The Field That Wasn't There: The Aharonov-Bohm Effect and the Reality of Potentials

The Hook: An Electron That Feels Something It Never Touches

Picture an experiment that any physics student understands in principle. A beam of electrons is split in two; the two partial beams pass an obstacle on the left and on the right and are recombined behind it. On the screen, an interference pattern of bright and dark fringes appears — the classic demonstration that electrons are waves.

Now the modification. A magnetic field is locked inside the obstacle. Not loosely, but perfectly: a long, thin solenoid whose field stays entirely inside. Outside — everywhere the electrons actually travel — the magnetic field is exactly zero. No field, no Lorentz force, no deflection. Classical mechanics therefore predicts that the electrons cannot possibly notice anything. Whether the solenoid is on or off must make no difference.

The interference pattern shifts anyway.

Yakir Aharonov and David Bohm, both then at the H. H. Wills Physics Laboratory of the University of Bristol, published this claim on 1 August 1959 in Physical Review under the title "Significance of Electromagnetic Potentials in the Quantum Theory." The manuscript had been received on 28 May 1959, the revised version on 16 June. Their own abstract states the outrage with remarkable calm:

"We shall show that, contrary to the conclusions of classical mechanics, there exist effects of potentials on charged particles, even in the region where all the fields (and therefore the forces on the particles) vanish."

That sentence contradicts one of the deepest intuitions of classical physics: that action is local and mediated by forces. Here is an action without a force. Victor Weisskopf is said to have captured the profession's typical reaction — as the Birkbeck physicist Basil Hiley reports in his reconstruction of the early history — in a single line: "The first reaction to this work is that it is wrong; the second is that it is obvious."

Between those two reactions lie almost three decades of dispute, several generations of electron microscopes, and finally a ring magnet six micrometers across, wrapped in a superconductor. This article tells what the effect is, why it was so hard to accept, how it was finally demonstrated beyond doubt — and what it teaches us about the difference between a bookkeeping quantity and a physical one.

Part 1: The Core Concept — Why Potentials Are More Than Computational Aids

Gauge Freedom and the Classical Verdict

Electrodynamics has two descriptive levels. One is the fields: the electric field E and the magnetic field B. They are what you measure, and they are what exerts forces — the Lorentz force depends exclusively on E and B.

The other level is the potentials: the scalar potential φ and the vector potential A. The fields follow from them by differentiation, B = ∇ × A. Historically they were introduced because they make Maxwell's equations more elegant to write and easier to solve.

The decisive point is that the potentials are not unique. You can add the gradient of any scalar function to A without changing B at all — the curl of a gradient vanishes. This freedom is called gauge freedom. Two physicists can therefore compute with entirely different vector potentials and arrive at identical fields and identical forces.

Classical physics drew the obvious conclusion: if a quantity can be changed at will without altering any measurement, then it is not a physical quantity but a mathematical convenience. The fields are real; the potentials are bookkeeping.

The Phase Is What Counts

Quantum mechanics overturns this order — but more subtly than the proposal by Aharonov and Bohm quoted below suggests. In the Schrödinger equation for a charged particle it is not the force that appears but the momentum, and there the vector potential enters directly. An electron traveling along a path therefore accumulates an additional phase proportional to the line integral of A.

Aharonov and Bohm draw exactly this parallel themselves: "Indeed, the potentials play a role, in Schrodinger's equation, which is analogous to that of the index of refraction in optics." The vector potential acts like a refractive index for matter waves.

For two paths that separate and meet again, what matters is the difference of the accumulated phases. And that difference is a loop integral of A around the closed circuit. By Stokes' theorem this loop integral equals the magnetic flux Φ through the area enclosed by the loop. In SI units the result reads:

Δφ = q · Φ / ħ

Aharonov and Bohm write it in the then-customary Gaussian units as (S₁ − S₂)/ħ = (e/ħc) ∮ A · dx = (e/ħc) Φ.

That is the whole effect, and mathematically it is trivial. The disturbing part is the interpretation: on the right-hand side stands the flux through an area the electron never enters. On the paths themselves, B = 0. Aharonov and Bohm say so unambiguously: "This effect will exist, even though there are no magnetic forces acting in the places where the electron beam passes." And: "Thus, there is a physical effect of the potentials even though no force is ever actually exerted on the electron."

