WorksPhysics and Astronomy
Polaron Conductivity in -Fe2O3 Quenched by Adsorbed NO2
Abstract
Polaron-mediated charge transport in α-Fe2O3 plays a central role in its performance as a gas-sensing material, yet the atomistic interaction between surface adsorbates and polarons remains insufficiently understood. Here, density functional theory with Hubbard-U correction (DFT+U) combined with nudged elastic band calculations is used to investigate polaron formation, migration, and quenching at the Fe-terminated α-Fe2O3 (0001) surface. The calculated activation energy for small-polaron hopping in bulk α-Fe2O3 is found to be 0.12 eV, in excellent agreement with experimental measurements, confirming the validity of the computational approach. Slab calculations show that migration of the polaron from bulk to the surface lowers the energy by 0.12 eV, indicating preferential localization of charge carriers at the gas-solid interface. Adsorption of NO2 induces substantial electron transfer (0.72 e-) from the oxide to the molecule, eliminating the localized Fe2+ polaron state and thereby suppressing polaronic conductivity. These results provide a direct microscopic explanation for the resistance increase of hematite-based sensors upon exposure to oxidizing gases. More broadly, the study establishes how surface adsorption can modulate charge transport α-Fe2O3 through control of polaron populations, offering design principles for improved iron oxide gas sensors.
Introduction
Iron oxides are widely studied transition metal oxides for sensing applications due to their chemical stability, abundance, and rich electronic properties [1, 2, 3]. In particular, (hematite) has attracted significant attention as a functional material for gas sensing, photocatalysis, and energy-related applications [4, 5, 6]. Its sensitivity to oxidizing and reducing gases is largely governed by surface-mediated charge transfer processes, which directly influence electrical conductivity [7]. Understanding the microscopic mechanisms underlying charge transport and its interaction with adsorbates is therefore essential for improving sensing performance.
Charge transport in is dominated by small polaron hopping rather than band-like conduction [8, 9]. In this mechanism, charge carriers localize on Fe sites, forming centers, and migrate via thermally activated hopping between neighboring lattice sites. The activation energy and mobility of these polarons are key parameters controlling the electrical response of hematite-based sensors. Previous studies have shown that polaron transport can be strongly affected by structural distortions, defect states, and surface interactions, which are all highly relevant under sensing conditions.
From a sensing perspective, has demonstrated promising performance for detection of gases such as , acetone, and other volatile compounds [10, 11]. The sensing mechanism is typically attributed to adsorption-induced charge transfer, which modifies the concentration or mobility of charge carriers near the surface. In particular, oxidizing gases such as are known to extract electrons from the oxide, thereby increasing the resistance of metal oxide semiconductors with n-type characteristics [12]. However, a detailed atomistic understanding of how such adsorption processes interact specifically with polaronic charge carriers remains incomplete.
In earlier experimental work, it was observed that iron oxide systems can initially form (maghemite) rather than [11]. While is often metastable, it is well established that it transforms into the thermodynamically stable phase, especially at elevated temperatures relevant for gas sensing operation [1, 13]. This phase transition underscores the importance of understanding charge transport and surface interactions in , as it is the phase most likely to dominate under realistic device conditions.
Despite extensive experimental studies, there remains a need for atomistic insight into the coupling between polaron transport and gas adsorption at surfaces. In particular, the extent to which adsorbates can modify or suppress polaron conduction is a critical question for interpreting sensing signals. Density functional theory (DFT) combined with harmonic transition state theory provides a powerful framework to address this issue by explicitly resolving charge localization, migration pathways, and adsorption-induced electronic changes [14].
In this work, we investigate polaron formation and migration in bulk – and at the Fe-terminated (0001) surface using DFT+U calculations. The activation energy for polaron hopping is determined using the nudged elastic band method. Furthermore, the interaction between surface polarons and molecules is analyzed to elucidate the microscopic origin of sensing behavior. The results provide direct insight into how adsorbates can quench polaron conductivity, thereby establishing a clear link between atomistic charge transport and macroscopic sensing response.
