Photophysical and electrochemical properties of thienylnaphthalimide dyes with excellent photostability

Article abstract of DOI:10.1021/jp502535n

The development of robust dyes is a highly important theme for any applications of dyes. Here we present photophysical and electrochemical characterization of a set of robust dyes based on the thienylnaphthalimide unit. The set is comprised of the thienylnaphthalimide derivatives with phenyl- (Ph-), 4-nitrophenyl- (NO₂Ph-), and 4-(diphenylamino)phenyl (Ph ₂NPh-) substituents as exemplars covering electron-withdrawing to electron-donating groups. The fluorescence quantum yields of the Ph-TNI increases as the solvent polarity increases, while that of Ph₂NPh-TNI showed the opposite trend. Changes in the rates of nonradiative decay were found to be a major factor for these contrasting behaviors. Cyclic voltammetry showed that the substituent effects were more apparent for the HOMO energies rather than the LUMO energies. Density functional theory calculations showed that the first singlet excited state of these compounds is a ¹π,π* state with a significant charge transfer character. Ph-TNI and Ph₂NPh-TNI are much more stable against photodegradation than coumarin and fluorescein dyes.

Full text of DOI:10.1021/jp502535n

Subscriber access provided by Swem Library | College of William and Mary & VIVA (Virtual Library of Virginia)
Article
Photophysical and Electrochemical Properties of
Thienylnaphthalimide Dyes with Excellent Photostability
Takatoshi Inari, Minori Yamano, Ayaka Hirano, Kosuke Sugawa, and Joe Otsuki
J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/jp502535n • Publication Date (Web): 24 Jun 2014
Downloaded from http://pubs.acs.org on June 28, 2014
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.
The Journal of Physical Chemistry A 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 40
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
Photophysical and Electrochemical Properties of
Thienylnaphthalimide Dyes with Excellent
Photostability
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
Takatoshi Inari, Minori Yamano, Ayaka Hirano, Kosuke Sugawa, and Joe Otsuki*
College of Science and Technology, Nihon University, 1-8-14 Kanda Surugadai, Chiyoda,
Tokyo 101-8308, Japan
S Supporting Information
ABSTRACT: The development of robust dyes is a highly important theme for any applications
of dyes. Here we present photophysical and electrochemical characterization of a set of robust
dyes based on thienylnaphthalimide unit. The set comprises of the thienylnaphthalimide
derivatives with phenyl- (Ph-), 4-nitrophenyl- (NO2Ph-), and 4-(diphenylamino)phenyl (Ph2NPh-
) substituents as exemplars covering electron-withdrawing to electron donating groups. The
fluorescence quantum yields of the Ph-TNI increases as the solvent polarity increases, while that
of Ph2NPh-TNI showed the opposite trend. Changes in the rates of nonradiative decay were
found to be a major factor for these contrasting behaviors. Cyclic voltammetry showed that the
substituent effects were more apparent for the HOMO energies rather than the LUMO energies.
Density functional theory calculations showed that the first singlet excited state of these
ACS Paragon Plus Environment
1

The Journal of Physical Chemistry
Page 2 of 40
1
2
3
4
compounds is a 1π,π* state with a significant charge transfer character. Ph-TNI and Ph2NPh-TNI
5
6
7
are much more stable against photodegradation than coumarin and fluorescein dyes.
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
Keywords: Absorption, Excited State, Fluorescence, Fluorophore, Push-pull
INTRODUCTION
Organic fluorescent dyes are used for a wide range of applications including lasers,1 sensors,2-4
and solar cells,5-6 and so on. Rylene dyes have attracted attention since they possess prominent
chemical, thermal, and photo stability.7 Rylenes are based on naphthalene units linked in the peri
positions. They are often used in the forms of di- or tetra-carboxylic acids and their derivatives.7-9
1,8-naphthalic acid is the simplest member of the rylene dicarboxylic acid series.6, 10 Fluorescent
dyes based on 1,8-naphthalic acid may have the following advantages. (i) The parent structure is
simple enough, which makes it possible to synthesize a wide variety of derivatives with
minimum cost. (ii) It can be assumed that they are less prone to π-π aggregation compared to
larger rylene derivatives. (iii) The dicarboxylic acid group can be used to tag the fluorophore
through the formation of the imide bond to another functional fragment such as a unit for
molecular recognition. Indeed, 1,8-naphthalimide (NI) based chromophores have been developed
for various applications including DNA-oxidizing agents,11-12 pH indicators,13-14 sensors for
various cationic and anionic species,15-27 and emitters for organic light-emitting diodes.28 As the
π-system of NI is relatively small, it only absorbs at ~340 nm in the UV region,29 although
substitution by an amino group at the 4-position brings the absorption up to ~420 nm. It is
desirable to bring the absorption further to the red so that visible light can be used to excite the
dye and autofluorescence or background fluorescence can be avoided in the case of biological
ACS Paragon Plus Environment
2

Page 3 of 40
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
and materials applications. It is also desirable to develop a multiple-color set of dyes to enable
multicolored experiments as well as Förster-type resonance energy transfer experiments.
Thus we set out to prepare longer-wavelength absorbing dyes on the basis of NI core. To make
the compound to absorb light in a longer wavelength region, we decided to extend the π-
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
conjugation of the NI core with a thiophene unit rather than with use of perylene-3,4-
dicarboximide. Potential advantages of this choice includes a higher solubility or less
aggregation tendency, and more facile synthesis of 4-thienyl-1,8-naphthalimide and its
derivatives than perylene-3,4-dicarboximide and its derivatives. Thiophene and oligothiophenes
are chemically stable and have attracted attention as functional π-systems for optoelectronic
applications.10, 30 Despite the desirable properties of thiophene, only a few investigations have
been reported for thienylnaphthalimide systems.31-34 Our basic design for the dyes in this study is
presented in Figure 1, in which a set of "tuning" groups were introduced into the parent
thienylnaphthalimide structure. For the tuning, phenyl (Ph-TNI), 4-nitrophenyl (NO2Ph-TNI),
and 4-(diphenylamino)phenyl (Ph2NPh-TNI) groups were introduced, which covers from
electron-withdrawing to electron-donating groups. Here we report the preparation, photophysical
and redox properties, and photostability of these new dyes. While photophysical properties
including solvent effects were extensively investigated for the parent NI and its derivatives, in
which substituents are directly introduced onto the naphthalene core,35-42 neither detailed
photophysics nor solvent effects have been reported for π-extended naphthalimide chromophores
as far as we are aware of.
ACS Paragon Plus Environment
3

