The Yukawa potential

The Yukawa idea is not a historical relic; it is the load-bearing template for how every force in the Standard Model is computed. In nuclear physics the long-range tail of the nucleon–nucleon interaction is still the one-pion-exchange Yukawa potential, and modern chiral effective field theory builds the full force as a controlled sum of pion and heavier-meson exchanges — one-boson-exchange potentials that fit scattering data to high precision, with QCD understood as the deeper layer beneath. In electroweak physics the massive propagators of the $W$ and $Z$ are Yukawa potentials of $\sim10^{-3}\ \mathrm{fm}$ range, and their low-energy contraction to a contact interaction is Fermi theory, with the mediator mass fixing $G_F$. The very same mathematics reappears far from particle physics: screened Coulomb (Debye–Hückel) potentials in plasmas and electrolytes, and Thomas–Fermi screening of impurities in metals, are Yukawa potentials $e^{-r/\lambda}/r$ with the screening length $\lambda$ playing the role of $1/m_X$. And every internal line in every Feynman diagram — the photon propagator, the gluon propagator, the Higgs propagator — is the momentum-space Yukawa object $1/(q^2-m^2)$. A separate modern usage borrows Yukawa's name for the Yukawa couplings between the Higgs field and fermions, which generate fermion masses after electroweak symmetry breaking; this is historically and conceptually distinct from Yukawa's meson-mediated nuclear-force mechanism.

The exchange picture lives in two frames at once. In the amplitude frame it is a sum over histories: the mediator $X$ can be launched by $a$ and caught by $b$, or launched by $b$ (as its antiparticle) and caught by $a$, and quantum mechanics adds these two time-ordered stories. Neither story alone is Lorentz-invariant, and in each the exchanged quantum flagrantly violates energy conservation; only their sum is covariant, and in the sum the two energy-denominators fuse into the single propagator $1/(q^2-m_X^2)$. In the potential frame the same propagator, read as scattering off a static source, is a screened Coulomb law whose screening length is the mediator's Compton wavelength. The first figure shows the two time orderings combining into one covariant propagator; the second shows how the mediator's mass sets the range of the force.

The exchanged quantum in each ordering carries a definite three-momentum but the "wrong" energy — that mismatch $E_j\neq E_i$ is exactly what makes it virtual, and it is why no detector ever catches an exchange particle in flight. What survives into the covariant amplitude is not the two illegal intermediate states but their sum, the propagator, whose denominator $q^2-m_X^2$ is a Lorentz scalar. When that amplitude is fed back into the Born formula for scattering off a static source, its Fourier transform is a potential — and the mediator's mass, buried in the denominator, resurfaces as an exponential screening factor. The second figure is the payoff: the Yukawa potential of a real pion beside the Coulomb potential it would become if the pion were massless.

Everything in this article can be run in either direction along one chain: field equation → potential, or amplitude → propagator → potential. We work in natural units $\hbar=c=1$, where mass, energy, momentum, and inverse length all share a single unit. Start with the field a static point source of a massive scalar sets up: it obeys the time-independent Klein–Gordon equation with a delta-function source of strength $g$.

where: $\phi$ — the static scalar field (potential per unit test charge); $m_X$ — the mediator mass; $g$ — the coupling (source strength) at the vertex; $\delta^3(\mathbf r)$ — a point source at the origin. Reading it: this is Poisson's equation $\nabla^2\phi=-g\delta^3$ with one extra term, $-m_X^2\phi$. That single mass term is the whole difference between an infinite-range and a short-range force. Solving equation (1) gives the Yukawa potential — the massive cousin of the Coulomb potential. Multiplying by the coupling $g$ of a second particle placed in the field gives the interaction energy between them.

where: $r=|\mathbf r|$ — separation; $g^2/4\pi$ — the dimensionless coupling strength (the scalar analogue of $\alpha$); the sign is negative because scalar exchange is attractive between like sources. Reading it: the familiar $1/r$ is multiplied by the screening factor $e^{-m_Xr}$, which crushes the potential to zero once $r$ exceeds $1/m_X$. The same physics arrives from the scattering side. Following $a+b\to c+d$ through both time orderings of the exchange and summing the amplitudes gives a manifestly covariant matrix element — the propagator sandwiched between two vertex couplings.

