Color Decoherence: QCD, Quantum Codes & Materials

• Color decoherence is the loss of coherent phase relations in color-charged systems, leading to independent radiation in QCD and altered error properties in quantum codes.
• Analytical methods such as Wilson line correlators and effective field theory formulations quantify decoherence times and dynamic angular scales in jet evolution.
• In solid-state platforms like diamond color centers, environmental interactions (nuclear spins or phonons) narrow resonances and reduce coherence times, affecting sensing and networking.

Searching arXiv for recent and foundational papers on “color decoherence” across the relevant domains. Color decoherence denotes the suppression of coherent phase relations associated with a quantity labeled “color,” but the term is not monosemous. In high-energy QCD it refers to the progressive destruction of interference among color-correlated partons in a medium, most cleanly in a $q\bar q$ antenna traversing a deconfined plasma, where multiple scattering suppresses interference and can drive the system to independent radiation and memory loss of the initial color state (Mehtar-Tani et al., 2011). In quantum information it denotes decoherence applied to the color code, where structured XX noise on red links produces an intrinsic mixed-state topological order whose modular quotient is closely related to a single toric code (Kataoka et al., 24 Apr 2026). In solid-state quantum platforms it refers to decoherence processes in diamond color centers, where nuclear-spin and phonon environments broaden resonances, shorten coherence times, and constrain sensing and networking protocols (Parker et al., 2015).

1. Vacuum coherence and its loss in a QCD medium

In perturbative QCD, the canonical setting is a soft gluon emitted from a quark–antiquark antenna. In vacuum, the radiation field is the coherent superposition of emissions from the two color charges, and interference enforces angular ordering: successive soft emissions are confined to angles smaller than the opening angle of the antenna, $\theta < \theta_{12}$. For a color singlet, destructive interference suppresses radiation at angles larger than $\theta_{12}$; for an initially colored antenna, coherent radiation off the total charge dominates at large angles. Schematically, the vacuum spectrum may be written as

$$\omega\, dN_{\rm vac} \propto \alpha_s \left[R_q + R_{\bar q} - 2J\right],$$

with $J$ the interference term that removes radiation outside the antenna cone (Mehtar-Tani et al., 2011).

In a deconfined medium, multiple scattering with medium color charges randomizes the color phases acquired by the quark and antiquark along their eikonal trajectories. The survival probability of coherence is encoded in a Wilson-line correlator,

$$S(L,r_\perp)=\exp\!\left(-\frac{1}{12}\hat q\,L\,r_\perp^2\right),\qquad \Delta_{\rm med}=1-S,$$

with $r_\perp \simeq \theta_{12}L$, equivalently

$$S(L)=\exp\!\left[-\frac{1}{12}\hat q\,L^3\,\theta_{12}^2\right],\qquad \Delta_{\rm med}=1-\exp\!\left[-\frac{1}{12}\hat q\,L^3\,\theta_{12}^2\right].$$

As $\Delta_{\rm med}$ grows from $0$ to $\theta < \theta_{12}$0, the coherent term fades, the soft gluon no longer “sees” the total charge, and radiation proceeds independently from the two legs (Mehtar-Tani et al., 2011).

The corresponding decoherence time follows from setting the exponent to unity,

$\theta < \theta_{12}$1

If $\theta < \theta_{12}$2, the gluon forms before decoherence and vacuum-like angular ordering persists. If $\theta < \theta_{12}$3, coherence is lost before or during formation, emissions proceed independently, and soft radiation can populate angles larger than $\theta < \theta_{12}$4. In this regime the literature describes anti-angular ordering, enhanced large-angle radiation, jet broadening, and energy transport to large angles.

The opaque limit is the limiting case of complete color-memory loss. For large $\theta < \theta_{12}$5 and/or large antenna size, $\theta < \theta_{12}$6, interference vanishes, and

$\theta < \theta_{12}$7

In this limit, “radiation off the total charge vanishes”: antennas initialized in different color representations radiate identically. This is the sense in which the medium induces memory loss of initial color correlations (Mehtar-Tani et al., 2011).

