In Situ Formation of Heterojunctions in Modified Graphitic Carbon

Sep 19, 2014 - In situ construction of fibrous AgNPs/g-C 3 N 4 aerogel toward light-driven CO x -free methanol dehydrogenation at room temperature. Yu...
4 downloads 14 Views 3MB Size
Download PDF
Article pubs.acs.org/cm

In Situ Formation of Heterojunctions in Modified Graphitic Carbon Nitride: Synthesis and Noble Metal Free Photocatalysis Menny Shalom,*,† Miguel Guttentag,‡ Christian Fettkenhauer,† Sahika Inal,§,∥ Dieter Neher,§ Antoni Llobet,‡ and Markus Antonietti† †

Max-Planck Institute of Colloids and Interfaces, Department of Colloid Chemistry, Research Campus Golm, 14424 Potsdam, Germany ‡ Institut Català d’Investigació Química, Paisos Catalans 16, Tarragona 43007, Spain § Institute of Physics and Astronomy, University of Potsdam, Karl-Liebknecht-Strasse 24-25, 14476 Potsdam, Germany ∥ Department of Bioelectronics, Ecole Nationale Supérieure des Mines CMP-EMSE, MOC, Gardanne, France S Supporting Information *

ABSTRACT: Herein, we report the facile synthesis of an efficient roll-like carbon nitride (C 3 N 4 ) photocatalyst for hydrogen production using a supramolecular complex composed of cyanuric acid, melamine, and barbituric acid as the starting monomers. Optical and photocatalytic investigations show, along with the known red shift of absorption into the visible region, that the insertion of barbituric acid results in the in situ formation of in-plane heterojuctions, which enhance the charge separation process under illumination. Moreover, platinum as the standard cocatalyst in photocatalysis could be successfully replaced with first row transition metal salts and complexes under retention of 50% of the catalytic activity. Their mode of deposition and interaction with the semiconductor was studied in detail. Utilization of the supramolecular approach opens new opportunities to manipulate the charge transfer process within carbon nitride with respect to the design of a more efficient carbon nitride photocatalyst with controlled morphology and optical properties.

■

INTRODUCTION The production of solar fuels through water splitting is highly desired because of the growing demand for clean and renewable energy.1 In general, the semiconductor photocatalyst should possess a suitable band gap (>1.23 eV) with the right position of the conduction and valence band.2 In addition, the photocatalytic performance is strongly dependent on the charge transfer processes in the semiconductor under illumination (electron and hole separation and internal recombination rate). One of the ways to manipulate the charge separation of excitons in many light-driven reactions (e.g., in solar cells,3 in light-emitting diodes,4 or for photocatalysis5) is the formation of a heterojunction that can be carefully designed to direct the charge flow in a given direction. While most of the research in this field was focused on metal-based semiconductors (metal oxides,6,7 sulfides,8 and nitrides9) as photocatalysts, in the past several years metal free graphitic carbon nitride (C3N4) materials have attracted widespread attention because of their outstanding (electro)catalytic and photocatalytic activity.10 C3N4 materials demonstrate high activity in many energy-related fields: as catalysts in heterogeneous catalysis and as metal free photocatalysts and electrocatalysts in water splitting reactions,11−13 in the degradation of pollutants,14 and many more. The growing © XXXX American Chemical Society

interest in C3N4 materials as photo- and electrocatalysts is mainly due to their excellent chemical and photophysical properties. Apart from its high chemical and thermal stability (up to 400 °C), energy level positions of C3N4 materials are suitable for water splitting under illumination (both proton reduction and water oxidation). The photocatalytic activity of C3N4 is strongly dependent on its morphology, size, surface area, defects, and energy states along with its electronic properties (band gap, exciton lifetime, etc.). In the past year, many modifications of the common synthesis were developed, including the introduction of new heteroatoms (sulfur and phosphor)15 into the C3N4 structure, to manipulate its electronic, optical, and catalytic properties. However, despite great progress in C3N4 synthesis, the traditional solid state synthesis of C3N4 results in a disorganized texture, with small grain size and surface defects, which leads to low catalytic activity. Recently, other groups and ours used the supramolecular chemistry approach to synthesize ordered structures of C3N4 such as hollow boxes, spheres, and spherical macroscopic Received: September 2, 2014 Revised: September 18, 2014

