Monopole and Relic Problems

Grand unified theories are the most compelling extrapolation particle physics offers: the three Standard-Model couplings, run up in energy, very nearly meet near $M_{\mathrm{GUT}}\sim10^{16}\,\mathrm{GeV}$ , inviting a simple gauge group — $SU(5)$ , $SO(10)$ — that breaks to the Standard Model in the first instant of the hot Big Bang. But that elegant unification carries a cosmological time-bomb. Whenever a simple group breaks to a subgroup containing an unbroken electromagnetic $U(1)$ — which it must, since electromagnetism survives today — the vacuum manifold acquires a non-trivial second homotopy group, $\pi_2\neq0$ . Topology then guarantees stable, point-like knots of trapped field carrying quantized magnetic charge: 't Hooft–Polyakov magnetic monopoles, each carrying $\sim10^{16}$ proton masses. The Kibble mechanism fixes how many form. Because the vacuum cannot be correlated across regions never in causal contact, at least one defect is trapped per horizon at the transition. The GUT horizon is minuscule — about $10^{-27}\,\mathrm{cm}$ — so "one per horizon" is a staggering density. Once formed they are non-relativistic and absolutely stable; their comoving number freezes, so they redshift as matter ( $n_M\propto a^{-3}$ ) while radiation thins faster ( $a^{-4}$ ). Their share of the budget can only grow, and extrapolated to today it overshoots the critical density by more than eleven orders of magnitude. That flat contradiction with the observed near-critical Universe is the monopole problem — the archetype of a class of relic catastrophes, and its resolution, cosmic inflation, is one of the three pillars on which the inflationary paradigm was built.

Key takeaways

Grand unified theories predict a symmetry-breaking transition at $T\sim10^{16}\,\mathrm{GeV}$ whose vacuum topology ( $\pi_2\neq0$ ) forces stable magnetic monopoles of mass $m_M\sim10^{16}$ – $10^{17}\,\mathrm{GeV}$ , roughly $10^{16}$ proton masses.
The Kibble mechanism sets the abundance at $\gtrsim1$ monopole per causal horizon, and because the GUT horizon is tiny the number density is enormous.
Monopoles are non-relativistic and stable, so $n_M\propto a^{-3}$ while radiation dilutes as $a^{-4}$ — their share of the budget grows with time, so $\Omega_M$ overshoots critical today by at least eleven orders of magnitude, contradicting the observed $\Omega\approx1$ .
The same logic condemns any stable superheavy relic formed with horizon-scale abundance — gravitinos, moduli, cosmic strings, domain walls: the general relic problem.
Cosmic inflation cures it by diluting any pre-inflationary relic density by $e^{-3N}\sim10^{-78}$ for $N\gtrsim60$ e-folds, provided reheating stays below $M_{\mathrm{GUT}}$ so monopoles are not regenerated.
Key idea: A stable, superheavy relic formed at roughly one per causal horizon at $T\sim10^{16}\,\mathrm{GeV}$ dilutes only as $a^{-3}$ ; because the GUT horizon is minuscule and the relic mass is huge, its present density overshoots critical by $\sim10^{11}$ unless a phase of accelerated expansion dilutes it away.

Where this lives today

The monopole and relic problems remain load-bearing in modern cosmology. On the theory side, the reheating bound $T_{\mathrm{reheat}}\lesssim M_{\mathrm{GUT}}$ constrains inflationary model-building, and its cousin — the gravitino problem — caps $T_{\mathrm{reheat}}\lesssim10^{9}\,\mathrm{GeV}$ in many supergravity scenarios, shaping viable baryogenesis and leptogenesis. Experimentally, the search continues: MoEDAL at the LHC hunts collider-produced monopoles with nuclear-track detectors and trapping arrays; IceCube and ANITA set cosmic-monopole flux limits orders of magnitude below the Parker bound; and the legacy MACRO null result still anchors the induction-search limits that began with Blas Cabrera's single unreproduced candidate event in 1982. Cosmic strings, the one-dimensional cousin, are a live target for pulsar-timing arrays and gravitational-wave observatories constraining the tension $G\mu$ . And the framework feeds forward: the same expansion that sweeps away monopoles stretches quantum fluctuations into the CMB perturbations, so the monopole problem is inseparable from the observational case for inflation.

