Characterization of peroxo reaction intermediates in the water oxidation process on hematite surfaces

  • Lodvert Tchibota Poaty
  • Kanchan Ulman
  • Nicola Seriani
  • Bernard M’Passi-Mabiala
  • Ralph GebauerEmail author
Original Paper
Part of the following topical collections:
  1. International Conference on Systems and Processes in Physics, Chemistry and Biology (ICSPPCB-2018) in honor of Professor Pratim K. Chattaraj on his sixtieth birthday


We use density functional theory-based calculations to study structural, electronic, and magnetic properties of two key reaction intermediates on a hematite, \(\alpha \)-Fe2O3, photoanode during the solar-driven water splitting reaction. Both intermediates contain an oxygen atom bonded to a surface iron atom. In one case, the adsorbed oxygen also forms a peroxo bond with a lattice oxygen from hematite; in the second case no such bond is formed. Both configurations are energetically equivalent and are related to the overpotential-determining step in the oxygen evolution reaction. The calculated reaction path for the breaking of the peroxo bond shows a barrier of about 0.86 eV for the transformation between the two intermediates. We explain this high barrier with the drastically different electronic and magnetic structure, which we also analyze using maximally localized Wannier functions. Photo-generated electron holes are shown to localize preferentially close to the reaction center at the surface in both configurations. In the case of the oxo species, this localization favors subsequent electron transfer steps during the oxygen evolution cycle. In the case of the peroxo configuration, this fact together with the high barrier for breaking the oxygen–oxygen bond indicates a possible loss mechanism due to hole trapping.

Graphical Abstract

Calculated spin density at a hematite surface with peroxo intermediate


Hematite Water splitting Density functional theory 


Using sunlight to split water into hydrogen and oxygen is a very promising way to harvest clean energy. Unlike other sustainable forms of energy conversion, like photovoltaics or wind energy, water splitting does not produce electricity, but leads to chemical storage of the energy. As a consequence, some of the most notorious issues related to renewable energies are mitigated: the intermittent nature of solar radiation and wind, as well as the use of electricity for transportation, which requires efficient electrical energy storage. Green chemical fuels, for which hydrogen is an essential ingredient, can be used within the existing technological infrastructure, and can easily store large amounts of energy, even for long periods.

While the upsides of clean chemical energy carriers like hydrogen are evident, the actual splitting of water using solar radiation is still a daunting task. The reaction can only take place in the presence of a photo-catalyst, and a lot of effort is currently under way to find efficient catalytic materials for the photoelectrochemical (PEC) water splitting reaction. The overall reaction (2 H2O → O2 + 2 H2) is performed in to half-reactions. The so-called oxygen evolution reaction (OER), 2 H2O → O2 + 4 H+ + 4 e, takes place at the photo-anode, while the hydrogen evolution reaction (HER), 4 H+ + 4 e → 2 H2, takes place at the photo-cathode.

Of the two half-reactions, the OER is by far the most difficult to catalyze efficiently, because it involves four distinct proton-coupled electron transfer (PCET) steps. In the following, we will therefore focus on the OER.

One heavily investigated candidate material as a photo-anode is hematite (iron oxide, \(\alpha \)-Fe2O3) [1]. This semiconductor is extremely stable, including in water and oxygen-rich environments, is earth-abundant, and exhibits a band gap of 2.1 eV, which is ideal for capturing visible light [2]. Hematite has many limitations as an efficient anode material, which include low lifetimes of electron-hole pairs [3], high recombination rates at surfaces [4], and charge-carrier trapping [5]. Such problems are addressed with various strategies, ranging from doping [6, 7, 8, 9] to nanostructuring [10] and thin overlayers [11, 12].

One particular feature that sets iron oxide apart from many other transition metal oxides is its magnetic nature: Bulk hematite has an anti-ferromagnetic ordering, where Fe atoms in neighboring layers have opposite magnetic moments. The magnetism in this material is entirely due to spin polarization of the electrons.

