Photosensitized Water Reduction Mediated by Semiconductors

7

Photosensitized Water R e d u c t i o n M e d i a t e d by Semiconductors Dispersed i n M e m b r a n e M i m e t i c Systems

Downloaded by NORTH CAROLINA STATE UNIV on May 3, 2015 | http://pubs.acs.org Publication Date: May 14, 1985 | doi: 10.1021/bk-1985-0278.ch007

YVES-M. TRICOT, RAFAEL RAFAELOFF,ÅSAEMEREN, and JANOS H. FENDLER Department of Chemistry, Institute of Colloid and Surface Science, Clarkson University, Potsdam, NY 13676 Colloidal semiconductor particles were in situ generated and coated by catalysts in reversed micelles, surfactant vesicles and polymerized surfactant vesicles. Sodium bis-2-ethylhexylsulfosuccinate (AOT) was used to form reversed micelle-entrapped water pools in isooctane. Platinized CdS, in situ generated in those water pools, sensitized water photoreduction by thiolphenol dissolved in the organic phase. Colloidal 30-50 Å diameter CdS particles were also in situ generated in 800-1000 Å diameter single bilayer vesicles prepared from dihexadecylphosphate (DHP), dioctadecyl dimethylammonium chloride (DODAC) and from the polymerizable surfactant [C H CO (CH ) ] N (CH )CH C H CH=CH , CI (1). Band gap excitation of the vesicle-embedded CdS resulted in weak fluorescence which could be quenched by electron donors and acceptors. Electron transfer was also examined by laser flash photolysis. Visible-light excitation of in situ rhodium-coated DHP-, DODAC- and 1-entrapped CdS particles led to sustained hydrogen production at the expense of sacrificial electron donors. Generation of CdS in polymerized 1 offered a number of advantages. Optimization of These systems and their potential in artificial photosynthesis are described +

15

31

2

2

2

2

3

2

-

6

4

2

The goal of artificial photosynthesis is to lay the foundation- for efficient and economically viable conversion of sunlight to chemical energy. Considerable advances towards this goal have been made by different laboratories around the world (1-4). One general approach has involved the separate studies of sacrificial reduction and oxidation half cells. Hydrogen has been generated in reduction half cells, following light harvesting by a sensitizer (S), electron transfer through relays (R), and catalytic water reduction at the expense of a sacrificial electron donor (D), Sacrificial photosensitized water reduction has been extensively investigated both in homogeneous solutions and in the presence of organized assemblies (micelles, microemulsions, polyelectrolytes, polymers, vesicles and 0097-6156/85/0278-0099$06.00/0 © 1985 American Chemical Society

In Organic Phototransformations in Nonhomogeneous Media; Fox, M.; ACS Symposium Series; American Chemical Society: Washington, DC, 1985.


100

O R G A N I C P H O T O T R A N S F O R M A T I O N S IN N O N H O M O G E N E O U S M E D I A

even clay particles) (5.6). O r g a n i z e d assemblies mimic the functions of the thylakoid membrane. THey provide compartments of potentially controllable microenvironments for the sensitizers, relays, electron donors and catalysts and allow v e c t o r i a l charge separations. T y p i c a l l y , ruthenium complexes and porphyrins have been used as sensitizers, viologens as relays (Figure 1A), ascorbic a c i d , thiols and E D T A as e l e c t r o n donors, and colloidal platinum and palladium as catalysts. Although this approach continues to be used and improved upon, replacement of the sensitizer and the relay by dispersed colloidal semiconductors has gained widespread popularity (Figure IB) (4). Colloidal semiconductors, l i k e their more robust solid-state analogs, a r e c h a r a c t e r i z e d by their band gap, the energy difference between the filled valence band (VB) and the vacant conduction band (CB). Photoexcitation of the semiconductor (SC) at a wavelength corresponding to the band gap results in the promotion of an e l e c t r o n from the valence to the conduction band, and hence in charge separation: SC - J u l

e" + h

>

+

(1)

