Laser flash photolysis and CIDNP studies of 1-naphthyl acetate photo-fries rearrangement

Article abstract of DOI:10.1021/jp952315k

The steady-state and time-resolved CIDNP and flash photolysis methods were used in a detailed study of the photo-Fries rearrangement of 1-naphthyl acetate (I) in acetonitrile and methanol. The main reaction channel is the decay of I through the excited singlet state with the quantum yields 0.17 ±0.02 in acetonitrile and 0.42 ± 0.04 in methanol at room temperature. The absorption spectra of the naphthoxyl radical and triplet state of 1-naphthyl acetate were detected. The quantum yield of triplet was estimated as 0.4 ±0.2 and 0.35 ± 0.17 in acetonitrile and methanol, respectively. It has been established that the triplet-born radical pairs make a main contribution to the CIDNP of the photo-Fries rearrangement products. The involvement in the process of two different triplet states of I was supposed. The main decay channel of the lowest triplet state is the triplet-triplet annihilation, while the CIDNP of photo-Fries rearrangement products results from the decay of the upper triplet state of I with a lifetime of a few nanoseconds. The kinetics of CIDNP formation in reaction products has been analyzed, and the rate constants of the rearrangement of the preceding intermediates at room temperature have been estimated.

Full text of DOI:10.1021/jp952315k

4
448
J. Phys. Chem. 1996, 100, 4448-4458
Laser Flash Photolysis and CIDNP Studies of 1-Naphthyl Acetate Photo-Fries
Rearrangement
Nina P. Gritsan*
Institute of Chemical Kinetics and Combustion and NoVosibirsk State UniVersity, 630090, NoVosibirsk, Russia
Yuri P. Tsentalovich,†,‡ Alexandra V. Yurkovskaya, and Renad Z. Sagdeev
†
†
International Tomography Center, 630090, NoVosibirsk, Russia, and Physikalisch-Chemisches Institut der
UniVersit a¨ t Z u¨ rich, Winterthurerstrasse 190, CH-8057, Switzerland
X
ReceiVed: August 8, 1995; In Final Form: December 8, 1995
The steady-state and time-resolved CIDNP and flash photolysis methods were used in a detailed study of the
photo-Fries rearrangement of 1-naphthyl acetate (I) in acetonitrile and methanol. The main reaction channel
is the decay of I through the excited singlet state with the quantum yields 0.17 ( 0.02 in acetonitrile and 0.42
(
0.04 in methanol at room temperature. The absorption spectra of the naphthoxyl radical and triplet state
of 1-naphthyl acetate were detected. The quantum yield of triplet was estimated as 0.4 ( 0.2 and 0.35 (
.17 in acetonitrile and methanol, respectively. It has been established that the triplet-born radical pairs
0
make a main contribution to the CIDNP of the photo-Fries rearrangement products. The involvement in the
process of two different triplet states of I was supposed. The main decay channel of the lowest triplet state
is the triplet-triplet annihilation, while the CIDNP of photo-Fries rearrangement products results from the
decay of the upper triplet state of I with a lifetime of a few nanoseconds. The kinetics of CIDNP formation
in reaction products has been analyzed, and the rate constants of the rearrangement of the preceding
intermediates at room temperature have been estimated.
Introduction
SCHEME 1
The photo-Fries rearrangement was first discovered by
1
Anderson and Reese in 1960. Over the years this phenomenon
has been extensively studied and has been observed not only
1
-5
in aryl esters
but also in various aromatic compounds,
2,6-8
9
2,10
including acetanilides,
carbamates, sulfamates, cinnamates, etc. Since the pio-
neering works,
sulfonanilides, aryl sulfonates,
1
1
12
13
constants of elementary stages are lacking. The absorption of
1
4,15
23
attempts have been made to elucidate the
two intermediates was detected by Kalmus and Hercules upon
photo-Fries rearrangement mechanism. On the basis of the facts
that product quantum yields are not affected by typical triplet
the pulse excitation of phenyl acetate. One of the intermediates
was spectroscopically identified as the phenoxyl radical. The
other intermediate was assigned to cyclohexadienone which
transforms to o-hydroxyacetophenone. The transformation of
the cyclohexadienone intermediates to o- and p-hydroxy ac-
2
,3,7,8
2
quenchers
and the triplet sensitization is absent, some
authors have assumed that the photo-Fries rearrangement occurs
through the excited singlet state. This assumption has been
1
6,17
24
supported by the analysis of the signs of CIDNP spectra
etophenones has been recently described in detail by Arai.
detected during the photoirradiation of aryl esters. The appear-
This paper describes a detailed kinetic study of the pho-
totransformation of 1-naphthyl acetate (I). It has been known
that the irradiation of I in various solutions gives rise to the
products typical of the photo-Fries rearrangement: 2-acetyl-1-
ance of the CIDNP effects has also revealed the radical
mechanism of the photo-Fries rearrangement.1
6,17
However,
some recent data indicate that in the case of phenylsulfamates
and cinnamates the photo-Fries rearrangement occurs through
the upper triplet states.1
2,4,5
naphthol (II), 4-acetyl-1-naphthol (III), and 1-naphthol (IV)
2,13
(
Scheme 1).
The photo-Fries rearrangement has much in common with
such photochemical reactions as the photo-Claisen rearrange-
5
Nakagaki has detected the effect of an external magnetic
13
field on the yield of product II for C-labeled 1-naphthyl acetate
13CdO). The value of the external magnetic effect (1.03 (
.01) suggests that the in-cage product originates from a singlet
1
8,19
and the â-cleavage of phenoxy ketones.20,21 The
ment
(
0
primary process of these phototransformations is the homolytic
dissociation of the carbon-heteroatom bond, yielding a pair of
free radicals. The recombining radicals form rearrangement
products. A theoretical study of the homolytic photodissociation
radical pair. The absorption of the intermediate in the laser
5
excitation of I was attributed to the 1-naphthoxyl radical.
However, this attribution seems to be justified insufficiently.
of the carbon-oxygen bond in R-O-Ph compounds has
recently been performed by Grimme.22
This work was aimed at determining the role of the excited
singlet and triplet states in the photo-Fries rearrangement of
compound I as well as at establishing the nature, spectral
characteristics, and reactivity of the intermediates. To make
the results more valid, we used complementary methods,
including laser flash photolysis and steady-state and flash
CIDNP.
There are many works concerned with the investigation of
the photo-Fries rearrangement; however, data on the rate
†
International Tomography Center.
Physikalisch-Chemisches Institut der Universit a¨ t Z u¨ rich.
Abstract published in AdVance ACS Abstracts, February 15, 1996.
‡
X
0022-3654/96/20100-4448$12.00/0 © 1996 American Chemical Society

1
-Naphthyl Acetate Photo-Fries Rearrangement
J. Phys. Chem., Vol. 100, No. 11, 1996 4449
TABLE 1: Long-Wavelength Maxima in the Absorption
Spectra (νmax), the Absorption Coefficients (Emax), and the
Chemical Shifts of Methyl Protons (δ) in Methanol and the
Melting Points of the Compounds under Study
TABLE 2: Quantum Yields of Products II-IV (æII-æIV) in
the Photolysis of 1-Naphthyl Acetate (Measurement
Accuracy (10%)
solvent
æ
II
æ
III
æ
IV
æ
ν
max
-
,
ꢀ
max
-1
,
acetonitrile
methanol
0.10
0.21
0.05
0.07
0.02
0.14
0.17
0.42
1
-1
compound
cm
M
cm
δ, ppm
Tml, K
I
II
1-naphthyl acetate
2-acetyl-1-naphthol 27180
31930
360
5400
11600
3000
2.44
2.70
2.66
317-318
371-372
470-471
368-369
see ref 28). A sample inserted in a commercial cylindrical Pyrex
ampule was irradiated inside the probehead of the spectrometer
by an excimer laser beam (Lambda Physik EMG 101 MSC
excimer laser, 308 nm, pulse energy up to 100 mJ). The light
was supplied to the sample from the side of the ampule
throughout an optical system consisting of two quartz lenses, a
prism, and a light guide. Time-resolved CIDNP spectra were
III 4-acetyl-1-naphthol 30470
IV 1-naphthol
30820
Experimental Section
Materials. 1-Naphthyl acetate from Chemapol was purified
by sublimation in vacuo. Commercially available 1-naphthol
was recrystallized from a water-ethanol mixture (1:9). 2-Acetyl-
and 4-acetyl-1-naphthol needed for the identification of products
29
detected according to the conventional pulse sequence: satura-
2
5
were synthesized and isolated as described by Stoughton.
tion-laser pulse-delay detection. The incident laser power was
Ethanol and methanol were boiled over NaOH and distilled.
