Divalent activation in temporary phosphate tethers: Highly selective cuprate displacement reactions

Article abstract of DOI:10.1021/ol061756r

(Chemical Equation Presented) Desymmetrization of a readily derived psuedo-C₂-symmetric monocyclic phosphate via highly diastereoselective anti-S_N2′ allylic displacement reactions is reported. This method utilizes of a wide variety o

Full text of DOI:10.1021/ol061756r

ORGANIC
LETTERS
2006
Vol. 8, No. 22
5025-5028
Divalent Activation in Temporary
Phosphate Tethers: Highly Selective
Cuprate Displacement Reactions
Alan Whitehead, James P. McParland, and Paul R. Hanson*
Department of Chemistry, UniVersity of Kansas, 1251 Wescoe Hall DriVe,
Lawrence, Kansas 66045-7582
phanson@ku.edu
Received July 17, 2006
ABSTRACT
Desymmetrization of a readily derived psuedo-C2-symmetric monocyclic phosphate via highly diastereoselective anti-SN2′ allylic displacement
reactions is reported. This method utilizes of a wide variety of zinc-derived organocuprates to afford E-1,2-syn-configured phosphate acid
building blocks. Extension of this protocol to unsymmetric monocyclic phosphates exclusively yields 1,2-anti-configured products. Within this
study, stereoelectronic factors, coupled with allylic strain, ultimately govern regio- and diastereoselective cuprate reactions, further substantiating
the Corey mechanism for organocuprate additions into allylic esters.
We have recently reported use of tripodal-phosphate tethers
for coupling of both simple and complex allylic alcohols
using ring-closing metathesis (RCM).1 This method high-
lights the ability of a phosphate tether to bestow multivalent
activation throughout phosphate ester appendages, thereby
providing latent leaving group ability yet possessing or-
thogonal stability. We report herein a new strategy employing
phosphate tethers in which a phosphate ester serves a dual
role as both tether for coupling two allylic alcohols via ring-
closing metathesis (RCM) and as a subsequent leaving group
in selective anti-SN2′ displacement reactions with organo-
cuprate nucleophiles.
An underlying principle of this approach utilizes the well-
documented superiority of phosphates as leaving groups in
copper-mediated anti-SN2′ displacement reactions, initially
highlighted by Yamamoto and co-workers.2 This feature,
where the stereochemistry of an allylic phosphate is inverted
in an anti-SN2′ displacement, has been successfully employed
in previous strategies that exploit high stereospecificity for
chirality transfer.3 Further advances in this area have recently
uncovered enantioselective anti-SN2′ displacements with
chiral Schiff base cuprates as a viable means of desymme-
trizing meso-1,3-syn allylic phosphates,4 as well as reagent-
controlled asymmetric allylic phosphate displacements cata-
lyzed by copper.5 The method we report herein initially
focuses on the use of symmetry-breaking cuprate additions
to readily prepared pseudo-C2-symmetric monocyclic phos-
phate 1. To the best of our knowledge, this system is the
(2) Initial report of the superiority of phosphate ester leaving groups in
anti-SN2′ displacements: (a) Yanagisawa, A.; Noritake, Y.; Nomura, N.;
Yamamoto, H. Synlett 1991, 4, 251-253. Proposed mechanism for allylic
phosphate displacement via organocuprates: (b) Corey, E. J.; Boaz, N. W.
Tetrahedron Lett. 1984, 25, 3063-3066.
(3) For chirality transfer in anti-SN2′ allylic phosphate displacements,
see: (a) Belelie, J. L.; Chong, M. J. Org. Chem. 2001, 66, 5552-5555. (b)
Calaza, M. I.; Hupe, E.; Knochel, P. Org. Lett. 2003, 5, 1059-1061 and
references therein.
(4) (a) Piarulli, U.; Daubos, P.; Claverie, C.; Roux, M.; Gennari, C.
Angew. Chem., Int. Ed. 2003, 42, 234-236. (b) Piarulli U.; Claverie, C.;
Daubos, Gennari, C.; Minnaard, A. J.; Feringa, B. L. Org. Lett. 2003, 5,
4493-4496.
(1) Whitehead, A.; McReynolds, M. D.; Moore, J. D.; Hanson, P. R.
Org. Lett. 2005, 7, 3375-3378.
10.1021/ol061756r CCC: $33.50
© 2006 American Chemical Society
Published on Web 09/29/2006