In order to guide such efforts to improve the catalytic performance of hematite and to better understand the effect of the experimental strategies at an atomic level, computer simulations play an increasingly important role. In particular, calculations based on density functional theory (DFT) have been heavily used to elucidate the elementary processes on hematite surfaces. We refer to Ref. [13] for a recent overview of DFT-based calculations of water splitting on hematite. In this context, much attention has been paid to the relative stabilities of different terminations of the hematite (0001) surface. While earlier studies have concentrated on fully hydroxylated surfaces [8, 14], it has become clear later [15, 16] that this is not the most stable surface termination in an aqueous environment and under illumination, when the anode is electrically connected to a photo-cathode (PEC-conditions). In such conditions, oxygen-rich terminations are more stable [16]. More recently, some of the authors of the present paper have shown the importance of reaction intermediates where lattice oxygen from hematite is bound to oxygen from water [17]. One of the key findings is that there are various possible terminations and reaction pathways which lead to water splitting which are nearly indistinguishable from each other, due to similar energetics and reaction potentials. Very likely, those pathways coexist in a real device.

One of the relevant pathways is represented in Fig. 1. That pathway, like all others identified in Ref. [17], is characterized by four subsequent PCET steps:
$$\begin{array}{@{}rcl@{}} 2 \text{H}_{2}\text{O} + ()^{*} &\overset{A}{\longrightarrow}&\text{H}\text{O}^{*} + \text{H}_{2}\text{O} + (\text{H}^{+} + e^{-}) \end{array} $$
$$\begin{array}{@{}rcl@{}} &\overset{B}{\longrightarrow}&\text{O}^{*} + \text{H}_{2}\text{O} + 2(\text{H}^{+} + e^{-}) \end{array} $$
$$\begin{array}{@{}rcl@{}} &\overset{C}{\longrightarrow}&\text{H}\text{O}\text{O}^{*} + 3(\text{H}^{+} + e^{-}) \end{array} $$
$$\begin{array}{@{}rcl@{}} &\overset{D}{\longrightarrow}&()^{*} + \text{O}_{2} + 4(\text{H}^{+} + e^{-}). \end{array} $$
Here, the asterisks () indicate reaction intermediates attached to the surface, and \(()^{*}\) is the active site of the surface—in the case shown in Fig. 1 an under-coordinated surface Fe atom. In each of the four PCET steps A to D, one proton is moving to the electrolyte and one electron to the anode material. The reaction is driven by sunlight, which is absorbed in the semiconductor material and is creating electron-hole pairs there. While the excited electrons are used at the cathode to reduce protons in the HER, the holes are neutralized at the anode surface by the electrons provided in steps A to D of the OER [16, 17, 18, 19, 20, 21, 22, 23, 24].
Fig. 1

Water oxidation cycle, starting from a hydroperoxo group attached to an iron-terminated surface. Color scheme for atomic spheres: Fe (blue), O (white), and H (black)

On an ideal catalytic surface, the free energy difference between each of the four configurations is 1.23 eV, amounting for the 4.92 eV needed to split two water molecules. In reality, the energy differences between the configurations are never equal, giving rise to an overpotential (and therefore a loss of energy), which is needed to overcome the step with the highest free energy difference—the overpotential-determining step.

In the case shown in Fig. 1, the potential-determining step is B, which involves the deprotonation of a surface-attached hydroxyl group. Interestingly, two possible final states, indicated as \(S_{3}\) and \(S_{3}^{\prime }\) in the figure, have been identified in Ref. [17]. Both configurations are very similar in energy, (S3 is only 32 meV higher in energy than \(S_{3}^{\prime }\)). The existence of \(S_{3}^{\prime }\) is a very intriguing fact: an oxygen–oxygen bond is formed between a lattice oxygen from the hematite and an adsorbed oxygen from the solution. The bond length of 1.47 Å is much larger than the one in an \(\text {O}_{2}\) molecule (1.2 Å), which seems to indicate that here the bonding is in the peroxo regime. This fact would explain various experimental hints [25, 26, 27] at the presence of peroxo reaction intermediates in water-splitting reactions on hematite and other oxide surfaces.

The main focus and novelty of this paper is to closer examine the nature of this important reaction intermediate, which is related to the overpotential-determining step in the OER, and therefore of key importance for water splitting on hematite. We analyze the electronic and magnetic structure of the two configurations \(S_{3}\) and \(S_{3}^{\prime }\) more closely, with the help of maximally localized Wannier functions. Also, we explicitly introduce electron holes in those structures. We also identify the transition state and reaction pathway for a hypothetical thermally activated crossing between \(S_{3}\) and \(S_{3}^{\prime }\). We also characterize the nature of the transition state.

