2-Alkylidenephosphiranes

Article abstract of DOI:10.1021/ja971340w

Three stable alkylidenephosphiranes have been synthesized from the addition of the terminal phosphinidene complex Ph-P-W(CO)₅ to allene, 1,1-dimethylallene, and tetramethylallene. Isopropylidenephosphirane 16b was characterized by a single-crystal X-ray structure determination. Demetalation of its W(CO)₅ group provides the uncomplexed compound. The addition reaction with tetramethylallene also yields vinylphosphirane epimers, which rearrange to phospholene 20. Ab initio MP2/6-31G* structures and energies are presented for the parent uncomplexed methylenephosphirane and its dimethyl derivatives.

Full text of DOI:10.1021/ja971340w

8432
J. Am. Chem. Soc. 1997, 119, 8432-8437
2-Alkylidenephosphiranes
Steffen Krill, Bing Wang, Jui-Te Hung, Chris J. Horan, Gary M. Gray, and
Koop Lammertsma*
Contribution from the Department of Chemistry, UniVersity of Alabama at Birmingham,
UAB Station, Birmingham, Alabama 35294-1240, and Faculty of Chemistry, Vrije UniVersiteit,
De Boelelaan 1083, 1081 HV Amsterdam, The Netherlands
ReceiVed April 28, 1997X
Abstract: Three stable alkylidenephosphiranes have been synthesized from the addition of the terminal phosphinidene
complex Ph-P-W(CO)5 to allene, 1,1-dimethylallene, and tetramethylallene. Isopropylidenephosphirane 16b was
characterized by a single-crystal X-ray structure determination. Demetalation of its W(CO)5 group provides the
uncomplexed compound. The addition reaction with tetramethylallene also yields vinylphosphirane epimers, which
rearrange to phospholene 20. Ab initio MP2/6-31G* structures and energies are presented for the parent uncomplexed
methylenephosphirane and its dimethyl derivatives.
Ring strain augmented by the presence of an exocyclic double
Only recently did Yoshifuji et al.9 report the first synthesis
of a congested methylenephosphirane, 5, via the addition of
dichlorocarbene to a sterically protected 1-phosphaallene (eq
1). No structural details were provided. Subsequently, Manz
bond makes the heteroatom analogues of methylenecyclopro-
panes (1) fascinating compounds.1,2 Strain underlies, for
example, the biradical interconversion of the valence isomers
2 and 3, which consequently have received particular attention.3-5
While synthesizing such systems is inherently challenging, their
presence as reactive and versatile intermediates has been well
established. Expectantly, the stability of the methylene deriva-
tives of the aziridines, oxiranes, and thiiranes increases with
bulky substituents on the heteroatom, ring, and/or double bond.
These sterically congested systems are typically generated via
ring closure reactions, either thermally or photochemically
induced, rather then by, e.g., the epoxidation of or the nitrene
addition to allenes.1,6 Our interest in these heterocycles and
their even more strained higher homologues, the radialenes 4,7
arose because of the noticeable elusive phosphorus analogues.8
and Maas10 reported two similarly congested methylene-
phosphiranes 6a,b. These stable compounds were obtained
from thermal N2-extrusion of diazaphosphole precursors, thereby
extending the method explored in detail by Quast and co-
workers11 for the syntheses of the O-, S-, and N-analogues.
Product formation is presumed to occur via ring closure of the
reactive 2-phosphacumulene intermediates (eq 2). The yields
* To whom correspondence should be addressed at Vrije Universiteit.
X Abstract published in AdVance ACS Abstracts, August 15, 1997.
(1) For a review, see: L’abbe´, G. Angew. Chem., Int. Ed. Engl. 1980,
19, 276.
(2) Doering, W. v. E.; Sachdev, K. J. Am. Chem. Soc. 1974, 96, 65.
Doering, W. v. E.; Roth, H. D. Tetrahedron 1970, 26, 2825. Gajewski, J.
J. J. Am. Chem. Soc. 1968, 90, 7178. Dolbier, W. R.; Alonso, J. H. J. Am.
Chem. Soc. 1973, 95, 4421. Roth, W. R.; Wegener, G. Angew. Chem., Int.
Ed. Engl. 1975, 14, 758. Dowd, P. Acc. Chem. Res. 1972, 5, 242. Berson,
J. A. Acc. Chem. Res. 1978, 11, 446.
(3) Osamura, Y.; Borden, W. T.; Morokuma, K. J. Am. Chem. Soc. 1976,
98, 7081. Wasserman, H. H.; Berdahl, D. R.; Lu, T.-J. In The Chemistry of
the Cyclopropyl Group; Rappoport, Z., Ed.; Wiley: New York, 1987; Part
2, pp 1455-1532. Turecek, F.; Drinkwater, D. E.; McLafferty, F. W. J.
Am. Chem. Soc. 1991, 113, 5950.
(4) Block, E.; Schwan, A.; Dixon, D. A. J. Am. Chem. Soc. 1992, 114,
3492. Block, E.; Penn, R. E.; Ennis, M. D.; Owens, T. A.; Yu, S. L. J. Am.
Chem. Soc. 1978, 100, 7436. Jongejan, E.; Buys, Th. S. V.; Steinberg, H.;
de Boer, Th. Recl. TraV. Chim. Pays-Bas 1978, 97, 215. Ando, W.; Itami,
A.; Furuhata. T.; Tokitoh, N. Tetrahedron Lett. 1987, 28, 1787. Ando, W.;
Choi, N.; Kabe, Y. J. Am. Chem. Soc. 1990, 112, 4574. Schhaumann, E.;
Behr, H.; Adiwidjaja, G.; Tangerman, A.; Lammerink, B. H. M.; Zwanen-
burg, B. Tetrahedron 1981, 37, 219.
are modest due to competing biradical H-abstractions. The
difference between the ab initio geometry of the parent
methylenephosphirane, C3H5P (7),12 and the single-crystal X-ray
structure of 6a emphasizes the effect of steric crowding (see
also later).10 These two synthetic routes are, however, rather
limited in scope due to the high sensitivity of phosphaallene
and 2-phosphacumulenes.8 A third route has been attempted,
(7) For sila- and thia[3]radialenes, see: Ando, W.; Hanyu, Y.; Kumamoto,
Y.; Takata, T. Tetrahedron Lett. 1986, 42, 1989. Yamamoto, T.; Kabe, Y.;
Ando, W. Organometallics 1993, 12, 1996. For a review on carbon
[n]radialenes, see: Hopf, H.; Maas, G. Angew. Chem., Int. Ed. Engl. 1992,
31, 931.
