🔬 materials science

Generation and spectroscopic characterization of continuous-wave 148.4-nm light in strontium tetraborate

This paper reports the generation of continuous-wave 148.4-nm light via random quasi-phase matching in strontium tetraborate and its spectroscopic characterization using a custom prism spectrometer, demonstrating temperature-dependent power output up to 5.7 nW to guide the optimization of solid-state $^{229}$ Th nuclear clocks.

Original authors: Ming Guan, Yang Wang, Heng Tu, Kaifgeng Zhou, Zhengrong Xiao, Zhong Zuo, Songpo Xu, Shaogang Yu, Xin Tong, Linqiang Hua, Xiaojun Liu

Published 2026-11-02

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Ming Guan, Yang Wang, Heng Tu, Kaifgeng Zhou, Zhengrong Xiao, Zhong Zuo, Songpo Xu, Shaogang Yu, Xin Tong, Linqiang Hua, Xiaojun Liu

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

Deep inside the atom, the nucleus usually behaves like a heavy, sluggish anchor, largely ignoring the frantic activity of the electrons swirling around it. But in the rare isotope of thorium known as thorium-229, the nucleus holds a secret: it possesses a tiny, low-energy state that can be nudged by light. This transition happens at a specific color of light that is invisible to the human eye, sitting in the vacuum ultraviolet range. Because this nuclear jump is so isolated from the messy environment of the atom's outer shell, it promises to be the most stable, precise tick imaginable. If scientists can build a clock that counts these nuclear ticks, it would be far more accurate than today's best atomic clocks, potentially revolutionizing how we measure time, navigate the globe, and test the fundamental laws of physics. The catch is that the light needed to trigger this nuclear jump is extremely difficult to produce and even harder to measure, as it is easily swallowed by the air around us and requires specialized crystals to generate.

A team of researchers in China has now taken a significant step toward solving this problem by creating a steady beam of this elusive light and building a custom tool to study it. They focused on generating light at a wavelength of 148.4 nanometers, the exact frequency required to drive the thorium-229 transition. To do this, they used a crystal made of strontium tetraborate, a material that does not naturally align its internal structure to convert light efficiently. Instead of forcing the crystal to align perfectly, the researchers relied on the random, chaotic arrangement of tiny regions inside the crystal to do the work. They shined a powerful beam of ultraviolet light, at 296.8 nanometers, into the crystal. As this light passed through the disordered internal structure, a small fraction of it was converted into the shorter, higher-energy 148.4-nanometer wavelength they needed. This process, known as random quasi-phase matching, allowed them to generate a continuous stream of the target light without needing the complex, perfect alignment usually required for such tasks.

To understand exactly what they had created, the team built a compact spectrometer, a device designed to separate different colors of light and measure their intensity. Because the target light is so easily absorbed by oxygen in the air, the entire optical path was sealed inside a chamber filled with dry nitrogen. They used a prism made of calcium fluoride, a material transparent to this deep ultraviolet light, to bend the beam. Since the target light bends more sharply than the original pump light, the prism spread them out into two distinct lines. A sensitive detector, shielded from the bright pump light by mirrors and a narrow slit, scanned across the angle where the new light appeared. This setup allowed the researchers to confirm that the light they generated was indeed the correct wavelength and to measure its power with high precision. They found that the new light was separated from the pump light by a clear margin, proving that their method successfully isolated the desired signal from the overwhelming background noise.

The measurements revealed that the amount of light produced depends heavily on how the crystal is treated. When the researchers varied the power of the input laser, the output of the new light increased in a predictable, quadratic fashion, confirming that the process was working as a true frequency doubler. However, the most surprising discovery came when they looked at the effect of temperature. They expected the crystal to produce the most light at one specific, smooth temperature peak. Instead, they found a jagged landscape of peaks and valleys. As they warmed the crystal from room temperature up to about 52 degrees Celsius, the output of the light rose and fell multiple times, with distinct maximums appearing at 34, 42, and 47 degrees, separated by sharp drops just a few degrees away. This means that the crystal is not a single, uniform converter but rather a collection of many tiny, independent converters, each responding to slightly different temperatures. If the temperature drifts by just a couple of degrees, the output can plummet from a maximum to a minimum.

The researchers also discovered that the crystal is fragile under prolonged exposure. When they ran the experiment in a vacuum for several hours, the crystal slowly turned yellow-brown, a sign that the intense light was damaging its structure and creating defects that absorbed the light. This "coloring" effect reduced the amount of usable light over time. By switching to a nitrogen atmosphere, they protected the crystal from this rapid degradation, allowing for more stable operation. In their best conditions, with the crystal held at its optimal temperature, the source produced about 5.7 nanowatts of power at 148.4 nanometers. While this amount of light is incredibly faint—far too dim to be seen by the human eye—it is sufficient for the delicate task of probing the thorium nucleus. The study concludes that for a future nuclear clock to work, the light source cannot simply be kept at a constant temperature; it must be locked precisely onto one of those specific temperature peaks, and the crystal must be protected from the damaging effects of prolonged irradiation. These findings provide the practical roadmap needed to stabilize and optimize the light sources required to bring the world's most precise clock to life.

