Photodeposition of Ag2S Quantum Dots and Application to

May 9, 2011 - Photodeposition of Ag2S Quantum Dots and Application to Photoelectrochemical Cells for Hydrogen Production under Simulated Sunlight...
6 downloads 0 Views 2MB Size
ARTICLE pubs.acs.org/Langmuir

Photodeposition of Ag2S Quantum Dots and Application to Photoelectrochemical Cells for Hydrogen Production under Simulated Sunlight Kazuki Nagasuna,† Tomoki Akita,‡ Musashi Fujishima,† and Hiroaki Tada*,† †

Department of Applied Chemistry, School of Science and Engineering, Kinki University, 3-4-1, Kowakae, Higashi-Osaka, Osaka 577-8502, Japan ‡ National Institute of Advanced Industrial Science and Technology, Midorigaoka 1-8-31, Ikeda, Osaka 563-8577, Japan ABSTRACT: UV light irradiation of TiO2 (λ > 320 nm) in a mixed solution of AgNO3 and S8 has led to the formation of Ag2S quantum dots (QDs) on TiO2, while Ag nanoparticles (NPs) are photodeposited without S8. Photoelectrochemical measurements indicated that the Ag2S photodeposition proceeds via the preferential reduction of Agþ ions to Ag0, followed by the chemical reaction with S8. The application of this in situ photodeposition technique to mesoporous (mp) TiO2 nanocrystalline films coated on fluorine-doped SnO2 (FTO) electrodes enables formation of Ag2S QDs (Ag2S/mp-TiO2/FTO). Ag2S/mpTiO2/FTO has the interband transition absorption in the whole visible region, while in the spectrum of Ag/mp-TiO2/FTO, a localized surface plasmon resonance absorption of Ag NPs is present centered at 490 nm. Ag2S QD-sensitized photoelectrochemical cells using the Ag2S/mp-TiO2/FTO and Ag/mp-TiO2/FTO photoanodes were fabricated. Under illumination of one sun, the Ag2S photoanode cell yielded H2 at a rate of 0.8 mL 3 h1 with a total conversion efficiency of 0.29%, whereas the Ag/mp-TiO2/FTO photoanode is inactive.

I. INTRODUCTION Production of hydrogen from water is of great importance because hydrogen gas is a clean, sustainable and storable energy source. Narrow-gap semiconductor (NGS) quantum dot (QD)sensitized mesoporous TiO2 nanocrystalline films (mp-TiO2) can work as a photoanode in photoelectrochemical (PEC) cells for converting solar energy into hydrogen energy.13 The crucial point in this case is enhancing the charge separation through efficient interfacial charge transfer (ICT) between NGS and TiO2, which needs both the intimate contact of the semiconductors 4,5 and the control of the band offset between NGS and TiO2 or the driving force for the ICT.6 Currently, the photodeposition (PD) technique is being revealed to have a wide possibility for the preparation of heteronanojunctions consisting of TiO2 and NGS such as Se,7 CdSe,79 PbS,911 PbSe,8 and CuSx.10 The application of this in situ deposition technique to mp-TiO2 allows us to obtain two unique features: one is that efficient ICT is inherently guaranteed because TiO2 photocatalysis is utilized, and the other is that a large amount of NGS QDs can be deposited on not only the external surfaces but also the inner surfaces of the mesopores with excellent reproducibility.12 Tian and Tatsuma13 have recently reported a solar cell using Ag nanoparticle (NP)-loaded mp-TiO2 (Ag/mp-TiO2) as a photoanode. In this cell, the electron injection from Ag NPs to TiO2 is induced by excitation of the localized surface plasmon resonance. However, Ag NPs absorb only a part of visible light, and the absorption intensity is fairly small. On the other hand, Ag2S, r 2011 American Chemical Society

having absorption in the whole visible region, is a promising material as a photosensitizer for photocatalysts14,15 and PEC cells.16 Methods for preparing the Ag2STiO2 coupling system are limited to the successive ionic layer adsorption and reaction17 and single-molecule precursor methods.18 Here we report a PD technique for forming Ag2S QDs on TiO2 (Ag2S/TiO2), and the Ag2S quantum dot sensitized photoelectrchemical (QD-SPEC) cell for producing hydrogen from water containing S2 ions as electron donor under illumination of simulated sunlight. To our knowledge, this is the first report of photodeposition of Ag2S on TiO2.

