Heptaphosphanortricyclenes with oligosilyl substituents: Syntheses and reactions

Article abstract of DOI:10.1002/ejic.201000749

Three new nortricyclic P₇R₃ derivatives with R = (SiMe₃)₂MeSi- (2), (SiMe₃)₂PhSi- (3), and cyclo-Si₆Me₁₁- (4) were synthesized from red phosphorus, sodium/potassium al

Full text of DOI:10.1002/ejic.201000749

FULL PAPER
DOI: 10.1002/ejic.201000749
Heptaphosphanortricyclenes with Oligosilyl Substituents: Syntheses and
Reactions
Pierre Noblet,[a] Vittorio Cappello,[a] Günter Tekautz,[a] Judith Baumgartner,[a][‡] and
Karl Hassler*[a]
Keywords: Heptaphosphanes / Oligosilyl groups / X-ray structural analysis / NMR spectroscopy / Silicon
Three new nortricyclic P7R3 derivatives with R = (SiMe3)2-
MeSi– (2), (SiMe3)2PhSi– (3), and cyclo-Si6Me11– (4) were
synthesized from red phosphorus, sodium/potassium alloy,
and a chlorooligosilane, and their structures were elucidated
with X-ray diffraction. Reactions of 2 and 3 and of tri(hyper-
The silylene immediately inserted into a P–P bond of the
basal P3 ring of 1a, thereby forming two new isomeric anions
{Hyp2P7[Si(SiMe3)2]}–, 1b and 1c. The structure of 1c could
be elucidated by X-ray crystallography, whereas 1b was
characterized by 31P NMR spectroscopy. The proposed reac-
tion mechanism could be further confirmed by performing
the reaction between 1 and KOtBu in the presence of a large
excess amount of dimethylbutadiene, which trapped the sil-
ylene. Just the formation of anion 1a was observed in that
case. Anion 1a was also obtained quantitatively when KOtBu
was replaced by LiOtBu. Compound 1a reacted with tri-
fluoroacetic acid to give Hyp2P7H (1d), which was success-
fully crystallized and examined by X-ray diffraction. It is the
first structural example of a stable and neutral nortricyclic P7
derivative, R2P7H. All compounds were chracterized further
by 31P and 29Si NMR spectroscopy and elemental analyses.
silyl)heptaphosphane
1 [hypersilyl = (SiMe3)3Si–] with
KOtBu and LiOtBu were performed, which led to different
results depending on the size of the substituent. With KOtBu,
Si–P bonds were cleaved in 2 and 3, and the mono- and di-
anions [P7R2]– [R = (SiMe3)2MeSi: 2a; R = (SiMe3)2PhSi: 3a]
and [P7R]2– [R = (SiMe3)2MeSi: 2b] formed. Crystals of the
[K+.18-crown-6]+ salt of 2b suitable for X-ray diffraction could
be grown successfully. Compound 1 reacted in an unprece-
dented way with KOtBu at –60 °C. An Si–Si bond was
cleaved, and a transient silyl anion formed, which immedi-
ately rearranged into a transient heptaphosphanide anion
[Hyp2P7]– (1a) under expulsion of bis(trimethylsilyl)silylene.
Introduction
The tricyclic cage of the heptaphosphide anion [P7]3– is
a structural unit that is not only interesting and important
for solid-state chemistry, but also for molecular chemistry.[1]
It is present in various Zintl phases, for instance, Li3P7.[2]
Its molecular structure has also been determined from the
solvate Li3P7(tmeda)3 (tmeda = N,N,NЈ,NЈ-tetramethyleth-
ylenediamine).[3] Solvated Li3P7 is fluxional in solution as
well as in the solid state; it undergoes a Cope rearrangement
similar to bullvalene.[4]
Scheme 1.
The molecular derivatives P7R3 feature a nortricyclane been reported, albeit without isolation of the cages.[10] The
structure with either C3 or C3v symmetry of the P7 unit number of As7R3 molecules described in the literature is
closely related to the structure of P4S3 (see Scheme 1).
much smaller, and no neutral molecular derivatives Sb7R3
A variety of substituents R such as CnHm (Me, Et, Bu, or Bi7R3 have been described yet, despite the fact that the
iPr, benzyl),[5] SiMe3,[6] SiPh3,[7] SitBu3,[8] SnPh3, GePh3, or anions [Sb7]3– and [Bi7]3– are well known as constituents of
PbPh3 (Ph = C6H5),[9] as well as phosphanyl groups[1] have some Zintl phases.[11]
been introduced into the P7 system over the last thirty years.
There are numerous examples of relatively stable poly-
It is noteworthy that even the formation of P7Br3 and P7I3 phosphorus anions that can be isolated,[12] which contrasts
in the reaction of white phosphorus P4 with Br2 or I2 has with a relatively small number of isolable neutral hydrogen
polyphosphanes that are inherently unstable.[1,13] So far, no
[a] Institute of Inorganic Chemistry, University of Technology,
neutral derivatives R2P7H or RP7H2 have been isolated and
Stremayrgasse 16, 8010 Graz, Austria
characterized structurally. However, the structures of the
E-mail: karl.hassler@tugraz.at
[‡] X-ray structure analysis
anions [H2P7]– and [HP7]2–, which in solution also possess
Eur. J. Inorg. Chem. 2011, 101–109
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
101

P. Noblet, V. Cappello, G. Tekautz, J. Baumgartner, K. Hassler
FULL PAPER
the marked tendency to disproportionate, could be deter- accumulate a negative charge. Consequently, derivatives
mined in their tetraphenylphosphonium salts.[14,15] Further such as [RP7]2–, [R2P7]–, and P7R3 are expected to possess
studies revealed that alkyl group transfer between anions exceptional electronic stability for substituents R that are
allowed the preparation of partially alkylated species capable of supporting sufficient positive charge. A further
[R2P7]– and [RP7]2–, and that the reaction of P7(SiMe3)3 kinetic stabilization of these species can be achieved if in
with nBuLi resulted in the formation of [P7(SiMe3)2]–.[16,17] addition R is a sterically demanding group. As both these
The first crystal structure of an alkylated anion [R2P7]– properties can be found in oligosilyl groups, we have set out
(R = benzyl) was not reported before 1998.[18] So far, no to develop a method for the synthesis of oligosilylhepta-
dianion [RP7]2– could be characterized structurally.
