Positional disorder manifested as compositional in a pseudo-C 2-symmetrical Pd complex

Article abstract of DOI:10.1107/S0108270108002503

The title compound {2-[3,5-bis-(trifluoro-meth-yl)-1H-pyrazol-1-ylmeth-yl]- 6-(3,5-dimethyl-1H-pyrazol-1-ylmeth-yl)-pyridine}-meth-yl-palladium(II) tetra-kis[3,5-bis-(trifluoro-meth-yl)phen-yl]-borate, [Pd(C18H18F6N5)][B(C8H3F6) 4], crystallizes as discrete cations and anions. The cation possesses a pseudo-twofold axis about which positional disorder of the tridentate ligand is exhibited. The four substituents on the two pyrazole rings exhibit CH3/CF3 disorder, while all other atoms are ordered. Thus, this disorder can be conveniently described locally as compositional, while globally for the entire tridentate ligand it is positional. The anion also exhibits typical rotational positional disorder in three of the CF3 groups. All disordered CF3 groups were modeled with idealized C 3v geometry.

Full text of DOI:10.1107/S0108270108002503

metal-organic compounds
Acta Crystallographica Section C
Crystal Structure
Communications
complex composition where only one pyrazolyl unit binds to
palladium (Ojwach et al., 2007). The strong binding of the
pyrazolyl units to palladium in the cationic species was
believed to be responsible for the inability of the cationic
species to catalyze ꢀ-ole®n reactions.
ISSN 0108-2701
During our studies of the factors that render cationic
palladium complexes with strongly coordinated bis(pyrazol-1-
ylmethyl)pyridine ligands inactive for ꢀ-ole®n oligomerization
and polymerization reactions, we replaced the CH3 groups on
one of the pyrazolyl units in the bis(3,5-dimethylpyrazol-1-
ylmethyl)pyridine ligand with CF3 groups. We aimed to
synthesize a hemilabile ligand, which upon complexation with
palladium would allow ole®n coordination to the metal center
(Jeffrey & Rauchfuss, 1979; Shi et al., 2002). We succeeded in
preparing the desired palladium precursor complexes and the
expected cationic palladium species (see reaction scheme
below).
Positional disorder manifested as
compositional in a pseudo-C2-
symmetrical Pd complex
Ilia A. Guzei,a* Lara C. Spencer,a Stephen O. Ojwachb
and James Darkwab
aDepartment of Chemistry, University of Wisconsin±Madison, 1101 University
Avenue, Madison, WI 53706, USA, and bDepartment of Chemistry, University of
Johannesburg, Auckland Park Kingsway Campus, Johannesburg 2006, South Africa
Correspondence e-mail: iguzei@chem.wisc.edu
Received 16 January 2008
Accepted 23 January 2008
Online 9 February 2008
The title compound {2-[3,5-bis(tri¯uoromethyl)-1H-pyrazol-1-
ylmethyl]-6-(3,5-dimethyl-1H-pyrazol-1-ylmethyl)pyridine}-
methylpalladium(II) tetrakis[3,5-bis(tri¯uoromethyl)phenyl]-
borate, [Pd(C18H18F6N5)][B(C8H3F6)4], crystallizes as discrete
cations and anions. The cation possesses a pseudo-twofold axis
about which positional disorder of the tridentate ligand is
exhibited. The four substituents on the two pyrazole rings
exhibit CH3/CF3 disorder, while all other atoms are ordered.
Thus, this disorder can be conveniently described `locally' as
compositional, while `globally' for the entire tridentate ligand
it is positional. The anion also exhibits typical rotational
positional disorder in three of the CF3 groups. All disordered
CF3 groups were modeled with idealized C3v geometry.
Unfortunately, the cationic species showed no catalytic
activity in ole®n reactions. Spectroscopic data for the cation
indicate that the tridentate ligand forms strong PdÐN bonds,
whereas the CF3 substituents do not make the pyrazolyl unit
labile. We con®rmed the coordination of the ligand by X-ray
crystallography and report here the structure of this complex,
(I).
Comment
Cationic palladium(II) complexes of symmetrical ꢀ-diimine
ligands with bulky aryl substituents on the N atoms are known
to ef®ciently catalyze ꢀ-ole®n oligomerization and poly-
merization reactions (Brookhart et al., 1995; Ittel et al., 2000;
Gibson & Spitzmesser, 2003). These cationic ꢀ-diimine
palladium complexes are often prepared by direct halide
abstraction using silver or alkali metal salts of weakly coor-
dinating ligands. The weakly coordinating ligands do not
compete with an incoming monomer for the vacant coordi-
nation site of an active catalyst (Mecking, 2000; Mecking et al.,
1998; Bianchini et al., 2006). The role of the weakly coordi-
nating ligands is therefore only to stabilize the active catalysts.
In our attempts to utilize bis(pyrazol-1-ylmethyl)pyridine±
palladium±chloromethyl complexes to catalyze ole®n reac-
tions, we found that such complexes yielded inactive cationic
species when the Cl atom in the precursor complexes was
abstracted (Ojwach et al., 2007). We were able to establish by
X-ray crystallography that the cationic species formed after
chloride abstraction had both pyrazolyl units strongly bound
to palladium. This is in contrast to the chloromethyl precursor
Ionic compound (I) crystallizes as discrete anions and
cations. The Pd center in the cationic complex bears a large ꢁ3-
ligand and a methyl group (Fig. 1). Both the cation and the
tetrakis[3,5-bis(tri¯uoromethyl)phenyl]borate anion exhibit
rotational positional disorder.
The most interesting aspect of the cationic structure is the
disorder. The complex contains a pseudo-twofold axis passing
m114 # 2008 International Union of Crystallography
DOI: 10.1107/S0108270108002503
Acta Cryst. (2008). C64, m114±m116

metal-organic compounds
along the PdÐC1 vector and exhibits positional rotational
disorder about it. A 180ꢀ rotation about the axis superimposes
the two pyrazole rings to make the CH3 and CF3 groups
appear disordered, whereas all other non-H atoms remain
ordered. Since there is no ambiguity as to the composition of
the ꢁ3-ligand, the CH3/CF3 disorder is positional; however, it is
manifested as compositional and was modeled as such. The
CH3/CF3 disorder ratio is 86.5 (3):13.5 (3)%. This disorder can
be thought of as positional in a `global' molecular sense and
compositional in a `local' substituent sense. The geometries of
the disordered CF3 groups were modeled with a C3v idealized
arrangement based on a density functional theory (DFT)
computation for 3-tri¯uoromethyl-1H-pyrazole.
The overall geometry of the Pd cation cannot be higher than
C1, and even with a symmetrical substitution pattern of the
pyrazole rings it cannot exceed Cs because of the ligated
methyl group. The Pd1 coordination polygon de®ned by atoms
C1, N1, N3 and N5 is irregular and slightly distorted from a
number of possible geometries, such as D2d and D4h. The four
a methyl group. The PdÐNpy distance in (I) is similar to the
Ê
distance of 2.128 (2) A in (II). The metal complex in (II)
differs from the cation in (I) only in that it has all methyl
groups substituted on the pyrazole rings (Ojwach et al., 2007).
The dihedral angles between the best ®tting least-squares
planes de®ned by the ring pyrazole atoms, ring pyridine atoms
and four ligating atoms are given in Table 3. These angles are
similar to those for other compounds with the same tridentate
pyrazolylmethylpyridine ligands; thus a pseudo-C2 geometry
of the ligand is preferred to a pseudo-Cs geometry.
The bond distances and angles in the tetrakis[3,5-bis(tri-
¯uoromethyl)phenyl]borate anion (Fig. 2) are typical, as
con®rmed by a Mogul structural check (Bruno et al., 2004).
The geometry around the B atom is slightly distorted tetra-
hedral. The average angle around atom B1 is 109.5 (11)ꢀ, but
the angles range from 107.8 (2) to 110.6 (2)ꢀ. The F atoms on
three of the CF3 groups in the anion show rotational positional
disorder over three positions each. The F atoms attached to
atom C26 are disordered in a 42.0 (4):29.2 (3):28.8 (3)% ratio,
those attached to C42 are disordered in a 52.3 (3):33.3 (5):-
14.3 (4)% ratio, and those attached to atom C50 are disor-
dered in a 54.0 (3):38.3 (4):7.7 (3)% ratio. All of these disor-
dered F atoms in the anion were re®ned isotropically, and the
CF3 groups were re®ned with restraints. The geometries of the
disordered CF3 groups were modeled with a C3v idealized
arrangement based on a DFT computation for PhCF3. It is
quite common for the tetrakis[3,5-bis(tri¯uoromethyl)phen-
yl]borate anion in crystal structures to contain some disorder.
Our data mining of the CSD revealed that in 295 out of 512
instances of this anion in the CSD there was some disorder in
the anion.
Ê
ligating atoms are coplanar within 0.07 A. The eccentricity of
Ê
Ê
atom Pd1 is 0.078 (3) A and it is displaced by 0.0139 (14) A
from the least-squares plane of the ligated atoms. The
geometry about the Pd atom is typical and can be summarized
as follows: the Pd atom has a distorted square-planar
geometry, with angles around the central Pd atom ranging
from 85.77 (10) to 93.62 (12)ꢀ and an average value of 90 (4)ꢀ
(Table 1). Atoms Pd1, N1, N3, N5 and C1 are coplanar within
Ê
0.06 A. The average PdÐNpz (pz is pyrazole) distance and
NpzÐPdÐNpy (py is pyridine) angle in (I) are statistically
similar to the averages for the nine relevant compounds
(Table 2) in the Cambridge Structural Database (CSD;
Version 1.10; Allen, 2002); however, the PdÐNpy distance in
(I) is signi®cantly longer than the average for the similar
compounds in the CSD. This difference can be attributed to
the trans in¯uence of the methyl group. In seven of the nine
related compounds, the ligand trans to the pyridine ring is a Cl
atom, in one it is another pyridine group, and in [2,6-bis(3,5-
dimethylpyrazolylmethyl)pyridine]methylpalladium(II) tetra-
kis[3,5-bis(tri¯uoromethyl)phenyl]borate, compound (II), it is
Figure 2
The molecular structure of the anion of (I). Displacement ellipsoids are
shown at the 50% probability level. All H atoms and all minor
components of disordered atoms have been omitted for clarity. The
disordered F atoms in the anion that were re®ned isotropically are shown
with crosshatched circles.
Figure 1
The molecular structure of the cation of (I). Displacement ellipsoids are
shown at the 50% probability level. All H atoms and all minor
components of disordered atoms have been omitted for clarity.
ꢁ
Acta Cryst. (2008). C64, m114±m116
Guzei et al.
[Pd(C18H18F6N5)](C32H12BF24
)
m115

metal-organic compounds
Table 3
Experimental
Dihedral angles (ꢀ) between least-squares planes in the cation.
To a J-Young NMR tube containing a solution of the palladium
precursor complex (4.00 mg, 0.007 mmol) in CDCl3 (0.2 ml) was
added a solution of NaBAr4 [Ar is 3,5-(CF3)2C6H3] (6.00 mg,
0.007 mmol) in CDCl3 (0.2 ml) (see reaction scheme in the
Least-squares plane
N1/N2/C3±C5
N4/N5/C15±C17
N3/C8±C12
C1/N1/N3/N5
N1/N2/C3±C5
N4/N5/C15±C17
51.14 (11)
±
±
39.44 (11)
87.86 (12)
±
42.48 (12)
70.32 (10)
61.06 (12)
1
Comment), and the H NMR spectrum was acquired after vigorous
shaking. The solution was left to stand at room temperature for
several days to afford colorless single crystals of (I) suitable for X-ray
analysis. 1H NMR (CDCl3): ꢂ 1.27 (s, 3H, CH3, PdÐMe), 2.17 (s, 3H,
CH3, pz), 2.33 (s, 3H, CH3, pz), 5.16 (d, 2H, CH2, 2JHH = 15.6 Hz), 5.63
(s, 1H, pz), 5.98 (s, 1H, pz), 6.20 (d, 2H, CH2, 2JHH = 17.4 Hz), 6.71 (d,
1H, py, 3JHH = 8.2 Hz), 7.23 (d, 2H, py, 3JHH = 8.6 Hz), 7.34 (t, 1H, py,
3JHH = 8.3 Hz), 7.48 (s, 4H, BAr4 ), 7.66 (s, 8H, BAr4 ). 19F{1H} NMR
(CDCl3): ꢂ 62.6 (s, BAr4 ), 60.2 (s, CF3, pz), 56.0 (s, CF3, pz).
13C{1H} NMR (CDCl3): ꢂ 5.7, 11.5, 15.0, 52.1, 55.8, 109.2, 117.5, 122.6,
124.7, 134.7, 142.0, 48.8, 152.2, 162.6.
1.5Ueq(C) for methyl H atoms and 1.2Ueq(C) for all other H atoms].
Default effective CÐH distances were adopted (secondary Csp3ÐH =
3
2
Ê
Ê
Ê
0.99 A, primary Csp ÐH = 0.98 A and tertiary Csp ÐH = 0.95 A).
The geometry of all disordered CF3 groups was modeled on the
basis of pbepbe/6±311++G(df,pd) DFT computations to conform to
C3v symmetry. This was achieved with the following restraints: the
CÐF distances were allowed to re®ne as one free variable FVAR2;
the FÁ Á ÁF separations were restrained to 1.607 (3) times FVAR2; the
CipsoÁ Á ÁF separations were restrained to 1.746 (11) times FVAR2.
This was achieved with DFIX commands in SHELXTL (Sheldrick,
2008). Attempts to conduct the re®nement with SAME instructions
did not result in computationally stable re®nements. All disordered F
atoms, except F1, F2 and F3, were re®ned isotropically.
Crystal data
3
Ê
[Pd(C18H18F6N5)](C32H12BF24)
Mr = 1388.00
V = 5308.2 (4) A
Z = 4
Monoclinic, P21=n
Mo- Kꢀ radiation
ꢄ = 0.50 mm
T = 105 (2) K
0.36 Â 0.31 Â 0.26 mm
1
Ê
a = 10.9194 (4) A
Ê
b = 18.0256 (7) A
There are two large peaks (ꢂ2 e A 3) in the ®nal difference map
Ê
Ê
c = 27.1635 (11) A
ꢃ = 96.865 (1)ꢀ
in the vicinity of the disordered group at C50, which may represent
additional positions of the F atoms. No attempt to re®ne this CF3
group as disordered over four positions was made.
Data collection
Bruker SMART 1000 CCD area-
detector diffractometer
Absorption correction: multi-scan
(SADABS; Bruker, 2007)
Tmin = 0.841, Tmax = 0.881
92342 measured re¯ections
14211 independent re¯ections
12423 re¯ections with I > 2ꢅ(I)
Rint = 0.030
Data collection: SMART (Bruker, 2000); cell re®nement: SAINT-
Plus (Bruker, 2007); data reduction: SAINT-Plus; program(s) used to
solve structure: SHELXTL (Sheldrick, 2008); program(s) used to
re®ne structure: SHELXTL; molecular graphics: SHELXTL; soft-
ware used to prepare material for publication: SHELXTL, publCIF
(Westrip, 2008) and modiCIFer (Guzei, 2007).
Re®nement
R[F2 > 2ꢅ(F2)] = 0.058
wR(F2) = 0.154
S = 1.01
14211 re¯ections
823 parameters
109 restraints
H-atom parameters constrained
3
Ê
Áꢆmax = 2.16 e A
Supplementary data for this paper are available from the IUCr electronic
archives (Reference: SK3199). Services for accessing these data are
described at the back of the journal.
3
Ê
1.63 e A
Áꢆmin
=
Table 1
Selected geometric parameters (A, ).
ꢀ
Ê
References
Allen, F. H. (2002). Acta Cryst. B58, 380±388.
Pd1ÐN5
Pd1ÐN1
2.030 (3)
2.049 (3)
Pd1ÐC1
Pd1ÐN3
2.061 (3)
2.127 (3)
Bianchini, C., Giambastiani, G., Rios, I. G., Mantovani, G., Meli, A. & Segarra,
A. M. (2006). Coord. Chem. Rev. 250, 1391±1418.
Brookhart, M., Johnson, L. K. & Killian, C. M. (1995). J. Am. Chem. Soc. 117,
6414±6415.
Bruker (2000). SMART. Version 5.622. Bruker AXS Inc., Madison, Wisconsin,
USA.
N5ÐPd1ÐN1
N5ÐPd1ÐC1
N1ÐPd1ÐC1
172.77 (10)
93.62 (12)
93.07 (11)
N5ÐPd1ÐN3
N1ÐPd1ÐN3
C1ÐPd1ÐN3
87.76 (11)
85.77 (10)
175.40 (12)
Bruker (2007). SADABS and SAINT-Plus. Bruker AXS Inc., Madison,
Wisconsin, USA.
Bruno, I. J., Cole, J. C., Kessler, M., Luo, J., Motherwell, W. D. S., Purkis, L. H.,
Smith, B. R., Taylor, R., Cooper, R. I., Harris, S. E. & Orpen, A. G. (2004). J.
Chem. Inf. Comput. Sci. 44, 2133±2144.
Gibson, V. C. & Spitzmesser, S. K. (2003). Chem. Rev. 103, 283±315.
Guzei, I. A. (2007). modiCIFer. University of Wisconsin±Madison, USA.
Ittel, S., Johnson, L. K. & Brookhart, M. (2000). Chem. Rev. 100, 1169±1204.
Jeffrey, J. C. & Rauchfuss, T. B. (1979). Inorg. Chem. 18, 2658±2666.
Mecking, S. (2000). Coord. Chem. Rev. 203, 325±351.
Mecking, S., Johnson, L. K., Wang, L. & Brookhart, M. (1998). J. Am. Chem.
Soc. 120, 888±898.
Table 2
ꢀ
Ê
Distances and angles (A, ) in (I) compared with those for nine similar
compounds in the CSDa.
(I)
CSD average
PdÐNpz
PdÐNpy
2.039 (13)
2.127 (3)
86.8 (14)
2.03 (2)
2.04 (4)
86.4 (16)
NpzÐPdÐNpy
Ojwach, S. O., Guzei, I. A., Darkwa, J. & Mapolie, S. F. (2007). Polyhedron, 26,
851±861.
(a) Cambridge Structural Database (Version 1.10; Allen, 2002).
Sheldrick, G. M. (2008). Acta Cryst. A64, 112±122.
Shi, P.-Y., Liu, Y.-H., Peng, S.-M. & Liu, S.-T. (2002). Organometallics, 21,
3203±3207.
All H atoms were placed in idealized locations and re®ned as
riding with appropriate displacement parameters [Uiso(H)
=
Westrip, S. P. (2008). publCIF. In preparation.
ꢁ
m116 Guzei et al. [Pd(C18H18F6N5)](C32H12BF24
)
Acta Cryst. (2008). C64, m114±m116