II. EXPERIMENTAL SECTION II.A. Photodeposition of Ag2S QDs and Ag NPs on TiO2. A paste containing anatase TiO2 particles with a mean size of 20 nm (PST18NR, Nikki Syokubai Kasei) was coated on glass substrates with a fluorine-doped SnO2 film (FTO, 12 Ω/0) by a squeegee method, and the sample was heated in air at 773 K to form mesoporous TiO2 films (mp-TiO2). A solution containing AgNO3 and S8 (solvent = CH3CN/ H2O/C2H5OH = 10:1:1 v/v/v) was placed in a double-jacket-type Pyrex reaction vessel (inner diameter = 85 mm, height = 140 mm), and then the mp-TiO2 substrate was immersed in the solution. After the solution had been bubbled with argon for 0.5 h in the dark, irradiation Received: September 10, 2010 Revised: April 12, 2011 Published: May 09, 2011 7294

dx.doi.org/10.1021/la200587s | Langmuir 2011, 27, 7294–7300

Langmuir was carried out from the direction of the mp-TiO2 film with a highpressure mercury lamp (I320400nm = 3.0 mW 3 cm2). The temperature of the reaction solution was maintained at 298 K by circulating thermostated water through the outer jacket. To identify the deposits and examine the dispersion state on the TiO2 surface by transmission electron microscopy (TEM), particulate samples were also prepared by the same procedure except for the use of 1 g of anatase TiO2 particles (A-100, Ishihara Sangyo) in place of mp-TiO2. Unless otherwise noted, the concentrations of AgNO3 and S8 were 2.76  103 M and 1.76  104 M, respectively. After the solution had been bubbled with argon for 0.5 h in the dark, irradiation (λ > 320 nm, I320400nm = 3.0 mW 3 cm2) was carried out for a given period with a high-pressure mercury lamp at 298 K. For comparison, the Ag nanoparticle-loaded TiO2 samples were prepared by the photodeposition method.19 One gram of the TiO2 particles or mp-TiO2 had been added to a 2.76  103 M aqueous solution of AgNO3. After the solution had been bubbled with argon for 0.5 h in the dark, irradiation (λ > 320 nm, I320400nm = 3.0 mW 3 cm2) was carried out for a given period with a high-pressure mercury lamp at 298 K. The resulting Ag2S/TiO2 and Ag/TiO2 samples were washed with ethanol and dried under vacuum. For the quantification of Ag in the deposits, the resulting sample was dispersed in concentrated HNO3, and the deposits were thoroughly dissolved by stirring for 1 h. Then the Ag concentration in the solution was determined by inductively coupled plasma spectroscopy (ICPS-7500, Shimadzu).

II.B. Characterization of Ag2S Quantum Dot-Loaded TiO2. Diffuse reflectance UVvis spectra of the resulting sample were recorded on a Hitachi U-4000 spectrometer mounted with an integral sphere at room temperature. The reflectance (R¥) was recorded with respect to a reference of BaSO4, and the KubelkaMunk function [F(R¥)] expressing the relative absorption coefficient was calculated by the equation F(R¥) = (1  R¥)2/2R¥. High-resolution transmission electron microscopic observation and X-ray energy-dispersive spectroscopic measurements were performed by use of a JEOL JEM-3000F and Gatan imaging filter at an applied voltage of 300 or 297 kV. X-ray photoelectron spectroscopic (XPS) measurements were performed on a Kratos Axis Nova X-ray photoelectron spectrometer with a monochromated Al KR X-ray source operated at 15 kV and 10 mA. II.C. Photoelectrochemical Measurements. Cyclic voltammograms (CV) of mp-TiO2/FTO were measured in the reaction solution containing 0.1 M NaClO4 supporting electrolyte under deaerated conditions with glassy carbon and Ag as a counterelectrode and a quasi-reference electrode, respectively. The electrode potential was swept from 0.6 to þ1.0 V (vs Agþ/Ag). Photochronopotentiometry (PCP) measurements were carried out in a 0.1 M NaClO4 electrolyte solution. After a constant potential had been reached by argon bubbling for 0.5 h in the dark, irradiation (λ > 300 nm, I320400nm = 7 mW 3 cm2) was started by using a 300-W Xe lamp as a light source (Wacom HX500). Electrochemical response with irradiation was followed for the PEC cells connected with a potentio/galvanostat (HZ-5000, Hokuto Denko). II.D. Photoconversion Efficiency Measurements. Ag2S quantum dot-sensitized solar cells (QD-SSC) consisting of Ag2S/mp-TiO2/ FTO and Ag/mp-TiO2/FTO (photoanode, apparent surface area = 7.5 cm2)|0.05 M Na2S þ 0.05 M Na2SO3 þ 0.1 M NaClO4 (aqueous electrolyte solution, pH 13)|Pt cathode|Ag/AgCl ([Cl] = 3.33 M) reference electrode were fabricated. The aqueous electrolyte solution was used after argon bubbling to remove the oxygen in the solution. Under illumination by a solar simulator (PEC-L10, Peccell technologies, Inc.) at one sun (AM 1.5, 100 mW 3 cm2) for the Ag2S QD-SPEC cells, the amount of hydrogen evolved at the Pt cathode at E was quantified by volumetric analysis, and the JE characteristics were recorded on the potentio/galvanostat (HZ-5000, Hokuto Denko): in all cases, only visible light (λ > 430 nm) was irradiated from the front side of the mp-TiO2 film by cutting off the UV light by use of an optical filter (Y-45, Toshiba)

