The Force from Nothing: The Casimir Effect and the Pressure of the Quantum Vacuum

The Hook: Two Mirrors, an Empty Space, and a Force That Should Not Exist

Take two perfectly smooth, electrically neutral metal plates. Put them in a vacuum as good as technology allows: no air, no dust, no residual charge, no temperature differences. Hold them parallel, one micrometer apart – a hundredth of the thickness of a human hair. And then do nothing.

According to everything classical physics has to say, nothing happens now. The plates are uncharged, so there is no Coulomb force. They are non-magnetic. The gravitational pull between two pieces of sheet metal is laughably small. Between them there is – nothing. And yet: the two plates attract each other. Not much, but measurably. At a separation of one micrometer the pressure is about a thousandth of a pascal. Reduce the gap to ten nanometers and it grows to more than one bar – the pressure of Earth's atmosphere at sea level, produced by nothing.

This force bears the name of the Dutch physicist Hendrik Casimir, who predicted it in 1948, and it ranks among the strangest predictions quantum physics has ever made. Because at its core it claims that empty space is not empty. That the vacuum has a structure, an energy, a seething of fields that one normally does not notice – but which turns into a real, mechanical force as soon as one imposes boundaries on it.

For almost fifty years the Casimir effect remained a textbook curiosity, theoretically elegant, experimentally inaccessible. Then, in the late 1990s, it was measured with a precision that left no room for doubt. Today it is a nuisance that makes micromachines stick together, a tool for moving nanostructures, a testing ground for theories of gravity at the smallest scales – and, as we shall see, the starting point of one of the most embarrassing open questions in all of physics: if the vacuum has energy, why does that energy not tear the universe apart?

This article takes you from Casimir's desk at Philips in 1948, via the torsion pendulums and atomic force microscopes of the 1990s, to superconducting circuits that in 2011 quite literally produced light from nothing. And it ends with the honest question of what the effect actually proves – and what it does not.

Part 1: The Vacuum That Is Not Empty

Zero-Point Energy – the Price of Uncertainty

To understand the Casimir effect, one first needs a picture of what quantum mechanics means by "vacuum." In classical physics the vacuum is the absence of everything: no particles, no fields, zero energy. Quantum mechanics does not permit such a state.

The reason is Heisenberg's uncertainty principle. A pendulum hanging perfectly still in its resting position would simultaneously have an exactly determined position (at the bottom) and an exactly determined momentum (zero). This is precisely what quantum mechanics forbids. A quantum harmonic oscillator therefore retains, even in its lowest state, a residual energy – the zero-point energy – of ½ħω, where ħ is the reduced Planck constant and ω the oscillator's angular frequency. It trembles even when it is "at rest."

Now, the electromagnetic field is nothing other than an infinite collection of such oscillators: every possible vibration pattern of the field, every wavelength, every direction of propagation, every polarization is its own oscillator, a "mode." In the vacuum all these modes are in their ground state – unexcited, without a single photon. But each of them carries its zero-point energy ½ħω. Summing over all modes yields the vacuum energy of the field. And because there are infinitely many modes of arbitrarily high frequency, this sum is infinite.

That sounds like a catastrophe for the theory, and in a sense it is – we will return to it. But for the everyday business of physics it is not a problem, because only energy differences are measurable. An infinite constant that is the same everywhere drops out of every measurement. By convention one sets it to zero and carries on.

Casimir's Insight: Boundaries Change the Sum

This is where Hendrik Casimir comes in. His decisive question was: what happens if the infinite constant is not the same everywhere? What if one imposes boundary conditions on the field so that the sum over the modes comes out differently in one place than in another?

That is exactly what two parallel, perfectly conducting plates do. At the surface of a perfect conductor the tangential electric field must vanish. Between the plates, therefore, only those modes are allowed that "fit" into the gap: standing waves whose half-wavelength is an integer fraction of the plate separation d. Anyone who has ever looked at a guitar string knows the principle – it can vibrate only at its fundamental and its overtones, not at arbitrary frequencies. Outside the plates, by contrast, all frequencies are allowed, a continuum.

