Modification of the Surface Adsorption Properties of Alumina

The electrocatalytic hydrogenation (ECH) of phenol has been studied using palladium supported on γ-alumina (10% Pd−Al2O3) catalysts. The catalyst p...
0 downloads 0 Views 266KB Size
Download PDF
6414

Langmuir 2006, 22, 6414-6421

Modification of the Surface Adsorption Properties of Alumina-Supported Pd Catalysts for the Electrocatalytic Hydrogenation of Phenol Ciprian Mihai Cirtiu, Hicham Oudghiri Hassani, Nicolas-Alexandre Bouchard, Paul A. Rowntree,* and Hugues Me´nard* De´ partement de Chimie, Centre de Recherche en EÄ lectrochimie et EÄ lectrocatalyse, UniVersite´ de Sherbrooke, Sherbrooke, Que´ bec, Canada J1K 2R1 ReceiVed July 13, 2005. In Final Form: April 11, 2006

The electrocatalytic hydrogenation (ECH) of phenol has been studied using palladium supported on γ-alumina (10% Pd-Al2O3) catalysts. The catalyst powders were suspended in aqueous supporting electrolyte solutions containing methanol and short-chain aliphatic acids (acetic acid, propionic acid, or butyric acid) and were dynamically circulated through a reticulated vitreous carbon cathode. The efficiency of the hydrogenation process was measured as a function of the total electrolytic charge and was compared for different types of supporting electrolyte and for various solvent compositions. Our results show that these experimental parameters strongly affect the overall ECH efficiency of phenol. The ECH efficiency and yields vary inversely with the quantity of methanol present in the electrolytic solutions, whereas the presence of aliphatic carboxylic acids increased the ECH efficiency in proportion to the chain length of the specific acids employed. In all cases, ECH efficiency was directly correlated with the adsorption properties of phenol onto the Pd-alumina catalyst in the studied electrolyte solution, as measured independently using dynamic adsorption isotherms. It is shown that the alumina surface binds the aliphatic acids via the carboxylate terminations and transforms the catalyst into an organically functionalized material. Temperature-programmed mass spectrometry analysis and diffuse-reflectance infrared spectroscopy measurements confirm that the organic acids are stably bound to the alumina surface below 200 °C, with coverages that are independent of the acid chain length. These reproducibly functionalized alumina surfaces control the adsorption/desorption equilibrium of the target phenol molecules and allow us to prepare new electrocatalytic materials to enhance the efficiency of the ECH process. The in situ grafting of specific aliphatic acids on general purpose Pd-alumina catalysts offers a new and flexible mechanism to control the ECH process to enhance the selectivity, efficiency, and yields according to the properties of the specific target molecule.

Introduction Chemisorbed hydrogen formed on electrocatalysts during the electrolysis of water may be used for the hydrogenation of functional groups of organic species. It is known that the efficiency of electrocatalytic hydrogenation (ECH) depends sensitively on the nature of the electrode materials.1-3 Metals such as Pd, Ni, Rh, and Pt are most often used for the ECH of phenol,4,5 the target molecule considered in this work. In the ECH process, chemisorbed atomic hydrogen (MHads) is formed in situ by the electroreduction of water (i.e., the Volmer reaction: reaction 1) on the metal surface. The freshly generated chemisorbed hydrogen can then react with coadsorbed unsaturated organic target species (YdZ) to form hydrogenated organic molecules. The competing hydrogen evolution reactions (HER) (i.e., Heyrovsky and Tafel steps: reactions 2 and 3) become the dominant processes when the organic molecule is thermodynamically or kinetically difficult to hydrogenate. * Corresponding authors. (P.A.R.) E-mail: [email protected]. Tel: (819) 821-7006. Fax: (819) 821-8017. (H.M.) E-mail: hugues.menard@ usherbrooke.ca. Tel: (819) 821-7084. Fax: (819) 821-8017. (1) Amouzegar, K.; Savadogo, O. Electrochim. Acta 1994, 39, 557. (2) Martel, A.; Mahdavi, B.; Lessard, J.; Brossard, L.; Me´nard, H. Can. J. Chem. 1997, 75, 1862. (3) Chapuzet, J. M.; Lasia, A.; Lessard, J. In Electrocatalysis; Lipkowski, J., Ross, P. N., Eds.; Wiley-VCH: New York, 1998; Chapter 4. (4) Dube, P.; Brossard, L.; Menard, H. Can. J. Chem. 2002, 80, 345. (5) Ilikti, H.; Rekik, N.; Thomalla, M. J. Appl. Electrochem. 2002, 32, 603.

