Intuition The one core idea
A silicon crystal is nothing more than one atom's bonding pattern — 4 arms pointing to the corners of a tent — copied endlessly in every direction . Everything else on the parent page (angles, dot products, atom counts, densities) is just careful bookkeeping about that single repeated shape.
This page assumes nothing . Before we can read the parent note on the silicon lattice , we need to earn every symbol it throws at us. We build them in an order where each one leans only on the ones before it.
Definition Atom (working picture)
An atom is a tiny ball with a positive core (the nucleus) surrounded by electrons. For crystals, forget the inside — picture it as a sphere with a few "hands" (bonds) sticking out . Silicon has exactly 4 hands .
Intuition Why we only care about the hands
A crystal is decided by how atoms connect , not what's inside them. So we shrink each atom to a dot and draw its bonds as lines. That's the whole game.
Definition Valence electron
A valence electron is an outer-shell electron an atom can use to bond. Silicon is in group IV of the periodic table, which is just a label meaning it has 4 of them.
The picture: four short line segments leaving the atom-dot. Each one is a hand looking for another hand to hold.
Why the topic needs it: those 4 hands are the reason silicon bonds to 4 neighbours, and 4 neighbours is what forces the whole geometry. Deep dive: Valence electrons and the octet rule .
silicon’s own 4 + one from each neighbour 4 = 8
Before the parent note says "bond angle ≈ 109. 5 ∘ ", you must know what an angle is .
Definition Angle and the degree symbol
∘
An angle measures how much you turn between two directions. A full turn is 36 0 ∘ ; a quarter turn (a square corner) is 9 0 ∘ . The little circle ∘ just means "degrees."
A 9 0 ∘ angle is a perfect square corner (look at the amber corner marker in the figure).
An angle bigger than 9 0 ∘ is called obtuse — it looks "opened up," leaning backwards.
The silicon bond angle 109. 5 ∘ is obtuse: the bonds lean away from each other. That "leaning away" is the physical fact that atoms want maximum space.
The parent note writes things like a = ( 1 , 1 , 1 ) . That bold letter is a vector .
A vector is an arrow: it has a direction (which way it points) and a length (how long it is). We write it in bold, like a , and we describe it by its three coordinates ( x , y , z ) — how far to step right, up, and out-of-the-page .
The picture: an arrow from the origin (the corner 0 , 0 , 0 ) to the point ( x , y , z ) .
a = ( 1 , 1 , 1 )
Start at the origin. Step + 1 right, + 1 up, + 1 out of the page. The arrow to that spot is a . A minus sign, as in b = ( 1 , − 1 , − 1 ) , just means step backwards along that axis.
Why the topic needs it: each of silicon's 4 bonds is a direction. To find the angle between two bonds, we treat them as two vectors — and there's exactly one tool for the angle between vectors.
The parent note writes ∣ a ∣ = 3 . Here's where that comes from.
Definition Magnitude (length)
∣ a ∣
The bars ∣ ⋅ ∣ mean "the length of." For a vector a = ( x , y , z ) :
∣ a ∣ = x 2 + y 2 + z 2
Intuition Why a square root? It's just Pythagoras in 3D
In a right triangle, the long side obeys c 2 = x 2 + y 2 , so c = x 2 + y 2 . In 3D you do it twice — once in the floor, once going up — and get x 2 + y 2 + z 2 . The square root is there because Pythagoras gives you length-squared , and you want plain length.
∣ a ∣ = 1 2 + 1 2 + 1 2 = 3
Why the topic needs it: the angle formula divides by both vectors' lengths, so we must be able to compute them.
The very next tool, the dot product, secretly carries an angle inside it. But an angle is a turn, and the dot product is a plain number — so we first need the gadget that converts one into the other. That gadget is the cosine.
Definition The angle between two vectors,
θ
When two arrows both start at the origin, ==θ == (the Greek letter "theta") is simply the angle between them — how far you'd swing one arrow to line it up with the other. Look back at the obtuse angle in the figure of Section 2: that opening is a θ .
cos θ
For an angle θ , ==cos θ == is a single number between − 1 and + 1 that says "how much two directions line up."
At θ = 0 ∘ (perfectly aligned) it is + 1 .
At θ = 9 0 ∘ (perpendicular) it is 0 .
