HBr or not HBr? That is the question: crystal structure of 6-hy­droxy-1,4-diazepane-1,4-diium dibromide redetermined

Piontek, M.; Morgenstern, B.; Steinbrück, N.; Oberhausen, B.; Kickelbick, G.; Hegetschweiler, K.

doi:10.1107/S2053229619005321

research papers

STRUCTURAL
CHEMISTRY

ISSN: 2053-2296

Volume 75| Part 6| June 2019| Pages 678-685

https://doi.org/10.1107/S2053229619005321

Open

access

HBr or not HBr? That is the question: crystal structure of 6-hydroxy-1,4-diazepane-1,4-diium dibromide redetermined

Mateusz Piontek,^a Bernd Morgenstern,^a Nils Steinbrück,^b Bastian Oberhausen,^b Guido Kickelbick ^b and Kaspar Hegetschweiler ^a ^*

^aInorganic Chemistry, Universität des Saarlandes, Campus C4.1, 66123 Saarbrücken, Germany, and ^bInorganic Solid State Chemistry, Universität des Saarlandes, Campus C4.1, 66123 Saarbrücken, Germany
^*Correspondence e-mail: hegetschweiler@mx.uni-saarland.de

Edited by M. Kubicki, Adam Mickiewicz University, Poland (Received 25 March 2019; accepted 17 April 2019; online 16 May 2019)

Liu et al. [Chin. J. Struct. Chem. (1996 ). 15, 371–373] reported the structure of 6-hydroxy-1,4-diazepane di(hydrogen bromide), C₅H₁₂N₂O·2HBr, which was interpreted in terms of neutral diazepane and HBr molecules. We found, however, ample evidence that the formation of an organic salt, consisting of a diammonium cation and two bromide anions, is more plausible. This interpretation is also in agreement with thermogravimetric analysis and with the observed solution behaviour. The crystal structure of 6-hydroxy-1,4-diazepane-1,4-diium dibromide, C₅H₁₄N₂O²⁺·2Br⁻, measured at 142 K, crystallized in the orthorhombic space group P2₁2₁2₁. The structure displays O—H⋯Br and N—H⋯Br hydrogen bonding. Contact distances are given. A search in the Cambridge Structural Database for the singly-bonded H—Br moiety revealed a total of 69 structures. The question, whether these structures really include HBr as neutral molecules or rather Br⁻ anions and a protonated substrate such as an amine, is addressed.

Keywords: dihydrobromide; crystal structure; protonated diamine; misinterpreted H atom; series-termination error.

CCDC reference: 1910784

1. Introduction

6-Hydroxy-1,4-diazepane (dazol) was first synthesized by Saari et al. (1971 ) and has been used as a tridentate facially coordinating metal-complexing agent (Liu et al., 1997a ). The free ligand has been isolated as a dihydrogen bromide, C₅H₁₂N₂O·2HBr, and its crystal structure has been reported [Liu et al., 1996; Cambridge Structural Database (CSD; Groom et al., 2016 ) refcode TOKTIW]. The authors postulated crystallization of the neutral diazepane as a free base together with two HBr molecules. From a chemical point of view, the formation of discrete HBr molecules beside a basic entity is surprising, even taking into account that the situation in the solid state does not necessarily reflect the well-known acid–base properties in aqueous solution. However, a search for molecular H—Br in the CSD (Version 5.20, 2018) gave a total of 69 hits and hence some support for the molecular

model. To shed light on this discrepancy, we have: (i) investigated the chemistry of dazol in aqueous solution using potentiometric titration experiments and pD-dependent ¹H NMR spectroscopy, and (ii) prepared a crystalline sample of the title compound (see Scheme

), repeated the structure determination reported by Liu et al. (1996

) and performed additional thermogravimetric measurements to elucidate the solid-state properties.

2. Experimental

2.1. Synthesis and crystallization

The synthesis of the title compound was performed following the protocol given by Saari et al. (1971) with minor modifications. N,N′-Dazolbis(toluenesulfonamide) was prepared as described by the reaction of N,N′-ethylenebis(toluenesulfonamide) with 2,3-dibromopropan-1-ol. However, in the next step, the detour via the acetate proved not necessary. The two toluenesulfonamide groups could be removed directly without any loss of yield by heating the bis(toluenesulfonamide) suspended in 48% aqueous HBr to 125 °C for 3 h. The clear bright-yellow solution was allowed to cool to room temperature and was then evaporated to dryness under reduced pressure. The resulting solid was washed with diethyl ether and ethanol to yield the title compound as a pale-gray solid (91%). Crystals were grown by slow diffusion of EtOH into an aqueous solution of the product which has been acidified with additional HBr.

