Microwave accelerated facile synthesis of fused polynuclear hydrocarbons in dry media by intramolecular Friedel-Crafts alkylation

Article abstract of DOI:10.1039/b311750m

Fused polynuclear tetrahydro arenes are synthesised in a fast, simple, high yielding and regiospecific procedure by an intramolecular Friedel-Crafts alkylation in dry media under microwave irradiation of the corresponding 1-bromo-4-arylbutanes immobilised on silica. The observed reactivity is rationalised by molecular modelling studies which suggest the occurence of a concerted mechanism.

Full text of DOI:10.1039/b311750m

View Article Online / Journal Homepage / Table of Contents for this issue
Microwave accelerated facile synthesis of fused polynuclear
hydrocarbons in dry media by intramolecular Friedel–Crafts
alkylation
a,b
b
b
Vanya B. Kurteva, António Gil Santos and Carlos A. M. Afonso*
a
Institute of Organic Chemistry, Bulgarian Academy of Sciences, Acad. G. Bonchev str.,
Block 9, 1113 Sofia, Bulgaria
b
REQUIMTE/CQFB, Departamento de Química, Faculdade de Ciências e Tecnologia,
Universidade Nova de Lisboa, 2829-516 Caparica, Portugal.
E-mail: cma@dq.fct.unl.pt; for modelling aspects; ags@dq.fct.unl.pt; Fax: ϩ 351 21 2948550;
Tel: ϩ 351 21 2948358
Received 23rd September 2003, Accepted 10th December 2003
First published as an Advance Article on the web 21st January 2004
Fused polynuclear tetrahydro arenes are synthesised in a fast, simple, high yielding and regiospecific procedure
by an intramolecular Friedel–Crafts alkylation in dry media under microwave irradiation of the corresponding
1
-bromo-4-arylbutanes immobilised on silica. The observed reactivity is rationalised by molecular modelling
studies which suggest the occurence of a concerted mechanism.
Introduction
Organic reactions performed under microwave irradiation
(
MWI) have some significant advantages in comparison with
the conventionally heated ones, such as much shorter reaction
1
times, better selectivities and higher yields. The use of a clean,
efficient and economical solvent-free procedure combined with
an immobilisation on solid support reagents technique, which
gives good dispersion of the active reagent sites, associated
selectivity and easier work-up, is an environmentally benign
procedure preventing release of reaction residues into the
environment.
Scheme
1
General method used for the preparation of fused
polynuclear hydrocarbons 3. i) n-BuLi (1.5 eq.), diethyl ether; α,ω-
dibromoalkane (4 eq.); ii) silica gel (2 g per mmol of bromide 2), MWI
2
(
800 W), 15 min.
followed by reaction with an excess of α,ω-dibromoalkane
using standard conditions.
6
Polynuclear hydroaromatic hydrocarbons are well known
3a
and widely investigated compounds.
They have been
The bromides 2 were impregnated onto the silica gel support
by prior preparation of dichloromethane solutions followed
by subsequent solvent removal, and then were submitted to
microwave irradiation in a domestic household oven (“dry
media” conditions). Initially, this transformation was studied
under different experimental conditions using 1-bromo-4-(9-
phenanthrenyl)butane 2a as a substrate model (Table 1). An
increase in the MW power gives higher conversion (entry 1, 600
W, 15% vs entry 7, 800 W, 82% and entry 4, 700 W, 61% vs entry
8, 800 W, 94%) and no reaction was observed when using basic
alumina, instead of silica (entry 1 vs entry 2). We also studied
the effect of several additives immobilised on the silica support:
while the basic salts K CO or AcONa gave a considerable
synthesised mainly by partial hydrogenation of polycyclic
arenes, as complicated mixtures of regioisomers of products
with different degrees of hydrogenation. Other methods
included the intramolecular Friedel–Crafts acylation followed
3b–c
by carbonyl reduction.
On the other hand, the reported
examples of intramolecular Friedel–Crafts alkylation have
3d–i
been mainly restricted to benzene and naphthalene.
Additionally, some photochemically irradiated transformations
of arylbromoalkanes have been reported on the model of
phenylbromobutane, where the cyclisation product, tetra-
hydronaphthalene, has been isolated in less than 20%, parallel
4
with reduction and elimination products.
2
3
In the course of our attempt to prepare the 4-(9-phen-
anthrenyl)butyl acrylic ester by nucleophilic substitution of
the corresponding alkyl bromide 2a with acrylic acid on a silica
gel support, assisted by a microwave (MW), we observed
instead the exclusive formation of the polynuclear hydrocarbon
reduction on the conversion (entry 11, 46% and entry 12, 50%
vs entry 8, 94%), the acid CF CO H originated only a small
3
2
reduction on the conversion (entry 13, 89%). On the other
hand, the radical inhibitor BHT as well as hexabutylditin also
effected a considerable reduction on the conversion (entry 9,
71% and entry 10, 63% vs entry 8, 94%). In contrast to the high
conversion using MWI on substrate 2a (Table 1, entry 8, 94%),
no reaction was observed by submitting the bromide 2a to
vigorous thermal conditions (138 ЊC, 6 h, Table 2, entry 1), or
on the attempt to effect the intermolecular assisted MW
Friedel–Crafts alkylation of phenanthrene with 1,4-dibromo-
butane (Table 2, entry 5). Interestingly, no reaction was
observed for the substrates 2b and 2c containing the propane
and pentane alkyl chains, respectively, or for substrate 2d (Table
5
3
a in high yield (91%). With such unexpected result in hand, we
studied this transformation in more detail. Herein we report a
simple, high yield and regiospecific synthesis of polycyclic
hydrocarbons from the corresponding aryl alkyl bromides,
using solid-supported reagents, in a microwave accelerated
solventless procedure (Scheme 1), with the reactivity being
rationalised by molecular modelling.
Results and discussion
2
, entries 2–4) which is in contrast to the high conversion
The starting 1-bromo-n-arylalkanes 2 were prepared from the
corresponding aryl bromide 1, by lithium–bromide exchange
observed for the butane alkyl chain 2a under similar experi-
mental conditions (Table 1, entry 8). In the case of ester 4
5
14
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 5 1 4 – 5 2 3
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4

View Article Online
a
Table 1 Conversion of 2a into 3a under different experimental conditions
b
c
Entry
Power/W
Time/min
Solid support
Additive
Conv. (%)
1
2
3
4
5
6
7
8
9
600
600
600
700
700
700
800
800
800
800
800
800
800
10
10
10
15
15
15
10
15
15
15
15
15
15
Silica
No
No
15
f
Basic alumina
Basic alumina
Silica
Silica
Silica
Silica
Silica
Silica
Silica
Nonef
K CO
None
61
2
3
No
MeCO Na
Aliquat 336
No
No
BHT
(n-Bu Sn)
71
56
82
94
71
63
46
50
2
®
d
e
10
11
12
13
3
2
Silica
Silica
Silica
K CO
2
3
MeCO Na
CF CO H
2
89
3
2
a
The position of the reaction flask in the microwave oven was prior optimised and fixed for all the examples presented here. In all experiments were
®
b
c
1
d
used 2.0 g of solid support/mmol of bromide 2a. When applied, was used 1.0 mol eq. of additive. Observed conversion by H NMR. Aliquat
3
e
f
1
36 (tricaprylmethylammonium chloride). BHT (2,6-di-t-butyl-4-methylphenol). None of the product 3a was detected by TLC and by H NMR.
Table 2 Attempts to effect the Friedel–Crafts alkylation
Entry
Substrate
Conditions
Result
a
1
2
3
4
5
2a
Xylene, reflux, 6 h
MW, 800 W, 75 min
MW, 800 W, 75 min
MW, 800 W, 30 min
Starting materiala
Starting material
2b
a
2c
2d
Starting materiala
Starting materiala
Starting material
Phenanthrene
1,4-Dibromobutane, MW, 800 W, 15 min
a
1
No reaction was observed by TLC and H NMR.
no formation of the desired cyclised product was observed,
giving instead the corresponding acrylate 5 in moderate yield
neighbouring the alkyl side chain, in all substrates tested,
which is in an accordance with the lack of reactivity of the
9-anthracenyl bromobutane 2d under the same conditions.
The transformation occurred in high yields 72–97% (Table 3)
when fused polynuclear aromatic hydrocarbons were formed
(
Scheme 2).
(
3a, 3e–3h) while low yields were found for 1,2,3,4-tetrahydro-
naphthalene 3i (13%) and 5-phenyl-1,2,3,4-tetrahydronaphth-
alene 3k (14%), where the aromatic substituents in the starting
bromides were of a simple phenyl type. In contrast, complete
conversion of 2j to 3j was observed. Interestingly, the substrate
2
g gives only the isomer 3f. These results suggest that this trans-
formation is strongly dependent on the aromatic ring structure
and the size of the alkyl chain. This reaction is applicable as a
high yield procedure for the synthesis of fused aromatic hydro-
carbons, if the annulating center is activated by an alkyl or aryl
substituent in the ortho position. It should also be mentioned
that in all substrates tested the transformation is very clean,
without any traces of other products apart from recovered
starting material 2 and product 3.
The liberation of hydrogen bromide from this reaction
is an important issue from a synthetic and environmental point
of view. However, we observed that the by-product of the
reaction, hydrogen bromide, is well adsorbed on silica gel
(above 70%). The residual content of HBr on silica was
determined by titration with aqueous NaOH solution of an
aqueous suspension of both the crude reaction mixture and
the silica gel, after extraction of the organic products with
dichloromethane.
Scheme 2 Attempt to effect the Friedel–Crafts alkylation on the ester
a
1
b
4
. Observed conversion by H NMR. Isolated yield after purification
by flash chromatography.
The scope of this transformation was further evaluated for
different substrates under optimised conditions (Table 3). The
proposed structures of the fused aromatic hydrocarbons were
based on comparisons with reported spectral and analytical
data and were confirmed by 2D NMR (COSY, NOESY
and HMQC). In fact, for the correct assignment of the non-
equivalent benzylic ethylene protons and the aromatic protons,
which show the place of annulation, the NOESY cross peaks
are of great importance.
The investigated microwave accelerated cyclisation proceeds
regiospecifically via substitution of the aromatic proton
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 5 1 4 – 5 2 3
515

