The nature of dark matter is one of the most pressing mysteries in fundamental physics. Among the particles proposed as dark matter constituents, ultralight photon-coupled bosons stand as promising candidates, including axions, axion-like particles (ALPs), and dark photons. In the hypothesis that these particles constitute the galactic dark matter halo, they can generate coherent electromagnetic fields oscillating at frequencies determined by their masses. In the microwave regime, these fields can be detected through resonant interactions with highly sensitive superconducting quantum devices. In particular, when a transmon qubit is resonant with the dark matter signal, it exhibits slow Rabi oscillations between its ground and excited state. The excitation probability after a fixed interaction time encodes information about the signal strength. Frequency sweeps of the qubit's resonant frequency enable the search for resonances with dark matter, covering a wide range of hypothetical dark particle masses. We introduce a quantum signal amplification protocol that post-processes the state of the sensing qubit after it interacts with dark matter via a programmable, gate-based enhancement circuit. The circuit we developed dramatically improves the detector's response to small dark matter signals. This enhancement leverages an auxiliary ancilla qubit, requiring its measurement and conditional post-selection on the outcome. Successful post-selection leads to a greatly improved sensitivity to the weak signal. A significant practical advantage over other quantum enhancement schemes existing in the literature, based on preparing fully-entangled multi-qubit states and letting them evolve under dark matter, is that our enhancement circuit is applied only after the sensing qubit is exposed to dark matter interaction, avoiding the requirement of maintaining multi-qubit entanglement for long sensing periods. Moreover, the protocol requires only one two-qubit gate, ensuring compatibility with modern quantum hardware limitations. Comprehensive experiment simulations, including realistic noise sources, were conducted, focusing on dark photon detection. Our results demonstrate that readout fidelity and state preparation accuracy heavily affect the sensitivity to dark matter. Under optimal (yet achievable) error conditions, a two-qubit enhanced detector achieves an integration time speedup factor of ~1.05 - 2.75 in the frequency range 2.5 GHz - 6.0 GHz, relative to the baseline Rabi-sampling approach. Scaling to multi-sensor architectures with shared ancilla for sequential enhancement increases the speedup to ~2.10 - 5.50. Sensitivity projections deploying 120 physical qubits for three years of data-taking yield an exclusion limit on the dark photon kinetic mixing of ε ≈ 1×10^(-14) at the 95% confidence level, uniformly across the frequency range 2.5 GHz - 6.0 GHz. Such projections substantially exceed both cosmological bounds and current experimental limits. The experimental side of this thesis involves the design, fabrication, and characterization of transmon-based architectures towards a dark photon sensor compatible with the enhanced sensing protocol. Design and simulation methodologies, addressing Hamiltonian engineering and decoherence mitigation, were developed and validated experimentally. Three test devices with increasing complexity, ranging from uncoupled qubits to qubits coupled dispersively through bus resonators, were characterized. Measured parameters confirm our predictions, though Purcell-limited relaxation emerged as the primary coherence bottleneck due to lower-than-expected resonator quality factors. This work demonstrates that programmable superconducting quantum processors represent a compelling platform for quantum-enhanced dark matter detection, establishing a bridge between quantum computing and fundamental physics.
Determinare la natura della materia oscura rappresenta uno dei più importanti problemi irrisolti della fisica fondamentale. Tra le particelle candidate come costituenti, vi sono bosoni ultraleggeri che ammettono un accoppiamento con il fotone ordinario, come assioni, axion-like particles (ALPs), e fotoni oscuri. Nell’ipotesi in cui tali particelle costituiscano l’alone oscuro, genererebbero un campo elettromagnetico oscillante ad una frequenza corrispondente alla loro massa. Nel regime delle microonde, questo segnale può essere rivelato utilizzando sensori quantistici superconduttivi. In particolare, quando un transmon qubit è in risonanza con il segnale prodotto dalla materia oscura, quest’ultimo subisce un’oscillazione di Rabi tra lo stato fondamentale e quello eccitato. La probabilità di eccitazione del qubit, trascorso un certo intervallo di tempo, codifica informazione sull’intensità del segnale. Variando la frequenza del qubit, dunque, è possibile effettuare una ricerca su ampio spettro di picchi nella probabilità di eccitazione, rivelando la presenza di campi elettromagnetici compatibili con quelli generati dalla materia oscura. Questa tesi introduce un protocollo di amplificazione del segnale, trasformando lo stato raggiunto dal qubit a causa dell’interazione con la materia oscura mediante un approccio basato su quantum gates. Quest’ultimo prevede l’utilizzo di un qubit ausiliario, il quale viene misurato, a cui segue un’operazione di post-selection sul risultato. Quando la post-selection ha successo, lo stato del qubit sensore acquista una notevole sensibilità al segnale proveniente dalla materia oscura. Un vantaggio considerevole di questo approccio rispetto ad altri schemi di amplificazione del segnale è il fatto che quest’ultimo si applica dopo l’interazione tra qubit e materia oscura, senza la difficoltà di preparare stati entangled e mantenerli a lungo. Inoltre, il circuito di amplificazione utilizza un solo gate a due qubit, rendendolo di fatto compatibile con le limitazioni degli hardware quantistici moderni. Sviluppando simulazioni numeriche che includono meccanismi di decoerenza e di rumore, quantifichiamo come l’accuratezza sulla lettura e sulla preparazione dello stato iniziale dei qubit influisca sulla sensibilità dell’esperimento al segnale proveniente dal fotone oscuro. Per un rivelatore soggetto a condizioni di rumore ottimali, simili allo stato dell’arte, l’utilizzo del protocollo permette di ridurre di un fattore ~1.05 - 2.75 (nell'intervallo 2.5 GHz - 6.0 GHz) il tempo di presa dati necessario per raggiungere lo stesso limite di esclusione sul parametro di mixing cinetico ε raggiungibile da un esperimento standard. Estendendo il protocollo a strutture con più qubit sensori accoppiati allo stesso qubit ausiliario, il guadagno sale fino a ~2.10 - 5.50. Stimiamo inoltre che utilizzando 120 qubit in tre anni di presa dati sia possibile raggiungere un limite di esclusione di ε ≈ 1×10^(-14) (95% C.L), uniformemente nell’intervallo 2.5 GHz - 6.0 GHz, superando sensibilmente i limiti attuali imposti da osservazioni cosmologiche e da esperimenti basati su haloscopi. Dal lato sperimentale, la tesi descrive strategie di design, simulazione e misure di dispositivi basati su transmon qubit finalizzati a realizzare un prototipo di detector adatto ad implementare lo schema di rivelazione. Sono stati fabbricati e misurati tre dispositivi, tra qubit disaccoppiati e qubit accoppiati dispersivamente. I parametri misurati risultano essere in accordo con le nostre previsioni, fatta eccezione per tempi di coerenza limitati a causa del ridotto fattore di qualità dei risonatori di lettura. Questo lavoro di tesi, posto all’intersezione fra quantum computing e fisica fondamentale, dimostra che dispositivi quantistici superconduttivi oggi disponibili possono implementare algoritmi per incrementare la sensibilità di rivelazione della materia oscura.