A

dx.doi.org/10.1021/cm503258z | Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials

Article

assemblies.14,16 In addition, this approach can also be applied to grow C3N4 on different substrates.11 Supramolecular chemistry provides a great opportunity for the synthesis of nanostructured materials without any further templating techniques. The supramolecular approach includes the use of noncovalent interactions such as hydrogen bonding to form order between building blocks for the desired synthesis. Hydrogen bonds are very useful for controlling molecular self-assembly because of the reversibility, specificity, and directionality of this class of interactions.17,18 The structure of the final products can be controlled by choosing the appropriate monomers and solvents for the synthesis. The starting monomers will organize into different structures according to their ability to form hydrogen bonds in the given solvent and form ordered and stable aggregates that consecutively define the resulting materials. Here we use the supramolecular approach to improve the efficiency of the C3N4 photocatalyst for hydrogen evolution. The modified carbon nitride was prepared by the pyrolysis of the supramolecular complex of cyanuric acid (C), melamine (M), and barbituric acid (B) as precursors in different molar ratios. We demonstrate that changes in the hydrogen bond network and the replacement of nitrogen atoms with carbon strongly influenced the electronic and chemical properties of C3N4. Moreover, lifetime measurements indicated the in situ generation of heterojunctions, which are favorable for photocatalytic applications. The chemical structure, morphology, and optical properties of the resulting carbon nitrides were characterized by X-ray diffraction (XRD), Fourier transform infrared (FT-IR) spectroscopy, scanning electron microscopy (SEM), elemental analysis, UV−vis absorption, and steady state and time-resolved fluorescence spectroscopy. The photocatalytic activities of the different modified C3N4’s were tested by measuring hydrogen evolution in combination with Pt, first row transition metal salts, and known water reduction catalysts as cocatalysts.

■

(modified Hamamatsu) multialkaline photomultiplier. The excitation wavelength was 405 nm. Raw decay data obtained as the logarithm of photon counts versus measurement time were analyzed with data analysis software from PicoQuant GmbH. The decay times were extracted by means of a reconvolution fit based on a triple-exponential −t/τi , where τi is the lifetime model. Considering that IPL(t) = ∑i=n i=1aie and ai is the amplitude of the ith component, the intensity-averaged fluorescence lifetime ⟨τ⟩ was calculated as i=n

⟨τ ⟩ =

i=n

∑ aiτi 2/∑ aiτi i=1

i=1

(1)

Hydrogen evolution reactions with Pt were performed using a side irradiated closed steel reactor equipped with a Teflon inlet, a thermocouple, a pressure sensor, a magnetic stir bar, and a thermostat, connected to a Schlenk line. Pt and triethanolamine (TEOA) were used as a cocatalyst for H2 generation and as a sacrificial hole scavenger, respectively. Being a good buffering agent, TEOA provides a stable pH of 10.8 for the duration of the experiment. The detailed description of the setup and the test procedure can be found in ref 19. Pt0 (3 wt %) was in situ photodeposited onto tested photocatalysts using hexachloroplatinic acid (H2PtCl6) as a precursor. During the experiment, the buildup of pressure was monitored as a function of irradiation time. A white LED array was used as the irradiation source. Hydrogen Evolution Setup and Measurement Procedure with Pt. All catalytic experiments were conducted under an argon atmosphere. The double-walled and thermostatically controlled reaction vessel was connected to a digital pressure sensor (TypeP30, DP = −0.1%, WIKA Alexander Wiegand SE & Co. KG) to monitor the pressure increase due to hydrogen evolution; 50 mg of the sample was placed inside the reactor. Then the reactor was evacuated and refilled with argon several times. H2O and TEOA were pretreated before being used. H2O was first degassed for 1 h under vacuum in an ultrasonic bath and then purged with argon for 1 h. TEOA was purged for 1 h with argon. The solvent mixture (38 mL), composed of water and triethanolamine (TEOA) in a 9:1 (v/v) ratio, and 39.4 μL of a H2PtCl6 solution, 8% in water (corresponds to a theoretical value of 3 wt % Pt loading onto the catalyst, total concentration of 0.207 mM), were added, and the temperature was maintained at 25 °C by a thermostat. After being stirred for 10 min to reach thermal equilibrium, the reaction mixture was started by switching on a 50 W (the illumination intensity on the photocatalyst surface is 10000 μmol s−1 m−2) white LED array (Bridgelux BXRA-50C5300; λ > 410 nm). The amount of evolved gas was continuously monitored by means of the time-dependent pressure increase (see the example in Figure S6 of the Supporting Information). The hydrogen evolution rate was calculated according to the ideal gas law:

