Induced Dipoles and Atomic Polarizability

A neutral atom should be electrically boring. Zero net charge, spherically symmetric electron cloud, zero dipole moment — every multipole moment vanishes, and the far-field machinery of the previous chapter says: no field, no story. Yet neutral argon atoms are deflected by charged rods, neutral dust sticks to charged screens, and a focused laser beam holds a neutral atom in place like a pair of tweezers. The resolution is that "zero dipole moment" is a statement about the atom left alone. Put it in a field and the field pulls the nucleus one way and the electron cloud the other; the two centers separate until the atom's own internal attraction balances the pull, and the atom sits there slightly stretched — a dipole made to order, pointing along the field that made it. This is the first of the two microscopic engines of dielectric response promised by the chapter opener, and the quantitative question is a single coefficient: for the small stretches that any laboratory field produces, the induced moment is strictly proportional, $\mathbf{p} = \alpha\mathbf{E}$ , and everything about the atom is packed into its polarizability $\alpha$ . We derive $\alpha$ for the classic crude model — a nucleus inside a uniform electron ball — and get the wonderfully simple answer $\alpha = 4\pi\varepsilon_0 a^3$ : polarizability is atomic volume, in disguise. The crude model earns its keep honestly (right size, right scaling, wrong factor — quantum mechanics supplies $4.5\,a_0^3$ for hydrogen where the ball model says $1.0$ ), and it connects straight back to a result we own from the images chapter: a conducting sphere has exactly $\alpha = 4\pi\varepsilon_0R^3$ — an atom is a leaky, spring-loaded version of that perfect screener. The energy of the stretch, $-\tfrac12\alpha E^2$ with its telltale factor of $\tfrac12$ , closes the article — it is the operating principle of the optical tweezer.

Key takeaways

A field separates a neutral atom's nucleus from its electron cloud until internal attraction balances the pull: an induced dipole $\mathbf{p} = \alpha\mathbf{E}$ , aligned with the field and proportional to it.
The crude model (point nucleus in a uniform electron ball of radius $a$ ) gives $\alpha = 4\pi\varepsilon_0 a^3$ — polarizability is essentially atomic volume, which is why big soft atoms polarize more.
Chemists' units make this vivid: $\alpha' \equiv \alpha/4\pi\varepsilon_0$ has units of volume — hydrogen $0.667\,\text{Å}^3$ , argon $1.64\,\text{Å}^3$ (quantum mechanics gives hydrogen exactly $4.5\,a_0^3$ ; the ball model's $a_0^3$ is the right scale, wrong factor).
The stretch is astonishingly small: at air's breakdown field, hydrogen's charge centers separate by $\sim 10^{-15}\,\mathrm{m}$ — a nuclear diameter — which is why the response is so perfectly linear.
A conducting sphere is the $\alpha = 4\pi\varepsilon_0R^3$ limit of the same physics (recalled from the image-charge chapter): perfect screening instead of spring-loaded partial screening.
The energy of an induced dipole is $U = -\tfrac12\alpha E^2$ — half the permanent-dipole value, because the moment had to be built as the field was applied — and its gradient force pulls polarizable matter toward strong field: the optical tweezer.
KEY IDEA: an atom in a field stretches into a dipole $\mathbf{p} = \alpha\mathbf{E}$ , with $\alpha \approx 4\pi\varepsilon_0 a^3$ set by atomic volume — the microscopic origin of induced polarization in every dielectric.

The core insight

An atom's induced dipole moment is linear in the applied field, $\mathbf{p}=\alpha\mathbf{E}$ ; the polarizability $\alpha\approx 4\pi\varepsilon_0 a^3$ is set by atomic size, and it is the microscopic origin of dielectric behavior.

