In Situ Surface-Enhanced Raman Spectroscopy of the Electrochemical

Subscriber access provided by Imperial College London | Library

Article

In Situ Surface-Enhanced Raman Spectroscopy of the Electrochemical Reduction of Carbon Dioxide on Silver with 3,5-Diamino-1,2,4-Triazole Kevin G. Schmitt, and Andrew A. Gewirth J. Phys. Chem. C, Just Accepted Manuscript • Publication Date (Web): 26 Jun 2014 Downloaded from http://pubs.acs.org on June 27, 2014

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry C is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

In Situ Surface-Enhanced Raman Spectroscopy of the Electrochemical Reduction of Carbon Dioxide on Silver with 3,5-Diamino-1,2,4-Triazole Kevin G. Schmitt and Andrew A. Gewirth* Department of Chemistry, University of Illinois at Urbana-Champaign, 600 South Mathews Avenue, Urbana, Illinois 61801, United States

*author to whom correspondence should be addressed: email: [email protected]; tel: +1217-333-8329; fax; +1-217-244-3186 1 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 39

ABSTRACT The influence of 3,5-diamino-1,2,4-triazole (DAT) on the electrochemical reduction of carbon dioxide to carbon monoxide on a silver electrode was studied via in situ surface-enhanced Raman spectroscopy (SERS). SERS bands obtained in the absence of DAT indicate potentialdependent adsorption of the CO product to bridge and 3-fold hollow sites. With the addition of DAT, CO adsorption to less-coordinated surface sites was found, including a physisorbed or non-coordinating site that exhibited no significant Stark vibrational shift. Raman peaks associated with adsorbed DAT observed at the same potentials as this species suggest that the ligand promotes weaker CO adsorption, which may be responsible for the high efficiency of the AgDAT catalyst. The observation of potential-dependent methylene stretching vibrations indicates the presence of surface hydrocarbon species while the presence of C-D stretches in deuterated electrolyte confirm that these hydrocarbons are generated as a CO2 reduction byproduct.

KEYWORDS Spectroelectrochemistry, Electrocatalysis, Surface Adsorption, Vibrational Stark Effect

2 ACS Paragon Plus Environment

Page 3 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


I. INTRODUCTION The impact of carbon dioxide on climate as a greenhouse gas and its abundance as a commonly emitted byproduct have led to interest in capturing, storing, and recycling CO2 for use as a potential source of fuel or other chemicals.1 In one scenario, excess electricity from intermittent renewable sources may be used to reduce CO2 either to CO (then reacted with H2 using water gas shift) or directly to hydrocarbons. These hydrocarbons would be either a storage medium for the intermittent energy or a feedstock for existing hydrocarbon demands. Due to the increase in atmospheric CO2 levels there has been much recent activity in electrocatalytic CO2 reduction.2-4 Hori provides a thorough review of CO2 electroreduction at metal electrodes, for which Au, Ag, and Zn are known to exhibit the greatest Faradaic efficiency for producing CO.5 Au and Ag are highly active for CO evolution because CO binds weakly, but the reduction rate is limited at these electrodes by CO2 activation. This behavior contrasts with that of metals such as Pt, Pd, Ni, and Rh, for which the CO2 to CO reaction is facile, but strong CO binding severely limits the catalytic efficiency.6 The mechanism of CO2 reduction at Ag surfaces is presumed to begin with single electron transfer to form CO2•-, stabilized by adsorption on the metal surface.7 Chemisorption of this anion through the carbon atom permits protonation of one of the available oxygen atoms, forming adsorbed COOH. With the addition of another electron from the Ag cathode, this adsorbed species then splits into free OH- and CO, which is weakly-bound and quickly desorbs. In aqueous electrolytes, the standard potential for the reduction of CO2 to CO2•- is -1.85 V8 or 1.90 V9, indicating a large activation barrier to product formation. This large activation barrier has led to suggestions that other species might be involved in CO2 reduction in aqueous solution, particularly HCO3-, since CO2 dissolved in water exists in equilibrium with HCO3-, which is far



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 39

more soluble. Indeed, the direct reduction of HCO3- to HCOO- has been previously reported on Pd10-11 and Cu12 electrodes, and the thermodynamic potential for formate generation (estimated at approx. -0.76 V vs SCE at pH 8.9)13 is much lower than that for CO2•-. In this paper we examine CO2 reduction on bare and decorated Ag electrodes using electrochemical and surface enhanced Raman scattering (SERS) methods. There have been numerous previous studies of CO2 reduction on Ag in a variety of environments, confirming the metal as an effective CO-producing catalyst.14-21 Because CO is the product of CO2 reduction on this electrode and there is general interest in CO ligation to metals, there is considerable spectroscopic work examining CO coordination to Ag in the electrochemical environment. The first Raman spectrum of CO adsorbed to a Ag electrode in aqueous solution showed a band near 2000 cm-1 in 0.1 M KCl or Na2SO4 with a potential-dependence (Stark shift) of approximately 58 cm-1 V-1.22-23 Additional features at 2100 cm-1 and 2167 cm-1 were assigned to weakly-held, easily-desorbed CO and CO adsorbed to a partially oxidized Ag surface, respectively. Other studies report CO stretches at 2021 cm-1, associated with CO linearly bound (atop site) to Ag,24 and at 2014 cm-1 on Ag held at -650 mV vs Ag/AgCl (4 M KCl).25 Addition of Cl- leads to CO stretches at -0.6 V vs SHE at 1998 cm-1. Additionally, a CO adsorption peak on Ag was visible at approximately 1950 cm-1, and this band exhibited higher frequencies with increasing potential.16 Finally, in situ FTIR spectroscopy of CO on unroughened Ag in electrolytes over a wide pH range produced an adsorption peak near 1970-2000 cm-1 for atop-adsorbed CO, and at neutral or alkaline pH, a peak ca. 1860-1900 cm-1 for bridge-bonded CO.23 Additional peaks at 2048 cm-1 or 2112 cm-1 (depending on pH) that did not exhibit a Stark vibrational shift were assigned to CO chemisorption on a Ag underpotential oxide.


Page 5 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


As these studies show, CO is able to adsorb to metal surfaces in several different ways.26 Figure 1 depicts the possible adsorption sites on the close-packed (111) surface of a metal with a face-centered cubic crystal structure, such as Ag. For each of these positions, adsorption occurs between the C atom and 1, 2, or 3 metal atoms, with the C-O axis oriented normal to the plane of the metal. Atop (A), 2-fold bridge (B), and 3-fold hollow adsorption involve coordination to 1, 2, and 3 metal atoms, respectively. CO may occupy 3-fold hollow sites that have either an fcc (C) or hcp (D) orientation, although adsorption to such similar structures can be difficult to differentiate experimentally.

C A

D B

Figure 1: Possible CO adsorption sites on the close-packed (111) face of a metal with an fcc crystal structure: atop (A), 2-fold bridge (B), fcc 3-fold hollow (C), and hcp 3-fold hollow (D).

DFT calculations of CO adsorption on Ag have estimated the vibrational frequencies for the atop, bridge, and hollow sites at approximately 2050 cm-1, 1935 cm-1, and 1895 cm-1, respectively.26 Furthermore, while UHV and DFT studies in the absence of solvent typically find CO adsorption to be most stable at the atop position, DFT calculations suggest that CO prefers adsorption on Ag to the 3-fold hollow site when water occupies the atop site.27 We and others have found that the catalytic activity of bare Ag toward CO2 reduction is improved by solution additives such as 1-ethyl-3-methylimidazolium tetrafluoroborate (EMIM-



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 39

BF4)28 and methylviologen.29 Recently, we showed that nitrogen-based Ag complexes formed from 3,5-diamino-1,2,4-triazole (DAT), pyrazole, or phthalocyanine all exhibited comparable performance for CO production as Ag alone, but with much lower mass loading of Ag. AgDAT, in particular, was found to exhibit the earliest onset potential and greatest Faradaic efficiency for CO evolution, and even improved electrocatalytic performance as a free ligand in solution above bare Ag particles supported on carbon.30 Our present work explores the electrochemical reduction of CO2 on Ag through in situ surface-enhanced Raman spectroscopy (SERS) and electrochemical measurements in order to better understand the effect of DAT on the reaction.

