Oxidation of propenol on nanostructured Ni and NiZn electrodes in alkaline solution

Electrochimica Acta 139 (2014) 345–355

Contents lists available at ScienceDirect

Electrochimica Acta journal homepage: www.elsevier.com/locate/electacta

Oxidation of propenol on nanostructured Ni and NiZn electrodes in alkaline solution Gert Göransson a , Jakub S. Jirkovsky´ b,1 , Petr Krtil b , Elisabet Ahlberg a,∗ a b

Department of Chemistry and Molecular Biology, University of Gothenburg, SE-412 96 Gothenburg, Sweden J. Heyrovsky Institute of Physical Chemistry, Academy of Sciences of the Czech Republic, Dolejskova 3, 18223 Prague, Czech Republic

a r t i c l e

i n f o

Article history: Received 7 January 2014 Received in revised form 23 June 2014 Accepted 30 June 2014 Available online 17 July 2014 Keywords: Electrodeposition DEMS Pulse plating Ni alloys Redox mediated reaction

a b s t r a c t Oxidation of propenol on nanostructured pulse plated Ni and NiZn alloys in alkaline solution gives the corresponding aldehyde, propenal, as the major product. The process is redox mediated by the NiOOH/Ni(OH)2 couple and starts as soon as NiOOH is formed on the surface. Depending on the applied potential, propenol oxidation takes place in parallel with water oxidation. The latter requires a higher surface concentration of NiOOH compared to propenol oxidation, indicating that a binuclear mechanism for oxygen evolution prevails at low overpotentials. The redox reaction of nickel oxide is markedly slower at the NiZn alloy than for solid Ni or pulse plated Ni, which also is reflected in the oxidation of water and propenol. Separation of all three oxidation currents (Ni(OH)2 , water and propenol) shows that pulse plated Ni is more efficient as catalyst than NiZn per NiOOH site, for both water and propenol oxidation. The role of Zn ions in the mixed NiZn hydroxide is discussed and a mechanism for propenol oxidation is suggested. © 2014 Elsevier Ltd. All rights reserved.

1. Introduction Selective oxidation of light alcohols in an environmentally sustainable as well as efficient way is of great interest and importance, both academically and industrially [1]. Many of the industrial oxidation processes used today include environmentally hazardous halogenated reactants and/or generate stoichiometric amounts of less valuable co-products [2,3]. In the light of green chemistry and cost efficiency, the search for new catalysts functioning in nonhalogenated aqueous solutions is of great importance. From the above perspective, heterogeneous electrochemical catalysts may offer an attractive alternative to many of the oxidation methods used today. The industrially important oxidation of propene to propene oxide is an example of a process that to a large extent still suffers from these environmental (chlorohydrine process) and co-product (oxirane process) obstacles, but is finally beginning to change. The well-studied HPPO (Hydrogen Peroxide Propene Oxide)-technology has since a few years been implemented in a production line in Antwerp, Belgium from which propene oxide is produced with, according to the company, drastically reduced

∗ Corresponding author. E-mail address: [email protected] (E. Ahlberg). 1 Argonne National Laboratory, Materials Science Division, Lemont, IL-, USA http://dx.doi.org/10.1016/j.electacta.2014.06.169 0013-4686/© 2014 Elsevier Ltd. All rights reserved.

environmental risks together with cost reduction [4]. The HPPOtechnology utilises the epoxidation ability of hydrogen peroxide by letting separately produced H2 O2 and propene react over a titanium silicate TS-1 catalyst [5–9]. A number of papers have also explored the HPPO technology and possibility to produce H2 O2 in-situ, using TS-1 mixed with a second material suitable for oxygen reduction to H2 O2 [10,11]. Epoxidation by the HPPO-technology is however an indirect process in the sense that the epoxide is formed by a chemical reaction between hydrogen peroxide and propene. In a direct epoxidation process the alkene is oxidised electrochemically to form the corresponding epoxide. The anode material needs to be catalytic with respect to breaking and oxidation of the double bond by an applied potential. Such system would ideally be very clean and the products easy to refine. However, only a few papers can be found reporting direct electrochemical epoxidation of propene on silver, steel, nickel and palladium electrodes [12–16]. None of these methods have yet, to our knowledge, changed the industrial epoxidation processes or even been reproduced in laboratory scale by others. Nickel and nickel alloys are interesting catalysts due to their durability and low cost, and have been investigated in the literature as possible electrocatalysts for methanol [17–20] and ethanol oxidation [21]. However, only few studies have been made with more complicated carbohydrates and with other applications in mind, e.g. electrochemical processing [22–25]. Even though it is well known since the beginning of the twentieth century that

346

G. Göransson et al. / Electrochimica Acta 139 (2014) 345–355

Table 1 Plating bath and parameters for galvanic square wave pulse plating. Chemicals

␻

C −1

Ni(pp) NiZn

NiSO4 ·6H2 O Na2 SO4 NiSO4 ·6H2 O ZnSO4 ·7H2 O Na2 SO4

ton

toff

ip −2

Emax a

Emax b

mol L

rpm

s

s

mA cm

V

V

0.30 1.2 0.30 0.028 1.2

200

0.66

7

28.4

-1.2

-1.1

1600

0.66

7

187.5

-1.6

-1.3

pH = 2.8, 40 pulse cycles for the standard RDEs, ␻ is the rotation rate of the RDEs. a resulting potential. b IR (23 ) compensated resulting potential.

organic species can be oxidised on nickel electrodes, there are only few papers in which the authors really try to deduce the mechanism behind the oxidation of organic alcohols and amines [26–29]. Fleischmann et al. was the first to suggest the oxidation process of various organic species to be redox mediated by the surface NiOOH species formed at positive potentials in alkaline electrolytes [26]. In the present work the oxidation of propenol, propanol and propene have been studied on Ni and NiZn alloys in alkaline solution with focus on propenol. Cyclic Voltammetry (CV) and Chronoamperometry (CA) in combination with Differential Electrochemical Mass Spectrometry (DEMS) have been used for mechanistic investigations and product analysis. The advantage of using DEMS in combination with traditional electrochemical analysis methods is that it offers an in-situ real time detection of volatile species by an online connected mass spectrometer [30]. 2. Experimental 2.1. The electrochemical setup The CV and CA experiments were performed in a onecompartment three electrode cell using a Ag/AgCl reference electrode (saturated KCl, E = 197 mV vs. she, Metrohm) in a doublejunction setup (containing 0.1 M KOH in the outer junction) and a platinum (Pt) mesh as the counter electrode. The DEMS measurements were performed in a 5 mL Kel-F single compartment cell with a Ag/AgCl wire as reference electrode and a Pt wire as counter electrode. The time constant of the cell, i.e. the time lag from the solution to the vacuum side, was in the order of 0.1 s. The cell construction was the same as described in previous work [31]. 2.2. Preparation of working electrodes The working electrodes used were solid Ni (99.999%, Puratronic®) and pulse plated Ni and NiZn on a Pt substrate. All the electrodes were cast in epoxy and functioned as rotating disc electrodes (RDE). The solid nickel electrode (0.1964 cm2 ) was pre-treated in the following order: ground on wet silicon carbide paper (#1000 to #4000, Struers), polished with diamond paste (6, 3 and 1 ␮m, Struers), and rinsed with ethanol and MilliQ (Millipore) water in an ultrasonic bath. The Ni and NiZn electrodes were made through galvanostatic square wave pulse plating (pp) on a Pt RDE (0.0707 cm2 ), which was pre-treated through polishing with diamond paste (6 ␮m, Struers) and ultrasonically cleaned in ethanol and MilliQ water. The pulse plating was performed using the parameters in Table 1. Subsequently, a potential sweep from -0.5 V to 0.5 V at 10 mVs−1 was performed on the freshly plated NiZn electrode directly in the plating solution. The pulse plated electrodes were potentiostatically cycled with 44 sweeps between 0.05 and -0.90 V in O2 saturated 0.1 M KOH at 10 mVs−1 . All electrodes were potentiostatically cycled within the water oxidation/reduction range in 0.1 M KOH until a steady response

