MgTaO2N Photocatalysts: Perovskite versus Ilmenite Structure. A

Subscriber access provided by READING UNIV

Article 2

MgTaON Photocatalysts: Perovskite vs. Ilmenite Structure. A Theoretical Investigation Ayako Kubo, Giacomo Giorgi, and Koichi Yamashita J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.7b08874 • Publication Date (Web): 29 Nov 2017 Downloaded from http://pubs.acs.org on November 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 33 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

MgTaO2N Photocatalysts: Perovskite vs. Ilmenite Structure. A Theoretical Investigation Ayako Kubo,a,* Giacomo Giorgi,b,* Koichi Yamashita,a,c,* a

Department of Chemical System Engineering, School of Engineering, The University of Tokyo, 7-3-1, Hongo, Bunkyo-ku, Tokyo, Japan. b

Dipartimento di Ingegneria Civile e Ambientale (DICA), Università degli Studi di Perugia, Via G. Duranti 93, 06125 Perugia, Italy. c

CREST-JST, 7 Gobancho, Chiyoda-ku, Tokyo 102-0076, Japan.

Abstract

In this study we discuss the applicability of magnesium tantalum oxynitride (in the ABO2N stoichiometry) as photocatalyst by calculating its fundamental properties from first-principles, fully analyzing structural factors and material polymorphism. We find that, due to their favorable band edge nature, most of the configurations of the perovskite-type MgTaO2N are suitable for water splitting, while ilmenite-type MgTaO2N, the structurally expected more favored polymorph here investigated, results mostly unsuitable for the overall water-splitting. Our results also show the strong correlation between the conduction band structures and octahedral tilting, i.e., the more tilted the octahedra, the larger the overlap between Ta 5d bands and anion 2p bands, with subsequent bandgap opening and dispersion reduction. Therefore, in the perovskite-type MgTaO2N material design for photocatalytic applications, reduced (if not totally suppressed) octahedral tilting (i.e., via A-site doping) is highly desirable, in order to improve the solar-to-energy performances of the material.

1. Introduction With the depletion of fossil fuels and the raise of the greenhouse effect associated with their usage, hydrogen fuels represent an alternative clean energy supply of increasing significance. Water splitting

ACS Paragon Plus Environment

The Journal of Physical Chemistry 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 2 of 33

photocatalysis is considered a promising hydrogen production process due to the prompt and ubiquitous availability and abundance of the two main elements that participate to the process: water and sunlight. In particular, the mass production of solar hydrogen requires highly effective overall water-splitting photocatalysts. 1, 2, 3 The three main requirements a water-splitting photocatalyst needs to fulfill are (i) a high stability in aqueous environment, (ii) band edge positions that straddle water redox potential, and (iii) a bandgap which matches with the energy of visible light. Conventional oxide photocatalysts satisfy the first and second requirements but have large bandgaps, thus not satisfying the third requirement. On the other hand, narrow-gap nitride photocatalysts fulfill the third requirement while do not satisfy the first one since generally unstable in water. Interestingly, oxynitrides have intermediate properties, such as smaller bandgap than conventional oxides and higher stability than conventional nitrides4, and thus are attracting wide attention as a group of promising candidates for novel photocatalyst materials. So far, photocatalytic reactions under visible light have been reported on a wide variety of oxynitrides such as, among the others, ATaO2N(A = Ca, Sr, Ba) 5, 6, 7, 8, LaTiO2N9 in their perovskite geometry, wurztize GaN-ZnO solid solution,

10, 11

but also baddeleyte-like TaON12, confirming that, regardless their

crystalline shape, oxynitride materials13,14 are now being largely explored as novel photocatalysts. Recently, with the impressively enhanced availability of computational resources, computational screening of novel materials of technological relevance has become an extremely powerful procedure to predict the possible existence of yet unexplored materials. 15, 16 For perovskitetype oxynitrides of formula ABO2N, a screening procedure on thermodynamic and electronic properties (bandgap and bandedge position) initially involving 2704 compounds has revealed four well established and investigated materials, CaTaO2N, SrTaO2N, BaTaO2N, LaTiO2N,5, 6, 7, 8, 9 and a new one, MgTaO2N, as promising candidates for photocatalysis. 17 However, with the enormous computational cost in mind, modeling of the compounds just in terms


Page 3 of 33 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


of thermodynamic stability and band edge positions may result not sufficient and a more detailed analysis is mandatory in order to shed light on the photocatalytic features of such materials. The first issue, indeed, is the oversimplified structural model used in the screening procedure: perovskite oxynitrides denoted by ABO2N were all considered in their cubic polymorph whose unit cell consists of only five atoms, and thus structural features, such as rotation of the octahedral unit TaO4N2 (octahedral tilting)18 and characteristic anionic distribution (anion ordering)4 in the perovskite crystals are not taken into account. These structural factors are known to impact on the structures of both conduction19,20 and valence bands21, respectively, and thus they need to be considered in modeling and calculating the properties of MgTaO2N. The second problem is that only perovskite-type MgTaO2N was structurally modeled and other polymorphs have not been considered. Generally, a compound with the chemical formula ABX 3 assumes the perovskite structure only when its tolerance factor t (reported in eq.(1)) is close to one.22, 23

𝑡=

𝑟A + 𝑟X √2(𝑟B + 𝑟X )

(1)

where ri (i = A, B, X) is the ionic radius of the three species. However, compared to the A site cation of previously reported tantalum oxynitrides such as Ca2+, Sr2+, and Ba2+, the ionic radius of Mg2+ is clearly smaller and thus MgTaO2N is characterized by crystal structures other than perovskite.

