Crystal structures, Hirshfeld atom refinements and Hirshfeld surface analyses of tris(4,5-dihydrofuran-2-yl)methylsilane and tris(4,5-dihydrofuran-2-yl)phenylsilane

The crystal structures of tris(4,5-dihydrofuran-2-yl)methylsilane (1) and -phenylsilane (2) display weak intermolecular C—H⋯O hydrogen-bonding interactions, which were analysed using Hirshfeld surface analysis. Futhermore, the crystal structures of (1) and (2) were refined using independent atom model (IAM) and Hirshfeld atom refinement (HAR) approaches

In the 1980s, Lukevits and co-workers first introduced the dihydrofuranyl group (DHF) as a substitutable silicon-carbon leaving group (Gevorgyan et al., 1989). The DHF group allows substitution by a number of nucleophiles including hydrides, lithiated amides, lithium alkyls and alcohols (Lukevits et al., 1993). Multiple nucleophilic substitutions using chlorosilanes show high reactivity and low selectivity. In general, the Si-O bond shows high reactivity and selectivity compared to the less or even non-reactive Si-C bond. Nonetheless, the DHF group shows a significant increase in reactivity and selectivity in the bond cleavage of Si-C bonds, which can extend the selectivity profile of functionalized organosilanes . Furthermore, the pre-coordination by a methoxy group plays an important role in the control of reactions with metalcontaining nucleophiles and leads to the question of whether this also applies to the DHF group (Barth et al., 2019). In order to understand the coordination possibilities, the alignment of the dihydrofuranyl group and thus the arrangement of the oxygen atoms in the crystal structure are interesting. In this ISSN 2056-9890 context, we here report the crystal structures of 1 and 2, both refined on basis of the independent atom model (IAM) and a Hirshfeld atom refinement (HAR) approach.

Structural commentary
The molecular structure of compound 1 is illustrated in Fig. 1, and selected bond lengths and angles using the results of IAM and HAR refinements are given in Table 1. In the molecule of 1, the Si-C bond lengths of the silicon-DHF groups are in a typical range and slightly longer than the silicon-methyl bond length. However, all Si-C bonds are as expected (Allen et al., 1987). The silicon atom in 1 has a slightly distorted tetrahedral environment, as shown by the deviation of the C-Si-C angles from the ideal value of 109.47 . This flexibility is often observed for Si-C single bonds (Otte et al., 2017;Glidewell & Sheldrick, 1971;Kü ckmann et al., 2005). The length of each of the C C double bonds of the DHF groups (C1 C2, C5 C6, C9 C10) also corresponds well with the literature (Allen et al., 1987).
The molecular structure of compound 2 is depicted in Fig. 2, and selected bond lengths and angles using the results of IAM and HAR refinements are collated in Table 2. The Si-C bond lengths and angles in the molecule of 2 differ only marginally from those of 1. In 2, there is a weak intramolecular C2-H2Á Á ÁO3 hydrogen-bonding interaction between the H2 atom of the C1 C2 group of one DHF molecule and the O3 atom of a neighbouring DHF group (Table 4), leading to a graph-set motif S 1 1 (6) (Etter et al., 1990). The Si-C bond lengths and C-Si-C angles of the IAM and HAR refinements coincide well. Slight deviations in the C C double bond of the DHF group can be observed and the trend shows that the double bonds from HAR refinement are slightly longer.

