We have characterized the concentration and time dependences of the attachment of colloidal CdSe quantum dots (QDs) to 16-mercaptohexadanoic acid (MHDA)-functionalized nanocrystalline TiO2 thin films. The amount of QDs and the extent of their agglomeration on MHDA-functionalized TiO2 films were characterized by transmission- and reflectance-mode UV/vis absorption spectroscopy and scanning electron microscopy. Optically transparent films with spatially homogeneous coloration and minimal agglomeration of QDs were prepared from 2.2 and 5.0 μM toluene dispersions of QDs at short reaction times (<5 h). In contrast, prolonged exposure of MHDA-functionalized TiO2 films to 22 μM dispersions of QDs yielded relatively opaque QD-functionalized films with spatially inhomogeneous coloration and substantial agglomeration of QDs. Agglomeration of QDs decreased the absorbed photon-to-current efficiencies of QD-sensitized solar cells (QDSSCs) by almost 3-fold. These results highlight the profound influence of agglomeration on the optical properties and interfacial electron-transfer reactivity of QD-functionalized TiO2 films prepared by in situ linker-assisted assembly as well as the photoelectrochemical performance of QDSSCs incorporating such films.

A library of six selenorhodamine dyes (4-Se–9-Se) were synthesized, characterized, and evaluated as photosensitizers of TiO2 in dye-sensitized solar cells (DSSCs). The dyes were constructed around either a bis(julolidyl)- or bis(half-julolidyl)-modified selenoxanthylium core functionalized at the 9-position with a thienyl group bearing a carboxylic, hydroxamic, or phosphonic acid for attachment to TiO2. Absorption bands of solvated dyes 4-Se–9-Se were red-shifted relative to the dimethylamino analogues. The dyes adsorbed to TiO2 as mixtures of monomeric and H-aggregated dyes, which exhibited broadened absorption spectra and increased light-harvesting efficiencies relative to the solvated monomeric dyes. Carboxylic acid-bearing dyes 4-Se and 7-Se initially exhibited the highest incident photon-to-current efficiencies (IPCEs) of 65–80% under monochromatic illumination, but the dyes desorbed rapidly from TiO2 into solutions of HCl (0.1 M) in a CH3CN:H2O mixed solvent (120:1 v:v). The hydroxamic acid- and phosphonic acid-bearing dyes 5-Se, 6-Se, 8-Se, and 9-Se exhibited lower IPCEs (49–65%) immediately after preparation of DSSCs; however, the dyes were vastly more inert on TiO2, and IPCEs decreased only minimally with successive measurements under constant illumination. Power-conversion efficiencies (PCEs) of the selenorhodamine-derived DSSCs were less than 1%, probably due to inefficient regeneration of the dyes following electron injection. For a given anchoring group, the bis(half-julolidyl) dyes exhibited higher open-circuit photovoltages and PCEs than the corresponding bis(julolidyl) dyes. The hydroxamic acid- and phosphonic acid-bearing dyes are intriguing photosensitizers of TiO2 in light of their aggregation-induced spectral broadening, high monochromatic IPCEs, and relative inertness to desorption into acidic media.

Colloidal semiconductor quantum dots (QDs) exhibit excitonic and surface states, both of which may participate in charge-transfer processes relevant to solar energy conversion. To explore this inherent complexity of the charge-transfer mechanisms of QDs, we used steady-state and time-resolved emission measurements to characterize excited-state electron transfer (ET) from core-only CdSe QDs and core/shell CdSe/ZnS QDs to TiO2 nanoparticles (NPs). Core-only QDs transferred electrons from both excitonic and surface states to TiO2 with rate constants of ET (ket) of approximately (1–3) × 108 s−1 and (4–7) × 107 s−1, respectively. Efficiencies of ET (ηet) from excitonic and surface states were approximately 71–82% and 64–76%, respectively. Thus, trapping of electrons lowered their potential energy but did not greatly affect the efficiency of their transfer to TiO2. Photogenerated holes were transferred from core-only CdSe QDs to adsorbed 3-mercaptopropionic acid (MPA), which linked the QDs to TiO2. We characterized core/shell CdSe/ZnS QDs as alternatives to core-only QDs. The ZnS shell eliminated the undesirable trapping of electrons and transfer of photogenerated holes to MPA. We measured ket of approximately (1–3) × 108 s−1 and ηet of approximately 66–85% for ET from excitonic states of core/shell CdSe/ZnS QDs to TiO2 NPs. The insensitivity of ket to the presence of the ZnS shell may have arisen from increased cross-linking of core/shell QDs to TiO2. Our results highlight the involvement of surface states in excited-state ET processes of core-only QDs and, for the heterostructures reported here, the improved performance of core/shell CdSe/ZnS QDs relative to core-only CdSe QDs.

