Twisted intramolecular charge transfer states in 2-arylbenzotriazoles: Fluorescence deactivation via intramolecular electron transfer rather than proton transfer

Article abstract of DOI:10.1021/jp021000j

Ultraviolet absorbers such as Tinuvin P (2-(2-hydroxy-5-methylphenyl)benzotriazole), 1, achieve their exceptional photostabilities as a result of deactivation of excited singlet states through excited state intramolecular proton transfer (ESIPT). Adding a methyl group to the 6′ position of 2-arylbenzotriazoles reveals an additional excited singlet state deactivation mechanism in this class of molecules which does not require intramolecular hydrogen bonding. Steady state fluorescence and fluorescence lifetime measurements for a series of 6′-methyl-2-arylbenzotriazoles provides compelling evidence for a twisted intramolecular charge transfer (TICT) mechanism of excited singlet state deactivation. Due to the steric requirements of the 6′-methyl group, conformations are favored in which the phenyl and triazole rings are no longer coplanar. In the case of compound 11 (2-(6-methoxy-2,3-dimethylphenyl)-2H-benzotriazole), the presence of a 2′-methoxy group enhances nonplanarity and results in large deactivation rates. Compound 12 (2-(6-methoxy-2,3-dimethylphenyl)-5-(trifluoromethyl)-2H-benzotriazole), which possesses both twist and enhanced donor/acceptor properties, undergoes the most efficient fluorescence quenching for the methoxyarylbenzotriazoles. Compounds with both a 6′-methyl and a hydroxy group on the phenyl ring exhibit diffusion controlled quenching (k_q = 2 × 10¹⁰ M^-1 s^-1) by DMSO. This quenching appears to result from either partial or complete excited state proton transfer to DMSO, which enhances TICT deactivation of the singlet excited state.

Full text of DOI:10.1021/jp021000j

7680
J. Phys. Chem. A 2002, 106, 7680-7689
Twisted Intramolecular Charge Transfer States in 2-Arylbenzotriazoles: Fluorescence
Deactivation via Intramolecular Electron Transfer Rather Than Proton Transfer
Ashok Maliakal,† George Lem,† Nicholas J. Turro,*,† Ravi Ravichandran,‡
Joseph C. Suhadolnik,‡ Anthony D. DeBellis,‡ Mervin G. Wood,‡ and Jacqueline Lau‡
Department of Chemistry, Columbia UniVersity, 3000 Broadway, New York, New York 10027, and
Ciba Specialty Chemicals, 540 White Plains Road, Tarrytown, New York 10591
ReceiVed: April 18, 2002; In Final Form: June 18, 2002
Ultraviolet absorbers such as Tinuvin P (2-(2-hydroxy-5-methylphenyl)benzotriazole), 1, achieve their
exceptional photostabilities as a result of deactivation of excited singlet states through excited state
intramolecular proton transfer (ESIPT). Adding a methyl group to the 6′ position of 2-arylbenzotriazoles
reveals an additional excited singlet state deactivation mechanism in this class of molecules which does not
require intramolecular hydrogen bonding. Steady state fluorescence and fluorescence lifetime measurements
for a series of 6′-methyl-2-arylbenzotriazoles provides compelling evidence for a twisted intramolecular charge
transfer (TICT) mechanism of excited singlet state deactivation. Due to the steric requirements of the 6′-
methyl group, conformations are favored in which the phenyl and triazole rings are no longer coplanar. In the
case of compound 11 (2-(6-methoxy-2,3-dimethylphenyl)-2H-benzotriazole), the presence of a 2′-methoxy
group enhances nonplanarity and results in large deactivation rates. Compound 12 (2-(6-methoxy-2,3-
dimethylphenyl)-5-(trifluoromethyl)-2H-benzotriazole), which possesses both twist and enhanced donor/acceptor
properties, undergoes the most efficient fluorescence quenching for the methoxyarylbenzotriazoles. Compounds
with both a 6′-methyl and a hydroxy group on the phenyl ring exhibit diffusion controlled quenching (kq )
2 × 1010 M-1 s-1) by DMSO. This quenching appears to result from either partial or complete excited state
proton transfer to DMSO, which enhances TICT deactivation of the singlet excited state.
Introduction
and the onset of both fluorescence and photoreactivity (loss of
stability) is observed.10,12,19,20 Thus, a paradigm has developed
which teaches that the best benzotriazole UVAs require both
coplanarity and a strong intramolecular hydrogen bond. Indeed,
parallels between intramolecular hydrogen bond strength and
performance as a UVA have been suggested as a means for
more efficient evaluation of promising lead compounds for
improved UVAs.11,14,21
The ability of certain classes of compounds to rapidly
dissipate excited state energy without irreversible photochem-
istry is a requirement for useful ultraviolet absorbers (UVAs).
These compounds have exceptional photostabilities with ex-
tremely small quantum yields of decomposition (10-6 or less).1-3
One class of such compounds is the 2-(2′-hydroxyaryl)benzo-
triazoles, of which Tinuvin P (1) is a member (see Figure 1).
Compound 1 has a very efficient pathway for deactivation of
S1 so that other excited state processes [i.e., intersystem crossing
(ISC), fluorescence, photoreaction] are unable to compete (see
Scheme 1).4,5
The mechanism of fluorescence deactivation of 1 and related
compounds has been the focus of several studies,5-12 whose
findings have converged on an excited state intramolecular
proton transfer mechanism (ESIPT) of fluorescence deactivation.
This mechanism is illustrated in Scheme 1. The excited singlet
state is capable of relaxing within ∼100 fs, via adiabatic transfer
of the phenolic proton to the nitrogen of the triazole ring
producing isomer K.13 The excited state isomer K, shows very
weak red shifted fluorescence under special conditions;4,9,11,14-18
however, subpicosecond internal conversion13 from the excited
“keto” form is the strongly favored process.9
Other mechanisms for ultrafast deactivation, such as twisted
intramolecular charge transfer (TICT),22,23 have been exploited
in applications such as laser dye absorbers used in mode-locked
lasers.24 In the TICT mechanism, rotation around a bond
connecting a donor group and an acceptor group can decouple
the orbitals of these two groups. This orbital decoupling provides
a means by which nearly complete electron transfer from the
donor to the acceptor can occur.25 A highly polar twisted excited
state results which, under appropriate solvent conditions, can
be preferentially stabilized with respect to a planar local excited
state. Since the highly twisted form is energetically destabilized
in the ground state, TICT formation brings the excited and
ground states closer, permitting more rapid internal conversion
(see Scheme 2). The concept of intramolecular electron transfer
as an alternative mode of action for ultraviolet stabilizers has
been suggested,26 and TICT states have been observed in ether
derivatives of hydroxyphenyltriazine UVAs.26,27 However, for
the triazine ethers, TICT states are longer lived than the locally
excited states, allowing more time for competitive photoreaction.
As a result, the ethers serve as poor UVAs. In benzotriazole
species, the elements of donor (phenol)28 and acceptor (tri-
azole)29 are present, and so the possibility also exists for TICT
states in these species.5 Therefore, we explored the effects of
The ESIPT mechanism has definite conformational require-
ments, the most important of which is a planar conformation
(closed form) that is capable of producing an intramolecular
hydrogen bond in the ground state. When the hydrogen bond is
broken (open form), the ESIPT deactivation mechanism is lost
† Columbia University.
‡ Ciba Specialty Chemicals.
10.1021/jp021000j CCC: $22.00 © 2002 American Chemical Society
Published on Web 07/31/2002

Twisted Intramolecular Charge Transfer States
J. Phys. Chem. A, Vol. 106, No. 34, 2002 7681
Figure 1. Arylbenzotriazoles studied. Ordered on the basis of the expected degree of twist along triazole C-N bond and donor/acceptor character.
