Internal conversion in 4-substituted 1-naphthylamines. Influence of the electron donor/acceptor substituent character

Article abstract of DOI:10.1039/a908924a

The thermally activated internal conversion (IC) taking place in 4- substituted 1-(dimethylamino)naphthalenes (14DMX) and 1-aminonaphthalenes (14ANX) with X = CN, Cl, H, CH₃ and OCH₃ was investigated in three solvents spanning the polarity scale, hexane, diethyl ether and acetonitrile. In both series 14DMX and 14ANX, the efficiency of the IC reaction decreases substantially when X changes from CN to OCH₃, the order in which the electron donor character of the 4-substituent increases. Considerably larger IC reaction rate constants are obtained for the first group of compounds. This difference is connected with the ground state structure of the amino group, which is more strongly twisted for 14DMX (ca. 60°) than for 14ANX (ca. 20°), whereas both sets of 1-naphthylamines are planarised in the S₁ excited state. The IC process slows down with increasing solvent polarity for each of the 14DMX and 14ANX molecules. The substituent X and the solvent polarity mainly affect the IC activation energy E(IC). With 14DMX in hexane, E(IC) increases from 10 kJ mol^-1 for X = CN to 34 kJ mol^-1 for X = OCH₃, whereas with, e.g., 14DMCL a solvent polarity dependent increase of E(IC) from 16 kJ mol^-1 in hexane to 28 kJ mol^-1 in acetonitrile is observed. The height of the barrier E(IC) is governed by the energy gap ΔE(S₁,S₂) between the two lowest excited singlet states. The influence of δE(S₁,S₂) on E(IC) is attributed to vibronic coupling caused by the proximity of the S₁ and S₂ states, which flattens the S₁ potential energy surface and thereby lowers the IC barrier when ΔE(S₁,S₂) becomes smaller. It is assumed that the IC reaction of the 1-naphthylamines passes through a conical intersection, which exists as a consequence of the relative displacement of the S₁ and S₀ surfaces caused by the different amino twist angles in the two states.

Full text of DOI:10.1039/a908924a

View Article Online / Journal Homepage / Table of Contents for this issue
Internal conversion in 4-substituted 1-naphthylamines. InÑuence of the
electron donor/acceptor substituent character
Kengo Suzuki,¤a Attila Demeter,”a Wolfgang Kuhnle,a Erich Tauer,a Klaas A. Zachariasse,*a
Seiji Tobitab and Haruo Shizukab
a Max-Planck-Institut fur biophysikalische Chemie, Spektroskopie und Photochemische Kinetik,
D-37070 Gottingen, Germany. E-mail: kzachar=gwdg.de; Fax: ]49 551 201 1501
b Department of Chemistry, Gunma University, Kiryu, Gunma 376-8515, Japan
Received 9th November 1999, Accepted 20th December 1999
The thermally activated internal conversion (IC) taking place in 4-substituted 1-(dimethylamino)naphthalenes
(14DMX) and 1-aminonaphthalenes (14ANX) with X \ CN, Cl, H, CH and OCH was investigated in three
3
3
solvents spanning the polarity scale, hexane, diethyl ether and acetonitrile. In both series 14DMX and 14ANX,
the efficiency of the IC reaction decreases substantially when X changes from CN to OCH , the order in which
3
the electron donor character of the 4-substituent increases. Considerably larger IC reaction rate constants are
obtained for the Ðrst group of compounds. This di†erence is connected with the ground state structure of the
amino group, which is more strongly twisted for 14DMX (ca. 60¡) than for 14ANX (ca. 20¡), whereas both sets
of 1-naphthylamines are planarised in the S excited state. The IC process slows down with increasing solvent
1
polarity for each of the 14DMX and 14ANX molecules. The substituent X and the solvent polarity mainly
a†ect the IC activation energy E . With 14DMX in hexane, E increases from 10 kJ mol~1 for X \ CN to 34
IC
IC
kJ mol~1 for X \ OCH , whereas with, e.g., 14DMCL a solvent polarity dependent increase of E from 16 kJ
3
IC
mol~1 in hexane to 28 kJ mol~1 in acetonitrile is observed. The height of the barrier E is governed by the
IC
energy gap *E(S ,S ) between the two lowest excited singlet states. The inÑuence of *E(S ,S ) on E is
1 2
1 2
IC
attributed to vibronic coupling caused by the proximity of the S and S states, which Ñattens the S potential
1
2
1
energy surface and thereby lowers the IC barrier when *E(S ,S ) becomes smaller. It is assumed that the IC
reaction of the 1-naphthylamines passes through a conical intersection, which exists as a consequence of the
1 2
relative displacement of the S and S surfaces caused by the di†erent amino twist angles in the two states.
1
0
*E(S ,S ) increases with increase in solvent polarity,5 which
Introduction
1 2
means that the S state of 1DMAN has a larger dipole
1
With 1-(dimethylamino)naphthalene (1DMAN) and other 1-
(dialkylamino)naphthalenes in alkane solvents such as hexane
at room temperature, efficient thermally activated internal
conversion (IC) takes place, which slows down with increasing
solvent polarity.1h5 The activation energy E of this reaction
increases when the solvent becomes more polar, for 1DMAN
moment than S . Note that with 4-(dimethylamino)benzoni-
2
trile (DMABN), in contrast, the opposite order holds for the
dipole moments: the S state has a smaller dipole moment
1
than S .6h8 As a consequence, *E(S ,S ) decreases with
2
1 2
increasing solvent polarity in the case of DMABN and intra-
molecular charge transfer (ICT) becomes faster in more polar
solvents.9,10
IC
from 18 kJ mol~1 in hexane to 31 kJ mol~1 in acetonitrile.3,5
The efficiency of the IC process in the 1-naphthylamines cor-
relates with the twist angle h of the amino group5 (deÐned
below) with respect to the plane of the naphthyl moiety. With
1-(N-azetidinyl)naphthalene, as an example, for which the
angle h is smaller (36¡) than that of 1DMAN (60¡), an IC rate
In this paper, a discussion is presented of the photophysics
of the two series of 4-substituted 1-naphthylamines 14DMX
and 14ANX in hexane, diethyl ether and acetonitrile, three
solvents spanning the polarity scale. In each of these solvents,
the IC efficiency strongly depends on the electron donor/
acceptor character of the substituent.
constant k of 4.0 ] 108 s~1 is observed in hexane at 25¡C as
IC
compared with 8.3 ] 109 s~1 for 1DMAN.5 As the 1-
naphthylamines such as 1DMAN become more planar in the
equilibrated S state than in the ground state, the amino sub-
stituent undertakes an anti-twist movement towards planarity
Experimental
1
1-Dimethylamino-4-cyanonaphthalene (14DMCN) and 1-
methylamino-4-cyanonaphthalene (14MCN) were synthesised
by methylation of 1-amino-4-cyanonaphthalene (14ANCN,
with the naphthalene moiety before the IC process sets in.3h5
In addition to the amino twist, the energy gap *E(S ,S )
1 2
between the two lowest excited singlet states plays a predomi-
nant role in the IC dynamics of 1DMAN, an increase in this
gap leading to a larger E .3h5 It has been shown that
IC
¤ On leave from the Department of Chemistry, Gunma University,
Kiryu, Gunma 376-8515, Japan.
” On leave from the Institute of Chemistry, Chemical Research
Center, Hungarian Academy of Sciences, P.O. Box 17, 1525 Budapest,
Hungary.
