Ab Initio Simulations of Copper Oxide Nanowires and Clusters on

Article pubs.acs.org/JPCC

Ab Initio Simulations of Copper Oxide Nanowires and Clusters on TiO2 (101) Anatase Surface Victor Meng’wa,†,‡ Nicholas Makau,† George Amolo,§ Sandro Scandolo,‡ and Nicola Seriani*,‡ †

Department of Physics, University of Eldoret, P.O Box 1125-30100, Eldoret, Kenya The Abdus Salam International Centre for Theoretical Physics, Condensed Matter and Statistical Physics Section, Strada Costiera 11, 34151 Trieste, Italy § Department of Physics and Space Science, The Technical University of Kenya, P.O Box 52428-00200, Nairobi, Kenya ‡

S Supporting Information *

ABSTRACT: Copper oxides deposited at titania surfaces have a beneficial effect on the photocatalytic activity of TiO2, but their role is not fully understood. In this work, possible nanostructures of copper oxide on TiO2 (101) have been investigated by simulations based on density functional theory. Various stoichiometries, from Cu2O to CuO, and morphologies, from clusters to nanowires, have been considered. Nanowire structures were consistently more stable than isolated clusters. In these structures, a Cu2O stoichiometry was found to be thermodynamically more stable than CuO at room conditions, in contrast to what happens in bulk copper. Occupied Cu 3d and O 2p states extend well into the band gap of titania, whereas the nature of the lowest-lying empty states depend on the stoichiometry: for Cu2O they consist mostly of Ti 3d orbitals, while in CuO unoccupied Cu 3d orbital ∼0.8 eV above the Fermi level are present. Thus, both oxides reduce the band gap of the system with respect to pure titania, but only Cu2O should be effective in separating photogenerated electrons and holes. These results provide insight into the role of copper oxides in the photocatalytic process.

■

INTRODUCTION Titanium dioxide, TiO2, is considered the most suitable semiconductor photocatalyst due to its powerful oxidation ability, superior charge transfer characteristics, long-term stability against photochemical corrosion, nontoxicity, high availability, and low production cost.1 It has therefore found a wide range of applications such as photocatalytic water splitting, CO2 photoreduction, and in the degradation of organic pollutants and in dye-sensitized solar cells, among others.2−4 However, the photocatalytic efficiency of TiO2 is limited by the high recombination rate of electron−hole pairs and weak photoabsorption in the visible range. For this reason, great efforts have been undertaken to modify titania, for example through doping or through addition of metals and semiconductors,5 resulting in remarkable improvement of the photocatalytic activity. Modifications can achieve this through different mechanisms, by extending light absorption to the visible region, by reducing the recombination of photogenerated electron−hole pairs, or by enhancing charge separation. Modification through copper is considered very promising, also because of the narrow band gap of its oxides,6 that could increase photoabsorption and, possibly, charge separation. Kumar and co-workers7 reported that titania-based photocatalysts containing copper in different states (Cu, Cu2O, and CuO) showed improved catalytic efficiency, and attributed the © 2017 American Chemical Society

enhancement to better absorption of visible light and to enhanced suppression of electron−hole recombination at the surface. It remains however unclear which copper phase is most active, how the different phases interact with titania, and how this affects the behavior of the overall system. In fact, both copper oxides are active photocatalysts by themselves. Cuprous oxide (Cu2O) is a p-type semiconductor with a direct band gap of 2.0 eV, and itself active for photocatalytic hydrogen production.8,9 When added to TiO2, a significantly enhanced photocatalytic H2 production under sunlight was observed.10 Also, CuO is an intrinsic p-type semiconductor with a narrow band gap (1.2−1.9 eV)11 and has been used as a cocatalyst on TiO2 for photocatalytic applications.12 It has been argued that such a semiconductor could form a p−n heterojunction with an n-type semiconductor like TiO2. Moreover, the conduction band of CuO lies higher than the H+/H2 redox level and can thus participate in the hydrogen evolution reaction.13 Quantum confinement in CuO clusters should push the conduction band even higher.14 However, the interface between nanostructured CuO and TiO2 has not been exhaustively studied, particularly, regarding relative band positions and bandgap. Moreover, different values Received: July 7, 2017 Revised: August 28, 2017 Published: August 28, 2017 20359

DOI: 10.1021/acs.jpcc.7b06681 J. Phys. Chem. C 2017, 121, 20359−20365

Article

The Journal of Physical Chemistry C

the vacuum between the TiO2 slab and its replica is ∼14 Å thick. Cu2O and CuO nanowires and clusters were present on one side of the slab. All atomic positions were relaxed using conjugated-gradient method for energy minimization until the force on each atom was smaller than 10 −3 a.u.

