Chlorinated polyhedral selenaboranes revisited by joint experimental/computational efforts: the formation of closo-1-SeB₉Cl₉ and the crystal structure of closo-SeB₁₁Cl₁₁

Abstract

The recent success in the formation of chlorinated telluraboranes and the reactivities of pnictogenaboranes prompted us to re-examine the vacuum co-pyrolysis of B₂Cl₄ with Se₂Cl₂ at various molar ratios and temperatures in order to search for the generation of other polyhedral selenaboranes than closo-SeB₅Cl₅ (1a) and closo-SeB₁₁Cl₁₁ (1b), the latter being observed earlier. Interestingly, a new compound with the elemental composition SeB₉Cl₉ (2) was detected, this time by high- and low-resolution mass spectrometry. Further characterization by 1- and 2-D ¹¹B-NMR spectroscopy suggests that 2 should adopt a closed bicapped square-antiprismatic geometry with selenium at the apical position. Moreover, vacuum sublimation gave suitable crystals of 1b, which were subjected to single-crystal X-ray structure determination. Crystallographic data analysis confirmed that 1b, consistent with its 26 skeletal electron count, adopts a distorted icosahedral structure close to the symmetry of C_5v. Computations at the DFT-D3 level have revealed that 33% of the total computed binding motifs in the grown 1b crystals are due to the very strong chalcogen bonding. Moreover, SAPT decomposition has shown that the bonding motifs in the crystals are stabilized mainly by dispersion and electrostatic terms. Homodecoupling and high resolution ¹¹B NMR and ⁷⁷Se NMR experiments have resolved both coupling constants ¹J(¹¹B¹¹B) and ¹J(⁷⁷Se¹¹B) as well as the ⁷⁷Se chemical shift of 1a and 1b, which are in reasonable agreement with the corresponding computed values. The computed ¹¹B chemical shifts of 2 were determined by the well-established DFT/GIAO/NMR structural tool based on its B3LYP/6-311+G** internal coordinates. They agree well with the experimental values and provide a good representation of the molecular structure of 2 in solution. The extraordinary downfield ¹¹B NMR chemical shift of B(10) in 2 has been ascribed to the intensive paramagnetic contribution to the shielding tensor in this bicapped square-antiprismatic motif. Calculations of the synproportionation free energies of smaller (n – 1) closo-selenaboranes with larger-sized (n + 1) ones support the extraordinary stability of octahedral, bicapped square-antiprismatic and icosahedral closo motifs in the SeB_nCl_n family (n = 4–12).

---

Introduction

Polyhedral borane and heteroborane clusters are known for the presence of delocalized electron-deficient bonding¹ and characterized by forming three-center, two-electron (3c–2e) bonds. This bonding is quite different from the organic chemistry that is dominated by classical two-center, two-electron (2c–2e) bonds. The trigonal faces of boranes and the so-called heteroboranes are assembled to create various three-dimensional shapes. There exist closo-, nido-, arachno- and hypho-type cages according to the so-called electron count.² The closo-systems belong to the most researched clusters predominantly represented, e.g. by octahedral closo-B₆H₆²⁻, bicapped square-antiprismatic closo-B₁₀H₁₀²⁻ and icosahedral closo-B₁₂H₁₂²⁻, recognized as well-designed building blocks of boron cluster chemistry.³ The systematic replacement of formally neutral BX vertices in the closo-B_nX_n²⁻ (X = H and/or halogens, n = 6, 10, 12) can lead to a variety of n-vertex closo-heteroboranes, either the parent or with terminal halogens. For example, isoelectrolobal vertices such as E (E = S, Se, Te) with a formal charge of +2 give rise to closo-1-EB_nX_n (n = 5, 9, 11), known as chalcogenaboranes.³ Conceivably, there is a sort of similarity of such 3D chalcogenaboranes with 2D heterocyclic compounds exemplified by thiophene, selenophene, etc. The presence of chalcogens in suitably designed structural moieties results in the detection of the so-called chalcogen bonds responsible for crystal packing not only for chalcogenaboranes.^4–6 However, the area of chalcogenaboranes turned out to be under-researched and just a few polyhedral selenaboranes Se_xB_yH_z are known.⁷ Recently, we have reported on the synthesis and DFT/GIAO/NMR study of the perhalogenated selenaboranes closo-SeB₅Cl₅ (1a) and closo-SeB₁₁Cl₁₁ (1b) by the co-pyrolysis reaction of a 9 : 1 mixture of B₂Cl₄ with Se₂Cl₂in vacuo at 330 °C.⁸ It is well known from the thermal disproportionation reactions of B₂Cl₄ with n-vertex boron subchlorides B_nCl_n (n = 8–12)⁹ that the size n and the distribution of products strongly depend on the reaction temperature and time under thermolytic conditions: higher pyrolysis temperatures augment the amount of active BCl species (see eqn (1)) and thus increase the probability of the formation of larger cages. In the co-pyrolysis of B₂X₄ (X = Cl, Br) with phosphorus trihalide PX₃ (X = Cl, Br), it has been shown that the formation of the main product, closo-1,2-P₂B₄X₄ (X = Cl,^10a X = Br ^10b,c), is accompanied by the larger homologs closo-1,10-P₂B₈Cl₈ ^10d and closo-1,7-P₂B₁₀X₁₀ (X = Cl,^10d X = Br ^10e) as well as conjuncto-3,3′-(1,2-P₂B₄X₃)₂ (X = Cl,^10d X = Br ^10e) when the temperature rises from 330 °C to 400 °C and when the molar B₂Cl₄ content is augmented. Furthermore, extensions of the geometric motifs formed by co-pyrolysis reactions have been found with exoskeletal dihaloboryl groups in perhalogenated carboranes.¹¹ In order to create new homologs of polyhedral selenaboranes, we have now reexamined the co-pyrolysis reaction of B₂Cl₄ with Se₂Cl₂ both by increasing the temperature and by altering the molar ratio of the reactants with respect to the conditions of our previous report. Under such conditions, a novel chlorinated selenaborane, i.e. closo-1-SeB₉Cl₉ (2) (Fig. 1), has now been detected. In addition, it was our intention to extend the still quite limited spectroscopic data of closo-SeB₅Cl₅ and closo-SeB₁₁Cl₁₁ available so far to allow comparisons with the data of the very recently reported related octahedral and icosahedral closo-telluraboranes.¹²

