The quantization process was always tightly connected to the Hamiltonian formulation of classical mechanics. For non-Hamiltonian systems, traditional quantization algorithms turn out to be unsuitable. Numerous attempts to quantize non-Hamiltonian systems have shown that this problem is nontrivial and requires the development of new approaches. In this paper, we present the quantization methods that do not depend upon the Hamiltonian formulation of classical mechanics. Two approaches to the quantization of mechanical systems are considered: axiomatic and hydrodynamic. It is shown that the formal application of these approaches to the classical Hamilton-Jacobi theory allows obtaining the wave equation for the corresponding quantum system in natural way. Examples are considered that show the effectiveness of the proposed approaches, both for Hamiltonian and non-Hamiltonian systems. The spinor form of the relativistic Hamilton-Jacobi theory for classical particles is considered. It is shown that it naturally leads to the Dirac equation for the corresponding quantum particle and to its non-Hamiltonian generalization, the bispinor relativistic Kostin equation.
Citation: Sergey A. Rashkovskiy. Quantization of Hamiltonian and non-Hamiltonian systems[J]. Communications in Analysis and Mechanics, 2023, 15(2): 267-288. doi: 10.3934/cam.2023014
[1] | Antoine Bendimerad-Hohl, Ghislain Haine, Laurent Lefèvre, Denis Matignon . Stokes-Lagrange and Stokes-Dirac representations of $ N $-dimensional port-Hamiltonian systems for modeling and control. Communications in Analysis and Mechanics, 2025, 17(2): 474-519. doi: 10.3934/cam.2025020 |
[2] | Henryk Żoła̧dek . An example in Hamiltonian dynamics. Communications in Analysis and Mechanics, 2024, 16(2): 431-447. doi: 10.3934/cam.2024020 |
[3] | Richard Cushman . Normalization and reduction of the Stark Hamiltonian. Communications in Analysis and Mechanics, 2023, 15(3): 457-469. doi: 10.3934/cam.2023022 |
[4] | Tianfang Wang, Wen Zhang . Existence and concentration of homoclinic orbits for first order Hamiltonian systems. Communications in Analysis and Mechanics, 2024, 16(1): 121-146. doi: 10.3934/cam.2024006 |
[5] | Andrea Brugnoli, Ghislain Haine, Denis Matignon . Stokes-Dirac structures for distributed parameter port-Hamiltonian systems: An analytical viewpoint. Communications in Analysis and Mechanics, 2023, 15(3): 362-387. doi: 10.3934/cam.2023018 |
[6] | Cheng Yang . On the Hamiltonian and geometric structure of Langmuir circulation. Communications in Analysis and Mechanics, 2023, 15(2): 58-69. doi: 10.3934/cam.2023004 |
[7] | Long Ju, Jian Zhou, Yufeng Zhang . Conservation laws analysis of nonlinear partial differential equations and their linear soliton solutions and Hamiltonian structures. Communications in Analysis and Mechanics, 2023, 15(2): 24-49. doi: 10.3934/cam.2023002 |
[8] | Sergey A. Rashkovskiy . Nonlinear Pauli equation. Communications in Analysis and Mechanics, 2024, 16(1): 94-120. doi: 10.3934/cam.2024005 |
[9] | Sergey A. Rashkovskiy . Nonlinear Schrödinger equation with a short-range compensating field. Communications in Analysis and Mechanics, 2025, 17(2): 520-549. doi: 10.3934/cam.2025021 |
[10] | Isaac Neal, Steve Shkoller, Vlad Vicol . A characteristics approach to shock formation in 2D Euler with azimuthal symmetry and entropy. Communications in Analysis and Mechanics, 2025, 17(1): 188-236. doi: 10.3934/cam.2025009 |
The quantization process was always tightly connected to the Hamiltonian formulation of classical mechanics. For non-Hamiltonian systems, traditional quantization algorithms turn out to be unsuitable. Numerous attempts to quantize non-Hamiltonian systems have shown that this problem is nontrivial and requires the development of new approaches. In this paper, we present the quantization methods that do not depend upon the Hamiltonian formulation of classical mechanics. Two approaches to the quantization of mechanical systems are considered: axiomatic and hydrodynamic. It is shown that the formal application of these approaches to the classical Hamilton-Jacobi theory allows obtaining the wave equation for the corresponding quantum system in natural way. Examples are considered that show the effectiveness of the proposed approaches, both for Hamiltonian and non-Hamiltonian systems. The spinor form of the relativistic Hamilton-Jacobi theory for classical particles is considered. It is shown that it naturally leads to the Dirac equation for the corresponding quantum particle and to its non-Hamiltonian generalization, the bispinor relativistic Kostin equation.
The basic equation of nonrelativistic quantum mechanics, the Schrödinger equation, is a phenomenological equation: it is not derived from any first principles, but is, in fact, postulated. The traditional operator "derivation" of the Schrödinger equation from classical Hamiltonian mechanics [1,2] is nothing more than a spectacular mathematical technique and cannot be considered as a physical justification for this equation. This is due to the fact that the transition from the physical parameters of a classical system (energy E and generalized momenta P) to the corresponding mathematical operators according to the correspondence rules E→iℏ∂∂t and P→ℏi∂∂q is formal and has no physical basis. The only justification for such an approach is that when a number of formalities [2] are fulfilled, it leads to correct results. However, there is no physical explanation for this.
Numerous attempts are known to derive the Schrödinger equation from some first principles [3,4,5,6,7], but none of them can be considered physically convincing.
