Antiparticles & negative-energy states

The Dirac equation was built to cure a disease — the negative probabilities of the Klein–Gordon theory — and it succeeded, delivering a positive density $\rho=\psi^\dagger\psi$ . But it did not cure the symptom everyone feared: its free plane-wave solutions still carry energies $E=\pm\sqrt{p^2+m^2}$ , and there is no way to keep only the positive root. Four independent solutions exist at each momentum, and squaring the Dirac operator returns the mass shell for every one, so two must carry $E>0$ and two $E<0$ . A decisive check makes the split unavoidable: the free Hamiltonian $\hat H=\boldsymbol\alpha\cdot\hat{\mathbf p}+\beta m$ is traceless, so its four eigenvalues sum to zero — exactly two positive, two negative. You cannot throw the negative ones away, because a quantum theory needs a complete basis to expand an arbitrary state. So the negative-energy solutions must be interpreted, and two great interpretations were offered. Dirac's 1930 picture fills every negative-energy level with electrons; the exclusion principle then blocks a positive-energy electron from cascading down, and a photon above $2m_e$ knocks a sea electron up, leaving a hole that behaves as a positive-energy, positively charged antiparticle. It predicted the positron, found by Anderson in 1932 — but it fails for bosons and gives the vacuum infinite charge. The modern reading, due to Stückelberg and Feynman, keeps none of that scaffolding: a negative-energy particle running backward in time is mathematically identical to a positive-energy antiparticle of opposite charge running forward. This article makes that equivalence precise and turns the discarded $u_3,u_4$ into the working spinors $v_1,v_2$ .

Key takeaways

The free Dirac Hamiltonian is a traceless $4\times4$ matrix, so its eigenvalues sum to zero — two positive and two negative energy states are mathematically unavoidable, not a removable defect.
The negative-energy solutions cannot be discarded: a complete basis is needed to span the state space, so they must be physically interpreted instead.
Dirac's hole theory (fill the sea, a hole = antiparticle) predicted the positron but fails for bosons and gives the vacuum infinite charge.
The modern Feynman–Stückelberg view: an $E<0$ particle running backward in time $\equiv$ an $E>0$ antiparticle of opposite charge running forward — the identity $e^{-iEt}=e^{-i(-E)(-t)}$ is its mathematical heart.
The discarded $u_3,u_4$ are repackaged as antiparticle spinors $v_{1,2}$ obeying $(\gamma^\mu p_\mu+m)v=0$ , normalised to $\bar v v=-2m$ , with charge conjugation $\hat C\psi=i\gamma^2\psi^*$ mapping $u\to v$ .
Because $v$ is still a negative-energy particle in disguise, the naive operators give $-E,-p$ ; recover physical values with $\hat H^{(v)}=-i\partial_t$ , $\hat p^{(v)}=+i\nabla$ , $\hat S^{(v)}=-\hat S$ .
Key idea: negative-energy solutions are not a defect to be removed but a feature to be reinterpreted — a backward-in-time negative-energy particle is precisely a forward-in-time positive-energy antiparticle of opposite charge.

Where this lives today

The negative-energy solutions are no longer a puzzle to be excused — they are the antiparticle creation operators of every quantum field theory, and the machinery runs continuously in the laboratory. At CERN's Antiproton Decelerator and ELENA, the ALPHA and AEgIS experiments trap antihydrogen — a positron bound to an antiproton — and compare its spectrum and free-fall against hydrogen, testing the CPT theorem that fixes particle and antiparticle to share a mass; in 2023 ALPHA-g confirmed that antimatter falls down under gravity. Positron annihilation into two 511 keV photons is the working principle of PET imaging in every hospital, while the diffuse 511 keV glow from the Galactic centre remains an unexplained positron source. And the reinterpretation is the default calculational rule, not a footnote: in every Feynman diagram an antiparticle line is drawn with its arrow pointing backward along the fermion flow, and the whole edifice of scattering amplitudes is bookkept that way. The deepest question it raises is cosmological — if the equations treat matter and antimatter so symmetrically, why is the universe made almost entirely of matter? The baryon asymmetry is still unsolved.

