betaFeedback

WorksPhysics and Astronomy

First-principles band alignment engineering in polar and nonpolar orientations for wurtzite AlN, GaN, and Bx_xAl1−x_{1-x}N alloys

Abstract

Boron aluminum nitride (Bx_xAl1−x_{1-x}N) is a promising material for next-generation electronic and optoelectronic devices due to its ultra-wide bandgap, high thermal stability, and compatibility with other III-nitride semiconductors. Despite its potential, the band alignments of Bx_xAl1−x_{1-x}N remain largely unexplored, although this information is essential for device design. In this study, we compute the valence and conduction band alignments of nonpolar (aa-plane) and polar (cc-plane) Bx_xAl1−x_{1-x}N, and compare them with those of AlN and GaN. Using density functional theory, many-body perturbation theory, GW0GW_0 method, and a novel passivation scheme, we find that they have near-zero valence band alignments for low-xx Bx_xAl1−x_{1-x}N/AlN, while higher compositions (x>x > 0.333) exhibit type I or II band alignments. The band alignments also show a notable dependence on surface polarity and the tetrahedral distortion of the Bx_xAl1−x_{1-x}N structures. Our computed offsets are in good agreement with available experimental data. Due to their low valence band alignments and higher conduction band alignments, the Bx_xAl1−x_{1-x}N/AlN heterostructures could be well suited for high-electron-mobility transistors and ultraviolet light-emitting diodes. The band alignments of Bx_xAl1−x_{1-x}N determined in this study provide essential design guidelines for integrating these ultra-wide bandgap alloys into advanced semiconductor technologies.

Boron aluminum nitride (BxAl1-xN) is a ternary group-III nitride material that has seen increasing interest in recent years due to its array of intriguing properties. In the wurtzite phase (w-), its tunable ultra-wide bandgap (6.2-7.4 eV) extends the bandgap range available to III-nitride electronic materials, making it particularly attractive for advanced electronic and optoelectronic devices in the deep-ultraviolet range[1, 2]. Its excellent thermal conductivity and high chemical stability mean it could be used in next-generation high-power and high-frequency devices[1, 2, 3]. In addition to these excellent properties, it has recently been demonstrated that as little as 2 % of boron incorporation into AlN yields ferroelectric switching capabilities[2, 3, 4, 5, 6], and it has also been shown to exhibit a dielectric constant much higher than AlN or BN, e.g. in B0.07Al0.93N a value of ∼16ε0\sim16\varepsilon_0 is predicted, compared to ∼9.7ε0\sim9.7\varepsilon_0 in AlN and 7.3ε07.3\varepsilon_0 in w-BN[2, 4, 7]. These properties could enable the use of BxAl1-xN for novel dielectric or ferroelectric layers[2, 4, 7], and thus BxAl1-xN has a huge potential in III-nitride compatible heterojunction-based devices.

Despite growing interest in BxAl1-xN for device applications, systematic studies of its electronic properties relevant to heterojunctions are still scarce.

Experimental investigation of such heterojunctions remains challenging due to the difficulty in obtaining high-purity, phase-pure, and high-crystallinity samples of BxAl1-xN[4, 5, 6, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18]. Additionally, theoretical studies of certain surfaces in wurtzite group-III-nitrides, such as the (0001) c-plane, are complicated due to the electric field generated by the polarization-induced charge displacement[19], which further limits the understanding of the fundamental factors that govern the properties of the heterojunctions.

Among various heterojunction properties, the band alignments at heterojunction interfaces critically determine carrier confinement, transport, and recombination behavior; thus, understanding these alignments in BxAl1-xN-based systems is essential for device design. However, only a few investigations have examined the band alignments of BxAl1-xN to date.

Experimentally, Sun et al. measured the band alignments in B0.14Al0.86N thin films with respect to AlN, GaN and Al0.7Ga0.3N[11]. Rather et al. measured the band alignments for B0.13Al0.87N/GaN and B0.13Al0.87N/AlN heterojunctions[20]. Only Ota et al. has published a theoretical study on band alignments in BxAl1-xN using the empirical solid-state energy (SSE) approach, which does not take into account effects of surfaces[21] which have significant impact for polar materials and interfaces[22, 23].

In this work, we make the first surface-dependent predictions of BxAl1-xN band alignment in the complete x=0x = 0 to 11 range using first-principles simulations.

For determination of the band alignment alignment, we follow the two-step process where in the first step, the valence-band maximum (VBM) and conduction-band minimum (CBM) energies are computed for each bulk structure.[24] Alongside 17 ground-state structures of BxAl1-xN[1, 2], we also investigate the band alignment of GaN, AlN, and w-BN. In a previous work, we determined the 17 ground-state structural configurations of BxAl1-xN using DFT and the cluster expansion formalism[1]. In this work, we also introduce x=0.083x = 0.083 and x=0.125x = 0.125 BxAl1-xN models to investigate the low-xx regime, with their structures taken to be the lowest formation energy structures from three random alloys. The VBM and CBM energies in the bulk structure are referenced to the macroscopic average potential in the bulk structure, V‾macroscopicbulk\overline{V}_{\mathrm{macroscopic}}^{\mathrm{bulk}}. In the second step, we simulate a finite-slab of the material to align V‾macroscopicslab\overline{V}_{\mathrm{macroscopic}}^{\mathrm{slab}} with the vacuum energy, EvacE_{\mathrm{vac}}, establishing a universal reference energy that allows comparison between different materials. Thus this two-step process leads to computation of a surface-dependent VBM energy, EVBMslabE_{\mathrm{VBM}}^{\mathrm{slab}}, as

EVBMslab=(EVBMbulk−V‾macroscopicbulk)+V‾macroscopicslab−Evac.(1)E_{\mathrm{VBM}}^{\mathrm{slab}} = \left(E_{\mathrm{VBM}}^{\mathrm{bulk}} - \overline{V}_{\mathrm{macroscopic}}^{\mathrm{bulk}}\right) + \overline{V}_{\mathrm{macroscopic}}^{\mathrm{slab}} - E_{\mathrm{vac}}. \tag*{(1)}

Similarly, the CBM energy, ECBMbulkE_{\mathrm{CBM}}^{\mathrm{bulk}}, can also be obtained. Figure 1 shows a schematic of some slab models of the cc-plane and aa-plane for AlN, B0.500_{0.500}Al0.500_{0.500}N, and w-BN, respectively. In total, 104 distinct surfaces were studied in this work.

Image not extracted

Figure 1. Slab structures in the cc-plane (left) and aa-plane (right) orientations for AlN, B0.500_{0.500}Al0.500_{0.500}N, and w-BN. The slab structures are fully passivated by pseudohydrogen (psH) atoms in the cc-plane. The aa-plane orientations are unpassivated. The dashed lines indicate the slab lattice with surface normals in the horizontal direction. Red, blue, green, and black spheres represent boron, aluminum, nitrogen, and psH atoms, respectively.