A close examination of the intermediates is given in “Oxo- and peroxo reaction intermediates”. The eventual possibility of a cross-over between the two configurations is discussed in “Oxygen bonding pathway”, while the relevant electronic structures are analyzed in terms of Wannier functions in “Electronic structure from Wannier functions in bulk hematite and in the peroxo configuration \(S_{3}^{\prime }\)”. Finally, we discuss how the presence of electron holes at the anode surface can alter the electronic structure in “Electron holes”.


All calculations are based on spin-polarized density functional theory (DFT) [28, 29]. The calculations are performed using the quantum-ESPRESSO suite of programs [30, 31] for the electronic structure and post-processing, Wannier functions are obtained using the program Wannier90 [32]. The exchange and correlation functional is approximated using the semi-local gradient corrected form proposed by Perdew, Burke, and Ernzerhof (PBE) [33]. This functional form is generally not well suited to correctly take into account the strongly correlated nature of localized electrons. For this reason, we add an effective Hubbard-like term, using the DFT+U scheme [34, 35]. Concretely, we choose for the U-parameter the value of 4.2 eV for the Fe d-states. This chosen value has been shown to lead to a band gap in agreement with the fundamental gap of Fe2O3 [18].

In a previous work [36], we have shown that also a Hubbard U-correction applied to the oxygen p-states is necessary in order to correctly localize electron holes in hematite. For this reason, we have carefully checked all results involving electron holes (see “Electron holes”) and performed those calculations also with a Hubbard U applied to oxygen (U = 4 eV on oxygen p-states and \(U = 5\) eV on Fe d-states). The result is that for the slab geometries considered here, there is no significant difference between the results when either of the two different schemes is employed.

The Kohn–Sham (KS) orbitals and the charge density are represented using a plane-wave basis set. The kinetic energy cutoff for the orbitals is 40 Ry, and for the charge density 320 Ry. It has been carefully checked that these values lead to well-converged energy differences and forces. The electron–ion interaction is modeled using Vanderbilt’s ultrasoft pseudopotentials (USPP) [37]. We use the pseudopotentials Fe.pbe-sp-van_ak.UPF, O.pbe-van_ak.UPF, and H.pbe-van_ak.UPF for iron, oxygen, and hydrogen, respectively. The pseudopotentials are available from the quantum-ESPRESSO Web site. The hematite surfaces are modeled using a 1×1 slab model, with 24 Å of vacuum between the periodic images of the surface. The Brillouin zone is sampled with a 4× 4 ×1 k-point grid.


Oxo- and peroxo reaction intermediates

Two possible configurations that involve an oxygen atom adsorbed to the hematite surface have been identified in Ref. [17] as reaction intermediates in the overpotential-determining step. Both configurations have a very similar energy (the energy difference is 32 meV) and are potentially present following the deprotonation of a surface hydroxyl group during the OER, see step B and \(B'\) in Fig. 1 and Eq. (2). The similarity in energy is in spite of the fact that the bonding patterns are drastically different in the two cases. In the configuration \(S_{3}\), see Fig. 2a, the adsorbed oxygen atom is bonded only to an undercoordinated Fe surface atom, while in configuration \(S_{3}^{\prime }\), see Fig. 2b, this oxygen atom is bonded to the surface Fe atom and to a lattice oxygen from hematite. This different bonding pattern drastically alters the electronic structure, a fact which is manifest in the different spin-polarization characteristics of the two configurations, which are shown in Fig. 2. Clearly, the anti-ferromagnetic nature of the hematite slab is visible, where Fe atoms in adjacent layers have opposite directions of magnetizations. Interesting is the case of the undercoordinated Fe atom at the surface. Its spin polarization is by 1.18 \(\mu _{B}\) larger in the peroxo-case (\(S_{3}^{\prime }\)), indicating a drastic electronic re-arrangement between the two configurations. Overall, the slab with the peroxo-group has a zero magnetic moment, while the oxo-terminated surface (S3) has a net magnetic moment of 1.89 \(\mu _{B}\). The adsorbed oxygen atom is essentially not spin-polarized in the configuration \(S_{3}^{\prime }\), but shows a net magnetic moment of 0.1 \(\mu _{B}\) in the case \(S_{3}\). This non-magnetic nature of the oxygen species together with the bond length of approximately 1.47 Å confirms its characterization as a peroxo group.
Fig. 2