The Journal of Physical Chemistry
Page 4 of 40
1
2
3
4
O
5
6
7
8
S
N R
O
R = 2,5-di-t-butylphenyl
X
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 tuning
For tagging
Electron withdrawing
π-System
Ph-TNI: X- = H-
NO2Ph-TNI: X- = O2N-
Ph2NPh-TNI: X- = Ph2N-
Figure 1. Thienylnaphthalimide dyes.
EXPERIMENTAL SECTION
Synthesis in General. Dehydrated tetrahydrofuran was purchased from Kanto Chemical. Flash
column chromatography was carried out using Kanto Chemical Silica Gel 60. Gel permeation
chromatography was performed using a Japan Analytical Industry LC-9201 system equipped
with columns, Jaigel-1H and Jaigel-2H. The retention volume VR is given for each compound. 1H
13
and C NMR spectra were measured on a JEOL JNM-ECX400 spectrometer (400 MHz for H
1
13
and 100 MHz for C). Atmospheric pressure chemical ionization (APCI) high-resolution mass
spectrometry (HRMS) were conducted on an Agilent LC/MCD TOF 1100 spectrometer. The
dyes were prepared according to Scheme 1. N-(2,5-di-t-butylphenyl)-4-bromo-1,8-
napththalimide was prepared from commercial 4-bromo-1,8-naphthalic anhydride according to a
literature procedure.43 All the reactions for the final steps described below were carried out in air.
ACS Paragon Plus Environment
4

Page 5 of 40
The Journal of Physical Chemistry
1
2
3
4
5
O2N
N
6
7
8
9
bromobenzene,
p-bromonitrobenzene,
or
Br
S
S
S
4-bromotriphenylamine
Br
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
S
Bu3Sn
SnBu3
2,5-di-t -butylaniline
O
N
O
t Bu
Zn(OAc)2
quinoline
200 °C, 18 h
80%
PdCl2(PPh3)2
THF, refulx, overnight
O
N
R
O
O
N
R
O
O
N
R
O
O
O
O
t Bu
Ph-TNI 26%
NO2Ph-TNI 8%
Ph2NPh-TNI 5%
R = 2,5-di-t -butylphenyl
Scheme 1. Preparation of thienylnaphthalimides.
Ph-TNI. A mixture of N-(2,5-di-t-butylphenyl)-4-bromo-1,8-napththalimide (500 mg, 1.08
mmol), bromobenzene (345 mg, 1.15 mmol), 2,5-bis(tributylstannyl)thiophene (728 mg, 1.10
mmol), and dichlorobis(triphenylphosphine)palladium(II) (70 mg, 0.10 mmol) in tetrahydrofuran
(120 mL) was refluxed overnight. The solvent was evaporated, leaving a residue, which was
purified through column chromatography (SiO2, hexane/chloroform = 1/1) and GPC to afford a
yellow solid (132 mg, 0.285 mmol, 26%).
TLC (SiO2, CHCl3): Rf = 0.51. GPC: VR = 179 mL. 1H NMR (CDCl3): δ (ppm) = 1.30 (9H, s),
1.33 (9H, s), 7.01 (1H, d, J = 2.4 Hz), 7.33–7.40 (2H, m), 7.42–7.45 (4H, m), 7.59 (1H, d, J = 7.3
Hz), 7.71 (2H, m), 7.83 (1H, t, J = 8.7 Hz), 7.91 (1H, d, J = 7.8 Hz), 8.67 (1H, d, J = 7.8 Hz),
13
8.71 (1H, d, J = 8.7 Hz), 8.81 (1H, d, J = 8.7 Hz). C NMR (CDCl3): δ (ppm) = 165.3, 165.0,
150.1, 147.0, 143.9, 139.3, 139.1, 133.8, 132.9, 132.7, 131.9, 131.3, 130.14, 130.08, 129.4, 129.2,
128.8, 128.5, 128.3, 127.8, 127.4, 126.3, 126.0, 124.0, 123.5, 122.3, 35.6, 34.3, 31.8, 31.3．
APCI–HRMS (CHCl3): m/z calculated for C36H34NO2S ([MH]+), 544.2310; found, 544.2339.
Analysis: calcd for C36H33NO2S, C 79.52, H 6.12, N 2.58%; found, C 79.51, H 6.26, N 2.60%.
ACS Paragon Plus Environment
5

