Iodine-catalyzed cycloalkenylation of dihydroquinolines and arylamines through a reaction with cyclic ketones under neat conditions

Article abstract of DOI:10.1016/j.tetlet.2013.10.081

An iodine-catalyzed direct cycloalkenylation of dihydroquinolines and arylamines has been developed. This method consists of a Friedel-Crafts reaction between dihydroquinolines (or arylamines) and cyclic ketones in which the double bond is selectively generated throughout the course of the reaction resulting in a direct cycloalkenylation, under neat conditions.

Full text of DOI:10.1016/j.tetlet.2013.10.081

Tetrahedron Letters 54 (2013) 7069–7073
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Iodine-catalyzed cycloalkenylation of dihydroquinolines
and arylamines through a reaction with cyclic ketones
under neat conditions
a
a
b
Jean Fotie a, , Suraj K. Ayer , Binit S. Poudel , Carolyn S. Reid
⇑
a Department of Chemistry and Physics, Southeastern Louisiana University, SLU 10878, Hammond, LA 70402-0878, USA
b Department of Chemistry and Biochemistry, University of Mount Union, 1972 Clark Avenue, Alliance, OH 44601-3993, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
An iodine-catalyzed direct cycloalkenylation of dihydroquinolines and arylamines has been developed.
This method consists of a Friedel–Crafts reaction between dihydroquinolines (or arylamines) and cyclic
ketones in which the double bond is selectively generated throughout the course of the reaction resulting
in a direct cycloalkenylation, under neat conditions.
Received 24 August 2013
Revised 14 October 2013
Accepted 17 October 2013
Available online 23 October 2013
Ó 2013 Elsevier Ltd. All rights reserved.
Keywords:
Dihydroquinolines and arylamines
Regioselective cycloalkenylation
Molecular iodine
Cyclic ketones
Friedel–Crafts reaction
Introduction
Results and discussion
Charles Friedel and James Mason Crafts initially reported the
alkylation and the acylation of aromatic rings using alkyl or acyl
chlorides respectively, in the presence of aluminum chloride in
1877.1 This breakthrough discovery has generated a lot of inter-
est over the years resulting in the synthesis of a wide range of
molecules through the modification of the original reaction con-
ditions. Throughout these years, a large number of catalysts
including Lewis acids such as BF3, BeCl2, TiCl4, SbCl5, or SnCl4,
As part of our continued effort to investigate the real mecha-
nism of the iodine-catalyzed version of the Skraup–Doebner–
Von-Miller quinoline synthesis,4 we discovered that molecular
iodine can catalyze a direct cycloalkenylation on dihydroquino-
lines. In fact, when reacting ethoxyquin with cyclohexanone or
cyclopentanone in the presence of a catalytic amount of iodine,
we obtained 8-cyclohexenyl-6-ethoxy-1,2-dihydro-2,2,4-trimeth-
ylquinoline (1a, 54%) or 8-cyclopentenyl-6-ethoxy-1,2-dihydro-
2,2,4-trimethylquinoline (1b, 52%), respectively (see Scheme 1).4a
Since these reaction conditions are unprecedented to the best of
our knowledge, we undertook further investigations with a goal
of expanding the scope of such a reaction.
Bronsted acids such as HFÁSbF5 or HSO3FÁSbF5, and even nano-
2À
TiO2/SO4
or Ph3PAuCl/AgOTf, to name a few, have been
explored.2 After more than a century of history, Friedel–Crafts
reactions are still at the forefront of organic synthesis, with the
emphasis geared toward the use of less toxic and non-corrosive
catalysts. One of these catalysts is molecular iodine (I2) which
has been used in a wide variety of transformations including
oxidative cyclization, cascade reactions, and direct C–H function-
alizations, to name just a few.3 Herein we report the iodine-cata-
lyzed direct cycloalkenylation of arylamines, through a reaction
between a cyclic ketone and an aromatic system in which the
double bond is selectively generated throughout the reaction
process, resulting in a direct C–H functionalization.
EtO
O
EtO
I
2 (5% mol.)
+
N
H
160°C, 72 h
neat, air
N
H
(CH2)n
5 eqv
n = 1 or 2
1a
: n = 2 (54%)
1b: n = 1 (52%)
(CH2)n
⇑
Corresponding author. Tel.: +1 985 549 5112; fax: +1 985 549 5126.
