Kasha's Rule: Why Emission Comes from the Lowest Excited State

The rule in one sentence—and what it actually claims

Kasha's rule states that appreciable photon emission (fluorescence or phosphorescence) occurs only from the lowest excited state of a given spin multiplicity. In Michael Kasha's own 1950 phrasing: "the emitting level of a given multiplicity is the lowest excited level of that multiplicity." For a typical closed-shell organic molecule that means fluorescence originates from S₁ (the lowest excited singlet) and phosphorescence from T₁ (the lowest triplet), regardless of which higher state—S₂, S₃, T₂—you initially populated.

The word appreciable matters. Kasha's rule is a statement about branching ratios, not an absolute prohibition. Emission from S₂ or T₂ is not forbidden by any symmetry or spin selection rule; it is simply outcompeted. Upper states drain nonradiatively to S₁ or T₁ so fast (picosecond timescale) that the radiative channel from those upper states—which needs nanoseconds—almost never fires. The rule is therefore fundamentally kinetic, a race between radiative and nonradiative rate constants, not thermodynamic.

A crucial corollary is that the emission spectrum is independent of the excitation wavelength. Pump a fluorophore anywhere in its absorption envelope—into S₁, S₂, or S₅—and you recover an identical fluorescence band shape and position. This constancy of the emission band shape follows directly from Kasha's rule: emission always issues from the same relaxed S₁ level. The closely related statement that the quantum yield is likewise independent of excitation wavelength is specifically Vavilov's rule. Both are direct experimental fingerprints of the same underlying photophysics.

The kinetic machinery: internal conversion and vibrational relaxation win the race

Consider the fate of a molecule promoted to a vibrationally hot level of, say, S₂. Two fast nonradiative processes act first. Vibrational relaxation dissipates excess vibrational energy to the surrounding solvent bath on a timescale of ~10⁻¹²–10⁻¹³ s, dropping the molecule to the vibrationless (v=0) level of S₂. Then internal conversion (IC)—a horizontal, isoenergetic, spin-conserving crossing between electronic surfaces—carries S₂(v=0) into a high vibrational level of S₁, again in ~10⁻¹²–10⁻¹³ s. These relaxations repeat down the manifold until the molecule reaches S₁(v=0).

Now compare rate constants. Radiative decay from S₁ has a rate constant k_r ~ 10⁷–10⁹ s⁻¹ (fluorescence lifetimes of nanoseconds). Internal conversion between upper states runs at k_IC ~ 10¹²–10¹³ s⁻¹. The branching ratio for emission from an upper state is roughly k_r/(k_r + k_IC) ≈ 10⁸/10¹² = 10⁻⁴. So even though S₂ could fluoresce, only about one molecule in ten thousand does before IC empties the state. That factor of ~10⁴ is Kasha's rule made quantitative.

Why is IC so much faster between upper states than between S₁ and S₀? The answer is the energy gap law. The nonradiative rate scales with the Franck–Condon-weighted density of accepting vibrational states, which falls off exponentially with the electronic energy gap: k_IC ∝ exp(−γ·ΔE), where γ is a molecule-specific parameter. Upper electronic states are closely spaced (gaps often <5,000 cm⁻¹), so their vibrational manifolds overlap densely and IC is ultrafast. The S₁–S₀ gap, by contrast, is large (often >15,000 cm⁻¹), the Franck–Condon overlap is tiny, and S₁→S₀ internal conversion slows to ~10⁷–10⁹ s⁻¹—finally slow enough that radiative decay can compete. That is precisely why S₁ is the level that emits.

Reading it off a Jablonski diagram—and a worked estimate

The Jablonski diagram is the visual bookkeeping of Kasha's rule. Vertical solid arrows denote radiative transitions (absorption up, fluorescence/phosphorescence down); horizontal squiggly arrows denote nonradiative transitions (internal conversion within a multiplicity, intersystem crossing between multiplicities). Kasha's rule is the observation that, no matter where the upward absorption arrow terminates, all the downward radiative arrows start from either S₁ or T₁.

