Curated map of Zoo entries in the Super-Semi family.
Entries
| Entry | Type | Status |
|---|---|---|
| all-semiconductor-superconducting-qubit | qubit | proposed |
| andreev-pair-qubit | qubit | demonstrated |
| andreev-spin-qubit | qubit | demonstrated |
| ferbo-qubit | qubit | proposed |
| gatemon | qubit | demonstrated |
| gatemonium | qubit | demonstrated |
Composition
- qubit: 6
Conceptual anchors
- josephson-junction-as-nonlinear-element is the shared superconducting backbone, even when the weak link is semiconductor-defined rather than oxide-barrier-defined.
- spin-orbit-coupling-for-qubit-control is the key separator between purely gate-tunable Josephson devices and spin-active Andreev hybrids.
- charge-noise-sweet-spot is the right lens for comparing how much electrical tunability each architecture gains without paying too much dephasing cost.
- noise-bias-and-asymmetric-error-channels is the extra routing note for the protection-first branch, where the claim is not just “higher coherence” but a deliberately skewed error channel.
Curated synthesis
The cleanest way to read this family is to ask where the semiconductor is doing the conceptual work. The same InAs/Al or group-IV vocabulary can mean three very different things.
-
Semiconductor as a tunable Josephson knob
- gatemon is the minimal case: transmon logic survives, but the weak link makes electrically tunable.
- gatemonium is the protected-circuit extension of the same move: the weak link still tunes the Josephson element, but now inside a fluxonium-style inductive circuit where gate control moves the device between different protection regimes.
- Read this branch primarily through josephson-junction-as-nonlinear-element and charge-noise-sweet-spot, not through spin language.
-
Semiconductor as part of the qubit subspace itself
- andreev-pair-qubit uses the even-parity empty/doubly occupied Andreev states; its microwave transition is controlled by superconducting phase and weak-link transmission.
- andreev-spin-qubit uses spin-split Andreev levels as the qubit degree of freedom, so spin-orbit physics and parity stability are the central story rather than side constraints.
- ferbo-qubit pushes further into the protection-first regime: the Andreev sector is part of the mechanism that suppresses relaxation before error correction, which is why it belongs adjacent to noise-bias-and-asymmetric-error-channels as well as to superconducting protected-circuit notes.
- This branch should feel closer to “engineered weak-link quantum states” than to ordinary tunable transmons.
-
Semiconductor as the full materials stack
- all-semiconductor-superconducting-qubit is the fabrication-endpoint branch. Its main claim is not a new control Hamiltonian but the possibility of moving the entire superconducting circuit, junctions included, inside one doped crystal.
- Compare it to gatemon or transmon on interface elimination and process integration, and to kane-qubit on atomic-precision fabrication lineage.
Semiconductor-superconductor routing table
| Entry | Keep it in super-semi-moc when the main claim is… | Hand off when the real question becomes… |
|---|---|---|
| gatemon | electrostatic tuning of an otherwise transmon-like Josephson element | ordinary circuit-QED scaling or baseline superconducting benchmarking in superconducting-moc |
| gatemonium | a gate-tunable weak link reshaping which fluxonium regime the same circuit can access | protected-circuit comparison against fluxonium, heavy-fluxonium-qubit, or 0-pi-qubit inside superconducting-moc |
| andreev-spin-qubit | odd-parity Andreev states and spin-dependent supercurrent are the qubit resource | nonlocal parity encoding or a topological phase claim in topological-moc |
| andreev-pair-qubit | even-parity Andreev pair states and phase-dependent weak-link spectroscopy are the resource | ordinary Josephson-circuit scaling in superconducting-moc |
| ferbo-qubit | even-parity Andreev structure is being used as a relaxation-bias / protection mechanism inside a fluxonium-like circuit | the comparison becomes purely protected-circuit taxonomy in superconducting-moc or topological parity protection in topological-moc |
| all-semiconductor-superconducting-qubit | monolithic crystal growth and interface elimination are the conceptual novelty | donor-style atomic fabrication lineage in semiconducting-moc or ordinary superconducting-circuit performance in superconducting-moc |
Weak-link participation ladder
The semiconductor weak link can enter the qubit at three different depths: as a tunable circuit coefficient, as the microscopic two-level system, or as one half of a deliberately hybridized protected mode. That distinction is more useful than grouping every device by material stack alone.
