Neutrino mass mechanisms

Every other fermion in the Standard Model gets its mass the same way: it exists in both left- and right-handed copies, and the Higgs field bridges them. The neutrino breaks this pattern twice over. Oscillation experiments proved it has a mass — but a mass some ten million times smaller than the next-lightest fermion, the electron, and the Standard Model as written has no right-handed neutrino to pair it with at all. So the question "how does the neutrino get its mass?" is really two questions stacked on top of each other: what term in the Lagrangian could do the job, and why is the answer so absurdly tiny. There are two candidate terms, and the neutrino's electric neutrality is what makes the second one possible. A Dirac mass, built like every other fermion's from a Yukawa coupling to the Higgs, works — but only if the neutrino's coupling is $10^{7}$ times feebler than the electron's, a number no one can explain. A Majorana mass exploits something unique: because a right-handed neutrino carries no charge of any kind, it is a total singlet under the gauge group, and you may hand it an arbitrarily large mass $M$ for free. Writing both terms together produces a $2\times2$ mass matrix whose diagonalisation performs a small miracle: one eigenvalue is pushed up to $M$ , the other pushed down to $m_D^2/M$ . Make $M$ enormous and the light neutrino becomes tiny — not by fine-tuning a coupling, but automatically. This is the seesaw mechanism, and it turns the embarrassment of the neutrino's smallness into a telescope pointed at physics near the grand-unified scale.

Key takeaways

Oscillations prove neutrinos are massive, so the Lagrangian needs a neutrino mass term — but the mass is $\sim10^{7}$ times below the electron's, demanding an explanation.
A Dirac mass $m_D=y_\nu v/\sqrt2$ copies every other fermion but needs a Yukawa coupling $y_\nu\sim3\times10^{-13}$ , unnaturally small.
Because a right-handed neutrino is a total gauge singlet, a Majorana mass $-\tfrac12 M\,\overline{\nu_R^c}\nu_R$ can be added for free — impossible for any charged fermion.
Together they give a mass matrix $\left(\begin{smallmatrix}0&m_D\\m_D&M\end{smallmatrix}\right)$ that diagonalises to a light state $m_\nu\approx m_D^2/M$ and a heavy state $m_N\approx M$ — the seesaw.
With $m_D\sim100$ GeV and $M\sim2\times10^{14}$ GeV one naturally gets $m_\nu\sim0.05$ eV; a Majorana nature would show up as neutrinoless double-β decay.
Key idea: because right-handed neutrinos are gauge singlets you can give them a huge Majorana mass M for free, and diagonalising sees-saws the observable neutrino mass down to m_D²/M — turning the puzzle of tiny masses into evidence for new physics at a very high scale.

The core insight

Because right-handed neutrinos are gauge singlets, you can give them a huge Majorana mass $M$ for free; diagonalising the combined mass matrix sees-saws the observable neutrino mass down to $m_D^2/M$ , turning the puzzle of tiny masses into evidence for new physics at a very high scale.

Where this lives today

The seesaw's most spectacular consequence is not in the laboratory but in the early universe. If the heavy Majorana neutrinos $N$ existed, they decayed shortly after the Big Bang — and because their Majorana mass violates lepton number and their couplings can violate CP, those decays could have produced a slight excess of leptons over antileptons. Electroweak "sphaleron" processes then convert part of that lepton asymmetry into a baryon asymmetry, and the result may be the entire reason matter survived to make galaxies. This is leptogenesis, and it ties the neutrino mass mechanism directly to the matter–antimatter asymmetry of the universe. On the experimental front, the decisive test is neutrinoless double-β decay ( $0\nu\beta\beta$ ): a nucleus emitting two electrons and no neutrinos, which can only happen if the neutrino is Majorana. The current leaders — KamLAND-Zen with $^{136}$ Xe and GERDA/LEGEND with $^{76}$ Ge — have pushed half-life limits past $10^{26}$ years, cornering the effective mass $m_{\beta\beta}$ below roughly $0.03$ – $0.15$ eV (the spread reflects nuclear-matrix-element uncertainty). Meanwhile KATRIN weighs the neutrino kinematically from the tritium β-decay endpoint, having tightened the direct bound to $m_\beta<0.8$ eV and aiming at $0.2$ eV, and cosmological surveys squeeze $\sum m_i$ from the other side. A single unambiguous $0\nu\beta\beta$ event would settle the Dirac-versus-Majorana question that Majorana himself opened, and point straight at the high-scale physics the seesaw predicts.

