Hydroquinone-Benzoquinone Redox Couple as a Versatile Element

Jul 29, 2013 - The possibility of controlling electron transport in a single molecule bridged between two metal electrodes represents the ultimate goa...
0 downloads 0 Views 1MB Size
Article pubs.acs.org/JPCC

Hydroquinone-Benzoquinone Redox Couple as a Versatile Element for Molecular Electronics Paolo Petrangolini,† Andrea Alessandrini,*,†,‡ and Paolo Facci*,‡ †

Dipartimento di Scienze Fisiche, Matematiche ed Informatiche, Università di Modena e Reggio Emilia, Via Campi 213/A, 41125 Modena, Italy ‡ CNR-Istituto Nanoscienze, S3, Via Campi 213/A, 41125 Modena, Italy S Supporting Information *

ABSTRACT: The possibility of controlling electron transport in a single molecule bridged between two metal electrodes represents the ultimate goal of molecular electronics. Molecular electronics aims also at introducing specific properties for the electron transport features both by controlling the structural details of the junction and by exploiting new chemical functionalities. Here we show that, in a molecular junction, where electrodes are represented by a gold substrate and the tip of a scanning tunneling microscope in electrochemical environment, the use of a single molecular species makes it possible to obtain different features for the tunneling current according to the structural details of the junction. In particular, molecules endowed with redox properties brought about by a hydroquinone/benzoquinone redox couple can show both transistor-like and negative differential resistance (NDR) effects. We discuss the mechanistic processes that might describe the different behavior in light of theories of electron transfer between metal electrodes and redox molecules. The results show, on the one hand, the great potential and flexibility that molecular electronics offer and, on the other hand, the need of controlling as much as possible the details of the tunneling junction in order to obtain reproducible results.



INTRODUCTION The main feature of molecular electronics lies in the exploitation of a few or a single molecule, which bridge the gap between two nanoscale electrodes and actively control the flow of electrons in the junction circuit.1,2 To fully exploit all the potentialities of this approach, contributions coming from different fields such as physics, chemistry, engineering, and also biology must join in a common effort. For example, chemical synthesis of specifically tailored molecules according to their conduction properties provides huge possibilities for devising new transport functionalities.3 Among the different functionalities that may be of interest, we can include switching ability (a different conductivity associated with a conformational change or a different redox state of the molecule), transistor effect (the possibility of controlling molecular conduction by a gate electrode), rectification effects (conduction asymmetry for an inversion of the bias voltage), and negative differential resistance (NDR; decrease of the current upon bias voltage increase). It would be a great advantage if a single molecule could provide different functionalities according to the particular situation or environment in which it is involved. A setup of choice for performing experiments in molecular electronics is the scanning tunneling microscope (STM).4 The suitability of this experimental tool to study the transport properties stems from the possibility of including a single molecule in the gap between the tip and the substrate which © 2013 American Chemical Society

play the role of contacting nanoelectrodes. Among the different classes of molecules that have been studied as candidates for molecular electronics applications, a special role is played by redox molecules.5 These molecules transfer electrons by changing their oxidation state and stabilizing a new charge state in the molecule by the interaction with the solvent. In molecular electronics, redox molecules have been proposed and experimentally investigated as transistor elements,5−8 NDR elements,9−11 rectification components, 12 switching elements,6,13,14 and elements with hysteretic and irreversible behavior.15 Their functional role is associated with their oxidation state which usually requires, to be properly controlled, an electrochemical environment. In this case, the STM is substituted for its electrochemical implementation, the electrochemical scanning tunneling microscope (EC-STM).16 Here, the electrochemical potential of the solution plays the role of the gate voltage, allowing a better control of the molecular electronic state with respect to other gating methods by virtue of its negligible electrostatic coupling with source and drain electrodes. Moreover, the electrical double layer effect allows apply strong electric fields directly to the molecules in solution enabling a very effective gating.17 Many experimental Received: June 4, 2013 Revised: July 25, 2013 Published: July 29, 2013 17451

dx.doi.org/10.1021/jp405516z | J. Phys. Chem. C 2013, 117, 17451−17461

The Journal of Physical Chemistry C



Article

EXPERIMENTAL METHODS Sample Preparation. Details of the synthesis of the two molecules used in this work (4-(2′,5′-dihydroxystyryl)benzyl thioacetate (1) and 2-(6-mercaptoalkyl)hydroquinone (2), Chart 1) together with the electrochemical characterization of

and theoretical studies have been performed exploiting redox molecules sandwiched between the tip and the substrate of an EC-STM. These studies demonstrated the possibility of the electrochemical gating effect, in which the electron transport through the molecule is controlled by the electrochemical potential of the solution.5,18−22 According to different situations (symmetric or asymmetric junctions) and depending on different kinds of molecules, a transistor-like effect and a switch effect have been observed. It has been suggested that the latter situation configures a so-called “soft” gating mechanism,22−24 where the molecule bound to both electrodes has a different behavior with respect to molecules in monolayers, the latter being the typical configuration for molecules in an asymmetric junction. In the context of a theoretical investigation of the tunneling current through a redox molecule sandwiched between two metal electrodes, it has been shown that the presence of many variables such as the rate constants of the electron transfer of the molecule with both tip and substrate, their comparison with the characteristic times for molecular relaxation, the position of the oxidized and reduced states with respect to the Fermi levels of the electrodes, the coupling of the molecule with both electrodes, and the presence of only one or more redox levels/ centers between the tip and the substrate can give rise to a multitude of features in the tunneling current as a function of the gate potential or as a function of the bias voltage.7,25−27 Many of these features have been foreseen by theoretical studies (see below), and it would be interesting to confirm them experimentally. In the last years, molecules involving hydroquinone/ benzoquinone or similar redox couples have raised a particular interest. In fact, experiments on these kinds of molecules have demonstrated the possibility of implementing transistor and switching effects.28−31 The former is consistent with the electrochemical gating effect, which has been already demonstrated for other redox molecules. However, the peculiarity of the exploited redox reaction (that involves exchange of 2 electrons and 2 protons) introduces, in the case of an asymmetric junction, the presence of two tunneling current resonance peaks as a function of the electrochemical gate and a strong dependence of the resonance potentials on the solution pH.28 These features point also to a subtle difference in the behavior of the hydroquinone/quinone redox couple in typical electrochemical experiments, such as cyclic voltammetry, and in the EC-STM that will be further discussed in this work. The second effect, the switching one, is related to a different conductance of the molecule in the two stable oxidation states, with the reduced state being more conductive than the oxidized one.32,33 It has been proposed that this behavior might be related to an off-resonance quantum conduction mechanism where the wave nature of the transmitting electrons can produce a destructive interference in the case of an oxidized molecule.30 In this work we will present new data on the transistor effect of the hydroquinone/benzoquinone redox couple, which allow a deeper understanding of the electron transfer mechanisms involved in the tunneling current resonance along with new evidence for a characteristic negative differential resistance effect in molecules bearing the hydroquinone/quinone redox moiety. A mechanistic rationalization of the electron transfer steps and the role of vibrational relaxation phenomena is provided for all the reported effects.