Methods and Model
The calculations were performed using density functional theory (DFT) employing the VASP software [15, 16] using the PBE exchange correlation functional [17] as well as the PBE+U approach. The valence electrons were described using a plane wave basis with a 500 eV cutoff while the effect of inner electrons was described using the PAW formalism. The addition of a Hubbard U term helps reduce the effect of the self-interaction error, which artificially increases the energy of localized electronic states with respect to localized states; without this correction, polaronic states are unstable. We chose the value eV as in earlier calculations of – [18]. In the structural relaxations, the energy was optimized with respect to atomic positions until the forces on the ions in the bulk unit cells were converged to 0.01 eV/Å. The criterion for convergence of the self-consistent charge density was a tolerance on the energy change of eV.
Bulk properties of – were calculated using a unit cell with k-point sampling. In order to find the optimized lattice parameter, the equilibrium cell shape for different volumes was calculated; the lattice parameters were extracted from the equilibrium cell shape that corresponds to the lowest energy.
A conventional periodic slab model was used for simulating the Fe-terminated (0001) surface. A slab of supercell was used for (0001) – surface. Supercells were generated using the optimized lattice parameters of the bulk –. In order to ensure that the periodic images do not interact, the box length along the -direction was set to ensure at least 12 Å of vacuum between the slab and its periodic images, and the -point sampling was reduced to . slabs were chosen in such a way that the top and the bottom layers were equivalent and symmetric with respect to the center of inversion located at the center of the slab. The symmetry prevented unphysical dipolar interaction between periodic images of the slab.
A polaron was formed by adding an excess electron in a super-cell, , with -point sampling. The excess negative charge was compensated by a uniform positive background.
The nudged elastic band (NEB) method [19, 20] with the climbing image extension [21] was used to find the activation energy for polaron migration.
Results
The calculated bulk atomic structure and the orientation of the magnetic moments of the Fe atoms is shown in Figure 1. In order to find the optimal lattice parameter, the lowest energy cell shape was identified. Calculated lattice parameters (Figure 1) are tabulated in Table I and compared with experimental values as well as PBE values without the U term. The agreement with experiment is good, within 1%.

Figure 1. (a) Atomic structure of the unit cell of the crystal and corresponding orientation of the magnetic moments. Brown and red spheres represent Fe and O atoms, respectively. (b) Energy of - with respect to volume change monitored as the change of the lattice constant .
| Structural Parameter | PBE Calculation | PBE+U () Calculation | Experiments [22] |
| (Å) | 5.02 | 5.08 | 5.04 |
| (Å) | 5.02 | 5.08 | 5.04 |
| (Å) | 13.91 | 13.90 | 13.77 |
| (deg) | 90.0 | 90.0 | 90.0 |
| (deg) | 90.0 | 90.0 | 90.0 |
| (deg) | 120.0 | 120.0 | 120.0 |
Table I. Calculated and experimental cell parameters of -FeO bulk using a unit cell of FeO.
The iron atoms in -FeO were found to be ferromagnetically coupled within the basal plane but were antiferromagnetically coupled along the c-axis, as shown in Figure 1. All the iron atoms in -FeO were in the oxidation state. The calculated magnetic moment of a Fe ions was 4.18 , similar to previous calculations at the same level of theory [18]. This also compares well with the experimental value of 4.6 . However, calculations using the PBE functional without the U term gave a magnetic moment of only 3.60 .