The physical picture

Picture the GUT Higgs field as an arrow that, once the symmetry breaks, must pick a direction in an internal space. Regions that have never exchanged a light signal cannot agree which way to point, so each causal patch settles into an independent, randomly oriented vacuum. Where patches meet, the arrows can be forced into a configuration no smooth deformation can flatten — a topological knot. The simplest is the hedgehog: arrows pointing radially outward, un-combable just as the hair on a sphere is. That trapped knot, carrying quantized magnetic charge, is the monopole. Since patches are horizon-sized, you get roughly one knot per horizon — the Kibble estimate.

Figure

Kibble mechanism: causally disconnected patches (each about one GUT horizon, d_H ∼ 10⁻²⁷ cm at T ∼ 10¹⁶ GeV) pick random vacuum directions; where the winding cannot be combed flat a monopole is trapped, giving ≳1 per horizon and a number density n_M ∼ 1/d_H³ ∼ T⁶/M_Pl³

The lethal step is arithmetic, not principle. "One per horizon" sounds modest until you see how small the GUT horizon is: at $T\sim10^{16}\,\mathrm{GeV}$ the causal distance is $\sim10^{-27}\,\mathrm{cm}$ , so a horizon fits in a volume $10^{81}$ times smaller than a cubic centimetre. Multiply that density by $\sim10^{16}$ proton masses each, dilute only as matter, and the monopoles dominate everything. A WIMP escapes this fate through annihilation — it stays in equilibrium and is Boltzmann-suppressed to a tiny freeze-out abundance. Monopoles are too rare and heavy to find their antiparticles, so annihilation switches off at once and their number is fixed by geometry, not thermodynamics.

Figure

The overclosure catastrophe and its cure: with no inflation the frozen Kibble abundance gives Ω_M ∼ 10¹¹ (overshooting critical Ω = 1 by at least eleven orders of magnitude), whereas inflation multiplies the relic number density by e^(−3N) ∼ 10⁻⁷⁸ for N ≳ 60 e-folds, driving Ω_M far below one

Inflation cures the disease at its root: it makes the entire observable Universe descend from a single pre-inflationary causal patch — one that held at most a handful of monopoles. Stretching that patch by a linear factor $e^{60}$ dilutes even that handful to far less than one monopole in the whole visible Universe. The same expansion that flattens space and seeds galaxies sweeps the monopoles away — a triple payoff that is much of why inflation was believed so quickly.

The core equations

The chain is: topology guarantees the defects; 't Hooft–Polyakov fixes their mass; Dirac fixes their charge; Kibble fixes their number; entropy conservation freezes it; and converting to a density parameter exposes the catastrophe. Each equation is one link. The monopole's existence is not a dynamical accident but a topological theorem, depending only on the shape of the vacuum manifold $G/H$ , not the potential.

\pi_2(G/H)\neq 0 \;\Longrightarrow\; \text{stable magnetic monopoles} \qquad (1)

where: $G$ — the GUT gauge group (e.g. $SU(5)$ ); $H$ — the residual group after breaking, containing $U(1)_{\mathrm{em}}$ ; $\pi_2$ — the second homotopy group of $G/H$ , counting distinct field configurations at spatial infinity. A non-trivial $\pi_2$ means some configurations cannot be unwound — those are the monopoles, and their charge is the winding number. Because the monopole is a soliton of size the inverse breaking scale, its rest mass is that scale divided by the coupling.

m_M \sim \frac{4\pi}{g^2}\,M_{\mathrm{GUT}} \sim \frac{M_{\mathrm{GUT}}}{\alpha_{\mathrm{GUT}}} \sim 10^{16}\text{–}10^{17}\,\mathrm{GeV} \qquad (2)

where: $m_M$ — monopole rest mass; $g$ — GUT gauge coupling; $\alpha_{\mathrm{GUT}}=g^2/4\pi\sim1/40$ ; $M_{\mathrm{GUT}}\sim10^{16}\,\mathrm{GeV}$ . Reading it: heavier than unification by the inverse coupling, so $m_M$ is a few $\times10^{17}\,\mathrm{GeV}$ — about $10^{-8}\,\mathrm{g}$ , macroscopic on the particle-physics ladder. The magnetic charge is not free: quantum consistency forces it to a fixed multiple of the fundamental unit — the Dirac quantization condition.

