CO2 Preactivation in Photoinduced Reduction via Surface

Jan 17, 2013 - Salicylate and salicylic acid derivatives act as electron donors via charge-transfer complexes when adsorbed on semiconducting surfaces...
0 downloads 0 Views 1MB Size
Download PDF
Letter pubs.acs.org/JPCL

CO2 Preactivation in Photoinduced Reduction via Surface Functionalization of TiO2 Nanoparticles Daniel Finkelstein-Shapiro,†,‡ Sarah Hurst Petrosko,§,▽ Nada M. Dimitrijevic,§,#,‡ David Gosztola,§ Kimberly A. Gray,*,⊥,‡ Tijana Rajh,*,§ Pilarisetty Tarakeshwar,∥ and Vladimiro Mujica*,∥,†,§ †

Department of Chemistry, Northwestern University, 2145 Sheridan Road, Evanston, Illinois 60208, United States Institute for Catalysis in Energy Processes, Northwestern University, 2145 Sheridan Road, Evanston, Illinois 60208, United States § Center for Nanoscale Materials, Argonne National Laboratory, 9700 South Cass Avenue, Argonne, Illinois 60439, United States ∥ Department of Chemistry and Biochemistry, Arizona State University, Physical Sciences Building, Room D-102, P.O. Box 871604, Tempe, Arizona 85287, United States ⊥ Department of Civil and Environmental Engineering, Northwestern University, 2145 Sheridan Road, Evanston, Illinois 60208, United States # Chemical Sciences and Engineering Division, Argonne National Laboratory, 9700 South Cass Avenue, Argonne, Illinois 60439, United States ‡

S Supporting Information *

ABSTRACT: Salicylate and salicylic acid derivatives act as electron donors via chargetransfer complexes when adsorbed on semiconducting surfaces. When photoexcited, charge is injected into the conduction band directly from their highest occupied molecular orbital (HOMO) without needing mediation by the lowest unoccupied molecular orbital (LUMO). In this study, we successfully induce the chemical participation of carbon dioxide in a charge transfer state using 3-aminosalicylic acid (3ASA). We determine the geometry of CO2 using a combination of ultraviolet−visible spectroscopy (UV−vis), surface enhanced Raman scattering (SERS), 13C NMR, and electron paramagnetic resonance (EPR). We find CO2 binds on Ti sites in a carbonate form and discern via EPR a surface Ti-centered radical in the vicinity of CO2, suggesting successful charge transfer from the sensitizer to the neighboring site of CO2. This study opens the possibility of analyzing the structural and electronic properties of the anchoring sites for CO2 on semiconducting surfaces and proposes a set of tools and experiments to do so. SECTION: Energy Conversion and Storage; Energy and Charge Transport

T

(SERS) through the chemical effect.8−12 This opens the possibility of probing surface molecular species spectroscopically to follow the details of a reaction or to detect a molecule of interest. Amine groups can bind CO2 to form stable carbamates13 and are thus a means for increasing the affinity of CO2 toward a surface.14−19 In nature, the lysine group of the enzyme RuBisCO is activated by intake of a CO2 molecule and formation of a carbamate.20 Amines grafted onto photocatalytic oxide substrates can efficiently adsorb CO2, but are degraded by the irradiated photocatalytic substrate and cannot convert CO2 to higher energy products.19 Recently, a pyridine p-GaP photoelectrochemical cell was shown to convert CO2 to methanol with high efficiencies.18 If we are to use amines tethered to metal oxide semiconductors, increased stability of

