Synthesis, Structure, and Magnetic ... - ACS Publications

Article pubs.acs.org/IC

Synthesis, Structure, and Magnetic Electrochemical Properties of a Family of Tungstoarsenates Containing Just CoII Centers or Both CoII and FeIII Centers Charyle S. Ayingone Mezui,† Pedro de Oliveira,† Anne-Lucie Teillout,† Jérôme Marrot,‡ Patrick Berthet,§ Mounim Lebrini,‡,∥ and Israel̈ M. Mbomekallé*,† †

Equipe d’Electrochimie et de Photo-électrochimie, Laboratoire de Chimie Physique, Université Paris-Sud, UMR 8000 CNRS, Université Paris-Saclay, Orsay, F-91405, France ‡ Institut Lavoisier de Versailles, Université de Versailles St. Quentin, UMR 8180 CNRS, Université Paris-Saclay, Versailles, F-78035, France § Equipe Synthèse Propriétés et Modélisation des Matériaux, ICMMO, Université Paris-Sud, UMR 8182 CNRS, Université Paris-Saclay, Orsay, F-91405, France S Supporting Information *

ABSTRACT: The three polyoxotungstates [(NaOH 2 ) 2 Co II 2 (As2W15O56)2]18− (1), [(NaOH2)(CoIIOH2)CoII2(As2W15O56)2]17− (2), and [(CoIIOH2)2CoII2(As2W15O56)2]16− (3) have been prepared in aqueous solution upon mixing cobalt(II) salts with the ligand [As2W15O56]12−. The reaction of 1 or 2 with the Fe3+ ion leads invariably to the same species [(FeIIIOH2)(CoIIOH2)CoII2(As2W15O56)2]15− (4) possessing three cobalt atoms and a single iron atom. However, if the Fecontaining homologue of compound 1, that is, the polyoxotungstate [(NaOH2)2FeIII2(As2W15O56)2]16− (5), is employed instead to react with the Co2+ ion, the species [(CoIIOH2)2FeIII2(As2W15O56)2]14− (6) is obtained, having two cobalt atoms and two iron atoms. The compounds 1, 2, 3, 4, and 6 are described for the first time and have been characterized by several physicochemical methods such as FTIR, UV−visible, ATG, and elemental analysis. Structural analysis by single-crystal X-ray diffraction has been carried out with compounds 2 (monoclinic space group P21/c, a = 17.0622(5) Å, b = 15.0828(4) Å, c = 32.0872(8) Å, β = 91.170(1)°, and Z = 2) and 3 (triclinic space group P1̅, a = 13.6137(7) Å, b = 13.8836(8) Å, c = 22.9276(6) Å, α = 89.906(3)°, β = 78.356(2)°, γ = 61.451(2)°, and Z = 1). Electrochemical studies undertaken with all the above-mentioned compounds and some of their homologues shed light on the influence of the chemical composition on their electrocatalytic properties toward substrates such as the nitrite ion and dioxygen. Magnetic measurements evidence anisotropic ferromagnetic interactions between Co2+ ions and antiferromagnetic interactions between Fe3+ ions. The nature and the strength of the Co2+−Fe3+ interactions depend on the relative orientations of their 3d orbitals. The effective magnetic moment of the Co2+ ions varies with the temperature and with the distortion of the octahedral sites in which they are located.

■

element.1,2 In this study, we concentrate on a family of tungstodiarsenic Dawson-type compounds4 of general formula [As2W18O62]6− which result from the union of two semianions [AsW9O34]9−, the final species having a plane of symmetry σ which separates its two identical moieties. It is possible to hydrolyze partially the [As2W18O62]6− species upon the action of conveniently chosen bases. Segments such as [WO]4+, [W3O6]6+, or [W6O14]8+ may be lost by the parent compound, resulting in the formation of lacunary POMs, since they possess one or several vacant sites represented by the symbol □ in the following formulas: [As2W17□O61]10−, [As2W15□3O56]12−, and [H2As2W12□6O48]12−, respectively.5 These lacunary POMs may be used as scaffolds to synthesize mixed compounds

INTRODUCTION Polyoxometalates (POMs for short) are inorganic compounds obtained upon acidification of solutions containing oxometalate ions. In fact, under certain experimental conditions (pH, concentration, temperature, and so on), the acidification of an oxoanion, [MOm]n−, aqueous solution leads to the formation of large molecular scaffolds by polycondensation whose nuclearity may vary from less than ten up to several hundred metal atoms M.1 In certain cases, the {MOx} entities combine or aggregate around moieties of a different nature, {XOx}, giving rise to POMs whose general formula is [XxMmOy]q−. The element M is found in its highest oxidation state and usually consists of tungsten, molybdenum, vanadium, and niobium,2 and more and more often noble metal elements such as palladium or gold.3 The heteroelement X is usually phosphorus, arsenic, silicon, or boron, but may also be any other nonmetallic or metallic © XXXX American Chemical Society

Received: October 26, 2016

A

DOI: 10.1021/acs.inorgchem.6b02593 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Scheme 1. Formation of “Sandwich-Type” Dawson POMs, [(NaOH2)2M2(As2W15O56)2]x− and [(M′OH2)2M2(As2W15O56)2]y−a

a

The green and red polyhedra represent the {AsO4} and the {WO6} groups, respectively. The yellow, the blue, and the purple spheres represent the two water molecules, the sodium atoms, and the atoms of the sandwiched M metal, respectively.

upon the insertion of “d” or “f” metal cations into the vacant sites. All these transformations allow the creation of a myriad of new molecules which have not yet been fully characterized. For example, we know that the trivacant derivative, [As2W15O56]12−, reacts with “d” and “f” metal cations in order to give rise to a family of POMs called “sandwich-type”. Indeed, in these species an equatorial plane cluster containing 2 to 4 identical or different metal cations is sandwiched between two [As2W15O56]12− fragments (Scheme 1).6−13 POMs exhibit several combinations of different properties which render them very interesting for applications in domains such as catalysis, magnetism, materials science, energy, and medicine. They are the object of many research projects, ranging from fundamental science to applied technology.14−25 As an example, we may mention the use of a Keggin-type POM as the cathode in a lithium battery.26 In fact, many POM molecules may reversibly accumulate and give back several electrons without decaying. The presence of electroactive metal cations in their structures makes them excellent models for the study of electron transfer mechanisms. In addition, the careful choice of the metal cations included in a POM may have a pronounced influence on its electrocatalytic features.17,27 Using several electrochemical techniques complemented with DFT calculations, we were recently able to rationalize the evolution of the redox potentials of the FeIII centers contained in sandwich-type POMs having the general formulas [(FeIIIOH)2M2(X2W15O56)2]n− or [(MOH2)2FeIII2(X2W15O56)2]n− (with X = AsV and PV and M = MnII, MnIII, CoII, NiII, CuII, and ZnII).28 In the present work, we have synthesized and characterized by different physicochemical techniques (UV−visible and infrared spectroscopies, elemental and thermogravimetric analysis) three new CoII center containing “sandwich-type” Dawson POMs: Na 18 [(NaOH 2 ) 2 Co II 2 (As 2 W 15 O 56 ) 2 ]·30H 2 O (1), Na 17 [(NaOH2 )(CoIIOH2)CoII2(As 2W15O56 )2 ]·31H2 O (2), and Na16[(CoIIOH2)2CoII2(As2W15O56)2]·30H2O (3). The crystallographic structure of compounds 2 and 3 was determined by single-crystal X-ray diffraction. Compounds 1 and 2 both reacted with Fe3+ ions and gave rise to the same reaction product, which contains three CoII centers and one FeIII center: Na15[(FeIIIOH2)(CoIIOH2)CoII2(As2W15O56)2]·

34H2O (4). In addition, Co2+ ions reacted with the species Na16[(NaOH2)2FeIII2(As2W15O56)2]·54H2O7 (5), leading to the formation of the compound Na14[(CoIIOH2)2FeIII2(As2W15O56)2]·44H2O (6), in which the two Na+ ions were replaced by two CoII centers, occupying the sites next to the two FeIII centers. All these new compounds were unambiguously identified and characterized by electrochemical techniques, namely, cyclic voltammetry and controlled potential coulometry coupled to UV−visible spectrophotometry. The latter allowed us to confirm, whenever required, the number and the relative positions between the FeIII and the CoII centers. We have conducted some tests in solution in order to verify if there was any synergic effect due to the mutual presence of FeIII and CoII centers toward the electrocatalytic efficiency of the compounds for the reduction of dioxygen and nitrite ion. Last but not least, the magnetic studies brought to light the existing interactions between the CoII centers, on the one hand, and the CoII and FeIII centers, on the other hand.

