🔬 optics

Highly-sensitive mid-infrared upconversion detection based on external-cavity pump enhancement

This paper presents a highly sensitive mid-infrared upconversion detection system that utilizes a low-loss external cavity to enhance a 1064 nm pump by a factor of 36, achieving a 22% conversion efficiency and a record-low noise equivalent power of 0.3 fW/Hz $^{1/2}$ for high-precision spectroscopy in photon-starved scenarios.

Original authors: Xiaohan Liu, Kun Huang, Wen Zhang, Ben Sun, Jianan Fang, Yan Liang, Heping Zeng

Published 2026-05-29

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Xiaohan Liu, Kun Huang, Wen Zhang, Ben Sun, Jianan Fang, Yan Liang, Heping Zeng

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to listen to a very faint whisper (a mid-infrared light signal) in a noisy, crowded room. Normally, the equipment we use to "hear" these whispers is either too clumsy, too slow, or needs to be kept in a freezer to work properly. This paper describes a clever new way to amplify that whisper so clearly that even the quietest one can be heard, all while keeping the equipment at room temperature.

Here is how the researchers did it, broken down into simple concepts:

1. The Problem: The "Faint Whisper"

Mid-infrared light is special because it carries the "fingerprints" of many molecules (like pollutants or specific chemicals). However, detecting this light is hard.

Old detectors are like trying to hear a whisper with a broken microphone: they are noisy and slow.
Super-sensitive detectors exist, but they are like expensive, high-tech microphones that only work if you keep them in a deep-freeze (cryogenic cooling), which is impractical for many uses.

2. The Solution: The "Translator" and the "Echo Chamber"

The team built a system that acts like a translator and an echo chamber combined.

The Translator (Upconversion): Instead of trying to listen to the faint infrared whisper directly, they use a special crystal to "translate" it into a different language: visible light (specifically, a color near infrared that our eyes or standard cameras can see easily). It's like taking a whisper in a foreign language and instantly translating it into clear, loud English.
The Echo Chamber (External Cavity): To make the translation efficient, they needed a very strong "voice" (a pump laser) to do the talking. But a standard laser wasn't strong enough. So, they built an optical "echo chamber" (a cavity made of mirrors).
- They shot the laser light into this chamber.
- The mirrors bounced the light back and forth thousands of times, building up the energy inside like sound building up in a concert hall.
- This boosted the laser power by 36 times without needing a massive, energy-hungry laser source.

3. The Result: Hearing the Unhearable

Because they amplified the "translator's" voice so effectively, the system became incredibly sensitive.

The Sensitivity: They achieved a level of sensitivity called a "Noise Equivalent Power" of 0.3 fW/Hz¹/². To put that in perspective, the paper claims this is ten times better than previous similar systems. It's like being able to hear a pin drop from a mile away.
The Efficiency: They managed to convert about 22% of the incoming infrared light into the new visible light, which is a very high success rate for this type of technology.
The "Dynamic Range" Upgrade: They also tested two types of "ears" (detectors) to hear the translated light. One was a single ear (a standard sensor), and the other was a "multi-pixel" ear (a sensor with thousands of tiny ears). The multi-pixel version could handle a much wider range of loudness, from a whisper to a shout, without getting overwhelmed. This improved the system's ability to handle different signal strengths by 15 to 30 decibels.

4. Why This Matters (According to the Paper)

The paper highlights that this setup is particularly good for:

Room Temperature Operation: No need for freezers.
High Precision: Because they used a very pure, single-frequency laser (like a perfect, steady tone), the system can create very accurate "maps" of what the light is made of. This is great for identifying specific molecules.
Photon-Starved Scenarios: It works even when there are very few photons (particles of light) available, such as in long-distance sensing or when you need to be very gentle with the sample (to avoid damaging it).

In short, the researchers built a room-temperature "super-microphone" for infrared light. They used a mirror-based echo chamber to boost the power of their signal, allowing them to detect incredibly faint infrared whispers with unprecedented clarity and speed.