H3O+ + e- + M T MHads + H2O (Volmer reaction) (reaction 1) H3O+ + MHads + e- T M + H2 + H2O (Heyrovsky reaction) (reaction 2) 2MHads T 2M + H2

(Tafel reaction) (reaction 3)

Laplante et al.6 have extended this reaction scheme in order to take into account the effect of the adsorption sites on the inorganic catalyst support because adsorption phenomena at the electrolyte/substrate interface can play an important role during many heterogeneous catalytic processes:

YdZ + A T (YdZ)adsA

(reaction 4)

(YdZ)adsA + 2MHads T (YH-ZH)adsA (YH-ZH)adsA T YH-ZH + A

(reaction 5) (reaction 6)

In agreement with these reactions, two distinct sites of adsorption must now be considered: the original metallic sites (M) where atomic hydrogen is generated (reaction 1) and the adsorption sites (A) located on the catalyst support where the organic target molecule (YdZ) is adsorbed. Enhanced binding (6) Laplante, F.; Brossard, L.; Menard, H. Can. J. Chem. 2003, 81, 258.

10.1021/la0519002 CCC: $33.50 © 2006 American Chemical Society Published on Web 06/01/2006

Surface Adsorption Properties of Pd Catalysts

at A sites increases the density of the target species in the immediate vicinity of the adsorbed atomic hydrogen, thus facilitating the overall hydrogenation process. Lateral diffusion of the target molecule across the surface toward the metal sites may further enhance the ECH efficiency. The efficiency of the process is defined as the amount of transformed target species (i.e., the phenol in the present case) versus the cumulative charge passed. The Coulombic efficiency of electrocatalytic process is determined by the competition between molecular hydrogenation (reaction 5) and hydrogen evolution (reactions 2 and 3). The adsorption/desorption equilibrium of the organic target molecules and parameters such as the nature of electrode, the current density, the pH, the solvent composition, and the temperature are directly related to the overall yield of the ECH process.3,7-9 The development of selective and efficient catalysts is a key requirement for the implementation of ECH in the chemical and pharmaceutical industries. Selectivity can be controlled by the structural, chemical, electronic, compositional, kinetic, and energy properties of the composite electrode material. Previous studies6 have shown that the inorganic solid matrix that supported the metallic electrocatalysts can enhance the adsorption of the target species. This proved to be an effective approach, but twocomponent electrode architecture is difficult to optimize according to the properties of specific target species and the mechanical requirements of the metal support. In this work, we demonstrate a new conceptsin situ functionalized materials for electrochemical processessfor use in enhancing selective ECH processes. The design of this new catalyst is based on the in situ modification of the alumina surface by the adsorption of monocarboxylic aliphatic acids.10 By modifying the nature of these adsorbed chains, we modify the adsorption properties of the target molecule. Furthermore, we show that the enhanced efficiency of the overall ECH process is directly correlated to the improved adsorption properties of the target molecule on the functionalized inorganic support. In addition to the development of new solid support materials, the present work shows that it is possible to modify the surfaces of existing materials to provide stronger (or more selective) adsorption sites that can enhance selectivity and/or ECH yields. By using this flexible architecture with easily substituted chain systems, true molecular recognition could be envisioned as the controlling factor in ECH; this would be a highly advantageous property of such a modified support surface. In addition to the nature and mechanisms of the functionalization of the alumina support, this work also shows the role of solvent polarity in controlling the adsorption processes, which can therefore also influence the overall ECH efficiency. Experimental Section Unless otherwise noted, all materials were obtained from commercial suppliers and used without further purification. Reticulated vitreous carbon (RVC, 100 pores/in.) was purchased from Electrolytica Inc.; phenol (99.5%), from Research Chemical Ltd.; glacial acetic acid, propionic acid, and butyric acid, from Caledon; and NaOH (10 N), from VWR. 3-Methylcyclohexanol (99%), chloroform (spectrophotometry grade, Fisher), and palladium 10% w/w on alumina powder were provided by Aldrich. High-purity water (Milli-Q, R ) 18 MΩ cm) was used for the preparation of solutions. (7) Miller, L. C.; Christensen, L. K. Org. Chem. 1978, 42, 2059. (8) Casadei, M. A.; Pletcher, D. Electrochim. Acta 1988, 33, 117. (9) Robin, D.; Comtois, M.; Martel, A.; Lemieux, R.; Cheong, A. K.; Belot, G.; Lessard, J. Can. J. Chem. 1990, 68, 1218. (10) Snyder, L. R. J. Chromatog. 1966, 23, 388.