Configurations corresponding to the reaction intermediates S3 a and \(S_{3}^{\prime }\) b shown in Fig. 1. The isosurfaces shown correspond to the spin density, where yellow is spin-up, and light blue is spin-down. Color scheme for atomic spheres: Fe (blue), O (white)

Oxygen bonding pathway

Given that both the oxo- and the peroxo-terminated surfaces are likely reaction intermediates following step B in Eq. 2, one important question is the possibility of interconversion between the two configurations.

How difficult is it to break the peroxo-bond between the adsorbed oxygen atom and the lattice oxygen? In order to answer this question, we performed nudged elastic band (NEB) [38] calculations of the reaction path connecting the two configurations. We included 17 images in the path, and relaxed the coordinates with a force convergence threshold of 0.05 eV/Å. The resulting minimum energy path is shown in Fig. 3. It results that the activation energy for the conversion between \(S_{3}\) and \(S_{3}^{\prime }\) is 0.86 eV, a high value that makes thermal crossing between the two configurations highly unlikely. This high barrier can be interpreted as a consequence of the drastically different electronic and spin structure of the two configurations, as discussed in “Oxo- and peroxo reaction intermediates”. Interestingly, the transition state is characterized by a spin magnetization of the surface Fe atom, which closely resembles that of the peroxo configuration (\(S_{3}^{\prime }\)), while the adsorbed oxygen atom is spin-polarized to a similar degree as in the oxo configuration (S3).
Fig. 3

Potential energy surface for transforming configuration S3 into \(S_{3}^{\prime }\), obtained from a NEB calculation

In a real device, the hematite surface is immersed in an aqueous electrolyte. One might argue that the presence of water molecules close to the active site might influence, and perhaps lower, the high reaction barrier found in this study. However, here we show that an electronic spin rearrangement must also go along with a transition between the two intermediates. Water molecules do not have any spin polarization and are therefore unlikely to qualitatively alter the nature or the high energy of the transition state.

Electronic structure from Wannier functions in bulk hematite and in the peroxo configuration \(S_{3}^{\prime }\)

One convenient way to visualize the electronic structure in complex materials is to examine the maximally localized Wannier functions (MLWF) [39]. Usually, those functions closely resemble atomic orbitals and because they exactly span the manifold of occupied orbitals, they allow one to rationalize the nature of a chemical bond. In Fig. 4, we illustrate this with MLWFs from bulk hematite, which indeed resemble d-orbitals on Fe and p-orbitals on oxygen. In the bulk, close to each Fe atom, there are 13 MLWFs, nine of majority spin, and four of the minority spin polarization, leading to a magnetization of roughly 5 \(\mu _{B}\) per Fe atom and an oxidation state Fe3+. For bulk O atoms, one finds eight MLWFs, four of each spin polarization, leading to non spin-polarized oxygens, formally O2−, as expected in Fe2O3.
Fig. 4

Examples of maximally localized Wannier functions in bulk hematite. a Wannier function centered at an iron atom. b Wannier function centered at an oxygen atom. Color scheme for atomic spheres: Fe (blue), O (white)

In the case of the peroxo reaction intermediate in configuration \(S_{3}^{\prime }\), the MLWFs allow one to visualize the nature of such an O-O bond: in Fig. 5 we show the two Wannier functions, which are centered on that bond. It is obvious that they resemble a textbook-case of a single bond, with rotational symmetry around the O-O bond. Moreover, the corresponding spin-up and spin-down MLWFs are essentially identical, explaining the non-magnetic character of this surface termination.
Fig. 5

Maximally localized Wannier functions on the peroxo bond in configuration \(S_{3}^{\prime }\). a Spin-up MLWF. b Spin-down MLWF. Color scheme for atomic spheres: Fe (blue), O (white)

Electron holes

In a photoelectrochemical cell under illumination, pairs of excited electrons and holes are created in the semiconductor material. During the OER, in each PCET step, one electron is transferred from to reactants to the semiconductor where it neutralizes one hole. For this mechanism to be viable, the holes must be localized close to the semiconductor–water interface, and there in vicinity to the active site [36].

