ADVANCED

STATISTICAL
MECHANICS

Mark Tuckerman
New York University
 New York University
Advanced Statistical Mechanics (Tuckerman)

Mark E. Tuckerman
This text is disseminated via the Open Education Resource (OER) LibreTexts Project
([Link] and like the thousands of other texts available within this powerful platform, it is
freely available for reading, printing, and "consuming."
The LibreTexts mission is to bring together students, faculty, and scholars in a collaborative effort to
provide an accessible, and comprehensive platform that empowers our community to develop, curate,
adapt, and adopt openly licensed resources and technologies; through these efforts we can reduce the
financial burden born from traditional educational resource costs, ensuring education is more accessible
for students and communities worldwide.
Most, but not all, pages in the library have licenses that may allow individuals to make changes, save,
and print this book. Carefully consult the applicable license(s) before pursuing such effects. Instructors
can adopt existing LibreTexts texts or Remix them to quickly build course-specific resources to meet the
needs of their students. Unlike traditional textbooks, LibreTexts’ web based origins allow powerful
integration of advanced features and new technologies to support learning.

LibreTexts is the adaptable, user-friendly non-profit open education resource platform that educators
trust for creating, customizing, and sharing accessible, interactive textbooks, adaptive homework, and
ancillary materials. We collaborate with individuals and organizations to champion open education
initiatives, support institutional publishing programs, drive curriculum development projects, and more.
The LibreTexts libraries are Powered by NICE CXone Expert and was supported by the Department of
Education Open Textbook Pilot Project, the California Education Learning Lab, the UC Davis Office of
the Provost, the UC Davis Library, the California State University Affordable Learning Solutions
Program, and Merlot. This material is based upon work supported by the National Science Foundation
under Grant No. 1246120, 1525057, and 1413739.
Any opinions, findings, and conclusions or recommendations expressed in this material are those of the
author(s) and do not necessarily reflect the views of the National Science Foundation nor the US
Department of Education.
Have questions or comments? For information about adoptions or adaptions contact
info@[Link] or visit our main website at [Link]
This text was compiled on 03/10/2026
TABLE OF CONTENTS
Licensing

1: Classical mechanics
1.1: The Lagrangian Formulation of Classical Mechanics
1.2: The Hamiltonian formulation of classical mechanics
1.3: The Microscopic Laws of Motion
1.4: Phase Space
1.5: Classical microscopic states or microstates and ensembles
1.6: Phase space distribution functions and Liouville's theorem

2: Foundations of classical statistical mechanics

2.1: The Ensemble Concept (Heuristic Definition)
2.2: Liouville's Theorem for non-Hamiltonian systems
2.3: The Liouville Operator and the Poisson Bracket
2.4: Equilibrium ensembles

3: The Microcanonical Ensemble

3.1: Basic Thermodynamics
3.2: The Partition Function
3.3: The Classical Virial Theorem (Microcanonical Derivation)

4: The Canonical Ensemble

4.1: Classical Virial Theorem (Canonical Ensemble Derivation)
4.2: Legendre Transforms
4.3: Relation between Canonical and Microcanonical Ensembles
4.4: Preservation of Phase Space Volume and Liouville's Theorem
4.5: Energy Fluctuations in the Canonical Ensemble

5: The Isothermal-Isobaric Ensemble

5.1: Basic Thermodynamics
5.2: Pressure and Work Virial Theorems
5.3: The partition function and relation to thermodynamics
5.4: Temperature and pressure estimators

6: The Grand Canonical Ensemble

6.1: Thermodynamics
6.2: Partition Functions
6.3: Ideal Gas
6.4: Particle Number Fluctuations

7: Distribution Functions and Liquid Structure

7.1: General Formulation of Distribution Functions
7.2: General Distribution Functions and Correlation Functions

1 [Link]
 7.3: Structure and distribution functions in classical liquids and gases
7.4: General Correlation Functions
7.5: Thermodynamic quantities in terms of g(r)
7.6: The Pair Correlation Function
7.7: Derivation of the Van der Waals equation

8: Rare-event sampling and free energy calculations

8.2: Free-energy Perturbation Theory
8.3: Adiabatic Switching and Thermodynamic Integration
8.4: Reaction Coordinates
8.5: Jarzynski's Equality and Nonequilibrium Methods
8.6: The "blue moon'' Ensemble Approach

9: Review of the basic postulates of quantum mechanics

9.1: Measurement
9.2: Physical Observables
9.3: The Fundamental Postulates of Quantum Mechanics
9.4: The Heisenberg Picture
9.5: The Heisenberg Uncertainty Principle
9.6: The Physical State of a Quantum System
9.7: Time Evolution of the State Vector

10: Fundamentals of quantum statistical mechanics

10.1: Principles of quantum statistical mechanics
10.2: The Density Matrix and Density Operator
10.3: Time evolution of the density operator
10.4: A simple example - the quantum harmonic oscillator
10.5: The quantum equilibrium ensembles
10.5.1: The microcanonical ensemble
10.5.2: The canonical ensemble
10.5.3: Isothermal-isobaric and grand canonical ensembles

11: Introduction to path integrals in quantum mechanics and quantum

statistical mechanics
11.1: Discretized and Continuous Path Integrals
11.1.1: Derivation of the Discretized Path Integral
11.1.2: Doing the Path Integral - the Free Particle
11.1.3: Dominant Paths in the Propagator and Density Matrix
11.1.4: The Continuous Limit
11.2: Calculation of observables from path integrals
11.2.1: Expectation values of observables
11.2.2: Path integrals for N-particle systems
11.2.3: Path integral molecular dynamics (optional reading)
11.2.4: Thermodynamics from path integrals
11.3: Expansion about the classical path and stationary phase
11.3.1: The harmonic Oscillator - Expansion about the Classical Path
11.3.2: The Stationary Phase Approximation

2 [Link]
12: Time-dependent Processes - Classical case
12.1: Perturbative solution of the Liouville equation
12.2: Kubo Transform Expression for the Time Correlation Function
12.3: Generalized Equations of Motion
12.4: General Properties of Time Correlation Functions
12.5: Perturbative solution of the Liouville equation
12.6: The Onsager Fluctuation Regression Theorem
12.7: Relation to Spectra
12.8: Time Correlation Functions and Transport Coefficients
12.8.1: Shear Viscosity
12.8.2: The Diffusion Constant

13: Time-dependent Processes - Quantum Case

13.1: Calculation of spectra from perturbation theory
13.1.1: The Hamiltonian
13.1.2: The Transition Rate
13.1.3: Examples
13.2: Iterative solution for the interaction-picture state vector
13.3: The Interaction Picture
13.4: Fermi's Golden Rule
13.5: Quantum Linear Response Theory

14: The Langevin and Generalized Langevin equations

14.1: The Harmonic Bath Hamiltonian
14.2: The Random Force Term
14.3: The Dynamic Friction Kernel
14.4: Relation between the Dynamic Friction Kernel and the Random Force
14.5: Derivation of the GLE
14.6: Mori-Zwanzig Theory- A more general derivation of the GLE
14.7: Example- Vibrational dephasing and energy relaxation

Index

Glossary

Detailed Licensing

Detailed Licensing

3 [Link]
Licensing
A detailed breakdown of this resource's licensing can be found in Back Matter/Detailed Licensing.

1 [Link]
 CHAPTER OVERVIEW

1: Classical mechanics
1.1: The Lagrangian Formulation of Classical Mechanics
1.2: The Hamiltonian formulation of classical mechanics
1.3: The Microscopic Laws of Motion
1.4: Phase Space
1.5: Classical microscopic states or microstates and ensembles
1.6: Phase space distribution functions and Liouville's theorem

1: Classical mechanics is shared under a not declared license and was authored, remixed, and/or curated by LibreTexts.

1
1.1: The Lagrangian Formulation of Classical Mechanics
In order to begin to make a connection between the microscopic and macroscopic worlds, we need to better understand the
microscopic world and the laws that govern it. We will begin placing Newton's laws of motion in a formal framework which will be
heavily used in our study of classical statistical mechanics.
First, we begin by restricting our discussion to systems for which the forces are purely conservative. Such forces are derivable from
a potential energy function by differentiation:

It is clear that such forces cannot contain dissipative or friction terms. An important property of systems whose forces are
conservative is that they conserve the total energy

To see this, simply differentiate the energy with respect to time:

where, the second line, the facts that (Newton's law) and (conservative force definition) have been used. This is
known as the law of conservation of energy.
For conservative systems, there is an elegant formulation of classical mechanics known as the Lagrangian formulation. The
Lagrangian function, , for a system is defined to be the difference between the kinetic and potential energies expressed as a
function of positions and velocities. In order to make the nomenclature more compact, we shall introduce a shorthand for the
complete set of positions in an -particle system: and for the velocities: . Then, the Lagrangian is defined
as follows:

In terms of the Lagrangian, the classical equations of motion are given by the so called Euler-Lagrange equation:

The equations that result from application of the Euler-Lagrange equation to a particular Lagrangian are known as the equations of
motion. The solution of the equations of motion for a given initial condition is known as a trajectory of the system. The Euler-
Lagrange equation results from what is known as an action principle. We shall defer further discussion of the action principle until
we study the Feynman path integral formulation of quantum statistical mechanics in terms of which the action principle emerges
very naturally. For now, we accept the Euler-Lagrange equation as a definition.
The Euler-Lagrange formulation is completely equivalent to Newton's second law. In order to see this, note that

Therefore,

1.1.1 [Link]
which is just Newton's equation of motion.
An important property of the Lagrangian formulation is that it can be used to obtain the equations of motion of a system in any set
of coordinates, not just the standard Cartesian coordinates, via the Euler-Lagrange equation (see problem set #1).

This page titled 1.1: The Lagrangian Formulation of Classical Mechanics is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by Mark Tuckerman.

1.1.2 [Link]
1.2: The Hamiltonian formulation of classical mechanics
The Lagrangian formulation of mechanics will be useful later when we study the Feynman path integral. For our purposes now, the
Lagrangian formulation is an important springboard from which to develop another useful formulation of classical mechanics
known as the Hamiltonian formulation. The Hamiltonian of a system is defined to be the sum of the kinetic and potential energies
expressed as a function of positions and their conjugate momenta. What are conjugate momenta?
Recall from elementary physics that momentum of a particle, , is defined in terms of its velocity by

In fact, the more general definition of conjugate momentum, valid for any set of coordinates, is given in terms of the Lagrangian:

Note that these two definitions are equivalent for Cartesian variables. In terms of Cartesian momenta, the kinetic energy is given by

Then, the Hamiltonian, which is defined to be the sum, , expressed as a function of positions and momenta, will be given by

where . In terms of the Hamiltonian, the equations of motion of a system are given by Hamilton's equations:

The solution of Hamilton's equations of motion will yield a trajectory in terms of positions and momenta as functions of time.
Again, Hamilton's equations can be easily shown to be equivalent to Newton's equations, and, like the Lagrangian formulation,
Hamilton's equations can be used to determine the equations of motion of a system in any set of coordinates.
The Hamiltonian and Lagrangian formulations possess an interesting connection. The Hamiltonian can be directly obtained from
the Lagrangian by a transformation known as a Legendre transform. We will say more about Legendre transforms in a later lecture.
For now, note that the connection is given by

which, when the fact that is used, becomes

Because a system described by conservative forces conserves the total energy, it follows that Hamilton's equations of motion
conserve the total Hamiltonian. Hamilton's equations of motion conserve the Hamiltonian

 Proof

1.2.1 [Link]
 QED. This, then, provides another expression of the law of conservation of energy.

This page titled 1.2: The Hamiltonian formulation of classical mechanics is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by Mark Tuckerman.

1.2.2 [Link]
1.3: The Microscopic Laws of Motion
Consider a system of classical particles. The particles a confined to a particular region of space by a "container'' of volume .
The particles have a finite kinetic energy and are therefore in constant motion, driven by the forces they exert on each other (and
any external forces which may be present). At a given instant in time , the Cartesian positions of the particles are )).
The time evolution of the positions of the particles is then given by Newton's second law of motion:

where are the forces on each of the particles due to all the other particles in the system. The notation .
Newton's equations of motion constitute a set of coupled second order differential equations. In order to solve these, it is
necessary to specify a set of appropriate initial conditions on the coordinates and their first time derivatives,
. Then, the solution of Newton's equations gives the complete set of coordinates and velocities for
all time .

This page titled 1.3: The Microscopic Laws of Motion is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
Mark Tuckerman.

1.3.1 [Link]
1.4: Phase Space
We construct a Cartesian space in which each of the coordinates and momenta is assigned to one of mutually orthogonal
axes. Phase space is, therefore, a dimensional space. A point in this space is specified by giving a particular set of values for the
coordinates and momenta. Denote such a point by

is a dimensional vector. Thus, the time evolution or trajectory of a system as specified by Hamilton's equations of motion, can
be expressed by giving the phase space vector, as a function of time.

Figure : Evolution of an ensemble of classical systems in phase space (top). The systems are a massive particle in a one-
dimensional potential well (red curve, lower figure). The initially compact ensemble becomes swirled up over time.
The law of conservation of energy, expressed as a condition on the phase space vector:

defines a dimensional hypersurface in phase space on which the trajectory must remain.

This page titled 1.4: Phase Space is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark Tuckerman.

1.4.1 [Link]
1.5: Classical microscopic states or microstates and ensembles
A microscopic state or microstate of a classical system is a specification of the complete set of positions and momenta of the
system at any given time. In the language of phase space vectors, it is a specification of the complete phase space vector of a
system at any instant in time. For a conservative system, any valid microstate must lie on the constant energy hypersurface,
. Hence, specifying a microstate of a classical system is equivalent to specifying a point on the constant energy hypersurface.
The concept of classical microstates now allows us to give a more formal definition of an ensemble. An ensemble is a collection of
systems sharing one or more macroscopic characteristics but each being in a unique microstate. The complete ensemble is specified
by giving all systems or microstates consistent with the common macroscopic characteristics of the ensemble.
The idea of ensemble averaging can also be expressed in terms of an average over all such microstates (which comprise the
ensemble). A given macroscopic property, , and its microscopic function , which is a function of the positions and
momenta of a system, i.e. the phase space vector, are related by

where is the microstate of the th member of the ensemble.

This page titled 1.5: Classical microscopic states or microstates and ensembles is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by Mark Tuckerman.

1.5.1 [Link]
1.6: Phase space distribution functions and Liouville's theorem
Given an ensemble with many members, each member having a different phase space vector corresponding to a different
microstate, we need a way of describing how the phase space vectors of the members in the ensemble will be distributed in the
phase space. That is, if we choose to observe one particular member in the ensemble, what is the probability that its phase space
vector will be in a small volume around a point in the phase space at time . This probability will be denoted

where is known as the phase space probability density or phase space distribution function. It's properties are as follows:

= Number of members in the ensemble

 Liouville's Theorem

The total number of systems in the ensemble is a constant. What restrictions does this place on ? For a given volume in
phase space, this condition requires that the rate of decrease of the number of systems from this region is equal to the flux of
systems into the volume.

Let be the unit normal vector to the surface of this region.

Figure : Divergence through a surface

The flux through the small surface area element, is just . Then the total flux out of volume is obtained by
integrating this over the entire surface that encloses :

which follows from the divergence theorem. is the dimensional gradient on the phase space

On the other hand, the rate of decrease in the number of systems out of the volume is

Equating these two quantities gives

1.6.1 [Link]
But this result must hold for any arbitrary choice of the volume , which we may also allow to shrink to 0 so that the result holds
locally, and we obtain the local result:

But

This equation resembles an equation for a "hydrodynamic'' flow in the phase space, with playing the role of a density. The
quantity , being the divergence of a velocity field, is known as the phase space compressibility, and it does not, for a general
dynamical system, vanish. Let us see what the phase space compressibility for a Hamiltonian system is:

However, by Hamilton's equations:

Thus, the compressibility is given by

Thus, Hamiltonian systems are incompressible in the phase space, and the equation for becomes

which is Liouville's equation, and it implies that is a conserved quantity when is identified as the phase space vector of a
particular Hamiltonian system. That is, will be conserved along a particular trajectory of a Hamiltonian system. However, if
we view is a fixed spatial label in the phase space, then the Liouville equation specifies how a phase space distribution function
evolves in time from an initial distribution .

This page titled 1.6: Phase space distribution functions and Liouville's theorem is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by Mark Tuckerman.

1.6.2 [Link]
 CHAPTER OVERVIEW

2: Foundations of classical statistical mechanics

2.1: The Ensemble Concept (Heuristic Definition)
2.2: Liouville's Theorem for non-Hamiltonian systems
2.3: The Liouville Operator and the Poisson Bracket
2.4: Equilibrium ensembles

2: Foundations of classical statistical mechanics is shared under a not declared license and was authored, remixed, and/or curated by LibreTexts.

1
2.1: The Ensemble Concept (Heuristic Definition)
For a typical macroscopic system, the total number of particles . Since an essentially infinite amount of precision is needed
in order to specify the initial conditions (due to exponentially rapid growth of errors in this specification), the amount of
information required to specify a trajectory is essentially infinite. Even if we contented ourselves with quadrupole precision,
however, the amount of memory needed to hold just one phase space point would be about 128 bytes = bytes for each
number or Gbytes. The largest computers we have today have perhaps Gbytes of memory, so we are off by
14 orders of magnitude just to specify 1 point in phase space.

 Example

Do we need all this detail?

Yes
There are plenty of chemically interesting phenomena for which we really would like to know how individual atoms are
moving as a process occurs. Experimental techniques such as ultrafast laser spectroscopy can resolve short time scale
phenomena and, thus, obtain important insights into such motions. From a theoretical point of view, although we cannot follow
particles, there is some hope that we could follow the motion of a system containing or particles, which might
capture most of the features of true macroscopic matter. Thus, by solving Newton's equations of motion numerically on a
computer, we have a kind of window into the microscopic world. This is the basis of what are known as molecular dynamics
calculations.

No
Intuitively, we would expect that if we were to follow the evolution of a large number of systems all described by the same set
of forces but having starting from different initial conditions, these systems would have essentially the same macroscopic
characteristics, e.g. the same temperature, pressure, etc. even if the microscopic detailed evolution of each system in time
would be very different. This idea suggests that the microscopic details are largely unimportant.

Since, from the point of view of macroscopic properties, precise microscopic details are largely unimportant, we might imagine
employing a construct known as the ensemble concept in which a large number of systems with different microscopic
characteristics but similar macroscopic characteristics is used to "wash out'' the microscopic details via an averaging procedure.
This is an idea developed by individuals such as Gibbs, Maxwell, and Boltzmann.

Ensemble
Consider a large number of systems each described by the same set of microscopic forces and sharing some common macroscopic
property (e.g. the same total energy). Each system is assumed to evolve under the microscopic laws of motion from a different
initial condition so that the time evolution of each system will be different from all the others. Such a collection of systems is called
an ensemble. The ensemble concept then states that macroscopic observables can be calculated by performing averages over the
systems in the ensemble. For many properties, such as temperature and pressure, which are time-independent, the fact that the
systems are evolving in time will not affect their values, and we may perform averages at a particular instant in time. Thus, let
denote a macroscopic property and let denote a microscopic function that is used to compute . An example of would be the
temperature, and would be the kinetic energy (a microscopic function of velocities). Then, is obtained by calculating the value
of in each system of the ensemble and performing an average over all systems in the ensemble:

where is the total number of members in the ensemble and is the value of in the th system.
The questions that naturally arise are:
1. How do we construct an ensemble?
2. How do we perform averages over an ensemble?

2.1.1 [Link]
 3. How many systems will an ensemble contain?
4. How do we distinguish time-independent from time-dependent properties in the ensemble picture?

Answering these questions will be our main objective in Statistical Mechanics.

This page titled 2.1: The Ensemble Concept (Heuristic Definition) is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by Mark Tuckerman.

2.1.2 [Link]
2.2: Liouville's Theorem for non-Hamiltonian systems
The equations of motion of a system can be cast in the generic form

where, for a Hamiltonian system, the vector function would be

and the incompressibility condition would be a condition on :

A non-Hamiltonian system, described by a general vector funciton , will not, in general, satisfy the incompressibility condition.
That is:

Non-Hamiltonian dynamical systems are often used to describe open systems, i.e., systems in contact with heat reservoirs or
mechanical pistons or particle reservoirs. They are also often used to describe driven systems or systems in contact with external
fields.
The fact that the compressibility does not vanish has interesting consequences for the structure of the phase space. The Jacobian,
which satisfies

will no longer be 1 for all time. Defining , the general solution for the Jacobian can be written as

Note that as before. Also, note that . Thus, can be expressed as the total time derivative of some function,
which we will denote W, i.e., . Then, the Jacobian becomes

Thus, the volume element in phase space now transforms according to

which can be arranged to read as a conservation law:

{
Thus, we have a conservation law for a modified volume element, involving a "metric factor'' . Introducing the
suggestive notation , the conservation law reads . This is a generalized version of
Liouville's theorem. Furthermore, a generalized Liouville equation for non-Hamiltonian systems can be derived which incorporates
this metric factor. The derivation is beyond the scope of this course, however, the result is

We have called this equation, the generalized Liouville equation Finally, noting that satisfies the same equation as J, i.e.,

2.2.1 [Link]
the presence of in the generalized Liouville equation can be eliminated, resulting in

which is the ordinary Liouville equation from before. Thus, we have derived a modified version of Liouville's theorem and have
shown that it leads to a conservation law for f equivalent to the Hamiltonian case. This, then, supports the generality of the
Liouville equation for both Hamiltonian and non-Hamiltonian based ensembles, an important fact considering that this equation is
the foundation of statistical mechanics.

This page titled 2.2: Liouville's Theorem for non-Hamiltonian systems is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by Mark Tuckerman.

2.2.2 [Link]
2.3: The Liouville Operator and the Poisson Bracket
From the last lecture, we saw that Liouville's equation could be cast in the form

The Liouville equation is the foundation on which statistical mechanics rests. It will now be cast in a form that will be suggestive
of a more general structure that has a definite quantum analog (to be revisited when we treat the quantum Liouville equation).
Define an operator

known as the Liouville operator ( - the i is there as a matter of convention and has the effect of making a Hermitian
operator). Then Liouville's equation can be written

The Liouville operator also be expressed as

where is known as the Poisson bracket between and :

Thus, the Liouville equation can be written as

The Liouville equation is a partial differential equation for the phase space probability distribution function. Thus, it specifies a
general class of functions that satisfy it. In order to obtain a specific solution requires more input information, such as an
initial condition on f, a boundary condition on f, and other control variables that characterize the ensemble.

This page titled 2.3: The Liouville Operator and the Poisson Bracket is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by Mark E. Tuckerman.

2.3.1 [Link]
2.4: Equilibrium ensembles
An equilibrium ensemble is one for which there is no explicit time-dependence in the phase space distribution function, . In
this case, Liouville's equation reduces to

which implies that must be a pure function of the Hamiltonian

The specific form that has depends on the specific details of the ensemble.
The integral over the phase space distribution function plays a special role in statistical mechanics:

It is known as the partition function and is equal to the number of members in the ensemble. That is, it is equal to the number of
microstates that all give rise to a given set of macroscopic observables. Thus, it is the quantity from which all thermodynamic
properties are derived.
If a measurement of a macroscopic observable is made, then the value obtained will be the ensemble average:

Equations and are the central results of ensemble theory, since they determine all thermodynamic and other observable
quantities.

This page titled 2.4: Equilibrium ensembles is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark
Tuckerman.

2.4.1 [Link]
 CHAPTER OVERVIEW

3: The Microcanonical Ensemble

The microcanonical ensemble is built upon the so called postulate of equal a priori probabilities:
Postulate of equal a priori probabilities: For an isolated macroscopic system in equilibrium, all microscopic states corresponding
to the same set of macroscopic observables are equally probable.
3.1: Basic Thermodynamics
3.2: The Partition Function
3.3: The Classical Virial Theorem (Microcanonical Derivation)

This page titled 3: The Microcanonical Ensemble is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark
Tuckerman.

1
3.1: Basic Thermodynamics
In the microcanonical ensemble, the entropy is a natural function of , and , i.e., . This can be inverted to give
the energy as a function of , , and , i.e., . Consider using Legendre transformation to change from to using
the fact that

The Legendre transform of is

The quantity is called the Hemlholtz free energy and is given the symbol and is the fundamental energy in
the canonical ensemble. The differential of is

However, from , we have

From the first law, is given by

Thus,

Comparing the two expressions, we see that the thermodynamic relations are

This page titled 3.1: Basic Thermodynamics is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark
Tuckerman.