E l e c t r o n (e") and hole (h ) recombinations, in the absence of quenching, result in radiationless transition (i.e., emission of heat) and, in some cases, in fluorescence (i.e., emission of ugnt): Δ e" + h < ^ ; (2; hv C h a r g e carriers escaping to the surface of the semiconductors can transfer an e l e c t r o n to an acceptor (A) or a c c e p t one from a donor (D): +

+

e" + A h

+

+ D

> >

A"

(3)

D

(4)

+

at the semiconductor interface. Advantage has been taken of these i n t e r f a c i a l electron transfers for mediating c a t a l y t i c photosensitized water reduction (Figure IB). C o l l o i d a l dispersed T i O C d S , CdSe, F e O , , S r T i O and their mixtures have been used as semi conductors either in i h e absence of in the presence of sensitizers. S a c r i f i c i a l e l e c t r o n donors used have included E D T A , thiols and alcohols and Pt, Pd and R h have been used as catalysts. There are a number of advantages of using colloidal, semiconductors in a r t i f i c i a l photosynthesis. They are relatively inexpensive. They have broad absorption spectra and high e x t i n c t i o n . c o e f f i c i e n t s at appropriate band gap energies. Nevertheless, they can be made optically transparent enough to allow direct flash photolytic investigations of e l e c t r o n transfers. They can be modified by derivatization or sensitizer a d s o r p t i o a Importantly, electrons produced by band gap excitation can be used directly without relays for c a t a l y t i c water reduction (Figure IB). Unfortunately, colloidal semiconductors also suffer from a number of disadvantages. T h e y are notoriously difficult to form reproducibly as small (smaller than 200 A in diameter) monodispersed particles. Such small particles are needed for obtaining high surface areas and achieving efficient electron transfer to catalysts and/or other molecules adsorbed at the semiconductor interfaces. However, a minimal size is necessary to obtain semiconductor properties and spatial separation of the e l e c t r o n from the hole. A s i z e range of 40 - 60 A has been suggested to be ideal for the use of colloidal semiconductors in photochemical solar energy c



7.

TRICOT ET A L .

Photosensitized

Water

101

Reduction

conversion (7). Semiconductors are equally difficult to maintain in solution for extendecT times in the absence of stabilizers which bound to a f f e c t their photoelectrical behavior unpredictably. T h e i r modification, and their coating by catalysts are, at present, more of an a r t than science. Furthermore, the lifetime of e l e c t r o n - h o l e pairs in semiconductors is orders of magnitude shorter than the e x c i t e d s t a t e lifetime of t y p i c a l organic sensitizers. This is due to a very fast undesirable electron-hole recombinatioa Quantum yields for charge separations in colloidal semiconductors are, therefore, disappointingly low. Incorporation of c o l l o i d a l semiconductors into polyurethane films (8) and Nafion membranes (9) were reported to have overcome some of these disadvantages. Research in our laboratories is focussed upon t h e utilization of membrane mimetic systems for r e a l i z i n g the full potential of colloidal semiconductors. We Ijave developed methologies for the in situ formation of small (ca, 40 A diameter) and uniform c a t a l y s t - c o a t e d colloidal semiconductors in reversed micelles (10), surfactant vesicles (11) and polymerized surfactant vesicles (12). U t i l i z a t i o n of these surfactant aggregate-stabilized semiconductor particles in photosensitized charge separation and hydrogen generation is the subject of the present report. Semiconductors in Reversed Micelles C o l l o i d a l CdS was in situ generated in reversed micelles formed by sodium bis-2-ethylhexyi sulfosuccinate ( A O T ) . Aqueous C d C l ^ or C d ( N O J was added to isooctane solutions of the surfactant to obtain a water το A O T ratio of 20 (10). Exposure to c o n t r o l l e d amounts of gaseous H S resulted in C d S formaTToa Excess H S was removed by A r g o n bubbling T h e onset of absorption of this optically clear C d S colloidal dispersion was approximately 500 nm (Figure 2), which is slightly blue-shifted compared to the band gap of 520 nm (2,40 eV) of bulk crystalline C d S (4). Such shifty in band gap have been shown to occur in CdS particles smaller than 100 A in ^iamet^er (13-15). A t ^00 nm, the e x t i n c t i o n coefficient was 900 ± 100 M" c m " , based on C d atomic absorption determinations. Dynamic light-scattering measurement prior and subsequent to H S exposure revealed only # slight increase in the apparent radius of trie water pool from 60 to 75 \ 2

