research papers\(\def\hfill{\hskip 5em}\def\hfil{\hskip 3em}\def\eqno#1{\hfil {#1}}\)

ISSN: 2052-2525

The origin of synthons and supramolecular motifs: beyond atoms and functional groups

crossmark logo

aCRM2, Université de Lorraine, CNRS, Nancy F-54500, France, bDepartment of Chemistry (NCI Laboratory), GITAM (Deemed to be University), 530045 Visakhapatnam, Andhra Pradesh, India, and cInstitut des Sciences Chimiques de Rennes, Université Rennes 1, UMR CNRS 6226, Campus de Beaulieu 35042, France
*Correspondence e-mail: enrique.espinosa@univ-lorraine.fr

Edited by A. Martín Pendás, Universidad de Oviedo, Spain (Received 20 December 2024; accepted 17 February 2025; online 7 April 2025)

This article is part of a collection of articles on Quantum Crystallography, and commemorates the 100th anniversary of the development of Quantum Mechanics. It is dedicated to Dr Slimane Dahaoui and Professor János G. Ángyán, who passed away before the preparation of this article.

A four-membered R22(4) supramolecular motif formed by S⋯S and S⋯I chalcogen-bonding interactions in the crystal structure of 4-iodo-1,3-di­thiol-2-one (C3HIOS2, IDT) is analysed and compared with a similar R22(4) motif (stabilized by Se⋯Se and Se⋯O chalcogen bonds) observed in the previously reported crystal structure of selena­phthalic anhydride (C8H4O2Se, SePA) through detailed charge density analysis. Our investigation reveals that the chalcogen-bonding interactions participating in the R22(4) motifs observed in the two structures have the same characteristic orientation of local electrostatic electrophilic⋯nucleophilic interactions while involving different types of atoms. We carried out Cambridge Structural Database searches for synthons and supramolecular motifs involving chalcogen-, halogen- and hydrogen-bonding (ChB, XB and HB) interactions. Geometrical characterizations and topological analyses of the electron density ρ(r) and its negative Laplacian function [L(r) = −∇2ρ(r)] indicate that all the bonding interactions forming the motifs are driven by local electrophilic⋯nucleophilic interactions between complementary charge concentration (CC) and charge depletion (CD) sites present in the valence shells of the atoms, regardless of the atoms and functional groups involved. The graph-set assignment Gda(n) (G = C, R, D or S), formerly developed by Etter [Acc. Chem. Res. (1990), 23, 120–126] for HB interactions, is a convenient way to describe the connectivity in supramolecular motifs based on electrophilic⋯nucleophilic interactions (such as ChB, XB and HB interactions), exchanging the number of atomic acceptors (a) and donors (d) with the number of nucleophilic (n: CC) and electrophilic (e: CD) sites, and the number of atoms building the motif n by m, leading to the new graph-set assignment Gen(m) (G = C, R, D or S). Geometrical preferences in the molecular assembly of synthons and other supramolecular motifs are governed by the relative positions of CC and CD sites through CC⋯CD interactions that, in most cases, align with the internuclear directions within a <15° range despite low interaction energies. Accordingly, beyond atoms and functional groups, the origin of recurrent supramolecular structures embedded within different molecular environments is found in the local electrostatic complementarity of electrophilic and nucleophilic regions that are placed at particular geometries, driving the formation and the geometry of synthons and supramolecular motifs by directional and locally stabilizing electrostatic interactions.

1. Introduction

Hydrogen bonding is the most commonly observed intermolecular interaction in crystals of organic compounds containing hydrogen atoms of acidic character. The use of rules about the formation of hydrogen bonds can aid in the design and analysis of multi-component crystals of organic compounds, such as co-crystals (Tiekink & Zukerman-Schpector, 2017[Tiekink, E. & Zukerman-Schpector, J. (2017). Editors. Multi-Com­ponent Crystals. Synthesis, Concepts, Function. Berlin: DeGruyter.]). However, so far, the formation of a co-crystal cannot be predicted using these rules, and it can only be verified after experimental analysis. Strategies put in place to build up co-crystals are mainly based on the formation of supramolecular building blocks called synthons (Desiraju, 1995[Desiraju, G. R. (1995). Angew. Chem. Int. Ed. Engl. 34, 2311-2327.]), which are defined as recurring intermolecular structural units found in crystal phases. Hence, studies on the competition between different supramolecular building blocks have been carried out to facilitate the prediction of co-crystal formation (Etter, 1991[Etter, M. C. (1991). J. Phys. Chem. 95, 4601-4610.]). The results of these studies suggest that the most energetically stable synthon is the one that appears in the arrangement of the co-crystal formed (Blagden et al., 2008[Blagden, N., Berry, D. J., Parkin, A., Javed, H., Ibrahim, A., Gavan, P. T., De Matos, L. L. & Seaton, C. C. (2008). New J. Chem. 32, 1659-1672.]). These hypotheses have been verified experimentally through comparisons of single-crystal structures (Vishweshwar et al., 2006[Vishweshwar, P., McMahon, J. A., Bis, J. A. & Zaworotko, M. J. (2006). J. Pharm. Sci. 95, 499-516.]).

A study carried out using the Cambridge Structural Database (CSD) (Allen, 2002[Allen, F. H. (2002). Acta Cryst. B58, 380-388.]) and the CONQUEST (version 1.17) software demonstrated that most co-crystals form preferentially by making use of heterosynthons rather than homosynthons (Groom & Allen, 2014[Groom, C. R. & Allen, F. H. (2014). Angew. Chem. Int. Ed. 53, 662-671.]). An homosynthon is characterized by intermolecular interactions between the same functional group, such as carb­oxy­lic acid⋯carb­oxy­lic acid or amide⋯amide interactions. An heterosynthon is characterized by intermolecular interactions between different but complementary functional groups, such as carb­oxy­lic acid⋯amide or carb­oxy­lic acid⋯pyridine. Since carb­oxy­lic acids are found in a large number of organic molecules and active pharmaceutical ingredients, they are among the most studied functional groups in crystal engineering (Desiraju, 1991[Desiraju, G. R. (1991). J. Appl. Cryst. 24, 265.]; Shan et al., 2002a[Shan, N., Bond, A. D. & Jones, W. (2002a). Tetrahedron Lett. 43, 3101-3104.]; Shan et al., 2002b[Shan, N., Batchelor, E. & Jones, W. (2002b). Tetrahedron Lett. 43, 8721-8725.]; Vishweshwar et al., 2005[Vishweshwar, P., McMahon, J. A., Peterson, M. L., Hickey, M. B., Shattock, T. R. & Zaworotko, M. J. (2005). Chem. Commun. 36, 4601-4603.]). Statistical analyses carried out using the CSD (Allen et al., 1999[Allen, F. H., Motherwell, W. D. S., Raithby, P. R., Shields, G. P. & Taylor, R. (1999). New J. Chem. 23, 25-34.]) and involving structures that only contain carb­oxy­lic acid functional groups have shown that the main molecular association follows the formation of the supramolecular homosynthon carb­oxy­lic acid⋯carb­oxy­lic acid. However, in the presence of competing functional groups such as amide or pyridine groups, carb­oxy­lic acid⋯pyridine and carb­oxy­lic acid⋯amide heterosynthons were found to be more favoured (Steiner, 2001[Steiner, T. (2001). Acta Cryst. B57, 103-106.]; Vishweshwar et al., 2003a[Vishweshwar, P., Nangia, A. & Lynch, V. M. (2003a). CrystEngComm, 5, 164-168.]; Vishweshwar et al., 2003b[Vishweshwar, P., Nangia, A. & Lynch, V. M. (2003b). Cryst. Growth Des. 3, 783-790.]; Nangia, 2010[Nangia, A. (2010). J. Chem. Sci. 122, 295-310.]). This empirical information on the complementarity between functional groups and their supramolecular building blocks could lead to guidelines for the design of co-crystals. Such guidelines would facilitate the selection of co-formers (namely, the appropriate molecules participating in the formation of co-crystals), which could then be used in experimental screens targeting forms of multi-component solids (Issa, 2011[Issa, N. (2011). PhD thesis. University College London, UK.]). The formation of co-crystals can be induced through the use of hydrogen-bond donors and acceptors that are part of the molecular entities forming the co-crystal, as well as the way they may interact through hydrogen bonding (Blagden et al., 2007[Blagden, N., de Matas, M., Gavan, P. T. & York, P. (2007). Adv. Drug Deliv. Rev. 59, 617-630.]). After a thorough study of preferential arrangements and HB patterns found in a large number of organic crystals, Etter made several observations to aid in the design of hydrogen bonds in solids and applied a graph-set assignment to define the morphology of hydrogen-bonded arrays differentiating the type of donors and acceptors that are present (Etter, 1990[Etter, M. C. (1990). Acc. Chem. Res. 23, 120-126.]). These graph-set symbols are of the form Gda(n), where G = C, R, D or S, denoting chain, ring, dimer (or other finite set) or intramolecular HB patterns, respectively, while the subscript d and superscript a indicate the number of donors and acceptors used in each motif, and n is the number of atoms in the unit.

The graph-set assignment Gda(n) is a very convenient way to describe the connectivity in supramolecular motifs based on electrophilic⋯nucleophilic interactions, generalizing the description of HB patterns to additional ones that originate from other types of non-covalent interactions. Indeed, after the initial use of synthons based on HB interactions, other supramolecular building units have been identified based on halogen bonds (Cavallo et al., 2016[Cavallo, G., Metrangolo, P., Milani, R., Pilati, T., Priimagi, A., Resnati, G. & Terraneo, G. (2016). Chem. Rev. 116, 2478-2601.]; Fourmigué, 2009[Fourmigué, M. (2009). Curr. Opin. Solid State Mater. Sci. 13, 36-45.]), and more recently on chalcogen bonds (Scilabra et al., 2019[Scilabra, P., Terraneo, G. & Resnati, G. (2019). Acc. Chem. Res. 52, 1313-1324.]; Vogel et al., 2019[Vogel, L., Wonner, P. & Huber, S. M. (2019). Angew. Chem. Int. Ed. 58, 1880-1891.]; Gleiter et al., 2018[Gleiter, R., Haberhauer, G., Werz, D. B., Rominger, F. & Bleiholder, C. (2018). Chem. Rev. 118, 2010-2041.]; Dhaka et al., 2024[Dhaka, A., Jeon, I.-R. & Fourmigué, M. (2024). Acc. Chem. Res. 57, 362-374.]; Dukhnovsky et al., 2024[Dukhnovsky, E. A., Novikov, A. S., Kubasov, A. S., Borisov, A. V., Sikaona, N. D., Kirichuk, A. A., Khrustalev, V. N., Kritchenkov, A. S. & Tskhovrebov, A. G. (2024). Int. J. Mol. Sci. 25, 3972.]). In all these cases, new supramolecular architectures exhibit specific electrophilic and nucleophilic regions at donor and acceptor molecules that lead to highly directional interactions. Electrophilic (charge-depleted: CD) sites in halogen bonds and chalcogen bonds are associated with σ-hole regions (Clark et al., 2007[Clark, T., Hennemann, M., Murray, J. S. & Politzer, P. (2007). J. Mol. Model. 13, 291-296.]; Politzer & Murray, 2017[Politzer, P. & Murray, J. S. (2017). Crystals, 7, 212.]) and are typically revealed by a plot of the molecular electrostatic potential (MESP) on isosurfaces of electron density in the external parts of molecules, which become significantly positive, as with acidic hydrogen atoms that are involved in HB interactions. On the other hand, good nucleophilic (charge concentration: CC) sites are typically associated with lone-pair regions of Lewis bases involved in HB, XB and ChB interactions, displaying negative MESP values and therefore ensuring the electrostatic complementarity with donor partners. Although additional types of non-covalent electrophilic⋯nucleophilic interactions have been described in the literature (Alkorta et al., 2020[Alkorta, I., Elguero, J. & Frontera, A. (2020). Crystals, 10, 180.]), this work will focus on ChB and XB interactions, and HB interactions to a lesser extent.

The IUPAC defined the chalcogen bond as a net `attractive interaction between an electrophilic region associated with a chalcogen atom in a molecular entity and a nucleophilic region in another, or the same, molecular entity' (Aakeroy et al., 2019[Aakeroy, C. B., Bryce, D. L., Desiraju, G. R., Frontera, A., Legon, A. C., Nicotra, F., Rissanen, K., Scheiner, S., Terraneo, G., Metrangolo, P. & Resnati, G. (2019). Pure Appl. Chem. 91, 1889-1892.]). While similar to the halogen bond as per the definition (Desiraju et al., 2013[Desiraju, G. R., Ho, P. S., Kloo, L., Legon, A. C., Marquardt, R., Metrangolo, P., Politzer, P., Resnati, G. & Rissanen, K. (2013). Pure Appl. Chem. 85, 1711-1713.]), characterizing a chalcogen bond can be complex in comparison. This is because the anisotropic distribution of ρ(r) around the Chsp3 atom can result in multiple CD sites suitable for ChB interactions, which can compete and sometimes merge (Shukla et al., 2020[Shukla, R., Dhaka, A., Aubert, E., Vijayakumar-Syamala, V., Jeannin, O., Fourmigué, M. & Espinosa, E. (2020). Cryst. Growth Des. 20, 7704-7725.]). In comparison, halogen atoms have only one CD region suitable for an XB interaction. This is the reason why intermolecular chalcogen bonds do not always form along the covalent bond direction of the chalcogen atom, as observed in halogen bonds. In the past few years, there has been a steady increase in both experimental and theoretical studies focusing on ChB interactions. This is because, in addition to their ability to form strong intermolecular and intramolecular interactions in crystal structures, chalcogen bonds also present potential applications in the fields of catalysis, organic synthesis and material design, among other fields (Mahmudov et al., 2017[Mahmudov, K. T., Kopylovich, M. N., Guedes da Silva, M. F. C. & Pombeiro, A. J. L. (2017). Dalton Trans. 46, 10121-10138.]; Bamberger et al., 2019[Bamberger, J., Ostler, F. & Mancheño, O. G. (2019). ChemCatChem, 11, 5198-5211.]; Dhaka et al., 2020[Dhaka, A., Jeannin, O., Jeon, I., Aubert, E., Espinosa, E. & Fourmigué, M. (2020). Angew. Chem. Int. Ed. 59, 23583-23587.]; Beau et al., 2023[Beau, M., Jeannin, O., Fourmigué, M., Auban-Senzier, P., Pasquier, C., Alemany, P., Canadell, E. & Jeon, I.-R. (2023). CrystEngComm, 25, 3189-3197.]). The formation of supramolecular synthons mediated by ChB interactions is also a very common phenomenon. However, this feature is relatively less explored compared with motifs involving hydrogen and halogen bonds. Investigations on chalcogenated molecules have shown the formation of robust Ch⋯O interactions with the electrophilicity of the chalcogen atom increasing in the order O < S < Se < Te (Brezgunova et al., 2013[Brezgunova, M., Lieffrig, J., Aubert, E., Dahaoui, S., Fertey, P., Lebègue, S., Ángyán, J. G., Fourmigué, M. & Espinosa, E. (2013). Cryst. Growth Des. 13, 3283-3289.]), which parallels the observed behaviour of X⋯O interactions along the series of halogen atoms F < Cl < Br < I (Geboes et al., 2015[Geboes, Y., Nagels, N., Pinter, B., De Proft, F. & Herrebout, W. A. (2015). J. Phys. Chem. A, 119, 2502-2516.]). Other studies have shown that the characteristics of the Ch⋯O interaction remained similar (Scilabra et al., 2019[Scilabra, P., Terraneo, G. & Resnati, G. (2019). Acc. Chem. Res. 52, 1313-1324.]) even when one type of chalcogen atom was replaced with another.

