Topics in Catalysis

, Volume 61, Issue 9–11, pp 1035–1042 | Cite as

Comparative Studies of Ethanol and Ethylene Glycol Oxidation on Platinum Electrocatalysts

  • Shalaka Dewan
  • David Raciti
  • Yifan Liu
  • David H. Gracias
  • Chao Wang
Original Paper


Ethanol represents a promising energy source for powering fuel cells. The development of direct ethanol fuel cells is however challenged by both the sluggish kinetics of the ethanol oxidation reaction and the poor selectivity toward complete oxidation. In this work, we combine spectroelectrochemical studies of extended surfaces using sum frequency generation (SFG) and product-resolved electrocatalytic measurements under potentiostatic conditions to investigate the electro-oxidation of alcohols. By comparing the electro-oxidation of ethanol and ethylene glycol, we illustrate the different catalytic mechanisms of C–C bond cleavage and identify the role of β carbon in hindering the complete oxidation of ethanol toward CO2. Our findings provide new insights into the development of efficient electrocatalysts for multi-carbon alcohol oxidation.


Ethanol oxidation Platinum electrocatalysts Fuel cells Sum frequency generation spectroscopy SFG 

1 Introduction

Ethanol represents a promising alternative to fossil fuels [1]. The large-scale production of ethanol from renewable sources (e.g., sugarcane) has been shown to be economically viable [2, 3]. Apart from being a promising substitute for gasoline in internal combustion engines, ethanol can also be fed into fuel cells. The complete oxidation of ethanol to carbon dioxide (CO2) delivers 12 electrons per molecule at an equilibrium potential of 0.084 V versus the reversible hydrogen electrode (RHE; the same potential scale is used in the paper), giving rise to an energy density of 6.7 kWh/L (vs. 9.5 kWh/L for gasoline) [4]. Ethanol is a liquid at room temperature and atmospheric pressure, and also considered nontoxic, making it more suitable than hydrogen and other liquid fuel alternatives (e.g., methanol and formic acid) for storage and transportation. Due to these advantages, direct ethanol fuel cells (DEFCs) have attracted significant interest for power applications [4, 5].

Despite the great potential, the practical implementation of DEFCs is challenged by the lack of efficient electrocatalysts for the ethanol oxidation reaction (EOR). Platinum (Pt) has been the most studied material for the EOR, but the sluggish kinetics still leads to high overpotentials (e.g., > 0.5 V) and poor selectivity toward CO2 [5, 6, 7, 8]. Previous studies indicate that the EOR on Pt takes parallel pathways toward partial and complete oxidation, forming acetic acid (or acetaldehyde) and CO2, respectively [8, 9, 10]. The partial oxidation pathway dominates the EOR on Pt in acid electrolytes at room temperature, with the Faradaic efficiency (FE) toward CO2 typically less than 10% [11, 12], although higher CO2 selectivity is achievable by raising the reaction temperature and/or electrolyte pH [13, 14, 15]. On the other hand, spectroscopic studies of the EOR electrocatalysis in situ show that Pt is capable of cleaving the C–C bond at low potentials (e.g., < 0.3 V) [11, 16, 17, 18, 19]. The inability of Pt to catalyze the complete oxidation of ethanol is attributed to high overpotentials (e.g., > 0.5 V) needed for the oxidation of strongly binding C1 intermediates such as CH x (x = 1, 2 or 3) and CO [11, 16, 17, 18, 19]. Although the formation of the C1 species is observed at low potentials, the energy barrier for C–C bond cleavage is much higher when the surface is covered by adsorbed hydroxide (*OH) at high potentials, making the partial oxidation pathway kinetically more favorable [20]. From this point of view, the introduction of a second metal (e.g., Sn and Ru) that is more oxophilic than Pt could promote the adsorption of *OH and thereby reduce the potential required for oxidation of the C1 species [21, 22, 23], enhancing the catalytic activity for the complete oxidation of ethanol. This view, however, is not well supported by experimental studies reported in the literature [24, 25, 26, 27, 28, 29, 30, 31, 32, 33], suggesting that more mechanistic studies are needed to understand the EOR electrocatalysis on Pt.

Here, we report comparative spectroelectrochemical studies of ethanol and ethylene glycol oxidation (EGO) on Pt electrocatalysts. Studies were first performed on polycrystalline platinum (Pt-poly) electrodes, where sum frequency generation (SFG) spectroscopy was used to track the potential-dependent adsorption of CO (*CO, where * denotes a surface site for adsorption) under reaction conditions, and then on high-surface-area platinum/carbon (Pt/C) catalysts where gas chromatography–mass spectrometry (GC–MS) was used to analyze the product distributions. Ethylene glycol (HO–CH2–CH2–OH) differs from ethanol (CH3–CH2–OH) by an additional –OH group, which has previously been suggested to facilitate the C–C bond cleavage [34, 35, 36]. By comparing the electro-oxidation of these two C2 alcohols, we sought to elucidate the role of the β carbon (in the –CH3 group of ethanol) in EOR electrocatalysis, including how it affects the interplay between the C–C bond cleavage and the C1 species oxidation, and ultimately the catalytic activity and selectivity.

