From 2-phosphino-2H-phosphirene to 1-phosphino-1H-phosphirene, lλ5,2λ3-diphosphete, and 1,2-dihydro-1λ3,2λ3-diphosphete: An experimental and theoretical study

Article abstract of DOI:10.1002/(sici)1521-3765(19990104)5:1<274::aid-chem274>3.0.co;2-p

The [bis(diisopropylamino)phosphino](trimethylsilyl)carbene 4 reacts cleanly with tert-butylphosphaalkyne 2 to give 2-phosphino-2H-phosphirene 5, which was spectroscopically characterized. Heterocycle 5 is thermally unstable and quantitatively rearranges after 3 h at room temperature into the 1λ⁵,2λ³-diphosphete 3. Irradiation of 5 at room temperature (λ=254 nm) with a Rayonnet photochemical reactor produces 1-phosphino-1H-phosphirene 6 (10 %), 1,2-dihydrodiphosphete 7 (3 %), and diphosphete 3 (87 %). Irradiation of 3 with a high pressure mercury lamp at λ = 254 nm affords the dihydrodiphosphete 7, which was isolated in 69 % yield. Calculations carried out on heterocycles 3 and 5-7 emphasize the crucial effect the amino substituents have on the stability of λ⁵-phosphacyclobutadiene derivatives, and show that the intrinsic difference in the thermodynamic stability between 1H and 2H- phosphirenes is rather small (68 kJ mol^-1).

Full text of DOI:10.1002/(sici)1521-3765(19990104)5:1<274::aid-chem274>3.0.co;2-p

FULL PAPER
From 2-Phosphino-2H-Phosphirene to 1-Phosphino-1H-Phosphirene,
1l5,2l3-Diphosphete, and 1,2-Dihydro-1l3,2l3-Diphosphete:
an Experimental and Theoretical Study
Muriel Sanchez,[a] ReÂgis ReÂau,[a] Colin J. Marsden,*[b]
Manfred Regitz,[c] and Guy Bertrand*[a]
Abstract: The [bis(diisopropylamino)-
phosphino](trimethylsilyl)carbene 4 re-
acts cleanly with tert-butylphosphaal-
kyne 2 to give 2-phosphino-2H-phos-
phirene 5, which was spectroscopically
characterized. Heterocycle 5 is thermal-
ly unstable and quantitatively rearrang-
es after 3 h at room temperature into the
1l5,2l3-diphosphete 3. Irradiation of 5 at
room temperature (l ˆ 254 nm) with a
Rayonnet photochemical reactor produ-
ces 1-phosphino-1H-phosphirene
(10%), 1,2-dihydrodiphosphete
(3%), and diphosphete 3 (87%). Irradi-
ation of 3 with a high pressure mercury
lamp at l ˆ 254 nm affords the dihydro-
6
7
diphosphete 7, which was isolated in
69% yield. Calculations carried out on
heterocycles 3 and 5 ± 7 emphasize the
crucial effect the amino substituents
have on the stability of l5-phosphacy-
clobutadiene derivatives, and show that
the intrinsic difference in the thermody-
namic stability between 1H and 2H-
Keywords: carbenes
´ diazo com-
pounds
phosphorus
´
phosphaalkynes
heterocycles
´
´
phosphirenes
is
rather
small
1
(68 kJmol ).
rearrangements
Introduction
2
[2‡2] Dimerization of alkynes and ring-expansion reactions
involving transient cyclopropenyl carbenes are well-known
methods for the synthesis of cyclobutadienes A, the archetype
of four-p-electron four-membered ring systems.[1] Similarly, in
the phosphorus series, transition metal-coordinated 1l3,3l3-
and 1l3,2l3-diphosphetes B and C[2] have been prepared by
inner-sphere dimerization of l3-phosphaalkynes, while it has
been shown that 2-phosphino-2H-azirines D rearrange into
1,2l5-azaphosphetes E[3] (Scheme 1).
We have already reported that the photolysis of [bis(diiso-
propylamino)phosphino](trimethylsilyl)diazomethane (1) in
the presence of tert-butylphosphaalkyne 2 afforded the
1l5,2l3-diphosphete 3 as the major product (90% yield).[4]
Since under photolysis conditions, derivative 1 is a precursor
of the l3-phosphinocarbene 4, which can also be regarded as
A
LyM
MLx
LyM
P
P
P
2
P
or
P
B
C
P
P
N
N
E
D
Scheme 1. Synthesis of unsaturated four-membered ring systems.