2.2. Refinement

Liu et al. (1996) reported an unambiguous location of the C-, N-, O- and Br-atom positions of one dazol moiety and two crystallographically independent Br atoms. It is clear that a reliable assignment of H-atom positions is more difficult, and might even be a highly questionable task, if the high electron density of the two heavy Br atoms is considered. Unfortunately, the data set of Liu et al. (1996) is of rather poor quality. According to the CSD, the data set was recorded at room temperature. Moreover, the information provided by these authors is not really conclusive. In their Table 1 (and in the CIF available from the CSD), the H-atom positions are all listed without standard deviations and the authors stated that `H atoms were located by geometric method except the hydroxyl one, which was oriented from difference Fourier map.' This statement seemingly indicates that the H(—Br), H(—C) and H(—N) positions have not been taken from a difference Fourier map. It is confusing that some of the C—H (1.13 Å) and N—H (1.15 Å) distances and H—C—H (93.2°) angles do not fall in expected ranges. Obviously, the authors did not apply the usual riding model with fixed angles and distances. The two H—Br lengths of 1.04 and 1.08 Å are also rather short.

Table 1
Experimental details

Crystal data
Chemical formula	C₅H₁₄N₂O²⁺·2Br⁻
M_r	278.00
Crystal system, space group	Orthorhombic, P2₁2₁2₁
Temperature (K/°C)	142/−131
a, b, c (Å)	7.7005 (4), 9.2774 (5), 12.6853 (6)
V (Å³)	906.25 (8)
Z	4
Radiation type	Mo Kα
μ (mm⁻¹)	8.89
Crystal size (mm)	0.23 × 0.07 × 0.02

Data collection
Diffractometer	Bruker APEXII CCD
Absorption correction	Multi-scan (SADABS; Krause et al., 2015 )
T_min, T_max	0.576, 0.746
No. of measured, independent and observed [I > 2σ(I)] reflections	4858, 2089, 1789
R_int	0.050
(sin θ/λ)_max (Å⁻¹)	0.652

Refinement
R[F² > 2σ(F²)], wR(F²), S	0.037, 0.063, 0.93
No. of reflections	2089
No. of parameters	106
No. of restraints	5
H-atom treatment	H atoms treated by a mixture of independent and constrained refinement
Δρ_max, Δρ_min (e Å⁻³)	0.61, −0.51
Absolute structure	Flack x determined using 646 quotients [(I⁺) − (I⁻)]/[(I⁺) + (I⁻)] (Parsons et al., 2013 )
Absolute structure parameter	0.05 (2)
Computer programs: APEX2 (Bruker, 2010 ), SAINT (Bruker, 2010), SHELXT2014 (Sheldrick, 2015a ), SHELXL2014 (Sheldrick, 2015b ), DIAMOND (Brandenburg, 2007 ) and PLATON (Spek, 2009 ).

To improve the quality of the data set, we performed a data collection at −131 °C. The space group and lattice parameters, as well as the positional parameters of the non-H atoms, were all in agreement with Liu's report (Liu et al., 1996). Crystal data, data collection and structure refinement details are summarized in Table 1. Inspection of a difference Fourier map unambiguously yielded all of the H(—C) and H(—O) hydrogens with meaningful bond lengths and angles. The positional parameters of these H atoms were now included in a subsequent refinement using a riding model, with an appropriate restraint for H(—O) and appropriate constraints for the H(—C) atoms. At this stage, the refinement provided agreement factors of R₁ = 3.94% and wR₂ = 7.47%. The two most intense peaks, with electron densities of 0.72 and 0.70 e Å⁻³, in the new difference Fourier map were located in proximity to atoms N2 and N1, respectively. They were interpreted as H(—N) positions. Further refinement confirmed this assignment and gave a slight drop of the R₁ (3.85%) and wR₂ (6.64%) values. A subsequent difference Fourier map exhibited more than 30 unassigned peaks with electron densities in the range 0.62–0.42 e Å⁻³. The four peaks with highest intensities (Q1–Q4) were located in proximity to atoms N1, Br2, N2 and Br1 (in this order). At this stage, two different models were considered for the final refinement. Model A comprised Q1 and Q3, which were both interpreted as H(—N) positions; Q2 and Q4 were disregarded. Model B comprised Q2 and Q4 as H(—Br) positions, with Q1 and Q3 now being neglected. Free refinement of model A resulted in a stable and meaningful result, yielding two NH₂⁺ groups, whereas the free refinement of model B collapsed: the H(—Br) atoms moved to positions with very short Br—H distances (<0.3 Å). The agreement factors of model A (R₁ = 3.72% and wR₂ = 6.31%) were marginally better than those of model B (R₁ = 3.84% and wR₂ = 6.63%). These results clearly show that the crystal structure analysis alone does not allow discrimination with certainty between the two models. However, the stable refinement of model A and the slightly better agreement factors may be regarded as a first sign for the ionic structure. The observed electron density in proximity to the Br atoms (Q2 and Q4) could be understood as well-known series-termination errors in the Fourier synthesis (Glusker et al., 1994 ).