g_M = \frac{2\pi}{e} = \frac{e}{2\alpha}\quad\Longrightarrow\quad g_M \approx 68.5\,e \qquad (3)

where: $g_M$ — magnetic charge; $e$ — elementary electric charge; $\alpha\approx1/137$ . Reading it: a monopole carries $\sim68.5$ Dirac units, so its electromagnetic coupling is enormous — it ionizes heavily and drains magnetic fields, the basis of every search strategy. The abundance is geometric: each horizon-sized patch traps of order one defect, so the density is one per horizon volume.

n_M \sim \frac{1}{d_H^{3}},\qquad d_H \sim H^{-1}\sim \frac{M_{\mathrm{Pl}}}{T_{\mathrm{GUT}}^{2}}\;\Longrightarrow\; n_M \sim \frac{T_{\mathrm{GUT}}^{6}}{M_{\mathrm{Pl}}^{3}} \qquad (4)

where: $n_M$ — number density at formation; $d_H$ — causal horizon; $H$ — Hubble rate; $T_{\mathrm{GUT}}$ — transition temperature; $M_{\mathrm{Pl}}\approx1.22\times10^{19}\,\mathrm{GeV}$ . Reading it: in radiation domination $H\propto T^2/M_{\mathrm{Pl}}$ , so the horizon is small and its inverse cube — the density — is large. Both the monopole number and the entropy dilute as $a^{-3}$ , so their ratio is a comoving invariant, fixed at formation and carried to the present.

Y_M\equiv\frac{n_M}{s} \sim \frac{T_{\mathrm{GUT}}^{6}/M_{\mathrm{Pl}}^{3}}{\tfrac{2\pi^2}{45}g_{*s}T_{\mathrm{GUT}}^{3}} \sim \frac{1}{g_{*s}}\!\left(\frac{T_{\mathrm{GUT}}}{M_{\mathrm{Pl}}}\right)^{3}\sim 10^{-11} \qquad (5)

where: $s=\tfrac{2\pi^2}{45}g_{*s}T^3$ — entropy density; $g_{*s}\sim100$ ; $Y_M$ — conserved comoving abundance. Reading it: with $T_{\mathrm{GUT}}/M_{\mathrm{Pl}}\sim10^{-3}$ the yield is $\sim10^{-11}$ — one monopole per $\sim10^{11}$ entropy quanta, frozen for the age of the Universe. Multiplying the frozen yield by the present entropy and the monopole mass, then dividing by critical density, gives the number that decides everything.

\Omega_M h^2 = \frac{m_M\,Y_M\,s_0}{\rho_{c}/h^2} \sim 10^{11}\text{–}10^{13} \qquad (6)

where: $\Omega_M$ — density in units of critical; $s_0\approx2.9\times10^{3}\,\mathrm{cm^{-3}}$ ; $\rho_c\approx1.05\times10^{-5}h^2\,\mathrm{GeV\,cm^{-3}}$ ; $h$ — reduced Hubble constant. Reading it: it overshoots $\Omega\approx1$ by at least eleven orders of magnitude — the monopole problem in one line (the exponent is cubically sensitive to $T_{\mathrm{GUT}}$ ; see Derivation 1). Observation confirms their absence independently through the Parker bound: monopoles must not drain the Galactic magnetic field faster than the dynamo rebuilds it.

F_M = \frac{n_M\,v}{4\pi} \lesssim 10^{-15}\,\mathrm{cm^{-2}\,s^{-1}\,sr^{-1}} \qquad (7)

where: $F_M$ — monopole flux; $v\sim10^{-3}c$ — typical Galactic monopole speed. Reading it: the allowed flux is astronomically below the naive GUT prediction, so whatever the history, it left essentially no monopoles. Inflation supplies the missing dilution: quasi-exponential expansion multiplies every pre-existing number density by the inverse cube of the expansion.

n_M \to n_M\,e^{-3N},\qquad N=\ln\frac{a_{\mathrm{end}}}{a_{\mathrm{start}}}\gtrsim 60 \qquad (8)

where: $N$ — number of e-folds; $e^{-3N}\sim10^{-78}$ for $N=60$ — the volume dilution factor. Reading it: the same $\gtrsim60$ e-folds demanded by horizon and flatness suppress the relic density by $\sim10^{78}$ , far more than enough.

Derivation AFrom the Kibble abundance to the overclosure catastrophe

The goal is to carry the conserved abundance $Y_M=n_M/s$ from the GUT transition to the present and convert it to $\Omega_M$ , making the eleven-order overshoot fall out of standard inputs. We work in natural units, estimate the horizon from the radiation-dominated Friedmann equation, impose one defect per horizon, freeze the ratio to entropy, and cash it out today with measured $s_0$ , $\rho_c$ , and $m_M$ .

