Fries rearrangement of dibenzofuran-2-yl ethanoate under photochemical and Lewis-acid-catalysed conditions

Article abstract of DOI:10.1016/j.tet.2004.05.060

The Fries rearrangement of dibenzofuran-2-yl ethanoate as a route to o-hydroxyacetyldibenzofurans has been investigated, both under thermal Lewis-acid catalysed and non-catalysed photochemical conditions. The reactions were examined theoretically at semi-empirical (PM3 and ZINDO/S) and density functional theory (DFT) levels. The correct selection of reaction conditions provides viable preparative routes to ortho-acylated hydroxydibenzofurans.

Full text of DOI:10.1016/j.tet.2004.05.060

Tetrahedron 60 (2004) 6145–6154
Fries rearrangement of dibenzofuran-2-yl ethanoate under
photochemical and Lewis-acid-catalysed conditions
a
b
Ana M. A. G. Oliveira,a Ana M. F. Oliveira-Campos,a M. Manuela M. Raposo, John Griffiths
,
*
and Antonio E. H. Machadoc
a
´
Centro de Quımica, Universidade do Minho, Campus de Gualtar, 4710 Braga, Portugal.
bDepartment of Colour Chemistry, University of Leeds, Leeds LS2 9JT, UK
ˆ ´ ´ ´
Universidade Federal de Uberlandia—Instituto de Quımica, Laboratorio de Fotoquımica/GFQM, P.O. Box 593, Uberlandia, 38400-089,
c
ˆ
Minas Gerais, Brazil
Received 27 January 2004; revised 3 May 2004; accepted 14 May 2004
Available online 11 June 2004
Abstract—The Fries rearrangement of dibenzofuran-2-yl ethanoate as a route to o-hydroxyacetyldibenzofurans has been investigated, both
under thermal Lewis-acid catalysed and non-catalysed photochemical conditions. The reactions were examined theoretically at semi-
empirical (PM3 and ZINDO/S) and density functional theory (DFT) levels. The correct selection of reaction conditions provides viable
preparative routes to ortho-acylated hydroxydibenzofurans.
q 2004 Elsevier Ltd. All rights reserved.
1. Introduction
In the present work, the photochemical and the Lewis acid
catalysed thermal Fries rearrangements of 2 under a variety
of conditions have been investigated and the results
compared with theoretical predictions.
Following previous work regarding the preparation of
benzopsoralen analogues,1 the synthesis of derivatives
containing an acetyl group in position 4 of the pyranone
ring was examined. This required the synthesis of acetyl-
substituted hydroxydibenzofurans, and the readily available
dibenzofuran-2-ol (1) was chosen as the starting material, as
its acetyl ester 2 should undergo Fries rearrangement and
potentially afford derivatives of this type (Fig. 1).
2. Experimental results
The synthesis of dibenzofuran-2-yl ethanoate (2) was
obtained in almost quantitative yield (98% yield) by
reaction of dibenzofuran-2-ol (1) with diethylmalonate, in
pyridine.9
The Fries rearrangement of benzene derivatives is well
known and can be induced either thermally or photo-
chemically.2,3 The thermal Fries reaction involves the
formation of a solvent caged ion pair,4 whereas the
photoreaction occurs via radical intermediates.2,5 In
the latter case, electronic excitation is followed by
homolytic cleavage to yield aryloxy and acyl radical
pairs,6 and calculations indicate that this involves the
excited singlet state.7 The photo-Fries rearrangement was
first observed by Anderson and Reese, in 1960.8 In both
processes, the resulting pairs of ions or radicals are
restrained by the solvent cage, until they combine to form
rearrangement products. This can involve reformation of the
original ester or formation of several isomeric ketone
intermediates. Alternatively, the reactive species can diffuse
apart to give other products.
The acid catalysed experiments were conducted either
without solvent or in dichloromethane solution, using AlCl3
or TiCl4 as Lewis acids, at various temperatures.
The photo-Fries reactions were carried out by sealing
solutions of the dibenzofuran (solvents: ethanol, cyclo-
hexane, dichloromethane and acetonitrile) in quartz tubes
and exposing them for the appropriate length of time to the
radiation from a 16 W low-pressure mercury lamp
(principal emission at 254 nm). The main products (Fig. 1)
were isolated by column chromatography and characterized
1
by H NMR, UV, IR and elemental analysis.
The product yields for both the dark and photochemical
reactions were estimated by separation of the products by
column chromatography or by high performance liquid
chromatography (HPLC), using a silica column (Fig. 2). In
order to estimate the yields for the main products by HPLC
Keywords: Fries rearrangement; Dibenzofuran-2-yl ethanoate; Acetylation.
*
Corresponding author. Tel.: þ35-253-604-386; fax: þ35-253-678-983;
e-mail address: amcampos@quimica.uminho.pt
0040–4020/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2004.05.060

6146
A. M. A. G. Oliveira et al. / Tetrahedron 60 (2004) 6145–6154
Figure 1. Products obtained in the Fries rearrangements.
Figure 2. HPLC chromatogram from experiment 4 (see Table 1). Conditions: Column: LichroCART (5 mm, silica 60). Dimensions: 25 cm£4 mm. Mobile
phase: AcOEt/hexane 20:80. Flow rate: 1.6 ml/min. Detector: UV absorption (290 nm, 0.64 AUFs). Sample volume: 5 ml.