The physical picture

The name is not a metaphor to be apologised for — it is the mechanism. Picture a playground see-saw. The pivot is the Dirac mass $m_D$ , the scale at which the Higgs would ordinarily set a fermion's mass. On the heavy end sits the right-handed neutrino with its enormous Majorana mass $M$ ; on the light end sits the neutrino we actually observe. Because the product of the two arms is fixed at $m_D^2$ , pushing the heavy end down — making $M$ larger — lifts the light end higher toward masslessness. The neutrino is light because its partner is heavy. The first figure makes this literal, with real numbers on both ends.

Figure

The seesaw with numbers: a Dirac scale m_D ≈ 1 GeV at the pivot, a heavy right-handed neutrino of mass M ≈ 10¹¹ GeV weighing down one end, and the observed light neutrino m_ν ≈ 0.01 eV riding high on the other. The product of the two masses is fixed at m_D², so m_ν ≈ m_D²/M — raise M and m_ν sinks toward zero

But a see-saw only explains why the mass is small once you accept that a Majorana mass $M$ can be large. The deeper picture is one of scales. Line up every fermion mass on a logarithmic ruler and the neutrino sits alone, seven orders of magnitude below the electron and utterly disconnected from the tidy ladder of charged-fermion masses that the Higgs vacuum value $v=246$ GeV anchors. The seesaw says that emptiness is not empty: the neutrino's mass is the electroweak scale $v$ reflected off a mass barrier $M$ far to the right, near the grand-unified scale. The second figure is that ruler, and it is the whole motivation for the mechanism in one glance.

Figure

The fermion mass ladder on a logarithmic scale from 0.01 eV to 10¹⁵ GeV. Charged fermions run from the electron (0.511 MeV) to the top quark (173 GeV), all near the Higgs scale v = 246 GeV. The neutrino sits far to the left at ~0.05 eV, seven orders below the electron. The seesaw ties it to a scale M ~ 10¹¹–10¹⁵ GeV on the far right through m_ν ≈ m_D²/M

The core equations

The Dirac option is the conservative one: treat the neutrino exactly like an up-type quark. Postulate a right-handed partner $\nu_R$ , couple it to the left-handed doublet through the conjugate Higgs field $\tilde\Phi=i\sigma_2\Phi^*$ , and let electroweak symmetry breaking do the rest. The Yukawa term becomes a mass linking the two chiralities.

\mathcal{L}_D=-y_\nu\,\overline{L}\,\tilde\Phi\,\nu_R+\text{h.c.}\ \xrightarrow{\ \langle\Phi\rangle=v/\sqrt2\ }\ \mathcal{L}_D=-m_D\left(\overline{\nu_L}\nu_R+\overline{\nu_R}\nu_L\right),\qquad m_D=\frac{y_\nu v}{\sqrt2} \qquad (1)

where: $y_\nu$ — the neutrino Yukawa coupling; $v=246$ GeV — the Higgs vacuum expectation value; $m_D$ — the Dirac mass; $\nu_L,\nu_R$ — the two chiral components. Reading it: this is an honest mass, but it conserves lepton number and forces $y_\nu\approx3\times10^{-13}$ to reach $m_\nu\sim0.05$ eV — a coupling with no explanation. The Majorana option uses a term forbidden to every charged fermion. Because $\nu_R$ is a complete gauge singlet, the bilinear $\overline{\nu_R^c}\nu_R$ carries no gauge charge and no Higgs is needed to make it invariant; you simply write down a bare mass $M$ .

\mathcal{L}_M=-\tfrac12 M\left(\overline{\nu_R^c}\,\nu_R+\overline{\nu_R}\,\nu_R^c\right),\qquad \nu_R^c=\hat C\overline{\nu_R}^{\,\top} \qquad (2)

where: $M$ — the Majorana mass, unbounded because no symmetry protects it; $\nu_R^c$ — the charge-conjugate field, the left-handed antineutrino. Reading it: this term couples a particle to an antiparticle, violating lepton number by two units ( $\Delta L=2$ ), and it makes the neutrino its own antiparticle. It is the crux of everything that follows. Keeping both terms, the neutrino mass Lagrangian is a quadratic form in the two-component vector $(\nu_L,\ \nu_R^c)$ . Its coefficients assemble into a symmetric $2\times2$ Majorana mass matrix — zero in the top-left because a bare $\nu_L$ Majorana mass is gauge-forbidden, $m_D$ off-diagonal, $M$ in the bottom-right.

