Ab Initio Simulations of Copper Oxide Nanowires and Clusters on

Subscriber access provided by UNIVERSITY OF CONNECTICUT

Article

Ab-Initio Simulations of Copper Oxide Nanowires and Clusters on TiO (101) Anatase Surface 2

Victor Meng'wa, Nicholas Makau, George Amolo, Sandro Scandolo, and Nicola Seriani J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.7b06681 • Publication Date (Web): 28 Aug 2017 Downloaded from http://pubs.acs.org on August 30, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry C is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

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, and Department of Physics and Space science, The Technical University of Kenya, P.O BOX 52428-00200, Nairobi, Kenya E-mail: [email protected]

∗

To whom correspondence should be addressed 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 †

1

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

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 Cu2 O to CuO, and morphologies, from clusters to nanowires, have been considered. Nanowire structures were consistently more stable than isolated clusters. In these structures, a Cu2 O 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 Cu2 O 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 Cu2 O 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 photo-chemical corrosion, non-toxicity, 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 mecha2


Page 2 of 23

Page 3 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


nisms, 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-workers 7 reported that titania based photocatalysts containing copper in different states (Cu, Cu2 O and CuO) showed improved catalytic efficiency, and attributed the enhancement to better absorption of visible light and to enhanced suppression of electronhole 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 behaviour of the overall system. In fact, both copper oxides are active photocatalysts by themselves. Cuprous oxide (Cu2 O) 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 co-catalyst 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 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

3



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

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 Lee 18 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 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 TiO2 surface can be reduced to metallic Cu during photoreactions due to trapped electrons, 20 Cu nanoparticles are easily oxidized to Cu2 O 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 Cu2 O 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 Cu2 O 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 favourable 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.

COMPUTATIONAL METHODS Spin polarized calculations were carried out in the framework of density functional theory, as implemented in the Quantum-Espresso code, 21 using a plane wave basis set. Vanderbilt

4


Page 4 of 23

Page 5 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ultrasoft pseudopotentials were used to describe core-valence interactions. 22 The study employed DFT+U 23 with the GGA-PBE 24 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 titanium 25,26 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 Cu2 O (experiment: 4.26 Å 28 ), with a Cu-O bond distance of 1.87 Å (experiment: 1.85 Å 28 ); for monoclinic CuO, a Cu-O bond distance of 1.95 Å was calculated (experiment: 1.96 Å 29 ). 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 Pack 30 was used. The TiO2 slabs have a thickness of three layers, corresponding to ∼10 Å, and the vacuum between the TiO2 slab and its replica is ∼14 Å thick. Cu2 O 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.

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 Cu2 O 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. 5



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 23

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 Fig. 1 and Fig. 2 for the Cu2 O and CuO stoichiometries, respectively. The most stable structure is the Cu2 O nanowire shown in Fig. 1a, with an almost planar geometry. Interestingly, planar structures were also found to be favourable for copper clusters in vacuum. 32 For each figure, also the difference in total energy (△E) at zero temperature between each structure and the most stable one for each stoichiometry is reported. For the structures with the Cu2 O stoichiometry, the stable nanowire structure lies lower than the second most stable structure by 1.15 eV, which is several times larger than kB T 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

Table 1: Average bond lengths in optimized structures of Cu2 O/TiO2 nanowire and clusters shown in Fig. 1. All in units of Å. nanowires/clusters a b c d e

Cu-Cu 2.53 2.52 2.44 2.60 2.59

Cu-O 1.96 1.91 1.90 1.84 1.88

Cu-Ti 2.79 2.68 2.62

O-Ti5C 2.02 1.82 1.76

Cu-O2C−surf. 1.92 1.88 1.98 1.90 1.90

copper oxide nanostructures is either Cu2 O or CuO. However, their structure might differ considerably from those of the bulk oxides. To show this, in Table 1 and Table 2 the average atomic distances are reported. For Cu2 O 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 Å to −0.01 Å, respectively, compared to the experimental Cu2 O bulk value of 1.85 Å, and longer than Cu-O length of 1.8 Å, from DFT+U calculations for bulk Cu2 O. 11 A Cu-Ti 6


Page 7 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1: Optimized Cu2 O nanowires and clusters on the TiO2 (101) surface with their respective total energy differences with respect to the most stable structure (a). (a) △E=0 eV; (b) △E=1.15 eV; (c) △E= 1.52 eV; (d) △E=2.10 eV; (e) △E=3.54 eV. Blue balls: titanium; red balls: oxygen; red-brown balls: copper. The two shown Cu2 O nanowires/clusters on TiO2 anatase (101) surface are replicas of each other. bond is found in the structures shown in Figs. 1c to 1e, with bond lengths ranging from 2.62 Å to 2.79 Å, considerably shorter than the Cu-Ti bond previously found in metallic copper clusters on TiO2 (2.92 Å 26 ). 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 Cu2 O structures have their total coordination ranging between 4- and 5-fold, and only

7



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

few have a 2-fold coordination like in the bulk. Three-fold coordinated O atoms are most prevalent, which is also different from bulk Cu2 O, where O is coordinated to 4 Cu atoms.

