On the origin of CP symmetry violations

Abstract

Experiments devoted to charge parity (CP) violation are normally interpreted by adjusting the elements of the Cabibbo-Kobayashi-Maskawa matrix to the measured violation parameters. However, the physical origin of these violations remains an open issue. To resolve this issue, the impact of Earth’s gravity on meson oscillations is analysed. The effect of gravity is to couple flavour oscillations to quark zitterbewegung oscillations, and this coupling induces a superposition of CP eigenstates. The three types of CP violation effects result from this gravity-induced mixing. The three associated violation parameters are predicted in agreement with experimental data. The amplitude of the violation is linear with respect to gravity, so this new mechanism allows us to envisage cosmological evolutions that provide the observed baryonic asymmetry of the universe.

I Introduction

Following the first observations of long-lived kaons anomalous decays [1], several observables associated with flavored neutral mesons Charge-Parity violation (CPV) have been identified, measured and interpreted during the past decades. Neutral mesons experiments dedicated to CPV are interpreted within the framework of the standard model (SM) through the adjustment between (i) the Kobayashi-Maskawa (KM) complex phase [2], in the Cabibbo-Kobayashi-Maskawa (CKM) matrix [2, 3], and (ii) the measured violation parameters. Despite the success of the SM-CKM interpretation based on the adjustment between the CKM matrix and the experimental data, the physical origin of CPV remains an open issue. To resolve this issue, we demonstrate that gravity induced CPV (GICPV) provides a pertinent and accurate framework to interpret the most documented experimental evidences of CPV and to predict the violation parameters in agreement with the experimental data. This set of new results does not rely on the adjustment of free parameters.

As a consequence of the accuracy of GICPV quantitative predictions, we conjecture that far from any massive object, i.e. in a flat Lorentzian space-time, the CKM matrix must be free from any CPV phase: CPV effects appears to be gravity induced near massive objects like Earth.

The three main types (i, ii and iii) of CPV experiments are analyzed here: (i) indirect CPV in the mixing observed with neutral kaons $K^{0}/\overline{K}^{0}$, the observable of these experiments is the parameter $\mathop{\rm Re}\varepsilon$ [4]; (ii) direct CPV in decays into one final state, the observable of these experiments is the parameter $\mathop{\rm Re}\left(\varepsilon^{\prime}/\varepsilon\right)$ [5]; (iii) CPV in interference between decays with and without mixing, the observable of these $B^{0}/\overline{B}^{0}$ experiments is the angle $\beta$ [5]. A forth (iv) experimental evidence of CPV is to be considered: the observed dominance of baryons over antibaryons in our universe because CPV is one of the necessary condition to build cosmological evolution models compatible with the baryon-antibaryon abundance asymmetry [6].

The SM provides a framework to interpret Earth based CPV experiments such as (i, ii and iii), but this interpretation, incorporated into cosmological evolution models, fails, by several orders of magnitude, to account for this (iv) major CPV evidence. To explain how our matter-dominated universe emerged during its early evolution we need to identify a CPV mechanism far larger than the KM one. Beside its potential to predict accurately the measured parameters ($\varepsilon$, $\varepsilon^{\prime}$, $\beta$) associated with types (i, ii and iii) CPV experiments on Earth, the new GICPV mechanism opens very interesting perspectives to set up cosmological models displaying an asymmetric baryogenesis compatible with the present state of our universe. Indeed, during the early stages of the cosmological evolution, gravity/curvature was far more larger than on Earth today and GICPV, which is a linear function of the gravitational field, opens an avenue to resolve the present contradiction between the very small SM-KM CPV and the very large CPV needed to build a pertinent model of our matter-dominated universe.

In this paper we demonstrate that the small coupling, induced by Earth’s gravity, between (i) fast quarks zitterbewegung oscillations, at the velocity of light, inside the mesons and (ii) strangeness oscillations $\Delta S=2$, or bottomness oscillations $\Delta B^{\prime}=2$, provides both a qualitative explanation of CPV and quantitative predictions of the CPV parameters $\varepsilon$, $\varepsilon^{\prime}$ and $\beta$ in agreement with the most recent experimental measurements reviewed in PDG 2024 [5].

The fact that the combination of gravity and quark zitterbewegung is a plausible candidate to explain the origin of CPV can be understood heuristically as follows. This heuristic argument is restricted to the $K^{0}/\overline{K}^{0}$ case and it can be easily extended to $B^{0}/\overline{B}^{0}$.

The Hamiltonian describing $\Delta S=2$ kaons oscillations in a flat Lorentzian space-time, far from any massive objects, is $\widehat{H}$. The observed meson eigenstates, $\left|M_{1}\right\rangle$ and $\left|M_{2}\right\rangle$, are solutions of $\widehat{H}\cdot\left|M_{1/2}\right\rangle=E_{1/2}\left|M_{1/2}\right\rangle$ where, in the meson rest frame, the eigenenergy, $E_{1}$ and $E_{2}$, are the inertial masses: $m+\delta m/2$ and $m-\delta m/2$. The Hamiltonian $\widehat{H}$ commute with the CP operator $\left[\widehat{H},\widehat{CP}\right]=0$, thus these energy eigenstates are also CP eigenstates $\widehat{CP}\left|M_{1/2}\right\rangle=\pm\left|M_{1/2}\right\rangle$. Consider now a CPV perturbation $\widehat{V}$ $\ll\widehat{H}$ that might slightly mix the CP eigenstates $\left|M_{1/2}\right\rangle$ and shift the eigenenergy $E_{1/2}$. With such a perturbation the observed eigenstates, $\left|M_{1}^{V}\right\rangle$ and $\left|M_{2}^{V}\right\rangle$, are solutions of

$$\left(\widehat{H}+\widehat{V}\right)\cdot\left|M_{1/2}^{V}\right\rangle=E_{1/2}^{V}\left|M_{V1/2}\right\rangle\text{.}$$ (1)

The observed shifted eigenvalues $E_{1/2}^{V}$ and mixed eigenstates $M_{V1/2}$, solutions of equation (1), are given by the usual first order perturbation theory:

$$\left|M_{1/2}^{V}\right\rangle=\left|M_{1/2}\right\rangle+\frac{\left\langle M_{2/1}\right|\widehat{V}\left|M_{1/2}\right\rangle}{E_{1/2}-E_{2/1}}\left|M_{2/1}\right\rangle\text{.}$$ (2)

On Earth, for a particle with a gravitational mass $m$, we consider the small Newtonian energy $V\sim$ $mG_{N}M_{\oplus}/R_{\oplus}$. The matrix elements $\left\langle M_{1/2}\right|\widehat{V}\left|M_{1/2}\right\rangle\sim mG_{N}M_{\oplus}/R_{\oplus}\sim$ $10^{-9}mc^{2}$ and, a priori, $\left\langle M_{2/1}\right|\widehat{V}\left|M_{1/2}\right\rangle=0$. This perturbation induces a small energy shift of the eigenenergy, which is not observable, and does not induces a CP eigenstates mixing.

To pursue the analysis of the GICPV hypothesis, we consider a CP violating vertical position fluctuation $x$ and use the first term of the Taylor expansion, with respect to $x\ll R_{\oplus}$, of the energy $mG_{N}$ $M_{\oplus}/\left(R_{\oplus}+x\right)$. The first term of the Taylor expansion is $mgx$ with $g$ = $G_{N}M_{\oplus}/R_{\oplus}^{2}$ = $9.8$ m/s². The Compton wavelength, $\lambda_{C}$ $=\hbar/mc\sim 4\times 10^{-16}$ m, provides an approximate upper bound of the size of any vertical fluctuations: $x\sim\lambda_{C}$. The expected, if any, matrix element $\left\langle M_{2/1}\right|\widehat{V}\left|M_{1/2}\right\rangle$ $\sim mg\lambda_{C}=\hbar g/c\sim 2.1\times 10^{-23}$ eV is rather small in front of $E_{1}-E_{2}\sim\delta mc^{2}$ $\sim 1.7\times 10^{-6}$ eV. Thus the expected GICPV mixing (2) is given by

$$\left|M_{1/2}^{V}\right\rangle\sim\left|M_{1/2}\right\rangle+\frac{\hbar g}{\delta mc^{3}}\left|M_{2/1}\right\rangle\text{.}$$ (3)

The order of magnitude $\hbar g/\delta mc^{3}$ $\sim 10^{-17}$ is far smaller than the experimental one $\left\langle M_{2/1}\right|\widehat{V}_{\text{exp}}\left|M_{1/2}\right\rangle/\delta mc^{2}\sim 10^{-3}$ [4]. This leads to the conclusion that this naive point of view does not provide an explanation to the origin of CPV. The small parameter $\hbar g/\delta mc^{3}$ has been identified in Ref. [7] by Fischbach who reaches the same negative conclusion. Fischbach also noted that the smallness of the parameter $\hbar g/\delta mc^{3}$ can be compensated by the large parameter $m/\delta m$ to provide a right order of magnitude, this puzzling remark was later used to explore the hypothesis of antigravity as the origin of CPV [8].

The previous negative conclusion about GICPV can be reevaluated if we consider the interplay between fast quarks zitterbewegung oscillations at the velocity of light, inside the mesons, and strangeness oscillations $\Delta S=2$.

Free or bound spin 1/2 fermions, like quarks, are well known to display the so called zitterbewegung (nonintuitive) behavior: a quiver (zitter) motion (bewegung), on a length scale given by the Compton wavelength, at an instantaneous velocity equal to the velocity light $c$ [9]. In addition to this zitterbewegung oscillation, the strangeness oscillation $K^{0}\rightleftharpoons\overline{K}^{0}$ takes place at the frequency $\delta mc^{2}/\hbar$ [4]. During one particle-antiparticle transition $K^{0}\rightleftharpoons\overline{K}^{0}$ the quarks zitterbewegung oscillation accumulate an energy $mgc\left(\hbar/\delta mc^{2}\right)$. The expected, if any, matrix element $\left\langle M_{2/1}\right|\widehat{V}\left|M_{1/2}\right\rangle\sim mg\hbar/\delta mc$ is responsible of a mixing of the CP eigenstates

$$\left|M_{1/2}^{V}\right\rangle\sim\left|M_{1/2}\right\rangle+\frac{m\hbar g}{\delta m^{2}c^{3}}\left|M_{2/1}\right\rangle\text{.}$$ (4)

The numerical value $m\hbar g/\delta m^{2}c^{3}$ $\sim 10^{-3}$ leads to the conclusion that this last point of view might provide an explanation to the origin of CPV and so requires a deeper analysis.