DOI: 10.1039/a908924a
Phys. Chem. Chem. Phys., 2000, 2, 981È991
This journal is ( The Owner Societies 2000
981

View Article Online
Aldrich) with CH I in methanol.11,12 1-Dimethylamino-4-
Results and discussion
3
chloronaphthalene (14DMCL) was obtained by reductive
methylation of 1-amino-4-chloronaphthalene (14ANCL,
Aldrich) with 37% aqueous formaldehyde and cyano-
borohydride in acetonitrile at room temperature.13 The reac-
tion product was washed with 1 M KOH and saturated
aqueous NaCl and extracted with diethyl ether. 1-
Fluorescence and absorption spectra
The Ñuorescence and absorption spectra of 1DMAN and four
4-substituted 1-(dimethylamino)naphthalenes 14DMX in
hexane at 25 ¡C are shown in Fig. 1. The spectra of the corre-
sponding 1-aminonaphthalenes 14ANX are also depicted. For
both series 14DMX and 14ANX in diethyl ether and acetoni-
trile, the shape of the absorption and Ñuorescence spectra is
similar to that in hexane and their maxima undergo a red shift
with increasing solvent polarity (Table 1). The Ñuorescence
spectra of the 1-naphthylamines (except 14DMCN) in the
three solvents consist of a single emission band, indicating
that dual Ñuorescence does not occur.5 With 14DMCN, but
not with 14ANCN, dual Ñuorescence consisting of a local
excited (LE) and an intramolecular charge transfer (ICT) emis-
sion band10,12,24 is observed in diethyl ether and acetonitrile
(Tables 1 and 2).25
Dimethylamino-4-methoxynaphthalene
(14DMMO)
was
made by methylation of 1-amino-4-naphthol hydrochloride
(14ANOH, Aldrich) with CH I and sodium methoxide in
3
methanol under a nitrogen atmosphere. 1-Amino-4-methyl-
naphthalene (14ANME) was synthesised by nitration of 1-
methylnaphthalene and subsequent reduction of 1-methyl-4-
nitronaphthalene with hydrazine hydrate and PdÈC in
ethanol. 1-Dimethylamino-4-methylnaphthalene (14DMME)
was made by methylation of 14ANME with CH I in
3
methanol. 1DMAN was obtained from Fluka. 1-
Aminonaphthalene (1AN) was purchased from Merck. 1-
(Methylamino)naphthalene (1MAN) was made as described
previously.5 With all 1-naphthylamines, HPLC was the last
puriÐcation step.
The energy E(S ) of the equilibrated S state relative to the
1
1
relaxed S ground state is determined from the intersection of
0
the absorption and Ñuorescence bands.5 The E(S ) data for
14DMX and 14ANX in hexane and acetonitrile are listed in
The solvent hexane (Merck) was used as received. Diethyl
ether (Merck, Uvasol) and acetonitrile (Baker) were chromato-
graphed over Al O just prior to use. The solutions, with an
1
Table 1. In the series 14DMX, the lowest energy E(S ) is
1
2
3
found for 14DMCN, increasing in the order X \ CN, OCH ,
optical density between 0.6 and 1.0 for the maximum of the
Ðrst band in the absorption spectrum, were deaerated by bub-
bling with nitrogen for 15 min.
3
Cl, CH , H, i.e., with decreasing mesomeric character26 of the
3
substituent.
The Ñuorescence spectra were measured with a quantum-
corrected Shimadzu RF-5000PC spectroÑuorimeter. Fluores-
cence quantum yields, with an estimated reproducibility of
2%, were determined with quinine sulfate in 0.5 M H SO as
Fluorescence quantum yields and decay times of 14DMX and
14ANX. InÑuence of substituent and solvent polarity
Hexane at 25 ÄC. As compared with the Ñuorescence
2
4
a standard (U \ 0.546 at 25 ¡C).14 The Ñuorescence decay
quantum yield U of 0.012 for the parent compound 1DMAN
f
times were determined with picosecond laser or nanosecond
f
in hexane at 25 ¡C, substantially smaller yields are observed
Ñashlamp single-photon counting (SPC) set-ups. These
systems were described previously.5,15,16 The picosecond SPC
equipment consists, in brief, of a mode-locked argon ion laser
(Coherent, Innova 100-10) and a synchronised dye-laser
system (Coherent 702-1CD, DCM). The outcoupled light is
under these conditions for 14DMCN (0.002) and 14DMCL
(0.006), whereas larger U values are obtained for 14DMME
f
frequency doubled (298 nm) in an LiIO crystal. The Ñuores-
3
cence and scatter (Ludox) records are measured with a Hama-
matsu R3809 MCP photomultiplier ([3000 V). Alternatively,
a
picosecond laser system consisting of a mode-locked
titaniumÈsapphire laser (Coherent, MIRA 900-F) pumped by
an argon ion laser (Coherent, Innova 415) was used. By fre-
quency doubling and sum-frequency generation, a wavelength
of 300 nm is reached. The instrument response function has a
half-width of 25È35 ps. The Ñuorescence is detected at the
magic angle (54.7¡). The analysis of the Ñuorescence decays
was carried out as reported earlier.17,18 The reproducibility of
the decay times is better than 2%.
The yield U of the intersystem crossing from the equili-
ISC
brated S state to the lowest triplet state T was measured by
1
1
T ÈT absorption, using a method based on triplet energy
transfer with perylene as the acceptor.19,20 As a reference sub-
stance, either N-methyl-1,8-naphthalimide in hexane and
diethyl ether or benzophenone in acetonitrile was used. For
the naphthalimide, a value of 0.96 for U was employed,21
ISC
whereas for benzophenone U
was taken to be equal to
ISC
unity.22,23 The solutions used in these experiments were
degassed employing the freezeÈpumpÈthaw method. A correc-
tion was applied for the contribution to the T ÈT absorption
coming from directly excited (at 308 nm) perylene molecules,
by comparison with a sample containing only the acceptor.
To take into account the singlet energy transfer from the
naphthylamines to perylene, a second correction was neces-
sary for molecules with a long Ñuorescence lifetime, such as
1AN. The correction was based on the fraction of photons
absorbed by each species as calculated from their respective
absorbance values. For this purpose, the Ñuorescence decay
time of samples with and without perylene was measured.
Fig. 1 Absorption and normalised Ñuorescence spectra of the 4-
substituted 1-naphthylamines 14DMX [R \ N(CH ) ], full lines, and
3 2
14ANX (R \ NH ), dotted lines, with X \ CN, Cl, H, CH and
2
3
OCH , in hexane at 25 ¡C.
3
982
Phys. Chem. Chem. Phys., 2000, 2, 981È991

View Article Online
Table 1 Absorption maximum lmax(abs), extinction coefficient emax at lmax(abs), Ñuorescence maximum lmax(Ñu), Stokes shift *l(St) and energy
E(S ) of the relaxed S state of the 4-substituted 1-naphthylamines 14DMX and 14ANX in hexane and acetonitrile at 25 ¡C
1
1
lmax(abs)/
103 cm~1
emax/
lmax(Ñu)/
*l(St)/
E(S )/
1
M~1 cm~1
103 cm~1
103 cm~1
103 cm~1
Hexane ( f [ f @ \ 0.003)aÈ
14DMCN
14DMCL
1DMAN
14DMME
14DMMO
14MCNb
1MANc
14ANCN
14ANCL
1AN
29.80
31.35
32.64
31.95
31.25
29.40
30.04
30.04
30.58
31.44
30.78
10300
6700
5100
5800
5700
13 100
7000
11800
8300
5500
6200
25.60
25.30
25.84
25.30
24.26
25.60
26.06
26.00
25.70
26.52
25.64
4.20
6.05
6.80
6.65
6.99
3.80
3.98
4.04
4.88
4.92
5.14
27.60
27.83
28.82
28.02
27.60
27.50
28.19
28.14
28.06
28.96
28.00
14ANME
Acetonitrile ( f [ f @ \ 0.305)aÈ
14DMCN
14DMCL
28.90
30.50
32.00
31.55
30.55
29.41
29.50
30.74
29.86
11300
6300
5300
6000
5800
13300
7800
6200
6600
23.96d
22.70
23.60
22.83
21.83
24.28
22.99
23.88
22.88
4.94
7.80
8.40
8.72
8.72
5.13
6.51
6.86
6.98
26.11
26.44
27.29
26.72
26.10
26.67
26.39
27.28
26.40
1DMAN
14DMME
14DMMO
14ANCN
14ANCL
1AN
14ANME
a Solvent polarity scale, see eqn. (4). b 14MCN: 1-methylamino-4-cyanonaphthalene. c 1MAN: 1-(methylamino)naphthalane. d The energy of
maximum of the locally excited (LE) Ñuorescence band. The intramolecular charge transfer (ICT) band of 14DMCN in acetonitrile has a
maximum at 18 500 cm~1.