They meet the obvious objection — that the electron might after all graze the field somewhere — directly in the text: "In order to avoid fully any possible question of contact of the electron with the magnetic field we note that our result would not be changed if we surrounded the solenoid by a potential barrier that reflects the electrons perfectly." That very move — making the field inaccessible — would become the experimental strategy 27 years later.

An order of magnitude for orientation: a phase shift of a full 2π — that is, a shift of the interference pattern by exactly one fringe — arises from an enclosed flux of Φ = h/e ≈ 4.14 · 10⁻¹⁵ weber. That is tiny. In a superconducting ring the flux is additionally quantized, in units of Φ₀ = h/2e ≈ 2.07 · 10⁻¹⁵ weber — the factor of two arises because the charge carrier in a superconductor is a pair of electrons. That number will matter again in the 1986 experiment.

Why Topology Enters the Picture

There is a more elegant way to understand the effect, and it explains at the same time why the vector potential cannot simply be gauged away. Aharonov and Bohm put it like this: "in a field-free multiply-connected region of space, the physical properties of the system still depend on the potentials."

The key phrase is "multiply-connected." The space in which the electron moves is field-free, but it has a hole: the solenoid. A loop around that hole cannot be continuously contracted to a point without crossing the forbidden region. And that is precisely why the usual argument fails. If the region were simply connected, a curl-free A could be written globally as a gradient and gauged away entirely. With the hole, that is impossible: the loop integral of A remains non-zero even though the curl vanishes everywhere.

The phase therefore does not depend on which exact path the electron takes, but only on how often and in which sense it encircles the hole. It is a topological quantity. That, far beyond this one experiment, is the pattern that would shape the physics of the following decades: measurable quantities that depend not on the details of geometry but on topology — the same grammar that underlies the quantum Hall effect.

Part 2: The 1959 Paper

What Aharonov and Bohm Proposed

The 1959 paper is short, clear, and unusually self-assured. It first discusses the solenoid case, derives the phase difference, and then proposes experiments. That the proposal was meant seriously is shown by a footnote in the original: "Dr. Chambers is now making a preliminary experimental study of this question at Bristol." The theorist and the experimentalist were in the same building.

At the end of the paper stands the conclusion that actually triggered the dispute. First the cautious version:

"But it does show that in a theory involving only local interactions (e.g., Schrodinger's or Dirac's equation, and current quantum-mechanical field theories), the potentials must, in certain cases, be considered as physically effective, even when there are no fields acting on the charged particles."

And then, a paragraph later, the bold one:

"It would therefore seem natural at this point to propose that, in quantum mechanics, the fundamental physical entities are the potentials, while the fields are derived from them by differentiations."

This second statement is a proposal ("it would seem natural … to propose"), not a proof — and that matters. Because the matter is not quite as simple as popular accounts often claim. We return to this in Part 5.

The Electric Variant

Less well known is that the same paper also describes an electric version of the effect. The idea: split an electron wave packet and send the two halves each through a metal tube — a Faraday cage. Inside such a tube the electric field is zero, but the potential can be controlled arbitrarily from outside. While the wave packets are deep inside, different time-dependent potentials are applied to the two tubes and then returned to zero before the packets emerge.

At no point does the electron experience an electric field, yet it accumulates a phase proportional to the time integral of the potential. The caption of the original figure reads simply: "Schematic experiment to demonstrate interference with time-dependent scalar potential."

For this electric variant Aharonov and Bohm report no experiment — and that has remained the case to this day. More on this in Part 6.

Part 3: The Lost Prehistory

Ehrenberg and Siday, 1949

The 1959 paper contains a conspicuous gap. Its reference list includes Marton, Möllenstedt, Düker, Brenner, Page, Bohm's own textbook, Kuper and two mathematical tables — but no mention of a work that had described the effect ten years earlier.

In 1949, Werner Ehrenberg and Raymond E. Siday had published a paper in the Proceedings of the Physical Society, Section B (volume 62, pages 8–21) under the thoroughly inconspicuous title "The Refractive Index in Electron Optics and the Principles of Dynamics." In it they defined a refractive index for electrons that contains the vector potential, calculated the phase difference between two paths as a function of the enclosed flux — and, as Hiley documents in his historical study, drew exactly the right conclusion:

"One might therefore expect wave-optical phenomena to arise which are due to the presence of a magnetic field but not due to the magnetic field itself, i.e., which arise whilst the rays are in field-free regions of space."