24

According to the previous literature focusing on ABO3 stoichiometry polymorphism, species with A and B ~ 0.7 Å are likely to take either “ilmenite” or “corundum” structures, and in case of large difference in charge and electronegativity between A- and B-site cations, the former polymorph becomes the preferential25, 26. Furthermore, the ionic radii of the MgTaO2N constituents (Ta5+: 0.64 Å, O2-: 1.38 Å, N3-: 1.46 Å)27 are quite similar to those of MgTiO3 (Ti4+: 0.605 Å, O2-: 1.38 Å ) 28, 29, 30



Page 4 of 33

which has ilmenite structure, and thus MgTaO2N is expected to have the same crystal structure. To confirm our idea, using the following ionic radii i.e., for O2- 1.38, for N3- 1.46, for Mg2+ 0.72, and for Ta4.5+ 0.66 Å (in this case we considered the algebraic average between +4 and +5 oxidation number of Ta), we calculate a value for t=0.73, which further supports the stability of the ilmenite‒like polymorph for MgTaO2N. Therefore, not only perovskite but also ilmenite has to be taken into account in order to fully predict the photocatalytic properties of MgTaO2N from first principles: of course, we are aware of the possible existence of other geometries for MgTaO2N, at the same time previous literature and calculated geometrical factors (the Goldschimdt one)22,

23

seem to safely support our

present choice. Here, in this study, we aim to disclose the properties of MgTaO2N considering octahedral tilting, anion ordering, and the interplay between perovskite and ilmenite polymorphs, and thereby discussing its applicability as material for photocatalysis. In particular, we calculate bandgap, band edge position, and carrier effective masses of MgTaO2N from first principles. We then provide a brief outline for MgTaO2N material design as a novel photocatalyst by identifying the factors that impact on each properties.

2. Methods 2.1. Structural models The initial structure of perovskite-type MgTaO2N (hereafter also p-MgTaO2N) is modeled by imposing the crystal coordinates of CaTaO2N31, whose unit cell has the space group Pnma, replacing Ca atoms with Mg ones. Since CaTaO2N already has intrinsic octahedral-tilting of TaO4N2, that of MgTaO2N was successfully modeled by imposing the former coordinates as the initial ones for the latter structure. Besides, since many experimental observations for N orderings in perovskite oxynitrides have shown that O/N arrangements in perovskite crystals are not fully random 31, 32, 33, we decided to directly model the possible anion arrangements instead of applying SQS (Special


Page 5 of 33 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


Quasirandom Structure) where the anions are assumed to be distributed randomly34. In particular, we prepared 13 different anion orderings (hereafter also aors) , from A1 to F4 as shown in Fig. S1 in S.I., similarly to our previous analysis on the aors of CaTaO2N.21 The initial structure of ilmenite-type MgTaO2N (hereafter also i-MgTaO2N) derives from the rhombohedral unit cell of MgTiO3 with space group R3̅,28 initially modelled replacing Ti atoms with Ta ones and one third of O atoms with N ones. Also, we modeled two different anion distributions, cis and trans (see Fig. 1), in order to relate the electronic properties with anion ordering. All the optimized structures in this study are attached as supporting material. Density functional theory, as implemented in the VASP code,35, 36, 37, 38 was employed in order to geometrically relax the initial structures. On top of the optimized geometries we calculated the electronic properties. We considered the convergence achieved once the forces on individual atoms were smaller than 0.01 eV/Å. The Brillouin Zone (BZ) was sampled with a 10 × 7 × 10 centered mesh for p-MgTaO2N and with a 11 × 11 × 4 one for i-MgTaO2N. In particular, we employed the GGAPBE exchange correlation functional39 and the projected augmented wave (PAW) method,40, 41 along with an energy cutoff for the plane wave of 500 eV.

Figure 1 Initial structures of i-MgTaO2N with two anion distributions. Brown: tantalum, red: oxygen, blue: nitrogen, orange: magnesium. 2.2. Calculation details



Page 6 of 33

In order to discuss the fundamental photocatalytic properties of both p-MgTaO2N and i-MgTaO2N, we have calculated bandgaps (Eg), carrier effective masses, and band edge positions. Additionally, we compared the absorption spectrum between some of the structural models in order to shed light on their adsorption features. Electronic properties were calculated on the optimized structures of both polymorphs, with a 12 × 9 × 12 (p-MgTaO2N) and a 13 × 13 × 5 (i-MgTaO2N) mesh of the BZ, until the difference in the total free energy between two consecutive steps was lower than 10-5 eV. Besides, HSE06 hybrid functional42 was also applied to compensate the methodological underestimation of the electronic properties of the bare DFT approach. Bandgaps are approximated as the Kohn-Sham ones, and absorption spectra estimated by calculating the frequency dependent dielectric functions within an independent particle framework. Due to the computational cost, such spectra are estimated only for the most thermodynamically stable system (characterized by one of the largest bandagaps), and also for the system with the smallest bandgap, for both ilmenite and perovskite polymorphs. Carrier effective masses were estimated using band dispersions (see eqs. (S1) and (S2) in S.I.). Band edge positions were estimated as the orbital energies referenced to the vacuum potential43 that was calculated by means of a slab approach. As already pointed in the previous studies, the calculated value of the vacuum potential strongly depends on the surface structure of the slab model.44 Therefore, similarly to our previous study on CaTaO2N, we only used one slab model to calculate the vacuum potential and define the “effective” vacuum potentials of the other anion orders, as in Eq. (2), by assuming that the potential of the bulk model equals that of the bulk region of the slab model21 (see also eq. (S3) and (S4) in S.I.), ′ 𝑉vac = 𝑉bulk + (𝑉vac−slab − 𝑉bulk−slab )

(2)

′ where 𝑉vac is the effective vacuum potential, 𝑉bulk the potential of the bulk model, 𝑉vac−slab that in the

vacuum region of the slab model, and 𝑉bulk−slab that in the bulk region of the slab model.


