Phospholes with reduced pyramidal character from steric crowding. 1. Synthesis and NMR characterization of 1-(2,4-di-tert-butyl-6-methylphenyl)-3-methylphosphole

Article abstract of DOI:10.1021/jo960640d

The sterically crowded 1-(2,4-di-tert-butyl-6-methylphenyl)-3-methylphosphole was synthesized by dehydrohalogenation of the corresponding 3,4-dibromophospholane, in order to probe the possibility that the steric congestion would cause some flattening of the phosphorus pyramid and an increase in electron delocalization. The phosphole was a recrystallizable solid with ³¹P NMR δ 1.8. Semiempirical calculations indicated that the pyramidal shape was retained but was noticeably flatter than in 1-phenylphosphole. In the low energy conformation, the phosphole and phenyl ring planes are approximately orthogonal, with the 2-tert-butyl group in the less crowded position that is syn to the lone pair on phosphorus. The 6-methyl group is positioned under the phosphole ring. This conformational prediction was amply confirmed by several chemical shift and coupling effects in the ¹³C NMR spectrum. The ¹H NMR spectrum displayed an unusually large four-bond coupling (6 Hz) of ³¹P to the m-phenyl proton syn to the lone pair (and none to the anti-meta proton), consistent with the orthogonal conformation. The oxide of the phosphole showed more stability than that of less crowded phospholes and gave a ³¹P NMR signal that was detectable over a several hour period at room temperature. The oxide proceeded to give the usual Diels-Alder dimer and also formed a cycloadduct with N-phenylmaleimide. The phosphoryl group of the latter was reduced with trichlorosilane to give the phosphine. This new 7-phosphanorbornene derivative gave the most downfield ³¹P NMR shift (δ 153.3) of any member of this family, all of which are characterized by remarkable deshielding in the syn isomer.

Full text of DOI:10.1021/jo960640d

J . Org. Chem. 1996, 61, 7801-7807
7801
P h osp h oles w ith Red u ced P yr a m id a l Ch a r a cter fr om Ster ic
Cr ow d in g. 1. Syn th esis a n d NMR Ch a r a cter iza tion of
1-(2,4-Di-ter t-bu tyl-6-m eth ylp h en yl)-3-m eth ylp h osp h ole†
Louis D. Quin,* Gyo¨rgy Keglevich,‡ Alexey S. Ionkin, Rajdeep Kalgutkar, and
Ga´bor Szalontai§
Department of Chemistry, University of Massachusetts, Amherst, Massachusetts 01003-4510
Received April 8, 1996X
The sterically crowded 1-(2,4-di-tert-butyl-6-methylphenyl)-3-methylphosphole was synthesized by
dehydrohalogenation of the corresponding 3,4-dibromophospholane, in order to probe the possibility
that the steric congestion would cause some flattening of the phosphorus pyramid and an increase
in electron delocalization. The phosphole was a recrystallizable solid with 31P NMR δ 1.8.
Semiempirical calculations indicated that the pyramidal shape was retained but was noticeably
flatter than in 1-phenylphosphole. In the low energy conformation, the phosphole and phenyl ring
planes are approximately orthogonal, with the 2-tert-butyl group in the less crowded position that
is syn to the lone pair on phosphorus. The 6-methyl group is positioned under the phosphole ring.
This conformational prediction was amply confirmed by several chemical shift and coupling effects
in the 13C NMR spectrum. The 1H NMR spectrum displayed an unusually large four-bond coupling
(6 Hz) of 31P to the m-phenyl proton syn to the lone pair (and none to the anti-meta proton), consistent
with the orthogonal conformation. The oxide of the phosphole showed more stability than that of
less crowded phospholes and gave a 31P NMR signal that was detectable over a several hour period
at room temperature. The oxide proceeded to give the usual Diels-Alder dimer and also formed
a cycloadduct with N-phenylmaleimide. The phosphoryl group of the latter was reduced with
trichlorosilane to give the phosphine. This new 7-phosphanorbornene derivative gave the most
downfield 31P NMR shift (δ 153.3) of any member of this family, all of which are characterized by
remarkable deshielding in the syn isomer.
Theoretical calculations by Chesnut and Quin1 have
recently been published that bring attention to the
importance of the geometry at phosphorus in controlling
the extent of electron delocalization in the phosphole ring
system and the relevance of this geometry in establishing
the 31P NMR shift of the system. These calculations, and
those of others,2 predict that delocalization would be
extensive in a phosphole with planarity at phosphorus,
but quite small in the natural form with pyramidal
phosphorus. The planar form is 25.8 kcal/mol higher in
energy.1 The low level of delocalization in phospholes is
clearly revealed by the chemical and physical properties
of typical members of the family, as is discussed in recent
reviews.3,4 The Bird Index,5 which is based on experi-
mental bond parameters, is a convenient way of assessing
the extent of delocalization in related cyclic systems, and
when applied to a phosphole for which structural data
are available (1-benzyl6), an Index of 35.5, below that of
furan (43; cf. to 59 for pyrrole and 66 for thiophene) was
obtained.7
In our theoretical studies,1 as the phosphorus atom was
gradually allowed to lose its pyramidal geometry and
approach planarity, the extent of delocalization was
observed to increase, and when full planarity was achieved
delocalization was extensive, as indicated by pronounced
shortening of the single bonds and lengthening of the
double bonds. We have attempted to perform flattening
of the phosphorus pyramid experimentally, by placing on
it a substituent of such size as to cause serious congestion
with the atoms of the phosphole ring. While it is unlikely
that this approach can force complete planarity at
phosphorus, it nevertheless could cause modification of
the measurable properties of the phosphole in directions
that would reveal an increase in electron delocalization.
