The Effect of Water on Electron Transfer through Conductive Oligo

Dec 8, 2010 - Center for Bio-Molecular Science and Engineering, Naval Research Laboratory, Washington, DC 20375, United States, Departments of Mathema...
0 downloads 8 Views 4MB Size
22710

J. Phys. Chem. C 2010, 114, 22710–22717

The Effect of Water on Electron Transfer through Conductive Oligo(phenylene vinylene) Quinones Nikolai Lebedev,*,† Igor Griva,‡ Gary S. Kedziora,§ Anders Blom,| and Joel M. Schnur⊥ Center for Bio-Molecular Science and Engineering, NaVal Research Laboratory, Washington, DC 20375, United States, Departments of Mathematical and Computational and Data Sciences, George Mason UniVersity, Fairfax, Virginia 22030, United States, High Performance Technologies, Wright Patterson Air Force Base, Ohio 45433-7802, United States, QuantumWise A/S, Lersø Parkalle´ 107, DK-2100 Copenhagen, Denmark, and College of Science, George Mason UniVersity, Fairfax, Virginia 22030, United States ReceiVed: September 16, 2010; ReVised Manuscript ReceiVed: October 29, 2010

Electron transfer (ET) through oxidized (QOPV) and reduced-protonated (HQOPV) forms of oligo(phenylene vinylene) quinone placed between two gold electrodes in the absence and presence of external water molecules is calculated using density functional theory with a nonequilibrium Green’s function method. The results show that the presence of hydrogen atoms as an internal component (in reduced-protonated form of the molecule) screens the stimulating effect of oxygen on QOPV conductivity and eliminates the efficiency of the molecule conductance at low bias voltages. The formation of a complex with water restores the HQOPV conductivity at low biases and opens several additional strong conductivity channels below the Fermi level, substantially improving the efficiency of ET between QOPV and the electrode, especially at longer distances between the molecule and the metal. This effect of water can be utilized in the construction of novel highly efficient electrochemically gated electronic devices. It also opens a possibility for the fine regulation of direction of ET in “soft” molecular electronic devices with flexible organization and in biological systems. Introduction The construction of novel soft/flexible and bioinspired electronic and sensing devices requires a fundamental understanding of molecular-electrode electron transfer (ET) in solutions.1 Solute molecules can be considered as a flexible soft gate electrode controlling the conductivity of an active molecule attached to the source and drain.2,3 So far, the theoretical analysis of molecular conductance was performed mainly for molecules in vacuum.4 Meanwhile, experiments show that interaction with solvent can substantially change the molecular ET parameters.1-5 Among the solvents, water has a very special position. First, it has a big dipole moment able to affect solute molecular electronic structure.5,6 Then it can form strong hydrogen bonds allowing for specific interaction and stabilization of solute-solvent complexes. In addition, water itself can form conductive channels between the electrodes and between the solute and the electrode. In the only few papers available in the literature on the effect of water on molecular conductance, water is treated as a solvent shell with statistical averaging of its effects.5-8 Meanwhile, specific interactions of water with particular atoms in the solute molecule can substantially change the mechanism and efficiency of molecular conductance.4,9 One of the promising molecules in molecular electronic devices is oligo(phenylene vinylene), OPV.10 Recently we have shown that the introduction of oxygen atoms in the OPV

molecule (substitution of one of the phenyl rings with quinone) leads to a substantial increase in the molecular conductivity.3,11,12 To identify the mechanism of this effect in the present work we calculate ET through OPVQ and its reduced and protonated derivative (HQOPV) placed between two gold electrodes in the absence and presence of water molecules. The structure of the complexes between water and QOPV/HQOPV is found from the first principles using energy minimization with a density functional theory (DFT) approach. The electron transmission through the complexes is calculated using NEGF. In these calculations, we consider only the specific interaction of water with QOPV or HQOPV. The solvent second shell effects have been shown to be negligible5,6,13 and are excluded from our considerations. Our results show that the specific interaction of water with the oxygens of the quinone headgroup of both oxidized (QOPV) and reduced-protonated (HQOPV) forms of the molecule leads to the activation of ET, especially though HQOPV at low bias voltages. This interaction also opens additional strong transmission channels below Fermi level switching the complex conductance from n-type to p-type. These new channels considerably improve the ET through (H)QOPVwater complexes allowing for their efficient conductivity at longer distances between the molecule and the electrode. They also allow for the switching of the direction of ET pathway at a single molecular level. Methods

* To whom correspondence should be addressed. E-mail: nikolai.lebedev@ nrl.navy.mil. † Naval Research Laboratory. ‡ Departments of Mathematical and Computational and Data Sciences, George Mason University. § Wright Patterson Air Force Base. | QuantumWise A/S. ⊥ College of Science, George Mason University.