iO2 photocatalyzes the conversion of CO2 to methane but suffers from a limited response to the solar spectrum because its band gap lies in the ultraviolet.1 Dyes can be used to extend its absorption range into the visible and add functionality by providing the possibility for tailored binding sites.2,3 Two mechanisms exist for charge injection from an adsorbed molecule into the semiconductor conduction band (CB): (i) excitation from the highest occupied molecular orbital (HOMO) to the lowest unoccupied molecular orbital (LUMO) of the molecule with subsequent injection into the CB, and (ii) direct excitation from the HOMO to the CB, that is, excitation of a charge transfer (CT) state. The first mechanism has been extensively studied in dye-sensitized solar cells while the second is just beginning to receive attention for its use in sensing and photocatalysis.4−7 Metal oxide nanoparticles functionalized with salicylic acid and salicylic acid derivatives represent examples of bioinorganic complexes with CT states. In these systems, the hydroxyl oxygens coordinate in a bidentate form to an undercoordinated Ti atom at the surface. Molecules that form a CT state with semiconductors also show surface enhanced Raman scattering © 2013 American Chemical Society

Received: December 7, 2012 Accepted: January 17, 2013 Published: January 17, 2013 475

dx.doi.org/10.1021/jz3020327 | J. Phys. Chem. Lett. 2013, 4, 475−479

The Journal of Physical Chemistry Letters

Letter

CO2 through a solution of pure 3ASA results in a precipitation of the compound, which is negligibly soluble in water below pH 7. Bubbling N2 instead of CO2 through the solution of TiO2− 3ASA complexes did not result in a spectral redshift. The behavior of structurally similar 4ASA and 5ASA is interesting. 5ASA forms a charge transfer complex with TiO2 nanoparticles, similar to that observed with 3ASA and, upon bubbling CO2, a red-shift is observed and the functionalized nanoparticles precipitate out of solution. Addition of 4ASA to a solution of TiO2 NPs, however, does not form a strong CT state. We focus on the case of TiO2−3ASA, because it is the conjugate with the strongest response and because it remains in colloidal suspension. We aimed to determine the structure of the complex formed upon CO2 intake. We measured the SERS signal for TiO2− 3ASA (1:50) before and after CO2 adsorption (Figure 2, InVia

the grafted amine or an alternate charge delivery mechanism that will not degrade the surface functional group is needed. In this Letter, we report on the chemical involvement of CO2 with 3-aminosalycilic acid (3-ASA, compound 1) on the surface of TiO2 nanoparticles via a CT state. We present a promising avenue for extending the response of TiO2 into the visible and for activating CO2 for photocatalytic reduction. (1) 3-Aminosalicylic Acid (3ASA); (2) 4-Aminosalicylic Acid (4ASA); (3) 5-Aminosalicylic Acid (5ASA); (4) Salicylic Acid (SA)

TiO2 anatase nanoparticles (TiO2 NPs, ∼ 2 (5) nm in diameter) were synthesized via the hydrolysis of titanium isopropoxide (titanium tetrachloride under nitrogen) (Sigma Aldrich) at pH = 11 (pH = 3.5). Addition of appropriate amounts of the adsorbate to these particles resulted in a CT state which red-shifted the photoresponse. This was accompanied by an enhancement of the Raman scattering of the adsorbed molecules due to the surface (SERS). Bubbling CO2 through the solution containing TiO2−3ASA conjugates resulted in a further red-shift. Figure 1 shows the change in Figure 2. SERS spectrum of TiO2−3ASA with and without CO2 (traces have been vertically offset for clarity).

Renishaw Microscope, λexc = 442 nm). Vibrational modes are particularly sensitive to changes in molecular structure. In addition, SERS on semiconductors is mostly observed for those modes that facilitate interfacial charge transfer. Monitoring the vibrational changes upon CO2 adsorption reveals modes that are involved in electron injection interactions and could be beneficial for photocatalysis, for example.21,22 On the basis of known amine chemistry, we expected to see the deprotonation of the amine with subsequent formation of a carbamate. We collected spectra for the TiO2−3ASA system in the presence and absence of CO2, with 3ASA containing a deuterated amine, at different pHs, with 13CO2, and upon addition of sodium bicarbonate (see SI). Table 1 lists the most

Figure 1. UV−vis of bare TiO2 NPs, TiO2−3ASA and TiO2−3ASA +CO2 (1:50 coverage) at pH 6.6.

