Qualitative Features of One-Dimensional Bound States

One-dimensional bound states have more structure than their explicit formulas suggest. Before solving a particular potential, one can often predict the ordering of states, the number of nodes, the role of parity, whether tails are long or short, and how the spectrum shifts when the potential is squeezed or deformed.

Consider a particle on a connected line or interval, with a real nonsingular scalar potential and Hamiltonian

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

with separated self-adjoint endpoint conditions on an interval, or decaying bound-state conditions on the line. Rigorous hypotheses matter for singular potentials and unusual domains; the mathematical framework is discussed in Sturm-Liouville Theory. The goal here is the practical physical picture used before and after exact calculations.

Helpful background. Time-Independent Schrödinger Equation to formulate an energy eigenvalue problem.

What Counts As A Bound State

A bound state is a normalizable energy eigenstate. On the real line this means

$$\int_{-\infty}^{\infty} \lvert\psi(x)\rvert^2\,dx =1,$$

and the wavefunction must decay sufficiently fast as $\lvert x\rvert\to\infty$. On a finite interval, normalizability is automatic for square-integrable functions, but the allowed states are still selected by the endpoint boundary conditions.

For a stationary bound state,

$$\hat H\psi_n=E_n\psi_n,$$

the probability density is time independent. General time-dependent bound motion comes from superpositions, as explained in Stationary States and Expansions.

Ordering And Nodes

For a broad class of regular one-dimensional bound-state problems, energy eigenstates can be ordered so that the $n$th state has $n$ interior nodes if the count begins at $n=0$:

$$\psi_0:\;0\ \text{nodes}, \qquad \psi_1:\;1\ \text{node}, \qquad \psi_2:\;2\ \text{nodes}, \quad \ldots$$

If a page labels the infinite-well states by $n=1,2,\ldots$, the same statement becomes: the $n$th state has $n-1$ interior nodes.

This node rule is not a decorative pattern. It expresses a basic cost: extra sign changes require extra curvature, and curvature contributes kinetic energy. For a real wavefunction in a region where boundary terms vanish,

$$\langle T\rangle = \frac{\hbar^2}{2m} \int \lvert \psi'(x)\rvert^2\,dx.$$

More oscillation usually means a larger kinetic-energy contribution. The potential-energy term may compensate in some regions, but the ordered node count remains a robust guide for ordinary one-dimensional bound spectra.

The Ground State Has No Nodes

The lowest bound state is strictly positive in the interior after a phase choice. To see the mechanism for a regular potential, choose a real minimizing wavefunction and replace it by its absolute value. Both its potential energy and kinetic quadratic form are unchanged, so a nonnegative minimizer exists. It solves the Schrödinger equation. If it vanished at an interior point, its derivative would also vanish there because that point is a minimum. Uniqueness for the second-order equation would then force the solution to vanish everywhere, contradicting normalization.

This is one of the most useful sanity checks in wave mechanics. If a proposed ground state crosses zero inside the allowed region, it is not the ground state of an ordinary one-dimensional bound problem.

There are caveats. Nodes may be forced by boundary conditions, by angular-momentum factors in radial problems after a change of variables, or by symmetry if one restricts attention to a particular odd sector. Those are not contradictions; they mean the word “ground” must refer to the full domain and the full set of allowed states.

Parity In Symmetric Potentials

If the potential is even,

$$V(x)=V(-x),$$

and the domain is invariant under reflection, then the Hamiltonian commutes with the parity operator. Bound-state eigenfunctions can be chosen with definite parity:

$$\psi_+(x)=\psi_+(-x), \qquad \psi_-(x)=-\psi_-(-x).$$

Even states have $\psi'(0)=0$ when they are smooth at the origin. Odd states have $\psi(0)=0$. This converts a symmetric full-line problem into a half-line problem with two possible boundary conditions at the origin.

For ordinary one-dimensional bound states, the parity alternates with energy:

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

The centered Infinite Square Well and the symmetric Finite Square Well show this pattern explicitly. The Parity page gives the symmetry principle behind it.