from the surface Ni(OH)2 /NiOOH redox couple was achieved (approximately 40 cycles). The working electrodes for the DEMS cell consisted of several pulse plated disc shaped Pt meshes with a diameter of 8.0 mm and a real surface area of approximately 0.3-0.4 cm2 (estimated by surface PtOy reduction) [32]. The meshes were washed in Aqua Regia, cleaned by flame annealing, and rinsed in MilliQ water, followed by the pulse plating of the Ni and NiZn layers as described for the RDE. When the current density for the electrodeposition was calculated from the area approximation it turned out to be too high (manifested in a massive hydrogen evolution). The perspective of the plating parameters had to be changed from the current density to a potential limited parameter, however, still galvanostatically controlled. The current density was adjusted so that the maximum plating potential at the first cycle (-1.5 V for NiZn and -1.15 V for Ni) remained the same as during the plating of the RDE. The second discrepancy from the method used for the RDE was the convection that in the case of the mesh was controlled by stirring the solution and keeping the electrode fixed. 2.3. Chemicals and instrumentation The electrolyte was prepared from MilliQ water and potassium hydroxide (KOH, 99.99% semiconductor grade, Aldrich); the concentration of KOH was 0.1 M and a volume of 100 mL was used in each experiment. The KOH electrolyte was saturated with nitrogen for at least 45 minutes prior to each experiment and an inert atmosphere over the electrolyte was retained by an overpressure of N2 . The propenol (prop-2-en-1-ol, 99%, Alfa Aesar) was added successively between each propenol oxidation experiment from 0.01 to 0.1 and 1 M concentrations in the de-aerated electrolyte. The potentiostats used for CV and pulse plating (pp) were an Autolab PGstat 12 (Metrohm) and a PAR 273A. The electrochemistry in the DEMS experiments was controlled by a PAR 263A potentiostat and the chemical analysis apparatus consisted of a Prisma TM QMS200 quadrupole mass spectrometer (Balzers) connected to a TSU071E turbo molecular drag pumping station (Balzers). 3. Results and Discussion 3.1. Surface characterisation For the CV and CA measurements in this study pulse plated Ni (ppNi) and NiZn alloys on Pt disks were used. The morphology and composition of these materials have been previously studied [33] and the short range structure has been determined by X-ray Absorption Fine Structure (EXAFS) in [34]. The characterisation of the ppNi and NiZn electrodes indicated an amorphous Ni layer and a considerably thicker nanostructured NiZn layer with two structural motives. The analysis also showed that pulse plated NiZn has a rougher surface that appears to be porous, compared to the smoother ppNi layer. The alloy composition of NiZn used in this study was determined by EDX to consist of 80 ± 2 at% Ni [33].

G. Göransson et al. / Electrochimica Acta 139 (2014) 345–355

347

Fig. 1. SEM images of pulse plated NiZn mesh (a) 20 kV, 9 mm x40k (b) 20 kV, 9 mm x200k. (c) EDX spectra (acc.V: 10 kV).

For the electrodes used in the DEMS investigations, a Pt mesh was used as substrate. Although the intention was to keep all of the deposition parameters the same as for the RDE electrodes this was not possible in the case of convection and current density as described in Section 2.2. SEM and EDX analyses of the Pt meshes pulse plated with NiZn were made, see Fig. 1, and showed some variation of the surface composition with an average of 70 ± 5 at% Ni. The SEM images of the NiZn plated mesh show a fairly rough surface at the ␮m level due to the roughness of the Pt mesh. On the nanometer scale the surface is built up of semi-spherical

a

particles of about 60-100 nm on which smaller particles of 1020 nm have started to grow (Fig. 1a,b). On this scale, the plated RDEs [33] and meshes have similar morphology with equally sized nanostructures. 3.2. Cyclic voltammetry in the absence and presence of propenol Cyclic voltammetry (CV) was performed on the different electrode materials prior to addition of the organics to generate information about the clean surfaces, Fig. 2. As previously reported

b

0.4

1.5

1 0.2

j /mAcm-2

j /mAcm-2

0.5

0

0

-0.5 -0.2 -1

-0.4

-1.5 0

0.2

0.4

0.6

0

E vs. Ag/AgCl /V Fig. 2. CV in 0.1 M KOH at 10 mVs−1 . (a)

0.2

0.4

0.6

E vs. Ag/AgCl /V Solid Ni disc and

ppNi on Pt disc. (b)

ppNi on Pt disc and

NiZn on Pt disc.

348

G. Göransson et al. / Electrochimica Acta 139 (2014) 345–355

[35–37] the clearly recognized peak couple around 0.5 V (positive) and 0.4 V (negative) can be ascribed to Ni(OH)2 oxidation to NiOOH and the reversed reduction, respectively, Eqn 1, where the standard potential comes from reference [38].

