Quantum Tunneling: State of the Art (2024–2025)

A research note for physicists and engineers approaching the field

1. What Is Quantum Tunneling?

Quantum tunneling is the phenomenon by which a particle crosses a potential energy barrier that it classically lacks sufficient energy to surmount — a direct consequence of the wave nature of quantum states and the nonzero probability amplitude that extends into classically forbidden regions. The transmission probability decays exponentially with barrier width and height, a relationship captured at leading order by the WKB approximation. Far from a curiosity, tunneling is a load-bearing mechanism in nature: it enables nuclear fusion in stars, determines enzyme reaction rates in biology, and is the operating principle of the transistors that may define the next generation of computing. The central tension in modern research is that tunneling is simultaneously an engineered resource (qubits, tunnel diodes, STM) and a fundamental nuisance (leakage currents, decoherence, gate-oxide breakdown).

2. SoTA Overview — The Four Frontiers

3. Domain Deep-Dives

3.1 Quantum Computing & Superconducting Qubits

The central enabling mechanism in superconducting quantum computing is Josephson junction tunneling: Cooper pairs tunnel coherently across a thin insulating barrier, producing a nonlinear inductance that gives transmon-style qubits their anharmonicity — the property that separates the |0⟩–|1⟩ transition from higher levels and makes selective qubit control possible. Tunneling is thus not a side effect but the engine of the qubit.

Key Breakthroughs (2024–2025)

• Flux-tunable superconducting qubits (UC Santa Barbara, 2024): coherence time of 1.5 ms — current record for superconducting qubits, achieved through improved materials and junction fabrication
• 3D Transmon architectures: coherence routinely >100 μs; cavity protection suppresses radiative loss
• Oxford tunneling qubits: >10 μs coherence
• Quantum dot systems: ~1.4 μs coherence in 5-qubit arrays; relevant for semiconductor integration
• Surface code error correction deployment (2024): Google, IBM, and Microsoft all deployed surface code protocols on real hardware — transition from proof-of-principle to engineered error suppression
• Topological qubit research (UC Berkeley, 2024): Majorana-based approaches active, promising inherent protection via non-Abelian anyons

Metrics Snapshot

Open Problems

• Scalability: extending to 1,000+ high-quality qubits while maintaining cross-chip connectivity
• Qubit crosstalk: parasitic tunneling couplings between neighboring qubits limit addressability
• Leakage: tunneling out of the |0⟩/|1⟩ subspace into higher transmon levels — not captured by standard qubit error models
• Josephson junction materials: two-level system (TLS) defects at the oxide interface remain the dominant decoherence source; microscopic origin not fully understood
• Error correction overhead: ~1,000 physical qubits per logical qubit means millions of physical qubits needed for practical fault-tolerant computation

Key insight: Tunneling is the mechanism (Josephson effect → qubit nonlinearity) and simultaneously the challenge (leakage, crosstalk, TLS noise). Progress is inseparable from deep materials-level control of tunneling.

3.2 Nanoscale Devices & Scanning Tunneling Microscopy

Tunnel FETs (TFETs)

TFETs exploit band-to-band tunneling at a reverse-biased p-n junction, replacing thermionic emission (the Boltzmann-limited mechanism that sets the 60 mV/decade floor in MOSFETs). TFETs can break this floor:

STM & Atomic-Scale Manipulation

• Sub-ångström lateral resolution, sub-meV energy resolution in tunneling spectroscopy
• High-speed STM: up to 100 Hz imaging; ~10 Hz for atomic-scale manipulation
• ESR-STM (Choi et al., Nanoscale Advances 7, 4551, 2025; arXiv:2505.10079): STM tunneling current drives electron spin resonance on individual atoms — atomic-scale spin sensing
• 2D heterostructures: MoS₂, WS₂, graphene interlayer tunneling active research area

Scaling Limits

3.3 Biological Quantum Tunneling

Evidence for tunneling in biology spans a wide credibility range:

Theoretical methods driving 2024 progress: QM/MM simulations (quantum subsystem up to ~200 atoms), instanton methods, ring polymer molecular dynamics (RPMD) — increasingly identifying tunneling contributions previously overlooked.

