Disorder-Induced Slow Relaxation of Phonon Polarization

Shuichi Murakami; Yuta Suzuki

REVIEW 3 major objections 4 minor 99 references

Disorder-Induced Slow Relaxation of Phonon Polarization

T0 review · 3 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This paper proposes that elastic scattering by isotropic disorder slows the relaxation of acoustic-phonon polarization through motional narrowing, so stronger disorder lengthens polarization lifetimes rather than shortening them.

desk verdict Clean derivation of a phonon-polarization motional-narrowing mechanism, but the mass-disorder example may be dominated by the very branch-flip scattering the model leaves out. read the letter →

arxiv 2501.07871 v1 pith:DT4LEIB6 submitted 2025-01-14 cond-mat.mes-hall

Yuta Suzuki , Shuichi Murakami This is my paper · ORCID

classification cond-mat.mes-hall

keywords phononangularmomentumpolarizationmotionalnarrowingDyakonov-Perelmechanismdisorderscatteringquantumkineticequationcircularlinear

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The authors argue that in disordered crystals, isotropic elastic scattering protects—rather than destroys—the polarization of acoustic phonons. The mechanism is motional narrowing: longitudinal–transverse band splitting makes polarization precess, and frequent collisions randomize the precession so its net effect averages away, leaving a slow decay. Their kinetic calculation gives a circular-polarization (phonon angular momentum) relaxation time $\tau_{\mathrm{AM}} \propto 1/\tau^{*}$, so more impurity scattering means longer-lived phonon angular momentum, and linear polarizations relax even more slowly. If correct, this provides a concrete route to extending phonon angular-momentum lifetimes in mass-disordered crystals such as solid argon.

What carries the argument

The engine of the argument is the quantum kinetic equation for the $3\times3$ phonon density matrix, expanded in the eight Gell-Mann matrices $\lambda_a$ (the SU(3) analogue of Pauli matrices): three antisymmetric components measure circular polarization, i.e. phonon angular momentum, and five symmetric components measure linear polarization. The collision integral is taken as elastic, isotropic, and polarization-independent, and the equation is expanded in spherical harmonics; the slowly relaxing $\ell=0$ components couple to fast-decaying $\ell=2$ multipoles, and adiabatic elimination produces the Hamiltonian-independent double-commutator formula $\partial\langle\rho_q\rangle/\partial t \simeq -(\tau^{*}/\hbar^2)\langle[H_q,[H_q,\langle\rho_q\rangle]]\rangle$. That formula, evaluated analytically for isotropic elastic bodies and numerically for a cubic crystal, yields the relaxation rates and the linear-slower-than-circular ordering.

What would settle it

Measure the circular- and linear-polarization relaxation times in a family of isotopically mass-substituted crystals (e.g., $^{36}$Ar$_{1-x}^{136}$Xe$_x$) at fixed temperature and wavevector $qa/2\pi\simeq0.4$, varying the substitution fraction $x$. The paper predicts $1/\tau_{\mathrm{AM}}$ and $1/\tau_{\mathrm{LP}}$ to decrease with $x$ and the diffusion lengths $\sqrt{D\tau}$ to stay constant; observing the opposite—rates increasing with impurity content—would falsify the mechanism.

Watch

Extended reading notes

Core claim

The central discovery is a relaxation mechanism for all acoustic-phonon polarization degrees of freedom in which disorder impedes relaxation. In the absence of scattering, the polarization vector oscillates because the longitudinal–transverse splitting acts as a wavevector-dependent effective field; when the scattering rate exceeds the oscillation frequency, adiabatic elimination of the fast multipole components yields slow decay with rates $1/\tau_{\mathrm{AM}} = \frac{2}{3}(c_L-c_T)^2 q^2 \tau_2$ for circular polarization and $1/\tau_{\mathrm{LP}} = \frac{2}{5}(c_L-c_T)^2 q^2 \tau_2$ for linear polarization in isotropic elastic bodies. Because these rates are proportional to the impurity relaxation time $\tau_2 \propto 1/\alpha$, increasing disorder strength lowers the rates, and the linear-polarization rates fall below the circular one. The paper validates the mechanism numerically for solid argon with Rayleigh-type mass-disorder scattering, reproducing the inverse proportionality to impurity concentration and the slower relaxation of linear polarizations.

Load-bearing premise

The load-bearing premise is that disorder scattering cares only about the phonon's energy and direction, never about which polarization or which acoustic branch it scatters; if real defects can tell the branches apart or flip polarization directly, the protective averaging can be bypassed and disorder would shorten lifetimes instead of lengthening them.

