Physical vapor deposition (PVD - sputtering): Visual walkthrough

We assume you know nothing but "things have mass" and "moving things carry energy." Everything else we build.

Step 1 — Two things you must picture first: mass and energy of a moving ball

WHAT. Before any collision, meet the two characters.

A projectile: our argon ion. It has a mass we call $M_{1}$ (a number saying "how much stuff is in it") and it is moving with speed $v_{1}$ .
A target atom: an aluminum atom sitting still on the surface. Mass $M_{2}$ , speed zero.

Two quantities describe motion:

Momentum $p = M v$ — "how hard it is to stop this thing." A heavy fast ball has lots of it.
Kinetic energy $E = \frac{1}{2} M v^{2}$ — "how much oomph it can deliver." Notice the speed is squared here but not in momentum; that difference is the whole secret of the derivation.

WHY these two. Nature keeps both books balanced in an ideal ("elastic") bounce: total momentum before = after, and total kinetic energy before = after. Two equations, so we can solve for two unknowns (the two speeds afterwards). That's why we picked exactly these tools and not, say, temperature.

PICTURE. The blue ion races in from the left; the yellow atom waits.

Figure — Physical vapor deposition (PVD - sputtering)

Step 2 — The two conservation laws, written out

WHAT. Let $v_{1}$ be the ion's speed before; after the hit, the ion has speed $u_{1}$ and the atom has speed $u_{2}$ . Write the two balance sheets.

$ion, before M_{1} v_{1} = ion, after M_{1} u_{1} + atom, after M_{2} u_{2}$

Each term is (mass)×(speed) — a momentum. The left side is all in the ion; the right side is shared.

$energy before \frac{1}{2} M_{1} v_{1}^{2} = ion after \frac{1}{2} M_{1} u_{1}^{2} + atom after \frac{1}{2} M_{2} u_{2}^{2}$

Same idea, but with the squared speeds — kinetic energy.

WHY. Two truths, two unknowns ( $u_{1}, u_{2}$ ). This is solvable algebra, no physics genius required past here.

PICTURE. The before/after strip: one arrow splits into two.

Step 3 — Solve for the atom's speed after impact

WHAT. Do the algebra (standard 1-D elastic result). Solving the two boxed equations gives the atom's post-hit speed:

$u_{2} = \frac{2 M _{1}}{M _{1} + M _{2}} v_{1}$

Read it: $u_{2}$ is $v_{1}$ scaled by the factor $\frac{2 M _{1}}{M _{1} + M _{2}}$ .

WHY it makes sense — check the extremes:

If the ion is very heavy ( $M_{1} ≫ M_{2}$ ): the factor $\to \frac{2 M _{1}}{M _{1}} = 2$ , so $u_{2} \to 2 v_{1}$ . A truck barely slows when it clips a marble; the marble flies off at nearly twice the truck's speed.
If the ion is very light ( $M_{1} ≪ M_{2}$ ): the factor $\to 0$ . A ping-pong ball can't budge a bowling ball.

PICTURE. The scaling factor plotted against the mass ratio.

Step 4 — Turn that speed into energy delivered to the atom

WHAT. The atom's kinetic energy after the hit is $E_{2} = \frac{1}{2} M_{2} u_{2}^{2}$ . Substitute $u_{2}$ from Step 3:

$E_{2} = \frac{1}{2} M_{2} (\frac{2 M _{1}}{M _{1} + M _{2}} v_{1})^{2} = \frac{4 M _{1} M _{2}}{( M _{1} + M _{2} ) ^{2}} \cdot E \frac{1}{2} M_{1} v_{1}^{2}$

The last piece, $\frac{1}{2} M_{1} v_{1}^{2}$ , is exactly the ion's incoming energy $E$ . So:

$E_{2} = γ E, γ \equiv \frac{4 M _{1} M _{2}}{( M _{1} + M _{2} ) ^{2}}$

$γ$ (Greek "gamma") is the fraction of the ion's energy that lands in the atom — a pure number between 0 and 1.
$E$ is what you dialed in with the target voltage.
$E_{2}$ is what the atom actually receives.

WHY the tool $γ$ . We named this messy factor because it will reappear everywhere. It answers one question: "of the energy I fired, how much reached the atom?"

PICTURE. $γ$ as a curve over the mass ratio, peaking at 1.

Step 5 — Why $γ$ peaks when the masses match

WHAT. Set $M_{1} = M_{2} = M$ . Then $γ = \frac{4 M ^{2}}{( 2 M ) ^{2}} = \frac{4 M ^{2}}{4 M ^{2}} = 1$ . All the energy transfers — the ion stops dead, like the front ball in Newton's cradle.