■

EXPERIMENTAL SECTION

General Methods and Materials. Pure water obtained with a Milli-Q Intregral 5 purification set was used throughout. All reagents were of high-purity grade and were used as purchased without further purification. CH3COOH (Carlo Erba), LiCH3COO·2H2O (Fluka), CoCl2·6H2O (Acros Organics), FeCl2·4H2O (Prolabo), NaCl (Prolabo), ethanol (Sigma-Aldrich), HCl (VWR), As2O5·xH2O (Acros), Na2WO4·2H2 O (Chempur), and KCl (Fluka) were commercial products. The composition of the various media was as follows: for pH 5.0 and 6.0, 1.0 M LiCH3COO + CH3COOH. The UV−visible spectrophotometric characterization of the different species was carried out in aqueous solutions with a PerkinElmer Lambda 19 using 10.00 mm optical path quartz cells. The method allowed us, on the one hand, to check for the pH stability domains of the compounds and, on the other hand, to determine the molar extinction coefficient, ε, of each compound at a given wavelength, λ. Elemental analysis was performed by the Service Central d’Analyze CNRS (Solaize, France). The thermogravimetric analysis (TGA) was carried out with an apparatus model TGA 2050, which records the mass loss of a solid sample as a function of the temperature. The samples are kept in an oxygen flow and heated from room temperature (ca. 20 °C) up to 600 °C at a rate of 5 °C per minute. The number of crystallization and of constitution water molecules of each compound B


Article

Inorganic Chemistry

(As2W15O56)2]·34H2O. IR: 861; 821; 517; 491. Anal. found (calcd) for Na15[(FeIIIOH2)(CoIIOH2)CoII2(As2W15O56)2]·34H2O: As, 3.30 (3.39); W, 61.30 (62.44); Co, 1.89 (2.00); Fe, 0.70 (0.63); Na, 4.30 (3.90) [MW 8832.86]. In the same experimental conditions, the reaction of compound 2 with an equivalent of FeIII always yielded compound 4, which possesses a single FeIII center and three CoII centers. Na14[(CoIIOH2)2FeIII2(As2W15O56)2]·44H2O (6). A sample of 1.0 g of pure compound Na16[(NaOH2)2FeIII2(As2W15O56)2]·54H2O (5) prepared as described in the literature7 was added to a solution composed of 2.5 mL of 0.5 M CH3COONa/0.5 M CH3COOH (pH = 4.6) + 2.5 mL of 1 M NaCl. The mixture was heated to about 80 °C, and after solubilization of 5, a sample of 55 mg of CoCl2·6H2O (0.23 mmol) was added in small aliquots and stirred for another 30 min at the same temperature, and finally filtered while still warm. The precipitate formed upon cooling was filtered on a fritted glass funnel, rinsed twice with a NaCl saturated solution and twice with ethanol, and dried in air. A mass of 0.28 g (28%) of a yellowish powder was obtained, corresponding to the compound Na14[(CoIIOH2)2FeIII2(As2W15O56)2]·44H2O. IR: 861; 822; 696; 518; 491. Anal. found (calcd) for Na14[(CoIIOH2)2FeIII2(As2W15O56)2]·44H2O: As, 3.40 (3.33); W, 62.60 (61.37); Co, 1.19 (1.31); Fe, 1.25 (1.24); Na, 4.60 (3.58) [MW 8986.94]. X-ray Crystallography. Intensity data collection was carried out with a Bruker Nonius X8 APEX 2 diffractometer equipped with a CCD bidimensional detector using the monochromatized wavelength λ(Mo Kα) = 0.71073 Å. The data were collected at 275 K for 2 and at 200 K for 3. The absorption correction was based on multiple and symmetry-equivalent reflections in the data set using the SADABS program29 based on the method of Blessing.30 The structures were solved by direct methods and refined by full-matrix least-squares using the SHELX-TL package.31 As usually observed for structures of polyoxometalates, which are large and highly negatively charged entities, it was not possible to locate all the water molecules and alkali counterions because of the disorder.32−34 As a consequence, all the structures have a discrepancy between the formulas determined by elemental analysis and the formulas deduced from the crystallographic atom list. Crystallographic data are given in Table 1. Comparisons of selected bond lengths are given in Table 2. Electrochemical Experiments. Electrochemical data was obtained using an EG&G 273A potentiostat driven by a PC with the M270 software. A one-compartment cell with a standard threeelectrode configuration was used for cyclic voltammetry experiments.

is derived from the curve corresponding to the sample mass loss as a function of the temperature. Infrared spectra were recorded on a Nicolet 6700 FT spectrophotometer driven by a PC with the OMNIC E.S.P. 8.2 software. Magnetic measurements were carried out on polycrystalline samples using a SQUID magnetometer (Quantum Design MPMS-5). Syntheses. The syntheses of the precursors K6[As2W18O62]· 14H 2 O, Na 12 [As 2 W 15 O 56 ]·21H 2 O, and Na 16 [(NaOH 2 ) 2 Fe III 2 (As2W15O56)2]·45H2O were carried out according to the procedures described in the literature, and their purity was confirmed by IR and cyclic voltammetry.5,7 Na18[(NaOH2)2CoII2(As2W15O56)2]·30H2O (1). A pure sample of the compound Na12[As2W15O56]·21H2O (10 g; 2.2 mmol) was suspended in 50 mL of 0.5 M CH3COONa/0.5 M CH3COOH (pH = 4.6). This suspension is treated with aliquots of crystallized CoCl2·6H2O (0.52 g; 2.2 mmol) which are slowly added to the mixture. The heterogeneous mixture progressively solubilized, and after about 30 min of stirring the almost clear solution obtained was paper filtered. The light pink filtrate obtained in a beaker was treated with a single portion of NaCl (6 g; 0.10 mol), which led to the immediate formation of a pink-colored precipitate. The latter was separated by filtration on a fritted glass funnel, rinsed twice with a 1 M NaCl solution and twice with ethanol, and dried in air. A mass of 8.46 g (86.2%) of a pinkish powder was obtained, corresponding to the compound Na18[(NaOH2)2CoII2(As2W15O56)2]·30H2O. IR: 939; 865, 840sh, 698, 472w. Anal. found (calcd) for Na18[(NaOH2)2CoII2(As2W15O56)2]·30H2O: As, 3.39 (3.42); W, 61.30 (62.95); Co, 1.40 (1.35); Na, 5.45 (5.25) [MW 8760.97]. Na17[(NaOH2)(CoIIOH2)CoII2(As2W15O56)2]·31H2O (2). A solution containing 0.14 g of CoCl2·6H2O (0.6 mmol) in 40 mL of 0.5 M NaCl was treated with a sample of 2.0 g of the previously prepared compound Na18[(NaOH2)2CoII2(As2W15O56)2]·30H2O (1) (0.2 mmol) added in small aliquots under stirring and moderate heating (ca. 60 °C). After 30 min, the solution is filtered while still warm and let cool in air. After 24 h, the precipitate formed was filtered on a fritted glass funnel, rinsed twice with a 1 M NaCl solution and twice with ethanol, and dried in air. A mass of 1.3 g (65%) of a pinkish powder was obtained, corresponding to the compound Na17[(NaOH2)(CoIIOH2)CoII2(As2W15O56)2]·31H2O. IR: 939; 865, 842sh, 694, 472w. Anal. found (calcd) for Na17[(NaOH2)(CoIIOH2)CoII2(As2W15O56)2]·31H2O: As, 3.28 (3.41); W, 61.90 (62.73); Co, 1.96 (2.01); Na, 5.10 (4.71) [MW 8791.94]. Na16[(CoIIOH2)2CoII2(As2W15O56)2]·30H2O (3). A solution containing 1.07 g of CoCl2·6H2O (4.5 mmol) in 100 mL of 1 M NaCl was treated with a sample of 10 g of Na12As2W15O56·21H2O (2.2 mmol) added in small aliquots under stirring. The resulting bright pink suspension was heated (ca. 60 °C) in a water bath until full solubilization was achieved, and then filtered while still warm. A pink crystalline material formed upon cooling. After 24 h, this precipitate was filtered on a fritted glass funnel, rinsed twice with a 1 M NaCl solution and twice with ethanol, and dried in air. A mass of 8.0 g (81.3%) of a pink powder was obtained, corresponding to the compound Na16[(CoIIOH2)2CoII2(As2W15O56)2]·30H2O. IR: 943; 865, 829sh, 694, 472w. Anal. found (calcd) for Na 16 [(Co II OH 2 ) 2 Co II 2 (As2W15O56)2]·30H2O: As, 3.35 (3.41); W, 62.05 (62.77); Co, 2.65 (2.68); Na, 4.80 (4.19) [MW 8786.88]. Na15[(FeIIIOH2)(CoIIOH2)CoII2(As2W15O56)2]·34H2O (4). It is important to note that our purpose was to synthesize the compound [(FeIIIOH2)2CoII2(As2W15O56)2]14− containing two FeIII centers. A sample of 1.0 g of previously synthesized pure compound 1 (0.11 mmol) was added to a medium constituted of 2.5 mL of 0.5 M CH3COONa/0.5 M CH3COOH (pH = 4.6) + 2.5 mL of 1 M NaCl. The resulting solution was heated to about 80 °C, and after solubilization of 1, a sample of 50 mg of FeCl2·4H2O (0.25 mmol) was added in small aliquots and stirred for another 30 min at the same temperature, and finally filtered while still warm. The precipitate formed upon cooling was filtered on a fritted glass funnel, rinsed twice with a 1 M NaCl solution and twice with ethanol, and dried in air. A mass of 0.51 g (50.6%) of a yellowish powder was obtained, corresponding to the compound Na15[(FeIIIOH2)(CoIIOH2)CoII2-

Table 1. Crystallographic Data for 2 and 3 empirical formula fw, g cryst syst space group a/Å b/Å c/Å α/deg β/deg γ/deg V/Å3 Ζ ρcalc/g cm−3 μ/mm−1 data/params Rint GOF R (>2σ(I)) a

R1 =

C

2

3

As4Co3Na18O154W30 8869.79 monoclinic P21/c 17.0622(5) 15.0828(4) 32.0872(8) 90 91.170(1) 90 8255.8(4) 2 3.568 22.05 24383/19943 0.056 1.107 R1a = 0.073, wR2b = 0.221

