Planck Units and Invariant Expansion Parameters

Planck units are the combinations of $G$, $\hbar$, $c$, and $k_{\mathrm B}$ that carry dimensions of length, time, mass, energy, temperature, area, curvature, and density. Their real use is not to label everything “Planckian,” but to turn a process or geometry into an invariant, dimensionless expansion parameter whose smallness can be checked.

Required background. Einstein Equations fixes the dimensions and normalization of $G$ in the gravitational field equation; Quantum Mechanics supplies $\hbar$ and the relation between energy, frequency, and wavelength.

Helpful background. Flat-Spacetime Quantum Field Theory explains loop and momentum expansions; Quantum Information helps distinguish an operationally specified measurement from a dimensional estimate.

Planck combinations and mass conventions

Restoring constants temporarily, dimensional analysis gives

$$\ell_{\mathrm P}=\sqrt{\frac{\hbar G}{c^3}},\qquad t_{\mathrm P}=\sqrt{\frac{\hbar G}{c^5}},\qquad M_{\mathrm P}=\sqrt{\frac{\hbar c}{G}}.$$

The first relation follows by seeking a monomial $G^a\hbar^bc^d$ with the dimensions of length. Matching independent powers of mass, length, and time gives $a=b=1/2$ and $d=-3/2$. Repeating the matching for time and mass yields the other two combinations. No dynamical equation enters this construction: Planck units are consequences of dimensional bookkeeping, not predictions that an observable becomes singular or discrete at precisely one of these values.

The corresponding energy and temperature are $E_{\mathrm P}=M_{\mathrm P}c^2$ and $T_{\mathrm P}=E_{\mathrm P}/k_{\mathrm B}$. The area, curvature, mass-density, and energy-density scales may be chosen as

$$A_{\mathrm P}=\ell_{\mathrm P}^2,\qquad \mathcal R_{\mathrm P}=\ell_{\mathrm P}^{-2},\qquad \rho_{\mathrm P}=\frac{M_{\mathrm P}}{\ell_{\mathrm P}^3},\qquad \varepsilon_{\mathrm P}=\rho_{\mathrm P}c^2.$$

In the site convention $c=\hbar=k_{\mathrm B}=1$, these reduce to $\ell_{\mathrm P}=t_{\mathrm P}=\sqrt G$, $M_{\mathrm P}=E_{\mathrm P}=T_{\mathrm P}=G^{-1/2}$, $\mathcal R_{\mathrm P}=G^{-1}$, and $\varepsilon_{\mathrm P}=G^{-2}$. This is the unreduced Planck mass. The reduced mass is

$$\overline M_{\mathrm P}=\frac{1}{\sqrt{8\pi G}}, \qquad \kappa=\sqrt{32\pi G}=\frac{2}{\overline M_{\mathrm P}}.$$

Dropping the bar or the factor $8\pi$ silently changes every power-counting coefficient. The rest of this volume uses $G$ and $\kappa$ explicitly.

It is also important to name the density. The Planck mass density and energy density differ by $c^2$ when constants are restored. Likewise, “Planck curvature” can mean a scalar with dimensions of inverse length squared or the square root of a quadratic invariant. Stating the invariant avoids equating a coordinate component, which may be made large by changing charts, with physical tidal curvature.

From scales to invariant control parameters

Suppose a local scattering experiment has Mandelstam invariants $s,t,u$. Its gravitational quantum expansion is organized by ratios such as

$$\epsilon_s=G\lvert s\rvert, \qquad \epsilon_t=G\lvert t\rvert,$$

with the usual loop factor and process-dependent logarithms added in an actual amplitude. For characteristic center-of-mass energy $E$, one often abbreviates this as $\epsilon_E=GE^2=(E/M_{\mathrm P})^2$. This statement is useful only after $E$ has been defined invariantly; a coordinate energy in a time-dependent or strongly curved region is not enough.