Acetonitrile was repeatedly dried with P2O5 for an hour and
distilled. CD3CN and CD3COD (99% enriched) from Isotope
were used as received. 1,3-Pentadiene (piperylene) from
Aldrich-Chemie, employed as a quencher, was distilled.
Carbazole and acetophenone from Merck, used for sensitization,
were additionally purified. The acetophenone was sublimated
in vacuo. The carbazole was recrystallized from a water-
ethanol mixture (3:7) two times. Phenathrene from Merck was
used as received.
measured by the photodecomposition of dibenzyl ketone in
28,30
benzene.
It has been determined that about 20% of the inlet
laser power reaches the sample.
To obtain the kinetics of CIDNP with the submicrosecond
time resolution, short NMR detection pulses (500 ns, flip angle
2
0°) were used. The detection-pulse width and shape were taken
31
into account by deconvolution procedure.
In the CIDNP
experiments, the optical density of the solutions at 308 nm was
about 0.3 (optical pathway inside the sample tube about 4 mm).
Since the piperylene content of the samples decreased under
bubbling, the real concentration of piperylene was tested by the
NMR spectra taken immediately prior to irradiation.
Steady-State Photolysis. A high-pressure mercury lamp
DRSh-500 was used as a radiation source. The mercury line
at 313 nm was isolated by a combination of UVS-2 and ZhS-3
glass filters. Absorption spectra were detected on a UV-vis
Specord spectrophotometer. On irradiating samples at 160-
Steady-state CIDNP spectra were taken on a Bruker MSL-
3
00 FT-NMR spectrometer by using the optical arrangement
similar to that described above for the flash-CIDNP experiments.
Samples were irradiated for 3 s at a repetition frequency of 7-10
Hz and NMR spectra were detected after the irradiation. CIDNP
effects were determined as a difference between the spectra of
irradiated and nonirradiated samples.
3
00 K, 2-mm-thick quartz cells were inserted into the quartz
tube of a Dewar flask with plane-parallel windows. Temper-
ature was varied with a thermostable nitrogen jet, measured by
a thermodiode, and maintained constant accurate to (0.5 K.
Radiation-induced changes in absorption spectra were detected
at room temperature. The concentration of dissolved oxygen
was varied by passing oxygen or an argon-oxygen mixture
Quantum Chemical Calculations. Experimental data were
interpreted by quantum chemical calculations using the
32
33
34
MNDO, AM1, and PM3 methods based on the modified
(2:3) through solution for 20 min. Oxygen was removed from
35
MNDO-85 program. The conventional Davidon-Fletcher-
Powell procedure was used to optimize the geometry. The
the solution by argon (99.99%) bubbling. Argon was passed
through a piperylene-containing solution at low temperature (243
K) to not reduce the piperylene concentration.
36
geometry of radicals were determined by the restricted Hartree-
37
Fock technique in a “half-electron” approximation.
The
The quantum yields of photolysis products II-IV were
determined spectroscopically using available absorption spectra
and absorption coefficients (Table 1). The photoreduction of
anthraquinone in ethanol (æ ) 0.98)26 was used as an actino-
metric reaction. The laser radiation intensity was measured by
an IMO-2 joulemeter. On the sensitization of the photolysis
of I by phenanthrene, carbazole, and acetophenone, the samples
were irradiated by an LGI-21 nitrogen laser (λ ) 337 nm).
Laser Flash Photolysis. The arrangement for time-resolved
optical studies has been described elsewhere.27 The solutions
38
INDO (UHF) method was applied to calculating the spin
density distribution and the hyperfine interaction constants using
39
the spin Hamiltonian program.
Results
1. Quantum Yields of Reaction Products. The UV
irradiation of 1-naphthyl acetate (I) leads to the formation of
three main reaction products: 2- and 4-acetyl derivatives of
2,4,5
1-naphthol (II, III) and unsubstituted 1-naphthol (IV)
(see
2
passing through the cell (inner dimensions 3 × 10 mm ) were
Scheme 1). Table 1 shows the positions of the absorption
maxima and the absorption coefficients of the products at the
wavelength maxima. These data were used in the spectroscopic
analysis of the irradiated reaction mixture and in determining
the quantum yields of the products.
irradiated by the light pulses of a Lambda Physik LPX 100
excimer laser (308 nm, pulse energy up to 100 mJ). The laser
output was monitored by a Gentec ED-500 joulemeter, and the
initial concentrations of excited species were determined as
27
described earlier. Oxygen was removed from all the solutions
by purging with helium for 1 h prior to use. In the flash
photolysis measurements, the optical density at 308 nm in the
intersection of the laser and monitoring beams (2 mm) was kept
below 0.2. All the data obtained by the flash photolysis
technique and given below refer to a room temperature of 23
It has been established that the quantum yields of the products
formed in acetonitrile are independent of the concentration of
-4
-2
I within 3 × 10 -6 × 10 M. It is seen from Table 2 that
the quantum yields of the products, in particular of 1-naphthol
(IV), in methanol are essentially higher than those in acetonitrile.
In both the solvents the yield of ortho product II is significantly
higher than that of para product III. Attempts have been
(
1 °C.
3
,12
CIDNP. Time-resolved CIDNP measurements were carried
made
to correlate the ratio of the ortho and para isomers
out on a Bruker AM-250 FT-NMR spectrometer (for details
arising in the photo-Fries rearrangement with the spin density

4
450 J. Phys. Chem., Vol. 100, No. 11, 1996
Gritsan et al.
Figure 2. Stern-Volmer plots for product quantum yields vs oxygen
concentration under low-intensity irradiation of a 1-naphthyl acetate
Figure 1. Temperature dependences of the quantum yield of 1-naphthyl
acetate photo-Fries rearrangement (squares) and of the ratio of 2-acetyl
and 4-acetyl-1-naphthol quantum yields (circles).
-3
solution (3 × 10 M). Quantum yield of II in acetonitrile (1) and in
methanol (3), quantum yield of III in acetonitrile (2) and in methanol
(4).
distribution in the aromatic radical. We have performed
quantum chemical calculations of the spin density distribution
in the naphthoxyl radical (see Appendix). The calculations show
that position 4 has the highest spin density; however, the ortho
product arises in a higher yield. Thus, in this case, the yield
ratio of the products is not related to the spin density distribution
but depends on solvent and temperature.
Our measurements show that the ratio of the quantum yields
of products II and III (æII/æIII) changes nonmonotonically with
temperature (Figure 1). In ethanol, for instance, the value of
the ratio first increases from 3 at room temperature to 15 at
2
00 K. With further decrease in temperature the value of the
ratio æII/æIII slightly decreases. The quantum yields of the
products decrease essentially with decreasing temperature. On
the assumption that the 1-naphthyl acetate is not formed in cage
during recombination and neglecting the radical escape from
the cage, we obtain for the total quantum yield of the products
Figure 3. Stern-Volmer plots for the quantum yield of 2-acetyl-1-
æ ) k /(k + kdeac
)
dis
dis
naphthol vs piperylene concentration under laser irradiation of 1-naph-
-
3
thyl acetate (3 × 10 M) in CH
3
OH (triangles), CD
3
OD (circles), and
CN (diamonds).
where kdis is the dissociation rate constant and kdeac is the rate
constant of the deactivation of the excited state. From the above
equation we derive the rate-constant ratio
under Hg lamp irradiation in CH
3
OH (squares), CH
3
products in different ways: via quenching triplet or singlet states
or via reacting with the radicals arising in the photolysis.