In the final refinement (model A), a riding model was used for the C-bonded H atoms. As suggested by Müller et al. (2006 ), the positional parameters of the O- and N-bonded H atoms were refined using isotropic displacement parameters, which were set at 1.5U_eq(O) or 1.2U_eq(N) of the pivot atom. In addition, restraints of 0.84 and 0.88 Å were used for the O—H and N—H distances, respectively.

3. Results and discussion

3.1. Chemical context

Liu et al. (1996) postulated the presence of neutral diazepane as a free base, together with two molecular HBr units in the solid-state structure. At first glance, such an interpretation is amazing, since HBr is known to react as a very strong acid, and the diazepane moiety – as an alicyclic diamine – is expected to react as a base. We investigated the protonation behaviour of dazol in aqueous solution (25 °C) and found – as expected – that an uptake of two protons occurred readily upon addition of acid. A series of potentiometric titration experiments (Fig. 1) revealed two pK_a values of 6.01 and 9.05 (0.1 M KCl) or 6.37 and 9.28 (1 M KNO₃) for H₂dazol²⁺. These values are in agreement with those reported for related amino alcohols (Martincigh & Marsicano, 1995 ). In addition, we also performed a ¹H NMR titration experiment in D₂O (Fig. 2) and observed characteristic pD-dependent resonances for the H(—C) protons upon addition of NaOD. This pD dependency could again be interpreted as a twofold deprotonation reaction of the dication. The evaluated pK_a values in this medium are 6.35 and 9.62. All these characteristics clearly indicate that addition of two equivalents of HBr to an aqueous solution of dazol results in a complete transformation into the H₂dazol²⁺ dication. Crystal growth of the title compound has indeed been performed in such an acidic aqueous medium. However, one must of course be aware that – in general – the solid state does not necessarily depict the equilibrium composition in solution.

Figure 1
Titration curve (25 °C, 0.1 M KCl) of H₂dazol²⁺. Squares refer to the measured values. The red line was calculated using the deprotonation constants −log K_a,1 = 6.01 and −log K_a,2 = 9.05.

Figure 2
pH* dependence of the various ¹H resonances for the nonlabile protons (pH* refers to the uncorrected pH-meter reading in D₂O for an electrode calibrated in H₂O). The observed resonances (δ_obs) are shown as closed circles: H1 (black), H2 (blue) and H3 (red). The lines were calculated [minimization of Σ(δ_obs − δ_calc)²]. Inset: The pattern of H1 at (i) pH* = 5, (ii) pH* = 10 and (iii) in 5 mol l⁻¹ NaOD.

A solid sample of the title compound was therefore investigated by IR spectroscopy, looking at around 2600 cm⁻¹ for any H—Br stretch vibration. However, these measurements were not conclusive, since a possible H—Br peak was covered by the intense and broad absorption between 2200 and 3500 cm⁻¹, caused by the various associated N—H and O—H stretching vibrations. Thermogravimetric measurements combined with an IR analysis of the gaseous products was more instructive (Fig. 3). A 20 mg sample was heated by a rate of 10 °C min⁻¹ from room temperature up to 800 °C and exposed to a steady stream of N₂ (20 ml min⁻¹). Complete degradation occurred almost quantitatively (>90%) in one single step in the range 300–400 °C. Evolution of HBr could readily be recognized in the IR spectrum (2400–2700 cm⁻¹) by its characteristic pattern for the two isotopomers with resolved transitions for the various rotamers (NIST, 2019 ). In addition, an organic component (O—H, N—H and C—H, and possibly C—C and C—O stretching vibrations, but no CO₂) was formed. These findings are in agreement with a predominant sublimation of the product. It is well known that NH₄Br and its organic derivatives, such as methylammonium bromide sublimate in the range 300–400 °C (Ivanov et al., 2019 ). The remaining small nonvolatile residue (5–10%) probably indicates some minor decomposition during the sublimation process (formation of elemental carbon, as indicated by a black coating inside the crucible). If a 1:1 stream of N₂ and O₂ (each 20 ml min⁻¹) is applied to the sample during heating, the degradation occurred in more than one step. Again, less than 1% weight loss was noted below 200 °C. Up to 250 °C, the sample weight decreased slightly by about 3%. A first significant step of decomposition was then observed in the range of about 250–330 °C, with a corresponding weight drop of 21%. This value is clearly smaller than the 29% required for a dissociation of one HBr molecule. The final part of the decomposition reaction occurred in two steps at 350–400 (−50%) and 400–550 °C (−22%). The IR spectra of the evolving gases showed the formation of H₂O during the entire decomposition reaction. At elevated temperatures (>300 °C), increasing formation of CO₂ and of some organic components (observation of C—H stretching vibrations) was also noted. Above 450 °C, CO₂ remained as the most significant decomposition product. Inspection of the spectra did again exhibit that small traces of HBr have been formed. Maximum HBr production was found around 350–400 °C. These observations do not indicate a simple and quantitative loss of HBr at low temperature. It rather appears that small amounts of HBr are formed in situ during the entire decomposition process, particularly at elevated temperatures. As a conclusion, evolution of HBr appears generally to be combined with a complete breakdown of the entire structure.