3.1.1 [Link]
3.2: The Partition Function
Consider two systems (1 and 2) in thermal contact such that

and the total Hamiltonian is just

Since system 2 is infinitely large compared to system 1, it acts as an infinite heat reservoir that keeps system 1 at a constant
temperature without gaining or losing an appreciable amount of heat, itself. Thus, system 1 is maintained at canonical conditions,
.
The full partition function for the combined system is the microcanonical partition function

Now, we define the distribution function, of the phase space variables of system 1 as

Taking the natural log of both sides, we have

Since , it follows that , and we may expand the above expression about . To linear order, the
expression becomes

where, in the last line, the differentiation with respect to is replaced by differentiation with respect to . Note that

where is the common temperature of the two systems. Using these two facts, we obtain

Thus, the distribution function of the canonical ensemble is

The prefactor is an irrelevant constant that can be disregarded as it will not affect any physical properties.
The normalization of the distribution function is the integral:

3.2.1 [Link]
where is the canonical partition function. It is convenient to define an inverse temperature . is the
canonical partition function. As in the microcanonical case, we add in the ad hoc quantum corrections to the classical result to give

The thermodynamic relations are thus,

This page titled 3.2: The Partition Function is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark
Tuckerman.

3.2.2 [Link]
3.3: The Classical Virial Theorem (Microcanonical Derivation)
Consider a system with Hamiltonian . Let and be specific components of the phase space vector.

 Theorem : Classical Virial Theorem

The classical virial theorem states that

where the average is taken with respect to a microcanonical ensemble.

To prove the theorem, start with the definition of the average:

where the fact that has been used. Also, the and dependence of the partition function have been suppressed. Note
that the above average can be written as

However, writing

allows the average to be expressed as

The first integral in the brackets is obtained by integrating the total derivative with respect to over the phase space variable .
This leaves an integral that must be performed over all other variables at the boundary of phase space where , as indicated by
the surface element . But the integrand involves the factor , so this integral will vanish. This leaves:

where is the partition function of the uniform ensemble. Recalling that we have

3.3.1 [Link]
which proves the theorem.

 Example

: and The virial theorem says that

Thus, at equilibrium, the kinetic energy of each particle must be . By summing both sides over all the particles, we obtain a
well know result

This page titled 3.3: The Classical Virial Theorem (Microcanonical Derivation) is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by Mark E. Tuckerman.

3.3.2 [Link]
 CHAPTER OVERVIEW

4: The Canonical Ensemble

4.1: Classical Virial Theorem (Canonical Ensemble Derivation)
4.2: Legendre Transforms
4.3: Relation between Canonical and Microcanonical Ensembles
4.4: Preservation of Phase Space Volume and Liouville's Theorem
4.5: Energy Fluctuations in the Canonical Ensemble

4: The Canonical Ensemble is shared under a not declared license and was authored, remixed, and/or curated by LibreTexts.

1
4.1: Classical Virial Theorem (Canonical Ensemble Derivation)
Again, let and be specific components of the phase space vector . Consider the canonical average

given by

But

Thus,

Several cases exist for the surface term :

1. a momentum variable. Then, since evaluated at clearly vanishes.
2. and as , thus representing a bound system. Then, also vanishes at .
3. and as , representing an unbound system. Then the exponential tends to 1 both at , hence the
surface term vanishes.
4. and the system is periodic, as in a solid. Then, the system will be represented by some supercell to which periodic
boundary conditions can be applied, and the coordinates will take on the same value at the boundaries. Thus, and
will take on the same value at the boundaries and the surface term will vanish.
5. and the particles experience elastic collisions with the walls of the container. Then there is an infinite potential at the
walls so that at the boundary and at the boundary.

Thus, we have the result

The above cases cover many but not all situations, in particular, the case of a system confined within a volume with reflecting
boundaries. Then, surface contributions actually give rise to an observable pressure (to be discussed in more detail in the next
lecture).

This page titled 4.1: Classical Virial Theorem (Canonical Ensemble Derivation) is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by Mark Tuckerman.

4.1.1 [Link]
4.2: Legendre Transforms
The microcanonical ensemble involved the thermodynamic variables , and as its variables. However, it is often convenient
and desirable to work with other thermodynamic variables as the control variables. Legendre transforms provide a means by which
one can determine how the energy functions for different sets of thermodynamic variables are related. The general theory is given
below for functions of a single variable.
Consider a function and its derivative

The equation defines a variable transformation from to . Is there a unique description of the function in terms of
the variable ? That is, does there exist a function that is equivalent to ?
Given a point , can one determine the value of the function given only ? No, for the reason that the function
for any constant will have the same value of as shown in Figure .

Figure : The Legendre transfer in action (Mark Tuckerman)

However, the value can be determined uniquely if we specify the slope of the line tangent to at , i.e., and the -
intercept, of this line. Then, using the equation for the line, we have

This relation must hold for any general :

Note that is the variable , and , where is the functional inverse of , i.e., . Solving for
gives

where is known as the Legendre transform of . In shorthand notation, one writes

however, it must be kept in mind that is a function of .

This page titled 4.2: Legendre Transforms is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark E.
Tuckerman.

4.2.1 [Link]
4.3: Relation between Canonical and Microcanonical Ensembles
We saw that the and could be related by a Legendre transformation. The partition functions and
can be related by a Laplace transform. Recall that the Laplace transform of a function is given by

Let us compute the Laplace transform of with respect to :

Using the -function to do the integral over :

By identifying , we see that the Laplace transform of the microcanonical partition function gives the canonical partition
function .

This page titled 4.3: Relation between Canonical and Microcanonical Ensembles is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by Mark Tuckerman.

4.3.1 [Link]
4.4: Preservation of Phase Space Volume and Liouville's Theorem
Consider a phase space volume element at t=0, containing a small collection of initial conditions on a set of trajectories. The
trajectories evolve in time according to Hamilton's equations of motion, and at a time t later will be located in a new volume
element as shown in the figure below:

Figure
How is related to dxdd ? To answer this, consider a trajectory starting from a phase space vector in and having a
phase space vector at time in . Since the solution of Hamilton's equations depends on the choice of initial conditions,
depends on :

Thus, the phase space vector components can be viewed as a coordinate transformation on the phase space from to time . The
phase space volume element then transforms according to

where is the Jacobian of the transformation:

where . The precise form of the Jacobian can be determined as will be demonstrated below.
The Jacobian is the determinant of a matrix ,

whose matrix elements are

Taking the time derivative of the Jacobian, we therefore have

The matrices and can be seen to be given by

4.4.1 [Link]
Substituting into the expression for gives

where the chain rule has been introduced for the derivative . The sum over i can now be performed:

Thus,

or

The initial condition on this differential equation is . Moreover, for a Hamiltonian system . This says
that and . Thus, . If this is true, then the phase space volume element transforms according to

which is another conservation law. This conservation law states that the phase space volume occupied by a collection of systems
evolving according to Hamilton's equations of motion will be preserved in time. This is one statement of Liouville's theorem.
Combining this with the fact that , we have a conservation law for the phase space probability:

which is an equivalent statement of Liouville's theorem.

This page titled 4.4: Preservation of Phase Space Volume and Liouville's Theorem is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by Mark Tuckerman.

4.4.2 [Link]
4.5: Energy Fluctuations in the Canonical Ensemble
In the canonical ensemble, the total energy is not conserved. ( ). What are the fluctuations in the energy? The energy
fluctuations are given by the root mean square deviation of the Hamiltonian from its average :

}}

Therefore

But

Thus,

Therefore, the relative energy fluctuation is given by

Now consider what happens when the system is taken to be very large. In fact, we will define a formal limit called the
thermodynamic limit, in which and such that remains constant.
Since and are both extensive variables, and ,

But would be exactly 0 in the microcanonical ensemble. Thus, in the thermodynamic limit, the canonical and microcanonical
ensembles are equivalent, since the energy fluctuations become vanishingly small.

This page titled 4.5: Energy Fluctuations in the Canonical Ensemble is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by Mark Tuckerman.

4.5.1 [Link]
 CHAPTER OVERVIEW

5: The Isothermal-Isobaric Ensemble

5.1: Basic Thermodynamics
5.2: Pressure and Work Virial Theorems
5.3: The partition function and relation to thermodynamics
5.4: Temperature and pressure estimators

5: The Isothermal-Isobaric Ensemble is shared under a not declared license and was authored, remixed, and/or curated by LibreTexts.

1
5.1: Basic Thermodynamics
The Helmholtz free energy is a natural function of and . The isothermal-isobaric ensemble is generated by
transforming the volume in favor of the pressure so that the natural variables are , , and (which are conditions under
which many experiments are performed, e.g., `standard temperature and pressure'. Performing a Legendre transformation of the
Helmholtz free energy

But

Thus,

where is the Gibbs free energy. The differential of is

But from , we have

but , thus

Equating the two expressions for , we see that

This page titled 5.1: Basic Thermodynamics is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark
Tuckerman.

5.1.1 [Link]
5.2: Pressure and Work Virial Theorems
As noted earlier, the quantity is a measure of the instantaneous value of the internal pressure . Let us look at the
ensemble average of this quantity

Doing the volume integration by parts gives

Thus,

This result is known as the pressure virial theorem. It illustrates that the average of the quantity gives the fixed pressure
that defines the ensemble. Another important result comes from considering the ensemble average

Once again, integrating by parts with respect to the volume yields

or

This result is known as the work virial theorem. It expresses the fact that equipartitioning of energy also applies to the volume
degrees of freedom, since the volume is now a fluctuating quantity.

This page titled 5.2: Pressure and Work Virial Theorems is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated
by Mark Tuckerman.

5.2.1 [Link]
5.3: The partition function and relation to thermodynamics
In principle, we should derive the isothermal-isobaric partition function by coupling our system to an infinite thermal reservoir as
was done for the canonical ensemble and also subject the system to the action of a movable piston under the influence of an
external pressure . In this case, both the temperature of the system and its pressure will be controlled, and the energy and volume
will fluctuate accordingly.
However, we saw that the transformation from to between the microcanonical and canonical ensembles turned into a Laplace
transform relation between the partition functions. The same result holds for the transformation from to . The relevant "energy''
quantity to transform is the work done by the system against the external pressure in changing its volume from to , which
will be . Thus, the isothermal-isobaric partition function can be expressed in terms of the canonical partition function by the
Laplace transform:

where is a constant that has units of volume. Thus,

The Gibbs free energy is related to the partition function by

This can be shown in a manner similar to that used to prove the . The differential equation to start with is

Other thermodynamic relations follow:

Volume:

Enthalpy:

Heat capacity at constant pressure

Entropy:

The fluctuations in the enthalpy are given, in analogy with the canonical ensemble, by

so that

5.3.1 [Link]
so that, since and are both extensive, which vanish in the thermodynamic limit.

This page titled 5.3: The partition function and relation to thermodynamics is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by Mark Tuckerman.

5.3.2 [Link]
5.4: Temperature and pressure estimators
From the classical virial theorem

we arrived at the equipartition theorem:

where are the Cartesian momenta of the particles in a system. This says that the microscopic function of the
momenta that corresponds to the temperature, a macroscopic observable of the system, is given by

The ensemble average of can be related directly to the temperature

\) is known as an estimator (a term taken over from the Monte Carlo literature) for the temperature. An estimator is
some function of the phase space coordinates, i.e., a function of microscopic states, whose ensemble average gives rise to a
physical observable.
An estimator for the pressure can be derived as well, starting from the basic thermodynamic relation:

with

The volume dependence of the partition function is contained in the limits of integration, since the range of integration for the
coordinates is determined by the size of the physical container. For example, if the system is confined within a cubic box of volume
, with the length of a side, then the range of each integration will be from 0 to . If a change of variables is made to
, then the range of each integration will be from 0 to 1. The coordinates are known as scaled coordinates. For containers
of a more general shape, a more general transformation is

To preserve the phase space volume element, however, we need to ensure that the transformation is a canonical one. Thus, the
corresponding momentum transformation is

With this coordinate/momentum transformation, the phase space volume element transforms as

Thus, the volume element remains the same as required. With this transformation, the Hamiltonian becomes

and the canonical partition function becomes

5.4.1 [Link]
Thus, the pressure can now be calculated by explicit differentiation with respect to the volume, :

Thus, the pressure estimator is

and the pressure is given by

For periodic systems, such as solids and currently used models of liquids, an absolute Cartesian coordinate is ill-defined. Thus,
the virial part of the pressure estimator must be rewritten in a form appropriate for periodic systems. This can be done by
recognizing that the force is obtained as a sum of contributions , which is the force on particle due to particle . Then, the
classical virial becomes

where is now a relative coordinate. must be computed consistent with periodic boundary conditions, i.e., the relative
coordinate is defined with respect to the closest periodic image of particle with respect to particle . This gives rise to surface
contributions, which lead to a nonzero pressure, as expected.

This page titled 5.4: Temperature and pressure estimators is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated
by Mark Tuckerman.

5.4.2 [Link]
 CHAPTER OVERVIEW

6: The Grand Canonical Ensemble

In the grand canonical ensemble, the control variables are the chemical potential , the volume and the temperature . The total
particle number is therefore allowed to fluctuate. It is therefore related to the canonical ensemble by a Legendre transformation
with respect to the particle number . Its utility lies in the fact that it closely represents the conditions under which experiments are
often performed and, as we shall see, it gives direct access to the equation of state.
6.1: Thermodynamics
6.2: Partition Functions
6.3: Ideal Gas
6.4: Particle Number Fluctuations

Template:HideTOC

This page titled 6: The Grand Canonical Ensemble is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
Mark Tuckerman.

1
6.1: Thermodynamics
In the canonical ensemble, the Helmholtz free energy is a natural function of and . As usual, we perform a
Legendre transformation to eliminate in favor of :

It turns out that the free energy is the quantity . We shall derive this result below in the context of the partition
function. Thus,

To motivate the fact that is the proper free energy of the grand canonical ensemble from thermodynamic considerations, we
need to introduce a mathematical theorem, known as Euler's theorem:

 Euler's Theorem

Let be a function such that

Then is said to be a homogeneous function of degree . For example, the function is a homogeneous function of
degree 2, is a homogeneous function of degree 3, however, is not a homogeneous
function. Euler's Theorem states that, for a homogeneous function ,

 Proof

To prove Euler's theorem, simply differentiate the the homogeneity condition with respect to lambda:

Then, setting , we have

which is exactly Euler's theorem.

Now, in thermodynamics, extensive thermodynamic functions are homogeneous functions of degree 1. Thus, to see how Euler's
theorem applies in thermodynamics, consider the familiar example of the Gibbs free energy:

The extensive dependence of is on , so, being a homogeneous function of degree 1, it should satisfy

Applying Euler's theorem, we thus have

6.1.1 [Link]
or, for a multicomponent system,

But, since

it can be seen that is consistent with the first law of thermodynamics.

Now, for the Legendre transformed free energy in the grand canonical ensemble, the thermodynamics are

But, since

the thermodynamics will be given by

Since, is a homogeneous function of degree 1, and its extensive argument is , it should satisfy

Thus, applying Euler's theorem,

and since

the assignment is consistent with the first law of thermodynamics. It is customary to work with , rather than , so
is the natural free energy in the grand canonical ensemble, and, unlike the other ensembles, it is not given a special name or
symbol!

This page titled 6.1: Thermodynamics is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark
Tuckerman.

6.1.2 [Link]
6.2: Partition Functions
Consider two canonical systems, 1 and 2, with particle numbers and , volumes and and at temperature . The systems
are in chemical contact, meaning that they can exchange particles. Furthermore, we assume that and so that
system 2 is a particle reservoir. The total particle number and volume are

The total Hamiltonian is

If the systems could not exchange particles, then the canonical partition function for the whole system would be

where

However, and are not fixed, therefore, in order to sum over all microstates, we need to sum over all values that can take
on subject to the constraint . Thus, we can write the canonical partition function for the whole system as

where is a function that weights each value of for a given .

Thus,
is the number of configurations with 0 particles in and particles in .
is the number of configurations with 1 particles in and particles in .
etc.

Determining the values of amounts to a problem of counting the number of ways we can put identical objects into 2
baskets. Thus,

etc. or generally,

6.2.1 [Link]
which is clearly a classical degeneracy factor. If we were doing a purely classical treatment of the grand canonical ensemble, then
this factor would appear in the sum for , however, we always include the ad hoc quantum correction in the expression
for the canonical partition function, and we see that these quantum factors will exactly cancel the classical degeneracy factor,
leading to the following expression:

which expresses the fact that, in reality, the various configurations are not distinguishable from each other, and so each one should
count with equal weighting. Now, the distribution function is given by

which is chosen so that

However, recognizing that , we can obtain the distribution for immediately, by integrating over the phase space of
system 2:

where the prefactor has been introduced so that

and amounts to the usual ad hoc quantum correction factor that must be multiplied by the distribution function for each ensemble to
account for the identical nature of the particles. Thus, we see that the distribution function becomes

Recall that the Hemlholtz free energy is given by

Thus,

or

But since and , we may expand:

Therefore the distribution function becomes

Dropping the "1'' subscript, we have

6.2.2 [Link]
We require that be normalized:

Now, we define the grand canonical partition function

Then, the normalization condition clearly requires that

Therefore is the free energy of the grand canonical ensemble, and the entropy is given by

We now introduce the fugacity defined to be

Then, the grand canonical partition function can be written as

which allows us to view the grand canonical partition function as a function of the thermodynamic variables and .
Other thermodynamic quantities follow straightforwardly:

Thus,

This page titled 6.2: Partition Functions is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark
Tuckerman.

6.2.3 [Link]
6.3: Ideal Gas
Recall the canonical partition function expression for the ideal gas:

Define the thermal wavelength as

which has a quantum mechanical meaning as the width of the free particle distribution function. Here it serves as a useful
parameter, since the canonical partition can be expressed as

The grand canonical partition function follows directly from :

Thus, the free energy is

In order to obtain the equation of state, we first compute the average particle number

Thus, eliminating in favor of in the equation of state gives

as expected. Similarly, the average energy is given by

where the fugacity has been eliminated in favor of the average particle number. Finally, the entropy

which is the Sackur-Tetrode equation derived in the context of the canonical and microcanonical ensembles.

This page titled 6.3: Ideal Gas is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark E. Tuckerman.

6.3.1 [Link]
6.4: Particle Number Fluctuations
In the grand canonical ensemble, the particle number is not constant. It is, therefore, instructive to calculate the fluctuation in this
quantity. As usual, this is defined to be

Note that

Thus,

In order to calculate this derivative, it is useful to introduce the Helmholtz free energy per particle defined as follows:

where is the volume per particle. The chemical potential is defined by

Similarly, the pressure is given by

Also,

Therefore,

and

6.4.1 [Link]
But recall the definition of the isothermal compressibility:

Thus,

and

and the relative fluctuation is given by

Therefore, in the thermodynamic limit, the particle number fluctuations vanish, and the grand canonical ensemble is equivalent to
the canonical ensemble.

This page titled 6.4: Particle Number Fluctuations is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
Mark Tuckerman.

6.4.2 [Link]
 CHAPTER OVERVIEW

7: Distribution Functions and Liquid Structure

7.1: General Formulation of Distribution Functions
7.2: General Distribution Functions and Correlation Functions
7.3: Structure and distribution functions in classical liquids and gases
7.4: General Correlation Functions
7.5: Thermodynamic quantities in terms of g(r)
7.6: The Pair Correlation Function
7.7: Derivation of the Van der Waals equation

7: Distribution Functions and Liquid Structure is shared under a not declared license and was authored, remixed, and/or curated by LibreTexts.

1
7.1: General Formulation of Distribution Functions
Recall the expression for the configurational partition function:

Suppose that the potential can be written as a sum of two contributions

where is, in some sense, small compared to . An extra bonus can be had if the partition function for can be evaluated
analytically.
Let
mtd

Then, we may express as

where means average with respect to only. If is small, then the average can be expanded in powers of :

The free energy is given by

Separating into two contributions, we have

where is independent of and is given by

mtd

and

We wish to develop an expansion for of the general form

mtd

where are a set of expansion coefficients that are determined by the condition that such an expansion be consistent with
.
Using the fact that

7.1.1 [Link]
we have that

Equating this expansion to the proposed expansion for , we obtain

This must be solved for each of the undetermined parameters , which can be done by equating like powers of on both sides of
the equation. Thus, from the term, we find, from the right side:

and from the left side, the and term contributes:

from which it can be easily seen that

Likewise, from the term,

and from the left side, we see that the and terms contribute:

Thus,

For , the right sides gives:

the left side contributes the and terms:

Thus,

Now, the free energy, up to the third order term is given by

In order to evaluate , suppose that is given by a pair potential

7.1.2 [Link]
Then,

The free energy is therefore given by

This page titled 7.1: General Formulation of Distribution Functions is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by Mark Tuckerman.

7.1.3 [Link]
7.2: General Distribution Functions and Correlation Functions
We begin by considering a general N-particle system with Hamiltonian

For simplicity, we consider the case that all the particles are of the same type. Having established the equivalence of the ensembles
in the thermodynamic limit, we are free to choose the ensemble that is the most convenient on in which to work. Thus, we choose
to work in the canonical ensemble, for which the partition function is

The 3N integrations over momentum variables can be done straightforwardly, giving

where is the thermal wavelength and the quantity is known as the configurational partition function

The quantity

represents the probability that particle 1 will be found in a volume element at the point , particle 2 will be found in a volume
element at the point ,..., particle N will be found in a volume element at the point . To obtain the probability associated
with some number n;SPMlt;N of the particles, irrespective of the locations of the remaining n+1,...,N particles, we simply integrate
this expression over the particles with indices n+1,...,N:

The probability that any particle will be found in the volume element at the point and any particle will be found in the
volume element at the point ,...,any particle will be found in the volume element at the point is defined to be

which comes about since the first particle can be chosen in N ways, the second chosen in N-1 ways, etc.
Consider the special case of . Then, by the above formula,

Thus, if we integrate over all , we find that

Thus, actually counts the number of particles likely to be found, on average, in the volume element at the point . Thus,
integrating over the available volume, one finds, not surprisingly, all the particles in the system.

7.2.1 [Link]
This page titled 7.2: General Distribution Functions and Correlation Functions is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by Mark Tuckerman.

7.2.2 [Link]
7.3: Structure and distribution functions in classical liquids and gases
So far, we have developed the classical theory of ensembles and applied it to the ideal gas, for which there was no potential of
interaction between the particles: . We were able to derive all the thermodynamics for this system using the various ensembles
available to us, and, in particular, we could compute the equation of state. Now, we wish to consider the general case that the
potential is not zero. Of course, all of the interesting physics and chemistry of real systems results from the specific interactions
between the particles. Real systems can exhibit spatial structure, undergo phase transitions, undergo chemical changes, exhibit
interesting dynamics, basically, a wide variety of rich behavior.
Consider the two snapshots below.

On the left is shown a configuration of an ideal gas, and on the right is shown a configuration of liquid argon. Can you see any
inherent structure in the snapshot of liquid argon? While it may not be readily apparent, there is considerable structure in the liquid
argon system that is clearly not present in the ideal gas. One way of quantifying spatial structure is through the use of the radial
distribution function g(r), which will be discussed in great detail later. For now, it is sufficient to know that g(r) is a measure of the
probability that a particle will be located a distance r from a another particle in the system. The figure below shows the function
g(r) for the ideal gas and for the liquid argon systems.

It can be seen that the radial distribution function for the ideal gas is completely featureless signifying that it is equally likely to
find a particle at any distance from a given particle. (Since the probability is uniform, for small . This is the particularly

7.3.1 [Link]
normalization condition on that gives rise to uniform probability. Hence its rapidly rising behavior for small .) For the liquid
argon system, exhibits several peaks, indicating that at certain radial values, it is more likely to find particles than at others.
This is a result of the attractive nature of the interaction at such distances. The plot of also shows that there is essentially zero
probability of finding particles at distances less than about 2.5 Å from each other. This is due to the presence of very strong
repulsive forces at short distances.
Sometimes, the structure can be readily seen in snapshots of configurations. Consider the following snapshot of a system of water
molecules:

The red spheres are oxygen atoms, the grey spheres are hydrogen atoms, the green lines are hydrogen bonds, and the reddish-grey
lines are covalent bonds. A good deal of structure can be seen in the form of a complex network of hydrogen bonds. This high
degree of structure is characteristic of water and gives rise to the ease which with water can form stable, organized structures
around other molecules. In water, one can ask several questions related to structure. For example, what is the probability that an
oxygen atom will be found at a distance away from another oxygen atoms? What is the probability that a hydrogen atom will be
located at a distance r from an oxygen atom, etc. The plot below shows the radial distribution functions corresponding to these two
scenarios.