That is the Aharonov-Bohm effect, ten years before Aharonov and Bohm. Ehrenberg and Siday even checked whether the vector potential could be gauged away after all, and concluded that this is impossible without a singularity that breaks Stokes' theorem — that is, they arrived at the same topological argument.

Why did the work disappear? Beyond the title, Hiley names two reasons: the specialized language of the "equivalent refractive index," which deterred readers outside electron optics, and publication in a journal that was not a leading one. Siday himself, asked around 1957 about the response, is reported to have answered: "No. We have not heard a bloody thing — not as much as a whisper. It has fallen to the bottom like a lump of lead."

That the omission was not deliberate is well documented. Hiley records that Bohm was unaware of the paper when the first article was written, and that in their follow-up paper "Further Considerations on Electromagnetic Potentials in the Quantum Theory" (Physical Review 123, 1511–1524, 1961) Aharonov and Bohm acknowledged that Ehrenberg and Siday had obtained the same results using a semi-classical treatment. Hiley also reports a scene from 1961, when he began working at Birkbeck and Ehrenberg approached him: "Ach Hiley, zis AB effect that you are discussing, is it the one that Siday and I discovered?" And Bohm himself, when the possibility of a shared Nobel Prize was raised, is said to have waved it off: "After all it is only a straight forward application of standard quantum mechanics, and anyway Ehrenberg and Siday were there first!"

Walter Franz, 1939

The trail may run back further still. In May 1939, Walter Franz gave a talk titled "Elektroneninterferenzen im Magnetfeld" at a meeting of the "Gauverein Ostland" in Danzig; the abstract appeared in the Verhandlungen der Deutschen Physikalischen Gesellschaft (volume 20, 1939, pages 65–66). Its key formula already contains the additional term eΦ/c.

It is unclear, however, whether Franz recognized that no field at the location of the rays is needed at all. Hiley points out that Franz also wrote that the path difference caused by the curvature of the trajectories in the magnetic field drops out because of the additional term — a formulation that rather clouds the picture. Hiley himself concludes that it is "still an open question as to whether Franz had spotted the effect." I am of the opinion that this is the most honest way to handle the evidence: Franz had the mathematics, but whether he saw the physical point cannot be decided from a two-page conference abstract.

Part 4: The Long Dispute Over Proof

Chambers 1960 — and the Objection

Only a year after the proposal, R. G. Chambers in Bristol delivered the first experiment: "Shift of an electron interference pattern by enclosed magnetic flux," Physical Review Letters 5 (1960), pages 3–5. He used a magnetic whisker — a hair-fine, needle-shaped single crystal — as a micromagnet and showed the expected shift.

The objection came immediately, and it was not unreasonable: a whisker is an open magnet with a stray field. If some residual field exists where the electrons fly, then the observed shift could be an ordinary force effect. In retrospect Hiley calls this objection "wishful thinking" — but as an objection it was methodologically sound, and it governed the next 26 years.

Möllenstedt and Bayh

The situation improved considerably with the Tübingen experiments of Gottfried Möllenstedt and Wolfgang Bayh. In 1962 Bayh published measurements with a tungsten coil in the Zeitschrift für Physik (volume 169, pages 492–510); the same year Möllenstedt and Bayh reported in Naturwissenschaften (volume 49, pages 81–82) on an experiment with an air coil and a mu-metal shield. Hiley judges that Bayh's results made it possible to rule out arguments attributing the effect to stray fields.

The community nonetheless remained divided. The problem was fundamental: as long as the magnet is not completely and demonstrably sealed off, a residue of doubt always remains, and that residue is enough to prevent so far-reaching a claim from counting as proven.

Tonomura 1986: The Superconducting Cage

The solution came from Japan, from Akira Tonomura and his group at Hitachi. The strategy was precisely the one Aharonov and Bohm had sketched as a thought experiment in 1959: the field must not merely be locked in, it must be provably locked in.