Langmuir, Vol. 22, No. 14, 2006 6415

Figure 1. Perspective view of the electrochemical dynamic cell used in this study. The electrochemical cell used in this study is shown in Figure 1 and has been described in detail elsewhere.11 The cell is equipped with a variable flow chemical pump (Fisher Scientific) connected to the cell by a PVC tube. The flow rate is maintained at 1072 mL/min, and the circulation of the solution is upward through the cathode. The cathode is a cylindrical piece of RVC (diameter ) 23 mm, width ) 13 mm, geometrical surface area ) 1770 mm2). To maximize the forced electrolyte flow through the RVC/catalyst system, the electrochemical cell is used as a punch to cut an electrode disk in the RVC, forcing a close press-fit tolerance. A carbon rod inserted into the RVC provides the mechanical support and electrical contact for the cathode; this support was coated with a thermoretractable polymer to prevent electrolytic processes on this conductor. The platinum mesh counter electrode (i.e., the anode) is introduced into a glass tube with an external diameter of 15 mm that was inserted into the cell from above. A Nafion 117 membrane is fixed at the lower end of the tube with a stopper thread to ensure the separation of the anodic compartment from the rest of the cell. Diffuse reflectance infrared Fourier transform (DRIFT) spectra of the electrocatalytic powders were recorded on a Nicolet Nexus 470 FTIR spectrometer equipped with a Harrick DRIFT accessory and an MCT-A detector. Data were collected with a resolution of 4 cm-1, a moving mirror speed of 1.98 cm/s, and Happ-Ganzel apodization; typical spectra required 256 mirror scans. The sample compartment and the optical bench were purged with Ar to exclude atmospheric interferences and to minimize contamination of the active surface. Thermal analysis-mass spectrometry (TA-MS) measurements were recorded with a home-built multicell reactor system with a temperature range of 25-1000 °C and a commercial system (Setaram SetSys 2400) operating from 25 to 600 °C. Functionalized catalyst powders were heated under an inert atmosphere (Ar), and the thermal decomposition of the organic adsorbate was detected downstream (principally as H2 and CO2) with a quadrupole mass spectrometer (Balzers, QMG-420, Pfeiffer GSD300T). TEM measurements were made on a Hitachi H-7500 instrument using an accelerating voltage of 80 kV. For the preparation of the samples, the Pd-alumina catalyst powders were embedded in an epoxy resin (Epon) and polymerized at 70 °C for 48 h. The samples were sectioned with a diamond knife to a thickness of approximately 70 nm. Adsorption isotherms were measured using an HPLC system (Agilent HPLC 1100 series) equipped with a diode array detector (11) Bannari, A.; Cirtiu, M. C.; Kerdouss, F.; Proulx, P.; Menard, H. Chem. Eng. Process 2006, 45, 471.

6416 Langmuir, Vol. 22, No. 14, 2006

Figure 2. TEM image of the as-received 10% Pd-alumina catalyst (TEM magnification ) 200 000×; HV) 80 kV). (Agilent DAD module 1100 series). The column was thermostated at 25 ( 0.1 °C (Agilent 1100 series thermostated column compartment). The solvent flow rate (buffer solution) was set to 1 mL min-1. The theoretical aspects of this method are described in detail elsewhere.12,13 This analytical technique allows us to measure the adsorption isotherms in the same media used in electrolysis experiments. All electrocatalytic experiments were carried out in the dynamic cell (Figure 1) under galvanostatic conditions (I ) 20 mA) using an Agilent galvanostat (model 6634B). The anodic compartment was filled with an acetate buffer solution (1 M). The cathodic compartment was filled with 29 mL of electrolyte solution (organic buffer solutions of the chosen aliphatic acid) adjusted to pH 5 with a NaOH solution. Prior to the ECH reaction, 200 mg of the 10% Pd-alumina catalyst powder was added to the catholyte solution, and 50 C was passed through the system to polarize the catalyst. This quantity of charge is necessary to condition the catalyst powder by the reduction of palladium oxides and to generate adsorbed hydrogen on the palladium aggregates. A volume of 1 mL of an aqueous phenol solution (25 mg/mL) was then added to the cathodic compartment, giving a total volume of 30 mL and a phenol concentration of 8.85 × 10-3 M. During the ECH process, 0.5 mL aliquots were withdrawn from the catholyte and saturated with NaCl, further extracted with 1 mL of chloroform, dried under anhydrous sodium sulfate, and then filtered. An external standard solution (50 µL of a 0.022 M 3-methylcyclohexanol solution) was added to 450 µL of the organic phase (chloroform), and the sample was analyzed by gas chromatography (GC). The GC analyses were carried out using an Agilent 6890 series chromatograph equipped with a flame ionization detector (FID) on a 30 m Agilent HP-5 column. The products were then identified by comparison with the retention time of prepared standards.

Results and Discussion It is well understood that the noble metals possess the ability to generate atomic hydrogen.3 Because the electrochemical process is a surface process, it is necessary to understand the state of metal dispersion on the support. The commercial 10% Pd-alumina catalyst used in the present study contains a large Pd loading. The TEM photographs of the catalyst particles (Figure 2) show palladium as small dark aggregates uniformly distributed on the alumina particles. We estimate that the average diameter of the metallic aggregates is ∼6 nm. (12) Chuduck, N. A.; Eltekov, Y. A.; Kiselev, A. V. J. Colloid Interface Sci. 1981, 84, 149. (13) Me´nard, H.; Noel, L.; Kimmerle F. M.; Lambert, M. Anal. Chem. 1984, 56, 1240.

Cirtiu et al.

Figure 3. DRIFT spectra of the Pd-alumina catalysts subjected to aqueous solutions of aliphatic acids: (a) acetic acid; (b) propionic acid; and (c) butyric acid. The reference data for each spectrum is the catalyst powder prior to exposure to the organic acid solution.