(8) (a) Regitz, M, Scherer, O. Eds. Multiple bonds and low coordination
in phosphorus chemistry; Georg Thieme Verlag: Stuttgart, 1990. (b)
Mathey, F. Chem. ReV. 1990, 90, 997.
(9) Yoshifuji, M.; Toyota, K.; Yoshimura, H.; Hirotsu, K.; Okamoto,
A. J. Chem. Soc. Chem. Commun. 1991, 124.
(5) Quast, H.; Risler, W. Angew. Chem., Int. Ed. Engl. 1973, 12, 414.
Quast, H.; Jacobi, H. Chem. Ber. 1991, 124, 1613.
(6) Bingham, E. M.; Gilbert, J. C. J. Org. Chem. 1975, 40, 224. Herges,
R.; Hoock, C. Synthesis 1991, 1151. Green, M.; Verkoczy, B.; Lown, E.
M.; Strausz, O. P. Can. J. Chem. 1985, 63, 667.
(10) Manz, B.; Maas, G. J. Chem. Soc., Chem. Commun. 1995, 25.
(11) Kinderman, M. K.; Poppek, R.; Rademacher, P.; Fuss, A.; Quast,
H. Chem. Ber. 1990, 123, 1161. Quast, H.; Fuss, A.; Jacobi, H. Chem. Ber.
1991, 124, 1747. Quast, H.; Hergenro¨ther, T. Chem. Ber. 1992, 125, 2095.
(12) Bachrach, S. M. J. Phys. Chem. 1993, 97, 4996.
S0002-7863(97)01340-1 CCC: $14.00 © 1997 American Chemical Society

2-Alkylidenephosphiranes
J. Am. Chem. Soc., Vol. 119, No. 36, 1997 8433
i.e., the simple addition of Bu3P to phenylallene, but the
intermediacy of a reactive derivative of 7, as suggested by
Gasparyan et al.,13 however, is speculative at best.
the use of this terminal phosphinidene complex for the syntheses
of several methylenephosphiranes, including its parent, will be
described.
Only two methylenediphosphiranes have been reported, both
of which are again highly congested. Yoshifuji et al.9 employed
also in this case the CCl2 carbene addition to a 1,3-diphos-
phaallene to form 8, while Baudler14 used a condensation route
instead for the synthesis of 9. Single-crystal X-ray structures
were reported for both. Likewise, there are only two reports
on the more strained phospha[3]radialenes. Yoshifuji et al.15
employed again the CCl2 carbene addition to phosphabutatriene
followed by a rearrangement to give 10, while Maercker and
Brieden16 used a condensation reaction to synthesize 11 from
dichlorophosphines and 3,4-dilithio-2,5-dimethyl-2,4-hexadiene.
Results and Discussion
Reaction of the phosphinidene precursor 12 with allene and
its 1,1-dimethyl and tetramethyl derivatives affords the new
methylenephosphiranes 16a, 16b, and 16c, respectively. These
are formed with surprising ease in reasonable yields and at a
modest temperature of ca. 55 °C. The crystalline products are
remarkably stable and can be stored for weeks in air at ambient
temperatures without showing signs of decomposition. They
We decided to explore an alternative access to the methyl-
enephosphiranes, namely, the addition of singlet, transition-
metal-complexed phosphinidenes (R-P-M) to allenes. Since
Mathey’s pioneering work on the development of such in situ
generated17 electrophilic phosphinidenes, their use toward the
syntheses of organophosphorus compounds has become a viable
option.18 While Gaspar, Cowley, and co-workers19 showed
recently by ESR spectroscopy that the parent Ph-P has a triplet
ground state, conforming to theoretical predictions,20 the transi-
tion-metal-complexed phosphinidenes R-P-W(CO)5 are be-
lieved to be singlet carbene-like species.21,22 Its cycloaddition
to olefins (eq 3) is a convenient route for the synthesis of
phosphiranes. Those of the parent 13 and the smallest phos-
phaspiro compound 14 are illustrative.23 In the present study,
1
are characterized by H, 13C, and 31P NMR spectroscopic data
and by a single-crystal X-ray structure determination for 16b.
The geometrical, electronic, and reactivity aspects of these
methylenephosphiranes will be discussed. We start with ad-
dressing the allene reactivities toward the phosphinidene
complex, followed by a structural analysis of 16b to illustrate
that stable methylenephosphiranes do not need to be heavily
substituted. We then elaborate on the demetalation of 16b and
use theoretical data for structure and energy evaluations.
Phosphinidene Addition. The carbene-like cycloaddition of
the in situ generated Ph-P-W(CO)5 to the parent allene (15a)
yields a stable methylenephosphirane under mild conditions. The
surprise is not in the ease of the addition, because Ph-P-
W(CO)5 adds with equal ease to gaseous ethylene and meth-
ylenecyclopropane, for example, to form the parent phosphirane
13 and spirophosphapentane 14, respectively.23 Rather, the
stability of the product 16a appears to stand out. The synthesis
of methylenephosphiranes via the addition of a phosphinidene
to allenes appears then as a viable alternative to the carbene
addition to phosphaallenes (eq 1) and to the cyclization of
phosphacumulenes (eq 2), the difference being that the phos-
phinidene precursors are inherently more stable than the
phosphaallenes and phosphacumulenes which typically require
bulky substituents to be synthetically accessible, thereby limiting
their synthetic scope. Because the transition metal group of
16 can be removed (see later) and its P-substituent can be varied
(e.g., alkyl, phenyl, alkoxy, amine),18 the phosphinidene-allene
addition is a versatile route.