The Journal of Physical Chemistry
Page 6 of 40
1
2
3
4
5
6
NO2Ph-TNI. A mixture of N-(2,5-di-t-butylphenyl)-4-bromo-1,8-naphthalimide (354 mg,
0.752 mmol)，2,5-bis(tributylstannyl)thiophene (500 mg, 0.762 mmol), p-bromonitrobenzene
7
8
9
(152 mg, 0.752 mmol), and dichlorobis(triphenylphosphine)palladium(II) (49 mg, 0.077 mmol)
in tetrahydrofuran (120 mL) was refluxed overnight. The solvent was evaporated, leaving a
residue, which was purified through column chromatography (SiO2, hexane/chloroform = 1/1)
and GPC to afford a yellow solid (33 mg, 0.057 mmol, 7.5%).
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
TLC (SiO2, CHCl3): Rf = 0.46. GPC: VR = 179 mL. 1H NMR (CDCl3): δ (ppm) = 1.30 (9H, s),
1.33 (9H, s), 7.01 (1H, d, J = 2.3 Hz), 7.42–7.50 (2H, m), 7.60 (1H, d, J = 8.7 Hz), 7.63 (1H, d, J
= 4.1 Hz), 7.80–7.89 (3H, m), 7.92 (1H, d, J = 7.8 Hz), 8.31 (2H, d, J = 8.7 Hz), 8.69 (1H, d, J =
7.3 Hz), 8.73 (2H, d, J = 8.2 Hz). 13C NMR (CDCl3): δ (ppm) = 165.2, 164.9, 150.2, 147.1, 143.9,
143.7, 142.0, 139.9, 138.4, 132.9, 132.3, 132.1, 131.2, 130.5, 130.1, 129.4, 128.9, 128.8, 127.8,
127.7, 126.5, 126.4, 126.2, 124.7, 123.6, 122.9, 35.6, 34.3, 31.8, 31.3. APCI–HRMS (CHCl3):
m/z calcd for C36H33N2O4S ([MH]+), 589.2161; found, 589.2136.
Ph2NPh-TNI. A mixture of N-(2,5-di-t-butylphenyl)-4-bromo-1,8-naphthalimide (348 mg,
0.750 mmol), 4-bromotriphenylamine (249 mg, 0.768 mmol), 2,5-bis(tributylstannyl)thiophene
(500 mg, 0.755 mmol), and dichlorobis(triphenylphosphine)palladium(II) (48 mg, 0.077 mmol)
in tetrahydrofuran (120 mL) was refluxed overnight. The solvent was evaporated, leaving a
residue, which was purified through column chromatography (SiO2, hexane/chloroform = 1/1)
and GPC to afford a red solid (23 mg, 0.032 mmol, 4.3%).
TLC (SiO2, CHCl3): Rf = 0.56. GPC: VR = 171 mL. 1H NMR (CDCl3): δ (ppm) = 1.30 (9H, s),
1.33 (9H, s), 7.01 (1H, d, J = 2.3 Hz), 7.04–7.19 (8H, m), 7.26-7.33 (4H, m), 7.34–7.40 (2H, m),
7.45 (1H, dd, J = 8.7, 2.3 Hz), 7.51–7.62 (3H, m), 7.82 (1H, t, J = 8.2 Hz), 7.90 (1H, d, J = 7.8
Hz), 8.66 (1H, d, J = 7.8 Hz), 8.71 (1H, d, J = 7.3 Hz), 8.81 (1H, d, J = 8.7 Hz). 13C NMR
ACS Paragon Plus Environment
6

Page 7 of 40
The Journal of Physical Chemistry
1
2
3
4
(CDCl3): δ (ppm) = 165.3, 165.0, 150.1, 148.0, 147.4, 147.1, 143.9, 139.5, 138.1, 133.0, 132.7,
5
6
7
8
9
131.9, 131.3, 130.2, 130.1, 129.5, 128.8, 128.4, 127.8, 127.5, 127.3, 126.8, 126.3, 124.9, 123.5,
123.4, 123.1, 122.1, 35.6, 34.3, 31.8, 31.3. APCI–HRMS (CHCl3): m/z calcd for C48H43N2O4S
([MH]+), 711.3045; found, 711.3104.
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
Physical Measurements and Computation. Absorption spectra were measured on a
Shimadzu UV-2400 spectrometer. The molar absorption coefficients were obtained from the
absorbance of 10 µM solutions prepared by weighing about 1 mg with a Mettler-Toledo UMX2
ultra-micro balance, whose readability is 0.1 µg. Fluorescence spectra were taken on a Jasco FP-
8600 spectrometer. The spectrum of the excitation source was corrected using a SIC-844
standard detector and the wavelength-dependent sensitivity of the detector was corrected using a
ESC-842 standard halogen lamp. Thus, the ordinate of the fluorescence spectra represents the
relative number of photons per unit wavelength interval. Fluorescence quantum yields φsample
were obtained by comparing the fluorescence spectra of the samples with that of rhodamine 6G
(φref = 0.94 in ethanol44) using eq 1,45
Isample
ns2ample
nr2ef
1−10−A
1−10−A
ref
φsample = φref
(1)
sample
Iref
where I is the integrated fluorescence spectrum which is proportional to the number of emitted
photons, A is the absorbance at the excitation wavelength, and n is the refractive index of the
solvent. The subscript 'sample' means the compound of interest and 'ref' means the reference
compound, which is rhodamine 6G in the present case. Hereafter, the subscript 'sample' is
replaced by 'f', which stands for fluorescence. Fluorescence lifetimes were measured on a
Hamamatsu Photonics Quantaurus-Tau C11367 system. The samples were excited with a light-
ACS Paragon Plus Environment
7

The Journal of Physical Chemistry
Page 8 of 40
1
2
3
4
emitting diode at 405 nm and the fluorescence decay was monitored at λmax (full width at half
5
6
7
8
9
maximum: ca. 10 nm). Spectroscopic grade toluene, dimethylsulfoxide (DMSO), methanol
(MeOH), 1,4-dioxane, and acetonitrile (MeCN) were purchased from Kanto Chemical and used
as received. All photophysical measurements were performed using air-equilibrated solvents.
Quantum yield and lifetime measurements were repeated for at least three separately prepared
sample solutions. The error ranges and error bars represent the confidence interval of 68.3%
obtained from Student's t-distribution.
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
Cyclic voltammetry was measured using a Hokuto Denko HZ-3000 electrochemical analyzer.
A 1 mM sample solution in MeCN containing 0.1 M tetrabutylammonium hexafluorophosphate
(N(Bu)4PF6) was purged with N2. Pt electrodes were used as the working and counter electrodes.
A Ag electrode was used as the quasi-reference electrode. The scan rate was 100 mV s–1. The
potential were referenced to the internal ferrocene/ferricenium couple added after each
measurement. Anhydrous MeCN for the electrochemistry was purchased from Kanto Chemical.
For the photostability evaluation, an Asahi Spectra LAX-Cute Xe-lamp (100 W, 400–700 nm)
was used as the light source. Density functional theory (DFT) calculations were performed with
Gaussian 09M46 and GaussView run on a personal computer. The functional and basis set as well
as options used are indicated in the relevant places in the Results and Discussion section.
RESULTS AND DISCUSSION
Synthesis. The three compounds were synthesized as shown in Scheme 1. The mixed Stille
coupling
of
N-(2,5-di-t-butylphenyl)-4-bromo-1,8-naphthalimide,
2,5-
bis(tributylstannyl)thiophene, and one of bromobenzene, p-bromonitrobenene, and 4-
bromotriphenylamine followed by purification with silica gel column chromatography and gel
ACS Paragon Plus Environment
8