SFG is a second order, non-linear optical process that only occurs in media where the inversion symmetry is broken, making it particularly suitable for studying reaction intermediates adsorbing at solid/liquid interfaces [37, 38]. SFG studies of ethanol oxidation have previously been reported [19, 39], but such studies have not been carried out on polyol alcohols such as ethylene glycol. Differential electrochemical mass spectrometer (DEMS) [12, 40, 41, 42] and on-line electrochemical mass spectroscopy (OLEMS) [43, 44, 45] studies have been reported for the EOR; in these studies, vacuum is used to draw the reaction products from the electrolyte for product analysis. These two methods are advantageous for capturing trace amounts of chemical species during transient and potentiodynamic processes, but can be limited by low collection efficiency. In this work, the catalytic selectivities of Pt/C are evaluated under potentiostatic, atmospheric-pressure and continuous-flow conditions, which has enabled us to better mimic steady-state conditions under fuel cell operation. The combination of SFG spectroscopic and electrocatalytic studies thus allows us to develop a more comprehensive understanding of the mechanisms involved in the electro-oxidation of C2 alcohols.

2 Experimental

2.1 Materials and Characterization

Both alcohols (99.8% for ethanol, 99.8% for ethylene glycol) and all chemicals used for the synthesis of Pt nanoparticles were purchased from Sigma-Aldrich. Optima-grade perchloric acid (70%) was purchased from J. T. Baker. Electrolyte solutions were prepared using deionized water (~ 18.2 MΩ). The Pt metal used for preparation of the disk electrode was obtained from Kurt J. Lesker. Inert gases (Ar and He) were ordered from Airgas.

The Pt nanoparticles were synthesized in organic solutions by modifying a previously reported method [46]. Typically, Pt(acac)2 (79 mg) and oleylamine (15 mL, 98%) were added to a round bottom flask. The obtained slurry was stirred under Ar and heated to 70 °C where it was held for 10 min to form a clear solution. Borane tert-butylamine (34 mg, dissolved in 2 mL of oleylamine) was injected into this solution, and the solution color turned from translucent yellow to an opaque black immediately. The reaction proceeded for another 30 min at 70 °C and the solution was then cooled to room temperature. The product was collected by adding methanol, centrifugating (10,000 rpm for 10 min) and then re-dispersing in hexane for future use. The morphology and size of the prepared nanoparticles were characterized using transmission electron microscopy (TEM), and the crystal structures were characterized using X-ray diffraction (XRD). TEM imaging was performed on a 120 kV, FEI Tecnai 12 TWIN microscope. XRD patterns were collected on a PANalytical X’Pert3 Powder X-ray diffractometer equipped with a Cu Kα source (λ = 0.15406).

The polycrystalline Pt (Pt-poly) electrode was prepared by machining a platinum slug into a disk (5 mm in diameter, 4 mm in height). Prior to the electrochemical studies, the Pt-poly electrode was polished to a mirror finish and/or subjected to radio frequency (RF) annealing in forming gas (5% H2/N2). Blank cyclic voltammograms were scanned before each spectroscopic or electrocatalytic measurement to verify that the surface was clean.

2.2 Electrochemistry

Electrochemical studies were performed in 0.1 M HClO4 using a three-electrode cell and an Autolab potentiostat (Metrohm PGSTAT302N). A Pt wire and a Hg/HgSO4 electrode were used as the counter and reference electrode, respectively. Measurements were collected with 85% iR correction. The product-resolved electrocatalytic measurements on Pt/C were performed using a customized gas-tight electrochemical cell (see the Supporting Information for more details of the setup and the protocols). The Pt/C catalyst was prepared by depositing the 3 nm synthesized Pt nanoparticles on carbon black (~ 900 m2/g, obtained from Tanaka) with a weight ratio of 2:1. The mixture was annealed at 185 °C in air overnight to remove the organic surfactant. The final ratio of Pt in the catalysts was determined by inductively coupled plasma-mass spectrometry (ICP-MS) measurements. The Pt/C catalyst was dispersed in deionized water by sonication at a concentration of 1 mg/mL. Nafion (5 wt% from Sigma-Aldrich) was added (1 vol%) into the solution to make a uniform catalyst ink. Aliquots of the obtained mixture (~ 1 mL) were deposited on a graphite plate, which was loaded into the electrochemical cell and used as the working electrode. During the measurements, the electrode was held at a given potential with a He flow (10 mL/min) continuously passing through the working electrode compartment. A gas chromatograph equipped with a mass spectrometer (GC–MS, Shimadzu QP2010SE) was used to periodically analyze the gas-phase products. A small amount of the electrolyte (~ 1 mL) was collected after the measurement at each potential, and the dissolved liquid products were analyzed using a 300 MHz nuclear magnetic resonance (NMR) spectrometer (Bruker Advance).

