Interaction of C1 Molecules with a Pt Electrode at Open Circuit

Publication Date (Web): March 10, 2014. Copyright © 2014 American Chemical Society. *E-mail: [email protected]; tel: +86-0551-63600035. Cite this:J...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/JPCC

Interaction of C1 Molecules with a Pt Electrode at Open Circuit Potential: A Combined Infrared and Mass Spectroscopic Study Qian Tao,† Yong-Li Zheng,† Dao-Chuan Jiang,† Yan-Xia Chen,*,† Zenonas Jusys,‡ and R. Juergen Behm‡ †

Hefei National Laboratory for Physical Science at Microscale and Department of Chemical Physics, University of Science and Technology of China, Hefei, 230026, China ‡ Institute of Surface Chemistry and Catalysis, Ulm University, D-89069 Ulm, Germany ABSTRACT: The interaction of CO, HCOOH, HCHO, and CH3OH molecules with a Pt surface initially covered with a layer of oxide under open circuit potential (OCP) is studied by combined infrared and mass spectroscopy. We found that after switching to the fuel containing solution and concomitantly switching off the electrode potential control at 1.2 V, (i) the OCP decays from 1.2 V down to values of 0.58 V (CO), 0.12 V (HCOOH), 0.08 V(HCHO) and 0.24 V(CH3OH); (ii) CO is the only adsorbate formed at the Pt surface from the fuels; (iii) the rates for the decay of OCP and for the buildup of COad adlayer decrease in the order of CO > HCHO > HCOOH > CH3OH; (iv) the rate of CO2 production and the total amount of CO2 produced decreases in the order of CO > HCOOH > HCHO > CH3OH; and (v) a significant amount of HCOOH is formed for the case with HCHO and the main by product from CH3OH is HCHO. Our results indicate that (i) the change in OCP is determined by the change of net charge at the electrode/ electrolyte interface due to the production of electrons from fuel oxidation and the consumption of electrons by Oad/OHad reduction; and (ii) even without an externally potential control, the reactions occurring at the interface are controlled by the electrochemical potential of the respective reactants and products. The implication from OCP transient to the local potential distribution and local activity in practical fuel cells are discussed.

1. INTRODUCTION Direct methanol fuel cells (DMFCs) and direct formic acid fuel cells (DFAFCs) are considered to be of great potential as energy devices for portable electronics. Due to the simplicity in their molecular structure, the electrochemical oxidation of C1 molecules has been taken as model system in electrocatalysis since the early 1960s.1−16 Two of the major challenges for such fuel cells based on proton exchange membranes (PEM) are the slow kinetics for the reactions at the anode and the crossover of fuel from the anode to the cathode. In order to mitigate such problems, extensive studies have been carried out for a better understanding of both the mechanisms of such reactions and the key parameters which limit their kinetics.15 Though extensive molecular level information has been gained during the last 30 years, however, even for the simplest case, i.e., the electrochemical oxidation of HCOOH, the detailed mechanism for this reaction is still under debate.12,13,17 Understanding how the reactants interact with the electrocatalysts will be of great help for the rational design of improved catalysts for the anode as well as to mitigate the problems caused by fuel crossover to the cathode. Most model studies carried out so far are under external potential or current control, under such conditions the reactions are accelerated by both the catalysts and the externally applied electrical driving force, where the contributions from both factors are not easily distinguished. Studies on the interaction of fuel molecules with the electrode/electrolyte interface under conditions without © 2014 American Chemical Society

applying potential externally better resembles the real situation in the operating fuel cell. Such studies will provide unambiguous information on how catalysts function under such circumstances. Only a few studies reported the interactions of fuels with preoxidized Pt surfaces under open-circuit conditions. Earlier in the 1960s, Oxley et al. reported that, at oxidized platinum black electrodes/5 N H2SO4 interface, methanol is spontaneously decomposed to CO2 and H2 at around 60 °C.18 Podlovchenko et al. studied the transients of the open-circuit potential (OCP) of all the C1 molecules in contact with oxide-covered platinum surfaces,19−24 it was concluded that oxide consumption is mainly attained by the “conjugated reactions” mechanism in which the electrochemical oxidation of methanol is coupled to the reduction of oxygen.19,25 Varela et al. have also investigated the effects of preoxidation time and the anions on the OCP transients at Pt oxide/methanol solution interface. 26,27 Recently, they also combined measurements of OCP transients with IR spectroscopy to follow the production of COad at the Pt surface and CO2 trapped in the HCOOH-containing electrolyte; they found that the fuel oxidation and the concomitant Ptoxide reduction display an autocatalytic effect.28 Received: December 6, 2013 Revised: February 24, 2014 Published: March 10, 2014 6799

dx.doi.org/10.1021/jp411982p | J. Phys. Chem. C 2014, 118, 6799−6808

The Journal of Physical Chemistry C

Article

s−1 under a continuous flow of N2-saturated 0.5 M H2SO4 solution, until reproducible cyclic voltammograms are obtained (Figure 1a, dotted line). Then, after the electrode potential is

Despite the significant contributions mentioned above, questions such as what are the major adsorbates formed at the electrode surface, what are the main products, how quick and how much are they produced, what about the products ratio, how do they differ with the nature of the fuel and so on, still remained open. To answer these questions, we report results of our recent study on the interaction of CO, HCOOH, HCHO, and CH3OH with an oxygen covered Pt electrode surface under open circuit conditions, which is monitored simultaneously by infrared spectroscopy under attenuated total reflection configuration (ATR-FTIRS), differential electrochemical mass spectroscopy (DEMS) and the open circuit transients. DEMS allows to monitor the volatile products formed at the electrode and which subsequently diffuse into bulk solutions,29 and the ATR-FTIRS detects the adsorbates/ intermediates produced and adsorbed at electrode surface.8,30 The OCP transient follows the change of the charge density and the potential difference across the electrode/electrolyte interface. Based on the simultaneously recorded information from three different methods, a molecular level understanding of the processes occurring at the electrode/electrolyte interface will be provided.