Let's put numbers on a representative dye. Suppose S₁ has k_r = 2×10⁸ s⁻¹ and a total nonradiative rate (IC to S₀ plus intersystem crossing to T₁) of k_nr = 3×10⁸ s⁻¹. The fluorescence quantum yield is Φ_F = k_r/(k_r + k_nr) = 2/(2+3) = 0.40, and the observed lifetime is τ = 1/(k_r + k_nr) = 1/(5×10⁸ s⁻¹) = 2.0 ns—entirely typical for a good fluorophore. Now put the same k_r on the S₂ state, but let its internal conversion to S₁ run at k_IC = 5×10¹² s⁻¹. The S₂ emission yield is Φ(S₂) = 2×10⁸/(2×10⁸ + 5×10¹²) ≈ 4×10⁻⁵. Four orders of magnitude weaker—invisible against the S₁ band.

The same logic applies to the triplet manifold. After intersystem crossing populates some upper triplet Tₙ, ultrafast triplet internal conversion funnels it to T₁ before it can phosphoresce, so phosphorescence is a T₁→S₀ event. This is why organic phosphors emit at a single, red-shifted, long-lived band (microseconds to seconds) characteristic of the T₁ energy, independent of how the triplet manifold was first accessed.

Why it matters: spectroscopy, quantum yield, and the mirror-image rule

Kasha's rule is the load-bearing assumption behind almost all of quantitative fluorescence spectroscopy. Because emission always starts from a single, thermally relaxed level (S₁, v=0), the fluorescence spectrum is a stable molecular signature: it does not drift with excitation wavelength, so a fluorophore has one characteristic emission band you can use for identification, sensing, and imaging. Every calibration of a fluorimeter, every FRET measurement, every single-molecule trace tacitly relies on this constancy.

The rule also underpins the mirror-image relationship between absorption and fluorescence. Since absorption is a vertical transition from S₀(v=0) to various vibrational levels of S₁, and fluorescence is a vertical transition from the relaxed S₁(v=0) down to various vibrational levels of S₀, the two spectra sample the same set of vibrational spacings from opposite ends. When the ground- and excited-state geometries and vibrational frequencies are similar, the emission band is an approximate mirror image of the lowest absorption band, offset by the Stokes shift—the energy lost to vibrational relaxation in both states. Perylene and fluorescein show textbook mirror symmetry; anthracene shows only approximate mirror symmetry, since a modest geometry change between S₀ and S₁ distorts the correspondence.

Finally, the rule clarifies what quantum yield means and why it is excitation-independent. Because upper-state relaxation to S₁ is essentially unit-efficient (Vavilov's rule), Φ_F reflects only the competition among decay channels out of S₁: radiative decay, S₁→S₀ internal conversion, and intersystem crossing to T₁. Engineering a bright emitter is therefore an exercise in maximizing k_r and suppressing the nonradiative pathways from S₁—rigidifying the chromophore, removing low-frequency accepting modes, and pushing up the S₁–S₀ gap all help, again via the energy gap law.

When Kasha's rule breaks: azulene and the anti-Kasha emitters

Kasha's rule is a rule, not a theorem, and its most celebrated violator is azulene, the deep-blue bicyclic isomer of naphthalene (a fused five-membered/seven-membered bicyclic—a cyclopentadiene-type ring fused to a cycloheptatriene-type ring—and a nonalternant aromatic). Azulene fluoresces from S₂, not S₁—an anti-Kasha emission first noted by M. Beer and H. C. Longuet-Higgins in 1955. The reason is a break in the assumptions of the energy gap law: azulene has an anomalously large S₂–S₁ gap of about 14,000 cm⁻¹ (~1.7 eV), while its S₁–S₀ gap is unusually small (~14,000 cm⁻¹ as well). The large S₂–S₁ separation throttles S₂→S₁ internal conversion (small Franck–Condon overlap), so S₂ survives long enough to radiate; meanwhile the tiny S₁–S₀ gap makes S₁→S₀ internal conversion ultrafast, quenching any S₁ fluorescence. Direct evidence: azulene's S₁ lifetime is only ~1–2 ps (about 1.9 ps), far too short for radiative decay to compete.