Chart 1. Structure of 4-(2′,5′-Dihydroxystyryl)benzyl Thioacetate (1) and of 2-(6-Mercaptoalkyl)hydroquinone (2)

self-assembled monolayers (SAM) of the molecules can be found in previous works.28,29 Au(111) substrates were prepared by evaporating a 150 nm-thick gold layer on high vacuum prebaked (450 °C) mica substrates and subsequently annealing them at 450 °C for 4 h in high vacuum. For assembling a molecular adlayer, a gold substrate was flame annealed and immediately immersed in a 10−5 M ethanol solution containing the specific molecules to form the SAM. The sample was then rinsed in ethanol and mounted in the EC-STM cell. The imaging buffer was 50 mM NH4Ac, pH 4.6 and 7.6. A similar procedure was executed when the molecular adlayers were assembled on STM gold tips. In this case, the adlayer was assembled on the tip after covering the tip with apiezon wax (see below) and the flame annealing procedure of the tip could not be executed. The wax-covered tip was simply immersed in the ethanol solution for 3 h and then was rinsed in ethanol. EC Scanning Tunneling Spectroscopy (EC-STS). ECSTM imaging was performed with a PicoSPM microscope (Molecular Imaging) equipped with a bipotentiostat. The bipotentiostat allowed us to control independently the potential of substrate and tip with respect to a quasi-reference silver wire. The tip was prepared by electrochemically etching a gold wire and it was coated with Apiezon wax in order to reduce leakage currents to values smaller than 5−10 pA. As a counterelectrode, a Pt wire was used. The potential of the silver wire was checked against a SCE reference electrode at the beginning and at the end of each experimental session and all the data in the following are referred to SCE. Two types of EC-STS experiments were performed. In a first type, a variation of the tunneling current was measured as a function of the substrate potential (Vs) at fixed height and lateral tip position with the molecular adlayer formed on a gold substrate. To this aim, tip− substrate approach was performed as usual at a given tunneling current set point in off-resonance conditions and the feedback loop was then switched off. The substrate potential was then swept alternatively toward the positive and negative directions at a scan rate of 0.05 V/s in order to cross the redox potential of the hydroquinone/benzoquinone couple for the two molecules. The comparison of the tunneling current at the beginning and at the end of the back and forth sweeps was used as a probe of vertical stability of the tip/substrate position. In a 17452

dx.doi.org/10.1021/jp405516z | J. Phys. Chem. C 2013, 117, 17451−17461

The Journal of Physical Chemistry C

Article

with respect to the oxidized state. As a consequence, this phenomenon can imply an electrochemically controlled switching behavior of the molecule. This aspect means that two conduction channels might be present in the case of redox molecules: a first channel related to the delocalization of electrons in the molecule according to the redox state of the molecule itself and a second channel related to the temporary population of the redox level of the molecule.37 Altogether, these phenomena can give rise to a variety of nonlinear behaviors of the tunneling current through redox molecules. The first experimental evidence of a resonance behavior for the current through a redox molecule in EC-STM setup was provided in a pioneering work by Tao in 1996.18 In that case, Tao showed that, by keeping the bias voltage constant, the tunneling current through Fe-porphyrins was strongly dependent on the potential of substrate on which they were adsorbed. In particular, a resonance-like effect was obtained for the tunneling current as a function of the substrate potential with a maximum near to the equilibrium potential of the redox molecules. This behavior resembles that of a single-molecule transistor. After this experiment, much other experimental evidence of a transistor-like effect has been obtained for different redox molecules, such as, for example, viologen,22,24 metalloproteins,19,38−40 carotenes,41 and transition metal complexes.20 The experiment by Tao, like many other experiments that followed, was based on repetitive imaging of the same sample area with constant tip/substrate bias but with different substrate potential values. This kind of experiment relies on the measurement of the change in apparent height of the molecular structures in the images, which is strictly related to the decay factor of the overall transport phenomenon in the tunneling gap. Moreover, the configuration of the studied molecular species is necessarily asymmetric due to the different interaction it establishes with tip and substrate. In other words, the gap between the tip and the substrate includes both the molecule and a water gap. In contrast, scanning tunneling spectroscopic techniques allow direct measure of the current through redox molecules. Here, the tip is brought near to the molecule until a given current set point is reached. Then, the STM feedback mechanism is switched off and, in the case of sufficiently stable configurations, the substrate voltage or the tip/substrate bias potential are changed while measuring the corresponding tunneling current at constant distance. A symmetrical configuration has been obtained by exploiting the STM break-junction technique developed by Tao.42 Here the current through the molecule is measured when a symmetrical configuration is achieved with the molecule covalently bound to both the tip and the substrate during a withdrawing process of an STM tip from an initial position near to the substrate. This procedure allows the measure of the current through the molecule as a function of the substrate potential and of the molecular oxidation state. The tunneling properties obtained for molecules bound to both electrodes might be different from the transport properties measured for the same molecules in asymmetric configuration. Indeed, in some cases, performing a sweep of the gate voltage at constant bias voltage, the tunneling current showed a clear resonance in asymmetric junctions whereas the transport properties for a symmetric junction showed a sigmoidal behavior characterized by a growth of the conductance across the region where the molecule changed its oxidation state, which is not followed by a decrease of the conductance.23,24 This behavior has been attributed to a so-called “soft gating” mechanism related to the

second type of EC-STS measurements, the molecular adlayer was assembled on a gold tip, the tip was set at a specific potential value with respect to the reference electrode and it was approached to the gold substrate until a specific current set point was established. Then, while keeping the tip potential constant, the substrate potential was swept to obtain an I−Vbias curve. Several test experiments were performed to be sure that, in this second experimental setup, the measured current was not due to capacitive contributions (see below). I−Vbias curves were measured both by acquiring separately the negative and positive Vbias ranges in two sweeps and by measuring the overall range in a single sweep. The potential was typically swept at scan rate of 0.05 V/s but different scan rates were also used to establish if the measured currents were due to capacitive contributions. Theoretical and Experimental Background. From a theoretical point of view the study of redox molecules in an in situ STM junction began in the early ‘90s with the works by Schmickler and Kutznesov and Ulstrup.34,35 The chosen approaches relied mainly on the electron transfer theories developed by Marcus.36 Depending on the degree of coupling of the molecule with the tip and the substrate, different scenarios were described.6,7 Briefly, in the case of weak coupling, the overall transport process of one electron between the two metal electrodes foresees two electron transfer steps. Depending on the initial positions of the energy levels, a first step can occur from the substrate to the molecule, which subsequently relaxes to its reduced state with a stabilization of the electron on the molecule exploiting also the presence of the solvent. A second electron transfer event promoted by fluctuations is then responsible for the transfer of the electron from the molecule to the tip, leaving the molecule again in its oxidized state. The stabilization of the electron on the molecule represents the prevalence of a hopping transport mechanism on a tunneling one in which the time of residence of an electron on the molecule is negligible and does not allow for reorganization of the molecule/solvent system. Intermediate situations between hopping and tunneling can be obtained if one varies the coupling between molecule and electrodes. In the case of a strong coupling (adiabatic limit), the first electron transfer can induce a partial relaxation of the molecule, with relevant role of the environment, opening a channel for the transfer of a great number of electrons from the substrate to the tip. The latter mechanism gives rise to an elevated transport current, which is compatible with many experimental results so far obtained.5 An analytical description (usually performed for low bias voltage) in the case of strong coupling is usually obtained by considering the electron transfer rates in the direction of the electron flow imposed by the bias voltage, neglecting the inverse transitions (even in the case of large bias voltage for weak coupling the inverse transitions can be neglected). The analytical description foresees the possibility of an electrochemical gating effect, with the maximum for the tunneling current occurring at a substrate potential near to the molecular redox equilibrium potential.20 For redox molecules, we have also to consider that, apart from the phenomenon of electron stabilization on the molecule, the stabilization of a change in the molecular redox state can by itself give rise to an increased or decreased efficiency for the transport of electrons between the tip and the substrate as a result of modified delocalized orbitals in the molecule. This means that, for example, a reduced state of a molecule can give rise to more delocalized orbitals, which ensure a more efficient conduction 17453