II. EXPERIMENTAL SECTION The electrolyte solution was composed of 1 M KOH prepared from semiconductor grade KOH (99.99% trace metals basis, Sigma-Aldrich) and 18.2 MΩ-cm Milli-Q water (Millipore Inc.). In order to minimize the amount of dissolved CO2 in the electrolyte, the water was boiled for approximately 15 minutes prior to use. Additionally, a saturating quantity of ACS reagent grade Ca(OH)2 (≥95.0%, Sigma-Aldrich) was added to bind CO2 that may be absorbed from the air over time as insoluble CaCO3.31 This electrolyte was used to prepare 10 mM sample solutions of DAT from 3,5-diamino-1,2,4-triazole (98%, Aldrich). Addition of CO2 decreases the pH of sample solutions from approximately 13.3 to approximately 7.8 by the conversion of OH- to HCO3-.5 For Ar control trials, the pH was reduced through the addition of HClO4 (Optima, Fischer Scientific) to match. A 1 M KOD deuterated electrolyte was prepared by diluting 40 wt % KOD (98 atom % D, Aldrich) in D2O (99.9% D, Cambridge Isotope Laboratories).


Page 7 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The working electrode for these measurements was a polycrystalline Ag disk which was progressively polished using 9 µm to 0.25 µm diamond suspensions (MetaDi Supreme, Buehler). After each polishing step, the electrode was sonicated in Milli-Q water and thoroughly rinsed. As Ag surfaces are highly susceptible to contamination by carbon species,32-33 the polished electrode surface was electrochemically cleaned by reducing surface carbon contaminants to more soluble hydrocarbon species.31 In particular, the electrode was held at -3 V vs a Ag/AgCl (3 M KCl) reference electrode for 30 minutes in the KOH/Ca(OH)2 electrolyte solution with a Ag wire as the counter electrode. After rinsing, the clean Ag surface was roughened by an oxidationreduction cycling process in 3 M KCl that has been previously described.34 A CHI760 potentiostat (CH Instruments) was used for all electrochemical experiments. A “no leak” Ag/AgCl electrode (3.4 M KCl, Cypress) was the reference electrode and a flameannealed Pt wire was the counter electrode. All potentials reported here are referenced with respect to Ag/AgCl. In situ SERS measurements utilized a spectroelectrochemical cell described previously.35 A 531.9 nm laser (B&W Tek) provided sample excitation approximately 45° relative to an 85 mm f/1.2 collection lens (Canon). The scattered radiation was then focused to the 50 µm slit of a SpectraPro 2300i monochromator (Princeton Instruments) with grating of 1200 grooves per mm. The CCD detector (Andor) was thermoelectrically cooled to -60° C. Typical acquisition times varied from 5s to 15 s.

III. RESULTS AND DISCUSSION 3.1 Voltammetry



Figure 2 displays cyclic voltammetry obtained from a roughened Ag electrode in a solution containing 1 M KOH + Ca(OH)2 (sat’d) pH-adjusted with HClO4 to pH = 7.8 and purged with Ar (black line).

0.0 2

Current Density (mA/cm )

2 2 Current Density (mA/cm ) ) Current Density (mA/cm

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 39

-0.5 -1.0 -1.5 -2.0

0.08 0.06 0.04 0.02 0.00 -0.02 -1.0

-0.5

0.0

Potential vs Ag/AgCl (V)

-2.5

Ar CO2

-3.0

CO2 + DAT

-1.5

-1.0

-0.5

0.0

Potential vs Ag/AgCl (V) Figure 2: Cyclic voltammetry (50 mV/s) of the roughened Ag electrode in 1 M KOH + Ca(OH)2 (sat’d): purged with Ar without DAT (black), purged with CO2 without DAT (red), and purged with CO2 with 10 mM DAT (blue). The inset shows the region of CO32- adsorption/desorption.

The voltammetry on the cathodic sweep is featureless until a potential of -1.1 V is reached and H2 evolution occurs. In a solution containing 1 M KOH + Ca(OH)2 (sat’d) continuously purged with CO2, the voltammetry exhibits increased cathodic current (Fig. 2 red line) which has previously been associated with CO and H2 production (as well as limited production of HCOO).7, 36 The CV obtained with this solution also exhibits small peaks between -0.4 V and -0.5 V as 8 ACS Paragon Plus Environment

Page 9 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


shown in the inset to Fig. 2. Originally, these features were attributed in part to the adsorption/desorption of OH- from the Ag surface based on previous work using electrolytes between pH 11 and 14.37-41 However, their absence in solutions not containing CO2 suggests these features are more likely associated with CO32- on the Ag surface at this pH, an inference supported by the SERS (vide infra). In 0.1 M KHCO3 electrolyte, these features have been previously assigned to the adsorption/desorption of KHCO3.36 The SERS data also indicate that bicarbonate is unlikely to be species responsible for these peaks. In a solution containing 1 M KOH + Ca(OH)2 (sat’d) + 10 mM DAT continuously purged with CO2, the voltammetry also exhibits cathodic current starting at ca. -1.1 V. A comparison of the onset potentials with changing DAT concentrations using an unroughened Ag electrode in an electrolyte without added Ca(OH)2 shows that the initial HER onset becomes more negative with increasing DAT concentrations, decreasing from approximately -1.1 V without DAT to -1.3 V at a concentration of 0.1 M. This suggests that adsorbed DAT may be inhibiting the onset of H2 evolution. Nonetheless, at more negative potentials where CO2 reduction to CO occurs (ca. -1.4 V), there is usually greater current density seen in the voltammetry of the samples containing DAT than those without, a result consistent with our earlier report.30 Additionally, the CO32peaks at ca. -0.5 V are absent in the presence of DAT, presumably due to site-blocking by the ligand. Figure 3 shows the cyclic voltammetry when the same experiment is performed absent Ag, using a polished glassy carbon disk as the working electrode.


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Current Density (mA/cm2)


Page 10 of 39

0.5 0.0 -0.5 -1.0 Without DAT With DAT

-1.5 -2.0 -2.5

-1.5

-1.0

-0.5

0.0


Figure 3: Cyclic voltammetry (50 mV/s) of CO2 reduction and HER on glassy carbon in 1 M KOH + Ca(OH)2 (sat’d): purged with CO2 without DAT (black) and purged with CO2 with 10 mM DAT (red).

DAT addition is observed to have negligible impact on the onset potential of the cathodic current and the magnitude of the current density is significantly lower. This behavior is consistent with previous observations,30 and shows that CO2 reduction is far more active with the combination of Ag with DAT than it is without the metal. 3.2 SERS 3.2.1 CO32In order to understand the origin of the increased CO production in the presence of DAT, we performed SERS on the roughened Ag electrode. Figure 4 displays the SER spectra obtained in the 1 M KOH + Ca(OH)2 (sat’d) solution that was continuously purged with CO2 during the cathodic sweep of staircase voltammetry.


Page 11 of 39

1

2 3 4

Intensity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


-0.05 V

-0.85 V

600

800

1000

1200

1400

1600

Raman Shift (cm-1)

Figure 4: SER spectra during cathodic polarization in 1 M KOH + Ca(OH)2 (sat’d) solution continuously purged with CO2 showing the potential dependence of CO32- desorption.

