de Broglie Matter Waves

In 1924 a French doctoral student handed his examiners a thesis so audacious they mailed it to Einstein for a second opinion. Light, the committee already knew, had been caught behaving like a hail of particles — Planck's quanta, Einstein's photons, Compton's recoiling X-rays. Louis de Broglie asked the question nobody had dared to invert: if waves can act like particles, do particles secretly travel as waves? His answer fit on one line. Every object of momentum $p$ carries a wavelength $\lambda = h/p$ — electrons, atoms, buckyballs, you. The claim sounds untestable until you put numbers in. Planck's constant is so small that a thrown baseball's wavelength is $10^{-34}$ metres — invisible forever. But slow, light particles are different: an electron dawdling through a 54-volt gun has a wavelength of $0.167$ nanometres, almost exactly the spacing between atoms in a crystal. Nature had left a window open, and in 1927 Davisson and Germer looked through it: electrons bouncing off nickel piled up at special angles, exactly as interfering waves must. Matter diffracts. Here the de Broglie relation becomes a working tool: you will derive the electron-gun formula and learn when relativity breaks it, read a crystal as a diffraction grating, and watch Bohr's mysterious quantization rule turn into a resonance condition. Everything the Schrödinger equation will soon formalize starts from this one inversion.

Key takeaways

Every particle of momentum $p$ has a de Broglie wavelength $\lambda = h/p$ ; equivalently $p = \hbar k$ and $E = \hbar\omega$ — the same dictionary that already worked for photons, now read in reverse.
For an electron accelerated through $V$ volts: $\lambda \approx 1.226\ \mathrm{nm}/\sqrt{V}$ — atomic-scale wavelengths near 100 V, which is why crystals diffract electrons.
Davisson and Germer (1927) confirmed it: 54 eV electrons off nickel peak at $50^\circ$ , matching $\lambda = h/p$ to about 1%; G. P. Thomson saw diffraction rings through foils the same year.
Bohr's rule $L = n\hbar$ stops being mysterious: it is the condition that a whole number of wavelengths fit around the orbit — quantization as resonance.
The wave hides for macroscopic objects because $h$ is tiny: a baseball's $\lambda \sim 10^{-34}\ \mathrm{m}$ can never be resolved by any aperture.
Key idea: momentum sets wavelength — $\lambda = h/p$ — and matter's wave face shows wherever that wavelength reaches the size of the obstacles it meets.

The core insight

Attach to every particle a wave whose wavelength is inversely proportional to momentum, $\lambda = h/p$ , with frequency $\nu = E/h$ . The dual relations $p = \hbar k$ and $E = \hbar\omega$ unify the wave and particle descriptions — for light and matter alike.

Where this lives today

Every transmission electron microscope is equation (4) with lenses: at 300 kV the 1.97 pm wavelength makes single atoms routine imaging targets, and cryo-electron microscopy — the 2017 Chemistry Nobel — rides the same physics to frozen proteins. Neutron diffraction, born from the 0.18 nm thermal wavelength you computed above, is the standard probe of magnetic structure and of light atoms that X-rays barely see; dedicated spallation sources exist on three continents to manufacture nothing but slow neutrons. Helium-atom scattering uses the 0.05 nm thermal wave of a cold beam to map surfaces so gently it cannot dent even an adsorbed hydrogen layer. At the frontier the question is how big a wave can get: molecule interferometers have pushed interference beyond 25,000 atomic mass units, and atom interferometers measure gravity precisely enough to survey aquifers. Cool a gas until $\lambda_{\mathrm{th}}$ overlaps the interparticle spacing and it becomes one collective matter wave — the Bose–Einstein condensate, a stat-mech story whose entry ticket is the formula at the top of this page.

