This journal is c the Owner Societies 2011 Phys. Chem. Chem. Phys., 2011, 13, 755–761 755

Photochemical primary process of photo-Fries rearrangement reaction of

1-naphthyl acetate as studied by MFE probe

Masao Gohdo,a Tadashi Takamasub and Masanobu Wakasa*a

Received 31st March 2010, Accepted 30th September 2010

DOI: 10.1039/c0cp00077a

Photo-Fries rearrangement reactions of 1-naphthyl acetate (NA) in n-hexane and in cyclohexane

were studied by the magnetic field effect probe (MFE probe) under magnetic fields (B) of 0 to 7 T.

Transient absorptions of the 1-naphthoxyl radical, T–T absorption of NA, and a short-lifetime

intermediate (t = 24 ns) were observed by a nanosecond laser flash photolysis technique.

In n-hexane, the yield of escaped 1-naphthoxyl radicals dropped dramatically upon application of

a 3 mT field, but then the yield increased with increasing B for 3 mT o B r 7 T. These observed

MFEs can be explained by the hyperfine coupling and the Dg mechanisms through the singlet

radical pair. The fact that MFEs were observed for the present photo-Fries rearrangement

reaction indicates the presence of a singlet radical pair intermediate with a lifetime as long as

several tens of nanoseconds.

1. Introduction

Magnetic field effects (MFEs) on photochemical reactions

through radical pairs (RPs) and biradicals have received

considerable attention during the past three decades, and the

mechanism of MFEs has been well clarified experimentally

and theoretically.1–3 In RPs generated by photochemical

reactions, the unpaired electron spins on each radical are

coupled, giving two different spin states: singlet (S) and triplet

(T). According to the Pauli principle, singlet RPs can react to

form a recombination product, whereas triplet RPs cannot

react with each other but instead form escaped radicals.

Magnetic fields interact with these spins and affect the reaction

of the RPs without changing other parameters such as reaction

rate of singlet RPs, activation barrier, and diffusion motion of

the radicals. Because the interaction between magnetic fields

and spins can be defined by quantum chemistry, MFE studies

on RPs provide valuable information on their kinetics and

dynamics and, in particular, on aspects of the reaction

mechanism such as the presence of precursors and intermediates.

Therefore, we refer to an MFE study on RPs as a magnetic

field effect probe (MFE probe).4 Recently, using the MFE

probe, we have reported the microviscosity of alcoholic

solutions5 and the nanoscale heterogeneous structure of ionic

liquids.6–8

Photo-Fries rearrangement reactions are believed to occur

via geminate RPs formed by C–O bond cleavage in the ps*state of the aryl esters that form the key intermediate of

cyclohexadienone.9–13 The reaction kinetics and intermediates

of photo-Fries reactions have been continuously studied by

means of product analysis, chemically induced dynamic nuclear

polarization (CIDNP), laser flash photolysis, and magnetic

isotope effects (MIEs).14–19 Nakagaki et al. reported that

MIEs were observed for a photo-Fries reaction of 13C-labeled

1-naphthyl acetate (NA) in acetonitrile under a magnetic field

of 0.64 T.14 Shine and Subotkowski reported an MIE of a

photo-Fries reaction of 4-methoxyphenyl acetate in ethanol.15

In contrast, Lochbrunner et al. measured broad and weak

femtosecond transient absorption spectra of a photo-Fries

reaction of 4-tert-butylphenyl acetate in cyclohexane and

reported a cyclohexadienone formation time of 13 ps.18

Because the rate constants for spin conversion of RPs are

109–108 s�1, MIEs should not be observed for the photo-Fries

reactions with a cyclohexadienone formation time of 13 ps.

Moreover, in non-viscous solvents, such as acetonitrile,

ethanol, and cyclohexane, the RPs should disappear within

several hundreds of picoseconds by diffusion, but the spin

conversion of the RPs is not fast enough to compete with the

diffusion process.20–22 As such, the mechanisms of photo-Fries

reactions are still unclear and, therefore, it is worthwhile

to study photochemical primary processes of photo-Fries

rearrangement reactions in non-viscous homogeneous

solutions. Recently, we reported preliminary results for the

photochemical primary process of the photo-Fries rearrangement

reaction of NA in n-hexane.4 In this paper, we provide a full

investigation of this reaction by means of the MFE probe.

