Metal coordination to the formal P=N bond of an iminophosphorane and charge-density evidence against hypervalent phosphorus(V)

Article abstract of DOI:10.1002/chem.200400163

The iminophosphorane Ph₂P(CH₂Py)(NSiMe₃) (1) was treated with deprotonating alkali metal reagents to give [(Et ₂O)Li{Ph₂P(CHPy)(NSiMe₃)}] (2), [{Ph ₂P(CH₂Py)(NSi-Me₃)}Li{Ph₂P(CHPy) (NSiMe₃)}] (3) and [{Ph₂P(CH₂Py)(NSiMe ₃)}Na(Ph₂P(CHPy)(NSiMe₃)}] (4). We report their coordination behaviour in solid-state structures and NMR spectroscopic features in solution. Furthermore, we furnish experimental evidence against hypervalency of the phosphorus atom in iminophosphoranes from experimental charge-density studies and subsequent topological analysis. The topological properties, correlated to the results from NMR spectroscopic investigations, illustrate that the formal P=N double bond is better written as a polar P⁺-N ^- single bond. Additionally, the effects of metal coordination on the bonding parameters of the iminophosphorane and the related anion are discussed.

Full text of DOI:10.1002/chem.200400163

FULL PAPER
Metal Coordination to the Formal P=N Bond of an Iminophosphorane and
Charge-Density Evidence against Hypervalent Phosphorus(v)°
Nikolaus Kocher, Dirk Leusser, Alexander Murso, and Dietmar Stalke*[a]
Abstract:
The
iminophosphorane
troscopic features in solution. Further-
more, we furnish experimental evi-
dence against hypervalency of the
phosphorus atom in iminophosphor-
anes from experimental charge-density
studies and subsequent topological
analysis. The topological properties,
correlated to the results from NMR
spectroscopic investigations, illustrate
Ph2P(CH2Py)(NSiMe3) (1) was treated
with deprotonating alkali metal re-
agents to give [(Et2O)Li{Ph2P(CH-
Py)(NSiMe3)}] (2), [{Ph2P(CH2Py)(NSi-
Me3)}Li{Ph2P(CHPy)(NSiMe3)}]
and [{Ph2P(CH2Py)(NSiMe3)}-
Na{Ph2P(CHPy)(NSiMe3)}] (4). We
report their coordination behaviour in
solid-state structures and NMR spec-
that the formal P=N double bond is
ꢀ
better written as a polar P+
single
ꢀ
N
(3)
bond. Additionally, the effects of metal
coordination on the bonding parame-
ters of the iminophosphorane and the
related anion are discussed.
Keywords: alkali metals
¥ bond
topology ¥ hypervalent compounds ¥
N ligands ¥ P ligands
Introduction
in iminophosphoranes are mostly described as a resonance
hybrid between a double-bonded ylene R3P=NR and a dipo-
Since the landmark work of Wittig and Geissler[1] and Stau-
dinger and Meyer[2] on phosphonium ylides and iminophos-
phoranes, respectively, these isoelectronic analogues of
phosphane oxides have been well established species in or-
ganometallic and organic syntheses.[3] Particularly, the Wittig
reaction and its extensions play an outstanding role in the
(stereo)selective transformation of ketones and aldehydes
into olefins. The P=N bond in polyphosphazenes is known
to be thermally very robust and gives access to inert poly-
lar ylidic form R3P+ N R.
Calculations on main group
ꢀ
[3,8]
ꢀ
™hypervalent∫ compounds suggest that although the pres-
ence of d-orbitals is necessary to successfully model any of
the heavier main group elements, they do not play a signifi-
cant role in bonding. It has been proposed instead that neg-
ative hyperconjunction may be responsible for any p-charac-
[7,9]
ꢀ
ꢀ
ꢀ
ter in the P O, P C or P N bonds.
These results have
been substantiated by recent calculations dealing with the
Wittig-type reactivity of phosphorus ylides[10] and imino-
phosphoranes[11] and by the experimental determination of
the charge density in a phosphane ylide.[12]
ꢀ
meric materials, but the Si N bond in silylated iminophos-
phoranes of the general type R3P=NSiMe3 can easily be
cleaved in reactions with main group[4] or transition metal
halides[5] leading to phosphoraneiminato complexes contain-
ing the [R3PN]ꢀ building block. Recently, Stephan et al.
proved the high catalytic activity of these complexes to-
wards ethylene polymerization.[6] Consequently, P=E bond-
ing (E = C, N, O) is an issue of experimental and theoreti-
cal debate because of the chemical importance of ylides,
iminophosphoranes and phosphane oxides.[7] The P=E bonds
Phosphorus-based ligand systems with one or more donat-
ing atoms in the periphery are gaining increasing impor-
tance in catalysis and in the design of self-assembling li-
gands.[13] The incorporation of heteroaromatic substituents
at the phosphorus centre instead of commonly employed
phenyl groups alters and augments the reactivity and coordi-
nation capability of the ligand system and leads to the
design of multidentate Janus head ligands.[14,15] The concept
of side-arm donation, which has already proved to be useful
in various catalytic reactions involving chelating phosphanes
with selectivity for hard/soft coordination sites,[16] can also
be achieved by means of ring heteroatoms (Scheme 1).
Similar to ketones, sulfones and hydrazones,[17] P-alkyl-
substituted iminophosphoranes are moderately acidic and
can be deprotonated at the Ca-position by lithium organyls.
Several alkali metal complexes of a-deprotonated imino-
phosphoranes have been obtained and structurally charac-
terised. The cations are C,N-chelated by the deprotonated
[a] Dr. N. Kocher, Dr. D. Leusser, A. Murso, Prof. Dr. D. Stalke
Institut f¸r Anorganische Chemie der Universit‰t W¸rzburg
Am Hubland, 97074 W¸rzburg (Germany)
Fax: ( +49)931-888-4619
E-mail: dstalke@chemie.uni-wuerzburg.de
[°] This work was supported by the Deutsche Forschungsgemeinschaft
(grant no. Sta 334/8-3 and Graduiertenkolleg 690 Elektronendichte–
Theorie und Experiment) and Chemetall, Frankfurt/Main.
Supporting information for this article is available on the WWW
under http://www.chemeurj.org/ or from the author.
3622
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/chem.200400163
Chem. Eur. J. 2004, 10, 3622 3631

3622 3631
behaviour in solid-state struc-
tures and their NMR spectro-
scopic features in solution. Fur-
thermore, we furnish experi-
mental evidence against hyper-
valency of the phosphorus atom
Scheme 1. Reactions implying P=N bond cleavage.
in iminophosphoranes from
high-resolution X-ray data and
subsequent topological analysis
carbon and imino nitrogen atoms. The steric demand of the
substituents and the nature of the donor solvents determine
the degree of aggregation, which ranges from monomers to
tetramers.[8,18] Hence, the introduction of alkyl bridges be-
tween the phosphorus atom and the donating heterocycle in
the ligand periphery of iminophosphoranes provides not
only the opportunity of greater geometrical adaptability due
to a wider bite but also to generate anionic species. Delocal-
isation of the carbanionic charge to the donating atoms
would enable the ligand to respond to the various require-
ments of metals differing in radius and polarisability.
Recently it was shown that
of the multipole refinement. The topological properties are
correlated to the results from NMR spectroscopic investiga-
tions.
Results and Discussion
Syntheses of 1 4: N-(Trimethylsilyl)diphenyl-2-picolylimino-
phosphorane (1) is readily available by the Staudinger reac-
tion of diphenyl-2-picolylphosphane[20] with trimethylsilyl-
azide (Scheme 2). Deprotonation at the methylene bridge of
N-(trimethylsilyl)diisopropyl-2-
picolyliminophosphorane
iPr2P(CH2Py)(NSiMe3) (Py=2-
pyridyl) is a valuable building
block in the generation of 1,3-
dimetallacyclobutanes
[M{C(iPr2PNSiMe3)(Py)}]2 (M
= GeII, SnII, PbII) by single de-
protonation with lithium organ-
yls and subsequent reaction
with
the
corresponding
Group 14 chloride.[19] Here we
report alkali metal complexes
derived from N-(trimethylsilyl)-
diphenyl-2-picolyliminophos-
phorane (1), their coordination
Scheme 2. Synthesis of 1, deprotonation of 1 with MeLi to give 2 and preparation of 3 and 4.
1 with one equivalent of MeLi in diethyl ether at ꢀ788C
Abstract
in
German:
Das
Iminophosphoran
gives the lithium complex[(Et O)Li{Ph2P(CHPy)(NSiMe3)}]
Ph2P(CH2Py)(NSiMe3) (1), wurde mit Alkalimetallbasen
2
(2) in high yield. The reaction of only 0.5 equiv of MeLi
gives the complex[{Ph 2P(CH2Py)(NSiMe3)}Li{Ph2P(CH-
Py)(NSiMe3)}] (3), in which the neutral starting material 1
and the related anion both chelate the same lithium atom.