where: $g_a,g_b$ — the couplings at the two vertices; $q^\mu$ — the four-momentum transferred by $X$; $q^2=(E_a-E_c)^2-(\mathbf p_a-\mathbf p_c)^2$ — its invariant square. Reading it: $1/(q^2-m_X^2)$ is the propagator, the amplitude for the virtual $X$ to travel from one vertex to the other. In elastic scattering the transfer is spacelike, $q^2=-\mathbf q^2<0$, so $\mathcal{M}=-g_ag_b/(\mathbf q^2+m_X^2)$ never hits the pole. The bridge between equations (2) and (3) is the Born approximation: the static-limit amplitude is minus the Fourier transform of the potential, so inverting the transform of the propagator reproduces the Yukawa form exactly.

where: the integrand is the momentum-space propagator times the coupling $g^2$; the transform is the standard three-dimensional integral $\int\frac{d^3q}{(2\pi)^3}\frac{e^{i\mathbf q\cdot\mathbf r}}{\mathbf q^2+m^2}=\frac{e^{-mr}}{4\pi r}$. Reading it: momentum space and position space hold the same information — a pole at $\mathbf q^2=-m_X^2$ is an exponential of scale $1/m_X$ in $r$. Finally, the two numbers that summarize a Yukawa force: its range and its strength. The range is fixed by the mediator mass alone; the strength by the dimensionless coupling.

where: $R$ — the range (the reduced Compton wavelength of $X$); $\alpha_X$ — the coupling constant (for the photon this is the fine-structure constant $\alpha\approx1/137$). Reading it: mass and range are reciprocals, and the massless limit smoothly restores the pure Coulomb law — the Yukawa potential contains Coulomb as its $m_X\to0$ special case.

1 · Weighing the pion from the range of the nuclear force (fully worked)

The long-range part of the nuclear force between nucleons extends over roughly 1.4 fm and is dominated by one-pion exchange. Modelling this long-range interaction with a Yukawa potential $V\propto e^{-m_\pi r}/r$, (a) what mediator mass does that range imply? (b) Invert with the measured pion mass $m_\pi=139.6\ \mathrm{MeV}/c^2$ to predict the range. Use $\hbar c=197.3\,{\rm MeV}\cdot{\rm fm}$.

Step 1. From equation (5) the range and mediator mass are reciprocals, $R=\hbar c/(m_Xc^2)$, so a mass is a range read backwards: $m_Xc^2=\hbar c/R$. The mass term in the field equation is nothing but the inverse range of the force it carries.

Step 2. Insert the observed range $R\approx1.4\ \mathrm{fm}$: $m_Xc^2=\dfrac{197.3\ \mathrm{MeV\cdot fm}}{1.4\ \mathrm{fm}}=\mathbf{141\ MeV}$. This is essentially Yukawa's 1935 estimate — a particle a few hundred times heavier than the electron, intermediate in mass between electron and proton, which is why he called it a meson ("middle-weight").

Step 3. Invert with the modern pion mass to check consistency: $R=\dfrac{\hbar c}{m_\pi c^2}=\dfrac{197.3\ \mathrm{MeV\cdot fm}}{139.6\ \mathrm{MeV}}=\mathbf{1.41\ fm}$. The measured pion reproduces the nuclear range on the nose.

Step 4. Interpret. Yukawa read a particle's mass off a force's reach a full twelve years before that particle was seen (the pion was found in cosmic rays in 1947). A Yukawa interaction mediated by a particle of mass $m_\pi$ has characteristic range $\hbar/(m_\pi c)$; a 1.4 fm range therefore implies $m_\pi\approx140\,{\rm MeV}$, remarkably close to the measured pion mass. The Yukawa potential turned a qualitative force into a quantitative prediction of a new particle.

2 · From the W-boson propagator to Fermi's constant

The charged-current weak interaction is mediated by the $W$ boson, with mass $M_W=80.4\,{\rm GeV}$ and coupling $g_W\approx0.65$. (a) Estimate its range. (b) Show that at low momentum transfer the massive propagator reduces to a contact interaction, and recover Fermi's constant $G_F$.

Step 1. Range from equation (5): $R=\dfrac{\hbar c}{M_Wc^2}=\dfrac{197.3\ \mathrm{MeV\cdot fm}}{80\,400\ \mathrm{MeV}}=2.45\times10^{-3}\ \mathrm{fm}=\mathbf{2.5\times10^{-18}\ m}$. The weak force reaches roughly a thousandth of a proton radius — effectively a point contact, which is why $\beta$ decay looked for decades like a zero-range interaction.

