betaFeedback

WorksPhysics and Astronomy

Challenging the pp-type Paradigm: Intrinsic nn-type Mobility in Antiferromagnetic Cr2_2O3_3

Abstract

Chromium oxide (Cr2_2O3_3) is widely considered a pp-type transparent conducting oxide despite ongoing debate regarding its intrinsic transport character. Here, we resolve this question by computing phonon-limited electron and hole mobilities using the ab initio Boltzmann transport equation. We find that electron mobility systematically exceeds hole mobility over a wide temperature range, demonstrating that Cr2_2O3_3 is intrinsically nn-type. Analysis of scattering mechanisms reveals that scattering with phonons affects electrons and holes similarly, and that the mobility asymmetry originates from the electronic structure, namely the larger effective mass and multi-valley character of the valence band. The intrinsic nn-type character, combined with moderate hole mobility, enables bipolar transport and revises the role of Cr2_2O3_3 in transparent electronics. Additionally, our results on mobility complement previous studies on defect formation indicating that the commonly observed pp-type behavior is extrinsic. These insights provide a complete chemical-transport paradigm for Cr2_2O3_3, re-evaluating its role in functional transparent electronic and magneto-optoelectronic applications

Main.pdf

Introduction

Transparent conducting oxides (TCO) have attracted interest in the last decade due to their potential in optoelectronic applications [1, 2, 3, 4]. These types of materials offer good electrical conductivity and are transparent in the visible range, making them interesting candidates for use on screens in mobile phones, tablets, or touchscreens [5, 6]. The most commonly used materials are nn-type semiconductors offering high conductivity and transparency, and a more stable doping [7, 8] [9, 10]. As a consequence, this means that in most practical applications of TCOs, nn-type semiconductors are preferred. Nonetheless, having a pp-type TCO is crucial for making pp−nn junctions for transparent light emitting diodes [11, 12], transparent photovoltaics [13–16] or to obtain transparent photo-detectors, logic devices and complementary metal oxide semiconductors (CMOS) [17–19]. Compounds such as Cr2O3\mathrm{Cr}_2\mathrm{O}_3 have been proposed to be good candidates for pp-type doping [20, 21], even suggesting Cr2O3\mathrm{Cr}_2\mathrm{O}_3 to be an intrinsic pp-type semiconductor [22, 23].

In that regard, several experimental groups have been successful in this pursuit, synthesizing pp-type Cr2O3\mathrm{Cr}_2\mathrm{O}_3 samples [23–30]. Nonetheless, it has been reported that pp-type doping of Cr2O3\mathrm{Cr}_2\mathrm{O}_3 is challenging [26–28] and that nn-type doping is also possible [30, 31], as well as some conflicting results with respect to oxygen deficiency and fabrication process [31–34]. Surprisingly, there are few theoretical studies [21, 22] that address the conducting properties of Cr2O3\mathrm{Cr}_2\mathrm{O}_3. Thus, the question of the intrinsic nature of Cr2O3\mathrm{Cr}_2\mathrm{O}_3 remains open.

In this study, we use spin polarized Density Functional Theory (DFT) calculations and the Boltzmann transport equation (BTE) [57–59], a methodology previously unexplored for this system, to compute the electron and hole mobilities of bulk antiferromagnetic Cr2O3\mathrm{Cr}_2\mathrm{O}_3. We verify its applicability in Supplementary Material S1 [60]. We find that (i) the electron mobility of Cr2O3\mathrm{Cr}_2\mathrm{O}_3 is consistently higher than the hole mobility in the 100-500 K temperature range, implying that Cr2O3\mathrm{Cr}_2\mathrm{O}_3 is an intrinsic nn-type semiconductor; (ii) the hole mobility (μh\mu^{\mathrm{h}}) of Cr2O3\mathrm{Cr}_2\mathrm{O}_3 is between 1−4 cm2/(V⋅s)1 - 4\ \mathrm{cm}^2/(\mathrm{V}\cdot\mathrm{s}) at room temperature, which in the context of pp-type TCOs makes it competitive with already established materials; and (iii) the conducting easy axis aligns with the magnetic one.

These findings provide an unexpected perspective on Cr2O3\mathrm{Cr}_2\mathrm{O}_3, shifting its description from a strictly pp-type compound to an intrinsic nn-type wide band gap semiconductor. Crucially, as we show in Table I, Cr2O3\mathrm{Cr}_2\mathrm{O}_3 exhibits