The peak in the O-O radial distribution function occurs at roughly 2.8Å which is the well known average hydrogen bond length in
water. Of course, one could also ask about structure from the point of view of a hydrogen atom and obtain two other representations

7.3.2 [Link]
of structure in water.

This page titled 7.3: Structure and distribution functions in classical liquids and gases is shared under a CC BY-NC-SA 4.0 license and was
authored, remixed, and/or curated by Mark Tuckerman.

7.3.3 [Link]
7.4: General Correlation Functions
A general correlation function can be defined in terms of the probability distribution function according to

Another useful way to write the correlation function is

i.e., the general n-particle correlation function can be expressed as an ensemble average of the product of -functions, with the
integration being taken over the variables .

This page titled 7.4: General Correlation Functions is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
Mark Tuckerman.

7.4.1 [Link]
7.5: Thermodynamic quantities in terms of g(r)
In the canonical ensemble, the average energy is given by

Therefore,

Since

Thus,

In order to compute the average energy, therefore, one needs to be able to compute the average of the potential . In general, this
is a nontrivial task, however, let us work out the average for the case of a pairwise-additive potential of the form

i.e., U is a sum of terms that depend only the distance between two particles at a time. This form turns out to be an excellent
approximation in many cases. U therefore contains N(N-1) total terms, and becomes

The second line follows from the fact that all terms in the first line are the exact same integral, just with the labels changed. Thus,

Once again, we change variables to and . Thus, we find that

7.5.1 [Link]
Therefore, the average energy becomes

Thus, we have an expression for E in terms of a simple integral over the pair potential form and the radial distribution function. It
also makes explicit the deviation from "ideal gas'' behavior, where E=3NkT/2.
By a similar procedure, we can develop an equation for the pressure P in terms of g(r). Recall that the pressure is given by

The volume dependence can be made explicit by changing variables of integration in to

Using these variables, becomes

Carrying out the volume derivative gives

Thus,

Let us consider, once again, a pair potential. We showed in an earlier lecture that

where is the force on particle i due to particle j. By interchaning the i and j summations in the above expression, we obtain

However, by Newton's third law, the force on particle i due to particle j is equal and opposite to the force on particle j due to
particle i:

Thus,

7.5.2 [Link]
The ensemble average of this quantity is

As before, all integrals are exactly the same, so that

Then, for a pair potential, we have

where u'(r) = du/dr, and . Substituting this into the ensemble average gives

As in the case of the average energy, we change variables at this point to and . This gives

Therefore, the pressure becomes

which again gives a simple expression for the pressure in terms only of the derivative of the pair potential form and the radial
distribution function. It also shows explicitly how the equation of state differs from the that of the ideal gas .
From the definition of g(r) it can be seen that it depends on the density and temperature T: . Note, however, that
the equation of state, derived above, has the general form

which looks like the first few terms in an expansion about ideal gas behavior. This suggests that it may be possible to develop a
general expansion in all powers of the density about ideal gas behavior. Consider representing as such a power series:

7.5.3 [Link]
Substituting this into the equation of state derived above, we obtain

This is known as the virial equation of state, and the coefficients are given by

are known as the virial coefficients. The coefficient is of particular interest, as it gives the leading order deviation from ideal
gas behavior. It is known as the second virial coefficient. In the low density limit, and is directly related
to the radial distribution function.

This page titled 7.5: Thermodynamic quantities in terms of g(r) is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by Mark Tuckerman.

7.5.4 [Link]
7.6: The Pair Correlation Function
Of particular importance is the case n=2, or the correlation function known as the pair correlation function. The explicit
expression for is

In general, for homogeneous systems in equilibrium, there are no special points in space, so that should depend only on the
relative position of the particles or the difference . In this case, it proves useful to introduce the change of variables

Then, we obtain a new function , a function of and :

In general, we are only interested in the dependence on . Thus, we integrate this expression over and obtain a new correlation
function defined by

For an isotropic system such as a liquid or gas, where there is no preferred direction in space, only the maginitude or is of
relevance. Thus, we seek a choice of coordinates that involves r explicitly. The spherical-polar coordinates of the vector is the
most natural choice. If then the spherical polar coordinates are

where and are the polar and azimuthal angles, respectively. Also, note that

where

Thus, the function that depends only on the distance between two particles is defined to be

7.6.1 [Link]
Integrating over the radial dependence, one finds that

The function is important for many reasons. It tells us about the structure of complex, isotropic systems, as we will see below,
it determines the thermodynamic quantities at the level of the pair potential approximation, and it can be measured in neutron and
X-ray diffraction experiments. In such experiments, one observes the scattering of neutrons or X-rays from a particular sample. If a
detector is placed at an angle from the wave-vector direction of an incident beam of particles, then the intensity that one
observes is proportional to the structure factor

where is the vector difference in the wave vector between the incident and scattered neutrons or X-rays (since neutrons and X-
rays are quantum mechanical particles, they must be represented by plane waves of the form ). By computing the
ensemble average (see problem 4 of problem set #5), one finds that and S(k) is given by

Thus, if one can measure , can be determined by Fourier transformation.

This page titled 7.6: The Pair Correlation Function is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
Mark E. Tuckerman.

7.6.2 [Link]
7.7: Derivation of the Van der Waals equation
As a specific example of the application of perturbation theory, we consider the Van der Waals equation of state. Let be given by
a pair potential:

with

This potential is known as the hard sphere potential. In the low-density limit, the radial distribution function can be shown to be
given correctly by or

is taken to be some arbitrary attractive potential, whose specific form is not particularly important. Then, the
full potential might look like:

Now, the first term in is

where is a number that depends on and the specific form of .

Since the potential is a hard sphere potential, can be determined analytically. If were 0, then would
describe an ideal gas and
mtd

However, because two particles may not approach each other closer than a distance between their centers, there is some
excluded volume:

If we consider two hard spheres at closest contact and draw the smallest imaginary sphere that contains both particles, then we find
this latter sphere has a radius :

Hence the excluded volume for these two particles is

7.7.1 [Link]
and hence the excluded volume per particle is just half of this:

Therefore is the total excluded volume, and we find that, in the low density limit, the partition function is given
approximately by

Thus, the free energy is

If we now use this free energy to compute the pressure from

we find that

This is the well know Van der Waals equation of state. In the very low density limit, we may assume that

hence

Thus,

from which we can approximate the second virial coefficient:

A plot of the isotherms of the Van der Waals equation of state is shown below:

The red and blue isotherms appear similar to those of an ideal gas, i.e., there is a monotonic decrease of pressure with increasing
volume. The black isotherm exhibits an unusual feature not present in any of the ideal-gas isotherms - a small region where the
curve is essentially horizontal (flat) with no curvature. At this point, there is no change in pressure as the volume changes. Below
this isotherm, the Van der Waals starts to exhibit unphysical behavior. The green isotherm has a region where the pressure decreases
with decreasing volume, behavior that is not expected on physical grounds. What is observed experimentally, in fact, is that at a
certain pressure, there is a dramatic, discontinuous change in the volume. This dramatic jump in volume signifies that a phase
transition has occurred, in this case, a change from a gaseous to a liquid state. The dotted line shows this jump in the volume. Thus,
the small flat neighborhood along the black isotherm becomes larger on isotherms below this one. The black isotherm just

7.7.2 [Link]
represents a boundary between those isotherms along which no such phase transition occurs and those that exhibit phase transitions
in the form of discontinuous changes in the volume. For this reason, the black isotherm is called the critical isotherm, and the point
at which the isotherm is flat and has zero curvature is called a critical point.
A critical point is a point at which

Using these two conditions, we can solve for the critical volume ( ) and critical temperature ( ):

and the critical pressure is therefore

Using these values for the critical pressure, temperature and volume, we can show that the isothermal compressibility, given by

diverges as the critical point is approached. To see this, note that

Thus,

It is observed that at a critical point, diverges, generally, as . To determine the

heat capacity, note that

so that

Then, since

it follows that

The heat capacity is observed to diverge as . Exponents such as and are known as critial exponents.
Finally, one other exponent we can easily determine is related to how the pressure depends on density near the critical point. The
exponent is called , and it is observed that

7.7.3 [Link]
What does our theory predict for
? To determine we expand the equation of state about the critical density and
temperature:

Unknown environment 'algin*'

The second and third terms vanish by the conditions of the critical point. The third derivative term can be worked out
straightforwardly and does not vanish. Rather

Thus, we see that, by the above expansion, .

The behavior of these quantities near the critical temperature determine three critical exponents. To summarize the results, the Van
der Waals theory predicts that

The determination of critical exponents such as these is an active area in statistical mechanical research. The reason for this is that
such exponents can be grouped into universality classes - groups of systems with exactly the same sets of critical exponents. The
existence of universality classes means that very different kinds of systems exhibit essentially the same behavior at a critical point,
a fact that makes the characterization of phase transitions via critical exponents quite general. The values obtained above for ,
and are known as the mean-field exponents and shows that the Van der Waals theory is really a mean
field theory.
These exponents do not agree terribly well with experimental values ( ). However, the simplicity of mean-field
theory and its ability to give, at least, qualitative results, makes it, nevertheless, useful. To illustrate universality classes, it can be
shown that, within mean field theory, the Van der Waals gas/liquid and a magnetic system composed of spins at particular lattice
sites, which composes the so called Ising model, have exactly the same mean field theory exponents, despite the completely
different nature of these two systems.
Another problem with the Van der Waals theory that is readily apparent is the fact that it predicts for certain values of the density
. Such behavior is unstable. Possible ways of improving the approximations used are the following:
1. Improve the approximation to .
2. Choose a better zeroth order potential .
3. Go to a higher order in perturbation theory.
Barker and Henderson have shown that going to second order in perturbation theory (see discussion in McQuarrie's book, chapter
14), yields well converged results for a square well fluid, compared to "exact'' results from a molecular dynamics simulation.

This page titled 7.7: Derivation of the Van der Waals equation is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by Mark Tuckerman.

7.7.4 [Link]
 CHAPTER OVERVIEW

8: Rare-event sampling and free energy calculations

Our treatment of the classical ensembles makes clear that the free energy is a quantity of particular importance in statistical
mechanics. Being related to the logarithm of the partition function, the free energy is the generator through which other
thermodynamic quantities are obtained, via differentiation. In many cases, the free energy difference between two thermodynamic
states is sought. Such differences tell, for example, whether or not a chemical reaction can occur spontaneously or requires input of
work and is directly related to the equilibrium constant for the reaction. Thus, for example, from free energy differences, one can
compute solvation free energies, acid ionization constants and associated values, or drug inhibition constants , that
quantify the ability of a compound to bind to the active site of an enzyme. Another type of free energy often sought is the free
energy as a function of one or more generalized coordinates in a system. An example might be the free energy surface as a function
of a pair of Ramachandran angles and in an oligopeptide. Such a surface would provide a map of the stable conformations of
the molecule, the relative stability of such conformations and the heights of barriers that need to be crossed in a conformational
change.
8.2: Free-energy Perturbation Theory
8.3: Adiabatic Switching and Thermodynamic Integration
8.4: Reaction Coordinates
8.5: Jarzynski's Equality and Nonequilibrium Methods
8.6: The "blue moon'' Ensemble Approach

8: Rare-event sampling and free energy calculations is shared under a not declared license and was authored, remixed, and/or curated by
LibreTexts.

1
8.2: Free-energy Perturbation Theory
We begin our treatment of free energy differences by examining the problem of transforming a system from one thermodynamic
state to another. Let these states be denoted generically as and . At the microscopic level, these two states are characterized by
potential energy functions and . For example, in a drug-binding study, the state might correspond to
the unbound ligand and enzyme, while would correspond to the bound complex. In this case, the potential would exclude all
interactions between the enzyme and the ligand and the enzyme, whereas they would be included in the potential .
The Helmholtz free energy difference between the states and is simply . The two free energies and are
given in terms of their respective canonical partition functions and , respectively by and ,
where

The free energy difference is, therefore,

where and are the configurational partition functions for states and , respectively,

The ratio of full partition functions reduces to the ratio of configurational partition functions because the
momentum integrations in the former cancel out of the ratio.
Equation is difficult to implement in practice because in any numerical calculation via either molecular dynamics or Monte
Carlo, we do not have direct access to the partition function only averages of phase-space functions corresponding to physical
observables. However, if we are willing to extend the class of phase-space functions whose averages we seek to functions that do
not necessarily correspond to direct observables, then the ratio of configurational partition functions can be manipulated to be in the
form of such an average. Consider inserting unity into the expression for as follows:

If we now take the ratio , we find

where the notation indicates an average taken with respect to the canonical configurational distribution of the state .
Substituting Equation into Equation , we find

8.2.1 [Link]
Equation is known as the free-energy perturbation formula; it should be reminiscent of the thermodynamic perturbation
formula used to derive the van der Waals equation. Equation can be interpreted as follows: We start with microstates
selected from the canonical ensemble of state and use these to compute by placing them in the state by simply
changing the potential energy from to . In so doing, we need to "unbias'' our choice to sample the configurations from the
canonical distribution of state by removing the weight factor from which the microstates are sample and
reweighting the states by the factor corresponding to state . This leads to Equation . The difficulty with this
approach is that the microstates corresponding to the canonical distribution of state may not be states of high probability in the
canonical distribution of state . If this is the case, then the potential enegy difference will be large, he exponential factor
will be negligibly small, and the free energy difference will be very slow to converge in an actual simulation.
For this reason, it is clear that the free-energy perturbation formula is only useful for cases in which the two states and are not
that different from each other.
If is not a small perturbation to , then the free-energy perturbation formula can still be salvaged by introducing a set of
intermediate states with potentials , where , corresponds to the state and corresponds to the
state . Let . We can now imagine transforming the system from state to state by passing through these
intermediate states and computing the average of in the state . Applying the free-energy perturbation formula to this
protocol yields the free-energy difference as

where means an average taken over the distribution . The key to applying Equation is choosing the
intermediate states so as to achieve sufficient overlap between the intermediate states without requiring a large number of them, i.e.
choosing the thermodynamic path between states and effectively.

This page titled 8.2: Free-energy Perturbation Theory is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
Mark Tuckerman.

8.2.2 [Link]
8.3: Adiabatic Switching and Thermodynamic Integration
The free-energy perturbation approach evokes a physical picture in which configurations sampled from the canonical distribution
of state are immediately "switched'' to the state by simply changing the potential from to . Such "instantaneous''
switching clearly represents an unphysical path from one state to the other, but we need not concern ourselves with this because the
free energy is a state function and, therefore, independent of the path connecting the states. Nevertheless, we showed that the free-
energy perturbation theory formula, Equation (6), is only useful if the states and do not differ vastly from one another, thus
naturally raising the question of what can be done if the states are very different.
The use of a series of intermediate states, by which Equation (7) is derived, exploits the fact that any path between the states can be
employed to obtain the free energy difference. In this section, we will discuss an alternative approach in which the system is
switched slowly or adiabatically from one state to the other, allowing the system to fully relax at each point along a chosen path
from state to state , rather than instantaneously switching the system between intermediate states, as occurs in Equation (7). In
order to effect the switching from one state to the other, we will employ a common trick in the form of an "external'' switching
parameter, . This parameter is introduced by defining a new potential energy function

The functions and are referred to as switching functions, and they required to satisfy the conditions ,
corresponding to the state , and , corresponding to the state . Apart from these conditions, and are
completely arbitrary. The mechanism embodied in Equation is one in which some imaginary external controlling influence
("hand of God''), represented by the parameter, starts the system off in state and slowly switches off the potential
while simultaneously switching on the potential . The process is complete when , when the system is in state . A simple
choice for the functions and is, for example, and .
In order to see how Equation can be used to compute the free energy difference , consider the canonical partition function
of a system described by the potential of Equation for a particular choice of :

This partition function leads to a free energy via

Recall, however, that the derivatives of the free energy with repsect to , and and lead to the chemical potential, pressure and
entropy, respectively. What does the derivative of the free energy with respect to represent? According to Equation

The reader should check that the expressions involving and are equivalent. Computing the derivative of with respect to , we
find

Now, the free energy difference can be obtained trivially from the relation

Substituting eqns. and into Equation , we obtain the free energy difference as

8.3.1 [Link]
where denotes an average over the canonical ensemble described by the distribution with fixed at a
particular value. The special choice of and has a simple interpretation. For this choice, Equation
becomes

The content of Equation (15) can be understood by recalling the relationship between work and free energy from the second law of
thermodynamics. If, in transforming the system from state to state , an amount of work is performed on the system, then

where equality holds only if the transformation is carried out along a reversible path. Since reversible work is related to a change in
potential energy, Equation is actually a statistical version of Equation for the special case of equality. Equation (15)
tells us that the free energy difference is the ensemble average of the microscopic reversible work needed to change the potential
energy of each configuration from to along the chosen -path. Note, however, that Equation (14), which is known as the
thermodynamic integration formula, is true independent of the choice of and , which means that Equation (14) always
yields the reversible work via the free energy difference. The flexibility in the choice of the -path, however, can be exploited to
design adiabatic switching algorithms of greater efficiency that can be achieved with the simple choice .
In practice, the thermodynamic integration formula is implemented as follows: A set of values of is chosen from the interval
, and at each chosen value , a full molecular dynamics or Monte Carlo calculation is carried out in order to generate the
average . The resulting values of , are then substituted into Equation (14), and the resulted is integrated
numerically to produce the free energy difference . Thus, we see that the selected values can be evenly spaced, for
example, or they could be a set of Gaussian quadrature nodes, depending on how is expected to vary with for the
chosen and .
As with free-energy perturbation theory, the thermodynamic integration approach can be implemented very easily. An immediately
obvious disadvantage of the method, however, is the same one that applies to Equation (7): In order to perform the numerical
integration, it is necessary to perform many simulations of a system at physically uninteresting intermediate values of where the
potential is, itself, unphysical. Only correspond to actual physical states and ultimately, we can only attach
physical meaning to the free energy difference . Nevertheless, the intermediate averages must be
accurately calculated in order for the integration to yield a correct result. The approach to be presented in the next section attempts
to reduce the time spent in such unphysical intermediate states and focuses the sampling in the important regions .

This page titled 8.3: Adiabatic Switching and Thermodynamic Integration is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by Mark Tuckerman.

8.3.2 [Link]
8.4: Reaction Coordinates
It is frequently the case that the progress of some chemical, mechanical, or thermodynamics process can be followed by following
the evolution of a small subset of generalized coordinates in a system. When generalized coordinates are used in this manner, they
are typically referred to as reaction coordinates, collective variables, or order parameters, often depending on the context and type
of system. Whenever referring to these coordinates, we will refer to them as reaction coordinates, although the reader should be
aware that the other two designations are also used in the literature.
As an example of a useful reaction coordinate, consider a simple gas-phase diatomic dissociation process

If and denote the Cartesian coordinates of atom A and B, then a useful generalized coordinate for following the progress of
the dissociation is simply the distance . A complete set of generalized coordinates that contains as one of the
coordinates is the set that contains the center of mass

the magnitude of the relative coordinate

and the two angles and , where , and are the components of the relative coordinate

Of course, in the gas-phase, where the potential between and likely only depends on the distance between and , is really
the only interesting coordinate. However, if the reaction were to take place in solution, then other coordinate such as and
become more relevant as specific orientations might change the mechanism or thermodynamic picture of the process, depending on
the complexity of the solvent, and averaging over these degrees of freedom to produce a free energy profile in alone will
wash out some of this information.
As another example, consider a gas-phase proton transfer reaction

Here, although the distance can be used to monitor the progress of the proton away from A and the distance can
be used to monitor the progress of the proton toward B, neither distance alone is sufficient for following the progress of the
reaction. However, the difference

can be used to follow the progress of the proton transfer from A to B and, therefore, is a potentially useful reaction coordinate. A
complete set of generalized coordinates involving can be constructed as follows. If , and denote the Cartesian coordinates
of the three atoms, then first introduce the center-of-mass

the relative coordinate between A and B, , and a third relative coordinate between H and the center-of-mass of A and
B,

Finally, is transformed into spherical polar coordinates, , and from and , three more coordinates are formed:

and the angle , which measures the "tilt'' of the plane containing the three atoms from the vertical. The coordinates are
known as confocal elliptic coordinates. These coordinates could also be used if the reaction takes place in solution. As expected,

8.4.1 [Link]
the generalized coordinates are functions of the original Cartesian coordinates. The alanine-dipeptide example above also employs
the Ramachandran angles and as reaction coordinates, and these can also be expressed as part of a set of generalized
coordinates that are functions of the original Cartesian coordinates of a system.
While reaction coordinates or collective variables are potentially very useful constructs, they must be used with care, particularly
when enhanced sampling methods are applied to them. Enhanced sampling of a poorly chosen reaction coordinate can bias the
system in unnatural ways, leading to erroneous predictions of free energy barriers and associated mechanisms. A dramatic example
of this is the autodissociation of liquid water following the classic reaction

the ostensibly only requires transferring a proton from one water molecule to another. If this notion of the reaction is pursued, then
a seemingly sensible reaction coordinate would simply be the distance between the oxygen and the transferring proton or the
number of hydrogens covalently bonded to the oxygen. These reaction coordinates, as it turns out, are inadequate for describing the
true nature of the reaction and, therefore, fail to yield reasonable free energies (and hence, values of the autoionization constant
). Chandler and coworkers showed that the dissociation reaction can only be considered to have occurred when the and
ions are sufficiently far apart that no contiguous or direct path of hydrogen-bonding in the liquid can allow the proton to
transfer back to the water or its origin. In order to describe such a process correctly, a very different type of reaction coordinate
would clearly be needed.
Keeping in mind such caveats about the use of reaction coordinates, we now proceed to describe a number of popular methods
designed to enhance sampling along pre-selected reaction coordinates. All of these methods are designed to generate, either directly
or indirectly, the probability distribution function of a subset of reaction coordinates of interest in a system. If these
reaction coordinates are obtained from a transformation of the Cartesian coordinates , , then the
probability density that these coordinates will have values in the canonical ensemble is

where the -functions are introduced to fix the reaction coordinates at values at . Once is known, the
free energy hypersurface in these coordinates is given by

This page titled 8.4: Reaction Coordinates is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark E.
Tuckerman.

8.4.2 [Link]
8.5: Jarzynski's Equality and Nonequilibrium Methods
In this section, the relationship between work and free energy will be explored in greater detail. We have already introduced the
inequality in Equation , which states that if an amount of work is performed on a system, taking from state to state ,
then . Here, equality holds only if the work is performed reversibly. The work referred to here is thermodynamic
quantity and, as such, must be regarded as an ensemble average. In statistical mechanics, we can also introduce the mechanical or
microscopic work performed on one member of the ensemble to drive it from state to state . Then, is simply an
ensemble average of . However, we need to be somewhat careful about how we define this ensemble average because the
work is defined along a particular path or trajectory which takes the system from state to state , and equilibrium averages do
not refer not to paths but to microstates. This distinction is emphasized by the fact that the work could be carried out irreversibly,
such that the system is driven out of equilibrium. Thus, the proper definition of the ensemble average follows along the lines
already discussed in the context of the free-energy perturbation approach, namely, averaging over the canonical distribution for the
state . In this case, since we will be discussing actual paths , we let the initial condition be the phase space vector for the
system in the (initial) state . Recall that is a unique function of the initial conditions. Then

and the Clausius inequality can be stated as .

From such an inequality, it would seem that using the work as a method for calculating the free energy is of limited utility, since the
work necessarily must be performed reversibly, otherwise one obtains only upper bound on the free energy. It turns out, however,
that irreversible work can be used to calculate free energy differences by virtue of a connection between the two quantities first
discovered in 1997 by C. Jarzynski that as come to be known as the Jarzynski equality. This equality states that if, instead of
averaging over the initial canonical distribution (that of state ), an average of is performed over the
same distribution, the result is , i.e.