Editorial extensions

If this is right

Larger mass-disorder strength $\alpha$ increases both $\tau_{\mathrm{AM}}$ and $\tau_{\mathrm{LP}}$, so isotope engineering can extend phonon polarization lifetimes without changing the phonon band structure.
Linear polarizations outlive circular ones in the same crystal, so phonon-angular-momentum probes see the shortest of the three rates while linear-polarization probes see longer decay.
Short-wavelength phonons dominate polarization transport: the rate $1/\tau_{\mathrm{Pol}}$ scales as $q^{-2}$ at small $q$ and $q^{-4}$ at large $q$, so injecting high-$q$ phonons should reveal the longest-lived polarizations.
The polarization diffusion length $l=\sqrt{D\tau_{\mathrm{Pol}}}$ is independent of impurity concentration because $D\propto\tau^{*}$ and $\tau_{\mathrm{Pol}}\propto 1/\tau^{*}$, an observable signature that does not require time-resolved detection.
In solid argon with 6% xenon mass substitution, the estimated polarization relaxation times exceed the impurity relaxation time near $qa/2\pi \simeq 0.4$ (about 40 K), indicating a realistic regime for observing long-lived phonon polarization.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

Editorial extension: the same double-commutator structure suggests that any three-component bosonic polarization degree of freedom whose band splitting acts as a $\mathbf{q}$-dependent field—electron orbital angular momentum, possibly magnon or optical-phonon polarizations—should exhibit disorder-protected relaxation under isotropic elastic scattering.
Editorial extension: the predicted linear-slower-than-circular hierarchy implies an experimental asymmetry test: preparing pure linear polarization states should yield longer-lived signals than circular ones, which could distinguish this mechanism from Elliott-Yafet-like direct depolarization.
Editorial extension: because polarization lifetimes grow with wavevector while heat is carried by small-$q$ phonons, the mechanism implies a spectral separation between heat transport and polarization transport that could be exploited in devices converting heat gradients into chiral-phonon currents.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Dear X,

Here's my take on Suzuki-Murakami. The paper's core is a real contribution: it sets up the eight-component polarization structure of acoustic phonons, derives the Dyakonov-Perel-type relaxation rates 1/τ_AM = (2/3)(cL−cT)^2 q^2 τ2 and 1/τ_LP = (2/5)(cL−cT)^2 q^2 τ2, and shows linear polarization relaxes slower than circular. The supplement is thorough, with explicit Gell-Mann and spherical-harmonic coefficients and a calculation for solid argon. Within the stated model, the math checks out.

The soft spot is exactly where the reader put it: the kinetic equation (3) uses a scalar, branch-averaged collision integral. That is a standard approximation in electron Dyakonov-Perel, but for phonons the 'spin' is the polarization vector itself, and isotope-mass disorder has a golden-rule matrix element proportional to e_{q,s}·e_{q',s'}. For L to T branch flips that overlap is order one, not suppressed by (cL−cT)/v. So there is a plausible Elliott-Yafet-like channel that relaxes polarization at a rate comparable to 1/τ*. The authors assert condition (ii), that polarization-conserving scattering dominates, but they don't verify it for the 36Ar/136Xe example. Footnote [75] explicitly concedes the approximation 'is not universally valid.' If a branch-flip rate B/τ* is present, the total AM decay rate is roughly A Ω^2 τ* + B/τ*, and for the q and α in Fig. 3(b) the second term can win. That would overturn the quantitative prediction for that system, though the motional-narrowing idea could still survive if a cleaner realization is found.

The multipole truncation at l=0,2 is a smaller issue; it's standard and the authors note the couplings.

Net: this is a serious, readable theory paper that deserves referee time. My recommendation is to send it out, with a referee asked to estimate the branch-flip scattering rate for mass disorder or otherwise justify condition (ii) for the specific example. If that can't be done, the quantitative claims should be softened to a model statement. As is, I'd cite it as the first derivation of this mechanism, with a caveat.

Best.

Referee Report

3 major / 4 minor

Summary. The manuscript proposes a motional-narrowing mechanism for the relaxation of phonon polarization in disordered crystals. The authors introduce eight Gell-Mann polarization parameters for acoustic phonons, show that band splitting causes coherent polarization oscillations, and derive from a quantum kinetic equation with a scalar elastic collision integral that the circular-polarization (AM) and linear-polarization relaxation rates are 1/τ_AM = (2/3)(c_L−c_T)^2 q^2 τ_2 and 1/τ_LP = (2/5)(c_L−c_T)^2 q^2 τ_2 for isotropic elastic bodies, together with a general formula (10) applicable to cubic crystals. They apply the formula to solid argon with mass-defect disorder, predict 1/τ_Pol ∝ q^{-2} to q^{-4}, and conclude that stronger disorder lengthens polarization lifetimes and that linear polarizations decay more slowly than circular polarizations. The Supplemental Material provides the Gell-Mann structure constants, spherical-harmonic expansion coefficients, explicit submatrices, and the mass-disorder strength estimate.