The initial optimization of the isolated molecular structures was performed using a DFT approach with the nonlocal exchange-correlation functional comprised of Becke’s threeparameter exchange functional and the Lee-Yang-Parr correlation functional (B3LYP) with a 6-31G(d) basis,3,11 using the Gaussian03 software.14 The optimized molecular structures

10.1021/jp108868z  2010 American Chemical Society Published on Web 12/08/2010

Effect of Water on Electron Transfer

J. Phys. Chem. C, Vol. 114, No. 51, 2010 22711

Figure 1. Optimized structures and molecular orbitals (MOs) of QOPV, HQOPV, and their water complexes. Only frontiers and some watercoupling MOs are shown. Em is a midpoint potential between the highest-occupied molecular orbital (HOMO) and lowest-unoccupied molecular orbital (LUMO). The colored arrows in the structures indicate the vectors of molecular dipole moments (all are in the plane of the molecules).

(parts A and B of Figure 1) were placed between two semiinfinite metallic electrodes having the atomic distances and angles of the gold cluster similar to those in a gold crystal with the Fm3m j group symmetry (space group number 225).15 The ET properties of the molecule in the junction were calculated using the nonequilibrium Green’s function (NEGF) approach.16 The optimization of the constructed molecular two-probe periodic structure was done using the Perdew-Burke-Ernzerhof (PBE) exchange correlation functional with the double-ζ basis set for the metal and the double-ζ polarized basis set for the other atom build in the ATK software.16,17 Compared to free molecules, the optimization using slabs representing the electrodes leads to some change in the dihedral angle between the gold-sulfur benzene carbons. For the cases with the water molecules, the water molecules do not move from the hydrogen bonding region with the quinone oxygens. In addition, the interaction with electrodes does not change the relative molecular energy level spacing.3 The surface cell size was 4 × 4 gold atoms, and two surface layers on each semi-infinite bulk gold electrode were included in the scattering area. The thiol group of the OPVQ was placed at the top of a trigonal pyramid with three gold atoms forming the base and with the sulfur atom 2.1 Å away from each of the gold atoms.18 The molecule was oriented normal to the electrode surfaces (Figure 2), and the number of Brillouin zone k-points were taken as 4,4,100 (with 100 k-points is in the direction of ET). The distance between two gold electrode was 25 Å; that corresponds to the shortest distance between the molecule (the hydrogen of quinone headgroup) and the right electrode equal to 4.13 Å. This means that the molecule is covalently bonded to the left electrode but weakly coupled to the right electrode. The current was calculated for a range of applied bias voltages using the Landauer-Bu¨ttiker formula19

I(V) )

∫ T(E, V)[nf(E - µL) - nf(E - µR)]dE

where nf is the temperature-dependent Fermi function and µL and µR are the electrochemical potentials of the left and right