The physical picture

Start with what you already own from the photon story: light of wavelength $\lambda$ delivers momentum in lumps of $p = h/\lambda$ . De Broglie's move — the full inversion argument lives in the Wave–Particle Duality article — was to read that equation right-to-left as a rule for matter: given a momentum, nature assigns a wavelength. The heavier or faster the object, the shorter its wave. That single reciprocal relationship decides everything about where quantum behaviour shows up, so before any derivation, walk the ladder.

Figure

The matter-wave ladder — where the wave face becomes visible

Read the ladder as a design chart. Diffraction happens when the wavelength is comparable to the structure it meets, and the green band marks atomic spacings in solids. Electrons at tens of volts, thermal neutrons, and cold helium atoms all land inside the band — each is now a workhorse diffraction probe. Push the electron to 300 kV and its wavelength drops a hundredfold below atomic spacing — exactly what a microscope wants. The baseball's wavelength is far too small to resolve in ordinary experiments, which is why its wave nature is not apparent in everyday classical physics. De Broglie's own favourite payoff was hydrogen. Bohr had decreed, without explanation, that only orbits with $L = n\hbar$ exist. Wrap a matter wave around the orbit and require it to return in phase with itself: constructive interference occurs only when the circumference contains an integer number of wavelengths, $2\pi r=n\lambda$ .

Figure

Only whole numbers of wavelengths survive on a closed orbit

Hold both pictures — ladder and loop — and the rest is bookkeeping: converting lab knobs into wavelengths, and checking those wavelengths against experiment.

The dictionary, read as matter's own

The de Broglie relation is one line, but read each symbol with care:

\lambda_{\mathrm{dB}} = \frac{h}{p} = \frac{h}{mv} \qquad (1)

where: $\lambda_{\mathrm{dB}}$ — the de Broglie wavelength; $h = 6.626\times10^{-34}\ \mathrm{J\,s}$ — Planck's constant; $p$ — the particle's momentum ( $mv$ non-relativistically). The reciprocal is the whole story: momentum up, wavelength down. Note which constant appears — the same $h$ that ran the photon's energy budget now runs matter's geometry. One constant, one dictionary, both directions. Alongside the wavelength comes a frequency, and together they are usually packaged in wavevector form. Since $k = 2\pi/\lambda$ and $\omega = 2\pi\nu$ , substituting equation (1) and $\nu = E/h$ gives

p = \hbar k, \qquad E = \hbar\omega \qquad (2)

where: $k$ — wavenumber; $\omega$ — angular frequency; $\hbar = h/2\pi = 1.055\times10^{-34}\ \mathrm{J\,s}$ . These two lines are the de Broglie–Einstein relations, and they are the form the Schrödinger equation will inherit: the simplest matter wave is the plane wave $\psi \propto e^{i(kx - \omega t)}$ , whose $k$ and $\omega$ carry the particle's momentum and energy. What such a wave means — and why it must be complex — is the business of The Wavefunction, three articles ahead. One honest caution before the toolkit. The matter wave's phase velocity $v_p = \omega/k$ is not the particle's speed; the particle rides in a wave packet travelling at the group velocity $v_g = d\omega/dk$ , which does equal the classical velocity. The full accounting belongs to Free Particle and Wave Packets and to Group and Phase Velocity, later in this chapter; here you only need the moral: don't read $\omega/k$ as the electron's speedometer. Which form of the wavelength you use depends on which physical quantity is given:

you are given	use	typical regime
momentum $p$	$\lambda = h/p$	always true — the definition
kinetic energy $K$ (slow)	$\lambda = h/\sqrt{2mK}$	$K \ll mc^2$
accelerating voltage $V$ (electron)	$\lambda \approx 1.226\ \mathrm{nm}/\sqrt{V}$	electron guns up to a few kV
kinetic energy $K$ (any speed)	$\lambda = hc/\sqrt{K^2 + 2K mc^2}$	electron microscopes, 100 kV and up
temperature $T$ (gas of mass $m$ )	$\lambda_{\mathrm{th}} = h/\sqrt{2\pi m k_B T}$	cold-atom and thermal-beam physics

Derivation AThe electron-gun formula and when it breaks

The most common conversion in this subject turns an accelerating voltage into a wavelength, because voltage is what a lab dial actually controls. We derive the non-relativistic formula, package it with its famous numerical coefficient, then find out where it quietly fails — at microscope voltages the electron is fast enough that ignoring relativity overstates the wavelength by double-digit percentages. The strategy is energy bookkeeping: voltage buys kinetic energy, kinetic energy buys momentum, momentum sets wavelength.