At θ = 18 0 ∘ (perfectly opposite) it is − 1 .
So a positive cosine means "pointing similarly," a negative cosine means "pointing apart."
Intuition Why introduce cosine now?
The next tool converts two coordinate-lists into a number that equals ∣ a ∣∣ b ∣ cos θ . We can't read that formula until we know what cos θ means — so we earn it first, before we ever write it down.
This is the heart of the derivation, so we build it slowly.
a ⋅ b
The dot product of a = ( a 1 , a 2 , a 3 ) and b = ( b 1 , b 2 , b 3 ) is a single number:
a ⋅ b = a 1 b 1 + a 2 b 2 + a 3 b 3
Multiply matching coordinates, then add.
Intuition Why THIS tool and not another? Because it secretly stores the angle
There is a second identity for the exact same number:
a ⋅ b = ∣ a ∣ ∣ b ∣ cos θ
where (from Section 5) θ is the angle between the two arrows and cos θ is the "how-much-they-line-up" number. So the dot product is the one operation that connects coordinates to the angle between them. That's why the parent note reaches for it — no other simple tool turns "two lists of numbers" into "the angle between two directions."
When two arrows point the same way , cos θ = 1 and the dot product is large positive.
When they are perpendicular (9 0 ∘ ), cos θ = 0 and the dot product is exactly 0 .
When they point apart (obtuse), cos θ is negative — and this is exactly what happens for silicon's bonds.
Recall Check your understanding of sign
Two arrows pointing roughly opposite directions: is their dot product positive or negative? ::: Negative — because cos θ < 0 for obtuse angles, and this is what confirms silicon's bonds lean away from each other.
The dot product hands us cos θ = − 3 1 . But we wanted the angle θ , not its cosine. We need the reverse button.
Definition arccos — the reverse question
==arccos == asks the reverse of cosine: "Which angle has this cosine?" If cos θ = − 3 1 , then θ = arccos ( − 3 1 ) hands you back the angle itself.
Intuition Why we need the reverse
Cosine goes angle → number ; arccos goes number → angle , just like a square root undoes a square. Feed in − 3 1 , get out ≈ 109.4 7 ∘ — the famous tetrahedral angle.
θ = arccos ( − 3 1 ) ≈ 109.4 7 ∘
Because − 3 1 is negative , θ lands between 9 0 ∘ and 18 0 ∘ — i.e. obtuse, i.e. bonds spread apart. Every case checks out.
A tetrahedron is a 4-cornered pyramid (a "tent" with a triangular base). Its 4 corners are the most spread-apart 4 points you can place around a centre.
The picture: put an atom at the centre, its 4 bonds reaching to the 4 corners of this tent. The angle between any two of those bonds is the 109. 5 ∘ we just derived.
Intuition Why a tetrahedron and not a flat square?
Four things that repel each other (electron pairs) want maximum distance apart. On the flat, a square gives 9 0 ∘ ; lifting into 3D as a tent opens that to 109. 5 ∘ — strictly more room. Nature always picks the roomier arrangement. See Diamond cubic structure .
Definition Lattice constant
a
The lattice constant a is the length of one edge of the repeating cube (the "unit cell"). For silicon a ≈ 5.43 A ˚ .
An Ångström is a length unit for atoms: 1 A ˚ = 1 0 − 10 m = 1 0 − 8 cm . Atoms are a few Å across, so it's the natural ruler here.
The body diagonal runs from one corner of the cube through its centre to the opposite corner. By 3D Pythagoras its length is a 2 + a 2 + a 2 = a 3 .
Common mistake Don't confuse the cube edge with the bond length
a = 5.43 A ˚ is the cube edge . The bond length (atom-to-nearest-atom) is one quarter of the body diagonal:
bond length = 4 3 a ≈ 2.35 A ˚
Same crystal, two different distances.
N means
We write ==N == for the total number of atoms that belong to one unit cell — the count we get after fairly dividing up every shared atom. It is the number we'll plug into the density formula.