(0.035) and 14DMMO (0.22) (see Table 2). A similar depen-
dence on the nature of the 4-substituent is found for the 1,4-
hexane at 25 ¡C is much shorter than that of 1DMAN (118
ps), whereas for 14DMME (414 ps) and especially for
14DMMO (3.20 ns) considerably longer decay times are
observed. With the 1-aminonaphthalenes 14ANX, the inÑu-
ence of the 4-substitution on the decay times shows the same
pattern as that observed with the 14DMX compounds. For
example, the decay time q is clearly shorter for 14ANCN (0.58
ns) than for 1AN (6.65 ns) (see Table 3 and Fig. 3). It should
be noted that the order CN, Cl, H, CH , OCH of the substit-
substituted NH derivatives 14ANX in comparison with the
2
U
of 0.47 for 1-aminonaphthalene (1AN): 0.08 (14ANCN)
f
and 0.30 (14ANCL) against 0.63 (14ANME) (see Table 3), with
clearly larger yields than for the 14DMX molecules. In the
case of the pair 1DMANÈ1AN, the di†erence in the quantum
yield U has been attributed to the larger ground state amino
f
twist angle of the former compound.3h5 The 4-substituent
3
3
likewise has a pronounced inÑuence on the Ñuorescence decay
uents X in which U and q become larger for both 14DMX
f
time q of the 1-(dimethylamino)naphthalenes 14DMX (see Fig.
and 14ANX corresponds to the change from electron acceptor
2 and Table 2), in a manner similar to that observed for U .
to donor character of these substituents,27,28 a change which
f
The decay time q of 14DMCN (5 ps) and 14DMCL (55 ps) in
is also represented by their Hammett constants p `.27,28
p
Table 2 Kinetic and thermodynamic data for the 4-substituted 1-(dimethylamino)naphthalenes 14DMX in hexane, diethyl ether and
acetonitrile at 25 ¡C
q/
k /
k
d/
k /
k0 /
E /
f
ISC
IC
IC
IC
ÈXa
U
ns
107 s~1
U
b
U
c
107 s~1
107 s~1
1013 s~1
kJ mol~1
f
ISC
IC
Hexane ( f [ f @ \ 0.003)eÈ
14DMCN
14DMCL
1DMAN
ÈCN
ÈCl
0.002
0.006
0.012
0.035
0.22
0.005
40
11
10
8.5
6.9
0.008
0.03
0.02
0.04
0.50
0.99
0.96
0.97
0.93
0.28
160
55
17
10
16
17200 ^ 200
1700 ^ 20
830 ^ 30
220 ^ 10
12 ^ 1
0.9
0.9
1.2f
1.8
10.0 ^ 0.5
15.5 ^ 0.3
18.2 ^ 0.8f
22.3 ^ 0.4
34.1 ^ 1.0
0.055
0.120f
0.414
3.20
ÈH
14DMME
14DMMO
ÈCH
3
ÈOCH
11
3
Diethyl ether ( f [ f @ \ 0.165)eÈ
14DMCN
14DMCL
1DMAN
14DMME
14DMMO
ÈCN
ÈCl
(0.008)g
0.014
0.032f
0.079
0.25
(0.20)h
0.03
0.025
0.04
(0.79)
0.96
0.94
0.88
0.18
0.171
0.440f
1.30
8.2
7.3
6.1
5.3
18
6
3
530 ^ 40
220 ^ 10
65 ^ 2
4i
7.6
13f
15
23.7 ^ 0.5
27.3 ^ 1.5e
30.6 ^ 0.8
ÈH
ÈCH
ÈOCH
3
4.70
0.57
12
3
Acetonitrile ( f [ f @ \ 0.305)eÈ
14DMCN
14DMCL
1DMAN
14DMME
14DMMO
ÈCN
ÈCl
(0.007)g
0.10
0.21e
0.38
(0.71)h
0.16
0.29
0.09
0.44
(0.28)
0.74
0.50
0.53
0.08
1.92
4.45f
9.19
5.2
4.7
4.1
3.6
8
7
1
3
36 ^ 2
11 ^ 2
5 ^ 1
1i
2.8
3.4f
5.8
27.9 ^ 0.6
31.4 ^ 2.0e
34.5 ^ 1.1
ÈH
ÈCH
ÈOCH
3
0.48
13.2
3
a The substituent at the 4-position of the naphthalene ring. b Determined by T ÈT absorption. c U \ 1 [ U [ U [eqn. (1)]. d k \ U /q.
IC ISC ISC ISC
e Solvent polarity parameter, see eqn. (4). f See ref. 5. g U \ ULE ] UICT, the sum of the LE and CT Ñuorescence quantum yields; see text. h The
f
f
f
f
triplets mainly originate from the ICT state; see text. i k \ U /q. The other k data were calculated from k0 and E
IC IC IC IC
.
IC
Phys. Chem. Chem. Phys., 2000, 2, 981È991
983

View Article Online
Table 3 Kinetic and thermodynamic data for the 4-substituted 1-aminonaphthalenes 14ANX in hexane and acetonitrile at 25 ¡C
q/
k /
k
d/
k /
k0 /
E /
f
ISC
IC
IC
IC
ÈXa
ns
U
U
b
U
c
107 s~1
107 s~1
107 s~1
1013 s~1
kJ mol~1
f
ISC
IC
Hexane ( f [ f @ \ 0.003)eÈ
14ANCN
14ANCL
1AN
ÈCN
ÈCl
0.580
0.080
0.30
0.47
0.63
0.015
0.58
0.40
0.25
0.91
0.12
0.13
0.12
14
3
16
6
160 ^ 10
1.8 ^ 0.5
1.4 ^ 0.6
0.08 ^ 0.04
2.7
3.0
6.4
6.5
24.1 ^ 0.5
35.5 ^ 1.0
38 ^ 8
3.64
6.65
9.87
8.2
7.1
6.4
ÈH
ÈCH
14ANME
3
45 ^ 3
3
Diethyl ether ( f [ f @ \ 0.165)eÈ
14ANCN
14ANCL
1AN
ÈCN
ÈCl
5.84
8.79
12.4
15.2
0.65
0.55
0.66
0.76
11
0.33
0.20
0.11
0.12
0.14
0.13
6.3
5.3
5.0
4
2
1
1.4f
1.1f
0.9f
ÈH
14ANME
ÈCH
3
Acetonitrile ( f [ f @ \ 0.305)eÈ
14ANCN
14ANCL
1AN
ÈCN
ÈCl
6.22
0.72
0.73
0.86
0.85
0.008
0.22
0.10
0.04
0.27
0.05
0.04
0.11
12
0.1
2
0.5
0.2
3.4 ^ 0.5
0.4f
0.2f
3.1
34.0 ^ 0.6
14.1
18.3
20.6
5.2
4.7
4.1
ÈH
14ANME
ÈCH
0.5f
3
a The substituent at the 4-position of the naphthalene ring. b Determined by T ÈT absorption. c U \ 1 [ U [ U . d k \ U /q. e Solvent
IC ISC ISC ISC
f
polarity parameter, see eqn. (4). f k \ U /q. The other k data were calculated from k0 and E
IC IC IC IC
.
IC
Polar solvents diethyl ether and acetonitrile at 25 ÄC. In each
of the two polar solvents diethyl ether and acetonitrile, the
presence of the 4-substituent in 14DMX and 14ANX exerts a
similar inÑuence on U and q to that observed for the less
f
polar solvent hexane (see Tables 2 and 3). For 14DMCL, U
f
and q are smaller than those of 1DMAN in diethyl ether
(0.014 and 171 ps as compared with 0.032 and 440 ps) and
also in acetonitrile (0.10 and 1.92 ns as compared with 0.21
and 4.45 ns), whereas larger quantum yields and decay times
are obtained for 14DMME and 14DMMO in both solvents
(see Table 2).
As has been reported previously, the Ñuorescence quantum
yield U and the decay time q of 1DMAN increase strongly
f
when the solvent polarity becomes larger.1h5 The quantum
yield increases from 0.012 in hexane to 0.032 in diethyl ether
and 0.21 in acetonitrile, the decay times following this trend
with 120 ps in hexane, 440 ps in diethyl ether and 4.45 ns in
acetonitrile.5 The same inÑuence of solvent polarity is
observed here for the 14DMX and 14ANX molecules. With
14DMME, as an example (Table 2), U and q increase from
f
0.035 and 0.41 ns in hexane to 0.38 and 9.2 ns in acetonitrile.
For the 1-aminonaphthalenes 14ANX
a similar solvent
polarity dependence of U and q is found (Table 3). With
f
14ANCN the following values are obtained for the Ñuores-
cence quantum yield and the decay time: 0.08 and 0.58 ns
(hexane), 0.65 and 5.84 ns (diethyl ether) and 0.72 and 6.22 ns
(acetonitrile).
Radiationless process. From the data on U and q presented
f
in this section, it is concluded that an important radiationless
process takes place with the 4-substituted 1-(dimethylamino)
naphthalenes in the three solvents studied, slowing down with
increasing solvent polarity. For the NH derivatives 14ANX,
2
the radiationless deactivation is less efficient than in the case
of the corresponding 14DMX molecules. To determine the
nature of this process, the yield of intersystem crossing (ISC)
must be known.