Figure 1. (a) i−E curve for CO stripping (solid line) and the base cyclic voltammogram (dotted line) and (b) the mass signal m/z = 44 for CO2 during CO stripping at Pt thin film electrode/0.5 M H2SO4 interface, potential scan rate, 20 mV/s.

2. EXPERIMENTAL SECTION Millipore Milli-Q water and sulfuric acid (suprapure, Merck) are used to prepare 0.5 M H2SO4, which is used as supporting electrolyte. Before the measurement, the supporting electrolyte is first purged with N2 (N5.0, Linde China) for 10 min, then it is bubbled with CO (N 4.7, Linde, China) for ca. 15 min after the solution is saturated with CO. The solution is constantly purged with CO during the measurements. 0.5 M H2SO4 + 0.1 M HCOOH, 0.5 M H2SO4 + 0.1 M CH3OH, 0.5 M H2SO4+0.1 M HCHO are prepared using formic acid (98%, Kanto Chemical Co., Inc.), HCHO (16%, CH3OH free, Ployscience, Inc.), CH3OH (for analysis, Fluka). Before the measurements, the HCOOH, CH3OH, and HCHO containing solutions are purged with N2 for 15 min and continuously purged with N2 during the measurements. The working electrode (WE) used in this study is a Pt thin film electrode with a thickness of ca. 50 nm deposited on the flat reflecting face of a hemicylindrical Si prism by electroless plating following the procedure described elsewhere.30 The roughness factor of the film electrode is ca. 7 as estimated from the charge for the oxidation of a saturated H adlayer formed in the potential region from 0.4 to 0.05 V. The Pt film deposited on the Si prism is very stable under electrocatalytic oxidation of small fuels, e.g., we have used such film electrode to study CO and HCOOH oxidation at various temperatures from 5 °C up to 70 °C under both potenstiostatic or potential sweeping conditions.31,32 A Pt gauge and a Pt wire are used as counter electrodes and a reversible hydrogen electrode (RHE) is used as a reference electrode. The dual thin-layer spectroelectrochemical flow-cell, used for combined DEMS and in situ ATR-FTIRS, has been described in detail in refs 8,9, and 14. The cell volume is ca. 10 μL (100 μm thick O-ring ×1 cm2 geometric surface area of the WE), the electrolyte flow rate is ca. 50 μL/s. The cell assembly is mounted into the sample chamber of the IR spectrometer and the backside of the cell is interfaced to a DEMS system through a porous Teflon membrane (Scimat, 60 μm thick, 50% porosity, 0.2 μm pore diameter). The electrode surface is precleaned by cycling in the potential region from 0.06 to 1.3 V at a scan rate of 100 mV

scanned from the lower limit to 1.2 V and subsequently the potential is held at 1.2 V for 1 min, the electrode potential control is switched off, and the electrolyte is switched to a C1 molecule-containing solution. The OCP, the mass spectrometric signal of CO2 (m/z = 44), and the IR spectra are recorded simultaneously. All potentials quoted in this study are given versus the RHE. The electrode potential is controlled using a potentiostat (CHI760 E, Shanghai ChenHua, China). The DEMS setup used in the present study is a HidenHPR-40 DSA Bench TopMembrane Inlet Gas Analysis System, mass signals are collected 20 points/sec. For reaction with CH3OH the mass signal for m/z = 60 for CH3COOCH29 is also recorded. The IR spectroscopic measurements were carried out using a Varian FTS-7000 spectrometer equipped with a mercury cadmium telluride (MCT) detector. A spectrum taken at 1.2 V in the supporting electrolyte is used as reference spectrum. All spectra are obtained with a resolution of 4 cm−1 and 1 spectrum/s. The spectra are presented in absorbance mode, i.e., log(R0/R), where R0 and R are the reflectance at reference and sample potential, respectively.

3. RESULTS AND DISCUSSION 3.1. Interaction of CO with Pt−O at Open Circuit Potential. In order to calibrate the mass signal of CO2 production, linear potential stripping of a saturated COad adlayer at Pt film electrode formed by adsorption of CO at 0.06 V for 10 min has been carried out, and the mass signal for m/z = 44 has been recorded simultaneously (Figure 1). From Figure 1, the integrated charge for oxidation of a saturated CO adlayer is Q = 3.76 mC, and the corresponding CO2 mass signal from oxidation of a saturated CO adlayer is 8.44 × 10−8 Torr s, from which the calibration constant for the CO2 mass signal is found to be k = 4.49 × 10−8 Torr mA−1. Figure 2 displays the change of the OCP, the mass signal for CO2, and the integral IR band intensity of linearly and multiply bonded adsorbed CO (denoted as COL and COM hereafter) during CO oxidation under open potential conditions as a function of reaction time 6800

dx.doi.org/10.1021/jp411982p | J. Phys. Chem. C 2014, 118, 6799−6808

The Journal of Physical Chemistry C

Article

Figure 2. Time evolution of (a) open circuit potential, (b) m/z = 44 mass signal for CO2 and (c) integral infrared band intensity of COL and COM at the Pt electrode in CO saturated 0.5 M H2SO4 solution after holding the potential at 1.2 V for 1 min and subsequent switching off the potential control at 0 s together with changing to CO saturated electrolyte. The right panel illustrates the signals at an expanded time scale from 7 to 18 s after switching off the potential control.