first of its kind where a central phosphate tether imparts
divalent activation to the unique doubly allylic phosphate
subunit. In the report that follows, intriguing conformational
effects have led us to extend the study into unsymmetric
phosphates, ultimately providing experimental insight into
the Corey mechanism of cuprate displacements.
Scheme 3. Possible Modes for Cuprate Addition
Our study began with the generation of (S,S)-monocyclic
phosphate 1, which was achieved in three steps in good
overall yields (Scheme 1). Allylic alcohol (S)-3 is readily
Scheme 1. Synthesis of Monocyclic Phosphate 1
moiety in two requisite anti-periplanar relationships outlined
in Scheme 3. In both conformers we initially anticipated that
energies related to the orthogonally aligned phosphate leaving
group would be roughly equal in energy, since both are
secondary allylic phosphates. When considering only elec-
tronic factors, this would imply transition states of equal
energy as described by the original Corey mechanism2b which
invokes a bidentate coordination of incoming cuprate to both
π* and σ* orbitals, vide infra. Attack of cuprate nucleophiles
on 1 can potentially occur through any of four diastereotopic
olefinic orbitals. Pseudosymmetry, however, presents two
(not four) stereodivergent pathways for anti-SN2′-allylic
displacement, ultimately leading to only two possible dia-
stereomeric products.9 Pathway A occurs through conformer
A with addition anti to the vicinal CH2OBn group generating
anti-(Z) diastereomer 7, while pathway B occurs via con-
former B with addition syn to the vicinal CH2OBn group
leading to the syn-(E) diastereomer 8.
generated on multigram scale using the Mioskowski-Christie
protocol6 with commercially available glycidol ether (S)-2.
Upon condensation of the alkoxide of 3 with (MeO)POCl2,
the phosphate triester tether mediates coupling of two olefins
via RCM in good yield using (IMesH2)(PCy3)(Cl)2RudCHPh
(cat.-B);7 use of catalyst (PCy3)2(Cl)2RudCHPh (cat.-A) gave
poor yields of 1. Optimized RCM conditions required
elevated temperatures (90 °C in toluene) with continuous
argon purging.8
We initially probed several conditions as part of our
overarching investigation of orthogonal reactivity patterns
within cyclic phosphates containing multiple allylic phos-
phate positions (Scheme 2). We then shifted focus toward
Scheme 2. Preparation of 1,4-Diols via Phosphate Cleavage
Pleasingly, anti-SN2′ displacement of allylic phosphate 1
using Et2Zn/CuCN‚2LiCl, after acid workup, led to formation
of phosphate acid 8b as a single diastereomer (>20:1) as
witnessed by 31P NMR (Scheme 4). Subsequent treatment
Scheme 4. Cuprate Addition/Phosphate Acid Cleavage
the anti-SN2′-allylic displacement studies using soft orga-
nocuprate nucleophiles. A prerequisite for the reaction
requires the leaving group to be orthogonal to the system,
i.e., coplanar alignment of the σ* and π* orbitals. Monocyclic
phosphate 1 possesses the ability to align the phosphate
with Red-Al led to good yields of homoallylic alcohol 9b.
Initial tentative assignment of the relative stereochemistry
within phosphate acid 8b was made based on the proposed
model for cuprate addition and the observed olefinic coupling
(5) (a) Kacprznski, M. A.; Hoveyda, A. H. J. Am. Chem. Soc. 2004 126,
10676-10681. (b) Larsen, A. O.; Leu, W.; Oberhuber, C. N.; Campbell, J.
E.; Hoveyda, A. H. J. Am. Chem. Soc. 2004, 126, 11130-11131.
(6) Davoille, R. J.; Rutherford, D. T.; Christie, S. D. R. Tetrahedron
Lett. 2000, 41, 1255-1259.
(7) Scholl, M.; Ding, S.; Lee, C. W.; Grubbs, R. H. Org. Lett. 1999, 1,
953-956.
(8) Nosse, B.; Schall, A.; Jeong, W. B.; Reiser, O. AdV. Synth. Catal.
2005, 347, 1869-1874.
(9) Cuprate addition leads to loss of the prochiral phosphorus center,
thereby providing only two potential products via four diastereomeric
pathways. These pathways correspond to addition syn- or anti- to the PdO
or OMe groups attached to the prochiral phosphorus atom. The other two
possible diastereomers, anti-(E)-7 and syn-(Z)-8 cannot be accessed from
(S,S)-1 via an anti-SN2′-allylic displacement mechanism. Use of meso-1
would be required to access these two diastereomers.
5026
Org. Lett., Vol. 8, No. 22, 2006