for the conduction band position relative to vacuum are reported.15 This lack of characterization adds uncertainty as to which copper oxide phase is the most active for photocatalytic applications. Yamashita et al.16 reported that addition of Cu2+ into the TiO2 could improve production of methanol from CO2 photoreduction process. Tseng et al.17 also observed that the formation of methanol was found to be much more effective on Cu-titania catalysts. The XPS spectra revealed that the copper species were predominantly Cu+ and a small fraction of Cu0. It is therefore important to investigate and shed more light on which copper species (Cu0, Cu+, or Cu2+) is primarily responsible and desired for photocatalytic applications. Wu and Lee18 reported that deposited Cu particles, which are in the oxidized state during reaction, cause up to a 10-fold enhancement in H2 production. In contrast, Cu ions embedded in the TiO2 lattice exhibited a negative effect. Liu et al. reported that Cu+ species on the catalyst surface could trap electrons more efficiently than Cu2+ species due to its higher reduction potential.19 Although Cu oxides loaded on the TiO2 surface can be reduced to metallic Cu during photoreactions due to trapped electrons,20 Cu nanoparticles are easily oxidized to Cu2O and CuO upon exposure to air or traces of oxygen molecules. It is therefore necessary also to understand how the relative stability of different oxide species is modified when going to nanosized structures at a TiO2 surface. In the present study we have investigated small clusters and nanowires of copper oxides on the (101) surface of TiO2 in the anatase crystal by density functional theory. Several structures were optimized, with stoichiometries corresponding to Cu2O and CuO, to search for the most stable one. Their thermodynamic stability in contact with an oxygen gas was calculated by ab initio thermodynamics. The most stable structure at room conditions is a Cu2O nanowire, for which the top of the valence band consists of Cu 3d and O 2p states, while the bottom of the conduction band is given by states of titanium. We argue that this should be a favorable situation to enhance electron−hole separation in this system. On the contrary, in the most stable CuO structure, also a nanowire, the bottom of the conduction band is given by states of the copper oxides.

■

RESULTS AND DISCUSSION In this section, we first present the results of structural search and optimization, then we discuss the thermodynamic stability of the nanostructures, and finally, we analyze the electronic structure of the most stable configurations. Atomic Structures. We performed an extensive structural search of possible structures with Cu2O and CuO stoichiometries, with four copper atoms in the simulation cell. We have followed different strategies to produce initial configurations for the optimization algorithm. First, we created clusters and more elongated structures by extracting them from the bulk oxide structures. Second, we added an appropriate amount of oxygen to the tetrahedron Cu4 cluster that was found to be stable in previous work on metallic clusters,26 and that had been shown to have a low barrier for conversion of CO2 to methane.31 Third, building on the outcome of these relaxations, educated guesses for further possible structures were considered. The optimized geometries are shown in Figure 1 and Figure 2 for the Cu2O and CuO stoichiometries, respectively. The most stable structure is the Cu2O nanowire shown in Figure 1a, with an almost planar geometry. Interestingly, planar structures were also found to be favorable for copper clusters in vacuum.32 For each figure, also, the difference in total energy (ΔE) at zero temperature between each structure and the most stable one for each stoichiometry is reported. For the structures with the Cu2O stoichiometry, the stable nanowire structure lies lower than the second most stable structure by 1.15 eV, which is several times larger than kBT at room temperature. On the contrary, in the case of CuO, the second lowest structure is only 0.13 eV higher in energy. The stoichiometry of the copper oxide nanostructures is either Cu2O or CuO. However, their structure might differ considerably from those of the bulk oxides. To show this, in Tables 1 and 2 the average atomic distances are reported. For Cu2O structures, the range of Cu−Cu bonds are very close to that of experimental bulk metal copper of 2.54 Å, but more than Cu−Cu lengths of the dimer (Cu2) that ranged between 2.28 Å, and 2.31 Å.26 The Cu−O bond lengths increased and decreased in the range of 0.11 to −0.01 Å, respectively, compared to the experimental Cu2O bulk value of 1.85 Å, and longer than Cu−O length of 1.8 Å, from DFT+U calculations for bulk Cu2O.11 A Cu−Ti bond is found in the structures shown in Figure 1c−e, with bond lengths ranging from 2.62 to 2.79 Å, considerably shorter than the Cu−Ti bond previously found in metallic copper clusters on TiO2 (2.92 Å26). However, it should be noted that the most stable structures have a O−Ti5C rather than a Cu−Ti bond. Regarding coordination, most of the Cu atoms in Cu2O structures have their total coordination ranging between 4- and 5-fold, and only few have a 2-fold coordination like in the bulk. 3-Fold coordinated O atoms are most prevalent, which is also different from bulk Cu2O, where O is coordinated to 4 Cu atoms. In the CuO nanostructures, the Cu−Cu bond lengths are about 0.1 Å longer than those of experimental bulk copper metal of 2.54 Å; on the contrary, in the Cu2O the distances were similar to those in bulk metallic copper. The Cu−O bonds are in most cases slightly contracted in a range of −0.02 to

