German Edition: DOI: 10.1002/ange.201704460Surface Analysis Very Important PaperInternational Edition: DOI: 10.1002/anie.201704460

Electrochemical TERS Elucidates Potential-Induced MolecularReorientation of Adenine/Au(111)Natalia Mart&n Saban8s, Tatsuhiko Ohto, Denis Andrienko, Yuki Nagata, and Katrin F. Domke*

Abstract: Electrochemical surface activity arises from theinteraction and geometric arrangement of molecules at elec-trified interfaces. We present a novel electrochemical tip-enhanced Raman spectroscope that can access the vibrationalfingerprint of less than 100 small, non-resonant moleculesadsorbed at atomically flat Au electrodes to study theiradsorption geometry and chemical reactivity as a function ofthe applied potential. Combining experimental and simulationdata for adenine/Au(111), we conclude that protonatedphysisorbed adenine adopts a tilted orientation at low poten-tials, whereas it is vertically adsorbed around the potential ofzero charge. Further potential increase induces adeninedeprotonation and reorientation to a planar configuration.The extension of EC-TERS to the study of adsorbatereorientation significantly broadens the applicability of thisadvanced spectroelectrochemical tool for the nanoscale char-acterization of a full range of electrochemical interfaces.

The adsorption geometry of (re)active molecules is a crucialparameter that determines surface reactivity or devicefunctionality in a large variety of applications, such asheterogeneous catalysis, electrochemical (EC) energy con-version, biotechnology,and molecular electronics. For exam-ple, it has been suggested that the spatial tilt of p-bonds withrespect to the substrate strongly influences the catalyticactivity or electron conductance of adsorbates.[1–3] Accessingadsorbate orientation in relation to specific surface sitesin situ or in operando is a first crucial step toward controllinginterfacial geometries for improved device architecture.However, suitable in situ techniques to study molecularorientation at well-defined adsorption sites are still scarce.Surface-specific in situ Raman- (EC-SERS) or IR-based (EC-SEIRAS) linear and nonlinear (EC sum frequency gener-ation, SFG) vibrational spectroscopies provide the requiredsensitivity and chemical specificity for investigating moleculargeometries at (potential-controlled) solid/liquid interfaces atthe ensemble level, that is, the signals represent an averageresponse from a large number of molecules and surface sites

present in the focus spot.[4–7] EC scanning probe microscopy(EC-SPM) enables the visualization of individual adsorbatestructures with nanometer spatial resolution, albeit at theexpense of chemical specificity.[8,9]

Tip-enhanced Raman spectroscopy (TERS) offers anelegant solution to achieve both the required surface (sub)-monolayer chemical sensitivity and nanometer spatial reso-lution. TERS is based on the combination of an SPM witha Raman optical platform in which a metallic SPM probelocated in close vicinity above a surface acts as a nanoantennato create a plasmon-enhanced field and, consequently, strongRaman scattering from adsorbates located in the tip–samplegap at a spatial resolution of typically below 10 nm.[10–12]

While TERS in UHV and ambient conditions is increasinglyused in surface science to interrogate adsorbate orienta-tion,[13–16] the study of solid/liquid interfaces has longremained a huge technical challenge.[17,18] Extending thetechnique to spectro-electrochemical experiments requiresadditional electrical contacting and control of the workingelectrode under investigation, thus rendering the setup evenmore complex and technically challenging. Recently, thegroups of Bin Ren and Richard Van Duyne reported EC-TERS setups that allow the in situ investigation of surfaceredox or chemical conversion processes steered by theelectrode potential.[19–21] While these works demonstrate theimpressive potential that EC-TERS holds for studyinginterfacial molecular processes in situ, the versatility of theinstruments was limited by either the required tilt of thesample[19] or the use of optically transparent substrates;[20] theformer complicates TERS mapping, while the latter is neededbecause of the chosen transmission illumination geometry.