Step 1. In the radiation era the Friedmann equation gives $H^2=\tfrac{8\pi G}{3}\rho$ with $\rho=\tfrac{\pi^2}{30}g_*T^4$ , so $H\sim g_*^{1/2}T^2/M_{\mathrm{Pl}}$ and $d_H\sim H^{-1}\sim M_{\mathrm{Pl}}/T_{\mathrm{GUT}}^2$ . At $T_{\mathrm{GUT}}=10^{16}\,\mathrm{GeV}$ this is $1.2\times10^{-13}\,\mathrm{GeV^{-1}}=2.4\times10^{-27}\,\mathrm{cm}$ . Why: the horizon is the only length scale at the transition, setting the coarseness of the vacuum pattern.

Step 2. Kibble traps of order one defect per horizon volume, so $n_M\sim d_H^{-3}\sim T_{\mathrm{GUT}}^6/M_{\mathrm{Pl}}^3$ — geometric, not thermal. Why: disconnected patches cannot align their vacua, so one un-combable winding survives per patch — a lower bound, not dialable.

Step 3. Divide by $s=\tfrac{2\pi^2}{45}g_{*s}T_{\mathrm{GUT}}^3$ to form $Y_M=n_M/s\sim g_{*s}^{-1}(T_{\mathrm{GUT}}/M_{\mathrm{Pl}})^3$ . With $T_{\mathrm{GUT}}/M_{\mathrm{Pl}}=8.2\times10^{-4}$ and $g_{*s}\sim100$ this is $Y_M\sim5\times10^{-12}\approx10^{-11}$ . Why: both $n_M$ and $s$ redshift as $a^{-3}$ , so their ratio is conserved — freezing the abundance.

Step 4. Present energy density $\rho_{M,0}=m_M\,Y_M\,s_0$ . With $m_M\sim10^{16}\,\mathrm{GeV}$ , $Y_M\sim10^{-11}$ , $s_0\approx2.9\times10^3\,\mathrm{cm^{-3}}$ this is $\sim3\times10^{8}\,\mathrm{GeV\,cm^{-3}}$ . Why: today's number is the frozen yield times present entropy, and each monopole contributes its full rest mass.

Step 5. Divide by $\rho_c\approx1.05\times10^{-5}h^2\,\mathrm{GeV\,cm^{-3}}$ : $\Omega_M\sim3\times10^{8}/(1.05\times10^{-5}h^2)\sim1.5\times10^{13}h^{-2}$ . A slightly lower scale $T_{\mathrm{GUT}}\approx2\times10^{15}\,\mathrm{GeV}$ drops this cubically to the widely quoted $\Omega_M\sim10^{11}$ . Either way the Universe is overclosed by at least eleven orders of magnitude. Why: all that matters is that the number is astronomically above one; no reasonable input rescues $\Omega\approx1$ .

Result.

\Omega_M \sim \frac{m_M\,s_0}{\rho_c}\left(\frac{T_{\mathrm{GUT}}}{M_{\mathrm{Pl}}}\right)^3\frac{1}{g_{*s}}\;\sim\;10^{11}\text{–}10^{13}\;\gg\;1 \qquad\text{(monopole catastrophe)}

The literal plug-in at $T_{\mathrm{GUT}}=10^{16}\,\mathrm{GeV}$ gives $1.5\times10^{13}h^{-2}$ , while the canonical $\Omega_M\sim10^{11}$ corresponds to $T_{\mathrm{GUT}}\approx2\times10^{15}\,\mathrm{GeV}$ — both mean overclosed by at least eleven orders of magnitude.

Go deeper · Derivation B · 1 minWhy monopole annihilation cannot rescue the abundance

Derivation BWhy monopole annihilation cannot rescue the abundance

The goal is to show that annihilation never brings the abundance below its Kibble value, because the annihilation rate per monopole is below the Hubble rate essentially at formation. This distinguishes a topological relic from a thermal one: for a WIMP, equilibrium annihilation Boltzmann-suppresses the abundance to a tiny freeze-out value, whereas here $\Gamma/H\ll1$ from the outset, so no such suppression operates.