Indeed, in the presence of oxygen, the absorption spectrum of
the product mixture changes: a protracted long-wave tail appears
corresponding to oxygen-containing products.
k /k ) æ/(1 - æ)
dis deac
Figure 1 shows the temperature dependence of the quantum
yield. It can be seen that the curve is well described by the
Arrhenius law
Unlike oxygen, 1,3-pentadiene (piperylene) quenches mainly
triplet states. However, the piperylene quenching of the excited
4
0
singlet states of carbonyl compounds and of the sodium salt
1
2
æ/(1 - æ) ) k /k ) (5.8 ( 0.3) ×
of phenylsulfamic acid has been reported. The addition of
piperylene to the solution of I up to a concentration of 0.05 M
does not affect significantly the quantum yield of the products,
while the higher piperylene concentrations decrease the product
yield (Figure 3). This can be described by Stern-Volmer plots
dis deac
exp(-(1900 ( 70)/T)
To study the dependence of the quantum yields on the exciting
light intensity, the solutions were irradiated by the light of the
1
6
-2 -1
-1
mercury lamp (∼1 × 10 quantum cm s ) or by the pulsed
with the slopes 7.8 ( 0.8 M in acetonitrile and 4.0 ( 0.3
24
-2 -1
-1
light of the excimer laser (4.3 × 10 quantum cm s ). The
quantum yields of II and III turned out to be independent of
the light intensity.
M
in methanol. Similar results have been obtained for
methanol and acetonitrile under lamp and laser radiation. No
changes in the ratio of product quantum yields (æII:æIII:æIV) in
the presence of piperylene have been revealed.
2
. Quencher Effect on Quantum Yield. Figure 2 shows
plots for the quantum yields of reaction products æII and æIII
versus oxygen concentration for methanol and acetonitrile
solutions. All the curves are satisfactory linear Stern-Volmer
dependences. Oxygen can affect the quantum yields of the
3. Sensitization of 1-Naphthyl Acetate Decomposition.
2
The triplet state energy of I is estimated as ET ) 251.8 kJ/
mol, which is lower than the corresponding value for naphtha-
4
1
lene (ET ) 254.7 kJ/mol). Phenanthrene (ET ) 258.5 kJ/

1
-Naphthyl Acetate Photo-Fries Rearrangement
J. Phys. Chem., Vol. 100, No. 11, 1996 4451
Figure 5. Transient absorption decay curves observed after excitation
-3
of a solution of 1 × 10 M 1-naphthyl acetate in acetonitrile in the
presence of piperylene (signal A in text). The piperylene concentrations
-
4
-4
-3
are 0 M (upper trace), 2 × 10 M, 5 × 10 M, 1 × 10 M, and 5
-3
×
10 M (lowest trace). The monitoring wavelength was 410 nm.
For calculated curves (solid lines) see text. Insert: the dependence of
the first-order rate constant k on piperylene concentration.
1
Figure 4. Transient absorption spectra observed 3 µs (circles) and 22
radiation power and can be described as a simple exponential
curve. The initial intensity of the signal practically does not
change until the quencher concentration becomes higher than
-
3
µs (squares) after laser pulse in a solution of 1 × 10 M 1-naphthyl
acetate in acetonitrile.
1
mM, whereas the rate of absorption decay increases signifi-
cantly. Such behavior proves that the spectrum shown in Figure
by circles, as well as the spectrum reported by Nakagaki et
4
1
42
mol), carbazole (ET ) 294.1 kJ/mol), and acetophenone
(
ET ) 307.9 kJ/mol),41 whose triplet states are known to be
4
4
2
quenchable by naphthalene, were used as sensitizers. The
sensitized photolysis was carried out under the radiation of the
nitrogen laser at 337 nm where the absorption of I is negligibly
small.
5
al., corresponds to the triplet state of I rather than to the
-naphthoxyl radical. Proceeding from the shape of the
1
5
absorption spectrum, Nakagaki et al. attributed the intermediate
absorption to the 1-naphthoxyl radical. However, we believe
that our spectrum (Figure 4, circles) as well as that reported by
Nakagaki et al. are much closer to the T-T absorption spectrum
of naphthalene.
The quantum yield of II was very small for all the three
sensitizers. The measurements for carbazole and acetophenone
were complicated by photodecomposition of the compounds,
followed by the formation of colored products. In sensitizing
by phenanthrene, the quantum yield of II formed from the triplet
5
4
3-45
This is no wonder since the long-wave
absorption spectrum of I is close to the absorption spectrum of
naphthalene, with the bathochromic shift being no more than
-3
state was estimated as æT e 4 × 10 (the yield of the sensitizer
-1
1
50 cm .
triplets was ΦT ) 0.8).41
At the same time, the intensity and time behavior of the signal
4
. Laser Flash Photolysis of 1-Naphthyl Acetate (I).
B change insignificantly in the presence of piperylene (Figure
). We attributed the weak and slowly decaying signal, whose
behavior only slightly depends on piperylene concentration, to
Figure 4 shows the absorption spectra of intermediates, obtained
during the photolysis of I in acetonitrile 3 µs (circles) and 22
µs (squares) after the laser flash. The first spectrum exhibits
characteristic maxima at 385 and 410 nm. The main peculiari-
ties of this spectrum coincide with those of the spectrum reported
6
4
6
the absorption of 1-naphthoxyl. The top trace (Figure 6),
obtained in the absence of quencher, exhibits the superposition
of the remains of the triplet I signal and the radical absorption.
The other traces correspond to 1-naphthoxyl radical: 10 µs after
the laser flash all triplet molecules should be quenched by
piperylene. The decrease in the radical signal intensity is
insignificant even at high piperylene concentrations.
Without triplet quencher, the triplet signal decays practically
by the pure second-order law. It means that the triplet state of
I is not very reactive, and the main channel of its decay is the
triplet-triplet annihilation. The evolution of the optical density
of the solution D(t) is described by the equation
5
by Nagakura et al. and attributed to the 1-naphthoxyl radical.
It is seen in Figure 4 that the absorption at 410 nm decays faster
than that at 385 nm. Figure 4 shows that more than one
intermediate absorbs in this region. A detailed analysis of the
kinetic traces obtained at the absorption maximum at 410 nm
allows us to conclude that at least two species contribute to the
absorption at this wavelength. Immediately after the laser pulse
the strong absorption signal (A) decays rapidly by the second-
order law (Figure 5, upper trace), which testifies to the high
initial concentration of the transient. In a long time scale the
second, weak and slowly decaying, signal (B) can be distin-
guished; its decay also obeys the second-order law (Figure 6,
upper trace).
D(t) ) ꢀ lT(t) ) ꢀ lT /(1 + k T t)
(1)
T
T
0
tt 0
where ꢀT is the triplet absorption coefficient, l is the optical
path (1 cm), T0 is the initial concentration of triplets, and ktt is
the rate constant of the T-T annihilation. In linear terms
The presence of piperylene even at small concentrations
about 10- M) strongly accelerates the decay of signal A. Figure
4
(
5
presents the kinetic traces obtained at 410 nm at different
concentrations of piperylene. At the piperylene concentration
1
× 10- M the decay rate becomes independent of laser
3
1/D(t) ) 1/D + k t/ꢀ l
(2)
0
tt
T

4
452 J. Phys. Chem., Vol. 100, No. 11, 1996
Gritsan et al.
low reactive triplet state. It is noteworthy that the reported2
quantum yield of intersystem crossing for I (0.29) is consistent
with our estimate.
