In Situ Photochemical Surface Passivation of CdSe ... - ACS Publications

0 downloads 0 Views 4MB Size Report
Dec 16, 2013 - Dots for Quantitative Light Emission and Enhanced Photocurrent ... of defects in QD by DTT results in the suppression of energy-wasting.
Article pubs.acs.org/JPCC

In Situ Photochemical Surface Passivation of CdSe/ZnS Quantum Dots for Quantitative Light Emission and Enhanced Photocurrent Response in Solar Cells Morihiko Hamada,†,‡ Norifumi Takenokoshi,†,‡ Keiji Matozaki,‡ Qi Feng,‡ Norio Murase,† Shin-ichi Wakida,† Shunsuke Nakanishi,‡ and Vasudevanpillai Biju*,†,§ †

Health Research Institute, National Institute of Advanced Industrial Science and Technology (AIST), Takamatsu, Kagawa, Japan Department of Advanced Materials Science, Kagawa University, Takamatsu, Kagawa, Japan § Precursory Research for Embryonic Science and Technology (PRESTO), Japan Science and Technology Agency (JST), Tokyo, Japan ‡

ABSTRACT: Suppression of energy-wasting nonradiative carrier relaxation in photoactivated semiconductor quantum dots (QDs) is an essential requirement for applications such as QD-sensitized solar cell (QDSSC). Here we report that in situ surface passivation of CdSe/ZnS QDs using dithiothreitol (DTT) results in the 4-fold enhancement of photoluminescence (PL) quantum efficiency of QDs. By the analyses of PL intensities and blinking statistics of single QDs, the PL enhancement is assigned to a combination of processes such as unexpected appearance of bright QDs, gradual increase in the PL intensity of QDs, and substantial suppression of blinking. These processes not only suggest the suppression of nonradiative carrier recombination in QDs passivated with DTT molecules but also enable us to classify the nonradiative carrier recombination centers in QDs into low-density defects in which the rates of nonradiative recombination largely exceed the rates of radiative relaxation and high-density defects in which the rates of nonradiative carrier recombination are comparable to the rates of radiative relaxations. In the former case, passivation of the defects by the spontaneous addition of one or a few DTT molecules to QDs with PL intensity below the detection limit results in the formation of highly luminescent QD-(DTT)n adducts. In the latter case, highly luminescent QD-(DTT)n adducts are formed from already luminescent QDs by the relatively slow addition of DTT molecules. These two types of surface passivation reactions are governed by both chemical and photochemical processes. Overall, the passivation of defects in QD by DTT results in the suppression of energy-wasting nonradiative relaxations, which enables us to drive ca. 33% excess photocurrent in a photoelectrochemical cell constructed using QD-(DTT)n as the sensitizer.



INTRODUCTION Colloidal semiconductor quantum dots (QDs), owing to their broad absorption and narrow emission bands, high photostability, and ability to activate multiple excitons,1−4 have become promising materials for light emitting diodes,5,6 solar energy harvesting devices,7−9 and bioimaging.10,11 Core-only cadmium chalcogenide nanocrystals synthesized in the early days by the nanocrystal nucleation−growth method12 show low photoluminescence (PL) quantum yields (QY) and poor stability against photoactivation and various chemical environments because of the enhanced nonradiative carrier relaxation in defects and surface etching. Nonradiative relaxations occur due to lattice strain, electron and hole vacancies, and danglingbonds on the surface.13−15 By the subsequent protection of the QD core with shell from higher band gap materials such as ZnS as reported by Hines and Guyot-Sionnest16 and Bawendi and co-workers17 resulted in the improvement of not only the PL QY but also the physical and chemical stabilities. Nevertheless, the strain at the core−shell interface, and defects in the core and shell remained. Such defects act as centers of nonradiative © 2013 American Chemical Society

carrier recombination, which not only lower the PL QY of QDs but also induce stochastic fluctuations in the PL intensity of single QDs, a phenomenon called blinking. Blinking is also contributed to a greater extent by the nonradiative Auger recombination in ionized QDs.18,19 Because of the growing interests in the applications of QDs in bioimaging and device technology, relations among surface defects, PL QY and blinking have been subjects of active investigation in the recent past. Although there have been many reports about blinking suppression in single QDs by either surface passivation17,20,21 or inducing soft quantum confinement,22 a complete solution to the suppression of nonradiative carrier recombination and blinking is yet to be identified. Different methods applied for PL enhancement in QDs are surface passivation using thiols1,23−25 or amines,12,26,27 surface modifications with large band gap materials,13,16,20,21,28−30 Received: August 21, 2013 Revised: December 9, 2013 Published: December 16, 2013 2178