2. Dynamic critical angles and effective-theory reformulations

A later development generalized the standard antenna picture by allowing the $\theta < \theta_{12}$8 antenna itself to be formed in the medium rather than inserted as a pre-existing object. In that case, medium interactions during the antenna-formation interval modify both the total emission rate and the interference pattern. The resulting decoherence factor depends not only on medium properties but also on antenna kinematics, and the notion of a universal critical angle is promoted to a dynamic quantity $\theta < \theta_{12}$9, different for every splitting. Depending on the region of parameter space, color decoherence can either be delayed or accelerated with respect to previous estimates (Abreu et al., 2024).

In the corresponding formulation, the medium-modified emission probability is multiplied by a factor $\theta_{12}$0, while the interference term is weighted by a generalized $\theta_{12}$1. The antenna formation time is

$\theta_{12}$2

so the onset of decoherence is explicitly splitting dependent. A plausible implication is that highly collinear splittings can remain effectively coherent longer than the standard medium-only estimate would suggest.

An EFT formulation recasts the same physics in a factorized language for inclusive jet production. The hierarchy is $\theta_{12}$3, with

$\theta_{12}$4

and the emergent angular scale is

$\theta_{12}$5

For a jet of radius $\theta_{12}$6, both the LPM effect and color decoherence are controlled by the same dimensionless parameter

$\theta_{12}$7

When $\theta_{12}$8, the medium cannot resolve the substructure and the jet behaves as a single color source; when $\theta_{12}$9, the medium resolves the prongs and interference terms vanish after path-length averaging (Vaidya, 27 Feb 2026).

This EFT also makes the renormalization-group structure nontrivial. Multi-sub-jet operators mix under $$\omega\, dN_{\rm vac} \propto \alpha_s \left[R_q + R_{\bar q} - 2J\right],$$0-evolution and rapidity evolution, and the coherent-to-decoherent crossover emerges as an operator-level feature rather than solely as a kinematic rule. This suggests that color decoherence is not only a geometric resolution effect but also a factorization-scale-sensitive component of in-medium jet evolution.

3. Jet-quenching phenomenology, color flow, and small systems

A phenomenological framework combining vacuum-like evolution and medium-induced radiation makes color decoherence quantitative at the level of resolved subjet multiplicity. In that construction, a jet produced at a hard scale $$\omega\, dN_{\rm vac} \propto \alpha_s \left[R_q + R_{\bar q} - 2J\right],$$1 first undergoes vacuum-like evolution in the double logarithmic approximation down to an infrared scale $$\omega\, dN_{\rm vac} \propto \alpha_s \left[R_q + R_{\bar q} - 2J\right],$$2, and each subjet at $$\omega\, dN_{\rm vac} \propto \alpha_s \left[R_q + R_{\bar q} - 2J\right],$$3 then loses energy independently according to BDMPS-Z quenching weights. In $$\omega\, dN_{\rm vac} \propto \alpha_s \left[R_q + R_{\bar q} - 2J\right],$$4–$$\omega\, dN_{\rm vac} \propto \alpha_s \left[R_q + R_{\bar q} - 2J\right],$$5 PbPb collisions at $$\omega\, dN_{\rm vac} \propto \alpha_s \left[R_q + R_{\bar q} - 2J\right],$$6, a $$\omega\, dN_{\rm vac} \propto \alpha_s \left[R_q + R_{\bar q} - 2J\right],$$7 analysis yields

$$\omega\, dN_{\rm vac} \propto \alpha_s \left[R_q + R_{\bar q} - 2J\right],$$8

with a mild degeneracy across $$\omega\, dN_{\rm vac} \propto \alpha_s \left[R_q + R_{\bar q} - 2J\right],$$9 and $J$0. The resulting $J$1 reproduces the ATLAS measurements for $J$2 and $J$3 jets, and the cone-size ordering is

$J$4

reflecting stronger decoherent energy loss for wider jets. The same framework separates coherent single-subjet configurations from decoherent multiple-subjet configurations, with the latter dominating at high $J$5 and at lower $J$6 for gluon jets than for quark jets (Duan et al., 23 Mar 2026).

Color decoherence also appears in color-differential opacity calculations of jet–medium interactions. Medium-induced color flow can color-decohere radiated gluons from the leading fragment, so that hadronization no longer proceeds through a single vacuum-like string. In the Lund picture, this can contribute to the quenching of leading hadron spectra and can increase strongly the yield of soft hadronic fragments from a jet, while the distribution of more energetic hadrons follows naturally the shape of a vacuum-like fragmentation pattern of lower total energy (Beraudo et al., 2012).

