J. Chem. SOC., Faraday Trans. I, 1989, 85(4), 813-827 Photoelectrochemical Investigations of Phenosafranine Dye bound to some Macromolecules R. Ramaraj and P. Natarajan" Department of Inorganic Chemistry, University of Madras, Guindy Campus, Madras 600 025, Tamil Nadu, India Macromolecular- bound phenosafranine dyes have been synthesized and characterized. The absorption spectra of the polymer-bound dyes show a bathochromic shift while the emission spectra shift slightly to higher frequency. Cyclic voltammetry of polymeric phenosafranines adsorbed at the carbon electrode show the subsequent two-electron redox reaction of the dye. Flash photolysis of the monomeric and polymeric dyes in the presence of EDTA produce the one-electron reduced dye which disproportionates ; the disproportionation rate constant was measured to be lo9 and lo8 dm3 mol-' s-l, respectively.Photoelectrochemical studies of the macromolecular phenosafranine indicate different behaviour depending upon the macro- molecule. In the case of an electrode coated with a film of poly(acry1- amidomethylphenosafranine-co-methylolacrylamide) cathodic behaviour was observed with reference to an inert electrode, while an electrode coated with a film of poly(acrylamidomethylphenosafranine-co-methylolacryl- amide-co-vinylp yridine) exhibited anodic polarity . A wa ter-spli tting regenerative cell was shown to operate using the polymeric phenosafranine- coated electrodes. The photoredox systems are receiving attention to utilize the solar energy as electricity and fuels by chemical methods.' The photochemistry in photosynthesis is known to involve one-electron transfer reactions from the excited state of chlorophyll species to an acceptor. In recent times many workers" are developing dye-sensitized systems as models for the primary processes of photosynthesis and as devices to convert and store solar energy.Among the systems studied so far, the photopotential values reported are 250 mV for the iron-thionine ~ystem,~ 476 mV for the proflavin-EDTA system,' 844 mV for the tolusafranine-EDTA system6 and 870 mV for the phenosafranine-EDTA ~ystem.~ Among these redox systems, the iron-thionine system is reversible and in the other systems the oxidation of EDTA is irreversible. The abovementioned photogalvanic systems are reported in homogeneous conditions, and in these systems the energy- wasting back reactions are found to reduce the efficiency of the photoenergy conversion.On the other hand a microheterogeneous reaction environment such as in micelles,' monolayer as~emblies,~ polymerslO and chemically modified electrode systems," facilitates unidirectional electron flow, and light absorbers show interesting charac- teristics, different from the corresponding monomeric light absorbers. '' In this paper we report the preparation and photoelectrochemistry of the different macromolecular- bound phenosafranines. We have constructed a photoelectrochemical cell by coating the polymer-dyes onto electrodes which show regenerative pho toelectrochemical behaviour. 813 28-2814 Macromolecular- bound Phenosafranine Dyes / H NH NH I CH20H Q / H Fig.1. Structure of polymer-bound phenosafranine dyes. Experiment a1 Materials Phenosafranine dye (3,7-diamino-5-phenylphenazinium chloride) was recrystallized from methanol. Acrylamide (B.D.H.), ethylenediaminetetra-acetic acid disodium salt, EDTA (B.D.H.), potassium chloride (B.D.H.) and sodium perchlorate (Merck) were analytical-grade chemicals and were used as such. Water was distilled twice adding alkaline KMnO,. Preparation of Polymer-dye Samples Poly(N-methylolacrylamide), P(MAAM), and poly(N-methylolacrylamide-co-4-vinyl- pyridine), P(MAAM-co-VP), were prepared by the following procedure as described earlier. l3 N-Methylolacrylamide was prepared by the reaction of paraformaldehyde with acrylamide. N-Methylolacrylamide was polymerized in aqueous solution using potassium peroxodisulphate as the initiator to obtain P(MAAM). The copolymer P(MAAM-co-VP) was prepared by copolymerizing N-methylolacrylamide with 4- vinylpyridine.The homopolymer P(MAAM) and poly(4-vinylpyridine) formed in the preparation of the copolymer P(MAAM-co-VP) were removed as described earlier. l3 Phenosafranine was condensed with P(MAAM) to give poly(acrylamidomethy1- phenosafranine-co-methylolacrylamide). Purified phenosafranine was added to an aqueous solution of the polymer