constant consistent with an (E)-configured olefin (J ) 15.7
Hz). Unambiguous determination of relative stereochemical
relationship at C(2,3) within 9 was ultimately achieved using
Karplus analysis of vicinal protons contained within the six-
membered ring derived from reductive ozonolysis/acetonide
formation of 9.10
Table 1. Cuprate Addition/Phosphate Acid Cleavage Sequence
The remarkable selectivity for this transformation can be
rationalized using Corey’s proposed concerted, asynchronous
mechanism2b for cuprate additions as highlighted in Figure
1. In this mechanism, Corey surmises that the reacting
Figure 1. Corey model for rationalizing stereoselectivity.
cuprate simultaneously coordinates both the π* orbital of
the olefin and σ* orbital of the phosphate ester leaving group.
The asynchronous nature of the transformation predicts a
transition state in which substantial bond-lengthening occurs
with respect to the σ* bonding orbital. When applied to
conformers A and B in 1, it is anticipated the σ* orbitals
are roughly equal in energy, since both are secondary allylic
phosphates. Therefore, the stereoselectivity displayed in the
cuprate reaction with 1 must ultimately be dictated by the
lower steric requirements of conformer B to attain the
requisite coplanar alignment of π* and σ* orbitals with the
cuprate d-orbital; the greater allylic 1,3-strain (A1,3 strain)
within conformer A due to the CH2OBn-side chain disfavors
cuprate addition via this conformer (Scheme 3).11
the phosphate acids were directly subjected to Red-Al to
afford corresponding homoallylic substrates 9 in good to
excellent yields.
Additional reactions of 1 with an array of zinc-based
organocuprates gave similar high yields and selectivities
when using 4-5 equiv of dialkylzinc-derived organocuprates
(entries 1-3, Table 1). Treatment of 1 with 8-9 equiv of
the mixed zinc-copper reagents using alkyl halide-derived,
functionalized organozinc reagents (entries 4-8, Table 1)
provided similar yields and diastereoselectivities. The speci-
fied excess reagents were used in all cases to ensure complete
consumption of starting materials and overall provided
cleaner post-workup mixtures.12 Following acidic workup,
While cleavage of the phosphate acid using Red-Al
afforded the majority of products 9 in good yields, lower
yields were observed for reductions in which substrates 8
contained functionally sensitive R1 groups. To circumvent
this problem, in situ methylation of the crude phosphate acid
using TMSCHN2 produced 10 prior to Red-Al reduction and
led to overall higher yields of 9 (Scheme 5).
We next turned our attention to the construction of
unsymmetric monocyclic phosphates containing similar steric
environments with regard to the incoming cuprate, but
disparate electronic energies at the σ* orbitals of the leaving
phosphate ester (i.e., primary versus secondary allylic
phosphate ester leaving groups). In contrast to phosphate 1,
an energetic bias for the corresponding allylic phosphates
(10) (a) See the Supporting Information for acetonide characterization
and analysis. (b) Analysis of JAC, JBC, JCD coupling constants provided
small coupling values (1.4-2.6 Hz) consistent with the assigned structure.
These values were identical to previously reported spectral data when R1
) Me; see: Ghosh, A. K.; Kim, J.-H. Org. Lett. 2003, 5, 1063-1066.
(11) (a) Hoffmann, R. W. Chem. ReV. 1989, 89, 1841-1860. (b) A1,3
strain in acyclic allylic phosphates, see: Harrington-Frost, N.; Leuser, H.;
Calaza, M. I.; Kneisel, F. F.; Knochel, P. Org. Lett. 2003, 5, 2111-2114.
(12) Acid quench and extraction provided clean 1H and 31P NMR spectra
of crude phosphate acids without column chromatography.
Org. Lett., Vol. 8, No. 22, 2006
5027

