CuBC: Copper Borocarbide & Superconductivity

• CuBC is a copper borocarbide defined by BC honeycomb layers bridged by C–Cu–C dumbbells, resulting in an atypical AA stacking order.
• Structural analyses reveal that the high-symmetry hP3 phase is dynamically unstable, leading to a more stable mP6 phase via subtle shear distortions.
• Partial Cu occupancy in CuₓBC enhances hole doping of BC-pₓ,y bands, markedly improving superconducting properties compared to the stoichiometric form.

Searching arXiv for the cited paper to ground the article. CuBC, in the context of layered borocarbides, denotes the copper analogue of AgBC: a metastable borocarbide comprising BC honeycomb sheets linked by C–Cu–C dumbbells rather than by simple intercalant planes. In "High-$T_{\rm c}$ Ag$_x$BC and Cu$_x$BC superconductors accessible via topochemical reactions" (Gochitashvili et al., 18 Jul 2025), CuBC is predicted to be metallic, structurally prone to shear, and only weakly superconducting at stoichiometric composition, whereas partially Cu-filled Cu$_x$BC phases are identified as more favorable realizations of a hole-doped BC network and therefore better superconducting candidates. The same source places CuBC within a broader design program for MgB$_2$-like superconductors based on covalent borocarbide frameworks accessible by topochemical ion exchange.

1. Structural identity and crystallographic motifs

All borocarbides considered in the study share a common structural backbone of 2D BC honeycomb layers stacked along $c$, with alternating B and C on a graphene-like lattice (Gochitashvili et al., 18 Jul 2025). In LiBC, NaBC, MgB$_2$C$_2$, and related compounds, the interlayer sites are occupied by cations located roughly at the centers of the hexagons between neighboring layers, giving AA$'$-stacked structures of honeycomb layers separated by intercalant planes. CuBC differs fundamentally from this motif.

Evolutionary structure searches and systematic decoration studies indicate that Cu, like Ag, prefers to form C–M–C dumbbells rather than occupy the middle of interlayer hexagons (Gochitashvili et al., 18 Jul 2025). These dumbbells connect adjacent BC layers and enforce AA stacking rather than AA$'$. The resulting structural motif is therefore not an intercalated layered material in the LiBC sense, but a layered borocarbide in which the BC honeycomb sheets are vertically bridged by C–Cu–C units.

For stoichiometric CuBC, the study identifies a high-symmetry hexagonal hP3 prototype analogous to hP3-LiBC but with linear C–Cu–C dumbbells along $_x$0 (Gochitashvili et al., 18 Jul 2025). However, this hP3 structure is dynamically unstable. Following the unstable phonon mode in a $_x$1 supercell produces a monoclinic mP6 phase that is dynamically stable and slightly lower in energy. The difference between hP3 and mP6 is structurally subtle but electronically consequential: in hP3 the BC layers sit directly above one another and the dumbbells are straight, whereas in mP6 adjacent BC layers are slightly shifted laterally and the dumbbells are tilted.

The energy landscape associated with this distortion is radially degenerate, described in the source as a Mexican-hat surface around the high-symmetry stacking (Gochitashvili et al., 18 Jul 2025). This suggests that real CuBC may exhibit stacking disorder rather than an ideally ordered monoclinic ground state. A plausible implication is that experimentally realized CuBC, if obtained, may be structurally heterogeneous even when the local dumbbell-bridged BC framework is preserved.

2. Stability, distortions, and comparison with AgBC and LiBC

The distinction between CuBC, AgBC, and LiBC is central to the interpretation of CuBC. LiBC in hP6 form contains Li in hexagon centers between AA$_x$2-stacked BC layers and is semiconducting in stoichiometric form (Gochitashvili et al., 18 Jul 2025). AgBC, by contrast, also adopts C–Ag–C dumbbells, but the linear hP3 arrangement is dynamically and thermally stable, with no imaginary phonons at $_x$3 GPa and stability in ab-initio molecular dynamics at $_x$4 K.