As early as 1982, Tonomura's group had used a holography electron microscope to measure the phase difference between one beam passing through the hole of a ring-shaped ferromagnet and another passing outside it. Hitachi describes the objection to that experiment in its own words without equivocation: because the electron beams touched the magnet, meaning there were leakage magnetic fields in space, the experiment did not prove the existence of the AB effect.

The 1986 solution consisted of a doughnut-shaped ferromagnet six micrometers in diameter, covered with a layer of the superconductor niobium. The reason is the Meissner-Ochsenfeld effect: a superconductor expels magnetic fields completely from its interior and can therefore confine them as a perfect shield. The magnet was cooled to 5 kelvin for this purpose. The structure became possible only through microfabrication techniques originally developed for semiconductor devices.

The result is a single image: the interference fringes inside and outside the ring are displaced against each other by exactly half a fringe spacing. A full fringe corresponds to 2π; half a fringe therefore corresponds to a phase difference of π. And where the electrons flew, there was demonstrably no magnetic field.

Hitachi's own verdict reads: "In this way, the long dispute on the AB effect was brought to a close by this one picture." The American journal Physics puts it more soberly but in the same direction: only with an experiment in which the electromagnet was shielded by a superconducting screen were the remaining doubts about the Aharonov-Bohm effect finally put to rest.

A bibliographic note that honesty requires: two relevant publications came out of this Hitachi group in 1986 — Tonomura et al., "Evidence for Aharonov-Bohm effect with magnetic field completely shielded from electron wave," Physical Review Letters 56, 792–795 (24 February 1986), and Osakabe et al., "Experimental confirmation of Aharonov-Bohm effect using a toroidal magnetic field confined by a superconductor," Physical Review A 34 (1986). The publisher pages of both papers were not retrievable for this article; the experimental details above therefore come from Hitachi's own account of the experiment.

Part 5: What the Effect Proves — and What It Does Not

Care pays off here, because this is exactly where popular accounts regularly overreach. One often reads that the Aharonov-Bohm effect proves the vector potential to be "real" and the magnetic field to be merely derived. That is a stronger claim than the experiment supports. The situation sorts into three readings.

First, what is established. A charged particle can demonstrably behave in a measurably different way depending on which magnetic flux is enclosed by its possible paths — even when it never enters a field region itself. Local fields at the particle's location are not sufficient to describe the quantum mechanics of charged particles. This is experimentally established and uncontested.

Second, what is an interpretation. The reading that the vector potential is the fundamental quantity is one possible interpretation — and it is exactly what Aharonov and Bohm formulated as a proposal, not as a proof. Against it stands the fact that A remains gauge-dependent. What is measurable is not A itself but the loop integral of A along a closed circuit, and that is gauge-invariant. The physically clean quantity is therefore a non-local but gauge-invariant one — in modern language: the holonomy, or the phase factor around the loop.

Third, what to take away for physics. The real lesson is not "potentials beat fields," but rather: the complete description of electrodynamics is neither purely local and field-based nor given by the potential at a point, but by gauge-invariant loop quantities. That sounds technical, but it had consequences: this principle stands at the beginning of modern gauge theories and of topological physics.

That the question of the correct formulation was debated among specialists as well is shown by an exchange referenced in the Physics account: in 2010 Michael Berry and Murray Peshkin published pieces in Physics Today (volume 63, issue 8) under the title "Aptly named Aharonov-Bohm effect has classical analogue, long history," in which — according to that summary — they argued that Aharonov and Bohm nevertheless deserve credit for properly understanding the effect that bears their name. The title betrays the tension: "aptly named" is a defense against the charge that the name is historically unjust.

Part 6: The Effect's Descendants

Mesoscopic Rings, Nanotubes, Flux Quantization

The exotic thought experiment has become a workhorse of solid-state physics. In 1985 Webb and co-workers demonstrated Aharonov-Bohm oscillations in ordinary, non-superconducting metallic rings: as the flux through the ring is increased, the electrical resistance oscillates periodically, because the two paths around the ring interfere alternately constructively and destructively. In 1999 Bachtold and co-workers detected the effect in carbon nanotubes. Research on geometric phases in mesoscopic rings is ongoing to this day.