(13) Gasparyan, G; Ovakimyan, M. Zh.; Abramyan, T. A.; Indzhikyan,
M. G. Arm. Khim. Zh. 1984, 37, 520; Chem Abstr. 1985, 102, 78993v. The
reported reaction of Bu3P with phenylallene followed by addition of HBr/
MeBr to give phosphonium salts probably occurs via betaines.
(14) Baudler, M.; Saykowski, F.; Hintze, M.; Tebbe, K.-F.; Heinlein,
T.; Vissers, A.; Fehe´r, M. Chem. Ber. 1984, 117, 1542.
(15) Toyota, K.; Yoshimura, H.; Uesugi, T.; Yoshifuji, M. Tetrahedron
Lett. 1991, 32, 6879.
(16) Maercker, A.; Brieden, W. Chem. Ber. 1991, 124, 933.
(17) (a) Marinetti, A.; Mathey F. Organometallics 1984, 3, 456. (b)
Marinetti, A.; Mathey, F.; Fischer, J.; Mitschler, J. J. Am. Chem. Soc. 1982,
104, 4484. (c) Marinetti, A.; Mathey, F. Organometallics 1982, 1, 1488.
(d) Marinetti, A.; Mathey, F.; Fischer, J.; Mitschler, J. J. Chem. Soc., Chem.
Commun. 1982, 667.
(18) (a) Mathey, F. Angew. Chem., Int. Ed. Engl. 1987, 26, 275. (b)
Mathey, F. Pure Appl. Chem. 1987, 59, 993.
(19) Li, X.; Weissman, S. I.; Lin, T.-S.; Gaspar, P. P.; Cowley, A. H.;
Smirnov, A. I. J. Am. Chem. Soc. 1994, 116, 7899.
(20) The triplet-singlet energy gap for PhP amounts to 24 kcal/mol at
CISD/6-31G*: Hamilton, T. P. Personal communication. See also: Cowley,
A. H.; Gabbai, F.; Schluter, R.; Atwood, D. J. Am. Chem. Soc. 1992, 114,
3142. For the triplet-singlet energy gap of 22.6 kcal/mol for CH3P, see:
Kim, S.-J.; Hamilton, T. P.; Schaefer, H. F. J. Phys. Chem. 1993, 97, 1872.
(21) (a) Cowley, A. H.; Barron, A. R. Acc. Chem. Res. 1988, 21, 81. (b)
Gonbeau, D.; Pfister-Guillouzo, G.; Marinetti, A.; Mathey, F. Inorg. Chem.
1985, 24, 4133.
(22) (a) Lammertsma, K.; Chand, P.; Yang, S.-W.; Hung, J.-T. Orga-
nometallics 1988, 7, 1875. (b) Hung, J.-T.; Yang, S.-W.; Chand, P.; Gray,
G. M.; Lammertsma, K. J. Am. Chem. Soc. 1994, 116, 10966. (c) Wang,
B.; Lammertsma, K. J. Am. Chem. Soc. 1994, 116, 10486.
Not surprisingly then, methylenephosphiranes can also be
generated from the 1,1-dimethyl- (15b) and tetramethylallenes
(15c). The phosphinidene complex adds exclusively to the
unsubstituted double bond of 15b to give 16b, but gives in the
case of 15c besides 16c a mixture of the anti/syn-vinylphos-
(23) Hung, J. T.; Yang, S.-W.; Gray, G. M.; Lammertsma, K. J. Org.
Chem. 1993, 58, 6786.

8434 J. Am. Chem. Soc., Vol. 119, No. 36, 1997
Krill et al.
phiranes 19. The chemical stability of the complexed methyl-
enephosphiranes is startling in comparison with methyleneox-
iranes, -aziridines, and -cyclopropanes, which are prone to
isomerization, polymerization, and decomposition reactions.1
The substituted derivatives 16b and 16c do not show any sign
of decomposition or rearrangement when heated in solution at
80 °C for two days. The absence of any formation of
phosphamethylenecyclopropane 17 contrasts the reported 2 h
3 biradical rearrangement of the analogous oxy, thio, and imino
derivatives.3,4,5 Attempts to alkylate the phosphirane ring of
16b in a manner analogous to that employed for methylene-
azirines24 proved fruitless.
Figure 1. Crystal structure of 16b (thermal ellipsoids with 50%
probability). Selected bond lengths (Å) and angles (deg): P-C1 1.85(1),
P-C2 1.776(8), C1-C2 1.47(1), C2-C3 1.31(1), C3-C4 1.47(1), C3-
C5 1.50(1), P-C6 1.799(7), W-P 2.500(2), C1-P-C2 47.8(4), W-P-
C6 124.0(3), C1-C2-C3 142.3(8), C2-C3-C4 123.0(7), C2-C3-
C5 120.4(7).
Table 1. Selected Bond Distances (Å) and CPC Angles (deg) for
the X-ray Structures of 16b, 6a, 14, and 13, the Microwave
Structure of Phopshirane C2H5P, and the Theoretical Structures of
C3H5P and C2H5P
The high intramolecular selectivity for Ph-P-W(CO)5
addition to 15b is somewhat surprising in light of the intermo-
lecular olefin competition reactions which showed isobutylene
to react slightly faster than other (poly)methylated olefins.22b
This selective formation of 16b over 18 may be thermodynami-
cally controlled, but it is also sterically the least demanding
pathway. Addition of Ph-P-W(CO)5 to 15c to give 16c
illustrates, however, that the higher substituted olefinic bond is
sterically accessible.