A. M. A. G. Oliveira et al. / Tetrahedron 60 (2004) 6145–6154
6147
Table 1. Products obtained from the catalysed Fries rearrangement of 2
under various experimental conditions
2). However, when a mixture of reactant and catalyst with
no solvent was heated for a short period of time (15 min),
rearrangement took place readily, and the main products,
apart from the phenol (1), formed by hydrolysis, were
2-hydroxydibenzofuran-3-yl methyl ketone (3a) and
8-hydroxydibenzofuran-2-yl methyl ketone (3b) (see
Table 1, experiments 4 and 5). Only traces of 2-hydroxy-
dibenzofuran-1-yl methyl ketone (3c) were observed. Other
products were also obtained and two were identified as 8-
acetyldibenzofuran-2-yl ethanoate (3d) and 8-hydroxy-
dibenzofuran-2,7-diyl dimethyl diketone (3e).
Exp.
Conditions
Product/Yield (%)
2
3a 3b 3c 3d 3e
1
1
2
3
4
5
2AlCl3, CH2Cl2, 7 days, 20 8C
2AlCl3, CH2Cl2, 30 min, reflux
2TiCl4, CH2Cl2, 3 h, refluxa
2AlCl3, 130 8C, 15 min
94
9
64
0
4
0
0
0
0
0
0
0
3
3
0
4
0
7
4
0
0
0
1
0
2
0
27
10
11
17 31 28
2AlCl3, 130 8C!150 8C, 15 min 19 27 16
Yield obtained by column chromatography.
a
2.2. Photochemical reactions
Table 2. Products obtained from the photo-Fries rearrangement of 2 in
different solvents
The photochemical Fries reaction proved to be more
efficient than the thermal acid-catalysed process. In contrast
to the thermal reaction, the photochemical process showed
that the two rearrangement products were 2-hydroxy-
dibenzofuran-1-yl methyl ketone (3c) and 2-hydroxy-
dibenzofuran-3-yl methyl ketone (3a), with 3c always
present in higher proportion (Table 2). Yields were
dependent on solvent and/or reaction time.
Solvent
Yield, %
Reaction time, min
2
3a
3c
1
Dichloromethane
Cyclohexane
Ethanol
5
19
21
20a
12b
41
32
27
16
10
2
12
12
300
1260
300
11
13
37
Acetonitrile
135
a
22% for 540 min of reaction.
14% for 210 min of reaction.
The best results were obtained in dichloromethane (41% of
3c and 19% of 3a) after 300 min exposure. Roughly similar
yields were obtained in ethanol and cyclohexane, but it was
noteworthy that significantly longer irradiation time was
needed in the latter solvent (Table 2). The reaction was
inefficient in acetonitrile and yields of 3a and 3c were low.
There was a tendency for the primary photoproducts to
undergo secondary reactions with increased irradiation
time. A similar trend was also observed in dichloromethane
and ethanol, but not in cyclohexane. In all cases, the analysis
of the products by HPLC showed formation of species with
lower retention times than 3a or 3c and, in dichloromethane,
also some species with higher retention times.
b
analysis, external calibration was used with solutions of
pure reference compounds and the peak height was plotted
as a function of the concentration of each compound (peak
areas were less useful and difficult to quantify due to the
overlap of peaks).
2.1. Dark reactions
Solutions of dibenzofuran-2-yl ethanoate in dichloro-
methane containing aluminium chloride or titanium
chloride as catalyst showed no significant rearrangement
reaction, either at room temperature or at reflux temperature
(Table 1, experiments 1–3). In some instances a high level
of decomposition of the reactant occurred (e.g. experiment
The formation of the rearrangement product 3a as a function
of time in these different solvents is shown in Figure 3. The
Figure 3. Photo-formation of 2-hydroxydibenzofuran-3-yl methyl ketone (3a), in different solvents: (a) cyclohexane, (b) dichloromethane, (c) ethanol, (d)
acetonitrile. In detail: consumption of dibenzofuran-2-yl ethanoate (2) in the photochemical reaction, using different solvents: (a) acetonitrile, (b) ethanol, (c)
cyclohexane, (d) dichloromethane.

6148
A. M. A. G. Oliveira et al. / Tetrahedron 60 (2004) 6145–6154
Figure 4. Ratios 3a/3c as a function of the conversion of 2.
detail in Figure 3 shows the parallel percentage decrease in
concentration of the starting material 2 with time. The
highest rate of disappearance of dibenzofuran-2-yl ethano-
ate (2) occurred in dichloromethane, and the lowest rate in
cyclohexane. It can be seen from Figure 3 that photo-
decomposition of 3a becomes evident after varying
irradiation times, depending on the solvent, and is most
pronounced in acetonitrile (after ca. 3 h), whereas in
cyclohexane formation of 3a is still increasing after 20 h.
conversion of 2 is shown in Figure 4. The ratio is below
unity and it is approximately constant throughout each
reaction until a relatively high conversion of 2 has been
reached, when the ratio increases sharply. This can be
attributed principally to the fact that 3c decomposes faster
than 3a towards the end of the reaction.
