Excited-State Conformational/Electronic Responses of Saddle-Shaped N, N ′-Disubstituted-Dihydrodibenzo[ a, c ]phenazines: Wide-Tuning Emission from Red to Deep Blue and White Light Combination

Article abstract of DOI:10.1021/jacs.5b03491

A tailored strategy is utilized to modify 5,10-dimethylphenazine (DMP) to donor-acceptor type N,N′-disubstituted-dihydrodibenzo[a,c]phenazines. The representative compounds DMAC (N,N′-dimethyl), DPAC (N,N′-diphenyl), and FlPAC (N-phenyl-N′-fluorenyl) reveal significant nonplanar distortions (i.e., a saddle shape) and remarkably large Stokes-shifted emission independent of the solvent polarity. For DPAC and FlPAC with higher steric hindrance on the N,N′-substituents, normal Stokes-shifted emission also appears, for which the peak wavelength reveals solvent-polarity dependence. These unique photophysical behaviors are rationalized by electronic configuration coupled conformation changes en route to the geometry planarization in the excited state. This proposed mechanism is different from the symmetry rule imposed to explain the anomalously long-wavelength emission for DMP and is firmly supported by polarity-, viscosity-, and temperature-dependent steady-state and nanosecond time-resolved spectroscopy. Together with femtosecond early dynamics and computational simulation of the reaction energy surfaces, the results lead us to establish a sequential, three-step kinetics. Upon electronic excitation of N,N′-disubstituted-dihydrodibenzo[a,c]phenazines, intramolecular charge-transfer takes place, followed by the combination of polarization stabilization and skeletal motion toward the planarization, i.e., elongation of the π-delocalization over the benzo[a,c]phenazines moiety. Along the planarization, DPAC and FlPAC encounter steric hindrance raised by the N,N′-disubstitutes, resulting in a local minimum state, i.e., the intermediate. The combination of initial charge-transfer state, intermediate, and the final planarization state renders the full spectrum of interest and significance in their anomalous photophysics. Depending on rigidity, the N,N′-disubstituted-dihydrodibenzo[a,c]phenazines exhibit multiple emissions, which can be widely tuned from red to deep blue and even to white light generation upon optimization of the surrounding media.

Full text of DOI:10.1021/jacs.5b03491

Subscriber access provided by UNIV OF MISSISSIPPI
Article
Excited-State Conformational/Electronic Responses of Saddle-
Shaped N,N´-Disubstituted-Dihydrodibenzo[a,c]phenazines: Wide-
Tuning Emission from Red to Deep Blue and White Light Combination
Zhiyun Zhang, Yu-Sin Wu, Kuo-Chun Tang, Chi-Lin Chen, Jr-Wei Ho, Jianhua Su, He Tian, and Pi-Tai Chou
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.5b03491 • Publication Date (Web): 15 Jun 2015
Downloaded from http://pubs.acs.org on June 17, 2015
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 13
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Excited-State Conformational/Electronic Responses of Saddle-
Shaped N,N′-Disubstituted-Dihydrodibenzo[a,c]phenazines: Wide-
Tuning Emission from Red to Deep Blue and White Light Combina-
tion
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Zhiyun Zhang,†,‡ YuꢀSin Wu,‡ KuoꢀChun Tang,‡ ChiꢀLin Chen,‡ JrꢀWei Ho,‡ Jianhua Su,† He Tian,*,†,
and PiꢀTai Chou*,‡
†Key Laboratory for Advanced Materials and Institute of Fine Chemicals, East China University of Science & Technology,
Shanghai 200237, P. R. China
‡Department of Chemistry, National Taiwan University, Taipei, 10617 Taiwan, R.O.C.
ABSTRACT: A tailored strategy is utilized to modify 5,l0ꢀdimethylphenazine (DMP) to donorꢀacceptor type N,N′ꢀdisubstitutedꢀ
dihydrodibenzo[a,c]phenazines. The representative compounds DMAC (N,N′ꢀdimethyl), DPAC (N,N′ꢀdiphenyl) and FlPAC (Nꢀ
phenylꢀN′ꢀfluorenyl) reveal significant nonꢀplanar distortions (i.e., a saddle shape) and remarkably large Stokesꢀshifted emission
independent of the solvent polarity. For DPAC and FlPAC with higher steric hindrance on the N,N′ꢀsubstituents, normal Stokesꢀ
shifted emission also appears, for which the peak wavelength reveals solventꢀpolarity dependence. These unique photophysical
behaviors are rationalized by electronic configuration coupled conformation changes en route to the geometry planarization in the
excited state. This proposed mechanism is different from the symmetry rule imposed to explain the anomalously longꢀwavelength
emission for DMP11 and is firmly supported by polarityꢀ, viscosityꢀ, and temperatureꢀdependent steadyꢀstate and nanosecond timeꢀ
resolved spectroscopy. Together with femtosecond early dynamics and computational simulation of the reaction energy surfaces,
the results lead us to establish a sequential, threeꢀstep kinetics. Upon electronic excitation of N,N′ꢀdisubstitutedꢀ
dihydrodibenzo[a,c]phenazines, intramolecular chargeꢀtransfer takes place, followed by the combination of polarization stabilizaꢀ
tion and skeletal motion toward the planarization, i.e., elongation of the πꢀdelocalization over the benzo[a,c]phenazines moiety.
Along the planarization, DPAC and FlPAC encounter steric hindrance raised by the N,N′ꢀdisubstitutes, resulting in a local miniꢀ
mum state, i.e., the intermediate. The combination of initial chargeꢀtransfer state, intermediate, and the final planarization state renꢀ
ders the full spectrum of interest and significance in their anomalous photophysics. Depending on rigidity, the N,N′ꢀdisubstitutedꢀ
dihydrodibenzo[a,c]phenazines exhibit multiple emissions, which can be widely tuned from red to deep blue and even to white light
generation upon optimization of the surrounding media.
1. Introduction
biphenyls and biaryls.6c, 6d Another case, which is relatively
rare, is the bentꢀtoꢀplanar7 motion involving outꢀofꢀplane
bending of a polycyclic aromatic molecule. Due to its scarcity,
any new finding of bentꢀtoꢀplanar motion accompanied by
unusual photophysical properties may add a new chapter in
this field to gain the associated fundamentals and applications.
Molecules having distinct differences in electronic configuꢀ
ration and/or structures between electronically ground and
excited states may display very distinct photophysical properꢀ
ties, such as large Stokesꢀshifted and/or multiple emissions,
which are of both fundamental and application interests.1ꢀ7
A
famous paradigm of the former case can be ascribed to the
excitedꢀstate electron (charge) transfer such that a large
change of dipole moment takes place between ground and
lowestꢀlying excited states, inducing solvent dipolar relaxation
for stabilization. As a result, the shift in emission spectra is a
function of solvent polarity, dubbed solvatochromism.2 The
latter can be demonstrated in a number of cases, such as excitꢀ
edꢀstate cisꢀtrans isomerization,3 ring opening/closure isomeriꢀ
zation,4 proton transfer tautomerism5 and geometry planarizaꢀ
tion,6, 7 resulting in a product that possesses a distinctly differꢀ
ent structure from the ground state. As for the excitedꢀstate
geometry planarization, the most typical case is twistedꢀtoꢀ
planar,6 which results from the rotation of the aryl rings
around the C–C (or C–N) single bond as in the initially twisted
Herein we report the photophysical studies of a series of
N,N'ꢀdisubstitutedꢀdihydrophenazines analogues exhibiting
anomalous spectral properties. Owing to highly electronꢀrich
and redoxꢀactive properties, dihydrophenazines have been
widely investigated for application in organic electronics as pꢀ
9
type materials.8, Meanwhile, the phenanthrene substitutes
also serve as useful building blocks to develop novel optoelecꢀ
tronic architectures.10 It is therefore demanding to develop
new N,N'ꢀdisubstitutedꢀdihydrophenazine derivatives comprisꢀ
ing phenanthrene (as a fused ring) and to explore their associꢀ
ated physical and chemical properties from both fundamental
and application perspectives. Upon pursuing holeꢀtransporting
and emissive materials based on arylamineꢀsubstituted pheꢀ
nanthrenes, in serendipity, we were very much attracted by an
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 13
intriguing phenomenon from the relevant compound 9ꢀ(9,9ꢀ the FranckꢀCondon excited state, accounts for the multiple
1
2
3
4
5
6
7
8
dimethylꢀ9Hꢀfluorenꢀ3ꢀyl)ꢀ14ꢀphenylꢀ9,14ꢀ
emission behaviors of the titled N,N′ꢀdisubstituedꢀ
dihydrodibenzo[a,c]phenazines.
dihydrodibenzo[a,c]phenazine (FlPAC, Scheme 1), which is
colorless in solution yet exhibit pronounced red emission. This
emission peak wavelength is independent of solvent polarity
with a Stokes shift, defined as the difference in energy beꢀ
tween absorption and emission peaks, of > 11,000 cmꢀ1.
2. Results and Discussion
2.1 Synthesis Strategy and Molecular Structure
N,N′ꢀDimethylꢀdihydrophenazines DMP and DMAC were
prepared by reductionꢀmethylation of phenazine derivatives
according to the standard literature procedure.13 DPAC and
FlPAC were prepared via a method of copperꢀpromoted couꢀ
pling starting from N9,N10ꢀdiphenylphenanthreneꢀ9,10ꢀdiamine
derivatives and aryl iodides (see Scheme S1 of the supporting
information (SI)). Details of the synthetic route are elaborated
in SI. All N,N′ꢀdisubstitutedꢀdihydrophenazines were fully
characterized with 1H NMR, 13C NMR and HRMS.
