Unusual Redox Catalysis in a Ruthenium Oxide±Prussian Blue Combined Material Annamalai Senthil Kumar and Jyh-Myng Zen*[a] RuVIIOx is (E8 = 1.35±1.45 V vs. RHE; RHE = reversible hydrogen electrode) a high-valent oxidant with the potential for use in various catalytic oxidations, including functional-group transformation in organic synthesis.[1±3] It was reported to be stable only in strong alkaline (pH > 12) solution and/or nonaqueous solutions of N-methyl-morpholine-N-oxide- or tetra-n-propylammonium perruthenate ionic compounds.[2, 3] In acidic solutions, RuVIIOx tends to convert into relatively stable lower oxidation states such as RuVIO3 and RuIVO2 by a disproportionation reaction, as illustrated in Figure 1.[2] This disproportionation re-

Figure 1. Conceptional representation of the dynamics of the oxyruthenium redox transitions against the solution pH, and their specific oxidative functions towards organic redox probes. The RuVII is stable only in strongly alkaline pH environments, while at pH < 12 it disproportionates to lower redox states. Note that only RuVII can oxidize all three organic compounds tested.

action is the reason that RuVII and its derived materials are restricted to the above-mentioned conditions for their applications. We report here the first evidence for the unusual electrogeneration and stabilization of RuVII in acidic medium in a ruthenium oxide±Prussian blue-based combined material (RuOx±PB) without a disproportionation±decomposition reaction. To our knowledge, this is the first observation of the stabilization of highly valent RuVII in acidic media. One of the RuOx±PB analogues, a mixed valent ruthenium oxide/ruthenium cyanide (mvRuOx±RuCN), has been taken [a] A. S. Kumar, Prof. J.-M. Zen Department of Chemistry, National Chung Hsing University Taichung 402 (Taiwan) Fax: (+ 886) 4-22862547 E-mail: [email protected]

ChemPhysChem 2004, 5, 1227 ±1231

DOI: 10.1002/cphc.200400068

¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

1227

here as a model system to explain the observed unique elecent pH values.[2, 7] It is noteworthy that both RuOx- and PBtrochemical behavior. The mvRuOx±RuCN was prepared by based compounds were ineffective for direct carbohydrate electrodeposition, using RuCl3 + [Ru(CN)6]4
in a pH = 2 KCl/HCl (glucose as a model here) oxidation in acidic media. In this study, CV responses towards the oxidation of glucose bath solution, in a similar procedure to that used for Prussian in pH = 2 Na2SO4/H2SO4 solution in systems of RuOx/SPE, PB/ blue (i.e., with FeCl3 + [Fe(CN)6]3
), which is used for various electrocatalytic applications.[4] Prussian blue (PB,
FeIII/II
CN
SPE, and SPE with [Ru(CN)6]4
in solution were compared. The FeIII/II
), a polynuclear mixed-valent compound with a zeolitemeasurements were carried out using a CHI 400 electrochemical analyzer together with SPE working electrodes, platinum type ordered framework,[5] was essential as a host for the stabi(0.07 cm2) counter electrodes, and Ag/AgCl reference electrolization of high-oxidation-state ruthenium oxide in acidic media. In brief, mvRuOx-RuCN was prepared according to two des. As shown in Figure 2, there were no oxidative responses different approaches;[4f, g] 1) by potential cycling (from
0.5 to for glucose in the individual systems of RuOx/SPE, PB/SPE, and SPE with [Ru(CN)6]4
in solution. These results indicate that the 1.2 V vs. Ag/AgCl at 20 mV s
1 for 20 cycles) and 2) by application of potentiostatic (at 1.1 V for 900 s) conditions on a above catalysts are not powerful enough to promote the glucleaned screen-printed carbon electrode (SPE, geometric cose oxidation. The RuOx±PB/SPE system, however, shows a area = 0.196 cm2) in 2 mm each of RuCl3 and [Ru(CN)6]4
with clear signal of glucose catalysis at  1.1 V vs. Ag/AgCl (i.e., A2/ C1). This is the first observation of direct glucose oxidation 0.3 m KCl/HCl (optimized), adjusted to pH 2. Both methods rewith RuOx-based materials at acidic pH values. According to sulted in almost similar electrochemical and catalytic function Figure 1, the RuOx±PB combinative material must undergo an (verified with glucose, ethanol, and formaldehyde); but apunusual electrogeneration of the high valent RuVIIOx state at proach 2 yielded a material with a more stable cyclic voltammetric (CV) response and hence this approach was uniformly  1.1 V and acidic pH values without a classical disproportionaadopted throughout these studies. Classical RuOx and PB were prepared using RuCl3 and FeCl3 + [Fe(CN)6]3
, similar to the reported procedures,[5d, 6] at the above bath standard conditions. Probing of the electrogenerated redox states in RuOx is an intriguing and challenging aspect of electrochemistry.[7] As reported earlier, a specific redox state of ruthenium gave selective in situ oxidation of carbohydrates, ethanol, and formaldehyde.[7b] Carbohydrates can be selectively oxidized by RuO2 and RuOx materials only under strong alkaline conditions, because of the limited availability of RuVII in these conditions (Figure 1).[8] Aliphatic alcohols (e.g., ethanol and methanol) and formaldehyde, on the other hand, can be oxidized in a wide pH range since even the lower RuVI oxidation state is enough to catalyze the reaction.[7] Hence, positive interactions between carbohydrates with a RuVII redox state can be taken here as a key piece of indirect evidence for the unusual characteristics of the RuOx±PBbased combined material. Figure 1 sketches the well-estab- Figure 2. Electrochemical evidences for the unusual redox4
behavior of RuOx±PB. Cyclic voltammetric responses for various SPE systems of PB/SPE, RuOx/SPE, SPE with [Ru(CN)6] in solution and RuOx±PB/SPE in the presence/absence of lished correlation between vari0.1 m glucose at a scan rate = 10 mV s
1 in pH 2 Na2SO4/H2SO4 solution (I = 0.1 m). The mvRuOx±RuCN is taken as a ous organic probes and specific model for the RuOx±PB system. Our new proposed structure of RuOx±PB (i.e., mvRuOx±RuCN) is sketched on the ruthenium redox states at differ- bottom right-hand side, beneath other conventional models.