3.5

NiOOH(s) + H + + e− ↔ Ni(OH)2 (s) E 0 = 0.48V

2.5

a 3.0

Eqn. 1

j /mAcm-2

2.0

1.5 1.0 0.5 0.0

-0.5

3.5

0.2

0.3

0.4

0.5

0.6

0.7

0.3

0.4

0.5

0.6

0.7

0.4

0.5

0.6

0.7

b 3.0 2.5

j /mAcm-2

2.0 1.5 1.0 0.5 0.0 -0.5 0.2

c

14 12 10 8

j /mAcm-2

The average between the position of the oxidation and reduction peaks marks the potential of the Ni(OH)2 /NiOOH redox couple. From Fig. 2a it can be seen that the CV of pulse plated nickel (ppNi) is comparable to solid nickel (Ni(s)) in the electrochemical response. Small differences in the redox potential for the Ni(OH)2 /NiOOH redox couple and for the onset of the O2 evolution can arise between different electrodes of the same material due to differences in the complex surface oxide. In these measurements, the reduction potential of NiOOH is generally more consistent than the potential for oxidation of Ni(OH)2 and O2 evolution, which has also been observed previously by Visscher and Barendrecht [39]. The characteristic response of the Ni(OH)2 /NiOOH redox couple for pulse plated Ni and NiZn is very similar (Fig. 2b). However, the morphology and porosity of the NiZn alloy generates a much larger real surface area, manifested in a current response that is 5 times larger than for ppNi. Oxidation of propenol was investigated on Ni(s), ppNi and NiZn electrodes (Fig. 3). Propenol was added sequentially (0.01 M ( ), 0.1 M( ) and 1 M ( )) to the 0.1 M KOH electrolyte and the current increases for each concentration at the potential of Ni(OH)2 oxidation. The CV response due to the addition of propenol from the three electrodes are similar, however the magnitude of the currents differ and is likely to be connected to the larger real area of the NiZn electrode. The current increase may be interpreted either as enhanced oxygen evolution at a less positive potential, propenol oxidation or both. It has previously been reported on BDD (Boron Doped Diamond) that the presence of alcohols can lower the overpotential for O2 evolution [40]. The sweep rate dependence is complex since several reactions take place in the same potential region, Fig. 4. The current for water oxidation is practically the same for all sweep rates used, Fig. 4a, which indicate that it is independent of mass transport and controlled by kinetics. The onset of oxygen evolution takes place well beyond the redox peak and indicates that more than one Ni(III) site is required for oxygen evolution to take place, in line with a binuclear mechanism for oxygen evolution, or that Ni(IV) is the active site [41,42]. The Tafel slope is ∼ 60 mV at low overpotentials and about 150 mV at higher overpotentials, which may reflect an increasing amount of active sites until a full surface coverage is obtained. For the Ni(OH)2 /NiOOH redox couple the current is proportional to the sweep rate, illustrated for NiZn in Fig. 4b where the current has been divided by the sweep rate. The peak separation increases with increasing sweep rate and was found to be 74 mV, 90 mV and 195 mV for 2, 10 and 100 mVs−1 , respectively. The increase in peak separation can be explained either by slow kinetics of the Ni(OH)2 /NiOOH couple or by a resistive component related to the hydroxide layer. For ppNi and Ni(s) the peak separations are smaller, 109 mV (ppNi) and 101 mV (Ni(s)) at 100 mVs−1 , showing that the presence of Zn in the hydroxide matrix slows down the redox reaction of Ni hydroxide or increases the resistance of the oxidised layer. In the presence of 10 mM propanol the voltammetric response resemble that in the absence of propanol, compare Fig. 4a and 4c. However, normalising with the sweep rate clearly shows some differences, compare Fig. 4d with 4b. At the lowest sweep rate the oxidation of Ni(OH)2 is masked by the oxidation of propenol and the reduction peak for NiOOH is small. With increasing sweep rate more of the Ni(OH)2 /NiOOH redox couple is visible. The fact that the oxidation of propenol starts at the onset of Ni(OH)2 oxidation shows

6 4 2 0 -2 0.2

0.3

E vs. Ag/ AgCl /V Fig. 3. CV in 0.1 M KOH at 10 mVs−1 with sequential addition of propenol, 0 M, 0.01 M, 0.1 M and 1 M propenol. (a) Solid Ni disc. (b) ppNi on Pt disc. (c) pp NiZn on Pt disc.

that Ni(III) is the active site in the reaction sequence as first reported by Fleischmann et al. [26,27]. The visibility of the reduction peak for surface NiOOH depends on both propenol concentration and sweep rate. For 0.1 M propenol it disappears at 2 mVs−1 , for 1 M at 10 mVs−1 , the reduction peak is, however, present for all

G. Göransson et al. / Electrochimica Acta 139 (2014) 345–355

15

349

200

a

b

150

10

C/mFcm-2

j/mAcm-2

100 5 0

50 0 -50

-5

-100

-10

-150 0.2

0.3

0.4

0.5

0.6

0.7

0.3

0.2

E vs. Ag/AgCl/V

0.5

0.6

0.7

0.6

0.7

E vs. Ag/AgCl/V

15

1 000

d

c 800

C/mFcm-2

10

j/mAcm-2

0.4

5 0 -5

600 400 200 0

-10

-200 0.2

0.3

0.4

0.5

0.6

0.7

E vs. Ag/AgCl/V

0.2

0.3

0.4

0.5

E vs. Ag/AgCl/V

Fig. 4. CV on NiZn in 0.1 M KOH at 100 mVs−1 , 10 mVs−1 and 2 mVs−1 . (a) and (b) current density and capacitance in the absence of propenol, (c) and (d) current density and capacitance in the presence of 10 mM propenol. The capacitance was calculated from the current density divided by the sweep rate. The dashed lines represent the positive to negative sweep direction.

concentrations at 100 mVs−1 . The sweep rate dependent disappearance of the NiOOH reduction peak has been observed previously for Ni electrodes in the presence of alcohols and amines [18,26]. These results indicate that the propenol oxidation on NiZn is not a simple electron transfer reaction and that the reaction rate is slow. The different current contributions to the cyclic voltammograms in Figs. 3 and 4 will be further discussed and elucidated in Section 3.3 and 3.6.

3.3. Separation of the oxidation reactions To better understand the propenol oxidation on nickel containing electrodes the three oxidation reactions (oxidation of Ni(OH)2 , H2 O and propenol) that are all active in the same potential region (above 0.4 V) need to be separated. To do this the first step was to determine the contribution of Ni(OH)2 oxidation to the overall current. This is fairly straightforward in electrolyte without propenol since only the second half of the oxidation peak is influenced by water oxidation as can be seen in Fig. 4a. Background currents not belonging to the redox couple were subtracted. In the presence of propenol, the charge of the NiOOH reduction peak was then used to determine the contribution from Ni(OH)2 oxidation to the overall oxidation charge. To separate the water oxidation from the propenol oxidation a thorough investigation of the O2 reduction currents (below 0.11 V), with and without addition of propenol, on ppNi and NiZn

was undertaken. In a potential sweep on the pulse plated electrodes, without propenol, there is a substantial amount of O2 produced at large positive potentials that is reduced during the subsequent negative going sweep (Fig. 4 and 5). However, as can be seen in Fig. 5, after the sequential addition up to 1 M of propenol the charge from reduced O2 is negligible. A detailed study of this phenomenon shows that if we include both the kinetic and diffusive contribution of O2 reduction during the sweep below -0.15 V (both negative and positive sweep directions), 88% and 42% of the O2 produced anodically is reduced on NiZn and ppNi, respectively, see also supplementary information. This information can then be used to calculate the amount of water oxidation also in the presence of propenol by the use of the oxygen reduction peak at different concentrations in relation to electrolyte without propenol. For these calculations the ratio of the O2 reduction products (H2 O2 and H2 O) used, were based on a previous study showing that the same ppNi and NiZn alloys gave 60% and 90% H2 O2 , respectively [33]. The large extent of O2 reduction on NiZn also indicates that the surface of the NiZn layer differs significantly from the surface on ppNi. If the charge from the oxidation of Ni(OH)2 to NiOOH is assumed to reflect available sites for water oxidation a comparison in oxidation efficiency between ppNi and NiZn can be calculated. The potential window for water oxidation is not exactly the same for both materials and therefore the charge from the O2 evolution reaction was calculated during 20 s at 10 mVs−1 from the onset potential of water oxidation at each material, i.e 0.55→0.65→0.55 V for NiZn and 0.50→0.60→0.50 V for Ni(pp). From this comparison it can be