dx.doi.org/10.1021/jp405516z | J. Phys. Chem. C 2013, 117, 17451−17461

The Journal of Physical Chemistry C

Article

different configurational fluctuations allowed to the molecule in the two experimental setups (molecules in monolayers vs isolated molecules). However, in other cases, even in a symmetrical configuration, a resonance-like behavior has been found. As already stated, the current measured through redox molecules as a function of the electrochemical potential of the solution might involve both an electronic restructuring and a contribution of molecular and solvent fluctuations around different redox states. The first contribution, which can be assimilated to an inner sphere contribution to the current, typically gives rise to a sigmoidal trend of the current as the redox state of the molecule is changed. The second contribution is more related to fluctuations and coupling of electrons to vibrational effects and typically gives rise to resonance-like features in the tunneling current. Depending on which of the two contributions is more relevant, the tunneling current across a redox molecule will display either feature. The two contributions can be present together at the same time. In this case, the sigmoidal behavior of the current will be superimposed to the resonance-like behavior. Recently, a great interest has arisen for systems bearing more than one redox state in the energy gap between tip and substrate.28,29,43−45 These levels can be brought about by the presence of a single level hosting progressively more than one electron in a stable configuration, similar to Coulomb blockade effects in quantum dots (the repulsion energy between the two electrons is not infinite, giving a situation in which the molecule can have three accessible oxidation states)46 or by the presence of two different redox levels associated with one or two molecules in the gap. At the same time, we must distinguish the case in which electron transfer proceeds via a sequential process between the two molecular levels, from that in which both levels can communicate directly with the two electrodes. From a theoretical point of view, even if the behavior of each redox level is similar to that occurring when only one redox level is present, the introduction of two redox levels within the gap, weakly or strongly coupled to the electrodes, the interaction between the two levels themselves, and the different effective potential at the two redox sites, introduce a rich variety of possibilities for the tunneling current as a function of the gate or bias voltage. In particular, negative differential resistance and rectification effects have been foreseen.27,47,48 In particular, in the limit of totally adiabatic conditions, for molecules bearing two redox levels, the presence of two regions of tunneling current enhancement has been predicted as a function of the substrate potential while keeping the bias voltage constant.27 In previous works, we demonstrated that the hydroquinone molecules in the gap between the tip and the substrate behave analogously to molecules with two redox levels strongly coupled to the electrodes.28,29 In fact, the tunneling current as a function of the overpotential gives rise to two regions of tunneling current enhancement. However, other interesting features of the current/overpotential and current/ bias voltage relations remain to be experimentally investigated to support the theoretical studies. Moreover, the role played by the structural aspects of the tunneling junction in determining the details of the tunneling current enhancement remains to be explored.

show a transistor-like effect when tip or substrate potential is changed with respect to the reference electrode while keeping the bias voltage constant. The redox reaction of hydroxyquinone involves the exchange of two electrons and two protons. The protons involvement imparts a marked dependence on the solution pH to the transistor-like effect. This is reflected in a dependence on the pH of the range of substrate potential at which the tunneling current enhancement occurs, configuring the possibility of a pH-gated transistor. Moreover, a particular feature of the tunneling current across hydroxyquinone molecules in an EC-STM setup is the presence of two regions of current enhancement as a function of the substrate potential. This behavior has been attributed to the presence of two electron transfer processes according to the overall number of exchanged electrons. However, the particular aspect for this molecule resides in the observation that in typical cyclic voltammetry experiments, the transfer of the two electrons occurs under a single voltammetric wave. In voltammetry, the first electron transfer event brings the quinone to a semiquinone state involving vibrational relaxation and solvent reorganization, which immediately favors the second electron transfer event, making the voltammetric detection of the semiquinone state impossible in aqueous solutions. Kano and Uno demonstrated49 that the voltammograms of adsorbed quinones can be interpreted on the basis of a two-step mechanism that involves a sequential transfer of the two electrons. Indeed, deviations from ideality of the voltammograms can be well interpreted assuming two separate events, each characterized by its redox equilibrium potential. In particular, if the two equilibrium potentials are very near to each other or they are in an inverted order (once the first electron transfer process occurs the system is already at a potential allowing the transfer of the second electron), the voltammogram presents only one oxidation and one reduction wave subtending the transfer of two electrons. The appearance of the two separate electron transfer events in EC-STM experiments has been ascribed to the simultaneous presence of two electrodes with different Fermi levels near to the redox molecule with the possibility for the molecule to exchange electrons with both of them.28 We also compared the behavior of two different molecules bearing the same hydroquinone/ benzoquinone redox group but linked to the gold substrate by different molecular segments. The choice of the two segments was guided by the aim of assuring a different electronic coupling between the electrode and the redox group. The main difference we found was related to the position of the regions of tunneling current enhancement which allowed a deeper understanding of the electron transfer steps involved in the transport phenomenon. Even if the redox levels associated with the two processes cannot be present in a stable configuration at the same time in the gap between the tip and the substrate, fluctuations in the molecular configuration enable EC-STM observation of the two levels. In practice, the system behaves as if two effective redox levels were present in the tunneling gap. This allowed for comparing the results we obtained with the theories developed in the case of multielectron transfer events. Here we consider another aspect of the tunneling junction which could affect the transported current. In particular, we study the difference in the tunneling current enhancement of the two molecules in the EC-STM setup. These have a similar length but different overall electron transport properties as a consequence of their linkers. If EC-STM/STS experiments are performed on the two molecules starting with the same current



RESULTS AND DISCUSSION Transistor Effect on Hydroxyquinone Molecules. In refs 28 and 29 we demonstrated that hydroxyquinone molecules in the gap between an STM tip and a gold substrate 17454