the l5-phosphaalkyne 4',[5] two mechanisms can account for
the formation of the non-antiaromatic four-p-electron four-
membered heterocycle 3:[6] the head-to-tail codimerization
(governed by steric factors) of the l5-phosphaalkyne 4' with
the l3-phosphaalkyne 2, or the [1‡2] cycloaddition of the l3-
phosphinocarbene 4 with 2, followed by a ring-expansion
reaction of the initially formed 2-phosphino-2H-phosphirene
Â
[a] Dr. G. Bertrand, Dr. M. Sanchez, Prof. R. Reau
Laboratoire de Chimie de Coordination du CNRS
205 route de Narbonne, F-31077 Toulouse Cedex (France)
Fax: (‡33)5-61-33-31-24
SiMe3
R2P
R2P
R2P
C
SiMe3
P
C
tBu
P
E-mail: gbertran@lcctoul.lcc-toulouse.fr
N2
tBu
1
2
3
[b] Prof. C. J. Marsden
Laboratoire de Physique Quantique, IRSAMC, UMR 5626-CNRS
Â
Universite Paul Sabatier, 118 route de Narbonne
R2P SiMe3
C
SiMe3
R2P
C
SiMe3
F-31062 Toulouse Cedex (France)
5
tBu
P
4
4'
[c] Prof. M. Regitz
Fachbereich Chemie der Universität
Erwin-Schrödinger-Strasse, D-67633 Kaiserslautern (Germany)
R = iPr2N
274
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0501-0274 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 1

274±279
in the absence of UV light, even on heating a toluene solution
at 508C for 10 h. Note, that in marked contrast, the
corresponding nonsilylated diazophosphine 9 reacted with 2
at room temperature to afford the diazaphosphole 10 in
quantitative yield. All attempts to remove dinitrogen from 10,
by photolysis or thermolysis, failed (Scheme 3).
5. Here we report a detailed experimental study of the
reaction leading to 3, as well as ab initio calculations on
different cyclic isomers of 3.
Results and Discussion
A careful examination of the 31P NMR spectrum of the crude
reaction mixture obtained by photolysis of the diazo deriva-
tive 1 with the phosphaalkyne 2 showed the presence of two
minor products 6 (3%) and 7 (6%), along with diphosphete 3
(90%). The 1-phosphino-1H-phosphirene structure of 6 was
established by comparison of its 31P NMR spectrum [d ˆ
171.5 (PC2), 85.4 (PN2) (J(P,P) ˆ 291.2 Hz)] with that of
the analogous compound 6a [d ˆ 164.6 (PC2), 83.3 (PN2)
(J(P,P) ˆ 290.1 Hz)], prepared by addition of the bis(diiso-
propylamino)trimethylstannylphosphine to the chlorophos-
phirene 8 (Scheme 2). The 1,2-dihydro-1l3,2l3-diphosphete
structure of 7 was established by comparison of its spectro-
scopic data with those of an authentic sample prepared by
another route described below.
P
C
2
tBu
no reaction
PR2
R2P
C
SiMe3
50°C, 10h
N2
1
N
+ 2
H C
N
R2P
C
H
r.t.
P
N2
tBu
9
10 (98% yield)
R=iPr2N
Scheme 3. Reaction of diazo derivatives 1 and 9 in the absence of UV light.
We then prepared the stable phosphinosilylcarbene 4/4' by
photolysis of 1. The reaction of 4/4' with a slight excess of 2
was monitored by multinuclear NMR spectroscopy at 308C.
After a few minutes the signals corresponding to both starting
materials disappeared, and new signals corresponding to only
one new compound, the 2-phosphino-2H-phosphirene 5, were
observed. The absence of a direct P ± P bond was indicated by
the small P ± P coupling constant [J(P,P) ˆ 26.6 Hz], while the
presence of a phosphaalkene moiety was apparent from a low
field doublet in the 13C NMR spectrum [d ˆ 195.6 (J(P,C) ˆ
38.6 Hz)]. The spectroscopic data for 5, as a whole, compared
well with those reported for the only other known stable 2H-
phosphirene 5a [13C NMR: d ˆ 215.4 (J(P,C) ˆ 50.9 Hz)].[7]
Therefore, the l3-phosphinocarbene 4 had reacted with the
phosphaalkyne 2 by a [1‡2]-cycloaddition process and not by
a [2‡2]-cycloaddition mechanism, as expected for the l5-
phosphaalkyne structure 4'. Indeed, the formation of 5 is
strictly analogous to that of 2H-phosphirene F, obtained by
reacting transient halogenocarbenes with 2[8] (Scheme 4).
PR2
SiMe3
tBu
SiMe3
tBu
hν
R2P
P
RP
RP
P
+
+
1
+
2
Me3Si
tBu
6 (3%)
7
3 (90%)
(6%)
PR2
P
Cl
P
R2PSnMe3
- Me3SnCl
Ph
tBu
Ph
tBu
6a
8
(92% yield)
R=iPr2N
Scheme 2. Reaction scheme for the preparation of 3, 6, 6a, and 7.
In order to prove whether l3-phosphinocarbene 4 or l5-
phosphaalkyne 4' was the precursor of products 3, 6, and/or 7,
we first checked that the diazo derivative 1 did not react with 2
- N2
+ 2
R2P SiMe3
1
R2P
C
SiMe3
hν
- 30°C
5
tBu
4
P
Abstract in French: Le [bis(diisopropylamino)phosphino]-
(trimØthylsilyl)carb›ne (4) rØagit avec le tert-butylphosphaacØ-
tyl›ne (2) en donnant le 2-phosphino-2H-phosphir›ne 5 qui a
ØtØ caractØrisØ par les mØthodes spectroscopiques usuelles.
LꢀhØtØrocycle 5 est instable thermiquement et se rØarrange
quantitativement apr›s 3 h à tempØrature ambiante en 1l5,2l3-
diphosph›te 3. Lꢀirradiation de 5 à tempØrature ambiante avec
un rØacteur photochimique Rayonnet (l ˆ 254 nm) conduit
aux 1-phosphino-1H-phosphir›ne 6 (10%), 1,2-dihydrodi-
phosph›te 7 (3%) et diphosph›te 3 (87%). Lꢀirradiation de
3 avec une lampe à mercure à haute pression à l ˆ 254 nm
donne le dihydrodiphosph›te 7 qui a ØtØ isolØ avec un
rendement de 69%. Des calculs menØs sur les hØtØrocycles 3
P
R'
X
tBu
+ 2
5a
R'
C
X
P
tBu
F
Scheme 4. Reaction scheme for the formation of 5.