Fig. 1 A molecular diagram of closo-1-SeB₉Cl₉ (2) of a symmetry of C_4v. Salient interatomic distances are (in Å at B3LYP/6-311+G**): r[Se–B(2)] 2.127, r[B(2)–B(3)] 2.013, r[B(2)–B(6)] 1.895, and r[B(6)–B(10)] 1.714.

Results and discussion

It has been shown in all previous formations of heteroboranes via co-pyrolysis of B₂X₄ with elemental halides in vacuo that the incorporation of heteroatoms can only occur in the gas phase,^10,12 and therefore the temperature during pyrolysis has to be high enough to vaporize both starting components, but not too high to cause the thermal degradation of the prospected heteroborane products (e.g. as a reverse reaction of eqn (4)). Under pyrolytic conditions, both Se₂Cl₂ and B₂Cl₄ are unstable and disproportionate to elemental selenium (eqn (3))¹³ and intermediate [BCl] species¹⁴ (eqn (1)), from which the latter can aggregate to boron subchlorides B_nCl_n (n = 8–12, eqn (2))⁹ in a homo-component reaction or incorporate selenium into condensing [BCl] particles to the selenaborane products (eqn (4) and (5)) in hetero-component reactions.

We have now reexamined the co-pyrolysis of B₂Cl₄ with Se₂Cl₂in vacuo (10⁻³ mbar) by altering both the temperature and molar conditions of our first report from 330 °C and 9 : 1 to (a) 350 °C and 7 : 1 or (b) 400 °C and 10 : 1. After the reaction, the product mixtures were thoroughly extracted with liquid BCl₃, after which the extractant was evaporated in vacuo. The residues were separated into the two fractions volatile upon heating to T ≈ 180 °C in vacuo or volatile between T ≈ 180 °C and T ≈ 350 °C under these conditions, and were analyzed by ¹¹B NMR spectroscopy and mass spectrometry.

Both experiments revealed the formation of closo-selenaboranes and boron subhalides B_nCl_n (n = 8–11), the latter consisting mainly of B₉Cl₉. The selenaborane portion in case (a) contained approximately equal amounts of closo-SeB₅Cl₅ (1a) in the more volatile fraction and closo-SeB₁₁Cl₁₁ (1b) together with smaller amounts of closo-1-SeB₉Cl₉ (2) in the less volatile fraction. In case (b), the more volatile fraction contained only very small amounts of 1a, while the less volatile fraction contained 1b in amounts that were approximately the same as in case (a). 2 could not be detected in (b). These results indicate that the generation of selenaborane 1a is favored both by a lower pyrolysis temperature and by a lower molar content of B₂Cl₄ in the starting mixture compared with the formation of 1b. 2 requires temperatures as high as 350 °C for formation. Due to their similar volatilities compared to those of the boron subhalide side-products B_nCl_n, the selenaboranes 1a, 1b and 2 could not be completely separated by vacuum fractionation, where they sublimed as white solids at approximately 50 °C (1a), 200 °C (2) and 350 °C (1b), respectively. The final crystallization step for the isolation of 1a and 1b afforded a slow cooling of the sealed, evacuated samples from 200 °C (for 1a) or 350 °C (for 1b) to room temperature over a period of 20 h. However, from the obtained crystalline materials, only crystals of 1b were suitable for X-ray diffraction analysis. Selenaborane 2 could only be enriched by the extraction of the less volatile fraction with BCl₃. Selenaboranes 1a, 1b and 2 are soluble in common aprotic hydrocarbon solvents, such as benzene, toluene, methylene chloride and chloroform. They are thermally stable at least up to their formation temperature, i.e. 400 °C.