Definition Where the 4 interior atoms sit
The diamond-cubic cell is two overlapping cubes-of-atoms; the second set drops 4 atoms fully inside the cube, each nestled in a tetrahedral interstitial site — the little pocket surrounded by 4 corner/face atoms. They sit at the positions ( 4 1 , 4 1 , 4 1 ) , ( 4 3 , 4 3 , 4 1 ) , ( 4 3 , 4 1 , 4 3 ) , ( 4 1 , 4 3 , 4 3 ) (in units of a ). There are exactly 4 of them because the offset second lattice places one such atom in each of the cube's four "alternate-corner" tetrahedral pockets. Being wholly inside, none is shared — each counts as 1 .
N = corners 8 × 8 1 + faces 6 × 2 1 + interior 4 × 1 = 1 + 3 + 4 = 8
Why fractions at all? If every cube claimed its corner atoms fully, we'd count the same atom 8 times over. Fractions stop the double-counting.
Definition Number density
n
==n == = how many atoms sit in a given volume, in units of atoms per cubic centimetre (atoms/cm 3 ). It's simply atoms per cell divided by volume per cell :
n = a 3 N = a 3 8
Definition Scientific notation
5.0 × 1 0 22 means "5.0 followed by 22 zeros' worth of size" — a shorthand for gigantic (or tiny) numbers. 1 0 − 8 means 0.00000001 .
n = ( 5.43 × 1 0 − 8 cm ) 3 8 ≈ 5.0 × 1 0 22 atoms/cm 3
Why the topic needs it: later, when we add impurities (Doping of silicon ), we compare against this ∼ 5 × 1 0 22 benchmark to see how few dopant atoms we actually add.
You'll meet three more names built on top of the above:
Face-centred cubic (FCC) lattice — a cube with atoms at corners and face-centres. Silicon = two of these interlocked.
Energy band gap in semiconductors — because the lattice repeats perfectly, electron energies clump into allowed "bands" with a forbidden gap E g between them.
Holes and electrons as charge carriers — when a bond breaks, the freed electron and the empty spot ("hole") both carry current.
Miller indices and crystal planes and Single-crystal silicon wafer manufacturing — how we name directions and grow the perfect crystal.
Valence electrons = 4 hands
Length via Pythagoras sqrt
Cosine turns angle into number
Dot product = coordinate sum
Dot product stores cos theta
Tetrahedron = most spread out
Bond length = sqrt3 over 4 times a
Sharing fractions 1 over 8 and half
N = 8 atoms per unit cell
Number density n = N over a cubed
Diamond cubic silicon lattice
Test yourself — you're ready for the parent note when you can answer every line.
What does a vector ( x , y , z ) tell you? A direction and length: step x right, y up, z out-of-page from the origin.
How do you find the length ∣ a ∣ of ( x , y , z ) ? x 2 + y 2 + z 2 — 3D Pythagoras.
What does cos θ measure, and what is θ ? θ is the angle between two arrows; cos θ is a number in [ − 1 , + 1 ] saying how much they line up (+1 aligned, 0 perpendicular, −1 opposite).
What is the dot product a ⋅ b in coordinates? Multiply matching coordinates and add: a 1 b 1 + a 2 b 2 + a 3 b 3 .
Why is the dot product the right tool for angles? Because a ⋅ b = ∣ a ∣∣ b ∣ cos θ links coordinates to the angle between the arrows.
What does arccos do? Reverses cosine — given a cosine value it returns the angle.
Why is silicon's bond angle obtuse (over 9 0 ∘ )? Because cos θ = − 3 1 is negative, so θ lies between 9 0 ∘ and 18 0 ∘ ; bonds spread apart.
Why does silicon form exactly 4 bonds? Group IV → 4 valence electrons; sharing one with each of 4 neighbours completes the octet of 8.
What does N stand for, and why does a corner atom count as 8 1 ? N = total atoms per unit cell; a corner atom is shared among 8 cubes so only 8 1 belongs to one cube.
Where do the 4 interior atoms of the silicon cell sit? In the 4 tetrahedral interstitial pockets (at quarter-positions), placed by the offset second FCC lattice; each is fully inside so counts as 1.
What is the lattice constant a vs the bond length? a is the cube edge (
5.43 Å); bond length is
4 3 a ≈ 2.35 Å.
What is 1 Å in metres? 1 0 − 10 m (i.e. 1 0 − 8 cm).
How do you get number density n ? Atoms per cell over cell volume: n = N / a 3 = 8/ a 3 ≈ 5 × 1 0 22 cm − 3 .