SCHEME 1: Paradigm for ESIPT Deactivationa
additional propensity to twist in these compounds,11,30 and the
addition of the trifluoromethyl group to the triazole ring, which
is expected to enhance donor/acceptor character.31 Furthermore,
the substitution patterns were varied to investigate the influence
of the degree of twist. The results indicate that a charge transfer
mechanism is involved in singlet deactivation of compounds
that possess both a large degree of twist and donor/acceptor
character. The mechanism for TICT deactivation is outlined in
Scheme 2. The locally excited (Franck-Condon) state rapidly
twists and undergoes intramolecular electron transfer to form
the TICT state. The TICT state deactivates by internal conver-
sion through either a close approach or a conical intersection
with the ground state.32
a G denotes different functional groups substituted on the phenyl
ring.
We have found that twisted hydroxybenzotriazoles such as
6 undergo dynamic fluorescence quenching in the presence of
weak bases such as dimethyl sulfoxide (DMSO). These results
are discussed in the context of deactivation via either partial or
complete intermolecular proton transfer to solvent (PTTS) from
the TICT state (see Scheme 3).33-37 In this process, either partial
or full proton transfer from the benzotriazole to DMSO results
in deactivation of the excited state. Although intermolecular
proton transfer has been observed previously with Tinuvin P,20
in this case, no significant fluorescence deactivation was
observed. Although the intermolecular proton transfer can be
reversible in the case of Tinuvin P, the long lifetime of the
phenolate anion (80 µs in argon purged DMSO) allows for side
reactions such as oxidation of the phenolate anion.20 In the case
of the twisted hydroxybenzotriazole, 6, initial photostability
results show that this compound undergoes reduced photo-
degradation in the presence of DMSO. The enhanced photo-
stability of 6 in DMSO supports the possibility of excited state
deactivation through an aborted proton transfer, although other
possibilities cannot be ruled out.
SCHEME 2: Paradigm for TICT Deactivationa
a G represents functional groups attached to the aryl ring (X ) H or
Me).
structural modification on the photophysics of various 2-aryl-
benzotriazoles.
Our results suggest that although donor/acceptor character is
significant for TICT deactivation, the degree of twist is the more
important parameter. The compounds in this study provide a
clear example of a TICT excited singlet state deactivation
mechanism for the benzotriazole UVAs and provide a potential
new deactivation manifold for exploration toward the preparation
of better UVAs.
In this paper we report the photophysical properties of a series
of 2-arylbenzotriazole compounds illustrated in Figure 1. The
compounds are ordered with respect to the expected degree of
twist and the degree of donor/acceptor character. The specific
features installed in these compounds in order to enhance a
potential TICT effect are the 6′-methyl group which provides

7682 J. Phys. Chem. A, Vol. 106, No. 34, 2002
Maliakal et al.
SCHEME 3: Paradigm for Deactivation through Intermolecular Proton Transfera
a The TICT excited state deactivates to the ground state (G.S.) through either a partial (aborted) or complete proton transfer to DMSO.
Figure 4. Steady state emission of compounds 11 and 12 in hexanes
and DMSO. Excitation at 290 nm.
Figure 2. UV-absorption spectra of compounds 6 and 3 in hexanes
and DMSO.
in the amount of absorption in the ∼345 nm band and a dramatic
increase in the amount of fluorescence. This result is consistent
with the current mechanistic paradigm for singlet excited state
deactivation of benzotriazole UVAs such as Tinuvin P (1, Figure
1). The decrease in the long wavelength absorption is interpreted
as a breakage of the intramolecular hydrogen bond. Without
the intramolecular hydrogen bond, the ESIPT deactivation
mechanism can no longer operate, resulting in fluorescence.38
Compound 6, an isomer of 3, with the methyl group at the 6′
position rather than the 4′ position, has an intramolecular
hydrogen bond in hexanes and, as expected, shows no fluores-
cence. However, in DMSO, the ∼345 nm band is absent,
showing that the intramolecular hydrogen bond is completely
broken; nevertheless, the compound still shows almost no
fluorescence (Φf ∼10-4). This result disagrees with the standard
paradigm for benzotriazole UVAs, which predicts that the loss
of the hydrogen bond will lead to the onset of significant
fluorescence.10,20
Figure 3. Steady state emission of compounds 6 and 3 in hexanes
and DMSO. Excitation at 290 nm.
The methyl ethers 11 and 12, neither of which possess an
intramolecular hydrogen bond, both display reduced fluores-
cence in DMSO as compared to hexanes (see Figure 4). In
addition, the fluorescence is strongly red shifted. The quenching
of fluorescence is more dramatic for 12 than for 11.
Fluorescence quantum yields are presented in Table 1. The
quantum yields are reported relative to o-methyl Tinuvin 2 in
DMSO (Φf ) 0.135).39 Comparison of compounds 5 and 8
indicates that in DMSO the addition of a trifluoromethyl group
to the triazole ring decreases Φf by an order of magnitude. On
the other hand, in hexanes, the electron withdrawing group has
no effect on Φf. Exceptionally low quantum yields of fluores-
Results
Steady State UV and Fluorescence Spectra. UV-vis
absorption and fluorescence spectra are illustrated in Figures 2
and 3 for compounds 6 and 3. The photophysics of compound
3 is completely analogous to Tinuvin P (1).17 The observation
of a long wavelength UV-vis absorption in hexanes at ∼345
nm is characteristic of an intramolecular hydrogen bond.38 Due
to the intramolecular hydrogen bond, 3 deactivates via the
standard ESIPT mechanism.17 Hence, it is essentially nonfluo-
rescent. In DMSO, the UV-vis spectra of 3 shows a decrease

Twisted Intramolecular Charge Transfer States
J. Phys. Chem. A, Vol. 106, No. 34, 2002 7683
TABLE 1: Quantum Yields of Fluorescence in Hexanes and
DMSO
compound
Φf hexanes
Φf DMSO
4
9
3
5
8
6
7
10
11
12
<10-5
5 × 10-4 ( 4 × 10-5
1.2 × 10-3 ( 2 × 10-4
<10-5
<10-5
0.27 ( 0.03
0.27 ( 0.01
<10-5
0.27( 0.05
0.31 ( 0.04
0.08 ( 0.04
0.10 ( 0.05
6 × 10-3 ( 5 × 10-4
(0.02 ( 3) × 10-3
(0.001 ( 4) × 10-4
2 × 10-4 ( 1 × 10-4
0.14 ( 0.03
0.07 ( 0.02
(0.02 ( 3) × 10-3
(0.01 ( 3) × 10-3
cence are found in DMSO for the hydroxy compounds with
both 2′ and 6′ groups on the aryl ring (compare compounds 6
and 9 with compound 3). In the case of isomers 3 and 6 in
DMSO, moving the methyl group from the 4′ to the 6′ position
results in a 30-fold decrease in Φf . Addition of the trifluoro-
methyl group (9) drops Φf below our detection limits in DMSO
(Φf < 10-5). The less twisted isomers 3 and 4 exhibit weak
fluorescence in DMSO (Φf ∼ 10-2-10-3).
Figure 5. Stern-Volmer plot of DMSO fluorescence quenching of 3
and 6 in acetonitrile. Excitation at 290 nm.
emission. The reduction in lifetime tracks roughly with the
reduction in steady state fluorescence, indicating a predomi-
nantly dynamic quenching process with DMSO. The effective-
ness of DMSO quenching increases with the number of ortho
substituents and also when an electron-withdrawing group is
present on the triazole ring (compare 6 and 8 with 5). Compound
3 shows identical behavior in DMSO and acetonitrile (vide
infra).