350

G. Göransson et al. / Electrochimica Acta 139 (2014) 345–355

a

b

0

0

j /mAcm-2

j /mAcm-2

-0.2

-0.1

-0.4

-0.6

-0.8

-1

-0.2 -0.5

-0.4

-0.3

-0.2

-0.1

-0.5

0.0

-0.4

-0.3

E

E vs. Ag/AgC l /V Fig. 5. CV in 0.1 M KOH at 10 mVs−1 , showing negative sweep direction, on Pt disc.

no propenol,

seen that ppNi is the more efficient water oxidation catalyst per site NiOOH by roughly 2.5 times compared with NiZn. The third oxidation reaction, the 2 e− oxidation of propenol to propenal (Section 3.5), is simply what is left of the oxidation current after the subtraction of the background, the nickel hydroxide oxidation and the water oxidation. In the evaluation of the propenol oxidation it was observed that the NiOOH reduction decreases linearly with the logarithm of propanol concentration. The ratio between the number of O2 molecules produced and the number of NiOOH sites in the absence and presence of the different propenol concentrations was calculated (Table 2). From Table 2 it can be seen that the ratio of O2 and NiOOH without propenol does not differ very much (given that this type of calculations will include some experimental fluctuations) from the overall average. This indicates that the peak corresponding to NiOOH reduction can be seen as a reflection of “free” NiOOH sites in the presence of propenol on which the water oxidation can proceed as without propenol. In the presence of 0.01, 0.1 and 1 M propenol, the propenol oxidation on NiZn and (ppNi) increases from 69 (77) to 88 (89) and to 99.6% respectively, in relation to the overall water and propenol oxidation (Fig. 5 and Table 3). The number of propenol molecules oxidised per NiOOH site that is “blocked” (not seen in the NiOOH reduction peak in the presence of propenol) is 22 on NiZn and 30 on Table 2 Moles of oxygen produced for each mole of NiOOH available for reduction in a potential sweep with and without propenol in the electrolyte. M/propenol

nO2 /nNiOOH ppNi

nO2 /nNiOOH NiZn

0 0.01 0.1 1 Average 0-1 M

4.4 3.0 3.1 3.9 ± 0.8

2.8 2.3 4.3 2.1 2.9 ± 1.0

0.01 M,

-0.2

-0.1

0

vs. Ag/AgC l /V

0.1 M and

1 M propenol. (b) pp Ni on Pt disc. (c) pp NiZn

ppNi and is constant at all concentrations (Table 3). These results clearly show that the large increase in the oxidation current is due to propenol oxidation and also that oxygen evolution is suppressed in the presence of propenol. The fact that the current response does not depend directly on the concentration suggests a more complicated reaction mechanism than a simple electron transfer reaction. The inhibition of the OER indicates that the surface is blocked by adsorption of propenol or a reaction intermediate. Since the number of O2 molecules produced on “free” NiOOH sites and the number of propenal molecules produced at “blocked” sites are constant for all concentrations, the mechanism is likely site specific. The results from this analysis also show that ppNi is more active than NiZn counted per NiOOH site for both oxygen evolution and propenol oxidation. 3.4. Chronoamperometry CV measurements in the presence of 1 M propenol show larger currents during the negative going scan compared to the positive going scan (Fig. 3). Such an activation of the electrode surface can be explained by oxidation of adsorbed species at high potentials, which then creates a larger number of active sites for the reversed reaction. At high propenol concentrations, the oxidation is sustained to lower potentials in line with the presence of Ni(III) at the surface. This phenomenon has previously been reported for oxidation of alcohols on nickel electrodes [19–21]. To investigate the nature of the increase in current, i.e. if it is a transient phenomenon during potentials cycling or if the current remains high as a function of time, chronoamperometry (CA) was used. The current was measured at three different potentials for the propenol/water oxidation on ppNi (Fig. 6). Each CA measurement was initiated with activation at 0.8 V for 4 s then stepped to one of the three potentials (0.53, 0.54 and 0.55 V, in that order) and the current was measured for 600 s. The results show that the current decreases with time, but

Table 3 Number of NiOOH sites and the fraction of aldehyde formation. M/propenol

NNiOOH ppNi

NNiOOH NiZn

NNiOOH NiZn/ppNi

0 0.01 0.1 1 Avg. 0-1 M

1.79 × 10−8 1.33 × 10−8 8.55 × 10−9 -

8.88 × 10−8 7.25 × 10−8 3.68 × 10−8 3.13 × 10−9

5.0 5.5 4.3

Nald /N(ald+O2) ppNi

Nald /N(ald+O2) NiZn

Nald /NNiOOH* ppNi

Nald /NNiOOH* NiZn

0.775 0.888 -

0.694 0.880 0.996

30.44 29.93 30.2 ± 0.4

23.06 21. 86 21.96 22.5 ± 0.8

4.9 ± 0.6

NiOOH* refers to the difference between NiOOH without and with propenol, calculated from its reduction peak. ald = aldehyde

G. Göransson et al. / Electrochimica Acta 139 (2014) 345–355

-100

400

t /s

900

140 0

␲-bonding may be involved in the surface adsorption also in the case of Ni containing surfaces.