The situation is even more complicated with the quantization of non-Hamiltonian systems, i.e., with the derivation of the wave equation for a quantum system whose classical counterpart is non-Hamiltonian. Currently, there are several attempts to solve this problem [8,9,10,11], however, all of them refer to special cases of non-Hamiltonian systems and do not give a general recipe for quantization of non-Hamiltonian systems.
In this paper, we propose a visual, but also formal "derivation" of the Schrödinger equation for Hamiltonian systems, close to the operator formalism and, in fact, justifying the operator formalism. As will be shown, this method is naturally generalized to a wide class of non-Hamiltonian systems, and in a number of cases allows obtaining the corresponding wave equations. Along with this approach, the hydrodynamic method of "deriving" the wave equations of nonrelativistic quantum mechanics and the method of transition from relativistic classical mechanics to relativistic quantum mechanics are also considered.
In the general case, the problem can be formulated as follows: there is a certain classical mechanical system described by known equations (for example, the Hamilton-Jacobi equation and/or Hamilton equations for Hamiltonian systems or, in the general case for a non-Hamiltonian system, Newton's equation). It is necessary to obtain the equations describing the quantum analogue of this system.
We will use the axiomatic approach. Let us introduce a number of postulates: (i) the state of a quantum system is described by a complex-valued wave function
ψ(t,q)=√ρexp(iS/iSℏℏ) | (1) |
which satisfies some partial differential equation (wave equation); here ρ(t,q) and S(t,q) are real-valued functions; (ii) the wave equation must be linear one with respect to the wave function ψ(t,q); (iii) in the limit ℏ→0, the wave equation must lead to the Hamilton-Jacobi theory for the corresponding classical system, i.e. it should decompose into the Hamilton-Jacobi equation for the function S(t,q), which plays the role of an action, and the continuity equation, in which the function ρ(t,q) plays the role of the density of the Hamilton-Jacobi ensemble [12]. Note that the Hamilton-Jacobi equation can be written for any Hamiltonian system [12].
Some "reason" for such an approach can be the analysis of the connection between the Schrödinger equation and the classical theory of Hamilton-Jacobi [13,14,15].
Let the classical Hamilton-Jacobi ensemble [12] be described by the Hamilton-Jacobi equation
∂S∂t+H(t,q,∂S∂q)=0 | (2) |
and the continuity equation in the configuration space
∂ρ∂t+∑k∂∂qk(ρ∂H(t,q,P)∂Pk)=0 | (3) |
where the Hamiltonian H(t,q,P), is in general an arbitrary function of the generalized momenta P=∂S∂q.
Then it is easy to show that the equation
[iℏ∂∂t−H(t,q,ℏi∂∂q)]ψ=0 | (4) |
which is linear with respect to the wave function ψ(t,q), in the limit ℏ→0 decomposes into the Hamilton-Jacobi equation (2) and the continuity equation (3).
Indeed, considering the operators ∂∂qk as ordinary variables, we expand the function H(t,q,ℏi∂∂q) formally in a power series:
H(t,q,ℏi∂∂q)=H(t,q,P0)+∑kHk(t,q,P0)(ℏi∂∂qk−P0k)++12∑k,nHkn(t,q,P0)(ℏi∂∂qk−P0k)(ℏi∂∂qn−P0n)+... |
where Hk(t,q,P0)=∂H(t,q,P)∂Pk|P=P0, Hkn(t,q,P0)=∂2H(t,q,P)∂Pk∂Pn|P=P0, P0(t,q) is some vector, the ellipsis means the terms of the expansion of the third and higher order with respect to (ℏi∂∂qk−P0k).
Then equation (4) can be written in the form
iℏ∂ψ∂t−H(t,q,P0)ψ−∑kHk(t,q,P0)(ℏi∂∂qk−P0k)ψ+−12∑k,nHkn(t,q,P0)(ℏi∂∂qk−P0k)(ℏi∂∂qn−P0n)ψ−...=0 |
Substituting here the wave function in the form (1), we obtain
iℏ1√ρ∂√ρ∂tψ−∂S∂tψ−H(t,q,P0)ψ−ψ∑kHk(t,q,P0)(ℏi1√ρ∂√ρ∂qk+∂S∂qk−P0k)+−12ψ∑k,nHkn(t,q,P0)ℏi∂∂qk(ℏi1√ρ∂√ρ∂qn+∂S∂qn−P0n)+−12ψ∑k,nHkn(t,q,P0)(ℏi1√ρ∂√ρ∂qk+∂S∂qk−P0k)(ℏi1√ρ∂√ρ∂qn+∂S∂qn−P0n)−...=0 |
Separating the real and imaginary parts, in the limit ℏ→0, one obtains
iℏ2ρ[∂ρ∂t+∑kHk(t,q,P0)∂ρ∂qk+ρ∑k,nHkn(t,q,P0)∂∂qk(∂S∂qn−P0n)++∑k,nHkn(t,q,P0)∂ρ∂qk(∂S∂qn−P0n)+...]−−[∂S∂t+H(t,q,P0)+∑kHk(t,q,P0)(∂S∂qk−P0k)++12∑k,nHkn(t,q,P0)(∂S∂qk−P0k)(∂S∂qn−P0n)+...]+o(ℏ2)=0 | (5) |
where o(ℏ2) means small terms of orders of ℏ2 and above. The terms in square brackets do not depend on ℏ.
It is easy to see that the square brackets contain the expansions in the series:
∑kHk(t,q,P0)∂ρ∂qk+ρ∑k,nHkn(t,q,P0)∂∂qk(∂S∂qn−P0n)++∑k,nHkn(t,q,P0)∂ρ∂qk(∂S∂qn−P0n)+...=∑k∂∂qk(ρ∂H(t,q,P)∂Pk) |
and
H(t,q,P0)+∑kHk(t,q,P0)(∂S∂qk−P0k)++12∑k,nHkn(t,q,P0)(∂S∂qk−P0k)(∂S∂qn−P0n)+...=H(t,q,P) |
where P=∂S∂q.