DMSO Quenching of the Fluorescence of 6. Stern-Volmer
Analysis. The steady state fluorescence of 3 and 6 was measured
in acetonitrile as a function of increasing DMSO concentration.
The Stern-Volmer plot of the results are presented in Figure
5. In the case of compound 3, the steady state fluorescence
changes only very slightly upon addition of DMSO. Compound
6, on the other hand, showed a dramatic decrease in fluorescence
emission upon addition of DMSO. The absorption spectra of 6
is not affected by added DMSO. The emission maximum is
slightly red shifted on addition of DMSO. The slope (kqτF) of
the Stern-Volmer curve for 6 is 19.4 ( 0.7 M-1 in acetonitrile.
Using the fluorescence lifetime (τF) as measured using single
photon counting, the dynamic quenching constant (kq) is
Fluorescence Lifetimes. Phenols. The fluorescence lifetimes
were measured by single photon counting (SPC) for 5, 6, and
8 under varying conditions in order to assess the relative
importance of polarity and basicity on the fluorescence decay.
The pulse width of the lamp was ca. 1.5 ns. All compounds
were excited at 290 nm. Emission was observed for 5 at 433
nm, for 6 at 420 nm, and for 8 at 440 nm. Lifetimes were
calculated from the decay traces using a reconvolution analysis
with the lamp profile, and the data were fit to a monoexponen-
tial. This analysis generated fits with reasonable ø2 for the longer
lifetime measurements. For compounds with shorter lifetimes
(<0.8 ns) the values quoted are estimates based on reconvolution
analysis. The measurements do indicate clearly and reproducibly
that the lifetime has decreased significantly; however, the
accuracy in the measured lifetimes is limited by the resolution
of our instrument. We estimate the uncertainty in the short
lifetime measurements to be approximately (0.2 ns.
Table 3 illustrates the special quenching effect of DMSO on
the fluorescence of these hydroxy compounds. In each case,
DMSO, whose polarity is similar to acetonitrile,40,41 significantly
reduces both the fluorescence lifetime and the steady state
calculated to be 2 × 1010 ( 1 × 109 M-1 s-1
.
For compound 3 the UV-vis spectrum changes with addition
of DMSO, resulting in a decrease of the amount of long
wavelength (∼340 nm) absorption. At 1.4 M DMSO, the
TABLE 2: Emission and Absorption Maxima for Compounds Studied in Different Solvents
hexanes cyclohexane ether THF
abs
DMSO
methanol
comp.
em
abs
em
abs
em
abs
em
abs
em
abs
em
4
9
3
5
8
6
7
10
11
12
15
300, 358
297
300, 346
292
298
296, 326
294
300
286
282
x
291
284
294
288
296
284
290
294
282
282
320
464
x
373
376
377
386
376
381
387
368
378
437
435
457
446
419
441
415
460
402
294
300
286
284
320
384
383
367
378
378
290
298
286
282
318
390
407
380
404
380
292
298
282
282
320
397
416
385
423
387
288
296
280
280
318
421
441
437
464
401
TABLE 3: Steady State Fluorescence Emission Intensities and Fluorescence Lifetimes for 6, 8, and 5 in Acetonitrile with
Varying Amounts of DMSOa
6
8
5
conditions
Ifrel
τf (ns)
Ifrel
τf (ns)
Ifrel
τf (ns)
0.8 ( 0.1
MeCN
2.5% DMSO
100% DMSO
2.3
1
X
0.8 ( 0.1
0.3 ( 0.2
4.5 ( 0.5
0.8 ( 0.1
0.4 ( 0.2
5.5 ( 0.5
1
X
X
1
X
X
X
0.2 ( 0.2
a X means that these measurements were not performed.

7684 J. Phys. Chem. A, Vol. 106, No. 34, 2002
Maliakal et al.
TABLE 4: Comparison of Steady State Fluorescence Quantum Yields versus Lifetimes for 12, 7, 10, and 11a
compound
hexanes
DMSO
MeCN
7
τf (ns)
1.7 ( 0.1
0.27 ( 0.05
1.6
1.8 ( 0.1
0.31 ( 0.04
1.7
1.1 ( 0.1
0.08 ( 0.04
0.7
1.7 ( 0.1
0.10 ( 0.05
0.6
1.0 ( 0.1
1.0 ( 0.1
Φf
0.14 ( 0.03
1.4
0.15 ( 0.01
Φf/τf (s-1 × 108)
1.5
X
X
X
X
X
X
10
11
12
τf (ns)
1.0 ( 0.1
0.07 ( 0.02
0.7
0.8 ( 0.1
(0.02 ( 3) × 10-3
0.25
1.3 ( 0.1
(0.01 ( 3) × 10-3
0.08
Φf
Φf/τf (s-1 × 108)
τf (ns)
Φf
Φf/τf (s-1 × 108)
τf (ns)
1.0 ( 0.1
Φf
(0.01 ( 3) × 10-3
Φf/τf (s-1 × 108)
0.1
a X means experiment was not performed.
TABLE 5: Results from Lippert-Mataga Plots
compound
slope (cm-1
)
intercept (cm-1
)
∆µ (D)
µg (D)a
µe (D)
12
11
10
7
18169 ( 2,400
15868 ( 4100
13812 ( 1700
9802 ( 2300
5768 ( 1600
8418 ( 528
7027 ( 893
7038 ( 1373
7649 ( 503
4560 ( 345
17.3 ( 1.1
16.2 ( 2.1
15.1 ( 0.9
12.7 ( 1.5
9.7 ( 1.4
4.3
1.6
3.8
1.3
1.4
21.6 ( 1.1
17.8 ( 2.1
18.9 ( 0.9
14.0 ( 1.5
11.1 ( 1.4
13.7b,45
15
DMABN (TICT)
DMABN (LE)
10.1b,45
a Dipole moments calculated using Spartan-HF/6-31G(d) scaled by 0.87. b Determined by same solvatochromic method.
absorption at 340 nm is 75% of its value in neat acetonitrile
(see Supporting Information for details). The change in the UV-
vis spectrum is indicative of an increase in the amount of open
form. Thus the ESIPT paradigm suggests an increase in the
amount of fluorescence upon addition of DMSO. Although the
increase in the amount of open form could obscure DMSO
induced quenching for compound 3, the open form fluorescence
would only have a minor effect on the Stern-Volmer plot.
Taking into account this interference, we estimate that the
maximal Stern-Volmer constant (kqτF) for 3 would be ca. 0.7
M-1
.
Methyl Ethers. SPC was used to measure fluorescence
lifetimes for the methyl ethers (see above for experimental
details). The SPC data for the methyl ethers were fit to
monoexponential decays, with reasonable ø2. The lifetimes (τF)
of the methyl ethers 7, 10, 11, and 12 were compared with the
steady state emission intensities If as a function of solvent (see
Table 4). Although a dramatic decrease in If is observed on
switching to more polar solvents (DMSO, acetonitrile), the
lifetimes vary only slightly as a function of solvent. The
difference in the ratio of the steady state emission intensity
versus fluorescence lifetime increases as the number of ortho
substituents is increased from one to two, and upon introduction
of an electron-withdrawing substituent to the triazole ring. In
the case of 12, the steady state emission intensity changes by
an order of magnitude going from hexanes to DMSO; how-
ever, the lifetime varies only slightly for this dramatic polarity
change.
