Diamond, the ultra-wide-bandgap semiconductor with exceptional thermal conductivity and breakdown field, is blocked for electronics chiefly by its doping problem: useful substitutional dopants are deep (B acceptor ≈0.37 eV, P donor ≈0.57 eV, N donor ≈1.7 eV), poorly soluble, and mobility-degrading at high concentration. Ruan, Li, and Yakobson (Adv. Funct. Mater. 34, 2408536, 2024) proposed bypassing this bottleneck by modulation doping: substituting an atomically thin cubic-BN layer into a diamond crystal plane creates a polar interfacial dipole and a multi-eV (≈3–4 eV) band offset near the [100] and [111] directions, which a closed-form planar-capacitor model with empirically assigned effective charges reproduces and screens over 4,096 crystallographic directions. The unresolved open problem is to convert that pristine-interface principle into a validated quantitative determination of whether diamond is actually modulation-doped by such an insert; its source-supported core is the self-consistent dopant-ionization, charge-transfer, and screening calculation yielding carrier densities and the modulation-doping threshold. Related aspects that bear on the accuracy and practical feasibility of such a determination - an ab initio-consistent, transferable parametrization of the band-offset model, quantified quasiparticle (GW) and temperature corrections, and the chemical vapor deposition (CVD) growth feasibility of the required atomically sharp insert - remain open in the same literature.

Background

Diamond is an ultra-wide-bandgap semiconductor (band gap 5.47 eV) with exceptional thermal conductivity, carrier mobility, and breakdown field, making it attractive for high-power and high-frequency electronics; its use is blocked chiefly by doping. The useful substitutional dopants are deep: the boron acceptor lies about 0.37 eV above the valence-band edge, the phosphorus donor about 0.57 eV below the conduction-band edge, and the nitrogen donor about 1.7 eV below it, so only a small fraction ionizes at ordinary temperatures; thermodynamic solubility is low, and heavy doping degrades mobility. Surface transfer doping (adsorbed molecular acceptors on hydrogen-terminated diamond inducing a surface hole channel without substitutional dopants) has so far produced mainly unstable p-type channels. These bottlenecks are surveyed in the source paper and in the ultra-wide-bandgap semiconductor roadmap of Tsao et al.

Modulation doping, the standard remedy in III-V heterostructures (e.g., the GaAs/AlGaAs high-electron-mobility transistor), spatially separates ionized dopants from the carriers they release: dopants sit in a neighboring layer and the released carriers fall into an undoped transport layer, avoiding ionized-impurity scattering. Implementing it inside diamond requires a polar, lattice-matched partner whose interfacial dipoles shift the band edges of adjacent diamond regions by more than the dopant activation energies. Cubic boron nitride (cBN), the zinc-blende, sp3-bonded polymorph of BN that is isoelectronic with diamond and lattice-mismatched by only about 1.4%, is the natural candidate: its polar B-N bonds carry a net dipole across a substituted layer.

Ruan, Li, and Yakobson studied a diamond|cBN|diamond 'insert': the minimum-thickness set of carbon atoms on a chosen crystallographic plane (hkl) replaced by bonded B-N pairs. Using Kohn-Sham density functional theory (DFT; Perdew-Burke-Ernzerhof exchange-correlation functional, projector-augmented-wave potentials, supercells longer than 25 angstroms along the interface normal) and macroscopic averaging of the electrostatic potential (smoothing the planar average over a window of 0.5-1.5 angstroms to expose the bulk-like baselines on each side), they extracted normalized band offsets of 4.6, 2.9, 2.2, 1.7, 1.2, and 1.2 eV for the (100), (111), (210), (221), (322), and (331) orientations - that is, multi-eV (about 3-4 eV) offsets in practice near the [100] and [111] directions. They also developed a closed-form 'planar capacitor' model: the effective interfacial charges smeared over infinite crystal planes give the electrostatic potential step Δ=2q(h+k)σ0d/ε0\Delta = 2q(h+k)\sigma_0 d/\varepsilon_0, with σ0=1/(a2h2+k2+l2)\sigma_0 = 1/(a^2\sqrt{h^2+k^2+l^2}) for lattice constant aa and B-N plane spacing dd, and effective charge qq on B (+) and N (-). Assigning q=0.46eq = 0.46e (from an experimental charge analysis) and fitted charges on the carbons adjacent to the insert (0.217e for planes with two even and one odd Miller index, 0.0929e for all-odd planes) - choices guided by DDEC charge partitioning, a scheme that partitions the computed electron density onto atoms - the model reproduces the DFT offsets to within about 0.1-0.2 eV. Screening 4,096 symmetry-inequivalent directions [hkl] (indices 0 to 15) locates the largest offsets near [100], [111], and [210]. For the (111) case, band structures projected onto the 2s and 2p orbitals of carbon atoms on each side of the insert show band edges shifted across the common dopant activation energies, suggesting both n-type (donors on the boron side) and p-type (acceptors on the nitrogen side) modulation-doping geometries. Earlier theory had addressed diamond-BN interfaces without the embedded-insert concept or dopants: Pickett's tight-binding superlattice study of the nonpolar (110) diamond-BN interface (1988), and the DFT study of Zhao et al. (2019) of band discontinuities at cBN/diamond junctions.