2. Experimental

1-Naphthyl acetate (NA, cica, SP grade) was purified by

passing through silica gel with n-hexane solution containing

5% of dichloromethane, and then the filtrate was evaporated

to obtain a white solid. The white solid was recrystallized twice

from n-hexane, yielding white needles. n-Hexane (cica, SP

grade), cyclohexane (cica, SP grade) and methylcyclohexane

(cica, SP grade) solvents were used as received. Viscosity was

measured with a viscometer (CBC, VM-10A-L) at 298 K.

Water content was measured by a Karl Fischer coulometer

(Metrohm, 831 KF Coulometer). The properties of n-hexane

and cyclohexane are listed in Table 1.23

aDepartment of Chemistry, Graduate School of Science andEngineering, Saitama University, 255 Shimo-okubo, Sakura-ku,Saitama-shi, Saitama 338-8570, Japan.E-mail: mwakasa@chem.saitama-u.ac.jp

bNational Institute for Materials Science (NIMS), 3-13 Sakura,Tsukuba, Ibaraki 305-0003, Japan

PAPER www.rsc.org/pccp | Physical Chemistry Chemical Physics

Dow

nloa

ded

by S

aita

ma

Dai

gaku

Fuz

oku

Tos

hoka

n on

20

Janu

ary

2011

Publ

ishe

d on

03

Nov

embe

r 20

10 o

n ht

tp://

pubs

.rsc

.org

| do

i:10.

1039

/C0C

P000

77A

View Online

756 Phys. Chem. Chem. Phys., 2011, 13, 755–761 This journal is c the Owner Societies 2011

A laser flash photolysis apparatus with a superconducting

magnet (Toshiba), which can generate magnetic fields up to

7 T, was newly constructed to accurately measure MFEs. The

excitation light source was the fourth harmonic of a Nd:YAG

laser (Spectra Physics, GCR-130-10, fwhm E 7 ns, 6.5 mJ per

pulse). A quartz cell was placed at the center of the magnet,

and a laser beam was introduced to the cell at an angle

perpendicular to a monitoring light beam. The optical path

of the monitoring light was 5 or 10 mm. The energy of the laser

beam was recorded for each excitation pulse by an energy

meter (Gentec, PRJ-M, with ED-100 thermopile head). The

fluctuation of the excitation energy was less than 3% during

each experiment. To measure the MFEs at lower magnetic

fields of 0 to 1.65 T, we used an electromagnet (Tokin,

SEE-10W). The laser flash photolysis apparatus equipped with

this electromagnet has been described previously.24 All laser

flash photolysis experiments were carried out at 298 K. The

sample solution was deoxygenated by bubbling with pure

nitrogen for 30 min and was pumped through a quartz flow

cell. The concentration of NA was 0.50 � 10�3 mol dm�3.

The fluorescence quantum yield was measured by an

absolute photoluminescence quantum yield measurement

system (Hamamatsu, C9920-02G) at room temperature. The

excitation wavelength for fluorescence measurements was 266 nm,

and the concentration of NA was 0.50 � 10�3 mol dm�3 in

n-hexane.

3. Results and discussion

3.1 Transient absorption spectra

Transient absorption spectra observed for the reaction of

1-naphthyl acetate (NA) in n-hexane at delay times of 0.03,

0.01, 1.0, 10, 20 and 50 ms after laser excitation are shown in

Fig. 1. The spectrum observed at a delay time of 1.0 ms afterlaser excitation had three peaks at 320, 390 and 410 nm,

whereas that observed at 50 ms had broad bands around 320

and 390 nm. The spectra observed at delay times of 1.0 ms andlater were essentially the same as those previously reported.14,16

Nakagaki et al. reported that the 1-naphthoxyl radical exhibits

transient absorption around 380–410 nm.14 However, Gritsan

et al. assigned this band to T–T absorption of NA, whereas the

absorption of 1-naphthoxyl radical was reported to exhibit a

slow decay around 380 nm.16 Thus the assignments of the

transient absorption spectra observed for the photo-Fries

rearrangement reaction of NA are somewhat complicated at

present.

Time profiles of the transient absorption, A(t), of NA

observed in n-hexane at 320, 380, 410, and 430 nm are shown

in Fig. 2. The A(t) curves observed at 380 and 410 nm agreed

well with an exponential fit (k = 1.41 � 105 s�1) combined

with a slow second-order component. Since the exponential

decay component was magnetically inactive, it was assigned to

the T–T absorption of NA. As discussed below, the second-order

component was magnetically active; thus, this component

could be safely assigned to the 1-naphthoxyl radical. The

second-order rate constant (k/el) was calculated to be

2.65 � 107 s�1 mol�1 dm3, where e is the extinction coefficient

and l is the illuminated path length. Using data reported by

Nakagaki et al. for NA in acetonitrile,14 we estimated the

second-order rate constant to be 1 � 107 s�1 mol�1 dm3. This

estimated value is similar to the present experimental data

obtained with an illuminated path length of about 5 mm.

