Heterogeneous and Homogeneous Routes in Water

DOI: 10.1002/asia.201501446

Full Paper

Electrocatalysis

Heterogeneous and Homogeneous Routes in Water Oxidation Catalysis Starting from CuII Complexes with Tetraaza Macrocyclic Ligands Andrea Prevedello,[a] Irene Bazzan,[a] Nicola Dalle Carbonare,[b] Angela Giuliani,[a] Sunil Bhardwaj,[c] Cristina Africh,[c] Cinzia Cepek,[c] Roberto Argazzi,[d] Marcella Bonchio,[a] Stefano Caramori,[b] Marc Robert,*[e] and Andrea Sartorel*[a]

of copper(II) complexes with two tetraaza macrocyclic ligands, distinguishing heterogeneous or homogeneous processes depending on the reaction media. In an alkaline aqueous solution, and upon application of an anodic bias to working electrodes, an active copper oxide layer is observed to electrodeposit at the electrode surface. Conversely, water oxidation in neutral aqueous buffers is not associated to formation of the copper oxide layer, and could be exploited to evaluate and optimize a molecular, homogeneous catalysis.

Abstract: Since the first report in 2012, molecular copper complexes have been proposed as efficient electrocatalysts for water oxidation reactions, carried out in alkaline/neutral aqueous media. However, in some cases the copper species have been recognized as precursors of an active copper oxide layer, electrodeposited onto the working electrode. Therefore, the question whether copper catalysis is molecular or not is particularly relevant in the field of water oxidation. In this study, we investigate the electrochemical activity

Introduction

the availability of high oxidation states, relevant in oxidation reactions.[15] The first example of a copper based WOC was reported in 2012;[2a] since then, few copper coordination complexes have been reported to be homogeneous, molecular water oxidation catalysts (Figure 1 and Table S1), upon application of an electrochemical anodic bias, in alkaline to neutral aqueous media (pH 7–12).[2–8] Among these reports, a CuII complex with a redox non-innocent tetraanionic amidate ligand displays a very low operating overpotential of 170 mV, albeit at the alkaline pH of 11.5;[5] water oxidation in neutral media was recently accessed with a dinuclear CuII species at 780 mV overpotential.[6] On the other hand, in more recent studies, several copper species were observed to be precursors of an active, heteroge-

In recent years, the development of water oxidation catalysts (WOC) based on transition metals has been the subject of intense investigation, due to their envisaged application in sunlight activated water splitting (WS).[1] Among the Earth-abundant first row transition metals, the investigation of copperbased WOCs is a novel and appealing area of research that is showing a promising potential.[2–12] The interest in copperbased WS is due to the low cost and low toxicity of this element, and the ability to form a large variety of stable complexes with well-defined coordination chemistry. Copper complexes are appealing candidates for water oxidation also because of their established biomimetic O2 chemistry[13–14] and to [a] A. Prevedello, Dr. I. Bazzan, A. Giuliani, Prof. M. Bonchio, Dr. A. Sartorel Department of Chemical Sciences University of Padova and Institute on Membrane Technology via Marzolo 1, 35131 Padova (Italy) E-mail: [email protected]

[e] Prof. M. Robert Universit Paris Diderot Sorbonne Paris Cit Laboratoire d’Electrochimie Molculaire UMR CNRS N8 7591 Btiment Lavoisier, 15 rue Jean de Baı¨f, 75205 Paris Cedex 13 (France) E-mail: [email protected]

[b] N. Dalle Carbonare, Prof. S. Caramori Department of Chemical and Pharmaceutical Sciences University of Ferrara via Fossato di Mortara 17–27, 44121 Ferrara (Italy)

Supporting information and the ORCID identification number(s) for the author(s) of this article can be found under http://dx.doi.org/10.1002/ asia.201501446.

[c] Dr. S. Bhardwaj, Dr. C. Africh, Dr. C. Cepek TASC-INFM National Laboratory S.S. 14 Area Science Park 34012 Basovizza (TS) (Italy)

This manuscript is part of a special issue on energy conversion and storage. A link to the Table of Contents of the special issue will appear here when the complete issue is published.