Significance. If the result holds beyond the scalar-collision approximation, the paper would establish a genuinely counterintuitive and falsifiable effect: polarization lifetimes increasing with disorder, with a clean separation between LP and CP rates. The analytic prefactors 2/3 and 2/5 are derived without fitted parameters, the multipole elimination is explicit, and the solid-argon estimate gives concrete numbers for experiment, including the prediction that the AM and LP diffusion lengths are independent of impurity concentration. These are real strengths. The main risk is not internal consistency of the derivation but the physical completeness of the collision model in Eq. (3), which the authors themselves flag in footnote [75] as not universally valid.

major comments (3)

[Eq. (3) and footnote [75]] The central quantitative claim rests on a collision integral in which K_{q,q'} is a scalar and energy conservation uses the averaged dispersion ℏvq. For the isotopic mass disorder used in Fig. 3, the exact golden-rule matrix element is proportional to e_{q,s}·e_{q',s'}, so longitudinal-to-transverse branch-flip scattering is not suppressed by (c_L−c_T)/v; after angular averaging the Elliott-Yafet-like channel has a rate 1/τ_EY comparable to 1/τ*. Since footnote [75] concedes that Eq. (3) is 'not universally valid,' the paper needs an explicit estimate of 1/τ_EY for the 36Ar/136Xe model, or a separate calculation with the exact matrix element, before the rates in Fig. 3(b) and the claim τ_AM ∝ 1/τ* can be accepted for mass disorder. This is load-bearing because the total AM decay rate would be roughly A(c_L−c_T)^2 q^2 τ* + B/τ_EY, and the second term grows with disorder.
[Condition (ii) after Eq. (10)] The authors assert that the polarization-conserving scattering rate 1/τ* is higher than the polarization-flip scattering rate, but no calculation of the latter is provided. For the solid-argon example the disorder model itself generates branch-flip scattering, so this condition is not automatically satisfied. The authors should either verify the inequality quantitatively for the mass-substitution parameters used in Fig. 3, or state the prediction conditionally on it; without this, the variational prediction and the experimental estimate based on 36Ar_{1−x}136Xe_x are incomplete.
[Eq. (10) and the definition of τ*] The replacement of τ_2 by τ* ≡ max{τ_l} in Eq. (10) is an uncontrolled approximation when the l=1 relaxation time differs strongly from τ_2. For Rayleigh scattering from mass defects, τ_1 and τ_2 are generally different, and the numerical rates in Fig. 3(b) would change if the actual τ_2 from the mass-disorder model were used instead of τ*. The authors should justify this replacement or use the appropriate τ_2 in the numerical calculation; this is secondary to the branch-flip issue but affects the quantitative predictions.

minor comments (4)

[Eq. (7)] The symbol O is used for zero submatrices in Eq. (7) but is not defined; please define it or use bold zero notation.
[After Eq. (2)] The sentence 'This is an analog of precession of an electron spin in the three-dimensional space' is informal; specify the analogy as precession on the eight-dimensional hypersphere described in the Supplement.
[Eq. (S46)] Please check the numerical prefactor and units in the mass-difference scattering rate in Eq. (S46); the surrounding text should specify the density-of-states convention used, since the conventional Tamura/Klemens result can differ by factors of π/2 or by the unit-cell volume factor.
[Fig. 3(b)] The three disorder strengths are described in the caption but the curves are not individually labeled in the figure; adding labels or a legend would improve readability.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the τ_AM ∝ 1/τ* result follows analytically from the stated kinetic model, and the numerical figure evaluates that same formula rather than inverting fitted data.

full rationale

The central derivation is self-contained. The relaxation rates (9a)-(9b) are obtained by adiabatic elimination of the l=2 multipoles in Eq. (7), giving 1/τ_AM = (2/3)(c_L−c_T)^2 q^2 τ2 and 1/τ_LP = (2/5)(c_L−c_T)^2 q^2 τ2. The inverse dependence on the impurity scattering time is a mathematical consequence of the assumed scalar, polarization-independent collision integral in Eq. (3), not a parameter fitted to polarization relaxation data. The solid-argon estimate uses a conventional mass-defect model for α (α(2π/a)^4 = 28 ps^-1 for x=6%) and then evaluates formula (10); Fig. 3(b) is therefore a consistency check of the analytic formula, not an independent empirical test, but the paper does not claim otherwise. The asserted condition (ii) that polarization-conserving scattering dominates polarization-flip scattering, and the caveat in footnote [75] that the scalar δ-function approximation 'is not universally valid,' are scope and robustness limitations of the model rather than circular steps. Self-citations (e.g., Refs. 12, 17, 18, 22, 39, 40) are background for phonon AM and are not load-bearing in the derivation; no uniqueness claim is imported from prior same-author work. No fitted quantity is renamed as a prediction, and no result is equivalent to its input by definition.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central claim rests on a small set of physical inputs: a symmetric phonon Hamiltonian, a scalar polarization-independent collision integral, and the overdamped regime. The five linear polarization parameters are not new physical entities but bookkeeping components of the 3x3 density matrix. The main unvalidated input is the scattering isotropy assumption.