A O T reversed micelle-entrapped, colloidal T i O particles were prepared by mixing anhydrous isooctane solutions of Ti-tetraisoprop oxide ( T i ( O C H ) ) with A O T / H Ο dispersions in isooctane having a r a t i o of 1Q H O r/er A O T molecule. Fast hydrolysis resulted in T i O formation. Particular c a r e had to be taken to a v o i d uncontrolled growth4>f the T i O particles. Due to the dynamic exchange between aggregates, initially dispersed T i O could form bigger particles by encountering other small particles, until they reached a point where they had a very low probability of c o l l i d i n g with each other. In 0.1 M A O T solution in isooctane and 1.0 M H O , 2 χ 10" M colloidal T i 0 remained stable and optically clear for several days. A t higher T i O concentrations, the turbidity increased with time and T i O particles eventually p r e c i p i t a t e d Reversed micelle-entrapped, colloidal C d S showed the c h a r a c t e r i s t i c weak fluorescence emission (Figure 2), previously observed in homogeneous solutions (16-19), However, the maximum emission intensity corresponded to full band gap emission (approximately 500 nm) and was not r e d - s h i f t e d as observed in homogeneous solution (17). This discrepancy might arise from the mode of preparation (H S instead of N a S ) , or from the s p e c i f i c effect of surfactant aggregates. A l t e r n a t i v e l y , this can be the result o f a size ?

2

2



102


F i g u r e 1. B a s i c features of s a c r i f i c i a l water reduction systems, (A) Homogeneous solution, with sensitizer, S, electron relay, R , s a c r i f i c i a l electron donor, D, and metal catalyst. (B) Catalyst-coated. colloidal semiconductor dispersion, obviating the need for electron relay.

F i g u r e 2. (Top) Stern-Volmer plots for the quenching of the fluorescence of c o l l o i d a l CdS in A O T - e n t r a p p e d water pools in isooctane by R M V (φ), M V ^ Π ) and PhSH (Q). (Bottom) Absorption and emission spe ctra of c o l l o i d a l CdS in A O T - e n t r a p p e d water pools in isooctane. T h e shoulder observed at 400 nm is due to a spectrometer a r t i f a c t . Z +

2 +


7.

TRICOT ET A L .

Photosensitized

Water

103

Reduction

distribution of the c o l l o i d a l CdS particles, which c o u l d produce different band gap energies and hence, different fluorescence spectra (13-15). Replacing CdCl by Cd(NOj did not change the fluorescence c h a r a c t e r i s t i c s o r colloidal C d S . Fluorescence intensities of these solutions decayed biexponentiaily with lifetimes oL 2.4 and 28.5 nanoseconds. Addition of 2.3 χ 10" M methylviologen, MV \ decreased the fluorescence lifetime of the J o n g - l i v e d component to 18.7 nanoseconds. Similarly, addition of MV and a surface-active viologen, C H =C(CH ) C O O ( C H ) (C H N ) CH B r " , I" ( R M V ) , as well as P h S H , tjuenchetl the émission intensity (Figure 2). Apparent Stern-Volmer constants for quenching the fluorescence of c o l l o i d a l C d S by M V , R M V and P h S H are 2.6 χ K r M ~ * 4.6 χ 10 M ,- and 11,6 Μ , respectively. T h e low quenching constant o ^ P r i S H arises from local concentration effects. Polar quenchers, such as M V and R M V , are locally concentrated in the water pool / A O T interface, whereas PhSH is homogeneously dissolved in isooctane. Due to the very fast electron-hole recombination, the quenching mechanism must be, at least in part, of s t a t i c nature and dependant upon the absorbed concentration of quencher at the CdS p a r t i c l e surface. The Stern-Volmer appearance of the quenching data in Figure 2 may be, therefore, coincidental. In fact, a different behavior is observed in surfactant vesicles (vide infra). To be a c t i v e in H -production, a catalyst had to be incorporated in the system and depositecr on the semiconductor surface. P l a t i n i z a t i o n was c a r r i e d out by adding aqueous Κ P t C l solutions to the reversed micelle entrapped, colloidal CdS and i r r a d i a t i n g b y a 450 W Xenon lamp under A r bubbling for 30 minutes, Platinization was monitored absorption spectrophoto metrically. Irradiation of degassed, reversed micelle-entrappec^ platinized C d S by visible light (450 W Xenon lamp, λ < 350 nm) resulted in hydrogen formation upon addition of 1.0 χ 1 0 " M P h S H , and it could be sustained for 12 h. This was the consequence of electron transfer from P h S H to t h e positive holes in the colloidal C d S , which diminished electron-hole recombinations (Figure 3). T h e reversed micelles had the s p e c i f i c advantage of providing a means for charge separation by continuously removing the product (PhSSPh) from the semiconductor, located in the water pool, to t h e organic solvent. However, due to water pool exchanges, the semiconductor particles c o u l d interact with each other, which limited the concentration of semiconductor tolerable by the system. A better insulation of inner compartment was provided by aqueous surfactant vesicles. 2 +