E-mail address: jean.fotie@selu.edu (J. Fotie).
Scheme 1. Reaction between ethoxyquin and cyclohexanone or cyclopentanone.4a
0040-4039/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tetlet.2013.10.081

7070
J. Fotie et al. / Tetrahedron Letters 54 (2013) 7069–7073
Table 1
O
Optimization of the reaction conditions
O
Entrya
Catalystb
Cyclohexanoneb
1a, Yieldc
(mol)
(%)
3
2
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
I2 (1% mol)
I2 (1% mol)
I2 (5% mol)
I2 (5% mol)
I2 (10% mol)
I2 (100% mol)
I2 (5% mol)
I2 (5% mol) + DPPH (5% mol)
I2 (5% mol) + DPPH (100% mol)
KI (5% mol)
CuI (5% mol)
NIS (5% mol)
AIBN (5% mol)
AIBN (100% mol)
I2 (5% mol), under nitrogen
I2 (5% mol), under oxygen
1
3
3
5
5
5
10
5
5
5
5
5
5
5
5
5
10
38
67
69
53
39
68
17
NR
NR
NR
67
NR
NR
NR
68
With the optimal reaction conditions in hand, we undertook to
explore the scope of the reaction starting with a series of 1,2-dihy-
droquinolines prepared through the iodine-catalyzed version of the
Skraup–Doebner–Von Miller synthesis.6 Toward this goal, 4-chloro-
aniline or aniline was allowed to react with acetone in refluxing tol-
uene to yield 6-chloro-1,2-dihydro-2,2,4-trimethylquinoline (4a,
38%) and 1,2-dihydro-2,2,4-trimethylquinoline (4b, 48%), respec-
tively (see Scheme 2).
These 1,2-dihydroquinoline derivatives were successively
subjected to the reaction conditions described in entry 4, in the
presence of cyclohexanone or cyclopentanone to yield the corre-
sponding cycloalkenylated dihydroquinoline derivatives. In fact,
under these conditions, 6-chloro-1,2-dihydro-2,2,4-trimethylquin-
oline (4a) reacts with cyclohexanone to yield 6-chloro-8-cyclohex-
enyl-1,2-dihydro-2,2,4-trimethylquinoline (5a) which totally
decomposes regardless of the method of purification and preserva-
tion. Thus, this compound was never fully characterized, although
the HRESI-MS spectrum was very decent. On the other hand, react-
ing 4a with cyclopentanone yielded 6-chloro-8-cyclopentenyl-1,2-
dihydro-2,2,4-trimethylquinoline (5b, 33%) which was very stable
and was fully characterized (see Scheme 3). The yield in these
cases was significantly lower than that obtained with ethoxyquin
under the exact conditions, suggesting that the deactivating effect
of the chlorine atom on the aromatic ring might be having a nega-
tive effect on the reaction.
More importantly, 1,2-dihydro-2,2,4-trimethylquinoline (4b), a
dihydroquinoline derivative in which the ortho- and the para-posi-
tions to the nitrogen are available, reacted under the same
conditions with cyclopentanone or cyclohexanone to yield 6,8-dicyc-
lohexenyl-1,2-dihydro-2,2,4-trimethylquinoline (6a, 61%) and 6,8-
dicyclopentenyl-1,2-dihydro-2,2,4-trimethylquinoline (6b, 58%),
respectively. These two compounds are cycloalkenylated at the
ortho- and para-positions to the nitrogen of the dihydroquinoline ring
as illustrated in Scheme 4. In the case of cyclopentanone, we also
isolated a side product (7b) in which only the ortho-position to the
nitrogen atom is substituted. This probably suggests that the
ortho-position to the nitrogen atom in these molecules is more
reactive than the para-position.
a
Each reaction mixture was allowed to stir at 160 °C for 72 h.
The percent mol and the mol equivalence are in relationship to 1 equiv of
b
ethoxyquin.
c
The percent yields refer to pure isolated products. NR = No Reaction.