CuBC occupies an intermediate but frustrated position. Relative to LiBC, its interlayer expansion is more moderate, about $_x$5, whereas AgBC expands by about $_x$6 to accommodate Ag (Gochitashvili et al., 18 Jul 2025). In CuBC, this smaller spacing and the details of Cu–C bonding lead to instability of the linear dumbbell arrangement. Phonon calculations reveal imaginary modes at the A point in hP3-CuBC corresponding to shearing of BC layers against one another. The stable mP6 structure is lower in energy than hP3 by about $_x$7 meV/atom with optB86b-vdW and by about $_x$8 meV/atom with r$_x$9SCAN+rVV10, while the A-point instability persists in both descriptions (Gochitashvili et al., 18 Jul 2025).

These facts define CuBC as a metastable layered borocarbide whose principal instability is not decomposition of the local BC–Cu connectivity itself, but a lateral shear of the stacked layers. The source further notes that the lowest-symmetry mP6 phase has only about $_x$0 shorter interlayer spacing than hP3-CuBC, so the decisive factor is not a large volume collapse but a small stacking adjustment that strongly alters the electronic structure (Gochitashvili et al., 18 Jul 2025). This makes CuBC a notable example in which weak crystallographic distortions govern whether an MgB$_x$1-like electronic configuration can be maintained.

3. Electronic structure and the origin of electronic frustration

Stoichiometric LiBC, NaBC, MgB$_x$2C$_x$3, BeB$_x$4C$_x$5, and ZnB$_x$6C$_x$7 are all described as semiconducting compounds satisfying an 8-electron rule, with HSE06 band gaps of $_x$8 eV for LiBC, $_x$9 eV for NaBC, $_x$0 eV for MgB$_x$1C$_x$2, $_x$3 eV for BeB$_x$4C$_x$5, and $_x$6 eV for ZnB$_x$7C$_x$8 (Gochitashvili et al., 18 Jul 2025). In those systems, superconductivity requires explicit hole doping of BC-$_x$9 bonds. CuBC and AgBC instead introduce a nearly free-electron metal $_2$0 band that interacts with BC-derived states and can act as an intrinsic hole-doping channel.

In hP3-CuBC, the Cu-$_2$1 band bottom lies just above $_2$2, so the Cu-$_2$3 band is only slightly occupied or even empty (Gochitashvili et al., 18 Jul 2025). As a result, it does not generate strong hole doping of BC-$_2$4. The C–Cu–C dumbbells do cause mixing among Cu-$_2$5, C-$_2$6, B-$_2$7, and Cu-$_2$8, but the alignment of the Cu-$_2$9 level is unfavorable. In the stable mP6 structure, the small shear-and-tilt distortion pushes the Cu-$c$0 band even higher, so it is no longer filled at all. This strongly reduces hole doping of BC-$c$1, leaving those bands almost filled and producing a pseudogap at $c$2.

The source quantifies this suppression through the BC-$c$3 projected density of states at the Fermi level: $c$4 states/(eV·atom) for mP6-CuBC (Gochitashvili et al., 18 Jul 2025). By contrast, hP3-AgBC has an Ag-$c$5 band bottom about $c$6 eV below $c$7 at $c$8, significant Ag-$c$9 occupation, and BC-$_2$0 states crossing $_2$1 with $_2$2 states/(eV·atom). The study attributes this difference primarily to the size difference between Cu and Ag and the corresponding interlayer spacing, which shifts the metal $_2$3 band relative to BC-derived bands.

The paper characterizes stoichiometric CuBC as metallic but electronically frustrated (Gochitashvili et al., 18 Jul 2025). That characterization is precise: the material is not semiconducting, but its metallicity does not arise from strong participation of the desired BC-$_2$4 states at the Fermi level. Instead, the stable structure nearly restores a closed BC-$_2$5 manifold and leaves only a low-density metallic state with a pseudogap. There is no discussion of Dirac points or flat bands; the defining feature is the near-elimination of MgB$_2$6-like BC-$_2$7-band activity at $_2$8.

4. Non-stoichiometric Cu$_2$9BC and superconducting response

The study extends beyond the stoichiometric end member to Cu$_2$0BC and mixed Li$_2$1Cu$_2$2BC phases for several Cu contents, including $_2$3, $_2$4, $_2$5, and $_2$6 (Gochitashvili et al., 18 Jul 2025). The structural trend changes with composition: at $_2$7, Cu prefers dumbbell sites, while at $_2$8 the lowest-energy phases place Cu in interstitial positions between the layers, more akin to Li-type intercalant sites. Intermediate compositions contain mixtures of these motifs, with some galleries fully filled by dumbbells and others half-filled by interstitials.