\mathcal{L}=-\tfrac12\begin{pmatrix}\overline{\nu_L^c}&\overline{\nu_R}\end{pmatrix}\begin{pmatrix}0&m_D\\ m_D&M\end{pmatrix}\begin{pmatrix}\nu_L\\ \nu_R^c\end{pmatrix}+\text{h.c.} \qquad (3)

where: the upper-left $0$ is enforced by $SU(2)_L\times U(1)_Y$ invariance; $m_D$ comes from the Higgs coupling (1); $M$ is the bare singlet mass (2). Reading it: diagonalising this matrix gives the two physical masses, exactly as one diagonalises the neutral-kaon or photon- $Z$ systems. In the physically relevant hierarchy $M\gg m_D$ the eigenvalues split dramatically. The light eigenvalue is suppressed by the ratio $m_D/M$ ; the heavy one is essentially $M$ itself.

m_{\text{light}}\approx\frac{m_D^2}{M},\qquad m_{\text{heavy}}\approx M,\qquad \tan 2\theta=\frac{2m_D}{M}\ \Rightarrow\ \theta\approx\frac{m_D}{M} \qquad (4)

where: $m_{\text{light}}$ — the observed neutrino, mostly $\nu_L$ ; $m_{\text{heavy}}$ — a new heavy Majorana neutrino $N$ , mostly $\nu_R$ ; $\theta$ — their tiny mixing. Reading it: the light neutrino is small not because a coupling is small but because a mass is large — the seesaw's whole payoff. At energies far below $M$ the heavy neutrino is invisible and can be integrated out, leaving a single dimension-five operator — the only one the Standard Model admits. It generates a Majorana mass for the light neutrino directly and is the model-independent face of the seesaw.

\mathcal{L}_5=\frac{c}{\Lambda}\left(\overline{L^c}\,\tilde\Phi\right)\!\left(\tilde\Phi^{\!\top}L\right)+\text{h.c.}\ \xrightarrow{\ \langle\Phi\rangle=v/\sqrt2\ }\ m_\nu=\frac{c\,v^2}{\Lambda},\qquad m_{\beta\beta}=\Big|\textstyle\sum_i U_{ei}^2\,m_i\Big| \qquad (5)

where: $\Lambda$ — the scale of new physics (in the seesaw, $\Lambda\to M$ with $c=y_\nu^2/2$ ); $c$ — a dimensionless coefficient; $m_{\beta\beta}$ — the effective Majorana mass probed by neutrinoless double-β decay, with $U_{ei}$ the mixing-matrix elements. Reading it: whatever the ultraviolet completion, light Majorana masses scale as $v^2/\Lambda$ , and their smoking gun is a nonzero $m_{\beta\beta}$ .

Derivation ADiagonalising the seesaw mass matrix

The goal is to take the combined Dirac-plus-Majorana mass matrix of equation (3), diagonalise it exactly, and then expand in the hierarchy $M\gg m_D$ to prove that one physical mass is driven down to $m_D^2/M$ while the other rises to $M$ , with a mixing angle of order $m_D/M$ . This is the calculation that converts the qualitative see-saw picture into the precise relation that underlies every numerical estimate in this article, and it is worth doing in full rather than quoting.

Step 1. Write the characteristic equation for the symmetric matrix $\bigl(\begin{smallmatrix}0&m_D\\ m_D&M\end{smallmatrix}\bigr)$ . Its eigenvalues $\lambda$ satisfy $\det\bigl(\begin{smallmatrix}-\lambda&m_D\\ m_D&M-\lambda\end{smallmatrix}\bigr)=\lambda^2-M\lambda-m_D^2=0$ . Why: the physical masses are the eigenvalues of the mass matrix, and everything else follows from this quadratic.

Step 2. Solve exactly: $\lambda_\pm=\tfrac12\left(M\pm\sqrt{M^2+4m_D^2}\right)$ . These are the two exact Majorana mass eigenvalues, valid for any ratio $m_D/M$ . Why: keeping the exact roots first lets us expand cleanly and see precisely which terms survive in the hierarchy limit.

Step 3. Expand for $M\gg m_D$ using $\sqrt{M^2+4m_D^2}=M\sqrt{1+4m_D^2/M^2}\approx M\left(1+2m_D^2/M^2\right)$ . The heavy root is $\lambda_+\approx M+m_D^2/M\approx M$ , and the light root is $\lambda_-\approx-\,m_D^2/M$ . Why: the two nearly-cancelling terms in $\lambda_-$ leave exactly the seesaw suppression $m_D^2/M$ , while $\lambda_+$ barely shifts from $M$ .

