remote interfacial electron transfer processes on...

Vol. 1 INTERNATIONAL JOURNAL OF PHOTOENERGY 1999

Remote interfacial electron transfer processeson nanocrystalline TiO2 sensitized

with polynuclear complexes

C. A. Bignozzi,1 M. Alebbi,1 E. Costa,1 C. J. Kleverlaan,1

R. Argazzi,1 and G. J. Meyer2

1Dipartimento di Chimica, Università di Ferrara, 44100 Ferrara Italy2Department of Chemistry, Johns Hopkins University, Baltimore MD 2128, USA

Abstract. The kinetic study of interfacial electron transfer in sensitized nanocrystalline semiconductor isessential to the design of molecular devices performing specific light induced functions in a microheteroge-neous environment. A series of molecular assemblies performing direct and remote charge injection to thesemiconductor have been discussed in the context of artificial photosynthesis. A particular attention in thisarticle has been paid to the factors that control the interfacial electron transfer processes in nanocrystallineTiO2 films sensitized with mononuclear and polynuclear transition metal complexes.

1. GENERAL REMARKS ON SENSITIZATION OFSEMICONDUCTORS

The general principles of dye sensitization of wide-band-gap semiconductors were already well estab-lished in the 1970s [1, 2, 3, 4] and advancements inthe application of such techniques to solar energy con-version have been initially very slow due to the poorlight absorption showed by monolayer of dyes on elec-trodes of small surface roughness. A substantial in-crease in light energy to electrical energy conversion ef-ficiencies, obtained with sensitized semiconductor elec-trodes, started with the development of high surfacearea nanocrystalline semiconductors [5] and of suitablemolecular sensitizers [5, 6, 7, 8, 9, 10, 11].

The accepted model for dye sensitization in regener-ative photoelectrochemical devices is shown in Figure 1

S∗/S+

S/S+

hνk1

e−

glass/SnO2 : F/TiO2 glass/SnO2 : F/Pt

e−

I−3

k7

k4 I−

k2

k5k3

CB

VB

Figure 1. Schematic representation of the kinetic processesoccurring in a regenerative photoelectrochemical cell.

[1, 2, 3, 4, 5, 8, 9, 10, 11]. The cell consists of a molec-ular sensitizer (S) anchored to the semiconductor sur-face, a solution containing a relay electrolyte (LiI, I2),and a counter electrode. Light excitation promotes thesensitizer to upper lying electronic excited-states thatconvert very rapidly and efficiently to the lowest-lying

electronic excited-state (S∗). The excited dye injects anelectron into the conduction band of the semiconduc-tor from a normal distribution of donor levels [10, 11]at a rate k2, and becomes oxidized. The oxidized dye(S+) is then reduced by an electron donor (I−) actingas relay electrolyte, at a rate k4. The electron flows (k5)through an external circuit to perform useful work. Re-duction of the oxidized donor (I−3 ) occurs at the counterelectrode (k7) and the solar cell is therefore regenera-tive. Radiative decay and nonradiative decay (k1) of theexcited dye molecule, and recombination of the pho-toinjected electron with the oxidized dye (k3), representloss mechanisms. Additional loss mechanisms, such asrecombination of the conduction band electrons withthe oxidized electron donor, are represented by the rateconstant k6 (not shown in Figure 1). Other loss mecha-nisms, as chemical reactions taking place from the ox-idized or excited dye sensitizer, are not shown in Fig-ure 1.

The meaningful parameters used to quantify on amacroscopic level the performance of a cell are the In-cident Photon to Current conversion Efficiency (IPCE),the open circuit photovoltage (Voc), and the overall ef-ficiency of the photovoltaic cell (ηcell).

The wavelength dependent IPCE term can be ex-pressed as a product of the quantum yield for chargeinjection (Φ), the efficiency of collecting electrons in theexternal circuit (η), and the fraction of radiant powerabsorbed by the material or «light harvesting efficien-cy» (LHE), equation (1):

IPCE(λ)= LHE(λ)φη. (1)

While Φ and η, can be rationalized on the basis of ki-netic parameters, LHE depends on the active surfacearea of the semiconductor and on the cross section forlight absorption of the molecular sensitizer. In practicethe IPCE measurements are performed with monochro-

2 C. A. Bignozzi et al. Vol. 1

matic light and calculated according to equation (2)

IPCE(λ)

= 12.4·103(V·nm)×photocurrent density(µA· cm−2)wavelength(nm)×photon flux(W·m−2)

.