Intersystem crossing and internal coversion
The intersystem crossing yield U
of the naphthylamines
ISC
14DMX and 14ANX was measured by employing the method
of T ÈT absorption as described in the Experimental section
(see Tables 2 and 3). These yields are relatively small for the
14DMX molecules at 25 ¡C in hexane (between 0.01 and 0.04)
and diethyl ether (between 0.02 and 0.04), except for
14DMMO, which has considerably larger yields of U \ 0.50
ISC
and 0.57 in the two solvents. In acetonitrile, relatively large
Fig. 2 Fluorescence decay curves with decay times
q of (a)
U
values are found for all 14DMX molecules, between 0.71
(14DMCN) and 0.09 (14DMME). With the 14ANX molecules
in hexane, U is larger than that of their 14DMX counter-
14DMCN, (b) 14DMCL, (c) 1DMAN, (d) 14DMME and (e)
14DMMO in hexane at 25 ¡C. The time-scales are 1.025 ps per
channel for (a)È(d) and 40.646 ps per channel for (e).
ISC
ISC
984
Phys. Chem. Chem. Phys., 2000, 2, 981È991

View Article Online
Fig. 3 Fluorescence decay curves with decay times q of (a) 14ANCN, (b) 1AN and (c) 14ANME in hexane at 25 ¡C. The long decay component
in parentheses in (a) is attributed to an impurity. The time scales are 2.050 ps per channel (picosecond laser system) for (a) and (b) and 83.11 ps per
channel (nanosecond system) for (c).
parts, whereas the opposite is observed in acetonitrile.4 From
these U data and U in Tables 2 and 3, the internal conver-
Fluorescence decay times and radiative rates as a function of
temperature
ISC
f
sion yield U can now be calculated by using the equation
IC
The Ñuorescence decays of 14DMX and 14ANX in hexane,
diethyl ether and acetonitrile are single exponential at all tem-
peratures investigated, indicating that dual Ñuorescence does
not take place.3h5 The decay times q of the 14DMX com-
pounds in hexane and in acetonitrile (without 14DMCN) are
presented in Fig. 5 as a function of temperature. With all
14DMX except 14DMMO, a strong temperature dependence
of q is observed, in accordance with the e†ect of temperature
U
\ 1 [ U [ U
(1)
IC
f
ISC
It is thereby assumed that no processes other than Ñuores-
cence, intersystem crossing and internal conversion take part
in the deactivation of the S state of the naphthylamines, as
1
evidence for the occurrence of photochemical reactions was
not found.5
From the data for U so obtained (Table 2), it follows that
IC
on U described in the previous section. For the 14ANX mol-
for 14DMCN, 14DMCL, 1DMAN and 14DMME (U
f
IC
ecules,5 a pronounced inÑuence of temperature on q is only
between 0.93 and 0.99) in the non-polar solvent hexane at
found with 14ANCN in hexane.
25 ¡C, internal conversion is the major process causing the
The temperature dependence of the ratio k /n2 of the radi-
strong decrease in the Ñuorescence quantum yield U and also
f
D
f
ative rate constant k and the square of the solvent refractive
in the decay time q of these molecules as described in the pre-
f
index n is plotted in Fig. 6 for 14DMCL, 1DMAN,
14DMME and 14DMMO in hexane and acetonitrile. The
vious section. In the case of 14DMMO, however, with a U of
D
IC
0.28, the much less efficient IC process does not prevent ISC
plots are presented in this way, as k depends on n (k D n2)30
from becoming the most important reaction pathway: U
\
f
D
f
D
ISC
0.50. In the more polar solvents diethyl ether and acetonitrile
at 25 ¡C (Table 2), the IC yields of 14DMX are smaller than
in hexane (see Table 2), the di†erence increasing when
the solvent polarity becomes larger. With the 1-
aminonaphthalenes 14ANX in hexane, the largest value for
U
(0.91) is observed for 14ANCN, decreasing for X in the
IC
order CN, Cl, H, CH , down to an IC yield of 0.12 for
14ANME (Table 3). Also with these compounds, the IC reac-
3
tion becomes less efficient with increasing solvent polarity (see
the data for acetonitrile in Table 3).
Temperature dependence of Ñuorescence quantum yields
The Ñuorescence quantum yield U of the 1-(dimethylamino)-
f
naphthalenes 14DMX was determined as a function of tem-
perature in hexane and acetonitrile (Fig. 4). In the case of
14DMCN, 14DMCL, 1DMAN3h5 and 14DMME in hexane
[Fig. 4(a)], U increases sharply with decreasing temperature.
f
This shows that the Ñuorescence quenching in these molecules
is strongly thermally activated.3h5 With 14DMMO,
U
f
reaches a plateau at ca. [20 ¡C. Such a temperature indepen-
dence is not observed for the other more strongly quenched
molecules over the temperature range determined by the
melting-point of hexane ([95 ¡C).29 In acetonitrile [Fig. 4(b)],
the Ñuorescence quantum yield of the 14DMX molecules like-
wise increases when the temperature is lowered. The compari-
son of the U data for hexane and acetonitrile in Fig. 4
f
indicates that for the 14DMX compounds Ñuorescence quen-
Fig. 4 Fluorescence quantum yields
U
of the 4-substituted 1-
f
ching by internal conversion still takes place in acetonitrile,
but has become less efficient in this polar solvent.
(dimethylamino)naphthalenes 14DMX with X \ CN, Cl, H, CH and
3
OCH in (a) hexane and (b) acetonitrile as a function of temperature.
3
Phys. Chem. Chem. Phys., 2000, 2, 981È991
985

View Article Online
function of temperature (see Fig. 5), by Ðtting the times q with
the equation
1
\ k ] k ] k
f
IC
ISC
q
A
[E B A[E
B
IC
ISC
\ k ] k0 exp
IC
] k0 exp
] k (0) (2)
f
ISC
ISC
RT
RT
The rate constants k and k in eqn. (2) are in Arrhenius
form and k (0) is the temperature-independent ISC rate con-
IC
ISC
ISC
stant.4,5 In the Ðtting procedure, a constant value of 5 kJ
mol~1 for the ISC activation energy E is adopted5,32 and
ISC
the minor temperature dependence of k (see Fig. 6) is not
f
taken into account. The Ðts for 14DMX in hexane and aceto-
nitrile are shown in Fig. 5 and the data for k0 and E so
IC
IC
obtained are given in Table 2. Similar Ðts for the decay times
of 14ANCN, 14ANCL, 1AN5 and 14ANME in hexane and of
14ANCN in acetonitrile resulted in the k0 and E data
IC
IC
reported in Table 3.
InÑuence of substituent X. With 14DMX in hexane, E
IC
increases from 10 kJ mol~1 for 14DMCN to 34 kJ mol~1 for
14DMMO (Table 2). The pre-exponential factors k0 show a
Fig. 5 Fluorescence decay times
q
of the 4-substituted 1-
(dimethylamino)naphthalenes 14DMX with X \ CN, Cl, H, CH and
IC
corresponding increase. In diethyl ether and acetonitrile a
3
OCH in (a) hexane and (b) acetonitrile as a function of temperature.
3
comparable dependence on the nature of the 4-substituent is
found (see Table 2). For the 1-aminonaphthalenes 14ANX in
The lines through the data points represent the Ðtting of the data with
eqn. (2), giving the activation energies E and the pre-exponential
hexane (Table 3), larger activation energies E are obtained
i
factors k0 (Table 2) of the internal conversion (IC) and intersystem
IC
i
than for their 14DMX counterparts, from 24 kJ mol~1 for
crossing (ISC) processes.
14ANCN against 10 kJ mol~1 for 14DMCN to 45 kJ mol~1
against 22 kJ mol~1 for the 4-methyl-substituted molecules
14ANME and 14DMME.
and n becomes smaller at higher temperatures.29,31 In both
D
solvents, k /n2 increases only slightly with temperature, in a
f
D
manner similar to that previously reported for 1DMAN.5
InÑuence of solvent polarity. For each of the 14DMX and
14ANX compounds studied, E becomes larger with increas-
IC
Activation energies of internal conversion from Ñuorescence
decay times
ing solvent polarity, similarly to what has been observed pre-
viously for 1DMAN and 1AN.5 In the case of 14DMCL, for
example, E increases from 16 kJ mol~1 in hexane, via 24 kJ
The activation energy E of the IC reaction of 14DMX and
IC
mol~1 in diethyl ether to 28 kJ mol~1 in acetonitrile. The IC
IC
14ANX in hexane, diethyl ether and acetonitrile was deter-
activation energy of 14DMMO in diethyl ether and acetonit-
rile could not be determined with sufficient accuracy, as its
decay time q decreases only slightly with increasing tem-
perature over the experimentally available temperature range
set by the boiling-point of the solvents hexane (86.2 ¡C) and
mined from their Ñuorescence decay times q measured as a
acetonitrile (80.1 ¡C).29 For 14ANCN in the series 14ANX, E
IC
likewise increases with solvent polarity, from 24 kJ mol~1 in
hexane to 34 kJ mol~1 in acetonitrile (Table 3).