Particularly suggestive of this effect would be the bond
parameters established by X-ray diffraction analysis, and
the ionization energies provided by photoelectron spec-
troscopy.
In selecting the bulky substituent to be placed on
phosphorus, we were guided by PM3 semiempirical
calculations, using the Spartan program (version 3.1,
Wavefunction, Inc.). We examined the bond angle at
phosphorus for carbon of the exocyclic substituent (â),
and the related parameter of the angle (R) by which the
exocyclic carbon atom falls below the plane of the C2-
P-C5 section of the ring. These angles are known for
1-benzylphosphole from X-ray analysis of the crystal6 to
be 106° and 66.9°, respectively, and the validity of our
calculations was confirmed by closely reproducing these
values for this model phosphole (106.3° and 66.5°,
respectively) when in the same conformation (with the
benzene ring lying under the phosphole ring) as found
in the crystal. Results of the calculations for several
groups on phosphorus are shown in Table 1 (the angles
* To whom inquiries should be directed.
† Presented at the International Conference on Phosphorus Chem-
istry, J erusalem, Israel, J uly 1995.
‡ Department of Organic Chemical Technology, Technical University
of Budapest, 1521 Budapest, Hungary.
§ NMR Laboratory, University of Veszpre´m, H8201 Veszpre´m,
Hungary.
X Abstract published in Advance ACS Abstracts, October 15, 1996.
(1) Chesnut, D. B.; Quin, L. D. J . Am. Chem. Soc. 1994, 116, 9638.
(2) Hughes, A. N.; Edgecombe, K. E. Heterocycles 1992, 33, 563.
Baldridge, K. K.; Gordon, M. S. J . Am. Chem. Soc. 1988, 110, 4204.
Bachrach, S. M. J . Org. Chem. 1993, 58, 5414; Nyulaszi, L. J . Phys.
Chem. 1995, 99, 586.
(3) Mathey, F. Chem. Rev. 1988, 88, 429.
(4) Hughes, A. N. In Handbook of Organophosphorus Chemistry;
Engel, R., Ed.; Marcel Dekker: New York, 1992; pp 512-514.
(5) Bird, C. W. Tetrahedron 1985, 41, 1409.
(6) Coggon, P.; Engel, J . F.; McPhail, A. T.; Quin, L. D. J . Am. Chem.
Soc. 1970, 92, 5779.
(7) Bird, C. W. Tetrahedron 1990, 46, 5697.
S0022-3263(96)00640-8 CCC: $12.00 © 1996 American Chemical Society

7802 J . Org. Chem., Vol. 61, No. 22, 1996
Quin et al.
Ta ble 1. Ca lcu la ted Effects of P -Su bstitu en t on th e P yr a m id a l Sh a p e a t P h osp h or u s in 1-R-p h osp h ole
Spartan version 3.1
POAV-2 method
θσπ -90, deg
R
P
oop, dega
R, degb
â, degc
C2-P-C5, deg
methyl
phenyl
1-adamantyl
benzyl
mesityl
2,4,6-trimethoxyphenyl
2,4-di-tert-butyl-6-methylphenyl
form 2A
form 2B
2,6-di-tert-butylphenyl
5.6
6.4
6.17
6.0
8.0
9.6
70.2
68.5
68.1
66.5d
65.2
61.7
103.8
105.0
105.3
106.3
107.1
109.5
90.1
90.2
90.1
90.3
90.4
90.6
28.23
27.55
27.42
26.80
26.29
24.89
8.9
9.1
10.0
63.2e
58.0e
56
108.5
90.5
90.4
90.8
25.55
23.75
22.74
109.9f
a
b
Angle by which P is out of plane (oop) (mean) of the phosphole ring carbon atoms. Deflection from C2-P-C5 plane. c C2-P-R bond
angle. The C5-P-R angle was essentially the same except as noted. Cf. to experimental value of 66.9°.6 e Confirmed by MOPAC (version
d
6.0) calculation. f C5-P-R angle 116.2°. Reference 8b.
g
should be considered as tentative and are to be refined
in later work with higher level programs). The predicted
value for R of 1-methylphosphole (70.2°) may be taken
as a starting point for discussion. Replacement by phenyl
or the more bulky 1-adamantyl reduced the value by only
about 2°. Mesityl caused a further reduction of 3°. The
flattening became quite pronounced for the 2,6-di-tert-
butylphenyl substituent, where the angle was reduced
to 56°. In these crowded aryl derivatives, it was found
that in the low-energy conformation the planes of the
phosphole and benzene rings are approximately perpen-
dicular, which causes placement of one ortho substituent
under the phosphole ring, the other above it. When the
benzene ring is unsymmetrically substituted, as in 2-tert-
butyl-6-methylphenyl, two conformations can be adopted,
with the smaller (methyl) substituent located under the
phosphole ring, as in 2A or above as in 2B. The energy
difference (3.96 kcal/mol) indicates a very strong prefer-
ence for 2A. As might be expected, the reduction in angle
R is larger (58°) with tert-butyl under the phosphole ring
than when methyl is under the ring (63.2°). In all
phospholes examined, the P atom was found to be forced
out of the plane of the four ring carbons, reaching a
maximum angle of deflection of 10.0° in 2,6-di-tert-
butylphenyl. The reality of this effect was demonstrated
years ago by the X-ray analysis of 1-benzylphosphole.6
in Table 1, using the calculated values for the bond angles
presented therein. The results are also recorded in Table
1. The value decreases steadily as the steric bulk of the
P-substituent increases.