In order to study the effect of the concentration of
starting material on the formation of the reaction
products, ethanolic solutions of 2 of three different
Similar results were also observed for formation of 3c, but
its tendency to undergo photodecomposition was higher. As
the sum of the amounts of 3a, 3c and unreacted 2 do not
reach 100%, it can be concluded that other species that were
not identified, are formed.
concentrations (1£1023, 2£1023, and 5£1023 mol dm23
)
were irradiated, and the percent formation of 3a determined
as a function of irradiation time. The results are shown in
Figure 5. It can be seen that for more dilute solutions,
decomposition starts earlier and consequently the yield
decreases. However, the maximum yield obtained is only
22%.
The variation of the ratios 3a/3c as a function of the
Figure 5. Effect of the dibenzofuran-2-yl ethanoate (2) concentration on the photo-formation of 2-hydroxydibenzofuran-3-yl methyl ketone (3a) in ethanol: (a)
1£1023 mol dm23, (b) 2£1023 mol dm23, (c) 5£1023 mol dm23
.

A. M. A. G. Oliveira et al. / Tetrahedron 60 (2004) 6145–6154
6149
Figure 6. Reaction coordinates, calculated using PM3, considering the formation of 2-hydroxydibenzofuran-3-yl methyl ketone (3a), for the catalysed and
uncatalysed reactions.
3. Discussion
tendency for the ion pair to undergo recombination, so
regenerating the starting material 2 and preventing
rearrangement. This can be overcome by an increase in
the reaction temperature, which will tend to displace the
transition state (ion pair) in the direction of the products.
The concentration of the starting material is also an
important parameter, as high concentrations will favour
the formation of intermolecular recombination products.
3.1. Dark reaction
The reaction coordinates for the production of 2-hydroxy-
dibenzofuran-3-yl methyl ketone (3a), calculated using
PM3 for the catalysed reaction and for the hypothetical non-
catalysed reaction are shown in Figure 6. The reaction
coordinate is the distance between the oxygen atom of
dibenzofuran-2-yl ethanoate and the acylium cation as the
transition state is approached. For the uncatalysed reaction
the coordinate of the reaction is shown until the bond is
formed in 3a. The reaction coordinate for the catalysed
reaction (AlCl3) is the bond distance of O–(CO)CH3 until
reaching the transition state. Comparing DHF calculated
values for the transition state, a decrease of 172 kJ/mol for
the catalysed reaction is indicated. As expected, the
activation energy needed for the catalysed reaction is
lower. The transition state indicated by the maximum of the
uncatalysed curve can be considered to represent the ion
pair formed between the acylium cation and the dibenzo-
furanoxide anion, and similarly, the maximum of the
catalysed curve can be considered to represent the complex
formed between the dibenzofuranoxide anion and the
Lewis’ acid. The activation energy for the uncatalysed
reaction was calculated to be 306 kJ/mol, whereas for the
catalysed reaction the corresponding value was 135 kJ/mol,
some 56% lower.
Significant amounts of dibenzofuran-2-ol (1) could also be
isolated in these Lewis acid catalysed reactions, and this can
be ascribed to protonation of the dibenzofuranoxide anion
after displacement of the acylium cation. Considering
positions in the dibenzofuran ring system that might be
attacked by the acylium ion, in an intramolecular process
the distance between the acylium cation and that position
will play an important role. If one considers the dibenzo-
furanoxide anion, the positions potentially most susceptible
to electrophilic attack by the acylium cation may be
predicted from the charge densities on the carbon atoms at
those positions. Calculated charge densities are shown in
Figure 7.
It can be seen that, on this basis, all the unsubstituted carbon
atoms in the dibenzofuranoxide anion have the potential to
undergo intramolecular acylation, but only products arising
from attack at positions 1, 3 and 8 were isolated. Other
minor products formed were due to intermolecular reac-
tions, as in the case of 8-acetyldibenzofuran-2-yl ethanoate
(3d) and 8-hydroxydibenzofuran-2,7-diyl dimethyl diketone
(3e) and other compounds in small amounts.
The best practical conditions to effect the thermal
rearrangement involves the use of a Lewis acid in
combination with heat. The use of a catalyst in refluxing
dichloromethane did not result in any significant rearrange-
ment as the boiling point of the solvent (40 8C) was too low
to reach an appropriate reaction temperature, and only
decomposition of the starting material was observed. It is
noteworthy that the complex formed between AlCl3 and 2
virtually eliminates the formation of the compound 3c.
The non-reactivity of position 4 may be attributed to the fact
that 4 is meta to the phenoxide group and as with all
electrophilic substitution reactions of phenols and phenolate
anions, this position will be much less reactive than the two
ortho positions 1 and 3 on simple resonance grounds.
Similar resonance considerations, may explain why there is
no reaction in position 7 and 9 in spite of their charge
densities. Thus positions 7 and 9 are meta to the furan
As the ions will possess limited mobility, there will be high

6150
A. M. A. G. Oliveira et al. / Tetrahedron 60 (2004) 6145–6154
Figure 7. Reaction pathways for the Lewis acid-catalysed Fries reaction of dibenzofuran-2-yl ethanoate (2). Total atomic charges over the dibenzofuranoxide
anion, estimated using DFT calculation (B3LYP/6-31G*), are also shown.
oxygen atom. In the case of position 9 there is also the
likelihood of steric inhibition, due the proximity of the
Lewis acid moiety. On charge density grounds and their
favourable ortho or para orientation with respect to the
furan oxygen, one would expect positions 6 and 8 to be both
appreciably reactive to acylation, and yet only the product
arising from reaction at position 8 can be detected. Again
this may be a steric effect, the furan oxygen making ortho
attack at position 6 less favourable.
a large steric effect, so inhibiting reaction at positions 1 and
3. However, position 1 will be additionally hindered to
attack because of the proximity of the hydrogen atom in the
9 position of the second benzene ring.