Herein, we report an alternative EC-TERS approach thatenables the study of the potential-dependent behavior ofsmall, non-resonant molecules adsorbed onto an opaquesubstrate. As a proof of concept, we monitor the EC-TERSresponse of a monolayer of the DNA base adenine adsorbedonto a well-defined Au(111) single crystal. Given theimportance of understanding the interactions between DNAbases and noble metals for biosensing applications and for thedevelopment of biocompatible materials,[22, 23] it is surprisingthat the DNA base orientation with respect to the Au surface(for example, upright, tilted, or flat) and the chemical state ofthe DNA base (for example, (de)protonated or anionic) hasremained a point of controversy, despite extensive efforts tocharacterize the system with traditional cyclo-voltammetric(CV) approaches in combination with SPM or (ensemble)spectro-electrochemistry.[24–32]

EC-TERS offers the unique advantage of placing thenear-field probe at a location of interest chosen according tothe SPM topographic information gathered simultaneously, in

[*] N. Mart&n Saban8s, Dr. Dr. D. Andrienko, Dr. Y. Nagata,Dr. K. F. DomkeMax Planck Institute for Polymer ResearchAckermannweg 10, 55128 Mainz (Germany)E-mail: domke@mpip-mainz.mpg.de

Dr. T. OhtoGraduate School of Engineering Science, Osaka University1-3 Machikaneyama, Toyonaka 560-8531 (Japan)

Supporting information for this article, including experimentaldetails, can be found under:https://doi.org/10.1002/anie.201704460.

AngewandteChemieCommunications

9796 T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Angew. Chem. Int. Ed. 2017, 56, 9796 –9801

contrast to EC-SERS and EC-SEIRAS studies in whichsubstrate roughness leads to an inherent variation of thegeometry of the surface adsorption sites and thus to a broaddistribution of probed molecular adsorption geometries. Ourapproach thus enables selective monitoring of potential-dependent TER spectral changes from less than 100 adeninemolecules adsorbed onto a Au(111) terrace with uniformadsorption geometry in which an averaging effect fromdifferent substrate/adsorbate geometries is limited. We com-plement the experimental EC-TERS results with densityfuncitonal theory (DFT) gas-phase approximations for dataanalysis and interpretation.

Figure 1A depicts our scanning tunneling microscope(STM-) based EC-TERS experimental configuration. Thepotentials of the sample (working electrode 1, WE1) and tip(WE2) can be controlled independently with a bipotentiostatversus a Pd/H2 reference electrode (RE) (all potentials arereported vs. Ag/AgCl). The TER signal is generated by side-illuminating the tip-sample gap in tunneling mode witha 632.8 nm excitation HeNe laser through a 50 X air-immer-sion objective (N.A. = 0.5) and is collected through the sameobjective in backscattering mode as described previously.[18]

The sample consists of an adenine self-assembled monolayeron a Au(111) single crystal in adenine-free 0.01m H2SO4. Fordetails about the EC-TERS setup and sample preparation,see Section 1 of the Supporting Information. All spectra wererecorded with the tip located in the center of a flat Au(111)terrace (Figure 1B, white dot) as confirmed by EC-STMimaging prior to and after spectral acquisition.

Figure 1C shows EC-TER spectra recorded at differentsample potentials, Esample, in chronological order of acquisitionfrom bottom to top. At low potentials, one prominent peak ispresent at circa 736 cm@1 (ring breathing), and weaker bandsappear at around 1320 cm@1 (CN and CC stretch) and1464 cm@1 (N7@C8 stretch, C8@H bend, and NH2 scissor), inagreement with previous reports for adenine/Au(111).[31,33–36]

At all potentials, a broad band is observed below 300 cm@1

that can be decomposed, with help of Lorentzian band fitting,

into three components (Figure 1 D): The two peaks located at172 and 220 cm@1 that are also observed in all far-field spectraacquired upon tip retraction and do not show any potential-dependent changes; as their molecular origin is unclear, theywill not be discussed further. Notably, the fitting revealsa near-field band at circa 260 cm@1 at potentials greater than0.6 V that has been assigned to Au–N interactions.[36]