In Monte Carlo studies of small collision systems, JEWEL 2.6.0 implements color coherence by rejecting medium scatterings off single partons when the medium cannot resolve the dipole. Turning color coherence on raises jet $J$7 by about $J$8 in the O+O-like geometry studied with TGlauberMC, and by about $J$9 in the $$S(L,r_\perp)=\exp\!\left(-\frac{1}{12}\hat q\,L\,r_\perp^2\right),\qquad \Delta_{\rm med}=1-S,$$0–$$S(L,r_\perp)=\exp\!\left(-\frac{1}{12}\hat q\,L\,r_\perp^2\right),\qquad \Delta_{\rm med}=1-S,$$1 “central” class. However, after retuning the incoherent initial temperature so that its $$S(L,r_\perp)=\exp\!\left(-\frac{1}{12}\hat q\,L\,r_\perp^2\right),\qquad \Delta_{\rm med}=1-S,$$2 matches the coherent case, the jet $$S(L,r_\perp)=\exp\!\left(-\frac{1}{12}\hat q\,L\,r_\perp^2\right),\qquad \Delta_{\rm med}=1-S,$$3 is the same as the coherent case, while hadron $$S(L,r_\perp)=\exp\!\left(-\frac{1}{12}\hat q\,L\,r_\perp^2\right),\qquad \Delta_{\rm med}=1-S,$$4 is somewhat larger with coherence turned on. At fixed charged-particle multiplicity, O+O and Pb+Pb give the same hadron and jet $$S(L,r_\perp)=\exp\!\left(-\frac{1}{12}\hat q\,L\,r_\perp^2\right),\qquad \Delta_{\rm med}=1-S,$$5 within uncertainties despite their different shapes (Kolbé et al., 20 Oct 2025).

A conceptually distinct formulation models the QGP as a composite CPTP quantum channel,

$$S(L,r_\perp)=\exp\!\left(-\frac{1}{12}\hat q\,L\,r_\perp^2\right),\qquad \Delta_{\rm med}=1-S,$$6

combining amplitude damping, $$S(L,r_\perp)=\exp\!\left(-\frac{1}{12}\hat q\,L\,r_\perp^2\right),\qquad \Delta_{\rm med}=1-S,$$7 depolarizing noise, and thermal singlet projection. In that picture, color decoherence is the suppression of off-diagonal color coherences under $$S(L,r_\perp)=\exp\!\left(-\frac{1}{12}\hat q\,L\,r_\perp^2\right),\qquad \Delta_{\rm med}=1-S,$$8, and the full channel produces monotonic entanglement degradation and purity loss (Twagirayezu, 2 Jul 2025).

4. Heavy-quark color decoherence as open-system dynamics

For heavy quarks in a high-temperature quark–gluon plasma, color decoherence takes the form of environment-induced loss of quantum coherence caused by coupling between the heavy quark’s $$S(L,r_\perp)=\exp\!\left(-\frac{1}{12}\hat q\,L\,r_\perp^2\right),\qquad \Delta_{\rm med}=1-S,$$9 color and thermal chromo-electric fluctuations. The reduced density matrix obeys a master equation in which the kernels

$r_\perp \simeq \theta_{12}L$0

are determined by the medium correlator

$r_\perp \simeq \theta_{12}L$1

A stochastic representation makes the coupling explicit: random color fields rotate the heavy-quark color and generate color-dependent momentum kicks, producing a macroscopic superposition of momentum states entangled with color (Akamatsu, 2015).

Tracing over color yields a color-averaged master equation with a Caldeira–Leggett-like decoherence term. For $r_\perp \simeq \theta_{12}L$2,

$r_\perp \simeq \theta_{12}L$3

or, with $r_\perp \simeq \theta_{12}L$4 restored, $r_\perp \simeq \theta_{12}L$5. Matching the configuration-space separation of the momentum branches to this decoherence time gives the central discretization scale

$r_\perp \simeq \theta_{12}L$6

On this timescale, interference between different momentum branches is destroyed and the heavy-quark momentum can be updated stochastically.

The resulting classical Langevin equation is

$r_\perp \simeq \theta_{12}L$7

with

$r_\perp \simeq \theta_{12}L$8

The heavy-quark color randomization time is $r_\perp \simeq \theta_{12}L$9, parametrically shorter than $$S(L)=\exp\!\left[-\frac{1}{12}\hat q\,L^3\,\theta_{12}^2\right],\qquad \Delta_{\rm med}=1-\exp\!\left[-\frac{1}{12}\hat q\,L^3\,\theta_{12}^2\right].$$0, so color behaves effectively as a rapidly randomized quantum variable that selects classical pointer states for momentum. In this setting, color decoherence is the mechanism enabling the quantum-to-classical transition in heavy-quark transport.