View Article Online
a
Table 3 MW assisted cyclisations of 4-aryl-1-bromobutanes 2
b
c
Substrate 2
Product 3
Conversion (%)
95
Yield (%)
91
a
e
Complete
97
f
94
91
96
72
g
h
3f
Complete
76
i
j
17
13
96
Complete
k
19
14
a
b
1
c
Silica gel (2 g per mmol of bromide 2), MW irradiation (800 W), 15 min. Observed conversion by H NMR. Isolated yield after purification by
flash chromatography.
The photochemistry of alkyl bromides and iodides, including
1
-bromo and 1-iodo-phenylbutane, have been studied in detail,
where it is assumed that, in general, the reaction occurs via
formation of radical and carbocation reactive intermediates,
4
respectively for iodides and bromides. Additionally, for the
Friedel–Crafts alkylation of aromatic halides or triflates
promoted by protic or Lewis acids, it is generally assumed that
the transformation occurs via the formation of a carbocation
7
intermediate. The experimental results described above for
this reaction, on the solid silica support under microwave
irradiation suggest that: i) The reaction should not occur via
a true carbocation intermediate because, in spite of the
observation that basic additives inhibit the transformation,
protic acids also do, however to a lesser extent and, if the
carbocation is formed it is expected that for the substrates
containing the propane 2b or pentane 2c alkyl chain and the
intermolecular attempt (phenanthrene and 1,4-dibromobutane,
Table 2, entry 5) some transformation of the starting bromide
to other products should occur, which was not the case. ii)
The true radical mechanism is also inconsistent, because low
conversion was observed in the presence of (n-Bu Sn) and, in
spite of considerable observed inhibition in the presence of
BHT, if the free radical intermediate was formed, it is expected
that the substrate 2b, containing the propane chain, should
also react, because of the more favourable 5-exo cyclization. On
the other hand, all results obtained are consistent with the
occurrence of a more concerted reaction of C–Br breaking and
C–Aryl bond formation, where the prior effective approach
should be crucial, as depicted in Scheme 3. The effect of the
Scheme 3 Possible mechanism for the cyclization. The energy of the ts
can be proportional to that of the carbocation.
additives could be rationalised from structural changes on
the active surface silica reaction environment, which effects the
stabilization of the reaction transition state.
The experimental results obtained clearly demonstrate that
the Friedel–Crafts alkylation of bromoarylalkanes, supported
in silica, under microwave irradiation, is extremely dependent
on the substrate structure. These results, motivated us to per-
form a theoretical study of such systems, with the aim of
understanding the reaction mechanisms. For that we used
several of the experimental structures, together with others we
needed for reasons which will be explained during the rest of
this discussion (Fig. 1).
3
2
5
16
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 5 1 4 – 5 2 3

View Article Online
Table 4 Relative energies, by comparison with one of the table entries, of the carbocations formed as possible intermediaries in the cyclization of
substrate 2 to product 3
Ϫ1
Ϫ1
Entry
2
3
Conversion (%)
∆G/kJ mol relative to j
∆G/kJ mol rel. to a, g or j
1
2
3
4
5
6
7
8
9
a
b
c
e
d
f
a
b
c
e
d
f
fЈ
f
g
n
h
i
94
Sm
Sm
Complete
Sm
94
—
Complete
—
—
76
Ϫ27.55
Ϫ12.24
Ϫ6.30
Ϫ21.23
Ϫ15.61
3.26
11.08
Ϫ17.59
11.90
7.36
1.52
41.60
0.00
29.04
68.52
62.11
0.00 a
15.31 a
21.25 a
6.32 a
20.85 g
28.67 g
0.00 g
29.49 g
24.95 g
f
g
g
n
h
i
j
k
l
1
1
1
1
1
1
1
0
1
2
3
4
5
6
17
41.60 j
0.00 j
29.04 j
68.52 j
62.11 j
j
k
l
Complete
19
—
—
m
m
Sm = starting material.
observed, one would expect elimination products, which were
never observed.
Other experimental data which makes us prefer the ionic S2
type mechanism, is the observation that the addition of ionic
substances, radical initiators or inhibitors does not seem to
change the reaction behaviour.
Our first approach in studying the S 2 type mechanisms, was

a calculation of the transition state structures (ts). However ts
structures are difficult to find and, usually, demand a large
amount of computer time. Due to this we took another
approach; in very similar structures, we can expect that the
energy of the carbocation formed after the cyclization will be
directly proportional to the energy of the respective transition
state (Scheme 3), since there is no kind of interaction between
the leaving bromide and the rest of the molecule. When we
perform the calculations in the gas phase, no ion stabilization
due to solvation is possible and, as a consequence, the energy of
the carbocation and the separated bromide will be much higher
than that of the transition state, where the charges partially
cancel themselves. Even so, if the energy increase due to charge
separation is similar in all structures, then the relative energies
of the carbocations should be proportional to the relative
energies of the transition states. The results† of our calcu-
lations are shown in Table 4.
Fig. 1 Additional compound structures used in the theoretical study.
A priori, we could imagine three different mechanisms which,
The values from Table 4 seemed to fluctuate so in order to
analyse them, we had to first make a few statements.
in turn, could be radical or ionic. The ionic ones could be S 1 or
S2 types. However, considering the experimental results, as
In spite of all the compounds being aromatic, we divided
the structures into benzene, naphthalene, phenanthrene,
anthracene or pyrene derivatives. Since the positive charge
is dissipated in the full aromatic system, we can expect that
for large systems the energy increase due to charge formation
would be less than for small molecules. This becomes even
more important if we think that there is no stabilization due
to solvation.
Since we wanted to make a qualitative comparison between
our theoretical results and the experimental observations (it
is not possible to make a kinetic study in the experimental con-
ditions we used, which means that we can only compare the
calculated energies with the experimental conversions), we
have to also consider that for some structures there are two
equivalent reactive carbons while, for others, there is only one.
Due to this statistical problem, after the same reaction time,

mentioned before, the radical or the ionic S 1 type mechanisms

should not occur, since, with one exception (compound 4), we
didn’t observe elimination products, even when the cyclisation
does not occur. In these situations, only initial material was
recovered. We have also to consider the medium where the
reaction is taking place. In fact when considering reactions in
solution, ionic mechanisms are in many cases possible due to
the strong stabilization arising from the ion solvation. Even so,
usually Friedel–Crafts alkylation or acylation reactions require
the use of strong Lewis acids, in order to promote the ion pair
formation. In our case, strong Lewis acids are not used and
strong solvation should not be occurring. This means that the
reaction should be an S 2 type mechanism, where the attacking

group approaches the aromatic ring, due to the high efficiency
of microwaves, to be converted into internal molecular energy.
The S 2 type mechanism could be either an attack on the α

†
All structure minimizations were calculated in the gas phase, using
22
carbon, as in Scheme 3, or an ipso attack, with the formation of
a 5 membered ring ts or intermediary, which would migrate
forming the final observed product. We dismissed this second
possibility since, in those cases where no cyclization was
Gaussian 98, and ab initio HF theory with a 6-31G** basis set. The
energies were calculated in the gas phase, as single points, using MP2
theory with a 6-31G** basis set and with thermal energy correction for
298.15 K and 1.00 atm.
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 5 1 4 – 5 2 3
517

