Stereoselective disposition of the geminal dimethyl group in the cyclization of geranyl acetate under zeolite confinement conditions

Article abstract of DOI:10.1002/ejoc.200600367

The stereochemistry in the acid-catalysed biomimetic cyclization of [8,8,8-D₃]geranyl acetate was examined in solution and under conditions of zeolite Y confinement. In the intrazeolite reaction the gem-allylic methyl group adopts a diastereose

Full text of DOI:10.1002/ejoc.200600367

FULL PAPER
DOI: 10.1002/ejoc.200600367
Stereoselective Disposition of the Geminal Dimethyl Group in the Cyclization
of Geranyl Acetate under Zeolite Confinement Conditions
Constantinos Tsangarakis[a] and Manolis Stratakis*[a]
Keywords: Zeolites / Terpenes / Stereoselectivity / Catalysis / Cyclization
The stereochemistry in the acid-catalysed biomimetic cycli-
zation of [8,8,8-D3]geranyl acetate was examined in solution
and under conditions of zeolite Y confinement. In the intra-
zeolite reaction the gem-allylic methyl group adopts a dia-
stereoselective disposition in the cyclization product
(64% dr). In contrast, the gem-dimethyl disposition in a
homogeneous medium (ClSO3H/2-nitropropane) proceeds
with negligible diastereoselectivity (dr Ͻ 5%). The enhanced
diastereoselection within the zeolite is attributed to the prox-
imity of the nucleophilic double bond to the intermediate
carbocation, as a result of confinement (entropy effect).
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2006)
[8,8,8-D3]geranate produced methyl [D3]-α-cyclogeranate, in
Introduction
which two equivalent absorptions for the diastereotopic ge-
1
The fascinating enzyme-catalysed cyclization of polyene
terpenoids has attracted much interest in the bioorganic
community over the past 50 years. Product selectivity has
been shown to be a delicate synergism between substrate
conformation within the enzyme cavities (cyclases) and suit-
able positioning of the amino acid residues that initiate and
terminate the proton-induced chain reactions.[1] Moreover,
it has been established through labelling experiments that
the stereochemistry of the terminal gem-dimethyl group (of
the acyclic terpene) is disposed in a highly diastereoselective
manner upon cyclization.[2]
Organic chemists have successfully accomplished non-en-
zymatic cyclizations of terpenoids by using either
Brønsted[3] or Lewis acids[4] as catalysts. Ishihara and co-
workers prepared a chiral acidic catalyst by combining a
typical Lewis acid (SnCl4) with a chiral Brønsted acid and
achieved cyclization of terpenoids in an enantioselective
manner for the first time.[5] Apart from the cationic cycliza-
tion, a catalytic free radical-based methodology for the cas-
cade polycyclization of epoxy polyene terpenoids has re-
cently been established[6] and has been applied as a key step
in a significant number of terpenoid syntheses.[7]
minal methyls appeared in the H NMR spectrum.
The use of acidic zeolites in the cyclization of terpenoids
in the literature is limited. Sen and co-workers[9] have used
the H-forms of the small-pore zeolites A and ZSM-5 as pro-
moters for cyclizations of silyl-substituted epoxide-contain-
ing polyenes with moderate to good product selectivities.
Presumably cyclization in such porous materials occurs at
the openings of the zeolite pores, since the sizes of those
epoxy polyenes should not allow them to diffuse into the
interiors of the cages. Our group’s interests have recently
been focused on the application of the slightly acidic[10] zeo-
lite NaY as both microreactor and catalyst to achieve bio-
mimetic cyclizations of terpenoids. Zeolite NaY is a crystal-
line aluminosilicate, the primary structure of which is made
4–
5–
up of edge-sharing SiO4 and AlO4 tetrahedral,[11] the
3D zeolite framework consisting of supercages of approxi-
mately 13 Å in interior diameter. The supercages are inter-
connected by four “windows”, which are tetrahedrally dis-
tributed around each cage and have pore entrance dia-
meters of around 7.5 Å.
