Mass-Energy Equivalence

Of all the equations in physics, $E=mc^2$ is the one everyone can recite and almost no one reads correctly. The popular reading — "mass turns into energy" — is a caricature of something deeper and simpler: energy and inertia are not two properties a body happens to have, but two names for one quantity seen from one place. A hot poker is heavier than a cold one. A wound spring outweighs a slack one. A helium nucleus weighs less than the four particles you built it from, by exactly the energy you had to carry away to bind them. In every case the bookkeeping is the same — add energy, add mass; remove energy, remove mass — with the conversion factor $c^2$ fixed once and for all by the geometry of spacetime. This article takes the invariant spacetime interval from aq-gr-spacetime-interval and the four-vector transformation law from aq-gr-four-vectors, applies them to energy and momentum, and reads off what falls out. The natural home of the relation is not the rest frame but the whole of Minkowski space. Energy and momentum assemble into a single four-vector $p^\mu=(E/c,\,\vec p)$ , and just as the interval fixes the invariant length of a displacement, the Minkowski metric fixes the invariant length of $p^\mu$ — and that length is the rest mass. Different observers disagree about $E$ and about $\vec p$ ; they never disagree about $m$ . Setting $\vec p=0$ collapses the invariant to the slogan $E_0=mc^2$ , but the full statement, the dispersion relation $E^2=(pc)^2+(mc^2)^2$ , is what governs photons, fast electrons, and colliding protons alike. And because every form of energy carries mass, and mass gravitates, the equivalence is the hinge on which general relativity turns: it forces the Newtonian mass density to be replaced by the full stress-energy tensor $T^{\mu\nu}$ , the true source of curvature in aq-gr-einstein-field-equations.

Key takeaways

Rest mass and rest energy are the same thing in different units: $E_0=mc^2$ , with $c^2$ a fixed conversion factor, not a rate of some process.
Energy and momentum are the components of one four-vector $p^\mu=(E/c,\,\vec p)$ ; its Minkowski length is the invariant rest mass, so $E^2=(pc)^2+(mc^2)^2$ .
Rest mass $m$ is a Lorentz invariant — all observers agree on it — while $E$ and $\vec p$ separately transform like the time and space parts of a four-vector.
Every internal energy counts toward rest mass: thermal energy, binding energy, field energy. A bound system weighs less than its parts by the binding energy over $c^2$ .
Both members of the identity conserve: in a closed system total energy and total invariant mass are conserved; "mass converts to energy" really means rest mass moves between subsystems as energy is exchanged.
Key idea: the rest mass of a system is the Lorentz-invariant magnitude of its energy-momentum four-vector, $m^2c^2=E^2/c^2-\vec p^{\,2}$ ; in the rest frame this is $E_0=mc^2$ , all energy weighs, and all energy therefore gravitates.

The core insight

The rest mass of a system is the Lorentz-invariant magnitude of its energy-momentum four-vector, $m^2c^2=E^2/c^2-\vec p^{\,2}$ ; in the rest frame it collapses to $E_0=mc^2$ . Mass is stored energy — interconvertible, additive as energy, and, because it weighs, universally gravitating.

Where this lives today

Mass-energy equivalence stopped being a curiosity the moment it became measurable, and today it is infrastructure. Nuclear power and nuclear weapons are, at their core, the controlled release of rest mass as binding energy; the mass defect of the fission or fusion products, multiplied by $c^2$ , is the entire energy budget. Every particle collider is an $E=mc^2$ machine run in reverse: the LHC pours kinetic energy into colliding protons precisely so that the dispersion relation can convert it into the rest masses of new particles — the Higgs boson at $125\ \mathrm{GeV}/c^2$ was, quite literally, energy made into mass. But the deepest modern descendant is gravitational. Because all energy weighs, the source of spacetime curvature is the stress-energy tensor $T^{\mu\nu}$ , and its components include pressure — so a hot, high-pressure star gravitates slightly more strongly than its rest mass alone would suggest, an effect that becomes decisive inside neutron stars and is written into every solution of the Einstein field equations, from black holes to the expanding universe.