Scheme 1. Chemical structures of N,N'-disubstituted-
dihydrophenazine and benzoannealated analogues.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For the parent compound DMP, earlier theoretical calculaꢀ
tion has pointed out the bent geometry (C2v) of its molecular
structure,11 and later it has been further confirmed8a to have a
single crystal structure, showing that the molecule is bent
along the N1–N2 axis and the intersecting planes of the two
aryl rings along the N1–N2 axis form a bay with a bent angle
Θb (∠C(1)–N(1)–N(2)–C(2) ~144°, see Figure 1). More speꢀ
cifically, the bent angle Θb is defined as the angle between two
planes (of two opposite aryl rings) bisected and hinged by the
two nitrogen atoms (see Figure 1). Furthermore, crystals of
DMAC and DPAC8h were also obtained by slow evaporation
of solvent from the dichloromethane/ethanol mixture solutions
and examined by Xꢀray analysis. The three angles of view of
these compounds’ single crystal structures are shown in Figure
1, together with the selected angles. Derived from DMP with
the addition of the nonlinear aromatic phenanthrene fused to
dihydroquinoxaline, DMAC inherits the bent geometry, with a
bent angle Θb at around 153°. The most pronounced difference
among the geometry of these two crystal structures is the angle
between the N1–C3 (N2–C4) and N1–N2 planes (∠C(3)–
N(1)–N(2) and ∠C(4)–N(2)–N(1), see Figure 1). The angle
∠C(3)–N(1)–N(2) (∠C(4)–N(2)–N(1)) of DMP is around
175°, while that of DMAC is reduced substantially to ~114°.
Obviously, the difference mainly results from the steric hinꢀ
drance between the substituted group on the N1 (N2) and pheꢀ
nanthrene ring (see Scheme 1 and Figure 1). On the other hand,
DPAC and FlPAC are designed by replacing the Nꢀmethyl
group of DMAC with aromatic rings. As shown in Figure 1c,
compared with DMAC, DPAC exhibits higher nonꢀplanar
distortions (Θb ~137° and ∠C(3)–N(1)–N(2) (∠C(4)–N(2)–
N(1)) ~110°) because of the bulkier substitutes and hence
greater steric hindrance being imposed.
We then performed a comprehensive literature search releꢀ
vant to the photophysics of dihydrophenazines. In 1974, Huber
and coꢀworkers reported11 the anomalous absorption and emisꢀ
sion behavior of 5,l0ꢀdimethylphenazine (DMP, see Scheme
1), 5ꢀmethylꢀ10ꢀphenylphenazine (MPP), and 5,l0ꢀ
diphenylphenazine (DPP). Exploiting symmetry analyses,
they attributed the great Stokes shift to the energy gap between
two different excited states, S2 and S1, responsible for the abꢀ
sorption and emission, respectively, since the S0 → S1 transiꢀ
tion is symmetrically forbidden and hence obscure in the abꢀ
sorption spectrum. Later, Schuster et al. reported12 the absorpꢀ
tion and emission behavior of benzoannealated analogues
DMP,
(DMAC,
9,14ꢀdimethylꢀ9,14ꢀdihydrodibenzo[a,c]phenazine
Scheme 1) and 6,13ꢀdimethylꢀ6,13ꢀ
dihydrodibenzo[b,i]phenazine (DMBI), and imposed a similar
symmetryꢀrule based explanation to rationalize the anomalous
Stokesꢀshifted emission for DMP and DMAC. On the other
hand, DMBI possesses symmetryꢀallowed S0 → S1 transition
and hence exhibits a normal Stokesꢀshifted emission.
If the mechanism discussed above were applicable to most
dihydrophenazines and the derivatives, we would not have
been puzzled by the enormously large Stokesꢀshifted emission
in, e.g., FlPAC, for which the C2 symmetry has been broken
down. Interestingly, since Schuster's work and up to the curꢀ
rent stage, no inꢀdepth photophysical studies along with the
excitedꢀstate
dynamics
of
the
N,N′ꢀdisubstitutedꢀ
dihydrophenazines have been conducted, despite their vast
usage in optoelectronic applications.8 We have thus strategiꢀ
cally designed and synthesized a series of N,N′ꢀdisubstitutedꢀ
dihydrophenazine and benzoannealated analogues (Scheme 1),
in which the representative DMAC, DPAC and FlPAC are
the electron donorꢀacceptor type of dyads (vide infra). Strucꢀ
tural analyses indicate that they all reveal significant nonꢀ
planar distortion of the dihydrodibenzo[a,c]phenazine frameꢀ
work as a result of steric hindrance between the N,N'ꢀ
disubstitutes and phenanthrene ring (Figure 1). Remarkably,
they all exhibit enormously large Stokesꢀshifted emission. In
this study, combining comprehensive solvent polarity and
viscosity dependent spectroscopy, timeꢀresolved measureꢀ
ments and computational approaches, we posit a generalized
mechanism incorporating structural relaxation, namely a bentꢀ
toꢀplanar transition of the Vꢀshaped dihydrodibenꢀ
zo[a,c]phenazine framework, to extend the πꢀconjugation.
This, together with the significant charge transfer character in
Upon electronic excitation of the above titled N,N′ꢀ
disubstitutedꢀdihydrodibenzo[a,c]phenazines,
energetically
favorable planarization takes place to elongate the πꢀ
conjugation. The bentꢀtoꢀplanar motion may encounter steric
hindrance. This, together with the inherited excitedꢀstate
charge transfer property, results in novel and distinctive phoꢀ
tophysical properties of these saddleꢀshaped molecules. Deꢀ
tails are provided in the following sections.
2
ACS Paragon Plus Environment

Page 3 of 13
Journal of the American Chemical Society
cmꢀ1, respectively, for DPAC and FlPAC (see enlarged area of
1
Figure 2c, d).
2
3
4
5
(a)
Toluene (Abs)
1.0
0.5
DMP
Cyclohexane (Em)
Toluene (Em)
Acetonitrile (Em)
6
7
8
9
0.0
1.0
(b)
DMAC
10
11
12
13
14
15
16
0.5
0.0
1.0
(c)
(d)
DPAC
FlPAC
×
60
0.5
Figure 1. Single crystal structures of (a) DMP, (b) DMAC and (c)
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
DPAC and their three angles of view together with the selected
angles. The bending angle between plane 1 (C1, N1, and N2)
and plane 2 (N1, N2, and C2) is denoted as Θb for all titled
compounds. Note that under this definition, a bending angle of
180° specifies the planarization between planes 1 and 2, while
0° indicates two parallel planes. Accordingly, from 0–180°, a
larger bending angle indicates more planarization.
0.0
1.0
×
10
0.5
0.0
1.0
(e)
2.2 Steady-State Emission Spectroscopy
DMP
77 K
DMAC
DPAC
FlPAC
The normalized absorption and emission spectra of DMP,
DMAC, DPAC, and FlPAC in various solvents from nonꢀ
polar cyclohexane to polar acetonitrile are shown in Figure 2.
The first absorption maximum of DMP is located at ~343 nm
(in toluene), while the maxima of phenanthreneꢀfused derivaꢀ
tives (DMAC, DPAC and FlPAC) are red shifted by ~10 nm
to that of DMP. The absorption onsets of DPAC and FlPAC
are at about 400 nm. In sharp contrast, the onset of DMAC’s
tails down to ~470 nm. This result affirms the above structural
analyses of a more planar structure for DMAC (cf., DPAC
and FlPAC) due to its smaller steric hindrance (see Figure 1),
rendering a greater πꢀdelocalization and hence lower energy
gap.
0.5
0.0
300
400
500
600
700
800
Wavelength (nm)
Figure 2. Steadyꢀstate absorption (in toluene, dashed line) and
photoluminescence (solid line) spectra of (a) DMP, (b) DMAC,
(c) DPAC and (d) FlPAC in various solvents at room temperature,
and (e) their photoluminescence spectra at 77 K in toluene glass.
The enlarged blue emission profiles of DPAC and FlPAC are
also shown in (c) and (d), respectively.
All titled compounds, including DMP, exhibit a large
Stokesꢀshifted emission. For DMP, the peak wavelength of
fluorescence is located at 465 nm in, e.g., cyclohexane with a
Stokes shift of ~8,000 cmꢀ1 (Figure 2a). For DMAC, DPAC
and FlPAC, shown in Figure 2b, c and d, respectively, the
emission peak wavelength is even farther red shifted, with the
value of Stokes shift calculated to be > 11,000 cmꢀ1. Imꢀ
portantly, the maximum of this anomalously long wavelength
emission is nearly independent of solvent polarity, being red
shifted by less than 15 nm (537 cmꢀ1) from cyclohexane to
acetonitrile for all titled compounds (Figure 2aꢀd).
We first pondered the mechanism of symmetry rule adopted
for DMP (vide supra). However, the breakdown of C2 symꢀ
metry in, e.g., FlPAC, makes the assignment of red emission
originating from the symmetry restriction rule impossible.
Also, the origin from solvatochromism has been eliminated
because of the nearly solventꢀpolarity independence of the red
emission (Figure 2). Alternatively and more plausibly, owing
to the Vꢀshaped nonꢀplanar geometry of the titled compounds
(Figure 1), we tentatively propose the occurrence of structural
bending motion along the N(1)–N(2) axis for DMAC, DPAC
and FlPAC in the electronically excited state, reaching a plaꢀ
nar configuration to a fully extended πꢀconjugation and hence
an energy minimum.