P(MAAM) in the desired molar ratio and the mixture was kept at 90-95 "C in a water bath for 6 h. Hydroquinone was added to this mixture to prevent crosslinking of the polymer. The uncondensed phenosafranine was removed by dialysing the solution for 15-20 days in water.Dialysis was continued until the solution outside the sack did not show any absorption of phenosafranine (Amax = 522 nm). Poly(acrylamidomethylphenosafranine-co-methylolacrylamide-co-vinylpyridine) wasR. Rarnaraj and P. Natarajan 815 prepared by following the same procedure as in the case of polymer dye P(AMPS+-co- MAAM). The structures of the macromolecular dyes used are shown in fig. 1. The phenosafranine dye condensed to the macromolecular chain was stable even under rigorous conditions. The polymer-dye solutions, on standing for months and subsequent dialysis against water, did not show any trace of dyes passing through the membrane, indicating the stability of the polymer-bound phenosafranine dye. Estimation of Phenosafranine A titrimetric method was used to estimate the phenosafranine dye bound to the macr0mo1ec~le.'~ A solution containing a known amount of titanous chloride in 4 mol dm-3 hydrochloric acid was titrated against the polymeric phenosafranine under a nitrogen atmosphere at 60 "C.The colour of the phenosafranine solution disappears owing to the reduction of phenosafranine by titanous ions in the presence of hydrochloric acid to leucophenosafranine. The end point is the appearance of a light brown colour. From the titration data the amount of phenosafranine present in the solution taken was calculated. A known volume of the original polymer-dye solution was evaporated and the weight of the residue was taken as the amount of polymer-dye complex present in the original solution.The number of phenosafranine units bound to the polymer chain consisting of a given number of monomer units (the m/d ratio) was calculated knowing the amount of polymer and the phenosafranine present per unit volume of the solution. Analytical Methods Absorption spectra of the polymer samples in aqueous solutions and in film state (coated onto a glass plate) were recorded using a Beckman 25 double-beam spectrophotometer. The emission spectra of the polymer-dye sample were recorded in a Perkin-Elmer MPF 44B fluorescence spectrophotometer. Flash-photolysis studies were carried out using Applied Photophysics Ltd, KN-020 model flash kinetic spectrometer. The experi- mental methods used for the flash-photolysis studies were described earlier." In the flash-photolysis studies the dye concentration in aqueous solution was maintained at ca.lo-' mol dmP3 and the dye solution in the flash cell was freed of oxygen by passing purified nitrogen through it. Transients were monitored using a tungsten-halogen lamp in the region 600-730 nm. In the outer jacket of the sample cell acetone was taken to filter off radiations below 300 nm. The monitoring beam was first passed through a filter cell containing phenosafranine solution to prevent any steady photolysis of pheno- safranine in the reaction mixture. The absorption spectra of semi-dye species were recorded in the region 60&730 nm, where the absorbance due to phenosafranine and polymer-bound phenosafranine are negligible. The absorbance of the transient was recorded after 0.5ms of the flash at different wavelengths and the spectrum of the transient species was obtained.The electrochemical data were obtained using PAR modules 173 potentiostat, 175 universal programmer and 176 current follower. The reference electrode (SCE) was connected to the potentiostat through a PAR 178 electrometer. The photopotential was measured using an Aplab digital voltmeter and the photocurrent was measured using a Radelkis- 105 universal polarograph. The cell consists of two compartments interconnected by a salt bridge or a single- compartment cell. A plain platinum electrode and polymer-dye-coated platinum or carbon electrode or both platinum electrodes coated with different polymer-dyes were employed. The distance between the two electrodes was maintained at 4+ 1 mm in the single-compartment cell.The polymer-bound phenosafranine was coated onto a platinum or carbon electrode by placing a solution of the macromolecular dye on the surface of the electrically cleaned and dried electrode surface by blowing hot air over the816 Macromolecular- bound Phenosafranine Dyes 