Figure 3. IR spectra of COL and COM at a Pt electrode as a function of contact time in 0.5 M H2SO4 with (a) saturated CO, (b) 0.1 M HCOOH, (c) 0.1 M HCHO, and (d) 0.1 M CH3OH solutions after holding the potential at 1.2 V for 1 min and subsequent switching off the potential control at 0 s, together with changing from 0.5 M H2SO4 to fuel containing electrolyte.

its intensity decreases with time and drops to ca. zero at 18 s after the switching solution supply (right panel in Figure 2). From Figure 2c it is seen that the IR signal for COad appears when the OCP drops to below 0.79 V, this is exactly the same potential where the mass signal for CO2 starts to decrease. After that, the band intensities for both COL and COM species increase with time and reach saturation within 3 s after their appearance. At that point, the electrode potential reaches ca. 0.58 V. The OCP transients obtained from present study are very similar to the previous observation under otherwise

after switching to CO saturated 0.5 M H 2 SO 4 and concomitantly switching off the potential control at 1.2 V. IR spectra of the stretching vibrations of COL and COM bands are given in Figure 3a. From Figure 2, it is obvious that after switching to CO saturated electrolyte the OCP decreases slowly within the first 9 s, and then drops sharply from 1.15 V at 9 s to 0.9 V at 10 s. Subsequently, it decays slowly to ca. 0.7 V at 13 s, and finally it gradually approaches a steady-state value (ca. 0.58 V) at ca. 35 s. The mass signal for CO2 appears at ca. 9 s after changing the solution, and reaches its maximum at ca. 11.7 s (at the same time, the potential drops to 0.79 V), Then 6801

dx.doi.org/10.1021/jp411982p | J. Phys. Chem. C 2014, 118, 6799−6808

The Journal of Physical Chemistry C

Article

identical condition,25 indicating that the results obtained in both laboratories are well reproducible. The OCP in N2 saturated 0.5 M H2SO4 solution is ca. 0.93 V;33 it represents a mixed potential for the following processes: Pt−Oad + H+ + e− ⇄ Pt−OHad

(1)

Pt−OHad + H+ + e− ⇄ Pt + H 2O

(2)

Pt2 + + 2e− ⇄ Pt

(3)

active sites by COad, the rate for OHad formation from the backward reaction in eq 4 will be significantly reduced. After that, the production rate for CO2 decrease and reaches zero when the surface is completely poisoned by COad. This is indicated by the IR spectroscopic data (Figure 2c and Figure 3). After reaching a saturated CO adlayer, the net charge at the electrode/electrolyte interface does not change anymore, consequently the OCP also does not change with time anymore. During the time period when the OCP drops from 1.2 to 0.79 V, the integral mass signal for CO2 produced is 7 × 10−8 Torr s and, until reaching a steady state OCP of 0.58 V, it is ca. 1.3 × 10−7 Torr s (Figure 8c), which is ca. 1.5 times of that from the oxidative stripping of a saturated COad layer. Since the coverage for a saturated CO adlayer is ca. 0.7 ML,36 the total amount of CO2 produced during the OCP transients by exposing oxide covered Pt electrode to CO saturated solution corresponds to the oxidation of a full monolayer of COad. From the reduction charge it is estimated that initial Oad coverage right after switching off the potential control at 1.2 V is ca. 0.5 ML.36 Considering that 1 ML Oad/OHad is consumed based on the amount of CO2 formation, but only 0.5 ML Oad/OHad/ Ooxide was present when switching off the potential, it appears that the formation of about 50% of the CO2 produced does not involve a net charge transfer, i.e.,

The OCP of the Pt/0.5 M H2SO4 interface is mainly determined by the equilibrium potential of reaction 2 since it has the highest exchange current density among all three reactions.33,34 When switching off the potential control at 1.2 V, the Pt surface is initially covered with Oad or at least locally transferred into a surface oxide (Pt2+ formation). Right after switching off the potential control, the electrochemical potential of Pt−Oad + H+ + e is higher than that of its counter parts OH ad or H 2O in the solution, so Pt−O ad will spontaneously reduce to Pt−OHad or even to bare metallic Pt through reactions 1 and 2. However, since the surface coverage of Oad is rather high, the Pt surface is slightly passivated by Oad, therefore the rate for the reduction of Pt− Oad to Pt−OHad is at least initially slow. On the other hand, since a significant amount of positive charges are accumulated at the Pt surface right after switching off the potential control (the potential of zero charge (PZC) for Pt electrode in 0.5 M H2SO4 is ca. 0.3 V).35 Under such high oxidative environment, some Pt atoms in the sublayer will be oxidized to Pt2+, as similar to the case without switching off the potential control. As a result, the net positive charge at the Pt electrode only decreases slowly within the first 9 s after switching off the potential control, and correspondingly the OCP drops only slightly from 1.2 to 1.15 V. When the Oad coverage is close to saturation, since there are no free sites for stable CO adsorption (as supported by the fact that no IR signal for adsorbed CO is observed within this time regime (Figure 2c), chances for a chemical reaction between COad and Oad through an Langmuir−Hinshelwood (LH) reaction mechanism are small. On the other hand, contributions from an Eley−Rideal (EL) mechanism, which theoretically would be possible by reaction of dissolved CO with Oad on an Oad covered Pt surface, can essentially be ruled out from the fact that no CO2 is detected in this OCP regime. After 9 s at open potential, a small portion of bare metallic Pt sites becomes available and CO can adsorb at such surface sites and reacts with the neighboring Oad or OHad promptly. As a result, significant amounts of CO2 are produced through Pt−OHad + COad → Pt + CO2 + H+ + e−

(4)

Pt−Oad + COad → Pt + CO2

(5)

CO + Oad → CO2

(6)

Another 0.5 ML O/OH must be supplied from the electrolyte, e.g., via the backward reaction of eq 1 and 2. For each CO2 formed this way, two electrons will be supplied to the electrode, according to the overall reaction equation: H 2O + CO → [OHad + H+ + e− + COad ] → CO2 + 2H+ + 2e−

(7)