Step 1. Voltage to kinetic energy. An electron falling through potential difference $V$ gains $K = eV$ . Why: the work–energy theorem for charge $e$ across voltage $V$ ; the gun geometry is irrelevant, only the total drop matters.

Step 2. Kinetic energy to momentum. Non-relativistically $K=\frac12mv^2$ , so for an electron accelerated through a voltage $V$ , $eV=\frac{p^2}{2m_e}$ and therefore $p=\sqrt{2m_e eV}$ . This is valid while $eV\ll m_ec^2=511\,\mathrm{keV}$ .

Step 3. Momentum to wavelength. Insert into equation (1):

\lambda = \frac{h}{\sqrt{2m_e eV}} \approx \frac{1.226\ \mathrm{nm}}{\sqrt{V/\mathrm{volt}}} \qquad (3)

Why: the packaged coefficient $h/\sqrt{2m_e e} = 1.226\ \mathrm{nm\,V^{1/2}}$ is worth memorizing — at $V = 150\ \mathrm{V}$ it gives $\lambda = 0.100\ \mathrm{nm}$ , dead on the atomic scale, which is why electron diffraction is a table-top experiment.

Step 4. Find the breaking point. At high voltage use the relativistic energy–momentum relation $E^2 = (pc)^2 + (m_ec^2)^2$ with $E = K + m_ec^2$ ; solving for $pc$ gives $pc = \sqrt{K^2 + 2K m_ec^2}$ and

\lambda = \frac{hc}{\sqrt{K^2 + 2K m_e c^2}} \qquad (4)

Why: equation (4) contains both worlds — for $K \ll m_ec^2$ the first term under the root is negligible and (3) reappears; for $K \gg m_ec^2$ it collapses to the photon-like $\lambda = hc/K$ .

Step 5. Put numbers on the failure. At $V = 300\ \mathrm{kV}$ , equation (3) predicts $2.24\ \mathrm{pm}$ while the honest (4) gives $1.97\ \mathrm{pm}$ — the non-relativistic formula overshoots by about 14%. Why: $eV$ is now 59% of $m_ec^2$ ; a microscope designer who used (3) would mis-focus every lens in the column. *Note: the coefficient $h/\sqrt{2m_e e} = 1.2264\ \mathrm{nm\,V^{1/2}}$ follows directly from CODATA values of $h$ , $m_e$ , and $e$ ; the 150 V and 300 kV results quoted above, and both limiting forms of equation (4), are consistent with those constants.*

The experiment that settled it

Davisson and Germer were not initially looking for matter waves. At Bell Labs, a 1925 accident involving a liquid-air system damaged the polycrystalline nickel target; subsequent heating produced larger crystalline regions, making the electron-scattering pattern much sharper. The scattered electrons promptly stopped behaving: instead of a smooth angular smear, sharp peaks appeared at special angles — the target had become a diffraction grating, and the "particles" were diffracting. By 1927 the numbers were precise: at 54 V, a strong peak at $50^\circ$ .

Figure

Davisson–Germer — surface atom rows act as a grating for the electron wave

Go deeper · Derivation B · 1 minReading the Davisson and Germer peak

Derivation BReading the Davisson and Germer peak

We predict where a beam of electrons of known voltage must pile up when scattered off a crystal surface, treating the top rows of atoms as a reflection grating. The route: convert the voltage to a wavelength with equation (3), write the constructive-interference condition for adjacent surface rows, and confront the 1927 numbers. Watch how few inputs there are — one voltage, one atomic spacing — and how little room the prediction leaves to hide.