In the current study, the crystal structure of 4-iodo-1,3-di­thiol-2-one (C3HIOS2, hereafter called IDT) is investigated via theoretical charge density analysis. The presence of hydrogen, chalcogen and halogen atoms in the small IDT molecule allows an in-depth investigation of HB, ChB and XB interactions present in the crystal structure of this molecule. A significant structural aspect of IDT on which we focus is the formation of a four-membered motif assembled from S⋯S and S⋯I chalcogen bonds. Considering the expected positions of electrophilic and nucleophilic regions around the atoms involved, the motif in IDT is extremely similar to the recurrent R22(4) motif in selena­phthalic anhydride (SePA) (Brezgunova et al., 2013[Brezgunova, M., Lieffrig, J., Aubert, E., Dahaoui, S., Fertey, P., Lebègue, S., Ángyán, J. G., Fourmigué, M. & Espinosa, E. (2013). Cryst. Growth Des. 13, 3283-3289.]), which is formed by Se⋯Se and Se⋯O chalcogen bonds (see the scheme below[link]). This feature raises the question of the origin of similar supramolecular motifs involving different atoms. Is the origin of the motif associated with the atom or the functional group involved in the formation of the interactions building the supramolecular structure, or with the appropriate orientation of particular electrophilic and nucleophilic regions present in the interacting atoms? The formation of a similar motif in IDT and SePA strongly suggests that the latter holds true. Hereafter, the graph-set assignment Gda(n) developed by Etter (1990[Etter, M. C. (1990). Acc. Chem. Res. 23, 120-126.]) for HB interactions is generalized to any non-covalent interaction (such as HB, XB or ChB interactions) using the new notation Gen(m) (G = C, R, D or S), where the number of atomic acceptors a and donors d are exchanged for the number of nucleophilic n (CC) and electrophilic e (CD) sites, and the number of atoms building the motif n for m (to avoid a misunderstanding with the number of nucleophiles).[link]

[Scheme 1]

With the aim of identifying the structural parameters that correspond to the geometrical preferences shown in synthons (and in a more general way, in the formation of structurally similar non-covalent interactions), a detailed CSD search has been carried out for other supramolecular R22(4), R33(3) and bifurcated intermolecular motifs, as well as for intramolecular motifs. In addition, to investigate the origin of recurrent non-covalent inter-/intramolecular motifs and their geometries in depth, the topology of L(r) = −∇2ρ(r) has been also investigated for selected motifs to determine local nucleophilic (CC) and electrophilic (charge depletion, CD) sites in the valence shells of the atoms and to quantify their geometric orientation with respect to internuclear directions; this procedure has already been applied in previous work and described by the angle between the CC⋯CD and internuclear directions (Brezgunova et al., 2013[Brezgunova, M., Lieffrig, J., Aubert, E., Dahaoui, S., Fertey, P., Lebègue, S., Ángyán, J. G., Fourmigué, M. & Espinosa, E. (2013). Cryst. Growth Des. 13, 3283-3289.]; Shukla et al., 2020[Shukla, R., Dhaka, A., Aubert, E., Vijayakumar-Syamala, V., Jeannin, O., Fourmigué, M. & Espinosa, E. (2020). Cryst. Growth Des. 20, 7704-7725.]). Overall, the investigation carried out in this study establishes that face-to-face CC⋯CD interactions drive atomic orientations and are definitively more important than the atoms or functional groups involved in the formation of different supramolecular synthons or motifs.

2. Experimental and theoretical methods

2.1. Synthesis and crystal growth

4-Iodo-1,3-di­thiol-2-one (C3HIOS2, IDT) was synthesized from the corresponding thione 4-iodo-1,3-di­thiol-2-thione (Le Gal et al., 2016[Le Gal, Y., Lorcy, D., Jeannin, O., Barrière, F., Dorcet, V., Lieffrig, J. & Fourmigué, M. (2016). CrystEngComm, 18, 5474-5481.]) in CH3Cl/AcOH with Hg(OAc)2, as shown in earlier work (Domercq et al., 2001[Domercq, B., Devic, T., Fourmigué, M., Auban-Senzier, P. & Canadell, E. (2001). J. Mater. Chem. 11, 1570-1575.]). The sublimation method was used to obtain IDT single crystals of good quality (70°C sublimation temperature).

2.2. X-ray diffraction

Low-temperature high-resolution single-crystal X-ray diffraction measurements were performed using an Mo Kα radiation source (λ = 0.71073 Å) on an Enraf-Nonius diffractometer with a Kappa-CCD detector at T = 100 (4) K (Oxford Cryosystems N2 cooling device). The data collection strategy was determined and data reduction was carried out with CrysAlisPro CCD/RED. After absorption correction, the measured reflections were then sorted, scaled and merged using the SORTAV program (Blessing, 1995[Blessing, R. H. (1995). Acta Cryst. A51, 33-38.]). The structures were solved with SHELXS (Sheldrick, 2008[Sheldrick, G. M. (2008). Acta Cryst. A64, 112-122.]) and refined with SHELXL (Sheldrick, 2015[Sheldrick, G. M. (2015). Acta Cryst. C71, 3-8.]) incorporated in the OLEX2 program (Dolomanov et al., 2009[Dolomanov, O. V., Bourhis, L. J., Gildea, R. J., Howard, J. A. K. & Puschmann, H. (2009). J. Appl. Cryst. 42, 339-341.]). The molecular packing of IDT was drawn using Mercury (Macrae et al., 2020[Macrae, C. F., Sovago, I., Cottrell, S. J., Galek, P. T. A., McCabe, P., Pidcock, E., Platings, M., Shields, G. P., Stevens, J. S., Towler, M. & Wood, P. A. (2020). J. Appl. Cryst. 53, 226-235.]). See Table S1 in the supporting information for data collection and structure refinement details.

2.3. Computational electron density

Owing to the presence of several heavy atoms (namely, two sulfur atoms and one iodine atom) in a small organic molecule that crystallizes in the triclinic system with the non-centrosymmetric space group P1, the experimental dataset for IDT was of limited quality for experimental charge density analysis (Fig. S1 of the supporting information). For this reason, only periodic theoretical calculations were performed on the crystal structure of IDT based on the experimental geometry and obtained after the independent atom model (IAM) refinement. Ab initio periodic calculations using density functional theory (DFT) were then performed on the crystal structure of IDT with the Vienna Ab initio Simulation Package (VASP) (Kresse & Furthmüller, 1996a[Kresse, G. & Furthmüller, J. (1996a). Comput. Mater. Sci. 6, 15-50.]; Kresse & Hafner, 1993[Kresse, G. & Hafner, J. (1993). Phys. Rev. B, 47, 558-561.]; Kresse & Hafner, 1994[Kresse, G. & Hafner, J. (1994). J. Phys. Condens. Matter, 6, 8245-8257.]; Kresse & Furthmüller, 1996b[Kresse, G. & Furthmüller, J. (1996b). Phys. Rev. B, 54, 11169-11186.]). More precisely, we used the functional of Perdew, Burke and Ernzerhof (Perdew et al., 1996[Perdew, J. P., Burke, K. & Ernzerhof, M. (1996). Phys. Rev. Lett. 77, 3865-3868.]) as the exchange–correlation functional, and to ensure convergence a value of 900 eV was set for the plane-wave expansion of the wavefunction, as well as a k-points grid of 16 × 12 × 8 to sample the first Brillouin zone. The valence electron density was represented on a regular grid comprising 160 × 224 × 280 points along the crystallographic axes and was used to compute the structure factors. Note that the comparison of a previous experimental charge density model for C6Cl6 with the corresponding periodic ab initio theoretical DFT model obtained with VASP, using the same method as in the present work, led to extremely close results in intermolecular regions (Aubert et al., 2011[Aubert, E., Lebègue, S., Marsman, M., Bui, T. T. T., Jelsch, C., Dahaoui, S., Espinosa, E. & Ángyán, J. G. (2011). J. Phys. Chem. A, 115, 14484-14494.]).

DFT calculations were also undertaken at the B3LYP/Def2TZVPP level of theory on isolated molecules (monomers, dimer, trimers) in order to characterize the intermolecular interactions. The molecular structures were extracted from crystal structures and the Gaussian 16 program package was used for the calculations (Frisch et al., 2019[Frisch, M. J., Trucks, G. W., Schlegel, H. B., Scuseria, G. E., Robb, M. A., Cheeseman, J. R., Scalmani, G., Barone, V., Petersson, G. A., Nakatsuji, H., Li, X., Caricato, M., Marenich, A. V., Bloino, J., Janesko, B. G., Gomperts, R., Mennucci, B., Hratchian, H. P., Ortiz, J. V., Izmaylov, A. F., Sonnenberg, J. L., Williams-Young, D., Ding, F., Lipparini, F., Egidi, F., Goings, J., Peng, B., Petrone, A., Henderson, T., Ranasinghe, D., Zakrzewski, V. G., Gao, J., Rega, N., Zheng, G., Liang, W., Hada, M., Ehara, M., Toyota, K., Fukuda, R., Hasegawa, J., Ishida, M., Nakajima, T., Honda, Y., Kitao, O., Nakai, H., Vreven, T., Throssell, K., Montgomery, J. A., Peralta, Jr., J. E., Ogliaro, F., Bearpark, M. J., Heyd, J. J., Brothers, E. N., Kudin, K. N., Staroverov, V. N., Keith, T. A., Kobayashi, R., Normand, J., Raghavachari, K., Rendell, A. P., Burant, J. C., Iyengar, S. S., Tomasi, J., Cossi, M., Millam, J. M., Klene, M., Adamo, C., Cammi, R., Ochterski, J. W., Martin, R. L., Morokuma, K., Farkas, O., Foresman, J. B. & Fox, D. J. (2019). Gaussian 16. Revision C.01. Gaussian Inc., Wallingford, Connecticut, USA.]). Topological analyses were also performed using AIMAll (Keith, 2019[Keith, T. A. (2019). AIMAll (version 19.10.12). TK Gristmill Software, Overland Park KS, USA. https://aim.tkgristmill.com.]).

2.4. Multipolar refinement

The theoretical multipolar refinement was performed using the MoPro/MoProviewer software package (Jelsch et al., 2005[Jelsch, C., Guillot, B., Lagoutte, A. & Lecomte, C. (2005). J. Appl. Cryst. 38, 38-54.]; Guillot et al., 2014[Guillot, B., Enrique, E., Huder, L. & Jelsch, C. (2014). Acta Cryst. A70, C279-C279.]). In the multipolar refinements against the theoretically calculated structure factors (sin θ/λmax = 1.0 Å−1), the Slater-type radial functions were expanded up to the hexadecapole level (l = 4) for the I (nl, ξ: 8, 8, 8, 8 for l = 1, 2, 3, 4; 4.84 bohr−1) and S atoms (nl, ξ: 4, 4, 6, 8 for l = 1, 2, 3, 4; 3.8513 bohr−1), and up to octapole level (l = 3) for C (nl, ξ: 2, 2, 3 for l = 1, 2, 3; 3.0 bohr−1) and O atoms (nl, ξ: 2, 2, 3 for l = 1, 2, 3; 4.5 bohr−1). In the fitting against theoretical data, core electrons were omitted from the MoPro atomic definitions, since the VASP structure factors correspond to valence electrons only. The multipolar models fitted against the theoretical structure factors converged to the agreement factors wR(F) = 0.023 and R(F) = 0.041 (Table S1). The residual density map of the best theoretical model shows residual peaks centred at the I and S atoms [Fig. 1[link](a)]. In the same plane, the static deformation density Δρ(r) and L(r) = −∇2ρ(r) maps of the theoretical multipolar model of IDT are shown in Figs. 1[link](b) and 1[link](c), respectively. The Δρ(r) map shows that the highest density accumulation is on the C2—C3 bond (0.6 e Å−3) and three is a polarized electron distribution towards the C atoms along the C—S and C—I bonds. The lone pairs of S1 merge in a banana-shaped distribution (their separation is not observed at any contour level), whereas those of S2 are separated [see inset figures in Fig. 1[link](b)]. However, it is noticeable that a polarization of the electron density takes place towards one of the lone pairs in both S atoms. This effect could be explained by the participation of the S atoms in different intermolecular interactions above and below the molecular plane (S1⋯O1, S2⋯O1 and S1⋯I1 contacts, which result in the stacking of the molecules in columns). The multipolar model shows the depletion of the electron density (δ+) extending along the C—S bonds (Δρ varies from −0.25 to −0.10 e Å−3). In general, the deformation density map of the S atom exhibits very similar features to those observed earlier for the Se atom in seleno­phtalic anhydride SePA (Brezgunova et al., 2013[Brezgunova, M., Lieffrig, J., Aubert, E., Dahaoui, S., Fertey, P., Lebègue, S., Ángyán, J. G., Fourmigué, M. & Espinosa, E. (2013). Cryst. Growth Des. 13, 3283-3289.]). On the other hand, the deformation density of the I atom in IDT is qualitatively close to those previously described for Cl and Br atoms (Bui et al., 2009[Bui, T. T. T., Dahaoui, S., Lecomte, C., Desiraju, G. R. & Espinosa, E. (2009). Angew. Chem. Int. Ed. 48, 3838-3841.]; Aubert et al., 2011[Aubert, E., Lebègue, S., Marsman, M., Bui, T. T. T., Jelsch, C., Dahaoui, S., Espinosa, E. & Ángyán, J. G. (2011). J. Phys. Chem. A, 115, 14484-14494.]; Brezgunova et al., 2012[Brezgunova, M. E., Aubert, E., Dahaoui, S., Fertey, P., Lebègue, S., Jelsch, C., Ángyán, J. G. & Espinosa, E. (2012). Cryst. Growth Des. 12, 5373-5386.]), namely a δ+ region on iodine located on the prolongation of the C3—I1 bond and δ regions in the plane perpendicular to the bonding direction.

[Figure 1]
Figure 1
Maps calculated in the molecular plane and in the two perpendicular planes bis­ecting the C—S—C angles containing the S-atom lone pairs (inset figures) of IDT (theoretical model): (a) residual electron density at the resolution range 0 ≤ (sin θ/λ) ≤ 0.9 Å−1, (b) static deformation density Δρ(r) and (c) L(r) = −∇2ρ(r). Contours for maps (a) and (b) are at the ±0.05 e Å−3 level: full lines in blue are positive, dotted lines in red are negative. Contours for (c) in e Å−5 are on a logarithmic scale: red lines are positive and blue lines are negative.

2.5. The topology of ρ(r) and L(r) functions

Within the QTAIM framework (Bader, 1990[Bader, R. F. W. (1990). Atoms in Molecules: A Quantum Theory. Oxford University Press.]), topological analysis of ρ(r) is performed in order to extract significant information on bonding interactions present in molecular crystals. Accordingly, we have evaluated the topological and local energetic properties of ρ(r) at bond critical points (BCPs) for a detailed description of the intermolecular and non-covalent intramolecular interactions investigated in this study. Topological analysis of L(r) = −∇2ρ(r), with the determination of its own CPs [different from those of ρ(r)], has also been performed in order to identify CD and CC regions present in the valence shells of the atoms participating in the interactions. The local electrostatic complementarity of CD and CC regions interacting through space will help in characterizing the electrophilic–nucleophilic interactions (CD⋯CC) which are the origin of preferred orientations in molecular assemblies.

In the valence shells of atoms, (3,−3)/(3,+3) CPs correspond to 3D nucleophilic/electrophilic sites [the L(r) distribution is locally concentrated/depleted at the CP along the three main directions], whereas the interpretation of saddle (3,+1)/(3,−1) CPs remains more complex because it depends on the relative positions of other (3,−3)/(3,−1)/(3,+1)/(3,+3) CPs around them. This feature is a consequence of the continuity of the L(r) function in space, leading to alternating local maxima and minima either along directions or on surfaces. For instance, a (3,−1) CP can be found as a maximum/minimum along a 1D gradient field line of the L(r) distribution linking two (3,+1) or two (3,−3) CPs. On the other hand, a (3,+1) CP can be found as a maximum/minimum along a 1D gradient field line of the L(r) distribution linking two (3,+3) or two (3,−1) CPs.

The saddle (3,−1) and (3,+1) CPs correspond to the local 1D depletion/2D concentration and 2D depletion/1D concentration, respectively, along the three main directions of the L(r) distribution at the CP. Thus, in surfaces including two (3,−3) CPs, a (3,+1) CP appears as a CD site in the valence shell of the atom where the interaction access towards the nucleus is more favourable for a nucleophile, whereas a (3,−1) CP appears in between two (3,−3) CPs along the gradient vector line of the L(r) distribution linking them. Accordingly, (3,−1) CPs generally perform less well than (3,+1) CPs as electrophilic centres, but this is however modulated by their observed L(r) values. The quality of (3,−1) CPs as electrophilic centres also depends on the relative positions of the (3,−3) CPs: the more distant from each other they are, the better is the electrophilic quality of the (3,−1) CP. Therefore, if the (3,−3) CPs are separated far enough, the (3,−1) CP marks the interaction access direction towards the nucleus for a nucleophile at the intersection of the perpendicular plane with the gradient field direction of the L(r) distribution linking the (3,−3) CPs. On the other hand, significant overlap of the electron distributions associated with the two (3,−3) CPs leads us to consider the (3,−1) CP as a nucleophilic indicator of the simultaneous contributions of both (3,−3) CPs, in particular if the former is placed in the intermolecular interaction plane rather than the latter two.