InÑuence of substituent X and solvent polarity on internal
conversion
From the values of the activation energy E in Tables 2 and
IC
3, it follows that the inÑuence of the substituent X and of the
solvent polarity on the efficiency of the IC reaction of 14DMX
and 14ANX is mainly caused by the di†erent activation ener-
gies E . Owing to the dependence of E on X, the IC rate
IC
IC
constant k for 14DMX and for 14ANX in hexane, diethyl
IC
ether and acetonitrile at 25 ¡C (Tables 2 and 3) increases when
X changes in the order OCH , CH , H, Cl, CN. This is the
3
3
order in which the Hammett constant p ` and hence the elec-
P
tron acceptor character of the substituent
X becomes
larger27,28 (see Fig. 7). The decrease in k on going from
IC
hexane to acetonitrile, shown in Fig. 7(a), is also caused by
larger E barriers in the more polar solvents.
IC
DE(S ,S ) energy gap model
1
2
Fig. 6 Plots of the ratio k /n2 for the 4-substituted 1-
In the case of 1DMAN and other 1-naphthylamines, it has
f
D
(dimethylamino)naphthalenes 14DMX with X \ CN, Cl, H, CH and
3
been postulated that the magnitude of E is directly related to
OCH in (a) hexane and (b) acetonitrile as a function of temperature.
IC
that of the energy gap *E(S ,S ) between the two lowest
1 2
3
The radiative rate constant k is divided by the square of the refractive
f
index n (see text and ref. 31).
excited singlet states S and S .5 In the context of this model,
D
1
2
986
Phys. Chem. Chem. Phys., 2000, 2, 981È991

View Article Online
14ANX, with 2200 cm~1 for 1AN and 1750 cm~1 for
14ANCN.
From the absorption and MCD spectra, information is
obtained on the FranckÈCondon (FC) states S (FC) and
1
S (FC) of the 1-naphthylamines, which states retain the same
2
molecular structure (e.g., the amino twist angle) as that
present in the S ground state. It is assumed that also for the
0
more planar5 relaxed S and S states in the series 14DMX
1
2
and 1ANX (see below), the gap *E(S ,S ) decreases in a
1 2
manner similar to that deduced from the absorption spectra
and the MCD data. This leads to the conclusion that a corre-
lation exists between the magnitude of the IC activation
energy E and the energy gap *E(S ,S ), as reported pre-
IC
1 2
viously for 1DMAN.5
Di†erent amino twist angles for 14DMX and 14ANX in S
0
As described in the preceding sections, considerably larger IC
rate constants k are observed with the 14DMX molecules
IC
than with their 1ANX derivatives in each of the three solvents
investigated here (Tables 2 and 3). Before an explanation of
this photophysical behaviour can be given, information on the
molecular structure of the two groups 14DMX and 14ANX is
presented. For the pair 1ANÈ1DMAN the twist angle h in the
ground state S (Fig. 8) increases upon methylation of the
0
amino group, from 22¡ (1AN) to 60¡ (1DMAN). These h
Fig. 7 Plots of log k at 25 ¡C of the 4-substituted 1-naphthylamines
values resulted from ab initio calculations.5 Experimental evi-
dence for the magnitude of the twist was obtained from
absorption spectra, 1H NMR chemical shifts and data on
dipole moments. The magnitude of the amino twist angle h in
the 1-naphthylamines determines the electronic coupling
between the amino and naphthyl groups, when the pyramidal
angle t (Fig. 8) of the amino nitrogen remains unchanged, as
discussed before with 1DMAN.5 Information on the twist
angle h in the present series 14DMX and 14ANX will now be
presented.
IC
14DMX (a) and 14ANX (b) with X \ CN, Cl, H, CH and OCH vs.
3
3
the Hammett constant r ` of the substituent X, in the solvents
hexane (a, b), diethyl ether (a) and acetonitrile (a).
p
invoking vibronic coupling between S and S (see below), the
1
2
observed increase in E with increasing solvent polarity is
IC
explained by assuming that *E(S ,S ) increases when the
1 2
solvent becomes more polar. This means that S must have a
1
larger dipole moment than S ,3,5 which di†erence may be
2
caused by mutually perpendicular polarisation directions in
Absorption spectra and extinction coefficients
the two states, along the short axis of the 1-naphthylamines
for S and in the long axis for S . The long-axis polarisation
For 1DMAN5 and also for 14DMME and 14DMCL, the
1
2
in the S state of the 1,4-substituted naphthylamines, perpen-
absorption spectrum in hexane is blue shifted relative to that
2
dicular to the 1,4-direction, minimises the inÑuence of the elec-
of their NH derivatives 1AN, 14ANME and 14ANCL (Fig.
2
tron donor and acceptor substituents on the S dipole
1). In the case of the pair 1DMANÈ1AN, this blue shift has
2
moments of 14DMX and 14ANX. A similar e†ect is found
been attributed to the larger amino twist angle (see above) of
the former compound.5 A similar conclusion regarding the
angle h is drawn here for 14DMME and 14DMCL relative to
14ANME and 14ANCL. In addition, the extinction coefficient
emax of the lowest-energy absorption maximum of the 14DMX
molecules in hexane is lower than that of the corresponding
14ANX derivatives (see Table 1 and Fig. 1). This decrease is
likewise attributed to the larger amino twist angle and the
consequently decreased aminoÈnaphthyl electronic coupling
for 14DMX.5
with DMABN, for which the short-axis polarised S state has
1
a smaller dipole moment than the S state with a polarisation
2
in the long-axis direction of the donor and acceptor substit-
uents.6,33 In the following section, information on the
*E(S ,S ) energy gaps of 14DMX and 14ANX is presented.
1 2
Energy gap DE(S ,S ) from absorption spectra and MCD
1
2
measurements
The low-energy part of the absorption spectrum of the
14DMX and 14ANX molecules in hexane (Fig. 1) consists of
two overlapping bands, as discussed previously for 1DMAN
and 1AN.5 In the spectrum of 14DMMO, the separation
between the S and S bands is still visible. From the spectra
The absorption spectrum in hexane of the cyano derivative
14DMCN, in contrast to the other 14DMX compounds, is red
1
2
of the other 14DMX molecules it is seen that the overlap of
these two bands increases from 14DMME to 14DMCN. This
already is an indication that in the series 14DMX the energy
gap *E(S ,S ) decreases when X changes in the order OCH ,
1 2
3
IC
CH , H, Cl, CN, the same order in which a decrease in E
3
takes place (Tables 2 and 3).
More quantitative information on the magnitude of
*E(S ,S ) is obtained from magnetic circular dichroism
1 2
Fig. 8 (a) Top view of an 1-aminonaphthalene used to deÐne the
amino twist angle h between the amino group and the plane of the
naphthyl moiety. (b) Partial side view of a 1-aminonaphthalene used
to deÐne the pyramidal angle t. N is the amino nitrogen, X(1) and
X(2) are the amino substituents and C(1) is the carbon atom in the
1-position of the naphthyl group.
(MCD) measurements.34 It is found from these experiments
that the energy di†erence between the maxima of the two
absorpton bands decreases from 2100 cm~1 for 14DMMO to
1500 cm~1 for 14DMCN, supporting the interpretation based
on the absorption spectra. Similar *E(S ,S ) values result for
1 2
Phys. Chem. Chem. Phys., 2000, 2, 981È991
987

View Article Online
shifted with respect to its amino counterpart 14ANCN (Fig.