Where this lives today

Polarizability is one of the most-tabulated numbers in physical science because it is the handle by which neutral matter is gripped. Optical tweezers — Ashkin's Nobel-winning generalization of worked example 2 to micron-scale beads, where $\alpha$ scales up with volume exactly as the ball model predicts — now measure piconewton forces inside living cells and stretch single DNA molecules. Optical lattices, standing waves whose $\langle E^2\rangle$ landscape traps ultracold atoms at the antinodes (or nodes, above resonance where $\alpha$ flips sign), are the substrate of today's atomic clocks and quantum simulators: every site of the lattice is an induced-dipole energy well. The same coefficient runs quieter machinery. London's dispersion force — the universal attraction between any two neutral atoms — is two induced dipoles inducing each other, with strength governed by the product of polarizabilities; it is why noble gases liquefy at all, and its $\alpha$ -scaling explains why xenon boils $160\,\mathrm{K}$ warmer than helium. Dielectrophoresis sorts living cells in microfluidic chips by their effective $\alpha$ contrast with the surrounding fluid, gripping them with $\nabla(E^2)$ exactly as Step 3 of the energy derivation prescribes. And in quantum chemistry, computed polarizability tensors are a standard benchmark of electronic-structure methods — the modern, many-decimal descendants of the uniform ball.

Origins · 1 minJoseph John Thomson and how this was discovered

Joseph John Thomson

1856–1940 · Cavendish Laboratory, Cambridge

J. J. Thomson discovered the electron in 1897 — cathode rays bent by fields, with a charge-to-mass ratio a thousandfold larger than any ion's — and thereby created the problem this article solves: if atoms contain movable charged parts, how do neutral atoms respond to fields? His 1904 "plum pudding" atom — electrons embedded in a uniform sphere of positive charge — was proposed as a real structural model and famously demolished by his own student Rutherford in 1911, when alpha-scattering revealed the nucleus. But inverted (point nucleus in a diffuse negative cloud), Thomson's uniform ball survives as the crude model of polarizability, the standard first calculation of $\alpha$ in every electrodynamics course — wrong as anatomy, enduring as a spring model whose interior linearity captures why dielectric response is linear. Thomson's Cavendish tenure was a dynasty: seven of his research assistants and his son won Nobel prizes, and his own 1906 prize for the electron was mirrored, with a symmetry physics students savor, by George Thomson's 1937 prize for showing the same electron is a wave.

History

1897

— Thomson isolates the electron: atoms have movable charged constituents, so fields can distort them

1904

— Thomson's uniform-sphere atom supplies the ball model whose spring constant yields alpha = 4 pi epsilon0 a cubed

1911

— Rutherford's nucleus kills the pudding as anatomy; the ball survives as the crude polarizability model

1926

— Wave mechanics delivers hydrogen's exact static polarizability, 4.5 Bohr-radii cubed, via perturbation theory

1930

— London derives the dispersion force: mutual induced dipoles attract all neutral matter, strength set by alpha

1986

— Ashkin's single-beam gradient trap (optical tweezers) turns minus one-half alpha E squared into a laboratory tool

The physical picture

Model the atom the way J. J. Thomson did in 1904: a point nucleus $+q$ sitting at the center of a uniform ball of electron charge $-q$ and radius $a$ . Switch on a field $\mathbf{E}$ . The nucleus is pushed along the field, the cloud pushed against it; the two centers separate by some distance $d$ . But the moment they separate, the atom fights back: a nucleus off-center inside a uniform negative ball feels a restoring attraction toward the center — and, crucially, one that grows in exact proportion to the displacement. The atom is a spring. Equilibrium is reached when the internal spring force balances the external pull, and since the spring is linear, the equilibrium stretch — and hence the induced moment $p = qd$ — is linear in $E$ . That is the entire content of $\mathbf{p} = \alpha\mathbf{E}$ ; the derivation below merely computes the spring constant.

