The nature of π-hole interactions between iodide anions and quinoid rings in the crystalline state

The investigated co-crystal of 3-chloro-N-methyl­pyridinium iodide with tetra­bromo­quinone (3-Cl-N-MePy·I·Br₄Q) reveals a π-hole interaction between an iodide anion and a quinoid ring involving an n → π* charge transfer. The quinoid ring has a partial negative charge (estimated to be in the range 0.08–0.11e) and a partial radical character, which is related to the black colour of the crystals (crystals of neutral tetra­bromo­quinone are yellow). A detailed X-ray charge density study revealed two symmetry-independent bond critical points between the iodide anions and carbon atoms of the ring. Their maximum electron density of 0.065 e Å⁻³ was reproduced by quantum chemical modelling. The energy of the interaction is estimated to be −11.16 kcal mol⁻¹, which is comparable to the strength of moderate hydrogen bonding (about −10 kcal mol⁻¹); it is dominantly electrostatic in nature, with a considerable dispersion component.

1. Introduction

π-hole interactions, i.e. interactions involving areas of electron depletion in π-electron systems, have attracted considerable attention in the fields of supramolecular chemistry and crystal engineering (Frontera et al., 2011; Wang & Wang, 2013; Kozuch, 2016; Angarov & Kozuch, 2018; Grounds et al., 2018; Jia et al., 2019; Kumar Seth et al., 2019). An electrostatic interaction occurs between the π-hole and an electron-rich group involving a lone electron pair [lone pair–π or lp⋯π interaction (Mooibroek et al., 2008; Singh & Das, 2015; Newberry & Raines, 2017; Kumar Seth et al., 2018; Angarov & Kozuch, 2018)] or an anion [anion–π interaction (Gamez et al., 2007; Schottel et al., 2008; Frontera et al., 2011; Wang & Wang, 2013; Bauzá et al., 2016; Lucas et al., 2016; Savastano et al., 2017)]. The interaction may also involve charge transfer, typically from a lone pair to an antibonding orbital of the π system [n → π* interaction (Mooibroek et al., 2008; Singh & Das, 2015; Newberry & Raines, 2017; Angarov & Kozuch, 2018)]. They are interesting due to their potential application in molecular recognition (Wang & Wang, 2013; Lucas et al., 2016; Zeng et al., 2019) and drug design (Singh & Das, 2015), and have also been used in crystal engineering (Bauzá et al., 2016; Kumar Seth et al., 2019; Bauza et al., 2019).

Even though the majority of studies have been carried out on electron-poor aromatic systems, quinoid rings, having electron-depleted carbonyl groups, are more promising acceptors of π-hole interactions (Molčanov et al., 2018; Kepler et al., 2019; Wilson et al., 2020; Milašinović & Molčanov, 2021). Quinones with four electronegative substituents have especially prominent π-holes (Molčanov et al., 2019; Vuković et al., 2019) and their interactions with halide anions often involve a charge transfer, indicated by a colour change of the compound (Molčanov et al., 2018; Milašinović & Molčanov, 2021). The common structure motif of the compound under investigation is a sandwich-like group involving two close contacts between a quinone and two halide anions, X⁻⋯Q⋯X⁻ (Fig. 1). The interaction, involving charge transfer is presumably of n → π* type, and is often related to the reduction of quinones to semi­quinone radicals. It may be assumed that the `sandwich' is an intermediate in the reduction of a quinone. Owing to its common formation it may be used in crystal engineering.

Figure 1 Sandwich-like I^−⋯Br4Q⋯I^− unit in 1 with geometric parameters indicated. Symmetry-independent atoms are labelled.

However, the nature and strength of this quinone–iodide interaction remains elusive, and a detailed study is difficult. The dark colour and opacity of the crystals impede studies that use optical spectroscopy, and the presence of heavy atoms (iodine and bromine) makes quantum chemical study challenging. The strong absorption limits the applicability of X-ray diffraction studies. Our initial work (Molčanov et al., 2018) on a series of similar compounds used structural characterization that coupled infrared and solid-state NMR spectroscopies and quantum chemical computation [MP2 and periodic density functional theory (DFT)]. The results highlighted the n → π* charge transfer and estimated the interaction energy to be 6−10 kcal mol⁻¹ (by the MP2 method). However, the degree of charge transfer remained an open issue. Most likely it is quite low (a few percent of an electron), which would be sufficient for a colour change, but not enough to be quantified by spectroscopic methods. A simple analysis of crystal structures using Hirshfeld surfaces (HSs), highest occupied molecular orbitals (HOMOs) and lowest unoccupied molecular orbitals (LUMOs) computed using the CrystalExplorer software (Spackman et al., 2021) and analysis of Voronoi–Dirichlet polyhedra (VDP) (Blatov, 2004) confirmed this model (Milašinović & Molčanov, 2021).