Between the plates, then, modes are missing that exist outside. Every missing mode means ½ħω less zero-point energy. The vacuum energy between the plates is therefore lower than outside – and the lower, the closer the plates stand, because then even fewer modes fit in. A system whose energy decreases when two parts approach each other experiences an attractive force. That is the Casimir effect.

One can also read the same picture as radiation pressure: the vacuum fluctuations outside the plates push from all sides with the full spectrum, while between the plates part of the spectrum is missing. The pressure from outside exceeds the pressure from inside, and the plates are pushed together – by the vacuum itself.

The Formula and Its Astonishing Properties

Casimir published his calculation in 1948 in a short paper entitled "On the attraction between two perfectly conducting plates" in the Proceedings of the Royal Netherlands Academy of Arts and Sciences (volume 51, p. 793). The result is disarmingly simple. The force per unit area between two perfectly conducting parallel plates at separation d is:

F/A = −π²ħc / (240 d⁴)

The minus sign denotes attraction. Three things about this formula are remarkable.

First, it contains only two constants of nature: ħ and the speed of light c. The elementary charge does not appear, nor any material property, nor any mass. The force appears to be a pure property of space, independent of what the plates are made of – an impression that will later turn out to be only half right.

Second, the distance dependence: the force falls off with the fourth power of the separation. Halve d and the pressure grows by a factor of 16. This explains why the effect is completely invisible at macroscopic distances and brutal at nanometers. The numbers: at d = 1 µm the pressure is about 1.3 mPa – the weight of a grain of sand spread over a square meter. At d = 100 nm it is already about 13 Pa. At d = 10 nm, roughly 130 kPa, or about 1.3 bar.

Third, the mathematical origin. Casimir had to subtract an infinite integral (the energy without plates) from an infinite sum (the energy between the plates). Both expressions diverge; only their difference is finite. This works with a physically motivated cutoff function: at very high frequencies (X-rays and beyond) real metals become transparent, the boundary condition no longer applies there anyway, and the corresponding modes contribute equally inside and outside. With this regularization, the Euler-Maclaurin formula leaves a finite remainder in which the value of the Riemann zeta function ζ(−3) = 1/120 appears – hence the 240 in the denominator. If you have ever seen the famous "equation" 1 + 2 + 3 + … = −1/12: it is the one-dimensional sister of the same calculation, ζ(−1) = −1/12, and it is to be understood in exactly this sense – not as a sum in the ordinary meaning, but as the finite remainder that is left when a physically justified infinity is cleanly subtracted.

A Conversation with Bohr

The origin story has a nice twist. In the 1940s Casimir worked at the Philips Research Laboratory in Eindhoven on a very practical problem: the stability of colloids, that is, fine particles suspended in a liquid. The van der Waals forces acting between such particles fell off faster at larger distances than the theory of the day predicted. Together with Dirk Polder, Casimir showed in 1948 in a paper in the Physical Review (volume 73, p. 360) that the cause is the finite speed of light: the fluctuations in one molecule need time to reach the other, and at large distances the interaction thereby loses strength. This has since been called the "retarded" van der Waals force or the Casimir-Polder force.

By Casimir's own account, he mentioned this result to Niels Bohr, who muttered that it must have something to do with zero-point energy. That remark gave Casimir the idea of calculating the force not from the fluctuating dipoles of the atoms but directly from the change in vacuum energy caused by the presence of boundaries. Thus the plate calculation was born – and with it the interpretation that this is a force of the vacuum itself. We will see at the end that this interpretation is contested to this day: whether one describes the effect as "vacuum energy" or as "retarded van der Waals force between the charges in the plates" is a matter of perspective, not of mathematics.

Part 2: Fifty Years of Waiting – the Long Road to Measurement

Sparnaay's Brave Attempt (1958)

Measuring a force of a few millipascals at a separation of one micrometer sounds feasible today. In the 1950s it was a nightmare. One had to hold two plates parallel to within a micrometer – for plates of square-centimeter size that means an angular tolerance that could hardly be guaranteed. One had to eliminate every residual charge, because even tiny electrostatic potential differences produce forces exceeding the Casimir effect by orders of magnitude. One had to control dust, roughness, and surface contamination.