Figure

The stretched atom — an induced dipole in equilibrium

Two consequences of the spring picture deserve headlines. First, the size of the effect: the internal fields holding atoms together are $\sim 5\times10^{11}\,\mathrm{V/m}$ (the scale audited in the chapter opener), so even a breakdown-strength laboratory field stretches the atom by parts per million of its radius — about a femtometer, literally the diameter of the nucleus. The dipoles that run all of dielectric physics are built from displacements smaller than anything a chemist would call a distortion. Linearity is not an approximation so much as an inevitability. Second, the scaling: the answer $\alpha = 4\pi\varepsilon_0 a^3$ says polarizability is set by atomic volume — soft, puffy, weakly-bound electron clouds polarize strongly. That single fact organizes the periodic table's response to fields: helium (small, tightly bound) has $\alpha' = 0.205\,\text{Å}^3$ ; cesium (one loose outer electron around a large core) reaches $59\,\text{Å}^3$ . And it connects to a theorem we already own: Image Charges for a Conducting Sphere proved that a conducting sphere of radius $R$ responds to a distant charge exactly as a point dipole with $\alpha = 4\pi\varepsilon_0R^3$ . The atom's ball-model answer is identical in form — because a conductor is simply the perfectly compliant limit of the same physics, screening completely where the atom's spring allows only a partial stretch. That the two calculations agree is not a coincidence; it is the first hint that "dielectric constant" and "conductor" sit on one continuous ladder.

The polarizability

The defining relation, for fields small compared to the atomic scale (i.e., all of them):

\mathbf{p} = \alpha\,\mathbf{E} \qquad (1)

where: $\mathbf{p}$ — the induced dipole moment (C·m); $\mathbf{E}$ — the field at the atom; $\alpha$ — the atomic polarizability (C·m²/V, equivalently F·m²). For a spherically symmetric atom $\alpha$ is a scalar and $\mathbf{p}$ aligns with $\mathbf{E}$ exactly; for molecules with structure (CO $_2$ polarizes almost twice as readily along its axis as across it) $\alpha$ becomes a symmetric tensor, $p_i = \alpha_{ij}E_j$ , and $\mathbf{p}$ tilts away from $\mathbf{E}$ — the anisotropic story is deferred by name to Anisotropic Dielectrics and the Permittivity Tensor.

Figure

Polarizability volumes across the periodic table — soft clouds polarize

Because $\alpha$ always appears in the combination $\alpha/4\pi\varepsilon_0$ , tables quote that quantity instead — and it has units of volume:

\alpha' \equiv \frac{\alpha}{4\pi\varepsilon_0}, \qquad \alpha'_{\rm H} = 0.667\,\text{Å}^3, \quad \alpha'_{\rm He} = 0.205\,\text{Å}^3, \quad \alpha'_{\rm Ar} = 1.64\,\text{Å}^3 \qquad (2)

where: $\alpha'$ — the polarizability volume (Griffiths Table 4.1 values). The ball model below explains why these numbers sit within a factor of a few of actual atomic volumes, and the trend — He < H < Ar, tight clouds less polarizable than loose ones — is the volume scaling in action.

Derivation AThe crude model — a nucleus in a uniform electron ball

The goal is $\alpha = 4\pi\varepsilon_0 a^3$ for the Thomson-style atom: point nucleus $+q$ , uniform ball of charge $-q$ , radius $a$ . The strategy is force balance: displace the nucleus by $d$ relative to the cloud's center, compute the restoring field it feels from the cloud, and set the external pull against it. The key input is the interior field of a uniform ball — a one-line recall from Gauss's Law: Applications and Symmetry, which is what makes this a two-step derivation rather than an integral. Watch for the moment linearity appears: it is a gift of the uniform ball's interior, not an added assumption.

Step 1. The restoring field. By the Gauss symmetry catalog, the field inside a uniform ball of total charge $-q$ at distance $d$ from its center is produced by the charge interior to $d$ alone, and grows linearly:

E_{\text{cloud}}(d) = \frac{1}{4\pi\varepsilon_0}\,\frac{(-q)\,d}{a^3}.