Figure 3
Thermogravimetric analyses of H₂dazolBr₂ (%weight versus T) together with IR characteristics of the evolving gases at temperatures as indicated. Heating of the sample (a) in a steady stream of N₂ only and (b) in a 1:1 mixture of N₂ and O₂. The inset in (b) shows an enlargement for the range 2400–2700 cm⁻¹, indicating the formation of HBr in small traces.

3.2. Structural commentary

Eliel et al. (1965 ) proposed high conformational flexibility for cycloheptane, with a twist–chair (TC) conformation being of lowest energy. We previously studied 6-amino-1,4-diazepane (daza) as a metal-complexing agent, and the adoption of such a TC conformation for the seven-membered 1,4-diazepane ring of H₃daza³⁺ could indeed be confirmed by crystal structure analysis (Romba et al., 2006 ; Neis et al., 2010 ). Its ¹H NMR spectrum exhibited a total of five resonances for the H(—C) protons, indicating a rapid interconversion of different conformations, yielding an averaged structure of higher symmetry (Longuet-Higgins, 2002 ). The molecular structure of daza and dazol is closely related and similar ¹H NMR characteristics have also been observed for dazol. The coupling pattern of the H—(C—X) proton (H₃daza³⁺: X = NH₃⁺; H₂dazol²⁺: X = OH) revealed, however, some characteristic differences for the two compounds. For H₃daza³⁺, this signal appeared as a triplet of triplets with one large coupling constant of 10.5 Hz and a second much smaller constant of 3.1 Hz. The large coupling of 10.5 Hz is indicative of a staggered orientation of the H—C—C(NH₃⁺)—H fragment, with a torsion angle close to 180°. Obviously, the primary ammonium group of H₃daza³⁺ is placed in an equatorial position. However, for H_xdazol^x+ (x = 0, 1, 2), the corresponding coupling constants are significantly smaller, with a value of 5.4/1.4 Hz at pD 5.5 and 5.8/4.4 Hz at pD 10 (Fig. 2). It thus appears that in solution the hydroxy group of H_xdazol^x+ (x = 0, 1, 2) is positioned axially. Interestingly, at a very high base concentration (5 mol l⁻¹ NaOD), the signal of this proton is again shifted by about 0.2 ppm to lower frequency and appeared as a quintet with one unique coupling constant of 4.0 Hz. Obviously, the hydroxy group of dazol becomes deprotonated in such a highly alkaline medium.

In agreement with our NMR study, the H₂dazol²⁺ cation also adopted a TC conformation in the crystal structure, with the pivot atom N1 located in the isoclinal position. The puckering parameters (Boessenkool & Boeyens, 1980 ) of the seven-membered diazepane ring are Q = 0.821 (7) Å, Q2 = 0.506 (7), Q3 = 0.647 (7), φ₂ = 86.5 (8)° and φ₃ = 92.0 (6)°. Also, in accordance with the solution structure, the hydroxy group adopted an axial position. The different orientation of the NH₃⁺ group of H₃daza³⁺ and the OH group of H_xdazol^x+ (x = 0, 1, 2) is remarkable. Since this difference is observed in the solid state, as well as in solution, the peculiar structure may be explained by the well-known attractive gauche effect (Entrena et al., 1997 ), which proposes the preferential adoption of a gauche rather than a trans conformation for such an X—C—C—OR fragment (X = O, N).

Similar to the work performed by Liu et al. (1996), the crystal structure analysis presented here exhibited low precision, i.e. large standard uncertainties for bond angles and distances. An inspection of the displacement parameters of the C, N and O atoms indicated some significant deviations for the displacement ellipsoids from a spherical shape (Fig. 4). Considering the high conformational flexibility of the seven-membered diazepane frame, we attribute these large deviations to minor disorder rather than to thermal motion (in contrast to Liu's work, we performed data collection at −131 °C). It was, however, not possible to resolve this slight disorder in terms of a superposition of distinct individual conformers.

Figure 4
The molecular structure of the H₂dazol²⁺·2Br⁻ unit, with the atom-numbering scheme and displacement ellipsoids drawn at the 50% probability level.