1228

¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemphyschem.org

ChemPhysChem 2004, 5, 1227 ±1231

tion±decomposition reaction, and with superior performance. Indeed, the fact that the catalytic oxidation potential observed here at  1.1 V vs. Ag/AgCl (i.e., 1.44 V vs. RHE {ERHE = [(EAg/AgCl + 0.06pH) + E8Ag/AgCl(0.22)]} lies exactly inside the reported redox window of the RuVIIO4
/RuVIO42
(perruthenate/ ruthenate) redox state with RuO2 anodes under alkaline conditions provides strong support to our above explanation.[2f, 7a] Meanwhile, in addition to glucose, both ethanol and formaldehyde (Figure 1) can also be oxidized with the RuOx±PB, with a sensitivity of  30 mA mm
1 (scan rate = 10 mV s
1) in pH = 2 solution. All the above electrochemical observations support the electrogeneration and stabilization of the unusual RuVII/VI redox state with RuOx-PB combinative material without a classical disproportionation±decomposition in acidic medium. The basic concept can also be illustrated using a ™balloon∫ model, as shown in Figure 3. The energized balloon expands

Figure 4. Typical potential segment analysis of the mvRuOx±RuCN in 0.1 m Na2SO4/H2SO4 of pH 2 solution at a scan rate of 10 mV s
1. Potentials were scanned in anodic (a) or cathodic directions (b±e). The elective cross-over reaction in c) implies electrogeneration of > RuVII=O coupled with a slow H2O oxidation reaction.

Figure 3. Balloon model cartoon explaining the concept of the unusual glucose catalysis and the novel stabilization of RuVII with RuOx±PB in acid media. The high-energy RuVII ions electrogenerated at 1.1 V vs. Ag/AgCl are stabilized in the RuOx±PB analogue, while they disproportionate±decompose to lower oxidation states with the classical RuOx matrix. The additional CN- linkage is uniquely found for the stabilization effect at RuOx±PB. G = glucose, GO = gluconolactone.