In addition to the dominant large Stokesꢀshifted band maxꢀ
imized at ~610 nm, careful examination of DPAC and FlPAC
also shows a weak but nonꢀnegligible normal Stokesꢀshifted
band around blue region. Both normal and large Stokesꢀshifted
emissions share the same origin, i.e., the same FranckꢀCondon
excited state, as evidenced by the identical excitation spectra,
which are also overlapping with the absorption spectrum (Figꢀ
ure S1 in SI for FlPAC). In stark contrast to the solventꢀ
polarity insensitive 610 nm emission, the normal emission
bands show substantial solvent polarity dependence, being red
shifted from cyclohexane to acetonitrile by 1,800 and 2,730
The largeꢀamplitude motion proposed above is expected to
be viscosity dependent. To support this viewpoint, we first
investigated their emission property in the rigid toluene matrix
at 77 K, in which the structural relaxation toward planarization
is impeded. The results shown in Figure 2e at 77 K offer firm
support; DMAC, DPAC and FlPAC all exhibit solely normal
Stokesꢀshifted emission at 525 nm, 400 nm, and 405 nm, reꢀ
spectively. Note that the additional emission bands of DPAC
and FlPAC that appear in the range of 500 to 600 nm in the 77
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 13
K toluene solid matrix are assigned to the phosphorescence emission was gradually tuned from red to green throughout the
1
2
3
4
5
6
7
8
due to its > 1 ms lifetime. In stark contrast, DMP still exhibits
a large Stokesꢀshifted fluorescence at 470 nm in the 77 K toluꢀ
ene solid matrix, the spectral feature of which is nearly identiꢀ
cal with that acquired at room temperature. Also, as shown in
Figure S2 (see SI), DMAC, DPAC and FlPAC reveal normal
Stokesꢀshifted emission in solid due to the lattice constraint,
whereas a similar large Stokesꢀshifted emission is still obꢀ
served in solid DMP. Unambiguously, the results lead us to
conclude different origins of anomalous emission between
DMP and its phenanthreneꢀfused derivatives DMAC, DPAC
and FlPAC. On the one hand, the anomalous longꢀwavelength
emission for DMP has been reported to originate from the
imposition of a symmetry rule (vide infra).11 On the other hand,
considering that the symmetry rule should not be temperaꢀ
ture/viscosity dependent, the latter must involve a significant
conformational change, plausibly toward planarization, in the
excited state, such that the normal and anomalous Stokesꢀ
shifted emissions are tentatively ascribed to bent and planar
forms, respectively.
temperature dependent study to 133 K (see Figure S3c and
Figure S4 for the change in color index in the CIE coordinate).
The steadyꢀstate approach described above thus suggests
that the planarization may couple with other electronic and/or
structural relaxation processes to account for the observed
multiple emissions that show various degrees of temperature
dependence among DMAC, DPAC and FlPAC. To shed light
on the associated mechanism, we then further performed naꢀ
nosecond time resolved emission spectra and their correspondꢀ
ing dynamics in the viscous solvents, described below.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The harnessing of the large amplitude internal motion can
be elegantly demonstrated by using nꢀbutanol as the media to
fineꢀtune the ratiometric emission because its viscosity is relaꢀ
tively high (2.948 cP at 293 K, cf. 0.343 cP for CH3CN at 298
K)14 and is temperature sensitive,15 suitable for probing the
ratiometric changes of the emission in a steady state manner.
Using FlPAC as a prototype, the decrease of temperature from
293 to 233 K is accompanied by the gradual increase of emisꢀ
sion intensity in both red and blue regions (see Figure S3a of
SI). Because the associated peak wavelengths remain unꢀ
changed, the result may qualitatively be rationalized by a
combination of slowing down the planarization process and
the suppression of the nonꢀradiative decay process. Further
lowering the temperature from 233 K to 173 K led to a deꢀ
crease in intensity and slight blue shift of the red emission
band from 610 nm to 595 nm, accompanied by an increase and
broadening of the blueꢀgreen emission band (~450–520 nm)
(Figure 3). As the temperature dropped toward 133 K, the red
emission band disappeared; concomitantly, the sky blueꢀgreen
emission band was shifted to shorter wavelengths and finally
showed a blue emission maximized at 425 nm at 133 K. Also
noteworthy is that the full width at half maximum (fwhm) of
the blueꢀgreen emission band was significantly reduced from
5,802 cmꢀ1 at 173 K to 3,215 cmꢀ1 at 133 K. Obviously, the
temperatureꢀdependent study of FlPAC in nꢀbutanol revealed
multiple bandꢀlike emissions covering the entire visible specꢀ
tral range from red to deep blue. The changes in emission colꢀ
or can be viewed with the naked eye (inset of Figure 3) as well
as by the color hue index of the Commission Internationale de
L’Eclairage (CIE) coordinate (Figure S4). At 188 K, the CIE
coordinate for FlPAC was calculated to be (0.33, 0.30), close
to the coordinate (0.33, 0.33) of the standard white light illuꢀ
mination.
Figure 3. Temperatureꢀdependent fluorescence spectra of FlPAC
in nꢀbutanol from 233 K to 133 K with various viscosities. The
arrows indicate the direction of fluorescence response upon reducꢀ
ing the temperature. (Inset: represented photographs under 365
nm UV light).
2.3 Nanosecond Time-Resolved Emission Spectroscopy
We realized that in a high viscosity system, such as in the
glycerol/ethanol (V/V 4:1) mixture with
η
~171 cP16 at 298 K,
DPAC and FlPAC exhibited a much enhanced blue emission
band versus the large Stokesꢀshifted red emission (see Figure
S5 of SI), inferring the slowdown of the structure relaxation
process, which makes feasible the timeꢀresolved emission
study at room temperature using the nanosecond time correlatꢀ
ed single photon counting (TCSPC) technique. The correꢀ
sponding relaxation dynamics can then be analyzed at each
wavelength to reconstruct the temporal evolution of the emisꢀ
sion spectrum. Accordingly, the spectral temporal evolutions
of DPAC and FlPAC were recorded in a 4:1 glycerol/ethanol
mixture at 298 K and are shown in Figures 4a and 4b, respecꢀ
tively. Upon FranckꢀCondon excitation, for the case of FlPAC,
a deepꢀblue emission band maximized at around 435 nm
promptly appeared (with a systemꢀresponseꢀlimit rise of ~50
ps), which revealed a nice mirror image with respect to the
absorption bands (see Figure 2d for the absorption). As
the time elapsed, the short wavelength emission bands (~400–
550 nm) were gradually decreased and red shifted, whereas the
long wavelength emission (~500–700 nm) continuously grew
and reached its maximum intensity at ~610 nm. A similar
temporal evolution pattern was observed in DPAC (see Figure
4a). It is interesting to note that the spectral temporal profiles
of DPAC and FlPAC in glycerol/ethanol (V/V 4:1) at 298 K
are similar to their temperature/viscosity dependent steadyꢀ
state spectra in nꢀbutanol (see Figure 3 and Figure S3), reafꢀ
firming the involvement of large amplitude motion for the
observed kinetics.
Temperatureꢀdependent emission properties were also obꢀ
served for DPAC and DMAC in nꢀbutanol (see Figure S3 and
S4 of SI). Nevertheless, there are discernible differences
among these three compounds. For DPAC, shown in Figure
S3b, the decrease in intensity of the red emission band and
hypsochromic shift of the blueꢀgreen emission took place at
lower temperature (~193 and ~163 K, respectively, cf. ~233
and ~173 K for FlPAC). As for DMAC, the emission intensiꢀ
ty was continuously increased, accompanied by the hypsoꢀ
chromic shift from room temperature to 133 K. As a result, the
4
ACS Paragon Plus Environment

Page 5 of 13
Journal of the American Chemical Society
gion (< 430 nm), lead us to propose a sequential threeꢀstate
1
2
3
4
reaction kinetics after electronic excitation, which incorporates
the initially excited state, an intermediate state and the finial
population state. The final population state possesses an enerꢀ
gy minimized geometry that plausibly contains a planarꢀlike
dibenzo[a,c]phenazine moiety.
5
6
7
2.4 Femtosecond Fluorescence Up-conversion and Relaxa-
tion Dynamics.
8
9
As described above, the multiple emission bands and their
corresponding relaxation dynamics in the nanosecond regime
could be well resolved in the relatively high viscous solvents
such as nꢀbutanol and glycerolꢀethanol mixture. Nevertheless,
the retardation of motion imposed by the mechanical force
(viscosity) makes the resolution of structural relaxation dyꢀ
namics per se complicated. Moreover, in low viscous solvents,
as shown by the emission spectra of DPAC and FlPAC (see
Figure 2c and d), despite the invariance of peak wavelength
for the major red emission band, the minor blue emission band
reveals a significant bathochromic shift upon increasing the
solvent polarity from cyclohexane to acetonitrile, indicating
that blue emission reveals a charge transfer character in nature.
Accordingly, in a low viscous environment, solvent polarity
and its associated relaxation may couple with structural relaxaꢀ
tion, which then channel into the overall reaction dynamics.
To shed light on the corresponding relaxation dynamics, we
then carried out femtoꢀpicosecond time resolved measurement
by use of the fluorescence upꢀconversion technique. In this
approach, FlPAC served as a prototype, and the study was
carried out in toluene and acetonitrile to elucidate the comparꢀ
ative relaxation dynamics. As shown in Figure 5a and Table 2,
in toluene, the relaxation dynamics of emission at the blue side
(420 nm) exhibited an instant rise (< 120 fs), followed by a
very fast and a fast decay component of ~5.6 ps and ~36 ps,
respectively. Upon monitoring at 450 nm, a finite rise compoꢀ
nent (negative preꢀexponential value) was fitted to be ~1.5 ps,
accompanied by the 5.6 ps and 38 ps decay components. At
500 nm, the rise component increased its weighing ratio (inꢀ
crease of % in preꢀexponential factor) and was lengthened to
~3.6 ps, accompanied by the decrease of the contribution from
the 5.6 ps decay component. At 550 nm, the 5.6 ps decay
component was not resolvable. Instead, in addition to the 38 ps
decay component, a very long decay appeared, which could
not be resolved but treated as a constant value, within the acꢀ
quisition time window of ~1000 ps. This long decay compoꢀ
nent was further resolved to be 7.1 ns by the TCSPC method
(see Figure S6). Upon monitoring at 600 nm, the kinetic trace
can be well fitted by two rise components of 5.6 ps and 38 ps
and a population decay of 7.1 ns (see Figure 5a). Note that the
rise times of 5.6 ps and 38 ps at 600 nm are the same as the 5.6
ps and 38 ps decay time constants monitored at 420 nm and <
550 nm, respectively. Finally, at the very red emission region
(650 nm), the fluorescence transient clearly exhibited only a
rise component of 38 ps and a population decay of 7.1 ns.
