Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

§III.B.2, eqs. (36)–(42), p. 6 — under pure dephasing, coherence can be held constant exactly on [0, t_b), t_b = v_z(0)²/(2γc), by fields that diverge at t_b

Proved
CoherenceTracking.Dephasing.coherence_tracking_breakdown

by mikedeng1 · Oct 22, 2026 · Mathlib 0df444a (Lean v4.33.1)

decoherencefinite-time-singularityp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1quantum-controltracking-control

Consider a qubit with density matrix ρ\rhoρ evolving under the master equation

∂ρ∂t=−iℏ[H(t),ρ]+γ2(σzρσz−ρ),H(t)=ℏ2(ω0(t)σz+ω1(t)σx−ω2(t)σy),\frac{\partial\rho}{\partial t}=-\frac{i}{\hbar}[H(t),\rho]+\frac{\gamma}{2}(\sigma_z\rho\sigma_z-\rho),\qquad H(t)=\frac{\hbar}{2}\bigl(\omega_0(t)\sigma_z+\omega_1(t)\sigma_x-\omega_2(t)\sigma_y\bigr),∂t∂ρ​=−ℏi​[H(t),ρ]+2γ​(σz​ρσz​−ρ),H(t)=2ℏ​(ω0​(t)σz​+ω1​(t)σx​−ω2​(t)σy​),

with ℏ>0\hbar>0ℏ>0 and dephasing rate γ>0\gamma>0γ>0. Let ρ0\rho_0ρ0​ be a qubit state with Bloch vector v⃗(0)=(vx(0),vy(0),vz(0))\vec v(0)=(v_x(0),v_y(0),v_z(0))v(0)=(vx​(0),vy​(0),vz​(0)), vα=Tr⁡(ρσα)v_\alpha=\operatorname{Tr}(\rho\sigma_\alpha)vα​=Tr(ρσα​), and coherence c=vx(0)2+vy(0)2>0c=v_x(0)^2+v_y(0)^2>0c=vx​(0)2+vy​(0)2>0. Let

tb=vz(0)22γc.t_b=\frac{v_z(0)^2}{2\gamma c}.tb​=2γcvz​(0)2​.

Then:

  1. (No-go.) For every choice of controls ω0,ω1,ω2\omega_0,\omega_1,\omega_2ω0​,ω1​,ω2​ and every solution ρ(t)\rho(t)ρ(t) on [0,T)[0,T)[0,T) with ρ(0)=ρ0\rho(0)=\rho_0ρ(0)=ρ0​ whose coherence stays equal to ccc, one has vz(t)2=vz(0)2−2γc tv_z(t)^2=v_z(0)^2-2\gamma c\,tvz​(t)2=vz​(0)2−2γct for all t∈[0,T)t\in[0,T)t∈[0,T), and T≤tbT\le t_bT≤tb​.
  2. (Achievability.) If vz(0)≠0v_z(0)\neq0vz​(0)=0, then for every function ω0\omega_0ω0​ the state ρ(t)=12(I+v⃗(t)⋅σ⃗)\rho(t)=\tfrac12(I+\vec v(t)\cdot\vec\sigma)ρ(t)=21​(I+v(t)⋅σ) with v⃗(t)=(vx(0),vy(0),sign⁡(vz(0))vz(0)2−2γct)\vec v(t)=(v_x(0),v_y(0),\operatorname{sign}(v_z(0))\sqrt{v_z(0)^2-2\gamma ct})v(t)=(vx​(0),vy​(0),sign(vz​(0))vz​(0)2−2γct​) starts at ρ0\rho_0ρ0​, is a qubit state for every t∈[0,tb)t\in[0,t_b)t∈[0,tb​), keeps vxv_xvx​, vyv_yvy​ and hence the coherence ccc fixed, and solves the master equation at every t∈[0,tb)t\in[0,t_b)t∈[0,tb​) with the fields ω1,ω2\omega_1,\omega_2ω1​,ω2​ of eqs. (41)–(42).
  3. (Divergence.) If vz(0)≠0v_z(0)\neq0vz​(0)=0, then for every ω0\omega_0ω0​ the fields of (41)–(42) satisfy ω1(t)2+ω2(t)2→+∞\omega_1(t)^2+\omega_2(t)^2\to+\inftyω1​(t)2+ω2​(t)2→+∞ as t→tb−t\to t_b^-t→tb−​.