In this work, we opted for a combined experimental and theoretical charge density study on a model system, a co-crystal of 3-chloro-N-methyl­pyridinium iodide and tetra­bromo­quinone [(3-Cl-N-MePy)₂I₂·Br₄Q, named 1, Fig. 2]. To avoid absorption problems, we used short-wavelength (0.6 Å) high-intensity synchrotron radiation for the X-ray diffraction studies. The analysis of the electron density obtained is further supported by extensive theoretical calculations both in the gas phase and as a crystalline structure. This approach allows us to study in detail the behaviour of molecular orbitals in the areas of crucial interactions, interaction energies as well as provide discussion on the topological analysis of charge density.

Figure 2 Tetra­bromo­quinone and 3-Cl-N-MePy cation.

2. Results and discussion

2.1. Crystal packing of 1

The asymmetric unit of 1 comprises a 3-Cl-N-MePy cation, an iodide anion and half of a centrosymmetric Br₄Q molecule; therefore, the molecular formula is (3-Cl-N-MePy)₂I₂·Br₄Q. The easily recognized motif is a sandwich-like I⁻⋯Br₄Q⋯I⁻ unit (Fig. 1) which we have identified in similar compounds (Molčanov et al., 2018; Milašinović & Molčanov, 2021). The distance from the iodide to the centroid of the quinoid ring is 3.947 Å, the distance to the ring mean plane is 3.727 Å and the angle α between the iodide–centroid axis and the ring plane is 72.0°. The iodide is offset by 1.220 Å approximately towards C2; the angle β defining the direction of the offset relative to the carbonyl–carbonyl axis is 76.8°. The only contact shorter than the sum of van der Waals radii for I and C (3.76 Å) is I1⋯C3 [3.7450 (7) Å].

Crystal packing can be described as cations inserted between I⁻⋯Br₄Q⋯I⁻ units (Fig. 3). Aside from the π-hole contact with Br₄Q, the iodide forms three halogen bonds, two with bromine from the Br₄Q molecules and one with a Cl from the cation (Table 1). It also forms two close contacts (3.70 Å) with N1 and C8 atoms from the cation, which are the result of electrostatic attraction between the cation and anion.

Table 1
Short halogen–halogen distances

	d (Å)	C—Br···I (°)	Symmetry operations on Br
C2—Br1⋯I1	3.4960 (1)	173.54 (2)	1 − x, −1/2 + y, 1/2 − z
C3—Br2⋯I2	3.5781 (1)	172.29 (2)	x, −1 + y, z
C5—Cl1⋯I1	3.6225 (2)	168.89 (3)	1 − x, 1 − y, 1 − z
C5—Cl1⋯Cl1†	4.2631 (4)	87.76 (5)	1 − x, 1 − y, 1 − z

†Type II contact.

Figure 3 Crystal packing viewed approximately in the 〈100〉 direction. Iodide anions are shown as spheres of arbitrary radii.

The quinone acts as electron donor of two symmetry-independent halogen bonds (a total of four; Table 1) and its oxygen atom accepts two weak hydrogen bonds from the cation (Table 2). A pair of inversion-related cations form a π-stacked pair with antiparallel C–Cl bonds (Fig. S9). The interplanar separation (the ring planes are parallel, so α = 0°) is 3.3879 (3) Å, but the rings are offset by 3.609 Å, so the centroid distance is rather large, 4.9503 (4) Å. This multitude of intermolecular interactions can be ranked according to their strength and importance as (i) cation–anion electrostatic interactions, (ii) iodide–quinone π-hole interactions, (iii) halogen bonding and (iv) stacking interactions between cations.

Table 2
Geometric parameters of hydrogen bonds

	D—H (Å)	H⋯A (Å)	D⋯A (Å)	D—H⋯A (°)	Symmetry operations on A
C4—H4⋯O1	1.08	2.11	3.1436 (13)	158	1 + x, y, z
C6—H6⋯Cl1†	1.08	2.78	3.8442 (8)	166	−x, 1 − y, 1 − z
C7—H7⋯I1	1.08	3.03	3.8402 (9)	132	−1 + x, y, z
C9A—H9A⋯O1†	1.08 (15)	2.41 (18)	3.4109 (14)	154 (7)	1 + x, y, z
C9—H9C⋯Br2†	1.1 (2)	2.8 (2)	3.8516 (9)	157 (6)	1 − x, −1/2 + y, 1/2 − z
C9—H9A⋯Br1‡	1.08 (15)	3.09 (4)	3.915 (10)	142 (11)	1 + x, y, z
C9—H9B⋯Br1‡	1.08 (11)	3.24 (10)	4.001 (10)	122 (2)	1 − x, −1/2 + y, 1/2 − z
C8—H8⋯Br1‡	1.08	3.21	4.168 (10)	147	1 − x, −1/2 + y, 1/2 − z

†Those found by geometric criteria, but lacking (3, −1) critical points.
‡Those with (3, −1) critical points but unfavourable geometry.