To quantify the spectral changes as a function of Esample, weplot amplitude and peak frequency of the 260, 736, and1464 cm@1 modes obtained from Lorentzian fits (Figure 2,details in the Supporting Information, Section 1). The corre-sponding CV recorded at 50 mVs@1 inside the EC-TERS cellis shown as an inset in Figure 2B. The CV exhibits the twowell-known peaks at 0.36 and 0.62 V (0.34 and 0.59 V) in theanodic (cathodic) scan direction.[26,27, 29] The first peak can beattributed to an increase in capacitance as a result of themolecular reorientation from a physisorbed tilted arrange-ment to a vertical geometry around the potential of zerocharge,[37] such that more Au surface is exposed to theadenine-free electrolyte, in agreement with previous stud-ies.[25, 27, 38] The second peak at around 0.6 V results fromdeprotonation and partial charge transfer from adenine to theAu electrode (around 0.68 electrons per molecule as deter-mined by Aldaz and coworkers[39]).[25, 27,38] Note that oxidationof adenine to its oxo forms occurs at potentials higher than1.2 V outside our potential window.[40, 41] Importantly, the CVwas identical before and after the EC-TERS experiment, thatis, after ramping up to 1.2 V several times, indicating that(partial) Au oxidation does not alter the behavior of adenineat the electrode (Supporting Information, Section 2).

The Au–N mode at circa 260 cm@1 only appears after thesecond anodic peak and blue-shifts from 245 cm@1 at 0.6 V to274 cm@1 at 1.2 V (Figure 2A, blue) while maintaining a con-stant intensity (variations within the noise level, Figure 2B,blue). In contrast, the intensity of the 736 cm@1 mode showsa maximum at 0.3 V (Figure 2B, purple). In the regionbetween the two anodic peaks, the band significantly fallssteeply in intensity and further decreases after the second

Figure 1. A) EC-TERS cell and electrode configuration; STM = scanning tunneling microscope; WE/CE/RE = working/counter/reference electrodes.B) Left: 228 W 228 nm2 STM image (Esample =0.5 V, Ebias =0.4 V, It = 1.35 nA); dot indicates EC-TERS position. Right: Au-step line profile as indicatedin STM image. C) EC-TER spectral evolution upon WE1 potential ramp (1.77 mW laser power, 5 s acquisition time). Inset: neutral adenine.D) Lorentzian band fitting examples for the low wavenumber region.

AngewandteChemieCommunications

9797Angew. Chem. Int. Ed. 2017, 56, 9796 –9801 T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.angewandte.org

peak at a less pronounced rate before disappearing atpotentials above 1 V. Interestingly, the band shows a small,but measurable blue shift from 734.6 cm@1 at 0.3 V to736.4 cm@1 at 1 V (Figure 2A, purple). The 1464 cm@1 modeshows a constant intensity between 0.2 and 0.6 V, it vanishesat 0.8 V, and reappears at 1 V (Figure 2, green). This modeblue-shifts by circa 10 cm@1 at high potentials.

Different phenomena, like changes in surface coverage,chemical conversion, molecular reorientation, and chargeeffects, could account for the observed potential-dependentspectral changes. Similar intensity variations of the ringbreathing mode have been attributed to changes in thecoverage in an adenine-containing electrolyte in an ensemblestudy.[31] As we work in an adenine-free electrolyte, anincrease in coverage would be uniquely due to moleculardiffusion. By integrating over 5 s per spectrum and averagingover 5 spectra per data point, however, we average overrandom fluctuations produced by diffusion into and out of thehot spot. Furthermore, the recovery of the 736 cm@1 uponsweep reversal (Supporting Information, Section 3) and theconstant intensity of the other Raman peaks as well as of theCV features (Supporting Information, Section 2) corroboratethat the surface coverage in the near-field region remainsconstant throughout the hour-long experiment set. Accordingto the literature, one adenine molecule occupies a surface