This supply of electrons is mainly responsible for the decrease in OCP. Since the numbers of electrons donated to the electrode through CO oxidation are limited, the positive charge accumulated at the Pt/0.5 M H2SO4 interface right after switching off the potential control at 1.2 V cannot be fully neutralized by the electrons produced through CO oxidation. Hence the final OCP with a value of ca. 0.6 V for Pt in CO saturated solution is still a bit higher than the potential of zero charge of this system (ca. 0.3 V).35 The time for the OCP transients to reach steady-state when exposing an Oad covered surface to a CO saturated solution is only ca. 1/10 of that reported by Podlovchenko et al, where more than 400 s were necessary in order to reach the steadystate OCP.20 We attribute this difference to the different Pt electrode pretreatment. In their study, before recording the current transient, they held the electrode at 1.5 V for 10 min, then jumped to 1.1 V. Hence, the initial Oad coverage on the Pt surface used for OCP transient study is much higher than the case in the present study; some Oad may even have diffused into the subsurface of the Pt lattice. Such differences reveal a lower reactivity of the highly Oad covered/oxidized Pt electrode, in agreement with expectation for a LH mechanism. Finally, it is also interesting to note that the OCP displays a small plateau with a value of ca. 0.7 V from 12.5 to 13.5 s after the solution switch. By careful repeating the experiments we found that this characteristic OCP plateau is well reproducible. The plateau coincides with a steep increase in COL and COM intensity and hence in COad coverage, while the CO2 formation rate

The free Pt site produced in reaction 4 can adsorb another CO, so more Pt−Oad can be reduced to Pt−OHad or Pt. On the other hand, when the OCP is above 0.6 V, the free Pt site produced in reaction 4 can also adsorb another OHad formed via the backward reaction in eq 2. Due to the accumulation of electrons at the electrode through eq 4 and the backward reaction in eq 2, the OCP drops promptly. Once the OCP is below 0.79 V, the rate for CO adsorption becomes higher than that for COad oxidation, COad starts to accumulate on the Pt surface. This is illustrated by the fact that the CO2 production rate reaches its maximum at ca. 0.79 V. Due to the blocking of 6802

dx.doi.org/10.1021/jp411982p | J. Phys. Chem. C 2014, 118, 6799−6808

The Journal of Physical Chemistry C

Article

Figure 4. Time evolution of (a) open circuit potential, (b) m/z = 44 mass signal for CO2 and (c) integral infrared band intensities of COL and COM at the Pt/0.5 M H2SO4 + 0.1 M HCOOH interface after holding the potential at 1.2 V for 1 min and subsequent switching off the potential control at 0 s together with changing to HCOOH containing electrolyte. The right panel illustrates the signals at an expanded time scale from 15 to 19 s after switching off the potential control.

decreases (Figure 2). Apparently the decrease in the accumulation of electrons from CO2 formation (reaction 4) at the electrode is compensated by a similar decrease in the consumption of electrons by OHad removal (reaction 2). 3.2. Interaction of HCOOH with Pt−O at Open Circuit Potential. Figure 4 displays the change of OCP (a), the mass signal for CO2 (b) and the integral IR band intensity of COL and COM (c) as a function of reaction time after switching to 0.5 M H2SO4 + 0.1 M HCOOH and concomitantly switching off the potential control at 1.2 V. The corresponding IR spectra of COL and COM are given in Figure 3b. In general, the transients of the OCP at Pt/HCOOH solution interface, the time evolution of the IR signal for COad, and that of the DEMS signal for CO2 are quite similar to those reported by Bastista et al. based on their infrared spectroscopic study.28 On the other hand, the transients for the OCP at Pt in contact with HCOOH containing solution observed in this study are much shorter than what was reported by Manzhos et al.19 Again we attribute the difference to the higher Oad coverage/oxidation state of the initial Pt surface in the latter study. Similar to the reaction in CO saturated solution (previous section), the OCP changes only very little from 1.2 to 1.15 V in the first 10 s, and no CO2 or COad is produced during this time period. A very small mass signal of CO2 appears at 11.8 to 16.4 s, which is accompanied by a decay of the OCP from 1.15 to 0.88 V. Based on the appearance of the mass signal for CO2, we conclude that during that time some HCOOH is oxidized to CO2, probably via Pt + HCOOH → CO2 + 2H+ + 2e−

(8)

Pt + HCOO− → CO2 + H+ + e−

(9)

Figure 5. Time evolution of (a) open circuit potential (square) (b) relative surface coverage of COad (θrel, triangle) and (c) the integrated mass signal for CO2 formation (star) at the Pt/0.5 M H2SO4 + 0.1 M HCOOH interface, θrel is derived from the IR spectral data given in Figure 4.

previously occupied by Oad, via reactions 1 and 2. After that, the CO2 signal increases sharply and reaches its maximum at ca. 16.9 s. At the same time, the OCP drops to 0.5 V. The steep increase in the rate for CO2 production and the steep drop in the OCP are due to the autocatalytic behavior of the system, both HCOOH adsorption and oxidation occur much faster on the metallic Pt sites after the ignition period. The oxidation of HCOOH in turn provides more electrons to reduce Pt−Oad and Pt−OHad, which in turn leads to more free sites for further oxidation of HCOOH.28 The continuous accumulation of

Together with reactions 1 and 2, this leads to the slow decay in OCP from 1.18 V at 11.8 s to 0.88 V at 16.4 s. During the period when the OCP drops from 1.2 to 0.88 V, the amount of CO2 produced corresponds to ca. 25% of that from oxidation of a full monolayer of COad (Figure 5). Since HCOOH oxidation also involves the release of two electrons, this corresponds to the formation of about 25% bare Pt surface sites, which were 6803

dx.doi.org/10.1021/jp411982p | J. Phys. Chem. C 2014, 118, 6799−6808

The Journal of Physical Chemistry C

Article

Figure 6. Time evolution of (a) open circuit potential, (b) m/z = 44 mass signal for CO2 and (c) integral infrared band intensities of COL and COM at the Pt/0.5 M H2SO4 + 0.1 M HCHO interface after holding the potential at 1.2 V for 1 min and subsequent switching off the potential control at 0 s together with changing from 0.5 M H2SO4 to HCHO containing electrolyte. The right panel illustrates the signals at an expanded time scale from 9 to 13 s after switching off the potential control.

