Letter www.acsami.org
Efficient Catalytic Electrode for CO2 Reduction Realized by Physisorbing Ni(cyclam) Molecules with Hydrophobicity Based on Hansen’s Theory Masakazu Murase, Gaku Kitahara, Tomiko M. Suzuki, and Riichiro Ohta* Toyota Central R&D Labs., Inc., 41-1 Yokomichi, Nagakute, Aichi 480-1192, Japan S Supporting Information *
ABSTRACT: An electrochemical electrode physisorbed with Ni(cyclam) complex molecules containing tetraphenylborate ions (BPh4−) as counteranions shows catalytic activity for the reduction reaction of CO2 to CO in an aqueous electrolyte, superior to that of an electrode physisorbed with conventional [Ni(cyclam)]Cl2 complex molecules. The BPh4−-containing Ni(cyclam) is inferred as having high hydrophobicity based on its Hansen solubility parameter (HSP), with an interaction sphere excluding HSPs of water in a three-dimensional vector space. The high hydrophobicity of BPh4−-containing Ni(cyclam) molecules inhibits their dissolution into aqueous electrolyte and retains their immobilization onto the electrode surface, which we believe to result in the improved catalytic activity of the electrode physisorbed with them. HSP analysis also provides an optimized mixing ratio of solvents dissolving BPh4−-containing Ni(cyclam) molecules. KEYWORDS: nickel cyclam, Hansen solubility parameter, hydrophobicity, CO2 reduction, electrochemistry introducing tetraphenylborate ions (BPh4−) as counteranions to inhibit desorption of Ni(cyclam) molecules from the electrode surface in an aqueous electrolyte. We used the counterion replacement strategy, not alkylation, which has been reported to degrade catalytic activity for the CO2 reduction reaction.14 [Ni(cyclam) (MeCN)2](BPh4)2 (1) was synthesized as described in the Supporting Information. Details of the solubility of 1, including its high hydrophobicity, are explained based on Hansen solubility parameter (HSP) in the following. HSP generally represents the intrinsic dissolution property of substances, denoted by the three-dimensional (3D) vector composed of dispersion (δD), polarization (δP), and hydrogen bonding (δH) solubility parameters.17,18 The HSPs of various common chemical substances have been clarified by Hansen and other developers of HSP and have been listed in the commercial HSP calculation software.19 Using the substances on the list as probe solvents, the software predicts the unknown HSP of a substance of interest using the Hansen sphere method, drawing a sphere that includes good solvents and which excludes poor solvents with the so-called interaction radius R0 in 3D vector space, in which the sphere center is defined as its HSP.17,18
A
rtificial conversion of CO2 to valuable organic resources has been a research subject of great importance for solving the greenhouse problem. Attempts have been made to reduce CO2 using electrochemical or photochemical methods to produce organic substances.1 Recent studies have demonstrated that complexes with ions of several precious metals such as Ru, Re, and Rh can function as catalysts, lending high efficiency and selectivity to photochemical reactions reducing CO2 to CO or to formic acid.2−5 Electrochemical CO2 reduction reactions activated by catalytic complexes containing ions of abundant metals such as Mn, Fe, Co, and Ni, without using precious metals, have also been investigated to develop inexpensive and low-resource-risk systems.6−16 Among them, Ni(cyclam) (cyclam =1,4,8,11tetraazacyclotetradecane) complexes have been demonstrated to present high selectivity and low activation overpotential in a reduction reaction of CO2 to CO,11−16 in which Ni(cyclam) complex molecules have been dissolved in electrolytes and activated using mercury11−13 or glassy carbon14,15 as working electrodes. However, the reaction rates in the systems where Ni(cyclam) complex molecules are dissolved in the electrolytes depend strongly on the frequency with which Ni(cyclam) complex molecules approach the electrode surface sufficiently close to enable electron transfer, where two electrons are required to transfer by two individual steps in the reduction reaction of CO2 to CO.16 For this study, we prepared an electrode with enhanced efficiency by immobilizing Ni(cyclam) complex molecules onto a Sn disk surface, where Ni(cyclam) molecules were provided with hydrophobicity by simply © 2016 American Chemical Society
