VSEPR Molecular Geometry
Turn a molecule in three dimensions and see why water is bent, with every lone pair drawn beside the bonds that hide behind them.
Visualiser
Drag the molecule to turn it, or use the arrow keys. Space plays and pauses.
- Molecular shape The shape of the bonds alone. Example: H2O.
- bent
- Electron geometry How all the domains are arranged, lone pairs included. Often not the same as the shape.
- tetrahedral
- Steric number 2 bonding, 2 lone.
- 4
- Bond angle Closed 5 degrees from the ideal 109.5, because lone pairs take more room than bonds.
- 104.5 deg
- Polarity The bond dipoles do not cancel, so the molecule has a net dipole.
- polar
- Net dipole The length of the vector sum of the bond directions, assuming identical substituents.
- 1.15 bond units
Every shape sharing an electron geometry sits together below. Switching between them without changing the arrangement is the fastest way to see that the shape is decided by how many domains hold a lone pair rather than by the arrangement itself.
- AX2 linear, CO2
- AX3 trigonal planar, BF3
- AX2E1 bent, SO2
- AX4 tetrahedral, CH4
- AX3E1 trigonal pyramidal, NH3
- AX2E2 bent, H2O
- AX5 trigonal bipyramidal, PCl5
- AX4E1 see-saw, SF4
- AX3E2 T-shaped, ClF3
- AX2E3 linear, XeF2
- AX6 octahedral, SF6
- AX5E1 square pyramidal, BrF5
- AX4E2 square planar, XeF4
Citing this tool
Last updated . Add the date you accessed it as well, which a citation of a page that can change asks for. If a specific result matters, cite the permalink from the tool’s share row instead of this page: it reproduces the exact parameters.
The equation
Gillespie and Nyholm (1957)
Two different questions, usually asked as one
Where are the electron domains, and what shape do the atoms make? Those have different answers, and treating them as one question is what turns this topic into a list to memorise.
Water and methane make the point. Both have four electron domains around the central atom, arranged tetrahedrally, and nothing whatsoever differs between them in that arrangement. Methane puts a bond in all four, so it is tetrahedral. Water puts bonds in two and lone pairs in the other two, so it is bent. Same geometry, different shape, because only the bonds are atoms you could see.
Switch between AX4, AX3E1 and AX2E2
in the tool and watch the tetrahedron stay exactly where it is while the
shape changes underneath it. That is the whole idea.
Why lone pairs squeeze the angles
A bonding pair is shared between two nuclei, so it is held in a fairly narrow region between them. A lone pair answers to only one nucleus and spreads wider. Being fatter, it pushes harder on its neighbours, so the bonds get shoved closer together.
| Molecule | Lone pairs | Bond angle |
|---|---|---|
| CH₄ | 0 | 109.5° |
| NH₃ | 1 | 107° |
| H₂O | 2 | 104.5° |
Roughly two and a half degrees per lone pair. Worth learning as one pattern rather than three unrelated numbers.
One honest limitation. VSEPR predicts the arrangement reliably and does not predict the size of that compression, which needs a proper calculation of where the electrons actually are. The distorted angles here are measured values, shown beside the ideal so you can see how far things have moved. A formula invented to fit water and ammonia would look more impressive and tell you less.
The trigonal bipyramid, and the one rule that generates three shapes
Five domains is the interesting case, because its positions are not all equivalent. Two sit on the axis and three around the equator, and those are genuinely different neighbourhoods.
Count the close contacts. An axial position has three other domains at 90 degrees to it. An equatorial position has only two, with the remaining two at a roomy 120. Since lone pairs are the more repulsive kind, they take the less crowded site, so lone pairs go equatorial.
That single rule produces three shapes without any further reasoning. One lone pair leaves a see-saw. Two leave a T. Three fill every equatorial position, so the only places left for bonds are the two axial ones, and the molecule comes out linear. Xenon difluoride is linear despite having five electron domains, which looks like an exception until you see where the lone pairs had to go.
Polarity is a vector sum, not a lone pair count
A molecule is non-polar when its bond dipoles cancel. That is a statement about directions, so add the bond vectors up and look at what you get. The tool draws that sum as an arrow, and when the sum is zero the arrow simply is not there.
This settles the cases that look like exceptions. Xenon difluoride has three lone pairs and is non-polar, because its two bonds point exactly opposite. Xenon tetrafluoride has two lone pairs and is non-polar, because its four bonds form a square and cancel in pairs. Bromine pentafluoride has only one lone pair and is polar, because five directions around an octahedron cannot cancel however you arrange them.
So the number of lone pairs tells you nothing about polarity on its own. Only the symmetry of the bonds does.