This two-step approach has long been used for computing band alignment of non-polar surfaces, even in polar materials.[24, 25, 26, 27] However, the spontaneous polarization, PsP_{\mathrm{s}}, induced electric fields interfere with the potential alignment, making this approach unsuitable for polar surfaces such as wurtzite cc-plane III-nitrides and Bx_xAl1−x_{1-x}N. We employ a recently developed passivation scheme by Yoo et al.¹⁹ to consider the electric fields due to spontaneous polarization in finite slabs, addressing the long-standing challenge of computing band alignments in materials with spontaneous polarization. The passivation involves adsorbing pseudohydrogen atoms (psH) with PsP_{\mathrm{s}}-dependent modified valency to one surface; unlike "conventional passivation scheme", which only uses psH with valency of the missing bond(s).[28] Further information on the passivation scheme is available in the Supplementary Information Figure S1 and S2. In the Supplementary Information, the DFT-computed PsP_{\mathrm{s}} with respect to the zincblende structure, PszbP_{\mathrm{s}}^{\mathrm{zb}}, and the modified psH for all the polar slabs are presented in detail. In addition, information on DFT computed PshexP_{\mathrm{s}}^{\mathrm{hex}} and piezoelectric constants is available in the Supplementary Information. We emphasize here that this work presents the first DFT-computed PszbP_{\mathrm{s}}^{\mathrm{zb}} in Bx_xAl1−x_{1-x}N in Figures S3 in the Supplementary Information.

All ab-initio simulations reported in this study were performed using the Vienna Ab Initio Simulation Package (VASP)[29, 30, 31, 32] package using the Projector-Augmented-Wave (PAW) formalism[33] with the PBE[33, 34, 35] exchange-correlation functional for all calculations. Slab calculations were performed without ionic relaxation, therefore predicted band alignments do not include the effects of strain due to heterojunction lattice mismatch, interfacial relaxation, or dependence on experimental growth conditions.

(PAW) formalism[33] with the PBE[33, 34, 35] exchange-correlation functional for all calculations. Slab calculations were performed without ionic relaxation, therefore predicted band alignments do not include the effects of strain due to heterojunction lattice mismatch, interfacial relaxation, or dependence on experimental growth conditions. A dipole correction[36] was implemented along the zz-axis to correct the spurious electric fields in the vacuum regions. For the slab calculations, the total energy was converged within 10−710^{-7} eV using Γ\Gamma-centered kk-grids of 1000 kk-points per Å3^3. An energy cutoff of 520 eV was used for all calculations. Thicknesses of 8 bilayers were used for all cc-plane (15.3–19.0 Å) slabs, which we have determined to be below the threshold for electric breakdown by analyzing the thickness-dependent electric fields. For nonpolar aa-plane slabs, a thicknesses of at least 13 layers (>>20.0 Å) was employed. These parameters allow convergence of V‾slabmacroscopic−E‾vac\overline{V}_{\mathrm{slab}}^{\mathrm{macroscopic}}-\overline{E}_{\mathrm{vac}} to within 0.01 eV.

Many unique slab configurations can often be constructed from the bulk Bx_xAl1−x_{1-x}N structures due to the varying fraction of atoms in the lattice. Therefore, we calculate band alignments for every slab for each xx generated using pymatgen[37], which generates slab models by rotating the bulk structure into the desired orientation and cleaving it to generate all unique surface terminations for a given thickness. We obtain V‾slabmacroscopic−E‾vac\overline{V}_{\mathrm{slab}}^{\mathrm{macroscopic}}-\overline{E}_{\mathrm{vac}} for each slab model along with EVBMbulk−V‾bulkmacroscopicE_{\mathrm{VBM}}^{\mathrm{bulk}}-\overline{V}_{\mathrm{bulk}}^{\mathrm{macroscopic}} from corresponding bulk calculations to calculate EVBMslabE_{\mathrm{VBM}}^{\mathrm{slab}} and ECBMslabE_{\mathrm{CBM}}^{\mathrm{slab}} for all slabs of the same xx and surface plane, with error bars indicating standard deviations. EVBMbulkE_{\mathrm{VBM}}^{\mathrm{bulk}}, ECBMbulkE_{\mathrm{CBM}}^{\mathrm{bulk}}, and other electronic properties of bulk AlN, w-BN, and ground-state bulk Bx_xAl1−x_{1-x}N configurations were computed in our previous work[2]. The mid-gap approximation was employed to obtain the GW0GW_0 EVBMbulkE_{\mathrm{VBM}}^{\mathrm{bulk}} and ECBMbulkE_{\mathrm{CBM}}^{\mathrm{bulk}} from the PBE values[38]. Given the severe band-gap underestimation in PBE[39], we used GW0GW_0 quasiparticle band gaps[2, 39, 40, 41], which give excellent agreement with experimental values and are available in the Supplementary Information Figure S4 and Table S1.

The results of our band alignment study are shown comprehensively in Figure 2, where a) and b) show the band alignments for the cc-plane N- and M-surfaces (where M=B, Al, or Ga), and c) shows the band alignments for the aa-plane surfaces. By taking the difference between the EvacE_{\mathrm{vac}}-aligned EVBMslabE_{\mathrm{VBM}}^{\mathrm{slab}} or ECBMslabE_{\mathrm{CBM}}^{\mathrm{slab}} (hereafter referred to as EVBME_{\mathrm{VBM}} and ECBME_{\mathrm{CBM}}) of different materials, we obtain the valence band offsets (VBO) or conduction band offsets (CBO) using an Anderson’s model type approach[24, 42]. All cc-plane slab models (Figures S5-S9) and their potential profiles (Figure S10-12) are available in the Supplementary Information.

Image not extracted

Figure 2. Band alignments for GaN and Bx_xAl1−x_{1-x}N, where EVBME_\mathrm{VBM} and ECBME_\mathrm{CBM} are represented by blue and red columns, respectively. The columns are shaded according to the average tetrahedrality in the bulk Bx_xAl1−x_{1-x}N structure, which are shown by the colormap. Blue and red dashed horizontal lines show the AlN EVBME_\mathrm{VBM} and ECBME_\mathrm{CBM}, respectively. The vertical dashed gray lines separate the GaN and Bx_xAl1−x_{1-x}N results. a) Shows the cc-plane N-surface slabs, b) the cc-plane M-surface slabs, and c) the aa-plane symmetric slabs. Error bars indicate standard deviations of the band alignments of the structures of the same composition.