The physical picture

Start from the bare fact and let it force the interpretation. The Dirac equation has four plane-wave solutions per momentum, and squaring its operator hands every one of them $E^2=p^2+m^2$ — so half sit on the upper branch $E=+\sqrt{p^2+m^2}$ and half on the lower branch $E=-\sqrt{p^2+m^2}$ . The cleanest proof needs no spinor algebra: the free Hamiltonian $\hat H=\boldsymbol\alpha\cdot\hat{\mathbf p}+\beta m$ is a sum of traceless matrices, so its four eigenvalues add to zero. With eigenvalues $\pm E$ that means two of each sign — no arrangement of the algebra gives four positive energies. The first figure carries that spectrum and the electron's real gap $2m_ec^2=1.022$ MeV.

Figure

Four solutions per momentum split two-and-two because the traceless Dirac Hamiltonian has eigenvalues summing to zero. The upper branch E = +√(p²+m²) holds the particle spinors u₁,u₂; the lower branch E = −√(p²+m²) holds u₃,u₄, reinterpreted as antiparticle spinors v₁,v₂. For the electron the gap is 2mₑc² = 1.022 MeV

Keeping the solutions means reading them. The surviving reinterpretation rests on one identity: the time factor of a negative-energy solution, $e^{-iEt}$ with $E<0$ , is indistinguishable from $e^{-i(-E)(-t)}$ — a positive energy $-E>0$ evolving in reversed time. Flip the charge to keep Maxwell's equations invariant under time reversal, and a negative-energy electron running backward becomes a positive-energy positron running forward. The second figure draws that conversion in spacetime, with the annihilation it explains: an electron and positron produce two 511 keV photons, the line PET scanners detect.

Figure

The Feynman–Stückelberg reinterpretation in spacetime: a negative-energy electron running backward in time (dashed, arrow down) is identical to a positive-energy positron running forward (solid, arrow up), because e^(−iEt) = e^(−i(−E)(−t)). At the vertex e⁺e⁻ annihilate into two photons, each carrying mₑc² = 511 keV

The core equations

The four free-particle solutions are labelled by momentum and two discrete choices; written together they are plane waves whose energies come in both signs. The first two are the particle spinors, the second two the ones we must reinterpret.

\psi_i=u_i(\mathbf p)\,e^{i(\mathbf p\cdot\mathbf x-Et)},\quad i=1,2,3,4,\qquad E^2=\mathbf p^2+m^2 \qquad (1)

where: $u_i$ — four-component spinor amplitude; $\mathbf p$ — momentum; $E$ — energy, taking both roots $\pm\sqrt{\mathbf p^2+m^2}$ . Reading it: squaring the operator sends each component to Klein–Gordon, so all four obey the mass shell and two are forced onto the negative branch. The reason the split cannot be dodged is a one-line trace argument: the free Hamiltonian is a sum of traceless matrices, so its eigenvalues vanish in sum.

\hat H=\boldsymbol\alpha\cdot\hat{\mathbf p}+\beta m,\qquad \mathrm{Tr}\,\hat H=0\ \Rightarrow\ \sum_{k=1}^{4}E_k=0 \qquad (2)

where: $\boldsymbol\alpha,\beta$ — Dirac matrices, each traceless; $E_k$ — the four eigenvalues, $+E,+E,-E,-E$ . Reading it: two positive and two negative energies is not a choice — it is fixed once the Hamiltonian is Hermitian and traceless. Rather than carry $E<0$ around, we rewrite the lower-branch solutions as antiparticle spinors $v$ in the physical positive energy and momentum, so the antiparticle plane wave reads with a $+$ sign in the exponent.

\psi=v(\mathbf p)\,e^{-i(\mathbf p\cdot\mathbf x-Et)},\qquad E=+\sqrt{\mathbf p^2+m^2}>0 \qquad (3)

where: $v(\mathbf p)$ — antiparticle spinor; the flipped sign in $e^{+iEt}$ (relative to the particle's $e^{-iEt}$ ) is the whole content of "running backward in time". Reading it: $v$ is built from $u_3,u_4$ , but now every symbol denotes the measurable, positive quantity of a real antiparticle. The two families satisfy momentum-space Dirac equations differing only by the sign of the mass term — the algebraic fingerprint that $v$ describes the opposite branch.

(\gamma^\mu p_\mu-m)\,u=0,\qquad (\gamma^\mu p_\mu+m)\,v=0 \qquad (4)

where: $\gamma^\mu$ — Dirac gamma matrices, $\{\gamma^\mu,\gamma^\nu\}=2g^{\mu\nu}$ ; $p_\mu=(E,\mathbf p)$ with $E>0$ . Reading it: $u$ needs $-m$ , $v$ needs $+m$ — the relative sign distinguishes particle from antiparticle at the spinor level. The spinors carry a Lorentz-invariant normalisation, and the antiparticle norm comes out negative — an invariant marker of the lower branch no change of frame can erase.