Marcus Sparnaay, also at Philips, made the attempt in 1958 and published in Physica (volume 24, p. 751). His measurements had an uncertainty of about 100 percent. His conclusion was correspondingly cautious: the observed attractions did not contradict Casimir's prediction. That was honest and correct – and far from a confirmation.

For almost forty years it stayed that way. The Casimir effect haunted textbooks as an elegant consequence of quantum electrodynamics whose experimental verification was "in principle" possible. One exception was the Casimir-Polder force between individual atoms and surfaces, which Sukenik and colleagues measured with good accuracy in 1993 using sodium atoms flying through a micrometer-wide slit. But the plate force itself, the actual centerpiece, remained unconfirmed.

Lamoreaux and the Torsion Pendulum (1997)

The breakthrough came in 1997 from Steve Lamoreaux, then at the University of Washington in Seattle. His paper "Demonstration of the Casimir Force in the 0.6 to 6 µm Range" appeared in Physical Review Letters (volume 78, p. 5) and is today regarded as the first unambiguous quantitative detection.

Lamoreaux sidestepped the parallelism problem with a trick that has since become standard: instead of two plates he used a plate and a weakly curved spherical lens. Between a sphere and a plane there is no tilt angle to control – the separation is simply the distance at the closest point. Both surfaces were gold-coated. The lens had a radius of curvature of a good ten centimeters; it hung from a torsion pendulum whose minute rotation revealed the force. A piezo element moved the plate, and a laser interferometer measured the separation to about ten nanometers.

For the sphere-plate geometry, a good approximation (the so-called proximity force approximation) gives F = −π³ħcR / (360 d³) – here the force falls off only with the third power of the separation, and it grows with the sphere radius R. With R ≈ 11 cm, this yields forces of about one nanonewton at 0.6 µm separation and still about one piconewton at 6 µm. Lamoreaux found agreement with theory at the level of roughly five percent. After almost half a century, the force from nothing was a measured fact.

The work was not flawless – the lens radius had to be corrected later, and the treatment of temperature and conductivity corrections triggered a debate that continues to reverberate. But that changed nothing about the core: the vacuum pushes.

The Atomic Force Microscope and the One-Percent Era (1998–2002)

A year later, Umar Mohideen and Anushree Roy at the University of California, Riverside, followed with an even finer instrument. In their paper "Precision Measurement of the Casimir Force from 0.1 to 0.9 µm" (Physical Review Letters, volume 81, p. 4549, 1998) they mounted a metallized polystyrene sphere of only 196 micrometers in diameter on the cantilever of an atomic force microscope. The deflection of the cantilever measured the force between the sphere and a metal-coated plate at separations down to 100 nanometers. For the first time, corrections for the finite conductivity of the metal, for surface roughness and for temperature were systematically included; agreement with theory was at the one-percent level.

In 2002 Giacomo Bressi, Giovanni Carugno, Roberto Onofrio and Giuseppe Ruoso in Padua finally achieved what Sparnaay had failed to do: a measurement in the original geometry with two parallel plates, in the range from 0.5 to 3 µm, with about 15 percent accuracy. Casimir's original prediction was thus confirmed in its own configuration.

The following table summarizes the milestones:

Part 3: Real Materials – Lifshitz and the Return of Matter

Why Gold Is Not Perfect

Casimir's formula holds for perfect conductors, which do not exist. Real gold reflects visible light superbly but becomes increasingly transparent in the ultraviolet; below its plasma frequency (about 9 eV, corresponding to a wavelength around 140 nm) it behaves like a mirror, above it it does not. Modes that penetrate the metal do not feel the boundary condition. At plate separations comparable to this wavelength – that is, precisely in the experimentally interesting range around 100 nm – this noticeably reduces the force, by ten to twenty percent or more.