Step 2. Take the low-$q$ limit of the propagator. For everyday weak processes $\mathbf q^2\ll M_W^2$, so from the limiting-cases table $\mathcal{M}=\dfrac{g_W^2}{\mathbf q^2+M_W^2}\to\dfrac{g_W^2}{M_W^2}$ — a constant, independent of $\mathbf q$. A momentum-independent amplitude is the Fourier transform of a delta function: the exchange has shrunk to a contact interaction. The heavy mediator hides the propagator's structure and mimics Fermi's original four-fermion vertex.

Step 3. Match to Fermi theory, whose contact coupling is $G_F/\sqrt2=g_W^2/(8M_W^2)$. Evaluate: $\dfrac{g_W^2}{8M_W^2}=\dfrac{(0.65)^2}{8\,(80.4\ \mathrm{GeV})^2}=\dfrac{0.4225}{8\times6464\ \mathrm{GeV^2}}=8.2\times10^{-6}\ \mathrm{GeV^{-2}}$, so $G_F=\sqrt2\times8.2\times10^{-6}=\mathbf{1.16\times10^{-5}\ GeV^{-2}}$.

Step 4. Compare with experiment: the measured Fermi constant is $G_F=1.166\times10^{-5}\ \mathrm{GeV^{-2}}$. The Yukawa propagator of an $80\ \mathrm{GeV}$ mediator reproduces Fermi's constant to three significant figures — the weak force is not intrinsically feeble, it is merely short-ranged, its apparent weakness at low energy being the $1/M_W^2$ suppression of a very heavy propagator.

The Yukawa potential is where a single equation predicted a single particle — a rare, clean instance of theory running ahead of experiment, and of a young physicist reading a new particle's mass straight off the reach of a force.

Texts. Thomson, Modern Particle Physics, §5.1–5.1.1 (time-ordered perturbation theory and the propagator) and §5.2 · Griffiths, Introduction to Elementary Particles, 2nd ed., Ch. 6 (the Feynman calculus) and §7.1 · Halzen and Martin, Quarks and Leptons, §4.5–4.6 (the propagator and the Yukawa potential) · Peskin and Schroeder, An Introduction to Quantum Field Theory, §4.7 (the Yukawa potential from the Born approximation) · Schwartz, Quantum Field Theory and the Standard Model, §3.4 · Aitchison and Hey, Gauge Theories in Particle Physics, Vol. 1, Ch. 5. The following key results tie the article together. The two time-ordered denominators recombine as $1/(\Delta-E_X)-1/(\Delta+E_X)=2E_X/(\Delta^2-E_X^2)$, with the $2E_X$ cancelling against the vertex normalisation to give $\mathcal{M}=g_ag_b/(q^2-m_X^2)$. The Fourier integral used throughout is $\int\frac{d^3q}{(2\pi)^3}\frac{e^{i\mathbf q\cdot\mathbf r}}{\mathbf q^2+m^2}=\frac{e^{-mr}}{4\pi r}$, and the field-equation identity $(\nabla^2-m_X^2)(e^{-m_Xr}/r)=-4\pi\delta^3(\mathbf r)$ confirms the static solution directly. Numerically: the pion range is $\hbar c/m_\pi c^2=197.3/139.6=1.41\ \mathrm{fm}$, and the weak-force contraction $G_F/\sqrt{2}=g_W^2/8M_W^2$ yields $G_F=1.166\times10^{-5}\ \mathrm{GeV^{-2}}$ for $M_W=80.4\ \mathrm{GeV}$ and $g_W=0.65$. The values $m_\pi=139.6\ \mathrm{MeV}$, $M_W=80.4\ \mathrm{GeV}$, and $\hbar c=197.3\ \mathrm{MeV\cdot fm}$ are standard tabulated quantities.

Prerequisites. Fermi's Golden Rule · Perturbation Theory and Particle Exchange · Natural Units · Relativistic Kinematics. Leads to. Feynman Diagrams and Virtual Particles · Introduction to QED · The Photon Propagator · The W-Boson Propagator · Rutherford and Mott Scattering.

Ask the tutor: Ask me to walk through how the same $1/(q^2-m^2)$ propagator gives Rutherford scattering for a massless photon and a contact four-fermion interaction for a heavy W — one formula, two famous force laws.