bond
16ba
6ab
14c
7d
13c
C2H5Pe
P-C(1)
P-C(2)
1.85(1)
1.967 1.855(7)
1.905 1.83(2) 1.867
1.818 1.80(2) 1.867
1.460 1.50(2) 1.502
1.330
1.776(8) 1.790 1.794(6)
C(1)-C(2) 1.47(1)
1.490 1.508(9)
1.331 1.470(1)
1.475(10)
C(2)-C(3) 1.31(1)
P-C(6)
1.799(7) 1.839 1.819(7)
46.4 48.6(3)
1.80(1)
48.6(7) 47.4
C(1)PC(2) 47.8(4)
51.2
The reaction with 15c is, as noted, accompanied by the
formation of the syn- and anti-phosphiranes 19. Formally, these
represent the (syn- and anti-) addition of Ph-P-W(CO)5 to
2,4-dimethyl-1,3-pentadiene, which we confirmed indepen-
dently. However, we can only speculate on their formation since
tetramethylallene does not rearrange to 2,4-dimethyl-1,3-pen-
tadiene at 70 °C in the presence of CuCl nor does isolated 16c
rearrange to 19. Interaction of the electrophilic Ph-P-W(CO)5
with the hyperconjugatively weakened C-CH3 single bond or
a biradical addition, each followed by a subsequent [1,3]-H-
shift, might explain the formation of these products.
a This study. b Reference 10; reported estimated standard deviations
are 0.003-0.004 Å for bond lengths and 0.1° for the bond angle.
c Reference 24. d This study, MP2(fc)/6-31G*. e Reference 30.
summarized in Table 1 together with those reported for the
X-ray structures of methylenephosphirane 6a,10 the related
W(CO)5-complexed spirophospha[2.2]pentane 14,23 and the
parent W(CO)5-complexed phosphirane 13.23 Comparison of
their important bond distances and bond angles illustrates that
the phosphirane ring of structure 16b differs from that of the
more congested 6a. Instead, the resemblance with the spiro
compound 14 is much better. For example, the P-C(1) bond
length of 16b of 1.85(1) Å is similar to the 1.855(7) Å bond
length of 14 but substantially shorter than the 1.967 Å bond
length reported for 6a. The much longer P-C(1) bond of the
latter must be attributed to its high degree of substitution,
reflecting the repulsion between the two phenyl groups and the
P-mesityl substituent, as noted by Manz and Maas.10 The other
phosphirane PC bond is in all three cases much shorter than
the P-C(1) bond. For 16b this difference is a significant 0.07
Å (4%). This shorter P-C(2) bond of 16b of 1.776(8) Å is
within experimental uncertainties the same as in 6a and 14, and
deviates little, if any, from those reported for W(CO)5-
complexed phosphiranes (ca. 1.81 Å) and phosphirenes (ca. 1.79
Å).8b,17b Thus, while the difference in hybridization contributes
to the difference in lengths between the P(sp3)-C(1)(sp3) and
P(sp3)-C(2)(sp2) bonds, this effect appears to be rather small.
Evidently, the phosphirane P-C(1) bond is elongated. Ad-
ditional strain resulting from the distal olefinic bond is the likely
Prolonged heating of the mixture of vinylphosphiranes 19
gives as the main product phospholene 20, which evidently
results from a [1,3]-sigmatropic shift. Recently, we25 and
others17a,26 have reported on such rearrangements.
Experimental Structures. The ORTEP plot of 16b is shown
in Figure 1. Its relevant bond distances and angles are
(24) Quast, H.; Weise, A. Angew. Chem., Int. Ed. Engl. 1974, 13, 342.
(25) (a) Wang, B.; Lammertsma, K. J. Am. Chem. Soc. 1996, 118, 1690.
(b) Lammertsma, K.; Hung, J.-T.; Chand, P.; Gray, P. M. J. Org. Chem.
1992, 57, 6557. (c) Hung, J.-T.; Lammertsma, K. J. Org. Chem. 1993, 58,
1800.
(26) Richter, W. J. Chem. Ber. 1985, 118, 1575.

2-Alkylidenephosphiranes
J. Am. Chem. Soc., Vol. 119, No. 36, 1997 8435
Table 2 Absolute (au) and Relative Energies (kcal/mol) of C3H5P
and C5H9P Isomers
cause of this effect; noteworthy is the similar influence of the
cyclopropyl ring. Small differences in the phosphirane CC bond
may suggest a conjugative effect with the exocyclic double bond.
Thus, the C(1)-C(2) bond of 16b of only 1.47(1) Å is shorter
than those of both 6a and 14, albeit that the differences fall
within the error margins.8b,23 However, the 1.31(1) Å short
double bond of 16b would argue against such an electronic
effect.
Because of its simple substitution pattern, 16b shows no signs
of steric distortion, which sets it apart not only from 6a but
also from the sterically congested methylenediphosphiranes 8
and 9. For example, structure 9 has a 2-fold axis, and the
exocyclic dCCl2 group of structure 8 is twisted by 20.4°. Also
the phenyl rings of 8 are deformed into boat forms which
underscores the presence of significant steric strain. Methyl-
enephosphirane 6a shows a similar twisting of the double bond;
i.e., the C(1)-C(2)-C(3)-C(4) torsion angle is 26.9(6)°. In
sharp contrast, the isopropylene group (C(3), C(4), and C(5))
of structure 16b is within 1.1° coplanar with the phosphirane
ring, and the P-phenyl ring is planar within 1.8°.
These geometrical comparisons indicate that the phosphirane
rings of 16b and spirophospha[2.2]pentane 14 are surprisingly
similar, which may suggest these systems to have similar ring
strain as well.
The 31P NMR chemical shifts, which are a sensitive electronic
probe, also suggest similarity between the methylenephos-
phiranes 16a and 16b and spirophospa[2.2]pentane 14. The
slightly more shielded chemical shifts of 16a (δ ) -164.9 ppm)
and 16b (δ ) -162.4 ppm) may suggest a marginally smaller
CPC angle as in 14 (δ ) -154.8 ppm).8b,27 The X-ray
structures point in the same direction (47.8(4)° for 16b), but
the differences are within the error margins.