3.3. Interaction of HCHO and CH3OH with Pt−O at Open Circuit Potential. The time evolution of the OCP, the mass signal for CO2 and the integral IR band intensity of COL and COM upon exposure of the Pt electrode to HCHO containing solution (Figure 6) are qualitatively similar to those observed for reactions with HCOOH and CO (Figures 2 and 4). Besides the similarities, a few differences are also noticed: (i) No CO2 mass signal is observed at OCP > 0.9 V, and the total amount of CO2 produced is only ca. 15% of that for the reaction with CO (Figure 8); (ii) after the ignition period (first 9 s), the OCP drop is much steeper than for the reaction with HCOOH; and (iii) the OCP at steady-state is just 0.08 V, which is significantly lower than upon reaction with CO (0.58 V) and also ca. 50 mV smaller than upon reaction with HCOOH. It should be noticed that in aqueous solution containing HCHO, the majority of the reaction components in the solution is in the form of H2C(OH)2, the equilibrium constant for HCHO + H2O ⇄ H2C(OH)2 is K = 2280.37,38 The absence of CO2 production at OCP > 0.90 V can be explained by the fact that at such high potentials, no active sites with an ensemble size large enough for H2C(OH)2 adsorption are available. This also indicates that larger ensembles of Pt atoms are required in order to get H2C(OH)2 adsorbed than that for CO. The fast decay of the OCP is probably due to the fact that once a few active metallic Pt sites for H2C(OH)2 adsorption are available, adsorbed H2C(OH)2 molecules can easily react according to eq 11 to produce CO2

electrons at the Pt surface drives the OCP further down to 0.5 V. In the meantime, the simultaneously recorded IR spectra reveal that COad appears when the OCP just reaches ca. 0.5 V, indicating that this reaction also takes place Pt + HCOOH → Pt−CO + H 2O

(10)

The change of COad coverage as a function of reaction time is also given in Figure 5. From that figure, it is clearly seen that COad only appears at E < 0.5 V. This is similar to our previous observation from systematic studies on the potential effect of HCOOH adsorption and oxidation at a Pt electrode.8,9 It should be mentioned that CO is also probably formed at higher potentials, but reacts immediately with OHad or Oad to CO2. Simultaneously with the fast buildup of COad, the OCP drops steeply and reaches a steady-state value of ca. 0.13 V at ca. 23 s. Under steady-state conditions, the intensity of the IR signal of COL is similar to that obtained when exposing to a CO saturated solution, while that for COM is ca. 2 times higher than in CO saturated solution. This can be explained by the difference in the steady-state OCP, which is 0.58 V in CO saturated solution, while it is ca. 0.13 V in HCOOH containing solution. It is well confirmed that at the lower potential COad prefers to adsorb at multiply bonded sites.3,14 A special feature with HCOOH is that even when the OCP is as low as 0.15 V (Figures 4 and 5), HCOOH can still be oxidized to CO2. Under these conditions, however, the Pt surface is soon poisoned by COad and the CO2 production rate drops sharply with increasing COad coverage. Finally, it stops once no active sites are available for HCOOH adsorption and reaction anymore. It should be mentioned that, in this study, we do not observe any signal for bridge-bonded formate during OCP transient at Pt/0.5 M H2SO4 + 0.1 M HCOOH interface. We think the lacking of the appearance of IR band of HCOOad is probably due to the heavy coverage of Oad/OHad and COad at high OCP and at the low OCP, respectively. The fast build-up of CO renders that there is no chance for weakly adsorbed species such as formate to adsorb.

Pt

H 2C(OH)2 → CO2 + 4H+ + 4e−

(11)

Alternatively, it loses two protons and two electrons to produce HCOOH via Pt

H 2C(OH)2 → HCOOH + 2H+ + 2e−

(12)

The occurrence of reaction 12 is supported by the fact that compared to the reaction with CO, the amount of CO2 produced for reaction with HCHO is only ca. 15%. This result is similar to findings in previous DEMS studies that the current 6804

dx.doi.org/10.1021/jp411982p | J. Phys. Chem. C 2014, 118, 6799−6808

The Journal of Physical Chemistry C

Article

Figure 7. Time evolution of (a) open circuit potential, (b) m/z = 44 mass signal for CO2, and (c) integral infrared band intensities of COL and COM at the Pt/0.5 M H2SO4 + 0.1 M CH3OH interface after holding the potential at 1.2 V for 1 min and subsequent switching off the potential control at 0 s together with changing from 0.5 M H2SO4 to CH3OH containing electrolyte. The right panel illustrates the signals at an expanded time scale from 21 to 25 s after switching off the potential control.

efficiency for CO2 production from HCHO at a Pt surface is less than 40%.39 Similar to the previous cases of reaction with CO or HCOOH, the electrons produced will consume Oad and OHad through reactions 1 and 2, this provides more metallic Pt sites for both reactions 11 and 12, and a rapid increase of the HCOOH and CO2 formation rate via this autocatalytic mechanism. For the reaction with CH3OH, the OCP drop after 10 s of ignition period is the slowest among the four C1 molecules, and the amount of CO2 produced is the least (ca. 5% of that obtained for the reaction with CO). The final OCP under steady-state conditions (0.23 V) is slightly higher than that obtained for reactions with HCOOH or HCHO (Figures 7 and 8). The slower rate for CO2 production and the slower decay in OCP for reaction with CH3OH is probably due to a combination of the following reasons: (i) the steric resistance for the reaction with CH3OH is the largest; (ii) the splitting of the C−H bond in CH3OH is more difficult than that in HCOOH and H2C(OH)2;1 (iii) in order to form CO2, consecutive splitting of C−H bonds, the transfer of six electrons as well as the formation of an additional C−O bond are necessary. This will only occur when the OCP drops to below 0.9 V where metallic Pt sites are available. Only under such conditions, efficient adsorption of CH3OH molecule and the splitting of its C−H bond is possible. This is in good agreement with findings in previous DEMS studies, where it was observed that at E > 0.9 V, the majority products from CH3OH oxidation at Pt are HCHO and HCOOH.29,40 We also followed the mass signal m/z = 60 for methylformate, which is often used as indicator for the production of HCOOH, which reacts with CH3OH in the solution to form CH3COOH. However, no such signal was detected (line with circle, Figure 7, note the signal level is magnified by 100 times comparing to the scale for m/z = 44), indicating that the amount of HCOOH produced under these conditions is negligible. This is in contrast to previous studies under potential control where significant amounts of HCOOH is formed, the reason for such difference may due to the fact that under present OCP condition Oad/OHad consumed cannot

