First isolation and characterization of an anti-apicophilic spirophosphorane bearing an oxaphosphetane ring: A model for the possible reactive intermediate in the Wittig reaction

Article abstract of DOI:10.1021/ja0170145

An anti-apicophilic phosphorane bearing an oxaphosphetane ring, in which the ring carbon is apical and the ring oxygen is equatorial (C-apical), has been prepared as a thermally less stable stereoisomer of a phosphorane with an ordinary equatorial carbon-

Full text of DOI:10.1021/ja0170145

Published on Web 06/11/2002
First Isolation and Characterization of an Anti-Apicophilic Spirophosphorane
Bearing an Oxaphosphetane Ring: A Model for the Possible Reactive
Intermediate in the Wittig Reaction
Satoshi Kojima, Masaya Sugino, Shiro Matsukawa, Masaaki Nakamoto, and Kin-ya Akiba*
Department of Chemistry, Graduate School of Science, Hiroshima UniVersity,
1-3-1 Kagamiyama, Higashi-Hiroshima 739-8526, Japan
Received September 6, 2001; Revised Manuscript Received March 12, 2002
Scheme 1
The Wittig reaction is a well-established method for introducing
double bonds in organic compounds.1 In this reaction, two isomeric
oxaphosphetane intermediates, int-a and int-b, are conceivable, and
on the basis of the general belief that P-C bond dissociation is
more advanced than that of P-O in the transition state, int-b bearing
an apical carbon in the oxaphosphetane ring has been assumed by
many to be the intermediate that gives the olefinic product (Scheme
1).1,2 Skepticism about this hypothesis has arisen from the fact that
species such as int-b could not be observed experimentally.
Furthermore, a theoretical calculation on model phosphoranes (H
Chart 1
in the place of Ph) showed that species such as int-b may only be
semistable, thus adding to the controversy.3
In the pursuit of int-b type compounds, the greatest advance
made thus far is the preparation of phosphorane 1a (Chart 1),4 which
can be considered a model for an intermediate in the imino-Wittig
reaction.5 Here, 1a has been geared to be thermodynamically as
stable as isomer 1b. For the oxygen series, even models for int-
a-type isomers are scarce.1f,g Among these, especially noteworthy
are the phosphoranes 2, which were the first to be fully structurally
characterized and which provided mechanistic insights into the
olefin-forming step.6 We have recently synthesized and isolated
3a, the first anti-apicophilic spirophosphorane, via a thermal
reaction7 and more recently by oxidation.8 By applying the latter
Scheme 2
methodology, we have succeeded in obtaining the anti-apicophilic
analogue of spirophosphorane 2, i.e., 5a, without altering the
intrinsic apicophilic property of the substituents about the phos-
phorus atom. Although the substitution pattern on the phosphorus
atom slightly differs from that of phosphoranes in the typical Wittig
reaction (int-a,b), 5a can be considered to be the closest model
for int-b reported to date.9 Herein, we report on the preparation
Scheme 3
and characterization of this unique compound.
Compound 4, the precursor to 5a, was obtained as shown in
Scheme 2. The use of the t-Bu group was based on our previous
finding that it was extremely effective in stabilizing anti-apicophilic
phosphoranes.7 t-BuPCl2 was first converted to sulfide 6, which in
a one-pot reaction was transformed into apical-H phosphorane 4
(47% yield from 6).10 Phosphorane 4 could not be obtained upon
3).10 Since the ratio could not be improved upon by changing
the use of the P oxide analogue of 6. All attempts to isolate
reaction conditions, this probably reflects the kinetic selectivity in
the ring-closing process. The desired anti-apicophilic 5a fortunately
crystallized out of the reaction mixture upon addition of hexane.
Unlike 3a, which could bear aqueous workup, 5a was found to be
labile to proton sources, resulting in facile stereomutation to 5b.
For instance, dissolution of 5a in anhydrous CDCl3 resulted in
complete conversion to 5b within minutes at room temperature,
presumably due to residual acid in the solvent. On the other hand,
in the presence of DBU (small excess) in diglyme, the half-life of
5a was ca. 50 min at 70 °C. Since there was a propensity for the
intermediates led to either decomposition or other undesirable
reactions.
The cyclization reaction of 4 proved to be less facile than in the
case of 3a. Cyclization could be effected at the temperature of
150 °C (in DMSO), only to furnish 5b and decomposition products
of 5b. However, the oxidation procedure using n-BuLi followed
by iodine in ether (0 °C) gave 5a as a 1:1 mixture with 5b (Scheme
* Address correspondence to this author. Present address: Advanced Research
Center for Science and Engineering, Waseda University, 3-4-1 Ohkubo,
Shinjuku-ku, Tokyo 169-8555, Japan.
9
7674
J. AM. CHEM. SOC. 2002, 124, 7674-7675
10.1021/ja0170145 CCC: $22.00 © 2002 American Chemical Society