The A(t) curve observed at 320 nm in n-hexane has an

exponential decay component and almost constant component.

Table 1 Properties of solvent

Solvent ea Z/mPa s Water content/ppm

n-Hexane 1.8799 0.2942a 22.4Cyclohexane 2.02 0.94 17.0

a Ref. 23.

Fig. 1 Transient absorption spectra observed for the photo-Fries

reaction of 1-naphthyl acetate (NA) in n-hexane (0.50� 10�3 mol dm�3).

Emissions observed around 320 nm are fluorescence of NA.

Fig. 2 Time profiles of transient absorption observed for NA in

n-hexane (0.50 � 10�3 mol dm�3) at 320, 380, 410, and 430 nm.

Dow

nloa

ded

by S

aita

ma

Dai

gaku

Fuz

oku

Tos

hoka

n on

20

Janu

ary

2011

Publ

ishe

d on

03

Nov

embe

r 20

10 o

n ht

tp://

pubs

.rsc

.org

| do

i:10.

1039

/C0C

P000

77A

View Online

This journal is c the Owner Societies 2011 Phys. Chem. Chem. Phys., 2011, 13, 755–761 757

Since the exponential decay component is magnetically

inactive and has a rate constant (1.32 � 105 s�1) similar to

that observed at 410 nm, it can be assigned to the T–T

absorption of NA. However, almost constant one may be

assigned to cyclohexadienone intermediate, because the life-

time of cyclohexadienone intermediates observed for the

photo-Claisen rearrangement reactions, which are similar

rearrangement reactions of allyl phenyl ethers, was reported

as ms time region.25,26

As shown in Fig. 1, the spectrum observed at a delay time of

0.03 ms had only a single peak at 410 nm, and the shape of the

spectrum, especially in the longer wavelength region, was

different from that of the spectrum observed at a delay time

of 1.0 ms. Moreover, each A(t) curve observed at 430–600 nm

had a fast decay component, as shown in Fig. 3. The lifetime

of the fast decay component was observed to be 24 ns

(k = 4.10 � 107 s�1), and this component may be attributed

to the RP complex of 1-naphthoxyl and acetyl radicals. This

assignment is discussed further in the next section.

The time profiles of the transient absorption observed in

n-hexane and cyclohexane at 410 and 430 nm are shown in

Fig. 4. As shown in Fig. 4a, the decay profile observed at

410 nm was dependent on the solvent composition: the decay

observed in cyclohexane (upper trace) was slower than that

observed in n-hexane (lower trace). It is reasonable to conclude

from these data that the lifetime of the triplet state and the

second-order reaction were strongly affected by the solvent

viscosity. In contrast, no change in the fast decay component

was observed for different solvents at 430 nm (Fig. 4b). Since

the first decay component was almost independent of the

solvent viscosity, this component may be ascribable to an

intra-molecular reaction.

3.2 Magnetic field effects

A(t) curves were measured at 410 nm for 1-naphthoxyl radical

in the absence and presence of magnetic fields up to 7 T. The

A(t) curves observed at 0 and 1.65 T in n-hexane are shown in

Fig. 5. The A(t) curves observed both at 0 and 1.65 T exhibit

exponential and second-order decay components due to the

T–T absorption of NA and 1-naphthoxyl radical, respectively.

Fig. 5 shows an appreciable MFE on the second-order decay

component of 1-naphthoxyl radical. The fact that MFEs were

observed for the present reaction in n-hexane strongly indicates

that the lifetime of the RPs of 1-naphthoxyl and acetyl radicals

are as long as several nanoseconds, because the spin conversion

process should compete with the disappearance of the RPs.

To elucidate the mechanism of the MFEs, we measured the

magnetic field dependence under magnetic fields up to 7 T.

Using the escaped radical yield Y(B) at a delay time of 3 msafter laser excitation, we calculated the relative radical yield as

R(B) = Y(B)/Y(0 T) = A(B)/A(0 T). The obtained R(B)Fig. 3 Time profiles of transient absorption observed at 380, 410, and

430 nm in the time range of 0–0.9 ms.