Unlike in the enzyme-catalysed reactions, however, the
stereochemical disposition of the gem-dimethyl groups in
the products of Brønsted acid-catalysed cyclizations of
polyene terpenoids seems to be completely non-diastereo-
selective. A Russian group reported over 30 years ago,[8] for
example, that the FSO3H-catalysed cyclization of methyl
Results and Discussion
We were pleased to find[12] that small acyclic terpenoids
such as geranyl, neryl or farnesyl derivatives undergo bio-
mimetic cyclization in moderate to excellent yield on ad-
sorption within the confined environment of NaY, dried at
120–130 °C under vacuum (10–4 Torr) for at least 6 h prior
to use. Hexane was the solvent of choice in the hetero-
geneous reaction slurry, as it has no affinity for the polar
intrazeolite environment and allows the reactant terpenes
to be adsorbed into the zeolite cavities. The reaction condi-
[a] Department of Chemistry, University of Crete,
71003 Voutes – Iraklion, Greece
E-mail: stratakis@chemistry.uoc.gr
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
Eur. J. Org. Chem. 2006, 4435–4439
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4435

C. Tsangarakis, M. Stratakis
FULL PAPER
tions resemble those already applied in other types of intra- the geminal dimethyl group in the cyclization of a model
zeolite reactions.[13] Geranyl acetate (1) cyclizes to give the terpenoid, namely [8,8,8-D3]geranyl acetate (3). The synthe-
corresponding γ- (1a) and α-cyclogeranyl (1b) isomers in sis of 3 was accomplished in nine steps, in 13% overall yield
excellent yield (Scheme 1). Initially, 1a predominates, iso- and with Ͼ97% geometrical purity with respect to the C6–
merizing to the thermodynamically more stable 1b with C7 double bond (Scheme 3), by a modification of known
prolonged reaction time or higher reaction temperature. In literature procedures.[2b,15]
addition, geranylacetone (2) is selectively transformed into
α-ambrinol (2c) by a tandem sequence involving a 1,5-diene
cyclization followed by a highly diastereoselective intramo-
lecular carbonyl-ene reaction of γ-cyclogeranylacetone (2a).
Similar cyclization results were reported almost simulta-
neously by a Chinese group[14] using zeolites NaY and FeY.
Scheme 3. Synthesis of [8,8,8-D3]geranyl acetate (3).
Initially we examined the cyclization of 3 in the presence
of ClSO3H[16] as the acid catalyst in solution in 2-nitropro-
pane containing traces of H2O[17] at –25 °C, in order to
probe the signals corresponding to the diastereotopic gem-
1
dimethyl groups in the cyclogeranyl products by H NMR.
The only products of the crude reaction mixture were the
two diastereomeric compounds 4a and 4b,[18] in a ratio of
52:48 (Scheme 4). This result concurs with stereochemical
studies previously reported for the cyclization of methyl ger-
anate.[8]
Scheme 1. Zeolite NaY-promoted cyclization of terpenoids.
The predominant formation of γ-cyclogeranyl acetate
(1a) when the reaction was carried out at ambient tempera-
ture could be envisioned as proceeding through kinetically
controlled deprotonation of the cyclized carbocation C
(Scheme 2) by oxygen in the Si–O–Si bonds in the interiors
of the zeolite supercages. The methyl group next to the cy-
clized carbocation is more accessible (than the hydrogen
atoms on the cyclohexyl ring) for proton abstraction by the
“zeolite wall”. The subsequent transformation of 1a to 1b
could be envisioned as a thermodynamically controlled
acid-catalysed process.
Scheme 4. Stereochemistry in the ClSO3H-catalysed cyclization of
3 in iPrNO2.
The lack of diastereoselection in the ClSO3H-catalysed
transformation of 3 into 4a+4b indicates that extensive ro-
tation around the C6–C7 bond of the initially formed terti-
ary carbocation CI is possible, thus forming rotamer CII
(Scheme 5). Subsequent cyclization of CI and CII via the
chair-like transition states TSI and TSII, respectively, fol-
lowed by trapping of the cyclized carbocations with the
traces of H2O from the solvent, would afford almost equi-
molar formation of 4a and 4b.