Page 7 of 33 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


To calculate the vacuum potential, we used a slab model with zero net dipole along the normal to the slab surface. The thickness of the vacuum region was set to 19 Å on all slab models in order to prevent Coulomb interactions among replicas of the slab along the same direction,43, 45 while the thickness of the slabs was set to 30 Å (=six unit layers) on p-MgTaO2N and 42 Å (= three unit layers) on i-MgTaO2N. We have considered six slab models for p-MgTaO2N (see Fig. S2a in S.I.) and two slab models for i-MgTaO2N (Fig. S2b in S.I.) and adopted that with the smallest surface energy as calculated from Eq.(3). All the ionic positions of each slab were fully relaxed with a 4 × 4 × 1

centered sampling of the BZ until the forces on single atoms were smaller than 0.05 eV/Å. On the optimized structures, electronic calculations were performed with a 4 × 6 × 1 -centered sampling of the BZ on the slab models of p-MgTaO2N and a 6 × 6 × 1 -centered sampling of the BZ for iMgTaO2N with a total free energy convergence criterion of 10-4 eV. For all the slab models the potential of the vacuum region was converged and the bandgaps at the bulk regions reproduced those of the corresponding bulk models within 0.12 eV. All the surface energies calculated as in 𝐸surf =

𝐸slab − 𝑛 × 𝐸bulk 2𝐴

(3)

where 𝐸slab is the energy of the slab, 𝐸bulk is the chemical potential of bulk MgTaO2N, n the number of MgTaO2N units in the slab, and A is the exposed area of the slab. The factor “2” means that we are dealing with a symmetric slab i.e., top and bottom surfaces of the slab are formally identical further reducing any possible residual dipole in the slab. The bandgaps of each slab model are listed on Table S1 in S.I section. 3. Results 3.1. p-MgTaO2N Structural parameters, bandgaps, and carrier effective masses of each p-MgTaO2N aors are listed in Table 1. For sake of completeness, Table S2 in S.I. section lists the formation energies of the aors



Page 8 of 33

which range within 90 meV/FU and where F2 is found to be the most stable one. Here, according to Table 1, compared to the initial structures i.e., those of CaTaO2N, as previously mentioned, lattice parameters are shorter by ~1-5 %, while bond angles ∠Ta-X-Ta (X = O, N) are narrower by ~12-15°. On the other hand, their bond lengths 𝑟Ta−N and 𝑟𝑇a−O remain almost unaltered, proving that the octahedral tilting increases in p-MgTaO2N, compared to CaTaO2N. Table 1 Structural parameters, bandgaps, and carrier effective masses of each p-MgTaO2N. Lattice parameters (Ang)

Bond angles (deg)

Bond length (Ang)

a

b

c

∠Ta-X-Ta

∠Ta-O-Ta

∠Ta-N-Ta

rTa-O

rTa-N

rTa-O

rTa-N

Eg (eV)

me*

mh*

A1

5.46

7.74

5.28

138.5

137.5

140.5

2.08

1.98

0.01

0.01

3.20

0.87

1.10

A2

5.47

7.73

5.30

138.5

137.1

141.4

2.09

1.97

0.04

0.02

2.61

0.68

0.82

B1

5.47

7.74

5.29

138.4

138.1

139.0

2.09

1.96

0.05

0.03

3.10

0.66

1.38

B2

5.45

7.75

5.29

138.6

135.7

144.3

2.08

1.98

0.08

0.02

3.11

0.73

1.31

C

5.43

7.79

5.27

137.1

135.4

140.4

2.09

1.99

0.12

0.10

3.70

2.87

2.08

D

5.44

7.80

5.28

136.9

134.8

141.1

2.10

1.99

0.16

0.15

3.64

1.44

0.84

E1

5.43

7.78

5.27

137.0

135.5

140.0

2.09

1.99

0.12

0.10

3.65

1.18

1.99

E2

5.45

7.75

5.28

138.4

137.1

140.9

2.08

1.98

0.04

0.05

3.31

0.93

1.53

E3

5.47

7.73

5.29

138.5

136.6

142.4

2.08

1.98

0.07

0.05

3.26

0.85

1.35

F1

5.46

7.77

5.28

137.6

136.7

139.3

2.10

1.98

0.16

0.11

3.63

2.24

1.30

F2

5.44

7.80

5.28

137.1

135.8

139.6

2.10

1.98

0.17

0.11

3.80

1.07

1.20

F3

5.46

7.77

5.28

137.6

136.7

139.4

2.09

1.98

0.15

0.11

3.62

1.19

1.42

F4

5.43

7.80

5.27

136.4

133.8

141.6

2.10

1.99

0.16

0.18

4.13

1.79

1.84

Ave

5.45

7.77

5.28

137.7

136.2

140.8

2.09

1.98

0.10

0.08

3.44

1.27

1.40

Initial

5.62

7.90

5.55

--

--

--

153.9 (average)

2.03 (average)

--

Bond angles ∠Ta-X-Ta (X = O, N) of p-MgTaO2N also vary among different aors: they differ up to 2.2°, proving that the magnitude of octahedral tilting also depends on the aors. Bond angles ∠Ta-XTa of A~B2, E2, and E3 are smaller than the average bond angles of all aors (137.7°), while those of the remaining ones, i.e., C, D, E1, F1~F4 are larger than the same average value. Thus, we will refer to the former aors as “less-tilted” and to the latter as “more-tilted”. For all aors, average bond angles ∠Ta-N-Ta are always larger than ∠Ta-O-Ta ones, while average bond lengths 𝑟Ta−N are always shorter than 𝑟𝑇a−O ones, in clear contrast with the order of the ionic radius of N3- and O2-. Additionally, most of the adjacent Ta-X bonds, i.e., Ta-X-Ta (X=O, N) differ by ~0.1 Å in average, as illustrated in Fig. 2a. These structural characteristics can be explained by


Page 9 of 33 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


assuming the existence of a partial covalent nature of the Ta-X bond mainly associated with a d-p interaction.33, 46, 47, 48 Nitrogen reduced electronegativity ‒ compared to oxygen ‒enhances the covalent nature of the Ta-N bond vs Ta-O bond, also resulting in bond angle opening and shortening the bond lengths.

