Dirac Equation

Definition

In 1928 Paul Dirac asked a deceptively simple question: what is the correct quantum wave equation for an electron moving fast enough that relativity matters? The answer reshaped physics. In its compact covariant form, the Dirac equation is:

(iγ^μ ∂_μ − m) ψ = 0

Here the γ^μ are four 4×4 matrices (the gamma matrices), ∂_μ is the four-gradient of spacetime, m is the particle's mass, and ψ is a four-component spinor — not a single number, but a column of four complex functions. (We work in natural units, ℏ = c = 1; restoring them puts factors of ℏc back in.) The equation is first-order in both space and time, which is the whole trick — and the source of everything that follows.

How it works — the square-root problem

The relativistic energy of a free particle obeys the famous relation

E² = (pc)² + (mc²)²

A naive quantum version (the Klein-Gordon equation) is second-order in time and produces negative probabilities for spin-1/2 matter. Dirac wanted something first-order in time, like Schrödinger's equation, but still consistent with that relation. Effectively he wanted to take a square root of the operator:

H = √( (cp)² + (mc²)² )   →   H = c·α·p + β·mc²

For this to square back to (cp)² + (mc²)², the coefficients α and β can't be ordinary numbers — they must anti-commute:

αᵢαⱼ + αⱼαᵢ = 2δᵢⱼ ,   αᵢβ + βαᵢ = 0 ,   β² = 1

No scalars satisfy this, but 4×4 matrices do. The smallest matrices that work are 4×4, and once your coefficients are 4×4 matrices, the object they act on — the wave function ψ — must have four components. That is not a modeling choice; it is forced by the algebra. In covariant notation the conditions become the defining Clifford algebra of the gamma matrices:

{γ^μ, γ^ν} = γ^μγ^ν + γ^νγ^μ = 2 η^μν · 𝟙

where η^μν = diag(+1, −1, −1, −1) is the Minkowski metric. The four components are not abstract bookkeeping: two of them describe the spin-up and spin-down states of the electron, and the other two describe the spin-up and spin-down states of its antiparticle.

Spin and g = 2 fall out for free

Before Dirac, spin was an empirical add-on. Goudsmit and Uhlenbeck (1925) had to postulate that the electron carries an intrinsic angular momentum of ℏ/2, and to fit the Zeeman effect they had to assign it a magnetic moment twice as large as a classical spinning charge would have — a gyromagnetic ratio g = 2 rather than g = 1. Nobody could say why.

When you couple the Dirac equation to an electromagnetic field (the minimal substitution ∂_μ → ∂_μ + ieA_μ) and take the non-relativistic limit, the Pauli equation appears automatically — including a magnetic-moment term with exactly the right coefficient:

μ = g · (e/2m) · S ,   with   g = 2   (exactly, from Dirac)

The factor of 2 that had to be guessed now emerges from the structure of the equation with zero free parameters. This is one of the most beautiful "free lunches" in physics: ask for relativity plus quantum mechanics, and spin-1/2 with g = 2 comes back unbidden.

The prediction of antimatter

Because the energy relation has two roots, ±√((pc)² + (mc²)²), the four-component spinor inevitably contains negative-energy solutions. That is a disaster on its face: a positive-energy electron could radiate photons and tumble forever down into ever more negative energies, and matter would be unstable.

Dirac's escape (1930) was the Dirac sea: imagine the vacuum is not empty but is an infinite ocean of filled negative-energy states. By the Pauli exclusion principle, no ordinary electron can fall in — every seat is taken. Now remove one negative-energy electron. The resulting hole in the sea behaves exactly like a particle with the electron's mass but opposite charge and positive energy: the positron.

Dirac first wondered whether the hole might be the proton, but the masses didn't match; by 1931 he committed to a brand-new "anti-electron." In 1932 Carl Anderson, photographing cosmic-ray tracks in a cloud chamber, found a particle that curved the wrong way in a magnetic field — the positron, just four years after the prediction. The same logic applies to every charged fermion: the equation that describes the electron also predicts the existence of its antiparticle. Pair production (γ → e⁻ + e⁺) and annihilation (e⁻ + e⁺ → 2γ) are the sea filling and emptying.

Worked example — the spectral line that proves it

Here is a concrete number you can check against measurement. Solve the Dirac equation for an electron bound in the Coulomb field of a proton (the hydrogen atom). The exact energy levels are:

E(n, j) = m c² · [ 1 + ( α / (n − (j+½) + √((j+½)² − α²)) )² ]^(−½)

where α ≈ 1/137.036 is the fine-structure constant and j is the total angular momentum. Expanding in powers of α:

E(n, j) ≈ −(13.6 eV)/n²  ·  [ 1 + (α²/n²)( n/(j+½) − 3/4 ) ]

The first term is the familiar Bohr/Schrödinger spectrum. The second term — of relative size α² ≈ (1/137)² ≈ 5.3 × 10⁻⁵ — is the fine structure. It tells you that levels with the same n but different j are split.

Take the n = 2 levels of hydrogen. The Dirac equation predicts that 2P_3/2 (j = 3/2) sits above 2S_1/2 and 2P_1/2 (both j = 1/2) by:

ΔE ≈ (13.6 eV) · α² / 16 ≈ 4.5 × 10⁻⁵ eV ≈ 45 μeV
   →  splitting of the Hα line at ≈ 10.9 GHz (≈ 0.36 cm⁻¹)

That number was measured, and it matched — Schrödinger's equation could not produce it at all, because it has no spin and no relativistic kinetic correction. Note one subtlety the Dirac equation gets almost right: it predicts 2S_1/2 and 2P_1/2 to be exactly degenerate. Experiment (Lamb, 1947) found a tiny splitting of about 1057 MHz — the Lamb shift — which is a quantum-field-theory effect beyond the single-particle Dirac equation. The Dirac equation is right to order α²; QED is needed for the next layer.