3.3. Supramolecular features

The cationic H₂dazol²⁺ gravicentres (dgcs) are arranged into layers oriented parallel to the crystallographic ac plane. In these layers, each dgc is surrounded by six neighbouring dgcs, forming a distorted hexagon. The layers are stacked along b in a staggered fashion (ABABAB…) and can thus be regarded as a distorted hexagonal packing. If the two adjacent layers are taken into account, each dgc receives 12 dgc neighbours, which form an anticuboctahedron (Fig. 5). However, some characteristic deviations from a regular shape are noted for this polyhedron. One reason for the distortion originates from the significant deviation of the H₂dazol²⁺ cations from a spherical shape; these cations should be regarded as disks rather than spheres. Consequently, the (dgc)₁₂ anticuboctahedron is compressed along the pseudohexagonal axis, as is expressed by the unequal edge lengths (5.59–7.87 Å) and short interlayer distances. Further distortion originates from the general position of the dgc and the reduced crystallographic symmetry. As a consequence, the dgc layers (and thus the equators of the anticuboctahedra) are puckered. The symmetry class of this layer group is p12₁1 (International Tables for Crystallography, 2002 ; Shubnikov & Koptsik, 1974 ). Interestingly, the Br1 ions (green spheres in Figs. 5 and 6) are located neither in the tetrahedral nor in the octahedral holes of this packing. They are placed almost straight within the pseudohexagonal dgc planes. Each Br1 anion is thus surrounded by three dazol dications. Only 50% of these triangular holes are occupied. Such a packing becomes understandable if the huge difference in size between Br⁻ and H₂dazol²⁺ is considered (Fig. 6). The entire geometry and coordination number of Br1 becomes evident if the staggered arrangement of the dgc layers is taken into account. Beside the three dgc neighbours of the triangular hole, the Br1 anions receive two additional dgc neighbours from adjacent dgc layers and the coordination polyhedron of Br1 can thus be described as a trigonal bipyramid. The Br2 ions (blue spheres) are located in the octahedral holes of the pseudohexagonal dgc packing. The coordination number of Br2 is thus six and the coordination geometry is a distorted octahedron. The H₂dazol²⁺ cations, in turn, are surrounded by 11 Br anions with Br⋯dgc distances ranging from 4.0 to 6.5 Å (Fig. 7a). The resulting Br₁₁ structure can be described as a distorted Edshammar polyhedron (Fig. 7b; Edshammar, 1969 ). Notably, the regular Edshammar polyhedron adopts D_3h symmetry (Fig. 7c) and is a space filler (Lidin et al., 1992 ). Each H₂dazol²⁺ entity forms N—H⋯Br and O—H⋯Br hydrogen bonds (Table 2). Br1 is bonded to H1N—N1, H3N—N2(−x + 1, y + $[{1\over 2}]$ , −z + $[{1\over 2}]$ ) and H4N—N2(−x + $[{1\over 2}]$ , −y + 1, z − $[{1\over 2}]$ ). Further Br1⋯H interactions (Br1⋯H2N—N1, Br1⋯H1—C1, Br1—H2A–C2 and Br1⋯H3B—C3), with Br⋯H—N or Br⋯H—C angles of 131–140° and Br⋯H distances of 2.95–3.05 Å, must be regarded as very weak hydrogen bonding if they are to be regarded as hydrogen bonding at all. The sum of the van der Waals radii of Br and H is 2.95 Å (Bondi, 1964 ). Br2 forms only two unambiguous hydrogen bonds to H1O—O1 and H2N—N1(x + $[{1\over 2}]$ , −y + $[{3\over 2}]$ , −z + 1). Again, the contacts Br2⋯H3A—C3, Br2⋯H4A—C4, Br2⋯H3N—N2, Br2⋯H2N—N1, Br2⋯H2B—C2, Br2⋯H5A—C5, Br2⋯H3A—C3, with Br⋯H distances of 2.88–3.05 Å and Br⋯H—N or Br⋯H—C angles of 117–162° may be considered as further weak interactions which stabilize the structure. A graph-set analysis (Bernstein et al., 1995 ) shows the hydrogen-bonding network in the [1 $[\overline{1}]$ 0] direction (Fig. 8). The Br1 and Br2 anions are μ₃-acceptors and the N—H and O—H hydrogens are donors in three distinct ring systems, i.e. R₇⁴(20), R₄²(8) and R₅³(16). Notably, there is no direct H₂dazol²⁺⋯H₂dazol²⁺ hydrogen bonding and, consequently, the hydroxy group of H₂dazol²⁺ does not act as an acceptor. All these structural features correspond well to the packing of large charged molecular entities, as observed for instance in Zintl phases (Lidin et al., 1992), and thus provide support for the ionic model.

Table 2
Hydrogen-bond geometry (Å, °)

D—H⋯A	D—H	H⋯A	D⋯A	D—H⋯A
O1—H1O⋯Br2	0.84 (1)	2.41 (2)	3.244 (4)	170 (8)
N1—H1N⋯Br1	0.88 (1)	2.36 (2)	3.230 (6)	167 (7)
N1—H2N⋯Br2ⁱ	0.88 (1)	2.79 (6)	3.323 (6)	120 (5)
N2—H3N⋯Br1ⁱⁱ	0.88 (1)	2.69 (5)	3.422 (6)	142 (6)
N2—H4N⋯Br1ⁱⁱⁱ	0.88 (1)	2.55 (3)	3.355 (6)	153 (6)
Symmetry codes: (i) $[x-{\script{1\over 2}}, -y+{\script{3\over 2}}, -z+1]$ ; (ii) $[-x+1, y-{\script{1\over 2}}, -z+{\script{1\over 2}}]$ ; (iii) $[-x+{\script{1\over 2}}]$ , $[-y+1, z+{\script{1\over 2}}]$ .