Step 1. Wavelength from the gun. At $V = 54\ \mathrm{V}$ , equation (3) gives $\lambda = 1.226/\sqrt{54} \approx 0.167\ \mathrm{nm}$ . Why: at this voltage the non-relativistic formula differs negligibly from the relativistic result.

Step 2. Grating condition. Adjacent surface rows, spaced by $D$ , scatter constructively toward an angle $\phi$ when the path difference between waves from neighboring rows is an integer multiple of the wavelength:

n\lambda = D\sin\phi, \qquad n = 1, 2, \dots \qquad (5)

Why: this is the same geometry as an optical reflection grating — the wave nature is assumed, and the experiment will vote on it. For nickel's surface net, $D = 0.215\ \mathrm{nm}$ , known independently from X-ray diffraction.

Step 3. Confront experiment. For first order ( $n = 1$ ) at the observed $\phi = 50^\circ$ , $D\sin 50^\circ = 0.215 \times 0.766 \approx 0.165\ \mathrm{nm}$ , compared with the de Broglie value $\lambda_{\mathrm{dB}} \approx 0.167\ \mathrm{nm}$ — a difference of about $1.3\%$ . Why: two independent routes to one number — a power supply plus $h$ , versus crystal geometry plus a protractor. Their meeting is the discovery.

Step 4. Make it predictive. Raise the voltage and the peak must march inward: at $V = 100\ \mathrm{V}$ , $\lambda = 0.123\ \mathrm{nm}$ and $\sin\phi = 0.123/0.215 = 0.570$ , so $\phi = 34.8^\circ$ . Why: a single peak could be luck; the full voltage sweep tracing $\sin\phi \propto 1/\sqrt{V}$ is what nailed the law. *The four numbers above follow from fundamental constants and the 1927 experimental data: $\lambda(54\ \mathrm{V}) = 0.16690\ \mathrm{nm}$ and $D\sin50^\circ = 0.16470\ \mathrm{nm}$ , a ratio agreement of 1.32%.*

Within months George Paget Thomson fired faster electrons through thin metal foils and photographed concentric diffraction rings; he and Davisson shared the 1937 Nobel Prize. The family irony is an examiners' favourite: J. J. Thomson won a Nobel for proving the electron a particle; his son won one for proving it a wave. Both were right — which is the point of this chapter.

Go deeper · Derivation C · 1 minBohr's quantization from a standing matter wave

Derivation CBohr's quantization from a standing matter wave

Bohr's 1913 model postulated the angular-momentum condition $L=n\hbar$ as part of its quantization rules, without the wave-based justification supplied later by de Broglie's matter-wave picture. De Broglie's thesis supplied the missing why in three lines: an electron wave living on a circular orbit must close on itself, and the closure condition is Bohr's rule verbatim. This was the argument that convinced Einstein the thesis was serious physics, and it is the conceptual bridge from "quantization as decree" to "quantization as boundary condition" — the theme the Infinite Square Well will industrialize.

Step 1. State the closure condition. A wave on a loop of radius $r$ reinforces itself only if the circumference holds a whole number of wavelengths: $2\pi r = n\lambda$ . Why: after one lap the wave must return in phase with itself. A half-integer mismatch makes the wave change sign after one circuit, so it is incompatible with the required single-valued standing-wave condition.

Step 2. Insert the de Broglie relation. With $\lambda = h/p$ : $2\pi r = nh/p$ . Why: the wavelength is not a free parameter — the electron's own momentum fixes it.