dx.doi.org/10.1021/jp405516z | J. Phys. Chem. C 2013, 117, 17451−17461

The Journal of Physical Chemistry C

Article

different values of the initial current set point. It is clear that for larger initial current set point the tip is brought nearer to the molecule, increasing the enhancement effect, as in the case of the two different molecules. Figure 2b shows the behavior of the tunneling current corresponding to the peak with respect to the current set point, whereas Figure 2c reports the trend of the enhancement factor (Ipeak/I0) as a function of the current set point (I0). In Figure 2b an almost linear increase of peak current as a function of the initial off-resonance current set point is found. Also, the enhancement factor increases with the current set point (Figure 2c) and it seems to reach a plateau after an initially steep increase. The first aspect has been found in other works dealing with the electrolyte gating effect of redox molecules, even if not with an almost linear trend as for our data.5,50−52 Our data suggest that the electron transport phenomenon does not reach a saturation effect while the set point current is increased, which means that a molecule, in these experimental conditions, will never be contacted by both electrodes. In fact, we expect saturation in the tunneling current, at constant bias, if the molecule starts being in contact with both leads, when only the tunneling barrier represented by the molecule will play a role. For the value of the enhancement factor, we observe in Figure 2c an initial large increase for small current values, followed by its stabilization.51 The observed trend could be interpreted as a transition from weak to strong coupling conditions between the not-functionalized electrode and the molecule. However, the increased coupling does not induce tunneling current saturation. In other cases it has been found that an increase of the current set point induces a decrease of the tunneling current enhancement factor, whereas we always found an increase. This behavior could be rationalized considering the different weights that the two tunneling barriers (that provided by the molecule and the one from the water gap) can have on the overall transport mechanism for different molecules. It is usually assumed that by decreasing the tip/molecule distance, the relative importance of the outer sphere reorganization contribution should decrease due to limited access of the solvent to the molecule (usually considered as a reduction in the reorganization energy of the molecule) and the conformational fluctuations of the molecule could be limited. Our results show that a molecule which is positioned nearer to an STM tip is able to produce a larger enhancement of the tunneling current with respect to a molecule positioned farther. It is difficult to establish a direct mechanistic correlation between the decrease of the reorganization energy and the value of the enhancement current. This is due to the fact that a variation of the tip/molecule distance can alter both the reorganization energy and the coupling parameters in the junction. The variation of these parameters can also make the variation of the involved potentials relatively more important. However, the results obtained here imply that the dominant role in establishing the value of the tunneling current is played by the particular structure of the tunneling junction. The above discussion assumes that the number of molecules involved in the tunneling transport does not change increasing the tunneling current set point. Indeed, a decrease in the water gap between the molecules and the second gold electrode could also induce a contribution to the current from more than one molecule in the gap. Especially in the case of the curve on molecule 2, it is found that a decreasing value of the tunneling current from negative to positive substrate potentials is overlaid to the two peaks. In

set point, different situations are at play in the tunneling gap and the structure of the tunneling junction itself gives rise to specific features. A more conductive linker produces a tunneling water gap between the molecule and the tip (more generally between the molecule and the electrode without the immobilized molecules), which is larger than in the case of the molecule with the less conductive linker. Indeed, at constant tunneling current set point, if the overall conductance of the molecule increases, the gap between the end of the molecule and the other electrode increases (see also Figure S1 in Supporting Information). This means that for the more conductive 4-(2′,5′-dihydroxystyryl)benzyl thioacetate (1), the molecule/tip gap is the rate limiting one. The overall β factor for the tunneling transport can be obtained from the product of the molecular β factor and that associated with the water gap in off-resonance condition in which the tip/substrate approach is obtained (for conduction dominated by a tunneling phenomenon). Due to the limited length of the linker, it is highly probable that also in the case of the lesser conductive 2-(6mercaptoalkyl)hydroquinone 2 the dominant barrier will be the gap between the molecule and the tip, but the molecule will be nearer to the gold electrode with respect to molecule 1. Accordingly, if the tip is very near to the molecule, a small variation of the electron transport properties of the molecule as a consequence of a variation of the substrate potential, will induce a strong enhancement effect on the tunneling current. Figure 1a shows the trend of the tunneling current in STS

Figure 1. Evidence of the different tunneling current enhancement effects for hydroquinone/quinone redox groups derivatized with linkers having different conductivity. The less conductive linker (molecule (2), filled squares/red line) produces an enhancement that is stronger than that observed in the case of a more conductive linker (molecule (1), open squares/blue line). The background current due to a change in the oxidation state of the molecule has been subtracted.

experiments performed as a function of substrate potential at constant bias voltage for the molecules 1 and 2 both immobilized on a gold substrate. It is evident that, by performing an initial tip−sample approach at the same current set point in an off-resonance region, the tunneling current enhancement is much higher in the case of the less conductive linker 2. These results are consistent with those obtained by Wandlowski et al. with redox-active N-thioalk(ano)ylferrocenes (FcN) molecules with N = 4, 6, and 8.50 In that work, Wandlowski et al. demonstrated that when the rate-limiting factor is given by the linker between the substrate and the redox group, the enhancement effect becomes negligible and not visible in the experimental curves. Similar concepts could apply to the case in which the tunneling current enhancement is performed for different initial current set-points. Figure 2a shows different STS curves obtained on a SAM of 2 for 17455

dx.doi.org/10.1021/jp405516z | J. Phys. Chem. C 2013, 117, 17451−17461

The Journal of Physical Chemistry C

Article

Figure 2. (a) Effect of the different initial off-resonance current set point on the tunneling current enhancement for molecule 2. Inset: Magnification of the behavior obtained for the lower current set point value (I0 = 0.5 nA (orange circles); 1 nA (purple open squares); 5 nA (green filled squares); 10 nA (blue triangles)). (b) Dependence of the value of the in-resonance peak current value as a function of the off-resonance initial tunneling current set point (c) Plot of the tunneling current enhancement factor (Ipeak/I0) as a function of I0. The lines in b and c are guides to the eye.

Figure 3. a) I−z curves obtained on a self-assembled monolayer of (2) for different values of substrate potential. Black solid lines refer to the curves of the oxidized molecules, whereas the dotted gray lines refer to the curves of the reduced molecules. The two straight lines refer to the two linear regions in the log I−z plot. The arrows points to the linear region chosen to obtain the effective β-factor. (b) The β factor obtained from the linear dependence of the log(I) vs Δz curves is reported as a function of the substrate potential. The dotted line is a guide to the eye (see Supporting Information for more details).

introduced to explain this effect, such as a conformational change of the molecule induced by the electrostatic field in the junction, and Coulomb-blockade effects, and it is likely that many of these causes act together in a typical experiment. In an electrochemical environment with a redox molecule bridging two metal electrodes, the alignment and misalignment of redox levels with the Fermi energy of the electrodes could result in an evident NDR effect also at room temperature.57 The most straightforward method to detect the presence of a NDR phenomenon is that of performing I−Vbias curves with the molecule or molecules of interest bridging the gap between two metal electrodes. In the STM context, this implies that the tip should approach the substrate until a given current set point is reached. The feedback mechanism is then switched off and, for a sufficiently stable setup, the bias voltage is swept while measuring the tunneling current at constant distance. In real EC-STM experiments this approach could be problematic. In fact, sweeping the bias could introduce in the overall measured current components coming from capacitive effects of the restructuring double layer.21 Typically, the presence of capacitive currents can be detected by changing the bias sweep rate. In fact, a systematic dependence of the measured current on the sweep rate could be ascribed to currents that do not originate from tunneling phenomena. The problems

this case the presence of the two channels discussed above is clearly evident. Moreover, the increased conductance of the molecule in the reduced form is also confirmed by measurements of the β factor of the overall tunneling junction in I−z experiments as a function of the substrate potential in offresonance regions (Figure 3 and other details in Supporting Information).53,54 An overall decrease of the β factor is observed as substrate potential gets more negative, even if the estimation of the β factor is to be considered only for offresonance conditions. Moreover, the numerical values obtained for the β factor are consistent with what has been found in previous works on a similar molecule.32 Negative Differential Resistance (NDR) and Switch Behavior. One of the nonlinear conductance effects which has received much attention since the pioneering studies on molecular electronics is Negative Differential Resistance (NDR).9,55 This effect consists in a decrease in current upon an increase in source/drain bias voltage. In solid-state electronics this property is associated, for example, with resonant-tunneling diodes.56 In the context of molecular electronics, the appearance of features resembling NDR is usually attributed to a matching and subsequent mismatching of specific levels of the bridge to the relevant electronic levels in the two metal contacts. Other mechanisms have been 17456

dx.doi.org/10.1021/jp405516z | J. Phys. Chem. C 2013, 117, 17451−17461

The Journal of Physical Chemistry C

Article

Figure 4. (a,b) Examples of the two types of curves obtained executing I−Vbias experiments on 4-(2′,5′-dihydroxystyryl)benzyl thioacetate molecules for different values of tip potential with respect to reference electrode. The reported curves are the results of averages over about 12 curves for each value of tip potential. (c) Plot of the conductance relative to the curve from (a) corresponding to Vtip = −0.3 V vs SCE. (d) Plot of the conductance relative to the curve from (b) corresponding to Vtip = −0.25 V vs SCE.