Step 1. Even with the strong magnetic Coulomb attraction, the annihilation cross-section is bounded by the inverse-mass-squared scale, $\sigma\sim\beta/m_M^2$ with $\beta$ order unity. Because $m_M\sim M_{\mathrm{GUT}}$ is huge, $\sigma$ is tiny. Why: a soliton of mass $m_M$ has size $\sim1/m_M$ , so its cross-section for finding a partner is $\sim1/m_M^2$ .

Step 2. Form the ratio with the Kibble density and $H\sim T^2/M_{\mathrm{Pl}}$ : $\Gamma/H=n_M\langle\sigma v\rangle/H\sim(T^6/M_{\mathrm{Pl}}^3)(1/m_M^2)/(T^2/M_{\mathrm{Pl}})\sim T^4/(m_M^2M_{\mathrm{Pl}}^2)$ . Why: annihilation matters only while partners meet faster than the Universe expands; $\Gamma$ vs $H$ is the freeze-out test.

Step 3. At formation, $T\sim T_{\mathrm{GUT}}\sim m_M$ : $\Gamma/H\sim(T_{\mathrm{GUT}}/M_{\mathrm{Pl}})^2\sim10^{-6}$ , far below one, and it falls further as $T^4$ . Why: monopoles are already too dilute to annihilate the instant they appear — there is no equilibrium epoch.

Step 4. With $\Gamma<H$ from the start, the WIMP freeze-out logic does not apply; the comoving number is fixed by topology, staying at the Derivation 1 value. Why: this is why the problem is robust — you cannot annihilate your way out of it.

Result.

\frac{\Gamma}{H}\bigg|_{T\sim T_{\mathrm{GUT}}}\sim\left(\frac{T_{\mathrm{GUT}}}{M_{\mathrm{Pl}}}\right)^{2}\sim10^{-6}\ll1 \qquad\text{(annihilation frozen)}

Go deeper · Derivation C · 1 minHow many e-folds of inflation are needed

Derivation CHow many e-folds of inflation are needed

The goal is to compute how much inflation sweeps the observable Universe clean of monopoles, and to show it coincides with the amount already demanded by the horizon and flatness problems. Inflation multiplies volumes by $e^{3N}$ , so any pre-existing density falls by $e^{-3N}$ ; we quantify $N$ , evaluate the suppression, and note the reheating condition that keeps the cure intact.

Step 1. The scale factor grows as $a\propto e^{Ht}$ , so after $N=\ln(a_{\mathrm{end}}/a_{\mathrm{start}})$ e-folds a number density dilutes as $n_M\to n_M\,e^{-3N}$ . Why: number is conserved while volume grows as $a^3=e^{3N}$ .

Step 2. The horizon and flatness problems independently require the observable Universe to trace back to one causal patch, fixing $N\gtrsim\ln(T_{\mathrm{reheat}}/T_0)\sim60$ . Why: the CMB's uniformity across the sky demands our whole Hubble volume descend from a single pre-inflationary patch — the same $\sim60$ e-folds that flatten space.

Step 3. At the minimal $N=60$ : $e^{-3N}=e^{-180}\approx10^{-78}$ , leaving far less than one monopole in the entire observable Universe. Why: $10^{-78}$ dwarfs the $10^{11}$ – $10^{13}$ overshoot by sixty-odd orders, so the cure is overwhelming.

Step 4. The cure survives only if reheating does not restore the GUT symmetry: $T_{\mathrm{reheat}}\lesssim M_{\mathrm{GUT}}$ , else the transition and monopoles recur. Why: reheating above $M_{\mathrm{GUT}}$ re-crosses the transition and undoes the dilution — a genuine upper bound on $T_{\mathrm{reheat}}$ .

Result.

N\gtrsim60\;\Longrightarrow\;n_M\,e^{-3N}V_{\mathrm{obs}}\ll1,\qquad e^{-180}\approx10^{-78} \qquad\text{(dilution to zero)}

Numerically, $e^{-180}=6.7\times10^{-79}\approx10^{-78}$ , and the same $N\approx60$ e-folds that solve the horizon and flatness problems supply this suppression.