sequence. Treatment of 14 with 3.0 equiv of diethylzinc-
derived organocuprate at -40 °C, and subsequent phosphate
cleavage, led to formation of a single product 15. Using the
aforementioned Karplus analysis of the resulting acetonide
of 15, unambiguous assignment of anti-relationship was
established.
Scheme 5. Methylation/Phosphate Cleavage
The regioselectivity of cuprate displacement in 14 is most
surprising in that the requisite conformer F leading to product
15 (Scheme 7) contains an unfavorable A1,3 interaction
generated by the gem-dimethyl moiety. However, the ob-
served selectivity is consistent with the results obtained with
12, in which the cuprate displaces the more substituted
phosphate leaving group. Since the Corey model for cuprate
addition predicts significant bond breakage through σ*, the
lower energy σ* (i.e., the most substituted carbon) in the
asynchronous concerted transition state dictates the reaction
pathway over substantial A1,3 steric interactions.
In rationalizing and summarizing regio- and diastereo-
selective cuprate additions with symmetric 1 and unsym-
metric phosphates 12 and 15, both steric and stereoelectronic
effects appear to play major roles. In the case of 12 and 15,
cuprate displacements proceeded via the more substituted
σ* positions, secondary and tertiary, respectively. When σ*
energies are equivalent as in 1, allylic strain appears to govern
the cuprate addition.
was thus anticipated. Dioxaphosphacycle 11 was generated
via the coupling of one equivalent of allylic alcohol 3 with
allyldiphenyl phosphate, followed by RCM to afford 11 in
good yields over two steps (Scheme 6). Treatment of 11 with
Scheme 6. Unsymmetric Monocyclic Phosphates
In conclusion, monocyclic phosphates undergo a highly
selective anti-SN2′ allylic phosphate displacement. Subse-
quent cleavage affords an array of syn-(E)-homoallylic
alcohols when pseudo-C2-symmetric monocyclic phosphates
are employed. In extending this method to unsymmetric
phosphates, cuprate addition proceeded through the more
substituted allylic phosphate position (lower σ* energy) with
the general reactivity pattern being 1 < 2 < 3, allowing
access to anti-configured homoallylic alcohols. In probing
these systems, the Corey mechanism for allylic organocuprate
displacements was examined and further substantiated.
3.0 equiv of diethylzinc-derived organocuprates, and sub-
sequent phosphate cleavage, led to formation of a single
product 12 produced via cuprate displacement at the second-
ary allylic phosphate position. In contrast to 1, no dominant
conformational preferences exist between the two described
reactive conformations of 11, indicating that regioselectivity
is dictated by electronic effects, vide infra.
This intriguing result led us to explore yet a third
paradigm, the unsymmetric cyclic phosphate 14, containing
both secondary and tertiary allylic phosphate positions
(Scheme 7). Phosphate 14 was assembled via differential
coupling of secondary allylic alcohol 3 and 2-methylbut-3-
en-2-ol with dichloromethyl phosphate. The resulting labile
acyclic phosphate was immediately subjected to metathesis
affording phosphate 14 in moderate yields over the two-step
Acknowledgment. This investigation was generously
supported by funds provided by the Petroleum Research Fund
(PRF 42457-AC, administered by the American Chemical
Society), the National Science Foundation (NSF CHE-
0503875), and the NIH Dynamic Aspects in Chemical
Biology Training Grant (A.W.). We thank Dr. David Vander
Velde and Sarah Neuenswander for assistance with NMR
measurements and Dr. Todd Williams for HRMS analysis.
We kindly acknowledge Daiso Co., Ltd., Fine Chemical
Department for donating 100 g of each antipode of both
benzyl- and trityl-protected glycidols (e-mail: akkimura@
daiso.co.jp) and Materia, Inc., for supplying metathesis
catalyst and helpful suggestions. In addition, we thank the
reviewers for helpful suggestions.
Scheme 7. Unsymmetric Monocyclic Phosphates
Note Added after ASAP Publication. The PRF number in
the Acknowledgment was incorrect in the version published
September 29, 2006; the corrected version was published
October 3, 2006.
Supporting Information Available: Experimental details
and spectroscopic data of new compounds 1 and 4-15 are
reported. This material is available free of charge via the
Internet at http://pubs.acs.org.
OL061756R
5028
Org. Lett., Vol. 8, No. 22, 2006