64
2 Molecular States
It is evident from Fig. 2.8 that electronic excitation of a monoalkene may easily
lead to isomerization: in fact, when the molecule goes back to the ground state from
a geometry close to the transition state, it follows the S 0 potential energy surface
ending with either a trans or cis final geometry, independently of the starting isomer
on S 0 .
The first band in the absorption spectrum of monoalkenes has mixed character
π → π
∗ and Rydberg and peaks around 180–200 nm. It is very broad, given the
different shapes of the S 0 and S 1 potential energy surfaces, as shown in Fig. 2.8.
The energetic separation between two π → π
∗ T 1 and S 1 states is large: in fact,
given the substantial overlap between |π | and |π
∗
|, the exchange integral K ππ ∗ is
not small, except when the double bond is twisted. The T 1 potential energy surface
has a behavior similar to S 1 with respect to the torsion around the double bond axis
and is closer to S 0 : population of the T 1 state, when feasible, may lead to cis–trans
isomerization as well.
2.6.5 Excited States n → π ∗
In nonconjugated molecules like imines, carbonyl compounds, azocompounds,
where the double bond involves an atom with a lone pair, the frontier orbitals are
n and π
∗ , so the first excited states have n → π
∗ character. There are significant
differences between n → π
∗ and π → π
∗ states. In particular, the singlet–triplet
energetic separation is smaller for n → π
∗ states than for π → π
∗ states. In fact, n
and π
∗ orbitals occupy quite different space regions, so that concerning the exchange
integrals we have K nπ ∗ K ππ ∗ . This has an important consequence on the photophysics, since with a smaller energy difference, the ISC S 1 → T 1 is easier. Therefore,
systems with n → π
∗ transitions have larger triplet quantum yield with respect to
molecules with π → π
∗ states only. For example, in acetone the n → π
∗ states S 1
and T 1 have minima at about 3.8 and 3.5 eV above the ground state, respectively.
Now, the spin–orbit coupling between S 1 and T 1 is very small because of the El-Sayed
rules (about 1 cm
−1 ), see Sect. 2.4. However, the π → π
∗ state T 2 , with minimum at
about 4.5 eV above S 0 , is not far from S 1 and the SO coupling S 1 /T 2 is not negligible
(about 60 cm
−1 ). As a result, the triplet quantum yield of acetone is close to 1 (at
least at excitation energies below the threshold for C-C bond dissociation, which is
a process in competition with ISC).
Another remarkable difference is due to the fact that, for symmetry reasons, the
μ nπ ∗ is zero or close to zero. Hence the radiative transition from S 0 is (almost) dipole
forbidden for n → π
∗ states (see Eq. (2.94)), at variance with the π → π
∗ states.
Let us consider, for example, a carbonyl compound: the C=O group has a local
C 2v symmetry, with n and π
∗ belonging to b 1 and b 2 irreducible representations, so
μ nπ ∗ = 0 (see also Sect. 3.7). The C 2v symmetry may be disrupted by the groups
linked to the carbonyl; however, μ nπ ∗ will be small, as n and π
∗ are localized on
the C=O, for a nonconjugated carbonyl compound. In the absorption spectrum the
2 Molecular States
It is evident from Fig. 2.8 that electronic excitation of a monoalkene may easily
lead to isomerization: in fact, when the molecule goes back to the ground state from
a geometry close to the transition state, it follows the S 0 potential energy surface
ending with either a trans or cis final geometry, independently of the starting isomer
on S 0 .
The first band in the absorption spectrum of monoalkenes has mixed character
π → π
∗ and Rydberg and peaks around 180–200 nm. It is very broad, given the
different shapes of the S 0 and S 1 potential energy surfaces, as shown in Fig. 2.8.
The energetic separation between two π → π
∗ T 1 and S 1 states is large: in fact,
given the substantial overlap between |π | and |π
∗
|, the exchange integral K ππ ∗ is
not small, except when the double bond is twisted. The T 1 potential energy surface
has a behavior similar to S 1 with respect to the torsion around the double bond axis
and is closer to S 0 : population of the T 1 state, when feasible, may lead to cis–trans
isomerization as well.
2.6.5 Excited States n → π ∗
In nonconjugated molecules like imines, carbonyl compounds, azocompounds,
where the double bond involves an atom with a lone pair, the frontier orbitals are
n and π
∗ , so the first excited states have n → π
∗ character. There are significant
differences between n → π
∗ and π → π
∗ states. In particular, the singlet–triplet
energetic separation is smaller for n → π
∗ states than for π → π
∗ states. In fact, n
and π
∗ orbitals occupy quite different space regions, so that concerning the exchange
integrals we have K nπ ∗ K ππ ∗ . This has an important consequence on the photophysics, since with a smaller energy difference, the ISC S 1 → T 1 is easier. Therefore,
systems with n → π
∗ transitions have larger triplet quantum yield with respect to
molecules with π → π
∗ states only. For example, in acetone the n → π
∗ states S 1
and T 1 have minima at about 3.8 and 3.5 eV above the ground state, respectively.
Now, the spin–orbit coupling between S 1 and T 1 is very small because of the El-Sayed
rules (about 1 cm
−1 ), see Sect. 2.4. However, the π → π
∗ state T 2 , with minimum at
about 4.5 eV above S 0 , is not far from S 1 and the SO coupling S 1 /T 2 is not negligible
(about 60 cm
−1 ). As a result, the triplet quantum yield of acetone is close to 1 (at
least at excitation energies below the threshold for C-C bond dissociation, which is
a process in competition with ISC).
Another remarkable difference is due to the fact that, for symmetry reasons, the
μ nπ ∗ is zero or close to zero. Hence the radiative transition from S 0 is (almost) dipole
forbidden for n → π
∗ states (see Eq. (2.94)), at variance with the π → π
∗ states.
Let us consider, for example, a carbonyl compound: the C=O group has a local
C 2v symmetry, with n and π
∗ belonging to b 1 and b 2 irreducible representations, so
μ nπ ∗ = 0 (see also Sect. 3.7). The C 2v symmetry may be disrupted by the groups
linked to the carbonyl; however, μ nπ ∗ will be small, as n and π
∗ are localized on
the C=O, for a nonconjugated carbonyl compound. In the absorption spectrum the