connected to this technique have limited the use of the I−Vbias spectroscopy in EC-STM experiments, in spite of the richness of information it could provide about molecular transport properties, allowing a sort of spectroscopic analysis of single molecules. To overcome the aforementioned difficulty, it is possible to perform a series of experiments at different bias voltages while keeping all the other parameters fixed. For example, with the introduction of the STM break-junction technique, it has been possible to measure the conductivity of single redox molecules for different values of applied bias voltage. However, it has been demonstrated that kinetic measurements (measuring current while sweeping potential) can lead to different results with respect to static ones (measuring current at constant potential values).31 In fact, measuring single molecule conductance at constant bias voltage and at different gate voltages while retracting the tip from the funzionalized substrate in the STM break-junction setup leads to a different behavior compared to the case in which the gate voltage is sweeped, at constant bias voltage, while the molecule is sandwiched between tip and substrate. In the case of static measurements, it has been found that, by increasing the bias voltage, a NDR event is observed for heptaaniline oligomers.10,58 In that specific case, the observed NDR effect was explained by considering that different bias voltages could induce a change in the redox state of the molecule, switching the molecule between two states of different conductivity. According to this interpretation, the effect on the tunneling current is provided by a different molecular electron delocalization and a possible different alignment of the molecular electronic states with the Fermi level of the electrodes. In the same work, a tip/substrate bias voltage

sweep was performed while the molecule was in contact with both electrodes. The obtained traces showed the presence of a NDR event overlaying a background current linearly increasing with bias voltage. Other experimental investigations demonstrated the presence of NDR effects when redox molecules were considered.41,59,60 Even in these cases, the NDR effect was ascribed to a different redox state induced by a fraction of the applied bias voltage at the redox center. For example, the appearance of a NDR effect in the work by Chen et al.59 has been attributed to a decrease of the molecular conductance when one electron was stabilized in the molecular bridge and a subsequent new increase of the molecular conductance when a second electron was stabilized on the bridge. To support their interpretation, Chen et al.59 considered that the width of the NDR effect is comparable to the distance between the two reduction peaks observed in cyclic voltammetry of the bridge molecules. In the work by Visoly-Fischer et al.,41 the usual twostep electron transport process described by Kuznetsov and Ulstrup could not fit the transport behavior of the redox molecule upon a change in the bias voltage. The authors concluded that the observed NDR effect was due to a change in the oxidation state of the molecule. Let us discuss the scenario of NDR in the case of a redox molecule bridging two electrodes in the case in which the tunneling current is mainly due to fluctuations of the redox levels.57 Suppose we start with a molecule in the oxidized state whose energy level is above the Fermi level of the electrode, which we assume to play the role of source electrode. Also, suppose that the molecule is more strongly coupled to the electrode whose overpotential is kept fixed, while the Fermi level of the other electrode is changed linearly in time. Due to the voltage drop across the region where the molecule is 17457

dx.doi.org/10.1021/jp405516z | J. Phys. Chem. C 2013, 117, 17451−17461

The Journal of Physical Chemistry C

Article

Figure 5. (a) Plot of the Vbias value corresponding to the switch between the two conductance state of the molecules obtained from Figure 4a. The continuous line is the corresponding linear fit. (b) 2D map of the tunneling current for different values of the tip potential and Vbias (Vtip range: from −0.3 to 0.3 V vs SCE; Vbias range: from 0 V to −1 V). (c) Plot of the Vbias value corresponding to the peak of the conductance plot obtained from Figure 4d. The continuous line is the corresponding linear fit.

values of tip potential. The curves show a nearly constant molecular resistance that, above a certain negative Vbias threshold (defined as the Vbias potential corresponding to the maximum slope in the conductance plot), strongly increases. The behavior is reminiscent of a NDR effect that is probably associated with a switch of the molecular redox state induced by bias voltage. The conductance obtained from a typical curve of Figure 4a is reported in Figure 4c. The curve shows a sigmoidal trend. In the case of single molecule conductance at constant Vbias, several experimental studies demonstrated a sigmoidal behavior for similar molecules. For example, in the case of a STM break-junction study on an anthraquinone-based norbornylogous bridge molecule,31 an increased conductance has been found for an increased negative value of the gating potential. The second type of curve we found is shown in Figure 4b. The representative conductance trend for this type of curve is reported in Figure 4d. In this case, the trend is composed by an almost constant value of the conductance coupled with a region of strong increase followed by a decrease featuring a resonancelike behavior. A similar behavior has been found by Gittins et al. for a redox molecule sandwiched between a gold substrate and a gold nanoparticle when studied by STM.13 In the case of the study of transport properties of single redox molecules, it has been reported that the features of the tunneling current as a function of the gating potential can be represented by both a resonance-like behavior and a sigmoidal trend (see above). Probably, the difference in the two cases stems from the degree of coupling between the molecule and the electrodes. For some molecules, in the case of a symmetric junction with both molecular ends covalently bound to the electrodes, the sigmoidal trend has been observed whereas, for the same molecule in an asymmetric junctions in which the tip is not covalently bound to the molecule, a resonance-like behavior has been found. Here, it is possible that the two different types of behavior correspond to a different degree of coupling between the molecule and both electrodes, even if the experiment was intended to have just an asymmetric junction. According to this interpretation, the trend of Figure 4a corresponds to the case of strong coupling in which the switching properties dominates, whereas the trend of Figure 4b correponds to a situation in which the transport is dominated by a partial localization of the electrons on the redox molecule in which fluctuations play a dominant role.