The new interpretation of CPV experiments presented below is based on the usual Hamiltonian of Lee, Oehme and Yang (LOY) [10, 11], completed here with Newtonian gravity. Neutral mesons oscillations such as $K^{0}\rightleftharpoons\overline{K}^{0}$   and $B^{0}\rightleftharpoons\overline{B}^{0}$ are very low energy oscillations ($10^{-6}-10^{-4}$ eV), so there is no need to rely on quantum field theory and the usual LOY Hamiltonian offers the pertinent framework to describe a low energy quantum oscillation between two quantum states slightly perturbed by gravity.

The study presented below complements a previous study based on two coupled Klein-Gordon equations describing $K^{0}/\overline{K}^{0}$ evolution on a Schwarzschild metric [12], rather than a Newtonian framework with two coupled Schrödinger equations used here. The results given by the Newtonian model are similar to those of this previous Einsteinian model [12], these results are thus model independent.

This paper is organized as follows, in the next section we briefly review the LOY Hamiltonian without CPV, then, in section 3, the experimental CPV parameters are defined. The impact of Earth’s gravity is considered in section 4 where, to describe neutral mesons oscillations $M^{0}\rightleftharpoons\overline{M}^{0}$on Earth, the CP conserving LOY Hamiltonian, presented in section 2, is completed with a gravity term.

The study of type (i), (ii) and (iii) GICPV are developed in sections 5, 6 and 7. We consider specifically type (i) and (ii) CPV for $K^{0}/\overline{K}^{0}\sim\left(d\overline{s}\right)/\left(\overline{d}s\right)$ and type (iii) CPV for $B^{0}/\overline{B}^{0}\sim\left(d\overline{b}\right)/\left(\overline{d}b\right)$. Section 8 provides a brief comment on others, $D^{0}/\overline{D}^{0}$ and $B_{s}^{0}/\overline{B_{s}}^{0}$, neutral mesons and gives our conclusions. In sections 2 and 4, $M^{0}/\overline{M}^{0}$ is either $K^{0}/\overline{K}^{0}$ or $B^{0}/\overline{B}^{0}$. In sections 5, 6 and 7 the experimental numerical values used to evaluate the expressions are taken from the most recent reference: PDG 2024 [5].

II Mass eigenstates without CPV

A generic flavored neutral meson state $\left|M\left(\tau\right)\right\rangle$ is a functions of the meson proper time $\tau$ and a linear superposition, with amplitudes $\left(a,b\right)$, of the two flavor eigenstates $\left|M^{0}\right\rangle$ and $\left|\overline{M}^{0}\right\rangle$ ($K^{0}/\overline{K}^{0}$ or $B^{0}/\overline{B}^{0}$). These flavor eigenstates provide an orthonormal basis of the mesons Hilbert space: $\left\langle M^{0}\right.\left|M^{0}\right\rangle=\left\langle\overline{M}^{0}\right.\left|\overline{M}^{0}\right\rangle=1$ and $\left\langle\overline{M}^{0}\right.\left|M^{0}\right\rangle=0$. Mesons states $\left|M\left(\tau\right)\right\rangle$ are unstable and decay to a set of final states $\left|f\right\rangle$. Because of these decays the Hilbert space should be extended to $\left\{\left|f\right\rangle\right\}$ and the time evolution should be described with an additional set of amplitudes $\left\{w_{f}\right\}$ as

$$\left|M\left(\tau\right)\right\rangle=a\left(\tau\right)\left|M^{0}\right\rangle+b\left(\tau\right)\left|\overline{M}^{0}\right\rangle+\sum_{f}w_{f}\left(\tau\right)\left|f\right\rangle.$$ (5)

Following Lee, Oehme and Yang, the Weisskopf-Wigner (WW) approximation [13] is used here to describe the coupling to $\left|f\right\rangle$ as an irreversible decay. Within the framework of this usual approximation [4, 5], rather than a unitary evolution with amplitudes $\left\{w_{f}\left(\tau\right)\right\}$, we introduce a non-Hermitian decay operator $j\widehat{\gamma}$ describing the ultimate $M\rightarrow f$ transitions as an irreversible process. It is to be noted that $f\rightarrow M$ transitions, avoided by the WW approximation, are in fact experimentally prohibited due to the very large phase space explored by the outgoing products $\left\{\left|f\right\rangle\right\}$. The irreversibility of the decay process gives rise to a violation of T that should not be attributed to fundamental interactions at the level of the CKM matrix of the SM.

The time evolution of $\left|M\left(\tau\right)\right\rangle$ can therefore be restricted to a two states Hilbert space: $\left|M^{0}\right\rangle,\left|\overline{M}^{0}\right\rangle$, at the cost of the loss of unitarity $d\left\langle M\right.\left|M\right\rangle/d\tau<0$ induced by $j\widehat{\gamma}$. This restriction of the Hilbert space to ($M^{0},\overline{M}^{0}$) leads to the LOY Hamiltonian $\widehat{H}_{Y}$ [10, 11]: the sum of the mass energy ($mc^{2}$), plus a $S=\pm 1$, or $B^{\prime}=\pm 1$, mixing operator ($\widehat{\delta m}c^{2}$), plus the irreversible decay:

$$\widehat{H}_{Y}=mc^{2}\widehat{I}-\widehat{\frac{\delta m}{2}}c^{2}-j\hbar\frac{\widehat{\gamma}}{2}\text{,}$$ (6)

where $\widehat{I}$ is the identity operator. The mixing and the decay operators, $\widehat{\delta m}$ and $\widehat{\gamma}$, are given by

$$\widehat{\delta m} = \delta m\left[\left|M^{0}\right\rangle\left\langle\overline{M}^{0}\right|+\left|\overline{M}^{0}\right\rangle\left\langle M^{0}\right|\right]\text{,}$$ (7)

$$\widehat{\gamma} = \Gamma\widehat{I}-\delta\Gamma\left[\left|M^{0}\right\rangle\left\langle\overline{M}^{0}\right|+\left|\overline{M}^{0}\right\rangle\left\langle M^{0}\right|\right]\text{,}$$ (8)

where $\delta m>0$ is the mass splitting between the heavy and light mass eigenstates and $\Gamma>0$, $\delta\Gamma<0$ are respectively the average and the splitting between the decay widths of the these eigenstates [4]. We take the convention $\widehat{CP}\left|M^{0}\right\rangle=\left|\overline{M}^{0}\right\rangle$. The evolution of the meson state $\left|M\left(\tau\right)\right\rangle$ is

$$j\hbar\frac{d\left|M\left(\tau\right)\right\rangle}{d\tau}=\widehat{H}_{Y}\cdot\left|M\left(\tau\right)\right\rangle\text{.}$$ (9)

The CP eigenstates $\left|M_{1}\right\rangle$ and $\left|M_{2}\right\rangle$ are related to the flavor eigenstates by

$$\left|M_{1/2}\right\rangle=\frac{\left|M^{0}\right\rangle}{\sqrt{2}}\pm\frac{\left|\overline{M}^{0}\right\rangle}{\sqrt{2}}=\pm\widehat{CP}\left|M_{1/2}\right\rangle\text{.}$$ (10)

These CP eigenstates are also mass eigenstates with eigenvalues $m\pm\delta m/2$.

The time evolution of these CP/mass eigenstates is given by the solutions of (9):

$$\left|M_{1/2}\left(\tau\right)\right\rangle=\left|M_{1/2}\right\rangle\exp-j\frac{c^{2}}{\hbar}\left[m\mp\frac{\delta m}{2}-j\hbar\frac{\Gamma\mp\delta\Gamma}{2c^{2}}\right]\tau\text{.}$$ (11)

The above symmetric picture, where $\widehat{CP}$ commute with $\widehat{H}_{Y}$, is no longer valid when the results of experiments dedicated to CPV are to be taken into account. The observed mass eigenstates are not the CP eigenstates $K_{1/2}$ or $B_{1/2}$ (10). The mass eigenvalues involved in (11) are not significantly changed by CPV.

III Observable violation parameters

The observed mass eigenstates are: the short-lived $S$ and the long-lived $L$ states $\left(K_{S/L}\right)$ for $K^{0}/\overline{K}^{0}$, and the light $L$ and heavy $H$ states $\left(B_{L/H}\right)$ for $B^{0}/\overline{B}^{0}$.

For $K^{0}/\overline{K}^{0}$ type (i) CPV, the observed mass eigenstates $\left|K_{S/L}\right\rangle$ are related to the CP eigenstates $\left|K_{1/2}\right\rangle$ (10) by

$$\left|K_{S/L}\right\rangle=\left|K_{1/2}\right\rangle+\varepsilon\left|K_{2/1}\right\rangle\text{.}$$ (12)

The quantity $\left\langle K_{S/L}\right.\left|K_{L/S}\right\rangle/2=\mathop{\rm Re}\varepsilon$ is an observable.

For $B^{0}/\overline{B}^{0}$ type (iii) CPV, it is convenient to introduce the angle $\beta$ and to consider that the mass eigenstates $\left|B_{L/H}\right\rangle$ are related to the CP eigenstates $\left|B_{1/2}\right\rangle$ (10) by

$$\left|B_{L/H}\right\rangle=\cos\beta\left|B_{1/2}\right\rangle+j\sin\beta\left|B_{2/1}\right\rangle\text{.}$$ (13)

Type (ii) direct CPV in the decay to one final state $\left\langle f\right|$ is also due to Earth’s gravity but $\varepsilon^{\prime}$, the associated CPV parameter, is not involved in the LOY Hamiltonian describing $\Delta S=2$ oscillations. The measurements of the direct violation parameter $\varepsilon^{\prime}$ are based on difficult and precise dedicated pions decays experiments. For the $2\pi^{0}$ decays of $K_{L}$ and $K_{S}$ the definition of $\varepsilon^{\prime}$ is related to the amplitude ratio $\eta$ ₀₀ by

$$\eta_{00}=\frac{\left\langle\pi^{0}\pi^{0}\right|{\cal T}\left|K_{L}\right\rangle}{\left\langle\pi^{0}\pi^{0}\right|{\cal T}\left|K_{S}\right\rangle}\equiv\varepsilon-2\varepsilon^{\prime}\text{,}$$ (14)

where $\mathop{\rm Re}\varepsilon\gg\mathop{\rm Re}\varepsilon^{\prime}$. CP symmetry is restored when $\varepsilon=0$, $\varepsilon^{\prime}=0$ and $\beta=0$.