View Article Online
Table 5 Transition state calculated energies (∆G‡ = ∆G Ϫ ∆G ϩ ∆GBrϪ
)
ts
reag
Ϫ1
Ϫ1
Ϫ1
2
3
Conv. (%)
∆G‡/kJ mol
∆G/kJ mol rel. to j
∆G/kJ mol rel. to a, g or j
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
a
b
c
e
d
f
a
b
c
e
d
f
fЈ
f
g
n
h
i
j
k
l
m
p
o
94
Sm
Sm
Complete
Sm
94
—
Complete
—
—
76
44.48
92.53
68.77
—
Ϫ10.87
37.18
13.42
—
18.00
7.37
19.84
Ϫ4.61
18.04
29.20
15.50
15.71
0.00
10.59
63.65
43.65
21.20
1.73
0.00 a
48.05 a
24.29 a
—
73.35
62.72
75.19
50.74
73.39
84.55
70.85
71.06
55.35
65.94
119.00
99.00
76.55
57.08
—
11.98 g
24.45 g
0.00 g
22.65 g
33.81 g
f
g
g
n
h
i
j
k
l
0
1
2
3
4
5
6
7
8
17
15.71 j
0.00 j
10.59 j
63.65 j
43.65 j
21.20 j
1.73 j
Complete
19
—
—
—
—
m
p
o
a
a
Sm = starting material. In case of the analog substrate 4 was observed the formation of the olefin 5 as the only reaction product.
even for similar reactivities, the yield for both types of
compounds can be considerably different.
Given that the intermediary approach does not seem to
perform very well, we decided to calculate the transition state
energies of all the derivatives in Table 4.
Even after these considerations, if one tries to analyse the
results in Table 4, using column 5, it is easy to conclude that a
direct comparison between different families of compounds is
not possible. In fact, we can see in entry 12 a conversion of 17%,
As stated before, we already assumed that the transition state
was of an ionic S 2 type. Nevertheless, after accepting that, we

discover another problem, which is the need for stabilization
of the ionic species formed along the reaction path. In fact,
one bromide and a proton are formed, at opposite sides of the
molecule, which will increase the transition state energy to
a value much higher than the real one, as it is exemplified in
Fig. 2. Because of this, we decided to use small molecules
or ions able to bind the transient ions in the transition state,
lowering their energy. Considering the medium where the
reactions are taking place, several possibilities were available,
but the one which most simplifies the calculation process, is the
use of bromide ion (Fig. 2). We believe that it is reasonable to
use this ion, as the formation of this species occurs during the
reaction. From a theoretical point of view, bromide is a bit
heavy, as we are worked at the ab initio level. On the other hand,
it does not increase the number of rotational or vibrational
degrees of freedom, like polyatomic molecules or ions would
do, and, on average, there was not a strong increase in the
computational demand. The results of our calculations‡ are
shown in Table 5.
Ϫ1
for a relative energy of 41.6 kJ mol , while the compound in
entry 2 does not react, even considering that the relative energy
Ϫ1
of its carbocation is Ϫ12.2 kJ mol . On the other hand, if we
analyse the results by families of compounds, as depicted in
column 6, it is possible to see a quite good agreement between
theory and practice.
For benzene derivatives we can see that structure 2j is
expected to be the most reactive one, as in fact it is. Structures
2
k and 2i show very similar experimental results, but the relative
energies of their respective carbocations have a difference of
about 11 kJ mol . Nevertheless, since structure 2i has two
Ϫ1
reactive carbons, the theoretical and experimental results can be
considered in good agreement.
For the naphthalene derivatives, the experimental results are
those in entries 6 to 8. Structure 2g seems to be the most
reactive one, in agreement with the theoretical results. In spite
Ϫ1
of its carbocation having near 21 kJ mol higher energy,
structure 2f is still very reactive. Structure 3fЈ would be formed
through an attack on the second aromatic ring. In practice this
was not observed, which is in agreement with the theoretical
results. However, the energy difference between the two possible
In contrast to the observations we made before for the calcu-
lation of carbocationic species, the values in columns 5 and 6 of
Table 5, now seem much more consistent. When we compare
Ϫ1
carbocations of structure 2f is only about 8 kJ mol . Structure
3
g would be formed from structure 2g, through an attack on
position 3, instead of position 1. The difference in energy of
Ϫ1
near 30 kJ mol explains why only compound 3f is formed.
For the phenanthrene derivatives, compound 2e is the experi-
mentally most reactive. The calculations say that it should be
structure 2a. In spite of this, the energy difference is small and
the result is acceptable. Structures 2b and 2c seem to not react at
all, in the experimental conditions we used. This result is in
agreement with the calculations, since one intermediary has
Ϫ1
around 15 kJ mol higher energy than that of intermediary of
Ϫ1
2
a, while the other has more than 21 kJ mol higher energy.
Fig. 2 Two possible ts structures for the cyclization. On the right side
is depicted the model used for all the calculations, which uses a second
bromide as a base for the proton removal.
Nevertheless it is interesting to compare the energies for
intermediaries 2b and 2c with those of intermediaries of 2l
and 2m, also with the formation of 5 and 7 membered rings
respectively. Not only the large absolute difference but also the
inversion of the relative energy differences (the formation of
the seven membered ring is more favourable in the benzene
derivative, while in the phenanthrene derivative it is the
opposite), show that this approach can be useful but only when
comparing very similar aromatic systems.
‡ All the transition states were minimized in the gas phase, using
2
2
Gaussian 98, at ab initio HF theory with a 3-21G* basis set. The
energies were calculated in the gas phase, as single points, at MP2
6
-31G** and thermal energy was corrected for 298.15 K and 1.00 atm.
The values were obtained after subtracting the energy of the transition
state structures from the energy of the reagents plus the energy of a
bromide ion.
5
18
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 5 1 4 – 5 2 3

View Article Online
entries 2 and 3 with entries 15 and 16 in Table 5, we can now see
that in both cases the energy needed for the formation of the
five membered ring is higher than that of the seven membered
ring. In Table 4, one can see that this was not the case.
From entry 17, Table 5, it is possible to conclude that
substrate 2p, which is a model of compound 4, is just on the
border between the reactive and the non-reactive molecules.
Since there is, in this compound, a second reaction path leading
to elimination products through a low energy transition state,
no cyclisation is observed for 4 but, instead, only the respective
olefin 5 was observed.
the experimental conditions we used, but at a much slower rate,
Ϫ1
with around a 16 kJ mol higher energy transition state, rel-
ative to structure 2j.
The HOMO orbital of compound 2k is shown in Fig. 4. It is
possible to see there is no density over carbon 6, which means
that the reacting orbital has to be a lower energy one. In this
case it is the HOMO-1, which has a coefficient of 0.24 over
carbon 6. In spite of being a high coefficient, this orbital is not
as reactive as the HOMO would be for the same coefficient.
Even so, considering the transition state energy of this
Ϫ1
compound (around 10.6 kJ mol higher than the ts energy of
Entry 18 in Table 5 corresponds to the transition state of a
secondary bromide. In practice we tested this reaction with
a phenanthrene derivative (entry 4, Table 5), but the transition
state we obtained was of low quality. Even for the ts in entry 18
the structure was not the best, as indicated by the imaginary
structure 2j), the low reactivity will be a result of its electronic
configuration, of having only one reactive carbon and also, due
to the presence of the large substituent, difficulty in the rotation
of the side chain.
Ϫ1
Ϫ1
frequency, of only around 90 cm , against more than 200 cm
for all the other transition states found. This happens because
the potential energy surface is extremely flat for compound 2o
and, even more, for compound 2e, making it very difficult to
find good transition structures in these situations. After follow-
ing a reaction path for structure 2o, we believe that the energy
Ϫ1
shown in entry 18 should be increased to ∼3 to 5 kJ mol . Even
so, we can conclude that the presence of the methyl group
does not increase the steric hindrance but, on the other hand,
stabilizes the positive charge formed during the reaction path
(
as soon as the C–Br bond length increases), lowering the
Ϫ
Ϫ3
Fig. 4 HOMO (left) and HOMO-1 (right) orbitals (0.032 e au ) for
structure 2k (MP2/6-31G**//HF/6-31G**).
transition state energy.
In spite of now being able to postulate the behaviour of these
types of reaction, it is also very important to understand the
differences or similarities in that behaviour. For compounds like
those in entries 12, 15 and 16 (Table 5), it is easy to understand
the relative behaviour, based on different ring strains. The five
membered ring is the most energy demanding, while the six
membered ring is the easiest one to be formed. But, when we
compare for instance, the compounds in entries 12, 13, 14 or
even 17 where all the transition states have six member rings, an
electronic analysis§ is necessary in order to understand the
observed differences.
In principle the reaction should take place between the
HOMO and the LUMO of the same molecule but, as we are
dealing with large conjugated systems, it is possible, as we will
see for some cases, that higher or lower energy orbitals have to
be considered.
Ϫ1
Compound 2p has a ts energy of around 21.2 kJ mol ,
relative to compound 2j, which is on the border of reacting/
non-reacting. Its HOMO orbital has a shape similar to that of
compound 2i, with equal coefficients over carbons 2 and 6
(0.15). Nevertheless, the anti-ligand orbital which can react is,
in this case, the LUMO ϩ 2, since both the LUMO and LUMO
ϩ 1 have very low coefficients over the carbon connected to the
bromine atom. Together with the higher ring strain due to the
strong planarity of the transition state, these factors render
a low reaction rate, associated with higher ts energy, when
compared with compound 2i (Fig. 5).
Starting with compound 2j from entry 13, we can see that the
HOMO orbital is located over carbons 2 and 6, the atoms
which will suffer the attack (Fig. 3). In these carbons the
HOMO coefficients are of around 0.25 and this structure has a
high reactivity and a low transition state energy.
Fig. 5 Comparison of the transition state structures of compound 2p
(
left) and 2i (right).
Another family of compounds which has to be studied
considering the electronic aspects, is the naphthalene deriv-
atives. If we compare the transition state structures in Fig. 6, we
can’t see any important geometrical or steric differences
explaining the reactivity. However, if we analyse the HOMO
orbitals of the two possible derivatives (Fig. 7), we can easily
understand the different behaviours.
For both isomers, the HOMO orbital has larger coefficients
over carbons 1, 4, 5, and 8. When the substituent is in the ring
position 1 (structure 2f ), the coefficient of carbon 1 is 0.21,
of carbon 2 is 0.15, of carbon 3 is 0.15 and of carbon 8 is 0.20.
When the substituent is in ring position 2 (structure 2g), the
coefficient of carbon 1 is 0.22, of carbon 2 is 0.17, of carbon 3 is
0.12 and of carbon 8 is 0.21. This large difference in the HOMO
Ϫ
Ϫ3
Fig. 3 HOMO orbitals (0.032 e au ) for structures 2j (left) and 2i
(
right) (MP2/6-31G**//HF/6-31G**).
If we compare the HOMO of compound 2i, entry 12, we can
see that the orbital is located also over carbons 2 and 6, but the
coefficients are much lower (0.15) (Fig. 3). Higher coefficients
are found in carbons 1 (0.26) and 4. This structure still reacts in
2
2
§
All structures were minimized in the gas phase, using Gaussian 98,
at ab initio HF theory with a 6-31G** basis set. The surfaces were
obtained from single point calculations, in the gas phase, at MP2
-31G**, using Spartan Ј02 for Macintosh, v.1.0.3.
2
3
6
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 5 1 4 – 5 2 3
519

