Fragmentation of carbohydrate anomeric alkoxy radicals: A new synthesis of chiral 1-halo-1-iodo compounds

Article abstract of DOI:10.1002/1521-3773(20010618)40:12<2326::AID-ANIE2326>3.0.CO;2-Y

One less carbon atom is found in 1-halo-1-iodo compounds obtained by C1-C2 radical fragmentation of carbohydrate 1,2-halohydrins. This fragmentation is achieved via the anomeric alkoxy radicals of the halohydrins, formed upon reaction with (diacetoxyiodo)

Full text of DOI:10.1002/1521-3773(20010618)40:12<2326::AID-ANIE2326>3.0.CO;2-Y

COMMUNICATIONS
Fragmentation of Carbohydrate Anomeric
Alkoxy Radicals: A New Synthesis of Chiral
1-Halo-1-iodo Compounds**
Â
Â
Concepcion C. Gonzalez, Alan R. Kennedy,
Â
Â
Elisa I. Leon, Concepcion Riesco-Fagundo, and
Ernesto Suarez*
Â
1,1-Dihaloalkanes, and specifically the highly reactive 1,1-
diiodoalkanes, are versatile compounds in organic synthesis.[1]
Several methods for the synthesis of these compounds have
been developed.[2] Among them are the iodolysis of 1,1-
bis(diisobutylalumino)alkanes,[2a] the alkylation of diiodo-
methyllithium or diiodomethylsodium with reactive electro-
philes,[2b] and the reaction of 1,1-bis(trifluoromethyl)sulfonyl-
oxy-alkanes with magnesium iodide.[2c] None of these ap-
proaches can be easily applied to the preparation of complex
or sensitive molecules. Consequently, in the majority of cases
the 1,1-diiodoalkanes obtained are derivatives of relatively
simple hydrocarbons.[3] A similar situation is found for the
synthesis of mixed halo-iodo species, for example 1-fluoro-1-
iodo,[4] 1-chloro-1-iodo,[5] and 1-bromo-1-iodo compounds.[6]
Moreover, there are sufficient drawbacks to most of these
procedures to justify the need for a general and practical
method of generating these types of compounds under mild
conditions.
Scheme 1. The mechanism of alkoxy radical fragmentation (ARF).
a) R1 ˆ O-alkyl; R2 ˆ protecting group. b) R1 ˆ OC(O)-alkyl, halogen;
R2 ˆ protecting group. See text for details.
The a-iodoalkyl esters thus formed may be interesting chiral
synthons.[7f]
With the above results in mind, we decided to introduce a
halogen atom at C2 in order to develop an advantageous
methodology for preparing 1-halo-1-iodo compounds. We
carried out the synthesis of 1,2-halohydrins of carbohydrates
in pyranose and furanose form as outlined in Table 1. 1,2-
Chlorohydrins, 1,2-bromohydrins, and 1,2-iodohydrins were
prepared from the corresponding 2-deoxy-hex-1-enitol by
reaction with N-chlorosuccinimide,[9] N-bromoacetamide,[10]
Table 1. Synthesis of 1-halo-1-iodo compounds.[a]
Earlier research from our laboratory[7] has shown the facile
formation of glycopyran-1-O-yl and glycofuran-1-O-yl radi-
cals by reaction of carbohydrate anomeric alcohols with
hypervalent iodine reagents in the presence of iodine. The
reactions presumably proceeds through an alkyl hypoiodite
intermediate.[8] Subsequently the alkoxy radical undergoes
fragmentation of the C1 C2 bond and gives rise to a C2
radical (Scheme 1). The substituent at C2 may have a strong
influence on the ultimate fate of the radical. When the
substituent is an ether group the radical is rapidly oxidized by
an excess of the hypervalent iodine reagent to give an
oxycarbenium ion (path a). Inter- or intramolecular trapping
by nucleophiles leads to a variety of modified carbohydrate
derivatives with one carbon atom less.[7] The presence of a
stronger electron-withdrawing group at C2 (e.g., an ester
group, path b) decreases the electron density at this position
and oxidation of the radical should be more difficult. This
opens the possibility of competitive trapping of the inter-
mediate radical by atoms of iodine from the reaction medium.