EXPERIMENTAL SECTION

Synthesis of CM1Bx−C3N4. The CMB complexes were prepared by using a 1:1 cynauric acid:melamine (CM) molar ratio and x molar barbituric acid (with respect to the total CM, x = 0.033, 0.05, 0.1, and 0.5) in 40 mL of water (e.g., for CM1B0.5 1 g of cyanuric acid, 1 g of melamine, and 1 g of barbituric acid were used). Then the complexes were mixed for 4 h using an automatic shaker. Then the white CMB complexes were filtered and washed several times with water. The resulting powders were dried at 60 °C in a vacuum oven overnight and then calcined at 550 °C for 4 h (heating rate of 2.3 °C/min) under an inert nitrogen atmosphere. Characterization. X-ray diffraction patterns were measured on a Bruker D8 Advance instrument using Cu Kα radiation. Nitrogen sorption measurements were taken with N2 at 77 K after the samples had been degassed at 150 °C under vacuum for 20 h using a Quantachrome Quadrasorb SI porosimeter. The apparent surface area was calculated by applying the Brunauer−Emmett−Teller (BET) model to the isotherm data points of the adsorption branch. Elemental analysis was accomplished as combustion analysis using a Vario Micro device. SEM images were recorded on a LEO 1550-Gemini instrument. FT-IR spectra for the characterization of the compounds were recorded on a Varian1000 FT-IR spectrometer. Optical absorbance spectra were measured using a Varian spectrophotometer equipped with an integrating sphere. The emission spectra were recorded on an LS-50B PerkinElmer instrument. Time-resolved fluorescence measurements were performed by using a single-photon counting setup (TCSPC) with a Becker&Hickl PML spectrometer (modified Oriel MS-125). As the excitation source, a picosecond laser diode at 405 nm was operated at a repetition rate of 20 MHz. The detector comprises a Becker&Hickl PML-16-C-1

ṅ =

ΔpV n = 105 t RTt

where ṅ is the hydrogen evolution rate (micromoles per hour), n the number of moles hydrogen (micromoles), t the illumination time (hours), Δp the pressure increase (bars) during irradiation time t, V the volume headspace above the reaction solution, R the universal gas constant (8.314 J mol−1 K−1), and T the temperature (298 K). To confirm that the evolved gas was hydrogen, the headspace of the reactor was analyzed by mass spectrometry (Pfeiffer Vacuum ThermoStar GSD 301 T gas analyzing system, using argon as the carrier gas) after the test. The external quantum efficiencies20 (EQEs) were calculated by

EQE =

amount of hydrogen per second amount of photons per second

Hydrogen Evolution Experiments with First Row Transition Metals and Complexes. CM1B0.05 was synthesized as described above. Cobalt complexes S2 and S3 were synthezised according to literature procedures. 21,22 Complex S1 was synthesized from [CoIIIDOHBr2] by removal of the two bromide ions with 2 equiv of AgTFMS in water in the complete absence of light. After filtration and B

dx.doi.org/10.1021/cm503258z | Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials

Article

clearly observed in the SEM images of all CM1Bx−C3N4 samples (Figure 1).