\bar u u=+2m,\qquad \bar v v=-2m,\qquad u^\dagger u=v^\dagger v=2E \qquad (5)

where: $\bar\psi=\psi^\dagger\gamma^0$ — Dirac adjoint; $u^\dagger u$ — the density, positive for both. Reading it: the density $\psi^\dagger\psi=2E$ is positive for particle and antiparticle alike (the Klein–Gordon disease cured), while the scalar $\bar v v=-2m$ quietly records which branch you are on. Finally, the operation that turns a particle solution into its antiparticle is charge conjugation, an explicit spinor operation in a given representation.

\hat C\psi=i\gamma^2\psi^*,\qquad \hat C:\ u_1\ \longrightarrow\ v_1 \qquad (6)

where: $\psi^*$ — complex conjugate of the spinor; $\gamma^2$ — the second gamma matrix. Reading it: charge conjugation maps every particle spinor to the antiparticle spinor of opposite charge — the concrete realisation of " $u_3,u_4\to v_1,v_2$ ".

Derivation AWhy negative-energy solutions are unavoidable and cannot be thrown away

The goal is to establish, with two independent arguments, that the free Dirac equation must possess negative-energy solutions and that a consistent quantum theory cannot simply delete them. The first is a linear-dependence count in the spirit of Thomson; the second is the traceless-Hamiltonian eigenvalue argument. Together they show the negative branch is structural, and completeness of the basis forbids its removal — which is why interpretation, not deletion, is the only way forward.

Step 1. Count the solutions. The Dirac equation is first order with a four-component wavefunction, so at each momentum $\mathbf p$ there are exactly four independent plane-wave solutions. Suppose, for contradiction, all four could share the same positive-energy exponent $e^{i(\mathbf p\cdot\mathbf x-Et)}$ , $E>0$ . Why: if the negative branch could be avoided, every solution would share one time dependence — a hypothesis we can test directly.

Step 2. Derive the contradiction. A common exponent reduces the equation to a fixed $4\times4$ algebraic condition on the amplitudes, which has only two independent solutions, not four. The other two are forced to carry the opposite exponent, $E<0$ — the negative branch is mandatory. Why: two positive-energy spinors cannot span a four-dimensional solution space, so the lower branch must supply the missing two.

Step 3. Confirm with the trace. The free Hamiltonian is built from traceless matrices, so $\mathrm{Tr}\,\hat H=0$ . Since the trace equals the sum of eigenvalues, and these are $\pm E$ , we need $E+E-E-E=0$ — exactly two positive and two negative. Why: the trace is basis-independent, so it pins the two-and-two split without solving for a single spinor.

Step 4. Forbid deletion. A quantum state is expanded on a complete basis; the four solutions at each $\mathbf p$ are that basis. Discarding the negative-energy pair leaves an incomplete set that cannot represent a general wave packet, and interactions scatter positive-energy states into the missing sector, so the theory is inconsistent without them. Why: completeness is not optional — the negative-energy solutions must be kept and interpreted, the subject of this article.

Go deeper · Derivation B · 1 minThe Feynman–Stückelberg reinterpretation of the negative branch

Derivation BThe Feynman–Stückelberg reinterpretation of the negative branch

The aim is to convert a negative-energy particle solution into a physical positive-energy antiparticle, using only the plane-wave exponent and the behaviour of charge under time reversal. We isolate the offending time factor, apply the identity that trades a negative energy for reversed time, restore a positive-energy description, and identify the result as an antiparticle of opposite charge. No Dirac sea and no exclusion principle appear, which is why the argument works for bosons too.

Step 1. Isolate the time dependence. A negative-energy solution evolves as $e^{-iEt}$ with $E<0$ ; write $E=-|E|$ so the factor is $e^{+i|E|t}$ , growing in phase the "wrong" way compared with an ordinary $e^{-i|E|t}$ . Why: the sign of the exponent is the only place the negative energy shows up in the observable phase, so the reinterpretation must act there.

