betaFeedback

WorksPhysics and Astronomy

An interpretation for Aharonov-Bohm effect with classical electromagnetic theory

Abstract

The magnetic Aharonov-Bohm effect shows that charged particles may be affected by the vector potential in regions without any electric or magnetic fields [1]. The Aharonov-Bohm effect was experimentally confirmed [2-3] and has been found in many situations [4-6]. A common explanation is based on quantum mechanics, which states that the wavefunctions associated with the charges will accumulate a phase shift due to the vector potential. However, consensus about its nature and interpretation has not been achieved [7-14]. We here propose a simple but reasonable interpretation based on the theory for electromagnetic radiation and couplings [15]. The energy associated with a pulse radiator is divided into a Coulomb-velocity energy and a radiative energy, together with a macroscopic Schott energy accounting for the energy exchange between them. All these energies are expressed with terms including the potentials, so are the mutual coupling energies. There exists a force acting on the moving charges even though the fields completely vanish. This force makes the charges pass through the magnetic solenoid in different velocity with different path length, causing a phase shift the same as that obtained with quantum mechanics. The theory is originally aimed for providing an interpretation for electromagnetic radiation and mutual coupling. It is derived directly from the Maxwell theory with no modification but only substitution and reorganization.

The Aharonov-Bohm effect was first discussed in 1959¹. As shown in Figure 1, the magnetic field is confined within the magnetic solenoid. There are no magnetic fields or electric fields outside the solenoid. When charged particles pass the outside of the solenoid, there is a phase shift due to the nonzero static vector potential A(r)\mathbf{A}(\mathbf{r}). Consider two electron paths with the same start point and the same end point. One path goes in the upper side of the solenoid, the other in the lower side. The phase difference of the electrons traveling along the two paths have a phase difference dependent on the magnetic flux Φ\Phi in the solenoid,

Figure 1. Magnetic Aharonov-Bohm effect. Coherent electrons from the two slits pass through a long solenoid from its upper and lower side, and form interference patterns on the screen. Without magnetic field in the solenoid, the interference pattern is like the yellow palette with bright strip at the center. With magnetic field, we show with the new interpretation that the electrons deflect due to mutual electromagnetic coupling, and form the interference pattern like the red palette. With the conventional interpretation based on quantum mechanics, the phase difference in the wave functions is directly related to the vector potential and causes the shift of the interference pattern, not necessarily requiring the deflection of the electrons.

Δφ=eℏ∮A⋅dl=eℏΦ(1)\Delta\varphi=\frac{e}{\hbar}\oint\mathbf{A}\cdot\mathrm{d}\mathbf{l}=\frac{e}{\hbar}\Phi \tag*{(1)}

Aharonov-Bohm effect is considered as an important quantum effect. It was experimentally confirmed by A. Tonomura²⁻³, and has been found in various situations. Aharonov-Bohm effect implies that potentials may have a physical meaning instead of just a pure mathematical tool. However, this role of potentials has not been convincingly demonstrated¹². The theory and interpretation for Aharonov-Bohm effect are still controversial¹²,¹³.

This paper is not to discuss the difficulty involved in Aharonov-Bohm effect. Instead, we attempt to explain Aharonov-Bohm effect with macroscopic Maxwell theory as we believe that Aharonov-Bohm effect is a classical electromagnetic phenomenon that could be interpreted with classical electromagnetic theory. We have proposed a theory for electromagnetic radiation and mutual coupling¹⁵, in which the energy of a pulse radiator is divided into three parts: a Coulomb-velocity energy that appears and disappears with its sources simultaneously, a radiative energy that leaves its sources after being emitted, and a reversible macroscopic Schott energy that is responsible for energy exchange. All energies are expressed with terms including potentials. The electromagnetic mutual couplings between two radiators can be handled in a similar way. With this formulation, the mutual coupling energies are not necessary to be zero when the force fields are all zeros. In the case of Aharonov-Bohm effect, the coupled electromagnetic energy between the electron and the solenoid can be evaluated with the vector potential instead of fields. The mutual energy varies with the motion of the electron, introducing a force on the electron to make it deflect, as shown in Figure 1. In the following we will show that this may possibly explain the Aharonov-Bohm effect in a simple but reasonable way.

Energies of a pulse radiator

For a pulse radiator in vacuum in time period of 0≤t≤T0\leq t\leq T, its energy can be divided into three parts¹⁵,¹⁶ (Appendix § 5),

Wtot(t)=WρJ(t)+WS(t)+Wrad(t)(2)W_{\mathrm{tot}}(t)=W_{\rho J}(t)+W_{\mathrm{S}}(t)+W_{\mathrm{rad}}(t) \tag*{(2)}

Where we denote

WρJ(t)=∫Vs[12J(r,t)⋅A(r,t)+12ρ(r,t)φ(r,t)]dr(3)W_{\rho J}(t)=\int_{V_s}\left[\frac{1}{2}\mathbf{J}(\mathbf{r},t)\cdot\mathbf{A}(\mathbf{r},t)+\frac{1}{2}\rho(\mathbf{r},t)\varphi(\mathbf{r},t)\right]\mathrm{d}r \tag*{(3)}
WS(t)=∫V∞12∂∂t[D(r,t)⋅A(r,t)]dr(4)W_{\mathrm{S}}(t)=\int_{V_\infty}\frac{1}{2}\frac{\partial}{\partial t}\left[\mathbf{D}(\mathbf{r},t)\cdot\mathbf{A}(\mathbf{r},t)\right]\mathrm{d}r \tag*{(4)}
Wrad(t)=−∫V∞D(r,t)⋅∂A(r,t)∂tdr(5)W_{\mathrm{rad}}(t)=-\int_{V_\infty}\mathbf{D}(\mathbf{r},t)\cdot\frac{\partial\mathbf{A}(\mathbf{r},t)}{\partial t}\mathrm{d}r \tag*{(5)}

In these equations, A(r,t)\mathbf{A}(\mathbf{r},t) and ϕ(r,t)\phi(\mathbf{r},t) are respectively the vector and scalar potential at position r\mathbf{r} and time tt associated with current density J(r,t)\mathbf{J}(\mathbf{r},t) and charge density ρ(r,t)\rho(\mathbf{r},t) in domain VsV_s. D(r,t)\mathbf{D}(\mathbf{r},t) is the electric flux density. The two potentials are subject to Lorentz Gauge and their zero points are at infinity.