Formally, the diethyl ether donor molecule in 2 is replaced
by the iminophosphorane 1. To obtain insight into the reac-
tivity and coordination flexibility of the ligand, 1 was depro-
tonated with sodium amide in THF at room temperature.
As the neutral donor ligand already proved to be a good bi-
dentate donor in the lithiated complex 3, the reaction was
performed with only 0.5 equiv of NaNH2, so that 1 acted as
reagent and donor at the same time, akin to 3.
umgesetzt. Die Komplexe [(Et2O)Li{Ph2P(CHPy)(NSiMe3)}]
(2),
(3),
[{Ph2P(CH2Py)(NSiMe3)}Li{Ph2P(CHPy)(NSiMe3)}]
und [{Ph2P(CH2Py)(NSiMe3)}Na{Ph2P(CHPy)(N-
SiMe3)}] (4) wurden auf ihr Koordinationsverhalten im Fes-
tkˆrper untersucht. Die NMR-spektroskopischen Eigenschaf-
ten in Lˆsung wurden in Beziehung zum Koordinationsver-
halten gesetzt. Dar¸ber hinaus zeigt die experimentelle Elek-
tronendichtebestimmung und die anschlie˚ende topologische
Analyse, dass zur Beschreibung der Bindungssituation in
Iminophosphoranen keine Hypervalenz des Phosphoratoms
nˆtig ist. Die topologischen Eigenschaften des anionischen
Liganden in ‹bereinstimmung mit den NMR-spektroskopi-
schen Ergebnissen zeigen, dass die formale P=N Doppelbin-
ꢀ
dung besser als polare P+ N -Einfachbindung zu beschrei-
ꢀ
Structures: The solid-state structures of the iminophosphor-
ane 1 and the metalated derivatives 2 4 are shown in Fig-
ures 1 4, respectively. Selected bond lengths and angles are
listed in Table 1; crystallographic details are given in
Table 5.
ben ist. Weiterhin werden Effekte der Metallkoordination auf
die Bindungsparameter des Iminophosphorans und seines
Anions abgeleitet.
Chem. Eur. J. 2004, 10, 3622 3631
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3623

D. Stalke et al.
FULL PAPER
Table 1. Selected bond lengths [pm] and angles [8] of 1 4.
Anionic ligand
3, M=Li
Neutral iminophosphorane
3, M=Li
2, M=Li
159.04(2)
4, M=Na
157.79(14)
168.69(14)
173.43(17)
141.1(2)
1
4, M=Na
156.28(13)
172.04(14)
183.48(17)
150.6(2)
ꢀ
ꢀ
P1 N1
158.96(15)
P2 N3
154.13(12)
168.08(12)
182.57(14)
150.39(19)
133.17(18)
156.55(15)
ꢀ
ꢀ
N1 Si1
170.67(2)
172.57(2)
140.13(3)
137.07(3)
194.37(7)
196.30(7)
127.114(19)
130.469(14)
170.69(16)
172.95(19)
140.5(3)
137.5(2)
206.9(3)
209.1(4)
123.33(15)
130.67(9)
N3 Si2
172.17(15)
181.46(19)
149.8(3)
134.3(2)
216.5(3)
210.2(3)
119.06(14)
132.91(9)
ꢀ
ꢀ
P1 C1
P2 C22
ꢀ
ꢀ
C1 C2
C22 C23
ꢀ
ꢀ
C2 N2
137.6(2)
C23 N4
ꢀ
M N3
134.1(2)
ꢀ
M N1
233.51(15)
239.16(15)
125.42(13)
140.06(9)
237.09(14)
240.00(15)
117.95(11)
130.47(8)
ꢀ
ꢀ
M N2
M N4
P1-C1-C2
P2-C22-C23
111.98(9)
143.91(8)
P1-N1-Si1
P2-N3-Si2
ꢀ
ꢀ
P1 CPh
av 182.24
av 183.15
av 182.85
P2 CPh
av 181.67
av 181.91
av 181.63
As expected, the central phosphorus atom in
Ph2P(CH2Py)(NSiMe3) (1) is tetrahedrally coordinated by
nated atoms. As expected from
the coordination behaviour,[24]
a
ꢀ
three carbon atoms and one nitrogen atom. The P2 CPh dis-
tances (av 181.67 pm) are on average 1 pm shorter than the
ꢀ
six-membered metallacycle is
formed. The conformation of
this ring is a distorted boat ar-
rangement with the deprotonat-
ed carbon atom and the metal
atom in the bow and stern posi-
tions. They lie 29 and 14 pm, re-
spectively, out of the best plane
defined by N1-P1-C2-N2.
P2 C22 bond (182.57(14) pm) because of the smaller radius
of an sp2- compared with an sp3-hybridised carbon atom.
ꢀ
The P2 N3 bond length of 154.13(12) pm is in the range
normally quoted for a formal P=N double bond in imino-
phosphoranes (147 162 pm).[3,8,21] As in all N-silyl-substitut-
ꢀ
ed iminophosphoranes, the N3 Si2 bond (168.08(12) pm) is
Figure 2. Solid-state structure
of 2; for clarity only the hydro-
gen atom at the bridging
carbon atom is depicted.
ꢀ
considerably shorter than
a
formal Si N single bond
(174 pm).[22] In the solid-state, 1 adopts a transoid conforma-
tion in which the two nitrogen atoms N3 of the imino group
and N4 of the pyridyl ring point in opposite directions due
The sum of the bond angles
at the imino nitrogen atom N1
is very close to 3608 indicating
ꢀ
to repulsion of the N lone pairs. N¥¥¥H C hydrogen bonding
is not observed (Figure 1).
a
planar environment. The
same is valid for the pyridyl ring nitrogen atom N2. In con-
trast to the chemically related monomeric bis(iminophos-
phorane)methanides
[(Et2O)Li{HC(Cy2PNSiMe3)2}],[25]
[(THF)Li{HC(Ph2PNSiMe3)2}][26]
and
the
dimeric
[Li{HC(Ph2PNSiMe3)2}]2,[27] the lithium cation in 2 shows no
contact to the deprotonated carbon atom C1 (Li¥¥¥C1
322.5 pm). The bond angles at C1 are close to 1208 and thus
indicate sp2-hybridisation. In contrast to the parent imino-
phosphorane 1, N,N-chelation to the metal atom results in a
ꢀ
N1/N2 cisoid conformation. Remarkably, the Li N1 distance
ꢀ
in 2 (194.37(7) pm) is only slightly shorter than the Li N2
distance of 196.30(7) pm. Both contacts are at the short end
ꢀ
of the range reported for Li N distances in lithium
amides,[28] lithium iminophosphoranates[14a,29] and lithium
aminidinates.[30] Metal coordination to N1 elongates the ele-
Figure 1. Solid-state structure of 1; for clarity only the methylene hydro-
gen atoms are depicted.
ꢀ
ment nitrogen bonds. Although the P1 N1 bond length in 2
is lengthened by about 5 pm with respect to 1, it is still in
Alkali metal complexes tend to form oligomeric aggre-
gates depending on the steric demand of the substituents,
the radius and polarisability of the metal atom and the
nature and amount of donor solvent.[23] Nevertheless, the
above-mentioned reactions of the iminophosphorane 1 with
alkali metal deprotonation reagents yielded exclusively
monomeric complexes.
the range generally quoted for formal P=N double bonds in
iminophosphoranes.[3,21] Likewise, the N1 Si1 bond
ꢀ
(170.67(2) pm) is elongated by 2.7 pm in comparison to the
ꢀ
N Si bond in the iminophosphorane 1. Similar elongations
relative to the related starting materials[31,32] are observed in
[14a]
[(Et2O)Li{(o-C6H4)Ph2PNSiMe3}]2
and lithiated bis(imi-
nophosphorane)methanides.[26,27]
In [(Et2O)Li{Ph2P(CHPy)(NSiMe3)}] (2), the trigonal-
planar coordination sphere of the lithium cation is made up
of the imino nitrogen atom N1, the pyridyl ring nitrogen
atom N2 and the oxygen atom of the diethyl ether donor
molecule (Figure 2). The metal cation is displaced by only
22.6 pm from the best plane determined by the three coordi-
Figure 3
shows
the
solid-state
structure
of
(3).
[{Ph2P(CH2Py)(NSiMe3)}Li{Ph2P(CHPy)(NSiMe3)}]
One deprotonated equivalent of the starting material 1 con-
tributes the monoanionic [N-P-C(H)-Py]ꢀ chelating ligand
also present in 2, while the second equivalent acts as neutral
bidentate N,N’-donor base towards the lithium cation.