Fig. 4 Time profiles of transient absorption observed in n-hexane and

cyclohexane (a) at 410 nm in the time range of 0–70 ms, and (b) at

430 nm in the time range of 0–0.9 ms.

Fig. 5 Time profiles of transient absorption observed at 410 nm in the

absence and presence of a magnetic field of 1.65 T. Inset: plots in the

time range of 2–4 ms.

Dow

nloa

ded

by S

aita

ma

Dai

gaku

Fuz

oku

Tos

hoka

n on

20

Janu

ary

2011

Publ

ishe

d on

03

Nov

embe

r 20

10 o

n ht

tp://

pubs

.rsc

.org

| do

i:10.

1039

/C0C

P000

77A

View Online

758 Phys. Chem. Chem. Phys., 2011, 13, 755–761 This journal is c the Owner Societies 2011

values are plotted against B in Fig. 6. The magnetic field

dependence on R(B) has an inverted feature. R(B) dropped

dramatically upon application of a 3 mT field, giving the

minimum observed R(B) of 0.97 � 0.02. Then R(B) gradually

increased with increasing B to 7 T. A cross field of R(B) = 1

was observed at 0.1 T. No saturation of R(B) was observed at

the maximum field of 7 T (R(7 T) = 1.07 � 0.02).

These observed MFEs can be classified by several

mechanisms as follows:1–3 (1) the Dgmechanism (DgM), which

is due to the difference between the isotropic g factors of two

radicals in a pair; (2) the hyperfine coupling mechanism

(HFCM), which is due to the isotropic hyperfine interaction

between electron and nuclear spins; (3) the level crossing

mechanism (LCM), which is due to crossing between the S

and T+1(or T�1) levels; (4) the relaxation mechanism (RM),

which is due to the anisotropic g tensor (dg), HFC (dHFC),

and the spin–spin dipolar interactions of RPs.

Among these mechanisms, the inverted MFEs on R(B) can

be explained by two possible mechanisms. One possible

mechanism is the MFEs due to the RM. In the present

reaction in n-hexane, however, the solvent viscosity was too

small to observe the MFEs by the RM. In other words, the

escaping rate of the RPs was larger than the spin relaxation

rate. Therefore, we can exclude the possibility of the RM.

Another possible mechanism to explain the inverted MFEs

on R(B) is the HFCM associated with the DgM. The

rate constant of S–T conversion by the HFCM (kHFCM) is

given by1–3

kHFCM = (1/2)gmBB1/2/h. (1)

Here, g and mB are the g-value of the free electron (2.00231)

and the Bohr magneton, respectively. B1/2 is the effective

hyperfine coupling (Weller’s half-field of the saturation) in

the RP, which can be expressed by27

B1/2 = 2 (B12 + B2

2)/(B1 + B2), (2)

Bi ¼Xk

a2ikIkðIk þ 1Þ( )1=2

ð3Þ

Here, aik is the isotropic hyperfine coupling constant of the kth

nuclei in radical (i = 1, 2). Ik is the nuclear spin. Using the

reported hyperfine coupling constants of 1-naphthoxyl28 and

acetyl radicals,29 kHFCM in the absence of a magnetic field was

estimated to be 9.16 � 107 s�1. In the presence of a magnetic

field, the triplet sublevel splits into three levels (T0, T+1, and

T�1) by means of the Zeeman interaction, and S–T+1

and S–T�1 spin conversions are obstructed. Thus the S–T

conversion rate by the HFCM decreases with increasing B and

reaches a value that is 3 fold of that observed in the zero field.

Since the obstruction of the spin conversions S–T+1 and

S–T�1 occurs at higher magnetic fields of the HFC, the

decrease in R(B) is saturated at those fields. The half-field of

the saturation (B1/2) for the present RP was calculated by

eqn (2) and (3) to be 1.04 mT.27,28 As shown in Fig. 6b, the

R(B) value reached its minimum at 3 mT, which may have

been the saturation field for the HFCM. Therefore, B1/2 was

roughly estimated to be 1–2 mT. This estimated value agrees

well with our calculated value of 1.04 mT. These results

indicate that the sharp drop in R(B) observed at 3 mT can

safely be explained by the HFCM.

The rate constant of S–T conversion by the DgM (kDgM) is

given by

kDgM = (1/2)DgmBB/h. (4)

Here, Dg is the difference between the g-values of two radicals.

