14 Luminescent Crystal–Control of Excited-State …
287
14.5.5 Theoretical Study on Polymorph-Dependent
Luminescence
14.5.5.1 Accuracy of TD-DFT Calculations
In order to obtain a deeper understanding of the polymorph-dependent ESIPT luminescence of 2, the ESIPT emission energies were evaluated using quantum chemical
calculations. First, a series of computations, TD-DFT, CASSCF, and MS-CASPT2,
were performed on the IPT form of the compound to validate the accuracies of the
computed S 1 –S 0 vertical transition energies. It was concluded that the TD-DFT level
geometry optimization was sufficient to obtain an appropriate geometry of 2, and
the Coulomb-attenuated functional CAM-B3LYP [91] was in quantitative agreement
with the S 1 –S 0 gap obtained from the MS-CASPT2 calculations, whereas the B3LYP
functional substantially underestimated the S 1 –S 0 gap.
14.5.5.2 Calculation of ESIPT Luminescence Energy
Two-layer ONIOM models [92] were applied to the molecular clusters extracted
from 2-Y, 2-O, and 2-R in order to calculate the ESIPT state in different molecular
aggregates. The effects of packing mode on ESIPT were taken into account through
QM:MM calculations within the finite model. Each cluster model consisted of 17
molecules, in which one IPT species (2 IPT ) was surrounded with sixteen enol species
(2 Enol ). The S 1 state of the center 2 IPT was geometrically optimized at the TDDFT(B3LYP)/6-31G(d) level of theory with fixed geometries of the surrounding
2 Enol molecules calculated at the semiempirical PM3 level.
The single point S 1 → S 0 emission energies were calculated at the TD(CAMB3LYP)/6-31G(d). In the case of isolated monomer, optimized 2 IPT without
surrounding enols, the order of the computed emission energies (2-O < 2-R < 2-Y) did
not coincide with the observed ESIPT luminescence (2-R < 2-O < 2-Y) (Table 14.3).
On the other hand, when the π-stacked dimer geometries consisted of the optimized
Table 14.3 S 1 –S 0 vertical energy (nm) of 2 IPT for optimized geometry at TD(B3LYP)/631G(d):PM3 (1 2 IPT +16 2 Enol ) ONIOM model
Isolated model a
ONIOM model
Measured
Monomer
Dimer
Monomer
Dimer
2-Y
548
541.8
506.8
540.8
505.9
2-O
570
592.8
519.5
590.7
518.3
2-R
585
557.0
587.8
555.8
586.4
Non-stacked dimer (2-O)
–
–
567.4
–
565.1
Mutai et al. [89]—Reproduced by permission of The Royal Society of Chemistry
a Geometries extracted from the optimized 17-molecule ONIOM model
287
14.5.5 Theoretical Study on Polymorph-Dependent
Luminescence
14.5.5.1 Accuracy of TD-DFT Calculations
In order to obtain a deeper understanding of the polymorph-dependent ESIPT luminescence of 2, the ESIPT emission energies were evaluated using quantum chemical
calculations. First, a series of computations, TD-DFT, CASSCF, and MS-CASPT2,
were performed on the IPT form of the compound to validate the accuracies of the
computed S 1 –S 0 vertical transition energies. It was concluded that the TD-DFT level
geometry optimization was sufficient to obtain an appropriate geometry of 2, and
the Coulomb-attenuated functional CAM-B3LYP [91] was in quantitative agreement
with the S 1 –S 0 gap obtained from the MS-CASPT2 calculations, whereas the B3LYP
functional substantially underestimated the S 1 –S 0 gap.
14.5.5.2 Calculation of ESIPT Luminescence Energy
Two-layer ONIOM models [92] were applied to the molecular clusters extracted
from 2-Y, 2-O, and 2-R in order to calculate the ESIPT state in different molecular
aggregates. The effects of packing mode on ESIPT were taken into account through
QM:MM calculations within the finite model. Each cluster model consisted of 17
molecules, in which one IPT species (2 IPT ) was surrounded with sixteen enol species
(2 Enol ). The S 1 state of the center 2 IPT was geometrically optimized at the TDDFT(B3LYP)/6-31G(d) level of theory with fixed geometries of the surrounding
2 Enol molecules calculated at the semiempirical PM3 level.
The single point S 1 → S 0 emission energies were calculated at the TD(CAMB3LYP)/6-31G(d). In the case of isolated monomer, optimized 2 IPT without
surrounding enols, the order of the computed emission energies (2-O < 2-R < 2-Y) did
not coincide with the observed ESIPT luminescence (2-R < 2-O < 2-Y) (Table 14.3).
On the other hand, when the π-stacked dimer geometries consisted of the optimized
Table 14.3 S 1 –S 0 vertical energy (nm) of 2 IPT for optimized geometry at TD(B3LYP)/631G(d):PM3 (1 2 IPT +16 2 Enol ) ONIOM model
Isolated model a
ONIOM model
Measured
Monomer
Dimer
Monomer
Dimer
2-Y
548
541.8
506.8
540.8
505.9
2-O
570
592.8
519.5
590.7
518.3
2-R
585
557.0
587.8
555.8
586.4
Non-stacked dimer (2-O)
–
–
567.4
–
565.1
Mutai et al. [89]—Reproduced by permission of The Royal Society of Chemistry
a Geometries extracted from the optimized 17-molecule ONIOM model