3624
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 3622 3631

Bonding in Iminophosphoranes
3622 3631
Figure 3. Solid-state structure of 3; for clarity only the hydrogen atoms at
the bridging carbon atoms are depicted.
Figure 4. Solid-state structure of 4; for clarity reasons only the hydrogen
atoms at the bridging carbon atoms are depicted.
In 3, the lithium cation is coordinated in a distorted tetra-
hedral fashion by the four nitrogen atoms (the N-Li-N
angles vary from 98.23(14) to 128.49(17)8). In the metalla-
spirocycle both six-membered LiN2C2P rings adopt a distort-
ed boat conformation. The lithium cation is displaced by
46 pm from the best plane of the anionic ligand, but only
12 pm from the plane of the neutral ligand. As expected, the
deprotonated sp2 carbon atom C1 is closer to the plane than
the methylene carbon atom C22 (displacements of 45 and
61 pm, respectively). The Li¥¥¥C1 distance in 3 is 324.3 pm,
that is, 2 pm longer than in 2. However, due to the more
pronounced boat conformation the lithium cation is 16 pm
closer to the deprotonated carbon atom C1 than to C22 of
the neutral ligand (Li¥¥¥C22 340.4 pm). The P1-C1-C2 angle
of 123.33(15)8 indicates sp2-hybridisation. In the neutral
donor ligand, the corresponding angle at C22 of 119.06(14)8
is only marginally smaller. The sum of the bond angles at
both imino nitrogen atoms N1 and N3 of close to 3608 indi-
cates a planar environment. For the pyridyl ring nitrogen
atoms N2 and N4, the sum of the angles shows slight dis-
placement of the metal atom from the plane of the heteroar-
As in 3 the metal atom is displaced further from the best
plane of the anion than from that of the neutral ligand (20
vs 17 pm), and the deprotonated carbon atom is closer to
that plane than the methylene carbon atom (34 vs 75 pm).
The imino nitrogen atoms N1 and N3 in 4 show trigonal-
planar coordination. The same is valid for the pyridyl ring
nitrogen atom N4 of the neutral ligand. However, the sum
of the angles at the pyridyl ring nitrogen atom N2 in the
anionic ligand of 343.48 indicates a displacement of the
sodium atom towards the p-electron density of the ring.
This more pronounced haptotropic shift of the softer metal
atom was observed earlier[36] and is emphasised by the grad-
ual increase in the ordering numbers of the alkali metals.[37]
ꢀ
The differences in the P N bond lengths in the anionic and
neutral ligands in 4 are only half as much as in the lithium
complex 3, and the elongation of the formal P=N double
bond compared to the starting material 1 is less pronounced.
ꢀ
ꢀ
The same is valid for the N Si distances. The Na N distan-
ces in 4 span the range of 233.51(15) to 240.00(15) pm and
are comparable to those of the structurally related mono-
meric [(THF)2Na{HC(Ph2PNSiMe3)2}][25] (241.6 pm) and di-
meric ([Na{HC(Ph2PNSiMe3)2}]2 (233.4 253.8 pm).[27] The
values correspond to those found in sodium amides rather
than those in complexes with coordinated amine donor
bases such as [(pmdeta)NaPh]2 (260.9 271.2 pm; pmdeta:
ꢀ
omatic ring (352.2 and 357.78, respectively). The P1 N1 dis-
tances in the anionic ligand of 2 (159.04(2) pm) and 3
(158.96(15) pm) are identical within the ESDs and about
5 pm longer than in the iminophosphorane
1
ꢀ
(154.13(12) pm). The P2 N3 bond length of 156.55(13) pm
MeN(CH2CH2NMe2)2).[38] However, the Na NPy distances
ꢀ
in the neutral donor ligand in 3 is half-way between them.
Interestingly, the N1 Si1 distance of 170.69(16) pm in the
are similar to those in, for example, [(C5H5N)3NaCp*][39]
(245.0 248.8 pm; Cp*=h5-C5Me5), [(C5H5N)Na{OtBu(Si-
Me2)NSiMe3}][40] (239.1 pm) and [(THF)3Na(PyCPh2)][36]
(241.4 pm). Figure 5 depicts a superposition plot of the core
structures of the 3 and 4.
ꢀ
anionic ligand of 3 is identical to that in 2. However, the
ꢀ
N2 Si2 distance (172.17(15) pm) is about 1.5 pm longer than
in 2 and even 4.1 pm longer than that in the iminophosphor-
ꢀ
ꢀ
ane 1. Thus, both P N and N Si bonds in the anionic and
neutral ligand are appreciably elongated on metal coordina-
tion at the imino nitrogen atom. As expected from the
higher coordination number of the lithium cation and the
ꢀ
In all compounds 1 4 the P CPh distances are very similar
and do not differ significantly from that of the iminophos-
phorane 1. In the iminophosphorane 1 and the neutral li-
ꢀ
ꢀ
ꢀ
greater steric bulk of the two ligands, both Li N bonds in 3
gands in 3 and 4, the P2 C22 (av 182.5 pm) and C22 C23
ꢀ
are longer than in 2. The Li Nimino bond distances in 3 vary
ꢀ
remarkably. Whereas the Li N1 distance to the anion is
only 206.9(3) pm, the Li N3 distance to the donor ligand is
9.6 pm longer. The imino N atom of the anion seems much
ꢀ
ꢀ
more attractive to the lithium cation. The Li NPy distances
in 3 differ only by about 1 pm and are in the normal
35]
range.[33
The
sodium
complex[{Ph
2P(CH2Py)(NSiMe3)}-
Na{Ph2P(CHPy)(NSiMe3)}] (4) crystallises as yellow blocks
from saturated THF solution. The solid-state structure of 4
is depicted in Figure 4. Compounds 3 and 4 are isomorphous
but not isostructural.
Figure 5. Superposition plot of the core structures of the lithium complex
3 (narrow lines) and the isomorphous sodium complex 4 (wide lines).
Chem. Eur. J. 2004, 10, 3622 3631
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3625

D. Stalke et al.
FULL PAPER
ꢀ
ꢀ
(av 150.3 pm) bond lengths are almost equal and are invari-
while polarisation of the P2 N3 bond gives a weaker M N3
bond.
ꢀ
ant to metal coordination. They are standard P C (185 pm)
[22]
ꢀ
ꢀ
and C C single bonds. However, in the anionic ligands of
This trend is also reflected in the N Si bond lengths. In 1,
the shortest N Si bond (168.08(12) pm) is present. In the
anionic ligands of the lithiated complexes 2 and 3, these
bonds are on average 2.6 pm longer than in the iminophos-
phorane 1. In 4, it is very similar to the distance in 1, that is,
the softer sodium atom is less polarising towards the anionic
ꢀ
ꢀ
2 4, the average P1 C1 bonds of 173.0 pm are about 10 pm
shorter than the corresponding bond in the neutral mole-
cules. In phosphaalkenes, the formal P=C double-bond
lengths vary from 161 to 171 pm, and the average value is
167 pm.[41] Ylidic P C bond lengths lie between 163 and
ꢀ
173 pm,[42] and Gilheany quotes an average value of
ligand. In comparison with standard N Si single bonds
ꢀ
169 pm.[7] Thus, the P1 C1 bond lengths in the anionic li-
(174 pm)[22] they all are inherently short. Interestingly, the
ꢀ
ꢀ
gands in this paper suggest partial double-bond character, as
N3 Si2 distances in the neutral ligands of 3 and 4 are about
4 pm longer than in the iminophosphorane 1 and are there-
depicted in resonance structure b of Scheme 3. At the same
ꢀ
fore almost in the range of standard N Si single bonds. In
the iminophosphorane 1, the negatively polarised imino ni-
trogen atom N3 is bonded to the electropositive phosphorus
ꢀ
and silicon atoms. These interactions result in a short N3
Si2 bond in 1. On metal coordination at N3, the negatively
polarised imino nitrogen atom has a further interaction with
ꢀ
the cationic centre. Thus, the N3 Si2 distances in 3 and 4
ꢀ
become longer. In the anionic ligands of 2 4 the N1 Si1
bond lengths are half-way between that in 1 and those of
the neutral ligands in 3 and 4. The negatively polarised
imino nitrogen atom N1 interacts with the cation, the elec-
tropositive silicon atom and a less electropositive phospho-
rus centre, due to the neighbouring negatively polarised de-
ꢀ
protonated carbon atom C1. Thus, the N Si distances are
longer than in iminophosphorane 1, but shorter than in the
neutral ligands of 3 and 4.
Scheme 3. Resonance forms of [M{Ph2P(CHPy)(NSiMe3)}]: a indicates a
carbanionic ylidic contribution, b shows the amidic ylenic resonance
structure, c emphasises the amidic olefinic resonance form and d visualis-
es the delocalisation of the negative charge.