[d] Dr. R. Argazzi ISOF-CNR c/o Dept. of Chemical and Pharmaceutical Sciences University of Ferrara via L. Borsari, 46-44121 Ferrara (Italy) Chem. Asian J. 2016, 00, 0 – 0

These are not the final page numbers! ÞÞ

1

2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

&

&

Full Paper Results and Discussion The choice of tetraaza macrocyclic ligands for copper is related to their unusual features, in particular: (i) a high binding constant for first row transition metals, by virtue of a tetra-dentate nitrogen-based ligand set;[16] (ii) a square planar coordination motif, with up to five different conformations, leaving the apical coordination sites of the metal available for water binding;[16, 17] (iii) the electron-donating character of the amine coordination sites, prompting to stabilize high valent copper intermediates;[18] (iv) their robustness, under WOC conditions.[19] Moreover, the use of a polydentate nitrogen ligand is expected to stabilize CuII centers, in agreement with literature reports (Figure 1). Cucyclam and CuMe4cyclam are readily synthesized as the perchlorate salts in 46–77 % yield (caution: perchlorate salts of transition metal complexes are potentially explosive and should be handled with care) by stoichiometric reaction of the ligands and Cu(ClO4)2 in tetrahydrofuran (see the Supporting Information).[17] The identity and purity of the complexes was verified by convergent electrospray ionization mass spectrometry (ESI-MS), UV-Vis spectroscopy, and elemental analysis. ESIMS registered under positive ion mode in aqueous solution shows base peaks at m/z = 362 ([CuC10H24N4·ClO4] + ) and at m/ z = 418 ([CuC14H32N4·ClO4] + ) (see Figure S1 in the Supporting Information). In the solid state, the copper atom displays Jahn– Teller distorted octahedral geometry, where the tetradentate ligand is coordinating the four equatorial positions and two perchlorate ions occupy the apical sites of the metal;[17] these latter ones are expected to be labile and susceptible to exchange in solution, in the presence of coordinating species. UV-Vis spectroscopy in neutral aqueous solutions (pH 7) shows broad absorptions ascribable to CuII d–d transitions,[20] at 513 nm (e = 104 L mol 1 cm 1) and 640 nm (e = 233 L mol 1 cm 1) for Cucyclam and CuMe4cyclam, respectively, (see Figure S2 in the Supporting Information), in agreement with literature data.[16, 17, 20] Under an alkaline environment (pH 12) the absorption maxima for the two complexes shift to 507 nm (e = 80 L mol 1 cm 1) and 652 nm (e = 1 228 L mol cm 1); this shift is most likely due to hydroxide coordination to copper.[20] Importantly, the spectra are not changing upon aging the solution after 15 h, indicating that the resulting CuII complexes are stable under the adopted conditions (vide infra). Water oxidation electrocatalysis was initially investigated with Cucyclam by cyclic voltammetry (CV) in alkaline media (NaOH/0.1 m NaOAc, pH 12) employing glassy carbon (GC) and Indium Tin Oxide (ITO) coated glass electrodes (Figure 2). With the GC electrode, an intense anodic current takes off at an onset potential of 0.6 V vs Ag/AgCl, suggesting the occurrence of water oxidation at the working electrode (red trace in Figure 2 top, compared to the black traces observed for the blank experiment, registered in the absence of Cucyclam). However, the first indication of electrodeposition of an active, heterogeneous layer under these conditions comes from the increase of the catalytic current under repeated cycles with ITO electrode (Figure 2 bottom), concomitant to a shift in the onset potential of the wave to lower values, up

Figure 1. Copper based, homogeneous molecular catalysts for water oxidation and precursors of a heterogeneous copper oxide active layer.

neous copper oxide phase, formed by electrodeposition under the application of the anodic bias employed for catalysis (Figure 1 and Table S1 in the Supporting Information).[9–12] Therefore, recognizing whether the catalysis nature is either molecular (homogeneous pathway) or driven by a “solid state” material (heterogeneous pathway) is mandatory for any possible optimization of the WOC system based on copper complexes. In this study, we report the electrocatalytic water oxidation by a CuII species with two tetraaza macrocyclic ligands: 1,4,8,11-tetraazacyclotetradecane (cyclam) and 1,4,8,11-tetramethyl-1,4,8,11-tetraazacyclotetradecane, (Me4cyclam).[16] In particular, the nature of the electrocatalytic process has been investigated under different reaction media (pH 7–12, Scheme 1), distinguishing between heterogeneous and homo-

Scheme 1. Water oxidation electrocatalysis with CuII complexes with tetraaza macrocyclic ligands: Cucyclam and CuMe4cyclam.

geneous pathways. While electrodeposition of an active copper oxide layer is observed in alkaline media (pH 9–12), leading to heterogeneous catalysis, in neutral solutions (0.2 m phosphate buffer, pH 7) catalysis is mediated by the molecular species under homogeneous conditions. Therefore, this media is indicated as the optimal environment to probe structure–reactivity correlations within this class of ligands, to optimize efficiency and stability of the copper(II) catalyst.

&

&

Chem. Asian J. 2016, 00, 0 – 0

www.chemasianj.org

2


ÝÝ These are not the final page numbers!