Page 9 of 40
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
permeation chromatography afforded Ph-TNI, NO2Ph-TNI, and Ph2NPh-TNI, respectively.
Identity and purity of the synthesized compounds were confirmed with 1H and 13C NMR, APCI–
HRMS, and GPC chromatograms; see Supporting Information for spectra and chromatograms.
Photophysical Properties. Experimental photophysical properties of the dyes are summarized
in Table 1, while the spectra are shown in Figure S1. The solvent used were toluene, DMSO, and
MeOH, as exemplars for nonpolar solvents, nonprotic polar solvents, and protic polar solvents,
respectively. The parent NI exhibits an absorption band peaking at ~330 nm in toluene.47 The
introduction of the phenylthienyl group makes the absorption peak significantly red-shifted by
~65 nm to 398 nm (Ph-TNI). Further introduction of a nitro group has little influence on the peak
position (392 nm for NO2Ph-TNI). Introduction of a diphenylamino group, on the other hand,
leads to a further red-shift by 42 nm, giving rise to the peak at 440 nm. The absorption maxima
of these compounds are hardly affected by the solvent polarity on going from toluene to MeOH,
as seen from the less than 8 nm changes.
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
Table 1. Photophysical Data for Ph-TNI, NO2Ph-TNI, and Ph2NPh-TNI
λab
ε × 10–4
/M–1 cm–1
1.53
λfl
Δν
φf
τf
Solvent
kr
knr
/ns–1
/ns–1
/cm–1
5423
6738
7331
4633
6413
6613
5336
–
/nm
398
405
403
392
400
393
440
445
/nm
493
557
572
479
538
531
575
–
/ns
Ph-TNI
Toluene
DMSO
MeOH
Toluene
DMSO
MeOH
Toluene
DMSO
0.06 ± 0.02
0.43 ± 0.02
0.34 ± 0.02
0.07 ± 0.02
0.17 ± 0.03
0.03 ± 0.00
0.56 ± 0.04
<0.01
0.21 ± 0.01
2.62 ± 0.05
2.53 ± 0.01
0.22 ± 0.01
0.55 ± 0.00
0.20 ± 0.01
2.60 ± 0.05
–
0.29 ± 0.10
0.16 ± 0.01
0.13 ± 0.01
0.32 ± 0.09
0.31 ± 0.05
0.15 ± 0.01
0.22 ± 0.02
–
4.28 ± 0.23
0.22 ± 0.01
0.26 ± 0.01
4.23 ± 0.21
1.51 ± 0.05
4.85 ± 0.24
0.17 ± 0.02
–
1.68
1.78
NO2Ph-TNI
Ph2NPh-TNI
2.95
2.73
2.11
1.83
1.52
ACS Paragon Plus Environment
9

The Journal of Physical Chemistry
Page 10 of 40
1
2
3
MeOH
447
1.35
–
–
<0.01
–
–
–
4
5
6
7
8
9
The parent NI emits fluorescence peaking at ~380 nm in toluene.47 Compared to this, extension
of the conjugate system as in Ph-TNI results in a large red shift in the fluorescence, which has a
peak at 493 nm. Further introduction of a nitro group (NO2Ph-TNI) slightly reduces this effect,
resulting in the fluorescence peaking at 479 nm. Introduction of a diphenylamino group, on the
other hand, results in a very large red shift, resulting in the fluorescence peaking at 575 nm.
In contrast to the absorption, the fluorescence from these compounds are very sensitive to the
solvent polarity both in terms of the transition energy and the quantum yield. For Ph-TNI, the
fluorescence red shifts on going from toluene (493 nm) to DMSO (557 nm), and further to
MeOH (572 nm). For NO2Ph-TNI, the fluorescence red shifts on going from toluene (479 nm) to
DMSO (538 nm), but a slight blue shift is observed on going from DMSO to MeOH (531 nm).
For Ph2NPh-TNI, the solvent dependent shift was not clearly observed because this compound
shows little fluorescence in DMSO and in MeOH. This point, therefore, was investigated in a
different solvent system as discussed later. It is noted that each of these compounds exhibits a
broad, structureless fluorescence band and none of them showed a dual fluorescence.48-49 The
absence of dual fluorescence is consistent with the single component exponential decays for
these compounds in time-resolved fluorescence measurements as described later.
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
General solvent effects may be approximated by Onsager's reaction field model, in which a
dipole is placed at the center of a vacuum cavity in a homogeneous dielectric medium.50-51 The
absorption (νab) and fluorescence (νfl) frequencies in this treatment are expressed by eqs 2 and 3,
2µG (µE − µG )
a3
hνab = hνa0b − Δf
(2)
ACS Paragon Plus Environment
10

Page 11 of 40
The Journal of Physical Chemistry
1
2
3
4
5
2µE (µE − µG )
a3
hνfl = hνf0l − Δf
(3)
(4)
6
7
8
9
where Δf is the orientation polarizability, which is defined in eq 4.
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
ε −1
2ε +1 2n2 +1
n2 −1
Δf =
−
Here, h is the Planck constant, νa0b and νf0l are the absorption and fluorescence frequency in the
vacuum, respectively, ε is the dielectric constant of the solvent, n is the refractive index of the
solvent, µG and µE are the dipole moments of the molecule in the ground and in the excited states,
respectively, and a is the radius of the cavity.52 The values of Δf of solvents used in this study are
0.013, 0.263, and 0.308 for toluene, DMSO, and MeOH, respectively. It should be noted that
these equations hold for molecules whose dipole moment in the ground state and that in the
excited state are oriented in the same direction. The assumption is likely to be valid for Ph-TNI
and Ph2NPh-TNI because the molecular structures, i.e., π-acceptor and donor-π-acceptor motifs,
respectively, are fairly straightforwardly related to polarization. However, the polarization
direction, in the ground state as well as in the excited state, is not obvious for NO2Ph-TNI
because of the acceptor–π-acceptor motif. The sensitivity of νab and νfl on Δf is represented by
the coefficients of Δf in eqs 2 and 3, respectively. Therefore, the observation that νab is almost
insensitive to the difference in Δf but that νfl becomes smaller with increasing Δf implies
qualitatively that the coefficient of Δf in eq 2 is small whereas that in eq 3 is large. This means
that the ratio µE/µG is significantly larger than one; the excited-state dipole moments are
significantly larger than the ground-state dipole moment for Ph-TNI and Ph2NPh-TNI.
ACS Paragon Plus Environment
11

