Research Article: Deciphering the hydrogen-bonding scheme in the crystal structure of tri­phenyl­methanol: a tribute to George Ferguson and co-workers

Date Published: September 01, 2019

Publisher: International Union of Crystallography

Author(s): Tomasa Rodríguez Tzompantzi, Aldo Guillermo Amaro Hernández, Rosa Luisa Meza-León, Sylvain Bernès.

http://doi.org/10.1107/S2053229619010714

Abstract

The disordered crystal structure of tri­phenyl­methanol features tetra­hedral chiral clusters formed through weak hydrogen bonds, leading to the formation of three-dimensional supra­molecular motifs having left or right handedness.

Partial Text

The hy­droxy group is known as one of the most efficient nodes for the formation of hydrogen bonds, as a consequence of the polarization of the O—H bond, and also because it can behave both as donor and acceptor for building intra- or inter­molecular bonds. In this context, the emblematic donor–acceptor mol­ecule is water, and many compounds have been crystallized as hydrates, in which the lattice water mol­ecules contribute to a significant part of the crystal free energy (Batsanov, 2018 ▸); currently, almost 13% of the structures deposited in the Cambridge Structural Database are hydrates (CSD, Version 5.40, updated February 2019; Groom et al., 2016 ▸). The situation is a bit less favourable in the case of alcohols (RO—H), especially for tertiary alcohols having the hy­droxy group surrounded by bulky hydro­carbon groups. For example, three hydrates for tert-butanol, (CH3)3COH, have been successfully characterized [namely the dihydrate and hepta­hydrate (Mootz & Stäben, 1993 ▸), and the deca­hydrate (Dobrzycki, 2018 ▸)], while tri-tert-butyl­methanol, [(CH3)3C]3­COH, has probably never been crystallized, although it has been studied in the solid state (Malarski, 1974 ▸). Although this mol­ecule is stable, it is not able to form stabilizing inter­molecular O—H⋯O hydrogen bonds, because of the steric hindrance of the three tert-butyl groups surrounding the OH donor group (Majerz & Natkaniec, 2006 ▸).

The asymmetric unit of the trigonal cell includes two disordered parts with site-occupancy factors converging at 153 K towards 0.7436 (17) (mol­ecules A and B hereafter) and 0.2564 (17) (mol­ecules C and D hereafter), close to the occupancies reported by Ferguson et al. of 0.71 and 0.29. Each part contains two independent mol­ecules, one of which has the σ C—O bond lying on the threefold axis in the space group R, while the other is located in a general position. The arrangement of these four disordered mol­ecules generates overlapped phenyl rings in the asymmetric unit, making the refinement of displacement parameters a tedious task (see §2.2). However, the mol­ecular structure based on data collected at 153 K can be considered as satisfactory, although the refinement was carried out with restrained geometry for the phenyl rings. The refinement based on data collected at 295 K is not as easy, since the scattering power of the crystal decreases dramatically: the fraction of observed data [I/σ(I) > 2] drops from 50% at 153 K to 36% at 295 K. However, the structure is essentially unmodified, and occupancies for the disordered parts refined to 0.761 (3) and 0.239 (3). At both temperatures, all non-H atoms could be refined anisotropically (see Fig. 2 ▸, right), after which phenyl H atoms were placed in idealized positions.

Tri­phenyl­methanol is a small simple mol­ecule with a structure unexpectedly difficult to refine compared, for example, to that of tri­phenyl­silanol (Bowes et al., 2002 ▸). It is worthwhile to consider the evolution of X-ray diffractometry over the last 25 years, using tri­phenyl­methanol as a benchmark (Table 4 ▸; data at room temperature were retained in order to avoid biases). The time taken for data collection is almost identical in the three cases, ca 20–30 h, and there is little doubt that the diffracted samples were of similar quality. Over time, a steady progress is noted. Development of new technologies for the detection of scattered X-ray photons seems to be the key point, in such a way that location of tiny fractions of electrons in the crystal space is now routinely affordable, outside the multipole density formalism. There is indeed a consensus that the hybrid pixel detectors with CdTe or GaAs sensors represent the ultimate state-of-art technology in this field, since they detect all scattered X-rays, with 100% efficiency and without any noise (Allé et al., 2016 ▸). For the herein presented study, a Pilatus detector was used, with a 1000 µm-thick silicon sensor, which has a quantum efficiency of 50% for 22.2 keV photons (Ag Kα radiation).

 

Source:

http://doi.org/10.1107/S2053229619010714

 

Leave a Reply

Your email address will not be published.