In the most successful reaction (no solvent; AlCl3 catalyst;
130 8C—experiment 4, Table 1), 2-hydroxydibenzofuran-3-
yl methyl ketone (3a) was the major product.
An estimate of the equilibrium distance between the
dibenzofuranoxide and acylium ions in the ion pair in the
˚
transition state gave a value of around 2.51 A. The estimated
distance between the nearest reaction centres (positions 1
˚
and 3) and the cation is, respectively, 3.88 and 4.58 A. The
formation of the compound 2-hydroxydibenzofuran-3-yl
methyl ketone (3a) must occur through electrophilic
substitution of the acylium cation in the 3-position followed
by proton transfer to oxygen. A summary of the reaction
Of the two dominant rearrangement products 3a and 3c, the
former is by far the most favoured (Table 1), even though
the 1-position has a calculated negative charge density
about 1.5 times higher than for position 3. This is surprising,
as simple electrophilic substitution reactions of 2-hydro-
xydibenzofuran (e.g. diazo coupling, nitration, bromination)
normally strongly favour 1-substitution. This may be due to
a steric effect. The Lewis acid moiety at position 2 will exert
Table 3. Calculated DHf values (PM3) for 2 and its various reaction products from the Fries reaction
Compound
DHf, kJ/mol
Observationa
Dibenzofuran-2-yl ethanoate (2)
2232.74
2250.11
2255.02
2245.31
2406.72
2 422.42
281.26
Starting reagent
Major product
Results from the escape of the cation
Traces
Results from the escape of the cation (intermolecular rearrangement)
From secondary reaction (traces)
Results from the loss of the acylium cation
2-hydroxydibenzofuran-3-yl methyl ketone (3a)
8-hydroxydibenzofuran-2-yl methyl ketone (3b)
2-hydroxydibenzofuran-1-yl methyl ketone (3c)
8-acetyldibenzofuran-2-yl ethanoate (3d)
8-hydroxydibenzofuran-2,7-diyl dimethyl diketone (3e)
Dibenzofuran-2-ol (1)
a
Applies to the dark reactions.

A. M. A. G. Oliveira et al. / Tetrahedron 60 (2004) 6145–6154
6151
sequences taking place in the non-photochemical Lewis
acid—catalysed reaction of dibenzofuran-2-yl ethanoate (2)
is shown in Figure 7.
marked contrast to Plank’s observations,10 in which he
noted that polar solvents such as methanol favoured radical
rearrangement, whereas non-polar solvents favoured radical
migration followed by hydrogen abstraction.
The calculated standard enthalpies of formation for the
isolated compounds, are listed in Table 3 and shows that the
products 3a, 3b and 3c are thermodynamically more stable
than the starting material, 2, as would be expected.
It is well accepted that the photo-Fries reaction proceeds
through radical pairs, formed by excitation of the starting
material to a S1(n,pp) state and subsequent homolytic
cleavage of the C–O bond.6,11 Our theoretical calculations
confirm the n,pp character of the S1 state of 2. DFT-TD,
PM3 and ZINDO/S calculations (these last two at a
configuration interaction level), show that this electronic
transition arises from an overlap of excited states involving
the carbonyl group and the dibenzofuran ring. As expected,
the TD-DFT calculations furnished a more quantitative
description of the electronic transitions, presenting an
absorption line at 295 nm for the S0!S1 transition for the
isolated molecule, very near to the estimated maximum for
this transition observed in ethyl acetate (Fig. 8).
3.2. Photochemical rearrangement
In contrast to the catalysed dark reaction in which, reaction
at position 1 is inhibited by the catalyst, the photochemical
rearrangement has no such inhibition and reaction at this
position is preferred. This must be due, at least in part, to its
higher electronic charge density compared to position 3 (see
Fig. 7). The most efficient reaction occurred with dichloro-
methane as the solvent, when the product 3c was isolated in
41% yield after 300 min exposure to UV light (Table 2).
The product distributions obtained in ethanol and cyclo-
hexane showed slightly lower yields. In both solvents the
maximum yield of 3c was around 30%. However, the
reaction time, in cyclohexane, needed to reach this yield was
about 4.2 times longer than in ethanol. Despite the longer
irradiation period needed to reach a yield of 3c in
cyclohexane comparable to that obtained in ethanol after
300 min, secondary reactions of 3a, 3c, and even 2, were
less prevalent than in other solvents. As solvent polarity
increases, there is an increasing tendency to form other
products, including secondary photoproducts from 3a and
(especially) from 3c (Figs. 3 and 4). This can be explained in
terms of increased stabilization of the radical species in
more polar solvents, thus favouring the formation of
rearrangement products within the solvent cage.6 Further-
more, the additional stabilization of the dibenzofuranoxy
radical, in a polar solvent will increase the probability of this
radical escaping from the solvent cage and undergoing
hydrogen abstraction to give 1. As can be seen (Table 2),
polar solvents favour this diffusion product. This is in
The five lines presented in Figure 8 correspond to the first
five electronic transitions for 2, and the band corresponding
to the S0!S1 transition appears as a shoulder in the
experimental spectrum, overlapped by the band correspond-
ing to the S0!S2 transition. The relatively low calculated
oscillator strength (0.18 for the isolated molecule) is
consistent with the n!pp character of this transition.