located, the potential at the site of the molecule results affected by the bias voltage via a coupling coefficient and, assuming for example that the coupling coefficient is lower than 1, the potential variation at the redox molecule will be lower than the variation imposed by the bias voltage. Depending on the bias polarity and on the initial overpotential, different scenarios can be envisaged. One of these possible scenarios considers that, when increasing the bias voltage, the molecule’s levels will reach the Fermi level of the tip and, upon an electron transfer event, the molecule will relax toward its reduced level. This situation will induce an increase in the tunneling current. Further increasing the bias voltage, the molecule will be stabilized in its reduced state which can be in a position lower than the Fermi level of the substrate. In this situation the tunneling current will decrease. Further increasing the bias voltage, the reduced level of the bridge will be again in the potential window between the Fermi levels of the two electrodes and the current will start to rise again. In this work we performed I−Vbias measurements for different values of the tip potential, which was kept constant. The molecules (4-(2′,5′-dihydroxystyryl)benzyl thioacetate) were covalently linked to the gold tip surface. Before measuring molecules, we executed a number of control experiments to be sure that the observed features could not be attributed to capacitive currents or voltammetric signals at the surface of the electrodes. We also measured the current as a function of the tip/substrate bias in the case of an insulated bare tip at tunneling distance from a bare substrate. In this case, no significative variation of the tunneling current was observed apart from a linear variation with bias voltage. In order to exclude a contribution from capacitive currents, we also performed I−Vbias measurements that included sweeping back and forth the bias voltage (see Supporting Information). In these experiments, capacitive currents must appear with different signs in the two traces. We typically found that the two traces had similar trends apart from a small hysteretic behavior. In light of these controls, we strongly believe that the features we observed in the case of tips modified with hydroquinone layers in tunneling contact with a substrate are mainly due to tunneling processes that involve the presence of the molecule. The experiments we performed resulted essentially in the appearance of two different trends for the I−Vbias curves. The first type of result is reported in Figure 4a, where we repeated the I−Vbias measurements for different 17458

dx.doi.org/10.1021/jp405516z | J. Phys. Chem. C 2013, 117, 17451−17461

The Journal of Physical Chemistry C In Figure 4a it is evident that the Vbias threshold for the conductance decrease depends on tip potential. In Figure 5a we reported the trend of the Vbias threshold as a function of the tip potential. The plot presented in Figure 5a shows a strong correlation between tip potential and threshold voltage. If the switching behavior is to be ascribed to a conversion between two redox states of the molecule induced by a fraction of the bias voltage, the plot in Figure 5a should provide an estimate for the coefficient which establishes the variation of the local potential at the redox center as a function of variation of the bias voltage. In Figure 5a we obtain a value around 1. This means that the applied bias is able to change the potential at the redox center with an efficiency similar to that of the interfacial polarization. In other words, there is not an efficient screening of the double layer which could prevent the effect of the bias voltage on the redox center. A value around 1 for this coefficient has been obtained also in electrochemically controlled break-junction experiments.59,61 In Figure 5b we reported a 2D representation of the behavior of the molecular junction as a function of both Vbias and tip potential. The color scale is representative of the current intensity. If a horizontal section of the plot in Figure 5b is extracted, the obtained curve should be similar to what is obtained in I−Vgate experiments for a constant value of the bias voltage. For a bias voltage around 0 V, it is possible to see an increase in tunneling current while going toward negative tip potentials and, for higher Vbias, a resonace-like behavior of the tunneling current is observed. Figure 5c shows the positions of the conductance peak in the case of curves represented by Figure 4b,d as a function of tip potential. In this specific case, ascribed to a weak coupling, there is not a definite trend in the position of the peak and the parameter described above should be around 0. In a previous report and also in the first part of this work, we showed that the molecule at issue here, presents two peaks when, at constant Vbias, the tip potential is swept with respect to a reference electrode. The relevant question is to understand if this particular feature is important in I−Vbias measurements. The molecule we used has three redox states and the intermediate one, which in aqueous environment is not stable (from what it appears by cyclic voltammetry), is stabilized when it is in tunneling contact with two electrodes. From a theoretical point of view, the situation of two redox levels in the same molecule sandwiched between two metal electrodes has been treated by Kuznetsov and Ulstrup without thermal activation but including vibrational relaxation.44,47 The predicted I−Vbias curves are strongly related to the possibility that only one or both levels enter the energy window defined by the increasing bias voltage. The predicted trends do not take into account the possible variation of the conduction of the transport channel associated with a different delocalization of the electrons on the molecule. In our results, in the case of the curves interpreted as due to strong molecule−electrodes coupling, it is likely that the change in the conductance due to that in the oxidation state prevails on the effect of a two-step electron transport mechanism. The situation is different in the weak coupling regime. Indeed, in the last circumstance, the successive entrance or exit of the two redox levels in the bias voltage window should affect the I−Vbias curve. In Figure 5b,d it is not clear if both levels provide a contribution to the tunneling current. Accordingly, further experiments will be needed to clarify this aspect considering the energy separation between the two levels and the explored bias voltage range.



CONCLUSIONS



ASSOCIATED CONTENT

Article

We studied the electron transport properties of redox molecules containing the hydroquinone/benzoquinone redox group in an EC-STM setup. We established that the structure of the tunneling gap plays a fundamental role in the phenomenon of the electrolyte gating. In particular, the nearer is the STM tip to the redox molecules responsible for the tunneling current enhancement, the higher is the overall enhancement factor. This situation has been verified considering two different molecular species endowed with the same redox group but with different linker conductivity and, for the same molecular species, by considering the dependence of the enhancing factor on the initial tunneling current set-point in off-resonance conditions, where the current is due only to direct tunneling. The entering of a second channel in the tunneling phenomenon, due to electron transfer steps in which the redox properties of the molecule are involved, is more effective in increasing the overall current if the barrier due to the water gap between the tip and the molecule is smaller. In practice, in the cases considered in this work, the rate-limiting step for current enhancement is mainly that through the water gap. We also presented I−Vbias spectroscopy data on the hydroquinone molecules endowed with a more conductive linker (4-(2′,5′-dihydroxystyryl)benzyl thioacetate) and we showed the appearance of two different behaviors. These were interpreted as due to a different coupling between the molecules and both electrodes. In the case characterized by a stronger coupling, the I−Vbias curves presented a NDR feature that could be interpreted as due to the molecular switching from a more conductive redox state to a less conductive state induced by the changing Vbias. Here, the transport channel associated with electron delocalization in the molecule prevails on that involving fluctuations of the molecular redox levels. The second behavior, which presents resonant-like trend for the conduction, has been interpreted as due to situations in which the molecule is weakly coupled with the electrode on which it has not been covalently bound. In this second case, the dominant effect is a resonant-like one, associated with the twostep electron transfer mechanism involving the relaxation of the molecular redox levels. This work demonstrates the rich variety of transport characteristic that a molecular junction can provide depending on the structural and electronic details of the junction. On the one hand, this aspect can be considered as an advantage of molecular electronics over traditional electronics but, on the other hand, it highlights the requirement for a fine-tuning of all the involved parameters in order to obtain reproducible characteristics.

S Supporting Information *

Details of the tunneling-junction structure, I−z measurements, and comparison between apparent height variation of hydroquinone molecules and effective β factor; back and forth I−V sweeps; and four figures. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]; paolo.facci@nano. cnr.it. 17459