+

2 +

Z +

V

Z +

3

Semiconductors in unpolymerized and polymerized surfactant

vesicles

CdS particles were in situ generated from C d and F L S in vesicles prepared from negatively charged dihexadecylphosphate, E m P , positively charged dioctadecyldimethylammonium chloride, D O D A C , and positively charged polymerizable [C H. C0 (CHJJN (CHJCH C.H.CH=CH ΟΓ L Sizes and sites of CdS p a r t i c l e s w er e c o n t r o l l e d b y n h ë amount o| C d adsorbed and by the method of preparation. Positively charged C d ions were readily a t t r a c t e d , o f course, to the negative surfaces of anionic D H P vesicles. Adsorption of C d ions to c a t i o n i c vesicles ( D O D A C and 1) was achieved by complexation with E D T A . In anionic D H P vesicles, up To one Cd ion per three surfactant molecules c o u l d be adsorbed. For cationic vesicles, up to one C d / E D T A complex per four surfactant molecules could be adsorbed, the difference arising presumably from s t e r i c hindrance. 2 +

1

0

+

9

2

z +

2 +

2 +



104

CdS particles were generated selectively inside, or outside, or on both sides of anionic D H P vesicles^ A f t e r nucleation, C d S particles grew by encounter with neighboring C d J o n s or with other C d S particles. In most experiments, a r a t i o of one C d ion per ten surfactant molecules was maintained. E l e c t r o n micrographs showed CdS particles to be in the 30 - 50 A s i z e range (11a). Band gap e x c i t a t i o n ( λ < 500 nm) of c o l l o i d a l C d S in anionic D H P vesicles resulted in c h a r a c t e r i s t i c weak fluorescence, due to e l e c t r o n - h o l e rec^mbinatioa The quantum y i e l d of fluorescence was estimated to be c a . 10" . E x c i t a t i o n at 330 nm gave fluorescence maximum at approximately 500 nm. Most surprisingly, d e t e c t a b l e fluorescence was produced only from colloidal CdS which was^ generated inside of Q H P vesicles (prepared by sonicating D H P with Cd and by removing C d from the outer vesicle surfaces by ion exchange chromatography prior to forming CdS) and also when the sonication p H was higher than 6. No fluorescence was observed on CdS generated from C d added to already formed vesicles, regardless of the p H and of the added C d concentration. It is believed that colloidal CdS on the outside of vesicles can "age" either by aggregation with other colloidal CdS particles during vesicle-vesicle collision, or by dissolution/recrystallization from C d and S ions in the outside aqueous solution (13). This aging c o u l d result in a decrease of the fluorescence quantum yield, although the mechanism of this e f f e c t is not clear. C o l l o i d a l C d S particles, when formed inside of the vesicles, are isolated from the outside bulk and hence are protected against these aging processes. Their fluorescence was very reproducible and independent o f the anions (which were excluded from the vesicles). The quenching of CdS fluorescence by methylviologen, MV , a m em bra ne-impermeable electron a c c e p t o r , is shown in F i g u r e 4. The highest quenching efficiency was obtained when both MV and C d S ^vere l o c a t e d inside of the vesicles. Lower quenching o c c u r r e d when MV and CdS were symmetrically distributed on bpth sides of the anionic vesicles, and no quenching was observed when MV was externally added, regardless of the location of C d S . This confirmed that only "inside" CdS fluoresced and, that M V did not diffuse through the D H P membrane. Observation of a time-dependent fluorescence quenching indicated a gradual relocation of CdS particles deeper inside the hydrocarbon part o f the membrane. Z +


