Long-Range Supramolecular Synthon Isomerism: Insight from a Case Study of Vinylic Tellurium Trihalides Cl(Ph)C=C(Ph)TeX₃ (X = Cl, I)

Abstract

A slight modification of the synthetic procedure resulted in a new (Cc) polymorph of vinylic tellurium-trichloride Z-Cl(Ph)C=C(Ph)TeCl₃ (1, β-form) which is stabilized by Te⋯Cl chalcogen bonds, assembling its molecules into the zigzag chains. Such a packing motive is in contrast to the known (Pca2₁) polymorph of Z-Cl(Ph)C=C(Ph)TeCl₃ (1, α-form, CCDC refcode: BESHOW), which is built upon Te⋯π(Ph) chalcogen bonded chains. We noted a similar case of [Te⋯halogen] vs. [Te⋯π(Ph)] supramolecular synthon polymorphism in its triiodide congener Z-Cl(Ph)C=CPh(TeI₃) (2, α and β-polymorphic forms). Quantum chemical calculations of the intermolecular interaction and lattice energies for 1α–β and 2α–β supported the assumption that α is thermodynamic while β is a kinetic form. Kinetic forms 1β and 2β are isostructural (Cc), while the thermodynamic forms 1α (Pca2₁) and 2α (P2₁/c) are not and feature an unusual example of long-range supramolecular synthon module isomerism. In other words, 1α–2α pairs demonstrate very similarly to isostructural Te⋯πPh ChB stabilized chains, which are further packed differently relative to each other, following different angular geometry of type-I Cl⋯Cl and type-II I⋯I halogen bonding. These structural considerations are backed by quantum chemical calculations that support the proposed hierarchy of primary and secondary supramolecular synthons and the assignment of α and β as thermodynamic and kinetic forms, respectively.

1. Introduction

A growing interest in halogen bonding (HaB) [1] and other σ-hole interactions [2,3] in the past two decades has provided a more profound understanding of chemical bonding and enriched the inventory and scope of crystal engineering [4,5,6,7,8]. In most case, these specific and directional interactions can be considered as the extended case of hypervalent 3c–4e interactions, allowing the consideration of intermolecular interactions in terms of molecular orbitals. This is particularly true and important for the heavy main group p-elements so that the structural chemistry of, say, organotellurides is extremely rich owing to the pervasive tendency of Te for the specific intermolecular interactions spanning the 3c–4e (hypervalent) and chalcogen bonding [9,10,11]. Attractive intermolecular interactions between the electrophilic atoms of Te and nucleophilic halogens (or other HaB-acceptors) are examples of chalcogen bonding (ChB, “a sister of halogen bond” [12]) and are so frequent in the solid state that their absence is sometimes more notable than their presence [13].

Recently, we have investigated the interaction of ferrocene with Ph₂C₂(Cl)TeCl₃ and described a series of ferrocenium cocrystalline salts with partly hydrolyzed Ph₂C₂(Cl)TeCl₃ [14]. We were quite surprised to notice that characteristic Te⋯Cl intermolecular interactions are absent in the crystal structure of Ph₂C₂(Cl)TeCl₃ as reported earlier [14] (1, hereinafter referred to as the α-form), so that only the Te⋯πPh-specific packing motif or supramolecular synthon (further abbreviated as SS) can be found in it. Later, during the routine PXRD analysis of Ph₂C₂(Cl)TeCl₃ sample synthesized for this project by a slightly modified procedure, we found (and were quite surprised once again) that we have a new polymorph of Ph₂C₂(Cl)TeCl₃ on our hands (hereinafter referred as the 1β), which, according to the SC-XRD analysis, features the anticipated Te⋯Cl SS instead of Te⋯πPh. This suggested the preparation of triiodide congeners Ph₂C₂(Cl)TeI₃ and a closer comparative examination of the polymorphism and peculiarities of the crystal structure in this series.

2. Materials and Methods

All reactions and manipulations were performed using the standard Schlenk techniques under an inert atmosphere of pure nitrogen or argon. Solvents and SO₂Cl₂ were purified, dried and distilled in argon atmosphere before use. Commercial reagents (Ph₂C₂, Te, KI) were used without additional purification. TeCl₄ was prepared according to the procedure reported by Petragnani et al. [15].