View Article Online
obtained for the benzene or naphthalene derivatives, does not
explain the theoretically expected and experimentally observed
higher reactivity. Also, since the phenanthrene and the naphth-
alene derivatives have only one reactive site for each isomer (all
structures have more than one possible reactive carbon but not
equivalents), opposite to some of the benzene derivatives, to get
similar final yields they must have lower energy transition states.
The ts energies agree with the experimental results, but they
don’t explain why. We believe that the differences between each
family of compounds is due to considerable differences in the
energy gap between the reactive ligand and the reactive anti-
ligand orbitals (as we said before, they are not always the
HOMO and the LUMO). Compound 2p has an energy gap
Ϫ1
of 1294 kJ mol . The average of the benzene derivatives
Ϫ1
(
excluding compound 2p) has an energy gap of 1210 kJ mol .
The average of the naphthalene derivatives has an energy gap
Ϫ1
of 1163 kJ mol . Finally, the phenanthrene derivative has
Ϫ1
an energy gap of 1115 kJ mol . These values are in good
agreement with the experimental results and with the ts
calculated values.
To finish this discussion we would like to point out a few more
aspects concerning some special characteristics of the transition
states, which can shed a bit more light over the total mechanism.
As we said in the beginning of our modelling discussion, we
used a second bromide ion in order to simulate a solvent effect,
reducing the transition state energy. Otherwise the charge
separation would enormously increase the energy, as seen
in Fig. 2 for the simplest benzene derivative. However, as the
reactions were performed in the solid state, the diffusion
of molecules or ions has to be slow, making it difficult for a
bromide ion to reach a substrate molecule and act as a base. Of
course, we can think about the solid matrix acting as a base,
abstracting the proton and lowering the ts energy. But, if one
looks at all the transition states we have calculated including
those in Fig. 8 we notice that the electrophilic carbon has, in the
transition state, a geometry away from planarity with valence
angles of around 100 degrees. Also, the bromine–carbon bond
Fig. 6 Three ts structures for the cyclization reaction of naphthalene
derivatives. 2g to 3g (top left); 2g to 3f (top right) and 2f to 3f (bottom).
Ϫ
Ϫ3
Fig. 7 HOMO orbitals (0.032 e au ) for structures 2f (left) and 2g
right) (MP2/6-31G**//HF/6-31G**).
(
Ϫ1
distribution can explain the difference of nearly 23 kJ mol
between the two possible transition states of compound 2g,
indicating that it should react preferentially through carbon 1,
as is experimentally observed. For isomer 2f, we would expect,
from the analysis of the HOMO distribution, that carbon 8
should be preferentially attacked. However, this is not what the
transition state energy or the experimental results suggest. In
this case, the main factor controlling the ts energy is the ring
strain, due to the ring size and to the three coplanar carbon
atoms (Fig. 8). The ring strain is even higher when a six
membered ring is formed (structure 3n, ts in Fig. 8). In this case,
and also due to the three coplanar carbon atoms, the transition
state adopts a boat conformation, while the transition states in
Fig. 6 all have chair conformations.
8
has a length of about 3.1 Å, while it should be around 2.5 Å.
These observations, together with the very flat transition state
energy surfaces, as stated before, give the indication that, in the
transition state, the bromine atom bond is almost ionic. This
means that the bromine atom can rotate around the substrate
(
with some limitations), without changing the potential energy
much. In this situation, we would suggest that the cyclisation
can occur as a uni-molecular process, where the electrophilic
carbon would approach the aromatic ring, followed by a
rotation of the bromine to a position where it can abstract the
hydrogen atom.
Conclusions
The study presented here clearly demonstrates that the intra-
molecular Friedel–Crafts alkylation of 4-aryl-1-bromobutanes
immobilised on silica, under microwave irradiation, in dry
media, is a simple, clean and efficient method for the synthesis
of a range of fused polynuclear hydrocarbons. The range of
substrate structures tested and the diverse experimental con-
ditions used prompted us to perform a more complete
modelling study of this transformation, which allowed a
rationalisation of the observed peculiar dependence of the
transformation on the substrate structure and, also, additional
support for a concerted ionic mechanism.
The study presented here enables the anticipation that taking
into consideration the effect of both n-alkyl bromide chains and
aryl functionalities on this transformation, a large range of
other fused polynuclear hydrocarbons can be synthesised by
the simple methodology described here. Additionally from
an environmental point of view, liberation of the harmful
hydrogen bromide as a by-product is also minimised by being
mainly absorbed on silica.
Fig. 8 ts Structures for the attack of carbon 8 in naphthalene
derivatives. Formation of a seven membered ring (3fЈ, left) and
formation of a six member ring (3n, right).
The phenanthrene and the anthracene derivatives are
identical cases to those of benzene and naphthalene structures
and do not need to be discussed in detail. The phenanthrene
derivative 2a has a HOMO orbital with a high density (0.21)
over carbon 10, the reacting carbon. This value, similar to those
5
20
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 5 1 4 – 5 2 3