This change in Cu filling is accompanied by a substantial increase in BC-$_2$9 spectral weight at the Fermi level. For Cu$'$0BC in hP16, $'$1 states/(eV·atom), and for Cu$'$2BC in hP21, $'$3 states/(eV·atom) (Gochitashvili et al., 18 Jul 2025). The source interprets this as a more favorable alignment of BC-$'$4 relative to $'$5 and a different balance among Cu-$'$6, Cu-$'$7, and BC-derived states, particularly when Cu is not confined entirely to dumbbell positions.

The superconducting properties are evaluated by electron–phonon coupling calculations and Eliashberg theory. For mP6-CuBC, the logarithmic average phonon frequency is $'$8 meV, the Allen–Dynes estimate gives $'$9 K for $'$0, and isotropic Migdal–Eliashberg gives $'$1 K for $'$2 (Gochitashvili et al., 18 Jul 2025). No anisotropic Migdal–Eliashberg calculation is reported for CuBC itself because the transition temperature is negligible within the framework used.

For reduced Cu content, the predicted response is markedly stronger. Cu$'$3BC has $'$4 meV with $'$5 K from Allen–Dynes and $'$6 K from isotropic Migdal–Eliashberg, while Cu$'$7BC has $'$8 meV with $'$9 K and $_x$00 K, respectively (Gochitashvili et al., 18 Jul 2025). Mixed Li–Cu phases are also superconducting in the calculations, but with lower values than the best Cu$_x$01BC ternaries: Li$_x$02Cu$_x$03BC has $_x$04 K in isotropic Migdal–Eliashberg, whereas Li$_x$05Cu$_x$06BC has $_x$07 K.

The following summary condenses the main stoichiometric and non-stoichiometric trends reported for Cu-based phases (Gochitashvili et al., 18 Jul 2025).

Phase	Electronic / phononic indicator	Superconducting estimate
CuBC (mP6)	$_x$08; $_x$09 meV	$_x$10 K (AD), $_x$11 K (iME)
Cu$_x$12BC (hP16)	$_x$13; $_x$14 meV	$_x$15 K (AD), $_x$16 K (iME)
Cu$_x$17BC (hP21)	$_x$18; $_x$19 meV	$_x$20 K (AD), $_x$21 K (iME)

These data support a consistent interpretation. Stoichiometric CuBC is a metallic end member in which the Cu-$_x$22 level is misaligned and the BC-$_x$23 manifold is nearly filled, while partially Cu-filled Cu$_x$24BC phases restore the hole-doped BC network required for MgB$_x$25-type conventional superconductivity. This suggests that Cu deficiency is not merely a perturbation but a central tuning parameter in the CuBC family.

5. Lattice dynamics and electron–phonon considerations

The lattice-dynamical discussion of CuBC is dominated by the instability of hP3-CuBC and the stabilization of mP6-CuBC (Gochitashvili et al., 18 Jul 2025). The imaginary A-point phonon in the hexagonal phase corresponds to a layer-shear mode. The source describes this qualitatively as conceptually similar to a Peierls-type or stacking instability, driven by electronic frustration and structural mismatch. Following that unstable mode removes the imaginary frequency and produces the dynamically stable monoclinic phase.

For stoichiometric CuBC, $_x$26 meV is low compared with MgB$_x$27 at $_x$28 meV, AgBC at $_x$29 meV, Cu$_x$30BC at $_x$31 meV, and Cu$_x$32BC at $_x$33 meV (Gochitashvili et al., 18 Jul 2025). The source associates the low $_x$34 with softer phonons and comparatively weaker coupling in modes that intersect the Fermi-level states, especially because BC-$_x$35 contributions at $_x$36 are minimal in stoichiometric CuBC.

