E

Sa

b

c

a

ARRAA

KSEUNE

1

ctdttrladA(awsatc

B

0h

Electrochimica Acta 110 (2013) 741– 753

Contents lists available at ScienceDirect

Electrochimica Acta

jo u r n al hom ep age: www.elsev ier .com/ locate /e lec tac ta

lectrochemically-gated single-molecule electrical devices

haoyin Guoc, Juan Manuel Artésb, Ismael Díez-Péreza,b,∗

Physical Chemistry Department, University of Barcelona, Martí i Franqués 1-11, 08028 Barcelona, SpainInstitute for Bioengineering of Catalonia (IBEC), Baldiri Reixac 15-21, 08028 Barcelona, SpainCenter for Bioelectronics and Biosensors, Biodesign Institute at Arizona State University, Tempe, AZ 85287, USA

r t i c l e i n f o

rticle history:eceived 14 December 2012eceived in revised form 25 March 2013ccepted 26 March 2013vailable online 4 April 2013

eywords:ingle-molecule junctionslectrochemical gatenipolar/ambipolar FETs

a b s t r a c t

In the last decade, single-molecule electrical contacts have emerged as a new experimental platformthat allows exploring charge transport phenomena in individual molecular blocks. This novel tool hasevolved into an essential element within the Molecular Electronics field to understand charge transportprocesses in hybrid (bio)molecule/electrode interfaces at the nanoscale, and prospect the implemen-tation of active molecular components into functional nanoscale optoelectronic devices. Within thisarea, three-terminal single-molecule devices have been sought, provided that they are highly desiredto achieve full functionality in logic electronic circuits. Despite the latest experimental developmentsoffer consistent methods to bridge a molecule between two electrodes (source and drain in a transistornotation), placing a third electrode (gate) close to the single-molecule electrical contact is still tech-

DRlectrochemical switches

nically challenging. In this vein, electrochemically-gated single-molecule devices have emerged as anexperimentally affordable alternative to overcome these technical limitations. In this review, the operat-ing principle of an electrochemically-gated single-molecule device is presented together with the latestexperimental methodologies to built them and characterize their charge transport characteristics. Then,an up-to-date comprehensive overview of the most prominent examples will be given, emphasizing onthe relationship between the molecular structure and the final device electrical behaviour.

. Introduction

Back in 1974, Aviram and Ratner firstly proposed a theoreti-al single-molecule device with one molecule connected betweenwo metallic beads that could rectify the current as in a standardiode configuration [1]. This was the kickoff of the single-moleculeransport studies in the rapidly growing field of Molecular Elec-ronics and, since then, large efforts were put in experimentallyealizing a single-molecule electrical contact. The beginning ofast decade brought us the first experimental examples [2–8],nd to date, different strategies have been designed to repro-ucibly obtain single-molecule conductance signatures [6,9–11].mong all these methodologies, Scanning Tunnelling Microscopy

STM) has been proven to be a powerful tool to in situ creatend analyze thousands of single-molecule contacts and come outith statistically meaningful values for the conductance of the

ingle-molecule device [9]. Generally, in order to reliably bridge

molecule between two electrodes and characterize its chargeransport, several conditions have to be met: (i) provide stablehemical anchoring between the two electrodes and the molecule

∗ Corresponding author at: Physical Chemistry Department, University ofarcelona, Martí i Franqués 1-11, 08028 Barcelona, Spain. Tel.: +34 93 4033707.

E-mail address: isma diez@ub.edu (I. Díez-Pérez).

013-4686/$ – see front matter © 2013 Elsevier Ltd. All rights reserved.ttp://dx.doi.org/10.1016/j.electacta.2013.03.146

© 2013 Elsevier Ltd. All rights reserved.

[11–21], (ii) identify signatures linked to the formation of a single-molecule bridge [9,22,23] and (iii) design experimental protocols tostatistically analyze the observed single-molecule conductance sig-natures [9,24–26]. Points (i) and (iii) have been particularly decisivefor the advance of this field; on the one hand, different anchor-ing chemistry has allowed to stably held molecules between twomacroscopic electrodes through strong covalent chemical bondssuch as metal-S [12,13] and, more recently, metal-C [20]. On theother hand, statistically meaningful results of single-molecule con-ductance values have been provided, which is a requirement inview of the observed dispersion in conductance results from deviceto device [9,24]. The dispersion mainly stems from the uncertaintyin the atomic details of the molecule–electrode contact geometry[27]. The improved statistical methodologies in measuring single-molecule conductance have prompted extensive studies regardinghow charge transport in single molecular devices is affected bymolecular chemistry [28–33] and conformation [25,34–36]. Thesemeasurements, together with theoretical efforts, have providedthe bulk of our current knowledge about electron transport inmolecules. For further revision on single-molecule charge trans-port characterization on simple two-terminal devices, we address

the reader to recent comprehensive reviews [37–41] and books[42–44].

The main block of experimental work on single-molecule trans-port has been developed in two-terminal device configurations,

742 S. Guo et al. / Electrochimica A

Fig. 1. Schematic representations of three-terminal single-molecule devices with(A) a solid-state back gate and (B) an electrochemical gate. Dot lines in (A) representelectric field lines and encircled signs in (B) represent solvated anions (negative sign)ai

mbbdmmb

review, we will introduce the main operating parameters of

Fmb

S

nd cations (positive sign) in a polar solvent. EDLs at the source-drain/electrolytenterfaces in (B) are omitted for simplicity.

eaning devices with one unique molecule bridging between twoiased (Vbias) electrodes. Although a huge advance in the field haseen achieved with such a simple configuration, three-terminalevices (Fig. 1A) are highly desirable when it comes to the require-

ents of the actual technological applications; not only activeolecular components generating specific electrical behaviours,

ut also logic device functionalities [45]. To date, measurements

ig. 2. Transition voltage spectroscopy (TVS) to evaluate gate efficiency in three-termiarked by arrows represent the transition bias voltage Vtrans. (B) Top panel represents sim

arrier at Vtrans. Bottom graph shows the linear relationship between the transition and g

ource: Reprinted from [48] by permission from Macmillan Publishers Ltd. Copyright 200

cta 110 (2013) 741– 753

on single-molecule three-terminal devices are limited, owing tothe experimental difficulty of placing a third electrode in closeproximity to a single-molecule electrical contact. In this vein,pioneer works performed at cryogenic temperatures by Park [8]and Bjørnholm [46] demonstrated the experimental feasibility ofadding a third electrode into a single-molecule electrical device andmeasuring gate-dependent transport. Despite the experimentaladvances in designing solid-state three-terminal single-moleculedevices during the last decade [47–49], their microfabricationinvolves fairly complex clean room processes, which significantlydiminishes the success rate of fabrication. Moreover, an importantlimitation of solid-state three-terminal single-molecule devices istheir low gate efficiency (˛), being defined as the ratio betweenthe molecular orbital energy shift (EHOMO/LUMO) and the actualapplied gate voltage (Vg). By measuring ˛, one can evaluate theamount of electric field from the gate electrode that is actuallyfelt by the confined molecule (Fig. 1A). is typically evaluatedfrom the slope of the transition voltage (Vtrans) as a function of Vg

(Fig. 2B, bottom panel), where Vtrans is extracted from the mini-mum in the Transition Voltage Spectroscopy (TVS, Fig. 2A) [50,51].TVS plots display a minimum (Vtrans) that represents an electrontransfer transition from tunnelling to a Fowler-Nordheim or fieldemission regime (Fig. 2B, top panel) [50]. Fig. 2B shows val-ues of 0.25, which means that 25% of the gate field only reachesthe single molecule device, i.e. to shift the molecular energy levelsinvolved in charge transport by 250 mV, at least 1 V of gate voltageis needed. This constitutes one of the largest reported values insuch devices, which typically hits much lower values in the order of≤0.1 [52].

Electrochemically-gated three-terminal single-moleculedevices (Fig. 1B) offer an alternative to overcome previousdifficulties in the fabrication and operation of their solid-state homologues. In such devices, the molecular junction isimmersed in an electrolyte and the gate potential is appliedthrough the electrochemical double layer developed at thedifferent electrodes/electrolyte interfaces. In the present

an electrochemically-gated single-molecule device and willillustrate the most recent advances through the latest reportedexamples.

nal single-molecule devices: (A) TVS plots at different gate voltages. The minimaplified band diagrams of the electron transfer process through a square tunnelling

ate voltages. The slope is the gate efficiency.