We have proceeded to attempt the synthesis of phos-
pholes with a P-aryl group bearing the ortho substi-
tutents seen in 1 and 2. We have been successful with
the latter substitution pattern, and some of the properties
for the compound prepared (1-(2,4-di-tert-butyl-6-meth-
ylphenyl)-3-methylphosphole, 9, in Scheme 1) are de-
scribed in this paper. In the following paper, the
photoelectron spectrum and its use in assessing the
degree of flattening of the phosphorus pyramid are
presented. The more crowded 2,6-di-tert-butylphenyl
substitution pattern of 1 has so far proved difficult to
create in a phosphole. As will be seen, however, the
properties of 9 are of considerable interest, prompting
continued attempts to synthesize this and other crowded
phospholes in future work.
Syn th esis of 1-(2,4-Di-ter t-bu tyl-6-m eth ylp h en yl)-
3-m eth ylp h osp h ole (9). The preferred method of syn-
thesis of phospholes is that described by Mathey in which
the cycloadduct of a diene and a phosphorus(III) halide
is subjected to dehydrohalogenation with an amine,
usually a pyridine derivative.9 However, the cycloaddi-
tion reaction is retarded by steric hindrance in the
phosphorus halide,10 and an alternative route to 9 had
to be used in which the 3-phospholene ring is synthesized
with a P-chloro substituent11 that is then replaced by
attack of an aryl Grignard reagent. To convert the
phospholene to the phosphole, the phosphorus atom is
first protected as the oxide, bromine is added to the
double bond, and the oxide returned to the phosphine by
reduction with trichlorosilane. Dehydrohalogenation
with sodium methoxide then provides the phosphole. This
method has been used in the synthesis of a number of
phospholes,10,12 and its application to the synthesis of
phosphole 9 is outlined in Scheme 1.
The change in the pyramidal shape at phosphorus as
the steric bulk of the exocyclic substituent is increased
can also be expressed by the POAV (π-orbital axis vector)
treatment, recently introduced to describe the pyramidal
shape of a carbon center.8 This treatment provides a
single parameter (θσπ - 90, which for sp3C is 19.47° and
for sp2C is 0.0°. We have applied a simplified version
(POAV-28b) to calculate the change in the pyramidal
shape at the P center of the series of phospholes found
Phosphole 9 was isolated by silica gel chromatography
and obtained as an odorless solid that could be recrystal-
(9) Breque, A.; Mathey, F.; Savignac, P. Synthesis 1981, 983.
(10) Quin, L. D., The Heterocyclic Chemistry of Phosphorus; Wiley-
Interscience: New York, 1981; Chapt. 2.
(11) Quin, L. D.; Szewczyk, J . Phosphorus Sulfur 1984, 21, 161.
(12) (a) Quin, L. D.; Borleske, S. G.; Engel, J . F. J . Org. Chem. 1973,
38, 1858. (b) Deschamps, E.; Mathey, F. J . Org. Chem. 1990, 55, 2494.
(c) Deschamps, E.; Ricard, C.; Mathey, F. Heteroatom Chem. 1991, 2,
377.
(8) (a) Haddon, R. C. J . Am. Chem. Soc. 1990, 112, 3385. (b) Haddon,
R. C. Science 1993, 261, 1545.

Phospholes with Reduced Pyramidal Character
J . Org. Chem., Vol. 61, No. 22, 1996 7803
Sch em e 1
substituent. With one exception the intermediates had
the expected properties. Dibromophospholane oxide 7,
which was obtained after chromatography as a 4:1
mixture of diasteroisomers from addition of bromine
(presumed to be trans) to the phospholene oxide, had 31P
NMR shifts of δ 72.0 (major) and 71.5 (minor), but when
7 was prepared by oxidation of dibromo phospholane 8,
the major isomer had a shift of δ 50.5 (CDCl3). A similar
shift (δ 49.4) was obtained when a sample of the dibro-
mide formed from 6 was exposed briefly to triethylamine
at room temperature. When this oxide (designated 7A)
was mixed with an equal amount of the oxide (designated
7B) from the halogen addition to 6, a single peak was
obtained (δ 54.2 in CDCl3). The 13C NMR spectra of
samples 7A and 7B were nearly the same, establishing
the authenticity of both as dibromo oxide 7, and as will
be seen both underwent dehydrobromination to give the
same product. This peculiar NMR effect requires further
study before an explanation can be developed.
NMR Sp ectr a l Ch a r a cter iza tion of P h osp h ole 9.