Full Paper

Figure 3. Electrolysis current (top) and dissolved oxygen measurement (bottom) for CPE experiments at 1.04 V with an ITO electrode immersed in a 1 mm Cucyclam solution (red traces) and the same ITO electrode recovered after the first CPE experiment, immersed in a copper-free solution (blue traces). Experiments in NaOH/0.1 m NaOAc, pH 12; auxiliary electrode: Pt wire, Reference electrode: Ag/AgCl.

Figure 2. Cyclic voltammetry of 1 mm Cucyclam (red traces compared to traces in black registered in the absence of copper species) in NaOH/0.1 m NaOAc, pH 12 with GC (top) and ITO (bottom) working electrodes. Auxiliary electrode: Pt wire, Reference electrode: Ag/AgCl, scan rate: 100 mV s 1.

ments (Figure 3 and Table 1). Indeed, by applying a constant potential of 1.04 V vs Ag/AgCl to an ITO electrode (1 cm2 active geometric surface, this applied potential corresponds to an overpotential of 700 mV) dipped in a deaerated 1 mm Cucyclam solution in NaOH/0.1 m NaOAc, a stationary current of 3.7 mA was registered, with a concomitant dissolved oxygen increase of 0.058 mg L 1 min 1, accounting for an overall 26 % Faradaic yield for oxygen production (red traces in Figure 3, and entry 1 in Table 1). However, a dark coating formed at the ITO surface duringthe CPE experiment (maintained for a total time of 30 min). This recovered ITO electrode was then dipped in a copper-free NaOH/0.1 m NaOAc media, and used in a second CPE experiment under the same above-mentioned conditions. In this case, the stationary electrolysis current and the dissolved oxygen production were 2.2 mA and 0.250 mgl 1 min 1, respectively, with a quantitative Faradaic Yield for oxygen production (blue traces in Figure 3, and values in parentheses in entry 1, Table 1). Therefore, the dark layer electrodeposited at the ITO surface is responsible for electrocatalytic oxygen production, thus confirming a heterogeneous route occurring under these conditions. Concerning the nature of the active layer, X-ray photoelectron spectroscopy (XPS) confirmed the presence of copper oxide,[10] as the typical

Table 1. Controlled potential electrolysis experiments. General conditions: a 1 mm copper species was used; traces in brackets indicated as recovered ITO, refer to experiments where the same electrode employed in the experiment in the presence of the copper species, was reused in a copper-free solution. Working electrode: ITO coated glass slide (8–12 W surface resistivity, 1 cm2 surface area); auxiliary electrode: Pt wire; reference electrode: Ag/AgCl. #

Cu species

pH

Applied E, V

Faradaic Yield

1

Cucyclam (recovered ITO) CuMe4cyclam (recovered ITO) Cucyclam (recovered ITO) Cucyclam (recovered ITO) Cucyclam (recovered ITO) CuMe4cyclam (recovered ITO)

12

1.04

12

1.04

9

1.20

9

1.75

7

1.75

7

1.75

26 ( > 98) – ( ) 75 ( > 98) 88 (92) 78 ( ) 80 ( )

2 3 4 5 6

to 0.6 V vs Ag/AgCl.[9] Electrodeposition of an active copper oxide layer was confirmed by controlled potential electrolysis (CPE) experiments, coupled to dissolved oxygen (DO) measureChem. Asian J. 2016, 00, 0 – 0