What the source study deliberately does not do defines the open problem: above all, its band alignments contain no dopants or free carriers; in addition, its model parameters are empirical rather than derived from the computed charge redistribution, its energy scales carry no many-body or temperature corrections, and the growth of the idealized insert is untested.

Problem Statement

Establish, quantitatively, whether an atomically thin cBN insert embedded in diamond can modulation-dope diamond - that is, whether donors (or acceptors) placed on one side of the insert ionize and release carriers that accumulate and conduct in the undoped diamond on the other side. The source study establishes only the pristine-interface principle: multi-eV band offsets computed with no dopants or free carriers present, with the actual doping capability asserted only as an expectation ('expected to facilitate effective modulation doping of diamond'). The open problem is the self-consistent determination that closes the gap between the demonstrated offset and an actual doping capability: a self-consistent calculation of the coupled dopant ionization, free-carrier redistribution, and electrostatic screening for realistically doped diamond|cBN|diamond structures (dopants on one side, in both the n-type geometry with donors on the boron side and the p-type geometry with acceptors on the nitrogen side), yielding the transferred carrier densities, the expected transport behavior, and the threshold insert thickness, band offset, and doping conditions below which modulation doping fails - including the back-action of the screening space charge on the very band offsets that drive the transfer.

Three related aspects bear on the accuracy and practical reach of this determination and remain open in the same literature; they are recorded here as context rather than as jointly necessary completion requirements. First, an ab initio-consistent mapping from the computed electronic charge redistribution to the effective-charge parameters of the band-offset model - replacing the empirical assignment of 0.46e0.46e to B/N and the fitted adjacent-carbon charges - with demonstrated transferability across cBN insert thicknesses, strain states, and interface reconstructions, or an equivalent predictive model of comparable screening speed validated against explicit first-principles calculations. Second, quantified corrections to the band-edge positions, band offsets, and dopant activation energies from many-body quasiparticle effects (GW-type corrections, i.e., the standard many-body perturbation-theory upgrade of DFT band energies) and from temperature, and their consequence for the predicted modulation-doping threshold. Third, the stability and growth kinetics of an atomically sharp, controlled-thickness, phase-pure cBN insert buried in diamond by chemical vapor deposition (CVD) - interdiffusion of B, N, and C, interfacial defect formation and reconstruction, and crystallographic-orientation control - which determines whether the idealized structures studied in theory can be realized.

These components share one scientific objective: converting the pristine-insert band-offset principle into a validated quantitative prediction (and ultimately a testable realization) of modulation-doped diamond. The self-consistent dopant-carrier electrostatics is the crux - an answer that leaves it uncoupled does not resolve the problem, however accurate its band offsets. A resolution states whether charge transfer occurs, at what carrier density (comparable, for instance, to what direct substitutional doping achieves in diamond), and under what insert and doping conditions.

