Strained metallocenophanes with late transition metals in the bridge: Syntheses and structures of nickel- and platinum-bridged [1]ferrocenophanes

Article abstract of DOI:10.1002/anie.200705672

(Chemical Equation Presented) Late and unexpected: The first strained metallocenophanes with a single late transition-metal atom in the bridge have been prepared (see scheme). The nickeland platinum-bridged complexes are the first [1]ferrocenophanes where the bridging atom has a distorted square-planar geometry, and the compounds are substantially more tilted and electron-rich than early transition-metal analogues.

Full text of DOI:10.1002/anie.200705672

Communications
DOI: 10.1002/anie.200705672
Metallocenophanes
Strained Metallocenophanes with Late Transition Metals in the Bridge:
Syntheses and Structures of Nickel- and Platinum-Bridged
[1]Ferrocenophanes
George R. Whittell, Benjamin M. Partridge, Oliver C. Presly, Christopher J. Adams, and
Ian Manners*
Until recently, few classes of strained rings containing
transition elements had been explored, but these interesting
species are now attracting substantial attention.[1] In partic-
ular, strained metallocenophanes (1) and related species
to strained, unexpectedly electron-rich species containing
square-planar d8 centers.
The scarcity of metal-bridged [1]ferrocenophanes is at
first sight somewhat surprising, given that the main synthetic
route to these molecules would appear applicable to any
element for which a dichloride is available.[2a] We have found,
however, that reaction of many transition-metal dihalides
with dilithioferrocene–tmeda complex (tmeda = N,N,N’,N’-
tetramethylethylenediamine) results in reduction of the metal
center. For instance, treatment of Li2[Fe(h5-C5H4)2]·tmeda
(7)[9] with trans-[NiCl2(PMe3)2],[10] in hexane at À208C, only
leads to the precipitation of metallic nickel. In striking
contrast, the analogous reaction of 7 with trans-[NiCl2-
(PnBu3)2] (8, Scheme 1)[11] resulted in a cloudy, dark red
containing p-hydrocarbon ligands (e.g. 2–5)[2] have been the
focus of recent investigations due to the fascinating questions
that surround their electronic structures, bonding, and
reactivity. Furthermore, the ability of these compounds to
yield functional metallopolymers by ring-opening polymeri-
sation (ROP) has provided an extra impetus to the field.[3]
The nature of the transannular bridge, ERx, in such species is
of pivotal importance as the size and geometric preferences of
the bridging element directly influence the strain present.
Moreover, in the ring-opened polymers 6 the metal–metal
interactions are modulated by the spacer unit.[4] To date, most
work in the area of strained metallorings has involved
[1]ferrocenophanes (e.g. 1, M = Fe) containing bridging non-
metallic main group elements E such as Si, P, and more
recently S and B.[2a] The incorporation of main group metals
(Sn,[5] Al,[2c,6] and Ga[2c,7]) and early d-block metals (Ti, Zr,
and Hf)[8] has also been successful. However, to date no
transition metals other than the aforementioned Group 4 d0
tetrahedral centers have been successfully introduced into the
bridge. Herein, we report examples of [1]ferrocenophanes
with the Group 10 metals Ni and Pt in the bridge, which leads
Scheme 1. Preparation of metal-bridged [n]ferrocenophanes.
solution with a single new product detected by 31P{1H} NMR
spectroscopy (d = 13.5 ppm, ca. 20% conversion). Work-up
afforded a dark red crystalline solid in modest yield (11%)
that was subsequently characterized as the first nickel-bridged
[1]ferrocenophane [{Fe(h5-C5H4)2}Ni(PnBu3)2] (9).[12] It is of
interest that the NiII component readily undergoes a trans-to-
cis isomerisation during the reaction and that no product
containing the original trans configuration was observed.