Similar kinetic patterns were obtained for FlPAC in acetoniꢀ
trile (see Figure 5b and Table 2), except that the early fast
decay component was decreased from 5.6 ps in toluene to
~1.8–2.0 ps in acetonitrile; the other two time constants were
measured to be ~34 ps and 4.3 ns, which are slightly smaller
than those of 38 ps and 7.1 ns obtained in toluene.
10
11
12
13
14
15
16
17
Figure 4. Reconstructed transient fluorescence of (a) DPAC and
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(b) FlPAC at different times in a 4:1 glycerol/ethanol mixture
with viscosity of ~171 cP at room temperature by timeꢀcorrelated
single photon counting (TCSPC) at room temperature. (c) Fluoꢀ
rescence spectra decay traces recorded for FlPAC in the 4:1 glycꢀ
erol/ethanol mixture at room temperature, of various wavelengths
as depicted. See Table 1 for the fitting results.
Table 1. Fitting results of TCSPC spectroscopic measure-
ments for FlPAC in a 4:1 glycerol/ethanol mixture at room
temperature.
Population decay (ns)
τ1: 0.35 (0.71), τ2: 2.07 (0.29)
λ
probe (nm)
430
450
τ1: 0.35 (0.53), τ2: 2.00 (0.47)
500
τ1: 0.35 (ꢀ1.00), τ2: 0.35 (0.98), τ3: 2.22 (0.02)
τ1: 0.35 (ꢀ0.27), τ2: 2.06 (0.04), τ3: 4.58 (0.96)
τ1: 0.35 (ꢀ0.19), τ2: 2.06 (ꢀ0.70), τ3: 4.58 (1.00)
τ1: 2.06 (ꢀ0.92), τ2: 4.83 (1.00)
550
600
650
We then performed comprehensive spectral/dynamic analꢀ
yses by plotting the rise and decay profiles at each crucial
wavelength for FlPAC. The results are depicted in Figure 4c,
while pertinent data are listed in Table 1. At the deep blue
region of 430 nm, the time resolved fluorescence consists of
two decay components of 0.35 ns and 2.07 ns (see Table 1).
Upon monitoring at 500 nm, a rise time of 0.35 ns appears, as
evidenced by a negative preꢀexponential factor (see Table 1),
which, within experimental error, correlates well with the 0.35
ns decay component detected at 430 nm. Although the emisꢀ
sion at these two wavelengths reveals the same decay compoꢀ
nent of ~2 ns, the difference lies in that its contribution in inꢀ
tensity and hence population, monitored by the preꢀ
exponential value (see Table 1), is substantially decreased at
the bluer wavelength. This result indicates the spectral overlap
of temporal evolution from 430 to 500 nm (Figure 4b). At the
longest wavelength monitored at 650 nm, the overall kinetics
can be well fitted by a rise (~2.06 ns) and a decay (~4.83 ns)
component, in which the rise of ~2.06 ns matches exactly with
the decay component (~2.06 ns) of the short wavelength at,
e.g., 500 nm. The resolved 0.35 ns, 2.0 ns and 4.8 ns time conꢀ
stants throughout the entire emission spectra, together with the
systemꢀresponseꢀlimited rise component at the very blue reꢀ
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 13
rise times of 1.5 ps (450 nm), 3.6 ps (500 nm) and 4.5 ps (550
1
2
3
4
5
6
7
8
nm) can thus be rationalized by the overlap between decay of
R* and rise of I* species in this region; the cancellation of
each other by different ratios of intensity results in shorting of
the fitted rise time. Moreover, as for the R* emission of
FlPAC in the blue, the steadyꢀstate spectrum (see Figure 2d)
has revealed a charge transfer property with a pronounced
solvatochromism effect. Therefore, solvent relaxation should
play a role, which may channel into the overall reaction coorꢀ
dinates; in this case, the early relaxation dynamics is subject to
polarization environments. This viewpoint is supported by the
shorter time constant (τ1) in toluene (5.6 ps) than in acetoniꢀ
trile (1.8–2.0 ps).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Femtosecond fluorescence transients (open circles) of
FlPAC in aerated (a) toluene and (b) acetonitrile at room temꢀ
perature observed at various wavelengths as depicted. The black
solid lines denote the fitting curves. See Table 2 for the fitting
results.
Table 2. Fitting results of femtosecond fluorescence up-
conversion measurement for FlPAC in toluene and ace-
tonitrile. a
Relaxation time constants
In toluene
420 nm [τ1: 5.6 ps (0.34); τ2: 36 ps (0.66)]
450 nm [τ1: 1.5 ps (ꢀ0.47); τ2: 5.6 ps (0.53); τ3: 38 ps (0.47)]
500 nm [τ1: 3.6 ps (ꢀ0.75); τ2: 5.6 ps (0.32); τ3: 38 ps (0.68)]
550 nm [τ1: 4.5 ps (ꢀ0.98); τ2: 38 ps (0.75); τp: 7.1 ns (0.25)]
600 nm [τ1: 5.6 ps (ꢀ0.40); τ2: 38 ps (ꢀ0.42); τp: 7.1 ns (1.00)]
650 nm [τ1: 38 ps (ꢀ1.00); τp: 7.1 ns (1.00)]
In acetonitrile
440 nm [τ1: 1.8 ps (0.81); τ2: 33 ps (0.19)]
480 nm [τ1: 0.5 ps (ꢀ0.37); τ2: 29 ps (1.00)]
500 nm [τ1: 0.5 ps (ꢀ0.41); τ2: 31 ps (1.00)]
560 nm [τ1: 2.0 ps (ꢀ0.64); τ2: 35 ps (0.77); τp: 4.3 ns (0.23)]
600 nm [τ1: 3.4 ps (ꢀ0.73); τ2: 33 ps (ꢀ0.05); τp: 4.3 ns (1.00)]
620 nm [τ1: 2.0 ps (ꢀ0.30); τ2: 33 ps (ꢀ0.58); τp: 4.3 ns (1.00)]
a The time constant τp of the population decay is determined from
the TCSPC result and is used to the fitting of upꢀconversion sigꢀ
nals.
Figure 6. Reconstructed plots of FlPAC in acetonitrile at room
temperature resolved by femtosecond fluorescence upꢀconversion:
(a) spectral temporal evolution. (b) A 3ꢀD surface plot provides a
more intuitive way to perceive the temporal/spectral evolution and
the excited state dynamics. Note that the time delay in (b) is in
logarithm.
According to the above fluorescence upꢀconversion data, for
simplicity, the overall structural relaxation process of FlPAC
can be qualitatively described by a sequential kinetic pattern
expressed as
Figure 6a depicts the reconstructed femtoꢀpicosecond specꢀ
tral temporal evolution of FlPAC in acetonitrile. Several reꢀ
marks can be pointed out from the layout of the spectra. First
of all, the clearly resolved temporal emission region around
450–550 nm is obscure in the steady state emission (cf. Figure
2d). Secondly, as the time elapsed, the overall spectrum reꢀ
veals multiple bands as well as gradual bathochromic shift.
The spectral peak quickly evolves from 460 nm (R*) to 520
nm (I*) within first ~10 ps, and then finally the population
emission band settles down at ~610 nm (P*) after ~40 ps. A
similar trend of spectral temporal evolution was observed for
FlPAC in toluene (see Figure S7), except that the spectral red
shift is slower in the time domain than that in acetonitrile.
Importantly, these femtoꢀpicosecond spectral temporal evoluꢀ
τ1
τ2
R* →I* →P*
(1)
where R* specifies the initial excited state, I* is an intermediꢀ
ate with the local minimum energy, and P* denotes the final
state with the global minimum energy along the structural
relaxation process. R*, I* and P* are all assumed to be the
emissive states. In toluene, we reasonably assign the time conꢀ
stants of the R* → I* and I* → P* processes to be 5.6 ps (τ1)
and 38 ps (τ2), respectively, followed by a population decay of
the P* state of 7.1 ns. On this basis, the relatively complicated
6
ACS Paragon Plus Environment

Page 7 of 13
Journal of the American Chemical Society
tions in low viscous solvent exactly echo the nanosecond timeꢀ of the phenyl derivative in DPAC during the structural relaxaꢀ
1
2
3
4
5
6
7
8
resolved data and temperature/viscosity dependent spectra in
the high viscous solvents. The results manifest the occurrence
of the structure evolution, which may also couple with the
solvent relaxation at early dynamics, giving rise to the obꢀ
served time resolved multiple emission bands with complex
kinetics. Figure 6b provides another angle of view to intuitiveꢀ
ly perceive the excited state dynamics of the proposed consecꢀ
utive reaction mechanism (R* → I* → P*). Also, for clarity,
various 3ꢀD surface plots from other viewing angles are shown
in Figure S8.
tion. On the other hand, the system response limiting decay (<
150 fs) in the blue region (440 nm) of DMAC implies its fast
structural relaxation with negligible steric hindrance being
imposed. Such a fast deactivation rate also well explains the
lack of a blue emission band for DMAC in toluene (see Figure
2b).