Figure 5
The distorted anticuboctahedron formed by the gravicentres of 12 H₂dazol²⁺ cations (gray spheres) together with the central H₂dazol²⁺ cation. The two crystallographically independent Br⁻ anions are placed in 50% of the trigonal holes of a hexagonal layer (Br1 = green spheres) and in all of the octahedral holes (Br2 = blue spheres).

Figure 6
View of the (010) plane (space-filling model). The H₂dazol²⁺ cations form a pseudohexagonal layer with the Br1 anions (green spheres) located in every second trigonal hole of this layer. The Br2 anions are shown as light-blue spheres and are placed in the octahedral holes above and below this layer. C, H, N and O atoms are shown as black, light-gray, dark-blue and red spheres, respectively.

Figure 7
(a) The H₂dazol²⁺ cation is surrounded by 11 Br⁻ anions. A coordination number (CN) of 11 is generated by the six Br2 anions forming a trigonal prism (tp, thin solid lines) and the five Br1 anions forming a trigonal bipyramid (tbipy, broken lines). The five vertices of the tbipy are placed over the mid-points of the five planes of the tp. (b) The observed distorted and (c) the idealized undistorted D_3h model (Edshammar polyhedron).

Figure 8
The hydrogen-bonding network observed for H(—N) and H(—O) H atoms as donors, and Br1 (green) and Br2 (blue) as acceptors. The plane perpendicular to the [1 $[\overline{1}]$ 0] direction is shown. The resulting cyclic structures are characterized by the descriptors R₇⁴(20), R₄²(8) and R₅³(16). [Symmetry code: (iii) −x + $[{1\over 2}]$ , −y + 1, z + $[{1\over 2}]$ .]

3.4. Database survey

By studying some related literature, we became aware that the report of Liu et al. (1996) might not be the only example where the nature of `HBr' in a crystal structure should be questioned. A search in Version 5.20 (2018) of the Cambridge Structural Database (CSD; Groom et al., 2016), looking for molecular H—Br (i.e. for an H atom directly connected to a Br atom by a single bond), revealed a total of 69 entries (see supporting information for the full list). Some of the listed entries do not provide three-dimensional coordinates and are thus not relevant for this discussion. Furthermore, it appears that the technical term `hydrobromide' may have led to confusion, since it has been used by some of the authors for a salt of a protonated organic molecule and bromide as counter-ion, but appeared in our search as molecular H—Br. For some of the listed compounds, the presence of HBr molecules in the structure might be quite feasible from a chemical point of view. However, there remained still a total of 13 entries [CSD refcodes BEPQIY (Růžička et al., 2013 ), EVAMIX (Cocco et al., 2004 ), GICSOC (Aureggi et al., 2013 ), KEKQAS (Wang et al., 1999 ), KONVEO (Lin et al., 1990 ), MOMVIV (Surendra Dilip & Gowri, 2014 ), MOMVIV01 (Gowri et al., 2015 ), MUFKON (Liang et al., 2002 ), NAVFOI01 (Zhao et al., 2017 ), NIJGOC (Liu et al., 1997b ), SOCZUH (Banothu et al., 2014 ), UNESAI (Zhang & Shen, 2011 ) and YOTSIJ (Monte et al., 1995 )], where the simultaneous presence of HBr with a basic moiety (in KEKQAS, HBr and OH⁻!) is proposed, with H—Br bond lengths ranging from 0.79 to 1.84 Å. It is clearly beyond the scope of this contribution to decide whether these structural assignments are correct. However, it appears that: (i) the maintainers of the database should carefully clarify whether the expression `hydrobromide' refers to an ionic or rather a molecular model, and (ii) reporting such a crystal structure, authors should take the required care to clarify the bonding mode when postulating incorporation of undissociated HBr together with a basic moiety. Similar considerations might also be applicable for the so called `hydrochlorides' (292 entries for a H—Cl molecule) and `hydroiodides' (20 entries for a H—I molecule).

4. Conclusion

The solution and solid-state properties of the title compound do not provide evidence for a simple incorporation of HBr as intact molecules into the solid-state structure. An ionic model, comprising one H₂dazol²⁺ cation and two Br⁻ anions is clearly a better explanation. According to the structure postulated by Liu et al. (1996), the two HBr molecules would not be bonded to other moieties by strong interactions (such as Coulombic forces or hydrogen bonding). The (Br1—)H⋯H(—Br2) separation of 2.24 Å roughly corresponds to the sum of the van der Waals radii of two H atoms (2.20 Å; Bondi, 1964) and, as a consequence, the two HBr molecules would not act as H-atom donors in hydrogen bonds. One could thus expect that liberation of HBr should occur readily even at moderate temperatures.