Step 4. Deal with the sign and read off the mixing. The negative sign of $\lambda_-$ is unphysical — it is removed by a phase (chiral) redefinition $\nu\to i\gamma_5\nu$ of the Majorana field, leaving the physical mass $m_{\text{light}}=m_D^2/M>0$ . The rotation that diagonalises the matrix has $\tan2\theta=2m_D/(M-0)$ , so $\theta\approx m_D/M$ is tiny. Why: this fixes the light state as almost pure $\nu_L$ with only an $O(m_D/M)$ admixture of the heavy singlet, which is why the heavy neutrino is so hard to see directly. *As a concrete check: for $m_D=1$ , $M=1000$ the exact eigenvalues are $\{-9.99999\times10^{-4},\ 1000.001\}$ , in precise agreement with $-m_D^2/M=-10^{-3}$ and $M$ ; the mixing angle is $\theta\approx m_D/M=10^{-3}$ .*

Go deeper · Derivation B · 1 minThe seesaw as the leading Weinberg operator

Derivation BThe seesaw as the leading Weinberg operator

The aim is to show that the seesaw is not an isolated trick but the tree-level realisation of the single lowest-dimension operator by which any heavy new physics can give the neutrino a mass — the dimension-five Weinberg operator of equation (5). Establishing this makes the result robust: whatever sits at the scale $M$ , integrating it out at low energy produces the same $m_\nu\sim v^2/\Lambda$ scaling, and the specific seesaw simply fixes the coefficient. We integrate the heavy neutrino out and match.

Step 1. Start from the seesaw Lagrangian with the heavy field: a Yukawa term $-y_\nu\,\overline{L}\tilde\Phi\,\nu_R$ and a Majorana mass $-\tfrac12 M\,\overline{\nu_R^c}\nu_R$ . The equation of motion for the heavy $\nu_R$ , dropping its kinetic term at energies $\ll M$ , gives $\nu_R\approx (y_\nu/M)\,\tilde\Phi^\dagger L$ . Why: below $M$ the heavy neutrino has no independent dynamics; it is slaved to the light fields, and substituting it back removes it from the theory.

Step 2. Substitute this back into the Yukawa term. The two powers of $y_\nu$ and one power of $1/M$ combine into an effective operator $\mathcal{L}_{\text{eff}}=\dfrac{y_\nu^2}{2M}\left(\overline{L^c}\tilde\Phi\right)\!\left(\tilde\Phi^{\!\top}L\right)$ , which is exactly the Weinberg operator (5) with $c/\Lambda=y_\nu^2/(2M)$ . Why: this is the matching condition — it identifies the abstract coefficient $c/\Lambda$ with concrete seesaw parameters, so the low-energy operator and the high-energy model agree.

Step 3. Insert the Higgs vacuum value $\langle\Phi\rangle=v/\sqrt2$ to read off the neutrino mass: $m_\nu=\dfrac{y_\nu^2 v^2}{2M}$ . Using $m_D=y_\nu v/\sqrt2$ , i.e. $m_D^2=y_\nu^2 v^2/2$ , this is identically $m_\nu=m_D^2/M$ . Why: the operator route and the matrix-diagonalisation route give the same answer, a strong internal consistency check on both.

Step 4. Note what is model-independent. The operator has dimension five, so it is suppressed by exactly one power of the new scale, $m_\nu\propto v^2/\Lambda$ — the unique way a renormalisable extension can make a small Majorana mass. It also necessarily violates lepton number by two units. Why: this is why any observation of a Majorana neutrino mass points to lepton-number violation at a high scale, independent of which seesaw (type I, II, or III) is realised in nature.

Go deeper · Derivation C · 2 minWhy a Majorana mass is allowed for the singlet but forbidden for the doublet

Derivation CWhy a Majorana mass is allowed for the singlet but forbidden for the doublet

The purpose here is to make precise the single most important and most misunderstood fact in the whole subject: a bare Majorana mass is gauge-invariant for the right-handed neutrino but strictly forbidden for the left-handed one, which is exactly the origin of the zero in the top-left corner of the seesaw matrix. We do this by tracking the gauge quantum numbers of each candidate mass bilinear under $SU(2)_L\times U(1)_Y$ .

Step 1. List the quantum numbers. The right-handed neutrino $\nu_R$ is a total singlet: $SU(3)$ singlet, $SU(2)_L$ singlet, hypercharge $Y=0$ , electric charge $Q=0$ . The left-handed neutrino lives inside the lepton doublet $L=(\nu_L,\ e_L)$ with $SU(2)_L$ doublet and $Y=-\tfrac12$ . Why: a mass term is only allowed if it is invariant under every gauge symmetry, so the quantum numbers of the bilinear decide its fate.

Step 2. Test the singlet Majorana term. The bilinear $\overline{\nu_R^c}\nu_R$ carries $2\times$ (the charges of $\nu_R$ ) $=0$ for every gauge factor — colour, weak isospin, and hypercharge all vanish. It is therefore gauge-invariant on its own, with no Higgs field required. Why: this is precisely what lets $M$ be arbitrarily large; it is not tied to electroweak symmetry breaking, so it can sit at any scale up to the Planck mass.