ARTICLE

Figure 1. (a) Annular dark-field scanning transmission electron microscopic image of Ag(tp = 1 h)/TiO2 (λ > 320 nm, I320400nm = 3.0 mW 3 cm2). (b) TEM image, (c) HRTEM image, and (d) energydispersive X-ray spectrum for a sample prepared by irradiating TiO2 in a solution containing AgNO3 and S8 at 298 K for 2 h (λ > 320 nm, I320400nm = 3.0 mW 3 cm2). to suppress the photodegradation of Ag2S QDs. The total conversion efficiency (η) was calculated as a percentage by the following equation:20 η ¼ fð1:23  jEapp jÞJph =Ig  100

Eoc e Emeas e Edark

where |Eapp| (volts) is the absolute value of the potential difference between the open-circuit potential under illumination (Eoc) and the potential at which the hydrogen evolution is measured (Emeas = 0.52 V), and I is the light intensity (watts per square centimeter).

III. RESULTS AND DISCUSSION Annular dark-field scanning transmission electron microscopic image of Ag/TiO2(A-100) prepared at irradiation time (tp) = 1 h (Figure 1a) indicates that Ag particles with a mean size of 1.7 nm are highly dispersed on the TiO2 surface. After the AgPD, the particles turned from white to grayish purple. On the other hand, UV light irradiation (λ > 320 nm) to TiO2 in a deaerated solution containing AgNO3 and S8 (solvent = CH3CN/ H2O/C2H5OH = 10:1:1 v/v/v) at 298 K changed its color to black, while no change occurred without TiO2. Figure 1b shows a transmission electron microscopic (TEM) image of a sample prepared at tp = 2 h. Larger particles are deposited on TiO2 in a highly dispersed state. Figure 1 panels c and d show highresolution transmission electron microscopic (HRTEM) image and X-ray energy-dispersive (ED) spectrum for the sample, respectively. The HRTEM image exhibits clear lattice fringes with the nearest distance of 0.239 nm, which is in agreement with the value for the (220) plane of Ag2S (PDF 01-071-0995). Also, the Ag2S particle is in contact with the TiO2 surface with a large contact area. As shown in Figure 1c, the Ag2S nanoparticle has a hemispheric shape, which is true for most particles observed by TEM. The small contact angle suggests that the adhering energy of the Ag2S particle to the TiO2 surface is large. In the ED spectrum, the Ag and S signals are observed at 3.0 and 2.4 keV, respectively, besides the Ti and O signals. The Cu signals arise 7295

dx.doi.org/10.1021/la200587s |Langmuir 2011, 27, 7294–7300

Langmuir

ARTICLE

Figure 2. (A) S8 concentration dependence of the rate of Ag2S photodeposition on TiO2 particles: initial concentration of AgNO3 ([AgNO3]0) = 2.76  103 M. (B) AgNO3 concentration dependence of the rate of Ag2S photodeposition on TiO2 particles: initial concentration of S8 ([S8]0) maintained at [S8]0 = 1.73  104 M.

