Investigation of Cobalt Phthalocyanine at the Solid/Liquid Interfaceby Electrochemical Tip-Enhanced Raman SpectroscopySong Jiang,† Zhu Chen,† Xu Chen,‡ Duc Nguyen,† Michael Mattei,†,∥ Guillaume Goubert,†,§

and Richard P. Van Duyne*,†,‡

†Department of Chemistry and ‡Applied Physics Graduate Program, Northwestern University, Evanston, Illinois 60208, UnitedStates

*S Supporting Information

ABSTRACT: Electrochemical tip-enhanced Raman spectrosco-py (EC-TERS) with three excitation wavelengths in combinationwith in situ electrochemical scanning tunneling microscopy (EC-STM) and absorption measurements has been employed toprovide comprehensive insights into the electrochemicalprocesses of cobalt phthalocyanine (CoPc) at the solid/liquidinterface supported on a Au(111) substrate. As the substratepotential becomes more negative, CoPc molecules form a highlyordered monolayer on the Au(111) surface (>0.1 V) until theordered-to-diffusing phase transition is triggered (<0.1 V). CoPcmolecules in the ordered phase are reduced during cathodicscanning, which leads to a redshift in the resonance conditionand gives rise to distinct EC-TERS behaviors which depend onexcitation wavelengths. The ordered-to-diffusing phase transitionof CoPc molecules results in the disappearance of the EC-TERS signal. The catalytic activity of CoPc for the oxygen reductionreaction (ORR) was not visible in the EC-STM and has negligible effect on the EC-TERS measurements. The comprehensiveevidence from EC-TERS, EC-STM, and absorption spectroelectrochemistry clearly demonstrates that partially reduced CoPcmolecules are the dominant species under steady state measurements during the oxygen reduction reaction.

■ INTRODUCTION

Unravelling the electrochemical processes at the solid/electrolyte interface can greatly improve device efficiency andminimize performance loss in energy conversion and storagetechnologies such as fuel cells and metal air batteries.1,2 Inelectrochemistry, surface heterogeneity, solvent interactions,and the transport of reactants and products create a dynamiclocal environment, which presents a significant challenge forunderstanding the mechanisms of electrochemical processes.3,4

The ability to provide accurate and relevant insights into themechanisms of an electrochemical process can only beachieved through in situ probing of the fundamentalinteractions under electrochemical (EC) conditions. EC-scanning tunneling microscopy (EC-STM) is a powerful toolto obtain potential-dependent nanoscale topographic informa-tion at solid/liquid interfaces.5,6 However, topographic imagesprovide only limited chemical information regarding thespecies undergoing electrochemical reactions at the solid/liquid interface. Surface-enhanced Raman spectroscopy(SERS) can provide characteristic vibrational spectroscopyfor the chemical species that are involved in a catalytic processbut does not provide topographic information.7 By integratingRaman spectroscopy with STM, tip-enhanced Raman spec-troscopy (TERS) can offer an alternative way to provide localchemical and topographic information with high spatial

resolution.8,9 In TERS, the highly confined plasmonic field inthe tip−sample junction can enhance the detection sensitivityand improve the spatial resolution. These unique features haveattracted significant research interest.10−12 By exploitation ofthe fundamental merits of TERS, single-molecule sensitivityand nanometer spatial resolution have been successfullyachieved in TERS studies under ultrahigh vacuum (UHV)conditions13−17 where low background pressure reducescontamination on the samples and tips and provides greatstability of the tip−sample junction. In recent studies, TERShas been used to reveal critical chemical identification underliquid and EC conditions for a number of importantreactions.18−22 However, molecularly resolved images havenot been achieved in these studies, and the lack of cleartopographic information could complicate the interpretation ofthe acquired signal. The combination of EC-TERS and EC-STM has the potential to simultaneously provide molecularlyresolved topographic information and chemical identification,which is especially suited for understanding electrocatalysis.Cobalt-based molecular catalyst is a promising candidate

that finds application in electrochemical reactions such as the

Received: January 17, 2019Revised: March 27, 2019Published: April 1, 2019

Article

pubs.acs.org/JPCCCite This: J. Phys. Chem. C 2019, 123, 9852−9859

© 2019 American Chemical Society 9852 DOI: 10.1021/acs.jpcc.9b00513J. Phys. Chem. C 2019, 123, 9852−9859

Dow

nloa

ded

via

NO

RT

HW

EST

ER

N U

NIV

on

July

9, 2

019

at 2

2:41

:31

(UT

C).

See

http

s://p

ubs.

acs.

org/

shar

ingg

uide

lines

for

opt

ions

on

how

to le

gitim

atel

y sh

are

publ

ishe

d ar

ticle

s.

oxygen reduction reaction (ORR),23,24 which is an importantreaction for fuel cell applications. The state-of-the-art industrialcatalysts for proton exchange membrane (PEM) fuel cells25

often consist of platinum- (Pt-) based catalysts. Compared tothese catalysts, cobalt-based molecular catalysts show advan-tages in terms of cost and chemical stability. A large body ofempirical and theoretical studies have been carried out toexamine the catalytic activity of cobalt-based molecularcatalysts and their interaction with O2. For example, well-defined monolayers of CoPc molecules supported on Au(111)have been demonstrated by EC-STM.26,27 Additionally,chemisorption of O2 on well-ordered CoPc monolayers hasprovided strong evidence for O2 activation by CoPc whensupported on the Ag(111) surface under UHV conditions.28

However, simultaneous acquisition of kinetic data, chemicalinformation, and molecular configuration in situ for CoPc hasnot been accomplished. Herein, we combine EC-TERS andEC-STM to investigate CoPc molecules during the oxygenreduction reaction. Following the onset of ORR, EC-STMresults illustrate a rapid consumption of O2 molecules thatwere initially adsorbed on CoPc, which is consistent with itshigh activity for hydrogen peroxide production via a two-electron reduction pathway in ORR. As the substrate potentialbecomes more negative, distinct trends in the EC-TERSmeasurements with three excitation wavelengths are observed.This phenomenon is explained by the changes in the resonantconditions of CoPc with different excitation wavelengthsfollowing a redshift in the absorption band of these moleculesupon reduction, which is confirmed by spectroelectrochemicalmeasurements. Our results provide a molecular description ofCoPc under steady state conditions during ORR andadditional insights of the electrochemical reactions catalyzedby CoPc at the solid/liquid interface.