The applied potential was decreased from -0.05 V to -1.85 V in 50 mV steps. Figure 4 displays a subset of this data from -0.05 V to -0.85 V in 100 mV increments. Bands observed at 1014 cm-1 and 1355 cm-1 are attributable to the bicarbonate in the electrolyte.42 Peaks at 685cm-1 (1) and 1049 cm-1 (3) correspond to vibrational modes of CO32- in aqueous solution,42 and peaks at 722 cm-1 (2) and 1500 cm-1 (4) are assigned to chemisorbed CO32- on the Ag surface.43-44 At lower frequency, shown later in Figure 6A, two bands at approximately 284 cm-1 (5) and 392 cm-1 (6) also represent CO32- adsorption to the Ag electrode.44 All these bands decrease in intensity with applied negative potential, and are not observed below ca. -0.55 V. The potential dependence of these peak intensities is in agreement with the voltammetry presented in Fig. 2, indicating that the CO32- is desorbed at approximately -0.5 V. 3.2.2 CO Figure 5 shows the in situ SER spectra collected during the electroreduction of CO2 in the spectral region where CO appears. Staircase voltammetry was performed from -0.05 V to -1.85



V, with spectral acquisition for 5 s at each 50 mV step. For brevity, each figure displays the spectra at 100 mV intervals. 8

8

A

B

-0.05 V

Intensity

7

Intensity

7

-0.05 V

-1.85 V

-1.85 V 1800

1900

2000

2100

2200

1800

-1

C

8

1900

2000

2100

2200

-1

Raman Shift (cm )

Raman Shift (cm )

9 10

D

8

10 9

7

11

-0.05 V

Intensity

-0.05 V Intensity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 39

-1.85 V 1800

1900

2000

2100 -1

Raman Shift (cm )

2200

-1.85 V 1800

1900

2000

2100

2200

-1

Raman Shift (cm )

Figure 5: In situ SER spectra during cathodic and anodic polarization without DAT (A, B) and with DAT (C, D).

The potential was cycled three times in the 1 M KOH + Ca(OH)2 (sat’d) solution saturated with CO2, and the data shown was collected during the third consecutive cycle. Fig. 5A shows the cathodic polarization. Here, adsorbed CO from previous cycles is observed near 1880 cm-1 (7) and 1945 cm-1 (8) at -0.05 V. As the potential is decreased, the corresponding decrease in intensity of these bands suggests CO desorption, associated with CO production onset at -1.4 V.


Page 13 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


These peaks also exhibit significant Stark vibrational shifts to lower frequencies with negative potential, indicating a weakening of the intramolecular C-O bond. As the potential is made more negative, the intensity of peak 8 increases again between approximately -1.4 V and -1.85 V indicating generation of new CO from the reduction of CO2. Assignment of the spectral features seen here is made with reference to prior work. There have been numerous vibrational spectroscopy studies of CO adsorption on Ag surfaces in both UHV and aqueous environments, in which the CO vibration frequency at specific adsorption sites varies considerably with the experimental conditions.22-25, 45-49 The chemisorption of molecular CO on a metal surface is generally described by the Blyholder model.50 In this model, electron density is donated from the CO 5σ molecular orbital to the Ag metal surface. In return, backbonding occurs from the Ag to the CO 2π* anti-bonding orbital, which decreases the strength of the CO bond. Thus, the observed frequencies for adsorbed CO are lower than that of free, gaseous CO (2145 cm-1).51 The greatest degree of backbonding occurs at CO adsorption sites with the highest coordination to the Ag.52 Thus, more-highly coordinated CO will have weaker internal CO bonding and lower vibrational frequency. Changing the applied potential changes the degree of backbonding. This sensitivity of the frequency of the CO vibration to potential is known as a vibrational Stark shift. Typically, a more negative potential increases the amount of backbonding, further weakening the internal CO bond and causing the CO vibration to shift to lower wavenumbers.53 Taken together, these considerations show that the most highly-coordinated molecules will exhibit both lower vibration frequency and greater Stark slopes. Such behavior has been



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 39

observed with SERS for CO adsorption to atop and bridge sites on Au@Pt nanoparticle electrodes over a wide potential range, with Stark shifts ranging from 32 cm-1/V to 137 cm-1/V.54 Given this information, the peaks shown in Fig. 5A may be assigned to CO adsorbed at a 3-fold hollow site (7) and a bridge site (8) on the Ag surface. As the potential sweep is reversed, peaks associated with CO produced on the surface are retained, as shown in Fig. 5B. At -0.05 V, the most prominent signals are assigned to the return of peaks 7 and 8, found at 1870 cm-1 and 1948 cm-1, respectively. As the potential is swept in the positive direction, the quantity of CO generated decreases dramatically. However, the increasing Ag surface charge likely causes a higher proportion of CO molecules that had been liberated to the solution to re-adsorb. This increasing adsorption at higher potential yields the increase in intensity seen in peaks 7 and 8. The data presented here show both similarities and differences with prior work examining CO2 reduction on Ag surfaces. In SNIFTER spectra of CO2 reduction on Ag in 0.05 M Na2SO4, a peak at 1954 cm-1 found at approximately -0.7 V vs SHE and was attributed to atop-adsorbed CO.16 In 3.5 M KCl, SERS of the Ag surface revealed peak growth at approximately 1940 cm-1 at -1.0 V vs SHE, which shifted to 1896 cm-1 at -1.6 V vs SHE. During the anodic potential sweep, the CO peak disappeared between -0.7 and -0.5 V while large peak growth was observed at 249 cm-1 from Ag-Cl at -0.6 V vs SHE. In another study, SERS spectra in 0.1 M KHCO3 revealed two bands near 2130 cm-1 and 2180 cm-1, attributed to adsorbed CO and [Ag(CO)2]+, respectively. 14 Both peaks decreased in intensity between 0 and -0.3 V vs Ag/AgCl, beyond which they were no longer apparent. In addition, several potential-dependent bands were assigned to HCOO-. The band assignments reported here differ from these previous reports due to both the absence of Cl- in the electrolyte and the prevention of Ag oxidation by restricting the potential to those below 0 V vs Ag/AgCl.


Page 15 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Fig. 5C and 5D display the cathodic and anodic scan SER spectra obtained from solutions containing 1 M KOH + Ca(OH)2 (sat’d) + 10 mM DAT continuously purged with CO2. The spectra here are substantially more complicated than those found absent DAT. As before, a band associated with bridge-bound CO is observed at 1924 cm-1 (8), but this band occurs with both lower vibrational frequency and intensity than in the absence of DAT. However, the potential dependence of the intensity decrease and Stark shift of peak 8 in the solution containing DAT are similar to that found absent DAT. The lower energy peak 7 near 1880 cm-1 is no longer observable. Fig. 5C shows that there are two new CO adsorption peaks seen at 2049 cm-1 (9) and 2099 cm-1 (10) in the presence of DAT. Like peak 8, peak 9 decreases in intensity with cathodic potential. However, peak 10 first increases to a maximum near -0.75 V before ultimately decreasing as the potential becomes more negative. Peak 9 is at an energy associated previously with atop-bound CO on the Ag electrode surface.24, 26 Peak 10 is at an energy where the CO must be physisorbed or only weakly coordinated with the underlying electrode surface.22-23 This band not only possesses a higher CO stretching frequency than atop-adsorbed CO, but also lacks a significant Stark shift. The diminished intensity of this peak at potentials more positive than 0.75 V may reflect the displacement of this weakly-adsorbing species by solution anions such as HCO3- and CO32-. As the potential is swept to more negative values, the growth of peak 8 at 1820 cm-1 is associated with the onset of CO gas generation near -1.4 V. During the anodic scan (Fig. 5D) peaks 8, 9, and 10 increase in intensity with increasing potential, demonstrating the recovery of CO on the electrode surface. A new feature (11) is also observed in the presence of DAT during anodic polarization that grows at approximately -0.85 V (1962 cm-1), peaks near -0.65 V (1970 cm-1), and disappears at -0.35 V (1970 cm-1). This peak,



intermediate in energy between adsorption to a top site and a typical bridge site, could possibly be assigned to another type of 2-fold bridge adsorption. One such adsorption site could be the long-bridge of a (110) crystal face, but this would be expected at a much lower frequency based on calculations of CO adsorption to bcc metal surfaces (Fe(110), Mo(110), and W(110)).55 However, the surface structure of the roughened Ag, the presence of DAT, and solvent effects could all have a large influence on the available adsorption sites and their adsorption properties. At lower frequencies, the metal-CO stretching region also demonstrates the differences in CO adsorption from the addition of DAT. Figure 6 shows the SER spectra every 100 mV during cathodic polarization of the electrolyte solution saturated with CO2. A