Ideal interfaces between two cc-plane III-nitrides consist of a shared nitrogen layer. Therefore we compute band alignments for cc-plane N-surface slabs of each III-nitride to model each side of the interface. III-nitride heterointerfaces are used in a variety of device structures, e.g. AlN/Alx_xGa1−x_{1-x}N and GaN/Inx_xGa1−x_{1-x}N heterojunctions[43, 44]. Additionally, N-polar surfaces are recommended for Ohmic contacts due to lower barrier heights[45]. Therefore, a detailed investigation of the N-surface band alignments is necessary, the results of which are shown in Figure 2a for GaN and Bx_xAl1−x_{1-x}N where E−EvacE - E_{\mathrm{vac}}, i.e. EVBMslabE_{\mathrm{VBM}}^{\mathrm{slab}} or ECBMslabE_{\mathrm{CBM}}^{\mathrm{slab}}, is the energy of the band edges with respect to the vacuum energy.

To benchmark our calculations, we consider the well-studied AlN/GaN heterointerface. Prior studies agree that EVBME_{\mathrm{VBM}} of GaN is above that of AlN, indicating a type-I band alignment, tabulated in Table I. However, there is some dependence of the VBO on the lattice strain[44] and the experimental growth temperatures[46]. We obtain a VBO of 0.44 eV for AlN/GaN, which is in excellent agreement with the previously reported computed values of 0.44 eV[44], as well as with XPS measurements of 0.5±0.30.5 \pm0.3 eV[46, 47]. The agreement can be attributed to the fact that contributions from lattice mismatch and interfacial atomic relaxation are small in AlN/GaN cc-plane interfaces[43]. Similarly, these contributions are also expected to be small in low-xx Bx_xAl1−x_{1-x}N/AlN interfaces due to the low lattice mismatch[1].

The common-anion rule states that VBO are small between materials with common anions since the valence wavefunctions are typically dominated by anionic contributions in group III-nitrides[48, 49]. Similar to AlN/GaN, we observe that EVBME_{\mathrm{VBM}} in Bx_xAl1−x_{1-x}N and AlN are nearly equal, resulting in small VBO (Figure 2a). However, there are variances with respect to xx. At low xx (x<0.33x < 0.33), EVBME_{\mathrm{VBM}} is similar to that of AlN with near-0 VBO. The low lattice distortion, low lattice mismatch, and similar bandgaps in low-xx Bx_xAl1−x_{1-x}N results in similar EVBME_{\mathrm{VBM}} and ECBME_{\mathrm{CBM}} to AlN. In intermediate xx (0.333≤x≤0.6670.333 \leq x \leq0.667) Bx_xAl1−x_{1-x}N, EVBME_{\mathrm{VBM}} rises above that of AlN. Here, we find a correlation between the average tetrahedrality in the Bx_xAl1−x_{1-x}N structures and EVBME_{\mathrm{VBM}}.

The tetrahedrality is a measure of the bonding environment of an atomic site, where 0 and 1 signify a non-tetrahedral and perfect tetrahedral bonding environment, respectively. Therefore, reduced tetrahedrality means the bonds deviate from that of ideal wurtzite bonding[1, 2]. The average tetrahedrality is lowest in the intermediate xx alloys[1], resulting in elevated EVBME_{\mathrm{VBM}} and ECBME_{\mathrm{CBM}}. However, EVBME_{\mathrm{VBM}} becomes lower than that of AlN for high-xx (x>0.667x > 0.667). These lower EVBME_{\mathrm{VBM}} values are due to high boron content and the low lattice distortion (due to high tetrahedrality)[2]. Therefore, Bx_xAl1−x_{1-x}N/AlN and Bx_xAl1−x_{1-x}N/GaN interfaces could provide tunable band alignments through modulation of the boron fraction in Bx_xAl1−x_{1-x}N, particularly if higher xx (x≥0.3x \geq0.3) Bx_xAl1−x_{1-x}N is realized experimentally.

In spite of the growth challenges with Bx_xAl1−x_{1-x}N, there are a few existing experimental measurements of the band alignments in Bx_xAl1−x_{1-x}N/GaN and Bx_xAl1−x_{1-x}N/AlN heterojunctions (Table I). Measurements of B0.13_{0.13}Al0.87_{0.87}N/GaN[20] and B0.14_{0.14}Al0.86_{0.86}N/GaN[13] heterojunctions have shown major disagreement at nearly the same xx, with reported VBOs of 1.0±0.11.0 \pm0.1 eV and 0.2±0.20.2 \pm0.2 eV, respectively. The calculated VBOs for the similar xx B0.125_{0.125}Al0.875_{0.875}N/GaN and B0.167_{0.167}Al0.833_{0.833}N/GaN interfaces are 0.63±0.150.63 \pm0.15 and 0.40±0.260.40 \pm0.26 eV respectively. Notably, the average B0.125_{0.125}Al0.875_{0.875}N/GaN VBO (0.63±0.150.63 \pm0.15 eV) compares very favorably to the average VBO reported between the two experimental studies (0.60.6 eV). Thus, our calculation scheme provides a useful prediction of the experimental VBO in cc-plane Bx_xAl1−x_{1-x}N/GaN interfaces. Moreover, the discrepancy observed in calculations and experimental measurements for similar xx Bx_xAl1−x_{1-x}N/GaN heterojunctions are likely due to differing interfacial stoichiometry near the interface. The effect of interfacial stoichiometry has been studied in Alx_xGa1−x_{1-x}N and Inx_xGa1−x_{1-x}N[24].

Further study of our cc-plane slab models revealed that a higher B composition near the surface leads to lower (more negative) EVBME_{\mathrm{VBM}} and ECBME_{\mathrm{CBM}}. Intuitively, this is more similar to the high-xx Bx_xAl1−x_{1-x}N EVBME_{\mathrm{VBM}} and ECBME_{\mathrm{CBM}}.

The III-nitride M-surfaces will have different electronic properties than N-surfaces due to electronegativity differences in the surface atoms. Therefore, an understanding of how the M-surface band alignments differ is needed. For example, heterointerfaces of materials with shared group-III cations (i.e., AlN/Al2_2O3_3)[52] could be understood through analysis of the M-surface band alignments. Therefore, we report the cc-plane M-surface band alignments for GaN, AlN, w-BN, and Bx_xAl1−x_{1-x}N in Figure 2b.