2.3 Spectroelectrochemical Studies

The spectroelectrochemical studies were performed in a customized cell made of Teflon, according to the previously reported design for similar experiments [47]. The cell was covered with a 3 mm-thick calcium fluoride (CaF2) window with the distance to the electrode surface fixed at 12 µm using a Teflon spacer (Harrick Corporation). A Pt wire and Ag/AgCl (saturated with KCl), were used as the counter and reference electrodes, respectively. SFG spectra in the range of 1850–2120 cm−1 were collected using the PPP polarization combination under potentiostatic conditions in the region of 0.05–0.9 V. For this polarization combination, the orientation of the electric field vector is parallel to the plane of incidence for all the sum frequency, visible and infrared beams. More details about the SFG principle and spectrum fitting are provided in the Supporting Information.

3 Results and Discussion

Figure 1a is a typical cyclic voltammogram (CV) recorded on the Pt-poly electrode used in this study. The CV exhibits peaks associated with the underpotential deposition of hydrogen (Hupd) at E < 0.4 V and hydroxyl adsorption at 0.8–0.9 V, with a capacitive double-layer region located in between. The pronounced peak at ~ 0.8 V in the cathodic scan can be assigned to the reduction of surface oxides and desorption of *OH [48]. The CO stripping curve recorded on the Pt-poly electrode exhibits a sharp peak at 0.73 V together with a relatively wide peak at 0.68–0.71 V (Fig. 1b), which can be assigned to the oxidation of *CO on various types of Pt sites, including low-index facets and stepped surfaces [49, 50, 51]. A much less-pronounced shoulder peak was also observed at 0.5–0.6 V in the CO stripping, which could be associated with the pre-oxidation on defect sites [52, 53].

Fig. 1

(a) Typical CV recorded on the Pt-poly electrode. (b) Polarization curves for the electro-oxidation of ethanol and ethylene glycol (electrolyte contains 0.1 M of HClO4 and 0.1 M of alcohol). The CO stripping curve is also shown for comparison. The scan rate is 50 mV/s for all

With either alcohol (0.1 M of ethanol or ethylene glycol) in the electrolyte, a wide, pronounced oxidation peak appears in the potential region of 0.5–1.0 V, whose onset potential is in line with the pre-oxidation potential as observed in the CO stripping (Fig. 1b). The maximum oxidation current was reached at 0.79 V for ethanol and 0.75 V for ethylene glycol, with the values being rather consistent at 1.47 and 1.38 mA/cm2, respectively. For both the EOR and EGO, non-zero anodic currents were observed at potentials below 0.5 V, together with a relatively small peak consistently displaying at ~ 0.30 V (Fig. 1b). These features were not observed in the CO stripping curve and their assignments will be discussed later following the spectroscopic studies.

To probe and track the adsorbing intermediates during the electrocatalytic processes, SFG spectra were recorded on the Pt-poly electrode in the wavelength region of 1850–2120 cm−1 at different anodic potentials (Fig. 2a, b). For comparison, the spectroscopic studies were also carried out for CO stripping, where the electrode potential was positioned at 0.05 V to allow for the adsorption of CO before the spectrum collection (Fig. 2c). The main feature exhibited in the SFG spectra is a peak 2065–2070 cm−1, which can be assigned to linearly bound CO (COL) on Pt [47, 54]. The COL peak appears at potentials as low as 0.05 V for both ethanol and ethylene glycol oxidation. The spectra for the EOR also show a broad peak of low intensity at ~ 1975 cm− 1, which can be assigned to bridge-bonded CO (COB) on Pt [17, 19]. This feature is much less visible in the spectra recorded for the CO stripping and EGO. The formation of *CO at potentials as low as 0.05 V indicates the presence of active sites on the Pt-poly electrode for the cleavage of the C–C bond, which is consistent with previously reported SFG studies for the EOR [17, 55]. It has been shown by density functional theory (DFT) calculations that the C–C cleavage has quite large energy barriers on low-index Pt facets, including (111), (100) and (211), varying from 0.5 to 1.0 eV [56]. It has been postulated that compared to such low-index facets, undercoordinated sites associated with high-index steps [55, 57] and defects [11] (adatoms, ad-islands, vacancies, etc.) are more likely to account for the C–C cleavage at low potentials.