As
a starting point for these investigations, the reaction
conditions described in Scheme 1 were optimized using the
commercially available 6-ethoxy-1,2-dihydro-2,2,4-trimethylquin-
oline (ethoxyquin) and cyclohexanone as starting materials, and
the obtained results are shown in Table 1. It appeared that the
use of 3 to 5 equivalents of the ketones and 0.05 equivalent of
iodine (entries 3 and 4) produced the best yield, and the use of a
larger amount of ketone under the same conditions (entry 7) did
not produce any further improvement of the yield. While the low
amount of iodine is to be blamed for the low yield obtained in
entry 2, the use of a larger amount of iodine (>5% mol) appeared
to negatively impact the yield of the reaction (entries 5 and 6).
Furthermore, the low amount of cyclohexanone (1:1 mol of eth-
oxyquin) is also to be blamed for the poor yield observed in entry
1 when compared to entry 2.
As we previously reported,4a cyclohexanone polymerizes
when heated in the presence of a catalytic amount of iodine to
yield predominantly b,
cyclopentanone reacts under the same conditions to produce
mainly ,b-unsaturated ketone oligomers (3), with the dimer
c-unsaturated ketone oligomers (2), while
a
being the major product in each case. This polymerization is
exacerbated by an increased amount of iodine,4a and this is
consistent with the use of molecular iodine as a catalyst in aldol
condensations.5 As a result, a mixture of oligomer derivatives of 2
and 3 has always been observed as side products throughout the
course of this work. In order to make sure that the self-polymer-
ization of the cyclic ketones does not affect the reaction outcome,
an excess of ketones (5:1 mol of ethoxyquin) was used through-
out these investigations. On the other hand, the use of KI (entry
10) or CuI (entry 11) as a substitute for molecular iodine did
not produce the expected product. However, the use of N-iodo-
succinimide (NIS, entry 12) as a substitute for molecular iodine
produces similar yield as in entries 3 and 4. As a result, the reac-
tion conditions described in entry 4 were used as the standard
conditions for subsequent reactions. It also appeared that the
reaction does not take place under a nitrogen environment (entry
15) or when the system is closed to the atmospheric air, but
proceeds just fine when run under an oxygen environment (entry
16). These data suggest that, although oxygen is needed for this
reaction to proceed, atmospheric oxygen is enough to fulfill that
requirement. As a result, the next series of reactions were run
under an open air environment.
R1
R1
O
+
I
2 (5% mol)
NH2
toluene, reflux
N
H
R2
4a:
R1 = Cl, R2 = H
R2
4b: R1 = R2 = H
Scheme 2. Synthesis of dihydroquinoline derivatives.
Cl
O
Cl
+
I
2 (5% mol.)
N
H
160°C, 72 h
neat, air
N
H
(CH2)n
5 eqv
n = 1 or 2
5a
: n = 2 (38%)
5b: n = 1 (33%)
(CH2)n
Scheme 3. Reaction between 6-chloro-1,2-dihydro-2,2,4-trimethylquinoline and
cyclohexanone or cyclopentanone.

J. Fotie et al. / Tetrahedron Letters 54 (2013) 7069–7073
7071
(CH2)n
O
+
I
2 (5% mol.)
+
N
H
160°C, 72 h
neat, air
N
H
(CH2)n
5 eqv
n = 1 or 2
N
H
6a: n = 2 (61%)
7b (13%)
6b: n = 1 (58%)
(CH2)n
Scheme 4. Reaction between 1,2-dihydro-2,2,4-trimethylquinoline and cyclohexanone or cyclopentanone.
In order to further expand the scope of the reaction, a series of
aniline derivatives were submitted to the same reaction conditions
aniline derivatives is also substituted, we concluded that steric
hindrances resulting from the presence of two methyl groups on
the nitrogen of the aniline must have precluded the ortho-cycloalk-
enylation. To further examine this hypothesis, 3-methoxy-N-
methylaniline and 2-methoxy-N-methylaniline, two molecules
bearing only one methyl group on the nitrogen atom, were allowed
to react with cyclopentanone under the conditions described
above. In these cases, the nitrogen and the para-position to the
nitrogen appeared to be systematically cyclopentenylated, but no
ortho-cyclopentenylation was observed. In fact, the reaction
between 2-methoxy-N-methylaniline and cyclopentanone yielded
N,4-dicyclopentenyl-2-methoxyaniline (10a, 42%) while the
reaction with 3-methoxy-N-methylaniline yielded N,4-dicyclopen-
tenyl-3-methoxy-N-methylaniline (10b, 36%), respectively, as
illustrated in Scheme 6. It appeared that the reaction between
the monosubstituted nitrogen atom and the ketone resulting in a
cycloalkenylated arylamine is faster than any ortho or para-alkeny-
lation of the benzene ring. This situation creates more steric
hindrances at the ortho-positions than that observed with N,N-
dimethylaniline and thus makes any ortho-alkenylation impossi-
ble. In fact, the reaction between the nitrogen atom of an aniline
derivative and the carbonyl of a ketone resulting in the formation
of a Schiff base is considered to be the very first step toward the
preparation of dihydroquinoline through a Skraup–Doebner–Von
Miller reaction.4,6 However, since the nitrogen is mono-methylated
in this case, a Schiff base cannot be produced, and instead, a
N-cyclopentenylaniline derivative is generated. Furthermore, in
the case of N,4-dicyclopentenyl-2-methoxyaniline (10a), an unex-
pected demethylation on the nitrogen atom was observed.