For solar energy conversion, not only must a semiconductor absorb incident solar radiation efficiently but also its photoexcited electron—hole pairs must further be separated and transported across interfaces. Charge transfer across interfaces requires consideration of both thermodynamic driving forces as well as the competing kinetics of multiple possible transfer, cooling, and recombination pathways. In this work, we demonstrate a novel strategy for extracting holes from photoexcited CdSe quantum dots (QDs) based on interfacing with β-Pb0.33V2O5 nanowires that have strategically positioned midgap states derived from the intercalating Pb2+ ions. Unlike midgap states derived from defects or dopants, the states utilized here are derived from the intrinsic crystal structure and are thus homogeneously distributed across the material. CdSe/β-Pb0.33V2O5 heterostructures were assembled using two distinct methods: successive ionic layer adsorption and reaction (SILAR) and linker-assisted assembly (LAA). Transient absorption spectroscopy measurements indicate that, for both types of heterostructures, photoexcitation of CdSe QDs was followed by the transfer of electrons to the conduction band of β-Pb0.33V2O5 nanowires and holes to the midgap states of β-Pb0.33V2O5 nanowires. Holes were transferred on time scales less than 1 ps, whereas electrons were transferred more slowly on time scales of ∼2 ps. In contrast, for analogous heterostructures consisting of CdSe QDs interfaced with V2O5 nanowires (wherein midgap states are absent), only electron transfer was observed. Interestingly, electron transfer was readily achieved for CdSe QDs interfaced with V2O5 nanowires by the SILAR method; however, for interfaces incorporating molecular linkers, electron transfer was observed only upon excitation at energies substantially greater than the bandgap absorption threshold of CdSe. Transient absorbance decay traces reveal longer excited-state lifetimes (1–3 μs) for CdSe/β-Pb0.33V2O5 heterostructures relative to bare β-Pb0.33V2O5 nanowires (0.2 to 0.6 μs); the difference is attributed to surface passivation of intrinsic surface defects in β-Pb0.33V2O5 upon interfacing with CdSe.

Semiconductor heterostructures for solar energy conversion interface light-harvesting semiconductor nanoparticles with wide-band-gap semiconductors that serve as charge acceptors. In such heterostructures, the kinetics of charge separation depend on the thermodynamic driving force, which is dictated by energetic offsets across the interface. A recently developed promising platform interfaces semiconductor quantum dots (QDs) with ternary vanadium oxides that have characteristic midgap states situated between the valence and conduction bands. In this work, we have prepared CdS/β-Pb0.33V2O5 heterostructures by both linker-assisted assembly and surface precipitation and contrasted these materials with CdSe/β-Pb0.33V2O5 heterostructures prepared by the same methods. Increased valence-band (VB) edge onsets in X-ray photoelectron spectra for CdS/β-Pb0.33V2O5 heterostructures relative to CdSe/β-Pb0.33V2O5 heterostructures suggest a positive shift in the VB edge potential and, therefore, an increased driving force for the photoinduced transfer of holes to the midgap state of β-Pb0.33V2O5. This approach facilitates a ca. 0.40 eV decrease in the thermodynamic barrier for hole injection from the VB edge of QDs suggesting an important design parameter. Transient absorption spectroscopy experiments provide direct evidence of hole transfer from photoexcited CdS QDs to the midgap states of β-Pb0.33V2O5 NWs, along with electron transfer into the conduction band of the β-Pb0.33V2O5 NWs. Hole transfer is substantially faster and occurs at <1-ps time scales, whereas completion of electron transfer requires 5—30 ps depending on the nature of the interface. The differentiated time scales of electron and hole transfer, which are furthermore tunable as a function of the mode of attachment of QDs to NWs, provide a vital design tool for designing architectures for solar energy conversion. More generally, the approach developed here suggests that interfacing semiconductor QDs with transition-metal oxide NWs exhibiting intercalative midgap states yields a versatile platform wherein the thermodynamics and kinetics of charge transfer can be systematically modulated to improve the efficiency of charge separation across interfaces.