This remarkable result not only provides a foundation for the development of nonequilibrium free-energy methods but also has
profound implications for thermodynamics, in general.
The Jarzynski equality be proved using different strategies. Here, however, we will present a proof that is most relevant for the
finite-sized systems and techniques employed in molecular dynamics calculations. Consider a time-dependent Hamiltonian of the
form

For time-dependent Hamiltonian's, the usual conservation law no longer holds, which can be seen by computing

where the phase space vector has been introduced. Integrating both sides over time from to a
final time , we find

Equation can be regarded as a microscopic version of the first law of thermodynamics, in which the first and second terms
represent the heat absorbed by the system and the work done on the system over the trajectory, respectively. Note that the work is
actually a function of the initial phase-space vector , which can be seen by writing this term explicitly as

8.5.1 [Link]
where the fact that the work depends explicitly on in Equation is indicated by the subscript. In the present discussion, we
will consider that each initial condition, selected from a canonical distribution in , evolves according to Hamilton's equations in
isolation. In this case, the heat term , and we have the usual addition to Hamilton's equations .
With the above condition, we can write the microscopic work as

The last term is also . Thus, the ensemble average of the exponential of the work becomes

The numerator in this expression becomes much more interesting if we perform a change of variables from to . Since the
solution of Hamilton's equations for the time-dependent Hamiltonian uniquely map the initial condition onto , when , we
have a new set of phase-space variables, and by Liouville's theorem, the phase-space volume element is preserved

When the Hamiltonian is transformed, we find . Consequently,

thus proving the equality. The implication of the Jarzynski equality is that the work can be carried out along a reversible or
irreversible path, and the correct free energy will still be obtained.
Note that due to Jensen's inequality:

Using Jarzynski's equality, this becomes

which implies, as expected, that

This page titled 8.5: Jarzynski's Equality and Nonequilibrium Methods is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by Mark E. Tuckerman.

8.5.2 [Link]
8.6: The "blue moon'' Ensemble Approach
The term "blue moon'' in the present context describes rare events, i.e. events that happen once in a blue moon. The blue moon
ensemble approach was introduced by Ciccotti and coworkers as a technique for computing the free energy profile along a reaction
coordinate direction characterized by one or more barriers high enough that they would not likely be crossed in a normal
thermostatted molecular dynamics calculation.
Suppose a process of interest can be monitored by a single reaction coordinate \( q_1=f_1({\bf r}_1,...,{\bf r}_N) \) so that eqns.
(29) and (30) reduce to
\[ \begin{align} P(s) &= { {C_N \over Q(N,V,T)}\int\;d^N{\bf p}\;d^N{\bf r}e^{-\beta H({\bf p},{\bf r})}\delta(f_1({\bf r}_1,...,
{\bf r}_N)-s) } \\[4pt] &= { {1 \over N!\lambda^{3N} Q(N,V,T)}\int\;d^N{\bf r}e^{-\beta U({\bf r})}\delta(f_1({\bf r}_1,...,{\bf
r}_N)-s) }\\[4pt] A (s) &=-kT\ln P(s) \label{31} \end{align}\]
The "1'' subscript on the value \(s\) of \( {q_1} \) is superfluous and will be dropped throughout this discussion. In the second line,
the integration over the momenta has been performed giving the thermal prefactor factor \(\lambda ^{3N} \). In the blue moon
ensemble approach, a holonomic constraint \( \sigma({\bf r}_1,...,{\bf r}_N) =f_1({\bf r}_1,...,{\bf r}_N)-s \) is introduced in a
molecular dynamics calculation as a means of "driving'' the reaction coordinate from an initial value \( {s_i} \) to a final value \(
{s_f} \) via a set of intermediate points \( {s_1,...,s_n } \) between \( {s_i} \) and \( {s_f} \). Unfortunately, the introduction of a
holonomic, constraint does not yield the single \(\delta \)-function condition \( \delta(\sigma({\bf r}) =\delta(f_1({\bf r})-s) \),
where \( {{\bf r}\equiv {\bf r}_1,...,{\bf r}_N }\) required by Equation \ref{31} but rather the product of \(\delta \)-functions \
(\delta(\sigma({\bf r}))\delta(\dot{\sigma}({\bf r},{\bf p})) \), since both the constraint and its first time derivative are imposed in a
constrained dynamics calculation. We will return to this point a bit later in this section. In addition to this, the blue moon ensemble
approach does not yield \(A(s) \) directly but rather the derivative
\[ {dA \over ds} = -{kT \over P(s)}{dP \over ds} \label{32}\]
from which the free energy profile \(A (q) \) along the reaction coordinate and the free energy difference \( \Delta A = A(s_f)-A(s_i)
\) are given by the integrals
\[ A(q) = A(s_i) + \int_{s_i}^q ds {dA \over ds}\;\;\;\;\;\;\;\;\;\;\Delta A = \int_{s_i}^{s_f} ds {dA \over ds} \label{33}\]
In the free-energy profile expression \(A (s_i)\) is just an additive constant that can be left off. The values \( {s_1,...,s_n }\) at which
the reaction coordinate is constrained can be chosen at equally-spaced intervals between \( {s_i} \) and \( {s_f} \), in which a
standard numerical quadrature can \(q=f_1({\bf r}) \) be applied for evaluating the integrals in Equation \ref{33}, or they can be
chosen according to a more sophisticated quadrature scheme.
We next turn to the evaluation of the derivative in Equation \ref{32}. Noting that \( P(s) = \langle \delta(f_1({\bf r})-s)\rangle \), the
derivative can be written as
\[ {1 \over P(s)}{dP \over ds} ={C_N \over Q(N,V,T)}{\int\;d^N {\bf p} d^N {\bf r} e^{-\beta H_{(p,r)}} {\partial \over \partial
s}\delta(f_1({\bf r})-s)\over \langle \delta(f_1({\bf r})-s)\rangle} \label{34}\]
In order to avoid evaluating the derivative of the \(\delta \)-function, an integration by parts can be used. First, we introduce a
complete set of \(3N \) generalized coordinates:
\[ q_{\alpha} = f_{\alpha}({\bf r}_1,...,{\bf r}_N) \label{35}\]
and their conjugate momenta \( {p_{\alpha} } \). Such a transformation has a unit Jacobian so that \( d^N{\bf p}\;d^N{\bf r}=
d^{3N}p\;d^{3N}q \). Denoting the transformed Hamiltonian as \( {\tilde{H}(p,q) }\), Equation \ref{34} becomes
\[ {1 \over P(s)}{dP \over ds} ={C_N \over Q(N,V,T)}{\int\;d^{3N} P d^{3N} q e^{-\beta \tilde {H} (p, q)}{\partial \over \partial
s}\delta(q_1-s)\over \langle \delta(q_1-s)\rangle} \label{36}\]
Changing the derivative in front of the \(\delta \)-function from \(\partial/\partial s \) to \(\partial/\partial q_1\), which introduces an
overall minus sign, and then integrating by parts yields
\[\begin{align} {1 \over P(s)}{dP \over ds} &= {C_N \over Q(N,V,T)}{\int\;d^{3N}p\;d^{3N}q\;\left[{\partial \over \partial
q_1}e^{-\beta \tilde{H}(p,q)}\right]\delta(q_1-s)\over \langle \delta(q_1-s)\rangle} \\[4pt] &= -{\beta C_N \over Q(N,V,T)}

8.6.1 [Link]
{\int\;d^{3N}p\;d^{3N}q\;{\partial\tilde {H} \over \partial q_1}e^{-\beta \tilde{H}(p,q)}\delta(q_1-s)\over \langle \delta(q_1-
s)\rangle} \\[4pt] & = -\beta {\left<\left({\partial \tilde{H} \over \partial q_1}\right)\delta(q_1-s)\right>\over \langle \delta(q_1-
s)\rangle} \label{37} \end{align}\]
The last line defines a new ensemble average, specifically an average subject to the condition (not constraint) that the coordinate \(
{q_1} \) have the particular value \(s\). This average will be denoted \(\langle\cdots\rangle^{\rm cond}_{s} \). Thus, the derivative
becomes
\[ {1 \over P(s)}{dP \over ds} =-\beta\left<{\partial \tilde{H} \over \partial q_1}\right>^{\rm cond}_s \label{38}\]
Substituting Equation \ref{38} yields a free energy profile of the form
\[ A(q) = A(s_i) + \int_{s_i}^q\;ds\;\left<{\partial \tilde{H} \over \partial q_1}\right>^{\rm cond}_s \label{39}\]
from which \(\Delta A \) can be computed by letting \( {q = s_f} \). Given that - \( {\langle \partial \tilde{H}/\partial
q_1\rangle^{\rm cond}_s } \) is the expression for the average of the generalized force on \( {q_1} \) when \( {q_1 = s}\), the
integral represents the work done on the system, i.e. the negative of the work done by the system, in moving from \( {s_i}\) to an
arbitrary final point \( {q} \). Since the conditional average implies a full simulation at each fixed value of \( {q_1}\), the
thermodynamic transformation is certainly carried out reversibly, so that Equation \ref{39} is consistent with the Clausius
inequality.
Although Equation \ref{39} provides a very useful insight into the underlying statistical mechanical expression for the free energy,
technically, the need for a full canonical transformation of both coordinates and momenta is inconvenient since, from the chain rule
\[ {\partial \tilde{H} \over \partial q_1} =\sum_{i=1}^N\left [ {\partial H \over \partial {\bf p} _i } \cdot {\partial {\bf p}_i \over
\partial q_1} + {\partial H \over \partial {\bf r}_i }\cdot{\partial {\bf r}_i \over \partial q_1}\right] \label{40}\]
A more useful expression results if the momenta integrations are performed before introducing the transformation to generalized
coordinates. Starting again with Equation \ref{34}, we carry out the momentum integrations, yielding
\[ {1 \over P(s)}{dP \over ds} ={1 \over N!\lambda^{3N}Q(N,V,T )} {\int d^N {\bf r} e^{-\beta U (r)} {\partial \over \partial s}
\delta(f_1({\bf r})-s)\over \langle \delta(f_1({\bf r})-s)\rangle} \label{41}\]
Now, we introduce only the transformation of the coordinates to generalized coordinates \( q_{\alpha} = f_{\alpha}({\bf r}_1,...,
{\bf r}_N) \). However, because there is no corresponding momentum transformation, the Jacobian of the transformation is not
unity. Let \( J(q)\equiv J(q_1,...,q_{3N}) =\partial ({\bf r}_1,...,{\bf r}_N)/\partial (q_1,...,q_{3N}) \) denote the Jacobian of the
transformation. Then, Equation \ref{41} becomes
\[\begin{align} {1 \over P(s)}{dP \over ds} &= { {1 \over N!\lambda^{3N}Q(N,V,T)}{\int\;d^{3N}q\;J(q)e^{-\beta {\tilde U}
(q)}{\partial \over \partial s}\delta(q_1-s)\over \langle \delta(q_1-s)\rangle}} \\[4pt] &= {1 \over N!\lambda^{3N}Q(N,V,T)}
{\int\;d^{3N}q\;e^{-\beta \left(U(q) - kT J(q) \right )}{\partial \over \partial s}\delta(q_1-s)\over \langle \delta(q_1-s)\rangle}
\label{42} \end{align}\]
where, in the last line, the Jacobian has been exponentiated. Changing the derivative \(\partial/\partial s \) to \( \partial/\partial q_1 \)
and performing the integration by parts as was done in Equation \ref{37}, we obtain
\[\begin{align} {1 \over P(s)}{dP \over ds} &= {1 \over N!\lambda^{3N}Q(N,V,T)}{\int\;d^{3N}q\;{\partial \over \partial
q_1}e^{-\beta \left (\tilde{U}(q)-kT\ln J(q)\right)}\delta(q_1-s)\over \langle \delta(q_1-s)\rangle} \\[4pt] &= -{\beta \over
N!\lambda^{3N}Q(N,V,T)}{\int\;d^{3N}q\;\left[{\partial \tilde {U} \over partial q_1} -KT{\partial \over \partial q_1} \ln J (q)
\right]e^{-\beta \left (\tilde{U}(q)-kT\ln J(q)\right)}\delta(q_1-s)\over \langle \delta(q_1-s)\rangle} \\[4pt] &= {-
\beta\left<\left[{\partial\tilde{U} \over \partial q_1}-kT{\partial \over \partial q_1}\ln J(q)\right]\right>^{\rm cond}_s } \label{43}
\end{align}\]
Therefore, the free energy profile becomes
\[ A(q) = A(s_i) +\int_{s_i}^q\;ds\;\left<\left[{\partial\tilde {U} \over \partial q_1} - KT {\partial \over \partial q_1}\ln
J(q)\right]\right>^{\rm cond}_s \label{44}\]
Again, the derivative of \( {\tilde U} \), the transformed potential, can be computed form the untransformed potential via the chain
rule

8.6.2 [Link]
\[ {\partial \tilde{U} \over \partial q_1} =\sum_{i=1}^N {\partial U \over \partial {\bf r}_i}\cdot {\partial {\bf r}_i \over \partial
q_1} \label{45}\]
Equation \ref{44} is useful for simple reaction coordinates in which the full transformation to generalized coordinates is known.
We will see shortly how the expression for \(A (q) \) can be further simplified in a way that does not require knowledge of the
transformation at all. First, however, we must tackle the problem alluded to earlier of computing the conditional ensemble averages
from the constrained dynamics employed by the blue moon ensemble method.

This page titled 8.6: The "blue moon'' Ensemble Approach is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated
by Mark Tuckerman.

8.6.3 [Link]
 CHAPTER OVERVIEW

9: Review of the basic postulates of quantum mechanics

9.1: Measurement
9.2: Physical Observables
9.3: The Fundamental Postulates of Quantum Mechanics
9.4: The Heisenberg Picture
9.5: The Heisenberg Uncertainty Principle
9.6: The Physical State of a Quantum System
9.7: Time Evolution of the State Vector

Template:HideTOC

This page titled 9: Review of the basic postulates of quantum mechanics is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by Mark Tuckerman.

1
9.1: Measurement
The result of a measurement of the observable must yield one of the eigenvalues of . Thus, we see why is required to be a
hermitian operator: Hermitian operators have real eigenvalues. If we denote the set of eigenvalues of by , then each of the
eigenvalues satisfies an eigenvalue equation

where is the corresponding eigenvector. Since the operator is hermitian and is therefore real, we have also the left
eigenvalue equation

The probability amplitude that a measurement of will yield the eigenvalue is obtained by taking the inner product of the
corresponding eigenvector with the state vector , . Thus, the probability that the value is obtained is given by

Another useful and important property of hermitian operators is that their eigenvectors form a complete orthonormal basis of the
Hilbert space, when the eigenvalue spectrum is non-degenerate. That is, they are linearly independent, span the space, satisfy the
orthonormality condition

and thus any arbitrary vector can be expanded as a linear combination of these vectors:

By multiplying both sides of this equation by and using the orthonormality condition, it can be seen that the expansion
coefficients are

The eigenvectors also satisfy a closure relation:

where is the identity operator.

Averaging over many individual measurements of gives rise to an average value or expectation value for the observable , which
we denote and is given by

That this is true can be seen by expanding the state vector in the eigenvectors of :

where are the amplitudes for obtaining the eigenvalue upon measuring , i.e., . Introducing this expansion into
the expectation value expression gives

9.1.1 [Link]
The interpretation of the above result is that the expectation value of is the sum over possible outcomes of a measurement of
weighted by the probability that each result is obtained. Since is this probability, the equivalence of the
expressions can be seen.
Two observables are said to be compatible if . If this is true, then the observables can be diagonalized simultaneously to
yield the same set of eigenvectors. To see this, consider the action of on an eigenvector of . . But if this
must equal , then the only way this can be true is if yields a vector proportional to which means it must also be an
eigenvector of . The condition can be expressed as

that is

where, in the second line, the quantity is know as the commutator between and . If , then and
are said to commute with each other. That they can be simultaneously diagonalized implies that one can simultaneously predict the
observables and with the same measurement.
As we have seen, classical observables are functions of position and momentum (for a one-particle system). Quantum analogs
of classical observables are, therefore, functions of the operators and corresponding to position and momentum. Like other
observables and are linear hermitian operators. The corresponding eigenvalues and and eigenvectors and satisfy the
equations

which, in general, could constitute a continuous spectrum of eigenvalues and eigenvectors. The operators and are not
compatible. In accordance with the Heisenberg uncertainty principle (to be discussed below), the commutator between and is
given by

and that the inner product between eigenvectors of and is

Since, in general, the eigenvalues and eigenvectors of and form a continuous spectrum, we write the orthonormality and
closure relations for the eigenvectors as:

The probability that a measurement of the operator will yield an eigenvalue in a region about some point is

9.1.2 [Link]
The object is best represented by a continuous function often referred to as the wave function. It is a representation
of the inner product between eigenvectors of with the state vector. To determine the action of the operator on the state vector
in the basis set of the operator , we compute

The action of on the state vector in the basis of the operator is consequential of the incompatibility of and and is given by

Thus, in general, for any observable , its action on the state vector represented in the basis of is

This page titled 9.1: Measurement is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark Tuckerman.

9.1.3 [Link]
9.2: Physical Observables
Physical observables are represented by linear, hermitian operators that act on the vectors of the Hilbert space. If is such an
operator, and is an arbitrary vector in the Hilbert space, then might act on to produce a vector , which we express as

Since is representable as a column vector, is representable as a matrix with components

The condition that must be hermitian means that

or

This page titled 9.2: Physical Observables is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark
Tuckerman.

9.2.1 [Link]
9.3: The Fundamental Postulates of Quantum Mechanics
The fundamental postulates of quantum mechanics concern the following questions:
1. How is the physical state of a system described?
2. How are physical observables represented?
3. What are the results of measurements on quantum mechanical systems?
4. How does the physical state of a system evolve in time?
5. The uncertainty principle.

This page titled 9.3: The Fundamental Postulates of Quantum Mechanics is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by Mark Tuckerman.

9.3.1 [Link]
9.4: The Heisenberg Picture
In all of the above, notice that we have formulated the postulates of quantum mechanics such that the state vector evolves in
time, but the operators corresponding to observables are taken to be stationary. This formulation of quantum mechanics is known as
the Schrödinger picture. However, there is another, completely equivalent, picture in which the state vector remains stationary and
the operators evolve in time. This picture is known as the Heisenberg picture. This particular picture will prove particularly useful
to us when we consider quantum time correlation functions.
The Heisenberg picture specifies an evolution equation for any operator , known as the Heisenberg equation. It states that the time
evolution of is given by

While this evolution equation must be regarded as a postulate, it has a very immediate connection to classical mechanics. Recall
that any function of the phase space variables evolves according to

where is the Poisson bracket. The suggestion is that in the classical limit ( small), the commutator goes over to the Poisson
bracket. The Heisenberg equation can be solved in principle giving

where is the corresponding operator in the Schrödinger picture. Thus, the expectation value of at any time is computed from

where is the stationary state vector.

Let's look at the Heisenberg equations for the operators and . If is given by

then Heisenberg's equations for and are

Thus, Heisenberg's equations for the operators and are just Hamilton's equations cast in operator form. Despite their innocent
appearance, the solution of such equations, even for a one-particle system, is highly nontrivial and has been the subject of a
considerable amount of research in physics and mathematics.
Note that any operator that satisfies will not evolve in time. Such operators are known as constants of the motion. The
Heisenberg picture shows explicitly that such operators do not evolve in time. However, there is an analog with the Schrödinger
picture: Operators that commute with the Hamiltonian will have associated probabilities for obtaining different eigenvalues that do
not evolve in time. For example, consider the Hamiltonian, itself, which it trivially a constant of the motion. According to the
evolution equation of the state vector in the Schrödinger picture,

9.4.1 [Link]
the amplitude for obtaining an energy eigenvalue at time upon measuring will be

Thus, the squared modulus of both sides yields the probability for obtaining , which is

Thus, the probabilities do not evolve in time. Since any operator that commutes with can be diagonalized simultaneously with
and will have the same set of eigenvectors, the above arguments will hold for any such operator.

This page titled 9.4: The Heisenberg Picture is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark
Tuckerman.

9.4.2 [Link]
9.5: The Heisenberg Uncertainty Principle
Because the operators and are not compatible, , there is no measurement that can precisely determine both and
simultaneously. Hence, there must be an uncertainty relation between them that specifies how uncertain we are about one quantity
given a definite precision in the measurement of the other. Presumably, if one can be determined with infinite precision, then there
will be an infinite uncertainty in the other. Recall that we had defined the uncertainty in a quantity by

Thus, for and , we have

These quantities can be expressed explicitly in terms of the wave function using the fact that

and

Similarly,

and

Then, the Heisenberg uncertainty principle states that

which essentially states that the greater certainty with which a measurement of or can be made, the greater will be the
uncertainty in the other.

This page titled 9.5: The Heisenberg Uncertainty Principle is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated
by Mark Tuckerman.

9.5.1 [Link]
9.6: The Physical State of a Quantum System
The physical state of a quantum system is represented by a vector denoted which is a column vector, whose components are
probability amplitudes for different states in which the system might be found if a measurement were made on it.
A probability amplitude is a complex number, the square modulus of which gives the corresponding probability

The number of components of is equal to the number of possible states in which the system might observed. The space that
contains is called a Hilbert space . The dimension of is also equal to the number of states in which the system
might be observed. It could be finite or infinite (countable or not). must be a unit vector. This means that the inner product:

In the above, if the vector , known as a Dirac "ket'' vector, is given by the column

then the vector , known as a Dirac "bra'' vector, is given by

so that the inner product becomes

We can understand the meaning of this by noting that , the components of the state vector, are probability amplitudes, and
are the corresponding probabilities. The above condition then implies that the sum of all the probabilities of being in the various
possible states is 1, which we know must be true for probabilities.

This page titled 9.6: The Physical State of a Quantum System is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by Mark Tuckerman.

9.6.1 [Link]
9.7: Time Evolution of the State Vector
The time evolution of the state vector is prescribed by the Schrödinger equation

where is the Hamiltonian operator. This equation can be solved, in principle, yielding

where is the initial state vector. The operator

is the time evolution operator or quantum propagator. Let us introduce the eigenvalues and eigenvectors of the Hamiltonian
that satisfy

The eigenvectors for an orthonormal basis on the Hilbert space and therefore, the state vector can be expanded in them according to

where, of course, , which is the amplitude for obtaining the value at time if a measurement of is performed.
Using this expansion, it is straightforward to show that the time evolution of the state vector can be written as an expansion:

Thus, we need to compute all the initial amplitudes for obtaining the different eigenvalues of , apply to each the factor
and then sum over all the eigenstates to obtain the state vector at time .
If the Hamiltonian is obtained from a classical Hamiltonian , then, using the formula from the previous section for the action
of an arbitrary operator on the state vector in the coordinate basis, we can recast the Schrödiner equation as a partial
differential equation. By multiplying both sides of the Schrödinger equation by , we obtain

If the classical Hamiltonian takes the form

then the Schrödinger equation becomes

which is known as the Schrödinger wave equation or the time-dependent Schrödinger equation. In a similar manner, the eigenvalue
equation for can be expressed as a differential equation by projecting it into the basis:

9.7.1 [Link]
where is an eigenfunction of the Hamiltonian.

This page titled 9.7: Time Evolution of the State Vector is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
Mark Tuckerman.

9.7.2 [Link]
 CHAPTER OVERVIEW

10: Fundamentals of quantum statistical mechanics

10.1: Principles of quantum statistical mechanics
10.2: The Density Matrix and Density Operator
10.3: Time evolution of the density operator
10.4: A simple example - the quantum harmonic oscillator
10.5: The quantum equilibrium ensembles
10.5.1: The microcanonical ensemble
10.5.2: The canonical ensemble
10.5.3: Isothermal-isobaric and grand canonical ensembles

Template:HideTOC

This page titled 10: Fundamentals of quantum statistical mechanics is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by Mark Tuckerman.

1
10.1: Principles of quantum statistical mechanics
The problem of quantum statistical mechanics is the quantum mechanical treatment of an -particle system. Suppose the
corresponding -particle classical system has Cartesian coordinates

and momenta

and Hamiltonian

Then, as we have seen, the quantum mechanical problem consists of determining the state vector from the Schrödinger
equation

Denoting the corresponding operators, and , we note that these operators satisfy the commutation relations:

and the many-particle coordinate eigenstate is a tensor product of the individual eigenstate :

The Schrödinger equation can be cast as a partial differential equation by multiplying both sides by :

where the many-particle wave function is . Similarly, the expectation value of an operator
is given by

This page titled 10.1: Principles of quantum statistical mechanics is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by Mark Tuckerman.

