Ramsauer-Townsend Effect

Fire slow electrons through a bottle of argon and something absurd happens: near an energy of about 1 eV the gas turns almost transparent. The total scattering cross section, instead of falling smoothly with energy the way kinetic theory demands, collapses into a sharp minimum and then climbs back up. Electrons at that one energy sail through as if the atoms were barely there. Carl Ramsauer saw it in 1921, and John Townsend with V. A. Bailey saw it independently in 1922; no classical picture of electron–atom collisions can produce such a window. The whole effect lives inside one term of the partial-wave sum. At low energy the centrifugal barrier freezes out every partial wave except $\ell=0$ , so the cross section is essentially $\sigma_0=(4\pi/k^2)\sin^2\delta_0$ . When the s-wave phase shift $\delta_0$ passes through an integer multiple of $\pi$ , that term switches off and the dominant channel goes dark. This is the exact three-dimensional cousin of the one-dimensional unit-transmission law owned by Resonant Transmission and Tunnelling Resonances — which already deeded the noble-gas story to this chapter. This article owns three things. First, the s-wave transparency theorem: why $\delta_0=n\pi$ empties the channel and why only $\ell=0$ matters at threshold. Second, the square-well phase-shift anatomy $\delta_0=-ka+\arctan[(k/k')\tan k'a]$ , derived by matching. Third — the machine crown — a model-atom laboratory that puts a realistic polarization tail on a square core, tunes it to the measured e–Xe scattering length, and lets the Ramsauer minimum emerge near 1 eV rather than be fitted; it then computes the honest p-wave floor that keeps the minimum deep but not exactly zero. We recall the partial-wave sum, phase-shift machinery, and low-energy scattering-length grammar by name from their owners; the actual dip we build and measure here.

Key takeaways

Slow electrons pass almost freely through argon near 1 eV because the s-wave phase shift $\delta_0$ hits $\pi$ , and $\sigma_0\propto\sin^2\delta_0$ vanishes. At threshold only $\ell=0$ survives the centrifugal barrier, so emptying the s-wave nearly empties the cross section. It needs a deep, attractive well — argon, krypton, xenon (large polarizabilities, negative scattering lengths) qualify; helium and neon (repulsive exchange cores, positive scattering lengths) do not, so they show no dip. A model well tuned to the real e–Xe scattering length reproduces the minimum at $\sim1$ eV, and the residual p-wave floor makes the dip deep, not perfectly transparent.

The core insight

A cross section built from $\sin^2\delta_\ell$ has zeros, not just minima. When the low-energy world is s-wave-only and $\delta_0$ sweeps through $\pi$ , the target becomes a hole in energy space — an interference null you can buy in a gas bottle.

Where this lives today

The Ramsauer–Townsend minimum is not a museum piece. Electron transport in gas-filled radiation detectors and in low-temperature plasmas depends sensitively on the argon/xenon cross-section minimum, because near 1 eV the electron mean free path becomes anomalously long — proportional counters and time-projection chambers are engineered around it. The same physics governs electron mobility in noble-gas discharges and the design of gas mixtures for switching and lighting. And the s-wave transparency has a direct cold-atom cousin: a Feshbach-tuned two-body scattering length can be driven through zero, making an ultracold gas transparent to its own collisions — the atomic analogue of the argon window. Realistic electron–atom modelling keeps exactly the polarization tail $-\alpha/2r^4$ we used, which fixes the near-threshold behaviour and the exact minimum energy through the O'Malley threshold laws.

The physical picture

Classically a slow electron should scatter more as it slows down: it spends longer near the atom, deflects more, and the cross section rises monotonically toward threshold. The data do the opposite in the noble gases — a dip. The resolution is that a quantum cross section is not a geometric shadow; it is $\sin^2$ of a phase, and phases can return to multiples of $\pi$ where the sine vanishes.

Figure

The s-wave channel switching off: as δ₀ sweeps through π, sin²δ₀ returns to zero and the dominant low-energy channel goes dark

Two things must line up for the window to open at an accessible energy. The potential must be attractive and deep enough that $\delta_0$ actually climbs to a multiple of $\pi$ as the electron energy rises from threshold — a repulsive or shallow interaction never gets there. And the low-energy world must be genuinely s-wave-dominated so that killing $\ell=0$ kills most of the cross section. The polarizable, closed-shell heavy noble atoms present exactly such a deep attractive well to a slow electron — reflected in their large negative scattering lengths. Helium and neon present a repulsive exchange core with a small positive scattering length, which is the entire content of the Ar/Kr/Xe-yes, He/Ne-no rule.