free parameters (2)

Impurity relaxation times τ_l (τ2 and τ*) = not fitted; defined by 1/τ_l = (2π/ℏ)g(ω_q)∫ K_{q,q'}[1-P_l(cos ϑ)] d(cos ϑ)/2
The derived rates (9a,b) are proportional to τ2, with τ* used as a typical value in Eq. (10). These are inputs from the assumed scattering model, not fitted to the target polarization relaxation.
Disorder strength α in solid argon = α(2π/a)^4 = 1, 10, 100, 28 ps^-1
Chosen for illustration and estimated for 36Ar0.94 136Xe0.06 via Eq. (S46). It affects quantitative rates in Fig. 3(b) but not the functional scaling.

assumptions (4)

domain assumption Nonmagnetic centrosymmetric crystals have a real symmetric dynamical matrix, so H_q = h0(q) I3 + Σ_{a∈S} h_a(q) λ_a.
Supplement S2 derives this from time-reversal symmetry and centrosymmetry. It restricts the theory to nonmagnetic centrosymmetric crystals.
domain assumption The collision integral uses scalar, polarization-independent elastic scattering with an averaged dispersion δ(ℏ v q - ℏ v q').
Eq. (3) and footnote [75]. This is the load-bearing assumption for motional narrowing; the authors state it is not universally valid because longitudinal/transverse degeneracy is lifted.
ad hoc to paper Overdamped regime: H'_q τ*/ℏ ≪ 1, with l=2 multipoles eliminated quasistatically and l≠0,2 multipoles omitted.
Eq. (7) and text after Eq. (10). The truncation is justified only at second order in small H'_q τ*/ℏ and is not fully controlled in the supplement.
domain assumption Mass-difference scattering rate follows Rayleigh scaling, 1/τ* = α q^4.
Supplement S7 uses the conventional mass-disorder model. This is an input for the numerical estimate, not derived in the paper.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Disorder-Induced Slow Relaxation of Phonon Polarization." pith.science (2026). https://pith.science/paper/DT4LEIB6

@misc{pith2026250107871,
  author       = {Pith},
  title        = {Pith review of: Disorder-Induced Slow Relaxation of Phonon Polarization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DT4LEIB6}},
  note         = {Machine review of arXiv:2501.07871}
}

read the original abstract

The role of the polarization degree of freedom in lattice dynamics in solids has been underlined recently. We theoretically discover a relaxation mechanism for both linear and circular polarizations of acoustic phonons. In the absence of scattering, the polarization exhibits oscillatory behavior. This behavior leads to a counterintuitive result: unlike linear momentum, more frequent scattering events cause slower polarization relaxation due to motional narrowing. We validate this mechanism using the quantum kinetic equation. We derive the relaxation rates of polarizations analytically for isotropic elastic bodies and numerically for a cubic crystal. Remarkably, we reveal that linear polarizations relax more slowly than circular ones. Our findings provide a pathway to extend the lifetime of phonon angular momentum.

Figures

Figures reproduced from arXiv: 2501.07871 by the authors.

**Figure 2.** FIG. 2. Schematic list of eight parameters [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗

**Figure 3.** FIG. 3. Phonon properties and relaxation dynamics in dis [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

99 extracted references · 69 canonical work pages

[75]

J. Sohn, J. M. Lee, and H.-W. Lee, Phys. Rev. Lett.132, 246301 (2024)

2024
[1]

S. V. Vonsovskii and M. S. Svirskii, Fiz. Tverd. Tela3, 2160 (1961), [Sov. Phys. Solid State3, 7 (1962)]

1961
[2]

denotes a vector of multipole compo- nents ρa 2m(q, t) with a ∈ AS (or a ∈ S) coupled to the dynamics of ρa 00(q, t). We omitted other multipole com- ponents ρa lm with l ̸= 0, 2appearinginEq.(7), sincethese components do not directly couple to the dynamics ofρa 00 whereas they may couple to that ofρAS/S 2 . The submatri- ces Ω, eΩ, and Ω′ consist of the ...
[3]

