Principles of Quantum Mechanics (Albert Chap 2) I. Principles of Quantum Mechanics
Principles of Quantum Mechanics (Albert Chap 2) I. Principles of Quantum Mechanics
We can now stipulate how vectors and operators represent quantum states and properties. We’ll
do it in terms of 5 principles:
ASIDE: More precisely, a physical system is represented by an inner-product space. And even more
precisely, the most general type of physical system is represented by a special type of inner-product space, called a
Hilbert space. A Hilbert space is roughly an inner-product space in which every converging sequence of vectors
(every “Cauchy” sequence) converges to a limit vector belonging to the space. This “Cauchy completeness”
technical property only becomes important when we need to talk about properties (operators) that have a
continuum of values.
Note: A property (i.e., hardness, position) of a system has a value (i.e., hard/soft, 5 feet
from origin/10 feet from origin) in a given state of the system. So we need a way of
specifying how properties and their values relate to states:
Eigenvector/Eigenvalue Rule:
A state possesses the value λ of a property represented by operator O if and only if that
state is an eigenvector of O with eigenvalue λ.
1
Application of Principles (A) and (B)
Physical system: electron
We want the states of our electron to be unit length vectors. And if we want to measure the
Hardness property, we need states in which the electron is hard (“hardness” state) and in which it
is soft (“softness” state). We only need 2 since there are only 2 values of the Hardness property.
These states should be orthogonal since they should be eigenvectors of a Hermitian operator with
different eigenvalues. One way to implement this is the following:
“hardness”
1 Check: These are orthonormal vectors:
state
|hard〉 =
0 〈hard|hard〉 = 1
〈soft|soft〉 = 1
“softness” 0 〈hard|soft〉 = 0
state |soft〉 =
1 Note: These form an orthonormal basis for a 2-dim vector space called
“spin space” -- the space of all possible spin states of an electron.
SO: H|hard〉 = +1|hard〉 Translation: When an electron is in the state |hard〉, it has the
Hardness value of hard (i.e., it is hard). When an electron is in
H|soft〉 = −1|soft〉 the state |soft〉, it has a Hardness value of soft (i.e., it is soft).
√ √
|black〉 = |white〉 =
√ −√
Have: 〈black|black〉 = 1
〈white|white〉 = 1
〈black|white〉 = 0
2
The following relations also hold:
Check:
|black〉 = √|hard〉 + √|soft〉
1 0 √ 0
|white〉 = √|hard〉 − √|soft〉 √ + √ = +
0 1 0 √
|hard〉 = √|black〉 + √|white〉
√
= = |black〉
|soft〉 = √|black〉 − √|white〉 √
SO: A white electron (an electron in the state represented by |white〉) is represented as a linear
sum (superposition) of hard and soft states.
Moreover: Is a white electron definitely hard or definitely soft? (Does it have a Hardness
value?) Recall that experiments indicate the answer should be “no”.
Recall Eigenvector/Eigenvalue Rule: To say a white electron has a Hardness value (hard or
soft), it must be in an eigenstate of the Hardness operator.
BUT: The state represented by |white〉 is not an eigenstate of the operator H representing the
Hardness property:
1 0 √ √
H|white〉 = = ≠ λ|white〉 for any value of λ
0 −1 −√ √
So the vector space mathematical formalism and the translation rules given in Principles (A) and
(B) allow us to represent experimental results involving electrons!
Note, finally, an interesting consequence of Principle (B) due to the characteristics of Hermitian
operators:
3
H = “Hamiltonian”
(C) Dynamics: States evolve in time via the Schrödinger equation. operator: represents
the energy of a system
(D) Born Rule: Connection with Experiment (how probabilities enter QM).
Recall Eigenvector/Eigenvalue Rule: A state possesses the value λ of a property represented
by O just when that state is an eigenvector of O with eigenvalue λ.
What about states that aren’t eigenvectors of O?
(2) The probability that |ψ〉 possesses the value bi of the property represented by B is
defined to be:
4
Two notes:
(a) Since |ψ〉 has length 1, The sum of the square of all its expansion coefficients in any basis
must equal 1. This is a mathematical characteristic of probabilities.
(b) When |ψ〉 is itself an eigenvector |bi〉 of B, then the probability that it possesses the value bi
is equal to 1 (as it should be, according to the Eigenvector/Eigenvalue Rule):
If:
|ψ〉 = |bi〉
then:
|〈ψ|bi〉|2 = |〈bi|bi〉|2 = 1
After measurement, suppose we get the value bi. Then the Projection Postulate says that
the state collapses to the state represented by |bi〉:
|ψ〉 ⎯⎯⎯→
collapse
|bi〉
Motivation: Measurements are repeatible. If we measure the value bi once, then we should
get the same value bi on a second measurement (provided the system is not
interferred with in between).
5
Recap: 5 Principles of Quantum Mechanics
(A) States are represented by vectors of length 1.
(B) Properties are represented by Hermitian operators.
Eigenvector/Eigenvalue Rule:
A state possesses the value λ of a property represented by operator
O if and only if that state is an eigenvector of O with eigenvalue λ.
|ψ(t1)〉 ⎯⎯⎯→
Schrödinger
|ψ(t2)〉
evolution
Now consider another state space for “coordinate” properties (Position, Momentum, Energy, etc.):
• Infinite continuum of values.
• State space (“coordinate” space) is an infinite-dimensional vector space.
• Many orthonormal bases; each associated with a coordinate property (Position, Momentum,
Energy, etc); all of which are mutually incompatible.
X|5〉 = 5|5〉
eigenvector eigenvalue
of X of X
6
Suppose |ψ〉 represents the state of an electron located at some position.
Can expand |ψ〉 in eigenvectors of position operator X:
ASIDE: The sum here should
|ψ〉 = a1|1〉 + a1.00001|1.00001〉 + ... + a72.93|72.93〉 + ... really be an integral, since x
takes on a continuum of values.
where ax = 〈ψ|x〉, x runs from −∞ to +∞
expansion coefficients
Again: All of these infinite number of position eigenvectors are orthogonal to each other. They
form a basis for coordinate space.
ASIDE: The wave function is really just a way of encoding the expansion coefficients of a
state vector |ψ〉 in a given coordinate basis. We can also expand |ψ〉 in other coordinate
space bases; a momentum basis, for instance (we would then get a momentum wave
function ψ(p)). In general, we can form a wave function for any state vector expanded in a
basis of eigenvectors of an operator with a continuous range (or “spectrum”) of eigenvalues.
Born Rule:
Pr(electron is located at position x1 in state |ψ〉) = |〈ψ|x1〉|2 = |ψ(x1)|2