Step 3. Rearrange into angular momentum. For a circular orbit $L = rp$ , so

L = rp = \frac{nh}{2\pi} = n\hbar \qquad (6)

Why: Bohr's rule drops out exactly — not approximately — with $n$ counting the standing wave's crests around the loop. *It is straightforward to confirm symbolically that substituting $\lambda = h/p$ into $2\pi r = n\lambda$ forces $rp = n\hbar$ exactly. The Bohr model itself — energies, radii, and its limits — is cross-track material (atomic physics); here it serves as the resonance exhibit.*

Limiting cases

Case	Result	Meaning
macroscopic object	$\lambda = h/p \to 10^{-34}\ \mathrm{m}$ scale	no aperture can resolve it — classical mechanics emerges intact
slow particle, $K \ll mc^2$	$\lambda = h/\sqrt{2mK}$	the working formula of electron guns and thermal beams
ultra-relativistic, $K \gg mc^2$	$\lambda \to hc/K$	the photon relation reappears — the dictionary is one law, not two
$\lambda \sim$ atomic spacing	strong diffraction	electrons at 30–150 V, thermal neutrons, cold atoms: the probe regime
thermal gas at temperature $T$	$\lambda_{\mathrm{th}} = h/\sqrt{2\pi m k_B T}$ grows as $T$ falls	when $\lambda_{\mathrm{th}}$ reaches the interparticle spacing, quantum statistics take over (stat-mech territory)

Common trap“The matter wave travels along with the electron like a wake in a medium”

Three interpretations to avoid. The de Broglie wave is not a classical ripple in a material medium, and it is not a physical wake dragged behind the particle. It is a quantum wave description associated with the particle, replacing a definite classical trajectory with a wavefunction. And its phase velocity $\omega/k$ — $v/2$ non-relativistically, faster than light relativistically — is not the particle's speed and carries no signal; the particle travels at its packet's group velocity, which equals the classical $v$ . What oscillates is a complex amplitude whose meaning — probability — arrives with the Born rule two articles ahead. Treat $\lambda = h/p$ as exact bookkeeping and postpone the metaphysics.

Worked examples

1 · Predicting the next peak (confidence builder)

Davisson and Germer raise their gun voltage from 54 V to 100 V. Where should the first-order peak move? (Nickel surface rows: $D = 0.215\ \mathrm{nm}$ .)

Step 1. New wavelength from the gun formula (3): $\lambda = \dfrac{1.226}{\sqrt{100}} = 0.123\ \mathrm{nm}$ .

Step 2. Grating condition (5) at first order: $\sin\phi = \dfrac{\lambda}{D} = \dfrac{0.123}{0.215} = 0.570$ .

Step 3. Peak angle: $\phi = \arcsin(0.570) = \boxed{34.8^\circ}$ — the peak marches toward the incident beam as the wavelength shrinks.

Step 4. Sanity-check the trend: $\sin\phi \propto \lambda \propto 1/\sqrt{V}$ , so quadrupling the voltage from 54 V would halve $\sin\phi$ . A whole voltage sweep traces this curve — the signature that convinced the skeptics in 1927.

2 · The microscope column (exam level)

A transmission electron microscope runs at $300\ \mathrm{kV}$ . Find the electron wavelength (a) with the non-relativistic formula, (b) honestly, and (c) compare both with a thermal neutron at $k_BT = 0.025\ \mathrm{eV}$ — which probe "sees" finer detail?

Step 1. Non-relativistic attempt (3): $\lambda = \dfrac{1.226\ \mathrm{nm}}{\sqrt{3\times10^5}} = \dfrac{1.226}{547.7} = 2.24\ \mathrm{pm}$ .

Step 2. Relativistic check first: $eV=300\,\mathrm{keV}$ compared with $m_ec^2=511\,\mathrm{keV}$ , so the kinetic energy is about $59\%$ of the rest energy. The nonrelativistic formula (3) is no longer sufficiently accurate; use the relativistic expression (4).

