The photoelectric effect gets most of the attention in the "light behaves like particles" story, but the Compton effect is arguably the more direct proof. It doesn't just show light carrying energy in discrete packets — it shows those packets carrying actual, measurable momentum, behaving in a collision exactly the way a particle would.

A photon scattering off an electron, losing energy to it

Not absorption — a genuine collision

In the photoelectric effect, a photon gets fully absorbed, handing over all its energy at once, and effectively ceases to exist. Compton scattering is a different animal: the photon bounces off an electron, keeps existing afterward, but comes away with less energy and a longer wavelength, having handed some momentum to the electron in the process. Genuinely a scattering event, not an absorption event — more like two billiard balls colliding than like a photon being swallowed.

The shift depends only on angle — nothing else

Δλ = (h/mₑc)(1 − cosθ)

Work through the relativistic momentum and energy conservation equations for this collision and something genuinely elegant falls out: the incident photon's own energy cancels out of the final expression entirely. The wavelength shift depends on nothing but the scattering angle θ and a cluster of fundamental constants (h, the electron's mass, c). A harder X-ray photon and a softer one, scattered through the same angle, pick up exactly the same Δλ. Not obvious in advance, but it's what the math says, and experiment confirms it.

The quantity h/m_ec — 2.43 picometres — is called the Compton wavelength, and it sets the scale for everything here. Maximum possible shift, at 180° direct backscatter, comes out to exactly twice that: 4.86 pm, regardless of what photon you started with.

Why X-rays show it and visible light doesn't

That 2.43 pm figure is a fixed absolute number, and whether it's noticeable depends entirely on what you're comparing it to. X-ray wavelengths run somewhere around 10 to 100 picometres, so a shift of a couple of picometres is a substantial fraction of the whole — genuinely easy to measure. Visible light sits around 400 to 700 nanometres, thousands of times longer, so that same few-picometre shift becomes a fractional change too small to detect against the baseline. Same physics, same shift — it's just swamped by scale in one case and clearly visible in the other.

Where the electron actually goes

Momentum has to balance in both directions, not just along the original beam. Since the incoming photon carries zero sideways momentum, whatever sideways momentum the scattered photon picks up has to be exactly canceled by the recoiling electron — which forces the electron onto the opposite side from wherever the photon scattered. It also means the electron can never recoil straight backward; it always has some forward component, because it has to help carry forward the momentum the photon gave up. Neither of these is a separate rule — both fall directly out of the same single momentum-conservation equation.

Why this mattered so much in 1923

Compton discovered the shift experimentally while scattering X-rays off graphite, and no amount of careful analysis under classical wave theory could account for what he saw. Treating the collision as a genuine relativistic particle interaction — photon carrying real momentum p=h/λ, colliding with an electron exactly like two billiard balls — matched the data. That match, more than the photoelectric effect alone, forced physicists to take the photon concept seriously as something carrying real, particle-like momentum rather than just energy delivered in packets. Compton won the Nobel Prize for it in 1927, and the same scattering mechanism remains genuinely important today in radiation therapy dosimetry and in modeling how X-rays and gamma rays interact with matter in astrophysical environments.

Try it yourself

Compton Effect Photoelectric Effect