Figure

Punchline — σ(E) for a Xe-like well vs a He-like well on one plot: the deep argon-class dip near 1 eV with its small p-wave floor, and the helium curve with no dip at all

Why only the s-wave matters at threshold

The total cross section is the partial-wave sum owned by Partial-Wave Expansion, recalled here as a working tool. Before we can talk about a single channel switching off, we need the fact that at low energy there is essentially only one channel. The centrifugal barrier $\ell(\ell+1)\hbar^2/2mr^2$ holds higher partial waves away from a short-range potential, and their phase shifts vanish as a power of $k$ .

\sigma=\frac{4\pi}{k^2}\sum_{\ell=0}^{\infty}(2\ell+1)\sin^2\delta_\ell

where: $k=\sqrt{2mE}/\hbar$ is the electron wavenumber, $\delta_\ell(k)$ is the phase shift of the $\ell$ -th partial wave, and the factor $(2\ell+1)$ counts magnetic substates. Physically each term is the area the target presents to that angular-momentum channel. For a potential of range $R$ , the low-energy phase shifts obey $\delta_\ell\sim(kR)^{2\ell+1}$ — the threshold law owned by Scattering Length and Low-Energy Scattering. So $\delta_1$ is suppressed by $(kR)^3$ relative to $\delta_0$ , $\delta_2$ by $(kR)^5$ , and so on. At $ka\lesssim1$ the sum is s-wave to excellent accuracy:

\sigma\;\xrightarrow{\;ka\ll1\;}\;\sigma_0=\frac{4\pi}{k^2}\sin^2\delta_0

where: this single term now carries the whole cross section. The consequence is immediate and is the thesis of the article: a zero of $\sin\delta_0$ is a zero of the cross section. When $\delta_0$ passes through $0,\pi,2\pi,\dots$ , the gas becomes transparent.

The square-well phase shift, by matching

To see when $\delta_0$ reaches $\pi$ we need its actual functional form. Start with the cleanest solvable model — an attractive square well of depth $V_0$ and radius $a$ . The reduction $u=r\psi$ to a one-dimensional radial problem is the machinery of Phase Shifts and the Radial Equation; here we solve the specific $\ell=0$ attractive-well instance and read off $\delta_0$ .

Derivation AThe s-wave phase shift of an attractive square well

We want the closed form $\delta_0=-ka+\arctan[(k/k')\tan k'a]$ from first principles, so that the transparency condition $\delta_0=n\pi$ becomes a concrete equation in the well parameters. The plan is to solve the radial equation inside and outside the well, match the logarithmic derivative at the edge, and isolate $\delta_0$ . What to watch: the interior wavenumber $k'$ is larger than $k$ because the well is attractive, and it is the ratio $k/k'$ that controls how far the phase can swing.

Step 1. Radial equations for $u=r\psi$ . For $\ell=0$ the Schrödinger equation becomes a 1D equation for $u(r)$ with $u(0)=0$ :

u''+k'^2u=0\ (r<a),\qquad u''+k^2u=0\ (r>a)

where $k'=\sqrt{2m(E+V_0)}/\hbar$ inside and $k=\sqrt{2mE}/\hbar$ outside. Why $u(0)=0$ : regularity of $\psi=u/r$ at the origin.

Step 2. Solutions. The regular interior solution and the asymptotically phase-shifted exterior solution are

u_{\rm in}=A\sin k'r,\qquad u_{\rm out}=C\sin(kr+\delta_0)

Why this form: $\sin k'r$ is the only solution vanishing at $r=0$ ; $\sin(kr+\delta_0)$ is the free s-wave with its phase displaced by exactly the quantity we want.

Step 3. Match the logarithmic derivative at $r=a$ . Continuity of $u$ and $u'$ means $u'/u$ is continuous, which removes the amplitudes:

k'\cot k'a = k\cot(ka+\delta_0)

Why the log-derivative: it is the amplitude-independent, physically meaningful matching quantity — the same object that fixes bound-state energies.

Step 4. Solve for $\delta_0$ . Invert the cotangent:

\boxed{\ \delta_0=-ka+\arctan\!\Big[\tfrac{k}{k'}\tan k'a\Big]\ }

Why the $-ka$ : it is the free-particle phase you would have with no well at all (a hard point would give $\delta_0=-ka$ ); the arctan is the well's contribution. As $E\to0$ Levinson's theorem fixes the branch: $\delta_0(0)=n_b\pi$ , where $n_b$ is the number of s-wave bound states the well supports. Verified: the boxed closed form is reproduced by an independent variable-phase (Calogero) integration of the same square well to $<10^{-6}$ (mod $\pi$ ) at three wavenumbers; verify script check 1.