z +

2 +

z+

+

z +

z +

+

+

Z +

Fluorescence quenching by M V is the consequence of e l e c t r o n transfer from the conduction band of e x c i t e d CdS to MV , with formation of radical cation M V (Figure 5). Due to the v e r y short ( < 1 ns) lifetime of the e l e c t r o n - hole pair in C d S , only adsorbed M V on the C d S surface was able to capture an electron. This is r e f l e c t e d in the non Stern-Volmer behavior of the quenching of c o l l o i d a l C d S by co-entrapped MV . In other configurations, the quenching apaears to be Stern-Volmer, but it is again misleading since not all the MV participates in the quenching process. D i r e c t evidence of this electron transfer was substantiated in anionic vesicles by laser flash photolysis (Figure 5). C d S fluorescence was also observed in c a t i o n i c D O D A C and 1 vesicles, albeit w i t h much less intensity than that found in D H P vesicles. The preparation pH influenced the CdS fluorescence in all vesicles. Interestingly, C d S fluorescence was somewhat stronger in unpolymerized than polymerized vesicles prepared from 1. This was not a consequence of the uv irradiation applied to polymerize tne vesicles, since such irradiation had no e f f e c t o n CdS fluorescence in nonpolymerizable surfactant vesicles (such as D H P or D O D A C ) . However, this effect correlated with the ability of E D T A to a c t as a s a c r i f i c i a l electron donor in hydrogen production experiments (vide infra). Z +

+ #

Z +

+

+



7.

TRICOT ET A L .

Photosensitized

Water

Reduction

105

F i g u r e 3. A n idealized model for t h e CdS-sensitized water photo reduction by P h S H in A O T reversed micelles in isooctane, V B = valence band, C B = conduction band.

10 χ M V

Z +

concentration,

M

F i g u r e d . Quenching of C d S (2 χ 10"^ M) fluorescence in D H P (2 χ 10" M) vesicles by MV in d i f f e r e n t locations: (Êk) : C d S and MV only inside; ( Q ) : C d S and M V are on b o t h s i d e s ; 30 min. after C d S formationjTu)) ι same samples, 8 days after C d S formation; i/\J' : C d S inside only, M V outside only; ( A ) :. CdS on both sides, MV outside only, λ = 330 n m, λ = 500 η m. Z +

