The tunneling two-level system (TTLS) model has described the universal low-temperature acoustic properties of glasses for five decades, but its key quantitative success - the small, nearly material-independent ultrasonic absorption Q−1≈3×10−4Q^{-1}\approx 3\times 10^{-4} - rests on parameter cancellations that the model itself does not explain. Vural and Leggett (2011) proposed an alternative in which a glass is modeled as elastically coupled mesoscopic 'generic blocks' whose non-phononic degrees of freedom have arbitrary non-harmonic spectra; a real-space renormalization treatment of the phonon-mediated dipole interaction between blocks then yields a quasi-universal resonant-regime absorption of the right magnitude without postulating two-level systems. The paper's own conclusion, however, leaves the theory's central questions open, and as of 2026 they remain substantially so: the linear ultrasonic response in the low-frequency relaxation regime - which the authors call 'particularly puzzling' because its TTLS explanation appears to rely on parameter dependencies specific to that model - has not been derived in any generic framework; no complete or controlled renormalization treatment exists to decide whether the output stress response takes the TTLS form or preserves the logarithmic structure Q−1(ω)∝[ln⁡(U0/ω)]−1Q^{-1}(\omega)\propto[\ln(U_0/\omega)]^{-1}; and of the nonlinear phenomena (saturation, echoes, hole burning) only phonon echoes have been shown to be generic in published multi-level models, with saturation treated only in unpublished preprints and hole burning unaddressed. The problem is to complete - or refute - this generic-block account of the low-temperature acoustic phenomenology of glasses, comprising the relaxation-regime response, the nonlinear response, and the renormalization self-consistency question.

Background

Below about 1 K, virtually all amorphous solids ('glasses') exhibit nearly identical thermal and acoustic behavior: a heat capacity approximately linear in temperature, a thermal conductivity approximately quadratic in temperature, and a small, nearly material-independent ultrasonic attenuation. These universalities have been known since the measurements of Zeller and Pohl (1971) and are reviewed by Pohl, Liu and Thompson (2002). The standard theoretical description is the tunneling two-level system (TTLS) model, introduced independently by Anderson, Halperin and Varma and by Phillips in 1972: atoms or small groups of atoms tunnel between the two minima of random double-well potentials, with broadly distributed energy asymmetries and tunneling amplitudes. The dimensionless attenuation coefficient Q−1Q^{-1} (the inverse acoustic quality factor, defined as λ/(2π2ℓ)\lambda/(2\pi^2\ell) in terms of the phonon wavelength λ\lambda and mean free path ℓ\ell) is predicted by TTLS to be nearly frequency- and temperature-independent in two very different regimes: the high-frequency 'resonance' regime, where kBT≪ℏωk_B T\ll\hbar\omega and phonons are absorbed by resonant excitation of tunneling systems, and the low-frequency 'relaxation' regime, where ωτ≪1\omega\tau\ll 1 (with τ\tau a typical thermal relaxation time) and a phonon is attenuated because it periodically perturbs the energy splitting of thermally occupied tunneling systems, driving their populations out of thermal equilibrium. TTLS also predicts the nonlinear ultrasonic phenomena - saturation of the absorption at high driving intensity, phonon echoes (a spontaneous pulse emitted after two applied pulses, in close analogy with spin echoes), and spectral hole burning (a narrow dip burned into the absorption spectrum by a strong monochromatic pulse) - which are widely regarded as among the strongest evidence for the model. The central unexplained fact is quantitative: the measured Q−1Q^{-1} is not only universal but strikingly small, (3±2)×10−4(3\pm 2)\times 10^{-4}, in nearly all non-metallic glasses, whereas in TTLS it is a product of several independent parameters whose apparent cancellation across materials has no basis in the model, a point long emphasized by Leggett. Approaches that keep tunneling systems but add their mutual interaction attribute the small value to the phonon-mediated elastic dipole-dipole interaction that strain-coupled defects acquire by exchanging virtual phonons (Joffrin and Levelut 1975; Burin and Kagan 1996).