Fig. 2

SFG spectra collected in the PPP mode on the Pt-poly electrode at various potentials over the course of, (a) EOR, (b) EGO and, (c) CO stripping

In Fig. 3a, we plot the variation of the SFG intensity associated with the COL peak in PPP as a function of the electrode potential. For the EOR, the peak intensity rises gradually as the potential is raised from 0.05 to 0.2 V, followed by a sharp increase at 0.3 V. We observed that the oxidation of ethylene glycol behaved quite differently in this potential region, with a sharp increase from 0.05 to 0.1 V and then flattening from 0.1 to 0.3 V. In comparison, the signal for CO stripping is consistent in the potential region of 0.05–0.3 V. The faster formation of *CO in the EGO than in the EOR indicates the role of the second –OH group in activating the C–C bond cleavage step in the EGO [35]. It is suggested that the complete oxidation of ethanol involves bidentate-binding intermediates such as *CH2CO* [20, 55], versus 2-hydroxyacetyl (*CO–CH2OH) with single binding sites for the EGO [34, 35]. Although direct comparison of the energy barriers for these two pathways has not been reported yet, it has been suggested that the C–C cleavage for ethanol may require very specific atomic configurations, such as those at step sites [55, 57], whereas the single-site adsorption and dissociation of 2-hydroxyacetyl could be more facile and take place on a variety of surface sites.

Fig. 3

Plot of the variation of SFG intensity corresponding to the COL peak in the PPP configuration on the Pt-poly electrode as a function of electrode potential. See the Supporting Information for details of the curve fitting for the SFG spectra

It should be pointed out that the SFG spectra were collected under potentiostatic conditions, while the polarization curves shown in Fig. 1b were recorded using potentiodynamic methods. The anodic currents observed at 0.05–0.3 V in the latter case can originate from the desorption of *H, dehydrogenation of adsorbed alcohols, and/or oxidation of the C1 species produced from the C–C bond cleavage in both the EOR and EGO. The anodic peak exhibited at ~ 0.3 V in the polarization curves (Fig. 1b) indicates the termination of those processes at this potential, which also coincides with reaching a maximum *CO signal in the SFG spectra.

At potentials higher than 0.3 V, we observed that the peak intensity starts to drop for both the EOR and the EGO, but the EOR exhibits a noticeable lag as compared to the EGO (Fig. 3a). From 0.3 to 0.5 V, the SFG intensity decreases by ~ 40% for the EGO, versus < 10% for the EOR. In both cases, the SFG peak nearly disappears by 0.7 V. We notice that the dissociation of ethanol produces both *CO and *CH x (x = 1–3) [17], whereas ethylene glycol only gives rise to oxygenated species (*CO, *CHO, *COH, etc.) [34]. Considering that the behavior of CO stripping resembles that for the EGO, we postulate that the delayed *CO signal observed in the EOR is due to the sluggish oxidation of *CH x generated from the β carbon. The sluggish conversion of *CH x –*CO was initially proposed by Lai et al. [17] based on a surface-enhanced Raman spectroscopy (SERS) study, and later confirmed by Kutz et al. [19] in their investigation of isotopically labelled ethanol using SFG spectroscopy. In particular, the latter report shows that the *CO produced from the β carbon persists up to ~ 0.6 V, versus 0.4–0.5 V for the *CO from the α carbon. Our observation of *CO signal retention in the potential region of 0.3–0.5 V is in line with these reports.

The dissimilar dissociation mechanisms of C–C bond cleavage for ethanol and ethylene glycol, as well as the slow conversion of *CH x –*CO caused by the presence of β carbon in ethanol, are expected to affect the electrocatalytic performances of Pt for the electro-oxidation of these two alcohols. To examine this effect, we further performed electrocatalytic studies on high-surface-area Pt/C catalysts and compared the product distributions between the two reactions.

The Pt/C catalyst employed in this study was prepared by depositing 3 nm Pt nanoparticles derived from organic solution synthesis on carbon black (Figs. 4a and S1; see the "Experimental" section). The organic surfactant was removed by applying a mild thermal treatment in air [58]. The CV recorded on the Pt/C catalyst is similar to that for the Pt-poly electrode, indicating the exposure of a clean Pt surface (Fig. 4a). The CO stripping curve recorded on the Pt/C catalyst exhibits a strong peak at 0.85 V, together with a weaker peak at 0.75 V. The shoulder peak at 0.5–0.6 V observed on Pt-poly is nearly invisible on Pt/C. The cause for the higher CO oxidation potential on Pt/C as compared to Pt-poly is unclear, but similar differences have previously been seen between extended surfaces and nanoscale catalysts [59]. In line with the CO stripping, the Pt/C catalyst also exhibits a positive shift of the primary oxidation peak in the electro-oxidation of the two alcohols, with the maximum specific currents (per mass of Pt) reached at 0.8–0.9 V (Fig. 4b). Besides the primary oxidation peak, EGO shows an additional shoulder peak at ~ 0.6 V, which is nearly invisible in the EOR. The EGO also exhibited larger anodic currents than the EOR in the potential region of E < 0.3 V, but the small anodic peak at ~ 0.3 V observed on the Pt-poly electrode is not seen on the Pt/C catalyst for both the EOR and EGO, possibly due to the lack of large ensembles of ordered terrace sites on the nanoscale catalyst.