■ EXPERIMENTAL SECTIONSample Preparation. CoPc (Figure 1a) was purchased

from Sigma-Aldrich Co. and used without further purification.For UHV-STM-TERS studies, the Au(111) single crystals

(Princeton Scientific Corporation) were cleaned by multiplesputter−annealing cycles. STM was used to confirm a cleanand well-defined Au(111) surface. The adlayer of CoPc wasprepared by sublimation of CoPc molecules at ∼650 K ontothe clean Au(111) surface under UHV conditions.For EC-STM-TERS studies, the Au(111) single crystal was

used as the substrate and chemically cleaned in freshly

prepared piranha solution (1:3 mixture of H2O2 and H2SO4by volume) for 20 min, followed by thorough rinsing withplenty of Milli-Q water, and immersion in pure ethanol for 10min to remove the surface oxide. After chemical cleaning, thesingle crystal was annealed with a butane torch and cooleddown to room temperature in air.The adlayer of CoPc was prepared by immersing a flame-

annealed Au(111) crystal in a CoPc saturated solution of N,N-dimethylformamide (DMF) for 5 min. After incubation, theCoPc−Au(111) electrode was rinsed with Milli-Q water,transferred into an electrochemical cell and immersed inelectrolyte solution for cyclic voltammetry (CV), EC-STM andEC-TERS measurements. All CV measurements, EC-STM andEC-TERS experiments were carried out in the same cell, exceptfor the spectroelectrochemical absorption measurements. Thesupporting electrolyte solution (0.1 M HClO4) was preparedwith HClO4 (Sigma-Aldrich, 70%) and Milli-Q water (18.2MΩ·cm, total organic carbon <5 ppb) to minimize thecontamination on samples and tips during experiments.

Experimental Setups. The UHV-STM TERS experimentswere performed at room temperature in a home-built UHV-STM system under a base pressure of ∼2 × 10−11 Torr,equipped with side-illumination optical system as detailed inprevious studies.28,29 A 632.8 nm laser (He−Ne laser, ResearchElectro-optics) was employed in the UHV-TERS experiments.The EC-STM and EC-TERS experiments were performed

using a home-built apparatus, equipped with a side-illumination optical system (see Figure S1 in the SupportingInformation for more details). A platinum (Pt) wire was usedas the counter electrode. A silver (Ag) wire was used as thequasi-reference electrode (QRE). All electrode potentialsdiscussed are referenced to the reversible hydrogen electrode(RHE) unless noted otherwise. The potential of silver wire vsRHE is 0.5 V. Fluctuations in the silver wire potential beforeand after experiment was less than 30 mV. Three excitationwavelengths were used in the EC-TERS measurements: 632.8nm (diode laser, PD-LD), 700 and 720 nm (SolsTiS, MSquared Laser Systems). All Raman spectra presented herewere not corrected for the wavelength dependent sensitivity ofphoton detection systems.The STM images were obtained with W tips, and the TER

spectra were obtained with Ag tips. All the UHV-STM andTERS measurements were obtained under constant currentmode with the sample biased. For EC-STM and EC-TERSmeasurements, the tips were insulated with nail polish to

Figure 1. (a) Chemical structure of CoPc molecule. (b) STM image showing a well-ordered CoPc monolayer on Au(111) under UHV conditions.The inset shows a higher magnification image overlaid with a square unit cell of 1.4 nm lattice constant. Imaging conditions: bias = −0.5 V, I = 0.5nA. (c) Tip-engaged (blue), tip-retracted (black), and TERS (red) spectra of CoPc on Au(111) under UHV conditions with 632.8 nm excitation.TERS conditions: bias = 0.1 V, I = 0.5 nA, 0.4 mW. The integration time (Itime) for each spectrum is 180 s.

The Journal of Physical Chemistry C Article

DOI: 10.1021/acs.jpcc.9b00513J. Phys. Chem. C 2019, 123, 9852−9859

9853

reduce the faradaic current, and the STM bias is defined as thedifference between the sample potential (Esample) and the tippotential (Etip), i.e., bias = Esample − Etip. The sample potentialand the STM bias can be controlled independently, in whichthe former was controlled by a potentiostat (EC301, StanfordResearch Systems) and the latter was maintained by the STMelectronics controller (R9, RHK Technology). As a result, thetip potential is defined as Etip = Esample − bias. Different tipswere used for each excitation wavelength.Spectroelectrochemical Absorption Measurements.

Spectroelectrochemical absorption measurements of CoPcmolecules were performed on a double-beam spectropho-tometer (Cary 5000, Agilent) with a potentiostat (660D, CHInstruments) in pyridine with 0.3 M tetrabutylammoniumperchlorate (TBAP, Sigma, 99%) as the supporting electrolyte.Pyridine used for spectroelectrochemical absorption and cyclicvoltammetry measurements was previously dried with NaOHpellets for several hours, followed with nitrogen gas purging for0.5 h. A quartz cell (10 mm optical length, Starna Cells, Inc.)was used with a Pt mesh (Bioanalytical Systems, Inc.) asworking electrode, a Pt wire (Alfa Aesar, 99.99%) as counterelectrode, and a leakfree Ag/AgCl unit as reference electrode.

