 and Fig. 2). When taking the
RQH compound as reference, the introduction of fluorine
atoms stabilizes the frontier electronic levels for both isomers,
with a more pronounced effect for the Fcycle. This stabilization
Table 1 Geometric parameters of the open isomer substituted on the photochromic core, as calculated at the B3LYP level. The torsion angle characterizing
the loss of coplanarity of the thiophene rings in the open isomer is defined from
the vectors connecting atoms 1–3, 3–4, and 4–6 in Fig. 1. The C1–C6 is the
distance between atoms 1 and 6. The bond angle involves atoms 2, 3, and 4 of
the ethylene unit. In the last four columns, we report the shift of the frontier
electronic levels upon substitution, with RQH chosen as a reference
Bond Closed
Torsion C1–C6 angle HOMO
Substituent (1)
(Å)
(1)
(eV)
Closed Open Open
LUMO HOMO LUMO
(eV)
(eV)
(eV)
H
F
CN
CH3
Fcycle
Hcycle
0
0.22
1.04
+0.15
0.78
+0.07
46
47
51
57
50
48
3.6
3.7
3.6
3.8
3.7
3.7
130
131
127
123
130
129
0
0.22
0.78
+0.11
0.52
+0.14
This journal is
c
0
0.14
0.76
+0.06
0.62
+0.14
0
0.28
1.77
+0.36
0.84
+0.16
the Owner Societies 2013
View Article Online
Paper
PCCP
closed HOMO level has no electronic density on the external
C–S bonds that will be attached to the gold surface in the
junction due to the breaking of the electronic communication
introduced by the meta connection. The meta linkers are thus
expected to strongly reduce the coherent transmission through
the molecule due to a poor electronic coupling between gold
and the molecule.55–58
Downloaded by University of Groningen on 04 March 2013
Published on 30 January 2013 on http://pubs.rsc.org | doi:10.1039/C3CP44132F
b
Fig. 2 Energy spectrum and HOMO–LUMO gaps of each substituted photochromic core, as obtained at the DFT level. Values for the closed (left) and open
(right) isomers are displayed. We have also superimposed the shape of the
HOMO level of octatetraene and hexatriene at the B3LYP level reflecting the
HOMO pattern in the closed and open forms of all core-substituted derivatives.
is further enhanced in the presence of cyano substituents
yielding shifts as large as 0.76–0.78 eV for the HOMO levels.
The CH3 and Hcycle substitutions induce a little destabilization
of the HOMO level; the larger destabilization is 0.14 eV in both
the open and closed isomers for the Hcycle substitution. As
expected, the HOMO–LUMO gaps are systematically smaller in
the closed isomers due to the larger delocalization of the
conjugated pathway. The band gap is further modulated by
the introduction of the substituents. Using H-substitution as
reference, the gap for the closed and open CH3-substituted
compounds is increased by 0.04–0.30 eV, respectively, while a
decrease by 0.26–1.01 eV is observed for the closed and open
CN-substituted isomers.
The chemical potential also varies significantly upon derivatization. The values lie in between 4.16 eV and 3.12 eV for
the closed form and 4.70 eV and 3.22 eV for the open form.
Lower values are obtained for fluorinated and CN-substituted
compounds. For a given structure, the chemical potential is
always lower for the open isomer.
Lateral linkers. The introduction of the linkers does not
affect the geometry of the central core. The core is still
characterized by torsion angles between the thiophene rings
around 451 and a C1–C6 distance of 3.6 Å in the open isomer.
The dihedral angles between the linkers and the thiophene
rings of the photochromic core are reasonably small, up to 271
for the meta-phenylene linker. This is expected to ensure a good
delocalization between the linker and the core and hence a
significant conductance across the molecular junctions.31 The
HOMO–LUMO gap is reduced by 0.50–0.72 eV for the closed
form and 0.41–0.85 eV for the open form upon introduction of
the linkers due to the increase in the size of the conjugated
pathway. The range of chemical potentials is much less
extended than for the core substitution, varying between
3.32 eV and 3.55 eV for the closed isomer and between
3.21 eV and 3.50 eV for the open form. The shape of
the HOMO orbitals displayed in ESI† confirms the good delocalization of this level. Interestingly, with the meta-phenylene linker, the
This journal is
c
the Owner Societies 2013
Junctions at equilibrium
The fundamental property of a junction in the coherent transport regime is the transmission probability for an electron
to cross it. In this section, we firstly analyze the way the
transmission spectrum at zero bias, that is at the Fermi level
and in the absence of electric field, is affected by the core and
lateral substituents.
Core substituents. Fig. 3a shows that the transmission
curves are practically identical in the vicinity of the Fermi level
for all substituted molecules in both closed and open forms, in
spite of the significant differences in the frontier orbital
energies in the isolated molecules. Both isomers are characterized
by a wide and intense transmission peak in the vicinity of the
Fermi level. The HOMO peak is perfectly aligned with the Fermi
level in the closed form whereas the transmission peak of the
open isomer is shifted to lower energy such that only its tail is
aligned with the Fermi level. The high transmission intensity (on
the order of 101) and the considerable width of the transmission
peak of the open form are hardly compatible with a simplistic
view of this isomer as a non-conjugated unit; the conjugated
pathway might be affected by cross-conjugation effects. For
comparison, NEGF-DFT calculations on saturated C8 and C12
alkanedithiol self-assembled monolayers (SAMs) (with a thickness
similar to the length of the photochromic cores) between gold
electrodes yield a transmission raising up to 103 at the Fermi
level.59
In order to get a deeper insight into the transmission
spectra, we start by performing the MPSH analysis. The open
RQH isomer is displayed in Fig. 3b. Many MPSH levels are
states concentrated only on gold atoms that do not offer a
pathway for transmission. The relevant levels are those delocalized over the molecular part that we can identify from the
shape of the orbitals computed for the isolated molecules.