A. D. Levine, Il Nuovo Cimento26, 190 (1962)

1962
[4]

D. L. Portigal and E. Burstein, Phys. Rev. 170, 673 (1968)

1968
[5]

A. S. Pine, Phys. Rev. B2, 2049 (1970)

1970
[6]

Ishii, Bussei-Kenkyu23, 217 (1975)

T. Ishii, Bussei-Kenkyu23, 217 (1975)

1975
[7]

McLellan, J

A. McLellan, J. Phys. C: Solid State Phys. 21, 1177 (1988)

1988

Show all 99 references

[8]

Zhang and Q

L. Zhang and Q. Niu, Phys. Rev. Lett. 112, 085503 (2014)

2014
[9]

Zhang and Q

L. Zhang and Q. Niu, Phys. Rev. Lett. 115, 115502 (2015)

2015
[10]

S.-Y. Chen, C. Zheng, M. S. Fuhrer, and J. Yan, Nano Lett. 15, 2526 (2015)

2015
[11]

H. Zhu, J. Yi, M.-Y. Li, J. Xiao, L. Zhang, C.-W. Yang, R. A. Kaindl, L.-J. Li, Y. Wang, and X. Zhang, Science 359, 579 (2018)

2018
[12]

Kishine, A

J. Kishine, A. S. Ovchinnikov, and A. A. Tereshchenko, Phys. Rev. Lett.125, 245302 (2020)

2020
[13]

Zhang and S

T. Zhang and S. Murakami, Phys. Rev. Res.4, L012024 (2022)

2022
[14]

Kato and J.-i

A. Kato and J.-i. Kishine, J. Phys. Soc. Jpn.92, 075002 (2023)

2023
[15]

Ishito, H

K. Ishito, H. Mao, Y. Kousaka, Y. Togawa, S. Iwasaki, T. Zhang, S. Murakami, J.-i. Kishine, and T. Satoh, Nat. Phys. 19, 35 (2023)

2023
[16]

Kusunose, J

K.Ishito, H.Mao, K.Kobayashi, Y.Kousaka, Y.Togawa, H. Kusunose, J. Kishine, and T. Satoh, Chirality35, 338 (2023)

2023
[17]

Oishi, Y

E. Oishi, Y. Fujii, and A. Koreeda, Phys. Rev. B109, 104306 (2024)

2024
[18]

Hamada, E

M. Hamada, E. Minamitani, M. Hirayama, and S. Mu- rakami, Phys. Rev. Lett.121, 175301 (2018)

2018
[19]

Hamada and S

M. Hamada and S. Murakami, Phys. Rev. B101, 144306 (2020)

2020
[20]

Park and B.-J

S. Park and B.-J. Yang, Nano Lett.20, 7694 (2020)

2020
[21]

Zhang, C

H. Zhang, C. Xu, C. Carnahan, M. Sretenovic, N. Suri, D. Xiao, and X. Ke, Phys. Rev. Lett.127, 247202 (2021)

2021
[22]

H. Chen, W. Wu, J. Zhu, Z. Yang, W. Gong, W. Gao, S. A. Yang, and L. Zhang, Nano Lett.22, 1688 (2022)

2022
[23]

Yao and S

D. Yao and S. Murakami, Phys. Rev. B 105, 184412 (2022)

2022
[24]

C. P. Romao, R. Catena, N. A. Spaldin, and M. Matas, Phys. Rev. Res.5, 043262 (2023)

2023
[25]

V. L. Korenev, M. Salewski, I. A. Akimov, V. F. Sapega, L. Langer, I. V. Kalitukha, J. Debus, R. I. Dzhioev, D. R. Yakovlev, D. Müller, C. Schröder, H. Hövel, G. Kar- czewski, M. Wiater, T. Wojtowicz, Y. G. Kusrayev, and M. Bayer, Nat. Phys.12, 85 (2016)

2016
[26]

T. F. Nova, A. Cartella, A. Cantaluppi, M. Först, D. Bossini, R. V. Mikhaylovskiy, A. V. Kimel, R. Merlin, and A. Cavalleri, Nat. Phys.13, 132 (2017)

2017
[27]

Holanda, D

J. Holanda, D. Maior, A. Azevedo, and S. Rezende, Nat. Phys. 14, 500 (2018)

2018
[28]

Sasaki, Y

R. Sasaki, Y. Nii, and Y. Onose, Nat. Commun.12, 2599 (2021)

2021
[29]

S. G. Jeong, J. Kim, A. Seo, S. Park, H. Y. Jeong, Y.-M. Kim, V. Lauter, T. Egami, J. H. Han, and W. S. Choi, Sci. Adv.8, 4 (2022)

2022
[30]