Figure 3. CV curves of mp-TiO2/FTO electrodes in a 1.38  102 M AgNO3 solution containing 0.1 M NaClO4, (a) without and (b) with S8 (6.90  103 M) (solvent = CH3CN/H2O/C2H5OH = 10:1:1 v/v/v): n denotes the cycle number.

from the copper grid used for analysis. Furthermore, X-ray photoelectron spectroscopic (XPS) measurements were performed to characterize the deposits. The Ag3d-binding energy (EB) for the sample were 368.0 (3d5/2) and 374.1 eV (3d3/2) with a splitting of 6.1 eV due to the spinorbit coupling, which are in agreement with the values for authentic Ag2S. In the S2p XPS spectra, two weak signals were observed at EB = 161.3 and 162.5 eV, corresponding to the S2p3/2 and S2p1/2 levels of Ag2S. Evidently, UV light irradiation of TiO2 in the solution of AgNO3 and S8 causes the deposition of Ag2S at 298 K to form the Ag2STiO2 heteronanojunction. The mean particle size of Ag2S (d, nanometers) for the sample increased from 9.7 nm at tp = 0.5 h to 13.2 nm at tp = 2 h. The effect of the initial concentrations of the starting materials on the Ag2S PD rate was examined. Figure 2A shows the initial S8 concentration ([S8]0) dependence of the PD rate, with the initial concentration of AgNO3 ([AgNO3]0) maintained at 2.76  103 M. Interestingly, Ag PD is drastically enhanced by the addition of S8, whereas the presence of excess S8 (S/Ag g 0.5) hardly affects the rate. On the other hand, Figure 2B shows the [AgNO3]0 dependence of the Ag2S PD rate at [S8]0 = 1.73  104 M. Under these conditions, the Agþ ions were confirmed to be entirely consumed at tp = 2 in each case. In the range of

Ag/S e 2, the initial PD rate is the same, and when all the Agþ ions in the solution are consumed, the PD stops at 0.1 mmol 3 g1 for Ag/S = 1 and at twice as high an amount for Ag/S = 2. Consequently, [AgNO3]0 determines the saturated amount of Ag deposited, of which the ratio in the reaction systems a:c:b is 1:5:10. These results indicate that Ag2S PD proceeds stoichiometrically in the presence of sufficient S8, and the amount of Ag2S deposited can be controlled by [AgNO3]0. To clarify the Ag2S PD mechanism, cyclic voltammograms (CV) were measured for mp-TiO2/FTO electrodes. In all the PEC measurements, an Ag electrode was used as a quasireference electrode, and the electrode potential (E) is shown with respect to that of Agþ/Ag. Figure 3 shows the CV curves of the mp-TiO2/FTO electrodes in a deaerated 1.38  102 M AgNO3 solution containing 0.1 M NaClO4 (solvent = CH3CN/ H2O/C2H5OH = 10:1:1 v/v/v), (a) without and (b) with S8 (6.90  103 M) in the potential range between 0.6 and þ1.0 V: n denotes the cycle number. At n = 1 in curve a, a pair of Agþ/ Ag0 redox current peaks is present at 0.47 and þ0.27 V. The oxidation current peak weakens with a successive shift toward the negative direction with increasing n, while the reduction current peak shifts to 0.38 V at n g 2. This latter finding may reflect the fact that the formation of Ag nuclei on TiO2 is energetically more 7296

dx.doi.org/10.1021/la200587s |Langmuir 2011, 27, 7294–7300

Langmuir difficult than the growth of Ag particles.21 Meanwhile at n = 1 in curve b, the current peaks due to the Agþ/Ag0 redox reactions are observed at 0.52 and þ0.42 V. On cycling the potential scan, the oxidation current rapidly declines to disappear at n > 5. In contrast to the case without S8, the reduction potential is invariant, which suggests that not the growth of Ag particles but the formation of Ag nuclei occurs in each potential cycle. Furthermore, photochronopotentiometric (PCP) measurements were carried out for the mp-TiO2/FTO electrode.7 Figure 4 shows the PCP profiles in a deaerated solution containing 0.1 M NaClO4 (solvent = CH3CN/H2O/C2H5OH = 10:1:1 v/v/v): an Ag electrode was used as a quasi-reference electrode, and E is shown with respect to that of Agþ/Ag. On irradiation (λ > 320 nm), the E gives rise to a drastic shift from 0 to ca. 0.5 V, which can be attributed to the Fermi energy upward shift due to the current doubling effect of ethanol.22 Note that the E is negative enough for reducing Agþ [E0(Agþ/Ag) = 0 V] but insufficient for reducing S8 [E0(S/S2) = 1.25 V]. As shown in curve a, the first addition of AgNO3 causes an abrupt potential shift to ca. 0.05 V, which is invariant by the subsequent addition of S8. Upon reversing the order of addition (curve b), E undergoes only a slight negative shift with the addition of S8, gradually