Step 2. Apply the identity. Note $e^{-iEt}=e^{-i(-E)(-t)}$ : a state of energy $E<0$ moving forward in $t$ has the same phase as one of energy $-E>0$ moving in $-t$ . The spatial factor behaves the same way, $e^{i\mathbf p\cdot\mathbf x}=e^{i(-\mathbf p)\cdot(-\mathbf x)}$ , so momentum flips alongside time. Why: this single equivalence is the mathematical heart of the reinterpretation — nothing is added, the same wave is read in reverse.

Step 3. Restore positive energy and flip the charge. A negative-energy electron of momentum $\mathbf p$ running backward is thus identical to a positive-energy particle of momentum $-\mathbf p$ running forward. To keep the electromagnetic coupling invariant under this reversal, the charge must reverse: the forward-running object has charge $-q$ . Why: reversing a worldline reverses the current it carries, so the forward-time reading describes a particle of opposite charge — an antiparticle.

Step 4. Name it. The forward-in-time, positive-energy, opposite-charge object is the antiparticle: for the electron ( $q=-e$ ) it is the positron ( $q=+e$ ), with the same mass because $|E|$ and $|\mathbf p|$ are unchanged. Charge conjugation $\hat C\psi=i\gamma^2\psi^*$ realises the map at the spinor level, sending $u\to v$ . Why: this identifies the reinterpreted solution with a real, detectable particle — closing the loop from "unwanted $E<0$ " to "physical antimatter".

Go deeper · Derivation C · 1 minBuilding the antiparticle spinors and their negative norm

Derivation CBuilding the antiparticle spinors and their negative norm

This derivation constructs the antiparticle spinors $v_{1,2}$ explicitly and shows they satisfy $(\gamma^\mu p_\mu+m)v=0$ with the Lorentz-invariant normalisation $\bar v v=-2m$ , while the density stays positive, $v^\dagger v=2E$ . The negative scalar norm is the invariant fingerprint of lower-branch solutions in physical variables — it is what forces the sign-flipped operators of the warning below.

Step 1. Write the standard forms. In the Dirac–Pauli representation the positive-energy spinor is $u=\sqrt{E+m}\,\big(\chi,\ \tfrac{\boldsymbol\sigma\cdot\mathbf p}{E+m}\chi\big)^{\!\top}$ ; the antiparticle spinor from the negative branch is $v=\sqrt{E+m}\,\big(\tfrac{\boldsymbol\sigma\cdot\mathbf p}{E+m}\chi,\ \chi\big)^{\!\top}$ , with $\chi$ a Pauli spinor and $E=+\sqrt{\mathbf p^2+m^2}$ . Why: the "large" and "small" blocks swap between $u$ and $v$ — the spinor-level signature of the opposite energy branch.

Step 2. Check the equation of motion. Acting with $(\gamma^\mu p_\mu+m)$ on $v$ and using $(\boldsymbol\sigma\cdot\mathbf p)^2=\mathbf p^2=E^2-m^2$ gives identically zero, so $v$ solves the $+m$ Dirac equation of (4). Why: the $+m$ sign, versus $-m$ for $u$ , is the algebra's way of tagging the antiparticle, and it must hold exactly for $v$ to be a genuine solution.

Step 3. Compute the density. With $N^2=E+m$ , $v^\dagger v=N^2\big(\tfrac{\mathbf p^2}{(E+m)^2}+1\big)=N^2\,\tfrac{\mathbf p^2+(E+m)^2}{(E+m)^2}$ . Using $\mathbf p^2=E^2-m^2$ the numerator is $2E(E+m)$ , so $v^\dagger v=2E$ — positive, like the particle. Why: a positive density on both branches is the cure the Dirac equation delivered; the antiparticle carries no negative probability.

Step 4. Compute the scalar. Insert $\gamma^0=\mathrm{diag}(\mathbb 1,-\mathbb 1)$ : $\bar v v=v^\dagger\gamma^0 v=N^2\,\tfrac{\mathbf p^2-(E+m)^2}{(E+m)^2}$ . The numerator is $-2m(E+m)$ , so $\bar v v=-2m$ — negative, in contrast to $\bar u u=+2m$ . Why: the Lorentz-invariant $\bar v v=-2m$ is the frame-independent marker of the antiparticle branch, and its sign is what makes the naive operators return $-E,-p$ on $v$ . *As a concrete check in natural units with $m=3$ , $p=4$ , $E=5$ , $\mathbf p\parallel\hat z$ : $v=(\sqrt2,0,2\sqrt2,0)$ gives $v^\dagger v=10=2E$ , $\bar v v=-6=-2m$ , and $(\gamma^\mu p_\mu+m)v=0$ component by component.*