■ RESULTS AND DISCUSSIONUHV provides a highly controlled environment for TERSstudies. Prior to EC-TERS studies, UHV-TERS measurementswere performed to obtain reference spectra of CoPc adsorbedon Au(111) in a controlled environment. Figure 1b shows alarge-scale STM image of a well-ordered monolayer of CoPcformed on a Au(111) surface under UHV conditions. Thehigh-resolution image in the inset shows that the CoPcmolecules self-assemble into a square unit cell with a latticeconstant of ∼1.4 nm, which is in good agreement with previousstudies on CoPc molecules.28,30 The ordered adlayer formationis attributed to the competition between the intermolecularinteractions and molecule−substrate interactions.The UHV-TER spectra acquired on this well-ordered

molecular sample are shown in Figure 1c. When the tip isengaged, the acquired near-field signal (blue spectrum) showsclear vibrational fingerprints of CoPc molecules. When the tipis retracted from the surface, only a featureless background isobserved in the far-field signal (black spectrum), whichsuggests that the tip is free of contamination. The TERSsignal only from CoPc can be obtained by subtracting the far-field signal from the near-field signal, which is shown in red.After we measured a reference TER spectrum for CoPc

molecules on Au(111), TERS experiments were subsequentlyperformed on CoPc molecules under EC conditions. A typicalEC-STM image of the flame-annealed Au(111) in 0.1 MHClO4 is shown in Figure 2a. When the Au(111) electrode isheld at 0.2 V, the electrochemically induced (22 ×√3) surfacereconstruction can be observed. The inset shows an area largerthan 300 × 300 nm2 with a single step edge. Clear herringbonestructures and step edges suggest a well-defined Au(111)surface with low density of defects at the nanoscale. Figure 2bshows a typical CV of Au(111) in HClO4 solution with twooxidation peaks (1.53 and 1.33 V) and one reduction peak(1.13 V) with two small waves as shoulders.31 These featuressuggest that the Au(111) surface is well-defined over a largescalean ideal starting substrate for the subsequent EC-TERSexperiments. The single electrochemical wave at low potentialis assigned as the oxygen reduction reaction catalyzed byAu(111) surface.

Molecular self-assembly at the solid/liquid interface isguided by intermolecular interactions, molecule−substrateinteractions, molecule-solvent interactions, and the solvent−substrate interactions.32,33 We successfully obtained a highlyordered monolayer of CoPc molecules on Au(111) under ECconditions as shown in Figure 2c. The detailed structure of theself-assembled adlayer can be observed at higher magnification(Figure 2c, inset). The CoPc molecules form a square unit cellwith a lattice constant of 1.4 nm, in good agreement with theordered structures obtained under UHV conditions in Figure1b.Our EC-TERS study was performed on a well-ordered

CoPc−Au(111) sample with Ag tips, which provide a largerfield enhancement than the commonly used gold tips.20,21,34,35

On the basis of our EC-TERS experiment of CoPc onAu(111), about 25 out of 120 tips can provide TERS signalwith signal-to-noise ratio more than 0.1 ADU/mW/s underEC conditions.Typical TER spectra of CoPc at the solid/liquid interface are

shown in Figure 2d. With the Ag tips retracted from thesurface, only one peak at 931 cm−1 is observed (blackspectrum), which is attributed to the symmetric stretching ofthe ClO4

− in the electrolyte solution. The absence of CoPcRaman bands in the tip-retracted spectrum indicates that thetip is free of contamination. When the tip is engaged, theacquired near-field signal (blue spectrum) shows clearvibrational fingerprints of CoPc and ClO4

−. Upon backgroundsubtraction, the TERS signal from only CoPc can be obtained,

Figure 2. (a) EC-STM image of a clean Au(111) at 0.2 V in 0.1 MHClO4 with herringbone reconstruction. The inset shows a large-scaleimage of the Au(111) surface with a single step edge. Imagingconditions: Esample = 0.2 V, Etip = 0.3 V, bias = −0.1 V, I = 0.3 nA. (b)CV of Au(111) in 0.1 M HClO4 saturated by air recorded at 20 mV/s.(c) EC-STM image of CoPc on Au(111) in 0.1 M HClO4 with theAu(111) held at 0.45 V. The inset shows the zoom-in imageoverlapped with a square unit cell with lattice constant as ∼1.4 nm.Imaging conditions: Esample = 0.45 V, Etip = 0.7 V, bias = −0.25 V, I =0.3 nA. (d) Tip-engaged (blue), tip-retracted (red), and TERS (red)spectra of CoPc adsorbed on Au(111) in aqueous solution with 632.8nm laser excitation. TERS conditions: Esample = 0.45 V, Etip = 0.5 V,bias = −0.05 V, I = 0.3 nA, 0.7 mW. The integration time (Itime) foreach spectrum is 10 s.