2.1. Preparation of Trichloro (Z)-2-Chloro-l,2-diphenylvinyl-tellurium(IV) Z-Cl(Ph)C=C(Ph)TeCl₃ (1α and 1β)

Powdered Te (1.37 g, 10 mmol) was refluxed with neat SO₂Cl₂ (10 mL, excess) for 3 h. After washing the solid residue with dried hexanes, as described in the original procedure for the preparation of TeCl₄ [16], the whole portion of crude TeCl₄ was refluxed with Ph₂C₂ (1.9 g, 10 mmol) in CCl₄ (5 mL) for 2 h. Further slow cooling of the reaction mixture afforded a yellow crystalline conglomerate, which was crushed and washed with cold hexane (3 × 10 mL), then dried in vacuum. It consisted of the crystals suitable for a single crystal and powder XRD analysis and was defined as a new (β polymorph) form of Z-Cl(Ph)C=C(Ph)TeCl₃ (1β). Crude 1β resulted in 1α upon recrystallization from hot CCl₄.

2.2. Preparation of Triiodo [(Z)-2-Chloro-l,2-diphenylvinyl-tellurium(IV) Cl(Ph)C=C(Ph)TeI₃ (2)

The light yellow solution of 1β (0.45 g, 1 mmol) in 5 mL of acetone was stirred with powdered KI (1.7 g, 10 mmol) for 12 h. The resulting dark red reaction mixture was dried in a vacuum, washed with hexane and extracted with CH₂Cl₂ (5 × 3 mL). The dark red extract was concentrated with hexane (3 mL) to 1/4 of the initial volume and kept at 4 °C for 12 h. Dark red crystalline precipitate was separated, washed with cold hexane, dried in a vacuum and used for single crystal X-ray investigation. The sample was defined as the isomorphic to 1β polymorphic form of Z-Cl(Ph)C=C(Ph)TeI₃ (hereinafter referred to as 2β). Further concentration and cooling of the mother liquor produced an additional quantity of dark red crystalline precipitate of 2β.

Recrystallization of 2β from a diluted CH₂Cl₂/hexane (1:3) solution, as well as slow vapour diffusion of pentane into its solution in CH₂Cl₂ at 4 °C, afforded dark red crystals of a new polymorph of 2, featuring structural resemblance with 1α (hereinafter referred as 2α).

2.3. X-ray Crystallography

2.3.1. Single-Crystal XRD

Suitable X-ray quality crystals of 1–2 were obtained directly during the preparation or recrystallization procedures (see preparation details). A Bruker SMART APEXII diffractometer equipped with a graphite-monochromated Mo Kα radiation (0.71070 Å) was used for cell determination and intensity data collection for the cocrystals 1–2. The data were collected by the standard phi–omega scan techniques and were reduced using SAINT v8.37A (Bruker, Billerica, MA, USA, 2015). The SADABS (Bruker, 2016) software was used for scaling and absorption correction. The structures were solved by direct methods and refined by full-matrix least-squares against F² using Olex2 and SHELXTL software (Bruker, Madison, WI, USA) [17,18]. Non-hydrogen atoms were refined with anisotropic thermal parameters. All hydrogen atoms were geometrically fixed and refined using a riding model. Atomic coordinates and other structural parameters of the reported cocrystals have been deposited at the Cambridge Crystallographic Data Center CCDC 2125950 (1β), CCDC 2125951 (2α) and CCDC 2125949 (2β) contains the supplementary crystallographic data for this paper.

2.3.2. PXRD

Powder XRD patterns were obtained at ambient conditions on a Bruker D8 Advance automatic X-ray diffractometer equipped with a Vario attachment and a Vantec linear coordinate detector (CuKα1 radiation, λ = 1.54063 Å, a curved Johansson monochromator; the X-ray tube mode was 40 kV and 40 mA). The samples were ground and deposited onto a silicon plate without strong pressing. The diffraction patterns were recorded in the Bragg–Brentano geometry for 2θ ranges of 5–90° or 5–60° with a step size of 0.008° and 1 or 4 s per step collection time. The samples were rotated in their planes at a rate of 15 rpm to eliminate the influence of preferred orientation and average data. The PXRD diffraction data were processed using the EVA program package (Bruker AXS, 2005) and Bruker TOPAS 5 software [19]. For the calculation of theoretical diffraction patterns, the data from X-ray single-crystal experiments were used. Crystallographic data for 1α were taken from the CSD (refcod BESHOW).

2.4. Intermolecular Energy Computations

Total pairwise energies of interactions between molecules for 1–2α and 1–2β and subsequent energy framework [20] generation was performed in Crystal Explorer 21.5 (TONTO, B3LYP-DGDZVP) Crystal Explorer [21] for all unique molecular pairs in the first coordination sphere of a molecule (5 Å) using experimental crystal geometries. The lattice energy of 1–2α, 1–2β was estimated as the sum of all unique bimolecular interactions for each independent molecule in the 25 Å coordination sphere [22].