Technical Summary: Generation and Spectroscopic Characterization of Continuous-Wave 148.4-nm Light in Strontium Tetraborate

Problem Statement
The $^{229}$ Th nucleus possesses a low-lying isomeric state at 8.4 eV (148.4 nm), representing the only nuclear transition accessible to table-top laser technology. Realizing a solid-state nuclear clock or an ion clock based on this transition requires narrow-line, continuous-wave (CW) radiation at 148.4 nm with controlled power, spectral purity, and stability. The most direct source for this radiation is the second-harmonic generation (SHG) of 296.8-nm light. However, characterizing such sources is experimentally challenging due to the strong absorption of 148.4-nm light by molecular oxygen (requiring vacuum or inert gas environments), the low quantum efficiency of solar-blind detectors, and the overwhelming brightness of the residual pump light (typically 8–11 orders of magnitude stronger than the harmonic). Furthermore, the performance of solid-state sources for clock applications depends critically on long-term stability, spatial uniformity, and temperature dependence over hours of operation, parameters that have not been comprehensively characterized for 148.4-nm generation in strontium tetraborate (SBO).

Methodology
The authors generated CW 148.4-nm light by frequency-doubling 296.8-nm pump light in a strontium tetraborate (SrB $_4$ O $_7$ , SBO) crystal. Since SBO cannot be birefringently phase-matched in this range, the experiment utilized random quasi-phase matching (RQPM) provided by the crystal's antiparallel ferroelectric domains. The SBO crystal was grown via the Czochralski method, cut as a 1 mm thick $a$ -cut plate, and used without coatings.

To characterize the output, the authors constructed a compact prism spectrometer designed specifically for this wavelength range. The optical layout included:

Beam Separation: A CaF $_2$ prism (60° apex angle) dispersed the beam, separating the 148.4-nm harmonic from the 296.8-nm pump based on their different deflection angles.
Pump Rejection: A pair of dichroic mirrors (DCM) and a baffle tube attenuated the residual pump by approximately six orders of magnitude before dispersion.
Detection: A solar-blind CsI photomultiplier tube (PMT) mounted on a rotation stage recorded the intensity as a function of angle ( $\theta_{PMT}$ ).
Environment: The entire optical path was enclosed in a sealed chamber that was evacuated and back-filled with dry nitrogen to prevent oxygen absorption of the VUV light.
Calibration: The wavelength scale was calibrated using a deuterium lamp and a reference VUV spectrometer, mapping the PMT angle to wavelength via a fourth-order polynomial derived from Sellmeier data for CaF $_2$ .

Key Contributions and Results

Spectral Characterization and Pump Rejection: The spectrometer successfully resolved the 148.4-nm harmonic and the 296.8-nm pump, which were separated by 11.2°. The system achieved a spectral resolution of $\delta\lambda \approx 0.6$ nm (limited by beam divergence rather than the slit width). The pump line was approximately two orders of magnitude brighter than the harmonic, yet the background contribution at the harmonic's angle was limited to less than 7%.
Absolute Power Measurement: By integrating the photon counts across the spectral line and accounting for the total detection efficiency (approximately 0.9%, including crystal transmission, optics, and PMT quantum efficiency), the authors determined the absolute power. At room temperature with ~1 W of pump power, the source delivered 1.9 nW at 148.4 nm.
Temperature Dependence: A critical finding was the complex temperature dependence of the conversion efficiency. Rather than a single smooth optimum, the conversion efficiency exhibited a multi-peaked structure between 30°C and 50°C. The signal displayed several pronounced maxima (e.g., near 34°C, 42°C, and 47°C) separated by minima only a few degrees apart. This behavior suggests that different groups of randomly oriented domains satisfy the quasi-phase-matching condition at slightly different temperatures. Consequently, the output power is highly sensitive to small temperature excursions, and stabilization requires holding the crystal on a specific maximum rather than merely maintaining a constant temperature.
Spatial and Temporal Stability: The RQPM signal was found to be spatially non-uniform, concentrated in a region roughly 1 mm across. Long-term irradiation in vacuum caused the crystal to develop yellow-brown color centers, reducing the output. However, operating in a nitrogen atmosphere mitigated this effect and aided heat dissipation.
Optimized Performance: By tuning the crystal to its optimal temperature, the output power increased by a factor of three, reaching 5.7 nW.

Significance
The paper provides practical guidance for the optimization and stabilization of 148.4-nm sources intended for solid-state $^{229}$ Th nuclear clocks. The authors demonstrate that while the source delivers power levels (1.9–5.7 nW) relevant for absorption spectroscopy of the isomer in a host crystal, the operational parameters are non-trivial. Specifically, the multi-peaked temperature response and spatial inhomogeneity of the RQPM process mean that the crystal position and temperature must be actively controlled and stabilized on specific local maxima. The developed compact prism spectrometer offers a robust method for characterizing VUV sources, separating harmonic and pump light without complex filter stacks, and establishing absolute power calibration. These results pave the way for the future development of high-performance nuclear optical clocks by defining the necessary constraints for source stability and characterization.

More like this