4. Conclusions
The Fries rearrangement of dibenzofuran-2-yl ethanoate (2)
as a route to o-hydroxyacetyldibenzofurans has been
investigated, both under thermal Lewis-acid (AlCl3 and
TiCl4) catalysed and non-catalysed photochemical con-
ditions. Reaction products have been isolated and charac-
terized, and the effects of temperature, solvent, type of acid,
concentration and time on product distributions have been
investigated.
Figure 8. UV–Vis spectrum of dibenzofuran-2-yl ethanoate (2) in ethyl acetate (1£1025 mol dm23). Vertical lines represent calculated electronic transitions
for the isolated molecule (TD-DFT).

6152
A. M. A. G. Oliveira et al. / Tetrahedron 60 (2004) 6145–6154
The efficiency of the Lewis-acid catalysed Fries reaction is
critically dependent on choice of reaction conditions. The
reaction proceeds best at temperatures above about 130 8C,
and in the absence of solvent. The principal rearrangement
products are 2-hydroxydibenzofuran-3-yl methyl ketone
(3a) and 8-hydroxydibenzofuran-2-yl methyl ketone (3b).
The formation of 3a possibly occurs by a concerted
intramolecular mechanism, mediated by the Lewis’ acid,
whereas formation of 3b can still be regarded as
intramolecular, but will involve discrete separation of the
acylium cation from the complexed dibenzofuranoxide
anion within the solvent cage. Minor products are also
observed where the separated acylium ion takes part in
intermolecular reactions. Steric effects caused by the
association between the catalyst and the aryloxy group
help explain why 2-hydroxydibenzofuran-1-yl methyl
ketone (3c) is not formed as the dominant product. A
theoretical evaluation of the reaction coordinate, using the
PM3 method, shows that the activation energy for
rearrangement is about 306 and 135 kJ/mol for the
uncatalysed and catalysed processes.
Dibenzofuran-2-yl ethanoate (2) was obtained as colourless
crystals (3.35 g, 94%); mp: 113.5–114.5 8C (ethanol) (lit.12
115–116 8C, n-propanol). UV: 284, (4.30). IR (Nujol): 1755
1
(CvO), 1224, 1158, 1112, 929, 892, 834, 745, 724. H
NMR: 7.91 (1H, br d, J¼7.8 Hz, H-9); 7.69 (1H, d,
J¼2.7 Hz, H-1); 7.58 (1H, br d, J¼8.1 Hz, H-6), 7.56 (1H,
d, J¼9.0 Hz, H-4); 7.49 (1H, dt, J¼1.5, 7.8 Hz, H-7); 7.35
(1H, dt, J¼0.9, 7.5 Hz, H-8); 7.17 (1H, dd, J¼2.4, 8.7 Hz,
H-3); 2.37 (3H, s, CH3). Anal. calcd for C14H10O3: C, 74.32;
H, 4.46%. Found: C, 74.04; H, 4.68%.
5.2. Lewis-acid catalysed reaction
A. With solvent. To a solution of dibenzofuran-2-yl
ethanoate (2) (0.23 g, 1.0 mmol) in dichloromethane
(10 ml) was added the Lewis acid (2.0 mmol of AlCl3,
solid, or TiCl4, 1 mol dm23, in CH2Cl2). The mixture was
stirred at room temperature or refluxed. Hydrochloric acid
(2 M) and crushed ice were added and the mixture was
stirred for 15 min. The aqueous layer was extracted with
CH2Cl2 (3£25 ml) and the combined organic extracts were
dried (MgSO4). The solvent was evaporated and a solid was
obtained. A solution of the final solid, in ethyl acetate, was
prepared and diluted to a known volume for HPLC analysis.
For the photo-Fries reaction, the major products are
2-hydroxydibenzofuran-1-yl methyl ketone (3c) and
2-hydroxydibenzofuran-3-yl methyl ketone (3a), with 3c
generally dominating. The yields of products in the
photochemical reaction are very sensitive to solvent polarity
and, as occurs with the dark reaction, the experimental data
show that there is a competition between intramolecular and
intermolecular mechanisms. The best preparative results
were obtained for very low polarity solvents. Photo-
rearrangement is also particularly slow in cyclohexane,
suggesting that polar solvents favour homolytic cleavage of
the ester to give the radical pair.
B. Without solvent. A mixture of dibenzofuran-2-yl
ethanoate (2) (0.45 g, 2.0 mmol) and AlCl3 (0.53 g,
4.0 mmol) was heated under argon for 15 min, at the
temperature indicated in Table 1. After cooling, crushed ice
and conc. HCl (5.0 ml) were added. The resulting mixture
was stirred for 15 min and extracted with diethyl ether
(3£30 ml). The combined organic extracts were dried
(MgSO4) and the solvent removed under reduced pressure.