These experimental parameters, $\varepsilon$, $\varepsilon^{\prime}$ and $\beta$, are used to construct the CKM matrix elements. Rather than adjusting the CPV part of the CKM matrix to these measured parameters, a new interpretation of the CPV experiments is proposed below. The final quantitative results predicted with this new interpretation leads to the conclusion that CPV effects observed in the three canonical types of flavored neutral mesons experiments (i, ii and iii) are gravity induced.

IV Impact of Earth’s gravity on neutral mesons oscillations

We assume that the Schrődinger equation (9) is pertinent far from any massive object, and, on Earth, we consider an additional energy term $mg\ \widehat{x}\left(\tau\right)$ in equation (6) such that the evolution becomes

$$j\hbar\frac{d\left|M\right\rangle}{d\tau}=\widehat{H}_{Y}\cdot\left|M\right\rangle+mg\ \ \widehat{x}\left(\tau\right)\cdot\left|M\right\rangle\text{.}$$ (15)

The vertical position operator $\ \widehat{x}\left(\tau\right)$ is associated with the vertical zitterbewegung internal motion inherent to all, free and bound, spin 1/2 fermions like quarks [9]. At the level of the quark structure, the flavor eigenstates, $M^{0}$ and $\overline{M}^{0}$, are stationary diquarks bound states ($K^{0}/\overline{K}^{0}\sim\left(d\overline{s}\right)/\left(\overline{d}s\right)$ and $B^{0}/\overline{B}^{0}\sim\left(d\overline{b}\right)/\left(\overline{d}b\right))$, ultimately described by Dirac spinors, $\left|q^{\prime}\overline{q}\right\rangle$ and $\left|\overline{q}^{\prime}q\right\rangle$, for one light quark $q^{\prime}$ and one heavier quark $q$ combined into singlet spin zero states : $M^{0}\sim\left|q^{\prime}\overline{q}\right\rangle$ and $\overline{M}^{0}$ $\sim$ $\left|\overline{q}^{\prime}q\right\rangle$. It is to be noted that $\left|M^{0}\right\rangle/\left|\overline{M}^{0}\right\rangle$ and $\left|q^{\prime}\overline{q}\right\rangle/\left|\overline{q}^{\prime}q\right\rangle$ belong to different Hilbert spaces.

A Dirac Hamiltonian $\widehat{H}_{D}$, describing quarks confinement, operate in the diquark Hilbert space. The position operator $\widehat{{\bf x}}\left(\tau\right)$, operating in the diquark spinor Hilbert space, fulfils Heisenberg’s equation:

$$j\hbar\frac{d\widehat{{\bf x}}}{d\tau}=\left[\widehat{{\bf x}},\widehat{H}_{D}\left(\widehat{{\bf x}},\widehat{{\bf p}}\right)\right]=j\hbar c{\bm{\alpha}}\text{.}$$ (16)

We have introduced the usual $4\times 4$ alpha matrices: ${\bm{\alpha}}=\left(\alpha_{x},\alpha_{y},\alpha_{z}\right)$ which can be expressed in terms of the $2\times 2$ Pauli matrices ${\bm{\sigma}}=\left(\sigma_{x},\sigma_{y},\sigma_{z}\right)$.

The non-zero matrix elements of ${\bm{\alpha}}c=d\widehat{{\bf x}}/d\tau$ are either $\pm c$ or $\pm jc$.

It is important to note that the zitterbewegung relation $d\widehat{{\bf x}}/d\tau=c{\bm{\alpha}}$ (16) is independent of the charge and mass of the fermions as well as of the shape and strength of the effective confinement potential involved in the Dirac Hamiltonian $\widehat{H}_{D}$ [9].

A zitterbewegung position operator $\widehat{{\bf x}}\left(\tau\right)$ operates also in the meson Hilbert space and is represented in this space by the following (unknown) four matrix elements $\left\langle{}\right|\widehat{{\bf x}}\left|{}\right\rangle$ of the stationary Dirac spinors diquarks states $\left|q^{\prime}\overline{q}\right\rangle$

$$\widehat{x}\left(\tau\right) = \left\langle q^{\prime}\overline{q}\right|\widehat{x}\left(\tau\right)\left|q^{\prime}\overline{q}\right\rangle\left|M^{0}\right\rangle\left\langle M^{0}\right|$$ (17)

$$+\left\langle\overline{q}^{\prime}q\right|\widehat{x}\left(\tau\right)\left|\overline{q}^{\prime}q\right\rangle\left|\overline{M}^{0}\right\rangle\left\langle\overline{M}^{0}\right|$$

$$+\left\langle\overline{q}^{\prime}q\right|\widehat{x}\left(\tau\right)\left|q^{\prime}\overline{q}\right\rangle\left|\overline{M}^{0}\right\rangle\left\langle M^{0}\right|$$

$$+\left\langle q^{\prime}\overline{q}\right|\widehat{x}\left(\tau\right)\left|\overline{q}^{\prime}q\right\rangle\left|M^{0}\right\rangle\left\langle\overline{M}^{0}\right|\text{.}$$

The Compton wavelength of the meson $\lambda_{C}$ provides an approximate upper bound of the matrix elements $\left|\left\langle{}\right|\widehat{x}\left|{}\right\rangle\right|$ in (17) as quarks are bound states inside the volume of a meson. The small numerical value of the energy $mg\lambda_{C}=\hbar g/c\sim 10^{-23}$ eV in front of $\delta mc^{2}\sim 10^{-6}-10^{-4}$ eV leads to the occurrence of a strong ordering between $mg\left|\left\langle{}\right|\widehat{x}\left|{}\right\rangle\right|$ $\sim\hbar g/c$ $\ll\delta mc^{2}$ and the other LOY matrix elements involved in $\widehat{H}_{Y}$. This ordering allows to set up a perturbative expansion of (15) with respect to the small expansion parameter $\hbar g/\delta mc^{3}$ $\sim 10^{-19}-10^{-17}$.

To do so we first define $\left|N\left(\tau\right)\right\rangle$ and $\left|n\left(\tau\right)\right\rangle$ such that

$$\left|M\left(\tau\right)\right\rangle=\left|N\left(\tau\right)\right\rangle\exp-j\frac{mc^{2}\tau}{\hbar}+\left|n\left(\tau\right)\right\rangle\exp-j\frac{mc^{2}\tau}{\hbar}\text{.}$$ (18)

The evolution of $\left|N\right\rangle+\left|n\right\rangle$ fulfils

$$j\hbar\frac{d}{d\tau}\left[\left|N\right\rangle+\left|n\right\rangle\right]=\left[\widehat{H^{\prime}}_{Y}+mg\ \ \widehat{x}\left(\tau\right)\right]\cdot\left[\left|N\right\rangle+\left|n\right\rangle\right]\text{,}$$ (19)

where the Hamiltonian $\widehat{H^{\prime}}_{Y}$ is given by

$$\widehat{H^{\prime}}_{Y}=\widehat{H}_{Y}-mc^{2}\widehat{I}=-\widehat{\delta m}c^{2}/2-j\hbar\widehat{\gamma}/2\text{.}$$ (20)

Then the states $\left|N\right\rangle$ and $\left|n\ \right\rangle$ are ordered according to: $\ \left|N\text{ }\right\rangle\sim$ $O\left(\hbar g/\delta mc^{3}\right)^{0}$ and $O\left(\hbar g/\delta mc^{3}\right)^{1}$ $\leq\left|n\right\rangle\ll O\left(\hbar g/\delta mc^{3}\right)^{0}$. With this expansion scheme Schrödinger’s equation (19) becomes

$$j\hbar\frac{d\left|N\right\rangle}{d\tau} = \widehat{H^{\prime}}_{Y}\cdot\left|N\right\rangle\text{,}$$ (21)

$$j\hbar\frac{d\left|n\right\rangle}{d\tau} = \widehat{H^{\prime}}_{Y}\cdot\left|n\right\rangle+mg\ \widehat{x}\cdot\left|N\right\rangle\text{.}$$ (22)

We introduce the inverse of the Hamiltonian $\widehat{H^{\prime}}_{Y}$ to define the operators $\widehat{Z}_{x}$ and $\widehat{Z}_{c}=\hbar d\widehat{Z}_{x}/d\tau$

$$\widehat{Z}_{x}\left(\tau\right)=jmg\ \widehat{x}\left(\tau\right)\cdot\widehat{H^{\prime}}_{Y}^{-1},\widehat{Z}_{c}=jmg\hbar\frac{d\widehat{{x}}}{d\tau}\cdot\widehat{H^{\prime}}_{Y}^{-1}\text{.}$$ (23)

To evaluate the orders of magnitudes of $Z_{x}$ and $Z_{c}$ we (i) anticipate the specific cases of $K^{0}/\overline{K}^{0}$ and $B^{0}/\overline{B}^{0}$ where we will use respectively $\widehat{\gamma}_{K}=\widehat{0}$ and $\gamma_{B}\sim\delta m_{B}$ leading to $H_{Y}^{\prime}\sim O\left(\delta mc^{2}\right)$ and (ii) use the fact that $d\widehat{{\bf x}}/d\tau=c{\bm{\alpha}}$ imply that the values of the matrix elements of $d\widehat{x}/d\tau$ are independent of $\tau$ and equal to $\pm c$, $\pm jc$ or $0$. The orders of magnitude of typical matrix elements of $\widehat{Z}_{x}$ and $\widehat{Z}_{c}$, between two normalized states, are thus $Z_{x}\sim O\left(\hbar g/\delta mc^{3}\right)$ and $Z_{c}\sim O\left(mc\hbar g/\delta mc^{2}\right)$. With the definitions (23) the relations (21,22) give

$$j\hbar\frac{d\left|n\right\rangle}{d\tau}=\widehat{H^{\prime}}_{Y}\cdot\left|n\right\rangle-\widehat{Z}_{c}\ \cdot\left|N\right\rangle+\hbar\frac{d}{d\tau}\left[\widehat{Z}_{x}\left(\tau\right)\cdot\left|N\right\rangle\right]\text{.}$$ (24)

We introduce the state $\left|n^{\prime}\right\rangle=\left|n\right\rangle+j\widehat{Z}_{x}\cdot\left|N\right\rangle$, such that

$$j\hbar d\left|n^{\prime}\right\rangle/d\tau=\widehat{H^{\prime}}_{Y}\cdot\left|n^{\prime}\right\rangle-\left[\widehat{Z}_{c}+j\widehat{H^{\prime}}_{Y}\cdot\widehat{Z}_{x}\right]\cdot\left|N\right\rangle\text{,}$$ (25)

