Flavor-Chiral Oscillations in Dirac Neutrinos

• Flavor-chiral oscillations are phenomena coupling neutrino flavor mixing with chirality transitions in massive Dirac fermions, where standard oscillations emerge in the ultrarelativistic limit and chiral suppression occurs otherwise.
• The framework employs Dirac-wave and quantum field-theoretic treatments that elucidate rapid u–v interference terms and the trade-off between exact flavor definition and exclusive positive-energy propagation.
• Chiral oscillations influence neutrino detection in nonrelativistic regimes and astrophysical settings, affecting experimental interpretations of neutrino mass, mixing, and potential flavor-chirality entanglement.

Flavor-chiral oscillations are the coupled phenomena that arise when neutrino flavor mixing is treated together with the non-conservation of chirality for massive Dirac fermions. In this setting, flavor fields are superpositions of mass eigenfields, while weak charged-current and neutral-current interactions create and detect neutrinos approximately as left-chiral states. The resulting dynamics combines the usual flavor phase differences with a Dirac-sector left-right admixture that is absent in scalar treatments. In the ultrarelativistic limit the standard oscillation formula is recovered, but outside that limit the theory predicts chiral suppression factors, possible rapid $u$–$v$ interference terms, and, in some formulations, an unavoidable tension between exact initial flavor definition and purely positive-energy propagation (Bernardini et al., 2010).

1. Kinematic and field-theoretic foundations

For each mass eigenstate $i$, vacuum propagation is governed by the free Dirac equation

$(i\,\gamma^\mu \partial_\mu - m_i)\,\psi_i(x)=0,$

and flavor fields are defined by a unitary mixing matrix. In the two-flavor case,

$$\begin{pmatrix}\nu_e\ \nu_\mu\end{pmatrix} = \begin{pmatrix} \cos\theta & \sin\theta\ -\sin\theta & \cos\theta \end{pmatrix} \begin{pmatrix}\psi_1\ \psi_2\end{pmatrix}.$$

Weak interactions project onto

$$P_L=\frac{1-\gamma_5}{2},\qquad P_R=\frac{1+\gamma_5}{2},$$

so created and detected neutrinos are approximately left-chiral. For massive fermions, however, chirality is not conserved in free motion, whereas helicity is conserved. One compact statement of this distinction is

$$\frac{d}{dt}\langle h\rangle=0,\qquad \frac{d}{dt}\langle \gamma_5\rangle=2 i m\,\langle \gamma^0\gamma_5\rangle.$$

Accordingly, chirality and helicity coincide only in the massless or ultrarelativistic limit, and flavor evolution for Dirac neutrinos must distinguish them (Bernardini et al., 2010, Bittencourt et al., 2023).

At the single-mass level, the Dirac mass term mixes left and right chiral components. For a mass eigenstate $i$ with momentum $p$ and energy $E_i=\sqrt{p^2+m_i^2}$, the chiral amplitudes can be written as

$v$0

which implies

$v$1

Thus the wrong-chirality component is suppressed by $v$2 and oscillates with the fast scale $v$3, since $v$4 (Bittencourt et al., 2023).

2. Dirac-wave descriptions and the flavor-chiral probability formulas

In first-quantized Dirac treatments, flavor oscillations and chiral oscillations are intertwined because each mass component is a four-spinor rather than a scalar wave packet. For two-flavor mixing, the chirality-resolved flavor amplitudes can be written as

$v$5

so that

$v$6

If chirality and spin are not measured, tracing them out yields a reduced flavor density matrix whose relativistic limit reproduces the standard flavor probability. Away from that limit, the functions $v$7 and $v$8 encode the $v$9 and $i$0 corrections induced by chirality (Bittencourt et al., 2023).

A representation-independent Dirac treatment uses the flavor-space kernel

$i$1

In momentum space the mixed element separates into a standard term and a rapid term: $i$2 with

$i$3

The first term reduces to the standard oscillation kernel after momentum smearing; the second is the rapid $i$4–$i$5 interference contribution, with frequency approximately $i$6 (Bernardini et al., 2010).

A central structural result of the causal Dirac analysis is a dichotomy. If exact flavor creation at $i$7 is imposed by taking equal mass-eigenstate wave packets, then both positive- and negative-frequency components are required, and rapid oscillations appear. If one instead restricts the initial state to positive-energy components only, rapid oscillations disappear, but the flavor is not exactly defined at $i$8: a tiny wrong-flavor component of order $i$9 is unavoidable. For unequal masses, these two requirements cannot be simultaneously satisfied without reducing the solution to the trivial one (Bernardini et al., 2010).