Received: June 27, 2016 Accepted: September 1, 2016 Published: September 8, 2016 24315
DOI: 10.1021/acsami.6b07665 ACS Appl. Mater. Interfaces 2016, 8, 24315−24318
Letter
ACS Applied Materials & Interfaces
water, and where R0 is the interaction radius of each complex. The HSPs of good solvents for each complex are expected to be located in the interior of the interaction sphere of each complex, with REDs expected to be 1.17,18 For [Ni(cyclam)]Cl2, one of the REDs from HSPs of water, which have been obtained previously by thermodynamic calculation,17,18 where the premise has been that water should exist as a single molecule without clustering,17,18 was 1.4, whereas the other REDs from HSPs of water, which have been obtained previously by dissolution experiments and Hansen sphere method,17,18 were 1, predicting its perfect immiscibility with water. We also confirmed experimentally that [Ni(cyclam)]Cl2 molecules were dissolved well in water, but 1 molecules were not dissolved at all. It is noteworthy that, particularly addressing REDs and Ra between HSPs of each complex and CO2 (see Table S1), although the RED (= 1.3) between HSPs of 1 and CO2 was slightly smaller than the RED (= 1.6) between HSPs of [Ni(cyclam)]Cl2 and CO2, the Ra (= 7.7 MPa1/2) between HSPs of 1 and CO2 was considerably smaller than the Ra (= 21.4 MPa1/2) between HSPs of [Ni(cyclam)]Cl2 and CO2, which might indicate that the affinity of 1 for CO2 is higher than that of [Ni(cyclam)]Cl2, possibly improving efficiency of the CO2 reduction reaction. 1 molecules were coated onto the surface of Sn disks by dispensing a droplet of solution dissolving 1 molecules on the surface and then drying out the solvent. We sought a solvent which can well-dissolve 1 and which can easily evaporate to facilitate the coating process by the RED between HSPs of 1 and solvents as follows. Acetonitrile (MeCN) and tetrahydrofuran (THF) are highly volatile solvents, but their REDs are, respectively, 0.99 and 1.45. By mixing MeCN and THF, the HSPs of the mixture solvents, which can be derived by internal division of the HSPs of each solvent in the volume mixing ratio, shifted the interior of HSP sphere of 1; therefore, the RED decreased and minimized at the ratio of 6:4 MeCN/THF (see Figure S4). We also experimentally attempted to dissolve 10 mM of 1 molecules and confirmed that MeCN left undissolved residue of 1 molecules and that THF was a poor solvent, whereas 6:4 MeCN/THF mixture solvent was able to dissolve 1 completely, which we used as the actual solvent for coating 1 molecules. The electrode in which 1 molecules were coated onto the Sn disk was catalytically active for the electrochemical reduction reaction of CO2 in an aqueous electrolyte, as shown in cyclic voltammograms (CVs) in Figure 2. In this figure, the irreversible peak that originated from the increase in current density (CD) below −0.95 V by CO2 reduction and the decrease in CD below −1.25 V by diffusion limitation of CO2 was observed only for the condition in which 1 molecules were coated and CO2 was present in the electrolyte, but not for the condition in which 1 molecules and CO2 were absent. The CD per molarity at the peak was approximately 3.7 × 103 mA cm−2 mM−1, which is 3 orders of magnitude larger than that reported previously for [Ni(cyclam)]Cl2 dissolved in an aqueous− organic electrolyte and evaluated using a glassy carbon electrode with similar conditions.14 The improved efficiency of reduction reaction for 1 is attributable to (i) promotion of electron-transfer between complex molecules and Sn surface, which is our initial aim, or (ii) enhancement of affinity of
Figure 1 depicts the 3D vector space, showing the calculated HSPs with their interaction spheres of 1 and [Ni(cyclam)]Cl2,
Figure 1. 3D vector space showing calculated HSPs of 1 (green plot) and [Ni(cyclam)]Cl2 (gray plot) with interaction spheres. Blue and black plots respectively indicate HSPs of good solvents for 1 and [Ni(cyclam)]Cl2, referred from the HSP software.19 Purple, orange, and red plots indicate the HSPs of water, referred from the HSP software,19 which have been obtained respectively by the following: (a) thermodynamic calculation; (b) Hansen sphere method, which has categorized solvents that can be dissolved more than 1% in water as good solvents; and (c) Hansen sphere method, which has categorized solvents having complete miscibility with water as good solvents.17,18