For further clarification of this question, we grew single crystals of the title compound and repeated the X-ray analysis. We have shown that it is possible to refine the crystal structure in terms of an ordinary ammonium salt (see §2.2, Refinement). There is thus no need to postulate the rather exotic incorporation of molecular HBr into the crystal structure. We think that, in such a case, it is more advisable to chose the model which also directly explains the observed chemical properties (acid–base behaviour and breakdown of the structure upon HBr elimination above 300 °C).

Supporting information

CCDC reference: 1910784

Crystal structure: contains datablock I. DOI: https://doi.org/10.1107/S2053229619005321/ku3242sup1.cif

Structure factors: contains datablock I. DOI: https://doi.org/10.1107/S2053229619005321/ku3242Isup2.hkl

Details of the CSD search. DOI: https://doi.org/10.1107/S2053229619005321/ku3242sup3.pdf

Computing details top

Data collection: APEX2 (Bruker, 2010); cell refinement: SAINT (Bruker, 2010); data reduction: SAINT (Bruker, 2010); program(s) used to solve structure: SHELXT2014 (Sheldrick, 2015a); program(s) used to refine structure: SHELXL2014 (Sheldrick, 2015b); molecular graphics: DIAMOND (Brandenburg, 2007); software used to prepare material for publication: PLATON (Spek, 2009).

6-Hydroxy-1,4-diazepane-1,4-diium dibromide top

Crystal data top

C₅H₁₄N₂O²⁺·2Br⁻	D_x = 2.038 Mg m⁻³
M_r = 278.00	Mo Kα radiation, λ = 0.71073 Å
Orthorhombic, P2₁2₁2₁	Cell parameters from 1133 reflections
a = 7.7005 (4) Å	θ = 3.8–25.7°
b = 9.2774 (5) Å	µ = 8.89 mm⁻¹
c = 12.6853 (6) Å	T = 142 K
V = 906.25 (8) Å³	Needle, colourless
Z = 4	0.23 × 0.07 × 0.02 mm
F(000) = 544

Data collection top

Bruker APEXII CCD diffractometer	1789 reflections with I > 2σ(I)
Radiation source: sealed tube	R_int = 0.050
φ and ω scans	θ_max = 27.6°, θ_min = 2.7°
Absorption correction: multi-scan (SADABS; Krause et al., 2015)	h = −8→10
T_min = 0.576, T_max = 0.746	k = −10→12
4858 measured reflections	l = −16→15
2089 independent reflections

Refinement top

Refinement on F²	Secondary atom site location: difference Fourier map
Least-squares matrix: full	Hydrogen site location: mixed
R[F² > 2σ(F²)] = 0.037	H atoms treated by a mixture of independent and constrained refinement
wR(F²) = 0.063	w = 1/[σ²(F_o²)] where P = (F_o² + 2F_c²)/3
S = 0.93	(Δ/σ)_max < 0.001
2089 reflections	Δρ_max = 0.61 e Å⁻³
106 parameters	Δρ_min = −0.51 e Å⁻³
5 restraints	Absolute structure: Flack x determined using 646 quotients [(I+)-(I-)]/[(I+)+(I-)] (Parsons et al., 2013)
Primary atom site location: structure-invariant direct methods	Absolute structure parameter: 0.05 (2)

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

	x	y	z	U_iso*/U_eq
C1	0.5844 (9)	0.5995 (7)	0.5028 (5)	0.0123 (15)
H1	0.6966	0.6299	0.5357	0.015*
C2	0.5905 (8)	0.6493 (8)	0.3890 (5)	0.0133 (15)
H2A	0.6493	0.7441	0.3860	0.016*
H2B	0.6615	0.5803	0.3479	0.016*
C3	0.2761 (9)	0.5574 (9)	0.3664 (5)	0.0152 (16)
H3A	0.1864	0.5590	0.3106	0.018*
H3B	0.2205	0.5892	0.4329	0.018*
C4	0.3390 (9)	0.4052 (8)	0.3801 (5)	0.0144 (17)
H4A	0.2443	0.3376	0.3617	0.017*
H4B	0.4370	0.3874	0.3313	0.017*
C5	0.5688 (8)	0.4385 (8)	0.5208 (5)	0.0139 (16)
H5A	0.6604	0.3890	0.4797	0.017*
H5B	0.5902	0.4181	0.5963	0.017*
N1	0.4167 (7)	0.6626 (8)	0.3381 (4)	0.0133 (13)
H1N	0.433 (8)	0.645 (8)	0.2703 (17)	0.016*
H2N	0.382 (9)	0.753 (3)	0.340 (5)	0.016*
N2	0.3970 (9)	0.3769 (6)	0.4909 (5)	0.0139 (14)
H3N	0.403 (10)	0.2828 (18)	0.498 (6)	0.017*
H4N	0.321 (7)	0.410 (7)	0.537 (4)	0.017*
O1	0.4508 (6)	0.6777 (6)	0.5558 (3)	0.0180 (12)
H1O	0.491 (9)	0.661 (9)	0.616 (3)	0.027*
Br1	0.48660 (9)	0.54448 (8)	0.10235 (5)	0.01472 (18)
Br2	0.61165 (9)	0.65666 (9)	0.79166 (5)	0.01602 (18)