phosphanes and to investigate their reactivities. The intro-
Two methods are commonly used for the preparation of duction of oligosilyl groups offers two further advantages
these species. First, melts of the composition K3P7 and that are not found in organoheptaphosphanes. First, Si–P
K3As7 can be used by fusing the elements at approximately bonds can be cleaved at low temperatures with various rea-
1000 °C in sealed tubes, a procedure for which special gents such as BuLi, tBuOK, or C2Cl6, and second, Si–Si
equipment is needed. Moreover, it is extremely dangerous bonds also can be cleaved at temperatures as low as –70 °C,
because polypnictides are known to spontaneously detonate for instance, with tBuOK.[25]
even under anaerobic conditions, thus making this method
All cages P7R3 with R = SiMe(SiMe3)2 (2), SiPh-
unattractive for preparative purposes. However, the use of (SiMe3)2 (3), and Si6Me11 (4), were prepared from red phos-
the pure Zintl phases allowed the preparation and struc- phorus, sodium/potassium alloy and a halosilane RX. For
tural characterization not only of derivatives P7R3, but also R = cyclo-Si6Me11, iodoundecamethylcyclohexasilane was
of [H2P7]–, [HP7]2–, [(C6H5CH2)2P7]–, and [(C6H5CH2)2- used due to its increased reactivity; chlorosilanes were used
As7]–.[18,19]
The second method starts from elemental phosphorus, Experimental Section.
in all other cases. A general procedure is presented in the
preferably (but not necessarily) from the red modification
X-ray structures of the cages 2, 3, and 4, which crystallize
due to safety reasons. In a solvent such as dimethoxyethane, racemically as mixtures of enantiomers, are shown in Fig-
red or white phosphorus reacts with sodium/potassium al- ure 1 with hydrogen atoms omitted for clarity. Details of
loy to form a nonstoichiometric phosphide NaxP/KxP, the structures can be found in the Experimental Section.
which gives excellent yields of (SiMe3)3P7 (78%) when
As all bond lengths and bond angles are in a normal
treated with ClSiMe3.[20] With this method, we were able to range, we will not discuss them here, but focus briefly on the
prepare the first oligosilyl-substituted heptaphosphane distance h from the apical phosphorus atom to the center of
P7R3 with R = Si(SiMe3)3 (1).[21] In the present work, we the basal triangular plane, which depends on the type of
describe the synthesis of the cages with R = SiMe(SiMe3)2 substituents (Scheme 2).
(2), SiPh(SiMe3)2 (3), and cyclo-Si6Me11 (4), as well as some
reactions of 1, 2, and 3. Preliminary results of these studies
have already been reported elsewhere.[22,23] It is worth
mentioning that substituted P7 systems such as
[HypP7(PHyp)]2– are formed in the treatment of white
phosphorus with KSi(SiMe3)3.[24]
Scheme 2.
Results and Discussion
It has been pointed out by von Schnering that this dis-
Syntheses and X-ray Structures of P7R3
tance is minimal for [P7]3– (ca. 300 pm[26]) due to the repul-
The stability of the heptaphosphanide [P7]3– is well sion of the three negative charges. It increases by about
known and certainly reflects the tendency of phosphorus to 15 pm for substituents such as SiMe3 (315 pm[26]), Si(tBu)3
Figure 1. ORTEP drawings (30% probabilities, top views) of the molecular structures of P7[SiMe(SiMe3)2]3 (2, left), P7[SiPh(SiMe3)2]3 (3,
center), and P7[Si6Me11]3 (4, right). Hydrogen atoms are omitted for clarity.
102
www.eurjic.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2011, 101–109

Heptaphosphanortricyclenes with Oligosilyl Substituents
(316 pm[8]), and Si(SiMe3)3 (318 pm[21]). The values for the Reaction of 1 with tBuOK and tBuOLi
cages 2, 3, and 4 prepared in this work are (315Ϯ1) pm,
close to that for SiMe3. From group-electronegativity con-
As reported earlier, tBuOK can be used for cleaving
siderations, the partial negative charge on the equatorial Si–P bonds to give a potassium phosphanide and a silyl
phosphorus atoms is expected to decrease in the series Si- ether. For instance, P(SiMe3)3 reacts with tBuOK to form
(SiMe3)3 Ͼ Si6Me11 Ͼ SiMe(SiMe3)2 Ͼ SiPh(SiMe3)2
Ͼ
KP(SiMe3)2 and tBuOSiMe3,[28] and experiments with
SiMe3. Accordingly, the cages should become more elon- (SiMe3)3SiP(SiMe3)2 have shown that it is the P-trimethyl-
gated, which is not observed. On the contrary, tri(hyper- silyl group that is cleaved exclusively with concomitant for-
silyl)heptaphosphane 1 [hypersilyl = Hyp = Si(SiMe3)3] pos- mation of (SiMe3)3SiP(SiMe3)K.[29] Unshielded Si–P bonds
sesses the largest h value, which we attribute to nonbonded react much more quickly with tBuOK than Si–Si bonds do,
van der Waals interactions between the hypersilyl groups.[21] which is easily understood from bond polarities.
The smallest H···H nonbonded distances in 1 are close to
Due to the effective shielding of the Si–P bonds by the
the twofold value of the van der Waals radius for hydrogen, hypersilyl groups, the treatment of tri(hypersilyl)heptaphos-
given in the literature as 120–145 pm.[27] The three hyper- phane with 1 equiv. of tBuOK in thf proceeds by means of
silyl groups therefore act like a belt that compresses and Si–Si bond cleavage (Scheme 3). The resulting silyl anion
elongates the cage. Quite obviously, these interactions are immediately rearranges into a heptaphosphanide anion
somewhat smaller for the other oligosilyl groups.
[Hyp2P7]– (1a) and bis(trimethylsilyl)silylene, which is in-
Scheme 3. Reaction of tri(hypersilyl)heptaphosphane 1 with tBuOK and tBuOLi.