View Article Online
Experimental
in COSY: 2.329/3.259, 2.329/3.477, 7.567/7.811, 7.567/8.627,
7
7
2
1
1
.622/8.069, 7.622/8.714, NOESY: 3.259/7.594, 3.259/8.069,
General remarks
13
.594/7.811, 8.627/8.714; C NMR: 31.55 (CH -3), 32.83 (CH -
2
2
), 33.55 (CH -1), phenanthrenyl CH-carbons 122.43 (CH-4),
Diethyl ether was pre-dried over sodium wire and distilled
from sodium/benzophenone under a nitrogen atmosphere.
Dichloromethane was distilled over calcium hydride powder
under a nitrogen atmosphere. Flash chromatography was
carried out using silica MN-Kiesel-gel 60M gel (230–400 mesh
ASTM, Art. 815381). All eluents were distilled prior to use. The
NMR spectra were recorded on a Bruker AMX 400 in deutero-
chloroform; chemical shifts (δ) were quoted in ppm against
tetramethylsilane (TMS) and coupling constants in Hz. Infra-
red spectra (IR) were recorded on a Mattson FTIR Satellite
2
23.30 (CH-5), 124.21 (CH-8), 126.17 (CH-7), 126.27 (CH-6),
26.68 (CH-2, CH-3, CH-10), 128.09 (CH-1).
1
0
1
-Bromo-5-(9-phenanthrenyl)pentane (2c). 43% Yield; mp
1
7
1
4–75 ЊC (lit. 72–74.5 ЊC); H NMR: 1.542 (m, 2H, CH -3),
2
.767 (m, 2H, CH -4), 1.863 (m, 2H, CH -2), 3.042 (t, 2H, J 7.4,
2
2
CH -5), 3.347 (t, 2H, J 6.7, CH -1), phenanthrenyl protons
2
2
7
.510 (s, 1H, CH-10), 7.541 (overlapped signals, 2H, CH-2,
CH-3), 7.598 (overlapped signals, 2H, CH-6, CH-7), 7.778 (d,
H, J 6.7, CH-1), 8.025 (d, 1H, J 7.8, CH-8), 8.596 (d, 1H, J 7.0,
1
3
000 spectrometer as thinly dispersed films (from dichloro-
CH-4), 8.679 (d, 1H, J 6.8, CH-5); important cross peaks in
COSY: 1.542/1.767, 1.542/1.863, 1.767/3.042, 1.863/3.347,
methane). Melting points (uncorrected) were determined on
a Electrothermal Mod. IA 6304 capillary melting point
apparatus. Microanalyses were carried out at the Laboratório
de Análises do CQFB/FCT/UNL. The microwave irradiated
reactions (MWI) were carried out in a domestic household oven
Balay 3WM-2121.
7
3
2
3
1
1
.541/7.778, 7.541/8.596, 7.598/8.025, 7.598/8.679, NOESY:
13
.042/7.510, 3.042/8.025, 7.510/7.778, 8.596/8.679; C NMR:
8.34 (CH -3), 29.31 (CH -4), 32.72 (CH -2), 33.22 (CH -5),
2
2
2
2
3.80 (CH -1), phenanthrenyl CH-carbons 122.31 (CH-4),
2
23.12 (CH-5), 124.19 (CH-8), 125.82 (CH-2), 125.90 (CH-3),
26.00 (CH-10), 126.37 (CH-7), 126.47 (CH-6), 127.89 (CH-1).
The starting bromides and the cyclisation products were
identified by comparison of their physical and spectroscopic
data with those of commercially available compounds or those
known from the literature. The assignment of the signals in the
NMR spectra of the compounds was based on their character-
istic chemical shifts and the multiplicity of the signals and were
confirmed by 2D NMR techniques (COSY, NOESY, HMQC).
As there are two groups of similar signals, methylene groups
and aromatic protons, NOESY cross peaks are of great
importance to a correct assignment. The signals of the protons
of the methylene group, connected with the bromine atom, in
the starting bromides are in an area free of other signals which
provides the possibility of determining the conversions using
6c
1
-Bromo-4-(9-anthracenyl)butane
(2d). 42% Yield; mp
1
1
2
14.5–116 ЊC (lit. 115–116 ЊC); H NMR: 1.995 (m, 2H, CH -3),
2
.136 (m, 2H, CH -2), 3.503 (t, 2H, J 6.8, CH -4), 3.664 (t, 2H,
2
2
J 8.0, CH -1), anthracenyl protons 7.491 (overlapped signals,
2
2
8
H, CH-3, CH-6), 7.549 (overlapped signals, 2H, CH-2, CH-7),
.030 (d, 2H, J 8.0, CH-4, CH-5), 8.273 (d, 2H, J 8.8, CH-1,
CH-8), 8.365 (s, 1H, CH-10); important cross peaks in COSY:
1
3
.995/3.503, 2.136/3.664, 7.491/8.030, 7.549/8.273, NOESY:
13
.503/8.273, 8.030/8.365; C NMR: 26.90 (CH -3), 29.60
2
(
CH -4), 32.98 (CH -2), 33.46 (CH -1), anthracenyl CH-
2 2 2
carbons 124.17 (CH-1, CH-8), 124.79 (CH-3, CH-6), 125.55
1
H NMR spectra of the crude reaction mixtures.
(
CH-2, CH-7), 125.84 (CH-10), 129.23 (CH-4, CH-5).
6
ꢀ
-Arylbromoalkanes. General procedure
2-Bromo-5-(9-phenanthrenyl)pentane (2e). 47% Yield; mp
7
6–77 ЊC; IR: 705, 738, 896, 1265, 1422, 2305, 2945, 2987, 3054
To a stirred solution of arylbromide (10 mmol) in anhydrous
diethyl ether (20 mL) under an argon atmosphere, n-butyl-
lithium (1.5 eq., 1.6 M in hexane) was slowly added at 0 ЊC
followed by the addition of α,ω-dibromoalkane (4.0 eq.) which
was then refluxed for 2 hours. The reaction mixture was cooled
to room temperature, partitioned between diethyl ether (40 mL)
and water (30 mL). The aqueous phase was extracted with
diethyl ether (2 × 20 mL), the combined organic phase were
Ϫ1
1
cm ; H NMR: 1.722 (d, 3H, J 6.6, CH ), 1.981 (m, 2H,
CH -3), 2.058 (m, 2H, CH -4), 3.148 (m, 2H, CH -5), 4.211 (m,
3
2
2
2
1
H, CH-2), phenanthrenyl protons 7.581 (s, 1H, CH-10), 7.591
overlapped signals, 2H, CH-2, CH-3), 7.656 (overlapped
signals, 2H, CH-6, CH-7), 7.838 (d, 2H, J 7.1, CH-1), 8.093
d, 1H, J 6.8, CH-8), 8.658 (d, 1H, J 8.0, CH-4), 8.738 (d, 1H,
J 6.6, CH-5); important cross peaks in COSY: 1.722/4.211,
(
(
1
7
8
.981/4.211, 2.058/3.148, 7.591/7.838, 7.591/8.658, 7.656/8.093,
.656/8.738, NOESY: 3.148/7.581, 3.148/8.093, 7.581/7.838,
.658/8.738; C NMR: 26.48 (CH ), 28.23 (CH -4), 32.65
dried over MgSO , evaporated in vacuo and purified by flash
4
chromatography on silica gel using hexane as a mobile phase.
13
3
2
(
CH -5), 40.99 (CH -3), 51.46 (CH-2), phenanthrenyl CH-
1
0
2
2
1
-Bromo-4-(9-phenanthrenyl)butane (2a). 63% Yield; mp
1
carbons 122.42 (CH-4), 123.24 (CH-5), 124.29 (CH-8), 126.00
CH-10), 126.14 (CH-2, CH-3), 126.58 (CH-6), 126.60 (CH-7),
1
15–116 ЊC (lit. 114.5–116.5 ЊC); H NMR: 1.977 (m, 2H,
CH -2), 2.015 (m, 2H, CH -3), 3.154 (t, 2H, J 7.2, CH -4), 3.482
t, 2H, J 6.4, CH -1), phenanthrenyl protons 7.587 (s, 1H,
2
CH-10), 7.614 (overlapped signals, 2H, CH-2, CH-3), 7.662
overlapped signals, 2H, CH-6, CH-7), 7.839 (d, 1H, J 7.5,
CH-1), 8.107 (d, 2H, J 7.0, CH-8), 8.664 (d, 1H, J 8.4, CH-4),
.752 (d, 1H, J 7.1, CH-5); important cross peaks in COSY:
.977/3.482, 2.015/3.154, 7.614/7.839, 7.614/8.664, 7.662/8.752,
.662/8.107, NOESY: 3.154/7.587, 3.154/8.107, 7.587/7.839,
.664/8.752; C NMR: 28.57 (CH -3), 32.48 (CH -4), 32.69
CH -2), 33.59 (CH -1), phenanthrenyl CH-carbons 122.44
CH-4), 123.27 (CH-5), 124.28 (CH-8), 126.04 (CH-10), 126.19
CH-2, CH-3), 126.56 (CH-6), 126.63 (CH-7), 128.04 (CH-1).
(
1
2
2
2
28.03 (CH-1); Anal calc. for C H Br: C 69.73, H 5.85; Found
19 19
(
C 69.91, H 5.81%.
(
9
1
-Bromo-4-(1-naphthyl)butane (2f ). 39% Yield; colourless
1
oil; H NMR: 1.844 (m, 2H, CH -3), 1.899 (m, 2H, CH -2),
3
2
2
8
1
7
8
(
(
(
.032 (t, 2H, J 6.9, CH -4), 3.369 (t, 2H, J 6.5, CH -1), naphthyl
2
2
protons 7.262 (d, 1H, J 6.9, CH-2), 7.357 (dd, 1H, J 7.0, 8.4,
CH-3), 7.434 (dd, 1H, J 6.8, 8.5, CH-6), 7.479 (dd, 1H, J 6.9,
8.4, CH-7), 7.681 (d, 1H, J 8.5, CH-4), 7.816 (d, 1H, J 8.5,
CH-5), 7.987 (d, 1H, J 8.5, CH-8); important cross peaks in
COSY: 1.899/3.369, 1.844/3.032, 7.262/7.357, 7.357/7.681,
13
2
2
2
2
7
7
.434/7.816, 7.479/7.987, NOESY: 3.032/7.262, 3.032/7.987,
13
.681/7.816; C NMR: 29.21 (CH -3), 32.14 (CH -4), 32.71
2
2
6
a
1
-Bromo-3-(9-phenanthrenyl)propane (2b). 22% Yield; mp
1
(
CH -2), 33.64 (CH -1), naphthyl CH-carbons 123.73 (CH-8),
2
2
5
(
6–57 ЊC (lit. 55–57 ЊC); H NMR: 2.329 (m, 2H, CH -2), 3.259
2
125.54 (CH-3, CH-6), 125.87 (CH-7), 126.02 (CH-2), 126.79
(CH-4), 128.86 (CH-5).
t, 2H, J 7.2, CH -3), 3.477 (t, 2H, J 6.4, CH -1), phenanthrenyl
2
2
protons 7.576 (overlapped signals, 2H, CH-2, CH-3), 7.594 (s,
H, CH-10), 7.622 (overlapped signals, 2H, CH-6, CH-7), 7.811
d, 1H, J 7.2, CH-1), 8.069 (d, 1H, J 6.9, CH-8), 8.627 (d, 1H,
J 7.9, CH-4), 8.714 (d, 1H, J 6.9, CH-5); important cross peaks
1
1
1
(
1-Bromo-4-(2-naphthyl)butane (2g). 44% Yield; colourless
1
oil; H NMR: 1.876 (m, 2H, CH -3), 1.916 (m, 2H, CH -2),
2.798 (t, 2H, J 7.2, CH -4), 3.423 (t, 2H, J 6.4, CH -1), naphthyl
2
2
2
2
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 5 1 4 – 5 2 3
521