(2)

The light harvesting efficiency can be related to thedye molar extinction coefficient (ε(λ), 1·mol−1· cm−1;λ = wavelength) and to the surface coverage (Γ ,mol·cm−2) by equation (3)

LHE(λ)= 1−10−[1000ε(λ)Γ]. (3)

An LHE factor of unity is ideal for a solar energy deviceas all the incident radiant power is collected. The devel-opment of high surface area nanocrystalline TiO2 hasproduced an enhancement of LHE [8].

In the absence of photodecomposition reactions, thequantum yield for electron injection from the excitedsensitizer to the semiconductor is given by equation (4)

Φ = k2/(k1+k2)= k2τ, (4)

where k1 is the radiative and nonradiative rate constantof the excited dye molecule (k1 = kr+knr), k2 is therate constant for the charge injection process and τrepresents the excited-state lifetime of the dye sensi-tizer anchored to the surface of the semiconductor.

Charge injection from the excited dye will be acti-vated if the donor energy is positive with respect to theconduction band-edge. Electron injection will be, on thecontrary, activationless if the donor level has an energyequal to, or more negative than, the conduction band-edge. For an activationless charge injection process, itis generally accepted that electron transfer from the ex-cited sensitizer to the conduction band is irreversibledue to extremely fast thermalization processes to theconduction band edge.

The fraction of injected charges, which percolatethrough the TiO2 membrane and reach the back con-tact of the photoanode, is represented by the factor η,equation 5

η= k4[I−]k3[e−]+k4[I−]

or η= 1− k3[e−]k3[e−]+k4[I−]

. (5)

The η term can be evaluated from equation 5 by timeresolved transient absorption measurements on TiO2

films, by monitoring the decay of the spectral featuresof the nascent oxidized sensitizer in absence (k3) orpresence (k4) of iodide [12].

The maximum open-circuit photovoltage, attainablein the dye sensitized solar cell, is the difference betweenthe Fermi level of the solid under illumination and theNernst potential of the redox couple in the electrolyte.However, for these devices this limitation has not beenrealized and Voc is in general much smaller. It appearsthat Voc is kinetically limited and for an n-type semi-conductor in a regenerative cell the diode equation (6)can be applied [13].

Voc =(kTe

)ln

(Iinj

n∑i ki[Ai]

), (6)

where Iinj is the electron injection flux, n is the concen-tration of electrons in TiO2, and the summation is forall electron transfer rates to acceptors Ai. One success-ful strategy for increasing Voc has been to add pyridinederivatives to the electrolyte [9]. Pyridine is thought toadsorb on the TiO2 surface and to inhibit recombina-tion of injected electrons with I−3 (k6). An alternative ap-proach involve the use of supramolecular systems con-taining a sensitizer unit covalently bound to a suitableelectron donor unit that allows to translate the positivecharge from the oxidized form of the sensitizer.

The overall efficiency of the photovoltaic cell, ηcell isgiven by equation (7)

ηcell = iph×Voc×ffIs

, (7)

where iph is the integrated photocurrent density, ffthe cell fill factor, and Is the intensity of the inci-dent light. The integrated photocurrent density repre-sents the overlap between the solar emission and themonochromatic current yield [14]. The maximum over-all efficiencies reported so far are in the 7–11% range,depending on the fill factor of the cells. Under optimalcurrent collection geometry, minimizing ohmic lossesdue to the resistance of the conductive glass, and underreduced solar irradiance, fill factors of 0.8 have beenobtained [15].

2. MONONUCLEAR MLCT SENSITIZERS

In order to be useful in a photoregenerative cell, themolecular sensitizer should fulfil several requirements,including:

a) The ability to adsorb firmly on the semiconductor.

b) Efficient light absorption in the visible region.

c) An excited-state redox potential negative enoughfor electron injection into the conduction band.

d) A ground-state redox potential as positive as pos-sible, compatible with b) and c).

e) Small reorganizational energy for excited- andground-state redox processes, so as to minimizefree energy losses in primary and secondary elec-tron transfer steps.

The most successful sensitizers, used so far, arebased on polypyridine complexes of d6 metal ions,such as Ru(II), Os(II), and Re(I). The absorption spec-tra of these complexes are dominated by intense metal-to-ligand charge transfer (MLCT) absorption bands inthe visible region. Excitation into these bands populateMLCT excited-states that are mainly singlet in charac-ter. This process is then followed by fast (k > 1012 s−1)and quantitative intersystem crossing to lower-lying,mainly triplet in character, MLCT excited-states [16].