P h osp h or u s-31. The 31P NMR shift of phosphole 9 in
CDCl3 was δ 1.8, somewhat upfield of that of the 1-phenyl
derivative (δ 7.59). This is in accord with the presence
of ortho substituents on the phenyl ring, which are
γ-related to phosphorus. Shielding is a common occur-
rence of γ substituents in both aliphatic and aromatic
compounds, as is seen from tabulated data15 for some
o-tolyl derivatives (cf. diphenyl-o-tolylphosphine, δ -13.0,
to triphenylphosphine, δ -4.7; the effects are additive,
and tri-o-tolylphosphine has δ -29.9). The effect on the
31P NMR shift of increased delocalization through any
flattening of the phosphorus pyramid is not expected to
be significant. Thus the calculated value for planar
1-methylphosphole is only 10.9 ppm downfield of that for
pyramidal 1-methylphosphole.1 The deshielding from a
moderate approach toward a flatter pyramid in 9 could
be offset by the γ shielding effect.
Ta ble 2. Ma ss Sp ectr a l Da ta a for P h osp h ole 9, Dim er 11,
a n d Cycloa d d u ct 12
Ca r bon -13. 13C NMR spectra generally reveal little
about cyclic electron delocalization but are quite sensitive
to influences from steric interactions. The spectrum of
phosphole 9 in fact abounds in both steric-derived
coupling and shift effects which uniquely define an
important feature of the geometry of the molecule,
namely, the preferred near-perpendicular relation of the
planes of the phosphole and phenyl rings as predicted
by calculations and shown in two conformations in Figure
1.
The preferred conformations shown in Figure 1 were
used in the interpretation of the spectrum (Table 3).
Shifts and 31P-13C coupling constants for the 3-meth-
ylphosphole moiety of 9 resembled, in general, those
reported16 for this moiety in 3-methyl-1-phenylphosphole,
also recorded in Table 3. However, there was a pro-
nounced upfield shift of the two R-carbons in 9 (C-2 by
3.5 ppm, C-5 by 3.7 ppm) relative to those in 3-methyl-
1-phenylphosphole. Upfield shifts frequently arise from
steric interaction with γ-related carbons; here the γ
carbons would be those at positions 8 or 12 of the phenyl
ring, in which case the effect should be reciprocal. This
is indeed present in a rather striking way for one of the
phenyl carbons; C-8 was assigned a shift of δ 122.8,
significantly upfield from the position found in a model
lized from acetone in analytically pure form. Unfortu-
nately, crystals suitable for X-ray diffraction analysis
have not yet been produced, and attempts at obtaining
structural parameters have failed. Phosphole 9 was
stable in the atmosphere, but was best preserved under
nitrogen. These properties are unlike those for the
corresponding 1-phenyl compound,13 an oxidizable, mal-
odorous liquid. The phosphole had the characteristic
feature14 of giving an intense peak (92% relative abun-
dance) in its mass spectrum (Table 2) for the molecular
ion. NMR spectral characterization is presented in the
following section.
The reactions of Scheme 1 proceeded in normal fashion
and showed no special influences from the large P-
(13) Mathey, F.; Mankowski-Favelier, R. Org. Magn. Reson. 1972,
4, 171.
(14) Quin, L. D.; Bryson, J . G.; Moreland, C. G. J . Am. Chem. Soc.
1969, 91, 3308.
(15) Tebby, J . C. Handbook of Phosphorus-31 Nuclear Magnetic
Resonance Data; CRC Press: Boca Raton, FL, 1991; Chapt. 6.
(16) Bundgaard, T.; J akobsen, H. J . Tetrahedron Lett. 1972, 3353.

7804 J . Org. Chem., Vol. 61, No. 22, 1996
Quin et al.
F igu r e 1. Computed minimum energy conformations of phosphole 9.
Ta ble 3. 13C NMR Sp ectr a l Da ta for P h osp h ole 9a
Sch em e 2
9
3-methyl-1-phenylphosphole15
carbon
δ
J PC, Hz
δ
J PC, Hz
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
125.1
148.7
135.8
132.3
18.9
143.1
122.8
125.5
152.5
121.4
157.0
18.7
3.0
5.7
15.4
0
128.56
147.38
140.07
136.0
1.2
8.8
7.6
5.3
3.7
10.5
19.0
8.0
0
mb
18.53
17.6
17.0
1.4
2.0
12.3
28.7
mb
2.8
11.5
0.5
0
130.62
132.98
127.99
128.09
127.99
132.98
bond coupling to the methyl at C-8 should also be small
in this conformation, and indeed this carbon shows no
coupling to P. Two-bond coupling is also controlled in
the same way by orientation of the lone pair, and this is
8.0
19.0
2
reflected here in the larger J P-C value for C-12 (28.7 Hz)
37.8
33.1
34.8
31.2
than for C-8 (17.0 Hz). Another consequence of the
conformational biasing is seen in the four-bond coupling
to C-15 of the o-tert-butyl group, which is remarkably
large at 11.5 Hz. The size of this coupling supports a
a
At 75.1 MHz in CDCl3 at 20 °C. 31P-13C couplings were
b
4
prediction18 that J P-C will be large in the structural
confirmed at 20 MHz. In deuterotoluene, C6 was a doublet with
J PC ) 0.7 Hz, and C13 was a doublet with J PC ) 3.6 Hz. No other
significant differences were noted.
