Postulates of Quantum Mechanics and Operator Algebra
CSIR NET chemistry: the postulates, acceptable wavefunctions, linear and Hermitian operators, eigenvalues, commutators, expectation values and uncertainty.
- 24 min
On this page · 19 sections
- Prerequisites
- The Hook: Three Polarisers
- The Postulates
- Acceptable Wavefunctions
- Operators
- Eigenvalue Equations
- Hermitian Operators
- Commutators
- The Uncertainty Principle
- Expectation Values and Superpositions
- Time Evolution and Stationary States
- Parity
- Memory Tricks
- Don’t Confuse
- Common Mistakes
- Practice
- Quick Revision Table
- Key Takeaways
- What’s Next
Prerequisites
- Complex numbers (\( z^*z = \lvert z\rvert^2 \)), differentiation, integration by parts and the idea of a basis in linear algebra.
- BSc-level old quantum theory: Planck’s relation \( E = h\nu \), de Broglie’s \( \lambda = h/p \) and the Bohr atom. The physics treatment is in Wave–Particle Duality and the Schrödinger Equation.
- The Dirac-notation version of this page, written for the Physical Sciences paper, is Operators, Commutators, Dirac Notation and Uncertainty. Reading both sides helps.
The Hook: Three Polarisers
The Postulates
Textbooks (Levine, McQuarrie, Atkins) number them differently but say the same things. Here is the form most useful for the exam.
| # | Postulate | What it gives you |
|---|---|---|
| 1 | The state of a system is completely described by a wavefunction \( \Psi(\mathbf{r}, t) \). \( \Psi^*\Psi\, d\tau \) is the probability of finding the particle in the volume element \( d\tau \) (Born interpretation). | Normalisation; probability densities |
| 2 | Every observable corresponds to a linear Hermitian operator, built from the classical expression by replacing \( x \) with \( \hat{x} = x \) and \( p_x \) with \( \hat{p}_x = -i\hbar\, \partial/\partial x \). | The Hamiltonian, momentum, angular momentum operators |
| 3 | The only values a single measurement of an observable can give are the eigenvalues \( a_n \) of its operator: \( \hat{A}\phi_n = a_n\phi_n \). | Quantisation |
| 4 | For a system in a normalised state \( \Psi \), the average of many measurements is the expectation value \( \langle A\rangle = \int \Psi^*\hat{A}\Psi\, d\tau \). If \( \Psi = \sum_n c_n\phi_n \), a measurement gives \( a_n \) with probability \( \lvert c_n\rvert^2 \), after which the system is in \( \phi_n \). | Probabilities, averages, collapse |
| 5 | The state evolves according to the time-dependent Schrödinger equation, \( i\hbar\, \partial\Psi/\partial t = \hat{H}\Psi \). | Stationary states; time evolution |
Some texts add a sixth postulate: the wavefunction of a system of identical fermions (electrons) is antisymmetric under exchange of any two of them, which gives the Pauli principle. That belongs to many-electron atoms and is treated on the physics side in Identical Particles, Pauli Principle and Spin-Statistics.
graph TD
A["State Psi
(Postulate 1)"] --> B["Observable A
Hermitian operator
(Postulate 2)"]
B --> C["Expand Psi in eigenfunctions
Psi = sum c_n phi_n"]
C --> D["Measurement gives a_n
with probability c_n squared
(Postulates 3, 4)"]
D --> E["State is now phi_n
(collapse)"]
A --> F["Between measurements
Psi evolves by the TDSE
(Postulate 5)"]
F --> CAcceptable Wavefunctions
Because \( \Psi^*\Psi \) is a probability density, \( \Psi \) must be well behaved:
- Single-valued: one probability at each point. On a ring this forces \( \Psi(\phi + 2\pi) = \Psi(\phi) \), which is what quantises \( m \) for angular momentum.