The importance of contacts with the iodide anion is illustrated by the HS of the Br₄Q molecule (Fig. 4): C⋯I contacts comprise 7.4% of the surface, whereas Br⋯I contacts (representing halogen bonding) comprise a further 7.3%. This is slightly higher than in a previously studied series of co-crystals [where the C⋯I contacts comprised 6–7% of the HS (Milašinović & Molčanov, 2021)]. The non-localized nature of the iodide–quinone interaction is also noted when the surface of a Br₄Q molecule is constructed using VDP (Blatov, 2004): 12 individual faces corresponding to 12 individual C⋯I contacts (six for each symmetry-independent C⋯I interaction, Fig. S10) have a total area of 16.64 Ų or 3.5% of the VDP surface. These contact areas are similar to those found in another series of compounds [typically 6–7% of the HS and 3–3.8% of the VDP surface (Milašinović & Molčanov, 2021)].

Figure 4 Hirshfeld surface of a Br4Q molecule with closest-neighbour atoms colour coded: H – grey, C – black, O – red, Cl – green, Br – brown and I – purple.

2.2. Analysis of intermolecular electron density and iodide–quinone contacts

Intermolecular critical points (Fig. 5, Table 3) are mostly in agreement with the geometric analysis of crystal packing. There are two (3, −1) critical points between the iodide and the quinoid ring, with respective electron densities of 0.040 and 0.014 e Å⁻³ (Table 3); the stronger one corresponds to a bond path between I1 and the most electron-depleted atom of the quinone C1 (Fig. 5). The same critical points have also been reproduced by a periodic DFT model, with somewhat higher electron densities of 0.065 and 0.042 e Å⁻³, respectively (Table 3). This is consistent with the electrostatic nature of the iodide–quinone interaction, thus the covalent component is likely negligible (see below).

Table 3
Experimental and calculated (italic) intermolecular bonding critical points in the crystal structure of 1

A⋯B	d (Å)	Electron density (eÅ^−3) ρcp	Laplacian (eÅ^−5)	Type	Symmetry operation on B
π-hole
I1⋯C1	4.1060 (7)	0.0141 0.0424	0.15 0.40	(3, −1)	1 + x, −1 + y, z
I1⋯C3	3.7450 (7)	0.0405 0.0655	0.42 0.67	(3, −1)	1 + x, −1 + y, z
Halogen bonding
I1⋯Cl1	3.6225 (2)	0.0600 0.0594	0.61 0.72	(3, −1)	1 − x, 1 − y, 1 − z
Cl1⋯Cl1	4.2631 (4)	0.0112 NO BCP FOUND	0.15	(3, −1)	1 − x, 1 − y, 1 − z
I1⋯Br1	3.4960 (1)	0.0743 0.0971	0.73 0.90	(3, −1)	1 − x, −1/2 + y, 1/2 − z
I1⋯Br2	3.5781 (1)	0.0637 0.0813	0.62 0.78	(3, −1)	x, −1 + y, z
Hydrogen bonding
H4⋯O1	2.11	0.0693 0.1161	1.79 1.69	(3, −1)	1 + x, y, z
H9A⋯Br1	3.09 (4)	0.0228 0.0378	0.35 0.48	(3, −1)	1 + x, y, z
H8⋯Br1	3.21	0.0261 0.0290	0.32 0.24	(3, −1)	1 − x, −1/2 + y, 1/2 − z
H9B⋯Br1	3.24 (10)	0.0341 0.0321	0.39 0.34	(3, −1)	1 − x, −1/2 + y, 1/2 − z
H9C⋯Br2	2.8 (2)	0.0290 0.0607	0.47 0.72	(3, −1)	1 − x, −1/2 + y, 1/2 − z
H7⋯I1	3.8402 (9)	0.0559 NO BCP FOUND	0.49	(3, −1)	−1 + x, y, z
C—H⋯π
H9B⋯C6	2.88 (5)	0.0457 0.0320	0.46 0.36	(3, −1)	1 + x, y, z
π-stacking
Cl1⋯C4	3.4361 (7)	0.0452	0.54	(3, −1)	1 − x, 1 − y, 1 − z

Figure 5 Symmetry-independent intermolecular critical points in 1. (3, −1) critical points are shown as red spheres and (3, +1) critical points as blue spheres in the interatomic space; bond paths are shown as red lines.

The analysis of theoretical charges (Table 4) indicates a partial charge transfer from the iodide to the quinoid ring of −0.077 to −0.109e, implying a partial negative charge of the Br₄Q. This corresponds with the black colour of the crystals and confirms our previous conclusion (Molčanov et al., 2018). It also provides a more reliable estimate of the degree of charge transfer, as our previous tentative computations were severely overestimated (Molčanov et al., 2018). Therefore, the interaction is of the n → π* type donates electrons into an empty π* (i.e. LUMO) orbital of the quinone, which manifests in an overlap of LUMOs in Fig. 6. Similar behaviour can be observed in the highest binding orbital HOMO-6, where a slight overlap between the iodine orbitals and the quinone ring supports our suggestion that a non-covalent interaction of the π-hole type is present.