The values of ktt and ꢀΤ obtained above were applied to the
description of the curves recorded in the presence of piperylene
(Figure 5) as a sum of the pseudo-first-order and second-order
decays:
2
dT/dt ) -k T - k T
(3)
(4)
1
tt
k ) k + k C
q q
1
0
where k0 is the rate constant of triplet state deactivation in the
absence of piperylene and T-T annihilation, kq is the quenching
rate constant, and Cq is the quencher concentration. In this case,
the evolution of the optical density of the T-T absorption D(t)
is determined by the equation
k1t
k1t
Figure 6. Transient absorption decay curves observed after excitation
D(t) ) ꢀ lT(t) ) k ꢀ lT /[k e + k T (e - 1)] (5)
T 1 T 0 1 tt 0
-
3
of a solution of 1 × 10 M 1-naphthyl acetate in acetonitrile in the
presence of piperylene (signal B in text). The piperylene concentrations
The initial triplet concentration T0 and k1 were the fitting
parameters. The results for different piperylene concentrations,
presented in the insert of Figure 5, give the values kq ) (1.9 (
-
4
-4
-3
are 0 M (upper trace), 2 × 10 M, 5 × 10 M, and 5 × 10
lowest trace). The monitoring wavelength was 410 nm. Calculated
curves are the second-order fit [R] ) [R] /(1 + k [R] t) + C, k
) 8 ×
M
(
0
t
0
t
9
-1 -1
5
-1
9
-1 -1
0.4) × 10 M
s
and k0 ) (2.2 ( 0.6) × 10 s for
1
0 M
s ; C is constant.
acetonitrile.
the slope of the straight line is determined by the ratio ktt/ꢀΤ.
The experimentally determined value of ktt/ꢀΤ was independent
of the light intensity varied within a factor of 5 and was equal
An analysis of the absorption decay of the radicals not
involved in the geminate recombination (Figure 6) has shown
that the kinetics obey the second-order law. It is known that
the reaction of radical termination is diffusion-controlled. We
assumed that the rate constants of the bimolecular reactions of
radicals in acetonitrile and methanol are typical for the termina-
6
6
51
to (1.5 ( 0.3) × 10 cm/s in acetonitrile and (1.2 ( 0.3) × 10
cm/s in methanol.
Unfortunately, our experiments do not allow us to determine
the values of ktt and ꢀT separately. Nevertheless, we can
compare the obtained value of ktt/ꢀT with those reported in the
5
1
9
-1 -1
tion rate constants kt ) 8 × 10 M s . Fitting the equation
for the second-order decay to the experimental curves for
different laser powers, we obtained the radical absorption
literature. It has been known that the T-T annihilation is a
diffusion-controlled process;4
4,47,48
however, its rate can be
coefficients ꢀR ) (4.7 ( 1.0) × 10 M cm and ꢀR ) (4.9
3
-1
-1
48,49
3
-1
-1
essentially lower than that of the diffusion
literature value of ktt for naphthalene in ethanol is (1.8 ( 0.3)
(ktt < 2kdif). The
( 1.0) × 10
M
cm for acetonitrile and methanol,
4
4
respectively. The values of initial radical concentrations
extracted from this fitting allow one to estimate the quantum
yields of the radicals not involved in the geminate recombination
for the samples with high piperylene content as æR ) (2.6 (
0.8) × 10 for acetonitrile and æR ) (6 ( 2) × 10 for
methanol. The radical yield in methanol is higher than that in
acetonitrile, which is consistent with our quantum yield values
determined for these two solvents (Table 2).
1
0
-1 -1
50
10
-1
×
10
M
s , whereas 2kdif for ethanol is 2.4 × 10
M
-
1
s . Assuming that the latter relationship holds for I in
acetonitrile, we may estimate the rate constant of the T-T
10
-1 -1
-2
-2
annihilation in acetonitrile: ktt ) (1.5-2.0) × 10
M
s ,
and, consequently, the triplet absorption coefficient ꢀT ) (1.2
4
-1
-1
(
0.6) × 10 M cm .
Since the spectrum we detected is similar to the T-T
absorption spectrum of naphthalene, it may be assumed that
the values of the corresponding maximum absorption coef-
ficients are close. For naphthalene, the maximum absorption
5. CIDNP. Before proceeding to the consideration of the
results obtained by the CIDNP method, we should note the
following. An important feature of the chemical system under
investigation is the formation of strongly absorbing products.
Under intense laser radiation the optical density of the solution
noticeably increases, which leads to a secondary photochemical
reaction and complicates the quantitative interpretation of the
experimental results. It has been established that irradiating
the solution of each reaction product (II, III, and IV) does not
lead to CIDNP. However, these compounds are photoactive
4
5
coefficient at 414 nm, averaged over 16 literature values, is
-
1
-1
2
2 500 ( 5230 M cm . For 1-methylnaphthalene the
extinction coefficient averaged for six literature sources is
-1
somewhat lower than that for naphthalene (≈15 000 M
-1 45
cm ). The value of ꢀT for 1-naphthyl acetate seems to be
4
-1
-1
about 10 M cm . Thus, the above estimation is in good
agreement with the literature data.
4
6
The triplet-state quantum yield can be estimated by the
comparison of the initial triplet concentration, extracted from
the fit of eq 1, and the number of the light quanta absorbed by
the initial compound during one laser pulse (for the last
procedure see ref 27). For both the solvents, acetonitrile and
methanol, the measurements were performed at three different
laser pulse intensities (typical values 53, 21, and 10 mJ) and
yielded similar results: æT ) 0.4 ( 0.2 and æT ) 0.35 ( 0.15
for acetonitrile and methanol, respectively. This can explain
the relatively low yields of the products (Table 2): after the
absorption of a light quantum, a significant part of the excited
molecules undergo an intersystem crossing and transfer to a
and when irradiated can give rise to free radicals.
Thus,
despite the insignificant conversion of an initial compound
(below 5-7%), the experimental conditions at the beginning
and at the end of irradiation could differ. Therefore we did not
perform the quantitative analysis of the amplitude of CIDNP
effects in the homogeneous prosesses.
The intensities of the CIDNP signals in the spectra obtained
in the steady-state experiments depend on laser radiation energy
and pulse repetition rate as well as on the time of irradiation
and the nuclear relaxation times of protons. Therefore the
results obtained by this method could be treated only qualita-
tively. The time-resolved CIDNP technique is free from these

1
-Naphthyl Acetate Photo-Fries Rearrangement
J. Phys. Chem., Vol. 100, No. 11, 1996 4453
Figure 7. NMR 1H spectra obtained before irradiation (upper
spectrum), under steady-state irradiation (middle spectrum), and under
steady-state irradiation in the presence of 0.2 M piperylene (lower
-
3
4
spectrum) of 10 M 1-naphthyl acetate solution in methanol-d .
disadvantages. However, under steady-state irradiation, the
signal-to-noise ratio is much better than that in the time-resolved
CIDNP experiments. It allows one to detect weak signals and
reveal minor reaction pathways in the steady-state experiments.
The CIDNP spectra obtained under steady-state laser irradia-
tion of I in methanol-d4 are shown in Figure 7. A comparison
between the CIDNP spectra and the NMR spectra of initial
compound I and reaction products II-IV allows one to assign
the polarized lines. The most intense emissive signals at 2.70
and 2.66 ppm belong to the methyl protons of II and III,
correspondingly. In the aromatic region one can see three
groups of emissive polarized signals. One of them is a doublet
at 7.30 ppm, which corresponds to overlapping signals of the
fourth protons of II and IV. The second group involves
emissive lines at 6.8 ppm, which are attributed to the partially
overlapping signals of the 2-protons of III (the doublet at 6.85
ppm with a splitting of 8.13 Hz) and the 2-protons of IV (the
doublet at 6.79 ppm with a splitting of 7.15 Hz). The third
group consists of weak emissive lines at 7.72-7.76 ppm, which
are attributed to the 5-protons of IV and to the 3-protons of II,
the positions of these lines coinciding. The emissive line at
Figure 8. CIDNP spectra obtained 2 µs (upper spectrum) and 1000
-
3
µs (lower spectrum) after the laser flash in a solution of 10
M
4
1-naphthyl acetate in methanol-d .
formation of a negative polarization of the in-cage products for
the protons of the radicals with the lower g factor and a positive
hfi constant in the case of triplet precursors. The emissive
polarization of the acetyl groups of both II and III was detected.
The hfi constants of the 2- and 4-protons of 1-naphthoxyl are
negative (see Appendix) and correspond to the radical with the
higher g factor. Thus, the emissive polarization of these protons
in products II-IV is consistent with the assumption about the
triplet precursor.
