Giacomo Corrielli

To assess the capabilities of photonic quantum systems, it is essential to develop analytical and experimental methods for probing the underlying optical dynamics. Here, we investigate fundamental constraints arising when multi-photon states evolve through a quantum linear optical network, showing how such limitations can be overcome when introducing measurement-induced nonlinearities. Specifically, we analyze bounds related to photon-bunching behavior at the output of a linear interferometer and test them via a hybrid photonic platform realizing experiments with up to four photons. When the analysis is extended to a nonlinear regime, such bounds can be violated. We experimentally validate such behavior by employing an optical architecture featuring measurement-induced nonlinearity, thus introducing dynamics that go beyond standard linear optics. Our results highlight how the constraints implied by the linear optical formalism can be used as tools to test and characterize photonic models going beyond quantum linear optical dynamics.

Femtosecond Laser Micromachining (FLM) is a powerful technology for the fabrication of photonic devices. In this context, the integration of resonant elements within the platform represents a key advancement, enhancing both its versatility and its compatibility with a wide range of optical and fluidic components specifically enabled by this technique. Here, we report the realization of a tunable racetrack resonator fabricated by FLM and operating at telecom wavelengths. Leveraging low-loss waveguides, we obtained a Q factor of the resonator as high as 8 × 105 at critical coupling. Moreover, by integrating two thermo-optic phase shifters, we achieved both resonance tuning and dynamic control of the Q factor. This capability makes the device highly versatile for applications requiring dynamic spectral control, such as tunable filters, gyroscopes, and sensors.

Over the past few years, various methods have been developed to engineer and to exploit the dynamics of photonic quantum states as they evolve through linear optical networks. Recent theoretical works have shown that the underlying Lie algebraic structure plays a crucial role in the description of linear optical Hamiltonians, as such formalism identifies intrinsic symmetries within photonic systems subject to linear optical dynamics. Here, we experimentally investigate the role of Lie algebras applied to the context of Boson sampling, a pivotal model to the current understanding of computational complexity regimes in photonic quantum information. Performing experiments of increasing complexity, realized within a fully reconfigurable photonic circuit, we show that sampling experiments do indeed fulfill the constraints implied by a Lie algebraic structure. In addition, we provide a comprehensive picture about how the concept of Lie algebraic invariant can be interpreted from the point of view of n th -order correlation functions in quantum optics. Our work shows how Lie algebraic invariants can be used as a benchmark tool for the correctness of an underlying linear optical dynamics within typical interferometric architectures for photonic quantum information. This opens avenues for the use of algebraic-inspired methods as analysis tools for photon-based quantum computing protocols.

To date, most integrated quantum photonics experiments rely on single-photon detectors operating at cryogenic temperatures coupled to photonic integrated circuits (PICs) through single-mode optical fibers. This approach presents significant challenges due to the detection complexity, as cryogenic conditions hinder the development of scalable systems. In addition, going towards fully-integrated devices or, at least, removing the optical fibers would be also advantageous to develop compact and cost-efficient solutions featuring a high number of optical modes. This work reports on the direct coupling of a PIC, fabricated by femtosecond laser writing (FLW), and a silicon single-photon avalanche diode (SPAD) array, fabricated in a custom planar technology and compatible with the operation at room temperature. The effectiveness of this solution is shown by achieving perfect coupling and a system detection efficiency as high as 41.0% at a wavelength of 561 nm, which is the highest value reported to date among both heterogeneous/hybrid integrated and directly coupled systems. We also show the robustness of the coupling to misalignments, demonstrating that costly alignment procedures are not needed. Finally, we exploit the SPAD array to characterize a reconfigurable Mach-Zehnder interferometer, i.e., the basic building block of multimode reconfigurable PICs. This solution provides a new avenue to the design and implementation of quantum photonics experiments, especially effective when compact and cost-efficient systems are needed.

The implementation of large-scale universal quantum computation represents a challenging and ambitious task on the road to quantum processing of information. In recent years, an intermediate approach has been pursued to demonstrate quantum computational advantage via non-universal computational models. A relevant example for photonic platforms has been provided by the Boson Sampling paradigm and its variants, which are known to be computationally hard while requiring at the same time only the manipulation of the generated photonic resources via linear optics and detection. Beside quantum computational advantage demonstrations, a promising direction towards possibly useful applications can be found in the field of quantum machine learning, considering the currently almost unexplored intermediate scenario between non-adaptive linear optics and universal photonic quantum computation. Here, we report the experimental implementation of quantum machine learning protocols by adding adaptivity via post-selection to a Boson Sampling platform based on universal programmable photonic circuits fabricated via femtosecond laser writing. Our experimental results demonstrate that Adaptive Boson Sampling is a viable route towards dimension-enhanced quantum machine learning with linear optical devices.

We demonstrate a sensitive optically pumped magnetometer using rubidium vapor and 0.75 amg of nitrogen buffer gas in a sub-mm-width sensing channel excavated by femtosecond laser writing followed by chemical etching. The channel is buried less than 1 mm below the surface of its fused silica host material, which also includes reservoir chambers and micro-strainer connections, to preserve a clean optical environment. Using a zero-field-resonance magnetometry strategy and a sensing volume of 2.25 mm3, we demonstrate a sensitivity of ≈1pT/Hz at 10 Hz. The device can be integrated with photonic structures and microfluidic channels with 3D versatility. Its sensitivity, bandwidth, and stand-off distance will enable detection of localized fields from magnetic nanoparticles and μL NMR samples.

Integrated photonics offers compact and stable manipulation of optical signals in miniaturized chips, with the possibility of changing dynamically their functionality by means of integrated phase shifters. Cryogenic operation of these devices is becoming essential for advancing photonic quantum technologies, accommodating components like quantum light sources, single photon detectors and quantum memories operating at liquid helium temperatures. In this work, we report on a programmable glass photonic integrated circuit (PIC) fabricated through femtosecond laser waveguide writing (FLW) and controlled by thermo-optic phase shifters both in a room-temperature and in a cryogenic setting. By taking advantage of a femtosecond laser microstructuring process, we achieved reliable PIC operation with minimal power consumption and confined temperature gradients in both conditions. This advancement marks the first cryogenically-compatible programmable FLW PIC, paving the way for fully integrated quantum architectures realized on a laser-written photonic chip.

Optical comb lasers could increase data transmission rates in telecommunications.