Table 4
Mulliken and Bader charges calculated in the gas-phase (g) as well as in a periodic system (p)

	Br4Q (g)		I^−Br4Q (g)		I^−Br4Q (p)
	Mulliken	Bader	Mulliken	Bader	Bader
C1	0.239	1.055	0.257	1.022	1.061
O1	−0.223	−1.057	−0.281	−1.048	−1.129
C2	−0.054	−0.049	−0.007	−0.054	−0.096
C3	−0.006	−0.047	0.045	−0.054	−0.094
Br1	0.048	0.048	−0.042	0.019	0.073
Br2	0.050	0.050	−0.049	0.028	0.076
I1			−0.923	−0.765	−0.684
Total Br4Q	0.000	0.000	−0.077	−0.087	−0.109
Total I^−B4Q			−1.000	−0.852	−0.793

Figure 6 Calculated HOMOs and LUMOs (at an isovalue of 0.001 e au^−3) for the trimer of a tetra­bromo­quinone molecule and two iodine anions, located at a distance of 3.947 Å from the centroid of the quinoid ring (as observed in the crystal structure).

The total binding energy calculated for an I⁻⋯Br₄Q⋯I⁻ unit (Fig. 1) is −95.36 kJ mol⁻¹ and hence for a single I⁻⋯Br₄Q it is −46.68 kJ mol⁻¹ or −11.16 kcal mol⁻¹. This value significantly exceeds our previous estimate of 6–10 kcal mol⁻¹ (Molčanov et al., 2018) and is comparable to intermolecular interactions such as hydrogen bonds (Steiner, 2002) and halogen bonds (Stilinović et al., 2017; Eraković et al., 2019). SAPT energy decomposition (Fig. 7) shows that the dominant component of the total interaction is electrostatic (−86.69 kJ mol⁻¹) followed by dispersion (−57.09 kJ mol⁻¹).

Figure 7 Ternary plot of the SAPT2+3 decomposition of the interaction energy between the Br4Q ring and two iodine anions.

2.3. Other intermolecular contacts

The AIM (Atoms In Molecule) analysis of intermolecular electron density shows that the highest electron density is found in halogen bonds (Table 3); for C—Br⋯I it exceeds 0.06 e Å⁻³ (0.09 e Å⁻³ in the theoretical model), and it is slightly lower in the C—Cl⋯I bond. This is in agreement with previous studies, which showed that the strength of halogen bonding involving Br as a donor is comparable to hydrogen bonding (Stilinović et al., 2017) and that it involves a non-negligible covalent component (Eraković et al., 2019). However, despite higher electron density, the C—Br⋯I halogen bonds are local interactions, whereas the I⋯quinone interaction is non-localized, dispersed between several centres of the Br₄Q molecule. Therefore, it can be concluded that the halogen bonds are of lesser importance in the crystal packing.

Hydrogen bonding (with the exception of C4—H4⋯O1) is weaker, with a maximum electron density below 0.06 e Å⁻³ (0.11 e Å⁻³ in the theoretical model, Table 3). Note that two contacts, which satisfy geometric criteria [Table 2 (see also Steiner, 2002)] do not have a corresponding (3, −1) critical point, therefore they should not be considered as hydrogen bonds. However, three C—H⋯Br bonds with a D⋯A distance exceeding 3.9 Å (Tables 2 and 3) have (3, −1) critical points with electron densities of about 0.03 e Å⁻³. This discrepancy between geometric and AIM criteria for weak hydrogen bonding has been noted previously (Milašinović et al., 2020).

3. Conclusions

This work confirmed the nature of the iodide–quinone interaction as a π-hole interaction involving n → π* charge transfer. The contact is strongly attractive (its strength and importance in crystal packing are second only to cation–anion electrostatic attraction), with an estimated interaction exceeding −11 kcal mol⁻¹, and its dominant component is electrostatic with a significant dispersion contribution. However, a relatively low electron density (not exceeding 0.045 e Å⁻³, Table 3) found between the iodide and the quinone indicates that the interaction is not localized but dispersed between multiple centres (as shown by the HS and the VDP, Figs. 4 and S10). The estimated degree of charge transfer between the iodide and the quinone is −0.077 to −0.109 e, consistent with the black colour of the crystals.

Since π-hole interactions between iodide and quinone occur frequently [so far we described more than 20 analogous compounds (Molčanov et al., 2018; Milašinović & Molčanov, 2021)], we expect that they can be employed in crystal engineering. However, since the sandwich-like moiety I⁻⋯Br₄Q⋯I⁻ is probably formed as a stable intermediate in the reduction of the quinone, it can be expected that more electronegative quinones will be reduced to radicals, while the less electronegative ones will not have sufficiently large π-holes. To test the applicability of this interaction in crystal engineering, a larger number of compounds should be tested, including quinones with different substituents (with different electron-withdrawing capabilities) and different nucleophiles (bromine and other halides as well as similar anions such as cyanate, iso­cyanate, thio­cyanate etc.).

4. Experimental

4.1. Preparation and basic characterization

All reagents and solvents were purchased from commercial sources (Merck, Sigma–Aldrich, Kemika), were of p.a. purity and were used without further purification.