In contrast, the cyclization of 3 under zeolite confine-
ment conditions (NaY) afforded, after 5 h at ambient tem-
perature, a mixture of the deuterated γ-cyclogeranyl and α-
cyclogeranyl acetates in 80% isolated yield and in a relative
ratio of ca. 3:1 (Scheme 6, Figure 1). The diastereomeric
product ratios of 5a/5b for γ-cyclogeranyl acetate, and 6a/
6b for the α-cyclogeranyl acetate were identical (82:18).
When the reaction was performed at 70 °C for 1 h, α-cyclo-
Scheme 2. Possible origin of the intrazeolite product selectivity.
We sought to study the influence of the confined intra- geranyl acetate (6) was mainly formed, yet the dia-
zeolite environment on the stereochemical disposition of stereomeric ratio (6a/6b Ϸ 80:20) remained almost un-
4436
www.eurjoc.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2006, 4435–4439

Stereoselective Cyclization of Geranyl Acetate on Zeolite Confinement
FULL PAPER
To ensure that all the reacting substrate had been ad-
sorbed into the zeolite cavities, low loading levels were used
(n = 0.1–0.2). This loading essentially corresponds to 1
molecule of geranyl acetate per 5–10 zeolite supercages. The
intrazeolite results were identical when zeolite NaY from
three different commercial sources (Aldrich, Degussa, Zeo-
lyst) was used, while the reaction mass balance was Ͼ82%
in all experiments apart from in the reaction with HY,
where it was moderate (65%). In one experiment, the reac-
tion was stopped before completion, and the unreacted 3
was examined, but no isomerization of the C6–C7 double
bond was detectable by 1H NMR spectroscopy. The stereo-
chemistry of all the products was assigned by nOe experi-
ments. Upon irradiation of the more shielded gem-methyl
group of γ-cyclogeranyl acetate (δ = 0.87 ppm) and of the
more deshielded one in the case of α-cyclogeranyl acetate
(δ = 0.94 ppm), significant signal enhancement of the sig-
nals of the diastereotopic methylene hydrogen atoms of
the –CH2OAc functionality was observed (see Supporting
Information), indicative of a cis arrangement between
them.[19]
Scheme 5. Mechanism for the ClSO3H-catalysed cyclization of 3.
In addition, for the reaction in solution (ClSO3H, 2-ni-
tropropane), we wanted to exclude the possibility that the
gem-dimethyl group might initially dispose stereoselectively
(but not stereospecifically), but that the cyclogeranyl carbo-
cation might revert to a mixture of CI and CII through a
possible reversibility in the cyclization step (Scheme 5), with
gradual loss of diastereoselectivity thus taking place. For
this purpose we treated a mixture containing mainly 5a and
5b (dr = 64%) produced from the intrazeolite reaction of
3 at ambient temperature (Scheme 6) with ClSO3H in dry
iPrNO2 under experimental conditions strictly identical to
those applied in the cyclization of 3 in solution. A mixture
of 6a/6b was isolated in a ratio of ca. 80:20, which implies
the lack of such a mechanistic scenario.
changed (see Supporting Information). Moreover, a ratio of
6a/6b ca. 80:20 was again observed when 3 was confined
within the highly acidic zeolite HY (25 °C, 1 h).
We propose that this interesting enhanced diastereoselec-
tivity seen in the cyclization of 3 in NaY and HY might be
attributable to a change in the energy reaction profile on
going from the homogeneous solution to the zeolite envi-
ronment. In solution, the initially formed tertiary carbo-
cation CI (Scheme 5) presumably has a barrier to rotation
around the C6–C7 bond approximately 2 kcalmol–1 lower
than the activation energy of its accompanying cyclization
step, so almost 95% of the CI population undergoes revers-
ible rotation around the C6–C7 bond prior to cyclization,
and essentially a product ratio of 4a/4b = 52:48 results. In
the NaY or HY, however, the stereochemical outcome
would require that around 40% of the CI population should
undergo such a rotation prior to the nucleophilic attack by
the C2–C3 double bond. In other words, the barrier to rota-
tion around the C6–C7 bond of CI in the intrazeolite reac-
tion is slightly higher than the activation energy for the cyc-
lization. These energy profile changes can be interpreted in
terms of an entropy effect. In the confined zeolite cavity,
the C2–C3 double bond is on average closer to the initially
formed carbocation than is the case in the reaction in solu-
tion. This proximity contributes to a significant decrease in
the activation energy of the cyclization step. We postulate
Scheme 6. Intrazeolite cyclization of 3 at 20 °C.