The energies can be numerically evaluated with integrations over the source domains[15] from which their non zero duration in time domain can be determined. As illustrated in Figure 2, the macroscopic radiative source exists in the region within the dot circle. We assume that the source is static for t<0t<0, becomes radiative for 0<t<T0<t<T, and disappears for t>Tt>T. WρJ(t)W_{\rho J}(t) is the Coulomb-velocity energy associated with the Coulomb fields and the velocity fields. It appears and disappears with its sources simultaneously. For t<0t<0, the static source has a Coulomb energy spreading in the whole space, and there is no velocity energy. For 0<t<T0<t<T, the source is radiative and the Coulomb-velocity energy varies, with its fluctuation transmitting like waves with light velocity in free space. As shown in Appendix § 6, WS(t)W_S(t) is similar to the Schott energy in the charged particle theory, so it is defined as macroscopic Schott energy. It associates with radiative sources and is zero for static sources. Different from the Coulomb-velocity energy, the Schott energy does not disappear immediately with its radiative sources but continue to exist for a short period for t<T+tmax⁡/2t<T+t_{\max}/2, where tmax⁡t_{\max} is the largest traveling time between two source points in VsV_s. Wrad(t)W_{\mathrm{rad}}(t) is the radiative energy at time tt. Its distribution region expands with the radiative wave to far region. However, the total radiative energy remains constant for t≥T+tmax⁡/2t\geq T+t_{\max}/2. There is a real radiative power flow passing through the observation surface always away from the source, together with a pseudo power flow due to the fluctuation of the Coulomb-velocity energy. The pseudo power flow may cross the observation surface forward if the Coulomb-velocity energy increases and backward if decreases. Note that all terms in the integrands in (3)-(5) consist of two quantities: the one in the left side is a source distribution or its fields, the one in the right side is the potential.

Schematic diagram of the energy distribution and transmission of a pulse radiator

Figure 2. Schematic diagram of the energy distribution and transmission of a pulse radiator. a, for 0≤t≤T0\leq t\leq T, the Coulomb energy spreads over the whole space (blue area). The radiative energy spreads over the yellow region, the fluctuation of the Coulomb-velocity energy also occurs in this region. The radiative fields interact with sources in their way propagating outward through the source region (red area). There is a Schott energy responsible for energy exchange in this period. b, for t≥T+tmax⁡/2t\geq T+t_{\max}/2, the sources have disappeared and the total Coulomb-velocity energy becomes zero, but its fluctuation travels with radiative fields (yellow region). The total Schott energy is zero. The zero-field region (white area) expands with time. c,d, the pseudo power flow due to the fluctuation of the Coulomb-velocity energy passes through the observation surface away from the source when the Coulomb-velocity energy increases, and toward source when the Coulomb-velocity energy decreases. The radiative power always crosses the observation surface away from the sources. The radiative fields travel to far region with a constant total radiative energy for t≥T+tmax⁡/2t\geq T+t_{\max}/2 until they encounter other sources.

Consider a system of two sources ρi(ri,t),Ji(ri,t)\rho_i(\mathbf{r}_i,t),\mathbf{J}_i(\mathbf{r}_i,t) in domain VsiV_{si}, i=1,2i=1,2. The total energy of the system consists of self-energies (i=ji=j) and mutual coupling energies (i≠ji\ne j),

Wtot(t)=∑i=12∑j=12Wtotij(t)=∑i=12∑j=12[WρJij(t)+WSij(t)+Wradij(t)](6)W_{\mathrm{tot}}(t)=\sum_{i=1}^{2}\sum_{j=1}^{2}W_{\mathrm{tot}}^{ij}(t)=\sum_{i=1}^{2}\sum_{j=1}^{2}\left[W_{\rho J}^{ij}(t)+W_S^{ij}(t)+W_{\mathrm{rad}}^{ij}(t)\right] \tag*{(6)}

where

WρJij(t)=∫Vsi[12Ji(ri,t)⋅Aj(ri,t)+12ρi(ri,t)ϕj(ri,t)]dri(7)W_{\rho J}^{ij}(t)=\int_{V_{si}}\left[\frac{1}{2}\mathbf{J}_i(\mathbf{r}_i,t)\cdot\mathbf{A}_j(\mathbf{r}_i,t)+\frac{1}{2}\rho_i(\mathbf{r}_i,t)\phi_j(\mathbf{r}_i,t)\right]d\mathbf{r}_i \tag*{(7)}
WSij(t)=∫Vsi12∂∂t[Di(ri,t)⋅Aj(ri,t)]dr(8)W_S^{ij}(t)=\int_{V_{si}}\frac{1}{2}\frac{\partial}{\partial t}\left[\mathbf{D}_i(\mathbf{r}_i,t)\cdot\mathbf{A}_j(\mathbf{r}_i,t)\right]d\mathbf{r} \tag*{(8)}
Wradij(t)=−∫VsiDi(ri,t)⋅∂Aj(ri,t)∂tdr(9)W_{\mathrm{rad}}^{ij}(t)=-\int_{V_{si}}\mathbf{D}_i(\mathbf{r}_i,t)\cdot\frac{\partial\mathbf{A}_j(\mathbf{r}_i,t)}{\partial t}d\mathbf{r} \tag*{(9)}

It can be interpreted that source-jj affects source-ii through its potentials. Note that if we consider the two radiators as a whole and treat them like a single radiator, (6)-(9) are completely in consistent with (2)-(5). Moreover, these formulae can be extended to systems consisting of multiple radiators.

An electromagnetic mutual coupling problem between two sources is shown in Figure 3a. Source-1 in Vs1V_{s1} is static for t≤0t\leq0 and t>Tt>T, and radiative for 0≤t≤T0\leq t\leq T, while source-2 is static itself. Consider a part of source (the yellow star in Figure 3a) in the source region Vs1V_{s1}. For 0≤t≤T0\leq t\leq T, the radiative fields from the yellow star propagate outwards and interact with other sources that they have encountered in region Vs1V_{s1} in their journey. Part of their radiative energy is transferred to those sources. Only after they have completely left the source region, the radiative fields from the yellow star can propagate with constant radiative energy, until they encounter source-2 and repeat the same interaction: change the Coulomb-velocity energy of source-2 and excite source-2 to emit radiative fields, as shown in Figure 3b. The macroscopic Schott energy is nonzero only in the energy exchanging stage. Although not illustrated in Figure 3, the variation of the Coulomb-velocity energy of the yellow star also affects the other sources in Vs1V_{s1} and source-2 when the fluctuation reaches there. As has been discussed in previous section, the radiative energy of source-1 induces a real radiative power flow and the fluctuation of the Coulomb-velocity energy of source-1 causes a pseudo power flow. They compose the total power flow that can be represented by the Poynting vector.

The radiative fields in electromagnetic radiation and mutual coupling

Figure 3. The radiative fields in electromagnetic radiation and mutual coupling. Source-1 is radiative in 0≤t≤T0\leq t\leq T and static for t≤0t\leq0 and t≥Tt\geq T, while source-2 is static. Their Coulomb energy spreads over the whole space (blue area) for t≤0t\leq0 and t≥Tt\geq T, while source-1 is radiative and static mutual couplings. a, for 0≤t≤T0\leq t\leq T, the radiative energy of source-1 has not yet reached source-2. The radiative fields from the yellow star interact with other sources which they have encountered in the way of traveling out of the source region Vs1V_{\mathrm{s1}}. A nonzero macroscopic Schott energy exists in this period responsible for energy exchange. b, source-1 has turned back to be static. The radiative fields of source-1 have reached source-2 and interact with it. Source-2 could be excited to create a radiative energy and its Coulomb-velocity energy may change too. There is also a mutual macroscopic Schott energy in this period responsible for energy exchanging.

The analysis also reveals that the electromagnetic radiation and mutual coupling are the same issue and can be handled in the same way. Multiple sources can be collectively regarded as a single source with multiple nonconnected parts. The energies involved are the same as those in the radiation process, as expressed in (2)-(9).