Figure 2. ORTEP plots (30% probabilities, hydrogen atoms not shown) of the molecular structure of the K·18-crown-6 salt of 1c shown
without (left) and with SiMe3 groups (right). Selected distances [pm] and angles [°]: P(1)–P(2) 223.2(3), P(1)–P(3) 222.3(3), P(1)–P(4)
230.0(11), P(2)–P(5) 224.8(12), P(3)–P(6) 226.3(7), P(4)–P(6) 229.3(17), P(7)–P(5) 214.8(14), P(5)–Si(3) 230.7(12), P(6)–Si(3) 233.5(7),
P(4)–K 364.5(12), P(7)–K 366.3(11), P(2)–Si(2) 229.6(3), P(3)–Si(1) 227.5(2); P(2)–P(1)–P(3) 96.07(10), P(2)–P(1)–P(4) 101.0(3), P(3)–
P(1)–P(4) 90.5(3), P(1)–P(2)–P(5) 104.3(3), P(1)–P(3)–P(6) 90.6(3), P(1)–P(4)–P(7) 107.7(5), P(1)–P(4)–P(6) 87.9(4), P(5)–P(7)–P(4) 66.6(5),
P(6)–P(4)–P(7) 109.6(6), P(5)–Si(3)–P(6) 105.9(3), P(1)–P(2)–Si(2) 103.73(11), P(1)–P(3)–Si(1) 100.13(10).
Eur. J. Inorg. Chem. 2011, 101–109
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
103

P. Noblet, V. Cappello, G. Tekautz, J. Baumgartner, K. Hassler
FULL PAPER
serted into the P–P bond of the basal P3 ring that lies oppo- transformed into a five-membered SiP4 ring with concomi-
site to the P atom, which formally bears the negative charge. tant formation of a P4 ring. Scheme 1 summarizes the reac-
A novel phosphanide anion [Hyp2P7Si(SiMe3)2]– (1b) is tion sequence, which occurs at a temperature as low as
formed, which then rearranges into anion 1c through P–P –70 °C. To the best of our knowledge, this is the first exam-
bond shift. The four-membered SiP3 ring present in 1b is ple of a silylene insertion into a P–P bond of the P7 nor-
Scheme 4. Isomers of [R2P7]–, top views. C is chiral.
Scheme 5. Numbering of the P atoms in 1, 1a, 1b, and 1c; R = SiMe3.
Figure 3. 31P{1H} NMR spectra (δ [ppm]) of 1, of the potassium salt 1a, of 1b, and of an approximately 20:80 mixture between 1b and
1c. From top to bottom.
104
www.eurjic.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2011, 101–109

Heptaphosphanortricyclenes with Oligosilyl Substituents
tricyclane system. With regard to previously reported reac- of isomer B is unlikely due to steric interaction of the two
tions between P4 and silylenes R2Si, it is not surprising that hypersilyl groups.
the insertion occurs into a P–P bond of the P3 ring.[30]
The lithium salt of 1a can be easily obtained by replacing
When the reaction between 1 and tBuOK is carried out tBuOK by tBuOLi, which possesses much less tendency to
in the presence of 18-crown-6, 1c crystallizes from toluene cleave Si–Si bonds. Again, isomer C is formed (Scheme 4).
as the [K.18-crown-6]+ salt. Figure 2 presents two views of
Due to magnetic inequivalence of the P atoms, all signals
the molecular structure of the anion and shows that the in the proton-decoupled 31P NMR spectra of cages 1–1c
orientation of a hypersilyl group has changed due to inver- are second-order multiplets. Scheme 5 presents the number-
sion at a P atom. So far, we have been unable to obtain ing of the P atoms used in the assignment of the resonances,
crystals of 1a or 1b, because only 1c crystallized from the and Figure 3 displays the spectra of pure 1, 1a, and 1b and
reaction mixture. By employing low-temperature 31P NMR of a mixture between 1b and 1c, which contained about
spectroscopy, we were, however, able to record spectra of 80% of 1c.
pure 1b prior to the isomerization to 1c. When crystals of
The peak assignments correspond to the atom number-
the [K·18-crown-6]+ salt of 1c are dissolved in toluene or ing used in Scheme 5 and are supported by connectivities
thf, the equilibrium between 1b and 1c is reestablished im- obtained from COSY spectra.
mediately; all attempts to obtain 31P NMR spectra of pure
1c failed. It is also noteworthy that neither 1b nor 1c do
react with a further equivalent of KOtBu at temperatures
Formation of Hyp2P7H (1d)
below 30 °C.
By treatment of 1a with CF3COOH, colorless single crys-
tals of the neutral heteroleptic cage Hyp2P7H could be ob-
tained from n-hexane, which we were able to characterize
by X-ray diffraction. It is the first known structure of a
neutral heptaphosphane that bears a hydrogen atom, and
at the same time of a neutral heteroleptically substituted
heptaphosphane. An ORTEP drawing is presented in Fig-
ure 4. The compound is stable at room temperature and
shows no tendency to undergo hydrogen/silyl exchange re-
actions.