The 2-phosphino-2H-phosphirene 5 appeared to be rather
unstable and quantitatively rearranged after 3 h at room
temperature into the 1l5,2l3-diphosphete 3; no trace of the
1H-phosphirene 6 and dihydrodiphosphete 7 was detected.
Interestingly, irradiation (l ˆ 254 nm) at room temperature of
a toluene solution of the 2H-phosphirene 5 afforded diphos-
phete 3 (87%), in addition to small amounts of the 1H-
phosphirene 6 (10%) and 1,2-dihydrodiphosphete 7 (3%)
(according to 31P NMR spectroscopy). Therefore, in contrast
to most of 2H-phosphirenes, which isomerize thermally into
Ã
et 5 ± 7 mettent en Øvidence le role dØterminant jouØ par les
substituants amino sur la stabilitØ des dØrivØs l5-phosphacy-
clobutadiØniques et montrent que la diffØrence intrins›que de
stabilitØ thermodynamique entre les 1H- et 2H-phosphir›nes
1
est relativement faible (68 kJmol ).
Chem. Eur. J. 1999, 5, No. 1
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0501-0275 $ 17.50+.50/0
275

G. Bertrand, C. J. Marsden et al.
FULL PAPER
1H-phosphirenes,[8, 9] 5 only rearranges into 6 under UV
irradiation. In order to confirm that the 1H-phosphirene 6 was
not an intermediate in the rearrangement of 5 to diphosphete
3, 1H-phosphirene 6a was heated and photolyzed under
various conditions; however, no trace of the ring-expansion
product, namely the 1l5,2l3-diphosphete 3a, was detected.
Lastly, it was of interest to explain the formation of 1,2-
dihydrodiphosphete 7 in the photolysis of 5. We found that
irradiation of the 1l5,2l3-diphosphete 3 with a high-pressure
mercury lamp at l ˆ 254 nm afforded the dihydrodiphosphete
7, which was isolated in 69% yield (Scheme 5).
SiMe3
S8
R2P
P
S
S
11 (51 and 90% yield)
tBu
5 and 3
SiMe3
12 (87 and 95% yield)
BH3
R2P
P
H3B
tBu
R=iPr2N
Scheme 7. Reaction scheme for the formation of 11 and 12.
DZP level of derivatives 3b, 5b, 6b, and 7b (where R ˆ H2N
instead of iPr2N used in the experimental work) is shown in
Figure 1,[10] and the relative energies (kJmol ) at the B3LYP/
1
SiMe3
DZP level of these isomers, as well as those of the parent
r. t.
R2P
P
tBu
R2P SiMe3
3
P
PR2
tBu
SiMe3
tBu
hν
5
RP
RP
+
+
3 (87%)
P
Me3Si
tBu
6 (10%)
(3%)
7
Ph
hν or ∆
R2P
P
6a
3
X
tBu
3a
7 (69% yield)
hν
R=iPr2N
Scheme 5. Reaction scheme for the formation of 7.
To summarize, it is clear that 2-phosphino-2H-phosphirene
5 is a rather unstable species, which thermally rearranges into
1l5,2l3-diphosphete 3, and photochemically into the 1-phos-
phino-1H-phosphirene 6. Compound 6 does not undergo a
thermal or photochemical ring-expansion reaction to give 3.
Finally, irradiation of diphosphete 3 gives rise to the isomeric
1,2-dihydrodiphosphete 7 (Scheme 6).
SiMe3
R2P SiMe3
∆
R2P
P
3
5
P
tBu
tBu
hν
hν or ∆
hν
X
PR2
SiMe3
tBu
RP
RP
P
Me3Si
tBu
7
6
Figure 1. Optimized geometry at the SCF/DZP level for derivatives 3b,
5b, 6b, and 7b with selected bond lengths [Š] and angles [8].
R=iPr2N
Scheme 6. Rearrangement of 5.
compounds 3c, 5c, 6c, and 7c (R ˆ H), are given in Scheme 8.
Not surprisingly, in both series, dihydrodiphosphetes 7b and
7c are by far the most stable isomers. However the order of
stability of the three other isomers is dependent on the nature
of the substituents. The NH2 substituents dramatically stabi-
lized the diphosphete 3b, which is in agreement with the
postulated structure of four-p-electron four-membered ring
systems containing a l5-phosphorus and where a positive
charge is located on the phosphorus.[6] Because derivatives 7
are substantially more stable than 3, it might appear surprising
that 3 can be prepared and that it is kinetically stable. A
Since, as mentioned above, there is only one known
example of a stable 2H-phosphirene,[7] we attempted to
isolate 5 as a thiophosphoranyl derivative or a BH3 adduct;
instead, new diphosphete derivatives 11 and 12, were isolated
in 51 and 87% yield, respectively. The same adducts 11 (90%
yield) and 12 (95% yield) were also prepared directly from
the diphosphete 3 (Scheme 7).