Entry Substrate
t [h] Product
Yield [%] (dr)
1
2
3
4
1 R ˆ F
2 R ˆ Cl
3 R ˆ Br
4 R ˆ I
1
3
1.5
5 R ˆ F
6 R ˆ Cl
7 R ˆ Br
96 (1:1)
95 (1:1)
99 (1:1)
92
0.75 8 R ˆ I
5
6
7
9 R ˆ Cl
10 R ˆ Br
11 R ˆ I
1.5
2
0.5
12 R ˆ Cl
96 (1:1)
98 (1:1)
84
13 R ˆ Br
14 R ˆ I
Â
Â
Â
[*] Prof. E. Suarez, Dr. C. C. Gonzalez, Dr. E. I. Leon, C. Riesco-Fagundo
Instituto de Productos Naturales y Agrobiología del C.S.I.C.
Carretera de La Esperanza 3, 38206 La Laguna, Tenerife (Spain)
Fax : (‡34)922-260135
8
9
10
15 R ˆ Cl
16 R ˆ Br
17 R ˆ I
1
1.5
1
18 R ˆ Cl (R and S)
19 R ˆ Br (R and S)
20 R ˆ I
92 (3:2)
93 (3:2)
90
E-mail: esuarez@ipna.csic.es
Dr. A. R. Kennedy
Department of Pure and Applied Chemistry
University of Strathclyde, Glasgow (UK)
[**] This work was supported by the Investigation Programme no. PB96-
11
12
13
21 R ˆ Cl
22 R ˆ Br
23 R ˆ I
0.5
0.5
0.5
24 R ˆ Cl
25 R ˆ Br
26 R ˆ I
83 (3:2)
92 (3:2)
69
Â
Â
Â
1461 of the Direccion General de Investigacion Científica y Tecnica,
Â
Spain. C.C.G. and C.R.-F. thank the Ministerio de Educacion y
Â
Ciencia, Spain, and the Direccion General de Universidades
e
Â
Investigacion del Gobierno de Canarias, respectively, for fellowships.
[a] Halohydrin (1 mmol) in CH2Cl2 (50 mL) containing DIB (1.5 mmol) and iodine
Supporting information for this article is available on the WWW under (1.5 mmol) was irradiated with two 80-W tungsten filament lamps at reflux
http://www.angewandte.com or from the author.
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
temperature.
2326
1433-7851/01/4012-2326 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2001, 40, No. 12

COMMUNICATIONS
and N-iodosuccinimide,[11] respectively, in THF/water. The
fluorohydrin 1 was synthesized by reaction of 3,4,6-tri-O-
acetyl-1,5-anhydro-2-deoxy-d-arabino-hex-1-enitol with xe-
non difluoride,[12] followed by acid hydrolysis of the anomeric
fluoride, peracetylation, and selective anomeric O-deacetyla-
tion with hydrazine acetate.[13] As detected spectroscopically,
the halohydrins were diastereoisomeric mixtures in most
cases.[10] Chromatographically homogeneous products which
gave correct elemental analyses were used in the fragmenta-
tion reaction.
Scheme 2. Alkoxy radical fragmentation of halohydrin 27. a) DIB
(1.5 equiv), iodine (1.5 equiv), CH2Cl2, hn, reflux, 1 h, 93%. DIB ˆ (di-
acetoxyiodo)benzene.
The alkoxy radical fragmentation (ARF) reactions were
performed under the conditions stated in Table 1, with
(diacetoxyiodo)benzene and iodine in CH2Cl2 at reflux
temperature and irradiation with two 80-W tungsten filament
lamps. Complete consumption of the starting material was
observed in all cases. In entries 1 ± 4 of Table 1 we compare
the reaction of several halohydrins of the 2-deoxy-2-halo-
glucopyranose type. The reaction proceeded smoothly in
excellent yield to give 1-halo-1-iodo-arabinitol derivatives 5 ±
8, without affecting the stereogenic integrity of the adjacent
center. As expected, diastereoselection was very low for the
mixed halogen compounds, which were obtained as a
chromatographically inseparable mixture of isomers. These
compounds seem to be stable for long periods of time and
were handled without any special precautions apart from
avoiding overexposure to light or heat. The galacto halohy-
drins 9 ± 11 provided a new series of 5-halo-5-iodo-arabinitol
derivatives 12 ± 14, also in excellent yields and without any
diastereoselectivity (entries 5 ± 7).
configuration at C1 for the mixed 1-halo-1-iodo compounds
prepared previously.