Materialμe (cm2/(V⋅s))\mu^{\mathrm{e}}\ (\mathrm{cm}^2/(\mathrm{V}\cdot\mathrm{s}))μh (cm2/(V⋅s))\mu^{\mathrm{h}}\ (\mathrm{cm}^2/(\mathrm{V}\cdot\mathrm{s}))μf/μs\mu^{\mathrm{f}}/\mu^{\mathrm{s}}Eg (eV)E_{\mathrm{g}}\ (\mathrm{eV})
In2_2O3_350-100 [35]0.1<< [36]500>>2.7 [37, 38]
SnO2_220-200 [39]1-4 [39]5-503.7 [40]
ITO50-100 [41]0.1<< [42]500>>4.0 [40]
ZnO10-200 [43]0.1<< [44, 45]100>>3.4 [46]
Ga2_2O3_3100-200 [47]1-5 [48]40-1004.6-4.8 [49]
IGZO20-50 [50]--3.0-3.2 [51]
CuAlO2_2-0.1-1 [52, 53]-3.5 [54]
SnO0.2 [42]1-7 [55]5-152.7 [56]
This work
Cr2O3\mathrm{Cr}_2\mathrm{O}_35-101-42-33.1 [16]

Table I. Electrical and optical properties, Band gap EgE_{\mathrm{g}}, electron (hole) mobilities μe\mu^{\mathrm{e}} (μh\mu^{\mathrm{h}}) and ratio μf/μs\mu^{\mathrm{f}}/\mu^{\mathrm{s}} between the fastest μf\mu^{\mathrm{f}} and slowest μs\mu^{\mathrm{s}} carrier type mobility, of common TCO materials compared with Cr2O3\mathrm{Cr}_2\mathrm{O}_3, where ITO is In2_2O3_3 and SnO2_2, and IGZO is In2_2O3_3, Ga2_2O3_3 and ZnO.

a near unity electron-hole mobility ratio μe/μh=2−3\mu_e/\mu_h = 2-3 absent in all high-performance nn-type TCOs. This intrinsic transport parity provides a unique opportunity for balanced charge extraction in all-oxide bipolar architectures. By establishing a solid link between electronic transport properties and the underlying thermodynamic defect landscape, this work outlines targeted synthesis routes to exploit the true, intrinsic mobile carrier channels of this stable oxide. Furthermore, the alignment of the magnetic and conducting easy axes, provides a unique platform for magneto-optoelectronic tuning, bridging the gap between transparent circuitry and antiferromagnetic spintronics.

Methods

We use DFT with the Quantum ESPRESSO v7.6 software [61, 62]. Scalar-relativistic norm-conserving pseudopotentials [63] from the standard accuracy table of the PseudoDojo project [64] are used with the Pardew Burke Ernzerhof (PBE) functional [65]. The plane-wave cutoff is set to 90 Ry, a total energy convergence threshold of 10−1210^{-12} Ry, a 434^3 kk-point grid, and a 636^3 qq-point grid for phonons. We relax the crystal structure in its R3cR3c ground state geometry with the Néel antiferromagnetic order (AFM) as shown in Figure 1(a), obtaining a=4.948a = 4.948 Å and c=13.781c = 13.781 Å lattice parameters, close to the experimental values of aexp=4.953a_{\mathrm{exp}} = 4.953 Å and cexp=13.578c_{\mathrm{exp}} = 13.578 Å [66]. The phonons are computed with density functional perturbation theory (DFPT) [67–70]. The maximally localized Wannier functions are computed using the wannier90 package version 3.1 [71] with Cr dd and O pp initial projections. The phonon limited transport is computed with the EPW package [59, 72, 73] which will be released in the v6.1. Mobilities are converged within 1% using 636^3 coarse and 32332^3 fine k/qk/q grids. We provide all data to re-produce the results on the Materials Cloud Archive [60].

Image not extracted

Figure 1. (a) Crystal structure of antiferromagnetic Cr2_2O3_3 with the arrows depicting the collinear magnetic order. (b) Intrinsic electron (in-plane μ∥e\mu_{\parallel}^{\mathrm{e}}, dash mauve, and out-of-plane μ⊥e\mu_{\perp}^{\mathrm{e}}, mauve line) and hole (in-plane μ∥h\mu_{\parallel}^{\mathrm{h}}, dash green, and out-of-plane μ⊥h\mu_{\perp}^{\mathrm{h}}, green line) mobility for bulk polycrystalline Cr2_2O3_3 as a function of temperature obtained with the Boltzmann transport equation. Solid star symbols correspond to the experimental drift mobility from Refs. [77, 78] at room temperature.

The intrinsic electron and hole mobility of Cr2_2O3_3 are calculated by means of the iterative ab initio BTE [58, 74] in the range of 100-500 K. Although Cr2_2O3_3 presents a Néel temperature of TNeˊel=307T_{\mathrm{N\acute{e}el}} = 307 K [75, 76], there are no significant changes in the experimental conductivity that may signal a different trend [23–30]. Consequently, the given trends in mobility on the AFM phase would still hold in the paramagnetic phase. We consider that each spin channel σ\sigma contributes with a spin resolved mobility tensor μαβσ\mu_{\alpha\beta}^{\sigma} defined as [59]