In order to gain more insight into the rearrangement
processes of Scheme 6, theoretical calculations were carried
out on model derivatives. The optimized geometry at the SCF/
276
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0501-0276 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 1

Phosphorus Heterocycles
274±279
than 6e by 36 kJmol [D(C H) ˆ 414 kJmol 1; D(P H) ˆ
1
X
P
X
PR2
P
PR2
1
R2P
P
RP
RP
326 kJmol ). Therefore, with regard to the difference in the
P
P
order of stability of 5b and 6b versus 5c and 6c, it can be
concluded that amino substituents on the P atom strongly
increase the stability of the P C bond compared with a P P
bond.
5b,c
6b,c
3b,c
7b,c
5d,e
6d,e
3
5
6
7
5
6
b: R = NH2
c: R = H
85
113
105
127
0
d: X = Cl
e: X = H
68
0
122
95
0
0
36
Conclusion
Scheme 8. The relative energies (kJmol 1) at the B3LYP/DZP level of
derivatives 3b, 5b, 6b, and 7b (where R ˆ H2N instead of iPr2N used in the
experimental work) as well as those of the parent compounds 3c, 5c, 6c,
and 7c (R ˆ H).
The formation of the 2-phosphino-2H-phosphirene 5 demon-
strates that the phosphinosilylcarbene 4 reacts as a normal
transient singlet carbene, although it is known that it features
a P C multiple bond. The rearrangement of 5 into diphos-
phete 3 illustrates the broad range of application of the ring-
expansion reaction method for the preparation of four-p-
electron four-membered heterocyles. The calculations em-
phasize the crucial role played by the amino substituents on
the stability of l5-phosphacyclobutadiene derivatives. More-
over, they show that the intrinsic difference in the thermody-
namic stability between 1H- and 2H-phosphirenes is rather
search for the transition state linking 3c and 7c showed that it
1
lies 85 kJmol above 3c at the B3LYP/DZP level of theory.
While this value should not, of course, be taken as accurate to
1
within 1 kJmol , we are confident that it is sufficiently large
to show that kinetic stability for 3 is to be expected. The
geometries of 3c, 7c, and the intermediate transition state
(TS) are shown in Figure 2. Interestingly, although the 2H-
1
1
small (68 kJmol ) and can be overcome by the appropriate
choice of substituents.
phosphirene 5c appears to be higher in energy by 10 kJmol
than the isomeric 1H-phosphirene 6c, the P-amino analogues
Experimental Section
All experiments were performed under an atmosphere of dry argon.
Melting points are uncorrected. 1H, 13C, 11B, and 31P NMR spectra were
recorded on Brucker AC80, AC200, WM250, or AMX400 spectrometers.
1H and 13C chemical shifts are reported relative to Me4Si as an external
standard. 11B and 31P NMR downfield chemical shifts are expressed relative
to external BF3, Et2O, and 85% H3PO4, respectively. Infrared spectra were
recorded on a Perkin ± Elmer FT-IR Spectrometer 1725X. Mass spectra
were obtained on a Ribermag R1010E instrument. Conventional glassware
was used.
Photolysis of diazophosphine 1 with phosphaalkyne 2: A solution of
[bis(diisopropylamino)phosphino](trimethylsilyl)diazomethane
(1,[5a]
1.09 g, 3.15 mmol) in ether (30 mL) and tert-butylphosphaalkyne (2,[11]
0.35 g, 3.47 mmol) was irradiated at l ˆ 254 nm for 5 h at room temper-
ature. According to the 31P NMR spectrum of the resulting mixture, two
other products 6 (3%) and 7 (6%) were present in addition to the 1l5,2l3-
diphosphete 3 (90%). The major product 3 was isolated and characterized
as already reported.[4, 12] The 1H-phosphirene 6 and 1,2-dihydrodiphos-
phete 7 were identified by 31P NMR.
Figure 2. Optimized geometry and selected bond lengths [Š] at the
B3LYP/DZP level of theory for 3c, 7c, and the transition state (TS) 7c.
1H-Phosphirene 6: 31P NMR (CDCl3, 81.015 MHz): d ˆ 171.5 (d,
present an inverse order of stability: 5b is lower in energy by
14 kJmol than 6b. As the instability of 2H-phosphirenes
towards their isomerization into 1H-phosphirene was be-
1
J(P,P) ˆ 291.2 Hz, PC2), 85.4 (d, J(P,P) ˆ 291.2 Hz, PN2).
Dihydrodiphosphete 7: 31P NMR (CDCl3, 81.015 MHz): d ˆ 3.0 (d,
J(P,P) ˆ 144.1 Hz, J(P,H) ˆ 13.6 Hz), 11.2 (d, J(P,P) ˆ 144.1 Hz).
ˆ
lieved to be the result of to the disappearance of the P C
1-Phosphino-1H-phosphirene (6a): A solution of [bis(diisopropylami-
no)](trimethylstannyl)phosphine[13] (0.94 g, 2.37 mmol) in ether (3 mL)
was added, at 788C, to a solution of 1-chloro-1H-phosphirene (8,[14]
0.53 g, 2.37 mmol) in ether(2 mL). The solution was allowed to warm to
room temperature then the solvent and Me3SnCl were removed under
vacuum at 408C. Derivative 6a was obtained as a very viscous pale yellow
oil. Yield: 0.92 g (92%); 31P NMR (C6D6, 162.000 MHz): d ˆ 164.6 (d,
double bond in favor of the more thermodynamically stable
C C double bond,[9] the later result was rather surprising. We
ˆ
therefore investigated the relative energy of simple 1H- and
2H-phosphirenes 5d, 5e, 6d, and 6e. The C Cl and P Cl
1
bond energies are almost identical (326 kJmol
and
1
331 kJmol , respectively), so that the relative stability of
the two isomers 5d and 6d indicates the intrinsic stabilities of
the respective ring systems. Because the 1H-phosphirene 6d is
only some 68 kJmol 1 more stable than the 2H-isomer 5d, it is
evident that if the C X bond energy is stronger than the
corresponding P X bond by at least this amount, 2H-
phosphirenes will be stable towards isomerization.