A displaced nucleus therefore feels a pull back toward the center of magnitude $qE_{\text{cloud}} = q^2d/4\pi\varepsilon_0a^3$ — a Hooke's-law spring with stiffness $k_{\rm spring} = q^2/4\pi\varepsilon_0 a^3$ . Why linear: interior to a uniform ball, enclosed charge grows as $d^3$ while Coulomb's law falls as $1/d^2$ ; the product is proportional to $d$ . The same interior-linearity powered the tunnel-through-the-Earth oscillator in mechanics; here it powers all of linear dielectrics.

Step 2. Balance against the applied field. In equilibrium the nucleus is stationary: $qE = q^2d/4\pi\varepsilon_0a^3$ , so the stretch is $d = 4\pi\varepsilon_0a^3E/q$ , and the induced moment is

p = qd = 4\pi\varepsilon_0 a^3\,E \qquad\Longrightarrow\qquad \alpha = 4\pi\varepsilon_0\,a^3, \quad \alpha' = a^3.

The atom's polarizability volume is (the cube of) its radius. No adjustable constants survive.

Step 3. Audit against reality. For hydrogen, take $a = a_0 = 0.529\,\text{Å}$ : the model predicts $\alpha' = a_0^3 = 0.148\,\text{Å}^3$ ; experiment (and exact quantum perturbation theory) gives $\alpha'_{\rm H} = 4.5\,a_0^3 = 0.667\,\text{Å}^3$ . Right order of magnitude, right volume scaling, factor $4.5$ too stiff — the real electron cloud is exponentially fuzzy-edged and yields more easily than a hard-edged ball. That is the correct way to hold a crude model: trust its scalings, not its coefficients.

Step 4. The conductor check. Replace the spring-loaded cloud by a perfectly conducting sphere of radius $R$ : the far-field response, computed by images in Image Charges for a Conducting Sphere, was a dipole with $\alpha = 4\pi\varepsilon_0R^3$ — the same formula. The conductor is the infinitely compliant atom: its "cloud" rearranges until the interior field vanishes entirely, and remarkably the ball model, which only partially screens its interior, lands on the identical polarizability. (The agreement is a peculiarity of the uniform ball; the lesson that survives generally is $\alpha \sim 4\pi\varepsilon_0\times\text{volume}$ .) Verified: Griffiths Ex. 4.1 (the model and its factor-of-4.5 honesty); interior field of the ball per GAUSA; the stretch at E = 3 MV/m is d = αE/e = 1.4×10⁻¹⁵ m for real hydrogen — parts per hundred thousand of a0, confirming the linear-zone assumption a posteriori.

The number that follows from the balance deserves its own line, because it is the article's quiet punchline. In a field of $3\times10^6\,\mathrm{V/m}$ — the strongest air sustains — hydrogen's charge centers separate by

d = \frac{\alpha E}{e} = \frac{(7.42\times10^{-41})(3\times10^{6})}{1.60\times10^{-19}} \approx 1.4\times10^{-15}\,\mathrm{m} \qquad (3)

where: we used the measured $\alpha_{\rm H} = 4\pi\varepsilon_0\times0.667\,\text{Å}^3 = 7.4\times10^{-41}$ F·m². A femtometer — the atom's dipole is built from a displacement the size of its own nucleus. Multiply such moments by $10^{25}$ – $10^{28}$ atoms per cubic meter, and you get the macroscopic polarization the rest of the chapter is built on; the census argument from the chapter opener, now with its microscopic numerator computed.

Go deeper · Derivation B · 2 minThe energy of an induced dipole — where the one-half comes from

Derivation BThe energy of an induced dipole — where the one-half comes from

The goal is $U = -\tfrac12\alpha E^2$ for an induced dipole, and specifically the factor $\tfrac12$ that distinguishes it from the permanent dipole's $U = -\mathbf{p}\cdot\mathbf{E}$ (derived in Torque and Force on a Dipole in a Field). The strategy is to charge the field up from zero and integrate the work, crediting honestly the energy spent stretching the atom's internal spring. This half is not bookkeeping pedantry: it halves every induced-dipole force in physics, from optical tweezers to van der Waals attractions, and forgetting it is the standard exam and research error alike.