9.

ica A

2

ematstsrgst(tpetefictptmgt(ipFDoed

cttmmposb

tbtmpisTgecfihelai

S. Guo et al. / Electrochim

. Electrolyte gating effect in single-molecule devices

Fig. 1B shows a schematic representation of a single-moleculelectrical contact with an electrochemical gate. The single-olecule junction is immersed in an electrolyte that can be either

n aqueous environment or a polar organic solvent, e.g. acetoni-rile containing milli-molar concentrations of a dissolved organicalt that increases the dielectric constant and decreases the resis-ance across it. A third electrode (counter) immersed within theame electrolyte acts as the gating electrode to modulate the cur-ent flowing through the molecule. The way the electrochemicalate is implemented in the actual experimental setup is by using atandard electrochemical bipotentiostat so that independent elec-rochemical potentials can be applied to both junction electrodesFig. 1: W1 and W2 or source (S) and drain (D) in a transistor nota-ion, Section 4). The Vbias will be defined as the electrochemicalotential difference between the W1 and W2 electrodes. When anlectrochemical potential is applied to the single-molecule junc-ion, an electrochemical double layer (EDL) capacitor builds up atvery electrode/electrolyte interface. The two plates of this inter-acial capacitor are composed by the charged electrode interfacen one side, and the solvated ions in solution counteracting thisharge in the other (Fig. 1B). The gate voltage applied throughhe reference electrode (for now on electrochemical (echem) gateotential) will drop across the two main EDLs; the one formed athe gate electrode/electrolyte interface and the one at the metal-

olecule-metal/electrolyte interface (Fig. 1B). The electrochemicalate potential will entirely fall across the EDLs, which correspondso an effective gate thickness of the order of a few solvated ionsencircled +/− signs in Fig. 1B), thus resulting in a large highly local-zed gate field. The electrochemical gate principle has been alreadyroven to be very efficient (with values close to 1) in tuning theermi energy of electrodes in solid-state EDL transistors [53,54].espite this, today’s microelectronics industry still relies mostlyn solid-state gates to build FETs, which highlights the need ofxploring new methods to implement EDL-FETs into real functionalevices.

From the setup sketched in Fig. 1B, a screening effect issuean be inferred; in order to avoid total or partial screening ofhe electrochemical gate potential at the single-molecule junc-ion, the EDL must be completely developed along the main

olecular junction axis, i.e. the junction gap imposed by theolecule length must be longer than the typical outer Helmholtz

lane distance [37,55,56]. This fact may explain the absencef gate effect observed in single-molecule contact built withhort (<1 nm) molecular wires such as 4-4′-bipyridine or 1,4-enzenedithiol [56].

Another limitation of the electrochemical gate approach regardso the narrow potential gate window imposed by the stability ofoth the solvent and anchoring chemistry. At high gate poten-ials, electrochemical oxidation/reduction of the solvent as well as

olecule/electrode bond can take place. In aqueous solutions, theotential range of electrochemically-gated single-molecule exper-

ments have been typically limited to roughly ±0.8 V versus atandard Ag/AgCl reference electrode at nearly neutral pHs [55–59].he use of room-temperature ionic liquids (RTIL) as electrolyticate has significantly improved this limitation thanks to theirnhanced electrochemical stability [60,61], spanning the electro-hemical gate potential range up to ±few volts [62,63]. Otheractors that may limit the electrochemical gate potential rangen electrochemically-gated single-molecule measurements are theigh capacitive currents and/or secondary faradaic processes gen-

rated at the source/drain electrodes surface. This issue has beenargely alleviated by developing methods to minimize the exposedrea of the source/drain electrodes to the gating electrolyte: (i)n a STM configuration, the current is limited by the smaller

cta 110 (2013) 741– 753 743

source/drain electrode, i.e. a sharp metallic probe, which can beeffectively insulated from the electrolyte by using a number ofpolymers [64–66]. (ii) In other configurations where both sourceand drain electrodes are equivalent in size, such as in electro-chemical Mechanically Controllable Break-Junction (MCBJ) setups,fairly complex microfabrication methods are commonly employedto minimize the source/drain electrodes size down to the �m2 level[67].

3. Measuring transport in single-molecule devices

As introduced above, electrochemically-gated single-moleculedevices have been experimentally realized by using mainly two dif-ferent experimental setups: an electrochemical MCBJ [67] and anelectrochemical STM [12,56]. Both systems use piezo-transducersable to control the gap distance between two electrodes downto the sub-nanometer level, thus allowing the accommodationof a molecule in between them, provided the right molecularanchoring chemistry (Fig. 1). While in the STM configuration, thelarge electrode surface and sharp metal probe are being usedas the source/drain electrodes [56], in a MCBJ setup, two con-fronted metal tips do the same function [48,67]. Both setups areequipped with a small electrochemical cell, where the whole elec-trode/molecule/electrode junction is immersed in the workingelectrolyte. Immersed in the same liquid cell, a long Pt wire typ-ically sewed around the liquid enclosure serves as the counterelectrode, and a miniaturized Ag/AgCl electrode as the referencefor electrochemical potential control.

Next, we will describe the latest protocols employed toidentify single-molecule bridge formations and characterize gate-controlled transport.

3.1. AC assisted DC-blinking method: fishing molecules

Fig. 3A illustrates a regular DC-blinking experiment performedwith an electrochemical STM configuration. The two electrodes arebrought to a constant gap distance of the order of the length ofthe target molecule. The approaching is achieved with the helpof a piezo-transducer connected to one of the two electrodesand a feedback loop to the tunnelling current flowing betweenthem. Once the distance is set, the feedback loop is opened andthe DC current is monitored. Stochastic events corresponding tosingle-molecule bridge formations are observed as sudden (ON-OFF) jumps in the DC current signal [22], which identifies theexact moment of the single-molecule junction formation. Thismethod has been recently complemented with a simultaneousAC modulation stage [23,32]. A small sinusoidal AC mechanicalperturbation of ∼1 kHz is applied to the piezo transducer so thatthe electrode–electrode distance is continuously modulated. AnAC current component can be then measured on top of the DCcurrent signal due to the electrode–electrode gap modulation. Con-trarily to the DC current signal, when a molecule is suddenlybridged between the two electrodes, the AC current responseexperiences a drop in amplitude (Fig. 3B), which is explained bythe constrain imposed by the rigid molecular backbone [23,68].Simultaneous monitoring of both DC and AC current responseswill give us a univocal signature of single-molecule bridgesformation.

The lifetime of some of the ON-states in Fig. 3A often extendsup to the second scale under electrochemical conditions, depend-ing upon the mechanical stability of the working setup, which

provides enough time to prompt typically several electrochemicalgate potential ramps (few 100 ms of duration each) and measurecharge transport characteristics before the single-molecule con-tact breaks down. The process can be automatized so that up to

744 S. Guo et al. / Electrochimica Acta 110 (2013) 741– 753

Fig. 3. Identifying the formation of single-molecule electrical contacts: (A) DC current transient during the formation/breakdown of a single-molecule bridge. TelegraphicO as (A)a

te

3

lsmifbc(

FpGstr

S

N-OFF in the DC-current signal serves to identify single-molecule events. (B) Same

mplitude drops of the AC-current signal.

ens of curves can be accumulated in a few hours of DC-blinkingxperiments.

.2. Break-junction assisted method

In Section 1, we introduced an STM-based method as a revo-utionary tool in the Molecular Electronics field that has broughttatistically meaningful values of the conductance of single-olecule devices [9]. The so-called break-junction method, today

mplemented in both STM and MCBJ configurations, is based on the

ast formation of thousands of single-molecule junctions that cane identified as plateaus in individual current vs. distance tracesollected during successive electrode–electrode separation cyclesFig. 4A, top panel). When such traces are added up into a histogram,

ig. 4. Characterizing single-molecule conductance: (A) summary of a regular break-junlateaus ascribed to single-molecule junction formations. Bottom graph is the resultin0 = 77.4 �S denotes conductance quantum units. (B) Break-junction assisted method to

hows a detail of an individual trace in (A). The coloured dots indicate the positions whereraces, hundreds of I(V)s are collected and represented in a two-dimensional histogram (beader is referred to the web version of this article.)

ource: Adapted with permission from [24]. Copyright 2011, American Chemical Society.

for the AC-assisted current method. Single-molecule events are observed as sudden

peaks show up corresponding to the conductance of single/multiplemolecular devices (Fig. 4A, bottom panel). We can now use this rou-tine to simultaneously collect hundreds of I(V) characteristics fromthe same molecular device. To this aim, automatic algorithms canbe developed to identify the early formation of a plateau (molecularbridge) and run several to few tens (depending upon experimen-tal conditions) of I(V) characteristics (coloured dots in Fig. 4B, toppanel) along a single current plateau [24]. This novel methodologyallows the collection of hundreds of I(V) curves of single-moleculedevices in a matter of tens of minutes, and the data is typically

represented in two-dimensional I(V) histograms (Fig. 4B, bottompanel). The break-junction assisted method is particularly usefulto record current versus electrochemical gate characteristics fromsingle-molecule devices (see Sections 4 and 5).

ction experiment: top panel shows individual retraction traces displaying currentg conductance histogram built out of thousands of individual (top panel) traces.record current–voltage characteristics from the single-molecule contact: top panel

individual I(V)s were recorded. By repeating the process for consecutive individualottom panel).(For interpretation of the references to color in this figure legend, the

ica A

4t

etetidmbct

dnidvsAmutmciplmw(tselid

4

efbrthmmtIsstvdcceproiO

S. Guo et al. / Electrochim

. Electrochemically-gated single-molecule field-effectransistors (FETs)

The building blocks of silicon-based microelectronics are fieldffect transistors (FET), whose basic function is controlling elec-rical current between two electrodes (named source and drainlectrodes) with a third electrode called gate. Under this scheme,he ability to control charge transport through a single molecules essential for the future development of molecular electronicevices. As introduced in Section 2, Fig. 1B sketches a single-olecule EDL-FET type of configuration, where the two electrodes

ridging the molecule are the source and drain electrodes and theounter electrode in the electrochemical configuration will func-ion as the gate electrode.