Lippert-Mataga Plots. Lippert-Mataga analysis was per-
formed on the methyl ether compounds in order to estimate the
magnitude of the excited state dipole moment in these com-
pounds. The energy difference of the fluorescence maximum
relative to the absorption maximum as a function of the solvent
orientation parameter (∆f) has been used as a measure of the
change in dipole moment between the fluorescing excited state
and the ground state.42,43 Equation 1, from Lippert-Mataga
theory, shows that the energy difference (νabs - νflu in cm-1) is
proportional to the orientation parameter (∆f).43 The Onsager
radius44 (a) is estimated from computer models (using Chem
Figure 6. Lippert-Mataga plot for 12 with ∆f ) orientation
parameter.42 Solvents: cyclohexane (1), ether (2), tetrahydrofuran (3),
dimethyl sulfoxide (4), methanol (5).
3D) to be 5.5 Å, h is Planck’s constant, and c is the speed of
light:
νjabs - νjflu ) [2(µe - µg)2/hca3]∆f + const
(1)
Figure 6 plots νabs - νflu in wavenumbers for compound 12.
The slope of this line can be used to calculate the change in
dipole moment on excitation of 12 using the Lippert-Mataga
theory. These values are presented for the methyl ethers 7, 10,
11, 12, and 15 in Table 5. These values can be compared with
dimethylaminobenzonitrile (DMABN), an exemplar TICT com-
pound, which is also listed in Table 5.45 Notice that for 10, 11,
and 12, the change in dipole on excitation is much larger than
the reported value for DMABN’s TICT state as determined by
the same method.45 The value calculated for the nontwisted
compound 15, is similar to that of the locally excited FC state
of DMABN. Unlike DMABN, in the case of all the compounds
in this study, only one fluorescence was observed.
Furans. The furan compounds 13 and 14 possess both planar
and nonplanar conformations (from UV-vis spectroscopy).
DMSO appears to shift the equilibrium more toward the
nonplanar state (see Figure 7). Interestingly, these results indicate

Twisted Intramolecular Charge Transfer States
J. Phys. Chem. A, Vol. 106, No. 34, 2002 7685
SCHEME 4: Paradigm for Photophysics of Methyl
Ethers
Our measurements of the excited state dipole moments in
the methyl ethers (see Table 5) using the Lippert-Mataga theory
support the hypothesis of a TICT excited state. The compounds
containing both 2′ and 6′ ortho substituents (11 and 12) have
the largest changes in dipole moment upon excitation, 16.2 and
17.3 D, respectively, leading to extremely large excited state
dipole moments. Compound 10, which has a sizable ground-
state dipole due to the presence of a trifluoromethyl group, also
exhibits a large excited state dipole moment that is comparable
in size to compound 11. Compound 7, however, has a lower
dipole moment, due to the fact that it has only one ortho
substituent, and does not have an electron withdrawing tri-
fluoromethyl group on its triazole ring. The ordering of excited
state dipole moments is consistent with the ordering of polar
solvent assisted excited singlet state deactivation (see Table 1),
and is in agreement with the principles for stabilization of TICT
states.23
Comparison of the fluorescence quantum yields and the
lifetimes for 10, 11, and 12 (see Table 4) indicate that the
variations in Φf are not parallel with the variations of τf on
going from a nonpolar to a polar solvent. The results suggest
that a large degree of the fluorescence is quenched by a process
occurring faster than the response time of our instrument. For
7, which has both a lower degree of twist and donor/acceptor
character, Φf/τf is constant for the different solvents.
We propose that compounds 10, 11, and 12 possess two
conformational states, a faster deactivating state (shorter than
our instrument’s response) and a longer lived fluorescent state
(τf ∼1-2 ns). Similar results were observed for the TICT states
of the triazine ethers.27 However, in that case, the shorter lifetime
component corresponded to the locally excited state, whereas
the longer lifetime component was assigned to the TICT state.
We suggest that in our system, both states are TICT states. It is
likely that the two states are conformers with a varying degree
of twist around the benzotriazole-aryl C-N bond and/or
difference in conformation of the methoxy group (see Scheme
4). The strong Stokes shift in the steady state fluorescence for
compounds 7, 10, 11, and 12 and the large excited state dipole
moments support the TICT nature of the longer lived excited
state. The correlation of the magnitude of deactivation with
greater twist and greater donor-acceptor character supports the
hypothesis that the shorter lived excited state is also TICT in
nature.
We propose that the behavior of Φf and τf can be explained
based upon Scheme 4. In this scheme, we postulate two ground
state conformations. Upon excitation of either of these conform-
ers, the initial locally excited states S1(LE) and S1′(LE) quickly
relax to the TICT states S1(TICT) and S1′(TICT). The state
S1(TICT) fluoresces. The fluorescence from S1(TICT) was used
to measure the excited state dipole moment (Table 5). The
presence of S1′(TICT) is inferred from a discrepancy between
Figure 7. UV-vis absorption spectra of furans 13 and 14.
that in the absence of steric repulsion, these compounds favor
a planar structure even without the aid of an intramolecular
hydrogen bond. The shift to more open form in DMSO suggests
the possibility of charge transfer in the ground state.
Discussion
The incorporation of a methyl group in the 6′ position of the
2-(2′-hydroxyphenyl)benzotriazole structure radically alters the
photophysics of this class of UVAs by enhancing a new
fluorescence deactivation mechanism which competes with or
replaces the standard ESIPT mechanism. Of the several possible
mechanisms of deactivation, the effect is attributed to a twisted
intramolecular charge transfer (TICT) process as indicated by
several pieces of evidence: (1) more efficient fluorescence
deactivation in polar solvents, (2) more efficient fluorescence
deactivation upon introduction of donor groups on the phenyl
ring and acceptor groups on the benzotriazole ring, (3) more
efficient fluorescence deactivation upon introduction of 6′
substituents in the phenyl ring (consistent with the rule of
minimum overlap in TICT states), and (4) large Stokes shifts
in fluorescence on changing solvent from hexanes to DMSO
(∼4700 cm-1 for 12) corresponding to a large change in excited
state dipole of (17.3 ( 1.1 D).22 Some other possibilities for
deactivation have been ruled out. Minor changes in absorption
spectra going from nonpolar to polar solvents and the observa-
tion of dynamic quenching effects rule out ground state
complexation. Energy transfer to solvent is not possible with
either DMSO or acetonitrile. ESIPT mechanisms in DMSO are
ruled out by either structure or by UV-vis data (no closed forms
in DMSO for 6′-methyl compounds). Phosphorescence meas-
urements at 77 K of 6 and 11 (not shown) showed similar or
smaller emission intensities compared to Tin P, indicating that
intersystem crossing is no more important in the 6′-methyl
compounds than in Tin P.
In the case of the methyl ethers (compounds 7, 10, 11, and
12), polarity is the most important solvent parameter in the new
deactivation scheme, since both the ESIPT mechanism and the
intermolecular proton transfer are not possible. For the hydroxy
compounds (compounds 4, 5, 6, and 8), the situation is more
complicated. Although polarity is important, solvent basicity
and hydrogen bond acceptor strength are also important. The
hydroxy compounds have the possibility of both an intra-
molecular and intermolecular proton transfer deactivation mech-
anism (Scheme 3), each of which is influenced by solvent.
With regard to the substrate, the two issues of donor/acceptor
effects and twist appear important in both the qualitative and
quantitative aspects of fluorescence deactivation.