To corroborate the hypothesis that bis(trimethylsilyl)silyl-
ene is an intermediate, the reaction between 1 and KOtBu
was repeated in the presence of an excess amount of di-
methylbutadiene, with the expectation that the silylene
would be trapped and that anion 1a would be formed,
which it did. Though we were unsuccessful in growing crys-
tals of 1a, the 31P NMR spectra are unambiguous. More-
over, treatment of 1 with CsF in thf also resulted in the
formation of 1a through Si–P bond cleavage, thereby giving
an identical 31P NMR spectrum. As six signals in the inten-
sity ratio 1:1:1:1:1:2 are observed for 1a, the asymmetric
isomer C formed, as shown below. This can be deduced
from the presence of two signals with triplet structures in
the range for the equatorial, Si-substituted P atoms at δ =
Treatment of 1c + 1d with Chlorosilanes
When we treated a mixture of the anions 1c and 1d with
–10 and –50 ppm. Formation of 1a through Si–P bond ClSiMe2SiMe2Cl, two heptaphosphane cages connected
cleavage thus involves no inversion at a P atom. Formation with an SiMe2SiMe2 bridge formed, which was corrobo-
Figure 4. ORTEP drawing (30% probabilities, hydrogen atoms not shown) of the molecular structure of Hyp2P7H (1d). Selected distances
[pm] and angles [°]: P(1)–P(2) 215.8(2), P(1)–P(3) 216.7(2), P(1)–P(4) 216.9(2), P(2)–P(5) 219.3(2), P(3)–P(7) 218.9(2), P(4)–P(6) 218.3(2),
P(5)–P(6) 228.0(2), P(5)–P(7) 220.3(2), P(6)–P(7) 219.9(2), P(2)–H(99) 115(4), P(3)–Si(1) 226.6(2), P(4)–Si(5) 228.0(2); P(3)–P(1)–P(2)
99.60(9), P(3)–P(1)–P(4) 98.12(8), P(2)–P(1)–P(4) 95.67(8), P(1)–P(2)–P(5) 103.32(9), P(1)–P(3)–P(7) 102.28(8), P(1)–P(4)–P(6) 103.03(8),
P(2)–P(5)–P(6) 105.94(9), P(2)–P(5)–P(7) 100.27(8), P(3)–P(7)–P(5) 108.22(9), P(3)–P(7)–P(6) 100.69(8), P(4)–P(6)–P(5) 100.10(8), P(4)–
P(6)–P(7) 107.73(8), P(5)–P(7)–P(6) 60.06(7), P(5)–P(6)–P(7) 60.07(8), P(6)–P(5)–P(7) 59.87(7), P(1)–P(3)–Si(1) 111.16(8), P(1)–P(4)–Si(5)
110.54(8).
Eur. J. Inorg. Chem. 2011, 101–109
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
105

P. Noblet, V. Cappello, G. Tekautz, J. Baumgartner, K. Hassler
FULL PAPER
Figure 5. ORTEP plots (30% probabilities, hydrogen atoms omitted for clarity) showing the molecular structure of [(SiMe3)2-
MeSiP7]2–2[K+·18-crown-6] (2b). Left: top view; right: side view. Selected distances [pm] and angles [°]: P(1)–P(2) 217.4(1), P(1)–P(3)
217.0(1), P(1)–P(4) 221.6(1), P(2)–P(5) 215.4(1), P(3)–P(6) 213.4(1), P(4)–P(7) 218.7(1), P(5)–P(6) 225.7(1), P(5)–P(7) 226.2(1), P(6)–P(7)
226.5(1), P(4)–Si(1) 227.0(1), P(2)–K(1) 336.5(1), P(3)–K(1) 359.5(1), P(5)–K(1) 367.9(1), P(6)–K(1) 383.3(1), P(2)–K(2) 334.3(1), P(5)–
K(2) 356.2(2), P(4)–K(2) 367.6(2); P(3)–P(1)–P(2) 105.02(6), P(3)–P(1)–P(4) 103.03(5), P(2)–P(1)–P(4) 93.47(6), P(1)–P(2)–P(5) 99.11(5),
P(1)–P(3)–P(6) 99.07(5), P(1)–P(4)–P(7) 101.88(5), P(5)–P(7)–P(6) 60.04(4), P(6)–P(7)–P(5) 59.80(4), P(7)–P(5)–P(6) 60.16(4), P(2)–P(5)–
P(7) 106.80(5), P(3)–P(6)–P(5) 106.42(5), P(4)–P(7)–P(6) 103.19(5).
rated by the 31P NMR spectra. No crystals could be ob- with formation of Li3P7, LiP(SiMe3)2, P(SiMe3)3.[17] Inter-
tained yet. Hyp2P7–SiMe2SiMe2–P7Hyp2 could also be ob- estingly, isomer C is obtained when tBuOK is used for the
tained by treating sodium/potassium phosphide first with Si–P bond cleavage, but isomer A with tBuOLi. At room
1 equiv. of ClSiMe2SiMe2Cl and in a second step with temperature, both the lithium and potassium salts are stable
4 equiv. of (SiMe3)3SiCl. A mixture between R2P7SiMe2Si- in solution for days.[32]
Me2P7R2 and 1 formed, from which 1 could not be sepa-
As the 31P NMR spectra are very similar to those of 1a,
rated quantitatively through fractional crystallization from no pictures are presented here. In the presence of 18-crown-
toluene. With (SiMe3)3SiCl, 1c + 1d give back cage 1 exclu- 6, needle-like crystals of 2a could be grown from toluene.
sively. Quite obviously, bis(trimethylsilyl)silylene is expelled Unfortunately, they were too thin for X-ray diffraction.
under the reaction conditions, and the cyclotriphosphane
Treatment of 2 with 2 equiv. of tBuOK afforded the di-
ring is reformed. This is not surprising, as reactions of silyl- anion [RP7]2– (2b), which in the presence of 18-crown-6
enes are known to be reversible.[31] Most noteworthy, the gave crystals suitable for a structure determination by em-
treatment of 1c + 1d with chlorotrimethylsilane proceeds ploying X-ray diffraction. Results are presented in Figure 5.
without extrusion of (SiMe3)2Si, as several second-order
As can be deduced from the NMR spectra, 2b is flux-
signals between δ = +120 and 0 ppm prove. So far, we have ional in solution and undergoes a Cope-like rearrangement.
been unable to separate silylated 1c from silylated 1d by At room temperature, just broad signals are observed,
fractional crystallization. Therefore, detailed results of these which split into seven second-order resonances at low tem-
ongoing investigations will be reported elsewhere.
peratures (–60 °C). An analogous degenerate Cope re-
arrangement between two identical valence tautomers was
also observed for [HP7]2– by employing temperature-de-
pendent 31P NMR spectroscopy.[33]
Treatment of 2 and 3 with tBuOK
As the Si–P bonds in 2 and 3 are less shielded than in 1
due to the smaller size of the oligosilyl groups, the treat-
ments with tBuOK proceed by means of Si–P cleavage.