B

5 6

-0.05 V

-0.05 V Intensity

Intensity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 39

250

300

12

13

350

400 -1

Raman Shift (cm )

-1.85 V 450

13 250

300

350

400

-1.85 V 450

-1

Raman Shift (cm )

Figure 6: In situ SER spectra during cathodic polarization without DAT (A) and with DAT (B).

Fig. 6A was obtained in the absence of DAT, whereas 10 mM DAT was used for the spectra shown in Fig. 6B. As discussed earlier, there are two potential-dependent bands (5 and 6) assignable to chemisorbed CO32- in Fig. 6A which are not seen when DAT is added. As the potential is made more negative, peak growth is observed at 353 cm-1 (12) and 406 cm-1 (13). Although these values are higher than expected for CO adsorption to Ag based on previous DFT


Page 17 of 39

calculations, they do fall within the general ranges observed experimentally for CO binding to other transition metal surfaces (330 cm-1 to 480 cm-1).26 Spectroscopic evidence of CO adsorption to transition metals has long established the inverse relationship between the level of coordination at surface sites and the frequency of CO vibration.56 Ag-CO stretching in hollow site adsorption is therefore expected at a lower Raman shift than at bridge site adsorption. Further, the intensity ratio between the higher frequency, higher intensity bridge peak and the lower frequency, lower intensity hollow peak observed in the C-O stretching region echoes the ratio of the peaks seen here in the Ag-CO stretching region. Therefore, we attribute peak 12 to the Ag-CO stretch of hollow site adsorption and band 13 to bridge adsorption. Fig. 6B shows that bridge adsorption is retained with the addition of DAT by the presence of band 13 at approximately 402 cm-1. There is no peak assignable to hollow site adsorption, which agrees with what was observed in the C-O stretching region during cathodic polarization (Fig. 5C). Similar to the frequency of adsorbed C-O stretching, the Ag-CO stretching mode also increases in frequency with decreasing surface coordination.26 Figure 7 depicts the potential dependence of the frequency of the CO bands at various adsorption sites.

2100 Raman Shift (cm-1)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


2050 2000

With DAT Without DAT

1950 1900 1850 1800 -2.0

-1.5

-1.0

-0.5

0.0




1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 39

Figure 7: Peak frequency as a function of potential for adsorbed CO during anodic polarization.

Peak frequency as a function of potential is plotted without DAT (black) and with DAT (red) for the anodic scans seen in Fig. 5B and 5D. As shown, the addition of DAT promotes CO adsorption in three new ways previously unobserved in the absence of the molecule. The Stark shift for each unique CO site is reported in Table 1. Table 1. Peak frequency ranges and Stark slopes of CO adsorbed at various Ag surface sites during anodic polarization from -1.85 to -0.05 V vs Ag/AgCl Peak

DAT Present

Adsorption Site

Frequency Range (cm-1)

Stark Slope (cm-1/V)

7

Yes

3-Fold Hollow

1815-1840

87

7

No

3-Fold Hollow

1818-1870

91

8

Yes

Bridge

1838-1921

84

8

No

Bridge

1858-1950

88

9

Yes

Atop

2044-2051

16

10

Yes

Physisorbed

2093-2100

4.3

11

Yes

Other Bridge

1960-1977

20

The most remarkable feature of the effect of DAT is to produce a CO stretching feature at high energy which does not exhibit a Stark shift. The lack of a Stark shift means that direct communication with the Ag surface must be inhibited. 3.2.3 DAT Figure 8A shows normal Raman obtained from a neat AgDAT complex synthesized from DAT and Ag2SO4 (black).


Page 19 of 39

A

B Neat AgDAT Ag + 10 mM DAT Neat DAT

Intensity

Intensity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


-0.05 V

-1.85 V 400

800

1200

1600

Raman Shift (cm-1)

2000

400

800

1200

1600

2000

Raman Shift (cm-1)

Figure 8: SER spectra of (A) the neat AgDAT complex (black), the silver electrode with 10 mM DAT dissolved in the electrolyte (red), and the the neat DAT ligand (blue), and (B) the Ag electrode immersed in the electrolyte solution with 10 mM DAT during cathodic potential application.

The complex exhibits peaks and shoulder bands between 1000 and 1700 cm-1 associated with a variety of vibrational modes for the ring and its amine groups according to Raman spectra and vibrational assignments for DAT reported previously for 1-H DAT57 and 4-H DAT58. To our knowledge, the SER spectra of the AgDAT complex has not been previously reported, but the Ag(DAT)NO3 complex has been characterized using IR.59 This study suggested that the complex exists as a polymer, with Ag coordinated to nitrogens on either side of the ring. The structure of the complex at our electrode surface has not been determined. Fig. 8A also shows that similar modes are seen in the SERS spectrum when the Ag electrode is first immersed in an electrolyte solution containing 10 mM DAT (red). The presence of these features shows that the AgDAT complex spontaneously forms at open circuit potential. Also shown in fig. 8A is the normal Raman spectrum of neat DAT (blue), which demonstrates the spectral differences between the 19 ACS Paragon Plus Environment


silver complex and the free ligand. Fig. 8B shows that peaks associated with the AgDAT complex diminish as the potential is swept to more negative values. The potential was changed in 50 mV steps, but the spectra are shown at 300 mV intervals for clarity. The peaks associated with the AgDAT complex re-appear once the electrode returns to the open circuit potential, but remain absent as the electrode is cycled over the negative potential region used here. The decreased intensity of peaks associated with the AgDAT complex during cathodic polarization indicates that the ligand is dissociating from the negatively-polarized surface. However, there are indications that part of the ligand is retained on the surface at potentials where CO adsorption is altered by the presence of DAT. Fig. 9 shows the potential dependence of three bands associated with the AgDAT complex at negative potentials in solutions absent CO2. Potential vs Ag/AgCl (V) -0.05 -0.15 -0.25 -0.35 -0.45 -0.55 -0.65 -0.75 -0.85 -0.95 -1.05 -1.15 -1.25

16 Intensity

15

Intensity

Intensity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 39

14 700

800

900 -1

Raman Shift (cm )

1150

1200

1250 -1

Raman Shift (cm )

1550

1600

1650

Raman Shift (cm-1)

Figure 9: SER spectra during cathodic polarization of the Ar-purged electrolyte containing 10 mM DAT.


Page 21 of 39

The figure shows the presence of bands 14, 15, and 16 at approximately 779 cm-1, 1212 cm-1, and 1606 cm-1. The first two bands may be assigned to the tertiary amine C-N-C stretching modes, and the 1606 cm-1 is attributable to the NH2 scissor mode.60 In contrast to the other AgDAT-associated bands which consistently decrease in intensity with cathodic potential, these peaks demonstrate growth before ultimately diminishing as gas production begins. The potentialdependent growth of these peaks therefore indicates that the ligand does interact with the electrode surface at the potentials just above the onset potential for HER. DAT remains bound to the surface at least this far during cathodic polarization with a possible change in surface orientation relative to the neat AgDAT complex. In the presence of CO2, these bands were obscured. However, Figure 10A shows the baseline-corrected difference spectra of the normalized CO2 reduction spectra during cathodic polarization with DAT minus the spectrum without DAT.