- Finite everywhere (more precisely, square-integrable: \( \int \lvert\Psi\rvert^2 d\tau \) is finite, so it can be normalised).
- Continuous, and with a continuous first derivative wherever the potential is finite. The derivative may jump only at an infinite wall or a delta-function potential.
| Function | Region | Acceptable? | Reason |
|---|---|---|---|
| \( e^{-ax^2} \) (\( a > 0 \)) | all \( x \) | Yes | finite, smooth, square-integrable |
| \( e^{-ax} \) | all \( x \) | No | diverges as \( x \to -\infty \) |
| \( e^{-ax} \) | \( x \ge 0 \) | Yes | decays; this is the form of a radial function |
| \( \tan x \) | all \( x \) | No | infinite at \( x = \pi/2 \) |
| \( \sin(n\pi x/L) \) | \( 0 \le x \le L \), zero outside | Yes | the particle in a box; the derivative jumps only at the infinite walls |
| \( e^{i\phi/2} \) | \( 0 \le \phi < 2\pi \) | No | not single-valued: \( \phi \) and \( \phi + 2\pi \) give different values |
| \( e^{ikx} \) | all \( x \) | Not normalisable | used for free particles as a limit (box or delta-function normalisation) |
Normalisation. If \( \int \psi^*\psi\, d\tau = 1 \) the function is normalised. Multiplying \( \psi \) by any constant, including a phase \( e^{i\alpha} \), describes the same physical state.
Operators
An operator turns one function into another: \( \hat{A}f = g \).
Linear operators
\( \hat{A} \) is linear if \( \hat{A}(c_1f_1 + c_2f_2) = c_1\hat{A}f_1 + c_2\hat{A}f_2 \) for any constants and functions.
| Linear | Not linear |
|---|---|
| \( d/dx \), \( d^2/dx^2 \), multiplication by \( x \) or \( V(x) \), \( \int dx \), \( \hat{H} \), \( \hat{L}_z \) | \( \sqrt{\ } \), squaring \( (\ )^2 \), \( \log \), \( \sin(\ ) \), complex conjugation (antilinear) |
All quantum-mechanical observables use linear operators; that is what makes superposition work.
Operator algebra
- Sum: \( (\hat{A} + \hat{B})f = \hat{A}f + \hat{B}f \).
- Product: \( \hat{A}\hat{B}f = \hat{A}(\hat{B}f) \): apply the right-hand operator first.
- Powers: \( \hat{A}^2 f = \hat{A}(\hat{A}f) \). So \( (d/dx)^2 = d^2/dx^2 \), but \( \hat{A}^2 \) is not “square the result of \( \hat{A} \)”.
- Order matters: in general \( \hat{A}\hat{B} \ne \hat{B}\hat{A} \).
The operators you need
| Observable | Classical | Operator (one dimension, or as stated) |
|---|---|---|
| Position | \( x \) | \( \hat{x} = x \) (multiply) |
| Momentum | \( p_x \) | \( \hat{p}_x = -i\hbar\,\dfrac{\partial}{\partial x} \) |
| Kinetic energy | \( p_x^2/2m \) | \( \hat{T} = -\dfrac{\hbar^2}{2m}\dfrac{\partial^2}{\partial x^2} \); in 3D \( -\dfrac{\hbar^2}{2m}\nabla^2 \) |
| Potential energy | \( V(x) \) | \( \hat{V} = V(x) \) (multiply) |
| Total energy | \( T + V \) | \( \hat{H} = \hat{T} + \hat{V} \) |
| z-angular momentum | \( xp_y - yp_x \) | \( \hat{L}_z = -i\hbar\left(x\dfrac{\partial}{\partial y} - y\dfrac{\partial}{\partial x}\right) = -i\hbar\dfrac{\partial}{\partial\phi} \) |
Eigenvalue Equations
If \( \hat{A}f = af \) with \( a \) a constant, \( f \) is an eigenfunction of \( \hat{A} \) with eigenvalue \( a \). The time-independent Schrödinger equation \( \hat{H}\psi = E\psi \) is the eigenvalue equation of the Hamiltonian.