Step 1. Raise the field from $0$ to $E$ in increments $\mathrm{d}E'$ . At intermediate field $E'$ the atom carries moment $p' = \alpha E'$ . The interaction energy of a dipole in a field changes by $\mathrm{d}U_{\rm int} = -p'\,\mathrm{d}E'$ as the field grows (the alignment energy deepens as both the field and the moment it induces grow together):

U_{\rm int} = -\int_0^E \alpha E'\,\mathrm{d}E' = -\tfrac12\,\alpha E^2 \;+\; \big(\text{spring energy accounting, Step 2}\big).

Step 2. Audit the books the long way, where the $\tfrac12$ is explicit. At full field the dipole's naive alignment energy is $-pE = -\alpha E^2$ . But building the moment stretched the internal spring, storing $+\tfrac12 k_{\rm spring}d^2 = +\tfrac12(qd)(E) = +\tfrac12\alpha E^2$ (using the equilibrium condition $k_{\rm spring}d = qE$ ). Net:

U = -\alpha E^2 + \tfrac12\alpha E^2 = -\tfrac12\,\alpha E^2.

Why permanent dipoles keep the full $-pE$ : their moment pre-exists — no spring was stretched to make it — so nothing is withheld. Half of an induced dipole's alignment payoff is always mortgaged to its own creation.

Step 3. The force this energy implies. In a non-uniform field, $\mathbf{F} = -\nabla U = \tfrac12\alpha\nabla(E^2)$ : induced dipoles are pulled toward strong field, regardless of the field's direction (the energy depends on $E^2$ ). This is the gradient force — the reason neutral dust migrates to charged surfaces and the reason a tightly focused laser, whose intensity peaks at the focus, traps a neutral particle at the focus. For oscillating fields one replaces $E^2$ by its time average $\langle E^2\rangle = E_{\rm pk}^2/2$ (valid at frequencies below the atom's resonances, where $\alpha$ retains its static sign).

Step 4. Put in tweezer numbers. A $1\,\mathrm{\mu m}$ -waist infrared beam at intensity $10^{12}\,\mathrm{W/m^2}$ carries peak field $E_{\rm pk} = \sqrt{2I/\varepsilon_0 c} = 2.7\times10^{7}\,\mathrm{V/m}$ . Trap depth for an argon atom: $U = \tfrac14\alpha_{\rm Ar}E_{\rm pk}^2 = 3.4\times10^{-26}\,\mathrm{J} \approx k_B\times2.5\,\mathrm{mK}$ . Millikelvin-deep — irrelevant to room-temperature atoms ( $k_BT = 4\times10^{-21}\,\mathrm{J}$ , five orders deeper than the trap), decisive for laser-cooled atoms at microkelvin. Optical dipole traps work only because atomic physics first learned to make atoms cold. Verified: the charging integral and the spring audit agree; the factor-1/2 statement matches Griffiths §4.1 (Eq. 4.6 discussion) and Jackson §4.5–4.6; tweezer numbers recomputed (E_pk = 2.74×10⁷ V/m, U/k_B = 2.5 mK at I = 10¹² W/m²).