as
a
replacement for 1,2-dihydroquinolines. Under these
conditions, reacting a primary arylamine with a cyclic ketone will
result in a dihydroquinoline derivative as we previously reported.4a
Thus, a series of N,N-dimethylaniline derivatives were used as
starting materials instead. N,N-dimethylaniline reacted with
cyclohexanone to yield mainly 4-cyclohexenyl-N,N-dimethylani-
line (8, 54%) and 4-(1-(4-(dimethylamino)phenyl)cyclohexyl)-
N,N-dimethylaniline (9, 27%), a compound in which two molecules
of N,N-dimethylaniline are attached to the same carbon of the
cyclohexane through the para-position to the nitrogen atom
(Scheme 5). No ortho-cycloalkenylated derivative was obtained
for this reaction. The structure of compound 9 was unambiguously
determined by single crystal X-ray diffraction, and the obtained
molecular formula is shown in Figure 1.
Furthermore, the reaction between 4-methoxy-N,N-dimethyl-
aniline, 4-bromo-N,N-dimethylaniline, or 4,N,N-trimethylaniline
with either cyclopentanone or cyclohexanone under the same
conditions yielded only the starting material and the polymerized
derivatives of the cyclic ketone used; no cycloalkenylation was
observed in any of these cases. Since the para-position in all these
O
I
2 (5% mol.)
+
+
160°C, 72 h
neat, air
N
N
9 (27%)
N
On the other hand, when reacting 2-chloro-N-ethylaniline or
2-methoxy-N-ethylaniline with cyclohexanone under these condi-
tions, 2-chloro-4-cyclohexenyl-N-ethylaniline (11a) and 4-cyclo-
8 (54%)
N
hexenyl-N-ethyl-2-methoxyaniline
(11b)
were
obtained,
Scheme 5. Reaction between N,N-dimethylaniline and cyclohexanone.
respectively (Scheme 7). Unlike with cyclopentanone, no reaction
(cycloakenylation) was observed on the nitrogen atom although
it was also monosubstituted. The structural difference between
cyclopentanone and cyclohexanone might be the key factor behind
the observed difference in reactivity, with cyclohexanone being
more hindered than cyclopentanone.
Several control experiments were conducted to gain some in-
sights into the mechanism of this reaction. The addition of
5% mol of 1,1-diphenyl-2-picrylhydrazyl (DPPH)—a radical inhibi-
tor—to the reaction mixture resulted in a significantly lower yield
for 1a, dropping from 69% (Table 1, entry 4) to just 17% (Table 1,
entry 8), while the addition of 100% mol of DPPH (Table 1, entry
9) totally shut down the reaction. In fact, the GC–MS analysis of
the mixture obtained in entry 9 showed no sign of the product.
These observations suggest a possible involvement of radical inter-
mediates in the mechanism of this reaction. It is then not surpris-
ing that the substitution of I2 in the reaction mixture by KI (Table 1,
entry 10) or CuI (Table 1, entry 11) did not produce any product.
Interestingly, the substitution of I2 by azobisisobutyronitrile (AIBN)
which is a free radial initiator (Table 1, entries 13 and 14) even at a
ratio of 100% mol in relationship with 1 equivalent of ethoxyquin,
under the same reaction conditions, failed to produced the
Figure 1. Anisotropic representation of compound 9 drawn at 50% probability level.

7072
J. Fotie et al. / Tetrahedron Letters 54 (2013) 7069–7073
O
I2 (5% mol.)