As4Co4Na16O169W30 9122.74 triclinic P1̅ 13.6137(7) 13.8836(8) 22.9276(6) 89.906(3) 78.356(2) 61.451(2) 3705.5(4) 1 4.088 24.67 21580/19340 0.051 1.003 R1 = 0.043, wR2 = 0.140

∑ |Fo| − |Fc| b . wR2 ∑ |Fc|

=

∑ w(Fo2 − Fc 2)2 ∑ w(Fo2)2

. DOI: 10.1021/acs.inorgchem.6b02593 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

elemental analysis results.6 We obtained similar high yields following a slightly different procedure which we previously described and have used for several years for the synthesis of “sandwich-type” Dawson POMs.8,9 We did not find it convenient to acidify a solution of the lacunary compound [As2W15O56]12− with 6 M aqueous hydrochloric acid. The progressive addition of small aliquots of the lacunary compound [As2W15O56]12− to a solution containing an excess of Co2+ ions implies that the concentration ratio [Co2+]/ [As2W15O56] always equals 2 at least. This prevents the decay of [As2W15O56]12−, since the reaction kinetics of the latter with “d” cations is relatively fast. The synthesis of compound 1 was carried out in buffered CH3COONa medium, in which the decay kinetics of [As2W15O56]12− is slower than that of its reaction with Co2+, leading to the formation of the two “d” center structure [(NaIOH2)2CoII2(As2W15O56)2]18−. In order to avoid the evolution of the latter toward the more stable three CoII center derivative, compound 1 was rapidly isolated in a solid form upon precipitation with NaCl. Recrystallization usually is a slow process but may yield convenient crystals for single-crystal X-ray diffraction analysis. In the present case, all attempts resulted in the formation of compound 2 crystals, the ultimate form of the evolution of compound 1 in solution:

Table 2. Selected Bond Lengths (Å) for 2 and 3 Associated with the Representations in Figure 1 2a Co1−O36 Co1−O41 Co1−O45 Co1−O55 Co1−O56*b Co1−O56*

2.080 2.052 2.037 2.061 2.125 2.146

(12) (13) (12) (12) (11) (11)

Co1−O41 Co1−O45 Co1−O48 Co1−O52 Co1−O57* Co1−O57*

2.029 2.078 2.099 2.022 2.143 2.153

(6) (6) (6) (6) (5) (5)

Co2−O36 Co2−O41 Co2−O48 Co2−O52 Co2−O56 Co2−O57**

2.276 (13) 2.238 (14) 2.189 (15) 2.193 (14) 2.233 (13) 1.95 (4)

Co2−O36 Co2−O45 Co2−O48 Co2−O55 Co2−O56** Co2−O57

2.063 2.113 2.142 2.062 2.064 2.140

3 (6) (6) (6) (6) (7) (6)

a

Co2 corresponds to external sites statistically occupied by 50% Co and 50% Na. b*Equivalent oxygen atoms; **oxygen atom of the water molecule. The reference electrode was a saturated calomel electrode (SCE) and the counter electrode a platinum gauze of large surface area; both electrodes were separated from the bulk electrolyte solution via fritted compartments filled with the same electrolyte. The working electrode was a ca. 10 × 10 × 2 mm stick (GC, Mersen, France). The pretreatment of this electrode before each experiment has been described elsewhere.35 The polyanion concentration was 2 × 10−4 M. Prior to each experiment, solutions were thoroughly deaerated for at least 30 min with pure argon. A positive pressure of this gas was maintained during subsequent work. All cyclic voltammograms were recorded at a scan rate of 10 mV s−1 unless otherwise stated. All experiments were performed at room temperature, which is controlled and fixed for the lab at 20 °C. Results were very reproducible from one experiment to another, and slight variations observed over successive runs are attributed to the uncertainty associated with the detection limit of our equipment (potentiostat, hardware, and software) rather than to the working electrode pretreatment or to possible variations in temperature.

[(Na IOH 2)2 CoII 2(As 2 W15O56 )2 ]18 − (1) → [(NaOH 2)(CoIIOH 2)CoII 2(As 2 W15O56 )2 ]17 − (2)

This explains why all attempts to synthesize the species having the theoretical formula [(Fe I I I OH 2 ) 2 Co I I 2 (As2W15O56)2]14− upon substitution of the two NaI centers occupying external sites by two FeIII centers led invariably to compound 4, which has three CoII centers and a single FeIII center. Compound 4 formally derives from compound 2 upon reaction with one equivalent of Fe3+, but it was also obtained upon mixing in solution Fe3+ ions with compound 1, since the latter evolved into the more stable form corresponding to compound 2. Compound 5 has the same structure as compound 1, but possesses two FeIII centers instead of two CoII centers and is more stable than the latter.36 It is stable enough in solution to react with Co2+ ions and lead to compound 6, which has two FeIII centers associated with two CoII centers. We were, then,

■

RESULTS AND DISCUSSION Syntheses. Compound 3 was first described by En-Bo Wang and collaborators, who proposed the formula Na16[(CoIIOH2)2CoII2(As2W15O56)2] based exclusively on

Figure 1. Representation of the structures of the polyanions. (A) [(CoOH2)2Co2(As2W15O56)2]16− (3) and (B) [(NaOH2)(CoOH2)Co2(As2W15O56)2]17− (2). Polyhedral and exploded view representation: the green and blue polyhedra represent the {AsO4} and the {WO6} groups, respectively. The spheres correspond to cobalt atoms (purple); the positions statistically occupied at 50% either by a cobalt or by a sodium atom (blue); oxygen atoms of the tungstic scaffold (yellow); oxygen atoms in water molecules (red). D


Article

Inorganic Chemistry able to isolate and characterize two new “sandwich-type” species containing both iron and cobalt: [(Fe IIIOH2)(Co IIOH 2)Co II2(As 2W15O 56) 2] 15− and [(CoIIOH2 ) 2Fe III2(As2W15O56)2]14−. They differ not just on the number of FeIII and CoII centers but also on their relative locations in the equatorial plane: while the sole FeIII center in compound 4 occupies an external site and is bound to a water molecule, the two FeIII centers of compound 6 are located in internal sites and share oxygen atoms with the tungstic skeleton of the molecule (vide inf ra, crystal structure section). Crystal Structure. It was not possible to obtain good quality crystals with compounds 1, 4, and 6 to carry out a structural study by X-ray diffraction. The structure of compound 5 had already been described in previous studies,7 so just the compounds 2 and 3 have been analyzed by singlecrystal X-ray diffraction. These analyses led to the structures presented in Figure 1. The two structures differ just in the composition of the equatorial metal cluster sandwiched between two [As2W15O56]12− fragments. In the case of 3, the cluster is composed of four cobalt atoms in octahedral coordination, being equivalent in pairs. The Co1 atoms occupy the so-called internal sites and are bound the oxygen atoms exclusively belonging to the tungstic scaffold of the molecule. The Co2 atoms occupy the so-called external sites, and their sixth ligand is an oxygen from a water molecule (O56, in red in Figure 1A), unlike the five others, which belong to the tungstic scaffold of the molecule. The {Co4O14(H2O)2} cluster and the two {As2W15O56} fragments are joined together by an αββαtype junction.37 In the case of 2, and despite the fact that the composition of the equatorial metal cluster is different, {NaCo3O14(H2O)2}, the same αββα-type junction is found, being almost exclusively the one observed for Dawson sandwich-type structures, a consequence of their greater stability.37 In the Co1 internal sites, there are two cobalt atoms, whereas the Co2 external sites are statistically occupied by 50% Co and 50% Na, matching the Na−Co3 stoichiometry and the elemental analysis results. A supplementary confirmation of the difference between the sites Co1 and Co2 in compound 2 comes from their disparate Co−O bond lengths, indicative of the fact that the external sites are not 100% occupied by a cobalt atom (see Table 2). Magnetic Studies. The temperature dependence of the magnetization was studied in a 0.1 T field from 2 to 300 K. Except for 5, the magnetization regularly decreases with increasing temperature. For 5, the decrease exhibits a shoulder between 10 and 50 K. As in our previous study,12 this particular behavior is attributed to the existence of a small contribution of free iron ion. This contribution was subtracted before further analysis; it represents around 6% of all the iron contained in the sample under study. As the features exhibited by the products Xmol·T generally give useful indications on the magnetic interactions existing between magnetic cations of a polynuclear complex, these products were calculated from the experimental data. They are plotted in Figure 2, where two types of behavior may be observed. Four clusters containing only two, three, or four Co2+ ions and no more than one Fe3+ ion (compounds 1 to 4), a sharp maximum is observed at low temperature near 5 K. It is followed by a flat minimum preceding a smooth increase leading to an asymptotic value which is reached between 200 and 250 K. For clusters containing two Fe3+ ions (compounds 5 and 6), the products Xmol·T regularly increase with the temperature without reaching an asymptotic value at 300 K. This last type of behavior is characteristic of dominant

Figure 2. Plot of the Xmol·T product as a function of temperature for the compounds under study.