Alumina powder is recognized as an efficient adsorbent for most organic molecules, including simple carboxylic acids. The role of the adsorbed organic acid species in these studies depends critically upon its stability on the alumina support. We have used infrared spectroscopy to probe the catalyst surface after immersing it in the supporting electrolyte solution containing the organic acid of choice (i.e., acetic acid, propionic acid, or butyric acid), all at 0.5 M. Washing the catalyst powder with water and methanol removed the soluble compounds and salts, but as shown below, the acids remain firmly bound to the alumina substrate. The solution that contains the catalyst was filtered, and the powder was dried at 50 °C for 20 h and then analyzed by DRIFT spectroscopy. A catalyst powder that has not been exposed to the organic acids was used to obtain the reference channel for the infrared spectra. The infrared spectra of carboxylic acids are generally characterized by the strong absorptions at 1750-1650 and 13001200 cm-1 associated with the CdO and C-O bonds of the carboxyl group,14 whereas the carboxylate ion has strong antisymmetric (νas) and relatively strong symmetric (νs) COOstretch absorption at 1650-1510 and 1420-1280 cm-1, respectively.15 In our results (Figure 3), the loss of the acetic, propionic, and butyric acid bands (ν CdO) following adsorption onto the Al2O3 indicates that the surface species are acetate, propionate, and butyrate, respectively, and this is supported by the appearance of the antisymmetric stretching of COO- at 1570 cm-1 and symmetric stretching at 1420-1400 cm-1. The symmetric deformation of the CH3 group appears at 1340 cm-1 for acetic acid and at 1312 cm-1 for propionic and butyric acids, whereas the antisymmetric deformation of the CH3 group appears at 1440 cm-1. Although the intensities of the DRIFT results are only qualitative, the C-H stretching bands (∼2800-3000 cm-1) increase in proportion to the chain length of the organic acid, whereas the C-O bands are of approximately constant intensity for the three samples. This suggests that the coverage of the alumina supports by the organic acids is, to a first approximation, independent of the chain length. This was confirmed by thermal analysis, as discussed below. Many previous studies on the adsorption of carboxylic acid onto alumina confirm the ability of the carboxylic acids to adsorb strongly on aluminum oxide.16-23 Several mechanisms are (14) Mehrotra, R. C.; Bohra, R. In Metal Carboxylates, Academic Press: New York, 1983. (15) Nakamoto, K. In Infrared and Raman Spectra of Inorganic and Coordination Compounds, 5th ed.; Wiley: New York, 1997. (16) Chatterji, A. C.; Singhal, G. S. J. Sci. Ind. Res. 1960, 19B, 298.

Surface Adsorption Properties of Pd Catalysts

possible for the adsorption of carboxylic acid on alumina. Chatterji16 proposed a mechanism based on the hydrogen bridge theory in which a pair of electrons from an oxygen atom in the alumina structure is liable to be bound to the hydrogen of a carboxyl group. In turn, a pair of electrons from the oxygen of the carboxylic group is associated with the hydrogen in the OH group of the alumina substrate. Using inelastic electron tunneling spectroscopy (IETS), Evans and Weinberg17 concluded that acetic acid chemisorbs as a bidentate symmetrical bridging acetate, in agreement with the results obtained by Hall and Hansma.18 They concluded that formic acid and acetic acid adsorbed predominantly as bridging species on the alumina surface. The effect of the preadsorption of HCl on the subsequent adsorption on alumina of carboxylic acids has been examined by IETS.19 An untreated surface has Al+-O- sites and relatively basic OH groups that react readily with carboxylic acids. The exposure of the surface to HCl reduces the basicity of these OH groups, which do not react as easily with carboxylic acids. HCl also occupies the Al+O- sites and blocks the adsorption on the Al+ Lewis acid sites of carboxylate ions. The formation of the organic surface phase is characterized by complicated kinetics in which surface and/or monolayer defects, as well as impurities, play an important role.20,21 However, Dobson and McQuillan22 have used ATRFTIR to study the adsorption of aliphatic mono- and dicarboxylic acids to metal oxides (Al2O3, TiO2, ZrO2, Ta2O5) from aqueous solutions, but their results are inconsistent with those obtained by most other groups and with our own results. They concluded that acetic acid adsorbs onto ZrO2 and weakly onto Ta2O5, but no adsorption was detected on TiO2 or Al2O3. However, our DRIFT study shows that all monocarboxylic acids (acetic, propionic, and butyric, respectively) adsorb strongly onto alumina. To test the thermal stability of the organic acids adsorbed onto the catalyst surfaces and to confirm the stable adsorption of even the shortest-chain systems, we performed thermal-mass spectrometry analysis for each acid/catalyst system. The samples are introduced into the reactor system, and a temperature ramp from 25 to 700 °C at 10 °C/min is employed. The results show that the organic acids bound to the alumina surface are stable until 200 °C using the inert carrier gas. Above this temperature, the carboxylic acids begin to decompose; CO2 and H2 are identified as the principal decomposition products of the organic phase, as shown in Figure 4. The initial loss of CO2 at 200 °C corresponds to desorption of the acids from the catalyst surface, accompanied by molecular fragmentation of the acid termination from the aliphatic chain. Similar degradation profiles were observed for acetic acid and propionic acid (data not presented here). Therefore, we can conclude that the aliphatic acids grafted onto the Pd/ alumina catalysts powders are stable for all temperatures below 200 °C. The quantity of CO2 released from the samples over the 50-600 °C range (using inert carrier gases and normalized according to the sample mass) varied by less than 6% among the samples with the three organic acids, with no trends observed as a function of the chain length; Pd/alumina samples that had not been exposed to organic acids exhibited negligible CO2 desorption. These behaviors allow us to affirm that the surface of alumina is modified with organic acids prior to initiating the ECH process and presents a stable organic surface at ambient (17) Evans, H. E.; Weinberg, W. H. J. Chem. Phys. 1979, 71, 4789. (18) Hall, J. T.; Hansma, K. P. Surf. Sci. 1978, 77, 61. (19) Cheveigne, S.; Gauthier, S.; Guinet, C.; Lebrun, M. M.; Klein, J.; Belin, M. J. Chem. Soc, Faraday Trans. 2 1985, 1375. (20) Allara, D. L.; Nuzzo, R. G. Langmuir 1985, 1, 45. (21) Allara, D. L.; Nuzzo, R. G. Langmuir 1985, 1, 52. (22) Dobson, K. D.; McQuillan, A. J. Spectrochim. Acta 1999, 55A, 1395. Ferri, D.; Burgi, T.; Baiker, A. HelV. Chim. Acta 2002, 85, 3639. (23) Maxted, E. B.; Ali, S. J. Chem. Soc. 1961, 83, 4137.