The MPSH level associated with the transmission peak of the
closed-H form is very reminiscent of the HOMO level of
the isolated molecule (and will be referred in the following as
the MPSH HOMO). This level acts as the conducting channel in
agreement with previous studies29–34 and exhibits a significant
electron density on all atoms connecting the two electrodes,
thus explaining the high transmission probability. The MPSH
open-H HOMO has a smaller density on the central two carbon
atoms than the MPSH open-H LUMO, which translates into a
higher transmission probability for the latter. Generating an
intense peak of transmission thus requires a good delocalization on all atomic pz orbitals along the conjugated pathway. On
the other hand, a good hybridization between the orbitals of
gold and of the external sulfur atoms is observed in the closedH and open-H HOMO levels whereas no electronic density is
Phys. Chem. Chem. Phys., 2013, 15, 4392--4404
4395
View Article Online
Downloaded by University of Groningen on 04 March 2013
Published on 30 January 2013 on http://pubs.rsc.org | doi:10.1039/C3CP44132F
PCCP
Paper
Fig. 3 (a) Transmission spectra (in log scale) of each photochromic core around the Fermi level (set to zero for convenience) for the closed (left) and open (right)
isomers; (b) transmission spectrum of the open-H isomer in a linear scale. The MPSH spectrum is superimposed in red to highlight the good match between the
transmission peaks and MPSH levels. MPSH orbitals of closed-H HOMO, open-H HOMO, and open-H LUMO are also displayed from top to bottom (isovalue = 0.04).
found at the interfacial region in LUMO open-H. This translates
into a much larger width of the HOMO transmission peak
compared to the LUMO peak in Fig. 3b. A small hybridization
will not yield high conductivities since hybridization plays a
major role in broadening the resonant transmission peaks and
thus in increasing the current upon integration.
The larger hybridization between the gold and molecular
orbitals for the open isomer is consistent with the fact that the
switching process operates only from the closed to open
geometry at the experimental level for dithienyl derivatives11,15
since the excited-state lifetime is expected to be reduced in the
open form due to the strong coupling with the metal. The large
width of the open-H HOMO transmission peak thus generates a
high conductance signal and yields irreversibility of the photochromic process. In this respect, the lateral linkers might act as
anti-quenching groups by reducing the extent of hybridization
and could further increase the ON/OFF ratio of the molecular
junction. The strong hybridization between the gold and molecular orbitals also promotes a Fermi level pinning, as reflected
by the fact that the energy difference between the HOMO
level and the Fermi level is almost the same for all derivatives,
4396
Phys. Chem. Chem. Phys., 2013, 15, 4392--4404
i.e. 0.13–0.14 eV for the closed isomers and 0.19–0.25 eV for the
open isomers. This pinning originates from charge-transfer
processes between the molecule and the electrodes and implies
that the amount of charge transferred varies with the nature of
the substituents.
In order to visualize the charge transfer, we have computed
the dr profiles for several closed isomers, according to the
procedure described in the methodology section, together with
the charge integrated over the molecular part, see Fig. 4. The
charge transfer primarily occurs at the interface. The surface
gold layers lose a significant fraction of electrons, as evidenced
by the dip in the density between the first two reference lines,
whereas the sulfur gains some negative charge though less than
the neighboring carbon atom in comparison to the biradical.
This reflects that the charge transfer from the carbon to the
sulfur within the C–S bonds of the biradical is partially cancelled upon creation of the gold–sulfur bond. The charge
distribution in the central molecular part is similar for all
closed isomers. The molecular backbone as a whole gains
electrons, with the excess charge integrated between the two
external carbon atoms on the order of 0.10–0.14|e|. The results
This journal is
c
the Owner Societies 2013
View Article Online
Downloaded by University of Groningen on 04 March 2013
Published on 30 January 2013 on http://pubs.rsc.org | doi:10.1039/C3CP44132F
Paper
PCCP
Fig. 4 (Left) a plane-averaged profile of dr due to the electronic reorganization
upon molecular adsorption. The three vertical reference lines correspond to the
position of the gold surface layer, the chemisorbed sulfur atom and the carbon
atom bonded to the sulfur. (Right) schematic explanation of the alignment
process. The blue arrow depicts the shift of the electronic potential energy Uel
in the central region induced by the bond dipoles on both sides and felt by the
molecular levels.
demonstrate that the Fermi level pinning is established by the
creation of interface dipoles whose amplitude varies with the
nature of the substituent. The two bond dipoles create an
upward shift of the electrostatic potential at the interfaces.54
The electronic density localized over the central molecular
backbone experiences this shift, leading to a destabilization
of the HOMO level with a shift depending on the strength of the
dipoles balancing the variation of the chemical potential
among the derivatives (see the scheme in Fig. 4).