| Function | Operator | Eigenfunction? | Eigenvalue |
|---|---|---|---|
| \( e^{ikx} \) | \( \hat{p}_x \) | Yes | \( \hbar k \) |
| \( e^{ikx} \) | \( d^2/dx^2 \) | Yes | \( -k^2 \) |
| \( \sin kx \) | \( d^2/dx^2 \) | Yes | \( -k^2 \) |
| \( \sin kx \) | \( \hat{p}_x \) | No | (it is \( (e^{ikx} - e^{-ikx})/2i \): a mixture of \( +\hbar k \) and \( -\hbar k \)) |
| \( e^{-ax^2} \) | \( d^2/dx^2 - 4a^2x^2 \) | Yes | \( -2a \) |
| \( x^n \) | \( x\, d/dx \) | Yes | \( n \) |
| \( e^{im\phi} \) | \( \hat{L}_z \) | Yes | \( m\hbar \) |
The \( e^{-ax^2} \) line is the harmonic oscillator ground state in disguise: \( f' = -2axf \) and \( f'' = (4a^2x^2 - 2a)f \).
Degeneracy. If several linearly independent eigenfunctions share one eigenvalue, that eigenvalue is degenerate, and any linear combination of them is also an eigenfunction with the same eigenvalue. This is why real orbitals such as \( p_x \) (a combination of \( m = +1 \) and \( m = -1 \)) are still eigenfunctions of \( \hat{H} \) for hydrogen, but not of \( \hat{L}_z \).
Hermitian Operators
An operator is Hermitian if, for all well-behaved \( f \) and \( g \),
$$\int f^*\,(\hat{A}g)\, d\tau = \int g\,(\hat{A}f)^*\, d\tau$$- The eigenvalues of a Hermitian operator are real (so they can be measured).
- Eigenfunctions belonging to different eigenvalues are orthogonal: \( \int \phi_m^*\phi_n\, d\tau = 0 \) for \( a_m \ne a_n \). Degenerate eigenfunctions can always be chosen orthogonal (Schmidt orthogonalisation).
- The eigenfunctions form a complete set: any well-behaved function in the same space can be expanded as \( \sum_n c_n\phi_n \) with \( c_n = \int \phi_n^*\Psi\, d\tau \).
Which are Hermitian?
| Operator | Hermitian? | Why |
|---|---|---|
| \( x \), \( V(x) \) (real) | Yes | multiplication by a real function |
| \( d/dx \) | No (anti-Hermitian) | integration by parts gives a minus sign |
| \( i\, d/dx \), \( -i\hbar\, d/dx \) | Yes | the \( i \) cancels the minus sign |
| \( d^2/dx^2 \) | Yes | two integrations by parts, two minus signs |
| \( \hat{A}\hat{B} \) with \( \hat{A}, \hat{B} \) Hermitian | Only if \( [\hat{A}, \hat{B}] = 0 \) | \( (\hat{A}\hat{B})^\dagger = \hat{B}\hat{A} \) |
| \( \hat{A}\hat{B} + \hat{B}\hat{A} \) | Yes | symmetric combination |
| \( [\hat{A}, \hat{B}] \) | No (anti-Hermitian) | \( i[\hat{A}, \hat{B}] \) is Hermitian |
| \( \hat{x}\hat{p}_x \) | No | \( \hat{x} \) and \( \hat{p}_x \) do not commute |
Commutators
$$[\hat{A}, \hat{B}] = \hat{A}\hat{B} - \hat{B}\hat{A}$$Evaluate a commutator by letting it act on a test function \( f \), then removing \( f \) at the end.
- \( [\hat{A}, \hat{B}] = -[\hat{B}, \hat{A}] \); \( [\hat{A}, \hat{A}] = 0 \); any operator commutes with a constant.