Eqn (4) and (5) describe the overall conversion, but not the course of molecular formation and possible side reactions. Based on eqn (5), the gross yield of 2 was only 1%.

nB₂Cl₄ → n[BCl] + nBCl₃ (1)

n[BCl] → B_nCl_n (n = 8–12) (2)

2Se₂Cl₂ ⇋ 3Se + SeCl₄ (3)

Se + n[BCl] → SeB_nCl_n (4)

19B₂Cl₄ + Se₂Cl₂ → 2SeB₉Cl₉(2) + 20BCl₃ (5)

In order to predict the relative stability of each member of the closo-SeB_nCl_n family (n = 4–12), we calculated the “synproportionation” energies of reactions between a smaller (n − 1) and a larger (n + 1) selenaborane according to eqn. (6)–(12) to form the member with the number n of B–Cl vertices in between at the B3LYP/6-311+G**+ZPE level. The computations of these energy balances use the most stable positional isomers within a particular closo-selenaborane (for the individual molecular shapes, see also ref. 15). These computations strongly support the extraordinary stabilities of octahedral, bicapped square-antiprismatic and icosahedral closo motifs, which is fully consistent with the trend observed for perhydrogenated carboranes¹⁶ and perchlorinated phosphaboranes.^10d

1-SeB₄Cl₄ + 2-SeB₆Cl₆ → 2 SeB₅Cl₅(1a) −19.7 (6)

SeB₅Cl₅(1a) + 1-SeB₇Cl₇ → 2 2-SeB₆Cl₆ 0.3 (7)

2-SeB₆Cl₆ + 1-SeB₈Cl₈ → 2 1-SeB₇Cl₇ 1.8 (8)

1-SeB₇Cl₇ + 1-SeB₉Cl₉(2) → 2 1-SeB₈Cl₈ 13.7 (9)

1-SeB₈Cl₈ + 2-SeB₁₀Cl₁₀ → 2 1-SeB₉Cl₉(2) −18.8 (10)

1-SeB₉Cl₉(2) + SeB₁₁Cl₁₁(1b) → 2 2-SeB₁₀Cl₁₀ 38.2 (11)

2-SeB₁₀Cl₁₀ + 2-SeB₁₂Cl₁₂ → 2 SeB₁₁Cl₁₁(1b) −68.8(12) (12)

The ¹¹B experimental and computed chemical shifts of 1a and 1b have already been given in our previous report, but as experimental ones without specifications. In analogy with closo-TeB₅Cl₅,¹² the ¹¹B NMR spectrum of 1a shows the reported two broad signals in a 1 : 4 ratio at 23.2 ppm (h_1/2 ≈ 110 Hz) for B(6) and at 4.6 ppm for B(2–5) with the expected cross-peak in the COSY ¹¹B¹¹B NMR spectrum. The latter signal resolves to a quartet either by homodecoupling or by applying the Lorentz–Gaussian transformation with a ¹J(¹¹B¹¹B) of 27 Hz (the computed J value is 30.6 Hz, see below and the ESI†), thereby indicating coupling with B(6). Satellites of the signal at 4.6 ppm assume the coupling of B(2–5) with selenium, given by ¹J(⁷⁷Se¹¹B) ≈ 36.5 Hz (the computed K value is 22 Hz, see below and the ESI†). The ⁷⁷Se NMR signal of 1a is very broad (h_1/2 ≈ 236 Hz) due to the scalar interaction of ⁷⁷Se with the ¹¹B (and ¹⁰B) of B(2–5) at −149 ppm (see below for the computed value).

The ¹¹B NMR spectrum of 1b shows the reported three signals in a 5 : 5 : 1 ratio at −3.3 (B7–11), 3.9 ppm (B2–6) and 17.3 ppm (B12) and the cross-peaks between B(7–11)/B(2–6) and B(7–11)/B(12) in the ¹¹B¹¹B COSY NMR spectrum. The individual values indicate the same trend as in the parent closo-SeB₁₁H₁₁.^7e The ⁷⁷Se NMR signal of 1b is very broad (h_1/2 ≈ 220 Hz) due to the scalar interaction of ⁷⁷Se with the ¹¹B (and ¹⁰B) of B(2–6) at −31 ppm (see below for the computed value).

The ¹¹B NMR spectrum of 2 consists of three signals in an approximately 4 : 4 : 1 ratio at 1.7 ppm, 16.8 ppm and 53.0 ppm (for the computed values, see Table 3). The signal at 1.7 ppm shows ¹¹B¹¹B cross-peak correlation with the other two signals and thus can be assigned to B(6–9), the signal at 16.8 ppm can be assigned to B(2–5), and the intensity-one signal at 53.0 ppm (also indicative of its very strong antipodal downfield shift) can be assigned to B(10). The chemical shifts and assignments of 2 are in agreement with those established for the corresponding parent selenaborane closo-1-SeB₉H₉,^7f namely −19.9 ppm (B(6–9)), −5.4 ppm (B(2–5)) and +73.3 ppm (B(10)), with the last one representing an extreme value of antipodal deshielding. This effect has already been explained in relation to the δ(¹¹B) of B(10) in closo-1-SB₉H₉ to the much less pronounced downfield chemical shift of B(12) in closo-SB₁₁H₁₁.¹⁷

The mass spectrum of selenaborane 2 exhibits a strong parent ion envelope with minor cut-offs, indicating the stepwise abstraction of BCl₂ and BCl₃ fragments, whose intensity patterns are consistent with the calculated spectra based on natural isotopic abundances.