Step 3. Honest wavelength from (4): $K^2 + 2Km_ec^2 = (300)^2 + 2(300)(511) = 9.00\times10^4 + 3.07\times10^5 = 3.97\times10^5\ \mathrm{keV^2}$ , so $pc = \sqrt{3.97\times10^5}\ \mathrm{keV} = 630\ \mathrm{keV}$ and $\lambda = \dfrac{hc}{pc} = \dfrac{1.240\ \mathrm{keV\,nm}}{630\ \mathrm{keV}} = \boxed{1.97\ \mathrm{pm}}$ . The shortcut overshoots by 14%.

Step 4. Thermal neutron: $\lambda = \dfrac{h}{\sqrt{2m_nK}}$ with $K = 0.025\ \mathrm{eV}$ gives $\lambda = 0.181\ \mathrm{nm} = 181\ \mathrm{pm}$ .

Step 5. Compare: the microscope electron's wavelength is about ninety times shorter than the neutron's, which illustrates why electrons can achieve much finer spatial resolution. In a TEM, however, the practical resolution is also limited by lens aberrations and other instrumental effects. Neutrons play a different role: a wavelength of about $0.18\,\mathrm{nm}$ is comparable to typical crystal spacings, making neutron diffraction useful for probing bulk structure and magnetic order. The common currency is $\lambda=h/p$ .

Origins · 1 minLouis de Broglie and how this was discovered

Louis de Broglie

1892–1987 · Paris

A prince of a family that expected diplomacy, de Broglie came to physics through his brother Maurice's X-ray laboratory and through reading wartime radio-engineering reports. His 1924 doctoral thesis proposed matter waves on symmetry grounds alone — no experiment, no wave equation, just the conviction that nature would not give light two faces and matter one. His examiners were unsure whether it was profound or absurd; Paul Langevin sent a copy to Einstein, who replied that de Broglie had "lifted a corner of the great veil." Three years later the Davisson–Germer peaks printed his wavelength in nickel, and in 1929 he became the first physicist to win the Nobel Prize for a doctoral thesis. Schrödinger's wave mechanics began, by Schrödinger's own account, as an attempt to answer the question a colleague fired at him after a seminar on de Broglie's waves: if electrons are waves, what is their wave equation?

History

1923

— Compton's X-ray scattering forces physicists to take photon momentum

p = h/\lambda

literally.

1924

— De Broglie's thesis inverts the relation: every particle of momentum

p

carries

\lambda = h/p

; Einstein endorses it.

1925

— A burst liquid-air bottle at Bell Labs oxidizes Davisson's nickel target; the recrystallizing anneal accidentally builds the diffraction grating.

1927

— Davisson & Germer measure the 54 V peak at 50°; G. P. Thomson photographs diffraction rings through foils.

1929

— De Broglie receives the Nobel Prize; Davisson and Thomson share theirs in 1937.

1999

— Arndt's group in Vienna diffracts C₆₀ fullerenes — sixty-atom molecules — through a grating, wavelength 2.5 pm.

Going further

Texts. Griffiths, Introduction to Quantum Mechanics, 3rd ed., Ch. 1 · Sakurai & Napolitano, Modern Quantum Mechanics, 2nd ed., Ch. 1 · Shankar, Principles of Quantum Mechanics, 2nd ed., Ch. 1 · Cohen-Tannoudji, Diu & Laloë, Vol. 1, Ch. 1 · Townsend, A Modern Approach to Quantum Mechanics, Ch. 2 · Eisberg & Resnick, Quantum Physics, Ch. 3. Key numerical anchors: $h=6.626\times10^{-34}\,\mathrm{J\,s}$ ; electron wavelength at $150\,\mathrm{V}\approx1.0\,\text{Å}$ , at $54\,\mathrm{V}\approx1.67\,\text{Å}$ , at $100\,\mathrm{V}\approx1.23\,\text{Å}$ , all from $\lambda=h/\sqrt{2m_e eV}$ . Davisson–Germer nickel spacing $2.15\,\text{Å}$ and $50^\circ$ peak are from their 1927 paper. The group and phase velocity relations $v_g v_p=c^2$ and $v_g=v$ are standard results.