and HSPs of water, which have been calculated previously with three premises,17,18 referred from the HSP software.19 All HSPs of good and poor solvents for each complex are plotted in 3D vector spaces in Figures S2 and S3. The HSP of 1 was calculated as δD = 16.9 MPa1/2, δP = 13.4 MPa1/2, δH = 6.7 MPa1/2, R0 = 5.6 MPa1/2, where δH and R0 are smaller than those of [Ni(cyclam)]Cl2 (δD = 18.6 MPa1/2, δP = 14.3 MPa1/2, δH = 24.6 MPa1/2, R0 = 13.5 MPa1/2). The large δH of [Ni(cyclam)]Cl2 is attributable to the high hydrogen bonding ability of its chlorine sites or its ionic states after dissociation of chloride ions. The decrease in δH by replacement of Cl− to BPh4− would originate from the low hydrogen bonding ability of BPh4−. The solubility of each complex into water is explained below with relative energy differences (REDs = Ra/R0)17,18 between the HSPs of each complex and water shown in Table 1, where Ra is the distances between the HSPs of each complex and Table 1. Calculated HSPs of 1 and [Ni(cyclam)]Cl2, the HSPs of Water, Which Have Been Calculated Previously with Three Premises (a−c),17,18 Referred from the HSP Software,19 and REDs from HSPs of Water REDs from HSPs of watera
HSPs (MPa1/2) substances
δD
δP
δH
R0
(a)
(b)
(c)
1 [Ni(cyclam)]Cl2 water (a)19 water (b)19 water (c)19
16.9 18.6 15.5 15.1 18.1
13.4 14.3 16.0 20.4 17.1
6.7 24.6 42.3 16.5 16.9
5.6 13.5
6.4 1.4
2.2 0.9
2.0 0.6
18.1 13.0
a
Calculation methods: (a) thermodynamic calculation; (b) Hansen sphere method, 1% solubility; and (c) Hansen sphere method, complete miscibility.17,18 24316
DOI: 10.1021/acsami.6b07665 ACS Appl. Mater. Interfaces 2016, 8, 24315−24318
Letter
ACS Applied Materials & Interfaces
Figure 3. Differentiated CD (dCD) of negative sweep curves in CVs of (a) the bare Sn disk, (b) 1-coated Sn disk, and (c) [Ni(cyclam)]Cl2coated Sn disk electrodes in 0.1 M KCl, pH 10 aqueous electrolytes: scan rate = 0.1 V s−1.
Figure 2. CVs of the bare Sn disk and 1-coated Sn disk electrodes in 0.1 M KCl, pH 10 aqueous electrolyte: scan rate =0.1 V s−1. Peaks observed above −0.95 V for all conditions have been reported to appear because of oxidation and reduction of Sn.20 Further interpretation of these peaks is given in the Supporting Information. The increase in CD below approximately −1.25 V for bare Sn disk electrode was elevated more than three times by introducing CO2, which might be predominantly attributable to the promoted H2 production reaction by increase in H+ concentration, which was inferred based on the high faradaic efficiency of H2.
approximately −1.0 V for bare Sn disk electrode is mainly attributable to an elevated increase in H2 production rate, whereas the peak at approximately −1.15 V in dCD for the 1coated electrode corresponds to the inflection point of the peak appearing in CV, which originated from the reduction and diffusion limitation of CO2. However, the dCD for [Ni(cyclam)]Cl2-coated electrode comprised a small peak of reduction and diffusion limitation of CO2 at approximately −1.2 V, followed by an increase of dCD with a similar profile to that of the bare Sn disk electrode below approximately −1.3 V. Therefore, we assume that most [Ni(cyclam)]Cl2 molecules were desorbed from the Sn surface because of their hydrophilicity and were dissolved into the electrolyte; therefore, Ni(cyclam) units were activated only when they approached the Sn surface; the H2 production reaction typical for the bare Sn disk electrode was not inhibited. In conclusion, hydrophobicity provided by the introduction of BPh4− to Ni(cyclam) complex as counteranions enabled retention of the layer of physisorbed complex molecules on the Sn disk electrode surface in the aqueous electrolyte, which enhanced the efficiency of CO2 reduction reaction and which inhibited H2 production reaction on the Sn surface. Elucidation of interactions among species by HSP analysis should be useful for designing electrolyte solutions and electrodes to improve the efficiency of versatile electrochemical systems.