(a)

(b)

(c)

(d)

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) △E=0 eV; (b) △E=0.13 eV; (c) △E= 0.49 eV; (d) △E=1.23 eV. Colour code as in Fig. 1. The two shown CuO nanowires/clusters on TiO2 anatase (101) surface are replicas of each other. 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 Cu2 O 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 Å to −0.01 Å. A Cu-Ti bond is present in one of the structures (Fig. 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 Å and 2.03 Å, 33 but for one value of 1.81 Å. The Cu-O2C bond, 8


Page 8 of 23

Page 9 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


which is between Cu atom and two-fold coordinated oxygen atom (O2C ) on the TiO2 surface range between 1.89 Å 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 Cu2 O, in agreement with other cluster studies, 34,35 and Cu3 units are significantly present in CuO nanowire and cluster formation. Additionally, the Cu3 triangular framework is present in almost all the Cu2 O and CuO structures. Table 2: Average bond lengths in the optimized structures of CuO/TiO2 nanowires and clusters shown in Fig. 2. The bond lengths are in Å. nanowires/clusters a b c d

Cu-Cu 2.68 2.65 2.68 2.62

Cu-O 1.94 1.96 1.95 1.95

Cu-Ti 2.78

O-Ti5C 1.95 1.89 2.00 1.81

Cu-O2C−surf. 2.13 1.91 1.96 1.89

We also considered most stable 3-layer (3L) (10 Å× 7.5 Å) with vacuum 14 Å, of relaxed Cu2 O/TiO2 (see Figs. 1a, 1b and 1c) 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 2layers (2L) to reduce the computational costs. We performed structural relaxation and three most stable optimized 2L Cu2 O/TiO2 and CuO/TiO2 structures arranged in the order of decreasing stability are shown in Figs. 3 and 4, respectively, SI. We also calculated formation energies of three most stable 3L and 2L Cu2 O/TiO2 and CuO/TiO2 and reported them in Table 1 and Table 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

9



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 23

remained consistent.

Thermodynamic stability of nanowires Using the most stable Cu2 O/TiO2 and CuO/TiO2 nanowires shown in Figs. 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 gas ∆GfCux O/T iO2 (T, P ) = gCux O/T iO2 − gCu4 /T iO2 − gO 2 2

(1)

36 where ggas and g O2 is the free energy of a dioxygen molecule obtained as described in Ref.

can be written as g=flatt +fvibr +pv. flatt is the total DFT energy per Cu atom. The term fvibr is the vibrational contribution and is not taken into account 36–39 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 atmosphere (atm), 10−5 atm and 10−10 atm of oxygen pressure are shown in Fig. 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 Cu2 O/TiO2 . This indicates that the Cu2 O nanowire on TiO2 (101) surface is thermodynamically more stable than the CuO nanowire on the same surface, unlike in bulk Cu oxides (see Fig. S2), where CuO is the most stable phase compared to Cu2 O 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 Cu2 O and CuO over a nanometer scale, found cubic Cu2 O to be more stable than monoclinic CuO. This was attributed to the fact that Cu2 O 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

10


Page 11 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


phase becomes more stable. 41 The same results could also be interpreted as a consequence of the lower surface energy of Cu2 O surfaces. Theoretical works 42–44 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: Cu2 O(110):CuO and oxygen-poor terminations: Cu2 O(111)-CuCU S . 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 Figs. 3b and 3c), the transition temperature at p=10−5 atm and p=10−10 atm is 871 K and 680 K, respectively.

11



(a) 1

Energy (eV)

p=1 atm

0

-1

-2

200

400

600

800

400

600

800

400

600

800

Temperature (K)

(b) 1 -5

Energy (eV)

p=10 atm 0

-1

-2

200

Temperature (K)

(c) 1 p=10

Energy (eV)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

-10

atm

0

-1

-2

200

Temperature (K)

Figure 3: Calculated Gibbs free energy of CuO (orange dashed line) and Cu2 O (green dotteddashed line) with respect to Cu (violet solid line) nanowires on TiO2 (101) anatase surface at (a) 1 atm, (b) 10−5 atm and (c) 10−10 atm oxygen pressure as a function of temperature.