For Ssp3 atoms, it is common to observe the lone pairs overlapping each other in the region where a (3,−1) CP is exhibited, whereas for Sesp3 atoms their open valence-shell structure – attributed to large σ-hole regions – results in a (3,−1) CP simultaneously positioned in between two (3,−3) CPs (lone pairs) and two (3,+1) CPs, lying approximately at the intersection of the lone-pair plane and the plane containing both σ holes in a region of significant depletion of electron density. As shown in this work, the last trend is observed for Sesp3 atoms and the (3,−1) CP is considered to be a CD site, while due to the closer proximity of the (3,−3) CPs (lone pairs) of the Ssp3 atoms, the (3,−1) CP takes the role of a CC site in IDT because it is placed in the intermolecular interaction plane, representing the accommodation of the contribution of each lone pair that is also involved in another more straightforward interaction [see the interlayer interactions in Fig. 2[link](b)]. For Osp3 atoms, the overlap of lone pairs is even more significant because they are closer to each other (Brezgunova et al., 2013[Brezgunova, M., Lieffrig, J., Aubert, E., Dahaoui, S., Fertey, P., Lebègue, S., Ángyán, J. G., Fourmigué, M. & Espinosa, E. (2013). Cryst. Growth Des. 13, 3283-3289.]). In that case, due to their proximity, the two (3,−3) CPs can contribute as two nucleophilic centres involved in the same intermolecular interaction, without consideration of the less performant (3,−1) CP between them.

[Figure 2]
Figure 2
Crystal structure of IDT. (a) View from above of layers showing the four-membered fragment R22(4) involving two ChB interactions (S⋯S and S⋯I in blue) and a V-type supramolecular motif involving HB (grey) and XB (orange) interactions. (b) Side view of layers showing intercolumn contacts in green.

3. Cambridge Structural Database searches

The Cambridge Structural Database (CSD) ConQuest module (version 2021.1.0) was used to perform a search for the intermolecular and non-covalent intramolecular interactions focused on in this study. In addition to several geometrical parameters that have been defined to characterize the interactions, we also applied some constraints for all the searches performed in this study. Hence, only structures whose 3D coordinates are fully determined and error free were involved in the search query. Furthermore, disordered structures, polymers, ionic and powder structures were excluded from the search. Finally, to ensure that only high-quality supramolecular motifs were part of the working set of crystal structures, the agreement R factor of the crystal structures, which represents the goodness of fit between the crystallographic model and the experimental diffraction data, was kept at ≤0.1 for all searches.

4. Results and discussion

4.1. Crystal structure of IDT

IDT crystallizes in the triclinic system (P1 space group) with one molecule in the unit cell (see Table S1 for crystallographic data as well as data for spherical and multipolar refinements). In the crystal structure, IDT molecules form layers where the molecules are connected by two types of bifurcated interactions [Fig. 2[link](a)]. Along the c-axis direction, each molecule is connected to neighbouring ones by relatively short S2⋯S1 and S2⋯I1 ChB interactions [3.8308 (1) and 3.8035 (9) Å, respectively], forming chains. The value of the reduction ratio (RR), which measures the penetration (RR < 1) or separation (RR > 1) of van der Waals (vdW) spheres, reaches 1.06 and 1.01 for the S2⋯S1 and S2⋯I1 ChB interactions, respectively. The molecules are also assembled in the layers by HB (O1⋯H2) and XB (O1⋯I1) interactions between the chains described above, with distances [2.235 Å and 2.9342 (3) Å, respectively] that are significantly shorter than the sum of their corresponding vdW radii (RR = 0.82 and 0.84). The relatively short distance between the layers (3.339 Å) is a result of the intra-column S1⋯I1, S1,2⋯O1 and S2⋯C2 (lp⋯π) contacts [Fig. 2[link](b)] between stacked molecules that are slipped within each column.

4.2. Chalcogen-bonding interactions

4.2.1. Comparison between chalcogen bonding in IDT and SePA

The ChB interactions in the crystal structure of IDT show features very similar to those observed in the crystal structure of SePA (Fig. 3[link]) (Brezgunova et al., 2013[Brezgunova, M., Lieffrig, J., Aubert, E., Dahaoui, S., Fertey, P., Lebègue, S., Ángyán, J. G., Fourmigué, M. & Espinosa, E. (2013). Cryst. Growth Des. 13, 3283-3289.]). Indeed, despite different atoms and relative positions of the interacting molecules, IDT and SePA exhibit similar types of intermolecular contacts involving two Chsp3 atoms and form a geometrically similar synthon-type fragment that is based on a cyclic four-centre motif. Based on the generalized graph-set notation, the four-membered fragments observed in IDT and SePA (Figs. 2[link] and 3[link]) can be classified as R22(4) supramolecular motifs, with two electrophilic regions and two nucleophilic regions in the four-membered rings.

[Figure 3]
Figure 3
Crystal structure of SePA. (a) Side view projected on the ac plane showing the four-membered R22(4) fragment involving two ChB interactions (Se⋯Se and Se⋯O in blue) and intracolumn interactions (green). (b) View from above projected on the bc plane showing two R22(10) fragments, each involving two HB interactions (grey).

Table 1[link] presents the observed Ch⋯Y (Y = Ch, X, where X represents a halogen) distances in IDT and SePA, which are compared with the corresponding sums of the vdW radii by means of the RR value, as well as the contact angles ζ1 and ζ2, indicating the directionality of the interactions (Fig. 4[link]). From the calculated RR parameters, it is observed that Se atoms in the Se⋯Se contact approach closer than the S atoms in the S⋯S contact. This trend has already been noted for halogen atoms, where the heavier bromine atoms approach closer in Br⋯Br contacts than the lighter chlorine atoms in Cl⋯Cl contacts (Brezgunova et al., 2012[Brezgunova, M. E., Aubert, E., Dahaoui, S., Fertey, P., Lebègue, S., Jelsch, C., Ángyán, J. G. & Espinosa, E. (2012). Cryst. Growth Des. 12, 5373-5386.]).

Table 1
Structural parameters (see Fig. 4[link]) characterizing Ch(δ+)⋯(δ)Y (Ch = S, Se; Y = O, S, I) chalcogen-bonding interactions in IDT and SePA

Parameters with units: distance (d) Ch⋯Y (Y = Ch, X) (Å); RR, defined as the Ch⋯Y distance over the sum of vdW radii (dimensionless); contact angles C1A—Ch1Y, ζ1A (°), C1B—Ch1Y, ζ1B (°), Ch1Y—C2A, ζ2A (°) and Ch1Y—C2B, ζ2B (°); molecular angles C1A—Ch1—C1B, α1 (°) and C2A—Ch2—C2B, α2 (°); degree of planarity angle in the σi plane ϕi = ζiA + ζiB + αi, where i = 1, 2 (°); ζ1 = max(ζ1A, ζ1B) and ζ2 = max(ζ2A, ζ2B). vdW radii = 1.52, 1.80, 1.90 and 1.98 Å for O, S, Se and I, respectively.

Compound Ch(δ+)⋯(δ)Y d RR ζ1A/ζ1B α1 ϕ1 ζ2A/ζ2B α2 ϕ2 ζ1ζ2 ϕ1ϕ2
IDT S1i⋯S2iii 3.8308 (1) 1.06 161.4/97.9 96.0 355.3 126.6/135.4 96.6 358.6 26.0 −3.3
  S2iii⋯I1i 3.8035 (3) 1.01 171.0/84.9 96.6 352.5 92.7 78.3
SePA Se1i⋯Se1ii 3.8220 (2) 1.01 166.3/79.1 87.4 332.8 87.4/122.0 87.4 296.8 44.3 36
  Se1i⋯O1ii 3.3552 (1) 0.98 119.0/145.4 87.4 351.8 111.2 34.2
†Symmetry codes. IDT: (i) x, y, z; (iii) x, y, −1 + z. SePA: (i) x, y, z; (ii) 1/2 − x, −1/2 + y, 3/2 − z.
[Figure 4]
Figure 4
(a) Schematic of the expected orientation of electrophilic (δ+) and nucleophilic (δ) regions around the Ch atom within the C—Ch—C σ plane and in the perpendicular π plane. (b) Structural angles used in the characterization of the C—Ch1⋯Ch2—C interactions (Ch1/Ch2 = S/Se/Te).

The structural angles ζ1 and ζ2 (Table 1[link]) allow us to roughly estimate the expected locations of δ+ and δ regions in Ch atoms that are involved in the formation of δ+δ interactions (Fig. 4[link]). The magnitudes of the angles show several differences in the interaction geometries. Thus, for S1⋯S2, Se1⋯Se1 and S2⋯I1 interactions, one of the two ζ1 angles tends to be close to 180° (linearity), predicting the location of the δ+ region close to the direction of the C—Ch bond. In contrast, for Se1⋯O2, the ζ1 values are close to each other, indicating that the location of the δ+ region is expected to be approximately along the direction bisecting the C—Se—C angle in the σ1 plane (Fig. 4[link]), as pointed out in the earlier study of SePA (Brezgunova et al., 2013[Brezgunova, M., Lieffrig, J., Aubert, E., Dahaoui, S., Fertey, P., Lebègue, S., Ángyán, J. G., Fourmigué, M. & Espinosa, E. (2013). Cryst. Growth Des. 13, 3283-3289.]). On the other hand, the inverse situation is observed with the ζ2 parameter. Indeed, for Se1⋯Se1, S2⋯I1 and Se1⋯O2, ζ2 ranges between 90 and 110°, which approximately corresponds to the expected angular location of the lone pairs (δ regions) in I, O and Se atoms. In the case of the S1⋯S2 contact, the ζ2A and ζ2B values are close to each other, indicating that the δ region is roughly located along the direction bis­ecting the C—S2—C angle in the σ2 plane. It is important to highlight that the observation of a region of either δ+ or δ character along the direction bis­ecting the C—Chsp3—C angle is the consequence of the angular opening of the Chsp3 lone pairs. Thus, a δ region is often observed for lighter chalcogens (O, S), whereas a region of δ+ character is more likely to be observed for heavier chalcogens (Se, Te), in which the lone pairs are more separated from each other.

A in-depth structural analysis of chalcogen–chalcogen interactions is needed for the description of geometrical features beyond the standard geometrical contact angles ζ1 and ζ2, also taking into account the relative orientation of molecules participating in the interactions. This is because not all Ch⋯Ch contacts involve electrophilic⋯nucleophilic (δ+δ) ChB interactions, so it is important to differentiate between several cases. For an sp3-hybridized Ch atom, the σ plane (defined as the C—Ch—C plane) represents the region where the chalcogen atom is expected to develop δ+ regions. On the other hand, the π plane, which bis­ects the α(C—Ch—C) angle and is perpendicular to the σ plane, represents the region where the lone pairs of Ch should show participation with δ regions [Fig. 4[link](a)]. Accordingly, the ζi and αi angles, and the relative position of the σi and πi planes of both molecules (i = 1, 2) geometrically characterize the relative orientation of the Ch1⋯Ch2 contact [Fig. 4[link](b)]. In fact, if Ch1 is participating via a δ+ region in the C1A/1B—Ch1⋯Ch2 interaction, Ch2 should be close to the σ1 plane, meaning that the magnitude of ϕ1 (defined as ϕ1 = α1 + ζ1A + ζ1B) should approach 360°. Similarly, if Ch2 participates via a δ+ region in the interaction, then Ch1 should be close to the σ2 plane and ϕ2 ≃ 360°. Other significant geometrical parameters are ζ1 = max(ζ1A, ζ1B) and ζ2 = max(ζ2A, ζ2B), which characterize the directionality of the Ch1⋯Ch2 interaction. Hence, based on the parameters ϕ1, ϕ2, ζ1 and ζ2, chalcogen–chalcogen interactions can be easily differentiated between chalcogen bonding (where regions of different electronic nature δ+δ are facing each other) and chalcogen contacts involving regions of similar electronic nature. The importance of these parameters and their use will be further discussed during the analysis of the CSD search results below.

4.2.2. Topological analysis of ρ(r) in chalcogen-bonding regions of R22(4) motifs in IDT and SePA

Based on the topological analysis of ρ(r) (Bader, 1990[Bader, R. F. W. (1990). Atoms in Molecules: A Quantum Theory. Oxford University Press.]), bonding interactions between atoms are identified by the existence of a bond path between their nuclei and the concomitant BCP at the intersection of the bond path with the interatomic zero-flux surface S(r) [∇ρ(r) · n(r) = 0, ∀rS, with n being the unit vector ⊥ S at r]. Chalcogen-bonding interactions in IDT and SePA were thus identified by the observed bond paths and BCPs in their S1⋯S2, S2⋯I1, Se1⋯Se1 and Se1⋯O2 regions (Fig. 5[link]).

[Figure 5]
Figure 5
Intermolecular Ch⋯Ch and Ch⋯Y contacts (Ch = S, Se; Y = I, O) with their bond paths (dotted lines) and their corresponding BCPs (small green circles) between the interacting atoms in the dimers of (a) IDT and (b) SePA.

In order to quantitatively characterize the intermolecular interactions observed in IDT and SePA, we focus on the topological and energetic properties of ρ(r) at their BCPs. Table 2[link] presents the values of the electron density ρ and the Laplacian 2ρ, as well as those of the local electron kinetic (G) and potential (V) energy densities at the BCPs of the selected interactions. For the sake of comparison between IDT and SePA, as well as for a coherent comparison with data throughout this work, values are reported for interacting dimers calculated at the DFT B3LYP/Def2TZVPP level of theory using experimental geometries (experimental data for SePA, and periodic theoretical VASP data for IDT and SePA are presented in Table S2 of supporting information, showing quite similar values). For the four intermolecular interactions, the magnitudes of ρ and ∇2ρ are low (0.028 < ρ < 0.050 e Å−3, 0.32 < ∇2ρ < 0.62 e Å−5) (Table 2[link]) and typically fall into the range determined (theoretically and experimentally) for other weak intermolecular interactions, such as weak H⋯O (Espinosa et al., 1999[Espinosa, E., Souhassou, M., Lachekar, H. & Lecomte, C. (1999). Acta Cryst. B55, 563-572.]), H⋯N (Mata et al., 2007[Mata, I., Molins, E., Alkorta, I. & Espinosa, E. (2007). J. Phys. Chem. A, 111, 6425-6433.]) and H⋯F (Espinosa et al., 2002[Espinosa, E., Alkorta, I., Elguero, J. & Molins, E. (2002). J. Chem. Phys. 117, 5529-5542.]) hydrogen bonds, and weak XX halogen-bonding interactions (Brezgunova et al., 2012[Brezgunova, M. E., Aubert, E., Dahaoui, S., Fertey, P., Lebègue, S., Jelsch, C., Ángyán, J. G. & Espinosa, E. (2012). Cryst. Growth Des. 12, 5373-5386.]).

Table 2
Structural and topological parameters for Ch(δ+)⋯(δ)Y (Ch = S, Se; Y = O, S, Se, I) chalcogen-bonding interactions in IDT and SePA

α is the angle between the CD⋯CC and internuclear directions. Parameters with units: α (°), topological and local energetic properties ρ(e Å−3), ∇2ρ (e Å−5), G and V (kJ mol−1 bohr−3), and |V|/G (dimensionless), calculated at BCPs using the AIMAll software (Keith, 2019[Keith, T. A. (2019). AIMAll (version 19.10.12). TK Gristmill Software, Overland Park KS, USA. https://aim.tkgristmill.com.]) with interacting dimers calculated at experimental geometries. vdW radii = 1.52, 1.80, 1.90 and 1.98 Å for the O, S, Se and I atom, respectively.

Compound Ch(δ+)⋯(δ)Y ρ 2ρ G V |V|/G Eint α§
IDT S1i⋯S2iii 0.028 0.32 6.6 −4.2 0.64 2.1/– 10.1
S2iii⋯I1i 0.048 0.45 10.0 −7.6 0.76 3.8/– 8.7
Dimer Eint           6.4/5.9/5.9  
SePA Se1i⋯Se1ii 0.041 0.36 8.1 −6.3 0.78 3.2/4.7 17.3/15.0
Se1i⋯O1ii 0.050 0.60 12.9 −9.4 0.73 4.7/5.9 11.7
Dimer Eint           9.4/8.3/7.9/10.6  
†See Table 1[link] for symmetry codes.
‡For individual interactions, −Eint (kJ mol−1) is estimated as (left) −V/2 and (right) −0.375V + 2.366 [for ChB interactions involving Se atoms, see Bauzá & Frontera (2020[Bauzá, A. & Frontera, A. (2020). ChemPhysChem, 21, 26-31.])]. The interaction energy of the dimers (difference between the energy of the dimer and those of the monomers) has been calculated at the B3LYP-D3/Def2TZVPP level of theory: values (in kJ mol−1) correspond to −Eint_dimer/−Eint_dimer_corrected_BSSE/−ΣV/2/−ΣEint_(Bauzá & Frontera, 2020[Bauzá, A. & Frontera, A. (2020). ChemPhysChem, 21, 26-31.]).
§In the case of the Se1i⋯Se1ii interaction, one CC seems to interact simultaneously with two CD sites, leading to two α angles.