1
Figure 1. Part of the H NMR spectrum of a mixture of products
5a, 5b, 6a and 6b in the 0.8–1.1 ppm region.
Eur. J. Org. Chem. 2006, 4435–4439
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4437

C. Tsangarakis, M. Stratakis
FULL PAPER
that the remarkable stereochemical dispositions[2] of the ter-
minal gem-dimethyl groups in enzyme-catalysed cyclizations
of terpenoids imply barriers to rotation around the
C–Ccarbocation bonds in the transient dimethyl-substituted
carbocations at least 3 kcalmol–1 higher than the activation
energies of the accompanying C–C bond-formation steps.
The confined environment of an enzyme cavity should fav-
our such an energy profile, due to the close proximity of
the reacting double bond and the carbocation, so the cycli-
zation step would be almost barrierless.[1e–1f]
An alternative explanation for the observed diastereose-
lection in the intrazeolite cyclization reaction might be the
major diastereomers (5a or 6a) arising from chair-like tran-
sition states, while the minor diastereomers (5b and 6b)
arise from boat-like transition states (Scheme 7). This sce-
nario would require that rotation around the C6–C7 bond
of CI be negligible (Scheme 5). Although it is difficult to
distinguish between the two proposed mechanisms, the sec-
ond explanation is less likely since the energy difference be-
tween a boat- and a chair-like transition state would be ex-
pected to be much higher than 0.8 kcalmol–1.[20] Neverthe-
less, the current results clearly establish that, within zeolite
Y, rotation around the C6–C7 bond of the initially formed
carbocation is slower than its cyclization rate, as a result of
the proximity of the nucleophilic C2–C3 double bond.
anol (2×5 mL for 30 min each time) and then filtered again. The
combined solvent extracts were evaporated to afford the cyclized
products α- and γ-cyclogeranyl acetate (16 mg, 80%). The 1H
NMR spectroscopic data for the labelled γ-cyclogeranyl and α-cy-
clogeranyl acetate (5a/5b and 6a/6b, respectively) matched the spec-
troscopic data for 1a and 1b reported earlier by us,[12] with the only
differences appearing in the region of 0.8–1.1 ppm, where the gem-
methyl group absorptions reflect the ratios of 5a/5b and 6a/6b.
Cyclization of [8,8,8-D3]Geranyl Acetate (3) Promoted by ClSO3H
in 2-Nitropropane: [8,8,8-D3]Geranyl acetate (20 mg, 0.1 mmol) was
added at –25 °C to a solution of ClSO3H (0.18 µL, 0.26 mmol) in 2-
nitropropane (as obtained from commercial sources, 0.3 mL). After
30 min, triethylamine (0.1 mL) dissolved in diethyl ether (2 mL)
was added and the solution was washed with brine. The organic
layer was dried with MgSO4, the solvent was removed under vac-
uum, and the oily residue was chromatographed (hexane/ethyl ace-
tate, 4:1) to afford the diastereomeric (2-hydroxy-2,6,6-trimethylcy-
clohexyl)methyl [D3]acetate (4a/4b, 12 mg, 53% isolated yield) as a
1
colourless oil. The H NMR spectroscopic data for the mixture of
the labelled 4a/4b matched the spectroscopic data of the unlabelled
cyclic alcohol reported by us earlier.[12]
Supporting Information (see also the footnote on the first page of
this article): Experimental details for the synthesis of [8,8,8-D3]ger-
1
anyl acetate (3). Copies of H and 13C NMR spectra for the inter-
mediate compounds in the synthesis of 3, cyclization reactions, and
nOe experiments.
Acknowledgments
This project was co-funded by the European Social Fund and
national resources (program PYTHAGORAS-II).
[1] a) E. E. van Tamelen, Acc. Chem. Res. 1975, 8, 152–158; b)
W. S. Johnson, Bioorg. Chem. 1976, 5, 51–98; c) I. Abe, M.
Rohmer, G. D. Prestwich, Chem. Rev. 1993, 93, 2189–2206; d)
C. A. Lesburg, J. M. Caruthers, C. M. Paschall, D. W. Chris-
tianson, Curr. Opin. Struct. Biol. 1998, 8, 695–703; e) K. U.