ARTICLE

approaching 0.05 V by the following addition of Agþ. A profile similar to that of curve a is obtained by simultaneous addition of AgNO3 and S8 (curve c). In every case, the mp-TiO2/FTO electrode turned black after the measurements, which is indicative of Ag2S formation on TiO2. Thus, the mechanism of Ag2S PD at the initial stage not involving the photochemical events of Ag2S can be summarized as follows (eqs 17). UV light irradiation of TiO2 triggers the excitation of electrons in the valence band (vb) to the conduction band (cb). In the anodic process, the vb holes oxidize C2H5OH, and the resulting CH3CHOH radicals, with a high reducing power, inject another electron into the cb.2224 In the cathodic process, Agþ ions adsorbed (Agþad) are preferentially reduced to Ag0. The Ag0 can spontaneously react with S to yield Ag2S, of which the standard reaction Gibbs energy is 47.1 kJ 3 mol1.25 TiO2 þ Agþ T Agþ ad 3 3 3 TiO2

ð1Þ

TiO2 þ hν f e cb þ hþ vb

ð2Þ

e cb þ hþ vb f TiO2

ð3Þ

hþ vb þ C2 H5 OH f CH3 CHOH þ Hþ

ð4Þ

CH3 CHOH f e cb þ CH3 CHO þ Hþ

ð5Þ

2Agþ ad þ 2e cb f 2Ag0

ð6Þ

2Ag0 þ Sx f Ag2 S þ Sx  1

Figure 4. PCP profiles of mp-TiO2/FTO electrode in reaction solution containing 0.1 M NaClO4 (solvent = CH3CN/H2O/C2H5OH = 10:1:1 v/v/v): (a) AgNO3 addition at tp = 20 min/S8 addition at tp = 40 min; (b) S8 addition at tp = 20 min/AgNO3 addition at tp = 40 min; (c) simultaneous addition of AgNO3 and S8 at tp = 20 min.

ðx e 8Þ

ð7Þ

Figure 5a shows UVvis absorption spectra of Ag2S/mpTiO2/FTO prepared by changing tp [Ag2S(0 e tp e 6 h)/mpTiO2]: F(R¥) denotes the KubelkaMunk function. New absorption grows at λ < 1200 nm with an increase in tp, whereas anatase TiO2 has absorption only in the UV region. In every case, the absorption intensity increases with decreasing λ, and the absorption edge significantly red-shifts as a result of the increase in tp. For comparison, the absorption spectra of Ag/mp-TiO2/ FTO are shown in the inset. Ag/mp-TiO2/FTO has broad absorption with a peak around 490 nm due to the localized surface plasmon resonance absorption of the Ag NPs. Figure 5b shows the indirect

Figure 5. (a) UVvis absorption spectra of Ag2S(0 e tp e 6 h)/mp-TiO2/FTO prepared by changing tp: F(R¥) denotes the KubelkaMunk function. (Inset) Absorption spectra of Ag(tp = 0.25, 0.5, 2 h)/mp-TiO2/FTO. (b) Plots of the Ag2S band gap (Eg) for Ag2S(0 e tp e 6 h)/mp-TiO2/FTO vs irradiation time (tp). 7297

dx.doi.org/10.1021/la200587s |Langmuir 2011, 27, 7294–7300

Langmuir band gap (Eg) determined for Ag2S(tp)/mp-TiO2/FTO from the Tauc plots.26 At tp < 4 h, the Eg value increases relative to the bulk value of ca. 1 eV.27 A similar trend was observed also in the photodeposition of CdS.28 The steep increase in Eg at tp < 2 h possibly results from the size quantization effect of Ag2S particles, since the effective masses of the electrons (≈0.15m0)29 and holes (≈0.30m0)30 are fairly small. Recently, the blue shift in the absorption edge in the Ag2S colloid system with a particle size smaller than 10 nm has been attributed to the quantum confinement effect.31 Currentpotential (JE) curves were measured for the Ag2S QD-SPEC cells using Ag2S/mp-TiO2/FTO photoanode (Scheme 1): E is shown with respect to the standard Ag/AgCl ([Cl] = 3.33 M) electrode potential below. Figure 6A shows the JE curves for cells using Ag2S(tp)/mp-TiO2/FTO photoanodes in the dark (curve a) and under illumination of one sun (AM 1.5, 100 mW 3 cm2) from the direction of the Ag2S/mpTiO2 film (curves bd). The photocurrent (Jph) rises at E > ca. 0.8 V, while in the dark, the electrolytic current due to the oxidation of S2 ions starts to flow at E ≈ 0.52 V (Edark). Also, the maximum Jph is obtained for Ag2S(tp =1 h)/mp-TiO2/FTO. Figure 6B shows time courses for the H2 generation in the Ag2S Scheme 1. Structure of Ag2S QD-SPEC Cell