+
R2
N
R
160°C, 72 h
neat, air
R2
N
H
R1
5 eqv
R1
10a: R1 = OMe, R = R2 = H (42%)
10b: R = CH3, R1 = H, R2 = OMe (36%)
Scheme 6. Reaction between 2-methoxy-N-methylaniline or 3-methoxy-N-methylaniline and cyclopentanone.
produced in the previous steps to restore the catalyst and generate
O
the enol needed for the aldolization reaction as illustrated in
Scheme 8. This hypothesis is consistent with the use of molecular
iodine as a catalyst in aldol condensations.5
I
2 (5% mol.)
+
N
H
160°C, 72 h
neat, air
N
H
R
Alternatively, the radical intermediate (I) can react with dihy-
droquinoline (or arylamine) derivatives to generate the cycloalke-
nylated product as illustrated in Scheme 9. In this case, the iodine
radical produced in the first step of the reaction reacts with the
dihydroquinoline (or arylamine) intermediate to generate an imine
radical (II) which couples with the cyclohexyl hypoiodite radical (I)
to generate intermediate III. This latter intermediate will aroma-
tize through a standard sigmatropic rearrangement ([1,3]-H shift),
resulting in IV. The involvement of free radical intermediate simi-
lar to II in iodine-catalyzed reactions has previously been
reported.3d The reaction between IV and HI followed by the elimi-
nation leads to the product 1a. Finally, the hydriodic acid and the
hypoiodic acid react in the presence of oxygen to restore the
catalyst. This is a tentative mechanistic pathway, with several
parameters that still need to be proven through much deeper
investigations.
5 eqv
R
11a: R = OMe (56%)
11b: R = Cl (40%)
Scheme 7. Reaction between 2-chloro-N-ethylaniline or 2-metoxy-N-ethylaniline
and cyclohexanone.
O
I2
O
I2
OH
O
O
I
HI
O
H
(I)
+
I
I
cyclohexenyl hypoiodite
Conclusion
Scheme 8. A tentative mechanism for the observed cyclic ketone self-aldolization.
A straightforward iodine-catalyzed regioselective cycloalkeny-
lation of dihydroquinolines and arylamines through a direct reac-
tion between a cyclic ketone and an aromatic system resulting in
a direct C–H functionalization has been developed. The major chal-
lenge for this work has been the auto-oxidation of some of the
products as they turned into dark gummy substances regardless
of the method of purification and preservation. This behavior is
consistent with previous reports on the stability of dihydroquino-
line derivatives.4a,6 Nevertheless, we were able to obtain decent 1H
NMR, 13C NMR, and HRESI-MS spectra for almost all these
compounds (see Supplementary data). This reaction offers an easy
and metal-free method for direct cycloalkenylation on arylamines
in which the double bond is selectively generated throughout the
expected product as indicated by GC–MS analysis of the reaction
mixture. This latter observation suggests that I2 is doing more than
just initiating the formation of radical intermediates in the reaction
mixture.
Based on these experimental observations and considering the
amount of dimerized cyclic ketone present in the reaction mixture
throughout the entire investigations, we hypothesized that during
the course of this reaction, I2 reacts first with the ketone to gener-
ate a free radical intermediate (I) which reacts with the iodine
radical generated in the same step, to produce a cyclohexenylhy-
poiodide. This latter entitity then reacts with hydriodic acid
OEt
O
OEt
O
OEt
N
H
H
I
I2
N
+
I
O
(III)
H
N
I
I
(I)
HI
(II)
O
OEt
OEt
OEt
I
O
H
N
H
HI +
H
I
+
N
H
1a
N
H
H
O
hypoiodous acid
(IV)
O
HI
I
H2O + I2
I
Scheme 9. A tentative reaction mechanism.

J. Fotie et al. / Tetrahedron Letters 54 (2013) 7069–7073
7073
reaction process. This reaction also enables easy structural diversi-
fications through the appropriate selection of starting materials,
which will be of interest in medicinal chemistry. Further investiga-
tions into the mechanism of the reaction that can help in broaden-
ing the scope of this reaction and turn it into a versatile catalytic
system are ongoing.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.tetlet.2013.10.081.
References and notes
1. (a) Friedel, C.; Crafts, J. M. Compt. Rend. 1877, 84, 1450–1454; (b) Friedel, C.;
Crafts, J. M. Compt. Rend. 1877, 84, 1392–1395.