dx.doi.org/10.1021/jp405516z | J. Phys. Chem. C 2013, 117, 17451−17461

The Journal of Physical Chemistry C

Article

Notes

(22) Haiss, W.; Albrecht, T.; van Zalinge, H.; Higgins, S. J.; Bethell, D.; Höbenreich, H.; Schiffrin, D. J.; Nichols, R. J.; Kuznetsov, A. M.; Zhang, J.; Chi, Q.; Ulstrup, J. Single-Molecule Conductance of Redox Molecules in Electrochemical Scanning Tunneling Microscopy. J. Phys. Chem. B 2007, 111, 6703−6712. (23) Leary, E.; Higgins, S. J.; van Zalinge, H.; Haiss, W.; Nichols, R. J.; Nygaard, S.; Jeppesen, J. O.; Ulstrup, J. Structure−Property Relationships in Redox-Gated Single Molecule JunctionsA Comparison of Pyrrolo-tetrathiafulvalene and Viologen Redox Groups. J. Am. Chem. Soc. 2008, 130, 12204−5. (24) Li, Z.; Han, B.; Meszaros, G.; Pobelov, I.; Wandlowski, T.; Błaszczyk, A.; Mayor, M. Two-Dimensional Assembly and Local Redox-Activity of Molecular Hybrid Structures in an Electrochemical Environment. Faraday Discuss. 2006, 131, 121−43. (25) Friis, E. P.; Kharkats, Y. I.; Kuznetsov, A. M.; Ulstrup, J. In Situ Scanning Tunneling Microscopy of a Redox Molecule as a Vibrationally Coherent Electronic Three-Level Process. J. Phys. Chem. A 1998, 102, 7851−7859. (26) Zhang, J.; Chi, Q.; Albrecht, T.; Kuznetsov, A. M.; Grubb, M.; Hansen, A. G.; Wackerbarth, H.; Welinder, A. C.; Ulstrup, J. Electrochemistry and Bioelectrochemistry Towards the SingleMolecule Level: Theoretical Notions and Systems. Electrochim. Acta 2005, 50, 3143−3159. (27) Kuznetsov, A. M.; Medvedev, I. G. A Theory of Redox-Mediated Electron Tunneling Through an Electrochemical Two-Center Contact. J. Phys.: Condens. Matter 2008, 20, 374112. (28) Petrangolini, P.; Alessandrini, A.; Berti, L.; Facci, P. An Electrochemical Scanning Tunneling Microscopy Study of 2-(6Mercaptoalkyl)hydroquinone Molecules on Au(111). J. Am. Chem. Soc. 2010, 132, 7445−53. (29) Petrangolini, P.; Alessandrini, A.; Navacchia, M. L.; Capobianco, M. L.; Facci, P. Electron Transport Properties of Single-MoleculeBearing Multiple Redox Levels Studied by EC-STM/STS. J. Phys. Chem. C 2011, 115, 19971−19978. (30) Darwish, N.; Díez-Pérez, I.; Da Silva, P.; Tao, N.; Gooding, J. J.; Paddon-Row, M. N. Observation of Electrochemically Controlled Quantum Interference in a Single Anthraquinone-Based Norbornylogous Bridge Molecule. Angew. Chem., Int. Ed. 2012, 51, 3203−6. (31) Darwish, N.; Díez-Pérez, I.; Guo, S.; Tao, N.; Gooding, J. J.; Paddon-Row, M. N. Single Molecular Switches: Electrochemical Gating of a Single Anthraquinone-Based Norbornylogous Bridge Molecule. J. Phys. Chem. C 2012, 116, 21093−21097. (32) Tsoi, S.; Griva, I.; Trammell, S. A.; Blum, A. S.; Schnur, J. M.; Lebedev, N. Electrochemically Controlled Conductance Switching in a Single Molecule: Quinone-Modified Oligo(phenylene vinylene). ACS Nano 2008, 2, 1289−95. (33) Li, J.; Zhao, Z.; Yu, C.; Wang, H.; Zhao, J. Theoretical Investigation on the Transportation Behavior of Molecular Wires with Redox Reaction. J. Comput. Chem. 2012, 33, 666−72. (34) Schmickler, W.; Widrig, C. The Investigation of Redox Reactions with a Scanning Tunneling Microscope: Experimental and Theoretical Aspects. J. Electroanal. Chem. 1992, 336, 213−217. (35) Kuznetsov, A. M.; Sommer-Larsen, P.; Ulstrup, J. Resonance and Environmental Fluctuation Effects in STM Currents Through Large Adsorbed Molecules. Surf. Sci. 1992, 275, 52−60. (36) Marcus, R. A.; Sutin, N. Electron Transfer in Chemistry and Biology. Biochim. Biophys. Acta 1985, 811, 265−322. (37) Migliore, A.; Nitzan, A. Nonlinear Charge Transport in Redox Molecular Junctions: A Marcus Perspective. ACS Nano 2011, 5, 6669− 85. (38) Chi, Q.; Farver, O.; Ulstrup, J. Long-Range Protein Electron Transfer Observed at the Single-Molecule Level: In Situ Mapping of Redox-Gated Tunneling Resonance. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 16203−16208. (39) Alessandrini, A.; Gerunda, M.; Canters, G. W.; Verbeet, M. P.; Facci, P. Electron Tunnelling Through Azurin is Mediated by the Active Site Cu Ion. Chem. Phys. Lett. 2003, 376, 625−630. (40) Facci, P.; Alliata, D.; Cannistraro, S. Potential-Induced Resonant Tunneling Through a Redox Metalloprotein Investigated by Electro-

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS The authors acknowledge partial financial support from Italian Ministry of Research FIRB Project “Italnanonet”. REFERENCES

(1) Aviram, A.; Ratner, M. A. Molecular Rectifiers. Chem. Phys. Lett. 1974, 29, 277−283. (2) Joachim, C.; Ratner, M. A. Molecular Electronics Special Feature: Molecular Electronics: Some Views on Transport Junctions and Beyond. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 8801−8808. (3) Tao, N. J. Electron Transport in Molecular Junctions. Nat. Nanotechnol. 2006, 1, 173−181. (4) Gimzewski, J. K.; Joachim, C. Nanoscale Science of Single Molecules Using Local Probes. Science 1999, 283, 1683−1688. (5) Pobelov, I. V.; Li, Z.; Wandlowski, T. Electrolyte Gating in Redox-Active Tunneling JunctionsAn Electrochemical STM Approach. J. Am. Chem. Soc. 2008, 130, 16045−54. (6) Zhang, J.; Kuznetsov, A. M.; Medvedev, I. G.; Chi, Q.; Albrecht, T.; Jensen, P. S.; Ulstrup, J. Single-Molecule Electron Transfer in Electrochemical Environments. Chem. Rev. 2008, 108, 2737−91. (7) Alessandrini, A.; Corni, S.; Facci, P. Unravelling Single Metalloprotein Electron Transfer by Scanning Probe Techniques. Phys. Chem. Chem. Phys. 2006, 8, 4383−97. (8) Nichols, R. J.; Haiss, W.; Higgins, S. J.; Leary, E.; Martin, S.; Bethel, D. The Experimental Determination of the Conductance of Single Molecules. Phys. Chem. Chem. Phys. 2010, 12, 2801−15. (9) Chen, J.; Reed, M. A.; Rawlett, A. M.; Tour, J. M. Large On−Off Ratios and Negative Differential Resistance in a Molecular Electronic Device. Science 1999, 286, 1550−1552. (10) Chen, F.; He, J.; Nuckolls, C.; Roberts, T.; Klare, J. E.; Lindsay, S. M. A Molecular Switch Based on Potential-Induced Changes of Oxidation State. Nano Lett. 2005, 5, 503−506. (11) Xiao, X.; Nagahara, L. A.; Rawlett, A. M.; Tao, N. Electrochemical Gate-Controlled Conductance of Single Oligo(phenylene ethynylene)s. J. Am. Chem. Soc. 2005, 127, 9235−9240. (12) Metzger, R. M. Unimolecular Electrical Rectifiers. Chem. Rev. 2003, 103, 3803−3834. (13) Gittins, D. I.; Bethell, D.; Schiffrin, D. J.; Nichols, R. J. A Nanometer-Scale Electronic Switch Consisting of a Metal Cluster and Redox-Addressable Groups. Nature 2000, 408, 67−69. (14) van der Molen, S. J.; Liljeroth, P. Charge Transport Through Molecular Switches. J. Phys.: Condens. Matter 2010, 22, 133001. (15) Migliore, A.; Nitzan, A. Irreversibility and Hysteresis in Redox Molecular Conduction Junctions. J. Am. Chem. Soc. 2013, 135, 9420− 32. (16) Lustenberger, P.; Rohrer, H.; Christoph, R.; Siegenthaler, H. Scanning Tunneling Microscopy at Potential Controlled Electrode Surface in Electrolytic Environment. J. Electroanal. Chem. 1988, 243, 225−235. (17) Corni, S. A Theoretical Study of the Electrochemical Gate Effect in an STM-Based Biomolecular Transistor. IEEE Trans. Nanotechnol. 2007, 6, 561−570. (18) Tao, N. J. Probing Potential-Tuned Resonant Tunneling Through Redox Molecules with Scanning Tunneling Microscopy. Phys. Rev. Lett. 1996, 76, 4066−4069. (19) Alessandrini, A.; Salerno, M.; Frabboni, S.; Facci, P. SingleMetalloprotein Wet Biotransistor. Appl. Phys. Lett. 2005, 86, 133902. (20) Albrecht, T.; Guckian, A.; Ulstrup, J.; Vos, J. G. Transistor-Like Behavior of Transition Metal Complexes. Nano Lett. 2005, 5, 1451− 1455. (21) Li, C.; Mishchenko, A.; Li, Z.; Pobelov, I.; Wandlowski, T.; Li, X. Q.; Würthner, F.; Bagrets, A.; Evers, F. Electrochemical GateControlled Electron Transport of Redox-Active Single Perylene Bisimide Molecular Junctions. J. Phys.: Condens. Matter 2008, 20, 374122. 17460