Strictly speaking, the chemical potentials of the neutral
molecules cannot be directly confronted to the chemical
potential of gold (5.26 eV at both the experimental and theoretical levels50,60) since the corresponding value for the biradical
(without the terminal H atoms) coupled to the gold surface
matters in the junction.61 Still, it is reasonable to assume that
the relative values of the chemical potentials will be similar
for the isolated molecules and for biradicals hybridized systematically in the same way with the gold surface. A similar
scenario prevails for the open isomers. It should be noted that
the amplitude of the interface dipole might also be affected by
the anchoring position of the sulfur atom on the gold surface.
This is, however, beyond the scope of the present paper focusing on the amplitude of the ON/OFF ratios of photoisomers in
the same contact geometry. The calculated interface dipole
does depend on core substitution: it grows along the same
order CN > Fcycle > H (Fig. 4) as the chemical potentials
calculated for the isolated molecules (Section C.a: 4.16 eV for
closed-CN > 3.90 eV for closed F-cycle > 3.25 eV for closed-H).
Larger interface dipoles for the molecules with lower lying
frontier MOs lead to increased destabilization of these levels in
the junction, thus counteracting the initial level positioning.
This mechanism dominates the metal–molecule level alignment
and is responsible for the similar alignment of the series of
molecules with different frontier MO energies.
Lateral linkers. We analyze here the impact of introducing
three aromatic linkers between the photochrome and the
anchoring thiol unit on the properties discussed in the previous
This journal is
c
the Owner Societies 2013
section. The transmission spectra of the linker-substituted
photochromic cores are reported in Fig. 6a. The meta-phenylene
linker has clearly a distinct behaviour, yielding an extremely
narrow transmission peak at the Fermi level for the closed
isomer, in agreement with previous theoretical studies on
benzene-dithiol junctions.55–58 The narrowness of this peak
originates from the lack of hybridization between the gold and
the frontier molecular p orbitals and gives rise to a negligible
current across the junction. For the other two linkers, the
transmission peaks at the Fermi level are quite similar to
closed-H. The linker substitution modifies significantly the
intensity of the transmission peak of the open-H isomers: for
the para-phenylene and thienyl linkers, the intensity decreases
while it increases in the presence of the meta-phenylene linker.
In the latter case, the changes can be detrimental for the current
switching performance if the intense peak enters in the transmission window under bias. For all linker-substituted derivatives, we also observe a second intense transmission peak due to
the reduction in the energy difference between the HOMO and
HOMO 1 levels of the isolated molecules following the
increase in the length of the conjugated path. This second peak
is not expected to contribute to the I–V curve since it lies far away
from the Fermi level (more than 0.5 eV).
The MPSH spectra confirm that it is again the HOMO level
of the three compounds which is aligned with the Fermi level.
For the closed isomers, the introduction of the para-phenylene
and thienyl linkers preserves the alignment of the HOMO level,
lying at 0.13 eV below the Fermi level. As seen in Fig. 6b, these
levels are well delocalized throughout the backbone, with a
significant weight on the terminal sulfur atoms and the gold
surface. This description is similar to that prevailing for the
compounds without linkers and explains the little changes in
the transmission spectra at the Fermi level. For the open
isomers, the thienyl and para-phenylene linkers exhibit a
HOMO aligned in the same range of energy compared to the
results obtained in the previous section (0.22 eV and 0.28 eV
below the Fermi level, respectively). The alignment energy at
exactly the same value for the reference RQH compound and
that featuring a thiophene-based linker is rationalized by the
fact that the interfacial hybridization and charge reorganization
are expected to be very similar in both cases due to the identical
anchoring groups.
Since the charge transfer between the gold surface and the
molecule was found in the previous section to mostly occur at
the interface, it is likely that the anti-quenching groups will act
as a shield against electronic reorganization in the photochromic
core. When plotting the plane-averaged profile of dr (see Fig. 5),
we do observe a reduction in the amplitude of the oscillations over
the central core, thus pointing to a limited electronic reorganization. The excess charge on the photochromic core is about 0.05–
0.07|e| after integration with the thienyl and para-phenylene
linkers and is lower compared to the unsubstituted photochromic
core (0.10–0.14|e|). The amplitude of the interface dipole does not
correlate with the chemical potential of the isolated molecules
since each linker hybridizes in a different way with the gold
surface. As expected from the identical nature of the anchoring
Phys. Chem. Chem. Phys., 2013, 15, 4392--4404
4397
View Article Online
Downloaded by University of Groningen on 04 March 2013
Published on 30 January 2013 on http://pubs.rsc.org | doi:10.1039/C3CP44132F
PCCP
Paper
open isomers, respectively. This transfer is equivalent to that
occurring with the other linkers for the closed isomer but is
reduced among the open forms.
The degree of hybridization and coupling between the
molecule and gold can be indirectly assessed from the widths
of the transmission peaks: the larger the width, the larger the
degree of coupling, and hence the smaller the lifetime of the
excited state on the molecular backbone. In turn, a reduced
lifetime will lower the probability for the geometric reorganization in the excited state to occur. The Full Width at Half
Maximum is directly connected to the interfacial coupling
strength for a simple model based on a single resonant transport level.5 FWHM of the transmission peaks reported in Fig. 6
is collected in Table 2. The determined FWHM values demonstrate that the nature of the linker strongly influences the width
of both the HOMO and LUMO transmission peaks. These peaks
are the most relevant since the lowest optical transition in
diarylethene compounds primarily involves an electronic transition between the HOMO and LUMO levels.31,37 The LUMO
peak of the open form is always rather narrow with or without
linkers. In contrast, the width of the LUMO peak in the closed
form is strongly reduced with linkers, which are thus expected
to facilitate the closed to open switch in the junction by
increasing the lifetime of the excited state. All together, we
conclude that the meta-phenylene linker is the best candidate
among the three to ensure the reversibility of the mechanism of
photoswitching.