Langmuir, Vol. 22, No. 14, 2006 6417

Figure 4. Thermal analysis-mass spectrometry data for butyric acid-modified Pd-alumina catalysts. The temperature ramp rate was 10 °C/min under flowing Ar gas.

Figure 5. Adsorption isotherms of (2) phenol and (9) acetic acid (pH 5) in water using 10% Pd-alumina supports.

temperatures. Equally important, the carboxylic acid surface coverage of the Pd/alumina supports is independent of the chain length, as is expected for such a strong adsorbate-surface interaction. We note, however, that the exchange of the carboxylic acids can occur in aqueous solutions, showing that the adsorption is reversible and apparently quantitative. Figure 5 compares the adsorption isotherms for acetic acid and phenol, measured separately on the 10% Pd-alumina powders used in most of this work. The quantity of adsorbed acid reaches saturation levels for solution concentrations of ∼10-3 M. Figure 5 shows that the quantity of adsorbed acid is approximately 10 times greater than the quantity of adsorbed phenol for identical solution-phase concentrations. This is consistent with the group adsorption energies tabulated by Snyder10 for adsorption on alumina, where the adsorption energies of the carboxylate group associated with an alkyl chain are ∼3 times greater than those of the phenolic -OH adsorption. In anticipation of the following discussion, we note that in all of the experiments reported herein where both phenol and the organic acid are present in solution (i) the organic acid supporting electrolytes are present at concentrations that are at least 50 times greater than the phenol concentration and (ii) the quantity of acid is several orders of magnitude greater than that required to reach saturation coverage on the Pd-alumina surface. As such, it is certain that each of the alumina surfaces employed in this study is saturated with the

6418 Langmuir, Vol. 22, No. 14, 2006

Figure 6. Remaining fraction of phenol in an aqueous medium as a function of the total charge. Catalysts: (9) submicrometric Pd powder; (b) 10% Pd-alumina; (0.5 M acetic buffer solution, pH 5; I ) 20 mA).

Figure 7. Remaining fraction of phenol in an aqueous medium as a function of the total charge for different concentrations of acetic acid: (9) 1 M; (b) 0.5 M; (2) 0.1 M; and (1) 0.05 M (I ) 20 mA).

organic acids and that the direct interaction between the phenol and the alumina support is negligible. The importance of the functionalized alumina support was made clear by a comparative study of phenol ECH using a pure Pd submicrometer powder and a 10% Pd-alumina-supported catalyst. Figure 6 presents the remaining phenol in the ECH process as a function of the total charge, with a pure Pd submicrometer powder and a 10% Pd-alumina-supported catalyst, under identical experimental conditions. The pure Pd powder provided only 6% of the phenol depletion after 6F/mol of charge passed as compared to the supported catalyst, where the phenol depletion is 100%. It is clear that the unsupported pure Pd powder is far less effective in the ECH of phenol than the supported Pd-alumina catalyst. The same low efficiency was observed for the ECH of phenol using a Pd foil.6 However, the ECH process is highly effective on Pd-alumina when the alumina is covered with the organic acid, suggesting that the interaction of the target phenol molecule with the organic surface is contributing to the greater efficiency. This sensitivity to the presence of the functionalized support makes clear the role of the adlineation point, as originally suggested by Maxted and Ali23 to distinguish between the metal sites and the adsorbent sites. Tests conducted across a wide range of supporting organic acid electrolyte concentrations, as shown in Figure 7, present only small differences in the ECH efficiency in the 50-150 C regime. As will be shown below, these slight differences are most likely due to changes in the solvent properties rather than changes in the surface characteristics. The insensitivity to the

Cirtiu et al.