On the basis of the NMR data and simple skeletal-electron counting rules,²2 should adopt a bicapped square-antiprismatic geometry with selenium on apical position 1 contributing four electrons and each B–Cl unit two electrons to the cluster bonding.

Vacuum sublimation has given suitable crystals of 1b, which could be subjected to one of the rare examples of the single-crystal X-ray structure determination of a reported selenaborane.

The compound crystallizes in the non-centrosymmetric space group P2₁ with two crystallographically independent molecules of essentially identical geometry in the asymmetric unit (Fig. 2). In contrast to the structure of closo-TeB₅Cl₅,¹² there are no signs of modulation.

Fig. 2 ORTEP plots of the molecular structure of 1b. For a compilation of salient bond lengths (r, Å), see Tables 1 and 2.

As expected, the crystal packing forces in the crystals of 1b cause a slight deviation from the symmetry of C_5v adopted for a single molecule of this perchlorinated icosahedral selenaborane. Some important bond lengths are collected in Tables 1 and 2. There are two notable geometrical trends: first, a decrease in the B–B bond lengths with increasing distance from the Se atom, and second, a shortening of the B–Cl distances involving boron atoms bound to Se with respect to the others. Se–B bonds range from 2.166(2) to 2.175(2) Å [mean 2.169(3) Å] in molecule 1 and from 2.162(2) to 2.172(2) Å [mean 2.168(3) Å] in molecule 2. These values are significantly longer than those reported for the gas-phase structure of the parent compound closo-SeB₁₁H₁₁ (2.129(2) Å),^7e see also Fig. 2.

Table 1 B–B bond lengths [Å] in 1b. The values in square brackets refer to the second independent molecule

Atoms involved	Range	Mean
{B(2)⋯B(6)}–{B(2)⋯B(6)}	1.928(3) [1.935(3)]⋯1.945(3) [1.946(3)]	1.938(6) [1.942(4)]
{B(7)⋯B(11)}–{B(7)⋯B(11)}	1.805(4) [1.810(3)]⋯1.819(3) [1.814(3)]	1.812(5) [1.812(1)]
{B(7)⋯B(11)}–B(12)	1.789(4) [1.791(3)]⋯1.797(4) [1.802(3)]	1.793(3) [1.796(4)]
{B(2)⋯B(6)}–{B(7)⋯B(11)}	1.769(4) [1.772(4)]⋯1.779(3) [1.782(3)]	1.774(3) [1.776(3)]

---

Table 2 B–Cl bond lengths [Å] in 1b. The values in square brackets refer to the second independent molecule

Bond(s)	Range	Mean
B(2)–Cl(2)⋯B(6)–Cl(6)	1.740(3) [1.736(2)]⋯1.746(2) [1.744(2)]	1.744(2) [1.741(3)]
B(7)–Cl(7)⋯B(11)–Cl(11)	1.757(2) [1.761(2)]⋯1.775(2) [1.772(2)]	1.765(6) [1.766(5)]
B(12)–Cl(12)	1.768(3) [1.762(2)]

---

The shape of the solid-state structure of 1b agrees with that already computed at the B3LYP/6-311+G** level of theory⁸ and with the gas-phase structures of the corresponding icosahedral chalcogenaborane hydrides closo-EB₁₁H₁₁ (E = Se, S) based on gas-phase electron diffraction (E = Se,^7e S ^18a) and microwave spectroscopy (E = S ^18b). However, due to the entirely different physical meanings of the interatomic distances derived using the structural tools mentioned above, the direct comparison must be taken with caution.

We have also examined the bonding motifs and electrostatic properties of 1a, 1b and 2 in bonding in terms of employing intrinsic bond orbitals (IBOs) and electrostatic potential (ESP) molecular surfaces in line with the earlier use of this class of materials.^19,20 As shown in Fig. 3, the ESP computations of 1a, 1b and 2 provide different patterns. The IBO charges on Se are 0.43 for 1a, 0.57 for 1b and 0.69 for 2. The application of the IBO approach has revealed the nature of bonding in these chlorinated selenaboranes (see Fig. 4 for example). The selenium atom seems to play a decisive role in the three kinds of IBOs in each of these three systems. According to the expansion coefficients (ECs: contributions from individual atoms to a particular IBO to illustrate the nature of such orbitals) associated with the contributions of individual atoms to these orbitals, they may be grouped as follows: there are two emerging IBOs with ECs (the contributing atoms in parentheses) such as 1.05 (Se), 0.52 (B) and 0.25 (B) as well as one IBO with ECs of 1.19 (Se), 0.57 (B) and 0.10 (B) in 1a. Whereas the first pair of IBOs may be considered of almost a 3c–2e nature, the second one is more or less of the 2c–2e-type. In 1b, the three IBOs in which Se is involved have the following ECs (the contributing atoms in parentheses): 1.07 (Se), 0.52 (B), 0.20 (B), and 0.12 (B); 1.07 (Se), 0.42 (B), and 0.37 (B); 1.07 (Se), 0.49 (B) and 0.27 (B). While the first IBO may be viewed as an orbital somewhere between 4c–2e and 2c–2e, the other two are of a 3c–2e nature. Finally, the bonding in the newly formed 2 is also based on the three IBOs grouped in two patterns, i.e. one of a 2c–2e nature with the ECs 1.15 (Se) and 0.56 (B) and two of the 3c–2e-type with the ECs 1.1 (Se), 0.54 (B) and 0.26 (B). The bonding in the other hemispheres (and between the hemispheres in 1b and 2) is predominantly of a 3c–2e nature, and the corresponding ECs are in the range of ca. 0.50–0.70. All the relevant computational details are provided in the ESI.†