K. Kim, E. Vetter, L. Yan, C. Yang, Z. Wang, R. Sun, Y. Yang, A. H. Comstock, X. Li, J. Zhou, L. Zhang, W. You, D. Sun, and J. Liu, Nat. Mater.22, 322 (2023)

2023
[31]

H. Ueda, M. García-Fernández, S. Agrestini, C. P. Ro- mao, J. van den Brink, N. A. Spaldin, K.-J. Zhou, and U. Staub, Nature618, 946 (2023)

2023
[32]

K. Ohe, H. Shishido, M. Kato, S. Utsumi, H. Matsuura, and Y. Togawa, Phys. Rev. Lett.132, 056302 (2024)

2024
[33]

D. M. Juraschek and N. A. Spaldin, Phys. Rev. Mater. 3, 064405 (2019)

2019
[34]

Hamada and S

M. Hamada and S. Murakami, Phys. Rev. Res.2, 023275 (2020)

2020
[35]

D. M. Juraschek, T. c. v. Neuman, and P. Narang, Phys. Rev. Res.4, 013129 (2022)

2022
[36]

S. R. Tauchert, M. Volkov, D. Ehberger, D. Kazenwadel, M. Evers, H. Lange, A. Donges, A. Book, W. Kreuzpaint- ner, U. Nowak,et al., Nature602, 73 (2022)

2022
[37]

Fransson, Phys

J. Fransson, Phys. Rev. Res.5, L022039 (2023)

2023
[38]

Chaudhary, D

S. Chaudhary, D. M. Juraschek, M. Rodriguez-Vega, and G. A. Fiete, Phys. Rev. B110, 094401 (2024)

2024
[39]

Funato, M

T. Funato, M. Matsuo, and T. Kato, Phys. Rev. Lett. 132, 236201 (2024)

2024
[40]

Yao and S

D. Yao and S. Murakami, J. Phys. Soc. Jpn.93, 034708 (2024)

2024
[41]

Suzuki, S

Y. Suzuki, S. Sumita, and Y. Kato, arXiv:2409.08874 [cond-mat.mes-hall]

arXiv
[42]

T. Lin, X. Chen, R. Xu, J. Luo, and H. Zhu, Nano Lett. 24, 10592 (2024). 6

2024
[43]

E.M.Gershenzon, N.M.Pevin,andM.S.Fogel’son,Fiz. Tekh. Poluprov.3, 748 (1969), [Sov. Phys. Semicond.3 634 (1969)]

1969
[44]

M. I. D’yakonov and V. I. Perel’, Zh. Eksp. Teor. Fiz60, 1954 (1971), [Sov. Phys. JETP33 1053 (1971)]

1971
[45]

M. I. D’yakonov and V. I. Perel’, Fiz. Tverd. Tela13, 3581 (1971), [Sov. Phys. Solid State13, 3023 (1972)]

1971
[46]

Fabian, A

J. Fabian, A. Matos-Abiague, C. Ertler, P. Stano, and I. Žutić, Acta Phys. Slovaca. Reviews and Tutorials57, 565 (2007), arXiv:0711.1461

2007 arXiv
[47]

P. G. Klemens, Proc. Phys. Soc. A68, 1113 (1955)

1955
[48]

Tamura, Phys

S.-I. Tamura, Phys. Rev. B27, 858 (1983)

1983
[49]

G. P. Srivastava,The Physics of Phonons (Adam Hilger, Bristol, Philadelphia, and New York, 1990)

1990
[50]

J. M. Ziman, Electrons and Phonons: the Theory of Transport Phenomena in Solids , International Series of Monographs on Physics (Clarendon Press, Oxford, UK, 1960)

1960
[51]

For electron spins, a similar process where spin-flip scattering from impurities accelerates AM relaxation is known as the Elliott–Yafet mechanism [45, 91, 92]
[52]

Bloembergen, E

N. Bloembergen, E. M. Purcell, and R. V. Pound, Phys. Rev. 73, 679 (1948)

1948
[53]

P. W. Anderson and P. R. Weiss, Rev. Mod. Phys.25, 269 (1953)

1953
[54]

J. H. Van Vleck, Phys. Rev.74, 1168 (1948)

1948
[55]

Kubo and K

R. Kubo and K. Tomita, J. Phys. Soc. Jpn.9, 888 (1954)

1954
[56]

M. Toda, R. Kubo, N. Saito, and N. Hashitsume,Sta- tistical Physics II: Nonequilibrium Statistical Mechanics , 2nd ed., Springer Series in Solid-State Sciences, Vol. 31 (Springer, Berlin, 1991)

1991
[57]