Step 3. Test the doublet Majorana term. The bilinear $\overline{\nu_L^c}\nu_L$ carries twice the doublet's charges: it transforms as an $SU(2)_L$ triplet with $Y=-1$ . It is not invariant, so a bare $m_L\,\overline{\nu_L^c}\nu_L$ is forbidden; generating it requires either a Higgs triplet with $Y=1$ (the type-II seesaw) or the dimension-five operator built from two Higgs doublets. Why: this forbidden entry is the zero in the top-left of the mass matrix, and its absence is what makes the seesaw suppression $m_D^2/M$ rather than a generic mix of scales.

Step 4. Contrast with a charged fermion. For the electron, a Majorana term $\overline{e^c}e$ carries electric charge $2$ , so it would permit charge-violating transitions like $e^-\!\leftrightarrow e^+$ and is absolutely forbidden by the exactly conserved electromagnetic $U(1)$ . Only the neutrino, being neutral, escapes this. Why: electric neutrality is the one property that distinguishes the neutrino and opens the Majorana door — no other Standard Model fermion can be its own antiparticle.

Limiting cases

Case	Result	Meaning
Pure Dirac ( $M=0$ )	$\nu_L,\nu_R$ pair into one Dirac fermion of mass $m_D$	lepton number conserved; needs $y_\nu\sim10^{-13}$ , unexplained
Pure Majorana ( $m_D=0$ )	$\nu_L$ massless, $\nu_R$ has mass $M$ ; the two decouple	no light mass generated; a sterile heavy state
Seesaw ( $M\gg m_D$ )	$m_{\text{light}}\approx m_D^2/M$ , $m_{\text{heavy}}\approx M$	tiny mass from a heavy scale, no small coupling
Decoupling ( $M\to\infty$ )	$m_{\text{light}}\to0$ , $\nu_R$ removed	strict Standard Model with massless neutrinos recovered
Pseudo-Dirac ( $M\sim m_D$ )	two nearly-degenerate Majorana states, splitting $\sim M$	a Dirac neutrino is the $M\to0$ limit of this
Weinberg operator	$m_\nu=c\,v^2/\Lambda$	model-independent $\Delta L=2$ mass, any UV completion

Common trapA Majorana mass must be forbidden by gauge invariance, just as it is for the electron

This is the central misconception, and getting it right is the whole point. For a charged fermion a Majorana mass term $\overline{\psi^c}\psi$ carries twice the particle's electric charge, so for the electron it would violate charge conservation by two units and permit $e^-\!\leftrightarrow e^+$ — absolutely forbidden by the exactly conserved electromagnetic $U(1)$ . It is natural to assume the same veto applies to neutrinos. It does not. The neutrino is electrically neutral, and the right-handed neutrino is neutral under every Standard Model gauge factor — colour, weak isospin, and hypercharge alike. The bilinear $\overline{\nu_R^c}\nu_R$ therefore carries no conserved gauge charge at all, and nothing in the gauge symmetry forbids a bare mass $M$ of any size. The only quantity it violates is lepton number, which is not a gauge symmetry but a mere accidental global symmetry of the renormalisable Standard Model — there is no gauge boson coupled to it and no theorem protecting it. So the Majorana door is open precisely for the neutrino and shut for everything else, and that single asymmetry is what makes the seesaw possible. (A second trap: the left-handed Majorana term is genuinely forbidden — but by $SU(2)_L\times U(1)_Y$ , not by electromagnetism — which is why it appears only at dimension five.)

Worked examples

1 · The seesaw scale from a top-like Dirac mass (fully worked)

Suppose the neutrino's Dirac mass is of order the top quark's, $m_D\approx100$ GeV, and the observed neutrino mass is set by the atmospheric splitting, $m_\nu\approx\sqrt{\Delta m^2_{31}}\approx0.05$ eV. What Majorana mass $M$ does the seesaw require, and what does that scale suggest?

Step 1. Rearrange the seesaw relation $m_\nu\approx m_D^2/M$ for the unknown: $M\approx m_D^2/m_\nu$ . This is the only equation needed; everything is fixed once $m_D$ and $m_\nu$ are chosen. The heavy scale is $M\approx m_D^2/m_\nu$ .

Step 2. Put everything in electronvolts. $m_D=100$ GeV $=1.0\times10^{11}$ eV, so $m_D^2=1.0\times10^{22}\ \text{eV}^2$ . With $m_\nu=0.05$ eV, $M=1.0\times10^{22}/0.05=2.0\times10^{23}$ eV. Converting back, $M\approx2\times10^{14}$ GeV.