0.8 550 650 450 a/nm 350 Fig. 2. Absorption spectra of phenosafranine and P(AMPS+-co-MAAM) : (a) P(AMPS+- m/d = 41 ; ( d ) phenosafranine. CO-MAAM), m/d = 16 ; (b) P(AMPS+-co-MAAM), m/d = 38 ; (c) P(AMPS+-co-MAAM), 0.8 0.6 8 B 0.4 4 0.2 0 400 500 600 A/nm Fig. 3. Absorption spectra of phenosafranine and P(AMPS+-co-MAAM-co-VP) : (a) P(AMPS+- co-MAAM-co-VP), m/d = 21 ; (b) P(AMPS+-co-MAAM-co-VP), m/d = 32; ( c ) phenosafranine.R. Ramaraj and P .Natarajan 817 605 625 645 665W 625 645 665 685 (b), (c) Llnm Fig. 4. Luminescence spectra of (a) phenosafranine, (b) P(AMPS+-co-MAAM) and (c) P(AMPS+-co-MAAM-co-VP). surface. The evaporation of the solvent leaves behind a film on the electrode, In the case of polymer-dye P(AMPS+-co-MAAM-co-VP) the electrodes were immersed in the cell containing 10 cm3 of 6:4 CH,CN:H,O (v/v) and potassium chloride or sodium perchlorate as supporting electrolyte. The pH of the solution was adjusted with perchloric acid or sodium hydroxide. The solutions were deaerated by passing oxygen- free nitrogen for 30 min. The irradiation source was 1000 W tungsten lamp and a 5 cm water filter cell was used to cut off infrared radiations; the design of the lamp is such that the light output used for irradiating the electrodes is more like a 300 W lamp. Results The absorption spectra of poly(acrylamidomethylphenosafranine-co-methylolacryla- mide) with different m/d ratios obtained in solution are shown in fig.2. The absorption spectra of poly(acrylamidomethyIphenosafranine-co-methylolacrylamide-co-vinyl- pyridine) with different m/d ratios are shown in fig. 3. The emission spectra of phenosafranine dyes are shown in fig. 4. Typical cyclic voltammograms of phenosafranine and polymer-bound phenosafranine systems are shown in fig. 5 . The cathodic peak potentials (E& anodic peak potentials (E,& separation of peak potentials (AE,) and the cathodic peak current (I,,) at various scan rates are given in table 1. The absorption spectra of the transients were obtained following flash photolysis of phenosafranine under various conditions and the results are shown in fig.6. The absorbance of semiphenosafranine and polymer bound semiphenosafranine with different m/d ratios at different time intervals after the flash are measured and the corresponding plots for the decay kinetics of the transients are shown in fig. 7 and the rate constants are given in table 2.818 Macromolecular-bound Phenosafranine Dyes 0 -0.2 -0.4 -0.6 EIVvs. SCE -1.2 -1.0 -0.8 - 0.6 - 0.4 - 0.2 0 0.2 0.4 0.6 0.2 -0.2 -0.6 -1.0 EIV us. SCE 0 -0.2 -0.4 -0.6 -0.8 EIVvs. SCE EIVvs. SCE Fig. 5. Cyclic voltammograms of phenosafranine and polymer-bound phenosafranines. Working electrode, carbon plate ; supporting electrolyte, KCl.(a) Phenosafranine in homogeneous solution, (b) phenosafranine adsorbed on carbon, (c) P(AMPS+-co-MAAM) coated onto carbon and ( d ) P(AMPS+-co-MAAM-co-VP) coated onto carbon. Potential scan rate, v/mV s-' : (I) 10, (11) 20, (111) 50, (IV) 100 and (V) 200. Table 1. Cyclic-voltammetric data of phenosafranine and polymer-bound phenosafranine" (A) PS+ (present in solution) 10 -0.36 -0.320 40 125 20 -0.36 -0.315 45 220 50 -0.37 -0.310 60 520 100 -0.38 -0.305 75 970 (B) PS+ (adsorbed on carbon) 10 -0.44 -0.32 120 20 -0.46 -0.28 180 - 50 -0.48 -0.24 240 100 -0.54 -0.62 380 200 -0.62 -0.10 520 - (C) P(AMPS+-co-MAAM) (coated on carbon), m/d = 41 - - - 10 -0.35 -0.44 -0.33 -0.43 20 10 20 -0.35 -0.44 -0.32 -0.43 30 10 50 -0.36 -0.46 -0.31 -0.43 50 30 100 -0.38 -0.47 -0.30 -0.42 80 50 (D) P(AMPS+-co-MAAM-co-VP) (coated on carbon), 10 -0.36 -0.60 -0.31 -0.56 50 40 20 -0.38 -0.61 -0.31 -0.56 70 50 50 -0.40 -0.63 -0.30 -0.54 100 90 m/d = 21 100 -0.43 -0.66 -0.27 -0.53 160 130 a [KCl] = 0.1 mol dm-3, solvent = 8 : 2 = CH3CN : H,O (v/v), potential range = + 0.0 to - 0.8 V, working electrode = carbon.R. Ramaraj and P.Natarajan 819 7.0 6.4 5.8 5.2 4.6 5 4.0 3 3.4 2.8 2.2 I .6 P 3 P 1.8 I .7 1.6 1.5 a 4 X 1.4 8 1.3 4 1.2 1. I 1.0 0.9 P 1 I I 1 1 1 ' 0.8 600 620 640 660 680 700 720 740 Alnm 1.0 1 ' Fig. 6. Absorption spectra of semiphenosafranine formed from A, phenosafranine with EDTA ; ., phenosafranine without EDTA; 0, P(AMPS+-co-MAAM), m / d = 16 with EDTA; A, P(AMPS+-co-MAAM), m/d = 16 without EDTA; 0, P(AMPS+-co-MAAM), m/d = 41 with EDTA; 0, P(AMPS+-co-MAAM-co-VP), m/d = 21 with EDTA.0 : 160 140 I20 too 2 80 60 40 20 a I I I I I I 0 2 4 6 8 10 12 14 m o