Achieving directional charge transfer across semiconductor interfaces requires careful consideration of relative band alignments. Here, we demonstrate a promising tunable platform for light harvesting and excited-state charge transfer based on interfacing β-PbxV2O5 nanowires with CdSe quantum dots. Two distinct routes are developed for assembling the heterostructures: linker-assisted assembly mediated by a bifunctional ligand and successive ionic layer adsorption and reaction (SILAR). In the former case, the thiol end of a molecular linker is found to bind to the quantum dot surfaces, whereas a protonated amine moiety interacts electrostatically with the negatively charged nanowire surfaces. In the alternative SILAR route, the surface coverage of CdSe nanostructures on the β-PbxV2O5 nanowires is tuned by varying the number of successive precipitation cycles. High-energy valence band X-ray photoelectron spectroscopy measurements indicate that “mid-gap” states of the β-PbxV2O5 nanowires derived from the stereoactive lone pairs on the intercalated lead cations are closely overlapped in energy with the valence band edges of CdSe quantum dots that are primarily Se 4p in origin. Both the midgap states and the valence-band levels are in principle tunable by variation of cation stoichiometry and particle size, respectively, providing a means to modulate the thermodynamic driving force for charge transfer. Steady-state and time-resolved photoluminescence measurements reveal dynamic quenching of the trap-state emission of CdSe quantum dots upon exposure to β-PbxV2O5 nanowires. This result is consistent with a mechanism involving the transfer of photogenerated holes from CdSe quantum dots to the midgap states of β-PbxV2O5 nanowires.

We synthesized quantum dot (QD) heterostructures via the N,N′-dicyclohexylcarbodiimide-mediated formation of amide bonds between capping ligands on CdS QDs and CdSe QDs. Products of ligand-exchange and coupling reactions were characterized by FTIR, 1H NMR, transmission electron micrscopy, and electronic absorption and emission spectroscopy. This cross-linking strategy yields exclusively heterostructures and prohibits the undesired formation of homostructures consisting of a single type of QD. The ground-state absorption spectra of the presynthesized colloidal QDs were unperturbed upon incorporation into heterostructures. Photoexcited CdS QDs transferred holes to molecularly tethered CdSe QDs, as evidenced by significant dynamic quenching of the trap-state emission from CdS QDs and the rapid (<10–8 s) growth of a broad and long-lived (>10–5 s) transient absorption band in the visible region. These spectral signatures were absent for mixed dispersions of noninteracting CdS and CdSe QDs. Our results reveal that carbodiimide coupling chemistry can be used to tether colloidal QDs selectively and covalently to each other and that the resulting heterostructures can undergo efficient photoinduced interfacial charge transfer.

Whereas a large number of sensitized polyoxotitanate clusters have been reported, information on the electrochemical properties of the fully structurally defined nanoparticles is not available. Bridging of this gap will allow a systematic analysis of the relation between sensitizer-cluster binding geometry, electronic structure and electron injection properties. Ti17O28(OiPr)16(FeIIPhen)2 is a member of a doubly-doped series of nanoclusters in which the phenanthroline is attached to the surface-located transition metal atom. The visible spectrum of a dichloromethane solution of the studied sample shows a series of absorption bands in the 400–900 nm region. Theoretical DOS and TDDFT calculations indicate that the bands in increasing wavelength order correspond essentially to metal-to-core charge transfer (MCCT) at ∼460 nm, metal-to-ligand charge transfer (MLCT) at ∼520 nm and d–d metal-atom transitions. Exposure of a thin layer of the sample to light in a photoelectrochemical cell produces an electric current in the 400 to ∼640 nm region. The fit of the wavelength range of the electron injection with the results of the calculations suggests that charge injection into the FTO anode occurs both from the TiO cluster and from the phenanthroline ligand. Injection from the phenanthroline via the cluster orbitals is ruled out by the lower energy of the phenanthroline orbitals.

