Energy-Time Uncertainty

Equivalent forms

\Delta E \approx \hbar/\tau

Why spectral lines have width, why unstable particles have a mass spread, and why the vacuum can flicker — all one inequality.

Symbol	Name	SI unit	Dimension	Range
$\Delta E$	Energy uncertaintyoutput Spread/width in energy	J	ML^2T^-2	—
$\Delta t$	Time / lifetime Characteristic timescale or state lifetime	s	T	1e-9 – 0.000001
$\hbar$	Reduced Planck constant h/2π, the quantum of action	J*s	ML^2T^-1	—

Where it holds

\Delta t

is a timescale of evolution or a state lifetime, not a measured 'time observable.' The rigorous Mandelstam–Tamm form bounds how fast any observable can change.

Dimensional analysis

\Delta E \Delta t \to [M L^2\,\mathrm{T}^-2][T] = [M*L^2*T^-1]

\hbar /2 \to [M*L^2*T^-1]

Energy

\times

time has the dimensions of action, matching

\hbar

.

\checkmark

Discovery

Werner Heisenberg · 1927

Heisenberg stated the energy-time relation alongside position-momentum in his 1927 uncertainty paper. Its interpretation was subtle — since time is a parameter, not an operator — and was clarified by Mandelstam and Tamm (1945), who linked Δt to how fast an observable actually changes.

Research status: stable

Real-world applications

Natural linewidths in spectroscopy and lasers
Particle 'widths' as inverse lifetimes at colliders
Virtual particles and the range of forces

Common misconceptions

Time is not an operator; this is not a symmetric partner of [x̂,p̂]
It does not mean energy conservation is violated in measurements — only that short-lived states have energy spread
$\Delta t$ is a lifetime/evolution time, not measurement duration in general

Experimental verification

The natural linewidth of spectral lines

(\Gamma \tau \approx \hbar )

, and the mass-width of unstable particles (e.g. the Z boson'

s \sim 2.5\,\mathrm{GeV}

width from

its \sim 10^{-25} s

lifetime) confirm

\Delta E \approx \hbar /\tau

precisely.

Derivation

Mandelstam–Tamm: for any observable A,

\Delta E\cdot \Delta A \ge (\hbar /2)|d

⟨A⟩/dt|.

Define

\Delta t = \Delta A/|d

⟨A⟩/dt| as the time for ⟨A⟩ to change by one standard deviation.

Substituting gives

\Delta E\cdot \Delta t \ge \hbar /2

.

For an exponentially decaying state with lifetime

\tau

, the Fourier transform of

e^{-t/2\tau }

has an energy width (linewidth

) \Gamma = \hbar /\tau

.

What if…

What if the state is perfectly stable

(\tau \to \infty )

?

$\Delta E\to 0$ : a stationary state has an exactly sharp energy and an infinitely narrow spectral line.

What if a particle lives only

10^{-23} s

?

Its energy (mass) is uncertain $by \sim MeV$ –GeV — this is literally the 'width' quoted for resonances like $the \Delta$ baryon.

1

Linewidth of a 10 ns state

Given ·

τ:: 1e-8 s

Find · energy width

\Delta E

Steps

$\Delta E \approx \hbar /\tau = 1.0546e-34 / 1e-8$
$= 1.05e-26\,\mathrm{J}$
In frequency, $\Delta E/h \approx 16\,\mathrm{MHz}$

Result ·

\Delta E \approx \hbar /\tau \approx 1.05e-26\,\mathrm{J} \approx 66\,\mathrm{neV}

Heisenberg Uncertainty Principle→Fermi's Golden Rule→natural linewidth