The kDgM value at 0 T is zero, but this value increases

proportionally with increasing B. Since Dg was estimated to

be 0.00381 from the reported g-values of 1-naphthoxyl radical

(2.00431)30 and acetyl radical (2.0005),29 kDgM at 7 T was

calculated to be 1.17 � 109 s�1. From these calculations, we

can estimate the magnetic field dependence of the overall S–T

conversion (kS�T) as follows:

kS�T = kDgM + (1/3) kHFCM. (5)

The rate constant of S–T conversion decreased dramatically

at low fields and then increased with increasing B to reach

1.20� 109 s�1 at 7 T. This observed behavior indicates that the

present MFEs can be qualitatively interpreted by the HFCM

associated with the DgM.

This spin conversion process by the HFCM and the DgMcompetes with the escaping process of the RPs. Therefore, we

also considered the rate constant of the escaping process. The

rate of escape (kesc) from a solvent cage depends on the

solvent’s viscosity. The simplest expression of the kesc is

described by the Stokes–Einstein relationship as follows:

kesc = 1/t = kBT/12pr3Z. (6)

Fig. 6 (a) Magnetic field dependence on the relative radical yield

(R(B)) of 1-naphthoxyl radical observed at 410 nm. Circles and

diamonds denote the used transient absorption apparatus with an

electromagnet (EM) and that with a superconducting magnet (SCM),

respectively. (b) Plots in the lower magnetic field range of 0–0.12 T.

Dow

nloa

ded

by S

aita

ma

Dai

gaku

Fuz

oku

Tos

hoka

n on

20

Janu

ary

2011

Publ

ishe

d on

03

Nov

embe

r 20

10 o

n ht

tp://

pubs

.rsc

.org

| do

i:10.

1039

/C0C

P000

77A

View Online

This journal is c the Owner Societies 2011 Phys. Chem. Chem. Phys., 2011, 13, 755–761 759

Here, kB is the Boltzmann constant, T is the temperature, and

Z is the viscosity of the solvent. The lifetime (t) of a neutral RP

is defined as t = 4r2/D, where r and D are the van der Waals

radii of the component radicals and the sum of the diffusion

coefficients of each radical, respectively.31 In the present study,

kesc was estimated to be 3.0 � 109 s�1 at 298 K using

Z = 0.2942 � 10�3 Pa s, rnaphthoxyl = 0.71 nm, and racetyl =

0.29 nm.32,33 Comparison of this value with the calculated S–T

conversion rate of 1.20 � 109 s�1 at 7 T reveals that the escape

rate is much larger than the S–T conversion rate. Therefore,

the MFEs observed in this study cannot be explained by the

normal diffusion process calculated from the Stokes–Einstein

relationship. Instead, a much slower escape process involving

the RP complex is suggested.

To clarify the formation of such an RP complex, the A(t)

curves observed in the time range of 0–0.9 ms were examined in

detail. As shown in Fig. 3, in the time range of 0–0.1 ms, a fast

decay component was observed at 430 nm concomitant with a

fast rise component observed at 380 and 410 nm. Such fast

decay component was also observed at 430–600 nm. The rate

constants of both decay (at 430 nm) and rise (at 410 nm) were

4.10 � 107 s�1. Since the transient absorption observed at

380 and 410 nm consists of both the 1-naphthoxyl radical and

the T–T absorption of NA, the fast decay component observed

at 430–600 nm could be assigned to the singlet–singlet (S–S)

absorption of NA or to an absorption of a parent intermediate

of the 1-naphthoxyl radical, such as an RP complex. To clarify

the assignment of the fast decay component, the rate

constant of fluorescence (kF) was measured in n-hexane and

in cyclohexane at 323 nm under the same conditions as those

used for the transient absorption measurement. The kF values

observed in n-hexane and cyclohexane were 3.99 � 107 and

5.29 � 107 s�1, respectively. As shown in Fig. 4b, the fast

decay component was independent of solvent composition.

Therefore, the fast decay component cannot be assigned to the

S–S absorption of NA, because the kF values in n-hexane

and in cyclohexane were different. From these results, we

concluded that the fast kinetics of decaying at 430–600 nm

could be safely assigned to the disappearance of the

1-naphthoxyl radical’s parent intermediate, namely an RP

complex. We further concluded that the fast kinetics of rising

observed at 380 and 410 nm could be attributed to the

formation of free 1-naphthoxyl radicals.

