Valence Bond Theory (VBT) — hybridization (sp, sp², sp³, sp³d, sp³d²): Visual walkthrough

We will use ideas from Atomic Orbitals (s, p, d shapes) and the "balloons push apart" picture behind VSEPR Theory. Everything else we build from zero.

Step 1 — The problem: four hands that must be identical

WHAT. Carbon in methane $C H_{4}$ grabs four hydrogen atoms. We want to know: at what angle do these four bonds sit relative to one another?

WHY start here. Before any maths, we must see the object. A carbon sits at the centre; four bonds shoot outward. If we knew nothing, we might guess the bonds sit at $90°$ (like the corners of a plus-sign in 3D). We will show that guess is wrong and find the true angle.

PICTURE. The red dot is the carbon nucleus. The four black arrows are the four identical bonds. The single angle $θ$ between any two of them (they are all equal) is what we are hunting.

Figure — Valence Bond Theory (VBT) — hybridization (sp, sp², sp³, sp³d, sp³d²)

Step 2 — Why not just use the raw s and p orbitals?

WHAT. Carbon owns one round $s$ orbital and three dumbbell $p$ orbitals ( $p_{x}, p_{y}, p_{z}$ ) that point along the three perpendicular axes — like the three edges of a room's corner.

WHY this matters. If carbon bonded with these directly, three bonds would sit at $90°$ (the $p$ orbitals are perpendicular) and the fourth (from the round $s$ ) would have no direction at all. That gives unequal bonds — contradicting experiment, which says all four are identical. So the raw orbitals cannot be the answer; the atom must average them. That averaging is hybridization.

PICTURE. Left: the three $p$ orbitals at stiff $90°$ angles plus a spherical $s$ — a mismatched set (red marks the odd one out, the $s$ ). Right: after blending, four identical fat lobes spread out evenly. We now need the angle of that even spread.

Step 3 — The trick: hide the tetrahedron inside a cube

WHAT. The four "maximally separated" directions are the corners of a regular tetrahedron. A tetrahedron is hard to write coordinates for directly — so we place it inside a cube, where the corners have dead-simple coordinates.

WHY a cube. A cube's eight corners are just all combinations of $+ 1$ and $- 1$ in three axes. If we pick four alternating corners (never two that share an edge), those four are all equally far apart — they are a perfect tetrahedron. This converts a hard geometry problem into easy arithmetic. This is the whole reason we reach for coordinates: to turn "angle" into "a division we can compute."

PICTURE. A cube centred at the origin. Four corners are marked in red; they alternate, so no two chosen corners touch along an edge. The red corners are our four bond directions.

Step 4 — The tool we need: the dot product (why this and not something else)

WHAT. To get the angle between two arrows we use the dot product. For two vectors $A$ and $B$ :

$A \cdot B = multiply matching parts, add them A_{x} B_{x} + A_{y} B_{y} + A_{z} B_{z} = lengths \times c o s (angle between) ∣ A ∣ ∣ B ∣ cos θ$

WHY the dot product specifically. We have coordinates, and we want an angle. The dot product is the one tool that connects the two: its left form is pure arithmetic on the coordinates, and its right form contains exactly the $cos θ$ we want. Setting them equal lets $θ$ fall out. No other single operation ties coordinates to an angle so directly.

Term by term in the right-hand form:

$∣ A ∣$ ::: the length of arrow $A$ (how long, ignoring direction).
$∣ B ∣$ ::: the length of arrow $B$ .
$cos θ$ ::: the cosine of the angle between them — this is the unknown we solve for.

PICTURE. Two arrows from a shared origin with angle $θ$ marked in red. The figure shows: when arrows point the same way $θ = 0$ and $cos θ = 1$ (biggest dot product); at right angles $θ = 90°$ and $cos θ = 0$ ; pointing apart, $cos θ$ goes negative. That sign is exactly what will tell us the bonds point away from each other.

Step 5 — Compute the top: the dot product of two bonds

WHAT. Plug $A = (1, 1, 1)$ and $B = (1, - 1, - 1)$ into the arithmetic form:

$A \cdot B = x : + 1 (1) (1) + y : - 1 (1) (- 1) + z : - 1 (1) (- 1) = 1 - 1 - 1 = - 1$

WHY it's negative. A negative dot product means $cos θ < 0$ , which means $θ > 90°$ — the bonds lean away from each other, past a right angle. This is our first proof the naive " $90°$ " guess is wrong: the true angle is obtuse.

Term by term:

$(1) (1) = + 1$ ::: both arrows share the same $x$ -direction, contributing agreement.
$(1) (- 1) = - 1$ ::: they disagree in $y$ (one up, one down).
$(1) (- 1) = - 1$ ::: they disagree in $z$ too.
Sum $= - 1$ ::: disagreement wins → the arrows spread apart.

PICTURE. The two red arrows $A$ and $B$ drawn from the origin inside the cube, visibly opening past $90°$ . The obtuse gap is shaded.