The radiative fields of source-1 have no influence on source-2 when they have left Vs1V_{s1} but have not yet reached Vs2V_{s2}. They just keep propagating in space and have no influence on sources in the space. However, when they have encountered any source in the space, all sources in the space may be affected soon or later through mutual couplings.

Schott energy and radiative energy

Schott energy was introduced first by Schott1 in 1912. The Schott energy of a moving charge ee in vacuum is (eq. (16) in Ref.18, eq. (4) in Ref.19)

Es(t)=−14πε02e23c3a⋅v(10)E_{\mathrm{s}}(t)=-\frac{1}{4\pi\varepsilon_{0}}\frac{2e^{2}}{3c^{3}}\mathbf{a}\cdot\mathbf{v} \tag*{(10)}

where v(t)\mathbf{v}(t) is its velocity, a(t)\mathbf{a}(t) is its acceleration, cc and ε0\varepsilon_{0} are respectively the light velocity and permittivity in the vacuum. As shown in Appendix § 6, by applying the Lienard-Wiechert potentials[20, 21] to the moving charge, the macroscopic Schott energy has the same form of (10). Making use of the relationship E=−∇ϕ−∂A/∂t\mathbf{E}=-\nabla\phi-\partial\mathbf{A}/\partial t and Lorentz Gauge ∇⋅A+c−2∂ϕ/∂t=0\nabla\cdot\mathbf{A}+c^{-2}\partial\phi/\partial t=0, Ws(t)W_{\mathrm{s}}(t) can be converted to an integral of the potentials,

Ws(t)=−ε04d2dt2∫V∞[c−2ϕ2(r,t)+A(r,t)⋅A(r,t)]dr(11)W_{\mathrm{s}}(t)=-\frac{\varepsilon_{0}}{4}\frac{d^{2}}{dt^{2}}\int_{V_{\infty}}\left[c^{-2}\phi^{2}(\mathbf{r},t)+\mathbf{A}(\mathbf{r},t)\cdot\mathbf{A}(\mathbf{r},t)\right]d\mathbf{r} \tag*{(11)}

Choosing a spherical coordinate system and expanding the potentials in power series of (n⋅v)(\mathbf{n}\cdot\mathbf{v}), it can be verified that Ws(t)≈Es(t)W_{\mathrm{s}}(t)\approx E_{\mathrm{s}}(t) for v=∣v∣≪cv=|\mathbf{v}|\ll c. Based on this relationship, and noting that both Ws(t)W_{\mathrm{s}}(t) and Es(t)E_{\mathrm{s}}(t) are full time derivatives, we consider that Ws(t)W_{\mathrm{s}}(t) is corresponding to Es(t)E_{\mathrm{s}}(t), and is termed as macroscopic Schott energy.

Interpretation for Aharonov-Bohm effect

As shown in Figure 4, an infinitely long magnetic solenoid with radius rsolr_{\mathrm{sol}}, is put in free space with its axis coinciding with the zz-axis. The magnetic field is B=μ0Jz^\mathbf{B}=\mu_{0}J\hat{\mathbf{z}} for r<rsolr<r_{\mathrm{sol}}, and B=0\mathbf{B}=0 for r>rsolr>r_{\mathrm{sol}}, where JJ is the static current density in the solenoid. The vector potential is found to be Asol=(A0/r)ϕ^\mathbf{A}_{\mathrm{sol}}=(A_{0}/r)\hat{\boldsymbol{\phi}}, with A0=0.5μ0Jrsol2A_{0}=0.5\mu_{0}Jr_{\mathrm{sol}}^{2}. An electron with charge ee and rest mass mm passes the solenoid with velocity v(t)\mathbf{v}(t). As the vector potential and the current density in the magnetic solenoid is static, the electromagnetic energy of the solenoid remains unchanged in the whole process. Except the fields of the electron itself, there is no external fields in its path, hence no Lorentz force acting on the electron. The kinetic energy of the electron will be changed due to the electro-magnetic mutual coupling between the electron and the solenoid. In this case, the electromagnetic energy of the electron can be expressed by

Figure 4. Deflection of electrons due to Aharonov-Bohm effect. a, coordinate system. The axis of the infinite long magnetic solenoid is put on the zz-axis, with an inner magnetic field in zz-direction. An electron travels from point r1\mathbf{r}_1 to r2\mathbf{r}_2 along two paths: l+l_{+} and l−l_{-}, with tangential unit l^\hat{\mathbf{l}}. b & c, the trajectories of the electron from the two slits. v(0)=2v(0)=2 cm/s for all electrons. Current density J\mathbf{J} in the solenoid: 20 mA/m for b, 40 mA/m for c.

W(r,t)=Wel(r,t)+Wcoup(r,t)(12)W(\mathbf{r},t)=W_{\mathrm{el}}(\mathbf{r},t)+W_{\mathrm{coup}}(\mathbf{r},t) \tag*{(12)}

where Wel(r,t)W_{\mathrm{el}}(\mathbf{r},t) is the total electromagnetic energy of the electron itself, and Wcoup(r,t)W_{\mathrm{coup}}(\mathbf{r},t) is the energy coupled from the solenoid to the electron. r\mathbf{r} is the position of the electron on its trajectory at time tt. The coupling energy Wcoup(r,t)W_{\mathrm{coup}}(\mathbf{r},t) can be considered as an external work WextW_{\mathrm{ext}} done on the electron by an external force. The energy balance equation on the electron is then[18],

dWextdt=ddt[EK+Es+Erad](13)\frac{dW_{\mathrm{ext}}}{dt}=\frac{d}{dt}\left[E_{K}+E_{\mathrm{s}}+E_{\mathrm{rad}}\right] \tag*{(13)}

which is the equation (17) in Ref. 18. EK=0.5mv⋅vE_{K}=0.5m\mathbf{v}\cdot\mathbf{v} is the kinetic energy. EsE_{\mathrm{s}} is the Schott energy given in (10). EradE_{\mathrm{rad}} is the radiative energy. Its time derivative is given by the famous Larmor formula,

dEraddt=14πε02e23c3a⋅a(14)\frac{dE_{\mathrm{rad}}}{dt}=\frac{1}{4\pi\varepsilon_{0}}\frac{2e^{2}}{3c^{3}}\mathbf{a}\cdot\mathbf{a} \tag*{(14)}

The mass of the electron is approximately expressed by (equation (2-2) in Ref. 21)

m=f14πε0e2r0c2(15)m=f\frac{1}{4\pi\varepsilon_{0}}\frac{e^{2}}{r_{0}c^{2}} \tag*{(15)}

where ff is a numerical constant of order 1 and r0r_{0} is the radius of the electron. For the moving electron, ∣a∣r0/c≪∣v∣|a|r_{0}/c\ll|v|, ∣a˙∣r0/c≪∣a∣|\dot{a}|r_{0}/c\ll|a|. Therefore, we have