www.chemasianj.org


3


&

&

Full Paper show a base peak at m/z = 308 ([CuC10H24N4·HCOO] + ; the formate anion is originated by neutralization of the alkaline electrolysis solution with formic acid before ESI-MS analysis). After 30 min of electrolysis, a new peak was observed at m/z = 306 ([CuC10H22N4·HCOO] + ), consistent with dehydrogenation from Cucyclam (see Figure S5 in the Supporting Information). Further confirmation is that, under the same conditions the electrochemical behavior of CuMe4cyclam was markedly different, as the anodic CV scan did not show any appreciable anodic current (see Figure S6 in the Supporting Information), while CPE experiments failed to produce oxygen and to electrodeposit active layers at the ITO surface (Table 1, entry 2). This confirms that alkylation at the nitrogen centers can increase the stability of the copper complexes, hampering the possibility of imine formation. The steady formation of copper oxide, at pH 12, is in part related to the low solubility product constant of Cu(OH)2 (2.2 10 20).[21] Electrodeposition of the active copper oxide layer with Cucyclam as a source of copper ions was also observed in borate buffer, pH 9, which was recently employed in the investigation of copper-based water oxidation catalysis.[10] Indeed, CV traces of Cucyclam in this buffer reveal the presence of cross-waves when employing ITO as a working electrode (see Figure S7 in the Supporting Information). CPE experiments (see Figure S8 in the Supporting Information) with an ITO working electrode set at 1.20–1.75 V vs Ag/AgCl revealed O2 formation with Cucyclam in solution (75–88 % Faradaic yield, Table 1, entries 3 and 4), concomitant to the deposition of a greenish film at the electrode surface. The same electrode retains the electrocatalytic activity, showing > 92 % Faradaic yield towards oxygen production when used in the same media, at the same potentials, but in a copper-free solution (Table 1, entries 3 and 4; values in brackets). A further electrochemical investigation was conducted at pH 7, as neutral aqueous solutions are considered the ideal compromise for operating conditions with respect to an integrated water splitting device. Recently, phosphate buffer at pH 7, was employed to investigate the activity towards water oxidation of a dinuclear copper species.[6] A CV scan under an anodic potential of 1 mm Cucyclam in 0.2 m phosphate buffer with GC working electrode (Figure 5, top) revealed an irreversible anodic peak at 0.85 V that can be likely attributed to oxidation of the metal center; its irreversible nature could be due to significant structural change of the copper coordination geometry, by passing from CuII (d9) to CuIII (d8).[4] A second, intense wave due to water oxidation occurs then at potentials above 1.5 V (overpotential 880 mV). Interestingly, in the case of CuMe4cyclam (Figure 5, bottom) the first anodic wave, ascribed to CuII oxidation, is observed at Epa = 1.4 V vs Ag/AgCl, displaying an approximate 550 mV positive shift with respect to the the Cucyclam derivative.[22] A similar effect was already observed for the NiIII/NiII complexes, where the Me4cyclam derivative displayed a 540 mV potential gap compared to the nonmethylated ligand (1.22 V versus 0.68 V vs SCE). This effect has been ascribed to the fact that tertiary amine ligands are poor sigma-donors, so that N-alkylation of the cyclam ligand turns to a favorable stabilization of the metal center in its low valent

Figure 4. XPS analysis (top) of the recovered ITO electrode after 30 min of CPE experiment at a potential of 1.04 V vs Ag/AgCl in a deaerated 1 mm Cucyclam solution in NaOH/0.1 m NaOAc, and comparison with an XPS analysis of a copper oxide coated electrode prepared by the literature procedure.[10] The solid blue component of the XPS spectra is the Cu2 + contribution, the striped blue components are its satellite, while the yellow component is the Cu + 1 contribution. SEM image of the surface (bottom).

bands of CuII 2p3/2 levels at 935 eV binding energy with characteristic satellite peaks are prominent (Figure 4, top), and observed also in the case of a CuO layer prepared according to the literature procedure proposed by Sun and co-workers.[10] A scanning electron microscopy (SEM) image of the electrode revealed the accumulation of non-homogeneous aggregates of up to 30–50 nm dimensions (Figure 4, bottom). Interestingly, copper oxide nanoparticles were observed also in solution upon 16 hours of electrolysis, as confirmed by a drastic change in the UV-Vis spectra, where diagnostic absorptions at 360 and 530 appeared (see Figure S3 in the Supporting Information),[9] while dynamic light scattering experiments indicated an average size of hundreds of nm, likely formed by aggregation of small particles (see Figure S4 in the Supporting Information). Concerning the primary site of Cucyclam degradation, a reasonable hypothesis is dehydrogenation of amine groups to imine within the macrocycle. This was supported by ESI-MS analysis of Cucyclam along the CPE experiment: before applying the anodic potential, ESI-MS analysis of the pristine species &

&

Chem. Asian J. 2016, 00, 0 – 0

www.chemasianj.org

4



Full Paper Figure 2 above), that showed a progressive increase of anodic currents upon repeated scans, in concomitance with the electrodeposition of the copper oxide film. Therefore, these results suggest no electrodeposition of active layers at the electrode surface under these conditions. The absence of electrodeposition of active layers in phosphate buffer pH 7, was further confirmed by electrolysis experiments. CPE experiments with Cucyclam and CuMe4cyclam 1 mm in phosphate buffer solution (pH 7), by applying a constant potential E = 1.75 V vs Ag/AgCl (corresponding to an overpotential of 1.13 V), showed intense electrolysis currents (ca. 1.5 mA, see Figure S12 in the Supporting Information) concomitant to oxygen evolution accounting for Faradaic yields of 78 and 80 % for Cucyclam and CuMe4cyclam, respectively (Table 1, entries 5 and 6). In both cases, the recovered ITO electrodes after CPE experiments, gave a very low electrolysis current (< 50 mA) in copper-free buffer, therefore failing to produce oxygen. Therefore, under these conditions, water oxidation electrocatalysis occurs as a molecular process under homogeneous conditions. Several points support a decisive role of the macrocyclic ligands dictating a molecular catalytic pathway: (i) CuII aqua ions are not active in these conditions, due to precipitation of Cu3(PO4)2 (solubility product constant KSP = 1.4 10 37);[9] (ii) when black copper oxide nanoparticles prepared with a literature procedure[25] were added to a 0.2 m phosphate buffer, immediate precipitation of a light-blue solid of Cu3(PO4)2 was observed; (iii) in both cases, the electrolysis currents undergo a significant decrease over the course of approximately 3 h, concomitantly to a partial degradation of the complexes upon the application of the anodic potential. For CuMe4cyclam the current drop was lower (ca. 1/3 of the initial value after 2 h, while in the same timeframe the current with Cucyclam was reduced to 1/10 of the initial value, see Figure S12 in the Supporting Information), thereby indicating a higher stability for the N-methylated derivative (see discussion above). For this species, the intensity of the absorption at 640 nm was reduced by 50 % after 18 charge equivalents were passed during the CPE experiment (see Figure S13 in the Supporting Information). Precipitation of insoluble Cu3(PO4)2 was concomitantly observed, as the product of the catalyst deactivation.[26]