1). This does not mean, however, that the dimethylamino
group of 14DMCN is twisted to a smaller extent than that of
the other 14DMX molecules. The absorption band of the
monomethylated derivative 14MCN is red shifted relative to
that of 14ANCN (Fig. 9), in a manner similar to that found for
1-(methylamino)naphthalene (1MAN) as compared with 1AN,
which is to be expected for molecules having similar twist
angles h.5 For 14DMCN, however, the absorption spectrum is
in fact blue shifted with respect to that of 14MCN, in the same
way as observed for 1DMAN as compared with 1MAN.5 This
shows that the amino twist angle of 14DMCN is indeed larger
than those of 14MCN and 14ANCN, with a h value similar to
that of the other 14DMX molecules in Fig. 1.
decrease in d(H2). Instead, a downÐeld shift to 7.09 ppm takes
place, in the direction of the d(H2) of naphthalene (7.48 ppm,
Table 4).38 This observation has been explained by the elec-
tronic decoupling of the dimethylamino and naphthyl moieties
in 1DMAN, caused by its larger twist angle (60¡) as compared
with 1MAN and 1AN. A similar behaviour is found here
(Table 4) for the series 14DMCN, 14MCN and 14ANCN, for
which d(H2) equals 6.97, 6.53 and 6.72 ppm, respectively. It
hence follows that both 14DMCN and 1DMAN have larger
twist angles than their NH(CH ) and NH derivatives. This
3
2
conclusion supports the interpretation given in the previous
section of the shifts in the absorption spectra of the three
cyano-substituted 1-naphthylamines (Fig. 9). With the other
14DMX molecules 14DMCL, 14DMME and 14DMMO,
d(H2) ranges between 6.96 and 7.02 ppm, whereas values
Support for this conclusion is obtained from the extinction
coefficients emax of the three 4-cyano-1-naphthylamines (see
Fig. 9 and Table 1). The presence of one methyl substituent in
the amino group of 14MCN leads to an increase in emax
(13 100 M~1 cm~1) as compared with 14ANCN (11 800 M~1
between 6.63 and 6.69 ppm are found for their NH deriv-
2
atives (see Table 4). These results mean that the presence of
the dimethylamino group in the 14DMX compounds studied
here leads to considerably larger twist angles h (ca. 60¡) than
those of the 14ANX molecules (ca. 20¡).
cm~1), whereas upon further methylation (14DMCN)
a
decrease in emax to 10 300 M~1 cm~1 is observed. A similar
dependence of emax on the methyl substitution of the amino
group has been found for the three molecules 1AN, 1MAN
and 1DMAN (see Table 1), with twist angles h of 22¡, 15¡ and
60¡, respectively.5
Dipole moments l (S ) and l (S ) of 14DMX and 14ANX.
g
0
e 1
Solvatochromic data
Ground state dipole moment l (S ). InÑuence of amino
g
0
twist angle. The ground state dipole moment k (S ) of
g
0
1DMAN (1.05 D) is considerably smaller than those of 1AN
(1.55 D) and 1MAN (1.8 D).5 This di†erence has been attrib-
uted to the larger amino twist angle h of 1DMAN (60¡) as
compared with 1MAN (15¡) and 1AN (22¡).5,39,40 As experi-
1H NMR spectra
For 1DMAN and a number of other 1-naphthylamines, it has
been reported5 that the amino twist angle h has an inÑuence
on the chemical shifts d in their 1H NMR spectra. The twist
angle can then be monitored by measuring the NMR
spectra35h37 [see the data for d(H2) in Table 4]. The chemical
shift d(H2) of 1MAN has a lower value (6.65 ppm) than that of
1AN (6.78 ppm). This upÐeld shift has been attributed to the
electron donating nature of the methyl substituent. For both
1-aminonaphthalenes similar twist angles were calculated: 22¡
(1AN) and 15¡ (1MAN).5 The presence of the dimethylamino
group in 1DMAN, however, does not result in a further
mental k data are not available for the other 14DMX and
g
14ANX molecules, calculated41 values are used (see Table 5).
Support for the reliability of these dipole moments is derived
from plots of the calculated k (14DMX) and k (14ANX) data
g
g
against the experimental dipole moments k of the corre-
g
sponding 4-substituted N,N-dimethylanilines 14DX and anil-
ines 14AX (Fig. 10).42
Dipole moments l (S ). The dipole moment k (S ) of the
e
1
e 1
14DMX and 14ANX molecules in the equilibrated S state
1
was determined from solvatochromic measurements.43 In
order to minimise the inÑuence of speciÐc solvent e†ects, the
Ñuorescence band maxima lmax(Ñu) of these compounds in a
number of solvents were plotted against those of 1DMAN or
1AN in each medium.44 From the slope of these plots (Table
5), the dipole moment k (S ) can be determined by using the
e
1
equation45,46
[1
2
lmax(Ñu) \
k (k [ k )( f [ f @) ] constant (3)
e
e
g
4ne hco3
0
where
e [ 1
2e ] 1 2n2 ] 1
n2 [ 1
f [ f @ \
[
(4)
In eqn. (3), o is the equivalent spherical radius (Onsager
radius) of the solute and e is the vacuum permittivity of the
0
solvent. It is assumed that the dipole moment kFC of the
g
FranckÈCondon ground state reached upon emission from the
Fig. 9 Absorption spectra of the three 4-cyano-1-naphthylamines
14DMCN, 14MCN and 14ANCN in hexane at 25 ¡C.
relaxed S state is identical with the ground state dipole
1
Table 4 1H NMR chemical shifts d(H2)a for the proton H2 of the 4-substituted 1-naphthylamines 14DMX and 14ANX in CDCl at room
3
temperature
14DMX
d(H2) (ppm)
14ANX
d(H2) (ppm)
14MX
1MCN
1MAN
d(H2) (ppm)
6.53
14DMCN
14DMCL
1DMAN
14DMME
14DMMO
6.97
6.98
7.09
6.96
7.02
14ANCN
14ANCL
1AN
14ANME
14ANMO
6.72
6.63
6.78
6.69
6.65b
6.65
a For naphthalene a d(H2) of 7.48 ppm was determined; see ref. 38. b From ref. 37.
988
Phys. Chem. Chem. Phys., 2000, 2, 981È991

View Article Online
Table 5 Ground state [k (S )] and excited state [k (S )] dipole moments of the 4-substituted 1-naphthylamines 14DMX and 14ANX
g
0
e 1
oa/
k (S )b/
k (S )/
*kc/
g
0
e
1
pm
D
D
D
Sloped
14DMCN
14DMCL
1DMAN
14DMME
14DMMO
14ANCN
14ANCL
1AN
427
434
410f
419
430
405
413
370f
396
4.5
2.6
1.05f
0.9
1.3
6.2
3.6
1.55f
1.3
8.5
9.2
7.1f
7.5
8.1
9.8
9.8
7.2f
8.1
4.0
6.6
6.1
6.6
6.8
3.6
6.1
5.7
6.8
0.655e
1.181
(1.000)
1.076
1.080
0.663
1.059
(1.000)
1.097
14ANME
a Onsager radius, see eqns. (3) and (4). b See text (ref. 40). c *k \ k (S ) [ k (S ). d Slope of the plot of the Ñuorescence maxima lmax(Ñu) of
e
1
g 0
14DMX vs. those of 1DMAN and of 14ANX vs. those of 1AN in each solvent. e The value for 1-(N-azetidinyl)naphthalene is listed. The
Ñuorescence spectrum of this molecule consists of a single emission band in all solvents investigated, in contrast to 14DMCN, for which dual
Ñuorescence is observed in solvents more polar than alkanes. f Experimental data; see ref. 5.
moment k (S ).5,44 The slope of the plots given in Table 5 is
in the S ground state.3h5 This conclusion on the planarity of
g
0
0
equal to the ratio of the expression k (S )[k (S ) [ k (S )]/o3
S
has been derived from a comparison of 1DMAN with
e
1
e
1
g
0
1
[see eqn. (3)] for both pairs 14DMXÈ1DMAN and 14ANXÈ
1AN.
1AN, the molecule of which has a planar structure in the S
1
state.48 The same conclusion is drawn here for the 14DMX
and also for the 14ANX compounds, based on the similarity
of their dipole moments k (S ) with those of 1DMAN and
1AN (Table 5), and on the spectral position of the Ñuorescence
bands (Fig. 1). This means that for the naphthylamines
For the reference molecules 1DMAN and 1AN, dipole
moments k of 7.1 and 7.2 D have been determined by plot-
e
e 1
ting the energy lmax(Ñu) of the Ñuorescence band maxima
against the solvent polarity parameter f [ f @ [eqn. (4)].5 Obvi-
ously, the value taken for o strongly inÑuences the results
obtained for k . The k (S ) data in Table 5 are based on an
14DMX, strongly twisted in S , a large change in the twist
0
angle h has taken place upon reaching the equilibrated S
e
e
1
1
state. The structural similarity of the S state of 14DMX and
1
Onsager radius o calculated by using the assumption that the
molecular density equals unity.47 The dipole moments k (S )
14ANX together with the di†erent amino twist angles in their
e
1
of the 14DMX molecules have values between 7 and 9 D,
ground state S is reÑected by the considerably larger Stokes
0
whereas slightly larger dipole moments k (S ), between 7 and
shift *l(St) observed with the former molecules (Table 1).
e
1
10 D, are obtained for 14ANX (see Table 5).
Summary of amino twist angles in S and S
0
1
Structure of the relaxed S state
1
In the S ground state of the 14DMX molecules investigated,
0
In the relaxed S state, the dimethylamino group of 1DMAN
the amino twist angle h is larger (ca. 60¡) than for the corre-
1
is much less twisted with respect to the naphthyl moiety than
sponding 14ANX compounds (ca. 20¡). In the equilibrated S
1
state, 14DMX and 14ANX have a similar molecular structure
with a largely planar amino group.