Then equation (5) takes the form
iℏ2ρ[∂ρ∂t+∑k∂∂qk(ρ∂H(t,q,P)∂Pk)]−[∂S∂t+H(t,q,P)]+o(ℏ2)=0 |
and in the limit ℏ→0 splits into two real-valued equations (2) and (3).
Thus, equation (4) satisfies the postulates formulated above and can be considered as a wave equation describing the corresponding quantum system.
The above reasoning can be considered as some "substantiation" of the operator formalism in the derivation of the wave equations of quantum mechanics.
Note that this method of reasoning leads to a wave equation describing a quantum system if the Hamilton-Jacobi equation for the corresponding classical system is known.
The approach described above can also be applied to a more general class of the system which are non-Hamiltonian systems if the requirement that the wave equation be linear with respect to the wave function is abandoned.
Let the classical Hamilton-Jacobi ensemble be described by a general Hamilton-Jacobi equation [12]
∂S∂t+F(t,q,S,∂S∂q)=0 | (6) |
Note that the function F(t,q,S,∂S∂q) in Eq. (6) is not a Hamiltonian function of the system under consideration.
We assume that in this case the corresponding quantum system is also described by the wave function (1). Then, as is easy to show by analogy with the case of the Hamiltonian system considered above, the equation
[iℏ∂∂t−F(t,q,−iℏln(ψ/ψ|ψ||ψ|),ℏi∂∂q)]ψ=0 | (7) |
in the limit ℏ→0 splits into the Hamilton-Jacobi equation (6) and the continuity equation
∂ρ∂t+∑k∂∂qk(ρ∂F(t,q,S,P)∂Pk)=0 | (8) |
for some classical Hamilton-Jacobi ensemble. For this reason, Eq (7) should be considered as a "wave" equation describing the quantum analogue of the classical system (6). In this case, obviously, Eq (7) is no longer linear.
According to (8), the velocity of the Hamilton-Jacobi ensemble in the configuration space is determined by the relation
˙qk=∂F(t,q,S,P)∂Pk |
which generalizes the first equation of Hamilton [16].
As an example, we apply the described method to a nonrelativistic particle moving in a potential field in the presence of drag force. Let us consider the case when the drag force depends linearly on the momentum of the particle: F=−∂U∂q−(β/βmm)P, where U is the potential energy of the particle in an external field; β is the constant drag factor; m is the particle mass.
The Hamilton-Jacobi equation for this system has the form [12]
∂S∂t+12m|∇S|2+βmS+U(t,r)=0 | (9) |
Equation (9) formally has the form (6), where
F(t,r,S,∇S)=12m|∇S|2+βmS+U(t,r) |
As a result, wave equation (7) takes the form
iℏ∂ψ∂t=−ℏ22m∇2ψ−iℏβmψln(ψ/ψ|ψ||ψ|)+U(t,r)ψ=0 | (10) |
This equation can be rewritten as
iℏ∂ψ∂t=−ℏ22m∇2ψ+Uψ−iβℏ2mψln(ψ/ψψ∗ψ∗) | (11) |
and is the known Kostin equation [17,18], where ψ∗ is the complex conjugate wave function.
The axiomatic approach described above, as well as the standard operator method of "deriving" the Schrödinger equation, are applicable only to those systems for which the Hamilton-Jacobi equation is known for the classical analogue. In particular, these methods are not applicable to arbitrary non-Hamiltonian systems that do not belong to the class (6).
Various approaches have been proposed for quantization of non-Hamiltonian systems [8,9,10,11,19,20,21,22], however, it seems to us that the most natural and understandable from a physical point of view is the hydrodynamic approach [20], which can be easily substantiated on the basis of the theory Hamilton-Jacobi ensemble [12].
Consider first a single quantum particle moving in an external potential field.
Substituting the wave function in the form (1) into the corresponding Schrödinger equation, and separating the real and imaginary parts, we obtain the equation
∂S∂t+12m(∇S)2+U−Uq=0 | (12) |
and the continuity equation
∂ρ∂t+div(1mρ∇S)=0 | (13) |
where
Uq=−ℏ22m∇2√ρ√ρ | (14) |
is the so-called quantum potential.
Equation (12) is formally the Hamilton-Jacobi equation for a particle, which is affected by an external classical potential U and an additional quantum potential (14), which has no classical analogue.
Equations (12), (13) describe an ensemble of identical noninteracting particles, which, by analogy with [12], we will call the quantum Hamilton-Jacobi ensemble.
Thus, we come to the conclusion that the quantum Hamilton-Jacobi ensemble moves under the action of an additional potential force
Fq=−∇Uq | (15) |
This allows formulating a general quantization rule for any mechanical nonrelativistic systems [20].
Let there be a classical mechanical system whose motion is described by Newton's equation
mdvdt=F(t,r,v) | (16) |
where F(t,r,v) is the external force;
drdt=v | (17) |
is the velocity of the particle.
The classical Hamilton-Jacobi ensemble corresponding to this system is described by the Euler equation [12]
m(∂v∂t+(v∇)v)=F(t,r,v) | (18) |
and the continuity equation
∂ρ∂t+divρv=0 | (19) |
An additional "quantum force" (15) acts on the corresponding quantum particle, and formally the motion of a quantum particle can be described by Newton's equation
mdvdt=F(t,r,v)−∇Uq |
Then the corresponding quantum Hamilton-Jacobi ensemble is described by the Euler equation
m(∂v∂t+(v∇)v)=F(t,x,v)−∇Uq | (20) |
with quantum potential (14) and continuity equation (19).