View Article Online
protons 7.321 (d, 1H, J 8.5, CH-3), 7.416 (dd, 1H, J 7.2, 8.1,
CH-6), 7.453 (dd, 1H, J 7.3, 8.4, CH-7), 7.605 (s, 1H, CH-1),
CH-1), 7.990 (d, 2H, J 8.0, CH-8), 8.675 (d, 1H, J 8.1, CH-4),
8.726 (d, 1H, J 8.3, CH-5); important cross peaks in COSY:
3.272/3.998, 7.631/7.864, 7.631/8.675, 7.714/7.990, 7.714/8.726,
7
.772 (d, 2H, J 8.1, CH-4, CH-8), 7.805 (d, 1H, J 7.7, CH-5);
13
important cross peaks in COSY: 1.876/2.798, 1.916/3.423,
NOESY: 7.563/7.864, 8.675/8.726; C NMR: 37.78 (CH -1),
2
7
2
3
.321/7.772, 7.416/7.805, 7.453/7.772, NOESY: 2.798/7.605,
38.92 (CH -2), 149.45 (C᎐O), phenanthrenyl CH-carbons
117.69 (CH-10), 121.78 (CH-8), 122.63 (CH-4), 122.98 (CH-5),
126.56 (CH-2), 126.96 (CH-3), 127.06 (CH-7), 127.28 (CH-6),
2
13
.798/7.321, 1.876/7.321, 7.605/7.772; C NMR: 29.65 (CH -3),
2
2.20 (CH -2), 33.64 (CH -1), 35.08 (CH -4), naphthyl CH-
2
2
2
carbons 125.18 (CH-6), 125.94 (CH-7), 126.40 (CH-1), 127.16
128.48 (CH-1); FABϩ [M] 329 (3.54), [M Ϫ CO ] 285 (18.52),
2
(
CH-3), 127.39 (CH-8), 127.59 (CH-4), 127.95 (CH-5).
[M Ϫ Br] 249 (23.04), [M Ϫ CH᎐CHBr] 223 (38.64),
[McLaferty rearrangement] 194 (100.00).
12
1
-Bromo-4-(1-pyrenyl)butane (2h). 28% Yield; mp 71–72 ЊC;
1
H NMR: 2.026 (m, 4H, CH -2, CH -3), 3.372 (t, 2H, J 7.0,
General procedure for the microwave irradiation (MWI)
reactions
2
2
CH -4), 3.467 (t, 2H, J 6.2, CH -1), pyrenyl protons 7.859 (d,
2
2
1
H, J 7.8, CH-2), 7.996 (dd, 1H, J 7.6, CH-7), 8.031 (over-
lapped signals, 2H, CH-4, CH-5), 8.115 (overlapped signals,
H, CH-3, CH-9), 8.168 (d, 1H, J 7.4, CH-6), 8.174 (d, 1H,
To a solution of arylbromoalkane (1.0 mmol) in dichloro-
methane (5 mL) silica gel (2.0 g) was added and the solvent was
removed in vacuo. The powder mixture was heated in a MW
oven at 800 W for 15 min. After extraction from silica gel with
2
J 7.4, CH-8), 8.268 (d, 1H, J 9.2, CH-10); important cross
peaks in COSY: 2.026/3.372, 2.026/3.467, 7.859/8.115, 7.996/
8
8
(
1
dichloromethane, the conversions were determined by H NMR
.168, 7.996/8.174, 8.115/8.268, NOESY: 3.372/7.859, 3.372/
13
spectroscopy of the crude reaction mixtures. The products were
purified by flash chromatography on silica gel using hexane as a
mobile phase.
.268, 8.031/8.115, 8.031/8.168, 8.115/8.174; C NMR: 30.16
CH -3), 32.55 (CH -2, CH -4), 33.52 (CH -1), pyrenyl CH-
2 2 2 2
carbons 123.18 (CH-10), 124.73 (CH-6), 124.76 (CH-8), 124.88
(
(
CH-3), 125.80 (CH-9), 126.64 (CH-7), 127.15 (CH-2), 127.31
CH-4), 127.44 (CH-5).
1
6
1
,2,3,4-Tetrahydrotriphenylene (3a). Starting bromide; 1-
bromo-4-(9-phenanthrenyl)butane (2a); 95% conversion;
isolated yield 91%; mp 119–120 ЊC (lit. 120–121 ЊC); H NMR:
1
13
1
1
-Bromo-4-phenylbutane (2i). 42% Yield; colourless oil; H
2
.000 (m, 4H, CH -2, CH -3), 3.164 (t, 4H, J 6.4, CH -1,
NMR: 1.782 (m, 2H, CH -3), 1.892 (m, 2H, CH -2), 2.642 (t,
2
7
COSY: 1.892/3.417, 1.782/2.642, NOESY: 2.642/7.197;
NMR: 29.844 (CH -3), 32.247 (CH -2), 33.643 (CH -1), 34.974
(
2
2
2
2
2
CH -4), 7.604 (overlapped signals, 4H, CH-6, CH-7, CH-10,
CH-11), 8.065 (d, 2H, J 7.4, CH-5, CH-12), 8.695 (d, 2H, J 7.4,
CH-8, CH-9); important cross peaks in COSY: 2.000/3.164,
.604/8.065, 7.604/8.695, NOESY: 3.164/8.065; C NMR:
2.91 (CH -2, CH -3), 26.85 (CH -1, CH -4), 122.70 (CH-8,
H, J 7.6, CH -4), 3.417 (t, 2H, J 6.7, CH -1), 7.178 (1H, p-Ph),
2
2
2
.197 (2H, o-Ph), 7.285 (2H, m-Ph); important cross peaks in
13
C
1
3
7
2
2
2
2
CH -4), 125.91 (o-Ph), 128.38 (m-Ph, p-Ph).
2
2
2
2
2
CH-9), 123.29 (CH-5, CH-12), 125.48 (CH-6, CH-11), 126.52
(CH-7, CH-10).
14
1
-Bromo-4-(3,5-dimethylphenyl)butane
(2j). 56% Yield;
2
1
colourless oil; H NMR: 1.735 (m, 2H, CH -3), 1.881 (m, 2H,
CH -2), 2.555 (t, 2H, J 7.6, CH -4), 3.434 (t, 2H, J 6.8, CH -1),
2
ant cross peaks in COSY: 1.735/2.555, 1.881/3.434, NOESY:
2
(
(
1
-Methyl-1,2,3,4-tetrahydrotriphenylene (3e). Starting brom-
ide; 1-bromo-1-methyl-4-(9-phenanthrenyl)butane (2e); 100%
conversion; isolated yield 97%; mp 94–95 ЊC; IR 705, 739, 896,
265, 1429, 1495, 2304, 2872, 2935, 2968, 3053, 3079 cm
NMR: 1.436 (d, 3H, J 6.8, CH ), 1.962 (m, 2H, CH -2), 2.096
m, 2H, CH -3), 3.056 (m, 1H, CH -4), 3.307 (dd, 1H, J 5.6,
2 2
7.2, CH -4), 3.720 (m, 1H, CH-1), 7.614 (overlapped signals,
4H, CH-6, CH-7, CH-10, CH-11), 8.081 (d, 1H, J 6.0, CH-5),
8.149 (d, 1H, J 7.6, CH-12), 8.702 (overlapped signals, 2H, CH-
8, CH-9); important cross peaks in COSY: 1.436/3.720, 1.962/
3.720, 2.096/3.056, 2.096/3.307, 7.614/8.081, 7.614/8.702, 7.614/
8.149, NOESY: 1.436/8.149, 3.720/8.149, 3.056/8.081, 3.307/
8.081; C NMR: 17.549 (CH -3), 21.474 (CH ), 26.501 (CH -
4), 28.480 (CH-1), 29.752 (CH -2), 122.63 (CH-9), 122.88 (CH-
8), 123.48 (CH-5), 123.92 (CH-12), 125.25 (CH-10), 125.52
(CH-11), 126.40 (CH-7), 126.50 (CH-6); Anal. calc. for C H
C 92.63, H 7.37; Found C 92.52, H 7.28%.
2
2
2
.283 (s, 6H, CH ), 6.790 (2H, o-Ph), 6.824 (1H, p-Ph); import-
3
Ϫ1
1
13
; H
1
.555/6.790; C NMR: 21.33 (CH ), 29.96 (CH -3), 32.41
3
2
CH -2), 33.70 (CH -1), 34.88 (CH -4), 126.27 (o-Ph), 127.57
3
2
2
2
2
(
1
p-Ph).
2
15
1
-Bromo-4-(2-biphenyl)butane (2k). 22% Yield; colourless
1
oil; H NMR: 1.581 (m, 2H, CH -3), 1.705 (m, 2H, CH -2),
2
2
2
.592 (t, 2H, J 7.7, CH -4), 3.216 (t, 2H, J 6.8, CH -1), biphenyl
2
2
protons 7.296 (d, 1H, J 7.8, CH-3), 7.317 (overlapped signal,
H, CH-9), 7.330 (overlapped signal, 1H, CH-4), 7.364 (over-
lapped signal, 1H, CH-5), 7.387 (d, 2H, J 7.7, CH-7, CH-11),
.426 (d, 1H, J 6.5, CH-6), 7.483 (dd, 2H, CH-8, CH-10);
important cross peaks in COSY: 1.581/2.592, 1.705/3.216,
1
13
2
3
2
7
2
7
2
3
.296/7.330, 7.364/7.426, 7.317/7.483, 7.387/7.483, NOESY:
19
18
13
.592/7.296, 2.592/7.387, 7.387/7.426; C NMR: 29.69 (CH -3),
2
2.03 (CH -4), 32.38 (CH -2), 33.50 (CH -1), biphenyl CH-
2
2
2
1
7
carbons 125.91 (CH-3), 126.90 (CH-6), 127.49 (CH-4), 128.15
CH-8, CH-10), 129.23 (CH-5, CH-7, CH-11), 130.14 (CH-9).
1,2,3,4-Tetrahydrophenanthrene (3f ). a) Starting bromide;
1-bromo-4-(1-naphthyl)butane (2f ); 94% conversion; isolated
yield 91%; mp 31–32 ЊC (lit. 32–33 ЊC); H NMR: 1.860 (m, 2H,
CH -2), 1.970 (m, 2H, CH -3), 2.899 (t, 2H, J 5.8, CH -1), 3.101
(t, 2H, J 6.2, CH -4), 7.189 (d, 1H, J 8.5, CH-10), 7.412 (dd,
(
1
(
9-Phenanthrenyl)-3-bromopropionate (4). To a stirred solu-
tion of 9-phenanthrol (194 mg, 1.0 mmol) and pyridine
0.4 mL, 5.0 mmol) in anhydrous dichloromethane (3 mL)
2
2
2
2
(
1H, J 8.2, 8.6, CH-7), 7.481 (dd, 1H, J 8.4, 8.5, CH-6), 7.598 (d,
1H, J 8.5, CH-9), 7.778 (d, 1H, J 8.5, CH-8), 7.955 (d, 1H, J 8.5,
CH-5); important cross peaks in COSY: 1.860/2.899, 1.970/
3.101, 7.189/7.598, 7.412/7.778, 7.481/7.955, NOESY: 2.899/
at room temperature and under an argon atmosphere, 3-bromo-
propionyl chloride (0.2 mL, 1.75 mmol) was added. After 2
hours the reaction mixture was partioned between dichloro-
methane (20 mL) and hydrochloric acid (10 mL). The organic
phase was washed with brine, dried over MgSO , evaporated in
vacuo and purified by flash chromatography on silica gel using
1
3
7.189, 3.101/7.955, 7.598/7.778; C NMR: 22.96 (CH -2), 23.24
2
4
(CH -3), 25.67 (CH -1), 30.45 (CH -4), 122.82 (CH-5),124.53
2
2
2
(CH-7), 124.80 (CH-9), 125.72 (CH-6), 128.24 (CH-10), 128.44
(CH-8). b) Starting bromide; 1-bromo-4-(2-naphthyl)butane
(2g); 100% conversion; isolated yield 96%; identical mp and
NMR data to the product described in a).
1
0% ethyl acetate in hexane as a mobile phase to give (4)
(
30 mg, 9%); mp 102–103 ЊC; IR: 705, 896, 1135, 1265, 1422,
Ϫ1
1
1
765, 2305, 2987, 3054, 3418 cm ; H NMR: 3.272 (t, 2H,
CH -1), 3.998 (t, 2H, CH -2), phenanthrenyl protons 7.563 (s,
1
(
2
2
18
H, CH-10), 7.631 (overlapped signals, 2H, CH-2, CH-3), 7.714
overlapped signals, 2H, CH-6, CH-7), 7.864 (d, 1H, J 7.6,
7,8,9,10-Tetrahydrobenzo[a]pyrene (3h). Starting bromide;
1-bromo-4-(1-pyrenyl)butane (2h); 76% conversion; isolated
5
22
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 5 1 4 – 5 2 3