The Journal of Physical Chemistry
Page 12 of 40
1
2
3
4
The dependence of the fluorescence quantum yield φf on the solvent is very characteristic to
5
6
7
8
9
the respective compound under study. Ph-TNI is weakly fluorescent in the nonpolar solvent
(0.07±0.02 in toluene) but strongly fluorescent in polar solvents (0.43±0.02 in DMSO; 0.34±0.02
in MeOH) whether the solvent is protic or nonprotic. NO2Ph-TNI is weakly fluorescent
irrespective of the solvent. Ph2NPh-TNI is highly fluorescent in the nonpolar solvent (0.56±0.04
in toluene) but nearly nonfluorescent in polar solvents (<0.01 in DMSO, MeOH) whether the
solvent is protic or nonprotic. The φf is expressed in eq. 5,
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
kr
φf =
(5)
kr + knr
where kr and knr are the radiative and nonradiative decay rate constants from the singlet excited
state, respectively. Thus a change in quantum yield can be ascribed to a change either in kr or knr.
These rate constants can be evaluated by measuring the excited-state (fluorescence) lifetime τf in
addition to the quantum yield as indicated by eqs 6 and 7.
φf
kr =
(6)
(7)
τf
1− φf
knr =
τf
We performed the time-resolved fluorescence measurements to obtain the lifetimes. Every trace
could be fitted by a single exponential decay with χ2 values in a range of 1.05–1.56. The data
thus obtained are shown in Figure S2 and the results are summarized in Table 1.
For Ph-TNI, it is now clear that the major reason for the increase of φf on going from nonpolar
to polar solvent is mainly due to a decrease in knr. The knr for Ph-TNI decreases nearly 20-fold
from 4.28±0.23 in toluene to 0.22±0.01 in DMSO and 0.26±0.01 in MeOH. For NO2Ph-TNI,
ACS Paragon Plus Environment
12

Page 13 of 40
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
changes in kr and knr are both modest. For Ph2NPh-TNI, the fluorescence in DMSO and in MeOH
were so weak that the determination of kr and knr was impossible in these solvents. Hence, the
polarity dependence of kr and knr for Ph2NPh-TNI was examined separately as described later.
Similar decrease in knr on going from nonpolar to polar environment has been observed for the
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
1
parent NI without an extended π system as well.35,36 The energies of the lowest π,π* state and
the lowest n,π* state are quite close for the parent NI.39 On going from nonpolar to polar
1
medium, the energy of the 1π,π* is lowered, while the energy of the 1n,π* state is raised, because
1
the π,π* state is more stabilized than the ground state by polar solvents and the opposite is the
1
1
case for the n,π* state. As a result, as the solvent polarity increases, the gap between the n,π*
1
state and the π,π* state widens. The literature agrees up to this point. As a consequence of the
widened gap, some authors, as the reason for the increase of φf in more polar solvents, invoked
1
less facile intersystem crossing due to the less contribution of the n,π* state to the first singlet
excited state (S1),36 while others invoked a smaller proximity effect.40 The proximity effect is a
1
1
nuclear motion-mediated coupling of the upper and lower electronic states, the n,π* and π,π*
states in this particular case.38 When the energies between the two electronic states are closer, the
lower state potential curve becomes wider due to the vibronic interaction. Consequently, the
Franck-Condon factor with the ground state becomes larger and hence knr also becomes larger.
1
These considerations may also be applied for Ph-TNI, because the S1 is a π,π* state with some
charge transfer character, which is more stabilized in more polar solvents, resulting in a widened
gap between this 1π,π* state and an upper lying 1n,π* state. It should be noted, however, that the
1
gap between the n,π* state and the π,π* state is much wider already in the vacuum or in
1
nonpolar solvents in the case of Ph-TNI than in the parent NI; see below in the section of DFT
ACS Paragon Plus Environment
13

The Journal of Physical Chemistry
Page 14 of 40
1
2
3
4
1
calculations. Apart from the solvent polarity-dependent widening of the gap between the n,π*
5
6
7
and 1π,π* states, an additional possibility of the solvent dependent change in knr is that the height
8
9
of the energy barrier on excited state potential energy surfaces is solvent dependent.53 This
barrier separates strongly emitting planar geometry and a low emitting, twisted geometry. For
Ph-TNI, the potentially rotatable bonds are the phenyl–thienylene bond and the thienylene–
naphthalimide bonds.
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
To probe more accurately the effects of the solvent polarity on the photophysical behaviors of
these compounds, the solvent polarity was incrementally changed by using a mixed solvent
system consisting of 1,4-dioxane (ε = 2.25; n = 1.42) and MeCN (ε = 37.5; n = 1.34) and
wavenumbers, νab and νfl , were recorded. The dielectric constants and the refractive indices of a
binary mixture, which are needed for obtaining Δf values, are approximated by eqs 8 and 9,
respectively,54
εbinary = x1ε1 + x2ε2
(8)
(9)
nbinary = x1n12 + x2n22
where x1 and x2 are the mole fractions of solvent 1 and solvent 2, respectively.
The results are summarized as the Lippert-type plot in Figure 2a, in which wavenumbers are
plotted in place of frequencies in eqs 2 and 3. All data are fitted on straight lines quite well,
demonstrating that the simple Onsager model is good enough to explain the polarization
dependent absorption and fluorescence changes. The absorption wavenumbers for all the
compounds show slightly negative slopes, while the fluorescence wavenumbers for all the
compounds show more pronounced negative slopes. A particularly large negative slope is
1
observed for Ph2NPh-TNI. These observations indicate that the fluorescence is due to π,π*
ACS Paragon Plus Environment
14