dx.doi.org/10.1021/jp4083882 | J. Phys. Chem. C 2014, 118, 2178−2186

The Journal of Physical Chemistry C



metals nanoparticles and surfaces,31−33 and photoactivation.34−39 Although comprehensive discussion about PL enhancement in QDs is difficult because of the sample-bysample variations in the surface properties and PL QY of QDs, multiple factors such as surface chemical and photochemical etching, surface oxidation, thermal annealing of point defects, chemical reconstruction of surface bonds, physicochemical changes induced by electron and energy transfer, pH, solvent polarity, and the presence of water play crucial roles on the PL enhancement factor.40 The role of thiols on surface passivation of QDs is extensively investigated at ensemble and single molecule levels.23,25,41−44 For example, Hohng and Ha have reported the blinking suppression of CdSe/ZnS QDs supplemented with β-mercaptoethanol.23 Similarly, Weiss and co-workers have reported the blinking suppression for QDs embedded in dithiothreitol (DTT) enfolded polyvinyl alcohol (PVA) film.25 Park and co-workers have seen PL enhancement of silica-coated QDs supplemented with DTT and activated with light.43 In these cases, the blinking suppression and PL enhancement are assigned to surface passivation of QDs via electron transfer from thiols. Similar surface passivation effects are observed for core or core/shell QDs treated with various amines.15,45,46 By the systematic analyses of the effects of butylamine, dibutylamine, and tributylamine on the PL properties of CdSe QDs, Bullen and Mulvaney have reported that the enhancement of PL intensity as a result of surface passivation using primary amines is greater than that using tertiary amines.26 In contrast with these reports about QD PL enhancement by surface passivation using thiols22−25 and amines,12,26,27 quenching of PL by amines is observed in many studies.47−50 Protection of the QD core using large band gap semiconductors is the most efficient method for the passivation of surface defects and enhancement and stabilization of PL.13 Yet another method for improvement of QD PL is photoactivation, during which surface etching and surface passivation by solvent or ligand molecules suppress the density of nonradiative carrier recombination centers.15,35−39 Some of the methods mentioned above for PL enhancement also induce blinking suppression. Ideal agents for blinking suppression are thiols,23,25 amines,21 fullerene (C60),28−30 and silver,2,31 gold32,33,51, and TiO252,53 nanoparticles, which in certain cases aggravate blinking because of efficient photoinduced electron transfer.28,54−56 Here we report that chemical and photochemical surface passivations of CdSe/ZnS QDs using DTT as an electron donor result in the steady increase in the PL intensity of QDs at ensemble and single molecule levels, which is due to the formation of QD-(DTT)n adducts. Such surface passivation allows us to not only obtain quantitative light emission from QDs at ensemble level but also suppress the undesired blinking of single QDs. Also, by the analyses of time-correlated single molecule PL intensity trajectories and images of QDs treated with or without DTT, we successfully categorize the rate of nonradiative carrier recombination in photoactivated QDs. Further, we construct photoelectrochemical solar cells sensitized with pristine QDs or QD-(DTT)n adducts and evaluate the photocurrent response. Here, as a result of the chemical and photochemical passivation of energy-wasting nonradiative carrier recombination centers in QDs using DTT, we extract ca. 33% excess photocurrent in the QD-(DTT)nsensitized solar cell when compared with QDSSC.

Article

MATERIALS AND METHODS CdSe/ZnS QDs as seen in Figure 1A,B with PL λmax ca. 655 nm and DTT are obtained from Invitrogen Corporation and

Figure 1. (A,B) TEM images of CdSe/ZnS QDs. The scale bars are (A) 25 nm and (B) 5 nm. (C,D) PL spectra of 5 nM QD solutions in o-xylene and in the presence or absence of DTT. (C) bottom to top are PL spectra recorded with increase in the concentration of DTT from 0 μM to 160 μM at 16 μM intervals and under continuous photoactivation (10 μW/cm2, 607 nm). (D) bottom to top are spectra recorded at 1 min intervals with continuous photoactivation (10 μW/ cm2, 607 nm) in the presence of 77 μM DTT. Insets: PL peak intensities of QD solutions as a function of DTT concentration (Ca) with and (Cb) without photoactivation, (Da) in the absence of DTT and with photoactivation, (Db) in the presence of 26 μM DTT but without photoactivation, (Dc) in the presence of 26 μM DTT and under photoactivation, (Dd) in the presence of 52 μM DTT and under photoactivation, and (De) in the presence of 77 μM DTT and under photoactivation.