190 0

3.6. DEMS measurements

2.7

2.2

0.55 V

0.55 V

j /mAcm-2

351

1.7

0.54 V

1.2

0 rpm

3000 rpm

0.53 V 0.7

0.2

-0.3 0.2

0.3

0.4

0.5

0.6

E vs. Ag/ AgCl /V Fig. 6. CV (at 10 mVs−1 ) and CA on ppNi in 0.1 M KOH, show the current decay with time at a certain potential and the electrode activation to higher potential in no propenol, 1 M propenol, CA at 0 and the presence of propenol. CA at different potentials proceeded with activation 3000 rpm no activation, to 0.8 V for 4 s, the CA potentials denoted in the anodic sweep direction of the CV.

also that the electrode can be reactivated with an anodic potential step. The potential or time for the reactivation step was not critical but some current efficiency was lost even after the activation. Comparing the CV with CA currents as a function of time shows a fairly good agreement at a scan rate of 10 mVs−1 (Fig. 6). These results show that the blocking of propenol oxidation on the positive going scan is likely due to adsorbed propenol and that the surface gradually changes and gets less active. The change of the surface is permanent and the original voltammetric behaviour with a well separated Ni(OH)2 /NiOOH redox couple cannot be retained with cycling in the absence of propenol, see Supplementary information for details. The lack of a Ni(OH)2 peak shows inhibition of Ni(OH)2 oxidation and activation of oxygen evolution. The reason for enhance oxygen evolution might either be a change in mechanism from binuclear to mononuclear at low overpotentials or the presence of Ni(III) sites that are stabilised by organic fragments. The lack of proper mass transport dependence (blue curve inserted in Fig. 6) also supports a site specific mechanism involving adsorbed species. 3.5. Propene and propanol oxidation on ppNi and NiZn Despite previously reported direct epoxidation on Ni and other metal electrodes [12–16], no propene oxidation was observed on ppNi and NiZn electrodes cycled in a propene saturated 0.1 M KOH. The oxidation of propanol takes place approximately at the same potential as propenol on both Ni and NiZn. A notable difference is that propenol shows higher current on the negative going sweep compared to propanol (Supplementary information Fig. S2). The results from propene and propanol oxidation clearly show that the double bond does not take active part in the oxidation process and it is the OH-group that is oxidised during the experimental conditions used in this work. For Pt electrodes ␲-bonding is believed to take part in the surface adsorption [43]. The strong adsorption observed for propenol on Ni and NiZn shown in Figure S2 indicates that

Despite the efforts described above, the electrochemical approach alone does not provide conclusive information on the product distribution of the oxidation process. The complex nature of the propenol oxidation can be characterized in greater depth by a combination of cyclic voltammetry and differential electrochemical mass spectrometry (DEMS). DEMS data characterizing the anodic behaviour of pulse plated Ni and NiZn electrodes in absence and presence of 0.1 M of propenol are summarized in Fig. 7. The CV and the mass response are plotted together to visualize the recorded current and the simultaneously recorded mass fraction abundances of the two products observed, propenal (m/z 56, green diamonds) and oxygen (m/z 32, red dots). The mass data were normalized to a zero baseline and smoothened. In pure 0.1 M KOH, for both ppNi and NiZn (Fig. 7a and b), it was observed that oxygen evolution starts once the surface is at least half covered with Ni(II), i.e. at the peak (0.25 V for ppNi and 0.4 V for NiZn). As discussed in Section 3.2, this indicates a binuclear mechanism for oxygen evolution at low overpotentials. The need for sufficient amounts of Ni(III) on the surface is also illustrated for NiZn, where the oxygen evolution starts 150 mV more positive than on Ni due to the slower kinetics for Ni(OH)2 oxidation in the NiZn matrix. After the addition of 0.1 M propenol, water oxidation is delayed with about 60 mV for ppNi and 120 mV for NiZn, Fig. 7c and d. The presence of propenol has a pronounced effect on the nature of the catalytic process represented in the measured current. The added propenol changes the voltammetric behaviour in the same way as in the standard CV procedures described in section 3.2. The observed anodic process involves at least two different electrode reactions leading to formation of the corresponding aldehyde, propenal (acrolein), reflected in a potential dependent increase of abundances for fragments with mass to charge ratio (m/z) of 56, 27, 26 and oxygen (m/z 32). The formation of both reaction products is complemented with a decrease of the DEMS signal for propenol (m/z 57). No fragments attributable to higher oxidation products like acids, esters or carbon dioxide were detected. It could be argued that detection of these products may be hindered by low volatility (2-propenoic acid) or by immobilization of the products by a follow-up chemical reaction (polymers, acids, carbonates). Calibration experiments were, however, made with 2-propenoic acid using the corresponding DEMS setup as for the propenol oxidation. The results showed small but detectable fragments of the acid in both neutral and alkaline media. The detected reaction product was propenal for both Ni and NiZn but with a slightly different potential dependence (Fig. 7). The potential range active for propenol oxidation can be divided into three different regions. In the first region (from 0.16 to 0.31 V for ppNi and 0.21 to 0.52 V for NiZn) propenal represents the only reaction product. In the second region (from 0.31 to 0.38 V for ppNi and 0.52 to 0.58 V for NiZn) O2 evolution starts and both propenal and oxygen is formed as a function of increasing potential. In the third region (from 0.38 to 0.60 V for ppNi and 0.58 to 0.60 V for NiZn) the oxygen response continues to increase with increasing potential but the propenal response levels off and becomes more or less potential independent. For NiZn the propenal response does not clearly level off as for Ni, see Fig. 7c and d, but at potentials above 0.58 V the signal declines slightly. Taking into account the small amount of oxygen produced at this potential one can expect a similar behaviour as for Ni. Additional experiments to higher potentials confirmed that this was the case. The DEMS experiment clearly shows that the oxidation of propenol starts immediately after the

352

G. Göransson et al. / Electrochimica Acta 139 (2014) 345–355

Fig. 7. DEMS on ppNi and NiZn in 0.1 M KOH at 10 mVs−1 , the mass spectrometer responses were carefully smoothened and the potentials were fitted to a Ag/AgCl (KCl) reference electrode. (a) pp Ni in KOH (b) NiZn in KOH, (c) pp Ni in KOH and 0.1 M propenol, (d) NiZn in KOH and 0.1 M propenol. m/z 32 O2 , m/z 56 propenal.

first formation of the surface NiOOH which has also been seen in the oxidation of methanol on nickel hydroxide electrodes [18]. The permanent change of the surface of NiZn after oxidation in propenol containing electrolyte, Supplementary information, was further explored in a DEMS experiment. A NiZn electrode used in an oxidation experiment with 0.1 M propenal was directly transferred into clean electrolyte and a potential sweep was made in the DEMS setup. Propenal was detected at 0.32 V and O2 at 0.42 V in a similar behaviour as with propenol present in the solution (Supplementary information Fig. S1). This result clearly shows that there is some propenol adsorbed and that the same oxidation product is formed from adsorbed propenol as for propenol in solution. To investigate the potential dependence of propenol oxidation further, potential step measurements were employed, Fig. 8. The potential was stepped from 0.2 V (240 s) to 0.4 V (240 s) to 0.5 V (240 s) and finally to 0.65 V (120 s), repeated four times. From start at 0.2 V no electrochemical reaction takes place. The exponential

decrease seen in the mass signals (Fig. 8) during the initial 500 s is due to levelling of pressure misbalances and not due to any electrochemical reaction. When the potential is stepped to 0.4 V the formation of propenal (m/z 56) is clearly observed while the signal for oxygen (m/z 32) still decreases. The production of propenal increases when going from 0.4 to 0.5 V and oxygen starts to evolve. During the last step to 0.65 V, oxygen evolution increases drastically while the propenal formation remains constant. The consumption of propenol (m/z 31) is harder to follow in the first step sequence due to the steady decrease of the signal and the small changes observed when propenol is oxidised. The intensity of the propenol depletion is significantly weaker than the increase in the signal for the reaction product, propenal. This is due to differences in boiling point and thereby the vapour pressure for the two compounds. The boiling point for propenol and propenal is 97 and 53 ◦ C, respectively, and the vapour pressure differs by one order of magnitude. During the following sequences of steps the signals are stabilised