We have investigated the properties of holes in the two key configurations \(S_{3}\) and \(S_{3}^{\prime }\). In Figs. 6 and 7, we show the Kohn–Sham densities of state (DOS) in both cases—with and without holes. In the case with the oxo termination, \(S_{3}\), Fig. 6, in the electrically neutral state, there are three spin-down states localized in the vicinity of the active site. Those orbitals are labeled E1, E2, and E3 and correspond to the highest occupied molecular orbital (HOMO), the lowest unoccupied molecular orbital (LUMO), and the LUMO+ 1, respectively. When an electron is removed from the system, the three corresponding states are E\(_{1}^{\prime }\), E\(_{2}^{\prime }\) and E\(_{3}^{\prime }\), which now form the LUMO, LUMO+ 1 and LUMO+ 2 of the configuration. From this analysis and the plot of E1 and E\(_{1}^{\prime }\) (the molecular orbital which is depopulated in presence of a hole), it is evident that electron holes will tend to be localized close to the active site and therefore be available as a driving force for the next PCET step.
Fig. 6

Left panel: Kohn–Sham density of states for the configuration S3 with an oxo-group. The DOS is plotted for the electrically neutral and for the case with an electron hole. Right panel: isosurfaces of selected frontier orbitals in the neutral case and with an electron hole. Color scheme for atomic spheres: Fe (blue), O (white). Yellow and gray colors of the isosurfaces denote the spin up and spin down states, respectively

Fig. 7

Left panel: Kohn–Sham density of states for the configuration \(S_{3}^{\prime }\) with a peroxo-group. The DOS is plotted for the electrically neutral and for the case with an electron hole. Right panel: isosurfaces of selected frontier orbitals in the neutral case and with an electron hole. Color scheme for atomic spheres: Fe (blue), O (white)

The situation is similar when one considers the surface termination containing the peroxo intermediate (\(S_{3}^{\prime }\), Fig. 7). In this case, one molecular orbital is localized on the peroxo group, it is the HOMO of the neutral system, labeled E in Fig. 7. When an electron is removed from the system, the corresponding state E’ is lying above the Fermi energy. This indicates a localization of electron holes in the \(S_{3}^{\prime }\) configuration on the peroxo bond between the adsorbed oxygen and the lattice oxygen atoms.


We have used density functional theory-based calculations to examine the electronic and magnetic properties of two key reaction intermediates of the OER on a hematite photoanode. The two intermediates are linked to the overpotential determining step in the water splitting process, and are characterized by an adsorbed oxygen atom on an undercoordinated surface iron atom. In one case, the adsorbed oxygen is bonded only to the iron, while in the other case it also forms a peroxo bond with a sub-surface lattice oxygen. Both configurations are very similar in energy.

We show that in spite of their apparent similarity, the two adsorption modes differ very strongly in their electronic structure and spin polarization. These diverse electronic properties manifest themselves also in a high reaction barrier, which makes a thermally activated cross-over between the two configurations highly unlikely.

Both kinds of surface groups are likely to co-exist in a functioning device, are predicted to be thermodynamically feasible reaction intermediates for the water oxidation reaction. We also show here that the electron holes (which are the driving force for the subsequent PCET steps) are localized in both cases on the oxo and the peroxo species, respectively.

Though the direct modeling of reaction barriers involved in the PCET step cannot itself can be obtained, our study can provide indirect but important implications regarding the kinetic barriers across the two pathways involving the oxo and the peroxo intermediates. The reaction proceeding along pathway C\(^{\prime }\) (see Fig. 1) requires breaking of the peroxo bond to form the hydroperoxo reaction intermediate. The evidence of large barrier to break the peroxo bond to go from peroxo to oxo species indirectly implies that the water oxidation reaction proceeding along the pathway C\(^{\prime }\) (though thermodynamically feasible) will also be severely hindered by kinetics. Further, the affinity of the hole to localize on the peroxo species means that it is likely to act as a hole trapping site and merely a spectator to the water oxidation reaction. Other reaction pathways involving this species as a spectator, have already been discussed in Ref. [17]. Quantitative estimation of reaction kinetic barrier, however, requires a computational tool, which is capable of simulating the electronic dynamics along the reaction coordinate, which is one of the goals of future research. It will also be interesting to apply the analysis shown here to other transition metal oxide systems, where similar reaction intermediates might exist and where other (or absent) magnetic orderings might lead to different conclusions.