Kittel, Introduction to Solid State Physics , 8th ed

C. Kittel, Introduction to Solid State Physics , 8th ed. (Wiley, New York, US, 2004)

2004
[58]

However, in isotropic elas- tic bodies, the AM relaxation time τ AM remains unaf- fected by this correction term [64]

If we define phonon AM by the statistical average of me- chanical AM of all atoms [1, 2, 5–7], other interband term appears off equilibrium [7, 67]. However, in isotropic elas- tic bodies, the AM relaxation time τ AM remains unaf- fected by this correction term [64]
[59]

Gell-Mann, Phys

M. Gell-Mann, Phys. Rev.125, 1067 (1962)

1962
[60]

Roman, Il Nuovo Cimento13, 974 (1959)

P. Roman, Il Nuovo Cimento13, 974 (1959)

1959
[61]

Barakat, Opt

R. Barakat, Opt. Commun.23, 147 (1977)

1977
[62]

Setälä, A

T. Setälä, A. Shevchenko, M. Kaivola, and A. T. Friberg, Phys. Rev. E66, 016615 (2002)

2002
[63]

J. C. Samson, Geophys. J. R. Astr. Soc.34, 403 (1973)

1973
[64]

Sugic, M

D. Sugic, M. R. Dennis, F. Nori, and K. Y. Bliokh, Phys. Rev. Res.2, 042045 (2020)

2020
[65]

See Supplemental Material at [url] for further details, in- cluding the list of Gell-Mann matrices, symmetry proper- ties of the phonon Hamiltonian, invariants in undamped polarization dynamics, the definition of phonon AM, the expansion of the quantum kinetic equation, calcul...
[66]

R. E. Peierls, Quantum Theory of Solids , International Series of Monographs on Physics (Clarendon Press, Ox- ford, UK, 1955)

1955
[67]

Simoncelli, N

M. Simoncelli, N. Marzari, and F. Mauri, Nat. Phys.15, 809 (2019)

2019
[68]

Zhong, H

J. Zhong, H. Sun, Y. Pan, Z. Wang, X. Xu, L. Zhang, and J. Zhou, Phys. Rev. B107, 125147 (2023)

2023
[69]

Han, H.-W

S. Han, H.-W. Lee, and K.-W. Kim, Phys. Rev. Lett. 128, 176601 (2022)

2022
[70]

Kohn and J

W. Kohn and J. M. Luttinger, Phys. Rev. 108, 590 (1957)

1957
[71]

E. L. Ivchenko, Yu B. Lyanda-Geller, and G. E. Pikus, Pis’ma Zh. Eksp. Teor. Fiz.50, 156 (1989), [JETP Lett. 50 175 (1989)]

1989
[72]

E. L. Ivchenko, Yu B. Lyanda-Geller, and G. E. Pikus, Zh. Eksp. Teor. Fiz98, 989 (1990), [Sov. Phys. JETP71 550 (1990)]

1990
[73]

A. V. Shytov, E. G. Mishchenko, H.-A. Engel, and B. I. Halperin, Phys. Rev. B73, 075316 (2006)

2006
[74]

Kailasvuori, J

J. Kailasvuori, J. Stat. Mec.2009, P08004 (2009)

2009
[76]

While our approximation in Eq

To be exact, the collision term consists of symmetrized product of the matrix delta function δ(Hq − Hq′ ) and the density matrix difference(ρq − ρq′ ) [71]. While our approximation in Eq. (3) is not universally valid—since degeneracy between longitudinal and transverse branche...
[77]

[45, 71]

Similar discussion for electron spins in terms of the Dyakonov–Perel mechanism [42–44] is found in Refs. [45, 71]
[78]

In this approximation, we also regard the linear- momentum relaxation timeτ1 as τ ∗
[79]

V. N. Gridnev, JETP Lett.74, 380 (2001)

2001
[80]

A. A. Burkov and L. Balents, Phys. Rev. B69, 245312 (2004)

2004
[81]

Žutić, J

I. Žutić, J. Fabian, and S. Das Sarma, Rev. Mod. Phys. 76, 323 (2004)

2004
[82]

Culcer and R

D. Culcer and R. Winkler, Phys. Rev. B 76, 195204 (2007)

2007
[83]

Winkler, D

R. Winkler, D. Culcer, S. J. Papadakis, B. Habib, and M. Shayegan, Semicond. Sci. Technol.23, 114017 (2008)

2008
[84]

Boross, B

P. Boross, B. Dóra, A. Kiss, and F. Simon, Sci. Rep.3, 3233 (2013)

2013
[85]

A. Kiss, L. Szolnoki, and F. Simon, Sci. Rep.6, 22706 (2016)

2016
[86]