Prerequisites. Breakdown of classical physics · Photoelectric effect and the photon · Blackbody radiation and the Planck quantum. Leads to. Wave–particle duality · Double-slit experiment · The wavefunction · Time-dependent Schrödinger equation.

If everything has a wavelength, why has nobody ever seen a person diffract through a doorway?

Put numbers on it. Walking through a door at 1 m/s with mass 70 kg, your wavelength is $h/p \approx 10^{-35}$ m — about $10^{34}$ times smaller than the doorway. Diffraction angles scale as $\lambda/d$ , so the bending is one part in $10^{34}$ : no measurement in the lifetime of the universe resolves that. The wave face never switches off; it just scales below all visibility. This is also why the classical limit works — see the ladder figure: macroscopic momenta bury $\lambda$ twenty decades under the smallest structures that exist.

Is the de Broglie wave a real physical wave or just a calculating trick?

It is real enough to steer electrons into nickel-crystal peaks and buckyballs into interference fringes — patterns you can photograph. What it is not is a mechanical vibration of a medium, or a smearing-out of the electron's charge. The modern answer: the wave is the wavefunction, a complex amplitude whose modulus squared gives detection probabilities — Born's reading, two articles ahead. De Broglie himself spent decades on a more literal "pilot wave" picture, which survives as Bohmian mechanics.

Does a composite object like a C₆₀ molecule have sixty wavelengths or one?

One — set by the total momentum of the centre of mass: $\lambda = h/Mv$ with $M$ the full molecular mass. Internal structure rides along coherently as long as nothing measures it; in the 1999 Vienna experiment the molecules were even internally hot, glowing faintly, yet each C₆₀ interfered as a single 2.5 pm wave. The caveat: if the internal state records path information (say, by emitting a photon that could reveal the route), coherence degrades — the theme the Double-Slit Experiment article picks up.

Which energy do I use for the thermal de Broglie wavelength — kT, 3kT/2, or something else?

Convention, and you must declare it. Setting $K = k_BT$ gives $\lambda = h/\sqrt{2mk_BT}$ (1.78 Å for a 300 K neutron); the most-probable-speed and mean-energy choices differ by factors of order one. Statistical mechanics standardizes on $\lambda_{\mathrm{th}} = h/\sqrt{2\pi m k_BT}$ (1.00 Å for the same neutron) because that combination makes partition-function formulas clean. For estimates, any of them tells you the same physics — an ångström-scale wave for thermal neutrons; for exams, quote which definition you are using.

Why does λ = h/p use momentum rather than energy or speed?

Because momentum is what the wave's geometry encodes: $p = \hbar k$ says momentum is (up to $\hbar$ ) the spatial phase gradient — how fast the phase winds per metre. Energy pairs with time instead, $E = \hbar\omega$ . Two different-mass particles with equal momenta have identical wavelengths and diffract identically off the same crystal; equal energies would not. Relativity seals the argument: $(E, \mathbf{p})$ and $(\omega, \mathbf{k})$ are both four-vectors, so $p = \hbar k$ and $E = \hbar\omega$ are one covariant statement.

If the phase velocity exceeds c, why doesn't that violate relativity?

A single plane wave has no beginning, no end, and no marker on it — an infinite sinusoid conveys nothing, so its crest speed can be anything without carrying information. An actual particle is a wave packet whose envelope — the thing that can carry a signal — moves at the group velocity $v_g = d\omega/dk$ , which equals the classical velocity and never exceeds $c$ . (For relativistic matter waves, neatly, $v_p v_g = c^2$ .) Full derivation in Group and Phase Velocity, later in this chapter.

de Broglie Matter Waves

Where this lives today

The physical picture

The dictionary, read as matter's own

The experiment that settled it

Limiting cases

Worked examples

1 · Predicting the next peak (confidence builder)

2 · The microscope column (exam level)

Going further

Before this

Where this leads