Roof-shaped halide-bridged bimetallic complexes via ring expansion reaction

Article abstract of DOI:10.1021/ic060700q

Binucleating behavior of rigid triptycene-based ligands has been studied. It has been demonstrated that trans-spanned transition-metal mononuclear complexes bearing 1,8-bis(diisopropylphosphino)triptycene (L1) and 1-diisopropylphosphino-8-diphenylphosphinotriptycene (L2) react with an appropriate transition-metal precursor via a ring-expansion pathway to form unusual bimetallic quasi-closed structures. New palladium and rhodium complexes featuring strongly bent (ca. 115°) M₂(μ-Cl₂) cores with very closely spanned metal centers (less than 3 A) have been prepared using the described ring-expansion reaction and have been fully characterized. Despite a constrained arrangement of the binuclear system, halogen bridges in all new compounds were stable in both the solid state and solution showing no tendency for dissociation even in the presence of added Lewis bases. Spontaneous resolution of the dissymmetric Pd₂(μ-Cl)₂Cl ₂(1-diisopropylphosphino-8-diphenylphosphinotriptycene) (2) into enantiopure antipodes is discussed as well.

Full text of DOI:10.1021/ic060700q

Inorg. Chem. 2006, 45, 7010−7017
Roof-Shaped Halide-Bridged Bimetallic Complexes via Ring Expansion
Reaction
Clarite Azerraf, Shmuel Cohen, and Dmitri Gelman*
Department of Organic Chemistry, The Hebrew UniVersity of Jerusalem, Jerusalem, 91904 Israel
Received April 25, 2006
Binucleating behavior of rigid triptycene-based ligands has been studied. It has been demonstrated that trans-
spanned transition-metal mononuclear complexes bearing 1,8-bis(diisopropylphosphino)triptycene (L1) and 1-di-
isopropylphosphino-8-diphenylphosphinotriptycene (L2) react with an appropriate transition-metal precursor via a
ring-expansion pathway to form unusual bimetallic quasi-closed structures. New palladium and rhodium complexes
featuring strongly bent (ca. 115°) M2(µ-Cl2) cores with very closely spanned metal centers (less than 3 Å) have
been prepared using the described ring-expansion reaction and have been fully characterized. Despite a constrained
arrangement of the binuclear system, halogen bridges in all new compounds were stable in both the solid state
and solution showing no tendency for dissociation even in the presence of added Lewis bases. Spontaneous
resolution of the dissymmetric Pd2(µ-Cl)2Cl2(1-diisopropylphosphino-8-diphenylphosphinotriptycene) (2) into enantiopure
antipodes is discussed as well.
Introduction
short-bridged polymetallic structures are of particular interest,
because of a possible cooperation between the metals that
The synthesis of structurally defined polymetallic com-
plexes is a great challenge in inorganic and organometallic
chemistry due to their potential use as enzyme-mimicking1
and conventional2 catalysts, new materials,3 and pharma-
ceuticals.4 Complexes featuring potentially “communicating”
often finds an expression in unique physical and chemical
properties.5 Clearly, the better “communication” can be
achieved when a M2(µ-X2) core exists in the face-to face
(Figure 1a) or strongly bent bridged forms (e.g., Figures
1b-d where the M1X1X2 and M2X1X2 interplanar angle is
lower than 160°), so that axial orbitals of the metal centers
converge in the space produced by X-M-X planes.
Lately, a variety of edge-sharing d8 [M2(µ-X2)X2L2]
complexes have been prepared.6 Although a majority of
structurally characterized bimetallic compounds of this type
possess a planar M2(µ-X2) core, examples of the bent
bimetallic units are known in the literature.7
* To whom correspondence should be addressed. Tel.: +972-2-6584588.
Fax: +972-2-6585345. E-mail: dgelman@chem.ch.huji.ac.il.
(1) (a) Durand, R. R.; Benscosme, C. S.; Collman, J. P.; Anson, F. C. J.
Am. Chem. Soc. 1983, 105, 2710. (b) McCollum, D. G.; Bosnich, B.
Inorg. Chim. Acta 1998, 270, 13. (c) Catalysis by Di- and Polynuclear
Metal Cluster Complexes; Adams, R. D., Cotton, F. A., Eds.; Wiley-
VCH: New York, 1998. (d) Jedrzejas, M. J.; Setlow, P. Chem. ReV.
2001, 101, 607. (d) Gavrilova, A. L.; Bosnich, B. Chem. ReV. 2004,
104, 349. (e) Justice, A. K.; Linck, R. C.; Rauchfuss, T. B. Inorg.
Chem. 2006, 45, 2406.
(2) (a) Qin, C. J.; Gavrilova, A.; Bosnich, B. Pure Appl. Chem. 2001, 73,
221. (b) Severin, K. Chem.sEur. J. 2002, 8, 1514. (c) Multimetallic
Catalysts in Organic Synthesis; Shibasaki, M., Yamamoto, Y., Eds.;
Wiley-VCH: Weinheim, Germany, 2004. (d) Bukhaltsev, E.; Frish,
L.; Cohen, Y.; Vigalok, A. Org. Lett. 2005, 7, 5123. (e) Haas, M.;
Solari, E.; Nguyen, Q. T.; Gautier, S.; Scopelliti, R.; Severin, K. AdV.
Synth. Catal. 2006, 348, 439.
(3) (a) Magnetism: Molecules to Materials IV; Miller, J. S., Drillon, M.,
Eds.; Wiley-VCH: Weinheim, Germany, 2003. (b) Braunstein, P.;
Frison, C.; Oberbeckmann-Winter, N.; Morise, X.; Messaoudi, A.;
Benard, M.; Rohmer, M.-M.; Welter, R. Angew. Chem., Int. Ed. 2004,
43, 6120.
(4) (a) Milkevitch, M.; Storrie, H.; Brauns, E.; Brewer, K. J.; Shirley, B.
W. Inorg. Chem. 1997, 36, 4534. (b) Metcalfe, C.; Webb, M.; Thomas,
J. A. Chem. Commun. 2002, 2026. (c) Tabassum, S.; Bhat, I. H. Trans.
Met. Chem. 2005, 30, 998. (d) Boerner, L. J. K.; Zaleski, J. M. Curr.
Opin. Chem. Biol. 2005, 9, 135.