The general theory that incorporates such material properties is due to Evgeny Lifshitz in 1956, extended in 1961 together with Dzyaloshinskii and Pitaevskii. Lifshitz theory computes the force between two arbitrary bodies from their frequency-dependent dielectric functions ε(ω) – that is, from how strongly a material responds to electric fields at each frequency. Casimir's result emerges as the limiting case ε → ∞. And the ordinary van der Waals force at very small separations, where retardation plays no role, emerges as another limiting case. Lifshitz theory thus unifies all three phenomena – van der Waals, Casimir-Polder and Casimir – into a single force, today often called the Casimir-Lifshitz force.

This is more than a correction. It is a change of perspective: the force does after all depend on matter. It is not a pure property of space but the result of an interplay between the fluctuations of the field and the charges in the bodies that respond to those fluctuations.

The Temperature Controversy

At finite temperature, thermal photons join the zero-point fluctuations and also contribute to the force. At room temperature the thermal wavelength is about 7 µm; at separations below that, the thermal contribution should be small. But exactly how small depends on how one models the electrical conductivity of the metal at very low frequencies – and here the field has been split for more than twenty years.

The Drude model takes into account that conduction electrons in a real metal suffer collisions and dissipate energy. The plasma model ignores this dissipation. The two models are practically identical at optical frequencies but differ dramatically in the limit ω → 0 – and it is precisely this limit that determines the thermal contribution. A series of precision experiments, in particular those by Ricardo Decca and colleagues using micro-torsional oscillators, agree better with the plasma model, even though the Drude model appears physically more plausible. Other experiments, including one from Lamoreaux's group at larger separations, favor the Drude model. The conflict is unresolved, and it is the reason why the Casimir effect is pursued today not only as a confirmation but as a sensitive test of the electrodynamics of real metals. I am of the opinion that this "Casimir puzzle" is a fine example of how a precision experiment puts pressure on a seemingly settled theory at exactly the spot where it glues two convenient approximations together.

Part 4: From Curiosity to Engineering Quantity – Casimir in the Microworld

Stiction: When Machines Stick to the Vacuum

As long as machines were centimeters in size, one could safely ignore the Casimir effect. With the advent of microelectromechanical systems (MEMS) – tiny mirrors, accelerometers, microswitches, gyroscopes, as found in every smartphone – that changed. Moving parts in MEMS have separations of hundreds of nanometers, and the force that explodes like d⁻⁴ at these separations suddenly becomes a relevant item in the force budget.

The best-known symptom is called stiction (from "static friction" and "stick"): two components come too close to each other during fabrication or operation, the Casimir force (together with capillary and electrostatic forces) pulls them together, and the restoring force of the tiny springs is not enough to separate them again. The component is permanently stuck – useless. For the MEMS designer, then, the vacuum is not a philosophical concept but an adhesive against which one must design.

Chan et al. 2001: The Vacuum as a Drive

That the same force can also be useful was shown by Ho Bun Chan, Vladimir Aksyuk, Raymond Kleiman, David Bishop and Federico Capasso at Bell Labs in 2001. In their paper "Quantum Mechanical Actuation of Microelectromechanical Systems by the Casimir Force" (Science, volume 291, p. 1941) they built a microscopic seesaw: a polysilicon plate half a millimeter on a side, suspended from two thin torsional rods. When a gold-coated sphere approached the plate, the plate tilted – purely through Casimir attraction. The tilt angle as a function of separation agreed with theory. For the first time the vacuum force had not merely been detected but deployed as an actuator. In a follow-up paper the same group showed that the Casimir force turns the micro-oscillator into a nonlinear, bistable system – a hint that the effect could be exploited deliberately for the construction of switches and sensors.

Since then a research field of its own has emerged, called "Casimir engineering." A 2024 review in the journal Micromachines summarizes the state of the art: researchers investigate how geometry (gratings, cylinders, nanostructured surfaces) and choice of materials (metals, semiconductors, graphene, metamaterials) can strengthen, weaken, or reverse the force. A particularly elegant finding comes from Munkhbat and colleagues (Nature, 2021): two gold flakes in a liquid find, entirely on their own, an equilibrium separation at which Casimir attraction and electrostatic repulsion balance – and thereby form a self-assembled optical resonator a few hundred nanometers wide, without anyone having had to assemble it.