| Branch | Where the computational state primarily lives | Role of Andreev occupation / parity | Main control and readout layer | Failure mode that must stay visible |
|---|---|---|---|---|
| gatemon | The collective charge-phase mode of a transmon-like circuit | The weak-link spectrum is folded into while the junction is assumed to remain on its even-parity ground branch; parity is not the encoding | Gate voltage tunes the junction; ordinary circuit-QED drive and dispersive readout operate the qubit | Gate noise, hysteresis, and junction-intrinsic loss can erase the benefit of voltage tunability; accidental weak-link occupation appears as instability rather than a logical operation |
| gatemonium | Fluxonium-like plasmon / fluxon modes set by , , external flux, and the gate-tuned weak-link potential | Andreev channel transparencies reshape the nonlinear potential, but the demonstrated device does not encode information in a chosen quasiparticle-parity manifold | Gate control of is combined with flux bias and circuit-QED spectroscopy / control | Array inductive loss, nonsinusoidal weak-link calibration, and gate-noise susceptibility must be separated from the ideal fluxonium sweet-spot story |
| andreev-pair-qubit | The two even-parity states of one Andreev doublet | The logical transition changes pair occupation while remaining in the even sector; a parity switch leaves that computational manifold | Phase-dependent microwave and cavity coupling address the pair transition | Quasiparticle trapping, parity switching, and relaxation out of the selected doublet are first-class state-lifecycle errors |
| andreev-spin-qubit | The spin-split odd-parity states of one trapped quasiparticle | A fixed odd-parity sector is the prerequisite for the spin qubit; spin-dependent supercurrent supplies the readout signal | Spin-orbit-enabled control plus dispersive sensing of the junction response | Quasiparticle poisoning changes the parity sector, while spin relaxation and single-shot-readout fidelity remain separate bottlenecks |
| ferbo-qubit | A proposed hybrid of one bosonic fluxonium-like mode and an even-parity Andreev pseudospin | Disjoint support in the Andreev degree of freedom is intended to suppress selected bosonic relaxation matrix elements; the fermionic sector is part of the protection mechanism | High-transmission weak-link engineering and a zero-flux operating point define the protected regime | The predicted bias must still survive experimental state preparation, parity stability, high-impedance fabrication, and readout |
Use this ladder to avoid calling every fixed-parity condition “parity protection.” In andreev-pair-qubit and andreev-spin-qubit, parity stability is primarily a prerequisite that keeps the device inside its chosen computational sector. In ferbo-qubit, structure within the even-parity Andreev space is proposed to suppress a particular relaxation matrix element, which is a stronger but still theory-stage claim. Read that final row through noise-bias-and-asymmetric-error-channels rather than treating it as generic immunity, and use divincenzo-criteria to keep preparation and readout evidence separate from Hamiltonian protection.
Routing rule inside the family
- Start with gatemon or gatemonium when the interesting question is what electrostatic control does to an otherwise familiar superconducting circuit.
- Start with andreev-spin-qubit or ferbo-qubit when the weak link’s internal spin, parity, or Andreev structure is the qubit resource or the protection mechanism.
- Start with all-semiconductor-superconducting-qubit when the value proposition is monolithic fabrication or interface elimination rather than a mature coherence result.
- For the non-hybrid dot and donor branch of the same broader materials ecosystem, continue in semiconducting-moc.
Boundary with the superconducting family
This family shares a lot of vocabulary with superconducting-moc, especially around Josephson elements, circuit-QED readout, and sweet-spot language, but the routing rule is simple: stay here only when the semiconductor weak link is doing conceptual work. gatemon and gatemonium belong here because gate-tunable junction physics is the point, and andreev-spin-qubit / ferbo-qubit belong here because Andreev-level and parity structure enter the qubit itself. If the main story is instead protected-circuit geometry, bosonic encoding, or ordinary superconducting coherence scaling, route back to superconducting-moc even when the fabrication stack overlaps.
Boundary with the topological family
This family shares hardware vocabulary with topological-moc such as InAs/Al heterostructures, Josephson weak links, and strong spin-orbit coupling, but it makes a different promise. gatemon, andreev-spin-qubit, and ferbo-qubit stay here because their value comes from tunability, Andreev structure, or circuit-level protection inside an otherwise conventional control stack. Once the central claim becomes nonlocal parity encoding or a phase-biased topological channel that is supposed to suppress errors before code overhead, route the reader to topological-moc instead.
Scope boundary
This facet is about hybridization, not mere material substitution: the semiconductor must change the circuit physics, control surface, or protection story.