Vol. 1 Remote interfacial electron transfer . . . 3

The energies of these states can be varied systemati-cally by changing the substituents at the chromophoricligands. Electron-withdrawing substituents tend to de-crease the energy of the π∗ orbitals of the polypyridineligand, while the opposite effect is observed with elec-tron donating substituents. The electronic effects arealso partially transmitted to the metal centres [17].

Energy tuning of MLCT states can also be obtainedby changing the non-chromophoric ligands. For bis-chelate complexes of the type cis-[RuII(bpy)2(X)2], aswell as for the analogous OsII complexes, MLCT absorp-tion bands shift to higher energy changing X fromπ do-nating ligands to π accepting ligands. The main effect,in this case, is a direct perturbation of the electronicdensity at the metal centre [17].

In order to obtain maximum conversion of visiblephotons into electrons, it is important that the absorp-tion spectrum of the dye sensitizer has significant over-lap with the spectrum of the solar radiation. Addition-ally, charge injection to the semiconductor and elec-tron transfer processes between the oxidized dye andthe relay must compete favourably with excited-staterelaxation processes and with chemical reactions tak-ing place from the excited dye and its ground-state ox-idized form. These requisites have all been met by us-ing ruthenium polypyridine complexes. The presenceof carboxylic acid functions on the bpy ligand allowsa stable linkage with TiO2 and promotes a strong elec-tronic coupling between the π∗ orbitals of the bipyri-dine ligand and the conduction band of the semicon-ductor, as evidenced by long term stability, in anhy-drous solvents, of the surface adsorbed dyes [18] andthe very high charge injection rates observed for dif-ferent MLCT sensitizers [19, 20, 21, 22, 23]. The highcoupling is thought to be mediated by π electrons inthe bridging group leading to increased delocalizationof the π∗ levels of the bipyridine ligand.

In the series of complexes of the type cis-[Ru(dcbH2)2(X)2] (dcbH2= 4,4′dicarboxy-2,2′ bipyri-dine; X = Cl−,Br−, I−,NCS−andCN−), MLCT absorp-tion and emission maxima were found to shift to-wards longer wavelength by decreasing the ligand fieldstrength of the ligand X, with E1/2 Ru(III)/(II) decreas-ing in the expected order, CN>NCS> halides. Thesecomplexes were generally found to act as efficient sen-sitizers for nanocrystalline TiO2. In particular, the per-formances of the NCS complex in photoelectrochemi-cal solar cells were found to be outstanding and un-matched by any other sensitizer [9, 11]. The complexshowed a photoaction spectrum dominating almost theentire visible region, with IPCE of the order of 90% be-tween 500–600 nm. Short-circuit photocurrents exceed-ing 17 mA/cm2 in simulated A.M. 1.5 sunlight and open-circuit photovoltages of the order of 0.7 V, were ob-tained by using iodide as redox electrolyte. For the firsttime a photoelectrochemical device was found to givean overall conversion efficiency of 10% [9].

In an attempt to extend the spectral sensitivity ofmononuclear dyes towards the red, new cis-[Ru(5,5′-

dcbH2)2(X)2] complexes based on the ligand 5,5′-dicarboxy-2,2′-bipyridine (X = Cl−,NCS−,CN−) weresynthesized [11]. The change of the carboxylic func-tion position from 4,4′ to 5,5′ causes a lowering ofthe energy of the π∗ accepting orbitals. A sustainedconversion of light to electricity was demonstrated for[Ru(5,5′-dcbH2) (NCS)2] achieving the goal of extend-ing the photoelectrochemical cell responsivity to longerwavelengths.