The Journal of Physical Chemistry C Article

DOI: 10.1021/acs.jpcc.9b00513J. Phys. Chem. C 2019, 123, 9852−9859

9854

shown as the red spectrum in Figure 2d. No significantdifference can be found between the TER spectra acquiredunder UHV and EC conditions, as detailed in the SupportingInformation, Table S1. We note that two vibrational modes at1105 and 1305 cm−1 are observed in EC-TER spectra but notin the UHV-TERS results. However, careful examinationreveals that these two peaks, which are related to Raman activemodes of CoPc molecules, do exist in the UHV-TERS dataalbeit at a low signal-to-noise ratio. Additionally, smallvariations in the peak positions (1−3 cm−1) in UHV- andEC-TERS are within the spectral resolution of our instruments.The discrepancies in the relative intensity of peaks in two TERspectra are attributed to the variation in plasmonic enhance-ment, which can be different between individual tips anddielectric environments. The direct comparison of the STMand TERS results of CoPc on Au(111) under UHV and ECconditions reveals that the electrochemically controlledenvironment can provide a sufficiently stable medium forTERS studies.One major advantage of EC-STM TERS over UHV-STM

TERS is that it replicates a realistic reaction environment for insitu measurements. In addition, the substrate potential can beadjusted independently of the STM bias, which enablesreaction rate control by changing the substrate potential.Figure 3a shows EC-TER spectra of CoPc adsorbed onAu(111) as a function of substrate potential with threeexcitation wavelengths. Recent research suggests that the hugeenhancement in TERS is resulted from the atomic structure of

the tip apex.36,37 In order to minimize the possible reactions(e.g., Ag oxidation, surface reconstruction) that would changethe atomic structure on the tip apex, the Ag tips were held at0.5 V vs RHE. The faradaic current flowing through the tip wasless than 10 pA during EC-TERS measurements. We wouldlike to note that dendritic growths could occur on the Ag tipseven under the open circuit potential conditions. In any givenTERS experiment, the Ag tips were held at the 0 V vs Ag QREand used for no longer than 3 h to minimize side reactions thatmight change the geometry of the Ag tips. No significantdendrite formation on the Ag tips is observed based on thescanning electron micrographs (SEM) of the tip before andafter the EC experiment, as detailed in Figure S2 in theSupporting Information. Additionally, tip-retracted spectrawere acquired after tip-engaged ones under each potential tomake sure the tips were not contaminated by CoPc molecules.The background-subtracted TER spectra are shown in Figure3a (see Figure S3 in the Supporting Information for all dataincluding the tip-retracted spectra).The left panel of Figure 3a shows the EC-TER spectra of

CoPc at different substrate potentials acquired with 632.8 nmexcitation. The TER spectra are observed over a wide range ofpotentials from 0.45 to 0.2 V. No significant change in peakpositions is observed in the EC-TER spectra. However, adecrease in the TERS intensity is observed as the potentialbecomes more negative. Furthermore, EC-TER spectra excitedby 700 and 720 nm lasers show potential-dependent behaviordifferent from the 632.8 nm case. As the substrate potential

Figure 3. (a) EC-TERS study of CoPc on Au(111) acquired at different potentials with three excitation wavelengths in 0.1 M HClO4. Ag tips wereheld at 0.5 V vs RHE during the EC-TERS measurement to minimize the faradaic current. TERS conditions: 632.8 nm laser (Etip = 0.5 V, I = 0.5nA, Itime = 10 s, 0.7 mW), 700 nm laser (Etip = 0.5 V, I = 0.5 nA, Itime = 20 s, 1.3 mW), 720 nm laser (Etip = 0.5 V, I = 0.3 nA, Itime = 20 s, 2 mW).Esample is as labeled. (b) Sequential STM images of the CoPc adlayer adsorbed on Au(111) at different substrate potentials in 0.1 M HClO4. The Wtips used for EC-STM measurement were held at 0.5 V vs RHE. Imaging condition: I = 0.3 nA. Esample is as labeled.

The Journal of Physical Chemistry C Article

DOI: 10.1021/acs.jpcc.9b00513J. Phys. Chem. C 2019, 123, 9852−9859

9855

becomes increasingly negative, the TERS intensity increaseswith 700 nm excitation, but remains stable with 720 nmexcitation. The different trends in the EC-TERS experimentswith three excitation wavelengths also suggest that the biasbetween the tip and substrate has negligent effects on theobserved EC behaviors. Additionally, the TERS signal almostcompletely disappears when the substrate potential is held at0.1 V for all three excitation wavelengths.On the basis of the results from Figure 3a, we summarize

two main observations. First, a common characteristic amongthe EC-TER spectra collected at different excitation wave-lengths is the disappearance of EC-TER spectrum at 0.1 V.Second, a potential-dependent changes in the EC-TERSintensity were observed and the trend of such potentialdependence varies based on the excitation wavelength.One explanation for these distinct behaviors in the EC-

TERS experiments with three excitation wavelengths could berelated to the adsorption behavior of CoPc molecules at thesolid/liquid interface. In a recent study, the TERS spectralchange of adenine molecules due to different applied potentialswas attributed to a potential-dependent orientation change ofthis molecule.21 In order to examine the possibility of apotential-dependent change in the adsorption configuration ofCoPc on Au(111), EC-STM was performed under different

substrate potentials. During the EC-STM measurements, Wtips were used and held at 0.5 V vs RHE to reproduce theconditions during EC-TERS measurements. A series of STMimages in Figure 3b illustrate that a well-ordered CoPc adlayerforms on Au(111) in the potential range from 0.45 to 0.2 V,while a transition from the ordered phase to a diffusing phase isobserved at 0.1 V. The diffusing phase of CoPc molecules athigh cathodic potential (<0.1 V) is in good agreement with theabsence of EC-TERS signal at 0.1 V in Figure 3a. Underelectrochemical conditions, shifting the substrate potentialnegatively would increase the surface charge density and resultin a weaker molecule−substrate interaction, which can causethe ordered-to-diffusion phase transition.38,39 The ordered-to-diffusing phase transition is also dependent on the STM bias.As shown in Figure S4 in the Supporting Information, theordered-to-diffusing phase transition of CoPc occurred at −0.1V when the STM bias was fixed at 0.1 V.The ordered-to-diffusing phase transition provides a

satisfactory explanation for the disappearance of TERSspectrum at large negative applied potentials (<0.1 V), but itdoes not explain the three trends in the potential-dependentTERS intensity with different excitations prior to the phasetransition (i.e., ordered phase, >0.1 V). We explain this

Figure 4. (a) Typical CVs of Au(111) electrodes (dash line) and CoPc−Au(111) electrodes (solid line) in 0.1 M HClO4 recorded at 50 mV/ssaturated by air (blue profile) and argon (red profile), respectively. (b) Time evolution of UV−visible absorption spectra of CoPc in pyridine whenthe working electrode potential was applied at the first reduction peak (−0.7 V vs Ag/AgCl reference electrode). Supporting electrolyte: 0.3 MTBAP. (c) High-resolution STM images and corresponding line profiles of CoPc−Au(111) under different potentials in 0.1 M HClO4. Imagingconditions: Esample as labeled, bias = 0.1 to 0.25 V, I = 0.3 to 1 nA.