5. Decoherence of the color code and mixed-state topological order

In quantum information, “color decoherence” can refer to decoherence of the two-dimensional color code on the honeycomb lattice. Qubits live on vertices, plaquettes are $$S(L)=\exp\!\left[-\frac{1}{12}\hat q\,L^3\,\theta_{12}^2\right],\qquad \Delta_{\rm med}=1-\exp\!\left[-\frac{1}{12}\hat q\,L^3\,\theta_{12}^2\right].$$1-colorable, and the stabilizer Hamiltonian is

$$S(L)=\exp\!\left[-\frac{1}{12}\hat q\,L^3\,\theta_{12}^2\right],\qquad \Delta_{\rm med}=1-\exp\!\left[-\frac{1}{12}\hat q\,L^3\,\theta_{12}^2\right].$$2

with

$$S(L)=\exp\!\left[-\frac{1}{12}\hat q\,L^3\,\theta_{12}^2\right],\qquad \Delta_{\rm med}=1-\exp\!\left[-\frac{1}{12}\hat q\,L^3\,\theta_{12}^2\right].$$3

The decoherence channel studied in this context acts only on red links,

$$S(L)=\exp\!\left[-\frac{1}{12}\hat q\,L^3\,\theta_{12}^2\right],\qquad \Delta_{\rm med}=1-\exp\!\left[-\frac{1}{12}\hat q\,L^3\,\theta_{12}^2\right].$$4

At $$S(L)=\exp\!\left[-\frac{1}{12}\hat q\,L^3\,\theta_{12}^2\right],\qquad \Delta_{\rm med}=1-\exp\!\left[-\frac{1}{12}\hat q\,L^3\,\theta_{12}^2\right].$$5, this is equivalent to projectively measuring $$S(L)=\exp\!\left[-\frac{1}{12}\hat q\,L^3\,\theta_{12}^2\right],\qquad \Delta_{\rm med}=1-\exp\!\left[-\frac{1}{12}\hat q\,L^3\,\theta_{12}^2\right].$$6 without recording outcomes (Kataoka et al., 24 Apr 2026).

The resulting mixed state is not merely a noisy version of the pure color code. XX decoherence on red links proliferates the anyon $$S(L)=\exp\!\left[-\frac{1}{12}\hat q\,L^3\,\theta_{12}^2\right],\qquad \Delta_{\rm med}=1-\exp\!\left[-\frac{1}{12}\hat q\,L^3\,\theta_{12}^2\right].$$7, rendering it transparent. Anyons that braid nontrivially with $$S(L)=\exp\!\left[-\frac{1}{12}\hat q\,L^3\,\theta_{12}^2\right],\qquad \Delta_{\rm med}=1-\exp\!\left[-\frac{1}{12}\hat q\,L^3\,\theta_{12}^2\right].$$8 become confined in the mixed-state sense, while those that braid trivially survive as deconfined excitations. Factoring the surviving deconfined sector by the transparent $$S(L)=\exp\!\left[-\frac{1}{12}\hat q\,L^3\,\theta_{12}^2\right],\qquad \Delta_{\rm med}=1-\exp\!\left[-\frac{1}{12}\hat q\,L^3\,\theta_{12}^2\right].$$9 yields a modular quotient equivalent to a single toric code. Operationally, the decohered color code inherits “half” of its topological properties: two logical qubits survive, and two non-contractible red-link $\Delta_{\rm med}$0 loops enter the center as strong $\Delta_{\rm med}$1-form symmetries.

The principal diagnostic is topological entanglement negativity. The negativity is

$\Delta_{\rm med}$2

and the topological entanglement negativity is

$\Delta_{\rm med}$3

For the pure color code,

$\Delta_{\rm med}$4

whereas for the maximally decohered state,

$\Delta_{\rm med}$5

This drop is the mixed-state signature of the emergent single-toric-code sector. As the decoherence strength is tuned from $\Delta_{\rm med}$6 to $\Delta_{\rm med}$7, $\Delta_{\rm med}$8 evolves smoothly from $\Delta_{\rm med}$9 toward $0$0, and the variance of $0$1 shows a large, broad peak for intermediate noise strengths $0$2–$0$3, nearly independent of system size. Raw negativities depend strongly on whether subsystems are commensurate with the emergent triangular lattice, but the cancellation-based TEN remains robust.