Limiting cases

Case	Result	Meaning
weak field, $E \ll E_{\rm at} \sim 6\times10^{11}$ V/m	$p = \alpha E$ exactly	all laboratory fields; stretch $\sim$ ppm of atomic size
strong field, $E \to E_{\rm at}$	hyperpolarizabilities, then field ionization	nonlinear optics territory; the spring's linear zone ends
spherical atom	$\alpha$ scalar, $\mathbf{p}\parallel\mathbf{E}$	H, noble gases; one number suffices
structured molecule	$p_i = \alpha_{ij}E_j$ , tensor	CO $_2$ : $\alpha_\parallel \approx 2\alpha_\perp$ ; deferred to the anisotropy article
conducting sphere	$\alpha = 4\pi\varepsilon_0R^3$	perfect compliance; same formula as the uniform ball
soft/large atom	$\alpha' \sim$ atomic volume	He 0.205 → Cs 59 Å³: loose electrons polarize
oscillating field, below resonance	replace $E^2 \to \langle E^2\rangle$	optical dipole traps; above resonance $\alpha$ flips sign

Common trapThe induced-dipole energy is minus one-half alpha E squared — not minus alpha E squared

Writing $U = -pE = -\alpha E^2$ for an induced dipole double-counts: it pockets the full alignment payoff while ignoring the $+\tfrac12\alpha E^2$ that went into stretching the atom to make the moment. Every induced-dipole result inherits the half — trap depths, van der Waals coefficients, dielectrophoretic forces — and the error is seductive because for permanent dipoles $U = -\mathbf{p}\cdot\mathbf{E}$ is correct with no half. The rule: if the moment exists before the field ( $\mathbf{p}$ fixed), no half; if the field builds the moment ( $\mathbf{p} = \alpha\mathbf{E}$ ), half. Mixed cases (polar and polarizable molecules — i.e., most of chemistry) take both terms, each with its own rule.

Worked examples

1 · How stretched is a real atom? (confidence builder)

An argon atom ( $\alpha' = 1.64\,\text{Å}^3$ ) sits in a uniform field $E = 1.0\times10^6\,\mathrm{V/m}$ . Find $\alpha$ in SI units, the induced moment, the charge-center displacement, and compare with water's permanent moment.

Step 1. Convert the polarizability volume: $\alpha = 4\pi\varepsilon_0\alpha' = 4\pi(8.854\times10^{-12})(1.64\times10^{-30}) = 1.82\times10^{-40}\,\mathrm{F\cdot m^2}$ .

Step 2. Induced moment: $p = \alpha E = 1.82\times10^{-40}\times10^{6} = 1.8\times10^{-34}\,\mathrm{C\cdot m}$ .

Step 3. Effective displacement, treating the 18-electron cloud as displacing rigidly: $d = p/(18e) = 1.8\times10^{-34}/(18\times1.6\times10^{-19}) \approx 6\times10^{-17}\,\mathrm{m}$ — a few percent of a nuclear diameter. Even chalking the whole moment to one electron gives only $1.1\times10^{-15}\,\mathrm{m}$ .

Step 4. Compare: water's permanent moment is $6.2\times10^{-30}\,\mathrm{C\cdot m}$ — thirty-four thousand times larger than what a megavolt-per-meter field induces in argon. Where permanent dipoles exist, they dominate the response utterly — which is why the next article's orientational mechanism, not this one, rules water, and why the two mechanisms are experimentally separable (Debye's trick: only the orientational part fades with temperature).

2 · Design check on an optical dipole trap (exam level)

A laser of wavelength $1064\,\mathrm{nm}$ is focused to intensity $I = 1.0\times10^{12}\,\mathrm{W/m^2}$ . Estimate the trap depth for ground-state argon and decide which atoms it can hold: room-temperature gas, or a laser-cooled cloud at $50\,\mathrm{\mu K}$ ?

Step 1. Peak field from intensity: $E_{\rm pk} = \sqrt{2I/\varepsilon_0c} = \sqrt{2\times10^{12}/(8.854\times10^{-12}\times3\times10^8)} = 2.7\times10^{7}\,\mathrm{V/m}$ — ten times air's breakdown field, routine inside a focused beam.

Step 2. Time-averaged well depth: $U = \tfrac12\alpha\langle E^2\rangle = \tfrac14\alpha E_{\rm pk}^2 = \tfrac14(1.82\times10^{-40})(2.7\times10^{7})^2 = 3.4\times10^{-26}\,\mathrm{J}$ .