2 +



106

In the presence oï a suitable electron donor, c a t a l y s t - c o a t e d , vesicle stabilized, in situ-generated colloidal C d S particles produced hydrogen in deaerated solutions under visible-light i r r a d i a t i o n (λ < 350 nm). In anionic D H P vesicles, rhodium was used as catalyst, whose precursor, R h , was co-adsorbed with C d on the surface of the vesicles. R e d u c t i o n of Rrr to Rh° could be achieved either by uv irradiation in deaerated solutions prior to addition of the e l e c t r o n donor, o r by visible-light irradiation in the presence of the e l e c t r o n donor. A slight induction period was observed in the latter case, but after 2 h of irradiation the same amount of hydrogen had been produced, and the r a t e was the same. F i g u r e 6 shows the hydrogen production rate -from P h S H as s a c r i f i c i a l electron donor in D H P vesicle-stabilized, rhodium-coated, colloidal CdS. Various catalyst concentrations were used, and the maximum efficiency was obtained with 8 χ 10° M R h , with 2 χ 10 M C d S generated on both sides of 2 χ 10~ M D H P vesicles. Several blanks demonstrated that the presence of a l l components of the system were necessary to observe hydrogen production. Hydrogen production could be sustained for about 15 h, until at least 90% of the PhSH had been c o n s u m e d Room temperature irradiation (2J>°C) was found to be optimal. The site of CdS generation (formed from C d n the inside or at the outside surface) was not very c r i t i c a l for hydrogen production. However, lower rates of hydrogen production with poorer reproducibility were usually obtained on i r r a d i a t i n g C d S particles generated from C d adsorbed only on the outside surfaces of D H P vesicles. A reproducibility better Jthan 10% obtained in systems prepared by e r a d i c a t i n g Çd Rh and D H P . A p e r f e c t l y homogeneous mixing of Cd and R h on the surfaces of the vesicles was easily obtained when those ions were cosonicated with the surfactant. A d d i t i o n o l E D T A to c a t i o n i c vesicles containing already (non-adsorbed) C d and R h resulted in the formation of R h - c o a t e d C d S exclusively on the outer surface. The hyorogen production efficiency in this system was the same as in systems prepared by cosonicating C d R h , E D T A , and the c a t i o n i c surfactants ( D O D A C or 1).


2 +

3

z +

w

+

f

a

0

s

+

+

+

Hydrogen production from P h S H in colloidal, rhodium-coated CdS located selectively in the inside of D H P vesicles, was reproducible and had about the same efficiency that symmetrically distributed C d S particles, at equivalent CdS concentration. It seemSj therefore, that homogeneously distributed C d S particles contributed to hydrogen production even if they d i d not exhibit a d e t e c t a b l e fluorescence. F i g u r e 7 illustrates the mechanism of photosensitized Η formation from P h S H in rhodium-coated colloidal CdS in nonpolymerized rationic vesicles. The proposed position of the C d S particle (partially buried in t h e v e s i c l e bilayer) is supported by the following observations, the last two made specifically in D H P vesicles: (a) CdS particles generated from externally adsorbed C d ions d i d not p r e c i p i t a t e , even after months; therefore, they had to remain bound to the v e s i c l e interface. (b) CdS fluorescence was efficiently quenched by P h S H , which was l o c a t e d in the hydrophobic membrane; therefore, the colloidal C d S particles had a direct contact with the inner part o f the membrane. (c) T h e C d S particle retained access to the surface where it originated, since entrapped polar electron acceptors such as MV while unable to penetrate the D H P membrane, c o u l d also quench the fluorescence of inside-generated CdS particles. 9

+

+

f


7.

TRICOT ET A L .

Photosensitized

Water

107

Reduction


.O

Sri iftS

F i g u r e 5. A n idealized mechanism of photoinduced electron transfer from C d S conduction band to methylviologen (MV ) resulting in formation of methylviologen radical cation ( M V ) , The colloidal C d S p a r t i c l e , as represented, was generated at t h e inside surface of the D H P vesicle. Its exact l o c a t i o n is based on fluorescence quenching ex periments (Figure 5). Insert: oscilloscope t r a c e showing the formation of M V by the absorbance change at 396 nm, after a laser pulse at 355 nm. z+

9

+

to Irradiation time, h

F i g u r e 6· Hydrogen production at 25°C in deaerated solutions as a function of catalyst (rhodium) c o n c e n t r a t i o n during the first two hours of irradiation using 350 nm c u t - o f f and water filters. Plotted are t h e amount o f hydrogen produced in 25 ml D H P v e s i c l e solution .and measured in the gas phase (16 ml) by G C : 2 χ 10" M D H P , 2 χ 10" M C d S symmetrically distributed on both sides of the vesicles, and 10" M P h S H as e l e c t r o n donor; pH approximately 7 a t so ni c a t i o n and during photolysis. Concentrations of the catalyst, reduced by uv irradiation prior to visible light photolysis: ( φ ) : R h 1.2 χ 10 Μ; ( Ο ) : R h 0.8 χ 10"* Μ: ( Π ) : RHfiA χ 10"* M ; ( • ) no R h ; ( Q ) no R h and no C d S (only PhSH); ( ζ ) ) : R h (unreduced) and no C d S (with PhSH). J +