μαβσ=−eVucncσ∑n∫ΩBZd3kΩBZvnkασ∂Eβfnkσ,(1)\mu_{\alpha\beta}^{\sigma} = \frac{-e}{V_{\mathrm{uc}}n_c^{\sigma}}\sum_n \int_{\Omega_{\mathrm{BZ}}}\frac{d^3k}{\Omega_{\mathrm{BZ}}}v_{n\mathbf{k}\alpha}^{\sigma}\partial_{E_\beta}f_{n\mathbf{k}}^{\sigma}, \tag*{(1)}

where VucV_{\mathrm{uc}} is the volume of the unit cell, ΩBZ\Omega_{\mathrm{BZ}} is the volume of the Brillouin Zone, ∂Eβfnkσ\partial_{E_\beta}f_{n\mathbf{k}}^{\sigma} is the perturbed occupation for an applied electric field EβE_\beta in the direction β\beta, vnkασv_{n\mathbf{k}\alpha}^{\sigma} is the α\alpha component of the spin-resolved electronic band velocity and ncσn_c^{\sigma} is the carrier density of a specific spin channel such that the relation with the conductivity tensor σαβ\sigma_{\alpha\beta} is

σαβ=nc↑μαβ↑+nc↓μαβ↓.(2)\sigma_{\alpha\beta}=n_c^{\uparrow}\mu_{\alpha\beta}^{\uparrow}+n_c^{\downarrow}\mu_{\alpha\beta}^{\downarrow}. \tag*{(2)}

We note that the directly measurable quantities are the total conductivity σαβ\sigma_{\alpha\beta} and its related mobility defined as

μαβeff≡σαβnc=∑σncσncμαβσ,(3)\mu_{\alpha\beta}^{\mathrm{eff}} \equiv\frac{\sigma_{\alpha\beta}}{n_c}=\sum_{\sigma}\frac{n_c^{\sigma}}{n_c}\mu_{\alpha\beta}^{\sigma}, \tag*{(3)}

where nc=nc↑+nc↓n_c=n_c^{\uparrow}+n_c^{\downarrow}. In an antiferromagnet such as Cr2_2O3_3 where the bands are spin degenerate, the carrier density on each spin channel is the same nc↑=nc↓n_c^{\uparrow}=n_c^{\downarrow}. In practice, this means that the effective mobility is μαβeff=(μαβ↑+μαβ↓)/2\mu_{\alpha\beta}^{\mathrm{eff}}=(\mu_{\alpha\beta}^{\uparrow}+\mu_{\alpha\beta}^{\downarrow})/2. In addition, due to the magnetic sub-lattice site symmetry of Cr2_2O3_3, one has that μαβeff=μαβ↑=μαβ↓\mu_{\alpha\beta}^{\mathrm{eff}}=\mu_{\alpha\beta}^{\uparrow}=\mu_{\alpha\beta}^{\downarrow}, see Supplementary material S2 [60].

Results and Discussion

Intrinsic mobility

The crystal structure and the electron and hole mobility of Cr2_2O3_3 as a function of temperature are presented in Figure 1. Our results reveal a fundamental transport characteristic: the electron mobility is greater than the hole mobility at all temperatures, indicating that Cr2_2O3_3 would be an intrinsic nn-type semiconductor, a finding that provides a new perspective on the materials carrier transport properties.

To assess the reliability of the calculations, we compare the calculated hole mobilities with the experimental values of the hole drift mobility at room temperature reported in Refs. [77, 78] as solid stars in Figure 1(b). Comparing the calculated hole mobility (1-4 cm2^2/(V·s)) with the experimental values of the hole drift mobility ranging from 1.5 cm2^2/(V·s) to 11 cm2^2/(V·s), we find that the calculations agree rather well with experiments. The observed experimental spread is largely attributed to the device architecture employed, organic field effect transistors (OFETs) and thin film transistors (TFT), where the gate and the electrode contact resistance can induce a much higher nominal mobility μnom\mu^{\mathrm{nom}} than the intrinsic one μint\mu^{\mathrm{int}} [79, 80]. For that reason, we also provide the saturation mobility μsat\mu^{\mathrm{sat}} of Ref. [77] (red star in Figure 1(b)), which following Ref. [79] should be smaller than the one obtained in the linear region (green and yellow stars in Figure 1(b)). This agreement is further supported by comparing the temperature dependence of the hole drift mobility with the experimental effective Hall mobility that can be extracted from Ref. [81], see Supplementary Material S3 [60], confirming the robustness of the BTE approach.