G. Göransson et al. / Electrochimica Acta 139 (2014) 345–355 0.75

R CHO

II

Ni (OH ) 2

0.5

0.65

0.4

0.55

0.3

0.45

R H 2C O Ni OOH

0.2

0.35

e + H 2O

0.1

0.25

0

0.15

E /V

Im/z /a.u.

0.6

353

OH

e + H 2O

III

III

Ni OOH R CH2OH

OH 0

500

1000

1500

t /s

2000

2500

II

R H 2 C O Ni (OH)2

Rearrangement

Fig. 8. DEMS on NiZn in 0.1 M KOH in a potential step experiment, the mass specm/z 32 O2 , trometer response (MS) was recounted into arbitrary units. E/V, m/z 56 propenal, m/z 31 propenol.

and the changes due to the oxidation reactions are clearly seen. The mass response of oxygen after the 0.65 V plateau is likely to be a consequence of the massive O2 evolution which can create bubbles close to the membrane, thereby causing a delay in the signal. Also CO2 (m/z 44) and propenoic acid (m/z 45, 72) as possible products were investigated but no response from these fragments was recorded. Similar experiments were performed with propanol to elucidate any differences in oxidation products compared with propenol. The corresponding oxidation product for both alcohols was detected by DEMS (propenal and propanal) as well as the inhibition of the OER in the presence of the alcohol. The same DEMS setup was also tested for propene oxidation but as expected from the voltammetric results no oxidation of propene was observed. Formation of propenal has been reported for propenol oxidation on gold and platinum in acidic media [44–46]. The formation of propenal was attributed to the oxidation of propenol in solution while the oxidation of adsorbed species was supposed to form CO2 [44–46]. This is in contrast to what was found in the present work where the DEMS results from a used electrode in pure 0.1 M KOH clearly shows that propenal is the only product formed from oxidation of adsorbed propenol (Supplementary information Fig. S1). 3.7. The role of Zn It is clear that the Ni(OH)2 /NiOOH redox couple has a decisive role for both oxygen evolution and alcohol oxidation and that the formation of Ni(III) sites marks the start for the alcohol oxidation while oxygen evolution requires at least half coverage of Ni(III). The question is then what the role of zinc is in the NiZn alloy. Firstly, zinc is critical in the plating of the unique structure and morphology of this nickel rich alloy. The higher amount of nickel available for NiZn in the Ni(OH)2 /NiOOH redox process increases the number of active sites (or real surface area) by a factor of five. As the propenal oxidation seems to be dependent on the NiOOH species the pulse plated NiZn surfaces generate a high catalytic efficiency per electrode geometric area. Also, the hardness [47] and the stability in alkaline environment [33] makes the alloy a durable catalyst. Secondly, Zn changes the properties of the oxidised Ni sites in a way that makes the nickel oxidation more sluggish. When it comes to water oxidation on Ni it has been suggested that the effective oxidation state needs to approach Ni(IV) for the oxidation to proceed [41,48], or that adjacent Ni(III) sites are formed that allow a combination reaction [41,42,49]. It seems like Zn hinders Ni from reaching the higher

II

R H 2 C(H) O Ni OOH

Scheme I. Redox mediated oxidation of propenol.

oxidation state or hinders the formation of adjacent Ni(III) sites and thereby inhibits the oxygen evolution. The special properties of NiZn also make the propenol oxidation to proceed over a wider potential range before the O2 evolution starts. 3.8. Product distribution and reaction mechanism A general mechanism for oxidation of primary alcohols on nickel anodes in alkaline solution was proposed in early 1970’s by Fleischman et al. [26]. Oxidation of primary alcohols was suggested to be mediated by the Ni(II)/Ni(III) redox couple and it was proposed that the final product was the corresponding acid. For primary and secondary butanol the formation of the corresponding acid was confirmed by FTIR [23]. In the present study, where cyclic voltammetry and differential electrochemical mass spectrometry were used to investigate the oxidation of propenol, the only oxidation product found was the corresponding aldehyde, propenal. Propenal has also been found on other electrode materials such as Pt [45], Pd [50] and Au [44] in acid solutions, together with formation of carbon dioxide. However, in all cases of detected aldehyde the method of analysis has been DEMS, and when the corresponding acid has been detected the method of analysis has been gas chromatography (GC) [26,27] or infrared spectroscopy (FTIR) [23]. An important difference between these methods is that in DEMS the products are immediately extracted from the cell. In other types of analysis the products are left to react further, chemically or electrochemically, in the cell for shorter or longer periods. Experiments by Pastor et al. in acid electrolyte indicate that this can lead to different product distributions for propenol oxidation on Pt [51]. In the present work alkaline solutions were used and the immediate removal of propenal hinders the well-known polymerisation that takes place in strongly alkaline solutions. Also, the decrease in near surface pH due to oxygen evolution may help in avoiding polymerisation. The results herein show that also for propanol oxidation the aldehyde is exclusively formed. The lack of any indication (in CV or DEMS) of oxidation of propene is surprising since others have shown this to be possible. Good yield of propene oxide on both steel and Ni in alkaline electrolyte has been reported [13,14,16]. In analogy with previous studies [26,27] a redox mediated mechanism is suggested for propenol oxidation including an intermediate formation of a surface nickel ester, Scheme I. 1. Oxidation of nickel (II) hydroxide II

III

Ni(HO)2 + OH − → NiOOH + e− + H2 O

(E1)

354

G. Göransson et al. / Electrochimica Acta 139 (2014) 345–355

2. Adsorption of propenol by charge transfer between the alcohol and the surface group. One free electron pair on the oxygen in the alcohol group is added to Ni(III), which will be reduced to Ni(II). Oxygen will be “positively charged” and a proton will immediately leave or be incorporated in the surface hydroxide, see (C3). III

II

⊕

NiOOH + RCH2 OH → HOONi − OHCH2 R

II

II

⊕

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.electacta. 2014.06.169. (C3)

4. Oxidation of the intermediate. II

III

(HO)Ni − OCH2 R + OH − → HOONi − OCH2 R + e− + H2 O

(E4)

5. Formation of the oxygen-carbon double bond and release of a proton, which will be incorporated in the Ni(OH)2 structure. This step closes the redox cycle, Scheme I. III

II

HOONi − OCH2 R → Ni(HO)2 + RCHO

This work was supported by the European Union project “Nanostructures for Energy and Chemicals Production” (NENA, Contract No. NMP3-CT-2004-505906).

