Same delta, wildly different brightness
In the previous guide you learned to read a complex's color as a map of its [[d-d-transition|d-d transitions]] — an electron leaping from a t2g orbital up to an eg orbital across the gap delta-o, absorbing exactly the photon whose energy matches that gap, so the complex wears the [[complementary-color|complementary color]] of the light it swallows. That tells you *where* a band sits on the spectrum: its wavelength. But hold two real spectra side by side and a second question jumps out. The pale-violet hexaaquatitanium ion [Ti(H2O)6]3+ has its band almost exactly where the deep-purple permanganate [MnO4]- has one — yet permanganate is so intensely colored that a single crystal tints a whole beaker, while you need a concentrated solution of the titanium complex to see any color at all.
The wavelength of a band is set by *how far* an electron must jump. Its intensity — how much light it actually drinks — is set by something different: *how easy* the jump is. Chemists measure that ease with the molar absorptivity, the symbol epsilon, the constant in the Beer-Lambert law that says absorbance equals epsilon times concentration times path length. A faint d-d band in a typical octahedral complex has epsilon of only a few to a few tens; an intense charge-transfer band like permanganate's runs into the thousands or tens of thousands — a thousandfold gap. Those numbers are not random. They are dictated by two selection rules: pass them and the transition is bright, break them and it is forbidden and dim. Understanding why those rules exist, and how complexes cheat them, is the whole of this guide.
The Laporte rule: parity must flip
The first and stricter brake on d-d intensity is the [[laporte-selection-rule|Laporte selection rule]], also called the parity rule. To feel it, you need one idea from the symmetry rung: a center of inversion. A molecule has a [[center-of-inversion|center of inversion]] if, from a point in the middle, every atom has an identical twin straight through on the opposite side. A regular octahedron has one — the metal sits at the center, and each ligand is mirrored by the ligand directly across from it. Such a molecule is called centrosymmetric, and in it every orbital carries a strict parity label: gerade (g, German for 'even') if it looks unchanged when you invert it through the center, ungerade (u, 'odd') if inverting flips its sign. All five d orbitals are gerade — they are symmetric under inversion. The p orbitals are ungerade — invert a p orbital and its two lobes swap sign.
Now the rule itself: light is absorbed when a photon's electric field shoves the electron cloud lopsided to one side — a so-called electric-dipole transition. The mathematics of that shove demands that the parity of the electron's home must *flip* during the jump: a g orbital can only absorb its way into a u orbital, and vice versa. In shorthand, g-to-u and u-to-g are allowed; g-to-g and u-to-u are Laporte-forbidden. A d-d transition is the textbook g-to-g case — both the starting t2g and the destination eg are made of gerade d orbitals — so in a perfect, centrosymmetric octahedron the d-d transition is forbidden, and its band is weak. This is the deep reason most octahedral complexes wear only gentle, dilute colors.
The spin rule: don't flip the spin
Laporte is not the only gatekeeper. The second is the [[spin-selection-rule|spin selection rule]], and it is even simpler to state: during an electronic transition the total number of unpaired electrons must not change. In symbols, delta-S equals zero — the spin multiplicity is conserved. The photon's electric field pushes on an electron's *position*; it has essentially no handle on the electron's *spin*. So an electron may jump to a new orbital, but it must keep the spin it had. A transition that would require an electron to flip its spin on the way up is spin-forbidden, and that costs you another large factor in intensity, layered on top of any Laporte penalty.
The most beautiful demonstration is the pale, ghostly pink of a manganese(II) salt — solid MnCl2, or the [Mn(H2O)6]2+ ion in solution. Mn2+ is a high-spin d5 ion: every one of the five d orbitals holds a single electron, all five spins aligned the same way (this is the high-spin arrangement you met when delta was smaller than the pairing energy, in the high-spin / low-spin choice). To move any electron into the upper eg set, it would have to land in an orbital that already contains an electron of the opposite spin — so it must pair up, which means flipping a spin. *Every possible* d-d transition for high-spin d5 is therefore spin-forbidden as well as Laporte-forbidden. Both rules broken at once is why Mn2+ salts are the palest of all — epsilon around 0.01 to 0.1, a hundred times fainter than an ordinary Laporte-forbidden band.