- \( [\hat{A}, \hat{B} + \hat{C}] = [\hat{A}, \hat{B}] + [\hat{A}, \hat{C}] \)
- \( [\hat{A}, \hat{B}\hat{C}] = [\hat{A}, \hat{B}]\hat{C} + \hat{B}[\hat{A}, \hat{C}] \)
- \( [\hat{A}\hat{B}, \hat{C}] = \hat{A}[\hat{B}, \hat{C}] + [\hat{A}, \hat{C}]\hat{B} \)
Results to know by heart (one dimension unless stated):
| Commutator | Value |
|---|---|
| \( [\hat{x}, \hat{p}_x] \) | \( i\hbar \) |
| \( [\hat{x}, \hat{p}_y] \), \( [\hat{y}, \hat{p}_x] \) | \( 0 \) (different coordinates) |
| \( [\hat{x}^n, \hat{p}_x] \) | \( i\hbar\, n x^{n-1} \) |
| \( [f(x), \hat{p}_x] \) | \( i\hbar\, f'(x) \) |
| \( [\hat{x}, \hat{p}_x^2] \) | \( 2i\hbar\,\hat{p}_x \) |
| \( [\hat{H}, \hat{x}] \) | \( -i\hbar\,\hat{p}_x/m \) |
| \( [\hat{H}, \hat{p}_x] \) | \( i\hbar\, dV/dx \) |
| \( [\hat{L}_x, \hat{L}_y] \) | \( i\hbar\,\hat{L}_z \) (and cyclic) |
| \( [\hat{L}^2, \hat{L}_z] \) | \( 0 \) |
Compatible observables
| Pair | Commute? | Consequence |
|---|---|---|
| \( \hat{x} \), \( \hat{p}_x \) | No | position and momentum along the same axis cannot both be sharp |
| \( \hat{x} \), \( \hat{p}_y \) | Yes | independent directions |
| \( \hat{H} \), \( \hat{p}_x \) (free particle, \( V = 0 \)) | Yes | free-particle states \( e^{ikx} \) have sharp \( E \) and \( p \) |
| \( \hat{H} \), \( \hat{p}_x \) (particle in a box) | No | box states have sharp \( E \) but not sharp \( p \) |
| \( \hat{H} \), \( \hat{L}^2 \), \( \hat{L}_z \) (any central potential) | Yes, all three | hydrogen orbitals labelled by \( n, l, m_l \) |
| \( \hat{L}_x \), \( \hat{L}_z \) | No | only one component of \( \mathbf{L} \) can be sharp |
| \( \hat{H} \), parity \( \hat{\Pi} \) (symmetric \( V \)) | Yes | states of a symmetric well are even or odd |
The Uncertainty Principle
For any two observables, the standard deviations \( \Delta A = \sqrt{\langle A^2\rangle - \langle A\rangle^2} \) obey
$$\Delta A\,\Delta B \ge \frac{1}{2}\left\lvert\langle[\hat{A}, \hat{B}]\rangle\right\rvert$$With \( [\hat{x}, \hat{p}_x] = i\hbar \):
$$\boxed{\Delta x\,\Delta p_x \ge \frac{\hbar}{2}}$$The bound is reached only by Gaussian wavefunctions (the oscillator ground state is one). It is a statement about the spread of results over many identically prepared systems, not about clumsy measurement.
Energy and lifetime. Time is not an operator in ordinary quantum mechanics, but a state with lifetime \( \tau \) has an energy spread of order \( \hbar/\tau \). In spectroscopy this is lifetime broadening: the linewidth in wavenumbers is \( \delta\tilde{\nu} \approx 1/(2\pi c\tau) \approx 5.3\ \text{cm}^{-1}/(\tau/\text{ps}) \). A state living 1 ps gives a line about 5 cm⁻¹ wide.