Casimir as a Window onto New Physics

There is one more reason why precision measurements of the Casimir force are of interest far beyond micromechanics. Many theories beyond the Standard Model – for example those with extra spatial dimensions or with new light particles – predict deviations from Newton's law of gravity at distances below a millimeter, typically in the form of an additional Yukawa-type term. At distances from nanometers to micrometers the Casimir force is the dominant background against which such deviations would have to be sought. Every measurement of the Casimir force that agrees with theory therefore excludes a region of parameter space for "fifth forces." The exclusion maps derived from Casimir experiments are today part of the standard repertoire in the search for new physics at small scales. That, too, is one of the surprising careers of this effect: from colloid chemistry to the hunt for extra dimensions.

Part 5: Repulsion – Can the Vacuum Also Push?

Boyer's Prediction and the Kenneth-Klich Theorem

Is the Casimir force always attractive? For two identical plates, yes – but not in general. As early as 1974 Timothy Boyer showed that a perfectly conducting plate and a perfectly magnetically permeable plate would have to repel each other: the magnetic boundary condition alters the allowed mode spectrum in the opposite way. In practice this foundered on the fact that there are no materials that are appreciably magnetic at optical frequencies.

A deep result on this comes from Oded Kenneth and Israel Klich (2006). Their theorem carries the apt title "Opposites attract" and states: two bodies that are mapped into each other by a reflection – that is, two identical objects facing each other as in a mirror – always attract in vacuum, regardless of shape and material. Anyone who wants repulsion must break the symmetry: different materials, a medium in between, or both.

Munday, Capasso and Parsegian 2009: The First Measured Repulsion

That is exactly the route taken by Jeremy Munday, Federico Capasso and Adrian Parsegian at Harvard. Lifshitz theory predicts: if two bodies made of materials 1 and 2 are separated by a medium 3, the force becomes repulsive when the dielectric functions, over the relevant frequency range, follow the ordering ε₁ > ε₃ > ε₂ – the medium must lie "between" the two solids. Munday and colleagues chose gold (high ε), silicon dioxide (low ε), and between them bromobenzene, whose dielectric response lies right in the middle.

Using an atomic force microscope they measured the force between a gold-coated sphere and a silica plate in this liquid. The result, published as "Measured long-range repulsive Casimir-Lifshitz forces" in Nature (volume 457, p. 170, 2009): a repulsive force of a few tens of piconewtons at separations down to about 40 nanometers. As a control they replaced the silica plate with a gold plate – and the force became attractive, as predicted. The vacuum, or more precisely the fluctuating field in the medium, really can push.

The consequence is tantalizing: a repulsive Casimir effect could enable frictionless bearings at the nanoscale – components floating on a cushion of quantum fluctuations. Zhao and colleagues demonstrated such a "quantum trap" in Science in 2019: a gold flake levitated in ethanol above a Teflon-coated gold surface in a stable equilibrium between attraction and repulsion.

A Warning from Recent Research

How difficult this field is was shown by an episode in 2024/25. A group at the University of Science and Technology of China reported in Nature Physics (Zhang et al., volume 20, p. 1282, 2024) that it had reversibly switched the Casimir force between a gold sphere and a silica plate in a ferrofluid – a suspension of magnetic nanoparticles – from attraction to repulsion by means of an external magnetic field. That would be a remarkable advance: a Casimir switch at the push of a button.

But in April 2025 a "Matters Arising" contribution by Moazzami Gudarzi and Aboutalebi appeared in the same journal under the title "Re-examining magnetic tuning of Casimir forces," questioning both the theoretical framework and the interpretation of the measurement data – among other reasons because in such a nanoparticle suspension forces arise that have nothing to do with the Casimir effect (electrical double layers, depletion forces) but can produce similar signatures. The original authors have replied; the question remains open. For us this is a valuable lesson: in liquids the Casimir-Lifshitz force is only one of many forces on the nanometer scale, and isolating it requires extraordinary care. Where a spectacular result rests on a difficult separation of effects, a measure of skepticism is warranted until independent replications are available.