Limiting cases

Case	Result	Meaning
No inflation (naive GUT)	$\Omega_M\sim10^{11}\text{–}10^{13}$	Kibble abundance diluted only by ordinary $a^{-3}$ expansion overcloses the Universe by at least eleven orders of magnitude.
Lower GUT scale	$Y_M\propto(T_{\mathrm{GUT}}/M_{\mathrm{Pl}})^3$	Reducing $T_{\mathrm{GUT}}$ shrinks the abundance cubically, but cannot reach $\Omega\lesssim1$ without spoiling gauge-coupling unification.
Inflation, $T_{\mathrm{reheat}}<M_{\mathrm{GUT}}$	$\Omega_M\to\Omega_M\,e^{-3N}\approx0$	Monopoles diluted below one per observable Universe and never regenerated: problem solved.
Inflation, $T_{\mathrm{reheat}}>M_{\mathrm{GUT}}$	monopoles re-form after reheating	Restored GUT symmetry re-triggers the transition; this caps the allowed reheat temperature.
General heavy relic	$\Omega_X\propto m_X Y_X$	Any stable superheavy state (gravitino, modulus, defect) with horizon-scale abundance faces the same overclosure.

Common trapassuming monopoles annihilate away like a thermal relic

The most common error is to treat the monopole abundance like a WIMP's, imagining that annihilation Boltzmann-suppresses it to a small freeze-out value. It does not. Annihilation requires the species to first be dense enough that $\Gamma=n_M\langle\sigma v\rangle$ exceeds $H$ — but for monopoles $\Gamma/H\sim(T_{\mathrm{GUT}}/M_{\mathrm{Pl}})^2\sim10^{-6}$ already at formation, because they are so rare and heavy. The channel is off from the outset, so the number freezes at the geometric Kibble value, not a thermodynamic one. A related trap is to conflate this with the WIMP freeze-out calculation: the two share notation ( $Y=n/s$ , comparing $\Gamma$ to $H$ ) but have opposite logic — the WIMP abundance is set by decoupling from equilibrium, the monopole's by never reaching it.

Worked examples

1 · The overclosure number, worked end to end

Estimate $\Omega_M$ for $T_{\mathrm{GUT}}=10^{16}\,\mathrm{GeV}$ , with $g_{*s}\sim100$ , $m_M\sim10^{16}\,\mathrm{GeV}$ , $s_0=2.9\times10^{3}\,\mathrm{cm^{-3}}$ , $\rho_c=1.05\times10^{-5}h^2\,\mathrm{GeV\,cm^{-3}}$ .

Step 1. Form and cube the ratio: $T_{\mathrm{GUT}}/M_{\mathrm{Pl}}=10^{16}/(1.22\times10^{19})=8.2\times10^{-4}$ , so $(8.2\times10^{-4})^3=5.5\times10^{-10}$ — the geometric smallness of the GUT horizon.

Step 2. Divide by $g_{*s}\sim100$ : $Y_M=n_M/s\sim\mathbf{5.5\times10^{-12}}$ , about one monopole per $2\times10^{11}$ entropy quanta.

Step 3. Present energy density: $\rho_{M,0}=m_M\,Y_M\,s_0=(10^{16})(5.5\times10^{-12})(2.9\times10^{3})=\mathbf{1.6\times10^{8}\,GeV\,cm^{-3}}$ .

Step 4. Divide by critical: $\Omega_M h^2=1.6\times10^{8}/(1.05\times10^{-5})=\mathbf{1.5\times10^{13}}$ . So $\Omega_M\sim10^{13}h^{-2}$ literally — and a slightly lower $T_{\mathrm{GUT}}\approx2\times10^{15}\,\mathrm{GeV}$ recovers the textbook $\Omega_M\sim10^{11}$ . The Universe is overclosed by at least eleven orders of magnitude — the monopole problem.

2 · How heavy, how charged is one monopole?

Express the monopole mass in grams and proton masses, and its magnetic charge in Dirac units.

Step 1. In grams: $m_M\sim10^{16}\,\mathrm{GeV}\times1.78\times10^{-24}\,\mathrm{g/GeV}=\mathbf{1.8\times10^{-8}\,g}$ — a hundredth of a microgram, macroscopic for a point defect.

Step 2. In proton masses: $m_M/m_p=10^{16}/0.938\approx\mathbf{1.1\times10^{16}}$ — each outweighs ten quadrillion protons, which is why "one per tiny horizon" is so lethal.