At a high piperylene concentration (0.2 M), the polarization
sign changes from emission to absorption for the acetyl protons
of both II and III (Figure 7), with the protons of I remaining
unpolarized. At this concentration of piperylene, the signals
in the aromatic region are too weak to be detected. The
alteration of the CIDNP sign at high concentrations of piperylene
indicates that I decomposes with the formation of radical pair
through the singlet state as well.
Figure 8 shows the time-resolved CIDNP spectra obtained 2
µs (top spectrum) and 1000 µs (bottom spectrum) after the laser
pulse. As in the steady-state experiments described above, the
acetyl protons of II and III exhibit the most intense polarization.
The kinetic analysis was performed only for these spectral lines.
Figure 9 shows the kinetics of the net nuclear polarization of
the protons of II (triangles) and III (squares), obtained during
the photolysis of I in acetonitrile. Both the curves are
characterized by an initial growth of the polarization. On
passing the maximum Im, the polarization decays in the
microsecond time scale to the stationary value I∞. The ratio of
Im and I∞ decreases with decreasing laser intensity, and the
decaying section of the curve flattens. The CIDNP signals of
II, obtained in acetonitrile at different laser intensities, show
maxima 1.5 µs after the laser pulse; for III the kinetics is a
maximum 5 µs after the flash. The kinetic curves for the nuclear
2
.08 ppm probably belongs to the ketene CH2CO which is
5
formed in the photolysis of I. The chemical shifts of the
compounds under study in acetonitrile are slightly different from
those given above; however, the corresponding CIDNP spectrum
exhibits signals of the same products. Only in the steady-state
CIDNP experiments did we succeed to detect the absorption
signals of the minor product acetaldehyde (aldehyde proton, 9.89
ppm).
The addition of the triplet quencher piperylene (>10 mM)
decreases the intensity of the CIDNP signals. The strong
dependence of CIDNP on the presence of piperylene clearly
indicates that the nuclear polarization appears in the radical pairs
formed from triplet precursors. The same conclusion follows
from the analysis of the polarization signs. The g factor of the
52
acetyl radical (2.0005) is lower than that of naphthoxyl radical
5
3
(
2.00431), and the hfi constant of acetyl protons is positive.54
From time-resolved CIDNP experiments (see below) we know
that the emission of the methyl protons of II and III is formed
in cage. According to Kaptein’s rules,55 this leads to the

4
454 J. Phys. Chem., Vol. 100, No. 11, 1996
Gritsan et al.
may be disregarded. The intensity of the geminate nuclear
polarization is proportional to the number of the radical pairs
(RP) resulting from the decomposition of the triplet states. The
rate of RP formation is described by the equation
d[RP]/dt ) k T(t)
(6)
r
Hence, the intensity P of geminate polarization is defined
by the following integral:
∞
P ) ê 0
∫
T(t) dt
(7)
where ê is the polarization per one triplet-born radical pair (RP),
and kr is the rate constant of RP formation. In the general case,
the time evolution of triplet states is described according to eq
5
with replacement of k1 by kr. Integration of (7) yields the
expression for the CIDNP amplitude as a function of quencher
concentration:
kr
k T
tt 0
P(C ) ) ê ln 1 +
(8)
q
(
)
ktt
k + k C
r q q
Figure 9. CIDNP kinetics (in arbitrary units) obtained during the
photolysis of 1-naphthyl acetate in acetonitrile-d for methyl protons
3
All constants in eq 8 have been specified above. The
dependence of the CIDNP signal intensity on quencher con-
centration is determined as follows:
of 2-acetyl-1-naphthol (triangles) and 4-acetyl-1-naphthol (squares).
Solid lines is biexponential fit.
k T
tt
0
P(0)/P(C ) ) ln(1 + k T /k )/ln 1 +
(9)
q
tt
0
r
(
)
k + k C
q q
r
However, in our experiments (Figure 10), the dependence of
CIDNP intensity on quencher concentration is linear. This
corresponds to the condition kttT0/kr , 1. In this case the eq 9
takes the form of the Stern-Volmer equation:
P(0)/P(C ) ) 1 + k C /k
r
(10)
q
q
q
Using the quenching rate constants kq ) 1.9 × 10 M s-1 for
9
-1
9
-1 -1
acetonitrile and kq ) 4 × 10 M
s
for methanol, determined
in the flash photolysis experiments, we obtain from eq 10 the
values kr ) 3.3 × 10 s for acetonitrile and kr ) 5.2 × 10
for methanol.
7
-1
6
-1
s
Discussion
Figure 10. Stern-Volmer plots for the CIDNP intensity of methyl
As one can see from Figures 3, 5, and 10, piperylene as a
quencher affects the product quantum yields, the decay of
triplets, and the CIDNP formation at different concentrations.
Only very high piperylene concentrations (g0.1 M) affect the
quantum yields of products; it means that the reaction proceeds
from the extremely short-lived triplet state or, more likely, the
product is formed directly from the first excited singlet state.
In this case, the effect of piperylene can be regarded as the
quenching of the excited singlet states that may occur at high
protons of 2-acetyl-1-naphthol vs piperylene concentration obtained
4
during the photolysis of I in methanol-d (triangles) and in acetonitrile-
d
4
(squares).
polarization in methanol are similar to those corresponding to
acetonitrile, but the polarization decay is less pronounced and
the stationary value of CIDNP differs from its maximum value
by no more than 25%. The solid line in Figure 9 is the result
of biexponential fitting for the evolution of CIDNP signals. It
is seen that the experimental and calculated curves are in good
agreement, with the characteristic rise times being τ2 ) 0.9 (
1
2,40
quencher concentrations.
Indeed, at sufficiently high
piperylene concentrations, when most of the triplet molecules
are quenched, we observe the alternation of the CIDNP sign as
0
.1 µs and τ3 ) 2.2 ( 0.2 µs for II and III, respectively.
In the presence of piperylene, the intensities of the CIDNP
5
a manifestation of the singlet-born polarization. Earlier the
signals of acetyl protons of II and III decay drastically. Figure
conclusion that the products are formed in the singlet RP was
drawn from the value of magnetic effect.
1
0 shows the Stern-Volmer plots for CIDNP intensity at the
kinetic curve maximum as a function of piperylene concentration
On the other hand, the strong dependence of CIDNP effects
on the presence of piperylene in much lower concentrations
(g0.002 M) and the signs of CIDNP signals clearly indicate
that the nuclear polarization is created mainly in the radical pairs
formed from triplet precursors. The change in the quantum yield
of the reaction at such piperylene concentrations is negligibly
small. Thus, we can conclude that both the singlet and triplet
excited states of I participate in the formation of the initial
for methanol and acetonitrile. The slopes of the curves are kQ
1
-1
)
302 ( 10 M- and kQ ) 75 ( 5 M for methanol and
acetonitrile, respectively.
Consider now the formation of the triplet geminate CIDNP
in the presence of a triplet quencher. By the time τ (1.5 µs)
the geminate recombination is almost completed and the nuclear
polarization arising in the homogeneous processes in the bulk

1
-Naphthyl Acetate Photo-Fries Rearrangement
J. Phys. Chem., Vol. 100, No. 11, 1996 4455
radical pairs. The triplet pathway of the reaction is of secondary
importance. From the accuracy of the determination of quantum
yields, one can estimate that the contribution of the triplet
channel to the product formation is below 10%. However, it
is this channel that is responsible for CIDNP formation. This
is not surprising since usually the polarization formed through
the singlet channel is weaker than that formed via the triplet
precursor.
probably the nπ* triplet state. Estimations (see Appendix) show
that in energy the nπ* triplet state is to be between the singlet
and triplet ππ* states. Moreover, one may expect that T2 can
be quenched by piperylene with a rate constant kq close to the
1
0
-1 -1
diffusion one (about 1.5 × 10
M
s
in acetonitrile and
5
0
methanol). Using this value of kq, we estimate the lifetime
of the reactive triplet state T2 as 3 and 6 ns for acetonitrile and
methanol, respectively. Further studies which are to validate
or invalidate this hypothesis are in progress.