fragment PsCdCsCsC if the lone pair is close to the
terminal C.
o-tolylphosphine (e.g., tri-o-tolylphosphine, δ 134.417). The
magnitude of the upfield shift of about 12 ppm suggests
a γ effect of about 6 ppm associated with each of the
phosphole ring carbons, which would be reasonable if
conformation 9A of Figure 1 dominated the conforma-
tional equilibrium. This conformation receives ample
documentation from coupling phenomena. Thus, the
magnitude of three-bond P-C coupling is empirically
related to the orientation of the lone pair on P in rigid
systems, being large when close to the coupled C, small
or even absent when remote. It is for this reason that
3J P-C to C-11 (12.3 Hz) is much larger than that to C-9
(1.4 Hz), since as seen in Figure 1 the restricted rotation
holds C-11 in closer proximity to the lone pair. The three-
The repositioning of the aryl Me and tert-butyl groups
can be accomplished either by rotation around the C-P
bond or by inversion of the phosphorus pyramid (Scheme
2). Pyramidal inversion in phospholes has been proposed
to have a significantly lower energy barrier (about 16
kcal/mol) than that in phosphines (about 35 kcal/mol);19
if sterically hindered phospholes have a similar barrier,
inversion could take place at or near room temperature.
However, in the present case of an unsymmetrically
substituted P-phenyl substituent in a plane orthogonal
to the phosphole ring plane, a larger energy difference
would be present between the two participants in the
(18) Quin, L. D. In Phosphorus-31 NMR Spectroscopy in Stereo-
chemical Analysis; Verkade, J . G.; Quin, L. D., Eds., VCH Publishers:
Deerfield Beach, FL, 1987; pp 417-420.
(19) Egan, W.; Tang, R.; Zon, G.; Mislow, K. J . Am. Chem. Soc. 1971,
93, 6205.
(17) J ohnson, L. F.; J ankowski, W. C. Carbon-13 NMR Spectra;
Wiley-Interscience: New York, 1972; p 489.

Phospholes with Reduced Pyramidal Character
J . Org. Chem., Vol. 61, No. 22, 1996 7805
Ta ble 4. 1H NMR Sp ectr a l Da ta for P h osp h ole 9a
9
1,3-Dimethylphosphole11
proton
δ
couplings, Hz
δ
couplings, Hz
31P, 37.4; H4, 1.6; CH3, ∼1.5
31P, 16.7; H5, 7.0; H2, 1.6
31P, 37.4; H4, 7.0; H2, 2.4
31P, 6; H2, ∼1.5
6.52
6.93
7.01
2.31
31P, 41; CH3, 1.5; H4, 1.33
31P, 12.15; H5, 7.5; H2, 2.5
31P, 40; H4, 7.0; H2, 2.4
31P, 3.3; H2, 1.5
b
H2
H4
H5
6.46
6.84
6.91
2.24
1.72
6.94
1.28
7.41
1.73
c
3-CH3
8-CH3
H9
10-C(CH3)3
H11
31P ∼0.5
H11, 2.3
31P, 6; H9, 2.3
d
12-C(CH3)3
31P, ∼2
a
b
At 300 MHz, in CDCl3 at 20 °C. For numbering, see Table 3. Confirmed by NOE of 5-6% to 3-CH3. c Confirmed by NOE of 4-5%
to 3-CH3. Confirmed by NOE of 10-12% to both C(CH3)3 groups.
d
inversion process, leading to a biasing toward the lower
energy form just as would be seen in a rotational
equilibrium. The uncertainty about the nature of the
mechanism for any interconversion that might take place
has no impact on the conclusions of the present study.
3-methyl group are not directly observable due to the
strong tendency to dimerize by a Diels-Alder mechanism
immediately upon formation. However, effects from
substituents can influence phosphole stability; with 3,4-
dimethyl substitution, the lifetime of the 1-phenyl de-
rivative is long enough that 31P NMR signals (δ 42.3) can
still be observed after several days of standing in the
refrigerator.20 We have proceeded to explore the effect
of the steric demand of the large aryl substituent in
3-methyl phosphole 9 on the stability of its oxide.
The oxide 10 was synthesized by oxidation of the
phosphole at room temperature with tert-butyl hydrogen
peroxide in chloroform. There was immediate formation
of a 31P NMR signal at δ 53.4. The signal rapidly began
to diminish with the appearance of the characteristic
doublet of doublets for the Diels-Alder dimer 11 (δ 59.0
and 85, J ) 39.5 Hz, also obtained by dehydrohalogena-
tion of dibromophospholane oxide 7 with triethylamine
in refluxing benzene), but was still observable after
several hours. It is concluded that the large aryl sub-
stituent has indeed provided some stabilization of the
phosphole oxide ring, but not enough to make compound
10 isolable.
The very strong preference for conformation 9A, pre-
dicted by computation and consistent with the 13C-31P
coupling effects, was confirmed by variable temperature
13C NMR studies. The temperature was reduced to -50
°C, with no indication of line broadening or the develop-
ment of a second set of signals that could be associated
with conformation 9B that might be in equilibrium with
the major conformation (9A) as established by NMR
spectroscopy. Similarly, the 13C NMR spectrum was
unchanged when recorded at 100 °C. Phosphole 9
therefore appears to be conformationally homogeneous
over the range -50 to 100 °C and is properly represented
by a single conformer, 9A.