We assert that the system of equations (14), (19) and (20) is an analogue of the Schrödinger equation for any (including non-Hamiltonian) nonrelativistic systems.
To demonstrate this, we consider a few examples.
In this case F=−∇U and equation (20) takes the form
m(∂v∂t+(v∇)v)=−∇U−∇Uq | (21) |
By analogy with [12], we reduce Eq. (21) to the form
m(∂v∂t−[vrotv])=−∇(12mv2+U+Uq) | (22) |
This equation has a potential solution v(t,r)=1m∇S(t,r), where the function S(t,r) satisfies equation (12), (14). Equations (12)-(14) are equivalent to the Schrödinger equation for the wave function (1), which proves the formal equivalence of equations (14), (19) and (20) to the Schrödinger equation.
Consider a charged particle subjected to the Lorentz force F=e(E+1c[vH]).
In this case, equation (20), by analogy with [12], can be written as
∂∂t(mv+qcA)+∇(mV22+eφ+Uq)=[vrot(mv+ecA)] | (23) |
This equation has a potential solution mv+ecA=∇S, where the function S(t,r) satisfies the equation
∂S∂t+12m(∇S−ecA)2+eφ+Uq=0 | (24) |
It is easy to check that equation (24) with quantum potential (14), together with the continuity equation (19), is equivalent to the Schrödinger equation for a charged particle in an electromagnetic field
iℏ∂ψ∂t=[12m(ℏi∇−ecA)2+eφ]ψ |
If the particle is under the action of a classical force F(t,r,v)=−∇U−βv, then equation (20), by analogy with [12], can be written as
m(∂v∂t−[vrotv])=−∇(12mv2+U+Uq)−βv |
This equation has a potential solution v(t,r)=1m∇S(t,r), where the function S(t,r) satisfies the equation
∂S∂t+12m(∇S)2+U+βmS+Uq=0 | (25) |
It is easy to see that equation (25) with quantum potential (14) together with the continuity equation (19) are equivalent to one complex-valued equation for the wave function (1)
iℏ∂ψ∂t=−ℏ22m∇2ψ+Uψ+βmSψ |
This equation can also be written in the form (10) or (11) and is the Kostin equation [17,18].
Equations of motion of the classical particles have the form
midvidt=Fi(t,r1,...rN,v1,...,vN) | (26) |
dridt=vi | (27) |
where Fi(t,r1,...rN,v1,...,vN) is the force which acts on i th particle.
The Hamilton-Jacobi ensemble for a system of N interacting particles is characterized by a density ρ(t,q) in the 3N -dimensional configuration space q=(r1,...,rN). This density satisfies the continuity equation
∂ρ∂t+N∑i=1∇i(ρvi)=0 | (28) |
We assume that the particles are in an external classical potential field U(t,q), which includes the interaction of particles with each other.
For the corresponding quantum system of many particles, the continuum equation of motion (20) has the form [12]
mi(∂vi∂t+N∑k=1(vk∇k)vi)=−∇iU−∇iUq | (29) |
where
Uq=−ℏ22N∑i=11mi∇2i√ρ√ρ | (30) |
is the quantum potential of a system of quantum particles, which is a generalization of quantum potential (14).
Equation (29) has a potential solution vi(t,q)=1mi∇iS(t,q),i=1,...,N, where the function S(t,q) satisfies the equation
∂S∂t+N∑k=112mk|∇kS|2+U+Uq=0 | (31) |
It is easy to check directly that equation (31) with quantum potential (30), together with the continuity equation (28), is equivalent to the many-particle Schrödinger equation
iℏ∂ψ∂t=−ℏ22N∑i=11mi∇2iψ+U(t,q)ψ |
for the wave function (1) defined in 3N -dimensional configuration space q=(r1,...,rN).
The description of physical objects in relativistic quantum mechanics is fundamentally different from the description of their classical counterparts.
In classical physics, all parameters describing physical objects are real-valued, and, depending on the nature of the physical object, they can be scalar, vector, or, in general, tensor functions of spatial coordinates and time.
In quantum mechanics, on the contrary, all physical objects are described by complex-valued wave functions, which, depending on the nature of the object, are either scalars or spinors of various ranks.
In classical physics, there is not a single physical object that would be described by spinors in natural way. For this reason, it is believed that the spinor form of writing the equations of motion is a feature of the exclusively quantum mechanical description of microworld objects, and has no analogues in classical physics. This is primarily due to the fact that the rank of a spinor that describes a quantum object is related to the value of the spin of this object, a purely quantum property that classical particles do not have.
At the same time, the spinor form of writing the laws of nature is in fact not an exclusive property of quantum objects, but is only one of the alternative forms of writing equations describing physical objects.
For example, Maxwell equations describing the classical electromagnetic field can be written in spinor form, similar to the Dirac equation [23,24].
In this connection, it is of interest to consider the spinor formulation of classical mechanics. This allows establishing a closer connection between relativistic classical mechanics and relativistic quantum mechanics, in particular with the Dirac equation.
In classical relativistic mechanics, the 4-velocity of a particle uμ=dxμds, by definition, satisfies the condition [25]
uμuμ=1 | (32) |
where uμ=(u0,v/vcc), uμ=gμνuν, gμν=diag(1,−1,−1,−1), μ,ν=0,1,2,3.