Fig. 4

(a) CV recorded for the 3 nm Pt/C catalyst with the inset showing a typical TEM image of the catalyst. (b) Polarization curves recorded on Pt/C for the electro-oxidation of ethanol and ethylene glycol (electrolyte contains 0.1 M of HClO4 and 0.1 M of alcohol) with the CO stripping curve also shown for comparison. The scan rate is 50 mV/s for all scans

Figure 5 summarizes the results of electrocatalytic studies on Pt/C employing potentiostatic measurements at E > 0.5 V while the product CO2 was monitored by GC (see the Supporting Information for more details of the experimental methods). As the potential increases, the electrode current increases similarly for both the EOR and EGO, until ~ 850 mV where the former is about 33% higher (Fig. 5a). However, the EOR gives much lower Faradaic efficiencies toward CO2 \(\left( {{\text{F}}{{\text{E}}_{{\text{C}}{{\text{O}}_{\text{2}}}}}} \right)\), the product for complete oxidation, than the EGO (Fig. 5b). At ~ 0.55 V, the EOR has ca. 5% \({\text{F}}{{\text{E}}_{{\text{C}}{{\text{O}}_{\text{2}}}}}\), compared to 21% for the EGO. For both the EOR and the EGO, the \({\text{F}}{{\text{E}}_{{\text{C}}{{\text{O}}_{\text{2}}}}}\) drops as the potential increases, down to 7 and 0.2% at 0.85 V, respectively. As a result of both smaller currents and lower FEs, the EOR produces much smaller specific currents for the complete oxidation than the EGO throughout the potential region (Fig. 5c). The maximum partial current for CO2 production \(\left( {{i_{{\text{C}}{{\text{O}}_{\text{2}}}}}} \right)\) of the EOR is recorded to be merely 0.03 A/gPt (specific current normalized by the mass of platinum) at 0.65 V. In contrast, \({i_{{\text{C}}{{\text{O}}_{\text{2}}}}}\) reaches up to 0.6 A/gPt at ~ 0.75 V for the EGO. Other products besides CO2 are identified to be acetaldehyde and acetic acid for the EOR, and glycolic acid and glycolaldehyde for the EGO (see the Supporting Information; Fig. S7 and Table S1).

Fig. 5

Electrocatalytic studies of ethanol and ethylene glycol oxidation: (a) Total electrode currents, (b) Faradaic efficiencies (\({\text{F}}{{\text{E}}_{{\text{C}}{{\text{O}}_{\text{2}}}}}\)) and (c) partial currents (\({i_{{\text{C}}{{\text{O}}_{\text{2}}}}}\)) toward the complete-oxidation product, CO2

The rather low \({\text{F}}{{\text{E}}_{{\text{C}}{{\text{O}}_{\text{2}}}}}\) for the electro-oxidation of C2 alcohols indicates more favorable kinetics of partial oxidation than for complete oxidation [20, 55, 57]. Our observed trend of \({i_{{\text{C}}{{\text{O}}_{\text{2}}}}}\) depending on the electrode potential is also consistent with previously reported DEMS studies of ethanol [12, 42] and ethylene glycol [40, 60] oxidation. For the electro-oxidation of both alcohols, the onset of CO2 product is found to be consistently at ~ 0.5 V. This observation seems to contradict the persistent *CO signal at 0.5 V as observed in the SFG spectra (Fig. 3). Considering the slow conversion of *CH x –*CO, we suggest that the CO2 generated around 0.5 V in the EOR is a result of the oxidation of *CO produced from the α carbon during the C–C cleavage, whereas the oxidation of *CH x (originated from the β carbon) to *CO and CO2 requires higher potentials (e.g., ~ 0.6 V). We also notice that \({i_{{\text{C}}{{\text{O}}_{\text{2}}}}}\) is non-zero and rather consistent at 0.02–0.03 A/gPt in the potential region of 0.5–0.9 V, indicating that the complete oxidation of ethanol still takes place at the high potentials. Presumably, it is the same type of active sites, such as steps or defects, that contribute to both the C–C cleavage at E < 0.3 V and complete oxidation of ethanol at E > 0.5 V. The much lower \({\text{F}}{{\text{E}}_{{\text{C}}{{\text{O}}_{\text{2}}}}}\) and \({i_{{\text{C}}{{\text{O}}_{\text{2}}}}}\) of the EOR than for the EGO is likely a result of the different C–C bond cleavage mechanisms. As discussed above, the dissociation pathway for ethanol may be restricted to specific surface sites (steps and/or defects), whereas the complete oxidation of ethylene glycol is not likely to be subjected to such a restriction. Thereby the different numbers of active sites give rise to lower CO2 selectivity of the EOR than for the EGO.