Figure 5. CV of Cucyclam (top) and CuMe4cyclam (bottom), 0.1–1.5 mm, in 0.2 m phosphate buffer, pH 7, with a GC working electrode. Color code: black: blank; orange: 0.1 mm copper; red: 0.4 mm; pink: 0.6 mm; violet: 0.8 mm; blue: 1 mm; green: 1.5 mm. Auxiliary electrode: Pt wire, Reference electrode: Ag/AgCl, scan rate: 100 mV s 1. Inset: plot of the current at 1.75 V vs copper concentration.

state.[20b, 23] Also in this case, a second wave due to water oxidation is observed at potentials above 1.5 V (overpotential of 880 mV). In both cases, the current at 1.75 V depends linearly on the copper concentration (0.1–1.5 mm), suggesting a single copper site molecular catalysis[4] rather than formation of active colloids or small oligomers.[6] The rate constant (kcat) for CuMe4cyclam, that displayed the higher catalytic current, was roughly estimated as 7 s 1 from the ratio of the quasi-plateau current at 1.75 V and the peak cathodic current at E = 0.55 vs Ag/AgCl, relative to a bielectronic reduction of CuII to Cu0 (see Figure S10 in the Supporting Information).[24] This value is comparable to other TOF numbers reported for copper species (see Table S1 in the Supporting Information). The wave due to water oxidation with Cucyclam and CuMe4cyclam in 0.2 m phosphate buffer, pH 7 was also observed in repetitive CV scans that employed ITO as working electrode (see Figure S11 in the Supporting Information). In both cases, under repetitive scans the intensity of the anodic current progressively goes down, likely ascribable to a decrease in the active electrode area due to oxygen bubbles formation at the surface. This is a significant difference with respect to CV experiments conducted in alkaline conditions for Cucyclam (see Chem. Asian J. 2016, 00, 0 – 0

www.chemasianj.org


Conclusions In this study, we have investigated the electrocatalytic activity towards water oxidation of two copper(II) complexes with tetraaza macrocyclic ligands: 1,4,8,11-tetraazacyclotetradecane (cyclam) and 1,4,8,11-tetramethyl-1,4,8,11-tetraazacyclotetradecane (Me4cyclam). Under alkaline conditions (pH 9–12), electrodeposition of an active copper oxide layer occurs readily upon application of an anodic bias; this leads to heterogeneous catalysis. Conversely, under neutral media (phosphate buffer pH 7), formation of oxide layers is disfavored, and water oxidation catalysis occurs under homogeneous conditions with both Cucyclam and CuMe4cyclam at overpotential of approximately 880 mV. Inspection of literature results indicate that this is unique for mononuclear copper species and only 120 mV 5


&

&

Full Paper respectively as working, auxiliary, and reference electrodes. 0.1 m S2O82 was used as a sacrificial oxidant at the auxiliary electrode, that was separated from the solution by a frit.

above that of the current benchmark with a dinuclear copper species.[6] By comparison of Cucyclam and CuMe4cyclam behavior, the presence of alkyl groups at the nitrogen binding sites of the macrocyclic ligand has proven to tune the redox potentials of the metal ion,[20b, 23] and to influence the stability of the complex towards oxidation. Moreover, these ligands offer the opportunity to introduce functional groups aimed at tuning the redox potentials of the metals, or at achieving specific proximal effects of water coordination, to improve catalytic performance by lowering overpotential and increasing catalytic currents, therefore powering the design of molecular complexes for water oxidation catalysis.