2. (a) Ruepeng, M.; Nachtsheim, B. J. Beilstein, J. Org. Chem. 2010, 6; (b) Aikawa, H.;
Tago, S.; Umetsu, K.; Haginiwa, N.; Asao, N. Tetrahedron 2009, 65, 1774–1784; (c)
Devi, K. R. P. S.; Sreeja, P. B.; Sugunan, S. Int. Nano Lett. 2012, 2, 30–39.
3. (a) Kloeckner, U.; Finkbeiner, P.; Nachtsheim, B. J. J. Org. Chem. 2013, 78, 2751–
2756; (b) Finkbeiner, P.; Nachtsheim, B. J. Synthesis 2013, 45, 979–999; (c)
Puligoundla, R. G.; Karnakanti, S.; Bantu, R.; Nagaiah, K.; Kondra, S. B.;
Nagarapu, L. Tetrahedron Lett. 2013, 54, 2480–2483; (d) Dhineshkumar, J.;
Lamani, M.; Alagiri, K.; Prabhu, K. R. Org. Lett. 2013, 15, 1092–1095; (e) Gao,
Q.-H.; Fei, Z.; Zhu, Y.-P.; Lian, M.; Jia, F.-C.; Liu, M.-C.; She, N.-F.; Wu, A.-X.
Tetrahedron 2013, 69, 22–28; (f) Bakavoli, M.; Bagherzadeh, G.; Vaseghifar, M.;
Shiri, A.; Pordel, M.; Mashreghi, M.; Pordeli, P.; Araghi, M. Eur. J. Med. Chem.
2010, 45, 647–650.
4. (a) Fotie, J.; Kemami Wangun, H. V.; Fronczek, F. R.; Massawe, N.; Bhattarai, B. T.;
Rhodus, J. L.; Singleton, T. A.; Bohle, D. S. J. Org. Chem. 2012, 77, 2784–2790; (b)
Fotie, J.; Kemami Wangun, H. V.; Dreux, K.; Sommerfeld, T.; Pittman, J. Magn.
Reson. Chem. 2012, 50, 68–73; (c) Fotie, J.; Fronczek, C. F.; Burns, K. A.; Fronczek,
F. R.; Bain, C.; Bohle, D. S.; Poudeu, F. P. Acta Cryst. 2010, E66, 1660.
5. Das, B.; Thrirupathi, P.; Mahender, I.; Reddy, K. R. J. Mol. Catal. A 2006, 247, 182–185.
6. (a) Edwards, J. P.; Ringgenberg, J. D.; Jones, T. K. Tetrahedron Lett. 1998, 39, 5139–
5142; (b) Fotie, J.; Kaiser, M.; Delfín, D. A.; Manley, J.; Reid, C. S.; Paris, J. M.;
Wenzler, T.; Maes, L.; Mahasenan, K. V.; Li, C.; Werbovetz, K. A. J. Med. Chem.
2010, 53, 966–982; (c) Reid, C. S.; Patrick, D. A.; He, S.; Fotie, J.; Premalatha, K.;
Tidwell, R. R.; Wang, M. Z.; Liu, Q.; Gershkovich, P.; Wasan, K. M.; Wenzler, T.;
Brun, R.; Werbovetz, K. A. Bioorg. Med. Chem. 2011, 19, 513–523.
Acknowledgment
The authors are grateful to Dr. Frank R. Fronczek for the single
crystal X-ray diffraction analysis and Dr. Debra Dolliver for her
support.
Supplementary data
The reaction procedures and the characterization of all the com-
pounds, the 1H NMR and the 13C NMR spectra of all the compounds
described in this report, the HRESI-MS of 1a, 1b, 2, 3, 5a, 5b, 6a, 6b,
8, 9, 10b, 11a and 11b, and the EI-MS (GC-MS) of 7b and 10a are
provided assupplemental data. This material is available free of
charge via the internet at http://www.journals.elsevier.com/tetra-
hedron Letters/. CCCDC 950965 contains the supplementary crys-
tallographic data for 9. Copies of these materials can be obtained
free of charge from CCDC, 12 Union Road, Cambridge CB2 1EZ,
UK, Fax. (+44) 1223 336 033; e-mail: deposit@ccdc.cam.ac.uk.