In conclusion, all the geometrical features of the anionic
ligands in the metalated complexes 2 4 suggest canonical
formulas as depicted in Scheme 3. However, all of them re-
quire valence expansion at the phosphorus atom, at variance
with the eight-electron rule, but chemical reactivity supports
neither P=N nor P=C double bonds, because both are easily
cleaved in various reactions.[14]
ꢀ
time, the C1 C2 bonds (av 140.6 pm) are shortened by ap-
proximately 10 pm in the anionic versus the neutral ligands,
and this reflects a partial double-bond character, as depicted
in c of Scheme 3.
ꢀ
Furthermore, the fact that the Cipso NPy bonds of the
anionic ligands (av 137.4 pm) are on average 3.5 pm longer
than the corresponding bonds in the neutral ligands indi-
cates a perturbation of the aromatic ring system. This might
be interpreted as the result of transfer of the negative
charge from the deprotonated carbon atom to the pyridyl
ring nitrogen atom.
NMR spectroscopic investigations: To elucidate the charge
distribution in 1 4, we performed several NMR experiments
in solution. The chemical shifts determined for the 31P and
15N nuclei are summarised in Table 2.
Table 2. 31P and 15N NMR shifts [ppm] of 1 4.
ꢀ
The P N bond lengths in all isolated compounds are short
Anionic ligands
Nimino
Neutral ligands
Nimino
and range from 154.13(12) pm in 1 to 159.04(2) pm in 2.
Even in the neutral donor ligands of 3 and 4 they are about
2.3 pm longer than in iminophosphorane 1. In accordance
with the harder Lewis-acidic character of the lithium cation
P
NPy
P
NPy
1
2
3
4
ꢀ0.32
ꢀ343
ꢀ61
18.03
15.36
11.30
ꢀ331
ꢀ334
ꢀ345
ꢀ145
ꢀ139
ꢀ137
8.46
ꢀ0.60
ꢀ339
ꢀ332
ꢀ68
ꢀ63
ꢀ
in comparison to sodium, the P1 N1 bonds in the lithium
complexes 2 (159.04(2) pm) and 3 (158.96(15) pm) are
longer than in the sodium derivative 4 (157.79(14) pm).
Metal coordination at the imino nitrogen atom of the neu-
tral ligand shifts electron density from the P=N bonding
region to the metal, lengthening the P=N bond. Metal coor-
dination at N1 together with deprotonation at the methylene
bridge next to the phosphorus atom P1 leads to a further
elongation of the P=N bond, as the additional negative
carbon atom competes for the charge density of the electro-
positive phosphorus atom. Charge accumulation at N1 in
Deprotonation at the Ca-atom in the iminophosphorane
Ph2P(CH2Py)(NSiMe3) (1) results in 31P NMR downfield
shifts relative to 1 and the neutral ligands in 3 and 4. This
shift seems to be counterintuitive to the idea of P=C bond
generation in the anions.
1
An interesting result was obtained in the H,15N HMBC
NMR experiments. In the starting material 1 and in the neu-
tral ligands of 3 and 4, the pyridyl nitrogen atoms resonate
at similar frequencies (d=ꢀ61 to ꢀ68). The observed small
ꢀ
the anionic ligands furnishes a strong M N1 interaction,
3626
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 3622 3631

Bonding in Iminophosphoranes
3622 3631
differences arise from metal coordination at the ring nitro-
gen atom. Remarkably, they reflect the polarising abilities
of the metal, and hence the upfield shift due to sodium coor-
dination is less pronounced than the changes induced by the
lithium cation. However, the 15N NMR signals for the pyrid-
yl ring nitrogen atoms in the anionic ligands are considera-
bly shifted to higher fields (d=ꢀ137 to ꢀ145) with respect
to 1 (d=ꢀ61) and to the neutral ligands in 3 and 4. These
NMR spectroscopic shifts can be explained by a considera-
ble charge transfer from the
presented in terms of the charge density 1(r), the Laplacian
2
5 1(r) and the charges obtained from the integration over
the atomic basins according to Bader×s theory of atoms in
molecules (AIM).[44] All bond critical points (BCPs) at the
ꢀ
P E bonds are displaced towards the electropositive phos-
phorus atom. A comparison of the densities 1(rBCP), the al-
gebraic sum of the eigenvalues li of the Hessian matrix
2
5 1(rBCP) of these bonds and the integrated charges are pre-
sented in Table 4.
carbanionic atom C1 to the
heteroaromatic ring and accu-
mulation at the pyridyl nitro-
[a]
ꢀ
ꢀ
Table 4. Topology of the P N and P C bonds in 2.
2
ꢀ
A B
ꢀ
ꢀ
ꢀ
d(A B)
d(A BCP)
d(BCP B)
1(rBCP
)
5 1(rBCP
)
Charge at B
ꢀ
P N1
ꢀ
P C1
1.5903
1.7252
1.8294
1.8147
0.6566
0.7753
0.8047
0.7784
0.9337
0.9499
1.0248
1.0364
1.508(10)
1.336(8)
1.134(6)
1.151(8)
5.874(28)
ꢀ8.325(22)
ꢀ6.402(15)
ꢀ6.145(16)
ꢀ1.98
ꢀ0.54
ꢀ0.31
ꢀ0.34
gen atom. Additionally, the
almost invariant 15N NMR
ꢀ
P C7
shift of the imino nitrogen
atom in all compounds indi-
cates that the charge density
at this nitrogen atom is less af-
ꢀ
P C13
ꢀ
ꢀ
ꢀ
[a] d(A B): distance [ä] between atoms A and B along the bond path; d(A BCP), d(BCP B): distance be-
tween the BCP and atoms A and B, respectively; 1(rBCP): charge density [eäꢀ3] at the BCP; 5 1(rBCP): Lapla-
2
cian of 1(r) at the BCP [eäꢀ5]; Charges of B from integration over atomic basins in electrons, charge of the
phosphorus atom A: +2.20.
fected by deprotonation.
Charge transfer to the pyr-
idyl substituent in the anionic
ligands of 2 4 is also evident from the 1H and 13C NMR
spectroscopic experiments. In Table 3, these chemical shifts
are compared with those of the starting material 1 and the
neutral ligands in 3 and 4. Deprotonation at Ca in imino-
phosphorane 1 results in a upfield shift for the pyridyl hy-
drogen atoms relative to the corresponding atoms in the
neutral ligands of 3 and 4 and the parent iminophosphorane
1. The charge density is accumulated in the pyridyl substitu-
ent. In the 13C NMR spectrum, the deprotonated C1 atoms
in the anionic ligands resonate at lower fields compared to
the neutral ligands. Additionally, the carbon atom atoms ad-
jacent to the ring nitrogen atoms are shifted downfield.
However, in quintessence neither the geometrical parame-
ters from the structure determination nor the NMR spectro-
scopic data provide a conclusive criterion to decide which
resonance form in Scheme 3, if any, contributes most to the
bonding in alkali metal complexes 1 4.
The properties at the BCPs of the aromatic C C bonds in
the phenyl rings and those in the SiMe3 group are in the ex-
ꢀ
ꢀ
pected range. The two topologically analysed P CPh bonds
in 2 are comparable to those in the triphenylphosphane
ligand of a transition metal complex.[45] In contrast to the
ꢀ
P CPh bonds in triphenylphosphonium benzylide Ph3P=
ꢀ
C(H)Ph, no variance in the topology between the two P CPh
bonds in 2 could be observed.[12] Compared to the latter, the
ꢀ
P C1 bond displays higher charge density and a more nega-
tive Laplacian at the BCP. An inspection of the Laplacian
along the P C bond paths (Figure 6) in 2 reveals an almost
equal charge distribution in the phosphorus basin, while the
main difference concerning the Laplacian in the basin of the
carbon atoms is related to the shorter distance between C1
ꢀ
ꢀ
and BCPPꢀC1. This rather short P C bond results from dis-
tinct electrostatic interactions between the negatively polar-
ised deprotonated C1 and the electropositive phosphorus
atom, as reflected in the charges of ꢀ0.52e for C1 and
+2.20e for P from integration over the atomic basins.[46]
The properties of the exocyclic C1 C2 bond, which are
almost the same as those of an aromatic C C bond, as well
as the aromatic character of the C2 N2 bond of the pyridine
ring fuels the idea of delocalisation of the negative charge in
the C1-C2-N2 residue. This exocyclic delocalisation slightly
Charge-density distribution in [(Et2O)Li{Ph2P(CHPy)(N-
SiMe3)}] (2): To describe the charge-density distribution in
[(Et2O)Li{Ph2P(CHPy)(NSiMe3)}] (2), we performed a mul-
tipole refinement based on the formalism of Hansen and
Coppens[43] with 100 K high-resolution data up to (sinq/
l)max =1.145 äꢀ1. The results of the topological analysis are
ꢀ
ꢀ
ꢀ
1
Table 3. Selected H and 13C NMR shifts [ppm] of 1 4.