2e23c3∣a⋅a∣≪2e23c2r0∣a⋅v∣2e23c3∣a⋅v∣≪r0c2e23c2r0∣a⋅v∣≪2e23c2r0∣a⋅a∣(16)\begin{aligned} \frac{2e^{2}}{3c^{3}}|\mathbf{a}\cdot\mathbf{a}|&\ll\frac{2e^{2}}{3c^{2}r_{0}}|\mathbf{a}\cdot\mathbf{v}|\\ \frac{2e^{2}}{3c^{3}}|\mathbf{a}\cdot\mathbf{v}|&\ll\frac{r_{0}}{c}\frac{2e^{2}}{3c^{2}r_{0}}|\mathbf{a}\cdot\mathbf{v}|\ll\frac{2e^{2}}{3c^{2}r_{0}}|\mathbf{a}\cdot\mathbf{a}| \tag*{(16)} \end{aligned}

Noting that ∂EK/∂t=ma⋅v\partial E_{K}/\partial t=m\mathbf{a}\cdot\mathbf{v}, we can immediately obtain that

dEraddt≪dEKdt,dEsdt≪dEKdt(17)\frac{dE_{\mathrm{rad}}}{dt}\ll\frac{dE_{K}}{dt},\qquad\frac{dE_{\mathrm{s}}}{dt}\ll\frac{dE_{K}}{dt} \tag*{(17)}

Consequently, we have

dWextdt≈dEKdt(18)\frac{dW_{\mathrm{ext}}}{dt}\approx\frac{dE_{K}}{dt} \tag*{(18)}

In Aharonov-Bohm situation, Wext=WcoupW_{\mathrm{ext}}=W_{\mathrm{coup}}. It is further checked in Appendix § 7 that Wcoup(t)=0.5v(t)⋅Asol(r)W_{\mathrm{coup}}(t)=0.5\mathbf{v}(t)\cdot\mathbf{A}_{\mathrm{sol}}(\mathbf{r}), hence, the power balance becomes

ddt[12mv(t)⋅v(t)]≈ddt[12v(t)⋅Asol(r)](19)\frac{d}{dt}\left[\frac{1}{2}m\mathbf{v}(t)\cdot\mathbf{v}(t)\right]\approx\frac{d}{dt}\left[\frac{1}{2}\mathbf{v}(t)\cdot\mathbf{A}_{\mathrm{sol}}(\mathbf{r})\right] \tag*{(19)}

Assume that the electron travels along its trajectory from (r1,t1)(\mathbf{r}_{1},t_{1}) to (r,t)(\mathbf{r},t). Integrating (19) over time [t1,t][t_{1},t] yields

mv2(t)−mv2(0)=ev(t)⋅Asol(r)−ev(0)⋅Asol(r1)(20)m\mathbf{v}^{2}(t)-m\mathbf{v}^{2}(0)=e\mathbf{v}(t)\cdot\mathbf{A}_{\mathrm{sol}}(\mathbf{r})-e\mathbf{v}(0)\cdot\mathbf{A}_{\mathrm{sol}}(\mathbf{r}_{1}) \tag*{(20)}

We consider the de Broglie wave for the electron in a diffraction situation. Substituting the momentum of the electron p=mv=ℏkp=m v=\hbar k into (20) yields

k(t)v(t)−k(0)v(0)=eℏv(t)⋅A(r)−eℏv(0)⋅A(r1)(21)k(t)\mathbf{v}(t)-k(0)\mathbf{v}(0)=\frac{e}{\hbar}\mathbf{v}(t)\cdot\mathbf{A}(\mathbf{r})-\frac{e}{\hbar}\mathbf{v}(0)\cdot\mathbf{A}(\mathbf{r}_1) \tag*{(21)}

where kk is the wavenumber. Note that on the trajectory of the electron we have dl=vdtd l=v d t, and v(t)⋅A(r)=∓v(t)⋅A(r)\mathbf{v}(t)\cdot\mathbf{A}(\mathbf{r})=\mp\mathbf{v}(t)\cdot\mathbf{A}(\mathbf{r}) for l±l_{\pm}, where l+l_{+} denote the upper path and l−l_{-} the lower path. k(t)v(t)dt=kdlk(t)\mathbf{v}(t)d t=k d l represents the phase shift of the electron over dld l. Integrating (21) over time along the two trajectories and subtracting the two results gives the phase difference of the electron traveling through the two paths,

Δφ=φ+−φ−=eℏ∮lA⋅dl=eℏΦ(22)\Delta\varphi=\varphi_{+}-\varphi_{-}=\frac{e}{\hbar}\oint_l\mathbf{A}\cdot d\mathbf{l}=\frac{e}{\hbar}\Phi \tag*{(22)}

which exactly agrees with (1).

We have calculated the trajectories of the electrons for the system using finite difference method to solve (20) with an initial velocity of 2 cm/s for all electrons. The results are shown in Fig.4b,c. As expected, larger deflection is observed for larger current in the solenoid.

Discussions

The theory proposed in Ref. 15 is originally aimed for interpreting the electromagnetic radiation and mutual coupling problems[22, 23, 24, 25, 26]. It has been shown here that a simple explanation for Aharonov-Bohm effect might be obtained based on the theory. Although there is no Lorentz force because of the absence of the fields, the analysis shows that the electrons indeed deflect due to the mutual electromagnetic coupling between the electrons and the solenoid, as has been predicted and confirmed by researchers[27, 28, 29]. We consider that this force is a kind of reactive force to the Lorentz force exerting on the current in the solenoid by the retarded magnetic field of the moving electron[7]. It can be justified by noting that for a moving electron in static vector potential, the energy coupled from the solenoid to the electron is not zero, i.e., Wreactcoup(t)+Wscoup(t)≠0W_{\mathrm{react}}^{\mathrm{coup}}(t)+W_{\mathrm{s}}^{\mathrm{coup}}(t)\ne0, as shown in Appendix C. This coupling energy can be considered as the external work done on the electrons by the reactive force. If the solenoid is shielded by a super conductor[2], the mutual energy coupled from the solenoid to a moving electron is not affected since the vector potential of the solenoid is not affected and the energy in this case is evaluated very close to the electron, as shown in (45).

If the moving electrons are replaced by a current distribution J(r,t)\mathbf{J}(\mathbf{r},t) outside the solenoid, we can show that the coupling energy can also be nonzero. Assume that the magnetic field and the electric flux density associated with J(r,t)\mathbf{J}(\mathbf{r},t) are respectively H(r,t)\mathbf{H}(\mathbf{r},t) and D(r,t)\mathbf{D}(\mathbf{r},t). Recalling that ∇×Asol=0\nabla\times\mathbf{A}_{\mathrm{sol}}=0 only at the region outside the solenoid, we can see that the energy coupled from the solenoid to the current source is not zero,