17

18 B

Intensity

A

Intensity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


-0.5 -1.0 -1.5 180 200 220 240 -1

Raman Shift (cm )


-0.5 -1.0 -1.5 160

200

240

280 -1

Raman Shift (cm )


Figure 10: SER difference spectra of cathodic reduction with DAT minus that without DAT in the Ag-N stretching region with CO2 (A) and with Ar (B).

In this region, a peak is observed at approximately 204 cm-1 (17) assigned to Ag-N stretching on the basis of similarity with other complexes.61-63 The intensity of this mode decreases continually 21 ACS Paragon Plus Environment


until approximately -1 V. Fig. 10B shows that in Ar, a Ag-N stretching mode is also observed with potential-dependence very similar to those of peaks 14, 15, and 16. However, the peak frequency of this band is shifted to approximately 249 cm-1 (18). Figure 11 shows the potential dependence of the normalized, baseline-corrected peak intensity for the stretching frequency of the non-coordinated CO peak at 2099 cm-1 that appears when DAT is added for CO2 reduction.

1.0 Normalized Intensity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 39

0.8 0.6

10: 14: 15: 16: 17:

CO DAT DAT DAT Ag-N

0.4 0.2 0.0 -2.0

-1.5

-1.0

-0.5

0.0


Figure 11: Normalized, baseline-corrected intensity as a function of cathodic potential during CO2 reduction for the non-coordinated CO stretching band (10) and four bands related to surface-bound DAT (14, 15, 16, 17).

Overlaid in this spectrum is the potential dependence of the four bands (14, 15, 16, 17) associated with DAT retention on the Ag surface. The figure shows that modes associated with DAT (or possibly its degradation product) are present on the electrode surface at the same potentials where the non-coordinated CO band is present. The presence of these modes with intensity mirroring that for non-coordinated CO indicates that a surface ligand may be present and preventing stronger Ag-CO adsorption. The diminished (or absent) Ag-CO coordination is


Page 23 of 39

suggested as the origin of the enhanced CO production in the presence of the DAT ligand. Fig. 11 also shows that beyond the potential of the onset of HER at ca. -1.1 V, neither the noncoordinated CO nor the ligand are very strongly bound to the Ag. 3.2.4 Alkanes In addition to CO adsorption, significant Raman bands were observed in the C-H stretching region. Figure 12 shows the potential dependent Raman spectra collected during staircase voltammetry in the 1 M KOH + Ca(OH)2 (sat’d) electrolyte.

22 19

20 21

B

22 19

-1.5 -1.0 -0.5

2600


3000 Potential vs 2600 Ag/AgCl (V)

Intensity

20 21

A

Intensity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


-1.5 -1.0 -0.5


3000 Potential vs

Ag/AgCl (V)

Figure 12: SER spectra in the C-H stretching region during cathodic (A) and anodic (B) polarization in the CO2-saturated 1 M KOH + Ca(OH)2 (sat’d) electrolyte.

The spectra shows that beginning at about -0.5 V as the potential is swept in the cathodic direction, CH2 symmetric (near 2850 cm-1) (21) and asymmetric (near 2920 cm-1) (22) stretching modes are visible.64 The intensity of these peaks increase considerably until a maximum is reached near -1.2 V, beyond which they decrease to a minimum at the most negative potential, 1.85 V. As the potential decreases, these two peaks experience a slight lower-energy shift to 2847 cm-1 and 2913 cm-1, respectively. In addition, two more peaks develop near 2715 cm-1 (19)



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 39

and 2820 cm-1 (20), which do not change frequency as the potential is modified. The CH2 scissor mode at 1440 cm-1 is also visible (not shown), and the potential dependence of its intensity mirrors that seen in Fig. 12. The reverse pattern is observed during the anodic scan, although the peak intensity is generally lower and the methylene stretching peaks 21 and 22 are retained all the way back to the highest potential (2853 cm-1 and 2924 cm-1 at -0.05 V). These potential-dependent changes in intensity and the appearance of new features at 2715 cm-1 and 2820 cm-1 are indicative of the presence of long chains of methylene groups growing on the surface. In work examining surfactant molecules on Ag, the presence of peaks 19 and 20 has been associated with the methylene chains becoming more parallel to the Ag plane and increasingly rigid with negative potential, until they are ultimately desorbed.65-66 A peak near 2720 cm-1 is observable in solid surfactant samples, indicating that the applied cathodic potential produces a rigid, crystalline hydrocarbon layer at the surface. The band 20 is attributable to C-H stretching in either the gauche or trans conformations. The increase in intensity among all C-H stretching modes likely results from the reorientation of methylene chains towards the surface, aligning the C-H stretching axis with the enhanced electromagnetic field perpendicular to the surface.66 We wanted to clarify the source of hydrogen in the hydrocarbon growing on the surface. Figure 13 shows the results during cathodic polarization from a similar experiment performed in deuterated electrolyte (1 M KOD in D2O).


B

20

21

19

22

-1.5 -1.0 -0.5 2000

2100

2200 -1


Raman Shift (cm )

C

23

-1.5 -1.0 -0.5 2700

2800

3000 Potential vs

2900

Ag/AgCl (V)

-1

Raman Shift (cm )

24

D 20 21 22

-1.5 -1.0 -0.5

-1.5 -1.0 -0.5 2000

2100

2200

Raman Shift (cm-1)

Intensity

Intensity

A

Intensity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



Intensity

Page 25 of 39

2700

2800

2900

3000 Potential vs

Raman Shift (cm-1)

Ag/AgCl (V)

Figure 13: SER spectra in the C-D (A,C) and C-H (B,D) stretching regions during cathodic polarization in CO2-saturated electrolyte. Figures A and B were collected during the first cycle, while C and D represent the third cycle.

Fig. 13A showing the C-D stretching region, obtained during the first potential cycle, displays no demonstrable intensity, while Fig. 13B, also obtained during the first cycle, shows that C-H stretching is initially observed. The origin of the hydrogen is likely residual water retained on the Ag surface during electrode preparation. However, Fig. 13C and 13D show the spectral results from same system after two complete voltammetric cycles. With successive cycling, the C-H peaks decrease in intensity



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 39

whereas peaks 23 and 24 in the C-D stretching region have grown. These stretches at approximately 2069 cm-1 (23) and 2106 cm-1 (24) correspond to the symmetric and asymmetric stretching modes of paraffinic CD2, respectively.67 These data show conclusively that hydrocarbons are being generated on the Ag electrode surface, with the ultimate source of hydrogen coming from water. The hydrocarbon peaks are not the result of adventitious contamination, but rather form as a consequence of CO2 reduction. Prior work using vibrational spectroscopy has observed similar hydrocarbon formation variously described as parrafinic hydrocarbons, branched hydrocarbons, aldehydes, and formate, as well as graphitic surface carbon.14, 17-18, 31, 68 This work reports the first observation of a methylene peak shift upon deuteration. Finally, in work using on-line mass spectrometry, Jaramillo confirmed the formation of hydrocarbons during CO2 reduction on Cu.69 The cycle dependence of the intensities in C-D and C-H regions shows that continued cycling leads to desorption of the hydrocarbon. We wanted to discriminate more carefully between hydrocarbons on the Ag surface that might occur as a result of contamination and those that grow as a result of CO2 reduction. Figure 14 shows that in 1 M KOH + Ca(OH)2 (sat’d) purged with Ar, these same C-H stretching modes were also observed. Fig. 14A shows that the intensity of these bands decreases considerably after several successive cycles.


