Polarization dynamics of a diamond Raman laser governed by local stress and pump polarization perturbation
Paper details
Section titled “Paper details”| Field | Value |
|---|---|
| Journal | Optics Express |
| Published | 2026-05-06 |
| Authors | Wanxiao Gao, Hui Chen, Yunpeng Cai, Aokun Shi, jiaming Zhang |
| DOI | 10.1364/oe.601904 |
| Citations | 1 |
The preferred crystal regime for high-power, high-efficiency Raman lasing had birefringence intensity below 5 × 10⁻⁶ and a fast axis aligned with [111] in this study. The authors mapped five local regions and measured their response to linear, elliptical, and circular pumping. High stress caused abrupt Stokes polarization switching. Low stress with a favorable axis supported Raman gain but allowed stimulated Brillouin scattering to redirect the polarization mode and cap power scaling. These results place crystal screening and pump-polarization control on the same level as cavity design.
Residual stress mapping identifies local birefringence conditions
Section titled “Residual stress mapping identifies local birefringence conditions”Division-of-focal-plane polarization imaging measured phase retardation and fast-axis orientation on the diamond end face. The map showed low birefringence near the center and high birefringence near the periphery. The authors selected five regions with different stress levels and axis orientations.
| Region | Birefringence intensity | Fast-axis direction |
|---|---|---|
| A | 3.14 × 10⁻⁶ | -65° |
| B | 9.4286 × 10⁻⁶ | -88° |
| C | 8.57 × 10⁻⁶ | 2° |
| D | 4.29 × 10⁻⁶ | -80° |
| E | 1.43 × 10⁻⁶ | -85° |
Fast-axis directions use the crystallographic reference direction defined in the paper. Region B had about three times the birefringence intensity of region A. Region C had similar stress but a fast axis nearly perpendicular to the Raman maximum-gain direction. Equal stress therefore did not produce equal polarization loss or mode selection.
Pump, cavity, and measurement conditions
Section titled “Pump, cavity, and measurement conditions”The pump was a 1064 nm quasi-continuous-wave laser operating at 100 Hz with 165 µs pulses. Average power was 4.2 W during the reported polarization comparisons. A 100 mm lens focused the beam to a 42 µm waist at the diamond center. The 128 mm cavity used a 50 mm input coupler and a 75 mm output coupler. The input coupler transmitted at 1064 nm and reflected at 1240 nm. The output coupler reflected at 1064 nm and transmitted 1% at 1240 nm. The gain medium was a 2 × 2 × 7 mm³ diamond with end faces anti-reflection coated at 1064 and 1240 nm. A translation stage positioned each interaction region with 0.01 mm accuracy. A quarter-wave plate set the pump ellipticity, and a half-wave plate rotated its direction. Manual wave plate adjustment had an angular uncertainty of ±1°. The authors measured the Stokes angle, degree of linear polarization, degree of polarization, effective gain, and threshold. Repeated measurements under identical conditions showed good repeatability.
Linear pumping exposes the stress dependence
Section titled “Linear pumping exposes the stress dependence”An ideal Raman tensor predicts a smooth Stokes rotation with pump angle. The maximum-gain direction is [111], while orthogonal Raman modes become degenerate near the paper’s reference direction. Region A followed this behavior under linear pumping because its phase retardation was too weak to break the gain symmetry.
Regions B and C instead produced polarization jumps because local birefringence competed with anisotropic Raman gain. The jump angles differed because B’s fast axis approached [111], while C’s was nearly perpendicular to it. Regions D and E combined low stress with favorable axes, but SBS entered their mode competition. Under linear pumping, the first-order spectra also showed whether SBS was present.
Elliptical pump perturbations cause stepwise mode switching
Section titled “Elliptical pump perturbations cause stepwise mode switching”The paper defines pump ellipticity as the minor-to-major axis ratio of the polarization ellipse. A value of 0.1 was used in the elliptical-pump measurements. Earlier tensor simulations predicted that weak ellipticity would reduce gain, raise threshold, and break equal-gain boundaries. The experiment confirmed those changes and produced stepwise Stokes evolution in all five regions.
Region A changed from a smooth linear-pump response to stepwise switching under elliptical pumping. In regions B and C, abrupt Stokes flipping began after the pump rotated about 20° to 30°. Region B switched across Stokes angles from 20° to 70° over selected pump intervals. Region C mainly switched between 60° and 140°. These results show that birefringence remained dominant in the high-stress regions.
The degree of linear polarization stayed close to unity across the measured pump-angle range. However, the minimum degree of polarization in regions B and C fell to about 0.4 when rotation reduced effective Raman gain. Region E retained a high degree of polarization because its residual stress was low. Elliptical pumping also shifted the SBS onset angles in regions D and E.
SBS redirects polarization in low-stress high-gain regions
Section titled “SBS redirects polarization in low-stress high-gain regions”Regions D and E combined low birefringence with fast axes close to [111]. This geometry supports Raman gain but can raise intracavity intensity above the SBS threshold. SBS depleted the favored Raman mode, so the laser switched to a mode that SBS suppressed less. The spectrum from region E consistently included a clear SBS component, unlike region B under either pumping condition.