1
Compound 9 was fully characterized by H, 13C{1H}, and
31P{1H} NMR spectroscopy. It is noteworthy that all three
spectra displayed resonances at predictable chemical shifts
and as such were consistent with the incorporation of a
diamagnetic square-planar nickel(II) center. The 1H NMR
spectrum of 9 is typical of [1]ferrocenophanes with the two
sets of cyclopentadienyl protons experiencing markedly
different shielding (Dd = 0.3 ppm).[2a] Further indication of
the strained nature of 9 can be taken from the 13C{1H} NMR
spectrum. The ipso-carbon resonance of the carbocyclic rings
(d = 54.1 ppm) appears at higher field than the signal
observed for ferrocene (d = 69 ppm).[13] This is attributed to
an increase in sp3 character of the ipso-carbons, which results
from a strain-induced distortion of these atoms away from the
initial trigonal-planar geometry. For the only other transition-
[*] Dr. G. R. Whittell, B. M. Partridge, Dr. O. C. Presly, Dr. C. J. Adams,
Prof. I. Manners
School of Chemistry, University of Bristol
Cantock’s Close, Bristol BS81TS (UK)
Fax : (+44)117-929-0509
E-mail: ian.manners@bristol.ac.uk
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
4354
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 4354 –4357

Angewandte
Chemie
Table 1: Summary of pertinent structural data for metal-bridged [1]ferro-
cenophanes 9 and 11.
metal-bridged [1]ferrocenophanes to have been prepared, the
reverse effect was reported with [{Fe(h5-C5H4)2}M(h5-
C5H4tBu)2] (M = Ti, Zr and Hf) exhibiting 13C{1H} resonances
at d = 179.8, 159.0, and 153.3 ppm, respectively.[8] The solid-
state structure of the sole crystallographically characterized
example (M = Zr), however, suggested that these were only
modestly strained (a = 68). The lowest-energy electronic
transition of ferrocene (440 nm)[14] has been shown to be
sensitive to the degree of ring-tilt and the formation of
[1]ferrocenophanes is classically accompanied by a bath-
ochromic shift of this band.[2a] The UV/Vis spectrum of 9
exhibited a corresponding transition at 478 nm and thus is
Parameter
M=Ni (9)
M=Pt (11)
a [8]
b [8]
28.4
25.5
25.2
26.4
25.9
26.2
d [8]
q [8]
157.3
80.7(2)
1.62
159.4
78.36(16)
1.64
Fe–Cp(centroid) [Š]
1.63
Fe···M [Š]
3.03
3.14
consistent with the formation of
a [1]ferrocenophane.
Unfortunately, no molecular ion peak was observed when 9
was subjected to EI-MS.
the solid state. It is noteworthy that both 9 and 11 possess
higher tilts than the only other crystallographically charac-
terized [1]ferrocenophane to contain a transition metal in the
bridge, [{Fe(h5-C5H4)2}Zr(h5-C5H4tBu)2] (a = 68).[8] This
clearly arises from a combination of the more constraining
square-planar geometries of the Group 10 metals and the
contraction in covalent radius that occurs on traversing a
period of the d-block. The Fe–Cp(centroid) distances (1.62–
1.64 Š) are comparable to those found in the sila[1]ferroce-
nophane, [Fe(h5-C5H4)2SiMe2] (1: M = Fe, ERx = SiMe2)
(both 1.63 Š).[19] This parameter, however, has been shown
To investigate the generality of the synthetic procedure,
cis-[PtCl2(PnBu3)2][15] (10) was treated with 7. This experi-
ment also afforded
a metal-bridged [1]ferrocenophane,
namely [{Fe(h5-C5H4)2}Pt(PnBu3)2] (11), although in this
instance the availability of a cis-dichloride precluded the
need for isomerisation at the bridging metal center. It
therefore appears that formation of a [1]ferrocenophane is
not exclusive to nickel but can be extended to platinum. The
NMR and UV/Vis spectroscopic data for 11 does not differ
significantly from that for 9 and therefore will not be
discussed further. The observation of a platinum–phosphorus
coupling constant [1J(Pt,P) = 2182 Hz], however, is notewor-
thy as it is consistent with the presence of a square-planar PtII
center.[16]
À
to be relatively insensitive to the strength of Fe Cp bonding.
In cases where the Cp rings are more firmly bound,
deformation occurs preferentially at the Cp ipso-carbon
atoms.[20] This is due to the higher energy penalty associated
with tilting the Cp rings and is manifested structurally in
higher b and lower a angles. In 9 and 11 the converse is true,
with the b angles proving considerably less marked (25.5 and
25.98 for 9, and 26.4 and 26.28 for 11) than in 1 (M = Fe, ERx =
Both compounds 9 and 11 were characterized by single
crystal X-ray diffraction (see Figure 1 for 9) and all pertinent
structural parameters are included in Table 1.[17] Upon cooling
concentrated hexane solutions to À408C, the compounds
crystallized as red blocks in the monoclinic space group P21/c.
The crystals were isomorphous and contained one molecule
per asymmetric unit with no short intermolecular interactions.
By virtue of the constraining ansa-bridge, the Cp ligands
subtended an angle (a) of 28.48 and 25.28 in 9 and 11,
respectively. As would be expected given the increased
covalent radius[18] of platinum (1.29 Š) with respect to
nickel (1.15 Š), compound 11 exhibits the lower a angle in
SiMe2) [both 37.0(6)8].[19] It would therefore appear that Fe
À
Cp bonding is in fact weaker in the former compounds. The
geometry about the nickel and platinum centers is best
described as distorted square-planar with cis phosphine
ligands. It is noteworthy that these represent the first
ferrocenophanes to contain an element with a square-planar
geometry in the ansa-bridge. The angle subtended by the two
À
Ni C(ipso) bonds (q) [80.7(2)8] is smaller than the ideal angle
(908). When taken together, these observations suggest that it
is distortion of the iron coordination environment, and not the
ipso-carbons or Group 10 center, that mainly accommodates
the strain of the system.
Compounds 9 and 11 were subjected to a cyclic voltam-
metric study in order to probe any electronic effect that the
bis(phosphine)metal(II) fragments may have on the ferro-
cene/ferrocenium redox couple. The platinum-containing
compound 11 exhibited an electrochemically reversible
redox couple, as was previously reported for sila[1]-,[20]
germa[1]-,[21] and stanna[1]ferrocenophanes,[5c] whereas oxi-
dation proved irreversible in the case of 9. The most striking
feature of each voltammogram was the oxidation potential,
which occurred at À0.821 (Ep) and À0.682 V (E8) versus the
ferrocene/ferrocenium redox couple (scan rate of 250 mVsÀ1
)
for compounds 9 and 11, respectively. Despite an appreciable
tilt angle [20.8(5)8],[19] the half-wave oxidation potential for 1
(M = Fe, ERx = SiMe2) is identical to that of ferrocene and,
thus, it would appear that this property is insensitive to
Figure 1. Left: Molecular structure of 9 as determined by single crystal
X-ray diffraction (thermal ellipsoids are set at 50% probability). All
hydrogen atoms have been omitted for clarity. Right: Definition of
parameters used to quantify distortion in [1]ferrocenophanes.
Angew. Chem. Int. Ed. 2008, 47, 4354 –4357ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4355

Communications
distortion of the ferrocenediyl group. Cathodic shifts have,
Keywords: iron · metallacycles · nickel · platinum ·
ring-opening polymerization
.
however, been observed upon increased methylation of the
Cp rings,[20] but these have always been attributed to an
increase in electron-density at the iron center. Furthermore,
such changes have been correlated with decreasing a and
increasing b angles, whereas analysis of the solid-state
structures here shows the converse to be true. It is conceivable
that a dative interaction from the bis(phosphine)metal(II)
fragment to the iron center may stabilize the resulting
iron(III) complex relative to the ferrocenium ion, which
could lead to the observed shift in the redox potential. The
spatial separation between the two metal centers in each
compound (Table 1), however, is at least 27% larger than the
sum of the covalent radii. It is noteworthy that in 1 (M = Fe,
ERx = SiMe2) the Fe···Si interaction is also larger than the
sum of covalent radii (by 18%), although the 57Fe Mössbauer
spectrum suggested a possible dative interaction between
silicon and iron.[20] This feature, however, did not result in the
expected anodic shift of the oxidation potential and thus it
difficult to invoke such an argument for the marked change in
E8 for 11 given the larger effective interatomic separation.
From the X-ray data, the two new metal-bridged [1]fer-
rocenophanes 9 and 11 appear to be highly strained species
that should undergo ring-opening reactions. In the initial
studies, given that the large a angle suggested a weakening of
[1] a) C. Laporte, C. Böhler, H. Schönberg, H. Grützmacher, J.
Organomet. Chem. 2002, 641, 227; b) C. Laporte, T. Büttner, H.
Rüegger, J. Geier, H. Schönberg, H. Grützmacher, Inorg. Chim.
Acta 2004, 357, 1931; c) C. Li, M. E. Cucullu, R. A. McIntyre,
E. D. Stevens, S. P. Nolan, Organometallics 1994, 13, 3621; d) L.
Luo, N. Zhu, N.-J. Zhu, E. D. Stevens, S. P. Nolan, Organo-
metallics 1994, 13, 669; e) E. Lozano, M. Nieuwenhuyzen, S. L.
James, Chem. Eur. J. 2001, 7, 2644; f) N. L. S. Yue, M. C.
Jennings, R. J. Puddephatt, Dalton Trans. 2006, 3886; g) metal-
locyclopropanes have been well-explored but in many cases the
unstrained metal–olefin interaction resonance structure contrib-
utes significantly, see M. L. Steigerwald, W. A. Goddard III, J.
Am. Chem. Soc. 1985, 107, 5027.
[2] For a recent review see a) D. E. Herbert, U. F. J. Mayer, I.
Manners, Angew. Chem. 2007, 119, 5152; Angew. Chem. Int. Ed.
2007, 46, 5060; for specific examples see b) U. Vogel, A. J.
Lough, I. Manners, Angew. Chem. 2004, 116, 3383; Angew.
Chem. Int. Ed. 2004, 43, 3321; c) C. L. Lund, J. A. Schachner,
J. W. Quail, J. Mueller, Organometallics 2006, 25, 5817; d) C. L.
Lund, J. A. Schachner, J. W. Quail, J. Müller, J. Am. Chem. Soc.
2007, 129, 9313; e) A. Berenbaum, I. Manners, Dalton Trans.
2004, 2057; f) H. Braunschweig, M. Homberger, C. Hu, X.
Zheng, E. Gullo, G. Clentsmith, M. Lutz, Organometallics 2004,
23, 1968; g) C. Elschenbroich, E. Schmidt, R. Gondrum, B. Metz,
O. Burghaus, W. Massa, S. Wocadlo, Organometallics 1997, 16,
4589; h) H. Braunschweig, T. Kupfer, K. Radacki, Angew. Chem.
2007, 119, 1655; Angew. Chem. Int. Ed. 2007, 46, 1630; i) M.
Tamm, A. Kunst, T. Bannenberg, S. Randoll, P. G. Jones,
Organometallics 2007, 26, 417; j) H. Braunschweig, M. Lutz, K.
Radacki, A. Schaumlöffel, F. Seeler, C. Unkelbach, Organo-
metallics 2006, 25, 4433; k) H. Braunschweig, T. Kupfer, M. Lutz,
K. Radacki, J. Am. Chem. Soc. 2007, 129, 8893; l) A. Bartole-
Scott, H. Braunschweig, T. Kupfer, M. Lutz, I. Manners, T.-l.
Nguyen, K. Radacki, F. Seeler, Chem. Eur. J. 2006, 12, 1266;
m) P. Chadha, J. L. Dutton, M. J. Sgro, P. J. Ragogna, Organo-
metallics 2007, 26, 6063.
[3] a) G. R. Whittell, I. Manners, Adv. Mater. 2007, 19, 3439; b) V.
Bellas, M. Rehahn, Angew. Chem. 2007, 119, 5174; Angew.
Chem. Int. Ed. 2007, 46, 5082; c) I. Manners, Chem. Commun.
1999, 857.
[4] a) D. A. Foucher, C. H. Honeyman, J. M. Nelson, B. Z. Tang, I.
Manners, Angew. Chem. 1993, 105, 1843; Angew. Chem. Int. Ed.
Engl. 1993, 32, 1709; b) R. Rulkens, D. P. Gates, D. Balaishis,
J. K. Pudelski, D. F. McIntosh, A. J. Lough, I. Manners, J. Am.
Chem. Soc. 1997, 119, 10976; c) G. E. Southard, M. D. Curtis,
Organometallics 2001, 20, 508.
[5] a) R. Rulkens, A. J. Lough, I. Manners, Angew. Chem. 1996, 108,
1929; Angew. Chem. Int. Ed. Engl. 1996, 35, 1805; b) F. Jäkle, R.
Rulkens, G. Zech, D. A. Foucher, A. J. Lough, I. Manners,
Chem. Eur. J. 1998, 4, 2117; c) H. K. Sharma, F. Cervantes-Lee,
J. S. Mahmoud, K. H. Pannell, Organometallics 1999, 18, 399.
[6] J. A. Schachner, C. L. Lund, J. W. Quail, J. Müller, Organo-
metallics 2005, 24, 785.
À
the Fe Cp bonds, we explored the use of NaCp-initiated
photolytic ROP.[22] This process proceeds through the cleav-
À
age of Fe Cp bonds, and as such differs from the “dark”
variation, initiated with alkyl or aryl lithiums, where the
[23]
À
bridging atom Cp bond is broken.
However, under the
same conditions that proved successful for the photolytic
ROP or stoichiometric ring-opening of 1 (M = Fe, ERx =
SiMe2), compound 11 remained unchanged as evidenced by
31P{1H} NMR spectroscopy. Further attempts to ring-open 11
by reaction with PhLi and nBuLi also resulted in no reaction,
even when 11 was stirred at 258C with 10 equivalents of
nBuLi for 18 h. We investigated whether compound 11 could
be ring-opened thermally. After heating in the melt at 1508C
for 18 h, 31P{1H} NMR spectroscopy indicated that the starting
material was completely unchanged. Further heating at 2108C
for 12 h, however, resulted in a brown solid forming from the
red melt. Analysis of this material by 31P{1H} NMR spectros-
copy revealed dissociated PnBu3 to be the only phosphorus-
containing component.
In conclusion, the application of the PnBu3 ligand has
enabled the synthesis and characterization of the novel d8
metal-bridged [1]ferrocenophanes, 9 and 11, which display
substantial structural differences from early d0 transition-
metal analogues and unexpected redox properties. Com-
pound 11 decomposed through phosphine dissociation prior
to undergoing any well-defined ring-opening chemistry. We
are currently investigating the use of stronger donors such as
bidentate diphosphines to circumvent the problems associ-
ated with ligand loss and prepare mixed-metal polymetallo-
cenes by ROP.
[7] J. A. Schachner, C. L. Lund, J. W. Quail, J. Müller, Organo-
metallics 2005, 24, 4483.
[8] R. Broussier, A. Da Rold, B. Gautheron, Y. Dromzee, Y.
Jeannin, Inorg. Chem. 1990, 29, 1817.
[9] D. A. Rider, K. A. Cavicchi, K. N. Power-Billard, T. P. Russell, I.
Manners, Macromolecules 2005, 38, 6931.
Received: December 11, 2007
Revised: January 17, 2008
Published online: April 29, 2008
[10] O. Dahl, Acta Chem. Scand. 1969, 23, 2342.
[11] J. P. Ducoux, P. Le Menez, N. Kunesch, E. Wenkert, J. Org.
Chem. 1993, 58, 1290.
4356
www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 4354 –4357

Angewandte
Chemie
[12] Full synthetic procedures, spectroscopic characterization and
electrochemical data can be found in the Supporting Informa-
tion.
[13] P. C. Lauterbur, J. Am. Chem. Soc. 1961, 83, 1838.
[14] Y. S. Sohn, D. N. Hendrickson, H. B. Gray, J. Am. Chem. Soc.
1971, 93, 3603.
[15] G. B. Kauffman, L. A. Teter, Inorg. Synth. 1963, 7, 24.
[16] S. O. Grim, R. L. Keiter, W. McFarlane, Inorg. Chem. 1967, 6,
1133.
eter: Bruker Smart Apex CCD; MoKa radiation, 100(2) K,
2qmax = 100.6308; 39427 reflections, 8073 unique (Rint = 0.0673),
direct methods; absorption correction SADABS (m =
4.498 mmÀ1); refinement (against F2o) with SHELXTL-97, 349
parameters, 0 restraints, R1 = 0.0382 (I > 2s) and wR2 = 0.0772
(all data), Goof = 1.080, max/min residual electron density:
1.397/À1.692 eŠÀ3. CCDC 670448 (9) and CCDC 670449 (11)
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.
[17] Crystal structure data for 9: red blocks from hexane;
C34H62FeNiP2, monoclinic, space group P21/c; a = 18.170(4),
b = 9.5255(19), c = 20.973(4) Š, a = 90, b = 107.40(3), g = 908,
V= 3463.9(12) Š3; Z = 4, 1calcd = 1.241 gcmÀ3; crystal dimen-
sions: 0.2 ” 0.1 ” 0.05 mm; diffractometer: Bruker Smart Apex
CCD; MoKa radiation, 100(2) K, 2qmax = 55.968; 38899 reflec-
tions, 7936 unique (Rint = 0.1421), direct methods; absorption
correction SADABS (m = 1.075 mmÀ1); refinement (against Fo2)
with SHELXTL-97, 349 parameters, 0 restraints, R1 = 0.0827 (I >
2s) and wR2 = 0.1693 (all data), Goof = 1.143, max/min residual
electron density: 0.589/À1.013 eŠÀ3. Crystal structure data for
11: red blocks from hexane; C34H62FeP2Pt, monoclinic, space
group P21/c; a = 18.301(4), b = 9.6527(19), c = 20.902(4) Š, a =
[18] J. G. Stark, H. G. Wallace, Chemistry Data Book, 2nd ed., John
Murray, London, 1982.
[19] W. Finckh, B. Z. Tang, D. A. Foucher, D. B. Zamble, R.
Ziembinski, A. Lough, I. Manners, Organometallics 1993, 12,
823.
[20] J. K. Pudelski, D. A. Foucher, C. H. Honeyman, A. J. Lough, I.
Manners, S. Barlow, D. OꢀHare, Organometallics 1995, 14, 2470.
[21] M. Castruita, F. Cervantes-Lee, J. S. Mahmoud, Y. Zhang, K. H.
Pannell, J. Organomet. Chem. 2001, 637–639, 664.
[22] a) M. Tanabe, I. Manners, J. Am. Chem. Soc. 2004, 126, 11434;
b) M. Tanabe, G. W. M. Vandermeulen, W. Y. Chan, P. W. Cyr, L.
Vanderark, D. A. Rider, I. Manners, Nat. Mater. 2006, 5, 467.
[23] Y. Ni, R. Rulkens, I. Manners, J. Am. Chem. Soc. 1996, 118, 4102.
90, b = 107.77(3), g = 908, V= 3516.4(12) Š3; Z = 4, 1calcd
1.480 gcmÀ3; crystal dimensions: 0.3 ” 0.1 ” 0.05 mm; diffractom-
=
Angew. Chem. Int. Ed. 2008, 47, 4354 –4357ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4357