The physical picture

Picture the energy-momentum four-vector $p^\mu$ as an arrow drawn in spacetime. Its shadow on the time axis is the energy $E/c$ ; its shadow on the space axes is the momentum $\vec p$ . Tilt your head — boost to another inertial frame — and both shadows change length, exactly as a stick's horizontal and vertical projections change when you rotate it. But the arrow itself has an intrinsic length that no tilt can alter, and that length is the rest mass. Mass is simply what is left of the arrow when you rotate away all the motion. The first figure makes this concrete: every observer finds the body somewhere on one hyperbola — the mass shell — whose vertex is the invariant $mc^2$ .

Figure

The mass shell E² = (pc)² + (mc²)² for an electron. Every inertial observer sees the body somewhere on this one hyperbola, and its vertex — the invariant rest energy mc² = 0.511 MeV — is the same for all of them; a Lorentz boost only slides the body up along the curve

The second lesson is that internal energy has weight. Assemble two protons and two neutrons into a helium-4 nucleus and you must carry off $28.3\ \mathrm{MeV}$ of binding energy — as gamma rays, as the kinetic energy of the products. The bound nucleus is left lighter than the four free particles by exactly that energy over $c^2$ , a mass defect of about $0.0304\ \mathrm{u}$ . This is not an exotic quantum effect; it is the same accounting as the hot poker, done where the numbers are big enough to weigh. Any system that has surrendered energy has surrendered mass, and the second figure shows the ledger for the most important reaction in the universe.

Figure

Binding lowers rest mass. Four free nucleons carry a combined rest energy Σmᵢc² = 3755.67 MeV, but the bound ⁴He nucleus carries only 3727.38 MeV — the 28.3 MeV difference is radiated away, so the nucleus weighs Δm = ΔE/c² = 0.0304 u less than the parts it was built from

The final step is the one that reaches into gravity. Inertial mass equals gravitational mass — the equivalence principle — so if every form of energy contributes to inertia, then every form of energy must also gravitate: not only rest mass, but kinetic energy, radiation, and even pressure. The Newtonian idea that "mass density is the source of gravity" cannot survive this. Its relativistic successor must package energy density, momentum flux, and stress into one object, the stress-energy tensor, and that is precisely the road from $E=mc^2$ to Einstein's field equations.

The core equations

Everything begins with the observation that energy and momentum are not separate conserved things but the four components of one conserved four-vector. In the standard convention its time component is the energy over $c$ and its space components are the ordinary three-momentum.

p^\mu=\left(\frac{E}{c},\,\vec p\right)=m\,u^\mu \qquad (1)

where: $p^\mu$ — energy-momentum four-vector, index $\mu=0,1,2,3$ ; $E/c$ — time component; $\vec p$ — spatial three-momentum; $u^\mu=\mathrm{d}x^\mu/\mathrm{d}\tau$ — four-velocity; $m$ — invariant rest mass. Reading: energy and momentum are one geometric object, the rest mass times the four-velocity. Because $p^\mu$ is a four-vector, the Minkowski metric assigns it an invariant squared length — a number every inertial observer computes to be the same, exactly as they all agree on the spacetime interval. With signature $(-,+,+,+)$ that invariant is negative and equal to minus the rest energy squared over $c^2$ .

p^\mu p_\mu=-\left(\frac{E}{c}\right)^2+\vec p^{\,2}=-(mc)^2 \qquad (2)

where: $p^\mu p_\mu$ — Minkowski self-contraction of the four-momentum; the right-hand side $-(mc)^2$ is frame-independent. Reading: whatever $E$ and $\vec p$ an observer measures, this particular combination always returns the rest mass. Rearranging (2) with a factor of $c^2$ gives the single most useful equation in relativistic dynamics, the dispersion relation connecting energy, momentum, and mass. It holds for anything — massive or massless, slow or ultra-relativistic.

E^2=(pc)^2+(mc^2)^2 \qquad (3)

where: $E$ — total energy; $p=|\vec p|$ — momentum magnitude; $m$ — rest mass. Reading: energy, momentum, and mass form a right triangle. Set $p=0$ and you get $E_0=mc^2$ ; set $m=0$ and you get $E=pc$ for light. Setting the momentum to zero picks out the rest frame and isolates the famous special case: the energy a body has purely by existing, with no motion at all.