Vural and Leggett (2011) asked how much of the low-temperature phenomenology can be derived without postulating tunneling systems at all. They model the glass as mesoscopic blocks (linear size L∼50L\sim 50 Å) each carrying 'non-phononic' degrees of freedom (the block's internal degrees of freedom other than its ordinary acoustic modes) described by a Hamiltonian H^0\hat H_0 that is arbitrary except that its spectrum must not be harmonic-oscillator-like. Each block couples to the strain field eije_{ij} through its stress tensor T^ij=∂H^/∂eij\hat T_{ij}=\partial\hat H/\partial e_{ij}, and the central object of the theory is the block's stress-stress response function χ(m)(ω)\chi^{(m)}(\omega) (the linear response of the block's stress to an imposed oscillatory strain, expressible through matrix elements of T^ij\hat T_{ij} between exact many-body eigenstates of block mm), whose thermal average gives the absorption Q−1Q^{-1} of a phonon at frequency ω\omega. Because the elastic strain field is shared, two blocks acquire a mutual interaction V^\hat V of dipole form (decaying as 1/r31/r^3). Vural and Leggett then carry out a real-space renormalization procedure in which neighboring blocks are repeatedly combined into superblocks, using second-order perturbation theory in V^\hat V and two uncontrolled assumptions: (A) matrix elements of V^\hat V connecting many-body eigenstates with energy below a high-energy cutoff U0U_0 to states above it are negligible, an assumption carried over to subsequent iterations of the renormalization, and (B) cross terms between the stress matrix elements of the two blocks being combined (of the form ⟨n∣T^ij(1)∣m⟩⟨m∣T^ij(2)∣n⟩\langle n|\hat T^{(1)}_{ij}|m\rangle\langle m|\hat T^{(2)}_{ij}|n\rangle) are on average small compared with same-block terms of the form ∣⟨m∣T^ij(s)∣n⟩∣2|\langle m|\hat T^{(s)}_{ij}|n\rangle|^2. Their results are: (i) the frequency-averaged absorption is quasi-universal, depending only on the ratios cl/ctc_l/c_t of longitudinal to transverse sound velocities and of the corresponding responses (both observed to vary by only about 20% between materials) and logarithmically on a cutoff U0U_0; (ii) with an additional postulated 'random' ansatz for the input form of χ(m)(ω)\chi^{(m)}(\omega), the renormalized frequency dependence is Q−1(ω)∝[ln⁡(U0/ω)]−1Q^{-1}(\omega)\propto[\ln(U_0/\omega)]^{-1}, which at 1 MHz gives 2.7×10−42.7\times 10^{-4}, close to the canonical experimental value; (iii) the predicted temperature dependence of Q−1Q^{-1}, of the logarithmic velocity shift, and a thermal conductivity K∝T2ln⁡(U0/T)K\propto T^2\ln(U_0/T) are all consistent with data. The authors are explicit that these arguments are heuristic.

The same conclusion section states what the paper deliberately does not do. It does not treat the nonlinear ultrasonic phenomena (saturation, echoes, hole burning), which are generally regarded as strong evidence for TTLS. It does not treat the linear response in the low-frequency relaxation regime, which the authors call 'particularly puzzling': the successful TTLS prediction of the observed frequency- and temperature-independence (and even the magnitude) of the relaxation-regime absorption appears to follow from parameter dependencies specific to the TTLS distribution (in TTLS the relaxation and resonant contributions are equal up to a factor of order unity, itself a consequence of the distribution of relaxation rates being proportional to the rate). And it leaves open the self-consistency question: whether a complete renormalization calculation would drive the renormalized ('output') form of χ(m)(ω)\chi^{(m)}(\omega) to the TTLS form - in which case the generic model would inherit all TTLS successes, including the nonlinear ones - or whether the TTLS spectrum must instead be present already at the 'input' block scale, with the renormalization serving only to generate the small universal value of Q−1Q^{-1} (a picture close to that of Burin and Kagan). The heuristic second-order treatment plus ansatz cannot decide this, and the quantitative conclusions are correspondingly sensitive to the assumed input form of χ(m)(ω)\chi^{(m)}(\omega) and to the accuracy of the approximate treatment of the inter-block interaction.

Problem Statement

Determine whether the universal low-temperature acoustic phenomenology of glasses can be derived from the generic-block model - mesoscopic blocks with arbitrary non-harmonic non-phononic spectra, coupled by the 1/r31/r^3 elastic dipole interaction - without postulating tunneling two-level system structure at any scale; that is, complete or refute the generic-block theory. The generic-block account is refuted if the model cannot reproduce the relaxation-regime response or one of the nonlinear phenomena; determining whether it can is the objective. The three sub-questions below are jointly posed and are linked by the common question of whether TTLS structure is emergent:

(1) Relaxation regime. Derive, within the generic model, the linear ultrasonic response in the low-frequency relaxation regime (ωτ≪1\omega\tau\ll 1, kBT≫ℏωk_B T\gg\hbar\omega, where τ\tau is a typical thermal relaxation time of the low-energy excitations) - or prove that the model cannot reproduce it. The target is the experimentally observed near frequency- and temperature-independence of Q−1Q^{-1} in this regime and its magnitude, comparable within a factor of order two to the resonant-regime value of order 3×10−43\times 10^{-4}, obtained without parameter cancellations specific to the TTLS distribution.

(2) Nonlinear response. Derive, within the generic framework, the nonlinear ultrasonic phenomena - saturation of the absorption at high driving intensity, phonon echoes, and spectral hole burning, including the effect of the mutual interaction between blocks - or prove that the model cannot reproduce them. (Echoes have been shown to be generic in multi-level models; single-system saturation exists only in unpublished preprints; hole burning is unaddressed.)

(3) Renormalization self-consistency. Carry the block renormalization beyond the heuristic second-order treatment - relaxing or justifying assumption (A) (that matrix elements of the inter-block interaction connecting many-body eigenstates below the cutoff U0U_0 to states above it are negligible, carried over to subsequent iterations of the renormalization) and assumption (B) (that cross terms between the stress matrix elements of the two combined blocks are on average negligible compared with same-block terms), including the neglected renormalization of the stress matrix elements and a more accurate treatment of the inter-block interaction - to determine whether the renormalized output stress response χ(m)(ω)\chi^{(m)}(\omega) takes the TTLS form or instead preserves the logarithmic structure Q−1(ω)∝[ln⁡(U0/ω)]−1Q^{-1}(\omega)\propto[\ln(U_0/\omega)]^{-1}, where U0U_0 is the high-energy cutoff of the block spectrum. This settles at once whether the theory's conclusions depend essentially on the assumed input form of χ(m)(ω)\chi^{(m)}(\omega), and whether the generic-block picture is self-consistent with the phenomenology it was designed to explain.

