arXiv subjects
Publications and source records attributed to Mingchen Liu.
We introduce cross-comb correlation spectroscopy, which correlates signals from asynchronous sum-frequency sampling of mid-IR incoherent light with two near-IR combs. We experimentally measure the frequency of phase-randomized mid-IR light with 100-MHz resolution using near-IR detectors.
Advanced Planning and Scheduling (APS) systems have become indispensable for modern manufacturing operations, enabling optimized resource allocation and production efficiency in increasingly complex and dynamic environments. While algorithms for solving abstracted scheduling problems have been extensively investigated, the critical prerequisite of specifying manufacturing requirements into formal constraints remains manual and labor-intensive. Although recent advances of generative models, particularly Large Language Models (LLMs), show promise in automating constraint specification from heterogeneous raw manufacturing data, their direct application faces challenges due to natural language ambiguity, non-deterministic outputs, and limited domain-specific knowledge. This paper presents a constraint-centric architecture that regulates LLMs to perform reliable automated constraint specification for production scheduling. The architecture defines a hierarchical structural space organized across three levels, implemented through domain-specific representation to ensure precision and reliability while maintaining flexibility. Furthermore, an automated production scenario adaptation algorithm is designed and deployed to efficiently customize the architecture for specific manufacturing configurations. Experimental results demonstrate that the proposed approach successfully balances the generative capabilities of LLMs with the reliability requirements of manufacturing systems, significantly outperforming pure LLM-based approaches in constraint specification tasks.
Industrial designers have long sought a natural and intuitive way to achieve the targeted control of prototype models -- using simple natural language instructions to configure and adjust the models seamlessly according to their intentions, without relying on complex modeling commands. While Large Language Models have shown promise in this area, their potential for controlling prototype models through language remains partially underutilized. This limitation stems from gaps between designers' languages and modeling languages, including mismatch in abstraction levels, fluctuation in semantic precision, and divergence in lexical scopes. To bridge these gaps, we propose an interface architecture that serves as a medium between the two languages. Grounded in design principles derived from a systematic investigation of fast prototyping practices, we devise the interface's operational mechanism and develop an algorithm for its automated domain specification. Both machine-based evaluations and human studies on fast prototyping across various product design domains demonstrate the interface's potential to function as an auxiliary module for Large Language Models, enabling precise and effective targeted control of prototype models.
Self-driving laboratories have begun to replace human experimenters in performing single experimental skills or predetermined experimental protocols. However, as the pace of idea iteration in scientific research has been intensified by Artificial Intelligence, the demand for rapid design of new protocols for new discoveries become evident. Efforts to automate protocol design have been initiated, but the capabilities of knowledge-based machine designers, such as Large Language Models, have not been fully elicited, probably for the absence of a systematic representation of experimental knowledge, as opposed to isolated, flatten pieces of information. To tackle this issue, we propose a multi-faceted, multi-scale representation, where instance actions, generalized operations, and product flow models are hierarchically encapsulated using Domain-Specific Languages. We further develop a data-driven algorithm based on non-parametric modeling that autonomously customizes these representations for specific domains. The proposed representation is equipped with various machine designers to manage protocol design tasks, including planning, modification, and adjustment. The results demonstrate that the proposed method could effectively complement Large Language Models in the protocol design process, serving as an auxiliary module in the realm of machine-assisted scientific exploration.
Traditional absorption spectroscopy has fundamental difficulty in resolving small absorbance from strong background due to the instability of laser sources. Existing background-free methods in broadband vibrational spectroscopy help to alleviate this problem but face challenges in realizing either low extinction ratios or time-resolved field measurements. Here, we introduce optical-parametric-amplification-enhanced background-free spectroscopy, in which the excitation background is first suppressed by an interferometer and then the free-induction decay that carries molecular signatures is selectively amplified. We show that this method can further improve the limit of detection in linear interferometry by order(s) of magnitude without requiring lower extinction ratios or time-resolved measurement, which can benefit sensing applications in detecting trace species.
The precision of phase estimation with interferometers can be greatly enhanced using non-classical quantum states, and the SU(11) interferometer is an elegant scheme, which generates two-mode squeezed state internally and also amplifies the signal. It has been shown in [Phys. Rev. A {\bf 95}, 063843 (2017)] that the photon-number measurement can achieve the Heisenberg limit, but only for estimating a small phase shift. We relax the constraint on the phase size by considering two adaptive protocols: one also uses the photon-number measurement with a specially tuned sequence of feedback phase; the other implements the yet-to-be-realised optimal measurement but without fine tuning.