In stark contrast to the GaN N-surface, the GaN M-surface EVBME_{\mathrm{VBM}} and ECBME_{\mathrm{CBM}} are dramatically higher than the GaN N-surface EVBME_{\mathrm{VBM}} and ECBME_{\mathrm{CBM}} by 1.62 eV. This large difference (>1.4> 1.4 eV[53] and 2 eV[45]) between Ga-surface and N-surface electron affinities has been experimentally observed in GaN. As a result, the predicted AlN/GaN M-surface VBO is 1.85 eV, much higher than the 0.44 eV calculated for the N-surface structures. Interestingly, many experimental measurements of the AlN/GaN VBO have found higher values[46, 50, 51, 54], for example, a VBO of 1.36 ±0.07\pm0.07 eV was measured by Waldrop et al.[55]. The higher measured VBO could be observed in AlN/GaN heterostructures as a result of interfacial Al/Ga

cc-plane interfaceVBO (eV)VBO (eV)VBO (eV)
AlN/GaNThis work0.44
Ref. Theory0.44−0.730.44 - 0.7344a0.300.3024b0.28−0.710.28 - 0.7150a
Ref. Exp.0.5±0.30.5 \pm0.3[46]0.540.54[7]0.70±0.240.70 \pm0.24[51]
Bx_xAl1−x_{1-x}N/GaNThis work0.48±0.180.48 \pm0.18 (x=0.083x = 0.083)0.63±0.150.63 \pm0.15 (x=0.125x = 0.125)0.40±0.260.40 \pm0.26 (x=0.167x = 0.167)
Ref. Exp.1.0±0.11.0 \pm0.1[20] (x=0.13x = 0.13)0.2±0.20.2 \pm0.2[13] (x=0.14x = 0.14)
Bx_xAl1−x_{1-x}N/AlNThis work0.04±0.180.04 \pm0.18 (x=0.083x = 0.083)0.19±0.150.19 \pm0.15 (x=0.125x = 0.125)−0.04±0.26-0.04 \pm0.26 (x=0.167x = 0.167)
Ref. Exp.0.5±0.10.5 \pm0.1[20] (x=0.13x = 0.13)

Table 1. The relative valence and conduction band alignments for III-nitride heterostructures calculated using the N-surface cc-plane band alignments and their comparison to existing experimental and theoretical results in the literature. a^a DFT study, b^b Hybrid functional study

Overall, the M-surface EVBME_\mathrm{VBM} trends in Bx_xAl1−x_{1-x}N are similar to that of the N-surface EVBME_\mathrm{VBM} in Bx_xAl1−x_{1-x}N, and nearly constant with xx. This suggests that the B or Al composition near the M-surface, as well as the tetrahedrality of the structure, has little impact on EVBME_\mathrm{VBM}. However, we find that the w-BN EVBME_\mathrm{VBM} is below that of AlN by 0.34 eV.

Construction of non-polar III-nitride heterojunctions could be useful through the elimination of the piezoelectric polarization effects[56]. While both the aa-plane and mm-plane surfaces are non-polar, in our calculations we found that the mm-plane EVBME_\mathrm{VBM} of AlN, GaN, and w-BN are within the aa-plane EVBME_\mathrm{VBM} by only 0.2 eV. Therefore, we consider only the aa-plane band alignments in this study, shown in Figure 2c for AlN, Bx_xAl1−x_{1-x}N, and GaN. All of the lowest formation energy aa-plane slab models (Figures S13–S16) and their potential profiles (Figure S17) are available in the Supplementary Information.

In Bx_xAl1−x_{1-x}N aa-plane surfaces, based on the average EVBME_\mathrm{VBM}, we find a strong trend where EVBME_\mathrm{VBM} lowers roughly linearly with xx. The w-BN EVBME_\mathrm{VBM} is roughly 0.7 eV below that of AlN. Additionally, the increasing bandgap of Bx_xAl1−x_{1-x}N with xx means ECBME_\mathrm{CBM} stays relatively constant with xx. We observe generally type-I band alignments of Bx_xAl1−x_{1-x}N with AlN, where the AlN band energies are located within the Bx_xAl1−x_{1-x}N energy gap.

In general, the average aa-plane EVBME_\mathrm{VBM} and ECBME_\mathrm{CBM} are systematically higher in energy (∼2.3\sim2.3 eV) than those of the cc-plane interfaces. This trend has been observed experimentally in GaN[57] and theoretically in InN[58] nonpolar and polar surfaces. We attribute this effect to differences in electronic structure from metallic and semiconducting surface states in polar and nonpolar surfaces, respectively[58]. Interestingly, this leads to negative electron affinity (EA) for x>0.167x > 0.167 surfaces, while the EA are all positive for the Bx_xAl1−x_{1-x}N surfaces. Thus, high-xx aa-plane Bx_xAl1−x_{1-x}N surfaces provide a route for spontaneous electron emitters that can be interfaced with other III-nitrides.

However, the band alignments with Bx_xAl1−x_{1-x}N depend strongly on the surface stoichiometry for the aa-plane surfaces. There are huge standard deviations (≤1.4\leq1.4 eV) in the predicted EVBME_\mathrm{VBM} and ECBME_\mathrm{CBM} of Bx_xAl1−x_{1-x}N structures for the same xx. The standard deviations are higher in Bx_xAl1−x_{1-x}N slabs with lower tetrahedrality[1] and peak at x=0.6x = 0.6. Therefore, this can be largely attributed to higher deviations in the tetrahedral bonds, particularly near the surface. Bond distortion induces dipoles

mixing near the cc-plane interface rather than an abrupt interface consisting of a single nitrogen layer. Thus, while Bx_xAl1−x_{1-x}N/AlN heterojunctions have similar VBO for N- and M-surfaces, this large difference between GaN M- and N-surface EVBME_\mathrm{VBM} and ECBME_\mathrm{CBM} provides a route for band alignment tuning in Bx_xAl1−x_{1-x}N/GaN heterojunctions by modulation of the interfacial stoichiometry. near the surface due to differences in Al-N and B-N bond lengths as well as rotations in the bond angles. These effects dramatically influence the predicted band alignments, leading to high standard deviations.

Additionally, this effect is also related to the Al/B stoichiometry near the surface. We find that higher presences of surface B atoms at the surface lower the band alignment energies, similar to our observations for the cc-plane slabs. Therefore, there is a wide range of possibilities for band alignment tuning in BxAl1−xN\mathrm{B}_{x}\mathrm{Al}_{1-x}\mathrm{N} heterostructures by controlling the interfacial stoichiometry. We find that both type-I and type-II band alignments in BxAl1−xN\mathrm{B}_{x}\mathrm{Al}_{1-x}\mathrm{N}/AlN and BxAl1−xN\mathrm{B}_{x}\mathrm{Al}_{1-x}\mathrm{N}/GaN nonpolar heterostructures could be realized.