Recent work (Ottosson and co-workers, J. Am. Chem. Soc. 2023) sharpened the picture using excited-state aromaticity: azulene's S₁ is antiaromatic (obeying Baird's rule for triplet/excited-singlet aromaticity) with an easily accessible antiaromaticity-relief distortion that opens a fast nonradiative funnel, while S₂ is comparatively aromatic and stable—both effects reinforcing anti-Kasha behavior. Other confirmed exceptions include certain thioketones and thiones (S₂ fluorescence), some porphyrin and metalloporphyrin systems showing S₂ emission, and engineered anti-Kasha organic dyes designed with deliberately large upper-state gaps.

• Large upper-state gap — slows Sₙ→Sₙ₋₁ IC so the upper state can compete radiatively (azulene, thioketones).
• Small S₁–S₀ gap — accelerates S₁→S₀ IC, killing the normal S₁ emission channel (azulene again).
• Symmetry-forbidden lower transition — if S₁→S₀ is dipole-forbidden but S₂→S₀ is allowed, the upper state can dominate emission despite Kasha.

History, scope, and the fine print

Michael Kasha (1920–2013), a physical chemist trained under G. N. Lewis at Berkeley and later a founder of the Institute of Molecular Biophysics at Florida State, articulated the rule in Discussions of the Faraday Society 9, 14–19 (1950), in the paper "Characterization of electronic transitions in complex molecules." It arrived alongside the broader mid-century development of the Jablonski diagram (Aleksander Jabłoński, 1933) and Sergey Vavilov's earlier observations that quantum yield is excitation-wavelength independent. Kasha's contribution was to unify these into a predictive, mechanistic generalization about which state emits and why.

The rule's domain of validity is important. It holds superbly for medium-sized and large organic molecules in condensed phases (solution, glasses, solids), where a dense vibrational manifold and an efficient solvent bath guarantee ultrafast relaxation to S₁ or T₁. It weakens for: (i) small molecules and atoms in the gas phase, which have sparse vibrational level structure and can radiate from upper states; (ii) molecules with anomalously large upper-state energy gaps (the azulene class); and (iii) cases where the lowest transition is symmetry- or spin-forbidden, removing S₁'s radiative advantage.

It is worth stressing what Kasha's rule is not. It is not a selection rule—no transition is forbidden by it. It is not thermodynamic—the emitting state is kinetically selected, not the free-energy minimum among excited states in any equilibrium sense. And it is not a statement that upper states are dark; they simply funnel their energy downward faster than they can shine. Understood as a race between k_r (~10⁸ s⁻¹) and k_IC (~10¹² s⁻¹), Kasha's rule is one of the most reliable and mechanistically transparent generalizations in all of molecular photophysics.

Why upper-state emission (Sₙ, n≥2) loses to the lowest excited state S₁

Property	Upper state Sₙ (n≥2)	Lowest excited state S₁
Sₙ–Sₙ₋₁ energy gap	Small (typ. <5,000 cm⁻¹)	Large S₁–S₀ gap (typ. >10,000 cm⁻¹)
Internal conversion rate	Fast, ~10¹²–10¹³ s⁻¹	Slow, ~10⁷–10⁹ s⁻¹
Franck–Condon vibrational overlap	Large (dense, near-resonant levels)	Small (energy gap law suppresses it)
Competes with radiative decay?	No — nonradiative wins by ~10³–10⁴×	Yes — k_r ~ 10⁸ s⁻¹ competes
Observed emission yield	Negligible (Φ ≈ 10⁻⁴ or less)	Appreciable (Φ up to ~1)