Technical Summary: Highly-sensitive mid-infrared upconversion detection based on external-cavity pump enhancement

Problem Statement
Sensitive mid-infrared (MIR) detection is critical for applications ranging from remote sensing and environmental monitoring to biomedical diagnosis. However, achieving high-performance passive MIR sensing remains challenging. Commercial semiconductor-based detectors (e.g., HgCdTe, InSb) typically suffer from high dark noise and low response speeds, resulting in noise equivalent power (NEP) levels around pW/Hz $^{1/2}$ . While frequency upconversion offers a pathway to room-temperature, high-sensitivity detection by transferring MIR signals to the visible/near-infrared regime, existing continuous-wave (CW) schemes face a trade-off: single-pass configurations require high pump powers (tens of watts) for efficient conversion, while intra-cavity enhancement schemes often suffer from increased noise due to spectral overlap between the pump and the upconverted signal, limiting NEP to the tens of fW/Hz $^{1/2}$ range.

Methodology
The authors propose and implement a MIR upconversion detection system utilizing external-cavity pump enhancement to overcome the limitations of high-power requirements and noise. The experimental setup involves three main stages:

MIR Signal Generation: A single-frequency Er-doped fiber laser (1550 nm) and an Yb-doped fiber laser (1064 nm) are amplified and combined to drive difference-frequency generation (DFG) in a 50-mm periodically poled lithium niobate (PPLN1) crystal, producing a tunable CW MIR signal at 3.4 µm.
External-Cavity Pump Enhancement: The 1064 nm pump beam is injected into a bowtie-type standing-wave optical cavity with a finesse of approximately 109. The cavity is constructed with four mirrors (two flat, two curved) and houses a 25-mm PPLN2 crystal for sum-frequency generation (SFG). A digital locking system based on a Field Programmable Gate Array (FPGA) stabilizes the cavity length via a piezoelectric actuator (PZT) using a dither-locking technique. This configuration decouples the pump source from the enhancement cavity, allowing the use of compact, low-noise single-frequency fiber lasers.
Detection and Filtering: The upconverted signal (810 nm) is extracted, filtered through a multi-stage spectral filter (notch, long-pass, short-pass, and band-pass filters) to suppress the 1064 nm pump by ~190 dB, and coupled into a single-mode fiber. Detection is performed using either a single-photon counting module (SPCM) or a multi-pixel photon counter (MPPC) to extend the dynamic range.

Key Contributions and Results

Pump Enhancement: The external cavity successfully enhances the intra-cavity pump power by a factor of 36. With an input pump power of 1.53 W, the circulating power reaches approximately 55 W.
Conversion Efficiency: At the peak intra-cavity power of 55 W, the system achieves a peak conversion efficiency of 22%. The total detection efficiency (including filtering, fiber coupling, and detector quantum efficiency) is approximately 6.8%.
Noise Performance: The system achieves a Noise Equivalent Power (NEP) as low as 0.3 fW/Hz $^{1/2}$ . This represents a ten-fold improvement over previously reported values for CW-pumped MIR upconversion detectors (which were typically in the pW/Hz $^{1/2}$ range) and is significantly lower than intra-cavity schemes that suffer from pump-signal spectral overlap.
Dynamic Range: By employing a multi-pixel photon counter (MPPC) with photon-number-resolving capabilities, the detection dynamic range is extended to 30 dB, a 15 dB improvement over the 15 dB range of a single-pixel SPCM.
Stability: The FPGA-based digital locking unit enables automatic stabilization and fast re-locking capabilities, ensuring long-term operation with pump power fluctuations estimated at only ~1%.

Significance and Claims
The paper claims that this external-cavity enhancement scheme effectively addresses the stringent power requirements of passive MIR sensing while maintaining ultra-low noise. By separating the pump source from the enhancement cavity, the system avoids the spectral filtering difficulties inherent in intra-cavity schemes, thereby achieving a background noise level that allows for a NEP of 0.3 fW/Hz $^{1/2}$ .

The authors highlight that the use of single-frequency pumping is particularly advantageous for high-fidelity spectral mapping, making the system suitable for high-precision MIR upconversion spectroscopy in "photon-starved" scenarios. The architecture is presented as a versatile platform that can be adapted for imaging and extended to longer wavelengths (up to 12 µm) by utilizing different nonlinear crystals (e.g., BaGa $_4$ Se $_7$ or AgGaS $_2$ ). The work demonstrates that combining high-performance silicon photon counters with external-cavity pump enhancement can realize highly sensitive, room-temperature MIR detection suitable for trace analysis and remote sensing.

1. The Problem: The "Faint Whisper"

2. The Solution: The "Translator" and the "Echo Chamber"

3. The Result: Hearing the Unhearable

4. Why This Matters (According to the Paper)

Technical Summary: Highly-sensitive mid-infrared upconversion detection based on external-cavity pump enhancement

More like this