electrodes, respectively. The total transmission at each energy is given by20,21

T(E, V) ) Tr[ΓLGΓRG+] where we note that the transmission probability is a function of both energy and the applied bias for each converged state of the entire system. G and G+ are the retarded and advanced Green’s functions, which describe the dynamics of the electrons for the central scattering region that includes the OPVQ molecule and the two layers of gold for each electrode.22 ΓL and ΓR are broadening matrices, which describe the strength of the coupling of the electrodes to the conducting scattering region. They are calculated as the imaginary parts of self-energy matrices of the left and right electrodes. We use the total transmission spectrum and the density of states to interpret the conductance properties of our systems. Though the Fermi energy is calculated independently for each electrode-molecule combination we note that its variation induced by the molecules is not more than 0.06 eV. All calculations are done for the same (H)QOPV isoform with an axis connecting oxygen atoms aligned parallel to the double bond between the phenyl rings. In our calculations we ignore any change in the surrounding medium. All dissipative/phase-breaking processes are assumed to be limited to the contacts. We also assume that the molecular orbitals are delocalized well enough so that the system does not fall in the Coulomb blockade regime. We are modeling ballistic conductance, which roughly corresponds to having delocalized conducting channel orbitals in or near in energy to the voltage bias window centered at the Fermi level.22 The NEGF method self-consistently matches the orbital energies of the molecule in the scattering region with the band energies of the biased gold electrodes. Furthermore, since we are using the real-space SIESTA method for the DFT,23 we increase the radial cutoff of the atomic orbital basis set and placed the molecule close enough to the electrodes to be sure that the atom-centered basis functions extend far enough to allow conduction through

22712

J. Phys. Chem. C, Vol. 114, No. 51, 2010

Lebedev et al.

Figure 2. Q (red) and Q + H2O (blue) electron transmission spectra (ETS, central panel) and molecular projected eigenstates of the ETS peaks (side panels) along with the structures of Q and Q + H2O between gold electrodes (bottom panels). The energies of the projected eigenstates are indicated above the images. The green dashed line with horizontal scale at the top of the graph shows the ETS for the molecule without water when the distance between the molecule and the right electrode is 1.72 Å.

vacuum and to give a better description of the surface dipole of the gold electrodes.24 Results and Discussion Effect of QOPV Reduction-Protonation on Molecular Conductance. Our electrochemical STM experiments have shown that the reduction-protonation of QOPV and the presence of water substantially affects the molecular conductance.2,3 Following them, in the present work we analyze the effect of water on both oxidized (QOPV) and reduced-protonated (HQOPV) forms of QOPV. To this end, we perform our calculations for the molecules that are well separated from the right electrode allowing the space for the positioning of water. The difference between QOPV and HQOPV lies in two additional electrons and two additional protons localized at the oxygen atoms.25 The optimized structures of these two molecules are shown in Figure 1 (structures I and III). The calculations show that in the optimized structures the presence of two additional hydrogens does not change (within ∼4%) the total length of the molecules. Meanwhile, the molecular energy diagram (Figure 1) shows that the reduction-protonation leads to a considerable increase in the LUMO-HOMO band gap and shifts these frontier MOs and the corresponding midpoint

potential (Em) to the higher energy (from -4.39 to -3.56 eV). It also results in a better LUMO delocalization throughout the HQOPV molecule. These changes also lead to the reorientation of the molecular dipole moments (Figure 1). To analyze the effect of the reduction-protonation on the ET ability of QOPV and HQOPV we calculate their ETS. Surprisingly, the spectra show that the narrow transmission band around the Fermi level, which is typical for QOPV, is missing from HQOPV (Figures 2 and 3). The other ETS bands are present, and their relative intensities and positions change only slightly. The disappearance of the narrow band is not due to the increase in the distance to the right electrode since the band is absent even if the molecule is placed close to the right electrode (Figure 3). Also, similar calculations performed for QOPV show no substantial distance dependence of the ETS (Figure 2). This effect is opposite to the results observed in our electrochemical STM experiments. To understand the origin of this effect we perform similar calculations in the presence of water. Role of Water in OPVQ/HQOPV Molecular Conductance. The optimized structures of the QOPV-H2O and HQOPV-H2O complexes are shown in Figure 1 (structures II and IV). Both structures have a single sharp global energy minimum in the optimization indicating the only positions of water. In these,

Effect of Water on Electron Transfer

J. Phys. Chem. C, Vol. 114, No. 51, 2010 22713

Figure 3. HQ (red) and HQ + H2O (blue) ETS (central panel) and molecular projected eigenstates of the ETS peaks (side panels) along with the structures of HQ and HQ + H2O between gold electrodes (bottom panels). The energies of the projected eigenstates are indicated above the mages. The energies of projected eigenstates are indicated above the images. The green dashed line with horizontal scale at the top of the graph shows the ETS for the molecule without water when the distance between the molecule and the right electrode is 1.72 Å.