1.0
9
FlPAC
DPAC
0.8
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The femtoꢀpicosecond time resolved measurements were alꢀ
so carried out for DPAC and DMAC (see Figure 7 and Figure
S9 of SI) to gain insights into the retardation of structural reꢀ
laxation caused by the steric effect. Figure 7 shows the upꢀ
converted fluorescence transients of DMAC, DPAC, and
FlPAC probed in the very blue region in toluene. The use of
the same solvent rules out any discrepancy due to solvent visꢀ
cosity or the polarity effect. The results, on the one hand,
clearly indicate that the decay of FlPAC (5.6 ps, τ1, see Table
2) is longer than that of DPAC (2.8 ps, τ1, see Table S1 in SI),
implying that the swing space of the bulky side chain, such as
the fluorene derivatives in FlPAC, is more hindered than that
DMAC
Fitting
0.6
0.4
0.2
0.0
0
10
20
30
40
50
Delay Time (ps)
Figure 7. Femtosecond fluorescence transients of FlPAC (420
nm), DPAC (420 nm) and DMAC (440 nm) in aerated toluene
Table 3. The experimental and calculated optical characteristics for the titled compounds.a
Stokes shift (cmꢀ1)
Absorption S0 → S1
exp(nm)b (ε/Mꢀ1cmꢀ1)
calc(nm)
Emission S1 → S0
exp(nm)b
τf (ns)c
Q.Y.
b
f
F
λ
λ
λ
λcalc(nm)
ꢀEexp
ꢀEcalc
0.58
0.13
103 (84)d
9.5
343 (13,600)
353 (6,100)
354
391
0.0001
0.0779
473
612
423
613
433
618
471 0.0001
603 0.0572
460 0.1676
564 0.1295
493 0.1521
567 0.1614
8013
11989
5177
7017
8992
4100
8109
4654
7301
DMP
DMAC
0.001 <50 ps [450nm]
347 (9,600)
387
0.1079
DPAC
12505
4606
0.22
11.6 [650nm]
<50 ps [450nm]
11.0 [650nm]
0.005
0.17
361 (19,600)
401
0.1473
FlPAC
11520
a Data were recorded in degassed toluene at room temperature; see details in supporting information for fluorescence quenching in aerated
toluene. λexp = experimental lowestꢀlying absorption or emission wavelengths, λcalc = calculated lowestꢀlying absorption or emission waveꢀ
lengths, ε = molar extinction coefficient, f = oscillator strengths, ꢀEexp = experimental Stokes shift, ꢀEcalc = calculated Stokes shift, Q.Y. =
emission, quantum yield, τf = population decay time. b The experimental values are taken from the peak wavelength of the absorption and
c
emission. The population decay time constants, which are measured by the timeꢀcorrelated single photon counting (TCSPC, Edinburgh
FL 900). d Reference 12, determined in argon purged benzene at room temperature.
Table 4. Frontier molecular orbital contours of the titled compounds in optimized ground-state geometries. Also shown are
FlPAC with relaxed structures in the S1 excited state at the local (133°) and global (160°) minima.
DMP
DMAC
DPAC
FlPAC
146°
151°
138°
136°
133°
160°
Bending Angle (Θb)
LUMO
HOMO
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 13
2.5 Computational Approach
possible the existence of an intermediate state with local minꢀ
imum energy described below.
1
2
3
4
5
6
7
8
We then performed the computational approach in an aim to
establish the associated mechanism, i.e., the excitedꢀstate conꢀ
formational/electronic responses of these unique saddleꢀ
shaped compounds. In this study, we applied the timeꢀ
dependent density functional theory (TDꢀDFT) approach (see
experimental section) to access the excited state properties, in
which the S0 → S1 absorption is taken directly from the
FranckꢀCondon excitation of the geometry optimized ground
state, while the emission energy gap is calculated via the vertiꢀ
cal S1 → S0 transition from the structurally optimized S1 state.
The pertinent energy gaps of each calculated transition, toꢀ
gether with experimental absorption and emission data in toluꢀ
ene, are listed in Table 3.
Using FlPAC as a paradigm, Figure 8 depicts the simulated
potential energy surface (PES) for FlPAC scanning along Θb
for both ground and lowest excited states, for which the geomꢀ
etry has been fully optimized at each value of Θb. As a result,
at the ground state, the Θb value for the geometry optimized
structure is 136°. We denote this geometry optimized ground
state as R. The vertical electronic transition of the R state is
calculated to be 401 nm (see Table 3), which is denoted as the
FranckꢀCondon excited state R* (also defined in eq. (1)).
Scanning Θb along the S1 PES, the global minimum was locatꢀ
ed at Θb = 160° and is represented by P*. Importantly, in addiꢀ
tion to the global minimum along Θb in the S1 state, FlPAC
was found to have a local minimum at Θb = 133°, denoted as
I*, along the structural relaxation pathway (see Figure 8). The
vertical transition from these two excitedꢀstate minima, i.e., I*
→ Iand P* → P, are calculated to be 493 and 567 nm, respecꢀ
tively. On the one hand, the estimated 401 nm excitation and
493 nm emission for FlPAC correlates well with the steadyꢀ
state ~420–550 nm emission resolved from the temperature
dependent study in nꢀbutanol (Figure 3) as well as nanosecond
(Figure 4b) and femtoꢀpicosecond (Figure 6) time resolved
spectra. On the other hand, the calculated 567 nm emission
wavelength matches the onset of the large Stokesꢀshifted red
emission (λmax ~610 nm) measured in all solvents applied in
this study.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For these molecules, the results of calculation show that S0
to S1 electronic transition corresponds to moving one electron
from the highest occupied molecular orbital (HOMO) to the
lowest unoccupied molecular orbital (LUMO) rather than
some linear combination of orbitals (See details in Table S2).
As shown in Table 4, the electron density for both HOMO and
LUMO of parent compound DMP are delocalized over the
entire molecular framework. In sharp contrast, upon extending
one of the phenyleneꢀfused rings to phenanthrene in DMP,
forming DMAC, DPAC and FlPAC, due to the steric hinꢀ
drance and structure nonꢀplanarity, HOMO and LUMO, to
certain degrees, are then localized at the central moiety N,N′ꢀ
disubstitutedꢀdihydropyrazine and the phenanthrene ring, reꢀ
spectively. For example, the HOMO of DMAC is mainly loꢀ
calized in the center moiety N,N′ꢀdimethylꢀdihydropyrazine
and partially extended fusing rings (phenylene and phenanꢀ
threne) on the two sides. Replacing N,N′ꢀdimethyl of DMAC
with two bulky phenyls, forming DPAC, results in the further
bending of two planes along the N1–N2 axis and hence the
reduction of the πꢀconjugation, such that HOMO of DPAC is
greatly localized in the center moiety N,N′ꢀdiphenylꢀ
dihydropyrazine. Further replacing one N,N′ꢀdiphenyl of
DPAC to fluorenꢀ3ꢀyl, forming FlPAC, the corresponding
HOMO was then mainly localized in the Nꢀfluorenꢀ3ꢀylꢀ
dihydropyrazine moiety. On the other hand, LUMO of DMAC,
DPAC and FlPAC are all mainly localized at the fused pheꢀ
nanthrene ring (see Table 4). The result of computation clearly
demonstrates that the initial excited state for DMAC, DPAC
and FlPAC, in part, is associated with the charge transfer
character.
70
FlPAC
68
R∗
I∗
66
P∗
64
40
30
20
10
0
R
130
140
150
160
170
Bending angle,Θb (deg.)
Figure 8. Calculated potential energy surfaces (PES) scanned
along the change of bending angles Θb from 130 to 170 degrees in
an increment of 1° for FlPAC in the first excited state (S1(ꢀ)),
and ground state, in which S0(ꢀ) denotes the vertical transition
from S1(ꢀ) in a specific Θb. S0(ꢁ) is a groundꢀstate optimized
state at the same Θb. Open data symbols ꢂ and ꢁ mark the globꢀ
al (or local) minimum of the PES for S0(ꢁ) and S1(ꢀ), respecꢀ
tively. Note: R* denotes the initial prepared S1 state, from which
only structural relaxation along Θb to the intermediate I* (see text)
is depicted. For simplicity, other relaxation processes such as
vibrational deactivation and solvent relaxation processes are omitꢀ
ted. More explicit depiction and delineation of the corresponding
relaxation pathways can be found in Figure 10.
In the excited state, as unveiled by the fully optimized geꢀ
ometry, we learned that the molecular framework tends toward
planarity from an initially prepared bending structure so that
the two nitrogen lone pair electrons of the N,N′ꢀphenazines
participate in the elongation of πꢀconjugation and hence the
stabilization, inducing a skeletal relaxation along Θb (Figure 1).
For a full planar conformation (Θb ~180°), the πꢀelectron deꢀ
localization within the molecular framework can thus be greatꢀ
ly extended. We then performed potential energy surface (PES)
simulation scanning along the bending angle Θb for all pheꢀ
nanthreneꢀfused molecules from 100 to 180° in both S0 and S1
states. Clearly, in the lowest lying electronically excited state,
the structural relaxation toward planarization is thermodynamꢀ
ically favorable for all phenanthreneꢀfused compounds. We
also realize, however, that amid the change of Θb toward a
global minimum energy, the planarization process also induces
the steric effect of the hinged N,N'ꢀdisubstitutes; hence, their
concomitant structural relaxation is necessary, which makes
Further detailed frontier analyses (see Table 4) also indicate
that the R* state prepared from the vertical excitation can be
ascribed to a HOMO → LUMO transition with an obvious
charge transfer character from the Nꢀfluorenꢀ3ꢀylꢀN,N′ꢀ
dihydropyrazine moiety (donor) to the phenanthreneꢀfused
ring (acceptor). This viewpoint is also supported by the calcuꢀ
8
ACS Paragon Plus Environment

Page 9 of 13
Journal of the American Chemical Society
lated dipole moment of 15.24 D in R*, which is significantly structural relaxation caused by the steric effect was observed
1
2
3
4
5
6
7
8
larger than that (2.33 D) of the S0 state. Moreover, as shown
by its rather localized frontier orbitals, i.e., electron density
distribution in Nꢀfluorenꢀ3ꢀylꢀN,N′ꢀdihydropyrazine and pheꢀ
nanthreneꢀfused ring for HOMO and LUMO, respectively (see
Table 4), the charge transfer property of the I* state is apparꢀ
ent, the dipole moment of which is calculated to be as large as
17.84 D. On the other hand, for the P* state, due to the planarꢀ
ization of the phenazine chromophore, the extensive πꢀ
electron delocalization significantly reduces the ππ* energy
gap, which leads to the dispersal of the local charge transfer
character. This is evidenced by the drastically different contriꢀ
bution of the frontier orbitals (HOMO/LUMO) for FlPAC in
P*, in which the population of electron density is spread over
the entire πꢀframework of the phenanzineꢀfused ring rather
than localized at the Nꢀfluorenꢀ3ꢀyl side chain as in the case of
R*(136°) and I*(133°). Additional support is the relatively
small dipole moment calculated for P* (6.74 D) and the nearly
solvent polarity independence of the P* emission peak waveꢀ
length (Figure 2d), despite its anomalously large Stokes shift.
