Geometric formulation of quantum mechanics

# Geometric formulation of quantum mechanics

## Abstract

Quantum mechanics is among the most important and successful mathematical model for describing our physical reality. The traditional formulation of quantum mechanics is linear and algebraic. In contrast classical mechanics is a geometrical and non-linear theory that is defined on a symplectic manifold. However, after invention of general relativity, we are convinced that geometry is physical and effect us in all scale. Hence the geometric formulation of quantum mechanics sought to give a unified picture of physical systems based on its underling geometrical structures, e.g., now, the states are represented by points of a symplectic manifold with a compatible Riemannian metric, the observables are real-valued functions on the manifold, and the quantum evolution is governed by a symplectic flow that is generated by a Hamiltonian function. In this work we will give a compact introduction to main ideas of geometric formulation of quantum mechanics. We will provide the reader with the details of geometrical structures of both pure and mixed quantum states. We will also discuss and review some important applications of geometric quantum mechanics.

## 1 Introduction

In the geometrical description of classical mechanics the states are represented by the points of a symplectic manifold which is called the phase space [Abraham_1978]. The space of observables consists of the real-valued and smooth functions on the phase space. The measurement of an observable in a state is given by . The space of observables is equipped with the structure of a commutative and associative algebra. The symplectic structure of the phase space also provides it with the Poisson bracket. An observable is associated with a vector field . Hence, flow on the phase space is generated by each observable. Moreover, the dynamics is given by a particular observer called the Hamiltonian and the flow generated by the Hamiltonian vector field describes the time evolution of the system on the phase space.

In quantum mechanics the systems correspond to rays in the Hilbert space , and the observables are represented by hermitian/self-adjoint linear operators on . Moreover, the space of observables is a real vector space equipped with two algebraic structures, namely the Jordan product and the commutator bracket. Thus the space of observables is equipped with the structure of a Lie algebra. However, the measurement theory is different compare with the classical mechanics. In the standard interpretation of quantum mechanics, the measurement of an observable in a state gives an eigenvalue of . The observable also gives rise to a flow on the state space as in the classical theory. But the flow is generated by the 1-parameter group that preserves the linearity of the Hilbert space. The dynamics is governed by a specific observable, called the Hamiltonian operator .

One can directly see that these theories both have several points in common and also in difference. The classical mechanical framework is geometric and non-linear. But the quantum mechanical framework is algebraic and linear. Moreover, the standard postulates of quantum mechanics cannot be stated without reference to this linearity. However, some researcher think that this difference seems quite surprising [Gunter_1977, Kibble_1979, Ashtekar_etal1998, Brody_etal1999] and deeper investigation shows that quantum mechanics is not a linear theory either. Since, the space of physical systems is not the Hilbert space but it is the projective Hilbert space which is a nonlinear manifold. Moreover, the Hermitian inner-product of the Hilbert space naturally equips the projective space with the structure of a Kähler manifold which is also a symplectic manifold like the classical mechanical phase space . The projective space is usually called quantum phase space of the pure quantum states.

Let be a Hamiltonian operator on . Then we can take its expectation value to obtain a real function on the Hilbert space which admits a projection to . The flow is exactly the flow defined by the Schrödinger equation on . This means that Schrödinger evolution of quantum theory is the Hamiltonian flow on . These similarities show us that the classical mechanics and quantum mechanics have many points in common. However, the quantum phase space has additional structures such as a Riemannian metric which are missing in the classical mechanics (actually Riemannian metric exists but it is not important in the classical mechanics). The Riemannian metric is part of underlying Kähler structure of quantum phase space. Some important features such as uncertainty relation and state vector reduction in quantum measurement processes are provided by the Riemannian metric.
In this work we will also illustrate the interplay between theory and the applications of geometric formulation of quantum mechanics. Recently, many researcher [Adler2000, Anandan1990b, Anandan1991, Gibbson1992, Marsden1999, Schilling1996, Marmo1, Marmo2, Montgomery1991, Hosh2, Levay] have contributed to development of geometric formulation of quantum mechanics and how this formulation provide us with insightful information about our quantum world with many applications in foundations of quantum mechanics and quantum information theory such as quantum probability, quantum uncertainty relation, geometric phases, and quantum speed limit.
In the early works, the most effort in geometric quantum mechanics were concentrated around understanding geometrical structures of pure quantum states and less attention were given to the mixed quantum states. Uhlmann was among the first researcher to consider a geometric formulation of mixed quantum states with the emphasizes on geometric phases [Uhlmann1986, Uhlmann(1989), Uhlmann(1991)]. Recent attempt to uncover hidden geometrical structures of mixed quantum states were achieved in the following works [GP, MB, DD, GQE, GUR, QSL]. Some researcher also argue that geometric formulation of quantum mechanics could lead to a generalization of quantum mechanics [Ashtekar_etal1998]. However, we will not discuss such a generalization in this work. Instead we concentrate our efforts to give an introduction to its basic structures with some applications. In particular, in section 2 we review some important mathematical tools such as Hamiltonian dynamics, principal fiber bundles, and momentum map. In section LABEL:sec2 we will discuss the basic structures of geometric quantum mechanics including quantum phase space, quantum dynamics, geometric uncertainty relation, quantum measurement, geometric postulates of quantum mechanics, and geometric phase for pure quantum states. In section LABEL:sec3 we will extend our discussion to more general quantum states, namely the mixed quantum states represented by density operators. Our review on the geometric quantum mechanics of mixed quantum states includes purification, symplectic reduction, symplectic and Riemannian structures, quantum energy dispersion, geometric uncertainty relation, geometric postulates of quantum mechanics, and geometric phase. Finally in section LABEL:sec4 we give a conclusion and an outlook. Note that we assume that reader are familiar with basic topics of differential geometry.