Mode-locked lasers (MLLs) have enabled ultrafast sciences and technologies by generating ultrashort pulses with peak powers substantially exceeding their average powers. Recently, tremendous efforts have been focused on realizing integrated MLLs not only to address the challenges associated with their size and power demand, but also to enable transforming the ultrafast technologies into nanophotonic chips, and ultimately to unlock their potential for a plethora of applications. However, till now the prospect of integrated MLLs driving ultrafast nanophotonic circuits has remained elusive because of their typically low peak powers, lack of controllability, and challenges with integration with appropriate nanophotonic platforms. Here, we overcome these limitations by demonstrating an electrically-pumped actively MLL in nanophotonic lithium niobate based on its hybrid integration with a III-V semiconductor optical amplifier. Our MLL generates $\sim$4.8 ps optical pulses around 1065 nm at a repetition rate of $\sim$10 GHz, with pulse energy exceeding 2.6 pJ and a high peak power beyond 0.5 W. We show that both the repetition rate and the carrier-envelope-offset of the resulting frequency comb can be flexibly controlled in a wide range using the RF driving frequency and the pump current, paving the way for fully-stabilized on-chip frequency combs in nanophotonics. Our work marks an important step toward fully-integrated nonlinear and ultrafast photonic systems in nanophotonic lithium niobate.
Optical solitons have long been of interest both from a fundamental perspective and because of their application potential. Both cubic (Kerr) and quadratic nonlinearities can lead to soliton formation, but quadratic solitons can practically benefit from stronger nonlinearity and achieve substantial wavelength conversion. However, despite their rich physics, quadratic cavity solitons have been used only for broadband frequency comb generation, especially in the mid-IR. Here, we show that the formation dynamics of mid-IR quadratic cavity solitons can be effectively leveraged to enhance molecular sensing. We demonstrate significant sensitivity enhancement while circumventing constraints of traditional cavity enhancement mechanisms. We perform experiments sensing CO$_2$ using quadratic cavity solitons around 4 $\mu$m and achieve an equivalent path length enhancement of 6000. Additionally, we demonstrate large sensitivity at high concentrations of CO$_2$, beyond what can be achieved using an equivalent high-finesse linear cavity by orders of magnitude. Our results highlight a path for utilizing quadratic cavity nonlinear dynamics and solitons for molecular sensing beyond what can be achieved using linear methods.
Optical frequency comb is an enabling technology for a multitude of applications from metrology to ranging and communications. The tremendous progress in sources of optical frequency combs has mostly been centered around the near-infrared spectral region while many applications demand sources in the visible and mid-infrared, which have so far been challenging to achieve, especially in nanophotonics. Here, we report frequency combs tunable from visible to mid-infrared on a single chip based on ultra-widely tunable optical parametric oscillators in lithium niobate nanophotonics. Using picosecond-long pump pulses around 1 $\mu$m and tuning of the quasi-phase matching, we show sub-picosecond frequency combs tunable beyond an octave extending from 1.5 $\mu$m up to 3.3 $\mu$m with femtojoule-level thresholds. We utilize the up-conversion of the infrared combs to generate visible frequency combs reaching 620 nm on the same chip. The ultra-broadband tunability and visible-to-mid-infrared spectral coverage of our nanophotonic source can be combined with an on-chip picosecond source as its pump, as well as pulse shortening and spectral broadening mechanisms at its output, all of which are readily available in lithium niobate nanophotonics. Our results highlight a practical and universal path for the realization of efficient frequency comb sources in nanophotonics overcoming their spectral sparsity.
Powerful and efficient optical frequency combs in the mid-infrared (MIR) spectral region are highly desirable for a broad range of applications. Despite extensive efforts utilizing various techniques, MIR frequency comb sources are still lacking power, efficiency, or bandwidth for many applications. Here, we report the generation of an intrinsically locked frequency comb source centered at 4.18 {\mu}m from an optical parametric oscillator (OPO) operating in the simulton regime, in which formation of purely quadratic solitons lead to enhanced performance. We show advantages of operation in the simulton regime in direct experimental comparisons to the conventional regime, which are also supported by simulation and theory. We achieve 565 mW of average power, 900 nm of instantaneous 3-dB bandwidth, 350% of slope efficiency, and 44% of conversion efficiency; a performance that is superior to previous OPO demonstrations and other sources in this wavelength range. This work opens a new avenue toward MIR frequency comb generation with high power and efficiency and suggests the great potential of soliton generation based on quadratic nonlinearity in the MIR spectral region.
Dual-comb spectroscopy has been proven a powerful tool in molecular characterization, which remains challenging to implement in the mid-infrared region due to difficulties in the realization of two mutually locked comb sources and efficient photodetection. Moreover, the detection capability of dual-comb spectroscopy is fundamentally limited by the strong excitation background and detector saturation. Here we introduce a variant of dual-comb spectroscopy called cross-comb spectroscopy, in which a mid-infrared comb is upconverted via sum-frequency generation with a near-infrared comb of a shifted repetition rate and then interfered with a spectral extension of the near-infrared comb. We show that cross-comb spectroscopy can have superior signal-to-noise ratio, sensitivity, dynamic range, and detection efficiency compared to other dual-comb-based methods and avoid the limits of the background excitation and detector saturation. We experimentally demonstrate a proof-of-concept measurement of atmospheric CO2 around 4.25 $\mu$m, with a 233-$cm^{-1}$ instantaneous bandwidth, 28000 comb lines, a single-shot SNR of 167, and a figure of merit of $2.4\times10^{6} Hz^{1/2}$. Cross-comb spectroscopy can be realized using up- or down-conversion and offers an adaptable and powerful spectroscopic method outside the well-developed near-IR region. This approach opens new avenues to high-performance molecular sensing with wavelength flexibility, which can impact a wide swath of applications.