Monopole Problem

What the problem actually is

Run any plausible grand unified theory through the standard hot Big Bang. At a cosmic temperature of T ~ 10¹⁶ GeV — at cosmic time t ~ 10⁻³⁶ s — the GUT symmetry group must spontaneously break to the Standard Model group. Whenever a continuous symmetry breaks down to a smaller subgroup whose second homotopy class is non-trivial, the field theory contains a topological soliton: a localised, stable, magnetically charged lump. For any simple GUT group that contains the Standard Model — SU(5), SO(10), E₆ — this homotopy condition is satisfied. The 't Hooft-Polyakov monopole (1974) is the explicit solution.

How many of these monopoles are produced? Tom Kibble worked this out in 1976. At the phase transition, the order parameter — the GUT Higgs field — randomises its vacuum direction within each region too small for causal communication. Regions with topologically inequivalent choices cannot match smoothly at their boundaries, and the topological defect is what fills the gap. The minimum density is one monopole per causal horizon volume, which at t ~ 10⁻³⁶ s is a sphere of radius c·t ~ 10⁻²⁸ m. So

n_M(t_GUT) ≳ 1 / (c · t_GUT)³  ≈  10⁸⁰ m⁻³

With each monopole carrying ~10¹⁶ GeV ≈ 2 × 10⁻⁸ kg of rest mass, the energy density at formation is enormous. Trace that density forward through standard FLRW expansion (matter scales as a⁻³, radiation as a⁻⁴). Monopoles are non-relativistic almost immediately after production, so they scale as matter. By recombination (t ~ 4 × 10⁵ yr), the monopole density should be ~10¹⁵ times the observed total matter density of the universe. The universe would have closed under its own weight and recollapsed long before forming stars.

A worked numerical example

Take a fiducial calculation. At the GUT phase transition:

Quantity	Value	Notes
Temperature T_GUT	10¹⁶ GeV ≈ 10²⁹ K	Symmetry breaking scale
Cosmic time t_GUT	~10⁻³⁶ s	From t = (1/H), H ~ T²/M_Pl
Horizon length c·t	~3 × 10⁻²⁸ m	Causally connected size
Horizon volume	~10⁻⁸³ m³	(c·t)³
Monopole density n_M	~10⁸³ m⁻³	≥1 per horizon
Entropy density s	~10⁹⁴ m⁻³	g_*·T³ at GUT
Ratio n_M / s	~10⁻¹¹	Conserved by adiabatic expansion
Today: monopole mass density	~10⁻²² kg/m³	If preserved
Critical density today	~9 × 10⁻²⁷ kg/m³	For comparison

The mass density today would be ~10¹⁵ times the critical density. That is a universe that has collapsed back to a singularity many times over. The mismatch with observation is not subtle.

The Parker bound: galactic magnetic fields know there are no monopoles

Even before inflation was proposed, observational limits on the monopole flux through Earth were stringent. Eugene Parker noticed in 1970 that magnetic monopoles, if present in the galaxy, would systematically reduce the galactic magnetic field by sliding along the field lines and converting magnetic field energy into monopole kinetic energy. The galactic field is observed to have strength ~3 μG and a coherence time of order the galactic rotation period, ~10⁹ years. For the field not to have been short-circuited by monopoles, the monopole flux must satisfy

F < F_Parker ≈ 10⁻¹⁵ cm⁻² s⁻¹ sr⁻¹

This corresponds to a present-day monopole number density less than ~10⁻²⁰ m⁻³ — orders of magnitude below what a naive GUT prediction would give. The 'extended Parker bound', applied to magnetic fields in galaxy clusters with longer survival times, tightens this by another order of magnitude. Direct searches confirm: MACRO at Gran Sasso (1989-2000) found zero candidates in 5.6 × 10⁻¹⁶ cm⁻² s⁻¹ sr⁻¹ of accumulated exposure, comfortably below Parker. IceCube has reached ~10⁻¹⁸ cm⁻² s⁻¹ sr⁻¹. Every search has come back empty.