area between 0.41 nm2 (upright) and 0.55 nm2 (flat).[39] TheTERS spatial resolution for small, non-resonant molecules ina Au-tip/Au-sample gap is on the order of a few nanometers,in which the near-field is sufficiently strong to providedetectable TER scattering;[42–45] with an effective TERSscattering diameter of 6 nm, we can estimate the number ofprobed adenine molecules to be between 50 and 70,demonstrating the extreme sensitivity our EC-TERS setupholds particularly for non-resonant species.

In acidic media, (de)protonation from N1-protonated toneutral adenine has been suggested to occur upon potentialsweep at around 0.6 V, facilitating chemisorption with a partialcharge transfer from N lone-pairs of adenine to the d*-bandof Au(111) at higher potentials.[26–28, 31] As a transition betweenprotonated, neutral, and partially oxidized adenine mayinduce band shifts, we have performed gas-phase DFTcalculations (details in the Supporting Information, section 1)to compare the Raman shifts predicted for cationic, neutral,and N1-protonated molecules. We assign the experimentalmode at 736 cm@1 to the simulated ring-breathing modearound 723 cm@1 and the experimental 1464 cm@1 band to thecalculated mode at circa 1513 cm@1 based on the modedisplacements returned by the calculations (SupportingInformation, Section 4) in agreement with previousreports.[33,34] DFT predicts a blue-shift from 721.69(1511.89) cm@1 for the protonated molecule to 725.71(1514.03) cm@1 for the neutral case to 731.72 (1569.53) cm@1

for the cation. Note that adenine is not fully oxidized withinour potential window as discussed earlier; the cation calcu-lation merely serves to confirm the trend of the Raman shiftdue to deprotonation and partial charge transfer uponchemisorption. This trend is in qualitative agreement withthe blue-shifts of 2 and 10 cm@1 observed experimentally forthe 736 and 1464 cm@1 modes, respectively. In addition,previous pH-dependent SERS experiments showed that atlow pH, the intensity ratio of the 1320 and 1350 cm@1 bands ishigher for the protonated species than for neutral adenine.[36]

We observe the same trend upon potential increase (Support-ing Information, Section 5), which further supports theconclusion that at potentials below 0.6 V, protonated adeninemolecules are present before deprotonation occurs at morepositive potentials. While potential-induced deprotonationexplains the observed band shifts and small variations inrelative band intensities, it cannot account for the drasticintensity changes of the prominent ring-breathing mode.

Since the Raman scattering intensity is proportional to theforth power of the polarizability, potential-dependent molec-ular reorientation is expected to manifest itself in strongintensity changes as observed in the EC-TER spectra,provided the tensor elements exhibit anisotropy that is largelymaintained upon adsorption. Note that in TERS, the excita-tion (near-)field is oriented parallel to the surface normal.[14]

As such, with knowledge of the respective Raman polar-izability tensors of the vibrational modes of interest, themolecular orientation of the few molecules in the nanometer-confined near-field spot on the flat Au(111) terrace can bededuced from the observed intensity changes.[13,14] As a firstapproximation, we have calculated the transition polarizabil-ity tensors of the 723 and 1513 cm@1 modes of an isolated

Figure 2. A) Raman shift and B) integrated intensity of the the260 cm@1 (blue), 736 cm@1 (purple), and 1464 cm@1 (green) bands asa function of sample potential. Error bars result from fittings. CV ofa monolayer of adenine on Au(111) in 0.01m H2SO4, v = 50 mVs@1 (B,inset).