Fig. 3 Computed electrostatic potentials (ESPs) on the 0.001 a.u. molecular surfaces of closo-SeB₅Cl₅ (1a, A), closo-1-SeB₉Cl₉ (2, B) and closo-SeB₁₁Cl₁₁ (1b, C). The ESP color range is in kcal mol⁻¹.

Fig. 4 Examples of bonding in closo-SeB₅Cl₅ (1a, A), closo-1-SeB₉Cl₉ (2, B) and closo-SeB₁₁Cl₁₁ (1b, C) as revealed by the application of the IBO computational approach. Color coding: blue – 2c–2e, pink – 3c–2e, and yellow – 4c–2e.

It is the σ-hole concept that connects the popular halogen bonding with chalcogen, pnictogen bondings and other related noncovalent interactions.²¹ The σ-hole is characterized by its magnitude, abbreviated as V_s,max. V_s,max is defined as the value of the most positive ESP of an electron density surface. Thus, the V_s,max values of closo-SeB₅Cl₅ (1a), closo-1-SeB₉Cl₉ (2) and closo-SeB₁₁Cl₁₁ (1b) are computed to be 25.8, 41.1 and 44.8 kcal mol⁻¹, respectively. The last result shows that the icosahedral perchlorinated selenaborane has a more positive V_s,max than closo-SeB₁₁H₁₁ (V_s,max of 29.5 kcal mol⁻¹).²² The last value is comparable with that of nido-7,8,9,11-Sb₂C₂B₇H₉ (V_s,max of 42.7 kcal mol⁻¹)²³ and explains why the chalcogen bonding²⁴ in 1b is so strong in terms of the crystal packing of the corresponding solid-state structure.

The above B3LYP/6-311+G** geometry of 2 has also been used in NMR shift computations using the GIAO-PBE1PBE model chemistry with the same basis set as in the derivation of the molecular geometry of 2, likewise successfully utilized in previous studies.^12,20 The most striking feature of the molecular geometry of 2 is the considerable expansion of the square belt adjacent to selenium (see Fig. 1), with the analogical pentagonal belt adjacent to Se in 1b exhibiting the same pattern.

We have also provided the computed ⁷⁷Se NMR chemical shifts (in ppm, with respect to Me₂Se; experimental values of compounds 1a and 1b are given in parentheses; see Table 3),²⁵ which are nicely related to the experimental values. The ¹¹B NMR chemical shifts of 2 at the same GIAO-PBE1PBE//B3LYP/6-311+G** level are computed as shown in Table 3. Note that B(10) in 2 resonates at ca. 20 ppm less (in the upfield direction) with respect to the parent closo-1-SB₉H₉.¹⁷ Basically, there is very good agreement with the computed values, i.e. the B3LYP/6-311+G** internal coordinates represent a very good approximation of the molecular geometry of 2 in solution.

Table 3 Computed^a and experimental ⁷⁷Se and ¹¹B chemical shifts with respect to BF₃.OEt₂ and Me₂Se, respectively, for a series of bicapped square-antiprismatic, octahedral, and icosahedral selenaboranes

Bicapped square-antiprismatic motifs
	Se	B(2–5)	B(6–9)	B(10)
2	−511	16.6	−1.9	51.2
Exp.	—	16.8	1.7	52.3
closo-SeB9H9	−231	−5.9	−23.5	78.0
Exp.	—	−5.4	−19.9	73.3
10-Cl-closo-SeB9H8	−519	−9.4	−24.7	76.0

---

Octahedral motifs
	Se	B(2–5)	B(6)
1a	−216	5.5^c	18.2^c
Exp.	−146.2	4.6	23.2
closo-SeB5H5	45	4.8	31.9
6-Cl-closo-SeB5H4	−406	1.7	26.6