10.1.1 [Link]
10.2: The Density Matrix and Density Operator
In general, the many-body wave function is far too large to calculate for a macroscopic system. If we wish to
represent it on a grid with just 10 points along each coordinate direction, then for , we would need total points, which
is clearly enormous.
We wish, therefore, to use the concept of ensembles in order to express expectation values of observables without requiring
direct computation of the wavefunction. Let us, therefore, introduce an ensemble of systems, with a total of members, and each
having a state vector , . Furthermore, introduce an orthonormal set of vectors ) and expand the state
vector for each member of the ensemble in this orthonormal set:

The expectation value of an observable, averaged over the ensemble of systems is given by the average of the expectation value of
the observable computed with respect to each member of the ensemble:

Substituting in the expansion for , we obtain

Let us define a matrix

and a similar matrix

Thus, is a sum over the ensemble members of a product of expansion coefficients, while is an average over the ensemble of
this product. Also, let . Then, the expectation value can be written as follows:

where and represent the matrices with elements and in the basis of vectors . The matrix is known as the density
matrix. There is an abstract operator corresponding to this matrix that is basis-independent. It can be seen that the operator

and similarly

have matrix elements when evaluated in the basis set of vectors .

10.2.1 [Link]
Note that is a hermitian operator

so that its eigenvectors form a complete orthonormal set of vectors that span the Hilbert space. If and represent the
eigenvalues and eigenvectors of the operator , respectively, then several important properties they must satisfy can be deduced.
Firstly, let be the identity operator . Then, since , it follows that

Thus, the eigenvalues of must sum to 1. Next, let be a projector onto an eigenstate of , . Then

But, since can be expressed as

and the trace, being basis set independent, can be therefore be evaluated in the basis of eigenvectors of , the expectation value
becomes

However,

Thus, . Combining these two results, we see that, since and , , so that satisfy the properties of
probabilities.
With this in mind, we can develop a physical meaning for the density matrix. Let us now consider the expectation value of a
projector onto one of the eigenstates of the operator . The expectation value of this operator is given by

10.2.2 [Link]
But is just probability that a measurement of the operator in the th member of the ensemble will yield the
result . Thus,

or the expectation value of is just the ensemble averaged probability of obtaining the value in each member of the ensemble.
However, note that the expectation value of can also be written as

Equating the two expressions gives

The interpretation of this equation is that the ensemble averaged probability of obtaining the value if is measured is equal to
the probability of obtaining the value in a measurement of if the state of the system under consideration were the state ,
weighted by the average probability that the system in the ensemble is in that state. Therefore, the density operator (or )
plays the same role in quantum systems that the phase space distribution function plays in classical systems.

This page titled 10.2: The Density Matrix and Density Operator is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by Mark Tuckerman.

10.2.3 [Link]
10.3: Time evolution of the density operator
The time evolution of the operator can be predicted directly from the Schrödinger equation. Since is given by

the time derivative is given by

where the second line follows from the fact that the Schrödinger equation for the bra state vector is

Note that the equation of motion for differs from the usual Heisenberg equation by a minus sign! Since is constructed from
state vectors, it is not an observable like other hermitian operators, so there is no reason to expect that its time evolution will be the
same. The general solution to its equation of motion is

The equation of motion for can be cast into a quantum Liouville equation by introducing an operator

In term of , it can be seen that satisfies

What kind of operator is ? It acts on an operator and returns another operator. Thus, it is not an operator in the ordinary sense,
but is known as a superoperator or tetradic operator (see S. Mukamel, Principles of Nonlinear Optical Spectroscopy, Oxford
University Press, New York (1995)).
Defining the evolution equation for this way, we have a perfect analogy between the density matrix and the state vector. The two
equations of motion are

We also have an analogy with the evolution of the classical phase space distribution , which satisfies

with being the classical Liouville operator. Again, we see that the limit of a commutator is the classical Poisson
bracket.

10.3.1 [Link]
This page titled 10.3: Time evolution of the density operator is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by Mark Tuckerman.

10.3.2 [Link]
10.4: A simple example - the quantum harmonic oscillator
As a simple example of the trace procedure, let us consider the quantum harmonic oscillator. The Hamiltonian is given by

and the eigenvalues of are

Thus, the canonical partition function is

This is a geometric series, which can be summed analytically, giving

The thermodynamics derived from it as as follows:

1.
Free energy:
The free energy is

2.
Average energy:
The average energy is

3.
Entropy
The entropy is given by

Now consider the classical expressions. Recall that the partition function is given by

Thus, the classical free energy is

In the classical limit, we may take to be small. Thus, the quantum expression for becomes, approximately, in this limit:

and we see that

10.4.1 [Link]
The residual (which truly vanishes when ) is known as the quantum zero point energy. It is a pure quantum effect and is
present because the lowest energy quantum mechanically is not but the ground state energy .

This page titled 10.4: A simple example - the quantum harmonic oscillator is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by Mark Tuckerman.

10.4.2 [Link]
10.5: The quantum equilibrium ensembles
At equilibrium, the density operator does not evolve in time; thus, . Thus, from the equation of motion, if this holds, then
, and is a constant of the motion. This means that it can be simultaneously diagonalized with the Hamiltonian and can
be expressed as a pure function of the Hamiltonian

Therefore, the eigenstates of , the vectors, we called are the eigenvectors of the Hamiltonian, and we can write and as

The choice of the function determines the ensemble.

This page titled 10.5: The quantum equilibrium ensembles is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated
by Mark Tuckerman.

10.5.1 [Link]
10.5.1: The microcanonical ensemble
Although we will have practically no occasion to use the quantum microcanonical ensemble (we relied on it more heavily in
classical statistical mechanics), for completeness, we define it here. The function , for this ensemble, is

where is the Heaviside step function. This says that is 1 if and 0 otherwise. The partition function
for the ensemble is , since the trace of is the number of members in the ensemble:

The thermodynamics that are derived from this partition function are exactly the same as they are in the classical case:

etc.

This page titled 10.5.1: The microcanonical ensemble is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
Mark Tuckerman.

[Link] [Link]
10.5.2: The canonical ensemble
In analogy to the classical canonical ensemble, the quantum canonical ensemble is defined by

Thus, the quantum canonical partition function is given by

and the thermodynamics derived from it are the same as in the classical case:

etc. Note that the expectation value of an observable is

Evaluating the trace in the basis of eigenvectors of (and of ), we obtain

The quantum canonical ensemble will be particularly useful to us in many things to come.

This page titled 10.5.2: The canonical ensemble is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark
Tuckerman.

[Link] [Link]
10.5.3: Isothermal-isobaric and grand canonical ensembles
Also useful are the isothermal-isobaric and grand canonical ensembles, which are defined just as they are for the classical cases:
isothermal-isobaric:

grand canonical ensembles

This page titled 10.5.3: Isothermal-isobaric and grand canonical ensembles is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by Mark Tuckerman.

[Link] [Link]
 CHAPTER OVERVIEW

11: Introduction to path integrals in quantum mechanics and quantum statistical

mechanics
11.1: Discretized and Continuous Path Integrals
11.1.1: Derivation of the Discretized Path Integral
11.1.2: Doing the Path Integral - the Free Particle
11.1.3: Dominant Paths in the Propagator and Density Matrix
11.1.4: The Continuous Limit
11.2: Calculation of observables from path integrals
11.2.1: Expectation values of observables
11.2.2: Path integrals for N-particle systems
11.2.3: Path integral molecular dynamics (optional reading)
11.2.4: Thermodynamics from path integrals
11.3: Expansion about the classical path and stationary phase
11.3.1: The harmonic Oscillator - Expansion about the Classical Path
11.3.2: The Stationary Phase Approximation

11: Introduction to path integrals in quantum mechanics and quantum statistical mechanics is shared under a not declared license and was
authored, remixed, and/or curated by LibreTexts.

1
 SECTION OVERVIEW

11.1: Discretized and Continuous Path Integrals

11.1.1: Derivation of the Discretized Path Integral

11.1.2: Doing the Path Integral - the Free Particle

11.1.3: Dominant Paths in the Propagator and Density Matrix

11.1.4: The Continuous Limit

Template:HideTOC

This page titled 11.1: Discretized and Continuous Path Integrals is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by Mark Tuckerman.

11.1.1 [Link]
11.1.1: Derivation of the Discretized Path Integral
We begin our discussion of the Feynman path integral with the canonical ensemble. The expressions for the partition function and
expectation value of an observable are, respectively

It is clear that we need to be able to evaluate traces of the type appearing in these expressions. We have already derived expressions
for these in the basis of eigenvectors of . However, since the trace is basis independent, let us explore carrying out these traces in
the coordinate basis. We will begin with the partition function and treat expectation values later.
Consider the ensemble of a one-particle system. The partition function evaluated as a trace in the coordinate basis is

We see that the trace involves the diagonal density matrix element . Let us solve the more general problem of any
density matrix element .
If the Hamiltonian takes the form

then we cannot evaluate the operator explicitly because the operators for kinetic and potential energies do not
commute with each other, being, respectively, functions of momentum and position, i.e.,

In this instance, we will make use of the Trotter theorem, which states that given two operators and , such that , then
for any number ,

Thus, for the Boltzmann operator,

and the partition function becomes

Define the operator in brackets to be :

Then,

In between each of the factors of , the coordinate space identity operator

is inserted. Since there are factors, there will be such insertions. the integration variables will be labeled . Thus,
the expression for the matrix element becomes

[Link] [Link]
The next step clearly involves evaluating the matrix elementx

Note that in the above expression, the operators involving the potential act on their eigenvectors and can thus be replaced by
the corresponding eigenvalues:

In order to evaluate the remaining matrix element, we introduce the momentum space identity operator

Letting , the matrix remaining matrix element becomes

Using the fact that

it follows that

The remaining integral over can be performed by completing the square, leading to the result

Collecting the pieces together, and introducing the limit, we have for the density matrix

The partition function is obtained by setting , which is equivalent to setting and integrating over , or equivalently
. Thus, the expression for becomes

where we have introduced a "frequency''

When expressed in this way, the partition function, for a finite value of , is isomorphic to a classical configuration integral for a
-particle system, that is a cyclic chain of particles, with harmonic nearest neighbor interactions and interacting with an external

[Link] [Link]
potential . That is, the partition function becomes

where

Thus, for finite (if large) the partition function in the discretized path integral representation can be treated as any ordinary
classical configuration integral. Consider the integrand of in the limit that all points on the cyclic chain are at the same
location . Then the harmonic nearest neighbor coupling (which is due to the quantum kinetic energy) vanishes and
, and the integrand becomes

which is just the true classical canonical position space distribution function. Therefore, the greater the spatial spread in the cyclic
chain, the more "quantum'' the system is, since this indicates a greater contribution from the quantum kinetic energy. The spatially
localized it is, the more the system behaves like a classical system.
It remains formally to take the limit that . There we will see an elegant formulation for the density matrix and partition
function emerges.

This page titled 11.1.1: Derivation of the Discretized Path Integral is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by Mark Tuckerman.

[Link] [Link]
11.1.2: Doing the Path Integral - the Free Particle
The density matrix for the free particle

will be calculated by doing the discrete path integral explicitly and taking the limit at the end.
The density matrix expression is

Let us make a change of variables to

The inverse of this transformation can be worked out explicitly, giving

The Jacobian of the transformation is simply

Let us see what the effect of this transformation is for the case . For , one must evaluate

According to the inverse formula,

Thus, the sum of squares becomes

From this simple example, the general formula can be deduced:

Thus, substituting this transformation into the integral gives

[Link] [Link]
where

and the overall prefactor has been written as

Now each of the integrals over the variables can be integrated over independently, yielding the final result

In order to make connection with classical statistical mechanics, we note that the prefactor is just , where

is the kinetic prefactor that showed up also in the classical free particle case. In terms of , the free particle density matrix can be
written as

Thus, we see that represents the spatial width of a free particle at finite temperature, and is called the "thermal de Broglie
wavelength.''

This page titled 11.1.2: Doing the Path Integral - the Free Particle is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by Mark Tuckerman.

[Link] [Link]
11.1.3: Dominant Paths in the Propagator and Density Matrix
Let us first consider the real time quantum propagator. The quantity appearing in the exponential is an integral of

which is known as the Lagrangian in classical mechanics. We can ask, which paths will contribute most to the integral

known as the action integral. Since we are integrating over a complex exponential , which is oscillatory, those paths away
from which small deviations cause no change in (at least to first order) will give rise to the dominant contribution. Other paths
that cause to oscillate rapidly as we change from one path to another will give rise to phase decoherence and will ultimately
cancel when integrated over. Thus, we consider two paths and a nearby one constructed from it and demand that
the change in between these paths be 0

Note that, since and , , since all paths must begin at and end at . The change in is

Expanding the first term to first order in , we obtain

The term proportional to can be handled by an integration by parts:

because vanishes at and , the surface term is 0, leaving us with

Since the variation itself is arbitrary, the only way the integral can vanish, in general, is if the term in brackets vanishes:

This is known as the Euler-Lagrange equation in classical mechanics. For the case that , they give

which is just Newton's equation of motion, subject to the conditions that , . Thus, the classical path and those near it
contribute the most to the path integral.
The classical path condition was derived by requiring that to first order. This is known as an action stationarity principle.
However, it turns out that there is also a principle of least action, which states that the classical path minimizes the action as well.
This is an important consideration when deriving the dominant paths for the density matrix, which takes the form

The action appearing in this expression is

[Link] [Link]
which is known as the Euclidean action and is just the integral over a path of the total energy or Euclidean Lagrangian .
Here, we see that a minimum action principle is needed, since the smallest values of will contribute most to the integral. Again,
we require that to first order . Applying the same logic as before, we obtain the condition

which is just Newton's equation of motion on the inverted potential surface , subject to the conditions , .
For the partition function , the same equation of motion must be solved, but subject to the conditions that , i.e.,
periodic paths.

This page titled 11.1.3: Dominant Paths in the Propagator and Density Matrix is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by Mark E. Tuckerman.

[Link] [Link]
11.1.4: The Continuous Limit
In taking the limit , it will prove useful to define a parameter

so that implies . In terms of , the partition function becomes

We can think of the points as specific points of a continuous functions , where

such that :

Figure 1:
Note that

and that the limit

is just a Riemann sum representation of the continuous integral

Finally, the measure

[Link] [Link]
represents an integral overa all values that the function can take on between and such that . We write
this symbolically as . Therefore, the limit of the partition function can be written as

The above expression in known as a functional integral. It says that we must integrate over all functions (i.e., all values that an
arbitrary function may take on) between the values and . It must really be viewed as the limit of the discretized
integral introduced in the last lecture. The integral is also referred to as a path integral because it implies an integration over all
paths that a particle might take between and such that , where the paths are parameterized by the variable
(which is not time!). The second line in the above expression, which is equivalent to the first, indicates that the integration is taken
over all paths that begin and end at the same point, plus a final integration over that point.
The above expression makes it clear how to represent a general density matrix element :

which indicates that we must integrate over all functions that begin at at and end at at :

Similarly, diagonal elements of the density matrix, used to compute the partition function, are calculated by integrating over all
periodic paths that satisfy :

[Link] [Link]
Note that if we let , then the density matrix becomes

which are the coordinate space matrix elements of the quantum time evolution operator. If we make a change of variables in
the path integral expression for the density matrix, we find that the quantum propagator can also be expressed as a path integral:

Such a variable transformation is known as a Wick rotation. This nomenclature comes about by viewing time as a complex
quantity. The propagator involves real time, while the density matrix involves a transformation to the imaginary time axis.
It is because of this that the density matrix is sometimes referred to as an imaginary time path integral.

This page titled 11.1.4: The Continuous Limit is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark E.
Tuckerman.

[Link] [Link]
 SECTION OVERVIEW

11.2: Calculation of observables from path integrals

Topic hierarchy

11.2.1: Expectation values of observables

11.2.2: Path integrals for N-particle systems

11.2.3: Path integral molecular dynamics (optional reading)

11.2.4: Thermodynamics from path integrals

Template:HideTOC

This page titled 11.2: Calculation of observables from path integrals is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by Mark Tuckerman.

11.2.1 [Link]
11.2.1: Expectation values of observables
Recall the basic formula for the expectation value of an observable :

Two important cases pertaining to the evaluation of the trace in the coordinate basis for expectation values will be considered
below:

Case 1: Functions only of position

If , i.e., a function of the operator only, then the trace can be easily evaluated in the coordinate basis:

Since acts to the left on one of its eigenstates, we have

which only involves a diagonal element of the density matrix. This can, therefore, be written as a path integral:

However, since all points are equivalent, due to the fact that they are all integrated over, we can make equivalent cyclic
renaming of the coordinates , etc. and generate equivalent integrals. In each, the function or , etc.
will appear. If we sum these equivalent integrals and divide by , we get an expression:

This allows us to define an estimator for the observable . Recall that an estimator is a function of the variables whose
average over the ensemble yields the expectation value of :

Then

where the average on the right is taken over many configurations of the variables (we will discuss, in the nex lecture, a
way to generate these configurations).
The limit can be taken in the same way that we did in the previous lecture, yielding a functional integral expression for the
expectation value:

Case 2: Functions only of momentum

Suppose that , i.e., a function of the momentum operator. Then, the trace can still be evaluated in the coordinate basis:

[Link] [Link]
However, acting to the left does not act on an eigenvector. Let us insert a coordinate space identity between
and :

Now, we see that the expectation value can be obtained by evaluating all the coordinate space matrix elements of the operator and
all the coordinate space matrix elements of the density matrix.
A particularly useful form for the expectation value can be obtained if a momentum space identity is inserted:

Now, we see that acts on an eigenstate (at the price of introducing another integral). Thus, we have

Using the fact that , we find that

In the above expression, we introduce the change of variables

Then

Define a distribution function

Then, the expectation value can be written as

which looks just like a classical phase space average using the "phase space'' distribution function . The distribution
function is known as the Wigner density matrix and it has many interesting features. For one thing, its classical limit is

which is the true classical phase space distribution function. There are various examples, in which the exact Wigner distribution
function is the classical phase space distribution function, in particularly for quadratic Hamiltonians. Despite its compelling
appearance, the evaluation of expectation values of functions of momentum are considerably more difficult than functions of
position, due to the fact that the entire density matrix is required. However, there are a few quantities of interest, that are functions
of momentum, that can be evaluated without resorting to the entire density matrix. These are thermodynamic quantities which will
be discussed in the next section.

This page titled 11.2.1: Expectation values of observables is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated
by Mark Tuckerman.

[Link] [Link]
11.2.2: Path integrals for N-particle systems
If particle spin statistics must be treated in a given problem, the formulation of the path integral is more complicated, and we will
not treat this subject here. The extension of path integrals to -particle systems in which spin statistics can safely be ignored,
however, is straightforward, and we will give the expressions below.
The partition function for an -particle system in the canonical ensemble without spin statistics can be formulated essentially by
analogy to the one-particle case. The partition function that one obtains is

Thus, it can be seen that the -particle potential must be evaluated for each imaginary time discretization, however, there is no
coupling between separate imaginary time slices due arising from the potential. Thus, interactions occur only between particles in
the same time slice. From a computational point of view, this is advantageous, as it allows for easily parallelization over imaginary
time slices.
The corresponding energy and pressure estimators for the -particle path integral are given by

This page titled 11.2.2: Path integrals for N-particle systems is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by Mark Tuckerman.

[Link] [Link]
11.2.3: Path integral molecular dynamics (optional reading)
Consider once again the path integral expression for the one-dimensional canonical partition function (for a finite but large value of
):

(
1
)

(the condition is understood). Recall that, according to the classical isomorphism, the path integral expression for the
canonical partition function is isomorphic to the classical configuration integral for a certain -particle system. We can carry this
analogy one step further by introducing into the above expression a set of momentum integrations:

(
2
)

Note that these momentum integrations are completely uncoupled from the position integrations, and if we were to carry out these
momentum integrations, we would reproduce Eq. (1) apart from trivial constants. Written in the form Eq. (2), however, the path
integral looks exactly like a phase space integral for a -particle system. We know from our work in classical statistical mechanics
that dynamical equations of motion can be constructed that will generate this partition function. In principle, one would start with
the classical Hamiltonian

derive the corresponding classical equations of motion and then couple in thermostats. Such an approach has certainly been
attempted with only limited success. The difficulty with this straightforward approach is that the more "quantum'' a system is, the
large the paramester must be chosen in order to converge the path integral. However, if is large, the above Hamiltonian
describes a system with extremely stiff nearest-neighbor harmonic bonds interacting with a very weak potential . It is,
therefore, almost impossible for the system to deviate far harmonic oscillator solutions and explore the entire available phase space.
The use of thermostats can help this problem, however, it is also exacerbated by the fact that all the harmonic interactions are
coupled, leading to a wide variety of time scales associated with the motion of each variable in the Hamiltonian. In order to
separate out all these time scales, one must somehow diagonalize this harmonic interaction. One way to do this is to use normal
mode variables, and this is a perfectly valid approach. However, we will explore another, simpler approach here. It involves the use
of a variable transformation of the formed used in previous lectures to do the path integral for the free-particle density matrix.
Consider a change of variables:

where

The inverse of this transformation can be worked out in closed form:

[Link] [Link]
and can also be expressed as a recursive inverse:

The term here can be used to start the recursion. We have already seen that this transformation diagonalized the harmonic
interaction. Thus, substituting the transformation into the path integral gives:

The parameters are given by

Note also that the momentum integrations have been changed slightly to involve a set of parameters . Introducing these
parameters, again, only changes the partition function by trivial constant factors. How these should be chosen will become clear
later in the discussion. The notation indicates that each variable is a generally a function of all the new variables
.
A dynamics scheme can now be derived using as an effective Hamiltonian:

which, when coupled to thermostats, yields a set of equations of motion

(
= 3
)

These equations have a conserved energy (which is not a Hamiltonian):

Notice that each variable is given its own thermostat. This is done to produce maximum ergodicity in the trajectories. In fact, in
practice, the chain thermostats you have used in the computer labs are employed. Notice also that the time scale of each variable is
now clear. It is just determined by the parameters . Since the object of using such dynamical equations is not to produce real
dynamics but to sample the phase space, we would really like each variable to move on the same time scale, so that there are no
slow beads trailing behind the fast ones. This effect can be produced by choosing each parameter to be proportional to
Finally, the forces on the variables can be determined easily from the chain rule and the recursive inverse given above. The
result is

where the first of these expressions starts the recursion in the second equation.
Later on, when we discuss applications of path integrals, we will see why a formulation such as this for evaluating path integrals is
advantageous.

[Link] [Link]
This page titled 11.2.3: Path integral molecular dynamics (optional reading) is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by Mark Tuckerman.

[Link] [Link]
11.2.4: Thermodynamics from path integrals
Although general functions of momentum are difficult (though not intractable) to evaluate by path integration, certain functions of
momentum (and position) can be evaluated straightforwardly. These are thermodynamic quantities such as the energy and pressure,
given respectively by

We shall derive estimators for these two quantities directly from the path integral expression for the partition function. However, let
us work with the partition function for an ensemble of 1-particle systems in three dimensions, which is given by

Using the above thermodynamic relation, the energy becomes

where

is the thermodynamic estimator for the total energy. Similarly, an estimator for the internal pressure can be derived using
. As we have done in the past for classical systems, the volume dependence can be made explicitly by introducing
the change of variables:

In terms of the scaled variables , the partition function expression reads:

Evaluating the derivative with respect to volume gives the internal pressure:

[Link] [Link]
where

is the thermodynamic estimator for the pressure. Clearly, both the energy and pressure will be functions of the particle momenta,
however, because they are related to the partition function by thermodynamic differentiation, estimators can be derived for them
that do not require the off-diagonal elements of the density matrix.

This page titled 11.2.4: Thermodynamics from path integrals is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by Mark Tuckerman.

[Link] [Link]
 SECTION OVERVIEW

11.3: Expansion about the classical path and stationary phase

11.3.1: The harmonic Oscillator - Expansion about the Classical Path

11.3.2: The Stationary Phase Approximation

Template:HideTOC

This page titled 11.3: Expansion about the classical path and stationary phase is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by Mark Tuckerman.