Preliminary attempts to use these compounds in a Takai
(E)-olefination[15] of aldehydes were unsuccessful. For exam-
ple, when compounds 8 and 20 were treated with hydro-
cinnamaldehyde with the complex chromium(ii) chloride ±
DMF as reagent, only a mixture of the corresponding (Z)-
and (E)-1-iodoalkenes was obtained by b elimination. Never-
theless, when pentitol 31[16] was submitted to the above
conditions the expected (E)-olefin 32 was obtained in good
yield (Scheme 3). It is worth noting the stability of the highly
Several differently protected galactopyranose halohydrins,
15 ± 17, were prepared in order to study the influence of the
protecting groups on the selectivity of the fragmentation
reaction (entries 8 ± 10). A modest increment in the diastero-
selectivity is observed as the steric demand of the starting
halohydrin increases (compare entries 5 vs. 8 and 6 vs. 9). The
mixed 1-chloro-1-iodo and 1-bromo-1-iodo diastereoisomers
18 and 19 can now be separated by chromatotron chromatog-
raphy. The configuration at C1 was tentatively assigned on the
basis of the significant deshielding of the C3 resonance
(dS ± dR ˆ 1.8 ± 2.8) observed in the 13C NMR spectra of the S
series (see below).
To further extend the scope of the described method, we
investigated the feasibility of applying it to the five-membered
glucofuranose halohydrins 21 ± 23 (entries 11 ± 13). In all cases
the reaction proceeded analogously to give a new set of halo-
iodo-arabino derivatives 24 ± 26 with a very different protec-
tion pattern.
Since neither of the diastereoisomers of 18 or 19 could be
crystallized, we prepared 3,5-dinitrobenzoate 27 (Scheme 2)
in order to establish the absolute configuration at C1 by X-ray
diffraction experiments. The diastereoisomeric mixture ob-
tained (dr 2:1) after the ARF reaction was separated by
chromatography to provide the 5-bromo-5-iodo-arabinitol
derivatives 28 and 29. Crystallization of 28 from EtOAc/
n-hexane yielded colorless crystals suitable for X-ray
studies, which confirmed the R configuration at C1.[14] The
observed deshielding of the C3 signal in the 13C NMR
spectrum of 29, possibly due to restricted rotation around
the C1 C2 bond, allowed us to tentatively assign the
Scheme 3. Alkoxy radical fragmentation of halohydrin 30. a) DIB
(1.5 equiv), iodine (1.5 equiv), CH2Cl2, hn, reflux, 30 min, 94%; b) CrCl2
(4 equiv), hydrocinnamaldehyde (Ph(CH2)2CHO; 2 equiv), DMF
(4 equiv), THF, RT, 30 min, 70%.
sensitive formyl ester under the reaction conditions. In the
majority of cases, this reaction has been used in synthetic
organic chemistry as an (E)-ethylenation of aldehydes with
diiodoethane.[17]
The new ARF reaction presented here offers special
advantages for the synthesis of gem-dihalo compounds,
including high efficiency, procedural simplicity, and mild
reaction conditions that are compatible with the protecting
groups most frequently used in carbohydrate chemistry.[18] It is
hoped that these 1-halo-1-iodo compounds will be powerful
building blocks for organic synthesis since the carbohydrate
would potentially be amenable to prior manipulation to
provide more specific or complex systems.
Received: December 28, 2000
Revised: March 27, 2001 [Z16345]
[1] P. Helquist in Comprehensive Organic Synthesis, Vol. 4 (Eds.: B. M.
Trost, I. Fleming, M. F. Semmelhack), Pergamon, Oxford, 1991,
pp. 961 ± 976.
[2] a) L. Aufauvre, P. Knochel, I. Marek, Chem. Commun. 1999, 2207 ±
2208; b) P. Charreau, M. Julia, J. N. Verpeaux, Bull. Soc. Chim. Fr.
Â
1990, 127, 275 ± 282; c) A. G. Martínez, A. H. Fernandez, R. M.
