Double ring-opening reactions of bis-cyclopropanes promoted by iron salts

Article abstract of DOI:10.1055/s-1999-2582

Certain bis-cyclopropanes were prepared from the corresponding dienol silane and, on treatment with ferric salts, were found to undergo two sequential C-C bond cleavages to give products corresponding to overall two carbon ring-expansion. In one case the nature of the product obtained was found to depend upon the types of ferric salts employed.

Full text of DOI:10.1055/s-1999-2582

LETTER
237
Double Ring-Opening Reactions of Bis-Cyclopropanes Promoted by Iron Salts
Adrian J. Highton,a Tahir N. Majidb and Nigel S. Simpkinsa*
aSchool of Chemistry, The University of Nottingham, University Park, Nottingham, NG7 2RD, UK
Fax +44(0115) 9513564; E-mail: Nigel.Simpkins@Nottingham.ac.uk
bRhône-Poulenc Rorer Limited, Rainham Road South, Dagenham, Essex RM10 7XS, UK
Received 20 November 1998
Abstract: Certain bis-cyclopropanes were prepared from the corre-
sponding dienol silane and, on treatment with ferric salts, were
found to undergo two sequential C–C bond cleavages to give prod-
ucts corresponding to overall two carbon ring-expansion. In one
case the nature of the product obtained was found to depend upon
the types of ferric salts employed.
Scheme 2
Key words: cyclopropane, ring expansion, iron salts
Our plan was to elaborate readily available dienol silanes
A protocol for overall ring expansion of a cyclic ketone by
(Danishefsky dienes)7 such as 4 by initial cyclopropana-
one carbon was described by Saegusa and co-workers in
tion of the most electron-rich double bond, and then to in-
stall either a cyclopropane, epoxide or aziridine by
reaction of the remaining double bond. With 5 (X = CH2,
O or NR) in hand we would explore reactions with iron
1976.1 Initial enol silane formation and cyclopropanation
serves to convert a ketone such as cyclohexanone 1 into
the corresponding silyloxycyclopropane 2, which is then
treated with FeCl3 in DMF to effect C–C bond cleavage to
give 3, Scheme 1.
salts with the aim of effecting two sequential C–C bond
cleavages to give a product 6 which has undergone overall
two atom ring expansion.8
This approach appeared particularly attractive, bearing in
mind the availability of six-membered starting materials
and the prospect of their conversion into eight-membered
carbocycles or heterocycles.9 In fact, to date we have ex-
plored only bis-cyclopropane substrates - 5 (X = CH2).10
Scheme 1
Preparation of two suitable bis-cyclopropanes fused to
six-membered systems was achieved as indicated in
Scheme 3.11
The key reaction is thought to involve single electron
transfer from the cyclopropane to FeCl3, thus generating a
ring-expanded carbon-centred radical that reacts with fur-
ther FeCl3 to install the chlorine atom in the product. The
initially formed b-chloroketone normally undergoes fac-
ile elimination of HCl to give the corresponding a,b-un-
saturated ketone.
Recently this reaction has been developed and extended
by Booker-Milburn,2 who has demonstrated the utility of
alternative iron salts in the reaction (see later),3 applied
the reaction to alternative types of substrate,4 and has
shown that the carbon-centred radical resulting from the
C–C bond cleavage can be cyclised onto pendant unsatur-
ated side-chains.5
A crucial aspect of the reaction is the observed preference
for cleavage of the endocyclic cyclopropane C–C bond,
which contrasts with the stereoelectronically controlled
exocyclic bond cleavage observed in many related exam-
ples involving cyclopropylcarbinyl radicals.6 We were in-
trigued by the prospect of extending this reaction to bis-
cyclopropanes, as outlined in Scheme 2, and describe
herein our preliminary results in this area.
Scheme 3
In our hands, access to 8 free from minor regioisomeric
impurities was best achieved by silylation of the b,g-un-
saturated cyclohexanone, which is available by controlled
hydrolysis of the vinyl ether 7.12 Double cyclopropanation
to give 9 as a single diastereoisomer was accomplished by
Synlett 1999, No. 2, 237–239 ISSN 0936-5214 © Thieme Stuttgart · New York