The ratio |V|/G < 1 classifies the four contacts as `pure' closed-shell interactions of weak intensity (Espinosa et al., 2002[Espinosa, E., Alkorta, I., Elguero, J. & Molins, E. (2002). J. Chem. Phys. 117, 5529-5542.]). Although the empirical estimation of the interaction energy Eint (kJ mol–1) = V (kJ mol−1 bohr−3)/2 (Espinosa et al., 1998[Espinosa, E., Molins, E. & Lecomte, C. (1998). Chem. Phys. Lett. 285, 170-173.]) was previously derived for weak-to-medium HB interactions, it has also been used to provide a reasonable estimate for other weak XB (C6Br5OH, C6Cl5OH) and ChB (SePA) interactions, which were thus compared in the same framework to HB interactions in the same crystal structures (Brezgunova et al., 2012[Brezgunova, M. E., Aubert, E., Dahaoui, S., Fertey, P., Lebègue, S., Jelsch, C., Ángyán, J. G. & Espinosa, E. (2012). Cryst. Growth Des. 12, 5373-5386.]; Brezgunova et al., 2013[Brezgunova, M., Lieffrig, J., Aubert, E., Dahaoui, S., Fertey, P., Lebègue, S., Ángyán, J. G., Fourmigué, M. & Espinosa, E. (2013). Cryst. Growth Des. 13, 3283-3289.]). In order to gauge the goodness of the estimates for halogen bonds and chalcogen bonds, we have also calculated the Eint values from other relationships found in the literature and derived for XB and ChB interactions (Bauzá & Frontera, 2020[Bauzá, A. & Frontera, A. (2020). ChemPhysChem, 21, 26-31.]). Interaction energies in both approaches are similar to each other (except for the S⋯S and S⋯I interactions, which cannot be estimated from the latter reference) and fall within a narrow range of small magnitudes (−Eint = 2–6 kJ mol−1, Table 2[link]). The addition of individual interaction contributions to estimate the interaction energy of the system in which they are embedded compares very well to the interaction energy of each dimer calculated at the B3LYP-D3/Def2TZVPP level of theory (Table 2[link]), which exhibits low Eint magnitudes (∼6 kJ mol−1 for IDT and between 8 and 10 kJ mol−1 for SePA). Low interaction energies contrast with the formation of similar supramolecular motifs embedded in different crystalline environments, despite the different atoms and functional groups involved.

Comparing the estimated Eint values derived for interactions involving homochalcogens, the Se⋯Se interaction is roughly 30% more energetic than S⋯S (3.8 kJ mol−1 versus 2.7 kJ mol−1, as averaged from experimental and theoretical values). The observed tendency, supported by the calculated |V|/G magnitudes (Table 2[link]), is in accordance with theoretical studies (Bleiholder et al., 2007[Bleiholder, C., Gleiter, R., Werz, D. B. & Köppel, H. (2007). Inorg. Chem. 46, 2249-2260.]), showing that the strength of these contacts decreases in the order Te⋯Te > Se⋯Se > S⋯S > O⋯O and reflects the tendency discussed earlier for XX interactions [−Eint = 4.9 ± 1.2 kJ mol−1 for Br⋯Br versus −Eint = 4.3 ± 1.1 kJ mol−1 for Cl⋯Cl (Brezgunova et al., 2012[Brezgunova, M. E., Aubert, E., Dahaoui, S., Fertey, P., Lebègue, S., Jelsch, C., Ángyán, J. G. & Espinosa, E. (2012). Cryst. Growth Des. 12, 5373-5386.])]. It is notable that, paralleling the increase of polarizability from chalcogen to halogen atoms (along the same periodic table row), the estimated magnitude Eint of individual interactions contributing to the interaction energy of motifs increases from S⋯S to Cl⋯Cl and from Se⋯Se to Br⋯Br.

The deformation of the electron density distribution Δρ(r), which is defined as the difference between the crystalline electron density model and the IAM [Δρ(r) = ρcryst(r) − ρIAM(r)], aims to show the redistribution of electrons from isolated atoms to their configuration in the crystal environment as a consequence of all intra- and intermolecular interactions. In Fig. 6[link], the theoretical 3D static Δρ(r) map is plotted in the chalcogen-bonding regions of IDT and SePA. In both crystal structures, the map clearly shows δ (Δρ > 0, blue) and δ+ (Δρ < 0, red) regions directed towards each other in the intermolecular regions S2⋯I1 and Se1⋯Se2, while a less clear situation is observed for the S1⋯S2 and Se1⋯O2 contacts involving regions located just in front of the Ch atoms. Paralleling the progressive separation of the lone pairs along the series O < S < Se < Te (Brezgunova et al., 2013[Brezgunova, M., Lieffrig, J., Aubert, E., Dahaoui, S., Fertey, P., Lebègue, S., Ángyán, J. G., Fourmigué, M. & Espinosa, E. (2013). Cryst. Growth Des. 13, 3283-3289.]), a δ region is more likely to be observed with a banana shape in the π plane for O and S atoms, while a δ+ region is instead seen with a banana shape in the σ plane from the contributions of the σ holes roughly along the C—Ch directions for Se and Te atoms. The former scenario corresponds to S1 in IDT, whereas the polarization of the electron distribution towards one lone pair in S2 leads to the splitting of the δ banana-shape distribution, permitting the emergence of a δ+ region [see Fig. 1[link](b) and insets]. In the plane of the R22(4) motif [Fig. 6[link](a)], the S2⋯S1 ChB interaction (δδ+) involves a depleted region δ+ of S1 roughly along the C1—S1 bond (close to the S1—C3 BCP) and a small δconcentration region for S2 (resulting after splitting) close to the intersection of the σ and π planes (in the vicinity of the δ banana-shape distribution). In SePA, the Se atom also exhibits a banana-shape distribution in the π plane [Fig. 6[link](b)], but due to the increased separation of the lone pairs with respect to S atoms, a small δ+ region appears in the σ plane, relatively close to the δ distribution in the π plane.

[Figure 6]
Figure 6
Static deformation density Δρ(r) maps in intermolecular regions of (a) IDT and (b) SePA. Isosurfaces are drawn at ±0.05 e Å−3 levels: δ and δ+ isosurfaces are shown in blue and red, respectively. The blue isosurface in front of the S2 atom corresponds to a δ region in the plane bis­ecting the C1—S2—C2 angle perpendicularly at S2. Maps and relevant topological CPs of the function L(r) [L(r) = −∇2ρ(r)] for (c) IDT and (d) SePA: CC sites correspond to (3,−3) CPs (yellow) for Se, I and O atoms, and to the (3,−1) CP (green) for S atoms; CD sites correspond to the (3,−1) CP (green) for the Se atom, and to the (3,+1) CPs (pink) for S and Se atoms. All the represented CPs are close to interaction planes. Bond paths (dotted lines) and BCPs (small green circles) associated with the synthon are shown in (c) and (d) and drawn with the AIMAll software (Keith, 2019[Keith, T. A. (2019). AIMAll (version 19.10.12). TK Gristmill Software, Overland Park KS, USA. https://aim.tkgristmill.com.]).

In order to clarify these observations, the interactions that are focused on have been further characterized using the topological CPs of the function L(r) [Figs. 6[link](c) and 6[link](d)]. In the valence shells of atoms, (3,−3) CPs lying outside covalent bonds indicate 3D CC sites and are associated with lone pairs, whereas (3,+1) CPs are mostly found on (or close to) intermolecular interaction planes (along the C—X bonding directions for halogen atoms, and in the σ plane roughly along the C—Chsp3 bonding directions for chalcogen atoms). The (3,+1) CPs are typically associated with CD sites, because the function L(r) is depleted along two main directions on surfaces where (3,−3) CPs remain. On the other hand, topological CPs of the (3,−1) type indicate that the function L(r) is depleted along only one main direction that links two (3,−3) CPs, also as a consequence of the continuity of the function in space. Accordingly, close to the π plane of Chsp3 atoms, a (3,−1) CP is exhibited between two (3,−3) CPs that correspond to the CC sites of the lone pairs. Note that the separation between the two (3,−3) CPs and the sign of the function L(r) in the region of the (3,−1) CP are irrelevant for the topological emergence of this CP. Therefore, (3,−1) CPs can show as either weak CC or weak CD sites because this feature only depends on the separation of the (3,−3) CPs, which are associated with the relative position of the lone pairs and therefore to the contribution of a small region of either δ or the δ+ character in the vicinity of the intersection of the σ and π planes. Hence, the (3,−1) CP of the S2 atom in Fig. 6[link](c) represents a CC site, whereas in the case of the Se1 atom in Fig. 6[link](d), it represents a CD site. Accordingly, in both crystal structures, the synthon shows facing CD⋯CC interactions, while aligning the corresponding electrophilic⋯nucleophilic interactions with internuclear directions, and are observed close to bond paths. Indeed, the angle between the directions α(CD1⋯CC2; at1⋯at2) is small (10.1, 8.7, 11.7 and 17.3° for S⋯S, S⋯I, Se⋯O and Se⋯Se, respectively), as previously observed in another crystal structure involving ChB interactions (Shukla et al., 2020[Shukla, R., Dhaka, A., Aubert, E., Vijayakumar-Syamala, V., Jeannin, O., Fourmigué, M. & Espinosa, E. (2020). Cryst. Growth Des. 20, 7704-7725.]). Note that, while a high similarity is observed between the positions of CC and CD sites within the same motif found in SePA and IDT, due to the extended δ+ region around the Se atom in the molecular σ plane a further (3,+1) CP is also observed on the other side of the bond path [upper position in Fig. 6[link](d)]. This CP remains roughly along the same C–Se direction and belongs to the same σ-hole region, with α = 15.0°. Accordingly, the CC site seems to interact simultaneously with two CD sites of the same Se atom, with α = 17.3 and 15°. This simultaneous interaction of one CC site with two CD sites observed on either side of the same bond path is perhaps at the origin of slightly higher α values than those typically observed with all the other interactions found in this work (<15°), as a consequence of the accommodation of two CD⋯CC interactions. Altogether, the reported results support the assumption that the relative orientation of atoms involved in intermolecular interactions, and therefore the supramolecular motifs they form, are driven by the local electrophilic and nucleophilic regions found in the valence shells of atoms, which are governed by specific CD and CC sites determined by the topological CPs of the function L(r).

4.2.3. CSD search on R22(4) supramolecular motifs

Fig. 7[link] shows the geometrical parameters and criteria used for the CSD search. Results indicate that the R22(4) supramolecular motif is primarily present with the sulfur atom in the Ch1 and Ch2 positions (Ch1 = S1 and Ch2 = S2) with either a halogen atom (X = F/Cl/Br/I) (219 unique motifs) [Figs. 7[link](a) and 7[link](c)] or a chalcogen atom (Ch3 = O/S/Se) (513 unique motifs) [Figs. 7[link](b) and 7[link](d)]. Focusing on the S1⋯S2 interaction with X in the R22(4) motif [Fig. 8[link](a)], 57 hits present ϕ1 = ϕ2 and ζ1 = ζ2. These motifs are chalcogen⋯chalcogen contacts similar to those previously reported as halogen⋯halogen type-I contacts (Desiraju & Parthasarathy, 1989[Desiraju, G. R. & Parthasarathy, R. (1989). J. Am. Chem. Soc. 111, 8725-8726.]). These motifs (set 1) are all located exactly at the origin of the ϕ1ϕ2 versus ζ1ζ2 plot (represented in black). In addition, 54 motifs (set 2) with −20° < ϕ1ϕ2 < 20° and −10° < ζ1ζ2 < 10° (represented in red) can be also considered as type-I contacts, because they show both similar directionality (ζ1ζ2) and relative situation of the S1 and S2 atoms with respect to the σ1 and σ2 planes (ϕ1ϕ2). A further 31 motifs (set 3) show both ϕ1 and ϕ2 < 320° (represented in light blue). The significant difference of ϕ1 and ϕ2 with respect to 360° implies that neither S2 nor S1 is close to the σ plane of the other S atom, indicating that their δ+ regions are not involved in their contacts and therefore suggesting that a primarily molecular stacking builds the motifs. The remaining 77 motifs show typical behaviour for chalcogen bonds involving electrophilic and nucleophilic regions in the interaction (δ+δ). In 31 of these motifs (set 4), ϕ1 > ϕ2 (ϕ1ϕ2 > 0) and ζ1 > ζ2 (ζ1ζ2 > 0) (represented in green) are observed. A positive magnitude of ϕ1ϕ2 implies that S2 is in the σ1 plane of S1, while a positive magnitude of ζ1ζ2 implies that S1 is participating via a δ+ site, therefore indicating an S1δ+⋯S2δ interaction. For 27 motifs (set 5), both ϕ1ϕ2 and ζ1ζ2 are negative (represented in dark blue) and they correspond to S1δ⋯S2δ+ interactions by invoking similar reasons. In the remaining 19 motifs (set 6), ϕ1ϕ2 < 0 and ζ1ζ2 > 0 (represented in orange). Here, while the positive ζ1ζ2 value means that S1 is participating in the interaction with a δ+ region, a negative ϕ1ϕ2 value implies that S1 is in the σ2 plane of S2. This is the case for IDT (ϕ1ϕ2 < −3.3° and ζ1ζ2 > 26.0°, Table 1[link]), which corresponds to a ChB situation where S2 is participating with a δ region roughly located in the vicinity of its σ2 plane, along the direction bis­ecting the C1A—S2—C1B angle, and influenced by the contribution of one lone pair that is polarized in this region.

[Figure 7]
Figure 7
R22(4) supramolecular motifs used for the CSD search: Ch1/Ch2 = S/Se/Te, X = F/Cl/Br/I, Ch3 = O/S/Se (no fragments were found with Ch3 = Te). Geometrical criteria used in the search for motifs: distances d = Ch1⋯Ch2/Ch2X/Ch2⋯Ch3 ≤ sum (vdW) + 0.4 Å, structural angles ∠C—Ch1⋯Ch2/∠C—Ch2⋯Ch1/∠C—Ch2X/∠C—X⋯Ch2/∠C—Ch2⋯Ch3/∠C=Ch3⋯Ch2 within the range 60–180°. For C1A/1B—Ch1⋯Ch2—C2A/2B interactions: ζ1A/1B = ∠C1A/1B—Ch1⋯Ch2 and ζ2A/2B = ∠C2A/2B—Ch2⋯Ch1. For C2A/2B—Ch2X/Ch3—C1B interactions: ζ1A/1B = ∠C2A/2B—Ch2X and ζ2 = ∠C1BX/Ch3⋯Ch2. Note that, in addition to Ch1⋯Ch2 interactions, the nomenclature ζ1A/1B and ζ2A/2B also requires Ch2X interactions to be consistent with that of Table 1[link]. X-atom vdW radii = 1.47, 1.75, 1.85 and 1.98 Å for F, Cl, Br and I, respectively. Ch-atom vdW radii = 1.52, 1.80 and 1.90 Å for O, S and Se, respectively.
[Figure 8]
Figure 8
(a) Scatter plot of ζ1ζ2 versus ϕ1ϕ2 with Ch1 = Ch2 = S and X = F/Cl/Br/I [Figs. 7[link](a) and 7[link](c)], see text for colour code. Frequency plots of geometrical angles: (b) ζ1A/1B = ∠C2A/2B—S2X defined in Fig. 7[link](c) for the whole set of 219 motifs, (c) ζ1A/1B = ∠C2A/2B—S2X defined in Fig. 7[link](c) for the 77 motifs showing Ch1⋯Ch2 ChB interactions, (d) ζ2A = ∠C1BX⋯S2 (X = F/Cl/Br/I) defined in Fig. 7[link](c) with X = F (yellow), X = Cl (orange), X = Br (red) and X = I (purple).

In order to extend the scope of our analysis, we also performed a detailed CSD search for motifs represented in Figs. 7[link](b) and 7[link](d), with Ch1 = Ch2 = S and Ch3 = O/S/Se [Fig. 9[link](a)]. The search resulted in 513 unique motifs that were analysed similarly to Fig. 8[link](a). In terms of the previously defined subsets of data, the analysis shows the presence of 130 motifs belonging to set 1 (black data, all of them placed at the origin of the graph), 104 to set 2 (red data), 69 to set 3 (light blue data), 91 to set 4 (green data), 87 to set 5 (dark blue data) and 32 to set 6 (orange data), leading to a data distribution which is very similar to that depicted in Fig. 8[link](a). Also, as in Fig. 8[link](a), sets 4–6 in Fig. 9[link](a) represent the motifs formed with Ch1⋯Ch2 ChB interactions, while sets 1–3 concern type-I and stacking contacts. Note that searching with Ch1/Ch2 = S/Se/Te (excluding Ch1 = Ch2 = S) and either X = F/Cl/Br/I or Ch3 = O/S/Se/Te resulted in very few hits (only 43 and 8 unique motifs, respectively) and hence no in-depth analysis was performed for this dataset, as it was too small.

