The work stands apart: there is almost no one nearby, the nearest neighbor sits at a distance of 0.545, and nothing is directly adjacent. The thematic group of the map is semiconductor, thermal_conductivity, carbon, superconductivity (2 of 8 neighbors). Around it lie studies of thermal conductivity and phonon dynamics in semiconductors, including work on laser cooling, superconductivity and the phonon bottleneck in ultrafast diffusion. But none of the neighboring works resolves the dynamics of individual acoustic branches with such precision or ties it directly to heat transport. BAs has drawn attention since 2013, when theory predicted its record thermal conductivity due to a large acoustic-optical gap that suppresses three-phonon scattering. However, thermal conductivity reflects only the aggregate effect near equilibrium and does not show how scattering governs the relaxation of individual modes. The coexistence of three- and four-phonon channels further muddles the microscopic pathways of energy redistribution. Resolving mode dynamics is necessary to understand the characteristic times and pathways of nonequilibrium phonon relaxation in BAs.
We used optical pumping and X-ray probing at the XPP (X-ray Pump Probe) instrument at the Linac Coherent Light Source (LCLS) at SLAC. A high-quality BAs crystal was oriented with its [111] surface and excited by 400 nm pulses (3.1 eV) of 50 fs duration with a fluence of 22 mJ/cm². Probing used X-ray pulses at 10.9 keV at a grazing angle of 0.76° to match the penetration depths of the optical and X-ray beams. Scattered radiation was recorded by an area detector with 75 µm pixels. To isolate the acoustic branches, we used calculations of one-phonon structure factors: the transverse acoustic branch (TA2) has maximum sensitivity at θ=115°, and the longitudinal one (LA) at θ=35°. In these directions we extracted the time dependences of the diffuse scattering intensity change and fitted them with an exponential rise plus oscillations. The time resolution was about 80 fs, and jitter was corrected frame by frame.
The measurements revealed an unusually slow rise of diffuse scattering near the (220) reflection, over tens of picoseconds. Fitting that accounted for Bose-Einstein occupation gave thermalization times of τ_TA = 29.2 ± 2.0 ps for the transverse acoustic branch and τ_LA = 13.9 ± 1.3 ps for the longitudinal one. The ratio τ_TA/τ_LA ≈ 2.1 agrees with calculations that predicted slower thermalization for TA. Even the faster LA branch thermalizes at least three times more slowly than in diamond (4.4 ps), GaN (3.0 ps), Si (2.0 ps), Ge (2.7 ps) and PbTe (2.8 ps). With 1 ps time binning, coherent oscillations appeared at 0.19 and 0.27 THz, corresponding to the TA and LA modes near the center of the Brillouin zone. Both components show virtually no damping throughout the entire 50 ps observation window. Wavelet analysis confirmed it: the spectral power remains nearly constant, with no measurable decay. By comparison, in PbTe the coherence amplitude after 50 ps drops to less than 1%, in Si to 42%, in GaN to 71%, while in diamond, as in BAs, no damping is visible.
The work differs from its neighbors in that it is the first to directly resolve the dynamics of individual acoustic branches in BAs with femtosecond time resolution and momentum resolution across the entire Brillouin zone, whereas most neighboring studies are limited either to integral thermal conductivity measurements or to optical methods that average over branches and are available only near the zone center. This is the first experimental test of the predicted mode-selective relaxation in BAs. The observation of slow thermalization and long coherence provides a direct dynamical signature of the weak phonon scattering that underlies the exceptional thermal conductivity of BAs. The results challenge standard two-temperature models that assume instantaneous thermalization of the phonon bath and highlight the importance of four-phonon processes in nonequilibrium dynamics.
The topic could move toward more detailed mapping of mode dynamics at different temperatures, excitation densities and crystal orientations. It would be interesting to study how long-lived acoustic phonon coherence affects heat transport in nanoscale devices and interactions with electrons. Experiments with coherent control of phonons using sequences of optical pulses may emerge. Theoretical work could refine the contribution of four-phonon scattering and predict new materials with even weaker scattering.
The results matter for phonon physics, thermoelectrics, heat removal in microelectronics and the design of materials with tunable heat transport. They also provide a benchmark for testing first-principles calculations of phonon dynamics.
Repeat the measurements at various temperatures and excitation levels to build a complete map of relaxation times. Investigate other crystallographic orientations and compare with first-principles calculations.
The work is directly tied to an unsolved problem in phonon transport: how weak scattering and long-lived coherent modes coexist with high thermal conductivity. It also touches on the limits of validity of the two-temperature model and the role of four-phonon processes in nonequilibrium dynamics.
Boron arsenide has a thermal conductivity above 1400 W/(m·K) at room temperature, comparable to diamond, but BAs is far cheaper and easier to process. The coherent oscillation frequencies of 0.19 and 0.27 THz lie in the terahertz range, which is being actively developed for 6G wireless communications.
Arthur C. Clarke's novel "The Fountains of Paradise" describes a material with exceptional thermal conductivity for a space elevator. BAs, with its record thermal conductivity and long-lived phonons, brings science fiction closer to reality.