In order to compute the ground state properties of -FeO in the presence of a polaron, an excess electron was added to a -FeO supercell and the size of the simulation box was optimized again. The structure was fully relaxed to compute lattice properties in presence of the excess charge; the optimized lattice parameters are tabulated in Table II. The presence of the polaron had negligible effect on the simulation box size, indicating that the latter was large enough. Results of calculations using the PBE functional without the U term are also shown in Table II for comparison. The excess electron was localized at an iron site and reduced one of the Fe ions to an Fe ion. The electron density difference is shown in Figure 2, where the presence of the excess electron breaks the symmetry of the lattice. The distortion of the bond lengths is observed near the Fe ion, where the polaron is localized. The angles of the simulation cell distort slightly in the presence of the polaron, as listed in Table II.
| Structural Parameter | PBE Calculation | PBE+U (U=4.3) Calculation |
| (Å) | 10.04 | 10.15 |
| (Å) | 10.04 | 10.15 |
| (Å) | 13.93 | 13.90 |
| (deg) | 90.04 | 90.10 |
| (deg) | 89.96 | 90.00 |
| (deg) | 119.95 | 120.00 |
Table II. Calculated cell parameters of α-FeO in the presence of a polaron in the system, calculated using both PBE and PBE+U.

Figure 2. Atomic structure and electron density difference of the polaron in α-FeO calculated using PBE+U ((a) and (b)), and using PBE without the U term ((c) and (d)). (a) and (c): supercell with an excess electron viewed along the c-axis. The electron density difference, , is shown, where blue indicates increased density and yellow indicates decreased density (where the overscore identifies the structure with the polaron present). The excess electron is located at the atom labeled Fe0. The distances to other iron atoms in the same basal plane, labeled Fe1, Fe2 and Fe3, are marked and indicated with blue dashed line. (b) and (d) Length of bonds between Fe and O atoms near the Fe ion.
The minimum energy path for the migration of the polaron from one Fe ion to an adjacent one within the basal plane was calculated using the NEB method. The results are shown in Figure 3. The rise in the energy along the minimum energy path gives an activation energy of 0.12 eV. This is in excellent agreement with the experimental estimate of 0.118 eV [23].

Figure 3. Minimum energy path found using NEB calculations for polaron migration between two adjacent Fe ions, Fe0 and Fe1, within the same basal plane (see fig:2a) in the crystal.
An excess electron was then added to the slab to generate a polaron at the Fe terminated (0001) surface. The energy of the polaron is lower at the surface than in the interior of the crystal by 0.12 eV. A polaron will, therefore, tend to migrate to the surface due to energetic preference, but entropic effects will counter that to some extent due to the large number of sites within the crystal compared with the number of surface sites (fig:4a). An NO molecule was then adsorbed on the (0001) surface at the site of the polaron (fig:4b). The electron transfer, as determined by Bader analysis, is 0.72 electrons (see Table III). The presence of the admolecule therefore removes the polaron from the α-FeO. This would largely eliminate the electrical conductivity due to presence of polarons preferentially located at surface sites.
| Electron gain by NO | Electron loss by Fe | |||||
| O1 | N | O2 | Fe0 | Fe1 | Fe2 | Fe3 |
| 0.32 | 0.12 | 0.28 | 0.48 | 0.07 | 0.06 | 0.07 |
Table III. Change in electron density integrated over Bader volumes due to adsorption of NO on Fe-terminated (0001)– -FeO surface with a polaron. The total negative charge on the NO is 0.72 electrons.

Figure 4. (a) Electron density difference, , between a charged and a neutral (0001)-FeO surface, due to the presence of excess electron at the Fe-terminated (0001) surface. (b) Electron density difference, , between a NO-adsorbed (0001)--FeO surface with a polaron and a clean (0001)--FeO surface with a polaron plus an isolated NO molecule. Electron densities are illustrated as the isosurfaces with . Blue isosurfaces indicate a net gain of electrons, whereas yellow isosurfaces indicate a net loss of electrons. Fe, O, and N atoms are depicted brown, red, and blue, respectively.