L. Tchibota Poaty is greatful to the OFID Postgraduate Fellowship Programme at ICTP and to the ICTP-IAEA Sandwich Training Educational Programme under which this work has been performed.


  1. 1.
    Grave DA, Yatom N, Ellis DS, Toroker MC, Rothschild A (2018) The ”Rust” Challenge: On the Correlations between Electronic Structure, Excited State Dynamics, and Photoelectrochemical Performance of Hematite Photoanodes for Solar Water Splitting. Advanced Materials, pp 1706577Google Scholar
  2. 2.
    Sivula K, Formal FL, Grȧtzel M (2011) Solar water splitting: progress using hematite (α-Fe(2) O(3)) photoelectrodes. ChemSusChem 4(4):432–49CrossRefGoogle Scholar
  3. 3.
    Barroso M, Pendlebury SR, Cowan AJ, Durrant JR (2013) Charge carrier trapping, recombination and transfer in hematite (α-Fe2O3) water splitting photoanodes. Chem Sci 4(7):2724CrossRefGoogle Scholar
  4. 4.
    Formal FL, Pastor E, Tilley SD, Mesa CA, Pendlebury SR, Grȧtzel M, Durrant JR (2015) Rate Law Analysis of Water Oxidation on a Hematite Surface. Journal of the American Chemical SocietyGoogle Scholar
  5. 5.
    Zhou Z, Liu J, Long R, Li L, Guo L, Prezhdo OV (2017) Control of charge carriers trapping and relaxation in hematite by oxygen vacancy charge: ab initio non-adiabatic molecular dynamics. J Am Chem Soc 139 (19):6707–6717CrossRefGoogle Scholar
  6. 6.
    Zhang M, Luo W, Li Z, Yu T, Zou Z (2010) Improved photoelectrochemical responses of Si and Ti codoped \(\alpha \)-Fe2O3 photoanode films. Appl Phys Lett 97(4):042105CrossRefGoogle Scholar
  7. 7.
    Meng XY, Qin GW, Li S, Wen XH, Ren YP, Pei WL, Zuo L (2011) Enhanced photoelectrochemical activity for Cu and Ti-doped hematite: the first principles calculations. Appl Phys Lett 98(11):112104CrossRefGoogle Scholar
  8. 8.
    Liao P, Keith JA, Carter EA (2012) Water oxidation on pure and doped hematite (0001) surfaces: prediction of Co and Ni as effective dopants for electrocatalysis. J Am Chem Soc 134(32):13296–309CrossRefGoogle Scholar
  9. 9.
    Liao P, Carter EA (2012) Hole transport in pure and doped hematite. J Appl Phys 112(1):013701CrossRefGoogle Scholar
  10. 10.
    Lin Y, Sa Z, Sheehan SW, Wang D (2011) Nanonet-based hematite heteronanostructures for efficient solar water splitting. J Am Chem Soc 133(8):2398–401CrossRefGoogle Scholar
  11. 11.
    Malara F, Fabbri F, Marelli M, Naldoni A (2016) Controlling the Surface Energetics and Kinetics of Hematite Photoanodes Through Few Atomic Layers of NiO x. ACS Catalysis, pp 3619–3628Google Scholar
  12. 12.
    Kiejna A, Pabisiak T (2012) Surface properties of clean and Au or Pd covered hematite (α-Fe(2)O(3)) (0001). J Phys Condens Matter: Instit Phys J 24(9):095003CrossRefGoogle Scholar
  13. 13.
    Seriani N (2017) Ab initio simulations of water splitting on hematite. J Phys: Condens Matter 29(46):463002Google Scholar
  14. 14.
    Trainor TP, Chaka AM, Eng PJ, Newville M, Waychunas GA, Catalano JG, Brown GE (2004) Structure and reactivity of the hydrated hematite (0001) surface. Surf Sci 573(2):204–224CrossRefGoogle Scholar
  15. 15.