Wave-packet analyses refine this picture without changing its qualitative content. In the intermediate wave-packet approach, the interference factor takes the form

$(i\,\gamma^\mu \partial_\mu - m_i)\,\psi_i(x)=0,$0

where $(i\,\gamma^\mu \partial_\mu - m_i)\,\psi_i(x)=0,$1 contains slippage-induced damping and $(i\,\gamma^\mu \partial_\mu - m_i)\,\psi_i(x)=0,$2 contains an additional phase from second-order energy expansion. These corrections survive spin inclusion, while the rapid terms remain bounded and vanish for $(i\,\gamma^\mu \partial_\mu - m_i)\,\psi_i(x)=0,$3 (Bernardini et al., 2010).

3. Quantum-field-theoretic formulations

Second-quantized formulations eliminate some of the ambiguities of first-quantized flavor states by working directly with Fock-space operators and flavor charges. In a simple QFT construction, flavor states are coherent superpositions of mass-eigenstate wave packets,

$(i\,\gamma^\mu \partial_\mu - m_i)\,\psi_i(x)=0,$4

and the flavor charge operator is

$(i\,\gamma^\mu \partial_\mu - m_i)\,\psi_i(x)=0,$5

For equal momentum distributions $(i\,\gamma^\mu \partial_\mu - m_i)\,\psi_i(x)=0,$6, the oscillation probability contains no rapid $(i\,\gamma^\mu \partial_\mu - m_i)\,\psi_i(x)=0,$7–$(i\,\gamma^\mu \partial_\mu - m_i)\,\psi_i(x)=0,$8 interference term, but the initial flavor violation survives: $(i\,\gamma^\mu \partial_\mu - m_i)\,\psi_i(x)=0,$9 In pion decay, the intrinsic flavor-violating branching fractions remain very small but calculable, for example

$$\begin{pmatrix}\nu_e\ \nu_\mu\end{pmatrix} = \begin{pmatrix} \cos\theta & \sin\theta\ -\sin\theta & \cos\theta \end{pmatrix} \begin{pmatrix}\psi_1\ \psi_2\end{pmatrix}.$$0

and are larger than loop-induced lepton-flavor-violating decays in the Standard Model (Bernardini et al., 2010).

A more explicit QFT treatment diagonalizes the non-conserved chiral charges by Bogoliubov transformations. For each mass $$\begin{pmatrix}\nu_e\ \nu_\mu\end{pmatrix} = \begin{pmatrix} \cos\theta & \sin\theta\ -\sin\theta & \cos\theta \end{pmatrix} \begin{pmatrix}\psi_1\ \psi_2\end{pmatrix}.$$1,

$$\begin{pmatrix}\nu_e\ \nu_\mu\end{pmatrix} = \begin{pmatrix} \cos\theta & \sin\theta\ -\sin\theta & \cos\theta \end{pmatrix} \begin{pmatrix}\psi_1\ \psi_2\end{pmatrix}.$$2

and the chiral ladder operators mix particle and antiparticle operators with time-dependent phases. The resulting chiral vacuum is a condensate of particle-antiparticle pairs with definite masses and helicities, and in the infinite-volume limit it is orthogonal to the energy vacuum. In the two-flavor theory, combining the chiral Bogoliubov transformation with the flavor rotation yields flavor-chiral annihilation operators and a unified survival probability

$$\begin{pmatrix}\nu_e\ \nu_\mu\end{pmatrix} = \begin{pmatrix} \cos\theta & \sin\theta\ -\sin\theta & \cos\theta \end{pmatrix} \begin{pmatrix}\psi_1\ \psi_2\end{pmatrix}.$$3

whose ultrarelativistic limit is the Pontecorvo formula and whose full form reproduces the first-quantized Dirac result (Bittencourt et al., 2024, Blasone et al., 13 Jul 2025).

The same physics can be expressed in propagator language. In the two-point-function formalism, one inserts chiral projectors at the source and detector,

$$\begin{pmatrix}\nu_e\ \nu_\mu\end{pmatrix} = \begin{pmatrix} \cos\theta & \sin\theta\ -\sin\theta & \cos\theta \end{pmatrix} \begin{pmatrix}\psi_1\ \psi_2\end{pmatrix}.$$4