The authors linked this result to an earlier study in which SBS disappeared after a ±1 mm cavity-length change in a cavity on the order of 10 cm. In regions D and E, SBS limited gain amplitude and raised the threshold baseline. High-stress regions had a different loss, because large phase retardation reduced output power and increased threshold. This distinction identifies birefringence and SBS as separate power-scaling bottlenecks.
Circular pumping locks high-stress axes and lowers gain
Section titled “Circular pumping locks high-stress axes and lowers gain”Circular pumping supplied equal orthogonal components with a π/2 phase difference. In high-stress region B, left- and right-handed pumps both locked the Stokes angle at 62°. Region C locked near 70° for right-handed pumping and 74° for left-handed pumping. These fixed angles show that birefringence determined the output under strong stress. In low-stress regions D and E, Raman tensor anisotropy selected a favored polarization direction.
Table 2 shows the right-handed subset of the paper’s circular-pump results.
| Region | Stokes angle | Gain (W) | Threshold (W) | DOP |
|---|---|---|---|---|
| A | 84° | 1.204 | 1.495 | 0.7928 |
| B | 62° | 1.256 | 1.15 | 0.778 |
| C | 70° | 1.104 | 1.516 | 0.6141 |
| D | 130° | 1.342 | 1.05 | 0.625 |
| E | 132° | 1.084 | 1.19 | 0.8639 |
For right-handed pumping, gain ranged from 1.084 W in region E to 1.342 W in region D. Thresholds ranged from 0.98 W in region D to 1.516 W in region C. Region C had the lowest degree of polarization, 0.6141, while region E reached 0.8639. The authors report that circular pumping generally reduced effective gain and increased threshold compared with optimal linear pumping. The lower projection of distributed pump energy onto [111] explains that penalty.
Crystal selection must control stress, axis, and SBS
Section titled “Crystal selection must control stress, axis, and SBS”Crystal selection must balance birefringence loss against SBS onset. The authors prefer birefringence intensity below 5 × 10⁻⁶ and a fast axis well aligned with [111]. Regions D and E fall within that stress criterion and have favorable Raman axes. Yet both can trigger SBS, which suppresses further power gain. If SBS cannot be controlled, the authors suggest moderately low stress with a nonoptimal fast axis. Such regions may offer a better balance between conversion efficiency and operational stability.
For device engineering, the measured maps support screening candidate crystals for phase retardation and fast-axis orientation. The results also support limiting pump ellipticity and controlling SBS through cavity design.
The evidence covers one 2 × 2 × 7 mm³ crystal, five regions, and a 4.2 W average, 100 Hz pump with 165 µs pulses. The study does not establish continuous-wave high-power scaling, long-term threshold stability, or performance across other crystals. No cavity-length sweep was reported. The underlying data were unavailable publicly at publication and could be requested from the authors. The measured stress limit and axis requirement are selection criteria, but each cavity needs separate SBS validation.
Materials and support
Section titled “Materials and support”6CCVD’s diamond materials catalog includes single-crystal diamond grown by MPCVD, matching the anisotropic Raman gain crystal required in this study. The study needs a 2 × 2 × 7 mm³ crystal with low birefringence, a defined fast-axis orientation, and anti-reflection coatings at 1064 and 1240 nm. Questions about matching that specification can be directed to engineering support for custom fabrication support.
View Original Abstract
Stress in diamond crystals is a critical factor affecting Raman gain and the polarization state of the output Stokes. This study systematically investigates the multi-field coupling among residual-stress-induced birefringence in diamond, initial pump ellipticity, and stimulated Brillouin scattering (SBS). By combining theoretical simulations, division-of-focal-plane polarization imaging, and Raman laser experiments, we elucidate the mechanisms responsible for the stepwise evolution of the Stokes polarization direction. In high-stress regions, birefringence-induced phase distortion dominates the polarization dynamics, leading to abrupt polarization switching. In low-stress regions, SBS acts as a parasitic loss channel that forces polarization-mode transitions and limits power scaling. The results further confirm that even slight pump polarization perturbations can degrade the Raman output polarization state, reduce the effective gain, and increase the laser threshold. These theoretical and experimental findings provide practical guidance for overcoming performance bottlenecks in diamond Raman lasers, screening high-quality crystals, and developing efficient coherent light sources with precisely controllable polarization states.
Single Crystal Cvd Diamond Plate
View Specs →
Boron Doped Single Crystal Diamond Plate
View Specs →
Customized Single Crystal Cvd Diamond Plate
View Specs →References we cover
Section titled “References we cover”- 213.4 W Continuous-Wave Diamond Raman Laser at 1240 nm with Polarization-Combined Pumping
- High-power free-running single-longitudinal-mode diamond Raman laser enabled by suppressing parasitic stimulated Brillouin scattering
- Widely-tunable single-frequency diamond Raman laser
- High Power Diamond Raman Lasers