Electronic structure The total density of states (TDOS) and partial densities of electronic states (PDOS) of the most stable Cu2 O/TiO2 and the metastable CuO/TiO2 nanowire have been investigated, to shed light on the behaviour of the system in dependence of the degree of oxidation of the

12


Page 12 of 23

Page 13 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


copper nanostructures. The PDOS of the Cu2 O nanowire is shown in Fig. 4. The upper valence band from −5 eV 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 Cu2 O 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 Cu2 O nanowire states extended into the titania bandgap. The energy level alignment favours electron transfer from the CB of Cu2 O nanowire into TiO2 CB. This is in agreement with Dongliang et al., 46 who established that for Cu2 O/TiO2 coupled system, visible light excitation forces photogenerated electrons to be transferred from the CB of Cu2 O to the CB of TiO2 . Meanwhile, valence band alignment drives holes towards Cu2 O, 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 Cu2 O/TiO2 sample is attributable to visible light charge transfer from Cu2 O to TiO2 . The first metastable Cu2 O structure (Fig. 1b) displays similar features, except for the small shoulder peak at the bottom of the CB of TiO2 (see Fig. S5a). We observed polarons (see Fig. S6) localized at 6- fold coordinated Ti atom located below Cu2 O 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 Fig. 5. Also here, the top of the 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

13



4

PDOS (arb. units)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 23

Total-DOS Cu-3d Ti-3d

2 O-2p 0 -8

-6

-4

-2 0 2 ENERGY (eV)

4

6

8

Figure 4: PDOS of the Cu2 O 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. 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 Cu2 O structures. 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 Cu2 O structures, where only holes should be transferred to the copper oxide. Similar to the case of Cu2 O on TiO2 (101) anatase surface at 0 K, we observed that no charge is donated from CuO nanowire to TiO2 surface atoms (see Table S4 and Table S6). The qualitative features of TDOS and PDOS of the CuO/TiO2 cluster (see Fig. S5b) which has △E= 0.49 eV (see Fig. 2b) are similar to those of most stable CuO/TiO2 nanowire. Our results suggest that for Cu nanostructured oxides on TiO2 , Cu2 O/TiO2 is more efficient for charge separation than CuO/TiO2 . This is in qualitative agreement with Ref., 19

14


Page 15 of 23

Cu-3d

Total-DOS

O-2p

Ti-3d

2 PDOS (arb. units)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


0

-2 -6

-4

-2

0 2 4 ENERGY (eV)

6

8

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. 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 co-workers, 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 towards separation of photogenerated electrons and holes in comparison to Cu2+ . Moreover, we find that nanostructured Cu2 O/TiO2 is thermodynamically more stable than CuO/TiO2 and therefore more likely to be present under reaction conditions.

CONCLUSIONS Copper oxides nanostructures deposited on TiO2 anatase (101) surface were investigated by density functional theory and ab-initio thermodynamics. The Cu2 O nanowires were more 15



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

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 Cu2 O 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 Cu2 O. 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 Cu2 O nanowire, and to CuO in the case of the CuO nanowire. This suggests that the Cu2 O nanostructures should favour 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. Supporting Information. It contains the results of bulk calculations, surface structure, atomic charges and magnetic moments for the stable structures.

ACKNOWLEDGMENTS Victor Meng’wa 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 CuOTiO2 hollow microspheres for highly efficient photodriven reduction of CO2 to CH4 . ACS Sustainable Chemistry & Engineering 2015, 3, 2381–2388. 16


Page 16 of 23

Page 17 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(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. Japanese Journal of Applied Physics 2005, 44, 8269. (5) Zaleska, A. Doped-TiO2 : a review. Recent Patents on Engineering 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. Journal of Alloys and Compounds 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 Cux O/TiO2 nanostructured catalyst for proficient hydrogen production under solar light irradiation. Solar Energy Materials and Solar 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 Cu2 O nanocrystals as a novel integrated photocatalytic adsorbent. Journal of Materials Chemistry A 2014, 2, 14563–14570. (9) Jin, Z.; Fei, G. T.; Hu, X. Y.; Wang, M.; De Zhang, L. Synthesis and photocatalytic activity of Cu2 O/TiO2 double wall nanotube arrays. Journal of Nanoengineering and Nanomanufacturing 2012, 2, 49–53. (10) Lalitha, K.; Sadanandam, G.; Kumari, V. D.; Subrahmanyam, M.; Sreedhar, B.; Hebalkar, N. Y. Highly stabilized and finely dispersed Cu2 O/TiO2 : a promising visi-