This intrinsic nn-type behavior stands in contrast with the common trend of pp-type doping Cr2_2O3_3 of many studies, where it is assumed to be a pp-type semiconductor [23–30]. The origin of this discrepancy comes the fact that Cr2_2O3_3 is a wide band gap semiconductor where experimental reports show an optical band gap of 3.1 − 3.3 eV [16, 66, 82]. As a result, the intrinsic number of carriers will be relatively small, making it sensitive to defects, impurities, and oxidation [31]. Recent calculations by Wang and Mu [83] show that under oxygen-rich growth conditions, chromium vacancies (VCr_{\mathrm{Cr}}) exhibit exceptionally low formation energies, acting as acceptors and yielding the observed pp-type character. Completing this picture, Medasani et al. [84, 85] made a similar analysis further showing that Cr-rich environments oxygen vacancies (VO_{\mathrm{O}}) are dominant acting as donors (nn-type doping). This aligns perfectly with our transport predictions. While the pristine lattice fundamentally favors electron mobility (μe/μh=2−3\mu_{\mathrm{e}}/\mu_{\mathrm{h}} = 2 - 3) due to its electronic structure, optimizing chemical synthesis towards oxygen reduced and Cr-rich environments favor donor defects to unlock this intrinsic nn-type channel. This unified thermodynamic and carrier transport picture directly explains the experimental success in producing nn-type Cr2_2O3_3 films via metallic Cr interfaces [30] or Ti-doping under reduced oxygen environments [31]. Thus, the systematically found pp-type character of Cr2_2O3_3 is extrinsic. We verify that these trends are confirmed when using DFT+U, see Supplementary Material S4 [60].

Beyond its intrinsic mobility, Cr2_2O3_3 exhibits a conducting easy axis anisotropy along the zz-axis yielding μxx=μyy≠μzz\mu_{xx} = \mu_{yy} \ne\mu_{zz} in line with experimental reports of the conductivity [23]. Comparing the in plane μ∥=μxx=μyy\mu_{\parallel} = \mu_{xx} = \mu_{yy} and out of plane μ⊥=μzz\mu_{\perp} = \mu_{zz} mobilities of both electrons and holes, we find that μ⊥∼1.06μ∥\mu_{\perp} \sim1.06\mu_{\parallel}, see Figure 1(b). This anisotropy is particularly interesting because it would be aligned with its magnetic and magnetoelectric (ME) easy axis. Moreover, we note that although the easy axis anisotropy is qualitatively well captured by DFT calculations, using a DFT+U approach does not lead to better results, see Supplementary Material S4 [60].

Advantages of Cr2_2O3_3 as a TCO and beyond

In the context of TCO applications, Cr2_2O3_3 presents an electron mobility ranging between μe∼5−10\mu^{\mathrm{e}} \sim5 - 10 cm2^2/(V s) which is rather mediocre compared with other commonly used nn-type TCOs such as ITO (In2_2O3_3 and SnO2_2) or ZnO, see Table I. In contrast, the hole mobility μh∼1−4\mu^{\mathrm{h}} \sim1 - 4 cm2^2/(V s) is suitable for TCO applications since commercially available pp-type TCOs such as SnO, or CuAlO2_2 show similar mobilities [86–88], see Table I. Crucially, Cr2_2O3_3 exhibits a near parity electron and hole mobility ratio μe/μh=2−3\mu^{\mathrm{e}}/\mu^{\mathrm{h}} = 2 - 3, a feature that is strikingly absent in conventional high-performance TCOs, see Table I. While materials like In2_2O3_3 or ZnO offer higher absolute electron mobilities, their extreme carrier asymmetry (often exceeding 100:1) represents a fundamental bottleneck for complementary logic. Compared to SnO – a rare example of pp-type oxide with a relatively balanced mobility ratio μh/μe=5−15\mu^{\mathrm{h}}/\mu^{\mathrm{e}} = 5 - 15– Cr2_2O3_3 offers wider optical window with a bandgap of 3.1-3.3 eV which ensures high transparency in the visible range, whereas SnO (2.7-3.0 eV) often suffers from a yellowish tint or absorption at the blue end of the spectrum. This makes Cr2_2O3_3 the more viable candidate for high-fidelity transparent displays and UV-photo-detectors. Furthermore, Cr2_2O3_3 offers a distinct chemical processing advantage over low-dimensional or narrow gap alternatives. While pp-type favored options like SnO are prone to phase structure instability and unintended oxidation into SnO2_2 during high temperature processing, Cr2_2O3_3 is a thermodynamically stable oxide. This structural robustness and stability ensures that its balanced transport properties can be preserved during standard high-temperature oxide processing. While the absolute mobilities are not particularly high, Cr2_2O3_3 could be a good candidate for specialized targeted applications like monolithic, fully transparent optoelectronic circuits or ultraviolet (UV) photo-detectors requiring symmetric carrier extraction. Thus, using materials like Cr2_2O3_3 with sub-10 mobility ratios would challenge the current fast-electron/slow-hole paradigm of the TCO industry.

In addition, the conducting and magnetic easy axis alignment in Cr2_2O3_3 provides a direct link between charge transport and magnetic order, suggesting that the carrier transport might be sensitive to the AFM state. Combining this alignment of AFM Cr2_2O3_3 with its large optical bandgap, reasonable p-type mobility and uniquely balanced electron and hole mobilities, places Cr2_2O3_3 as a multifunctional material allowing to build transparent, bipolar ME-driven AFM switching devices.