Discussion
The calculated activation energy for polaron migration (0.12 eV) is in close agreement with the experimental estimate, confirming that the DFT+U framework accurately captures the small-polaron transport mechanism in –FeO [8, 9, 14]. This low activation barrier indicates that charge transport is feasible at moderate temperatures, consistent with the operating conditions of many gas sensors based on hematite. From a sensing standpoint, this implies that the electrical conductivity is highly sensitive to perturbations that affect either the polaron population or their mobility. We point out, however, that the PBE+U approximation with U = 4.2 eV gives an underestimation of the band gap by about 0.5 eV. A better approach, but also significantly more computationally intensive, would be to apply explicit self-interaction correction [24, 25] or use a hybrid functional where screened Fock exchange is mixed in with the PBE functional [26].
A key finding of this work is the energetic preference of polarons for the surface relative to the bulk. The stabilization of the polaron by 0.12 eV at the Fe-terminated (0001) surface suggests that charge carriers naturally accumulate near the surface region, where gas adsorption occurs. This is particularly important for sensing applications, as it implies that the active charge carriers are inherently localized at the interface where interaction with analyte molecules takes place. Such surface localization enhances the sensitivity of the material, since even small perturbations in surface chemistry can significantly influence the overall conductivity.
The adsorption of NO has a pronounced effect on the electronic structure of the system. Bader charge analysis [27] shows substantial electron transfer (0.72 e) from the oxide to the adsorbed molecule. This charge transfer effectively removes the localized electron responsible for polaron formation, thereby eliminating the corresponding Fe site. As a result, the polaron conduction pathway is disrupted. This mechanism provides a clear atomistic explanation for the experimentally observed increase in resistance upon exposure to oxidizing gases such as NO.
Importantly, the quenching of polaron conductivity is not merely a local effect but can have a global impact on transport. Since conduction in –FeO occurs via hopping between localized states, the removal of even a fraction of these states can significantly reduce the connectivity of the conduction network. In systems where polarons preferentially reside at the surface, adsorption-induced charge transfer can therefore lead to a strong modulation of electrical resistance, forming the basis of the sensing signal.
These findings also highlight the importance of surface structure and termination. The Fe-terminated (0001) surface stabilizes polarons and facilitates strong interaction with NO. Variations in surface morphology, defect density, or crystallographic orientation could therefore lead to substantial differences in sensing performance. This is consistent with experimental observations that nanostructured –FeO (e.g., nanotubes, nanocubes) exhibits enhanced sensitivity due to increased surface area [3, 10, 11, 28].
Furthermore, the results provide insight into the role of phase stability. While –FeO may initially form in some synthesis routes (such as gas-phase synthesis of nanoparticles [11, 29]), its metastability and transformation to at elevated temperatures imply that the sensing behavior in practical devices will largely be governed by hematite [30], unless it can be ensured that the sensing material remains maghemite during the sensor lifetime. The polaron-based mechanism identified here is therefore directly relevant for interpreting sensing responses under realistic operating conditions.
Conclusion
In this work, density functional theory calculations have been used to investigate polaron formation, migration, and interaction with adsorbates in . The calculated activation energy for polaron hopping is in excellent agreement with experimental measurements, confirming the validity of the small-polaron transport model in this material.
A key result is the preferential stabilization of polarons at the Fe-terminated (0001) surface, which implies that charge carriers are naturally localized in the region where gas adsorption occurs. This surface localization enhances the sensitivity of to adsorbates. The adsorption of oxidizing gases such as leads to significant electron transfer from the oxide to the molecule, effectively removing the polaron and quenching polaron-mediated conductivity.
This mechanism provides a clear microscopic explanation for the sensing behavior of toward oxidizing gases, where adsorption results in increased electrical resistance. The results emphasize the critical role of polaron dynamics in determining sensor response and highlight the importance of surface structure in optimizing sensing performance. To this end, further studies are needed to compare the effect of different surface orientations from a theoretical perspective.
Overall, the findings establish a direct link between atomistic charge transport processes and macroscopic sensing properties in . This understanding can guide the design of improved iron oxide-based sensors through control of surface structure, defect engineering, and phase composition.
Acknowledgments
This work was supported by the Icelandic Research Fund (grant no. 2410644). P.G. was supported by a Beatriz Galindo Senior Fellowship (BG23/00144).