■

COMPUTATIONAL METHODS Spin polarized calculations were carried out in the framework of density functional theory, as implemented in the QuantumEspresso code,21 using a plane wave basis set. Vanderbilt ultrasoft pseudopotentials were used to describe core−valence interactions.22 The study employed DFT+U23 with the GGAPBE24 exchange-correlation functional and the Hubbard U correction in the version of Dudarev et al.23 The values of U were taken from the literature: a value of 4.2 eV was used for titanium25,26 and 5.2 eV for copper.26,27 Cutoffs of 40 and 480 Ry were employed for wave functions and charge density, respectively. A lattice constant of 4.32 Å was obtained for bulk cubic Cu2O (experiment: 4.26 Å28), with a Cu−O bond distance of 1.87 Å (experiment: 1.85 Å28); for monoclinic CuO, a Cu−O bond distance of 1.95 Å was calculated (experiment: 1.96 Å29). The copper oxide nanowires and clusters were deposited on the anatase-TiO2 (101) surface. The (101) surface was modeled with a rectangular (1 × 2) surface cell of size (10.27 Å × 7.57 Å). A k-point grid of (1 × 2 × 1) with the method of Monkhorst and Pack30 was used. The TiO2 slabs have a thickness of three layers, corresponding to ∼10 Å, and 20360


Article


Figure 2. Optimized CuO nanowire and clusters on the TiO2 (101) surface with their respective total energy differences referenced to the most stable stable structure: (a) ΔE = 0 eV; (b) ΔE = 0.13 eV; (c) ΔE = 0.49 eV; (d) ΔE = 1.23 eV. Color code as in Figure 1. The two shown CuO nanowires/clusters on TiO2 anatase (101) surface are replicas of each other.

Table 1. Average Bond Lengths in Optimized Structures of Cu2O/TiO2 Nanowire and Clusters Shown in Figure 1a

Figure 1. Optimized Cu2O nanowires and clusters on the TiO2 (101) surface with their respective total energy differences with respect to the most stable structure (a): (a) ΔE = 0 eV; (b) ΔE = 1.15 eV; (c) ΔE = 1.52 eV; (d) ΔE = 2.10 eV; (e) ΔE = 3.54 eV; blue balls, titanium; red balls, oxygen; red-brown balls, copper. The two shown Cu2O nanowires/clusters on TiO2 anatase (101) surface are replicas of each other.

nanowires/clusters

Cu−Cu

Cu−O

a b c d e

2.53 2.52 2.44 2.60 2.59

1.96 1.91 1.90 1.84 1.88

a

−0.01 Å. A Cu−Ti bond is present in one of the structures (Figure 2b) with a Cu−Ti distance of 2.78 Å. The O−Ti5C bond distances are in good agreement with O−Ti bond lengths seen in surface X-ray diffraction (SXRD) of anatase-TiO2 (101) surface that ranged between 1.89 and 2.03 Å,33 but for one value of 1.81 Å. The Cu−O2C bond, which is between Cu atom and 2-fold coordinated oxygen atom (O2C) on the TiO2 surface range between 1.89 and 2.13 Å, within ranges almost equivalent to O−Ti bond lengths in SXRD anatase-TiO2 (101) surface.33 A greater number of the Cu atoms in the CuO structures are 4-, 5-, and 6-fold coordinated. Except for one 4-fold coordinated Cu atom, nearly all Cu atoms are overcoordinated compared to Cu in bulk CuO. In contrast, almost all O atoms are undercoordinated with the 3-fold coordination O being the most frequent. The triangular frameworks of Cu−Cu−Cu and Cu−Cu−O are the most dominant patterns in Cu2O, in agreement with other cluster studies,34,35 and Cu3 units are significantly present in CuO nanowire and cluster formation.

Cu−Ti

2.79 2.68 2.62

O−Ti5C

Cu−O2C‑surf

2.02 1.82 1.76

1.92 1.88 1.98 1.90 1.90

All in units of Å.

Table 2. Average Bond Lengths in the Optimized Structures of CuO/TiO2 Nanowires and Clusters Shown in Figure 2a nanowires/clusters

Cu−Cu

Cu−O

a b c d

2.68 2.65 2.68 2.62

1.94 1.96 1.95 1.95

a

Cu−Ti 2.78

O−Ti5C

Cu−O2C‑surf

1.95 1.89 2.00 1.81

2.13 1.91 1.96 1.89

The bond lengths are in Å.