the two water molecules are located next to the quinone oxygens and have hydrogens in between. The distance between the water and quinone oxygens is about 3 Å allowing for efficient formation of hydrogen bonds. In both cases, the presence of water substantially increases the molecular dipole moment, although without changing its orientation (Figure 1, bottom). The calculated energy diagrams and MOs show that the band gap and the spatial delocalization of the frontiers MOs are similar for the molecules with and without water, but their Em shifts to the lower energies in QOPV-H2O and to the higher energies in HQOPV-H2O, compared to nonhydrated molecules (Figure 1). At the same time, substantial changes in molecular electronic structures are seen at the area of the potential of water HOMO (-7.924 eV). For both QOPV and HQOPV association with water leads to a splitting of the corresponding MOs indicating strong electronic coupling (Figure 1). However, if in QOPV-H2O the coupling does not substantially change the energy of the coupled nonbonding MOs, in HQOPV-H2O this coupling leads their broader separation (by 2.44 eV; data not shown). In addition, well-delocalized MOs with unchanged energy that partially penetrate into water can be identified in the HQOPV-H2O.

A strong hybridization between QOPV and HQOPV with H2O is seen around the water HOMO level (Figure 1). This hybridization leads to a slight lowering of their energies as well as MO delocalization between the water and the quinone headgroup (Figure 1). Meanwhile, comparing HQOPV-H2O to QOPV-H2O shows that for HQOPV-H2O the effect is more complex. It leads to the hybridization of at least three HQOPV MOs with the water HOMO (Figure 1). In all cases, the hybridization with water tends to shift MO spatial distribution toward the quinone end of the molecules. To analyze the ET abilities of QOPV and HQOPV in the presence of water, we calculate their ETS. The calculation reveals that the presence of water does not considerably affect the QOPV-H2O narrow transmission band (as well as the most of other bands); it just slightly shifts its position to the lower energy (Figure 2). At the same time, interaction with water induces the formation of new strong transmission bands below the Fermi level (Figure 2). The position of these bands below the Fermi level indicates their p-type conductivity. Molecular projected eigenstates for these bands are similar to those identified for MOs of the QOPV-H2O complex (Figure 1) and include delocalization through the QOPV quinone head and both

22714

J. Phys. Chem. C, Vol. 114, No. 51, 2010

Lebedev et al.

Figure 4. Calculated (circles) and fitted with Lorentzian curves for the narrow transmission bands in the Q, Q-H2O, and HQ-H2O ETS.

TABLE 1: Lorentzian Fit and Calculated Parameters of ET through the Narrow Transmission Band in Q, Q-H2O, and HQ-H2O Q Q + H 2O HQ + H2O

εband, eV

ΓL × 10-5, eV

ΓR × 10-5, eV

KL × 1012, s-1

KR × 1012, s-1

+0.1802 +0.0760 -0.1312

693.34 885.21 2062.1

0.7420 1.0080 1.1918

10.53 13.45 31.33

0.011 0.015 0.018

water molecules. Similar strong bands appear in HQOPV-H2O (data not shown). They are consistent with MO calculations for the free HQOPV-water complex (Figure 1) and located at lower energy. The most intriguing result is the effect of water on the HQOPV-H2O ETS. First, the presence of water leads to the restoration of a narrow transmission band that is missing from HQOPV (Figure 3). The width and intensity of this restored band is compatible with that observed for QOPV with and without water. The spatial distribution of the projected eigenstate of this band is also similar to that calculated for QOPV. Meanwhile, the position of the band below the Fermi level indicates a shift from n-type to p-type conductivity. To estimate the role of molecular coupling to the electrodes in the formation of the narrow transmission band we analyze its shape. In all three cases where this band is seen, the shape of the narrow band is fitted well by a Lorentzian curve11