Origin of the nn-type nature of Cr2_2O3_3

In order to rationalize the difference between electron and hole mobilities, we compute the phonon dispersion and the frequency resolved scattering rate ∂τ−1/∂ω\partial\tau^{-1}/\partial\omega for both electrons and holes. As seen in Figure 2, most of the scattering for both electron and hole conduction come from optical phonons. We quantify the relative importance of each modes by computing the integrated scattering rate τ−1(ω)\tau^{-1}(\omega). We find that for both electrons and holes, 90% of the total scattering rate is given by optical phonons. Focusing on holes, we find that 50% of the total scattering comes from the highest energy phonons at energies 75–82 meV. The remaining 40% is mainly given by slightly lower energy optical phonons of 67–70 meV

Image not extracted

Figure 2. (a) Phonon dispersion of Cr2_2O3_3, and spectral decomposition of the contribution of each phonon with energy ωqν\omega_{q\nu} to the total scattering rate at the band edges τ−1\tau^{-1} for (b) hole mobility and (c) electron mobility. The total scattering rates are τCB−1=851\tau^{-1}_{\mathrm{CB}} = 851 meV and τVB−1=1020\tau^{-1}_{\mathrm{VB}} = 1020 meV for electrons and holes, respectively.

(25% of the total scattering rate) and 47–51 meV (10% of the total scattering rate), see Figure 2(a) and 2(b).

A similar analysis can be made in the case of the electrons where mainly optical phonons at similar energy scales produce the higher scattering rates with slightly less pronounced contributions from the phonons at energies 47–51 meV than in the hole case, see Figure 2(a) and 2(c). In addition, the scattering rate distributions shown in Figure 2(b) and 2(c) are not only qualitatively but quantitatively similar. This implies that phonon scattering affects in a similar way both electrons and holes with only a 15% difference. As a consequence, the specific phonon modes that produce scattering cannot explain the disparity between the electron and hole mobility.

Therefore, we turn our attention to the electronic properties of Cr2O3\mathrm{Cr_2O_3}. We compute the band structure and show in Figure 3 the conduction and valence band edges relevant to the mobility in the given temperature range. We observe that the conduction band has 3 main valleys along the Γ−T\Gamma- \mathrm{T}, Γ−S0\Gamma- \mathrm{S_0} and Γ−F\Gamma- \mathrm{F}, with the latter hosting the conduction band minimum (CBM), see point 1 in Figure 3. In contrast, the valence band has 4 valleys along the Γ−T\Gamma- \mathrm{T}, Γ−L\Gamma- \mathrm{L}, Γ−S0\Gamma- \mathrm{S_0} and Γ−F\Gamma- \mathrm{F} paths, with the latter 2 being quasi-degenerate and hosting the VBM, see points 2 and 3 in Figure 3. This multi-valley structure of the valence band edge implies that for the holes the intra-valley scattering would be more pronounced than for the electrons.

Image not extracted

Figure 3. Electronic dispersion of antiferromagnetic Cr2_2O3_3. Conduction and valence band manifolds in the interesting region for mobility.

Additionally, we note that the bands in the conduction band manifold are more dispersive than in the valence band. This indicates that in the conduction band we have higher band velocities and lower effective masses than in the valence. We compute the effective masses of the band edge of the valleys 1 and 2 in Figure 3 that host the CBM and the VBM, respectively. We find that in the case of the electrons the effective mass is lighter than in the hole case. For the CBM m∗∼4mem^* \sim4m_e while for the VBM m∗∼6mem^* \sim6m_e. Thus, the main origin of the disparity is electronic with a minor role for the phonons. This can be further verified from the DFT+U calculations since the main difference are changes in band curvature and relative valley energy, see Supplementary Material S4 [60] for further discussion.

Conclusions

In summary, our ab initioab\ initio BTE calculations reveal that the phonon limited mobility of pristine Cr2O3\mathrm{Cr_2O_3} inherently favors nn-type over pp-type behavior, featuring a well balanced carrier mobility ratio (μe/μh=2−3\mu_e/\mu_h = 2 - 3). By contextualizing the intrinsic electron and hole mobility, its wide band gap, and the established thermodynamic and defect models, we solve a long-standing debate in the literature: the pp-type character of Cr2O3\mathrm{Cr_2O_3} is an extrinsic feature dictated by oxygen rich defect thermodynamics, whereas the pristine crystal is an intrinsic nn-type semiconductor. Hence, the current trend of pp-type doping Cr2O3\mathrm{Cr_2O_3} is only partially justified, since the hole mobility is around μh∼1−4 cm2/(Vs)\mu_h \sim1 - 4\ \mathrm{cm^2/(Vs)}, Cr2O3\mathrm{Cr_2O_3} and it could still be appealing in the context of pp-type TCOs since they show lower mobility. Thus, the combination of a wide band gap, chemical and thermodynamic phase stability and the near-parity electron-hole mobilities positions Cr2O3\mathrm{Cr_2O_3} as a unique platform for transparent bipolar electronics and functional oxide heterostructures, shifting the design focus on oxygen reduction of Cr-rich processing regimes. In addition, we uncover the importance of the electronic degrees of freedom over the scattering with the phonons in Cr2O3\mathrm{Cr_2O_3} when determining charge carrier mobility. As a consequence, small changes in Cr2O3\mathrm{Cr_2O_3} electronic structure are expected to impact the mobility which hints at an exploitable response to external stimuli such as strain. Importantly, the alignment of the conducting easy axis of Cr2O3\mathrm{Cr_2O_3} with its known ME easy axis for both electron and hole mobility, suggests that Cr2O3\mathrm{Cr_2O_3} could be an interesting platform for transparent ME-FET and antiferromagnetic spintronic optoelectronic devices. Hence reshaping the role of Cr2O3\mathrm{Cr_2O_3} as a TCO and presenting itself as well as an interesting platform for bipolar transparent AFM spintronics, and overall as a multifunctional material.