o 0 0' tlms Fig. 7. Kinetics of the decay of macromolecular semiphenosafranine dyes. A, Phenosafranine with EDTA; 0, phenosafranine without EDTA; 0, P(AMPS+-co-MAAM), m/d = 16 with EDTA; ., P(AMPS+-co-MAAM), m/d = 16 without EDTA; A, P(AMPS+-co-MAAM), m/d = 41 without EDTA; 0, P(AMPS+-co-MAAM-co-VP), m/d = 21 with EDTA.820 Macromolecular- bound Phen osafran ine Dyes Table 2. The disproportionation rate constant of semipheno- safranine and polymer-bound semiphenosafranines sample k,/ los dm3 m/d EDTA mo1-'s-l phenosafranine (PS') - absent 10.10 phenosafranine (PS') present 3.19 P(AMPS+-co-MAAM) 16 absent 1.98 P(AMPS+-co-MAAM) 41 absent 1.63 P(AMPS+-co-MAAM) 16 present 1.40 - P(AMPS+-co-MAAM-co-VP) 2 1 present 1.02 Table 3.Electrodic behaviour of polymer-bound phenosafranine coated onto platinum electrodes in a single-compartment cell in the presence of light sample AqJmV i/pA electrolyte P(AMPS+-CO-MAAM)~ + 22 + 121 + 24 + 120 + 40" P(AMPS+-CO-MAAM-CO-VP)~ -25 -111 - 33 - 125 - 37" +0.16 + 0.66 +0.16 + 0.63 + 0.50 -0.10 - 6.00 -0.16 - 6.00 - 0.40 0.1 mol dm-3 KC1 0.1 mol dm-3 KCl + 0.1 mol dm-3 NaClO, 0.1 mol dmT3 NaClO, + lop3 mol dmP3 0.1 mol dm-3 KCI + 0.1 mol dm-3 KCl 0.1 mol dm-3 KC1 + lop3 mol dmP3 0.1 mol dm-3 NaC10, 0.1 mol dm-3 NaClO, + 0.1 mol dm-3 KCl+ lop3 mol dmP3 mol dmP3 HCl HClO, mol dm-3 HCl NaOH mol dm-3 NaOH NaOH a Air-equilibrated solution. 6 : 4 CH3CN : H,O (v/v) medium.Coated onto carbon electrodes. 7 : 3 CH3CN : H,O (v/v) medium. Table 4. Effect of acid concentration on the electrodic behaviour of polymer- bound pheno- safranine, P(AMPS+-co-MAAM) coated onto a platinum electrode in a single-compartment cell in the presence of light HCl/mol dm-3 AE:,/mV i/PA 0 73 0.06 10-4 89 0.13 10-3 135 0.27 1 o-, 82 0.14 a Electrolyte = 0.1 mol dm-3 KCl. 6 : 4 CH3CN : H,O (v/v) medium.R. Ramaraj and P. Natarajan 82 1 Table 5. Effect of sodium hydroxide on the electrodic behaviour of polymer-bound phenosafranine, P(AMPS+-co-MAAM-co-VP) coated onto platinum electrode in a single- compartment cell in the presence of light 0 - 25 -0.10 10-4 - 40 - 1.00 10-3 - 1 1 1 - 6.00 1 o-, - 70 -2.10 0.6 0.4 0.2 a Electrolyte = 0.1 mol dm-3 KCI.7: 3 CH3CN: H,O (v/v) medium. 1.8 1.2 3 > 0.6 Fig. 8. (a) Photocurrent for a P(AMPS+-co-MAAM)-coated electrode without EDTA ; (b) photocurrent for a P(AMPS+-co-MAAM-co-VP)-coated electrode without EDTA ; (c) photocurrent for one electrode coated with P(AMPS+-co-MAAM) and the other coated with P(AMPS+-co-MAAM-co-VP) coupled and kept in a two-compartment cell without EDTA ; ( d ) photocurrent for the P(AMPS+-co-MAAM jEDTA system: L,, ‘light on’ condition for a P(AMPS+-co-MAAM)-coated electrode ; D,, ‘ light off’ condition for a P(AMPS+-co-MAAM)- coated electrode; L,, ‘light on ’ condition for a P(AMPS+-co-MAAM-co-VP)-coated electrode; D,, ‘ light off’ condition for a P(AMPS+-co-MAAM-co-VP)-coated electrode. The open-circuit photopotential and photocurrent for polymer-bound phenosafranine coated onto platinum and carbon electrode systems measured in single- and two- compartment cells are given in table 3.The photopotential and photocurrent of the electrodes measured at different concentrations of hydrochloric acid and sodium hydroxide are given in tables 4 and 5, respectively. The photocurrent was measured at different times of irradiation of the polymer-coated electrodes and the results are shown in fig. 8. The photocurrent was measured in the photoelectrochemical cell with polymer- bound phenosafranine-coated carbon electrode systems with and without EDTA and plain platinum electrode at different applied potential with respect to saturated calomel electrode (SCE) and are shown in fig. 9. Discussion Interest in the redox behaviour of phenazine-based compounds stems from their use as dyestuffs as well as from the recognition of the biological importance of some phenazine derivatives.Prior to any recorded voltammetric measurements of phenazine, it has been shown that a stable, one-electron reduction product (semiquinone) is formed during822 Macromolecular- bound Phenosafranine Dyes 20 1 I I 1 I I 1 1 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 a 0 -0.1 -0.2 -0.3 -0.4 -0.5 -0.6 -0.7 -0.8 -0.9 o 80 60 - 40 - 20 - 3n 1 I I 1 I I 1 1 1 I LV 0.2-0.1 o 7 6 5 4 3 2 I 0 - I 4 3 EIVvs. SCE 70 60 3 50 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 EIVvs. SCE Fig. 9. A (a) Photocurrent plotted against applied voltage (us. SCE) across the working- electrode-reference-electrode for a P(AMPS+-co-MAAM)-coated electrode-EDTA system.