40.4: Finite Wells

Physics: A Foundation for Scientists and Engineers · LibreTexts

0%

Contents

    Text

    Size
    100%
    Theme
    Typeface
    Line length
    Spacing

    Read aloud

    How

    Speed: 1×
    Light up
    Highlight

    Symbols: Mulberry Symbols, © 2018–2026 Steve Lee, CC BY-SA 4.0 (licence, the set), the site's own copies.

    From LibreTexts, under CC BY 4.0, as the source states it; its equations are written in TeX, which the PDF shows typeset. This page at LibreTexts.

    40.4: Finite Wells

    Real traps do not have infinitely high walls. A more realistic model is the finite well: the potential energy inside is still zero, but outside it rises only to some finite value \(U\) (the well depth), with the trapped particle’s energy \(E\) less than \(U\). We split space into three regions: region I to the left of the well (\(x < 0\)), region II inside (\(0 < x < L\)), and region III to the right (\(x > L\)).

    Finite potential well of depth U with the ground-state wave function.
    Figure 40.5. Ground-state wave function of a finite well (gold) of depth \(U\). Inside (region II) the wave function is sinusoidal; in the classically forbidden regions I and III it does not vanish but decays exponentially over a depth \(\sim 1/C\). The leakage means the particle can be found, with small probability, outside the well.
    Figure descriptionA finite potential well, drawn in gold: the potential is flat at height U on both sides and drops to zero between x equals 0 and x equals L, dividing space into region I on the left, region II inside, and region III on the right. The blue ground-state wave function is a single smooth arch inside the well, but instead of falling to zero at the walls it continues outward into regions I and III, where it decays away exponentially rather than stopping. That leakage into the classically forbidden regions is the whole point of the figure.

    The crucial difference from the infinite well is that the wave function no longer drops abruptly to zero at the walls. It leaks into the classically forbidden regions I and III, so there is a small but nonzero chance of finding the particle outside a well it does not have the energy to escape. To see why, rearrange the Schrödinger equation (40.1) as

    \[\frac{\mathrm{d}^{2}\psi}{\mathrm{d}x^{2}} = \frac{2m(U - E)}{\hbar^2}\,\psi. \tag{40.10}\]

    Inside the well \(U = 0 < E\) and the right-hand side is negative, giving oscillating (sinusoidal) solutions just as before. Outside, \(U > E\), so the right-hand side is positive, and the solutions are exponentials rather than oscillations. Defining

    \[C = \frac{\sqrt{2m(U - E)}}{\hbar}, \qquad\text{so that}\qquad \frac{\mathrm{d}^{2}\psi}{\mathrm{d}x^{2}} = C^2\psi, \tag{40.11}\]

    the general solution outside is \(\psi = Ae^{Cx} + Be^{-Cx}\). Physical sense (the probability cannot blow up far from the well) forces us to keep only the decaying exponential in each outside region:

    \[\psi_{\mathrm{I}}(x) = Ae^{Cx} \ \ (x < 0), \qquad \psi_{\mathrm{III}}(x) = Be^{-Cx} \ \ (x > L), \tag{40.12}\]

    each dying away over a characteristic penetration depth \(1/C\). Inside, region II is sinusoidal, \(\psi_{\mathrm{II}}(x) = F\sin kx + G\cos kx\) (Figure 40.5).

    We do not solve for all the constants here: that requires matching \(\psi\) and its slope at both walls,

    \[\begin{aligned} \psi_{\mathrm{I}}(0) &= \psi_{\mathrm{II}}(0), &\qquad \psi_{\mathrm{I}}'(0) &= \psi_{\mathrm{II}}'(0),\\ \psi_{\mathrm{II}}(L) &= \psi_{\mathrm{III}}(L), &\qquad \psi_{\mathrm{II}}'(L) &= \psi_{\mathrm{III}}'(L), \end{aligned} \tag{40.13}\]

    together with normalization, and leads to a transcendental equation best handled numerically. What matters physically is the consequence: because the wave function spreads a little beyond the walls, its effective wavelength is slightly longer than in an infinite well of the same width, so each energy level sits a little lower than the corresponding infinite-well level. A finite well also holds only a finite number of bound states; climb high enough and the particle is no longer trapped at all.

    Checkpoint 40.4

    In an infinite well the wave function is exactly zero at and beyond the walls; in a finite well it is not. Use this difference to explain, qualitatively, why the finite well’s energy levels are lower than the infinite well’s for the same width \(L\).

    Answer

    Quantum Tunneling

    Quantum tunneling: a matter wave of energy E meeting a barrier of height U and width L.
    Figure 40.6. Quantum tunneling. A matter wave of energy \(E\) strikes a barrier of height \(U > E\) and width \(L\). Its amplitude decays exponentially inside the barrier (red) but does not reach zero, so a reduced-amplitude wave emerges on the far side and the particle may be transmitted. Transmission falls off steeply with barrier width.
    Figure descriptionQuantum tunneling through a barrier. A gold rectangular barrier of height U and width L stands between two flat regions, with a dashed horizontal line marking the particle's energy E, which is below U. On the left a blue wave of constant amplitude travels toward the barrier, labeled incident. Inside the barrier a red curve decays exponentially but does not reach zero before the far wall. On the right the wave resumes oscillating, labeled transmitted, at a visibly smaller amplitude than the incident wave. Some probability gets through a wall the particle has too little energy to climb.