Fig. 5 Plane averaged profiles of dr due to the electronic reorganization upon
adsorption for the closed (up) and open (down) isomers. The two vertical
reference lines correspond to the position of the interfacial gold layer and the
chemisorbed sulfur atom. The photochromic core part extending between the
carbon atoms bonded to the linkers is highlighted in red.
unit, the electronic reorganization upon adsorption is similar
for the reference RQH compound and upon introduction of
the thienyl linker.
In the presence of the meta-phenylene linker, the MPSH
HOMO of the closed form does not bear a weight on the sulfur
atoms and is aligned differently with the Fermi level (at 0.06 eV).
The first observation readily explains why the transmission peak is
so narrow and indicates that the linker hybridizes in a different way
with the gold surface. For the same reason, the open isomer with
the meta-phenylene linker also aligns differently, i.e. at 0.38 eV
below the Fermi level. It is, however, surprising that the metaphenylene linker promotes an efficient transmission peak in the
open form. This is, however, coherent with the shape of the MPSH
orbital showing a significant weight on the sulfur atoms and
delocalization throughout the conjugated pathway. Fortunately
for the switch, this peak is quite low in energy and is not likely
to enter the transmission window and contributes to a significant
current for the open form. The meta-phenylene linker is the best
shield to protect the central core, as evidenced by lower fluctuations
of the charge density wave over the central part. The excess charge
transferred to the central core delimited by the two carbon atoms
connected to the linkers is 0.05|e| and 0.02|e| for the closed and
4398
Phys. Chem. Chem. Phys., 2013, 15, 4392--4404
c
Junctions under bias
Having considered the properties of our junctions at zero bias
in the previous section, we will now study them under bias,
which opens the way to the simulation of I–V curves. We are
now interested in the transmission probability for an electron
to cross the junction within a transmission energy window
(defined by the applied bias) around the Fermi level. For the
calculation of the full I–V characteristics, it is often considered
that the transmission spectrum calculated at equilibrium (zero
bias) remains in first approximation unaltered when a bias is
applied. Therefore, only this single transmission spectrum is
integrated for different transmission windows to obtain the I–V
curve of a junction. This approach is generally too simplistic
since the polarization of orbitals induced by the applied bias
can significantly modify the electronic properties of the junction, as it will be shown below. Moreover, the pinning of the
molecular orbitals with the most stable Fermi level under bias
cannot be grasped without a self-consistent treatment. In this
section, we discuss the transmission spectra calculated under
bias from 0 V to 1.05 V by steps of 0.15 V, and integrated to
obtain the current flowing across the molecular junctions.
Core substituents. The transmission spectra calculated at
0.90 V are shown in Fig. 7a for both the closed and open
isomers together with the transmission window at 0 K
characterized by Fermi–Dirac step-like statistics. The HOMO
transmission peaks of the closed isomers responsible for the
current keep their shape and intensity. In contrast, the HOMO
peak of the open isomers under bias experiences a significant
This journal is
c
the Owner Societies 2013
View Article Online
Downloaded by University of Groningen on 04 March 2013
Published on 30 January 2013 on http://pubs.rsc.org | doi:10.1039/C3CP44132F
Paper
PCCP
Fig. 6 (a) Transmission spectra of the linker-substituted photochromic core around the Fermi level for the closed (left) and open (right) isomers; (b) orbitals
corresponding to the HOMO of the linker-substituted isomers.
Table 2 Full Width at Half Maximum (FWHM, in meV) of the HOMO and LUMO
transmission peaks for the reference compound RQH and for each linkersubstituted derivative
Peak
H
m-Phenylene
p-Phenylene
Thienyl
Closed HOMO
Closed LUMO
Open HOMO
Open LUMO
180
177
333
o12
o3
12
114
24
147
66
255
42
189
72
228
51
decrease in intensity down to half the value at equilibrium.
Moreover, a fraction of this broad peak is shifted outside the
transmission window due to the pinning of the HOMO level of
the open isomer to the lower Fermi level (i.e., the positive
electrode). As a result, the current through the open isomers
under bias is much lower than through their closed equivalents
and significantly lower than what could have been expected
from their zero-bias transmission spectra. This has a positive
impact on the closed/open current ratio.
In order to shed light on the different behaviour of the
isomers, we also report in Fig. 7b the MPSH orbitals associated
with the main transmission peaks at 0.90 V. The shape of the
MPSH HOMO orbital of the closed form is very reminiscent of
that found at equilibrium and is therefore hardly modified by
the electric field. Strikingly, the MPSH HOMO orbital of the
This journal is
c
the Owner Societies 2013
open form is completely polarized towards one side of the
junction. The loss of delocalization of this orbital explains the
significant drop in intensity of the transmission peak. Furthermore, the strong confinement on one side of the junction
promotes a larger hybridization with the gold surface which
preserves the energy difference between the MPSH HOMO
orbital and the most stable Fermi level under bias (Fermi
pinning). These results clearly demonstrate that a fully selfconsistent treatment is essential to provide a reliable description of the I–V characteristics.