108


However, this quenching decreased with time, showing a gradual penetration of the C d S towards the middle of the bilayer. (d) CdS particles generated at the vesicle interiors remained on their original side of the membrane, since externally added quencher such as M V and R h , while adsorbed on the outside surface of D H P vesicles, did not quench "inside" C d S fluorescence, even after several weeks, A different situation was encountered on using polymerized vesicles prepared from 1_ as hosts for the c a t a l y s t - c o a t e d colloidal semiconductors. P o l y m e r i z a t i o n of 1_ vesicles was shown to result in pulling together some 20 styrene headgroups"at the v e s i c l e surfaces, thereby c r e a t i n g clefts of some 15 A diameter (20). These relatively aqueous c l e f t s were proposed to be the site of C d S particles (Figure 8). The observed behavior of E D T A in the photolysis of R h - c o a t e d C d S particles embedded in 1^ vesicles were in a c c o r d with this interpretation. In nonpolymerized 1^ vesicles, E D T A was unable to act as a s a c r i f i c i a l electron donor; no photosensitized hydrogen formation was observed ( l i b ) . Conversely, in polymerized vesicles prepared from 1^ efficient photosensitized hydrogen formation was observed from E D T A (12). This effect is rationalized in terms of improved access of EDTA to the CdS surface in polymerized vesicles, due to surface irregularities (Figure 8). The observed increase in fluorescence intensity of CdS in polymerized vesicles as compared to that in their nonpolymerized counterparts also supports this r a t i o n a l i z a t i o a C o n t r o l l e d degrees of vesicle surface polymerization provide, therefore, a means for improving the efficiencies of electron donors. It should be pointed out, however, that E D T A is not an ideal e l e c t r o n donor since in high concentrations it destabilized the vesicles. In an attempt to overcome the problem of accumulation of the oxidized electron donor, we have incorporated a r e c y c l a b l e s u r f a c e - a c t i v e electron donor in D O D A C vesicles (12). This electron donor contains a sulfide moiety which dimerizes upon light-induced oxidation. Simultaneously, hydrogen is evolved via v e s i c l e - s t a b i l i z e d , c a t a l y s t - c o a t e d , colloidal C d S particles. The dimer could be chemically reduced for additional hydrogen formation. Figure 9 is an idealized view of this c y c l i c process (12).


2 +

Conclusion Our initial efforts to develop the use of surfactant aggregate-stabilized, c a t a l y s t - c o a t e d , colloidal semiconductors in a r t i f i c i a l photosynthesis are promising. In situ generation of the particles provides an unprecedented control over their morphologies. C o l l o i d a l semiconductors embedded in reversed micelles, vesicles or polymerized vesicles show remarkable long-term stabilities and diminish electron-hole recombination rates. These systems have the potential of decreasing photocorrosion and separating charges. Subsequent to optimization of the hydrogen generating systems, sacrificial nitrogen and carbon dioxide reductions can be a c c o m p l i s h e d Our ultimate goal is to combine oxidation and reduction half cells in a bilayer vesicle which would efficiently split water upon solar irradiation. F i g u r e 10 represents a highly i d e a l i z e d system. It would require the immobilization of CdS (or other suitable semiconductors) particles in polymerized vesicles in such a way that they span the bilayer. C o a t i n g the opposite sides of the vesicle-embedded semiconductor by catalysts would result in c y c l i c water splitting. The hydrogen and oxygen formed c o u l d then be harvested in compartments separated by the bilayer (Figure 10).



7.

TRICOT ET A L .

®ZZZZ1

Photosensitized

-

Water

109

Reduction

Dioctadecyldimethylammonium

cation

F i g u r e 7. A n idealized model for CdS sensitized photo reduction of water by P h S H in aqueous D O D A C or D O D A B vesicles. The exact position of the c o l l o i d represented here as generated on the outside surface of the vesicles, is based on fluorescence quenching experiments performed in anionic D H P vesicles, assuming similar i n t e r * actions of the C d S particles with both types of vesicles.