During the electrocatalytic studies of EOR on Pt/C, we noticed the detection of a mass signal with m/z = 15 throughout the potential region of 0.5–0.9 V (Fig. S5), which can be assigned to the CH3+ fragment generated from methane. Our observation is consistent with previous reports on DEMS studies of the EOR and suggests the likelihood of cathodic desorption of *CH x produced from the C–C bond cleavage [12, 42]. Interestingly, the detection of methane is accompanied with the production of CO2, confirming the inference above that the *CO derived from the α carbon in ethanol gets oxidized fast, whereas the process for oxidizing the *CH x derived from the β carbon is rather slow. It also reiterates the role of the β carbon in causing the low selectivity of EOR toward CO2 as compared to the EGO, since the oxygenated C1 intermediates involved in the latter case are less likely subject to the desorption loss.

4 Conclusions

We conducted comparative studies of ethanol and ethylene glycol electro-oxidation on Pt. SFG spectroscopy was used to track the formation of the surface-bound key reaction intermediate, *CO, on polycrystalline electrodes and elucidate the dependence of C–C bond cleavage on potential. Electrocatalytic studies were further performed on high-surface-area Pt/C catalysts under potentiostatic conditions to resolve the distinct catalytic activities and selectivities of the two catalytic processes. Our findings emphasize the role of the β carbon in causing the poor selection of ethanol oxidation toward CO2 due to hindered C–C bond cleavage and slow conversion of *CH x to *CO, in contrast to the second hydroxyl group which facilitates the dissociation and complete oxidation of ethylene glycol. Nevertheless, active sites for the EOR are confirmed to be present on Pt that account for the formation of C1 adsorbates at potentials as low as < 0.3 V and the production of CO2 at potentials higher than ~ 0.5 V. Our findings suggest that to improve the performance of EOR electrocatalysts, one would need to increase the exposure of such active sites for dissociation, and meanwhile tailor the surface properties to accelerate the conversion of *CH x to *CO and oxidation of *CO to CO2.



This work was supported by the Young Investigator Award of Army Research Office (W911 NF-15-1-0123) and the Discovery Award of the Johns Hopkins University.

Supplementary material

11244_2018_930_MOESM1_ESM.docx (857 kb)
Supplementary material 1 (DOCX 857 KB)