238
A. J. Highton et al.
LETTER
the Furukawa procedure using the modified conditions ably the result of very rapid radical trapping, whereas
described by Denmark.13 Worries concerning the volatili- when Fe(NO3)3 is employed alone the intermediate ho-
ty of intermediates and products prompted us to prepare a moallylic radical is quenched (by hydrogen atom abstrac-
more highly substituted substrate 12, starting with enone tion from the solvent) more slowly, allowing time for C=C
10. In this case, adequate purification of the final bis-cy- isomerisation and transannular ring closure onto the ke-
clopropane could only be achieved following desilylation tone function.
and, since such cyclopropanols are known to participate in
The results in Scheme 4 also allow us to make other, albeit
preliminary observations. Firstly, in line with the observa-
tions of Booker-Milburn, the Fe(NO3)3 systems seem to
Saegusa-type reactions,14 we used 12 without recourse to
re-silylation.
Reaction of either 9 or 12 under typical Saegusa condi- give improved results compared to the original Saegusa
tions, or under the modified conditions described by protocol. Secondly, the results for 15 are significantly im-
Booker-Milburn, gave the products 13 and 14 respective- proved compared to the smaller ring sized systems 9 and
ly, as shown below.15
12. This may be due to the more flexible nature of the larg-
er ring allowing for better bond alignment for endocyclic
bond cleavage in intermediate radicals, which results in
more efficient channelling of intermediates towards a sin-
gle product.
In conclusion, we have demonstrated for the first time the
viability of a two carbon ring expansion protocol which
utilises iron-mediated bond cleavage reactions of bis-cy-
clopropanes. The results provide further evidence of the
utility of the Booker-Milburn variants of the Saegusa re-
action as well as highlighting, in one case, a remarkable
product dependency on the nature of the iron salt em-
ployed.
Pleasingly, the anticipated eight-membered b,g-unsaturat-
ed ketones were formed as we had planned, although our
success in achieving the double ring cleavage process to
access these substituted cyclooctenones was tempered by
the moderate yields obtained. In the case of 13 (X = H or
Cl) this was in part due to volatility, but separation of
products from uncharacterised byproducts was also a
problem.
Acknowledgement
We are grateful to the University of Nottingham, and to Rhône-Pou-
lenc Rorer Limited, Dagenham, UK, for support of A.J.H.
References and Notes
We went on to consider a larger ring system 15, which
was readily prepared from cyclooctadiene, Scheme 4.16
(1) (a) Ito, Y.; Fujii, S.; Saegusa, T. J. Org. Chem. 1976, 41,2073.
(b) Ito, Y.; Fujii, S.; Nakatsuka, M.; Kawamoto, F.; Saegusa,
T. Org. Synth. Coll. Vol. 6, 1988, 327.
(2) For the most recent contributions from this group, see
(a) Booker-Milburn, K. I.; Dainty, R. F. Tetrahedron Lett.
1998, 39,5097. (b) Booker-Milburn, K. I.; Barker, A.; Brails-
ford, W. Tetrahedron Lett. 1998, 39, 4373.
(3) Booker-Milburn, K. I.; Thompson, D. F. Synlett 1993, 592.
(4) Booker-Milburn, K. I.; Cox, B.; Mansley, T. E. Chem. Com-
mun. 1996, 2577.
(5) Booker-Milburn, K. I. Synlett 1992, 809.
(6) (a) For a pertinent review, see Dowd, P.; Zhang, W. Chem.
Rev.1993, 93, 2091. (b) For a recent contribution, see Lee, P.
H.;Lee, J. Tetrahedron Lett. 1998, 39, 7889.
(7) For a recent article describing seminal contributions, see
Danishefsky, S. J. Tetrahedron 1997, 53, 8689.
(8) In the case of radical opening of epoxides it has been demon-
strated that either C–O or C–C bond clevage is possible,
depending upon the substitution pattern, see Murphy, J. A.;
Patterson, C. W.Tetrahedron Lett. 1993, 34, 867.
(9) (a) Eight-membered carbocycles, see Petasis, N. A.; Patane,
M.A. Tetrahedron 1992, 48, 5757. (b) Saturated ethers, inclu-
ding eight-membered derivatives, see Elliott, M. C. Contemp.
Org.Synth. 1997, 4, 238. (c) The eight-membered azacyclic
system is of interest mainly in the context of the manzamine
family of natural products, see Magnier, E.; Langlois, Y. Te-
trahedron 1998, 54, 6201.
Scheme 4
This system gave a remarkable outcome in that changing
the nature of the iron salt resulted in a dramatic change in
the type of product observed. Using Fe(NO3)3 alone we
obtained the decalin 16 as a mixture of diastereomers (ca.
2:1 ratio),17 whereas with FeCl3 or combinations of
Fe(NO3)3 with NCS or diphenyldisulfide we obtained the
ten-membered ring ketones 17 or 18 (both as exclusively
the E-isomer).18 In the latter cases the product is presum-
(10) For previous examples of bis-cyclopropanes from silyloxydi-
enes,see Girard, C.; Conia, J-M. J. Chem. Res. 1978, 182.
Synlett 1999, No. 2, 237–239 ISSN 0936-5214 © Thieme Stuttgart · New York