Expectation Values and Superpositions
For a normalised state \( \Psi \):
$$\langle A\rangle = \int \Psi^*\hat{A}\Psi\, d\tau$$The operator goes between \( \Psi^* \) and \( \Psi \); it acts on \( \Psi \) only. For a multiplicative operator such as \( x \) or \( V(x) \) the order does not matter; for \( \hat{p}_x \) or \( \hat{T} \) it does.
If \( \Psi = \sum_n c_n\phi_n \) with orthonormal eigenfunctions \( \phi_n \) of \( \hat{A} \):
$$\langle A\rangle = \sum_n \lvert c_n\rvert^2 a_n, \qquad \sum_n \lvert c_n\rvert^2 = 1$$- A single measurement gives one eigenvalue, \( a_n \), never the average (unless the average happens to equal an eigenvalue).
- After the measurement the state is \( \phi_n \); an immediate second measurement gives \( a_n \) again.
- If \( \Psi \) is already an eigenfunction, \( \langle A\rangle = a \) and \( \Delta A = 0 \).
Real and imaginary wavefunctions. For any real, normalisable wavefunction in one dimension, \( \langle p_x\rangle = 0 \). Momentum needs a complex wavefunction such as \( e^{ikx} \). The box states \( \sin(n\pi x/L) \) are real, so \( \langle p_x\rangle = 0 \) but \( \langle p_x^2\rangle = 2mE_n \ne 0 \): the particle moves both ways with equal probability.
Time Evolution and Stationary States
If \( \hat{H} \) does not depend on time, the TDSE separates. Each energy eigenfunction evolves as
$$\Psi_n(x, t) = \psi_n(x)\, e^{-iE_nt/\hbar}$$The phase factor has modulus 1, so \( \lvert\Psi_n\rvert^2 = \lvert\psi_n\rvert^2 \) does not change with time, and neither does the expectation value of any time-independent operator. These are stationary states. “Stationary” does not mean the electron is at rest (\( \langle T\rangle > 0 \)); it means nothing measurable changes.
A superposition of states with different energies is not stationary:
$$\lvert\Psi\rvert^2 = \lvert c_1\rvert^2\psi_1^2 + \lvert c_2\rvert^2\psi_2^2 + 2\,c_1c_2\,\psi_1\psi_2\cos\!\left(\frac{(E_2 - E_1)t}{\hbar}\right)$$(for real \( c_1, c_2, \psi_1, \psi_2 \)). The density sloshes back and forth with period \( h/(E_2 - E_1) \). This is the quantum picture of a wavepacket and of a transition dipole oscillating at the transition frequency.
Ehrenfest’s theorem says expectation values obey the classical equations: \( d\langle x\rangle/dt = \langle p_x\rangle/m \) and \( d\langle p_x\rangle/dt = -\langle dV/dx\rangle \). It follows directly from the commutators \( [\hat{H}, \hat{x}] \) and \( [\hat{H}, \hat{p}_x] \) in the table above.
Parity
The parity operator reflects through the origin: \( \hat{\Pi}f(x) = f(-x) \). Since \( \hat{\Pi}^2 = 1 \), its eigenvalues are \( +1 \) (even functions) and \( -1 \) (odd functions). If \( V(x) = V(-x) \), \( \hat{\Pi} \) commutes with \( \hat{H} \) and every non-degenerate energy eigenfunction is either even or odd. In three dimensions inversion sends \( \mathbf{r} \to -\mathbf{r} \); atomic orbitals have parity \( (-1)^l \), the origin of the g and u labels used in Bonding in Homonuclear Molecules.
An integral of an odd function over a symmetric range is zero. This one fact kills many expectation values in a line: \( \langle x\rangle = 0 \) for every energy eigenstate of a symmetric well centred at the origin (a superposition of an even and an odd state can have \( \langle x\rangle \ne 0 \)).