The equations of motion of a charged particle in an electromagnetic field in a four-dimensional form are [25]
duμds=emc2Fμνuν | (33) |
where e is the electric charge of the particle, m is its rest mass,
Fμν=∂Aν∂xμ−∂Aμ∂xν | (34) |
is the electromagnetic field tensor; Aμ is the 4-potential of the electromagnetic field.
Let us pass to the relativistic Hamilton-Jacobi ensemble [12]. This means that we consider a 4-velocity field uμ(x) and hence duμds=dxνds∂uμ∂xν=uν∂uμ∂xν. Equation (33) corresponds to the Lagrangian description of the ensemble; it can be rewritten in the form of the 4-dimensional Euler equation
uν∂uμ∂xν=emc2Fμνuν | (35) |
To this equation, it is necessary to add the continuity equation for the Hamilton-Jacobi ensemble
∂Jμ∂xμ=0 | (36) |
where Jμ=ρ0uμ is the particle flux density in the Hamilton-Jacobi ensemble; ρ0 is a 4-scalar, which is related to the density ρ of the Hamilton-Jacobi ensemble in the considered inertial reference frame by the relation [25]
ρ=ρ0cdtds |
Equations (35) and (36) are a general relativistic formulation of the Hamilton-Jacobi theory for a charged particle in an electromagnetic field.
From them, it is easy to obtain the well-known relativistic Hamilton-Jacobi equation.
Taking into account (34), we rewrite (35) in the form
uν[∂∂xν(uμ+emc2Aμ)−emc2∂Aν∂xμ]=0 | (37) |
It follows from relation (32) that
uν∂uν∂xμ=0 | (38) |
Subtract from the left side of (37) the expression uν∂uν∂xμ; taking into account (38), one obtains
uν[∂∂xν(uμ+emc2Aμ)−∂∂xμ(uν+emc2Aν)]=0 | (39) |
Obviously, one of the solutions to equation (39) is
mcuμ+ecAμ=−∂S∂xμ |
where S is the 4-scalar.
Thus
mcuμ=−∂S∂xμ−ecAμ | (40) |
Substituting (40) into (32) and (36), we obtain the relativistic Hamilton-Jacobi equation [25]
gμν(∂S∂xμ+ecAμ)(∂S∂xν+ecAν)=m2c2 | (41) |
and the continuity equation for the Hamilton-Jacobi ensemble
gμν∂∂xν[ρ0(∂S∂xμ+ecAμ)]=0 | (42) |
Equations (41) and (42) are the traditional formulation of the relativistic Hamilton-Jacobi theory for a classical charged particle, which naturally follows from the general formulation of the theory of motion of the Hamilton-Jacobi ensemble (35) and (36).
It is easy to check that the Klein-Gordon equation
gμν(iℏ∂μ−ecAμ)(iℏ∂ν−ecAν)ψ=m2c2 | (43) |
with respect to the wave function
ψ=√ρ0exp(iS/iSℏℏ) | (44) |
obtained by the formal replacement ∂S∂xμ→−iℏ∂μ in the Hamilton-Jacobi equation (41), satisfies the postulates of Section 2: at ℏ→0, equation (43) splits into the Hamilton-Jacobi equation (41) and the continuity equation (42), which describe the classical Hamilton-Jacobi ensemble.
Let us show that the relativistic theory of the Hamilton-Jacobi ensemble for a classical charged particle, considered in the previous section, can be written in the spinor form, which is completely equivalent to its traditional formulation (41), (42).
Consider the equation
(γμuμ−1)Ψ=0 | (45) |
where 1 is the identity matrix; γμ are the Dirac matrices; Ψ=col(ψ1, ψ2, ψ3, ψ4) is the some column matrix.
Relations (45) can be considered as a system of linear homogeneous algebraic equations with respect to unknown parameters ψ1,ψ2,ψ3,ψ4.
If relation (45) is assumed to be Lorentz invariant, then it follows from the transformation properties of uμ and γμ that the matrix Ψ is a bispinor.
Using α -matrices
α0=(100−1),αα=(0σσσσ0) |
equation (45) can be rewritten as [26]
(u0−ααu−α0)Ψ=0 | (46) |
In expanded form, the system of equations (46) looks as follows:
(u0−1)ψ1−u3ψ3−(u1−iu2)ψ4=0(u0−1)ψ2−(u1+iu2)ψ3+u3ψ4=0u3ψ1+(u1−iu2)ψ2−(u0+1)ψ3=0(u1+iu2)ψ1−u3ψ2−(u0+1)ψ4=0 | (47) |
The determinant of this system of equations is
Δ=(uμuμ−1)2 |
In order for the system of equations (47) to have nontrivial solutions Ψ≠0, the determinant of the system must be equal to zero: (uμuμ−1)2=0.
As a result, we come to the conclusion that the system of equations (45) has nontrivial solutions Ψ≠0 only if condition (32) is satisfied. In other words, for nonzero matrices Ψ, relation (32) is a trivial consequence of equation (45).
The same result can be obtained more formally using the standard representation [26].
We introduce spinors φ=(ψ1ψ2) and χ=(ψ3ψ4). Then Ψ=(φχ). Taking this into account, equation (46) decomposes into two spinor equations
(u0−1)φ−uσσχ=0 | (48) |
(u0+1)χ−uσσφ=0 | (49) |
From equation (49), one obtains
χ=uσσφu0+1 | (50) |
Substituting χ from (50) into equation (48), one obtains
(u0−1)φ−(uσσ)(uσσ)φu0+1=0 |
Taking into account that (uσσ)(uσσ)=u2 [26], one obtains
(uμuμ−1)φ=0 |
For φ≠0 we again arrive at equation (32).