when an amount of air, x, is introduced. An externally protected balloon can retain its energetic state. Similarly, at 1.1 V (i.e., x = 1.1 V), RuOx is in a highly energetic RuVIIOx level but is converted into lower oxidation states, RuVIOx and RuIVO2, in acid media and hence no catalysis occurs with glucose (Figure 2 b). The RuVIIOx state can only be stabilized in the RuOx±PB combined matrix, which assists the glucose oxidation (Figure 2 d). Note that the
CN
linkages in RuOx±PB are unique in this observed stabilization effect. Potential segment CV analysis at a slow scan rate also implies the existence of high valent RuVII oxidation state in mvRuOx-RuCN. As can be seen in Figure 4 c, instead of a reversible/quasireversible response, a crossover type of electrochemistry was noticed especially in the potential window of 1.2!0.8 V in a pH = 2 blank solution. The electrogenerated RuVIIOx slowly oxidized H2O in the absence of analyte; this is similar to the behavior of solution-phase perruthenate ions (RuVIIO4
) under alkaline conditions.[2d] Control experiments with RuOx and PB analogues did not show any such behavior. X-ray photoelectron spectroscopic (XPS) studies can provide direct evidence to support our hypothesis (Figure 5). The ChemPhysChem 2004, 5, 1227 ±1231 www.chemphyschem.org

measurements (American Physical Electronics) were carried out with both mvRuOx±RuCN/SPE and pelleted RuO2 powder (Aldrich) samples and the binding energies of 397.8 eV (-CN-, N 1s) and 280.9 eV (RuO2, Ru 3d5/2) were taken as internal standards.[9a, b, 4g] The potentiostatic preparation (i.e., case 2) condition of mvRuOx±RuCN at 1.1 V (where the RuVII is electrogenerated) was found to store a fraction of the RuVII intermediates in the surface of the electrode. During the electrodeposition condition the SPE promotes mvRuOx±RuCN nucleation, the coupled H2O oxidation was thus absent at this stage. No external potential was supplied during the XPS measurements. After the mvRuOx±RuCN film preparation, the system was removed from the bath and quickly washed and stored in a dry box. High resolution XPS analysis indicated two types of important linkages; -Ru-O-Ru- (O 1s at 531.8 eV)[9c] and -CN- (N 1s at 397.8 eV) with the RuOx±PB combinative system (Figure 4); while the RuO2 spectrum contains the response from RuIVO2 (Ru 3d5/2 = 280.9 eV, O 1 s = 529.6 eV) and RuVIO3 (Ru 3d5/2 = 282.1 eV, O 1 s = 530.7 eV) traces only.[9a, b] Some of the Ru 3d5/2 fractions were specifically identified at the high binding energy of 283.0 eV with the mvRuOx±RuCN material; this value lies between the reported values for the RuVIOx (282.1 eV) and RuVIIIOx (283.3 eV) species.[9a] This observation strongly supports the existence of > RuVII=O type (as an oxo-complex) highvalent species, unusually, within the RuOx-PB system. The high-resolution O 1s spectra also provide evidence for the existence of strong electronegative oxygen atoms (533.3 eV) that

¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

1229

C2’ and A1/C1) did not participate in any catalysis in this work, which further implies that the redox states must be lower than the RuVI/RuIV level. Explanations of the behavior of A2/C2 are highly inconsistent in the literature as it varies from RuIIIOx to RuVIOx.[4] Hence the unusual catalysis and novel stabilization of > RuVII=O in this work opens up a new avenue for catalytic applications. In conclusion, we identified an unusual high-valent RuVII/VI Figure 5. Evidence for the > RuVII=O trace in the mvRuOx±RuCN network. High-resolution XPS of the RuOx-PB/SPE (i.e., redox state in a ruthenium mvRuOx-RuCN) and pelleted RuO2 powder samples confirms the population of the Ru 3d, N 1s and O 1s energy levels. oxide±Prussian blue-based maThe mvRuOx±RuCN was prepared at SPE by potentiostatic polarization condition by maintaining the potential at terial in acidic media, without a 1.1 V vs. Ag/AgCl (where RuVII was electrogenerated) of 900 s in 2 mm each of RuCl3 + [Ru(CN)6]4
containing 0.3 m KCl/ classical disproportionation±deHCl-pH 2 bath solution (case 2). composition reaction. The achievement is based on the correspond to the > RuVII/VI=O species, in addition to the surknowledge of two independent studies of PB and RuO2. The face -OH and -H2O (534.4 eV) moieties. These surface functional three unique bridging features, -CN-, -Ru-O-Ru-, and the strong internal hydrogen bonding, make the RuOx±PB comgroups should provide strong intermolecular hydrogen bondbined material more rigid and free from disproportionation± ing within the internal structure (Figure 2). Overall, the -CN-, decomposition reaction. The marked catalytic activity of this -Ru-O-Ru- and strong internal hydrogen bonding effects confer system offers a good opportunity to extend into diverse applimore integrity and rigidity on the structure of the RuOx±PB. cation fields of organic synthesis, chemical sensor, fuel cell, etc. In this unique and stable structure, the electrogenerated Numerous applications can readily be imagined from this > RuVII=O can be retained in the internal network without the study. classical disproportionation±decomposition reaction. The mvRuOx±RuCN film prepared under potential cycling conditions (as in case 1) from
0.20 to 1.10 V did not show Acknowledgements signs of RuVII.[4g] The final potential of
0.20 V may reduce RuVII VI to Ru . Kˆtz et al studied the Ru oxidation states on RuO2/Ti Financial support of this work provided by the National Science anodes by XPS under different electrochemical conditions.[9d] Council of Taiwan is gratefully acknowledged. Prior to the XPS measurements, the anodes were subjected to different applied potentials from 0 to 2 V vs. SCE (i.e.,
0.03 to 1.98 V vs. Ag/AgCl) in 0.5 m H2SO4 for 15 min. An average value Keywords: catalysis ¥ electrochemistry ¥ glucose ¥ oxidation of 280 eV was obtained for the Ru 3d5/2 core energy level while states ¥ Prussian blue ¥ ruthenium oxide potentials were maintained in the window of 0 to 1.1 V. After [1] M. Hudlicky¬, Oxidations in Organic Chemistry. ACS monograph 186, Amerthis potential window, the energy level value shifted to a maxiican Chemical Society, Washington, DC, 1990. mum of 282.4 eV, which corresponds to the RuIV and RuVI oxi[2] a) R. E. Connick, C. R. Hurley, J. Am. Chem. Soc. 1952, 74, 5012; b) M. D. dation states. This observation rules out the possibility of the Silverman, H. A. Levy, J. Am. Chem. Soc. 1954, 79, 3319; c) R. P. Larsen, stabilization of > RuVII=O with classical systems, according to L. E. Ross, Anal. Chem. 1959, 31, 176; d) K. W. Lam, K. E. Johnson, D. G. Figures 1 and 3. Lee, J. Electrochem. Soc. 1978, 125, 1069 ± 1076; e) A. J. Bailey, W. P. Griffith, S. I. Mostafa, P. A. Sherwood, Inorg. Chem. 1993, 32, 268; f) A. J. Bard, Figure 2 also sketches the proposed network structure for R. Parsons, J. Jordan, Standard Potentials in Aqueous Solutions, IUPAC, the presented system in comparison with PB, RuOx and Marcel Dekker, New York, 1983, p. 413; g) A. S. Kumar, K. C. Pillai, J. Solid [Ru(CN)6]4
. The CV signal peaks at 0.0 (A1/C1), 0.8 (A2’/C2’) State Electrochem. 2000, 4, 408. and 1.1 V (A2/C2) are three different redox pairs of the [3] a) W. P. Griffith, J. M. Jolliffe, S. V. Ley, D. J. Williams, J. Chem. Commun. 1990, 1219; b) A. Goti, M. Romani, Tetrahedron Lett. 1994, 35, 6567; mvRuOx±RuCN system. Based on XPS and classical PB studies,[5] c) K. J. Tony, M. V. Rajaram, C. S. Swamy. React. Kinet. Catal. Lett. 1997, 62, III/II A1/C1 and A2’/C2’ correspond to the -Ru - species from the 105; d) R. Lenz, S. V. Ley, J. Chem. Soc. Perkin Trans. 1 1997, 3291; e) B. Hi™-RuIII/II-N∫ and ™-RuIII/II-C∫ environments, respectively. It is obvinezen, S. V. Ley. J. Chem. Soc. Perkin Trans. 1 1997,1907; f) Y. Tokunaga, ous that the other A2’/C2’ redox behavior is rather reversible M. Ihara, K. Fukumoto, J. Chem. Soc. Perkin Trans. 1 1997, 207; g) B. Hinezen, S. V. Ley. J. Chem. Soc. Perkin Trans. 1 1998, 1 ± 2; h) I. E. Marko, A. for the solution [Ru(CN)6]3
/4. A2/C2 is unique and may origiGautier, S. M. Brown, Angew. Chem. 1999, 111, 2126; I. E. Marko, A. GautiVII/VI nate from the separate oxo/hydroxy-Ru species interlinked er, S. M. Brown, Angew. Chem. Int. Ed. 1999, 38, 1960; i) K. R. Flower, A. P. in ™-Ru-O-Ru-∫, as shown in the proposed model structure Lightfoot, H. Wan, A. Whiting, Chem. Commun. 2001, 1812; j) R. Cirimin(Figure 2). Except A2/C2, the lower oxidation states (i.e., A2’/ na, M. Pagliaro, Chem. Eur. J. 2003, 9, 5067.