General trends in the formation of such butterfly-like
dimeric compounds (Figure 1b) and factors governing the
spontaneous bending have been studied theoretically and
(5) (a) Bosnich, B. Inorg. Chem. 1999, 38, 2554. (b) Bitterwolf, T. E.
Coord. Chem. ReV. 2000, 206-207, 416. (c) Szacilowski, K. T.; Xie,
P.; Malkhasian, A. Y. S.; Heeg, M. J.; Udugala-Ganehenege, M. Y.;
Wenger, L. E.; Endicott, J. F. Inorg. Chem. 2005, 44, 6019.
(6) (a) Praefcke, K.; Bilgin, B.; Pickardt, J.; Borowski, M. J. Organomet.
Chem. 1999, 592, 155. (b) Otto, S.; Roodt, A.; Elding, L. I. Dalton
Trans. 2003, 2519. (c) Speiser, F.; Braunstein, P.; Saussine, L.; Welter,
R. Organometallics 2004, 23, 2613. (d) Gusev, O. V.; Peganova, T.
A.; Kalsin, A. M.; Vologdin, N. V.; Petrovskii, P. V.; Lyssenko, K.
A.; Tsvetkov, A. V.; Beletskaya, I. P. J. Organomet. Chem. 2005,
690, 1710. (e) Yu, J. O.; Lam, E.; Sereda, J. L.; Rampersad, N. C.;
Lough, A. J.; Browning, C. S.; Farrar, D. H. Organometallics 2005,
24, 37.
7010 Inorganic Chemistry, Vol. 45, No. 17, 2006
10.1021/ic060700q CCC: $33.50
© 2006 American Chemical Society
Published on Web 07/26/2006

Roof-Shaped Halide-Bridged Bimetallic Complexes
Figure 1. Possible modes of communicating metal centers.
have been well-reviewed by Aullo´n et al.8,9 However, their
synthesis is often fortuitous.
Figure 2. Structure of 1.
“Sterically induced bending” (Figure 1c), which appears
a more rational strategy to the synthesis of closely bridged
bimetallic complexes, was recently documented.9 For ex-
ample, roof-shaped alkyldithiolato-10 or alkyldiphosphido-
bridged11 binuclear complexes have been prepared and
characterized. In these cases, the presence of the alkyl linker
enforces them sterically to adopt the desired conformation.
toluenesulfinyl)phenoxy}-3-oxapentane complexes.15 To the
best of our knowledge, Pd2X2 rhombuses in other structurally
defined ligand-bridged compounds of this type (Figure 1d)
are planar or only slightly distorted from planarity.16
Recently, we reported about the isolation and characteriza-
tion of an unusual quasi-closed chloride-bridged dipalladium
complex 1 bearing a strongly curved 1,8-bis(diisopropy-
lphosphino)triptycene ligand (L1; Figure 2).17a We found that
the Pd2Cl4 core is bent (ca. 122°) and the corresponding Pd‚
‚‚‚Pd distance is as short as 3.036 Å. On the other hand,
incorporation of the Pd2Cl4 unit between the phosphine
donors results in a significant increase in the P‚‚‚P distance
(5.907 Å). This becomes possible owing to a drastic deviation
of the phosphine groups from the planes of the corresponding
aromatic rings (planarity deviation is ca. 0.54 Å) at the
expense of an increased strain in the 10-membered ring.
A priori, because the observed intermetallic distance in 1
is short, a cooperative effect between the two metal centers
may be anticipated if applied as a catalyst. The monoligated
quasi-closed structure may also provide an unrivaled op-
portunity to create a sterically less-demanding active bime-
tallic site (for example, in comparison to a majority of doubly
ligated A-frame complexes).18 Finally, an undersaturated
(intrinsically reactive) catalyst or catalyst precursor may form
owing to the unique ligation mode.
Another possible route for the construction of the desired
bent structures, which also may be classified as “sterically
enforced” bending, could be the utilization of a somewhat
rigid ligand bearing adequately remote donor groups and
capable of binucleation (as represented in Figure 1d). For
instance, a class of trans-chelating ligands (TransPhos,
SpanPhos, etc.)12 may be seen as suitable candidates.
However, only a limited number of structurally characterized
bimetallic compounds (almost exclusively of rhodium)13
exemplify this approach. With regard to dipalladium com-
plexes, only two structures where a bidentate ligand sterically
enforces the Pd2(µ-X2) unit to adopt a bent conformation
have been previously reported. Guzei et al. reported that 1,3-
bis(3,5-dialkylpyrazolyl-1-carbonyl)benzenes form the cor-
responding [Pd2(µ-Cl2)Cl2LkL] compounds featuring a Pd2-
(µ-Cl2) interplanar angle of ca. 138° and a metal-metal
separation of ca. 3.21 Å.14 Similar parameters have been
previously observed by Hambley et al. for 1,5-bis{o-(p-
Although, in our early communication, we reported that
complex 1 could be isolated from a ring-expansion reaction
of the corresponding monometallic palladium precursor
(Scheme 1),17a other aspects of its formation, properties, and
stability, as well as the possibility to apply the described
(7) (a) Garcia-Anton, J.; Pons, J.; Solans, X.; Font-Bardia, M.; Ros, J.
Eur. J. Inorg. Chem. 2002, 3319. (b) Kuang, S.-M.; Fanwick, P. E.;
Walton, R. A. Inorg. Chem. 2002, 41, 1036. (c) Kuang, S.-M.;
Fanwick, P. E.; Walton, R. A. Inorg. Chim. Acta 2002, 338, 219. (d)
O’Keefe, B. J.; Steel, P. J. Organometallics 2003, 22, 1281. (e) Smith,
D. C., Jr.; Lake, C. H.; Gray, G. M. Dalton Trans. 2003, 2950. (f)
Pamplin, C. B.; Rettig, S. J.; Patrick, B. O.; James, B. R. Inorg. Chem.
2003, 42, 4117. (g) Chandrasekaran, P.; Mague, T.; Balakrishna, M.
S. Organometallics 2005, 24, 3780.
(8) Aullon, G.; Ujaque, G.; Lledos, A.; Alvarez, S. Chem.sEur. J. 1999,
5, 1391.