As a complementary strategy to polypyridyl sub-stitution, several attempts have been made to findsuitable ancillary ligands for complexes of thetype cis-[Ru(dcbH2)2X2] in order to have bettermononuclear ruthenium sensitizers. The complexescis-[Ru(dcbH2)2(isq)2] (isq= isoquinoline) [24], cis-[Ru(dcbH2)2(ppy)2] and cis-[Ru(dcbH2)2(H2O)(ppy)][25] (ppy = 4-phenylpyridine) have shown good photo-electrochemical properties, despite their photochemi-cal instability in solution, with IPCE values ranging from40% for the isoquinoline and bis(4-phenylpyridine)complexes to 50% for [Ru(dcbH2)2(H2O)(ppy)]. In an-other study, dithiocarbamate ligands [26] were con-sidered to be good candidates to extend the spec-tral sensitivity of mononuclear Ru sensitizers to-wards red. New violet dyes of general formula[Ru(dcbH)(dcbH2)(L)], where L is diethyldithiocarba-mate, dibenzyldithiocarbamate or pyrrolidinedithio-carbamate, have been synthesized, characterized andtested on TiO2 based photoelectrochemical regenera-tive cells using [Ru(dcbH2)2(NCS)2] as a reference com-pound. These dyes showed intense dπ−π∗ MLCT tran-sitions in the visible with the lower energy band slightlyred shifted with respect to the thiocyanate complex.Contrary to what observed for [Ru(dcbH2)2(NCS)2], thecyclic voltammery of the dithiocarbamate complexesgave reversible waves both in the protonated and an-ionic forms testifying a greater stability of the oxi-dized species towards ligand loss or substitution, asexpected from the chelating nature of these ancillaryligands. The photoelectrochemical behaviour of regen-erative cells with LiI 0.5 M/I2 0.05 M in acetonitrile gaveIPCE values for the violet dyes consistently lower thanthe reference complex, being, on the average, around50%. In order to rationalize this different behaviour,the kinetics of the primary processes of charge injec-tion (k2) and recombination (k3 and k4) was exploredby nanosecond transient absorption spectroscopy. Ab-sorption difference spectra recorded after 532 nm exci-tation of the complexes bound to TiO2 in argon satu-rated 1 M LiClO4/CH3CN solutions, showed mainly thebleaching and recovery of the MLCT absorption of thedyes. In all cases the formation of the oxidized sen-sitizer occured within the instrument response func-tion (k2 > 108 s−1) and the recombination rate (k3) wasfound to be approximately the same for the dithiocar-bamate complexes and the reference compound. Thekinetics of reduction of the oxidized sensitizer by io-dide ions (k4) was examined with the same techniqueat different I− concentrations. Stern-Volmer analysis


of the integrated area under the absorption transientsgave KSV = 100M−1 for [Ru(dcbH2)2(NCS)2] and KSV =30M−1 for the three dithiocarbamate sensitizers. Theslower reduction rate of the oxidized dithiocarbamatespecies by I− is thought to be consistent with the ob-served trend in IPCE values by virtue of a reducedcharge collection efficiency η equation (5).

Since one of the problems encountered with the car-boxy polypyridine class of sensitizers is the slow des-orption from the semiconductor surface in the pres-ence of water, the search for new anchoring group isadvisable. Along this line of research, we have recentlyprepared complexes based on the derivatization of 2,2′-bipyridine with a phenylboronic functionality [27]. Thephotoaction spectra of TiO2 electrodes sensitized withthe [Ru(4-phenylboronic-2,2′-bipyridine)2(CN)2] com-plex showed IPCE values comparable to those observedfor [Ru(dcbH2)2(CN)2] [10, 11], indicating that the newtype of linkage does not reduce the electron insectionyield.

Another different approach, recently explored [28],is based on the electropolymerization of molecularassemblies containing the ligand 4-methyl-4′-vinyl-2,2′-bipyridine (vbpy). Mononuclear as well as polynu-clear metal complexes such as [Ru(vbpy)3]2+ and[ ( bpy )2(CN)Ru(NC)Ru(vbpy)2 (NC)Ru(CN)(bpy)2]2+

form very stable and uniform films on TiO2 particlesupon electrochemical reduction of π∗(vbpy) followedby radical polymerization. Electropolymerized speciescan be advantageously used as overlayer structures toenhance the stability of previously adsorbed sensitiz-ers. Considering that the electronic coupling betweenthe polymerized sensitizer and the semiconductor isnot comparable to that of chemically adsorbed species,the IPCE values were found lower with respect to thosegiven by analogous complexes based on the dcb lig-and. The reduction of IPCE in overlayed structures isascribed also to excited state quenching due to trapsites formation in the polymer network.

