Molecular geometry — Foundations

Organic molecules are three-dimensional objects, and most of the interesting things about them — why one enzyme binds a drug and its mirror image does nothing, why attacking a carbon from one side turns it inside out like an umbrella in the wind, why a ring prefers one shape over another — are facts about shape. This section is where the flat drawings from the previous two sections become objects in space.

The idea

VSEPR — valence shell electron pair repulsion — rests on one observation: electron groups repel each other, so they arrange themselves as far apart as possible around the central atom. An electron group is a bond (single, double or triple, each counting once) or a lone pair. This is exactly the same count you used for hybridization, which is not a coincidence: hybridization tells you which orbitals the central atom uses, and VSEPR tells you the shape those orbitals settle into. Two descriptions of one physical arrangement.

There is one complication that trips up everyone the first time. The name of a molecular shape describes only the positions of atoms. Lone pairs are still there and still pushing, but they are invisible in the name. So the same four-group tetrahedral arrangement produces three differently-named shapes depending on how many of those four groups happen to be lone pairs.

4 groups (sp³) → tetrahedral base · 3 groups (sp²) → trigonal planar base · 2 groups (sp) → linear

The shapes you need

Every shape in the table, with the lone pairs drawn in as ghosts. This is the one thing that trips everyone: the lone pairs are there and still pushing, but the shape's name describes only where the atoms are. Follow the bottom row of three — four groups every time, the same tetrahedral arrangement every time — and watch the name change from tetrahedral to pyramidal to bent purely because more of the corners have gone invisible.

Groups	Lone pairs	Shape	Angle	Example
2	0	Linear	180°	CO₂
3	0	Trigonal planar	120°	BF₃, an alkene carbon
3	1	Bent	<120°	SO₂
4	0	Tetrahedral	109.5°	CH₄
4	1	Trigonal pyramidal	107°	NH₃
4	2	Bent	104.5°	H₂O

Worked examples

Methane, CH₄. Four bonds, no lone pairs. All four groups are visible atoms → tetrahedral, 109.5°.

Ammonia, NH₃. Three bonds plus one lone pair — still four groups, so still a tetrahedral arrangement. But the name counts only the three N–H bonds, which form a trigonal pyramidal shape, like a tripod with nitrogen at the apex.

Water, H₂O. Two bonds plus two lone pairs, again four groups. Only the two O–H bonds are visible → bent.

Notice that all three have the same underlying tetrahedral geometry and the same sp³ hybridization. The only thing that changes is how many corners happen to be occupied by lone pairs.

The same three molecules as solids. Every one of them is a tetrahedron — four electron groups, four corners — and the only thing that changes down the row is how many of those corners hold a lone pair (the paired dots) instead of an atom. The shape's name counts only the atoms, which is why one arrangement collects three names; and because a lone pair is held by one nucleus rather than shared between two, it takes more room and squeezes the angle shut as you go.

A lone pair is held by one nucleus rather than shared between two, so it spreads out closer to the central atom and takes up more angular room than a bonding pair does. That squeezes the remaining bonds together: ammonia's H–N–H angle is 107° and water's H–O–H angle is 104.5°, both compressed below the ideal 109.5°. Water, with two lone pairs doing the squeezing, is compressed further than ammonia with one — a small quantitative prediction that the measurements confirm exactly.

Three molecules with four electron groups each, so all three are built on the same tetrahedron and all three are sp³. The only difference is how many corners hold a lone pair instead of an atom — and because a lone pair answers to one nucleus rather than two, it spreads out and takes more angular room (shaded). Each one it gains costs the remaining bonds a couple of degrees. The measured angles land exactly where that argument says they should.each added lone pair takes more room than a bond, and squeezes the rest closer

Reading geometry off an organic structure

In practice you rarely apply VSEPR to a whole molecule. You apply it atom by atom, at whichever center you care about, and a molecule of any size has several different local geometries at once.

Take ethanol, CH₃CH₂OH. The methyl carbon: 4 groups, tetrahedral. The CH₂ carbon: 4 groups, tetrahedral. The oxygen: 2 bonds plus 2 lone pairs, bent, with a C–O–H angle near 105°. Or take acetic acid: the methyl carbon is tetrahedral, the carbonyl carbon has 3 groups and is trigonal planar at 120°, and the hydroxyl oxygen is bent. The molecule as a whole has no single shape — it has a shape at every atom.

The practical payoff is that the trigonal planar carbons are the flat parts, and flatness has a specific consequence: a flat carbon has two faces, and they are open. Anything approaching it can arrive from above or from below, and unless something else in the molecule distinguishes those two directions, both are equally available. A tetrahedral carbon offers no such choice — all four positions are already occupied. That difference, between a site with two open faces and a site with none, decides the three-dimensional outcome of a great many reactions later on, and every one of them traces back to a group count you can do in five seconds.

Two recurring errors. Forgetting that lone pairs count as groups — this is what makes an amine pyramidal rather than planar, and it is the most common geometry mistake in Module 1. And counting a double bond as two groups; it points at one neighbor, so it occupies one direction. An alkene carbon has three groups, not four, and is therefore flat.

Beyond four groups

Five groups with no lone pairs gives a trigonal bipyramidal shape — three equatorial positions at 120° in a plane, plus two axial positions at 90° above and below, as in PCl₅. Six groups gives an octahedral shape, four equatorial and two axial, all at 90°, as in SF₆. These are rare in day-to-day organic chemistry but worth recognizing in sulfur and phosphorus compounds.

The note from Hybridization applies here as well: these shapes are real and experimentally well established, but the old "sp³d / sp³d²" rationale for why period 3+ atoms can adopt them has been superseded. Atomic size and the ionic character of the extra bonds do the explaining, not genuine d-orbital mixing.

Drawing three dimensions on paper

The standard convention for showing a tetrahedral center on a flat page uses two plain lines in the plane of the paper, a solid wedge for a bond coming toward you, and a dashed wedge for one going away. A correctly drawn tetrahedral center therefore has exactly two plain bonds, one wedge and one dash, with the wedge and dash adjacent to each other rather than opposite.

The object and the notation, side by side — because the notation only makes sense once you have seen what it is standing in for. All four bonds on the left are identical; the drawing on the right has to fake that with two different kinds of line, and the wedge and dash land next to each other rather than opposite precisely because the two bonds they represent do. Read it the other way and you have drawn the mirror image.

The notation that puts three dimensions on a flat page. It looks like a drawing convention and it is, but it is also the input to a real calculation: in Module 5 you will assign R and S configurations by reading which group points at you and which points away, and a centre drawn with its wedge and dash on opposite sides is not a tetrahedron at all. Practise it now, while nothing depends on it.A correct tetrahedral centre has exactly two plain bonds, one wedge and one dash — and the wedge and the dash sit NEXT to each other, never opposite.

This is worth practising deliberately now, because in Module 5 you will be assigning R and S configurations from these drawings, and the whole exercise depends on reading wedges and dashes correctly. You can rotate real structures in three dimensions with the 3D viewer tool, which is the fastest way to build the intuition that a flat drawing is a projection of something solid.

What carries forward

Geometry is the input to the very next section, where bond dipoles either cancel or add depending on shape. It is the whole of Module 4, where the question is which rotational and ring shapes a molecule actually adopts. It is the foundation of Module 5, since chirality is a purely geometric property. And it explains mechanism after mechanism: the backside attack of SN2, the anti-periplanar requirement of E2, the face selectivity of carbonyl addition.