There are very few VBO measurements involving nonpolar AlN, GaN, or BN interfaces. In one aa-plane VBO measurement for AlN/GaN, the VBO in aa-plane AlN/GaN (GaN/AlN) was measured to be 0.73±0.160.73 \pm0.16 eV (1.33±0.161.33 \pm0.16 eV)[59], which is much higher than our calculations–we calculate a VBO of 0.200.20 eV for the AlN/GaN heterojunction in the aa-plane surface. This suggests that experimental effects such as intermixing in the AlN/GaN non-polar interfaces could influence the determination of their band alignments strongly, in line with findings from previous theoretical work[60].

To compare the different approaches for calculating band alignments, we compute the natural band alignments for GaN and BxAl1−xN\mathrm{B}_{x}\mathrm{Al}_{1-x}\mathrm{N} similar to the solid-state energy (SSE) approach as implemented to study BxAl1−xN\mathrm{B}_{x}\mathrm{Al}_{1-x}\mathrm{N} by Ota et al.21 shown in Figure 3. In the SSE model, EVBME_{\mathrm{VBM}} and ECBME_{\mathrm{CBM}} are calculated using only the bandgap and the SSE. The SSE depends only on the stoichiometry and a set of empirical parameters for each element, making this an extremely efficient method. For the cations, the SSE was taken to be the average electron affinities of 3, 5, and 5 main group compounds for B, Al, and Ga, respectively. For N, the SSE was taken to be the average ionization potential for the 5 main group compounds of N. Thus we obtained the SSE values as 2.18, 3.14, 3.82, and 7.0 eV, with standard deviations of 1.74, 0.80, 0.47, and

Image not extracted

Figure 3. Natural band alignments for GaN and BxAl1−xN\mathrm{B}_{x}\mathrm{Al}_{1-x}\mathrm{N} calculated using the SSE approach, where EVBME_{\mathrm{VBM}} and ECBME_{\mathrm{CBM}} are represented by blue and red columns respectively, shaded according to the average tetrahedrality in the structure. Blue and red dashed lines show the AlN EVBME_{\mathrm{VBM}} and ECBME_{\mathrm{CBM}} respectively. Dashed gray lines separate the GaN and BxAl1−xN\mathrm{B}_{x}\mathrm{Al}_{1-x}\mathrm{N} results. The opacity of the columns is proportional to the average tetrahedrality of the corresponding bulk structure.

1.09 eV for B, Al, Ga and N, respectively[21, 61]. The EVBME_{\mathrm{VBM}} and ECBME_{\mathrm{CBM}} can then be computed using EVBM=−ξM−0.5EgE_{\mathrm{VBM}}=-\xi_{\mathrm{M}}-0.5E_{\mathrm{g}} and ECBM=−ξM+0.5EgE_{\mathrm{CBM}}=-\xi_{\mathrm{M}}+0.5E_{\mathrm{g}}, where EgE_{\mathrm{g}} is the bandgap, and ξM\xi_{\mathrm{M}} is the geometric mean of the SSE of the constituent atoms[21, 49].

Figure 3 shows that the EVBME_{\mathrm{VBM}} and ECBME_{\mathrm{CBM}} predicted from the SSE approach are systematically lower than those found in cc-plane calculations but higher than those found in aa-plane calculations. Since the SSE approach calculates EVBME_{\mathrm{VBM}} and ECBME_{\mathrm{CBM}} without consideration of interfacial effects (i.e. natural band alignments), the polar and non-polar contributions are effectively averaged.

In BxAl1−xN\mathrm{B}_{x}\mathrm{Al}_{1-x}\mathrm{N}, EVBME_{\mathrm{VBM}} is nearly constant with xx, while ECBME_{\mathrm{CBM}} increases linearly with xx due to the increasing bandgap. However, the EA values are always positive, with w-BN having the lowest predicted EA of 0.170.17 eV. Similar trends can be found in the natural band alignment of AlN, GaN, and w-BN[62].

Using SSE values, the AlN/GaN VBO is 0.980.98 eV, which is within the large range of experimental measurements. In comparison with the BxAl1−xN\mathrm{B}_{x}\mathrm{Al}_{1-x}\mathrm{N}/GaN system however, we find that the B0.125Al0.875N\mathrm{B}_{0.125}\mathrm{Al}_{0.875}\mathrm{N}/GaN VBO of 0.870.87 eV from SSE agrees much more poorly with the cc-plane VBO measurement of 0.2±0.20.2 \pm0.2 for B0.14Al0.86N\mathrm{B}_{0.14}\mathrm{Al}_{0.86}\mathrm{N}/GaN[13] than our cc-plane calculation. There is better agreement with the B0.13Al0.87N\mathrm{B}_{0.13}\mathrm{Al}_{0.87}\mathrm{N}/GaN measured VBO of 1.0±0.11.0 \pm0.1 by Rather et al., although the B0.13Al0.87N\mathrm{B}_{0.13}\mathrm{Al}_{0.87}\mathrm{N}/AlN VBO of 0.5±0.10.5 \pm0.1 eV compares less favorably to the value of −0.11-0.11 eV for B0.125Al0.875N\mathrm{B}_{0.125}\mathrm{Al}_{0.875}\mathrm{N}/AlN found using the SSE approach[20].

Therefore, while the SSE approach provides a computationally inexpensive method for estimating band alignments in bulk materials, it is limited by the available data. Surface-dependent effects are important to consider for reliable band alignment prediction[60] and can change EVBME_{\mathrm{VBM}} by up to several eV in this system. In particular, this is important for B-containing compounds. The SSE for B is calculated using a low number of data points (3) with a high standard deviation (1.74 eV), which helps to explain the large discrepancies between measured and calculated VBO in BxAl1−xN\mathrm{B}_{x}\mathrm{Al}_{1-x}\mathrm{N}/GaN and BxAl1−xN\mathrm{B}_{x}\mathrm{Al}_{1-x}\mathrm{N}/AlN systems, as well as the non-linearities observed in our surface plane-dependent calculations.