Figure 8. Adsorption isotherms of phenol in 0.5 M acetic acid buffer solution. Supports: (9) 63 µm Pd powder and (b) 10% Pdalumina (pH 5, T ) 298 K).

acid concentration is consistent with the strong adsorption of the acid on alumina supports and the large excess of acid relative to saturation coverage on the alumina. The key physical difference between these samples is the lack of strong adsorption sites for the phenol on the pure Pd surface. In the absence of a high adsorbed-phenol density in the vicinity of the hydrogen-bearing Pd surface, the ECH process is kinetically limited. To test this hypothesis, the adsorption isotherms of phenol were determined for the pure Pd catalyst (63 µm average diameter particles) and the similarly sized alumina-supported Pd catalyst in the same media as the electrolysis media employed in the previous experiment. Figure 8 shows that the functionalized aluminasupported Pd catalyst adsorbs significantly more phenol than a Pd unsupported catalyst, at all concentrations studied. Thus, we conclude that coupling the adsorption properties of the acidcovered alumina support with the hydrogen-generation capabilities of Pd creates a promising composite electrode for the ECH of phenol. These results demonstrate that the adsorption of phenol (i.e., reaction 4 identified above) is a crucial step in the ECH process, yet clearly the adsorption on pure Pd is less effective than on the organically treated alumina. It is reasonable to expect that this step could be strongly influenced by the polarity of the solvent because this affects the solvation/adsorption equilibrium in the electrolytic cell. To establish the role of the nature and concentration of the cosolvent on the ECH efficiency (and hence to optimize the ECH process), the yields for the ECH of phenol were determined for four different methanol-water mixtures. Two measures of the solvent polarity can be used to quantify the evolution of the solvent’s properties as a function of its composition. The polarity index (P′) of a binary mixture, as defined in the field of chromatography, varies linearly between the polarities of the pure solvent components according to the v/v ratio. An alternative Lewis acidity-based definition, ENT, is calculated from the absorption band positions of standard chromophores in various solvents24,25 (normalized to the band position in pure water). Both criteria show that the polarity of binary mixtures evolves smoothly and monotonically as a function of the composition, with no local minima or anomalous features. In the present context, the polarity of the solvent has been modified from a strongly polar (100% H2O, P′ ) 10.2; ENT ) 1) to a less strongly polar solvent (100% methanol, P′ ) 5.1; ENT ) 0.77). (24) Kellner, R.; Mermet, J. M.; Otto, M.; Valcarcel, M.; Widmer, H. M. Analytical Chemistry, 2nd ed., Wiley-VCH Press: Great Britain, 2004; p 570. (25) Krygowski, T. M.; Wrona, P. K.; Zielkowska, U.; Reichardt, C. Tetrahedron 1985, 41, 4519.

Surface Adsorption Properties of Pd Catalysts

Figure 9. Remaining fraction of phenol in an aqueous medium as a function of the total charge for different quantities (v/v) of methanol as cosolvent: ([) 0%; (b) 20%; (9) 50%; and (2) 60%. (0.5 M acetic buffer solution, pH 5; I ) 20 mA).

Figure 10. Adsorption isotherms of phenol on 10% Pd-alumina for different quantities of methanol (v/v) as cosolvent: ([) 0%; (b) 5%; (9) 20%; and (2) 50%; (0.5 M acetic buffer solution; pH 5; T ) 323 K).

The polarity of a solvent mixture24,25 can be calculated from the values for the pure solvents. The lower-polarity (methanol-rich) solutions reduce the adsorption of the organic molecule onto the functionalized alumina, which should strongly affect the ECH efficiency. Figure 9 shows the phenol consumption trends for four concentrations of the methanol cosolvent. It is clear that as the methanol concentration increases, the hydrogenation of phenol becomes less efficient. Again we suggest that this is principally due to the differing adsorption properties of the functionalized support for the phenol target molecule, which modulates the ECH process via reaction 4. To test this hypothesis, we have measured the adsorption isotherms of phenol on the functionalized Pd-alumina catalyst powder in the four water-methanol mixtures. The phenol adsorption is clearly reduced for methanolrich solutions (Figure 10). The increasing sequence of solvent polarity that directly correlates to the increasing ECH activity (Figure 9) is also directly correlated to the increasing adsorption onto the functionalized Pd-alumina catalyst powders. The combination of the ECH yield measurements and the direct adsorption measurements confirms that the solvent composition is controlling the ECH efficiency via the solvation versus adsorption properties of the catalyst because of the importance of reaction 4 in the overall ECH process. A similar, but less pronounced, effect was observed by St. Pierre et al. in their study of the ECH of cyclohexanone with a nickel supported on functionalized-silica catalyst.26 In our study, ethanol and propanol

Langmuir, Vol. 22, No. 14, 2006 6419

Figure 11. Remaining fraction of phenol in an aqueous medium as a function of the total charge using water-methanol solution (80:20 v/v) in the presence of different electrolytes: (9) acetic acid; (b) propionic acid; and (2) butyric acid. (0.5 M; I ) 20 mA).