Step 6 — Compute the bottom: the lengths

WHAT. The length of any arrow is the square root of (sum of its squared parts) — the 3D Pythagoras:

$∣ A ∣ = 1^{2} + 1^{2} + 1^{2} = 3, ∣ B ∣ = 1^{2} + (- 1)^{2} + (- 1)^{2} = 3$

WHY. The dot-product formula divides by both lengths, so we must know them. Squaring kills the minus signs (that's why $∣ B ∣$ is also $3$ despite the negatives) — length never cares about direction. Both arrows reach a cube corner from the centre, so of course they are the same length, $3$ .

Term by term:

$1^{2} + 1^{2} + 1^{2} = 3$ ::: each coordinate is $\pm 1$ , its square is $1$ , three of them add to $3$ .
$3$ ::: undo the squaring to get the true length of the arrow.

PICTURE. One red arrow to a corner, with a right-triangle drawn on the cube showing how $1^{2} + 1^{2}$ (a face diagonal) then combines with the last edge to give $3$ (the body diagonal to the corner).

Step 7 — Put it together: the angle appears

WHAT. Substitute the top ( $- 1$ ) and bottom ( $3 \cdot 3 = 3$ ) into the rearranged formula:

$cos θ = \frac{A \cdot B}{∣ A ∣ ∣ B ∣} = \frac{- 1}{3 \cdot 3} = \frac{- 1}{3}$

Now we ask the reverse question — "which angle has cosine $- \frac{1}{3}$ ?" — using $cos^{- 1}$ (arccosine), the button that undoes cosine:

$θ = cos^{- 1} (- \frac{1}{3}) = 109.47° \approx 109.5°$

WHY $cos^{- 1}$ . Our formula handed us $cos θ$ , but we want $θ$ itself. $cos^{- 1}$ is precisely the inverse: feed it a cosine value, it returns the angle that produced it. It is the only step that converts "the ratio" back into "the angle."

Term by term:

$\frac{- 1}{3}$ ::: negative → obtuse angle (confirmed from Step 5's sign).
$cos^{- 1} (\cdot)$ ::: the "which angle?" operator, undoing the cosine.
$109.47°$ ::: the exact tetrahedral bond angle — derived, not memorised.

PICTURE. The cosine curve from $0°$ to $180°$ . A red horizontal line at height $- \frac{1}{3}$ cuts the curve at exactly $109.47°$ , shown by a dropped red vertical. This makes "which angle has this cosine" a literal picture.

Recall Where does this angle live in real molecules?

Because the derivation used only "four maximally separated directions," the $109.5°$ appears in every $s p^{3}$ centre with four identical positions: $C H_{4}$ , the C–C skeleton of diamond, the tetrahedral core of $N H_{4}^{+}$ . When lone pairs distort the picture (water $104.5°$ , ammonia $107°$ ), it's the same tetrahedron squeezed — the ideal we just built is the starting point.

Step 8 — Edge cases: what if it isn't four?

WHAT. The identical logic ("arrows to maximum separation, then $cos^{- 1}$ ") handles the other steric numbers. Their answers:

Directions	Shape	Angle	Where the number comes from
$2$ ( $s p$ )	Linear	$180°$	Two arrows apart = a straight line; $cos^{- 1} (- 1) = 180°$
$3$ ( $s p^{2}$ )	Trigonal planar	$120°$	Three arrows in a plane, $360°/3 = 120°$
$4$ ( $s p^{3}$ )	Tetrahedral	$109.5°$	$cos^{- 1} (- \frac{1}{3})$ — derived above

WHY show these. A reader must never meet a case we didn't cover. The two-arrow and three-arrow cases are simpler than four (they live in a line and a flat plane), which is why they give the clean $180°$ and $120°$ — no cube needed.

Degenerate check ( $n = 1$ or $n = 0$ ). With only one bond there is no angle to speak of (nothing to compare against), and with zero bonds there is no geometry at all. The formula $SN \geq 2$ is where "angle" first becomes meaningful — below that, the question itself is empty.

PICTURE. Three mini-panels side by side: two arrows (red, straight line, $180°$ ), three arrows (flat, $120°$ ), four arrows (3D tetrahedron, $109.5°$ ) — the family the same method produces.

The one-picture summary

WHAT. One figure compresses the whole chain: cube → pick 4 alternating corners → two red arrows → dot product $- 1$ over lengths $3$ → $cos^{- 1}$ → $109.5°$ .

Recall Feynman: the whole walkthrough in plain words

We wanted the angle between the four bonds of methane. The raw round-and-dumbbell orbitals would give an ugly, uneven set, so the atom blends them into four identical fat lobes that shove each other as far apart as possible — like balloons on a knot. "As far apart as possible for four" is a shape called a tetrahedron, and the sneaky way to get its coordinates is to hide it inside a cube: pick four corners that never touch, and each corner is just a mix of $+ 1$ and $- 1$ . To turn those corners into an angle we use the dot product, the one tool linking coordinates to angles: multiply matching parts and add ( $1 - 1 - 1 = - 1$ on top), divide by the two arrow lengths ( $3 \times 3 = 3$ on the bottom), giving $cos θ = - \frac{1}{3}$ . The minus sign already tells us the bonds lean apart past a right angle. Finally we press the "which angle has this cosine?" button, $cos^{- 1}$ , and out drops $109.5°$ — not a magic number from a table, but a fact you can now rebuild from a cube and one division.