antiferromagnetic interactions as evidenced in studies of similar Mn2+- or Fe3+-containing compounds.12,38−40 Conversely, the first one is typical of the occurrence of mainly ferromagnetic interactions between the magnetic ions of the cluster, as already observed for similar Co 2+ - or Ni 2+ -containing compounds.12,41−44 To extend the analysis of these curves, it is necessary to consider the structure of the cluster containing the magnetic centers. In this cluster four cations are found at the corners of a slightly distorted rhomboid. The cations found on the internal sites (Fe3+ in 5 and 6; Co2+ in other compounds) are separated by the shortest diagonal of the rhomboid and linked by two oxygen anions allowing superexchange interactions. These cations are also linked in a similar way to the cations occupying the external sites (Na+ in 1 and 5, Co2+ or Fe3+ in other compounds). Therefore, the magnetic properties of the cluster depend on the magnetic moment of each cation and of the interactions existing along the sides of the rhomboid and that along its shortest diagonal. For several cations of transition metals like Mn2+ or Fe3+, the orbital magnetic moment is quenched and there is no magnetic contribution related to the spin−orbit interaction. Therefore, their molar magnetic susceptibility Xmol follows a Curie law taking into account only the spin contribution. Assuming an ideal value of 2 for the Landé factor g, this can be written, in the cgs system, Xmol·T = (1/2)S(S + 1) where S is the spin of the cation. In such a case, the magnetic susceptibility of the cluster can be calculated from its magnetic energy described by the following Heisenberg-type Hamiltonian: Ĥ = −2J1(S1̂ ·S2̂ + S2̂ ·S3̂ + S3̂ ·S4̂ + S4̂ ·S1̂ )‐2J2 S1̂ ·S3̂

In this expression J1 and J2 are two exchange constants describing respectively the interactions along the sides of the rhomboid and that along its shortest diagonal. The intrinsic susceptibility of the cluster is given by the following equation: Xcluster = g 2Nβ 2 ST (ST + 1) /3kT

in which the mean value ⟨ST(ST + 1)⟩ depends on the temperature which determines the population of the energy levels resulting from the individual spin coupling. The calculation applied to other compounds including a different sample of 5 was presented in previous papers.12,40 It will not be developed here since it cannot be used for clusters containing Co2+ ions. The reason is that in most compounds single Co2+ ions do not follow a Curie law: their magnetic behavior cannot E


Article

Inorganic Chemistry

and the resulting susceptibility of the cluster to which they belong over the whole measured temperature range. However, the shape of the Xmol·T curves provides some interesting indications on these properties. The maxima presented by the Xmol·T products of the compounds 1, 2, and 3 suggest an anisotropic ferromagnetic interaction between the Co2+ ions of the clusters, since they are similar to those observed for the analogous clusters mentioned above.43,44,47 The maximum values of Xmol·T are reached between 4 and 5 K and followed by a rapid decrease. The decrease following the maximum is due to the thermal population of excited levels corresponding to less ordered magnetic configurations inside the cluster. After a flat minimum observed near 40 K, the Xmol·T products increase slowly as a consequence of the increase of the individual effective moments. It is worth noting that the maximum values of Xmol·T, respectively 6.8, 12.6, and 17.4 emu·K/mol for 1, 2, and 3, are clearly different from those obtained for the analogous clusters, respectively 5.8, 10.3, and 25.0 emu·K/mol. These differences are always larger than those observed at room temperature and mentioned above. They clearly indicate that the magnetic properties of the clusters significantly evolve with slight differences in their geometries. The magnetic cluster found in compound 4 contains 3 Co2+ and 1 Fe3+ ions. At 300 K the value of the Xmol·T product is 11.25 emu·K/mol; assuming the usual contribution of 4.375 emu·K/mol for Fe3+, that of the 3 Co2+ ions is only 6.875 emu· K/mol, significantly lower than that observed for compound 2 yet higher than that expected for spin only contributions of these ions. At 2 K, retaining only the spin contributions, a ferromagnetic coupling of all the ions would give S = 7 and Xmol·T = 28 emu·K/mol, whereas an antiferromagnetic coupling of Fe3+ ion with all the Co2+ ions would give S = 2 and Xmol·T = 3 emu·K/mol. The experimental value is 12.74 emu·K/mol, only a little higher than that found for compound 2; it precludes both the S = 2 and the S = 7 configurations. An intermediate spin configuration may result from two different interactions of Fe3+ with Co2+ ions along the rhomboid sides. As Fe3+ ions have three half-filled t2g orbitals whereas Co2+ ions have only one, the other two being fully filled, the interaction along the rhomboid sides depends on the orientation of the Co2+ orbitals relative to the Fe3+ orbitals. Therefore, one of these interactions may be ferromagnetic and the other antiferromagnetic. Two Fe3+ ions are found on the internal sites of compounds 5 and 6 with respectively two Na+ ions and two Co2+ ions on the external sites. The shapes of the curves representing the Xmol·T product are similar for both compounds. They are typical of antiferromagnetic interactions between the Fe3+ ions. A simulation of the curve obtained for 5 was performed using the procedure described previously for another sample of this compound;12 it gives J2 = −2.70 cm−1, g = 2.01, and X0 = −0.36 × 10−3 emu/mol, in very good agreement with our previously reported results. For 5, at 2 K, the value of the Xmol·T product, close to 0, reflects the S = 0 ground state of the magnetic cluster. For 6 we found Xmol·T = 1.66 emu·K/mol at this temperature, therefore it appears that the magnetic moments of the Co2+ ions are not antiparallel, which results likely from interactions of similar type, ferromagnetic or antiferromagnetic, with Fe3+ ions along the rhomboid sides. Assuming that Fe3+ ions are in an antiferromagnetic configuration, it is impossible for the moment of each Co2+ ion to adopt a similar orientation in relation to that of their two Fe3+ neighbors. This kind of impossibility is known to lead to frustrated configurations.

be described with a spin only contribution since spin−orbit interaction and site distortion effects cannot be neglected.45,46 Moreover, for the analysis of the magnetic properties of similar Co2+ clusters an anisotropic exchange model is required, as evidenced by Casan-Pastor et al.42 Calculations made with such a model are limited to temperatures lower than 30 K where the Co2+ magnetic moment may be represented by a S = 1/2 effective spin, which corresponds to the lowest energy level resulting from the site distortion together with the spin−orbit interaction. In this temperature range, this level (which is the lowest Kramers doublet) is the only one populated. This model requires the determination of 4 parameters for 1 and at least 8 parameters for each of the other compounds containing Co2+ ions. To obtain significant values of these parameters, magnetic measurements carried out under only one field are not sufficient. They have to be completed by runs under different fields and experiments using different techniques such as heat capacity measurements, electron spin resonance, or inelastic neutron scattering. Such developments are beyond the aim of the present study. The high values of the Xmol·T product at high temperature for clusters containing only Co2+ ions (compounds 1, 2, and 3) clearly reveal the specific character of the magnetic contribution of these ions. If it was exhibiting a spin only contribution with a Landé factor equal to 2, each ion would contribute with 1.875 emu·K/mol. In this case the expected value for Xmol·T would be respectively 3.750, 5.625, and 7.500 emu·K/mol for compounds 1, 2, and 3. Experimentally, we observed nearly the double, respectively 7.15, 10.08, and 13.26 emu·K/mol with an uncertainty due to diamagnetic contributions lower than 0.2 emu·K/mol. For 1 and 2 the experimental values, respectively 6.38 and 9.4 emu·K/mol, are a little higher than those reported for an analogous Keggin phosphotungstate44 and an analogous Dawson diphosphotungstate.43 For 3 it is 13.1 emu·K/mol, being close to that obtained between 100 and 140 K for an analogous Keggin phosphotungstate.42,47 The comparison of the room-temperature values suggests that the contributions of the Co2+ ions on external sites are lower than those on internal sites. From the value determined for 1, the contribution of Co2+ ion on the internal site appears close to 3.5 emu·K/mol, whereas the Xmol·T product increases only around 3.0 emu·K/ mol for each Co2+ ion added on an external site. This difference appears as a consequence of the different distortions of the octahedra surrounding the magnetic cations. An examination of the crystallographic data describing compounds 2 and 3 clearly evidences the different distortions of the internal and external sites, but it also reveals that their geometry evolves with the number of sites occupied by Co2+ ions. For both sites of 3, the mean Co−O distance is close to 2.09 Å, but the distribution of the individual Co−O distances is larger for the internal site, 2.022 to 2.153 Å, than for the external site, 2.063 to 2.142 Å. The O−Co−O angles are also different, with values ranging from 82.7 to 96.2° (82.3 to 97.5°) and from 169.2 to 173.3° (171.9 to 176.8°) for the internal (external) site. The interaction paths are also slightly different: two identical Co− O−Co angles of 97.3° are found between the internal sites, whereas four O−Co−O angles ranging from 95.4° to 98.1° are found between the internal and the external sites which are linked by two pairs of slightly different sides (3.16 Å for one pair and 3.19 Å for the other, compared to 3.22 Å for the short diagonal). The lack of symmetry of all these sites makes it practically impossible to calculate their individual magnetic contributions F