[Figure 9]
Figure 9
(a) Scatter plot of ζ1ζ2 versus ϕ1ϕ2 with Ch1 = Ch2 = S and Ch3 = O/S/Se [Figs. 7[link](b) and 7[link](d)] (no fragments were observed with Ch3 = Te), see text for colour code. Frequency plots of geometrical angles: (b) ζ1A/1B = ∠C2A/2B—S2⋯Ch3 defined in Fig. 7[link](d) for the whole set of 513 motifs, (c) ζ1A/1B = ∠C2A/2B—S2⋯Ch3 defined in Fig. 7[link](d) for the 210 motifs showing Ch1⋯Ch2 ChB interactions, (d) ζ2A = ∠C1B—Ch3⋯S2 (Ch3 = O/S/Se) defined in Fig. 7[link](d) with Ch3 = O (yellow), Ch3 = S (orange) and Ch3 = Se (red) (no fragments were observed with Ch3 = Te).

Overall, the observations made here with respect to S⋯S ChB interactions and contacts are supported by the geometrical parameters ϕ1ϕ2 and ζ1ζ2 (Fig. 8[link]). Hence, while sets 1 to 3 correspond to S⋯S contacts of type I (roughly |ϕ1ϕ2| < 20° and |ζ1ζ2| < 10°, red data) and stacking geometries (mainly with 15° < |ϕ1ϕ2| < 60° and |ζ1ζ2| > 30°, light blue data, even if some data points can spread out of these limits), sets 4 and 5 are associated with ChB δ+δ interactions (either both ϕ1ϕ2 and ζ1ζ2 are positive or negative, green or dark blue data). The particular S⋯S chalcogen contacts in set 6 (ϕ1ϕ2 < 0 and ζ1ζ2 > 0, orange data) correspond to δ2δ1+ ChB interactions. However, they do not have counterpart δ2+δ1 ChB interactions in the region where ϕ1ϕ2 > 0 and ζ1ζ2 < 0 due to the presence of the Ch2X contacts, which involve the corresponding electrophilic region of Ch2 and the nucleophilic region of X in the δ2+δX ChB interaction, as observed in IDT.

The high similarity between Figs. 8[link](a) and 9[link](a) seems to indicate that the structural features involved in the R22(4) motifs that were analysed are equivalent when either halogen or chalcogen atoms are present at the analogous position (X or Ch3) in the R22(4) motif. Despite this significant trend, plotting the frequency of the C2A/C2B—S2X [Fig. 8[link](b)] and C2A/C2B—S2⋯Ch3 [Fig. 9[link](b)] angles for all 219 and 513 motifs, respectively, observed for each dataset does not reveal any particular angular preference. However, selecting the 77 and 210 motifs that we consider to be ChB interactions within each dataset, a clear pattern appears in both frequency plots. Two peaks are found in the 90–100° and 160–170° regions for R22(4) motifs with X [Fig. 8[link](c)] and two others apear in the 80–90° and 140–150° regions for R22(4) motifs with Ch3 [Fig. 9[link](c)], which correspond to the expected angular geometries of C2B—S2X/Ch3 and C2A—S2X/Ch3, respectively, associated with the δ+ region of the S atom along the C2A—S2 direction. These peaks in frequency are similar to those previously reported for a CSD search on S/Se/Te⋯O ChB interactions (Brezgunova et al., 2013[Brezgunova, M., Lieffrig, J., Aubert, E., Dahaoui, S., Fertey, P., Lebègue, S., Ángyán, J. G., Fourmigué, M. & Espinosa, E. (2013). Cryst. Growth Des. 13, 3283-3289.]). Furthermore, a third lower-frequency peak appears in the ranges 120–130°/110–120° [Figs. 8[link](c) and 9[link](c)] for each dataset, which mainly corresponds to a δ+ region (smaller than the δ+ regions that are approximately along C—S bonding directions) placed in the σ plane along the direction bis­ecting the C2A—S2—C2B angle and making a ChB interaction with X/Ch3. In this case, we have not, however, excluded the possibility of finding motifs where a δ region appears to be involved instead of a δ+ region, making a type-I interaction between S and X/Ch3 atoms, even if this situation should be, in principle, less favourable for the stability of the motif. Figs. 8[link](d) and 9[link](d) show the frequency plots for the C1BX/Ch3⋯S2 angles [Figs. 7[link](c) and 7[link](d)], depicted separately for each X or Ch3 atom type. They point the expected angular geometry at which the lone pairs of the X/Ch3 atoms are involved in the interaction. Accordingly, while the peak of the frequency plot is observed at 80–90° for X = Cl, Br and I, and for Ch3 = S and Se, the peak shifts towards 110–120° for X = F and to 100–110° for Ch3 = O.

To conclude, the geometrical analysis above shows that the S⋯S ChB interaction in IDT belongs to set 6 of the CSD search results (ϕ1ϕ2 < 0 and ζ1ζ2 > 0), whereas the Se⋯Se ChB interaction in SePA belongs to set 1 (ϕ1ϕ2 > 0 and ζ1ζ2 > 0) (Table 1[link]). Both occur simultaneously with a further ChB interaction, involving either a halogen or chalcogen atom as the acceptor. Overall, both ChB interactions build a supramolecular R22(4) motif, the geometry of which fits with the expected positions of the electrophilic and nucleophilic regions of the atoms in the interaction. Additionally, the electronic descriptors that characterize the electrophilic (CD) and nucleophilic (CC) sites in the valence shells of the atoms are observed in a CD⋯CC interaction within the supramolecular R22(4) motif, facing each other closely along the internuclear directions. These trends suggest that the formation and the geometry of the synthon in the crystal structures of IDT and SePA are exhibited regardless of the molecule and are highly dependent on the appropriate orientation of CC and CD sites present in the motif. Hence, with the aim of extending this investigation, hereafter we analyse other supramolecular motifs involving different intermolecular and non-covalent intramolecular interactions.

4.3. Other synthons and motifs

4.3.1. Triangular X3 synthons [R33(3) motifs]

Previous studies on the crystal structures of the halogenated molecules hexa­chloro­benzene, C6Cl6 (HCB), hexa­bromo­benzene, C6Br6 (HBB), penta­chloro­phenol, C6Cl5OH (PCP), and penta­bromo­phenol, C6Br5OH (PBP), have shown the formation of triangular X3 synthons (Bui et al., 2009[Bui, T. T. T., Dahaoui, S., Lecomte, C., Desiraju, G. R. & Espinosa, E. (2009). Angew. Chem. Int. Ed. 48, 3838-3841.]; Brezgunova et al., 2012[Brezgunova, M. E., Aubert, E., Dahaoui, S., Fertey, P., Lebègue, S., Jelsch, C., Ángyán, J. G. & Espinosa, E. (2012). Cryst. Growth Des. 12, 5373-5386.]; Aubert et al., 2011[Aubert, E., Lebègue, S., Marsman, M., Bui, T. T. T., Jelsch, C., Dahaoui, S., Espinosa, E. & Ángyán, J. G. (2011). J. Phys. Chem. A, 115, 14484-14494.]), which can be labelled as R33(3) ring motifs following the new graph-set notation, where each halogen atom simultaneously plays an amphoteric role involving one electrophilic and one nucleophilic region in the X3 synthon. The bonding pattern in these motifs was established to be driven by local electrostatic electrophilic⋯nucleophilic (δ+δ) interactions. In terms of geometry, the main difference between a standard single halogen⋯halogen bond and a triangular XB interaction in the X3 synthon is the relative orientation of the electrophilic (δ+) and nucleophilic (δ) regions involved in the formation of the halogen bonds. Thus, in a standard single halogen⋯halogen bond, δ+ and δ sites are roughly orthogonal to each other, resulting in an angular difference |Δζ| = |ζ1ζ2| ≃ 90° [Fig. 10[link](a)]. On the other hand, in the case of X3 synthons, the angular difference |Δζ| is significantly less and ideally close to 60° in order to simultaneously accommodate three electrophilic (δ+) and nucleophilic (δ) regions [Fig. 10[link](b)].

[Figure 10]
Figure 10
Geometrical representation of XX halogen bonds (X = Cl/Br/I) (a) for a single interaction (ζ1 ≃ 180°, ζ2 ≃ 90°) and (b) within a triangular X3 synthon (ζ1 ≃ 180°, ζ2 ≃ 120°). (c) Ch⋯Ch chalcogen bonds (Ch = S/Se/Te) shown within a triangular Ch3 synthon. Δζ is expected at roughly 90 and 60° in (a) and (b), whereas in (c) it is difficult to estimate due to the presence of multiple δ+/δ sites in Chsp3 atoms [see Fig. 4[link](a)]. CSD searches for X3 and Ch3 synthons [ R33(3) motifs] involve the following criteria: distances XX and Ch⋯Ch ≤ sum (vdW) + 0.4 Å, and angles ∠C—XX and ∠C—Ch⋯Ch within the ranges 100–180° and 60–180°, respectively. X-atom vdW radii = 1.75, 1.85 and 1.98 Å for Cl, Br and I, respectively. Ch-atom vdW radii = 1.80, 1.90 and 2.06 Å for S, Se and Te, respectively.

Table 3[link] presents the geometrical parameters of the XX interactions within X3 synthons in C6Cl6, C6Cl5OH, C6Br6 and C6Br5OH. The RR parameter ranges from 0.95 to 1.05, indicating intermolecular distances between the interacting atoms that are close to the sums of their vdW radii. In addition, the magnitude of |Δζ|, which ranges from 47.9 to 59.1° in all but two interactions in C6Cl5OH, further confirms the characteristic XB interactions δ+δ under investigation. The two interactions that lie out of the expected |Δζ| range in C6Cl5OH concern a specific molecular orientation, different to those found in the other compounds, which is due to a particular competition between XB and HB interactions.

Table 3
Halogen-bond geometries of X3 synthons in C6Cls6 (HCB), C6Cl5OH (PCP), C6Br6 (HBB) and C6Br5OH (PBP)

α is the angle between the CD⋯CC and internuclear directions. Parameters with units: d (Å), angles (°), RR (dimensionless). vdW radii = 1.75 and 1.85 Å for Cl and Br atoms, respectively.

Molecule X(δ+)⋯(δ)X d RR ζ1 ζ2 |Δζ| α
C6Cl6 Cl1i⋯Cl2ii 3.4466 (1) 0.98 175.0 116.7 58.3 7.6
Cl2ii⋯Cl3iii 3.4701 (1) 0.99 174.7 124.2 50.5 9.3
Cl3iii⋯Cl1i 3.6662 (1) 1.05 171.2 123.3 47.9 7.3
C6Cl5OH Cl4i⋯Cl2ii 3.4095 (1) 0.97 175.1 78.3 96.8 13.5
Cl2ii⋯Cl3iii 3.6476 (2) 1.04 126.4 110.3 16.1 10.8
Cl3iii⋯Cl4i 3.6197 (2) 1.03 177.5 122.1 55.4 5.5
C6Br6 Br1i⋯Br2ii 3.5412 (2) 0.96 174.2 115.1 59.1 12.2
Br2ii⋯Br3iii 3.5551 (4) 0.96 176.7 123.8 52.9 10.5
Br3iii⋯Br1i 3.7761 (2) 1.02 171.2 122.8 48.4 10.8
C6Br5OH Br2i⋯Br3ii 3.5066 (2) 0.95 173.3 117.3 56.0 10.3
Br3ii⋯Br4iii 3.6576 (2) 0.99 173.0 122.6 50.4 9.0
Br4iii⋯Br2i 3.7127 (3) 1.00 175.7 120.7 55.0 11.7
†Symmetry codes. C6Cl6: (i) x, y, z; (ii) 1/2 − x, 1/2 + y, 1/2 − z; (iii) −1/2 − x, −1/2 + y, 1/2 − z; (iv) −1 + x, 1 + y, z. C6Cl5OH: (i) x, y, z; (ii) x, 3 − y, −1/2 + z; (iii) −x, y, 1/2 − z; (iv) −x, 3 − y, −z. C6Br6: (i) x, y, z; (ii) 1.5 − x, 1/2 + y, 1/2 − z; (iii) −1/2 + x, 1/2 − y, 1/2 + z; (iv) 2 − x, 1 − y, −z. C6Br5OH: (i) x, y, z; (ii) 2 − x, 1/2 + y, 1 − z; (iii) 1 − x, 1/2 + y, 1 − z; (iv) −1 + x, 1 + y, z.

All the reported XX interactions in C6Cl6, C6Cl5OH, C6Br6 and C6Br5OH (see Fig. 11[link]) were found to be of `pure' closed-shell type, since they showed |V|/G < 1 at their corresponding BCPs (Brezgunova et al., 2012[Brezgunova, M. E., Aubert, E., Dahaoui, S., Fertey, P., Lebègue, S., Jelsch, C., Ángyán, J. G. & Espinosa, E. (2012). Cryst. Growth Des. 12, 5373-5386.]). The local properties of ρ(r) at BCPs indicate that the Cl⋯Cl and Br⋯Br interactions are weak, as shown by the small magnitudes of ρ and ∇2ρ (0.028 < ρ < 0.058 e Å−3 and 0.38 < ∇2ρ < 0.62 e Å−5 for Cl⋯Cl contacts, and 0.042 < ρ < 0.066 e Å−3 and 0.41 < ∇2ρ < 0.66 e Å−5 for Br⋯Br contacts, including both experimental and theoretical data). According to the local electronic properties at the BCPs, it was found that Br(δ+)⋯(δ)Br interactions are more intense than Cl(δ+)⋯(δ)Cl, owing to the larger electrophilic δ+ character of the heavier Br atom, even if the nucleophilic δ power of the lone pairs is smaller in the Br atom that for the lighter Cl atom.

[Figure 11]
Figure 11
Maps and topological CPs of the function L(r) [= −∇2ρ(r)] for (a) HCB, (b) PCP, (c) HBB and (d) PBP. L(r) plots are calculated in the planes defined by the positions of the three atoms involved in each synthon. CC and CD sites correspond to (3,−3) CPs (yellow) and (3,+1) CPs (pink) for all Cl and Br atoms. Intermolecular Cl⋯Cl and Br⋯Br contacts are shown with their bond paths (dotted lines) and the corresponding BCPs (small green circles) between the interacting atoms.

The CSD search for homo-X3 synthons (X = Cl/Br/I) resulted in 855 motifs and therefore in a total of 2565 XX interactions (see the caption for Fig. 10[link] for the geometrical criteria). In Fig. 12[link], the peak in frequency observed for |Δζ| in the 0–10° range mainly corresponds to type-I halogen⋯halogen contacts, regardless of the type of halogen atom involved [Figs. 12[link](d)–12[link](f)]. In addition, the frequency plot for |Δζ| shows an extended maximum around 40–50° for X = Cl (1887 Cl⋯Cl interactions) and a clearly defined maximum at 50–60° for X = Br and I (546 Br⋯Br and 132 I⋯I interactions), as expected for XB interactions taking place in X3 synthons (Fig. 10[link]). On the other hand, the frequency distribution of the XX distances shows a peak for each type of halogen atom, appearing at 3.5–3.8, 3.6–3.9 and 3.8–4.0 Å for X = Cl, Br and I, respectively [Figs. 12[link](a)–12[link](c)]. These correspond approximately to the ranges 1 < RRCl < 1.3, 0.9 < RRBr < 1.2 and 0.8 < RRI < 1 (vdW radii = 1.75, 1.85 and 1.98 Å for X = Cl, Br and I, respectively), following the expected increase in interaction strength along the series Cl⋯Cl < Br⋯Br < I⋯I, which parallels the increasing polarizability Cl < Br < I of the halogen atoms.

[Figure 12]
Figure 12
Frequency distributions for XX (X = Cl/Br/I) XB interactions within homo-X3 synthons. For X = Cl, Br and I, the XX distance and the |Δζ| angular orientation are represented in (a) and (d), (b) and (e), and (c) and (f), respectively. The motif shown in Fig. 10[link](b) was used for the CSD search with the geometrical parameters d(XX) ≤ sum(vdW) + 0.4 Å and ζ1 and ζ2 angles within the range 100–180°. vdW radii = 1.75, 1.85 and 1.98 Å for Cl, Br and I atoms, respectively.

The characterization of the topological CPs of the function L(r) shows that the CD⋯CC interactions are almost collinear with internuclear directions. Indeed, for the 12 XX interactions found in the X3 synthons (X = Cl, Br) of the four crystal structures, the ranges for the angle between the two directions α(CD1⋯CC2; at1⋯at2) are 7.3–7.6, 10.5–12.2, 5.5–13.5 and 9.0–11.7° for HCB, HBB, PCP and PBP, respectively. This result further supports the fact that the relative orientation of atoms involved in intermolecular interactions, and therefore the supramolecular motifs they form, are driven by the local electrophilic⋯nucleophilic interactions between CD and CC sites, as previously observed in IDT, SePA and other crystal structures involving ChB interactions (Shukla et al., 2020[Shukla, R., Dhaka, A., Aubert, E., Vijayakumar-Syamala, V., Jeannin, O., Fourmigué, M. & Espinosa, E. (2020). Cryst. Growth Des. 20, 7704-7725.]).