Compd
MP2(fc)/6-31G*
ZPVEa
∆Eb
23
24
25
26
-457.64570
-457.63957
-535.99143
-535.98761
45.83
46.20
83.52
83.60
0.00
3.52
0.00
2.47
a Zero-point vibrational energies at HF/6-31G* (kcal/mol). b Relative
MP2(fc)/6-31G* with inclusion of 0.89-scaled ZPVE.
For comparison, δ(31P) for the W(CO)5-complexed triphen-
ylphosphirene is -161.4 and -190.3 ppm for the uncomplexed
form.28a
The demetalation is also reflected in the NMR coupling
constants. Thus, while 22b has sizable 1JP-C couplings of 40.8
Hz for the olefinic carbon and 21.3 Hz for the methylene carbon
of the phosphirane ring, the W(CO)5-complexed parent 16a and
its dimethyl derivative 16b display an absence of these
couplings.8b Such a difference is not unexpected as transition
1
metal complexation significantly reduces the negative JP-C
coupling constant due to an increase in the phosphorus
coordination number from 3 to 4. In the case of 16a and 16b
1
this apparently results in a coincidental absence of JP-C for
1
the ring carbons. The large JP-C coupling constant of 40.8
Hz for the olefinic carbon of uncomplexed 22b compares well
with the 47.6 Hz coupling constant of 6a and the 43.9 Hz
coupling constant in triphenylphosphirene.28a Likewise, the
smaller coupling of 21.3 Hz for the methylene carbon in 22b is
similar to the 21.3 Hz coupling reported for the CPh2 carbon in
6a.
Theoretical Structures and Stabilities. Ab initio quantum
mechanical calculations were performed at MP2(fc)/6-31G* for
2-methylenephosphirane (23), its isomeric phosphamethylenecy-
clopropane (24), and its dimethyl derivatives 25 and 26. For
Demetalation. Embolded by the chemical stability of 16b,
we attempted decomplexation of its W(CO)5 group.28 Succes-
sive addition of iodine at -30 °C in the dark, and monitoring
of the reaction by 31P NMR for complete conversion to 21b (δ
) -183.3 ppm), followed by quenching with N-methyl imida-
zole yielded after chromatography the parent isopropylidene-
phosphirane 22b, which was characterized by NMR spectro-
scopic methods. The expected reduced stability of this uncom-
plexed compound is reflected in its slow decomposition at room
temperature, but 22b keeps reasonably well in solution below
-10 °C. Maercker and Brieden16 reported a similar air
sensitivity for 2,3-diisopropylidene-1-phenylphosphirane, which
is a phosphirane with an exocyclic allene.
The W(CO)5 decomplexation is clearly reflected in the NMR
features. The change in phosphorus coordination is evident from
the 31P NMR chemical shifts. Thus, decomplexation results in
a 39.7 ppm upfield shift of the 31P NMR chemical shift from δ
) -162.4 ppm for 16b to δ ) -202.1 ppm for 22b. This
resonance is at much higher field than those of the highly
substituted methylenephosphiranes 5 (δ ) -93.4 ppm) and 6a
(δ ) -134.5 ppm), which is in accordance with expectations.
computational simplification, these parent systems have no
transition metal complex and the P-atom carries a hydrogen
instead of a phenyl substituent. Bachrach12 has reported earlier
on the geometry of 23, its P-H inversion barrier, and its ring
strain energy. The MP2(fc)/6-31G* structures of this species
and 24-26 are shown with relative energies (kcal/mol) in
parentheses. Absolute and relative energies with 0.89-scaled
zero-point vibrational energy (HF) corrections (ZPVE) are listed
in Table 2.
All methylenephosphiranes have PsC(1) bonds that are ca.
0.1 Å longer and therefore presumably also weaker than the
PsC(2)d bonds in accordance with the X-ray structure of 16b.
The differences between this structure and that calculated for
25 are modest. Both PsC bonds of the experimental structure
are shorter by 0.04-0.06 Å, which is mainly attributed to
(27) Goldwhite, H.; Rowsell, D.; Vertal, L. E.; Bowers, M. T.; Cooper,
M. A.; Manatt, S. L. Org. Magn. Reson. 1983, 21, 494.
(28) (a) Marinetti, A.; Mathey, F.; Fischer, J.; Mitschler, J. J. Chem.
Soc., Chem. Commun. 1984, 45. (b) Marinetti, A.; Fischer, J.; Mathey, J.
F. J. Am. Chem. Soc. 1985, 107, 5001.

8436 J. Am. Chem. Soc., Vol. 119, No. 36, 1997
Krill et al.
positive. Mass spectra were recorded on a HP 5985 at 70 eV. Melting
points were determined on an electrothermal melting point apparatus
and are uncorrected. Elemental analyses were performed by Atlantic
Microlab, Inc., Norcross, GA. All materials were handled under an
atmosphere of dry, high-purity nitrogen. Reagents and solvents were
used as purchased, except for THF, which was distilled from sodium-
benzophenone prior to use, and toluene, which was dried over molecular
sieves. Chromatographic separations were performed on silica gel
columns (230-400 mesh, EM Science). The synthesis of [5,6-
dimethyl-2,3-bis(methoxycarbonyl)-7-phenyl-7-phosphanorbornadiene-
]pentacarbonyltungsten (12) is described in ref 17c.
complexation by the W(CO)5 group and to a lesser extent to
P-phenyl stabilization.29 The same difference in PsC bond
lengths exists between the X-ray structure of the W(CO)5-
complexed phosphirane 1323 and the microwave structure30 of
the parent phosphirane C2H5P. This “tightening” effect in the
W(CO)5-complexed phosphiranes is in line with their remarkable
stabilities. The comparison of the phosphirane bond lengths
indicates that the observed methylenephosphirane 16b is not
particularly strained. This is also supported by the reported
“modest” 30 kcal/mol ring strain energy (HF/6-31G*) for 23.12
For phosphirane itself (C2H5P) Bachrach12 reported a ring strain
energy of only 20 kcal/mol. For comparison, methylenecyclo-
propane has a ring strain energy of 37 kcal/mol and cyclopro-
panone of 26 kcal/mol, and those of methyleneaziridine and
aziridine are 39 and 27 kcal/mol, respectively (same level).