3. Results and Discussion

3.1. Preparation and Crystal Structure of New Polymorph of Ph(Cl)C=C(Ph)TeCl3 (1β)

Modification of the reported preparation of Cl(Ph)C=C(Ph)TeCl₃ [16] by reducing the quantity of the solvent (CCl₄) produced a solid agglomerate of quite well-formed crystals of Cl(Ph)C=CPh(TeCl₃) (1β, Cc) after cooling of the refluxed reaction mixture down to room temperature. Such a condition, usually considered to favour the formation of kinetic crystals, allowed the new polymorph, featuring the 1D catemer chains assembled by Te⋯Cl ChBs (3.449(2)Å, Figure 1b), in addition to the known, presumably thermodynamic polymorph (1α, Pca2₁, BESHOW [16])), which has Te⋯π(Ph) chalcogen bonded chains (Figure 1a).

3.2. Preparation and Crystal Structure of Ph(Cl)C=C(Ph)TeI₃

Tellurium tetrachloride (TeCl₄) is the only tellurium tetrahalide which is sufficiently reactive for addition to the C≡C triple bond of Ph₂C₂, affording Z-Cl(Ph)C=C(Ph)TeCl₃ (1). We have prepared corresponding triiodide Z-Cl(Ph)C=C(Ph)TeI₃ (2) by treatment with acetone solution of 1 with powdered KI at room temperature (Figure 2).

Crystallization of the CH₂Cl₂ extract of the reaction mixture afforded the isomorphic to the 1β form of Cl(Ph)C=C(Ph)TeI₃ (2β, Figure 3a). Similar to 1β, its recrystallization afforded 2α crystals with a similar Te⋯π(Ph) chain pattern. The latter is not exactly isomorphic to 1α, but can be considered as a thermodynamic form of Cl(Ph)C=C(Ph)TeI₃ (Figure 3b, see Section 3.3 and Section 3.4)

It is noteworthy that trichloride 1 reacts with acetone in a moderate yield of 54% only at reflux for 4 h [23] so that only known polymorphic form 1α (BESHOW [16]) was recovered after the overnight stirring of 1β solution in acetone at room temperature and no side reactions with acetone were noticed during the overnight stirring of powdered KI in 1β solution in acetone (see Figure 2 and Section 2.1).

As a final remark on the molecular structure of 1–2, we mention that Te atoms in these molecules seem to have pseudo-trigonal bipyramid (Ψ-TBP) surroundings, where the fifth position is occupied by intermolecular Te1⋯π(Ph) (α) or Te1⋯X3 (X = Cl (1β), I (2α) chalcogen bonds (see Figure 1 and Figure 3). However, a closer look, motivated by our recent observations on Te geometry [9] drew attention to the possible intramolecular Te(1) ⋯Cl(1) ChB (average distance ~3 Ǻ), so that the sixth position is occupied or shielded by vinylic Cl atom and Te geometry in molecules 1–2 can be described as distorted octahedron, but not Ψ-TBP.

3.3. Supramolecular Organization Isomerism

Conformations of the isolated (Cl)PhC=CPh(TeX₃) molecules in the polymorphic pairs 1α (Pca2₁)/1β (Cc) and 2α (P2₁/c)/2β (Cc) are nearly identical (Figure 4), so we can consider them not as a case of conformational isomerism, but of supramolecular synthon isomerism [24,25].

At the same time, crystals 1β (Cc) and 2β (Cc) are isomorphic, while 1α (Pca2₁) and 2α (P2₁/c) are not. Crystals 1α and 2α feature isostructural, [Te⋯πPh] stabilized chains (Figure 5), but these very isostructural chains are further packed differently relative to each other, following different angular geometry of type-I Cl⋯Cl HaB and type-II I⋯I HaBs (Figure 6).

Although 1α and 2α are not isomorphic, the isostructural chain modules (Figure 5a) result in close values of the respective unit cell dimension (8.2010(8) and 8.2395(5)) with which they are parallel (Figure S1, Supplementary Materials). Further aggregation of these chain modules into the 3D structure is achieved by inter-chain Cl⋯Cl and I⋯I HaBs, but owing to the different tendencies of iodine and chlorine atoms for HaB, the arrangement of the chain modules is different in 1α and 2α (Figure 6). Easily polarized iodine atoms provide type-II I1⋯I2 XBs in 2α (Te1-I1⋯I2 ~106° Te1-I2⋯I1 ~174°, see Figure 6b), which direct the stretching of the respective unit cell dimension in 2α (27.9101(17)Å) as compared to 21.031(3) Å in 1α, where type-I Cl2⋯Cl3 XBs does not have any notable structure-directing effect (Te1-Cl2⋯Cl3~128°, Te1-Cl3⋯Cl2~123°, Figure 6a).