Variants and regimes

The Dirac equation is the parent of several specialized forms, each appropriate to a different physical regime:

Form	What it describes	Components	Key feature
Dirac equation	Massive spin-1/2 particle (electron)	4-spinor	Mass term mixes particle & antiparticle, both chiralities
Weyl equation	Massless spin-1/2 (chiral fermion)	2-spinor	Definite handedness; once thought to fit neutrinos
Majorana equation	Fermion that is its own antiparticle	4-spinor (real)	No distinct antiparticle; candidate for neutrinos
Pauli equation	Non-relativistic spin-1/2 in a field	2-spinor	Low-velocity limit of Dirac; carries the g = 2 term
Schrödinger equation	Non-relativistic, spinless	1 scalar	Drops antimatter and spin entirely
Dirac in curved spacetime	Fermion near a black hole / cosmology	4-spinor + vierbein	γ^μ become position-dependent; spin connection added
Lattice / condensed-matter Dirac	Electrons in graphene, topological insulators	2-spinor (effective)	Emergent "relativistic" quasiparticles, c → Fermi velocity

The condensed-matter row is more than an analogy: near the corners of graphene's Brillouin zone, electrons obey a 2D massless Dirac equation with an effective light speed of roughly 10⁶ m/s — about 1/300 of c. The "Dirac sea" picture and pair-production-like physics become directly observable on a lab bench.

Common pitfalls and misconceptions

• "ψ is a 4-vector." It is not. A four-vector transforms with the Lorentz matrices Λ^μ_ν; a Dirac spinor transforms under a different (spinor) representation, and crucially it returns to itself only after a 720° rotation, not 360°. The four components are not the four spacetime directions.
• "The Dirac sea is literally real and has infinite negative charge." The sea is brilliant intuition but a flawed literal model. Modern QFT discards the infinite reservoir and instead treats antiparticles via field operators; the predictions (pair creation, annihilation, the positron) survive without the paradoxes.
• "g = 2 is exact." Dirac gives g = 2 at tree level. QED loop corrections give the anomalous moment, so g = 2.00231930436... The 0.1% correction is itself one of the most precisely verified numbers in science — but g is not exactly 2.
• "Antiparticles travel backward in time." This is the Stückelberg-Feynman bookkeeping convenience for drawing diagrams, not a literal claim. A positron is an ordinary particle moving forward in time with positive energy.
• "The Dirac equation explains all of hydrogen's spectrum." It nails the fine structure to order α² but misses the Lamb shift and hyperfine structure, which require QED and nuclear spin respectively.
• "Negative-energy solutions are unphysical and should be thrown away." Discarding them breaks completeness and Lorentz invariance. They are physical — reinterpreted as antiparticles.

Where the Dirac equation shows up

• Quantum electrodynamics (QED). The Dirac field is the matter field of QED — every electron line in a Feynman diagram is a Dirac propagator. QED is the most precisely tested theory in physics.
• The Standard Model. Every fundamental matter particle — electrons, muons, taus, neutrinos, all six quarks — is a Dirac (or Majorana) spinor. The whole fermion sector is built on this equation.
• Atomic and chemical physics. Relativistic Dirac-based calculations explain why gold is yellow and mercury is liquid at room temperature — heavy-atom electrons move fast enough that relativity shifts their orbitals.
• Medical and industrial imaging. Positron Emission Tomography (PET) scans rely directly on positron–electron annihilation producing back-to-back 511 keV gamma rays — the antimatter the equation predicted, used millions of times a year.
• Materials science. Graphene, topological insulators, and Weyl semimetals host emergent Dirac fermions; their exotic transport (the quantum Hall effect, Klein tunneling) is engineered using the Dirac equation.
• Astrophysics. Electron degeneracy pressure in white dwarfs and the equation of state of neutron-star matter use relativistic spin-1/2 statistics rooted in the Dirac framework.

Derivation and structure analysis

Why exactly four components, and not two or eight? Count the physics. A spin-1/2 particle has 2 spin states. Relativity pairs every particle with an antiparticle, doubling that to 4. So the minimal Lorentz-covariant object describing a massive spin-1/2 field has exactly 2 × 2 = 4 complex components. The algebra confirms it: the smallest irreducible representation of the Clifford algebra {γ^μ, γ^ν} = 2η^μν in 3+1 dimensions is 4-dimensional. The number 4 is dictated jointly by the dimension of spacetime and the demand for both spin and antimatter.

The chiral (Weyl) decomposition makes the internal structure vivid. In the chiral basis the four-spinor splits into two two-spinors:

ψ = ( ψ_L )      mass term couples L ↔ R:
    ( ψ_R )      iσ̄·∂ ψ_L = m ψ_R ,   iσ·∂ ψ_R = m ψ_L

If m = 0, the two halves decouple into independent massless Weyl fermions of fixed handedness — which is exactly what the Standard Model uses before electroweak symmetry breaking gives particles mass. The mass term is precisely what stitches left- and right-handed fields into a single Dirac particle, and what forbids a Dirac fermion from having a single definite chirality. This is why "where does mass come from?" and "why is the weak force chiral?" are questions you can ask directly of the spinor structure.

A final structural point about cost and precision. The Dirac equation is a single 4-component first-order PDE, computationally heavier than Schrödinger's scalar equation, but it buys you spin, fine structure, and antimatter at once rather than as three separate patches. Its predictions are not merely qualitative: g = 2 corrected by QED to g/2 = 1.00115965218..., agreement with experiment to ~12 significant figures; the positron mass predicted equal to the electron's 0.511 MeV/c² and confirmed; fine-structure splittings of order α² verified across the periodic table. Few equations earn their complexity so completely.