absorption for a surface coverage of 1:50 TiO2 NPs:3ASA (4.0 × 1014 molecules/cm2). We can clearly see the redshift caused by the introduction of 3ASA and a further redshift resulting from bubbling CO2. Bubbling CO2 has two effects: (1) the pH of the solution decreases, increasing the adsorption of the molecules to the surface, and (2) CO2 adsorbs onto the complex, modifying its electronic properties. The pH of the solution after functionalization with a surface coverage of 1:50 is pH = 9. At this pH, less than half of the 3ASA is bound to the NP, with full binding occurring below pH = 7 (see Supporting Information (SI)). Bubbling of CO2 lowers the pH of the solution to 6.6 through the formation of a carbonic acid buffer, causing more 3ASA to adsorb to the nanoparticle surface. We thus adjusted the pH of the samples without CO2 to a pH of 6.6 with HCl. Controls run on salicylic acid (compound 2) and bare NPs showed no redshift upon bubbling with CO2 (see SI). Bubbling

Table 1. Raman Vibrations for 3ASA−TiO2 in Different Environments

476

without CO2 (cm−1)

with CO2 (cm−1)

deuterated (cm−1)

deuterated with CO2 (cm−1)

theory (without CO2 see SI) (cm−1)

1200

1200

1200

1200

1182

1216

1212

1206

1210

1217

1237

1237

1241

1241

1251

1254

1258

1252

1252

1332 1352 1371

1328

1330

1331

1352 1371

1379

1374

1386

assignment νCC; A′(8b) νasCH; A′(13) δCH; A′(3) νCC; A′(14)

νCOO−

dx.doi.org/10.1021/jz3020327 | J. Phys. Chem. Lett. 2013, 4, 475−479

The Journal of Physical Chemistry Letters

Letter

intense modes. Because these are skeletal modes of the aromatic ring, the shifts are smaller than if we were monitoring N−H bands; however, all measurements were repeated several times resulting in consistent shifts every time. Because of the difficulty in ascertaining which modes involve vibrations of the N−H bonds, we selectively deuterate 3ASA at the amine position (SI), revealing modes 1216, 1254, 1332, and 1371 as the clearest indicators of the binding of this functional group. We observe that in all experiments (3ASA, 3ASA+CO2, deuterated 3ASA, deuterated 3ASA+CO2; see SI for traces) the addition of CO2 causes a decrease in the effective mass of the nitrogen group. We conclude from this analysis that instead of the formation of a carbamate N−C bond, we have a partial or complete deprotonation of the amine by an oxygen atom of CO2. If CO2 interacts indirectly and weakly with the N atom, we expect the surface to play a key role in binding the CO2. To confirm the presence of surface-bound CO2, we carried out 13C NMR (500 MHz Bruker). Two sets of 2 nm nanoparticles in D2O, one unfunctionalized and one functionalized with 3ASA, were bubbled with 13CO2 for 6 min. Forty microliters of ethanol were added to 460 μL of the bubbled solution as an internal standard, which renders an adsorbate density of approximately 30 CO2 molecules/NP. In both cases, two new distinct peaks were observed compared to the case where CO2 was absent: a feature at 160.85 PPM, and a feature at 125.29 PPM. Bubbling N2 after CO2 eliminates the signal at 125.3 PPM but does not affect the intensity of the peak at 160.9 PPM (even though the pH is raised to 8), and so we conclude that it is this peak that corresponds to strongly bound CO2. The other peak, at 125 PPM, has been previously assigned to linear, unbound CO2.16 The 13C NMR shift observed for both bare NPs and TiO2−3ASA is consistent with a carbonic acid form,23 and the absence of chemical shift for the 13C resonance between a bare NP and an NP with the 3ASA linker excludes a direct bonding between the amine and the carbon, in contrast with other reports of CO2 adsorption on surfaces with grafted amines. For solid-state MAS NMR of gaseous CO2 adsorbed on amine-grafted zeolites a carbamate shift has been measured and calculated at 160 PPM.13 This binding is unexpected; it is also desirable because it places the CO2 in a more favorable position for electron transfer from the nanoparticle. We now turn to electron paramagnetic resonance (EPR) measurements to map out the electron dynamics. X-band EPR measurements (Bruker Elexsys E580 spectrometer equipped with an Oxford CF935 helium flow cryostat with an ITC-5025 temperature controller) were conducted to determine the nature of the excited state and to understand its interaction with CO2. Light induced CW-EPR on 3ASATiO2 (Figure 3) shows a strong cation radical signal centered on the aromatic ring with hyperfine contributions from the protons on the ring as well as the nitrogen (SI). We can also observe lattice bound electrons from their parallel and perpendicular component. The electron parallel component is more sensitive to the electron environment and indicates the energy of the trapping site. Upon irradiation, we observe the presence of two components: near-surface shallow trapping sites (g = 1.959) and a motionally narrowed partially delocalized component (g = 1.963). Because the energy of the state corresponding to g = 1.963 is close to the energy of the conduction band, an ambient temperature at 4 K (0.3 meV)