Limiting cases

Case	Result	Meaning
Rest frame ( $\mathbf p=0$ )	$u$ pure upper block, $v$ pure lower block; $E=\pm m$	particle and antiparticle at rest sit on the two branches, split by the gap $2m$
Non-relativistic ( $p\ll m$ )	small component $\tfrac{\boldsymbol\sigma\cdot\mathbf p}{E+m}\to 0$	$u$ becomes the upper two components, $v$ the lower two — ordinary spin- $\tfrac12$ QM recovered
Ultra-relativistic ( $p\gg m$ )	$\tfrac{\boldsymbol\sigma\cdot\mathbf p}{E+m}\to$ helicity	spinors approach helicity/chirality eigenstates; $\bar v v=-2m$ becomes negligible next to $2E$
Massless ( $m\to0$ )	$\bar v v=-2m\to0$	the scalar marker vanishes; particle vs antiparticle is carried only by helicity (Weyl, and the Majorana question)
Bosons (Klein–Gordon)	no exclusion principle	Dirac's sea fails, but Feynman–Stückelberg still works — antiparticles are not a fermion-only story

Common trapThinking the antiparticle spinor v(E,p) literally carries energy E and momentum p

It does not, and this trips up almost everyone. The spinor $v(E,p)$ is written with physical, positive $E$ and $\mathbf p$ for bookkeeping, but it is still a negative-energy particle solution in disguise — its plane wave carries $e^{+iEt}$ , not $e^{-iEt}$ . So the ordinary operators return the wrong sign: $\hat H\psi=-i\partial_t\big(v\,e^{+iEt}\big)$ gives $-E$ , and $\hat p\,\psi$ gives $-\mathbf p$ . To read off the physical energy, momentum and spin you must use the sign-flipped operators $\hat H^{(v)}=-i\partial_t\to+E$ , $\hat p^{(v)}=+i\nabla\to+\mathbf p$ , and $\hat S^{(v)}=-\hat S$ — a positron's measured quantum numbers are minus the eigenvalues the naive electron operators return on $v$ . Forgetting this is the most common error in antiparticle bookkeeping: it flips the sign of helicity in $\beta$ -decay problems and the charge in current calculations.

Worked examples

1 · Pair production and annihilation energetics (fully worked)

The positron has the same mass as the electron, $m_ec^2=0.511$ MeV. Find the threshold photon energy to create an $e^+e^-$ pair, and the energy of each photon when a slow $e^+e^-$ pair annihilates at rest into two photons. Why is the second number the one PET scanners look for?

Step 1. Pair creation must supply at least the rest energy of both particles: $E_{\min}=2m_ec^2=2\times0.511=\mathbf{1.022\ MeV}$ . A single real photon cannot do this in vacuum (energy and momentum cannot both balance), so pair production needs a nearby nucleus to absorb recoil — but the threshold is the $2m_ec^2$ gap of the first figure.

Step 2. Annihilation at rest is the reverse: total energy $E_{\rm tot}=2m_ec^2=1.022$ MeV and total momentum zero. Momentum conservation forbids a single final photon, so the minimum is two photons emitted back-to-back.

Step 3. By symmetry the two photons share the energy equally: each carries $E_\gamma=\tfrac12(2m_ec^2)=m_ec^2=\mathbf{0.511\ MeV=511\ keV}$ , flying in opposite directions.

Step 4. That 511 keV back-to-back signature is unmistakable. A PET tracer emits positrons; each annihilates with a tissue electron into two 511 keV photons $180^\circ$ apart, and a ring of detectors registers the coincident pair to reconstruct the emission line. Pair-creation threshold 1.022 MeV; each annihilation photon 511 keV — the line PET detects.

2 · Constructing an antiparticle spinor and checking its norm (fully worked)

Work in natural units with $m=3$ and $\mathbf p=4\,\hat z$ . Build the antiparticle spinor $v$ and verify $v^\dagger v=2E$ (positive) while $\bar v v=-2m$ (negative), the invariant marker of the lower branch.

Step 1. Energy and normalisation. $E=\sqrt{p^2+m^2}=\sqrt{16+9}=\mathbf{5}$ , so $E+m=8$ and $N=\sqrt{E+m}=2\sqrt2$ . With $\mathbf p\parallel\hat z$ , $\boldsymbol\sigma\cdot\mathbf p=p\,\sigma_z$ , and $\chi=(1,0)^\top$ : $v=N\big(\tfrac{p}{E+m},0,1,0\big)^\top=2\sqrt2\,(0.5,0,1,0)^\top=\mathbf{(\sqrt2,\,0,\,2\sqrt2,\,0)^\top}$ .