Sigma-Aldrich, respectively. Ensemble PL experiments are carried out using 5 nM QD solutions in o-xylene as a function of the concentrations of added DTT solution. Single-molecule samples are prepared by the tethering of QDs on glass coverslips. First, glass coverslips are washed using 0.1 M NaOH solution, water and acetone, which is followed by soaking of the cover glasses for 30 min at room temperature in 0.5% 3mercaptopropyl triethoxysilane (MPTES) dissolved in acetone. During this step, the silanol groups on the glass surface react with MPTES and form a layer. Subsequently, 10 pM solution of CdSe/ZnS QDs in o-xylene is placed uniformly on the silanized coverslip, which resulted in the disulfide bond formation between the thiol layer tethered on glass coverslips and ZnS shells on QDs. This coupling reaction is allowed for ca. 15 min followed by the removal of free QDs by repeated washing with o-xylene under ultrasonication. During the construction of photoelectrochemical cells, TiO2-coated electrodes are prepared by the doctor-blade technique by depositing TiO2 nanoparticles paste on fluorine-doped indium tin oxide (FTO) glass plate (25 mm × 25 mm). TiO2 nanoparticles 2179

dx.doi.org/10.1021/jp4083882 | J. Phys. Chem. C 2014, 118, 2178−2186

The Journal of Physical Chemistry C

Article

paste is prepared by dispersing TiO2 nanoparticles in a 0.2 M HNO3 solution supplemented with 3 wt % of Triton X-100, 5 wt % of acetylacetone, and 10 wt % of polyethylene glycol (PEG, 20000 of molecular weight). The TiO2 content in the paste is adjusted to about 15−18 wt % to control the thickness of the TiO2 film.9 We fabricated three types of QDSSC with the working electrode as TiO2, TiO2 adsorbed with QDs, or TiO2 adsorbed with QD-(DTT)n. In the QDSSC fabrication, 100 μL of 1 μM QD solution is adsorbed on each TiO2 plate with or without the addition of 130 μM DTT solution to the electrolyte (I3−/I−). The ensemble PL spectra are recorded using a Hitachi Fluorescence Spectrophotometer F-4500. PL decay curves are recorded using a combination of a polychromator and a streak-scope (Hamamatsu, Japan). The excitation light source used in the PL lifetime measurements is 400 nm (200 kHz) femtosecond laser pulses from an optical parametric amplifier (OPA). The OPA is pumped by 200 kHz pulses from a regenerative amplifier (RegA 9000, Coherent, Japan), which is seeded by 76 MHz pulses from a mode-locked Ti:Sapphire laser (Mira 900, Coherent, Japan). The PL images of single QDs are obtained using an inverted optical microscope (IX 70, Olympus) and a iXon3 EMCCD (Andor Technology). The excitation light source used for single quantum dot imaging is a 532 nm cw laser (Millennia IIs, Spectra Physics). The photocurrent−voltage and photocurrent−time characteristic curves for the QDSSCs are measured using a BAS100B electrochemical analyzer (Hokuto-Denko, Japan) under photoirradiation with simulated sunlight of AM 1.5 (100 mW/cm2) generated using a TSSE40 sunlight simulator (Yamashita Denso, Japan) and 0.25 cm2 mask.



Figure 2. PL decay profiles of 5 nM solutions of CdSe/ZnS QDs in oxylene (A) with 77 μM (A) and (B) without DTT and recorded at 1 min intervals with 1 min photoactivation.

RESULTS AND DISCUSSION At first, we investigate the effect of thiols such as DTT on the ensemble PL QY of CdSe/ZnS QDs as functions of the concentration of DTT and time under photoactivation. Figure 1C shows the PL spectra of a 5 nM QD solution in o-xylene and in the presence of different concentrations of DTT and under continuous photoactivation. Interestingly, the PL QY of QD gradually increases from 24 to ca. 94% with increase in the concentration of DTT up to 160 μM and under photoactivation. This ca 4-fold enhancement of PL QY can be attributed to a combination of chemical and photochemical passivation of surface defects in QDs by DTT. Inset of Figure 1C shows the PL profile of a QD solution (5 nM) as a function of increase in the concentration (0−160 μM) of DTT with (a) or without (b) photoactivation. Here, the PL intensity enhances up to ca. 230% for QDs treated with DTT but without any photoactivation, which suggests DTT chemically passivates the surface defects and enhances the PL QY of QD. On the other hand, the PL QY increases up to 420% as a QD solution is photoactivated in the presence of equivalent amounts of DTT (trace “a”, inset of Figure 1C). In other words, the surface passivation of QDs with DTT and the related PL enhancement is efficient under the combined chemical (ca. 230%) and photochemical (ca. 190%) passivation. In general, as synthesized core and core−shell QDs show low PL QY, which is due to electronic defects in the core and shell, dangling-bonds on the core or shell surface, and strain at the core/shell interface or within the shell.13 However, the PL QY can be considerably improved under high-intensity illumination with UV or visible light, which is called photoenhancement or photobrightening.40 Postsynthesis factors that modify the PL

quantum efficiency of QDs include surface etching, chemicaland photo-oxidation, temperature, pH, solvent polarity, chemical reconstruction of surface, and physicochemical changes.40 Sample-by-sample variations in the PL QY of assynthesized QDs, various combinations among the above factors, and photoactivation make it difficult to figure out the exact mechanism underlying the PL enhancement. However, direct photobrightening of QDs without DTT is minimized in the current work by maintaining the excitation light intensity at a low value (