Memory Tricks
Don’t Confuse
| This | vs | That |
|---|---|---|
| Eigenvalue: a possible result of one measurement | Expectation value: average of many measurements; need not be an eigenvalue | |
| Linear operator: \( \hat{A}(f + g) = \hat{A}f + \hat{A}g \) | Hermitian operator: real eigenvalues, orthogonal eigenfunctions (every Hermitian operator is linear; not every linear one is Hermitian, e.g. \( d/dx \)) | |
| Stationary state: \( \lvert\Psi\rvert^2 \) constant in time | Particle at rest: never; \( \langle T\rangle > 0 \) in every bound state | |
| Degenerate eigenfunctions: any combination is still an eigenfunction | Non-degenerate: combinations of different eigenvalues are not eigenfunctions | |
| \( \sin kx \): eigenfunction of \( d^2/dx^2 \) and of \( \hat{T} \) | \( \sin kx \): not an eigenfunction of \( \hat{p}_x \) | |
| \( \Delta x\Delta p \ge \hbar/2 \) (standard deviations) | \( \Delta x\Delta p \ge h/4\pi \) is the same thing written with \( h \) |
Common Mistakes
Practice
All questions below are original practice questions written for this page. They are not from any CSIR NET paper.
Level 1: Foundation
- What is the eigenvalue of \( \hat{p}_x \) acting on \( e^{-3ix} \) (x in metres)?
- Is \( \sqrt{\ } \) (take the square root) a linear operator?
- What is \( [\hat{x}, \hat{p}_y] \)?
Answers
- \( \hat{p}_x e^{-3ix} = -i\hbar(-3i)e^{-3ix} = -3\hbar\, e^{-3ix} \). Eigenvalue \( -3\hbar \) (in kg m s⁻¹ when \( \hbar \) is in J s): momentum along \( -x \).
- No. \( \sqrt{f + g} \ne \sqrt{f} + \sqrt{g} \).
- Zero. \( x \) and \( p_y \) refer to independent coordinates.
Level 2: Part B Style
Practice Q1. Which function is an eigenfunction of \( d^2/dx^2 \) but not of \( d/dx \)?
(a) \( e^{2ix} \) (b) \( \cos 2x \) (c) \( e^{-2x} \) (d) \( x^2 \)
Answer: (b) cos 2x
\( d^2/dx^2(\cos 2x) = -4\cos 2x \) (eigenvalue −4), but \( d/dx(\cos 2x) = -2\sin 2x \), not a multiple of \( \cos 2x \).
(a) and (c) are eigenfunctions of both (\( 2i \) and \( -4 \); \( -2 \) and \( 4 \)). (d) is an eigenfunction of neither.
Practice Q2. The commutator \( [\hat{x}^3, \hat{p}_x] \) equals
(a) \( 3i\hbar x^2 \) (b) \( -3i\hbar x^2 \) (c) \( i\hbar x^3 \) (d) \( 3\hbar x^2 \)
Answer: (a)
\( [f(x), \hat{p}_x] = i\hbar f'(x) \) with \( f = x^3 \) gives \( 3i\hbar x^2 \). Option (b) is \( [\hat{p}_x, \hat{x}^3] \), the reversed order.
Practice Q3. Which of the following operators is Hermitian?
(a) \( d/dx \) (b) \( x\, d/dx \) (c) \( i\, d/dx \) (d) \( \hat{x}\hat{p}_x \)
Answer: (c)
\( d/dx \) is anti-Hermitian; multiplying by \( i \) makes it Hermitian (it is \( -\hat{p}_x/\hbar \)). (b) and (d) are products of non-commuting factors and are not Hermitian.