(C2)

3. Formation of an intermediate surface ester. HOONi − O(H)CH2 R → 2 (HO)Ni − OCH2 R

Acknowledgements

(C5)

Step C5 demonstrates why the NiOOH reduction peak disappears, depending on the sweep rate and concentration of propenol (Section 3.2).

4. Conclusions • DEMS measurements clearly show that for oxidation of propenol and propanol in alkaline solution the corresponding aldehydes are exclusively formed. No propene oxidation could be detected. • Oxygen evolution in the absence of propenol starts at potentials where more than half of the surface is converted to Ni(III). This was discussed in the context of a binuclear mechanism for oxygen evolution. In the presence of propenol, oxygen evolution is inhibited and takes off when the propenol oxidation starts to level off. The inhibition is likely due to lack of active sites (Ni(IV) or adjacent Ni(III) sites) for oxygen evolution. • CV at different sweep rates shows that the redox reaction of nickel hydroxide is significantly slower at the NiZn alloy than for pure Ni(s) or ppNi. The same behaviour is also reflected in the oxidation of water and propenol. • The plating procedure for NiZn creates a porous alloy, which, after potential cycling in alkaline solution, generates five times higher charge for the Ni(OH)2 /NiOOH redox couple compared with the same redox couple on ppNi. As a consequence a higher total efficiency, normalised to geometric area, is observed for NiZn. However, separation of the three oxidation currents (Ni(OH)2 , water and propenol) shows that ppNi is more efficient than NiZn per NiOOH site, for both water and propenol oxidation. • CA measurements show that the oxidation efficiency in the presence of propenol decreases with time but that the surface can be reactivated by a positive potential step. Activation of the surface at high potentials was also observed in cyclic voltammetry at high propenol concentrations, resulting in significantly higher oxidation current during the reversed scan. These experiments together with DEMS measurements show that adsorbed propenol is oxidised to propenal at high potentials, leaving the surface free for oxidation of propanol in solution. • A redox mediated reaction cycle with a surface nickel-ester intermediate is proposed for the propenol oxidation.

References [1] R.A. Sheldon, J. Dakka, Heterogeneous catalytic oxidations in the manufacture of fine chemicals, Catal. Today 19 (1994) 215–246. [2] R. Sheldon, Organic peroxygen chemistry, in: W.A. Herrmann (Ed.), Topics in current chemistry, vol. 164, Springer-Verlag Berlin Heidelberg, Berlin, 1993. [3] P.T. Anastas, J.C. Warner, Green chemistry: Theory and practice, Oxford University Press New York, 1998. [4] HPPO–a groundbreaking technology for more environmental compatibility and efficiency, in, http://www.basf.com, 2013. [5] M.G. Clerici, G. Bellussi, U. Romano, Synthesis of propylene oxide from propylene and hydrogen peroxide catalyzed by titanium silicalite, J. Catal. 129 (1991) 159–167. [6] I.W.C.E. Arends, R.A. Sheldon, Recent developments in selective catalytic epoxidations with H2 O2 , Top. Catal. 19 (2002) 133–141. [7] M.G. Clerici, Catalytic oxidations with hydrogen peroxide: New and selective catalysts, Heterogeneous Catalysis and Fine Chemicals III (1993) 21–33. [8] C. Perego, A. Carati, P. Ingallina, M.A. Mantegazza, G. Bellussi, Production of titanium containing molecular sieves and their application in catalysis, Appl. Catal A 221 (2001) 63–72. [9] M.G. Clerici, P. Ingallina, Epoxidations of lower olefines with hydrogen peroxide and titanium silicalites, J. Catal. 140 (1993) 71–83. [10] M.G. Clerici, P. Ingallina, Oxidation reactions with in situ generated oxidants, Catalysis Today 41 (1998) 351–364. [11] A. Zimmer, D. Mönter, W. Reschetilowski, Catalytic epoxidation with electrochemically in situ generated hydrogen peroxide, J. App. Electrochem. 33 (2003) 933–937. [12] L.L. Holebrook, H. Wise, Electrooxidation of olefins at a silver electrode, J. Catal. 38 (1975) 294–298. [13] T.-C. Chou, J.-C. Chang, Anodic oxidation of propylene on a screen electrode, Chem. Eng. Sci. 35 (1980) 1581–1590. [14] C.F. Oduoza, K. Scott, Aspects of the direct electrochemical oxidation of propylene, Inst. Chem. Eng. Symp. Series 127 (1992) 37–48. [15] K. Otsuka, I. Yamanaka, Electrochemical cells as reactors for selective oxygenation of hydrocarbons at low temperature, Catal. Today 41 (1998) 311–325. [16] Y.P. Sun, O. Qiu, W.L. Xu, K. Scott, A study of the performance of a sparged packed bed electrode reactor for the direct electrochemical oxidation of propylene, J. Electroanal. Chem. 503 (2001) 36–44. [17] M.A. Abdel Rahim, R.M. Abdel Hameed, M.W. Khalil, Nickel as a catalyst for the electro-oxidation of methanol in alkaline medium, J. Power Sources 134 (2004) 160–169. [18] A.A. El-Shafei, Electrocatalytic oxidation of methanol at a nickel hydroxide:Glassy carbon modified electrode in alkaline medium, J. Electroanal. Chem. 471 (1999) 89–95. [19] S. Maximwitch, G. Bronoel, Oxidation of methanol on nickel-zinc catalysts, Electrochim. Acta 26 (1981) 1331–1338. [20] J. Taraszewska, G. Roslonek, Electrocatalytic oxidation of methanol on a glassy carbon electrode modified by nickel hydroxide formed by ex situ chemical precipitation, J. Electroanal. Chem. 364 (1994) 209–213. [21] J.-W. Kim, S.-M. Park, In situ xanes studies of electrodeposited nickel oxide films with metal additives for the electro-oxidation of ethanol, J. Electrochem. Soc. 150 (2003) E560–E566. [22] S. Berchmans, H. Gomathi, G.P. Rao, Electrooxidation of alcohols and sugars catalysed on a nickel oxide modified glassy carbon electrode, J. Electroanal. Chem. 394 (1995) 267–270. [23] A. Kowal, S.N. Port, R.J. Nichols, Nickel hydroxide electrocatalysts for alcohol oxidation reactions: An evaluation by infrared spectroscopy and electrochemical methods, Catal. Today 38 (1997) 483–492. [24] A.S. Vaze, S.B. Sawant, V.G. Pangarkar, Electrochemical oxidation of isobutanol to isobutyric acid at nickel oxide electrode: Improvement of the anode stability, J. Appl. Electrochem. 27 (1997) 584–588. [25] Q. Yi, J. Zhang, W. Huang, X. Liu, Electrocatalytic oxidation of cyclohexanol on a nickel oxyhydroxide modified nickel electrode in alkaline solutions, Catal. Commun. 8 (2007) 1017–1022. [26] M. Fleischmann, K. Korinek, D. Pletcher, The oxidation of organic compounds at a nickel anode in alkaline solution, J. Electroanal. Chem. 31 (1971) 39–49. [27] M. Fleischmann, K. Korinek, D. Pletcher, The kinetics and mechanism of the oxidation of amines and alcohols at oxide-covered nickel, silver, copper, and cobalt electrodes, J. Chem. Soc. Perkin II 2 (1972) 1396–1403.