View Article Online
1
yield 72%; mp 112–113 ЊC (lit. 113 ЊC); H NMR: 1.986 (m, 2H,
Notes and references
CH -8), 2.092 (m, 2H, CH -9), 3.232 (t, 2H, J 6.1, CH -7), 3.439
t, 2H, J 6.3, CH -10), 7.895 (s, 1H, CH-6), 7.937 (overlapped
2
2
2
1
Reviews: (a) S. Caddick, Tetrahedron, 1995, 51, 10403; (b)
(
2
S. Deshayes, M. Liagre, A. Loupy, J.-L. Luche and A. Petit,
Tetrahedron, 1999, 55, 10851; (c) P. Lidstrom, J. Tierney, B. Wathey
and J. Westman, Tetrahedron, 2001, 57, 9225; (d ) A. Kirschning,
H. Monenschein and R. Wittenberg, Angew. Chem., Int. Ed., 2001,
40, 650; (e) M. Larhed, C. Moberg and A. Hallberg, Acc. Chem.
Res., 2002, 35, 717 and references cited therein.
signal, 1H, CH-4), 7.943 (overlapped signal, 1H, CH-2), 7.956
overlapped signal, 1H, CH-1), 8.085 (d, 1H, J 9.6, CH-12),
.108 (d, 1H, J 8.7, CH-5), 8.138 (d, 1H, J 7.6, CH-3), 8.233 (d,
H, J 9.2, CH-11); important cross peaks in COSY: 1.986/
.232, 2.092/3.439, 7.937/8.108, 7.943/7.956, 7.943/8.138, 8.085/
.233, NOESY: 3.232/7.895, 3.439/8.233, 7.895/8.108, 7.937/
.138, 7.956/8.085; C NMR: 23.05 (CH -8), 23.51 (CH -9),
6.40 (CH -10), 30.89 (CH -7), 122.74 (CH-11), 124.63 (CH-3,
CH-5), 125.28 (CH-2), 125.83 (CH-6), 126.43 (CH-4), 127.01
CH-1), 127.16 (CH-12).
(
8
1
3
8
8
2
2
Reviews: (a) A. Loupy, A. Petit, J. Hamelin, F. Texier-Boullet,
P. Jacquault and D. Mathe, Synthesis, 1998, 1213; (b) R. S. Varma,
Green Chemistry, 1999, 43 and references cited therein.
13
2
2
2
2
3 (a) E. Clar, Polycyclic Hydrocarbons, Academic Press,New York,
1964, vol. I, II; (b) F. Oesch, G. Stillger, H. Frank and K. L. Platt,
J. Org. Chem., 1982, 47, 568; (c) A. T. Balaban, M. D. Gheorghiu,
A. Schiketanz and A. Necula, J. Am. Chem. Soc., 1989, 111, 734;
(d ) R. M. Roberts, A. A. Khalaf, Friedel–Crafts Alkylation
Chemistry. A Century of Discovery, Dekker, New York, 1984; (e)
G. A. Olah, R. Krishnamurti, G. K. S. Prakash, Friedel–Crafts
Alkylations, in Comprehensive Organic Synthesis, ed. B. M. Trost,
Pergamon Press, Oxford, 1991, vol. 3, p. 293; ( f ) B. M. Trost and
M. R. Ghandiri, J. Am. Chem. Soc., 1984, 106, 7260; (g) S. T. Bright,
J. M. Coxon and P. J. Steel, J. Org. Chem., 1990, 55, 1338; (h) S. R.
Angle and M. S. Louie, J. Org. Chem., 1991, 56, 2853; (i) I. Hachiro,
M. Moriwaki and S. Kobayashi, Bull. Chem. Soc. Jpn., 1995, 68,
(
19
1
,2,3,4-Tetrahydronaphthalene
(Tetralin) (3i). Starting
bromide; 1-bromo-4-phenylbutane (2i); 17% conversion; 13%
isolated yield; colourless oil; H NMR: 1.787 (m, 4H, CH -2,
CH -3), 2.753 (t, 4H, J 6.5, CH -1, CH -4), 7.056 (m, 4H, CH-5,
CH-6, CH-7, CH-8); important cross peak in COSY: 1.787/
.753, NOESY: 2.753/7.056; C NMR: 23.27 (CH -2, CH -3),
9.43 (CH -1, CH -4), 125.43 (CH-6, CH-7), 129.16 (CH-5,
1
2
2
2
2
13
2
2 2
2
2 2
CH-8).
2
053.
20
4 (a) V. K. Bhalerao, B. S. Nanjundiah, H. R. Sonawane and P. M.
Nair, Tetrahedron, 1986, 42, 1487; (b) K. V. Subbarao, N. P.
Damodaran and S. Dev, Tetrahedron, 1987, 43, 2543; (c) P. J. Kropp
and R. L. Adkins, J. Am. Chem. Soc., 1991, 113, 2709.
5 A. Loupy, P. Pigeon and M. Ramdani, Tetrahedron, 1996, 52,
6705.
6 (a) A. Sugimoto, K. Sumi, K. Urakawa, M. Ikemura, S. Sakamoto,
Sh. Yoneda and Y. Otsuji, Bull. Chem. Soc. Jpn., 1983, 56, 3118;
(b) H. Cao, Y. Fujiwara, T. Haino, Y. Fukazawa, C.-H. Tung and
Y. Tanimoto, Bull. Chem. Soc. Jpn., 1996, 69, 2801; (c) C. A. M.
Afonso and J. P. S. Farinha, J. Chem. Res., 2003, 584.
7 J. March, Advanced Organic Chemistry: Reactions, Mechanisms, and
Structures, John Wiley & Sons, New York, 1992.
5
,7-Dimethyl-1,2,3,4-tetrahydronaphthalene
bromide; 1-bromo-4-(3,5-dimethylphenyl)butane (2j); 100%
conversion; isolated yield 96%; colourless oil; H NMR: 1.747
(3j). Starting
1
(
m, 2H, CH -2), 1.811 (m, 2H, CH -3), 2.576 (t, 2H, J 6.5, CH -
2
2
2
4
3
), 2.726 (t, 2H, J 6.2, CH -1), 2.171 (s, 3H, CH -5), 2.245 (s,
2 3
H, CH -7), 6.756 (s, 1H, CH-8), 6.801 (s, 1H, CH-6); import-
ant cross peak in COSY: 1.747/2.725, 1.811/2.575; NOESY:
.726/6.756, 2.576/2.171, 2.171/6.801, 2.245/6.756; C NMR:
9.37 (CH -5), 20.79 (CH -7), 23.01 (CH -2), 23.57 (CH -3),
6.36 (CH -4), 30.06 (CH -1), 127.43 (CH-8), 127.96 (CH-6).
3
1
3
2
1
2
3
3
2
2
2
2
21
5
-Phenyl-1,2,3,4-tetrahydronaphthalene
bromide; 1-bromo-4-(2-biphenyl)butane (2k); 19% conversion;
4% isolated yield; colourless oil; H NMR: 1.629 (m, 2H, CH -