(b) Photocurrent plotted against applied voltage (us. SCE) across the working-electrode-reference- electrode for the PS+-EDTA system. B (a) Photocurrent plotted against applied voltage (us. SCE) across the working-electrode-reference-electrode for a P(AMPS+-co-MAAM-co-VP)- coated electrode without EDTA. (b) Photocurrent plotted against applied voltage (us. SCE) across the working-electrode-reference-electrode for a P(AMPS+-co-MAAM)-coated electrode without EDTA.R. Ramaraj and P . Natarajan 823 potentiometric titrations in different solvents. l6 Muller l7 has shown that the information on the formal potentials and semiquinone intermediate of a number of compounds could be obtained from polarograms. Previous studies on phenazine16* l7 indicate the following reaction sequence for solutions, in which single cathodic and anodic waves were observed : +Hf -Hf PS+ + e- e PSH'+ +Hf -Hf PSH'+ + e- G= PSH;.In aqueous solution, phenazine (PS+) undergoes two one-electron reduction to give leucophenazine (PSH;) via a one-electron intermediate semiphenazine (PSH'+) which is a protonated species. For the phenosafranine dye at ca. pH 6.5 only one cathodic and one anodic wave are observed [fig. 5 (a)] as established for phenazine derivatives in neutral solutions. 179 la Cyclic-voltammograms run for the dye adsorbed on carbon electrodes [fig. 5(b)] show one cathodic and one anodic wave. However, the peaks are broader and the peak separation (AE,) is large (table l), indicating that the electrodic process is not very reversible at the carbon electrode.It is observed that with the increase in the scan rate of potential, the cathodic curve shifts to more negative potentials (table 1). It is also observed that the separation of peak potentials increases with an increase in the potential scan rate. This observation is explained on the basis of an electron-transfer-electron- transfer mechanism (E-E mechanism) coupled with a fast protonation reaction. l9 In contrast to the behaviour of the monomeric dye, polymer-bound phenosafranine shows [fig. 5(c) and (d)] two anodic and two cathodic waves. The peaks are broader and the peak separation is larger for polymeric phenosafranine as compared to that of monomeric phenosafranine (table 1). When the dye is attached to the macromolecule, a shift in the peak potentials is observed.The cyclic voltammograms of phenosafranine and polymer-bound phenosafranine are not found to be similar, and the two-electron reduction of semiphenosafranine could be separately observed when polymer chain is attached to the dye [table 1 and fig. 5(c) and (d)]. During the reduction of polymer bound phenosafranine, protonation takes place indicating a slow chemical reaction and the process can be described in terms of an electron-transfer-chemical-reaction-electron- transfer (ECE) mechanism. Absorption Spectra of Semiphenosafranine attached to Macromolecular Chains Flash-photolysis studies of the safranine-0,-EDTA system have been 2o The dye phenosafranine is easily photoreduced by electron donors such as EDTA.Semiphenosafranine radicals were produced upon irradiation of a solution containing phenosafranine and EDTA. 21 The photoreduction involves electron transfer from EDTA to the excited triplet state of the phenosafranine molecule. In the present investigation the flash-photolysis experiments of phenosafranine-EDTA and polymer- bound phenosafranine-EDTA systems and the monomeric and polymeric dye in the absence of EDTA were carried out to investigate the effect of polymer backbone on the properties of the semiphenosafranine radical. The absorption spectrum of the free semiphenosafranine radical produced on flash photolysis of the phenosafranine-EDTA system shows an absorption maximum around 640nm as shown in fig. 6. The absorption spectra of semiphenosafranines produced on flash irradiation of polymer- bound phenosafranine systems with EDTA as shown in fig.6 indicate absorption maxima at 680 nm. The polymer-dye P(AMPS+-co-MAAM-co-VP) does not give any transient absorption in the absence of EDTA in solution. The shift in the absorption maxima for the polymer-bound semiphenosafranines compared to that for free824 Macromolecular- bound Phenosafranine Dyes semiphenosafranine is presumably due to the -CH,- group present in the polymer chain. The absorption maxima of the transient obtained for phenosafranine in the presence and in the absence of EDTA during flash photolysis are the same (640 nm). The absorption spectra of the polymer-bound semiphenosafranines obtained for macromolecular-bound phenosafranines P(AMPS+-co-MAAM) and P(AMPS+-co- MAAM-VP) in the presence and absence of EDTA are the same (680 nm).It is concluded that the transient formed from P(AMPS+-co-MAAM) in the absence of EDTA is polymer-bound semiphenosafranine. The absorption maxima of semi- phenosafranine attached to two different polymer chains seen around the same wavelength (680 nm) suggest that the absorption spectrum of the semiphenosafranine radical attached to a polymer chain is not significantly influenced by the nature of the polymer backbone. Failure to see any transient species on flash photolysis of P(AMPS+-co-MAAM-co-VP) may be due to the formation of an adduct between the excited dye centre and the pyridine groups present in the polymer chain, indicating that the photochemical behaviour of P(AMPS+-co-MAAM-co-VP) is different from that of P(AMPS+-co-MAAM).Kinetics of the Decay of the Polymer-bound Semiphenosafranines The decay of semiphenosafranine radical produced by electron transfer from EDTA to excited phenosafranine observed on flash excitation is represented by the equations PS+ + H+ + EDTA -+ PSH'+ + EDTA,, 2PSH'+ -+ PSHg + PS+ (3) (4) where PSH'+ = semiphenosafranine and PSH; = leucophenosafranine. Semipheno- safranine is also produced on flash photolysis of phenosafranine without EDTA present in solution. The second-order rate constant k, for the disproportionation reaction of semiphenosafranine (PSH'+) produced with and without EDTA present in the solution was found to be 1.01 x lo9 s-l. The decay of the transient formed from the polymer-dye also followed equal concentration second-order kinetics.The formation of semireduced dye from excited phenosafranine and polymer-bound phenosafranine in the absence of EDTA is attributed to the reaction *PSH2+ + PS+ -+ 'PSH+ + 'PS2+. ( 5 ) This observation is supported by earlier studies Kosui et al. and Baumgartner et al. carried out for monomeric phenosafranine.209 22 The latter group20 was unable to detect the transient absorption which can be attributed to the semioxidized dye (PS2+) even in the presence of Fe3+ or H202. The failure to detect the semioxidized dye may reflect the rapid reduction by the solvent. The rate constants for the decay of the monomeric semi- reduced dye and polymer-bound semireduced dye are found to be ca. lo9 dm3 mol-' and los dm3 mol-l, respectively (table 2).The rate constant for the disproportionation reaction of macromolecular semiphenosafranine are ca. 10 times slower than that of the monomeric semidye. A similar observation is reported in the case of thionine and polymer-bound thionine ~ystems.'~ This is due to the difference in the diffusional rate of the monomeric phenosafranine and polymer-bound phenosafranine. The polymer-bound dye P(AMPS+-co-MAAM-co-VP) does not give any transient on flash photolysis in the absence of EDTA, whereas in the presence of EDTA a transient similar to that for polymer dye P(AMPS+-co-MAAM) is observed. The flash-photolysis results also show that addition of 0.1 mol dm-3 pyridine to the experimental polymeric dye solution quenches the excited phenosafranine dye.This may be due to the formation of an adduct between the excited dye centre and the pyridine units. In the polymer-dye P(AMPS+-co-MAAM-co-VP) the adduct formation may be enhanced by the polymerR. Ramaraj and P. Natarajan 825 backbone, which brings the dye centre and pyridine units to a closer and more rigid environment. The flash-photolysis studies of polymer-dye P(AMPS+-co-MA AM) and P(AMPS+-co-MAAM-co-VP) in the absence of EDTA show a difference in photochemical reaction. It is suggested that this difference in photochemical behaviour is due to the presence of pyridine units in the polymer chain along with the dye centre. Photoelectrochemical Cell to Produce Fuels like Hydrogen using Macromolecular- bound PhenosafranineEDTA Systems The phenosafranine-EDTA system was suggested as a possible pho toelectrochemical cell to convert photoenergy into electricity or to evolve hydrogen from water.21 The dye phenosafranine is easily photoreduced by electron donors such as EDTA.Its