Figure 3. CW-EPR of 3ASA-TiO2 during and after illumination (top) and of 3ASA−TiO2 and illuminated 3ASA−TiO2−CO2 during illumination. Insets show the parallel component of the lattice electrons. See text for details.

is sufficient to excite electrons to the conduction band, partially delocalizing them and participating in the narrowing of the signal. When the irradiation is stopped, we see a depletion of the mobile electrons while the near-surface trapped electrons survive. This behavior is expected given the electrons contributing to the motionally narrowed signal can move at 4 K and recombine with photogenerated holes trapped at the sensitizer. The same experiment carried out in the presence of CO2 shows a different behavior. In the presence of CO2, we see that the signal of holes localized on 3ASA changes, showing a loss in the hyperfine splitting attributed to the N atom (see SI). We attribute the loss of the hyperfine contribution to a disappearance of spin density at the N position. This could reflect the partial deprotonation of the amine due to its interaction with CO2, which could also be enhanced in the excited state. In the presence of CO2, even during irradiation, the signal from the shallow trapped electrons has disappeared. This type of behavior could arise because shallow trapped electrons that are formed in the near-surface region can react with adsorbed molecules. When irradiation is stopped, both components disappear (not shown). To follow more explicitly the electron dynamics, timeresolved EPR was carried out on the perpendicular component of the electron spin. The electron resonance was monitored while the light was turned on (Figure 4). Without CO2, we

Figure 4. Light ON (top) and light OFF (bottom) experiments in the absence and presence of CO2. 477

dx.doi.org/10.1021/jz3020327 | J. Phys. Chem. Lett. 2013, 4, 475−479

The Journal of Physical Chemistry Letters

Letter

to map the structure and dynamics that occur when CO2 interacts with a metal oxide, (ii) we have detected a defect surface state created by the adsorption of CO2 that is expected to play an important role in charge transfer and chemical activation, (iii) we have synthesized a system that can potentially undergo a proton-coupled electron transfer whereby an electron is photoexcited into the semiconductor along with a deprotonation of the amine. These concepts and methods can readily be extended to other oxides: future work will also look to calculate and measure how defect sites created by CO2 adsorption change as a function of oxide surface and auxiliary ligand and how to increase the number of electrons available for activation. Having demonstrated that a close proximity and interaction between the sensitizer and the substrate is possible, we would want to use these to alter the thermodynamics of charge transfer to CO2 to make higher energy products. Achieving this overarching goal is a notoriously difficult problem, one to whose solution our investigation can prove to be of substantial value.

observe the expected build-up of electrons in the lattice. Addition of CO2 results in a build-up of charge with a subsequent depletion of the lattice electron. We attribute the depletion of the excited state to electron transfer toward the CO2. The idea of an additional pathway for electron transfer is further backed-up by a faster disappearance for the lattice electron in the presence of CO2 when the light was turned off. (Figure 4) In order to allow a higher yield of electron transfer, these materials were irradiated at 77 K for 40 min (Figure 5). At this