A solution of the final solid, in ethyl acetate, was prepared
and diluted for analysis by HPLC.
The remaining solids of several experiments were combined
and submitted to column chromatography (silica, ethyl
acetate/light petroleum ether, 4–60%).
5. Experimental
5.1. Syntheses
5.2.1. 2-Hydroxydibenzofuran-3-yl methyl ketone (3a).
The title compound was the first compound eluted and was
obtained as intense yellow crystals, mp: 170.0–172.0 8C
(ethanol) (lit.12 168–169 8C, ethanol/n-propanol/water).
UV: 365 (3.71), 315 (4.44), 280 (4.01). IR (KBr): 1650,
1631 (strong, CvO), 1609, 1590, 1459, 1426, 1370, 1328,
1218, 865, 817, 749, 660. 1H NMR: 12.25 (1H, s, OH); 7.95
(1H, br d, J¼7.8 Hz, H-9); 7.90 (1H, s, H-4); 7.57–7.53
(1H, m, H-7); 7.48 (1H, s, H-1); 7.41–7.31 (2H, m, H-6 and
H-8); 2.74 (3H, s, CH3). Anal. calcd for C14H10O3: C, 74.32;
H, 4.46%. Found: C, 74.09; H, 4.70%.
Light petroleum refers to solvent boiling in the range
40–60 8C. Column chromatography (CC) was performed on
Merck silica gel 60 (230–400 mesh). Melting points were
determined on a Gallenkamp apparatus and are uncorrected.
Ultraviolet spectra were recorded in ethyl acetate on a
HITACHI 2000 and data are presented in lmax (nm), log 1
(mol21 dm3 cm21). Infrared spectra were recorded on a
Diffus-IR Bomem MB-Series FTIR spectrometer in cm21
.
1H NMR spectra were obtained on a Varian Unity Plus at
300 MHz and the assignments were based on irradiation
experiments. The solvent was CDCl3 (if not stated
otherwise) and d is in ppm, relative to internal SiMe4.
Elemental analyses were carried out with a LECO
CHNS-932.
5.2.2. Dibenzofuran-2-yl ethanoate (2). The second
compound eluted was obtained as colourless crystals; mp
and spectroscopic data were identical to those of an
authentic sample of 2.
5.1.1. Dibenzofuran-2-yl ethanoate (2). A mixture of
dibenzofuran-2-ol (1) (2.78 g, 15.1 mmol), pyridine (25 ml)
and acetic anhydride (7.1 ml, 75.3 mmol) was refluxed for
3 h. Crushed ice (100 g) was added and the mixture was
stirred until a formation of a beige precipitate is observed.
The mixture was filtered and the solid was recrystallized
from ethanol.
5.2.3. Dibenzofuran-2-ol (1). The third compound eluted
was obtained as beige crystals; mp: 133.5–134.5 8C (lit.13
134 8C, ethanol) UV: 324 (3.73), 290 (4.19). IR (KBr): 3272
(OH), 1600, 1482, 1447, 1361, 1334, 1308, 1282, 1214,
1191, 1168, 1150, 1017, 869, 842, 801, 743. 1H NMR: 7.89
(1H, br d, J¼7.5 Hz, H-9); 7.55 (1H, br d, J¼8.1 Hz, H-6);

A. M. A. G. Oliveira et al. / Tetrahedron 60 (2004) 6145–6154
6153
7.46 (1H, dt, J¼1.2, 8.2 Hz, H-7); 7.44 (1H, d, J¼9 Hz,
H-4); 7.38 (1H, d, J¼2.4 Hz, H-1); 7.32 (1H, dt, J¼1.2,
7.8 Hz, H-8); 6.97 (1H, dd, J¼2.4, 8.7 Hz, H-3); 5.01 (1H,
br s, OH).
were rotated about throughout the irradiation period.
Samples (0.5 ml) of the solutions were withdrawn at
different photolysis times for HPLC analysis, until the
concentration of the main products started to decrease after
reaching a maximum value.
5.2.4. 2-Hydroxydibenzofuran-1-yl methyl ketone (3c).
The fourth compound eluted was obtained as pale yellow
crystals, mp: 154.5–155.5 8C (ethanol/hexane) (lit.12 105–
110 8C). UV: 302 (4.02). IR (KBr): 3177 (OH), 1654
(CvO), 1592, 1454, 1432, 1368, 1277, 1263, 1079, 889,
809, 739, 730. 1H NMR: 11.25 (1H, s, OH); 8.03 (1H, br d,
J¼8.4 Hz, H-9); 7.69 (1H, d, J¼9.0 Hz, H-4); 7.62 (1H, br
d, J¼8.4 Hz, H-6); 7.51 (1H, dt, J¼1.2, 7.2 Hz, H-7); 7.36
(1H, dt, J¼1.2, 8.4 Hz, H-8); 7.12 (1H, d, J¼9.0 Hz, H-3);
2.91 (3H, s, CH3). Anal. calcd for C14H10O3: C, 74.32; H,
4.46%. Found: C, 74.52; H, 4.81%.
In order to study the effect of concentration, several
ethanolic solutions of 2 of different concentrations
(5.0£1023 mol dm23, 2.0£1023 mol dm23 and 1.0£1023
mol dm23) were also irradiated.