(9) Aullon, G.; Ujaque, G.; Lledos, A.; Alvarez, S.; Alemany, P. Inorg.
Chem. 1998, 37, 804.
(15) Hambley, T. W.; Raguse, B.; Ridley, D. D. Aust. J. Chem. 1985, 38,
1445.
(16) On the basis of the search within Cambridge Structural Database (CSD
version 5.27 with May 2006 updates), only two bent complexes could
be found that correspond to a monoligated Pd2(µ-X2)X2LkL structure
(refs 14 and 15). Examples of the flat complexes of this type can be
found in: (a) Smith, D. C., Jr.; Gray, G. M. Chem. Commun. 1998,
2771. (b) Garcia-Anton, J.; Pons, J.; Solans, X.; Font-Bardia, Me.;
Ros, J. Eur. J. Inorg.Chem. 2002, 12, 3319. (c) Kuang, S.-M.; Fanwick,
P. E.; Walton, R. A. Inorg. Chem. 2002, 41, 1036. (d) Kuang, S.-M.;
Fanwick, P. E.; Walton, R. A. Inorg. Chim. Acta 2002, 338, 219. (e)
Pamplin, C. B.; Rettig, S. J.; Patrick, B. O.; James, B. R. Inorg. Chem.
2003, 42, 4117. (f) Smith, D. C., Jr.; Lake, C. H.; Gray, G. M. Dalton
Trans. 2003, 14, 2950.
(17) (a) Grossman, O.; Azerraf, C.; Gelman, D. Organometallics 2006,
25, 375. (b) Grossman, O.; Gelman, D. Org. Lett. 2006, 8, 1189.
(18) (a) Woodcock, C.; Eisenberg, R. Inorg. Chem. 1984, 23, 4207. (b)
Stockland, R. A., Jr.; Janka, M.; Hoel, G. R.; Rath, N. P.; Anderson,
G. K. Organometallics 2001, 20, 5212. (c) Tsukada, N.; Tamura, O.;
Inoue, Y. Organometallics 2002, 21, 2521. (d) Koshevoy, I. O.;
Grachova, E. V.; Tunik, S. P.; Haukka, M.; Pakkanen, T. A.; Heaton,
B. T.; Iggo, J. A.; Podkorytov, I. S. Dalton Trans. 2004, 3893. (e)
Braun, T.; Steffen, A.; Schorlemer, V.; Neumann, B.; Stammler, H.-
G. Dalton Trans. 2005, 3331.
(10) Mizuta, T.; Aoki, S.; Nakayama, K.; Miyoshi, K. Inorg. Chem. 1999,
38, 4361.
(11) Meij, R.; Stufkens, D. J.; Vrieze, K.; Brouwers, A. M. F.; Overbeek,
A. R. J. Organomet. Chem. 1978, 155, 123.
(12) (a) DeStefano, N. J.; Johnson, D. K.; Lane, R. M.; Venanzi, L. M.
HelV. Chim. Acta 1976, 59, 2674. (b) Kapoor, P. N.; Venanzi, L. M.
HelV. Chim. Acta 1977, 60, 2824. (c) Bessel, C. A.; Aggarwal, P.;
Marschilok, A. C.; Takeuchi, K. J. Chem. ReV. 2001, 101, 1031. (d)
Freixa, Z.; Beentjes, M. S.; Batema, G. D.; Dieleman, C. B.; van
Strijdonck, G. P. F.; Reek, J. N. H.; Kamer, P. C. J.; Fraanje, J.;
Goubitz, K.; van Leeuwen, P. W. N. M. Angew. Chem., Int. Ed. 2003,
42, 1284.
(13) (a) Eisler, D. J.; Puddephatt, R. J. Can. J. Chem. 2004, 82, 1423. (b)
Hierso, J.-C.; Lacassin, F.; Broussier, R.; Amardeil, R.; Meunier, P.
J. Organomet. Chem. 2004, 689, 766. (c) Freixa, Z.; Kamer, P. C. J.;
Lutz, M.; Spek, A. L.; van Leeuwen, P. W. N. M. Angew. Chem., Int.
Ed. 2005, 44, 4358.
(14) Guzei, I. A.; Li, K.; Bikzhanova, G. A.; Darkwa, J.; Mapolie, S. F.
Dalton. Trans. 2003, 715.
Inorganic Chemistry, Vol. 45, No. 17, 2006 7011

Azerraf et al.
Scheme 1. Synthesis of 1 by the Ring-Expansion Reaction
Figure 3. 31P NMR changes during the formation of 1 by the ring expansion.
Scheme 2. Equilibrium between the Monometallic and the Bimetallic
Complexes
ring expansion to the synthesis of other bimetallic com-
pounds, were not clear. In the present report, we wish to
address these issues.
Results and Discussion
Stability of [Pd2(µ-Cl2)Cl2-(1,8-Bis(diisopropylphosphi-
no)triptycene)] (1). Taking into account the constrained and
visually unfavorable arrangement of the bimetallic complex
1, we initially attributed its formation to the difference in
solubility between the monometallic precursor (highly soluble)
and the bimetallic product (sparingly soluble). If indeed this
were the case, the formation of such bimetallic complexes
would be questionable under high-dilution conditions and 1
could remain just a chemical curiosity. To test the stability
of the bent structure, the reaction between the trans-[1,8-
bis(diisopropylphosphino)triptycene]palladium dichloride
(PdCl2L1) and one additional equivalent of the palladium
dichloride bis-acetonitrile complex (Scheme 1) was per-
formed under high-dilution conditions and was followed by
31P{1H} NMR. Fortunately, the conversion of the monome-
tallic palladium precursor into 1 was clean and quantitative
at 45 °C over 24 h: a singlet ascribed to the monometallic
trans-PdCl2L1 (33.6 ppm) was gradually transformed into a
singlet centered at 56.4 ppm, which is attributed to 1 (Figure
3). No formation of byproducts was observed either by 31P
likely to form because of the thermodynamic stability of the
bridge despite a conformational tension in the spanning
ligand.
Curiously, we found that the affinity of L1 to palladium
is so high that it is capable of scavenging palladium salts
1
from solutions containing other metals. For instance, if /2
equivalent of L1 is added to the chloroform solution of NiCl2-
(diglyme), PdCl2(CH3CN)2, and PtCl2(CH3CN)2, a yellow
precipitate of 1 forms after ca. 24 h at 45 °C. On the basis
of the weight of the isolated 1, we were able to remove 84%
of the initial palladium loading from the original solution.