7686 J. Phys. Chem. A, Vol. 106, No. 34, 2002
Maliakal et al.
the steady state and fluorescence lifetime measurements for
compounds 10, 11, and 12 (see Table 4). As structural features
are added, which push the equilibrium of the ground state to
the postulated high twist form, the decrease in Φf becomes larger
than the decrease in τf. This result supports the hypothesis that
the rapidly deactivating state is, in fact, also a TICT state. In
all cases the dynamic component of fluorescence quenching (i.e.,
the change in lifetime as a function of solvent polarity) results
from the deactivation pathway in which the S1 conformer
converts on the excited state surface to the S1′ conformer
followed by internal conversion (see Scheme 4).
Φf). The enhanced twist, therefore, appears to be the most
important factor in the TICT deactivation pathway. The donor/
acceptor character, nevertheless, is also important, and the largest
deactivation rates are observed for compound 12, which
incorporates both the twist (2′ + 6′ substitution) and charge
transfer (trifluoromethyl group in triazole ring) features.
Intermolecular and intramolecular hydrogen bonding in the
twisted state offer additional pathways for deactivation for the
phenolic benzotriazoles. In the intermolecular process, the
excited state can relax to a conical intersection32,36,46,47 through
partial or complete proton transfer to solvent,34 resulting in
ultrafast return to the ground state (see Scheme 3). With 2-(2′-
hydroxyphenyl)benzotriazoles, such as Tinuvin P, it has been
shown by laser flash photolysis that the deprotonation from the
excited state occurs in basic solvents such as DMSO.20 However
for Tinuvin P, DMSO serves to destabilize the planarity of the
molecule, resulting in an enhancement of fluorescence through
the loss of the competing ESIPT mechanism. In compounds in
which a 6′-methyl group is installed, e.g., 6 or 9, although
DMSO disrupts the intramolecular hydrogen bond, virtually no
fluorescence is observed (Φf ∼10-4-10-5). The effect is
peculiar to the 6′-methyl compounds (compare compounds 3
and 6 in Table 1 and see Figure 3), suggesting that a twist
between the benzotriazole and phenyl rings is essential for
increasing the efficiency of this mechanism of deactivation.
In compounds with a lower degree of twist (compare 5 and
8), the introduction of an electron withdrawing group into the
BZT-ring and a polar environment triggers the TICT deactiva-
tion mechanism.
In the case of compound 12, the conformation of the methoxy
group (methyl syn or anti to the benzotriazole ring) may be the
critical parameter determining the conversion between the two
states (see Scheme 4). When the methyl group is pointing away
from the triazole ring, a more planar structure may be favored,
taking advantage of energetic stabilization via conjugation of
the benzotriazole ring with the phenyl ring. When the methyl
group is pointing toward the benzotriazole ring, the twisted state
may be favored for both steric reasons as well as electronics, if
there is some charge transfer character in the ground state. This
hypothesis is supported by data for the furan compounds, 13
and 14, which show a long wavelength band in their UV
absorption spectra consistent with a significant population of a
planar conformer (see Figure 7). As in the case of Tinuvin P,
the presence of the long wavelength band is indicative of a
planar conformer. In DMSO, the long wavelength band is
reduced (see Figure 7), indicating a decrease in the amount of
the planar conformer. The ability of the polar solvent to increase
the amount of nonplanar conformer is consistent with the
presence of some charge transfer character in the ground state
of this conformer.
By studying the fluorescence of compound 6 as a function
of DMSO concentration and comparing this with lifetime
measurements, it is seen from the Stern-Volmer plot (Figure
5) that dynamic quenching occurs with kq ) 2 × 1010 M-1 s-1
.
Compound 12 has both a 2′-methoxy and a 6′-methyl which
favor the high twist side of the equilibrium in the ground state.
The presence of the trifluoromethyl group in the benzotriazole
ring would significantly stabilize charge transfer, and also favor
the higher twist side of the equilibrium. In the case of 7, which
has only one ortho group and no electron withdrawing group
on the triazole ring, Φf and τf both decrease by a factor of 2
going from hexanes to DMSO. This fact can be explained by
the lower degree of twist, resulting in a lower equilibrium
amount of high twist compound. Furthermore, the absence of
the trifluoromethyl group in the benzotriazole ring would reduce
the amount of stabilization of the TICT state, decreasing its
rate of deactivation. Compound 10, which has a trifluoromethyl
group in the benzotriazole ring in addition to the 6′-methyl
group, shows a 4-fold decrease in Φf on switching from hexanes
to DMSO, yet only a 2-fold decrease in τf . Compound 11, which
has two ortho groups on the phenyl ring, shows a 5-fold decrease
in Φf on switching from hexanes to DMSO, yet only a 1.5-fold
decrease in τf.
Note that for the isomer 3, which does not possess a 6′-methyl
(i.e., similar to Tinuvin P), no significant DMSO induced
deactivation is observed. These data are all consistent with
fluorescence deactivation through either partial or complete
excited state proton transfer to DMSO (see Scheme 3). Partial
proton transfer from the substrate to DMSO could lead the
substrate toward a conical intersection and hence permit fast
deactivation.36,48 Although intermolecular proton transfer could
potentially reduce photostability by producing a reactive
phenolate anion,20 initial stability studies indicate that 6 is
actually photostabilized in the presence of DMSO.49 The
enhanced photostability of 6 in DMSO supports the possibility
of excited state deactivation through an aborted proton transfer
rather than full PTTS.
The results for the methyl ethers corroborate these findings.
In the methyl ether series, DMSO and acetonitrile behave
similarly in their effects on fluorescence, as would be expected
considering their identical polarities (as defined by Kosower Z
values 71.1 and 71.3, respectively).40,41
Upon analysis of the quantum yields of fluorescence in Table
1, we can conclude that ortho substitution is more important
than donor/acceptor character for the TICT deactivation mech-
anism in methyl ethers. Notice that for the compounds 10 and
7, each with only one ortho group, the fluorescence deactivation
is approximately the same in hexanes. In DMSO, the tri-
fluoromethyl group in the benzotriazole ring reduces Φf for 10
only by a factor of 2. In combination with a 2′-methoxy group,
the 6′-methyl effect is much more dramatic. Comparing
compounds 7 and 11, we see a drop in Φf in both hexanes and
DMSO on moving the methoxy group from the 4′ to the 2′
position (∼3.5× and 7× reduction in Φf, respectively). Com-
pounds 10 and 12 mirror this trend (3× and 7× reduction in
Conclusion
The 2-(hydroxy or methoxy-6′-methylaryl)benzotriazoles
present a new class of compounds exhibiting TICT excited
states. The twisted-hydroxy benzotriazoles have an interesting
photochemical trigger for excited state deactivation via inter-
molecular proton transfer to basic solvents. The methoxy
compounds have an interesting equilibrium between a roughly
planar and a highly twisted form, each of which has unique
photophysical properties. The highly twisted form can deactivate
through charge transfer which accelerates the rate of internal
conversion. The planar form is capable in the excited state of

Twisted Intramolecular Charge Transfer States
J. Phys. Chem. A, Vol. 106, No. 34, 2002 7687
either fluorescence or conversion to the more highly twisted
form followed by deactivation. Donor/acceptor character in the
phenyl and benzotriazole portions is important to the rates of
deactivation. However, the steric effects which generate a twist
in the ground and excited states are essential for increasing the
efficiency of this mechanism of deactivation.
The complex photophysics and photochemistry of these
compounds will, no doubt, provide material for several further
studies and applications, and prove to be an interesting area for
future UVA research. A thorough understanding of the interplay
between ESIPT, excited state intermolecular proton transfer, and
intramolecular electron transfer mechanisms (TICT) will allow
for the tailoring of compounds with very useful properties as
UVAs and potentially in other applications. For example, the
ability of the (hydroxyaryl)benzotriazoles to switch between
fluorescent and nonfluorescent states as a function of environ-
ment could be exploited in applications such as molecular probes
and switches.
temperature for 2 h. To the reaction mixture, allyl bromide (4.84
g, 0.039 mol) and potassium iodide (0.34 g, 0.002 mol) were
added portion wise, after which the reaction mixture was heated
to 85 °C. After holding at 85 °C for 4.5 h, the solvent was
removed by distillation and replaced with 100 mL of heptane.