1230

¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemphyschem.org

ChemPhysChem 2004, 5, 1227 ±1231

[4] a) L. Huang, H. Shen, M. A. Atkinson, R. T. Kennedy, Proc. Natl. Acad. Sci. USA 1995, 92, 9608; b) J. A. Cox, P. J. Kulesza, Anal. Chem. 1984, 56, 1021; c) J. A. Cox, R. K. Jaworski, P. J. Kulesza, Electroanalysis 1991, 3, 869; d) P. J. Kulesza, M. Bandoch, J. Electroanal. Chem. 1992, 323, 131; e) T. R. I. Cataldi, C. Campa, D. Centonze, Anal. Chem. 1995, 67, 740; f) T. R. I. Cataldi, D. Centonze, A. Guerrieri, Anal. Chem. 1995, 67, 101; g) T. R. I. Cataldi, A. M. Salvi, D. Centonze, L. Sabbatini, J. Electroanal. Chem. 1996, 406, 91; h) P. J. Kulesza, B. Grzybowska, M. A. Malik, M. Chojak, K. Miecznikowski, J. Electroanal. Chem. 2001, 512, 110; i) S.-M. Chen, S.-H. Hsueh, J. Electroanal. Chem. 2004, 566, 291; j) B. Loetanantawong, C. Suracheep, W. Surareungchai, Anal. Chem. 2004, 76, 2266. [5] a) K. Itaya, I. Uchida, V. D. Neff, Acc. Chem. Res. 1986, 19, 162; b) O. Kahn, Nature, 1999, 399, 21; c) N. R. de Tacconi, K. Rajeshwar, Chem. Mater. 2003, 15, 3046; d) A. S. Kumar, P.-Y. Chen, J.-M. Zen, Electroanalysis 2004, 16, 242. [6] C.-C. Hu, Y.-H. Huang, J. Electrochem. Soc. 1999, 146, 2465. [7] a) L. D. Burke, J. F. Healy, J. Electroanal. Chem. 1981, 124, 327; b) J.-M. Zen, A. S. Kumar, J.-C. Chen, J. Mol. Catal. A: Chemical. 2001, 126, 177, and references therein; c) J.-M. Zen, A. S. Kumar, Acc. Chem. Res. 2001, 34, 772. [8] a) M. E. G. Lyons, C. A. Fitzgerald, M. R. Smyth, Analyst 1994, 119, 855; b) J. Wang, Z. Taha, Anal. Chem. 1990, 62, 1413; c) V. Dharuman, K. C. Pillai, Bull. Electrochem. 1999, 15, 476; d) H. S. El Khadem, Carbohydrate Chemistry, Academic Press, New York, 1988, p. 144. [9] a) K. S. Kim, N. Winograd, J. Catal. 1974, 35, 66; b) J. Augustynski, L. Balsenc, J. Hinden, J. Electrochem. Soc. 1978, 125, 1093; c) A. Mekki, D. Holland, C. F. McConville, M. Salim, J. Non-cryst. Solids 1996, 208, 267; d) R. Kˆtz, H. J. Lewerenz, S. Stucki, J. Electrochem. Soc. 1983, 130, 825. Received: February 19, 2004 Revised: June 1, 2004

ChemPhysChem 2004, 5, 1227 ±1231 www.chemphyschem.org

¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

1231