A large dimensionful scale alone is not by itself a breakdown criterion. For example, the total energy of a spatially extended coherent state may exceed $M_{\mathrm P}$ while every momentum transfer and local curvature invariant remains small. Conversely, a low-energy process can become sensitive to gravity through a long interaction time, a very large occupation number, an infrared enhancement, or a near-degenerate energy denominator. The correct procedure is to identify all ratios that can enhance the particular observable and then bound the neglected terms.

A curved background needs geometric curvature diagnostics. In a region where the Ricci scalar is representative, $G\lvert R\rvert$ is dimensionless. In Ricci-flat geometry it vanishes even when tidal curvature does not. For backgrounds such as Schwarzschild, where the Kretschmann scalar $K$ is positive and representative, a useful diagnostic is

$$\epsilon_K=G\sqrt{\lvert K\rvert}, \qquad K=R_{\mu\nu\rho\sigma}R^{\mu\nu\rho\sigma}.$$

This is not a universal curvature norm. Lorentzian contractions need not be positive, and some nonflat plane waves have vanishing scalar curvature invariants. In such cases use tidal components in a physically specified orthonormal frame, together with the probe’s energy and wavelength. A zero scalar cannot establish that every experiment sees weak curvature.

Derivative expansions additionally compare wavelengths with the curvature radius, while classical nonlinear dynamics can be controlled by ratios such as $GM/b$. These are distinct axes: $GM/b$ may be order unity for a macroscopic black hole even though quantum curvature corrections remain tiny.

One should also distinguish a local expansion from an integrated error. If a correction to a phase accumulates over $N$ nearly identical cycles, a local parameter $\epsilon\ll1$ can produce an order-$N\epsilon$ effect. That observation does not make the microscopic theory strongly coupled; it says that the observable demands a resummation or a longer-time error estimate. This distinction will matter again in waveform, cosmological, and interferometric applications.

Donoghue 2023, §§4, 6 develops the low-energy graviton expansion and its kinematic and curvature limits.

A discriminating black-hole check

For Schwarzschild geometry,

$$K\equiv R_{\mu\nu\rho\sigma}R^{\mu\nu\rho\sigma} =\frac{48G^2M^2}{r^6}.$$

At the horizon $r=2GM$,

$$\epsilon_K =G\sqrt K =\frac{\sqrt3}{2}\left(\frac{M_{\mathrm P}}{M}\right)^2.$$

Thus a large black hole has a strong classical potential at its horizon but parametrically small curvature in Planck units. This independent limit checks both dimensions and physical interpretation: “horizon” does not mean “microscopic quantum gravity.” Near a singular classical solution, by contrast, $\epsilon_K$ can grow without bound, but that shows failure of a curvature expansion, not which microscopic theory replaces it.

The same calculation supplies a useful scaling test. At fixed $r/(GM)$, the Kretschmann scalar scales as $M^{-4}$ and its square root as $M^{-2}$. Increasing the mass therefore makes the horizon less curved, even though the horizon radius and total energy increase. Any proposed criterion that predicts the opposite solely because $M/M_{\mathrm P}$ is large has confused an extensive scale with the invariant local expansion parameter.

The page does not infer string, loop, holographic, or alternative-gravity dynamics from these ratios. Those constructions require their own degrees of freedom, observables, and correspondence limits; semiclassical thermodynamics and generic flat-spacetime field theory remain in their separate treatments.

Exercises

A Schwarzschild black hole has mass $M=10^6M_{\mathrm P}$. Estimate $\epsilon_K$ at its horizon. Does the result justify treating the horizon as a Planck-curvature region?

Solution

Substitution gives $\epsilon_K=(\sqrt3/2)10^{-12}\approx8.7\times10^{-13}$. The invariant tidal curvature is therefore extremely small in Planck units even though $GM/r=1/2$ at the horizon. Classical gravity is nonlinear there, but local quantum-curvature corrections are not forced to be large. One would need a different, explicitly defined observable or accumulation parameter to argue otherwise.

References

• John F. Donoghue, “Quantum General Relativity and Effective Field Theory,” arXiv:2211.09902v2 (2023), Open manuscript.