Figure 2. Asymmetric bond lengths between the consecutive Ta-N bonds. The magnitude of d-p  interaction can be described as the difference between the two bond lengths that connect X anion and Ta, i.e., ∆𝑟Ta−O and ∆𝑟Ta−N . These values differ among the aors and interestingly correlate with the bond angles ∠Ta-X-Ta (see Figure 3) showing that the more tilted the octahedron the larger the Ta-X d-p  interaction. The octahedral tilting also impacts on bandgaps. Those of p-MgTaO2N are larger, ranging from 2.6 eV to 4.1 eV, than those of analogous oxynitrides such as BaTaO2N (2.0 eV), SrTaO2N (2.1 eV), and CaTaO2N (2.5 eV)5 whose octahedral tilting is rather small. Here, such a wide range of bandgaps among different aors was also indicated from the estimated absorption spectrum (Figure S3a), where the onsets differed up to 1.0 eV between the two aors, A2 and F2, considered. Besides, in our case more-tilted aors show much larger bandgaps. Both trends suggest the non-negligible positive correlation between bandgaps and octahedral tilting. A similar trend has also been reported in the comparison among a series of perovskite oxynitride pigments, where the bandgaps increased with reducing rotation angles between their octahedral units49.



(a)

Page 10 of 33

(b)

Figure 3 Correlation between bond angles (∠Ta-X-Ta) and bond lengths differences in (a) Ta-O bond (∆𝑟Ta−O ) and (b) Ta-N bond (∆𝑟Ta−N ). ∆𝑟Ta−O and ∆𝑟Ta−N shows the larger value when ∠Ta-X-Ta is rather small, indicating the increase of the covalent interaction.

The calculated carrier effective mass of p-MgTaO2N also indicates the strong relationship between octahedral tilting and the band dispersion, especially in the conduction region. For pMgTaO2N, electron effective mass of me* range between 0.66 to 2.87 m0 (1.27 m0 in average) while the hole effective mass mh* between 0.82 and 2.08 m0 (1.40 m0 in average). Compared to the effective mass of CaTaO2N (me*: 0.63 m0 in average, mh*: 1.75 m0 in average),21 electron effective masses of pMgTaO2N are quite larger, while its hole effective masses are rather similar. In addition, the bond angles ∠Ta-X-Ta (X = O, N) of p-MgTaO2N strongly correlate with the electron effective mass, while negligible correlation is observed with the hole effective mass, as shown in Figure 4. Thus, the picture that emerges from our analysis is that the octahedral tilting mainly impacts on the structures of


Page 11 of 33 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


conduction bands (CBs). The correlation between octahedral tilting and the CB structures are also confirmed by the pMgTaO2N band edge positions reported in Figure 5(a) for different aors, with VBMs lying between +1.16 and +2.10 V, while CBMs between -1.45 and -2.04 V (vs NHE), most of which straddle the water redox potentials. As shown in Figures 5(b) and 5(c), CBM positions are strongly correlated with the ∠Ta-X-Ta bond angles, while VBM positions only slightly do it. All these results show that the octahedral tilting is the main factor for the different CBM positions among the aors.

(a)

(b)

Figure 4 Correlation between bond angles (∠Ta-X-Ta) and carrier effective mass of (a) electrons and (b) holes. The difference in CB electronic structures is also observed in the Density of States (DOS) and Projected Density of States (PDOS) of p-MgTaO2N, reported in Figure 6. CB dispersions of less-tilted aors i.e., A1~B2, E2, and E3, are relatively smooth, while those of more-tilted ones are sharp. Besides, the bottom regions of the CBs of the formers mainly consist of Ta 5d, while additional contribution of O 2p is observed in the same regions of the latter ones. These characteristics indicate that the uplift of the CBM positions with the increasing octahedral tilting originates from the band dispersion reduction that is caused by the d-p  interaction between Ta 5d and O 2p in the CBs. Similar trends are also



Page 12 of 33

reported in previous report by Eng et al.,19 where the relationship between the band dispersion reduction together with the octahedral tilting increase are ascribed to the increase of the antibonding character at the CBM.

(a)

(b)

(c)

Figure 5 (a) Band edge positions of p-MgTaO2N. The vertical axis was set to the normal hydrogen electrode (NHE) potential. (b) Correlation between bond angles (∠Ta-X-Ta) and CBM positions. (c) Correlation between bond angles (∠Ta-X-Ta) and VBM positions.


Page 13 of 33 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 6 DOS and PDOS of p-MgTaO2N. Their energy positions are aligned with the band edge positions depicted in Figure 5. Black dotted line: total DOS, Red dashed line: O 2p, Blue dashed line: N 2p, Brown solid line: Ta 5d.



Page 14 of 33

3.2. i-MgTaO2N Structural parameters, bandgaps, and carrier effective masses of each i-MgTaO2N are reported on Table 2. We also compared their formation energies to that of p-MgTaO2N in the most stable aor, and find that they are more stable by 249 meV/FU and 174 meV/FU in cis and trans arrangement, respectively, as listed in Table S2. In both i-MgTaO2N, average bond lengths 𝑟Ta−N are always shorter than 𝑟𝑇a−O , like in p-MgTaO2N, confirming the existence of a partial covalent nature of the bond in such class of compounds. In this case, at variance with the p-MgTaO2N, adjacent Ta‒X bondlengths i.e., Ta‒X‒Ta (X = O, N), are extremely different. Such differences (∆𝑟Ta−X , X = O, N) were more than twice larger than the same in p-MgTaO2N, and thus the enhanced covalent nature, that is, the enhanced overlap between tantalum and anion orbitals, are even larger in i-MgTaO2N. Such difference between the two polymorphs is ascribed to their different bond angles ∠Ta-X-Ta (X = O, N). In p-MgTaO2N, the overlap between Ta 5d and anion 2p is much hindered, since there is essentially a symmetry mismatch between such orbitals when they are in linear fashion,19, 21 while the reduced bond angles in i-MgTaO2N diminish the symmetry mismatch and thus lead to the enhanced overlap. Table 2 Structural parameters, bandgaps, and carrier effective masses of each of i-MgTaO2N. Lattice parameters (Ang)