Indeed, for the parent compound, 5e is more stable
1
J(P,P) ˆ 290.1 Hz, PC2), 83.3 (d, J(P,P) ˆ 290.1 Hz, PN2); H NMR (C6D6,
400 MHz): d ˆ 1.09 (d, J(H,H) ˆ 6.6 Hz, 6H, CH3CHN), 1.23 (d, J(H,H) ˆ
6.6 Hz, 6H, CH3CHN), 1.31 (d, J(H,H) ˆ 6.8 Hz, 6H, CH3CHN), 1.43 (s,
9H, CH3C), 1.47 (d, J(H,H) ˆ 6.8 Hz, 6H, CH3CHN), 3.55 (septd,
J(H,H) ˆ 6.8 Hz, J(P,H) ˆ 8.8 Hz, 2H, CH3CHN), 3.94 (septd, J(H,H) ˆ
6.6 Hz, J(P,H) ˆ 7.7 Hz, 2H, CH3CHN), 7.14 (t, J(H,H) ˆ 7.2 Hz, 1H,
CHarom.), 7.27 (dd, J(H,H) ˆ 7.2, 8.1 Hz, 2H, CHarom.), 7.75 (d, J(H,H) ˆ
8.1 Hz, 2H, CHarom.); 13C NMR (C6D6, 100.630 MHz): d ˆ 24.80 (d,
J(P,C) ˆ 5.7 Hz, CH3CHN), 24.85 (d, J(P,C) ˆ 5.7 Hz, CH3CHN), 25.87 (d,
Chem. Eur. J. 1999, 5, No. 1
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0501-0277 $ 17.50+.50/0
277

G. Bertrand, C. J. Marsden et al.
FULL PAPER
J(P,C) ˆ 9.7 Hz, CH3CHN), 25.95 (d, J(P,C) ˆ 10.2 Hz, CH3CHN), 30.20 (d,
J(P,C) ˆ 1.5 Hz, CH3C), 34.53 (dd, J(P,C) ˆ 1.0, 8.9 Hz, CH3C), 49.34 (dd,
J(P,C) ˆ 1.0, 8.2 Hz, CH3CHN), 51.93 (dd, J(P,C) ˆ 6.5, 9.9 Hz, CH3CHN),
116.17 (dd, J(P,C) ˆ 10.1, 47.0 Hz, PhC), 128.35 (s, Carom.), 128.78 (s,
CHarom.), 130.82 (d, J(P,C) ˆ 1.5 Hz, CHarom.), 131.24 (dd, J(P,C) ˆ 2.4,
7.1 Hz, CHarom.), 134.56 (dd, J(P,C) ˆ 15.3, 50.7 Hz, C-tBu); C24H42N2P2
(420.56): C 68.54, H 10.06, N 6.66; found: C 69.10, H 10.34, N 6.90.
CH3C), 4.11 (septd, J(H,H) ˆ 6.8 Hz, J(P,H) ˆ 10.2 Hz, 4H, CH3CHN);
13C NMR (CDCl3, 62.896 MHz): d ˆ 3.85 (s, CH3Si), 24.70 (d, J(P,C) ˆ
4.3 Hz, CH3CHN), 25.34 (d, J(P,C) ˆ 3.8 Hz, CH3CHN), 31.54 (d, J(P,C) ˆ
4.1 Hz, CH3C), 41.47 (dd, J(P,C) ˆ 12.9, 48.1 Hz, CH3C), 50.00 (d, J(P,C) ˆ
3.5 Hz, CH3CHN), 136.31 (dd, J(P,C) ˆ 21.6, 24.5 Hz, PCSi), 200.31 (dd,
‡
J(P,C) ˆ 37.6, 64.8 Hz, PC-tBu); CIMS(CH4): m/e ˆ 481 [M ];
C21H46N2P2S2Si (480.77): C 52.46, H 9.64, N 5.83; found: C 52.54, H 9.77,
N 5.79.