Dissolved oxygen measurements were performed using an Ocean Optics FOXY-R fluorescence probe coupled with NeoFox-GT phase fluorimeter and NeoFox Viewer software. The oxygen probe was calibrated using the two points calibration procedure of NeoFox Viewer software, employing a 2.0 m sodium sulfite as 0 mg l 1 dissolved oxygen value, and an aerated aqueous solution for which the amount of dissolved oxygen is reported in the instrument handbook, depending on the temperature and pressure. XPS spectra were acquired in ultra high vacuum at room temperature in normal emission geometry using a conventional Mg X-ray source (hn = 1253.6 eV) and a 1808 spherical sector electron energy analyzer (total energy resolution 0.8 eV). The photoelectron binding energy is referenced to the Fermi edge. The spectra are normalized to the incident photon flux and have been analyzed by performing a non-linear mean square fit of the data following. We used a Shirley background and reproduced the photoemission intensity by using asymmetric Doniach–Sunjic functions.

Experimental Section Caution: perchlorate salts of transition metal complexes are potentially explosive and should be handled with care! Synthesis of Cucyclam(ClO4)2 was performed by adapting the literature procedure.[17] In a 50 mL round-bottom flask, Cu(ClO4)2·6 H2O (201 mg, 0.54 mmol) was dissolved in THF (30 mL). The solution was light blue colored. When 1,4,8,11-tetraazacyclotetradecane (107 mg, 0.54 mmol) was added to the solution it turned to dark pink and appeared cloudy. After 2 h of stirring at room temperature, the reaction mixture was filtered and the solution obtained was maintained at room temperature until, after four days, from slow evaporation of the solvent, the crystals of Cucyclam as the perchlorate salt form. The solid was separated by centrifugation and air dried (46 % yield). Elemental analysis (calc.): C 26.30 % (calc. 25.95); H 5.48 (5.23); N 11.92 (12.11). ESI-MS (FIA, flow: CH3CN, m/ z): 362 [{Cu(C10H24N4)}ClO4] + . FT-IR (KBr): n˜ = 3240 (s), 3170 (m), 2930 (w), 2880 (w), 1430 (w), 1090 (vs), 998 (m), 883 (w), 625 cm 1 (m). UV-Vis: lmax = 513 nm (e = 104 L mol 1 cm 1 in 0.2 m phosphate buffer pH 7); lmax = 507 nm (e = 80 L mol 1 cm 1 in NaOH pH 12).

Acknowledgements Financial support from the Italian Ministero dell’Universit e della Ricerca (MIUR), (FIRB RBAP11C58Y, “NanoSolar” and PRIN 2010 “Hi-Phuture” and COST actions (CM1205 “CARISMA: CAtalytic RoutInes for Small Molecule Activation” and CM1202 “PERSPECT-H2O”) are gratefully acknowledged. I.B. acknowledges the CARISMA Action for a short term scientific mission at Laboratoire d’Electrochimie Moleculaire, LEM, Paris Diderot. Keywords: copper · electrocatalysis · homogeneous/ heterogeneous · molecular catalysis · oxidation

Synthesis of CuMe4cyclam(ClO4)2 was performed by adapting the literature procedure.[17] In a 50 mL round-bottom flask, Cu(ClO4)2·6 H2O (201 mg, 0.54 mmol) was dissolved in THF (30 mL). The solution was light blue colored. When 1,4,8,11-tetramethyl1,4,8,11-tetraazacyclotetradecane (138 mg, 0.54 mmol) was added, the solution turned to dark blue and appeared cloudy. After 2 h of stirring at room temperature, the mixture was filtered, the solution was colorless and the solid was violet blue. The solid compound was air dried, weighed and characterized (77 % yield). Elemental analysis (calc.): C 32.14 % (calc. 32.41); H 6.51 (6.22); N 10.33 (10.80). ESI-MS (FIA, flow: CH3CN + HCOOH 0.1 %, m/z): 418 [{Cu(C14H32N4)}ClO4] + . FT-IR (KBr): n˜ = 2920 (w), 2874 (w), 2020 (w), 1478 (m), 1462 (m), 1098 (vs), 960 (m), 808 (m), 732 (w), 624 cm 1 (s). UV-Vis: lmax = 640 nm (e = 233 L mol 1 cm 1 in 0.2 m phosphate buffer pH 7); lmax = 652 nm (e = 228 L mol 1 cm 1 in NaOH pH 12).