1
P r oton NMR. The H NMR spectrum (Table 4) for
the 3-methylphosphole moiety of 9 was interpreted with
reference to that reported for 1,3-dimethylphosphole,12a
and with the aid of NOE measurements. There was
strong resemblance in the shifts for the two phospholes
and in their complex coupling patterns, and no special
effects were present. In the phenyl ring of 9, however,
an unusually large four-bond coupling (6 Hz) to 31P was
noticed for the meta proton (H-11) that was syn to the
lone pair orbital. No coupling was observed to the other
meta proton (H-9). This conformational effect provides
1H NMR spectral support to the conclusion that phosp-
hole 9 has adopted conformation 9A.
The reactivity of the phosphole oxide toward the
dienophile N-phenylmaleimide, commonly used as a
trapping agent for phosphole oxides,21 was also tested by
including it in the reaction with triethylamine in reflux-
ing benzene (Scheme 3). The Diels-Alder adduct 12 was
obtained as a crystalline solid in 72% yield. Thus the
steric congestion around P has exerted no influence on
the cycloaddition reactions.
An incidental observation made in the dehydrohalo-
genation studies was that the same results were obtained
when samples of the dibromide in form 7A or form 7B
with different 31P NMR shifts, as discussed earlier, were
used, attesting further to the similarity of these forms.
Phosphines in the 7-phosphanorbornene system are
characterized by extraordinary downfield shifting of the
31P NMR signal when the 7-substituent is in the syn
position, an effect which has recently received theoretical
interpretation.22 We have proceeded to remove the
oxygen from phosphorus of Diels-Alder adduct 12 in the
usual way23 with trichlorosilane, thus synthesizing the
7-phosphanorbornene 13, unique because of the presence
of steric crowding at phosphorus. The 31P NMR shift of
δ 153.3 was found to be even further downfield than that
We have considered the possibility of using the nuclear
Overhauser effect (NOE) that might prevail between
close-lying protons of the phosphole ring and the sub-
stituents (methyl as well as tert-butyl) on the phenyl ring,
to provide further insight into the conformational prefer-
ence of phosphole 9. However, the semiempirical calcu-
lations on the preferred conformation 9A (Figure 1)
showed that the two ring planes were not exactly
orthogonal, to relieve the congestion. In this conforma-
tion, one C-H of the tert-butyl group is 2.96 Å from H-5,
and a C-H of the methyl group is 2.8 Å from H-2.
Neither distance is fixed, however, since the conformation
allows some oscillation around the P-C bond axis. One
would therefore expect very similar NOE effects from
both methyl and tert-butyl groups, and this was seen in
the NOE experiments. However, the proximity of the
8-CH3 and 12-(CH3)3C signals (∆δ 0.01 ppm) made
selective irradiation difficult, and the observed NOE
values (all in the range 1-3%) were not meaningful.
(20) Quin, L. D.; Wu, X.-P. Heteroatom Chem. 1991, 2, 359.
(21) Quin, L. D. Rev. Heteroat. Chem. 1990, 3, 39.
(22) Chesnut, D. B.; Quin, L. D.; Moore, K. D. J . Am. Chem. Soc.
1993, 115, 11984.
(23) Quin, L. D.; Hughes, A. N. In The Chemistry of Organophos-
phorus Compounds; Hartley, F. R., Ed.; Wiley-Interscience: New York,
1990; pp 313-318.
Ch a r a cter istics of th e Oxid e of P h osp h ole 9.
Phosphole oxides with the only ring substituent being a

7806 J . Org. Chem., Vol. 61, No. 22, 1996
Quin et al.
Sch em e 3
h. Solvent was evaporated and the residue extracted four
times with 50 mL portions of n-pentane. Solvent of the
combined extracts was evaporated to give 16.0 g (98%)
of phosphine 5 (31P NMR (CDCl3) δ -31.3; MS, m/ z (rel
int) 302 (M+, 100), 287 (38), 234 (51), 219 (58), 178 (97),
57 (64)). A chloroform (60 mL) solution of phosphine 5
was slowly treated with 6.7 g (0.059 mol) of 30% hydrogen
peroxide at 0 °C with vigorous stirring. After 1 h, the
mixture was washed with 4 × 20 mL of water. The
organic phase was dried (MgSO4) and the solvent evapo-
rated to give 16.7 g (97%) of oxide 6: 31P NMR (CDCl3)
1
δ 59.3; H NMR (CDCl3) δ 1.26 (s, 9 H), 1.56 (s, 9 H),
1.63 (s, 3 H), 2.33 (s, 3 H), 2.55-2.98 (m, 4 H), 5.57 (d,
3J PH ) 31.3 Hz, 1 H), 6.80-7.53 (m, 2 H); MS, m/ z (rel
int) 318 (M+, 94), 303 (53), 250 (64), 235 (36), 204 (11),
189 (45), 57 (100); Anal. Calcd for C20H31OP: C, 75.43;
H, 9.81. Found: C, 75.83; H, 9.85.