Additionally, the Cu3 triangular framework is present in almost all the Cu2O and CuO structures. We also considered the most stable 3-layer (3L; 10 Å × 7.5 Å) with vacuum 14 Å of relaxed Cu2O/TiO2 (see Figure 1a−c) and CuO/TiO2 structures and increased the supercell size to (10 Å × 15.15 Å) with vacuum of 18 Å. We made the bigger supercells to be 2-layers (2L) to reduce the computational 20361


Article

The Journal of Physical Chemistry C costs. We performed structural relaxation and three most stable optimized 2L Cu2O/TiO2 and CuO/TiO2 structures arranged in the order of decreasing stability are shown in Figures 3 and 4, respectively, SI. We also calculated formation energies of three most stable 3L and 2L Cu2O/TiO2 and CuO/TiO2 and reported them in Tables 1 and 2, respectively, SI. We noted that for low density/larger distance between the nanowires and clusters on TiO2 2L (101) surface, the formation energy values change a bit compared to the values at high density of wires/ clusters (3L) but the stable structures remained consistent. Thermodynamic Stability of Nanowires. Using the most stable Cu2O/TiO2 and CuO/TiO2 nanowires shown in Figures 1a and 2a, respectively, we investigated their thermodynamic stability at finite temperature. The Gibbs free energies in equilibrium with gas-phase O2 at a specified pressure (p) and temperature (T) were calculated using ab initio thermodynamics. In particular, the Gibbs free energy of formation was calculated as the difference x f ΔGCu (T , P) = gCu O/TiO − gCu /TiO − gOgas xO/TiO2 x 2 4 2 2 2 (1)

ggas O2

where is the free energy of a dioxygen molecule obtained as described in ref 36 and g can be written as g = f latt + f vibr + pv. f latt is the total DFT energy per Cu atom. The term f vibr is the vibrational contribution and is not taken into account36−39 in the solid phases, since it results in a small shift, within the error bar associated with other DFT approximations, and the pv term is neglected in solid phases as well. Plots of Gibbs free energy against temperature at 1, 10−5, and 10−10 atm of oxygen pressure are shown in Figure 3. At both low and high temperature and oxygen pressure, the CuO/TiO2 nanowire was found to have higher values of Gibbs free energy compared to Cu2O/TiO2. This indicates that the Cu2O nanowire on TiO2 (101) surface is thermodynamically more stable than the CuO nanowire on the same surface, unlike in bulk Cu oxides (see Figure S2), where CuO is the most stable phase compared to Cu2O at ambient conditions.11 This is also consistent with results of Ayyub et al.,40 where their studies on the effect of reducing the size of the particles on the crystal structure of Cu2O and CuO over a nanometer scale, found cubic Cu2O to be more stable than monoclinic CuO. This was attributed to the fact that Cu2O is more symmetric and more ionic than CuO. The nanometer particle size increases the ionic nature of the copper oxides and makes the lattice less directional, and hence, the more symmetric phase becomes more stable.41 The same results could also be interpreted as a consequence of the lower surface energy of Cu2O surfaces. Theoretical works42−44 predict that CuO (111), which is the most stable CuO surface, has a higher surface energy of 0.048 eV/Å2 compared to 0.003 eV/Å2 and 0.025 eV/Å2 for the oxygen-rich: Cu2O (110)/CuO and oxygen-poor terminations: Cu2O (111)−CuCUS. This might explain our result since, for nanoparticles, the surface energy contribution is very important in determining stability.45 We noted a corresponding decrease of transition temperature with pressure for the Cu oxides nanowires to Cu metal on TiO2 (101) anatase surface. For the case of CuO to Cu (see Figure 3b,c), the transition temperature at p = 10−5 atm and p = 10−10 atm is 871 and 680 K, respectively. Electronic Structure. The total density of states (TDOS) and partial densities of electronic states (PDOS) of the most stable Cu2O/TiO2 and the metastable CuO/TiO2 nanowire

Figure 3. Calculated Gibbs free energy of CuO (orange dashed line) and Cu2O (green dotted-dashed line) with respect to Cu (violet solid line) nanowires on TiO2 (101) anatase surface at (a) 1, (b) 10−5, and (c) 10−10 atm oxygen pressure as a function of temperature.

have been investigated, to shed light on the behavior of the system in dependence of the degree of oxidation of the copper nanostructures. The PDOS of the Cu2O nanowire is shown in Figure 4. The upper valence band from −5 to −1.25 eV is mainly occupied by Cu 3d states with moderate contributions from O 2p. The 3d states of copper atoms and O 2p extend the valence band edge by ∼1.5 eV into titania energy gap. The lower valence band consists mainly of hybridized states of O 2p, Cu 3d, and Ti 3d. The conduction band is predominantly composed of empty Ti 3d states. The Löwdin atomic charges and spin magnetic moments show no charge transfer between the Cu2O nanowire and the TiO2 anatase (101) surface atoms (see Table S3 and Table S5). A narrowed energy gap of 1.61 eV was obtained because the Cu2O nanowire states extended into the titania bandgap. The energy level alignment favors electron transfer from the CB of Cu2O nanowire into TiO2 CB. This is in agreement with Dongliang et al.46 who established that, for Cu2O/TiO2 20362