    The leakage of the wave function into a forbidden region has a spectacular consequence. Suppose the forbidden region is not infinitely thick but a barrier of finite width \(L\): a wall of potential energy \(U\) higher than the particle’s energy \(E\). The exponentially decaying wave function does not reach zero before the far side of the barrier; a small amplitude survives, and beyond the barrier the wave resumes oscillating. The particle has a real, nonzero chance of appearing on the far side of a wall it classically could never surmount. This is quantum tunneling (Figure 40.6).

    Classically this is impossible: a ball rolled at a hill too tall always rolls back. Quantum mechanically, the wave function reaches into the forbidden region, so the particle has some chance of being found beyond it, and that chance is appreciable only if the barrier is thin.

    We describe the outcome with two probabilities: the transmission coefficient \(T\), the probability the particle gets through, and the reflection coefficient \(R\), the probability it bounces back. Since it must do one or the other,

    \[T + R = 1. \tag{40.14}\]

    When the barrier is thick or tall enough that \(T \ll 1\), an excellent approximation for the transmission coefficient is

    Key Equation: Barrier transmission (tunneling)

    \[T \approx e^{-2CL}, \qquad C = \frac{\sqrt{2m(U - E)}}{\hbar}, \tag{40.15}\]

    with \(L\) the barrier width. The exponential makes \(T\) ferociously sensitive to both the width \(L\) and the energy shortfall \(U - E\): a small change in either can swing the transmission by many orders of magnitude.

    Why it matters. This exponential sensitivity is the working principle of the scanning tunneling microscope (STM), an instrument that images individual atoms. A fine metal tip is held so close to a surface that electrons tunnel across the vacuum gap; because \(T\) depends so steeply on the gap width, even an atom-sized bump changes the tunneling current sharply, and tracking that current maps the surface atom by atom (Figure 40.7). Tunneling also lets alpha particles escape nuclei and lets protons fuse in the Sun’s core despite their mutual repulsion: it is one of the most consequential effects in all of physics.

    Scanning tunneling microscope, side view: a sharp tip above a row of surface atoms.
    Figure 40.7. A scanning tunneling microscope. Electrons tunnel across the vacuum gap between a sharp tip and the surface. Because \(T \approx e^{-2CL}\), the current changes sharply with the gap, so a feedback loop that holds the current constant must raise and lower the tip over each atom. The tip’s path (dashed) is a map of the surface.
    Figure descriptionA scanning tunneling microscope, side view. A row of nine gray circles represents surface atoms. A sharp gray metal tip points straight down at one atom, separated from it by a small vacuum gap marked with a double arrow. Two small green dots in the gap are labeled electrons tunnel across the gap. A blue dashed line runs just above the row of atoms, rising over each atom and dipping between them, and passes through the tip's apex; it is labeled tip path at constant current, with an arrow labeled scan pointing to the right.
    Example 40.5: Tunneling through a thin and a thick barrier

    A \(30\,\mathrm{eV}\) electron is incident on a square barrier of height \(40\,\mathrm{eV}\).

    1. Find the probability that it tunnels through if the barrier is \(1.0\,\mathrm{nm}\) wide.
    2. Find the probability if the barrier is only \(0.10\,\mathrm{nm}\) wide.
    Solution

    The energy shortfall is \(U - E = 40 - 30 = 10\,\mathrm{eV} = 1.602\times10^{-18}\,\mathrm{J}\). The decay constant is

    \[C = \frac{\sqrt{2m_e(U-E)}}{\hbar} = \frac{\sqrt{2(9.11\times10^{-31})(1.602\times10^{-18})}}{1.055\times10^{-34}} = 1.62\times10^{10}\,\mathrm{m^{-1}}.\]

    (a) For \(L = 1.0\times10^{-9}\,\mathrm{m}\), the exponent is \(2CL = 2(1.62\times10^{10})(1.0\times10^{-9}) = 32.4\), so

    \[T \approx e^{-32.4} = \boxed{8.6\times10^{-15}}.\]

    About one electron in \(10^{14}\) gets through: effectively never, for a barrier a nanometer thick.

    (b) Shrinking the barrier tenfold to \(L = 1.0\times10^{-10}\,\mathrm{m}\) cuts the exponent to \(2CL = 3.24\):

    \[T \approx e^{-3.24} = \boxed{0.039}.\]

    Now nearly \(4\%\) of the electrons tunnel through. A tenfold reduction in width raised the transmission by more than twelve orders of magnitude: the very sensitivity that lets a scanning tunneling microscope feel out single atoms.

    Checkpoint 40.5

    A scanning tunneling microscope’s tip drifts very slightly farther from the surface. Using Eq. (40.15), explain qualitatively whether the tunneling current rises or falls, and why even a tiny drift produces a large, easily measured change.

    Answer

    Answers to Checkpoints

    Checkpoint 40.4. In a finite well the wave function does not stop dead at the walls; it leaks a short way into them, decaying exponentially. The electron therefore effectively occupies a region a little wider than \(L\). A wider region fits a standing wave of longer wavelength, and a longer wavelength means a smaller momentum and a lower kinetic energy. So each level sits somewhat below the corresponding level of an infinite well of the same width.

    Checkpoint 40.5. It falls. The probability of tunneling through a barrier of thickness \(L\) goes as \(e^{-2CL}\). Moving the tip farther away makes the gap (the barrier) thicker, so the current decreases. Because the dependence is exponential, with \(C\) of order \(10^{10}\,\mathrm{m^{-1}}\) for typical barriers, a change in gap of a fraction of an atom’s diameter changes the current by a large factor. That sensitivity is what lets an STM map a surface atom by atom.


    This page titled 40.4: Finite Wells was last modified on Thu, 12 Nov 2026 04:27:13 GMT and is shared under a CC BY 4.0 license and was authored, remixed, and/or curated by Michael Ragsdale.