The operator $H_{Y}^{\prime}\cdot Z_{x}\sim O\left(\delta mc^{2}\right)\cdot O\left(\hbar g/\delta mc^{3}\right)$ can be neglected in front $Z_{c}\sim O\left(m\hbar g/\delta mc\right)$ as $\delta m/m\sim 10^{-15}-10^{-13}$. To evaluate the interplay between zitterbewegung and $\Delta S=2$, or $\Delta B^{\prime}=2$, oscillations we have to solve:

$$j\hbar\frac{d\left|n^{\prime}\right\rangle}{d\tau}=\widehat{H^{\prime}}_{Y}\cdot\left|n^{\prime}\right\rangle-\widehat{Z}_{c}\cdot\left|N\right\rangle\text{.}$$ (26)

The last step is to express on the flavor basis $\left(M^{0},\overline{M}^{0}\right)$ the zitterbewegung instantaneous velocity operator $d\widehat{{x}}/d\tau$ involved in $\widehat{Z}_{c}$. As a consequence of (16) the eigenvectors of $d\widehat{{\bf x}}/d\tau=c{\bm{\alpha}}$ can be identified. Without loss of generality we consider the Dirac spinors eigenvectors of $\alpha_{x}$:

$$\frac{1}{\sqrt{2}}\left[\begin{tabular}[]{c}$1$\\ $0$\\ $0$\\ $1$\end{tabular}\right]\text{, }\frac{1}{\sqrt{2}}\left[\begin{tabular}[]{c}$0$\\ $1$\\ $1$\\ $0$\end{tabular}\right]\text{, }\frac{1}{\sqrt{2}}\left[\begin{tabular}[]{c}$1$\\ $0$\\ $0$\\ $-1$\end{tabular}\right]\text{, }\frac{1}{\sqrt{2}}\left[\begin{tabular}[]{c}$0$\\ $1$\\ $-1$\\ $0$\end{tabular}\right]\text{.}$$ (27)

The usual physical interpretation of these four spinors (27) is as follows [9].

Starting from the left, the first spinor and the second one describe a symmetric superpositions of one fermion and one antifermion: $\left|g\right\rangle=\left(\left|q\right\rangle+\left|\overline{q}\right\rangle\right)/\sqrt{2}$. These two symmetric superpositions (27) are eigenstates of $\alpha_{x}$ with eigenvalue $1$, and of $d\widehat{x}/d\tau$ with eigenvalue $+c$.

The last two spinors, on the right, describe an antisymmetric superpositions of one fermion and one antifermion: $\left|u\right\rangle=\left(\left|q\right\rangle-\left|\overline{q}\right\rangle\right)/\sqrt{2}$. These two antisymmetric superpositions (27) are eigenstates of $\alpha_{x}$ with eigenvalue $-1$, and of $d\widehat{x}/d\tau$ with eigenvalue $-c$.

Note that $\left\langle u\right.\left|g\right\rangle=0$. The symmetric CP eigenstate $M_{1}$ (10) is a combination of quarks spinors (27) of the $\left|g\right\rangle$ type and $M_{2}$, the antisymmetric CP eigenstate (10), is a combination of quarks spinors of the $\left|u\right\rangle$ type. Thus, in the two states LOY Hilbert space, on the ($M_{1},M_{2}$) CP basis (10), the representation of the zitterbewegung velocity operator $d\widehat{x}/d\tau$ is given by

$$\frac{d\widehat{x}}{d\tau}=c\left|M_{1}\right\rangle\left\langle M_{1}\right|-c\left|M_{2}\right\rangle\left\langle M_{2}\right|\text{.}$$ (28)

On the flavor basis ($M^{0},\overline{M}^{0}$) (28) gives the relations $\left\langle\overline{M}^{0}\right|d\widehat{x}/d\tau\left|M^{0}\right\rangle$ = $c$ and $\left\langle M^{0}\right|d\widehat{x}/d\tau\left|\overline{M}^{0}\right\rangle$ = $c$ and the two others matrix elements are equal to zero. In the LOY Hamiltonian (6) the mass $m$ of the antiparticle is positive like the mass of the particle, although, in a Dirac representation, the antiparticle are negative mass solutions. This point is resolved through the Feynman interpretation of an antiparticle as a particle propagating backward in time. To construct the LOY representation of $d\widehat{x}/d\tau$, on the flavor basis ($M^{0},\overline{M}^{0}$), the usual Feynman prescription leads to the following zitterbewegung velocity operator

$$\frac{d\widehat{x}}{d\tau}=c\left|M^{0}\right\rangle\left\langle\overline{M}^{0}\right|-c\left|\overline{M}^{0}\right\rangle\left\langle M^{0}\right|\text{.}$$ (29)

In two previous studies [12, 14], we have given two demonstrations of this result (29) with two different methods. Flavored neutral mesons pairs $K^{0}/\overline{K}^{0}$ and $B^{0}/\overline{B}^{0}$ display different $m$, $\delta m$, $\Gamma$ and $\delta\Gamma$ and the impact of Earth gravity on their behavior is to be analyzed specifically. In the following we keep the notations of (23, 26) with an additional index $K$ or $B$ when needed.

V Type (i) CPV in the mixing of $K^{0}/\overline{K}^{0}$

The ordering associated with the specific case of a $K^{0}/\overline{K}^{0}$ pair is given by: $\delta m_{K}/m_{K}\sim 10^{-15}$. The first step to interpret $K^{0}/\overline{K}^{0}$ experiments is to consider a unitary evolution where both particles are regarded as stable, $j\hbar\widehat{\gamma}=\widehat{0}$. Then, as the lifetime of $K_{L}$ is 577 times longer than the lifetime of $K_{S}$, we will set up a steady state-balance between: (i) the gravity induced small $K_{S}$ component regenerated from a given $K_{L}$ and (ii) this $K_{S}$ component fast decay.

Considering first a unitary evolution, we have to solve (21,26)

$$j\hbar\frac{d\left|n^{\prime}\right\rangle}{d\tau}=-\frac{\widehat{\delta m_{K}}}{2}c^{2}\cdot\left|n^{\prime}\right\rangle+2jm_{K}c^{-2}g\hbar\frac{d\widehat{x}}{d\tau}\ \cdot\widehat{\delta m_{K}}^{-1}\cdot\left|N\right\rangle$$ (30)

The operator $\widehat{\delta m_{K}}$ is given by (7) and the operator $d\widehat{x}/d\tau$ by (29). As $\widehat{\delta m_{K}}\cdot\widehat{\delta m_{K}}=\delta m_{K}^{2}\widehat{I}$ the actions of $\widehat{Z}_{c}$ and $\widehat{\delta m_{K}}$ on the CP eigenstates $\left|K_{1}\right\rangle$ and $\left|K_{2}\right\rangle$ are

$$m_{K}g\hbar c^{-2}\frac{d\widehat{x}}{d\tau}\ \cdot\widehat{\delta m_{K}}^{-1}\cdot\left|K_{2/1}\right\rangle = \kappa\frac{\delta m_{K}}{2}c^{2}\left|K_{1/2}\right\rangle\text{,}$$

$$\frac{\widehat{\delta m_{K}}}{2}\cdot\left|K_{2/1}\right\rangle = \mp\frac{\delta m_{K}}{2}\left|K_{2/1}\right\rangle\text{,}$$ (31)

where the small parameter $\kappa$ is defined by

$$\kappa=2m_{K}g\hbar/\delta m_{K}^{2}c^{3}=1.7\times 10^{-3}\text{.}$$ (32)

If we consider the following CP eigenstates, which are also the $\left(m_{K}\pm\delta m_{K}/2\right)$ mass eigenstates without CPV

$$\left|N\left(\tau\right)\right\rangle=\left|K_{2/1}\right\rangle\exp\left(\mp j\delta m_{K}c^{2}\tau/2\hbar\right)\text{,}$$ (33)

they fulfils Eq. (21) and the associated solution of (30) is

$$\left|n^{\prime}\left(\tau\right)\right\rangle=\pm j\kappa\left|K_{1/2}\right\rangle\exp\left(\mp j\delta m_{K}c^{2}\tau/2\hbar\right)\text{.}$$ (34)

Thus, on Earth, the mass eigenstates $\left|K_{L/S}^{\oplus}\right\rangle$ are not the CP eigenstates $\left|K_{2/1}\right\rangle$, but

$$\left|K_{L/S}^{\oplus}\right\rangle=\left|K_{2/1}\right\rangle\pm j\kappa\left|K_{1/2}\right\rangle\text{.}$$ (35)

We neglect the $O[10^{-6}]$ correction needed for normalization $\left\langle K_{L/S}^{\oplus}\right.\left|K_{L/S}^{\oplus}\right\rangle$ = $1$, and we have neglected the term $-j\widehat{Z}_{x}\cdot\left|K_{2/1}\right\rangle\sim O\left(\hbar g/\delta m_{K}c^{3}\right)$ in front of $\kappa\left|K_{2/1}\right\rangle\sim O\left(m_{K}\hbar g/\delta m_{K}^{2}c^{3}\right)$ as $\delta m_{K}/m_{K}\sim 10^{-15}$. At the fundamental level of a unitary evolution, without decays, the impact of Earth’s gravity appears as a CPT violation, with T conservation, because the indirect violation parameter $\left\langle K_{S}^{\oplus}\right.\left|K_{L}^{\oplus}\right\rangle$ = $2j\kappa$ is imaginary [4], rather than a CP and T violation with CPT conservation requiring a non zero real value of $\left\langle K_{S}^{\oplus}\right.\left|K_{L}^{\oplus}\right\rangle$ [4].

Usually, the three types of CPV experimental evidences are interpreted under the assumption of CPT conservation. The CPT theorem is demonstrated within the framework of three hypothesis: Lorentz group invariance, spin-statistics relations and local field theory. In the rest frame of a meson interacting with a massive spherical object, like Earth, the first hypothesis is not satisfied. Thus, when Earth influence is taken into account, we must not be surprised that CPT theorem, apparently, no longer holds. Within the framework of a GICPV mechanism Earth’s gravity is described as an external field and the evolution of a meson state $\left|M\right\rangle$ alone, as a linear superposition of two flavor eigenstates $\left|M^{0}\right\rangle$ and $\left|\overline{M}^{0}\right\rangle$, does not provide the complete picture of the dynamical system and so can not be considered as a good candidate displaying CPT invariance. Of course there are no CPT violation stricto sensu, CPT is restored for the global three bodies $\left(M^{0}/\overline{M}^{0}/\oplus\right)$ quantum evolution of the state $\left|M\left(\tau\right),\oplus\right\rangle$ describing both the meson-antimeson pair and Earth.