to define $$\begin{pmatrix}\nu_e\ \nu_\mu\end{pmatrix} = \begin{pmatrix} \cos\theta & \sin\theta\ -\sin\theta & \cos\theta \end{pmatrix} \begin{pmatrix}\psi_1\ \psi_2\end{pmatrix}.$$5 amplitudes. The $$\begin{pmatrix}\nu_e\ \nu_\mu\end{pmatrix} = \begin{pmatrix} \cos\theta & \sin\theta\ -\sin\theta & \cos\theta \end{pmatrix} \begin{pmatrix}\psi_1\ \psi_2\end{pmatrix}.$$6 channel reproduces the usual flavor oscillation probability, whereas the $$\begin{pmatrix}\nu_e\ \nu_\mu\end{pmatrix} = \begin{pmatrix} \cos\theta & \sin\theta\ -\sin\theta & \cos\theta \end{pmatrix} \begin{pmatrix}\psi_1\ \psi_2\end{pmatrix}.$$7 channel is suppressed by $$\begin{pmatrix}\nu_e\ \nu_\mu\end{pmatrix} = \begin{pmatrix} \cos\theta & \sin\theta\ -\sin\theta & \cos\theta \end{pmatrix} \begin{pmatrix}\psi_1\ \psi_2\end{pmatrix}.$$8 in amplitude and by $$\begin{pmatrix}\nu_e\ \nu_\mu\end{pmatrix} = \begin{pmatrix} \cos\theta & \sin\theta\ -\sin\theta & \cos\theta \end{pmatrix} \begin{pmatrix}\psi_1\ \psi_2\end{pmatrix}.$$9 in probability. This formulation is especially useful when widths, unstable intermediate states, or nonstandard dispersion relations are important (Martone et al., 2011).

4. Limiting regimes, observables, and phenomenology

For left-chiral production and detection, flavor probabilities acquire a small subtraction corresponding to leakage into undetected right-chiral components. In the two-flavor case,

$$P_L=\frac{1-\gamma_5}{2},\qquad P_R=\frac{1+\gamma_5}{2},$$0

For a single mass eigenstate this reduces to

$$P_L=\frac{1-\gamma_5}{2},\qquad P_R=\frac{1+\gamma_5}{2},$$1

The leading suppression is therefore $$P_L=\frac{1-\gamma_5}{2},\qquad P_R=\frac{1+\gamma_5}{2},$$2. In laboratory regimes the effect is negligible: the vacuum right-chiral component is described as utterly negligible for neutrino energies relevant to present oscillation experiments, with $$P_L=\frac{1-\gamma_5}{2},\qquad P_R=\frac{1+\gamma_5}{2},$$3 depending on the mass scale and energy (Bernardini et al., 2010).

Rapid terms are likewise negligible under ordinary oscillation conditions. The $$P_L=\frac{1-\gamma_5}{2},\qquad P_R=\frac{1+\gamma_5}{2},$$4–$$P_L=\frac{1-\gamma_5}{2},\qquad P_R=\frac{1+\gamma_5}{2},$$5 interference piece proportional to $$P_L=\frac{1-\gamma_5}{2},\qquad P_R=\frac{1+\gamma_5}{2},$$6 is largest near $$P_L=\frac{1-\gamma_5}{2},\qquad P_R=\frac{1+\gamma_5}{2},$$7 and for $$P_L=\frac{1-\gamma_5}{2},\qquad P_R=\frac{1+\gamma_5}{2},$$8, but realistic baselines and packet widths wash it out. In wave-packet language the coherence scale is

$$P_L=\frac{1-\gamma_5}{2},\qquad P_R=\frac{1+\gamma_5}{2},$$9

and the scalar-like limit is recovered when $$\frac{d}{dt}\langle h\rangle=0,\qquad \frac{d}{dt}\langle \gamma_5\rangle=2 i m\,\langle \gamma^0\gamma_5\rangle.$$0 (Bernardini et al., 2010).

The nonrelativistic regime is qualitatively different. For momenta comparable to the masses, the left-chiral content of each mass eigenstate oscillates strongly, and the time-averaged flavor survival probability is depleted relative to the standard formula. A Dirac-bispinor calculation finds that in the non-relativistic regime this depletion can be as large as $$\frac{d}{dt}\langle h\rangle=0,\qquad \frac{d}{dt}\langle \gamma_5\rangle=2 i m\,\langle \gamma^0\gamma_5\rangle.$$1, while it is negligible in the ultrarelativistic regime (Bittencourt et al., 2020). This motivates renewed interest in low-energy settings such as the cosmic neutrino background. In that context, realistic detection schemes such as PTOLEMY-like setups must account for chiral oscillations, because for $$\frac{d}{dt}\langle h\rangle=0,\qquad \frac{d}{dt}\langle \gamma_5\rangle=2 i m\,\langle \gamma^0\gamma_5\rangle.$$2 they can deplete the measured left-chiral flux and affect the interpretation of Dirac-versus-Majorana signatures (Bittencourt et al., 2023).

External fields introduce an additional channel for spin-flip. In a static magnetic field transverse to the momentum,

$$\frac{d}{dt}\langle h\rangle=0,\qquad \frac{d}{dt}\langle \gamma_5\rangle=2 i m\,\langle \gamma^0\gamma_5\rangle.$$3

whereas for $$\frac{d}{dt}\langle h\rangle=0,\qquad \frac{d}{dt}\langle \gamma_5\rangle=2 i m\,\langle \gamma^0\gamma_5\rangle.$$4 the correction to the vacuum chiral dynamics is only of order $$\frac{d}{dt}\langle h\rangle=0,\qquad \frac{d}{dt}\langle \gamma_5\rangle=2 i m\,\langle \gamma^0\gamma_5\rangle.$$5. Such effects can become relevant in strong astrophysical environments, but are negligible for typical laboratory neutrinos (Bernardini et al., 2010).