Bond angles (deg)

Bond lengths (Ang)

A

b

c

∠Ta-O-Ta

∠Ta-N-Ta

∠Ta-X-Ta

rTa-O

rTa-N

rTa-O

rTa-N

Eg (eV)

me*

mh*

cis

5.33

5.33

14.38

99.4

94.3

97.7

2.12

1.99

0.26

0.16

4.28

3.01

1.62

trans

5.29

5.28

14.45

96.2

98.9

97.1

2.11

2.01

0.24

0.20

4.22

3.33

3.31

Ave

5.31

5.31

14.42

97.8

96.6

97.4

2.11

2.00

0.25

0.18

4.25

3.17

2.47

Bandgaps of both i-MgTaO2N are larger than 4.2 eV, markedly larger than those of p-MgTaO2N, with the material resulting a large gap semiconductor. Such larger bandgaps relative to p-MgTaO2N are also observed in their estimated absorption spectra in Figure S3b. Carrier effective masses are also larger in i-MgTaO2N, most of which are over 3.0 m0. Additionally, the band edge positions reveal i-


Page 15 of 33 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


MgTaO2N to be unsuitable for both oxygen and hydrogen evolution reaction (OER and HER, respectively) according to their positioning towards the NHE (Figure 7). PDOS of i-MgTaO2N reported in Figure 8 shows that their VB top and CB bottom mainly consist of anion 2p and Ta 5d, respectively, as for p-MgTaO2N. Nevertheless, for i-MgTaO2N, Ta 5d contribution is also present in the top region of VBs, while that of O 2p and N 2p is similarly present in the bottom region of CBs. These results indicate that the overlap between Ta 5d and anion 2p occurs in both bands, similarly witnessing a reduced selectivity of the carriers with subsequent reduced applicability in devices for photoconversion. By comparing the DOS of p-MgTaO2N and i-MgTaO2N, dispersions are clearly small in the latter one, and based on previous results, the i-MgTaO2N bandgap opening is ascribed to the band dispersion decrease which is consequence of the Ta-X interaction increase.

Figure 7. Band edge positions of i-MgTaO2N. The vertical axis was set to the normal hydrogen electrode (NHE) potential.

In addition to the increased interaction between Ta and X, the decrease of band dispersion in iMgTaO2N is mainly related to the layered (2D) nature of the structures: in the crystal of p-MgTaO2N, TaO4N2 octahedral units share their apices forming a 3D network, while they share their sides in the



Page 16 of 33

crystal of i-MgTaO2N, thus resulting in a quasi 2D structure. As shown in Fig. S4, the layered structures of i-MgTaO2N reduce their ionic density of both Ta cations and O, N anions compared to those of pMgTaO2N.

Figure 8 DOS and PDOS of i-MgTaO2N. Their energy positions are in line with the band edge positions depicted in Figure 7. Black dotted line: total DOS, Red dashed line: O 2p, Blue dashed line: N 2p, Brown solid line: Ta 5d.

4. Discussion Based on the obtained results, here we briefly discuss the possible applicability and the material design guidelines of the novel compound MgTaO2N. First, as mentioned in the previous section, most of the p-MgTaO2N configurations have suitable band edge positions for overall water splitting, since VBM positions are lower than the OER potential and CBM positions are higher than the HER potential. In particular, larger potential differences between CBM positions and HER potential are fingerprint of their strong tendency towards the HER. Nevertheless, their bandgaps are still larger than the energy of visible light, except for A2, whose VBM does not exceed the oxygen evolution potential. Accordingly, the bandgaps of p-MgTaO2N should be reduced mainly modifying their CBM positions. As demonstrated in the previous section, CBM positions are strongly correlated with octahedral tilting of p-MgTaO2N i.e., the suppression of the tilting would lead to a CBM reduction. Furthermore, these less-tilted p-MgTaO2N are also promising as water splitting photocatalyts since the electron effective mass diminishes as the octahedral tilting decreases, as reported in Figure 4. Therefore, a modified,


Page 17 of 33 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


tilting reduced p-MgTaO2N, is highly desirable: doping/alloying by replacing Mg2+ with different cations whose ionic radii is larger than Mg2+ will reduce the octahedral tilting in p-MgTaO2N offering suitable band structures for visible light driven water splitting, as previously shown in the case of perovskite-like AxA’1-xBO3 alloys, where an increased content of the larger cation is accompanied by the monotonic decrease of the bandgap.50 Concerning the stability of our investigated material it is worth mentioning that perovskite structure itself is known to be a thermally robust polymorph: analogous oxynitrides such as BaTaO2N, SrTaO2N, SrNbO2N, and LaTiO2N do not decompose up to 673 K. 49 On the other hand, while in principle oxynitrides are expected to undergo defect formation due to their ionic nature, recent experimental procedures have been reported where the synthetic procedure of ATaO2N (A=Ba) from an A-rich precursor and prepared under KCl flux

51

is likely to highly suppress the defects formation.

51

Accordingly we expect that following a similar procedure (Mg-rich precursor) the density of defects could be highly reduced. To confirm our idea new calculations are nowadays in progress. On the other hand, the bandgap of both i-MgTaO2N models is larger than 4.0 eV, which is markedly larger than the energy of visible radiation. Also, since the carrier effective mass of i-MgTaO2N is noticeably larger, its carrier mobility will clearly result reduced. Although the CBM positions of iMgTaO2N are negative enough to exceed the HER potential, their VBM positions are not positive enough to exceed the OER potential, and thus they are expected to be less performing in oxygen evolution ability. Therefore, in contrast to p-MgTaO2N, i-MgTaO2N is not suitable as water splitting photocatalyst.