Substituting relation (40) into equation (45), for the classical Hamilton-Jacobi ensemble, one obtains
(γμ(∂S∂xμ+ecAμ)+mc)Ψ=0 | (51) |
The equivalence of equations (32) and (45) also means the equivalence of the Hamilton-Jacobi equation (41) and the spinor equation (51). Thus, equation (51) for nonzero matrices Ψ is an equivalent (spinor) form of the relativistic Hamilton-Jacobi equation for a classical charged particle in an external electromagnetic field.
As is known [2,26], from the bispinor Ψ and the Dirac matrix γμ, one can construct a real-valued 4-vector
Jμ≡(J0,J)=¯ΨγμΨ | (52) |
where ¯Ψ=Ψ+γ0, Ψ+ is the matrix Hermitian conjugate to the matrix Ψ.
Using the spinors φ and χ, as well as the connection between them (50), after simple transformations one obtains
J0=2|φ|2u0+1u0,J=2|φ|2u0+1u |
or
Jμ=2|φ|2u0+1uμ | (53) |
Similarly, a 4-scalar can be formed from the bispinor Ψ and the Dirac matrices γμ [2,26]
ρ0=¯ΨΨ |
which can be converted to the form
ρ0=2|φ|2u0+1 | (54) |
Using (53) and (54), one can write
Jμ=ρ0uμ | (55) |
Taking into account (40), one obtains
Jμ=−ρ0mc(∂S∂xμ+ecAμ) | (56) |
Equation (45) implies that the bispinor Ψ is defined up to a factor, which can be an arbitrary 4-scalar function. This allows us to choose the 4-scalar ρ0 in such a way that the 4-vector (55) determines the particle flux density in the Hamilton-Jacobi ensemble, which satisfies the continuity equation (36) or, in expanded form, equation (42).
Taking into account (54) and (40), relation (56) can be rewritten using the spinor φ in the form
Jμ=2|φ|2(∂S∂xμ+ecAμ)∂S∂x0+ecA0−mc | (57) |
If the action S is known (for example, found by solving the Hamilton-Jacobi Eq (41)), then the continuity equation (36) with the 4-vector (57) allows finding the modulus of the spinor φ, and, using relation (50), the modulus of the spinor χ.
Thus, we have shown that for a classical charged particle moving in an external electromagnetic field, the Hamilton-Jacobi theory can be formulated in spinor form, while the corresponding Hamilton-Jacobi ensemble is described by the Hamilton-Jacobi equation in the spinor representation (51) and the continuity equation (36) for the particle flux density (57), which is expressed in terms of the bispinor Ψ and the Dirac matrix by the same expression (52) as in the Dirac theory.
Note that in the non-relativistic limit, when |u|<<1 and 1−u0<<1, it will be χ<<φ, taking into account that the density of the Hamilton-Jacobi ensemble ρ=ρ0u0, one obtains
ρ≈ρ0≈|φ|2 |
The connection of the spinor form of writing (36), (51) and (52) of the theory of the classical Hamilton-Jacobi ensemble with the Dirac equation is obvious.
Indeed, we represent the components of the bispinor Ψ as
ψn=|ψn|exp(iαn+iSn/iSnℏℏ) | (58) |
where Sn and αn are real-valued functions having finite limits at ℏ→0; n=1,2,3,4.
We set the conditions:
limℏ→0ℏ∂μ|ψn|=0,limℏ→0ℏ∂μSn=0,limℏ→0ℏ∂μαn=0,limℏ→0ℏαn=0 | (59) |
Then, it is easy to show that the Dirac equation
(γμ(iℏ∂μ−ecAμ)−mc)Ψ=0 |
obtained by a formal replacement ∂S∂xμ→−iℏ∂μ in equation (51), at ℏ→0 decomposes into equation (51), which is the spinor form of the Hamilton-Jacobi equation for a classical particle, and the continuity equation (36), (52).
Thus, we see that the postulates introduced in Section 2 are valid both in the nonrelativistic and relativistic cases, which is the justification for the operator formalism in relativistic quantum mechanics.
In the general case, the relativistic equation of particle motion has the form [25]
mcduμds=Fμ | (60) |
where the 4-force Fμ satisfies the condition
uμFμ=0 | (61) |
In the case of particle motion in an external electromagnetic field, the 4-force (33), (34) automatically satisfies condition (61).
We consider a relativistic motion of a particle in an external electromagnetic field in the presence of drag force. The corresponding nonrelativistic case was considered in [12].
In the relativistic case, equation (60), taking into account (33), has the form
mcduμds=ecFμνuν+fμ | (62) |
where the drag 4-force fμ=(f0,f), according to (61), satisfies the condition
uμfμ=0 | (63) |
The simple 4-vector fμ generalizing the nonrelativistic drag force [12] and satisfying condition (63) has the form
fμ=βuν[(uν+emc2Aν)Uμ−(uμ+emc2Aμ)Uν] | (64) |
where Uμ=(U0,V/Vcc) is the constant 4-vector; β is the constant 4-scalar (drag factor).
It is easy to check that the 4-force (64) in the non-relativistic approximation (v2/v2c2c2<<1 and U2/U2c2c2<<1) transforms into a linear resistance law
f=−βmc[m(v−V)+ecA] | (65) |
considered in [12] for a particle moving in an external electromagnetic field.
Obviously, the system described by equations (62), (64) is non-Hamiltonian one.
For the Hamilton-Jacobi ensemble, using the continuum description, we reduce equation (62) to the Euler equation
uν∂uμ∂xν=emc2Fμνuν+1mcfμ | (66) |
Equations (66), (34) together with the continuity Eq (36) form the Hamilton-Jacobi theory [12] for the relativistic motion of a particle in the presence of a "linear" drag force.