These energies underscore that the methylenephosphiranes do
not need stabilizing substituents to be synthetically accessible.
It is also evident that blocking the reactive, tricoordinate
phosphorus with a transition metal group limits its reactivity.
2-Isopropylidenephosphirane (25) is a modest 2.5 kcal/mol
more stable than the isomeric 2,2-dimethyl-3-methylenephos-
phirane (26), which is in line with the observed formation of
16b. The minor geometrical differences between the two
isomers are as expected. Both P-C bonds of structure 25 are
shorter than the corresponding ones of its isomer 26. Introduc-
tion of two methyl groups on the ring carbon lengthens the
P-C(1) bond by nearly 0.02 Å. It is not surprising that large
sterically repulsive, but electronically stabilizing groups, as in
6a, lengthen this bond even more. Cleavage of the P-C(1)
bond would give a phosphatrimethylene biradical from which
a phosphamethylenecyclopropane could result. However, this
is not a likely event because 24 is 3.5 kcal/mol less stable than
23 at MP2(fc)/6-31G* + ZPE. Dimethyl substitution will not
influence this energy difference much.
Conclusions. The important observation of this work is that
stable methylenephosphiranes can be synthesized with surprising
ease using the carbene-like addition of the phosphinidene
complex Ph-P-W(CO)5 to allenes. These additions have been
accomplished with the parent allene, and its 1,1-dimethyl and
tetramethyl derivatives. The X-ray structure of 2-isopropy-
lidene-1-phosphirane 16b shows no signs of steric distortion.
Its geometrical parameters compare, in fact, well with those of
the reported X-ray structure of the W(CO)5-complexed
spirophospha[2.2]pentane 14. Demetalation of the W(CO)5
group from 16b was accomplished to give the uncomplexed
product 22b. Theoretical ab initio calculations on model
systems are in agreement with the experimental data, underscor-
ing the stability of these systems.
(2-Methylene-1-phenyl-1-phosphirane)pentacarbonyltungsten (16a).
Allene (15a) was passed through a 30 mL toluene solution containing
complex 12 (0.77 mmol, 0.50 g) and CuCl (30 mg, 0.3 mmol) at 55-
60 °C for 2.5 h and monitored by TLC. Following evaporation of the
solvent, chromatography of the reaction mixture on silica (hexane-
benzene, 4:1) and crystallization of the product gave 210 mg (55%) of
16a as colorless crystals. Mp: 102-103 °C. 31P NMR (C6D6): δ )
2
-164.9 (1JP-W ) 251.1 Hz). 13C NMR (C6D6): δ ) 197.7 (d, JP-C
2
) 29.3 Hz, trans-CO), 195.7 (d, JP-C ) 7.2 Hz, cis-CO), 132.1 (s,
H2Cd), 131.4 (d, 2JP-C ) 17.7 Hz, o-phenyl), 130.7 (s, p-phenyl), 128.9
(m, m-phenyl), 121.1 (s, dC-), 18.4 (s, PCH2). 1H NMR (C6D6): δ
) 7.29-6.91 (m, phenyl, 5 H), 5.88 (dm, 3JP-H ) 13.5 Hz, 1H, dCH2),
5.76 (dm, 3JP-H ) 31.5 Hz, 1H, )CH2), 1.51 (d, 2JH-H ) 12.0 Hz, 1H,
PCH2), 1.29-1.33 (m, 1H, PCH2).
(2-Isopropylene-1-phenyl-1-phosphirane)pentacarbonyltung-
sten (16b). Complex 12 (1 mmol, 0.64 g) and a 3-4-fold excess of
1,1-dimethylallene (15b) were heated at 55-60 °C in 5 mL of benzene
with CuCl (10 mg, 0.1 mmol) for 2 h in a sealed autoclave to give
after workup a yield of 385 mg (77%) of 16b as colorless crystals.
Mp: 84-86 °C. 31P NMR (C6D6): δ ) -162.4 (1JP-W ) 252.0 Hz).
2
13C NMR (C6D6): δ ) 198.0 (d, JP-C ) 29.5 Hz, trans-CO), 195.7
(d, 2JP-C ) 6.9 Hz, cis-CO), 142.8 (s, Me2Cd), 134.3 (d, 1JP-C ) 23.8
Hz, P-phenyl), 131.3 (d, 2JP-C ) 13.8 Hz, o-phenyl), 130.4 (s, p-phenyl),
128.8 (3JP-C ) 9.9 Hz, m-phenyl), 116.4 (s, Me2CdC-), 25.0 (d, 3JP-C
3
) 3.3 Hz, CH3), 23.3 (d, JP-C ) 8.4 Hz, CH3), 17.9 (s, PCH2). 1H
NMR: δ ) 7.37-7.3 and 6.96-6.93 (m, phenyl, 5 H), 1.86 (d, 4JP-H
4
) 2.7 Hz, CH3), 1.62 (d, JP-H ) 3.0 Hz, CH3), 1.60-1.53 (m, 1H)
and 1.44-1.37 (m, 1H, CH2). MS (184W): m/z (rel intens) 500 (M,
21.1), 444 (M - 2CO, 18.0), 416 (M - 3CO, 19.9), 388 (M - 4CO,
25.9), 376 (PhPW(CO)3, 4.2), 360 (M - 5CO, 93.6), 348 (PhPW(CO)2,
50.4), 320 (PhPW(CO), 46.9), 292 (PhPW, 100). Anal. Calcd for
C16H13PWO5: C, 38.41; H, 2.60. Found: C, 38.50; H, 2.62.
