Molecular And Electron Geometry Table

How to Actually Use a Molecular And Electron Geometry Table Without Confusing Yourself

Most people treat these tables like cheat sheets, but they don't really work that way unless you understand what you're looking at. I used to hand students a printed one and tell them to memorize it. That approach failed about eighty percent of the time because nobody actually retained which shape corresponds to how many lone pairs. What actually works is understanding the mapping logic so you can reconstruct the table from scratch when you forget it.

Electron geometry and molecular geometry come from the same VSEPR framework but answer different questions. Electron geometry describes the arrangement of all electron domains around the central atom—bonding pairs, lone pairs, single bonds, double bonds, triple bonds, everything. Molecular geometry describes only the positions of the atoms themselves, ignoring where the lone pairs sit. That distinction is where most mistakes happen. Here is the actual table I reference when I need to look something up. This is the complete mapping you need, organized by the number of electron domains: 2 electron domains: Electron geometry is linear. If both domains are bonding pairs, molecular geometry is also linear. Example: BeCl, CO.
3 electron domains: Electron geometry is trigonal planar. With three bonding pairs and zero lone pairs, molecular geometry stays trigonal planar. With two bonding pairs and one lone pair, it becomes bent or V-shaped with an ideal angle of 120 degrees. Example: SO for the bent case.
4 electron domains: Electron geometry is tetrahedral. Four bonding pairs gives tetrahedral molecular geometry like CH. Three bonding plus one lone pair gives trigonal pyramidal like NH. Two bonding plus two lone pairs gives bent with an angle closer to 104.5 degrees, as in HO.
5 electron domains: Electron geometry is trigonal bipyramidal. Five bonding pairs gives trigonal bipyramidal like PCl. Four bonding plus one lone pair gives seesaw shape, as in SF. Three bonding plus two lone pairs gives T-shaped like ClF. Two bonding plus three lone pairs gives linear like I.
6 electron domains: Electron geometry is octahedral. Six bonding pairs gives octahedral like SF. Five bonding plus one lone pair gives square pyramidal like BrF. Four bonding plus two lone pairs gives square planar like XeF.

That is the whole table. Eight distinct molecular geometries across five electron geometries. The pattern you should notice is that molecular geometry always matches electron geometry when there are zero lone pairs, and diverges from it only when lone pairs are present. Lone pairs occupy more space than bonding pairs, which is why bond angles compress below their ideal values. I ran into a real problem with this last year when a student was trying to determine the geometry of XeOF and kept getting the wrong answer because the standard table assumes all bonding domains are equivalent. In XeOF, you have four bonding pairs to fluorine atoms, one bonding domain to oxygen (a double bond), and one lone pair on xenon. That is six electron domains total, so the electron geometry is octahedral. But the double bond to oxygen exerts greater repulsion than the single bonds to fluorine, which distorts the square pyramidal molecular geometry slightly. The lone pair sits opposite the oxygen atom, and the F-Xe-F angles in the basal plane compress a bit from the ideal 90 degrees. The table won't tell you that. You have to apply the refined VSEPR priority order: lone pair-lone pair repulsion is strongest, followed by lone pair-bonding pair, then bonding pair-bonding pair. Double bonds count as a single domain for geometry classification but still exert stronger repulsion than single bonds when angles are being predicted. The shortcut most people miss is that you don't actually need to memorize the table if you can count domains correctly. The first step that breaks for everyone is counting electron domains. A double bond counts as one domain, not two. A triple bond counts as one domain. A lone pair counts as one domain. An unpaired electron counts as one domain. You draw the Lewis structure first, count everything around the central atom, and the domain count tells you the electron geometry directly. Then you remove the lone pairs from the picture and name the shape formed by the atoms only.

Here is a nuance that textbooks rarely emphasize. Hypervalent molecules like SF or XeF are perfectly valid within the VSEPR model, but the underlying quantum chemistry is more complicated than introductory courses suggest. The d-orbital participation that older textbooks claimed was responsible for expanded octets is largely considered a myth now. Modern computational chemistry shows that hypervalency is better explained through three-center four-electron bonding and polar covalent interactions rather than d-orbital hybridization. The geometry predictions remain accurate, but if someone tells you that sp³d² hybridization is the physical reason SF is octahedral, that explanation is outdated. The geometries are correct regardless. The justification is not. Another common pitfall is assuming molecular geometry alone determines polarity without checking the symmetry of the atomic arrangement. BF is trigonal planar and nonpolar because the three identical B-F bonds are symmetrically arranged at 120 degrees. NF is also trigonal pyramidal in shape, but it is polar because the lone pair creates an asymmetric charge distribution. Same basic shape family, different polarity outcomes because of that lone pair. The table will not tell you this. You have to visualize the dipole vectors. There is also a practical limitation to keep in mind. The Molecular And Electron Geometry Table approach breaks down for transition metal complexes, large clusters, and molecules where steric effects dominate over electronic effects. For example, bulky ligands around a metal center can force geometries that deviate significantly from what VSEPR predicts. If you are working with organometallic compounds or coordination chemistry, VSEPR-based geometry prediction is essentially useless and you need crystal field theory or ligand field theory instead. The table covers main group elements fairly well. Beyond that, it stops being reliable.

If you want a downloadable version of the table, most general chemistry textbooks include it as a reference chart. You can also generate your own from any VSEPR calculator tool online by inputting common molecules and recording the results. That process of building the table yourself actually reinforces the domain-counting logic better than printing one out and highlighting it.

Quick Reference for Common Molecules

CO: 2 domains, 0 lone pairs. Linear/linear.
SO: 3 domains, 1 lone pair. Trigonal planar/bent.
CH: 4 domains, 0 lone pairs. Tetrahedral/tetrahedral.
NH: 4 domains, 1 lone pair. Tetrahedral/trigonal pyramidal.
HO: 4 domains, 2 lone pairs. Tetrahedral/bent.
PF: 5 domains, 0 lone pairs. Trigonal bipyramidal/trigonal bipyramidal.
SCl: 5 domains, 1 lone pair. Trigonal bipyramidal/seesaw.
BrF: 5 domains, 2 lone pairs. Trigonal bipyramidal/T-shaped.
I: 5 domains, 3 lone pairs. Trigonal bipyramidal/linear.
SF: 6 domains, 0 lone pairs. Octahedral/octahedral.
BrF: 6 domains, 1 lone pair. Octahedral/square pyramidal.
XeF: 6 domains, 2 lone pairs. Octahedral/square planar. The method takes about thirty seconds per molecule once you are comfortable with Lewis structures. The hardest part is always getting the Lewis structure right, particularly for ions and molecules with formal charges. Get the Lewis structure wrong and every step after that is wrong too. I still see people drawing incorrect Lewis structures for sulfate and phosphate ions, then wondering why their geometry comes out wrong. The table is only as good as the input you put into it.