Article

Inorganic Chemistry At 300 K, the value of Xmol·T reaches 8.10 emu·K/mol for 5 and 11.68 emu·K/mol for 6. The difference, 3.58 emu·K/mol, which is nearly constant above 200 K, comes from the moments of the two Co2+ ions found on the external sites of 6. It roughly corresponds to the spin only contributions of a 3d7 configuration, which is quite unexpected in comparison to the highest moments found for Co2+ ions in the other compounds under study. UV−Visible Spectrophotometry and Determination of the Molar Extinction Coefficient. All the compounds strongly absorb in the UV and scarcely in the visible range, resulting in almost colorless solutions. Sufficiently concentrated solutions (between 10−3 M and 5 × 10−3 M) were prepared in order to get useful absorbance values in the visible range, which were measured at a wavelength of λ = 560.11 nm, roughly corresponding to the band maxima (Figure S3). The Beer− Lambert law, A = εcl, via a linear regression, allowed determination of the value of ε for each species. When they are compared for compounds 1, 2, and 3, which possess 2, 3, and 4 CoII centers, respectively, it is clear that ε increases with the number of CoII centers present in the molecule (see Supporting Information), meaning that these centers are implicated in the absorption in the visible range, which is attributed to O → Co2+ charge transfer. This was clearly visible to the naked eye when the hues of solutions of the same concentration were compared, becoming more intense when going from compound 1, than 2, and finally 3. The absorbance of the solutions made of either of the Fe-containing POMs was too weak to allow for a reliable determination of ε values, since the concentrations used fell beyond the validity of the Beer− Lambert law. Stability in Solution and Electrochemistry Studies. Before studying the redox and the electrocatalytic properties of the POMs, their stability in solution as a function of the pH has to be checked. In fact, the solution pH influences the stability of POMs, and their frameworks may evolve into new forms or totally decay.2 The stability tests consist of comparing UV− visible spectra of POM-containing solutions recorded every 30 min over a total period of 24 h at least. Such a period is far longer than the time required to carry out an electrochemical characterization by cyclic voltammetry (several hours), but is justified by experiments where electrocatalytic applications are envisaged. The POM concentration in solution was 10−4 M, and spectra were recorded at wavelengths between 1200 and 200 nm at a scan rate of 240 nm/min. All complexes were studied in solution at pH values 5 and 6. Corroboration of stability was obtained by electrochemistry, the cyclic voltammograms (CV) of the given compounds being reproducible after several hours. Table S1 compiles the results of the stability studies as a function of the electrolyte pH carried out with all the compounds. A comprehensive cyclic voltammetry (CV) electrochemical study of the compounds was done in pH 6.0/ 1.0 M LiCH3COO + CH3COOH, since they were stable enough in this medium. Characterization of [As2W18O62]6− and [As2W15O56]12− at pH 6: Influence of the Extra Charge Created by the Departure of a W3O66+ Moiety. The purpose here was to compare the redox behavior between these two compounds and to highlight the observed similarities and/or differences. The electrochemical behavior of the two compounds has already been gauged in several previous studies,48,49 but this was the first time that such a comparative study was carried out at pH 6, in which the species [As2W18O62]6− is relatively

unstable and decays with time (Figure S4). At pH 6, the CV of the saturated compound [As2W18O62]6−, obtained immediately after solubilization and recorded between +0.6 and −1.2 V vs SCE, exhibited four single-electron, reversible redox waves, followed by a fifth reversible wave which involves the transfer of two electrons. Within the same potential range, the CV of the lacunary compound [As2W15O56]12− showed a first fourelectron, composite wave, followed by a two-electron, reversible redox wave. The two species [As 2 W 1 8 O 6 2 ] 6 − and [As2W15O56]12− exchange exactly the same number of electrons (six in total), but involving a different number of steps taking place at distinct redox potential values (Figure 3 and Table 3).

Figure 3. CVs of [As2W18O62]6− (black) and [As2W15O56]12− (red) in 1.0 M LiCH3COO + CH3COOH/pH 6.0. POM concentration: 0.4 mM. Working electrode: glassy carbon. Reference electrode: SCE. Auxiliary electrode: platinum. Scan rate: 10 mV·s−1.

Table 3. Midpoint Redox Potentials, E°′ = (Epa + Epc)/2 (Epa Being the Anodic Peak Potential and Epc the Cathodic Peak Potential), of the Compounds As2W18 and As2W15a As2W18 E°′1 V vs SCE no. of electrons

V vs SCE no. of electrons

0.07 1e−

E°′2 −0.11 1e− As2W15

E°′3

E°′4

E°′5

−0.50 1e−

−0.66 1e−

−0.89 2e−

E°′1

E°′2

E°′3

ca. −0.62 2e−

−0.68 2e−

−0.82 2e−

a

All the values were taken from the CVs recorded in 1.0 M LiCH3COO + CH3COOH/pH 6.0 at a scan rate of 10 mV·s−1.

The first redox step was observed at a midpoint redox potential, E°′, of +0.07 V vs SCE for the species [As2W18O62]6−, whereas for the lacunary compound [As2W15O56]12− a far more negative value had to be reached (Epc = −0.57 V vs SCE) for electron transfer to ensue. It is important to note that [As2W15O56]12− has a negative charge which is double that of [As2W18O62]6−, −12 vs −6. In conclusion, [As2W15O56]12− exhibits two partially merged two-electron redox steps (Epc = −0.57 V and −0.66 V vs SCE) which correspond to the four electrons exchanged by [As2W18O62]6− at the four single-electron redox steps taken together. Characterization of 1 at pH 6. Comparison with [As 2 W 15 O 56 ] 12− . The insertion of the metal cluster {(NaOH 2 ) 2 Co 2 O 14 } between two lacunary fragments [As2W15O56]12− led to marked changes in the CV of the latter when compared to that of the new sandwich-type compound [(NaOH2)2CoII2(As2W15O56)2]18− (1). The composite characG


Article

Inorganic Chemistry

Figure 4. (A) CVs of As2W15 (black) and 1 (red). (B) CVs of 1 showing two consecutive cycles with the potential range extended up to +1.2 V vs SCE on the positive side. (C) CVs of 1 showing consecutive cycling between +0.2 V and +1.2 V, the black arrow indicating the wave corresponding to Co2+ oxidation. (D) Electrocatalytic oxidation of water on a glassy carbon (blank) electrode in the absence (black) and in the presence (red) of 1. Electrolyte: 1.0 M LiCH3COO + CH3COOH/pH 6.0. POM concentration: 0.4 mM. Working electrode: glassy carbon. Reference electrode: SCE. Auxiliary electrode: platinum. Scan rate: 10 mV·s−1.

ter of the first reduction wave is more obvious in the CV of 1. In fact, the separation between the first two redox steps attributed to the reduction of WVI centers became more apparent when we compared the CV of compound [As2W15O56]12−, where there is a slight shoulder on the first wave, with that of 1, where two steps are clearly distinct (Figure 4A). However, when we compare the midpoint redox potentials corresponding to the first step for the two compounds, it is surprising to realize that 1 is reduced at less negative potentials than the [As2W15O56]12− species (see Table 4), despite the fact

and also a stronger basic character, resulting in a more pronounced influence of the proton concentration on its redox behavior. At higher pH values, its CV waves were more prone to splitting than those of [As2W15O56]12−. After the first reduction step, there was a redox potential inversion for the second and the third reduction steps, 1 becoming more difficult to reduce than [As2W15O56]12−, as expected. The scan range was extended up to +1.2 V vs SCE in order to probe the oxidation of the CoII centers. Three statements may be made regarding this aspect: (1) a redox couple corresponding to Co2+/3+ was observed at around +1.0 V vs SCE (Figure 4C); (2) the oxidation of the CoII centers led to the formation of oxide species which modify the surface of the working electrode, rendering the reduction process more sluggish upon the second scan cycle (blue curve in Figure 4B); (3) when the potential applied to the working electrode reached +1.35 V vs SCE, a rather high catalytic current was observed, indicative of the electrocatalytic oxidation of water by compound 1. Preliminary tests confirmed the evolution of dioxygen (see Supporting Information, Figures S5−S7), in good agreement with previous results obtained with Cocontaining POMs.50,51 Potential values as high as +1.5 V vs SCE had to be applied in order to observe similar current densities on a GC electrode in the absence of compound 1 (Figure 4D).50,51 Comparison of 1, 2, and 3. The comparison of the CVs of 1 and 2 (Figure 5A) shows that the first two reduction steps attributed to the W centers are identical for the two compounds. Bearing in mind the observations and conclusions

Table 4. Midpoint Redox Potentials, E°′ = (Epa + Epc)/2, for the Compounds [As2W15O56]12− and 1, and Epa and Epc Values for Couple Co2+/3+ in 1a V vs SCE

As2W15 1

E°′1

E°′2

E°′3

Epa

W(2e) −0.62 −0.61

W(2e) −0.68 −0.74

W(2e) −0.82 −0.90

Co2+/3+ +1.00

Epc

+0.97

a


that the latter has a smaller negative charge than 1 (−12 for [As2W15O56]12− vs −18 for 1). These two observations, which in the first place are not related nor expected, may be rationalized by differences in the acid/base properties of the two compounds, which mainly derive from their distinct electric charges.9,12 Compound 1 has a higher negative charge H


Article

Inorganic Chemistry

Figure 5. (A) CVs of 1 and 2. (B) CVs of 1 and 3. Electrolyte: 1.0 M LiCH3COO + CH3COOH/pH 6.0. POM concentration: 0.4 mM. Working electrode: glassy carbon. Reference electrode: SCE. Auxiliary electrode: platinum. Scan rate: 10 mV·s−1.