Hirshfeld atom refinements
The independent atom model (IAM) approach for crystalstructure refinement cannot reliably model bonding electrons or any distortion of the electron density. An approach that takes this into consideration is Hirshfeld atom refinement The molecular structure of compound 2 with displacement ellipsoids drawn at the 50% probability level.
In previous (unpublished) structure refinements of compounds with dihydrofuranyl rings performed by our group, we observed slight disorders of the oxygen atom and the methine atom of the dihydrofuranyl ring. Therefore, results of HARs for such compounds are interesting in order to draw conclusions about the residual electron densities to exclude and/or model disorder. For 1 and 2, the minimum and maximum values of residual electron density are significantly lower than those of IAM results (1: IAM Á min = À0.21 e Å À3 , Á max = 0.55 e Å À3 ; HAR Á min,max = AE0.21 e Å À3 ; 2: IAM Á min = À0.23 e Å À3 , Á max = 0.47 e Å À3 ; HAR Á min = À0.17 e Å À3 , Á max = 0.26 e Å À3 ). In all cases, the residual densities do not indicate any disorder. For compound 1, the residual electron density on the basis of the HAR refinement is close to O1 and H8A and for 2 is near C15 and H3B. Another aim of the Hirshfeld atom refinement was the accurate localization of hydrogen atoms. From a comparison of the C-H bond lengths of the methine groups using IAM and HAR approaches, it can be clearly observed that the C-H bonds of the HAR model are significantly longer than those of the AIM model (Table 5). Woiń ska et al. (2016) have already reported that the positions of hydrogen atoms and their corresponding bond lengths show a significantly improved agreement with neutron diffraction by refinement with HAR.
When using HAR, an improved R 1 value of 0.023 was observed for compound 1, compared to the refinement using IAM with an R 1 value of 0.035 (compound 2: R 1 for HAR = 0.024 versus IAM = 0.037).
HÁ Á ÁH, CÁ Á ÁH/HÁ Á ÁC and especially OÁ Á ÁH/HÁ Á ÁO are significant contributors, suggesting the relevance of these contacts in the packing arrangement of the crystal structure. The crystal packing of compound 2 is illustrated in Fig. 6 and shows a ribbon-like supramolecular network structure propagating along the b-axis direction. The molecules are linked by a C-HÁ Á ÁO hydrogen bond between the O2 i atom of a DHF group and the C16-H16 para group of the phenyl ring (Table 4), leading to the formation of chains with graphset motif C 1 1 (8 Hirshfeld surface analysis of 1 showing close contacts in the crystal. The weak hydrogen bond between oxygen atom O2 and the H11A hydrogen atom is labelled. [Symmetry codes: (i) Àx + 1 2 , y + 1 2 , Àz + 1 2 ; (ii) Àx + 1 2 , y À 1 2 , Àz + 1 2 ].   The crystal packing of compound 2 in a partial view along the b axis. C-HÁ Á ÁO hydrogen bonds are shown as dotted lines. [Symmetry code: (i) Àx + 1, y + 1 2 , Àz + 3 2 ].

Refinement
Crystal data, data collection and structure refinement details are summarized in Table 6. For the IAM approach using SHELXL (Sheldrick, 2015b), the H atoms were positioned geometrically (C-H = 0.95-1.00 Å ) and were refined using a riding model, with U iso (H) = 1.2U eq (C) for CH 2 and CH hydrogen atoms and U iso (H) = 1.5U eq (C) for CH 3 hydrogen atoms. Hydrogen atoms H6, H8A,B and H11A,B for compound 1 and H2 and H16 for compound 2 were refined independently.
HARs were performed with the HARt implementation in OLEX2 (Dolomanov et al., 2009), using the restricted Khom-Sham method with the basis set x2c-TZVP. The results of previous IAM refinements using -served as an input (Fugel et al., 2018). For the HAR approach, all H atoms were refined anistropically and independently.

Tris(4,5-dihydrofuran-2-yl)methylsilane (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 )  Special details Refinement. HAR makes use of tailor-made aspherical atomic form factors calculated on-the-fly from a Hirshfeldpartitioned electron density (ED) -not from spherical-atom form factors. The ED is calculated from a gaussian basis set single determinant SCF wavefunction -either SCF or DFT -for a fragment of the crystal embedded in an electrostatic crystal field. If constraints were applied they are defined by zero eigenvalues of the least-squares hessian, see the value of _refine_ls_SVD_threshold. Specify symmetry and Friedel pair averaging. Only reflections which satisfy the threshold expression are listed below, and only they are considered observed, thus the *_gt, *_all and *_total data are always the same.

Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Å 2 )
x y z U iso */U eq    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 Special details Refinement. HAR makes use of tailor-made aspherical atomic form factors calculated on-the-fly from a Hirshfeldpartitioned electron density (ED) -not from spherical-atom form factors. The ED is calculated from a gaussian basis set single determinant SCF wavefunction -either SCF or DFT -for a fragment of the crystal embedded in an electrostatic crystal field. If constraints were applied they are defined by zero eigenvalues of the least-squares hessian, see the value of _refine_ls_SVD_threshold. Specify symmetry and Friedel pair averaging. Only reflections which satisfy the threshold expression are listed below, and only they are considered observed, thus the *_gt, *_all and *_total data are always the same.