1H
13C
1
3
4
5
6
1
2
3
4
5
6
1
2
3.56
3.58
4.11
3.78
3.39
3.85
7.18
6.42
6.60
6.85
6.18
7.15
7.07
6.71
7.00
7.00
6.64
7.44
6.55
5.77
5.95
6.42
5.51
6.98
8.28
7.12
7.33
8.46
7.15
8.31
43.2
55.8
59.0
39.7
58.2
42.5
152.3
167.8
166.3
150.7
168.0
155.3
125.6
118.9
118.4
126.6
117.9
125.8
129.5
134.8
134.2
128.3
133.5
131.1
121.6
106.9
103.6
121.5
103.9
122.1
149.6
147.5
148.0
152.2
147.8
149.8
3a[a]
3b[a]
4a[a]
4b[a]
[a] a: anionic ligand, b: neutral ligand.
Chem. Eur. J. 2004, 10, 3622 3631 www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3627

D. Stalke et al.
FULL PAPER
ꢀ
ꢀ
Figure 6. a) Laplacian along the bond paths of the P C and P N bonds; b) Laplacian in the P1-C2-Li plane (blue lines indicate charge concentrations,
ꢀ
ꢀ
red lines indicate charge depletions); c) Laplacian along the bond paths of the Li N and Li O bonds. The zero value of the x axis indicates the position
of the BCP.
perturbs the aromaticity of the heteroaromatic ring. The
bond lengths and electron densities in the pyridyl substitu-
ent differ from those in undisturbed rings.[47]
2
ꢀ
The Laplacian distribution 5 1(r) along the P N1 bond is
ꢀ
completely different to that of the P Cipso bonds. The Lapla-
cian at the BCP is positive and charge density is exclusively
concentrated in the nitrogen basin (Figure 6a). This indi-
cates a severe contribution of electrostatic interaction to the
bonding energy, further substantiated by the integration of
the atomic basins.[46] The two basins related to the nitrogen
atoms give distinct negative values. The imino nitrogen
atom N1 is bonded to three electropositive neighbours, and
therefore the charge of ꢀ1.91e is higher than that of ꢀ1.11e
for the ring nitrogen atom N2. The charge of the deproto-
nated carbon atom C1 is about 0.2e higher than those of the
two phosphorus bonded ipso carbon atoms (C7: ꢀ0.30e,
ꢀ
C13: ꢀ0.34e). All these findings support an ylidic P+
si-
ꢀ
C
ꢀ
multaneous with a P+
bond, not yet present in the reso-
ꢀ
N
nance forms of Scheme 3. The determination of the (3;ꢀ3)
2
critical points in the spatial distribution of ꢀ5 1(r) and
quantification of the valence shell charge concentrations
(VSCCs) around the electronegative atoms (N1, N2, O1,
C1) confirms this view. Isosurface presentations around N1
and a contour plot in the plane containing the (3;ꢀ3) critical
points are presented in Figure 7a. They indicate sp3-hybridi-
sation at N1 with two lone pairs, both oriented towards the
Li cation. The same is valid for O1 (Figure 7b). The angle
between the lone pairs LP1-N1-LP2 of approximately 718 is
much more acute than the expected 1098, but can be ration-
alised by the closer approach of the critical points by simul-
taneous interaction of both VSCCs with the lithium cation.
Clearly, the lone pairs at N1 and O1 act as a bifurcated
donor to the electropositive metal acceptor. We found the
same arrangement at S-bonded sp3-hybridised nitrogen
atoms in an intramolecular hydrogen bond.[48]
2
Figure 7. Isosurface maps at constant 5 1(r) values indicating bonding
and nonbonding charge concentrations around the displayed atoms, and
contour plots in a plane with the two lone pairs oriented towards Li1 for
N1 (a) and O1 (b) in 2; a) N1, 5 1(r)=ꢀ48 eäꢀ5 (bonding VSCCs) and
2
ꢀ41 eäꢀ5 (nonbonding VSCCs); b) O1, 5 1(r)=ꢀ125 eäꢀ5 (bonding
2
VSCCs) and ꢀ105 eäꢀ5 (nonbonding VSCCs). Blue contours indicate
charge concentration, and red contours charge depletion.
double bond must be written as polar P+
N C
and P+
ꢀ
ꢀ
ꢀ
ꢀ
single bonds augmented by electrostatic contributions. As
predicted from calculations, a hypervalent central phospho-
rus atom is not required to describe the bonding. It is much
more appropriate to assign charges in the resonance formu-
la, even for the starting material Ph2P(CH2Py)(NSiMe3) (1).
The almost invariant imino nitrogen signals in the 15N NMR
spectra of 1 4 further supplies credibility to these canonical
Conclusion
forms. The iminophosphorane 1 is better been written as a
ꢀ
zwitterionic phosphonium amide Ph2(PyCH2)P+ NSiMe3.
ꢀ
The experimental charge-density distribution in [(Et2O)-
Li{Ph2P(CHPy)(NSiMe3)}] (2) clearly proves that the formal
P=N imino double bond and the potential ylenic P=C
This polar single bond corresponds best with the reactivi-
ty: metal organyls in polar solvents can more easily cleave
3628
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 3622 3631

Bonding in Iminophosphoranes
3622 3631
and storage at ꢀ168C gave 1 as colourless plates. M.p. (DTA): 448C;
31P NMR ([D8]toluene): d=ꢀ0.32 (s); 29Si NMR ([D8]toluene): d=
ꢀ10.61 (s); 1H,15N HMBC ([D8]toluene): d=ꢀ343 (s, NSiMe3), ꢀ61 (s,
2
pyN); 1H NMR ([D8]toluene): d=0.23 (s, 9H, SiMe3), 3.56 (d, JPH
=
14.1 Hz, 2H, H1), 6.55 (dddd, 1H, H5), 7.07 (m, 1H, H4), 7.18 (d, 1H,
H3), 8.28 (d, 1H, H6), 7.00 7.05 (m, 6H, m-, p-PhH), 7.65 7.81 (m, 4H,
3
o-PhH); 13C NMR ([D8]toluene): d=4.8 (d, JSiC =3.0 Hz, SiMe3), 43.2 (d,
C1), 121.6 (d, C5), 125.6 (d, C3), 129.5 (d, C4), 149.6 (d, C6), 152.3 (d,
C2), 129.2 (m-PhC), 131.0 (p-PhC), 131.8 (o-PhC), 134.9 (ipso-PhC); ele-
mental analysis (%) calcd for C21H25N2PSi: C 69.44, H 7.68, N 7.36;
found: C 69.24, H 7.72, N 7.43.
this bond rather than the wrongly assigned P=N double
bond. Therefore, deimination or the retro-Staudinger reac-
tion of iminophosphoranes seems an unorthodoxbut suita-
ble synthetic route to phosphanes.
The isolation of the complexes [{Ph2P(CH2Py)(NSi-
Me3)}M{Ph2P(CHPy)(NSiMe3)}] (M=Li: 3, Na: 4), in which
both anionic and neutral iminophosphorane are bound to
the cation, allowed a detailed discussion of the geometrical
parameters in the ligands and we were able to deduce the
effects of metal coordination on the structural parameters of
the neutral iminophosphorane and its anionic derivative.
[(Et2O)Li{Ph2P(CHPy)(NSiMe3)}] (2): Compound 1 (0.50 g, 1.37 mmol)
was dissolved in Et2O (10 mL) and cooled to ꢀ788C. To this solution
MeLi (0.94 mL, 1.6m in Et2O, 1.51 mmol) was added dropwise. After
warming to RT and stirring the yellow solution for 2 h the volume of the
solvent was reduced. After 48 h crystalline 2 (0.48 g, 1.08 mmol, 79%)
was isolated. M.p. (DTA): 1048C; 31P NMR ([D8]toluene): d=18.03 (s);
7Li NMR ([D8]toluene): d=1.50 (s); 29Si NMR ([D8]toluene): d=ꢀ8.63
(s); 1H,15N HMBC NMR ([D8]toluene): d=ꢀ331 (NSiMe3), ꢀ145 (pyN);
1H NMR ([D8]toluene): d=0.11 (s, 9H, SiMe3), 3.58 (d, 2JPH =18.2 Hz,
1H, H1), 5.77 (dd, 1H, H5), 6.42 (d, 1H, H3), 6.71 (dd, 1H, H4), 7.12 (d,
1H, H6), 7.01 7.06 (m, 6H, m-,p-PhH), 7.79 7.82 (m, 4H, o-PhH), 3.08
(q, 4H, OCH2CH3), 0.93 (t, 6H, OCH2CH3); 13C NMR ([D8]toluene): d=
4.9 (d, 3JSiC =3.8 Hz, SiMe3), 55.8 (d, C1), 106.9 (s, C5), 118.9 (d, C3),
134.8 (d, C4), 147.5 (s, C6), 167.8 (d, C2), 128.8 (m-PhC), 130.3 (p-PhC),
132.5 (o-PhC), 137.4 (ipso-PhC); elemental analysis (%) calcd for
C25H34LiN2OPSi: C 67.54, H 7.71, N 6.30 found: C 67.42, H 7.72, N 6.38.