Among the published reports on the MFEs of photo-

chemical reactions, there are several reports of MFEs in

non-viscous homogeneous solvents; such MFEs are classified

as the d-type triplet mechanism (TM).34–36 This class of MFEs

is observed only from the triplet precursor and is often

observed for RPs containing heavy atoms such as d-block

metals,34 sulfur35 or phosphorus.36 The key intermediates for

the d-type TM are excited triplet molecules or RP complexes.

For the present reaction, although this class of MFEs can

safely be excluded, it is worthwhile to consider the character of

the RP complex. In this study, the MFEs due to the LCMwere

not observed. Therefore, the exchange interaction should have

been small enough to observe MFEs due to the HFCM and

the DgM. Even in a non-viscous solvent, such complexation

enables long-lifetime RPs. Moreover, since the fast decayed

transient absorption observed at 430–600 nm was very broad

and not very strong, the RP complex may be characterized as

similar to a charge transfer complex or an electron-donor–

acceptor complex. Theoretical calculations may be useful to

elucidate the ultraviolet-visible spectrum of such complex.

However, those calculations are beyond the scope of the

present experimental study of the photochemical primary

process.

3.3 Time-resolved EPR measurements

Next, to directly measure the RP complex, we carried out

time-resolved electron paramagnetic resonance (EPR)

measurements on an n-hexane solution of NA. The concentration

of the NA solution was 2.0 � 10�3 mol dm�3 and the solution

was pumped through a quartz cell at a flow rate of 4 mL min�1.

No signal of spin-correlated RPs was observed under

excitation by the fourth harmonic of a Nd:YAG laser. This

lack of signal might have been due to the fast rearrangement

reaction of the RP complex. The time resolution of our

apparatus is ca. 100 ns, and the spin polarization of the RP

is generated within a similar time range. If the lifetime of the

RP complex was as short as 24 ns (k = 4.10 � 107 s�1),

it would be difficult to observe the signal of the RPs by

time-resolved EPR.

3.4 Photochemical primary process of the photo-Fries

rearrangement reaction

Upon irradiation of a degassed n-hexane solution of NA

(4.0 � 10�3 mol dm�3) with an Ushio 500 W Xe lamp,

2-acetyl-1-naphthol was obtained as the main product. For

an aerated solution of NA, the yield of 2-acetyl-1-naphthol

decreased by only 8 � 2%. Such a small decrease in yield

indicates that the present reaction occurred fast enough to

avoid quenching by oxygen. Nakagaki et al. have also

reported that dissolved oxygen has no effect upon the initial

transient absorption during laser flash photolysis of NA.14

From the results of the present study, the reaction mechanism

can be described as follows (Scheme 1):

Here, 1,3NA*, NpO�, Ac�, and 1,3(NpO� �Ac) denote the

singlet or triplet excited state of NA, 1-naphthoxyl radical,

acetyl radical, and the singlet or triplet RP, respectively. Upon

irradiation of NA by the fourth harmonic of the nanosecond

Nd:YAG laser, 1NA* is immediately generated. Fluorescence

Scheme 1 Mechanism of the photo-Fries rearrangement reaction.

Dow

nloa

ded

by S

aita

ma

Dai

gaku

Fuz

oku

Tos

hoka

n on

20

Janu

ary

2011

Publ

ishe

d on

03

Nov

embe

r 20

10 o

n ht

tp://

pubs

.rsc

.org

| do

i:10.

1039

/C0C

P000

77A

View Online

760 Phys. Chem. Chem. Phys., 2011, 13, 755–761 This journal is c the Owner Societies 2011

and phosphorescence of NA were observed in n-hexane and in

methylcyclohexane around 310–390 nm (peaks: 317.8, 322.6,

328.0, 332.2, 337.6, 342.0, 348.0 nm) at room temperature and

around 470–630 nm (peaks: 479.0, 491.0, 514.6, 555.0, 603.2 nm)

at 77 K, respectively. The excitation energies estimated from

the 0–0 band were 384 kJ mol�1 for S1 and 255 kJ mol�1 for

T1. Fluorescence is emitted with a quantum yield of 0.17 at

room temperature, and intersystem crossing spontaneously

occurs to form 3NA*. T–T absorption was observed at

380–410 nm. Since no decomposition reaction occurs from3NA*,37–39 we concluded that 3NA* is likely deactivated by

thermal relaxation at room temperature or by phosphorescence

radiation at 77 K. 1NA* generates the singlet RP complex of1(NpO� �Ac). The RP complex of 1(NpO� �Ac) can convert to

a triplet RP complex of 3(NpO� �Ac), and this spin conversion

process is affected by the magnetic field.1–3 From the singlet

RP complex of 1(NpO� �Ac), rearrangement products (in-cage

products) are generated. In contrast, 3(NpO� �Ac) generates

escaped radicals that subsequently form escaped products,

because the triplet RP cannot recombine.