    Hellman A, Pala RGS (2011) First-principles study of photoinduced water-splitting on Fe2O3. J Phys Chem C 115(26):12901–12907CrossRefGoogle Scholar
  16. 16.
    Nguyen M-T, Seriani N, Piccinin S, Gebauer R (2014) Photo-driven oxidation of water on \(\alpha \)-Fe2O3 surfaces: an ab initio study. J Chem Phys 140(6):064703CrossRefGoogle Scholar
  17. 17.
    Ulman K, Nguyen M-T, Seriani N, Piccinin S, Gebauer R (2017) A unified picture of water oxidation on bare and gallium oxide-covered hematite from density functional theory. ACS Catal 7(3):1793–1804CrossRefGoogle Scholar
  18. 18.
    Nguyen M-T, Seriani N, Gebauer R (2013) Water adsorption and dissociation on \(\alpha \)-Fe2O3(0001): PBE+U calculations. J Chem Phys 138(19):194709CrossRefGoogle Scholar
  19. 19.
    Nguyen M-T, Gebauer R (2014) Graphene supported on hematite surfaces a density functional study. J Phys Chem C 118(16):8455–8461CrossRefGoogle Scholar
  20. 20.
    Nguyen M-T, Seriani N, Gebauer R (2014) Defective \(\alpha \)-Fe2 O3 (0001): an ab initio study. Chemphyschem: Eur J Chem Phys Phys Chem 15(14):2930–5CrossRefGoogle Scholar
  21. 21.
    Nguyen M-T, Piccinin S, Seriani N, Gebauer R (2015) Photo-oxidation Of water on defective hematite(0001). ACS Catal 5(2):715–721CrossRefGoogle Scholar
  22. 22.
    Nguyen M-T, Camellone MF, Gebauer R (2015) On the electronic, structural, and thermodynamic properties of Au supported on \(\alpha \)-Fe2O3 surfaces and their interaction with CO. J Chem Phys 143(3):034704CrossRefGoogle Scholar
  23. 23.
    Nguyen M-T, Seriani N, Gebauer R (2015) Nitrogen electrochemically reduced to ammonia with hematite: density functional insights. Phys Chem Chem Phys: PCCP 17(22):14317–22CrossRefGoogle Scholar
  24. 24.
    Ulman K, Nguyen M-T, Seriani N, Gebauer R (2016) Passivation of surface states of α-Fe2O3(0001) surface by deposition of Ga2O3 overlayers: a density functional theory studyGoogle Scholar
  25. 25.
    Nakamura R, Nakato Y (2004) Primary intermediates of oxygen photoevolution reaction on TiO2 (rutile) particles, revealed by in situ FTIR absorption and photoluminescence measurements. J Am Chem Soc 126(4):1290–8CrossRefGoogle Scholar
  26. 26.
    Zhang M, de Respinis M, Frei H (2014) Time-resolved observations of water oxidation intermediates on a cobalt oxide nanoparticle catalyst. Nat Chem 6(4):362–367Google Scholar
  27. 27.
    Zandi O, Hamann TW (2016) Determination of photoelectrochemical water oxidation intermediates on haematite electrode surfaces using operando infrared spectroscopy. Nat Chem 8(8):778–783CrossRefGoogle Scholar
  28. 28.
    Hohenberg P, Kohn W (1964) Inhomogeneous electron gas. Phys Rev 136(3B):B864–B871CrossRefGoogle Scholar
  29. 29.
    Kohn W, Sham LJ (1965) Self-consistent equations including exchange and correlation effects. Phys Rev 140(4A):A1133–A1138CrossRefGoogle Scholar
  30. 30.
    Giannozzi P, Baroni S, Bonini N, Calandra M, Car R, Cavazzoni C, Ceresoli D, Chiarotti GL, Cococcioni M, Dabo I, Dal Corso A, de Gironcoli S, Fabris S, Fratesi G, Gebauer R, Gerstmann U, Gougoussis C, Kokalj A, Lazzeri M, Martin-Samos L, Marzari N, Mauri F, Mazzarello R, Paolini S, Pasquarello A, Paulatto L, Sbraccia C, Scandolo S, Sclauzero G, Seitsonen AP, Smogunov A, Umari P, Wentzcovitch RM (2009) QUANTUM ESPRESSO: a modular and open-source software project for quantum simulations of materials. J Phys Condens Matter: Inst Phys J 21(39):395502Google Scholar