3. SUPRAMOLECULAR SENSITIZERS

Studies on sensitization of nanocrystalline TiO2 withsupramolecular species may provide fundamental in-sights into interfacial electron transfer processes thatwould not be gained with simple molecular com-pounds. A supramolecular species possesses in gen-eral the following attributes: (i) the intrinsic proper-ties of the molecular components are not significantlyperturbed, and (ii) the properties of the supramolec-ular system are not simply the superposition of theproperties of the molecular components, but there is asupramolecular function. Upon substitution of one ofthe molecular components by a condensed phase, i.e. ananocrystalline semiconductor, a heterosupramolecu-lar system is formed. Two simple supramolecular dyadsystems, containing a chromophoric molecular compo-nent (photosensitizers, P), covalently linked acceptor(A) or donor (D) components on a semiconductor sur-face, are schematized in Figure 2.

Since one of the components is a condensed phase,this system can be considered as a “heterotriad”. In or-der to give the heterosupramolecular systems the func-tions as depicted in Figure 2, several nontrivial prob-lems must be solved. Apart from various important is-sues related to molecular architecture (choice of theappropriate molecular components so that each stepis thermodynamically allowed, assembling of the com-ponents via suitable connectors in the right sequence,binding of the supramolecular system to the semicon-ductor surface in the appropriate orientation, etc.), del-icate problems of kinetic nature must be addressed. Asa matter of fact, the kinetics of each of the electrontransfer steps must be optimized, so as to bring to 100%efficient charge separation. In particular, the key to theproblem is likely to be the competition between the sec-ondary electron transfer step (2 in Figure 2a) and theprimary charge recombination process (4 in Figure 2a).

a

b

3

5

4

VB

VB

CB

CB

TiO2

TiO2

12

12

A P

P D

3

4

Figure 2. Intramolecular and interfacial electron transferprocesses in dyads adsorbed on a semiconductor.

It is evident that a good deal of control on the fac-tors (driving force, reorganizational barriers, electronicfactors), that govern electron transfer rates, must bereached before a successful supramolecular device ofthis kind is developed. Both systems in Figure 2 are de-signed to translate the electron hole away from the sur-face and reduce the rate of electron-hole recombination(k3 in Figure 1), and thus increase the overall cell effi-ciency. In principle, an extension from dyads to largersystems can be envisioned keeping in mind that eachadditional charge separation step will reduce the driv-ing force which can be stored in the interfacial chargeseparated state.


4. CHROMOPHORE-ELECTRON DONOR SYSTEM

The vectorial translation of the electron-hole, awayfrom the surface, can be accomplished by using a sys-tem of the type sketched in Figure 2b. The dyad sys-tem, Ru-PTZ, which was prepared to model this func-tion, is shown in Figure 3. The chromophoric unit isrepresented by the Ru(II) moiety (with two dcbH andone dmb ligands), while the covalently bound phenoth-iazine (PTZ) acts as the electron donor [29, 30].

Irradiation with visible light in solution results inthe population of the Ru(III)(dcbH−) 3MLCT excited-state, which is followed by reductive quenching of theexcited-state by the PTZ moiety. The reductive quench-ing of the excited-state (Figure 4) is moderately exer-gonic (by ca. 0.25 eV) and has a rate constant of ca.2.5× 108 s−1 in methanol, as measured from the life-time of the residual ∗Ru(II) emission.

CB

VB

TiO2 O

O

O

O

O

O

OH

HOe−

e−

N

N

N

N

N

NN

S

hν

Ru

Figure 3. Structural formula of the Ru-PTZ chromophore-electron donor system and related kinetic processes onTiO2.

1∗Ru(II)-PTZ

Ru(II)-PTZ

3∗Ru(II)-PTZ

Ru(II)−-PTZ+

<1 ps

<1 ns

4 ns

200 ns

Figure 4. Excited state electron transfer processes in the Ru-PTZ chromophore-electron donor system.

The corresponding charge recombination step, presum-ably faster than the forward one, leads the system backto the ground-state without appreciable transient accu-mulation of the electron transfer product.

When the dyad system is attached to TiO2, excitationof the Ru chromophore can result in a new charge sep-arated state, TiO2(e−)-Ru-PTZ+. In principle there aretwo possible pathways available to reach this chargeseparation. In pathway 1 (Figure 5), charge injectionis followed by the charge shift to the donor, while in

TiO∗2 -Ru(II)-PTZ

TiO2-Ru(II)-PTZ

TiO−2 -Ru(II)-PTZ+

TiO−2 -Ru(III)-PTZ TiO2-Ru(II)−-PTZ+

1 2

Figure 5. Possible intramolecular and interfacial electrontransfer processes in the TiO2-Ru-PTZ system.

pathway 2, reductive quenching by the PTZ moiety isfollowed by charge injection.

