# D17 — E4-H006: The CDW amplitude mode is least damped exactly where its phason gap reaches half its frequency

**Batch:** 1 · **Selected rank:** 17/30 · **Tractability:** 8/10 · **Significance:** 8/10

## 1. Original hypothesis and refined claim

The original proposes that amplitude (Higgs) mode damping in a charge-density wave is non-monotonic in impurity pinning: two-phason decay closes once the phason gap Δ_p exceeds ω_H/2, so raising pinning first lowers the linewidth, and impurity scattering eventually raises it. The physics survives; three details do not.

The entry writes the two-phason term as A(ω_H² − 4Δ_p²)^½/ω_H, a three-dimensional two-body density of states; in the quasi-one-dimensional or layered CDWs it actually treats, that density of states is a step at threshold. It also asserts a "frequency dip through level repulsion", but for two discrete coupled modes the upper branch is pushed up; what the model requires is the Kramers–Kronig partner of the two-phason self-energy, which depresses ω_H below threshold and releases it above, vanishing at threshold. And it moves the minimum to lower concentration with temperature because "the phason gap closes", which alone does the opposite: if Δ_p falls at fixed ω_H, more pinning is needed to reach Δ_p = ω_H/2.

**Refined claim.** With u ≡ 4Δ_p²/ω_H² and Δ_p² = c n_i,

  Γ_H = Γ_0 + A(1 − u)^{1/2} Θ(1 − u) + (B/c)Δ_p²,

the linewidth has a one-sided cusp minimum located exactly at Δ_p = ω_H/2, of depth A(1 − 1/(2ε)) with ε ≡ 2Ac/(Bω_H²). The minimum exists and is global only for ε > 1/2; the approach is monotone, with no intervening local maximum, only for ε > 1. The observable is a non-analytic kink whose left slope diverges as −(ω_H − 2Δ_p)^{−1/2}.

## 2. Refined scientific description

Amplitude and phase are the real and imaginary parts of one complex order parameter, so the quartic Ginzburg–Landau term that sets the amplitude-mode frequency also fixes the amplitude → phason-pair vertex. The decay is a phase-space gate and pinning is the handle. Raising pinning is normally read as adding disorder and hence width; here it removes a channel. The relation is anti-Matthiessen, and pinning acts as a coherence knob rather than a loss term.

## 3. Verified supporting literature and prior-art assessment

[EST] A CDW has an amplitude and a phase mode, and impurity pinning gaps the phase mode. [EST] Two-particle kinematics forbid a decay whose products are individually heavier than the parent, blocking A → 2p for 2Δ_p > ω_H. [EST] The mean-field amplitude mode lies near 2Δ, degenerate with the quasiparticle continuum edge, so Landau damping is generic and is the default explanation of its width. [EST] The two-body joint density of states of a d-dimensional gapped dispersion goes as (ω − 2Δ_p)^{(d−2)/2}: square root in three dimensions, a step in two and one.

[REC] Terahertz and femtosecond driving and detection of CDW amplitude modes in several compounds. [REC] Pressure-tuned commensurate–incommensurate transitions at which the pinning frequency collapses. [REC] Depinning threshold fields from transport, tracking the pinning frequency.

[UNV] That measured amplitude-mode widths are dominated by impurities alone — the entry's target. [UNV] The A → 2p vertex magnitude in any compound. [UNV] That the channel is symmetry-allowed in the compounds measured so far.

**Prior art.** The decay into Goldstone pairs is textbook; the new content is pinning as the gate and the cusp shape. Nearest corpus collisions: E1-H126 (phasons and negative thermal expansion) and E3-H115 (central peak); neither gates a decay channel.

## 4. Proposed physical or mathematical model

Ginzburg–Landau functional for ψ = (ψ_0 + δρ)e^{iθ}: a|ψ|² + (b/2)|ψ|⁴ + K|∇ψ|², plus a pinning term gapping θ. Expanding |ψ|²(∇θ)² gives the vertex g = 2Kψ_0. The zero-momentum self-energy is Σ(ω) = g² ∫ d^d q/(2π)^d [ω − ω_p(q) − ω_p(−q) + i0⁺]^{−1} with ω_p(q) = (Δ_p² + v_p²q²)^½. In three dimensions Im Σ ∝ ω(ω²/4 − Δ_p²)^{1/2}θ(ω − 2Δ_p), reproducing the entry's square root; in two dimensions it is a step, in one a constant. Adding B n_i with Δ_p² = c n_i gives section 1, and every qualitative outcome is a statement about the single group ε.