Wendt, G. E. Schulz, E. J. Corey, D. R. Liu, Angew. Chem.
2000, 112, 2930–2952; Angew. Chem. Int. Ed. 2000, 39, 2812–
2833; f) R. Rajamani, J. Gao, J. Am. Chem. Soc. 2003, 125,
12768–12781; g) R. A. Yoder, J. N. Johnston, Chem. Rev. 2005,
105, 4730–4756.
Scheme 7.
Conclusions
[2] a) S. S. Mohanty, P. Uebelhart, C. H. Eugster, Helv. Chim. Acta
2000, 83, 2036–2053; b) M. L. Wise, H.-J. Pyun, G. Helms, B.
Assink, R. M. Coates, R. B. Croteau, Tetrahedron 2001, 57,
5327–5334; c) M. L. Wise, M. Urbansky, G. L. Helms, R. M.
Coates, R. Croteau, J. Am. Chem. Soc. 2002, 124, 8546–8547;
d) Q. Jin, D. C. Williams, M. Hezari, R. Croteau, R. M.
Coates, J. Org. Chem. 2005, 70, 4667–4675.
In conclusion, we have shown that the confined environ-
ment of the zeolite Y significantly alters the reaction energy
profile for the cyclization of a model terpenoid (geranyl ace-
tate) relative to the reaction in solution.
[3] P. F. Vlad, Pure Appl. Chem. 1993, 65, 1329–1336, and refer-
Experimental Section
ences therein.
[4] H. Yamamoto, K. Futatsugi, Angew. Chem. 2005, 117, 1958–
1977; Angew. Chem. Int. Ed. 2005, 44, 1924–1942.
General: Nuclear magnetic resonance spectra were obtained on a
500 MHz instrument. Isomeric purities were determined by 1H
NMR, 13C NMR and by GC or GC-MS on an HP-5 capillary
column. All spectra reported here were taken in CDCl3.
[5] a) K. Ishihara, S. Nakamura, H. Yamamoto, J. Am. Chem.
Soc. 1999, 121, 4906–4907; b) H. Ishibashi, K. Ishihara, H.
Yamamoto, J. Am. Chem. Soc. 2004, 126, 11122–11123.
[6] a) J. Justicia, A. Rosales, E. Bunuel, J. L. Oller-Lopez, N. Valdi-
via, A. Haidour, J. E. Oltra, A. F. Barrero, D. J. Cardenas, J. M.
Cuerva, Chem. Eur. J. 2004, 10, 1778–1788; b) J. Justicia, J. L.
Oller-Lopez, A. G. Campana, J. E. Oltra, J. M. Cuerva, E. Bu-
nuel, D. J. Cardenas, J. Am. Chem. Soc. 2005, 127, 14911–
14921.
Intrazeolite Cyclization of [8,8,8-D3]Geranyl Acetate (3): [8,8,8-D3]-
Geranyl acetate (3, 20 mg) was added to a slurry containing NaY
(0.5 g), previously dried[12] at 120 °C for at least 6 h under vacuum
(10–4 Torr), in hexane (5 mL), and the heterogeneous mixture was
either stirred at room temperature for 5–6 h or heated to 70 °C for
1 h. After that period, the reaction mixture was filtered, and the
filtrate was kept. The solid material was further washed with meth-
[7] a) A. F. Barrero, J. F. Q. del Moral, E. M. Sanchez, J. F. Arte-
aga, Eur. J. Org. Chem. 2006, 10, 1627–1641; b) J. M. Cuerva,
4438
www.eurjoc.org
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2006, 4435–4439

Stereoselective Cyclization of Geranyl Acetate on Zeolite Confinement
FULL PAPER
J. Justicia, J. L. Oller-Lopez, J. E. Oltra, Top. Curr. Chem. 2006,
264, 63–91.
[8] M. T. Mustafaeva, V. A. Smit, A. V. Semenovskii, V. F. Kuch-
erov, Izv. Akad. Nauk. SSSR-Ser. Khim. 1973, 5, 1151–1153;
Chem. Abstr. 1973, 79, 77795.