■

ASSOCIATED CONTENT

* Supporting Information S

Experimental section, Raman spectra of 4- and 5ASA. UV−vis for SA-TiO2, and bare NP with and without CO2. Binding of 3ASA on TiO2 at different pHs. Fluorescence of TiO2-3ASA at different pHs. Raman of isotopically labeled 13CO2 on TiO23ASA. EPR simulations. This material is available free of charge via the Internet at http://pubs.acs.org.

Figure 5. CW-EPR of TiO2−3ASA with and without CO2 irradiated at 77 K for 40 min (top). Binding geometry and electronic levels of the states involved in the system (bottom).

temperature, lattice electrons have enough energy (6.3 meV) to diffuse to the surface and fill deep trapping sites, which would result in a very broad signal that cannot be observed. Interestingly, when 3ASA−TiO2 was illuminated in the presence of CO2, a state with g = 1.956 was observed and was found to lack a 12C/13C isotope effect. We assign this species to a surface Ti-centered electron coupled to CO2, as similar signals have been measured for carboxyl group chelated Ti(III) ions. The Ti involved most likely exists in a tetragonally distorted octahedral field.24−26 We believe this state can be resolved by EPR because the CO2 changes the symmetry of the surface states while lowering the energy of the electron generating an electron sink.27 The experiments indicate that upon population of the conduction band by photoexcited electrons, a fraction of them migrate to the surface to a defect site generated by the CO2. We summarize the geometry of the adsorbed CO2 along with its corresponding energy levels in Figure 5. Our group, and others, have shown that the reduction of CO2 on TiO2 is possible and occurs coupled to water oxidation.1,27−29 The stable products of this photoreduction require at least two electrons. For our functionalized material to be active for the full conversion of CO2 to hydrocarbons, a renewable proton source and a pathway to reduce the oxidized sensitizer are needed, preferably coming from a water oxidation reaction. This direction of research, analyzing the chemistry of sacrificial electron donors, including water, with catechol and salicylic acid derivatives to increase the availability of electrons is currently being pursued to complete the photocatalytic circuit. Sensitized photocatalysts for CO2 reduction require the transfer of charge from the sensitizer to the substrate mediated by the conduction band of the semiconductor. In this communication, (i) we propose a combination of theoretical analysis with various experimental techniques, i.e., SERS, ultraviolet−visible spectroscopy (UV−vis), 13NMR, and EPR

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected] (V.M.); [email protected] (T.R.); [email protected] (K.A.G.). Present Address ▽

Department of Chemistry and International Institute of Nanotechnology, Northwestern University, 2145 Sheridan Rd., Evanston, IL 60208, U.S.A. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Dr. Alon Danon for useful discussions and current literature. S.J.H.-P. acknowledges Argonne National Laboratory for a Director’s Postdoctoral Fellowship. This work was performed under the auspices of Argonne National Laboratories (Center for Nanomaterials, User Proposal 23792), the U.S. Department of Energy, under Contracts DEFG02-03 ER 15457/A003 and DE-AC02-06CH11357 (ICEP) and NSF CHE-1124895. This work made use of the IMSERC NMR facility at Northwestern University (NSF CHE-9871268 (1998)).