ARTICLE

QD-SPEC cells under illumination of one sun (AM 1.5, 100 mW cm2) at E = 0.52 V. The Ag2S(tp)/mp-TiO2/FTO photoanode yields H2, whose amount is almost proportional to tp at tp > 10 min. This indicates that S2 ions operate as good hole(Ag2S) scavengers to restrict its photocorrosion.32 On the other hand, the Ag/mp-TiO2/FTO photoanode is inactive. Mild oxidation ability can be induced in Ag/TiO2 by visible light irradiation;33 however, it may be insufficient for the oxidation of S2 ions. As shown in Scheme 1, upon visible light illumination (λ > 430 nm) of the Ag2S/mp-TiO2/FTO photoanode, the electrons in the vb(Ag2S) are excited to the cb(Ag2S). The electrons in the cb(Ag2S) are injected into the cb(TiO2), while the holes in the vb(Ag2S) oxidize S2 ions to S22 ions. Iintimate direct coupling between TiO2 and Ag2S formed by the PD technique would be preferable to interfacial electron transfer from Ag2S to TiO2. The electrons are transported to the underlying FTO electrode in competition with the back reaction with S22 ions. SO32 ions prevent the back reaction by reducing S22 ions back to S2 ions.34 The electrons transferred to the Pt counterelectrode via the external circuit reduce Hþ to H2 with the aid of the high electrocatalytic activity. In Figure 6B, the H2 generation rate (vH2) strongly depends on tp for Ag2S deposition, and the Ag2S(tp = 1 h)/mp-TiO2/FTO photoanode gives a maximum value of 0.8 mL 3 h1. The increase in vH2 with increasing tp at tp < 1 h results from the increase in light absorption efficiency. On the other hand, from the ionization potential (þ0.5 V)35 and the Eg for bulk Ag2S, the cb edge (Ecb) of Ag2S is estimated to be ca. 0.7 V, which is more positive than Ecb(TiO2) at pH 13 (ca. 1.1 V). This means that the rise in the Ecb(Ag2S) due to size quantization is necessary for efficient interfacial electron transfer from Ag2S to TiO2,12 in which the template effect of mp-TiO2 plays a crucial role. Additional interpretation of the decrease in vH2 at tp > 1 h is possible. When an excess amount of Ag2S is deposited, while the incident light is completely absorbed, the excited electrons generated in the surface layer are subject to recombination rather than injection into TiO2. As a result, the efficiency of H2 evolution is reduced. Consequently, the optimum tp for preparing the photoanode should be determined by the balance between light absorption efficiency and interfacial electron transfer efficiency. Figure 7 shows the total conversion efficiency under illumination of one sun (η) at E = 0.52 V as a function of tp: mp-TiO2 film thicknesses are (a) 2.6 and (b) 5.0 μm. Since sulfur compounds

Figure 6. (A) JE curves for cells using Ag2S(tp)/mp-TiO2/FTO photoanodes in the dark (curve a) and under illumination of one sun (AM 1.5, 100 mW cm2) (curve b, tp = 0.25 h; curve c, tp = 1 h; curve d, tp = 6 h). (B) Time courses for the H2 generation in the Ag NP-SPEC cell (curve a) and Ag2S QDSPEC cells (curves bd) under illumination of one sun (AM 1.5, 100 mW 3 cm2) at E = 0.52 V (vs Ag/AgCl): (b) tp = 0.25 h, (c) tp = 1 h, (d) tp = 6 h. 7298

dx.doi.org/10.1021/la200587s |Langmuir 2011, 27, 7294–7300

Langmuir

ARTICLE

’ ACKNOWLEDGMENT We thank H. Yamamoto for the XPS measurements. This work was supported by a Grant-in-Aid for Scientific Research (B) 20350097 from the Ministry of Education, Science, Sport, and Culture, Japan. ’ REFERENCES