Compound 1 (Fig. 2) was prepared using an analogous procedure similar to previously studied co-crystals (Molčanov et al., 2018; Milašinović & Molčanov, 2021): an excess of solid 3-chloro-N-methyl­pyridinium iodide was added to a cold (5°C) saturated solution of tetra­bromo­quinone in acetone. Diffraction-quality single crystals were grown overnight.

4.2. X-ray diffraction and refinement details

Single-crystal XRD data were collected at the Rossendorf Beamline [ESRF, Grenoble, France (Scheinost et al., 2021)] equipped with an Si(111) monochromator and two Pd-coated mirrors. The single-crystal data were recorded with a Pilatus3 X 2M detector (Dectris) with an excitation energy of 20000 eV per 0.6200926 Å. The monochromator energy was calibrated against the first inflection of the K-absorption edge of an Mo metal foil point, tabulated as 20000 eV. The diffraction measurements were performed in shutterless mode with an angular step size of 0.1° and a counting time of 0.1 s per frame. The detector geometry parameters were calibrated with PyFAI (Kieffer & Wright, 2013) using a powder pattern of the NIST 660b standard LaB₆. Experimental data were collected using the Pylatus software (Dyadkin et al., 2016) and treated using the SNBL ToolBox (Dyadkin et al., 2016) and CrysAlis PRO (Rigaku OD, 2019).

A total of 131 167 reflections were collected, up to a maximum θ of 40.9° (d = 0.475 Å). The multiple integrated reflections were averaged for the space group P2₁/c using SORTAV (Blessing, 1987) adapted to the area detector data.

The structure was solved using SHELXT (Sheldrick, 2015) and a spherical-atom model was refined using SHELXL2017 (Sheldrick, 2015). Multipolar refinement was carried out versus all reflections F² with the program package MoPro (Jelsch et al., 2005). Halogen atoms were modelled as hexadecapoles, O, N and C as octupoles and hydrogens as dipoles; loose restraints were used for multipoles and exponential κ coefficients of chemically equivalent atoms. Vibrations of halogen atoms were refined as anharmonic using fourth-order Gram–Charlier coefficients. Anisotropic parameters for hydrogen atoms were calculated by the SHADE3 server (Madsen, 2006) and kept fixed in the multipolar atom refinement; aromatic C—H bond lengths were restrained to 1.077 (2) Å and methyl C−H bond lengths to 1.083 (2) Å. Geometry and charge-density calculations and analysis of HSs were performed using MoPro (Jelsch et al., 2005); molecular graphics were prepared using MoProViewer (Guillot, 2012) and CCDC-Mercury (Macrae et al., 2020). Crystallographic and refinement data are shown in Table 5.

Table 5
Crystallographic data collection and charge-density refinement details

Compound	1
Empirical formula	C18H14Br4Cl2I2N2O2
Formula weight (g mol^−1)	934.61
Crystal dimensions (mm)	0.12 × 0.09 × 0.09
Space group	P21/c
a (Å)	6.62900 (10)
b (Å)	11.07760 (10)
c (Å)	17.50330 (10)
α (°)	90
β (°)	97.6870 (10)
γ (°)	90
Z	2
V (Å^3)	1273.776 (14)
Dcalc (g cm^−3)	2.438
μ (mm^−1)	6.226
θ range (°)	1.59–40.93
T (K)	100 (2)
Radiation wavelength	0.62009
Detector type	Dectris Pilatus3 X 2M
Range of h, k, l	–13 < h < 13; −23 < k < 23; −36 < l < 36
Reflections collected	131167
Independent reflections	12766
Reflections with I ≥ 2σ	12766
Absorption correction	Analytical
Tmin, Tmax	0.32961, 1.00000
Rint	0.0289
Spherical refinement
Weighting scheme	w = 1/[σ^2(Fo2) + (0.0292P)^2 + 0.2885P], where P = (Fo^2 + 2Fc^2)/3
R (F)	0.0190
Rw (F^2)	0.0547
Goodness of fit	1.101
Hydrogen atom treatment	Constrained isotropic
No. of parameters	137
No. of restraints	0
Δρmax, Δρmin, Δρr.m.s. (eÅ^–3)	1.745; −1.071; 0.115
Multipolar refinement
Weighting scheme	w = 1/[6σ^2(Fo^2)]
R (F)	0.0164
Rw (F^2)	0.0403
Goodness of fit	1.125
H atom treatment	Constrained anisotropic
No. of parameters	709
No. of restraints	477
Δρmax, Δρmin, Δρr.m.s. (e Å^–3)	0.488; −0.992; 0.084

Topological bond orders were calculated using the formula (Zarychta et al., 2011)

Coefficients a, b, c and d were taken from the literature: for C—C bonds, a = −0.522, b = −1.695, c = 0.00, d = 8.473 (Howard & Lamarche, 2003); for C—O bonds, a = −0.427, b = −0.240, c = 0.280, d = 6,464 (Tsirelson et al., 2007); for C—N bonds, a = −0.284, b = 0.331, c = 0.559, d = 6.569; (Howard & Lamarche, 2003); for C—H bonds a = −0.153, b = 0.481, c = 0.983, d = 8.087. (Zhurova et al., 2007).