5. Concluding remarks Computational screening procedures have revealed MgTaO2N a promising candidate for visible light responsive water splitting photocatalysis. However, some structural features such as octahedral



Page 18 of 33

tilting, anion ordering, and polymorphism have not been considered in the previous studies, and thus the properties of MgTaO2N as a photocatalyst material have been largely overlooked. We have here calculated the properties of MgTaO2N analyzing the mentioned features on different polymorphs and thereby discussed the applicability of MgTaO2N as a novel photocatalyst. Additionally, we have suggested a guideline to control and improve the properties of MgTaO2N in agreement with the results of the present analysis. Accordingly, we have shown that, in contrast with the intrinsic structural instability, most of the perovskite-type MgTaO2N have band structures highly suitable for overall water splitting. In particular, p-MgTaO2N with reduced octahedral tilting similarly has reduced bandgap and band edge positions that straddle the water redox potentials, although its bandgap is still slightly larger than the energy of visible light. On the other hand, MgTaO2N in ilmenite polymorph, the thermodynamically more stable one, has a bandgap extremely larger than visible light energy and band edge positions that do not match with the water redox potentials. As consequence, i-MgTaO2N performances in solar-to-energy devices will result quite poor. By comparing the structural features and the calculated properties, octahedral tilting results to highly impact on the conduction band structures. In particular, the larger the octahedral tilting the more marked the interaction between Ta 5d and anion 2p in the conduction band. Doping/alloying the A-site cation of MgTaO2N in the perovskite polymorph with a larger ionic radius cation, A2+, forming isovalent A’xMg1-xTaO2N alloys is a way to reduce their octahedral-tilting, improving the performances of the final photocatalyst. Acknowledgments Theoretical calculations in this study were performed using facilities at the Supercomputer Center in Institute for Solid State Physics (ISSP) of University of Tokyo and at Research Center for Computational Science in Institute for Molecular Science (IMS) in Okazaki, Japan. A.K. was supported by Japan Society for the Promotion of Science through Program for Leading Graduate


Page 19 of 33 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


Schools (MERIT). This research was supported by CREST (No. JPMJCR12C4) and MEXT as "Priority Issue on Post-K computer” (Development of new fundamental technologies for highefficiency energy creation, conversion/storage and use). G.G. wants to thank CINECA (ISCRA project ref HP10CN7DI0) and acknowledge PRACE for awarding us access to resource Marconi based in Italy at CINECA (Grant No. Pra14_3664). G. G. is similarly grateful to CARIT project “Progetto per l’applicazione delle attività di ricerca pubblica nell’area di crisi complessa ternana. Valutazione della possibilità di utilizzo di materiali metallici innovativi per applicazioni antisismiche, automobilistiche ed energetiche (ref. FCARITR17FR)” for supporting this research. References (1)

Kudo, A.; Miseki. Y. Heterogeneous Photocatalyst Materials for Water Splitting. Chem. Soc. Rev. 2009, 38, 253-278.

(2)

Yerga, R. M. N.; Galván, M. C. Á.; del Valle, F.; de la Mano, J. A. V.; Fierro, J. L. G. Water Splitting on Semiconductor Catalysts under Visible-Light Irradiation. ChemSusChem 2009, 2, 471−485.

(3)

Tong, H.; Ouyang, S.; Bi, Y.; Umezawa, N.; Oshikiri, M.; Ye, J. Nano-photocatalytic Materials: Possibilities and Challenges. Adv. Mater. 2012, 24, 229−251.

(4)

Attfield, J. P. Principles and Applications of Anion Order in Solid Oxynitrides. Cryst. Growth Des. 2013, 13, 4623−4629.

(5)

Yamasita, D.; Takata, T.; Hara, M.; Kondo, J. N.; Domen, K. Recent Progress of VisibleLight-driven Heterogeneous Photocatalysts for Overall Water Splitting. Solid State Ionics 2004, 172, 591−595.

(6)

Xu, J.; Pan, C.; Takata, T.; Domen, K. Photocatalytic Overall Water Splitting on the Perovskite-type Transition Metal Oxynitride CaTaO2N under Visible Light Irradiation. Chem. Commun. 2015, 51, 7191−7194.

(7)

Matoba, T.; Maeda, K.; Domen, K. Activation of BaTaO2N Photocatalyst for Enhanced Non-Sacrificial Hydrogen Evolution from Water under Visible Light by Forming a Solid Solution with BaZrO3. Chem. Eur. J. 2011, 17, 14731 – 14735.



Page 20 of 33

(8)

Maeda, K.; Domen, K. Water Oxidation Using a Particulate BaZrO3-BaTaO2N SolidSolution Photocatalyst That Operates under a Wide Range of Visible Light. Angew. Chem. Int. Ed. 2012, 51, 9865 –9869.

(9)

Kasahara, A.; Nukumizu, K.; Hitoki, G.; Takata, T.; Kondo, J. N.; Hara, M.; Kobayashi, H.; Domen, K. Photoreactions on LaTiO2N under Visible Light Irradiation. J. Phys. Chem. A 2002, 106, 6750−6753.