Why anyone cares about the shape
Because shape decides polarity, and polarity decides most of what you can measure. Carbon dioxide is more than twice as heavy as water, yet it turns to gas at minus 78 °C while water boils at 100 °C. Carbon dioxide is linear, so its two polar bonds cancel and the molecules barely attract each other. Water is bent, so its bonds do not cancel, and the result is hydrogen bonding, a liquid at room temperature, and oceans.
Shape also decides how molecules fit together, which is the entire basis of enzymes recognising their substrates and drugs binding their targets.
Common mistakes
- Naming the electron geometry when asked for the shape. Water’s electron geometry is tetrahedral and its shape is bent. Both answers are correct to different questions.
- Forgetting lone pairs when counting domains. The steric number is bonds plus lone pairs. Miss a lone pair and every prediction after it is wrong.
- Putting lone pairs axially in a trigonal bipyramid. They go equatorial, where there are two 90 degree neighbours instead of three.
- Assuming lone pairs make a molecule polar. Xenon difluoride and xenon tetrafluoride both have lone pairs and neither is polar. Check whether the bonds cancel.
- Treating double bonds as two domains. A double bond is one region of electron density and counts once. Carbon dioxide has two domains, not four, which is why it is linear.
- Expecting exact angles from VSEPR. It predicts shapes, not precise angles. Use it to know a molecule is bent, not to claim 104.5 without measuring.
Common questions
What is the difference between electron geometry and molecular geometry?
Electron geometry describes how all the electron domains around the central atom are arranged, counting bonds and lone pairs alike. Molecular geometry describes only the shape formed by the bonds, because those are the atoms you can see. Water and methane are the clearest illustration: both have four electron domains arranged tetrahedrally, and nothing about that arrangement differs between them. Methane uses all four for bonds, so it is tetrahedral. Water uses two for bonds and two for lone pairs, so it is bent. Conflating the two is what makes the topic feel like a list to memorise rather than one idea applied repeatedly.
Why is the bond angle in water 104.5 degrees and not 109.5?
Because a lone pair takes up more room than a bonding pair. A bonding pair is shared between two nuclei and so is held in a relatively narrow region between them, while a lone pair is attached to only one nucleus and spreads wider. It therefore repels neighbouring pairs more strongly, and the bonds get pushed closer together. Methane, with no lone pairs, sits at the ideal tetrahedral 109.5 degrees. Ammonia has one lone pair and closes to about 107. Water has two and closes to 104.5. Each lone pair costs roughly a couple of degrees, and that sequence is worth remembering as a pattern rather than three separate facts.
Why do lone pairs go in equatorial positions in a trigonal bipyramid?
Because equatorial positions are less crowded, and lone pairs are the more repulsive kind so they take the roomier site. Count the close neighbours: an axial position has three other domains at 90 degrees to it, while an equatorial position has only two, with the remaining two at a comfortable 120. Putting the lone pair equatorially therefore minimises the number of strong 90 degree repulsions. This one rule generates three shapes: sulfur tetrafluoride becomes a see-saw, chlorine trifluoride becomes T-shaped, and xenon difluoride becomes linear because all three equatorial sites end up holding lone pairs and only the two axial positions are left for bonds.
How can a molecule with lone pairs be non-polar?
Because polarity is decided by whether the bond dipoles cancel, not by whether lone pairs are present. Xenon difluoride has three lone pairs and is non-polar, because its two bonds point in exactly opposite directions and their dipoles sum to zero. Xenon tetrafluoride has two lone pairs and is also non-polar, because its four bonds form a square whose dipoles cancel in pairs. Bromine pentafluoride, by contrast, has only one lone pair and is polar, because five bonds around an octahedron cannot possibly cancel. The tool draws the vector sum of the bonds as an arrow, and it simply disappears when the sum is zero, which is what non-polar looks like.
Does VSEPR predict bond angles exactly?
No, and it is worth being clear about what it does and does not do. VSEPR predicts the arrangement reliably: it will tell you water is bent and sulfur hexafluoride is octahedral, and it is right. It does not predict by how much a lone pair closes an angle, because that requires calculating electron distributions properly. The angles reported here for the distorted shapes are experimental measurements, shown beside the ideal value so the size of the distortion is visible. A model that invented a formula for the compression and happened to fit water and ammonia would be worse than one that admits the number is measured.
Why does the shape matter?
Because it decides polarity, and polarity decides most of the behaviour you can observe. Whether a molecule has a net dipole determines its boiling point relative to similar molecules, what it dissolves in, and how it interacts with an electric field. Carbon dioxide is more than twice as heavy as water, yet it turns to gas at minus 78 °C while water boils at 100 °C, and the reason is geometry: carbon dioxide is linear so its two polar bonds cancel, while water is bent so its bonds do not. Shape also governs how molecules fit together, which is the whole basis of how enzymes recognise their substrates and how drugs bind their targets.