ꢀ
The electrostatic contributions to the P N bonding and the
ꢀ
negatively charged imino nitrogen atom result in short P N
ꢀ
and N Si distances in iminophosphorane 1. The negative
charge at the imino nitrogen atom in 1 is stabilised by the
positive phosphorus centre and the electropositive silicon
[{Ph2P(CH2Py)NSiMe3}Li{Ph2P(CHPy)NSiMe3}] (3): Compound
1
ꢀ
ꢀ
atom, and this results in short P N and N Si contacts. The
observed elongations of these bonds in the neutral metal-co-
ordinating ligands are a result of polarisation of the negative
(0.50 g, 1.37 mmol) was dissolved in Et2O (30 mL) and cooled to ꢀ788C.
To this solution MeLi (0.43 mL, 1.6m in Et2O, 0.69 mmol) was added
dropwise. After warming to RT and stirring for 8 h the clear yellow solu-
tion was allowed to stand at RT for 3 d to yield 3 (0.80 g, 1.08 mmol,
79%) as yellow blocks. M.p. (DTA): 778C (decomp); 31P NMR ([D6]ben-
zene): d=8.46 (brs, P*), 15.36 (s, P); 7Li NMR ([D6]benzene): d=1.70
(s); 1H, 29Si HMBC NMR ([D6]benzene): d=ꢀ9.5 (s, Si), ꢀ1.82 (s, Si*);
1 H,15N HMBC NMR ([D6]benzene): d=ꢀ339 (Me3SiN*), ꢀ334
ꢀ
charge by the cations. The elongation of the P N distances
found in the anionic ligands are explained by the interaction
of the positively charged phosphorus centre with the nega-
tively polarised Ca atom, which weakens the electrostatic
1
(Me3SiN), ꢀ139 (PyN), ꢀ68 (PyN*); H NMR ([D6]benzene): d=0.39 (s,
ꢀ
contribution to the P N bonds.
9H, SiMe3*), 0.50 (s, 9H, SiMe3), 3.78 (d, 2JPH =13.9 Hz, 2H, H1*), 4.11
(d, 2JPH =23.6 Hz, 1H, H1), 5.95 (dd, 1H, H5), 6.42 (dd, 1H, H5*), 6.60
(d, 1H, H3), 6.85 (d, 2H, H3*), 7.00 (m, 2H, H4, H4*), 7.33 (d, 1H, H6),
8.46 (d, 1H, H6*), 7.25 7.30 (m, 6H, m-, p-PhH), 7.21 7.24 (m, 6H, m-,
p-PhH*), 7.52 7.70 (m, 4H, o-PhH*), 8.08 8.25 (m, 4H, o-PhH);
Experimental Section
3
13C NMR ([D6]benzene): d=2.9 (d, 3JSiC =4.5 Hz, SiMe3*), 3.8 (d, JSiC
=
4.5 Hz, SiMe3), 39.7 (d, C1*), 59.0 (d, C1), 103.6 (s, C5), 118.4 (d, C3),
121.5 (d, C5*), 126.6 (s, C3*), 128.3 (d, C4*), 134.2 (d, C4), 148.0 (s, C6),
150.7 (s, ipso-PyC*), 152.2 (s, C6*), 166.3 (s, ipso-PyC), 128.3 (m-PhC),
128.5 (m-PhC*), 129.8 (p-PhC), 130.7 (p-PhC*), 131.4 (o-PhC*), 132.8 (o-
PhC), 137.7 (ipso-PhC), 138.5 (ipso-PhC*); elemental analysis (%) calcd
for C42H49LiN4P2Si2: C 68.64, H 6.72, N 7.62 found: C 69.01, H 6.52, N
7.73.
All reactions were performed under an inert atmosphere of dry N2 with
Schlenk techniques or in an argon glove box. All solvents were dried
over Na/K alloy and distilled prior to use. NMR spectra were recorded at
room temperature on a Bruker DRX 300 spectrometer at 300.1 (1H),
75.5 (13C), 121.5 (31P), 59.6 (29Si), 155.5 (7Li), 79.4 (23Na) and 30.4 MHz
(1H,15N HMBC). Chemical shifts d are relative to the solvent for 1H, 13C
and 1H,15N HMBC NMR, to H3PO4 (85%) for 31P, to external saturated
7
LiCl solution for Li, to 0.1m NaCl in D2O for 23Na and to SiMe4 for 29Si
[{Ph2P(CH2Py)NSiMe3}Na{Ph2P(CHPy)NSiMe3}] (4): A suspension of
NaNH2 (0.11 g, 2.75 mmol) in THF (5 mL) was added to a solution of 1
(2.00 g, 5.49 mmol) in THF (40 mL) at RT. After 3 d, the yellow reaction
mixture was filtered and the volume of the solution was reduced by evap-
oration. Storage of the clear yellow solution at RT for several days gave
NMR. Elemental analyses were performed by the Microanalytisches
Labor der Universit‰t W¸rzburg.
The NMR shifts were assigned according to the following scheme:
The heteronuclei are assigned an asterisk in the neutral ligands of com-
pounds 3 and 4.
4
(3.05 g, 4.06 mmol, 74%) as yellow blocks. M.p. (DTA): 338C
(decomp); 31P NMR ([D8]THF): d=ꢀ0.6 (brs, P*), 11.3 (brs, P); 23Na
NMR ([D8]THF): d=5.02 (brs); 29Si NMR ([D8]THF): d=ꢀ13.9 (d,
2JSiP =8.3 Hz, Si), ꢀ12.8 (d, 2JSiP =23.3 Hz, Si*); 1H,15N HMBC NMR
([D8]THF): d=ꢀ345 (Me3SiN*), ꢀ332 (Me3SiN), ꢀ137 (PyN), ꢀ63
(PyN*); 1H NMR ([D8]THF): d=ꢀ0.14 (s, 9H, SiMe3*), ꢀ0.20 (s, 9H,
SiMe3), 3.39 (d, 2JPH =22.3 Hz, 1H, H1), 3.85 (d, 2JPH =14.1 Hz, 2H,
H1*), 5.51 (dd, 1H, H5), 6.18 (d, 1H, H3), 6.64 (dd, 1H, H4), 6.98 (dd,
1H, H5*), 7.14 7.17 (m, 2H, H6, H3*), 7.18 7.25 (m, 8H, m-, p-PhH, p-
PhH*), 7.44 (dd, 1H, H4*), 7.69 7.77 (m, 8H, o-PhH, o-PhH*), 7.31 7.38
(m, 4H, m-PhH*), 8.31 (d, 1H, H6*); 13C NMR ([D8]THF): d=3.87 (d,
3JSiC =4.5 Hz, SiMe3), 3.69 (d, 3JSiC =3.0 Hz, SiMe3*), 42.5 (d, C1*), 58.2
(d, C1), 103.9 (s, C5), 117.9 (d, C3), 122.1 (d, C5*), 125.8 (d, C3*), 131.1
(d, C4*), 133.5 (d, C4), 147.8 (s, C6), 149.8 (d, C6*), 155.3 (d, ipso-
PyC*),168.0 (d, ipso-PyC), 128.2 (m-, p-PhC), 128.8 (m-PhC*), 129.5 (p-
PhC*), 132.0 and 132.8 (o-PhC, o-PhC*), 137.3 (ipso-PhC*), 141.3 (ipso-
Ph2P(CH2Py)(NSiMe3), (1): N3SiMe3 (0.46 g, 3.97 mmol) was added to
Ph2PCH2Py (1.00 g, 3.61 mmol). The reaction mixture was heated under
refluxfor 3 h. Evaporation of the excess of N 3SiMe3 and distillation of
the crude product under vacuum gave pure Ph2P(CH2Py)(NSiMe3) (1;
1.29 g, 3.53 mmol, 98%) as a colourless oil. Addition of hexane to the oil
Chem. Eur. J. 2004, 10, 3622 3631
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3629

D. Stalke et al.
FULL PAPER
PhC); elemental analysis (%) calcd for C42H49N4P2Si2: C 67.17, H 6.58, N
7.46 found: C 66.36, H 6.70, N 7.21.
Acknowledgement
This work was supported by the Deutsche Forschungsgemeinschaft
(grant no. Sta 334/8-3 and Graduiertenkolleg 690 Elektronendichte–The-
orie und Experiment) and Chemetall, Frankfurt/Main.