Â
Alvarez, A. G. Fraile, J. B. Calderon, J. O. Barcina, Synthesis 1986,
Angew. Chem. Int. Ed. 2001, 40, No. 12
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001 1433-7851/01/4012-2327 $ 17.50+.50/0
2327

COMMUNICATIONS
1076 ± 1078; d) D. H. R. Barton, G. Bashiardes, J.-L. Fourrey, Tetrahe-
dron 1988, 44, 147 ± 162.
[3] a) J. A. Rechka, J. R. Maxwell, Tetrahedron Lett. 1988, 29, 2599 ± 2600;
b) A. B. Charette, J. Lemay, Angew. Chem. 1997, 109, 1163 ± 1165;
Angew. Chem. Int. Ed. Engl. 1997, 36, 1090 ± 1092.
[4] P. D. Bailey, S. R. Baker, A. N. Boa, J. Clayson, G. M. Rosair,
Tetrahedron Lett. 1998, 39, 7755 ± 7758; P. D. Bailey, A. N. Boa,
G. A. Crofts, Tetrahedron Lett. 1989, 30, 7457 ± 7460.
A Schizophrenic Water-Soluble Diblock
Copolymer**
Shiyong Liu, Norman C. Billingham, and
Steven P. Armes*
Recently we reported the first example of a water-soluble
diblock copolymer capable of existing in three states in
aqueous solution, namely, as conventional micelles, reverse
micelles, and molecularly dissolved (non-micellar) chains.[1]
This diblock copolymer was based on two tertiary amine
methacrylates, 2-(diethylamino)ethyl methacrylate (DEA)
and 2-(N-morpholino)ethyl methacrylate (MEMA), and was
synthesized using group transfer polymerization,[2, 3] a type of
anionic polymerization which is particularly well suited to the
living polymerization of methacrylates at room temperature.
Formation of micelles with DEA cores was achieved merely
by adjusting the solution pH value, but formation of the
reverse micelles with MEMA cores required the addition of a
large amount of electrolyte to selectively ªsalt outº the
MEMA chains. To date, this remains the only well-docu-
mented example of such a ªschizophrenicº block copolymer.[4]
Since its discovery in 1995,[5] atom-transfer radical polymer-
ization (ATRP) has proved to be a reliable and versatile
method for the synthesis of functional, controlled-structure
copolymers.[6] This free-radical polymerization chemistry is
ªpseudo-livingº and particularly tolerant of monomer func-
tionality; it has been used to polymerize a wide range of
hydrophilic monomers with narrow molecular weight distri-
butions and controlled architectures.[7]
Herein we describe the facile ATRP synthesis of a new
diblock copolymer based on poly(propylene oxide) (PPO)
and DEA (Scheme 1a). This diblock copolymer dissolves
molecularly in cold aqueous solution but undergoes reversible
micellar self-assembly to give either PPO-core micelles or
DEA-core micelles. Unlike the DEA ± MEMA diblock co-
polymer reported previously, both types of micelles can be
formed solely by the judicious selection of solution pH value
and solution temperature.
It is well known that PPO with an Mn of around 2000
dissolves in cold, dilute aqueous solution but becomes
insoluble at 208C; its lower critical solution temperature
(LCST) lies between 108C and 208C, depending on the
solution concentration.[8] Similarly, we have recently shown
that DEA homopolymer is soluble in acidic solution as a weak
cationic polyelectrolyte (due to protonation of the tertiary
amine residues), but precipitates from solution at around
neutral pH. We have recently published several papers
[5] R. C. Neuman, M. L. Rahm, J. Org. Chem. 1966, 31, 1857 ± 1859.
[6] J. E. Lyons, J. Chem. Soc. Chem. Commun. 1975, 418 ± 419.
Â
[7] a) P. Armas, C. G. Francisco, E. Suarez, Angew. Chem. 1992, 104, 746 ±
748; Angew. Chem. Int. Ed. Engl. 1992, 31, 772 ± 774; b) P. Armas,
Â
C. G. Francisco, E. Suarez, J. Am. Chem. Soc. 1993, 115, 8865 ± 8866;
Â
Â
Â
c) R. Hernandez, E. I. Leon, P. Moreno, E. Suarez, J. Org. Chem. 1997,
Â
62, 8974 ± 8975; d) C. G. Francisco, R. Freire, C. C. Gonzalez, E.