dx.doi.org/10.1021/jp405516z | J. Phys. Chem. C 2013, 117, 17451−17461

The Journal of Physical Chemistry C

Article

chemical Scanning Probe Microscopy. Ultramicroscopy 2001, 89, 291− 298. (41) Visoly-Fisher, I.; Daie, K.; Terazono, Y.; Herrero, C.; Fungo, F.; Otero, L.; Durantini, E.; Silber, J. J.; Sereno, L.; Gust, D.; Moore, T. A.; Moore, A. L.; Lindsay, S. M. Conductance of a Biomolecular Wire. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 8686−8690. (42) Xu, B. Q.; Tao, N. J. Measurement of Single-Molecule Resistance by Repeated Formation of Molecular Junctions. Science 2003, 301, 1221−3. (43) Kay, N. J.; Higgins, S. J.; Jeppesen, J. O.; Leary, E.; Lycoops, J.; Ulstrup, J.; Nichols, R. J. Single-Molecule Electrochemical Gating in Ionic Liquids. J. Am. Chem. Soc. 2012, 134, 16817−26. (44) Kuznetsov, A. M.; Ulstrup, J. Theory of Electron Tunneling Through a Bridge Molecule with Two Electronic Levels at Low Temperature. Russ. J. Electrochem. 2006, 42, 760−766. (45) Schmickler, W.; Rampi, M. A.; Tran, E.; Whitesides, G. M. Electron Exchange Between Two Electrodes Mediated by Two Electroactive Adsorbates. Faraday Discuss. 2004, 125, 171−7. (46) Kuznetsov, A. M.; Medvedev, I. G.; Ulstrup, J. Coulomb Repulsion Effect in Two-Electron Nonadiabatic Tunneling Through a One-Level Redox Molecule. J. Chem. Phys. 2009, 131, 164703. (47) Kuznetsov, A. M.; Ulstrup, J. Single-Molecule Electron Tunnelling Through Multiple Redox Levels with Environmental Relaxation. J. Electroanal. Chem. 2004, 564, 209−222. (48) Kuznetsov, A. M.; Medvedev, I. G. Effect of Coulomb Interaction Between the Electrons on Two-Electron Redox-Mediated Tunneling. Electrochem. Commun. 2008, 10, 1191−4. (49) Kano, K.; Uno, B. Surface-Redox Reaction Mechanism of Quinones Adsorbed on Basal-Plane Pyrolytic Graphite Electrodes. Anal. Chem. 1993, 65, 1088−93. (50) Rudneva, A. V.; Pobelov, I. V.; Wandlowski, Th. Structural Aspects of Redox-Mediated Electron Tunneling. J. Electroanal. Chem. 2011, 660, 302−308. (51) Albrecht, T.; Guckian, A.; Kuznetsov, A. M.; Vos, J. G.; Ulstrup, J. Mechanism of Electrochemical Charge Transport in Individual Transition Metal Complexes. J. Am. Chem. Soc. 2006, 128, 17132−8. (52) Li, Z.; Liu, Y.; Mertens, S. F.; Pobelov, I. V.; Wandlowski, T. From Redox Gating to Quantized Charging. J. Am. Chem. Soc. 2010, 132, 8187−93. (53) Albrecht, T.; Moth-Poulsen, K.; Christensen, J. B.; Guckian, A.; Bjørnholm, T.; Vos, J. G.; Ulstrup, J. In Situ Scanning Tunnelling Spectroscopy of Inorganic Transition Metal Complexes. Faraday Discuss. 2006, 131, 265−79. (54) Nagy, G.; Wandlowski, T. Double Layer Properties of Au(111)/ H2SO4 (Cl)+Cu2+ from Distance Tunneling Spectroscopy. Langmuir 2003, 19, 10271−10280. (55) Collier, C. P.; Wong, E. W.; Belohradsky, M.; Raymo, F. M.; Stoddart, J. F.; Kuekes, P. J.; Williams, R. S.; Heath, J. R. Electronically Configurable Molecular-Based Logic Gates. Science 1999, 285, 391−4. (56) Esaki, L. New Phenomenon of Narrow Germanium p−n Junctions. Phys. Rev. 1958, 109, 603−604. (57) Kuznetsov, A. M. Negative Differential Resistance and Switching Behavior of Redox-Mediated Tunnel Contact. J. Chem. Phys. 2007, 127, 084710. (58) Chen, F.; Nuckolls, C.; Lindsay, S. In Situ Measurements of Oligoaniline Conductance: Linking Electrochemistry and Molecular Electronics. Chem. Phys. 2006, 324, 236−243. (59) Chen, J.; Wang, W.; Reed, M. A.; Rawlett, A. M.; Price, D. W.; Tour, J. M. Room-Temperature Negative Differential Resistance in Nanoscale Molecular Junctions. Appl. Phys. Lett. 2000, 8, 1224−1226. (60) Seminario, J. M.; Zacarias, A. G.; Tour, J. M. Theoretical Study of a Molecular Resonant Tunneling Diode. J. Am. Chem. Soc. 2000, 122, 3015−20. (61) He, J.; Fu, Q.; Lindsay, S.; Ciszek, J. W.; Tour, J. M. Electrochemical Origin of Voltage-Controlled Molecular Conductance Switching. J. Am. Chem. Soc. 2006, 128, 14828−35.

17461

dx.doi.org/10.1021/jp405516z | J. Phys. Chem. C 2013, 117, 17451−17461