  1. 1.
    Goldemberg J (2007) Science 315:808–810CrossRefPubMedGoogle Scholar
  2. 2.
    Farrell AE, Plevin RJ, Turner BT, Jones AD, O’Hare M, Kammen DM (2006) Science 311:506–508CrossRefPubMedGoogle Scholar
  3. 3.
    Demirbas A (2007) Prog Energy Combust 33:1–18CrossRefGoogle Scholar
  4. 4.
    Badwal SPS, Giddey S, Kulkarni A, Goel J, Basu S (2015) Appl Energy 145:80–103CrossRefGoogle Scholar
  5. 5.
    Lamy C, Lima A, LeRhun V, Delime F, Coutanceau C, Leger JM (2002) J Power Sources 105:283–296CrossRefGoogle Scholar
  6. 6.
    Antolini E (2007) J Power Sources 170:1–12CrossRefGoogle Scholar
  7. 7.
    Wang Q, Sun GQ, Jiang LH, Xin Q, Sun SG, Jiang YX, Chen SP, Jusys Z, Behm RJ (2007) Phys Chem Chem Phys 9:2686–2696CrossRefPubMedGoogle Scholar
  8. 8.
    Braunchweig B, Hibbitts D, Neurock M, Wieckowski A (2013) Catal Today 202:197–209CrossRefGoogle Scholar
  9. 9.
    Lai SCS, Koper MTM (2008) Faraday Discuss 140:399–416CrossRefPubMedGoogle Scholar
  10. 10.
    Colmati F, Tremiliosi-Filho G, Gonzalez ER, Berna A, Herrero E, Feliu JM (2008) Faraday Discuss 140:379–397CrossRefPubMedGoogle Scholar
  11. 11.
    Leung LWH, Chang SC, Weaver MJ (1989) J Electroanal Chem 266:317–336CrossRefGoogle Scholar
  12. 12.
    Wang H, Jusys Z, Behm RJ (2004) J Phys Chem B 108:19413–19424CrossRefGoogle Scholar
  13. 13.
    Rao V, Hariyanto, Cremers C, Stimming U (2007) Fuel Cells 7:417–423CrossRefGoogle Scholar
  14. 14.
    Lai SCS, Kleijn SEF, Ozturk FTZ, Vellinga VCV, Koning J, Rodriguez P, Koper MTM (2010) Catal Today 154:92–104CrossRefGoogle Scholar
  15. 15.
    Sun S, Heinen M, Jusys Z, Behm RJ (2012) J Power Sources 204:1–13CrossRefGoogle Scholar
  16. 16.
    Shao MH, Adzic RR (2005) Electrochim Acta 50:2415–2422CrossRefGoogle Scholar
  17. 17.
    Lai SCS, Kleyn SEF, Rosca V, Koper MTM (2008) J Phys Chem C 112:19080–19087CrossRefGoogle Scholar
  18. 18.
    Colmati F, Tremiliosi G, Gonzalez ER, Berna A, Herrero E, Feliu JM (2009) Phys Chem Chem Phys 11:9114–9123CrossRefPubMedGoogle Scholar
  19. 19.
    Kutz RB, Braunschweig B, Mukherjee P, Behrens RL, Dlott DD, Wieckowski A (2011) J Catal 278:181–188CrossRefGoogle Scholar
  20. 20.
    Kavanagh R, Cao XM, Lin WF, Hardacre C, Hu P (2012) Angew Chem Int Ed 51:1572–1575CrossRefGoogle Scholar
  21. 21.
    Gasteiger HA, Markovic N, Ross PN, Cairns EJ (1993) J Phys Chem 97:12020–12029CrossRefGoogle Scholar
  22. 22.
    Kabbabi A, Faure R, Durand R, Beden B, Hahn F, Leger JM, Lamy C (1998) J Electroanal Chem 444:41–53CrossRefGoogle Scholar
  23. 23.
    Kakati N, Maiti J, Lee SH, Jee SH, Viswanathan B, Yoon YS (2014) Chem Rev 114:12397–12429CrossRefPubMedGoogle Scholar
  24. 24.
    de Souza JPI, Queiroz SL, Bergamaski K, Gonzalez ER, Nart FC (2002) J Phys Chem B 106:9825–9830CrossRefGoogle Scholar
  25. 25.
    Neto AO, Giz MJ, Perez J, Ticianelli EA, Gonzalez ER (2002) J Electrochem Soc 149:A272–A279Google Scholar
  26. 26.
    Zhou WJ, Li WZ, Song SQ, Zhou ZH, Jiang LH, Sun GQ, Xin Q, Poulianitis K, Kontou S, Tsiakaras P (2004) J Power Sources 131:217–223CrossRefGoogle Scholar
  27. 27.
    Colmenares L, Wang H, Jusys Z, Jiang L, Yan S, Sun GQ, Behm RJ (2006) Electrochim Acta 52:221–233CrossRefGoogle Scholar
  28. 28.
    Sen S, Sen F, Gokagac G (2011) Phys Chem Chem Phys 13:6784–6792CrossRefPubMedGoogle Scholar
  29. 29.
    Zhu MY, Sun GQ, Xin Q (2009) Electrochim Acta 54:1511–1518CrossRefGoogle Scholar
  30. 30.
    Kim JH, Choi SM, Nam SH, Seo MH, Choi SH, Kim WB (2008) Appl Catal B Environ 82:89–102CrossRefGoogle Scholar
  31. 31.
    Camara GA, de Lima RB, Iwasita T (2004) Electrochem Commun 6:812–815CrossRefGoogle Scholar
  32. 32.
    Rousseau S, Coutanceau C, Lamy C, Leger JM (2006) J Power Sources 158:18–24CrossRefGoogle Scholar