Page 27 of 39

A

B

1st cycle 2nd cycle 3rd cycle

5 min. CO2 sparging 45 min. CO 2 sparging

Intensity

Intensity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


2600

2800

3000

3200

2600

-1

2800

3000

3200

-1

Raman Shift (cm )

Raman Shift (cm )

Figure 14: SER spectra at -1.1 V of hydrocarbon formation from CO2 reduction during cathodic sweeps: in an Ar-purged electrolyte for three cycles (A) and in the same electrolyte after several more cycles in Ar and after the introduction of CO2 (B).

The Ar-purged system was cycled several additional times and the peaks continue to show diminished intensity cycle after cycle. Then, CO2 was introduced to the solution, and staircase voltammetry was once again collected after 5 min. and after 45 min of CO2 sparging. Figure 12B shows that re-growth of the C-H peaks was observed after a sufficient quantity of CO2 has been added. The carbon source in the initial sample is therefore most likely atmospheric CO2 that adsorbs to the Ag during experimental setup. The influence of DAT on the C-H stretching peaks is shown in Figure 15.



1.0

A

B

No additive With DAT

0.8 0.6

No additive With DAT Intensity

Normalized Intensity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 39

0.4 0.2 0.0 -2.0

-1.5

-1.0

-0.5


2600

2800

3000

3200

-1

Raman Shift (cm )

Figure 15: Differences in potential dependence (A) and spectral intensity (B) of C-H stretching during cathodic polarization between trials without and with DAT.

Figure 15A compares the normalized, baseline-corrected intensities of the C-H symmetric stretching modes during cathodic polarization in solutions with and without DAT. As shown, the onset of growth and decline in the band intensity with DAT occurs at potentials slightly more negative than that without. Figure 15B shows SERS obtained from the C-H stretching band region at a potential of -1.25 V during the first cathodic scan with and without DAT. The figure shows that the peaks have much greater intensity when the ligand molecule is absent. These data suggest that the presence of DAT on the surface inhibits the formation and/or adsorption of the hydrocarbon species, which might again favor CO production over other reaction pathways.

CONCLUSIONS This work shows that adding DAT to a Ag electrode gives rise to a number of significant changes in CO2 reduction activity. The voltammetry shows that CO2 reduction occurs with slightly less overpotential in the presence of DAT. While in situ SER spectra obtained during the electroreduction of CO2 on a Ag electrode reveal several sites for adsorption of the generated CO 28 ACS Paragon Plus Environment

Page 29 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


product, the addition of DAT is associated with observed CO adsorption to sites with less surface coordination, even after dissociation of most of the initial AgDAT complex. In particular, SER spectra show that CO is much more weakly bound on DAT-exposed Ag relative to bare Ag. This weakly-bound CO does not exhibit a Stark shift, again emphasizing the lack of significant interaction with the underlying Ag surface. The observation of weakly-bound or physisorbed CO implies that the product is able to leave easily, increasing the efficiency of the Ag catalyst. The presence of DAT-related peaks at the potentials where physisorbed CO is observed suggests that it is adsorbed DAT or a ligand decomposition product that is disrupting the Ag surface and preventing more stable CO coordination. The explicit mechanism by which DAT suppresses stable CO coordination, and/or encourages weaker CO coordination is not yet known. The Ag requirement for significant catalytic activity suggests strongly that Ag sites are involved in the reactivity. The DAT may blocks sites where more stable CO coordination can occur, leading to higher reduction rates because the product CO removal is enhanced. Addition of DAT has consequences for the formation of hydrocarbon species also observed on bare and DAT-decorated Ag. Absent DAT, hydrocarbon byproducts are observed to form, the origin of which is CO2 bubbling into the solution and protons from water decomposition. With DAT, hydrocarbon formation is suppressed and occurs only at ca. 0.2 V more negative potentials relative to the bare surface. This result again suggests that DAT is blocking a reaction pathway existent on the bare Ag surface.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 39

AUTHOR INFORMATION Corresponding Author *E-mail: [email protected]. Tel.: +1 (217) 333-8329. Fax: +1 (217) 244-3186. Notes The authors declare no competing financial interest.

ACKNOWLEDGMENTS K.G.S. gratefully acknowledges a University Fellowship award from the Graduate College of the University of Illinois at Urbana-Champaign. This work was supported by the NSF (CHE13-09731). We also acknowledge the International Institute for Carbon Neutral Energy Research (WPI-I2CNER), which is sponsored by the World Premier International Research Center Initiative (WPI) of MEXT in Japan.

REFERENCES (1)

Appel, A. M.; Bercaw, J. E.; Bocarsly, A. B.; Dobbek, H.; Dubois, D. L.; Dupuis, M.;

Ferry, J. G.; Fujita, E.; Hille, R.; Kenis, P. J.; et al. Frontiers, Opportunities, and Challenges in Biochemical and Chemical Catalysis of CO2 Fixation. Chem. Rev. 2013, 113, 6621-6658 (2)

Benson, E. E.; Kubiak, C. P.; Sathrum, A. J.; Smieja, J. M. Electrocatalytic and

Homogeneous Approaches to Conversion of CO2 to Liquid Fuels. Chem. Soc. Rev. 2009, 38, 8999 (3)

Costentin, C.; Robert, M.; Saveant, J. M. Catalysis of the Electrochemical Reduction of

Carbon Dioxide. Chem. Soc. Rev. 2013, 42, 2423-2436


Page 31 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(4)

Whipple, D. T.; Kenis, P. J. A. Prospects of CO2 Utilization Via Direct Heterogeneous

Electrochemical Reduction. J. Phys. Chem. Lett. 2010, 1, 3451-3458 (5)

Hori, Y. Electrochemical CO2 Reduction on Metal Electrodes. In Modern Aspects of

Electrochemistry, Vayenas, C.; White, R.; Gamboa-Aldeco, M., Eds. Springer: New York, 2008; Vol. 42, pp 89-189. (6)

Hansen, H. A.; Varley, J. B.; Peterson, A. A.; Norskov, J. K. Understanding Trends in the

Electrocatalytic Activity of Metals and Enzymes for CO2 Reduction to CO. J. Phys. Chem. Lett.

2013, 4, 388-392 (7)

Hori, Y.; Wakebe, H.; Tsukamoto, T.; Koga, O. Electrocatalytic Process of CO

Selectivity in Electrochemical Reduction of CO2 at Metal Electrodes in Aqueous Media.

Electrochim. Acta 1994, 39, 1833-1839 (8)

Surdhar, P. S.; Mezyk, S. P.; Armstrong, D. A. Reduction Potential of the Carboxyl

Radical Anion in Aqueous Solutions. J. Phys. Chem. 1989, 93, 3360-3363 (9)

Schwarz, H. A.; Dodson, R. W. Reduction Potentials of CO2- and the Alcohol Radicals. J.

Phys. Chem. 1989, 93, 409-414 (10)

Spichiger-Ulmann, M.; Augustynski, J. Electrochemical Reduction of Bicarbonate Ions at

a Bright Palladium Cathode. J. Chem. Soc., Faraday Trans. 1 1985, 81, 713-716 (11)

Podlovchenko, B. I.; Kolyadko, E. A.; Lu, S. G. Electroreduction of Carbon Dioxide on

Palladium Electrodes at Potentials Higher Than the Reversible Hydrogen Potential. J.

Electroanal. Chem. 1994, 373, 185-187 (12)

Kortlever, R.; Tan, K. H.; Kwon, Y.; Koper, M. T. M. Electrochemical Carbon Dioxide

and Bicarbonate Reduction on Copper in Weakly Alkaline Media. J. Solid State Electrochem.