AngewandteChemieCommunications

9798 www.angewandte.org T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Angew. Chem. Int. Ed. 2017, 56, 9796 –9801

molecule in protonated and neutral forms by using DFT.[46]

The complete tensors are shown in the Supporting Informa-tion, Section 6. In our calculations, the z-axis is perpendicularto the ringsQ plane (containing the x-/y-axes, with y (short axis)parallel to the C4@C5 bond; Figure 3A, inset). The diagonal

elements of the Raman tensors for the 723 cm@1 mode showa pronounced anisotropy between the x,y-plane of theadenine rings and the z-component for both protonated andneutral species with a ratio of axx :ayy :azz = 3:3:1, while theRaman tensor of the 1513 cm@1 mode shows a pronouncedanisotropy in the x,y-plane with a ratio of 14:1:1 for theneutral form of adenine and 20:2:1 for the protonated case.

Figure 3A displays the simulated intensity variation of the723 and 1513 cm@1 modes for a vertical to planar transitionwith the y-axis of the molecular plane initially parallel to thepolarization direction of the (plane-wave) excitation field (Y’,or to the surface normal in the experiment) and with themolecular plane tilting relative toward the Z’X’ plane(represented by the electrode surface in the experiment) atan angle q together with the experimental data points. Thering-breathing mode intensity is highest for a tilt angle of q =

088 and slowly decreases with molecular reorientation to a flatorientation at q = 9088 (Figure 3A). The simulation results forprotonated (purple solid line) and neutral adenine (blue solidline) show the same trend and describe perfectly the

experimental data (black solid circles), including the depro-tonation occuring between peaks in the CV (gray shadedarea). The lower band intensity of the ring breathing mode at0.2 V (red solid circle) below the potential of zero charge,which has also been observed in previous EC-SERS experi-ments,[31, 47] is consistent with a tilted physisorbed adsorbateorientation as deduced from EC experiments and discussedabove. In other words, the experimentally observed 736 cm@1

band intensity behavior as a function of increasing potentialcan be explained by the reorientation of adenine from tiltedto upright to flat with respect to the Au(111) surface.

The simulation results for the 1513 cm@1 mode show thatfor both the protonated and neutral cases, the peak intensitydrops to zero at intermediate angles between circa 4088 and6088, while at smaller or larger angles, the peak is expected todisplay a higher intensity (Figure 3B). The experimental datashow such behavior; the peak vanishes at 0.8 V, while at lowerand higher potentials the measured intensity is constantlyhigh. Note that we found no qualitative agreement betweensimulation and experiment for an initial orientation with themoleculeQs x-axis parallel to the excitation field (SupportingInformation, Section 7). The quantitative deviation betweenthe absolute intensity values of experiments and simulationsfor this mode may be related to the fact that the gas-phasesimulations do not take into account molecule–molecule ormolecule–substrate interactions. The 1513 cm@1 modeinvolves the NH2 moiety that has been suggested to enableadenine adsorption onto Au.[32] Therefore, this vibration isexpected to be more affected by the exclusion of the electrodein the simulations than the delocalized ring-breathing mode.Simulations of adenine at solid/liquid interfaces beyond gas-phase calculations could give polarizability tensors that betterdescribe the experimental results.

Adenine reorientation from vertical to horizontal upona potential increase above 0.3 V is further supported by theanalysis of the low-wavenumber mode at circa 260 cm@1. Thisband has previously not been investigated experimentally indetail, despite the wealth of information it contains about theadsorbate–substrate interaction. Interestingly, the Au@Nmode only appears at potentials above 0.6 V after deproto-nation. The significant blue-shift from 245 to 274 cm@1 can beattributed to a strengthening of the Au@N bond withincreasing potential, that is, an increasing interaction betweenthe N lone pairs and the metal d*-band that results in a planarorientation of adenine. Coordination of neutral adeninethrough the amino group N10 and N1 at potentials abovethe anodic peaks in the CV is consistent with all observedtrends. The lack of the Au@N mode at potentials below 0.6 Vis in line with weakly physisorbed adenine in which themonolayer is stabilized by p-d* adsorbate–substrate or p–p

intermolecular interactions at potentials below or around0.3 V, respectively.[25, 30, 32] Figure 4 summarizes the adsorption/reaction model of the potential-dependent, reversible ade-nine (de)protonation and reorientation as deduced from ourEC-TERS results.