Some authors have reported a cyclohexadienone inter-

mediate as the key species of photo-Fries rearrangement

reaction, and they have further reported that a subsequent

proton shift is generally slow even if the cyclohexadienone

intermediate is unstable.26

1NA* - 1(NpO� Ac�), (7)

1(NpO� Ac�) - cyclohexadienone(s), (8)

Cyclohexadienone(s) - acetyl naphthol(s). (9)

Considering the results obtained from the present MFE

measurements, we concluded that the key species are the

RPs, because magnetic fields only affected the spin conversion

process of the RPs. According to Lochbrunner et al., the

formation time of cyclohexadienone is 13 ps for 4-tert-

butylphenyl acetate in cyclohexane.18 If the geminate singlet

RPs are depleted by the formation of cyclohexadienone within

several tens of picoseconds, then no MFEs would be observed

for the RPs. Therefore, because MFEs were indeed observed

for the RPs, we concluded that the RP complex rather than the

cyclohexadienone intermediate should be considered the

primary intermediate in this reaction. In the previous section,

we assigned the fast-decaying component observed at 430 nm

to the disappearance of the RP complex. It is reasonable to

conclude that the transient signal of the 1-naphthoxyl radical

observed at 410 nm increased as the signal at 430 nm from the

RP complex decreased.

4. Conclusion

Laser flash photolysis under magnetic fields up to 7 T was

carried out for the photo-Fries rearrangement reactions of

1-naphthyl acetate in n-hexane and cyclohexane. The transient

absorption of the 1-naphthoxyl radical was observed at

380–410 nm in both solvents, and appreciable MFEs on its

escaped yields were observed for the first time. Observation of

these MFEs shows the existence of the RP intermediates in

the present reaction, but the escape of the RP cannot be

explained by the normal diffusive process. The presence of

an RP complex, which had a lifetime of 24 ns, was strongly

suggested by the slow escaping process of the RP.

Consequently, we concluded that the observed MFEs can be

explained by the HFCM associated with DgM.

Acknowledgements

We thank Professor Yoshio Sakaguchi of RIKEN for

measuring time-resolved EPR spectrum. This work was

partially supported by a Grant-in-Aid for Scientific Research

(No. 2003002) in the Priority Area ‘‘High Field Spin Science in

100 T’’ (No. 451) from the Ministry of Education, Culture,

Sports, Science, and Technology (MEXT) of Japan.

References

1 U. E. Steiner and T. Ulrich, Chem. Rev., 1989, 89, 51–147.2 S. Nagakura, H. Hayashi and T. Azumi, Dynamic Spin Chemistry,Kodansha-Wiley, Tokyo, NY, 1998.

3 H. Hayashi, Introduction to Dynamic Spin Chemistry, WorldScientific, Singapore, 2004.

4 M. Gohdo and M. Wakasa, Chem. Lett., 2010, 39, 106–107.5 A. Hamasaki, T. Yago and M. Wakasa, J. Phys. Chem. B, 2008,112, 14185–14192.

6 M. Wakasa, J. Phys. Chem. B, 2007, 111, 9434–9436.7 A. Hamasaki, T. Yago, T. Takamasu, G. Kido and M. Wakasa,J. Phys. Chem. B, 2008, 112, 3375–3379.

8 M. Wakasa, T. Yago and H. Hamasaki, J. Phys. Chem. B, 2009,113, 10559–10561.

9 S. Grimme, Chem. Phys., 1992, 163, 313–330.10 M. A. Miranda and F. Galindo, CRC Handbook of Organic

Photochemistry and Photobiology, CRC Press, Boca Raton,2nd edn, 2004, ch. 42, pp. 42-1.