Two readings follow immediately. The transparency needs $\delta_0$ to reach a multiple of $\pi$ beyond its Levinson starting value at a finite, accessible energy — so the interaction must be attractive and deep enough. And the phase must swing by roughly $\pi$ over the low-energy range. The connection between the transparency and the bound-state count is a Levinson reading; the pole structure that makes those bound states is owned by Bound States as Poles of the Scattering Amplitude.

The model-atom laboratory

Now the crown. A bare square well is a cartoon; a real atom presents a long-range polarization potential $-\alpha/2r^4$ to a slow electron (the electron polarizes the atom, which pulls back). We add that tail to a square core, tune the model to the measured e–Xe scattering length, and let the Ramsauer dip emerge — a physical prediction, not a fit to the dip energy.

Go deeper · Derivation B · 1 minThe model-atom laboratory: the dip that emerges, and its honest floor

Derivation BThe model-atom laboratory: the dip that emerges, and its honest floor

The goal is a falsifiable numerical exhibit: an electron–xenon model tuned only to the known scattering length, whose s-wave phase shift then passes through $\pi$ near 1 eV on its own — reproducing the argon-class transparency — together with the honest p-wave floor. The strategy uses the variable-phase (Calogero) method, which integrates the absolute phase shift outward and builds in the Levinson branch automatically. What to watch: we fix the model by $a_s$ alone; the dip energy is a consequence.

Step 1. The potential and its one knob. Square core of radius $1.0\,\text{Å}$ plus the real xenon polarization tail:

V(r)=\begin{cases}-V_0,&r<a_{\rm core}\\[2pt]-\alpha/2r^4,&r>a_{\rm core}\end{cases}\qquad \alpha_{\rm Xe}=27.29\,a_0^3

Why $\alpha$ is fixed, not free: it is the measured static dipole polarizability of xenon; only $V_0$ is tuned.

Step 2. Tune to the measured scattering length. Adjust $V_0$ so the zero-energy phase gives the literature value $a_s=-6.5\,a_0$ for slow electrons on xenon:

V_0=30.16\,\text{eV}\ \Rightarrow\ a_s=-6.50\,a_0

Why negative: a deep attractive well over-binds the low-energy wave, pulling the scattering length negative — the signature of the noble-gas attraction.

Step 3. The dip emerges. With the model now fixed, integrate $\delta_0(E)$ . It passes through $\pi$ at

E^*=0.963\,\text{eV},\qquad k^*a_{\rm core}=0.503

Why this is a prediction: we never tuned the dip energy — it fell out of matching the scattering length, and it lands right where argon's Ramsauer minimum sits. The wavenumber $k^*a_{\rm core}=0.503$ confirms we are deep in the s-wave world.

Step 4. The honest p-wave floor. At the dip the s-wave cross section is zero to machine precision, but $\ell=1$ contributes

\sigma_1(E^*)=\frac{12\pi}{k^2}\sin^2\delta_1=2.22\,a_0^2 \quad\text{vs}\quad 4\pi a_s^2=530.9\,a_0^2

Why the dip is so deep: the p-wave floor is only $0.4\%$ of the threshold cross section — the polarization tail obeys the O'Malley threshold law $\tan\delta_1\to\pi\alpha k^2/15$ , keeping $\delta_1$ small at the dip energy. The transparency is dramatic — a $234\times$ drop from threshold to dip — but not perfect: the minimum is deep, not zero. Verified: verify script §2–5 — $a_s=-6.50\,a_0$ by Brent root-find, the emergent dip $E^*=0.963$ eV, the p-wave floor $2.22\,a_0^2$ , and the O'Malley law ( $\tan\delta_1/[\pi\alpha k^2/15]\to1$ monotonically as $k\to0$ ) all recomputed; the dip's quadratic profile confirmed; 22-check suite green.

The 3D transparency at $\delta_0=\pi$ is the exact analogue of the 1D unit-transmission law of Resonant Transmission and Tunnelling Resonances, whose overview opens with Ramsauer's 1921 argon measurement and which already stated the deed we cash here: heavier noble gases (Ar, Kr, Xe) have wells deep enough to do it; helium and neon do not — which is why they show no Ramsauer minimum. Our He-like model makes the point numerically: helium presents a repulsive exchange core (its closed shell resists an intruding electron by the exclusion principle) with a small positive scattering length $a_s=+1.2\,a_0$ . Tuned to that value, $\delta_0(0)=0$ (no bound state) and $\delta_0$ stays small and negative, never reaching $\pi$ . No dip.