Intensity ladder (rough molar absorptivity epsilon, L/mol/cm): spin- AND Laporte-forbidden d-d ~0.01 - 1 e.g. [Mn(H2O)6]2+ (palest) Laporte-forbidden d-d (Oh) ~1 - 50 e.g. [Ti(H2O)6]3+ (pale) Laporte-relaxed d-d (Td, no centre) ~50 - 500 e.g. [CoCl4]2- (deep) fully allowed charge transfer ~1000 - 50000 e.g. [MnO4]- (vivid) Each broken rule buys roughly 1-3 orders of magnitude of brightness.
Cheating Laporte: vibrations and lost symmetry
If d-d transitions in an octahedron are Laporte-forbidden, why do ordinary octahedral complexes like the rose-pink [Co(H2O)6]2+ or violet [Ti(H2O)6]3+ show *any* color? Because the rule has a loophole, and the loophole is motion. Molecules are never frozen still; they constantly bend and stretch through their vibrations. Some of those vibrations are themselves ungerade — they momentarily destroy the center of inversion, lifting the metal off-center or buckling the octahedron into a lopsided shape for an instant. In that fleeting asymmetric moment the metal's d orbitals can mix a whisper of p character into themselves, and p orbitals are ungerade. The d-d transition borrows that breath of u-ness and partly escapes the parity ban. This rescue is called [[vibronic-coupling|vibronic coupling]] — 'vibronic' fusing vibrational and electronic — and it is what gives every centrosymmetric complex its soft, real color instead of none.
Vibronic coupling is a part-time cheat: it only works during those asymmetric wobbles, so it lets a forbidden band through faintly rather than fully — hence epsilon in the range of a few to a few tens, still a hundred times below an allowed band. But it also makes a testable prediction. Cool the complex down and the atoms vibrate less, the symmetry-breaking wobbles weaken, and the color *fades*. Many d-d bands genuinely lose intensity when chilled toward absolute zero — a direct fingerprint that vibration, not the static structure, is doing the smuggling. A static rule explained by a dynamic loophole, confirmed by watching the color dim in the cold: this is the model earning its keep.
Why tetrahedral complexes blaze
Now comes the payoff that ties the rung together. Recall from the geometry guide that a tetrahedron has no center of inversion — there is no atom sitting straight across the middle from each ligand, because a tetrahedron's four corners do not come in opposite pairs. With no inversion center, the labels gerade and ungerade simply cannot be defined, so there is nothing for the Laporte rule to forbid. On top of that, in a tetrahedral field the metal's d orbitals are forced to mix permanently with the metal's p orbitals (they share the same symmetry label there), lacing genuine, full-time u character into the d-based orbitals. The Laporte ban is not merely cheated part-time as in the octahedron — it is structurally lifted. The result is dramatic: a tetrahedral d-d band runs roughly a hundred times more intense than the corresponding octahedral one.
You can see this with the naked eye in a single element. Cobalt(II) chloride in water is the pale rose [Co(H2O)6]2+ — six water ligands, octahedral, centrosymmetric, Laporte-forbidden, faint. Add concentrated hydrochloric acid and the water is driven off to give the [[tetrahedral-field-splitting|tetrahedral]] [CoCl4]2- ion, and the solution flares an intense, deep sapphire blue — the very color change in those humidity-indicator cards and in cobalt-glass. Both are d7 cobalt with d-d transitions; the octahedral one is dim because Laporte forbids it, the tetrahedral one is brilliant because the missing inversion center sets Laporte aside. (Part of the shift to blue is also that delta-t is smaller than delta-o, moving the band — but the leap in *brightness* is pure selection-rule physics.)
One honest caveat keeps this from being a fairy tale. The deepest, most saturated inorganic colors of all — permanganate, dichromate, the blood-red iron(III) thiocyanate — are *not* d-d bands at all, no matter how cleverly relaxed. They are [[charge-transfer-band|charge-transfer bands]], in which an electron leaps bodily from a ligand to the metal or back, rather than shuffling among the metal's own d orbitals. Charge transfer changes the orbital's parity outright and keeps the spin, so it obeys *both* selection rules and is fully allowed, with epsilon in the thousands. So the full ranking of intensity is a three-step staircase: faint forbidden octahedral d-d, then the hundredfold-brighter relaxed tetrahedral d-d, and far above both, the blazing fully-allowed charge-transfer band — the subject of the next guide.