Triangular R33(3) motifs involving chalcogen atoms are also found in the literature. For instance, they appear in some of the crystal structures of the families of 3H-1,2-benzodi­thiole-3-one and 3H-1,2-benzodi­thiole-3-thione heterocycles, and their selenium analogs (Shukla et al., 2020[Shukla, R., Dhaka, A., Aubert, E., Vijayakumar-Syamala, V., Jeannin, O., Fourmigué, M. & Espinosa, E. (2020). Cryst. Growth Des. 20, 7704-7725.]). In order to expand the investigation of X3 synthons (X = Cl, Br, I) [Fig. 10[link](b)] to Ch3 synthons [Fig. 10[link](c)], we also performed a detailed CSD search for R33(3) motifs involving Ch = S, Se and Te atoms in sp3 hybridization (Fig. 13[link]). The search resulted in 1071, 56 and 55 R33(3) motifs with Ch = S, Se and Te atoms, leading to 3213 S⋯S, 168 Se⋯Se and 165 Te⋯Te interactions, respectively. The maxima in the frequencies for S⋯S, Se⋯Se and Te⋯Te distances [Figs. 13[link](a)–13[link](c)] are seen at 3.8–4.0, 3.9–4.0 and 4.1–4.2 Å, corresponding to 1.06 < RRS < 1.11, 1.03 < RRSe < 1.05 and 1.00 < RRTe < 1.02 (vdW radii = 1.80, 1.90 and 2.06 Å for Ch = S, Se and Te, respectively). The peak in frequency observed for |Δζ| in the 0–10° range (for Ch = S/Te) corresponds mainly to type-I Ch⋯Ch contacts [Figs. 13[link](d) and 13[link](f)]. Besides this peak, the |Δζ| frequency plots show an extended maximum at 20–30° for Ch = Se, a small maximum at 30–40° for Ch = S and a very well defined maximum at 60–70° for Ch = Te. The RR results are in line with those of X3 synthons, whereas the frequency plots of |Δζ| do not show peaks at high |Δζ| angles for all Ch atoms that could be systematically associated to electrophilic⋯nucleophilic interactions, as in X3 synthons. However, as discussed in Section 4.2.3[link] and due to the presence of multiple δ+/δ sites in Chsp3 atoms [Fig. 4[link](a)], the structural descriptors associated with Chδ+δCh ChB interactions need to be handled with care. According to our previous analysis of Ch⋯Ch contacts, |Δζ| ≥ 50° mainly encompasses electrophilic⋯nucleophilic (δ+δ) ChB interactions, even if some type-I and stacking interactions can still be found in this particular region [Figs. 8[link](a) and 9[link](a)]. Hence, due to the very well defined frequency maxima occurring at |Δζ| = 60–70°, Te is found to be the most suitable amongst the chalcogens for forming Ch3 synthons.

[Figure 13]
Figure 13
Frequency distributions in Ch⋯Ch (Ch = S/Se/Te) ChB interactions within homo-Ch3 synthons. For Ch = S, Se and Te, the Ch⋯Ch distance and the |Δζ| angular orientation are represented in (a) and (d), (b) and (e), and (c) and (f), respectively. The motif shown in Fig. 10[link](c) was used for the CSD search with the geometrical parameters d (Ch⋯Ch) ≤ sum(vdW) + 0.4 Å and ζ1 and ζ2 angles within the range 60–180°. vdW radii = 1.80, 1.90 and 2.06 Å for S, Se and Te atoms, respectively.

On the other hand, besides X3 synthons and Ch3 synthons, other triangular R33(3) motifs based on pnictogen-bonded cyclic trimers (PH2Y)3 (Y = F, Cl, OH, CN, NC, CH3, H and BH2) with C3h symmetry and bonding energies ranging from −17 to 63 kJ mol−1 have also been described in the literature from theoretical calculations (Alkorta et al., 2013[Alkorta, I., Elguero, J. & Del Bene, J. E. (2013). J. Phys. Chem. A, 117, 4981-4987.]). However, they are beyond the scope of this study, which mainly focuses on ChB, XB and HB interactions, and will not be addressed here.

4.3.2. Bifurcated V-type motifs

The most common V-type atomic motifs are found in HB interactions H⋯Y (Y = O, N), bifurcated at either Y or H. They involve local δ+δ interactions, where δ+ corresponds to a CD region along the extended bonding direction of H and δ corresponds to a lone-pair region of Y. Accordingly, several very common situations can appear: (i) the δ+ region of a single H atom interacts with δ regions of two Y atoms, (ii) a single δ region of one Y atom interacts with each δ+ region of two H atoms, and (iii) two δ regions of a single Y atom interact with δ+ regions of two H atoms. Hereafter we will see that V-type motifs can also be found with other type of atoms, as long as they contribute with appropriate δ+ or δ regions and have a convenient angular disposition in the δ+δ interactions building the motif.

In the crystal structure of IDT, a bifurcated V-type motif involving HB (H⋯O) and XB (I⋯O) interactions has been identified [Figs. 2[link](a) and 14[link](a)]. Table 4[link] presents the structural parameter characteristics of this motif. Here, the O atom forms a bifurcated interaction with H and I atoms, showing almost equal ζ2 angles (∠C—O⋯H/I = 137.8/137.5°) that are larger than those typically observed for the location of the oxygen lone pairs in sp2 hybridization (∼120°). This can be explained by the need to respond to the arrangement of two neighbouring molecules involved in the intermolecular interactions. On the other hand, as expected, the angle ζ1 for C—I⋯O is closer to linear geometry (170.3°) than for C—H⋯O (158.6°). With respect to the RR parameter, the two interactions present very similar values (0.82 and 0.84).

Table 4
Structural parameters, topological and energetic properties at BCPs, and estimated interaction energies for Xδ+δ−Y (X = H, I; Y = O, Cl) interactions in the V-type motifs present in the crystal structures of IDT and PCP

α is the angle between the CD⋯CC and internuclear directions. Parameters with units: d (Å), RR (dimensionless), ζ1 and ζ2 (°), α (°), and topological properties as in Table 2[link]. For individual interactions, −Eint (kJ mol−1) is estimated as (left) −V/2 and (right) −0.556V − 2.697 [for XB interactions involving I atoms, see Bauzá & Frontera (2020[Bauzá, A. & Frontera, A. (2020). ChemPhysChem, 21, 26-31.])]. The interaction energies in IDT and PCP dimers have been calculated at the B3LYP-D3/Def2TZVPP level of theory: values (in kJ mol−1) correspond to −Eint_system/−Eint_system_corrected_BSSE/−ΣV/2/−ΣEint_(Bauzá & Frontera, 2020[Bauzá, A. & Frontera, A. (2020). ChemPhysChem, 21, 26-31.]). vdW radii = 1.20, 1.52, 1.75 and 1.98 Å for H, O, Cl and I atoms, respectively.

Compound X(δ+)⋯(δ)Y d RR ζ1 ζ2 ρ 2ρ G V |V|/G Eint α
IDT C2—H2⋯O1—C1 2.235 0.82 158.7 137.9 0.089 1.25 28.4 −22.6 0.80 11.3/– 8.5
C3—I1⋯O1—C1 2.9342 (3) 0.84 170.3 135.5 0.112 1.49 35.6 −30.6 0.86 15.3/14.3 1.8
Dimer Eint                   33.4/31.6/26.6  
PCP O1—H1⋯O1—C6 2.018 0.74 151.3 125.4 0.136 1.69 43.6 −41.2 0.95 20.6/– 6.5/15.7
O1—H1⋯Cl5—C5 2.943 1.00 114.7 89.3 0.038 0.51 10.8 −7.7 0.71 3.8 7.0
Dimer Eint                   24.8/22.3/24.4  
†For the calculation of the interaction energy of the motif, the IDT trimer is considered to be a dimer formed by the molecule involving an O atom in the V-type motif and the supra-molecular rigid unit formed by the two assembled molecules that involve the I and H atoms in the V-type motif.
[Figure 14]
Figure 14
Maps and topological CPs of the function L(r) [= −∇2ρ(r)] for (a) IDT and (b) PCP. CC sites correspond to (3,−3) CPs (yellow) for O and Cl atoms, CD sites correspond to either (3,+3) (violet) CPs for H atoms, or (3,−1) (dark green) CPs for H and I atoms. Intermolecular HB and XB interactions (Xδ+δO, with X = H, I) are shown with their bond paths (dotted lines) and the corresponding BCPs (small light-green circles) between the interacting atoms. Planes are defined by the interacting atoms in (a) and by the CPs of L(r) focused on in (b).

Fig. 14[link](b) shows another bifurcated V-type motif involving O—H⋯O—C and O—H⋯Cl—C interactions, previously observed in the reported crystal structure of penta­chloro­phenol (C6Cl5OH, PCP) (Brezgunova et al., 2012[Brezgunova, M. E., Aubert, E., Dahaoui, S., Fertey, P., Lebègue, S., Jelsch, C., Ángyán, J. G. & Espinosa, E. (2012). Cryst. Growth Des. 12, 5373-5386.]). The O—H⋯O—C interaction is a short hydrogen bond (RR = 0.74), with structural angles of ζ1 = 151.3° and ζ2 = 125.4° indicating the expected electrophilic and nucleophilic regions close to the O—H bonding direction and around the oxygen lone pair in sp2 hybridization. The O—H⋯Cl—C interaction exhibits a distance close to a vdW contact (RR = 0.99), with structural angles of ζ1 = 114.7° and ζ2 = 89.3°, suggesting that the δ+ region of the H atom is involved in a lateral interaction with the δ region of the Cl atom lone pair that is perpendicular to the C—Cl bonding direction (Table 4[link]).

The bond paths and BCPs of the V-type motifs found in the trimer of IDT and in the dimer of PCP are depicted in Fig. 14[link], along with the relevant CC and CD sites found from the topological analysis of the function L(r). Structural data, electron properties at BCPs, estimated interaction energies and the angular descriptor α between the CC⋯CD and internuclear directions are presented in Table 4[link].

In both bifurcated V-motifs the interactions are of `pure' closed-shell type (|V|/G < 1) (Table 4[link]) (Espinosa et al., 2002[Espinosa, E., Alkorta, I., Elguero, J. & Molins, E. (2002). J. Chem. Phys. 117, 5529-5542.]). In IDT, the topological (ρ, ∇2ρ) and local energetic properties (G, V, |V|/G) at BCPs are larger for O⋯I than for O⋯H, indicating a stronger interaction in the former. In addition, from the estimated energetic contributions to the dimer interaction energy (Eint), the O⋯I bonding is more energetic than O⋯H. On the other hand, in PCP, the magnitudes of ρ, ∇2ρ, G, V and |V|/G are significantly larger for H⋯O than for H⋯Cl (Table 4[link]), in particular when looking at their estimated energy contributions to the interaction energy of the dimer (Eint). As in the previous sections dealing with R22(4) and R33(3) motifs, the characterization of the topological CPs of the function L(r) in the V-type motifs shows local electrophilic⋯nucleophilic interactions (CD⋯CC) that are remarkably well aligned with internuclear directions. Indeed, α < 10° for the four interactions in the V-type motifs. A second value of α is also given for the H⋯O interaction in PCP to emphasize that the π plane containing both lone pairs is in fact directed towards H (the second α value corresponds to the second CC site of O). It is also noticeable that, despite the very weak H⋯Cl interaction (pointed out by all the topological properties at the BCP), the angle α is lower than 10°. This result seems to indicate that directionality is not necessarily linked to a significant interaction energy. A similar finding is observed, for instance, in X3 synthons. In the I⋯O interaction of IDT, a (3,+1) CP of L(r) is also plotted in Fig. 14[link](a), which is found further inside the electron shell than the (3,−1) CP associated with the CD site of I. The two CPs are connected by a gradient line of L(r) that indicates the direction of electrophilicity of the atom, while the angle α is very similar for both of them. Their joint representation aims to show that, in cases of heavier atoms such as I, the CD site is better described by a (3,−1) CP of L(r) in a region of large depletion of electron density rather than the usual (3,+1) CP, which is found in a region of significant electron concentration [∇2ρ(r) < 0].

4.3.3. Intramolecular interactions

The results of CSD searches for intramolecular HBs and ChB interactions are shown in Fig. 15[link]. Differing from the former notation of Etter (1990[Etter, M. C. (1990). Acc. Chem. Res. 23, 120-126.]), where the graph-set notation S(n) denotes an intramolecular HB pattern of n members and subscript d and superscript a do not appear because they equal 1 in all cases, in the new graph-set notation involving CC and CD sites, it is of interest to keep both the subscript e and superscript n to describe the number of electrophilic and nucleophilic sites in the interaction, as shown later.

[Figure 15]
Figure 15
CSD searches performed for non-covalent intramolecular interactions involving (a) H⋯O hydrogen bonds belonging to S11(5) motifs, (b) H⋯O hydrogen bonds belonging to S11(6) motifs, (c) H⋯Br hydrogen bonds belonging to S11(5) motifs, (d) H⋯Br hydrogen bonds belonging to S11(6) motifs, (e) Ch⋯O chalcogen bonds (Ch = S, Se) belonging to S11(5) motifs with X = any atom, and (f) Ch⋯O chalcogen bonds (Ch = S, Se) belonging to S11(6) motifs with X = any atom. Dashed lines crossing the bonds represent bonds of any type (single, double, aromatic etc.). Geometric criteria used for the CSD search were: distances H⋯O/H⋯Br/Ch⋯O ≤ sum(vdW radii), angles ζ1 = ∠O—H⋯O/∠O—H⋯Br, ζ2 = ∠C=O⋯H/∠C=O⋯Br/∠C=O⋯Ch, ζ1A = ∠X—Ch⋯O and ζ1B = ∠C—Ch⋯O ranging from 60 to 180° (vdW radii = 1.20, 1.52, 1.80, 1.90 and 1.85 Å for H, O, S, Se and Br atoms, respectively).

The most common non-covalent intramolecular interactions concern H⋯O hydrogen bonds, forming S11(6) motifs. S11(5) motifs involving H⋯O hydrogen bonds are also observed, but are in less number. Supporting this observation, a CSD search for these motifs gave rise to 3596 S11(5) and 9793 S11(6) hits. On the other hand, a similar search involving H⋯Br hydrogen bonds resulted in 295 S11(5) and 10 S11(6) hits. Based on the total number of hits, these results indicate that O is more commonly observed as a hydrogen-bond acceptor than Br for either the S11(5) or the S11(6) motif. In the case of Br, the larger number of hits involving S11(5) motifs [with respect to S11(6) motifs] may indicate that larger atoms accommodate better with the geometry of five-membered rings than smaller atoms, such as O, for which the intramolecular non-covalent interaction appears to be favoured within a six-membered ring geometry, as shown by the number of hits for S11(6) and S11(5) motifs involving H⋯O contacts. A similar feature is observed in intramolecular chalcogen bonding, where larger donor atoms (such as Ch = S, Se with rvdW = 1.80, 1.90 Å, respectively) favour the obervation of a greater number of five-membered rings with respect to six-membered rings. Indeed, the search for intramolecular Ch⋯O chalcogen bonds resulted in 1501 and 183 hits for S11(5) and S11(6) motifs, respectively, with Ch = S, whereas the search resulted in 152 and 5 hits for S11(5) and S11(6) motifs with Ch = Se. Hence, with S, Se and Br atoms, the number of S11(6) motifs are only 12, 3 and 3% of the number of observed S11(5) motifs with the same type of atoms, respectively, whereas the opposite is observed for O atoms where the number of S11(5) motifs are 37% of the number of S11(6) motifs.

Fig. 16[link] shows the frequency distribution of the ζ1 and ζ2 angles defined in Fig. 15[link] for H⋯O and H⋯Br HB interactions forming S11(5) and S11(6) motifs. For both types of hydrogen bonds, the frequency peaks shift towards larger angles from S11(5) (ζ1 = 100–110° and ζ2 = 80–90° for H⋯O, and 110–120° and ζ2 = 60–70° for H⋯Br) to S11(6) motifs (ζ1 = 140–150° and ζ2 = 100–110° for H⋯O, and ζ1 = 130–140° and ζ2 = 70–80° for H⋯Br). A similar result is observed for the corresponding frequency distribution of the ζ1A/1B and ζ2 angles defined in Fig. 15[link] for S⋯O and Se⋯O ChB interactions (Fig. 17[link]). Indeed, they shift to larger angles from S11(5) (for S⋯O: ζ1A = 160–170°, ζ1B = 70–80° and ζ2 = 90–100°; for Se⋯O: ζ1A = 160–180°, ζ1B = 70–80° and ζ2 = 90–110°) to S11(6) motifs (for S⋯O: ζ1A = 170–180°, ζ1B = 80–90° and ζ2 = 110–120°; for Se⋯O: ζ1A = 170–180°, ζ1B = 80–90° and ζ2 = 90–100°), even if the very few cases of S11(6) motifs with Se⋯O interactions cannot be considered statistically significant. While the ζ1A angle of ∼180° in both S⋯O and Se⋯O interactions mostly seems to indicate that the electrophilic region along the C—Ch bond [in both S11(5) and S11(6) motifs] is responsible for the chalcogen bond, the electrophilic region along the O—H bonding direction is clearly better oriented in S11(6) motifs (ζ1 = 140–150°) than in S11(5) motifs (ζ1 = 100–110°), in line with the previous observation that showed a larger number of hits with the former motif. Additionally, in both S11(5) and S11(6) motifs involving either HB or ChB interactions, the ζ2 angle approximately corresponds to the expected position of a lone pair of the acceptor atom (O or Br) in its particular hybridization.