FET-like behaviour has been already demonstrated in nanoscaleevices such as carbon nanotubes [69,70], semiconductoranowires [71] and even on small graphene sheets [72,73]. Theoret-

cal studies predicted gate-voltage modulation of a single-moleculeevice conductance in a similar way of what is observed in con-entional solid-state FETs [74,75]. As pointed in the Section 1,olid-state single-molecule FET setups are experimentally tricky.lternatively, single-molecule EDL-FET devices have started to gainore protagonism, evolving in a number of interesting contrib-

tions within the last decade. In this chapter, we will reviewhe most prominent examples of electrochemically-gated single-

olecule FETs. In general, the current through the active molecularomponent in such devices is modulated either through energet-cally proximal low energy molecular orbitals (direct tunnellingrocesses), giving rise to linear conductance devices such as unipo-

ar/ambipolar FETs (Sections 4.1 and 4.2), or through reversibleolecular redox transitions (sequential two-electron processes),hich results in non-linear conductance devices displaying NDR

Section 4.3). Due to the larger mechanistic complexity of chargeransport in nanoscale molecular devices, as compare to their solid-tate homologues, different mechanisms have been often used toxplain the same single-molecule FET behaviour and some over-apping among mechanisms is frequently found. As a result, thellustrative examples presented in each section below will beescribed case-by-case.

.1. Unipolar

Unipolar FET behaviour has been observed in a number oflectrochemically-gated single-molecule devices built with rigidused conjugated backbones such as perylenes [55,57,76,77] andenzocoronenes [62], as well as with other well-known flexibleedox molecules like viologen [78,79]. Fig. 5 summarizes the chargeransport characteristics obtained for such devices. Perylene blocksave undoubtedly brought the largest maximum ON-OFF currentodulation (Imax/Imin ∼ 103) in an electrochemically-gated single-olecule device (Fig. 5A,B), and have prompted numerous studies

o better understand their fundamental operation [55,57,76,77].n general, the charge transport through electrochemically-gatedingle-perylene junctions is understood in analogy to an n-typeolid-state FET, where a net current increase is detected whenhe electrochemical gate potential is scanned towards negativealues. The exact mechanistic picture in the molecular deviceiffers though from the classical picture of the n-type chargearriers moving along the band diagram structure of the solid semi-onductor. When a single-molecule is constrained between twolectrodes, particular energy level positions of the lowest unoccu-ied (LUMO) and highest occupied (HOMO) molecular orbitals with

espect to the electrodes Fermi energies (EF) will arise as a resultf the molecule–electrodes hybridization. The final energy pos-tions will determine the charge transport type and the maximumN-OFF current modulation observed for a particular molecular

cta 110 (2013) 741– 753 745

junction. For single-molecule devices built of perylene tetra-caboxylic diimide (PTCDI) derivatives, the LUMO level is believedto energetically locate closer to the electrodes EF [56,76], beingresponsible for the observed n-type transport. This model has beensupported by first-principle calculations [76,80]. The fact that thecurrent increases monotonically with the applied gate potential,and no current maximum is reached, is explained by the largepolarization of the molecule induced by the strong gate electricfield [80], which results in a progressive shift of the molecu-lar energy levels as the electrochemical gate potential is mademore negative. As a consequence, the electron transfer mecha-nism through the molecule always occurs via the LUMO tail, sothat fully resonance conditions are never achieved. This particularmechanism may be a consequence of the predicted large molec-ular orbital hybridization with the supporting metal electrodesof the junction when thiolated anchor groups (T-PTCDI structurein Fig. 5A) are used [76]. The later reasoning is also experimen-tally supported by the fact that uncoupling (i.e. detaching) oneof the junction electrodes from the molecule results in a radi-cally different charge transport scenario displaying a maximum inthe tunnelling current [76], i.e. NDR effect, that will be treated inSection 4.3.

From an electrochemical point of view, accessing the PTCDILUMO energy level implies the electrochemical reduction ofthe molecule, giving rise to the corresponding radical anionPTCDI•–[81] (see also voltammetry studies in [55,57,76]). The factthat a transient radical anion may be formed during the chargetransport process foresees a strong dependence with the elec-trolytic environment, where a more polar electrolyte would helpto stabilize the charged molecular state. This is, indeed, exper-imentally evidenced when comparing transport measurementsdone in an aqueous environment (Fig. 5A) with those performedin a less polar organic medium such as acetonitrile (Fig. 5B). Inthe later case, the ON-OFF current ratio drops down to valuesof ∼10, two orders of magnitude less than in 5 A. Temperaturedependence transport on single-PTCDI junctions also highlightsthe importance of stabilizing the charged intermediate state in thetransport mechanism. While in an aqueous environment a markedtemperature-dependent conductance is measured, thus suggestingan incoherent two-step tunnelling process [82,83] as the dominat-ing transport mechanism, in a non-polar environment, a coher-ent temperature-independent tunnelling transport is observed[57].

Similar unipolar n-type transport behaviour has been alsoachieved using hexa-peri-hexabenzocoronene (HBC) blocks insingle-molecule EDL-FETs (Fig. 5C). Such devices display fairlylarge maximum ON-OFF current modulation ratios of ∼102 inorganic electrolytes [62]. The transport mechanism proposed hereis, however, different than the previously described for peryleneblocks. The significantly larger HOMO-LUMO energy gap in HBCblocks, ∼3.2 eV, as compare to ∼2.4 eV for unsubstituted PTCDI [56],leaves single-HBC junctions energetically far from full LUMO res-onance conditions within the experimental gate potential range.The observed n-type transport is then discussed in terms of res-onance with the large LUMO tail that is observed in the UV–visspectra of highly diluted solutions of the compound [62]. An inter-esting comparison between single-HBC FETs and graphene-basedFETs is pointed out in this work. High charge carrier mobility val-ues measured in large (�m length) graphene sheets have promisedrevolutionary applications in fast electronics [72], however, one ofthe main drawbacks remains in the low maximum ON-OFF currentratios measured in graphene-based FETs [73], mainly due to the

absence of band gap in the graphene energy band structure. HBCblocks are then presented as a bottom-up alternative to graphene-based FETs; large conjugated coronene blocks can be synthesizedwith precise size [84], giving us direct control over their electronic

746 S. Guo et al. / Electrochimica Acta 110 (2013) 741– 753

Fig. 5. Unipolar electrochemically-gated single-molecule FETs built with different molecular blocks: (A) and (B) devices built with perylene derivatives (PTCDI). The experi-ments were done in water and acetonitrile respectively. Open squares in (A) represent independent conductance values extracted from break-junction histograms (Fig. 4A).Grey shadow in (B) represents the experimental standard deviation. (C) Device built with a hexa-peri-hexabenzocoronene (HBC) block. A control experiment (open junction)is also shown. (D) Device built with a viologen derivative (6V6). The solid line is a mere visual guidance and the filled squares correspond to the experimental conductancevalues extracted from break-junction histograms.

S respeM

sf�

bavereitt(tmbasebaunTjtt

ource: (A), (B) and (D) were adapted from [55,79,91]. Copyrights 2005, 2012, 2008acmillan Publishers Ltd. Copyright 2010.

tructure and, therefore, over FETs performance, and benefitingrom the high charge carrier mobility given by the large delocalized-electron system.

A different example of unipolar single-molecule EDL-FET haseen exploited using redox viologen blocks (6V6, Fig. 5D) as thective molecular component [12,77–79]. The electrochemistry ofiologens has been extensively studied [85], showing two fast one-lectron reduction processes that go through the formation of theadical cation (6V6•+, Fig. 5D, bottom inset). The unipolar n-typelectrochemical gate behaviour observed for the single-6V6 devicen Fig. 5D has a completely different interpretation as compared tohe previous two examples. Despite the first electrochemical reduc-ion process occurs at relatively small electrochemical potentials∼ −0.45 V versus a reference SCE), full resonant tunnelling withhe LUMO energy level is, however, not observed, and the current

onotonically increases at electrochemical potentials <−0.45 V,eyond the redox reduction potential of the molecule. Nichols etl. came out with an elegant explanation based on the compari-on with other redox and non-redox backbones bearing the samexact thiolated anchor groups (Fig. 5D [79]): when the viologenlock is being electrochemically reduced, the torsional dihedralngle between the two pyridine rings will be changing from val-es around 34◦ [34], depending upon the ionic environment, to aear planar conformation (from bottom to top in Fig. 5D inset).

he exact tunnelling current flowing through the single-viologenunction as a function of the electrochemical gate potential will behen determined by the instantaneous nuclear configuration of itswisting mode.

ctively, American Chemical Society. (C) was adapted from [62] by permission from

4.2. Ambipolar

Unlike its unipolar counterpart whose conduction is dominatedby one type of charge carriers, electrons or holes, the conduction inan ambipolar FET can be dominated both by electrons and holes,depending on the gate voltage polarity. The unique ambipolarbehaviour in molecular devices promises new design principles forboth analog circuit and digital applications, and new opportunitiesfor one to understand the role of the HOMO and LUMO in the con-duction through the same molecule. Despite ambipolar behaviouris already well known in post-silicon devices such as nanotubes[86], graphene [73,87] and organic crystals [88], ambipolar single-molecule FETs are rare [46,47]. In order to achieve ambipolar effectin a molecular device, compounds with low HOMO-LUMO energygap are sought [89].