with the same trend (see Figure 7, vide supra).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
As for DPAC, the PES simulation scanning along Θb (Figꢀ
ure S10 in SI) in the S1 state also resulted in a local minimum
I* at Θb = 134° and a global minimum P* at Θb = 161°. For I*,
the corresponding localized frontier orbitals for HOMO (at
N,N′ꢀdiphenylꢀdihydropyrazine) and LUMO (at phenanthreneꢀ
fused ring, Table S3), together with the calculated large dipole
moment (~12.40 D), also led to the conclusion of its charge
transfer property. The charge transfer character is also apparꢀ
ent from the vertical excitation of the geometry optimized
ground state (R*, Θb = 138°). The smaller dipole moment of
R* for DPAC (12.34 D, cf. FlPAC (15.24 D)) may qualitaꢀ
tively be rationalized by the weaker electron donating ability
of benzene (DPAC) than fluorene in FlPAC. The global minꢀ
imum P* (Θb = 161°) of DPAC also possesses a planar pheꢀ
nanthrene moiety that extends the ππ* (HOMO/LUMO) conꢀ
figuration (see Table S3 of SI for frontier orbital analysis) with
a smaller dipole moment (7.80 D).
Figure 9. The DFT optimized geometry of FlPAC (a) in the S0
ground state at the global minimum (Θb = 136°), (b) in the S1
excited state at the local minimum (Θb = 133°), and (c) in the S1
excited state at the global minimum (Θb = 160°), and three views
together with the selected angles.
2.6 Discussion
In this study, we have performed comprehensive studies on
steadyꢀstate polarity, viscosity and temperature dependent
spectroscopy, nanosecond to femtosecond timeꢀresolved specꢀ
troscopy, and dynamics in various viscosity and polarity enviꢀ
ronments. Together with the computational approach elaboratꢀ
ed in the above section, we are now able to propose a threeꢀ
step kinetics mechanism using FlPAC as a paradigm. Starting
from excitation of the global minimum S0 (Θb = 136°) in the
ground state, Figure 10 only depicts internal structural relaxaꢀ
tion in the S1 state, in which the PES along the planarization
angle Θb is singled out, while the rest of the coordinates releꢀ
vant to structural motions are simply summed up and denoted
as φ. The mechanical force resulting from viscosity and solꢀ
vent polarity effects, which may affect the corresponding reꢀ
laxation dynamics, are not shown in this plot. Also, limited by
the 3ꢀD plot, it is not possible to draw the solvent polarization
coordinate in Figure 10, so internal conversion (the vibrational
relaxation, VR) and solvent relaxation (SR) are simply comꢀ
bined in the initial structural relaxation and depicted by use of
a zigzag arrow symbol.
In addition to the difference in Θb among R*, I* and P*, alꢀ
so noted is the changes of space/orientation on the N,N′ꢀ
disubstituted side chain among the three states. For example,
from the R* state at Θb = 136° to the local minimum I* at Θb =
133°, the results of calculation for FlPAC, shown in Figure 9,
indicate the conformational change at the N,N′ꢀdiaryls site via
swinging and rotating round their jointed nitrogen atoms, i.e.,
an increase of ∠C3–N1–N2 from 118° to 127° and a decrease
of ∠C5–C3–N1–N2 from 100° to 88°. The results indicate that
FlPAC, along the planarization, may encounter steric hinꢀ
drance raised by the N,N′ꢀdisubstituted side chain. Thus, the
dynamics of structural relaxation should be affected by the
N,N′ꢀdisubstituted side chain motion, which is slowed down
upon increases in the bulky side chain with more steric hinꢀ
drance (Scheme 1 and Figure 1). Qualitatively, this viewpoint
could be supported by the computational approach. Upon
scanning Θb along the planarization (I* → P*), internal energy
barriers of 2.0 kcal/mol (Θb ~144°) and 0.7 kcal/mol (Θb ~141°)
were calculated for FlPAC and DPAC, respectively, while
virtually no barrier could be resolved for DMAC (see PES in
Figure S10 of SI). The results are consistent with the increasꢀ
ing bulkiness of the sideꢀchain in the order of DMAC <
DAPC < FlPAC. Meanwhile, the level/functional of current
computation did not allow us to identify any barriers along the
R* → I* process. Experimentally, however, the retardation of
Upon vertical excitation of FlPAC (Θb =136°), intramoꢀ
lecular charge transfer takes place simultaneously from the
N,N′ꢀdisubstituted chromophore to the phenazine moiety, deꢀ
noted by a charge transfer state R* (see Figure 10). The vibraꢀ
tionally hot and unfavorable polarization configuration is subꢀ
sequently subject to a rapid internal conversion and solvent
relaxation, the time scale of which may couple with the small
structural relaxation in the low viscous environment, reaching
a structurally relaxed local minimum I*. Note that amid the R*
→ I* process, despite the small variation of Θb, prominent
conformational change takes place at the N,N′ꢀdiaryls side
chains, which swing and rotate round their jointed nitrogen
atoms (coordinate φ in Figure 10). Subsequently, driven by the
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 13
thermodynamically favorable πꢀdelocalization over the benꢀ
zo[a,c]phenazine moiety, I* → P* planarization takes place.
Along the planarization, FlPAC (and DPAC) encounters steꢀ
ric hindrance raised by the N,N′ꢀdisubstituted side chain, reꢀ
sulting in a barrier at Θb ~144o (~141o for DPAC). The combiꢀ
nation of initial charge transfer state, intermediate, and the
final planarization state renders the full spectrum to account
for their anomalous emission and the associated kinetics. Usꢀ
ing a simplified sequential kinetic model R* → I* → P*, in
low viscous solvents, the rates of structure relaxation of R*
and I* are measured to be fast, 2–6 ps and 33–38 ps, respecꢀ
tively, in toluene or acetonitrile. Their associated steadyꢀstate
emission intensity is thus rather weak, for which the I* emisꢀ
sion is hindered under the intense P* population emission (see
Figure 2d), but has been resolved in the highly viscous solꢀ
vents as well as by the femtosecond fluorescence upꢀ
conversion experiment in toluene and acetonitrile.
change (146° → 170°), stabilization energy, and hence Stokesꢀ
shifted emission. However, similar to Huber11 and Schuster’s12
cryogenic experiments, we also observed a large Stokesꢀ
shifted ~470 nm fluorescence for DMP in the 77 K toluene
solid matrix (vide supra), which is quite different from all
other titled compounds that exhibit normal emission in the 77
K toluene solid matrix.
1
2
3
4
5
6
7
8
For a small molecule like DMP, we then wondered if the
prominent structural change (146° → 170°) could take place in
the excited state in the 77 K rigid toluene solid matrix. In this
regard, Wan and coꢀworkers have reported a nonplanar comꢀ
pound dibenz[b,f]oxepin with a large Stokesꢀshifted emission
both in solution and rigid condition.7a They thus proposed a
change from a nonplanar ground state to a more planar fluoꢀ
rescent excited singlet state for rationalization. Meanwhile,
they considered that the torsional motion might have a very
low activation barrier for dibenz[b,f]oxepin at 77 K in a 2ꢀ
propanol/ether glass. This explanation may be applicable to
DMP. However, more direct and convincing evidences are
needed to prove that large amplitude motion during planarizaꢀ
tion can still take place for DMP in the 77K rigid matrix. Supꢀ
port of this viewpoint unfortunately is still pending. Nevertheꢀ
less, it is evident that the symmetry rule imposed previously
by Huber and coꢀworkers should explain, in part, the observed
large Stokesꢀshifted emission for DMP.
It is also worthy to note that Schuster et al.12 also studied
DMAC and observed large Stokesꢀshifted emission (~600 nm
in benzene). Similar to DMP, they invoked the symmetry forꢀ
bidden S0 → S1 transition for rationalization. In this study, our
TDꢀDFT calculation gives an S0 → S1 oscillator strength of
0.07 (Table 3), which is considered to be a partially allowed
transition. In addition, the emission spectrum at ~520 nm of
DMAC in the 77 K toluene solid matrix is much blueꢀshifted
with respect to its room temperature emission (~610 nm, see
Figure 2e). Considering that the symmetry rule should not be
temperature/viscosity dependent, the imposition of forbidden
transition is not sufficient for rationalizing the large Stokesꢀ
shifted emission of DMAC. Therefore, a significant conforꢀ
mational change, i.e., the planarization of the benꢀ
zo[a,c]phenazines moiety, should also take place in the excited
state for DMAC (vide supra).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 10. The proposed structural relaxation mechanism of
FlPAC in the first electronically excited states along the skeletal
motions in toluene. VR: vibrational relaxation, SR: solvent relaxaꢀ
tion. Limited by the 3ꢀdimensional plot, it is not possible to draw
the solvent polarization coordinate. The SR process is only deꢀ
picted by a decrease in energy using a zigzag arrow symbol.
Finally, among the titled compounds investigated, we have
to point out the perhaps exceptional case, DMP, according to
previous report.11,12 In our computational approach, the S0 →
S1 (354 nm) transition for DMP is optically inaccessible (1A1
1
3. Conclusion
→ A2 in the C2 point group) with an oscillator strength (f) of
In summary, we report on the unusual photophysical behavꢀ
ior and comprehensive investigation of a series of saddleꢀ
shaped N,N′ꢀdisubstitutedꢀdihydrodibenzo[a,c]phenazine doꢀ
norꢀacceptor dyads (DMAC, DPAC and FlPAC), derived
from 5,l0ꢀdimethylphenazine (DMP) by extending one of the
phenyleneꢀfused rings to phenanthrene. Anomalously large
Stokesꢀshifted emission (up to ~12000 cmꢀ1), with slightly
solvent polarity dependence, has been observed for all comꢀ
pounds; additionally, a weak normal Stokesꢀshifted band with
an obvious solvatochromic effect has been found for the N,N′ꢀ
diarylꢀdihydrodibenzo[a,c]phenazines (DPAC and FlPAC).