Compounds 1, 2, and 3 possess each a tetranuclear metal cluster respectively composed of 2 CoII ions and 2 NaI ions, 3 CoII ions and 1 NaI ion, and 4 CoII ions. As a consequence, the replacement of Co2+ by Na+ results in the reinforcement of the basic character of the POMs, the order of increasing basicity being 3 < 2 < 1. In the reduction processes of these compounds, electron transfer is concomitant with proton exchange and will be more favorable to occur (that is, at less reducing potential values) for compounds exhibiting a more pronounced basic character. Indeed, compounds 1 and 2 are more prone to take up protons than compound 3 and are therefore easier to reduce than the latter. Study of the Mixed Iron−Cobalt Compounds. The NaI centers in species 1 are rather labile in aqueous solution. Therefore, they may easily be replaced by Co2+ or Fe3+ ions, as described in the synthesis section, yielding compounds 3, 4, and 6. The basic characters of these three species are close to each other, as suggested by great similarity of their CVs (Figure 6A). The reduction waves of the WVI centers of the three compounds have both the same shape and very close potential peak values (Table 6). This behavior renders plausible the hypothesis mentioned in the previous paragraph concerning the influence of the composition of the central metal cluster both on the acid/base and on the redox properties of sandwich-type molecules of the tungstic Dawson family. In each of the three compounds, the central metal cluster surrounded by two [As2W15O56]12− fragments is composed of four “d” metal centers, both CoII and FeIII or just CoII, no alkaline cation being present (NaI in this case). Despite the different electric charges, the redox behavior attributed to the tungstic framework is the

mentioned above, this indicates that the two species have very similar acid/base properties. This was expected since these two compounds are very close as far as their chemical composition and electrical charge are concerned. Despite the fact that the first two reduction steps are almost superimposable, there is a slight difference on the third reduction step: the reduction wave obtained with compound 2 has a drawn out shape, indicative of a slow electron transfer process. The comparison of the CVs of compounds 1 and 3 reveals more marked differences, especially for the third reduction wave, the one of compound 3 having a rather sharper shape than those of the other two species (Figure 5B). Compound 3 possesses the smallest negative charge of all (−16 instead of −17 and −18 for 2 and 1, respectively), but unexpectedly is the most difficult to reduce in any of the three successive steps studied in the present work (Figures 5 and Table 5). Table 5. Midpoint Redox Potentials, E°′ = (Epa + Epc)/2, for Compounds 1, 2, and 3a V vs SCE

1 2 3

E°′1

E°′2

E°′3

W(2e) −0.61 −0.62 −0.62

W(2e) −0.74 −0.74 −0.77

W(2e) −0.90 −0.92 −0.96

a


Figure 6. (A) CVs of 3 (black), 4 (red), and 6 (blue). (B) CVs of 1 (black) and 5 (red). Electrolyte: 1.0 M LiCH3COO + CH3COOH/pH 6.0. POM concentration: 0.4 mM. Working electrode: glassy carbon. Reference electrode: SCE. Auxiliary electrode: platinum. Scan rate: 10 mV·s−1. I


Article

Inorganic Chemistry Table 6. Midpoint Redox Potentials, E°′ = (Epa + Epc)/2, for Compounds 3, 4, and 6a V vs SCE

3 4 6

E°′1

E°′2

E°′3

W(2e) −0.62 −0.62 −0.62

W(2e) −0.77 −0.76 −0.77

W(2e) −0.96 −0.95 −0.95

a


same for the three compounds. The presence of either one or two FeIII centers which are reduced just before the WVI centers (see Figure 6 and Tables 6 and 7) seems not to affect the Table 7. Midpoint Redox Potentials, E°′ = (Epa + Epc)/2, for Compounds 4, 5, and 6a

Figure 7. CVs of 4 (red), 5 (black), and 6 (blue) restricted to the first waves attributed to the reduction of FeIII centers. Electrolyte: 1.0 M LiCH3COO + CH3COOH/pH 6.0. POM concentration: 0.4 mM. Working electrode: glassy carbon. Reference electrode: SCE. Auxiliary electrode: platinum. Scan rate: 10 mV·s−1.

V vs SCE

4 5 6

E°′1

E°′2

Fe(1e)

Fe(1e) −0.17 ca. −0.21 −0.31

−0.06 −0.21

potential is scanned toward more negative values and according to the increase of their negative electric charge. Electrocatalytic Properties: Reduction of Nitrite. We carried out a comparative study between compounds 1, 2, and 3 regarding their electrocatalytic properties toward the reduction of nitrite ion. It mainly concerned cyclic voltammetry experiments in homogeneous phase in the medium 1.0 M LiCH3COO + CH3COOH/pH 5.0, so that the predominant species in solution was NO2− (HNO2 = H+ + NO2−, pKA = 3.2). All CVs were recorded at a scan rate of 2 mV s−1. Figures 8A, 8B, and 8C show the CVs of compounds 1, 2, and 3 in the absence and in the presence of an excess of nitrite ions. For the three species, the electrocatalytic reduction of nitrite is effective from the second reduction wave of the WVI centers onward, i.e., for potential values lower than −0.60 V vs SCE. This corresponds to a potential gain of about 0.5 V with respect to the reaction carried out on the same glassy carbon but without any of the POMs in solution (Figure 8D). The performance of each compound toward this electrocatalytic reaction is measured by the catalytic efficiency, ICAT, which corresponds to a normalized catalytic reduction current calculated by the following formula: ICAT = [(Ix − I0)/I0] × 100, I0 and Ix being the reduction currents at Ec = −0.85 V vs SCE obtained with the POM in the absence and in the presence of nitrite ions, respectively. The values in Table 8 show that ICAT increases with the number of CoII centers present in the molecule. It is rather surprising to realize that the CoII centers, which are not electroactive in the potential range explored, strongly influence the electrocatalytic reduction of nitrite ions. When compound 1 is replaced by compound 3, the value is multiplied by 4.25. Several previous works described the formation of nitrosyl adducts in the case of the reduction of nitrite by Fe3+-containing POMs. The formation of these adducts is concomitant with the reduction of the FeIII centers according to the electrochemical reaction represented by Fe3+ + e− = Fe2+. In the present case, we may emit the hypothesis that a complex forms in solution between NO2− and the CoII centers in the POM, rendering the reduction of the former more favorable.

a


position nor the shape of the reduction waves of the latter, as evidenced upon comparing with those obtained with the species having no FeIII centers. Ultimately, just the total number of “d” metal centers (2, 3, or 4) contained in the equatorial metal cluster of the three sandwich-type compounds will have a determining influence on the redox behavior of their tungstic centers. In fact, compounds 1 and 5, which have the same numerical and spatial distribution of “d” metal centers in their meridional clusters (2 “d” centers in internal positions and 2 Na+ centers in external positions, each binding a water molecule), give rise to CVs whose reduction waves attributed to the WVI centers have similar shapes and very close redox peak potential values (Figure 6B). Reduction of the FeIII Centers. The CVs of compounds 4 and 6 exhibit a reversible redox wave each whose reduction peak potentials are at −0.24 V and at −0.27 V vs SCE, respectively. This wave is undoubtedly attributed to the FeIII/II couple. Controlled potential coulometry carried out at the level of these waves confirmed that both one and two moles of electrons were required to reduce fully one mole of compounds 4 (one FeIII center) and 6 (two FeIII centers), respectively. This outcome was anticipated upon comparing the reduction peak current intensities measured on the respective CVs (Figure 6A and Figure 7), that of species 6 being roughly double that of compound 4. In Figure 7, the CVs of compounds 4, 5, and 6 are compared. The two FeIII centers present in compound 5 are reduced at more negative potentials than those of compound 6. The presence of two CoII ions decreases the negative electric charge of the polyanion from −16 to −14. As a consequence, compound 6 is more prone to accept extra electrons than compound 5. In other words, and taking into account just the redox processes attributed to the FeIII centers, the three compounds are reduced following the order 6 > 4 > 5, as the J


Article

Inorganic Chemistry

Figure 8. (A) CVs of 1 in the absence (black) and in the presence (red) of nitrite ion. (B) CVs of 2 in the absence (black) and in the presence (red) of nitrite ion. (C) CVs of 3 in the absence (black) and in the presence (red) of nitrite ion. (D) CVs of nitrite alone (black) and in the presence of either 1 (red), 2 (blue), or 3 (green). Electrolyte: 1.0 M LiCH3COO + CH3COOH/pH 5.0. POM concentration: 0.4 mM. Working electrode: glassy carbon. Reference electrode: SCE. Auxiliary electrode: platinum. Scan rate: 2 mV·s−1.

■

Table 8. ICAT Values for NO2− Electroreduction by Compounds 1, 2, and 3 at Ec = −0.85 V vs SCE ICAT (%)

1

2

3

154

434

657

CONCLUSION In this work, five new sandwich-type, tungstic Dawson POMs containing CoII and FeIII centers were synthesized. These new species were characterized by several physicochemical techniques, namely, IR and UV−vis spectrophotometry, ATG, elemental analysis, and single-crystal X-ray diffraction, which allowed determination or confirmation of their structures. Magnetic measurements evidence anisotropic ferromagnetic interactions between Co2+ ions and antiferromagnetic interactions between Fe3+ ions. The nature and the strength of the Co2+−Fe3+ interactions depends on the relative orientation of their 3d orbital. The effective magnetic moment of the Co2+ ions varies with the temperature and with the distortion of the octahedral sites in which they are located. On the one hand, UV−visible spectrophotometry allowed us to assess their stabilities in aqueous solutions as a function of the pH. We selected the medium 1 M CH3COOLi + CH3COOH/pH 6 as the most appropriate to carry out a comparative study of their electrochemical properties. On the other hand, UV−visible spectrophotometry was a convenient technique to evaluate the influence of metal cations such as CoII and FeIII on the values of the molar extinction coefficients, ε, of each compound. We were able to show that ε increased linearly with the number of CoII centers included in the molecule, but was not much influenced by the presence of FeIII centers. The cyclic voltammetry electrochemical study revealed the effect of the charge on the acid/base properties and on the redox behavior of the POM molecules. Controlled potential coulometry allowed us to determine the number of exchanged electrons per molecule at each redox process and to confirm the number of FeIII centers for the species containing them. All