11.3.1 [Link]
11.3.1: The harmonic Oscillator - Expansion about the Classical Path
It will be shown how to compute the density matrix for the harmonic oscillator:

using the functional integral representation. The density matrix is given by

As we saw in the last lecture, paths in the vicinity of the classical path on the inverted potential give rise to the dominant
contribution to the functional integral. Thus, it proves useful to expand the path about the classical path. We introduce a
change of path variables from to , where

where satisfies

subject to the conditions

so that .
Substituting this change of variables into the action integral yields

An integration by parts makes the cross terms vanish:

where the surface term vanishes because and the second term vanishes because \(xcl\) satisfies the classical
equation of motion.
The first term in the expression for is the classical action, which we have seen is given by

Therefore, the density matrix for the harmonic oscillator becomes

where is the path integral

Note that does not depend on the points and and therefore can only contribute an overall (temperature dependent) constant
to the density matrix. This will affect the thermodynamics but not any averages of physical observables. Nevertheless, it is

[Link] [Link]
important to see how such a path integral is done.
To compute , we note that it is a functional integral over functions that vanish at and . Thus, they are a special
class of periodic functions and can be expanded in a Fourier sine series:

where

Thus, we wish to change from an integral over the functions to an integral over the Fourier expansion coefficients . The two
integrations should be equivalent, as the coefficients uniquely determine the functions . Note that

Thus, terms in the action are:

Since the cosines are orthogonal between and , the integral becomes

similarly,

The measure becomes

which, is not an equivalent measure (since it is not derived from a determination of the Jacobian), but is chosen to give the correct
free-particle ( ) limit, which can ultimately be corrected by attaching an overall factor of .
With this change of variables, becomes

The infinite product can be written as

the product in the square brackets is just the infinite product formula for , so that is just

Finally, attaching the free-particle factor , the harmonic oscillator density matrix becomes:

[Link] [Link]
Notice that in the free-particle limit , and , so that

which is the expected free-particle density matrix.

This page titled 11.3.1: The harmonic Oscillator - Expansion about the Classical Path is shared under a CC BY-NC-SA 4.0 license and was
authored, remixed, and/or curated by Mark Tuckerman.

[Link] [Link]
11.3.2: The Stationary Phase Approximation
Consider the simple integral:

Assume has a global minimum at , such that . If this minimum is well separated from other minima of
and the value of at the global minimum is significantly lower than it is at other minima, then the dominant contributions to the
above integral, as will come from the integration region around . Thus, we may expand about this point:

Since , this becomes:

Inserting the expansion into the expression for gives

Corrections can be obtained by further expansion of higher order terms. For example, consider the expansion of up to fourth
order:

Substituting this into the integrand and further expanding the exponential would give, as the lowest order nonvanishing correction:

This approximation is known as the stationary phase or saddle point approximation. The former may seem a little out-of-place,
since there is no phase in the problem, but that is because we formulated it in such a way as to anticipate its application to the path
integral. But this is only if is taken to be a real instead of an imaginary quantity.
The application to the path integral follows via a similar argument. Consider the path integral expression for the density matrix:

We showed that the classical path satisfying

is a stationary point of the Euclidean action , i.e., . Thus, we can develop a stationary phase or saddle point
approximation for the density matrix by introducing an expansion about the classical path according to

where the correction , satisfying has been expanded in a complete set of orthonormal functions , which
are orthonormal on the interval and satisfy as well as the orthogonality condition:

[Link] [Link]
Setting all the expansion coefficients to 0 recovers the classical path. Thus, we may expand the action (the "E'' subscript will
henceforth be dropped from this discussion) with respect to the expansion coefficients:

Ambiguous use of \over

Since

the expansion can be worked out straightforwardly by substitution and subsequent differentiation:

where the fourth and eighth lines are obtained from an integration by parts. Let us write the integral in the last line in the suggestive
form:

Missing or unrecognized delimiter for \right

which emphasizes the fact that we have matrix elements of the operator with respect to the basis functions.
Thus, the expansion for can be written as

and the density matrix becomes

\[ \rho(x,x';\beta) = {\cal N}\int \prod_j {dc_j \over \sqrt{2\... ...x,x';\beta)} e^{-{1 \over 2}\sum_{j,k}c_j\Delta_{jk}c_k/\hbar} \]
where . is an overall normalization constant. The integral over the coefficients becomes a generalized
Gaussian integral, which brings down a factor of :
\[ \rho(x,x';\beta) = {\cal N}e^{-S_{\rm cl}(x,x';\beta)}{1 \over \sqrt

฀ ParseError: invalid DekiScript (click for details)

\) \( {\cal N}e^{-S_{\rm cl}(x,x';\beta)} {1 \over \sqrt

[Link] [Link]
 ฀ ParseError: invalid DekiScript (click for details)
\]
where the last line is the abstract representation of the determinant. The determinant is called the Van Vleck-Pauli-Morette
determinant.
If we choose the basis functions to be eigenfunctions of the operator appearing in the above expression, so that they satisfy

Then,

and the determinant can be expressed as a product of the eigenvalues. Thus,

The product must exclude any 0-eigenvalues.

Incidentally, by performing a Wick rotation back to real time according to , the saddle point or stationary phase
approximation to the real-time propagator can be derived. The derivation is somewhat tedious and will not be given in detail here,
but the result is

Missing \left or extra \right

where satisfies
\] m\ddot{x}_{\rm cl} = -\left.{\partial U \over \partial x}\right\vert _{x=x_{\rm cl}}\) \( x_{\rm cl}(0) = x\]

and is an integer that increases by 1 each time the determinant vanishes along the classical path. is called the Maslov index. It
is important to note that because the classical paths satisfy an endpoint problem, rather than an initial value problem, there can be
more than one solution. In this case, one must sum the result over classical paths:
\( U(x,x';t) = \sum_{\rm classical\ paths} e^

฀ ParseError: EOF expected (click for details)

\)

This page titled 11.3.2: The Stationary Phase Approximation is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or
curated by Mark Tuckerman.

[Link] [Link]
 CHAPTER OVERVIEW

12: Time-dependent Processes - Classical case

12.1: Perturbative solution of the Liouville equation
12.2: Kubo Transform Expression for the Time Correlation Function
12.3: Generalized Equations of Motion
12.4: General Properties of Time Correlation Functions
12.5: Perturbative solution of the Liouville equation
12.6: The Onsager Fluctuation Regression Theorem
12.7: Relation to Spectra
12.8: Time Correlation Functions and Transport Coefficients
12.8.1: Shear Viscosity
12.8.2: The Diffusion Constant

12: Time-dependent Processes - Classical case is shared under a not declared license and was authored, remixed, and/or curated by LibreTexts.

1
12.1: Perturbative solution of the Liouville equation
As in the classical case, we assume a solution of the form

where

and we will assume

Substituting into the Liouville equation and working to first order in small quantities, we find

which is a first order inhomogeneous equation that can be solved by using an integrating factor:

(Note that we have chosen the origin in time to be , which is an arbitrary choice.)
For an observable , the expectation value is

when the solution for is substituted in, this becomes

where the cyclic property of the trace has been used and the Heisenberg evolution for has been substituted in. Expanding the
commutator gives

where the cyclic property of the trace has been used again. Define a function

called the after effect function. It is essentially the antisymmetric quantum time correlation function, which involves the
commutator between and . Then the linear response result can be written as

12.1.1 [Link]
which is the starting point for the theory of quantum transport coefficients. If we choose to measure the operator , then we find

This page titled 12.1: Perturbative solution of the Liouville equation is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by Mark Tuckerman.

12.1.2 [Link]
12.2: Kubo Transform Expression for the Time Correlation Function
We shall derive the following expression for the quantum time correlation function

known as a Kubo transform relation. Since is given by the Heisenberg equation:

it follows that

Evaluating the expression at gives

Thus,

By performing the trace in the basis of eigenvectors of , we obtain

But

Therefore,

which proves the relation. The classical limit can be deduced easily from the Kubo transform relation:

Note further, by using the cylic properties of the trace, that

12.2.1 [Link]
This page titled 12.2: Kubo Transform Expression for the Time Correlation Function is shared under a CC BY-NC-SA 4.0 license and was
authored, remixed, and/or curated by Mark Tuckerman.

12.2.2 [Link]
12.3: Generalized Equations of Motion
The most general way a system can be driven away from equilibrium by a forcing function is according to the equations of
motion:

where the functions and are required to satisfy the incompressibility condition

in order to insure that the Liouville equation for is still valid. These equations of motion will give rise to a distribution
function satisfying

with . (We assume that is normalized so that .)

What does the Liouville equation say about the nature of in the limit that and are small, so that the displacement away
from equilibrium is, itself, small? To examine this question, we propose to solve the Liouville equation perturbatively. Thus, let us
assume a solution of the form

Note, also, that the equations of motion take a perturbative form

and as a result, the Liouville operator contains two pieces:

where and is assumed to satisfy

means the Hamiltonian part of the equations of motion

For an observable , the ensemble average of is a time-dependent quantity:

which, when the assumed form for is substituted in, gives

where means average with respect to .

This page titled 12.3: Generalized Equations of Motion is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
Mark Tuckerman.

12.3.1 [Link]
12.4: General Properties of Time Correlation Functions
Define a time correlation function between two quantities and by

The following properties follow immediately from the above definition:

Property 1

Property 2

Thus, if , then

known as the autocorrelation function of , and

If we define , then

which just measures the fluctuations in the quantity .

Property 3
A time correlation function may be evaluated as a time average, assuming the system is ergodic. In this case, the phase space
average may be equated to a time average, and we have

which is valid for . In molecular dynamics simulations, where the phase space trajectory is determined at discrete time steps,
the integral is expressed as a sum

where is the total number of time steps, is the time step and .

Property 4: Onsager regression hypothesis

In the long time limit, and eventually become uncorrelated from each other so that the time correlation function becomes

For the autocorrelation function of , this becomes

Thus, decays from at to as .

An example of a signal and its time correlation function appears in Figure . In this case, the signal is the magnitude of the
velocity along the bond of a diatomic molecule interacting with a Lennard-Jones bath. Its time correlation function is shown
beneath the signal:

12.4.1 [Link]
 Figure
Over time, it can be seen that the property being autocorrelated eventually becomes uncorrelated with itself.

This page titled 12.4: General Properties of Time Correlation Functions is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by Mark Tuckerman.

12.4.2 [Link]
12.5: Perturbative solution of the Liouville equation
Substituting the perturbative form for into the Liouville equation, one obtains

Recall . Thus, working to linear order in small quantities, one obtains the following equation for :

which is just a first-order inhomogeneous differential equation. This can easily be solved using an integrating factor, and one
obtains the result

Note that

But, using the chain rule, we have

Define

which is known as the dissipative flux. Thus, for a Cartesian Hamiltonian

where is the force on the th particle, the dissipative flux becomes:

In general,

Now, suppose is a canonical distribution function

then

12.5.1 [Link]
so that

Thus, the solution for is

The ensemble average of the observable now becomes

Recall that the classical propagator is . Thus the operator appearing in the above expression is a classical propagator of the
unperturbed system for propagating backwards in time to . An observable evolves in time according to

Now, if we take the complex conjugate of both sides, we find

where now the operator acts to the left on . However, since observables are real, we have

which implies that forward evolution in time can be achieved by acting to the left on an observable with the time reversed classical
propagator. Thus, the ensemble average of becomes

where the quantity on the last line is an object we have not encountered yet before. It is known as an equilibrium time correlation
function. An equilibrium time correlation function is an ensemble average over the unperturbed (canonical) ensemble of the
product of the dissipative flux at with an observable evolved to a time . Several things are worth noting:
1. The nonequilibrium average , in the linear response regime, can be expressed solely in terms of equilibrium averages.
2. The propagator used to evolve to is the operator , which is the propagator for the unperturbed,
Hamiltonian dynamics with . That is, it is just the dynamics determined by .
3. Since is a function of the phase space variables evolved to a time , we must now specify over
which set of phase space variables the integration is taken. The choice is actually arbitrary, and for convenience, we choose
the initial conditions. Since is a function of the initial conditions , we can write the time correlation function as

This page titled 12.5: Perturbative solution of the Liouville equation is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by Mark Tuckerman.

12.5.2 [Link]
12.6: The Onsager Fluctuation Regression Theorem
Suppose that is of the form

which adiabatically induces a fluctuation in the system for and the lets the system evolve in time according to the unperturbed
Hamiltonian for . How will the induced fluctuation evolve in time? Combining the Kubo transform relation with the linear
response result for , we find that

where the change of variables has been made. Taking the limit , and performing the integral over , we find

Since we assumed that , we have . Thus, dividing by , we find

Thus at long times in the classical limit, the fluctuations decay to 0, indicting a complete regression or suppression of the induced
fluctuation:

This page titled 12.6: The Onsager Fluctuation Regression Theorem is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by Mark Tuckerman.

12.6.1 [Link]
12.7: Relation to Spectra
Suppose that is a monochromatic field

where the parameter insures that field goes to 0 at . We will take at the end of the calculation. The expectation
value of then becomes

where the change of integration variables has been made.

Define a frequency-dependent susceptibility by

then

If we let , then we see immediately that

i.e., the susceptibility is just the Laplace transform of the after effect function or the time correlation function.
Recall that

Under time reversal, we have

Thus,

and if , then

Therefore

12.7.1 [Link]
From the properties of it follows that

so that is positive for and negative for . It is a straightforward matter, now, to show that the energy
difference derived in the lecture from the Fermi golden rule is related to the susceptibility by

This page titled 12.7: Relation to Spectra is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark
Tuckerman.

12.7.2 [Link]
 SECTION OVERVIEW

12.8: Time Correlation Functions and Transport Coefficients

12.8.1: Shear Viscosity

12.8.2: The Diffusion Constant

This page titled 12.8: Time Correlation Functions and Transport Coefficients is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by Mark Tuckerman.

12.8.1 [Link]
12.8.1: Shear Viscosity
The shear viscosity of a system measures is resistance to flow. A simple flow field can be established in a system by placing it
between two plates and then pulling the plates apart in opposite directions. Such a force is called a shear force, and the rate at
which the plates are pulled apart is the shear rate. A set of microscopic equations of motion for generating shear flow is

where is a parameter known as the shear rate. These equations have the conserved quantity

The physical picture of this dynamical system corresponds to the presence of a velocity flow field shown in the figure.
The flow field points in the direction and increases with increasing -value. Thus, layers of a fluid, for example, will slow past
each other, creating an anisotropy in the system. From the conserved quantity, one can see that the momentum of a particle is the
value of plus the contribution from the field evaluated at the position of the particle

Figure
Such an applied external shearing force will create an asymmetry in the internal pressure. In order to describe this asymmetry, we
need an analog of the internal pressure that contains a dependence on specific spatial directions. Such a quantity is known as the
pressure tensor and can be defined analogously to the isotropic pressure that we encountered earlier in the course. Recall that an
estimator for the pressure was

and in equilibrium. Here, is the volume of the system. By analogy, one can write down an estimator for the pressure
tensor :

and

where is a unit vector in the direction, . This (nine-component) pressure tensor gives information about spatial
anisotropies in the system that give rise to off-diagonal pressure tensor components. The isotropic pressure can be recovered from

[Link] [Link]
which is just 1/3 of the trace of the pressure tensor. While most systems have diagonal pressure tensors due to spatial isotropy, the
application of a shear force according to the above scheme gives rise to a nonzero value for the component of the pressure tensor
. In fact, is related to the velocity flow field by a relation of the form

where the coefficient is known as the shear viscosity and is an example of a transport coefficient. Solving for we find

where is the nonequilibrium average of the pressure tensor estimator using the above dynamical equations of motion.
Let us apply the linear response formula to the calculation of the nonequilibrium average of the component of the pressure
tensor. We make the following identifications:

Thus, the dissipative flux becomes

According to the linear response formula,

so that the shear viscosity becomes

Recall that means average of a canonical distribution with . It is straightforward to show that for an
equilibrium canonical distribution function. Finally, taking the limit that in the above expression gives the result

which is a relation between a transport coefficient, in this case, the shear viscosity coefficient, and the integral of an equilibrium
time correlation function. Relations of this type are known as Green-Kubo relations. Thus, we have expressed a new kind of
thermodynamic quantity to an equilibrium time correlation function, which, in this case, is an autocorrelation function of the
component of the pressure tensor.

This page titled 12.8.1: Shear Viscosity is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark
Tuckerman.

[Link] [Link]
12.8.2: The Diffusion Constant
The diffusive flow of particles can be studied by applying a constant force to a system using the microscopic equations of motion

which have the conserved energy

Since the force is applied in the direction, there will be a net flow of particles in this direction, i.e., a current . Since this current
is a thermodynamic quantity, there is an estimator for it:

and . The constant force can be considered as arising from a potential field

The potential gradient will give rise to a concentration gradient which is opposite to the potential gradient and related
to it by

However, Fick's law tells how to relate the particle current to the concentration gradient

where is the diffusion constant. Solving for gives

Let us apply the linear response formula again to the above nonequilibrium average. Again, we make the identification:

Thus,

In equilibrium, it can be shown that there are no cross correlations between different particles. Consider the initial value of the
correlation function. From the virial theorem, we have

which vanishes for . In general,

[Link] [Link]
Thus,

In equilibrium, being linear in the velocities (hence momenta). Thus, the diffusion constant is given by, when the limit
is taken,

However, since no spatial direction is preferred, we could also choose to apply the external force in the or directions and
average the result over the these three. This would give a diffusion constant

The quantity

is known as the velocity autocorrelation function, a quantity we will encounter again in other contexts.

This page titled 12.8.2: The Diffusion Constant is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark
Tuckerman.

[Link] [Link]
 CHAPTER OVERVIEW

13: Time-dependent Processes - Quantum Case

13.1: Calculation of spectra from perturbation theory
13.1.1: The Hamiltonian
13.1.2: The Transition Rate
13.1.3: Examples
13.2: Iterative solution for the interaction-picture state vector
13.3: The Interaction Picture
13.4: Fermi's Golden Rule
13.5: Quantum Linear Response Theory

13: Time-dependent Processes - Quantum Case is shared under a not declared license and was authored, remixed, and/or curated by LibreTexts.

1
 SECTION OVERVIEW

13.1: Calculation of spectra from perturbation theory

13.1.1: The Hamiltonian

13.1.2: The Transition Rate

13.1.3: Examples

Template:HideTOC

This page titled 13.1: Calculation of spectra from perturbation theory is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by Mark Tuckerman.

13.1.1 [Link]
13.1.1: The Hamiltonian
Consider a quantum system with a Hamiltonian . Suppose this system is subject to an external driving force such that the
full Hamiltonian takes the form

where is an operator through which this coupling occurs. This is the situation, for example, when the infrared spectrum is
measured experimentally - the external force is identified with an electric field and is identified with the electric dipole
moment operator. If the field is inhomogeneous, then takes the more general form

where the sum is taken over Fourier modes. Often, is an operator such that, if , then

Suppose we take to be a monochromatic field of the form

Generally, the external field can induce transitions between eigenstates of in the system. Consider such a transition between an
initial state and a final state , with energies and , respectively:

(see figure below).

This transition can only occur if

This page titled 13.1.1: The Hamiltonian is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark
Tuckerman.

[Link] [Link]
13.1.2: The Transition Rate
In the next lecture, we will solve the quantum Liouville equation

perturbatively and derive quantum linear response theory. However, the transition rate can actually be determined directly within
perturbation theory using the Fermi Golden Rule approximation, which states that the probability of a transition's occurring per unit
time, , is given by

The -function expresses the fact that energy is conserved. This describes the rate of transitions between specific states and .
The transition rate between any initial and final states can be obtained by summing over both and and weighting the sum by the
probability that the system is found in the initial state :

where is an eigenvalue of the density matrix, which we will take to be the canonical density matrix:

Using the expression for , we find

Note that

This quantity corresponds to a time-reversed analog of the absorption process. Thus, it describes an emission event with
, i.e., emission of a photon with energy . If can also be expressed as a process by recognizing that

or

Therefore

If we now interchange the summation indices, we find

where the fact that has been used. Comparing this expression for to that for , we find

[Link] [Link]
which is the equation of detailed balance. We see from it that the probability of emission is less than that for absorption. The
reason for this is that it is less likely to find the system in an excited state initially, when it is in contact with a heat bath and
hence thermally equilibrated. However, we must remember that the microscopic laws of motion (Newton's equations for classical
systems and the Schrödinger equation for quantum systems) are reversible. This means that

The conclusion is that, since , reversibility is lost when the system is placed in contact with a heat bath, i.e., the
system is being driven irreversibly in time.
Define

then

Now using the fact that the -function can be written as

becomes

Recall that the evolution of an operator in the Heisenberg picture is given by

if the evolution is determined solely by . Thus, the expression for becomes

which involves the quantum autocorrelation function .

In general, a quantum time correlation function in the canonical ensemble is defined by

In a similar manner, we can show that

[Link] [Link]
since

in general. Also, the product is not Hermitian. However, a hermitian combination occurs if we consider the energy
difference between absorption and emission. The energy absorbed per unit of time by the system is , while the emitted into
the bath by the system per unit of time is . The energy difference is just

But since

it follows that

or

Note, however, that

where is known as the anticommutator:

The anticommutator between two operators is, itself, hermitian. Therefore, the energy difference is

The quantity is the symmetrized quantum autocorrelation function. The classical limit is now manifest (
):

The classically, the energy spectrum is directly related to the Fourier transform of a time correlation function.

This page titled 13.1.2: The Transition Rate is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark
Tuckerman.

[Link] [Link]
13.1.3: Examples
Define

which is just the frequency spectrum corresponding to the autocorrelation function of . For different choices of ,
corresponds to different experimental measurements. Consider the example of a molecule with a transition dipole moment vector .
If an electric field is applied, then the Hamiltonian becomes

If we take , then

Identifying , the spectrum becomes

or for a general electric field, the result becomes

These spectra are the infrared spectra.

As another example, consider a block of material placed in a magnetic field in the direction. The spin of each particle will
couple to the magnetic field giving a Hamiltonian

The net magnetization created by the field is given by

so that

Identify (the extra factor of just expresses the fact that is extensive). Then the spectrum is

which is just the NMR spectrum. In general for each correlation function there is a corresponding experiment that measures its
frequency spectrum.
To see what some specific lineshapes look like, consider as an ansatz a pure exponential decay for the correlation function :

The spectrum corresponding to this time correlation function is

and doing the integral gives

[Link] [Link]
which is shown in the figure below:

Figure : Copy and Paste Caption here. (Copyright; author via source)
We see that the lineshape is a Lorentzian with a width . As a further example, suppose is a decaying oscillatory function:

which describes well the behavior of a harmonic diatomic coupled to a bath. The spectrum can be shown to be

which contains two peaks at as shown in the figure below:

Figure : Copy and Paste Caption here. (Copyright; author via source)

This page titled 13.1.3: Examples is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark Tuckerman.