LETTER
Double Ring-Opening Reactions of Bis-Cyclopropanes Promoted by Iron Salts
239
(11) The bis-cyclopropane 9 was isolated as a single diastereo-
mer,which we assume has the two three-membered rings in an
anti arrangement. Compound 12 was isolated as a mixture of
isomers.
(12) Fleming, I.; Goldhill, J.; Paterson, I. Tetrahedron Lett.
1979,3205.
(13) Denmark, S. E.; Edwards, J. P. J. Org. Chem. 1991, 56,6974.
(14) (a) Sun U, J.; Lee, J.; Cha, J. K. Tetrahedron Lett. 1997,
38,5233. (b) Iwasawa, N.; Funahashi, M.; Hayakawa, S.;
Narasaka,K. Chem. Lett. 1993, 545.
(15) Representative experimental procedure - 14 (X = SPh) from
12.
phy (5% Et2O in petroleum ether) to give 14 as a white solid
(193 mg, 57%), m.p. 133–134°C (from cyclohexane) nmax
/
cm-1 (CHCl3) 1689; dH (400 MHz;CDCl3) 1.19 (3H, s, CH3),
1.22 (3H, s, CH3) 1.95 (1H, dd, J 4 and 12, 3-H), 2.39 (1H, dd,
J 12 and 14, 3-H), 2.73 (2H, app.d, J 5, 5-H), 3.16 (1H, dd, J
5 and 16, 8-H), 3.60 (1H, dddd,J 4, 4, 4 and 14, 4-H), 3.70 (1H,
dd, J 5 and 16, 8-H), 5.80 (1H, dd, J 5 and 5, 7-H), 6.95–6.99
(2H, m, Ar-H) and 7.13-7.28 (8H, m, Ar-H); dC (100 MHz;
CDCl3) 22.8 (CH3), 29.0 (CH3), 31.9 (5-CH2), 42.3 (8-CH2),
42.4 (3-CH2), 46.3 (4-CH), 48.1 (2-C), 125.1 (7-CH), 126.7
(CH), 126.8 (CH), 126.9 (CH), 128.1 (CH), 128.7 (CH), 132.0
(CH), 134.7 (C), 137.5 (C), 144.8 (C) and 216.3 (1-C); m/z
(Found: M+Na+359.1423.C22H23OS requires M+Na,
359.1446).
A solution of dried ferric nitrate was prepared by stirring a so-
lution of Fe(NO3)3.9H2O (0.89g, 2.2 mmol) in dry DMF (10
ml) over 4Å molecular sieves under an atmosphere of argon,
for 16h. This solution was then added via syringe pump, over
a period of 30min, to a solution of 12 (0.23g, 1.0 mmol) and
PhSSPh (0.65g,3.0 mmol) in dry de-gassed DMF (5 ml). The
mixture was stirred until TLC showed complete consumption
of starting material (1.5h). The reaction mixture was then pou-
red into water (100 ml) and the product extracted into EtOAc
(3 x 50 ml). The organic extract was dried (Na2SO4) and eva-
porated under reduced pressure to give the crude product as a
yellow solid. This material was purified by flash chromatogra-
(16) Compound 15 was prepared from cyclooctadiene by the follo-
wing sequence of reactions: (i) MeCO3H, Na2CO3; (ii) LiAlH4
(57%two steps); (iii) TPAP, NMO (55%); (iv) LDA, Me3SiCl
(76%);(v) Et2Zn, ClCH2I (77%).
(17) Both diastereomers show O–H str. in the IR but no C=O
absorption and no carbonyl in the 13C NMR. For the major,
lesspolar, isomer (Found: M+,152.1201. C10H16O requires
M,152.1202).
(18) The structure of 17 was confirmed by X-ray crystallography;
full details will be reported elsewhere.
Synlett 1999, No. 2, 237–239 ISSN 0936-5214 © Thieme Stuttgart · New York