This body of work established color-code decoherence as a concrete route to intrinsic mixed-state topological order. A plausible implication is that decoherence, rather than merely destroying topological order, can reorganize it into a nonmodular phase with a modular quotient of lower rank.

6. Diamond color centers: magnetic-noise and phonon-induced decoherence

In solid-state spin physics, color decoherence is most often discussed for diamond color centers. For NV$0$4 centers in $0$5 $0$6C-enriched diamond, the dominant mechanism is magnetic noise from the nuclear-spin bath. The ground-state Hamiltonian is

$0$7

with $0$8. Near level anti-crossings, hyperfine-mediated mixing produces decoherence-protected transitions between mixed hyperfine sublevels within the $0$9 manifold. In the reported measurements, at magnetic field strengths ranging from $\theta < \theta_{12}$00, linewidths as low as $\theta < \theta_{12}$01 were observed at $\theta < \theta_{12}$02, compared with an average linewidth $\theta < \theta_{12}$03 for conventional electron-type transitions and a maximum observed $\theta < \theta_{12}$04. This corresponds to linewidth narrowing factors up to $\theta < \theta_{12}$05 and $\theta < \theta_{12}$06, with minimum effective gyromagnetic ratio $\theta < \theta_{12}$07 and minimum curvature $\theta < \theta_{12}$08 (Parker et al., 2015).

Hahn-echo dynamics of an individual NV center in a $\theta < \theta_{12}$09C bath provide another operational notion of color decoherence. The revival period is set predominantly by the nuclear Larmor frequency and obeys the fitted relation

$\theta < \theta_{12}$10

with $\theta < \theta_{12}$11 in ms when $\theta < \theta_{12}$12 is in Gauss, while

$\theta < \theta_{12}$13

Among the three characteristic timescales $\theta < \theta_{12}$14, $\theta < \theta_{12}$15, and $\theta < \theta_{12}$16, the paper identifies $\theta < \theta_{12}$17 as the most sensitive and robust magnetic-field marker. The proposed vector-magnetometry protocol achieves single-NV sensitivity $\theta < \theta_{12}$18, and with $\theta < \theta_{12}$19C enrichment and pulse optimization the reported figure is $\theta < \theta_{12}$20 (Li et al., 2017).

For group-IV color centers, the dominant channel is phonon-induced decoherence of a spin-orbit split orbital manifold. In the effective description, the acoustic-phonon spectral density is super-Ohmic,

$\theta < \theta_{12}$21

and the orbital relaxation rate scales as

$\theta < \theta_{12}$22

Representative ground-state orbital splittings are

$\theta < \theta_{12}$23

so heavier group-IV centers suppress thermal occupation of the relevant phonon modes more strongly at fixed temperature (Harris et al., 2023). A complementary Born–Markov treatment for SiV-class centers derives a Lindblad equation with rates proportional to $\theta < \theta_{12}$24 and $\theta < \theta_{12}$25, yielding exponential decay of coherences with rate $\theta < \theta_{12}$26 and a direct connection between phonon occupation and entanglement loss in heralded spin–spin states (Dhara et al., 2023).

At the architectural level, memory decoherence in NV-based distributed quantum computing has been modeled explicitly in a distributed toric surface code. For diamond color centers, coherence times during entanglement generation are orders of magnitude lower than coherence times of idling qubits. Incorporating these realistic timescales predicts error probability thresholds for gate and measurement reduced by at least a factor of three compared to prior work with more idealized noise models, together with a threshold of $\theta < \theta_{12}$27 in the ratio between the entanglement generation and the decoherence rates (Bone et al., 2024).

Across these solid-state settings, color decoherence is not a single mechanism but a family of environment-induced processes—nuclear-spin dephasing, phonon-assisted orbital relaxation, and operation-induced memory loss—acting on spin or spin-orbital degrees of freedom hosted by color centers. The common structure is open-system suppression of coherence; the controlling bath, timescale, and observable depend on whether the platform is an NV ensemble, a single NV under ODMR or Hahn echo, a group-IV spin–photon interface, or a distributed-network node.