The analysis of the VDP was achieved using the Topos PRO program package (Blatov, 2004).

4.3. Computational details

Gas-phase calculations were carried out in order to obtain more insight into the nature of the quinone–iodide interaction by means of molecular orbitals, atomic charges and interaction energies. Single-point DFT calculations at the B3LYP/def2-SVPD level of theory (Pritchard et al., 2019) were performed using the GAUSSIAN 16.C.01 program package (Frisch et al., 2016). Grimme D3 dispersion correction (Grimme et al., 2010) was applied in conjunction with the Becke–Johnson damping function and the core electrons for the iodide anions were approximated using pseudopotential functions (Peterson et al., 2003). Bader charges were obtained with the AIMAll software (Keith, 2019). The SAPT2+3 level was performed with the Psi4 software (version 1.3.2; Turney et al., 2012) symmetry-adapted perturbation theory [SAPT (Jeziorski et al., 1994)] using the same basis set as in single-point DFT calculations.

The choice of a smaller def2-SVPD basis set was found to be a good compromise between the efficiency and accuracy of calculations performed. A benchmark study (Parker et al., 2014) found the gold standard of SAPT calculations to be the SAPT2+(3)δMP2 using the aug-cc-pVTZ basis set. Unfortunately, this basis set, and the other basis sets included in the benchmark study, did not achieve SCF convergence in the single-point calculations when we tried to predict the molecular orbitals with the experimental structure of the chosen fragment, i.e. one Br₄Q molecule and two iodine anions. Moreover, the triple zeta basis sets would have been rather computationally expensive for SAPT calculations beyond SAPT0. Choosing a basis set that is similar to those in the above-mentioned benchmark study, we found that the def2-SVPD did achieve convergence, was small enough to complete SAPT2+3 calculations in a timely fashion and is accurate enough to describe non-covalent interactions correctly, as already discussed in the literature (Witte et al., 2016; 2017). Although we could have extended the single-point calculations that produced the molecular orbital diagrams to a larger def2 basis set, we choose instead to keep the same level of theory for both the molecular orbital calculations and the calculation of the interaction energies. It is also important to point out here that SAPT2+3 calculations are designed to compute energies of dimers. In our studies, we chose a slightly different approach and instead of taking only one iodide anion and quinone ring for consideration, we opted for a trimer, where we treated the Br₄Q molecule as a single unit, and two iodide anions as a second unit. Such a choice was motivated by the specific crystal packing where indeed the interaction between the quinone molecule and two iodide anions occurs simultaneously. Although such an approach is not commonly used, one may find examples in the literature when more than two units were considered (Yourdkhani et al., 2016; Steber et al., 2017).

To further support the discussion on the nature of intra- and intermolecular interactions in the crystalline state, periodic DFT was engaged with the use of the CRYSTAL17 software (Dovesi et al., 2018). Calculations were performed on the PBE0/POB-DZVP level of theory (Vilela Oliveira et al., 2019) applying an additional Grimme's D3 correction (Grimme et al., 2010). Atomic coordinates were taken from the X-ray diffraction experiment and were kept frozen during modelling. Periodic wavefunctions obtained in such a way were further used to carry out the topological analysis of periodic electron densities. The QTAIM approach was adopted (Bader, 1990) using the TOPOND14 program (Gatti & Casassa, 2017) integrated with CRYSTAL17.

Supporting information

Computing details

Data collection: CrysAlis PRO 1.171.42.24a (Rigaku OD, 2021); cell refinement: CrysAlis PRO 1.171.42.24a (Rigaku OD, 2021); data reduction: CrysAlis PRO 1.171.42.24a (Rigaku OD, 2021); program(s) used to solve structure: SHELXT 2018/2 (Sheldrick, 2018); program(s) used to refine structure: MoPro (J. Appl. Cryst. 2005, 38, 38-54); molecular graphics: MoPro (J. Appl. Cryst. 2005, 38, 38-54); software used to prepare material for publication: MoPro (J. Appl. Cryst. 2005, 38, 38-54).

(I)

Crystal data

C6Br4O2·2(C6H7ClN)·2(I)	F(000) = 864
Mr = 934.61	Dx = 2.438 Mg m^−3
Monoclinic, P21/c	Synchrotron radiation, λ = 0.62009 Å
a = 6.629 (1) Å	Cell parameters from 27738 reflections
b = 11.0776 (1) Å	θ = 1.6–40.9°
c = 17.5033 (1) Å	µ = 6.23 mm^−1
β = 97.687 (1)°	T = 100 K
V = 1273.78 (1) Å^3	Prism, black
Z = 2	0.12 × 0.09 × 0.09 mm