[Link] [Link]
13.2: Iterative solution for the interaction-picture state vector
The solution to Equation can be expressed in terms of a unitary propagator , the interaction-picture propagator, which
evolves the initial state according to

Substitution of Equation into Equation yields an evolution equation for the propagator :

The initial condition on Equation is . In developing a solution to Equation , we assume that is a small
perturbation, so that the solution can take the form of a sum of powers of .
A solution of this form can be generated by recognizing that Equation can be solved formally in terms of an integral
equation:

It is straightforward to verify this form solution for . Computing the time derivative of both sides gives

Thus, Equation is a valid expression of the solution. The implicit nature of the integral equation means that an iterative
procedure based on the assumption that is a small perturbation can be easily developed. We start with a zeroth-order solution
by setting in Equation , which gives the trivial result

This solution is now fed back into the right side of Equation to develop a first-order solution:

The first order solution is fed back into the right side of Equation to develop a second-order solution:

and so forth, such that the th-order solution is always generated from the st-order solution according to the recursion
formula:

Thus, the third-order solution is given by

The exact solution is then just a sum of the solutions obtained at each order:

13.2.1 [Link]
Having seen how to generate a solution for the propagator in the interaction picture to arbitrarily high orders in the perturbation, the
time evolution of the state vector in the interaction picture can be determined from

and from this expression, the time evolution of the original state vector in the Schrödinger picture can be determined

where we have used the fact that and, in the last line, the full propagator in the Schrödinger picture is identified as

From Equation , the structure of the full propagator for the time-dependent system reveals itself. Let us use Equation
to generate the first few lowest order terms in the propagator. Substituting Equation into Equation yields the lowest
order contribution to :

Thus, at zeroth order, Equation implies that the system is to be propagated using the unperturbed propagator as if
the perturbation did not exist. At first order, we obtain

where, in the second line, the definition of in terms of the original perturbation Hamiltonian has been used. What
Equation says is that at first order, the propagator is composed of two terms. The first term is simply the unperturbed
propagation from to . In the second term, the system undergoes unperturbed propagation from to and at , the perturbation
is allowed to act. From to , the system undergoes unperturbed propagation. Finally, we need to integrate over all possible
intermediate times .
In a similar manner, it can be shown that up to second order, the full propagator is given by

Thus, at second order, the new term involves unperturbed propagation from to , action of at , unperturbed propagation
from to , action of at and, finally, unperturbed propagation from to . Again, the intermediate times and must be
integrated over. The picture on the left side of the equation indicates that the perturbation causes the system to undergo some
undetermined dynamical process between and . The terms on the right show how that process is broken down in terms of the
action of the perturbation at specific intermediate times. At the th order, the perturbation Hamiltonian acts on the system at
specific instances in time. Because of the limits of integration, these time instances are ordered chronologically.
The specific ordering of the instances in time when acts on the unperturbed system raises an important point. At each order the
expansion for , the order in which the operators , , etc. are multiplied is important. The reason for this is that
the operator does not commute with itself at different instances in time

13.2.2 [Link]
Thus, in order to remove any possible ambiguity when specifying the order in which operators are to be applied in a time series, we
introduce the time-ordering operator, . The purpose of is to take a product string of time-dependent operators
which act at different instances in time and order the operators in the product such that they act
chronologically in time from the earliest time to the latest time. For example, the action of on two operators and is

Let us now apply the time-ordering operator to the second-order term. First write the double integral as a sum of two terms
generated simply interchanging the names of the dummy variables and :

The same region can be covered by choosing and . With this choice, Equation becomes

In the first term on the right side of Equation , and acts first, followed by . In the second term,
and acts first followed by . The two terms can, thus, be combined with both and lying in the interval if the
time-ordering operator is applied:

The same analysis can be applied to each order in Equation , recognizing that the number of possible time orderings of a
product of operators is . Thus, Equation can be rewritten in terms of the time-ordering operator as

The sum in Equation resembles the power-series expansion of an exponential, and, indeed, we can write the sum
symbolically as

which is known as a time-ordered exponential. Equation is really a symbolic representation of Equation , in which it
is understood that the time-ordering operator acts to order the operators in each term of the expansion of the exponential.
Given the formalism of time-dependent perturbation theory, we now seek to answer the following question: If the system is initially
in an eigenstate of with energy , what is the probability as a function of time that the system will undergo a transition to a
new eigenstate of with energy ? From the statement of the question, it is clear that the initial state vector is simply the
eigenstate of with energy

The amplitude as a function of time that the system will undergo a transition to the eigenstate is obtained by propagating this
initial state out to time with the propagator and then taking the overlap of the resultant state with the eigenstate :

and the probability is just the square magnitude of this complex amplitude:

13.2.3 [Link]
Consider, first, the amplitude at zeroth order in perturbation theory. At this order, , and the amplitude is simply

which clearly vanishes if . Thus, at zeroth order, the only possibility is the trivial one in which no transition occurs.
The lowest nontrivial order is first order, where the transition amplitude is given by

Define a transition frequency by

Then, taking the absolute square of the last line of Equation , we obtain the probability at first-order

At first order, the probability depends on the matrix element of the perturbation between the initial and final eigenstates. Thus far,
the formalism we have derived is valid for any perturbation Hamiltonian . If we consider the use of an external perturbation to
probe the eigenvalue spectrum of , then the specific type of probe determines the form of , as we saw in the first section
and will explore in the next subsection.

This page titled 13.2: Iterative solution for the interaction-picture state vector is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by Mark Tuckerman.

13.2.4 [Link]
13.3: The Interaction Picture
Consider a quantum system described by a time-dependent Hamiltonian of the form

In the language of perturbation theory, is known as the unperturbed Hamiltonian and describes a system of interest such as a
molecule or a condensed-phase sample such as a pure liquid or solid or a solution. is known as the perturbation, and it often
describes an external system, such as a laser field, that will be used to probe the energy levels and other properties of .
We now seek a solution to the time-dependent Schrödinger equation

subject to an initial state vector . In order to solve the equation, we introduce a new state vector related to by

The new state vector is an equally valid representation of the state of the system. In Chapter 10, we introduced the concept of
pictures in quantum mechanics and discussed the difference between the Schrödinger and Heisenberg pictures. Equation
represents yet another picture of quantum mechanics, namely the interaction picture. Like the Schrödinger and Heisenberg
pictures, the interaction picture is a perfectly valid way of representing a quantum mechanical system. The interaction picture can
be considered as "intermediate'' between the Schrödinger picture, where the state evolves in time and the operators are static, and
the Heisenberg picture, where the state vector is static and the operators evolve.
However, as we will see shortly, in the interaction picture, both the state vector and the operators evolve in time, however, the time-
evolution is determined by the perturbation . Equation specifies how to transform between the Schrödinger and
interaction picture state vectors. The transformation of operators proceeds in an analogous fashion. If denotes an operator in the
Schrödinger picture, its representation in the interaction picture is given by

which is equivalent to an equation of motion of the form

Substitution of Equation into the time-dependent Schrödinger equation yields

According to Equation , the is the interaction-picture representation of the

perturbation Hamiltonian, and we will denote this operator as . Thus, the time-evolution of the state vector in the interaction
picture is given a Schrödinger equation of the form

The initial condition to Equation , is, according to Equation , also . In the next section, we will develop an
iterative solution to Equation , which will reveal a rich structure of the propagator for time-dependent systems.

This page titled 13.3: The Interaction Picture is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark
Tuckerman.

13.3.1 [Link]
13.4: Fermi's Golden Rule
In the first section, we saw how to formulate the Hamiltonian of a material system coupled to an external electromagnetic field.
Moreover, we obtained solutions for the electromagnetic field in the absence of sources or physical boundaries, namely, solutions
of the free-field wave equations. In this chapter, we will focus primarily on weak fields. We will also focus on a class of
experiments in which the wavelength of electromagnetic radiation is taken to be long compared to the size of the sample under
investigation. In this case, the spatial dependence of the electromagnetic field can also be neglected, since

and

in the long-wavelength limit. In this case, it is sufficient to consider to be of the general form

where is a Hermitian operator.

Although we could use and to express the perturbation, the form in Equation is a particularly convenient one, and
since we will be seeking probabilities of transitions, the results we obtain will be real in the end.

Again, the question we seek to answer is given this form for the perturbation, what is the probability that the material system will
be excited from an initial eigenstate with energy to a final state with energy ? However, since the perturbation is
periodic in time, what we really seek to know is if the perturbation is applied over a long time interval, what is the probability per
unit time or rate at which transitions will occur. Thus, in order to make the calculation somewhat easier, let us consider a time
interval and choose and . At first order, the transition rate is just the total probability divided by
the interval length :

For finite , the integral can be carried out explicitly yielding

Thus, the transition rate can be expressed as

In the limit of very large, this expression becomes highly peaked only if . Otherwise, as , the expression vanishes.
The condition is equivalent to the condition , which is a statement of energy conservation. Since is the
energy quantum of the electromagnetic field, the transition can only occur if the energy of the field is exactly "tuned'' for the the
transition, and this "tuning'' depends on the frequency of the field. In this way, the frequency of the field can be used as a probe of
the allowed transitions, which then serves to probe the eigenvalue structure of .
Now, let us consider the more carefully. We shall denote the rate in this limit simply as . In this limit, the integral
becomes

13.4.1 [Link]
Therefore, the expression for the rate in this limit can be written as

where we have dropped the "(1)'' superscript (it is understood that the result is derived from first-order perturbation theory), and
indicate explicitly the dependence on the frequency . When one the first integral is replaced by the -function, the remaining
integral becomes simply , which cancels the in the denominator. Thus, the expression for the rate is finally

which is known as Fermi's Golden Rule. It states that, to first-order in perturbation theory, the transition rate depends only the
square of the matrix element of the operator between initial and final states and includes, via the -function, an energy-
conservation condition. We will make use of the Fermi Golden Rule expression to analyze the application of an external
monochromatic field to an ensemble of systems in order to derive expressions for the observed frequency spectra.

This page titled 13.4: Fermi's Golden Rule is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark
Tuckerman.

13.4.2 [Link]
 SECTION OVERVIEW

13.5: Quantum Linear Response Theory

Consider again the Hamiltonian for a system coupled to a time-dependent field

We wish to solve the quantum Liouville equation

in the linear regime where is small.

Template:HideTOC

This page titled 13.5: Quantum Linear Response Theory is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated
by Mark Tuckerman.

13.5.1 [Link]
 CHAPTER OVERVIEW

14: The Langevin and Generalized Langevin equations

14.1: The Harmonic Bath Hamiltonian
14.2: The Random Force Term
14.3: The Dynamic Friction Kernel
14.4: Relation between the Dynamic Friction Kernel and the Random Force
14.5: Derivation of the GLE
14.6: Mori-Zwanzig Theory- A more general derivation of the GLE
14.7: Example- Vibrational dephasing and energy relaxation

Template:HideTOC

This page titled 14: The Langevin and Generalized Langevin equations is shared under a CC BY-NC-SA 4.0 license and was authored, remixed,
and/or curated by Mark Tuckerman.

1
14.1: The Harmonic Bath Hamiltonian
In the theory of chemical reactions, it is often possible to isolate a small number or even a single degree of freedom in the system
that can be used to characterize the reaction. This degree of freedom is coupled to other degrees of freedom (for example, reactions
often take place in solution). Isomerization or dissociation of a diatomic molecule in solution is an excellent example of this type of
system. The degree of freedom of paramount interest is the distance between the two atoms of the molecule - this is the degree of
freedom whose detailed dynamics we would like to elucidate. The dynamics of the "bath'' or environment to which is couples is
less interesting, but still must be accounted for in some manner. A model that has maintained a certain level of both popularity and
success is the so called "harmonic bath'' model, in which the environment to which the special degree(s) of freedom couple is
replaced by an effective set of harmonic oscillators. We will examine this model for the case of a single degree of freedom of
interest, which we will designate . For the case of the isomerizing or dissociating diatomic, could be the coordinate , where
is the distance between the atoms. The particular definition of ensures that . The degree of freedom is assumed to couple
to the bath linearly, giving a Hamiltonian of the form

where the index runs over all the bath degrees of freedom, are the harmonic bath frequencies, are the harmonic bath
masses, and are the coupling constants between the bath and the coordinate . is a momentum conjugate to , and is the mass
associated with this degree of freedom (e.g., the reduced mass in the case of a diatomic). The coordinate is assumed to be
subject to a potential as well (e.g., an internal bond potential). The form of the coupling between the system ( ) and the bath (
) is known as bilinear.
Below, using a completely classical treatment of this Hamiltonian, we will derive an equation for the detailed dynamics of alone.
This equation is known as the generalized Langevin equation (GLE).

This page titled 14.1: The Harmonic Bath Hamiltonian is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
Mark Tuckerman.

14.1.1 [Link]
14.2: The Random Force Term
Within the context of a harmonic bath, the term "random force'' is something of a misnomer, since is completely deterministic
and not random at all!!! We will return to this point momentarily, however, let us examine particular features of from its
explicit expression from the harmonic bath dynamics. Note, first of all, that it does not depend on the dynamics of the system
coordinate (except for the appearance of ). In this sense, it is independent or "orthogonal'' to within a phase space picture.
From the explicit form of , it is straightforward to see that the correlation function

i.e., the correlation function of the system velocity with the random force is 0. This can be seen by substituting in the expression
for and integrating over initial conditions with a canonical distribution weighting. For certain potentials that are even in
(such as a harmonic oscillator), one can also show that

Thus, is completely uncorrelated from both and , which is a property we might expect from a truly random process. In fact,
is determined by the detailed dynamics of the bath. However, we are not particularly interested or able to follow these detailed
dynamics for a large number of bath degrees of freedom. Thus, we could just as well model by a completely random process
(satisfying certain desirable features that are characteristic of a more general bath), and, in fact, this is often done. One could, for
example, postulate that act over a maximum time at discrete points in time , giving values of
, and assume that takes the form of a gaussian random process:

where the coefficients and are chosen at random from a gaussian distribution function. This might be expected to be
suitable for a bath of high density, where strong collisions between the system and a bath particle are essentially nonexistent, but
where the system only sees feels the relatively "soft'' fluctuations of the less mobile bath. For a low density bath, one might try
modeling as a Poisson process of very strong collisions.
Whatever model is chosen for , if it is a truly random process that can only act at discrete points in time, then the GLE takes the
form of a stochastic (based on random numbers) integro-differential equation. There is a whole body of mathematics devoted to the
properties of such equations, where heavy use of an calculus is made.

This page titled 14.2: The Random Force Term is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark
Tuckerman.

14.2.1 [Link]
14.3: The Dynamic Friction Kernel
The convolution integral term

is called the memory integral because it depends, in general, on the entire history of the evolution of . Physically it expresses the
fact that the bath requires a finite time to respond to any fluctuation in the motion of the system ( ). This, in turn, affects how the
bath acts back on the system. Thus, the force that the bath exerts on the system presently depends on what the system coordinate
did in the past. However, we have seen previously the regression of fluctuations (their decay to 0) over time. Thus, we expect that
what the system did very far in the past will no longer the force it feels presently, i.e., that the lower limit of the memory integral
(which is rigorously 0) could be replaced by , where is the maximum time over which memory of what the system
coordinate did in the past is important. This can be interpreted as a indicating a certain decay time for the friction kernel . In
fact, often does decay to 0 in a relatively short time. Often this decay takes the form of a rapid initial decay followed by a slow
final decay, as shown in the figure below:
Consider the extreme case that the bath is capable of responding infinitely quickly to changes in the system coordinate . This
would be the case, for example, if there were a large mass disparity between the system and the bath ( ). Then, the bath
retains no memory of what the system did in the past, and we could take to be a -function in time:

Then

and the GLE becomes

This simpler equation of motion is known as the Langevin equation and it is clearly a special case of the more generalized
equation of motion. It is often invoked to describe Brownian motion where clearly such a mass disparity is present. The constant
is known as the static friction and is given by

In fact, this is a general relation for determining the static friction constant.
The other extreme is a very sluggish bath that responds slowly to changes in the system coordinate. In this case, we may take to
be a constant , at least, for times short compared to the response time of the bath. Then, the memory integral becomes

and the GLE becomes

where the friction term now manifests itself as an extra harmonic term added to the potential. Such a term has the effect of trapping
the system in certain regions of configuration space, an effect known as dynamic caging. An example of this is a dilute mixture of
small, light particles in a bath of heavy, large particles. The light particles can get trapped in regions of space where many bath
particles are in a sort of spatial 'cage.'' Only the rare fluctuations in the bath that open up larger holes in configuration space allow
the light particles to escape the cage, occasionally, after which, they often get trapped again in a new cage for a similar time
interval.

14.3.1 [Link]
This page titled 14.3: The Dynamic Friction Kernel is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by
Mark E. Tuckerman.

14.3.2 [Link]
14.4: Relation between the Dynamic Friction Kernel and the Random Force
From the definitions of and , it is straightforward to show that there is a relation between them of the form

This relation is known as the second fluctuation dissipation theorem. The fact that it involves a simple autocorrelation function of
the random force is particular to the harmonic bath model. We will see later that a more general form of this relation exists, valid
for a general bath. This relation must be kept in mind when introducing models for and \zeta(t). In effect, it acts as a constraint
on the possible ways in which one can model the random force and friction kernel.

This page titled 14.4: Relation between the Dynamic Friction Kernel and the Random Force is shared under a CC BY-NC-SA 4.0 license and was
authored, remixed, and/or curated by Mark Tuckerman.

14.4.1 [Link]
14.5: Derivation of the GLE
The GLE can be derived from the harmonic bath Hamiltonian by simply solving Hamilton's equations of motion, which take the
form

This set of equations can also be written as second order differential equation:

In order to derive an equation for , we solve explicitly for the dynamics of the bath variables and then substitute into the equation
for . The equation for is a second order inhomogeneous differential equation, which can be solved by Laplace transforms. We
simply take the Laplace transform of both sides. Denote the Laplace transforms of and as

and recognizing that

we obtain the following equation for :

or

can be obtained by inverse Laplace transformation, which is equivalent to a contour integral in the complex -plane around a
contour that encloses all the poles of the integrand. This contour is known as the Bromwich contour. To see how this works,
consider the first term in the above expression. The inverse Laplace transform is

The integrand has two poles on the imaginary -axis at . Integration over the contour that encloses these poles picks up both
residues from these poles. Since the poles are simple poles, then, from the residue theorem:

By the same method, the second term will give . The last term is the inverse Laplace transform of a product of
and . From the convolution theorem of Laplace transforms, the Laplace transform of a convolution gives the product of
Laplace transforms:

14.5.1 [Link]
Thus, the last term will be the convolution of with . Putting these results together, gives, as the solution for :

The convolution term can be expressed in terms of rather than by integrating it by parts:

The reasons for preferring this form will be made clear shortly. The bath variables can now be seen to evolve according to

Substituting this into the equation of motion for , we find

We now introduce the following notation for the sums over bath modes appearing in this equation:

1.
Define a dynamic friction kernel

2.
Define a random force

Using these definitions, the equation of motion for reads

(1)

Eq. (1) is known as the generalized Langevin equation. Note that it takes the form of a one-dimensional particle subject to a
potential , driven by a forcing function and with a nonlocal (in time) damping term , which depends, in
general, on the entire history of the evolution of . The GLE is often taken as a phenomenological equation of motion for a
coordinate coupled to a general bath. In this spirit, it is worth taking a moment to discuss the physical meaning of the terms
appearing in the equation.

This page titled 14.5: Derivation of the GLE is shared under a CC BY-NC-SA 4.0 license and was authored, remixed, and/or curated by Mark
Tuckerman.

14.5.2 [Link]
14.6: Mori-Zwanzig Theory- A more general derivation of the GLE
A derivation of the GLE valid for a general bath can be worked out. The details of the derivation are given in the book by Berne
and Pecora called Dynamic Light Scattering. The system coordinate and its conjugate momentum are introduced as a column
vector:

and, in addition, one introduces statistical projection operators and that project onto subspaces in phase space parallel and
orthogonal to . These operators take the form

These operators are Hermitian and satisfy the property of idempotency:

Also, note that

The time evolution of is given by application of the classical propagator:

Note that the evolution of is unitary, i.e., it preserves the norm of :

Differentiating both sides of the time evolution equation for gives:

Then, an identity operator is inserted in the above expression in the form :

The first term in this expression defines a frequency matrix acting on :

where

14.6.1 [Link]
In order to evaluate the second term, another identity operator is inserted directly into the propagator:

Consider the difference between the two propagators:

If this difference is Laplace transformed, it becomes

which can be simplified via the general operator identity:

Letting

we have

or

Now, inverse Laplace transforming both sides gives

Thus, multiplying fromthe right by gives

Define a vector

so that

Because is completely orthogonal to , it is straightforward to show that

14.6.2 [Link]
Then,

Thus,

Finally, we define a memory kernel matrix:

Then, combining all results, we find, for :

which equivalent to a generalized Langevin equation for a particle subject to a harmonic potential, but coupled to a general bath.
For most systems, the quantities appearing in this form of the generalized Langevin equation are

It is easy to derive these expressions for the case of the harmonic bath Hamiltonian when

For the case of a harmonic bath Hamiltonian, we had shown that the friction kernel was related to the random force by the
fluctuation dissipation theorem:

For a general bath, the relation is not as simple, owing to the fact that is evolved using a modified propagator
. Thus, the more general form of the fluctuation dissipation theorem is

so that the dynamics of is prescribed by the propagator . This more general relation illustrates the
difficulty of defining a friction kernel for a general bath. However, for the special case of a stiff harmonic diatomic molecule

14.6.3 [Link]
interacting with a bath for which all the modes are soft compared to the frequency of the diatomic, a very useful approximation
results. One can show that

where is the Liouville operator for a system in which the diatomic is held rigidly fixed at some particular bond
length (i.e., a constrained dynamics). Since the friction kernel is not sensitive to the details of the internal potential of the diatomic,
this approximation can also be used for diatomics with stiff, anharmonic potentials. This approximation is referred to as the rigid
bond approximation (see Berne, et al, J. Chem. Phys. 93, 5084 (1990)).

This page titled 14.6: Mori-Zwanzig Theory- A more general derivation of the GLE is shared under a CC BY-NC-SA 4.0 license and was
authored, remixed, and/or curated by Mark Tuckerman.

14.6.4 [Link]
14.7: Example- Vibrational dephasing and energy relaxation
Recall that the Fourier transform of a time correlation function can be related to some kind of frequency spectrum. For example,
the Fourier transform of the velocity autocorrelation function of a particular degree of freedom of interest

where , gives the relevant frequencies contributing to the dynamics of , but does not give amplitudes. This "frequency''
spectrum is simply given by

That is, we take the Laplace transform of using . Since carries information about the relevant frequencies of the
system, the decay of in time is a measure of how strongly coupled the motion of is to the rest of the bath, i.e., how much
of an overlap there is between the relevant frequencies of the bath and those of . The more of an overlap there is, the more mixing
there will be between the system and the bath, and hence, the more rapidly the motion of the system will become vibrationally "out
of phase'' or decorrelated with itself. Thus, the decay time of , which is denoted is called the vibrational
dephasing time.
Another measure of the strength of the coupling between the system and the bath is the time required for the system to dissipate
energy into the bath when it is excited away from equilibrium. This time can be obtained by studying the decay of the energy
autocorrelation function:

where is defined to be

The decay time of this correlation function is denoted .

The question then becomes: what are these characteristic decay times and how are they related? To answer this, we will take a
phenomenological approach. We will assume the validity of the GLE for :

and use it to calculate and .

Suppose the potential is harmonic and takes the form

Substituting into the GLE and dividing through by gives

where

An equation of motion for can be obtained directly by multiplying both sides of the GLE by
and averaging over a canonical ensemble:

Recall that

14.7.1 [Link]
and note that

also

Thus,

Combining these results gives an equation for

which is known as the memory function equation and the kernel is known as the memory function
or memory kernel. This type of integro-differential equation is called a Volterra equation and it can be solved by Laplace
transforms.
Taking the Laplace transform of both sides gives

However, it is clear that and also

Thus, it follows that

In order to perform the inverse Laplace transform, we need the poles of the integrand, which will be determined by the solutions of

which we could solve directly if we knew the explicit form of .

However, if is sufficiently larger than , then it is possible to develop a perturbation solution to this
equation. Let us assume the solutions for can be written as
Substituting in this ansatz gives

Since we are assuming is small, then to lowest order, we have

so that . The first order equation then becomes

or

14.7.2 [Link]
Note, however, that

Thus, stopping the first order result, the poles of the integrand occur at

Define

Then

and is then given by the contour integral

Taking the residue at each pole, we find

which can be simplified to give

Thus, we see that the GLE predicts oscillates with a frequency and decays exponentially. From the
exponential decay, we can directly read off the time :

That is, the value of the real part of the Fourier (Laplace) transform of the friction kernel evaluated at the renormalized frequency
divided by gives the vibrational dephasing time! By a similar scheme, one can easily show that the position
autocorrelation function decays with the same dephasing time. It's explicit form is

The energy autocorrelation function can be expressed in terms of the more primitive correlation functions
and . It is a straightforward, although extremely tedious, matter to show that the relation, valid for the
harmonic potential of mean force, is

Substituting in the expressions for and gives

so that the decay time can be seen to be

14.7.3 [Link]
and therefore, the relation between and can be seen immediately to be

The incredible fact is that this result is also true quantum mechanically. That is, by doing a simple, purely classical treatment of the
problem, we obtained a result that turns out to be the correct quantum mechanical result!
Just how big are these times? If is very large compared to any typical frequency relevant to the bath, then the friction kernel
evaluated at this frequency will be extremely small, giving rise to a long decay time. This result is expect, since, if is large
compared to the bath, there are very few ways in which the system can dissipate energy into the bath. The situation changes
dramatically, however, if a small amount of anharmonicity is added to the potential of mean force. The figure below illustrates the
point for a harmonic diatomic molecule interacting with a Lennard-Jones bath. The top figure shows the velocity autocorrelation
function for an oscillator whose frequency is approximately 3 times the characteristic frequency of the bath, while the bottom one
shows the velocity autocorrelation function for the case that the frequency disparity is a factor of 6.

Figure 1:

This page titled 14.7: Example- Vibrational dephasing and energy relaxation is shared under a CC BY-NC-SA 4.0 license and was authored,
remixed, and/or curated by Mark Tuckerman.