Figure 7. Total conversion efficiency under illumination of one sun (η) at E = 0.52 V as a function of tp: thicknesses of mp-TiO2 films are (a) 2.6 and (b) 5.0 μm.

are abundant in nature, the use of S2 ions as a sacrificial electron donor should not devaluate this QD-SPEC cell. However, the range of Emeas is restricted between Eoc and Edark in the presence of the sacrificial agents (S2 and SO32 ions), while it can be changed within 1.23 V in a pure water-splitting system.20,36,37 The conversion efficiency strongly depends on the thickness of the mp-TiO2 film, and the η values for the 5.0 μm samples are much higher than those for the 2.6 μm samples. This is probably due to the larger light absorption by Ag2S QDs in the former sample. Both the plots show volcano-shaped curves, reaching maxima of 0.13% in plot a and 0.29% in plot b at Ag loading of 0.52 and 0.70 μmol 3 cm2, respectively. The same discussion on the data in Figure 6B would also be valid in this case.

IV. CONCLUSIONS Ag2S QDs have been deposited on TiO2 by irradiation of TiO2 (λ > 320 nm) in a mixed solution of AgNO3 and S8 (solvent = CH3CN/H2O/C2H5OH = 10:1:1 v/v/v). In this photodeposition technique, the loading amount and particle size of Ag2S can be controlled by irradiation time. Photoelectrochemical measurements showed that the photodeposition of Ag2S proceeds via preferential reduction of Agþ ions to Ag0, followed by chemical reaction with S8. By applying this in situ photodeposition technique to mesoporous TiO2 nanocrystalline films (mpTiO2), significant size quantization of Ag2S appears due to the template effect of mp-TiO2. Ag2S QD-sensitized photoelectrochemical (QD-SPEC) cells consisting of Ag2S/mp-TiO2/FTO (photoanode, apparent surface area = 7.5 cm2)|0.1 M NaClO4 þ 0.05 M Na2S þ 0.05 M Na2SO3 (electrolyte solution)|Pt (cathode)|Ag/AgCl (reference electrode) yields H2 with a constant rate, whereas the Ag/mp-TiO2/FTO photoanode is inactive. Under optimum conditions, this Ag2S QD-SPEC cell produces hydrogen with a rate of 0.8 mL 3 h1 or a total conversion efficiency of 0.29%. This simple in situ photodeposition technique for preparing the Ag2STiO2 heterojunction would be useful for applications of PEC nanodevices and photocatalysts. ’ AUTHOR INFORMATION Corresponding Author

*Telephone þ81-6-6721-2332; fax þ81-6-6727-2024; e-mail [email protected].