Localization Costs Kinetic Energy

Confinement raises energy because a sharply localized wavefunction must contain short-wavelength components. A rough uncertainty estimate gives

$$\Delta x\,\Delta p\gtrsim\frac{\hbar}{2},$$

so a state localized over a length $L$ has a typical kinetic-energy scale

$$E_{\text{kin}} \sim \frac{(\Delta p)^2}{2m} \sim \frac{\hbar^2}{2mL^2}.$$

This scaling is not just dimensional analysis. In the Infinite Square Well,

$$E_1=\frac{\pi^2\hbar^2}{2mL^2}.$$

Changing the shape of the well changes numerical constants and wavefunction shapes, but the lesson survives: tighter confinement raises the zero-point energy.

Tunneling Tails

In a classically forbidden region where $V(x)>E$, the stationary Schrödinger equation locally gives exponential behavior. For a constant forbidden-region height,

$$\psi(x)\sim e^{-\kappa x}, \qquad \kappa= \frac{\sqrt{2m(V-E)}}{\hbar}.$$

The tail length is therefore

$$\ell_{\text{tail}}=\frac{1}{\kappa}.$$

Higher barriers, larger masses, and lower energies make the tail shorter. A state close to a binding threshold has a small outside decay constant and a long tail. This is why weakly bound states can be spatially large even when the attractive region is narrow.

The Finite Square Well is the clean first model of bound-state tails. Scattering tunneling uses the same forbidden-region exponential in a different normalization and current setting.

Level Spacing And Confinement

Level spacing is controlled by available length scale and by the classical speed through the allowed region. Narrower wells generally have larger spacings. Wider wells generally have denser spectra.

For a box of length $L$,

$$E_n = \frac{n^2\pi^2\hbar^2}{2mL^2},$$

so all levels move upward as $L$ shrinks. Widening a finite square well lowers its confinement scale and can add bound states. Deepening it also lowers energies relative to the outside threshold and can add states, but need not decrease level spacings. At fixed width, a deep square well approaches the hard-wall spacings as its tails shrink.

Smooth potentials show the same idea in a less rigid form. Near a stable minimum,

$$V(x)\approx V(x_0) +\frac{1}{2}m\omega^2(x-x_0)^2,$$

so low-lying states resemble oscillator states with spacing approximately $\hbar\omega$. A steeper minimum has larger $\omega$ and wider spacing.

Deforming A Well

When a potential is changed continuously, bound-state energies usually move continuously. One-dimensional node ordering gives a strong constraint: the ground state remains the lowest nodeless state, the first excited state remains the one-node state, and so on, unless a state leaves or enters the bound spectrum at threshold.

Two common deformations are worth separating.

A symmetric deformation preserves parity, but it does not allow the exact bound levels of this connected scalar one-dimensional problem to cross. The eigenvalues are nondegenerate, and their node ordering persists. For two solutions at the same energy, the Wronskian is constant; the endpoint or decay conditions force it to zero, making the solutions proportional. Symmetry-protected crossings familiar from more general systems therefore require hypotheses outside this setting, such as internal degrees of freedom or disconnected wells.

A useful approximation instead starts with two localized, orthonormal states, one in each well. Their uncoupled energies may cross as the wells are biased. A real two-state Hamiltonian has the form

$$H_{\mathrm{eff}}(\lambda) = \begin{pmatrix} E_a(\lambda) & K(\lambda) \\ K(\lambda) & E_b(\lambda) \end{pmatrix}.$$

The eigenvalue separation is

$$\Delta E = \sqrt{ \left(E_a-E_b\right)^2 +4K^2 }.$$

If $K\ne0$, the levels avoid crossing. This is the same finite-dimensional mixing logic used in the Double-Well Potential and in degenerate perturbation theory.