In this study we consider only the evolution of $\left|M\right\rangle$ and Earth’s effect is described as an external static field so that CPT will appear to be violated because of this restricted two bodies $\left(M^{0}/\overline{M}^{0}\right)$ model of a three bodies system $\left(M^{0}/\overline{M}^{0}/\oplus\right)$.

Moreover, with GICPV there is no T violation at the microscopic level. As demonstrated in the next section, the observed T violation stems from the irreversible decay of the short-lived kaons $K_{S}$ continuously regenerated from the long-lived one $K_{L}$ by gravity.

We must now take into account the $K_{S}$ fast decay. This decay will change the picture, qualitatively: an apparent CP and T violation, with CPT conservation rather than a CPT violation, and quantitatively: with the right prediction of $\mathop{\rm Re}\varepsilon$.

The lifetime of the $K_{L}$ is 577 times larger than the lifetime of $\ K_{S}$. The previous results (35) allows to calculate the gravity induced transition rate $\Omega_{L\rightarrow S}^{\oplus}$ describing the transition amplitude per unit time from the state $\left|K_{L}^{\oplus}\right\rangle$ $\exp-j\delta m_{K}c^{2}\tau/2\hbar$ to the state $\left|K_{S}^{\oplus}\right\rangle$ $\exp j\delta m_{K}c^{2}\tau/2\hbar$:

$$\Omega_{L\rightarrow S}^{\oplus}=\left\langle\frac{dK_{L}^{\oplus}}{d\tau}\left|K_{S}^{\oplus}\left(\tau\right)\right.\right\rangle=\kappa\frac{\delta m_{K}c^{2}}{\hbar}\exp j\frac{\delta m_{K}c^{2}}{\hbar}\tau\text{.}$$ (36)

This can be viewed as a gravity induced regeneration competing with the short-lived kaon irreversible decay. This decay, to the set of final states $\left\{\left|f\right\rangle\right\}$, takes place at a rate $\Gamma_{1\rightarrow f}/2=\left(\Gamma_{K}-\delta\Gamma_{K}\right)/2$ $\sim$ $\Sigma_{f}\left|\left\langle f\right|{\cal T}\left|K_{1}\right\rangle\right|^{2}$.

We consider now a typical experiment dedicated to indirect CPV. Experimentally $K_{1}$ and $K_{2}$ are initially produced together in equal amounts. Then, after few $1/$ $\Gamma_{1\rightarrow f}$ decay times, the initial content of $\left|K_{1}\right\rangle$ disappears and a pure $\left|K_{2}\right\rangle$ state is expected. In fact, the state $\left|K_{L\text{obs}}\left(\tau\right)\right\rangle$ observed in such an experiment is not a pure $\left|K_{2}\right\rangle$ state. This observed state $\left|K_{L\ \text{obs}}\left(\tau\right)\right\rangle$ is a linear superposition of $\left|K_{2}\right\rangle$, plus a small amount of $\left|K_{1}\right\rangle$,

$$\left|K_{L\text{obs}}\left(\tau\right)\right\rangle=a_{2}\left(\tau\right)\left|K_{2}\right\rangle+a_{1}\left(\tau\right)\left|K_{1}\right\rangle\text{,}$$ (37)

resulting from the balance between gravity induced regeneration (36) and the fast irreversible decay of the $K_{1}$ component. We assume that the $K_{2}$ component is stable and that the depletion of its amplitude associated with the gravitational regeneration of $K_{1}$ is negligible so that $\left|a_{2}\left(\tau\right)\right|$ = $1$ and

$$a_{2}\left(\tau\right)=\exp-j\delta m_{K}c^{2}\tau/2\hbar\text{.}$$ (38)

The amplitude $a_{1}$ of $K_{1}$ in (37) is given by the steady-state balance between a decay at the rate $\Gamma_{1\rightarrow f}/2$ on the one hand, and a gravity induced regeneration at the rate $\Omega_{L\rightarrow S}^{\oplus}$ (36) from $K_{2}$ on the other hand. This steady-state balance reads

$$a_{2}\left(\tau\right)\Omega_{L\rightarrow S}^{\oplus}=a_{1}\left(\tau\right)\frac{\Gamma_{1\rightarrow f}}{2}\text{.}$$ (39)

The solution is this equation is

$$a_{1}\left(\tau\right)=\frac{\delta m_{K}c^{2}}{\hbar\Gamma_{S}/2}\kappa\exp j\delta m_{K}c^{2}\tau/2\hbar\text{,}$$ (40)

where we have dropped the index $1\rightarrow f$ in $\Gamma$ to simplify the notation and used $\Gamma_{S}$. The short-lived $\left|K_{1}\right\rangle$ component is observed through its two pions decay [1].

Thus the observed long-lived mass eigenstate $\left|K_{L\text{obs}}\right\rangle$, obtained after few $1/$ $\Gamma_{S}$ decay times away from a neutral kaons source, must be represented by

$$\left|K_{L\text{obs}}\right\rangle=\left|K_{2}\right\rangle+\frac{\delta m_{K}c^{2}}{\hbar\Gamma_{S}/2}\kappa\left|K_{1}\right\rangle\text{.}$$ (41)

This is the usual CPV parametrization (12) of the mass eigenstates used under a CPT assumption.

The observed value of the indirect GICPV parameter,

$$\mathop{\rm Re}\varepsilon_{\text{obs}}=\frac{\delta m_{K}c^{2}}{\hbar\Gamma_{S}/2}\frac{2m_{K}g\hbar}{\delta m_{K}^{2}c^{3}}=1.66\times 10^{-3}\text{,}$$ (42)

is in agreement with the most recent experimental value [5]:

$$\mathop{\rm Re}\varepsilon_{PDG2024}=\left(1.66\pm 0.02\right)\times 10^{-3}\text{.}$$ (43)

To complete the previous analysis, we can also take into account the decay of the other mass eigenstate, and this will reveal a phenomenological dissipative phase of $\varepsilon$.

Considering the decay rates $\Gamma_{S}=\Gamma_{K}-\delta\Gamma_{K}$ for $K_{S}$, and $\Gamma_{L}$ = $\Gamma_{K}+\delta\Gamma_{K}$ for $K_{L}$ $(\delta\Gamma_{K}<0$), beside their usual definitions in terms of transition amplitudes, $\Gamma_{S/L}$ = $\Sigma_{f}\left|\left\langle f\right|{\cal T}\left|K_{S/L}\right\rangle\right|^{2}$, Bell and Steinberger have demonstrated a general relation based on global unitarity starting from the evaluation of $\ d\left\langle M\right.\left|M\right\rangle/d\tau$ at $\tau=0$ [15].

Using the fact that, for $K_{S}$, the sum $\Sigma_{f}$ over the final states is dominated (99.9%) by $K_{S}\rightarrow 2\pi$ decays, more precisely by the $K_{S}$ $\rightarrow I_{0}$ decays (95%) to the isospin-zero combination of $\left|\pi^{+}\pi^{-}\right\rangle$ and $\left|\pi^{0}\pi^{0}\right\rangle$, the Bell-Steinberger’s unitarity relations can be written [15]:

$$j\frac{\delta m_{K}c^{2}}{\hbar}+\frac{\Gamma_{S}}{2}=\frac{\left\langle I_{0}\right|{\cal T}\left|K_{L}\right\rangle\left\langle I_{0}\right|{\cal T}\left|K_{S}\right\rangle^{*}}{\left\langle K_{S}\right.\left|K_{L}\right\rangle}\text{.}$$ (44)

The restriction of $\sum_{f}\left|f\right\rangle$ to $\left|I_{0}\right\rangle$ reduces the $K_{S}$ width to $\Gamma_{S}$ = $\left\langle I_{0}\right|{\cal T}\left|K_{S}\right\rangle\left\langle I_{0}\right|{\cal T}\left|K_{S}\right\rangle^{*}$ so that

$$\frac{\left\langle I_{0}\right|{\cal T}\left|K_{L}\right\rangle}{\left\langle I_{0}\right|{\cal T}\left|K_{S}\right\rangle}=\frac{\left\langle I_{0}\right|{\cal T}\left|K_{L}\right\rangle\left\langle I_{0}\right|{\cal T}\left|K_{S}\right\rangle^{*}}{\Gamma_{S}}\text{.}$$ (45)

This expression is then substituted in Bell-Steinberger’s relation (44) to get the final expression

$$\frac{\left\langle I_{0}\right|{\cal T}\left|K_{L}\right\rangle}{\left\langle I_{0}\right|{\cal T}\left|K_{S}\right\rangle}=\frac{\left\langle K_{S}\right.\left|K_{L}\right\rangle}{2}\left(1+j\frac{2\delta m_{K}c^{2}}{\hbar\Gamma_{S}}\right)\text{.}$$ (46)

The left hand side of (46) can be considered as the definition of a complex indirect CPV parameter $\varepsilon$ and $\left\langle K_{S}\right.\left|K_{L}\right\rangle/2=\mathop{\rm Re}\varepsilon$ (42), thus the argument of this CPV complex parameter $\varepsilon$ is:

$$\arg\varepsilon=\arctan\left(2\delta m_{K}c^{2}/\hbar\Gamma_{S}\right)=43.4^{\circ}\text{,}$$ (47)

in agreement with the experimental result $43.5^{\circ}$ [5]. This last relation (47) complements (42) and confirms that GICPV provides a global and pertinent framework to interpret $K^{0}/\overline{K}^{0}$ indirect CPV experiments.

It is very important to note that the fundamental parameter describing indirect CPV is associated with the unitary evolution overlap of the mass eigenstates induced by Earth’s gravity:

$$\frac{\left\langle K_{S}^{\oplus}\right.\left|K_{L}^{\oplus}\right\rangle}{2}=j\frac{2m_{K}g\hbar}{\delta m_{K}^{2}c^{3}}\text{,}$$ (48)

and, as explained above, the measurements of the complex CPV parameter given by

$$\varepsilon=\frac{2m_{K}g\hbar}{\delta m_{K}^{2}c^{3}}\left[\frac{2\delta m_{K}c^{2}}{\hbar\Gamma_{S}}\left(1+j\frac{2\delta m_{K}c^{2}}{\hbar\Gamma_{S}}\right)\right]\text{,}$$ (49)

is due to a dissipative dressing of the fundamental overlap (48), dissipative dressing resulting from the finite lifetime of the mesons. This dissipative dressing is not stricto sensu a CPV effects but is inherent to experiments with unstable particles, this point is important to interpret CPV experiments and to understand the nature of GICPV.