We have characterized the concentration and time dependences of the attachment of colloidal CdSe quantum dots (QDs) to 16-mercaptohexadanoic acid (MHDA)-functionalized nanocrystalline TiO2 thin films. The amount of QDs and the extent of their agglomeration on MHDA-functionalized TiO2 films were characterized by transmission- and reflectance-mode UV/vis absorption spectroscopy and scanning electron microscopy. Optically transparent films with spatially homogeneous coloration and minimal agglomeration of QDs were prepared from 2.2 and 5.0 μM toluene dispersions of QDs at short reaction times (<5 h). In contrast, prolonged exposure of MHDA-functionalized TiO2 films to 22 μM dispersions of QDs yielded relatively opaque QD-functionalized films with spatially inhomogeneous coloration and substantial agglomeration of QDs. Agglomeration of QDs decreased the absorbed photon-to-current efficiencies of QD-sensitized solar cells (QDSSCs) by almost 3-fold. These results highlight the profound influence of agglomeration on the optical properties and interfacial electron-transfer reactivity of QD-functionalized TiO2 films prepared by in situ linker-assisted assembly as well as the photoelectrochemical performance of QDSSCs incorporating such films.

We have synthesized water-dispersible cysteinate(2–)-capped CdSe nanocrystals and attached them to TiO2 using one-step linker-assisted assembly. Room-temperature syntheses yielded CdSe magic-sized clusters (MSCs) exhibiting a narrow and intense first excitonic absorption band centered at 422 nm. Syntheses at 80 °C yielded regular CdSe quantum dots (RQDs) with broader and red-shifted first excitonic absorption bands. Cysteinate(2–)-capped CdSe MSCs and RQDs adsorbed to bare nanocrystalline TiO2 films from aqueous dispersions. CdSe-functionalized TiO2 films were incorporated into working electrodes of quantum dot-sensitized solar cells (QDSSCs). Short-circuit photocurrent action spectra of QDSSCs corresponded closely to absorptance spectra of CdSe-functionalized TiO2 films. Power-conversion efficiencies were (0.43 ± 0.04)% for MSC-functionalized TiO2 and (0.83 ± 0.11)% for RQD-functionalized TiO2. Absorbed photon-to-current efficiencies under white-light illumination were approximately 0.3 for both MSC- and RQD-based QDSSCs, despite the significant differences in the electronic properties of MSCs and RQDs. Cysteinate(2–) is an attractive capping group and ligand, as it engenders water-dispersibility of CdSe nanocrystals with a range of photophysical properties, enables facile all-aqueous linker-assisted attachment of nanocrystals to TiO2, and promotes efficient interfacial charge transfer.Keywords: cadmium selenide; cysteine; electron injection; linker-assisted assembly; magic-sized clusters; quantum dot-sensitized solar cell;

Chalcogenopyrylium monomethine dyes were prepared via condensation of a 4-methylchalcogenopyrylium compound with a chalcogenopyran-4-one bearing a 4-(diethoxyphosphoryl)phenyl substituent (or the phosphonic acid derivative). The dyes have absorbance maxima of 603–697 nm in the window where the solar spectrum is most intense. The dyes formed H-aggregates on TiO2, increasing the light-harvesting efficiency of the dyes. Shortcircuit photocurrent action spectra were acquired to evaluate the influence of dye structure on the photoelectrochemical performance.

•We have used transient absorption spectroscopy to measure relative electron-injection yields from selenorhodamine dyes to titanium dioxide.•Electrons were injected approximately twice as efficiently from dyes anchored to TiO2 via carboxylate linkages than via phosphonate linkages.•For a given anchoring mode, electrons were injected approximately twice as efficiently from H-aggregated dyes than from non-aggregated dyes.We used transient absorption spectroscopy to characterize excited-state electron injection from a 2,7-bis(dimethylamino)-9-(5-phosphonothien-2-yl)selenoxanthylium dye (3-Se) into TiO2. Dye 3-Se adsorbed to TiO2 via the phosphonic acid group as a mixture of H-aggregates and monomers. Injection of electrons from photoexcited 3-Se into TiO2 yielded the dication radical (3-Se+) and an associated transient absorption at wavelengths shorter than 540 nm, the amplitude of which was proportional to the quantum yield of electron injection (ϕinj). Our data revealed that ϕinj from H-aggregated 3-Se was (2.0 ± 1.3)-fold greater than from monomeric 3-Se; therefore, H-aggregation increased the efficiencies of both light-harvesting and electron injection. Comparison with our reported data for the analogous carboxylic acid-functionalized dye (1-Se) revealed that ϕinj via the carboxylate linkage was (2.3 ± 1.1)-fold greater than via the phosphonate linkage. Thus, electron-injection reactivity is sensitive to both the aggregation state and the surface-anchoring mode of these chalcogenorhodamine dyes. The decrease of ϕinj for 3-Se is offset by its enhanced stability and persistence on TiO2, rendering the phosphonic acid-functionalized and H-aggregated dye a particularly attractive sensitizer.