The observation that crystals 1α and 2α are built of isostructural chains suggests that [Te⋯πPh] may be a primary supramolecular synthon and therefore, these [Te⋯πPh] chains themselves are the primary supramolecular synthon modules. Their further association is built upon the secondary [Cl⋯Cl (type-I)] and [I⋯I (type-II)] SSs, respectively, and owing to the difference between these secondary SSs, we can speak of the long-range synthon module isomerism in 1–2. This is also an illustrative example of the 1D-to-2D stage of the Kitaigorodsky Aufbau Principle (KAP) [26,27,28], which anticipates the stepwise growth of dimensionality in the process of crystal formation.

3.4. Energy Frameworks of Ph(Cl)C=C(Ph)TeX₃ (X = Cl, I)

Certainly, short intermolecular contacts themselves do not imply a strong interaction [29] or a crystal structure-determining motif [30]. The functionality of Crystal Explorer software [21] favourably combines fast calculations of pairwise intermolecular interaction energies (B3LYP DGDZVP) and visualization of their magnitude in molecular clusters as Energy Frameworks [20]. This allows the construction of energetically based molecular packing patterns that may match, and thus support, patterns derived from evident short contacts or, conversely, can reveal overlooked or non-obvious supramolecular patterns [30]. Such calculation of the intermolecular interaction energy (B3LYP-DGDZVP TONTO/Crystal Explorer 21.5) in 1 and 2, demonstrates that short Te⋯Cl and Te⋯C_Ph contacts are observed in the pairs of molecules with the maximum binding energy (1β) or in the third-ranking (1α) in the top three of the strongest intermolecular interactions in the respective crystal (Figure 7a and Figure 8b). Energy frameworks visualize these interactions as straight (1α) and zigzag (1β) chains (Figure 6.)

The orthogonal-to-chain direction cut of the energy framework of 1α and 2α (Figure 6) shows the interactions between the chains. built upon the secondary [Cl⋯Cl (type-I)] and [I⋯I (type-II)] supramolecular synthons, respectively. It also demonstrates that strong electrostatic interactions (mostly contributed by type-II I⋯I HaBs) dominate over the dispersion interactions in 2α (Figure 8).

Although the Te⋯Cl pair in 1α appears as the third-ranking interaction (−24,6 kJ/mol, Figure 7a and Figure 8b, Table S2), it arises in the simple translation-related 1D chain. Such a fundamental symmetric relationship to the translation or screw axis is another powerful structure-determining factor [31]. Thus chemically meaningful and specific Te⋯Cl and Te⋯πPh chalcogen bonds in 1–2 could be the structure determining interactions which provide the chemical recognition, which in turn governs the supramolecular association at the earliest, kinetic stages of crystal genesis. This suggests that structure-directing factors are defined not merely by the strongest intermolecular interactions, but by their combination with the symmetry operators resulting from the straightest chains (e.g., simple translation).

It should be mentioned that Te⋯X and Te⋯πPh interactions in 1–2 are the strongest, but not outstanding, and are just 2–4 kJ/mol stronger than the second-ranking interactions (Tables S2–S5).

The resulting energy framework for 1, 2, built with a slightly lower cut-off (20 kJ/mol, Figure 9) presents a dense network of strong intermolecular interactions with no clear “weak links” where a second component can be inserted, providing a stronger lattice In contrast to the crystals, which have a significant gap between the chains or layers of the strongest interactions in their energy framework, so allowing the penetration of the guest molecules [32], we can assume that 1–2 are poor conformers for the design of co-crystals stabilized by X⋯A XBs or Te⋯X ChBs with the second ChB/XB-donor or acceptor component. Indeed, an attempted co-crystallization of 1 and 2 with the iconic HaB-donor (1,4-diiodo tetrafluorobenzene) and HaB-acceptor (pyridine) showed their low affinity for the formation of the ChB/HaB-stabilized two-component crystals.

Intermolecular energy calculations (TONTO CE-B3LYP/DGDZVP) demonstrated the ~1 kcal/mol difference in the lattice stabilization energies in 1α/1β and 2α/2β pairs (see Table S1). Such a difference is typical for the polymorphs’ and is supported by preliminary (based on their crystallization conditions) assignment of α and β as thermodynamic and kinetic forms, respectively.

Although the triiodide polymorphs 2α and 2β can be prepared by varying the crystallization conditions, 1β looks elusive, not to say, disappearing polymorph [33,34].

4. Conclusions

In this work, two new polymorphic pairs of Ph₂C₂(Cl)TeI₃ (1α and 1β) and Ph₂C₂(Cl)TeI₃ (2α and 2β) were structurally characterized. Although their presumably kinetic forms (1α and 2α) are isostructural and isomorphic, the structures of the respective thermodynamic forms (1β and 2β) demonstrate only partial similarity. They reveal isostructural primary chain-like modules which are subsequently packed differently following the different geometry and energetics of intermolecular I⋯I and Cl⋯Cl HaBs. This phenomenon can be defined as long-range aufbau supramolecular synthon modules isomerism.