17



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ble sensitive photocatalyst for continuous production of hydrogen from glycerol: water mixtures. The Journal of Physical Chemistry C 2010, 114, 22181–22189. (11) Heinemann, M.; Eifert, B.; Heiliger, C. Band structure and phase stability of the copper oxides Cu2 O, CuO, and Cu4 O3 . Physical Review B 2013, 87, 115111. (12) Pozan, G. S.; Isleyen, M.; Gokcen, S. Transition metal coated TiO2 nanoparticles: synthesis, characterization and their photocatalytic activity. Applied Catalysis B: Environmental 2013, 140, 537–545. (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. Materials Research Bulletin 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. Chemical Reviews 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 H2 O on TiO2 and Cu/TiO2 catalysts. Research on Chemical Intermediates 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. Applied Catalysis B: Environmental 2002, 37, 37–48. (18) Wu, N.-L.; Lee, M.-S. Enhanced TiO2 photocatalysis by Cu in hydrogen production

18


Page 18 of 23

Page 19 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


from aqueous methanol solution. International Journal of 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. Applied Catalysis B: Environmental 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. International Journal of 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. Journal of Physics: Condensed Matter 2009, 21, 395502. (22) Vanderbilt, D. Soft self-consistent pseudopotentials in a generalized eigenvalue formalism. Physical Review B 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. Physical Review B 1998, 57, 1505. (24) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized gradient approximation made simple. Physical Review Letters 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 H2 O: Structures, energies, thermodynamics, and dynamics from PBE+U. Physical Review B 2011, 84, 035413. (26) Seriani, N.; Pinilla, C.; Crespo, Y. Presence of gap states at Cu/TiO2 anatase sur-

19



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

faces: consequences for the photocatalytic activity. The Journal of Physical Chemistry 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−x Crx O2 . Physical Review B 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. Physica 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 Crystallographica Section B: Structural Crystallography and Crystal Chemistry 1970, 26, 8–15. (30) Monkhorst, H. J.; Pack, J. D. Special points for Brillouin-zone integrations. Physical Review 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. Journal of the American Chemical Society 2015, 137, 8676– 8679. (32) Baishya, K.; Idrobo, J. C.; Öğüt, S.; Yang, M.; Jackson, K. A.; Jellinek, J. Firstprinciples absorption spectra of Cun (n= 2-20) clusters. Physical Review B 2011, 83, 245402. (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). Physical Review B 2017, 95, 075416.

20


Page 20 of 23

Page 21 of 23

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(34) Yang, F.; Sun, Q.; Ma, L.; Jia, Y.; Luo, S.; Liu, J.; Geng, W.; Chen, J.; Li, S.; Yu, Y. Magnetic Properties of Cum On Clusters: A First Principles Study. The Journal of Physical Chemistry A 2010, 114, 8417–8422. (35) Jadraque, M.; Martín, M. DFT calculations of clusters: Evidence for Cu2 O building blocks. Chemical Physics Letters 2008, 456, 51–54. (36) Reuter, K.; Scheffler, M. Composition, structure, and stability of RuO2 (110) as a function of oxygen pressure. Physical Review B 2001, 65, 035406. (37) Seriani, N.; Pompe, W.; Ciacchi, L. C. Catalytic oxidation activity of Pt3 O4 surfaces and thin films. The Journal of Physical Chemistry 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. Physical Review B 1995, 51, 6135. (41) Han, S.; Chen, H.-Y.; Chu, Y.-B.; Shih, H. C. Phase transformations in copper oxide nanowires. Journal of Vacuum Science & Technology B: Microelectronics and Nanometer Structures Processing, Measurement, and Phenomena 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. Physical Review B 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 75, 125420 (2007)]. Physical Review B 2007, 76, 129902. 21



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(44) Mishra, A. K.; Roldan, A.; de Leeuw, N. H. CuO surfaces and CO2 activation: a dispersion-corrected DFT+U study. The Journal of Physical Chemistry 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. Physical Review B 2007, 76, 155421. (46) Dongliang, H.; Jiahai, H.; Long, Q.; Jiangrui, P.; Zhenji, S. Optical and Photocatalytic Properties of Cu-Cu2 O/TiO2 Two-Layer Nanocomposite Films on Si Substrates. Rare Metal Materials and Engineering 2015, 44, 1888–1893. (47) Jagminas, A.; Kovger, J.; R˙eza, A.; Niaura, G.; Juodkazyt˙e, J.; Selskis, A.; Kondrotas, R.; Šebeka, B.; Vaiči¯ unien˙e, J. Decoration of the TiO2 nanotube arays with copper suboxide by AC treatment. Electrochimica Acta 2014, 125, 516–523. (48) Yu, J.; Hai, Y.; Jaroniec, M. Photocatalytic hydrogen production over CuO-modified titania. Journal of Colloid and Interface Science 2011, 357, 223–228.

22


Page 22 of 23

Page 23 of 23

Table of Contents graphic

4 PDOS (arb. units)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Total-DOS Cu-3d Ti-3d

2 O-2p 0 -8

-6

-4

-2 0 2 ENERGY (eV)

23

4


6

8

Ab Initio Simulations of Copper Oxide Nanowires and Clusters on

Recommend Documents