induced changes to the ECH efficiency and adsorption properties that are similar to those induced by methanol, as described above. However, for 1,2-propandiol as the cosolvent, only a slight decrease in the ECH process efficiency was observed (data not presented here). The demonstration that adsorption onto the catalyst support can strongly modulate the overall ECH efficiency suggests other mechanisms by which adsorption can be affected. To view how the electrolyte composition can influence the ECH of phenol, a comparative study of several organic acids was performed using a 20% methanol solution. This methanol-water mixture was used to ensure the solubility of all of the organic acids under consideration. It is well known that organic buffers have a detrimental effect on the ECH efficiency of fused polycyclic aromatic compounds at Raney Ni electrodes.27 The authors assumed that the detrimental effect is caused by the adsorption of organic acids onto Raney Ni electrodes that block important adsorption sites for the target molecule. Of course, there are significant differences between the morphology, electronic structure, and ECH properties between the Raney Ni and PdAl2O3 catalysts, and the role of the organic acids need not be comparable in these two systems. Figure 11 shows the residual phenol versus cumulative charge curves for acetic acid-, propionic acid-, and butyric acid-based electrolytes, all of which were present at 0.5 M. These results show that the longer-chain acids have an increased propensity to promote the ECH reaction; the remaining phenol at 150 C of cumulative charge is approximately 4-fold lower for the butyric acid-containing solutions. As has been demonstrated above for the influence of the methanolwater ratio and the presence of the alumina support, these variations in ECH efficiency are also due to the changes in the phenol adsorption properties onto the Pd-alumina catalyst in the presence of these acids, as characterized by the adsorption isotherms of phenol onto the functionalized Pd-alumina catalysts powders. Figure 12 shows that electrolytes containing the longerchain acids show enhanced adsorption properties for the phenol target molecule relative to the shorter-chain analogues. In fact, electrolytes containing butyric acid show a 2.5-fold increased phenol adsorption relative to the acetic acid-based solutions. Again, the measured ECH efficiencies correlate well with the measured adsorption properties of the various acid-based solutions. (26) St-Pierre, G.; Chagnes, A.; Bouchard, N. A.; Harvey, P. D.; Brossard, L.; Menard, H. Langmuir 2004, 20, 6365. (27) Robin, D.; Martel, A.; Lemieux, R.; Cheong, A. K.; Belot, G.; Lessard, J. Can. J. Chem. 1990, 68, 1218.

6420 Langmuir, Vol. 22, No. 14, 2006

Figure 12. Adsorption isotherms of phenol on 10% Pd-alumina catalysts in water-methanol solutions (80:20 v/v) in the presence of different electrolytes (0.5 M): (b) acetic acid; (9) propionic acid; and (2) butyric acid.

From the above discussion, it is clear that the supporting electrolyte must intervene in the ECH process. As shown in Figures 3 and 4, simple carboxylic acid molecules are strongly adsorbed onto the alumina surface and functionalize the catalyst surface, thus changing the adsorption properties of the phenol target molecule toward the composite catalyst powder. Carboxylic acid adsorption can arise at metal oxide surfaces from the ability of the carboxyl group to act as a ligand for vacant coordination sites of surface metal ions. This in situ molecularly functionalized surface becomes only weakly polar in the presence of the bounded acids, which influences the adsorption phenomena. Although the influence of the organic acid chain length is less pronounced than that of the solvent composition (as shown in Figure 9) the use of such a simple and easily tailored surface property can be exploited to enhance the selectivity of the adsorption process for various target molecules. The adsorption aspects of these systems resemble chromatographic systems in reversed mode (partition chromatography) where the stationary organic phase is nonpolar (or less polar) and the mobile phase is a relatively polar solvent. In reverse-mode chromatography, the most polar components are eluted first because of their weak affinity for the support, and an increase in the polarity of the mobile phase increases its retention time;28 these retention times are only indirectly affected by the relative diffusion coefficients on the surface because the net mass flow from the injector to the detector in a chromatographic application is principally due to the forced flow of the sample (10-30 cm/s), which greatly exceeds the diffusional time scales for macroscopic movement on the surface (10-4 to 10-6 cm/s). In the context of the present study, more polar solvents (and longer carboxylic acid supports) would be expected to increase the adsorption on the acid-functionalized alumina, as observed using the various water-methanol solvents and carboxylic acids studied herein. The adsorption of the phenol (as well as the intermediate cyclohexanone and product cyclohexanol for the ECH process) is on the new organic phase that covers the alumina surface, and the strength of this adsorption is clearly greater than for phenol on the bare Pd-alumina, as seen in the comparison of the results shown in Figures 5 and 12. Evidently, diffusion in and on the 2D organic phases is more effective in transporting the phenol to the Pd catalyst (the site where hydrogenation occurs) than diffusion across the pure alumina support or diffusion in the 3D (28) Skoog, D. A.; West, D. M.; Holler, F. J.; Crouch, S. R. Analytical Chemistry: An Introduction, 7th ed.; Sauders College Publishing: Fort Worth, TX, 1999; p 690.