The mixture was washed three times with 40 mL of water. The
solvent was then removed by distillation to yield 14.2 g (99%)
of the corresponding O-allyl ether as an off-white solid. 1H NMR
(CDCl3): δ 8.35 (s, 1H), 8.20-8.12 (d, 1H), 7.67-7.61 (m,
2H), 7.54-7.49 (dd, 1H), 7.11-7.06 (d, 1H), 6.00-5.86 (m,
1H), 5.34-5.16 (m, 2H), 4.65-4.60 (d, 2H), 1.77 (s, 1H), 1.41
(s, 6H), 0.78 (s, 9H).
The O-allyl compound (14.2 g) as prepared above was added
to a round-bottom flask, heated to 190-195 °C, and held at
that temperature for 5 h. The residue was purified by flash
column chromatography with silica gel (ethyl acetate/heptane
gradient) to give 12.2 g (86%) of the C- allyl substituted
1
benzotriazole as a yellow oil. MS m/z 432 (M+H); H NMR
(CDCl3): δ 11.17 (s, 1H), 8.35-8.29 (m, 2H), 8.12-8.05 (d,
1H), 7.72-7.66 (dd, 1H), 7.34-7.29 (d, 1H), 6.18-6.02 (m,
1H), 5.20-5.06 (m, 2H), 3.64-3.53 (d, 2H), 1.81 (s, 2H), 1.46
(s, 6H), 0.78 (s, 9H).
Experimental Section
Materials and Solvents. DMSO, cyclohexane, and hexanes
were of spectrophotometric grade from Aldrich. THF (HPLC
grade) was from ACROS. Acetonitrile and methanol were
Optima grade from Fisher. Diethyl ether was from Mallinckrodt.
Solvents were tested for background absorption and emission
prior to use.
Measurements. UV-vis absorption spectra were recorded
on an HP8452A UV spectrometer. Fluorescence spectra were
measured with a Spex FluoroMax-2 spectrometer using a 1 cm
× 1 cm quartz cell and concentrations such that the optical
density was approximately 0.1 at the excitation wavelength (290
nm). Time-resolved fluorescence lifetimes were performed by
single photon counting on a OB900 fluorometer (Edinburgh
Analytical Instruments).
A mixture of 5-(trifluoromethyl)-2-(2-hydroxy-3-allyl-5-tert-
octylphenyl)-2H-benzotriazole (12.16 g, 0.028 mol), methane-
sulfonic acid (3.34 g, 0.034 mol), and chlorobenzene (50 mL)
was heated to 140 °C and held there for 3 h. After removal of
solvent, the brown oil was chromatographed twice on silica gel
(19:1 hexane/ethyl acetate) to yield 1.21 g (10%) as a yellow
viscous resin. 1H NMR (CDCl3): δ 8.36 (s, 1H), 8.12 (d, 1H),
7.87 (s, 1H), 7.60 (d, 1H), 7.35 (s, 1H), 5.21 (m, 1H), 3.47 (dd,
1H), 2.94 (dd, 1H), 1.79 (s, 2H), 1.57 (d, 3H), 1.44 (s, 6H), 0.8
(s, 9H); MS m/z 432 (M+H). Anal. Calcd for C24H28F3N3O:
C, 66.81; H, 6.54; N, 9.74. Found: C, 66.10; H, 6.25; N, 9.76.
2-[2-Methyl-5-(1,1,3,3-tetramethylbutyl)-2,3-dihydrobenzo-
furan-7-yl]-2H-benzotriazole (13). The synthesis of this com-
pound followed the same synthetic conditions used for com-
pound 14. A white solid (0.61 g, 5%) was obtained with a
melting point of 128-128.5 °C. 1H NMR (CDCl3): δ 8.00 (m,
2H), 7.85 (s, 1H), 7.42 (m, 2H), 7.30 (s, 1H), 5.22 (m, 1H),
3.43 (dd, 1H), 2.93 (dd, 1H), 1.76 (s, 2H), 1.56 (s, 3H), 1.42
(s, 6H), 0.80 (s, 9H). MS m/z 364 (M+H). Anal. Calcd for
C23H29N3O: C, 76.00; H, 8.04; N, 11.56. Found: C, 75.94; H,
7.98; N, 11.42.
Synthetic Procedures. General Details. Unless otherwise
specified, all reagents were purchased from commercial sup-
pliers and used without further purification. H NMR (300.08
1
and 499.84 MHz, respectively) spectra were obtained on a
Varian Gemini 300 MHz w/ASM-100 or a Unity 500 MHz
spectrometer. All 1H chemical shifts are reported in ppm relative
to tetramethylsilane, where a positive sign is downfield from
the standard. 19F NMR spectra (282.33 MHz) were obtained
on a Varian Gemini 300 MHz w/ ASM-100. 19F NMR chemical
shifts are reported in ppm relative to CF3COOH at -70.0 ppm.
Mass spectra were obtained on a Fisons VG Platform II in APcI
mode using a Hewlett-Packard LC II for sample delivery.
Elemental analyses were obtained from Robertson Microlit
Laboratories, Madison, NJ. Merck silica gel 60 (200-400 mesh)
was used for column chromatography. Merck precoated (0.25
mm) silica gel F-254 plates were used for TLC. Solvents were
dried prior to use when necessary with appropriate drying agents.
Reactions were carried out in dried apparatus under a dry inert
atmosphere of nitrogen.
General synthetic procedures for benzotriazole ring construc-
tion are found in U.S. Patents 5,977,219 and 6,166,218.
Compounds for photophysical studies were purified sequentially
by chromatography (silica gel) and sublimation, except 14,
which was chromatographed twice on silica gel and sublimed.
2-[2-Methyl-5-(1,1,3,3-tetramethylbutyl)-2,3-dihydrobenzo-
furan-7-yl]-5-(trifluoromethyl)-2H-benzotriazole (14). A mixture
of 5-(trifluoromethyl)-2-(2-hydroxy-5-tert-octylphenyl)-2H-ben-
zotriazole (13.01 g, 0.033 mol), potassium hydroxide (2.37 g,
0.036 mol), and ethanol (60 mL) was stirred at ambient
3,4-Dimethyl-2-(5-(trifluoromethyl)benzotriazol-2-yl)phenol
(9). A light yellow solid (0.36 g, 42%) with a melting point of
1
127.5-128.5 °C. H NMR (CDCl3): δ 8.38 (s, 1H), 8.15 (d,
1H), 7.71 (s, 1H), 7.68 (dd, 1H), 7.25 (d, 1H), 6.96 (d, 1H),
2.35 (s, 3H), 2.27 (s, 3H). MS m/z 308 (M+H). Anal. Calcd
for C15H12F3N3O: C, 58.63; H, 3.94; N, 13.67. Found: C, 58.66;
H, 3.71; N, 13.60.
2,3-Dimethyl-4-(5-(trifluoromethyl)benzotriazol-2-yl)phenol
(8). A white solid (15 g, 46%) with a melting point of 135-
1
136 °C. H NMR (CDCl3): δ 8.33 (d, 1H), 8.09 (d, 1H), 7.65
(dd, 1H), 7.31 (d, 1H), 6.78 (d, 1H), 5.30 (broad s, 1H), 2.29
(s, 3H), 2.14 (s, 3H). 19F NMR (CDCl3): -68.9 ppm. Anal.