2013, 17, 1843-1849



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(13)

Page 32 of 39

Stalder, C. J.; Chao, S.; Wrighton, M. S. Electrochemical Reduction of Aqueous

Bicarbonate to Formate with High Current Efficiency Near the Thermodynamic Potential at Chemically Derivatized Electrodes. J. Am. Chem. Soc. 1984, 106, 3673-3675 (14)

Ichinohe, Y.; Wadayama, T.; Hatta, A. Electrochemical Reduction of CO2 on Silver as

Probed by Surface-Enhanced Raman Scattering. J. Raman Spectrosc. 1995, 26, 335-340 (15)

Kaneco, S.; Iiba, K.; Ohta, K.; Mizuno, T.; Saji, A. Electrochemical Reduction of CO2 at

an Ag Electrode in KOH-Methanol at Low Temperature. Electrochim. Acta 1998, 44, 573-578 (16)

Oda, I.; Ogasawara, H.; Ito, M. Carbon Monoxide Adsorption on Copper and Silver

Electrodes During Carbon Dioxide Electroreduction Studied by Infrared Reflection Absorption Spectroscopy and Surface-Enhanced Raman Spectroscopy. Langmuir 1996, 12, 1094-1097 (17)

Yano, H.; Shirai, F.; Nakayama, M.; Ogura, K. Electrochemical Reduction of CO2 at

Three-Phase (Gas∣Liquid∣Solid) and Two-Phase (Liquid∣Solid) Interfaces on Ag Electrodes. J.


Mcquillan, A. J.; Hendra, P. J.; Fleischmann, M. Raman Spectroscopic Investigation of

Silver Electrodes. J. Electroanal. Chem. 1975, 65, 933-944 (19)

Azuma, M.; Hashimoto, K.; Hiramoto, M.; Watanabe, M.; Sakata, T. Carbon Dioxide

Reduction at Low Temperature on Various Metal Electrodes. J. Electroanal. Chem. 1989, 260, 441-445 (20)

Thorson, M. R.; Siil, K. I.; Kenis, P. J. A. Effect of Cations on the Electrochemical

Conversion of CO2 to CO. J. Electrochem. Soc. 2013, 160, F69-F74 (21)

Hori, Y.; Ito, H.; Okano, K.; Nagasu, K.; Sato, S. Silver-Coated Ion Exchange Membrane

Electrode Applied to Electrochemical Reduction of Carbon Dioxide. Electrochim. Acta 2003, 48, 2651-2657


Page 33 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(22)

Mahoney, M. R.; Howard, M. W.; Cooney, R. P. Raman Spectra of Carbon Monoxide

Adsorbed on Silver Electrodes. J. Electroanal. Chem. 1984, 161, 163-167 (23)

Orozco, G.; Perez, M. C.; Rincon, A.; Gutierrez, C. Adsorption and Electrooxidation of

Carbon Monoxide on Silver. Langmuir 1998, 14, 6297-6306 (24)

Ikezawa, Y.; Saito, H.; Matsubayashi, H.; Toda, G. Comparative-Study of CO Adsorbed

on Pt, Pd, Au and Ag Electrodes in Neutral Solution by IR Reflection Absorption Spectroscopy.

J. Electroanal. Chem. 1988, 252, 395-402 (25)

Furukawa, H.; Takahashi, M.; Ito, M. SERS and FT-IR Studies of CO Adsorbed on

Underpotential Deposited Ag/Pt Electrodes. J. Electroanal. Chem. 1990, 280, 415-423 (26)

Marek, G.; Andreas, E.; Jürgen, H. CO Adsorption on Close-Packed Transition and

Noble Metal Surfaces: Trends from ab-Initio Calculations. J. Phys.: Condens. Matter 2004, 16, 1141-1164 (27)

Su, H. Y.; Yang, M. M.; Bao, X. H.; Li, W. X. The Effect of Water on the CO Oxidation

on Ag(111) and Au(111) Surfaces: A First-Principle Study. J. Phys. Chem. C 2008, 112, 1730317310 (28)

Rosen, B. A.; Salehi-Khojin, A.; Thorson, M. R.; Zhu, W.; Whipple, D. T.; Kenis, P. J.;

Masel, R. I. Ionic Liquid-Mediated Selective Conversion of CO2 to CO at Low Overpotentials.

Science 2011, 334, 643-644 (29)

Zheng, J. W.; Lu, T. H.; Cotton, T. M.; Chumanov, G. Photoelectrochemical Reduction

of CO2 Mediated with Methylviologen at Roughened Silver Electrodes. J. Electroanal. Chem.

2002, 518, 6-12



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(30)

Page 34 of 39

Tornow, C. E.; Thorson, M. R.; Ma, S.; Gewirth, A. A.; Kenis, P. J. Nitrogen-Based

Catalysts for the Electrochemical Reduction of CO2 to CO. J. Am. Chem. Soc. 2012, 134, 1952019523 (31)

Mahoney, M. R.; Howard, M. W.; Cooney, R. P. Carbon Dioxide Conversion to

Hydrocarbons at Silver Electrode Surfaces. Chem. Phys. Lett. 1980, 71, 59-63 (32)

Taylor, C. E.; Garvey, S. D.; Pemberton, J. E. Carbon Contamination at Silver Surfaces:

Surface Preparation Procedures Evaluated by Raman Spectroscopy and X-Ray Photoelectron Spectroscopy. Anal. Chem. 1996, 68, 2401-2408 (33)

Tsang, J. C.; Demuth, J. E.; Sanda, P. N.; Kirtley, J. R. Enhanced Raman Scattering from

Carbon Layers on Silver. Chem. Phys. Lett. 1980, 76, 54-57 (34)

Ambrosio, R. C.; Gewirth, A. A. Characterization of Water Structure on Silver Electrode

Surfaces by SERS with Two-Dimensional Correlation Spectroscopy. Anal. Chem. 2010, 82, 1305-1310 (35)

Schultz, Z. D.; Feng, Z. V.; Biggin, M. E.; Gewirth, A. A. Vibrational Spectroscopic and

Mass Spectrometric Studies of the Interaction of Bis(3-Sulfopropyl)-Disulfide with Cu Surfaces.

J. Electrochem. Soc. 2006, 153, C97-C107 (36)

Hoshi, N.; Kato, M.; Hori, Y. Electrochemical Reduction of CO2 on Single Crystal

Electrodes of Silver Ag(111), Ag(100) and Ag(110). J. Electroanal. Chem. 1997, 440, 283-286 (37)

Jovic, B. M.; Jovic, V. D.; Stafford, G. R. Cyclic Voltammetry on Ag(111) and Ag(100)

Faces in Sodium Hydroxide Solutions. Electrochem. Commun. 1999, 1, 247-251 (38)

Savinova, E. R.; Wasle, S.; Doblhofer, K. Structure and Activity Relations in the

Hydrogen Peroxide Reduction at Silver Electrodes in Alkaline NaF/NaOH Electrolytes.

Electrochim. Acta 1998, 44, 1341-1348


Page 35 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(39)

Savinova, E. R.; Kraft, P.; Pettinger, B.; Doblhofer, K. In Situ Raman Spectroscopy

Studies of the Interface Between Silver(111) Electrodes and Alkaline NaF Electrolytes. J.


Horswell, S. L.; Pinheiro, A. L.; Savinova, E. R.; Danckwerts, M.; Pettinger, B.; Zei, M.