Acknowledgments

Acknowledgments–S. P. is a Research Associate of the Fonds de la Recherche Scientifique - FNRS. This work was supported by the Fonds de la Recherche Scientifique - FNRS under Grants number T.0183.23 (PDR) and T.W011.23 (PDR-WAVE). Computational resources have been provided by the supercomputing facilities of the Université catholique de Louvain (CISM/UCL) and the Consortium des Équipements de Calcul Intensif en Fédération Wallonie Bruxelles (CÉCI) funded by the Fond de la Recherche Scientifique de Belgique (F.R.S.-FNRS) under convention 2.5020.11 and by the Walloon Region. The present research benefited from additional computational resources made available on Lucia, the Tier-1 supercomputer of the Walloon Region, infrastructure funded by the Walloon Region under the grant agreement n°1910247. In addition, we acknowledge EuroCC Belgium for awarding this project access to the LUMI supercomputer, owned by the EuroHPC Joint Undertaking, hosted by CSC (Finland) and the LUMI consortium through EuroCC Belgium (UCLouvain-MODL-EPIMS-Bench-T0447)

Supplementary Material: Challenging the pp-type Paradigm: Intrinsic nn-type Mobility in Antiferromagnetic Cr2_2O3_3

Álvaro Adrián Carrasco Álvarez; Samuel Poncé

Polaron radius

In this section, we estimate if Cr2_2O3_3 is in the large or small polaron regime, following Ref. [1]. To that end, we compute the static dielectric constant and the Born effective charges as well as the effective mass of both electrons and holes and use it to make an estimate of the coupling strength of the Feynman polaron model α\alpha defined as:

α≡e24πε0ℏm∗2ℏωeff(1ε(∞)−1ε(0))(4)\alpha\equiv\frac{e^2}{4\pi\varepsilon_0\hbar}\sqrt{\frac{m^*}{2\hbar\omega_{\mathrm{eff}}}}\left(\frac{1}{\varepsilon(\infty)}-\frac{1}{\varepsilon(0)}\right) \tag*{(4)}

where m∗m^* is the effective mass, ωeff\omega_{\mathrm{eff}} is the characteristic optical phonon frequency, ε0\varepsilon_0 is the vacuum permittivity, ε(∞)\varepsilon(\infty) is the high frequency dielectric constant (this is the dielectric constant ε(ω)\varepsilon(\omega) when ω→∞\omega\to\infty) and ε(0)\varepsilon(0) is the static dielectric constant, (this is the dielectric constant ε(ω)\varepsilon(\omega) when ω=0\omega= 0) which we estimate as

ε(0)=ε(∞)+1Vucε0∑νIνων2(5)\varepsilon(0) = \varepsilon(\infty) + \frac{1}{V_{\mathrm{uc}}\varepsilon_0}\sum_\nu\frac{I_\nu}{\omega_\nu^2} \tag*{(5)}

with VucV_{\mathrm{uc}} being the volume of the unit cell, IνI_\nu the infra red activities of the phonon mode with frequency ων\omega_\nu. Regarding the effective phonon frequency ωeff\omega_{\mathrm{eff}}, we take an average phonon frequency based on their infrared activity

ωeff=∑νωνIν∑νIν(6)\omega_{\mathrm{eff}} = \frac{\sum_\nu\omega_\nu I_\nu}{\sum_\nu I_\nu} \tag*{(6)}

obtaining a value of ωeff=63\omega_{\mathrm{eff}} = 63 meV. Although the dielectric constant is a tensor and in Cr2_2O3_3 has 2 independent directions, we take an isotropic average for both static ε\varepsilon and ε∞\varepsilon_\infty to give an estimate of the coupling constant. For the effective mass, we consider the effective masses at the conduction band minimum (CBM) and valence band maximum (VBM). We obtain m∗∼4mem^* \sim4m_e for electrons and m∗∼6mem^* \sim6m_e for holes which yields a coupling constant of α=1.15\alpha= 1.15 for electrons and α=1.41\alpha= 1.41 for holes, which falls in the weak to intermediate coupling regime. In addition, we can estimate the polaron radius as rp=3.4ℏ/m∗αωeffr_p = \sqrt{3.4\hbar/m^*\alpha\omega_{\mathrm{eff}}} obtaining values of 9.5 Å and 7 Å for electron and holes respectively, corresponding to 1.8 and 1.3 lattice parameters encompassing 23 and 9 unit cells respectively. This falls in the large polaron regime following Ref. [2]. Thus, we conclude that in Cr2_2O3_3, the transport properties can be well described within the Boltzmann transport equation since the polaron radius is in the large-intermediate regime, where band transport dominates over polaron hopping.