Step 3. Interpret the scale. This lands within a couple of orders of the grand-unified scale $M_{\text{GUT}}\sim10^{15\text{–}16}$ GeV — not a coincidence anyone imposed, but an output of demanding a top-like Dirac coupling and the measured neutrino mass. A natural $O(1)$ Yukawa plus a GUT-scale $M$ reproduces the neutrino mass with no small numbers.

Step 4. Check the opposite convention. If instead $m_D\approx1$ GeV (a charm-like coupling), then $M=m_D^2/m_\nu=(10^9)^2/0.05=2\times10^{19}$ eV, i.e. $2\times10^{10}$ GeV; for the lighter $m_\nu=0.01$ eV the same coupling gives $M=10^{20}\,\text{eV}=10^{11}$ GeV. The required scale slides between $10^{11}$ and $10^{15}$ GeV depending on $m_D$ — always far above anything a collider can reach, which is why the heavy neutrino is not seen directly.

2 · How unnatural the pure-Dirac option is

If the neutrino were a pure Dirac fermion with $m_\nu=0.05$ eV, what Yukawa coupling would it need, and how does that compare with the electron and top?

Step 1. Invert equation (1): $y_\nu=\sqrt2\,m_\nu/v$ . With $m_\nu=0.05$ eV $=5\times10^{-11}$ GeV and $v=246$ GeV, $y_\nu=1.414\times5\times10^{-11}/246$ . This gives $y_\nu\approx2.9\times10^{-13}$ .

Step 2. Compare with the electron: $y_e=\sqrt2\,m_e/v=1.414\times0.511\times10^{-3}/246\approx2.9\times10^{-6}$ . The neutrino Yukawa would be about $10^{7}$ times smaller than the electron's — itself already the smallest charged-fermion coupling.

Step 3. Compare with the top: $y_t=\sqrt2\times173/246\approx0.99\approx1$ . So the pure-Dirac picture spreads the Yukawa couplings across $13$ orders of magnitude, from $y_t\sim1$ to $y_\nu\sim10^{-13}$ , with no principle explaining the tiny end — exactly the "naturalness" objection the seesaw removes by trading a small coupling for a large mass.

3 · From oscillations to an absolute mass floor

Oscillations measure only mass-squared differences: $\Delta m^2_{21}=7.5\times10^{-5}\,\text{eV}^2$ and $|\Delta m^2_{31}|=2.5\times10^{-3}\,\text{eV}^2$ . What is the minimum possible mass of the heaviest neutrino, and the minimum sum $\sum m_i$ , and how do they compare with the cosmological bound?

Step 1. The heaviest state must carry at least the largest splitting: $m_{\text{heaviest}}\ge\sqrt{|\Delta m^2_{31}|}=\sqrt{2.5\times10^{-3}}$ . This gives $m_{\text{heaviest}}\ge0.050$ eV — a hard floor set purely by oscillation data.

Step 2. In the normal ordering, take the lightest state $m_1\to0$ ; then $m_2=\sqrt{\Delta m^2_{21}}=8.7\times10^{-3}$ eV and $m_3=\sqrt{\Delta m^2_{31}}=0.050$ eV. The minimum sum is $\sum m_i\approx0+0.0087+0.050\approx0.059$ eV.

Step 3. Compare with cosmology. The Planck-plus-BAO bound is $\sum m_i<0.12$ eV. The oscillation floor $0.059$ eV and the cosmological ceiling $0.12$ eV bracket the absolute scale into a narrow window — one that upcoming experiments will close in on, and that no mass mechanism can predict without fixing $M$ and $m_D$ separately.

Physics of the portrait

The whole edifice rests on one 1937 idea from a physicist who vanished the following year: that a neutral fermion can be its own antiparticle. Everything downstream — the Majorana mass, the seesaw, neutrinoless double-β decay, leptogenesis — is a consequence of Ettore Majorana's symmetric theory.