longest- wavelength absorption is centred at 522 nm. Anaerobic photolysis in the presence of EDTA leads to disappearance of the absorption peak at 522 nm, with a new peak appearing around 640 nm for monomeric phenosafranine and 680 nm for polymeric phenosafranine. The current-potential curves presented in fig. 9 show that the photogenerated reductant is easily oxidized at the carbon anode at potentials around 0.1 V in 0.5 mol dmP3 H,SO, catholyte. The cell is given as carbonlphenosafranine, EDTA, pH 71 (0.5 mol dm-3 H,SO,IPt. Cyclic-voltammetric studies of phenosafranine in water showed that the midpoint potential for the first electron reduction is located at -0.3 V us.SCE. The current-potential curve of P(AMPS+-co-MAAM) coated onto the carbon electrode (fig. 9) shows that the photochemically produced semi-dye is easily oxidized at the carbon anode at potentials around -0.1 V vs. SCE. The cyclic-voltammetric data for P(AMPS+-co-MAAM) or P(AMPS+-co-MAAM-co-VP) coated onto carbon electrodes indicate the occurrence of a two-electron redox process which corresponds to the formation of semi-reduced dye at -0.54 V us. NHE and leucodye at -0.69 V us. NHE. In the cathode compartment, protons are getting reduced at ca. pH 1. Photoelectrochemical Reactions at Macromolecular-bound Phenosafranine-coated Electrode : Example of a Regenerative-type Water-splitting Cell In this section certain properties of platinum or carbon electrodes coated with polymer- bound phenosafranine dye on irradiation with visible radiation are discussed.The electrode coated with P(AMPS+-co-MAAM) shows a photopotential in air-equilibrated solutions either in the single compartment or in the two-compartment cell with an uncoated platinum plate as the counter-electrode. Under the same conditions an increase in the acid concentration leads to an increase in the open-circuit photopotential and photocurrent [fig. 8 (a)]. As far as the P(AMPS+-co-VP) coated electrode is concerned, it shows opposite polarity compared to an analogous cell with P(AMPS+-co-MAAM) [fig. 8 ( b ) ] . In this case, however, in alkaline (pH 10) solution, an enhancement in the photocurrent and photopotential is observed.In experiments with two compartment cells where one of the electrodes is coated with P(AMPS+-co-MAAM) and the other coated with P(AMPS+-co-MAAM-VP), coupled and kept in different compartments connected by a salt bridge, the observed current is a sum of that observed for the individual electrodes kept separately along with a plain platinum electrode in a single- compartment cell [fig. s(~)]; in this case both the electrodes were exposed to light. The current-voltage diagram for the electrodes coated with P(AMPS+-co-MAAM) and P(AMPS+-co-MAAM-co-VP) under illumination are given in fig. 9 B. Cyclic-voltammetric data for the macromolecular phenosafranines P(AMPS+-co- MAAM) and P(AMPS+-co-MAAM-co-VP) coated onto a carbon electrode indicate a two-electron redox process in both the cathodic and anodic waves which corresponds to826 Macromolecular-bound Phenosafranine Dyes the formation of semireduced dye at a potential of -0.54 V us.NHE and the leucodye at a potential of -0.69 V vs. NHE. Photoexcitation of the monomeric dye in homogeneous solution leads to the oxidation of EDTA or ascorbic acid present in solution.'^ 22, 23 The semireduced dye undergoes a disproportionation reaction to give back phenosafranine and leucophenosafranine. The results in the present investigation indicate the excitation of the macromolecular dye in the film, which produces either a cathodic or anodic reaction at the electrode, depending upon the nature of the macromolecule. Poly(acrylamidomethylphenosafranine-co-methylolacrylamide) acts as a photocathode whereas poly(acrylamidomethylphenosafranine-co-methylolacrylam- ide-co-vinylpyridine) behaves as an anode with respect to a plain platinum electrode with the electrodes dipped in an aqueous electrolyte solution.The photoelectrochemical reactions are proposed to involve water and oxygen in the redox processes at the electrodes. The photopotential depends upon the extent of oxygen present in the solution, which suggests that the electrode coated with macromolecular dye reduces oxygen to water at the cathode and oxidizes water to oxygen at the anode. The direction of the current is reversed depending upon the nature of the polymer-dye film. The redox potentials for