Spin resolved mobility

In this section, we calculate the relative contributions of each spin channel in Cr2_2O3_3 showing that due to the lattice site symmetry the spin resolved mobilities are the same μαβ↑=μαβ↓\mu_{\alpha\beta}^{\uparrow} = \mu_{\alpha\beta}^{\downarrow} as well as the carrier density nc↑=nc↓n_c^{\uparrow} = n_c^{\downarrow}. We first compute the band structure for the two spin channels in Figure S1 which is a hallmark of the lattice site symmetry. Additionally, we compute the spin resolved mobility and the relative carrier density as a function of temperature in Figure S1.

Image not extracted

Figure S1. Electronic band dispersion for bulk Cr2O3\mathrm{Cr}_2\mathrm{O}_3 (a), and carrier density ncn_c for different temperatures (b). Red solid lines correspond to the spin up while dashed blue lines correspond to the spin down. The total carrier density is set to 1010 cm−310^{10}\ \mathrm{cm}^{-3} at all temperatures.

We see that the eigenvalues are spin degenerate εnk↑=εnk↓\varepsilon_{nk}^{\uparrow} = \varepsilon_{nk}^{\downarrow}. This yields an equivalent relative carrier density between the two spin channels nc↑=nc↓n_c^{\uparrow} = n_c^{\downarrow}. This is not the case in general even if the system is antiferromagnetic. The combination of these two factors yields mobilities that are equivalent as well μαβ↑=μαβ↓\mu_{\alpha\beta}^{\uparrow} = \mu_{\alpha\beta}^{\downarrow}, see Figure S2 and Figure S2 for the in plane and out of plane electron and hole mobilities, respectively.

Image not extracted

Figure S2. Electron and hole mobility for bulk Cr2O3\mathrm{Cr}_2\mathrm{O}_3. (a) In plane μ∥=μxx=μyy\mu_{\parallel} = \mu_{xx} = \mu_{yy} and (b) out of plane μ⊥=μzz\mu_{\perp} = \mu_{zz}. Solid lines correspond to the spin up (red and orange for electron and hole, respectively) while dashed lines correspond to the spin down (dark and light blue for electron and hole mobility, respectively). The total carrier density is set to 1010 cm−310^{10}\ \mathrm{cm}^{-3}at all temperatures.

Temperature dependence of the Hall mobility

We compare the hole Hall mobility from the experimental data of Ref. [3], with the drift mobility obtained with the Boltzmann Transport Equation, see Figure S3.

Image not extracted

Figure S3. Hole mobility of antiferromagnetic Cr₂O₃. Blue squares correspond to the Hall mobility computed as μH=σRH\mu_{\mathrm{H}} = \sigma R_{\mathrm{H}} and scaled by a factor 14\frac{1}{4}. Green solid line correspond to the calculated hole drift mobility.

The experimental Hall mobility μH=σRH\mu_{\mathrm{H}} = \sigma R_{\mathrm{H}} is obtained as the product of the Hall resistance RHR_{\mathrm{H}} and the conductivity σ\sigma. Additionally, it is common for the Hall mobility to be higher than the drift mobility since the Hall factor μH/μ=rH\mu_{\mathrm{H}}/\mu= r_{\mathrm{H}} is often greater than one. Although strictly speaking, the Hall factor is temperature dependent, its variation is not extremely pronounced [4] and taking it to be constant is reasonable approximation. In that regard, we scale the experimental Hall values computed as μH=σRH\mu_{\mathrm{H}} = \sigma R_{\mathrm{H}} [3] by a constant factor of 14\frac{1}{4}, see blue squares in Figure S3, to better compare the temperature dependence. In doing so, we find that the temperature dependence is well described by our calculations.