  31. 31.
    Giannozzi P, Andreussi O, Brumme T, Bunau O, Buongiorno Nardelli M, Calandra M, Car R, Cavazzoni C, Ceresoli D, Cococcioni M, Colonna N, Carnimeo I, Dal Corso A, de Gironcoli S, Delugas P, DiStasio RA, Ferretti A, Floris A, Fratesi G, Fugallo G, Gebauer R, Gerstmann U, Giustino F, Gorni T, Jia J, Kawamura M, Ko H-Y, Kokalj A, Küçükbenli E, Lazzeri M, Marsili M, Marzari N, Mauri F, Nguyen NL, Nguyen H-V, Otero-de-la Roza A, Paulatto L, Poncé S, Rocca D, Sabatini R, Santra B, Schlipf M, Seitsonen AP, Smogunov A, Timrov I, Thonhauser T, Umari P, Vast N, Wu X, Baroni S (2017) Advanced capabilities for materials modelling with Quantum ESPRESSO. J Phys: Condens Matter 29(46):465901Google Scholar
  32. 32.
    Mostofi AA, Yates JR, Pizzi G, Lee Y-S, Souza I, Vanderbilt D, Marzari N (2014) An updated version of Wannier90. A tool for obtaining maximally-localised Wannier functions. Comput Phys Commun 185(8):2309–2310CrossRefGoogle Scholar
  33. 33.
    Perdew JP, Burke K, Ernzerhof M (1996) Generalized gradient approximation made simple. Phys Rev Lett 77(18):3865–3868CrossRefGoogle Scholar
  34. 34.
    Cococcioni M, de Gironcoli S (2005) Linear response approach to the calculation of the effective interaction parameters in the LDA+U method. Phys Rev B 71(3):035105Google Scholar
  35. 35.
    Anisimov VI, Zaanen J, Andersen OK (1991) Band theory and Mott insulators: Hubbard U instead of Stoner I. Phys Rev B 44(3):943–954CrossRefGoogle Scholar
  36. 36.
    Ansari N, Ulman K, Camellone MF, Seriani N, Gebauer R, Piccinin S (2017) Hole localization in Fe2O3 from density functional theory and wave-function-based methods. Phys Rev Mater 1(3):035404CrossRefGoogle Scholar
  37. 37.
    Vanderbilt D (1990) Soft self-consistent pseudopotentials in a generalized eigenvalue formalism. Phys Rev B 41(11):7892–7895CrossRefGoogle Scholar
  38. 38.
    Henkelman G, Uberuaga BP, Jónsson H (2000) A climbing image nudged elastic band method for finding saddle points and minimum energy paths. J Chem Phys 113(22):9901–9904CrossRefGoogle Scholar
  39. 39.
    Marzari N, Vanderbilt D (1997) Maximally localized generalized Wannier functions for composite energy bands. Phys Rev B 56(20):12847–12865CrossRefGoogle Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2018

Authors and Affiliations

  • Lodvert Tchibota Poaty
    • 1
    • 2
  • Kanchan Ulman
    • 1
  • Nicola Seriani
    • 1
  • Bernard M’Passi-Mabiala
    • 2
    • 3
  • Ralph Gebauer
    • 1
    Email author
  1. 1.The Abdus Salam International Centre for Theoretical Physics (ICTP)TriesteItaly
  2. 2.Groupe de Simulations numériques en Magnétisme et CatalyseUniversité Marien Ngouabi, Faculté des Sciences et TechniquesBrazzavilleCongo
  3. 3.Unité de Recherche en Nanomatériaux et NanotechnologiesInstitut National de Recherche en Sciences Exactes et Naturelles (IRSEN)BrazzavilleCongo

Personalised recommendations