References
- [1]Cornell, R.M., Schwertmann, U. The iron oxides: Structure, properties, reactions, occurrences and uses. Wiley-VCH 2003. https://doi.org/10.1002/3527602097
- [2]Hjiri, M. An overview of room temperature Fe₂O₃ gas sensors. Micro Nanostructures 2025, 206, 208223. https://doi.org/10.1016/j.micrna.2025.208223
- [3]Mirzaei, A., Hashemi, B., Janghorban, K. α–Fe₂O₃ based nanomaterials as gas sensors. J Mater Sci: Mater Electron 2016, 27, 3109–3144. https://doi.org/10.1007/s10854-015-4200-z
- [4]Kusior, A., Synowiec, M., Zakrzewska, K., Radecka, M. Surface-controlled photocatalysis and chemical sensing of TiO₂, α–Fe₂O₃, and Cu₂O nanocrystals. Crystals 2019, 9, 163. https://doi.org/10.3390/cryst9030163
- [5]Kay, A., Cesar, I., Grätzel, M. New benchmark for water photooxidation by nanostructured α-Fe₂O₃ films. J Am Chem Soc 2006, 128, 15714–15721. https://doi.org/10.1021/ja064380l
- [6]Reddy, M.V., Yu, T., Sow, C.-H., Shen, Z.X., Lim, C.T., Subba Rao, G.V., Chowdari, B.V.R. α-Fe₂O₃ nanoflakes as an anode material for Li-ion batteries. Adv Funct Mater 2007, 17, 2792–2799. https://doi.org/10.1002/adfm.200601186
- [7]Aronniemi, M., Saino, J., Lahtinen, J. Characterization and gas-sensing behavior of an iron oxide thin film prepared by atomic layer deposition. Thin Solid Films 2008, 516, 6110–6115. https://doi.org/10.1016/j.tsf.2007.11.011
- [8]Rettie, A.J.E., Chemelewski, W.D., Emin, D., Mullins, C.B. Unravelling small-polaron transport in metal oxide photoelectrodes. J Phys Chem Lett 2016, 7, 471–479. https://doi.org/10.1021/acs.jpclett.5b02143
- [9]Iordanova, N., Dupuis, M., Rosso, K.M. Charge transport in metal oxides: A theoretical study of hematite α–Fe₂O₃. J Chem Phys 2005, 122, 144305. https://doi.org/10.1063/1.1869492
- [10]Kim, D.H., Shim, Y.-S., Jeon, J.-M., Jeong, H.Y., Park, S.S., Kim, Y.- W., Kim, J.-S., Lee, J.-H., Jang, H.W. Vertically ordered hematite nanotube array as an ultrasensitive and rapid response acetone sensor. ACS Appl Mater Interfaces 2014, 6, 14779–14784. https://doi.org/10.1021/am504156w
- [11]Vernieres, J., Steinhauer, S., Zhao, J., Chapelle, A., Menini, P., Dufour, N., Diaz, R.E., Nordlund, K., Djurabekova, F., Grammatikopoulos, P., Sowwan, M. Gas phase synthesis of multifunctional Fe-based nanocubes. Adv Funct Mater 2017, 27, 1605328. https://doi.org/10.1002/adfm.201605328
- [12]Yamazoe, N. Toward innovations of gas sensor technology. Sens Actuators B Chem 2005, 108, 2–14. https://doi.org/10.1016/j.snb.2004.12.075
- [13]Sharma, M., Murugavel, S., Shukla, D.K., De Groot, F.M.F. Reversal in the lattice contraction of α–Fe₂O₃ nanoparticles. J Phys Chem C 2018, 122, 9292–9301. https://doi.org/10.1021/acs.jpcc.8b00550
- [14]Rosso, K.M., Smith, D.M.A., Dupuis, M. An ab initio model of electron transport in hematite (α–Fe₂O₃) basal planes. J Chem Phys 2003, 118, 6455–6466. https://doi.org/10.1063/1.1558534
- [15]Kresse, G., Hafner, J. Ab initio molecular dynamics for liquid metals, Phys Rev B 1993, 47, 558(R). https://doi.org/10.1103/PhysRevB.47.558