With 1 equiv. of tBuOK, the monoanions [R2P7]– form
quantitatively [R = SiMe(SiMe3)2: 2a; R = SiPh(SiMe3)2:
Conclusion
Heptaphosphanes P7R3 functionalized with various oli-
3a]. From the 31P NMR spectra we deduce that the asym- gosilyl groups R can be synthesized from red phosphorus,
metric isomers C are formed in both cases, despite the fact sodium/potassium alloy, and a suitable halooligosilane.
that only five resonances with intensities 1:1:2:1:2 are ob- They show unusual thermal stability and can be used for
served in each case because of overlapping signals. Two sig- further derivatizations, either by Si–P or Si–Si bond cleav-
nals with a triplet structure are observed in the region be- age.
tween δ = 0 and –30 ppm, which is typical for the equatorial
When R is a trisilanyl group, one or two Si–P bonds can
P atoms that bear silyl substituents (see the Experimental be cleaved by KOtBu with concomitant formation of the
Section). Due to substituent interactions, formation of iso- heptaphosphanides [P7R2]– or [P7R]2– and a silyl ether
mers B can safely be excluded.
It is of some note that in the reaction of P7(SiMe3)3 with (SiMe3)2]–Li+, which decomposes quickly, the potassium
tBuLi or LiP(SiMe3)2, the symmetric isomer of salts are stable, crystallize readily in the presence of 18-
[P7(SiMe3)2]– is initially formed, which decomposes quickly crown-6, and can be used for further reactions.
tBuOR. Contrary to lithium heptaphosphanide [P7-
A
106
www.eurjic.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2011, 101–109

Heptaphosphanortricyclenes with Oligosilyl Substituents
fractometer by using graphite-monochromated Mo-Kα radiation (λ
= 0.71073 Å). The data were reduced to Fo and corrected for ab-
When R is the hypersilyl group (SiMe3)3Si, the Si–P
bonds in P7R3 are shielded very effectively, and a trimethyl-
silyl group is cleaved off by KOtBu, even at temperatures of
–60 °C. The resulting silyl anion immediately rearranges
into a heptaphosphanide anion and bis(trimethylsilyl)silyl-
ene, which then inserts into a P–P bond of the P7 cage. Two
isomeric novel potassium heptaphosphanides with P7Si
backbones are formed, which are in equilibrium with each
other through a P–P bond shift. They may also be used for
2
sorption effects with SAINT[34] and SADABS,[35] respectively. The
structures were solved by direct methods and refined by full-matrix
least-squares method (SHELXL97).[36] If not noted otherwise, all
non-hydrogen atoms were refined with anisotropic displacement
parameters. All hydrogen atoms were located in calculated posi-
tions to correspond to standard bond lengths and angles, except
for H(99) in 1d. All diagrams were drawn with 30% probability
thermal ellipsoids, and all hydrogen atoms were omitted for clarity.
further reactions with halosilanes. With LiOtBu, only Si–P Crystallographic data (excluding structure factors) for the struc-
tures of compounds 1c, 1d, 2, 2b, 3, and 4 reported in this paper
are summarized in Table 1. CCDC-612400 (1c), -612415 (1d),
-612402 (2), -612401 (2b), -612399 (3), and -612403 (4) contain the
supplementary crystallographic data for this paper. These data can
be obtained free of charge from The Cambridge Crystallographic
Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
bonds are cleaved, and the heptaphosphanide salt
[P7R2]–Li+ with R = hypersilyl is obtained.
These results open up doors for further investigations,
which are in progress.
Synthesis: All syntheses and manipulations were carried out under
either N2 or Ar by using standard Schlenk techniques. Solvents
were first dried with Al2O3 by using a column purification system
from Innovative Technology and then distilled from sodium, potas-
sium, sodium/potassium alloy, or LiAlH4 to remove traces of
moisture and oxygen. 18-crown-6 and KOtBu were subjected to
vacuum sublimation prior to use. Chloro- and iodosilanes such as
(Me3Si)3SiCl, (Me3Si)2SiMeCl, (Me3Si)2SiPhCl, and (Me2Si)5Si-
MeI, were prepared and purified according to procedures described
in the literature. Elemental analyses were performed with a Heraeus
Vario Elementar.
Experimental Section
Spectroscopy: NMR spectra were recorded with a Bruker MSL 300
or a Varian Unity Inova 300 spectrometer. The 29Si{1H} and
31P{1H} spectra were measured in solutions in thf or toluene in
10 mm tubes with capillaries containing D2O that served as exter-
nal lock, except when C6D6 was used. The insensitive nuclei en-
hanced by polarization transfer (INEPT) pulse sequence was used
for the 29Si spectra. Due to coupling over two and three bonds,
1
which causes line broadening, the accuracy of JSi,P values is not
better than an estimated Ϯ5 Hz.
X-ray Structure Analysis: For X-ray structure analyses, the crystals
were mounted onto the tip of glass fibers, and data collection was
General Procedure for the Synthesis of P7R3: Typically, red phos-
phorus (8.06 g, 260.00 mmol) was suspended in dimethoxyethane
(dme; 150 mL). Then sodium/potassium alloy (3.40 g) prepared
performed with
a Bruker-AXS SMART APEX CCD dif-
Table 1. Crystallographic data for compounds 1c, 1d, 2, 2b, 3, and 4.