[Figure 16]
Figure 16
ζ1 frequency distribution for HB interactions: (a) S11(5) H⋯O, (b) S11(6) H⋯O, (c) S11(5) H⋯Br and (d) S11(6) H⋯Br. ζ2 frequency distribution for HB interactions: (e) S11(5) H⋯O, (f) S11(6) H⋯O, (g) S11(5) H⋯Br and (h) S11(6) H⋯Br.
[Figure 17]
Figure 17
ζ1A/1B frequency distribution for ChB interactions: (a) S11(5) S⋯O, (b) S11(6) S⋯O, (c) S11(5) Se⋯O and (d) S11(6) Se⋯O interactions. ζ2 frequency distribution for ChB interactions: (e) S11(5) S⋯O, (f) S11(6) S⋯O, (g) S11(5) Se⋯O and (h) S11(6) Se⋯O.

The crystal structure of PBP (Brezgunova et al., 2012[Brezgunova, M. E., Aubert, E., Dahaoui, S., Fertey, P., Lebègue, S., Jelsch, C., Ángyán, J. G. & Espinosa, E. (2012). Cryst. Growth Des. 12, 5373-5386.]) exhibits an intramolecular H⋯Br HB interaction [Fig. 18[link](a)]. As presented in Table 5[link], the interaction shows a significantly low reduction ratio parameter RR = 0.75, while the associated structural angles ζ1 and ζ2 (146.1 and 62.9°) indicate an HB interaction with an O—Hδ+ electrophilic region well oriented towards the nucleophilic region of the Br lone pair, forming an S11(5) motif [Figs. 16[link](c) and 16[link](g)]. In line with the large ζ1 angle, the short HB distance and the small RR value, the H⋯Br interaction exhibits an incipient degree of covalence |V|/G >1 with a significant electron density magnitude at the BCP (ρ = 0.232 e Å−3).

Table 5
Structural parameters, topological and energetic properties at BCPs, and estimated interaction energies for intramolecular interactions X+)⋯(δ)Y (X = H, C, Se, Br; Y = O, C, Br) in PBP, GOHJAQ and AHEQIN

α is the angle between the CD⋯CC and internuclear directions. Parameters with units: d (Å), ζ1 (°), ζ2 (°), α (°), topological properties as in Table 2[link]. Interaction energies −Eint (kJ mol−1) are estimated as in Tables 2[link] and 4[link].

Compound X(δ+)⋯(δ)Y d RR ζ1 ζ2 ρ 2ρ G V |V|/G Eint α
PBP O—H⋯Br—C 2.156 0.71 146.1 62.9 0.235 1.70 58.2 −70.1 1.20 35.1/– 15.0
AHEQIN C—Se⋯O—C 2.561 0.75 170.3/76.6 103.0 0.212 2.35 64.9 −65.7 1.01 32.9/27.0 3.1
C—H⋯C(π)—N 2.354 0.81 110.7 90.3 0.101 1.22 28.9 −24.6 0.85 12.3/– 13.8
GOHJAQ C—Br⋯O—C 2.909 0.86 174.5 115.7 0.101 1.38 31.9 −26.4 0.83 13.2/12.0 5.1
C—C(π)⋯Br—C 3.516 0.99 98.4 134.1 0.040 0.46 10.1 −7.6 0.76 3.8/- 12.9
[Figure 18]
Figure 18
Maps and topological CPs of the function L(r) [= −∇2ρ(r)] for (a) H⋯Br in PBP, (b) Se⋯O and H⋯C in AHEQIN and (c) Br⋯O and C⋯Br in GOHJAQ. CC sites correspond to (3,−3) CPs (yellow) for O and Br atoms and to (3,−1) CPs (dark green) for the C atom; CD sites correspond to (3,+1) (pink) and (3,−1) CPs (dark green) for H atoms, (3,+1) CPs (pink) for the C atom, (3,−1) CPs (dark green) for the Se atom and (3,+1) CPs (pink) for the Br atom. Intramolecular HB, ChB, XB and TrB interactions [X(δ+)⋯(δ)Y with X = H, Se, Br; Y = O, Br] are shown with their bond paths (dotted lines) and their corresponding BCPs (small light-green circles) between the interacting atoms.

For comparison with other intramolecular HB patterns, we have also performed topological analysis on intramolecular chalcogen and hydrogen bonding [Fig. 18[link](b)], and halogen and tetrel bonding [Fig. 18[link](c)]. Thus, two S11(5) motifs sharing a C—Se moiety (CSD code AHEQIN; Jones et al., 2002[Jones, P. G., Wismach, C., Mugesh, G. & du Mont, W.-W. (2002). Acta Cryst. E58, o1298-o1300.]) and a pair of S11(11) and S11(9) motifs intersecting on a secondary S22(4) motif (GOHJAQ, Widner et al., 2014[Widner, D. L., Knauf, Q. R., Merucci, M. T., Fritz, T. R., Sauer, J. S., Speetzen, E. D., Bosch, E. & Bowling, N. P. (2014). J. Org. Chem. 79, 6269-6278.]) were studied. Note that all of them are found in the molecular planes. The scheme below[link] shows the expected positions of the electrophilic and nucleophilic regions building the two S11(5) motifs in AHEQIN, and the S11(11) and S11(9) motifs [intersecting on a secondary S22(4) motif] in GOHJAQ, as discussed below. From the structural parameters (Table 5[link]), the significantly low RR values demonstrate that these intramolecular interactions are very short (except for the tetrel bond, TrB), while the ζ1 and ζ2 angles further confirm the expected geometry of the electrophilic⋯nucleophilic interaction involved in the intramolecular chalcogen and halogen bonding under investigation. In Fig. 18[link](c), the V-type bifurcated interaction at the amphoteric halogen atom involves its electrophilic σ-hole region along the C—Br bonding direction in a halogen bond (C—Brδ+δO) and its nucleophilic lone-pair region in the tetrel bond (C—Cδ+δBr), leading to an S22(4) motif. The existence of these non-covalent intramolecular interactions is established by the presence of bond paths and the concomitant BCPs (Fig. 18[link]). In AHEQIN and GOHJAQ, the three intramolecular XB, HB and TrB interactions are of `pure' closed-shell type (|V|/G < 1), whereas the ChB interaction shows an incipient degree of covalence (|V|/G > 1). The topological parameters at the BCP of the Se⋯O chalcogen bond in AHEQIN are roughly similar to those calculated for the H⋯Br hydrogen bond in PBP, suggesting similar strength. In comparison, the intramolecular XB (Br⋯O) and TrB (C⋯Br) interactions in GOHJAQ are observed to be significantly weaker (Table 5[link]).[link]

[Scheme 2]

As previously pointed out, the assignment of CC and CD sites to specific CPs of the function L(r) should be taken with care. Hence, in addition to the CPs that are associated with CC and CD sites, surrounding CPs are plotted in Fig. 18[link] for discussion. Nucleophilic sites are clearly identified by the (3,−3) CPs that are associated with the atomic lone pairs of Br and O atoms, whereas in π systems they appear as (3,−1) CPs, such as in the H⋯C interaction of AHEQIN. On the other hand, π systems also develop electrophilic sites that are described by (3,+1) CPs, such as in the C⋯Br interaction of GOHJAQ. Because of the development of (3,−1) and (3,+1) CPs in π systems, the local electrostatic complementarity of electrophilic⋯nucleophilic interactions also takes place in ππ interactions and drives the mutual arrangement of these systems. This trend, being beyond the scope of this work, has not been investigated in depth here and will be addressed in future research. The electrophilic site of an H atom is typically described by a (3,+3) CP that appears along the X—H bond direction. Concomitantly, it is very common to find two (3,+1) CPs, one on each side of this (3,+3) CP, such as in AHEQIN. These (3,+1) CPs are typically involved in lateral HB interactions, as observed in the H⋯C interaction of this compound and in the orientation of the (3,+1)⋯(3,−3) CPs with the internuclear direction H⋯O of GOHJAQ. In the case of PBP, and due to the strong H⋯Br interaction that pulls the O—H group toward the intramolecular region, the electron distribution around the H atom follows the effect of the surrounding atoms, leading to the coalescence of the (3,+3) and (3,+1) CPs, which disappear while the observed (3,−1) CP emerges in the interaction with the (3,−3) CP of the Br atom. The σ-hole interaction of Se with the lone pair of O in AHEQIN involves a (3,−1) CP that behaves as a CD site because the lone pairs of the Se atom are sufficently separated, as explained in previous sections. This feature permits the development of an extended σ plane of electrophilic character in front of the O atom where, in addition to the (3,−1) CP, we also found two (3,+1) CPs, one on each side of the former, as observed in SePA and other chalcogenated molecules (Shukla et al., 2020[Shukla, R., Dhaka, A., Aubert, E., Vijayakumar-Syamala, V., Jeannin, O., Fourmigué, M. & Espinosa, E. (2020). Cryst. Growth Des. 20, 7704-7725.]). In GOHJAQ, the σ hole of Br shows an electrophilic region more directed toward the lone pair of the O atom. The corresponding CD site is assigned to a (3,+1) CP because the (3,−1) CP found for the heavier I atom in IDT does not appear here for Br.

The topological CPs of the function L(r) in the intramolecular motifs analysed again show that CD⋯CC interactions are almost collinear with internuclear directions (α ≤ 15° for all of them, see Table 5[link]). Note that, in addition to the observations made for intermolecular motifs and synthons, they also indicate that local electrophilic⋯nucleophilic interactions between CD and CC sites facing each other drive the atomic orientation in intramolecular motifs.

5. Conclusions

It is quite difficult to predict the geometrical preferences of intermolecular interactions involving chalcogen atoms because the position of their electrophilic (δ+) and nucleophilic (δ) regions can appear enlarged and/or shifted depending on the chalcogen atom type and on its atomic environment. Indeed, in addition to a σ-hole region that can be found enlarged in the σ plane (encompassing a wider range of contact angles than, for instance, in XB interactions, and therefore blurring the expected directionality of the ChB interactions along bonding directions), the relative position of its lone pairs in the π plane can lead to either a δ+ or a δ region close to the intersection of the π and σ planes. The structural angles are very useful for characterizing the geometrical particulars of their intermolecular contacts and therefore for analysing electronic interactions involving chalcogen atoms in sp3 hybridization. To this end, the geometrical descriptors Δζ = ζ1ζ2 and Δϕ = ϕ1ϕ2 defined in this work and based on (i) the intermolecular contact angles ζi = max(ζiA, ζiB), (ii) the intramolecular angles αi = CiA—Chi—CiB and (iii) the planarity degree angle in the σi plane ϕi = ζiA + ζiB + αi (i = 1, 2), have shown their usefulness. With this respect, the Δϕ versus Δζ plot is a very convenient way to differentiate between different Ch⋯Ch contacts, and in particular to assess δ+δ ChB interactions.

The CSD searches and frequency plots of geometrical descriptors of synthons and other supramolecular motifs highlight the particular orientations of intermolecular interactions in all the supramolecular structures investigated in this work. These specific orientations match the expected positions of electrophilic (δ+) and nucleophilic (δ) regions in either ChB, XB or HB interactions.

Molecular δ+ and δ regions are observed in the deformation Δρ(r) maps in the external part of the molecules. The electronic sites responsible for these features are found in the valence shells of the atoms and are assigned by the critical points of the L(r) function, which determine their either CD or CC character along their three main topological directions. The topological analyses of ρ(r) and L(r) functions point out that the observation of particular supramolecular motifs does not depend on the type of atom and functional group involved in donor and acceptor sites, and therefore on the bonding type (such as, for instance, ChB, XB, HB or TrB). Indeed, once relevant electrophilic and nucleophilic regions are placed facing each other along the internuclear directions of the intermolecular interactions that assemble the structural motif, the motif forms in a similar way. As an example, the supramolecular four-membered motif observed in IDT and SePA is electronically equivalent in the two crystal structures, building a synthon that is independent of the atoms involved.

The proposed graph-set assignment Gen(m) (G = C, R, D or S) is a very convenient way to describe the connectivity in motifs based on electrophilic⋯nucleophilic interactions, where n and e denote the number of nucleophilic (CC) and electrophilic (CD) sites, and m is the number of atoms building the motif. In all the R22(4), R33(3), V-type, and S11(6), S11(5) and S22(4) motifs explored, involving either inter- or intramolecular ChB, XB, HB or TrB interactions, the orientation of atoms and functional groups is systematically driven by local electrostatic electrophilic⋯nucleophilic interactions between CD⋯CC sites, as suggested by the small α angle between CD⋯CC and the internuclear directions (in most cases α < 15°). Hence, CC and CD sites drive the relative positions of molecules by orienting their atoms. They are thus at the origin of particular molecular assemblies, governing the geometrical preferences of synthons and other supramolecular motifs despite the fact that they can exhibit modest interaction energies. According to the features described in this work, the local complementary nature of CD⋯CC interactions makes them the building blocks of recurring supramolecular motifs, even if made from different atoms and functional groups and embedded in different molecular environments.

This work will be followed by further investigations dealing with descriptors measuring the nucleophilic/electrophilic power of CC/CD sites and the interaction between them. In addition to synthons and supramolecular motifs, we would also like to extend this analysis to ππ stacking interactions in the same framework. In our future work, a central point of interest will concern the sensitivity of the geometrical position of CC/CD sites, and their nucleophilic/electrophilic power, to molecular environments and conformations.

Supporting information


Computing details top

(I) top
Crystal data top
C3HIOS2Z = 1
Mr = 244.06F(000) = 112
Triclinic, P1Dx = 2.726 Mg m3
a = 4.09358 (10) ÅMo Kα radiation, λ = 0.71073 Å
b = 5.63601 (16) ÅCell parameters from 1403 reflections
c = 6.64675 (17) Åθ = 1.8–17.9°
α = 103.946 (2)°µ = 5.96 mm1
β = 91.017 (2)°T = 100 K
γ = 91.983 (2)°Block, colorless
V = 148.69 (1) Å30.11 × 0.06 × 0.03 mm
Data collection top
Enraf-Nonius
diffractometer
Rint = 0.082
Absorption correction: analytical
[Sheldrick, G.M. (2014). SADABS Bruker AXS Inc., Madison, Wisconsin, USA]
θmax = 65.8°, θmin = 3.2°
h = 1010
66473 measured reflectionsk = 1414
10013 independent reflectionsl = 1617
8709 reflections with I > 2σ(I)
Refinement top
Refinement on F2H-atom parameters constrained
Least-squares matrix: full w = 1/[σ2(Fo2) + 0.137P]
where P = (Fo2 + 2Fc2)/3
R[F2 > 2σ(F2)] = 0.035(Δ/σ)max < 0.001
wR(F2) = 0.079Δρmax = 2.42 e Å3
S = 1.14Δρmin = 3.20 e Å3
10013 reflectionsExtinction correction: SHELXL, Fc*=kFc[1+0.001xFc2λ3/sin(2θ)]-1/4
65 parametersExtinction coefficient: 0.149 (6)
3 restraintsAbsolute structure: Flack x determined using 3669 quotients [(I+)-(I-)]/[(I+)+(I-)] (Parsons, Flack and Wagner, Acta Cryst. B69 (2013) 249-259).
Hydrogen site location: inferred from neighbouring sitesAbsolute structure parameter: 0.034 (11)
Special details top

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 (Å2) top
xyzUiso*/Ueq
I10.14030 (2)0.03280 (2)0.07300 (2)0.01392 (3)
S10.27413 (18)0.52300 (13)0.18084 (10)0.01393 (10)
S20.2680 (2)0.40676 (14)0.58866 (11)0.01562 (10)
O10.5754 (7)0.7961 (5)0.5163 (4)0.0202 (4)
C20.0697 (7)0.1986 (5)0.3844 (4)0.0147 (3)
H20.03700.05350.40360.018*
C30.0737 (6)0.2517 (5)0.1973 (4)0.0125 (3)
C10.4070 (6)0.6117 (5)0.4437 (4)0.0137 (3)
Atomic displacement parameters (Å2) top
U11U22U33U12U13U23
I10.01271 (4)0.01446 (5)0.01236 (4)0.00292 (3)0.00251 (3)0.00045 (3)
S10.0179 (2)0.0131 (2)0.01009 (19)0.00416 (17)0.00256 (17)0.00234 (16)
S20.0208 (3)0.0159 (2)0.00957 (19)0.0047 (2)0.00250 (18)0.00288 (17)
O10.0254 (10)0.0178 (8)0.0151 (7)0.0101 (7)0.0040 (7)0.0014 (6)
C20.0176 (8)0.0139 (8)0.0121 (7)0.0037 (7)0.0005 (6)0.0026 (6)
C30.0129 (7)0.0117 (7)0.0117 (7)0.0024 (5)0.0010 (5)0.0010 (6)
C10.0154 (8)0.0130 (7)0.0115 (7)0.0032 (6)0.0027 (6)0.0013 (6)
Geometric parameters (Å, º) top
I1—C32.078 (2)S2—C11.760 (3)
S1—C31.738 (3)O1—C11.219 (3)
S1—C11.767 (2)C2—C31.348 (4)
S2—C21.734 (3)
C3—S1—C196.01 (13)S1—C3—I1117.61 (15)
C2—S2—C196.55 (13)O1—C1—S2123.8 (2)
C3—C2—S2117.0 (2)O1—C1—S1123.4 (2)
C2—C3—S1117.52 (18)S2—C1—S1112.83 (13)
C2—C3—I1124.87 (19)
 

Acknowledgements

We thank the PMD2X X-ray diffraction facility of the Université de Lorraine for X-ray diffraction measurements. RS also thanks the Gandhi Institute of Technology and Management (GITAM, a Deemed University) for infrastructure and research facilities. We thank Julien Lieffrig (Rennes) for the preparation of IDT.

Funding information

This work has been supported by the French National Agency for Research (grant Nos. ANR-17-CE07-0025 and ANR-21-CE07-0014). RS thanks the ANR for 1 year and 4 month postdoctoral contracts. The EXPLOR mesocenter is thanked for providing access to their computing facility (project No. 2021CPMXX2483).

References

First citationAakeroy, C. B., Bryce, D. L., Desiraju, G. R., Frontera, A., Legon, A. C., Nicotra, F., Rissanen, K., Scheiner, S., Terraneo, G., Metrangolo, P. & Resnati, G. (2019). Pure Appl. Chem. 91, 1889–1892.  Web of Science CrossRef CAS Google Scholar
First citationAlkorta, I., Elguero, J. & Del Bene, J. E. (2013). J. Phys. Chem. A, 117, 4981–4987.  CrossRef CAS PubMed Google Scholar
First citationAlkorta, I., Elguero, J. & Frontera, A. (2020). Crystals, 10, 180.  Web of Science CrossRef Google Scholar
First citationAllen, F. H. (2002). Acta Cryst. B58, 380–388.  Web of Science CrossRef CAS IUCr Journals Google Scholar
First citationAllen, F. H., Motherwell, W. D. S., Raithby, P. R., Shields, G. P. & Taylor, R. (1999). New J. Chem. 23, 25–34.  Web of Science CrossRef CAS Google Scholar
First citationAubert, E., Lebègue, S., Marsman, M., Bui, T. T. T., Jelsch, C., Dahaoui, S., Espinosa, E. & Ángyán, J. G. (2011). J. Phys. Chem. A, 115, 14484–14494.  Web of Science CrossRef CAS PubMed Google Scholar
First citationBader, R. F. W. (1990). Atoms in Molecules: A Quantum Theory. Oxford University Press.  Google Scholar
First citationBamberger, J., Ostler, F. & Mancheño, O. G. (2019). ChemCatChem, 11, 5198–5211.  Web of Science CrossRef CAS PubMed Google Scholar
First citationBauzá, A. & Frontera, A. (2020). ChemPhysChem, 21, 26–31.  PubMed Google Scholar
First citationBeau, M., Jeannin, O., Fourmigué, M., Auban-Senzier, P., Pasquier, C., Alemany, P., Canadell, E. & Jeon, I.-R. (2023). CrystEngComm, 25, 3189–3197.  CSD CrossRef CAS Google Scholar
First citationBlagden, N., Berry, D. J., Parkin, A., Javed, H., Ibrahim, A., Gavan, P. T., De Matos, L. L. & Seaton, C. C. (2008). New J. Chem. 32, 1659–1672.  Web of Science CSD CrossRef CAS Google Scholar
First citationBlagden, N., de Matas, M., Gavan, P. T. & York, P. (2007). Adv. Drug Deliv. Rev. 59, 617–630.  Web of Science CrossRef PubMed CAS Google Scholar
First citationBleiholder, C., Gleiter, R., Werz, D. B. & Köppel, H. (2007). Inorg. Chem. 46, 2249–2260.  Web of Science CrossRef PubMed CAS Google Scholar
First citationBlessing, R. H. (1995). Acta Cryst. A51, 33–38.  CrossRef CAS Web of Science IUCr Journals Google Scholar
First citationBrezgunova, M., Lieffrig, J., Aubert, E., Dahaoui, S., Fertey, P., Lebègue, S., Ángyán, J. G., Fourmigué, M. & Espinosa, E. (2013). Cryst. Growth Des. 13, 3283–3289.  CSD CrossRef CAS Google Scholar
First citationBrezgunova, M. E., Aubert, E., Dahaoui, S., Fertey, P., Lebègue, S., Jelsch, C., Ángyán, J. G. & Espinosa, E. (2012). Cryst. Growth Des. 12, 5373–5386.  Web of Science CSD CrossRef CAS Google Scholar
First citationBui, T. T. T., Dahaoui, S., Lecomte, C., Desiraju, G. R. & Espinosa, E. (2009). Angew. Chem. Int. Ed. 48, 3838–3841.  Web of Science CSD CrossRef CAS Google Scholar
First citationCavallo, G., Metrangolo, P., Milani, R., Pilati, T., Priimagi, A., Resnati, G. & Terraneo, G. (2016). Chem. Rev. 116, 2478–2601.  Web of Science CrossRef CAS PubMed Google Scholar
First citationClark, T., Hennemann, M., Murray, J. S. & Politzer, P. (2007). J. Mol. Model. 13, 291–296.  Web of Science CrossRef PubMed CAS Google Scholar
First citationDesiraju, G. R. (1991). J. Appl. Cryst. 24, 265.  Google Scholar
First citationDesiraju, G. R. (1995). Angew. Chem. Int. Ed. Engl. 34, 2311–2327.  CrossRef CAS Web of Science Google Scholar
First citationDesiraju, G. R., Ho, P. S., Kloo, L., Legon, A. C., Marquardt, R., Metrangolo, P., Politzer, P., Resnati, G. & Rissanen, K. (2013). Pure Appl. Chem. 85, 1711–1713.  Web of Science CrossRef CAS Google Scholar
First citationDesiraju, G. R. & Parthasarathy, R. (1989). J. Am. Chem. Soc. 111, 8725–8726.  CrossRef CAS Web of Science Google Scholar
First citationDhaka, A., Jeannin, O., Jeon, I., Aubert, E., Espinosa, E. & Fourmigué, M. (2020). Angew. Chem. Int. Ed. 59, 23583–23587.  CSD CrossRef CAS Google Scholar
First citationDhaka, A., Jeon, I.-R. & Fourmigué, M. (2024). Acc. Chem. Res. 57, 362–374.  CSD CrossRef CAS PubMed Google Scholar
First citationDolomanov, O. V., Bourhis, L. J., Gildea, R. J., Howard, J. A. K. & Puschmann, H. (2009). J. Appl. Cryst. 42, 339–341.  Web of Science CrossRef CAS IUCr Journals Google Scholar
First citationDomercq, B., Devic, T., Fourmigué, M., Auban-Senzier, P. & Canadell, E. (2001). J. Mater. Chem. 11, 1570–1575.  Web of Science CSD CrossRef CAS Google Scholar
First citationDukhnovsky, E. A., Novikov, A. S., Kubasov, A. S., Borisov, A. V., Sikaona, N. D., Kirichuk, A. A., Khrustalev, V. N., Kritchenkov, A. S. & Tskhovrebov, A. G. (2024). Int. J. Mol. Sci. 25, 3972.  CSD CrossRef PubMed Google Scholar
First citationEspinosa, E., Alkorta, I., Elguero, J. & Molins, E. (2002). J. Chem. Phys. 117, 5529–5542.  Web of Science CrossRef CAS Google Scholar
First citationEspinosa, E., Molins, E. & Lecomte, C. (1998). Chem. Phys. Lett. 285, 170–173.  Web of Science CrossRef CAS Google Scholar
First citationEspinosa, E., Souhassou, M., Lachekar, H. & Lecomte, C. (1999). Acta Cryst. B55, 563–572.  Web of Science CrossRef CAS IUCr Journals Google Scholar
First citationEtter, M. C. (1990). Acc. Chem. Res. 23, 120–126.  CrossRef CAS Web of Science Google Scholar
First citationEtter, M. C. (1991). J. Phys. Chem. 95, 4601–4610.  CrossRef CAS Web of Science Google Scholar
First citationFourmigué, M. (2009). Curr. Opin. Solid State Mater. Sci. 13, 36–45.  Google Scholar
First citationFrisch, M. J., Trucks, G. W., Schlegel, H. B., Scuseria, G. E., Robb, M. A., Cheeseman, J. R., Scalmani, G., Barone, V., Petersson, G. A., Nakatsuji, H., Li, X., Caricato, M., Marenich, A. V., Bloino, J., Janesko, B. G., Gomperts, R., Mennucci, B., Hratchian, H. P., Ortiz, J. V., Izmaylov, A. F., Sonnenberg, J. L., Williams-Young, D., Ding, F., Lipparini, F., Egidi, F., Goings, J., Peng, B., Petrone, A., Henderson, T., Ranasinghe, D., Zakrzewski, V. G., Gao, J., Rega, N., Zheng, G., Liang, W., Hada, M., Ehara, M., Toyota, K., Fukuda, R., Hasegawa, J., Ishida, M., Nakajima, T., Honda, Y., Kitao, O., Nakai, H., Vreven, T., Throssell, K., Montgomery, J. A., Peralta, Jr., J. E., Ogliaro, F., Bearpark, M. J., Heyd, J. J., Brothers, E. N., Kudin, K. N., Staroverov, V. N., Keith, T. A., Kobayashi, R., Normand, J., Raghavachari, K., Rendell, A. P., Burant, J. C., Iyengar, S. S., Tomasi, J., Cossi, M., Millam, J. M., Klene, M., Adamo, C., Cammi, R., Ochterski, J. W., Martin, R. L., Morokuma, K., Farkas, O., Foresman, J. B. & Fox, D. J. (2019). Gaussian 16. Revision C.01. Gaussian Inc., Wallingford, Connecticut, USA.  Google Scholar
First citationGeboes, Y., Nagels, N., Pinter, B., De Proft, F. & Herrebout, W. A. (2015). J. Phys. Chem. A, 119, 2502–2516.  CrossRef CAS PubMed Google Scholar
First citationGleiter, R., Haberhauer, G., Werz, D. B., Rominger, F. & Bleiholder, C. (2018). Chem. Rev. 118, 2010–2041.  Web of Science CrossRef CAS PubMed Google Scholar
First citationGroom, C. R. & Allen, F. H. (2014). Angew. Chem. Int. Ed. 53, 662–671.  Web of Science CrossRef CAS Google Scholar
First citationGuillot, B., Enrique, E., Huder, L. & Jelsch, C. (2014). Acta Cryst. A70, C279–C279.  Web of Science CrossRef IUCr Journals Google Scholar
First citationIssa, N. (2011). PhD thesis. University College London, UK.  Google Scholar
First citationJelsch, C., Guillot, B., Lagoutte, A. & Lecomte, C. (2005). J. Appl. Cryst. 38, 38–54.  Web of Science CrossRef IUCr Journals Google Scholar
First citationJones, P. G., Wismach, C., Mugesh, G. & du Mont, W.-W. (2002). Acta Cryst. E58, o1298–o1300.  CSD CrossRef IUCr Journals Google Scholar
First citationKeith, T. A. (2019). AIMAll (version 19.10.12). TK Gristmill Software, Overland Park KS, USA. https://aim.tkgristmill.comGoogle Scholar
First citationKresse, G. & Furthmüller, J. (1996a). Comput. Mater. Sci. 6, 15–50.  CrossRef CAS Web of Science Google Scholar
First citationKresse, G. & Furthmüller, J. (1996b). Phys. Rev. B, 54, 11169–11186.  CrossRef CAS Web of Science Google Scholar
First citationKresse, G. & Hafner, J. (1993). Phys. Rev. B, 47, 558–561.  CrossRef CAS Web of Science Google Scholar
First citationKresse, G. & Hafner, J. (1994). J. Phys. Condens. Matter, 6, 8245–8257.  CrossRef CAS Web of Science Google Scholar
First citationLe Gal, Y., Lorcy, D., Jeannin, O., Barrière, F., Dorcet, V., Lieffrig, J. & Fourmigué, M. (2016). CrystEngComm, 18, 5474–5481.  Web of Science CSD CrossRef CAS Google Scholar
First citationMacrae, C. F., Sovago, I., Cottrell, S. J., Galek, P. T. A., McCabe, P., Pidcock, E., Platings, M., Shields, G. P., Stevens, J. S., Towler, M. & Wood, P. A. (2020). J. Appl. Cryst. 53, 226–235.  Web of Science CrossRef CAS IUCr Journals Google Scholar
First citationMahmudov, K. T., Kopylovich, M. N., Guedes da Silva, M. F. C. & Pombeiro, A. J. L. (2017). Dalton Trans. 46, 10121–10138.  Web of Science CrossRef CAS PubMed Google Scholar
First citationMata, I., Molins, E., Alkorta, I. & Espinosa, E. (2007). J. Phys. Chem. A, 111, 6425–6433.  Web of Science CrossRef PubMed CAS Google Scholar
First citationNangia, A. (2010). J. Chem. Sci. 122, 295–310.  Web of Science CrossRef CAS Google Scholar
First citationPerdew, J. P., Burke, K. & Ernzerhof, M. (1996). Phys. Rev. Lett. 77, 3865–3868.  CrossRef PubMed CAS Web of Science Google Scholar
First citationPolitzer, P. & Murray, J. S. (2017). Crystals, 7, 212.  CrossRef Google Scholar
First citationScilabra, P., Terraneo, G. & Resnati, G. (2019). Acc. Chem. Res. 52, 1313–1324.  Web of Science CrossRef CAS PubMed Google Scholar
First citationShan, N., Batchelor, E. & Jones, W. (2002b). Tetrahedron Lett. 43, 8721–8725.  CSD CrossRef CAS Google Scholar
First citationShan, N., Bond, A. D. & Jones, W. (2002a). Tetrahedron Lett. 43, 3101–3104.  CSD CrossRef CAS Google Scholar
First citationSheldrick, G. M. (2008). Acta Cryst. A64, 112–122.  Web of Science CrossRef CAS IUCr Journals Google Scholar
First citationSheldrick, G. M. (2015). Acta Cryst. C71, 3–8.  Web of Science CrossRef IUCr Journals Google Scholar
First citationShukla, R., Dhaka, A., Aubert, E., Vijayakumar-Syamala, V., Jeannin, O., Fourmigué, M. & Espinosa, E. (2020). Cryst. Growth Des. 20, 7704–7725.  Web of Science CSD CrossRef CAS Google Scholar
First citationSteiner, T. (2001). Acta Cryst. B57, 103–106.  Web of Science CrossRef CAS IUCr Journals Google Scholar
First citationTiekink, E. & Zukerman-Schpector, J. (2017). Editors. Multi-Com­ponent Crystals. Synthesis, Concepts, Function. Berlin: DeGruyter.  Google Scholar
First citationVishweshwar, P., McMahon, J. A., Bis, J. A. & Zaworotko, M. J. (2006). J. Pharm. Sci. 95, 499–516.  Web of Science CrossRef PubMed CAS Google Scholar
First citationVishweshwar, P., McMahon, J. A., Peterson, M. L., Hickey, M. B., Shattock, T. R. & Zaworotko, M. J. (2005). Chem. Commun. 36, 4601–4603.  CSD CrossRef Google Scholar
First citationVishweshwar, P., Nangia, A. & Lynch, V. M. (2003a). CrystEngComm, 5, 164–168.  Web of Science CSD CrossRef CAS Google Scholar
First citationVishweshwar, P., Nangia, A. & Lynch, V. M. (2003b). Cryst. Growth Des. 3, 783–790.  Web of Science CSD CrossRef CAS Google Scholar
First citationVogel, L., Wonner, P. & Huber, S. M. (2019). Angew. Chem. Int. Ed. 58, 1880–1891.  Web of Science CrossRef CAS Google Scholar
First citationWidner, D. L., Knauf, Q. R., Merucci, M. T., Fritz, T. R., Sauer, J. S., Speetzen, E. D., Bosch, E. & Bowling, N. P. (2014). J. Org. Chem. 79, 6269–6278.  Web of Science CSD CrossRef CAS PubMed Google Scholar

This is an open-access article distributed under the terms of the Creative Commons Attribution (CC-BY) Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original authors and source are cited.

ISSN: 2052-2525