---

Icosahedral motifs
	Se	B(2–6)	B(7–11)	B(12)
a GIAO-PBE1PBE//B3LYP/6-311+G**. b From ref. 7f. c The computed ^11B chemical shifts are compatible with those derived at GIAO-B3LYP//B3LYP/6-311+G** given in ref. 1. d The experimental chemical shifts are compatible with those given in ref. 8, see also the ESI.† e The computed ^11B chemical shifts are compatible with those given in ref. 7e and with the experimental values reported therein.
1b	−92	3.4^c	−7.0^c	14.4^c
Exp.	−31	3.9	−3.3^d	17.3
closo-SeB11H11	213	−3.8^e	−5.9^e	25.2^e
12-Cl-closo-SeB11H10	−57	−5.4	7.0	29.1

---

The most striking feature of the electronic structure of 2 is the resonation of the B(10) vertex in the ¹¹B NMR spectrum at a frequency of ca. 52 ppm. This is a result of the intensive paramagnetic contribution to the shielding tensor in this bicapped square-antiprismatic system. Such a contribution arises from the coupling of suitably occupied and unoccupied MOs, i.e. the HOMO and the LUMO, with large coefficients on the antipodal atom B(10). For this overlap in 2, see Fig. 5. This explanation is also valid in the case of closo-1-SB₉H₉, where the overlap is even more pronounced, which is in line with the higher downfield ¹¹B NMR chemical shift of this atom, namely ca. 75 ppm. Interestingly, there is a significant correlation between δ's (⁷⁷Se) and δ's (¹¹B-antipodal) for 1a, 1b, and 2 (see Scheme 1), which indicates a direct transformation of the electron density from an obviously deshielded antipodal boron atom towards the shielded selenium through the corresponding body diagonal. The latter observation might serve as an additional proof of the antipodal effect for these three structural motifs. On the selenium bicapped square-antiprismatic front, we have also looked at closo-1-SeB₉H₉, for which the ¹¹B chemical shifts are known.^7f The downfield shift of B(10) in the latter compound is even larger than that in 2 and also that in closo-1-SB₉H₉ mentioned earlier, see also Table 3. In contrast, the computed upfield shift of ⁷⁷Se is less pronounced than that in 2, see also Scheme 1 that also shows a nice correlation between δ's (⁷⁷Se) and δ's (¹¹B-antipodal) for SeB₅H₅, 1-SeB₉H₉, and SeB₁₁H₁₁. In contrast, there is no such a correlation between δ's (⁷⁷Se) and δ's (¹¹B-antipodal) for 4-Cl-SeB₅H₄, 10-Cl-SeB₉H₈, and 12-Cl-SeB₁₁H₁₀ (r² = 0.43).

Fig. 5 The HOMO (A) and the LUMO (B) of closo-1-SeB₉Cl₉ (2).

Scheme 1 Correlation between δ's (⁷⁷Se) and δ's (¹¹B-antipodal) for 1a, 1b, and 2 (in black, r² = 0.96) and that for their hydrogen-based analogues (in red, r² = 0.93).

The coupling constants ¹J(⁷⁷Se¹¹B) are computed to be zero for each of the Se–B and Se⋯B pairs for all the compounds under scrutiny, obviously due to a very large amount of Se isotopes. Therefore, the coupling constants K are computed instead, i.e. exclusively those between the ⁷⁷Se and ¹¹B nuclei. Since spins of these nuclei differ, viz 3/2 for ¹¹B and 1/2 for ⁷⁷Se, the former exhibits four spin functions with eigenvalues of −3/2, −1/2, 1/2, and 3/2 in contrast to just two such functions for ⁷⁷Se (−1/2, 1/2). For the algebraic derivation of such function for ¹¹B, see ref. 26. On this basis the coupling constants K (⁷⁷Se ¹¹B) are relatively small. However, computations of the coupling constants of the J-type between the B atoms were performed (the ¹⁰B isotope is also included in the computations) and the values obtained are obviously larger than the K ones.

Interactions in the crystal structure of 1b were examined using a cluster model around molecule 2 (according to X-ray labeling). Two-body and many-body interaction energy (ΔE² and ΔE^MB) values were computed between the selected central molecule 2 and two layers of the surrounding molecules (see Fig. 6; individual molecules are labeled as A, B, and C). The surroundings molecules within 5 Å of the central molecule 2 formed the first layer; the second layer was formed by molecules within 5 Å of the first layer. The obtained overall binding of molecule 2 is summarized in Table 4. Molecule 1 is expected to exhibit very similar energetic balances. Concerning the energy criteria in the crystals of 1b, the ΔE² values of the most favorable interaction motifs in the reported X-ray crystal structure of 1b are summarized in Table 5. The chalcogen-bonding motifs (i.e. the motifs 2:A and 2:B) had ΔE² values of −7.93 and −7.20 kcal mol⁻¹ at the DFT-D3 level, thus accounting for about 33% of the total computed binding of 1b. The SAPT0 decomposition showed that the chalcogen-bonding motifs were stabilized mainly by dispersion, which formed approximately 57% of the attractive terms. The second most important term was electrostatic, forming about 31% of the attractive terms. The remaining motifs had considerably less negative ΔE² values with a very large contribution of the dispersion energy (about 84%).

Fig. 6 The cluster model of the reported crystals of closo-SeB₁₁Cl₁₁ (up). The central molecule is colored by the element: the 1^st layer in black and the 2^nd layer in gray. The most stable interaction motifs from the crystals (down). Close contacts below the sum of van der Waals radii are in Å. The atom color coding is as follows: orange – selenium, green – chlorine, and forest green – boron.