Wreactcoup(t)+Wscoup(t)=∫V∞12[J(r,t)+∂∂tD(r,t)]⋅Asol(r) dr=∫V∞12∇×H(r,t)⋅Asol(r) dr=∫V∞12H(r,t)⋅∇×Asol(r) dr=∫Vsol12H(r,t)⋅∇×Asol(r) dr≠0(23)\begin{aligned} W_{\mathrm{react}}^{\mathrm{coup}}(t)+W_{\mathrm{s}}^{\mathrm{coup}}(t) =\int_{V_{\infty}}\frac{1}{2}\left[\mathbf{J}(\mathbf{r},t)+\frac{\partial}{\partial t}\mathbf{D}(\mathbf{r},t)\right]\cdot\mathbf{A}_{\mathrm{sol}}(\mathbf{r})\,d r \\ &=\int_{V_{\infty}}\frac{1}{2}\nabla\times\mathbf{H}(\mathbf{r},t)\cdot\mathbf{A}_{\mathrm{sol}}(\mathbf{r})\,d r \\ &=\int_{V_{\infty}}\frac{1}{2}\mathbf{H}(\mathbf{r},t)\cdot\nabla\times\mathbf{A}_{\mathrm{sol}}(\mathbf{r})\,d r \\ &=\int_{V_{\mathrm{sol}}}\frac{1}{2}\mathbf{H}(\mathbf{r},t)\cdot\nabla\times\mathbf{A}_{\mathrm{sol}}(\mathbf{r})\,d r\ne0 \tag*{(23)} \end{aligned}

which may be interpreted that the magnetic fields of the source enter the solenoid and interact with the magnetic field inside the solenoid. If the solenoid is shielded by a super conductor, then the integration area is limited to the region outside the solenoid,

Wreactcoup(t)+Wscoup(t)=∫V∞−Vsol12[J(r,t)+∂∂tD(r,t)]⋅Asol(r) dr=∫V∞−Vsol12∇×H(r,t)⋅Asol(r) dr=−∮Ssol12[H(r,t)×Asol(r)]⋅n^ dS=−∮Ssol12[n^×Asol(r)]⋅H(r,t) dS(24)\begin{aligned} W_{\mathrm{react}}^{\mathrm{coup}}(t)+W_{\mathrm{s}}^{\mathrm{coup}}(t) =\int_{V_{\infty}-V_{\mathrm{sol}}}\frac{1}{2}\left[\mathbf{J}(\mathbf{r},t)+\frac{\partial}{\partial t}\mathbf{D}(\mathbf{r},t)\right]\cdot\mathbf{A}_{\mathrm{sol}}(\mathbf{r})\,d r \\ &=\int_{V_{\infty}-V_{\mathrm{sol}}}\frac{1}{2}\nabla\times\mathbf{H}(\mathbf{r},t)\cdot\mathbf{A}_{\mathrm{sol}}(\mathbf{r})\,d r \\ &=-\oint_{S_{\mathrm{sol}}}\frac{1}{2}\left[\mathbf{H}(\mathbf{r},t)\times\mathbf{A}_{\mathrm{sol}}(\mathbf{r})\right]\cdot\hat{\mathbf{n}}\,dS \\ &=-\oint_{S_{\mathrm{sol}}}\frac{1}{2}\left[\hat{\mathbf{n}}\times\mathbf{A}_{\mathrm{sol}}(\mathbf{r})\right]\cdot\mathbf{H}(\mathbf{r},t)\,dS \tag*{(24)} \end{aligned}

which could also be nonzero as Asol(r)\mathbf{A}_{\mathrm{sol}}(\mathbf{r}) has nonzero tangential component on the surface.

The derivation reveals that the result of (22) is valid for magnetic solenoids with arbitrary shapes as long as the vector potential is time invariant. As a matter of fact, the phase difference exists for any pair of paths that form a closed loop circling the solenoid. As the traveling time of the electron in different paths depends not only on the path length but also on the velocity, it is not necessary to bring a time delay between two electrons traveling through two different paths. Therefore, the phase difference is not explicitly dependent on the time delay as observed by researchers[29].

The interpretation could be applied for explaining the Aharonov-Bohm effect in cases with magnetic whiskers or rings[2, 31], only to note that an additional phase difference by the magnetic flux has to be added to the phase of the electrons without the magnetic flux. If the phase difference is π\pi, then the bright-dark strips in the interference pattern will exchange. Since the mutual electro-magnetic coupling energies expressed with potentials are also valid for time varying situations, it may provide a possible explanation for Aharonov-Bohm effect with time varying vector potentials. For Aharonov-Bohm effect in materials, we may replace the materials with equivalence sources at first, then handle the coupling problems with the equivalence sources in free space, and finally interpret the Aharonov-Bohm effect involved. It can be seen from (3) that the mutual coupling energy for an electron in a static scalar potential is ϕ(r)\phi(\mathbf{r}), so the main part of the force acting on the electron is e∇ϕ(r)=eEe\nabla\phi(\mathbf{r})=e\mathbf{E}, which is exactly the Coulomb force. This implies that electrons cannot be affected with the scalar potential alone in the absence of electric fields. Hence, the first version of electric Aharonov-Bohm effect2 may not exist unless it is described in other formulations.

The energies of a pulse radiator

The total electromagnetic energy in vacuum is

Wtot(t)=∫V∞12(E⋅D+H⋅B) dr(25)W_{\mathrm{tot}}(t)=\int_{V_{\infty}}\frac{1}{2}(\mathbf{E}\cdot\mathbf{D}+\mathbf{H}\cdot\mathbf{B})\,d\mathbf{r} \tag*{(25)}

Substituting E=−∇ϕ−∂A/∂t\mathbf{E}=-\nabla\phi-\partial\mathbf{A}/\partial t, and B=∇×A\mathbf{B}=\nabla\times\mathbf{A} into (17) yields

Wtot(t)=∫V∞(12ϕρ+12A⋅J)dr+∫V∞12∂∂t(D⋅A)dr−∫V∞D⋅∂A∂tdr+∮S∞12(A×H−ϕD)⋅n^dS(26)W_{\mathrm{tot}}(t)=\int_{V_{\infty}}\left(\frac{1}{2}\phi\rho+\frac{1}{2}\mathbf{A}\cdot\mathbf{J}\right)d\mathbf{r}+\int_{V_{\infty}}\frac{1}{2}\frac{\partial}{\partial t}(\mathbf{D}\cdot\mathbf{A})d\mathbf{r} -\int_{V_{\infty}}\mathbf{D}\cdot\frac{\partial\mathbf{A}}{\partial t}d\mathbf{r}+\oint_{S_{\infty}}\frac{1}{2}(\mathbf{A}\times\mathbf{H}-\phi\mathbf{D})\cdot\hat{\mathbf{n}}dS \tag*{(26)}

For a pulse radiator in period 0≤t≤T0\leq t\leq T, the surface integral in (26) is zero because the fields never reach S∞S_{\infty}. So we get (2)-(5).