complex molecules for CO2, which was assumed by HSP analysis. Alternatively, (iii) the surface of Sn might have a superior effect on activating complex molecules compared to the surface of glassy carbon because the electrode in which 1 molecules were coated onto a glassy carbon disk was confirmed to be less active (see Figure S5). Products of CO2 reduction reactions were revealed by controlled potential electrolysis (CPE) at −1.4 V, which is sufficiently lower than −0.95 V where CO2 reduction reaction started in the CV profile, in the aqueous electrolyte dissolved with CO2, in combination with gas chromatography analysis. CO and H2 were detected in the products by CO2 reduction reactions for 1-coated Sn disk electrode, similarly to previously reported products by CO2 reduction reactions for [Ni(cyclam)]Cl2 dissolved in the aqueous−organic solutions.11−15 For the 1-coated Sn disk electrode, the faradaic efficiencies were estimated as approximately 60% for CO and approximately 35% for H2 after 30 min of CPE, whereas those for the bare Sn disk electrode under similar conditions were approximately 2% for CO and approximately 84% for H2. Therefore, the reduction reaction of H2O to produce a large amount of H2, typical for the bare Sn disk electrode, was relatively inhibited by coating 1 molecules, indicating that the layer of 1 molecules prevented the approach of H2O to the Sn surface, even though they are only physisorbed. The CO production rate was almost stable until at least 20 min of CPE (see Figure S6). To investigate the effect of hydrophobicity further, negative sweep curves in CVs of (a) the bare Sn disk, (b) 1-coated Sn disk, and (c) [Ni(cyclam)]Cl2-coated Sn disk electrodes were compared by differentiating CD to clarify the inflections in the curves, as shown in Figure 3. The [Ni(cyclam)]Cl2 molecules were coated the same molar amount as 1 molecules on a Sn disk surface using a similar dispensing/drying method conducted for coating 1 molecules, but using MeOH instead of MeCN/THF as a solvent. According to our interpretation of CVs discussed above, the continuous increase in dCD below
■
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.6b07665. Experimental details, crystal structure, 3D vector spaces with plots of HSPs of good and poor solvents, REDs from HSPs of the mixture solvents, CVs of 1-coated Sn and 1-coated glassy carbon, CO production amount, Ra and REDs from HSP of CO2 (PDF) Crystallographic information file for C66H76B2N8Ni (CIF)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. 24317
DOI: 10.1021/acsami.6b07665 ACS Appl. Mater. Interfaces 2016, 8, 24315−24318
Letter
ACS Applied Materials & Interfaces Notes
(14) Froehlich, J. D.; Kubiak, C. P. Homogeneous CO2 Reduction by Ni(cyclam) at a Glassy Carbon Electrode. Inorg. Chem. 2012, 51, 3932−3934. (15) Froehlich, J. D.; Kubiak, C. P. The Homogeneous Reduction of CO2 by [Ni(cyclam)]+: Increased Catalytic Rates with the Addition of a CO Scavenger. J. Am. Chem. Soc. 2015, 137, 3565−3573. (16) Song, J.; Klein, E. L.; Neese, F.; Ye, S. The Mechanism of Homogeneous CO2 Reduction by Ni(cyclam): Product Selectivity, Concerted Proton−Electron Transfer and C−O Bond Cleavage. Inorg. Chem. 2014, 53, 7500−7507. (17) Hansen, C. M. The Three-Dimensional Solubility ParameterKey to Paint Component Affinities. I. Solvents, Plasticizers, Polymers, and Resins. J. Paint Technol. 1967, 1, 505−510. (18) Hansen, C. M. Hansen Solubility Parameters A User’s Handbook, second ed.; CRC Press: Boca Raton, FL, 2007; pp 1−510. (19) HSPiP, fourth ed.; Hansen Solubility Parameters, http://www. hansen-solubility.com; 2013. (20) Kapusta, S. D.; Hackerman, N. Anodic Passivation of Tin in Slightly Alkaline Solutions. Electrochim. Acta 1980, 25, 1625−1639.
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors thank Mr. Satoru Kosaka for inductively coupled plasma optical emission spectrometry analysis explained in the Supporting Information and Dr. Daisuke Nakamura for thoughtful discussions related to HSP analysis. Crystal structure analysis and elemental analysis were conducted at the Instrument and Research Technology Center, Nagoya Institute of Technology.