1d
2b
2
1c
3
4
Empirical formula
Mr
T [K]
C18H55P7Si8
713.13
100(2)
C45H85K2O12P7Si3 C21H63P7Si9
C36H87KO6P7Si11
1180.94
223(2)
C36H69P7Si9
971.51
223(2)
C33H99P7Si18
1218.53
100(2)
1197.39
100(2)
785.31
100(2)
Size [mm]
Crystal system
Space group
a [Å]
b [Å]
c [Å]
0.38ϫ0.28ϫ0.20
monoclinic
P2/c
15.712(3)
9.605(2)
26.969(5)
90
101.67(3)
90
3986(2)
4
0.42ϫ0.38ϫ0.28
monoclinic
P21/n
17.570(4)
20.443(4)
18.726(4)
90
105.37(3)
90
6485(2)
4
0.36ϫ0.28ϫ0.16
triclinic
P1
0.38ϫ0.26ϫ0.15
triclinic
P1
0.45ϫ0.32ϫ0.01
monoclinic
P21/c
21.069(4)
10.880(2)
26.624(5)
90
109.74(3)
90
5744(2)
4
0.44ϫ0.32ϫ0.20
trigonal
R3c
15.639(3)
15.639(3)
53.22(2)
90
90
120
11273(5)
6
1.077
¯
¯
9.689(2)
15.325(3)
15.900(3)
99.90(3)
91.02(3)
97.63(3)
2303(2)
2
13.297(6)
15.036(7)
20.819(10)
69.65(2)
79.46(2)
64.14(2)
3509(3)(2)
2
α [°]
β [°]
γ [°]
V [Å3]
Z
ρcalcd. [gcm–3]
1.188
0.561
1.226
0.423
1.132
0.516
1.118
0.456
1.123
0.426
µ [mm–1]
0.473
F(000)
1520
2544
840
1258
2064
3924
θ range [°]
Reflections collected/
unique
Completeness to θ [%]
Data/restraints/
parameters
Goodness of fit on F2
Final R indices
[IϾ2σ(I)]
R indices
1.32–24.72
1.50–24.71
1.71–25.00
1.78–24.71
1.03–23.25
1.69–26.36
27335/6820
99.9
44643/11042
99.9
10106/7401
91.1
23930/11625
97.1
34503/8243
99.9
28347/5117
99.9
6820/0/320
0.99
R1 = 0.063,
wR2 = 0.145
R1 = 0.116,
wR2 = 0.167
11042/0/631
1.10
R1 = 0.059,
wR2 = 0.113
R1 = 0.081,
wR2 = 0.122
7401/0/355
1.04
R1 = 0.043,
wR2 = 0.116
R1 = 0.050,
wR2 = 0.122
11625/909/934
1.13
R1 = 0.103,
wR2 = 0.257
R1 = 0.125,
wR2 = 0.270
8243/0/445
1.04
R1 = 0.078,
wR2 = 0.201
R1 = 0.118,
wR2 = 0.222
5117/1/187
1.13
R1 = 0.064,
wR2 = 0.124
R1 = 0.073,
wR2 = 0.128
(all data)
Largest difference
peak/hole [eÅ3]
0.69/–0.45
0.48/–0.31
0.73/–0.35
1.08/–0.58
0.98/–0.44
0.51/–0.44
Eur. J. Inorg. Chem. 2011, 101–109
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
107

P. Noblet, V. Cappello, G. Tekautz, J. Baumgartner, K. Hassler
from Na (1.47 g, 64.00 mmol) and K (1.86 g, 47.5 mmol) was dropwise to a solution of 1a (1.05 mmol) in toluene [10 mL, pre-
FULL PAPER
added, and the reaction mixture was heated to reflux for 20 h. The
dme was then removed by evaporation in vacuo and replaced by
toluene (ca. 200 mL). Then a solution of the required oligosilane
(111.5 mmol) in toluene (50 mL) was added dropwise at room tem-
perature [for 1: (SiMe3)3SiCl (31.57 g); for 2: (SiMe3)2MeSiCl
(25.08 g); for 3: (SiMe3)2PhSiCl (32.00 g); for 4: Si6Me11I (51.38 g)].
Subsequently, the mixture was again heated to reflux for 3 h to
complete the reaction, and the salts were then separated by fil-
tration. From the toluene solutions, colorless crystals of the hepta-
phosphanes with a quality suitable for X-ray analysis could be
grown at –80 °C. Yields ranged from 50 to 75%.
Compound 2: 31P NMR (toluene): δ = 0 (m, 3 P), –85 (m, 1 P),
–167 (m, 3 P) ppm. 29Si NMR (toluene): δ = –15.5 (SiMe3), –51.7
[1JSi,P = (83.5Ϯ5) Hz, SiMe] ppm. C21H63Si9P7 (785.30): calcd. C
33.02, H 8.81; found C 32.25, H 7.97.
Compound 3: 31P NMR (toluene): δ = 0 (m, 3 P), –98 (m, 1 P),
–165 (m, 3 P) ppm. 29Si NMR (toluene): δ = –13.9 (SiMe3), –51.0
[1JSi,P = (95.4Ϯ5) Hz, SiPh)] ppm. C36H69Si9P7 (971.51): calcd. C
44.51, H 7.16; found C 44.19, H 7.47.
pared from 1 and KOtBu (1.00 g)] at –70 °C. The reaction mixture
was then allowed to come to room temperature. After decantation
from the salts, the toluene was removed in vacuo and replaced by
n-hexane. At –30 °C, colorless crystals of 1d formed. Yield ca. 0.2 g
(27%). 31P NMR (C6D6): δ = +17 (t, 1 P), –5 (m, 2 P), –109 (m, 1
P), –155 (m, 2 P), –171 (m, 1 P) ppm. 29Si NMR (C6D6): δ = –8.8
1
1
(SiMe3), –93.9 [d, JSi,P
= (82Ϯ5) Hz], –95.1 [d, JSi,P =
(90Ϯ5) Hz] ppm. 1H NMR (C6D6): δ = 0.42 (9H, SiMe3), 0.39
(9H, SiMe3), 3.1 [1JP,H = (191Ϯ5) Hz, 1H] ppm. C18H55Si8P7
(713.13): calcd. C 30.32, H 7.77; found C: 30.72, H 7.24.
Treatment of 2 and 3 with KOtBu. Synthesis of 2a, 2b, and 3a: For
2a, KOtBu (0.166 g, 1.48 mmol) and 18-crown-6 (0.385 g,
1.48 mmol) were added with a spatula to a solution of 2 (1.16 g,
1.48 mmol) in toluene (20 mL) at room temperature. The solution
immediately turned orange and was stirred for an additional 1 h to
complete the reaction. At –80 °C, crystals of the [K·18-crown-6]+
salt of 2a were obtained, which unfortunately were not suitable for
X-ray structure determination due to their small size. From the 31
P
NMR spectra, a nearly 100% conversion of 2 into 2a could be
deduced. The reaction could also be carried out in thf without the
addition of 18-crown-6. 31P NMR (toluene): δ = –18 (m, 1 P), –32
(m, 1 P), –58 (m, 2 P), –92 (m, 1 P), –188 (m, 2 P) ppm. 29Si NMR
(toluene): δ = –8.5 (SiMe3), δ = –49.0 [1JP,Si = (89Ϯ5) Hz, SiMe]
ppm.
Compound 4: 31P NMR (toluene): δ = 0 (m, 3 P), –84 (m, 1 P),
–169 (m, 3 P) ppm. 29Si NMR (toluene): δ = –36.6, –38.2, –41.0
(SiMe2), –42.0 [1JSi,P = (110.0Ϯ5) Hz, SiMe] ppm. C33H99Si18P7,
(1218.49): calcd. C 32.53, H 8.19; found C 33.77, H 7.96.
Treatment of Hyp3P7 (1) with KOtBu. Synthesis of 1b and 1c:
KOtBu (0.06 g, 0.53 mmol) and 18-crown-6 (0.23 g, 0.53 mmol)
were added to a solution of 1 (0.50 g, 0.52 mmol) in toluene (ca.
20 mL) at a temperature of –60 °C. The solution slowly turned
orange and then red. After 3 h, no more starting material 1 could
be detected by 31P NMR spectroscopy. The solution that contained
a mixture of 1b and 1c was then cooled to –80 °C. After several
days, red crystals of the [K·18-crown-6]+ salt of 1c suitable for X-
ray diffraction were obtained. The reaction could also be per-
formed in thf solution in the absence of 18-crown-6 with identical
results. 31P NMR (toluene): δ = 130, 117, 48, 23, –27, –62, –74,
–170 ppm. 29Si NMR (toluene): δ = 13.0 (SiP2), –8.6 (SiMe3), –19.6
(SiMe3), –86 (SiSi3) ppm.
Compound 2b: KOtBu (0.33 g, 2.96 mmol) and 18-crown-6 (0.77 g,
2.96 mmol) were added to a solution of 2 (1.16 g, 1.48 mmol) in
toluene (20 mL) at room temperature. The solution, which turned
orange immediately, was stirred for another 1 h to complete the
reaction and then stored at –80 °C. After several days, yellow-
orange crystals of the [K·18-crown-6]+ salt of 2b were obtained,
which were suitable for X-ray diffraction experiments. Again, the
reaction was quantitative. 31P NMR (–60 °C, toluene): δ = –38 (m,
1 P), –60 (m, 1 P), –77 (m, 1 P), –83 (m, 1 P), –122 (m, 1 P), –125
(m, 1 P), –224 (m, 1 P) ppm.
Compound 3a: KOtBu (0.046 g, 0.41 mmol) and 18-crown-6 (0.11 g,
0.41 mmol) were added with a spatula to a solution of 3 (0.40 g,
0.41 mmol) in toluene (20 mL) at room temperature. The solution
immediately turned orange and was stirred for another 1 h to com-
plete the reaction. The solution was concentrated by evaporation
of the toluene in vacuo and stored at –80 °C for two weeks. How-
ever, no crystallization could be observed. As deduced from the
31P NMR spectra, the reaction proceeded quantitatively. 31P NMR
(toluene): δ = –7 (m, 1 P), –30 (m, 1 P), –68 (m, 2 P), –88 (m, 1 P),
–193 (m, 2 P) ppm. 29Si NMR (toluene): δ = –12.0 (SiMe3), –50.6
[1JP,Si = (95Ϯ5) Hz, SiPh] ppm.
Treatment of Hyp3P7 (1) with KOtBu in the Presence of Dimethyl-
butadiene. Formation of 1a:
A solution of KOtBu (0.03 g,
0.26 mmol) and 2,3-dimethylbutadiene (DMB; 0.28 g, 3.43 mmol,
13-fold excess amount) in thf (10 mL) was cooled to –70 °C. Then,
a solution of 1 (0.25 g, 0.26 mmol) in thf (10 mL) was added drop-
wise. The reaction mixture turned orange and was stirred at that
temperature for another 3 h. According to the 31P NMR spectrum,
1a formed almost quantitatively. Minor byproducts were 1b and
1c. 31P NMR (thf): δ = –10 (m, 1 P), –50 (m, 1 P), –67 (m, 1 P),
–76 (m, 1 P), –95 (m, 1 P), –188 (m, 2 P) ppm. 29Si NMR (thf): δ
Acknowledgments
1
= –11.9 (SiMe3), –99.9 [d, JSi,P = (101Ϯ5) Hz] ppm.
Treatment of Hyp3P7 (1) with LiOtBu. Synthesis of 1a: A solution
of LiOtBu (0.021 g, 0.26 mmol) in thf (10 mL) was added to a solu-
tion of 1 (0.25 g, 0.26 mmol) in thf (10 mL). The reaction mixture
was stirred at 25 °C for 10 d, and the progress of the reaction was
monitored by 31P NMR spectroscopy. Compound 1 was quantita-
tively converted into the lithium salt of 1a. Even at room tempera-
ture, [Hyp2P7]–Li+ showed no tendency to decompose noticeably
over a period of two weeks. 31P NMR (thf): δ = –10 (m, 1 P), –48
We gratefully acknowledge the financial support by the Fonds zur
Förderung der wissenschaftlichen Forschung, Vienna (FWF;
www.fwf.ac.at) (project P1-9167-N19).
[1] See, for instance, the review “Monocyclic and Polycyclic Phos-
phanes”: M. Baudler, K. Glinka, Chem. Rev. 1993, 93, 1623.
[2] V. Manriquez, W. Hönle, H. G. von Schnering, Z. Anorg. Allg.
Chem. 1986, 539, 95.
[3] W. Hönle, H. G. von Schnering, A. Schmidpeter, G. Burget,
Angew. Chem. Int. Ed. Engl. 1984, 23, 817.
[4] See, for instance: M. Baudler, H. Ternberger, W. Faber, J. Hahn,
Z. Naturforsch., Teil B 1979, 34, 1690; T. Sen, R. Poupko, U.
Fleischer, H. Zimmermann, Z. Luz, J. Am. Chem. Soc. 2000,
122, 889.