Substituting (34) and (64) into equation (66), we transform it by analogy with Section 5.1 to the form
uν(∂Pμ∂xν−∂Pν∂xμ−βmc(PνUμ−PμUν))=0 | (67) |
where
Pμ=mсuμ+ecAμ | (68) |
Equation (67) has a simple solution
Pμ=−∂S∂xμ−βmcSUμ | (69) |
where S is the 4-scalar function.
Taking into account (68), one obtains
mсuμ=−∂S∂xμ−ecAμ−βmcSUμ | (70) |
This solution generalizes relation (40) and passes into it for β=0.
Using (32), we obtain the relativistic Hamilton-Jacobi equation for a particle moving in an external electromagnetic field in the presence of a drag force (64)
gμν(∂S∂xμ+βmcSUμ+ecAμ)(∂S∂xν+βmcSUν+ecAν)=m2c2 | (71) |
This equation generalizes the Hamilton-Jacobi equation (41) to the case of drag force (64) and together with the continuity equation (36) describes the classical Hamilton-Jacobi ensemble.
Using the postulates formulated in Section 2, it is easy to obtain a relativistic wave equation for the scalar wave function (44) for the corresponding spinless quantum particle; to do this, it suffices to make the replacements ∂S∂xμ→−iℏ∂μ and S=−iℏln(ψ/ψ|ψ||ψ|). As a result, we obtain the equation
gμν(iℏ∂μ−ecAμ+iℏβmcUμln(ψ/ψ|ψ||ψ|))(iℏ∂ν−ecAν+iℏβmcUνln(ψ/ψ|ψ||ψ|))ψ=m2c2 | (72) |
which generalizes the Klein-Gordon equation (43) to the case of drag force (64) and is a relativistic version of the Kostin equation (10).
In the case of the spinor representation of the Hamilton-Jacobi equation (45), we substitute (70) into (45). As a result, one obtains
[γμ(∂S∂xμ+ecAμ+βmcSUμ)+mc]Ψ=0 | (73) |
The spinor equation (73) together with the continuity equation (36), (70) describes the classical Hamilton-Jacobi ensemble in the presence of drag force (64), and, as shown above, at Ψ≠0 is equivalent to equation (71).
To obtain the wave equation for the corresponding quantum particle, we introduce a unit spinor Φ satisfying the condition
¯ΦΦ=1 | (74) |
where ¯Φ=Φ+γ0, Φ+ is the matrix Hermitian conjugate to the matrix Φ.
By definition, ¯ΦΨ is a 4-scalar [26].
Obviously, the function −iℏln(¯ΦΨ/¯ΦΨ|Ψ||Ψ|), where |Ψ|2=¯ΨΨ, in the limit ℏ→0 is equivalent to the action S.
Then the corresponding wave equation for a quantum particle with spin ½ with respect to the wave function (58), (59) according to the postulates of Section 2 will be obtained by replacing ∂S∂xμ→−iℏ∂μ and S→−iℏln(¯ΦΨ/¯ΦΨ|Ψ||Ψ|) in Eq (73). As a result, one obtains a nonlinear equation
[γμ(iℏ∂μ−ecAμ+iℏβmcUμln(__ΦΨ/__ΦΨ|Ψ||Ψ|))−mc]Ψ=0 | (75) |
Equation (75) generalizes the Dirac equation to the case of a non-Hamiltonian quantum system under the action of the drag force (64) and can be called the bispinor Kostin equation or the Dirac-Kostin equation.
In this paper, we present the quantization methods that allow obtaining wave equations for both Hamiltonian and non-Hamiltonian systems. Using these methods, we have obtained nonlinear wave equations for both non-relativistic and relativistic non-Hamiltonian systems. We hope that these approaches can be useful for studying non-Hamiltonian (in particular, dissipative) quantum mechanical systems, such as mesoscopic systems.
In the next articles of this series, we plan to apply the considered quantization methods to complex non-Hamiltonian systems.
The author declares he has not used Artificial Intelligence (AI) tools in the creation of this article.
The present work was supported by the Ministry of Science and Higher Education within the framework of the Russian State Assignment under contract No. 123021700057-0.
The author declares there is no conflict of interest.