Step 2. Density (positive). $v^\dagger v=(\sqrt2)^2+(2\sqrt2)^2=2+8=\mathbf{10}$ . Compare $2E=2\times5=10$ — a match, so the antiparticle carries a positive density, no Klein–Gordon disease.

Step 3. Scalar (negative). With $\gamma^0=\mathrm{diag}(1,1,-1,-1)$ , $\bar v v=v^\dagger\gamma^0 v=(\sqrt2)^2(+1)+(2\sqrt2)^2(-1)=2-8=\mathbf{-6}$ . Compare $-2m=-6$ — a match, and the sign is negative, opposite to $\bar u u=+2m$ .

Step 4. Read the lesson. The same spinor gives a positive density but a negative scalar; the negative $\bar v v$ is Lorentz-invariant, so no boost turns this into a particle — it is the frame-independent stamp of the antiparticle branch, and the reason the operators on $v$ must be sign-flipped. $v=(\sqrt2,0,2\sqrt2,0)$ : $v^\dagger v=10=2E>0$ , $\bar v v=-6=-2m<0$ .

Physics of the portrait

The reinterpretation that made antiparticles ordinary — no infinite sea, no exclusion principle, working equally for electrons and pions — came not from Dirac but from a Swiss physicist who published it years ahead of the field and was ignored.

Origins · 1 minErnst Stückelberg and how this was discovered

Ernst Stückelberg

1905–1984 · the spacetime interpretation of antiparticles

Ernst Carl Gerlach Stückelberg was a Swiss theoretical physicist whose ideas repeatedly ran a decade ahead of their eventual fame. Working in Geneva and Lausanne, often in near-isolation and hampered by illness, he arrived at results others later rediscovered and were credited for. Around 1941–1942 he proposed reading the negative-energy solutions of relativistic wave equations as particles propagating backward in time — mathematically identical to positive-energy antiparticles propagating forward — precisely the picture that dispenses with Dirac's crowded sea and, crucially, works for bosons that have no exclusion principle to hold a sea in place. He also invented what is now the renormalization group with André Petermann and developed a covariant perturbation theory before Feynman's diagrams. Richard Feynman independently built the same backward-in-time reading of the positron into his 1949 formulation of quantum electrodynamics, and it is Feynman's version the world learned — so the reinterpretation is now called Feynman–Stückelberg, a rare correction of the record in Stückelberg's favour. He was nominated for the Nobel Prize but never received it; the modern, sea-free understanding of antimatter is substantially his.

History

1928

— Dirac's equation yields free solutions of negative energy E = −√(p²+m²)

1931

— Dirac's hole theory predicts an anti-electron of equal mass and opposite charge

1932

— Anderson photographs a positron track in a cloud chamber

1941

— Stückelberg reads negative-energy states as particles running backward in time

1949

— Feynman builds the same reinterpretation into his positron theory of QED

Going further

Texts. Thomson, Modern Particle Physics, §4.5–4.7 (negative-energy solutions, antiparticles, Feynman–Stückelberg) · Griffiths, Introduction to Elementary Particles, 2nd ed., §7.1–7.3 and App. C (Dirac spinors and charge conjugation) · Halzen & Martin, Quarks and Leptons, Ch. 5 (antiparticles and the hole theory) · Peskin & Schroeder, An Introduction to Quantum Field Theory, §3.3–3.5 (the Dirac field, $v$ -spinors, charge conjugation) · Schwartz, Quantum Field Theory and the Standard Model, Ch. 11 (spinor solutions and antiparticles). The key results of this article are: (1) $\mathrm{Tr}\,\hat H=0$ forces eigenvalues $+E,+E,-E,-E$ ; (2) the identity $e^{-iEt}=e^{-i(-E)(-t)}$ and the charge flip under time reversal; (3) $(\gamma^\mu p_\mu+m)v=0$ with $\bar v v=-2m$ , $v^\dagger v=2E$ , checked at $m=3,p=4,E=5$ giving $v=(\sqrt2,0,2\sqrt2,0)$ , $v^\dagger v=10$ , $\bar v v=-6$ ; (4) energetics $2m_ec^2=1.022$ MeV and annihilation photons $511$ keV.