A laser flash photolysis study of the heterotriadshowed that visible light excitation results in the Ru-based MLCT excited-state which rapidly injects an elec-tron into the semiconductor. The injection process isthen followed by electron transfer from PTZ to the ox-idized Ru centre (−∆G ca. 0.36 eV), resulting in thecharge separated state TiO2(e−)-Ru(II)-PTZ+. The re-combination of this state to the ground-state occurswith a rate of 3.6×103 s−1. Excitation of the model com-pound [Ru(dmb)(dcbH)2], gives rise to the formation ofthe charge separated state TiO2(e−)-Ru(III), whose re-combination kinetics is complex and can be analysedby a distribution model, with an average rate constantk3 = 3.9× 106 s−1. Translating the hole from the Rucentre to the pendant PTZ moiety produces a delay inrecombination rates (k3) by three order of magnitude[29, 30].

The dyad and model molecules were also tested in re-generative solar cells, with iodide as an electron donor.While the observed IPCE was of the order of 45% forboth systems, the open circuit photovoltage was ob-served to be higher for the dyad by 100 mV. The effectwas more pronounced in the absence of iodide with Voc

180 mV larger over 5 decades of irradiation. Applyingthe measured interfacial electron transfer rates to the


diode equation equation (6) gave the predicted increaseof Voc of 200 mV, which was in agreement with the ob-tained value (180 mV) [29, 30]. It is therefore encour-aging that an increase of the lifetime of the interfacialcharge separated state TiO2(e−)-Ru(II)-PTZ+ has a di-rect influence on the overall efficiency of the cell.

5. REMOTE CHARGE INJECTION PROCESSES

As discussed before, the binding to the semiconduc-tor surface is an important factor for solar cells, interms of stability and electronic coupling. Ligands withlinkage groups containing (CH2)n spacers were usedin molecular dyes [31]. Surprisingly, the reduced elec-tronic coupling between the excited sensitizer and theconduction band of TiO2 did not affect the performanceof the solar cell. This suggested that intimate linkagebetween the chromophoric unit and the semiconductorsurface is not a strict requirement for the design of amolecular sensitizer, and that remote charge injectionprocesses can be profitably used to achieve efficient dyesensitization.

A supramolecular approach for designing a molec-ular sensitizer, which can control the orientation ofthe component units on the semiconductor surface,is shown in Figure 6. The binuclear compound isbased on the -Re(I)(dcbH2)2 and -Ru(II)(bpy)2 moi-eties, undergoing ultra-fast and efficient photoin-duced Re∗→Ru energy transfer [32]. Due to the fa-cial geometry of the surface-bound Re-moiety, the-Ru(bpy)2 unit is forced to be close to the surface.

CB

VB

TiO2

e−

e−

O

O

O

O O

C

C

C

C

CRe

Ru

N

N

N

N

N

N

N

hν′

hν

O

O

N

Figure 6. Structural formula of the Re-Ru binuclear com-plex and related kinetic processes on TiO2.

Visible excitation of TiO2 photoanodes, loadedwith the binuclear sensitizer, resulted in an efficientphoton-to-current conversion. The photoaction spec-trum showed that the Ru-polypyridine moiety absorbsmost of the visible light with uncorrected maximumIPCE values of the order of 70%. The transient absorp-tion difference spectra for the sensitizers bound toTiO2 showed a broad bleach in the region from 400 to

600 nm, typical for the TiO2(e−)/Re(I)-Ru(III) state. Theformation of this state is promptly observed after thelaser pulse. This indicates that either remote electroninjection into TiO2, or intraligand (bpy•− dcbH2) elec-tron hopping from Ru(III) (bpy•−) to Re(I)(dcbH2), oc-curs within the laser pulse (k > 5×107). In conclusion,a rapid and efficient injection process is observed froma chromophoric group which is not directly coupled tothe semiconductor surface. These data indicate that di-rect covalent attachment is not strictly necessary fordesigning efficient molecular sensitizers.

The occurrence of remote interfacial electron trans-fer processes has been further confirmed by study-ing the photophysics of the binuclear complex[Ru(dcbH2)2(Cl)-BPA-Os(bpy)2Cl]2+ (BPA = 1,2-bis(4-pyridyl)ethane) on transparent TiO2 films [33]. Atmonolayer coverages the binuclear complex gives riseto 1/2 molecular adsorption with respect to the modelcomplex [Ru(dcbH2)2(Cl)(py)]+, indicating that the bin-uclear complex lies on the nanocrystals of TiO2 in amore or less extended conformation, as shown in Fig-ure 7.

