Scientists are pursuing fast, high-fidelity state transfer which is fundamental to scalable integrated photonics and quantum information processing. Adiabatic evolution offers inherent robustness in this field.
Evolution-Pause Synthesis enables ultrafast high-fidelity photonic state transfer
An 11.8-fold acceleration in state transfer has been demonstrated compared with traditional adiabatic processes, overcoming limits found in earlier techniques struggling to maintain fidelity during rapid transitions within photonic circuits. Conventional methods previously struggled because of a trade-off between evolution time and accuracy, requiring either slow driving speeds or complex Hamiltonian engineering. Evolution-pause synthesis (EPS) introduces a new shortcut protocol manipulating transient excitations through strategically inserted pauses into the active evolutionary process.
We validated accelerated state transfer using a device footprint reduced by nearly tenfold; specifically, we achieved high-fidelity transfers within $16μm compared to a traditional adiabatic coupler spanning 150μm. Extending beyond two-level systems, EPS successfully applied to a complex seven-level stimulated Raman adiabatic passage (STIRAP) process demonstrated an 11.76-fold acceleration in reaching over 99% fidelity without altering how the system changes or adding extra components. Carefully timed interruptions effectively cancel accumulated amplitudes from those transient states decoupling phase accumulation from parameter variations allowing for faster and more reliable information processing.
Interference Engineered Pauses Enable Rapid Adiabatic State Transfer
Fast coherent control across wave and quantum platforms is receiving increasing attention; adiabatic evolution offers a natural solution for complete state transfer but suffers slow evolution times limiting compactness, integration density, and scalability in photonic implementations where temporal evolution maps onto spatial propagation along z. Researchers have developed acceleration strategies as shortcuts to adiabaticity, yet often rely on non-native auxiliary couplings or delicate Hamiltonian engineering that can be difficult to implement practically. EPS achieves fast near-perfect state transfer strictly within the native system Hamiltonian by treating transient excitations as coherent resources cancelling their accumulated amplitudes via strategically interleaved pauses. This decoupling of relative dynamical phase accumulation from parameter variations steers open transition trajectories into a closed loop in complex amplitude space enabling perfect state transfer without auxiliary fields or complex detours.
The underlying mechanism was first elucidated using Landau, Zener dynamics and then extended it to more general state transfer processes demonstrating substantial shortening of evolution time compared with linear adiabatic process in a seven-level system. Experimental validation on a silicon photonic platform realises high-fidelity state transfer in a 16μm footprint representing a nearly tenfold reduction in device length when contrasted with the 150μm adiabatic reference. EPS offers a general hardware-compatible framework for fast practical coherent control across wave and quantum platforms; precise manipulation of coherent states is key for integrated photonics, optical information processing and programmable wave-based devices.
Complete state transfer, strong mode conversion and high-fidelity state preparation serve as fundamental building blocks for scalable photonic circuits, but conventional adiabatic evolution intrinsically provides insensitivity to moderate fabrication or control imperfections at the price of slow operation. Various acceleration strategies have been developed including shortcuts such as counterdiabatic driving and invariant-based inverse engineering which can reproduce adiabaticlike transfer in shorter times.
However, these approaches are usually Hamiltonian-centric accelerating evolution by introducing auxiliary interactions or designing delicate parameter trajectories, requirements often incompatible with realistic integrated platforms where non-native couplings may be inaccessible and complex paths difficult to fabricate precisely potentially increasing vulnerability to parameter noise.
This motivates a shift from Hamiltonian-centric control towards using the accumulated effects of nonadiabatic transitions. Rather than suppressing transient excitations, they can be used as coherent resources whose interference closes the accumulation trajectory at the final time.
Achieving this closure requires independent control over relative phases lacking in conventional protocols because phase accumulation remains tied to the parameter path; EPS provides such phase control through strategically placed pauses along the native Hamiltonian path preserving the geometric path optimising only its temporal parametrization by interleaving active segments with pauses during which the Hamiltonian is held fixed requiring no auxiliary fields or complicated detours in parameter space.
Perfect terminal transfer requires complex closure of the accumulated amplitudes in the target eigenstate.
The research demonstrated fast and accurate movement of quantum states using a technique called evolution-pause synthesis on silicon photonics. This method achieves state transfer within existing hardware limitations by utilising temporary excitations as resources and cancelling their effects with carefully timed pauses during operation.
The process accelerated state transfer by a factor of 11.8 compared to standard methods while maintaining high fidelity within a footprint of 16μm$. Researchers suggest this approach offers robustness against imperfections commonly found in integrated photonic circuits, potentially simplifying fabrication requirements.
👉 More information
🗞 Interference-engineered shortcut to perfect state transfer
✍️ Yichuan Zhang, Xuanyu Liu, Zemeng Lin, Wange Song and Shuang Zhang
🧠 ArXiv: https://arxiv.org/abs/2608.18794



See today’s quantum computing news on Quantum Zeitgeist for the latest breakthroughs in qubits, hardware, algorithms, and industry deals.