Figure 8. Time evolution of (a) open circuit potential, (b) m/z = 44 mass signal for CO2, and (c) integral m/z = 44 mass signal for CO2 at the Pt in 0.5 M H2SO4 with 0.1 M HCOOH (square) or HCHO (triangle) or CH3OH (circle) after holding the potential at 1.2 V for 1 min and subsequent switching off the potential control at 0 s together with changing to the fuel containing electrolyte.

be continuously replenished and electrons produced cannot be continuously transported away from the electrode. The results indicate that during the OCP transients when exposing Pt electrode to CH3OH, the main reaction product is HCHO via 6805

dx.doi.org/10.1021/jp411982p | J. Phys. Chem. C 2014, 118, 6799−6808

The Journal of Physical Chemistry C

Article

Pt

CH3OH → HCHO + 2H+ + 2e

reactions taking place at the Pt/electrolyte interface under OCP conditions follow similar mechanisms as with external potential control at identical potentials. The results clearly reveal that the behavior of the reactions occurring at the electrode/electrolyte interface are electrochemical in nature, the thermodynamics of the processes are governed by the electrochemical potential of the respective species involved in the reaction at the interface, no matter whether it takes place with or without external potential control. Under the OCP condition (when the connection to the external circuit is off), the accumulation or consumption of small amount electrons at the electrode/electrolyte interface via redox reaction, may cause significant change in the electrode potential across the electrode/electrolyte interface. In turn, this will greatly change the electrochemical potential of electrons, the structure of the interface as well as the reactivity of the desired reaction. If the electronic conductivity of the porous catalysts layer within the membrane electrode assembly in practical fuel cells is not good and uniform, the amount of electrons accumulated at local active sites may differ a lot. This may cause great difference in local potential distribution as well as the local activity.

(13)

This reaction involves only the transfer of two electrons. Due to the steric hindrance in the methanol adsorption and C−H bond breaking reaction step, the kinetics for reaction 13 is much slower than that for CO, HCOOH and H2C(OH)2, and hence the drop for the OCP is also slowest. On the other hand, as evident from Figure 7c for reaction with CH3OH, COad appears already at OCP in the potential region from 0.8 to 0.65 V, which is much higher than the values obtained for HCHO (H2C(OH)2) (ca. 0.6 V) and for HCOOH (ca. 0.45 V). This trend is also quite similar to what we observed during potentiodynamic oxidation of C1 fuels by cyclic voltammetry.30,31,41 The higher potential for the onset of COad accumulation during CH3OH oxidation as compared to HCOOH and HCHO oxidation may be related to the fact that for the reaction with CH3OH, consecutive splitting of C− H and O−H bonds will lead to COad. In contrast, for the reaction with HCOOH and for H2C(OH)2, reductive splitting of one C−O bond is necessary. The latter will only happen at lower potentials, when the surface is largely metallic in nature and when the channels for reactions 8, 9 and 11, 12 are kinetically unfavorable. The lower values of the final OCPs obtained for CH3OH and HCHO oxidation are explained by a combination of different effects. First, a less efficient COad poisoning at OCP values above 0.6 V may allow additional surface oxygen formation from H2O (reverse reaction in eq 2) and oxidation of CH3OH with these surface oxygen species. Second, formation of the partly oxidized reaction product HCHO, which generates 2e per reacting CH3OH, is possible without additional surface oxide. The same is possible for HCOOH formation from H2C(OH)2, which also generates 2e per reacting molecule. Apparently, these processes for current generation are more efficient for HCHO, where the lowest OCP (0.08 V) is observed, than for CH3OH oxidation. 3.4. Implications from the Comparison of OCP Transient at Pt for All Four C1 Fuels. From Figure 8 it is clearly seen that immediately after the ignition period, the slope for the OCP-time transients in the potential region from 1.1 to 0.8 V decreases in the order of CO > HCHO > HCOOH > CH3OH, although the concentration of CO in the solution saturated with CO is only around 1 mM, while those for other fuels are 0.1 M. This indicates that the interaction of the fuels with Oad/OHad is mainly kinetically controlled rather than mass transport limited. Hence, the slope of the OCP transients can be considered as an indicator that the kinetics of the C−H bond splitting reaction at the Oad/OHad covered Pt surface increases in the order of CH3OH HCHO > HCOOH > CH3OH, the amount of CO2 produced decays in the order of CO > HCOOH > HCHO > CH3OH. The OCP under steadystate decreases in the order of CO > CH3OH > HCOOH > HCHO. Our results demonstrate that the OCP changes sensitively when just a small amount of species is consumed in the redox reaction and the OCP is purely determined by the numbers of electrons accumulated at the electrode surface due to the oxidation of the C1 fuels and the reduction of Oad/OHad. The interfacial processes occurring during the OCP transients are identical to those observed under potential control at potentials close to the values of the respective OCPs.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]; tel: +86-0551-63600035. 6806