), 1.724 (m, 2H, CH -2), 2.512 (t, 2H, J 5.6, CH -4), 2.785
(3k).
Starting
8
A. G. Santos, S. X. Candeias, C. A. M. Afonso, K. Jenkins,
S. Caddick, N. R. Treweeke and D. Pardoe, Tetrahedron, 2001, 57,
6607.
1
1
3
2
9
R. H. F. Manske and A. E. Ledingham, Can. J. Res., 1939, 17B, 14
2
2
5
(
Chem. Abstr., 1939, 33, 5387 ).
(
t, 2H, J 6.5, CH -1), 6.961 (d, 1H, J 6.8, CH-6), 7.024 (d, 1H,
2
1
1
0 A. Sugimoto, R. Hiraoka, N. Fukada, H. Kosaka and H. Inoue,
J. Chem. Soc., Perkin Trans. 1, 1992, 2871.
1 (a) R. Huisgon and U. Rietz, Chem. Ber., 1957, 90, 2768; (b) P. Studt,
Ann. Chem., 1965, 685, 49.
J 6.8, CH-8), 7.086 (dd, 1H, CH-7), 7.213 (overlapped signal,
H, o-Ph), 7.251 (overlapped signal, 1H, o-Ph), 7.321 (over-
lapped signal, 2H, m-Ph); Important cross peaks in COSY:
2
1
2
2
1
1
.629/2.512, 1.724/2.785, 6.961/7.086, 7.024/7.086, NOESY:
12 K. Furuta, K. Tomokiyo, M. T. Kuo, T. Ishikawa and M. Suzuki,
1
3
Tetrahedron, 1999, 55, 7529.
3 (a) L. Cherkasov, Kh. Balyan and V. Kormer, Zh. Org. Khim., 1966,
.512/7.251, 2.785/7.024, 6.961/7.251; C NMR: 22.88 (CH -2),
2
1
3.32 (CH -3), 28.12 (CH -4), 29.90 (CH -1), 125.13 (CH-7),
26.60 (p-Ph), 127.03 (CH-6), 127.91 (m-Ph), 128.36 (CH-8),
29.15 (o-Ph).
2
2
2
2
, 1934; (b) K. Subbarao, N. Damodaran and S. Dev, Tetrahedron,
987, 43, 2543.
1
1
1
4 R. Abramovitch and A. Kress, J. Org. Chem., 1984, 49, 3114.
5 A. Gribble, R. Dolle, A. Shaw, D. McNair, R. Novelli, Chr. Novelli,
Br. Slingsby, V. Shah, D. Tew, B. Saxty, M. Allen, P. Groot, N. Pearce
and J. Yates, J. Med. Chem., 1996, 39, 3569.
(
9-Phenanthrenyl)acrylate (5). Starting bromide; (9-phenan-
threnyl)-3-bromopropionate (4); 66% conversion; 63% isolated
yield; mp 69–70 ЊC; IR 705, 737, 896, 1153, 1265, 1422, 1745,
16 (a) E. Bergmann and O. Blum-Bergmann, J. Am. Chem. Soc., 1937,
59, 1441; (b) Z. Marcinow, A. Sygula and P. W. Rabideau, J. Org.
Chem., 1988, 53, 3603.
Ϫ1
1
2
2
1
J
305, 2987, 3054, 3438 cm ; H NMR: 6.122 (dd, 1H, J
cis,trans
3
3
3.2, J
10.5, Hcis of CH ᎐), 6.514 (dd, 1H, J
10.5,
H,Htrans
2
᎐
H,Htrans
1
1
7 W. Bachmann and W. Struve, J. Org. Chem., 1939, 4, 472.
8 (a) L. F. Fieser and M. Fieser, J. Am. Chem. Soc., 1935, 57, 782;
(b) P. Fu, H. Lee and R. Harvey, J. Org. Chem., 1980, 45, 2797.
2
3
17.3, CH᎐), 6.761 (dd, 1H, J
1.2, J
H,Hcis
17.3, Htrans
H,Hcis
cis,trans
of CH ᎐), phenanthrenyl protons 7.588 (s, 1H, CH-10), 7.611
2
᎐
(
overlapped signals, 2H, CH-2, CH-3), 7.689 (overlapped sig-
nals, 2H, CH-6, CH-7), 7.848 (d, 1H, J 7.7, CH-1), 7.954 (d,
H, J 8.1, CH-8), 8.655 (d, 1H, J 8.0, CH-4), 8.706 (d, 1H, J 8.2,
CH-5); important cross peaks in COSY: 6.514/6.122, 6.514/
.761, 7.611/7.848, 7.611/8.655, 7.689/7.954, 7.689/8.706,
NOESY: 6.122/6.514, 6.122/6.761, 7.588/7.848, 8.655/8.706;
NMR: 127.74 (CH᎐), 133.04 (CH ᎐), 164.56 (C᎐O), phenan-
threnyl CH-carbons 117.70 (CH-10), 121.84 (CH-8), 122.64
CH-4), 122.98 (CH-5), 126.46 (CH-3), 126.90 (CH-2), 127.04
CH-7), 127.22 (CH-6), 128.49 (CH-1); Anal. Calc. for
C H O C 82.24, H 4.87; Found C 82.10, H 4.90%.
19 Aldrich catalog, No. 52,265–1.
20 E. de B. Barnett and F. G. Sanders, J. Chem. Soc., 1933, 434.
2
1 Y. Nakatsuji, T. Kubo, M. Nomura and Sh. Kikkawa, Bull. Chem.
Soc. Jpn., 1978, 51, 618.
2
2
2 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A. Montgomery,
Jr., R. E. Stratmann, J. C. Burant, S. Dapprich, J. M. Millam,
A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J. Tomasi,
V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli,
C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson, P. Y. Ayala,
Q. Cui, K. Morokuma, D. K. Malick, A. D. Rabuck,
K. Raghavachari, J. B. Foresman, J. Cioslowski, J. V. Ortiz,
B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi,
R. Gomperts, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham,
C. Y. Peng, A. Nanayakkara, C. Gonzalez, M. Challacombe,
P. M. W. Gill, B. G. Johnson, W. Chen, M. W. Wong, J. L. Andres,
M. Head-Gordon, E. S. Replogle and J. A. Pople, GAUSSIAN 98
6
13
C
᎐
2
(
(
1
7
12
2
Acknowledgements
We thank Fundação para a Ciência e Tecnologia, FEDER
Ref. POCTI/QUI/42983/2001 and Ref. SFRH/BPD/5531/2001)
and INVOTAN (Ref. OUTREACH) for financial support.
(
(
Revision A.7), Gaussian, Inc., Pittsburgh, PA, 1998.
23 Spartan Ј02, Wavefunction, Inc.Irvine, CA.
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 5 1 4 – 5 2 3
523