E_0=mc^2 \qquad (4)

where: $E_0$ — rest energy; $m$ — invariant rest mass; $c$ — speed of light, $299\,792\,458\ \mathrm{m\,s^{-1}}$ (exact). Reading: $c^2$ is a units conversion — in natural units with $c=1$ it reads $E_0=m$ , energy and mass literally identical. For a body actually in motion the total energy and the momentum both pick up the Lorentz factor $\gamma$ , so that the rest energy is only their $v\to 0$ floor. These two equations, together with (3), are internally consistent — squaring and subtracting reproduces the dispersion relation.

E=\gamma mc^2,\qquad \vec p=\gamma m\vec v,\qquad \gamma=\frac{1}{\sqrt{1-v^2/c^2}} \qquad (5)

where: $\gamma$ — Lorentz factor; $v$ — speed; $\vec v$ — velocity. Reading: energy and momentum are the same rest quantities $mc^2$ and $m\vec v$ stretched by $\gamma$ ; note $\vec p=(E/c^2)\vec v$ , so energy carries momentum whenever it moves. The part of the total energy that is not rest energy is the relativistic kinetic energy, and the small change in rest energy when a body absorbs or emits energy is the mass defect — the two faces of interconvertibility.

K=(\gamma-1)mc^2,\qquad \Delta m=\frac{\Delta E}{c^2} \qquad (6)

where: $K$ — kinetic energy (total minus rest); $\Delta m$ — change in rest mass; $\Delta E$ — energy absorbed ( $+$ ) or emitted ( $-$ ). Reading: kinetic energy is what is left of $E$ after removing $mc^2$ , and any energy leaving a system leaves as mass at the rate $1/c^2$ .

Derivation AFrom the four-momentum norm to the dispersion relation

The goal here is to show that the entire dynamical content — $E=mc^2$ , the dispersion relation, the invariance of mass — is not a set of separate postulates but a single geometric fact: the four-momentum has a fixed Minkowski length. We start from the definition of the four-velocity, use its metric normalization (established for any timelike worldline in the prerequisites), multiply through by the invariant mass, and contract. Every step is a one-line consequence of the metric, and the payoff is that the rest mass emerges as a frame-independent label carried along every worldline.

Step 1. Define the four-velocity as the derivative of position with respect to proper time, $u^\mu=\mathrm{d}x^\mu/\mathrm{d}\tau$ , and recall its normalization $u^\mu u_\mu=-c^2$ in signature $(-,+,+,+)$ . Why: proper time is the invariant parameter along the worldline, so differentiating with respect to it produces a genuine four-vector whose length is fixed by the metric, not by the observer.

Step 2. Multiply by the invariant rest mass to define the four-momentum, $p^\mu=m\,u^\mu$ , so that its time component is $\gamma mc = E/c$ and its space components are $\gamma m\vec v=\vec p$ . Why: mass is a Lorentz scalar, so multiplying a four-vector by it yields another four-vector — energy and momentum inherit the clean transformation law of $u^\mu$ .

Step 3. Contract with the metric: $p^\mu p_\mu=m^2\,u^\mu u_\mu=-m^2c^2$ . Why: the scalar $m^2$ pulls out of the contraction, and the normalization of Step 1 supplies the numerical value, giving an invariant that depends on nothing but the rest mass.