(2-Isopropylene-3,3-dimethyl-1-phenyl-1-phosphirane)penta-
carbonyltungsten (16c). Complex 12 (1 mmol, 0.65 g) and tetra-
methylallene (15c) (1 mmol, 0.09 g) were heated at 45-50 °C in 20
mL of toluene with CuCl (100 mg) under an atmosphere of nitrogen
for 1 h to give after workup 430 mg (58%) of a 1:1:1 isomeric mixture
of 16c, 19a, and 19b, as determined by their 31P NMR chemical shifts
at δ ) -134.8 (1JP-W ) 245 Hz), -137.4 (1JP-W ) 255.0 Hz), and
-142.5 (1JP-W ) 255.0 Hz) ppm, respectively. After chromatography
16c was obtained as a colorless oil. 31P NMR (CDCl3): δ ) -134.8
(1JP-W ) 245.0 Hz). 13C NMR (CDCl3): δ ) 195.0 (d, 2JP-C ) 12.1
This leads to the conclusion that the addition of phosphin-
idenes to allenes is a convenient and versatile synthetic route
to simply substituted, but stable, methylenephosphiranes. It is
a welcome alternative method to the carbene-1-phosphaallene
and 2-phosphacumulene cyclization routes, both of which are
more limited by their accessibility of the phosphorus-containing
unsaturated systems.
1
Hz, cis-CO), 125.9 (s, Cd), 124.2 (d, JP-C ) 7.7 Hz, PCd), 127.1-
131.4 (m, Ph), 29.3 (m, CH3), 24.3 (m, CH3), 21.9 (s, CH3), 19.1 (d,
1JP-C ) 45.9 Hz, PCMe2). 1H NMR (CDCl3): δ ) 7.54-7.20 (m,
3
Ph, 5H), 2.12 (s, CH3, 3H), 1.85 (d, JP-H ) 13.5 Hz, CH3, 3H), 1.50
3
(m, CH3, 3H), 1.16 (d, JP-H ) 10.7 Hz, CH3, 3H).
(2-(1′-Isobutylene)-2-methyl-1-phenyl-1-phosphirane)pentacar-
bonyltungsten (19a). Complex 12 (1 mmol, 0.65 g) and 2,4-
dimethylpenta-2,3-diene (15c) (2 mmol, 0.18 g) were heated at 55 °C
in 20 mL of toluene with CuCl (100 mg) under an atmosphere of
nitrogen for 1 h to give after workup 430 mg (58%) of a 1:1 isomeric
mixture of 19a and 19b, as determined by their 31P NMR chemical
shifts at δ ) -137.4 and -142.5, respectively. After chromatography
and crystallization 19a was obtained as a colorless oil. 31P NMR
(CDCl3): δ ) -137.4 (1JP-W ) 258.6 Hz). 13C NMR (CDCl3): ) δ
Experimental Section
NMR spectra were recorded on a Brucker ARX 300, wide-bore
spectrometer. Chemical shifts are referenced in parts per million to
1
internal Me4Si for the H and 13C NMR spectra and to external 85%
H3PO4 for the 31P NMR spectra. Downfield shifts are reported as
2
3
196.4 (d, JP-C ) 8.0 Hz, cis-CO), 140.5 (d, JP-C ) 10.1 Hz, Cd),
(29) For example, the P-C bond lengths of triphenylphosphirene (1.8197
Å, X-ray structure)28c are shortened by 0.03-0.05 Å on W(CO)5 complex-
ation (i.e., 1.7901 and 1.7868 Å, X-ray structure).17b
(30) Bowers, M.; Baudet, R. A.; Goldwhite, H.; Tang, R. J. Am. Chem.
Soc. 1969, 91, 17.
127.4 (s, CHd, 129.1-132.8 (m, Ph), 30.6 (d, 1JP-C ) 17.4 Hz, CH2),
1
25.8 (t, JP-C ) 11.0 Hz, C), 25.9 (m, CH3), 23.3 (s, CH3), 20.4 (s,
CH3). 1H NMR (CDCl3): δ ) 7.40-7.15 (m, Ph, 5H), 5.15 (d, 3JP-H
) 12.8 Hz, CHd), 1.83 (s, CH3, 3H), 1.67 (d, 4JH-H ) 4.41 Hz, CH3,

2-Alkylidenephosphiranes
J. Am. Chem. Soc., Vol. 119, No. 36, 1997 8437
3
3
3H), 1.47 (d, JP-H ) 19.8 Hz, CH2, 1H), 1.43 (d, JP-H ) 29.9 Hz,
148°, with index ranges -18 e h e 18, -8 e k e 0, e 0 -l e 22;
3272 were observed with I > 3σ(I), abs coeff ) 129.746 cm-1, 209
parameters. The data manipulation and structure solution were carried
out using the Enraf-Nonius MolEN software. Intensities were corrected
for Lorentz and polarization effects; no decay was observed, and an
empirical absorption correction (Tmax ) 99.35%, Tmin ) 42.25%) was
applied. Variances were assigned on the basis of standard counting
3
CH2, 1H), 1.10 (d, JP-H ) 11.8 Hz, CH3, 3H).
(2,2,4-Trimethyl-1-phenyl-3-phospholene)pentacarbonyltung-
sten (20). Monitoring of the reaction of 15c with complex 12 at 45
°C using 31P NMR showed the appearance not only of δ ) -134.8
ppm (16c) but also of two additional resonances at δ ) -137.4 (1JP-W
) 255 Hz) and -142.5 ppm (1JP-W ) 255 Hz) (19a,b). On continued
heating 19a,b disappeared with the simultaneous emergence of a
resonance at δ ) 40.0 ppm (1JP-W ) 231 Hz) (20). This phospholene
20 was separated by preparative thin layer chromatography. 31P NMR
(CDCl3): δ ) 40.0 (1JP-W ) 230.6 Hz). 13C NMR (CDCl3): δ )
2
statistics, with the addition of an instrumental uncertainty term, 0.04Fo .
The structure was solved by standard Patterson and difference Fourier
techniques and refined by weighted full-matrix least squares. The final
model contains anisotropic thermal parameters for all non-hydrogens,
and isotropic hydrogen atoms in calculated positions riding on their
attached carbon atoms. The final difference Fourier map contained
maximum ∆F values near the W atom (∆Fmax ) 0.741, ∆Fmin
)
-0.400), with no peaks interpretable as extra atoms. Final R ) 5.54%,
Rw ) 7.69%, GOF ) 2.526, ∆/σmax ) 0.00.
