Crystal structures of metallocene complexes with uranium–germanium bonds

The first structural examples of complexes with uranium–germanium bonds are presented. The two complexes both have a long U—Ge bond [distances of 3.0428 (7) and 3.0523 (7) Å.


Chemical context
While actinide complexes with heavier main-group elements have been studied with the chalcogen and pnictogen groups, the tetrel series has not been examined in nearly as much detail. Actinide-heavier main-group element bonds have been of interest to our group and others, for three primary reasons. First is the exploration of the energy-driven-covalency concept, which shows increased covalent-bonding character going down a group (Walensky et al., 2010;Neidig et al., 2013;Su et al., 2018). Second, despite increased covalency, bond strength does not scale with covalency, hence the weaker, more polarized bonds with heavier main-group elements should afford greater reactivity (Kaltsoyannis, 2013). Finally, the fundamental chemistry obtained by the structure, bonding, and reactivity of these understudied metals advances our knowledge of the periodic table and helps to elucidate new and exciting properties.
The coordination chemistry of f elements with heavier tetrel elements (Si, Ge, Sn, Pb) is quite rare (Ré ant et al., 2020b), and their reactivity is virtually unknown. With respect to the actinides, there are two reports of actinide-silicon bonds without structural data (King & Marks, 1995;Radu et al., 1995), and two structurally characterized uranium-silicon bonds with anionic silyl ligands (Diaconescu et al., 2001;Ré ant et al., 2020a) and two more with neutral silylene ligands (Brackbill et al., 2020). In the 1990s, the reaction of (C 5 H 5 ) 3 UCl with KEPh 3 , E = Si, Ge, Sn, was conducted by Porchia and co-workers to form uranium-tetrel bonds, and their reactivity with isocyanides was described (Porchia et al., 1986(Porchia et al., , 1987(Porchia et al., , 1989. Finally, the Boncella group has more recently reported a second uranium-tin bond (Winston et al., 2016). We have found that protonolysis reactions with primary pnictines have resulted in the formation of actinide-pnictido bonds (Behrle & Walensky, 2016;Vilanova et al., 2017;Tarlton et al., 2020Tarlton et al., , 2021, so we decided to utilize a secondary ISSN 2056-9890 germane in the same strategy. However, the issue is the protonic versus hydridic nature of the E-H bonds, and hence we used 3,5-(CF 3 ) 2 C 6 H 3 -substituted germane to obtain a more protonic hydrogen. Herein, we report the structural characterization of uranium-germanium bonds with a secondary germanido ligand. When attempting to form the germylene, a C-F bond-activated product is obtained, indicating the reactive nature of these weak uranium-germanium bonds.

Structural commentary
The solid-state structure of each complex was definitively determined by X-ray crystallography to elucidate the first uranium-germanium bond (Fig. 1). Both structures have similar geometries in which the U atom is coordinated to two 5 -coordinating Cp* ligands, a halide ligand, and the germanido ligand, which coordinates only through the germanium atom in both cases. Geometric parameters involving the U-Ge distances and Ge-centered angles are given in Table 1. Each U-Ge bond is within the sum of the covalent radii of 3.16 Å (Cordero et al., 2008). Both complexes are distorted tetrahedra with the Cp* ligands occupying single vertices. The angles between the Cp* mean planes are similar in both structures [133.4 (3) for 1 and 137.8 (3) for 2], which is significantly larger than the ideal tetrahedral angle as expected for two adjacent, sterically bulky ligands. The uraniumcentered angles between the halide and Ge atoms in both structures are consequently distorted to smaller angles [88.06 (2) for 1 and 88.92 (14) for 2]. The 3,5-(CF 3 ) 2 C 6 H 3 rings are oriented significantly differently in the two structures. In 1 the rings are roughly consistent with a mirror plane passing through the U, Ge, and I atoms, and their mean planes intersect at an angle of 72.8 (2) . In 2 they have an unsymmetrical orientation, which appears to be the result of rotation of the germanido ligand to reduce repulsion between the Cp* and Ge-CH 3 groups, and the mean planes of the rings intersect at an angle of 66.1 (1) .

Supramolecular features
Compound 1 crystallizes in the space group C2/c with Z 0 = 1. Each molecule makes short (less than the sum of the van der Waals radii) contacts to six neighboring molecules. Two of these neighbors interact through donating or receiving a weak  Table 1 Selected geometric parameters (Å , ) for 1 and 2.

Figure 2
Packing plot viewed down c showing complementary interactions between 3,5-(CF 3 ) 2 C 6 H 3 and Cp* rings in 1. Dashed green lines indicate short (less than the sum of the van der Waals radii) contacts. Elements color coded as in Fig. 1. Axes color coded as follows: a = red, b = green.