3,4-Dibr om o-1-(2,4-d i-ter t-bu tyl-6-m eth ylp h en yl)-
3-m eth ylp h osp h ola n e 1-Oxid e (7). A solution of 2.5
mL (0.049 mol) of bromine in 50 mL of chloroform (or
benzene) was added dropwise to 15.5 g (0.049 mol) of
phospholene oxide 6 in 100 mL of chloroform (or benzene)
at 0 °C. After the addition, the mixture was stirred at
room temperature for 1.5 h. Evaporation of the solvent
left a semisolid residue which was purified by column
chromatography (silica gel, 3% methanol in chloroform)
to give 16.5 g (71%) of 7 as a mixture of diastereomers;
31P NMR (CDCl3) δ 72.0 (75%) and 71.5 (25%); 1H NMR
(CDCl3) δ 1.27 (s, 9 H), 1.51 (s, 9 H), 2.13 (s, 3 H), 2.65
(s, 3 H), 2.78-3.66 (m, 4 H), 4.34-5.10 (m, 1 H), 6.97-
7.50 (m, 2 H); 13C NMR, partial, C2 δ 49.3 (J ) 62 Hz),
C3 66.3 (8.9), C4 57.2 (7.1), C5 43.9 (62.6); 13C NMR,
complete, Supporting Information Table 1): CI-MS: 477
(M+ + H); EI-MS, m/ z (rel int) 476 (M+, <0.5), 397 (67),
317 (100), 250 (41), 235 (28), 57 (83).
of a related phosphine with a 7-phenyl substituent (14,
δ 126.419). Indeed, the value is the most downfield ever
recorded for a 7-phosphanorbornene, just exceeding the
previous extreme value (δ 152.5) found for the benzo
derivative 15.24 It may be significant that this compound
Compound 7 was not stable on standing. New 31P
NMR signals developed at ca. δ 52 and 54 (CDCl3)
presumably due to loss of HBr.
Tr ea tm en t of Dibr om op h osp h ola n e Oxid e 7 w ith
Tr ieth yla m in e a t Room Tem p er a tu r e. A solution of
2.58 g (0.0054 mol) of 7 [31P δ 72.0 (75%) and 71.5 (25%)]
in 14 mL of dry benzene was treated with 1.54 mL
(0.0108 mol) of triethylamine at room temperature. After
a 1 h period of stirring, the mixture was filtered and
solvent of the filtrate evaporated. The recovered 7 so
obtained was purified by column chromatography (silica
gel, 3% methanol in chloroform) to give 1.55 g (60%) of 7
as a mixture of diastereomers with 31P NMR (CDCl3) δ
49.4 (80%) and δ 50.6 (20%); 1H NMR (CDCl3) δ 1.29 (s,
also has a sterically demanding group (tert-butyl) at P,
suggesting that in general steric congestion at the syn-7
position may be associated with substantial extra deshield-
ing. This point has not yet been treated theoretically.
Exp er im en ta l Section 25
1-(2,4-Di-ter t-b u t yl-6-m et h ylp h en yl)-3-m et h yl-3-
p h osp h olen e 1-Oxid e (6). A solution of 0.059 mol of
(2,4-di-tert-butyl-6-methylphenyl)magnesium bromide in
THF (prepared from 1.42 g (0.059 mol) of magnesium and
16.7 g (0.059 mol) of 1-bromo-2,4-di-tert-butyl-6-methyl-
benzene (1) in 40 mL of dry THF) was added dropwise
to 7.2 g (0.054 mol) of 1-chloro-3-methyl-3-phospholene
(4, prepared from phosphinic acid 3 with HSiCl311) in 40
mL of THF at 0 °C with stirring under an argon
atmosphere. After the addition, the cooling bath was
removed and the contents of the flask were stirred for 2
4
9 H), 1.53 (s, 9 H), 2.15 (s, 3 H), 2.67 (d, J (P,H) ) 1.5, 3
H), 2.80-3.70 (m, 4 H), 4.37-4.83 (m, 1 H), 6.96-7.52
(m, 2 H); 13C NMR, partial, C2 δ 49.7 (63.1 Hz), C3 66.5
(8.5), C4 57.5 (7.5), C5 43.9 (62.6); complete, Supporting
Information, Table 1: CI-MS: 477 (M+ + H); EI-MS, m/ z
(rel int): 476 (M+, <0.5), 397 (35), 317 (100), 250 (20),
235 (11), 57 (57).
1-(2,4-Di-ter t-bu tyl-6-m eth ylph en yl)-3-m eth ylph os-
p h ole (9). A solution of 4.1 g (0.00858 mol) of dibromo
compound 7 (from 6, with major isomer 31P δ 76.7) in 30
mL of dry benzene was added to a mixture of 2.29 mL
(0.0283 mol) of pyridine, 0.95 mL (0.00941 mol) of
trichlorosilane and 30 mL of benzene under a nitrogen
atmosphere. After a reflux period of 4 h, the precipitated
amine salt was filtered off. Concentration of the filtrate
in vacuum gave 3.96 g (100%) of dibromophospholane 8:
31P NMR (CDCl3) δ -39.2. Oxidation of a small sample
(24) Quin, L. D.; Bernhardt, F. C. J . Org. Chem. 1986, 51, 3235.