Step 3. Convert to temperature: $U/k_B = 3.4\times10^{-26}/1.38\times10^{-23} = 2.5\,\mathrm{mK}$ .

Step 4. Verdict. Room-temperature argon carries $k_BT = 4.0\times10^{-21}\,\mathrm{J}$ — a hundred thousand times the well depth; atoms fly through the focus unbothered. The $50\,\mathrm{\mu K}$ cloud carries $k_BT = 6.9\times10^{-28}\,\mathrm{J}$ , fifty times less than the depth: solidly trapped. A neutral atom held at a laser focus by $-\tfrac12\alpha\langle E^2\rangle$ — the crude ball model's coefficient, doing frontier physics. (Real traps use alkali atoms whose near-resonant $\alpha$ is thousands of times larger; the mechanism is identical.)

Going further

References. Griffiths, Introduction to Electrodynamics, 4th ed., §4.1 · Jackson, Classical Electrodynamics, 3rd ed., §4.5–4.6 · Purcell & Morin, Electricity and Magnetism, 3rd ed., §10.5 · Zangwill, Modern Electrodynamics, §6.2.

Provenance. Equations and the $\alpha=4\pi\varepsilon_0 a^3$ model follow Griffiths §4.1; energy and force expressions and the Clausius–Mossotti derivation follow Jackson §4.5–4.6 and Zangwill §6.2. A verifier should recompute: (1) the Gauss's-law internal field $E=qr/4\pi\varepsilon_0 a^3$ and the resulting $\alpha$ ; (2) the $-\tfrac12\alpha E^2$ integral; (3) the Clausius–Mossotti algebra from the Lorentz local field; (4) the unit conversion $\alpha_{\rm SI}=4\pi\varepsilon_0\alpha'$ with the argon/hydrogen tabulated $\alpha'$ values (Griffiths Table 4.1, ~ $1.6$ and ~ $0.67\,\mathrm{\AA}^3$ ). No nonstandard numbers are asserted. (The Clausius–Mossotti bridge itself is built in its own article later in this chapter; here it is provenance context only.) Worked numbers recomputed for this edition: $\alpha_{\rm Ar} = 1.82\times10^{-40}$ F·m², $p = 1.8\times10^{-34}$ C·m at 1 MV/m, hydrogen stretch $1.4\times10^{-15}$ m at 3 MV/m, crude vs exact hydrogen $a_0^3 = 0.148$ vs $4.5a_0^3 = 0.667$ Å³, tweezer depth $3.4\times10^{-26}$ J $= k_B\times2.5$ mK at $10^{12}$ W/m².

Builds on: Electric Dipole Moment · Gauss's Law: Applications and Symmetry · Image Charges for a Conducting Sphere · Dielectric Materials and Insulators

Leads to: Permanent Dipoles and Orientational Polarization · Polarization Density P · Clausius–Mossotti Relation · Local Field and the Lorentz Cavity · Anisotropic Dielectrics and the Permittivity Tensor

If the displacement is only a femtometer, why isn't the induced dipole hopelessly negligible?

Compared to a permanent molecular dipole it is small — worked example 1's factor of 34,000. But two things rescue it. First, the census: $10^{25}$ – $10^{28}$ atoms per cubic meter turn femtometer stretches into macroscopic surface charge (the chapter opener's $\sigma_b$ ). Second, universality: every atom and molecule has an $\alpha$ , polar or not, so induced polarization is the floor response of all matter — the reason even nonpolar liquids like hexane have $\kappa \approx 2$ , and the only response left at optical frequencies, where permanent dipoles cannot rotate fast enough to follow.

Why does the crude model use a uniform ball rather than the real electron distribution?