Step 4. Write the same contraction in components, $-(E/c)^2+\vec p^{\,2}=-(mc)^2$ , and multiply by $-c^2$ to obtain $E^2=(pc)^2+(mc^2)^2$ . Why: equating the coordinate form with the invariant value converts a geometric identity into the working dispersion relation; setting $\vec p=0$ then reads off $E_0=mc^2$ . *The derivation yields $E^2=(pc)^2+(mc^2)^2$ and $E_0=mc^2$ ; the $m=0$ limit gives $E=pc$ .*

Go deeper · Derivation B · 2 minEinstein's light-pulse argument for the inertia of energy

Derivation BEinstein's light-pulse argument for the inertia of energy

This is Einstein's own 1905 reasoning, and it is worth reconstructing in full because it derives $\Delta m=\Delta E/c^2$ without ever assuming it — it needs only energy conservation and how radiation energy transforms between frames. A body at rest emits two equal light pulses in opposite directions, so it stays at rest and only its internal energy changes. Comparing the energy budget in the rest frame with the budget in a frame where the body moves, the discrepancy can only be absorbed by a change in the body's mass. The argument is the cleanest demonstration that the inertia of a body tracks its energy content.

Step 1. In the rest frame $S$ , a stationary body emits two light pulses of energy $L/2$ each along $+x$ and $-x$ . Its momentum is unchanged (the pulses cancel), and by energy conservation its internal energy drops by $L$ : $E_{\text{after}}=E_{\text{before}}-L$ . Why: the symmetric emission keeps the body at rest, isolating a pure change of internal energy with no bulk motion to complicate the accounting.

Step 2. View the same event from a frame $S'$ moving at $-v$ , so the body drifts at $+v$ . Each pulse is Doppler-shifted, and the total emitted energy becomes $L'=\tfrac{L}{2}\gamma\!\left(1+\tfrac{v}{c}\right)+\tfrac{L}{2}\gamma\!\left(1-\tfrac{v}{c}\right)=\gamma L$ . Why: energy is the time component of the radiation's four-momentum, so it transforms with the same $\gamma$ that governs every boost — the forward and backward blueshift/redshift are asymmetric individually but sum to a clean $\gamma L$ .

Step 3. Subtract the two energy budgets. In $S$ the body loses $L$ ; in $S'$ it loses $\gamma L$ . The extra $(\gamma-1)L$ can only come from the body's kinetic energy, so $K_{\text{before}}-K_{\text{after}}=(\gamma-1)L$ . Why: the body's speed $v$ is identical before and after (it never moved in $S$ ), so a change in its kinetic energy at fixed $v$ can only mean a change in the mass multiplying $\tfrac12 v^2$ .

Step 4. Expand for small $v$ : $(\gamma-1)\approx \tfrac12 v^2/c^2$ , so $\Delta K=\tfrac12(\Delta m)v^2 = \tfrac{L}{c^2}\cdot\tfrac12 v^2$ , giving $\Delta m=L/c^2$ . Why: matching the kinetic-energy loss to the Newtonian form $\tfrac12 m v^2$ forces the mass to have dropped by exactly the emitted energy over $c^2$ — Einstein's conclusion that "the mass of a body is a measure of its energy content."

Go deeper · Derivation C · 1 minThe Newtonian limit of relativistic kinetic energy

Derivation CThe Newtonian limit of relativistic kinetic energy

A relativistic formula is only credible if it reduces to the familiar one where the familiar one works. The goal here is to show that the exact kinetic energy $K=(\gamma-1)mc^2$ contains $\tfrac12 mv^2$ as its leading term at low speed, and to read off the first relativistic correction — the term that starts to matter well before anything dramatic happens. This both validates the identification of $K$ as kinetic energy and locates precisely where Newtonian mechanics begins to under-count the energy.

Step 1. Start from the exact kinetic energy, the total energy minus the rest energy: $K=E-mc^2=(\gamma-1)mc^2$ . Why: rest energy is present whether or not the body moves, so the physically meaningful "energy of motion" is what remains after subtracting it.

Step 2. Binomial-expand the Lorentz factor for $v\ll c$ : $\gamma=(1-v^2/c^2)^{-1/2}\approx 1+\tfrac12\dfrac{v^2}{c^2}+\tfrac38\dfrac{v^4}{c^4}+\cdots$ . Why: the small parameter is $v^2/c^2$ , and the expansion is well-behaved for any sub-luminal speed, converging fast when $v$ is a modest fraction of $c$ .

