Farooq A. Khan and John E. Hansen
Farooq A. Khan and John E. Hansen
Farooq A. Khan and John E. Hansen
The Chemical Educator, Vol. 5, No. 3, S1430-4171(00)03379-9, 10.1007/s00897000379a, © 2000 Springer-Verlag New York, Inc.
The Dirac (Bracket) Notation in the Undergraduate Physical Chemistry Curriculum: A
Pictorial Introduction
Farooq A. Khan* and John E. Hansen
Abstract: The Dirac (bracket) notation is ubiquitous in the chemical literature, but it is rarely
introduced in the undergraduate chemistry curriculum. In this article we present a pictorial
approach to the bracket notation that we have successfully used for the past three years in a junior–
senior-level physical chemistry course. We have found that it requires roughly 75 minutes to
introduce this topic, and, upon integration into subsequent discussions, it prepares our
undergraduate students to routinely use this powerful tool in the study of chemical bonding and
spectroscopy. We believe that this approach, when introduced after the traditional integral
treatment, enhances student learning of the abstract subject of quantum mechanics.
Introduction
In a study of atomic structure, chemical bonding, and atomic and molecular spectroscopy, expressions such as eq 1
are encountered frequently. Manipulations of such expressions
by regarding the action of the operator on the ket as a vector and the as scalar
product of the two vectors.
In our experience, which spans three years, the use of this notation has served as a powerful tool to enhance student
learning in the abstract and often dreaded subject of elementary quantum mechanics. It also provides the student with
adequate preparation to start browsing through journal articles or more advanced texts without experiencing the
intimidation that most students feel when first encountering the ubiquitous bracket notation.
As we developed our pictorial approach, we did consider the possibility of adopting the presentation from any one
of a number of excellent physics textbooks available on the topic of bracket notation [2]. Most of these texts assume a
certain degree of mathematical maturity and may also require some familiarity with linear algebra. Not all chemistry
majors have the time to adopt the formal route to the bracket notation that physics majors typically use through a
sequence that includes mathematical methods and quantum mechanics. Thus, with a limited amount of time (typically
one semester) available to teach elementary quantum mechanics with applications to bonding and molecular
spectroscopy, we deemed it necessary to develop a very concise introduction.
In this article we present our approach in essentially the form that we have used in the classroom, along with a brief
discussion of some selected illustrative examples. In addition, we present a preliminary account of the learning
outcomes of this approach.
In the physical chemistry course that covers introductory quantum chemistry, we first present a historical
perspective and the detailed solution of Schrödinger's equation for the particle-in-a-box using the traditional integral
approach, which includes the explicit calculation of expectation values. We then introduce operators, matrices, and the
bracket notation, using the wavefunctions from the particle-in-a-box problem as illustrative examples. This is
followed by a presentation of atomic and molecular structure, and rotational, vibrational, and electronic spectroscopy.
The bracket notation is used extensively, particularly in the discussion of the harmonic oscillator and chemical
bonding. Our approach ensures that the students are adequately exposed to the use of appropriate integrals in the
process and are expected to evaluate them, for example, in calculations of expectation values. The power of the
bracket approach is that it recasts the integral approach in a format less formidable to students because it draws upon
their familiarity with vectors from general physics. It also ensures to a considerable degree that they do not lose sight
of the underlying physical principles, particularly when they spend considerable time and effort evaluating integrals
explicitly. In our opinion, both of these approaches (the integral and bracket) are pedagogically essential in an
introductory course.
Before we proceed further, we parenthetically note that because some of our students have little prior knowledge of
matrices, we introduce the concepts of matrix addition, multiplication, and inversion, as well as taking the adjoint of a
matrix as a brief digression prior to introducing the bracket notation. In the following paragraphs, we have,
essentially, reproduced our class notes, minus the review of matrices, refined over a three-year period. The entire
discussion including a review of matrices, the notation itself, and its properties in simple manipulations takes roughly
75 minutes.
The states of a quantum mechanical system are represented by ket vectors or simply kets, denoted as . For
example, ket is represented as . The ket vector belongs to an abstract space, referred to as the state space.
Furthermore, any quantum mechanical state can be expressed as a linear combination of the elements of the state
space. The properties of ket vectors can be introduced pictorially with the aid of Figure 1 by using the analogies with
vectors used in classical mechanics.
While presenting the simple pictorial introduction given above, we carefully make the following distinctions
between ordinary vectors and quantum mechanical states.
a) The two-state system in Figure 1 is chosen as a convenient, easily understandable starting point. (We note that
such a system is of particular interest in chemistry, for example, in NMR spectroscopy). In practice, the number
of orthogonal basis vectors in quantum mechanics may be large, even infinite.
b) In classical mechanics the coefficients and are real; however, in quantum mechanics this restriction is not
necessary and and are, in general, complex numbers.
Manipulations of Ket and Bra Vectors: A Few Rules
Many of the operations involving bras and kets introduced below are analogous to operations with vectors in
Euclidean space, with which students have considerable familiarity from general physics.
Manipulations using the bracket notation are governed by the following rules.
b) Ket vectors can be represented as column matrices, while bra vectors can be represented as row matrices.
c) For a ket vector , the corresponding bra vector is given by , where and
and
This is analogous to scalar products of unit vectors in classical mechanics,
and
e) The scalar product of two ket vectors and is given by .
f)
Illustrative Examples
Properties of Hermitean Operators. In this exercise, we illustrate that the eigenvalues of a Hermitean operator are
real, which is a key property of Hermitean operators.
chemical bonding. It also addresses questions that often arise in the students' minds during a
first course in quantum chemistry, "are there real applications of quantum mechanics?" or
"what is this material really good for?"