[1] a) S. Berardi, S. Drouet, L. Francs, C. Gimbert-Surinˇach, M. Guttentag, C. Richmond, T. Stoll, A. Llobet, Chem. Soc. Rev. 2014, 43, 7501 – 7519; b) M. D. Krks, O. Verho, E. V. Johnston, B. kermark, Chem. Rev. 2014, 114, 11863 – 12001; c) A. Sartorel, M. Bonchio, S. Campagna, F. Scandola, Chem. Soc. Rev. 2013, 42, 2262 – 2280; d) N. Armaroli, V. Balzani, Angew. Chem. Int. Ed. 2007, 46, 52 – 66; Angew. Chem. 2007, 119, 52 – 67. [2] a) S. M. Barnett, K. I. Goldberg, J. M. Mayer, Nat. Chem. 2012, 4, 498 – 502; b) T. Zhang, C. Wang, S. Liu, J.-L. Wang, W. Lin, J. Am. Chem. Soc. 2014, 136, 273 – 281; c) D. L. Gerlach, S. Bhagan, A. A. Cruce, D. B. Burks, I. Nieto, H. T. Truong, S. P. Kelley, C. J. Herbst-Gervasoni, K. L. Jernigan, M. K. Bowman, S. Pan, M. Zeller, E. T. Papish, Inorg. Chem. 2014, 53, 12689 – 12698. [3] M.-T. Zhang, Z. Chen, P. Kang, T. J. Meyer, J. Am. Chem. Soc. 2013, 135, 2048 – 2051. [4] M. K. Coggins, M.-T. Zhang, Z. Chen, N. Song, T. J. Meyer, Angew. Chem. Int. Ed. 2014, 53, 12226 – 12230; Angew. Chem. 2014, 126, 12422 – 12426. [5] P. Garrido-Barros, I. Funes-Ardoiz, S. Drouet, J. Benet-Buchholz, F. Maseras, A. Llobet, J. Am. Chem. Soc. 2015, 137, 6758 – 6761. [6] X.-J. Su, M. Gao, L. Jiao, R.-Z. Liao, P. E. M. Siegbahn, J.-P. Cheng, M-.T. Zhang, Angew. Chem. Int. Ed. 2015, 54, 4909 – 4914; Angew. Chem. 2015, 127, 4991 – 4996. [7] L.-L. Zhou, T. Fang, J.-P. Cao, Z.-H. Zhu, X.-T. Su, S.-Z. Zhan, J. Power Sources 2015, 273, 298 – 304. [8] L. Yu, X. Du, Y. Ding, H. Chen, P. Zhou, Chem. Commun. 2015, 51, 17443 – 17446. [9] a) Z. Chen, T. J. Meyer, Angew. Chem. Int. Ed. 2013, 52, 700 – 703; Angew. Chem. 2013, 125, 728 – 731; b) Z. Chen, P. Kang, M.-T. Zhang, B. R. Stoner, T. J. Meyer, Energy Environ. Sci. 2013, 6, 813 – 817. [10] F. Yu, F. Lei, B. Zhang, H. Li, L. Sun, ACS Catal. 2015, 5, 627 – 630.

Cyclic voltammetry experiments were performed using a BAS ECepsilon potentiostat and an AMEL Potentiostat-Galvanostat, model 7050. As the electrochemical cell, a standard three-electrode electrochemical cell and a custom designed air-tight cell were employed. As working electrode, a glassy carbon electrode (3 mm diameter) from BAS or Indium Tin Oxide coated glass slides (surface resistivity 8–12 W sq 1) from Sigma–Aldrich were used. In all experiments a Pt wire auxiliary electrode and Ag/AgCl/3 m NaCl reference electrodes were used. Electrolysis experiments were performed using an AMEL Potentiostat-Galvanostat, model 7050, in an assembled air-tight cell. An Indium Tin Oxide coated glass slide (surface resistivity 8–12 W sq 1) from Sigma–Aldrich, a Pt wire, and an Ag/AgCl/3 m NaCl were used