Structure determination of 1 4: Crystallographic data for 1 4 are listed
in Table 5. All data were measured at low temperature[49] with graphite-
monochromated MoKa radiation (l=71.073 pm) on a Bruker D8 goniom-
eter platform, equipped with a Smart ApexCCD detector. Cell parame-
ters were determined and refined using the SMART software.[50] Series
of w scans were performed at several f settings. Raw frame data were in-
tegrated using the SAINT program.[51] The structures were solved using
direct methods and refined by full-matrixleast-squares techniques on F2
using SHELXTL.[52] Data of 1 and 2 were subjected to empirical correc-
tion for absorption with SADABS2.[53] The hydrogen atoms H1 at C1 in
2, H1 at C1 and H22A, H22B at C22 in 3 and 4 were taken from the dif-
ference Fourier map and refined freely. All other hydrogen atoms were
refined using a riding model. The Uiso values for the hydrogen atoms of a
CH3 group were set to 150%, and those of all other hydrogen atoms to
120%, of the Ueq values of the corresponding C atoms. All non-hydrogen
atoms were refined anisotropically. The anisotropic displacement param-
eters (ADPs) in the Supporting Information were plotted at the 50%
probability level.
[1] G. Wittig, G. Geissler, Justus Liebigs Ann. Chem. 1953, 580, 44.
[2] H. Staudinger, J. Meyer, Helv. Chim. Acta 1919, 2, 635.
[3] A. W. Johnson, Ylides and Imines of Phosphorus, Wiley, New York,
1993.
[4] Review: K. Dehnicke, F. Weller, Coord. Chem. Rev. 1997, 158, 103.
[5] Review: K. Dehnicke, M. Krieger, W. Massa, Coord. Chem. Rev.
1999, 182, 19.
[6] a) R. E. v. H. Spence, S. J. Brown, R. P. Wurz, D. Jeremic, D. W. Ste-
phan, PCT Int. Appl. WO 2000-CA978 20000824, CA 99-2282070D,
2001; b) W. Stephan, J. C. Stewart, F. Guerin, S. Courtenay, J. Kick-
ham, E. Hollink, C. Beddie, A. Hoskin, T. Graham, P. Wie,
R. E. v. H. Spence, W. Xu, L. Koch, X. Gao, D. G. Harrison, Orga-
nometallics 2003, 22, 1937.
[7] Review: D. G. Gilheany, Chem. Rev. 1994, 94, 1339.
[8] a) Review: A. Steiner, S. Zacchini, P. I. Richards, Coord. Chem. Rev.
2002, 227, 193; b) J. F. Bickley, M. C. Copsey, J. C. Jeffery, A. P.
Leedham, C. A. Russell, D. Stalke, A. Steiner, T. Stey, S. Zacchini, J.
Chem. Soc. Dalton Trans. 2004, 989.
CCDC-231738 231741 contain the supplementary crystallographic data
for this paper. These data can be obtained free of charge via
www.ccdc.cam.ac.uk/conts/retrieving.html (or from the Cambridge Crys-
tallographic Data Centre, 12 Union Road, Cambridge CB21EZ, UK;
fax: (+44)1223-336-033; or deposit@ccdc.cam.uk).
[9] a) W. Kutzelnigg, Angew. Chem. 1984, 96, 262; Angew. Chem. Int.
Ed. Engl. 1984, 23, 272; b) E. A. Reed, P. v R. Schleyer, J. Am.
Chem. Soc. 1990, 112, 1434; c) E. J. Magnusson, J. Am. Chem. Soc.
1993, 115, 1051; d) A. Dobado, H. Martinez-GarcÌa, J. M. Molina,
M. R. Sundberg, J. Am. Chem.
Table 5. Crystallographic data for compounds 1 4.
Soc. 1998, 120, 8461; e) D. B.
Chesnut, J. Phys. Chem. A 2003,
107, 4307.
1
2[a]
3
4
formula
C21H25N2PSi
364.49
231738
173(2)
triclinic
C25H34LiN2OPSi C42H49LiN4P2Si2 C42H49N4NaP2Si2
[10] a) T. Naito, S. Nagase, H. Yama-
taka, J. Am. Chem. Soc. 1994,
116, 10080; b) A. A. Restrepo-
Cossio, C. A. Gonzalez, F. Mari,
J. Phys. Chem. A 1998, 102, 6993;
c) H. Yamataka, S. Nagase, J.
Am. Chem. Soc. 1998, 120, 7530;
d) J. A. Doblado, H. Martinez-
Garcia, J. M. Molina, M. R. Sund-
berg, J. Am. Chem. Soc. 2000,
122, 1144.
[11] a) W. C. Lu, C. B. Liu, C. C. Sun,
J. Phys. Chem. A 1999, 103, 1078;
b) C. W. Lu, C. C. Sun, Q. J.
Zang, C. B. Liu, Chem. Phys.
Lett. 1999, 311, 491; c) J. Koketsu,
Y. Ninomiya, Y. Suzuki, N. Koga,
Inorg. Chem. 1997, 36, 694.
[12] D. S. Yufit, J. A. K. Howard,
M. G. Davidson, J. Chem. Soc.
Perkin Trans. 2 2000, 249.
[13] a) G. Fries, J. Wolf, M. Pfeiffer,
D. Stalke, H. Werner, Angew.
Chem. 2000, 112, 575; Angew.
Chem. Int. Ed. 2000, 39, 564;
b) L. Mahalakshmi, D. Stalke in
Structure and Bonding–Group
Mr [gmolꢀ1
CCDC no.
T [K]
]
444.54
231739
100(2)
triclinic
734.91
231740
100(2)
triclinic
750.96
231741
173(2)
monoclinic
P21/n
1124.06(18)
1488.1(2)
2478.7(4)
crystal system
space group
a [pm]
b [pm]
c [pm]
≈
≈
≈
P1
P1
P1
905.67(5)
1064.73(6)
1145.95(6)
97.0510(10)
99.3490(10)
107.1020(10)
1.02496(10)
2
1038.40(2)
1047.83(2)
1285.27(2)
73.078
81.058
67.668
1.23586(4)
2
1.195
0.18
476
1011.46(6)
1104.73(7)
1825.93(11)
93.3760(10)
93.6010(10)
98.7470(10)
2.0077(2)
2
1.216
0.203
780
0.4î0.4î0.35
2.11 26.41
22724
a [8]
b [8]
101.283(3)
g [8]
V [nmꢀ3
]
4.0661(11)
4
1.227
0.211
1592
0.4î0.4î0.4
3.05 26.37
87301
Z
1calcd [Mgmꢀ3
]
1.181
0.198
388
m [mmꢀ1
F(000)
]
crystal size [mm]
q range [8]
0.4î0.4î0.05 0.60î0.57î0.55
1.83 26.37
13417
1.66 54.54
reflns collected, low batch
reflns collected, high batch
independent reflns (Rint), low batch
independent reflns (Rint), high batch
absorption correction
max/min transmission
data/restraints/parameters
GOF on F2
51465[b]
74909[c]
4165 (0.0189)
5045 (0.0246)
25005 (0.0269)
empirical
0.99/0.95
30050/0/384
1.048
8174 (0.0400)
none
8675 (0.0950)
none
empirical
0.98/0.91
4165/0/229
1.051
8170/0/478
0.922
8300/0/478
1.086
final R indices [I>2s(I)]
R1
wR2
R indices (all data)
R1
13 Chemistry
I
(Eds.: D. A.
0.0343
0.0940
0.0298
0.0899
0.0410
0.0859
0.0414
0.1150
Atwood, H. W. Roesky), Spring-
er, Heidelberg, 2002, 103, p. 85.
[14] a) A. Steiner, D. Stalke, Angew.
Chem. 1995, 107, 1908; Angew.
Chem. Int. Ed. Engl. 1995, 34,
1752; b) S. Wingerter, M. Pfeiffer,
A. Murso, C. Lustig, T. Stey, V.
Chandrasekhar, D. Stalke, J. Am.
Chem. Soc. 2001, 123, 1381.
0.0389
0.0973
336/ꢀ196
0.0378
0.0956
695/ꢀ285
0.0581
0.0890
479/ꢀ360
0.0482
0.1188
813/ꢀ276
wR2
largest diff. peak/hole [enmꢀ3
]
[a] For the sake of comparison, the data of the conventional IAM refinement are given in this table. For the
results of the multipole refinement of ED, see Supporting Information. [b] sinq/l<0.625 äꢀ1. [c] 0.625<sinq/
l<1.145 äꢀ1
.
3630
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 3622 3631

Bonding in Iminophosphoranes
3622 3631
[15] Review: F. Baier, Z. Fei, H. Gornitzka, A. Murso, S. Neufeld, M.
Pfeiffer, I. R¸denauer, A. Steiner, T. Stey, D. Stalke, J. Organomet.
Chem. 2002, 661, 111.