Mobility calculations with DFT+U

In this section we show the mobility calculations using a DFT+U [5] for both electron and hole mobility considering the three cases U=0,3,5U = 0, 3, 5 eV. The results are given in Figure S4 for relaxed lattice parameters for each UU value while keeping the same k/q meshes as with DFT. The general trends with respect to the nn-type behavior of Cr₂O₃ are not

Image not extracted

Figure S4. Intrinsic electron (red, magenta and purple lines) and hole (green, khaki and orange lines) mobility for bulk Cr₂O₃ as a function of temperature obtained with the Boltzmann transport equation (BTE). Dashed lines correspond to out of plane mobility while solid lines are in plane mobility values. A DFT+U approach is taken with U=0,3,5U = 0, 3, 5 eV corresponding to the curves with circles, triangles and squares respectively.

affected by the specific UU, giving the same qualitative picture. Nonetheless, we notice that at the quantitative level, the inclusion of the onsite UU repulsion produces different behaviors for the electron and hole mobilities. While for the hole mobility on site UU can both enhance or reduce the hole mobility, for the electron mobility there is a consistent enhancement regardless of the UU value. This is mainly due to the change in band topology of valence and conduction band manifolds, see Figure S5.

Image not extracted

Figure S5. Valence band (a) and conduction band (b) manifolds of Cr2_2O3_3. A DFT+U approach is used with U=0,3,5U=0,3,5 eV corresponding to blue, magenta and red bands.

By comparing the different bands in the Figure S5, it becomes clear that with DFT+U the inter-valley scattering is strongly affected by the specific UU values. The inclusion of non zero UU produces the disappearance of the Γ−T\Gamma-\mathrm{T} and Γ−F\Gamma-\mathrm{F} valleys when increasing the UU, and at the same time the appearance of two valleys along the H2−F\mathrm{H}_2-\mathrm{F} direction and more importantly Γ−L\Gamma-\mathrm{L} directions which for larger UU reaches almost the Fermi level. For the electron mobility, the inter-valley scattering is reduced for non-zero UU values because the valleys along the Γ−T\Gamma-\mathrm{T} and Γ−S0\Gamma-\mathrm{S}_0 directions are pushed upwards in energy, see Figure S5. In addition, we can find that in Figure S4 the conducting easy axis anisotropy disappears with a finite UU for the electron mobility while for the hole mobility the specific UU may change from easy to hard axis with larger UU values leading to a more prominent easy axis similar to experiments [6]. For the case of the electrons, we can understand the change from easy axis to hard axis anisotropy by increasing UU as a consequence of the tendency of DFT+U approaches to produce a much smaller band width, flatter bands and larger effective masses [7]. We can verify this in Figs. S5(a) and S5(b) where it is clear that along the out of plane Γ−T\Gamma-\mathrm{T} and Γ−F\Gamma-\mathrm{F} directions the bands are flatter and less dispersive with non zero UU when compared with the in plane Γ−L\Gamma-\mathrm{L} and Γ−S0\Gamma-\mathrm{S}_0 directions. A possible approach to fix the issue of the UU values is to compute the self consistent UU following Refs. [8]. We proceed with that approach and obtain a self consistent value of Usc=2.83±0.02U_{\mathrm{sc}}=2.83\pm0.02 eV which is quite similar to the U=3U=3 eV used here. Thus, although DFT+U approaches are sufficient when providing an effective tool to open band gaps, the band curvature is not always accurately predicted and in some cases can produce the wrong anisotropic behavior [7].

References

References

  1. [1]M. Schliff, S. Poncé, and F. Giustino, Carrier lifetimes and polaronic mass enhancement in the hybrid halide perovskite CH₃NH₃PbI₃ from multiphonon fröhlich coupling, Phys. Rev. Lett. 121, 086402 (2018).
  2. [2]P. M. M. de Melo, J. C. de Abreu, B. Guster, M. Giantomassi, Z. Zanolli, X. Gonze, and M. J. Verstraete, High-throughput analysis of fröhlich-type polaron models, npj Computational Materials 9, 147 (2023).arxiv.org/abs/2207.00364
  3. [3]M. Julkarnain, J. Hossain, K. Sharif, K. Khan, et al., Temperature effect on the electrical properties of chromium oxide (Cr₂O₃) thin films, Journal of Optoelectronics and advanced materials 13, 485 (2011).
  4. [4]S. Poncé, F. Macheda, E. R. Margine, N. Marzari, N. Bonini, and F. Giustino, First-principles predictions of hall and drift mobilities in semiconductors, Physical Review Research 3, 043022 (2021).DOI
  5. [5]W. Yang, S. Tiwari, F. Giustino, and Y.-W. Son, First-principles electron-phonon interactions with self-consistent hubbard interaction: Application to transparent conducting oxides, Phys. Rev. B 112, 075203 (2025).DOI
  6. [6]J. A. Crawford and R. W. Vest, Electrical conductivity of single-crystal Cr₂O₃, Journal of Applied Physics 35, 2413 (1964).DOI
  7. [7]S. Falletta, and A. Pasquarello, Hubbard u through polaronic defect states, npj Computational Materials 8, 263 (2022).arxiv.org/abs/2209.11341
  8. [8]I. Timrov, N. Marzari, and M. Cococcioni, Self-consistent hubbard parameters from density-functional perturbation theory in the ultrasoft and projector-augmented wave formulations, Phys. Rev. B 103, 045141 (2021).DOI

Paper details

vlnc:W_sqzq8dyc
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 1

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents