Tetraphosphacyclopentadienyl and triphosphaallyl ligands in iron complexes

Article abstract of DOI:10.1002/anie.200500194

(Chemical Equation Presented) Different faces of phosphorus chemistry: The reaction of the bicyclotetraphosphine complex 1 with fBuC≡P leads to the first 1,2,3,4-tetraphosphaferrocene complex 2 and an allylic P ₃-bridged complex 3. Cp? = η

Full text of DOI:10.1002/anie.200500194

Angewandte
Chemie
Polyphosphorus Ligands
Tetraphosphacyclopentadienyl and
Triphosphaallyl Ligands in Iron Complexes**
Manfred Scheer,* Shining Deng, Otto J. Scherer, and
Marek Sierka
Dedicated to Professor Gerd Becker
on the occasion of his 65th birthday
Since the first synthesis of [Ti(h5-P5)2]2ꢀ by Ellis et al.[1] the
search for decaphosphaferrocene has been a challenge in the
formation of soluble supramolecular aggregates[10] as well as
field of phosphorus-containing complexes.[2] In addition, the
1D and 2D polymers and networks.[11] Thus, we developed a
synthesis for the first completely characterized 1,2,3,4-tetra-
ꢀ
lower phosphorus-containing complexes with [(RC)nP5ꢀn
]
ligands (n = 1–3) are of particular interest since h5-phospholyl
phosphaferrocene complex D by a method using tBuC P as
ꢁ
complexes show activity in various homogeneous catalysis
starting material, the results of which are reported herein.
applications.[3] Among the series of ferrocene complexes with
The reaction of 1 with tBuC P leads to the desired
ꢁ
ꢀ
[(RC)nP5ꢀn
]
ligands (n = 0–5) the syntheses of A[4] and E[5]
complex in a mixture of products which have been separated
by chromatography [Eq. (1); Cp’’’ = h5-C5H2tBu3]. In each of
these thermolysis reactions a very broad signal appears at d =
have been benchmarks in the development of polyphospholyl
complex chemistry.[6]
Whereas the ferrocenes of type B[7] as well as B’ and C’[8]
were synthesized later using phosphaalkyne as starting
material for B’ and C’, the missing 1,2,3-triphosphaferrocene
C was only recently synthesized by Scherer et al. by the
reaction of the tetraphosphabicyclobutadiene complex 1 with
diphenylacetylene.[9] In the series shown, the hitherto missing
complex D is of interest for us, since our general goal lies in
the use of Pn ligand complexes as linking units for the
92 ppm (w = = 300 Hz) in the 31P NMR spectrum, which
1
2
probably represents a polymeric product. The main products
isolated are complexes 4 and 3, whereas 6 and the known
cyclo-P5 complex 5[12] could only be obtained in small
amounts.[13] The structure of the main products gives evidence
for a P3/P1 fragmentation[14] of the bicyclo-tetraphosphine
complex 1, with subsequent further reaction with one or two
equivalents of phosphaalkyne to give the products 3 and 4,
respectively. Similar observations were made for an indanyl-
substituted derivative of 1 and 1-methylcyclohexylphosphaal-
kyne by Scherer et al.[15]
[*] Prof. Dr. M. Scheer, Dipl.-Chem. S. Deng
Institut fꢀr Anorganische Chemie
Universitꢁt Regensburg
The crystals of complexes 3 and 6 are green and those of 4
are red. The compounds are readily soluble in toluene and
CH2Cl2 and slightly soluble in nonpolar solvents, such as n-
pentane. The mass spectrum of each compound shows the
appropriate molecular ion peak. The 31P{1H} NMR spectrum
of the major product 4 shows two groups of signals of an AM2
spin system, which is very similar to the values found in
complexes of type C’.[8] Furthermore, the molecular structure
of 4[16] (see Supporting Information) is similar to those of
other type C’ derivatives[8] and reveals the staggered con-
formation of the tBu groups of the five-membered rings to
minimize steric repulsion. The rings are tilted by only 0.78.
The 31P NMR spectrum of the target complex 3 reveals an
AA’MM’ spin system at d = 122.8 ppm for the PM and PM’
atoms, and at d = 80.4 ppm for the PA and PA’ atoms.
93040 Regensburg (Germany)
Fax: (+49)941-943-4441
E-mail: mascheer@chemie.uni-regensburg.de
Dr. M. Sierka
Institut fꢀr Chemie
Humboldt-Universitꢁt zu Berlin
Unter den Linden 6, 10099 Berlin (Germany)
Prof. Dr. O. J. Scherer
Fachbereich Chemie
Universitꢁt Kaiserslautern
Erwin-Schrꢂdinger-Strasse, 67663 Kaiserslautern (Germany)
[**] This work was comprehensively supported by the Deutsche
Forschungsgemeinschaft and the Fonds der Chemischen Industrie.
The authors thank Prof. Dr. H. Sitzmann for helpful discussions
about the synthesis of the starting material and Dr. E Matern for the
simulation of the 31P NMR spectra.
Complex 3 crystallizes in the monoclinic space group P21/
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
n with two independent molecules in the unit cell.[16] In
Angew. Chem. Int. Ed. 2005, 44, 3755 –3758
DOI: 10.1002/anie.200500194
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3755

Communications
Figure 1 the top view of both molecules is shown, revealing
ring are 1.763(3) and 2.115(1) ꢀ, respectively, and are there-
ꢀ
that in the solid state both enantiomers are present, these
become equivalent in solution owing to the free rotation of
the five-membered rings.
fore shorter than corresponding single bonds. The P P bonds
in 3 are slightly longer than the averaged bond lengths in the
cyclo-P5 rings of [Cp’’’Fe(h5-P5)] (2.079(8) ꢀ)[12] and [CpxFe-
(h5-P5)] (Cpx = h5-C5Me4Et) (2.096(3) ꢀ)[17] and thus more
comparable to those of the cyclo-P3C2-ring in [Cp’’’Fe(h5-
[9]
ꢀ
ꢀ
P5C3Ph2)] (av. P C 1.775(4) ꢀ, av. P P 2.124(2) ꢀ). Since
each of the angles in the tetraphospholyl ring of 3 deviate
from 1208 (av. P-P-P 102.88, P(1)-C(1)-P(4) 127.1(19)8) the
five-fold symmetry of the ring is distorted, which is similar to
the situation in the P3C2 ring of [Cp’’’Fe(h5-P3C2Ph2)].[9]
Complex 6 contains the first triphosphaallyl unit in a
binuclear complex. To date only cyclo-P3 ligands[18] and
organoallylic EtE3 units (E = P, As)[19] are the only known
phosphorus or allylic groups to bridge two metal centers. The
molecular structure of 6[16] (Figure 3) reveals a cis-orientation
Figure 1. Top-view of the two independent molecules of 3 in the crystal
structure. Hydrogen atoms are omitted for clarity. Selected bond
lengths [ꢃ] and angles [8]: P1-C1 1.759(3), P1-P2 2.115(1), P2-P3
2.115(1), P3-P4 2.116(1), P4-C1 1.769(3), P5-C23 1.766(3), P8-C23
1.759(3), P5-P6 2.1167(13), P6-P7 2.1190(13), P7-P8 2.1152(12); P1-
C1-P4 124.67(18), P1-P2-P3 104.02(5), P2-P3-P4 103.59(5), C1-P1-P2
103.75(11), C1-P4-P3 103.85(11), P5-C23-P8 125.05(17), P5-P6-P7
103.53(7), P6-P7-P8 104.03(5), C23-P5-P6 103.73(11), C23-P8-P7
103.58(11).
In the sandwich complex 3 both five-membered rings are
nearly parallel (Figure 2). The tilting angles between the five-
membered rings are 4.93(2)8 (molecule A) and 5.24(2)8
(molecule B). Since the structural data of both molecules
are similar only molecule A is considered hereafter. The
ꢀ
ꢀ
averaged P C and P P bond lengths of the tetraphospholyl
Figure 3. Molecular structure of 6. Hydrogen atoms are omitted for
clarity. Selected bond lengths [ꢃ] and angles [8]: Fe-Fe 2.589(1), P1(1’)-
P2 2.148(1), Fe-P1 2.225(1), Fe-P2 2.446(1); P1-P2-P1’ 100.71(7), Fe-
P1-Fe’ 71.51(4), Fe-P2-Fe’ 63.92(4).
of the Cp’’’ groups, whereas the allylic P3 unit is directed away
ꢀ
ꢀ
from the Fe Fe axis. The P P bond length of the allylic P3
moiety is 2.148(1) ꢀ and thus shorter than a single bond and
ꢀ
slightly longer than the P P double bond of the EtP3 unit in
[{(tripod)Co}2(m,h3-EtP3)]
(2.110 ꢀ;
tripod = CH3C-
(CH2PPh2)3).[19] The distance between the Fe atoms
(2.589(1) ꢀ) is in the same range as those of the various
[{Cp’Fe(CO)2}2] complexes.[20]
Since the electronic structure of 6 was not clear upon
initial inspection, we decided to carry out density functional
theory (DFT) calculations. Theoretical methods have proven
to be a useful tool in analyzing the structure and properties of
phosphorus-containing transition-metal complexes.[1,2,21] Cal-
culations are performed on the original complex 6 and,
additionally, to analyze the steric influence of the bulky tBu
groups, on a model compound 6a in which the tBu groups are
replaced by H atoms.
Structure optimizations of 6 and 6a yield close lying low-
spin doublet states (Table 1) with an unpaired electron
delocalized over the d orbitals of both Fe atoms (see
Supporting Information). Comparison of the results obtained
Figure 2. Molecular structure of 3 (molecule A). Hydrogen atoms are
omitted for clarity. Selected bond lengths [ꢃ] and angles [8] [mole-
cule B]: P1-C1 1.759(3) [1.766(3)], P1-P2 2.115(1) [2.117(1)], P2-P3
2.115(1) [2.119(1)], P3-P4 2.116(1) [2.115(1)], P4-C1 1.769(3)
[1.759(3)]; P4-Fe1-P3 53.36(3) [53.33(3)], C1-Fe1-P1 45.16(8)
[45.48(8)], P3-Fe1-P1 89.86(3) [89.99(4)], P4-Fe1-P2 89.96(4)
[89.86(3)], P3-Fe1-P2 52.94(3) [53.02(3)], C2-C1-Fe1 138.3(2)
[138.2(2)].
3756
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 3755 –3758

Angewandte
Chemie
Table 1: Comparison of relative energies of different electronic states DE [kJmolꢀ1] and selected
structural parameters ([ꢃ], [8]) for the central Fe2P3 unit of compounds 6a and 6 (BP86 calculations).
groups are two competing effects.
The DFT method may not be
accurate enough to properly de-
scribe the balance of the two inter-
actions, and the energy difference
between the two states is certainly
within its margin for error. The
delocalization of the unpaired elec-
tron over both Fe atoms is the
reason that signals for an AA’X
spin system are observed in the
31P NMR spectrum of 6. However,
the paramagnetism might cause the
large chemical shift differences in
both signals, which is especially
pronounced for the P1 and P1’
atoms.
6a (C2v)[a]
2B1
6 (C2)[a]
2B
Exp
2A2
2A
9.1
2.640
2.219
2.215
2.481
DE
Fe-Fe’
Fe-P1
Fe-P1’
ꢀ3.2
0.0
2.741
2.192
2.192
2.361
0.0
2.832
2.202
2.196
2.365
2.483
2.589
2.225
2.206
2.446
2.215
2.215
2.488
Fe-P2
Fe-C
2.114–2.126
2.189
99.89
68.18
59.86
2.093–2.120
2.272
94.46
77.40
70.95
2.090–2.166
2.191
97.99
73.07
64.28
2.080–2.157
2.263
93.40
80.19
73.57
2.102–2.158
2.148
100.71
71.51
63.92
P2-P1(1’)
P1-P2-P1’
Fe-P1-Fe’
Fe-P2-Fe’
[a] Symmetry point group.
for 6 and 6a indicates that the steric repulsion between tBu
In conclusion, the results have shown that the bicyclo-
tetraphosphine complex 1 is an excellent starting material for
the synthesis of novel polyphosphorus-ligand-containing
metallocenes. With the synthesis of the 1,2,3,4-tetraphospha-
ferrocene the path has been cleared for its application in the
formation of well-defined aggregates by intermolecular
assembly.
groups has an important influence on the electronic structure.
2
2
For 6a calculations yield two almost degenerate A2 and B1
states with the latter one being slightly less stable. These two
states differ by the occupancy of the a2 and b1 orbitals which
are composed of dxz and dxy orbitals of the Fe and Fe’ atoms
(with both Fe atoms located on the y axis). The (b1)2(a2)1
configuration of the 2A2 state facilitates a weak Fe Fe’
ꢀ
bonding interaction. Promoting one electron from the b1 to a2
orbital results in the 2B1 state ((b1)1(a2)2), which weakens the
Fe···Fe’ interaction and facilitates the P3!Fe p donation. In 6
the stability of the 2A and 2B states, which correspond to the
2A2 and 2B1 states of 6a, respectively, is reversed. The 2A state
Experimental Section
All manipulations were performed under an atmosphere of dry
nitrogen using a glovebox and Schlenk technique. Solvents were
purified and degassed by standard procedures.
ꢀ
is now less stable as a result of the stronger Fe Fe’ interaction
ꢁ
tBuC P (0.223 g, 2.23 mmol) was added to a solution of 1 (1.82 g,
which results in a higher repulsion between bulky Cp’’’
ligands.
2.23 mmol) in toluene (150 mL) at room temperature. The reaction
mixture was heated under reflux for 36 h. After removal of all volatile
material in vacuum, the residue was dissolved in dichloromethane
(10 mL) and transferred onto silica gel. Chromatographic work-up on
a silica gel column (40 ꢁ 2.5 cm) eluting with hexane/dichloromethane
(10:1) gave first a red fraction of 6, then a green fraction of 5, an
orange-red fraction of 4, and finally an olive-green fraction of 3. Each
product was recrystallized from n-hexane. A fifth band (green)
contained the proposed polymeric compound mentioned in the main
text. If, as in some cases, the separation was not complete after
column chromatography, the separation was completed in a glovebox
by thin layer chromatography (TLC) using a hexane/dichloromethane
(10:1) solvent mixture.
To rationalize the differences between different states of 6
and 6a we performed an atoms-in-molecules (AIM)[22]
topological analysis of the electronic charge density for both
systems (Figure 4; see Supporting Information). The bonding
of the 2B state of 6 is similar to the 2B1 state of 6a. However,
2
2
ꢀ
the A and A2 states differ because of the absence of the Fe
Fe’ bond in 6 which is a result of the repulsion between bulky
tBu groups.
1
3: Yield: 60 mg (11%); H NMR (250 MHz, C6D6): d = 1.62 (s,
9H), 1.20 (s, 9H), 1.35 (s, 18H), 4.22 ppm (s, 2H); 31P NMR
(250 MHz, C6D6, AA’MM’ spin system): d(PA) = d(PA’) = 81.6 (m,
2P), d(PM) = d(PM’) = 122.8 ppm (m, 2P), 1J(PAPM) = ꢀ431.2 Hz, 2J-
(PAPM’) = 9.4 Hz, 3J(PAPA’) = ꢀ54.7 Hz, 1J(PMPM’) = ꢀ425.2 Hz (simu-
lated values); EI-MS (1008C) m/z (%): 482 [M+] (100), 382
[M+ꢀPCtBu] (71), 121 [C5H4tBu]+ (23.8).
1
Figure 4. Bonding schemes of the compounds 6a (2A2 and 2B1 states)
and 6 (2A and 2B states) from the AIM analysis of electronic charge
densities (Cp denotes h5-C5H5 in 6a and h5-C5H2tBu3 in 6).
4: Yield: 110 mg (20%); H NMR (250 MHz, C6D6): d = 1.25 (s,
9H), 1.24 (s, 18H), 1.36 (s, 18H), 4.58 ppm (s, 2H); 31P NMR
(250 MHz, C6D6): d = 52.6 (t, 1P), 42.7 ppm (d, 2P), 2J(P,P) =
43.9 Hz; EI-MS (608C) m/z (%) 520 [M+] (4), 432 [M+ꢀPtBu]
(71.5), 295 [(C5H3tBu2)FeP2]+, (100), 233 ([Cp’’’]+ (28).
5: Yield: 4 mg (2%); 1H NMR (250 MHz, C6D6): d = 1.08 (s, 9H),
1.21 (s, 18H), 3.9 ppm (s, 2H); 31P NMR (250 MHz, C6D6): d =
165.4 ppm (s, 5P).
The structure parameters calculated for the slightly less
stable 2A state of 6 are in better agreement with experimental
2
data than those calculated for the B state. Thus, the m,h2:h2-
6: Yield: 5 mg (2%); 1H NMR (250 MHz, C6D6): d = 1.22 (s, 9H),
1.30 (s, 18H), 4.16 ppm (s, 2H); 31P NMR (250 MHz, C6D6): d(PA) =
d(PA’) = 677.8 (dd, 1P), d(PX) = ꢀ380.9 ppm (t, 2P), J(PA,PX) =
390.2 Hz, J(PA’,PX) = 32.6 Hz; EI-MS (908C) m/z 671 (%) [M+]
coordination of the allylic P3 ligand is the most probable one.
Results obtained for 6 and 6a show that the weak attractive
Fe···Fe’ interaction and repulsion between the bulky tBu
Angew. Chem. Int. Ed. 2005, 44, 3755 –3758
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3757

Communications
(5.7), 614 [M+ꢀtBu] (8.4), 382 [M+ꢀCp’’’Fe], (27.8). Fifth fraction:
structures were solved by direct methods with the program
SHELXS-97,[30a] and full matrix least-squares refinement on F2
in SHELXL-97[30b] was performed with anisotropic displace-
ments for non-hydrogen atoms. Hydrogen atoms were located in
idealized positions and refined isotropically according to the
riding model. In compound 4 the C atoms of all CH3 groups are
disordered; a merohedral twin could not be ruled out. 3:
C22H38FeP4, Mr = 482.25, crystal dimensions 0.30 ꢁ 0.15 ꢁ
0.05 mm3, monoclinic, space group P21/n (No. 14), a = 8.679(1),
b = 15.985(1) ꢀ, c = 35.579(2) ꢀ, b = 96.14(1)8, T= 150(1) K,
31
1
P NMR (250 MHz, C6D6): d = 92 ppm, w = = 300 Hz.
2
All theoretical calculations were performed using the TURBO-
MOLE program package.[23] We employ the triple zeta plus polar-
ization (TZVP) basis set on all atoms.[24] We considered BP86[25] and
B3LYP[26] exchange-correlation functionals. Comparison of structural
parameters calculated for 6 using both functionals (see Table 1 for
BP86 and the Supporting Information for B3LYP results) shows that
the BP86 provides
a more reliable description. To speed up
calculations with BP86 functional the Coulomb part was evaluated
using the MARI-J method.[27] To confirm that a calculated structure
and density are not the result of symmetry restrictions, the calcu-
lations are performed both by using symmetry and without symmetry.
The most stable spin states have been found employing the pseudo
Fermi smearing technique.[28] All nuclear coordinates have been fully
optimized. Convergence criteria were set to 10ꢀ3 a.u. for the norm of
the gradients and 10ꢀ5 a.u. for energy change. The AIM analysis was
performed with the AIMPAC program.[29]
Z = 8, V= 4907.7(6) ꢀ3, 1calcd = 1.305 Mgmꢀ3
0.881 mmꢀ1, 9640 independent reflexes (Rint = 0.0549, 2qmax
,
m(MoKa) =
=
53.188), 6244 observed with Fo = 4s(Fo), 511 parameters, R1 =
0.0400, wR2 = 0.0971. 4: C27H45FeP3, Mr = 518.39, crystal dimen-
sions 0.30 ꢁ 0.20 ꢁ 0.05 mm3, orthorhombic, space group Pnma
(No. 62), a = 16.361(3) b = 16.933(3) ꢀ, c = 10.102(2) ꢀ, T=
203(2) K, Z = 4, V= 2798.7(10) ꢀ3, 1calcd = 1.230 Mgmꢀ3
, m-
(MoKa) = 0.378 mmꢀ1, 2963 independent reflexes (Rint = 0.0988,
2qmax = 41.728), 2387 observed with Fo = 4s(Fo), 226 parameters,
R1 = 0.0704, wR2 = 0.1913. 6: C34H58Fe2P3, Mr = 671.41, crystal
dimensions 0.16 ꢁ 0.10 ꢁ 0.02 mm3, monoclinic, space group C2/c
(No. 15); a = 10.207(2), b = 15.635(3), c = 22.210(4) ꢀ, b =
Received: January 18, 2005
Published online: May 11, 2005
94.63(3)8, T= 100(1) K, Z = 4, V= 3532.8(12) ꢀ3, 1calcd
=
Keywords: density functional calculations · iron · metallocenes ·
1.262 Mgmꢀ3, m(AgKa) = 0.978 mmꢀ1, 3174 independent reflexes
(Rint = 0.0591, 2qmax = 50.728), 2247 observed with Fo = 4s(Fo);
186 parameters, R1 = 0.0448, wR2 = 0.1039. CCDC-258064 (3),
CCDC-258065 (4), and CCDC-258066 (6) contain the supple-
mentary 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.
[17] O. J. Scherer, T. Brꢂck, G. Wolmershꢃuser, Chem. Ber. 1988, 121,
935.
.
phosphorus · polyphosphorus ligands
[1] E. Urnezius, W. W. Brennessel, C. J. Cramer, J. E. Ellis, P. v. R.
Schleyer, Science 2002, 295, 832.
[2] J. Frunzke, M. Lein, G. Frenking, Organometallics 2002, 21, 3351.
[3] F. Mathey in Phosphorus-Carbon Heterocyclic Chemistry: The
Rise of a New Domain (Ed.: F. Mathey), Pergamon, Oxford,
2001, pp. 767; Homogeneous Catalysis with Organometallic
Compounds, Vol. 1 and 2 (Eds.: B. Cornils, W. A. Herrmann),
VCH, Weinheim, 1996; K. B. Dillon, F. Mathey, J. F. Nixon in
Phosphorus: The Carbon Copy, Wiley, New York, 1998; F.
Mathey, Angew. Chem. 2003, 115, 1616 – 1643; Angew. Chem.
Int. Ed. 2003, 42, 1578 – 1604.
[18] O. J. Scherer, Acc. Chem. Res. 1999, 32, 751 – 762; K. H.
Whitmire, Adv. Organomet. Chem. 1998, 42, 1 – 145.
[19] A. Barth, G. Huttner, M. Fritz, L. Zsolnai, Angew. Chem. 1990,
102, 956 – 958; Angew. Chem. Int. Ed. Engl. 1990, 29, 929 – 931.
[20] M. Scheer, K. Schuster, U. Becker, A. Krug, H. Hartung, J.
Organomet. Chem. 1993, 460, 105 – 110.
[21] M. Lein, J. Frunzke, G. Frenking, Inorg. Chem. 2003, 42, 2504 –
2511.
[22] R. F. W. Bader, Chem. Rev. 1991, 91, 893 – 928.
[23] a) R. Ahlrichs, M. Bꢃr, M. Hꢃser, H. Horn, C. Kꢄlmel, Chem.
Phys. Lett. 1989, 162, 165 – 169; b) O. Treutler, R. Ahlrichs, J.
Chem. Phys. 1995, 102, 346 – 354.
[24] a) A. Schꢃfer, H. Horn, R. Ahlrichs, J. Chem. Phys. 1992, 97,
2571 – 2577; b) A. Schꢃfer, C. Huber, R. Ahlrichs, J. Chem. Phys.
1994, 100, 5829 – 5835; c) K. Eichkorn, F. Weigend, O. Treutler,
R. Ahlrichs, Theor. Chem. Acc. 1997, 97, 119 – 124.
[25] a) A. D. Becke, Phys. Rev. A 1988, 38, 3098 – 3100; b) S. H.
Vosko, L. Wilk, M. Nusair, Can. J. Phys. 1980, 58, 1200 – 1211;
c) J. P. Perdew, Phys. Rev. B 1986, 33, 8822 – 8824; Erratum: J. P.
Perdew, Phys. Rev. B 1986, 34, 7406.
[26] a) A. D. Becke, J. Chem. Phys. 1993, 98, 5648 – 5652; b) C. Lee,
W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785 – 789.
[27] a) K. Eichkorn, O. Treutler, H. ꢅhm, M. Hꢃser, R. Ahlrichs,
Chem. Phys. Lett. 1995, 242, 652; b) M. Sierka, A. Hogekamp, R.
Ahlrichs, J. Chem. Phys. 2003, 118, 9136 – 9148.
[4] F. Mathey, Tetrahedron Lett. 1976, 4155 – 4158.
[5] O. J. Scherer, T. Brꢂck, Angew. Chem. 1987, 99, 59; Angew.
Chem. Int. Ed. Engl. 1987, 26, 59.
[6] F. Mathey, Coord. Chem. Rev. 1994, 137, 1 – 52.
[7] N. Maigrot, N. Avarvari, C. Charrier, F. Mathey, Angew. Chem.
1995, 107, 623 – 625; Angew. Chem. Int. Ed. Engl. 1995, 34, 590 –
592.
[8] a) C. Mꢂller, R. Bartsch, A. Fischer, P. G. Jones, R. Schmutzler,
J. Organomet. Chem. 1996, 512, 141 – 148; b) R. Bartsch, P. B.
Hitchcock, J. F. Nixon, J. Organomet. Chem. 1988, 340, C37 –
C39; c) R. Bartsch, P. B. Hichcock, J. F. Nixon, J. Chem. Soc.
Chem. Commun. 1987, 1146 – 1148; d) C. Mꢂller, R. Bartsch, A.
Fischer, P. G. Jones, J. Organomet. Chem. 1993, 453, C16 – C18.
[9] O. J. Scherer, T. Hilt, G. Wolmershꢃuser, Angew. Chem. 2000,
112, 1484 – 1485; Angew. Chem. Int. Ed. 2000, 39, 1425 – 1427.
[10] B. Junfeng, A. Virovets, M. Scheer, Science 2003, 300, 781 – 782.
[11] B. Junfeng, A. Virovets, M. Scheer, Angew. Chem. 2002, 114,
1808 – 1811; Angew. Chem. Int. Ed. 2002, 41, 1737 – 1740.
[12] O. J. Scherer, T. Hilt, G. Wolmershꢃuser, Organometallics 1998,
17, 4110 – 4112.
[28] P. Nava, M. Sierka, R. Ahlrichs, Phys. Chem. Chem. Phys. 2003,
5, 3372 – 3381.
[13] S. Deng, Diplom thesis, Karlsruhe 2002.
[14] For P3/P1 fragmentation of P4 phosphorus see M. Scheer, K.
Schuster, U. Becker, Phosphorus Sulfur and Silicon 1996, 109–
110, 114 – 141; M. Scheer, U. Becker, Chem. Ber. 1996, 129,
1307 – 1310; M. Scheer, U. Becker, J. Magull, Polyhedron 1998,
17, 1983 – 1989.
[29] F. W. Biegler-Kꢄnig, R. F. W. Bader, T. H. Tang, J. Comput.
Chem. 1982, 3, 317 – 328.
[30] a) G. M. Sheldrick, SHELXS-96, Universitꢃt Gꢄttingen, 1996;
b) G. M. Sheldrick, SHELXL-97, Universitꢃt Gꢄttingen, 1997.
[15] C. Eichhorn, PhD thesis, Kaiserslautern, 2003.
[16] Crystal structure analyses of 3, 4 and 6 were performed on a
STOE IPDS diffractometer with MoKa radiation (l = 0.71073 ꢀ)
for 3 and 6 and AgKa radiation (l = 0.56087 ꢀ) for 4. The
3758
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 3755 –3758