We have recently demonstrated ambipolar behaviour inan electrochemically-gated single-molecule FET by using apyrrolidine-substituted PTCDI block [90]. Pyrrolidine groups arestrong Lewis bases whose marked electron donation characterresults in both a substantial decrease of the perylene block HOMO-LUMO energy gap and an approaching of the HOMO level closerto the electrodes Fermi energies. These electronic effects aresupported by simple ab initio calculations and confirmed by voltam-metry and UV–vis experiments. Scanning the electrochemical gate

potential allows us to probe both HOMO and LUMO levels in thesingle-molecule device, giving rise to charge transport dominatedeither by holes or electrodes respectively (Fig. 6B). For visual guid-ance, Fig. 6A (same as 5B) is accompanied.

S. Guo et al. / Electrochimica Acta 110 (2013) 741– 753 747

Fig. 6. Ambipolar electrochemically-gated single-molecule FETs built with modified PTCDI derivatives: (A) same as 5B is added for visual comparison. The inset shows a controlexperiment from an open junction. (B) Ambipolar electrochemically-gated single-molecule FET built with a pyrrolidine-modified PTCDI block. The blue line corresponds toa KU model fit of the NDR peak observed in the p-type charge transport at positive electrochemical gate potentials.(For interpretation of the references to color in this figurelegend, the reader is referred to the web version of this article.)

S ciety.

tWilattctt

4

cndaipsiaefrtm

[dmipba

ource: Adapted with permission from [91]. Copyright 2012, American Chemical So

An interesting feature when comparing both HOMO and LUMOransports in Fig. 6B is the stark contrast between both mechanisms.

hile the electron transport (LUMO-related) displays a monotonicncrease of current, just as the one described in previous unipo-ar devices (Section 4.1), the hole transport (HOMO-related) shows

current maximum at an electrochemical gate potential close tohe anodic redox potential of the substituted-PTCDI, which is bet-er described by a sequential two-step tunnelling process (blueurve in 6B [90]). The latter mechanism constitutes a representa-ive example of NDR effect in a single-molecule device and will bereated in detail in the next section.

.3. Negative differential resistance

A region of decreasing current with increasing voltage in aurrent versus voltage characteristic of a device is referred to asegative differential resistance (NDR). The NDR effect was firstiscovered by Esaki [91], and inspired many applications suchs low-power memories. In an Esaki diode, the current initiallyncreases with the bias voltage as electrons tunnel through the–n junction barrier because the conduction band on the n-ide is aligned with the valence band on the p-side. As voltagencreases further the conduction and valence bands become mis-ligned and the current drops, resulting in NDR. Other devicesxhibiting NDR effect are semiconductor heterostructures con-orming quantum dots [92], where the maximum in current iseached when resonance conditions between the energy levels ofhe electrodes and the levels confined in the quantum well are

et.After the first examples of NDR in nanoscale molecular devices

5,93,94], several electrochemically-gated single-molecule devicesisplaying NDR have been presented [63,80,95–99]. Fig. 7A–C sum-arizes some of the most extensively studied cases. NDR behaviour

n a single-molecule device stands on different principles as com-are to its homologous solid-state device. The active molecularlock is typically a redox moiety with marked electrochemicalctivity, i.e. presenting a voltammetric signal with peaks assigned

to one or more oxidation/reduction states of the molecule. NDReffect is then explained as a tunnelling process mediated throughthe redox energy level of the molecule, which is typically locatedenergetically very close to the Fermi energy levels of the electrodes.Two different models have been successfully applied to describethe current maximum observed in the single-molecule currentversus electrochemical gate potential characteristics: a resonanttunnelling process proposed by Schmikler et al. [100,101] and asequential two-steps electron transport formulated by Kuznetsovand Ulstrup (KU) [82,83]. Both models rely in an electron transportmechanism assisted by the redox energy level of the molecule whenit lies within the applied bias window (Fig. 8). While the formerdescribes a coherent one-step tunnelling process that is assistedby the particular redox energy level of the molecule (Fig. 8A),the KU model proposes a sequential electron transfer where theelectron (or hole) first enters the molecule, which is momen-tarily reduced/oxidized, to subsequently leave the molecule aftera partial vibrational relaxation of the redox centre (Fig. 8B). Thisrelaxation process is directly related to the reorganization energy�, which gives an estimation of the energy difference betweenthe equilibrium redox potential and the energy levels of the oxi-dized/reduced molecular states (Fig. 8B). Independently of themechanistic details, both formalisms predict a bell-shape featurein the current versus electrochemical gate potential curves. A char-acteristic distinction between them can be intuitively deducedfrom Fig. 8: the position of the current maximum in the cur-rent vs. gate potential matches the formal redox potential in theKU case, while it appears to be shifted by ∼� in the former case[63,77].

Interesting examples of single-molecule NDR behaviourhave been recently observed in electrochemically-gated single-metalloprotein junctions such as juctions with blue copper Azurin(Fig. 7D [98]) and cytochrome b562 [99]. Single-protein junctions

are key tools to understand charge transport processes in lifeand to pave the way towards the design of bio-electronic devicesthat profit from biomolecular features such as the long-range elec-tron/energy transfer [102].

748 S. Guo et al. / Electrochimica Acta 110 (2013) 741– 753

Fig. 7. NDR effect in electrochemically-gated single-molecule FETs built with different molecular blocks: (A) device built with a pyrrolo-tetrathiafulvalene (pTTF) block. Thesolid line corresponds to a Gaussian fit. (B) Device built with a bipyridyl-dinitro oligophenylene–ethynelene (BPDN) block. (C) Device built with a polyaniline (PANI) block.T nce. (Dm er is r

( 12 Am

etrtFpteasta

Fa

he corresponding voltammetric signal has been embedded in blue as visual guidaodel fit.(For interpretation of the references to color in this figure legend, the read

A)–(D) adapted with permission from [79,97–99]. Copyrights 2008, 2006, 2005, 20

Multiple NDR features can also be observed inlectrochemically-gated devices built with pyrrolo-etrathiafulvalene (pTTF) blocks. TTF compounds display aich redox activity with multiple one-electron processes goinghrough the radical cation pTTF+•, in analogy with the 6V6 block inig. 5D. In aqueous environment, only the first oxidation processTTF → pTTF+• is accessible within the electrochemical range ofhe thiol bond stability in this medium (Fig. 7A [78,79]). Nicholst al. enlarged the working potential range by using RTIL, getting

ccess to the full TTF oxidation [63]. The result is a remarkableingle-molecule EDL-FET with two separate current maxima inhe transport characteristics (Fig. 9), which foresees excitingpplications in logic molecular devices.

A

Vbias

Utip US

vo

vo

Molecule

e-

EF

Eref

Eox

Ered

Resonant Tunnelling

E

Substrate STM tip

ig. 8. Schematic band diagrams of the electrode–molecule(redox)–electrode junction sknd (B) a sequential two-step (KU model) electron transport.

) Device built with a Cu-Azurin metalloprotein. The solid line corresponds to a KUeferred to the web version of this article.)

erican Chemical Society.

5. Electrochemically-gated single-molecule switches

Molecular Switches is a fascinating and extensive scientific area[103]. In this field, the optical/electrical/mechanical bi-stabilityof a molecular block is exploited to create a logic ON-OFF func-tionality in a molecular device. Robust examples of molecularelectrical switches have been already presented, being the cate-nanes/rotaxanes one of the most studied [104].

Contrarily to the single-molecule FET setup described in pre-

vious section, in a molecular electrical switch the current ismodulated as a result of a quasi-reversible chemical transformationof the active molecular component. The irreversible character ofthe process implies that the two molecular states involved in the

B

Vbias

Utip US

vo

vo

Molecule

e-

EF

e-

ref

2

Eox

Ered

2-step electron transport

Substrate STM tip

etching two different electron transfer scenarios: (A) a resonant tunnelling process

S. Guo et al. / Electrochimica A

Fig. 9. NDR effect in an electrochemically-gated single-PTCDI device displaying twocurrent maxima. Both peaks have been fitted to KU models (coloured solid lines).The experiment was performed using an ionic liquid as the gating electrolyte.(Forinterpretation of the references to color in this figure legend, the reader is referredto the web version of this article.)

SS

stthttt

ea

F(p

Aoc

ource: Adapted with permission from [63]. Copyright 2012, American Chemicalociety.

witching process are stable upon chemical reaction, and that a cer-ain energy barrier is imposed to reverse the reaction and return tohe initial state. This behaviour often results in the appearance of aysteresis loop in the transport characteristics. In a molecular elec-rical switch, the equivalent maximum ON-OFF switching ratio ofhe device will be determined by the difference in conductance of

he two molecular states.