The absorption spectrum of the triplet state of I was obtained
in the flash photolysis experiments. The validity of our
attribution is strongly supported by the fact that in the presence
Let us discuss the nature of the maximum observed in the
CIDNP kinetics (Figure 9). Usually the presence of a maximum
in CIDNP kinetics is associated with the fact that the same
products formed in the geminate cage and in the bulk have
-4
-3
of 10 -10 M of piperylene the signal decay rate increases
essentially, with the initial intensity of the signal being
unchanged. At such concentrations, piperylene does not affect
the quantum yield of the products. Our attribution of the
intermediate absorption is consistent with the similarity of the
spectrum of this intermediate to the T-T absorption spectrum
of naphthalene. In the flash photolysis experiments, the main
channel of the decay of the triplet states of I in the absence of
quencher is the T-T annihilation. Very low concentrations of
5
7
polarizations of opposite signs. The nuclear polarizations of
the geminate products and the radicals not involved in the in-
cage recombination are the same in intensity and opposite in
sign. If the nuclear paramagnetic relaxation time is long enough,
the polarization carried out of the initial cage by radicals is
transferred to the diamagnetic products in the course of the
radical termination. This competition between geminate and
5
7
-
4
homogeneous processes has been studied before. According
piperylene (about 10 M) appear to be sufficient to affect
noticeably the decay rate of the triplet states. Thus, piperylene
quenches the triplet state of I through T-T energy transfer,
and the quenching process is expected to be diffusion-
5
7
to theoretical calculations, the maximum of a CIDNP kinetic
-1
curve should be positioned at t1 ) (2kt[R]0) , and the intensity
of the CIDNP signal at the maximum should not be more than
50% higher than the value of geminate polarization. In our
case, the time behavior of the nuclear polarization differs from
4
1,50
controlled.
The rate constant of the piperylene quenching
of the triplet state of I, measured in the flash photolysis
5
7
9
-1 -1
9
that predicted by Vollenweider and Fischer: the intensity of
CIDNP at zero point is negligibly small. Thus, we can conclude
that the formation of the geminate polarization is protracted in
time. This protraction can arise from the following. First, the
absence of the signal at the initial time can be due to the nonzero
lifetime of the reactive triplet state. According to the above
estimations, the lifetime of the reactive triplet state is several
nanoseconds. Such a short lifetime cannot be responsible for
the position of the maximum of CIDNP kinetics.
experiments, kq (1.9 × 10 M
s
in acetonitrile and 4 × 10
-1
-1
M
s
in methanol) is approximately an order of magnitude
lower than the diffusion rate constant. However, this is no
wonder since in this case the T1 states of the donor (251.8 kJ/
mol) and acceptor (238 and 245 kJ/mol for cis- and trans-
piperylene, respectively)56 are similar. Note that the value of
kq for 1-naphthol (T1 energy 250 kJ/mol) is even smaller and
8
-1 -1
46
equals 4.4 × 10 M
s
in an ether/hexane mixture.
The lifetime of T1 of 1-naphthyl acetate (in the absence of
piperylene and T-T annihilation) is rather long: 4.5 ( 0.7 µs
in acetonitrile solution (actually, this result, extracted from
piperylene quenching of triplet states, could be affected by
remains of oxygen in solution; the real lifetime is probably even
longer). This value is inconsistent with the estimated lifetime
of the triplet state responsible for the formation of CIDNP. Using
the value of kq measured for quenching of T1, one can estimate
the lifetime of the triplets responsible for the formation of RP
and polarization as 30 ns for acetonitrile and 190 ns for
methanol. It is seen that these lifetimes are 2 orders of
magnitude shorter than those measured by flash photolysis. The
qualitative difference is that in the flash photolysis experiments
in the absence of quencher the T-T annihilation is the main
decay channel of triplet molecules and the addition of 1 mM of
piperylene decreases essentially the triplet lifetime (Figure 5).
In the CIDNP experiments, where the exciting light intensity
is the same, the T-T annihilation is no longer the main channel
for the decay of the reactive triplet states and the addition of 1
mM of piperylene practically has no appreciable influence on
CIDNP intensity.
The kinetics of the increasing geminate CIDNP can be
accounted for by another hypothesis. We assume that the
detected kinetics depends on the process of the formation of
compounds II and III from intermediate unstable adducts V
and VI, respectively. The estimation of the rate constants of
these processes as the rate of signal increase in flash-CIDNP
6
experiments (Figure 9) gives the values k2 ) (1.1 ( 0.1) × 10
-1
5 -1
s
and k3 ) (4.5 ( 0.5) × 10 s for acetonitrile. It has been
24
known that the lifetimes of cyclohexadienones, which are
intermediates of the photo-Fries rearrangements of phenyl
acetate, increase essentially in polar and particularly in protic
solvents. In alcohols (EtOH, MeOH), the rate constants for 1,3-
It would be reasonable to assume that the different methods
detect two different triplet states. A similar hypothesis about
two reactive triplet states has been proposed for the photo-Fries
rearrangement of phenyl sulfamates.12 In the experiments on
the flash photolysis of I we detect the spectrum of the lower
5
-1
and 1,5-hydrogen shifts reach values of about 10 s , which
are close to the results of our measurements.
The results obtained allow us to propose the mechanism for
the photo-Fries rearrangement of 1-naphthyl acetate (Scheme
2). The 1-naphthyl acetate (I) molecule that has absorbed a
(T1) nonreactive ππ* state, whose properties are similar to those
of the triplet state of naphthalene. In the photo-Fries rearrange-
ment, the reactive state is the upper short-lived triplet state (T2)
which decomposes to yield a radical pair of acyl and naphthoxyl
radicals, wherein the CIDNP effects arise. The T2 state is
1
photon passes into the excited singlet state I*. After that the
major portion of the excited singlet states (according to Table
2, from 0.17 ( 0.02 in acetonitrile to 0.42 ( 0.04 in methanol)

4
456 J. Phys. Chem., Vol. 100, No. 11, 1996
Gritsan et al.
SCHEME 2
Figure 11. Computer-generated drawing of the 1-naphthyl acetate
structure.
decays to the radical pairs of acetyl and naphthoxyl radicals.
The remaining excited molecules undergo the intersystem
3
3
of 1-naphthyl acetate, which is a typical photo-Fries rearrange-
ment reaction. The analysis of the effect of piperylene on the
quantum yield of reaction products and on the signs of
polarization observed in CIDNP spectra has validated the
assumption that the decomposition of 1-naphthyl acetate to
acetyl and naphthoxyl radicals through the singlet state is the
main channel of the reaction (not less than 90%). The analysis
of the CIDNP spectra shows that recombining radical pairs do
not form initial 1-naphthyl acetate.
Although the reaction goes mainly through the singlet
channel, the major contribution to CIDNP is made by the radical
pairs resulting from the decomposition of the upper triplet states
of 1-naphthyl acetate. Analyzing the kinetics of the polarization
of photo-Fries rearrangement products (2-acyl- and 4-acyl-1-
naphthol) we have established that the products are formed in
the geminate cage as well as in the bulk. The unusual CIDNP
kinetics, with the polarization absent at short times, indicates
that the formation of geminate polarization is protracted in time.
The most likely limiting stage of CIDNP formation in the
products is the isomerization of intermediates. The absence of
polarization of the 1-naphthyl acetate protons is indicative of
the fact that the recombination of the naphthoxyl and acyl
radicals to the initial compound is insignificant and, hence, is
not responsible for the essential difference of the quantum yield
of the photo-Fries rearrangement from unity.
crossing and transfer into the triplet states T2 ( I**) and T1 ( I*).
Most of the singlet-born radicals recombine in the initial cage
to form V and VI, which then are rearranged into II and III.
The rest radicals escape and decay in bulk reactions.
3
The arising triplets I* (T1) are low-reactive. The quantum
yields of the triplet states arising under laser flash radiation are
0
.4 ( 0.2 and 0.35 ( 0.17 for acetonitrile and methanol,
respectively. The T-T annihilation is the main channel of their
decay. The fact that the yield of the products under laser
irradiation coincides with that under lamp irradiation as well
as the extremely low quantum yield of the products in the
3
sensitized photolysis indicates that the triplet states I* mainly
decay without reaction even at the lower light intensities. The
results obtained by the CIDNP method suggest that at least some
3
I** (T2) react to form radical pairs. The CIDNP created in
such pairs is higher than the nuclear polarization formed in the
singlet-born pairs.