Reduction of Dioxygen. The electrocatalytic reduction of dioxygen, O2, by the FeIII-containing compounds 4, 5, and 6 was studied in the medium 1.0 M LiCH3COO + CH3COOH/ pH 5.0. For these three compounds, the electrocatalytic reduction of dioxygen is effective on the wave attributed to the FeIII/II couple (Figure 9). The catalytic efficiency assessed by ICAT (Table 9) allows us to draw the following conclusions: (1) The number of FeIII centers present in the POM molecule has a direct influence on the efficiency toward the electrocatalytic reduction of dioxygen. In fact, species 4, which contains just a single FeIII center, exhibits the smallest ICAT value. (2) When the number of FeIII centers is the same, the presence of CoII centers in the metal cluster makes the difference: compound 6, which has two CoII centers associated with its two FeIII centers, is more effective than compound 5, which has no CoII centers whatsoever (Table 9). In addition, the electrocatalytic reduction onset potential values, Eonset, follow the same trend, that is, 6 > 5 > 4, meaning that the less negative value is that of compound 6. By and large, this corresponds to a potential gain between 0.2 and 0.3 V compared to the Eonset value obtained on a glassy carbon electrode in the same experimental conditions, but in the absence of the POM. K


Article

Inorganic Chemistry

Figure 9. (A) CVs of 4 in the absence (black) and in the presence of O2 (red). (B) CVs of 5 in the absence (black) and in the presence of O2 (red). (C) CVs of 6 in the absence (black) and in the presence of O2 (red). Electrolyte: 1.0 M LiCH3COO + CH3COOH/pH 5.0. POM concentration: 0.4 mM. Working electrode: glassy carbon. Reference electrode: SCE. Auxiliary electrode: platinum. Scan rate: 2 mV·s−1.

■

Table 9. ICAT Values for O2 Electroreduction by Compounds 4, 5, and 6 at Ec = −0.20 V vs SCE ICAT (%)

4

5

6

200

310

400

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Tel: +33-169-154-159. Fax: +33-169-156-188. ORCID

Israël M. Mbomekallé: 0000-0003-3440-8066 Present Address ∥

compounds turned out to be efficient for the electrocatalytic reduction of nitrite, and the ICAT values increased with the number of CoII centers present in the molecule. We have also proved that the presence of CoII centers in POMs possessing mixed FeIII and CoII metal clusters improved ICAT with respect to the reduction of dioxygen, the reaction taking place at the level of the FeIII/II couple wave.

■

M.L.: Département Scientifique Interfacultaire, Université des Antilles Campus de Schoelcher, B.P. 7209, Schoelcher, F97275, France.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the Centre National de la Recherche Scientifique (UMR 8000, 8180, and 8182), the University Paris-Sud, and the University of Versailles.

ASSOCIATED CONTENT

S Supporting Information *

■

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.inorgchem.6b02593.

REFERENCES

(1) Contant, R.; Herve, G. The heteropolyoxotungstates: relationships between routes of formation and structures. Rev. Inorg. Chem. 2002, 22 (2), 63−111. (2) POPE, M. T. Heteropoly and Isopoly Oxometalates; SpringerVerlag: Berlin, 1983. (3) Barsukova-Stuckart, M.; Izarova, N. V.; Jameson, G. B.; Ramachandran, V.; Wang, Z.; van Tol, J.; Dalal, N. S.; Ngo Biboum,

X-ray crystallographic file of 2 (CIF) X-ray crystallographic file of 3 (CIF) TGA and IR spectra, results and electrochemical characterization (PDF) L


Article

Inorganic Chemistry R.; Keita, B.; Nadjo, L.; Kortz, U. Synthesis and characterization of the dicopper(II)-containing 22-palladate(II)[Cu(II)2Pd(II)22P(V)12O60(OH)8]20. Angew. Chem., Int. Ed. 2011, 50 (11), 2639−2642. (4) Contant, R.; Thouvenot, R. A reinvestigation of isomerism in the Dawson structure: syntheses and 183W NMR structural characterization of three new polyoxotungstates [X2W18O62]6− (X = PV, AsV). Inorg. Chim. Acta 1993, 212 (1), 41−50. (5) Contant, R.; Thouvenot, R. Hétéropolyanions de type Dawson. 2. Synthèses de polyoxotungstoarsénates lacunaires dérivant de l’octadécatungstodiarsénate. Étude structurale par RMN du tungstène-183 des octadéca(molybdotungstovanado)diarsénates apparentés. Can. J. Chem. 1991, 69 (10), 1498−1506. (6) Bi, L.-H.; Wang, E.-B.; Peng, J.; Huang, R.-D.; Xu, L.; Hu, C.-W. Crystal Structure and Replacement Reaction of Coordinated Water Molecules of the Heteropoly Compounds of Sandwich-Type Tungstoarsenates. Inorg. Chem. 2000, 39 (4), 671−679. (7) Mbomekalle, I. M.; Keita, B.; Nadjo, L.; Neiwert, W. A.; Zhang, L.; Hardcastle, K. I.; Hill, C. L.; Anderson, T. M. Lacunary WellsDawson sandwich complexes - synthesis, characterization, and stability studies of multi-iron species. Eur. J. Inorg. Chem. 2003, 2003 (21), 3924−3928. (8) Mbomekalle, I. M.; Keita, B.; Nadjo, L.; Berthet, P.; Neiwert, W. A.; Hill, C. L.; Ritorto, M. D.; Anderson, T. M. Manganous heteropolytungstates. Synthesis and heteroatom effects in WellsDawson-derived sandwich complexes. Dalton Trans. 2003, 13, 2646− 2650. (9) Mbomekalle, I. M.; Keita, B.; Nadjo, L.; Berthet, P.; Hardcastle, K. I.; Hill, C. L.; Anderson, T. M. Multi-Iron Tungstodiarsenates. Synthesis, Characterization, and Electrocatalytic Studies of αββα(FeIIIOH2)2FeIII2(As2W15O56)212. Inorg. Chem. 2003, 42 (4), 1163−1169. (10) Mbomekalle, I. M.; Cao, R.; Hardcastle, K. I.; Hill, C. L.; Ammam, M.; Keita, B.; Nadjo, L.; Anderson, T. M. Synthesis, structural characterization, and electrocatalytic studies of αββα(ZnIIOH2)2(FeIII)2(X2W15O56)214- (X = P or As). C. R. Chim. 2005, 8 (6−7), 1077−1086. (11) Anderson, T. M.; Fang, X.; Mbomekalle, I. M.; Keita, B.; Nadjo, L.; Hardcastle, K. I.; Farsidjani, A.; Hill, C. L. Structural and electrochemical studies of dicupric Wells-Dawson sandwich-type complexes. J. Cluster Sci. 2006, 17 (2), 183−195. (12) Mbomekalle, I. M.; Mialane, P.; Dolbecq, A.; Marrot, J.; Sécheresse, F.; Berthet, P.; Keita, B.; Nadjo, L. Rational Synthesis, Structure, Magnetism and Electrochemistry of Mixed Iron-NickelContaining Wells-Dawson-Fragment-Based Sandwich-Type Polyoxometalates. Eur. J. Inorg. Chem. 2009, 2009 (34), 5194−5204. (13) Lebrini, M.; Mbomekalle, I. M.; Dolbecq, A.; Marrot, J.; Berthet, P.; Ntienoue, J.; Secheresse, F.; Vigneron, J.; Etcheberry, A. Manganese(III)-Containing Wells-Dawson Sandwich-Type Polyoxometalates: Comparison with their Manganese(II) Counterparts. Inorg. Chem. 2011, 50 (14), 6437−6448. (14) Long, D.-L.; Tsunashima, R.; Cronin, L. Polyoxometalates: Building Blocks for Functional Nanoscale Systems. Angew. Chem., Int. Ed. 2010, 49 (10), 1736−1758. (15) Kozhevnikov, I. V. Catalysis by Heteropoly Acids and Multicomponent Polyoxometalates in Liquid-Phase Reactions. Chem. Rev. 1998, 98 (1), 171−198. (16) Mizuno, N.; Misono, M. Heterogeneous Catalysis. Chem. Rev. 1998, 98 (1), 199−218. (17) Sadakane, M.; Steckhan, E. Electrochemical Properties of Polyoxometalates as Electrocatalysts. Chem. Rev. 1998, 98 (1), 219− 238. (18) Mü l ler, A.; Peters, F.; Pope, M. T.; Gatteschi, D. Polyoxometalates: Very Large ClustersNanoscale Magnets. Chem. Rev. 1998, 98 (1), 239−272. (19) Coronado, E.; Gómez-García, C. J. Polyoxometalate-Based Molecular Materials. Chem. Rev. 1998, 98 (1), 273−296. (20) Yamase, T. Photo- and Electrochromism of Polyoxometalates and Related Materials. Chem. Rev. 1998, 98 (1), 307−326.