Radiative rate constants of 14DMX and 14ANX
From the data for the Ñuorescence quantum yields U and the
f
decay times q of 14DMX and 14ANX at 25 ¡C in Tables 2 and
3, the radiative rate constants k (\U /q) can be calculated. For
f
f
the series 14DMX in hexane, the largest k (ca. 4 ] 108 s~1) is
f
observed with 14DMCN, with k values around 1 ] 108 s~1
f
for the other molecules, decreasing in the order X \ CN, Cl,
H, CH , OCH , down to 7 ] 107 s~1 for 14DMMO. A
3
3
similar observation is made for 14ANCN in hexane and
diethyl ether as compared with the other 14ANX molecules in
Table 3. The di†erence in dipole moment *k \ k (S ) [ k (S )
e
1
g 0
between the S and S states has the smallest value for
1
0
14DMCN and for 14ANCN: 4.0 D as compared with 6È7 D
for the other 14DMX and 14ANX compounds (Table 5).
These di†erences in *k are responsible for the relatively large
k value for 14DMCN and 14ANCN as compared with the
f
other 14DMX and 14ANX molecules, as a radiative rate con-
stant decreases with increasing di†erence in the CT character
of the S and S states. This notion has been used to explain
1
0
the small k values (D107 s~1) of exciplexes 1(A~D`)49h51
f
and of the ICT state of DMABN.52
Mechanism of internal conversion. InÑuence of vibronic
Fig. 10 Plot of the calculated ground state dipole moments k (S )calc
coupling [DE(S ,S )] and amino twist angle
g
0
1
2
of the 4-substituted 1-naphthylamines 14DMX and 14ANX vs. the
experimental dipole moments k (S )exp of the 4-substituted anilines
The fast IC process occurring in 1DMAN and related mol-
ecules has been attributed to vibronic coupling between the
g
0
14DX and 14AX (see text).
Phys. Chem. Chem. Phys., 2000, 2, 981È991
989

View Article Online
the group 14DMX (ca. 60¡) than of the molecules 14ANX (ca.
20¡). In the relaxed S state, both series of 1-naphthylamines
1
have become largely planar, as deduced from the similarity of
their Ñuorescence spectra.
The substituent X and the solvent polarity exert their inÑu-
ence on E by changing the energy gap *E(S ,S ), which
IC
1 2
governs the vibronic coupling between the potential energy
surfaces of S and S . This coupling Ñattens the S surface
1
2
1
when *E(S ,S ) becomes smaller, leading to a lowering of the
1 2
barrier E for the reaction from S to the ground state.
IC
1
After relaxation of the primarily excited S (FC) state to the
1
equilibrated S state, the IC reaction pathway is assumed to
1
Fig. 11 Potential surfaces of the electronic ground state S and the
0
two lowest excited singlet states
S
and
S
for the 1-
pass through a conical intersection of the S and S potential
1
2
1
0
aminonaphthalenes undergoing internal conversion (IC). The reaction
energy surfaces. This conical intersection exists as a conse-
coordinate m involves the amino twist angle. Light is absorbed to a
quence of the horizontal shift *m of the S and S surfaces
FranckÈCondon S state and Ñuorescence occurs after relaxation. The
1
0
1
caused by the di†erent amino twist angles and bond lengths of
surface of S is Ñattened by vibronic coupling with S . E is the acti-
1
2
IC
the 1-naphthylamines in these states. The increase in E when
vation energy for the non-adiabatic IC reaction from the equilibrated
IC
S
state to S . *E(S ,S ) is the energy gap between S and S .
the solvent polarity becomes larger is attributed to a prefer-
1
0
1
2
1
2
ential stabilisation of the S state relative to S , due to the
1
2
larger dipole moment of the former state. This increase in
two lowest excited singlet states S and S caused by the small
1
2
*E(S ,S ) reduces the vibronic coupling between the two
energy gap *E(S ,S ).5 For the IC reaction taking place with
1 2
1 2
states, which leads to a larger barrier for the IC reaction.
the 14DMX and 14ANX molecules investigated here, the
same mechanism is adopted (see Fig. 11).
After excitation to the S (FC) state with the same amino
1
Acknowledgements
twist angle as in S , a fast equilibration process sets in, involv-
0
ing the planarisation of the amino group of the 1-
Thanks are due to J. Bienert, W. Bosch, H. Lesche and J.
Schimpfhauser, Max-Planck-Institut fur biophysikalische
Chemie, Gottingen, for expert technical assistance. We also
thank Professor J. Nishimura, Gunma University, for the syn-
thesis of 4-methoxy-1-(dimethylamino)naphthalene. This work
was supported in part by a Grant-in-Aid for ScientiÐc
Research from the Japanese Ministry of Education, Science
and Culture.
naphthylamines. The thermally activated IC reaction then
starts from the relaxed S state. The S potential energy
1
1
surface is Ñattened by vibronic coupling between S and S ,
1
2
when the energy gap *E(S ,S ) is sufficiently small,53h55
which lowers the IC energy barrier E . The height of the IC
activation energy E therefore depends on the strength of the
1 2
IC
IC
vibronic coupling, governed by the gap *E(S ,S ), and also on
1 2
the shift *m determined by the amino twist angle.5
The substituents X in 14DMX and 14ANX and the solvent
polarity exert their inÑuence on the IC reaction by modifying
References
the energy gap *E(S ,S ) and hence the height of the IC acti-
1
2
3
S. I. Druzhinin and B. M. Uzhinov, T heor. Exp. Chem. Engl.
T ransl., 1982, 18, 565.
1 2
vation energy E (Fig. 11). The E barrier decreases with
IC
IC
S. R. Meech, D. V. OÏConnor, D. Phillips and A. G. Lee, J. Chem.
Soc., Faraday T rans. 2, 1983, 79, 1563.
increasing electron acceptor character of the X substituent,
with the lowest value for the cyano-substituted naphthyl-
K. A. Zachariasse, M. Grobys, Th. von der Haar, A. Hebecker,
Yu. V. IlÏichev, O. Morawski, I. Ruckert and W. Kuhnle, J.
Photochem. Photobiol., A, 1997, 105, 373.
K. Suzuki, H. Tanabe, S. Tobita and H. Shizuka, J. Phys. Chem.
A, 1997, 101, 4496.
I. Ruckert, A. Demeter, O. Morawski, W. Kuhnle, E. Tauer and
K. A. Zachariasse, J. Phys. Chem. A, 1999, 103, 1958.
E. Lippert, W. Luder, F. Moll, H. Nagele, H. Boos, H. Prigge and
I. Siebold-Blankenstein, Angew. Chem., 1961, 73, 695.
K. A. Zachariasse, Th. von der Haar, U. Leinhos and W. Kuhnle,
Pure Appl. Chem., 1993, 65, 1745.
W. Schuddeboom, S. A. Jonker, J. M. Warman, U. Leinhos, W.
Kuhnle and K. A. Zachariasse, J. Phys. Chem., 1992, 96, 10809.
J. Hicks, M. Vandersall, Z. Babarogic and K. B. Eisenthal, Chem.
Phys. L ett., 1985, 116, 18.
amines. With increasing solvent polarity, *E(S ,S ) becomes
1 2
larger.3,5 This leads to larger E barriers and hence smaller
IC
IC rate constants.
4
5
6
7
8
9
It is assumed that the IC reaction pathway passes through a
conical intersection,56 which allows the system to proceed to
the ground state S . The conical intersection comes into exis-
0
tence by the horizontal shift (*m) of the S and S potential
1
0
energy surfaces (Fig. 11), this shift being proportional to the
di†erence in amino twist angle h of the 1-naphthylamines in
the S and S states.
1
0
Conclusion
10 K. A. Zachariasse, M. Grobys, Th. von der Haar, A. Hebecker,
Yu. V. IlÏichev, Y.-B. Jiang, O. Morawski and W. Kuhnle, J.
