Skip to content

Parity and Nodes

Parity and nodes are two of the most powerful organizing tools for one-dimensional bound states. Parity uses symmetry to split an even potential into even and odd sectors. Node counting orders the bound states by the number of times their wavefunctions cross zero.

This page is the focused one-dimensional solving guide. The broader qualitative overview is Qualitative Features of One-Dimensional Bound States, while the abstract symmetry principle is Parity.

Consider a real one-dimensional Hamiltonian

H^=−ℏ22md2dx2+V(x),\hat H = -\frac{\hbar^2}{2m}\frac{d^2}{dx^2} +V(x),

with boundary conditions that define a self-adjoint bound-state problem. A bound state satisfies

H^ψn=Enψn,∫∣ψn(x)∣2 dx<∞.\hat H\psi_n=E_n\psi_n, \qquad \int \lvert\psi_n(x)\rvert^2\,dx\lt\infty.

The statements below apply most directly to ordinary regular one-dimensional wells. Singular potentials, half-line problems, and radial reductions may require adjusted boundary conditions, but the same ideas remain useful diagnostics.

Parity acts as

(Πψ)(x)=ψ(−x).(\Pi\psi)(x)=\psi(-x).

If the potential is even,

V(−x)=V(x),V(-x)=V(x),

then

[H^,Π]=0.[\hat H,\Pi]=0.

Energy eigenstates can then be chosen as parity eigenstates:

Πψ=ηψ,η=±1.\Pi\psi=\eta\psi, \qquad \eta=\pm1.

In position space,

ψ(−x)=ψ(x)for even states,\psi(-x)=\psi(x) \quad \text{for even states},

and

ψ(−x)=−ψ(x)for odd states.\psi(-x)=-\psi(x) \quad \text{for odd states}.

Parity is not available unless the Hamiltonian and its domain are symmetric. An asymmetric square well can still have ordered nodes, but its eigenstates are not generally even or odd.

For a smooth even potential, parity converts the full-line problem into two half-line problems on x≥0x\ge0.

An even state obeys

ψ′(0)=0.\psi'(0)=0.

An odd state obeys

ψ(0)=0.\psi(0)=0.

Then one solves on x≥0x\ge0 and reflects the solution:

ψe(−x)=ψe(x),ψo(−x)=−ψo(x).\psi_{\mathrm e}(-x)=\psi_{\mathrm e}(x), \qquad \psi_{\mathrm o}(-x)=-\psi_{\mathrm o}(x).

This is why symmetric wells often have separate even and odd transcendental equations. In the Finite Square Well, for example, even states use qtan⁡(qa)=κq\tan(qa)=\kappa, while odd states use −qcot⁡(qa)=κ-q\cot(qa)=\kappa.

If the origin contains a singular interaction, such as a delta potential, the origin condition must be derived from the singular matching rule rather than copied from the smooth case.

Ordinary one-dimensional bound states are nondegenerate. A short Wronskian argument explains why.

Suppose ψ\psi and ϕ\phi solve the same real Schrödinger equation with the same energy:

−ℏ22mψ′′+Vψ=Eψ,−ℏ22mϕ′′+Vϕ=Eϕ.-\frac{\hbar^2}{2m}\psi''+V\psi=E\psi, \qquad -\frac{\hbar^2}{2m}\phi''+V\phi=E\phi.

Define the Wronskian

W(x)=ψ(x)ϕ′(x)−ψ′(x)ϕ(x).W(x)=\psi(x)\phi'(x)-\psi'(x)\phi(x).

Subtracting the two equations gives

W′(x)=0,W'(x)=0,

so WW is constant. For bound states on the line, both wavefunctions and their derivatives vanish at infinity fast enough that

W(∞)=0.W(\infty)=0.

Therefore W(x)=0W(x)=0 everywhere. The two solutions are linearly dependent, so they do not represent two independent bound states with the same energy.

This is why, in an even one-dimensional well, each bound-state energy has a definite parity. Degenerate subspaces do not appear for ordinary one-dimensional bound states, so there is no freedom to mix even and odd states at the same energy.