This is the main result of the paper: "At the breakdown time tbt_btb​ the control fields diverge … Thus, decay of the coherence can be prevented for t<tbt<t_bt<tb​", and not beyond. If vz(0)=0v_z(0)=0vz​(0)=0, then tb=0t_b=0tb​=0, and item 1 says that the coherence cannot be held on any interval of positive length.

Formalization Note γ>0\gamma>0γ>0 and c>0c>0c>0 are added because tbt_btb​ divides by 2γc2\gamma c2γc. A solution on [0,T)[0,T)[0,T) is continuous on [0,T)[0,T)[0,T) and satisfies the master equation at every t∈(0,T)t\in(0,T)t∈(0,T); the controls are arbitrary real functions, and item 1 does not assume vxv_xvx​, vyv_yvy​ frozen. The sign (±)(\pm)(±) of (37) is sign⁡vz(0)\operatorname{sign}v_z(0)signvz​(0). The Bloch vector is vα=Re⁡Tr⁡(ρσα)v_\alpha=\operatorname{Re}\operatorname{Tr}(\rho\sigma_\alpha)vα​=ReTr(ρσα​), following eq. (9) rather than the misprinted factor 12\tfrac1221​ of eq. (8).

Preamble
import Mathlib
import Definitions.Def_CoherenceTracking_Dephasing_Model
Formal statement
namespace CoherenceTracking.Dephasing
theorem coherence_tracking_breakdown (hbar γ : ℝ) (hhbar : 0 < hbar) (hγ : 0 < γ)
    (ρ0 : Qmat) (hρ0 : IsQubitState ρ0) (hc : 0 < coherence ρ0) :
    -- (i) no-go, (36)–(38): any controls, any constant-coherence solution of (32)
    (∀ (T : ℝ) (ω : ℝ → Fin 3 → ℝ) (ρ : ℝ → Qmat),
        ρ 0 = ρ0 → SolvesMaster hbar γ ω ρ T →
        (∀ t ∈ Set.Ico 0 T, coherence (ρ t) = coherence ρ0) →
        (∀ t ∈ Set.Ico 0 T,
            blochVec (ρ t) 2 ^ 2 = blochVec ρ0 2 ^ 2 - 2 * γ * coherence ρ0 * t) ∧
        T ≤ breakdownTime γ (blochVec ρ0)) ∧
    -- (ii) achievability, (37), (41)–(42): the tracked state solves (32) and keeps c on [0, t_b)
    (blochVec ρ0 2 ≠ 0 → ∀ ω0 : ℝ → ℝ,
        trackedState γ (blochVec ρ0) 0 = ρ0 ∧
        ∀ t ∈ Set.Ico 0 (breakdownTime γ (blochVec ρ0)),
          IsQubitState (trackedState γ (blochVec ρ0) t) ∧
          coherence (trackedState γ (blochVec ρ0) t) = coherence ρ0 ∧
          blochVec (trackedState γ (blochVec ρ0) t) 0 = blochVec ρ0 0 ∧
          blochVec (trackedState γ (blochVec ρ0) t) 1 = blochVec ρ0 1 ∧
          HasDerivAt (trackedState γ (blochVec ρ0))
            (masterRHS hbar γ (trackingFields γ (blochVec ρ0) ω0 t)
              (trackedState γ (blochVec ρ0) t)) t) ∧
    -- (iii) divergence of the control fields at t_b
    (blochVec ρ0 2 ≠ 0 → ∀ ω0 : ℝ → ℝ,
        Filter.Tendsto
          (fun t => trackingFields γ (blochVec ρ0) ω0 t 1 ^ 2 +
            trackingFields γ (blochVec ρ0) ω0 t 2 ^ 2)
          (nhdsWithin (breakdownTime γ (blochVec ρ0)) (Set.Iio (breakdownTime γ (blochVec ρ0))))
          Filter.atTop) := by sorry
end CoherenceTracking.Dephasing
Source
Lidar & Schneider, Stabilizing qubit coherence via tracking-control, arXiv:quant-ph/0410048v1, p. 6, §III.B.2, eqs. (36)–(42) and the sentence 'At the breakdown time t_b the control fields diverge'; abstract, p. 1
Human review
  • Endorsed by Shuze Chen · Oct 22, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 22, 2026

    Confirmed by the mission captain (proposal self-audit).

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me