CB

VB

TiO2

e−

e−O

O

O

O

O

O

OH

HO

Cl

Cl

Ru

Os

NN

NN

N

N

N

NN

N hν

hν′

Figure 7. Structural formula of the Ru-BPA-Os binuclearcomplex and related kinetic processes on TiO2.

Transient absorbance difference spectra following532 nm laser excitation, where both Ru and Os chro-mophores absorb, reveal the typical bleaching of thespin-forbidden MLCT transition localized on the Os(II)moiety. Spectral and kinetic analysis of the transientsignals are consistent with the formation of the chargeseparated state TiO2(e−)-Ru(II)-Os(III). This state caneither be formed through charge injection from theexcited Ru chromophore followed by intramolecularOs(II)→Ru(III) electron transfer, or via remote electrontransfer from the 3MLCT excited-state localized on theOs(II)(bpy)2 unit (Figure 8).

The occurence of the latter process is confirmed bytime resolved experiments in which selective laser exci-tation of the Os chromophore (at 683 nm) was obtained.

6. CONCLUSIONS

Some strategies for the design of mononuclear sensi-tizers and of artificial supramolecular systems, featur-ing functions such as photoinduced charge separation


and antenna effect for their use in sensitization of semi-conductors, have been discussed. These functions de-pend on the choice of specific molecular componentswhich may control the kinetics of the interfacial andintercomponent processes. On the basis of the knowl-edge gained in this field, molecular devices with pre-determined built-in functions can now be rationally de-signed and synthesized. Studies on model systems sug-gest the possibility to prevent interfacial charge recom-bination in sensitized semiconductor cells, by taking

--- ---

---

---

---

---

TiO2

TiO2

TiO2dcbRuII—OsIIbpy

dc−bRuIII—OsIIbpy

TiO−2 dcbRuIII—OsIIbpy

dcbRuII—OsIIIbpy

TiO2

dcbRuII—OsIIbpy

TiO−2

dcbRuII—OsIIIbp−y

hν hν

Figure 8. Possible intramolecular and interfacial electrontransfer processes in the TiO2-Ru-BPA-Os system.

advantage of photoinduced charge separation withinthe photosensitizer; they additionally indicate thatintramolecular energy transfer processes in polynu-clear complexes can be used to improve the over-all cross-section for light absorption of a sensitizer.

REFERENCES

[1] H. Gherisher, M. E. Michel-Beyerle, F. Rebentrost,and H. Tributsch, Electrochim. Acta 13 (1968),1509.

[2] H. Tributsch, Ber. Bunsen.-Ges. Phys. Chem. 73(1969), 582.

[3] R. Memming and H. Tributsch, J. Phys. Chem. 75(1971), 562.

[4] H. Gherisher, Photochem. Photobiol. 16 (1972),243.

[5] B. O’Regan and M. Grätzel, Nature 353 (1991), 737.

[6] R. Amadelli, R. Argazzi, C. A. Bignozzi, and F. Scan-dola, J. Am. Chem. Soc. 112 (1990), 7099.

[7] G. Smestad, C. A. Bignozzi, and R. Argazzi, Sol. En-ergy Mater. Sol. Cells 32 (1994), 259.

[8] M. K. Nazeeruddin, P. Liska, J. Moser, N. Vla-chopoulos, and M. Grätzel, Helv. Chim. Acta 73(1990), 1788.

[9] M. K. Nazeeruddin, A. Kay, I. Rodicio, R. Humphry-Baker, E. Muller, P. Liska, N. Vlachopoulos, andM. Grätzel, J. Am. Chem. Soc. 115 (1993), 6382.

[10] T. A. Heimer, C. A. Bignozzi, and G. J. Meyer, J.Phys. Chem. 97 (1993), 11987.

[11] R. Argazzi, C. A. Bignozzi, T. A. Heimer, F. N.Castellano, and G. J. Meyer, Inorg. Chem. 33 (1994),5741.

[12] A. C. Lees, B. Evrard, T. E. Keyes, J. G. Vos, C. J. Klev-erlaan, M. Alebbi, and C. A. Bignozzi, submitted.

[13] A. Kumar, P. G. Santangelo, and N. S. Lewis, J. Phys.Chem. B. 96 (1992), 834.

[14] A. Hagfeldt and M. Grätzel, Chem. Rev. 95 (1995),49.

