Module 9: Dark energy · Lesson 9.3 · about 16 min
The vacuum energy puzzle
Quantum theory says empty space holds energy that acts like a cosmological constant, but the estimate is about 10¹²⁰ times too big. The worst mismatch in physics.
Module 9 · about 16 minutes. Read first: 9.2 Einstein’s cosmological constant returns and 3.3 E = mc². Lesson 9.2 described Λ as an energy of empty space. Quantum physics says empty space really does hold energy. The trouble is how much. By the end you will be able to make the famous estimate yourself and explain why it is called the worst prediction in physics.
1. Empty space is not empty
Take a pendulum and remove every bit of energy you can. Classically it would hang perfectly still. Quantum mechanics forbids that: a position known exactly and a speed known exactly cannot go together (the uncertainty principle). So even in its lowest state the pendulum keeps a little jitter, and a little energy, the zero-point energy:
- ħ
- the reduced Planck constant, 1.055 × 10−34 J s: the size of quantum effects
- ω
- how fast the oscillator swings (angular frequency)
- Physical meaning
- The lowest possible energy is not zero. Every way something can wobble keeps half a "quantum" of energy for ever.
In modern physics every particle is a ripple in a field that fills space: the electromagnetic field for light, the electron field, and so on. Each field is like a mattress of countless tiny springs, one for every possible wave, and each keeps its zero-point jitter. So “empty” space should hold energy in every cubic metre.
And it would be exactly the right kind. Empty space must look the same to everyone, however they move (if the vacuum had a “rest frame”, you could measure your speed against nothing). The only energy and pressure that look identical to every moving observer have p = −ρc². So vacuum energy behaves exactly like Einstein’s Λ. On paper, dark energy is explained. Then you try to work out how much.
2. Adding it up
Each wave with frequency ω contributes ½ħω, and there are more and more possible waves at higher frequencies. Counting all of them up to some highest energy Ec, a “cut-off” beyond which we do not trust our theories, gives roughly:
- Ec
- the cut-off: the highest energy the theory is trusted to. The natural top end is the Planck energy, 1.22 × 1028 eV, where gravity itself needs a quantum theory
- Power of 4
- three powers from counting waves in three dimensions, one from the energy of each wave. Ten times the cut-off gives ten thousand times the energy
- Physical meaning
- This is one field's share; real fields add more, with different signs. It is an order-of-magnitude estimate, not a precise prediction, but no known physics makes it small.
With the Planck energy as the cut-off, the estimate is about 10111 J/m³. Dark energy is measured at about 5 × 10−10 J/m³. The gap is about 120 orders of magnitude. Even if new physics cuts the counting off at around 1 TeV, the energy reached by the Large Hadron Collider, the gap is still about 60 orders of magnitude.
Turn it round: the measured dark energy corresponds to an energy scale of only about 2 milli-electronvolts, hundreds of millions of times less than the energy locked in an electron’s mass. Nothing in particle physics lives there.
3. Is vacuum energy real?
Its effects are measured. Put two smooth metal plates very close together in a vacuum. Between them only waves that fit the gap are allowed; outside, all waves are. The difference in zero-point energy pulls the plates together: the Casimir effect, predicted by Hendrik Casimir in 1948 and measured accurately by Steve Lamoreaux in 1997. Vacuum jitter also shifts the energy levels of hydrogen slightly (the Lamb shift, measured in 1947).
So the jitter is real. What nobody knows is how the vacuum’s energy gravitates. Laboratories only ever measure energy differences, which do not care about a constant added everywhere. Gravity is the one thing that feels the total, and the total it feels is tiny.
4. Ways out
None of these is settled. They are the main ideas people work on:
- Something cancels it. Some unknown symmetry could force the total to exactly zero, with dark energy something else (Lesson 9.4). Supersymmetry would cancel matter fields against force fields, but if it exists at all it is broken above about a TeV, which still leaves the 60 orders.
- It is huge elsewhere, and we live where it is small (anthropic). In 1987 Steven Weinberg showed that if the vacuum energy were bigger than about a few hundred times the density of matter today, the expansion would have sped up before gravity could pull matter into galaxies, and nobody would be here to measure it. If many regions of a larger universe had different values (Lesson 11.7), observers would find a small one. He predicted a small, non-zero value a decade before the supernovae.
- Gravity is different on the largest scales. Some theories change general relativity so that a constant vacuum energy stops curving space-time. None works fully yet.
The worst prediction in physics.
- 1948 Hendrik Casimir predicts the force between plates from vacuum energy.
- 1967-68 Yakov Zel'dovich points out that quantum vacuum energy should act as a cosmological constant, and that the estimate is absurdly large.
- 1987 Steven Weinberg's anthropic bound.
- 1989 Weinberg's review "The cosmological constant problem" makes it famous.
- 1997 Lamoreaux measures the Casimir force to about 5%.
- 1998 Dark energy is measured: not zero, but 10120 times below the estimate.
“Quantum theory predicted dark energy and got the number wrong by 10120.” It is fairer to say we have no prediction at all. The estimate shows that the natural size is enormous; nobody knows the right calculation. Before 1998 most physicists expected the true answer to be exactly zero, by some unknown cancellation. Finding a small value that is not zero made the puzzle harder, not easier.
- Quantum fields keep a zero-point jitter, E0 = ½ħω per wave, so empty space holds energy.
- Vacuum energy must look the same to every observer, so p = −ρc²: it acts exactly like a cosmological constant.
- The rough estimate ρc² ≈ Ec⁴/(16π²(ħc)³) overshoots the measured value by about 120 orders of magnitude with a Planck cut-off, about 60 with a TeV cut-off.
- The measured dark energy corresponds to an energy of only about 2 meV.
- Vacuum effects (Casimir force, Lamb shift) are real; how the vacuum’s total energy gravitates is the unsolved part. Weinberg’s 1987 anthropic bound foresaw a small, non-zero value.
Check yourself
Sources
- S. Weinberg (1989), The cosmological constant problem, Reviews of Modern Physics 61, 1 https://doi.org/10.1103/RevModPhys.61.1
- S. Weinberg (1987), Anthropic bound on the cosmological constant, Physical Review Letters 59, 2607 https://doi.org/10.1103/PhysRevLett.59.2607
- S. K. Lamoreaux (1997), Demonstration of the Casimir force in the 0.6 to 6 μm range, Physical Review Letters 78, 5 https://doi.org/10.1103/PhysRevLett.78.5
- S. M. Carroll (2001), The cosmological constant, Living Reviews in Relativity 4, 1 https://doi.org/10.12942/lrr-2001-1
- J. Martin (2012), Everything you always wanted to know about the cosmological constant problem (but were afraid to ask), Comptes Rendus Physique 13, 566 https://doi.org/10.1016/j.crhy.2012.04.008
Your ticks stay in this browser. Save them as a file on the course home.
Updated 2026-11-22. Numbers are rounded for teaching; measured values come from the sources listed in each lesson. Simulations run in your browser; nothing you type is sent anywhere.