The potentials and the electric flux density are defined in their usual forms as follow,

ϕ(r,t)=∫Vsρ(r1,t−R1/c)4πε0R1dr1(27)\phi(\mathbf{r},t)=\int_{V_s}\frac{\rho(\mathbf{r}_1,t-R_1/c)}{4\pi\varepsilon_0R_1}d\mathbf{r}_1 \tag*{(27)}
A(r,t)=μ0∫VsJ(r1,t−R1/c)4πR1dr1(28)\mathbf{A}(\mathbf{r},t)=\mu_0\int_{V_s}\frac{\mathbf{J}(\mathbf{r}_1,t-R_1/c)}{4\pi R_1}d\mathbf{r}_1 \tag*{(28)}
D(r,t)=−ε0∇ϕ(r,t)−ε0∂∂tA(r,t)=−∫Vs∫−∞∞ρ(r1,t1)∇G(t−t1−R1/c) dt1dr1\begin{aligned} \mathbf{D}(\mathbf{r},t)=-\varepsilon_0\nabla\phi(\mathbf{r},t)-\varepsilon_0\frac{\partial}{\partial t}\mathbf{A}(\mathbf{r},t) \\ &=-\int_{V_s}\int_{-\infty}^{\infty}\rho(\mathbf{r}_1,t_1)\nabla G(t-t_1-R_1/c)\,dt_1d\mathbf{r}_1 \end{aligned}
−1c2∫Vs∫−∞∞J(r1,t1)G˙(t−t1−R1/c) dt1dr1(29)-\frac{1}{c^2}\int_{V_s}\int_{-\infty}^{\infty}\mathbf{J}(\mathbf{r}_1,t_1)\dot{G}(t-t_1-R_1/c)\,dt_1d\mathbf{r}_1 \tag*{(29)}

where μ0\mu_0 is the permeability in vacuum. The time domain Green’s function can be expressed using the Dirac delta function,

G1(r,r1;t−R1/c)=δ(t−R1/c)4πR1(30)G_1(\mathbf{r},\mathbf{r}_1;t-R_1/c)=\frac{\delta(t-R_1/c)}{4\pi R_1} \tag*{(30)}

The mutual coupling macroscopic Schott energy can then be written as

Wsij(t)=∫V∞12∂∂t(Di⋅Aj)dr=∫V∞{μ0∫Vsj∫−∞∞ρi(ri,t1)∇Gi dt1dri−c−2∫Vsj∫−∞∞J(ri,t1)G˙i dt1dri}⋅∫−∞∞J(rj,t2)Gj dt2 dr(31)\begin{aligned} W_s^{ij}(t)=\int_{V_{\infty}}\frac{1}{2}\frac{\partial}{\partial t}(\mathbf{D}_i\cdot\mathbf{A}_j)d\mathbf{r} \\ &=\int_{V_{\infty}}\left\{\mu_0\int_{V_{sj}}\int_{-\infty}^{\infty}\rho_i(\mathbf{r}_i,t_1)\nabla G_i\,dt_1d\mathbf{r}_i -c^{-2}\int_{V_{sj}}\int_{-\infty}^{\infty}\mathbf{J}(\mathbf{r}_i,t_1)\dot{G}_i\,dt_1d\mathbf{r}_i\right\}\cdot\int_{-\infty}^{\infty}\mathbf{J}(\mathbf{r}_j,t_2)G_j\,dt_2\,d\mathbf{r} \tag*{(31)} \end{aligned}

where Gi,j=G(t−ti,j−Ri,j/c)G_{i,j}=G(t-t_{i,j}-R_{i,j}/c) and Ri,j=∣r−ri,j∣R_{i,j}=|\mathbf{r}-\mathbf{r}_{i,j}|. The following explicit expression[15, 24] is useful to evaluate the integrations for energies,

I=∫V∞G(τ−R1/c)G(τ−R2/c) dτ=c28πr21(32)I=\int_{V_\infty}G(\tau-R_1/c)G(\tau-R_2/c)\,d\tau=\frac{c^2}{8\pi r_{21}} \tag*{(32)}

where r21=∣R1−R2∣r_{21}=|R_1-R_2|. The integration domain for (t1,t2)(t_1,t_2) is determined in deriving (32)[15, 24]. With careful manipulation, the macroscopic Schott energy can be transformed to an integration over the source domain,

Wδψ(t)=−18πε0∫Vsi∫Vsj1rij×[∫t−rij/ct{ρ(ri,t)ρ(rj,2t−τ−rijc)+c−2J(ri,2t−τ−rijc)⋅J(rj,τ)}dτ]drjdri(33)W_{\delta}^{\psi}(t)=-\frac{1}{8\pi\varepsilon_0}\int_{V_{si}}\int_{V_{sj}}\frac{1}{r_{ij}}\times\left[\int_{t-r_{ij}/c}^{t}\left\{\rho(r_i,t)\rho\left(r_j,2t-\tau-\frac{r_{ij}}{c}\right)+c^{-2}\mathbf{J}(r_i,2t-\tau-\frac{r_{ij}}{c})\cdot\mathbf{J}(r_j,\tau)\right\}d\tau\right]dr_jdr_i \tag*{(33)}

For a pulse source in 0≤t≤T0\leq t\leq T, it can be checked that the integral becomes zero when t≥T+rij,max⁡/2ct\geq T+r_{ij,\max}/2c or t≤rij,min⁡/2ct\leq r_{ij,\min}/2c.

The other mutual coupling energies can be derived in the same way. For example, we can check that

WρJψ(t)=18πε0∫Vsi∫Vsj1rij[c−2Ji(ri,t)⋅Jj(rj,t−rijc)+ρi(ri,t)ρj(rj,t−rijc)]drjdri(34)W_{\rho J}^{\psi}(t)=\frac{1}{8\pi\varepsilon_0}\int_{V_{si}}\int_{V_{sj}}\frac{1}{r_{ij}}\left[c^{-2}\mathbf{J}_i(r_i,t)\cdot\mathbf{J}_j\left(r_j,t-\frac{r_{ij}}{c}\right)+\rho_i(r_i,t)\rho_j\left(r_j,t-\frac{r_{ij}}{c}\right)\right]dr_jdr_i \tag*{(34)}

Schott energy

The Lienard-Wiechert potentials[20, 21] for a moving charge are,

ϕ(r,t)=14πε0[ceRc−R⋅v]t′=e4πε0[1R(1−n⋅β)]t′(35)\phi(r,t)=\frac{1}{4\pi\varepsilon_0}\left[\frac{ce}{Rc-\mathbf{R}\cdot\mathbf{v}}\right]_{t'}=\frac{e}{4\pi\varepsilon_0}\left[\frac{1}{R(1-\mathbf{n}\cdot\boldsymbol{\beta})}\right]_{t'} \tag*{(35)}
A(r,t)=μ04π[ecvRc−R⋅v]t′=e4πε0c[βR(1−n⋅β)]t′(36)\mathbf{A}(r,t)=\frac{\mu_0}{4\pi}\left[\frac{e c\mathbf{v}}{Rc-\mathbf{R}\cdot\mathbf{v}}\right]_{t'}=\frac{e}{4\pi\varepsilon_0c}\left[\frac{\boldsymbol{\beta}}{R(1-\mathbf{n}\cdot\boldsymbol{\beta})}\right]_{t'} \tag*{(36)}

where R=r−x(t′)\mathbf{R}=r-x(t'), R=∣r−x(t′)∣R=|r-x(t')|, and x(t′)x(t') is a point on the trajectory of the moving charge. t′=t−R/ct'=t-R/c is the retarded time. Note that the quantities at the righthand side of (35) and (36) are evaluated at t′t'. v(t′)=dx(t′)dt′\mathbf{v}(t')=dx(t')dt', β=v/c\boldsymbol{\beta}=\mathbf{v}/c, and n=R/R\mathbf{n}=\mathbf{R}/R. For v=∣v∣≪cv=|\mathbf{v}|\ll c, the potential terms can be expanded to the second order in (n⋅β)(\mathbf{n}\cdot\boldsymbol{\beta}),