T(E) ) ΓLΓR/[(E - εBAND)2 + (ΓL + ΓR)2/4] where E is the energy relative to the Fermi level, εBAND is the position of the band, T(E) is a local approximation of the transmission spectra, and ΓL and ΓR are electronic couplings to the left and right electrode, respectively (Figure 4); in all cases the two coupling coefficients (high and low) can be identified (Table 1). One of the coefficients is in the range 700-2000 × 10-5 eV (with the corresponding ET rate between the molecule and the electrode, Γ/p ) 10-30 × 1012 s-1). This value is similar to the estimated earlier for the molecule positioned near the electrodes.11 It can be ascribed to the QOPV, QOPV-H2O, or HQOPV-H2O coupling to the left electrode, to which the molecule is chemically bound. The other coefficient in the pairs is about 3 orders of magnitude lower (Table 1). It could be ascribed to the right electrode that is separated from the molecule by ∼4 Å. This assumption is in line with our previously calculations showing that the increase in the distance between a molecule and an electrode from a close position to about 4 Å leads to a decrease in the current by a factor of 1000.11 The comparison of QOPV with QOPV-H2O indicates that the formation of complex between the molecule and water does not substantially change the coupling through the narrow band

(it slightly increases the coupling to both electrodes) consistent with the absence of water MOs in eigenfunctions of this band. Reduction-protonation of QOPV increases the coupling by 2-3 times. Interestingly enough, a more pronounced effect of protonation is seen for the stronger coupling (due to interaction with the left electrode). The position of the narrow band shifts slightly to the lower level as the result of both protonation and hydration (Table 1). To identify the origin of the strong transmission bands appearing below the Fermi level in the presence of water, we calculate their projected eigenstates (Figure 3). Contrary to the narrow transmission band around the Fermi level, all these bands show delocalization between QOPV and water molecules indicating participation of water MOs in the ET. Their spatial delocalization is similar to that identified for the molecules without electrodes (Figure 1) indicating intermolecular interaction as the main factor of their generation. Moreover, these MOs show clear penetration into the right electrode, indicating the possibility for efficient ET from QOPV moiety through the water to the right electrode (Figure 2, right panel). Previously we obtained data indicating that the ET path operated in QOPV by the narrow transmission channels goes directly from a carbon atom in the quinone ring to the right electrode.11 In accordance with this, analysis of the molecular projected eigenstate performed in the present work for the narrow transmission band demonstrates that water MOs are not involved in this ET path. On the other hand, the spatial distribution of the MOs involved in the strong broad transmission channel below the Fermi level that appears in the presence of water clearly shows involvement of water MOs in their generation. The presence of shoulders in the strong p-type transmission band appearing in QOPV-H2O ETS indicates the possibility of existence of not one but several overlapping transmission bands (Figure 2). To evaluate that, we perform deconvolution of this band into sets of Lorentzian curves (Figure 5). The best fit is obtained with four bands of various intensities having widths from 0.020 to 0.075 eV. The number of these bands is similar to the number of MOs generated as the result of coupling between QOPV and waters (Figure 1). Although the strong overlapping of the bands does not allow for an analysis of their shape (and coupling to the electrodes), the half-width of each

Effect of Water on Electron Transfer

J. Phys. Chem. C, Vol. 114, No. 51, 2010 22715

Figure 5. Spectrum and Lorentzian deconvolution of a broad ETS band at about -1.5 eV in QOPV-H2O. The distribution of residuals is shown in the top panel.

TABLE 2: Lorentzian Fitting Parameters of a Broad Band at about -1.5 eV in Q-H2O amplitude position width

1

2

3

4

0.0134393 -1.56052 0.0339762

0.00258016 -1.50036 0.0427543

0.000948329 -1.45984 0.0194328

0.00103065 -1.39448 0.0748944

of the bands is close to what can be expected for thermal broadening (kBT ) 0.026 eV) at 300 K. All these overlapping bands have similar projected eigenstates with specific localization on the quinone headgroup and water MOs with strong delocalization between them and the right electrode (Figure 2). Similar MOs are seen in the QOPV-H2O complex without electrodes, indicating their origin to mainly be due to intermolecular interaction. ET Paths and Conditions for Their Operation. Previously we have shown that the ET path operated by the narrow transmission band in QOPV without water more likely goes directly through a carbon and a hydrogen of the quinone ring to the right electrode.11 We have also shown that quinone oxygens are necessary for the generation of this path. Comparison of molecular structures in the cases where the narrow transmission band appears (QOPV, QOPV-H2O, or HQOPVH2O) with those where it is missing (HQOPV) indicates that in all conductive cases an unshielded oxygen atom faces the ET path and/or the right electrode (Figures 2 and 3, bottom panels). To evaluate the mechanism of oxygen-induced activation of this path and to identify the paths for the strong p-type transmission bands in (H)QOPV-water complexes we calculate the spatial distribution of local electrostatic potentials. The spatial distribution of the electrostatic potential for QOPV-H2O placed between two gold electrodes at zero bias shows that the increasing the distance between the molecule and the right electrode leads to a dramatic increase in the ET barrier between the molecule and the right electrode (Figure 6). This is consistent with substantial reduction in the intensity of the narrow transmission band in the ETS spectrum. On the other hand, the presence of water reduces the barrier between the quinone headgroup and the right electrode by introducing several additional paths going through the water oxygen, either directly or via a water hydrogen (Figure 6). These paths have