dx.doi.org/10.1021/jp411982p | J. Phys. Chem. C 2014, 118, 6799−6808

The Journal of Physical Chemistry C

Article

Notes

Applications; Vielstich, W., Lamm, A., Gasteiger, H. A., Yokokawa, H., Eds.; John Wiley & Sons, Ltd.: New York, 2010; pp 1−9. (17) Joo, J.; Uchida, T.; Cuesta, A.; Koper, M. T.; Osawa, M. Importance of Acid−Base Equilibrium in Electrocatalytic Oxidation of Formic Acid on Platinum. J. Am. Chem. Soc. 2013, 135, 9991−9994. (18) Oxley, J.; Johnson, G.; Buzalski, B. The Anodic Oxidation of Methanol: Open-Circuit Transient Reactions. Electrochim. Acta 1964, 9, 897−910. (19) Podlovchenko, B.; Manzhos, R.; Maksimov, Y. M. Kinetics and Mechanism of Interaction Between Methanol and Adsorbed Oxygen on a Smooth Polycrystalline Platinum Electrode: Transients of the Open-Circuit Potential. Russ. J. Electrochem. 2006, 42, 1061−1066. (20) Podlovchenko, B.; Manzhos, R.; Maksimov, Y. M. Transients of the Open-Circuit Potential During Carbon Monoxide Interaction with Oxygen Preliminarily Adsorbed on Smooth and Platinized Platinum Electrodes. Russ. J. Electrochem. 2002, 38, 1292−1298. (21) Manzhos, R.; Podlovchenko, B.; Maksimov, Y. M. Specific Features of Interaction Between Formic Acid and Oxygen Adsorbed on Smooth Polycrystalline Platinum: Transients of the Open-Circuit Potential. Russ. J. Electrochem. 2006, 42, 658−664. (22) Manzhos, R.; Maksimov, Y. M.; Podlovchenko, B. Transients of the Open-Circuit Potential Observed During the Interaction of Formic Acid with Preliminarily Adsorbed Oxygen on a Platinized Platinum Electrode. Russ. J. Electrochem. 2005, 41, 832−837. (23) Manzhos, R.; Maksimov, Y. M.; Podlovchenko, B. Effect of the Pt/Pt Electrode Roughness on the Interaction of CO with Oads. Russ. J. Electrochem. 2004, 40, 563−568. (24) Podlovchenko, B.; Manzhos, R.; Maksimov, Y. M. Interaction of HCO-substances with Adsorbed Oxygen on Platinum Electrodes: Open-Circuit Transient Reactions of HCOOH and CO. Electrochim. Acta 2005, 50, 4807−4813. (25) Heinen, M.; Chen, Y.; Jusys, Z.; Behm, R. In situ ATR-FTIRS Coupled with On-line DEMS Under Controlled Mass Transport ConditionsA Novel Tool for Electrocatalytic Reaction Studies. Electrochim. Acta 2007, 52, 5634−5643. (26) Sitta, E.; Varela, H. On the Open-Circuit Interaction Between Methanol and Oxidized Platinum Electrodes. J. Solid State Electrochem. 2008, 12, 559−567. (27) Batista, B. C.; Sitta, E.; Eiswirth, M.; Varela, H. Autocatalysis in the Open Circuit Interaction of Alcohol Molecules with Oxidized Pt Surfaces. Phys. Chem. Chem. Phys. 2008, 10, 6686−6692. (28) Batista, B. C.; Varela, H. Open Circuit Interaction of Formic Acid with Oxidized Pt Surfaces: Experiments, Modeling, and Simulations. J. Phys. Chem. C 2010, 114, 18494−18500. (29) Jusys, Z.; Behm, R. Methanol Oxidation on a Carbon-Supported Pt Fuel Cell Catalyst a Kinetic and Mechanistic Study by Differential Electrochemical Mass Spectrometry. J. Phys. Chem. B 2001, 105, 10874−10883. (30) Chen, Y. X.; Miki, A.; Ye, S.; Sakai, H.; Osawa, M. Formate, An Active Intermediate for Direct Oxidation of Methanol on Pt Electrode. J. Am. Chem. Soc. 2003, 125, 3680−3681. (31) Chen, Y. X.; Ye, S.; Heinen, M.; Jusys, Z.; Osawa, M.; Behm, R. J. Application of in-situ Attenuated Total Reflection-Fourier Transform Infrared Spectroscopy for the Understanding of Complex Reaction Mechanism and Kinetics: Formic Acid Oxidation on a Pt Film Eectrode at Elevated Temperatures. J. Phys. Chem. B 2006, 110, 9534−9544. (32) Geng, B.; Cai, J.; Liang, S. Z.; Liu, S. X.; Li, M. F.; Chen, Y. X. Temperature Effects on CO Adsorption/Desorption at Pt Film Electrodes: An Electrochemical in situ Infrared Spectroscopic Study. Phys. Chem. Chem. Phys. 2010, 12, 10888−10895. (33) Tarasevich, M.; Sadkowski, A.; Yeager, E. Oxygen Electrochemistry. In Comprehensive Treatise of Electrochemistry; Springer: 1983; pp 301−398. (34) Li, M. F.; Liao, L. W.; Yuan, D. F.; Mei, D.; Chen, Y.-X. pH Effect on Oxygen Reduction Reaction at Pt (111) Electrode. Electrochim. Acta 2013, 110, 780−789.

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by National Natural Science Foundation of China (NSFC) (Project No. 21273215), 973 program from the Ministry of Science and Technology of China (Project No. 2010CB923302).