While all of the compounds but DMP exhibit prominent norꢀ
mal Stokes shifted emission in rigid environments, the interꢀ
pretation based on the symmetry selection rule11 must be toned
down. Instead, combining comprehensive spectroscopy, timeꢀ
resolved measurements, and computational approaches, we
conclude that the multiꢀfluorescence occurs from a sequential
threeꢀstate electronic/structural relaxation process described
above. As depicted in Figure 10, the interplay among solvatoꢀ
chromism at the early relaxation stage, the structural planariꢀ
~0.0001, and the calculated S0 → S2 (327 nm) transition is
allowed with an f of ~0.4436. Similar result was obtained by
Huber and coꢀworkers11 using a semiꢀempirical approach.
Experimentally, the molar extinction coefficient (ε) of the
absorption band at 343 nm is measured to be 13,600 Mꢀ1 cmꢀ1,
which can be assigned to the allowed S0 → S2 transition.
Therefore, the calculated S0 → S1 (354 nm) forbidden transiꢀ
tion does not seem to be sufficiently low in energy to account
for the large Stokes shifted emission peak wavelength at 473
nm. Previously, it was not possible for Huber and coꢀworkers
to calculate the geometry relaxed S1 → S0 emission. To gain
inꢀdepth insight into the origin of the anomalous Stokes shift
of DMP emission, we have performed geometry optimization
for both the ground state and the excited (emitting) state. As a
result, Θb values for the most stable ground/excited states
(S0/S1) are calculated to be 146°/170° for DMP. Furthermore,
for DMP the vertical (FranckꢀCondon) S1 → S0 emission from
the structurally relaxed excited state is calculated to be 471 nm
(see Table 3), which matches the observed 473 nm emission.
If so, after excitation, DMP also shows prominent structural
10
ACS Paragon Plus Environment

Page 11 of 13
Journal of the American Chemical Society
The solvent used in this study was spectral grade and was
zation, i.e., elongation of πꢀconjugation, accompanied by the
1
used upon receipt (from Merck Inc.) without further purificaꢀ
tion. The degassed solvent was obtained by 3 freezeꢀpumpꢀ
thaw cycles.
side chain relaxation and hence the possible steric hindrance,
harnesses the emission properties. As a result, depending on
the surrounding environment, a wideꢀtuning emission color
from red to deep blue or even to white light generation can be
achieved. These unusual and unique characters make feasible
the titled phenazines and their future analogues for probing
environment rigidity or local viscosity, or even for the fabricaꢀ
tion of light devices wherein host materials such as polymers
can be intentionally rather soft so that decent viscosity can be
attained for ratiometric tuning of the multiple emission bands
(for example, see Figure S14 of SI). The associated geometry
relaxed phenomena of these titled phenazines thus provide a
new paradigm en route to versatile lighting materials.
2
3
4
5
6
7
8
9
Time-resolved emission spectra (TRES).19 TRES are
constructed in way that we started with measurement of the
timeꢀresolved decays at a number of designated wavelengths
across the emission spectrum, I(λ,t). The intensity decays are
wavelength dependent. In order to calculate the TRES one
computes a new set of intensity decays, which are normalized
so that the time integrated intensity at each wavelength is
equal to the steadyꢀstate intensity at that wavelength. This is
done by the following equation
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
F(λ)
H(λ) =
∞
Experimental Section
I(λ,t)dt
∫
Synthesis and Characterization. All reagents and solvents
were purchased from commercial sources and were of analytiꢀ
cal grade. Details of the synthetic procedures are provided in
SI as well as compound characterization.
1H and 13C NMR spectra were recorded on a Bruker AM
400 spectrometer, operating at 400 and 100 MHz, respectively.
Highꢀresolution mass spectrometry (HRMS) spectra were recꢀ
orded with a Waters electrospray ionization (ESI) mass specꢀ
troscope. The intensity data of Xꢀray diffraction were collectꢀ
ed on a Rigaku RAXIS RAPID IP imaging plate system with
MoKα radiation (λ = 0.71073 Å).
0
where F(λ) is the steadyꢀstate emission spectrum. H(λ) is the
volume, which when multiplied by the timeꢀintegrated intensiꢀ
ty at the same wavelength is equal to the steadyꢀstate intensity
at that wavelength. Accordingly, the normalized intensity deꢀ
cay functions can be expressed as
, which
′
I (λ,t) = H(λ)I(λ,t)
can be used to calculate the intensity at any wavelength and
time to construct actual TRES.
Computational Methodology. For all title molecules, the
geometries of the ground state were optimized by the density
functional theory (DFT) method, and excited state structures
with related photophysical properties were calculated by the
timeꢀdependent density functional theory (TDꢀDFT) method
with B3LYP hybrid function in combination with polarizable
continuum model (PCM). The 6ꢀ31+G(d,p) basis set was emꢀ
ployed for all atoms. In addition, for better understanding the
molecular conformational changes inꢀbetween the electroniꢀ
cally ground and first excited states, potential energy surfaces
(PES) for the first excited state and ground state of all title
molecules were computed and scanned along the bending anꢀ
gle, which was defined between plane 1 (C1, N1, and N2) and
plane 2 (N1, N2, and C2) as a constraint (see Scheme 2). We
progressively scanned the bending angle spanning from 100 to
180 degrees for every ten degrees, and applied finer scanning
steps per degree between 130–170 degrees. All theoretical
calculations were performed using the Gaussian 09 program.20
Spectroscopic Measurements. Steadyꢀstate absorption and
emission spectra were recorded by a Hitachi (Uꢀ3310) specꢀ
trophotometer and an Edinburgh (FS920) fluorimeter, respecꢀ
tively, but temperatureꢀdependent fluorescence measurements
of the samples, from 293 K to 133 K in nꢀbutanol, were colꢀ
lected by an optical assembly at the entrance slit of a polyꢀ
chromator (blazed at 150 nm) coupled with a sensitive image
charge coupled detector (ICCD, Princeton Instruments, PIꢀ
MAX3).
A Linkam FTIR 600 liquidꢀnitrogen
cooling
stage was employed to hold the sample and control the temꢀ
perature.
Detailed timeꢀresolved spectroscopic measurements have
been reported previously elsewhere.18 In brief, nanosecond
timeꢀresolved studies were performed with an Edinburgh FL
900 timeꢀcorrelated single photonꢀcounting (TCSPC) system
with a pulsed hydrogenꢀ or nitrogenꢀfilled lamp as the excitaꢀ
tion light source. Data were fitted with sum of exponential
functions using the nonlinear leastꢀsquares procedure in comꢀ
bination with the convolution method.
ASSOCIATED CONTENT
Supporting Information. Additional crystallographic (including
CIF), computational, and spectroscopic data along with complete
reference are provided. This material is available free of charge
via the Internet at http://pubs.acs.org.
Ultrafast spectroscopic studies were performed with a
femtosecond fluorescence upꢀconversion system (FOG100,
CDP) pumped by the same femtosecond oscillator (Tsunami,
SpectraꢀPhysics) with a central output wavelength at 760 nm.
In this measurement, fluorescence from a rotating sample cell,
following the excitation at 380nm (SHG of 760 nm), was colꢀ
lected, focused, and frequency summed in a BBO crystal,
along with an interrogation gate pulse at a designated delay
time with respect to the pump pulse. A λ/2 waveplate was used
to set the pump polarization at the magic angle (54.7°) with
respect to gate pulse for preventing the fluorescence anisotroꢀ
py contributed by solute reorientation. Fluorescence upꢀ
conversion data were fitted to the sum of exponential funcꢀ
tions convoluted with the instrument response function (IRF).
The IRF was determined from the Raman scattering signal and
its profile was fitted to a Gaussian function with full width at
half maximum of ~150 fs.
AUTHOR INFORMATION
Corresponding Author
Email: tianhe@ecust.edu.cn (H. Tian);
Email: chop@ntu.edu.tw (P.ꢀT. Chou).
Author Contributions
Z. Zhang, and Y.ꢀS. Wu contributed equally.
Notes
The authors declare no competing financial interests.
ACKNOWLEDGMENT
11
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 13
P.ꢀT. Chou thanks National Science Council, Taiwan for the fiꢀ
7. (a) Shukla, D.; Wan, P., J. Am. Chem. Soc. 1993, 115, 2990ꢀ
1
2
3
4
nancial support. H. Tian thanks National 973 Program
(2013CB733700) and NSFC/China.
2991; (b) Yuan, C.; Saito, S.; Camacho, C.; Irle, S.; Hisaki, I.;
Yamaguchi, S., J. Am. Chem. Soc. 2013, 135, 8842ꢀ8845; (c)
Yuan, C.; Saito, S.; Camacho, C.; Kowalczyk, T.; Irle, S.;
Yamaguchi, S., Chem. Eur. J. 2014, 20, 2193ꢀ2200.
REFERENCES
8. (a) Holzapfel, M.; Lambert, C.; Selinka, C.; Stalke, D., J.
Chem. Soc., Perkin Trans. 2 2002, 1553ꢀ1561; (b) Miao, Q.;
Nguyen, T.ꢀQ.; Someya, T.; Blanchet, G. B.; Nuckolls, C., J. Am.
Chem. Soc. 2003, 125, 10284ꢀ10287; (c) Okamoto, T.; Terada, E.;
Kozaki, M.; Uchida, M.; Kikukawa, S.; Okada, K., Org. Lett.
2003, 5, 373ꢀ376; (d) Terada, E.; Okamoto, T.; Kozaki, M.;
Masaki, M. E.; Shiomi, D.; Sato, K.; Takui, T.; Okada, K., J. Org.