In conclusion, we have employed DFT methods to study the band alignments of AlN, GaN, and BxAl1−xN\mathrm{B}_{x}\mathrm{Al}_{1-x}\mathrm{N} for cc-plane and aa-plane slabs, finding generally good agreement with the available experimental data. We use GW0GW_{0} methods for accurate bandgap prediction without the use of any empirical parameters. These calculations constitute the first surface-dependent band alignment predictions for BxAl1−xN\mathrm{B}_{x}\mathrm{Al}_{1-x}\mathrm{N} and GaN. For the polar cc-planes, generally low VBO are found for low-xx BxAl1−xN\mathrm{B}_{x}\mathrm{Al}_{1-x}\mathrm{N}/AlN heterostructures, but there are non-linearities at intermediate xx (0.333≤x≤0.6670.333 \le x \le0.667) BxAl1−xN\mathrm{B}_{x}\mathrm{Al}_{1-x}\mathrm{N}. The nonlinearities are correlated with lower tetrahedrality in the bulk BxAl1−xN\mathrm{B}_{x}\mathrm{Al}_{1-x}\mathrm{N} structures. ECBME_{\mathrm{CBM}} of nearly all BxAl1−xN\mathrm{B}_{x}\mathrm{Al}_{1-x}\mathrm{N} structures is higher than that of AlN due to the increased bandgap, resulting in type-II (staggered) band alignments in BxAl1−xN\mathrm{B}_{x}\mathrm{Al}_{1-x}\mathrm{N}/AlN heterostructures. In the non-polar aa-plane surfaces, the average BxAl1−xN\mathrm{B}_{x}\mathrm{Al}_{1-x}\mathrm{N} EVBME_{\mathrm{VBM}} decreases with xx with a near constant ECBME_{\mathrm{CBM}}, indicating that aa-plane BxAl1−xN\mathrm{B}_{x}\mathrm{Al}_{1-x}\mathrm{N} heterojunctions have potential electron-conductive layer applications. However, there is a strong dependence on the distortion of Al-N and B-N tetrahedral bonds near the surface. We provide a comparison to band alignments calculated using the SSE approach, demonstrating that the surface effects are important to consider. These results mark a huge first step toward establishing the performance of BxAl1−xN\mathrm{B}_{x}\mathrm{Al}_{1-x}\mathrm{N}/AlN and BxAl1−xN\mathrm{B}_{x}\mathrm{Al}_{1-x}\mathrm{N}/GaN heterojunction-based devices.

Acknowledgments

This work was supported by ULTRA, an Energy Frontier Research Center funded by the U.S. Department of Energy (DOE), Office of Science, Basic Energy Sciences (BES), under Award No. DESC0021230. This research used computing resources from the San Diego Supercomputer Center under the NSF-XSEDE and NSF-ACCESS Award No. DMR150006, and the Research Computing facility at Arizona State University.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