It has been concluded that the total contribution of the triplet
channel to photo-Fries rearrangement is small (<10%). This
can be accounted for by the fact that the higher reactive triplet
3
state I** (T2) converts fast into the nonreactive triplet state
3
I* (T1) in the internal conversion, and this channel competes
3
with the I** decay to radicals. The nonreactive nature of the
triplet state T1 is unlikely to be determined by its orbital nature.
2
2
According to the recent theoretical work of Grimme, the
primary process in the photo-Fries rearrangement is treated as
the πσ*-induced photodissociation. The C-O bond fission can
start from both the nπ* and ππ* states. The photodissociation
In the flash photolysis experiments we have detected the
absorption spectra of 1-naphthyl acetate triplet and naphthoxy
radicals; their absorption coefficients have been estimated. It
has been established that the quantum yield of the triplet state
of 1-naphthyl acetate is sufficiently high and the main channel
of triplet decay is the triplet-triplet annihilation. This is
consistent with the fact that the quantum yield of the photo-
Fries rearrangement of 1-naphthyl acetate is much less than unity
(0.17 in acetonitrile and 0.42 in methanol).
process is activated, with the barriers formed due to the crossing
of nπ* or ππ* with repulsive πσ* potential curves.22 It has
been estimated (see Appendix) that the C-O bond dissociation
occurring through the excited singlet and second triplet states
is an essentially exothermic process. In the lower ππ* triplet
3
state ( I*) the enthalpy of the reaction is small. Therefore, in
Proceeding from the different effects of piperylene on the
decay kinetics of T-T absorption and on the yield of CIDNP
via the triplet channel, we propose that the upper triplet state
participates in the photo-Fries rearrangement of 1-naphthyl
acetate. Further investigations are carried out in order to validate
or invalidate this assumption.
this case, the activation barrier of the dissociation should be
very high and, hence, the dissociation should be low-efficient.
The essentially higher activation barrier of the photo-Fries
rearrangement of 1-naphthyl acetate with respect to the corre-
sponding value for phenyl acetate appears to be associated with
the lower energy of the excited singlet state of 1-naphthyl acetate
3
(
383.4 kJ/mol, see Appendix) as compared to the corresponding
22
Appendix. Quantum Chemical Calculations
energy value for phenyl acetate (421.5 kJ/mol).
Figure 11 shows the results of the AM1 calculation of
-naphthyl acetate geometry. It can be seen that the plane of
Conclusions
1
In this work, we have employed a number of experimental
methods for a detailed investigation of the phototransformation
the acyl group (COC atoms) is practically perpendicular (θ )
84°) to the plane of naphthalene rings. The rotation barrier of

1
-Naphthyl Acetate Photo-Fries Rearrangement
J. Phys. Chem., Vol. 100, No. 11, 1996 4457
TABLE 3: Calculated and Experimental Values of Maxima
in the Absorption Spectra of 1-Naphthyl Acetate (λmax),
Experimental Values of the Absorption Coefficients at the
Maxima (E), and Calculated Values of Oscillator Strengths
It is seen from Table 4 that the calculated values of negative
H hfi constants are in good agreement with the corresponding
experimental values, with the positive values of the hfi constants
being overestimated. Both theoretical and experimental data
suggest that the hfi constant with hydrogen in position 4 is higher
than that in position 2. This is consistent with the data of Table
1
(f)
expt λmax (log ꢀ)
calc λmax (f)
3
2
13 (2.60)
80 (3.80)
304 (0.011)
274 (0.100)
5
, which show that the spin density on carbon in position 4 is
higher than that in position 2.
TABLE 4: INDO (UHF) Calculations and Experimental
Values for H hfi Constants (a, mT) in 1-Naphthoxyl Radical
H2 H3 H4 H5 H6 H7 H8
1
Acknowledgment. This work has been supported by Council
of High Education of Russia (project 94-9.5-111) and by Russian
Foundation for Fundamental Research (project 93-03-18593 and
project 95-03-08920).
calc
-0.89 0.53 -0.95 -0.38 0.28 -0.35 0.32
5
9
exp (abs val) 0.825 0.175 1.075 0.25 0.065 0.25 <0.04
References and Notes
TABLE 5: INDO (UHF) Calculation for Spin Density
Distribution in 1-Naphthoxyl
(
(
(
1) Anderson, J. C.; Reese, C. B. Proc. Chem. Soc. 1960, 217.
2) Bellus, D. AdV. Photochem. 1971, 8, 109 and references therein.
3) Shizuka, H.; Morita, T.; Mori, Y.; Tanaka, I. Bull. Chem. Soc. Jpn.
C2
.26
C3
C4
C5
C6
C7
C8
O
0
-0.07
0.28
-0.05
0.07
-0.04
0.07
0.53
1
969, 42, 1831.
4) Ohto, Y.; Shizuka, H.; Sekiguchi, S.; Matsui, K. Bull. Chem. Soc.
(
Jpn. 1974, 47, 1209. Crouse, D. J.; Hurlbut, S. L.; Wheeler, D. M. S. J.
Org. Chem. 1981, 46, 374. Evreinov, V. I.; Pivovarov, A. P. Russ. Zh. Fiz.
Khim. 1980, 55, 1901.
(5) Nakagaki, R.; Hiramatsu, M.; Watanabe, T.; Tanimoto, Y.; Na-
gakura, S. J. Phys. Chem. 1985, 89, 3222.
the acyl group in 1-naphthyl acetate is 9.6 kJ/mol. The MNDO
and PM3 calculations also give large angles of rotation of the
acyl group with respect to the ring plane (80° and 70°,
respectively). The AM1 calculation of the heat of C-O bond
dissociation gives 227 kJ/mol (compare with 186 kJ/mol for
PM3 and 212 kJ/mol for MNDO).
(
(
(
(
6) Elad, D. Tetrahedron Lett. 1963, 873.
7) Shizuka, H.; Tanaka, T. Bull. Chem. Soc. Jpn. 1968, 41, 2343.
8) Shizuka, H. Bull. Chem. Soc. Jpn. 1969, 42, 52.
9) Nozaki, H.; Okada, R.; Noyori, R.; Kawanisi, M. Tetrahedron 1966,
On the basis of the similarity of the 1-naphthyl acetate and
naphthalene spectra and the known positions of the lower triplet
2
2, 2177.
(
(
10) Miranda, P. M.; Factor, A. J. Polym. Sci. Part A 1989, 27, 4427.
11) Herweh, J. E.; Hoyle, C. E. J. Org. Chem. 1980, 45, 2195.
41
and singlet states of naphthalene, one can estimate the positions
of the lower singlet and triplet ππ* states of 1-naphthyl acetate,
which are 383 and 252 kJ/mol, respectively. From these data
and the heat of the bond dissociation we obtain the heat of the
(12) Lally, J. M.; Spillane, W. J. J. Chem. Soc., Chem. Commun. 1987,
-9. Lally, J. M.; Spillane, W. J. J. Chem. Soc., Perkin. Trans. 2 1991,
03-807.
8
8
(13) Subramanian, P.; Greed, D.; Griffin, A. C.; Hoyle, C. E.; Venka-
S
T
dissociation ∆H ) -156 kJ/mol in the singlet state and ∆H
-24.7 kJ/mol in the lowest triplet state.
It follows from our experimental data that the lowest triplet
taram, K. J. Photochem. Photobiol. A: Chem. 1991, 61, 317-327.
(14) Kobsa, H. J. Org. Chem. 1962, 27, 2293.
)
(
15) Anderson, J. C.; Reese, C. B. J. Chem. Soc. 1963, 1781.
(16) Adam, W.; de Sanabia, J. A.; Fischer, H. J. Org. Chem. 1972, 38,
state is nonreactive in the photo-Fries rearrangement. The small
contribution of the triplet channel to the photorearrangement
could be explained by the participation of the nπ* triplet state.
The results of the CNDO/S calculation58 of the absorption
spectrum of 1-naphthyl acetate are listed in Table 3.
2571.