Here we will describe the most extensively studiedlectrochemically-gated single-molecule switches, whosectuation involves changes in molecular conductance as a

ig. 10. Electrochemically-gated single-molecule switches built with different molecularOPE-NO2). The solid line has been added as a visual guidance. The inset displays a Hammetarameter.

dapted from [56] with permission of The Royal Society of Chemistry (RSC). Copyright 200f a single-molecule switch based on the electrochemical reduction of diazonium anchoorrespond to conductance values from break-junction histograms. Error bars are represe

cta 110 (2013) 741– 753 749

function of an electrochemically-induced molecular redoxtransformation.

5.1. Hammett parameter (�)

The conductance of a single-molecule junction can be largelychanged by the electron-withdrawing nature of particular chemi-cal substituents in the active molecular backbone [28,95]. The trendhas been correlated to the so-called Hammett parameter (�) thatdescribes the electronic effect that a chemical substituent has in therate of an organic chemical reaction [105]. The nice inverse linearrelationship obtained in the single-molecule conductance versus �plot (Fig. 10A inset) evidences a clear correlation between the tun-nelling energy barrier and the electron-withdrawing character ofthe substituent (increasing �) [28,56]. This trend is in good agree-ment with previous theoretical estimations by Vedova-Brook et al.[106].

Tao and co-workers exploited this effect to fabricate anelectrochemically-gated single-molecule switch by using a nitro-substituted oligo(phenylene ethynylene) (OPE-NO2) molecularblock (Fig. 10A [95]). When a negative electrochemical gate poten-tial is applied to the single-OPE-NO2 junction, the nitro group( NO2) is electrochemically reduced to amine ( NH2) with a muchlower � value (Fig. 10A inset), thus increasing the single-moleculedevice conductance. The in situ formation of an OPE-NH2 block atnegative gate potentials is demonstrated by acidifying the elec-trolyte pH so that the protonation of the amine group NH3

+

resulted again in a larger � with the corresponding conductancedecrease.

This type of electrical switching can be understood as apotential-induced chemical switching and may be exploitedin other molecular backbones with redox active chemicalsubstitutions.

backbones: (A) device built with a nitro-substituted oligo(phenylene ethynylene)t plot of the single-molecule conductance versus the chemical substituent Hammett

5 RSC. (B) An anthraquinone (AQ)-based single-molecule switch. (C) Representationring groups. (D) Charge transport plot of the device sketched in (C). Filled squaresnted by experimental standard deviations.

7 ica A

5

citcpepeebebaRtiec

afbttcbeOsic

5

gor

meitdl

Fntardtmmtttmsrtt

50 S. Guo et al. / Electrochim

.2. Quantum interference

Quantum interference (QI) effects have been argued to explainhanges in the conductance of single-molecule contacts using sim-lar conjugated backbones. A representative example correspondso the differences in conductance measured for the same fusedonjugated ring when the anchor groups are placed at differentositions [107,108]. The effect is based on the wave nature oflectrons travelling along the molecule through different electronathways, which can, in some cases, result in destructive interfer-nces and give rise to nodes (or anti-resonant dips) at particularnergies in the transmission function [109]. Anthraquinone (AQ)-ased nanoscale molecular devices [110–114] are one prominentxample of QI effects in molecular transport. AQ blocks appear toe cross-conjugated along the keto oxygens in their central rings opposed to its linearly conjugated homologue anthracene (Ac).ecently, experimental results supported by theoretical calcula-ions have evidenced that cross-conjugation is responsible of a noden the transmission function due to QI effects [109,110], whichxplains the conductance drop in the AQ case versus the linearlyonjugated Ac [110–112].

Darwish et al. have recently exploited the redox activity ofn AQ block as a system that can be electrochemically switchedrom cross-conjugated to linearly conjugated in a single-moleculeridge (Fig. 10B). An applied negative electrochemical gate poten-ial results in the reduction of the central benzoquinone groupo hydrobenzoquinone, going from a low conductance cross-onjugated to a high conductance linearly conjugated molecularlock [113]. The result is a single-molecule switch based onlectrochemically controlled QI [113,114]. As in the previousPE example, the pH-dependence switching behaviour of the

ingle-molecule contact corroborates that the observed changen conductance is indeed due to the quinone–hydroquinoneonversion [113].

.3. Molecule/electrodes coupling

We have recently designed a new class of electrochemically-ated single-molecule switches based on the controlled formationf electrodes-molecule bonds through specific electrochemicaleactions with the molecular anchoring groups [115].

Diazonium groups ( N2+) can be electrochemically reduced to

olecular nitrogen leaving a radical at the same position. The gen-ration of such radicals in the vicinity of an electrode surface resultsn the formation of a stable covalent bond (Fig. 10C), a methodhat has been largely employed to functionalize solid surfaces withifferent chemical groups [116], and recently exploited to build

arge-scale molecular junctions [117].The newly developed switching concept represented in the

ig. 10C exploits the electro-reduction of two diazonium termi-al groups to controllably form single-molecule junctions betweenwo electrodes [115]. The single-molecule device can operate as

fast electrochemically-gated switch, where the ON state will beeached at negative electrochemical gate potentials where the bis-iazonium compound is being reduced to form covalent bondso both junction electrodes. Covalent attachment of the single-

olecule translates into a higher electrical coupling between theolecule and the electrodes and, therefore, high ON conduc-

ance values. At more positive electrochemical gate potentials thanhe redox reduction potential of the diazonium terminal groups,he molecule–electrodes attachment will go through low energy

etal-diazonium interactions resulting in a low OFF conductance

tate (Fig. 10D). Due to the high stability of the C(sp2)-Au bondesulting from the diazonium reduction, large reverse gate poten-ial pulses must be used to reverse the reaction and go back tohe OFF switching state. This particular switching configuration

cta 110 (2013) 741– 753

achieved through sharp variations of the molecule–electrodes cou-pling provides large ON-OFF switching ratios in the device currentwell over 103.

6. Conclusions

This review provides a summary of the latest studies onelectrochemically-gated single molecule contacts. After introduc-ing the operation principles of the EDL gate in a molecular device,a short survey through the latest experimental methodologies tobuild and characterize transport through single-molecule contactsis given. EDL is presented as a more efficient alternative to con-trol charge transport in a single-molecule electrical contact, whichcan be easily implemented through a standard bipotentiostaticcontrol of the junction electrodes. The large gate voltage efficien-cies achieved with this method allow using lower gate potentialranges for device operation, thus preserving the stability of theelectrode/molecule/electrode bridge.

Electrochemically-gated single-molecule contacts display avariety of FET behaviours resembling well-known solid-statedevices. The operation mechanisms are, however, very different.Unipolar/ambipolar FET behaviour is achieved by using redoxmolecular blocks displaying fast, reversible oxidation/reductiontransitions. A generalized model has not been found yet to fullydescribe the unipolar transport and a case-by-case casuistry isemployed. However, a general reasoning based on a coherent tun-nelling process with partial HOMO/LUMO resonant conditions withthe electrodes Fermi levels, is commonly used.

NDR effect is a better-understood phenomenon inelectrochemically-gated single-molecule devices. Molecularblocks with well-characterized oxidation/reduction states aretypically employed. The observed current maximum in the trans-port characteristics has been successfully described by a two-steptunnelling process with partial relaxation of the redox centre(KU model), where the molecule experiences a momentarilyoxidation/reduction transformation at every electron transfercycle.

Electrochemically-gated single-molecule switches are builtupon very different fundamental principles. First examplesexploited molecular blocks with redox active substituents. In thesecases, the switching mechanism of the single-molecule devices isbased on the oxidation/reduction of the substituent group thatdirectly affects its electron withdrawing character and modifies theelectronic structure of the entire molecular backbone. The Ham-mett parameter best describes the observed conductance changeinduced by the chemical transformation of the substituent.

More recently, quantum interference effects have been used todesign an electrochemically-gated single-molecule switch. To thisaim, electrochemical transition of the active molecular compoundfrom cross-conjugated to linearly conjugated results in low andhigh conducting single-molecule electrical contacts respectively.

Finally, a new class of electrochemically-gated single-moleculeswitches is presented, based on the electrochemical control ofthe molecule–electrodes binding. The electrochemical reduction ofdiazonium anchoring groups is used to covalently bridge the molec-ular compound to the two junction electrodes, and switch betweenlow and high single-molecule conductance states with remarkablyhigh maximum ON-OFF current ratio.

Electrochemically-gated single-molecule devices offer an excel-lent platform to understand charge transport in nanoscalemolecular devices. Although more theoretical efforts are needed

to fully understand the mechanisms behind, the already presentedexperimental results reveal a completely open field with a widefan of possibilities regarding structure-dependent molecular deviceoperations.

ica A

A

wRft2f

R

S. Guo et al. / Electrochim

cknowledgements

The authors acknowledge Prof. Nongjian Tao and Dr. Nadim Dar-ish for useful comments on the manuscript. I.D-P. thanks theamon y Cajal programme and the national grant (CTQ2012-36090)

rom the Spanish Ministry of Economy and Competitiveness, andhe EU International Reintegration Grant (FP7-PEOPLE-2010-RG-77182) for financial support. J.M.A. acknowledges a fellowshiprom the Generalitat de Catalunya (BE-DGR 2009).

eferences

[1] A. Aviram, M.A. Ratner, Molecular rectifiers, Chemical Physics Letters 29(1974) 277.

[2] M.A. Reed, C. Zhou, C.J. Muller, T.P. Burgin, J.M. Tour, Conductance of a molec-ular junction, Science 278 (1997) 252.