Step 3. Substitute and keep terms: $K\approx\left(\tfrac12\dfrac{v^2}{c^2}+\tfrac38\dfrac{v^4}{c^4}\right)mc^2=\tfrac12 mv^2+\tfrac38\dfrac{mv^4}{c^2}+\cdots$ . Why: the leading term is exactly Newton's $\tfrac12 mv^2$ , confirming the interpretation of $K$ ; the next term is the first relativistic correction, growing as $v^4$ .

Step 4. Estimate when the correction bites: the fractional error of the Newtonian form is $\tfrac{3}{4}v^2/c^2$ , so at $v=0.1c$ it is already $0.75\%$ , and it diverges as $v\to c$ while $\tfrac12 mv^2$ stays finite. Why: this pinpoints the failure — the Newtonian energy is bounded, so it can never forbid $v>c$ , whereas the exact $K$ blows up and does. *The expansion confirms the leading term $\tfrac12 mv^2$ and first relativistic correction $\tfrac38 mv^4/c^2$ .*

Limiting cases

The master relations $E=\gamma mc^2$ and $E^2=(pc)^2+(mc^2)^2$ collapse to familiar and useful forms in each physical corner:

Case	Result	Meaning
Rest frame ( $v=0$ , $\vec p=0$ )	$E_0=mc^2$	Pure rest energy — the whole energy of a stationary body is its mass times $c^2$ .
Low speed ( $v\ll c$ )	$E\approx mc^2+\tfrac12 mv^2$	Rest energy plus the Newtonian kinetic energy as the first correction; classical mechanics recovered.
Massless particle ( $m=0$ )	$E=pc$	Photons and gravitons carry energy entirely as momentum and move at $c$ ; $E=mc^2$ gives $0$ and cannot be used.
Ultra-relativistic ( $pc\gg mc^2$ )	$E\approx pc+\dfrac{(mc^2)^2}{2pc}$	A fast massive particle behaves almost like a massless one, with a vanishing mass correction — the regime of collider beams.
Bound composite	$M<\sum_i m_i$	The system mass is the total energy in its rest frame; binding energy makes the whole lighter than the sum of parts.

Common trapE = mc² means mass is destroyed and converted into pure energy, so mass is not conserved

This is the single most common misreading, and it inverts the physics. In a closed system both total energy and total invariant mass are separately conserved — nothing is destroyed. When a uranium nucleus fissions, the summed rest masses of the fragments are less than the parent's, but the missing rest mass reappears exactly as the kinetic energy and radiation of the products; enclose the whole reaction in a perfect box and weigh it, and the box's mass never changes. What people call "mass converted to energy" is really rest mass being redistributed: energy that was locked inside one subsystem as its rest mass has been handed to other subsystems as their kinetic energy. Rest mass is not conserved reaction by reaction; the invariant mass of the total isolated system is. (A related trap: do not speak of a velocity-dependent "relativistic mass" $\gamma m$ . Modern usage reserves mass for the invariant $m$ and writes $E=\gamma mc^2$ ; calling $\gamma m$ a mass hides the fact that $m$ is the same for every observer and wrongly suggests a fast object could collapse into a black hole, which it cannot — in its own rest frame it is perfectly ordinary.)

Worked examples

1 · How much mass the Sun burns each second (numeric)

The Sun radiates at a luminosity $L=3.846\times10^{26}\ \mathrm{W}$ . Find the rate at which it loses rest mass, and check it against the efficiency of the proton-proton fusion chain that powers it.

Step 1. Every joule of radiated energy carries off mass at the rate $\Delta m=\Delta E/c^2$ , so the mass-loss rate is $\dot m=L/c^2$ . Numerically $\dot m=\dfrac{3.846\times10^{26}}{(2.998\times10^{8})^2}=\dfrac{3.846\times10^{26}}{8.988\times10^{16}}=\mathbf{4.28\times10^{9}\ kg\,s^{-1}}$ — about 4.3 million tonnes of mass radiated away every second.

Step 2. Cross-check with the nuclear ledger. The chain fuses four protons into one helium-4 nucleus; four hydrogen atoms ( $4\times1.007825\ \mathrm{u}$ ) exceed one helium-4 atom ( $4.002602\ \mathrm{u}$ ) by $\Delta m=0.02870\ \mathrm{u}$ , an energy release of $\Delta m\,c^2=0.02870\times931.5=\mathbf{26.7\ MeV}$ per helium nucleus.