Page 15 of 40
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
transition with charge transfer character, whose contribution is particularly large for Ph2NPh-TNI.
These results are consistent with the solvent-dependent behaviors in toluene, DMSO, and MeOH
described above.
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
ACS Paragon Plus Environment
15

The Journal of Physical Chemistry
Page 16 of 40
1
2
3
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
(a)
(b)
(c)
ACS Paragon Plus Environment
16

Page 17 of 40
The Journal of Physical Chemistry
1
2
3
4
Figure 2. Dependence of photophysical properties on the orientation polarization Δf in a binary
5
6
7
8
9
solvent system consisting of 1,4-dioxane and MeCN. (a) Absorption and fluorescence maxima.
(b) Quantum yields. (c) Radiative and nonradiative rate constants for Ph2NPh-TNI.
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 dependence of the fluorescence quantum yield on Δf is shown in Figure 2b. The quantum
yield increases as the Δf increases for Ph-TNI, while the quantum yield is consistently low over
the range of Δf investigated for NO2Ph-TNI. The dependence of fluorescence quantum yield on
solvent polarity becomes now clear for Ph2NPh-TNI whose fluorescence is extremely weak in
DMSO and MeOH as described earlier. The quantum yield, starting from a high value of ca. 0.6
in 1,4-dioxane, gradually decreases as Δf increases, down to near zero values at Δf > 0.25. To see
whether this decrease is due to changes in the rate of radiative process or the nonradiative
process, kr and knr are plotted in Figure 2c. While kr decreases as Δf increases, a more prominent
increase in knr is apparent particularly at a large Δf region as the major reason for the decrease of
fluorescence quantum efficiency. Similar behaviors were also observed for N,N-dialkyl-4-amino-
1,8-naphthalimides, in which a dialkylamino group is directly introduced onto the 1,8-
naphthalimide core.41 The fluorescence from these parent compounds decreases on going from
nonpolar to polar solvent. The mechanism behind this behavior is not clear even for the parent
N,N-dialkyl-4-amino-1,8-naphthalimides. It is certain that the excited state is the strongly charge-
transferred state and hence is more stabilized in more polar solvents, as manifested in the
negative slope in the Lippert-type plot for νF of Ph2NPh-TNI. We just make a general remark
that the lower the energy of an excited state, knr becomes larger, which is in accord with the
energy-gap law.55
ACS Paragon Plus Environment
17

The Journal of Physical Chemistry
Page 18 of 40
1
2
3
4
5
6
7
8
9
Redox Properties. In addition to photophysical properties, redox properties of fluorophores
are important particularly for this class of compounds, because they could be used for
photoinduced electron-transfer sensors13-14 and DNA oxidation.11-12 The cyclic voltammograms
are presented in Figure 3. For Ph-TNI, two reduction waves appear at –1.59 V (vs
ferrocene/ferricenium) and –2.01 V, while an oxidation wave appears peaking at +1.17 V. The
reductions are reversible but the oxidation is irreversible. Each redox wave for NO2Ph-TNI (–
1.46 V and +1.63 V) is observed at more positive potentials than the corresponding wave for Ph-
TNI reflecting the electron withdrawing nature of the nitrophenyl substituent, except for the
second reduction wave, which was not observed under our conditions. The oxidation waves for
Ph2NPh-TNI (+0.50 V and +1.00 V) appear at less positive potentials than that for Ph-TNI,
reflecting the electron-donating nature of the triphenylamine unit. Both of these waves are now
(quasi)reversible. The reduction waves for Ph2NPh-TNI (–1.58 V and –2.17 V), however, appear
at potentials almost the same as those for Ph-TNI. In general, it is noticed from the
voltammograms that the differences for the reduction potentials among these compounds are
quite small and the effect of substituents are more manifested in the oxidation potentials.
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
ACS Paragon Plus Environment
18

Page 19 of 40
The Journal of Physical Chemistry
1
2
3
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
Figure 3. Cyclic voltammograms. Measurement conditions are as follows. Sample
concentration: 1 mM; electrolyte: 0.1 M N(Bu)4PF6 solution in MeCN; working electrode: Pt;
counter electrode: Pt; quasi-reference electrode: Ag; scan rate: 100 mV s–1. The potentials are
given versus the internal ferrocene/ferricenium couple.
DFT Calculations. DFT calculations were performed to probe the electronic structures of
these molecules both in the ground state and in the excited state. For the calculations, the
substituent on the imide nitrogen (the 2,5-t-butylphenyl group) was replaced by a hydrogen atom
to reduce the computation time. The ground state geometries were optimized using the B3LYP
hybrid functional56 combined with the 6-31+G* basis set, which contains the diffuse functions
and the polarization functions for atoms other than hydrogen. We then computed the transition
energies to the singlet excited electronic states on these ground-state geometries with TDDFT.
We compared the results of calculations obtained by using some different functionals, i.e.,
B3LYP,56 CAM-B3LYP,57 and BMK58 with the experimental data. The experimental absorption
ACS Paragon Plus Environment
19

The Journal of Physical Chemistry
Page 20 of 40
1
2
3
4
5
6
7
8
9
maxima for Ph-TNI, NO2Ph-TNI, and Ph2NPh-TNI are 398, 392, and 440 nm in toluene,
respectively (Table 1). The B3LYP/6-31+G* gave the values of 437, 425, and 555 nm,
respectively, significantly underestimating the transition energies (2000–4700 cm–1), particularly
for Ph2NPh-TNI, which is expected to have a large charge transfer character. The poor
performance of B3LYP, with its low fraction (20%) of the Hartree–Fock exchange, for charge
transfer excitations is well known.59-60 Increasing the fraction of the Hartree-Fock exchange in a
range-separated manner where the fraction of Hartree-Fock exchange depends on the electron–
electron distance (e.g., CAM-B3LYP) or in a manner through the kinetic energy density (e.g.,
BMK) alleviates the problem. For the present molecules, the CAM-B3LYP/6-31+G* gave the
values of 361, 359, and 390 nm as the transition wavelengths, respectively, significantly
overestimating the energies (2300–2900 cm–1). The BMK/6-31+G*, on the other hand, gave
reasonable values of 377, 371, and 429 nm, respectively, with deviations from experiments being
in a range of 500–1400 cm–1. Accordingly, we decided to use the BMK functional in the
following excited state calculations. The HOMOs and LUMOs obtained by BMK/6-31+G* are
displayed in Table 2 and the transition energies and the oscillator strengths are superimposed on
the experimental spectra in toluene in Figure 4.
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
ACS Paragon Plus Environment
20