To conclude, we have developed a novel EC-TERS setupto study the electrochemical behaviour of small non-resonantmolecules at electrified interfaces with a spatial chemicalresolution of a few nanometers and sensitivity below 100 mol-

Figure 3. Orientation-dependent calculated Raman intensities of theA) 723 cm@1 and B) 1513 cm@1 modes in neutral (A, purple; B, darkgreen) and protonated (A, blue; B, light green) forms. Inset in A:Schematic of the rotation considered. The experimental data points(black or red solid circles) of both modes are included for comparison.The gray shaded areas indicate the potential region in which deproto-nation is expected.

AngewandteChemieCommunications

9799Angew. Chem. Int. Ed. 2017, 56, 9796 –9801 T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.angewandte.org

ecules. For the adenine/Au(111) system, EC-TERS in combi-nation with DFT results allow us to deduce a reversibleadsorption/reaction model in which protonated adenine isadsorbed onto atomically flat Au(111) in a tilted geometry atlow potentials and in an upright configuration with its shortmolecular axis perpendicular to the surface around thepotential of zero charge. Upon potential increase, themolecule adopts a flat adsorption geometry with strong Au@N interactions between deprotonated adenine molecules andthe gold substrate. With this important extension of EC-TERS capabilities to evaluate adsorbate orientation andchemical conversion as a function of potential, in situ localchemical information at the nanoscale becomes accessible fora full range of electrochemical systems.

Acknowledgements

N.M.S. and K.F.D. acknowledge generous project fundingthrough the Emmy Noether Program of the DeutscheForschungsgemeinschaft (DO 1691/1-1) and the MarieCurie Actions of the European Union (ITN-FINON607842). We thank Leonie Driessen for technical support inthe cell design, Jgrgen Worm for creative and fast softwaresolutions, Marc-Jan van Zadel for technical support, andAmala Elizabeth and Ulmas Zhumaev for assistance duringexperiments and valuable discussions.

Conflict of interest

The authors declare no conflict of interest.

Keywords: adenine · electrochemistry · nucleobases ·surface science · tip-enhanced Raman spectroscopy

How to cite: Angew. Chem. Int. Ed. 2017, 56, 9796–9801Angew. Chem. 2017, 129, 9928–9933

[1] a) M. E. Chiu, D. J. Watson, G. Kyriakou, M. S. Tikhov, R. M.Lambert, Angew. Chem. Int. Ed. 2006, 45, 7530 – 7534; Angew.Chem. 2006, 118, 7692 – 7696.

[2] Z. Qiu, M. Zhang, D. Y. Wu, S. Y. Ding, Q. Q. Zuo, Y. F. Huang,W. Shen, X. D. Lin, Z. Q. Tian, B. W. Mao, ChemPhysChem2013, 14, 2217 – 2224.

[3] M. Lukas, K. Dçssel, A. Schramm, O. Fuhr, C. Stroh, M. Mayor,K. Fink, H. Von Lçhneysen, ACS Nano 2013, 7, 6170 – 6180.

[4] M. Osawa, Bull. Chem. Soc. Jpn. 1997, 70, 2861 – 2880.[5] S. Baldelli, Acc. Chem. Res. 2008, 41, 421 – 431.[6] Z.-Q. Tian, B. Ren, Annu. Rev. Phys. Chem. 2004, 55, 197 – 229.[7] S. Zaleski, A. J. Wilson, M. Mattei, X. Chen, G. Goubert, M. F.