References

  1. Qiyuan Ruan, Xingfu Li, Boris I. Yakobson, 'Electronic Band Offset in a Diamond|cBN|Diamond System for Modulation Doping', Advanced Functional Materials 34, 2408536 (2024). DOI: 10.1002/adfm.202408536, https://doi.org/10.1002/adfm.202408536
  2. Qiyuan Ruan, Xingfu Li, Boris I. Yakobson, 'Electronic band offset in a diamond|cBN|diamond system for modulation doping', ChemRxiv preprint posted 2024-05-16. DOI: 10.26434/chemrxiv-2024-0rkw6, https://doi.org/10.26434/chemrxiv-2024-0rkw6
  3. Warren E. Pickett, 'Thin superlattices and band-gap discontinuities: The (110) diamond-boron nitride interface', Physical Review B 38, 1316-1322 (1988). DOI: 10.1103/PhysRevB.38.1316, https://doi.org/10.1103/PhysRevB.38.1316
  4. Dehe Zhao, Wei Gao, Yujing Li, Yuyuan Zhang, Hong Yin, 'The electronic properties and band-gap discontinuities at the cubic boron nitride/diamond hetero-interface', RSC Advances 9, 8435-8443 (2019). DOI: 10.1039/c9ra00784a, https://doi.org/10.1039/c9ra00784a
  5. Ali Ebadi Yekta, Martha R. McCartney, David J. Smith, Norio Tokuda, Robert J. Nemanich, 'Band offset between cubic boron nitride and nitrogen-plasma terminated boron-doped diamond (111)', Applied Physics Letters 127, 171603 (2025). DOI: 10.1063/5.0278816, https://doi.org/10.1063/5.0278816
  6. Ting Cheng, Ksenia V. Bets, Boris I. Yakobson, 'Hydrogen-Driven Phase Differentiation in BN Nucleation on Diamond', Advanced Functional Materials 35, 2413675 (2024). DOI: 10.1002/adfm.202413675, https://doi.org/10.1002/adfm.202413675
  7. Lu Bai, Jiayin Li, Jiaxin Ming, Yuguo Chen, Wenhongxin Li, Pengcheng Zhao, Jiangnan Song, Tiansong Zhang, Wuxiao Han, Guoshuai Du, Kenji Watanabe, Takashi Taniguchi, Jian-guo Zhang, Tiange Bi, Yabin Chen, 'Controlled Synthesis and Interfacial Bonding of Hexagonal Diamond/Wurtzite Boron Nitride Covalent Heterostructures', Advanced Materials, e74216 (2026). DOI: 10.1002/adma.74216, https://doi.org/10.1002/adma.74216
  8. Shaohua Lu, Hanyang Wang, Yuhao Zheng, Xiaojun Hu, 'First-principles analysis of biaxial-strain reducing ionization energy in a doped diamond', Diamond and Related Materials 163, 113409 (2026). DOI: 10.1016/j.diamond.2026.113409, https://doi.org/10.1016/j.diamond.2026.113409
  9. Hanqing Wei, Haifei Zhan, Dominik Legut, Shihao Zhang, 'A dislocation perspective on heterointerfacial strengthening in nanostructured diamond and cubic boron nitride composites', Carbon 235, 120079 (2025). DOI: 10.1016/j.carbon.2025.120079, https://doi.org/10.1016/j.carbon.2025.120079
  10. J. Y. Tsao, S. Chowdhury, M. A. Hollis, D. Jena, N. M. Johnson, K. A. Jones, R. J. Kaplar, S. Rajan, C. G. Van de Walle, E. Bellotti, C. L. Chua, R. Collazo, M. E. Coltrin, J. A. Cooper, K. R. Evans, S. Graham, T. A. Grotjohn, E. R. Heller, M. Higashiwami, M. S. Islam, P. W. Juodawlkis, M. A. Khan, A. D. Koehler, J. H. Leach, U. K. Mishra, R. J. Nemanich, R. C. Pilawa-Podgurski, J. B. Shealy, Z. Sitar, M. J. Tadjer, A. F. Witulski, M. Wraback, J. A. Simmons, 'Ultrawide-Bandgap Semiconductors: Research Opportunities and Challenges', Advanced Electronic Materials 4, 1600501 (2018). DOI: 10.1002/aelm.201600501, https://doi.org/10.1002/aelm.201600501