[16] a) M. Albrecht, G. van Koten, Angew. Chem. 2001, 113, 3866;
Angew. Chem. Int. Ed. 2001, 40, 3750; b) H. Hagen, J. Boersma, G.
van Koten, Chem. Soc. Rev. 2002, 31, 357.
[33] M. F. Lappert, M. J. Slade, A. Singh, J. L. Atwood, R. D. Rogers, R.
Shakir, J. Am. Chem. Soc. 1983, 105, 302.
[34] A. M¸ller, M. Marsch, K. Harms, J. C. W. Lorenz, G. Boche, Angew.
Chem. 1996, 108, 1639; Angew. Chem. Int. Ed. Engl. 1996, 35, 1518.
[35] R. I. Papasergio, B. W. Skelton, P. Twiss, A. H. White, C. L. Raston,
J. Chem. Soc. Dalton Trans. 1990, 1161.
[17] G. Boche, Angew. Chem. 1989, 101, 286; Angew. Chem. Int. Ed.
Engl. 1989, 28, 277.
[36] U. Pieper, D. Stalke, Organometallics 1993, 12, 1201.
[37] D. Hoffmann, W. Bauer, P. v. R. Schleyer, U. Pieper, D. Stalke, Or-
ganometallics 1993, 12, 1193.
[38] U. Sch¸mann, U. Behrens, E. Weiss, Angew. Chem. 1989, 101, 481;
Angew. Chem. Int. Ed. Engl. 1989, 28, 476.
[18] a) A. M¸ller, B. Neum¸ller, K. Dehnicke, Chem. Ber. 1996, 129, 253;
b) A. M¸ller, B. Neum¸ller, K. Dehnicke, Angew. Chem. 1997, 109,
2447; Angew. Chem. Int. Ed. Engl. 1997, 36, 2350; c) P. B. Hitchcock,
M. F. Lappert, P. G. H. Uiterweerd, Z.-X. Wang, J. Chem. Soc.
Dalton Trans. 1999, 3413; d) I. Fernandez, G. Alvarez, M. Julia, N.
Kocher, D. Leusser, D. Stalke, J. Gonzalez, F. Lopÿz-Ortiz, J. Am.
Chem. Soc. 2002, 124, 15184.
[19] W.-P. Leung, Z.-X. Wang, H.-W Li, Q.-C. Yang, T. C. W. Mak, J.
Am. Chem. Soc. 2001, 123, 8123.
[20] a) W. J. Knebel, R. J. Angelici, Inorg. Chim. Acta 1973, 17, 713;
b) M. Alvarez, N. Lugan, R. Mathieu, J. Chem. Soc. Dalton Trans.
1994, 2755.
[21] a) A. F. Cameron, N. S. Hair, D. G. Norris, Acta Crystallogr. Sect. B
1974, 30, 221; b) E. Niecke, D. Gudat, Angew. Chem. 1991, 103, 251;
Angew. Chem. Int. Ed. Engl. 1991, 30, 217; c) A. Steiner, D. Stalke,
Inorg. Chem. 1993, 32, 1977; d) R. Fleischer, D. Stalke, Inorg. Chem.
1997, 36, 2413.
[39] G. Rabe, H. W. Roesky, D. Stalke, F. Pauer, G. M. Sheldrick, J. Or-
ganomet. Chem. 1991, 403, 11.
[40] M. Veith, J. Bˆhnlein, V. Huch, Chem. Ber. 1989, 122, 841.
[41] a) R. Appel, Pure Appl. Chem. 1987, 59, 977; b) D. G. Gilheany in
The Chemistry of Organophosphorus Compounds, Vol. 1 (Ed.: F. R.
Hartley), Wiley-Interscience, Chichester, 1990, Chapter 2, p. 9.
[42] a) J. C. J. Bart, J. Chem. Soc. B 1969, 350; b) H. Schmidbaur, A.
Schlier, C. M. F. Frazao, G. M¸ller, J. Am. Chem. Soc. 1986, 108,
976; c) H. Schmidbaur, J. Jeong, A. Schier, W. Graf, D. L. Wilkinson,
G. M¸ller, New J. Chem. 1989, 13, 341.
[43] a) N. K. Hansen, P. Coppens, Acta Crystallogr. Sect. A 1978, 34, 909;
b) T. Koritsanszky, S. Howard, T. Richter, Z. W. Su, P. R. Mallinson,
N. K. Hansen, XD–A Computer Program Package for Multipole
Refinement and Analysis of Electron Densities from Diffraction
Data. User Manual, Freie Universit‰t Berlin, 1995.
[44] R. F. W. Bader, Atoms in Molecules–A Quantum Theory, Clarendon
Press, Oxford 1990.
[45] Y. A. Abramov, L. Brammer, W. T. Klooster, R. M. Bullock, Inorg.
Chem. 1998, 37, 6317.
[46] A. Volkov, C. Gatti, Y. A. Abramov, Acta Crystallogr. Sect A. 2000,
56, 252.
[22] P. Rademacher, Strukturen organischer Molek¸le, VCH, Weinheim,
1987.
[23] a) W. N. Setzer, P. v. R. Schleyer, Adv. Organomet. Chem. 1985, 24,
353; b) C. Schade, P. v. R. Schleyer, Adv. Organomet. Chem. 1987,
27, 169; c) T. Stey, D. Stalke, Lead Structures in Lithium Organic
Chemistry in The Chemistry of Organolithium Compounds (Eds.: Z.
Rappoport, I. Marek), Wiley, Weinheim, 2003.
[24] a) A. Steiner, D. Stalke, J. Chem. Soc. Chem. Commun. 1993, 444;
b) H. Gornitzka, D. Stalke, Angew. Chem. 1994, 106, 695; Angew.
Chem. Int. Ed. Engl. 1994, 33, 693; c) A. Steiner, D. Stalke, Organo-
metallics 1995, 14, 2422.
[47] W. Scherer, P. Sirsch, D. Shorokhov, G. McGrady, S. A. Mason,
M. G. Gardiner, Chem. Eur. J. 2002, 8, 2324.
[48] D. Leusser, B. Walfort, D. Stalke, Angew. Chem. 2002, 114, 2183;
Angew. Chem. Int. Ed. 2002, 41, 2079.
[25] R. P. K. Babu, K. Aparna, R. McDonald, R. G. Cavell, Organometal-
lics 2001, 20, 1451.
[26] M. T. Gamer, P. W. Roesky, Z. Anorg. Allg. Chem. 2001, 627, 877.
[27] R. P. K. Babu, K. Aparna, R. McDonald, R. G. Cavell, Inorg. Chem.
2000, 39, 4981.
[28] a) K. Gregory, P. v. R. Schleyer, R. Snaith, Adv. Inorg. Chem. 1991,
37, 47; b) R. E. Mulvey, Chem. Soc. Rev. 1991, 20, 167; c) R. E.
Mulvey, Chem. Soc. Rev. 1998, 27, 339.
[29] a) P. B. Hitchcock, M. F. Lappert, Z.-X. Wang, Chem. Commun.
1997, 1113; b) A. M¸ller, B. Neum¸ller, K. Dehnicke, Angew. Chem.
1997, 109, 2447; Angew. Chem. Int. Ed. Engl. 1997, 36, 2350.
[30] a) T. Chivers, A. Downard, Y. P. A. Glenn, J. Chem. Soc. Dalton
Trans. 1998, 16, 2603; b) T. Chivers, A. Downard, M. Parvez, Inorg.
Chem. 1999, 38, 4347.
[49] a) T. Kottke, D. Stalke, J. Appl. Crystallogr. 1993, 26, 615; b) T.
Kottke, R. J. Lagow, D. Stalke, J. Appl. Crystallogr. 1996, 29, 465;
c) D. Stalke, Chem. Soc. Rev. 1998, 27, 171.
[50] Bruker, SMART-NT, Data Collection Software, Version 5.6, Bruker
Analytical X-ray Instruments Inc., Madison, Wisconsin (USA),
2000.
[51] Bruker, SAINT-NT, Data Reduction Software, Version 6, Bruker
Analytical X-ray Instruments, Inc., Madison, Wisconsin (USA),
1999.
[52] Bruker, SHELX-TL, Version 6, Bruker Analytical X-ray Instru-
ments Inc., Madison, Wisconsin (USA), 2000.
[53] G. M. Sheldrick, SADABS2, Empirical Absorption Correction Pro-
gram, University of Gˆttingen (Germany) 2001.
[31] A. Steiner, PhD thesis, Universit‰t Gˆttingen 1994.
[32] A. M¸ller, M. Mˆhlen, B. Neum¸ller, N. Faza, W. Massa, K. Deh-
nicke, Z. Anorg. Allg. Chem. 1999, 625, 1748.
Received: February 18, 2004
Published online: June 2, 2004
Chem. Eur. J. 2004, 10, 3622 3631
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3631