Chem. 2005, 70, 10073ꢀ10081; (e) Suzuki, S.; Takeda, T.;
Kuratsu, M.; Kozaki, M.; Sato, K.; Shiomi, D.; Takui, T.; Okada,
K., Org. Lett. 2009, 11, 2816ꢀ2818; (f) Wang, C.; Dong, H.; Hu,
W.; Liu, Y.; Zhu, D., Chem. Rev. 2011, 112, 2208ꢀ2267; (g)
Zheng, Z.; Dong, Q.; Gou, L.; Su, J.ꢀH.; Huang, J., J. Mater.
Chem. C 2014, 2, 9858ꢀ9865; (h) Huang, W.; Sun, L.; Zheng, Z.;
Su, J.; Tian, H., Chem. Comm. 2015, 51, 4462ꢀ4464.
9. (a) Hiraoka, S.; Okamoto, T.; Kozaki, M.; Shiomi, D.; Sato,
K.; Takui, T.; Okada, K., J. Am. Chem. Soc. 2004, 126, 58ꢀ59; (b)
Masuda, Y.; Kuratsu, M.; Suzuki, S.; Kozaki, M.; Shiomi, D.;
Sato, K.; Takui, T.; Hosokoshi, Y.; Lan, X.ꢀZ.; Miyazaki, Y., J.
Am. Chem. Soc. 2009, 131, 4670ꢀ4673; (c) Masuda, Y.; Takeda,
H.; Kuratsu, M.; Suzuki, S.; Kozaki, M.; Shiomi, D.; Sato, K.;
Takui, T.; Okada, K., Pure Appl. Chem. 2010, 82, 1025ꢀ1032.
10. (a) Arimori, S.; Bell, M. L.; Oh, C. S.; James, T. D., Org.
Lett. 2002, 4, 4249ꢀ4251; (b) Boden, B. N.; Jardine, K. J.; Leung,
A. C.; MacLachlan, M. J., Org. Lett. 2006, 8, 1855ꢀ1858; (c) Feng,
X.; Pisula, W.; Müllen, K., Pure Appl. Chem. 2009, 81, 2203ꢀ
2224.
11. Morris, J.; Brühlmann, U.; Serafimov, O.; Huber, J. R., Ber.
Bunsen. Phys. Chem. 1974, 78, 1348ꢀ1353.
12. Schuster, G. B.; Schmidt, S. P.; Dixon, B. G., J. Phys.
Chem. 1980, 84, 1841ꢀ1843.
13. (a) Smith, J. G.; Levi, E. M., J. Organomet. Chem. 1972, 36,
215ꢀ226; (b) Schmidt, S. P.; Schuster, G. B., J. Am. Chem. Soc.
1980, 102, 306ꢀ314.
14. Montalti, M.; Credi, A.; Prodi, L.; Gandolfi, M. T.,
Handbook of photochemistry. CRC press: 2006.
15. KorppiꢀTommola, J. E. I.; Hakkarainen, A.; Hukka, T.;
Subbi, J., J. Phys. Chem. 1991, 95, 8482ꢀ8491.
16. Bahri, M. A.; Heyne, B. J.; Hans, P.; Seret, A. E.;
MouithysꢀMickalad, A. A.; Hoebeke, M. D., Biophys. Chem. 2005,
114, 53ꢀ61.
17. Meijer, G.; Berden, G.; Meerts, W. L.; Hunziker, H. E.; de
Vries, M. S.; Wendt, H. R., Chem. Phys. 1992, 163, 209ꢀ222.
18. (a) Chou, P.ꢀT.; Chen, Y.ꢀC.; Yu, W.ꢀS.; Chou, Y.ꢀH.; Wei,
C.ꢀY.; Cheng, Y.ꢀM., J. Phys. Chem. A 2001, 105, 1731ꢀ1740; (b)
Chou, P.ꢀT.; Pu, S.ꢀC.; Cheng, Y.ꢀM.; Yu, W.ꢀS.; Yu, Y.ꢀC.;
Hung, F.ꢀT.; Hu, W.ꢀP., J. Phys. Chem. A 2005, 109, 3777ꢀ3787.
19. Lakowicz, J. R., Principles of fluorescence spectroscopy.
Springer Science & Business Media: 2007.
1. (a) Doroshenko, A. O., Theor. Exp. Chem. 2002, 38, 135ꢀ155;
(b) Vollmer, F.; Rettig, W.; Birckner, E., J. Fluoresc. 1994, 4, 65ꢀ
69.
5
6
7
8
2. (a) Chien, Y.ꢀY.; Wong, K.ꢀT.; Chou, P.ꢀT.; Cheng, Y.ꢀM.,
Chem. Commun. 2002, 2874ꢀ2875; (b) Wong, K.ꢀT.; Ku, S.ꢀY.;
Cheng, Y.ꢀM.; Lin, X.ꢀY.; Hung, Y.ꢀY.; Pu, S.ꢀC.; Chou, P.ꢀT.;
Lee, G.ꢀH.; Peng, S.ꢀM., J. Org. Chem. 2005, 71, 456ꢀ465; (c)
Terenziani, F.; Painelli, A.; Katan, C.; Charlot, M.; Blanchardꢀ
Desce, M., J. Am. Chem. Soc. 2006, 128, 15742ꢀ15755; (d)
Marini, A.; MuñozꢀLosa, A.; Biancardi, A.; Mennucci, B., J. Phys.
Chem. B 2010, 114, 17128ꢀ17135; (e) Liu, C.; Tang, K.ꢀC.; Zhang,
H.; Pan, H.ꢀA.; Hua, J.; Li, B.; Chou, P.ꢀT., J. Phys. Chem. A
2012, 116, 12339ꢀ12348.
3. (a) Sension, R. J.; Repinec, S. T.; Szarka, A. Z.; Hochstrasser,
R. M., J. Chem. Phys. 1993, 98, 6291ꢀ6315; (b) Gepshtein, R.;
Leiderman, P.; Huppert, D.; Project, E.; Nachliel, E.; Gutman, M.,
J. Phys. Chem. B 2006, 110, 26354ꢀ26364; (c) Stavrov, S. S.;
Solntsev, K. M.; Tolbert, L. M.; Huppert, D., J. Am. Chem. Soc.
2006, 128, 1540ꢀ1546; (d) Gil, M.; Waluk, J., J. Am. Chem. Soc.
2007, 129, 1335ꢀ1341; (e) Zhu, L.; Li, X.; Zhang, Q.; Ma, X.; Li,
M.; Zhang, H.; Luo, Z.; Ågren, H.; Zhao, Y., J. Am. Chem. Soc.
2013, 135, 5175ꢀ5182.
4. (a) Pullen, S. H.; Anderson, N. A.; Walker, L. A.; Sension, R.
J., J. Chem. Phys. 1998, 108, 556ꢀ563; (b) Tang, K.ꢀC.; Rury, A.;
Orozco, M. B.; Egendorf, J.; Spears, K. G.; Sension, R. J., J.
Chem. Phys. 2011, 134, 104503; (c) Kim, S.; Yoon, S.ꢀJ.; Park, S.
Y., J. Am. Chem. Soc. 2012, 134, 12091ꢀ12097.
5. (a) Waluk, J., Conformational analysis of molecules in
excited states. WileyꢀVCH: 2000; (b) Tanner, C.; Manca, C.;
Leutwyler, S., Science 2003, 302, 1736ꢀ1739; (c) Huynh, M. H.
V.; Meyer, T. J., Chem. Rev. 2007, 107, 5004ꢀ5064; (d) Fang, C.;
Frontiera, R. R.; Tran, R.; Mathies, R. A., Nature 2009, 462, 200ꢀ
204; (e) Hsieh, C.ꢀC.; Jiang, C.ꢀM.; Chou, P.ꢀT., Acc. Chem. Res.
2010, 43, 1364ꢀ1374; (f) Hsieh, C.ꢀC.; Chou, P.ꢀT.; Shih, C.ꢀW.;
Chuang, W.ꢀT.; Chung, M.ꢀW.; Lee, J.; Joo, T., J. Am. Chem. Soc.
2011, 133, 2932ꢀ2943; (g) Demchenko, A. P.; Tang, K.ꢀC.; Chou,
P.ꢀT., Chem. Soc. Rev. 2013, 42, 1379ꢀ1408.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
6. (a) Valat, P.; Wintgens, V.; Kossanyi, J.; Biczok, L.;
Demeter, A.; Berces, T., J. Am. Chem. Soc. 1992, 114, 946ꢀ953;
(b) Demeter, A.; Bérces, T.; Biczók, L.; Wintgens, V.; Valat, P.;
Kossanyi, J., J. Phys. Chem. 1996, 100, 2001ꢀ2011; (c) Grabner,
G.; Rechthaler, K.; Köhler, G., J. Phys. Chem. A 1998, 102, 689ꢀ
696; (d) Baumgarten, M.; Gherghel, L.; Friedrich, J.; Jurczok, M.;
Rettig, W., J. Phys. Chem. A 2000, 104, 1130ꢀ1140; (e) Inoue, Y.;
Jiang, P.; Tsukada, E.; Wada, T.; Shimizu, H.; Tai, A.; Ishikawa,
M., J. Am. Chem. Soc. 2002, 124, 6942ꢀ6949; (f) Keinan, S.; Hu,
X.; Beratan, D. N.; Yang, W., J. Phys. Chem. A 2006, 111, 176ꢀ
181; (g) Chen, Y.; Zhao, J.; Guo, H.; Xie, L., J. Org. Chem. 2012,
77, 2192ꢀ2206.
20. Gaussian 09, Revision A. 2; Gaussian, Inc.: Wallingford.
CT, 2009.
12
ACS Paragon Plus Environment

Page 13 of 13
Journal of the American Chemical Society
1
2
3
SYNOPSIS TOC.
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
13
ACS Paragon Plus Environment