the processes 0, + 4H+ + 4e- -+ 2H,O (6) 0, + e- -+ 0; (7) O;+e--+O;- (8) are, respectively, 0.82, -0.50 and - 1.80 V us.NHE.,* The redox chemistry of oxygen-water systems has been of interest for a long time, and the mechanism of electron-transfer processes depends upon a variety of factors including the nature of the medium. Thus, the reaction (9) is catalysed in the forward and reverse directions by the polymer films of P(AMPS+-co- MAAM-co-VP) and P(AMPS+-co-MAAM), respectively. The methylolacrylamide- bound phenosafranine photocatalyses the reduction reaction, and the same polymeric dye in the presence of pyridine in the copolymer units catalyses the oxidation reaction in the presence of light. The different photochemical behaviour observed for P(AMPS+-co-MAAM) and P(AMPS+-co-MAAM-co-V) by flash-photolysis experi- ment supports the results observed in this photoelectrochemical cell.The fact that the polymer-dye film itself is stable for several cycles indicates that organic molecules in the system are not consumed. Negligible photopotential or photocurrent observed in deaerated solutions shows that oxygen is involved in the photoelectrochemical reaction. In a single-compartment cell with both the electrodes coated by P(AMPS+-co-MAAM) and P(AMPS+-co-MAAM-co-VP) the cyclic reaction catalysed by light occurs at the electrode. In a two-compartment cell, one electrode coated with P(AMPS+-co-MAAM) and dipped in a solution containing supporting electrolyte at pH 3 and the other electrode coated with P(AMPS+-co-MAAM-co-VP) dipped in a solution containing supporting electrolyte at pH 10 shows the additive response of maximum photopotential and photocurrent under irradiated conditions [fig.8(c)]. This cell is perhaps one of the first examples of a regenerative-type photocell using water. In the case of the electrode coated with the polymer film P(AMPS+-co-MAAM), acrylamide becomes swollen because it is a hydrophilic polymer, and the swollen electrode acts as a cathode, whereas because the vinylpyrodine copolymer is hydrophobic, the electrode coated with the polymer film P(AMPS+-co-MAAM-co-VP) shows anodic polarity on irradiation. This type of chemically modified electrode has potential to devise new electrodes to bring about specifically a desired reaction at the electrode.2H,O + 0, + 4H+ + 4e-R. Ramaraj and P. Natarajan 827 The investigations reported here are partially supported by the D.S.T. Thrust Area S.E.R.C. programme and by the U.G.C. COSIST programme. References 1 Energy Resources Through Photochemistr,v and Catalysis, ed. M . Gratzel (Academic Press, New York, 2 K. Kalyanasundaram, Coord. Chem. Rev., 1982, 46, 159. 3 J. Kiwi, K. Kalyasundaram and M. Gratzel, Struct. Bonding, 1982, 49, 37. 4 E. Rabinowitch, J. Chem. Phys., 1940, 8, 551; 560. 5 M. Eisenberg and H. P. Silverman, Electrochim. Acta, 1961, 5, 1. 6 M. Kaneko and A. Yamada, J. Phys. Chem., 1977,81, 1213. 7 K. K. Rohatgi-Mukherjee, M. Bagchi and B. B. Bhowmick, Electrochim. Acta, 1983, 28, 293. 8 J. H. Fendler and E. J. Fendler, Catalysis in Micellar and Macromolecular Systems (Academic Press, 9 D. G. Whitten, Acc. Chem. Res., 1980, 13, 83. 1983). New York, 1975). 10 M. Kaneko and A. Yamada, Adu. Polym. Sci., 1984, 55, 1. 11 R. W. Murray, in Electroanalytical Chemistry, ed. A. J . Bard (Marcel Dekker, New York, 1984), 12 R. Tamilarasan and P. Natarajan, Nature (London), 1981, 292, 224. 13 R. Tamilarasan, R. Ramaraj, R. Subramanian and P. Natarajan, J. Chem. SOC., Faraday Trans. I , 14 A. I. Vogel, Quantitative Inorganic Analysis (ELBS, London, 1975), p. 329. 15 R. Ramaraj, R. Tamilarasan and P. Natarajan, J. Chem. SOC., Faraday Trans. I , 1985, 81, 2763. 16 L. Michaelis, Chem. Rev., 1935, 16, 243. 17 0. H. Muller and J. P. Baumberger, Trans. Electrochem. Soc., 1937, 71, 181. 18 D. N. Baile, D. M. Hercules and D. K. Roe, J . Electrochem. SOC., 1969, 116, 190. 19 R. Nichlolson and I. Shain, Anal. Chem., 1965, 37, 178. 20 C. E. Baumgartner, H. H. Richtol and D. A. Aikens, Photochem. Photobiol., 1981, 34, 17. 21 M. Neumann-Spallart and K. Kalyanasundaram, J. Phys. Chem., 1982, 86, 268. 22 N. K. Kosui, K. Uchida and Koizumi, Bull. Chem. SOC. Jpn, 1965, 38, 1958. 23 R. Bhardwaj, R. C. Pan and E. L. Grass, Photochem. Photobiol., 1981, 34, 215. 24 J. Wiltshire and D. T. Sawyer, Acc. Chem. Res., 1979, 12, 105. p. 191. 1984, 80, 2405. Paper 8/01079J; Received 15th March, 1988