Langmuir 1993,9,1893-1897
1893
Hydrophobic Effects in Surface-Confined Decaalkylferrocene Redox Polymers David Albagli and Mark S. Wrighton' Department of Chemistry, Massachusetts Imtitute of Technology, Cambridge, Massachusetts 02139 Received November 30,1992. In Finial Form: May 3,1993 Surface-confined redox-active polymers formed from a decaalkylferrocene derivative containing two triakoxysilylbutyl groups, 1, are durable, pinhole-free electroactive films in nonaqueous solvents (Eo' = -0.09 V v8 SCE). n-Si electrodes can be derivatized with 1 to give photoelectrodes that, when illuminated with light of energy greater than the band gap, cause the photooxidation of 1 with a photovoltage of 0.45 V. Electrodes derivatized with 1 show pronounced solvent effects upon switching from nonaqueous to aqueous electrolyte; in aqueous media current associated with oxidation of the polymer decreases significantly. Contact angle measurements of HzO on films of 1 approach 90°, showing that decaallrylferrocene polymer/solutioninterfaces are hydrophobic. In contrast, by covalently binding a cyanopropyl group into the decaalkylferrocene f i using (3-~yanopropyl)triethoxyeilane, the electrode response in aqueous media was simiiar to that in nonaqueous media. However, the contact angle of HzO is not significantly lower than for films of 1, suggesting the nitrile groups act to support ion transport within the film. We wish to report that electrode-surface-confined decaalkylferrocene redox polymers from 1 exhibit a
1
typical' electrochemical response in organic solvent (CH3CN, CH2Cl2, EtOH) electrolyte media but pass little or no faradaic current when examined in aqueous electrolyte. This effect is reversible. A copolymer film of 1 and the nitrile 2 yields a redox-active decaalkylferrocene system which shows nearly ideal response in either non-aqueous or aqueous electrolyte. (Et0)sSi-
C=N
2
Ferrocene derivatives represent an oftan-studied class of redox molecules and redox polymers. Ferrocene-based reagents are typically used as a prototypic redox center in work demonstrating new techniques or concepts or in studies designed to test theory.2 However, peralkylated derivatives of ferrocene have received little attention? and the properties of such a redox polymer have yet to be reported. One reason for increasing the number of alkyl (donor) substituents on the cyclopentadienyl rings is to tune the redox potential to more negative values. Decamethylferrocene has a redox potential of -0.1 V vs SCE, -0.5 V negative of f e r r ~ c e n e .Accordingly, ~ alkylferrocene derivatives may be useful components in setting up redox potential gradients for effecting charge separat i ~ n .Peralkylation, ~ however, increases the hydrophobicity of the ferrocene center. We show that this detrimentally affecta charge transport in the redox polymer f i as the water content of the electrolyte increases. However, this adverse effect can be overcome by adjusting the molecular composition of the surfaceconfined film. 0743-7463/93/2409-1893$04.00/0
Experimental Section General Procedures. All chemicalsused were reagentgrade. Tetrahydrofuran was distilled from LiAlK under N2. Glass distilled deionized HzO, absoluteEtOH, and glass-distilledCHsCN (EM Science),paseed through a column of activatedalumina, collected under N2, and stored over 4-A molecular sieves, were used in the electrochemical experiments. Anhydrous Eta0 and the solvents used in chromatography were wed as received. Alumina (neutral, Activity I) waa used in chromatography. [n-BuSJlClOcelectrolytewas recrystallizedfrom EtOH and dried in vacuo at 80 O C for 48 h. lH NMR spectra were recorded on a Bruker 250 MHz or a Varian XL-300 MHz FT spectrometer. lSCNMR spectra were recorded at 75.4 MHz. Electronic absorption spectra were obtained on a Hewlett-Packard8541A diode array spectrometer or a Cary 17 spectrometer using 1.00 cm path length quartz cuvettes. Maas spectrometry was done on a Finnigan MAT System 8200 with a double focusing magnetic sector by electron impact (70 eV). Melting points were obtained with a Thomas capillarymelting point apparatusand are uncorrected. Elemental analyses were done by Schwarzkopf Microanalytical Laboratory, Woodside, NY. Methyl 4-Pentenoate. 4-Pentenoic acid was esterified according to the literature procedure6using CHsOH, 2,2-dimethoxypropane, and a catalytic amount of p-TsOH.H20. The product waa purified by distillation and obtained in 80% yield as a clear, colorlees oil,bp 127-129 O C (lit.' bp 128O C ) . lH NMR (CDCh) 6 2.32 (m, 4 H), 3.60 (s,3 H), 4.94 (m, 2 H), 5.74 (m, 1 HI. 1 - ( 3 - B u t e n y l ) - 2 ~ , 4 , & t e t ~ e t h y l c y c l o ~ n ~Meth~ene. yl 4-pentenoate was reacted with 2-lithio-2-buteneaccording to
* Author to whom correspondence should be addressed.
(1) Bard, A. J.; Faulkner, L. R. ElectrochemicalMethode;Wiley: New York, 1980, pp 521-624. (2) For example: (a) Chidmy, C. E. D. Science 1991, 261, 919. (b) Pinkerton, M.J.; LeMest, Y.; Zhang, H.; Watanabe, M.;Murray,R. W. J. Am. Chem. SOC.1990,112,3730. (c) Wipf, D. 0.; Kriatensen, M. D.; Wightman, R. M.Anal. Chem. 1988,60,306. (d) Aoki, K.; Morita, M.; Niwa, 0.; Tabei, H. J. Electroanal. Chem. 1988,266, 269. (3) For an example of a pentamethylferrocene derivative see: Chao, S.; Robbine, J. L.; Wrighton, M.S. J . Am. Chem. SOC.1983,106, 181. J.Am.Chem. (4) Robbins,J.L.;Edelstein,N.;Spencer,B.;Smart,J.C.