Lateral dispersion forces induce the ordering of n-alkanoic acids on nanocrystalline TiO2 films and cause the compositions of mixed monolayers to change. The equilibrium formation of single-component monolayers of n-alkanoic acids and 6-bromohexanoic acid (Br6A) on TiO2 was well-modeled by the Langmuir adsorption isotherm. Surface adduct formation constants were 103–104 M–1, and saturation amounts of adsorbates per projected surface area of TiO2 were on the order of 10–7 mol cm–2. The adsorption of n-heneicosanoic acid (21A) followed Langmuir kinetics, whereas the net rates of adsorption of shorter n-alkanoic acids and Br6A were slower than predicted by simple Langmuir kinetics, suggesting that desorption was non-negligible. At high surface coverages, n-alkanoic acids with 14 or more methylene groups formed ordered, crystalline monolayers, as evidenced by shifts of asymmetric and symmetric CH2 stretching bands in IR spectra. The extent of ordering was similar to that of self-assembled monolayers of alkanethiols on gold. The formation of ordered monolayers was well-modeled by an idealized mechanism, in which adsorption and desorption followed Langmuir kinetics and ordering was first-order with respect to the fractional surface coverage of adsorbates. Dispersion forces and ordering affected the compositions of mixed monolayers of 21A and Br6A on TiO2 films that remained in contact with mixed coadsorption solutions. When the fractional surface coverage of 21A was sufficiently high to induce ordering, it displaced Br6A from TiO2. We propose that these compositional changes were driven by the stabilization of 21A via cohesive lateral dispersion forces. Our results reveal that mixed monolayers on nanocrystalline TiO2 films are dynamic and that noncovalent intermolecular interactions can profoundly influence their compositions and properties.

We have characterized the persistence and degradation of magic-sized CdSe nanocrystals (NCs) after their removal from the original reaction mixture and dispersion into basic aqueous solutions. Such studies are important given the myriad potential applications of semiconductor NCs and ongoing efforts to characterize the properties and reactivity of monodisperse suspensions of intact NCs. Correlated challenges are to elucidate the mechanisms by which NCs degrade and to establish conditions under which NCs persist. Our CdSe NCs degraded after dilution into aqueous NaOH, resulting in red-shifted excitonic absorption bands and eventual flocculation. Dilution of NCs into basic aqueous solutions of cysteinate resulted in degradation via a different mechanism with an absence of flocculation; kinetics varied with concentration of cysteinate. The chemical fate of NCs after dilution into basic aqueous solutions containing both Cd2+ and cysteinate varied with the cysteinate-to-Cd2+ molar ratio, which determined the relative solute mole fractions of various Cd2+–cysteinate complexes. CdSe NCs persisted on long time scales only when dispersed in solutions containing [Cd(cysteinate)3]4–. We present equilibria to account for the observed spectral changes after dilution of CdSe into various basic media. Cadmium(II)–cysteinate complex-formation equilibria influenced the temporal persistence of the NCs; the pathway through which CdSe NCs degraded depended on the concentration of free, uncoordinated cysteinate. Our findings indicate that solution-phase chemistry can determine whether NCs remain intact upon removal from their original reaction mixtures.Keywords: cadmium selenide; cysteine; magic-sized clusters; persistence and degradation; quantum dots; semiconductor nanocrystals;