Step 3. The fraction of mass that becomes energy is $0.02870/4.0313=\mathbf{0.71\%}$ — only seven parts per thousand of the fuel, yet at solar scale this converts $4.3\times10^{9}\ \mathrm{kg\,s^{-1}}$ . Over the Sun's $\sim\!10^{10}$ -year main-sequence life this is a tiny fraction of its $2\times10^{30}\ \mathrm{kg}$ total, which is why the Sun shines so long: a small mass fraction times an enormous $c^2$ is a vast energy budget.

2 · The binding energy and mass defect of helium-4 (numeric)

Given the constituent rest energies — proton $938.272\ \mathrm{MeV}$ , neutron $939.565\ \mathrm{MeV}$ — and the helium-4 nuclear rest energy $3727.379\ \mathrm{MeV}$ , find the binding energy, the mass defect in atomic mass units, and the binding energy per nucleon.

Step 1. Sum the free-constituent rest energies: $2(938.272)+2(939.565)=1876.544+1879.130=\mathbf{3755.674\ MeV}$ . This is what four separated nucleons weigh.

Step 2. The bound nucleus weighs less; the difference is the binding energy: $E_B=3755.674-3727.379=\mathbf{28.30\ MeV}$ . This is the energy that must be radiated to assemble the nucleus, and that must be resupplied to tear it apart.

Step 3. Convert the deficit to mass with $1\ \mathrm{u}=931.494\ \mathrm{MeV}/c^2$ : the mass defect is $\Delta m=28.30/931.494=\mathbf{0.0304\ u}$ , i.e. the nucleus is $0.0304\ \mathrm{u}$ lighter than its parts. Per nucleon the binding is $28.30/4=\mathbf{7.07\ MeV}$ , near the peak of the nuclear binding curve — which is exactly why helium is so stable and why hydrogen fusion is so exothermic.

3 · Why Newton's kinetic energy fails at the LHC (numeric)

A proton in the Large Hadron Collider is accelerated to a total energy $E=6.5\ \mathrm{TeV}$ . With proton rest energy $mc^2=0.938\ \mathrm{GeV}$ , find its Lorentz factor and speed, and compare the true kinetic energy with the Newtonian estimate.

Step 1. The Lorentz factor is the ratio of total to rest energy: $\gamma=E/mc^2=6500/0.938=\mathbf{6930}$ . The true kinetic energy is $K=E-mc^2=6500-0.938=\mathbf{6499\ GeV}$ , essentially all of the total energy.

Step 2. Solve $\gamma$ for the speed: $1-v/c\approx 1/(2\gamma^2)=1/(2\times6930^2)=\mathbf{1.0\times10^{-8}}$ , so $v=0.99999999\,c$ — the proton is within three metres per second of light speed.

Step 3. The Newtonian estimate uses $\tfrac12 mv^2\approx\tfrac12 mc^2=0.47\ \mathrm{GeV}$ . The ratio of truth to Newton is $6499/0.47\approx\mathbf{1.4\times10^{4}}$ : the Newtonian formula under-counts the proton's energy by more than four orders of magnitude, because it caps kinetic energy at $\tfrac12 mc^2$ while the true value grows without bound as $v\to c$ .

Physics of the portrait

The equation belongs to one man's remarkable year, and to a habit of mind — refusing to treat energy and inertia as separate — that runs straight from a Bern patent office to the source term of general relativity.