&

&

Chem. Asian J. 2016, 00, 0 – 0

www.chemasianj.org

6



Full Paper [11] X. Liu, H. Jia, Z. Sun, H. Chen, P. Xu, P. Du, Electrochem. Commun. 2014, 46, 1 – 4. [12] T.-T. Li, S. Cao, C. Yang, Y. Chen, X.-J. Lv, W.-F. Fu, Inorg. Chem. 2015, 54, 3061 – 3067. [13] L. M. Mirica, X. Ottenwaelder, T. D. P. Stack, Chem. Rev. 2004, 104, 1013 – 1046. [14] P. Haack, C. Limberg, Angew. Chem. Int. Ed. 2014, 53, 4282 – 4293; Angew. Chem. 2014, 126, 4368 – 4380. [15] a) N. Gagnon, W. B. Tolman, Acc. Chem. Res. 2015, 48, 2126 – 2131; b) C. J. Cramer, W. B. Tolman, Acc. Chem. Res. 2007, 40, 601 – 608. [16] J. Cho, R. Sarangi, W. Nam, Acc. Chem. Res. 2012, 45, 1321 – 1330. [17] a) B. Bosnich, C. K. Poon, M. L. Tobe, Inorg. Chem. 1965, 4, 1102 – 1108; b) P. A. Tasker, L. Sklar, J. Cryst. Mol. Struct. 1975, 5, 329 – 344; c) M. Bakaj, M. Zimmer, J. Mol. Struct. 1999, 508, 59 – 72; d) E. K. Barefield, F. Wagner, Inorg. Chem. 1973, 12, 2435 – 2439; e) M. Sarma, T. Chatterjee, S. K. Das, Inorg. Chem. Commun. 2010, 13, 1114 – 1117. [18] a) T. Hoppe, S. Schaub, J. Becker, C. Wrtele, S. Schindler, Angew. Chem. Int. Ed. 2013, 52, 870 – 873; Angew. Chem. 2013, 125, 904 – 907; b) M. Nappa, J. S. Valentine, A. R. Miksztal, H. J. Schugar, S. S. Isied, J. Am. Chem. Soc. 1979, 101, 7744 – 7746. [19] M. Zhang, M.-T. Zhang, C. Hou, Z.-F. Ke, T.-B. Lu, Angew. Chem. Int. Ed. 2014, 53, 13042 – 13048; Angew. Chem. 2014, 126, 13258 – 13264. [20] a) L. M. P. Lima, D. Esteban-Gmez, R. Delgado, C. Platas-Iglesias, R. Tripier, Inorg. Chem. 2012, 51, 6916 – 6927; b) G. Golub, H. Cohen, P. Paoletti, A. Bencini, L. Messori, I. Bertini, D. Meyerstein, J. Am. Chem. Soc. 1995, 117, 8353 – 8361. [21] P. Patnaik, Handbook of Inorganic Chemicals, McGraw-Hill, New York, 2002. [22] A difference in redox potentials of the two species was observed also in acetonitrile, where anodic waves were observed at 1.56 and 1.94 V

Chem. Asian J. 2016, 00, 0 – 0

www.chemasianj.org


[23] [24]

[25] [26]

vs Ag/AgCl for Cucyclam and CuMe4cyclam, respectively (see Figure S9 in the Supporting Information). Upon addition of 40 mm phosphate buffer (up to 40 % vol/vol), a shift of the potentials towards less positive values was observed (Figure S9). E. K. Barefield, G. M. Freeman, D. G. Van Derveer, Inorg. Chem. 1986, 25, 552 – 558. The kcat value was estimated from the following Equation: Icat/Ip = 2.242 (ncat/np) [(kcat RT)/(Fu)]1/2 ; where Icat is the catalytic current at 1.75 V, Ip is the peak current at E = 0.55 vs Ag/AgCl relative to the bielectronic reduction of CuII to Cu0 (the currents of the blank were subtracted to both Icat and Ip), ncat = 4 is the number of electrons required for water oxidation, np = 2 is the number of electrons involved in the reference redox event, F = 96485 Cmol 1 is the Faradaic constant, kcat is the turnover frequency, R = 8.314 JK 1 mol 1, T = 298 K, u is the scan rate = 0.1 Vs 1. A scan rate of 0.1 Vs 1 was exploited to avoid natural convection events, likely occurring at lower scan rates. A. S. Lanje, S. J. Sharma, R. B. Pode, R. S. Ningthoujam, Adv. Appl. Sci. Res. 2010, 1, 36 – 40. Moreover, some inactive copper phosphate precipitate was observed to form at the ITO surface, leading to its passivation and contributing to the abatement of electrolysis current. A similar observation was recently reported by C. Lu, J. Du, X.-J. Su, M.-T. Zhang, X. Xu, T. J. Meyer, Z. Chen, ACS Catal. 2016, 6, 77 – 83.

Manuscript received: January 10, 2016 Accepted Article published: February 17, 2016 Final Article published: && &&, 0000

7


&

&

Full Paper

FULL PAPER Electrocatalysis

The dark side of copper: Homogeneous water oxidation electrocatalysis promoted by copper (II) complexes with tetraaza macrocyclic ligands occurs in neutral aqueous media with turnover frequency up to 7 s-1, while heterogeneous routes are predominant in alkaline conditions, involving the formation of active copper oxide at the electrode surface.

Andrea Prevedello, Irene Bazzan, Nicola Dalle Carbonare, Angela Giuliani, Sunil Bhardwaj, Cristina Africh, Cinzia Cepek, Roberto Argazzi, Marcella Bonchio, Stefano Caramori, Marc Robert,* Andrea Sartorel* && – && Heterogeneous and Homogeneous Routes in Water Oxidation Catalysis Starting from CuII Complexes with Tetraaza Macrocyclic Ligands

&

&

Chem. Asian J. 2016, 00, 0 – 0

www.chemasianj.org

8