The relative stability of the bimetallic core was tested using
the reaction of 1 with pyridine. Previously, we showed that,
upon dissolution of 1 in coordinating solvents (DMSO, THF,
etc.), the parent monometallic trans-PdCl2L1 formed quan-
titatively. However, when 2 equiv of pyridine were added
to the chloroform solution of 1, no significant changes in
the 31P{1H} NMR spectrum were observed. The disintegra-
tion of the chloride bridge was very sluggish, even after 4 h
at room temperature; though, it can be completed upon the
addition of larger quantities of pyridine (100-fold excess)
and a prolonged reaction time (12 h) (Scheme 2). Interest-
ingly, the only observable signal detected by 31P{1H} NMR
1
or H NMR.
Because, as we reported earlier,17a the monometallic
precursor exhibited an exceptional stability in both the solid
state and solution in the absence of an excessive palladium
chloride (no equilibrium processes have been observed by
31P{1H} NMR at various temperatures), the complex 1
possessing the bent µ-chloride bridged bimetallic core is
7012 Inorganic Chemistry, Vol. 45, No. 17, 2006

Roof-Shaped Halide-Bridged Bimetallic Complexes
Figure 4. Synthesis of 2 by the ring expansion: 31P{1H} NMR spectra of trans-PdCl2(L2) (top) and 2 (bottom).
Scheme 3. Synthesis of 2 by the Ring-Expansion Reaction
Although the environment around each of the palladium
atoms is only slightly distorted from the square-planar
geometry with bond angles ranging within 84.37° and 94.62°,
the bimetallic rhombus is bent: the Pd(1)-Cl(1)-Cl(2) and
Pd(2)-Cl(1)-Cl(2) interplanar angle was calculated as
116.15°, and the Pd(1)-Pd(2) separation is as short as 2.969
Å (Figure 5). Both of these parameters are even smaller than
those observed previously in 1. The yet stronger twist in this
case is likely due to a slightly different trans influence by
electronically dissymmetric donor phosphine groups. For
example, the Pd(2)-Cl(1) bond (trans to the better σ-donor
groupsdialkylphosphine) is longer than the corresponding
Pd(1)-Cl(2) bond (trans to the diarylphosphine group), 2.406
Å versus 2.386, while the Pd(1)-Cl(2) and Pd(2)-Cl(2)
bonds are essentially equal (2.330 and 2.335 Å). Remarkably,
the incorporation of a Pd2Cl4 fragment between the phosphine
donors results in a significant increase in the P‚‚‚P distance
was attributed to the monometallic trans-PdCl2L1 complex.
These results suggest that L1 has a very good donor ability
forming very stable binuclear complexes.
Synthesis and Characterization of the New [M2(µ-Cl2)-
Cl2L] Complexes. The stability exhibited by 1 inspired us
to apply the described ring-expansion reaction to the
synthesis of other bimetallic complexes. Initially, we at-
tempted to employ a dissymmetric 1-diisopropylphosphino-
8-diphenylphosphinotriptycene (L2) to bridge between metal
centers (expectantly, in their bent form).
As in the previous example, we prepared the ligand using
our published procedure17a and allowed it to react with a
stoichiometric amount of PdCl2(CH3CN)2 in deuterated
chloroform at room temperature, resulting in the formation
of the corresponding monometallic trans-PdCl2(L2) complex
quantitatively as was confirmed by 31P{1H} NMR measure-
ments [the complex exhibits a characteristic AB pattern due
to the coupling established by two different phosphorus
nuclei coordinated to the same Pd center (38.06 and 14.47
ppm, JP-P ) 576 Hz)17a]. Then, an additional 1 equiv of
PdCl2(CH3CN)2 was added, and the progress of the reaction
described by Scheme 3 was followed by 31P{1H} NMR. As
in the aforementioned entry, the gradual formation of the
new material was observed (Figure 4).
The 31P{1H} NMR spectrum of the new product displayed
two sharp singlets with resonance frequencies of 32.74 and
56.54 ppm. By analogy with 1 (Figure 3), the low-field shift
of the signals relative to the monometallic complex could
indicate the formation of the desired bimetallic compound.
The absence of phosphorus-phosphorus coupling also
confirms the structuresdifferent phosphine groups do not
chelate the same palladium center.
Suitable single crystals of 2, grown by slow evaporation
of the saturated chloroform solution at room temperature,
were subjected to X-ray diffraction analysis to verify the
structure. We found that the resulting compound indeed
possesses the roof-shaped Pd2(µ-Cl)2Cl2 core. The crystal-
lographic parameters are presented in Table 1.
Figure 5. ORTEP drawing (50% probability ellipsoids) of the representa-
tive enantiomer of 2. Hydrogen atoms and solvent molecules are omitted
for clarity. Selected bond lengths (Å) and angles (deg): Cl(1)-Pd(1)
2.3860(8), Cl(1)-Pd(2) 2.4062(8), Cl(2)-Pd(2) 2.3302(8), Cl(2)-Pd(1)
2.3344(8), Cl(3)-Pd(1) 2.2837(8), Cl(4)-Pd(2) 2.2821(9), P(1)-Pd(1)
2.2179(8), P(2)-Pd(2) 2.2302(8), Pd(1)-Pd(2) 2.9687(4), C(4)-P(1)
1.821(3), C(10)-P(2) 1.838(3); P(1)-Pd(1)-Cl(3) 89.11(3), P(1)-Pd(1)-
Cl(2) 94.62(3), Cl(3)-Pd(1)-Cl(2) 173.84(3), P(1)-Pd(1)-Cl(1) 175.15(3),
Cl(3)-Pd(1)-Cl(1) 91.96(3), Cl(2)-Pd(1)-Cl(1) 84.73(3), P(1)-Pd(1)-
Pd(2) 124.21(2), Cl(3)-Pd(1)-Pd(2) 130.51(3), Cl(2)-Pd(1)-Pd(2) 50.41(2),
Cl(1)-Pd(1)-Pd(2) 52.03(2), P(2)-Pd(2)-Cl(4) 92.14(3), P(2)-Pd(2)-
Cl(2) 93.32(3), Cl(4)-Pd(2)-Cl(2) 173.05(3), P(2)-Pd(2)-Cl(1) 169.56-
(3), Cl(4)-Pd(2)-Cl(1) 91.01(3), Cl(2)-Pd(2)-Cl(1) 84.37(3), P(2)-
Pd(2)-Pd(1) 133.18(2), Cl(4)-Pd(2)-Pd(1) 122.54(2), Cl(2)-Pd(2)-Pd(1)
50.537(19), Cl(1)-Pd(2)-Pd(1) 51.41(2).
Inorganic Chemistry, Vol. 45, No. 17, 2006 7013

Azerraf et al.
Table 1. Crystal Data and Structure Refinement Details for Compounds 2, 2′, and 4
2
2′
4
empirical formula
fw
C40H38Cl10P2Pd2
1147.94
173(1)
0.710 73
orthorhombic
P212121
11.304(1)
15.752(2)
24.733(3)
90°
90°
90°
4403.8(9)
4
1.731
C40H38Cl10P2Pd2
1147.94
173(1)
0.710 73
orthorhombic
P212121
11.310(8)
15.757(1)
24.744(2)
90°
90°
90°
4409.8(6)
4
1.729
C40H52Cl2O3.50P2Rh
927.48
295(1)
0.710 73
triclinic
P1h
12.961(8)
13.725(8)
15.125 (9)
83.6920(10)°
66.6190(10)°
62.3430(10)°
2178.4(2)
2
temp (K)
wavelength (Å)
cryst syst
space group
a (Å)
b (Å)
c (Å)
R (deg)
â (deg)
γ (deg)
V (A3)
Z
calcd density (mg/m3)
1.414
0.988
948
abs coeff (mm-1
F(000)
)
1.526
2280
1.524
2280
cryst size (mm)
θ range, data
0.27 × 0.13 × 0.11
0.28 × 0.14 × 0.10
0.20 × 0.13 × 0.10
1.65 to 25.00°
1.65 to 25.00°
1.89 to 27.00°
collection (deg)
limiting indices
-13 e h e 13
-18 e k e 18
-29 e l e 29
43215
-13 e h e 13
-18 e k e 18
-29 e l e 29
43627
-16 e h e 16
-17 e k e 17
-19 e l e 19
25119
no. of reflns collected
no. of indep reflns
refinement method
7755
7762
9452
full-matrix
least-squares
on F2
full-matrix
least-squares
on F2
full-matrix
least-squares
on F2
no. of data/
7755/0/491
7762/0/491
9452/0/477
restraints/params
goodness of fit on F2
final R indices [I > 2σ(I)]
0.977
0.977
0.933
R1 ) 0.0221,
wR2 ) 0.0475
R1 ) 0.0239,
wR2 ) 0.0479
-0.033(16)
R1 ) 0.0281,
wR2 ) 0.0519
R1 ) 0.0307,
wR2 ) 0.0526
-0.04(2)
R1 ) 0.0375,
wR2 ) 0.0939
R1 ) 0.0548,
wR2 ) 0.0975
-
R indices (all data)
absolute structure
parameter
Largest diff. peak
and hole
0.490 and
0.634 and
0.631 and
-0.306 e Å-3
-0.319 e Å-3
-0.272 e Å-3
compared to that found in the monometallic trans-PdCl2-
(L2)17a (5.766 Å versus 4.502 Å). On the other hand, the
stronger bending of the Pd2Cl4 unit in 2 relative to that in 1
results in some reduction of this value (5.766 Å versus 5.907
Å). The terminal chlorine atoms, as expected, adapt a syn
conformation with Pd(1)-Cl(3) and Pd(2)-Cl(4) distances
of 2.282(8) and 2.284(8) Å, respectively. This and other
structural parameters are within the normal range.
Intriguingly, we found that chiral but racemic 2 crystallizes
in the acentrosymmetric space group P212121 (Table 1).19 If
so, its crystals should consist of one of two enantiomorphs.
Therefore, other crystals were analyzed as well. Flack
parameters calculated for enantiomeric crystal structures were
-0.033 and -0.04, illustrating that the separation of the two
enantiomers spontaneously occurred during the crystal-
lization.19c
[Rh2Cl2(CO)2L1] (4) can be efficiently prepared using the
ring-expansion reaction. The existence of bent Rh2(µ-Cl)2X4
complexes is not surprising: as was shown theoretically,
rhodium halide dimers are mostly bent because of the relative
diffuse nature of their atomic orbitals, in particular, if bearing
strong σ-donor ligands.20 Moreover, some quasi-closed
chloride-bridged compounds of the described type have been
structurally characterized.13 However, to the best of our
knowledge, all of these compounds were prepared kinetically
starting from the corresponding rhodium dimer.13,21 There-
fore, it was fascinating to attempt this synthesis using the
ring-expansion reaction.
To this end, we prepared and fully characterized the
monometallic RhCl(CO)L1 (3)22 compound to be used as a
(20) (a) Schnabel, R. C.; Roddick, D. M. Inorg. Chem. 1993, 32, 1513.
(b) Hofmann, P.; Meier, C.; Hiller, W.; Heckel, M.; Riede, J.; Schmidt,
M. U. J. Organomet. Chem. 1995, 490, 51.
(21) The following manuscripts report about the isolation of the halide-
bridged quasi-closed bimetallic species from the reaction mixtures,
apparently, as a result of a ligand elimination process and, thus, may
be seen as exceptions: (a) Thomas, C. M.; Mafua, R.; Therrien, B.;
Rusanov, E.; Stoeckli-Evans, H.; Su¨ss-Fink, G. Chem.sEur. J. 2002,
8, 3343. (b) Thomas, C. M.; Su¨ss-Fink, G. Coord. Chem. ReV. 2003,
243, 125.
(22) Synthesis of several rhodium complexes bearing triptycene-based
ligands have been reported in the past: Ahlers, W.; Roeper, M.;
Hofmann, P.; Warth, D. C. M.; Paciello, R. PCT Int. Appl. 2001, WO
2001058589. However, no detailed structural information has been
provided by the patent record.
Figure 5 represents the ORTEP drawings for (9R,10S)
isomer of 2.