Related, though not identical, is flux quantization in superconducting loops: the enclosed flux must be an integer multiple of Φ₀ = h/2e, because the Cooper-pair wavefunction must be single-valued. The same logic carries the SQUID, the most sensitive magnetometer known.

The Electric Variant: Still Open

The electric version proposed alongside in 1959 has to this day no uncontested demonstration. A 1998 ring experiment with tunnel barriers and a constant bias voltage does count as an observation — but in an arrangement in which the charges do traverse the electric field generated by the bias voltage. The original, time-dependent electric Aharonov-Bohm effect in its strict form — phase without any field at the particle's location — is therefore still awaiting confirmation, 67 years after the proposal. That is a remarkable open account.

Gravity: Overstreet et al. 2022

Perhaps the most spectacular continuation came in 2022. Chris Overstreet, Peter Asenbaum, Joseph Curti, Minjeong Kim and Mark A. Kasevich (Stanford) published "Observation of a gravitational Aharonov-Bohm effect" in Science (volume 375, issue 6577, pages 226–229, 14 January 2022).

The setup is an atom interferometer with ⁸⁷Rb atoms whose partial wave packets are separated by up to 25 centimeters. A source mass is brought close to one of the two arms: a 170° semicircular ring of tungsten (99.95% pure) with an inner radius of 6.8 cm, an outer radius of 7.8 cm, a height of 3 cm and a mass of 1.25 kg. Crucially, the group measured the deflection of each arm separately with gradiometers and subtracted it, in order to isolate the remaining, purely quantum-mechanical phase contribution. That contribution was −125 ± 24 mrad in one configuration and −182 ± 28 mrad in the other; the deviation from the classically expected midpoint prediction amounted to 13 and 19 standard deviations respectively.

The paper's abstract closes with the sentence: "These results show that gravity creates Aharonov-Bohm phase shifts analogous to those produced by electromagnetic interactions."

What matters here is the limitation the authors themselves state: "There is no configuration in which the interferometer trajectories are completely unperturbed." Unlike the superconductor-shielded ring magnet, gravity cannot be switched off; the field-free ideal case cannot be realized in the gravitational experiment, which is why the classical contribution has to be subtracted by calculation and measurement. This is therefore an analogue, not an exact repetition of the electromagnetic experiment.

Frameworks to Take Away: The Variants at a Glance

The Central Takeaway

The Aharonov-Bohm effect is in truth a lesson about representations. For decades the vector potential counted as "just mathematics" because it is not unique. Ambiguity was confused with meaninglessness. In fact the situation was more complicated: the potential at a point really is not a physical quantity — but a particular global combination of potential values, the loop integral around a circuit, very much is one, and it is measurable even though none of its local constituents are.

This is a pattern that carries far beyond physics, and it has an immediately practical form. In every complex system there are quantities that look locally arbitrary and are therefore dismissed as implementation details — and for which a globally invariant combination nonetheless exists that determines the observable behavior. Anyone who designs software architectures knows the phenomenon: the absolute ordering of events in a distributed system is no more unique than a gauge, and yet there is an invariant structure — the causal order — on which everything turns. Look only at the local timestamps and you see arbitrariness; look at the loop and you see the law.

The practical prompt: when you encounter a quantity in a system that you can "define away" without any measurement changing, do not immediately conclude that it is meaningless. Ask the second question instead: Is there a closed loop over which this quantity can be summed? And: Does the space I am moving in have a hole? Because that is exactly where — in multiply connected spaces — arbitrariness turns into invariance. The answer to those two questions shifted a worldview in 1959.

Cross-References in the Vault

The Aharonov-Bohm effect connects to several topics already covered here in more than one way. Topology as a source of exactly measurable quantities is the leitmotif of The Most Exact Number in Physics: The Quantum Hall Effect and the Topology of Matter; the Meissner-Ochsenfeld effect and flux quantization, without which the 1986 experiment would not have been possible, are explained in Resistance Is Futile: Superconductivity from Onnes to BCS to the Room-Temperature Dream. That quantum mechanics contains non-local relationships which cannot be reduced to local properties is the subject of The Wager Against Reality: Bell's Theorem and the End of Local Realism and Spooky Action at a Distance: Quantum Entanglement from Einstein to the Quantum Internet. Why such interference patterns disappear in the macroscopic world is explained by Why the World Turns Classical: Decoherence, Einselection, and Quantum Darwinism. That even "empty" vacuum is physically effective although seemingly nothing is there is shown by the related line of reasoning in The Force from Nothing: The Casimir Effect and the Pressure of the Quantum Vacuum. And the extreme phase sensitivity of modern interferometry, which makes experiments like Overstreet's possible in the first place, is the subject of The Clock That Feels a Centimetre: Optical Atomic Clocks and Measuring Time at the Quantum Limit.