Â
Suarez, Tetrahedron: Asymmetry 1997, 8, 1971 ± 1974; e) C. G. Fran-
Â
Â
cisco, C. C. Gonzalez, E. Suarez, J. Org. Chem. 1998, 63, 2099 ± 2109;
Â
Â
f) C. G. Francisco, C. C. Gonzalez, E. Suarez, J. Org. Chem. 1998, 63,
8092 ± 8093.
Â
[8] J. L. Courtneidge, J. Lusztyk, D. Page, Tetrahedron Lett. 1994, 35,
1003 ± 1006.
Á
[9] J. Rodriguez, J.-P. Dulcere, Synthesis 1993, 1177 ± 1205.
[10] C. H. Marzabadi, C. D. Spilling, J. Org. Chem. 1993, 58, 3761 ± 3766.
[11] M. Smietana, V. Gourverneur, C. Mioskowski, Tetrahedron Lett. 2000,
41, 193 ± 195.
[12] W. Korytnyk, S. Valentekovic-Horvath, C. R. Petrie III, Tetrahedron
1982, 38, 2547 ± 2550.
[13] G. Excoffier, D. Gagnaire, J.-P. Utille, Carbohydr. Res. 1975, 39, 368 ±
373.
[14] Crystallographic data (excluding structure factors) for the structure 28
reported in this paper have been deposited with the Cambridge
Crystallographic Data Centre as supplementary publication no.
CCDC-155002. Copies of the data can be obtained free of charge on
application to CCDC, 12 Union Road, Cambridge CB21EZ, UK (fax:
(‡44)1223-336-033; e-mail: deposit@ccdc.cam.ac.uk).
[15] T. Okazoe, K. Takai, K. Utimoto, J. Am. Chem. Soc. 1987, 109, 951 ±
953.
[16] The 3-deoxy-d-glucal precursor of 31 was prepared in two steps by
BF3-catalyzed rearrangement of commercially available tri-O-acetyl-
d-glucal and LiAlH4 reduction of the hex-2-enopyranoside formed
(84% overall yield): R. J. Ferrier, Methods Carbohydr. Chem. 1972, 6,
307 ± 311; B. Fraser-Reid, S. Y.-K. Tam, B. Radatus, Can. J. Chem.
1975, 53, 2005 ± 2016. A one-step method is available as well, also
starting from tri-O-acetyl-d-glucal, although in lower yield (58%): N.
Greenspoon, E. Keinan, J. Org. Chem. 1988, 53, 3723 ± 3731.
[17] For example, D. Meng, Q. Tan, S. J. Danishefsky, Angew. Chem. 1999,
111, 3393 ± 3397; Angew. Chem. Int. Ed. 1999, 38, 3197 ± 3201; C. M.
Blackwell, A. H. Davidson, S. B. Launchbury, C. N. Lewis, E. M.
Morrice, M. M. Reeve, J. A. R. Roffey, A. S. Tipping, R. S. Todd, J.
Org. Chem. 1992, 57, 5596 ± 5606; R. Baker, J. L. Castro, J. Chem. Soc.
Perkin Trans. 1 1990, 47 ± 65.
[18] Recently, the use of DIB/I2 for the deprotection of carbohydrate
benzyl ethers was described: J. Madsen, C. Viuf, M. Bols, Chem. Eur. J.
2000, 6, 1140 ± 1146.
[*] Prof. S. P. Armes, Dr. S. Liu, Prof. N. C. Billingham
School of Chemistry, Physics and Environmental Science
University of Sussex
Falmer, Brighton, E. Sussex, BN1 9QJ (UK)
Fax : (‡44)1273-677-196
E-mail: s.p.armes@sussex.ac.uk
[**] We thank EPSRC for a post-doctoral fellowship for S.Y.L. (GR/
N17409). Dr. P. McKenna of Laporte Performance Chemicals, Hythe
(UK), is thanked for the gift of the monohydroxy-capped PPO. The
reviewers are thanked for their helpful and constructive comments.
Supporting information for this article is available on the WWW under
http://www.angewandte.com or from the author.
2328
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2001
1433-7851/01/4012-2328 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2001, 40, No. 12