Part 6: The Dynamical Casimir Effect – Light from Nothing

A Mirror Faster Than the Vacuum

So far we have confined the vacuum in space. What happens if one disturbs it in time – if one does not hold a mirror still but moves it? Gerald Moore posed this question in 1970; Stephen Fulling and Paul Davies deepened it in 1976. The answer of quantum field theory: if a mirror moves uniformly, nothing special happens. But if it moves with acceleration, and so fast that the field can no longer follow adiabatically, then the virtual fluctuations of the vacuum are converted into real photons. The mirror "shakes" light out of nothing. This is called the dynamical Casimir effect.

Intuitively: one can picture the zero-point fluctuations as virtual photon pairs that arise for an immeasurably short time and vanish again. A mirror moving back and forth at an appreciable fraction of the speed of light can supply energy to such a pair before it annihilates – and the virtual pair becomes a real one that flies away. Characteristically, the photons are created in pairs, and their frequencies add up to the modulation frequency of the mirror.

The problem: to achieve a measurable effect, a mechanical mirror would have to oscillate at gigahertz frequencies with amplitudes bringing it to relativistic speeds. No material withstands that. For forty years the effect was considered correct in principle but unobservable.

Wilson et al. 2011: The Electrical Mirror

The solution was found by a group around Christopher Wilson and Per Delsing at Chalmers University in Gothenburg, published as "Observation of the dynamical Casimir effect in a superconducting circuit" in Nature (volume 479, p. 376, 2011). Their trick: do not move the mirror, but change the effective electrical length of a transmission line.

The experiment consisted of a superconducting coplanar microwave transmission line terminated by a SQUID – a superconducting quantum interference ring whose inductance can be controlled by a magnetic flux. Changing the inductance changes the boundary condition for the microwave field in the line, which electrically acts exactly as if one were shifting the end of the line – the "mirror." And this effective mirror can be modulated at more than ten gigahertz, so that it "moves" at a substantial fraction of the speed of light without a single atom stirring.

The result: from the line, cooled to a few millikelvin and containing not a single thermal photon, microwave photons emerged – in pairs, with frequencies distributed symmetrically around half the modulation frequency. Decisive was the detection of so-called two-mode squeezing in the emitted radiation: a correlation between the photon pairs that classical noise sources cannot produce and that demonstrates the quantum nature of the generation process. Light from nothing – produced by imposing a rapidly changing boundary on the vacuum.

The dynamical Casimir effect is conceptually closely related to two other famous predictions: the Hawking radiation of black holes and the Unruh effect, according to which an accelerated observer perceives the vacuum as a warm bath of particles. In all three cases real radiation arises because the definition of "vacuum" does not agree for different observers or at different times. The Casimir effect is the only one of these three phenomena that can be observed directly in the laboratory – which makes it a valuable analog model for the others.

Part 7: What Does the Casimir Effect Actually Prove? Jaffe and the Cosmological Constant

The Infinity That Does Not Go Away

Let us return to the infinity from the beginning. We dismissed the infinite vacuum energy as a harmless constant because only differences are measurable. That is true for all forces – except one: gravity. According to Einstein's general relativity, gravity couples to every form of energy, including a constant energy density of the vacuum. Such a vacuum energy density acts exactly like a cosmological constant Λ and drives the expansion of the universe.

Precisely such an accelerated expansion was observed in 1998, and the vacuum energy density inferred from it is about 10⁻⁹ joules per cubic meter – tiny, but positive. If one now naively adds up the zero-point energy of the electromagnetic field with a cutoff frequency at the Planck scale, one obtains a value that is about 120 orders of magnitude larger. Even with far more modest cutoff energies, a discrepancy of many dozens of orders of magnitude remains. Steven Weinberg, in a famous 1989 review article, described this "cosmological constant problem" as perhaps the worst failure of a theoretical prediction in the history of physics.