G. Göransson et al. / Electrochimica Acta 139 (2014) 345–355 [28] P.M. Robertson, On the oxidation of alcohols and amines at nickel oxide electrodes, mechanistic aspects, J. Electroanal. Chem. 111 (1980) 97–104. [29] G. Vértes, G. Horányi, Some problems of the kinetics of the oxidation of organic compounds at oxide-covered nickel electrodes, Electroanal. Chem. and Interf. Electrochem. 52 (1974) 47–53. [30] H. Baltruschat, Differential electrochemical mass spectrometry, J. Am. Soc. Mass Spectrom. 15 (2004) 1693–1706. ´ H. Hoffmannová, M. Klementová, P. Krtil, Particle size dependence [31] J. Jirkovsky, of the electrocatalytic activity of nanocrystallineRuO2 electrodes, J. Electrochem.Soc. 153 (2006) E111–E118. [32] G. Jerkiewicz, G. Vatankhah, J. Lessard, M.P. Soriaga, Y.-S. Park, Surface-oxide growth at platinum electrodes in aqueous H2 SO4 . Reexamination of its mechanism through combined cyclic-voltammetry, electrochemical quartz-crystal nanobalance, and auger electron spectroscopy measurements, Electrochim. Acta 49 (2004) 1451–1459. [33] G. Göransson, A., Johansson, F., Falkenberg, E. Ahlberg, Characterisation of pulse plated Ni and NiZn alloys, J. Electrochem. Soc. 161 (2014) D476–D483. [34] G. Göransson, M. Peter, J. Franc, V. Petrykin, E. Ahlberg, P. Krtil, Local structure of pulse plated Ni rich Ni-Zn alloys and its effect on the electrocatalytic activity in the hydrogen evolution reaction, J. Electrochem.Soc. 159 (2012) D555–D562. [35] H. Bode, K. Dehmelt, J. Witte, Zur kenntnis der nickelhydroxidelektrode I. Über das nickel II hydroxidhydrat, Electrochim. Acta 11 (1966) 1079–1087. [36] P. Oliva, J. Leonardi, J.F. Laurent, C. Delmas, J.J. Braconnier, M. Figlarz, F. Fievet, A. de Guibert, Review of the structure and the electrochemistry of nickel hydroxides and oxy-hydroxides, J. Power Sources 8 (1982) 229–255. [37] D.J. Singh, Characteristics and effects of NiOOH on cell performance and a method to quantify it in nickel electrodes, J. Electrochem. Soc. 145 (1998) 116–120. [38] A. Bard, R., Parsons, J. Jordan, Standard potentials in aqueous solution, International Union of Pure and Applied Chemistry, Marcel Dekker Inc., New York, 1985. [39] W. Visscher, E. Barendrecht, Investigation of thin-film ␣- and ß-Ni(OH)2 electrodes in alkaline solution, J. Electroanal. Chem. 154 (1983) 69–80. [40] I. Duo, P.-A. Michaud, W. Haenni, A. Perret, C. Comninellisa, Activation of borondoped diamond with IrO2 clusters, Electrochem. and Solid-State Lett. 3 (2000) 325–326.

355

[41] M. Busch, E. Ahlberg, I. Panas, Hydroxide oxidation and peroxide formation at embedded binuclear transition metal sites; tm = Cr, Mn, Fe, Co, PCCP 13 (2011) 15062–15068. [42] M. Busch, E. Ahlberg, I. Panas, Validation of binuclear descriptor for mixed transition metal oxide supported electrocatalytic water oxidation, Catalysis Today 202 (2013) 114–119. [43] J.Y. Gui, B.E. Kahn, C.-H. Lin, F. Lu, G.N. Salaita, D.A. Stem, D.C. Zapien, A.T. Hubbard, Studies of adsorbed unsaturated alcohols at well-defined Pt (111) electrode surfaces by cyclic voltammetry assisted by vibrational spectroscopy (eels), J. Electroanal. Chem. 252 (1988) 169–188. [44] E. Pastor, V.M. Schmidt, T. Iwasita, M.C. Arévalo, S. González, A.J. Arvia, The reactivity of primary c3 -alcohols on gold electrodes in acid media. A comparative study based on dems data, Electrochim. Acta 38 (1993) 1337–1344. [45] E. Pastor, S. Wasmus, T. Iwasita, Dems and in-situ FTIR investigations of c3 primary alcohols on platinum electrodes in acid solutions. Part II. Allyl alcohol, J. Electroanal. Chem. 353 (1993) 81–100. [46] G.M. Pereira, D.M. Jiménez, P.M. Elizalde, A. Manzo-Robledo, N. Alonso-Vante, Study of the electrooxidation of ethanol on hydrophobic electrodes by dems and hplc, Electrochim. Acta 49 (2004) 3917–3925. [47] I. Ahmed, Pulse plating of nanocrystalline nickel, zinc and Ni-Zn multi-layers on steel substrate, in: Report IM-2002-038 Swedish Institute for Metals Research, Stockholm, Sweden, 2002, pp. 67. [48] M.E.G. Lyons, M.P. Brandon, The oxygen evolution reaction on passive oxide covered transition metal electrodes in aqueous alkaline solution. Part 1-nickel, Int. J. Electrochem. Sci. 3 (2008) 1386–1424. ´ M. Busch, E. Ahlberg, I. Panas, P. Krtil, Switching on the electro[49] J.S. Jirkovsky, catalytic ethene epoxidation on nanocrystalline RuO2, J. Am. Chem. Soc. 133 (2011) 5882. [50] M.C. Arévalo, J.L. Rodríguez, E. Pastor, Elucidation of the reaction pathways of allyl alcohol at polycrystalline palladium electrodes, J. Electroanal. Chem. 505 (2001) 62–71. [51] E. Pastor, S. Wasmus, T. Iwasita, Spectroscopic investigations of c, primary alcohols on platinum electrodes in acid solutions. Part I. n-propanol, J. Electroanal. Chem. 350 (1993) 97–116.