The simulated I–V curves for the different photochromic
cores are reported in Fig. 7c. As already discussed for the
junctions at equilibrium, the different substitution patterns
have nearly no impact on the shape of the I–V curves and hence
on the current ratios. This is due to both: (i) the similar shape
of the conducting HOMO orbital at equilibrium or under bias
when varying the nature of the electroactive substituents or
torsion angles in the closed and open isomers; and (ii) the
Fermi pinning effect which makes the alignment insensitive to
changes in the chemical potential of the compounds. Different
I–V traces would be obtained by modulating in a stronger way
the shape of the HOMO orbital or by changing the chemistry at
the molecule–gold interface as a linker does. The current ratio
decreases with a growing bias because the transmission peak of
the closed form is quickly fully incorporated in the transmission
Phys. Chem. Chem. Phys., 2013, 15, 4392--4404
4399
View Article Online
Downloaded by University of Groningen on 04 March 2013
Published on 30 January 2013 on http://pubs.rsc.org | doi:10.1039/C3CP44132F
PCCP
Paper
Fig. 7 (a) Transmission spectra of the closed and open isomers under a bias of 0.9 V. The zero of energy is set as the average Fermi level of the left and right
electrodes. The horizontal lines depict the transmission window at 0 K. The dashed line corresponds to the transmission of the RQH core at equilibrium; (b) MPSH
states of the RQH photochrome at 0.9 V associated with the transmission peaks at 0.20 eV and 0.68 eV for the closed and the open isomer, respectively; (c) I–V
curves of each substituted photochromic core (left); the current ratio between the closed and open forms as a function of the voltage (right).
window, leading to saturation of the current, in contrast to the
broader transmission peak associated with the open form. As a
consequence, we predict that the low bias regime yields a higher
switching ratio for all photochromic cores without any linker.
However, the strong interfacial hybridization might preclude the
observation of the switching effect, thus motivating the introduction of linkers at the experimental level.
Lateral linkers. Fig. 8a shows the transmission spectra of the
three linker-substituted derivatives under a bias of 0.90 V. The
closed isomers with the para-phenylene and thienyl linkers
behave like the isolated photochromic core, with a small
shift of the transmission peaks associated with the different
hybridization of the phenylthiol with the gold surface. The shift
away from the center of the transmission window is expected to
reduce the current at low bias. The transmission peaks are
always governed by the HOMO level which does not polarize
under bias. For the corresponding open isomers, there is a
strong polarization of the HOMO level under bias which shifts
the level at quite the same energies as for the isolated photochromic cores albeit with a drop in intensity. This picture is
4400
Phys. Chem. Chem. Phys., 2013, 15, 4392--4404
fully corroborated by the MPSH orbitals reported in Fig. 8b. The
decrease in the transmission of the open isomer under bias and
the conservation of a strong transmission for the closed form
thus guarantee a high current ratio.
The situation is radically different for the meta-phenylene
linker. The transmission peak of the HOMO level of the closed
form is extremely narrow, as rationalized in the previous
section, and lies outside the transmission window at 0.90 V.
This translates into a very small current across the junction.
The transmission peak of the open form is also centred
outside the transmission window. In contrast to the other two
linkers, the polarization effects are softened. A high intensity of
the transmission peaks is observed, which is fully supported by
the MPSH analysis that features an orbital still delocalized over
both sides of the junction. The meta-phenylene linker is thus
acting as a shield against charge reorganization due to interfacial hybridization effects and against polarization effects
under bias. However, this linker is not an attractive candidate
for molecular switches since: (i) it creates a large contact
resistance, hence smaller current, by reducing hybridization
This journal is
c
the Owner Societies 2013
View Article Online
Downloaded by University of Groningen on 04 March 2013
Published on 30 January 2013 on http://pubs.rsc.org | doi:10.1039/C3CP44132F
Paper
PCCP
Fig. 8 (a) Transmission spectra of the closed and open linker-substituted isomers under a bias of 0.90 V. The zero of energy is set at the average Fermi level of the left
and right electrodes; (b) illustration of the electronic polarization of the conducting HOMOs of the open isomers with the different linkers under a bias of 0.90 V. The
MPSH states are located in energy at 0.62 eV, 0.74 eV and 0.66 eV for the meta-phenylene, para-phenylene and thienyl linkers, respectively; (c) I–V curves of each
linker-substituted photochromic core (left); current ratio between closed and open forms as a function of the voltage (right).
effects at the interface and; (ii) it reduces the polarization
effects that are favourable for high current ratios.
The previous considerations allow us to readily understand
the I–V curves and I–V ratios plotted in Fig. 8c. The calculated
current for a para-phenylene or thienyl linker is about the same
as that of the reference compound for the closed isomer,
whereas the amplitude is reduced by one order of magnitude
for the open isomers. This translates into ON/OFF ratios of
order 102 that are one order of magnitude larger than for the
photochromic core without a linker. On the other hand, the
meta-phenylene linker generates a current two to three orders
of magnitude lower for both isomers with similar performance
compared to the reference photochromic core. Since the HOMO
level is outside the transmission window with the meta-phenylene linker, the switching mechanism does not rely on
different shapes of the HOMO orbital under bias but rather
on subtle difference in the energy level alignment. The energy
alignment might be further affected by the nature of the
anchoring site and by the choice of the theoretical approach
which is expected to modulate less the extent of polarization
effects. Moreover, the poor performance predicted in the
coherent regime and the strong decoupling might open the
way to other charge transport mechanisms, for instance a
Coulomb blockade regime.