Table 4 Two-body and many-body interaction energies (ΔE² and ΔE^MB) computed at the DFT-D3/TPSS/TZVPP level in kcal mol⁻¹

Compound	∑ΔE^2 (1^st layer)	∑ΔE^MB (1^st layer)	∑ΔE^2 (2^nd layer)	Total
SeB11Cl11 (1b)	−46.87	5.46	−4.09	−45.50

---

Table 5 Interaction energies (ΔE²) computed at the DFT-D3/TPSS/TZVPP and SAPT0/jun-cc-pVDZ levels. ΔE² have been decomposed into electrostatic (E_elec), exchange (E_exch), induction (E_ind) and dispersion (E_disp) contributions using the SAPT methodology. All energies are in kcal mol⁻¹. The relative values in parentheses show the contribution to the sum of all the attractive terms

	DFT-D3	SAPT0
		Total	E elec	E exch	E ind	E disp
2:A	−7.93	−9.34	−8.06 (31.7%)	16.06	−2.95 (11.6%)	−14.39 (56.6%)
2:B	−7.20	−8.12	−6.61 (30.6%)	13.46	−2.51 (11.6%)	−12.47 (57.8%)
2:C	−3.98	−3.32	−0.85 (11.6%)	4.00	−0.31 (4.2%)	−6.16 (84.2%)

---

Conclusions

The successful chemistry of chlorinated telluraboranes and pnictogenaboranes initiated the repetition of the vacuum co-pyrolysis of B₂Cl₄ with Se₂Cl₂ at various molar ratios and temperatures in search of the generation of other polyhedral closo-selenaboranes than SeB₅Cl₅ (1a) and SeB₁₁Cl₁₁ (1b). Indeed, bicapped square-antiprismatic closo-1-SeB₉Cl₉ (2) was detected this time by 1- and 2-D ¹¹B NMR spectroscopy and exact mass spectrometry. In addition, vacuum sublimation provided suitable crystals of 1b, which were successfully diffracted by means of the single-crystal X-ray technique, and the first solid-state structure in the family of chlorinated selenaboranes was determined; in it, the structure of 1b slightly departs from the symmetry of C_5v, apparently due to the corresponding crystal packing. Interestingly, the computations at the DFT-D3 level have revealed that 33% of the total computed binding motifs in the 1b crystal are caused by very strong chalcogen bonding. The subsequent SAPT decomposition illustrates that the bonding motifs in the crystals are mainly attributable to the dispersion and electrostatic terms. The more refined techniques of ¹¹B NMR and ⁷⁷Se NMR spectroscopy have resolved the coupling constants and ⁷⁷Se chemical shifts of 1a and 1b, which are in good agreement with their computed values. The extraordinary downfield ¹¹B NMR chemical shift of B(10) in 2 has been ascribed to the intensive paramagnetic contribution to the shielding tensor in this bicapped square-antiprismatic motif. Attempts at obtaining other perhalogenated heteroboranes and the subsequent structural studies are in progress in our laboratories, in particular, in relation to 2D-aromatic heterocycles.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the Czech Science Foundation (project no. 23-05083S). The authors also acknowledge the support by the state of Baden–Württemberg through bwHPC and the German Research Foundation (DFG) through no. INST40/467-1FUGG (Justus cluster).