[3] H.-W. Fink, C. Schönenberger, Electrical conduction through DNA molecules,Nature 398 (1999) 407.

[4] J.K. Gimzewski, C. Joachim, Nanoscale science of single molecules using localprobes, Science 283 (1999) 1683.

[5] D.I. Gittins, D. Bethell, D.J. Schiffrin, R.J. Nichols, A nanometre-scale electronicswitch consisting of a metal cluster and redox-addressable groups, Nature408 (2000) 67.

[6] X.D. Cui, A. Primak, X. Zarate, J. Tomfohr, O.F. Sankey, A.L. Moore, T.A. Moore, D.Gust, G. Harris, S.M. Lindsay, Reproducible measurements of single-moleculeconductivity, Science 294 (2001) 571.

[7] D. Porath, A. Bezryadin, S. de Vries, C. Dekker, Direct measurements of elec-trical transport through DNA molecules, Nature 403 (2000) 635.

[8] W.J. Liang, M. Shores, M. Bockrath, J.R. Long, H. Park, Kondo resonance in asingle-molecule transistor, Nature 417 (2002) 725.

[9] B.Q. Xu, N.J. Tao, Measurements of single-molecule resistance by repeatedformation of molecular junctions, Science 301 (2003) 1221.

[10] T. Dadosh, Y. Gordin, R. Krahne, I. Khivrich, D. Mahalu, V. Frydman, J. Sper-ling, A. Yacoby, I. Bar-Joseph, Measurements of the conductance of singleconjugated molecules, Nature 436 (2005) 677.

[11] F. Moresco, L. Gross, M. Alemani, K.H. Rieder, H. Tang, A. Gourdon, C. Joachim,Probing the different stages in contacting a single molecular wire, PhysicalReview Letters 91 (2003) 36601.

[12] W. Haiss, H. van Zalinge, S.J. Higgins, D. Bethell, H. Höbenreich, D.J. Schiffrin,R.J. Nichols, Redox state dependence of single molecule conductivity, Journalof the American Chemical Society 125 (2003) 15294.

[13] F. Chen, X. Li, J. Hihath, Z.F. Huang, N.J. Tao, Effect of anchoring groupson single-molecule conductance: comparative study of thiol-, amine-, andcarboxylic-acid- terminated molecules, Journal of the American ChemicalSociety 128 (2006) 15874.

[14] Y.S. Park, A.C. Whalley, M. Kamenetska, M.L. Steigerwald, M.S. Hybertsen, C.Nuckolls, L. Venkataraman, Contact chemistry and single-molecule conduc-tance: a comparison of phosphines, methyl sulphides, and amines, Journal ofthe American Chemical Society 129 (2007) 15768.

[15] S.T. Schneebeli, M. Kamenetska, Z. Cheng, R. Skouta, R.A. Friesner, L.Venkataraman, R. Breslow, Single-molecule conductance through multi-ple (-(-stacked benzene rings determined with direct electrode-to-benzenering connections, Journal of the American Chemical Society 133 (2011)2136.

[16] A. Mishchenko, L.A. Zotti, D. Vonlanthen, M. Burkle, F. Pauly, J.C. Cuevas,M. Mayor, T. Wandlowski, Single-molecule junctions based on nitrile-terminated biphenyls: a promising new anchoring group, Journal of theAmerican Chemical Society 133 (2011) 184.

[17] W. Hong, D.Z. Manrique, P. Moreno-García, M. Gulcur, A. Mishchenko, C.J.Lambert, M.R. Bryce, T. Wandlowski, Single molecular conductance of tolanes:experimental and theoretical study on the junction evolution dependent onthe anchoring group, Journal of the American Chemical Society 134 (2012)2292.

[18] F. von Wrochem, D.Q. Gao, F. Scholz, H.G. Nothofer, G. Nelles, J.M. Wessels,Efficient electronic coupling and improved stability with dithiocarbamate-based molecular junctions, Nature Nanotechnology 5 (2010)618.

[19] S. Ahn, S.V. Aradhya, R.S. Klausen, B. Capozzi, X. Roy, M.L. Steigerwald, C.Nuckolls, L. Venkataraman, Electronic transport and mechanical stabilityof carboxyl linked single-molecule junctions, Physical Chemistry ChemicalPhysics 14 (2012) 13841.

[20] Z.-L. Cheng, R. Skouta, H. Vazquez, J.R. Widawsky, S. Schneebeli, W. Chen, M.S.Hybertsen, R. Breslow, L. Venkataraman, In situ formation of highly conduct-ing Au–C contacts for single molecule junctions, Nature Nanotechnology 6(2011) 353.

[21] T. Masateru, T. Makusu, S. Kohei, F. Hideki, K. Tomoji, Single-molecule junc-tions with strong molecule–electrode coupling, Journal of the American

Chemical Society 131 (2009) 14146.

[22] W. Haiss, R.J. Nichols, H. van Zalinge, S.J. Higgins, D. Bethell, D.J. Schiffrin,Measurement of single molecule conductivity using the spontaneous for-mation of molecular wires, Physical Chemistry Chemical Physics 6 (2004)4330.

cta 110 (2013) 741– 753 751

[23] J.L. Xia, I. Díez-Pérez, N.J. Tao, Electron transport in single molecules measuredby a distance-modulation assisted break junction method, Nano Letters 8(2008) 1960.

[24] S. Guo, J. Hihath, I. Díez-Pérez, N.J. Tao, Measurements and statistical analysisof single-molecule current–voltage characteristics, transition voltage spec-troscopy, and tunnelling barrier height, Journal of the American ChemicalSociety 133 (2011) 19189.

[25] I. Díez-Pérez, J. Hihath, T. Hines, Z.-S. Wang, G. Zhou, K. Müllen, N.J. Tao, Con-trolling single molecule conductance through lateral coupling of �-orbitals,Nature Nanotechnology 6 (2011) 226.

[26] P. Makk, D. Tomaszewski, J. Martinek, Z. Balogh, S. Csonka, M. Wawrzyniak,M. Frei, L. Venkataraman, A. Halbritter, Correlation analysis of atomic andsingle-molecule junction conductance, ACS Nano 6 (2012) 3411.

[27] X. Li, J. He, J. Hihath, B.Q. Xu, S.M. Lindsay, N.J. Tao, Conductance of single alka-nedithiols: conduction mechanism and effect of molecule–electrode contacts,Journal of the American Chemical Society 128 (2006) 2135.

[28] L. Venkataraman, Y.S. Park, A.C. Whalley, C. Nuckolls, M.S. Hybertsen, M.L.Steigerwald, Electronics and chemistry: varying single-molecule junctionconductance using chemical substituents, Nano Letters 7 (2007) 502.

[29] X. Xiao, B.Q. Xu, N.J. Tao, Measurement of single molecule conductance: ben-zenedithiol and benzenedimethanethiol, Nano Letters 4 (2004) 267.

[30] S.H. Choi, C. Risko, M.C.R. Delgado, B. Kim, J.-L. Brédas, C.D. Frisbie, Transitionfrom tunneling to hopping transport in long, conjugated oligo-imine wiresconnected to metals, Journal of the American Chemical Society 132 (2010)4358.

[31] T. Hines, I. Díez-Pérez, J. Hihath, H.M. Liu, Z.-S. Wang, J.W. Zhao, G. Zhou, K.Müllen, N.J. Tao, Transition from tunneling to hopping in single molecularjunctions by measuring length and temperature dependence, Journal of theAmerican Chemical Society 132 (2010) 11658.

[32] I. Díez-Pérez, J. Hihath, Y. Lee, L. Yu, L. Adamska, M.A. Kozhushner, I. Oleynik,N.J. Tao, Rectification and stability of a single molecular diode with controlledorientation, Nature chemistry 1 (2009) 635.

[33] R. Yamada, H. Kumazawa, T. Noutoshi, S. Tanaka, H. Tada, Electrical conduc-tance of oligothiophene molecular wires, Nano Letters 8 (2008) 1237.

[34] L. Venkataraman, J.E. Klare, C. Nuckolls, M.S. Hybertsen, M.L. Steigerwald,Dependence of single molecule junction conductance on molecular confor-mation, Nature 442 (2006) 904.

[35] A. Mishchenko, D. Vonlanthen, V. Meded, M. Bürkle, C. Li, I.V. Pobelov, A.Bagrets, J.K. Viljas, F. Pauly, F. Evers, M. Mayor, T. Wandlowski, Influence ofconformation on conductance of biphenyl-dithiol single-molecule contacts,Nano Letters 10 (2010) 156.

[36] J.S. Meisner, M. Kamenetska, M. Krikorian, M.L. Steigerwald, L. Venkataraman,C. Nuckolls, A single-molecule potentiometer, Nano Letters 11 (2011) 1575.

[37] N.J. Tao, Electron transport in molecular junctions, Nature Nanotechnology 1(2006) 173.

[38] F. Chen, J. Hihath, Z. Huang, X. Li, N.J. Tao, Measurements of single-moleculeconductance, Annual Review of Physical Chemistry 58 (2007) 535.

[39] F. Chen, N.J. Tao, Electron transport on single molecules: from benzene tographene, Accounts of Chemical Research 42 (2009) 429.