(Phosphino)diazaphosphole (10): Neat tert-butylphosphaalkyne (2, 0,18 g,
2.00 mmol) was added at room temperature to a solution of bis(diisopro-
pylamino)phosphinodiazomethane (9,[5a] 0.54 g, 2.00 mmol) in ether
(30 mL). The solution was allowed to stand for 1 h at room temperature
and then the solvent was removed under vacuum. Derivative 10 was
obtained as an orange oil. Yield: 0.73 g (98%); 31P NMR (CDCl3,
Diphosphete ´ BH3 adduct 12:
A solution of BH3 ´ SMe2 (0.08 mL,
1.70 mmol) in THF (2.0m) was added dropwise, at 788C, to a solution
of 5 (0.71 g, 1.70 mmol) in ether (1 mL). The solution was allowed to warm
to room temperature, then it was filtered, the solvent removed under
vacuum, and the residue was washed with pentane (3 Â 5 mL). Derivative
12, along with another unidentified product (<5% according to 31P NMR
spectroscopy), was isolated as a light brown powder from a CH2Cl2/pentane
1
81.015 MHz): d ˆ 76.5 (d, J(P,H) ˆ 42.0 Hz, PC), 93.7 (brs, PN); H NMR
(CDCl3, 200.133 MHz): d ˆ 1.05 (d, J(H,H) ˆ 6.7 Hz, 12H, CH3CHN), 1.24
(d, J(H,H) ˆ 6.7 Hz, 12H, CH3CHN), 1.39 (s, 9H, CH3C), 3.36 (septd,
J(H,H) ˆ 6.7 Hz, J(P,H) ˆ 12.8 Hz, 4H, CH3CHN), 8.73 (dd, J(P,H) ˆ
42.0 Hz, J(P,H) ˆ 5.9 Hz, 1H, NH); 13C NMR (CDCl3, 20.149 MHz): d ˆ
23.16 (d, J(P,C) ˆ 8.8, CH3CHN), 23.94 (d, J(P,C) ˆ 6.9, CH3CHN), 32.03
(d, J(P,C) ˆ 6.6 Hz, CH3C), 36.20 (d, J(P,C) ˆ 17.3 Hz, CH3C), 47.16 (d,
J(P,C) ˆ 13.7 Hz, CH3CHN), 159.05 (dd, J(P,C) ˆ 51.2, 21.7 Hz, PCH),
192.06 (d, J(P,C) ˆ 63.7 Hz, PC-tBu); C18H38N4P2 (372.47): C 58.04, H 10.28,
N 15.04; found: C 57.85, H 10.14, N 15.27.
solution at
308C. Yield: 0.62 g, (87%). By means of the same
experimental procedure, but starting from 3, derivative 12 along with the
same unidentified product observed previously was obtained in 95% yield.
31P NMR (CDCl3, 162.000 MHz): d ˆ 61.8 (d q, J(P,P) ˆ 17.6 Hz, J(P,H) ˆ
11.2 Hz, N2P), 123.7 (brd, J(P,P) ˆ 17.6 Hz, PBH3); 1H NMR (CDCl3,
200 MHz): d ˆ 0.41 (s, 9H, CH3Si), 1.39 (s, 9H, CH3C), 1.44 (d, J(H,H) ˆ
6.4 Hz, 12H, CH3CHN), 1.45 (d, J(H,H) ˆ 6.4 Hz, 12H, CH3CHN), 4.09
(septd, J(H,H) ˆ 6.4 Hz, J(P,H) ˆ 11.2 Hz, 4H, CH3CHN), BH3 was not
observed; 13C NMR (CDCl3, 50.323 MHz): d ˆ 3.69 (d, J(P,C) ˆ 2.4 Hz,
CH3Si), 24.68 (d, J(P,C) ˆ 3.8 Hz, CH3CHN), 25.20 (d, J(P,C) ˆ 3.8 Hz,
CH3CHN), 30.94 (dd, J(P,C) ˆ 3.3, 2.6 Hz, CH3C), 42.24 (dd, J(P,C) ˆ 13.0,
47.2 Hz, CH3C), 49.78 (d, J(P,C) ˆ 5.5 Hz, CH3CHN), 143.07 (dd, J(P,C) ˆ
3.2, 37.4 Hz, PCSi), 199.42 (dd, J(P,C) ˆ 8.6, 13.7 Hz, PC-tBu); 11B NMR
(CDCl3, 128.379 MHz): d ˆ 30.6 (brs).
2-Phosphino-2H-phosphirene (5): A solution of [bis(diisopropylamino)]-
(trimethylsilyl)diazomethane (1, 0.25 g, 0.73 mmol) in ether (30 mL) was
irradiated at l ˆ 254 nm for 3 h at room temperature. The solvent was
removed under vacuum, and the residue then dissolved in [D8]toluene
(0.5 mL). This solution was transferred to a NMR tube, and neat tert-
butylphosphaalkyne (2, 0.115 mL, 0.81 mmol) was added dropwise at
308C. Derivative 5 was characterized in solution. 31P NMR ([D8]toluene,
162.003 MHz, 243 K): d ˆ 48.1 (d, J(P,P) ˆ 26.6 Hz, s2-P), 94.4 (d, J(P,P) ˆ
26.6 Hz, s3-P); 13C NMR ([D8]toluene, 50.323 MHz, 243 K): d ˆ 1.68 (d,
J(P,C) ˆ 5.9 Hz, CH3Si), 27.53 (m, CH3CHN), 29.01 (m, CH3CHN), 31.00
(d, J(P,C) ˆ 1.0 Hz, CH3C), 38.96 (s, CH3C), 45.90 (m, CH3CHN), 195.56 (d,
J(P,C) ˆ 38.6 Hz, PC-tBu), the PCSi carbon atom was not observed.
Computational details: We have used standard, well-calibrated computa-
tional methods incorporated into the Gaussian92 program.[15] In view of the
size of the molecules concerned, typical approximations were made so that
the computations were actually performed on model compounds in which
the SiMe3, tBu, and iPr substituents were all replaced by hydrogen atoms.