References

1. R. N. Grimes, Carboranes, Academic Press, 3rd edn, 2016.
2. (a) R. E. Williams, Inorg. Chem., 1971, 10, 210; (b) K. Wade, Adv. Inorg. Chem. Radiochem., 1976, 18, 1; (c) R. E. Williams, Adv. Inorg. Chem. Radiochem., 1976, 18, 67; (d) R. W. Rudolph, Acc. Chem. Res., 1976, 9, 446; (e) R. E. Williams, in Electron Deficient Boron and Carbon Clusters, ed. G. A. Olah, K. Wade and R. E. Williams, Wiley and Sons, New York, 1991, vol. 11; (f) R. E. Williams, Chem. Rev., 1992, 92, 177.
3. D. Hnyk and D. A. Wann, Boron: the Fifth Element, in Challenges and Advances in Computational Chemistry and Physics, ed. D. Hnyk and M. L. McKee, 2016, vol. 20, ch. 2.
4. B. J. Eckstein, L. C. Bron, B. C. Noll, M. P. Moghadasnia, G. J. Balaich and C. M. McGuirk, J. Am. Chem. Soc., 2021, 143, 20207.
5. D. B. Werz, R. Gleiter and F. Rominger, J. Org. Chem., 2004, 69, 2945.
6. L. Vogel, P. Wonner and S. M. Huber, Angew. Chem., Int. Ed., 2019, 58, 1880.
7. (a) SeB₁₀H₁₂ and Se₂B₉H₉: J. L. Little, G. D. Friesen and L. J. Todd, Inorg. Chem., 1977, 16, 869; (b) Se₂B₆H₉⁻, Se₂B₇H₉, SeB₁₈H₂₁⁻, SeB₁₈H₁₉⁻, Se₂B₁₇H₁₈⁻, and Se₂B₁₈H₁₉⁻: J. Bould, M. G. S. Londesborough, M. Litecká, R. Macías, S. L. Shea, T. D. McGrath, W. Clegg and J. D. Kennedy, Inorg. Chem., 2022, 61, 1899; (c) SeB₁₆H₁₈: J. Bould, O. Tok, M. Litecká, M. G. S. Londesborough and M. Ehn, Inorganica Chim. Acta, 2022, 542, 121148; (d) Se₂B₉H₁₀⁻: J. Bould, O. Tok, W. Clegg, M. G. S. Londesborough, M. Litecká and M. Ehn, Inorganica Chim. Acta, 2023, 547, 121341; (e) SeB₁₁H₁₁: D. Hnyk, D. A. Wann, J. Holub, M. Bühl, H. E. Robertson and D. W. H. Rankin, Dalton Trans., 2008, 1, 96; (f) SeB₉H₁₂⁻ and SeB₉H₉: G. D. Friesen, R. L. Kump and L. J. Todd, Inorg. Chem., 1980, 19, 1485.
8. W. Keller and M. Hofmann, Z. Anorg. Allg. Chem., 2017, 643, 729.
9. For a selected review on the boron subhalides B_nX_n, see: J. A. Morrison, in Advances in Boron and the Boranes, ed. J. F. Liebman, A. Greenberg and R. E. Williams, VCH Publishers, New York, 1988, 151.
10. (a) W. Haubold, W. Keller and G. Sawitzki, Angew. Chem., Int. Ed. Engl., 1988, 27, 925; (b) W. Keller, L. G. Sneddon, W. Einholz and A. Gemmler, Chem. Ber., 1992, 125, 2343; (c) W. Keller, W. Einholz, D. Rudolph and Th. Schleid, Z. Anorg. Allg. Chem., 2017, 643, 664; (d) W. Keller, G. Sawitzki and W. Haubold, Inorg. Chem., 2000, 39, 1282; (e) W. Keller, Z. Anorg. Allg. Chem., 2017, 643, 517.
11. W. Keller, J. Conrad and M. Hofmann, ChemistrySelect, 2023, 8, e202301653.
12. W. Keller, J. Ballmann, M. B. Sárosi, J. Fanfrlík and D. Hnyk, Angew. Chem., Int. Ed., 2023, 62, e202219018.
13. M. Lamoureux and J. Milne, Can. J. Chem., 1989, 67, 1936.
14. R. Weinmann, PhD Thesis, Universität Stuttgart, 1985.
15. J. Holub, J. Vrána, A. Růžička, Z. Růžičková, J. Fanfrlík and D. Hnyk, ChemPlusChem, 2019, 84, 822.
16. P. v. R. Schleyer and K. Najafian, Inorg. Chem., 1998, 37, 3454.
17. D. Hnyk, D. A. Wann, J. Holub, S. Samdal and D. W. H. Rankin, Dalton Trans., 2011, 40, 5734.
18. (a) D. Hnyk, E. Vajda, M. Bühl and P. v. R. Schleyer, Inorg. Chem., 1992, 31, 2464; (b) H. Møllendal, S. Samdal, J. Holub and D. Hnyk, Inorg. Chem., 2003, 42, 3043.
19. P. Melichar, D. Hnyk and J. Fanfrlík, Phys. Chem. Chem. Phys., 2018, 20, 4666.
20. W. Keller, M. Hofmann, M. B. Sárosi, J. Fanfrlík and D. Hnyk, Inorg. Chem., 2022, 61, 16565.
21. T. Clark, M. Hennemann, J. S. Murray and P. Politzer, J. Mol. Model., 2007, 13, 29.
22. A. Pecina, M. Lepšík, D. Hnyk, P. Hobza and J. Fanfrlík, J. Phys. Chem. A, 2015, 119, 1388.
23. J. Holub, P. Melichar, Z. Růžičková, J. Vrána, D. A. Wann, J. Fanfrlík, D. Hnyk and A. Růžička, Dalton Trans., 2017, 46, 13714.
24. Such a value of V_s,max is responsible for the S⋯Ph contact (chalcogen bonding) in the crystal, see: J. Fanfrlík, A. Přáda, Z. Padělková, A. Pecina, J. Macháček, M. Lepšík, J. Holub, A. Růžička, D. Hnyk and P. Hobza, Angew. Chem., 2014, 126, 10303.
25. See, e.g. M. Bühl, W. Thiel, U. Fleischer and W. Kutzelnigg, J. Phys. Chem., 1995, 99, 4000.
26. J. Macháček, M. Bühl, J. Fanfrlík and D. Hnyk, J. Phys. Chem. A, 2017, 121, 9631.

---

Footnote

† Electronic supplementary information (ESI) available: Descriptions of experiments, computational and crystallographic details, ¹¹B and ⁷⁷Se NMR spectra, and ESI-MS spectra. CCDC 2251721. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d3dt02987e

---