5. Hyperentanglement and quantum-correlation structure

A more recent formulation emphasizes that the state of a massive Dirac neutrino produced in a weak interaction is not merely a flavor superposition. Because free Dirac evolution correlates flavor, chirality, and spin, the neutrino state can be written as

$$\frac{d}{dt}\langle h\rangle=0,\qquad \frac{d}{dt}\langle \gamma_5\rangle=2 i m\,\langle \gamma^0\gamma_5\rangle.$$6

with

$$\frac{d}{dt}\langle h\rangle=0,\qquad \frac{d}{dt}\langle \gamma_5\rangle=2 i m\,\langle \gamma^0\gamma_5\rangle.$$7

This state is hyperentangled in $$\frac{d}{dt}\langle h\rangle=0,\qquad \frac{d}{dt}\langle \gamma_5\rangle=2 i m\,\langle \gamma^0\gamma_5\rangle.$$8. The reduced flavor density matrix after tracing out chirality and spin has matrix elements controlled by the functions $$\frac{d}{dt}\langle h\rangle=0,\qquad \frac{d}{dt}\langle \gamma_5\rangle=2 i m\,\langle \gamma^0\gamma_5\rangle.$$9 and $i$0, and its purity is

$i$1

In the relativistic limit $i$2, so flavor effectively disentangles from chirality, whereas in the nonrelativistic limit sizable flavor-chirality entanglement oscillations appear (Bittencourt et al., 2023).

The same work applies complete complementarity relations to quantify the redistribution of predictability, coherence, and entanglement. For the flavor subsystem,

$i$3

As flavor superposition develops, predictability decreases while coherence and entanglement increase, with their sum constrained by the complementarity relation. This reframes flavor-chiral oscillations as a dynamical redistribution problem in multipartite quantum information rather than only as a probability-modulation effect (Bittencourt et al., 2023).

The framework extends to spin-entangled lepton-antineutrino pairs from pion decay. After projecting onto definite chiralities and evolving freely, the reduced two-spin density matrix has the form

$i$4

with the initial logarithmic-negativity amplitude

$i$5

Time evolution then transfers spin-spin entanglement into coherence and correlations involving flavor and chirality. This provides a possible indirect route to probing flavor-chiral dynamics through correlation observables, even when direct wrong-chirality detection is impractical (Bittencourt et al., 2023).

6. Interpretative disputes and related usages

The existence of vacuum chiral oscillations for physically produced neutrinos is disputed. One line of work argues that the apparent $i$6 oscillation arises only when one evolves the projected chiral spinor $i$7, which decomposes as

$i$8

and therefore contains a negative-energy component. In that analysis the state produced in a weak process must instead be computed from the production amplitude and contains only positive-energy solutions. Its helicity components propagate with the same energy in vacuum, so no space-time-dependent relative phase develops between them. The corresponding detection probability is

$i$9

which is interpreted as a zero-distance mismatch between production and detection helicity mixtures rather than as a dynamical vacuum oscillation (Smirnov, 9 May 2025).

Within that same amplitude-based viewpoint, matter can generate genuine chirality oscillations because left- and right-handed components experience different forward-scattering potentials. The induced phase is

$p$0

and the oscillation length is

$p$1

The oscillation depth nonetheless remains extremely small for ultrarelativistic neutrinos because the production and detection admixtures are themselves suppressed by $p$2 (Smirnov, 9 May 2025).

A different source of terminological ambiguity appears in dense-neutrino-gas studies. There, “flavor-chiral” can refer not to left-right conversion of Dirac chirality, but to flavor dynamics in a sector carrying a chiral lepton number. Fast flavor oscillation waves and flavor isospin waves are governed by mean-field equations for flavor density matrices and by the electron lepton number moments $p$3 and $p$4; the formulations explicitly omit spin/helicity coherence, magnetic-moment couplings, and mass-suppressed spin-flip operators. What propagates is a coherent flavor wave that redistributes electron lepton number, not an $p$5 Dirac-chirality oscillation (Martin et al., 2019, Duan et al., 2021).

Outside the standard complex Dirac-QFT literature, a quaternionic massless-Dirac model has proposed flavor oscillations even for massless neutrinos by identifying flavor with the quaternionic directions $p$6, $p$7, and $p$8. In that construction the solutions remain spin-$p$9 states with helicity tied to the sign of the energy, and chirality stays locked to helicity in the massless limit; flavor oscillates, but chiral oscillation does not (Welch, 2016).