■

REFERENCES

(1) Jarvi, T. D.; Stuve, E. M. Fundamental Aspects of Vacuum and Electrocatalytic Reactions of Methanol and Formic Acid on Platinum Surfaces. In Electrocatalysis; Lipkowski, J., Ross, P. N., Eds.; John Wiley & Sons, Ltd.: New York, 1998; pp 75−154. (2) Hamnett, A. Mechanism of Methanol Electro-Oxidation. In Interfacial Electrochemistry: Theory, Experiment, and Applications; Wieckowski, A., Ed.; Marcel Dekker, Inc.: New York, 1999; pp 843−884. (3) Marković, N.; Ross, P., Jr. Surface Science Studies of Model Fuel Cell Electrocatalysts. Surf. Sci. Rep. 2002, 45, 117−229. (4) Hamann, C. H.; Hamnett, A.; Vielstich, W. Electrocatalysis and Reaction Mechanisms. In Electrochemistry; John Wiley & Sons Ltd.: New York, 2007; pp 339−380. (5) Li, M.; Adzic, R. R. Low-Platinum-Content Electrocatalysts for Methanol and Ethanol Electrooxidation. In Electrocatalysis in Fuel Cells: A Non- and Low-Pt Approach; Shao, M. H., Ed.; Springer-Verlag: London, 2013; pp 1−26. (6) Rice, C. A.; Wieckowski, A. Electrocatalysis of Formic Acid Oxidation. In Electrocatalysis in Fuel Cells: A Non- and Low-Pt Approach; Shao, M. H., Ed.; Springer-Verlag: London, 2013; pp 43− 68. (7) Rice, C. A.; Bauskar, A.; Pickup, P. G. Recent Advances in Electrocatalysis of Formic Acid Oxidation. In Electrocatalysis in Fuel Cells: A Non- and Low-Pt Approach; Shao, M. H., Ed.; Springer-Verlag: London, 2013; pp 69−88. (8) Chen, Y. X.; Heinen, M.; Jusys, Z.; Behm, R. J. Kinetics and Mechanism of the Electrooxidation of Formic AcidSpectroelectrochemical Studies in a Flow Cell. Angew. Chem., Int. Ed. 2006, 45, 981−985. (9) Chen, Y. X.; Heinen, M.; Jusys, Z.; Behm, R. J. Kinetic Isotope Effects in Complex Reaction Networks: Formic Acid ElectroOxidation. ChemPhysChem 2007, 8, 380−385. (10) Liao, L. W.; Liu, S. X.; Tao, Q.; Geng, B.; Zhang, P.; Wang, C. M.; Chen, Y. X.; Ye, S. A Method for Kinetic Study of Methanol Oxidation at Pt Electrodes by Electrochemical in situ Infrared Spectroscopy. J. Electroanal. Chem. 2011, 650, 233−240. (11) Liu, S. X.; Liao, L. W.; Tao, Q.; Chen, Y. X.; Ye, S. The Kinetics of CO Pathway in Methanol Oxidation at Pt Electrodes, A Quantitative Study by ATR-FTIR Spectroscopy. Phys. Chem. Chem. Phys. 2011, 13, 9725−9735. (12) Xu, J.; Mei, D.; Yuan, D.-F.; Zhang, Z.-B.; Liu, S.-X.; Chen, Y.-X. A Revisit to the Role of Bridge-Adsorbed Formate in the Electrocatalytic Oxidation of Formic Acid at Pt Electrodes. Chin. J. Chem. Phys. 2013, 26, 321. (13) Xu, J.; Yuan, D.; Yang, F.; Mei, D.; Zhang, Z.; Chen, Y.-X. On the Mechanism of the Direct Pathway for Formic Acid Oxidation at a Pt (111) Electrode. Phys. Chem. Chem. Phys. 2013, 15, 4367−4376. (14) Heinen, M.; Chen, Y. X.; Jusys, Z.; Behm, R. J. Room Temperature COad Desorption/Exchange Kinetics on Pt Electrodes A Combined In Situ IR and Mass Spectrometry Study. ChemPhysChem 2007, 8, 2484−2489. (15) Climent, V.; Garcia-Araez, N.; Feliu, J. M. Clues for the Molecular-Level Understanding of Electrocatalysis on Single-Crystal Platinum Surfaces Modified by p-Block Adatoms. In Fuel Cell Catalysis: A Surface Science Approach; Koper, M., Ed.; John Wiley & Sons Ltd: Hoboken, NJ, 2009; pp 209−244. (16) Varela, H.; Sitta, E.; Batista, B. Methanol Oxidation on Oxidized Pt Surface. In Handbook of Fuel Cells − Fundamentals, Technology and 6807

dx.doi.org/10.1021/jp411982p | J. Phys. Chem. C 2014, 118, 6799−6808

The Journal of Physical Chemistry C

Article

(35) Climent, V. c.; Gómez, R.; Feliu, J. M. Effect of Increasing Amount of Steps on the Potential of Zero Total Charge of Pt (111) Electrodes. Electrochim. Acta 1999, 45, 629−637. (36) Chen, D.; Tao, Q.; Liao, L. W.; Liu, S. X.; Chen, Y. X.; Ye, S. Determining the Active Surface Area For Various Platinum Electrodes. Electrocatalysis 2011, 2, 207−219. (37) Guthrie, J. P. Carbonyl Addition Reactions: Factors Affecting The Hydrate-hemiacetal and Hemiacetal-acetal Equilibrium Constants. Can. J. Chem. 1975, 53, 898−906. (38) Smolin, A.; Podlovchenko, B.; Maksimov, Y. M. Formaldehyde Interaction with Adsorbed Oxygen on Open-circuit Platinum Electrodes in Aqueous Sulfuric Acid Solutions. Russ. J. Electrochem. 2009, 45, 246−251. (39) Jusys, Z.; Behm, R. J. Methanol, Formaldehyde, and Formic Acid Adsorption/Oxidation on a Carbon-Supported Pt Nanoparticle Fuel Cell Catalyst: A Comparative Quantitative DEMS Study. In Fuel Cell Catalysis: A Surface Science Approach; Koper, M., Ed.; John Wiley & Sons Ltd.: Hoboken, NJ, 2009; pp 411−464. (40) Wang, H.; Lö f fler, T.; Baltruschat, H. Formation of Intermediates During Methanol Oxidation: A Quantitative DEMS Study. J. Appl. Electrochem. 2001, 31, 759−765. (41) Heinen, M.; Jusys, Z.; Behm, R. Reaction Pathway Analysis and Reaction Intermediate Detection via Simultaneous Differential Electrochemical Mass Spectrometry (DEMS) and Attenuated Total Reflection Fourier Transform Infrared Spectroscopy (ATR-FTIRS). In Handbook of Fuel Cells − Fundamentals, Technology and Applications; Vielstich, W. A, Gasteiger, H., Lamm, A.,Yokokawa, H., Eds.; John Wiley & Sons, Ltd.: New York, 2010; pp 1−32.

6808

dx.doi.org/10.1021/jp411982p | J. Phys. Chem. C 2014, 118, 6799−6808