F i g u r e 8. A n idealized model for CdS sensitized photo reduction of water by E D T A in polymerized 1_ vesicles. The location of the p a r t i c l e is represented as similar tô that in unpolymerized vesicles, but should not be taken for granted in this case, however.


110



H

^ΞΞΝ-

a

Z^NC^SH

2

Dioctadecydimethylammonium =

Thiol

functionalized

cation

surfactant

Figure 9. A n i d e a l i z e d model for the c y c l i c oxidation-reduction process using t h i o l - f u n c t i o n a l i z e d surfactant, incorporated into D O D A C vesicles together with R h - c o a t e d colloidal C d S , for sustained hydrogen generation under visible light irradiation in one part of the c y c l e .

F i g u r e 10. A n hypothetical model for a c y c l i c water splitting system, based on a semiconductor particle immobilized in a polymerized mem brane and having access to both aqueous solutions on each side of the membrane. Specific and s e l e c t i v e c o a t i n g by catalysts leads to simul taneous and separate hydrogen and oxygen generation on e a c h side of the polymerized membrane.


7. tricot et al.

Photosensitized Water Reduction

111

Acknowledgment Support of this work by the Department of Energy is gratefully acknowledged. Literature Cited


1. 2. 3. 4.

Calvin, M. Acc. Chem. Res., 1978, 11, 369 - 374. Porter, G. Proc. R. Soc. London Ser Α., 1978, 362, 281 - 303. Fendler, J. H. J. Phys. Chem., 1980, 84, 1485 - 1491. Gratzel, M. "Energy Resources through Photochemistry and Catalysis", Academic Press, New York, 1983. 5. Fendler, J. H. "Membrane Mimetic Chemistry", Wiley-Interscience, New York, 1982. 6. Fendler, J. H. Chem. Eng. News, 1984, Jan. 2., 62, 25 - 38. 7. Duonhong, D.; Ramdsen, J.; Gratzel, M. J. Am. Chem. Soc., 1982, 104, 2977 - 2985. 8. Meissner, D.; Memming, R.; Kastening, B. Chem. Phys. Lett., 1983, 96, 34 - 37. 9. Krishnan, M.; White, J. R.; Fox, Μ. Α.; Bard, A. J. J. Am. Chem. Soc., 1983, 105, 7002 - 7003. 10. Meyer, M.; Wallberg, C.; Kurihara, K.; Fendler, J. H. J. Chem. Soc. Chem. Commun., 1984, 90 - 91. 11. (a) Tricot,Y.M.;Fendler, J. H. J. Am. Chem. Soc., 1984, in press; (b) Rafaeloff, R.; Tricot, Y. M.; Nome, F.; Fendler, J. H. J. Phys. Chem., 1984, in press. 12. Rafaeloff, R.; Tricot, Y. M.; Emeren, Å; Nome, F.; Fendler, J. H., unpublished results, 1984. 13. Rossetti, R.; Ellison, J. L.; Gibson, J. M.; Brus, L. E. J. Chem. Phys., 1984, 80, 4464 - 4469. 14. Brus, L. E. J. Chem. Phys., 1984, 80, 4403 - 4409. 15. Rossetti, R.; Nakahara, S.; Brus, L. E. J. Chem. Phys., 1983, 79, 1086 1088. 16. Henglein, A. Ber. Bunsenges. Phys. Chem., 1982, 86, 301 - 305. 17. Ramsden, J. J.; Gratzel, M. J. Chem.Soc.Faraday Trans. I, 1984, 80, 919 - 933. 18. Rosetti, R.; Brus, L. J. Phys. Chem., 1982, 86, 4470 - 4472. 19. Kuczyuski, J. P.; Milosavljeciv, B. H.; Thomas, J. Κ. J. Phys. Chem., 1983, 87, 3368 - 3370. 20. Nome, F.; Reed, W.; Politi, M.; Tundo, P.; Fendler, J. H. J. Am. Chem. Soc., 1984, 106, 0000 - 0000. RECEIVED January 10, 1985


Photosensitized Water Reduction Mediated by Semiconductors

Recommend Documents