Cardinal, K. A. Willets, R. P. Van Duyne, Acc. Chem. Res. 2016,49, 2023 – 2030.

[8] a) H. Wackerbarth, M. Grubb, J. Zhang, A. G. Hansen, J.Ulstrup, Angew. Chem. Int. Ed. 2003, 43, 198 – 203; Angew.Chem. 2003, 116, 200 – 205.

[9] R. Wen, B. Rahn, O. M. Magnussen, J. Phys. Chem. C 2016, 120,15765 – 15771.

[10] a) A. Hartschuh, Angew. Chem. Int. Ed. 2008, 47, 8178 – 8191;b) A. Hartschuh, Angew. Chem. 2008, 120, 8298 – 8312.

[11] G. Sharma, T. Deckert-Gaudig, V. Deckert, Adv. Drug DeliveryRev. 2015, 89, 42 – 56.

[12] N. Jiang, D. Kurouski, E. A. Pozzi, N. Chiang, M. C. Hersam,R. P. Van Duyne, Chem. Phys. Lett. 2016, 659, 16 – 24.

[13] R. Zhang, Y. Zhang, Z. C. Dong, S. Jiang, C. Zhang, L. G. Chen,L. Zhang, Y. Liao, J. Aizpurua, Y. Luo, J. L. Yang, J. G. Hou,Nature 2013, 498, 82 – 86.

[14] N. Jiang, N. Chiang, L. R. Madison, E. A. Pozzi, M. R. Wasie-lewski, T. Seideman, M. A. Ratner, M. C. Hersam, G. C. Schatz,R. P. Van Duyne, Nano Lett. 2016, 16, 3898 – 3904.

[15] D. Zhang, K. F. Domke, B. Pettinger, ChemPhysChem 2010, 11,1662 – 1665.

[16] X. Wang, K. Broch, F. Schreiber, A. J. Meixner, D. Zhang, Phys.Chem. Chem. Phys. 2016, 18, 15919 – 15926.

[17] T. Schmid, B.-S. Yeo, G. Leong, J. Stadler, R. Zenobi, J. RamanSpectrosc. 2009, 40, 1392 – 1399.

[18] N. Mart&n Saban8s, L. Driessen, K. F. Domke, Anal. Chem. 2016,88, 7108 – 7114.

[19] Z.-C. Zeng, S.-C. Huang, D.-Y. Wu, L.-Y. Meng, M.-H. Li, T.-X.Huang, J.-H. Zhong, X. Wang, Z.-L. Yang, B. Ren, J. Am. Chem.Soc. 2015, 137, 11928 – 11931.

[20] D. Kurouski, M. Mattei, R. P. Van Duyne, Nano Lett. 2015, 15,7956 – 7962.

[21] M. Mattei, G. Kang, G. Goubert, D. V. Chulhai, G. C. Schatz, L.Jensen, R. P. Van Duyne, Nano Lett. 2017, 17, 590 – 596.

[22] a) E. Katz, I. Willner, Angew. Chem. Int. Ed. 2004, 43, 6042 –6108; Angew. Chem. 2004, 116, 6166 – 6235.

[23] P. A. Sontz, N. B. Muren, J. K. Barton, Acc. Chem. Res. 2012, 45,1792 – 1800.

[24] N. L. Tao, J. A. DeRose, S. M. Lindsay, J. Phys. Chem. 1993, 97,910 – 919.

[25] Y.-J. Xiao, Y.-F. Chen, X.-X. Gao, Spectrochim. Acta 1999, 55,1209 – 1218.

[26] A. P. M. Camargo, H. Baumg-rtel, C. Donner, Phys. Chem.Chem. Phys. 2003, 5, 1657 – 1664.

[27] A. Martins, A. Queirls, F. Silva, ChemPhysChem 2005, 6, 1056 –1060.

[28] A. Rodes, M. Rueda, F. Prieto, C. Prado, J. M. Feliu, A. Aldaz, J.Phys. Chem. C 2009, 113, 18784 – 18794.