lyophilization [CoIIIDOH(H2O)2]TFMS2 (S1) was obtained as a brown powder. Synthesis of the bromo derivative has been described previously.23 Photocatalysis. Photocatalytic experiments were performed in a double-walled, thermostated glass reactor at a controlled temperature (25 °C) in a total volume of ∼6 mL. To a solution of 3 mL of 10 vol % TEOA and 0.5 mM cocatalyst was added 4 mg of slightly ground CM1B0.05. The solution was degassed with N2 for 30 min before irradiation with a xenon lamp (ABET Technologies, LS 150 SN 268). H2 was detected by means of a Clark electrode (Unisense Microsensor H2-N). The reactors were calibrated by injection of known quantities of H2 into the headspace of the degassed reaction vessels after each experiment. Analysis of Solids and Solutions after Catalysis. Solids from catalytic solutions were separated from the solutions either by centrifugation (in the case of recycling of the solid, SEM-EDX and TEM analyses; samples were washed with water until neutrality was reached) or by filtration (0.2 μm syringe filters; in the case of recycling of the solutions and ICP quantification measurements). Solids were then dried by lyophilization overnight, and solutions were directly used for further experiments or diluted 25-fold for ICP measurements. SEM-EDX experiments were performed on a Jeol 6400 electron microscope connected to a rX micro analyst (Inca Systems, Oxford Instruments), TEM experiments on a Jeol 1011 transmission electron microscope, and ICP experiments on an ICP-OES Spectro Arcos spectrometer.

■

RESULTS AND DISCUSSION Supramolecular CM1Bx complexes were prepared by mixing different molar ratios of the three precursors in water (equimolar amounts of melamine and cyanuric acid = CM1, x = 0.0033, 0.05, 0.1, and 0.5 molar ratios of barbituric acid with respect to CM1; 1:x CM:B). Although the solubility of the three monomers in water is generally low, it was sufficient to allow for the formation of organized aggregates after the monomers had been shaken for 4 h (Figure S1 of the Supporting Information). In the absence or in small amounts of barbituric acid (60% higher than that of mesoporous C3N4, which is one of the most photoactive C3N4 modifications reported to date. In addition, the photocatalyst demonstrated high stability in the hydrogen evolution reaction, as shown in Figures S4 and S6 of the Supporting Information. The enhanced photocatalytic activity can originate from several factors: (1) better light harvesting that can lead to more charge carriers that can be used in the catalytic process, (2) improvement in charge separation efficiency (with appropriate energy band positioning), and (3) a large surface area. From the absorption spectra and the hydrogen production rate, it can be estimated that there is no direct correlation between the light harvesting and the photocatalytic activity (e.g., the hydrogen production of CM1B0.1 and CM1B0.5 is lower than that of CM1B0.05 and CM1B0.033, although they absorb more light). To further study the excited state dynamics, we employed time-resolved fluorescence spectroscopy with a single photon counting setup.26 The intensity-averaged fluorescence lifetimes of all the C3N4 materials are shown in Figure 4b. In accordance with the trend in the steady state PL

Figure 3. (a) Diffuse reflectance absorption spectra and (b) emission spectra of CM1Bx−C3N4 obtained after condensation at 550 °C for 4 h. Figure 4. (a) TOFs for different CM1Bx−C3N4’s, bulk C3N4, and mesoporous graphitic (mpg) C3N4 under photocatalytic conditions (50 mg of g-C3N4, 0.21 mM H2PtCl6, 10 vol % TEOA, H2O, total volume of 38 mL). (b) Intensity-averaged fluorescence lifetimes of CM1Bx-C3N4’s. The samples were excited at 405 nm, and their emission was monitored at 475, 500, and 525 nm. (c and d) TOF vs. the surface area and TOF vs the average lifetime, respectively. For the best material (CM1B0.05), the TOF statistical error is 400 nm cutoff filter), and the evolving hydrogen was detected by means of a Clark electrode. Apart from TEOA, other known sacrificial reducing agents were tested [namely, ascorbic acid, 27 tris(2carboxyethyl)phosphine/ascorbic acid, 28 and hydroquinone29]; however, none of those systems produced hydrogen. Figure 5a shows the hydrogen evolution traces for all active cocatalysts, including the trace from a blank experiment, in which a degassed solution of only CM1B0.05 and TEOA was irradiated. Turnover frequencies (TOFs), turnover numbers (TONs), and initialization times (time between the beginning of irradiation and the start of H2 production) for all investigated cocatalysts are summarized in Table S4 of the Supporting Information. While metal salts Co(BF4)2, NiAc2, and the complex [CoIIIDOH(H2O)2]TFMS2 (S1)30 showed catalytic activity, CuTFMS2, Fe(SO4)2, ZnCl2, AlTFMS2, [CoIIICRCl2]Cl (S2),31 and [CoIITpyBr2] (S3)32 were inactive toward H2 production (Scheme S2 of the Supporting Information). In the absence of a suitable cocatalyst, the system did not produce any hydrogen. Complex S1 showed the highest initial rate with an initialization of 2.5 h, while Co(BF4)2 performed less efficiently but started H2 production after only 1 h. NiAc2 was less active than both cobalt species but started producing hydrogen almost instantaneously after the beginning of irradiation. This stands in contrast to a report by Sun et al. When using traditional gC3N4 with Acriflavine as a cophotosensitizer, nickel salts performed better than their cobalt analogues.33 The rather surprising complete inactivity of the two well-known WRCs, S231 and S3,32 gave a first indication of the mode of interaction between CM1B0.05 and its cocatalysts (vide inf ra). Although