Because the uniform ball is the crudest object that gets the physics qualitatively right, and its interior linearity makes the spring exact rather than approximate. The real payoff is the scaling $\alpha \sim 4\pi\varepsilon_0 a^3$ , which survives every refinement. The coefficient does not: hydrogen's true cloud is exponential ( $e^{-2r/a_0}$ ), softer at the edges, and yields $4.5\,a_0^3$ — a factor the ball cannot produce and second-order perturbation theory can. Treat the model as dimensional analysis with a mechanism attached, and it will never mislead you.

Is the polarizability related to the dielectric constant, and how do I get from one to the other?

Directly — that bridge is the chapter's destination. Dilute gases are easy: each molecule sees essentially the applied field, so the polarization is $N\alpha E$ and $\kappa - 1 = N\alpha/\varepsilon_0$ (small, like air's 0.0006). Dense matter is subtler because each molecule sits in the field of its polarized neighbors — the local field problem — and the honest accounting yields the Clausius–Mossotti relation connecting $\alpha$ to $\kappa$ . Both steps get their own articles (Linear Dielectrics; Clausius–Mossotti Relation; Local Field and the Lorentz Cavity). What is already true here: measure $\kappa$ of argon gas and you can extract the $\alpha'$ in our table, one atom at a time, from a capacitance bridge.

Does the induced dipole point along the field for every atom and molecule?

For any spherically symmetric charge distribution, yes — there is no other direction available, so $\alpha$ is a scalar. For structured molecules, no: CO $_2$ 's electrons slosh more easily along the O=C=O axis than across it, so its $\alpha$ is a tensor with distinct parallel and perpendicular eigenvalues, and a field applied obliquely induces a moment tilted toward the easy axis. The tilt has consequences — it gives a torque on the molecule even though the dipole is induced, and it makes crystal permittivity directional — developed in Anisotropic Dielectrics and the Permittivity Tensor.

What happens to alpha in an oscillating field — is the static value still right?

At frequencies far below the atom's internal resonances ( $\sim10^{15}$ – $10^{16}\,\mathrm{Hz}$ for electronic transitions), the cloud follows the field adiabatically and the static $\alpha$ is accurate — this covers radio through much of the infrared. Approaching a resonance, the driven-spring physics takes over: $\alpha$ grows, acquires an imaginary (absorptive) part, and changes sign above resonance — which is why blue-detuned optical traps repel atoms from intensity maxima while red-detuned ones attract. The full frequency-dependent story is dispersion, the electromagnetic-waves chapter's business; the spring model of this article is precisely the model that chapter drives sinusoidally.

Why is the conducting sphere's polarizability identical to the crude atom's — is that deep or accidental?

Half and half. Deep: both systems respond to a field by displacing negative charge until equilibrium, and dimensional analysis forces $\alpha = 4\pi\varepsilon_0\times(\text{size})^3\times(\text{dimensionless compliance})$ for any such object — so agreement in form is guaranteed. Accidental: the compliance factor happening to equal exactly 1 for both the perfectly screening conductor and the partially screening uniform ball is a curiosity of the ball's interior field. A real atom's factor is 4.5 (hydrogen); a dielectric sphere's is $3(\kappa-1)/(\kappa+2)$ (derived when we meet the Dielectric Sphere in a Uniform Field), which sweeps from 0 for vacuum up to the conductor's value of 3 only as $\kappa \to \infty$ — closing the ladder this chapter keeps climbing.

Do induced dipoles attract each other the way permanent dipoles can?

Yes — and that attraction holds the liquid and solid noble gases together. Atom A's instantaneous quantum fluctuation gives it a momentary dipole; its field polarizes atom B; the two correlated dipoles attract with energy scaling as $-\alpha_A\alpha_B/r^6$ — the London dispersion force, the universal floor of intermolecular attraction. Classically, a static induced-induced pair also attracts (each seeks the other's field gradient), but the quantum fluctuation version operates even between atoms in their ground states with no external field at all. Its $\alpha^2$ scaling is why polarizable xenon condenses at 165 K while stiff little helium holds out to 4 K.