Origins · 1 minEttore Majorana and how this was discovered

Ettore Majorana

1906–1938 · the Majorana equation and fermions that are their own antiparticles

Ettore Majorana was, by Enrico Fermi's own reckoning, a genius of the rank of Galileo and Newton — and one of the strangest figures in the history of physics. Born in Catania, Sicily, he joined Fermi's group in Rome (the "Via Panisperna boys") in 1928 and worked with a reluctance to publish that bordered on the pathological, solving problems in his head and throwing the results away. In 1932 he worked out a correct proton–neutron theory of the nucleus before Heisenberg published his, and declined to write it up. His last and most consequential paper, Teoria simmetrica dell'elettrone e del positrone (1937), asked a question no one had thought to ask: must a spin-½ field be complex, with distinct particle and antiparticle, as Dirac's equation assumes? Majorana showed it need not. For a neutral fermion one can impose a reality condition, $\psi=\psi^c$ , so that the particle is its own antiparticle — the Majorana fermion. At the time it was a curiosity with no known application; the neutrino was still hypothetical. Today it is the pivot of neutrino physics: if the neutrino is Majorana, it can carry the lepton-number-violating mass that drives the seesaw, and it must reveal itself through neutrinoless double-β decay. In March 1938 Majorana withdrew his savings, boarded a ship between Palermo and Naples, and vanished without trace at the age of thirty-one — a mystery never solved. The idea he left behind now anchors the leading explanation for why the neutrino has any mass at all.

History

1937

— Ettore Majorana formulates neutral fermions that are their own antiparticles

1939

— Furry proposes neutrinoless double-β decay as a test of the Majorana nature

1977

— Minkowski writes the type-I seesaw relation m_ν ≈ m_D²/M

1979

— Gell-Mann, Ramond, Slansky, Yanagida and Weinberg's dimension-five operator ground the mechanism

1998

— Super-Kamiokande's atmospheric oscillations prove neutrinos are massive

Going further

Texts. Thomson, Modern Particle Physics, §17.8–17.8.1 (neutrino masses and the seesaw), Eq. (17.53)–(17.54) · Griffiths, Introduction to Elementary Particles, 2nd ed., §11.7 (neutrino mass and mixing) · Halzen & Martin, Quarks and Leptons, §12.6 · Peskin & Schroeder, An Introduction to Quantum Field Theory, §3.4 (Majorana fermions) and §20.2 · Schwartz, Quantum Field Theory and the Standard Model, §11.3 and §29 (Majorana masses, the Weinberg operator). The key numerical results in this article are: the seesaw eigenvalues $\lambda_\pm=\tfrac12(M\pm\sqrt{M^2+4m_D^2})$ reduce to $m_D^2/M$ and $M$ in the hierarchy $M\gg m_D$ ; $M=m_D^2/m_\nu=2\times10^{14}$ GeV for $m_D=100$ GeV and $m_\nu=0.05$ eV, and $10^{11}$ GeV for $m_D=1$ GeV and $m_\nu=0.01$ eV; $y_\nu=\sqrt2\, m_\nu/v=2.9\times10^{-13}$ versus $y_e=2.9\times10^{-6}$ and $y_t\approx1$ ; $m_{\text{heaviest}}\ge\sqrt{2.5\times10^{-3}}=0.050$ eV and, in the normal ordering, $\sum m_i\ge0.059$ eV against the cosmological bound of $0.12$ eV; and the Weinberg-operator matching gives $c/\Lambda=y_\nu^2/(2M)$ , reproducing $m_\nu=m_D^2/M$ . The oscillation parameters $\Delta m^2_{21}=7.5\times10^{-5}\,\text{eV}^2$ and $|\Delta m^2_{31}|=2.5\times10^{-3}\,\text{eV}^2$ and the $0\nu\beta\beta$ limits quoted are standard 2020s global-fit values.

Prerequisites. Chirality and Weyl Spinors · The Higgs Mechanism · Majorana vs Dirac Neutrinos · Mass and Weak Eigenstates. Leads to. Neutrinoless Double-β Decay · Neutrino Mass Beyond the Standard Model · Matter–Antimatter Asymmetry · Grand Unification.

Ask the tutor: Ask me to diagonalise the seesaw matrix step by step, or to show why a Majorana mass is free for the singlet neutrino but forbidden for the electron.

Why can't the neutrino just get its mass from the Higgs like every other fermion?

It can, in principle — that is exactly the Dirac option. You postulate a right-handed neutrino $\nu_R$ , couple it to the Higgs, and after symmetry breaking you get a mass $m_D=y_\nu v/\sqrt2$ just as for the electron or top quark. Two things make physicists uneasy about stopping there. First, the required Yukawa coupling is about $3\times10^{-13}$ , some ten million times smaller than the electron's, and nothing explains why it should be so minute. Second, $\nu_R$ is a gauge singlet, so once you have introduced it there is no symmetry that forbids also giving it a large Majorana mass — and if that term is allowed, it is unnatural to set it to zero. Including it is what produces the seesaw. So the Higgs can do the job, but doing only that leaves the smallness unexplained and ignores a term the theory practically demands.

What is the difference between a Dirac and a Majorana neutrino, physically?