Article


Photogenerated electrons should preferentially be transferred to this CuO state from titania. This is in agreement with results of Yu et al.,48 who reported that the conduction band (CB) edge of TiO2 is higher than that of CuO clusters, therefore photogenerated electrons transfer from the CB of TiO2 into the CB of CuO clusters. The conduction band consists mainly of Ti 3d states. Since also the highest filled states belong to CuO, holes should be transferred there as well. This is an important difference with respect to Cu2O structures, where only holes should be transferred to the copper oxide. Similar to the case of Cu2O on TiO2 (101) anatase surface at 0 K, we observed that no charge is donated from CuO nanowire to TiO2 surface atoms (see Tables S4 and S6). The qualitative features of TDOS and PDOS of the CuO/TiO2 cluster (see Figure 5b) which has ΔE = 0.49 eV (see Figure 2b) are similar to those of most stable CuO/TiO2 nanowire. Our results suggest that for Cu nanostructured oxides on TiO2, Cu2O/TiO2 is more efficient for charge separation than CuO/TiO2. This is in qualitative agreement with ref 19, where it was observed that presence of Cu+ led to a higher photocatalytic activity, which was attributed to a better charge separation. Regarding electron transfer in CuO/TiO2, our results are in agreement with the experiments by Yu and coworkers,48 who conclude that the empty CuO states are below the bottom of the conduction band of TiO2 and filled states of CuO are above the top of the valence band of TiO2. Thus, a driving force for transfer of both electrons and holes to CuO is present. However, in the same work they observe that, while electrons are readily transferred to CuO, holes are present both in CuO and in TiO2. This points to different transfer rates for holes and electrons. Still, our findings suggest that Cu+ species should work better toward separation of photogenerated electrons and holes in comparison to Cu2+. Moreover, we find that nanostructured Cu2O/TiO2 is thermodynamically more stable than CuO/TiO2 and therefore more likely to be present under reaction conditions.

Figure 4. PDOS of the Cu2O nanowire on the TiO2 (101) surface: black solid line, total DOS; red line with circles, Cu-3d; blue line with triangles, O-2p; turquoise line with diamonds, Ti-3d. Fermi level is set at 0 eV. The total DOS has been reduced by a factor of 15 to be on the same scale as the PDOS.

coupled system, visible light excitation forces photogenerated electrons to be transferred from the CB of Cu2O to the CB of TiO2. Meanwhile, valence band alignment drives holes toward Cu2O, thus, fostering charge separation and suppressing recombination of photogenerated electrons and holes. The reduced energy gap extends the photoabsorption to visible range. Jagminas et al.47 similarly noted that the shift in the energy gap of titania nanotube film sample from ∼3.18 eV to visible range ∼2.3 eV for Cu2O/TiO2 sample is attributable to visible light charge transfer from Cu2O to TiO2. The first metastable Cu2O structure (Figure 1b) displays similar features, except for the small shoulder peak at the bottom of the CB of TiO2 (see Figure S5a). We observed polarons (see Figure S6) localized at 6- fold coordinated Ti atom located below Cu2O nanowire on TiO2 anatase (101) surface. The presence of polarons will enhance TiO2 photocatalytic activity. The PDOS of TiO2 anatase (101) surface modified by the CuO nanowire is shown in Figure 5. Also here, the top of the

■

CONCLUSIONS

Copper oxides nanostructures deposited on TiO2 anatase (101) surface were investigated by density functional theory and ab initio thermodynamics. The Cu2O nanowires were more stable than clusters with the same stoichiometry, and than all CuO structures. This is at variance from the case of bulk oxides, where CuO is more stable at room conditions. The atomic configurations of the low-energy structures are quite different from their bulk counterparts: in the Cu2O structures, Cu−Cu bond lengths are very close to that of experimental bulk metal copper (2.54 Å), while Cu−O bond lengths can differ by ∼0.1 Å from those of bulk Cu2O. In the most stable structure there is no direct Cu−Ti bond. The copper oxide nanostructures display filled electronic states that extend from the upper valence band into the band gap of titania, decreasing the band gap of the system by over 1 eV. The lowest-lying empty states belong to titania in the case of the Cu2O nanowire and to CuO in the case of the CuO nanowire. This suggests that the Cu2O nanostructures should favor hole transfer from titania to the copper oxide, while keeping photoexcited electrons in TiO2, contributing to charge separation and suppressing recombination. On the contrary, on the CuO nanostructures also the photoexcited electrons should be transferred to the copper oxide.