Cirtiu et al.

Figure 13. Correlation of the ECH efficiency of phenol on Qads, measured for 10% Pd-alumina catalysts in buffered acetic acid solutions using methanol-water solvents as discussed in the text. The data are extracted for conditions of total charge ) 100 C at an equilibrium concentration of 0.1 µmol/mL.

bulk solution, as witnessed in the trends of the ECH efficiency. It is well known that 2D diffusion is much more effective that its 3D analogue in delivering a target molecule to a finite-sized object if the diffusion coefficients are comparable.29 In the present context, the phenol is adsorbed according to its van der Waals interactions with the adsorbed phase, but its transport across this surface is relatively unimpeded by the lack of specific binding of the phenol to the alkyl chain terminations. As a result, in the absence of strong localized interactions that would trap the phenol at specific sites, the key parameter that increases the ECH efficiency is the increased surface concentration of the target phenol molecules, which is controlled by the nature of the bound organic phase. As has been shown repeatedly in this work, adsorption strength correlates extremely well with the observed ECH yields and efficiency. In fact, a direct dependence can be established between the overall current efficiency and the adsorption characteristics of the catalyst using the data of Figures 9 and 10. For a quantity of charge of 100 C and an equilibrium concentration of 0.1 µmol/mL, the efficiency of the ECH process is linearly related to the adsorption of phenol onto the functionalized catalyst, as shown in Figure 13; it should be noted that the linear trends will not, in general, be observed for all total delivered charges, but the overall correlation of stronger adsorption with greater ECH efficiency is always found for these systems. Although the ECH efficiency for any given molecule will certainly depend on the nature of the adsorbed phase in a rather complicated manner, the results presented herein demonstrate that general purpose commercial catalysts such as Pdalumina can be further optimized for specific ECH processes by a simple in situ protocol. This new concept (i.e., an in situ functionalized catalyst) can explain the results obtained in the study of phenol ECH as a function of the composition of the supporting electrolyte. The electrocatalytic hydrogenation process with the in situ functionalized catalyst is shown in schematic form in Scheme 1. As the chain length of the aliphatic acid increases, the surface becomes less polar (acetic acid < propionic acid < butyric acid), and the van der Waals interactions with the target molecules increase; this leads to increased adsorption of the phenol in and on the organic surface. As in reverse-phase chromatography, the retention may be influenced by the solvent composition (mixture of aqueous buffers with methanol) or by the nature of the stationary (29) Berg, O. G.; von Hippel, P. H. Ann. ReV. Biophys. Biophys. Chem. 1985, 14, 131.

Surface Adsorption Properties of Pd Catalysts Scheme 1. Schematic Representation of the Electrocatalytic Hydrogenation of an Unsaturated Molecule on the in Situ Functionalized Catalyst

(bound) phase. The retention of the target molecule is roughly dependent on the carbon load of the stationary phase (i.e., the amount and type of bonded phase on the support material). Sample retention normally increases for bonded phases of greater length (C18 > C8 > C3 > C1).24,28 In the present study, strong adsorption of phenol in the order of the increasing chain length of the adsorbed organic layer improves the ECH efficiency, as demonstrated experimentally in the results of Figure 7.

Conclusions The study has focused principally on the effect of solvent composition and the nature of the supporting electrolyte on the ECH process of phenol. We have presented and developed herein a new conceptsthat of in situ functionalized materials for electrocatalytic hydrogenation processes. This concept is supported by the results obtained for the phenol ECH with different

Langmuir, Vol. 22, No. 14, 2006 6421

supporting electrolytes. These new materials are based on the strong adsorption of aliphatic carboxylic acids (acetic acid, propionic acid, and butyric acid) onto the alumina surface. These surface modifications play a key role in the adsorption/desorption phenomena of the organic molecules onto the composite electrode surface. The adsorption isotherms confirm the role of the carboxylic acids, whereas the DRIFT spectra and TA-MS analyses confirm that the carboxylic acids are strongly adsorbed onto the catalyst surface. A direct correlation has been established between current efficiency and adsorption phenomena for the ECH of phenol; the applicability of this protocol to enhance the selectivity of the ECH process for other target species is now under investigation. This new approach opens the door for new challenges in electrocatalytic hydrogenation, such as the electrocatalytic hydrogenation of asymmetric organic molecules at specific chemically similar functional sites within a larger target molecule. It is probable that by coupling the selectivity of the adsorption process with the electrochemical hydrogenation such site-specific modifications could be achieved. Acknowledgment. This work was financially supported by the Natural Sciences and Engineering Research Council of Canada (NSERC) and the Fond Que´be´cois de la Recherche de la Nature et les Technologies (FQRNT). We thank Louis Brossard for useful discussions and Charles Bertrand (CHUS) for TEM analyses. C.M.C. thanks the Universite´ de Sherbrooke (Sherbrooke, Quebec, Canada) for the institutional fellowship. LA0519002