Step 3. Magnetic charge from Dirac quantization: $g_M=e/2\alpha=137e/2=\mathbf{68.5\,e}$ , with magnetic coupling $g_M^2/4\pi\approx34$ deep in the strong regime — monopoles interact ferociously with magnetic fields, the basis of the Parker bound and induction searches. Dilution check: killing $\Omega_M\sim10^{13}$ needs a factor $10^{-13}$ , while $N=60$ e-folds deliver $e^{-180}\approx10^{-78}$ — over-solving by $\sim10^{65}$ , so the reheating bound, not the e-fold count, is the operative constraint.

Physics of the portrait

The monopole problem has a precise author. In 1979 a graduate student turned an elegant topological prediction into a cosmological crisis, and in doing so handed inflation one of its founding motivations.

Origins · 1 minJohn Preskill and how this was discovered

John Preskill

born 1953 · cosmological production of GUT magnetic monopoles

John Preskill was a Harvard graduate student when, in 1979, he published Cosmological Production of Superheavy Magnetic Monopoles (Phys. Rev. Lett. 43, 1365) — the paper that turned a topological curiosity into a cosmological paradox. The ingredients existed: Dirac had shown in 1931 that a single magnetic charge would quantize electric charge; 't Hooft and Polyakov had shown in 1974 that broken gauge theories contain monopoles as smooth solitons; and Kibble had shown in 1976 how defects freeze out at roughly one per horizon. Preskill combined them, tracked the abundance forward, and found GUT monopoles would outweigh everything else by many orders of magnitude — with annihilation unable to rescue it. Alan Guth later recalled that this was the puzzle he was chasing when he hit upon inflation in 1980–81: that a supercooled transition could both suppress monopoles and drive exponential expansion was the seed of the inflationary Universe. Preskill went on to foundational work on monopoles, the "invisible axion," and quantum information — but the monopole problem remains the cleanest case of cosmology constraining physics far above any accelerator.

History

1931

— Dirac shows a single magnetic charge would quantize all electric charge

1974

— 't Hooft and Polyakov find monopole solutions in broken gauge theories

1976

— Kibble computes topological-defect formation in cosmological transitions

1979

— Preskill computes GUT monopole overproduction: the monopole problem

1981

— Guth proposes inflation, diluting relics and solving horizon and flatness

Going further

Texts. Ryden, Introduction to Cosmology, 2nd ed., Ch. 10 (Inflation), monopole-problem section · Dodelson & Schmidt, Modern Cosmology, 2nd ed., Ch. 6 (Inflation) · Kolb & Turner, The Early Universe, Ch. 7 (Monopoles) and Ch. 8 (Inflation) · Carroll & Ostlie, An Introduction to Modern Astrophysics, 2nd ed., Ch. 30 · Mo, van den Bosch & White, Galaxy Formation and Evolution, §3.6 (relics and inflation) · Weinberg, Cosmology, Ch. 4. Primary sources: 't Hooft (1974) Nucl. Phys. B79, 276; Polyakov (1974) JETP Lett. 20, 194; Kibble (1976) J. Phys. A9, 1387; Preskill (1979) Phys. Rev. Lett. 43, 1365; Guth (1981) Phys. Rev. D23, 347. Key numerical anchors: the horizon $H^{-1}\sim M_{\mathrm{Pl}}/T^2$ evaluates to $2.4\times10^{-27}\,\mathrm{cm}$ at $10^{16}\,\mathrm{GeV}$ ; the Kibble density is $n_M\sim T_{\mathrm{GUT}}^6/M_{\mathrm{Pl}}^3$ ; the yield is $Y_M\sim g_{*s}^{-1}(T_{\mathrm{GUT}}/M_{\mathrm{Pl}})^3\sim5\times10^{-12}$ . The literal plug-in at $T_{\mathrm{GUT}}=10^{16}\,\mathrm{GeV}$ gives $\Omega_M\sim1.5\times10^{13}h^{-2}$ , while the canonical $\Omega_M\sim10^{11}$ corresponds to $T_{\mathrm{GUT}}\approx2\times10^{15}\,\mathrm{GeV}$ — the result is cubically sensitive to the GUT scale, but both values mean overclosed by at least eleven orders of magnitude. The Dirac charge $g_M=2\pi/e=68.5\,e$ , mass $m_M\sim M_{\mathrm{GUT}}/\alpha_{\mathrm{GUT}}$ , and inflationary suppression $e^{-180}\approx10^{-78}$ are standard results.

Builds on: Thermal History of the Universe · Equilibrium Thermodynamics of the Early Universe · Effective Number of Relativistic Degrees of Freedom · Entropy Conservation and Comoving Entropy · Decoupling and Freeze-Out. Leads to: Baryon Asymmetry and Baryogenesis · Big Bang Nucleosynthesis.