For regular one-dimensional bound-state problems, the energy eigenstates can be ordered as

E0<E1<E2<⋯E_0\lt E_1\lt E_2\lt\cdots

and the corresponding eigenfunction ψj\psi_j has exactly jj interior nodes:

ψ0:  0 nodes,ψ1:  1 node,ψ2:  2 nodes,…\psi_0:\;0\ \text{nodes}, \qquad \psi_1:\;1\ \text{node}, \qquad \psi_2:\;2\ \text{nodes}, \quad \ldots

The physical intuition is kinetic energy. Additional nodes require additional curvature, and curvature contributes to

⟨T⟩=ℏ22m∫∣ψ′(x)∣2 dx\langle T\rangle = \frac{\hbar^2}{2m} \int \lvert\psi'(x)\rvert^2\,dx

when boundary terms vanish.

The mathematical statement is a Sturm–Liouville result: zeros of successive eigenfunctions interlace, and eigenvalues are ordered by oscillation count. The rigorous framework is Sturm–Liouville Theory.

In an even one-dimensional well, parity and node counting combine. The ground state is nodeless and can be chosen positive, so it is even. The first excited state has one node; by symmetry that node lies at the origin, so the state is odd. The next state has two nodes and is even, and so on:

even ground state,odd first excited state,even second excited state,odd third excited state,…\text{even ground state}, \quad \text{odd first excited state}, \quad \text{even second excited state}, \quad \text{odd third excited state}, \ldots

This alternation is a guide, not a replacement for boundary conditions. It assumes an ordinary symmetric one-dimensional bound problem. If the domain itself imposes a node or if the potential is singular at the symmetry point, check the domain before applying the smooth-well pattern.

For an infinite square well centered at the origin, −a<x<a-a\lt x\lt a, the potential is even. The eigenfunctions can be chosen as alternating cosine and sine functions:

ψe(x)∝cos⁡(kx),ψo(x)∝sin⁡(kx),\psi_{\mathrm e}(x)\propto \cos(kx), \qquad \psi_{\mathrm o}(x)\propto \sin(kx),

with hard-wall conditions at x=±ax=\pm a.

The ground state is even and has no interior node. The first excited state is odd and has one node at x=0x=0. The centered representation makes parity visible, while the interval 0<x<L0\lt x\lt L representation used in Infinite Square Well makes boundary conditions at the walls direct.

The finite square well is the first place parity saves serious algebra. The full-line matching problem separates into two equations:

qtan⁡(qa)=κeven,q\tan(qa)=\kappa \quad \text{even},

and

−qcot⁡(qa)=κodd.-q\cot(qa)=\kappa \quad \text{odd}.

The roots alternate by parity as the energy increases. Newly appearing bound states enter from threshold with the next node count, as explained in Bound-State Counting.

In a symmetric double well, the exact eigenstates have definite parity. The lowest pair is approximately

ψ+≈12(ϕL+ϕR),\psi_+ \approx \frac{1}{\sqrt2} \left( \phi_L+\phi_R \right),

and

ψ−≈12(ϕL−ϕR),\psi_- \approx \frac{1}{\sqrt2} \left( \phi_L-\phi_R \right),

where ϕL\phi_L and ϕR\phi_R are approximate localized states in the left and right wells.

The even state is lower because it has no node between the wells. The odd state has a central node and lies higher. The localized states ϕL\phi_L and ϕR\phi_R are useful approximations, but they are not exact stationary states of a symmetric finite double well.

The exactly solvable Double Delta Potential gives explicit even and odd equations; the smoother Double-Well Potential gives the qualitative tunneling picture.

If V(x)≠V(−x)V(x)\ne V(-x), parity no longer commutes with the Hamiltonian. The eigenstates generally have no definite parity, and the half-line reduction is unavailable.

The node theorem still orders ordinary one-dimensional bound states. The ground state is still nodeless, the first excited state still has one node, and so on. The nodes need not sit at the origin, and the wavefunction need not have any simple left-right symmetry.

This distinction is important in asymmetric wells: losing parity does not destroy one-dimensional ordering, but it does remove the even-odd sector decomposition.