The synthesis of bimetallic complexes of other metals was
sought as well. For example, we found that bimetallic
(19) (a) Jacques, J. J.; Collet, A.; Wilen, S. H. Enantiomers, Racemates
and Resolutions; Wiley: New York, 1981. (b) Collet, A.; Brienne,
M.-J.; Jacques, J. Chem. ReV. 1980, 80, 215. (c) To exclude the
possibility of an incorrect determination of the absolute configuration
of 2, the structure was inverted and the Flack parameter was
recalculated. A Flack parameter of 1.0263 was found for the
intentionally inverted (wrong) configuration of 2.
7014 Inorganic Chemistry, Vol. 45, No. 17, 2006

Roof-Shaped Halide-Bridged Bimetallic Complexes
Scheme 4. Synthesis of 4 by the Ring-Expansion Reaction from 3 or from L1 and the Corresponding Dimer
precursor to the bimetallic compound. The synthesis was
accomplished through the mixing of 1 equiv of L1 with /2
than Rh(1)-Cl(2) and Rh(2)-Cl(2) distances (2.41 Å vs 2.39
1
Å) and fall within the normal range. This difference can most
likely be attributed to the slightly stronger trans influence
of the phosphine over the carbonyl ligand.
equiv of Rh2Cl2(CO)4 in chloroform. 31P{1H} NMR mea-
surements indicated that the complexation of rhodium with
the ligand was complete within 45 min at room temperature.
The 31P{1H} NMR spectrum of complex 3 displays a doublet
with a resonance frequency of 39.16 ppm and a Rh-P
coupling constant of 127 Hz. As was observed previously
for monometallic palladium complexes, the signal ascribed
to the central methine hydrogen atom appears in the 1H NMR
spectrum at a very low field (9.11 ppm), suggesting a close
proximity of the hydrogen to the metal center. Unfortunately,
we do not have the X-ray evidence for the trans-spanned
geometry of the precursor 3; however, the splitting of the
phosphorus-coupled carbon signals into 1:2:1 triplets in the
13C NMR spectra of 3 may be a good indication of this
configuration. Numerous compounds forming AXX′ spin
systems, where the two phosphorus nuclei couple with a large
coupling constant (characteristic to the trans diphosphine
complexes),23 feature similar coupling.24
The deformation of the Rh2Cl2 rhombus is very strong:
the interplanar angle between the Rh(1)-Cl(1)-Cl(2) and
Rh(1)-Cl(1)-Cl(2) planes is 114.58°, while the Rh-Rh
separation is normal compared to the previously published
syn- and anti-Rh2(µ-Cl2)(CO)2 quasi-closed complexes.13
Complex 4 was found to be air- and moisture-stable and
perfectly soluble in common organic solvents (CH2Cl2,
toluene, THF, DMSO, etc.) without notable decomposition.
The complex is very stable even when reacted with 2 equiv
of pyridine for 3 h. However, the complete conversion into
the parent 3 takes place if allowed to react with an excess
of pyridine overnight. This behavior manifests a very stable
nature of the bent dirhodium bridges.
These results are particularly interesting because of the
Complex 3 was then allowed to react with an additional
1/2 equiv of Rh2Cl2(CO)4, and the reaction was monitored
by 31P{1H} NMR. The formation of a single new product
was observed (Scheme 4). The new compound exhibited a
signal that appeared as a deshielded doublet at 61.72 ppm
(JRh-P ) 171 Hz) in contrast to that observed for 3 (33.93
ppm). The conversion was quantitative after 12 h.
The compound possessing identical spectral characteristics
can successfully be prepared starting from ligand L1 and 1
equiv of the rhodium dimer Rh2Cl2(CO)4.
Suitable single crystals of 4, grown by the slow diffusion
of pentane into their saturated solutions in THF at room
temperature, were subjected to X-ray diffraction analysis (see
Table 1 for crystallographic parameters). As we expected,
the resulting 4 showed the bent arrangement of the Rh2Cl2
core (Figure 6).
The Rh atoms in the dinuclear complex 4 are in slightly
distorted square-planar configurations with the bond angles
around the metals ranging between 84.76 and 93.11 Å
(deviation from planarity is 0.05 Å). Terminal carbonyl
ligands adapt a syn conformation with Rh(1)-C(33) and
Rh(2)-C(34) distances of 1.806(8) and 1.799(8) Å, respec-
tively (within the normal range). Rh(1)-Cl(1) and Rh(2)-
Cl(1) (trans to phosphine) distances are only slightly longer
Figure 6. ORTEP drawing (50% probability ellipsoids) of the structure
of 4. Hydrogen atoms and solvent molecules are omitted for clarity. Selected
bond lengths (Å) and angles (deg): Cl(1)-Rh(1) 2.4085(10), Cl(1)-Rh-
(2) 2.4119(11), Cl(2)-Rh(1) 2.3904(9), Cl(2)-Rh(2) 2.3919(9), P(1)-Rh-
(1) 2.2361(10), P(2)-Rh(2) 2.2309(10), C(33)-O(1) 1.141(4), C(33)-Rh(1)
1.799(4), C(34)-O(2) 1.138(5), C(34)-Rh(2) 1.806(4), C(4)-P(1) 1.843(3),
C(10)-P(2) 1.839(3); Rh(1)-Cl(1)-Rh(2) 76.45(3), Rh(1)-Cl(2)-Rh(2)
77.18(3), C(33)-Rh(1)-P(1) 91.38(13), C(33)-Rh(1)-Cl(2) 175.50(13),
P(1)-Rh(1)-Cl(2) 93.11(3), C(33)-Rh(1)-Cl(1) 90.69(13), P(1)-Rh(1)-
Cl(1) 173.92(4), Cl(2)-Rh(1)-Cl(1) 84.87(3), C(34)-Rh(2)-P(2) 91.17(14),
C(34)-Rh(2)-Cl(2) 176.05(13), P(2)-Rh(2)-Cl(2) 92.31(3), C(34)-Rh-
(2)-Cl(1) 92.00(14), P(2)-Rh(2)-Cl(1) 172.76(4), Cl(2)-Rh(2)-Cl(1)
84.76(4).
(23) Redfield, D. A.; Cary, L. W.; Nelson, G. H. Inorg. Chem. 1975, 14,
50.
(24) (a) Fild, M.; Althoff, W. J. Chem. Soc., Chem. Commun. 1973, 933.