Data collection

Dectris-CrysAlisPro-abstract goniometer imported dectris images diffractometer	12766 independent reflections
Si(111) monochromator	12193 reflections with > 2.0σ(I)
Detector resolution: 5.8140 pixels mm^-1	Rint = 0.029
ω scans	θmax = 40.9°, θmin = 1.9°
Absorption correction: multi-scan CrysAlisPro 1.171.42.24a (Rigaku Oxford Diffraction, 2021) Empirical absorption correction using spherical harmonics, implemented in SCALE3 ABSPACK scaling algorithm.	h = −13→13
Tmin = 0.330, Tmax = 1.000	k = −23→23
131167 measured reflections	l = −36→36

Refinement

Refinement on F^2	Primary atom site location: structure-invariant direct methods
Least-squares matrix: full	Secondary atom site location: difference Fourier map
R[F^2 > 2σ(F^2)] = 0.016	Hydrogen site location: difference Fourier map
wR(F^2) = 0.041	H atoms treated by a mixture of independent and constrained refinement
S = 1.17	Weighting scheme based on measured s.u.'s
12766 reflections	(Δ/σ)max = 0.009
420 parameters	Δρmax = 0.49 e Å^−3
162 restraints	Δρmin = −0.99 e Å^−3

Special details

Refinement. Refinement of F² against reflections. The threshold expression of F² > 2sigma(F²) is used for calculating R-factors(gt) and is not relevant to the choice of reflections for refinement. R-factors based on F² are statistically about twice as large as those based on F, and R-factors based on ALL data will be even larger.

Fractional atomic coordinates and isotropic or equivalent isotropic displacement parameters (Ų)

	x	y	z	Uiso*/Ueq
I1	0.799833 (6)	0.205080 (4)	0.321463 (2)	0.020372 (5)
BR1	0.135979 (11)	0.855957 (7)	0.349054 (4)	0.020505 (7)
BR2	0.395036 (10)	1.105757 (7)	0.422424 (4)	0.019448 (6)
O1	−0.19990 (15)	0.80593 (10)	0.44221 (6)	0.03310 (9)
C1	−0.10861 (11)	0.89449 (7)	0.46891 (4)	0.01974 (6)
C2	0.07157 (10)	0.94096 (6)	0.43541 (4)	0.01729 (5)
C3	0.17258 (9)	1.03998 (6)	0.46399 (4)	0.01711 (5)
CL1	0.24568 (3)	0.62511 (2)	0.504800 (12)	0.025378 (19)
N1	0.54252 (10)	0.54752 (6)	0.33121 (3)	0.01804 (5)
C4	0.48158 (11)	0.60042 (6)	0.39371 (4)	0.01882 (6)
H4	0.56297	0.67789	0.41999	0.02259*
C5	0.31469 (11)	0.55523 (7)	0.42404 (4)	0.01959 (6)
C6	0.21027 (12)	0.45543 (7)	0.39042 (5)	0.02301 (6)
H6	0.08183	0.41801	0.41463	0.02762*
C7	0.27440 (13)	0.40492 (8)	0.32515 (5)	0.02379 (7)
H7	0.19435	0.32825	0.29719	0.02855*
C8	0.44136 (12)	0.45309 (7)	0.29595 (4)	0.02116 (6)
H8	0.49058	0.41467	0.24463	0.02539*
C9	0.72080 (13)	0.59553 (9)	0.29906 (5)	0.02491 (7)
H9A	0.784 (4)	0.6705 (16)	0.334 (14)	0.03736*
H9B	0.834 (3)	0.526 (14)	0.2994 (16)	0.03736*
H9C	0.675 (4)	0.625 (2)	0.241 (14)	0.03736*

Atomic displacement parameters (Ų)

	U^11	U^22	U^33	U^12	U^13	U^23
I1	0.019179 (17)	0.02187 (2)	0.020688 (17)	−0.002087 (11)	0.004943 (12)	0.001470 (11)
BR1	0.02465 (3)	0.01947 (3)	0.01893 (2)	−0.001924 (18)	0.008568 (18)	−0.001179 (16)
BR2	0.01761 (2)	0.02206 (3)	0.01938 (2)	−0.003742 (17)	0.005076 (17)	0.003582 (17)
O1	0.0320 (3)	0.0339 (4)	0.0370 (3)	−0.0182 (3)	0.0179 (3)	−0.0172 (3)
C1	0.0191 (2)	0.0206 (2)	0.0207 (2)	−0.00589 (18)	0.00682 (17)	−0.00394 (18)
C2	0.01780 (18)	0.0180 (2)	0.01683 (17)	−0.00266 (16)	0.00527 (15)	−0.00068 (15)
C3	0.01647 (17)	0.0184 (2)	0.01718 (17)	−0.00298 (16)	0.00493 (15)	0.00026 (16)
CL1	0.02758 (7)	0.02739 (7)	0.02340 (6)	0.00034 (6)	0.01163 (5)	0.00072 (6)
N1	0.01798 (18)	0.0186 (2)	0.01808 (17)	−0.00115 (16)	0.00440 (15)	−0.00093 (16)
C4	0.0193 (2)	0.0190 (2)	0.01897 (19)	−0.00259 (17)	0.00566 (17)	−0.00145 (17)
C5	0.0193 (2)	0.0197 (2)	0.0208 (2)	−0.00103 (18)	0.00636 (17)	0.00044 (18)
C6	0.0204 (2)	0.0215 (3)	0.0279 (3)	−0.0041 (2)	0.0064 (2)	−0.0001 (2)
C7	0.0231 (2)	0.0202 (3)	0.0283 (3)	−0.0043 (2)	0.0042 (2)	−0.0031 (2)
C8	0.0220 (2)	0.0197 (2)	0.0221 (2)	−0.0010 (2)	0.00403 (19)	−0.00318 (19)
C9	0.0226 (2)	0.0310 (3)	0.0227 (2)	−0.0049 (2)	0.0087 (2)	−0.0031 (2)