Moretti, R (2026). Gate-Based Enhancement of Superconducting Qubit Sensitivity to Ultralight Bosonic Dark Matter. (Tesi di dottorato, , 2026).
Gate-Based Enhancement of Superconducting Qubit Sensitivity to Ultralight Bosonic Dark Matter
MORETTI, ROBERTO
2026
Abstract
The nature of dark matter is one of the most pressing mysteries in fundamental physics. Among the particles proposed as dark matter constituents, ultralight photon-coupled bosons stand as promising candidates, including axions, axion-like particles (ALPs), and dark photons. In the hypothesis that these particles constitute the galactic dark matter halo, they can generate coherent electromagnetic fields oscillating at frequencies determined by their masses. In the microwave regime, these fields can be detected through resonant interactions with highly sensitive superconducting quantum devices. In particular, when a transmon qubit is resonant with the dark matter signal, it exhibits slow Rabi oscillations between its ground and excited state. The excitation probability after a fixed interaction time encodes information about the signal strength. Frequency sweeps of the qubit's resonant frequency enable the search for resonances with dark matter, covering a wide range of hypothetical dark particle masses. We introduce a quantum signal amplification protocol that post-processes the state of the sensing qubit after it interacts with dark matter via a programmable, gate-based enhancement circuit. The circuit we developed dramatically improves the detector's response to small dark matter signals. This enhancement leverages an auxiliary ancilla qubit, requiring its measurement and conditional post-selection on the outcome. Successful post-selection leads to a greatly improved sensitivity to the weak signal. A significant practical advantage over other quantum enhancement schemes existing in the literature, based on preparing fully-entangled multi-qubit states and letting them evolve under dark matter, is that our enhancement circuit is applied only after the sensing qubit is exposed to dark matter interaction, avoiding the requirement of maintaining multi-qubit entanglement for long sensing periods. Moreover, the protocol requires only one two-qubit gate, ensuring compatibility with modern quantum hardware limitations. Comprehensive experiment simulations, including realistic noise sources, were conducted, focusing on dark photon detection. Our results demonstrate that readout fidelity and state preparation accuracy heavily affect the sensitivity to dark matter. Under optimal (yet achievable) error conditions, a two-qubit enhanced detector achieves an integration time speedup factor of ~1.05 - 2.75 in the frequency range 2.5 GHz - 6.0 GHz, relative to the baseline Rabi-sampling approach. Scaling to multi-sensor architectures with shared ancilla for sequential enhancement increases the speedup to ~2.10 - 5.50. Sensitivity projections deploying 120 physical qubits for three years of data-taking yield an exclusion limit on the dark photon kinetic mixing of ε ≈ 1×10^(-14) at the 95% confidence level, uniformly across the frequency range 2.5 GHz - 6.0 GHz. Such projections substantially exceed both cosmological bounds and current experimental limits. The experimental side of this thesis involves the design, fabrication, and characterization of transmon-based architectures towards a dark photon sensor compatible with the enhanced sensing protocol. Design and simulation methodologies, addressing Hamiltonian engineering and decoherence mitigation, were developed and validated experimentally. Three test devices with increasing complexity, ranging from uncoupled qubits to qubits coupled dispersively through bus resonators, were characterized. Measured parameters confirm our predictions, though Purcell-limited relaxation emerged as the primary coherence bottleneck due to lower-than-expected resonator quality factors. This work demonstrates that programmable superconducting quantum processors represent a compelling platform for quantum-enhanced dark matter detection, establishing a bridge between quantum computing and fundamental physics.| File | Dimensione | Formato | |
|---|---|---|---|
|
phd_unimib_825617.pdf
accesso aperto
Descrizione: Gate-Based Enhancement of Superconducting Qubit Sensitivity to Ultralight Bosonic Dark Matter
Tipologia di allegato:
Doctoral thesis
Dimensione
11.71 MB
Formato
Adobe PDF
|
11.71 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.