Page 21 of 40
The Journal of Physical Chemistry
1
2
3
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
(a)
(b)
(c)
Figure 4. Oscillator strengths and absorption spectra. The oscillator strengths were obtained by
TDDFT calculations (BMK/6-31+G*) and the spectra are those in toluene. (a) Ph-TNI. (b)
NO2Ph-TNI. (c) Ph2NPh-TNI.
According to the TDDFT calculations, the lowest singlet excited states for all these
compounds have predominantly the HOMO→LUMO configuration. Thus, the S1 on the vertical
1
excitation are π,π* states. For Ph-TNI, the HOMO is largely delocalized over the entire π
system, while the LUMO is mainly localized on the naphthalimide unit. Therefore, the lowest
singlet excited state, which appears at 377 nm according to the TDDFT calculations and at 398
nm in the experiment in toluene, is a 1π,π* state with an appreciable charge transfer nature. The
ACS Paragon Plus Environment
21

The Journal of Physical Chemistry
Page 22 of 40
1
2
3
4
5
6
7
8
9
charge transfer nature is visualized by the difference map which is also displayed in Table 2. The
map was made by plotting the difference between the squares of the coefficients of the LUMO
and those of the HOMO. It is clearly seen from the difference map that the charge moves from
the phenylthienyl unit to the naphthalimide unit on going from the ground state to the S1.
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
ACS Paragon Plus Environment
22

Page 23 of 40
The Journal of Physical Chemistry
1
2
3
4
Table 2. LUMO, HOMO, and Difference Maps According to DFT (BMK/6-31+G*)
5
6
Ph–TNI
NO2Ph–TNI
Ph2NPh–TNI
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
LUMO
HOMO
S0→S1a
a On going from the S0 state to the S1 state, charge moves from the purple to the sky blue regions.
For NO2Ph-TNI, both the HOMO and LUMO are delocalized over the entire π system. It is an
interesting question which group withdraws electrons more strongly in this molecule, the nitro
group or the dicarboximide group. The calculated dipole moment of the ground state is oriented
ACS Paragon Plus Environment
23

The Journal of Physical Chemistry
Page 24 of 40
1
2
3
4
5
6
7
8
9
nearly perpendicularly with respect to the molecule's long axis, which indicates that the groups at
both ends of the molecule withdraw electrons nearly equally. This is also the case for the S1.
Both of the nitro group and the imide moiety at both ends of the molecule pull the electrons
nearly equally, as seen from the difference map. According to the TDDFT calculations, the
lowest singlet excited 1π,π* state is located at 371 nm.
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 Ph2NPh-TNI, a significant charge-transfer nature in the HOMO→LUMO transition is
apparent. The HOMO is localized on the diphenylaminophenylthienyl unit, while the LUMO is
mostly localized on the naphthalimide moiety. Therefore, the HOMO→LUMO transition is a
1π,π* transition with a large contribution of charge transfer. The charge transfer character is
clearly seen from the difference map, in which electrons move from the triphenylamine unit to
the naphthalimide unit. According to the TDDFT calculations, this transition to the S1 is located
at 429 nm.
The same set of calculations were performed on the parent NI as a reference to highlight the
effects of the extended π system of the present three dyes. The HOMO and LUMO of NI are π
and π* orbitals according to BMK/6-31+G*//B3LYP/6-31+G*. The n-orbital mainly localized
on the carbonyl oxygens is the HOMO–2 orbital lying below the HOMO by 1.17 eV. According
1
to the TDDFT BMK/6-31+G* calculations, the S1 is a π,π* state, in which an electron in the
HOMO is excited to the LUMO, with the transition at 308 nm. The second singlet excited state,
1
which is a n,π* state with a major configuration of HOMO–2 → LUMO, lies very close in
energy (298 nm) to the S1. This result indicating the closely separated 1π,π* and 1n,π*state agrees
with calculations in the literature.39
ACS Paragon Plus Environment
24

Page 25 of 40
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
To gain more insight into the excited states, we carried out TDDFT calculations focusing on
the excited states in different solvents. First, we optimized the excited state structures in toluene,
DMSO, and MeOH starting with the geometries optimized in vacuum as the initial input
structures. This was done by employing the polarizable continuum model and the self-consistent
reaction field options, which are implemented in the Gaussian 09 package with TDDFT
(BMK/6-31G). In short, the excited state in solution was found to be more planar than the ground
state in vacuum (B3LYP/6-31+G*). More specifically, the dihedral angle between the thienylene
ring and the naphthalimide plane is 47–50° for each of the three thienylnapththalimide
compounds in the ground state in vacuum, while this angle is reduced to 21–28° in the excited
state for every compound in every solvent. The dihedral angles between the thienylene ring and
the phenylene ring are 23–29° in the ground states in vacuum, while these angles are 3–5° in the
excited states in solution. Next, single point calculations were done on the optimized structure
using the state specific model with the key word ExternalIteration in the Gaussian packege.61-62
This model calculates the self-consistent reaction field with equilibrium solvation, in which both
the fast and slow components of the solvent polarization, which correspond to the electronic and
orientation polarization, respectively, are in equilibrium with the geometrical and electronic
structures of the solute molecule in the excited state.
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
Excited-state energies thus obtained are displayed in Figure 5. For NI, the lowest singlet
excited state is a 1π,π* state and the second lowest singlet excited state is a 1n,π* state. These two
states are separated by 0.28 eV in toluene. Upon increasing the solvent polarity from toluene to
1
DMSO or MeOH, the π,π* state is stabilized while the n,π* state is destabilized, resulting in
1
the widened gap up to 0.42 eV. Although the energies in DMSO and in MeOH are quite close,
we should note that specific interactions, particularly hydrogen bonding, are not taken into
ACS Paragon Plus Environment
25