(b) Alexson, D. E.; Holloway, C. E. J. Chem. Soc., Chem. Commun.
1973, 455.
Inorganic Chemistry, Vol. 45, No. 17, 2006 7015

Azerraf et al.
package.27 Refinement of the structure on F2 was carried out by
the SHELXTL software package.28 Further information may be
found within the CIF files provided as Supporting Information.
recent observations of an unusual reactivity of dirhodium
halide-bridged complexes in the methanol carbonylation
reaction.13c
[Pd2Cl4(L1)] (1). A solution of PdCl2(L1) (200 mg, 0.301 mmol)
and PdCl2(CH3CN)2 (78 mg, 0.301 mmol) in chloroform (6 mL)
was stirred at 45 °C for 24 h. The yellow precipitate formed was
filtered, washed with two portions (2-3 mL) of the same solvent,
and redissolved in ca. 30-50 mL of chloroform. The analytically
pure compound was obtained by the slow diffusion of pentane into
a saturated chloroform solution (125 mg, 49%). 1H NMR
(CDCl3): δ 8.08 (d, 1H, J ) 7 Hz), 7.85 (s, 1H), 7.61 (d, 2H, J )
5 Hz), 7.46 (d, 1H, J ) 7 Hz), 7.14-7.07 (m, 6H), 5.58 (s, 1H),
3.38-3.33 (m, 2H), 2.53-2.46 (m, 2H), 1.70 (dd, 6H, J ) 7 Hz,
J ) 11 Hz), 1.38 (dd, 6H, J ) 7 Hz, J ) 11 Hz), 0.54 (dd, 6H, J
) 7 Hz, J ) 13 Hz), 0.44 (dd, 6H, J ) 7 Hz, J ) 8 Hz). 31P NMR
(CDCl3): δ 56.49. Anal. Calcd for C32H40Cl4P2Pd2: C, 45.69; H,
4.79. Found: C, 45.41; H, 4.95.
Conclusion
A series of bimetallic chloride-bridged complexes pos-
sessing a rare strongly bent M2(µ-Cl2) core have been
prepared via the ring-expansion reaction and have been fully
characterized. We demonstrated that steric properties of
ligands can play a pivotal role in the synthesis of halide-
bridged compounds with potentially “communicating” metal
centers. The stability of the bridge was exceptionally high:
no equilibrium between the mono- and bimetallic species
was observed under high-dilution conditions. Moreover, we
found that the complexes were stable even in the presence
of limited quantities of added Lewis bases such as pyridine.
These observations are important in the context of potential
applications of the new bent quasi-closed bimetallic com-
plexes in catalysis: one can expect the catalytic site to remain
undissociated. On the other hand, the exceptional stability
of the bimetallic complexes despite the conformational strain
in the 10-membered ring implies stabilizing interactions
between the metal centers. Finally, the spontaneous resolution
of the dissymmetric 3 into enantiomerically pure antipodes
was demonstrated. The latter will be used as seed crystals
for the resolution of larger quantities of the racemic L2.
Further theoretical and experimental studies on the mech-
anism of the ring expansion (which is of interest by itself),
the new bimetallic compounds,25 their catalytic and physical
properties, and the synthesis of heterobimetallic compounds
are in progress.
Synthesis of 1 under High-Dilution Conditions. A solution of
PdCl2(L1) in chloroform-d (0.5 mL, 7.5 mM) was mixed with a
solution of PdCl2(CH3CN)2 in the same solvent (0.5 mL, 7.5 mM).
The progress of the reaction was followed by 31P{1H} NMR at 45
°C.
[Pd2Cl4(L2)] (2). A solution of PdCl2(L2) (200 mg, 0.273 mmol)
and PdCl2(CH3CN)2 (70 mg, 0.273 mmol) in chloroform (6 mL)
was stirred at 45 °C for 24 h. The yellow precipitate formed was
filtered, washed with two portions (2-3 mL) of the same solvent,
and redissolved in ca. 30-50 mL of chloroform. The analytically
pure compound suitable for X-ray analysis was obtained by the
slow evaporation of chloroform (125 mg, 73%). 1H NMR
(CDCl3): δ 7.64-7.82 (m, 4H), 7.40-7.59 (m, 7H), 7.16-7.29
(m, 4H), 7.12 (t, 1H, J ) 7 Hz), 6.94 (t, 1H, J ) 7 Hz), 6.87 (t,
1H, J ) 7 Hz), 6.55-6.73 (m, 2H), 6.31 (d, 1H, J ) 7 Hz), 5.39
(s, 1H), 3.21-3.36 (m, 1H), 2.65-2.83 (m, 1H), 1.80 (dd, 3H, J
) 7 Hz, J ) 11 Hz), 1.53 (dd, 3H, J ) 7 Hz, J ) 11 Hz), 0.99 (dd,
3H, J ) 7 Hz, J ) 13 Hz), 0.77 (dd, 3H, J ) 7 Hz, J ) 8 Hz). 31
P
Experimental Section
NMR (CDCl3): δ 56.53 (s), 32.74 (s). Anal. Calcd for C38H36-
All manipulations were performed using the standard Schlenk
technique under an atmosphere of dry N2. Anhydrous solvents and
metal precursors were purchased from Aldrich and used as
delivered. Ligands L1 [1,8-bis(diisopropylphosphino)triptycene] and
L2 (1-diisopropylphosphino-8-diphenylphosphino-triptycene) and
complexes PdCl2(L1) and PdCl2(L2) were prepared using our
published procedures.17 NMR spectra were recorded on a Bruker
instrument operating at 400 MHz for protons, 100 MHz for carbons,
and 121 MHz for phosphorus. Diffraction data were collected with
a Bruker APEX CCD instrument [Mo KR radiation (λ ) 0.710 73
Å)]. Crystals were mounted onto glass fibers using epoxy. Single-
crystal reflection data were collected on a Bruker APEX CCD X-ray
diffraction system controlled by a Pentium-based PC running the
SMART software package.26 The integration of data frames and
refinement of the cell structure were done by the SAINT+ program
Cl4P2Pd2: C, 50.19; H, 3.99. Found: C, 50.32; H, 3.95.
[RhCl(CO)L1] (3). A solution of L1 (400 mg, 0.82 mmol) and
Rh2Cl2(CO)4 (159.6 mg, 0.41 mmol) in chloroform (8 mL) was
stirred at room temperature for 24 h. The analytically pure
compound was obtained by the slow diffusion of pentane into a
1
saturated chloroform solution (446.1 mg, 83%). H NMR (THF-
d8): δ 9.10 (s, 1H), 7.47 (d, 2H, J ) 8 Hz), 7.41 (d, 1H, J ) 7
Hz), 7.31 (d, 1H, J ) 7 Hz), 7.01-7.13 (m, 5H), 6.96 (t, 1H, J )
8 Hz), 5.49 (s, 1H), 2.96-3.03 (m, 2H), 2.73-2.85 (m, 2H), 1.54
(dd, 6H, J ) 7 Hz, J ) 11 Hz), 1.44 (dd, 6H, J ) 7 Hz, J ) 11
Hz), 1.16 (dd, 6H, J ) 7 Hz, J ) 13 Hz), 1.10 (dd, 6H, J ) 7 Hz,
J ) 8 Hz). 13C (CDCl3): δ 187.07 (dt, JC-Rh ) 57 Hz, JC-P ) 15
Hz), 152.03 (t, J ) 8 Hz), 147.52, 147.36 (t, J ) 4 Hz), 144.08,
129.13 (t, J ) 19 Hz), 127.57, 125.83, 125.70, 125.37, 124.71,
123.77 (t, J ) 3 Hz), 122.96, 55.23, 54.96 (t, J ) 6 Hz), 24.21 (t,
J ) 10 Hz), 23.57 (t, J ) 10 Hz), 20.99, 20.47 (t, J ) 3 Hz), 18.25
(t, J ) 6 Hz), 16.05. 31P NMR (THF-d8): δ 38.16 (d, JP-Rh ) 127
Hz). Anal. Calcd for C33H40ClOP2Rh: C, 60.70; H, 6.17. Found:
C, 60.57; H, 6.01.
(25) Attempts to prepare similar diplatinum complexes have been under-
taken as well. Unfortunately, we cannot report about the isolation of
a structurally characterized compound of this type by now, though
we have a NMR indication that they form. For example, when 2 equiv
of the trans-PtCl2(CH3CN)2 were allowed to react with 1 equiv of
L1, the formation of two products was observed by 31P{1H} NMR.
The major signal was centered at 1.93 ppm flanked by 195Pt satellites
[Rh2Cl2(CO)2L1] (4). Method A. A solution of L1 (400 mg, 0.82
mmol) and Rh2Cl2(CO)4 (319.2 mg, 0.82 mmol) in THF (8 mL)
was stirred at room temperature for 24 h. Crystals suitable for X-ray
1
and had a JPt-P coupling constant of 1848 Hz. This signal is
undoubtedly ascribed to a monometallic PdCl2(L1) complex. The
appearance of the minor signal centered at a lower field of 38.11 ppm
may be attributed to the desired bimetallic compound by analogy with
previously described Pd and Rh complexes. Unfortunately, the isolation
of this product did not work well so far, but experiments are in
progress.
(26) SMART-NT, v. 5.6; Bruker AXS GMBH: Karlsruhe, Germany, 2002.
(27) SAINT-NT, v. 5.0; Bruker AXS GMBH: Karlsruhe, Germany, 2002.
(28) SHELXTL-NT, v. 6.1; Bruker AXS GMBH: Karlsruhe, Germany,
2002.
7016 Inorganic Chemistry, Vol. 45, No. 17, 2006

Roof-Shaped Halide-Bridged Bimetallic Complexes
analysis were obtained by slow evaporation of the resulting clear
solution (508.6 mg, 76%).
Reactions of the Bimetallic Complexes with Pyridine. The
solution of 1 or 4 (0.01 mmol) in chloroform-d (3 mL) was treated
with an appropriate amount of pyridine (1 M solution in chloroform-
d), and the 31P{1H} NMR spectrum was recorded at the desired
temperature.
Method B. A solution of [RhCl(CO)L1] (200 mg, 0.306 mmol)
and Rh2Cl2(CO)4 (59.5 mg, 0.153 mmol) in chloroform (8 mL)
was stirred at 45 °C for 24 h. The clear yellow solution resulted.
The volatiles were evaporated in vacuo, providing the desired
product as yellow powder (239.1 mg, 95%). 1H NMR (CDCl3): δ
8.59 (s, 1H), 7.99 (d, 1H, J ) 7 Hz), 7.51 (d, 2H, J ) 7 Hz), 7.39
(d, 1H, J ) 7 Hz), 7.17 (t, 2H, J ) 7 Hz), 6.98-7.12 (m, 4H),
5.44 (s, 1H), 2.03-2.17 (m, 2H), 2.53-2.46 (m, 2H), 1.57 (dd,
6H, J ) 6 Hz, J ) 18 Hz), 1.32 (dd, 6H, J ) 7 Hz, J ) 17 Hz),
0.49-0.46 (m, 12H). 13C (CDCl3): δ 185.93 (dd, JC-Rh ) 82 Hz,
JC-P ) 16 Hz), 151.25 (d, J ) 11 Hz), 147.74 (d, J ) 8 Hz), 144.68,
143.99, 129.24, 126.62, 126.13 (d, J ) 2 Hz), 124.99 (d, J ) 4
Hz), 123.99 (d, J ) 6.26 Hz), 123.58, 123.23, 123.10, 56.16, 51.38
(t, J ) 4 Hz), 30.36 (d, J ) 33 Hz), 24.75 (d, J ) 29 Hz), 20.83
(d, J ) 7 Hz), 19.83 (d, J ) 3 Hz), 19.13 (d, J ) 2 Hz), 16.39 (d,
J ) 6 Hz). 31P NMR (CDCl3): δ 61.71 (d, JP-Rh ) 171 Hz).
Acknowledgment. We thank the German-Israeli Foun-
dation for Scientific Research and Development (Grant No.
894/05) for financial support and Prof. S. Biali for helpful
discussions.
Supporting Information Available: X-ray crystallographic files
in CIF format and copies of NMR spectra for all structures.
This material is available free of charge via the Internet at
http://pubs.acs.org.
IC060700Q
Inorganic Chemistry, Vol. 45, No. 17, 2006 7017