(25) 31P NMR spectra (FT, 1H-decoupled) were recorded at 32.44
MHz; chemical shifts are referenced to 85% phosphoric acid and are
positive if downfield. 13C NMR spectra (FT, 1H-decoupled) were taken
at 75.1 MHz, and 1H NMR spectra at 80 or 300 MHz. EI and CI (with
isobutylene) mass spectra were obtained at 70 eV. Elemental analyses
were performed by the University of Massachusetts Analytical Labora-
tory. Manipulations involving phosphines were performed under
nitrogen or argon.

Phospholes with Reduced Pyramidal Character
J . Org. Chem., Vol. 61, No. 22, 1996 7807
of 8 in CDCl3 with 30% H2O2 gave 7 with 31P NMR δ 50.5
for the major isomer and 51.6 for the minor isomer.
A solution of 0.93 g (0.0172 mol) of sodium methylate
in 1.5 mL of methanol was added to a solution of 3.96 g
(0.00858 mol) of dibromophospholane 8 in 40 mL of dry
benzene. After 4 days of reflux in a nitrogen atmosphere,
the mixture was filtered and solvent was evaporated. The
residue so obtained was purified by column chromatog-
raphy (silica gel, 3% methanol in chloroform) to give 1.33
Diels-Ald er Rea ction of P h osp h ole Oxid e 10 w ith
N-P h en ylm a leim id e. Triethylamine (0.88 mL, 0.00628
mol) was added to a mixture of 1.5 g (0.00314 mol) of
dibromophospholane 7, 0.6 g (0.00346 mol) of N-phenyl-
maleimide, and 15 mL of benzene at reflux. After 22 h
of refluxing, the precipitated salt was filtered off and the
solvent evaporated. The crude product was purified by
column chromatography (silica gel, benzene-acetone 2:3
and then 6% methanol in chloroform) to afford 1.1 g (72%)
of 12: mp 182-183 °C (from diethyl ether-ethyl acetate
9:1); 31P NMR (CDCl3) δ 88.9; 1H NMR, Supporting
Information, Table 2; 13C NMR, Supporting Information,
Table 4; MS, Table 2. Anal. Calcd for C30H36NO3P: C,
73.60; H, 7.41. Found: C, 73.25; H, 7.12.
g (61%) of phosphole 9: mp 97-99 °C (from acetone); 31
P
1
NMR (CDCl3) δ 1.8; MS, Table 1; 13C NMR, Table 2; H
NMR, Table 3. Anal. Calcd for C20H29P: C, 79.94; H,
9.75. Found: C, 79.63; H, 9.32.
Phosphole 9 was formed in similar yield when the
above procedure was applied to a sample of dibromo oxide
7 with 31P δ 49.4 following the triethylamine treatment.
Oxid a tion of P h osp h ole 9. Oxidation of a small
sample of 9 in CDCl3 with one drop of 70% tert-butyl
hydroperoxide resulted in the appearance of an interme-
diate signal at δ 53.4, attributed to phosphole oxide 10.
Simultaneously with the reduction of this signal, the
shifts of the dimer (11, see below) appeared.
A portion of oxide 12 (100 mg) was reacted with
trichlorosilane/pyridine in benzene in the usual way.23
The product 13 was a solid that was recrystallized from
chloroform and had mp 189-190 °C; 31P NMR δ 153.3
(CDCl3); high-resolution FAB-MS 474.2554 (Calcd for M+
+ 1 (C30H37NO2P) 474.2562).
Dim er 11 of P h osp h ole Oxid e 10. Triethylamine
(0.88 mL, 0.00628 mol) was added to a solution of 1.5 g
(0.00314 mol) of dibromo compound 7 in 15 mL of
refluxing dry benzene. During the reaction, transient 31P
NMR signals at δ 52 and 54 were noted, presumably for
monobromo intermediates. After 22 h of stirring at
reflux, the mixture was filtered and the filtrate evapo-
rated. The crude product so obtained was purified by
column chromatography (silica gel, 4% methanol in
chloroform, then benzene-acetone 2:3) to give 0.77 g
(78%) of 11: mp 168-170 °C (from petroleum ether-
ethyl acetate 8:2); 31P NMR (CDCl3) δ 59.0 and 85.1 (3J PP
Ack n ow led gm en t. L.D.Q. thanks the Exxon Edu-
cation Foundation for an Award that made this inter-
national collaboration possible and the Army Research
Office for support. OTKA is thanked for awards to G.K.
(Grant T014917) and to G.Sz. (Grant T016260). Mass
spectral data were provided by K. Ujsza´szy, EGIS
Pharmaceuticals, Budapest, Hungary, and by the Mid-
west Center for Mass Spectrometry with partial support
by NSF Biology Division (Grant No. DIR 9017262). R.K.
thanks Professor Paul M. Lahti for assistance with the
semiempirical calculations which were performed on
computers made available from NSF (CHE-9204695)
and the Petroleum Research Fund of the American
Chemical Society. We thank a referee for suggesting
the calculation of the degree of pyramidalization angle
by the POAV treatment.
1
) 39.5 Hz); H NMR, Supporting Information, Table 2;
13C NMR, Supporting Information, Table 3; MS, Table
2. Anal. Calcd for C40H58O2P2: C, 75.92; H, 9.24.
Found: C, 76.32; H, 9.01. The reaction was faster in
refluxing toluene (16 h); when the reaction was conducted
with 2 equiv of NaOMe in benzene at room temperature,
dimer 11 formed within a few minutes (75%).
J O960640D