WHY it matters for real sputtering. This is the practical reason we pick argon ( $M_{1} = 40$ ) rather than, say, light helium: argon's mass is close to many target metals, so $γ$ is high and energy transfer is efficient. (For very heavy targets, even heavier ions like xenon do better.)

PICTURE. Newton's-cradle picture of the equal-mass "clean stop."

Step 6 — The atom must beat its "leash": surface binding energy

WHAT. An atom on the surface isn't free — its neighbors hold it with a surface binding energy $U_{s}$ (measured in electron-volts, eV). Think of $U_{s}$ as the depth of a hole the atom sits in: to escape, the energy it receives must at least equal $U_{s}$ .

Escape condition: $E_{2} \geq U_{s} ⟺ γ E \geq U_{s}$

WHY. Below this, the atom just jiggles and settles back — no ejection, no film. This is the physical origin of a threshold.

PICTURE. The atom in an energy "well" of depth $U_{s}$ ; a small kick fails, a big kick clears the rim.

Step 7 — Solve the inequality for the threshold energy

WHAT. Rearrange $γ E \geq U_{s}$ to find the smallest $E$ that works:

$E \geq \frac{U _{s}}{γ} ⟹ E_{t h} = \frac{U _{s}}{γ} = \frac{( M _{1} + M _{2} ) ^{2}}{4 M _{1} M _{2}} U_{s}$

Term by term:

$U_{s}$ on top — the deeper the leash, the more energy you need. Makes sense.
$γ$ on the bottom — the worse the energy transfer (small $γ$ ), the harder you must fire. So the threshold blows up when masses are badly mismatched, and is smallest ( $E_{t h} = U_{s}$ ) when $γ = 1$ .

WHY this is the parent's headline result. It's the exact boxed formula from the topic note, now earned from momentum + energy + a leash.

PICTURE. Threshold energy vs. mass ratio — a valley whose floor is $U_{s}$ .

Step 8 — Numbers: Ar → Al (checking the parent's worked example)

WHAT. $M_{1} = 40$ (Ar), $M_{2} = 27$ (Al), $U_{s} = 3.4$ eV.

$γ = \frac{4 ( 40 ) ( 27 )}{( 40 + 27 ) ^{2}} = \frac{4320}{4489} = 0.962$

$E_{t h} = \frac{U _{s}}{γ} = \frac{3.4}{0.962} \approx 3.53 eV$

WHY it's tiny compared to real voltages. In practice we accelerate ions to hundreds of eV — far above $3.5$ eV — so we sit deep in the linear-yield regime where each ion ejects many atoms. The threshold isn't the operating point; it's the floor below which nothing happens at all.

The one-picture summary

Everything at once: a ball of mass $M_{1}$ carrying energy $E$ hits a ball of mass $M_{2}$ ; a fraction $γ$ of that energy passes over; if it clears the well of depth $U_{s}$ , the atom escapes; the break-even point is $E_{t h} = U_{s} / γ$ .

Recall Feynman retelling — the whole walkthrough in plain words

A tiny cannonball (argon) flies in and smacks a marble (a metal atom) that's glued to a wall. When they hit, the cannonball hands over some of its "oomph" — but not all of it. How much it hands over depends only on their weights: if they weigh the same, the cannonball stops dead and gives the marble everything; if they're mismatched, most of the oomph bounces back and the marble barely moves. That handed-over fraction we called $γ$ . Now the marble is glued down with a certain strength $U_{s}$ — a hole it has to climb out of. If the oomph it got beats the glue, it pops free and flies off to coat the wafer. If not, it just wobbles and stays. Work backward from "just barely escapes" and you get the smallest cannonball energy that ever ejects anything: $E_{t h} = U_{s} / γ$ . That single line is the whole physics of why sputtering has a threshold.

Parent: Physical vapor deposition (PVD - sputtering)
The transport half of the story: Mean Free Path and Plasma Physics
Where these films go: Interconnect Metallization, Thin Film Deposition
The reverse process: Sputter Etching
Contrast method: Chemical Vapor Deposition

Flashcards

Post-collision speed of a struck stationary atom (1-D elastic)?

u_{2} = \frac{2 M _{1}}{M _{1} + M _{2}} v_{1}

Fraction of ion energy delivered to the atom?

γ = \frac{4 M _{1} M _{2}}{( M _{1} + M _{2} ) ^{2}}

, max

= 1

when

M_{1} = M_{2}

Threshold energy for ejection?

E_{t h} = U_{s} / γ

; smallest when masses match, blows up when mismatched.

Why argon, not helium, for aluminum?

Ar mass (40) is close to Al (27) →

γ \approx 0.96

; He gives

γ \approx 0.45

, doubling the threshold.