Key research groups: Klinman (UC Berkeley), Benkovic (Penn State), Engel (U Chicago), Hammes-Schiffer (UIUC)

3.4 Theory & Decoherence: Open Problems

The Tunneling Time Problem

One of the most genuinely unresolved foundational questions in quantum mechanics:

• Core question: How long does a particle spend inside a classically forbidden barrier? Is "tunneling time" even a well-defined observable?
• The Hartman effect: Phase time (group delay) saturates and becomes independent of barrier width for thick barriers — apparently implying superluminal traversal. No causal violation occurs, but the physical meaning is disputed.
• Competing definitions (all give different answers): phase time, dwell time, Larmor precession time, Wigner delay time, complex time. No experiment has cleanly discriminated between them.
• Experimental landmarks: Ramos et al. (Nature, 2020) used atomic spin precession in a magnetic barrier; Sainadh et al. (Nature, 2019) used attosecond streaking on atomic tunnel ionization. Both set constraints, neither resolved the debate.
• Status (2025): No consensus. The question of whether tunneling time is a measurable observable in the standard quantum mechanical sense remains open.

Many-Body & Macroscopic Tunneling

• Macroscopic quantum tunneling (MQT): Well-established in SQUIDs and Josephson junctions (Martinis, Devoret groups). Flux in a SQUID tunnels between flux states — same physics as qubit |0⟩↔|1⟩ under the barrier.
• Instanton methods: Saddle-point paths in imaginary time provide systematic semiclassical tunneling rate expansions; applied to vacuum decay, Floquet systems (Takayoshi & Oka, JPSJ 94, 111003, 2025; arXiv:2509.03674).
• Caldeira-Leggett model (Ann. Phys. 149, 374, 1983): Foundational framework for tunneling in dissipative environments; extensions to non-Markovian baths active.
• Decoherence from complex saddle points (Nishimura & Watanabe, PRL, 2025; arXiv:2408.16627): Connects instanton paths to decoherence rates — bridges tunneling dynamics and open-system physics.

Other Active Theoretical Directions

4. Cross-Cutting Themes & Tensions

1. Decoherence as the universal enemy Tunneling requires quantum coherence; the environment destroys it. This is the central tension in every domain: qubits need long coherence times but are embedded in macroscopic devices; biological tunneling occurs in warm, wet, noisy cells; TFET operation is classical by the time a signal reaches the circuit. The Caldeira-Leggett framework unifies these challenges — all four frontiers are fundamentally fighting the same battle against environmental dephasing.

2. Tunneling as resource vs. nuisance The same physical effect is engineered into devices (Josephson junctions → qubits; band-to-band tunneling → TFETs; elastic tunneling → STM current) and simultaneously constitutes the primary failure mode (gate-oxide leakage in CMOS; qubit leakage in transmons). Progress in each domain requires finely tuned control over the same tunneling parameters.

3. The scalability gap Every domain faces a version of the same scaling problem:

• QC: ~1,000 physical qubits per logical qubit; millions needed for practical fault-tolerant computation
• TFETs: drive current gap vs. CMOS not yet closed; no manufacturable process at scale
• STM: 10–100 atoms/hour manipulation throughput — not a viable manufacturing path
• Biological: QM/MM simulations max out at ~200 quantum atoms; full-protein quantum treatment remains intractable

4. Theory lags experiment in the time domain Attosecond experiments on tunneling ionization are now technically feasible (Sainadh 2019, Ramos 2020), but theory cannot yet provide an unambiguous, operationally defined tunneling time that all approaches agree on. Measurement is ahead of interpretation.

5. Mathematical convergence across domains Instanton methods (QFT), WKB (single-particle), RPMD (chemistry), and Floquet theory (driven systems) are converging on a common language for tunneling across energy and length scales. The 2025 papers by Nishimura & Watanabe and Takayoshi & Oka exemplify this trend toward unified treatment.

5. Reading Path: From Zero to SoTA

Interactive Reading Graph

Hover any node for a paper summary, HPC/QPU access notes, and a direct link. Edges show recommended reading order; dashed edges are skip-tier shortcuts.

Tier 1 — Foundations (Undergraduate)

Tier 2 — Core Theory (Graduate)

Tier 3 — Domain Reviews

Tier 4 — Cutting-Edge Papers (2023–2026)

Free Online Resources

Suggested Study Order

Note: Read Caldeira & Leggett (1983) at the start of Phase 4 regardless of domain — it is the shared theoretical vocabulary of all four frontiers.

6. Confidence Assessment