Photochem. Photobiol., A, 1996, 102, 59.
For the 4-substituted 1-naphthylamines 14DMX and 14ANX,
a thermally activated internal conversion reaction takes place
in hexane, diethyl ether and acetonitrile. The efficiency of this
IC reaction depends on the nature of the substituent X,
decreasing in the order CN, Cl, H, CH , OCH , in which the
character of the substituent changes from electron acceptor to
electron donor. For every 14DMX and 14ANX molecule, the
IC process is less efficient in diethyl ether and acetonitrile than
in hexane, slowing down when the solvent polarity becomes
larger. The inÑuence of the substituent X and of the solvent
11 C. F. Lane, Synthesis, 1975, 135..
12 A. A. Ayuk, W. Rettig and E. Lippert, Ber. Bunsen-Ges. Phys.
Chem., 1981, 85, 553.
13 R. F. Borch and A. I. Hassid, J. Org. Chem., 1972, 37, 1673.
14 J. N. Demas and G. A. Crosby, J. Phys. Chem., 1971, 75, 991.
15 K. A. Zachariasse, G. Duveneck and R. Busse, J. Am. Chem. Soc.,
1984, 106, 1045.
16 U. Leinhos, W. Kuhnle and K. A. Zachariasse, J. Phys. Chem.,
1991, 95, 2013.
17 G. Striker, in Deconvolution and Reconvolution of Analytical
Signals, ed. M. Bouchy, University Press, Nancy, 1982, p. 329.
18 K. A. Zachariasse, W. Kuhnle, U. Leinhos, P. Reynders and G.
Striker, J. Phys. Chem., 1991, 95, 5476.
19 I. Carmichael, W. P. Helman and G. L. Hug, J. Phys. Chem. Ref.
Data, 1987, 16, 239.
20 L. Biczok, T. Berces and H. Inoue, J. Phys. Chem. A, 1999, 103,
3837.
3
3
polarity is mainly caused by the IC activation energy E
which increases when X changes in the series from CN to
,
IC
OCH and with increasing solvent polarity. The IC rate con-
3
stants k are larger for 14DMX than for 14ANX in each of
IC
the three solvents, likewise owing to larger activation energies
for the latter group of molecules. This di†erence in IC effi-
ciency is attributed to the larger ground state twist angle h of
990
Phys. Chem. Chem. Phys., 2000, 2, 981È991

View Article Online
21 A. Demeter and S. Braslavski, unpublished results.
22 I. Carmichael and G. L. Hug, J. Phys. Chem. Ref. Data, 1986, 15,
1.
37 N. J. Bunce, S. R. Cater, J. C. Scaiano and L. J. Johnston, J. Org.
Chem., 1987, 52, 4214.
38 H. S. Freeman, W. N. Hsu, J. F. Esancy and M. K. Esancy, Dyes
Pigm., 1988, 9, 67.
23 A. A. Lamola and G. S. Hammond, J. Chem. Phys., 1965, 43,
2129.
39 H. Lumbroso, J. Berlot, D. Bertin and J. Renault, C. R. Acad.
Sci., Ser. C, 1970, 270, 1204.
40 R. J. W. Le Fevre and A. Sundaram, J. Chem. Soc., 1962, 4756.
41 (a) PCMODEL, Serena Software, Bloomington, IN; (b) J. J.
Gajewski, K. E. Gilbert and L. McKelvey, in Advances in Molec-
ular Modelling, ed. D. Liotta, JAI Press, Greenwich, CT, 1990,
vol. 2.
24 E. Lippert, A. A. Ayuk, W. Rettig and G. Wermuth, J. Photo-
chem., 1981, 17, 237.
25 O. Morawski, I. Ruckert, K. Suzuki and K. A. Zachariasse,
unpublished results.
26 J. N. Murrell, T he T heory of the Electronic Spectra of Organic
Molecules, Methuen, London, 1963, ch. 11.
27 N. S. Isaacs, Physical Organic Chemistry, Longman, Harlow, 2nd
edn., 1995, ch. 4.
42 (a) A. L. McLellan, T ables of Experimental Dipole Moments,
Freeman, San Francisco, 1963, vol. 1; (b) A. L. McLellan, T ables
of Experimental Dipole Moments, Rahara Enterprises, El Cerrito,
CA, 1974, vol. 2; (c) A. L. McLellan, T ables of Experimental
Dipole Moments, Rahara Enterprises, El Cerrito, CA, 1989, vol. 3.
43 W. Baumann, H. Bischof, J.-C. Frohling, C. Brittinger, W. Rettig
and K. Rotkiewicz, J. Photochem. Photobiol., A, 1992, 64, 49.
44 K. A. Zachariasse, M. Grobys and E. Tauer, Chem. Phys. L ett.,
1997, 274, 372.
28 S. L. Murov, I. Carmichael and G. L. Hug, Handbook of Photo-
chemistry, Marcel Dekker, New York, 1993, p. 68.
29 J. Timmermans, Physico-Chemical Constants of Pure Organic
compounds, Elsevier, New York, 1950, vol. I.
30 J. B. Birks, Photophysics of Aromatic Molecules, Wiley, London,
1970.
31 The data for n at the various temperatures were calculated from
D
the corresponding solvent densities q by using the LorentzÈ
45 E. Lippert, Z. Elektrochem., 1957, 61, 962.
Lorenz equation q D (n2 [ 1)/(n2 ] 2). The values for q and n
46 W. Liptay, in Excited States, ed. E. C. Lim, Academic Press, New
York, 1974, vol. 1, p. 129.
D
(25¡C) were taken from the following references: (a) Landolt-
Bornstein, Numerical Data and Functional Relationships in
Science and T echnology, New Series, ed. W. Martienssen,
Springer, Berlin, 1996, vol. III/38b; (b) F. D. Rossini, K. S. Pitzer,
R. L. Arnett, R. M. Braun and G. C. Pimentel, Selected V alues of
Physical and T hermodynamic Properties of Hydrocarbons and
Related Compounds, Carnegie Press, Pittsburgh, PA, 1953; (c) J.
Timmermans, Physico-Chemical Constants of Pure Organic
L iquids, Elsevier, New York, 1950, vol. I.
47 Yu. V. IlÏichev, W. Kuhnle and K. A. Zachariasse, Chem. Phys.,
1996, 211, 441.
48 G. Berden, W. L. Meerts, D. F. Plusquelic, I. Fujita and D. W.
Pratt, J. Chem. Phys., 1996, 104, 3935.
49 H. Beens and A. Weller, in Organic Molecular Photophysics, ed.
J. B. Birks, Wiley, London, 1975, vol. 2, p. 159.
50 P. Pasman, F. Rob and J. W. Verhoeven, J. Am. Chem. Soc.,
1982, 104, 5127.
32 An activation energy of intersystem crossing E of 5 kJ mol~1
51 I. R. Gould, R. H. Young, L. J. Mueller, A. C. Albrecht and S.
Farid, J. Am. Chem. Soc., 1994, 116, 3147.
ISC
was obtained from the analysis of the decay times s of 1DMAN
in isopentane in the low temperature region down to its melting
52 Yu. V. IlÏichev, W. Kuhnle and K. A. Zachariasse, J. Phys. Chem.
A, 1998, 102, 5670.
point ([160 ¡C) see ref. 5. By Ðtting the Ñuorescence decay times
s as a function of temperature with E values over the range
53 R. M. Hochstrasser and C. A. Marzzacco, in Molecular L umines-
cence, ed. E. C. Lim, Benjamin, New York, 1969, p. 631.
54 E. C. Lim and C.-S. Huang, J. Chem. Phys., 1974, 61, 736.
55 G. Fischer, V ibronic Coupling, Academic Press, London, 1984.
56 F. Bernardi, M. Olivucci and M. A. Robb, Chem. Soc. Rev., 1996,
25, 321.
ISC
5 ^ 3 kJ mol~1, the same results for the IC activation energies
E
and the pre-exponential factors k0 were obtained.
IC
IC
33 Z. R. Grabowski, K. Rotkiewicz, W. Rubaszewska and E.
Kirkor-Kaminska, Acta Phys. Pol. A, 1978, 54, 767.
34 K. Kilsa, B. Albinsson and K. A. Zachariasse, in preparation.
35 L. Ernst, Z. Naturforsch., B, 1975, 30, 794.
36 A. Zweig, J. E. Lancaster and M. T. Neglia, T etrahedron, 1967,
23, 2577.
Paper a908924a
Phys. Chem. Chem. Phys., 2000, 2, 981È991
991