Geometric parameters (Å, º)

BR1—C2	1.8779 (6)	C4—H4	1.0830
BR2—C3	1.8767 (6)	C5—C6	1.3924 (10)
O1—C1	1.2131 (9)	C6—C7	1.3888 (12)
C1—C2	1.4909 (8)	C6—H6	1.0830
C1—C3	2.4743 (8)	C7—C8	1.3865 (11)
C2—C3	1.3462 (9)	C7—H7	1.0830
CL1—C5	1.7261 (7)	C8—H8	1.0830
N1—C8	1.3469 (9)	C9—H9A	1.0770 (8)
N1—C4	1.3496 (8)	C9—H9B	1.0770 (9)
N1—C9	1.4745 (9)	C9—H9C	1.0770 (8)
C4—C5	1.3834 (9)
O1—C1—C2	120.82 (5)	C5—C6—C7	118.55 (6)
O1—C1—C3	148.50 (5)	C5—C6—H6	120.72
BR1—C2—C3	123.53 (4)	C7—C6—H6	120.72
BR1—C2—C1	115.08 (4)	C8—C7—C6	119.70 (6)
C3—C2—C1	121.33 (5)	C8—C7—H7	120.15
BR2—C3—C2	123.36 (4)	C6—C7—H7	120.15
BR2—C3—C1	154.30 (2)	N1—C8—C7	120.20 (6)
C8—N1—C4	121.60 (5)	N1—C8—H8	119.90
C8—N1—C9	118.72 (5)	C7—C8—H8	119.90
C4—N1—C9	119.66 (5)	N1—C9—H9A	109.47
N1—C4—C5	119.70 (5)	N1—C9—H9B	109.47
N1—C4—H4	120.15	N1—C9—H9C	109.47
C5—C4—H4	120.15	H9A—C9—H9B	109.47
CL1—C5—C4	117.72 (4)	H9A—C9—H9C	109.47
CL1—C5—C6	122.07 (5)	H9B—C9—H9C	109.47
C4—C5—C6	120.20 (5)
BR1—C2—C3—BR2	−1.00 (4)	C4—N1—C9—H9C	−118.61 (7)
BR1—C2—C3—C1	176.98 (7)	C4—C5—C6—C7	−1.58 (7)
BR1—C2—C1—O1	1.98 (7)	C4—C5—C6—H6	178.42
BR1—C2—C1—C3	−177.22 (9)	H4—C4—N1—C8	−178.19
BR2—C3—C2—C1	−177.98 (5)	H4—C4—N1—C9	0.19
BR2—C3—C1—O1	2.6 (2)	H4—C4—C5—C6	−179.83
BR2—C3—C1—C2	3.89 (18)	C5—C4—N1—C8	1.81 (7)
O1—C1—C2—C3	179.20 (10)	C5—C4—N1—C9	−179.81 (7)
O1—C1—C3—C2	−1.3 (3)	C5—C6—C7—C8	1.10 (8)
CL1—C5—C4—N1	179.12 (6)	C5—C6—C7—H7	−178.90
CL1—C5—C4—H4	−0.88	C6—C7—C8—H8	−179.20
CL1—C5—C6—C7	179.52 (6)	H6—C6—C7—C8	−178.90
CL1—C5—C6—H6	−0.48	H6—C6—C7—H7	1.10
N1—C8—C7—C6	0.80 (7)	C7—C8—N1—C9	179.31 (8)
N1—C8—C7—H7	−179.20	H7—C7—C8—H8	0.80
N1—C4—C5—C6	0.17 (7)	C8—N1—C9—H9A	179.82 (7)
C4—N1—C8—C7	−2.30 (7)	C8—N1—C9—H9B	−60.18 (7)
C4—N1—C8—H8	177.70	C8—N1—C9—H9C	59.82 (7)
C4—N1—C9—H9A	1.39 (7)	H8—C8—N1—C9	−0.69
C4—N1—C9—H9B	121.39 (7)

Hydrogen-bond geometry (Å, º)

D—H···A	D—H	H···A	D···A	D—H···A
C4—H4···O1^i	1.08	2.11	3.1436 (9)	158 (1)
C9—H9A···O1^i	1.08 (1)	2.41 (1)	3.4109 (10)	154 (1)

Symmetry code: (i) x+1, y, z.