VI Type (ii) CPV in the decay of $K^{0}/\overline{K}^{0}$

The analysis of type (ii) CPV in the decay to a final state $\left\langle f\right|$ rely on the measurement of the ratio $\eta_{f}=\left\langle f\right|{\cal T}\left|K_{L}\right\rangle/\left\langle f\right|{\cal T}\left|K_{S}\right\rangle$. To interpret the measurements of the direct violation parameter $\varepsilon^{\prime}$ we consider $\left\langle f\right|=\left\langle\pi^{0}\pi^{0}\right|$ and the $2\pi^{0}$ decays of $K_{L}$ and $K_{S}$ [16, 17, 18]. The definition of the direct CPV parameter $\varepsilon^{\prime}$, as a function of the amplitude ratio $\eta_{00}$, is given by

$$\eta_{00}=\frac{\left\langle\pi^{0}\pi^{0}\right|{\cal T}\left|K_{L}\right\rangle}{\left\langle\pi^{0}\pi^{0}\right|{\cal T}\left|K_{S}\right\rangle}\equiv\varepsilon-2\varepsilon^{\prime}\text{.}$$ (50)

The various bra and ket in a quantum model are defined up to an unobservable phase. The arbitrary conventional phases inherent to quantum theoretical models are to be eliminated to define phase-convention-independent observables.

The definition of $\eta_{00}$ is invariant under rephasing of the pions state $\left\langle\pi^{0}\pi^{0}\right|$, but not with respect to the rephasing of the kaons mass eigenstates $\left|K_{L/S}\right\rangle$. We can define a decay amplitude ratio which is a phase-convention-independent quantity through the multiplication of $\eta_{00}$ with the rephasing factor $\varphi_{K}$

$$\varphi_{K}=\frac{\left\langle K^{0}\right.\left|K_{S}\right\rangle}{\left\langle K^{0}\right.\left|K_{L}\right\rangle}.$$ (51)

Within the SM-CKM framework, CPT invariance is assumed and the observed mass eigenstates (12) are parametrized as:

$$\left|K_{S/L}\right\rangle=\frac{1+\varepsilon}{\sqrt{2}}\left|K^{0}\right\rangle\pm\frac{1-\varepsilon}{\sqrt{2}}\left|\overline{K}^{0}\right\rangle\text{.}$$ (52)

Within the GICPV framework, the mass eigenstates (35) display a different structure and are given by:

$$\left|K_{S/L}^{\oplus}\right\rangle=\frac{1\mp j\kappa}{\sqrt{2}}\left|K^{0}\right\rangle\pm\frac{1\pm j\kappa}{\sqrt{2}}\left|\overline{K}^{0}\right\rangle.$$ (53)

For the usual CPV parametrization (52) we obtain

$$\varphi_{K}=\frac{\left\langle K^{0}\right.\left|K_{S}\right\rangle}{\left\langle K^{0}\right.\left|K_{L}\right\rangle}=1\text{.}$$ (54)

For GICPV (53) we obtain

$$\varphi_{K}^{\oplus}=\frac{\left\langle K^{0}\right.\left|K_{S}^{\oplus}\right\rangle}{\left\langle K^{0}\right.\left|K_{L}^{\oplus}\right\rangle}=1-\left\langle K_{S}^{\oplus}\right.\left|K_{L}^{\oplus}\right\rangle\text{,}$$ (55)

where $O\left[10^{-6}\right]$ and higher orders terms are neglected.

The interaction between a $\left(\pi^{0},\pi^{0}\right)$ state and a neutral kaon state, $K^{0}$ or $\overline{K}^{0}$, can not differentiate the $K^{0}$ from the $\overline{K}^{0}$ (a final state phase can be absorbed by a proper phase convention between $K^{0}$ and $\overline{K}^{0}$), thus the amplitude of $K^{0}\rightarrow\pi^{0}\pi^{0}$ can be taken to be equal to the amplitude of $\overline{K}^{0}\rightarrow\pi^{0}\pi^{0}$.

Using equation (53), the ratio of amplitudes $\eta_{00}^{\oplus}$ associated with the unitary mass eigenstates resulting from GICPV is

$$\eta_{00}^{\oplus}=\frac{\left\langle\pi^{0}\pi^{0}\right|{\cal T}\left|K_{L}^{\oplus}\right\rangle}{\left\langle\pi^{0}\pi^{0}\right|{\cal T}\left|K_{S}^{\oplus}\right\rangle}=\frac{\left\langle K_{S}^{\oplus}\right.\left|K_{L}^{\oplus}\right\rangle}{2}.$$ (56)

We conclude that the (unitary $\widehat{\gamma_{K}}=\widehat{0}$) physical observable $\eta_{00}^{\oplus}\varphi_{K}^{\oplus}$ is given by

$$\eta_{00}^{\oplus}\varphi_{K}^{\oplus}=\frac{\left\langle K_{S}^{\oplus}\right.\left|K_{L}^{\oplus}\right\rangle}{2}\left[1-2\frac{\left\langle K_{S}^{\oplus}\right.\left|K_{L}^{\oplus}\right\rangle}{2}\right]\text{.}$$ (57)

This is the GICPV physical structure of the phase-convention-independent amplitude ratio $\eta_{00}$. However, finite lifetimes and decays are inherent to the experiments, this results in a measured amplitude ratio

$$\eta_{00\text{obs}}^{\oplus}\varphi_{K\text{obs}}^{\oplus}=\frac{\left\langle K_{S}\right.\left|K_{L}\right\rangle}{2}\left[1-2\frac{\left\langle K_{S}\right.\left|K_{L}\right\rangle}{2}\right].$$ (58)

Using the phase-convention-independent definition of $\varepsilon^{\prime}$ (50) within the SM-CKM framework, the measurement of $\eta_{00}$ is normally interpreted as a direct CPV

$$\eta_{00\text{obs}}^{\oplus}\varphi_{K\text{obs}}^{\oplus}=\varepsilon\left[1-2\frac{\varepsilon^{\prime}}{\varepsilon}\right]\text{.}$$ (59)

The relations (58) and (59) lead to the conclusion $\mathop{\rm Re}\left(\varepsilon^{\prime}/\varepsilon\right)=\mathop{\rm Re}\left(\varepsilon\right)$. The GI direct CPV parameter

$$\mathop{\rm Re}\left(\varepsilon^{\prime}/\varepsilon\right)_{\text{GICPV}}=\frac{\delta m_{K}c^{2}}{\hbar\Gamma_{S}/2}\kappa=1.66\times 10^{-3},$$ (60)

is in agreement with the most recent experimental value [5]:

$$\mathop{\rm Re}\left(\varepsilon^{\prime}/\varepsilon\right)_{\text{{\it PDG2024}}}=\left(1.66\pm 0.23\right)\times 10^{-3}.$$ (61)

The fact that $\mathop{\rm Re}\left(\varepsilon^{\prime}/\varepsilon\right)\sim\mathop{\rm Re}\left(\varepsilon\right)$ was considered, up to now, as a numerical coincidence and it finds here a simple explanation in term of a phase-convention-independent amplitude ratio within the framework of GICPV.

The precise definition of phase-convention-independent quantities, in order to clearly identify what is measured in an experiment, is also one of the key to interpret the experimental observation of interferences between mixing and decay in $B^{0}/\overline{B}^{0}$ CPV dedicated experiments.

VII Type (iii) CPV in the interference between mixing and decay

Up to 2001, the evidences of CPV where restricted to K mesons experiments and to the baryons asymmetry of the universe. In 2001 the first clear identification of CPV with B mesons experiments in B-factories was reported [19, 20]. The mass and width ordering associated with the $B^{0}/\overline{B}^{0}$ system are given by : $\delta m_{B}/m_{B}\sim 10^{-19}$ and $\delta m_{B}/\Gamma_{B}\sim 0.7$. The lifetime of the CP eigenstate $B_{1}$ is considered to be equal to the lifetime of the other CP eigenstate $B_{2}$ so that $\delta\Gamma_{B}=0$.

The most pronounced CPV effects in the $B^{0}/\overline{B}^{0}$ system are displayed through interference experiments dedicated to the study of the phase difference between the decay path $B_{0}\rightarrow f$ and the decay path $B_{0}\rightarrow\overline{B}^{0}\rightarrow f$ [21, 22, 23].

To set up an interpretation of these experiments we keep a finite lifetime $\Gamma_{B}^{-1}$ for both particles and consider the decay operator

$$\widehat{\gamma}_{B}=\Gamma_{B}\left[\left|B^{0}\right\rangle\left\langle B^{0}\right|+\left|\overline{B}^{0}\right\rangle\left\langle\overline{B}^{0}\right|\right]\text{,}$$ (62)

to describe the dissipative part of the $B^{0}/\overline{B}^{0}$ dynamics. Thus, we have to solve (26 )

$$j\hbar\frac{d\left|n^{\prime}\right\rangle}{d\tau}=\widehat{H^{\prime}}_{YB}\cdot\left|n^{\prime}\right\rangle-jm_{B}g\hbar\frac{d\widehat{x}}{d\tau}\ \cdot\widehat{H^{\prime}}_{YB}^{-1}\cdot\left|N\right\rangle\text{.}$$ (63)

The action of $\ jm_{B}g\hbar\frac{d\widehat{x}}{d\tau}\ \cdot\widehat{H^{\prime}}_{YB}^{-1}$ on the CP eigenstates $\left|B_{1}\right\rangle$ and $\left|B_{2}\right\rangle$ is

$$\ jm_{B}g\hbar\frac{d\widehat{x}}{d\tau}\ \cdot\widehat{H^{\prime}}_{YB}^{-1}\left|B_{2/1}\right\rangle=\pm\delta m_{B}c^{2}\varsigma\left(1\mp j\chi\right)\left|B_{1/2}\right\rangle\text{.}$$ (64)

where we have defined the real parameters $\chi=0.77$ and $\varsigma\sim O\left[10^{-6}\right]$ as

$$\chi=\delta m_{B}c^{2}/\hbar\Gamma_{B}\text{, }\varsigma=2m_{B}g\hbar/\delta m_{B}^{2}c^{3}\left(\chi+\chi^{-1}\right)\text{.}$$ (65)