different heights and widths indicating that the efficiency of ET through each of them can vary. The presence of these paths with barriers of different height and width correlates with the presence of several sub bands of various intensities in ET spectrum of this molecule at about -1.5 eV below the Fermi level (Figure 5). It is also consistent with electron delocalization between quinone and water in the projected eigenstates and the strong coupling of these states with the right electrode (Figure 2). To evaluate the efficiency of these paths in ET through the QOPV-H2O complex we calculate the molecular electrostatic potentials at different bias voltages. The results indicate that an increase in the bias voltage has two types of effects. First, it reduces the barrier height, and, second, it changes the potential to the left and the right sides of the barrier. The first effect indicates that the efficiency of ET through the barrier increases with the potential, as it should be for a conductive path. The second effect shows that the different electrodes control the potential on the left and the right side of the barrier. It means that despite the electron delocalization between the quinone and water molecules the potential of (H)QOPV when a bias voltage is applied is determined by the right electrode, while the potential of the water is determined by the left electrode. This result is in a good agreement with our previous estimation of the contribution of left and right electrodes in the control of the QOPV potential when the molecule is in close proximity to both electrodes.11 Thus, the high efficiency of ET through water is due to the presence of several ET channels allowing for their parallel operation and extended MO delocalization shortening the distance for through-space tunneling. Improving Efficiency of ET through QOPV by Water. The obtained results allow for a prediction of the I-V dependence of an electronic device constructed from (H)QOPV. For dry molecules at low source-drain biases (between 0.4 and 2.5 V), a turning off conductivity is expected when the molecule is reduced-protonated (Figure 7, dashed curve). Instead, for wet devices at these biases the reduction protonation will induce an increase in conductivity due to ET through the narrow conductive channel (Figure 7, solid curves). This coincides well with our direct measurement of HQOPV and QOPV conductivities in electrochemical STM experiment at these bias potentials.3 At biases between 3 and 4 V, when the ET paths through water molecules becomes operational, the difference in conductivity for the wet configuration will be orders in magnitude (Figure 7, inset). Conclusion In conclusion, our results show that the reduction and screening of the oxygen atoms of the QOPV quinone headgroup by internal hydrogens eliminate the ET through HQOPV at low biases. The presence of water suppresses this screening effect and restores the conductivity of the molecule at low biases. Moreover, the formation of a complex between QOPV and water molecules fuses some MOs of QOPV and HQOPV with the water HOMO which leads to the generation of several strong transmission bands below the Fermi level. This switches the mechanism of ET through QOPV from n-type to p-type and increases the efficiency of the ET between QOPV or HQOPV and the electrode allowing for their operation at longer distances from the surface. In this regime, with slightly increased bias voltages, the conductivity of QOPV-H2O surpasses the conductivity of the other forms by more than ten times. In addition to the increasing efficiency, the fusion of the molecule and water

22716

J. Phys. Chem. C, Vol. 114, No. 51, 2010

Lebedev et al.

Figure 6. (A) Spatially resolved electrostatic potential of QOPV-H2O between two gold electrodes at zero bias. (B) Potential profiles for the directions shown in (A). The letters below the profiles indicate atoms for profiles 1 (red) and 5 (light blue). The path through the rest of the molecule (black S to C) is the same in all cases.