The Journal of Physical Chemistry C Article

DOI: 10.1021/acs.jpcc.9b00513J. Phys. Chem. C 2019, 123, 9852−9859

9856

observation by considering the resonance conditions that arecritical to the TERS measurements.In a TERS measurement, achieving a resonance condition

between the excitation wavelength and molecular absorptioncan lead to a substantial increase in the overall signal intensity.Our previous EC-TERS study on Nile Blue demonstrates thatthe TERS intensity can appear or disappear entirely dependingon the spectral overlap of the excitation wavelength with themolecular resonance.19,35 Consequently, we expect that achange in the resonance condition induced by the appliedpotential would be reflected as a change in the TERS intensityfor CoPc as well. To confirm our hypothesis, CV andspectroelectrochemical absorption measurements were per-formed to offer further insight into the electrochemistry ofCoPc molecules at the solid/liquid interface, as shown inFigure 4.Figure 4a shows the CV of CoPc on Au(111) in 0.1 M

HClO4 without oxygen. In the Ar-saturated electrolyte, a redoxcouple is observed with formal potential E0

CV = 0.20 V, whichis assigned to the transition between the CoIIPc and the CoIPcanion. The absorbance measurement from the CoPcmonolayer adsorbed on metal surface was too weak formaking a solid conclusion about the molecular electronictransitions. Instead, Figure 4b illustrates the time evolution ofthe absorption spectrum of CoPc in pyridine solution when theworking electrode potential is −0.7 V vs Ag/AgCl (see the CVof CoPc in pyridine in Figure S5 in the SupportingInformation). Initially, the CoPc molecules dissolved inpyridine solution are neutral and show a Q-band centered at657 nm. As the reduction reaction progresses, a clearattenuation in the Q-band intensity is observed due to adecrease in the concentration of neural CoIIPc molecules andsimultaneously a new absorption feature located at 700 nmappears and increases in intensity as the population of reducedCoIPc anion grows. Such shifting in the Q-band of reducedCoIPc is in good agreement with the reported literatures.40

Similar to the change in the Q-band of the CoPc moleculedissolved in pyridine, we hypothesis that these moleculesadsorbed on Au(111) surface are likely to experience a redshiftin the Q-band as the substrate potential becomes morenegative. As a result, the red-shifted Q-band leads to adecreased overlap between the molecular electronic transitionand the 632.8 nm excitation wavelength, and a subsequentdecrease in TERS signal. In contrast, the resonance betweenthe electronic transition of a reduced CoPc and the excitationwavelength increases with 700 nm excitation, which leads to anincrease in the TERS intensity. For the 720 nm excitation, theelectronic resonance increases but the TERS intensity isunaffected as the substrate potential becomes more negative,which might be related to the interaction between moleculesand substrate.Furthermore, EC tip-SERS experiment with 632.8 nm

excitation was performed to study the electrochemical processby modifying the tip with CoPc molecules, as shown in theFigure S6 in the Supporting Information. A decreasing in theintensity is also observed in the tip-SERS signal as the potentialbecomes more negative. The E0

tip‑SERS acquired from the tip-SERS voltammogram is about 0.23 V, which is in goodagreement with the E0

CV value (0.20 V) as obtained from CVin Figure 4a. Such agreement further suggests that the threetrends in the potential-dependent TERS intensity withdifferent excitations are the result of CoPc reduction at morenegative applied potential.

In addition to the reduction of CoPc molecules as thesubstrate potential becomes more negative, the CoPc-catalyzedORR is also known to occur in the same potential window (0−0.5 V vs RHE). Our current EC-STM TERS setup is opened tothe air, which provides a constant supply of oxygen for theORR. As shown in Figure 4a, the onset for ORR on bareAu(111) surface is observed at 0.3 V, and the onset for ORRon the CoPc−Au(111) surface is observed at 0.45 V duringcathodic scanning in the electrolyte exposed to air. A clearelectrocatalytic reduction peak of O2 is found at 0.23 V,indicating that the ORR is catalyzed by CoPc. During ORR, aCoPc molecule is first oxidized by adsorbing an O2 to form aCoPc−O2 complex. Following this, the adsorbed O2 receivestwo electrons and protons to form H2O2, which quicklydesorbs and the oxidized CoPc molecule is simultaneouslyreduced by the substrate.41,42 Previous studies have reportedTERS signal related to adduct formation between different gasmolecules (CO, O2, etc.) and Co-based macrocycles.17,28,43

However, no spectral evidence of CoPc−O2 complex wasobserved in our EC-TERS experiment, which can be attributedto the combine effect of the low Raman cross section of oxygenrelated modes and the low population of the CoPc−O2complex due to the rapid consumption of O2 under ORRpotential. With respect to the second hypothesis, molecular-resolution EC-STM images were performed.On the basis of the studies on CoTPP and FePc by Wan’s

group,44,45 the CoPc−O2 complex would show a highercontrast in the EC-STM image, as shown in Figure S7 in theSupporting Information. We successfully achieved molecular-resolution EC-STM images at different substrate potentials bytuning the W tip potential, as shown in Figure 4c. The heightprofiles extracted along the marked lines are presented underthe corresponding STM images and shown consistent contrastin all STM images acquired at potentials between 0.45 and 0.2V (i.e., during ORR). The consistent contrast suggests a rapidconsummation of O2 and all observed molecules are CoPc.From the EC-TERS and EC-STM experiments, no evidenceabout the CoPc−O2 complex was observed under ORRpotential. In order to reveal the detailed ORR mechanismusing the EC-TERS technique, catalytic systems having stableintermediates with a strong metal−intermediate interactionand a large Raman cross-section that enables facile TERSacquisition are needed.