This journal is
c
the Owner Societies 2013
Some experimental results have been published for the
linker-substituted isomers involved in our theoretical
study.11,12,18 In mechanical break junctions based on gold
electrodes, the thienyl-based compound gave a ratio one order
of magnitude higher than in our calculations due to a smaller
measured current in the open form. A similar ratio (B102) is
obtained theoretically and experimentally for the para-phenylene linker anchored on a gold surface and probed by a gold
STM tip, although the calculated current is two orders of
magnitude higher. The meta-phenylene linker has been used
in large area junctions with a gold-dithienylethene–PEDOT/
PSS–gold architecture, leading to a switching ratio of one order
of magnitude. The nature of the interface with the PEDOT–PSS
layer is, however, far from being understood62 and the number
of bad contacts and non-switching molecules in the junction is
difficult to estimate. The lack of a unique experimental protocol
and the ill-defined nature of fabricated junctions make it
very difficult to critically compare theory and experiment at
this stage. This further motivates theoretical calculations on
well-defined architectures to better assist the interpretation of
the experimental curves as well as measurements on several
functional compounds using strictly the same experimental
protocol. However, it is worth stressing that the experimental
and calculated current ratios are generally in good quantitative
Phys. Chem. Chem. Phys., 2013, 15, 4392--4404
4401
View Article Online
PCCP
Paper
agreement and that the calculations do shed light on the
(ir)reversibility of the switching observed experimentally.
Downloaded by University of Groningen on 04 March 2013
Published on 30 January 2013 on http://pubs.rsc.org | doi:10.1039/C3CP44132F
D. Concluding remarks
We have reported transmission spectra and I–V characteristics
calculated at a fully self-consistent level for a series of molecular junctions incorporating dithienylethene photochromic
cores. We have modified the central part by substituents or
the lateral linkers in search of design rules for reversible
molecular switches with a high ON/OFF ratio.
Interestingly, it is the HOMO level that aligns in the same way in
the vicinity of the Fermi level of the electrodes, irrespective of the
chemical potential of the core-substituted isolated molecules that
can be quite different, hence pointing to Fermi level pinning.
Consequently, derivatization does not alter the shape of the
transmission peak responsible for conductivity since the shape of
the HOMO is not modified by the electroactive substituents.
The pinning effect is established via an interfacial charge
transfer from the electrode to the adjacent S–C bond. The
amplitude of charge transfer correlates with the chemical
potential of the isolated molecules. Two bond dipoles are
actually created on each side of the junction shifting up the
central electrostatic potential between the contacts and dictating
the final alignment of the molecular levels with the Fermi level of
the electrodes. Larger interface dipoles for molecules with lowerlying MOs are responsible for the similar energy level alignment of a
series of molecules with different frontier MO energies. On the other
hand, different alignments of the HOMO transmission peak can be
triggered by changing the electronic reorganization at the interface,
for instance by introducing a phenyl-thiol unit in a linker instead of
a thienylthiol unit. The present results thus demonstrate that the
design of molecules for the fabrication of functional molecular
junctions cannot be limited to the sole consideration of the isolated
molecules.
An intense transmission peak requires both good hybridization between the anchoring part of the molecule and the gold
surface and a well delocalized orbital with a significant electronic density all along the backbone (as revealed by MPSH
analysis). These two conditions do not necessarily go hand in
hand: the degree of hybridization of the HOMO was found to be
systematically higher for the lower conducting open isomers.
The introduction of linkers limits the extent of electronic
reorganization over the central photochromic core upon
adsorption. In addition, the meta-phenylene linker breaks the
communication between the core and the gold electrodes,
which is of interest for the reversibility of the switching process
but not beneficial for a high current across the junction.
We suggest that a sensible way to further improve the
current ratio performance of the photoswitches is to modulate
the nature of the linker and the anchoring group or to vary in a
more significant way the molecular topology of the central core
to yield HOMO levels with a distinctively new orbital pattern.
Under bias, the HOMO transmission peaks of the closed and
open isomers behave in a very different way for most of the
derivatives under study. A strong electronic polarization of the
4402
Phys. Chem. Chem. Phys., 2013, 15, 4392--4404
HOMO takes place in the open form, as evidenced by the MPSH
analysis, and drastically reduces the intensity of the transmission peak. On the other hand, the shape of the HOMO level and
hence the intensity of the transmission peak are preserved in
the closed form. This polarization effect is crucial for the actual
ON/OFF current ratio of the junction and cannot be properly
taken into account in calculations that are not fully selfconsistent. The introduction of thienyl or para-phenylene
linkers generates a drop in the intensity of the HOMO transmission peak by one order of magnitude in the open form
leading to a large exaltation of the ON/OFF ratio (B102). A metaphenylene linker, in contrast, shields the open isomer against
polarization and promotes both very small currents and low
ON/OFF ratios. In contrast to the pronounced linker effect, all
substitutions considered for the central photochromic core
yield identical I–V characteristics.