5.4. HPLC analyses
High performance liquid chromatography (HPLC) analyses
of the reaction mixtures were carried out using a JASCO
PU-980 pump with a RHEODYNE—7725i (20 ml) loop
valve, a JASCO UV-975 UV–Vis variable wavelength
detector, without scanning capability, and a Shimadzu
C-R6A Chromatopac recorder. The column was a Merck
LichroCART, 250£4 mm2 (Lichrospher Si 60, 5 mm). The
analyses were conducted at constant flow rate (1.6 ml/min),
with monitoring at l¼290 nm and with 0.64 AUFs. All the
samples were prepared in ethyl acetate and 5 ml aliquots
were injected for each analysis. The standards were
obtained by column chromatography and solutions of
different concentrations were prepared in order to obtain
an external calibration from peak height plotted as a
function of the concentration for each compound. The
mobile phase was a mixture of ethyl acetate/hexane of
analytical grade, in the proportion 20:80.
5.2.5. 8-Hydroxydibenzofuran-2,7-diyl dimethyl
diketone (3e). The fifth compound eluted was obtained as
yellow crystals, mp: 224.5–226.0 8C (ethanol). UV: 365
(3.70), 275 (4.67). IR (KBr): 1679 (CvO), 1651, 1628
(strong, CvO), 1423, 1359, 1324, 1239, 1213, 1128, 944,
823, 785. 1H NMR: 12.27 (1H, s, OH), 8.57 (1H, d,
J¼2.1 Hz, H-1); 8.20 (1H, dd, J¼1.8, 8.7 Hz, H-3); 7.94
(1H, s, H-6); 7.60 (1H, d, J¼8.7 Hz, H-4); 7.54 (1H, s, H-9);
2.76 (3H, s, 3-COCH3 or 9-COCH3); 2.73 (3H, s, 9-COCH3
or 3-COCH3). Anal. calcd for C16H12O4: C, 71.63; H,
4.52%. Found: C, 71.71; H, 4.67%.
5.2.6. 8-Acetyldibenzofuran-2-yl ethanoate (3d). The
sixth compound eluted was obtained as yellow crystals,
mp: 145.0–147.0 8C (ethyl acetate/hexane). UV: 292 (4.04),
249 (4.52). IR (Nujol): 1745 (CvO), 1672 (CvO), 1634,
5.5. Theoretical calculations
1
1596, 1287, 1245, 1216, 1155, 1015, 904, 823, 807. H
The quantum-mechanical calculations were carried out at
the semi-empirical (PM3 and ZINDO/S) and density
functional theory (DFT) levels. The DFT B3LYP method
was employed, using a Gaussian basis-function (6-31G*), to
refine the structure of the compound dibenzofuran-2-yl
ethanoate (2) after its modelling using the PM3 method
NMR: 8.52 (1H, d, J¼1.2 Hz, H-9); 8.12 (1H, dd, J¼2.1,
8.7 Hz, H-7); 7.73 (1H, d, J¼2.4 Hz, H-1) 7.59 (1H, d,
J¼8.4 Hz, H-6); 7.57 (1H, d, J¼8.7 Hz, H-4); 7.22 (1H, dd,
J¼2.4, 9.0 Hz, H-3); 2.71 (3H, s, COCH3); 2.38 (3H, s,
OCOCH3). Anal. calcd for C16H12O4: C, 71.63; H, 4.52%.
Found: C, 71.61; H, 4.75%.
˚
(UHF calculation, gradient 0.1000 kcal/A mol, Polak-
Ribiere optimization algorithm)14 and to evaluate the charge
distribution in the ground state of the 2-dibenzofuranoxide
anion and radical.
5.2.7. 8-Hydroxydibenzofuran-2-yl methyl ketone (3b).
The last compound to be eluted was obtained as brownish
yellow crystals, mp: 223.5–225.5 8C (ethyl acetate/hexane).
UV: 326 (3.70), 248 (4.62). IR (Nujol): 3193 (OH), 1660
(CvO) 1596, 1580, 1304, 1286, 1263, 1185, 1169, 1113,
The Berny analytical gradient was used in the optimization
using DFT. The requested convergence limit on RMS
density matrix was 1£1028 and the threshold values for the
maximum force and the maximum displacement were
0.000450 and 0.001800 au, respectively.
1
1019, 864, 824, 723. H NMR (d6-acetone): 8.73 (1H, d,
J¼2.1 Hz, H-1); 8.62 (1H, br s, OH); 8.18 (1H, dd, J¼1.8,
9.0 Hz, H-3); 7.68 (1H, d, J¼9.0 Hz, H-4); 7.64 (1H, d,
J¼2.4 Hz, H-9); 7.54 (1H, d, J¼9.0 Hz, H-6); 7.11 (1H, dd,
J¼2.4, 9.0 Hz, H-7); 2.72 (3H, s, CH3). Anal. calcd for
C14H10O3: C, 74.31; H, 4.46%. Found: C, 74.32; H, 4.50%.
Using time-dependent DFT calculations (B3LYP/6-31G*),
the electronic spectrum of the compound dibenzofuran-2-yl
ethanoate (2) was predicted15 and compared with experiment.
5.3. Photochemical experiments
The reaction coordinates for the formation of the principal
product of the dark—Fries reaction under the catalysed and
the uncatalysed conditions were calculated using PM3.16
PM3 was also used to estimate the standard DHf for the
starting reagent and all isolated products.