References

  1. [1]¹C. J. Milne, T. Biswas, and A. K. Singh, “Ab initio prediction of ultra-wide band gap BₓAl₁₋ₓN materials,” Advanced Electronic Materials 9, 2201197 (2023), https://onlinelibrary.wiley.com/doi/pdf/10.1002/aelm.202201197.DOI
  2. [2]²C. J. Milne, T. Biswas, and A. K. Singh, “Electronic properties of ultra-wide bandgap BₓAl₁₋ₓN computed from first-principles simulations,” Advanced Electronic Materials (2024), 10.1002/aelm.202400549.
  3. [3]³R. Kudrawiec and D. Hommel, “Bandgap engineering in III-nitrides with boron and group V elements: Toward applications in ultraviolet emitters,” Applied Physics Reviews 7, 041314 (2020).DOI
  4. [4]⁴J. Hayden, M. D. Hossain, Y. Xiong, K. Ferri, W. Zhu, M. V. Imperatore, N. Giebink, S. Trolier-McKinstry, I. Dabo, and J. P. Maria, “Ferroelectricity in boron-substituted aluminum nitride thin films,” Physical Review Materials 5 (2021), 10.1103/PhysRevMaterials.5.044412.DOI
  5. [5]⁵W. Zhu, J. Hayden, F. He, J. I. Yang, P. Tipsawat, M. D. Hossain, J. P. Maria, and S. Trolier-McKinstry, “Strongly temperature dependent ferroelectric switching in AlN, Al₁₋ₓScₓN, and Al₁₋ₓBₓN thin films,” Applied Physics Letters 119 (2021), 10.1063/5.0057869.DOI
  6. [6]⁶S. Calderon, J. Hayden, S. M. Baksa, W. Tzou, S. Trolier-McKinstry, I. Dabo, J.-P. Maria, and E. C. Dickey, “Atomic-scale polarization switching in wurtzite ferroelectrics,” Science 380, 1034–1038 (2023).DOI
  7. [7]⁷C. Savant, T.-S. Nguyen, K. Nomoto, S. Vishwakarma, S. Ma, A. Dhar, Y.-H. Chen, J. Casamento, D. J. Smith, H. G. Xing, and D. Jena, “Epitaxial high-K barrier AlBN/GaN HEMTs,” Applied Physics Letters 126, 112906 (2025).
  8. [8]⁸T. Akasaka, Y. Kobayashi, and T. Makimoto, “Nonpolar AlBN (11̄20) and (1̄100) films grown on SiC substrates,” Applied Physics Letters 91, 041914 (2007).
  9. [9]⁹X. Li, S. Sundaram, Y. E. Gmili, T. Moudakir, F. Genty, S. Bouchoule, G. Patriarche, R. D. Dupuis, P. L. Voss, J.-P. Salvestrini, and A. Ougazzaden, “BAIN thin layers for deep UV applications,” Physica Status Solidi (a) 212, 745–750 (2015).
  10. [10]¹⁰X. Li, S. Wang, H. Liu, F. A. Ponce, T. Detchprohm, and R. D. Dupuis, “AlN thin single-phase wurtzite BAlN films with boron contents over 10%,” Physica Status Solidi (b) 254 (2017), 10.1002/pssb.201600699.
  11. [11]¹¹H. Sun, Y. J. Park, K. H. Li, C. G. T. Castanedo, A. Alowayed, T. Detchprohm, R. D. Dupuis, and X. Li, “Band alignment of B₀.₁₄Al₀.₈₆N/Al₀.₇Ga₀.₃N heterojunction,” Applied Physics Letters 111 (2017), 10.1063/1.4999249.DOI
  12. [12]¹²H. Sun, F. Wu, Y. J. Park, T. M. A. Tahtamouni, C. H. Liao, W. Guo, N. Alfaraj, K. H. Li, D. H. Anjum, T. Detchprohm, R. D. Dupuis, and X. Li, “Revealing microstructure and dislocation behavior in BAlN/AlGaN heterostructures,” Applied Physics Express 11 (2018), 10.7567/apex.11.011001.DOI
  13. [13]¹³H. Sun, Y. J. Park, K. H. Li, X. Liu, T. Detchprohm, X. Zhang, R. D. Dupuis, and X. Li, “Nearly-zero valence band and large conduction band offset at AlN/GaN heterointerface for optical and power device application,” Applied Surface Science 458, 949–953 (2018).DOI
  14. [14]¹⁴T. B. Tran, C. H. Liao, F. Alqatari, and X. Li, “Demonstration of single-phase wurtzite BaInN with over 20% boron content by metalorganic chemical vapor deposition,” Applied Physics Letters 117 (2020), 10.1063/5.0019881.DOI
  15. [15]¹⁵P. Vuong, A. Mballo, S. Sundaram, G. Patriarche, Y. Halfaya, S. Karkachou, A. Srivastava, K. Krishnan, N. Y. Sama, T. Ayari, S. Gautier, P. L. Voss, J. P. Salvestrini, and O. Ouagzaden, “Single crystalline boron rich BAlN alloys grown by MOVPE,” Applied Physics Letters 116 (2020), 10.1063/1.5135505.DOI
  16. [16]¹⁶J. Sarkar, T. B. Tran, F. Alqatari, C. H. Liao, X. Li, and B. Mazumder, “Nanoscale compositional analysis of wurtzite BAlN thin film using atom probe tomography,” Applied Physics Letters 117 (2020), 10.1063/5.0027861.
  17. [17]¹⁷Q. Zhang, Q. Li, W. Zhang, H. Zhang, F. Zheng, M. Zhang, P. Hu, M. Wang, Z. Tian, Y. Li, Y. Liu, and F. Yun, “Phase transition and bandgap engineering in B₁−ₓAlₓN alloys: DFT calculations and experiments,” Applied Surface Science 575, 151641 (2022).DOI
  18. [18]¹⁸N. Wolff, S. Fichtner, B. Haas, M. R. Islam, F. Nickiel, M. Kessel, O. Ambacher, C. Koch, B. Wagner, F. Lofink, and L. Kienle, “Atomic scale confirmation of ferroelectric polarization inversion in wurtzite-type AlScN,” Journal of Applied Physics 129 (2021), 10.1063/5.0033205.DOI
  19. [19]¹⁹S. H. Yoo, M. Todorova, D. Wickramaratne, L. Weston, C. G. de Walle, and J. Neugebauer, “Finite-size correction for slab supercell calculations of materials with spontaneous polarization,” npj Computational Materials 7 (2021), 10.1038/s41524-021-00529-1.DOI
  20. [20]²⁰M. A. Rather, L. Ravi, T. Y. Yu, C. T. Wu, K. L. Lin, K. Y. Lai, and J. I. Chyi, “An experimental study of the energy band alignments of B(Al, Ga)N heterojunctions,” Applied Physics Letters 123 (2023), 10.1063/5.0116591.DOI
  21. [21]²¹Y. Ota, M. Imura, R. G. Banal, and Y. K. Yoide, “Natural band alignment of BAlN and BGaN alloys,” Journal of Physics D: Applied Physics 55 (2022), 10.1088/1361-6463/ac8fff.DOI
  22. [22]²²A. Biswas, T. S. Pieshkov, C. L. Milne, H. Gomez, C. Li, A. B. Puthirath, X. Zhang, T. Gray, T. Terlier, R. Vajtai, E. J. Garratt, A. G. Birdwell, T. G. Ivanov, A. K. Singh, B. P. Pate, M. R. Neupane, and P. M. Ajayan, “Growth of cubic boron nitride thin film on diamond surface,” Small n/a, e06840 (2025), https://onlinelibrary.wiley.com/doi/pdf/10.1002/smll.202506840.DOI
  23. [23]²³C. L. Milne, H. Gomez, A. Gupta, A. G. Birdwell, S. Rudin, E. J. Garratt, B. B. Pate, T. G. Ivanov, A. K. Singh, and M. R. Neupane, “Interface structure and electronic properties in cubic boron nitride — diamond heterostructures,” Materials Today Physics 59, 101887 (2025).
  24. [24]²⁴Y. C. Tsai and C. Bayram, “Band alignments of ternary wurtzite and zincblende III-nitrides investigated by hybrid density functional theory,” ACS Omega 5, 3917–3923 (2020).DOI
  25. [25]²⁵P. G. Moses, M. Miao, Q. Yan, and C. G. V. D. Walle, “Hybrid functional investigations of band gaps and band alignments for AlN, GaN, InN, and InGaN,” Journal of Chemical Physics 134 (2011), 10.1063/1.3548872.DOI
  26. [26]²⁶Y. Hinuma, A. Grüneis, G. Krese, and F. Oba, “Band alignment of semiconductors from density-functional theory and many-body perturbation theory,” Physical Review B 90, 155405 (2014).DOI
  27. [27]²⁷Y. Hinuma, Y. Kumagai, T. Tanaka, and F. Oba, “Band alignment of semiconductors and insulators using dielectric-dependent hybrid functionals: Toward high-throughput evaluation,” Physical Review B 95, 075302 (2017).DOI
  28. [28]²⁸K. Shiraishi, “A new slab model approach for electronic structure calculation of polar semiconductor surface,” Journal of the Physical Society of Japan 59, 3455–3458 (1990).DOI
  29. [29]²⁹G. Kresse and J. Hafner, “Ab initio molecular dynamics for liquid metals,” Physical Review B 47, 558 (1993).
  30. [30]³⁰G. Kresse and J. Hafner, “Ab initio molecular-dynamics simulation of the liquid-metal-amorphous-semiconductor transition in germanium,” Physical Review B 49, 14251 (1994).pubmed.ncbi.nlm.nih.gov/10010505