What One Dimension Hides

One-dimensional bound spectra are unusually ordered. In higher dimensions, degeneracies can arise from rotational symmetry, hidden symmetry, spin, or tensor-product structure. Nodes become nodal surfaces rather than isolated zeros. Radial equations also carry measures and effective centrifugal potentials. The one-dimensional intuition remains valuable, but it should not be mistaken for a universal theorem about all quantum systems.

Practical Checklist

Before solving a one-dimensional bound-state problem, ask:

• What is the domain and which boundary conditions define the Hamiltonian?
• Is the state normalizable, or is it a scattering state?
• Is the potential symmetric, so that parity can reduce the problem?
• How many nodes should the candidate state have?
• Where are the classically allowed and forbidden regions?
• How long should the evanescent tails be?
• What length scale sets the kinetic-energy cost?
• Does an apparent crossing come from uncoupled approximate levels, and what tunneling couples them?

This checklist catches many algebraic mistakes before the algebra begins.

Common Mistakes

• Drawing a one-dimensional ground state with an interior node.
• Forgetting that the node count depends on the chosen ordering convention.
• Treating parity as available when the potential is not symmetric.
• Setting finite-well wavefunctions to zero at finite walls.
• Calling a long evanescent tail “classically allowed” because it has nonzero probability density.
• Predicting an exact even–odd bound-level crossing in a regular connected one-dimensional scalar potential; nondegeneracy rules it out.
• Importing the simple one-dimensional node theorem into higher-dimensional or spinful systems without checking the quantum numbers.

Exercises

1. A proposed ground-state wavefunction in a smooth one-dimensional well changes sign once in the interior. What does the node rule tell you?

Solution

For an ordinary one-dimensional bound problem, the true ground state is nodeless. A wavefunction with one interior node is a candidate for an excited state, not the ground state. The proposed state may satisfy the differential equation for some higher energy, but it cannot be the lowest state unless the domain or symmetry restriction has forced that node.

2. In a symmetric potential, why can one solve only on $x\ge0$ if parity is fixed?

Solution

If $V(x)=V(-x)$, eigenstates can be chosen even or odd. An even state obeys $\psi'(0)=0$, while an odd state obeys $\psi(0)=0$. Once one of these origin conditions is imposed, the half-line solution determines the full-line solution by reflection.

3. A finite well is made wider while its depth and the particle mass are held fixed. What qualitative changes do you expect?

Solution

The kinetic-energy scale associated with confinement decreases roughly like $1/L^2$, so low-lying levels generally move downward relative to the outside threshold and become more closely spaced. A wider attractive region can also support additional bound states. Existing states tend to be less tightly curved in the well, while near-threshold states may have long exterior tails.

4. In a model of two orthonormal localized states, explain how coupling replaces a crossing of their uncoupled energies by an avoided crossing.

Solution

For a two-state Hamiltonian

$$H_{\mathrm{eff}} = \begin{pmatrix} E_a & K \\ K & E_b \end{pmatrix},$$

the eigenvalue separation is

$$\Delta E = \sqrt{(E_a-E_b)^2+4K^2}.$$

If $K=0$, the uncoupled model energies can cross when $E_a=E_b$. For constant $K\ne0$, their separation at that point is $2\lvert K\rvert$. In a connected finite double well, tunneling couples the localized states, and its exact bound levels remain nondegenerate. A zero splitting arises only in a decoupled limit such as an infinite separating barrier.

References

• C. Cohen-Tannoudji, B. Diu, and F. Laloë, Quantum Mechanics, Wiley, 1977.
• D. J. Griffiths and D. F. Schroeter, Introduction to Quantum Mechanics, 3rd ed., Cambridge University Press, 2018.
• L. D. Landau and E. M. Lifshitz, Quantum Mechanics: Non-Relativistic Theory, 3rd ed., Pergamon, 1977.
• R. Shankar, Principles of Quantum Mechanics, 2nd ed., Springer, 1994.
• G. Teschl, Mathematical Methods in Quantum Mechanics, 2nd ed., American Mathematical Society, 2014, Chapter 9 (separated boundary conditions and oscillation theory).