ϕ2(r,t)≈(e4πε0c)2[1R2(1+2n⋅β+3(n⋅β)2)]t′(37)\phi^2(r,t)\approx\left(\frac{e}{4\pi\varepsilon_0c}\right)^2\left[\frac{1}{R^2}\left(1+2\mathbf{n}\cdot\boldsymbol{\beta}+3(\mathbf{n}\cdot\boldsymbol{\beta})^2\right)\right]_{t'} \tag*{(37)}
A⋅A≈(e4πε0c)2[β⋅βR2]t′(38)\mathbf{A}\cdot\mathbf{A}\approx\left(\frac{e}{4\pi\varepsilon_0c}\right)^2\left[\frac{\boldsymbol{\beta}\cdot\boldsymbol{\beta}}{R^2}\right]_{t'} \tag*{(38)}

The integrand of (11) can be divided into two parts,

Ipo(r,t)=c−2ϕ2(r,t)+A(r,t)⋅A(r,t)=(e4πε0c)21R2[1+2n⋅β+3(n⋅β)2+β⋅β]=(e4πε0c)21R2+(e4πε0c)21R2[2n⋅β+3(n⋅β)2+β⋅β]\begin{aligned} I_{po}(r,t)=c^{-2}\phi^2(r,t)+\mathbf{A}(r,t)\cdot\mathbf{A}(r,t) \\ &=\left(\frac{e}{4\pi\varepsilon_0c}\right)^2\frac{1}{R^2}\left[1+2\mathbf{n}\cdot\boldsymbol{\beta}+3(\mathbf{n}\cdot\boldsymbol{\beta})^2+\boldsymbol{\beta}\cdot\boldsymbol{\beta}\right] \\ &=\left(\frac{e}{4\pi\varepsilon_0c}\right)^2\frac{1}{R^2}+\left(\frac{e}{4\pi\varepsilon_0c}\right)^2\frac{1}{R^2}\left[2\mathbf{n}\cdot\boldsymbol{\beta}+3(\mathbf{n}\cdot\boldsymbol{\beta})^2+\boldsymbol{\beta}\cdot\boldsymbol{\beta}\right] \end{aligned}
≜Ipo1+Ipo2(39)\triangleq I_{po1}+I_{po2} \tag*{(39)}

We are now to evaluate the Schott energy and the rate of radiation of the moving electron. We assume that the electron has nonzero velocity in a very short time period of [t′−dt′,t′][t'-dt',t'], and t′=t−ro/ct'=t-r_o/c. Here ror_o is a small distance from the electron at x(t′)x(t'). It can be checked that, at time tt, Ipo2I_{po2} is nonzero only within a thin shell VshellV_{shell}, which can be approximately modelled with a nonuniform spherical shell with radius ror_o to ro+c(1−n⋅β)dtr_o+c(1-\mathbf{n}\cdot\boldsymbol{\beta})dt, as shown in Figure 5. In the spherical coordinate system with origin locating at x(t′)x(t'), R=rR=r. We can write

Image not extracted

Figure 5. Integration region for evaluating Ws(t)W_s(t).

∫V∞Ipo2(r,t) dr=∫VshellIpo2(r,t) dr≈(e4πε0c)2∫02π∫0π∫roro+c(1−n⋅β)dt′{1r2[2n⋅β+3(n⋅β)2+β⋅β]}r2sin⁡θdrdθdφ≈{(e4πε0c)2∫02π∫0π[2n⋅β+(n⋅β)2+β⋅β]sin⁡θdθdφ}c dt′(40)\begin{aligned} \int_{V_\infty}I_{po2}(r,t)\,dr=\int_{V_{shell}}I_{po2}(r,t)\,dr \\ &\approx\left(\frac{e}{4\pi\varepsilon_0c}\right)^2\int_0^{2\pi}\int_0^\pi\int_{r_o}^{r_o+c(1-\mathbf{n}\cdot\boldsymbol{\beta})dt'}\left\{\frac{1}{r^2}\left[2\mathbf{n}\cdot\boldsymbol{\beta}+3(\mathbf{n}\cdot\boldsymbol{\beta})^2+\boldsymbol{\beta}\cdot\boldsymbol{\beta}\right]\right\}r^2\sin\theta d r d\theta d\varphi\\ &\approx\left\{\left(\frac{e}{4\pi\varepsilon_0c}\right)^2\int_0^{2\pi}\int_0^\pi\left[2\mathbf{n}\cdot\boldsymbol{\beta}+(\mathbf{n}\cdot\boldsymbol{\beta})^2+\boldsymbol{\beta}\cdot\boldsymbol{\beta}\right]\sin\theta d\theta d\varphi\right\}c\,dt' \tag*{(40)} \end{aligned}

A convenient choice is to put β\boldsymbol{\beta} on the zz-axis, then we have n⋅β=∣β∣cos⁡θ\mathbf{n}\cdot\boldsymbol{\beta}=|\boldsymbol{\beta}|\cos\theta and the integrand is symmetrical with zz-axis. By checking that ∫0πcos⁡θsin⁡θdθ=0\int_0^\pi\cos\theta\sin\theta d\theta=0, and ∫0πcos⁡2θsin⁡θdθ=2/3\int_0^\pi\cos^2\theta\sin\theta d\theta=2/3, we can obtain

∫VshellIpo2(r,t) dr=14πε04e23ε0c2β(t)⋅β(t′)dt′(41)\int_{V_{shell}}I_{po2}(r,t)\,dr=\frac{1}{4\pi\varepsilon_0}\frac{4e^2}{3\varepsilon_0c^2}\boldsymbol{\beta}(t)\cdot\boldsymbol{\beta}(t')dt' \tag*{(41)}

which also holds true when ro=0r_o=0 and t′=tt'=t, while we evaluate the integration at time tt with no retardation. Therefore, (41) can be rewritten as

∫VshellIpo2(r,t)dr=14πε04e23ε0c2β(t)⋅β(t)dt=14πε04e23ε0c3v⋅vdt(42)\int_{V_{shell}}I_{po2}(r,t)dr=\frac{1}{4\pi\varepsilon_0}\frac{4e^2}{3\varepsilon_0c^2}\boldsymbol{\beta}(t)\cdot\boldsymbol{\beta}(t)dt=\frac{1}{4\pi\varepsilon_0}\frac{4e^2}{3\varepsilon_0c^3}\mathbf{v}\cdot\mathbf{v}dt \tag*{(42)}

Ipo1I_{po1} is not directly related to the velocity. Its integration should be evaluated over the whole space instead of the shell. It is a constant and vanishes after differentiation with respect to time tt. Therefore, it can be deduced that

Ws(t)=−ε04ddt(ddt∫V∞Ipo2(r,t)dr)≈−14πε02e23c3a(t)⋅v(t)(43)W_s(t)=-\frac{\varepsilon_0}{4}\frac{d}{dt}\left(\frac{d}{dt}\int_{V_\infty}I_{po2}(r,t)dr\right)\approx-\frac{1}{4\pi\varepsilon_0}\frac{2e^2}{3c^3}\mathbf{a}(t)\cdot\mathbf{v}(t) \tag*{(43)}

Estimation of energies in Aharonov-Bohm Effect

The mutual energy coupled from the solenoid to the electron is

Wcoup(t)=WρJcoup(t)+WScoup(t)+Wradcoup(t)(44)W_{coup}(t)=W_{\rho J}^{coup}(t)+W_{S}^{coup}(t)+W_{rad}^{coup}(t) \tag*{(44)}