[40] M.-P. Kasper, B. Thomas, Molecular electronics with single molecules in solid-state devices, Nature Nanotechnology 4 (2009) 551.

[41] S. Karthaeuser, Control of molecule-based transport for future moleculardevices, Journal of Physics: Condensed Matter 23 (2011) 013001.

[42] J. Tomfohr, G.K. Ramachandran, O.F. Sankey, S.M. Lindsay, Making contactsto single molecules: are we there yet? in: G. Cuniberti, G. Fagas, K. Richter(Eds.), Introducing Molecular Electronics, 2, Springer, Germany, 2005, p. 301(Chapter 11).

[43] S.M. Lindsay, Introduction to Nanoscience, Oxford University Press, New York,2010.

[44] J.C. Cuevas, E. Scheer, Molecular Electronics: An Introduction to Theory andExperiments, World Scientific Publishing Co. Pte. Ltd, Singapore, 2010.

[45] S. Ami, M. Hliwa, C. Joachim, Molecular ‘OR’ and ‘AND’ logic gates integratedin a single molecule Chemical Physics Letters 367 (2003) 662.

[46] S. Kubatkin, A. Danilov, M. Hjort, J. Cornil, J.-L. Bredas, N. Stuhr-Hansen,P. Hedegard, T. Bjørnholm, Single-electron transistor of a single organicmolecule with access to several redox states, Nature 425 (2003) 698.

[47] Y.-V. Kervennic, J.M. Thijssen, D. Vanmaekelbergh, R. Dabirian, L.W.Jenneskens, C.A. van Walree, H.S.J. van der Zant, Charge transport inthree-terminal molecular junctions incorporating sulfur-end-functionalizedtercyclohexylidene spacers, Angewandte Chemie International Edition 118(2006) 2602.

[48] H. Song, Y. Kim, Y.H. Jang, H. Jeong, M.A. Reed, T. Lee, Observation of molecularorbital gating, Nature 462 (2009) 1039.

[49] C.A. Martin, J.M. Ruitenbeek, H.S.J. van der Zant, Sandwich-type gatedmechanical break junctions, Nanotechnology 21 (2010) 265201.

[50] J.M. Beebe, B. Kim, J.W. Gadzuk, C.D. Frisbie, J.G. Kushmerick, Transition fromdirect tunnelling to field emission in metal–molecule–metal junctions, Phys-ical Review Letters 97 (2006) 026801.

[51] J.M. Beebe, B. Kim, C.D. Frisbie, J.G. Kushmerick, Measuring relative barrierheights in molecular electronic junctions with transition voltage spec-troscopy, ACS Nano 2 (2008) 827.

[52] Z.K. Keane, J.W. Ciszek, J.M. Tour, D. Natelson, Three-terminal devices toexamine single-molecule conductance switching, Nano Letters 6 (2006) 1518.

[53] H. Yuan, H. Liu, H. Shimotani, H. Guo, M. Chen, Q. Xue, Y. Iwasa, Liquid-gatedambipolar transport in ultrathin films of a topological insulator Bi2Te3, NanoLetters 11 (2011) 2601.

7 ica A

52 S. Guo et al. / Electrochim

[54] J.T. Ye, S. Inoue, K. Kobayashi, Y. Kasahara, H.T. Yuan, H. Shimotani, Y. Iwasa,Liquid-gated interface superconductivity on an atomically flat film, NatureMaterials 9 (2010) 125.

[55] B.Q. Xu, X.Y. Xiao, X.M. Yang, L. Zang, N.J. Tao, Large gate current modulationin a room temperature single molecule transistor, Journal of the AmericanChemical Society 127 (2005) 2386.

[56] X. Li, B.Q. Xu, X. Xiao, X. Yang, L. Zang, N.J. Tao, Controlling charge transport insingle molecule using electrochemical gate, Faraday Discussions 131 (2006)111.

[57] X. Li, J. Hihath, F. Chen, T. Masuda, L. Zang, N.J. Tao, Thermally activated elec-tron transport in single redox molecules, Journal of the American ChemicalSociety 129 (2007) 11535.

[58] I. Pobelov, Z. Li, T. Wandlowski, Electrolyte gating in redox-active tunnel-ing junctions – an electrochemical STM approach, Journal of the AmericanChemical Society 130 (2008) 16045.

[59] A. Mishchenko, M. Abdualla, A. Rudnev, Y. Fu, A.R. Pike, T. Wandlowski,Electrochemical scanning tunnelling spectroscopy of a ferrocene-modified n-Si(1 1 1)-surface: electrolyte gating and ambipolar FET behaviour, ChemicalCommunications 47 (2011) 9807.

[60] E.I. Rogers, B. Sljukicı, C. Hardacre, R.G. Compton, Electrochemistry in room-temperature ionic liquids: potential windows at mercury electrodes, Journalof Chemical and Engineering Data 54 (2009) 2049.

[61] M. Armand, F. Endres, D.R. MacFarlane, H. Ohno, B. Scrosati, Ionic-liquid mate-rials for the electrochemical challenges of the future, Nature Materials 8(2009) 621.

[62] I. Díez-Pérez, Z. Li, J. Hihath, J. Li, C. Zhang, X. Yang, L. Zang, Y. Dai, X. Feng, K.Müllen, N.J. Tao, Gate-controlled electron transport in coronenes: bottom-upapproach towards graphene transistors, Nature communications 1 (2010) 31.

[63] N.J. Kay, S.J. Higgins, J.O. Jeppesen, E. Leary, J. Lycoops, J. Ulstrup, R.J. Nichols,Single-molecule electrochemical gating in ionic liquids, Journal of the Amer-ican Chemical Society 134 (2012) 16817.

[64] L.A. Nagahara, T. Thundat, S.M. Lindsay, Preparation and characterization ofSTM tips for electrochemical studies, Review of Scientific Instruments 60(1989) 3128.

[65] C.E. Bach, R.J. Nichols, W. Beckmann, H. Meyer, A. Schulte, J.O. Besenhard, P.D.Jannakoudakis, Effective insulation of scanning tunnelling microscopy tipsfor electrochemical studies using an electropainting method, Journal of theElectrochemical Society 140 (1993) 1281.

[66] A.G. Guell, I. Díez-Pérez, P. Gorostiza, F. Sanz, Preparation of reliable probesfor electrochemical tunnelling spectroscopy, Analytical Chemistry 76 (2004)5218.

[67] W. Hong, H. Valkenier, G. Mészáros, D.Z. Manrique, A. Mishchenko, A. Putz,P.M. García, C.J. Lambert, J.C. Hummelen, T. Wandlowski, An MCBJ casestudy: the influence of (-conjugation on the single-molecule conductance ata solid/liquid interface, Beilstein Journal of Nanotechnology 2 (2011) 699.

[68] Assuming a simple direct tunnelling process where the conductance G decaysexponentially with the electrode–electrode gap distance L as G = G0 exp (- L),being G0 = 77.4 �S the conductance quantum.

[69] S.J. Tans, A.R.M. Verschueren, C. Dekker, Room-temperature transistor basedon a single carbon nanotube, Nature 393 (1998) 49.

[70] P.C. Collins, M.S. Arnold, P. Avouris, Engineering carbon nanotubes and nano-tubes circuits using electrical breakdown, Science 292 (2001) 706.

[71] Y. Cui, C. Lieber, Functional nanoscale electronic devices assembled usingsilicon nanowire building blocks, Science 291 (2001) 851.

[72] X. Du, I. Skachko, A. Barker, E.Y. Andrei, Approaching ballistic transport insuspended graphene, Nature Nanotechnology 3 (2008) 491.

[73] L. Jiao, L. Zhang, X. Wang, G. Diankov, H. Dai, Narrow graphene nanoribbonsfrom carbon nanotubes, Nature 458 (2009) 877.

[74] M.D. Ventra, S.T. Pantelides, N.D. Lang, The benzene molecule as a molecularresonant-tunneling transistor, Applied Physics Letters 76 (2000) 3448.

[75] P. Damle, T. Rakshit, M. Paulsson, S. Datta, Current–voltage characteristicsof molecular conductors: two versus three terminal, IEEE Transactions onNanotechnology 1 (2002) 145.

[76] C. Li, A. Mishchenko, Z. Li, I. Pobelov, T. Wandlowski, X.Q. Li, F. Würthner,A. Bagrets, F. Evers, Electrochemical gate-controlled electron transport ofredox-active single perylene bisimide molecular junctions, Journal of Physics:Condensed Matter 20 (2008) 374122.

[77] Z. Li, B. Han, G. Mészáros, I. Pobelov, T. Wandlowski, A. Błaszczyk, M.Mayor, Two-dimensional assembly and local redox-activity of molecularhybrid structures in an electrochemical environment, Faraday Discussions131 (2006) 121.

[78] W. Haiss, T. Albrecht, H. van Zalinge, S.J. Higgins, D. Bethell, H. Höbenre-ich, D.J. Schiffrin, R.J. Nichols, A.M. Kuznetsov, J. Zhang, Q. Chi, J. Ulstrup,Single-molecule conductance of redox molecules in electrochemical scanningtunneling microscopy, Journal of Physical Chemistry B 111 (2007) 6703.