MOs changes the pathway of the ET, potentially allowing for its redirection. Acknowledgment. This work was supported in part by the Air Force Office of Scientific Research and Naval Research Laboratory base programs. Supporting Information Available: Chemical structures of the molecules and complexes used in the work are available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) van der Molen, S. J.; Liljeroth, P. J. Phys.: Condens. Matter 2010, 22.

Figure 7. I-V curves for the conductivity of QOPV (red) and HQOPV (blue) between two gold electrodes in the absence (dashed) and presence (solid) of water. The high current at increased biases is shown in the inset.

(2) Tsoi, S.; Griva, I.; Trammell, S. A.; Blum, A. S.; Schnur, J. M.; Lebedev, N. ACS Nano 2008, 2, 1289. (3) Tsoi, S.; Griva, I.; Trammell, S. A.; Kedziora, G.; Schnur, J. M.; Lebedev, N. Nanotechnology 2010, 21. (4) Nichols, R. J.; Haiss, W.; Higgins, S. J.; Leary, E.; Martin, S.; Bethell, D. Phys. Chem. Chem. Phys. 2010, 12, 2801. (5) Rungger, I.; Chen, X.; Schwingenschlogl, U.; Sanvito, S. Phys. ReV. B 2010, 81. (6) Cardamone, D. M.; Kirczenow, G. Nano Lett. 2010, 10, 1158. (7) Cao, H.; Jiang, J.; Ma, J.; Luo, Y. J. Am. Chem. Soc. 2008, 130, 6674. (8) Leary, E.; Hobenreich, H.; Higgins, S. J.; van Zalinge, H.; Haiss, W.; Nichols, R. J.; Finch, C. M.; Grace, I.; Lambert, C. J.; McGrath, R.; Smerdon, J. Phys. ReV. Lett. 2009, 102. (9) Kula, M.; Jiang, J.; Lu, W.; Luo, Y. J. Chem. Phys. 2006, 125.

Effect of Water on Electron Transfer (10) Kushmerick, J. G.; Holt, D. B.; Pollack, S. K.; Ratner, M. A.; Yang, J. C.; Schull, T. L.; Naciri, J.; Moore, M. H.; Shashidhar, R. J. Am. Chem. Soc. 2002, 124, 10654. (11) Lebedev, N.; Griva, I.; Trammell, S. A.; Kedziora, G. S.; Tender, L. M.; Schnur, J. J. Phys. Chem. C 2010, 114, 12341. (12) Trammell, S. A.; Lowy, D. A.; Seferos, D. S.; Moore, M.; Bazan, G. C.; Lebedev, N. J. Electroanal. Chem. 2007, 606, 33. (13) Cao, H.; Ma, J.; Luo, Y. Nano Research 2010, 3, 350. (14) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson,

J. Phys. Chem. C, Vol. 114, No. 51, 2010 22717 B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, revision B.05; Gaussian, Inc.: Wallingford, CT, 2004. (15) Gosavi, S.; Marcus, R. A. J. Phys. Chem. B 2000, 104, 2067. (16) Taylor, J.; Brandbyge, M.; Stokbro, K. Phys. ReV. Lett. 2002, 89. (17) Atomistix ToolKit version 2008. 10; QuantumWise A/S 2008. (18) Zhou, Y. X.; Jiang, F.; Chen, H.; Note, R.; Mizuseki, H.; Kawazoe, Y. Phys. ReV. B 2007, 75. (19) Buttiker, M.; Imry, Y.; Landauer, R.; Pinhas, S. Phys. ReV. B 1985, 31, 6207. (20) Brandbyge, M.; Mozos, J. L.; Ordejon, P.; Taylor, J.; Stokbro, K. Phys. ReV. B 2002, 65. (21) Jauho, A. P.; Wingreen, N. S.; Meir, Y. Phys. ReV. B 1994, 50, 5528. (22) Datta, S. Quantum Transport: Atom to Transistor; Cambridge University Press: Cambridge, 2009. (23) Stokbro, K.; Taylor, J.; Brandbyge, M.; Mozos, J. L.; Ordejon, P. Comput. Mater. Sci. 2003, 27, 151. (24) Lindsay, S. M.; Ratner, M. A. AdV. Mater. 2007, 19, 23. (25) Trammell, S. A.; Lebedev, N. J. Electroanal. Chem. 2009, 632, 127.

JP108868Z