C O M M U N I C A T I O N S
Thus, for compound 5a it is apparent that the bond strength of the
apical P-C bond of the oxaphosphetane ring makes the bond
cleavage process (Wittig reaction) much higher in energy than
stereomutation.
In summary, the preparation and characterization of 5a have been
accomplished. This success implies that the previously hypothesized
Wittig reaction intermediate with an apical carbon may actually
exist as a thermodynamically stable species. A question unanswered
here, whether the less stable isomer is actually the direct precursor
to the olefin product, may be clarified by modification of the
substituents upon the oxaphosphetane ring in 5a.12
Figure 1. ORTEP drawing of 5a and 5b showing the thermal ellipsoids at
the 30% probability level.
Acknowledgment. The authors are grateful to Central Glass
Co. Ltd. for a generous gift of hexafluorocumyl alcohol. Partial
support of this work through a Grant-in-Aid for Scientific Research
(No. 11304044) provided by the Ministry of Education, Science,
Culture, and Sports of the Japanese Government is heartily
acknowledged.
stereomutation of 5a to accelerate in the presence of acids and
deteriorate in the presence of DBU, it is very likely that the
isomerization pathway involves P-O bond dissociation-recom-
bination. A 31P NMR measurement (detection limit assumed at 1%)
of 5b at 120 °C (p-xylene) did not show the presence of 5a in
equilibrium, implying that the lower limit in energy difference is
∆G393 K ) 3.6 kcal mol-1. Thus, it is evident that 5a is thermo-
dynamically much less favorable than 5b.
Supporting Information Available: Experimental procedures for
the syntheses of 4, 5a, 5b, 6, 4′, 5b′, and 6′ and their NMR data, and
crystallographic data for 4, 5a, 5b, 4′, and 5b′, where primes stand for
2,4,6-mesityl compounds (PDF). This material is available free of charge
via the Internet at http://pubs.acs.org.
The NMR spectra of 5a (acetone-d6) were found to be very
characteristic of its supposed trigonal bipyramidal structure in
solution and were similar to those reported for 1a.10 The coupling
constants between the phosphorus and the apical bound carbon
nuclei (1JPC ) 29.4 Hz), and between the phosphorus nucleus and
References
(1) For leading reviews, see: (a) Wadsworth, W. S., Jr. Org. React. 1977,
25, 73-253. (b) Cadogan, J. I. G. Organophosphorus Reagents in Organic
Synthesis; Academic Press: New York, 1979. (c) Maryanoff, B. E.; Reitz,
A. B. Chem. ReV. 1989, 89, 863-927. (d) Kelly, S. E. In ComprehensiVe
Organic Synthesis; Trost, B. M., Fleming, I., Eds.; Pergamon Press:
Oxford, 1991; Vol. 1, pp 730-817. (e) Johnson, A. W. Ylides and Imines
of Phosphorus; Wiley-Interscience: New York, 1993. (f) Vedejs, E.;
Peterson, M. J. Top. Stereochem. 1994, 21, 1-157. (g) Vedejs, E.;
Peterson, M. J. AdV. Carbanion Chem. 1996, 2, 1-85.
the protons (2JPH ) 6.8, 11.2 Hz) on this carbon, were small
2
compared to their counterparts in 5b (1JPC ) 106.7 Hz; JPH
)
20.0, 23.0 Hz), as expected from the involvement of the weak
hypervalent bond.11 The proton ortho to P on the aryl ring (δH
)
7.76) of the Martin ligand was not shifted downfield as much as
for 5b (δH ) 8.30), which has a more polar apical bond. An NOE
of 11% was observed between the ortho proton mentioned above
and the proton on the oxaphosphetane ring facing this hydrogen