- [16]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. https://doi.org/10.1103/PhysRevB.54.11169
- [17]Perdew, J.P., Burke, K. Ernzerhof, M. Generalized gradient approximation made simple, Phys Rev Lett 1996, 77, 3865; Erratum: Phys Rev Lett 1997, 78, 1396. https://doi.org/10.1103/PhysRevLett.77.3865
- [18]Adelstein, N., Neaton, J.B., Asta, M., De Jonghe, L.C. Density functional theory based calculation of small-polaron mobility in hematite, Phys Rev B 2014, 89, 245115. https://doi.org/10.1103/PhysRevB.89.245115
- [19]Jónsson, H., Mills, G., Jacobsen, K.W. Nudged elastic band method for finding minimum energy paths of transitions, in Classical and Quantum Dynamics in Condensed Phase Simulations, edited by Berne, B.J., Ciccotti, G. Coker, D.F., pp. 385-404 (World Scientific, 1998). https://doi.org/10.1142/9789812839664_0016
- [20]Henkelman, G., Jónsson, H. Improved tangent estimate in the nudged elastic band method for finding minimum energy paths and saddle points. J Chem Phys 2000, 113, 9978. https://doi.org/10.1063/1.1323224
- [21]Henkelman, G. Uberuaga, B.P., Jönsson, H. A climbing image nudged elastic band method for finding saddle points and minimum energy paths. J Chem Phys 2000, 113, 9901. https://doi.org/10.1063/1.1329672
- [22]Blake, R., Hessevick, R., Zoltai, T., Finger, L.W. Refinement of the hematite structure. Am Mineral 1966 51(1-2), 123-129.
- [23]Zhao, B., Kaspar, T.C., Droubay, T.C., McCloy, J., Bowden, M.E., Shuthanandan, V., Heald, S.M., Chambers, S.A. Electrical transport properties of Ti-doped Fe₂O₃(0001) epitaxial films. Phys Rev B 2011, 84, 245325. https://doi.org/10.1103/PhysRevB.84.245325
- [24]Perdew, J.P., Zunger, A. Self-interaction correction to density-functional approximations for many-electron systems. Phys Rev B 1981, 23, 5048. https://doi.org/10.1103/PhysRevB.23.5048
- [25]Jönsson, H. Simulation of surface processes. Proc Natl Acad Sci U S A 2011, 108:3, 944-949. https://doi.org/10.1073/pnas.1006670108
- [26]Heyd, J., Scuseria, G.E., Ernzerhof, M. Hybrid functionals based on a screened Coulomb potential. J Chem Phys 2003, 118, 8207-8215. https://doi.org/10.1063/1.1564060
- [27]Henkelman, G., Arnaldsson, A., Jönsson, H. A fast and robust algorithm for Bader decomposition of charge density, Comput Mater Sci 2006, 36:3, 354-360. https://doi.org/10.1016/j.commatsci.2005.04.010
- [28]Liu, X., Cheng, S., Liu, H., Hu, S., Zhang, D., Ning, H. A survey on gas sensing technology. Sensors 2012, 12, 9635–9665. https://doi.org/10.3390/s120709635
- [29]Vernieres. J., Steinhauer, S., Zhao, J., Grammatikopoulos, P., Ferrando, R., Nordlund, K., Djurabekova, F., Sowwan, M. Site-specific wetting of iron nanocubes by gold atoms in gas-phase synthesis. Adv Sci 2019, 6:13, 1900447. https://doi.org/10.1002/advs.201900447
- [30]Gallagher, K.J., Feitknecht, W., Mannweiler, U. Mechanism of oxidation of magnetite to γ-Fe₂O₃ and conversion to α-Fe₂O₃. Nature 1968, 217, 1118–1121. https://doi.org/10.1038/2171118a0