SOC.1982,104,1882. (6) (a) Wrighton, M.S. Comments Znorg. Chem. 1981, 4 , 269. (b) Kaneko, M.;Wchrle, D. Adu. Polym. Sci. 1988,84,141. (c) Hopfield, J. J.; Onuchic, J.; Beraten, D. N. J . Phys. Chem. 1989,93,6350. (6) Lorette, N. B.;Brown, J. H., Jr. J. Org. Chem. 1919,24, 261. (7) McCreer, D. E.; Chiv, N. W. K.; Can. J. Chem. 1968,46, 2217.
0 1993 American Chemical Society
1894 Langmuir, Vol. 9, No. 7, 1993
Albagli and Wrighton
Contact Angle Measurements. Measurements were made the literature procedures to make the title compound. The on a Ram&-HartModel 100 contact angle goniometer equipped product was purified by column chromatography wing hexane with an environmental chamber using a 1 p L drop of doubly as the eluant. TLC Rf 0.95 (hexane); 'H NMR (CDCb) major distilled water. The chamber atmosphere was maintained at isomer: b 1.0(d, 3 H), 1.55 (m, 2 H), 1.78 (m, 9 H), 2.11 (m, 2 H), 100% humidity by f i i n g the wells in the chamber with distilled 2.27 (q, 1 H), 4.94 (m, 2 H), 5.82 (m, 1 H). water. The contact angle was determined by measuring the l,l'-Bis(3-butenyl)octamethylferrocene. The peralkylcytangent normal to the sessile drop at the liquid-eolid interface. clopentadiene was used to prepare the symmetric ferrocene The substrates used were Pt flag electrodes derivatized with accordingto the literature procedure? The product was purified triethoxysilyl reagents as described above. The electrodes were by sublimation at 95"C (0.07" H g ) : mp 67"C;'H NMR ( c a s ) f i i t examined electrochemicallyto mess the coverage and were 6 1.64 (e, 6 H), 1.67 (8, 6 H), 2.08 (m, 2 H), 2.30 (m, 2 H), 4.97 (m,1H)5.06(m,1H),5.84(m,1H);1sCNMR(CsD6)d9.73,9.84,removed from the cell in the reduced state and rinsed thoroughly with CHsCN and EtOH. After the contact angle was determined 25.75,35.63,78.21,79.03,82.42,114.53,139.05;MSm/e406 (M+), on the reduced samples, the f i i s were oxidized by soaking in 365 (M- CH&HICH2+), 350 (M - CH&H&H band gap = 1.1 eV) cyclic voltalayer on the surface by holding the potential at +1.30 V until the mmograms for the surface-bound redox couple are obcurrent decayed to the background level, and then removing it served. As shown in Figure 2, for an illuminated (15mW/ from potential control. Derivatization was accomplished by cm2 at 632.8 nm) n-Si/l electrode in EtOH/O.l M soaking pretreated electrodes in a 10 mM isooctane solution [n-Bu4NlC104, the anodic peak for 1 is at -0.490 V vs of the triethoxysilyl reagent(s) for times ranging from 30 s to 48 h. The electrodes were rinsed thoroughly with isooctane, CHsSCE, representing a negative shift of -460 mV compared CN, and EtOH prior to use. to 1on Pt. The 450-mV shift represents the photovoltage, n-Si electrodes were fabricated from single-crystal, P-doped or the extent to which light is used to drive the oxidation (100orientation), n-Si wafers with resistivities between 0.5 and of 1 in a thermodynamically uphill sense. If the light is 1.5 Q/cm that were obtained from Monmuhl' n-Sielectrodes turned off at the poaitive scan limit following photooxiwere pretreated for derivatization with triethoxysilyl reagents dation of 1,the reduction of 1in the dark is observed, but by soaking in concentrated HF for 308, rinsing with H10, soaking thereafter no faradaic current passes. The behavior of 1 in 10M NaOH for 608,and rinsing with H a , CHaCN, and hexane on n-Si is that expected for a molecule whose redox and were then immediatelyplaced into the derivatizationsolution. potential is between the conductionand valenceband edges Derivatized electrodes were rinsed thoroughly with hexane and of an n-type semiconductor electrode in contact with the EtOH before being used in electrochemical experiments. Photoelectrochemical experiments with n-Si were carried out by electrolyte solution." irradiating the electrodes with a 5-mW He-Ne laser emitting at At high surface coverages of 1 (>5 X 10-9 mol cm-2) the 632.8-nmwith the beam expanded to cover the electrode area polymer film resists interdiffusion of small-size neutral (0.25cm2). (8)Manriquez, J. M.; Fagan, P. J.; Schertz, L. D.;Marks, T. J. Inorg. (12)(a) Deschler, U.;Kleinschmit, P.; Danster, P. Angew. Chem., Znt. Ed. Engl. 1986,25,236.(b) Moses, P. R;Weir,L.; Murray, R. W. A m l . Synth. 1982,21,181.
-
~~
Chem. 1975,47,1882. (13)Electrodes derivetizedwith 1 can be left exposed to air for 2week or continuonsly cycled through ita redox wave for 18 h (-3600 cycles) with