Prerequisites. The Dirac Equation · Solutions of the Dirac Equation · The Klein–Gordon Equation. Leads to. Spin and Helicity States · Intrinsic Parity of Fermions · Covariant Dirac Formalism.

Ask the tutor: Ask me to run the trace argument that forces two negative-energy states, or to build the v-spinor and watch v̄v = −2m fall out.

Do the negative-energy solutions mean electrons can lose energy forever?

In the naive one-particle reading they would — a positive-energy electron could radiate photons and cascade down through the negative-energy states without limit, which is unphysical. That looming catastrophe is exactly what forced an interpretation. Dirac's answer was to fill every level so the exclusion principle blocks the cascade. The modern answer is cleaner: there is no cascade because the negative-energy solutions are not extra states an electron can fall into — they are antiparticle states, described by the $v$ -spinors and populated by positrons, not by low-lying electrons. Reinterpreting the branch, rather than filling it, removes the instability without burdening the vacuum.

What is the difference between Dirac's hole theory and Feynman–Stückelberg?

They explain the same experiments with very different pictures. Hole theory is a many-body construction: the vacuum is an infinite filled sea of negative-energy electrons, and an antiparticle is the absence of one — a hole — behaving as a positive-energy, positive-charge particle. It vividly pictures pair production (lift a sea electron, leave a hole) and predicted the positron, but it saddles the vacuum with infinite negative charge and, fatally, only works for fermions. Feynman–Stückelberg is a single-particle reinterpretation: a negative-energy particle running backward in time is a positive-energy antiparticle running forward, full stop. It needs no sea, works for every spin, and is the language of modern quantum field theory — the tool still in daily use.

Why can't we just discard the negative-energy solutions?

Because a quantum theory expands states on a complete basis, and at each momentum the four Dirac solutions are that basis. Keep only the two positive-energy ones and you can no longer represent a general localised wave packet. Worse, the moment you switch on any interaction (a potential step, an external field, an incoming photon) the equation couples positive-energy states to the negative-energy sector — the Klein paradox is exactly this coupling at a sharp barrier. So the negative-energy solutions are not passive baggage; leaving them out makes the mathematics incomplete and the physics inconsistent. The only coherent option is to keep them and give them a physical meaning — antiparticles.

Is the antiparticle really "the same particle running backward in time"?

As a calculational statement, yes and it is exact; as a claim about the flow of time, no. Inside a Feynman diagram the amplitude computed with a backward-pointing positive-energy positron is identical, term for term, to the one for a forward-pointing negative-energy electron — that equivalence is the content of $e^{-iEt}=e^{-i(-E)(-t)}$ . It is a superb bookkeeping device, physically correct for every observable. But no positron travels into your past; macroscopic time still runs one way, set by thermodynamics. "Backward in time" is the right way to organise the mathematics of antiparticles, not a statement that antimatter is a time machine.

Why does charge conjugation flip the charge but not the mass?

Charge conjugation $\hat C\psi=i\gamma^2\psi^*$ swaps a particle solution for its antiparticle by complex-conjugating the spinor and rotating it in Dirac space. The complex conjugation reverses the sign of the coupling to the electromagnetic field — physically, the charge, turning an electron ( $-e$ ) into a positron ( $+e$ ). The mass, however, enters as a real scalar $m$ untouched by conjugation or by the $\gamma^2$ rotation; the operation maps the $-m$ equation for $u$ into the $+m$ equation for $v$ at the same $|m|$ . This is why every antiparticle has exactly the mass of its partner but the opposite charge — and, more generally, the opposite sign of every additive quantum number, which CPT elevates to a theorem.

If matter and antimatter are so symmetric, why is the universe made of matter?

That is one of the great unsolved problems, and this article sharpens rather than solves it. The Dirac equation and its Feynman–Stückelberg reading treat particle and antiparticle almost perfectly symmetrically — same mass, opposite charge — and a naive early universe should have made equal amounts, which would then annihilate into radiation, leaving no galaxies. The survival of a small residue of matter (about one part in a billion) requires the three Sakharov conditions: baryon-number violation, C and CP violation, and a departure from thermal equilibrium. The Standard Model contains CP violation (seen in kaon and B-meson decays) but far too little to explain the observed asymmetry, so the shortfall points to physics beyond it. The tidy reinterpretation of a negative-energy solution becomes, at cosmological scale, a genuine mystery.