Following the treatment by Atkins, a standard text in Physical Chemistry [1], we construct the hybrid orbitals using
the and orbitals on each carbon atom
We can rewrite these in the bracket notation as
In this section we derive the gross and specific selection rules for vibrational spectroscopy with the aid of the
bracket notation. In addition, we determine the mean value for the potential energy of the harmonic oscillator for a
given state. The derivations contained in this section can also be arrived by explicit integration, using the properties of
Hermite polynomials, as illustrated in the appendix. Based on our experience with both methods, we have found the
bracket approach to be more understandable for the average student.
Gross Selection Rule. The wavefunctions for a harmonic oscillator can be represented as , where the quantum
number m is 0 or a positive integer. Consider a transition between vibrational states and with the same
electronic state. The transition dipole moment is given by , where is the dipole moment of the molecule
for that electronic state. The dipole moment is a function of the displacement x, as shown in eq 13
Therefore,
15)
In order that the right hand side of eq 15 is nonzero, must be nonzero. Thus, we obtain the gross selection
rule for vibrational spectroscopy: the dipole moment of the molecule must change with spatial displacement for a
vibrational mode to be active.
Specific Selection Rule. In order to obtain the specific selection rule, , we use the creation and
annihilation operators as illustrated below. The treatment is essentially that presented by Cohen-Tannoudji et al [2]
and Atkins [3].
The dimensionless operators and (upper case) are defined in terms of the position (x) and momentum (p)
operators,
(33)
Using an analogous process,
We have taught the introductory quantum chemistry course five times, without (1995, 1996) and with (1997, 1998,
1999) the bracket notation. In the past three years we have used the bracket notation with considerable success. We
believe that there are two attributes that contribute to its success. First, it lends itself to pictorial representation.
Second, it is very compact and intuitive when compared with the traditional integral formulation alone. For example,
the illustration that the eigenvalues of a Hermitean operator are real is very nonintuitive with integrals, but easily
handled by students with the bracket notation. In addition, students' comments reveal that the bracket notation enabled
them to visualize orthonormality, perhaps the single most important concept in introductory quantum mechanics,
much better than they would in the absence of the bracket approach. Feedback from students who have proceeded to
graduate school is that the two class periods used for introducing creation and annihilation operators were well spent
because they gave them a good foundation for further study in quantum mechanics at the graduate level.
When examining the usefulness of a new pedagogical method, it is customary to measure learning outcomes
through some external yardstick. We have measured the learning outcomes using the quantum chemistry examination
prepared by the American Chemical Society. The average results were 12 percentile points higher in the past three
years (58th percentile, compared to national norms) as compared to the preceding two years (46th percentile).
Admittedly, the sample size, (10 students in 1995 and 1996 combined and 29 students in 1997 through 1999
combined) is small; however, we believe that the increased scores are significant. We note, however, that the driving
force for the continued use of this method is our belief in the pedagogical soundness of this approach, as well as the
positive feedback that we have received from our former students in graduate school, as discussed above.
Many of our students have moved on from this course to undergraduate research projects using the software
package, Spartan, which makes extensive, albeit implicit, use of orthonormality. They carry out energy-minimization
calculations for small as well as large organic molecules that are of interest from a synthetic perspective. Students'
familiarity with the bracket notation allows for a greater intuitive understanding of orthonormality as they browse
through the Spartan manual, thus, it demystifies the "black box" nature of a prepackaged software program.
Conclusions
We have presented a pictorial approach that measurably enhances student learning of the abstract subject of
quantum mechanics. We have included several illustrative examples of the bracket notation that can be used alongside
the traditional presentation that uses integrals. Students are appreciative of this pictorial approach, which eases their
first journey into the rather elusive world of quantum mechanics.
Appendix
In this appendix, we have used the traditional integral approach to derive the expressions pertaining to the harmonic
oscillator that were previously derived using the bracket notation.
I. Evaluation of conditions for a non-zero value of
The analog of this expression in the bracket notation, , was evaluated while deriving the specific
selection rules for vibrational spectroscopy using the harmonic oscillator model. To evaluate this integral, we use the
functional form of the wavefunctions defined as [1]
(A1)
where Nn is the normalization constant
if m n
if m = n (A5)
(A6)
In accord with the notation adopted by Cohen-Tannoudji, we have used the subscript n as opposed to used in
many standard textbooks when labeling the harmonic oscillator wavefunctions.
(A7)
= (A8)
Rearranging eq A6, we obtain an expression for yHn
= (A11)
Given the properties of Hermite polynomials (eq A5), the first integral in eq A11 is zero unless m = n+1. The
second integral is zero unless m = n–1. In other words (m – n) must equal 1 in order to obtain a nonzero value for the
(A12)
(A14)
Substituting eq A14 in eq A13
The first and third integrals equal zero, eq A5. Substituting (eq A5) and the values for Nn
(eq A2) and (eq A4), we obtain
and the proof is complete.
Acknowledgment. The authors express their profound gratitude to the undergraduate chemistry majors at West
Georgia for their enthusiastic reception of the bracket notation, a topic not included in their textbook. They also wish
to thank their colleagues, Professors Andrew J. Leavitt and Spencer J. Slattery, for their insightful comments on this
manuscript. They are greatful to Ms. Brenda Shelnutt for typing this manuscript.