14.7.4 [Link]
Index
A free energy hypersurface Liouville equation
action principle 8.4: Reaction Coordinates 2.3: The Liouville Operator and the Poisson Bracket
1.1: The Lagrangian Formulation of Classical fugacity 12.5: Perturbative solution of the Liouville equation
Mechanics 6.2: Partition Functions Liouville operator
anticommutator functional integral 2.3: The Liouville Operator and the Poisson Bracket
13.1.2: The Transition Rate 11.1.4: The Continuous Limit Liouville's Theorem
1.6: Phase space distribution functions and
Liouville's theorem
C G 2.2: Liouville's Theorem for non-Hamiltonian
caging general correlation function systems
14.3: The Dynamic Friction Kernel 7.4: General Correlation Functions 4.4: Preservation of Phase Space Volume and
generalized equations of motion Liouville's Theorem
canonical partition function
4.3: Relation between Canonical and Microcanonical 12.3: Generalized Equations of Motion
Ensembles Generalized Langevin Equation M
Classical Virial Theorem 14.3: The Dynamic Friction Kernel
Maslov index
3.3: The Classical Virial Theorem (Microcanonical generalized Liouville equation 11.3.2: The Stationary Phase Approximation
Derivation)
2.2: Liouville's Theorem for non-Hamiltonian
memory integral
closure systems 14.3: The Dynamic Friction Kernel
9.1: Measurement microcanonical ensemble
collective variables H 3: The Microcanonical Ensemble
8.4: Reaction Coordinates Hamiltonian operator Microscopic Laws of Motion
configurational partition function 13.1.1: The Hamiltonian 1.3: The Microscopic Laws of Motion
7.2: General Distribution Functions and Correlation Harmonic bath microscopic state
Functions
14.1: The Harmonic Bath Hamiltonian 1.5: Classical microscopic states or microstates and
confocal elliptic coordinates ensembles
Heisenberg picture
8.4: Reaction Coordinates
9.4: The Heisenberg Picture
microstate
conjugate momenta 1.5: Classical microscopic states or microstates and
Heisenberg uncertainty principle
1.2: The Hamiltonian formulation of classical ensembles
9.5: The Heisenberg Uncertainty Principle
mechanics
Hemlholtz energy O
D 3.1: Basic Thermodynamics
Onsager fluctuation regression theorem
Density matrix Hermitian operators
12.6: The Onsager Fluctuation Regression Theorem
9.1: Measurement
10.2: The Density Matrix and Density Operator Onsager regression hypothesis
density operator Hilbert space
12.4: General Properties of Time Correlation
9.6: The Physical State of a Quantum System Functions
10.2: The Density Matrix and Density Operator
diffusion constant order parameters
12.8.2: The Diffusion Constant
I 8.4: Reaction Coordinates
distribution functions imaginary time path integral
7.1: General Formulation of Distribution Functions 11.1.4: The Continuous Limit P
interaction picture pair correlation function
E 13.2: Iterative solution for the interaction-picture 7.6: The Pair Correlation Function
state vector
ensemble 13.3: The Interaction Picture partition functions
1.5: Classical microscopic states or microstates and 6.2: Partition Functions
ensembles
J path integrals
2.1: The Ensemble Concept (Heuristic Definition)
11.1: Discretized and Continuous Path Integrals
equal a priori probabilities Jarzynski's equality
8.5: Jarzynski's Equality and Nonequilibrium
phase space
3: The Microcanonical Ensemble
Methods 1.4: Phase Space
equations of motion phase space distribution function
1.1: The Lagrangian Formulation of Classical
Mechanics
K 1.6: Phase space distribution functions and
Liouville's theorem
equilibrium time correlation function Kubo transform
12.2: Kubo Transform Expression for the Time
phase space probability density
12.5: Perturbative solution of the Liouville equation
Correlation Function 1.6: Phase space distribution functions and
Euclidean action Liouville's theorem
11.1.3: Dominant Paths in the Propagator and
L Poisson bracket
Density Matrix
2.3: The Liouville Operator and the Poisson Bracket
Euler's theorem Lagrangian
pressure tensor
6.1: Thermodynamics 1.2: The Hamiltonian formulation of classical
mechanics 12.8.1: Shear Viscosity
expectation value principle of least action
11.2.1: Expectation values of observables
Langevin equation
14.3: The Dynamic Friction Kernel 11.1.3: Dominant Paths in the Propagator and
Density Matrix
F Legendre Transforms
4.2: Legendre Transforms
Fermi's Golden Rule
13.4: Fermi's Golden Rule
Q stationary phase approximation transition rate
quantum Liouville equation 11.3.2: The Stationary Phase Approximation 13.1.2: The Transition Rate
13.1.2: The Transition Rate superoperator transport coefficient
quantum propagator 10.3: Time evolution of the density operator 12.8.1: Shear Viscosity
9.7: Time Evolution of the State Vector switching functions
8.3: Adiabatic Switching and Thermodynamic V
Integration van der Waals equation
R
reaction coordinate 7.7: Derivation of the Van der Waals equation
T Velocity autocorrelation function
8.4: Reaction Coordinates
tetradic operator 12.8.2: The Diffusion Constant
S 10.3: Time evolution of the density operator virial equation of state
thermal de Broglie wavelength
saddle point approximation 7.5: Thermodynamic quantities in terms of g(r)
11.3.2: The Stationary Phase Approximation
11.1.2: Doing the Path Integral - the Free Particle virial theorem
thermodynamic limit 4.1: Classical Virial Theorem (Canonical Ensemble
Schrödinger picture
4.5: Energy Fluctuations in the Canonical Ensemble Derivation)
9.4: The Heisenberg Picture
time evolution operator
shear force
9.7: Time Evolution of the State Vector W
12.8.1: Shear Viscosity
trajectory Wick rotation
shear viscosity
1.1: The Lagrangian Formulation of Classical 11.1.4: The Continuous Limit
12.8.1: Shear Viscosity
Mechanics work virial theorem
solvent cage
5.2: Pressure and Work Virial Theorems
14.3: The Dynamic Friction Kernel
Glossary
Sample Word 1 | Sample Definition 1
Detailed Licensing
Overview
Title: Advanced Statistical Mechanics (Tuckerman)
Webpages: 116
Applicable Restrictions: Noncommercial
All licenses found:
CC BY-NC-SA 4.0: 83.6% (97 pages)
Undeclared: 16.4% (19 pages)

By Page
Advanced Statistical Mechanics (Tuckerman) - CC BY-NC- 4.1: Classical Virial Theorem (Canonical Ensemble
SA 4.0 Derivation) - CC BY-NC-SA 4.0
Front Matter - Undeclared 4.2: Legendre Transforms - CC BY-NC-SA 4.0
TitlePage - Undeclared 4.3: Relation between Canonical and Microcanonical
InfoPage - Undeclared Ensembles - CC BY-NC-SA 4.0
Table of Contents - Undeclared 4.4: Preservation of Phase Space Volume and
Liouville's Theorem - CC BY-NC-SA 4.0
Licensing - Undeclared
4.5: Energy Fluctuations in the Canonical Ensemble -
1: Classical mechanics - Undeclared
CC BY-NC-SA 4.0
1.1: The Lagrangian Formulation of Classical 5: The Isothermal-Isobaric Ensemble - Undeclared
Mechanics - CC BY-NC-SA 4.0
5.1: Basic Thermodynamics - CC BY-NC-SA 4.0
1.2: The Hamiltonian formulation of classical
5.2: Pressure and Work Virial Theorems - CC BY-NC-
mechanics - CC BY-NC-SA 4.0
SA 4.0
1.3: The Microscopic Laws of Motion - CC BY-NC-
5.3: The partition function and relation to
SA 4.0
thermodynamics - CC BY-NC-SA 4.0
1.4: Phase Space - CC BY-NC-SA 4.0
5.4: Temperature and pressure estimators - CC BY-
1.5: Classical microscopic states or microstates and
NC-SA 4.0
ensembles - CC BY-NC-SA 4.0
1.6: Phase space distribution functions and Liouville's 6: The Grand Canonical Ensemble - CC BY-NC-SA 4.0
theorem - CC BY-NC-SA 4.0 6.1: Thermodynamics - CC BY-NC-SA 4.0
2: Foundations of classical statistical mechanics - 6.2: Partition Functions - CC BY-NC-SA 4.0
Undeclared 6.3: Ideal Gas - CC BY-NC-SA 4.0
6.4: Particle Number Fluctuations - CC BY-NC-SA 4.0
2.1: The Ensemble Concept (Heuristic Definition) -
CC BY-NC-SA 4.0 7: Distribution Functions and Liquid Structure -
2.2: Liouville's Theorem for non-Hamiltonian Undeclared
systems - CC BY-NC-SA 4.0 7.1: General Formulation of Distribution Functions -
2.3: The Liouville Operator and the Poisson Bracket - CC BY-NC-SA 4.0
CC BY-NC-SA 4.0 7.2: General Distribution Functions and Correlation
2.4: Equilibrium ensembles - CC BY-NC-SA 4.0 Functions - CC BY-NC-SA 4.0
7.3: Structure and distribution functions in classical
3: The Microcanonical Ensemble - CC BY-NC-SA 4.0
liquids and gases - CC BY-NC-SA 4.0
3.1: Basic Thermodynamics - CC BY-NC-SA 4.0
7.4: General Correlation Functions - CC BY-NC-SA
3.2: The Partition Function - CC BY-NC-SA 4.0
4.0
3.3: The Classical Virial Theorem (Microcanonical
7.5: Thermodynamic quantities in terms of g(r) - CC
Derivation) - CC BY-NC-SA 4.0
BY-NC-SA 4.0
4: The Canonical Ensemble - Undeclared

1 [Link]
 7.6: The Pair Correlation Function - CC BY-NC-SA 11.1: Discretized and Continuous Path Integrals - CC
4.0 BY-NC-SA 4.0
7.7: Derivation of the Van der Waals equation - CC 11.1.1: Derivation of the Discretized Path Integral
BY-NC-SA 4.0 - CC BY-NC-SA 4.0
8: Rare-event sampling and free energy calculations - 11.1.2: Doing the Path Integral - the Free Particle
Undeclared - CC BY-NC-SA 4.0
8.2: Free-energy Perturbation Theory - CC BY-NC-SA 11.1.3: Dominant Paths in the Propagator and
4.0 Density Matrix - CC BY-NC-SA 4.0
8.3: Adiabatic Switching and Thermodynamic 11.1.4: The Continuous Limit - CC BY-NC-SA 4.0
Integration - CC BY-NC-SA 4.0 11.2: Calculation of observables from path integrals -
8.4: Reaction Coordinates - CC BY-NC-SA 4.0 CC BY-NC-SA 4.0
8.5: Jarzynski's Equality and Nonequilibrium
11.2.1: Expectation values of observables - CC
Methods - CC BY-NC-SA 4.0 BY-NC-SA 4.0
8.6: The "blue moon'' Ensemble Approach - CC BY- 11.2.2: Path integrals for N-particle systems - CC
NC-SA 4.0 BY-NC-SA 4.0
9: Review of the basic postulates of quantum mechanics 11.2.3: Path integral molecular dynamics (optional
- CC BY-NC-SA 4.0 reading) - CC BY-NC-SA 4.0
9.1: Measurement - CC BY-NC-SA 4.0 11.2.4: Thermodynamics from path integrals - CC
9.2: Physical Observables - CC BY-NC-SA 4.0 BY-NC-SA 4.0
9.3: The Fundamental Postulates of Quantum 11.3: Expansion about the classical path and
Mechanics - CC BY-NC-SA 4.0 stationary phase - CC BY-NC-SA 4.0
9.4: The Heisenberg Picture - CC BY-NC-SA 4.0 11.3.1: The harmonic Oscillator - Expansion about
9.5: The Heisenberg Uncertainty Principle - CC BY- the Classical Path - CC BY-NC-SA 4.0
NC-SA 4.0
11.3.2: The Stationary Phase Approximation - CC
9.6: The Physical State of a Quantum System - CC
BY-NC-SA 4.0
BY-NC-SA 4.0
12: Time-dependent Processes - Classical case -
9.7: Time Evolution of the State Vector - CC BY-NC-
Undeclared
SA 4.0
12.1: Perturbative solution of the Liouville equation -
10: Fundamentals of quantum statistical mechanics - CC
CC BY-NC-SA 4.0
BY-NC-SA 4.0 12.2: Kubo Transform Expression for the Time
10.1: Principles of quantum statistical mechanics - Correlation Function - CC BY-NC-SA 4.0
CC BY-NC-SA 4.0 12.3: Generalized Equations of Motion - CC BY-NC-
10.2: The Density Matrix and Density Operator - CC SA 4.0
BY-NC-SA 4.0 12.4: General Properties of Time Correlation
10.3: Time evolution of the density operator - CC BY- Functions - CC BY-NC-SA 4.0
NC-SA 4.0 12.5: Perturbative solution of the Liouville equation -
10.4: A simple example - the quantum harmonic CC BY-NC-SA 4.0
oscillator - CC BY-NC-SA 4.0 12.6: The Onsager Fluctuation Regression Theorem -
10.5: The quantum equilibrium ensembles - CC BY- CC BY-NC-SA 4.0
NC-SA 4.0 12.7: Relation to Spectra - CC BY-NC-SA 4.0
10.5.1: The microcanonical ensemble - CC BY- 12.8: Time Correlation Functions and Transport
NC-SA 4.0 Coefficients - CC BY-NC-SA 4.0
10.5.2: The canonical ensemble - CC BY-NC-SA 12.8.1: Shear Viscosity - CC BY-NC-SA 4.0
4.0 12.8.2: The Diffusion Constant - CC BY-NC-SA
10.5.3: Isothermal-isobaric and grand canonical 4.0
ensembles - CC BY-NC-SA 4.0 13: Time-dependent Processes - Quantum Case -
11: Introduction to path integrals in quantum mechanics Undeclared
and quantum statistical mechanics - Undeclared

2 [Link]
 13.1: Calculation of spectra from perturbation theory 14.2: The Random Force Term - CC BY-NC-SA 4.0
- CC BY-NC-SA 4.0 14.3: The Dynamic Friction Kernel - CC BY-NC-SA
13.1.1: The Hamiltonian - CC BY-NC-SA 4.0 4.0
13.1.2: The Transition Rate - CC BY-NC-SA 4.0 14.4: Relation between the Dynamic Friction Kernel
13.1.3: Examples - CC BY-NC-SA 4.0 and the Random Force - CC BY-NC-SA 4.0
13.2: Iterative solution for the interaction-picture 14.5: Derivation of the GLE - CC BY-NC-SA 4.0
14.6: Mori-Zwanzig Theory- A more general
state vector - CC BY-NC-SA 4.0
derivation of the GLE - CC BY-NC-SA 4.0
13.3: The Interaction Picture - CC BY-NC-SA 4.0
14.7: Example- Vibrational dephasing and energy
13.4: Fermi's Golden Rule - CC BY-NC-SA 4.0
relaxation - CC BY-NC-SA 4.0
13.5: Quantum Linear Response Theory - CC BY-NC-
SA 4.0 Back Matter - Undeclared
14: The Langevin and Generalized Langevin equations - Index - Undeclared
CC BY-NC-SA 4.0 Glossary - Undeclared
14.1: The Harmonic Bath Hamiltonian - CC BY-NC- Detailed Licensing - Undeclared
SA 4.0 Detailed Licensing - Undeclared

3 [Link]
Detailed Licensing
Overview
Title: Advanced Statistical Mechanics (Tuckerman)
Webpages: 116
Applicable Restrictions: Noncommercial
All licenses found:
CC BY-NC-SA 4.0: 83.6% (97 pages)
Undeclared: 16.4% (19 pages)

By Page
Advanced Statistical Mechanics (Tuckerman) - CC BY-NC- 4.1: Classical Virial Theorem (Canonical Ensemble
SA 4.0 Derivation) - CC BY-NC-SA 4.0
Front Matter - Undeclared 4.2: Legendre Transforms - CC BY-NC-SA 4.0
TitlePage - Undeclared 4.3: Relation between Canonical and Microcanonical
InfoPage - Undeclared Ensembles - CC BY-NC-SA 4.0
Table of Contents - Undeclared 4.4: Preservation of Phase Space Volume and
Liouville's Theorem - CC BY-NC-SA 4.0
Licensing - Undeclared
4.5: Energy Fluctuations in the Canonical Ensemble -
1: Classical mechanics - Undeclared
CC BY-NC-SA 4.0
1.1: The Lagrangian Formulation of Classical 5: The Isothermal-Isobaric Ensemble - Undeclared
Mechanics - CC BY-NC-SA 4.0
5.1: Basic Thermodynamics - CC BY-NC-SA 4.0
1.2: The Hamiltonian formulation of classical
5.2: Pressure and Work Virial Theorems - CC BY-NC-
mechanics - CC BY-NC-SA 4.0
SA 4.0
1.3: The Microscopic Laws of Motion - CC BY-NC-
5.3: The partition function and relation to
SA 4.0
thermodynamics - CC BY-NC-SA 4.0
1.4: Phase Space - CC BY-NC-SA 4.0
5.4: Temperature and pressure estimators - CC BY-
1.5: Classical microscopic states or microstates and
NC-SA 4.0
ensembles - CC BY-NC-SA 4.0
1.6: Phase space distribution functions and Liouville's 6: The Grand Canonical Ensemble - CC BY-NC-SA 4.0
theorem - CC BY-NC-SA 4.0 6.1: Thermodynamics - CC BY-NC-SA 4.0
2: Foundations of classical statistical mechanics - 6.2: Partition Functions - CC BY-NC-SA 4.0
Undeclared 6.3: Ideal Gas - CC BY-NC-SA 4.0
6.4: Particle Number Fluctuations - CC BY-NC-SA 4.0
2.1: The Ensemble Concept (Heuristic Definition) -
CC BY-NC-SA 4.0 7: Distribution Functions and Liquid Structure -
2.2: Liouville's Theorem for non-Hamiltonian Undeclared
systems - CC BY-NC-SA 4.0 7.1: General Formulation of Distribution Functions -
2.3: The Liouville Operator and the Poisson Bracket - CC BY-NC-SA 4.0
CC BY-NC-SA 4.0 7.2: General Distribution Functions and Correlation
2.4: Equilibrium ensembles - CC BY-NC-SA 4.0 Functions - CC BY-NC-SA 4.0
7.3: Structure and distribution functions in classical
3: The Microcanonical Ensemble - CC BY-NC-SA 4.0
liquids and gases - CC BY-NC-SA 4.0
3.1: Basic Thermodynamics - CC BY-NC-SA 4.0
7.4: General Correlation Functions - CC BY-NC-SA
3.2: The Partition Function - CC BY-NC-SA 4.0
4.0
3.3: The Classical Virial Theorem (Microcanonical
7.5: Thermodynamic quantities in terms of g(r) - CC
Derivation) - CC BY-NC-SA 4.0
BY-NC-SA 4.0
4: The Canonical Ensemble - Undeclared

1 [Link]
 7.6: The Pair Correlation Function - CC BY-NC-SA 11.1: Discretized and Continuous Path Integrals - CC
4.0 BY-NC-SA 4.0
7.7: Derivation of the Van der Waals equation - CC 11.1.1: Derivation of the Discretized Path Integral
BY-NC-SA 4.0 - CC BY-NC-SA 4.0
8: Rare-event sampling and free energy calculations - 11.1.2: Doing the Path Integral - the Free Particle
Undeclared - CC BY-NC-SA 4.0
8.2: Free-energy Perturbation Theory - CC BY-NC-SA 11.1.3: Dominant Paths in the Propagator and
4.0 Density Matrix - CC BY-NC-SA 4.0
8.3: Adiabatic Switching and Thermodynamic 11.1.4: The Continuous Limit - CC BY-NC-SA 4.0
Integration - CC BY-NC-SA 4.0 11.2: Calculation of observables from path integrals -
8.4: Reaction Coordinates - CC BY-NC-SA 4.0 CC BY-NC-SA 4.0
8.5: Jarzynski's Equality and Nonequilibrium
11.2.1: Expectation values of observables - CC
Methods - CC BY-NC-SA 4.0 BY-NC-SA 4.0
8.6: The "blue moon'' Ensemble Approach - CC BY- 11.2.2: Path integrals for N-particle systems - CC
NC-SA 4.0 BY-NC-SA 4.0
9: Review of the basic postulates of quantum mechanics 11.2.3: Path integral molecular dynamics (optional
- CC BY-NC-SA 4.0 reading) - CC BY-NC-SA 4.0
9.1: Measurement - CC BY-NC-SA 4.0 11.2.4: Thermodynamics from path integrals - CC
9.2: Physical Observables - CC BY-NC-SA 4.0 BY-NC-SA 4.0
9.3: The Fundamental Postulates of Quantum 11.3: Expansion about the classical path and
Mechanics - CC BY-NC-SA 4.0 stationary phase - CC BY-NC-SA 4.0
9.4: The Heisenberg Picture - CC BY-NC-SA 4.0 11.3.1: The harmonic Oscillator - Expansion about
9.5: The Heisenberg Uncertainty Principle - CC BY- the Classical Path - CC BY-NC-SA 4.0
NC-SA 4.0
11.3.2: The Stationary Phase Approximation - CC
9.6: The Physical State of a Quantum System - CC
BY-NC-SA 4.0
BY-NC-SA 4.0
12: Time-dependent Processes - Classical case -
9.7: Time Evolution of the State Vector - CC BY-NC-
Undeclared
SA 4.0
12.1: Perturbative solution of the Liouville equation -
10: Fundamentals of quantum statistical mechanics - CC
CC BY-NC-SA 4.0
BY-NC-SA 4.0 12.2: Kubo Transform Expression for the Time
10.1: Principles of quantum statistical mechanics - Correlation Function - CC BY-NC-SA 4.0
CC BY-NC-SA 4.0 12.3: Generalized Equations of Motion - CC BY-NC-
10.2: The Density Matrix and Density Operator - CC SA 4.0
BY-NC-SA 4.0 12.4: General Properties of Time Correlation
10.3: Time evolution of the density operator - CC BY- Functions - CC BY-NC-SA 4.0
NC-SA 4.0 12.5: Perturbative solution of the Liouville equation -
10.4: A simple example - the quantum harmonic CC BY-NC-SA 4.0
oscillator - CC BY-NC-SA 4.0 12.6: The Onsager Fluctuation Regression Theorem -
10.5: The quantum equilibrium ensembles - CC BY- CC BY-NC-SA 4.0
NC-SA 4.0 12.7: Relation to Spectra - CC BY-NC-SA 4.0
10.5.1: The microcanonical ensemble - CC BY- 12.8: Time Correlation Functions and Transport
NC-SA 4.0 Coefficients - CC BY-NC-SA 4.0
10.5.2: The canonical ensemble - CC BY-NC-SA 12.8.1: Shear Viscosity - CC BY-NC-SA 4.0
4.0 12.8.2: The Diffusion Constant - CC BY-NC-SA
10.5.3: Isothermal-isobaric and grand canonical 4.0
ensembles - CC BY-NC-SA 4.0 13: Time-dependent Processes - Quantum Case -
11: Introduction to path integrals in quantum mechanics Undeclared
and quantum statistical mechanics - Undeclared

2 [Link]
 13.1: Calculation of spectra from perturbation theory 14.2: The Random Force Term - CC BY-NC-SA 4.0
- CC BY-NC-SA 4.0 14.3: The Dynamic Friction Kernel - CC BY-NC-SA
13.1.1: The Hamiltonian - CC BY-NC-SA 4.0 4.0
13.1.2: The Transition Rate - CC BY-NC-SA 4.0 14.4: Relation between the Dynamic Friction Kernel
13.1.3: Examples - CC BY-NC-SA 4.0 and the Random Force - CC BY-NC-SA 4.0
13.2: Iterative solution for the interaction-picture 14.5: Derivation of the GLE - CC BY-NC-SA 4.0
14.6: Mori-Zwanzig Theory- A more general
state vector - CC BY-NC-SA 4.0
derivation of the GLE - CC BY-NC-SA 4.0
13.3: The Interaction Picture - CC BY-NC-SA 4.0
14.7: Example- Vibrational dephasing and energy
13.4: Fermi's Golden Rule - CC BY-NC-SA 4.0
relaxation - CC BY-NC-SA 4.0
13.5: Quantum Linear Response Theory - CC BY-NC-
SA 4.0 Back Matter - Undeclared
14: The Langevin and Generalized Langevin equations - Index - Undeclared
CC BY-NC-SA 4.0 Glossary - Undeclared
14.1: The Harmonic Bath Hamiltonian - CC BY-NC- Detailed Licensing - Undeclared
SA 4.0 Detailed Licensing - Undeclared

3 [Link]