Figure 5. PDOS of a CuO nanowire on the TiO2 (101) surface: black solid line, total DOS; red line with circles, Cu-3d; blue line with triangles, O-2p; turquoise line with diamonds, Ti-3d. Fermi level is set at 0 eV. The total DOS has been reduced by a factor of 20 to be on the same scale as the PDOS.

valence band is predominantly composed of O 2p with significant contribution from Cu 3d states. The valence band is largely made up of Cu 3d states with moderate and small contributions of O 2p states and Ti 3d states, respectively. The TiO2 energy gap is occupied by O 2p states and Cu 3d states with a pronounced peak located at ∼0.8 eV above the Fermi level. This empty Cu 3d state near the bottom of the conduction band is not present in the Cu2O structures. 20363


Article


■

(13) Forcade, F.; Snyders, R.; Guisbiers, G.; Gonzalez, B.; Noirfalise, X.; Vigil, E. Impact of the chemical precursor on the crystalline constitution of nano-CuO/TiO2 films. Mater. Res. Bull. 2015, 70, 248− 253. (14) Gawande, M. B.; Goswami, A.; Felpin, F.-X.; Asefa, T.; Huang, X.; Silva, R.; Zou, X.; Zboril, R.; Varma, R. S. Cu and Cu-based nanoparticles: synthesis and applications in catalysis. Chem. Rev. 2016, 116, 3722−3811. (15) Roy, S. C.; Varghese, O. K.; Paulose, M.; Grimes, C. A. Toward solar fuels: photocatalytic conversion of carbon dioxide to hydrocarbons. ACS Nano 2010, 4, 1259−1278. (16) Yamashita, H.; Nishiguchi, H.; Kamada, N.; Anpo, M.; Teraoka, Y.; Hatano, H.; Ehara, S.; Kikui, K.; Palmisano, L.; Sclafani, A.; et al. Photocatalytic reduction of CO2 with H2O on TiO2 and Cu/TiO2 catalysts. Res. Chem. Intermed. 1994, 20, 815−823. (17) Tseng, I.-H.; Chang, W.-C.; Wu, J. C. Photoreduction of CO2 using sol-gel derived titania and titania-supported copper catalysts. Appl. Catal., B 2002, 37, 37−48. (18) Wu, N.-L.; Lee, M.-S. Enhanced TiO2 photocatalysis by Cu in hydrogen production from aqueous methanol solution. Int. J. Hydrogen Energy 2004, 29, 1601−1605. (19) Liu, L.; Gao, F.; Zhao, H.; Li, Y. Tailoring Cu valence and oxygen vacancy in Cu/TiO2 catalysts for enhanced CO2 photoreduction efficiency. Appl. Catal., B 2013, 134, 349−358. (20) Zhang, S.; Wang, H.; Yeung, M.; Fang, Y.; Yu, H.; Peng, F. Cu(OH)2-modified TiO2 nanotube arrays for efficient photocatalytic hydrogen production. Int. J. Hydrogen Energy 2013, 38, 7241−7245. (21) Giannozzi, P.; Baroni, S.; Bonini, N.; Calandra, M.; Car, R.; Cavazzoni, C.; Ceresoli, D.; Chiarotti, G. L.; Cococcioni, M.; Dabo, I.; et al. QUANTUM ESPRESSO: a modular and open-source software project for quantum simulations of materials. J. Phys.: Condens. Matter 2009, 21, 395502. (22) Vanderbilt, D. Soft self-consistent pseudopotentials in a generalized eigenvalue formalism. Phys. Rev. B: Condens. Matter Mater. Phys. 1990, 41, 7892. (23) Dudarev, S.; Botton, G.; Savrasov, S.; Humphreys, C.; Sutton, A. Electron-energy-loss spectra and the structural stability of nickel oxide: An LSDA+U study. Phys. Rev. B: Condens. Matter Mater. Phys. 1998, 57, 1505. (24) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized gradient approximation made simple. Phys. Rev. Lett. 1996, 77, 3865. (25) Camellone, M. F.; Kowalski, P. M.; Marx, D. Ideal, defective, and gold-promoted rutile TiO2 (110) surfaces interacting with CO, H2, and H2O: Structures, energies, thermodynamics, and dynamics from PBE+U. Phys. Rev. B: Condens. Matter Mater. Phys. 2011, 84, 035413. (26) Seriani, N.; Pinilla, C.; Crespo, Y. Presence of gap states at Cu/ TiO2 anatase surfaces: consequences for the photocatalytic activity. J. Phys. Chem. C 2015, 119, 6696−6702. (27) Scanlon, D. O.; Walsh, A.; Morgan, B. J.; Watson, G. W.; Payne, D. J.; Egdell, R. G. Effect of Cr substitution on the electronic structure of CuAl1−xCrx O2. Phys. Rev. B: Condens. Matter Mater. Phys. 2009, 79, 035101. (28) Meyer, B.; Polity, A.; Reppin, D.; Becker, M.; Hering, P.; Klar, P.; Sander, T.; Reindl, C.; Benz, J.; Eickhoff, M.; et al. Binary copper oxide semiconductors: from materials towards devices. Phys. Status Solidi B 2012, 249, 1487−1509. (29) Åsbrink, S.; Norrby, L.-J. A refinement of the crystal structure of copper (II) oxide with a discussion of some exceptional E.s.d.’s. Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst. Chem. 1970, 26, 8−15. (30) Monkhorst, H. J.; Pack, J. D. Special points for Brillouin-zone integrations. Phys. Rev. B 1976, 13, 5188. (31) Liu, C.; Yang, B.; Tyo, E.; Seifert, S.; DeBartolo, J.; von Issendorff, B.; Zapol, P.; Vajda, S.; Curtiss, L. A. Carbon dioxide conversion to methanol over size-selected Cu4 clusters at low pressures. J. Am. Chem. Soc. 2015, 137, 8676−8679. (32) Baishya, K.; Idrobo, J. C.; Ö ğüt, S.; Yang, M.; Jackson, K. A.; Jellinek, J. First-principles absorption spectra of Cun (n = 2−20) clusters. Phys. Rev. B: Condens. Matter Mater. Phys. 2011, 83, 245402.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpcc.7b06681. It contains the results of bulk calculations, surface structure, atomic charges, and magnetic moments for the stable structures (PDF).