■ CONCLUSIONSTo summarize, a systematic study that employs EC-TERS withthree excitation wavelengths, in situ EC-STM, and spectroelec-trochemistry absorption measurements has been designed andimplemented to understand the electrochemical processes onCoPc molecules at the solid/liquid interface. As the substratepotential becomes increasingly negative, the well-orderedadlayer of CoPc is stable on a clean, well-defined Au(111)surface until the ordered-to-diffusing phase transition istriggered. Upon reduction, the electronic transition of CoPcredshifts from 657 to 700 nm, changes the resonanceconditions at different excitation wavelengths and results indifferent trends in the EC-TERS measurements with threeexcitation wavelengths. The ordered-to-diffusing phase tran-sition of CoPc results in the disappearance of EC-TERS signalwith different excitation wavelengths. The CoPc-catalyzedORR process has a negligible effect either on the EC-STM oron EC-TERS measurements. Our results contribute newunderstanding to the electrochemical behavior of the catalyti-

The Journal of Physical Chemistry C Article

DOI: 10.1021/acs.jpcc.9b00513J. Phys. Chem. C 2019, 123, 9852−9859

9857

cally active CoPc molecules at the solid/liquid interface underrealistic ORR conditions.

■ ASSOCIATED CONTENT*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acs.jpcc.9b00513.

Experimental setup, peak assignments, SEM images ofthe silver tips before and after EC experiment, raw dataof EC-TERS with three excitation wavelengths, EC-STMon CoPc with constant bias (0.1 V), CV of CoPc inpyridine, EC tip-SERS, and ORR monitored by EC-STM (PDF)

■ AUTHOR INFORMATIONCorresponding Author*(R.P.V.D.) E-mail: vanduyne@northwestern.edu. Telephone:847-491-3516. Fax: 847-491-7713.ORCIDSong Jiang: 0000-0002-2223-8517Zhu Chen: 0000-0001-5889-5415Xu Chen: 0000-0002-2603-4837Duc Nguyen: 0000-0002-6591-6429Michael Mattei: 0000-0002-8276-5562Guillaume Goubert: 0000-0002-4325-0267Richard P. Van Duyne: 0000-0001-8861-2228Present Addresses∥Department of Chemistry, University of WisconsinMadison, Madison, WI 53706§Laboratory of Organic Chemistry, D-CHAB, ETH Zurich,Vladimir-Prelog-Weg 3, 8093 Zurich, SwitzerlandNotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThe authors acknowledge financial support from the Air ForceOffice of Scientific Research MURI (FA9550-14-1-0003). D.N.and R.P.V.D. acknowledge support from the National ScienceFoundation Center for Chemical Innovation dedicated toChemistry at the Space-Time Limit (CaSTL) Grant CHE-1414466. The authors thank Dr. Li-Jun Wan, Dr. Dong Wang,Dr. Allen J. Bard, Dr. Henry White, Dr. Martin Edwards, andDr. Katherine A. Willets for helpful discussions.

■ REFERENCES(1) Arico, A. S.; Bruce, P.; Scrosati, B.; Tarascon, J.-M.; vanSchalkwijk, W. Nanostructured Materials for Advanced EnergyConversion and Storage Devices. Nat. Mater. 2005, 4, 366−377.(2) Seh, Z. W.; Kibsgaard, J.; Dickens, C. F.; Chorkendorff, I.;Nørskov, J. K.; Jaramillo, T. F. Combining Theory and Experiment inElectrocatalysis: Insights into Materials Design. Science 2017, 355,No. eaad4998.(3) Zaera, F. Probing Catalytic Reactions at Surfaces. Prog. Surf. Sci.2001, 69, 1−98.(4) Zaera, F. Probing Liquid/Solid Interfaces at the Molecular Level.Chem. Rev. 2012, 112, 2920−2986.(5) Wan, L.-J. Fabricating and Controlling Molecular Self-Organization at Solid Surfaces: Studies by Scanning TunnelingMicroscopy. Acc. Chem. Res. 2006, 39, 334−342.(6) Zhang, J.; Welinder, A. C.; Chi, Q.; Ulstrup, J. ElectrochemicallyControlled Self-Assembled Monolayers Characterized with Molecular