As the vector potential is static, and there is no scalar potential, making use of (7)–(9), we can show that the mutual coupled energy only contains two terms,

Wcoup(t)=12ev(t)⋅Asol(r)+12∂∂t∫V∞Del(r,t)⋅Asol(r)dr(45)W_{coup}(t)=\frac{1}{2}e\mathbf{v}(t)\cdot\mathbf{A}_{sol}(r)+\frac{1}{2}\frac{\partial}{\partial t}\int_{V_\infty}\mathbf{D}_{el}(r,t)\cdot\mathbf{A}_{sol}(r)dr \tag*{(45)}

The second term in right hand side is the mutual macroscopic Schott energy. It can be estimated with the Lienard-Wiechert vector potential (36) and the nonrelativistic electric flux density of the electron,

D(r,t)=e4πc2{c2(n−β)R2(1−n⋅β)3+n×[(n−β)×a]R(1−n⋅β)3}(46)\mathbf{D}(\mathbf{r},t)=\frac{e}{4\pi c^{2}}\left\{\frac{c^{2}(\mathbf{n}-\boldsymbol{\beta})}{R^{2}(1-\mathbf{n}\cdot\boldsymbol{\beta})^{3}}+\frac{\mathbf{n}\times[(\mathbf{n}-\boldsymbol{\beta})\times\mathbf{a}]}{R(1-\mathbf{n}\cdot\boldsymbol{\beta})^{3}}\right\} \tag*{(46)}

Since Asol\mathbf{A}_{sol} is static, ∂[Del(r,t)⋅Asol]/∂t=Asol⋅∂Del(r,t)/∂t\partial[\mathbf{D}_{el}(\mathbf{r},t)\cdot\mathbf{A}_{sol}]/\partial t=\mathbf{A}_{sol}\cdot\partial\mathbf{D}_{el}(\mathbf{r},t)/\partial t is nonzero only on spheres with radius rr that satisfying r=c(t−t′)r=c(t-t'). As the equation (18) should be balanced at the same time, we have to choose r=ror=r_o. Consequently, the volume integral becomes a surface integral,

∫Vshell∂∂t[Del⋅Asol]dr=∮So∂Del∂t⋅AsoldS(47)\int_{V_{\mathrm{shell}}}\frac{\partial}{\partial t}\left[\mathbf{D}_{el}\cdot\mathbf{A}_{sol}\right]d r=\oint_{S_o}\frac{\partial\mathbf{D}_{el}}{\partial t}\cdot\mathbf{A}_{sol}dS \tag*{(47)}

in which the time derivative of the electric flux density can be obtained from (46). Keeping the terms including the first order of a\mathbf{a} or β\boldsymbol{\beta} gives

∂∂tD(r,t)=∂t′∂t∂∂t′D(r,t)≈e4πc{c2R[3n(n⋅β)−β]−a+3n(n⋅a)}(48)\begin{aligned} \frac{\partial}{\partial t}\mathbf{D}(\mathbf{r},t)=\frac{\partial t'}{\partial t}\frac{\partial}{\partial t'}\mathbf{D}(\mathbf{r},t) \\ &\approx\frac{e}{4\pi c}\left\{\frac{c^{2}}{R}\left[3\mathbf{n}(\mathbf{n}\cdot\boldsymbol{\beta})-\boldsymbol{\beta}\right]-\mathbf{a}+3\mathbf{n}(\mathbf{n}\cdot\mathbf{a})\right\} \tag*{(48)} \end{aligned}

Taking Asol\mathbf{A}_{sol} as a constant vector on the small sphere, and making use of the relationships of ∫0πcos⁡θsin⁡θdθ=0\int_{0}^{\pi}\cos\theta\sin\theta d\theta=0 and ∫0πcos⁡2θsin⁡θdθ=2/3\int_{0}^{\pi}\cos^{2}\theta\sin\theta d\theta=2/3 again, we can check that

Wscoup(t)=∮SoDel⋅AsoldS≈0(49)W_{s}^{\mathrm{coup}}(t)=\oint_{S_o}\mathbf{D}_{el}\cdot\mathbf{A}_{sol}dS\approx0 \tag*{(49)}

References

  1. [2]Tonomura, A., Osakabe, N., Matsuda, T., Kawasaki, T., Endo, J., Yano, S., Yamada, H. Evidence for Aharonov–Bohm effect with magnetic field completely shielded from electron wave. Phys. Rev. Lett. 56, 792 (1986)pubmed.ncbi.nlm.nih.gov/10033287
  2. [7]Boyer, T. H. Interpretation of the Aharonov–Bohm effect. Am. J. Phys. 40, 56 (1972)DOI
  3. [15]Xiao, G. B. A theory for electromagnetic radiation and coupling. TechRxiv. Preprint. https://doi.org/10.36227/techrxiv.16686112. v2, (2021)
  4. [18]Grøn, Ø. The significance of the Schott energy for energy-momentum conservation of a radiating charge obeying the Lorentz–Abraham–Dirac equation. Am. J. Phys. 79, 1, 115–122 (2011)DOI
  5. [20]Jackson, J. D. Classical Electrodynamics. 3rd ed., John Wiley & Sons: New York, NY, USA, (1998)
  6. [21]Rohrlich, F. Classical charged particles. 3rd ed., World Scientific Publishing: Singapore (2007)
  7. [22]Collin, R. E. & Rothschild, S. Evaluation of antenna Q. IEEE Trans. Antennas Propag. 12, 1, 23–27 (1964)DOI
  8. [23]Yaghjian, A. D. Internal energy, Q-energy, Poynting’s theorem, and the stress dyadic in dispersive material. IEEE Trans. Antennas Propag. 55, 6, 1495–1505 (2007)DOI
  9. [24]Vandenbosch, G. A. E. Radiators in time domain—Part I: Electric, magnetic, and radiated energies. IEEE Trans. Antennas Propag. 61, 8, 3995–4003 (2013)DOI
  10. [25]Vandenbosch, G. A. E. Radiators in time domain—Part II: Finite pulses, sinusoidal regime and Q factor. IEEE Trans. Antennas Propag. 61, 8, 4004–4012 (2013)DOI
  11. [26]Xiao, G. B., Xiong, C., Huang, S., Liu, R., Hu, Y. A new perspective on the reactive electromagnetic energies and Q factors of antennas. IEEE Access 8, 999565,17370-173803 (2020)DOI
  12. [27]Shelankov, A. L. Magnetic force exerted by the Aharonov-Bohm line. Europhys. Lett. 43, 623 (1998)arxiv.org/abs/cond-mat/9802158
  13. [28]Berry, M. V. Aharonov-Bohm beam deflection: Shelankov’s formula, exact solution, asymptotics and an optical analogue. J. Phys. A: Math. Gen. 32, 5627 (1999)DOI
  14. [29]Keating, J. P. & Robbins, J. M. Force and impulse from an Aharonov-Bohm flux line. J. Phys. A: Math. Gen. 34, 807–827 (2001)
  15. [31]Tonomura, A. The quantum world unveiled by electron waves, World Scientific, Hackensack, NJ (1998).

Paper details

vlnc:W_c53p7fwp
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents