Synthesis and characterization of a new binucleating tetraphosphine ligand based on 1,2-phenylene chelates and the structures of dinickel tetrachloride complexes of the ligand

Article abstract of DOI:10.1021/ic5019345

A new binucleating tetraphosphine ligand, rac-and meso-(Et₂P-1,2-C₆H₄)P(Ph)CH₂(Ph)P(1,2-C₆H₄PEt₂) (et,ph-P4-Ph), has been synthesized. Separation and purification of the ligand diastereomers have been accomplished via column chromatography. Ni₂Cl₄(et,ph-P4-Ph) complexes of both diastereomers have been prepared in high yield and crystallographically characterized.

Full text of DOI:10.1021/ic5019345

Communication
pubs.acs.org/IC
Synthesis and Characterization of a New Binucleating
Tetraphosphine Ligand Based on 1,2-Phenylene Chelates and the
Structures of Dinickel Tetrachloride Complexes of the Ligand
William J. Schreiter, Alexandre R. Monteil, Ekaterina Kalachnikova, Marc A. Peterson, Marshall D. Moulis,
Ciera D. Gasery, Gregory T. McCandless, Frank R. Fronczek, and George G. Stanley*
Department of Chemistry, Louisiana State University, Baton Rouge, Louisiana 70803-1804, United States
S
* Supporting Information
ABSTRACT: A new binucleating tetraphosphine ligand,
rac- and meso-(Et2P-1,2-C6H4)P(Ph)CH2(Ph)P(1,2-
C6H4PEt2) (et,ph-P4-Ph), has been synthesized. Separa-
tion and purification of the ligand diastereomers have been
accomplished via column chromatography. Ni2Cl4(et,ph-
P4-Ph) complexes of both diastereomers have been
prepared in high yield and crystallographically charac-
terized.
he tetraphosphine ligands rac- and meso-(Et2PCH2CH2)-
(Ph)PCH2P(Ph)(CH2CH2PEt2) (et,ph-P4), 1R and 1M
(Figure 1), were designed to bridge and chelate two metal
Figure 2. (Top) Two proposed rhodium complexes that are formed
from the fragmentation of catalytically active [Rh2H2(μ-CO)2(rac-et,ph-
P4)]2+ species during hydroformylation in acetone. (Bottom)
Ni2Cl4(meso-et,ph-P4) that completely converts to the bimetallic
double-P4 nickel complex shown when dissolved in a polar organic
solvent that contains more than 5% (by volume) water, even at room
temperature.9
T
of the rhodium products shown in Figure 2 is active for
hydroformylation. Their structures are based on similar rhodium
complexes that have been crystallographically characterized10,11
1
and on H and 31P{1H} NMR experiments of [Rh2(nbd)2(rac-
et,ph-P4)]2+ under H2/CO pressure.
Therefore, the synthesis of a new tetraphosphine ligand in
which the rotationally flexible ethylene linkage between the
internal and external phosphines is replaced by a rigid phenylene
group will lead to a much stronger chelate effect, inhibiting
phosphine arm dissociation. This should lead to a more robust
dirhodium catalytic system that still has high activity and
regioselectivity via bimetallic cooperativity. Herein we report the
synthesis and characterization of the new tetraphosphine ligands
rac- and meso-(Et2P-1,2-C6H4)P(Ph)CH2(Ph)P(1,2-C6H4PEt2)
(et,ph-P4-Ph), 2R and 2M (Figure 1), and the bimetallic nickel
complexes of 2R and 2M.
The synthesis of 2R and 2M is depicted in Scheme 1. In the
first step, (H)(Ph)PCH2P(Ph)(H) (3) is reacted with 2 equiv of
hexachloroethane12 and refluxed in diethyl ether for 24 h to give,
after workup, (Cl)(Ph)PCH2P(Ph)(Cl) (4) as a slightly pink
air-, moisture-, and heat-sensitive viscous liquid that has been
fully characterized (see the Supporting Information, SI).13,14 The
31P{1H} NMR spectrum of 4 shows two singlets at 81.3 and 81.0
Figure 1. (Top) Tetraphosphine ligands 1R and 1M. (Bottom) New
more strongly chelating tetraphosphine ligands 2R and 2M.
centers to promote bimetallic cooperativity during a catalytic
reaction.1,2 This goal was realized with the synthesis of
[Rh2(nbd)2(rac-et,ph-P4)][BF4]2 (nbd = norbornadiene),
which is a precursor to a very active and regioselective
hydroformylation catalyst.3−5 This catalytic system remains
one of the most dramatic examples of bimetallic cooperativity in
homogeneous catalysis.
Spectroscopic studies4,6−8 on the dirhodium system as well as
on a related nickel system9 have shown that these complexes can,
unfortunately, readily fall apart to form inactive mono- and
bimetallic complexes (Figure 2). This is likely due to the
rotational flexibility of the ethylene bridge, which allows one of
the PEt2 arms to dissociate, leading to the loss of a metal center,
even under mild conditions. Once one metal center is lost, either
the ligand can wrap around the remaining metal to form a
monometallic complex, or two monometallic complexes can
dimerize to form double-P4 ligand bimetallic complexes. Neither
Received: August 8, 2014
Published: September 10, 2014
© 2014 American Chemical Society
10036
dx.doi.org/10.1021/ic5019345 | Inorg. Chem. 2014, 53, 10036−10038

Inorganic Chemistry
Communication
distillation to remove 6 is eliminated because it usually comes out
of the separation column in the first 5 × 10 mL fractions,
followed by 2M and then 2R. The isolated yield of et,ph-P4-Ph
after chromatography is 60−65%.
Scheme 1. Synthesis of 2R and 2M
The 31P{1H} NMR of each diastereomer is shown in Figure 3
along with the 1H NMR of the methylene bridge region, which
ppm in benzene-d6 due to rac and meso diastereomers. However,
only one resonance can be observed at high concentrations or in
a nonaromatic solvent; e.g., the 31P NMR spectrum of 4 in
CD2Cl2 shows only one singlet at 82.0 ppm. This is presumably
due to intermolecular interactions that result in the accidental
degeneracy of the resonances depending on the solvent and/or
concentration.
The next stage involves a two-step process to make 1-(Et2P)-2-
iodobenzene (6).15 1,2-Diiodobenzene is reacted with iPrMgBr,
then treated with Et2PCl, and allowed to stir overnight. After
workup, 6 is isolated as an air-, moisture-, and light-sensitive
colorless liquid in 70−75% yield.16
Figure 3. (Top) 31P{1H} NMR of rac- and meso-et,ph-P4-Ph showing
the second-order patterns for each diastereomer. (Bottom) 1H NMR of
the methylene bridge region for each diastereomer.
The final step of the synthesis involves another Grignard-
mediated P−C coupling reaction. 6 is reacted with iPrMgBr
followed by the addition of 4 and allowed to stir overnight. An
initial workup of this reaction mixture allowed us to isolate a
white paste, which was identified as a 1:1 mixture of 2R and 2M
in 75−81% yield. Part of this initial workup involved short-path
vacuum distillation of the crude reaction mixture around 130 °C
to remove unreacted 6 and other volatile impurities. However,
1H NMR experiments monitoring the distinctive central
methylene protons of 2R and 2M demonstrate that the crude
reaction mixture contains mostly (>95%) the meso diastereomer.
It is only after the crude product is heated that a 1:1 mixture of
both diastereomers is obtained. Although 4 exists as an
approximately 1:1 mixture of diastereomers, the final P−C
coupling forms mainly the meso diastereomer for reasons not
currently understood. Heating 2M at 130 °C for only 1−2 h leads
to complete epimerization. Thus, if one wants 2M, only very mild
heating should be employed during the solvent removal and
ligand isolation steps. Currently, the only way to obtain 2R is to
racemize 2M and separate the individual diastereomers.
Purification and separation of crude 2M and 2R can be
achieved directly via column chromatography. The crude
mixture of the final epimerized product can be initially purified
using a neutral alumina column and eluting with CH2Cl2. This
removes a large portion of the impurities present in the crude
ligand mixture (19−25% impurities). A second longer neutral
alumina (grade IV) column using a 1:4 mixture of CH2Cl2/
hexanes allows for good separation of 2M and 2R with only a
small interface containing a mixture of both diastereomers, which
can be recycled. Additionally, the need for short-path vacuum
more clearly identifies the rac and meso ligand diastereomers. The
31P NMR for each diastereomer is a complex second-order
pattern because of the closeness of the chemical shifts and
coupling constants between the internal and external phosphorus
atoms. The 1H NMR of the methylene bridge hydrogen atoms is
much more useful for identifying the purity of each diastereomer.
The ligand was reacted with 2 equiv of NiCl2·6H2O to form
the bimetallic nickel complexes in order to more fully
characterize the ligand’s coordinating ability for making
bimetallic complexes. As with 1R and 1M, one can use a 1:1
mixture of 2R and 2M and the proper solvents to get a fairly clean
separation of the two dinickel P4 diastereomers.2,17 Dissolving
the mixed ligand in CH2Cl2 and adding this solution to NiCl2·
6H2O in 1-butanol produces an orange powder that precipitates
out of solution. After filtration to remove the orange powder, the
remaining dark-red solution is allowed to sit overnight, during
which small orange crystals precipitate out of solution.
NMR analysis of both the filtered orange powder and orange
crystals from the dark-red filtrate revealed them to be the pure
meso diastereomer, Ni2Cl4(meso-et,ph-P4-Ph) (7M). A workup
of the remaining dark-red solution affords an orange powder,
which consists of an 87:13 mixture of Ni2Cl4(rac-et,ph-P4-Ph),
7R, and 7M. Both powders can be recrystallized from either
acetonitrile or acetone to give crystals suitable for single-crystal
X-ray diffraction. The acetone-derived crystals have two
molecules of 7M and acetone per asymmetric unit. ORTEP
plots of 7R and 7M are shown in Figure 4.
10037
dx.doi.org/10.1021/ic5019345 | Inorg. Chem. 2014, 53, 10036−10038

Inorganic Chemistry
Communication
ASSOCIATED CONTENT
* Supporting Information
■
S
Synthetic details, spectroscopic data (NMR and IR spectra of P4
ligands; 31P NMR data on Ni2 complexes), and X-ray
crystallographic data. This material is available free of charge
via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: gstanley@lsu.edu.
■
Author Contributions
All authors have given approval to the final version of the
manuscript.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
Financial support from the NSF (Grant CHE-0111117), Sasol
North America, Louisiana Board of Regents (OPT-IN &
Graduate Fellowship Program), and a summer graduate student
research assistantship from Eastman Chemical are gratefully
acknowledged.
REFERENCES
■
(1) Laneman, S. A.; Fronczek, F. R.; Stanley, G. G. J. Am. Chem. Soc.
1988, 110, 5585−5586.
(2) Laneman, S. A.; Fronczek, F. R.; Stanley, G. G. Inorg. Chem. 1989,
28, 1872−1878.
(3) Broussard, M. E.; Juma, B.; Train, S. G.; Peng, W.-J.; Laneman, S.
A.; Stanley, G. G. Science 1993, 260, 1784−1788.
(4) Matthews, R. C.; Howell, D. K.; Peng, W.-J.; Train, S. G.;
Treleaven, W. D.; Stanley, G. G. Angew. Chem., Int. Ed. 1996, 35, 2253−
2256.
Figure 4. ORTEPs of 7R (top) and 7M (bottom; one molecule from the
acetone solvate structure). Hydrogen atoms omitted for clarity.
(5) Aubry, D. A.; Bridges, N. N.; Ezell, K.; Stanley, G. G. J. Am. Chem.
Soc. 2003, 125, 11180−11181.
(6) Matthews, R. C. In Situ Spectroscopic Studies of a Bimetallic
Hydroformylation Catalyst. Ph.D. Dissertation, Louisiana State
University, Baton Rouge, LA, 1999.
The crystal structures revealed the expected square-planar
arrangement of two Cl− ligands cis to each other along with two
phosphine moieties chelated through the phenylene linkage for
both metal centers. The bond lengths are very similar to those
seen for Ni2Cl4(1R) and Ni2Cl4(1M)2 and to one another (see
the SI). The biggest difference between 7R and 7M is the
rotational conformation about the central methylene bridge. For
7R, the nickel centers are symmetrically rotated away from one
another, adopting a completely open-mode geometry in which
the nickel centers are on opposite sides of the molecule with a
Ni−Ni separation of 5.9031(1) Å and a Ni1−P2···P3−Ni2
torsional angle of 130.14(1)°.
This conformation is similar to the solid-state structure of
Ni2Cl4(1R), which also adopts an open-mode conformation.
However, the Ni−Ni separation (5.417 Å) and Ni1−P···P′−Ni2
torsional angle (105.82°) are both smaller for Ni2Cl4(1R).2 7M
adopts a partially closed-mode geometry with a Ni−Ni distance
ranging between 4.272 and 4.434 Å and an average Ni1−P2···
P3−Ni2 torsional angle of 42.2°. This is facilitated by a weak
interaction between Ni1 and Cl3 with an average distance of
3.046 Å. This solid-state conformation is quite different from
Ni2Cl4(1M), which adopts an open-mode structure with a longer
Ni−Ni separation (6.272 Å) and a much larger Ni1−P2···P3−
Ni2 torsional angle (160°) compared to those of 7M.2
(7) Gueorguieva, P. G. Spectroscopic and Synthetic Studies Relating to
a Dirhodium Hydroformylation Catalyst. Ph.D. Dissertation, Louisiana
State University, Baton Rouge, LA, 2004.
(8) Polakova, D. Studies on a Dirhodium Tetraphosphine Hydro-
formylation Catalyst. Ph.D. Dissertation, Louisiana State University,
Baton Rouge, LA, 2012.
(9) Schreiter, W. J. Investigations into Alkene Hydration and
Oxidation Catalysis. Ph.D. Dissertation, Louisiana State University,
Baton Rouge, LA, 2013.
(10) Hunt, C., Jr.; Fronczek, F. R.; Billodeaux, D. R.; Stanley, G. G.
Inorg. Chem. 2001, 40, 5192−5198.
(11) Hunt, C., Jr.; Nelson, B. D.; Harmon, E. G.; Fronczek, F. R.;
Watkins, S. F.; Billodeaux, D. R.; Stanley, G. G. Acta Crystallogr., Sect. C
2000, 56, 546−548.
(12) Weferling, N. Z. Anorg. Allg. Chem. 1987, 548, 55−62.
(13) Gol, F.; Hasselkuß, G.; Knuppel, P. C.; Stelzer, O. Z. Naturforsch.,
̈
B: Chem. Sci. 1988, 43, 31−44.
(14) Schmidbaur, H.; Schnatterer, S. Chem. Ber. 1986, 119, 2832−42.
(15) Boymond, L.; Rottlander, M.; Cahiez, G.; Knochel, P. Angew.
Chem., Int. Ed. 1998, 37, 1701−1703.
̈
(16) Schreiter, W. J.; Monteil, A. R.; Peterson, M. A.; McCandless, G.
T.; Fronczek, F. R.; Stanley, G. G. Polyhedron 2013, 58, 171−178.
(17) Aubry, D. A.; Laneman, S. A.; Fronczek, F. R.; Stanley, G. G. Inorg.
Chem. 2001, 40, 5036−5041.
We believe that the far stronger chelate effect of this new P4
ligand will produce catalysts that will have the high activity and
regioselectivity observed with the previous ligand 1R but be far
more resistant to deactivating fragmentation reactions.
10038
dx.doi.org/10.1021/ic5019345 | Inorg. Chem. 2014, 53, 10036−10038