On 26 October, a team associated with the 11th Research Institute of China Electronics Technology Group Corporation reported in Applied Physics B (open access, CC BY 4.0) a method combining an integrating sphere with a multipass cell: a 1064 nm beam passes seven times through each sample, cutting the measurement uncertainty for low-loss materials to one seventh of the single-pass value. Comparing pure YAG and Yb:YAG (0.35 at%) ceramics against single crystals of identical composition gave scattering coefficients of 0.0046/cm for YAG ceramic and 0.0093/cm for Yb:YAG ceramic, with absorption coefficients of 0.0267/cm and 0.0221/cm respectively, close to the crystals. In a 10 mm rod model the ceramic’s parasitic loss was about 0.0314 versus 0.0108 for the crystal — roughly triple — yet at 40% output coupling the two loss factors remained close (0.927 versus 0.974).
01 / Why scattering and absorption resist separation
A laser gain material’s attenuation coefficient is the sum of two parts: absorption, where photons are genuinely consumed, and scattering, where photons are deflected out of the beam while their energy stays in the material. Conventional transmission measurements cannot separate them, because both dim the emerging light. Cavity ring-down spectroscopy is exquisitely sensitive but needs extra modelling and a high-finesse cavity to split the two; laser calorimetry measures absorption directly but says little about scattering; photothermal interferometry and angle-resolved scattering each target only one channel.
The integrating sphere has long been the workhorse for separating the two: collecting both transmitted and diffusely scattered light, it can in principle yield independent scattering and absorption coefficients. Its weakness is that light passes through the sample only once — and for the low-loss materials that matter for high-power lasers, the signal is too small for precision.
02 / Seven passes: the numbers and the check
The team placed each sample at the centre of a multipass cell inside the integrating sphere: two slightly tilted mirrors, highly reflective at 1064 nm, bounced the beam seven times through the material before it exited. The source was a Q-switched pulsed laser (1064 nm, 5 kHz, 120 ns) with a beam diameter of roughly 3 mm at the sample. Seven times the interaction length means seven times the accumulated loss signal, and the attenuation coefficient’s relative uncertainty falls to one seventh of the single-pass value — theory and experiment agreeing.
All four samples measured 15 × 5 × 10 mm and were purchased from the 11th Research Institute of China Electronics Technology Group Corporation: pure YAG ceramic (A1), Yb:YAG ceramic doped at 0.35 at% (A2), and single crystals of the same dimensions and doping (B1, B2). At 1064 nm, scattering was 0.0046/cm for YAG ceramic and 0.0093/cm for Yb:YAG ceramic — roughly double — with absorption of 0.0267/cm and 0.0221/cm respectively, close to the crystals. One conclusion deserves emphasis: absorption and scattering proved comparable in magnitude, so absorption cannot be neglected when evaluating ceramics. The authors estimate that intrinsic Yb absorption at 1064 nm accounts for only a small fraction of the measured value, attributing the rest to extrinsic host losses: structural defects, residual pores, scattering centres and residual impurities.
The validation step removed the two mirrors and repeated the measurements single-pass: attenuation trends matched, but the relative uncertainties of the absorption, scattering and attenuation coefficients were all larger single-pass — the attenuation uncertainty roughly seven times larger, exactly as multipass theory predicts. Wavefront measurements (He-Ne at 632.8 nm, Shack-Hartmann sensor) told the same story: ceramics at 0.028/0.037 μm RMS (A1/A2) versus crystals at 0.020/0.026 μm, with the Yb:YAG ceramic showing the largest deviation and clearest beam degradation.
03 / What it means for high-power ceramic lasers
Feeding the measured attenuation coefficients into a classical cavity-loss analysis, the authors estimate that for a 10 mm laser rod, material-related parasitic loss is about 0.0314 for the Yb:YAG ceramic versus 0.0108 for the crystal — the crystal roughly one third of the ceramic. Yet at an identical 40% output-coupler transmission, the corresponding loss factors are 0.927 (ceramic) and 0.974 (crystal), still close. The authors read this as evidence that high-quality, low-loss Yb:YAG ceramics can still offer favourable cavity-loss conditions and genuine potential for high-performance laser applications.
For laser processing and high-power solid-state lasers, the method may matter more than any single number: whether ceramics’ advantages over crystals — large size, high doping, cost-effective production — pay off depends on whether fabrication can press down pores and grain-boundary scattering, and that needs a yardstick for “how far is left.” The integrating-sphere multipass cell is such a yardstick: one configuration separating scattering from absorption, with precision improving linearly in the number of passes.
PRIMARY REFERENCES & LINKS
- High-precision characterization of scattering losses in YAG and Yb:YAG ceramics with an integrating sphere and multipass-cell
Full paper verified: publication date 26 October 2026, all coefficients, wavefront data and cavity-loss analysis come from the paper; Fig. 1 used under CC BY 4.0 with attribution.
Open source link ↗