Practice Q4. A particle in a one-dimensional box (\( E_n = n^2E_1 \)) is in the normalised state \( \Psi = \tfrac{1}{2}\psi_1 + \tfrac{\sqrt{3}}{2}\psi_3 \). The average energy and the probability of finding \( E_3 \) are
(a) \( 7E_1 \) and 3/4 (b) \( 7E_1 \) and \( \sqrt{3}/2 \) (c) \( 5E_1 \) and 3/4 (d) \( 2.5E_1 \) and 1/2
Answer: (a)
Probabilities: \( \lvert c_1\rvert^2 = 1/4 \), \( \lvert c_3\rvert^2 = 3/4 \). \( \langle E\rangle = \tfrac{1}{4}(1) + \tfrac{3}{4}(9) = \tfrac{1 + 27}{4} = 7 \), in units of \( E_1 \). Option (b) uses the amplitude instead of its square.
Practice Q5. If the position of an electron is known to within 0.10 nm, the minimum uncertainty in its speed is closest to
(a) \( 5.8 \times 10^3 \) m s⁻¹ (b) \( 5.8 \times 10^5 \) m s⁻¹ (c) \( 1.2 \times 10^6 \) m s⁻¹ (d) \( 3.2 \times 10^2 \) m s⁻¹
Answer: (b)
\( \Delta v \ge \hbar/(2m_e\Delta x) = 1.055 \times 10^{-34}/(2 \times 9.11 \times 10^{-31} \times 1.0 \times 10^{-10}) = 5.8 \times 10^5 \) m s⁻¹. Option (c) uses \( \hbar/(m\Delta x) \) without the 1/2; option (d) is the answer for a proton.
Level 3: Part C Style
Practice Q6. A trial function for a particle in a box of length \( L \) is \( \psi = Nx(L - x) \), \( 0 \le x \le L \). Which statements are correct?
- \( N = \sqrt{30/L^5} \).
- \( \langle E\rangle = 5\hbar^2/(mL^2) \).
- \( \langle E\rangle \) is about 1.3% above the true ground-state energy \( \pi^2\hbar^2/(2mL^2) \).
- \( \psi \) is an eigenfunction of \( \hat{H} \) for the box.
(a) 1, 2 and 3 only (b) 1 and 4 only (c) 2 and 3 only (d) 1, 2, 3 and 4
Answer: (a) 1, 2 and 3 only
- Correct. \( \int_0^L x^2(L - x)^2 dx = L^5/30 \) (worked above).
- Correct. Inside the box \( \hat{H} = -\frac{\hbar^2}{2m}\frac{d^2}{dx^2} \) and \( \psi'' = -2N \). So $$\langle E\rangle = -\frac{\hbar^2}{2m}N^2\int_0^L x(L - x)(-2)\,dx = \frac{\hbar^2}{m}N^2\cdot\frac{L^3}{6} = \frac{\hbar^2}{m}\cdot\frac{30}{L^5}\cdot\frac{L^3}{6} = \frac{5\hbar^2}{mL^2}$$
- Correct. Ratio \( \dfrac{5\hbar^2/mL^2}{\pi^2\hbar^2/2mL^2} = \dfrac{10}{\pi^2} = 1.013 \). The average is above the true ground state, as it must be for any trial function (the idea behind the variational method of the next unit).
- Wrong. \( \hat{H}\psi = (\hbar^2/m)N \), a constant, not a multiple of \( \psi \).
Practice Q7. Consider the statements:
- If \( \hat{A} \) and \( \hat{B} \) are Hermitian, \( i[\hat{A}, \hat{B}] \) is Hermitian.
- Eigenfunctions of a Hermitian operator with different eigenvalues are orthogonal.
- \( \hat{L}_z \) and \( \hat{L}^2 \) cannot have simultaneous eigenfunctions.
- For a particle in a box, energy and momentum can both be sharp in every eigenstate.
Which are correct?
(a) 1 and 2 only (b) 1, 2 and 3 only (c) 2 and 4 only (d) 1, 2 and 4 only
Answer: (a) 1 and 2 only
- Correct. \( [\hat{A}, \hat{B}] \) is anti-Hermitian; the \( i \) fixes it.
- Correct. A standard property.
- Wrong. They commute; the spherical harmonics \( Y_l^m \) are common eigenfunctions.