(1) Gr€atzel, M. Nature 2001, 414, 338. (2) Youngblood, W. J.; Lee, S. A.; Maeda, K.; Mallouk, T. E. Acc. Chem. Res. 2009, 42, 1966. (3) Ogisu, K.; Takanabe, K.; Lu, D.; Saruyama, M.; Ikeda, T.; Kanehara, M.; Teranishi, T.; Domen, K. Bull. Chem. Soc. Jpn. 2009, 82, 528. (4) Lawless, D.; Kapoor, S.; Meisel, D. J. Phys. Chem. 1995, 99, 10329. (5) Mukaihata, N.; Matsui, H.; Kawahara, T.; Fukui, H.; Tada, H. J. Phys. Chem. C 2008, 112, 8702. (6) Kongkanand, A.; Tyrdy, K.; Takeuchi, K.; Kuno, M.; Kamat, P. V. J. Am. Chem. Soc. 2008, 130, 4007. (7) Somasundaram, S.; Chenthamarakshan, C. R.; de Tacconi, N. R.; Ming, Y.; Rajeshwar, K. Chem. Mater. 2004, 16, 3846. (8) Chenthamarakshan, C. R.; Ming, Y.; Rajeshwar, K. Chem. Mater. 2000, 12, 3538. (9) Nguyen, V. N. H.; Amal, R.; Beydoun, D. J. J. Photochem. Photobiol., A 2006, 179, 57. (10) Zhukovskiy, M. A.; Stroyuk, A. L.; Shavalagin, V. V.; Smirnova, N. S.; Lytvyn, O. S.; Eremenko, A. M. J. Photochem. Photobiol., A 2009, 203, 137. (11) Jin-nouchi, Y.; Akita, T.; Tada, H. ChemPhysChem 2010, 11, 2349. (12) Jin-nouchi, Y.; Naya, S.-i.; Tada, H. J. Phys. Chem. C 2010, 114, 16837. (13) Tian, Y.; Tatsuma, T. Chem. Commun. 2004, 1810. (14) Neves, M. C.; Nogueira, J. M. F.; Trindade, T.; Mendonca, M. H.; Pereira, M. I.; Monteiro, O. C. J. Photochem. Photobiol., A 2009, 204, 168–173. (15) Xe, Y.; Heo, S. H.; Kim, Y. N.; Yoo, S. H.; Cho, S. O. Nanotechnology 2010, 21, 015703. (16) Nasrallah, T. B.; Dlala, H.; Amlouk, M.; Belgacem, S.; Bernede, J. C. Synth. Met. 2005, 151, 225. (17) Vogel, R.; Hoyer, P.; Weller, H. J. Phys. Chem. 1994, 98, 3183. (18) Neves, M. C.; Monteiro, O. C.; Hempelmann, R.; Silva, A. M. S.; Trindade, T. Eur. J. Inorg. Chem. 2008, 4380. (19) Nishimoto, S.-I.; Ohtani, B.; Kajiwara, H.; Kagiya, T. J. Chem. Soc., Faraday Trans. 1 1983, 79, 2685. (20) Khan, S. U. M.; Al-Shahry, M; Ingler, W. B., Jr. Science 2002, 297, 2243. (21) Henglein, A. Ber. Bunsen-Ges. Phys. Chem. 1990, 94, 600. (22) Morrison, S. R. Electrochemistry at Semiconductor and Oxidized Metal Electrodes, Plenum Press: New York, 1980. (23) Rajeshwar, K.; Chenthamarakshan, C. R.; Ming, Y.; Sun, W. J. Electroanal. Chem. 2002, 538539, 173. (24) Ming, Y.; Chenthamarakshan, C. R.; Rajeshwar, K. J. Photochem. Photobiol., A 2002, 147, 199. (25) Handbook of Chemistry and Physics, 83rd ed.; Lide, D. R., Ed.; CRC Press: Boca Raton, FL, 2002. (26) Tauc, J.; Grigorovich, R.; Vancu, A. Phys. Status Solidi 1966, 15, 627. (27) El-Nahass, M. M.; Farag, A. A. M.; Ibahim, E. M.; Abd-ElRahman, S. Vacuum 2004, 72, 453. (28) Fujii, M.; Nagasuna, K.; Fujishima, M.; Akita, T.; Tada, H. J. Phys. Chem. C 2009, 113, 16711. (29) Ohachi, T.; Imai, S.; Tanaka, T.; Yamai, H.; Taniguchi, I. Solid State Ionics 1988, 2830, 1160. (30) Ehrlich, S. H. Photogr. Sci. Eng. 1979, 23, 348. (31) Leon-Velazquez, M. S.; Irizarry, R.; Castro-Rosario, M. E. J. Phys. Chem. C 2010, 114, 5839. 7299

dx.doi.org/10.1021/la200587s |Langmuir 2011, 27, 7294–7300

Langmuir

ARTICLE

(32) Tachibana, Y.; Akiyama, H. Y.; Ohtsuka, Y.; Torimoto, T.; Kuwabata, S. Chem. Lett. 2007, 36, 88. (33) Naya, S.-i.; Inoue, A.; Tada, H. J. Am. Chem. Soc. 2010, 132, 6292. (34) Reber, J. F.; Rusek, M. J. Phys. Chem. 1986, 90, 824. (35) Kashida, S.; Watanabe, N.; Hasegawa, T.; Iida, H.; Mori, M.; Savrasov, S. Solid State Ionics 2003, 158, 167–175. (36) Mor, G. K.; Shankar, K.; Paulose, M.; Varghese, O. K.; Grimes, C. A. Nano Lett. 2005, 5, 191. (37) Gao, X.-F.; Li, H.-B.; Sun, W.-T.; Chen, Q.; Tang, F.-Q.; Peng, L.-M. J. Phys. Chem. C 2009, 113, 7531.

7300

dx.doi.org/10.1021/la200587s |Langmuir 2011, 27, 7294–7300