The invariability of the electrical behaviour of the junction
upon chemical modifications of the photochromic molecule
provides freedom to improve properties such as stability or
fatigue resistance without changing the ON/OFF ratios for the
devices.
Acknowledgements
This work has been supported by the European project ONE-P
(NMP3LA-2008-212311), the Interuniversity Attraction Pole
program of the Belgian Federal Science Policy Office, and the
Belgian National Fund for Scientific Research (FRS-FNRS).
C.V.D. and J.C. are FNRS research fellows.
Notes and references
1
2
3
4
5
6
7
8
9
10
11
12
13
14
A. Aviram and M. A. Ratner, Chem. Phys. Lett., 1974, 29, 277.
M. A. Ratner, Mater. Today, 2002, 5(2), 20.
J. Heath and M. A. Ratner, Phys. Today, 2003, 43(5), 43.
A. Salomon, D. Cahen, S. Lindsay, J. Tomfohr, V. Engelkes
and D. Frisbie, Adv. Mater., 2003, 15, 1881.
J. C. Cuevas and E. Scheer, in Molecular electronics: an
introduction to theory and experiment, World Scientific, 2010.
B. Feringa, in Molecular switches, Wiley-VCH, 2001.
S. J. van der Molen and P. Liljeroth, J. Phys.: Condens. Matter,
2010, 22, 133001.
M. Irie, Chem. Rev., 2000, 100, 1685.
H. Tian and S. Yang, Chem. Soc. Rev., 2004, 33, 85.
T. Tsujioka and M. Irie, J. Photochem. Photobiol., C, 2010,
11, 1.
D. Dulic, S. J. van der Molen, T. Kudernac, H. T. Jonkman,
J. J. D. de Jong, T. N. Bowden, J. van Esch, B. L. Feringa and
B. J. van Wees, Phys. Rev. Lett., 2003, 91, 207402.
J. He, F. Chen, P. A. Liddell, J. Andreasson, S. D. Straight,
D. Gust, T. A. Moore, A. L. Moore, J. Li, O. F. Sankey and
S. M. Lindsay, Nanotechnology, 2005, 16, 695.
M. Taniguchi, Y. Nojima, K. Yokota, J. Terao, K. Sato,
N. Kambe and T. Kawai, J. Am. Chem. Soc., 2006, 128, 15062.
N. Katsonis, T. Kudernac, M. Walko, S. J. van der Molen,
B. J. van Wees and B. L. Feringa, Adv. Mater., 2006, 18, 1397.
This journal is
c
the Owner Societies 2013
View Article Online
Downloaded by University of Groningen on 04 March 2013
Published on 30 January 2013 on http://pubs.rsc.org | doi:10.1039/C3CP44132F
Paper
PCCP
15 T. Kudernac, S. J. van der Molen, B. J. van Wees and
B. L. Feringa, Chem. Commun., 2006, 3597.
16 A. C. Whalley, M. L. Steigerwald, X. Guo and C. Nuckolls,
J. Am. Chem. Soc., 2007, 129, 12590.
17 W. R. Browne, T. Kudernac, N. Katsonis, J. Areephong,
J. Hjelm and B. L. Feringa, J. Phys. Chem. C, 2008, 112, 1183.
18 A. J. Kronemeijer, H. B. Akkerman, T. Kudernac, B. J. van
Wees, B. L. Feringa, P. W. M. Blom and B. de Boer, Adv.
Mater., 2008, 20, 1467.
19 F. Meng, Y.-M. Hervault, L. Norel, K. Costuas, C. Van Dyck,
V. Geskin, J. Cornil, H. H. Hng, S. Rigaut and X. Chen,
Chem. Sci., 2012, 3, 3113.
20 W. R. Browne and B. L. Feringa, Chimia, 2010, 64, 398.
21 Y. Imry and R. Landauer, Rev. Mod. Phys., 1999, 71, S306.
22 J. Taylor, H. Guo and J. Wang, Phys. Rev. B: Condens. Matter
Mater. Phys., 2001, 63, 245407.
23 P. A. Derosa and J. M. Seminario, J. Phys. Chem. B, 2001,
105, 471.
24 Y. Xue, S. Datta and M. A. Ratner, Chem. Phys., 2002,
281, 151.
25 M. Brandbyge, J.-L. Mozos, P. Ordejon, J. Taylor and
K. Stokbro, Phys. Rev. B: Condens. Matter Mater. Phys.,
2002, 65, 165401.
26 S. Lindsay and M. A. Ratner, Adv. Mater., 2007, 19, 23.
27 J. Li, G. Speyer and O. F. Sankey, Phys. Rev. Lett., 2004,
93, 248302.
28 G. Speyer, J. Li and O. F. Sankey, Phys. Status Solidi, 2004,
241, 2326.
29 M. Kondo, T. Tada and K. Yoshizawa, Chem. Phys. Lett.,
2005, 412, 55.
30 M. Zhuang and M. Ernzerhof, Phys. Rev. B: Condens. Matter
Mater. Phys., 2005, 72, 073104.
31 M. Zhuang and M. Ernzerhof, J. Chem. Phys., 2009,
130, 114704.
32 J. Huang, Q. Li and H. Ren, J. Chem. Phys., 2007,
127, 094705.
33 A. Staykov, D. Nozaki and K. Yoshizawa, J. Phys. Chem. C,
2007, 111, 3517.