References

  1. Dervis C. Vural and Anthony J. Leggett, "Universal sound absorption in amorphous solids: A theory of elastically coupled generic blocks", Journal of Non-Crystalline Solids 357(19-20), 3528-3537 (2011). DOI: 10.1016/j.jnoncrysol.2011.06.035; arXiv:1103.5530. URL: https://doi.org/10.1016/j.jnoncrysol.2011.06.035
  2. Anthony J. Leggett and Dervis C. Vural, "'Tunneling Two-Level Systems' Model of the Low-Temperature Properties of Glasses: Are 'Smoking-Gun' Tests Possible?", The Journal of Physical Chemistry B 117, 12966-12971 (2013). DOI: 10.1021/jp402222g; arXiv:1310.3387. URL: https://doi.org/10.1021/jp402222g
  3. Di Zhou and Anthony J. Leggett, "Random Nonlinear Infinite-Level-System Model for Amorphous Solid Phonon Echo and Saturation Phenomena", preprint, arXiv:1510.05538 (submitted October 2015; latest version v2 posted April 2016, manuscript dated March 2022). URL: https://arxiv.org/abs/1510.05538 (unpublished preprint)
  4. Di Zhou and Anthony J. Leggett, "A Generic Microscopic Theory for the Universality of TTLS Model Meissner-Berret Ratio in Low-Temperature Glasses", preprint, arXiv:1510.05528 (submitted October 2015; latest version v3 posted July 2016, manuscript dated June 2021). URL: https://arxiv.org/abs/1510.05528 (unpublished preprint by Zhou and Leggett; the published single-author journal account of the Meissner-Berret ratio result is the 2019 JPCM paper listed separately)
  5. Di Zhou, "Universal ratio of TTLS-phonon coupling constants in low-temperature amorphous solids", Journal of Physics: Condensed Matter 32, 055704 (2019; published online November 2019, print issue January 2020). DOI: 10.1088/1361-648X/ab4e5b. URL: https://doi.org/10.1088/1361-648X/ab4e5b
  6. Di Zhou, "Phonon echo from multilevel systems and many-body interactions in low-temperature glasses", Physical Review E 114(2), 025401 (2026). DOI: 10.1103/fb15-y2ls; arXiv:2602.14435. URL: https://doi.org/10.1103/fb15-y2ls
  7. Herve M. Carruzzo and Clare C. Yu, "Why Phonon Scattering in Glasses is Universally Small at Low Temperatures", Physical Review Letters 124, 075902 (2020). DOI: 10.1103/PhysRevLett.124.075902; arXiv:1909.09258. URL: https://doi.org/10.1103/PhysRevLett.124.075902
  8. Clare C. Yu and Hervé M. Carruzzo, "Two-Level Systems and the Tunneling Model: A Critical View", in Low-Temperature Thermal and Vibrational Properties of Disordered Solids (World Scientific, 2022). DOI: 10.1142/9781800612587_0004; arXiv:2101.02787. URL: https://doi.org/10.1142/9781800612587_0004
  9. M. Schechter, P. Nalbach and A. L. Burin, "Nonuniversality and strongly interacting two-level systems in glasses at low temperatures", New Journal of Physics 20, 063048 (2018). DOI: 10.1088/1367-2630/aac930; arXiv:1802.07283. URL: https://doi.org/10.1088/1367-2630/aac930
  10. Mihail Turlakov, "A fermion phenomenology of low-temperature strongly-noncrystalline solids", Journal of Non-Crystalline Solids 588, 121632 (2022). DOI: 10.1016/j.jnoncrysol.2022.121632; arXiv:2111.03578. URL: https://doi.org/10.1016/j.jnoncrysol.2022.121632
  11. Pragya Shukla, "Universality of ultrasonic attenuation coefficient of amorphous systems at low temperatures", Scientific Reports 12, 2662 (2022). DOI: 10.1038/s41598-022-06589-7. URL: https://doi.org/10.1038/s41598-022-06589-7
  12. P. W. Anderson, B. I. Halperin and C. M. Varma, "Anomalous low-temperature thermal properties of glasses and spin glasses", Philosophical Magazine 25, 1-9 (1972). DOI: 10.1080/14786437208229210. URL: https://doi.org/10.1080/14786437208229210
  13. W. A. Phillips, "Tunneling states in amorphous solids", Journal of Low Temperature Physics 7, 351-360 (1972). DOI: 10.1007/BF00660072. URL: https://doi.org/10.1007/BF00660072
  14. R. C. Zeller and R. O. Pohl, "Thermal Conductivity and Specific Heat of Noncrystalline Solids", Physical Review B 4, 2029-2041 (1971). DOI: 10.1103/PhysRevB.4.2029. URL: https://doi.org/10.1103/PhysRevB.4.2029
  15. Robert O. Pohl, Xiao Liu and EunJoo Thompson, "Low-temperature thermal conductivity and acoustic attenuation in amorphous solids", Reviews of Modern Physics 74, 991-1013 (2002). DOI: 10.1103/RevModPhys.74.991. URL: https://doi.org/10.1103/RevModPhys.74.991
  16. A. L. Burin and Yu. Kagan, "On the nature of the universal properties of amorphous solids", Physics Letters A 215, 191-196 (1996). DOI: 10.1016/0375-9601(96)00191-0. URL: https://doi.org/10.1016/0375-9601(96)00191-0
  17. J. Joffrin and A. Levelut, "Virtual phonon exchange in glasses", Journal de Physique 36, 811-822 (1975). DOI: 10.1051/jphys:01975003609081100. URL: https://doi.org/10.1051/jphys:01975003609081100