The photochemical experiments were conducted using a
16 W low-pressure mercury lamp (emission at 254 nm)
positioned in the center of a merry-go-round set-up
(Annular Photoreactor, Model APQ 40—PhotoChemical
Reactors Limited). Solutions of dibenzofuran-2-yl ethano-
ate (2) (5.0£1023 mol dm23), in the appropriate solvent,
were sealed in stoppered quartz tubes (23 ml, 1 cm
diameter), and placed at 4 cm from the lamp. The samples
The methods used are available in HYPERCHEM 5.11
Pro,14 AMPAC 6.56 PC16 and GAUSSIAN 98W,15 suites of
programs, installed in PC-compatible computers.

6154
A. M. A. G. Oliveira et al. / Tetrahedron 60 (2004) 6145–6154
8. Anderson, J. C.; Reese, C. B. Proc. Chem. Soc., London 1960,
217.
Acknowledgements
ˆ
9. Vogel, A. I. Vogel’s Textbook of Practical Organic Chemistry;
5th edn. Furniss, B. S., Hannaford, A. J., Smith, P. W. G.,
Tatchell, A. R., Eds.; Longman Group UK: UK, 1989; p 1248.
10. Plank, D. A. Tetrahedron Lett. 1968, 5423.
The authors thank Fundac¸a˜o para a Ciencia e Tecnologia,
Portugal for financial support through IBQF-UM and for a
scholarship to A.M.A.G.O. (PRAXIS XXI/BD/19707/99),
and Elisa Pinto for obtaining 1H NMR spectra and elemental
analyses. A.E.H.M. thank to FAPEMIG (Fundac¸a˜o de
11. Dickerson, T. J.; Tremblay, M. R.; Hoffman, T. Z.; Ruiz, D. I.;
Janda, K. D. J. Am. Chem. Soc. 2003, 125, 15395–15401.
12. Gilman, H.; Cook, T. H.; Hogg, J. A.; Swiss, J.; Johnson, R. G.
J. Am. Chem. Soc. 1954, 76, 5783–5784.
`
Amparo a Pesquisa do Estado de Minas Gerais — Brasil)
and CNPq (Conselho Nacional do Desenvolvimento
´ ´
Cientıfico e Tecnologico — Brasil) for financial support
and research grants.
13. Gilman, H.; Ess, P. R. V. J. Am. Chem. Soc. 1939, 61,
1365–1371.
14. HyperChem 5.11 Pro Computational Chemistry, Hypercube,
USA, 1999.
References and notes
15. Gaussian 98, Revision A.11.4 Frisch, M. J., Trucks, G. W.,
Schlegel, H. B., Scuseria, G. E., Robb, M. A., Cheeseman,
J. R., Zakrzewski, V. G., Montgomery Jr., J. A., Stratmann,
R. E., Burant, J. C., Dapprich, S., Millam, J. M., Daniels, A. D.,
Kudin, K. N., Strain, M. C., Farkas, O., Tomasi, J., Barone, V.,
Cossi, M., Cammi, R., Mennucci, B., Pomelli, C., Adamo, C.,
Clifford, S., Ochterski, J., Petersson, G. A., Ayala, P. Y., Cui,
Q., Morokuma, K., Rega, N., Salvador, P., Dannenberg, J. J.,
Malick, D. K., Rabuck, A. D., Raghavachari, K., Foresman,
J. B., Cioslowski, J., Ortiz, J. V., Baboul, A. G., Stefanov,
B. B., Liu, G., Liashenko, A., Piskorz, P., Komaromi, I.,
Gomperts, R., Martin, R. L., Fox, D. J., Keith, T., Al-Laham,
M. A., Peng, C. Y., Nanayakkara, A., Challacombe, M., Gill,
P. M. W., Johnson, B., Chen, W., Wong, M. W., Andres, J. L.,
Gonzalez, C., Head-Gordon, M., Replogle, E. S., Pople, J. A.,
Gaussian, Inc., Pittsburgh PA, 2002.
1. Oliveira, A. M. A. G.; Raposo, M. M. M.; Oliveira-Campos,
A. M. F.; Griffiths, J.; Machado, A. E. H. Helv. Chim. Acta
2003, 86(8), 2900–2907.
2. Cui, C.; Wang, X.; Weiss, R. J. Org. Chem. 1996, 61,
1962–1974.
3. Bonesi, S.; Erra-Balsells, R. J. Photochem. Photobiol. A:
Chemistry 1997, 110(3), 271–284.
4. Gerecs, A. Friedel-Crafts and Related Reactions; Olah, G. A.,
Ed.; Interscience: New York, 1964; Vol. III. Chapter XXXIII.
5. Kalmus, C. E.; Hercules, D. N. J. Am. Chem. Soc. 1974, 96(2),
449–456.
6. Haga, N.; Takayanagi, H. J. Org. Chem. 1996, 61, 735–745.
Park, K. K.; Lee, H. J.; Kim, E. H.; Kang, S. K. J. Photochem.
Photobiol. 2003, 159, 17–21.
7. Sandner, M. R.; Hedaya, E.; Trecker, D. J. J. Am. Chem. Soc.
1968, 90(26), 7249–7254.
16. AMPAC Version 6.56, Semichem, Inc., USA, 1999.