[15] M. Grätzel, Semiconductor Nanoclusters, Studiesin Surface Science and Catalysis (P. V. Kamat andD. Meisel, eds.), vol. 103, Elsevier Science B.V.,1996, p. 353.

[16] J. N. Demas and D. G. Taylor, Inorg. Chem. 18(1979), 3177.

[17] C. A. Bignozzi, J. R. Schoonover, and F. Scandola,Molecular Level Artificial Photosynthetic Materials,Progr. Inorg. Chem. 44 (1997), 1.

[18] O. Kohle, M. Grätzel, A. F. Meyer, and T. B. Meyer,Adv. Mater. 9 (1997), 904.

[19] Y. Tachibana, J. E. Moser, M. Grätzel, D. R. Klug, andJ. R. Durrant, J. Phys. Chem. B 100 (1996), 20056.

[20] T. Hannappel, B. Burfeindt, Strock W., and F. Willig,J. Phys. Chem. B 101 (1997), 6799.

[21] J. E. Moser, D. Noukakis, U. Bach, Y. Tachibana,D. R. Klug, J. R. Durrant, R. Humphry-Baker, andM. Grätzel, J. Phys. Chem. B 102 (1998), 3649.

[22] T. Hannappel, C. Zimmermann, B. Meissner, B. Bur-feindt, W. Strock, and F. Willig, J. Phys. Chem. B 102(1998), 3651.

[23] F. Willig, R. Kietzmann, and K. Schwarzburg, Proc.SPIE Conf. on Energy Efficiency and Solar EnergyConversion XI (Washington), 18-22 May, Toulouse,France, SPIE Bellingham, 1992.

[24] C. G. Garcia, N. Y. Murakami Iha, R. Argazzi, andC. A. Bignozzi, J. Braz. Chem. Soc. 9 (1998), 13.

[25] C. G. Garcia, N. Y. Murakami Iha, R. Argazzi,and C. A. Bignozzi, J. Photochem. Photobiol. 115(1998), 239.

[26] R. Argazzi, C. A. Bignozzi, G. M. Hasselmann, andG. J. Meyer, Inorg. Chem. 37 (1998), 4533.

[27] R. Argazzi, S. Bui, C. A. Bignozzi, L. De Cola, andB. Schlicke, Inorg. Chim. Acta, submitted.

[28] J. A. Moss, J. M. Stipkala, J. C. Yang, C. A. Bignozzi,G. J. Meyer, T. J. Meyer, X. Wen, and R. W. Linton,Chem. Mater. 10 (1998), 1748.

[29] R. Argazzi, C. A. Bignozzi, T. A. Heimer, F. N.Castellano, and G. J. Meyer, J. Am. Chem. Soc. 117(1995), 11815.

[30] R. Argazzi, C. A. Bignozzi, T. A. Heimer, F. N.Castellano, and G. J. Meyer, J. Phys. Chem. B 101(1997), 2591.

[31] T. A. Heimer, S. T. D’Arcangelis, F. Fazard, J. M.Stipkala, and G. J. Meyer, Inorg. Chem. 35 (1996),5319.


[32] R. Argazzi, C. A. Bignozzi, T. A. Heimer, and G. J.Meyer, Inorg. Chem. 36 (1997), 2.

[33] M. Alebbi, C. Kleverlaan, C. A. Bignozzi, G. M. Has-selmann, and G. J. Meyer, Inorg. Chem., submitted.

Submit your manuscripts athttp://www.hindawi.com

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Inorganic ChemistryInternational Journal of

Hindawi Publishing Corporation http://www.hindawi.com Volume 2014

International Journal ofPhotoenergy


Carbohydrate Chemistry

International Journal of


Journal of

Chemistry


Advances in

Physical Chemistry

Hindawi Publishing Corporationhttp://www.hindawi.com

Analytical Methods in Chemistry

Journal of

Volume 2014

Bioinorganic Chemistry and ApplicationsHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

SpectroscopyInternational Journal of


The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014

Medicinal ChemistryInternational Journal of


Chromatography Research International


Applied ChemistryJournal of



Theoretical ChemistryJournal of


Journal of

Spectroscopy

Analytical ChemistryInternational Journal of


Journal of


Quantum Chemistry


Organic Chemistry International

ElectrochemistryInternational Journal of

Hindawi Publishing Corporation http://www.hindawi.com Volume 2014


CatalystsJournal of

remote interfacial electron transfer processes on...

Documents