■

REFERENCES

(1) Dimitrijevic, N. M.; Vijayan, B. K.; Poluektov, O. G.; Rajh, T.; Gray, K. A.; He, H.; Zapol, P. Role of Water and Carbonates in Photocatalytic Transformation of CO2 to CH4 on Titania. J. Am. Chem. Soc. 2011, 133, 3964−3971. (2) Woolerton, T. W.; Sheard, S.; Reisner, E.; Pierce, E.; Ragsdale, S. W.; Armstrong, F. A. Efficient and Clean Photoreduction of CO2 to CO by Enzyme-Modified TiO2 Nanoparticles Using Visible Light. J. Am. Chem. Soc. 2010, 132, 2132−2133. (3) Anfuso, C. L.; Snoeberger, R. C.; Ricks, A. M.; Liu, W.; Xiao, D.; Batista, V. S.; Lian, T. Covalent Attachment of a Rhenium Bipyridyl 478

dx.doi.org/10.1021/jz3020327 | J. Phys. Chem. Lett. 2013, 4, 475−479

The Journal of Physical Chemistry Letters

Letter

CO2 Reduction Catalyst to Rutile TiO2. J. Am. Chem. Soc. 2011, 133, 6922−6925. (4) Hurst, S. J.; Fry, H. C.; Gosztola, D. J.; Rajh, T. Utilizing Chemical Raman Enhancement: A Route for Metal Oxide SupportBased Biodetection. J. Phys. Chem. C 2011, 115, 620−630. (5) Syres, K.; Thomas, A.; Bondino, F.; Malvestuto, M.; Grätzel, M. Dopamine Adsorption on Anatase TiO2(101): A Photoemission and NEXAFS Spectroscopy Study. Langmuir 2010, 26, 14548−14555. (6) Calzolari, A.; Ruini, A.; Catellani, A. Anchor Group versus Conjugation: Toward the Gap-State Engineering of Functionalized ZnO(1010̅ ) Surface for Optoelectronic Applications. J. Am. Chem. Soc. 2011, 133, 5893−5899. (7) Agrios, A. G.; Gray, K. A.; Weitz, E. Narrow-Band Irradiation of a Homologous Series of Chlorophenols on TiO2: Charge-Transfer Complex Formation and Reactivity. Langmuir 2004, 20, 5911−5917. (8) Musumeci, A.; Gosztola, D.; Schiller, T.; Dimitrijevic, N. M.; Mujica, V.; Martin, D.; Rajh, T. SERS of Semiconducting Nanoparticles (TiO2 Hybrid Composites). J. Am. Chem. Soc. 2009, 131, 6040−6041. (9) Rajh, T.; Saponjic, Z.; Liu, J.; Dimitrijevic, N. M.; Scherer, N. F.; Vega-Arroyo, M.; Zapol, P.; Curtiss, L. A.; Thurnauer, M. C. Charge Transfer Across the Nanocrystalline-DNA Interface: Probing DNA Recognition. Nano Lett. 2004, 4, 1017−1023. (10) Tarakeshwar, P.; Finkelstein-Shapiro, D.; Hurst, S. J.; Rajh, T.; Mujica, V. Raman Investigation of Nanosized TiO2: Effect of Crystallite Size and Quantum Confinement. J. Phys. Chem. C 2011, 115, 8994−8797. (11) Quagliano, L. G. Observation of Molecules Adsorbed on III-V Semiconductor Quantum Dots by Surface-Enhanced Raman Scattering. J. Am. Chem. Soc. 2004, 126, 7393−7398. (12) Sun, Z.; Zhao, B.; Lombardi, J. R. ZnO Nanoparticle SizeDependent Excitation of Surface Raman Signal from Adsorbed Molecules: Observation of a Charge-Transfer Resonance. Appl. Phys. Lett. 2007, 91, 221106−221109. (13) Danon, A.; Stair, P. C.; Weitz, E. FTIR Study of CO2 Adsorption on Amine-Grafted SBA-15: Elucidation of Adsorbed Species. J. Phys. Chem. C 2011, 115, 11540−11549. (14) Leal, O.; Bolívar, C.; Ovalles, C.; García, J. J.; Espidel, Y. Reversible Adsorption of Carbon Dioxide on Amine Surface-Bonded Silica Gel. Inorg. Chim. Acta 1995, 240, 183−189. (15) Aaron, D.; Tsouris, C. Separation of CO2 from Flue Gas: A Review. Sep. Sci. Technol. 2005, 40, 321−348. (16) Pinto, M. s. L.; Mafra, L. s.; Guil, J. M.; Pires, J.; Rocha. Adsorption and Activation of CO2 by Amine-Modified Nanoporous Materials Studied by Solid-State NMR and 13CO2 Adsorption. J. Chem. Mater. 2011, 23, 1387−1395. (17) Barton Cole, E.; Lakkaraju, P. S.; Rampulla, D. M.; Morris, A. J.; Abelev, E.; Bocarsly, A. B. Using a One-Electron Shuttle for the Multielectron Reduction of CO2 to Methanol: Kinetic, Mechanistic, and Structural Insight. J. Am. Chem. Soc. 2010, 132, 11539−11551. (18) Barton, E. E.; Rampulla, D. M.; Bocarsly, A. B. Selective SolarDriven Reduction of CO2 to Methanol Using a Catalyzed p-GaP Based Photoelectrochemical Cell. J. Am. Chem. Soc. 2008, 130, 6342−6344. (19) Tseng, C.-L.; Chen, Y.-K.; Wang, S.-H.; Peng, Z.-W.; Lin, J.-L. 2Ethanolamine on TiO2 Investigated by in Situ Infrared Spectroscopy. Adsorption, Photochemistry, and Its Interaction with CO2. J. Phys. Chem. C 2010, 114, 11835−11843. (20) Blankenship, R. E. Molecular Mechanisms of Photosynthesis; Blackwell Science: Oxford, U.K., 2002. (21) Lombardi, J. R.; Birke, R. L. A Unified Approach to SurfaceEnhanced Raman Spectroscopy. J. Phys. Chem. C 2008, 112, 5605− 5617. (22) Lombardi, J. R.; Birke, R. L.; Lu, T.; Xu, J. Charge-Transfer Theory of Surface Enhanced Raman Spectroscopy: Herzberg−Teller Contributions. J. Chem. Phys. 1986, 84, 4174−4180. (23) Mani, F.; Peruzzini, M.; Stoppioni, P. CO2 Absorption by Aqueous NH3 Solutions: Speciation of Ammonium Carbamate, Bicarbonate and Carbonate by a 13C NMR Study. Green Chem. 2006, 8, 995−1000.