S.; Ertl, G. A Comparative Study of Hydroxide Adsorption on the (111), (110), and (100) Faces of Silver with Cyclic Voltammetry, Ex Situ Electron Diffraction, and In Situ Second Harmonic Generation. Langmuir 2004, 20, 10970-10981 (41)

Iwasaki, N.; Sasaki, Y.; Nishina, Y. Ag Electrode Reaction in NaOH Solution Studied by

In-Situ Raman Spectroscopy. Surf. Sci. 1988, 198, 524-540 (42)

Davis, A. R.; Oliver, B. G. A Vibrational-Spectroscopic Study of the Species Present in

the CO2-H2O System. J. Solution Chem. 1972, 1, 329-339 (43)

Kai, S.; Chaozhi, W.; Guangzhi, X. Surface Enhanced Raman Spectra of Carbonate,

Hydrocarbonate, and Substituted Acetic Acids on Silver Hydrosols. Spectrochim. Acta, Part A

1989, 45, 1029-1032 (44)

Sasaki, Y.; Iwasaki, N.; Nishina, Y. Adsorption of CO3 on a Silver Electrode in Na2CO3

Solution Studied by SERS. Surf. Sci. 1988, 198, 541-554 (45)

Wood, T. H.; Klein, M. V. Raman Scattering from Carbon Monoxide Adsorbed on

Evaporated Silver Films. J. Vac. Sci. Tech. 1979, 16, 459-461 (46)

Müslehiddinoglu, J. CO Adsorption on Supported and Promoted Ag Epoxidation

Catalysts. J. Catal. 2003, 213, 305-320 (47)

Dilella, D. P.; Gohin, A.; Lipson, R. H.; Mcbreen, P.; Moskovits, M. Enhanced Raman

Spectroscopy of CO Adsorbed on Vapor-Deposited Silver. J. Chem. Phys. 1980, 73, 4282-4295



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(48)

Page 36 of 39

Qu, Z. P.; Zhou, S. T.; Wu, W. C.; Li, C.; Bao, X. H. CO Adsorption and Correlation

Between CO Surface Coverage and Activity/Selectivity of Preferential CO Oxidation Over Supported Ag Catalyst: An In Situ FTIR Study. Catal. Lett. 2005, 101, 21-26 (49)

Dumas, P.; Tobin, R. G.; Richards, P. L. Study of Adsorption States and Interactions of

CO on Evaporated Noble Metal Surfaces by Infrared Absorption Spectroscopy. Surf. Sci. 1986,

171, 555-578 (50)

Blyholder, G. Molecular Orbital View of Chemisorbed Carbon Monoxide. J. Phys.

Chem. 1964, 68, 2772-2777 (51)

Fenner, W. R.; Hyatt, H. A.; Kellam, J. M.; Porto, S. P. S. Raman Cross-Section of Some

Simple Gases. J. Opt. Soc. Am. 1973, 63, 73-77 (52)

Mason, S. E.; Grinberg, I.; Rappe, A. M. First-Principles Extrapolation Method for

Accurate CO Adsorption Energies on Metal Surfaces. Phys. Rev. B: Condens. Matter Mater.

Phys. 2004, 69, 161401 (53)

Lambert, D. K. Vibrational Stark Effect of Adsorbates at Electrochemical Interfaces.

Electrochim. Acta 1996, 41, 623-630 (54)

Zhang, P.; Cai, J.; Chen, Y.-X.; Tang, Z.-Q.; Chen, D.; Yang, J.; Wu, D.-Y.; Ren, B.;

Tian, Z.-Q. Potential-Dependent Chemisorption of Carbon Monoxide at a Gold Core−Platinum Shell Nanoparticle Electrode: A Combined Study by Electrochemical in Situ Surface-Enhanced Raman Spectroscopy and Density Functional Theory. J. Phys. Chem. C 2010, 114, 403-411 (55)

Abild-Pedersen, F.; Andersson, M. P. CO Adsorption Energies on Metals with Correction

for High Coordination Adsorption Sites – a Density Functional Study. Surf. Sci. 2007, 601, 1747-1753


Page 37 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(56)

Ishi, S.; Ohno, Y.; Viswanathan, B. An Overview on the Electronic and Vibrational

Properties of Adsorbed CO. Surf. Sci. 1985, 161, 349-372 (57)

Guennoun, L.; El jastimi, J.; Guedira, F.; Marakchi, K.; Kabbaj, O. K.; El Hajji, A.;

Zaydoun, S. Molecular Geometry and Vibrational Studies of 3,5-Diamino-1,2,4-Triazole Using Quantum Chemical Calculations and FT-IR and FT-Raman Spectroscopies. Spectrochim. Acta A

2011, 78, 347-353 (58)

Kumar, V. K.; Keresztury, G.; Sundius, T.; Xavier, R. J. Hydrogen Bonding and

Molecular Vibrations of 3,5-Diamino-1,2,4-Triazole. Spectrochim. Acta A 2005, 61, 261-267 (59)

Gabryszewski, M.; Wieczorek, B. Silver(I) Complexes with 1,2,4-Triazole, 1-Ethyl-

1,2,4-Triazole, 3-Amino-1,2,4-Triazole, 4-Amino-1,2,4-Triazole and 3,5-Diamino-1,2,4Triazole. Pol. J. Chem. 1998, 72, 2352-2355 (60)

Lin-Vien, D.; Colthup, N. B.; Fateley, W. G.; Grasselli, J. G. Chapter 10 - Compounds

Containing —NH2, —NHR, and —NR2 Groups. In The Handbook of Infrared and Raman

Characteristic Frequencies of Organic Molecules; Academic Press: San Diego, 1991; pp 155178. (61)

Pergolese, B.; Muniz-Miranda, M.; Bigotto, A. Study of the Adsorption of 1,2,3-Triazole

on Silver and Gold Colloidal Nanoparticles by Means of Surface Enhanced Raman Scattering. J.

Phys. Chem. B 2004, 108, 5698-5702 (62)

Wrzosek, B.; Cukras, J.; Bukowska, J. Adsorption of 1,2,4-Triazole on a Silver

Electrode: Surface-Enhanced Raman Spectroscopy and Density Functional Theory Studies. J.

Raman Spectrosc. 2012, 43, 1010-1017 (63)

Sarkar, J.; Chowdhury, J.; Talapatra, G. B. Adsorption of 4-Methyl-4H-1,2,4-Triazole-3-

Thiol Molecules on Silver Nanocolloids: FT-IR, Raman, and Surface-Enhanced Raman



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 38 of 39

Scattering Study Aided by Density Functional Theory. J. Phys. Chem. C 2007, 111, 1004910061 (64)

Lin-Vien, D.; Colthup, N. B.; Fateley, W. G.; Grasselli, J. G. Chapter 2 - Alkanes. In The

Handbook of Infrared and Raman Characteristic Frequencies of Organic Molecules; Academic Press: San Diego, 1991; pp 9-28. (65)

Sun, S. C.; Birke, R. L.; Lombardi, J. R. Surface-Enhanced Raman Spectroscopy of

Surfactants on Silver Electrodes. J. Phys. Chem. 1990, 94, 2005-2010 (66)

Foucault, R.; Birke, R. L.; Lombardi, J. R. SERS of Surfactants in Monolayer and

Multibilayer Forms on an Electrified Ag Surface. Langmuir 2003, 19, 8818-8827 (67)

Brown, R. G. Raman Spectra of Polyethylenes. J. Chem. Phys. 1963, 38, 221-225

(68)

Cooney, R. P.; Mahoney, M. R.; Howard, M. W. Intense Raman Spectra of Surface

Carbon and Hydrocarbons on Silver Electrodes. Chem. Phys. Lett. 1980, 76, 448-452 (69)

Kuhl, K. P.; Cave, E. R.; Abram, D. N.; Jaramillo, T. F. New Insights Into the

Electrochemical Reduction of Carbon Dioxide on Metallic Copper Surfaces. Energy Environ.

Sci. 2012, 5, 7050-7059


Page 39 of 39

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


TABLE OF CONTENTS IMAGE


In Situ Surface-Enhanced Raman Spectroscopy of the Electrochemical

Recommend Documents