Ask the tutor: Ask me to show how Preskill's "one monopole per horizon" becomes an eleven-order overclosure — and why sixty e-folds of inflation erase it with room to spare.

Why is "at least one monopole per horizon" a lower bound rather than an estimate?

It follows from causality alone. Regions separated by more than the horizon distance have never exchanged a signal, so they cannot have agreed on which vacuum to choose — the orientation is a random variable, uncorrelated beyond the horizon. Topology then guarantees that a fraction of the boundaries between these random domains traps a defect that cannot be unwound, and the winding argument gives of order one per horizon volume. A detailed lattice estimate can raise this somewhat, but you cannot lower it much without violating causality, which is why the problem is so robust and cannot be dialed away by tuning the potential.

If monopoles are so heavy, why doesn't gravity or expansion simply thin them out?

Expansion does thin them out — but only as $a^{-3}$ , exactly like ordinary matter, and that is the problem. What matters is the relative rate: radiation dilutes faster, as $a^{-4}$ , because each photon also redshifts. So the monopole-to-radiation ratio grows as $a$ , and the monopoles' fractional share of the density increases with time rather than decreasing. Gravity cannot help either — they are stable and never clump and annihilate. Only expansion faster than $a^{-3}$ dilution can outrun them, which is exactly what inflation provides by making number density fall as $e^{-3N}$ .

Is the monopole problem the same as the horizon and flatness problems?

They are three distinct puzzles with one shared cure. The horizon problem asks why disconnected patches of the CMB share a temperature; the flatness problem asks why the spatial curvature is so near zero today given that it grows under ordinary expansion; the monopole problem asks why the predicted GUT relics are absent. All three are solved at once by quasi-exponential expansion: inflation stretches one causal patch across the sky (horizon), drives the density parameter toward one (flatness), and dilutes pre-existing relics to negligibility (monopoles). This triple payoff is much of why inflation was accepted so rapidly.

Does the relic problem apply to more than magnetic monopoles?

Yes — the monopole is only the sharpest case. Any state that is stable, superheavy, and produced with horizon-scale or thermal abundance faces the same overclosure, since its density dilutes only as $a^{-3}$ while radiation thins faster. Supergravity gravitinos are the classic example: produced thermally after inflation, they overclose the Universe or spoil nucleosynthesis unless $T_{\mathrm{reheat}}\lesssim10^{9}\,\mathrm{GeV}$ . String moduli face the "moduli problem" for the same reason, and cosmic strings and domain walls have their own versions. The generic escapes are inflation, low reheating, or defect-eating dynamics — the monopole problem is the archetype.

What is the Parker bound, and how is it independent of the cosmological argument?

The Parker bound is a purely astrophysical limit with no dependence on cosmological history. The Galaxy hosts a few-microgauss field maintained by a dynamo; a passing monopole is accelerated by it and extracts energy, so a large enough flux would drain the field faster than the dynamo rebuilds it. Requiring the field to survive caps the flux at $F_M\lesssim10^{-15}\,\mathrm{cm^{-2}\,s^{-1}\,sr^{-1}}$ . This constrains the monopoles in our Galaxy today, whatever produced them — and lying far below the naive GUT prediction, it is an independent line of evidence that GUT-density monopoles are absent.

Could lowering the GUT scale solve the problem without inflation?

It helps, but not enough alone, and it costs you the thing GUTs are for. The yield scales as $Y_M\propto(T_{\mathrm{GUT}}/M_{\mathrm{Pl}})^3$ , so a factor-of-ten drop in $T_{\mathrm{GUT}}$ buys only three orders of magnitude. Shedding eleven orders would demand pushing $T_{\mathrm{GUT}}$ near $\sim10^{12}\,\mathrm{GeV}$ — contradicting the measured running of the gauge couplings, which unify near $10^{16}\,\mathrm{GeV}$ , and the proton-decay limits requiring a high scale. Lowering $T_{\mathrm{GUT}}$ trades one problem for another; inflation dilutes the relics without touching the successful unification.

Monopole and Relic Problems

Where this lives today

The physical picture

The core equations

Limiting cases

Worked examples

1 · The overclosure number, worked end to end

2 · How heavy, how charged is one monopole?

Physics of the portrait

Going further

Before this

Where this leads