01 Introduction
The research team led by Researcher Kun Huang and Professor Heping Zeng from the State Key Laboratory of Precision Spectroscopy, East China Normal University, has made significant progress in the field of mid-infrared single-photon measurement and control. They proposed a mid-infrared single-photon temporal ghost imaging technique based on nonlinear wavelength conversion, achieving high-precision reconstruction of mid-infrared single-photon temporal waveforms. This advancement provides strong support for fields such as quantum spectroscopy, single-molecule physics, and mid-infrared communications. The relevant results were recently published in Laser & Photonics Reviews under the title “Mid-Infrared Single-Photon Computational Temporal Ghost Imaging” (Figure 1). East China Normal University is the first completing institution, with graduate student Wen Zhang as the first author and Researcher Kun Huang and Professor Heping Zeng as corresponding authors.
Figure 1: Latest findings by the East China Normal University research team published in Laser & Photonics Reviews.
02 Research Background
High-resolution and high-sensitivity mid-infrared temporal detection and analysis technologies are essential tools for analyzing ultrafast dynamic processes and achieving precision measurements. They are widely applied in areas such as remote infrared atmospheric sensing, deep-space free-space communication, and single-molecule ultrafast dynamics. However, conventional mid-infrared detectors still fall short in terms of sensitivity, response speed, and operational temperature. There is an urgent need for new mid-infrared measurement and control methods that combine high temporal resolution and sensitivity under room temperature. In recent years, temporal ghost imaging technology has attracted extensive attention. By exploiting temporal correlations of the optical field, it enables the reconstruction of high-speed temporal signals using low-speed detectors, relaxing stringent requirements on detector bandwidth. Currently, most temporal ghost imaging studies operate in the visible/near-infrared spectrum, and extending this technique into broader electromagnetic bands, especially the mid-infrared region, is crucial to meet broader application needs. However, mid-infrared spectral regions face dual technical challenges: the lack of high-bandwidth and sensitive detectors, and immaturity of high-fidelity optical modulation. To date, no study has demonstrated mid-infrared single-photon-level temporal ghost imaging, highlighting the urgent need for high-sensitivity photon detection and precise temporal encoding techniques.
Figure 2: Conceptual diagram of mid-infrared single-photon computational temporal ghost imaging.
03 Innovative Research
To address these challenges, the research team proposed a mid-infrared temporal ghost imaging method based on nonlinear wavelength conversion. By using nonlinear optical difference and sum frequency generation processes, the team overcame limitations in high-precision modulation and sensitive detection within the mid-infrared band. They successfully used low-bandwidth silicon-based detectors to capture high-speed mid-infrared waveforms and demonstrated high-resolution reconstruction of mid-infrared single-photon waveforms beyond the temporal jitter limits of conventional detectors.
Specifically, high-speed temporal encoding was performed on near-infrared continuous-wave light via telecom-band intensity modulators. This temporal encoding was then transferred with high fidelity to the mid-infrared band through frequency down-conversion. The achieved temporal resolution was 80 ps, corresponding to an analysis bandwidth of 12.5 GHz, significantly exceeding the response bandwidth of existing commercial mid-infrared detectors. The encoded mid-infrared signals passed through the temporal target and were intensity-integrated by low-bandwidth detectors. Hadamard-encoded sequences were used, and their corresponding integrated intensity values were collected. Although the so-called ‘bucket detector’ lacks temporal resolution, the temporal information of the target can still be reconstructed by correlating the known encoding patterns with the detected signals, thereby realizing mid-infrared computational temporal ghost imaging.
Figure 3: Setup of mid-infrared single-photon computational temporal ghost imaging system.
To overcome sensitivity limitations of current mid-infrared detectors, the researchers employed external-cavity-pumped frequency upconversion detection to efficiently and low-noise convert mid-infrared signals to the visible range, allowing the use of high-performance silicon single-photon detectors. In experiments, the encoded mid-infrared pulse sequence was attenuated to 0.1 photon/bit. Despite such low light levels, accurate reconstruction of high-SNR mid-infrared temporal waveforms was achieved using photon accumulation within the coding window, thus demonstrating computational temporal ghost imaging at the single-photon level for the first time.
Notably, this method exceeds the temporal resolution limits imposed by jitter in traditional single-photon detection schemes. Combined with compressed sensing algorithms, this approach can reduce data acquisition time by over 90% under under-sampling conditions, greatly enhancing analysis efficiency in low-light environments.
Figure 4: Experimental results of mid-infrared single-photon computational temporal ghost imaging. (a) High-resolution mid-infrared single-photon temporal waveform; (b) Under-sampled waveform reconstruction using compressed sensing.
04 Summary and Outlook
By integrating nonlinear wavelength conversion and computational temporal ghost imaging, this work achieves, for the first time, high-speed mid-infrared temporal waveform reconstruction at single-photon sensitivity. It breaks through the limitations of traditional detector bandwidth and temporal jitter and provides a novel room-temperature method for high-precision measurement and control in ultrafast dynamics and deep-space communication.
In the future, the system's temporal resolution can be further improved by leveraging faster near-infrared modulators and incorporating ultrashort optical pulse shaping techniques. This could potentially achieve femtosecond-level temporal resolution, providing robust support for temporal waveform characterization at mid-infrared and even longer wavelengths.
In recent years, the team has achieved a series of innovative advances in nonlinear single-photon mid-infrared measurement and control, including mid-infrared nonlinear Fourier transform imaging [Optica 11, 1716 (2024)], low-threshold mid-infrared fiber parametric oscillator [Photon. Res. 12, 2123 (2024)], ultra-sensitive mid-infrared single-photon detection [Adv. Photon. Nexus 3, 046002 (2024); Photonics Res. 12, 1294 (2024)], broadband mid-infrared single-pixel spectroscopy [Laser Photon. Rev. 18, 2301272 (2024); Laser Photon. Rev. 18, 2401099 (2024)], and high-frame-rate mid-infrared hyperspectral imaging [Nat. Commun. 15, 1811 (2024)]. These achievements have been supported by the Ministry of Science and Technology, National Natural Science Foundation of China, Shanghai Science and Technology Commission, Chongqing Science and Technology Bureau, and East China Normal University.
Article Information: Wen Zhang, Kun Huang, Xu Wang, Ben Sun, Jianan Fang, Yijing Li, and Heping Zeng, “Mid-Infrared Single-Photon Computational Temporal Ghost Imaging,” Laser & Photonics Reviews, 19, 2402180 (2025).
DOI Link: Mid-Infrared Single-Photon Computational Temporal Ghost Imaging