Origins · 2 minAlbert Einstein and how this was discovered

Albert Einstein

1879–1955 · special relativity and the inertia of energy

Einstein arrived at mass-energy equivalence almost as an afterthought. His main 1905 paper, "On the Electrodynamics of Moving Bodies," built special relativity from two postulates and never mentioned mass and energy in the same breath. It was a three-page sequel that September, "Does the Inertia of a Body Depend Upon Its Energy Content?", that carried the light-pulse argument reconstructed above and reached the conclusion that a body emitting energy $L$ loses mass $L/c^2$ — written then as $m=L/V^2$ , in his notation for the speed of light. Einstein was cautious and correct about what he had shown: not that mass is a store of extractable energy in some engineering sense, but that inertia and energy are the same quantity, so that "the mass of a body is a measure of its energy content." He returned to the idea repeatedly, extending it in 1907 to bound systems and to the gravitational weight of energy, an extension that became the germ of general relativity: if energy has inertia and inertia equals weight, then energy gravitates, and the source of gravity can no longer be mass density but must be the full stress-energy tensor. Predecessors had brushed against pieces of it — Poincaré had assigned a momentum, and hence an effective mass, to electromagnetic radiation in 1900 — but only Einstein saw it as a universal statement about all energy, independent of its form. The relation he first wrote as a footnote to relativity is now the most tested equation in physics, confirmed from nuclear reactions to Penning-trap mass spectrometry to better than one part in a million.

History

1900

— Poincaré assigns a momentum and an effective mass E/c² to electromagnetic radiation

1905

— Einstein derives Δm = L/c² in his September Annalen der Physik paper

1907

— Planck sharpens the reasoning and Einstein extends inertia of energy to bound systems and to gravitation

1932

— Cockcroft and Walton confirm the relation in the first accelerator-driven nuclear disintegration

2005

— Rainville and colleagues verify E = mc² to better than one part in a million using nuclear γ-rays and Penning-trap masses

Going further

Texts. Misner, Thorne & Wheeler, Gravitation, §2.8–2.9 (four-momentum and the mass shell) · Wald, General Relativity, Ch. 4 · Carroll, Spacetime and Geometry, §1.3 (energy-momentum four-vector) · Schutz, A First Course in General Relativity, Ch. 2 · Hartle, Gravity, Ch. 5 (relativistic dynamics). The core relations $E^2=(pc)^2+(mc^2)^2$ and $E=\gamma mc^2$ follow from the four-momentum norm $p^\mu p_\mu=-(mc)^2$ , itself derived from the four-velocity normalization $u^\mu u_\mu=-c^2$ in signature $(-,+,+,+)$ . Key anchor values: $c=299\,792\,458\ \mathrm{m\,s^{-1}}$ (exact, SI definition); electron rest energy $0.511\ \mathrm{MeV}$ , pair-production threshold $1.022\ \mathrm{MeV}$ ; helium-4 binding energy $28.3\ \mathrm{MeV}$ and mass defect $0.0304\ \mathrm{u}$ , from nucleon rest energies $938.272$ and $939.565\ \mathrm{MeV}$ and nuclear rest energy $3727.379\ \mathrm{MeV}$ ; solar mass-loss rate $L/c^2=4.28\times10^{9}\ \mathrm{kg\,s^{-1}}$ from $L=3.846\times10^{26}\ \mathrm{W}$ ; p-p chain release $26.7\ \mathrm{MeV}$ , efficiency $0.71\%$ . The light-pulse argument for $\Delta m=\Delta E/c^2$ is Einstein's (1905); the sharpest experimental confirmation, at $4\times10^{-7}$ , is from Rainville et al., Nature (2005).

Prerequisites. Four-Vectors · Minkowski Metric and Signature · Lorentz Transformations · Spacetime Interval. Leads to. Proper Time · Stress-Energy Tensor · Einstein Field Equations.

Ask the tutor: Ask me to work any reaction as a four-momentum balance — from a decaying muon to two colliding protons — and watch where the rest mass goes.

Is rest mass really "turned into" energy in a nuclear reaction?

Not in the way the phrase suggests. In an isolated reaction, energy and invariant mass are each conserved; what changes is how rest mass is distributed among the pieces. Before fission the mass lives in one heavy nucleus; afterward some of it has moved out of the fragments' rest masses and into their kinetic energy and radiation. If you weighed a sealed, perfectly insulated box containing the whole reaction, its mass would be unchanged before and after. "Mass converted to energy" is a shorthand for "rest mass of the products is less than that of the reactants, and the difference shows up as kinetic energy" — a redistribution, not a destruction.