(10) Maeda, K.; Takata, T.; Hara, M.; Saito, N.; Inoue, Y.; Kobayashi, H.; Domen, K. GaN:ZnO Solid Solution as a Photocatalyst for Visible-Light-Driven Overall Water Splitting. J. Am. Chem. Soc. 2005, 127, 8286-8287. (11) Maeda, K.; Teramura, K.; Takata, T.; Hara, M.; Saito, N.; Toda, K.; Inoue, Y.; Kobayashi, H.; Domen, K. Overall Water Splitting on (Ga1-xZnx)(N1-xOx) Solid Solution Photocatalyst: Relationship between Physical Properties and Photocatalytic Activity. J. Phys. Chem. B 2005, 109 (43), 20504–20510. (12) Hara, M..; Takata, T.; Kondo, J. N.; Domen, K. Photocatalytic Reduction of Water by TaON under Visible Light Irradiation. Catal. Today 2004, 90, 313–317. (13)

Maeda, K.; Domen, K. New Non-oxide Photocatalysts Designed for Overall Water

Splitting under Visible Light. J. Phys. Chem. C 2007, 111, 7851–7861. (14) Maeda, K.; Domen, K.; Oxynitride Materials for Solar Water Splitting. MRS Bull. 2011, 36, 25−31. (15) Wu, Y.; Lazic, P.; Hautier, G.; Persson, K.; Ceder, G. First Principles High Throughput Screening of Oxynitrides for Water-splitting Photocatalysts. Energy Environ. Sci. 2013, 6, 157– 168. (16)

Zhuang, H. L.; Hennig, R. G. Computational Search for Single-Layer Transition-Metal

Dichalcogenide Photocatalysts. J. Phys. Chem. C 2013, 117, 20440−20445. (17) Castelli, I. E.; Olsen, T.; Datta, S. Landis, D. D.; Dahl, S.; Thygesen K. S.; Jacobsen, K. W. Computational Screening of Perovskite Metal Oxides for Optimal Solar Light Capture. Energy Environ. Sci. 2012, 5, 5814-5819.


Page 21 of 33 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


(18) Glazer, A. M. The Classification of Tilted Octahedra in Perovskites. Acta Cryst. 1972, B28, 3384–3392. (19) Eng, H. W.; Barnes, P. W.; Auer, B. M.; Woodward, P. M. Investigations of the Electronic Structure of d0 Transition Metal Oxides Belonging to the Perovskite Family. J. Solid State Chem. 2003, 175, 94–109. (20) Balaz, S.; Porter, S. H.; Woodward, P. M. Electronic Structure of Tantalum Oxynitride Perovskite Photocatalysts. Chem. Mater. 2013, 25, 3337−3343. (21) Kubo, A.; Giorgi, G.; Yamashita, K. Anion Ordering in CaTaO2N: Structural Impact on the Photocatalytic Activity. Insights from First-Principles. Chem. Mater. 2017, 29, 539−545. (22) Goldschmidt, V. M. Die Gesetze der Krystallochemie. Naturwissenschaften 1926, 14(21), 477-485. (23) Hoffman, A. Examinations of Compounds with Perovskite Structure. Z. Phys. Chem. B 1935, 28(1), 65-77. (24)

Li, W.; Ionescu, E.; Riedel, R. Gurlo, A. Can We Predict the Formability of Perovskite

Oxynitrides from Tolerance and Octahedral Factors? J. Mater. Chem. A 2013, 1, 12239-12245. (25) Roth, R. S. Classification of Perovskite and Other ABO3-Type Compounds. J. Res. Nat. Bur. Stand. 1957, 58(2), 75-88. (26) Giaquinta, D. M.; zur Loye, H.-C. Structural Predictions in the ABO3 Phase Diagram. Chem. Mater. 1994, 6, 365-372. (27) Kim, Y.-I.; Woodward, P. M.; Baba-Kishi, K. Z.; Tai, C. W. Characterization of the Structural, Optical, and Dielectric Properties of Oxynitride Perovskites AMO2N (A = Ba, Sr, Ca; M = Ta, Nb). Chem. Mater. 2004, 16, 1267−1276. (28) Wechsler, B. A.; Von Dreele, R. B. Structure Refinements of Mg2TiO4, MgTiO3 and MgTi2O5 by Time-of-Flight Neutron Powder Diffraction. Acta Cryst. 1989, B45, 542–549.



Page 22 of 33

(29) Baura-Peña, M. P.; Martínez-Lope, M. J.; García-Clavel, M. E. Synthesis of the Mineral Geikielite MgTiO3. J. Mater. Sci. 1991, 26, 4341–4343. (30) Shannon, R. D.; Prewitt, C. T. Effective Ionic Radii in Oxides and Fluorides. Acta Cryst. 1969, B25, 925-946. (31) Günther, E.; Hagenmayer, R.; Jansen, M. Structural Investigations on the Oxidenitrides SrTaO2N, CaTaO2N and LaTaON2 by Neutron and X-ray Powder Diffraction. Z. Anorg. Allg. Chem. 2000, 626, 1519−1525. (32) Logvinovich, D.; Bocher, L.; Sheptyakov, D.; Figi, R.; Ebbinghaus, S. G.; Aguiar, R.; Aguirre, H. M.; Reller, A.; Weidenkaff, A. Microstructure, Surface Composition and Chemical Stability of Partly Ordered LaTiO2N. Solid State Sci. 2009, 11, 1513-1519. (33) Yang, M.; Oró-Solé, J.; Rodgers, J. A.; Jorge, A. B.; Fuertes, A.; Attfield, J. P. Anion Order in Perovskite Oxynitrides. Nat. Chem. 2011, 3, 47−52. (34) Zunger, A.; Wei, S.-H.; Ferreira, L. G.; Bernard, J. E. Special Quasirandom Structures. Phys. Rev. Lett. 1990, 65, 353-356. (35)

Kresse, G.; Hafner, J. Ab-initio Molecular Dynamics for Liquid Metals. Phys. Rev. B

1993, 47, 558−561. (36) Kresse, G.; Hafner, J. Ab-initio Molecular-Dynamics Simulation of the Liquid Metal−Amorphous-Semiconductor Transition in Germanium. Phys. Rev. B 1994, 49, 14251−14269. (37) Kresse, G.; Furthmüller, J. Efficiency of Ab-initio Total Energy Calculations for Metals and Semiconductors Using a Plane-wave Basis Set. Comput. Mater. Sci. 1996, 6, 15−50. (38)

Kresse, G.; Furthmüller, J. Efficient Iterative Schemes for Ab-initio Total-Energy

Calculations Using a Plane-wave Basis Set. Phys. Rev. B 1996, 54, 11169−11186. (39) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77, 3865−3868.