- Wrong. In the box \( \hat{H} \) and \( \hat{p}_x \) do not commute (the walls break translational symmetry); \( \sin(n\pi x/L) \) is a mixture of \( \pm n\pi\hbar/L \).
Practice Q8. An electron in a one-dimensional box of length 1.0 nm is prepared in \( \Psi = (\psi_1 + \psi_2)/\sqrt{2} \) at \( t = 0 \). (\( h = 6.626 \times 10^{-34} \) J s, \( m_e = 9.109 \times 10^{-31} \) kg.) Which is correct?
(a) \( \lvert\Psi\rvert^2 \) is time-independent because each component is a stationary state. (b) \( \lvert\Psi\rvert^2 \) oscillates with period of about \( 3.7 \times 10^{-15} \) s. (c) \( \lvert\Psi\rvert^2 \) oscillates with period of about \( 1.1 \times 10^{-14} \) s. (d) An energy measurement gives \( 2.5E_1 \) with certainty.
Answer: (b)
\( E_1 = h^2/(8m_eL^2) = (6.626 \times 10^{-34})^2/(8 \times 9.109 \times 10^{-31} \times 10^{-18}) = 6.02 \times 10^{-20} \) J.
\( E_2 - E_1 = 3E_1 = 1.81 \times 10^{-19} \) J.
The cross term oscillates as \( \cos[(E_2 - E_1)t/\hbar] \), with period \( T = 2\pi\hbar/(E_2 - E_1) = h/(E_2 - E_1) = 6.626 \times 10^{-34}/1.81 \times 10^{-19} = 3.7 \times 10^{-15} \) s.
(a) is wrong: a superposition of different energies is not stationary. (c) uses \( h/E_1 \) instead of \( h/\Delta E \). (d) confuses the average with a measured value.
Quick Revision Table
| Idea | Fact |
|---|---|
| Born interpretation | \( \Psi^*\Psi\, d\tau \) = probability; normalise to 1 |
| Acceptable \( \Psi \) | single-valued, finite (square-integrable), continuous with continuous derivative where \( V \) is finite |
| Momentum operator | \( -i\hbar\, \partial/\partial x \); kinetic energy \( -(\hbar^2/2m)\nabla^2 \) |
| Linear | \( \hat{A}(c_1f + c_2g) = c_1\hat{A}f + c_2\hat{A}g \) |
| Hermitian | real eigenvalues, orthogonal eigenfunctions, complete set |
| \( d/dx \) | anti-Hermitian; \( i\,d/dx \) Hermitian; \( d^2/dx^2 \) Hermitian |
| Key commutator | \( [x, p_x] = i\hbar \); \( [f(x), p_x] = i\hbar f' \) |
| Commuting | common eigenfunctions; simultaneously sharp |
| Uncertainty | \( \Delta A\Delta B \ge \tfrac{1}{2}\lvert\langle[A, B]\rangle\rvert \); \( \Delta x\Delta p \ge \hbar/2 \) |
| Superposition | outcome \( a_n \) with probability \( \lvert c_n\rvert^2 \); \( \langle A\rangle = \sum\lvert c_n\rvert^2a_n \) |
| Stationary state | \( \psi e^{-iEt/\hbar} \); density constant in time |
| Beat period | \( h/(E_2 - E_1) \) for a two-state superposition |
| Lifetime broadening | \( \delta\tilde{\nu} \approx 5.3\ \text{cm}^{-1}/(\tau/\text{ps}) \) |
| Parity | eigenvalues \( \pm1 \); orbitals \( (-1)^l \) |
Key Takeaways
What’s Next
- Particle in a Box, Harmonic Oscillator and Tunnelling: the postulates applied to the three standard one-dimensional problems.
- Orbital and Spin Angular Momentum uses the commutator algebra of this page.
- Basic Principles of Quantum Mechanics: Quick Revision.
- Back to the chapter overview.