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Nicola Seriani: 0000-0001-5680-9747 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS V.M. acknowledges the financial support from OFID for ICTP Postgraduate Fellowship Programme provided under the ICTP/IAEA Sandwich Training Educational Programme (STEP). Computational resources were provided by Abdus Salam ICTP and Centre for High Performance Computing (CHPC). The authors are most grateful.

■

REFERENCES

(1) Fang, B.; Xing, Y.; Bonakdarpour, A.; Zhang, S.; Wilkinson, D. P. Hierarchical CuO-TiO2 hollow microspheres for highly efficient photodriven reduction of CO2 to CH4. ACS Sustainable Chem. Eng. 2015, 3, 2381−2388. (2) Fujishima, A.; Honda, K. Electrochemical photolysis of water at a semiconductor electrode. Nature 1972, 238, 37−38. (3) O’Regan, B.; Grätzel, M. A low-cost, high-efficiency solar cell based on dye-sensitized. Nature 1991, 353, 737−740. (4) Hashimoto, K.; Irie, H.; Fujishima, A. TiO2 photocatalysis: a historical overview and future prospects. Jpn. J. Appl. Phys. 2005, 44, 8269. (5) Zaleska, A. Doped-TiO2: a review. Recent Pat. Eng. 2008, 2, 157− 164. (6) Ma, Q.; Liu, S.; Weng, L.; Liu, Y.; Liu, B. Growth, structure and photocatalytic properties of hierarchical Cu-Ti-O nanotube arrays by anodization. J. Alloys Compd. 2010, 501, 333−338. (7) Kumar, D. P.; Reddy, N. L.; Srinivas, B.; Durgakumari, V.; Roddatis, V.; Bondarchuk, O.; Karthik, M.; Ikuma, Y.; Shankar, M. Stable and active CuxO/TiO2 nanostructured catalyst for proficient hydrogen production under solar light irradiation. Sol. Energy Mater. Sol. Cells 2016, 146, 63−71. (8) Jing, H.-Y.; Wen, T.; Fan, C.-M.; Gao, G.-Q.; Zhong, S.-L.; Xu, A.-W. Efficient adsorption/photodegradation of organic pollutants from aqueous systems using Cu2O nanocrystals as a novel integrated photocatalytic adsorbent. J. Mater. Chem. A 2014, 2, 14563−14570. (9) Jin, Z.; Fei, G. T.; Hu, X. Y.; Wang, M.; De Zhang, L. Synthesis and photocatalytic activity of Cu2O/TiO2 double wall nanotube arrays. J. Nanoeng. Nanomanuf. 2012, 2, 49−53. (10) Lalitha, K.; Sadanandam, G.; Kumari, V. D.; Subrahmanyam, M.; Sreedhar, B.; Hebalkar, N. Y. Highly stabilized and finely dispersed Cu2O/TiO2: a promising visible sensitive photocatalyst for continuous production of hydrogen from glycerol: water mixtures. J. Phys. Chem. C 2010, 114, 22181−22189. (11) Heinemann, M.; Eifert, B.; Heiliger, C. Band structure and phase stability of the copper oxides Cu2O, CuO, and Cu4O3. Phys. Rev. B: Condens. Matter Mater. Phys. 2013, 87, 115111. (12) Pozan, G. S.; Isleyen, M.; Gokcen, S. Transition metal coated TiO2 nanoparticles: synthesis, characterization and their photocatalytic activity. Appl. Catal., B 2013, 140, 537−545. 20364