34 A. Odell, A. Delin, B. Johansson, I. Rungger and S. Sanvito,
ACS nano, 2010, 4, 2635.
35 A. Odell, A. Delin, B. Johansson, K. Ulman, S. Narasimhan,
I. Rungger and S. Sanvito, Phys. Rev. B: Condens. Matter
Mater. Phys., 2011, 84, 165402.
36 J. Huang, L. Qunxiang, H. Su and J. Yang, Chem. Phys. Lett.,
2009, 479, 120.
37 M. Ashraf, N. A. Bruque, J. Tan, G. J. O. Beran and
R. K. Lake, J. Chem. Phys., 2011, 134, 024524.
38 C. Motta, M. I. Trioni, G. P. Brivio and K. L. Sebastian, Phys.
Rev. B: Condens. Matter Mater. Phys., 2011, 84, 113408.
39 A. Perrier, F. Maurel and J. Aubard, J. Photochem. Photobiol.,
A, 2007, 189, 167.
40 F. L. E. Jakobsson, P. Marsal, S. Braun, M. Fahlman,
M. Berggren, J. Cornil and X. Crispin, J. Phys. Chem. C,
2009, 113, 18396.
41 C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens. Matter
Mater. Phys., 1988, 37, 785.
This journal is
c
the Owner Societies 2013
42 S. F. Sousa, P. A. Fernandes and M. J. Ramos, J. Phys. Chem.
A, 2007, 111, 10439.
43 A. Goldberg, A. Murakami, K. Kanda, T. Kobayashi,
S. Nakamura, K. Uchida, H. Sekiya, T. Fukaminato,
T. Kawai, S. Kobatake and M. Irie, J. Phys. Chem. A, 2003,
107, 4982.
44 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, J. A. Montgomery, T. Vreven,
K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar,
J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani,
N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima,
Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox,
H. P. Hratchian, J. B. Cross, V. Bakken, C. Adamo,
J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev,
A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski,
P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador,
J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich,
A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick,
A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz,
Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski,
B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz,
I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, Al M.
A. Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe,
P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong,
C. Gonzalez and J. A. Pople, Gaussian 03, Revision C.02,
Gaussian, Inc., Wallingford, CT, 2004.
45 J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett.,
1996, 77, 3865.
46 Y. Zhang and W. Yang, Phys. Rev. Lett., 1998, 80, 890.
47 H. Kondoh, M. Iwasaki, T. Shimada, K. Amemiya,
T. Yokoyama, T. Ohta, M. Shimomura and S. Kono, Phys.
Rev. Lett., 2003, 90, 066102.
48 F. Remacle and R. D. Levine, Chem. Phys. Lett., 2004,
383, 537.
49 A. Perrier, F. Maurel and J. Aubard, J. Phys. Chem. A, 2007,
111, 9688.
50 N. Crivillers, A. Liscio, F. Di Stasio, C. Van Dyck, S. Osella,
D. Cornil, S. Mian, G. M. Lazzerini, O. Fenwick, E. Orgiu,
F. Reinders, S. Braun, M. Fahlman, M. Mayor, J. Cornil,
V. Palermo, F. Cacialli and P. Samorı̀, Phys. Chem. Chem.
Phys., 2011, 13, 14302.
51 R. Parr and W. Yang, in Density-Functional Theory of Atoms
and Molecules, Oxford University Press, 1989.
52 H. Vázquez, R. Oszwaldowski, P. Pou, J. Ortega, R. Pérez,
F. Flores and A. Kahn, EPL, 2004, 65, 802.
53 H. Vázquez, Y. Dappe, J. Ortega and F. Flores, J. Chem. Phys.,
2007, 126, 144703.
54 T. Tsujioka and H. Kondo, Appl. Phys. Lett., 2003, 83, 4978.
55 D. M. Cardamone, C. A. Stafford and S. Mazumdar, Nano
Lett., 2006, 6, 2422.
56 L.-Y. Hsu and B.-Y. Jin, Chem. Phys., 2009, 355, 177.
57 G. C. Solomon, C. Herrmann, T. Hansen, V. Mujica and
M. A. Ratner, Nat. Chem., 2010, 2, 223.
58 T. Markussen, R. Stadler and K. S. Thygesen, Nano Lett.,
2010, 10, 4260.
Phys. Chem. Chem. Phys., 2013, 15, 4392--4404
4403
View Article Online
PCCP
Paper
61 S. Braun, W. R. Salaneck and M. Fahlman, Adv. Mater., 2009,
21, 1450.
62 A. J. Kronemeijer, E. H. Huisman, I. Katsouras, P. A. van
Hal, T. C. T. Geuns, P. W. M. Blom, S. J. van der Molen and
D. M. de Leeuw, Phys. Rev. Lett., 2010, 105, 156604.
Downloaded by University of Groningen on 04 March 2013
Published on 30 January 2013 on http://pubs.rsc.org | doi:10.1039/C3CP44132F
59 T. Frederiksen, C. Munuera, C. Ocal, M. Brandbyge,
M. Paulsson, D. Sanchez-Portal and A. Arnau, ACS Nano,
2009, 3, 2073.
60 D. R. Lide, in Handbook of chemistry and physics, CRC Press,
1995.
4404
Phys. Chem. Chem. Phys., 2013, 15, 4392--4404
This journal is
c
the Owner Societies 2013