Figure 1
50% probability ellipsoid plots of compounds 1 (left) and 2 (right). The Ge-H hydrogen atom in 1 is shown as a green circle, all other H atoms and minor disordered parts are omitted for clarity. Elements are color coded as follows: C = black, F = yellow-green, Ge = dark blue, I = purple, U = dark green. Ge-HÁ Á ÁI hydrogen bond (Table 2), which forms the basis of an infinite chain parallel to c. The two 3,5-(CF 3 ) 2 C 6 H 3 rings bonded to the Ge atom form a cavity, which complements the shape of the two Cp* groups, resulting in two neighboring molecules encapsulating or residing within this cavity and forming chains parallel to the b-axis direction (Fig. 2). The layers formed by these two interactions stack along the c-axis direction with adjacent layers making contact through likelike interactions between -CF 3 or Cp* groups, which are likely only weakly attractive or repulsive in nature. The phenyl rings are unsymmetrical in their interactions with one making a larger number of short contacts; the ring which makes fewer contacts has rotational disorder of both -CF 3 groups, which could be modeled over two positions in one case [modeled at occupancies of 0.536 (8) and 0.464 (8)] and is indicated by the shape of the displacement ellipsoids in the other case. Compound 2 crystallizes in the monoclinic space group C2/c with Z 0 = 1. Each molecule makes short contacts to seven neighboring molecules. One neighboring molecule interacts to form a centrosymmetric dimer through ion-dipole interactions between Cp* -CH 3 and aromatic C atoms. A second neighboring molecule also interacts across an inversion center through similar interactions between the other Cp* ligand and one of the phenyl rings. The remaining molecules only interact through C-HÁ Á ÁF contacts ( Table 2). The interactions involving the Cp* ligands appear to be the strongest and organize the molecules into tightly packed layers which are parallel to the (001) family of planes (Fig. 3) , and these layers are bridged through the C-HÁ Á ÁF interactions into a threedimensional network. As with 1, one of the 3,5-(CF 3 ) 2 C 6 H 3 rings does not participate as strongly in intermolecular interactions and has disordered -CF 3 groups, one of which could be modeled over two positions (occupancies 0.75 and 0.25).

Figure 5
Synthesis of compound 2.

Figure 3
Packing plot showing formation of close-packed layers in 2. Dashed green lines indicate short contacts. Elements color coded as in Fig. 1; unit-cell axes color coded as in Fig. 2. and the product recrystallized from a saturated toluene solution at 248 K.
The reaction of (C 5 Me 5 ) 2 U(CH 3 ) 2 with one equivalent of H 2 E[3,5-(CF 3 ) 2 C 6 H 3 ) 2 ] (Bender IV et al., 1997) in toluene at 333 K produces a dark-red solution (Fig. 5). The solution was allowed to stir overnight after which the solvent was removed, and the product recrystallized from a saturated toluene solution at 248 K. No byproducts could be found that led us to a plausible mechanism of C-F bond activation.

Refinement
The crystal structure of 1 was solved by an iterative dual space approach as implemented in SHELXT. All full occupancy non-hydrogen atoms could be located from the difference map refined anisotropically. In one of the disordered -CF 3 groups, two sets of fluorine atoms could be located from the difference map. The other -CF 3 group on the same molecule has prolate ellipsoids, which indicates disorder of this group as well, but attempts to model additional F-atom positions using chemically reasonable difference map peaks resulted in extremely poor geometries and displacement parameters. All C-F distances for this molecule were restrained to 1.33 (1) Å , and the intramolecular FÁ Á ÁF distances were restrained to be equal within AE 0.01 Å . For the -CF 3 group modeled over two positions, the three pairs of related F atoms each had their anisotropic displacement parameters constrained to be equal.
The occupancies of the major and minor parts refined to 53.6 and 46.4% (AE 0.8%) and were constrained to sum to 100%. The H atom bonded to Ge was located from the difference map, its coordinates were allowed to refine, and its isotropic displacement parameter was constrained to ride on the carrier atom. The structure also contained large residual difference map peaks in chemically non-reasonable positions. Some of these peaks occur at distances from the U atom very close to the U-I bond and have x or y coordinates equal to the I atom. Given the layer packing of this structure, these peaks most likely correspond to packing defects where layers are occasionally shifted relative to each other resulting in superposition of molecules related by rotation or reflection. These peaks could not be modeled, but truncating the data to a resolution of 0.77 Å greatly reduced their intensity.
The crystal structure of 2 was solved by an iterative dual space method as implemented in SHELXT. All non-hydrogen atoms were located from the difference map and refined anisotropically. For the disordered -CF 3 group both sets of F atoms were located from the difference map. The occupancies were manually adjusted until the isotropic thermal parameters were approximately equal, which occurred at 75% for the major part and 25% for the minor part. The major part could be refined anisotropically without restraints; the minor part failed to converge in an anisotropic refinement and was left isotropic.

Bis[3,5-bis(trifluoromethyl)phenyl-2κC 1 ](hydrido-2κH)(iodido-1κI)bis[1,1(η 5 )pentamethylcyclopentadienyl]germaniumuranium(Ge-U) (1)
Crystal data  Special details Geometry. All esds (except the esd in the dihedral angle between two l.s. planes) are estimated using the full covariance matrix. The cell esds are taken into account individually in the estimation of esds in distances, angles and torsion angles; correlations between esds in cell parameters are only used when they are defined by crystal symmetry. An approximate (isotropic) treatment of cell esds is used for estimating esds involving l.s. planes.

Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Å 2 )
x y z U iso */U eq Occ.