Article

The Journal of Physical Chemistry C (33) Treacy, J. P.; Hussain, H.; Torrelles, X.; Grinter, D. C.; Cabailh, G.; Bikondoa, O.; Nicklin, C.; Selcuk, S.; Selloni, A.; Lindsay, R.; et al. Geometric structure of anatase TiO2 (101). Phys. Rev. B: Condens. Matter Mater. Phys. 2017, 95, 075416. (34) Yang, F.; Sun, Q.; Ma, L.; Jia, Y.; Luo, S.; Liu, J.; Geng, W.; Chen, J.; Li, S.; Yu, Y. Magnetic Properties of CumOn Clusters: A First Principles Study. J. Phys. Chem. A 2010, 114, 8417−8422. (35) Jadraque, M.; Martín, M. DFT calculations of clusters: Evidence for Cu2O building blocks. Chem. Phys. Lett. 2008, 456, 51−54. (36) Reuter, K.; Scheffler, M. Composition, structure, and stability of RuO2 (110) as a function of oxygen pressure. Phys. Rev. B: Condens. Matter Mater. Phys. 2001, 65, 035406. (37) Seriani, N.; Pompe, W.; Ciacchi, L. C. Catalytic oxidation activity of Pt3O4 surfaces and thin films. J. Phys. Chem. B 2006, 110, 14860−14869. (38) Seriani, N. Ab initio thermodynamics of lithium oxides: from bulk phases to nanoparticles. Nanotechnology 2009, 20, 445703. (39) McQuarrie, D. A.; Simon, J. D. Molecular Thermodynamics; University Science Books: Sausalito, CA, 1999. (40) Ayyub, P.; Palkar, V.; Chattopadhyay, S.; Multani, M. Effect of crystal size reduction on lattice symmetry and cooperative properties. Phys. Rev. B: Condens. Matter Mater. Phys. 1995, 51, 6135. (41) Han, S.; Chen, H.-Y.; Chu, Y.-B.; Shih, H. C. Phase transformations in copper oxide nanowires. J. Vac. Sci. Technol., B: Microelectron. Process. Phenom. 2005, 23, 2557−2560. (42) Soon, A.; Todorova, M.; Delley, B.; Stampfl, C. Thermodynamic stability and structure of copper oxide surfaces: A first-principles investigation. Phys. Rev. B: Condens. Matter Mater. Phys. 2007, 75, 125420. (43) Soon, A.; Todorova, M.; Delley, B.; Stampfl, C. Erratum: Thermodynamic stability and structure of copper oxide surfaces: A first-principles investigation [Phys. Rev. B 2007, 75, 125420]. Phys. Rev. B: Condens. Matter Mater. Phys. 2007, 76, 129902. (44) Mishra, A. K.; Roldan, A.; de Leeuw, N. H. CuO surfaces and CO2 activation: a dispersion-corrected DFT+U study. J. Phys. Chem. C 2016, 120, 2198−2214. (45) Seriani, N.; Jin, Z.; Pompe, W.; Ciacchi, L. C. A DFT study of platinum oxides: from infinite crystals to nanoscopic particles. Phys. Rev. B: Condens. Matter Mater. Phys. 2007, 76, 155421. (46) Dongliang, H.; Jiahai, H.; Long, Q.; Jiangrui, P.; Zhenji, S. Optical and Photocatalytic Properties of Cu-Cu2O/TiO2 Two-Layer Nanocomposite Films on Si Substrates. Rare Metal Materials and Engineering 2015, 44, 1888−1893. (47) Jagminas, A.; Kovger, J.; Rėza, A.; Niaura, G.; Juodkazytė, J.; Selskis, A.; Kondrotas, R.; Šebeka, B.; Vaiči unienė , J. Decoration of ̅ the TiO2 nanotube arays with copper suboxide by AC treatment. Electrochim. Acta 2014, 125, 516−523. (48) Yu, J.; Hai, Y.; Jaroniec, M. Photocatalytic hydrogen production over CuO-modified titania. J. Colloid Interface Sci. 2011, 357, 223−228.

20365


Ab Initio Simulations of Copper Oxide Nanowires and Clusters on

Recommend Documents