In order to solve equation (63) and to express the mass eigenstates on Earth, we consider the CP eigenstates

$$\left|N\left(\tau\right)\right\rangle=\left|B_{2/1}\right\rangle\exp\mp j\frac{\delta m_{B}c^{2}\mp j\hbar\Gamma_{B}}{2\hbar}\tau\text{,}$$ (66)

which are also $\left(m_{B}\pm\delta m_{B}/2\right)$ mass eigenstate without CPV. The associated solution of (63) is

$$\left|n^{\prime}\left(\tau\right)\right\rangle=-\varsigma\left(1\mp j\chi\right)\left|B_{1/2}\right\rangle\exp\mp j\frac{\delta m_{B}c^{2}\mp j\hbar\Gamma_{B}}{2\hbar}\tau\text{.}$$ (67)

Thus, on Earth, the mass eigenstates $\left|B_{L/H}^{\oplus}\right\rangle$ are not the CP eigenstates $\left|B_{1/2}\right\rangle$, but

$$\left|B_{L/H}^{\oplus}\right\rangle=\left|B_{1/2}\right\rangle-\varsigma\left(1\pm j\chi\right)\left|B_{2/1}\right\rangle\text{.}$$ (68)

Using the flavor basis $\left[\left|B^{0}\right\rangle,\left|\overline{B}^{0}\right\rangle\right]$, rather than the CP basis $\left[\left|B_{1}\right\rangle,\left|B_{2}\right\rangle\right]$, these mass eigenstates (68) are expressed as

$$\left|B_{L/H}^{\oplus}\right\rangle=\frac{1-\varsigma\left(1\pm j\chi\right)}{\sqrt{2}}\left|B^{0}\right\rangle\pm\frac{1+\varsigma\left(1\pm j\chi\right)}{\sqrt{2}}\left|\overline{B}^{0}\right\rangle\text{.}$$ (69)

This GICPV results requires two real number, $\varsigma$ and $\chi$, to express the mass eigenstates $\left|B_{L/H}^{\oplus}\right\rangle$ although type (iii) standard SM-CKM parametrization (13) is based on a single angle $\beta$ and display a different structure

$$\left|B_{L/H}\right\rangle=\frac{\exp+j\beta}{\sqrt{2}}\left|B^{0}\right\rangle\pm\frac{\exp-j\beta}{\sqrt{2}}\left|\overline{B}^{0}\right\rangle\text{.}$$ (70)

To interpret the CPV experiments we consider the decay into one final CP eigenstate $\left|f\right\rangle$. The amplitude ratio $\lambda_{f}$

$$\lambda_{f}=\frac{\left\langle\overline{B}^{0}\right.\left|B_{L}\right\rangle}{\left\langle B^{0}\right.\left|B_{L}\right\rangle}\frac{\left\langle f\right|{\cal T}\left|\overline{B}^{0}\right\rangle}{\left\langle f\right|{\cal T}\left|B^{0}\right\rangle}=\exp-2j\beta\frac{\left\langle f\right|{\cal T}\left|\overline{B}^{0}\right\rangle}{\left\langle f\right|{\cal T}\left|B^{0}\right\rangle}\text{,}$$ (71)

is observable within the SM-CKM framework because the SM-CKM parametrization (70) is reduced to a single angle which leads to the relations:

$$\frac{\left\langle\overline{B}^{0}\right.\left|B_{L}\right\rangle}{\left\langle B^{0}\right.\left|B_{L}\right\rangle}=-\frac{\left\langle\overline{B}^{0}\right.\left|B_{H}\right\rangle}{\left\langle B^{0}\right.\left|B_{H}\right\rangle}=\frac{\left\langle\overline{B}^{0}\right.\left|B_{L}\right\rangle}{\left\langle B^{0}\right.\left|B_{H}\right\rangle}$$

$$=-\frac{\left\langle\overline{B}^{0}\right.\left|B_{H}\right\rangle}{\left\langle B^{0}\right.\left|B_{L}\right\rangle}=\exp-2j\beta\text{.}$$ (72)

However, these four amplitudes ratios are different if we consider the gravity induced mass eigenstates (69)

$$\frac{\left\langle\overline{B}^{0}\right.\left|B_{L}^{\oplus}\right\rangle}{\left\langle B^{0}\right.\left|B_{L}^{\oplus}\right\rangle}\neq-\frac{\left\langle\overline{B}^{0}\right.\left|B_{H}^{\oplus}\right\rangle}{\left\langle B^{0}\right.\left|B_{H}^{\oplus}\right\rangle}\neq\frac{\left\langle\overline{B}^{0}\right.\left|B_{L}^{\oplus}\right\rangle}{\left\langle B^{0}\right.\left|B_{H}^{\oplus}\right\rangle}\neq-\frac{\left\langle\overline{B}^{0}\right.\left|B_{H}^{\oplus}\right\rangle}{\left\langle B^{0}\right.\left|B_{L}^{\oplus}\right\rangle}\text{.}$$ (73)

Despite this difference between (72) and (73), the experimental results analyzed within a SM-CKM framework can be understood and explained within the framework of GICPV. This situation is similar to the one encountered in the previous section devoted to the study of $\varepsilon^{\prime}$: if CPT is assumed the rephasing factors $\varphi=1$, and the interpretation of the experimental measurements is based on the hypothesis of direct violation and imply a CPV at the fundamental level of the CKM matrix. However, if Earth’s gravity effects are taken into account $\varphi\neq 1$ and the very same phase-convention-independent measured quantities agree with the experiments without any additional assumption.

The analysis below will use two different approaches to interpret the measurement of $\ \beta$, each providing the same final result.

The two issues addressed below are: (i) the invariance under rephasing of the mass eigenstates to define an observable and (ii) the invariance under rephasing of the flavor eigenstates to define an observable.

In order to accommodate the relation (71) with (72, 73), we consider a $\widetilde{\lambda}_{f}$ parameter constructed with the amplitude ratio $\left\langle\overline{B}^{0}\right.\left|B_{L}\right\rangle/\left\langle B^{0}\right.\left|B_{H}\right\rangle$ which is better suited to characterize the dynamics of oscillating $B_{L/S}$ as it takes into account all the eigenstates: the two flavor eigenstates and the two mass eigenstates involved in experiments. However, this $\widetilde{\lambda}_{f}$ parameter:

$$\widetilde{\lambda}_{f}=\frac{\left\langle\overline{B}^{0}\right.\left|B_{L}\right\rangle}{\left\langle B^{0}\right.\left|B_{H}\right\rangle}\frac{\left\langle f\right|{\cal T}\left|\overline{B}^{0}\right\rangle}{\left\langle f\right|{\cal T}\left|B^{0}\right\rangle}=\lambda_{f}\text{,}$$ (74)

is not phase-convention-independent with respect to the mass eigenstates.

To set up a fully phase-convention-independent parameter we introduce the symmetric rephasing factor $\varphi_{B}$:

$$\varphi_{B}=\sqrt{\frac{\left\langle B_{1}\right.\left|B_{H}\right\rangle}{\left\langle B_{1}\right.\left|B_{L}\right\rangle}\frac{\left\langle B_{2}\right.\left|B_{H}\right\rangle}{\left\langle B_{2}\right.\left|B_{L}\right\rangle}}=1\text{.}$$ (75)

We have used $B_{1/2}$ states because they are CP eigenstates like $f$. The amplitude ratio observed in the experimental measurement are given by phase-convention-independent product $\widetilde{\lambda}_{f}\varphi_{B}$

$$\widetilde{\lambda}_{f}\varphi_{B}=\frac{\left\langle\overline{B}^{0}\right.\left|B_{L}\right\rangle}{\left\langle B^{0}\right.\left|B_{H}\right\rangle}\frac{\left\langle f\right|{\cal T}\left|\overline{B}^{0}\right\rangle}{\left\langle f\right|{\cal T}\left|B^{0}\right\rangle}\varphi_{B}=\exp-2j\beta\frac{\left\langle f\right|{\cal T}\left|\overline{B}^{0}\right\rangle}{\left\langle f\right|{\cal T}\left|B^{0}\right\rangle}$$ (76)

which is equal to $\lambda_{f}$ (71).

When the same rephasing factor $\varphi_{B}^{\oplus}$ is calculated within the framework of GICPV with (68) this gives

$$\varphi_{B}^{\oplus}=\sqrt{\frac{\left\langle B_{1}\right.\left|B_{H}^{\oplus}\right\rangle}{\left\langle B_{1}\right.\left|B_{L}^{\oplus}\right\rangle}\frac{\left\langle B_{2}\right.\left|B_{H}^{\oplus}\right\rangle}{\left\langle B_{2}\right.\left|B_{L}^{\oplus}\right\rangle}}=\sqrt{\frac{1-j\chi}{1+j\chi}}\text{.}$$ (77)

The phase-convention-independent product $\widetilde{\lambda}_{f}^{\oplus}\varphi_{B}^{\oplus}$ is defined as

$$\widetilde{\lambda}_{f}^{\oplus}\varphi_{B}^{\oplus}=\frac{\left\langle\overline{B}^{0}\right.\left|B_{L}^{\oplus}\right\rangle}{\left\langle B^{0}\right.\left|B_{H}^{\oplus}\right\rangle}\frac{\left\langle f\right|{\cal T}\left|\overline{B}^{0}\right\rangle}{\left\langle f\right|{\cal T}\left|B^{0}\right\rangle}\varphi_{B}^{\oplus}\text{,}$$ (78)

and is given by

$$\widetilde{\lambda}_{f}^{\oplus}\varphi_{B}^{\oplus}=\exp\left(-j\arctan\chi\right)\frac{\left\langle f\right|{\cal T}\left|\overline{B}^{0}\right\rangle}{\left\langle f\right|{\cal T}\left|B^{0}\right\rangle}\left(1+O\left[10^{-6}\right]\right)\text{.}$$ (79)

To compare the interpretations based on the usual SM-CKM eigenstates $\left|B_{L/H}\right\rangle$ (70) with the gravity induced mass eigenstates $\left|B_{L/H}^{\oplus}\right\rangle$ (69), we must define $\beta$ such that $2\beta=$ $\arctan\left(0.77\right)$. If $\left\langle f\right|{\cal T}\left|\overline{B}^{0}\right\rangle/\left\langle f\right|{\cal T}\left|B^{0}\right\rangle$ is assumed real and equal to one the experiments dedicated to the measure of $\lambda_{f}$ give a measurement equal to

$$\sin 2\beta=\sin\left[\arctan\left(0.77\right)\right]=0.61\text{.}$$ (80)

The modes $\overline{b}\rightarrow\overline{s}s\overline{s}$ and $\overline{b}\rightarrow\overline{c}c\overline{s}$ have been studied in depth through $B_{0}\rightarrow\phi K_{S}^{0}$ and $B_{0}\rightarrow\psi K^{0}$ interferences. According to the data reported in [5] the present status of the values is $\sin 2\beta_{\phi K_{S}^{0}}=0.58\pm 0.12\text{, }\sin 2\beta_{\psi K^{0}}=0.701\pm 0.01$. Other neutral final states, such as $J/\psi K^{*0}$ and $K^{0}\pi^{0}$, giving $0.60\pm 0.24\pm 0.08$ and $0.64\pm 0.13$, are in good agreement with the GICPV result (80) if $\left\langle f\right|{\cal T}\left|\overline{B}^{0}\right\rangle$ = $\left\langle f\right|{\cal T}\left|B^{0}\right\rangle$. For the full set of final states $f$ studied up to now, the results are centered around (80) but deviate from this value. The difficulty to evaluate $\arg\left\langle f\right|{\cal T}\left|\overline{B}^{0}\right\rangle/\left\langle f\right|{\cal T}\left|B^{0}\right\rangle$ is one source of the dispersion, note also that the sign of $\left\langle f\right|\widehat{CP}\left|f\right\rangle$ is to be considered and the fact that (70) is assumed rather than (69) is probably also a source of dispersion. A clear understanding of the $\sin 2\beta$ distribution around $0.6-0.7$ requires to drop (70) and to adopt the mass eigenstates (69); a precise evaluation of $\ \left\langle f\right|{\cal T}\left|\overline{B}^{0}\right\rangle/\left\langle f\right|{\cal T}\left|B^{0}\right\rangle$ is also needed.