And here the Casimir effect enters – as an argument. In countless lectures and books one hears: vacuum energy is real, because the Casimir effect proves it. But if the zero-point energy is real and gravitates, why is its cosmological effect so absurdly small? The Casimir effect seemed to sharpen the problem by nailing down the reality of vacuum energy experimentally.

Jaffe's Objection (2005)

Robert Jaffe of MIT published in 2005 in Physical Review D (volume 72, 021301) a short paper with the sober title "Casimir effect and the quantum vacuum" that cuts through this chain of reasoning. His thesis: the Casimir effect does not prove the reality of vacuum energy. One can compute the Casimir force completely without ever speaking of zero-point energy – as a relativistic, quantum-mechanical force between the fluctuating charges and currents in the two plates, that is, essentially as a retarded van der Waals force, just as Casimir and Polder had originally derived it before Bohr's remark put them on the vacuum trail.

The decisive argument is the dependence on the fine-structure constant α, which measures the strength of the electromagnetic interaction. Casimir's formula does not contain α – but only because it describes the limiting case α → ∞ of perfect conductors. For real materials with finite conductivity the force depends on α, and in the limit α → 0, in which charges no longer interact with the field, the Casimir force vanishes. A force that disappears when one switches off the interaction between matter and field is a force between matter – not a property of empty space alone. In this view the zero-point energy is a convenient computational tool that gives the same answers in the limit of perfect conductors but claims no independent physical reality.

Jaffe by no means concludes from this that vacuum energy does not exist. His statement is more precise and more modest: the Casimir effect is not proof of it. Whether the vacuum possesses a gravitating energy density must be decided by gravity itself – and there the answer is the unsolved cosmological constant problem. Whoever says "Casimir proves vacuum energy, therefore the Λ problem is real" confuses two things that resemble each other formally but differ physically: the difference in field energy under changed boundary conditions (measurable, and measured via Casimir) and the absolute value of the field energy in free space (accessible only through gravity, and puzzling there).

I am of the opinion that this distinction is one of the most important lessons of the whole topic, precisely because it is so often glossed over. The Casimir effect is real, precisely measured, and technically relevant. But what it says about the vacuum is a matter of interpretation in which two mathematically equivalent descriptions – "the vacuum pushes" and "the charges attract each other with a delay" – lead to very different intuitions. Physics does not decide between them; only gravity could, and so far it remains silent.

A Small Table of Perspectives

Both columns give the same number for every experiment ever performed. That is exactly what makes the debate so instructive.

The Central Takeaway

The Casimir effect teaches several things at once, and the most important of them is perhaps not the physical but the methodological one: What one treats as "nothing" is usually just what one has not yet learned to measure. For fifty years the vacuum between two plates was empty because nobody could measure the force. Then it was measured to one percent, and suddenly engineers had to design their micromachines against the vacuum.

The second lesson is Jaffe's: A correct result does not automatically prove the story one uses to explain it. Casimir's formula is right, the measurements confirm it – and yet the popular interpretation "the vacuum has energy, and that is proven" is an over-interpretation, because an entirely different story yields the same formula. Anyone who has a model in their own work that explains the data should regularly ask: is there a second model that explains the same data but claims something different about the world? If so, one has proven less than one believes – and at the same time knows more precisely which experiment one would need next.

And the third lesson, from the ferrofluid episode of 2024/25: The more spectacular a result, the more carefully one must check whether it really is the claimed effect or a neighbor that looks similar. On the nanometer scale in a liquid, half a dozen forces act simultaneously; a sign change in the raw data is not yet a sign change of the Casimir force. This holds in every discipline in which one filters a small effect out of a large background – whether it concerns forces, security vulnerabilities, or performance regressions.

The practical prompt: pick a quantity in your own system that you have so far treated as "negligible" or "zero" – a latency, an error rate, an overhead, an assumption about trust. Ask yourself at what scale (what distance, what load, what data volume) it might come back with the fourth power. And ask yourself whether you could even measure it if it did.

