Covalent bonding in silicon crystals: Visual walkthrough

We assume you know nothing beyond: atoms have electrons, and heat means "jiggling." Everything else we build.

Step 1 — One silicon atom and its four spare electrons

WHAT. We draw a single silicon atom. Around its centre sit 4 valence electrons — the outermost, most loosely-held electrons. "Valence" just means "the ones available for bonding."

WHY start here. Every later step is about what these 4 electrons do. If we do not first see that silicon owns exactly 4 (not 3, not 8), nothing that follows makes sense. Silicon is atomic number 14, and only the outer shell matters for bonding — the inner 10 electrons are spectators.

PICTURE. The magenta dots are the 4 valence electrons. The violet ring is the outer shell, drawn with 8 empty slots because a full, comfortable shell wants 8 — silicon has filled exactly half.

Figure — Covalent bonding in silicon crystals

Step 2 — Four neighbours, four shared pairs → a full shell

WHAT. Place four more silicon atoms around our central one. Each neighbour lends one electron, and our atom lends one back to each — so between every pair sits a shared pair of electrons. That shared pair is the covalent bond.

WHY sharing. Our atom now "sees" $4 (its own) + 4 (borrowed) = 8$ electrons — a full shell. It never had to pay the huge energy cost of stealing or donating 4 electrons; sharing is the cheap route. This is the covalent network described in the parent note.

PICTURE. Four bonds radiate out; each orange lens is a shared electron pair. Count the electrons touching the centre: eight. This is a flat cartoon of the true 3D tetrahedron (bond angle $109.5°$ ).

Step 3 — The cold limit: every electron is locked (0 K)

WHAT. At absolute zero ( $T = 0$ K, no heat at all), every valence electron is trapped inside a bond. None are free to wander.

WHY it matters. A material conducts only if charges can move. With zero free electrons, silicon at 0 K is a perfect insulator — no current, full stop. This is the degenerate/limiting case we must nail before adding heat: it tells us the free-carrier count starts at zero and can only grow.

PICTURE. All bonds intact (solid orange), all electrons pinned. A battery is attached but the ammeter reads zero — nothing moves.

Recall Why "insulator at 0 K" is the anchor of the whole derivation

Because the free-carrier count $n_{i}$ must be 0 at $T = 0$ . ::: Any formula we derive must return $n_{i} \to 0$ as $T \to 0$ . We will check that the exponential does exactly this.

Step 4 — Heat breaks a bond and makes a pair

WHAT. Add heat. Heat is random jiggling energy. If one bond receives at least a threshold energy $E_{g}$ , its electron snaps free and wanders. But breaking one bond leaves an empty slot behind — a hole.

WHY a pair, not just an electron. You cannot free an electron without leaving a vacancy. So every broken bond makes exactly two things at once: a free electron and a hole. This "always in pairs" fact is the seed of the mysterious factor of 2 coming in Step 6.

PICTURE. One bond snaps (drawn broken/red). A magenta electron flies off; a hollow circle (the hole) stays behind. Count: 1 electron + 1 hole = 1 pair.

Step 5 — Why heavier heat means exponentially more broken bonds

WHAT. How many bonds break at temperature $T$ ? Not linearly — the fraction that have enough energy to jump the gap $E_{g}$ follows a Boltzmann factor: $fraction with energy \geq E_{g} \sim e^{- E_{g} / (k_{B} T)}$

WHY the exponential (and why this tool, not a straight line). Thermal energies are shared randomly among countless bonds. The chance that any one bond happens to collect an unusually large energy $E_{g}$ falls off exponentially as $E_{g}$ grows or as $T$ (the "typical" energy scale $k_{B} T$ ) shrinks. We use the exponential precisely because it answers "how rare is a rare, large fluctuation?" — a linear guess would badly under-count how sharply carriers switch on with heat.

PICTURE. A histogram of bond energies (the thermal spread). Only the sliver past the vertical $E_{g}$ line breaks free — and that sliver grows dramatically as the whole curve slides right with rising $T$ .

Step 6 — Where the "2" comes from (the pair splits the cost)

WHAT. The parent note's result has a 2 in the denominator: $n_{i} \propto e^{- E_{g} / (2 k_{B} T)}$ . This step earns it.

WHY the 2. Because bonds break in pairs (Step 4), the crystal must keep the number of free electrons $n$ equal to the number of holes $p$ — you can't have one without the other. This balance is the mass-action law: $n p = n_{i}^{2}$ Each population — electrons and holes — carries away roughly half the influence of the gap energy in the exponent. When you solve for $n_{i}$ (with $n = p = n_{i}$ ), the single Boltzmann factor $e^{- E_{g} / k_{B} T}$ gets a square root taken of it, and $e^{- E_{g} / k_{B} T} = e^{- E_{g} / (2 k_{B} T)}$ . The square root is the factor of 2.

PICTURE. Two side-by-side tanks: an "electron" tank and a "hole" tank, always kept at equal level (that's $n = p$ ). The gap energy $E_{g}$ pours into a splitter that feeds both tanks — each gets half the exponent.

Step 7 — Reading the finished formula (worked check)

WHAT. Put it all together and use it. Silicon: $E_{g} = 1.12$ eV. Compare 300 K and 400 K.

WHY. A formula you cannot plug numbers into is decoration. This confirms the parent note's Example 2 and shows the exponential's bite.

PICTURE. A curve of $n_{i}$ versus $T$ on a log axis — a rising line. Two dots marked at 300 K and 400 K; the vertical jump between them is the ~230× increase.

The one-picture summary

From bare atom → shared octet → cold lock → heat snaps a pair → rare bond-breaks follow $e^{- E_{g} / k_{B} T}$ → pairing ( $n p = n_{i}^{2}$ ) takes a square root → the final law:

$n_{i} \propto e^{- E_{g} / (2 k_{B} T)}$

Recall Feynman retelling — the whole walkthrough in plain words

A silicon atom has four spare "hands." It holds hands with four neighbours, so it feels a full set of eight — that hand-holding is the covalent bond (Steps 1–2). When it's freezing cold, everyone grips tight, nobody moves, no electricity (Step 3). Warm it up, and once in a while a pair of hands lets go: one hand runs off free (an electron), leaving an empty grip behind (a hole) — always two things at once (Step 4). How often does a grip get enough random shove to let go? That's rare, and rareness falls off exponentially — the Boltzmann factor $e^{- E_{g} / k_{B} T}$ (Step 5). Because the freed hand and the empty grip come as a matched pair that must stay equal in number, solving for the count takes a square root, which halves the exponent — and that is where the 2 in $e^{- E_{g} / (2 k_{B} T)}$ is born (Step 6). Plug in numbers and a 100-degree warm-up multiplies the free carriers about 200-fold (Step 7). Locked when cold, alive when warm — that in-between is exactly what makes silicon a semiconductor.

Connections

Band theory of solids — bonds become the valence and conduction bands; $E_{g}$ is the gap between them.
Electron-hole pair generation and recombination — Step 4's pair-making, in detail.
Intrinsic vs extrinsic semiconductors — $n_{i}$ is the intrinsic count before doping.
Doping with donors and acceptors — the other way to make carriers without heat.
Diamond cubic crystal structure — the true 3D tetrahedral lattice behind our flat cartoon.
Ionic vs covalent vs metallic bonding — why silicon shares instead of stealing.