■
REFERENCES
(1) Appel, A. M.; Bercaw, J. E.; Bocarsly, A. B.; Dobbek, H.; DuBois, D. L.; Dupuis, M.; Ferry, J. G.; Fujita, E.; Hille, R.; Kenis, P. J. A.; Kerfeld, C. A.; Morris, R. H.; Peden, C. H. F.; Portis, A. R.; Ragsdale, S. W.; Rauchfuss, T. B.; Reek, J. N. H.; Seefeldt, L. C.; Thauer, R. K.; Waldrop, G. L. Frontiers, Opportunities, and Challenges in Biochemical and Chemical Catalysis of CO2 Fixation. Chem. Rev. 2013, 113, 6621−6658. (2) Morimoto, T.; Nishiura, C.; Tanaka, M.; Rohacova, J.; Nakagawa, Y.; Funada, Y.; Koike, K.; Yamamoto, Y.; Shishido, S.; Kojima, T.; Saeki, T.; Ozeki, T.; Ishitani, O. Ring-Shaped Re(I) Multinuclear Complexes with Unique Photofunctional Properties. J. Am. Chem. Soc. 2013, 135, 13266−13269. (3) Sato, S.; Arai, T.; Morikawa, T.; Uemura, K.; Suzuki, T. M.; Tanaka, H.; Kajino, T. Selective CO2 Conversion to Formate Conjugated with H2O Oxidation Utilizing Semiconductor/Complex Hybrid Photocatalysts. J. Am. Chem. Soc. 2011, 133, 15240−15243. (4) Kou, Y.; Nabetani, Y.; Masui, D.; Shimada, T.; Takagi, S.; Tachibana, H.; Inoue, H. Direct Detection of Key Reaction Intermediates in Photochemical CO2 Reduction Sensitized by a Rhenium Bipyridine Complex. J. Am. Chem. Soc. 2014, 136, 6021− 6030. (5) Kobayashi, K.; Kikuchi, T.; Kitagawa, S.; Tanaka, K. Selective Generation of Formamides through Photocatalytic CO2 Reduction Catalyzed by Ruthenium Carbonyl Compounds. Angew. Chem., Int. Ed. 2014, 53, 11813−11817. (6) Bourrez, M.; Molton, F.; Noblat, S. C.; Deronzier, A. [Mn(bipyridyl) (CO)3Br]: An Abundant Metal Carbonyl Complex as Efficient Electrocatalyst for CO2 Reduction. Angew. Chem., Int. Ed. 2011, 50, 9903−9906. (7) Smieja, J. M.; Sampson, M. D.; Grice, K. A.; Benson, E. E.; Froehlich, J. D.; Kubiak, C. P. Manganese as a Substitute for Rhenium in CO2 Reduction Catalysts: The Importance of Acids. Inorg. Chem. 2013, 52, 2484−2491. (8) Costentin, C.; Drouet, S.; Robert, M.; Savéant, J. M. A Local Proton Source Enhances CO2 Electroreduction to CO by a Molecular Fe Catalyst. Science 2012, 338, 90−94. (9) Lacy, D. C.; McCrory, C. C. L.; Peters, J. C. Studies of CobaltMediated Electrocatalytic CO2 Reduction Using a Redox-Active Ligand. Inorg. Chem. 2014, 53, 4980−4988. (10) Chen, L.; Guo, Z.; Wei, X. G.; Gallenkamp, C.; Bonin, J.; Mallart, E. A.; Lau, K. C.; Lau, T. C.; Robert, M. Molecular Catalysis of the Electrochemical and Photochemical Reduction of CO2 with EarthAbundant Metal Complexes. Selective Production of CO vs HCOOH by Switching of the Metal Center. J. Am. Chem. Soc. 2015, 137, 10918−10921. (11) Fisher, B. J.; Eisenberg, R. Electrocatalytic Reduction of Carbon Dioxide by Using Macrocycles of Nickel and Cobalt. J. Am. Chem. Soc. 1980, 102, 7361−7363. (12) Beley, M.; Collin, J. P.; Ruppert, R.; Sauvage, J. P. Electrocatalytic Reduction of CO2 by Ni Cyclam2+ in Water: Study of the Factors Affecting the Efficiency and the Selectivity of the Process. J. Am. Chem. Soc. 1986, 108, 7461−7467. (13) Collin, J. P.; Jouaiti, A.; Sauvage, J. P. Electrocatalytic Properties of Ni(cyclam)2+ and Ni2(biscyclam)4+ with Respect to CO2 and H2O Reduction. Inorg. Chem. 1988, 27, 1986−1990. 24318
DOI: 10.1021/acsami.6b07665 ACS Appl. Mater. Interfaces 2016, 8, 24315−24318