Limiting cases

regime	behaviour	why
$ka\to0$ , deep attractive well	$\sigma\to4\pi a_s^2$	s-wave only; large negative $a_s$ sets a big plateau
$\delta_0\to n\pi$	$\sigma_0\to0$	the transparency window; total floored by $\sigma_1$
repulsive/shallow ( $a_s>0$ )	no dip	$\delta_0$ never reaches a multiple of $\pi$
$ka\gg1$	many partial waves	s-wave zero irrelevant; classical shadow limit
$E\to\infty$	monotone decrease resumes	high- $\ell$ average washes out the s-wave zero

Worked examples

Example 1 — The dip emerges, and how deep it is. Take the Xe model: core $1.0\,\text{Å}$ , polarization tail $\alpha_{\rm Xe}=27.29\,a_0^3$ , depth tuned to $a_s=-6.5\,a_0$ ( $V_0=30.16$ eV).

Step 1. Confirm the tuning. At $k=4\times10^{-3}\,a_0^{-1}$ the zero-energy phase gives $a_s=-\tan\delta_0/k=\mathbf{-6.50\ a_0}$ — the measured e–Xe value.

Step 2. Find the transparency. Integrating $\delta_0(E)$ , the phase reaches $\pi$ at $E^*=\mathbf{0.963\ eV}$ , with $k^*a_{\rm core}=0.503$ — squarely in the s-wave world, and right where argon's Ramsauer dip sits. We did not tune this energy; it emerged.

Step 3. The contrast. At the dip $\sigma(E^*)=\mathbf{0.62\ Å^2}$ ; near threshold $\sigma=145.4\,\text{Å}^2$ . The gas is $\mathbf{234\times}$ more transparent at the dip than just above threshold — the residual $2.2\,a_0^2$ is almost all p-wave.

Example 2 — Mean free path in the window. How far does a 1-eV electron travel through xenon at 1 Torr ( $n=3.22\times10^{22}\,\text{m}^{-3}$ )?

Step 1. Mean free path. $\lambda=1/(n\sigma)$ . At the dip, $\sigma=0.62\,\text{Å}^2=6.2\times10^{-21}\,\text{m}^2$ , so $\lambda=1/(3.22\times10^{22}\times6.2\times10^{-21})\approx\mathbf{0.5\ cm}$ .

Step 2. Compare to threshold. Just above threshold $\sigma\approx145\,\text{Å}^2$ , giving $\lambda\approx\mathbf{0.02\ mm}$ — 250 times shorter.

Step 3. The observable. An electron entering the gas at the dip energy travels centimetres before its first collision, while the same electron a fraction of an eV away is stopped in microns. That two-hundred-fold jump in mean free path is the Ramsauer–Townsend transparency, directly visible in swarm-transport data.

Origins · 1 minJohn Sealy Townsend and how this was discovered

John Sealy Townsend

1868–1957 · Galway, Cambridge, Oxford

Townsend was one of J. J. Thomson's first research students at the Cavendish, where in the late 1890s he made the earliest measurements of the electron charge by watching charged clouds fall — the germ of the oil-drop method. His life's work was the physics of electrons drifting and multiplying in gases: the Townsend discharge and the avalanche coefficient that bear his name underpin every gas detector and spark counter. In 1922, by then Wykeham Professor at Oxford, Townsend and V. A. Bailey measured the motion of slow electrons in argon and found the anomalous transparency — independently of Carl Ramsauer, whose 1921 Danzig beam experiments had seen the same effect first. The joint name records a genuine co-discovery from opposite experimental traditions: Ramsauer's collimated beams and Townsend's swarm-diffusion method.

History

1921

— Carl Ramsauer measures the cross section of slow electrons in argon (Danzig) and finds a deep minimum near 1 eV

1922

— Townsend and Bailey, using electron swarms in argon at Oxford, find the same transparency independently

1927

— Faxén and Holtsmark decompose the slow-electron amplitude into partial waves and phase shifts, explaining the effect

1930

— the effect is recognized as a direct manifestation of quantum phase shifts, unexplained by any classical model

1933

— Mott and Massey's Theory of Atomic Collisions codifies the partial-wave account of the Ramsauer–Townsend transparency