(21) Rhule, J. T.; Hill, C. L.; Judd, D. A.; Schinazi, R. F. Polyoxometalates in Medicine. Chem. Rev. 1998, 98 (1), 327−358. (22) Keita, B.; Nadjo, L. Polyoxometalate-based homogeneous catalysis of electrode reactions: Recent achievements. J. Mol. Catal. A: Chem. 2007, 262 (1−2), 190−215. (23) Long, D.-L.; Burkholder, E.; Cronin, L. Polyoxometalate clusters, nanostructures and materials: From self assembly to designer materials and devices. Chem. Soc. Rev. 2007, 36 (1), 105−121. (24) Kortz, U.; Müller, A.; van Slageren, J.; Schnack, J.; Dalal, N. S.; Dressel, M. Polyoxometalates: Fascinating structures, unique magnetic properties. Coord. Chem. Rev. 2009, 253 (19−20), 2315−2327. (25) Keita, B.; Liu, T.; Nadjo, L. Synthesis of remarkably stabilized metal nanostructures using polyoxometalates. J. Mater. Chem. 2009, 19 (1), 19−33. (26) Wang, H.; Hamanaka, S.; Nishimoto, Y.; Irle, S.; Yokoyama, T.; Yoshikawa, H.; Awaga, K. In Operando X-ray Absorption Fine Structure Studies of Polyoxometalate Molecular Cluster Batteries: Polyoxometalates as Electron Sponges. J. Am. Chem. Soc. 2012, 134 (10), 4918−4924. (27) Keita Bineta, N. L. Encyclopedia of Electrochemistry: Electrochemistry of Isopoly and Heteropoly Oxometalates; Wiley VCH: 2006; Vol. 7. (28) Doungmene, F.; Aparicio, P. A.; Ntienoue, J.; Mezui, C. S. A.; de Oliveira, P.; Lopez, X.; Mbomekalle, I. M. Electrochemical behaviour of mixed d metal-iron containing Wells-Dawson sandwich-type complexes: [(FeOH2)2M2(X2W15O56)2]n- and [(MOH2)2Fe2(X2W15O56)2]n- (M = CrIII, MnIII, MnII, CoII, NiII, CuII, ZnII, X = AsV or PV and n = 12 or 14). Electrochim. Acta 2014, 125, 674− 682. (29) Sheldrick, G. M. SADABS: University of Göttingen: Germany, 1997. (30) Blessing, R. An empirical correction for absorption anisotropy. Acta Crystallogr., Sect. A: Found. Crystallogr. 1995, 51 (1), 33−38. (31) Sheldrick, G. M. SHELX-TL version 5.03; Siemens Analytical Xray Instrument Division: Madison, WI, USA, 1994. (32) Loose, I.; Droste, E.; Bösing, M.; Pohlmann, H.; Dickman, M. H.; Rosu, C.; Pope, M. T.; Krebs, B. Heteropolymetalate Clusters of the Subvalent Main Group Elements BiIII and SbIII ∥. Inorg. Chem. 1999, 38 (11), 2688−2694. (33) Reinoso, S.; Vitoria, P.; Gutiérrez-Zorrilla, J. M.; Lezama, L.; San Felices, L.; Beitia, J. I. Inorganic−Metalorganic Hybrids Based on Copper(II)-Monosubstituted Keggin Polyanions and Dinuclear Copper(II)−Oxalate Complexes. Synthesis, X-ray Structural Characterization, and Magnetic Properties. Inorg. Chem. 2005, 44 (26), 9731−9742. (34) Zhang, Z.; Qi, Y.; Qin, C.; Li, Y.; Wang, E.; Wang, X.; Su, Z.; Xu, L. Two Multi-Copper-Containing Heteropolyoxotungstates Constructed from the Lacunary Keggin Polyoxoanion and the HighNuclear Spin Cluster. Inorg. Chem. 2007, 46 (20), 8162−8169. (35) Vilà, N.; Aparicio, P. A.; Sécheresse, F.; Poblet, J. M.; López, X.; Mbomekallé, I. M. Electrochemical Behavior of α1/α2-[Fe(H2O)P2W17O61]7− Isomers in Solution: Experimental and DFT Studies. Inorg. Chem. 2012, 51 (11), 6129−6138. (36) Belombe, M. M.; Nenwa, J.; Mbiangue, Y.-A.; Nnanga, G. E.; Mbomekalle, I.-M.; Hey-Hawkins, E.; Loennecke, P.; Majoumo, F. Unusual aquation of Ba2+ ions in the solid state: synthesis and X-ray structural and spectroscopic characterization of the novel polymeric complex salt of empirical formula {Ba6(H2O)17[Cr(ox)3]4}·7H2O (ox = oxalate dianion). Dalton Trans. 2003, 11, 2117−2118. (37) Anderson, T. M.; Zhang, X.; Hardcastle, K. I.; Hill, C. L. Reactions of Trivacant Wells−Dawson Heteropolytungstates. Ionic Strength and Jahn−Teller Effects on Formation in Multi-Iron Complexes. Inorg. Chem. 2002, 41 (9), 2477−2488. (38) Mbomekalle, I. M.; Keita, B.; Nadjo, L.; Berthet, P.; Neiwert, W. A.; Hill, C. L.; Ritorto, M. D.; Anderson, T. M. Manganous heteropolytungstates. Synthesis and heteroatom effects in WellsDawson-derived sandwich complexes. Dalton Transactions 2003, 13, 2646−2650. M


Article

Inorganic Chemistry (39) Mbomekalle, I. M.; Keita, B.; Nadjo, L.; Berthet, P.; Hardcastle, K. I.; Hill, C. L.; Anderson, T. M. Multi-Iron Tungstodiarsenates. Synthesis, Characterization, and Electrocatalytic Studies of αββα(FeIIIOH2)2FeIII2(As2W15O56)212. Inorg. Chem. 2003, 42 (4), 1163−1169. (40) Keita, B.; Mbomekalle, I. M.; Lu, Yu W.; Nadjo, L.; Berthet, P.; Anderson, T. M.; Hill, C. L. Electron Transfer Behavior of Multi-Iron Sandwich-Type Polyoxometalates and Electrocatalytic Reduction Reactions. Eur. J. Inorg. Chem. 2004, 2004 (17), 3462−3475. (41) Gomez-Garcia, C. J.; Borras-Almenar, J. J.; Coronado, E.; Ouahab, L. Single-Crystal X-ray Structure and Magnetic Properties of the Polyoxotungstate Complexes Na16[M4(H2O)2(P2W15O56)2].cntdot.nH2O (M = MnII, n = 53; M = NiII, n = 52): An Antiferromagnetic MnII Tetramer and a Ferromagnetic NiII Tetramer. Inorg. Chem. 1994, 33 (18), 4016−4022. (42) Casan-Pastor, N.; Bas-Serra, J.; Coronado, E.; Pourroy, G.; Baker, L. C. W. First ferromagnetic interaction in a heteropoly complex: [CoII4O14(H2O)2(PW9O27)2]10-. Experiment and theory for intramolecular anisotropic exchange involving the four Co(II) atoms. J. Am. Chem. Soc. 1992, 114 (26), 10380−10383. (43) Clemente-Juan, J. M.; Coronado, E.; Gaita-Ariño, A.; GiménezSaiz, C.; Güdel, H.-U.; Sieber, A.; Bircher, R.; Mutka, H. Magnetic Polyoxometalates: Anisotropic Exchange Interactions in the Moiety of [(NaOH2)Co3(H2O)(P2W15O56)2]17. Inorg. Chem. 2005, 44 (10), 3389−3395. (44) Hou, Y.; Xu, L.; Cichon, M. J.; Lense, S.; Hardcastle, K. I.; Hill, C. L. A New Family of Sandwich-Type Polytungstophosphates Containing Two Types of Metals in the Central Belt: M′2M2(PW9O34)212− (M′ = Na or Li, M = Mn2+, Co2+, Ni2+, and Zn2+). Inorg. Chem. 2010, 49 (9), 4125−4132. (45) Goodenough, J. B. Magnetism and the Chemical Bond; Interscience (Wiley): New York, 1963. (46) Kahn, O. Molecular Magnetism; VCH: New York, 1993. (47) Clemente-Juan, J. M.; Coronado, E. Magnetic clusters from polyoxometalate complexes. Coord. Chem. Rev. 1999, 193−195, 361− 394. (48) Keita, B.; Mbomekalle, I. M.; Nadjo, L.; Contant, R. [H4AsW18O62]7-, a novel Dawson heteropolyanion and two of its sandwich-type derivatives [Zn4(H2O)2(H4AsW15O56)2]18-, [Cu4(H2O)2(H4AsW15O56)2]18-: cyclic voltammetry and electrocatalytic properties towards nitrite and nitrate. Electrochem. Commun. 2001, 3 (6), 267−273. (49) Mbomekalle, I. M.; Keita, B.; Nadjo, L.; Contant, R.; Belai, N.; Pope, M. T. Rationalization and improvement of the syntheses of two octadecatungstoarsenates: the novel α-K7[H4AsW18O62]·18H2O and the well known symmetrical α-K6[As2W18O62]·14H2O. Inorg. Chim. Acta 2003, 342, 219−228. (50) Soriano-López, J.; Goberna-Ferrón, S.; Vigara, L.; Carbó, J. J.; Poblet, J. M.; Galán-Mascarós, J. R. Cobalt Polyoxometalates as Heterogeneous Water Oxidation Catalysts. Inorg. Chem. 2013, 52 (9), 4753−4755. (51) Stracke, J. J.; Finke, R. G. Water Oxidation Catalysis Beginning with 2.5 μM [Co4(H2O)2(PW9O34)2]10−: Investigation of the True Electrochemically Driven Catalyst at ≥ 600 mV Overpotential at a Glassy Carbon Electrode. ACS Catal. 2013, 3 (6), 1209−1219.

N


Synthesis, Structure, and Magnetic ... - ACS Publications

Recommend Documents