(24) Watanabe, T.; Fujiwara, S. EPR Studies of Ti(III) Chelates in Aqueous Solution; The Nature of Chemical Bonding and EPR Relaxation Mechanism. J. Magn. Reson. 1970, 2, 103−113. (25) Fujiwara, S.; Nagashima, K.; Codell, M. Mixed Chelate Compounds of Titanium(III) as Studied by Electron Paramagnetic Resonance and Spectrophotometry. Bull. Chem. Soc. Jpn. 1964, 37, 773−779. (26) Rajh, T.; Ostafin, A. E.; Micic, O. I.; Tiede, D. M.; Thurnauer, M. C. J. Phys. Chem. 1996, 100, 4538−4545. (27) Anpo, M.; Yamashita, H.; Ichihashi, Y.; Ehara, S. Photocatalytic Reduction of CO2 with H2O on Various Titanium Oxide Catalysts. J. Electroanal. Chem. 1995, 396, 21−26. (28) Pan, J.; Wu, X.; Wang, L.; Liu, G.; Lu, G. Q.; Cheng, H.-M. Synthesis of Anatase TiO2 Rods with Dominant Reactive {010} Facets for the Photoreduction of CO2 to CH4 and Use in Dye-Sensitized Solar Cells. Chem. Commun. 2011, 47, 8361−8363. (29) Liu, L.; Zhao, H.; Andino, J. M.; Li, Y. Photocatalytic CO2 Reduction with H2O on TiO2 Nanocrystals: Comparison of Anatase, Rutile, and Brookite Polymorphs and Exploration of Surface Chemistry. ACS Catal. 2012, 2, 1817−1828.

479

dx.doi.org/10.1021/jz3020327 | J. Phys. Chem. Lett. 2013, 4, 475−479