CdSe nanoparticles (NPs) capped with cysteinate (Cys), 3-mercaptopropionate (MP), and mercaptosuccinate (MS) were adsorbed to TiO2 from basic aqueous dispersions. Native capping groups served as molecular linkers to TiO2. Thus, the materials-assembly chemistry was simplified and made more reproducible and environmentally benign. The electronic properties of CdSe and the electron-transfer reactivity at CdSe-linker-TiO2 interfaces varied with the structure and functionality of the capping groups. Cys-capped CdSe NPs exhibited a narrow and intense first excitonic absorption band centered at 422 nm, suggesting that they were magic-sized nanocrystals (MSCs) with diameters less than 2 nm. MP- and MS-capped CdSe NPs had broader and lower-energy absorption bands, which are typical of regular quantum dots. Photocurrent action spectra of nanocrystalline TiO2 films functionalized with Cys-CdSe, MP-CdSe, and MS-CdSe overlaid closely with absorption spectra, indicating that excitation of CdSe gave rise to the injection of electrons into TiO2. Under white-light illumination, the global energy-conversion efficiency for Cys-capped CdSe ((0.45 ± 0.11)%) was 1.2-to-6-fold greater than for MP- and MS-capped CdSe. Similarly, the absorbed photon-to-current efficiency was 1.3-to-3.3-fold greater. These differences arose from linker-dependent variations of electron-injection and charge-recombination reactivity. Transient absorption measurements indicated that electron injection from Cys-capped CdSe was more efficient than from MS-capped CdSe. In addition, charge recombination at CdSe-MS-TiO2 interfaces was complete within hundreds of nanoseconds, whereas the charge-separated-state lifetime at CdSe-Cys-TiO2 interfaces was on the order of several microseconds. Thus, Cys-capped CdSe MSCs are readily attached to TiO2 and exhibit unusual electronic properties and desirable electron-transfer reactivity.Keywords: charge recombination; electron injection; interfacial electron transfer; linker-assisted assembly; quantum dot solar cell;

We used transient absorption spectroscopy to characterize excited-state electron injection from a 2,7-bis(dimethylamino)-9-(2-thienyl-5-carboxy)selenoxanthylium dye (1-Se) and a 2,7-bis(dimethylamino)-9-(3-thienyl-2-carboxy)selenoxanthylium dye (2-Se) to TiO2. Monolayers of 1-Se on nanocrystalline TiO2 films consisted of both H-aggregated and nonaggregated dyes, whereas 2-Se underwent little or no aggregation upon adsorption. Two transient spectral signals were correlated with the dication radicals (1-Se+ and 2-Se+) generated by electron injection: an absorption at wavelengths shorter than 540 nm and a bleach from approximately 540−650 nm. Relative quantum yields for electron injection (ϕinj) were calculated from the measured amplitudes of these signals. The value of ϕinj for H-aggregated 1-Se was approximately 2-fold greater than ϕinj for nonaggregated 1-Se and approximately 3-fold greater than ϕinj for nonaggregated 2-Se. Thus, H-aggregation can increase both the light-harvesting efficiencies and the electron-injection yields of rhodamine derivatives. Our findings suggest that controlled aggregation of organic dyes may represent an attractive strategy for improving the global energy-conversion efficiencies of organic dye-sensitized solar cells and photocatalysts.

Patterned mixed monolayers of porphyrins on nanocrystalline TiO2 films were fabricated by substrate-catalyzed monolayer photolithography. Tin(IV) protoporphyrin IX (SnPP), zinc(II) protoporphyrin IX (ZnPP), and iron(III) meso-tetra(4-carboxyphenyl)porphine (FeTCP) were adsorbed to TiO2 through the carboxyl groups, yielding saturation surface amounts per projected area of approximately 10−7 mol/cm2. Illumination of SnPP- and ZnPP-functionalized TiO2 films with 355 nm light caused the desorption of the porphyrins, most likely through oxidative decarboxylation. SnPP was removed more rapidly than ZnPP. The faster kinetics was due, in part, to the contribution of other photochemical pathways including TiO2-catalyzed photoreduction and direct photodegradation reactions. Patterned binary monolayers were prepared by the photoinduced desorption of a protoporphyrin, followed by the adsorption of FeTCP to previously illuminated regions of the surface.

We have characterized electron injection from photoexcited CdS quantum dots (QDs) to TiO2 nanoparticles as a function of the interparticle separation within molecularly linked assemblies. CdS QDs were tethered to TiO2 nanoparticles through bifunctional mercaptoalkanoic acids (MAAs). Electron injection and interfacial charge recombination were characterized by steady-state emission quenching, nanosecond time-resolved emission, and nanosecond transient absorption. The electron injection yield decreased with increasing MAA chain length and interparticle separation. Electron injection occurred on multiple timescales. A fast component (<10−8 s) accounted for the majority of injection, while the remainder occurred on the microsecond time scale. We attribute the multiexponential injection kinetics to electron transfer from a range of conduction-band and trap states. Interfacial charge recombination occurred on the microsecond time scale, and the kinetics were independent of the MAA chain length. Our findings reveal that the excited-state deactivation pathways and interfacial electron-transfer reactivity of tethered assemblies of nanoparticles can be tuned systematically by varying the interparticle separation.