[79] E. Leary, S.J. Higgins, H. van Zalinge, W. Haiss, R.J. Nichols, S. Nygaard, J.O.Jeppesen, J. Ulstrup, Structure–property relationship in redox-gated singlemolecule junctions – a comparison of pyrrolo-tetrathiafulvalene and viologenredox groups, Journal of the American Chemical Society 130 (2008) 12204.

[80] W. Su, J. Jiang, W. Lu, Y. Luo, First-principles study of electrochemicalgate-controlled conductance in molecular junctions, Nano Letters 6 (2006)

2091.

[81] J.H. Oh, S. Liu, Z. Bao, R. Schmidt, F. Würthner, Air-stable n-channelorganic thin-film transistors with high field-effect mobility based on N,N′ ′-bis(heptafluorobutyl)-3,4:9,10-perylene diimide, Applied Physics Letters 91(2007) 212107.

cta 110 (2013) 741– 753

[82] J. Zhang, Q. Chi, A.M. Kuznetsov, A.G. Hansen, H. Wackerbarth, H.E.M.Christensen, J.E.T. Andersen, J. Ulstrup, Electronic properties of functionalbiomolecules at metal/aqueous solution interfaces, Journal of Physical Chem-istry B 106 (2002) 1132.

[83] J. Zhang, A.M. Kuznetsov, I.G. Medvedev, Q. Medvedev, T. Albrecht, P.S.Jensen, J. Ulstrup, Single-molecule electron transfer in electrochemical envi-ronments, Chemical Reviews 108 (2008) 2737.

[84] L. Dossel, L. Gherghel, X. Feng, K. Müllen, Graphene nanoribbons by chemists:nanometer-sized, soluble, and defect-free, Angewandte Chemie InternationalEdition 50 (2011) 2540.

[85] C.L. Bird, A.T. Kuhn, Electrochemistry of the viologens, Chemical SocietyReviews 10 (1981) 49.

[86] S. Heinze, M. Radosavljevic, J. Tersoff, P.H. Avouris, Unexpected scaling ofthe performance of carbon nanotube schottky-barrier transistors, PhysicalReview B 68 (2003) 235418.

[87] K.S. Novoselov, A.K. Geim, S.V. Morozov, D. Jiang, Y. Zhang, S.V. Dubonos, I.V.Grigorieva, A.A. Firsov, Electric field effect in atomically thin carbon films,Science 306 (2004) 666.

[88] A. Dodabalapur, H.E. Katz, L. Torsi, R.C. Haddon, Organic heterostructure field-effect transistors, Science 269 (1995) 1560.

[89] Q. Ye, J. Chang, K.-W. Huang, C. Chi, Thiophene-fused tetracene diimide withlow band gap and ambipolar behavior, Organic Letters 13 (2011) 5960.

[90] I. Díez-Pérez, Z. Li, S. Guo, C. Madden, H. Huang, Y. Che, X. Yang, L. Zang,N.J. Tao, Ambipolar transport in an electrochemically gated single-moleculefield-effect transistor, ACS Nano 6 (2012) 7044.

[91] L. Esaki, New phenomenon in narrow germanium p-n junctions, PhysicalReview 109 (1958) 603.

[92] S. Biswas, M. Dutta, M.A. Stroscio, Negative differential resistance in con-ductive polymer and semiconducting quantum dot nanocomposite systems,Applied Physics Letters 95 (2009) 182102.

[93] J. Chen, M.A. Reed, A.M. Rawlett, J.M. Tour, Large on-off rations and negativedifferential resistance in a molecular electronic device, Science 286 (1999)1550.

[94] E. Tran, M.A. Rampi, G.M. Whitesides, Electron transfer in a Hg-SAM//SAM-Hg junction mediated by redox centers, Angewandte Chemie InternationalEdition 43 (2004) 3835.

[95] X. Xiao, L.A. Nagahara, A.M. Rawlett, N.J. Tao, Electrochemical gate-controlledconductance of single oligo(phenylene ethynylene)s, Journal of the AmericanChemical Society 127 (2005) 9235.

[96] F. Chen, J. He, C. Nuckolls, T. Roberts, J.E. Klare, S. Lindsay, A molecular switchbased on potential-induced changes of oxidation state, Nano Letters 5 (2005)503.

[97] J. He, Q. Fu, S. Lindsay, J.W. Ciszek, J.M. Tour, Electrochemical origin ofvoltage-controlled molecular conductance switching, Journal of the AmericanChemical Society 128 (2006) 14828.

[98] J.M. Artés, I. Díez-Pérez, P. Gorostiza, Transistor-like behaviour of single met-alloprotein junctions, Nano Letters 12 (2012) 2679.

[99] E.A. Della Pia, Q. Chi, J.E. Macdonald, J. Ulstrup, D.D. Jones, M. Elliott, Fastelectron transfer through a single molecule natively structured redox protein,Nanoscale 4 (2012) 7106.

[100] W. Schmickler, C.J. Widrig, The investigation of redox reactions witha scanning tunneling microscope: experimental and theoretical aspects,Electroanalytical Chemistry 336 (1992) 213.

[101] W. Schmickler, N.J. Tao, Measuring the inverted region of an electron transferreaction with a scanning tunneling microscope, Electrochimica Acta 42 (1997)2809.

[102] L. Sepunaru, I. Pecht, M. Sheves, D. Cahen, Solid-state electron transportacross Azurin: from a temperature-independent to a temperature-activatedmechanism, Journal of the American Chemical Society 133 (2011)2421.

[103] B.L. Feringa, Molecular Switches, Wiley-VCH Verlag GmbH, Germany, 2001.[104] N.N.P. Moonen, A.H. Flood, J.M. Fernandez, J.F. Stoddart, Towards a rational

design of molecular switches and sensors from their basic building blocks,Topics in Current Chemistry 262 (2005) 99.

[105] L.P. Hammett, The effect of structure upon the reactions of organic com-pounds. Benzene derivatives, Journal of the American Chemical Society 59(1937) 96.

[106] N. Vedova-Brook, N. Matsunaga, K. Sohlberg, Correlating substituent param-eter values to electron transport properties of molecules, Chemical Physics299 (2004) 89.

[107] M. Taniguchi, M. Tsutsui, R. Mogi, T. Sugawara, Y. Tsuji, K. Yoshizawa, T. Kawai,Dependence of single-molecule conductance on molecule junction symme-try, Journal of the American Chemical Society 133 (2011) 11426.

[108] Z. Li, E. Borguet, Determining charge transport pathways through singleporphyrin molecules using scanning tunneling microscopy break junction,Journal of the American Chemical Society 134 (2012) 63.

[109] T. Markussen, R. Stadler, K.S. Thygesen, The relation between structure andquantum interference in single molecule junctions, Nano Letters 10 (2010)4260.

[110] D. Fracasso, H. Valkenier, J.C. Hummelen, G.C. Solomon, R.C. Chiechi, Evidencefor quantum interference in SAMs of arylethynylene thiolates in tunneling

junctions with eutectic Ga–In (EGain) top-contacts, Journal of the AmericanChemical Society 133 (2011) 9556.

[111] V. Kaliginedi, P. Moreno-García, H. Valkenier, W. Hong, V.M. García-Suarez,P. Buiter, J.L.H. Otten, J.C. Hummelen, C.J. Lambert, T. Wandlowski, Correla-tion between molecular structure and single-junction conductance: a case

ica A

[

[

[

American Chemical Society 119 (1997) 201.

S. Guo et al. / Electrochim

study with oligo(phenylene–ethynylene)-type wires, Journal of the AmericanChemical Society 134 (2012) 5262.

112] C.M. Guédon, H. Valkenier, T. Markussen, K.S. Thygesen, J.C. Hummelen, S.J.van der Molen, Observation of quantum interference in molecular chargetransport, Nature Nanotechnology 7 (2012) 305.

113] N. Darwish, I. Díez-Pérez, P. Da Silva, N.J. Tao, J.J. Gooding, M.N. Paddon-Row,Observation of electrochemically controlled quantum interference in a singleanthraquinone-based norbornylogous bridge molecule, Angewandte Chemie

International Edition 51 (2012) 3203.

114] N. Darwish, I. Díez-Pérez, S. Guo, N.J. Tao, J.J. Gooding, M.N. Paddon-Row,Single molecule switches: electrochemical gating of a single anthraquinone-based norbornylogous bridge molecule, Journal of Physical Chemistry C 116(2012) 21093.

cta 110 (2013) 741– 753 753

[115] T. Hines, I. Díez-Pérez, H. Nakamura, T. Shimazaki, Y. Asai, N.J. Tao, Controllingformation of single-molecule junction by electrochemical reduction of diazo-nium terminal groups, Journal of the American Chemical Society 135 (2013)3319.

[116] P. Allongue, M. Delamar, B. Desbat, O. Fagebaume, R. Hitmi, J. Pinson, J.M.Saveant, Covalent modification of carbon surfaces by aryl radicals gener-ated from the electrochemical reduction of diazonium salts, Journal of the

[117] P. Martin, M.L. Della Rocca, A. Anthore, P. Lafarge, J.C. Lacroix, Organicelectrodes based on grafted oligothiophene units in ultrathin, large-areamolecular junctions, Journal of the American Chemical Society 134 (2012)154.