(m, 1 P), –70 (m, 1 P), –75 (m, 1 P), –95 (m, 1 P), –190 (m, 2 P)
1
ppm. 29Si NMR (thf): δ = –9.8 (SiMe3), –99.8 [d, JSi,P
=
(100Ϯ5) Hz] ppm.
Treatment of 1a with CF3COOH. Synthesis of 1d: A solution of
CF3COOH (0.12 g, 1.05 mmol) in toluene (10 mL) was added
108
www.eurjic.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2011, 101–109

Heptaphosphanortricyclenes with Oligosilyl Substituents
[5] See, for instance: M. Baudler, W. Faber, J. Hahn, Z. Anorg.
Allg. Chem. 1980, 469, 15; G. Fritz, K. D. Hoppe, W. Hönle,
D. Weber, C. Mujica, V. Manriquez, H. G. von Schnering, J.
Organomet. Chem. 1983, 249, 63; G. Fritz, H. W. Schneider, Z.
Anorg. Allg. Chem. 1990, 584, 12.
[6] G. Fritz, W. Hölderich, Naturwissenschaften 1975, 62, 573.
[7] C. Mujica, D. Weber, H.-G. von Schnering, Z. Naturforsch.,
Teil B 1986, 41, 991.
[8] I. Kovacs, G. Baum, D. Fenske, N. Wiberg, H. Schuster, K.
Karaghiosoff, Z. Anorg. Allg. Chem. 1993, 616, 453.
[9] D. Weber, C. Mujica, H. G. von Schnering, Angew. Chem. 1982,
94, 869.
[22] J. Baumgartner, V. Cappello, A. Dransfeld, K. Hassler, in Sili-
con Chemistry–From the Atom to the Extended Systems (Eds.:
P. Jutzi, U. Schubert), Wiley-VCH, Weinheim, 2004.
[23] G. Tekautz, Thesis, T.U. Graz, 2005.
[24] W. T. K. Chan, F. Garcia, A. D. Hopkins, L. C. Martin, M.
McPartlin, D. S. Wright, Angew. Chem. Int. Ed. 2007, 46, 3084.
[25] C. Marschner, Eur. J. Inorg. Chem. 1998, 221.
[26] W. Hönle, H.-G. v. Schnering, Z. Anorg. Allg. Chem. 1978, 440,
171.
[27] J. Huheey, E. Keiter, R. Keiter, Anorganische Chemie, Walter
de Gruyter, Berlin, 2003.
[28] U. Englich, K. Hassler, K. Ruhland-Senge, F. Uhlig, Inorg.
Chem. 1998, 37, 3532.
[10] B. W. Tattershall, N. L. Kendall, Polyhedron 1994, 13, 1517.
[11] Chemistry, Structure, and Bonding of Zintl Phases and Ions
(Ed.: S. M. Kauzlarich), VCH, Weinheim, 1996.
[12] See, for instance: H. G. von Schnering, in Rings, Chains and
Macromolecules of Main Group Elements (Ed.: A. L. Rhein-
gold), Elsevier, Amsterdam, 1977.
[29] V. Capello, J. Baumgartner, A. Dransfeld, K. Hassler, Eur. J.
Inorg. Chem. 2006, 4589.
[30] For experiments, see, for instance: Y. Xiong, S. Yao, M. Brym,
M. Driess, Angew. Chem. Int. Ed. 2007, 46, 4511; M. Driess,
A. D. Fanta, D. R. Powell, R. West, Angew. Chem. Int. Ed.
Engl. 1989, 28, 1038; M. Driess, Angew. Chem. Int. Ed. Engl.
1991, 121, 9722. Recent theoretical results can be found in: R.
Damrauer, S. E. Pusede, Organometallics 2009, 28, 1289.
[31] See, for instance: P. P. Gaspar, in Reactive Intermediates, vol. 1
(Eds.: M. Jones Jr., R. A. Moss), Wiley, New York, 1978.
[32] P. Noblet, A. Dransfeld, R. Fischer, M. Flock, K. Hassler, J.
Organomet. Chem. 2010, DOI: 10.1016/j.jorganchem.2010.
[33] M. Baudler, R. Heumüller, K. Langerbeins, Z. Anorg. Allg.
Chem. 1984, 514, 7.
[34] SAINTPLUS, Software Reference Manual, Version 6.45,
Bruker-AXS, Madison, WI, 1997–2003.
[35] R. H. Blessing, Acta Crystallogr., Sect. A 1995, 51, 33–38;
SADABS, version 2.1, Bruker AXS, 1998.
[36] G. M. Sheldrick, SHELX97, Programs for Crystal Structure
Analysis, release 97-2, University of Göttingen, Göttingen,
Germany, 1998.
[13] See, for instance: M. Baudler, K. Glinka, Chem. Rev. 1994, 94,
3.
[14] N. Korber, H. G. von Schnering, J. Chem. Soc., Chem. Com-
mun. 1995, 1713.
[15] J. C. Aschenbrenner, N. Korber, Z. Anorg. Allg. Chem. 2004,
630, 31.
[16] G. Fritz, J. Härer, E. Matern, Z. Anorg. Allg. Chem. 1983, 504,
38.
[17] G. Fritz, J. Härer, K. H. Schneider, Z. Anorg. Allg. Chem. 1982,
487, 44.
[18] S. P. Mattamana, K. Promprai, J. C. Fettinger, B. W. Eichhorn,
Inorg. Chem. 1998, 37, 6222.
[19] J. C. Aschenbrenner, N. Korber, Z. Anorg. Allg. Chem. 2004,
630, 31.
[20] H. Schmidbaur, A. Bauer, Phosphorus Sulfur Silicon Relat.
Elem. 1995, 102, 217.
[21] H. Siegl, W. Krumlacher, K. Hassler, Monatsh. Chem. 1999,
130, 139.
Received: July 9, 2010
Published Online: November 18, 2010
Eur. J. Inorg. Chem. 2011, 101–109
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
109