11 V. I. Stenberg, Org. Photochem., 1967, 1, 127.12 D. Bellus, Adv. Photochem., 1971, 8, 109–159.13 R. Martin, Org. Prep. Proced. Int., 1992, 24, 369–435.14 R. Nakagaki, M. Hiramatsu, T. Watanabe, Y. Tanimoto and

S. Nagakura, J. Phys. Chem., 1985, 89, 3222–3226.15 H. J. Shine and W. Subotkowski, J. Org. Chem., 1987, 52,

3815–3821.16 N. P. Gritsan, Y. P. Tsentalovich, A. V. Yurkovskaya and

R. Z. Sagdeev, J. Phys. Chem., 1996, 100, 4448–4458.17 I. F. Molokov, Y. P. Tsentalovich, A. V. Yurkovskaya and

R. Z. Sagdeev, J. Photochem. Photobiol., A, 1997, 110, 159–165.18 S. Lochbrunner, M. Zissler, J. Piel, E. Riedle, A. Spiegel and

T. Bach, J. Chem. Phys., 2004, 120, 11634–11639.19 T. Mori, M. Takamoto, H. Saito, T. Furo, T. Wada and Y. Inoue,

Chem. Lett., 2004, 254–255.20 R. M. Noyes, J. Am. Chem. Soc., 1956, 78, 5486–5490.21 E. N. Step, A. L. Buchachenko and N. J. Turro, J. Org. Chem.,

1992, 57, 7018–7024.22 N. J. Turro, A. L. Buchachenko and V. F. Tarasov, Acc. Chem.

Res., 1995, 28, 69–80.23 Organic Solvents, ed. J. A. Riddick, W. B. Bunger and

T. K. Sakano, Wiley, New York, 4th edn, 1986.24 M. Wakasa, K. Nishizawa, H. Abe, G. Kido and H. Hayashi,

J. Am. Chem. Soc., 1999, 121, 9191–9197 and references citedtherein.

25 F. Galindo, J. Photochem. Photobiol., C, 2005, 6, 123–138.26 M. C. Jimenez, M. A. Miranda, J. C. Scaiano and R. Tormos,

Chem. Commun., 1997, 1487–1488.27 A. Weller, F. Nolting and H. Staerk, Chem. Phys. Lett., 1983, 96,

24–27.28 W. T. Dixon, W. E. J. Foster and D. Murphy, J. Chem. Soc.,

Perkin Trans. 2, 1973, 2124–2127.29 J. Benett and B. Mile, Trans. Faraday Soc., 1971, 67, 1587–1597.30 M. Adams, M. S. Blois Jr. and R. H. Sands, J. Chem. Phys., 1958,

28, 774–776.31 K. M. Salikhov, Magnetic Isotope Effect in Radical Reactions,

Springer-Verlag, Weinheim, Germany, 1996.

Dow

nloa

ded

by S

aita

ma

Dai

gaku

Fuz

oku

Tos

hoka

n on

20

Janu

ary

2011

Publ

ishe

d on

03

Nov

embe

r 20

10 o

n ht

tp://

pubs

.rsc

.org

| do

i:10.

1039

/C0C

P000

77A

View Online

This journal is c the Owner Societies 2011 Phys. Chem. Chem. Phys., 2011, 13, 755–761 761

32 W. Gu, D. J. Abdallah and R. G. Weiss, J. Photochem. Photobiol.,A, 2001, 139, 79–87.

33 Each van der Waals radius was estimated with the reportedcrystallographic data of NA in ref. 32, assuming that bondcleavage will not cause big change to molecular structure. Fromthe EPR study reported in ref. 29, the structure of acetyl radical is abend form and C–C–O angle of the radical (1301 by EPR) is similarto that of NA (127.141 by XRD). Estimated radii were 0.71 nm fornaphthoxyl radical and 0.29 nm for acetyl radical defined by themaximum length of molecule.

34 U. E. Steiner, Chem. Phys. Lett., 1980, 74, 108–112.

35 P. Gilch, M. Linsenmann, W. Haas and U. E. Steiner, Chem. Phys.Lett., 1996, 254, 384–390.

36 Y. Sakaguchi and H. Hayashi, J. Phys. Chem. A, 2004, 108,3421–3429.

37 C. E. Kalmus and D. M. Hercules, J. Am. Chem. Soc., 1974, 96,449–456.

38 P. Subramanian, D. Creed, A. C. Griffin, C. E. Hoyle andK. Venkataram, J. Photochem. Photobiol., A, 1991, 61,317–327.

39 S. Grimme and H. Dreeskamp, J. Photochem. Photobiol., A, 1992,65, 371–382.

Dow

nloa

ded

by S

aita

ma

Dai

gaku

Fuz

oku

Tos

hoka

n on

20

Janu

ary

2011

Publ

ishe

d on

03

Nov

embe

r 20

10 o

n ht

tp://

pubs

.rsc

.org

| do

i:10.

1039

/C0C

P000

77A

View Online