Figure 5. (a) H2 evolution traces for photocatalytic experiments with different cocatalysts under standard conditions (4 mg of CM1B0.05, 0.5 mM metal salt or WRC, 10 vol % TEOA, H2O, total volume of 3 mL) and a blank experiment without a cocatalyst. (b) H2 evolution traces for recycling experiments with Co(BF4)2: black for a typical first run under standard conditions (corresponds to the red curve in panel a), green for recycled CM1B0.05 from the first run with a fresh TEOA solution (10 vol %), red for recycled CM1B0.05 from the first run with a fresh TEOA solution (10 vol %) and additional Co(BF4)2 (0.5 mM), blue for reuse of the solution from the first run after irradiation with fresh CM1B0.05 (4 mg) for 21 h, and cyan for reuse of the solution from the first run after irradiation with fresh CM1B0.05 (4 mg) for 70 h. (c) H2 evolution trace (red) for recycled CM1B0.05 with a fresh TEOA solution (10 vol %) after the first run was performed in the dark and a standard experiment (black) for comparison (corresponds to the red curve in panel a). (d) Amount of surface-bound cobalt, as a percentage of the initial concentration (0.5 mM), for catalytic experiments under standard conditions stopped after different irradiation times.

reduction by the semiconductor should be thermodynamically feasible, neither of the two catalysts produced hydrogen. To understand the nature of the catalytically active species when using Co(BF4)2 as a cocatalyst, a series of experiments in which either the solution or the solids were reused for a second photocatalytic experiment were performed (Figure 5b,c). At first, a catalytic standard solution was centrifuged after irradiation for 1 day, and the solid was washed and dried. Then, upon addition to a fresh solution of 10 vol % TEOA only, the catalytic activity was restored to ∼50% of the original initial rate (Figure 5b, green). When the procedure was repeated, full catalytic activity could be restored upon addition of additional Co(BF4)2 to the second run (Figure 5b, red). When the catalytic solution was filtered after irradiation for 1 day and the filtrate with fresh CM1B0.05 was reused, catalytic activity was restored to only a very small extent (Figure 5b, dark blue). These results strongly indicate a surface-bound catalytically active cobalt species. This was further underlined by an experiment in which the first cycle was performed in the dark (one night, no H2 detected). After separation from the first catalytic solution and subsequent washing and drying, the solid was used with a fresh solution of TEOA. The material still showed catalytic activity, suggesting the deposition or incorporation of cobalt into the semiconductor even in the absence of photoinduced reducing equivalents (Figure 5c, red). Because the reproducibilities of the second runs were in all cases very dependent on the duration of the first run, a sequence of standard experiments were stopped at different reaction times, and the cobalt concentrations of the filtered solutions were determined by ICP measurements. The differences between the initial cobalt concentration and the results from the ICP quantification were assumed to be the amount of metal deposited in some manner on the surface (Figure 5d). Until ∼25 h after the start of the reaction, the E

dx.doi.org/10.1021/cm503258z | Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials

Article

surface cobalt concentration steadily increased to the point at which all cobalt was bound to the semiconductor. This explained the low but still detectable activity in the second run of the recycling experiment, when fresh CM1B0.05 was added to the filtered catalytic solution, and the not fully restored activity of the recycled CM1B0.05 sample, because in both experiments the catalysis was stopped shortly before reaching complete deposition of the cobalt (small amount of cobalt still in solution). Therefore, when the first cycle was stopped after irradiation for 70 h (to ensure complete deposition of the metal) and then the solution with a fresh sample of CM1B0.05 was reused, no hydrogen was detected after irradiation for 1 day (Figure 5b, light blue). For the other metal salts investigated, the amount of deposited material was determined in the same manner with which the surface cobalt content was established. With the exception of AlTFMS2, all metal ions were deposited on the surface of the semiconductor or somehow incorporated into its framework. The Zn salt and the Cu salt almost completely disappeared from the solution and the Fe salt to ∼18%. Nickel was deposited to a final concentration of ∼28% after catalysis for 1 day. As described for Co(BF4)2, it probably gradually accumulated on CM1B0.05. The same applies for complex S1, which showed around 32% less cobalt in solution after the same irradiation period (Table S4 of the Supporting Information). The fact that the two highly stable WRCs, S2 and S3, were inactive toward catalysis strongly suggests that mostly deposited material is acting as active species and that a significant portion of the activity of S1 is due to its deposited decomposition products being subjected to a fate similar to that of Co(BF4)2. However, considering the high activity of S1 and reports of conventional g-C3N4 used with similar WRCs, the possibility of a homogeneous catalytic contribution cannot be discarded completely.34,35 To further corroborate the presence of cobalt in the solids and as well to determine the nature of the deposition (either in the form of nanoparticles or isolated, into the framework of the semiconductor-incorporated cobalt ions), SEM-EDX and TEM experiments were performed with the solids before and after catalysis (Figure 6). Although cobalt was found by EDX analysis in the samples from after the catalysis (up to 0.2 mol %, roughly consistent with ICP data), it was not possible to pinpoint the location of the metal. Distribution maps showed evenly dispersed cobalt over the whole of several of the investigated areas. In samples of fresh CM1B0.05, no cobalt was detected. In agreement with the lack of discrete cobalt sites on the surface, no large agglomerates (>5 nm) of particles could be found in TEM analysis either, suggesting a monatomic distribution of cobalt on or in the semiconductor. In addition, the absence of crystalline cobalt aggregates is also evidenced in the XRD analysis (Figure S7 of the Supporting Information) in which only the typical reflection peaks of C 3 N 4 are obtained.14,36,37

Figure 6. SEM pictures (top) of CM1B0.05 after catalysis (left, scale bar of 30 μm) and mapping of cobalt according to EDX (right, scale bar of 30 μm). TEM pictures (bottom) of CM1B0.05 before (left, scale bar of 100 nm) and after (right, scale bar of 10 nm) catalysis.

domains) is indirectly evident by fluorescence lifetime measurements. The formation of a heterojunction is proposed to improve the efficiency of charge formation, resulting in high photocatalytic activity at a low barbituric acid concentration. In addition, we also show the successful replacement of platinum as a cocatalyst in photocatalytic hydrogen production by more sustainable transition metals. The system is visible light responsive and performs well without an additional photosensitizer or use of UV light. Co and Ni ions and complex S1 act as cocatalysts for CM1B0.05 for photocatalytic hydrogen generation. S1 shows the highest initial rate, while Co(BF4)2 performed better than NiAc2. Recycling experiments and ICP measurements with Co(BF4)2 and CM1B0.05 suggest a deposited catalytically active species rather than a homogeneous catalyst. SEM-EDX, TEM, and XRD analysis show no evidence of particle formation above a 5 nm diameter, further suggesting very small aggregates or a molecular catalytically active species. This work opens the possibility for facile synthetic modification and optimization of C3N4-type semiconductors and semiconductor heteroassemblies, using only earth-abundant and low-cost starting materials. In addition, the substitution of platinum as a cocatalyst with first row transition metals corroborates the importance of the further study of these systems. Lowering the absorption edge further into the visible range and improving the interaction between heterogeneous cocatalysts and the material should allow for progressive improvement of photocatalytic efficiencies.