How inflation removes the problem

Alan Guth's 1981 inflationary universe paper had three problems in its motivation — horizon, flatness, monopole — but the monopole problem was the one he started from. The mechanism is simple. Suppose that for a brief window — perhaps 10⁻³⁵ s to 10⁻³² s after the Big Bang — the universe is in a metastable vacuum dominated by the potential energy of some scalar field, and that this vacuum drives exponential expansion a(t) ∝ exp(Ht). After N e-folds, every comoving volume has grown by a factor e^N in linear size, or e^3N in volume.

Any number density established before inflation gets divided by e^3N. The standard target is N ~ 60, motivated by the horizon problem; this gives a dilution factor of

dilution = e^(−180) ≈ 10⁻⁷⁸

Apply that to the GUT monopole density of ~10⁸³ m⁻³ at formation. After dilution to today's scale factor, the present-day number density is ~10²⁹ × 10⁻⁷⁸ × (today / inflation scale factor adjustments) ≪ 1 per observable universe. Even N = 50 — the minimum for a self-consistent inflationary scenario — gives a dilution of e^(−150) ~ 10⁻⁶⁵, more than enough.

The constraint becomes: inflation must end before the GUT phase transition can run again. If reheating after inflation exceeds T_GUT ~ 10¹⁶ GeV, monopoles are re-produced thermally and the problem returns. Standard slow-roll inflation gives T_RH well below 10¹⁵ GeV — comfortably safe.

Topological defects, ranked by danger

Defect	From	Cosmological hazard if produced	Status today
Domain wall	Discrete symmetry breaking, π₀ ≠ 0	Worst — energy density grows as a⁻¹, immediately dominates	Excluded; must be diluted
Magnetic monopole	Continuous → U(1) breaking, π₂ ≠ 0	Severe — overdense by 10¹⁵	Diluted by inflation; unseen
Cosmic string	U(1) breaking, π₁ ≠ 0	Marginal — ρ_string ∝ a⁻², similar to radiation	Bounded by CMB (Gμ ≤ 10⁻⁷); searches ongoing
Texture	Higher symmetry breaking, π₃ ≠ 0	Mild — unwind and radiate	Ruled out as primary structure source
Skyrmion	Chiral symmetry breaking, σ-model	Specialty; nuclear physics analogue	Used in baryon counting; cosmologically rare

Domain walls and monopoles together are the dangerous pair: their energy density falls slower than the universe expands, so any meaningful production rapidly dominates the dynamics. Inflation is the standard defense against both. Cosmic strings are the interesting middle case — they would not have been catastrophic, and they could plausibly form after inflation ends (during reheating), if the appropriate U(1) symmetry is broken at that scale.

Dirac's quantization condition: why monopoles are theoretically required

Long before any GUT, Paul Dirac argued in 1931 that magnetic monopoles must exist as a consistency condition of quantum mechanics applied to charged particles in arbitrary magnetic fields. Consider a quantum particle of electric charge e moving in the field of a magnetic monopole of charge g. The vector potential A_μ describing the monopole's field has a 'Dirac string' — a singular line along which the gauge potential is not defined. For the resulting wave function to be single-valued around the string, the integrated phase shift e·∮A·dl must be a multiple of 2π. Working this out gives

e · g = n · ℏc / 2,   n ∈ ℤ

This is the Dirac quantization condition. It has two extraordinary consequences. First, if monopoles exist anywhere, electric charge is automatically quantized — this matches the experimental fact that |q_proton| / |q_electron| − 1 < 10⁻²¹. Second, the minimum magnetic charge g_D = ℏc/(2e) is gigantic, roughly 68.5 times the elementary charge in Gaussian units. A single Dirac monopole would carry an enormous magnetic flux quantum and ionize anything it passed through.