Szolnoki, B

L. Szolnoki, B. Dóra, A. Kiss, J. Fabian, and F. Simon, Phys. Rev. B96, 245123 (2017)

2017
[87]

Szolnoki, A

L. Szolnoki, A. Kiss, B. Dóra, and F. Simon, Sci. Rep.7, 9949 (2017)

2017
[88]

Cosacchi, M

M. Cosacchi, M. Cygorek, F. Ungar, and V. M. Axt, Phys. Rev. B95, 195313 (2017)

2017
[89]

A typical example of the polarization-flip scattering is a three-phonon process due to lattice anharmonicity [48, 49, 65]
[90]

D. K. Christen and G. L. Pollack, Phys. Rev. B12, 3380 (1975)

1975
[91]

Fujii, N

Y. Fujii, N. A. Lurie, R. Pynn, and G. Shirane, Phys. Rev. B10, 3647 (1974)

1974
[92]

R. J. Elliott, Phys. Rev.96, 266 (1954)

1954
[93]

Yafet, in Solid State Physics , Advances in Research andApplications, Vol.14,editedbyF.SeitzandD.Turn- bull (Academic Press, New York and London, 1963) pp

Y. Yafet, in Solid State Physics , Advances in Research andApplications, Vol.14,editedbyF.SeitzandD.Turn- bull (Academic Press, New York and London, 1963) pp. 1–98. 1 Supplemental Materials for: Disorder-Induced Slow Relaxation of Phonon Polarization Yuta Suzuki and Shuichi Mu...

1963
[94]

D(q) is a positive semi-definite Hermitian matrix,D†(q) = D(q)
[95]

Time-reversal symmetry requires a reciprocityD(q) = D⊤(−q)
[96]

These conditions ensure thatD = D(q) is a real symmetric matrix

For centrosymmetric crystals,D(−q) = D(q). These conditions ensure thatD = D(q) is a real symmetric matrix. We can thus diagonalize it asD = O∆O⊤. Here, O is a real orthogonal matrix, and∆ is a real diagonal matrix. SinceD is positive semi-definite, all diagonal elements of ∆ ...
[97]

,Λ8) moves on a hypersphere of radius qP8 a=1 Λ2a embedded in the eight- dimensional Euclidean space

(S9) We conclude that the vector(Λ1, Λ2, . . . ,Λ8) moves on a hypersphere of radius qP8 a=1 Λ2a embedded in the eight- dimensional Euclidean space. S4. DEFINITION OF PHONON ANGULAR MOMENTUM In footnote [56] of the main text, we noted that phonon angular momentum (AM) off equi...
[98]

(S3)], and 6

The structure constantfacb with c ∈ S becomes zero fora, b∈ S or a, b∈ AS [see also Eq. (S3)], and 6
[99]

ρ0 q′ · I3 + 8X a=1 ∞X l=0 lX m=−l ρa lm(q)Ylm( ˆq′)λa # −

The Wigner 3j symbol l 2 l′ 0 0 0 vanishes unless triangle inequality|l − l′| ≤2 ≤ l + l′ holds and l + l′ is even. Moreover, the coefficientΩa;b lm;l′m′ satisfies a symmetry relation: Ωb;a l′m′;lm = Ωa;b lm;l′m′ ∗ , (S27) which is derived as follows: iΩb;a l′m′;lm = 2(cL − cT...

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.

[1] [75]

J. Sohn, J. M. Lee, and H.-W. Lee, Phys. Rev. Lett.132, 246301 (2024)

2024

[2] [1]

S. V. Vonsovskii and M. S. Svirskii, Fiz. Tverd. Tela3, 2160 (1961), [Sov. Phys. Solid State3, 7 (1962)]

1961

[3] [2]

denotes a vector of multipole compo- nents ρa 2m(q, t) with a ∈ AS (or a ∈ S) coupled to the dynamics of ρa 00(q, t). We omitted other multipole com- ponents ρa lm with l ̸= 0, 2appearinginEq.(7), sincethese components do not directly couple to the dynamics ofρa 00 whereas they may couple to that ofρAS/S 2 . The submatri- ces Ω, eΩ, and Ω′ consist of the ...

[4] [3]

A. D. Levine, Il Nuovo Cimento26, 190 (1962)

1962

[5] [4]

D. L. Portigal and E. Burstein, Phys. Rev. 170, 673 (1968)

1968

[6] [5]

A. S. Pine, Phys. Rev. B2, 2049 (1970)

1970

[7] [6]

Ishii, Bussei-Kenkyu23, 217 (1975)

T. Ishii, Bussei-Kenkyu23, 217 (1975)

1975

[8] [7]

McLellan, J

A. McLellan, J. Phys. C: Solid State Phys. 21, 1177 (1988)

1988