## 2 Mathematical structures

Mathematical structures are important in both classical and quantum physics. In algebraic description of quantum mechanics linear algebra and operator algebra are the most preferred structures for describing physical systems. However, in geometric quantum mechanics the most important mathematical structures are geometrical such as Hamiltonian dynamics, principal fiber bundles, and momentum maps. In this section we will give a short introduction to these topics.

### 2.1 Hamiltonian dynamics

In Hamiltonian mechanics the space of states or phase space is a differential manifold equipped with a symplectic form which plays an important role in describing the time evolution of the states of the system. Here we will give a short introduction to Hamiltonian dynamics. For a detail discussion of Hamiltonian dynamics we recommend the following classical book [Abraham_1978].
Let be a smooth manifold with and be the tangent space of . Moreover, let

 ω:Tp(M)×Tp(M)⟶R (1)

be a two-form on . Then is called symplectic if

1. is closed, , and

2. is non-degenerated, that is, for all whenever .

The pair are called a symplectic manifold. If is a smooth function on , then is a 1-form on . Moreover, let be a vector field. Then we define a contraction map by . The vector field is called symplectic if is closed. Furthermore, a vector field is called a Hamiltonian vector field with a Hamiltonian function if it satisfies

 ıXHω=ω(XH,⋅)=dH. (2)

A Hamiltonian system consists of the following triple . Let be a Hamiltonian vector field. Then generates the one-parameter group of diffeomorphism

 {ϕHt}:R×M⟶M, (3)

where satisfies

• with , and

• for all and .

For a Hamiltonian system each point corresponds to a state of system and the symplectic manifold is called the state space or the phase space of the system. In such a classical system, the observables are real-valued functions on the phase space. Let be a function. Then is constant along the orbits of the flow of the Hamiltonian vector field if and only if the Poisson bracket defined by

 {K,H}ζ=ωζ(XK,XH)=dK(XH)(ζ) (4)

vanishes for all . Assume is a Hamiltonian vector on , and let be a point of . Moreover, let be one-parameter group generated by in a neighborhood of the point . If we assume that the initial state is , then the evolution of the state can be described by the map defined by with initial state . Under these assumptions the trajectory of is determined by the Hamilton’s equations

 ζ′=XH(ζ). (5)

Note that

 LXHω=ıXHdω+d(ıXHω)=0+ddH=0, (6)

where is the Lie derivative, implies that the flow preserves the symplectic structure, that is . If the phase space is compact, then is an integral curve of the Hamiltonian vector field at the point .

###### Theorem 2.1.

Consider a Hamiltonian system . If is an integral curve of , then energy function is constant for all . Moreover, the flow of satisfies .

###### Proof.

By the Hamilton’s equation (5) we have

 dH(ζ(t))dt = dHζ(t)(ζ′(t))=dHζ(t)((XH)ζ(t))) = ω((XH)ζ(t),(XH)ζ(t))=0.

Thus we have shown that is constant for all . ∎

Let be a manifold. Then an almost complex structure on is an automorphism of its tangent bundle that satisfies . Moreover, the almost complex structure is a complex structure if it is integrable, meaning that a rank two tensor, usually called the Nijenhuis tensor vanishes [DaSilva].
Let be symplectic manifold. Then a Kähler manifold is symplectic manifold equipped with an integrable compatible complex structure. Moreover, being a Kähler manifold implies that is a complex manifold. Thus a Kähler form is a closed, real-valued, non-degenerated -form compatible with the complex structure.

###### Example.

Let be canonical coordinate for the symplectic form, that is . Then in these coordinates we have

 XH=(∂H∂pi,−∂H∂qi)=J⋅dH,    J=(0In−In0), (8)

where is a identity matrix.

In the following sections we will show that the quantum dynamics governed by Schrödinger and von Neumann equations can be described by Hamiltonian dynamics outlined in this section.

### 2.2 Principal fiber bundle

One important mathematical tool used in the geometric formulation of quantum physics is principal fiber bundles. In this section we will introduce the reader to the basic definition and properties of principal fiber bundles and in the following sections we will apply the tool to the quantum theory.
Let and be differentiable manifolds and be a Lie group. Then a differentiable principal fiber bundle

 (9)
You are adding the first comment!
How to quickly get a good reply:
• Give credit where it’s due by listing out the positive aspects of a paper before getting into which changes should be made.
• Be specific in your critique, and provide supporting evidence with appropriate references to substantiate general statements.
• Your comment should inspire ideas to flow and help the author improves the paper.

The better we are at sharing our knowledge with each other, the faster we move forward.
The feedback must be of minimum 40 characters and the title a minimum of 5 characters   