Some Cu$_x$37BC phases exhibit minor dynamical instabilities, described as soft modes near $_x$38, which the source notes are common in layered materials and may be cured by temperature, anharmonic effects, or slight structural disorder (Gochitashvili et al., 18 Jul 2025). Despite these soft features, the Eliashberg spectral functions of phases with elevated $_x$39 indicate substantial electron–phonon coupling and superconducting transition temperatures up to about $_x$40–$_x$41 K in isotropic calculations.

A notable contrast is drawn with AgBC, for which full anisotropic Migdal–Eliashberg analysis predicts two-gap superconductivity with a large gap of about $_x$42 meV on BC-$_x$43-dominated ellipsoidal Fermi surfaces, a small gap below $_x$44 meV on Ag-$_x$45 and BC-$_x$46 sheets, and $_x$47 K for $_x$48 (Gochitashvili et al., 18 Jul 2025). No analogous multigap prediction is made for CuBC. The source instead states that, given the small $_x$49 and small $_x$50, CuBC is expected to be a weak, likely single-gap superconductor, if superconducting at all under experimentally accessible conditions.

6. Metastability, topochemical accessibility, and design significance

CuBC is thermodynamically metastable with respect to the elemental Cu–B–C system. For stoichiometric CuBC, the formation energy from the elements is positive, about $_x$51 eV/atom, and Cu$_x$52BC phases lie at least $_x$53 eV/atom above the global convex hull defined by Cu, C, and B$_x$54C (Gochitashvili et al., 18 Jul 2025). This rules out equilibrium synthesis from the elements as the primary route. The same source, however, emphasizes that layered BC frameworks with comparable metastability, including BC$_x$55, have been synthesized by soft chemistry or topochemical routes.

The proposed synthetic strategy is topochemical Li$_x$56Cu ion exchange starting from LiBC or Li$_x$57BC precursors (Gochitashvili et al., 18 Jul 2025):

$_x$58

For full Li$_x$59Cu exchange to CuBC, the calculated net reaction energies are strongly exothermic: about $_x$60 kJ/mol with CuCl, about $_x$61 kJ/mol with CuBr, and about $_x$62 kJ/mol with CuI. The study further states that CuBC formation is about $_x$63 kJ/mol more exothermic than AgBC formation with the same halide, and that the more moderate interlayer expansion of CuBC relative to LiBC should be kinetically easier than the much larger expansion required for AgBC (Gochitashvili et al., 18 Jul 2025).

The analysis of partially delithiated precursors indicates that many mixed Li–Cu phases lie below the tie-line connecting Li$_x$64BC and Cu$_x$65BC, making them locally stable quaternaries relative to those end members (Gochitashvili et al., 18 Jul 2025). At $_x$66 K, $_x$67 Cu-based quaternaries are locally stable by $_x$68–$_x$69 meV/atom. Vibrational entropy at $_x$70 K shifts energies upward by about $_x$71 meV/atom on average, while configuration entropy from random Li/Cu mixing contributes about $_x$72 eV/atom at $_x$73 K for typical site fractions. Even so, the source concludes that the driving force for full Li$_x$74Cu exchange is strong, about $_x$75 eV/atom per atom of product, so reactions may tend to proceed all the way to CuBC or Cu-rich Cu$_x$76BC unless kinetic barriers or controlled conditions stabilize intermediate mixed phases.

In design terms, CuBC serves as a negative-control counterpart to AgBC. The study identifies the position of the metal $_x$77 band and the interlayer spacing or dumbbell geometry as decisive tuning parameters (Gochitashvili et al., 18 Jul 2025). For high-$_x$78 borocarbides, the metal-$_x$79 band should be partially occupied at $_x$80 and should hole-dope the BC-$_x$81 $_x$82 bands. Ag satisfies this condition in the stable structure; Cu does not, because the $_x$83 band sits too high and is emptied upon structural relaxation. CuBC therefore illustrates how small changes in ionic size, dumbbell tilt, and stacking registry can separate a high-$_x$84 two-gap superconductor from a pseudogapped weakly superconducting metal.

The term “CuBC” can be ambiguous outside this materials context because similar letter sequences are also used informally for the $_x$85 meson in high-energy physics (collaboration et al., 2014). In the borocarbide literature discussed here, however, CuBC specifically denotes the copper borocarbide derivative with BC honeycomb layers bridged by C–Cu–C dumbbells (Gochitashvili et al., 18 Jul 2025).