  33. 33.
    Jin JM, Sheng T, Lin X, Kavanagh R, Hamer P, Hu PJ, Hardacre C, Martinez-Bonastre A, Sharman J, Thompsett D, Lin WF (2014) Phys Chem Chem Phys 16:9432–9440CrossRefPubMedGoogle Scholar
  34. 34.
    Salciccioli M, Yu WT, Barteau MA, Chen JGG, Vlachos DG (2011) J Am Chem Soc 133:7996–8004CrossRefPubMedGoogle Scholar
  35. 35.
    Schnaidt J, Heinen M, Jusys Z, Behm RJ (2012) J Phys Chem C 116:2872–2883CrossRefGoogle Scholar
  36. 36.
    Aran-Ais RM, Herrero E, Feliu JM (2014) Electrochem Commun 45:40–43CrossRefGoogle Scholar
  37. 37.
    Shen YR (1996) Proc Natl Acad Sci USA 93:12104–12111CrossRefPubMedGoogle Scholar
  38. 38.
    Baldelli S, Gewirth AA (2006) Adv Electrochem Sci Eng 9:163–198Google Scholar
  39. 39.
    Gomes JF, Bergamaski K, Pinto MFS, Miranda PB (2013) J Catal 302:67–82CrossRefGoogle Scholar
  40. 40.
    Wang H, Jusys Z, Behm RJ (2006) J Electroanal Chem 595:23–36CrossRefGoogle Scholar
  41. 41.
    Delpeuch AB, Asset T, Chatenet M, Cremers C (2014) J Electrochem Soc 161:F918–F924CrossRefGoogle Scholar
  42. 42.
    Buso-Rogero C, Brimaud S, Solla-Gullon J, Vidal-Iglesias FJ, Herrero E, Behm RJ, Feliu JM (2016) J Electroanal Chem 763:116–124CrossRefGoogle Scholar
  43. 43.
    Housmans THM, Wonders AH, Koper MTM (2006) J Phys Chem B 110:10021–10031CrossRefPubMedGoogle Scholar
  44. 44.
    Wonders AH, Housmans THM, Rosca V, Koper MTM (2006) J Appl Electrochem 36:1215–1221CrossRefGoogle Scholar
  45. 45.
    Rizo R, Lazaro MJ, Pastor E, Koper MTM (2016) ChemElectroChem 3:2196–2201CrossRefGoogle Scholar
  46. 46.
    Peng S, Lee YM, Wang C, Yin HF, Dai S, Sun SH (2008) Nano Res 1:229–234CrossRefGoogle Scholar
  47. 47.
    Baldelli S, Markovic N, Ross P, Shen Y-R, Somorjai G (2000) Sum-frequency generation of CO on (111) and polycrystalline platinum electrode surfaces: evidence for SFG invisible surface CO. J Phys Chem 103:8920–8925Google Scholar
  48. 48.
    Markovic NM, Ross PN (2002) Surf Sci Rep 45:121–229CrossRefGoogle Scholar
  49. 49.
    Lopez-Cudero A, Cuesta A, Gutierrez C (2005) J Electroanal Chem 579:1–12CrossRefGoogle Scholar
  50. 50.
    Cuesta A, Couto A, Rincon A, Perez MC, Lopez-Cudero A, Gutierrez C (2006) J Electroanal Chem 586:184–195CrossRefGoogle Scholar
  51. 51.
    Lopez-Cudero A, Cuesta A, Gutierrez C (2006) J Electroanal Chem 586:204–216CrossRefGoogle Scholar
  52. 52.
    Strmcnik DS, Tripkovic DV, van der Vliet D, Chang KC, Komanicky V, You H, Karapetrov G, Greeley J, Stamenkovic VR, Markovic NM (2008) J Am Chem Soc 130:15332–15339CrossRefPubMedGoogle Scholar
  53. 53.
    Farias MJS, Camara GA, Feliu JM (2015) J Phys Chem C 119:20272–20282CrossRefGoogle Scholar
  54. 54.
    Dederichs F, Friedrich KA, Daum W (2000) J Phys Chem B 104:6626–6632CrossRefGoogle Scholar
  55. 55.
    Ferre-Vilaplana A, Buso-Rogero C, Feliu JM, Herrero E (2016) J Phys Chem C 120:11590–11597CrossRefGoogle Scholar
  56. 56.
    Wang H-F, Liu Z-P (2008) J Am Chem Soc 130:10996–11004CrossRefPubMedGoogle Scholar
  57. 57.
    Lai SCS, Koper MTM (2010) J Phys Chem Lett 1:1122–1125CrossRefGoogle Scholar
  58. 58.
    Li DG, Wang C, Tripkovic D, Sun SH, Markovic NM, Stamenkovic VR (2012) ACS Catal 2:1358–1362CrossRefGoogle Scholar
  59. 59.
    van der Vliet DF, Wang C, Li DG, Paulikas AP, Greeley J, Rankin RB, Strmcnik D, Tripkovic D, Markovic NM, Stamenkovic VR (2012) Angew Chem Int Ed 51:3139–3142CrossRefGoogle Scholar
  60. 60.
    Wang H, Jusys Z, Behm RJ (2009) Electrochim Acta 54:6484–6498CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC, part of Springer Nature 2018

Authors and Affiliations

  1. 1.Department of Chemical and Biomolecular EngineeringJohns Hopkins UniversityBaltimoreUSA

Personalised recommendations