Calcd for C15H12F3N3O: C, 58.63; H, 3.94; N, 13.67. Found:
C, 58.53; H, 3.88; N, 13.72.
2-(4-Methoxy-2,3-dimethylphenyl)-5-(trifluoromethyl)-2H-
benzotriazole (10). A white solid (0.30 g, 27%) with a melting
point of 109-111 °C. 1H NMR (CDCl3): δ 8.34 (s, 1H), 8.10
(d, 1H), 7.63 (dd, 1H), 7.42 (d, 1H), 6.88 (d, 1H), 3.92 (s, 3H),
2.27 (s, 3H), 2.15 (s, 3H). MS m/z 322 (M+H). Anal. Calcd
for C16H14F3N3O: C, 59.81; H, 4.39; N, 13.08. Found: C, 59.65;
H, 4.10; N, 13.08.

7688 J. Phys. Chem. A, Vol. 106, No. 34, 2002
Maliakal et al.
2-(6-Methoxy-2,3-dimethylphenyl)-2H-benzotriazole (11). A
white solid (0.46 g, 48%) with a melting point of 121.5-122.5
°C. 1H NMR (CDCl3): δ 8.00 (m, 2H), 7.44 (m, 2H), 7.30 (d,
1H), 6.84 (d, 1H), 3.72 (s, 3H), 2.30 (s, 3H), 1.81 (s, 3H). MS
m/z 254 (M+H). Anal. Calcd for C15H15N3O: C, 71.13; H, 5.97;
N, 16.59. Found: C, 71.01; H, 5.82; N, 16.75.
2-Benzotriazol-2-yl-3,4-dimethylphenol (6). A white solid
(0.22 g, 17%) with a melting point of 166.5-167.5 °C. 1H NMR
(CDCl3): δ 8.04 (s, 1H), 8.00 (m, 2H), 7.51 (m, 2H), 7.21 (d,
1H), 6.95 (d, 1H), 2.34 (s, 3H), 2.32 (s, 3H). MS m/z 240
(M+H). Anal. Calcd for C14H13N3O: C, 70.28; H, 5.48; N,
17.56. Found: C, 70.02; H, 5.58; N, 17.53.
References and Notes
(1) Otterstedt, J. J. Chem. Phys. 1973, 58, 5716.
(2) Decker, C.; Biry, S.; Zahouily, K. Polym. Degrad. Stab. 1995, 42,
111.
(3) Catalan, J.; Valle, J. D.; Fabero, F.; Garcia, N. Photochem.
Photobiol. 1995, 61, 118.
(4) Weichmann, M.; Port, H.; Frey, W.; Larmer, F.; Elsaesser, T. J.
Phys. Chem. 1991, 95, 1918.
(5) Goeller, G.; Rieker, J.; Maier, A.; Stezowski, J.; Daltrozzo, E.;
Neureiter, M.; Port, H.; Wiechmann, M.; Kramer, H. J. Phys. Chem. 1988,
92, 1452.
(6) Kummrow, A.; Dreyer, J.; Chudoba, C.; Stenger, J.; Nibbering, E.
T. J.; Elsaesser, T. J. Chin. Chem. Soc. (Taipei) 2000, 47, 721.
(7) Weichmann, M.; Port, H. J. Lumin. 1991, 48,49, 217.
(8) Estevez, C.; Bach, R.; Hass, K.; Schneider, W. J. Am. Chem. Soc.
1997, 119, 5445.
(9) Flom, S.; Barbara, P. Chem. Phys. Lett. 1983, 94, 488.
(10) Ghiggino, K.; Scully, A.; Leaver, I. J. Phys. Chem. 1986, 90, 5089.
(11) Rieker, J.; Lemmert-Schmitt, E.; Goeller, G.; Roessler, M.; Stueber,
G.; Schettler, H.; Kramer, H.; Stezowski, J.; Hoier, H.; Henkel, H.; Schmidt,
A.; Port, H.; Weichmann, M.; Rody, J.; Rytz, G.; Slongo, M.; Birbaum, J.
J. Phys. Chem. 1992, 96, 10225.
4-Benzotriazol-2-yl-2,3-dimethylphenol (5). A light yellow
1
solid (1.79 g, 14%) with a melting point of 189-190 °C. H
NMR (CDCl3): δ 7.94 (m, 2H), 7.46 (m, 2H), 7.30 (d, 1H),
6.78 (d, 1H), 5.19 (s, 1H), 2.28 (s, 3H), 2.14 (s, 3H). MS m/z
240 (M+H). Anal. Calcd for C14H13N3O: C, 70.28; H, 5.48;
N, 17.56. Found: C, 70.06; H, 5.51; N, 17.65.
(12) Woessner, G.; Goeller, G.; Rieker, J.; Hoier, H.; Stezowski, J.;
Daltrozzo, E.; Neureiter, M.; Kramer, H. J. Phys. Chem. 1985, 89, 3629.
(13) Frey, W.; Elsaesser, T. Chem. Phys. Lett. 1992, 189, 565.
(14) Dux, R.; Ghiggino, K.; Vogl, O. Aust. J. Chem. 1994, 47, 1461.
(15) O’Connor, D.; Scott, G.; Coulter, D.; Gupta, A.; Webb, S.; Yeh,
S.; Clark, J. Chem. Phys. Lett. 1985, 121, 417.
2-(4-Methoxy-2,3-dimethylphenyl)-2H-benzotriazole (7). A
white solid (0.17 g, 40%) with a melting point of 110-110.5
°C. 1H NMR (CDCl3): δ 7.94 (m, 2H), 7.44 (m, 2H), 7.41 (d,
1H), 6.85 (d, 1H), 3.91 (s, 3H), 2.28 (s, 3H), 2.14 (s, 3H). MS
m/z 254 (M+H). Anal. Calcd for C15H15N3O: C, 71.13; H, 5.97;
N, 16.59. Found: C, 71.03; H, 5.81; N, 16.69.
(16) Scully, A.; Bigger, S.; Ghiggino, K.; Vogl, O. J. Photochem.
Photobiol. A 1991, 55, 387.
2-Benzotriazol-2-yl-4,5-dimethylphenol (3). A white solid
(0.15 g, 11%) with a melting point of 143-143.5 °C. 1H NMR
(CDCl3): δ 11.04 (s, 1H), 8.14 (s, 1H), 7.95 (m, 2H), 7.46 (m,
2H), 7.01 (s, 1H), 2.32 (s, 3H), 2.31 (s, 3H). MS m/z 240
(M+H). Anal. Calcd for C14H13N3O: C, 70.28; H, 5.48; N,
17.56. Found: C, 69.98; H, 5.40; N, 17.63.
2-(6-Methoxy-2,3-dimethylphenyl)-5-(trifluoromethyl)-2H-
benzotriazole (12). A light yellow solid (0.35 g, 9%). 1H NMR
(CDCl3): δ 8.36 (s, 1H), 8.11 (d, 1H), 7.65 (dd, 1H), 7.35 (d,
1H), 6.86 (d, 1H), 3.72 (s, 3H), 2.31 (s, 3H), 1.81 (s, 3H). MS
m/z 322 (M+H). Anal. Calcd for C16H14F3N3O: C, 59.81; H,
4.39; N, 13.08. Found: C, 59.58; H, 4.03; N, 13.11.
(17) Weichmann, M.; Port, H.; Laermer, F.; Frey, W.; Elsaesser, T.
Chem. Phys. Lett. 1990, 165, 28.