Something to Think About

For decades physics regarded the vector potential as a mere computational aid because it is not uniquely determined — overlooking that a non-unique quantity can nonetheless possess a unique, measurable invariant. Which quantity in your own field of work are you currently treating as "just a convention," "just an implementation detail," or "just a representation"? And what would you find if you asked not about its value at a point, but about what it sums to around a closed loop?

Fact Check

For this article, 34 checkable claims were extracted (14 hard facts, 8 verbatim quotations, 6 source attributions, 6 synthesis claims) and checked against the primary and publisher sources opened during this run. 7 claims were corrected or weakened. The most important ones:

1. The core claim "the AB effect proves that the vector potential is more fundamental than the magnetic field" was deleted and replaced. The original paper explicitly formulates this as a proposal ("It would therefore seem natural at this point to propose …"), not as a proof; the gauge-invariant and thus physically robust quantity is the loop integral of A, not A itself. Part 5 of the article was rewritten accordingly, and "The Central Takeaway" now follows this weaker but correct reading.
2. The account of the Tonomura experiment was reduced to the verifiable details. The publisher pages of Physical Review Letters 56, 792 and Physical Review A 34 were not retrievable; statements about the material (permalloy), the accelerating voltage, and a dependence of the phase on the parity of flux quanta were therefore deleted. Only the following are documented and retained: ring shape, 6 µm diameter, niobium shielding, 5 K, half-fringe displacement — all from Hitachi's own account.
3. The gravitational AB effect is not presented as a field-free demonstration. The authors state themselves: "There is no configuration in which the interferometer trajectories are completely unperturbed." The article therefore describes the result as an analogue with the classical contribution subtracted, not as a gravitational repetition of the Tonomura experiment.

Further corrections concern Walter Franz's priority (an open question rather than an anticipation), the status of the electric AB effect (the strict time-dependent form is unconfirmed; a 1998 variant was observed), the attribution of the Ehrenberg-Siday quotation (quoted after Hiley; the 1949 original was not consulted directly), and an unsupported time span: the original wording "for forty years the vector potential counted as a mere computational aid" was weakened to "for decades," since no defensible starting point can be given for a specific number of years.

Note: Automated AI verification against primary sources; no human expert review.

Sources and Further Reading

• Y. Aharonov and D. Bohm, "Significance of Electromagnetic Potentials in the Quantum Theory," Physical Review 115, 485–491 (1 August 1959) — original paper, full-text scan: https://docente.unife.it/guido.zavattini/allegati/aharonov-bohm-physrev-115-485.pdf
• B. J. Hiley, "The Early History of the Aharonov-Bohm Effect," arXiv:1304.4736 — historical reconstruction with the evidence on Ehrenberg & Siday (1949) and Franz (1939): https://arxiv.org/pdf/1304.4736
• Hitachi Research & Development, "Aharonov-Bohm Effect" — account of the 1982 and 1986 experiments by the group that performed them: https://www.hitachi.com/rd/research/materials/quantum/aharonov-bohm/index.html
• David Lindley, "Quantum Milestones, 1959: Ghostly Influence of Magnetic Field," Physics 18, 48 (republished for the International Year of Quantum Science; original dated 22 July 2011): https://physics.aps.org/articles/v18/48
• C. Overstreet, P. Asenbaum, J. Curti, M. Kim and M. A. Kasevich, "Observation of a gravitational Aharonov-Bohm effect," Science 375 (6577), 226–229 (14 January 2022), open-access version: https://par.nsf.gov/servlets/purl/10340035
• Overview article on the effect with references to Webb et al. (1985), Bachtold et al. (1999) and the status of the electric variant: https://en.wikipedia.org/wiki/Aharonov%E2%80%93Bohm_effect

← All articles