When solving a one-dimensional bound problem, ask:

  • Is the potential and domain invariant under x↦−xx\mapsto -x?
  • If yes, should the state be even or odd?
  • For a smooth origin, should the half-line condition be ψ′(0)=0\psi'(0)=0 or ψ(0)=0\psi(0)=0?
  • How many nodes should this energy level have?
  • Does a claimed degeneracy contradict ordinary one-dimensional nondegeneracy?
  • Are localized left-right states being mistaken for exact parity eigenstates?

These checks often catch errors before numerical values are computed.

  • Assuming every one-dimensional potential has parity symmetry.
  • Using even-state boundary conditions for an odd state, or the reverse.
  • Forgetting that ordinary one-dimensional bound states are nondegenerate.
  • Drawing a ground-state wavefunction with an interior node.
  • Treating localized double-well states as exact energy eigenstates in a symmetric well.
  • Applying smooth-origin conditions at a singular point without deriving the matching rule.
  • Confusing the number of nodes with the principal quantum-number convention used by a particular page.
  • D. J. Griffiths and D. F. Schroeter, Introduction to Quantum Mechanics, 3rd ed., Cambridge University Press, 2018.
  • R. Shankar, Principles of Quantum Mechanics, 2nd ed., Springer, 1994.
  • L. D. Landau and E. M. Lifshitz, Quantum Mechanics: Non-Relativistic Theory, 3rd ed., Pergamon, 1977.
  • G. Teschl, Mathematical Methods in Quantum Mechanics, 2nd ed., American Mathematical Society, 2014.
  1. Show that a smooth even wavefunction obeys ψ′(0)=0\psi'(0)=0 and a smooth odd wavefunction obeys ψ(0)=0\psi(0)=0.
Solution

If ψ\psi is even, then

ψ(−x)=ψ(x).\psi(-x)=\psi(x).

Differentiate with respect to xx:

−ψ′(−x)=ψ′(x).-\psi'(-x)=\psi'(x).

Set x=0x=0 to get

−ψ′(0)=ψ′(0),-\psi'(0)=\psi'(0),

so ψ′(0)=0\psi'(0)=0.

If ψ\psi is odd, then

ψ(−x)=−ψ(x).\psi(-x)=-\psi(x).

Setting x=0x=0 gives

ψ(0)=−ψ(0),\psi(0)=-\psi(0),

so ψ(0)=0\psi(0)=0.

  1. Use the Wronskian argument to explain why two independent one-dimensional bound states cannot have the same energy.
Solution

Let ψ\psi and ϕ\phi solve the same real Schrödinger equation at energy EE. Their Wronskian

W=ψϕ′−ψ′ϕW=\psi\phi'-\psi'\phi

has derivative W′=0W'=0, so WW is constant. For bound states on the line, the wavefunctions and derivatives vanish at infinity, so W=0W=0 there. Hence W=0W=0 everywhere. A zero Wronskian means the two solutions are linearly dependent, not independent states.

  1. What parities should the first four bound states of a smooth even one-dimensional well have?
Solution

The ground state is nodeless and even. The first excited state has one node and is odd. The second excited state has two nodes and is even. The third excited state has three nodes and is odd. The sequence is therefore

even,odd,even,odd.\text{even},\quad \text{odd},\quad \text{even},\quad \text{odd}.
  1. In a symmetric double well, why are ϕL\phi_L and ϕR\phi_R not exact energy eigenstates?
Solution

The Hamiltonian has parity symmetry, and ordinary one-dimensional bound states are nondegenerate. Therefore exact energy eigenstates can be chosen with definite parity. A state localized only in the left well is not even or odd; parity maps it to a right-localized state. The exact low-energy states are approximately the even and odd combinations

12(ϕL+ϕR),12(ϕL−ϕR).\frac{1}{\sqrt2} \left( \phi_L+\phi_R \right), \qquad \frac{1}{\sqrt2} \left( \phi_L-\phi_R \right).
  1. An asymmetric well has three bound states. What can you say about their node counts?
Solution

Parity is unavailable because the potential is asymmetric, but ordinary one-dimensional node ordering still applies. Ordered by increasing energy, the three states have

0,1,20,\quad 1,\quad 2

interior nodes respectively. The nodes need not be placed symmetrically.