[9] a) S. E. Sen, Y.-Z. Zhang, S. L. Roach, J. Org. Chem. 1996, 61,
9534–9537; b) S. E. Sen, Y.-Z. Zhang, S. M. Smith, J. Org.
Chem. 1998, 63, 4459–4465.
[10] V. J. Rao, D. L. Perlstein, R. J. Robbins, P. H. Lakshminarasim-
han, H.-M. Kao, C. P. Grey, V. Ramamurthy, Chem. Commun.
1998, 269–270.
moderate signal enhancement for the tertiary allylic hydrogen
atom and a significant one for the methylene hydrogen atoms
of the –CH2OAc functionality. The opposite enhancement
trends were found upon irradiation of the more deshielded
methyl group, resonating at δ = 0.99 ppm. The nOe enhance-
ment trends presented above indicate that the methyl group
resonating at δ = 0.87 ppm is trans to the tertiary allylic hydro-
gen atom, and the methyl group absorbing at δ = 0.99 ppm is
cis. This argument can be explained by considering an equilib-
rium of conformers A and B for 1a, with A being predominant.
[11] A. Dyer, An Introduction to Zeolite Molecular Sieves, Wiley,
Bath, UK, 1988.
[12] C. Tsangarakis, M. Stratakis, Adv. Synth. Catal. 2005, 347,
1280–1284.
[13] V. Ramamurthy, P. Lakshminarasimhan, C. P. Grey, L. J. John-
ston, Chem. Commun. 1998, 2411–2418.
[14] W. Yu, F. Bian, Y. Gao, L. Yang, Z.-L. Liu, Adv. Synth. Catal.
2006, 348, 59–62.
[15] S. Muto, Y. Nishimura, K. Mori, Eur. J. Org. Chem. 1999,
The pseudoequatorial gem-methyl group (Me1) of conformer
A, which after ring inversion is converted to pseudoaxial in
conformer B, would be expected upon irradiation to enhance
the signal of the –CH2OAc hydrogens significantly and the sig-
nal of the tertiary allylic hydrogen moderately. On the other
hand, the pseudoaxial methyl group (Me2) of conformer A
(pseudoequatorial in conformer B) would be expected upon
irradiation to give a significant signal enhancement of the terti-
ary allylic hydrogen atom and a moderate one for the
–CH2OAc. Similar nOe analysis holds for 1b, which establishes
that the gem-methyl cis to the –CH2OAc resonates at δ =
0.94 ppm, and the trans one at δ = 0.92 ppm. Moreover, from
a combination of DEPT and 2D-NMR techniques it was found
that the tertiary allylic hydrogen atom of 1a resonates at δ =
2.16 ppm and that of 1b at δ = 1.72 ppm.
2159–2165.
[16] P. J. Linares-Palomino, S. Salido, J. Altarejos, A. Sanchez, Tet-
rahedron Lett. 2003, 44, 6651–6655.
[17] The 2-nitropropane was used as received from commercial
sources and contains traces of H2O.
[18] The perprotio analogue of the diastereomeric deuterated
alcohols 4a and 4b is a known compound,[12] and is formed
stereoselectively in the halosulfonic acid-catalysed cyclization
of geranyl acetate (see ref.[3]).
[19] The γ- and α-cyclogeranyl derivatives preferentially adopt con-
formations in which the –CH2OAc groups adopt pseudoaxial
orientations (difference of approximately 0.5 kcalmol–1 relative
to the pseudoequatorial). See: a) R. L. Snowden, J.-C. Eichen-
berger, S. M. Linder, P. Sonnay, C. Vial, K. H. Schulte-Elte, J.
Org. Chem. 1992, 57, 955–960; b) C. Fehr, Angew. Chem. 1998,
110, 2509–2512; Angew. Chem. Int. Ed. 1998, 37, 2407–2409.
Irradiation of the more shielded gem-methyl group in γ-cyclo-
geranyl acetate (1a), resonating at δ = 0.87 ppm, results in a
[20] An energy difference of 0.8 kcalmol–1 between two transition
states reflects a product ratio of around 80:20.
Received: April 27, 2006
Published Online: August 9, 2006
Eur. J. Org. Chem. 2006, 4435–4439
© 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4439