  31. [31]³¹G. Kresse and J. Furthmüller, “Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave set,” Computational Materials Science 6, 15 (1996).DOI
  32. [32]³²G. Kresse and J. Furthmüller, “Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set,” Physical Review B 54, 11169–11186 (1996).pubmed.ncbi.nlm.nih.gov/9984901
  33. [33]³³J. P. Perdew, K. Burke, and M. Ernzerhof, “Erratum generalized gradient approximation made simple,” Physical Review Letters 78, 1396 (1997).
  34. [34]P. E. Blochl, “Projector augmented-wave method,” Physical Review B 50, 17953 (1994).pubmed.ncbi.nlm.nih.gov/9976227
  35. [35]J. P. Perdew, K. Burke, and M. Ernzerhof, “Generalized gradient approximation made simple,” Physical Review Letters 77, 3865 (1996).pubmed.ncbi.nlm.nih.gov/10062328
  36. [36]C. Freysoldt, A. Mishra, M. Ashton, and J. Neugebauer, “Generalized dipole correction for charged surfaces in the repeated-slab approach,” Physical Review B 102, 045403 (2020).DOI
  37. [37]S. P. Ong, W. D. Richards, A. Jain, G. Hautier, M. Kocher, S. Cholia, D. Gunter, V. L. Chevrier, K. A. Persson, and G. Ceder, “Python Materials Genomics (pymatgen): A robust, open-source python library for materials analysis,” Computational Materials Science 68, 314–319 (2013).DOI
  38. [38]Y. Liang, S. Huang, R. Soklaski, and L. Yang, “Quasiparticle band-edge energy and band offsets of monolayer of molybdenum and tungsten chalcogenides,” Applied Physics Letters 103 (2013), 10.1063/1.4816517.DOI
  39. [39]T. Biswas and A. K. Singh, “pyGWBSE: a high throughput workflow package for GW-BSE calculations,” npj Computational Materials 9 (2023), 10.1038/s41524-023-00976-y.arxiv.org/abs/2210.00152
  40. [40]T. Biswas and A. K. Singh, “Excitonic effects in absorption spectra of carbon dioxide reduction photocatalysts,” npj Computational Materials 7 (2021), 10.1038/s41524-021-00640-3.arxiv.org/abs/2103.08384
  41. [41]T. Biswas, A. Gupta, and A. K. Singh, “Many-body physics and machine learning enabled discovery of promising solar materials,” RSC advances 15, 8253–8261 (2025).DOI
  42. [42]R. Anderson, “Germanium-gallium arsenide heterojunctions,” IBM Journal of Research and Development 4, 283–287 (1960).
  43. [43]M. B. Nardelli, K. Rapcewicz, and J. Bernholc, “Theory of interfaces and surfaces in wide-gap nitrides,” Journal of Vacuum Science & Technology B: Microelectronics and Nanometer Structures Processing, Measurement, and Phenomena 15, 1144–1147 (1997).
  44. [44]M. B. Nardelli, K. Rapcewicz, and J. Bernholc, “Strain effects on the interface properties of nitride semiconductors,” Physical Review B 55, R7323 (1997).DOI
  45. [45]P. Reddy, I. Bryan, Z. Bryan, W. Guo, L. Hussey, R. Collazo, and Z. Sitar, “The effect of polarity and surface states on the Fermi level at III-nitride surfaces,” Journal of Applied Physics 116 (2014).DOI
  46. [46]S. King, C. Ronning, R. Davis, M. Benjamin, and R. Nemanich, “Dependence of (0001) GaN/AlN valence band discontinuity on growth temperature and surface reconstruction,” Journal of Applied Physics 84, 2086–2090 (1998).DOI
  47. [47]J. Baur, K. Maier, M. Kunzer, U. Kaufmann, and J. Schneider, “Determination of the GaN/AlN band offset via the (-/0) acceptor level of iron,” Applied Physics Letters 65, 2211–2213 (1994).
  48. [48]A. Franciosi, L. Sorba, G. Bratina, and G. Biasiol, “Modification of heterojunction band offsets at III–V/IV/III–V interfaces,” Journal of Vacuum Science & Technology B: Microelectronics and Nanometer Structures Processing, Measurement, and Phenomena 11, 1628–1637 (1993).
  49. [49]Y. Ota, “Band alignment via atomic solid-state energy scale approach,” AIP Advances 10, 125321 (2020), https://pubs.aip.org/aip/adv/article-pdf/doi/10.1063/5.0031224/12945337/125321_1_online.pdf.
  50. [50]H. He and S. Yang, “First-principles study on improvement of two-dimensional hole gas concentration and confinement in AlN/GaN superlattices,” Chinese Physics B 31, 017104 (2022).DOI
  51. [51]G. Martin, A. Botchkarev, A. Rockett, and H. Morcoc, “Valence-band discontinuities in wurtzite GaN, AlN, and InN heterojunctions measured by x-ray photoemission spectroscopy,” Applied Physics Letters 68, 2541–2543 (1996).
  52. [52]S. Zoino, L. Borowik, B. Mohamad, E. Nowak, and P. Kempisty, “Ab initio investigations of two-dimensional carrier gas at interfaces in GaN/AlN and GaN/AlN/Al2O3 heterostructures,” Journal of Applied Physics 134, 155304 (2023), https://pubs.aip.org/aip/jap/article-pdf/doi/10.1063/5.0169332/18174401/155304_1_5.0169332.pdf.DOI
  53. [53]W.-C. Yang, B. J. Rodriguez, M. Park, R. Nemanich, O. Ambacher, and V. C. Cimalla, “Photoelectron emission microscopy observation of inversion domain boundaries of GaN-based lateral polarity heterostructures,” Journal of Applied Physics 94, 5720–5725 (2003).DOI
  54. [54]H. Li, X. Liu, L. Sang, J. Wang, D. Jin, H. Zhang, S. Yang, S. Liu, W. Mao, Y. Hao, et al., “Determination of polar c-plane and nonpolar a-plane AlN/GaN heterojunction band offsets by x-ray photoelectron spectroscopy,” Physica Status Solidi (b) 251, 788–791 (2014).DOI
  55. [55]J. Waldrop and R. Grant, “Measurement of AlN/GaN (0001) heterojunction band offsets by x-ray photoemission spectroscopy,” Applied physics letters 68, 2879–2881 (1996).DOI
  56. [56]B. Roul, M. Kumar, M. K. Rajpalke, T. N. Bhat, and S. Krupanidhi, “Binary group III-nitride based heterostructures: band offsets and transport properties,” Journal of Physics D: Applied Physics 48, 423001 (2015).DOI
  57. [57]M. Mishra, T. S. Krishna, N. Aggarwal, and G. Gupta, “Surface chemistry and electronic structure of nonpolar and polar GaN films,” Applied Surface Science 345, 440–447 (2015).DOI
  58. [58]A. Belabbes, J. Furthmüller, and F. Bechstedt, “Electronic properties of polar and nonpolar InN surfaces: A quasiparticle picture,” Physical Review B—Condensed Matter and Materials Physics 84, 203504 (2011).
  59. [59]L. Sang, Q. S. Zhu, S. Y. Yang, G. P. Liu, H. J. Li, H. Y. Wei, C. M. Jiao, S. M. Liu, Z. G. Wang, X. W. Zhou, W. Mao, Y. Hao, and B. Shen, “Band offsets of non-polar α-plane GaN/AlN and AlN/GaN heterostructures measured by x-ray photoemission spectroscopy,” Nanoscale Research Letters 9, 470 (2014).DOI
  60. [60]C. Kaewmeechai, Y. Laosiritaworn, and A. P. Jaroenjittichai, “Hybrid functional investigation of band offsets for non-polar, Ga-polar and Al-polar interfaces in GaN/AlN heterojunction,” Journal of Physics: Condensed Matter 33, 035005 (2020).DOI
  61. [61]B. D. Pelatt, J. F. Wager, and D. A. Keszler, “Elucidation of bonding trends from variability in atomic solid state energies,” Journal of Solid State Chemistry 274, 337–351 (2019).DOI
  62. [62]C. E. Dreyer, J. L. Lyons, A. Janotti, and C. G. V. de Walle, “Band alignments and polarization properties of BN polymorphs,” Applied Physics Express 031001 (2014), 10.7567/1882-0786/ab5e1c.DOI

Paper details

vlnc:W_tr5zz8cf
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 5

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