## 5. Variables, assumptions and boundary conditions

ω_H amplitude-mode frequency at q = 0; Γ_H its half width at half maximum; Δ_p phason gap; v_p phason velocity; n_i pinning-centre density; c the proportionality Δ_p² = c n_i; A two-phason coupling (frequency); B Matthiessen slope per centre; Γ_0 residual width from all other channels; ε the group above; K phase stiffness; T_CDW transition temperature.

Assumptions: one complex order parameter, single-q, so the vertex is fixed; weak pinning Δ_p ≪ Δ; perturbative self-energy A ≪ ω_H; Matthiessen additivity; underdamped mode ω_H > Γ_H; T < 0.9 T_CDW. Small parameters: Δ_p/ω_H ≤ 1/2 and A/ω_H. Domain: 0.2 < 2Δ_p/ω_H < 1.5. Δ_p depends on temperature as well as pinning, so the crossing is a ratio and both must be tracked.

## 6. Quantitative predictions

With ω_H/2π = 1 THz the cusp sits at Δ_p/2π = 0.5 THz exactly. Depth A(1 − 1/(2ε)); for A/2π = 0.1 THz — an estimate from the quartic-to-quadratic GL ratio — and ε = 2, the depth is 0.075 THz ≈ 2.5 cm⁻¹, a 7.5% effect. The left slope diverges, so the last tenth of the approach carries half the change.

Existence inequality: no minimum unless B < 4Ac/ω_H², i.e. ε > 1/2. This is the sharpest prediction because it says which materials cannot show the effect: any CDW whose broadening per unit pinning shift exceeds 4Ac/ω_H² stays monotone in n_i forever.

Onset shape: exponent 1/2 for a three-dimensional phason; a discontinuous jump of size A for a layered or chain CDW.

Frequency companion: ω_H is depressed below threshold and released above, so it rises by of order A across the crossing — opposite to the entry's dip.

## 7. Comparison against conventional explanations

**Landau damping into the quasiparticle continuum.** The amplitude mode sits at ≈2Δ, degenerate with the pair-breaking edge, so single-particle damping is generic and needs no phasons. *Separating observable:* Landau damping tracks the quasiparticle population and hence σ_1(ω_H) measured on the same sample at the same temperature, while the two-phason channel does not. Hold σ_1(ω_H) fixed to ±10% and change Δ_p by 30% with pressure: Landau damping predicts Γ unchanged, this hypothesis predicts 30–60%.

**Matthiessen impurity broadening.** Γ = Γ_0 + B n_i, monotone, width a materials constant. *Separating observable:* the slope reversal and, decisively, the non-analyticity. A monotone model cannot produce a kink, and a cusp is distinguishable from a quadratic minimum with six points per decade.

**Inhomogeneous broadening.** A distribution of pinning frequencies from strain or pressure gradients smears any sharp feature. *Separating observable:* that distribution's width, from the dielectric loss peak. If it exceeds A the cusp is unobservable and the test is void — a precondition, not a rival.

## 8. Specific experimental or computational method

Broadband terahertz time-domain transmission spectroscopy in a cryogenic piston pressure cell, on a crystal thin enough that the lowest Fabry–Perot fringe lies above the mode. Apparatus: THz-TDS with 0.2–3 THz bandwidth and a 60 ps scan; helium-transmitting pressure cell; 5–120 K cryostat. Dataset: complex conductivity at 40 pressures × 6 temperatures, plus an independent growth. Procedure: fit σ(ω) with a Drude term, a Fano line and a mid-infrared background, holding the Fano asymmetry q fixed at its 10 K ambient value — q is the nuisance parameter, and floating it transfers width into asymmetry. Measure Δ_p independently from the dielectric loss peak and the depinning threshold field E_T ∝ Δ_p². Control arm: the same series on a crystal with the CDW suppressed, where Γ must vary smoothly. Fixed throughout: incident field at the lowest fluence giving signal-to-noise ratio above 100.