(17) Vollenweider, J.-K.; Fischer, H. Chem. Phys. 1988, 124, 333.
Vollenweider, J.-K. Dissertation, Physikalisch-Chemisches Institut der
Universit a¨ t Z u¨ rich, 1987.
(18) Pinhey, J. T.; Schaffner, K. Aust. J. Chem. 1968, 21, 2265.
(19) Pohlers, G. P.; Grimme, S.; Dreeskamp, H. J. Photochem. Photobiol.
A: Chem. 1994, 79, 153.
The calculated spectrum fits well the experimental one (ππ*
states, Table 3). For the nπ* state the calculation gives λmax )
(
20) Schaffner, K. Pure Appl. Chem. 1967, 16, 75.
(21) Palm, W.-U.; Dreeskamp, H.; Bouas-Laurent, H.; Castellan, A. Ber.
2
99 nm. Hence, the ππ* singlet and triplet states are the lower
Bunsen-Ges. Physik. Chem. 1992, 96, 50.
22) Grimme, S. Chem. Phys. 1992, 163, 313.
(23) Kalmus, C. E.; Hercules, D. M. J. Am. Chem. Soc. 1974, 96, 449.
(
excited states of 1-naphthyl acetate. Assuming that the singlet-
4
1,42
3
-1
triplet splitting for the nπ* states3
is (2-4) × 10 cm , one
(
(
24) Arai., T.; Tobita, S.; Shizuka, H. Chem. Phys. Lett. 1994, 223, 521.
25) Stoughton, R. W. J. Am. Chem. Soc. 1935, 57, 202.
can estimate the position of the nπ* state, which is 327-352
kJ/mol. For the heat of the dissociation in this triplet state we
have ∆H ) -(100-126) kJ/mol.
(26) Carlson, S. A.; Hercules, D. M. Photochem. Photobiol. 1973, 17,
123.
27) Tsentalovich, Yu. P.; Fischer, H. J. Chem. Soc., Perkin Trans. 2
T
(
The reaction of radical recombination, yielding intermediate
adducts, is naturally an exothermic one. If adduct V is formed,
the AM1-calculated enthalpy of the reaction is -199 kJ/mol,
while the enthalpy of the formation of adduct VI is -191 kJ/
mol. The reactions of adducts V and IV to the final products
1
994, 729.
(28) Tsentalovich, Yu. P.; Yurkovskaya, A. V.; Sagdeev, R. Z.;
Obynochny, A. A.; Purtov, P. A.; Shargorodsky, A. A. Chem. Phys. 1989,
39, 307. Yurkovskaya, A. V.; Tsentalovich, Yu. P.; Lukzen, N. N.;
1
Sagdeev, R. Z. Res. Chem. Intermediates 1992, 17, 145. Tsentalovich, Yu.
P.; Yurkovskaya, A. V.; Sagdeev, R. Z. J. Photochem. Photobiol. A: Chem.
1993, 70, 9.
(
II and III) are also exothermic (∆H ) -53 and -41 kJ/mol,
(
29) Miller, R. J.; Closs, G. L. ReV. Sci. Instrum. 1981, 52, 1876.
respectively). However, the isomerization rate constants in-
(30) Engel, P. S. J. Am. Chem. Soc. 1970, 92, 6074. Robbins, W. K.;
crease essentially (by 5-6 orders of magnitude in proton-donor
Eastman, R. H. J. Am. Chem. Soc. 1970, 92, 6074.
solutions, e.g., in alcohols), which indicates that solvent protons
(31) Goez, M. Chem. Phys. Lett. 1990, 165, 11.
(32) Dewar, M. J. S.; Thiel, W. J. Am. Chem. Soc. 1977, 99, 4899, 4907.
are involved in the reaction.24
(33) Dewar, M. J. S.; Zoebich, E. G.; Healy, E. F.; Stewart, J. J. P. J.
1
The spin density distributions and H hfi constants for the
Am. Chem. Soc. 1985, 107, 3902.
1
-naphthoxyl radical have been calculated (Tables 4 and 5).
(34) Stewart, J. J. P. J. Comput. Chem. 1989, 10, 209, 221.
(
35) Bliznyuk, A. A.; Voityuk, A. A. Zh. Struct. Khim. 1986, 27, N4,
90.
36) Fletcher, R.; Powell, M. J. D. Comput. J. 1963, 6, 163. Davidon,
W. C. Comput. J. 1968, 10, 406.
1
(
(
37) Voityuk, A. A. Zh. Struct. Khim. 1983, 24, N3, 18.
(38) Pople, J. A.; Beverige, D. L.; Dobosh, R. A. J. Chem. Phys. 1967,
4
7, 2026.

4
458 J. Phys. Chem., Vol. 100, No. 11, 1996
Gritsan et al.
(
39) Plahutin, B. N.; Zhydomirov, G. M. Theor. Experim. Khim. 1988,
4, 149.
40) Turro, N. J. Modern Molecular Photochemistry; Benjamin Cam-
mings Publishing: Melto Park, 1978; p 476.
41) Birks, J. B. Photophysics of Aromatic Molecules; Wiley-Inter-
science: London, 1970; pp 193-297.
42) Ermolaev, V. L.; Bodunov, E. N.; Sveshnikova, E. B.; Shakhverdov,
(50) Saltiel, J.; Shannon, P. T.; Zafiriou, O. C.; Uriarte, A. K. J. Am.
Chem. Soc. 1980, 102, 6799.
(51) Schuh, H.; Fischer, H. HelV. Chem. Acta 1978, 61, 2130. Lehni,
H.; Fischer, H. Int. J. Chem. Kinet. 1983, 15, 733. Tsentalovich, Yu. P.;
Fischer, H. J. Chem. Soc., Perkin Trans. 2 1994, 729.
2
(
(
(52) Benett, J. E.; Mile, B. Trans. Faraday Soc. 1971, 67, 1587.
(
53) Adams, M.; Blois, M. S., Jr.; Sands, R. H. J. Chem. Phys. 1958,
(
2
8, 774.
T. A. Electronic Excitation Energy Radiationless Transfer; Nauka: Len-
ingrad, 1977; pp 94-105.
(
(
54) Paul, H.; Fischer, H. HelV. Chem. Acta 1973, 56, 1575.
55) Kaptein, R. Chem. Commun. 1971, 732. Salikhov, K. M.; Molin,
(43) Labhart, H.; Heizelmann, W. Triplet-Triplet Absorption Spectra
Yu. N.; Sagdeev, R. Z.; Buchachenko, A. L. Spin Polarization and Magnetic
Effects in Radical Reactions; Molin, Yu. N., Ed.; Elsevier: Amsterdam,
of Organic Molecules. In Organic Molecular Photophysics; Birks, J. B.,
Ed.; Wiley: New York, 1973; Vol. 1, pp 297-355.
(
1
984.
56) McGlynn, S. P.; Azumi, T.; Kinoshita, M. Molecular Spectroscopy
of the Triplet State; Prentice Hall: Englewood Cliffs, NJ, 1969.
44) Dempster, D. N.; Morrow, T.; Quinn, M. F. J. Photochem. 1973/
(
7
4, 2, 329.
(
(
(
(
45) Carmichael, I.; Hug, G. L. J. Phys. Chem. Ref. Data 1986, 15, 1.
(57) Vollenweider, J.-K.; Fischer, H. Chem. Phys. 1986, 108, 365.
(58) del Bene, J.; Jaffer, H. H. J. Chem. Phys. 1968, 48, 1807, 4050.
(59) Fischer, H., Ed. Magnetic Properties of Free Radicals; Landolt-
46) Hara, K.; Baba, H. J. Chem. Soc., Faraday Trans. 2 1975, 71, 1100.
47) Wyrsch, D.; Labhart, H. Chem. Phys. Lett. 1971, 12, 373.
48) Saltiel, J.; Marchand, G. R.; Smothers, W. K.; Stout, S. A.; Charlton,
B o¨ rnstein New Series, Vol. II 9c; Springer: Berlin, 1980.
J. L. J. Am. Chem. Soc. 1981, 103, 715.
49) Burri, J.; Fischer, H. Chem. Phys. 1991, 151, 279.
(
JP952315K