Does a hot object genuinely weigh more than a cold one?

Yes, though the effect is far too small to weigh directly. Any internal energy — thermal motion, compression, excitation — adds to a system's rest energy and therefore to its rest mass by $\Delta m=\Delta E/c^2$ . Heating a kilogram of water by a few degrees adds a few kilojoules and so about $10^{-14}\ \mathrm{kg}$ ; a charged battery outweighs a flat one by a comparable trifle. The principle is exact even where the number is unmeasurable: the mass of a composite body is the total energy in its rest frame divided by $c^2$ , and that total includes every form of internal energy, not just the masses of the constituents.

If mass and energy are the same, why keep two words?

Because they are the same quantity measured against different backgrounds. "Rest energy" emphasizes that $mc^2$ is on the same footing as kinetic or thermal energy and can be exchanged with them. "Mass" emphasizes that this particular energy — the energy in the rest frame — is a Lorentz invariant, a fixed label every observer agrees on, unlike the total energy which grows with $\gamma$ . Keeping both words keeps both facts in view: mass is the invariant magnitude of the energy-momentum four-vector, while energy is its frame-dependent time component. In natural units ( $c=1$ ) the distinction in units vanishes, but the distinction between an invariant and a component remains.

Why is the invariant mass of a system not the sum of its parts' masses?

Because mass is not additive — energy is, and mass is the rest-frame energy divided by $c^2$ . To combine particles you add their four-momenta component by component, then take the Minkowski length of the total. Kinetic energy and binding energy contribute to that total, so the system mass can be larger than the sum (two fast particles flying apart) or smaller (a bound nucleus that shed binding energy). Two photons, each massless, can even form a system with nonzero invariant mass if they travel in different directions. The rule is always the same: build $p^\mu_{\text{tot}}=\sum_i p^\mu_i$ first, and only then read off $M^2c^2=-p^\mu_{\text{tot}}p_{\mu,\text{tot}}$ .

Is "relativistic mass" simply wrong, or just old-fashioned?

It is not internally inconsistent, but it is a poor bookkeeping choice that modern physics has abandoned. Defining $m_{\text{rel}}=\gamma m$ lets you write $E=m_{\text{rel}}c^2$ and $\vec p=m_{\text{rel}}\vec v$ , which look tidy, but it disguises the one quantity that matters most — the invariant $m$ — and invites real errors, such as imagining that a fast particle's increased "mass" strengthens its gravity or could form a black hole (it cannot; in its own frame it is unremarkable). The clean convention keeps $m$ as the invariant rest mass and puts all the velocity dependence in explicit factors of $\gamma$ : $E=\gamma mc^2$ , $\vec p=\gamma m\vec v$ . This is why "mass" in relativity now means only rest mass.

How exactly does E = mc² lead into general relativity?

Through two links. First, the equivalence principle sets inertial mass equal to gravitational mass, so anything with inertia has weight. Second, $E=mc^2$ says every form of energy has inertia. Chaining them: every form of energy — kinetic, thermal, radiative, and the momentum flux and pressure that come with them — must gravitate. Newtonian gravity, sourced by mass density alone, cannot express this. Its relativistic replacement must be sourced by an object that packages energy density, momentum density, and stress into one covariant quantity — the stress-energy tensor $T^{\mu\nu}$ . That tensor is the right-hand side of the Einstein field equations, which is why mass-energy equivalence is the true entry point from the kinematics of special relativity to the curved-spacetime dynamics of general relativity.

Has E = mc² actually been measured, or is it purely theoretical?

It is among the most stringently tested relations in physics. Every nuclear mass measurement confirms it indirectly: binding energies from mass defects match calorimetric energy releases, and Cockcroft and Walton's 1932 lithium disintegration was an early direct check. The sharpest test compared the gamma-ray energy emitted when nuclei capture a neutron against the corresponding mass differences measured in a Penning trap; the two agreed to within $4\times10^{-7}$ , so $E=mc^2$ holds to better than one part in a million. Particle physics tests it constantly in reverse, converting collision energy into fresh rest mass. There is no credible evidence of any deviation.