(Scheme 3). The observed lower field 31P chemical shift for 5a (δP
) -6.3) compared to that for 5b (δP ) -10.9) could be attributed
to the σ-acceptor property of the equatorial oxygen substituent,
which should be more influential than that occupying the apical
position.
Indisputable evidence for the structure of the anti-apicophilic
spirophosphorane 5a was obtained from X-ray structural analysis,
as shown in comparison with that of 5b in Figure 1.10 Both the
group of three equatorial elements and phosphorus, and the group
of four atoms in the oxaphosphetane ring, were found to form
planes, respectively for both 5a and 5b, implying that both
compounds are essentially of trigonal bipyramid nature. As
expected, the apical P-C bond was found to be much longer (5a,
1.914 Å) than the corresponding P-C bond in 5b (1.820 Å), while
the opposite trend was observed for the four-membered ring P-O
bond (5a, 1.663 Å; 5b, 1.745 Å). The apical bond was also distorted
more in 5a than in 5b.
(2) Bestmann, H. J.; Roth, K.; Wilhelm, E.; Bo¨hme, R.; Burzlaff, H. Angew.
Chem., Int. Ed. Engl. 1979, 18, 876-877.
(3) Naito, T.; Nagase, S.; Yamataka, H. J. Am. Chem. Soc. 1994, 116, 10080-
10088.
(4) (a) Kawashima, T.; Soda, T.; Kato, K.; Okazaki, R. Phosphorus, Sulfur
Silicon 1996, 109-110, 489-492. (b) Kawashima, T.; Soda, T.; Okazaki,
R. Angew. Chem., Int. Ed. Engl. 1996, 35, 1096-1098.
(5) For examples on phosphorane stabilization using the Martin ligand, see:
(a) Martin, J. C. Science 1983, 221, 509-514. (b) Moon, C. D.; Chopra,
S. K.; Martin, J. C. In Phosphorus Chemistry, DeVelopments in American
Science; Walsh, E. N., Griffiths, E. J., Parry, R. W., Quin, L. D., Eds.;
ACS Symposium Series 486; American Chemical Society: Washington,
DC, 1992; pp 128-136.
(6) (a) Kawashima, T.; Kato, K.; Okazaki, R. J. Am. Chem. Soc. 1992, 114,
4008-4010. (b) Kawashima, T.; Kato, K.; Okazaki, R. Angew. Chem.,
Int. Ed. Engl. 1993, 32, 869-870. (c) Kawashima, T.; Okazaki, R. Synlett
1996, 600-608 and references therein.
(7) (a) Kajiyama, K.; Kojima, S.; Akiba, K.-y. Tetrahedron Lett. 1996, 37,
8409-8412. (b) Kojima, S.; Kajiyama, K.; Nakamoto, M.; Akiba, K.-y.
J. Am. Chem. Soc. 1996, 118, 12866-12867.
(8) Kajiyama, K.; Yoshimune, M.; Nakamoto, M.; Matsukawa, S.; Kojima,
S.; Akiba, K.-y. Org. Lett. 2001, 3, 1873-1875.
(9) For a phosphorane bearing an oxaphosphetane ring with an apically fixed
carbon atom, see: Vollbrecht, S.; Vollbrecht, A.; Jeske, J.; Jones, P. G.;
Schmutzler, R.; du Mont, W.-W. Chem. Ber. 1997, 130, 819-822.
(10) See Supporting Information for details.
(11) Cavell, R. G. In Phosphorus-31 NMR Spectroscopy in Stereochemical
Analysis; Verkade, J. G., Quin, L. D., Eds.; VCH Publishers: Deerfield
Beach, FL, 1987; pp 231-254.
(12) Following the suggestion of a reviewer, a phosphorane (4′) bearing a bulky
aromatic 2,4,6-trimethylphenyl group in the place of the tert-butyl group
in 4 was prepared and subjected to the cyclization conditions. NMR signals
corresponding to the C-apical isomer (5a′) could not be observed in the
reaction mixture (see Supporting Information).
Heating a solid sample of 5a at the melting point temperature
(ca. 120 °C) for 5 min gave only 5b as a consequence of
pseudorotation. Only after prolonged heating of 5b at the melting
point (140 °C) could olefin formation (quantitative) be observed.
JA0170145
9
J. AM. CHEM. SOC. VOL. 124, NO. 26, 2002 7675