A Dirac neutrino has a distinct antiparticle: the neutrino and antineutrino are genuinely different states, distinguished by a conserved lepton number, and the mass term links a left-handed neutrino to a separate right-handed one. A Majorana neutrino is its own antiparticle: there is no separate antineutrino, only the same particle in different helicity states, and lepton number is not conserved. The practical consequence is that a Majorana mass allows processes that turn a neutrino into an antineutrino — most importantly neutrinoless double-β decay. Because the neutrino is the only electrically neutral fundamental fermion, it is the only one that could be Majorana; all charged fermions must be Dirac, since a Majorana mass would violate electric charge conservation. Which one nature chose for the neutrino is still unknown and is one of the central open questions in particle physics.

Why is the top-left entry of the seesaw mass matrix exactly zero?

Because a Majorana mass for the left-handed neutrino is forbidden by the electroweak gauge symmetry. The left-handed neutrino sits inside the lepton doublet with hypercharge $-\tfrac12$ , so the bilinear $\overline{\nu_L^c}\nu_L$ transforms as a weak-isospin triplet with hypercharge $-1$ — it is not gauge-invariant, and you cannot write it down as a bare renormalisable term. That is the zero in the corner. The right-handed neutrino, by contrast, is a complete singlet, so its Majorana bilinear carries no gauge charge and the bottom-right entry $M$ is allowed and unconstrained. The off-diagonal $m_D$ is the ordinary Higgs-generated Dirac mass. This exact pattern — zero, $m_D$ , $M$ — is dictated entirely by gauge invariance, and it is what makes the light eigenvalue come out as the suppressed combination $m_D^2/M$ rather than some generic average.

Does the seesaw predict the value of the neutrino mass?

Not by itself. The relation $m_\nu\approx m_D^2/M$ has two unknown inputs — the Dirac mass $m_D$ and the heavy scale $M$ — and only their combination is fixed by the measured neutrino mass. So the seesaw explains why the mass is naturally small without a tiny coupling, but it does not pin down the absolute value; that requires assumptions about $m_D$ (often taken near the charged-fermion or GUT-scale masses) and $M$ . What the seesaw does predict, more usefully, is a scale: given a plausible $m_D$ between $1$ and $100$ GeV, the heavy neutrino must weigh $10^{11}$ to $10^{15}$ GeV — squarely in the range of grand unification. That is the mechanism's real content: it converts the smallness of the neutrino mass into a statement about very-high-energy physics, far beyond direct experimental reach.

Are there other kinds of seesaw besides the one with a heavy right-handed neutrino?

Yes. The version in this article — a heavy gauge-singlet fermion $N$ — is the type-I seesaw, the most economical. The type-II seesaw instead introduces a heavy scalar $SU(2)_L$ triplet with hypercharge $1$ , which couples directly to two lepton doublets and gives the left-handed neutrino a Majorana mass without any right-handed partner; the type-III uses a heavy fermion triplet instead of a singlet. All three produce the same low-energy dimension-five Weinberg operator and the same $m_\nu\sim v^2/\Lambda$ scaling — they differ only in what heavy particle is integrated out. This universality is why the Weinberg-operator viewpoint is so powerful: it captures every seesaw at once, and any experiment sensitive only to the light neutrinos cannot tell the three apart.

If neutrinos are Majorana, how would we ever know?

The cleanest way is neutrinoless double-β decay. In ordinary double-β decay a nucleus emits two electrons and two antineutrinos; if the neutrino is Majorana, the two antineutrinos can annihilate internally, leaving just two electrons. The tell-tale signature is a sharp spike in the summed electron energy right at the decay endpoint, where the neutrinoless mode deposits all the available energy. The rate is proportional to the square of the effective Majorana mass $m_{\beta\beta}=|\sum_i U_{ei}^2 m_i|$ , so a detection would both prove the Majorana nature and measure a combination of masses and phases. A deep result — the Schechter–Valle "black box" theorem — guarantees that any observation of $0\nu\beta\beta$ implies a Majorana mass, whatever the underlying mechanism. KamLAND-Zen and LEGEND are hunting for exactly this, so far without a signal.

Why does the seesaw connect to the matter–antimatter asymmetry?

Because the same heavy Majorana neutrinos that make the light neutrinos tiny can also explain why the universe is made of matter. A Majorana mass violates lepton number, and the heavy neutrinos' couplings can violate CP; if these states decayed out of thermal equilibrium in the early universe, they would have produced slightly more leptons than antileptons. Electroweak "sphaleron" processes then reprocess part of that lepton asymmetry into a baryon asymmetry. This scenario, leptogenesis, is one of the seesaw's most attractive features: a single mechanism accounts for the smallness of neutrino masses and for the existence of matter itself. It cannot be tested directly — the heavy neutrinos are far too massive — but confirming that neutrinos are Majorana (via $0\nu\beta\beta$ ) would strongly support it.