■

■

ASSOCIATED CONTENT

S Supporting Information *

CONCLUSION In summary, we demonstrate the synthesis of a highly efficient carbon nitride photocatalyst for hydrogen production by a careful design of supramolecular complexes. The resulting CM1Bx−C3N4 demonstrates a roll-like morphology with different chemical properties. In addition, the insertion of barbituric acid leads to an increase in the optical density along with a red shift into the visible region. The in situ formation of heterojunctions (composed from different semiconducting

Experimental details, SEM images, nitrogen sorption isotherms, FTIR, XRD, and elemental analysis data, and the structure of the catalysts investigated in this study. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. F

dx.doi.org/10.1021/cm503258z | Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials

Article

Author Contributions

(28) Bachmann, C.; Probst, B.; Guttentag, M.; Alberto, R. Chem. Commun. 2014, 50, 6737. (29) Rausch, B.; Symes, M. D.; Cronin, L. J. Am. Chem. Soc. 2013, 135, 13656. (30) Probst, B.; Guttentag, M.; Rodenberg, A.; Hamm, P.; Alberto, R. Inorg. Chem. 2011, 50, 3404. (31) Varma, S.; Castillo, C. E.; Stoll, T.; Fortage, J.; Blackman, A. G.; Molton, F.; Deronzier, A.; Collomb, M. N. Phys. Chem. Chem. Phys. 2013, 15, 17544. (32) Guttentag, M.; Rodenberg, A.; Bachmann, C.; Senn, A.; Hamm, P.; Alberto, R. Dalton Trans. 2013, 42, 334. (33) Dong, J. F.; Wang, M.; Li, X. Q.; Chen, L.; He, Y.; Sun, L. C. ChemSusChem 2012, 5, 2133. (34) Cao, S. W.; Liu, X. F.; Yuan, Y. P.; Zhang, Z. Y.; Fang, J.; Loo, S. C. J.; Barber, J.; Sum, T. C.; Xue, C. Phys. Chem. Chem. Phys. 2013, 15, 18363. (35) X, L.; Ward, A. J.; Masters, A. F.; Maschmeyer, T. Chem.Eur. J. 2014, 20, 1. (36) Zhang, J. S.; Grzelczak, M.; Hou, Y. D.; Maeda, K.; Domen, K.; Fu, X. Z.; Antonietti, M.; Wang, X. C. Chem. Sci. 2012, 3, 443. (37) Jin, J. T.; Fu, X. G.; Liu, Q.; Zhang, J. Y. J. Mater. Chem. 2013, 1, 10538.

M.S. and M.G. contributed equally to this work. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS M.S. is grateful for the financial support of the “Minerva Fellowship”. M.G. is grateful for a postdoctoral grant from the Swiss National Foundation. A.L. thanks MINECO (CTQ2013-49075-R) and “La Caixa Foundation” for financial support.