While these substituents are, no doubt, very important in a practical, kinetic
sense, we do not think that their electronic role is crucial for the bonding or
thermodynamic stabilities of the various isomers. Geometries were
optimized at the SCF level, initially with the compact 3-21G* basis set[16]
Thermal rearrangement of 5: A solution of 5 in [D8]toluene, prepared as
indicated above, was allowed to warm to room temperature. After 3 h, the
diphosphete 3 was the only product formed, according to 31P NMR
spectroscopy. After work up, 3 was isolated in 70% yield.
and subsequently with
a
more complete DZP basis.[17] Vibrational
frequencies were calculated for the various stationary points located, to
check that these are indeed true minima. Better estimates of relative
energies were then obtained by the use of a hybrid form of DFT theory,
usually known as B3LYP,[18] with the DZP basis in order to optimize the
geometries of those conformers of 3, 5, 6, and 7 which are most stable at the
SCF level. This method is generally thought to incorporate dynamic
correlation effects very effectively, and any nondynamic effects are at least
partially included, so that the relative energies will certainly be more
reliable than those obtained at the SCF level of theory.
Photochemical rearrangement of 5: A solution of 5 in [D8]toluene,
prepared as indicated above, was irradiated at l ˆ 254 nm at room
temperature for 3 h. According to 31P NMR spectroscopy, diphosphete 3
(87%), 1H-phosphirene 6 (3%), and 1,2-dihydrodiphosphete 7 (10%)
were formed.
1,2-Dihydro-1l3,2l3-diphosphete (7): A solution of 1l5,2l3-diphosphete (3;
0.75 g, 1.80 mmol) in ether (50 mL) was irradiated with a mercury vapor
lamp for 4 h at room temperature. The solvent was removed under vacuum
and the 1,2-dihydro-1l3,2l3-diphosphete 7 was isolated by column chro-
matography. Yield: 0.52 g (69%), colorless crystals; m.p. 1218C; 31P NMR
(CDCl3, 81.015 MHz): d ˆ 3.0 (d, J(P,P) ˆ 144.1 Hz, J(P,H) ˆ 13.6 Hz), 11.2
(d, J(P,P) ˆ 144.1 Hz); 1H NMR (CDCl3, 200 MHz): d ˆ 0.51 (s, 9H,
CH3Si), 1.05 (d, J(H,H) ˆ 6.7 Hz, 6H, CH3CHN), 1.23 (d, J(H,H) ˆ 6.5 Hz,
6H, CH3CHN), 1.43 (d, J(H,H) ˆ 6.8 Hz, 6H, CH3CHN), 1.46 (d, J(H,H) ˆ
6.7 Hz, 6H, CH3CHN), 1.49 (s, 9H, CH3C), 2.90 ± 3.15 (m, 4H, CH3CHN);
13C NMR (CDCl3, 50.323 MHz): d ˆ 2.25 (d, J(P,C) ˆ 2.8 Hz, CH3Si),
18.94 ± 26.38 (m, CH3CHN), 31.20 (s, CH3C), 38.21 (t-like, J(P,C) ˆ 13.0 and
13.0 Hz, CH3C), 44.19 ± 53.87 (m, CH3CHN), 142.09 (dd, J(P,C) ˆ 38.1,
34.3 Hz, PCSi), 171.75 (dd, J(P,C) ˆ 36.6, 26.7 Hz, PC-tBu); C21H46N2P2Si
(416.64): C 60.54, H 11.13, N 6.72; found: C 59.80, H 10.90, N 6.80.
Acknowledgements
We are grateful to the CNRS and the Centre National Universitaire Sud de
Calcul (project irs1013) for financial support of this work.
[1] For reviews on cyclobutadienes see: a) G. Maier, Angew. Chem. 1988,
100, 317; Angew. Chem. Int. Ed. Engl. 1988, 27, 309; b) M. Regitz, H.
Heydt, U. Bergsträsser, Adv. Strain Org. Chem. 1996, 5, 161; c) V. L.
Minkin, M. N. Glukhovtsev, B. Y. Simkin, Aromaticity and Antiaro-
maticity, Wiley, New York, 1994.
[2] For reviews on l3,l3-diphosphetes see: a) P. Binger, in Multiple Bonds
and Low Coordination in Phosphorus Chemistry (Eds.: M. Regitz,
O. J. Scherer), Thieme, Stuttgart, 1990, p. 90; b) J. F. Nixon, Coord.
Chem. Rev. 1995, 145, 201; c) R. Streubel, Angew. Chem. 1995, 107,
478; Angew. Chem. Int. Ed. Engl. 1995, 34, 436; d) J. F. Nixon,
Endeavour 1991, 15, 49; d) M. Regitz, Chem. Rev. 1990, 90, 191; See
also: e) M. D. Francis, D. E. Hibbs, M. B. Hursthouse, C. Jones,
K. M. A. Malik, Chem. Commun. 1996, 631; f) P. Binger, G. Glaser, S.
Albus, C. Krüger, Chem. Ber. 1995, 128, 1261.