2
2
198.9 (d, JP-C ) 21.7 Hz, trans-CO), 197.1 (d, JP-C ) 7.1 Hz, cis-
3
CO), 135.9 (s, dCsCH3),135.6 (d, JP-C ) 5.1 Hz, dCsH), 134.6
(d, 1JP-C ) 30.3 Hz, ipso-Ph), 131.7 (d, 2JP-C ) 10.2 Hz, o-Ph), 129.9
4
3
(d, JP-C ) 1.9 Hz, p-Ph), 128.1 (d, JP-C ) 8.8 Hz, m-Ph), 48.2 (d,
1JP-C ) 22.7 Hz, C), 39.2 (d, JP-C ) 27.8 Hz, CH2), 29.2 (d, J )
15.4 Hz, CH3), 24.2 (s, CH3), 18.7 (d, 3JP-C ) 5.5 Hz, dCsCH3). 1H
NMR (CDCl3): δ ) 7.72-7.12 (m, Ph, 5H), 5.17 (d, 3JP-H ) 16.9 Hz,
1
2
Computational Methods. The ab initio calculations30 were carried
out using the GAUSSIAN 92 and 94 suite of programs.31 Geometries
23-26 were optimized at the SCF level with the d-polarized split-
valence 6-31G* basis set and with inclusion of the effects of valence
electron correlation by using Møller-Plesset perturbation theory at
second order (MP2(fc)). The force constant matrices, vibrational
harmonic frequencies, and zero-point vibrational energies (ZPVE) were
calculated analytically for the HF-optimized geometries. All structures
are minima, characterized by real frequencies only.
3
CHd), 3.10 (m, CH2), 1.93 (s, CH3, 3H), 1.44 (d, JP-H ) 28.4 Hz,
CH3, 3H), 0.86 (d, JP-H ) 12.9 Hz, CH3, 3H).
3
2-Isopropylene-1-phenyl-1-phosphirane (22b). To a stirred solu-
tion of complex 16b (0.5 mmol, 250 mg) in 10 mL of CHCl3 was
added dropwise an equimolar amount of iodine in 10 mL of CHCl3 at
-30 °C in the dark over 1 h. The reaction was followed by 31P NMR
spectroscopy to minimize excess of iodine and to verify completion of
the reaction to the W(CO)4I2 complex 21b. 31P NMR (C6D6): δ )
-183.3 (1JP-W ) 173.2 Hz). The reaction mixture was allowed to
warm to room temperature, and a 6-fold excess of N-methylimidazole
was added over 5 min, after which the reaction mixture was chromato-
graphed over silica with hexane to yield 48 mg (54%) of 22b as a
light yellow oil. 31P NMR (CDCl3): δ ) -202.1. 13C NMR
Acknowledgment is made to the National Science Founda-
tion (Grant CHE-9500344) and the donors of the Petroleum
Research Fund, administered by the American Chemical Society,
for support of this study.
Supporting Information Available: Positional and thermal
parameters and a complete listing of bond lengths and angles
for 16b (10 pages). See any current masthead page for ordering
and Internet access instructions.
(CDCl3): δ ) 139.1 (d, 1JP-C ) 40.8 Hz, P-phenyl), 136.0 (d, 2JP-C
)
2
9.2 Hz, Me2Cd), 131.8 (d, JP-C ) 19.5 Hz, o-phenyl), 128.5 (s,
p-phenyl), 128.2 (d, 3JP-C ) 5.2 Hz, m-phenyl), 120.5 (d, 1JP-C ) 40.8
1
Hz, Me2CdC-), 24.6 (s, CH3), 23.8 (s, CH3), 15.7 (d, JP-C ) 21.3
Hz, PCH2). 1H NMR (CDCl3): δ ) 7.37-7.14 (m, phenyl, 5 H), 1.97
(s, CH3), 1.93 (s, CH3), 1.69-1.61 (m, 1H) and 1.34-1.16 ppm (m,
1H, CH2). MS (184W): m/z (rel intens) 176 (M, 44.9), 161 (M -CH3,
54.7), 134 (M - C3H6, 100), 107 (PC6H4, 80.7).
JA971340W
(31) Hehre, W. J.; Radom, L.; Schleyer, P.v.R.; Pople, J. A. Ab Initio
Molecular Orbital Theory, Wiley: New York, 1986.
(32) GAUSSIAN 92, Revision B; GAUSSIAN 94, Revision B.1: Frisch,
M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.; Johnson, B. G.;
Robb, M. A.; Cheeseman, J. R.; Keith, T.; Petersson, G. A.; Montgomery,
J. A.; Raghavachari, K.; Al-Laham, M. A.; Zakrzewski, V. G.; Ortiz, J. V.;
Foresman, J. B.; Cioslowski, J.; Stefanov, B. B.; Nanayakkara, A.;
Challacombe, M.; Peng, C. Y.; Ayala, P. Y.; Chen, W.; Wong, M. W.;
Andres, J. L.; Replogle, E. S.; Gomperts, R.; Martin, R. L.; Fox, D. J.;
Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart, J. P.; Head-Gordon, M.;
Gonzalez, C.; Pople, J. A., Gaussian, Inc., Pittsburgh, PA, 1995.
Crystal Structure Analysis of 16b. C16H13O5PW, M ) 500.10,
crystal size 0.28 × 0.56 × 0.15 mm, monoclinic, space group P21/c, a
) 15.151(1) Å, b ) 6.947(5) Å, c ) 17.791(2) Å, â ) 104.276(8)°, V
) 1814.8(3) Å3, Z ) 4, Fcalc ) 1.830 g‚cm-3, 4135 unique reflections
were measured on an Enraf-Nonius CAD4 diffractometer (Ni-filtered
monochromatized Cu KR radiation (λ ) 1.5418 Å)) using an ω/2θ
scan mode (scan width 1.14°, background/scan ratio 0.5) up to 2θ )