[1] | L. D. Landau, E. M. Lifshitz, Quantum Mechanics: Non-Relativistic Theory, Vol. 3 (3rd ed.), Pergamon Press, 1977. |
[2] | A. Messiah, Quantum Mechanics, Dover Publications Inc., New York, 1999. |
[3] |
E. Nelson, Derivation of the Schrödinger equation from Newtonian mechanics, Phys. Rev., 150 (1966), 1079–1085. https://doi.org/10.1103/PhysRev.150.1079 doi: 10.1103/PhysRev.150.1079
![]() |
[4] |
M. J. W. Hall, M. Reginatto, Schrödinger equation from an exact uncertainty principle, J. Phys. A, 35 (2002), 3289–3303. https://doi.org/10.1088/0305-4470/35/14/310 doi: 10.1088/0305-4470/35/14/310
![]() |
[5] |
L. Fritsche, M. Haugk, A new look at the derivation of the Schrödinger equation from Newtonian mechanics, Ann. Phys. (Leipzig), 12 (2003), 371–403. https://doi.org/10.1002/andp.200310017 doi: 10.1002/andp.200310017
![]() |
[6] |
G. Grössing, Sub-Quantum Thermodynamics as a Basis of Emergent Quantum Mechanics, Entropy, 12 (2010), 1975–2044. https://doi.org/10.3390/e12091975 doi: 10.3390/e12091975
![]() |
[7] |
S. A. Rashkovskiy, Eulerian and Newtonian dynamics of quantum particles, Progr. Theor. Exp. Phys., 2013 (2013), 063A02. https://doi.org/10.1093/ptep/ptt036 doi: 10.1093/ptep/ptt036
![]() |
[8] |
A. O. Bolivar, Quantization of non-Hamiltonian physical systems, Phys. Rev. A, 58 (1998), 4330–4335. https://doi.org/10.1103/PhysRevA.58.4330 doi: 10.1103/PhysRevA.58.4330
![]() |
[9] |
V. E. Tarasov, Quantization of non-Hamiltonian and dissipative systems. Phys. Let. A, 288 (2001), 173–182. https://doi.org/10.1016/S0375-9601(01)00548-5 doi: 10.1016/S0375-9601(01)00548-5
![]() |
[10] |
L. A. Gonçalves, L. S. F. Olavo, Foundations of Quantum Mechanics: Derivation of a dissipative Schrödinger equation from first principles, Ann. Phys., 380 (2017), 59–70. https://doi.org/10.1016/j.aop.2017.03.002 doi: 10.1016/j.aop.2017.03.002
![]() |
[11] |
L. Gonçalves, L. S. F. Olavo, Schrödinger equation for general linear velocity-dependent forces, Phys. Rev. A, 97 (2018), 022102. https://doi.org/10.1103/PhysRevA.97.022102 doi: 10.1103/PhysRevA.97.022102
![]() |
[12] |
S. A. Rashkovskiy, Hamilton-Jacobi theory for Hamiltonian and non-Hamiltonian systems, J Geom. Mech., 12 (2020), 563–583. https://doi.org/10.3934/jgm.2020024 doi: 10.3934/jgm.2020024
![]() |
[13] |
M. O. Scully, The time-dependent Schrödinger equation revisited. I: Quantum field and classical Hamilton-Jacobi routes to Schrödinger wave equation, J. Phys. Conf. Ser., 99 (2008), 012019. https://doi.org/10.1088/1742-6596/99/1/012019 doi: 10.1088/1742-6596/99/1/012019
![]() |
[14] |
L. Nottale, Generalized quantum potentials, J Phys. A: Math. Theor., 42 (2009), 275306. https://doi.org/10.1088/1751-8113/42/27/275306 doi: 10.1088/1751-8113/42/27/275306
![]() |
[15] |
W. P. Schleich, D. M. Greenberger, D. H. Kobe, M. O. Scully, Schrödinger equation revisited, PNAS 110 (2013), 5374–5379. https://doi.org/10.1073/pnas.1302475110 doi: 10.1073/pnas.1302475110
![]() |
[16] | L. D. Landau, E. M. Lifshitz, Mechanics, Vol. 1 (3rd ed.), Butterworth-Heinemann, 1976. |
[17] |
M. D. Kostin, On the Schrödinger-Langevin equation, J. Chem. Phys., 57 (1972), 3589–3591. https://doi.org/10.1063/1.1678812 doi: 10.1063/1.1678812
![]() |
[18] |
M. D. Kostin, Friction and dissipative phenomena in quantum mechanics, J. Stat. Phys., 12 (1975), 145–151. https://doi.org/10.1007/BF01010029 doi: 10.1007/BF01010029
![]() |
[19] |
H. J. Wagner, Schrödinger quantization and variational principles in dissipative quantum theory, Z Physik B - Condensed Matter, 95 (1994), 261–273. https://doi.org/10.1007/BF01312199 doi: 10.1007/BF01312199
![]() |
[20] |
R. J. Wysocki, Hydrodynamic quantization of mechanical systems, Phys. Rev. A, 72 (2005), 032113. https://doi.org/10.1103/PhysRevA.72.032113 doi: 10.1103/PhysRevA.72.032113
![]() |
[21] |
S. G. Rajeev, A canonical formulation of dissipative mechanics using complex-valued hamiltonians, Ann. Phys., 322 (2007), 1541–1555. https://doi.org/10.1016/j.aop.2007.02.004 doi: 10.1016/j.aop.2007.02.004
![]() |
[22] |
H. Majima, A. Suzuki, Quantization and instability of the damped harmonic oscillator subject to a time-dependent force, Ann. Phys., 326 (2011), 3000–3012. https://doi.org/10.1016/j.aop.2011.08.002 doi: 10.1016/j.aop.2011.08.002
![]() |
[23] |
H. E. Moses, A spinor representation of Maxwell's equations, Nuovo. Cim., 7 (1958), 1–18. https://doi.org/10.1007/BF02725084 doi: 10.1007/BF02725084
![]() |
[24] |
A. A. Campolattaro, New spinor representation of Maxwell's equations. I. Generalities, Int. J Theor. Phys., 19 (1980), 99–126. https://doi.org/10.1007/BF00669764 doi: 10.1007/BF00669764
![]() |
[25] | L. D. Landau, E. M. Lifshitz, The Classical Theory of Fields. Vol. 2 (4th ed.), Butterworth-Heinemann, 1975. |
[26] | V. B. Berestetskii, E. M. Lifshitz, L. P. Pitaevskii, Quantum Electrodynamics. Vol. 4 (2nd ed.), Butterworth-Heinemann, 1982. |
1. | Tianfang Wang, Wen Zhang, Existence and concentration of homoclinic orbits for first order Hamiltonian systems, 2024, 16, 2836-3310, 121, 10.3934/cam.2024006 | |
2. | S. A. Rashkovskiy, Theory of Brownian motion in the Hamilton–Jacobi form, 2025, 37, 1070-6631, 10.1063/5.0261126 |