(18) Woessner, G.; Goeller, G.; Kollat, P.; Stezowski, J.; Hauser, M.;
Klein, U.; Kramer, H. J. Phys. Chem. 1984, 88, 5544.
(19) DeBellis, A.; Iyengar, R.; Kaprinidis, N.; Rodebaugh, R.; Suha-
dolnik, J. “Proceedings of Service Life Methodologies and Metrologies
Symposium”; Service Life Methodologies and Metrologies Symposium,
1999, Monterey, CA.
(20) McGarry, P. F.; Jockusch, S.; Fujiwara, Y.; Kaprinidis, N. A.; Turro,
N. J. Phys. Chem. A. 1997, 101, 764.
(21) Dai, G.; Wu, S.; Sustic, A.; Vogl, O. Polym. Bull. 1988, 20, 67.
(22) Rettig, W. Top. Curr. Chem. 1994, 169, 254 and references therein.
(23) Grabowski, Z.; Dobkowski, J. Pure Appl. Chem. 1983, 55, 245.
(24) Vogel, M.; Rettig, W. Ber. Bunsen-Ges. Phys. Chem. 1985, 89,
962.
4,5-Dimethyl-2-(5-(trifluoromethyl)benzotriazol-2-yl)phenol
(4). A light yellow solid (4.02 g, 9%) with a melting point of
113-114 °C. 1H NMR (CDCl3): δ 10.75 (s, 1H), 8.29 (d, 1H),
8.16 (s, 1H), 8.06 (d, 1H), 7.67 (d, 1H), 7.02 (s, 1H), 2.33 (s,
3H), 2.32 (s, 3H). 19F NMR (CDCl3): -68.9 ppm; MS m/z 308
(M+H). Anal. Calcd for C15H12F3N3O: C, 58.63; H, 3.94; N,
13.67. Found: C, 58.55; H, 3.79; N, 13.60.
(25) Rettig, W. Appl. Phys. B 1988, 45, 145.
(26) Elbe, F.; Keck, J.; Fluegge, A.; Kramer, H.; Fischer, P.; Hayoz,
P.; Leppard, D.; Rytz, G.; Kaim, W.; Ketterle, M. J. Phys. Chem. A 2000,
104, 8296.
(27) Keck, J.; Roessler, M.; Schroeder, C.; Stueber, G. J.; Waiblinger,
F.; Stein, M.; LeGoumie´rec, D.; Kramer, H. E. A.; Hoier, H.; He´nkel, S.;
Fischer, P.; Port, H.; Hirsch, T.; Rytz, G.; Hayoz, P. J. Phys. Chem. B
1998, 102, 6975.
(28) Jovanovic, S.; Tosic, M.; Simic, M. J. Phys. Chem. 1991, 95, 10824.
(29) Kamiya, M.; Akahori, Y. Bull. Chem. Soc. Jpn. 1970, 43, 268.
(30) Schuddeboom, W.; Jonker, S. A.; Warman, J. M.; Leinhos, U.;
Kuehnle, W.; Zachariasse, K. A. J. Phys. Chem. 1992, 96, 10809.
(31) Suhadolnik, J.; DeBellis, A.; Hendricks-Guy, C.; Iyengar, R.; Wood,
M. “Unexpected Electronic Effects on benzotriazole UV absorber photo-
stability: Mechanistic Implications beyond excited state intramolecular
proton transfer”; ACS National Meeting, 2000, Washington, DC.
(32) Bernardi, F.; Olivucci, M.; Robb, M. Chem. Soc. ReV. 1996, 25,
321.
2-(4-Methoxyphenyl)-2H-benzotriazole (15). A white solid
1
(2.12 g, 24%) with a melting point of 111-112 °C. H NMR
(CDCl3): δ 8.29 (m, 2H), 7.93 (m, 2H), 7.42 (m, 2H), 7.07
(m, 2H), 3.91 (s, 3H). MS m/z 225, 210, 182. Anal. Calcd for
C13H11N3O: C, 69.32; H, 4.92; N, 18.65. Found: C, 69.10; H,
4.80; N, 18.65.
Acknowledgment. The authors at Columbia thank the
National Science Foundation (Grants CHE-98-12676 and CHE-
01-10655) for its generous support of this research. A.J.M.
thanks the NSF for a predoctoral fellowship and Drs. Mark
Kleinman, Steffen Jockusch, and Tracy Morkin for their
comments on the manuscript. We thank the reviewers for their
insightful comments and suggestions.
(33) Solntsev, K.; Huppert, D.; Agmon, N. J. Phys. Chem. A 1999, 103,
6984.
(34) Acuna, A.; Oton, J. J. Lumin. 1979, 20, 379.
(35) Dobkowski, J.; Herbich, J.; Galievsky, V.; Thummel, R.; Wu, F.;
Waluk, J. Ber. Bunsen-Ges. Phys. Chem. 1998, 102, 469.
(36) Nau, W.; Greiner, G.; Rau, H.; Olivucci, M.; Robb, M. Ber. Bunsen-
Ges. Phys. Chem. 1998, 102, 486.
(37) Shizuka, H. Acc. Chem. Res. 1985, 18, 141.
(38) Werner, T. J. Phys. Chem. 1979, 83, 320.
(39) Woessner, G. Photophysikalische und photochemische Desaktiv-
ierungsprozesse der UV-Stabilisatoren des Benzotriazol-Typs. Doctoral
thesis, University of Stuttgart, 1983.
(40) Kosower, E. J. Am. Chem. Soc. 1957, 80, 3253.
(41) Kosower, E. J. Am. Chem. Soc. 1957, 80, 3261.
(42) Stueber, G.; Kieninger, M.; Schettler, H.; Busch, W.; Goeller, B.;
Franke, J.; Kramer, H.; Hoier, H.; Henkel, S.; Fischer, P.; Port, H.; Hirsch,
T.; Rytz, G.; Birbaum, J. J. Phys. Chem. 1995, 99, 10097.
Supporting Information Available: UV-vis and emission
spectra for 6 and 3 in DMSO and acetonitrile. Table of optical
density of long wavelength (340 nm) absorption band in binary
DMSO/acetonitrile solvent system as a function of DMSO. This
information is available free of charge via the Internet at http://
pubs.acs.org.

Twisted Intramolecular Charge Transfer States
J. Phys. Chem. A, Vol. 106, No. 34, 2002 7689
(43) Liptay, W. Angew. Chem., Int. Ed. 1969, 8, 177.
(44) Onsager, L. J. Am. Chem. Soc. 1936, 99, 1486.
(45) Rettig, W.; Braun, D.; Suppan, P.; Vauthey, E.; Rotkiewicz, K.;
Luboradzki, R.; Suwinska, K. J. Phys. Chem. 1993, 97, 13500.
(46) Michl, J. J. Mol. Photochem. 1972, 4, 243.
(48) Nau, W.; Greiner, G.; Wall, J.; Rau, H.; Olivucci, M.; Robb, M.
Angew. Chem., Int. Ed. 1998, 37, 98.
(49) Compound 6 (OD >2 absorbance units) was irradiated at 300 nm.
The loss rate measured in neat acetonitrile was >0.7 absorbance units/
hour. In 3:1 acetonitrile/DMSO the loss rate drops to 0.16 absorbance units/
hour. Tinuvin P, under identical conditions, showed a loss rate of 0.08
absorbance units/hour in acetonitrile, and >0.7 abs. units/hour in 3:1
acetonitrile/DMSO.
(47) Sinicropi, A.; Pischel, U.; Basosi, R.; Nau, W.; Olivucci, M. Angew.
Chem., Int. Ed. 2000, 39, 4582.