■

REFERENCES

(1) Gray, H. B. Abstracts of Papers; American Chemical Society: Washington, DC, 2013; Vol. 246, p 1. (2) Maeda, K.; Domen, K. J. Phys. Chem. Lett. 2010, 1, 2655. (3) Scharber, M. C.; Wuhlbacher, D.; Koppe, M.; Denk, P.; Waldauf, C.; Heeger, A. J.; Brabec, C. L. Adv. Mater. 2006, 18, 789. (4) Konenkamp, R.; Word, R. C.; Godinez, M. Nano Lett. 2005, 5, 2005. (5) Xu, H.; Xu, Y. G.; Li, H. M.; Xia, J. X.; Xiong, J.; Yin, S.; Huang, C. J.; Wan, H. L. Dalton Trans. 2012, 41, 3387. (6) Batzill, M. Energy Environ. Sci. 2011, 4, 3275. (7) Ran, J.; Zhang, J.; Yu, J.; Liu, J.; Jaroniec, M.; Qiao, S. Z. Chem. Soc. Rev. 2014, DOI: 10.1039/c3cs60425j. (8) Yan, H. J.; Yang, J. H.; Ma, G. J.; Wu, G. P.; Zong, X.; Lei, Z. B.; Shi, J. Y.; Li, C. J. Catalysis 2009, 266, 165. (9) Kudo, A.; Miseki, Y. Chem. Soc. Rev. 2009, 38, 253. (10) Wang, X. C.; Maeda, K.; Thomas, A.; Takanabe, K.; Xin, G.; Carlsson, J. M.; Domen, K.; Antonietti, M. Nat. Mater. 2009, 8, 76. (11) Shalom, M.; Gimenez, S.; Schipper, F.; Herraiz-Cardona, I.; Bisquert, J.; Antonietti, M. Angew. Chem., Int. Ed. 2014, 53, 3654. (12) Zheng, Y.; Jiao, Y.; Chen, J.; Liu, J.; Liang, J.; Du, A.; Zhang, W. M.; Zhu, Z. H.; Smith, S. C.; Jaroniec, M.; Lu, G. Q.; Qiao, S. Z. J. Am. Chem. Soc. 2011, 133, 20116. (13) Zheng, Y.; Liu, J.; Liang, J.; Jaroniec, M.; Qiao, S. Z. Energy Environ. Sci. 2012, 5, 6717. (14) Shalom, M.; Inal, S.; Fettkenhauer, C.; Neher, D.; Antonietti, M. J. Am. Chem. Soc. 2013, 135, 7118. (15) Zhang, Y. J.; Mori, T.; Ye, J. H.; Antonietti, M. J. Am. Chem. Soc. 2010, 132, 6294. (16) Young-Si, J.; Eun Zoo, L.; Xinchen, W.; Won, H.; Stucky, G. D.; Thomas, A. Adv. Funct. Mater. 2013, 23, 3661. (17) Aida, T.; Meijer, E. W.; Stupp, S. I. Science 2012, 335, 813. (18) Arrachart, G.; Carcel, C.; Trens, P.; Moreau, J. J. E.; Man, M. W. C. Chem.Eur. J. 2009, 15, 6279. (19) Schwarze, M.; Stellmach, D.; Schroder, M.; Kailasam, K.; Reske, R.; Thomas, A.; Schomacker, R. Phys. Chem. Chem. Phys. 2013, 15, 3466. (20) Serpone, N.; Salinaro, A. Pure Appl. Chem. 1999, 71, 303. (21) Guttentag, M.; Rodenberg, A.; Bachmann, C.; Senn, A.; Hamm, P.; Alberto, R. Dalton Trans. 2013, 42, 334. (22) Varma, S.; Castillo, C. E.; Stoll, T.; Fortage, J.; Blackman, A. G.; Molton, F.; Deronzier, A.; Collomb, M. N. Phys. Chem. Chem. Phys. 2013, 15, 17544. (23) Probst, B.; Rodenberg, A.; Guttentag, M.; Hamm, P.; Alberto, R. Inorg. Chem. 2011, 50, 3404. (24) Jurgens, B.; Irran, E.; Senker, J.; Kroll, P.; Muller, H.; Schnick, W. J. Am. Chem. Soc. 2003, 125, 10288. (25) Zhang, J. S.; Chen, X. F.; Takanabe, K.; Maeda, K.; Domen, K.; Epping, J. D.; Fu, X. Z.; Antonietti, M.; Wang, X. C. Angew. Chem., Int. Ed. 2010, 49, 441. (26) Lakowica, J. R. Principles of Fluorescence Spectroscopy, 3rd ed.; Springer: 2009. (27) Guttentag, M.; Rodenberg, A.; Kopelent, R.; Probst, B.; Buchwalder, C.; Brandstätter, M.; Hamm, P.; Alberto, R. Eur. J. Inorg. Chem. 2012, 1, 59. G

dx.doi.org/10.1021/cm503258z | Chem. Mater. XXXX, XXX, XXX−XXX