Page 23 of 33 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


(40)

Blöchl, P. E. Projector Augmented-wave Method. Phys. Rev. B 1994, 50, 17953−17979.

(41) Kresse, G.; Joubert, D. From Ultrasoft Pseudopotentials to the Projector Augmentedwave Method. Phys. Rev. B 1999, 59, 1758−1775. (42) Krukau, A. V.; Vydrov, O. A.; Izmaylov, A. F.; Scuseria, G. E. Influence of the Exchange Screening Parameter on the Performance of Screened Hybrid Functionals. J. Chem. Phys. 2006, 125, 224106-5. (43)

Toroker, M. C.; Kanan, D. K.; Alidoust, N.; Isseroff, L. Y.; Liao, P.; Carter, E. A. First

Principles Scheme to Evaluate Band Edge Positions in Potential Transition Metal Oxide Photocatalysts and Photoelectrodes. Phys. Chem. Chem. Phys. 2011, 13, 16644−16654. (44) Huang, W. L. First-principles Determination of the Absolute Band-edge Positions of BiOX (X = F, Cl, Br, I). Comput. Mater. Sci. 2012, 55, 166−170. (45) Tasker, P. W. The Stability of Ionic Crystal Surfaces. J. Phys. C: Solid State Phys. 1979, 12, 4977−4984. (46) Clark, L.; Oró-Solé, J.; Knight, K. S.; Fuertes, A.; Attfield, J. P. Thermally Robust Anion-Chain Order in Oxynitride Perovskites. Chem. Mater. 2013, 25, 5004−5011. (47) Porter, S. H.; Huang, Z.; Woodward, P. M. Study of Anion Order/Disorder in RTaN2O (R = La, Ce, Pr) Perovskite Nitride Oxides. Cryst. Growth Des. 2014, 14, 117−125. (48)

Fuertes, A. Metal Oxynitrides as Emerging Materials with Photocatalytic and

Electronic Properties. Mater. Horiz. 2015, 2, 453−461. (49)

Aguiar, R; Logvinovich, D.; Weidenkaff, A.; Rachel, A.; Reller, A.; Ebbinghaus, S. G.

The Vast Colour Spectrum of Ternary Metal Oxynitride Pigments. Dyes and Pigments 2008, 76,70-75. (50) Tian, H. Y.; Luo, W. G.; Pu, X. H.; He, X. Y.; Qiu, P. S.; Ding, A. L.; Yang, S. H.; Mo D. Determination of the Optical Properties of Sol-Gel-derived BaxSr1-xTiO3 Thin Film by Spectroscopic Ellipsometry. J. Phys.: Condens. Matter 2001, 13, 4065-4074.



Page 24 of 33

(51) Dong, B.; Qi, Y.; Cui, J.; Liu, B.; Xiong, F.; Jiang, X.; Li, Z.; Xiao, Y.; Zhang, F.; Li, C. Synthesis of BaTaO2N Oxynitride from Ba-rich Oxide Precursor for Construction of VisibleLight-driven Z-scheme Overall Water Splitting. Dalton Trans. 2017, 46, 10707-10713.


Page 25 of 33 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


TOC 247x133mm (150 x 150 DPI)



Initial structures of i-MgTaO2N with two anion distributions. Brown: tantalum, red: oxygen, blue: nitrogen, orange: magnesium. 195x147mm (150 x 150 DPI)


Page 26 of 33

Page 27 of 33 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


Asymmetric bond lengths between the consecutive Ta-N bonds 124x72mm (150 x 150 DPI)



Correlation between bond angles (∠Ta-X-Ta) and bond lengths differences in (a) Ta-O bond (∆‫ݎ‬Ta−O) and (b) Ta-N bond (∆‫ݎ‬Ta−N). ∆‫ݎ‬Ta−O and ∆‫ݎ‬Ta−N shows the larger value when ∠Ta-X-Ta is rather small, indicating the increase of the covalent interaction 197x93mm (150 x 150 DPI)


Page 28 of 33

Page 29 of 33 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


Correlation between bond angles (∠Ta-X-Ta) and carrier effective mass of (a) electrons and (b) holes. 197x93mm (150 x 150 DPI)



(a) Band edge positions of p-MgTaO2N. The vertical axis was set to the normal hydrogen electrode (NHE) potential. (b) Correlation between bond angles (∠Ta-X-Ta) and CBM positions. (c) Correlation between bond angles (∠Ta-X-Ta) and VBM positions. 206x192mm (150 x 150 DPI)


Page 30 of 33

Page 31 of 33 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


(a) Band edge positions of p-MgTaO2N. The vertical axis was set to the normal hydrogen electrode (NHE) potential. (b) Correlation between bond angles (∠Ta-X-Ta) and CBM positions. (c) Correlation between bond angles (∠Ta-X-Ta) and VBM positions. 121x280mm (120 x 120 DPI)



Band edge positions of i-MgTaO2N. The vertical axis was set to the normal hydrogen electrode (NHE) potential. 104x110mm (120 x 120 DPI)


Page 32 of 33

Page 33 of 33 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


DOS and PDOS of i-MgTaO2N. Their energy positions are in line with the band edge positions depicted in Figure 7. Black dotted line: total DOS, Red dashed line: O 2p, Blue dashed line: N 2p, Brown solid line: Ta 5d. 130x63mm (120 x 120 DPI)


MgTaO2N Photocatalysts: Perovskite versus Ilmenite Structure. A

Recommend Documents