From a pile of operations to a group
In the previous guide you learned to spot the individual pieces: the rotation axes, the mirror planes, the center of inversion, the sneaky improper rotations. Each symmetry element is a geometric feature, and each move you can make using it — turning, reflecting, inverting — is a symmetry operation that leaves the molecule looking unchanged. Now comes the beautiful step. If you collect every symmetry operation a molecule has, the whole set is not just a bag of tricks. It is a group in the precise mathematical sense — and because all the elements pass through one common point at the molecule's center, it is called a point group.
Why insist on the word group? Because a group is not just any collection — it must obey four rules, and a molecule's operations obey all four automatically. First, doing one operation after another gives you a third operation that is also in the set (closure): reflect water in one plane, then the other, and the result is the same as a 180-degree rotation. Second, there is always a do-nothing operation, the identity E (leave the molecule exactly as it is). Third, every operation has an inverse that undoes it (rotate 120 degrees one way, then 120 degrees back). Fourth, the operations associate. You do not need to prove these every time — they come free with the geometry — but they are the reason the whole machinery of group theory, and the character tables you will meet next, can be brought to bear on chemistry.
Reading the labels: what C2v and Oh actually say
The symbols look cryptic until you crack the code, and then they read almost like words. The naming scheme used in chemistry is the Schoenflies system. The big capital letter tells you the backbone. A plain C with a number, like C2 or C3, means the molecule's only real symmetry is one rotation axis of that order — its single most important axis, the one of highest order, which we call the principal axis. A D means there is a principal Cn axis plus n two-fold axes lying perpendicular to it, like spokes on a wheel. The fancy capitals T, O and I name the high-symmetry shapes: T for tetrahedral, O for octahedral, I for icosahedral.
The small subscripts then tell you which mirror planes are present. A v (for vertical) means a mirror plane that contains the principal axis — you can almost picture slicing down through the axis. An h (for horizontal) means a mirror plane perpendicular to the principal axis, lying flat across it. A d (for dihedral) is a special vertical plane that bisects the angle between two of those perpendicular C2 axes. So C2v reads: a two-fold principal axis plus vertical mirror planes — exactly water's situation. And Oh reads: full octahedral symmetry with the horizontal mirror planes too — the symmetry of SF6 or an octahedral metal complex.
label reads as example ------- -------------------------------------------- ---------- C2v C2 axis + 2 vertical mirror planes H2O C3v C3 axis + 3 vertical mirror planes NH3 D3h C3 axis + 3 perp. C2 + horizontal plane BF3, PF5 D4h C4 axis + 4 perp. C2 + horizontal plane XeF4, PtCl4 2- Td tetrahedral (4 C3, 3 C2, 6 sigma_d, ...) CH4, [Ni(CO)4] Oh octahedral (3 C4, 4 C3, mirrors, i, ...) SF6, [Co(NH3)6]3+
The flowchart: assigning any molecule
You do not eyeball a point group and hope — there is a reliable decision tree, and point-group assignment is just walking it from the top. The trick is to ask a fixed sequence of yes-or-no questions, each one splitting the possibilities in half, until only one label survives. Have a 3-D picture of the molecule in your head (or, honestly, a model kit on the desk) and run the steps in order.
- Is it one of the special high-symmetry shapes? If the molecule is linear, it is C-infinity-v (no inversion center, like HCl) or D-infinity-h (with one, like CO2). If it is tetrahedral, octahedral, or icosahedral, you are done: Td, Oh, or Ih. Otherwise, go on.
- Find the principal axis — the rotation axis of highest order, Cn. If there is no proper rotation axis at all, check for a single mirror plane (Cs) or just an inversion center (Ci); with nothing but the identity it is C1, the trivial group of a lopsided molecule like CHFClBr.
- Are there n two-fold (C2) axes perpendicular to the principal Cn axis? If yes, you are in the D family; if no, you are in the C family. This single question is the great fork in the road.
- Now look for a horizontal mirror plane (perpendicular to the principal axis). If present, add the subscript h: Cnh or Dnh (so BF3 with its flat plane is D3h). If absent, look for vertical/dihedral planes containing the axis: that gives Cnv or Dnd (so NH3 with its three vertical planes but no horizontal one is C3v).