and Sub-Molecular Resolution. Phys. Chem. Chem. Phys. 2011, 13,5526−5545.(7) Dong, J.-C.; Zhang, X.-G.; Briega-Martos, V.; Jin, X.; Yang, J.;Chen, S.; Yang, Z.-L.; Wu, D.-Y.; Feliu, J. M.; Williams, C. T.; Tian,Z.-Q.; Li, J.-F. In situ Raman Spectroscopic Evidence for OxygenReduction Reaction Intermediates at Platinum Single-crystal Surfaces.Nat. Energy 2019, 4, 60−67.(8) Yeo, B. S.; Stadler, J.; Schmid, T.; Zenobi, R.; Zhang, W. H. Tip-Enhanced Raman Spectroscopy - Its Status, Challenges and FutureDirections. Chem. Phys. Lett. 2009, 472, 1−13.(9) Schmid, T.; Opilik, L.; Blum, C.; Zenobi, R. Nanoscale ChemicalImaging Using Tip-Enhanced Raman Spectroscopy: A CriticalReview. Angew. Chem., Int. Ed. 2013, 52, 5940−5954.(10) Deckert-Gaudig, T.; Taguchi, A.; Kawata, S.; Deckert, V. Tip-Enhanced Raman Spectroscopy - From Early Developments to RecentAdvances. Chem. Soc. Rev. 2017, 46, 4077−4110.(11) Wang, X.; Huang, S.-C.; Huang, T.-X.; Su, H.-S.; Zhong, J.-H.;Zeng, Z.-C.; Li, M.-H.; Ren, B. Tip-Enhanced Raman Spectroscopyfor Surfaces and Interfaces. Chem. Soc. Rev. 2017, 46, 4020−4041.(12) Pozzi, E. A.; Goubert, G.; Chiang, N.; Jiang, N.; Chapman, C.T.; McAnally, M. O.; Henry, A.-I.; Seideman, T.; Schatz, G. C.;Hersam, M. C.; Duyne, R. P. V. Ultrahigh-Vacuum Tip-EnhancedRaman Spectroscopy. Chem. Rev. 2017, 117, 4961−4982.(13) Zhang, R.; Zhang, Y.; Dong, Z. C.; Jiang, S.; Zhang, C.; Chen,L. G.; Zhang, L.; Liao, Y.; Aizpurua, J.; Luo, Y.; Yang, J. L.; Hou, J. G.Chemical Mapping of a Single Molecule by Plasmon-EnhancedRaman Scattering. Nature 2013, 498, 82−86.(14) Jiang, S.; Zhang, Y.; Zhang, R.; Hu, C.; Liao, M.; Luo, Y.; Yang,J.; Dong, Z.; Hou, J. G. Distinguishing Adjacent Molecules on aSurface Using Plasmon-Enhanced Raman Scattering. Nat. Nano-technol. 2015, 10, 865−869.(15) Chiang, N.; Chen, X.; Goubert, G.; Chulhai, D. V.; Chen, X.;Pozzi, E. A.; Jiang, N.; Hersam, M. C.; Seideman, T.; Jensen, L.; VanDuyne, R. P. Conformational Contrast of Surface-Mediated MolecularSwitches Yields Angstrom-Scale Spatial Resolution in UltrahighVacuum Tip-Enhanced Raman Spectroscopy. Nano Lett. 2016, 16,7774−7778.(16) Jiang, S.; Zhang, X.; Zhang, Y.; Hu, C.; Zhang, R.; Zhang, Y.;Liao, Y.; Smith, Z. J.; Dong, Z.; Hou, J. G. Subnanometer-ResolvedChemical Imaging via Multivariate Analysis of Tip-Enhanced RamanMaps. Light: Sci. Appl. 2017, 6, No. e17098.(17) Lee, J.; Tallarida, N.; Chen, X.; Liu, P.; Jensen, L.; Apkarian, V.A. Tip-Enhanced Raman Spectromicroscopy of Co(II)-Tetraphenyl-porphyrin on Au(111): Toward the Chemists’Microscope. ACS Nano2017, 11, 11466−11474.(18) Schmid, T.; Yeo, B. S.; Leong, G.; Stadler, J.; Zenobi, R.Performing Tip-Enhanced Raman Spectroscopy in Liquids. J. RamanSpectrosc. 2009, 40, 1392−1399.(19) Kurouski, D.; Mattei, M.; Van Duyne, R. P. Probing RedoxReactions at the Nanoscale with Electrochemical Tip-EnhancedRaman Spectroscopy. Nano Lett. 2015, 15, 7956−7962.(20) Zeng, Z.-C.; Huang, S.-C.; Wu, D.-Y.; Meng, L.-Y.; Li, M.-H.;Huang, T.-X.; Zhong, J.-H.; Wang, X.; Yang, Z.-L.; Ren, B.Electrochemical Tip-Enhanced Raman Spectroscopy. J. Am. Chem.Soc. 2015, 137, 11928−11931.(21) Martín Sabanes, N.; Ohto, T.; Andrienko, D.; Nagata, Y.;Domke, K. F. Electrochemical TERS Elucidates Potential-InducedMolecular Reorientation of Adenine/Au(111). Angew. Chem., Int. Ed.2017, 56, 9796−9801.(22) Kumar, N.; Su, W.; Vesely, M.; Weckhuysen, B. M.; Pollard, A.J.; Wain, A. J. Nanoscale Chemical Imaging of Solid-Liquid InterfacesUsing Tip-Enhanced Raman Spectroscopy. Nanoscale 2018, 10,1815−1824.(23) Sorokin, A. B. Phthalocyanine Metal Complexes in Catalysis.Chem. Rev. 2013, 113, 8152−8191.(24) Dey, S.; Mondal, B.; Chatterjee, S.; Rana, A.; Amanullah, S.;Dey, A. Molecular Electrocatalysts for the Oxygen ReductionReaction. Nat. Rev. Chem. 2017, 1, 0098.