Mixed monolayers of octanoic acid (OA) and 16-mercaptohexadecanoic acid (MHDA) were adsorbed to nanocrystalline TiO2 films from mixed solutions in tetrahydrofuran. For a range of solution compositions, the mole fraction of MHDA within the mixed monolayers (χMHDA,surf) exceeded that of the coadsorption solution. In addition, χMHDA,surf increased with time, while the sum of the surface coverages of MHDA and OA remained constant. To account for these effects, we propose a mechanism involving disulfide formation between the terminal thiol groups of surface-adsorbed MHDA molecules. Disulfide formation leads to an increase in the surface adduct formation constant (Kad) of dimeric MHDA, causing the gradual displacement of OA from the surface. The mechanism is supported by spectroscopic evidence and desorption kinetics. These are the first examples of mixed monolayers that undergo time-dependent compositional changes as a result of covalent bond formation between surfactants. Our findings illustrate that dimerization and other intermolecular interactions between surfactants may dramatically influence the composition and terminal functionalization of a wide range of mixed monolayer systems.

Colloidal semiconductor quantum dots (QDs) exhibit excitonic and surface states, both of which may participate in charge-transfer processes relevant to solar energy conversion. To explore this inherent complexity of the charge-transfer mechanisms of QDs, we used steady-state and time-resolved emission measurements to characterize excited-state electron transfer (ET) from core-only CdSe QDs and core/shell CdSe/ZnS QDs to TiO2 nanoparticles (NPs). Core-only QDs transferred electrons from both excitonic and surface states to TiO2 with rate constants of ET (ket) of approximately (1–3) × 108 s−1 and (4–7) × 107 s−1, respectively. Efficiencies of ET (ηet) from excitonic and surface states were approximately 71–82% and 64–76%, respectively. Thus, trapping of electrons lowered their potential energy but did not greatly affect the efficiency of their transfer to TiO2. Photogenerated holes were transferred from core-only CdSe QDs to adsorbed 3-mercaptopropionic acid (MPA), which linked the QDs to TiO2. We characterized core/shell CdSe/ZnS QDs as alternatives to core-only QDs. The ZnS shell eliminated the undesirable trapping of electrons and transfer of photogenerated holes to MPA. We measured ket of approximately (1–3) × 108 s−1 and ηet of approximately 66–85% for ET from excitonic states of core/shell CdSe/ZnS QDs to TiO2 NPs. The insensitivity of ket to the presence of the ZnS shell may have arisen from increased cross-linking of core/shell QDs to TiO2. Our results highlight the involvement of surface states in excited-state ET processes of core-only QDs and, for the heterostructures reported here, the improved performance of core/shell CdSe/ZnS QDs relative to core-only CdSe QDs.

Whereas a large number of sensitized polyoxotitanate clusters have been reported, information on the electrochemical properties of the fully structurally defined nanoparticles is not available. Bridging of this gap will allow a systematic analysis of the relation between sensitizer-cluster binding geometry, electronic structure and electron injection properties. Ti17O28(OiPr)16(FeIIPhen)2 is a member of a doubly-doped series of nanoclusters in which the phenanthroline is attached to the surface-located transition metal atom. The visible spectrum of a dichloromethane solution of the studied sample shows a series of absorption bands in the 400–900 nm region. Theoretical DOS and TDDFT calculations indicate that the bands in increasing wavelength order correspond essentially to metal-to-core charge transfer (MCCT) at ∼460 nm, metal-to-ligand charge transfer (MLCT) at ∼520 nm and d–d metal-atom transitions. Exposure of a thin layer of the sample to light in a photoelectrochemical cell produces an electric current in the 400 to ∼640 nm region. The fit of the wavelength range of the electron injection with the results of the calculations suggests that charge injection into the FTO anode occurs both from the TiO cluster and from the phenanthroline ligand. Injection from the phenanthroline via the cluster orbitals is ruled out by the lower energy of the phenanthroline orbitals.