- If there are no mirror planes at all, the last check is for an improper rotation axis S2n collinear with the principal axis (S4n in older notation gives Sn groups); failing even that, you land on the bare Cn or Dn group.
The six groups of inorganic chemistry, made concrete
Let the six workhorse groups settle into your hands one at a time. C2v belongs to bent and similar low-symmetry molecules: water H2O, sulfur dioxide SO2, the bent NO2. One C2 axis runs down the middle, two vertical mirror planes contain it, four operations in all (E, C2, and two reflections). C3v is the umbrella shape — ammonia NH3, the phosphine PH3, a tetrahedral fragment like CHCl3 where one corner differs from the other three. A C3 axis stands up through the apex, three vertical planes splay down between the legs, six operations in all.
Now the D groups, which add a horizontal mirror plane that the C groups lack. D3h is trigonal-planar or trigonal-bipyramidal: boron trifluoride BF3 lying flat as a pancake, the carbonate ion CO3 2-, the five-coordinate PF5. The molecule is the same flipped over, so the principal C3 axis is joined by three perpendicular C2 axes and the flat horizontal mirror, giving twelve operations. D4h is the signature of the square: square-planar [PtCl4]2- and the famously flat XeF4, with a C4 axis up through the center, four perpendicular C2 axes, a horizontal plane, and an inversion center. The square-planar geometry is everywhere in the chemistry of d8 metals like Pt(II), Pd(II) and Ni(II).
Finally the two cubic giants. Td is the full symmetry of the regular tetrahedron: methane CH4, the tetrahedral [Ni(CO)4], the sulfate ion SO4 2-, the perchlorate ClO4-. It has twenty-four operations — four C3 axes through the corners, three C2 axes through the edge-midpoints, and six dihedral mirror planes — but, importantly, no inversion center. Oh is the full symmetry of the octahedron, the richest group most inorganic chemists meet: SF6, and the great family of octahedral complexes like the cobalt hexammine ion [Co(NH3)6]3+. It packs forty-eight operations and does have an inversion center — a fact that will matter enormously when you ask why octahedral complexes show the colors they do.
What the label buys you — and honest warnings
Once you have the point group, it pays dividends without any extra work. Two physical properties read straight off it. A molecule can be polar only if its dipole survives every symmetry operation — which means the point group must have no operation that would flip the dipole onto itself reversed. In practice that allows only the C-type groups (Cn, Cnv, Cs): water (C2v) and ammonia (C3v) are polar, but the highly symmetric BF3 (D3h), XeF4 (D4h), CH4 (Td) and SF6 (Oh) all have their bond dipoles cancelled by symmetry and are nonpolar. This is the rigorous version of the hand-waving you did with shape and polarity back in VSEPR.
The other free gift is chirality. A molecule is chiral — non-superimposable on its mirror image, and so optically active — exactly when its point group contains no improper operation at all: no mirror plane, no inversion center, no improper rotation axis Sn. (The mirror plane is the special case S1 and the inversion center is S2, so really one test covers them all.) That is why a tris-chelate octahedral complex like [Co(en)3]3+, which keeps only rotation axes and falls in point group D3, comes in left- and right-handed forms — its lack of any improper element is the deep reason behind the optical isomerism you met in coordination chemistry.
Why this matters for the rest of the rung
It would be a waste to learn point groups just to slap a label on a molecule. The real payoff is that the point group is the doorway to the quantitative tools of the rest of this rung. Every point group has a fixed roster of ways that things — orbitals, vibrations, electron motions — can behave under its operations, and those behaviors are tabulated once and for all in a character table. The next guide opens that table and shows how to read it.
From there the dominoes fall. Knowing the point group of a complex lets you sort its d orbitals into symmetry sets — which is exactly why, in an octahedral field, the five d orbitals split into a lower set of three (those pointing into the gaps between ligands) and an upper set of two (those aimed straight at the ligands). It tells you which vibrations will show up in an infrared spectrum and which in a Raman spectrum. And it is the reason the inversion center hiding inside the Oh group enforces the Laporte rule, the selection rule that makes pure d-d transitions weak and octahedral complexes only faintly colored. Group theory is the hidden grammar; the point group is the first sentence you learn to parse.