Reflection Question

Casimir and Polder had first derived the force "down to earth" from the interactions of the atoms; only Bohr's casual remark led to the more elegant but interpretively riskier vacuum narrative – which then shaped the popular account for half a century. Which elegant explanation in your own field have you adopted because it is more beautiful than the cumbersome one – and have you ever checked whether both really predict the same thing, or whether elegance merely sold you a certainty the data do not support?

Cross-References in the Vault

• The Accelerating Cosmos: Type Ia Supernovae, Dark Energy, and Whether Einstein's Constant Is Wavering – the observed accelerated expansion and the cosmological constant, whose relation to vacuum energy is discussed in Part 7.
• Why the World Turns Classical: Decoherence, Einselection, and Quantum Darwinism – how the environment (and its fluctuations) affects quantum systems; the vacuum field is the most unavoidable environment of all.
• The Leap Through the Wall: Quantum Tunneling from Gamow's Alpha Decay to the Attosecond Riddle – another effect in which quantum mechanics makes the "impossible" real.
• Resistance Is Futile: Superconductivity from Onnes to BCS to the Room-Temperature Dream – superconductivity, without which the SQUID-based detection of the dynamical Casimir effect would not have been possible.
• Order from Noise: Quantum Error Correction and the Road to a Fault-Tolerant Quantum Computer – superconducting circuits at millikelvin temperatures, the same technology platform as in the Chalmers experiment.
• The Shadow of the Invisible: Black Holes, the Event Horizon, and the Event Horizon Telescope – black holes, whose Hawking radiation is the theoretical relative of the dynamical Casimir effect.
• The Wager Against Reality: Bell's Theorem and the End of Local Realism – another example of how a seemingly purely interpretive question of quantum physics became experimentally tangible.

Sources

• S. K. Lamoreaux, "Demonstration of the Casimir Force in the 0.6 to 6 µm Range," Physical Review Letters 78, 5 (1997): https://ui.adsabs.harvard.edu/abs/1997PhRvL..78....5L/abstract
• U. Mohideen & A. Roy, "Precision Measurement of the Casimir Force from 0.1 to 0.9 µm," Physical Review Letters 81, 4549 (1998); discussion and comment: https://link.aps.org/doi/10.1103/PhysRevLett.83.3340
• H. B. Chan, V. A. Aksyuk, R. N. Kleiman, D. J. Bishop & F. Capasso, "Quantum Mechanical Actuation of Microelectromechanical Systems by the Casimir Force," Science 291, 1941 (2001): https://www.science.org/doi/10.1126/science.1057984
• J. N. Munday, F. Capasso & V. A. Parsegian, "Measured long-range repulsive Casimir-Lifshitz forces," Nature 457, 170 (2009): https://www.nature.com/articles/nature07610
• C. M. Wilson et al., "Observation of the dynamical Casimir effect in a superconducting circuit," Nature 479, 376 (2011): https://www.nature.com/articles/nature10561
• R. L. Jaffe, "Casimir effect and the quantum vacuum," Physical Review D 72, 021301(R) (2005): https://link.aps.org/doi/10.1103/PhysRevD.72.021301
• Y. Zhang et al., "Magnetic-field tuning of the Casimir force," Nature Physics 20, 1282 (2024): https://www.nature.com/articles/s41567-024-02521-0 – and the critical reply M. Moazzami Gudarzi & S. H. Aboutalebi, "Re-examining magnetic tuning of Casimir forces," Nature Physics 21, 520 (2025): https://www.nature.com/articles/s41567-025-02845-5
• Review: "Casimir Effect in MEMS: Materials, Geometries, and Metrologies – A Review," Micromachines (2024): https://pmc.ncbi.nlm.nih.gov/articles/PMC11278474/
• A. Lambrecht, "The Casimir effect: a force from nothing," Physics World (2002): https://physicsworld.com/a/the-casimir-effect-a-force-from-nothing/

← All articles