The Journal of Physical Chemistry C Article

DOI: 10.1021/acs.jpcc.9b00513J. Phys. Chem. C 2019, 123, 9852−9859

9858

(25) Jung, N.; Chung, D. Y.; Ryu, J.; Yoo, S. J.; Sung, Y.-E. Pt-basedNanoarchitecture and Catalyst Design for Fuel Cell Applications.Nano Today 2014, 9, 433−456.(26) Yoshimoto, S.; Tada, A.; Suto, K.; Itaya, K. Adlayer Structuresand Electrocatalytic Activity for O2 of Metallophthalocyanines onAu(111): In Situ Scanning Tunneling Microscopy Study. J. Phys.Chem. B 2003, 107, 5836−5843.(27) Suto, K.; Yoshimoto, S.; Itaya, K. Two-Dimensional Self-Organization of Phthalocyanine and Porphyrin: Dependence on theCrystallographic Orientation of Au. J. Am. Chem. Soc. 2003, 125,14976−14977.(28) Nguyen, D.; Kang, G.; Chiang, N.; Chen, X.; Seideman, T.;Hersam, M. C.; Schatz, G. C.; Van Duyne, R. P. Probing Molecular-Scale Catalytic Interactions between Oxygen and Cobalt Phthalocya-nine Using Tip-Enhanced Raman Spectroscopy. J. Am. Chem. Soc.2018, 140, 5948−5954.(29) Jiang, N.; Foley, E. T.; Klingsporn, J. M.; Sonntag, M. D.;Valley, N. A.; Dieringer, J. A.; Seideman, T.; Schatz, G. C.; Hersam,M. C.; Van Duyne, R. P. Observation of Multiple Vibrational Modesin Ultrahigh Vacuum Tip-Enhanced Raman Spectroscopy Combinedwith Molecular-Resolution Scanning Tunneling Microscopy. NanoLett. 2012, 12, 5061−5067.(30) Takada, M.; Tada, H. Low temperature Scanning TunnelingMicroscopy of Phthalocyanine Multilayers on Au(111) Surfaces.Chem. Phys. Lett. 2004, 392, 265−269.(31) Hamelin, A. Cyclic Voltammetry at Gold Single-CrystalSurfaces. Part 1. Behaviour at Low-Index Faces. J. Electroanal. Chem.1996, 407, 1−11.(32) Yoshimoto, S. Molecular Assemblies of Functional Moleculeson Gold Electrode Surfaces Studied by Electrochemical ScanningTunneling Microscopy: Relationship between Function and AdlayerStructures. Bull. Chem. Soc. Jpn. 2006, 79, 1167−1190.(33) Wang, D.; Wan, L.-J. Electrochemical Scanning TunnelingMicroscopy: Adlayer Structure and Reaction at Solid/liquid Interface.J. Phys. Chem. C 2007, 111, 16109−16130.(34) Sabanes, N. M.; Elizabeth, A.; Pfisterer, J. H. K.; Domke, K. F.The Effect of STM Parameters on Tip-Enhanced Raman Spectra.Faraday Discuss. 2017, 205, 233−243.(35) Chen, X.; Goubert, G.; Jiang, S.; Van Duyne, R. P.Electrochemical STM Tip-Enhanced Raman Spectroscopy Study ofElectron Transfer Reactions of Covalently Tethered Chromophoreson Au(111). J. Phys. Chem. C 2018, 122, 11586−11590.(36) Benz, F.; Schmidt, M. K.; Dreismann, A.; Chikkaraddy, R.;Zhang, Y.; Demetriadou, A.; Carnegie, C.; Ohadi, H.; de Nijs, B.;Esteban, R.; Aizpurua, J.; Baumberg, J. J. Single-MoleculeOptomechanics in “Picocavities”. Science 2016, 354, 726−729.(37) Trautmann, S.; Aizpurua, J.; Gotz, I.; Undisz, A.; Dellith, J.;Schneidewind, H.; Rettenmayr, M.; Deckert, V. A ClassicalDescription of Subnanometer Resolution by Atomic Features inMetallic Structures. Nanoscale 2017, 9, 391−401.(38) He, Y.; Ye, T.; Borguet, E. Porphyrin Self-Assembly atElectrochemical Interfaces: Role of Potential Modulated SurfaceMobility. J. Am. Chem. Soc. 2002, 124, 11964−11970.(39) Tao, N. J.; DeRose, J. A.; Lindsay, S. M. Self-assembly ofMolecular Superstructures Studied by in situ Scanning TunnelingMicroscopy: DNA bases on Gold (111). J. Phys. Chem. 1993, 97,910−919.(40) Saka, E. T.; Sarkı, G.; Kantekin, H.; Koca, A. Electrochemical,Spectroelectrochemical and Catalytical Properties of New Cu(II) andCo(II) Phthalocyanines. Synth. Met. 2016, 214, 82−91.(41) Sun, S.; Jiang, N.; Xia, D. Density Functional Theory Study ofthe Oxygen Reduction Reaction on Metalloporphyrins and Metal-lophthalocyanines. J. Phys. Chem. C 2011, 115, 9511−9517.(42) Shao, M.; Chang, Q.; Dodelet, J.-P.; Chenitz, R. RecentAdvances in Electrocatalysts for Oxygen Reduction Reaction. Chem.Rev. 2016, 116, 3594−3657.(43) Domke, K. F.; Pettinger, B. In Situ Discrimination betweenAxially Complexed and Ligand-Free Co Porphyrin on Au(111) with

Tip-Enhanced Raman Spectroscopy. ChemPhysChem 2009, 10,1794−1798.(44) Cai, Z.-F.; Wang, X.; Wang, D.; Wan, L.-J. Cobalt-Porphyrin-Catalyzed Oxygen Reduction Reaction: A Scanning TunnelingMicroscopy Study. ChemElectroChem 2016, 3, 2048−2051.(45) Gu, J.-Y.; Cai, Z.-F.; Wang, D.; Wan, L.-J. Single-MoleculeImaging of Iron-Phthalocyanine-Catalyzed Oxygen Reduction Re-action by in Situ Scanning Tunneling Microscopy. ACS Nano 2016,10, 8746−8750.

The Journal of Physical Chemistry C Article

DOI: 10.1021/acs.jpcc.9b00513J. Phys. Chem. C 2019, 123, 9852−9859

9859