Let us now consider a second point of view. We will not consider the interpretation of interferences experiments and, rather than addressing the issue of $\lambda_{f}$, we address directly the issue of $\beta$ through a gedanken experiment. We consider the different mass eigenstates expansions on either CP or flavor eigenstates: (13, 70) for the SM-CKM framework, and (68, 69) for the GICPV framework.

In order to compare the usual eigenstates parametrization (70), based on a single angle $\beta$, with the gravity induced mass eigenstates (69), involving two parameters $\varsigma$ and $\chi$, we must define $\beta$ through a gedanken experiment providing $\exp j\beta$ as a phase-convention-independent expression. We consider the symmetric and complete combination

$$\rho_{B}=\frac{\left\langle B^{0}\right.\left|B_{L}\right\rangle}{\left\langle\overline{B}^{0}\right.\left|B_{L}\right\rangle}\frac{\left\langle B^{0}\right.\left|B_{H}\right\rangle}{\left\langle\overline{B}^{0}\right.\left|B_{H}\right\rangle}\text{,}$$ (81)

which takes into account the four components at work in the description. This definition of $\beta$ through $\rho_{B}$ takes into account all flavor and mass eigenstates but suffers from a lack of (unphysical) phase compensation with respect to the flavor eigenstates. All measured observables, independently of the interpretation of the measurement, are combinations of phase-convention-independent quantities. We introduce the coefficient $\varphi_{B}^{\prime}$ needed to provide a phase-convention-independent observable associated with $\rho_{B}$

$$\varphi_{B}^{\prime}=\frac{\left\langle\overline{B}^{0}\right.\left|B_{2}\right\rangle\left\langle B_{2}\right.\left|B_{H}\right\rangle}{\left\langle B^{0}\right.\left|B_{1}\right\rangle\left\langle B_{1}\right.\left|B_{L}\right\rangle}\frac{\left\langle\overline{B}^{0}\right.\left|B_{2}\right\rangle\left\langle B_{2}\right.\left|B_{L}\right\rangle}{\left\langle B^{0}\right.\left|B_{1}\right\rangle\left\langle B_{1}\right.\left|B_{H}\right\rangle}\text{,}$$ (82)

where we have chosen the two projection operators $\left|B_{1}\right\rangle\left\langle B_{1}\right|$ and $\left|B_{2}\right\rangle\left\langle B_{2}\right|$ because they commute with CP.

It can be checked that the product $\rho_{B}\varphi_{B}^{\prime}$ is phase-convention-independent and thus can be measured in a gedanken experiment which does not need to be described here.

If the usual parametrization of CPV effects (13, 70) is used , this rephasing factor $\varphi_{B}^{\prime}$ changes nothing because it is equal to one

$$\rho_{B} = -\exp j4\beta\text{,}$$ (83)

$$\varphi_{B}^{\prime} = 1\text{.}$$ (84)

If CPV is gravity induced, we replace $\left|B_{H}\right\rangle$ and $\left|B_{L}\right\rangle$ with $\left|B_{H}^{\oplus}\right\rangle$ and $\left|B_{L}^{\oplus}\right\rangle$ given by (68, 69), and the very same observable $\rho_{B}\varphi_{B}^{\prime}$ is the product of the following factors

$$\rho_{B}^{\oplus} = -1+O\left[10^{-6}\right]\text{,}$$ (85)

$$\varphi_{B}^{\prime\oplus} = \frac{1+j\chi}{1-j\chi}=\exp\left(2j\arctan\chi\right)\text{.}$$ (86)

We conclude that, within the framework of GICPV, the measurement of the phase-convention-independent observable $\rho_{B}\varphi_{B}^{\prime}$ on Earth gives

$$\rho_{B}^{\oplus}\varphi_{B}^{\prime\oplus}=-\exp\left(2j\arctan\chi\right)\text{,}$$ (87)

although if the measurement of the very same phase-convention-independent observable $\rho_{B}\varphi_{B}^{\prime}$ is interpreted within the usual SM-CKM framework it defines $\beta$ as

$$\rho_{B}\varphi_{B}^{\prime}=-\exp j4\beta\text{.}$$ (88)

The conclusion of this $\rho_{B}$ gedanken experiment measurement, with two frameworks of interpretation, is that $\arctan\chi=2\beta$ and $\sin 2\beta=\sin\left[\arctan\left(0.77\right)\right]=0.61$.

The measurement of the angle $\beta$ is still one of the major subjects at the forefront of the studies related to the physics of the SM and beyond.

VIII Conclusion

On the basis of highly accurate predictions of both $\varepsilon$ (42, 49) and $\varepsilon^{\prime}$ (60), and of a relevant prediction of $\beta$ (80), we can state that GICPV offers a pertinent framework to interpret $K^{0}/\overline{K}^{0}$ and $B^{0}/\overline{B}^{0}$ experiments dedicated to CPV and that the CKM matrix should be considered free from any CPV phase far from any massive object.

Gravity induced Charge-Parity violation not only explain (i, ii and iii) CPV effects, and predict the associated observables, but it also renews, in depth, the baryons asymmetry (iv) cosmological issue.

The previous calculations on the impact of Earth gravity on neutral mesons oscillations can be extended to $D^{0}/\overline{D}^{0}$ $\sim\left(c\overline{u}\right)/\left(\overline{c}u\right)$ and $B_{s}^{0}/\overline{B_{s}}^{0}\sim\left(s\overline{b}\right)/\left(\overline{s}b\right)$. The framework of analysis of the experimental data on $D^{0}/\overline{D}^{0}$ and $B_{s}^{0}/\overline{B_{s}}^{0}$ is similar to the methods presented in section 5, 6 and 7. The parameters $m_{D}g\hbar/\delta m_{D}^{2}c^{3}$ and $m_{B_{s}}g\hbar/\delta m_{B_{s}}^{2}c^{3}$ for both mesons systems are very small so a type (i) indirect violation will be extremely difficult to observe. However type (ii) and type (iii) CPV can be analyzed on the basis of a clear definition of phase-convention-independent observables, similar to those identified in section 6 and 7, and will be considered in a forthcoming study.

In any environment where a flavored neutral mesons $\left|M\right\rangle$, with mass $m$, mass spliting $\delta m$ and Compton wavelength $\lambda_{C}$, experiences a gravity ${\bf g}$, i.e. in any curved space-time environment, a CP eigenstates mixing with amplitude $j\left(m/\delta m\right)^{2}\left|{\bf g}\right|\lambda_{C}/c^{2}$ will be observed. The first factor $m/\delta m$ is associated with electroweak and strong interactions, the second one is the product of a (wave)length, an acceleration and $c$, quantities related to space-time geometry rather than to electroweak or strong interactions. For a massive spherical object, with radius $R_{0}$ and Schwarzschild radius $R_{S}$, the mixing amplitude is given by $j\left(m/\delta m\right)^{2}\left(R_{S}/R_{0}\right)(\lambda_{C}/R_{0})$. The proportionality to $\left|{\bf g}\right|$ or $R_{S}$ indicates that this new CPV mechanism allows to set up cosmological evolution models predicting the strong asymmetry between the abundance of matter and the abundance anti-matter in our present universe [6].

Beside the problem of early baryogenesis, neutrinos oscillations near a spherical massive object might be revisited to explore the impact of the interplay between gravity and mixing.

The type (i) CPV observed with $K^{0}/\overline{K}^{0}$ stems from a gravity induced interplay between vertical quarks zitterbewegung oscillations at the velocity of light on the one hand and the strangeness oscillations $\left(\Delta S=2\right)$ on the other hand.

The type (ii) small CPV observed with $K^{0}/\overline{K}^{0}$ is associated with the SM-CKM, CPT invariant, interpretation of a GICPV and is elucidated through a careful analysis of the rephasing invariance of the observable $\eta_{00}$.

The large type (iii) CPV observed with $B^{0}/\overline{B}^{0}$ is associated with the SM-CKM, CPT invariant, interpretation of a GICPV, displaying a very small modulus and a significant phase, and is elucidated through a careful analysis of the rephasing invariance of the observable $\beta$.

When the mesons are considered stables, the evolution is unitary and there is no T violation at the fundamental level. The observed T violation stems from very the large phase space offered to the final states of forward transitions $M\rightarrow f$ so that backward transitions $f\rightarrow M$ can not be observed. The amplitudes $w_{f}$ in equation (5) can be considered within the framework of the WW approximation [13] as irreversible decays in equation (6).

The very large type (iv) CPV observed in our universe remains an open issue within the SM-CKM framework of interpretation, whereas GICPV displays the potential to set up cosmological evolution models in agreement with the present state of our universe.

This set of new results is obtained without any speculative assumption on new fields, and without any free parameters adjustment. From this convergence of results, we can conclude that a CKM matrix free of CPV phase is to be considered as the core of the SM in a flat Lorentzian environment and Earth’s gravity is the sole source of $\ \varepsilon$, $\varepsilon^{\prime}$ and $\beta$ CPV effects observed in $K^{0}/\overline{K}^{0}$ and $B^{0}/\overline{B}^{0}$ dedicated experiments.