## 9. Necessary controls

Fluence series at fixed T, verifying that Γ is flat below a threshold fluence rather than measuring laser heating. Polarisation along the chain axis with a crossed run bounding anisotropy leakage. Sample thermometry, not cell-body thermometry. Hydrostaticity checked by the ruby linewidth. The suppressed-CDW arm. A zero-pressure repeat at the end, catching alignment drift that mimics a width change.

## 10. Sources of measurement or numerical error

Dominant: the finite scan window. At 60 ps the transform limit is ≈17 GHz, but a fitted damped oscillator recovers Γ to about ±5% at signal-to-noise ratio 100, i.e. ±50 GHz on a 1 THz mode — comparable to the predicted 75 GHz depth. The test therefore rests on shape, a factor 3–5 slope contrast across threshold, not on depth alone. Second: Fano asymmetry, a drift of 0.1 in q shifting fitted Γ by roughly 10%. Third: pressure moves ω_H itself, shifting the threshold, so both must be fitted jointly.

## 11. Falsification criteria

1. No non-analytic feature in Γ(Δ_p) within ±10% of Δ_p = ω_H/2 over 0.2 < 2Δ_p/ω_H < 1.5, the pinning-frequency distribution being narrower than A.
2. Γ monotone increasing in Δ_p across the window: the ε > 1/2 inequality is violated and the mechanism is excluded for that compound.
3. Γ unchanged to within 10% when Δ_p changes by 30% at fixed σ_1(ω_H), T and q — this kills the channel independently of any lineshape model.
4. Onset exponent inconsistent with sample dimensionality — 1/2 for a three-dimensional phason, a step for a layered one. An exponent near 1 excludes the vertex.
5. The suppressed-CDW control shows the same feature at the same reduced conditions.
6. No Kramers–Kronig-consistent ω_H anomaly at the same Δ_p, given a measured Γ change of at least 5%.
7. Dip depth failing to scale with the square of the order-parameter stiffness K across two materials, which excludes the GL vertex even if a dip exists.

## 12. Expected scientific contribution

Converts "amplitude-mode width is a materials constant" into spectroscopy of one decay channel, and supplies a rare anti-Matthiessen relation in which disorder suppresses rather than adds loss. Identifies pinning as a coherence knob for CDW oscillators and light-induced melting. If the effect is absent, the same data bound the A → 2p vertex and hence the quartic GL coefficient, a quantity rarely measured directly.

## 13. Resource and feasibility assessment

Estimates, not quotes: one THz-TDS system with a cryogenic pressure cell; 6–9 months of one researcher; two crystal growths; ~240 spectra at 20 minutes each, i.e. ~80 hours of spectrometer time. Verified: the fit is a two-oscillator least squares with fixed q, and the Kramers–Kronig step is a direct Hilbert transform; no new software. Dependencies: spectrometer, cryostat, cell windows, crystal growth, and — the true bottleneck — a depinning-threshold transport measurement on the same crystal measured optically.

## 14. Reproducibility plan

(1) Characterise T_CDW by transport and by amplitude-mode softening, and E_T(T) at ambient. (2) At 10 K record reference and sample traces at 40 pressures with fixed window and fluence. (3) Fit each spectrum with the three-component model, q fixed, Levenberg–Marquardt, 200 random starts; keep fits with residual autocorrelation below 0.1. (4) Convert E_T to Δ_p via E_T ∝ Δ_p². (5) Fit Γ(Δ_p) to the three-regime form with free A, B, c, Γ_0; bootstrap 1000 resamples, seed 20261101; report threshold, depth and both slopes. Tolerances: threshold within 15% of ω_H/2 from the jointly fitted ω_H; left slope exceeding the right by a factor 3. Replication: an independently grown crystal reproduces the threshold within 10% and the depth within a factor of two, with all seven criteria of section 11 evaluated.

## 15. Principal remaining uncertainties

The likeliest killer is that the amplitude mode is overdamped in precisely the dirty samples needed to raise Δ_p: driving ω_H/Δ_p to 2 by doping may push ω_Hτ below unity and leave no oscillator. Second, Landau damping may be so large that a 7% dip sits on a 40% background; the shape argument mitigates but does not remove this. Third, reach: in canonical CDWs the pinning frequency is one to two orders below the amplitude-mode frequency, so the crossing may need pinning that suppresses the CDW. Tuning Δ_p by pressure through a commensurate–incommensurate transition rather than by doping is a bet on material availability, not on the physics.