Going further

The mechanism sits at the intersection of one-dimensional resonant transmission (the $T=1$ law of Resonant Transmission) and three-dimensional partial-wave scattering. For the low-energy grammar — scattering length, effective range, the $(kR)^{2\ell+1}$ threshold law — see Scattering Length and Low-Energy Scattering; for the phase-shift machinery and the $u=r\psi$ radial reduction, Phase Shifts and the Radial Equation; for the partial-wave sum itself, Partial-Wave Expansion. The bound-state/Levinson connection is developed in Bound States as Poles of the Scattering Amplitude. Standard treatments: Griffiths §11.4, Sakurai & Napolitano Ch. 6, Cohen-Tannoudji Vol. II Ch. VIII. Provenance note (base corpus): A verifier should recompute: (1) the total cross section $\sigma=(4\pi/k^2)\sum(2\ell+1)\sin^2\delta_\ell$ from the partial-wave amplitude (Griffiths Sec. 11.4, Sakurai Ch. 6); (2) the 1D square-well transmission $T=[1+\frac{V_0^2}{4E(E+V_0)}\sin^2(2k'a)]^{-1}$ with resonance $2k'a=n\pi$ (Griffiths Ch. 2); (3) the 3D attractive-square-well s-wave phase shift $\delta_0=-ka+\arctan[(k/k')\tan k'a]$ and its zero-energy scattering length (Sakurai/Cohen-Tannoudji Vol. II); (4) Levinson's theorem $\delta_0(0)=n_b\pi$ (Mott & Massey). The experimental argon minimum near $\sim1\,\mathrm{eV}$ is a textbook-reported value (Ramsauer 1921; Townsend & Bailey 1922) cited qualitatively; no fabricated numerical cross-section values are quoted.

If the s-wave cross section drops to zero, why isn't the gas perfectly transparent?

Because "s-wave" is not "total." By the time $\delta_0$ reaches $\pi$ (here $ka\approx0.5$ ), the p-wave phase shift has grown enough to contribute a small cross section — in our model $2.2\,a_0^2$ , about $0.4\%$ of the threshold value. The transparency is a switching-off of the dominant channel (a $234\times$ drop), not an annihilation of all scattering.

Why put a polarization tail on the model instead of a bare square well?

Because it turns a fit into a prediction. A bare square well can be tuned to place the dip at 1 eV, but that is circular. Adding the measured polarizability $-\alpha/2r^4$ and tuning only to the measured scattering length, the dip then emerges at 0.96 eV on its own — the model is constrained by data that has nothing to do with the dip, and the dip still comes out right. That is the difference between reproducing and explaining.

How does the scattering length tell helium and argon apart?

Sign. Argon, krypton and xenon have large negative scattering lengths (deep attractive wells that over-bind the low-energy wave), which is exactly the condition for $\delta_0$ to swing up through $\pi$ . Helium and neon have small positive scattering lengths (their closed-shell exchange repulsion dominates), and a positive $a_s$ means $\delta_0$ starts at zero and stays small — it never reaches $\pi$ , so there is no dip.

Is the Ramsauer effect a resonance?

It is the mirror of a resonance. A resonance is a peak where $\delta_\ell$ passes rapidly through $\pi/2$ and $\sin^2\delta_\ell\to1$ ; the Ramsauer minimum is a dip where $\delta_0$ passes through $\pi$ (or $0$ ) and $\sin^2\delta_0\to0$ . Same phase-shift language, opposite extremum. Resonance shapes proper are owned by Breit–Wigner Resonances.

Why does classical kinetic theory get the trend exactly backwards?

Classically, a slower electron dwells longer in the potential and deflects more, so the cross section should rise toward threshold. That reasoning treats scattering as accumulated deflection. Quantum mechanically the cross section is $\sin^2$ of an accumulated phase, and a phase can wrap back through a multiple of $\pi$ where the amplitude vanishes. There is no classical quantity that behaves like $\sin^2\delta_0$ .

Do krypton and xenon show the effect at the same energy as argon?

They show it at somewhat different low energies (all in the rough $0.3$ – $1$ eV band), because the effective depth and polarizability differ. The heavier, more polarizable atoms have the deeper wells and larger negative scattering lengths that make the effect prominent. The Ar/Kr/Xe-yes, He/Ne-no rule is about sign and depth, not a shared energy.

Does the effect require quantum indistinguishability of the electrons?

No. This is single-particle potential scattering of one electron off a fixed atom; the transparency is entirely about the s-wave phase shift, not about exchange symmetry between projectile and any target electron. Identical-particle interference (adding amplitudes for exchanged histories) is a different effect, owned by Identical-Particle Scattering.