History: Cornell 1979, Guth's notebook, and Cabrera's lonely event

The chain of events is well documented because it produced inflation. In 1979 Alan Guth, a postdoc at Cornell, was working with Henry Tye on grand unified theories. They asked the natural question: are the cosmological consequences self-consistent? The answer was no. The Kibble mechanism gives 10⁻¹⁰ monopoles per photon at formation; the present-day baryon-to-photon ratio is ~10⁻⁹; so monopole and baryon densities should be comparable today, with monopole masses 10¹⁹ times that of the proton, giving monopole mass density 10¹⁰ times baryon mass density. Universe closes in seconds. They sent the paper around with a note saying 'something has to give'.

Guth then asked: what if the GUT phase transition were first-order, with the universe getting stuck in a metastable vacuum long enough to supercool below T_GUT? Once the universe finally tunnels to the true vacuum, you would still get monopoles, but the supercooled phase would inflate every relic away. He worked the math in a notebook in December 1979 and titled the entry 'spectacular realization'. The paper appeared in January 1981.

One year later, on Valentine's Day 1982, Blas Cabrera at Stanford recorded what looked like exactly one monopole passing through his superconducting loop. The signal was a sudden jump of one flux quantum, exactly the value predicted by Dirac. No other detector reproduced it. Cabrera himself has been agnostic about the event ever since. The likeliest explanation is a glitch — a thermal expansion shifting the superconducting ring — but the original interpretation could not be definitively excluded. The graph in his lab notebook of that single jump is one of the more famous artifacts of late-twentieth-century physics.

Where the searches stand today

• MoEDAL. Operates at the LHC's Point 8 since 2010, instruments around the LHCb interaction region. Looks for monopole pair production in proton-proton collisions, sensitive to monopole masses up to ~3 TeV. Has set the world's best collider bounds; no candidates.
• IceCube. One cubic kilometre of Antarctic ice instrumented with photomultipliers. A relativistic monopole would Cherenkov-light a track far brighter than a muon's. Sensitivity reaches flux ~10⁻¹⁸ cm⁻² s⁻¹ sr⁻¹ for v > 0.5c. No candidates.
• SLIM, ANITA, RICE. Specialised searches in mica, balloons, radio. All null.
• Cosmic-ray air showers. Auger and Telescope Array look for the distinctive high-energy showers a relativistic monopole would make. None seen.
• Cabrera-style superconducting loops. Several modern variants run continuously. None has reproduced the 1982 event.

Combined, current experiments rule out flux above F ~ 10⁻¹⁸ cm⁻² s⁻¹ sr⁻¹ over a broad mass range — strictly consistent with the inflationary prediction that the present-day monopole density is far less than one per observable universe.

Common pitfalls

• Conflating Dirac and 't Hooft-Polyakov monopoles. Dirac monopoles are heuristic point sources required by quantum-mechanical self-consistency given electric charge quantization. 't Hooft-Polyakov monopoles are explicit topological soliton solutions of specific gauge theories, with finite mass set by the symmetry-breaking scale. The cosmological problem concerns the latter.
• Thinking inflation produces monopoles. Inflation is what dilutes them. If reheating after inflation exceeded T_GUT, monopoles would be re-produced thermally. Standard slow-roll inflation has T_RH ≪ 10¹⁵ GeV, so this does not happen.
• Assuming the Parker bound rules out all monopoles. Parker bounds the flux at Earth. Many fewer monopoles would still evade detection but matter cosmologically. The cosmological bound from Ω_M is tighter than Parker by orders of magnitude.
• Treating the 1982 Cabrera event as confirmation. A single unrepeated event in a single detector is not evidence for a particle. The community's consensus is that it was almost certainly a glitch, though no definitive alternative explanation exists.
• Assuming all GUTs have the monopole problem equally. Some GUT-like models (notably those broken at scales below the inflation reheating temperature) avoid the issue entirely. The problem is specific to GUTs that break above the reheating temperature.