[29] M. Rueda, F. Prieto, A. Rodes, J. M. Delgado, Electrochim. Acta2012, 82, 534 – 542.

[30] C. Vaz-Dom&nguez, M. Escudero-Escribano, A. Cuesta, F.Prieto-Dapena, C. Cerrillos, M. Rueda, Electrochem. Commun.2013, 35, 61 – 64.

Figure 4. Proposed potential-dependent reversible adsorption/reactionmodel for adenine/Au(111).

AngewandteChemieCommunications

9800 www.angewandte.org T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Angew. Chem. Int. Ed. 2017, 56, 9796 –9801

[31] B.-Y. Wen, X. Jin, Y. Li, Y.-H. Wang, C.-Y. Li, M.-M. Liang, R.Panneerselvam, Q.-C. Xu, D.-Y. Wu, Z.-L. Yang, J.-F. Li, Z.-Q.Tian, Analyst 2016, 141, 3731 – 3736.

[32] F. Prieto, Z. Su, J. J. Leitch, M. Rueda, J. Lipkowski, Langmuir2016, 32, 3827 – 3835.

[33] R. Santamaria, E. Charro, A. Zacaras, M. Castro, J. Comput.Chem. 1999, 20, 511 – 530.

[34] B. Giese, D. McNaughton, J. Phys. Chem. B 2002, 106, 101 – 112.[35] K. F. Domke, D. Zhang, B. Pettinger, J. Am. Chem. Soc. 2007,

129, 6708 – 6709.[36] J. Kundu, O. Neumann, B. Janesko, D. Zhang, S. Lal, A.

Barhoumi, G. E. Scuseria, N. J. Halas, J. Phys. Chem. C 2009,113, 14390 – 14397.

[37] D. M. Kolb, J. Schneider, Electrochim. Acta 1986, 31, 929 – 936.[38] A. P. M. Camargo, H. Baumg-rtel, C. Donner, PhysChemComm

2002, 5, 151.[39] C. Prado, F. Prieto, M. Rueda, J. Feliu, A. Aldaz, Electrochim.

Acta 2007, 52, 3168 – 3180.[40] L. M. GonÅalves, C. Batchelor-McAuley, A. A. Barros, R. G.

Compton, J. Phys. Chem. C 2010, 114, 14213 – 14219.

[41] G. Dryhurst, P. J. Elving, J. Electrochem. Soc. 1968, 115, 1014 –1020.

[42] S. Jiang, Y. Zhang, R. Zhang, C. Hu, M. Liao, Y. Luo, J. Yang, Z.Dong, J. G. Hou, Nat. Nanotechnol. 2015, 10, 865 – 870.

[43] C. Zhang, B. Q. Chen, Z. Y. Li, J. Phys. Chem. C 2015, 119,11858 – 11871.

[44] F. Latorre, S. Kupfer, T. Bocklitz, D. Kinzel, S. Trautmann, S.Gr-fe, V. Deckert, Nanoscale 2016, 8, 10229 – 10239.

[45] T. Deckert-Gaudig, D. Kurouski, M. A. B. Hedegaard, P. Singh,I. K. Lednev, V. Deckert, Sci. Rep. 2016, 6, 33575.

[46] Gaussian 09 (Revision E.01), M. J. Frisch, et al., Gaussian Inc.Wallingford CT 2009.

[47] D. IbaÇez, A. Santidrian, A. Heras, M. Kalb#c, A. Colina, J.Phys. Chem. C 2015, 119, 8191 – 8198.

Manuscript received: May 1, 2017Revised manuscript received: June 7, 2017Accepted manuscript online: January 0, 0000Version of record online: July 11, 2017

AngewandteChemieCommunications

9801Angew. Chem. Int. Ed. 2017, 56, 9796 –9801 T 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.angewandte.org