The Journal of Physical Chemistry
Page 26 of 40
1
2
3
4
account in the present model calculations. As discussed above, the larger separation in energy
5
6
1
1
between the π,π* state and the n,π* states in a more polar environment will lead to a lower
7
8
9
nonradiative decay rate whether it is due to a less facile intersystem crossing or due to a smaller
proximity effect.
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 present three thienylnaphthalimide compounds, the general characteristics in the
1
solvent dependency are similar to those for NI; (i) the lowest singlet excited state S1 is a π,π*
1
state, (ii) the S1 gets lower in energy in more polar solvents, (iii) the n,π* state gets higher in
energy in more polar solvent. However, the S1 energies of the present three compounds are much
lower in energy than that of NI, resulting in much wider gaps between the 1n,π* and 1π,π* states
even in less polar toluene. Experimentally, the solvent dependence of fluorescence intensity is
the same for NI35-37, 40 and Ph-TNI; more intense fluorescence in more polar solvents. Thus, the
argument applied to NI may also be applied to Ph-TNI, with a caveat that the gap between the
1
1
1n,π* and π,π* states is already wide in toluene: less involvement of n,π* states in more polar
1
solvents leads to less nonradiative decay rates. For NO2Ph-TNI, a n,π* state involving not the
carbonyl moieties but the nitro group is found as the S2 or S3 states depending on the solvent. The
presence of this low-lying 1n,π* state could prompt the nonradiative process, which may account
for the relatively large nonradiative decay rates and low fluorescence yields for NO2Ph-TNI. For
Ph2NPh-TNI, the S1 state is even lower in energy than the other compounds. Just contrary to NI
and Ph-TNI, the nonradiative decay rates for Ph2NPh-TNI as well as for N,N-dialkyl-4-amino-
1,8-naphthalimide39-42 increase in more polar environments (Figure 2c). As described earlier, no
definite mechanism for this behavior seems to have been presented even for the simpler N,N-
dialkyl-4-amino-1,8-naphthalimide in the literature. Further complication arises in the case of
ACS Paragon Plus Environment
26

Page 27 of 40
The Journal of Physical Chemistry
1
2
3
4
5
6
7
8
9
Ph2NPh-TNI, because this molecule has five rotatable bonds. While the present optimization was
done for the nearly planar conformation, the excited-state molecule may sweep over the potential
energy surface and could find effective nonradiative paths to the ground state and the height of
the relevant energy barrier may depend on the solvent.63
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. Excited state energies in solvents obtained from a state specific model in TDDFT
1
(BMK/6-31G). Excited states from S1 up to the n,π* (CO) excited state are shown. Black bars
indicate 1π,π* states and blue bars indicate 1n,π* states.
Photostability. Although it is often advertized that naphthalimide dyes are stable, concrete
data for stability are rarely reported. Here we compare the photostability of our newly
ACS Paragon Plus Environment
27

The Journal of Physical Chemistry
Page 28 of 40
1
2
3
4
synthesized dyes with those of known commercial dyes. A broad band of visible light in a range
5
6
of 400–700 nm was continuously irradiated from a 100 W Xe-lamp to 1 µM solutions, during
7
8
9
which time the absorption and fluorescence spectra were monitored. The irradiation was carried
out in air. The results are presented in Figure 6. As a representative example for nonpolar
solvents, toluene was chosen. Coumarin 343, which has an absorption band over 370–470 nm,
was used as a reference for the experiments in toluene. One hour after under our irradiation
conditions, when the absorbance as well as the fluorescence of the coumarin 343 dye diminished
to about 20% of the original values, no sign of degradation was observed for the present three
dyes (Figures 5a and 5b). As a representative example for polar solvents, MeCN was chosen and
fluorescein isothiocyanate, which is one of the most commonly used fluorescent labels (λmax
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
~500 nm), was used as a reference. One hour after under our irradiation conditions, when the
absorbance as well as the fluorescence of fluorescein isothiocyanate diminished to ~15% of the
original values, no sign of degradation was observed for Ph-TNI and Ph2NPh-TNI. Note that
only the absorption was monitored for Ph2NPh-TNI, because this dye is almost nonfluorescent in
MeCN as described earlier. On the other hand, NO2Ph-TNI showed changes in absorption and
fluorescence spectra during the first 20 minutes of irradiation. The absorption changes were
accompanied by an isosbestic point. After 20-minute irradiation, both the new absorption and
fluorescence spectra remained constant. Absorption maximum was slightly red-shifted by 3 nm
from that of the original NO2Ph-TNI, while the fluorescence maximum was red-shifted by 30 nm.
Similar changes were also observed when a solution purged with Ar was irradiated. These
observations suggest that NO2Ph-TNI has changed into a new, stable fluorescent species upon
irradiation, although we have not identified the product at this moment.
ACS Paragon Plus Environment
28

Page 29 of 40
The Journal of Physical Chemistry
1
2
3
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
(a)
(b)
(c)
(d)
Figure 6. Stability of dyes under light irradiation (100 W Xe lamp, 400–700 nm). Light was
irradiated to a solution of 1 µM dye in toluene (a and b) and in MeCN (c and d), during which
time the relative absorbance at λab (a and c) and the fluorescence intensity at λfl (b and d) were
recorded.
CONCLUSIONS
We have prepared naphthalimide fluorescent dyes whose π system is extended via a thienylene
spacer by additional aromatic unit with or without an electron-withdrawing or electron-donating
substituent. By extending the π system, both absorption and fluorescence exhibit significant red-
1
shifts. The lowest energy transitions are π,π* transitions with a significant charge transfer
character, according to TDDFT calculations. The solvent dependent photophysics were
ACS Paragon Plus Environment
29