1l5,2l5-Diphosphete (11):
A suspension of elemental sulfur (0.16 g,
0.63 mmol) in ether was added at 788C to a solution of 5 (0.26 g,
0.63 mmol) in ether (1 mL). The solution was allowed to warm to room
temperature, the solvent was removed under vacuum, and the residue
treated with ether (3 Â 5 mL). Derivative 11 was isolated as a pale yellow
powder (0.15 g, 51% yield). By means of the same experimental procedure
but starting from 3, derivative 11 was obtained in 90% yield. M.p. 548C
(decomp); 31P NMR (CDCl3, 101.256 MHz): d ˆ 87.3 (qd, J(P,H) ˆ 10.2 Hz,
J(P,P) ˆ 23.9 Hz, N2P), 127.3 (d, J(P,P) ˆ 23.9 Hz, PS2); 1H NMR (CDCl3,
250 MHz): d ˆ 0.42 (s, 9H, CH3Si), 1.40 (d, J(H,H) ˆ 6.8 Hz, 12H,
CH3CHN), 1.47 (d, J(H,H) ˆ 6.8 Hz, 12H, CH3CHN), 1.56 (s, 9H,
278
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0501-0278 $ 17.50+.50/0
Chem. Eur. J. 1999, 5, No. 1

Phosphorus Heterocycles
274±279
[3] a) G. Alcaraz, U. Wecker, A. Baceiredo, F. Dahan, G. Bertrand,
Angew. Chem. 1995, 107, 1358; Angew. Chem. Int. Ed. Engl. 1995, 34,
1246; b) V. Piquet, A. Baceiredo, H. Gornitzka, F. Dahan, G.
Bertrand, Chem. Eur. J. 1997, 3, 1757.
[10] It is interesting to note that 3 adopts C1 symmetry with two NH2
substituents on P2, though the parent PH2 derivative has Cs symmetry.
Conversely, 6, which has two NH2 substituents, adopts Cs symmetry,
whereas its parent, which contains the PH2 group, has no symmetry.
[11] G. Becker, G. Gresser, W. Uhl, Z. Naturforsch. B 1981, 36, 16.
Â
[4] R. Armbrust, M. Sanchez, R. Reau, U. Bergsträsser, M. Regitz, G.
Â
Bertrand J. Am. Chem. Soc. 1995, 117, 10785.
[12] M. Sanchez, R. Reau, H. Gornitzka, F. Dahan, M. Regitz, G. Bertrand,
[5] a) A. Igau, H. Grützmacher, A. Baceiredo, G. Bertrand, J. Am. Chem.
Soc. 1988, 110, 6463; b) A. Igau, A. Baceiredo, G. Trinquier, G.
Bertrand, Angew. Chem. 1989, 101, 617; Angew. Chem. Int. Ed. Engl.
1989, 28, 621; c) M. Soleilhavoup, A. Baceiredo, O. Treutler, R.
Ahlrichs, M. Nieger, G. Bertrand, J. Am. Chem. Soc. 1992, 114, 10959;
d) M. T. Nguyen, M. A. McGin, A. F. Hegarty, Inorg. Chem. 1986, 25,
2185; e) D. A. Dixon, K. D. Dobbs, J. A. Arduengo, III, G. Bertrand, J.
Am. Chem. Soc. 1991, 113, 8782; f) G. Bertrand, R. Reed, Coord.
Chem. Rev. 1994, 137, 323.
[6] For reviews on four-p-electron four-membered phosphorus hetero-
cycles see: a) L. Weber, Angew. Chem. 1996, 108, 2779; Angew. Chem.
Int. Ed. Engl. 1996, 35, 2618; b) G. Bertrand, Angew. Chem. 1998, 110,
282; Angew. Chem. Int. Ed. 1998, 37, 270.
J. Am. Chem. Soc. 1997, 119, 9720.
[13] H. R. G. Bender, E. Niecke, M. Nieger, H. Westermann, Z. Anorg.
Allg. Chem. 1994, 620, 1194.
[14] A. Marinetti, F. Mathey, J. Fischer, A. Mitschler, J. Chem. Soc. Chem.
Commun. 1984, 45.
[15] Gaussian92, M. J. Frisch, G. W. Trucks, M. Head-Gordon, P. M. W.
Gill, M. W. Wong, J. B. Foresman, B. G. Johnson, H. B. Schlegel, M. A.
Robb, E. S. Replogle, R. Gomperts, J. L. Andres, K. Raghavachari,
J. S. Binkley, C. Gonzalez, R. L. Martin, D. J. Fox, D. J. Defrees, J.
Baker, J. J. P. Stewart, S. Topiol, J. A. Pople, Gaussian, Pittsburgh, PA,
1992.
[16] M. S. Gordon, J. S. Binkley, J. A. Pople, W. J. Pietro, W. J. Hehre J.
Am. Chem. Soc. 1982, 104, 2797.
[7] O. Wagner, G. Maas, M. Regitz, Angew. Chem. 1987, 99, 1328; Angew.
Chem. Int. Ed. Engl. 1987, 26, 1257.
[8] a) O. Wagner, M. Ehle, M. Regitz, Angew. Chem. 1989, 101, 227;
Angew. Chem. Int. Ed. Engl. 1989, 28, 225; b) H. Memmesheimer, M.
Regitz, Rev. Heteroat. Chem. 1994, 10, 61.
[17] T. H. Dunning, P. J. Hay, in Modern Quantum Chemistry (Ed.: H. F.
Schaefer), Plenum, New York, 1976, Vol. 3, chapter 1.
[18] a) A. D. Becke, J. Chem. Phys. 1993, 98, 1372 and 5648; b) C. Lee, W.
Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785.
[9] F. Mathey, Chem. Rev. 1990, 90, 997
Received: May 28, 1998 [F1175]
Chem. Eur. J. 1999, 5, No. 1
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999
0947-6539/99/0501-0279 $ 17.50+.50/0
279