Atomic displacement parameters (Å²) top

	U¹¹	U²²	U³³	U¹²	U¹³	U²³
C1	0.010 (4)	0.012 (4)	0.014 (3)	−0.003 (3)	−0.002 (3)	−0.002 (3)
C2	0.009 (3)	0.014 (4)	0.017 (3)	−0.003 (3)	0.000 (3)	0.002 (4)
C3	0.017 (4)	0.018 (4)	0.010 (3)	0.002 (4)	0.001 (3)	0.003 (4)
C4	0.010 (4)	0.017 (4)	0.016 (4)	−0.001 (3)	−0.006 (3)	−0.001 (3)
C5	0.010 (3)	0.017 (4)	0.014 (3)	−0.001 (3)	−0.003 (3)	0.005 (3)
N1	0.016 (3)	0.013 (3)	0.011 (3)	0.003 (3)	0.001 (2)	0.002 (3)
N2	0.014 (3)	0.014 (4)	0.013 (3)	0.006 (3)	−0.003 (2)	0.005 (3)
O1	0.019 (3)	0.021 (3)	0.013 (2)	0.007 (3)	−0.004 (2)	−0.005 (2)
Br1	0.0136 (3)	0.0172 (4)	0.0134 (3)	0.0003 (3)	−0.0008 (3)	−0.0018 (3)
Br2	0.0166 (4)	0.0195 (4)	0.0120 (3)	−0.0024 (4)	0.0007 (3)	0.0008 (4)

Geometric parameters (Å, º) top

C1—O1	1.427 (8)	C4—N2	1.498 (8)
C1—C5	1.516 (10)	C4—H4A	0.9900
C1—C2	1.517 (9)	C4—H4B	0.9900
C1—H1	1.0000	C5—N2	1.490 (9)
C2—N1	1.491 (8)	C5—H5A	0.9900
C2—H2A	0.9900	C5—H5B	0.9900
C2—H2B	0.9900	N1—H1N	0.883 (13)
C3—N1	1.502 (9)	N1—H2N	0.882 (13)
C3—C4	1.503 (10)	N2—H3N	0.879 (13)
C3—H3A	0.9900	N2—H4N	0.884 (13)
C3—H3B	0.9900	O1—H1O	0.839 (13)

O1—C1—C5	111.9 (6)	C3—C4—H4B	109.3
O1—C1—C2	108.4 (5)	H4A—C4—H4B	107.9
C5—C1—C2	116.5 (6)	N2—C5—C1	114.2 (6)
O1—C1—H1	106.5	N2—C5—H5A	108.7
C5—C1—H1	106.5	C1—C5—H5A	108.7
C2—C1—H1	106.5	N2—C5—H5B	108.7
N1—C2—C1	114.2 (5)	C1—C5—H5B	108.7
N1—C2—H2A	108.7	H5A—C5—H5B	107.6
C1—C2—H2A	108.7	C2—N1—C3	119.3 (6)
N1—C2—H2B	108.7	C2—N1—H1N	106 (4)
C1—C2—H2B	108.7	C3—N1—H1N	103 (5)
H2A—C2—H2B	107.6	C2—N1—H2N	110 (5)
N1—C3—C4	113.9 (6)	C3—N1—H2N	113 (5)
N1—C3—H3A	108.8	H1N—N1—H2N	104 (7)
C4—C3—H3A	108.8	C5—N2—C4	115.9 (6)
N1—C3—H3B	108.8	C5—N2—H3N	108 (5)
C4—C3—H3B	108.8	C4—N2—H3N	107 (5)
H3A—C3—H3B	107.7	C5—N2—H4N	106 (5)
N2—C4—C3	111.7 (6)	C4—N2—H4N	111 (4)
N2—C4—H4A	109.3	H3N—N2—H4N	109 (7)
C3—C4—H4A	109.3	C1—O1—H1O	94 (5)
N2—C4—H4B	109.3

O1—C1—C2—N1	−44.8 (8)	C1—C2—N1—C3	−35.3 (9)
C5—C1—C2—N1	82.5 (8)	C4—C3—N1—C2	−39.4 (8)
N1—C3—C4—N2	88.3 (7)	C1—C5—N2—C4	55.1 (8)
O1—C1—C5—N2	55.6 (8)	C3—C4—N2—C5	−75.7 (7)
C2—C1—C5—N2	−69.9 (8)

Hydrogen-bond geometry (Å, º) top

D—H···A	D—H	H···A	D···A	D—H···A
O1—H1O···Br2	0.84 (1)	2.41 (2)	3.244 (4)	170 (8)
N1—H1N···Br1	0.88 (1)	2.36 (2)	3.230 (6)	167 (7)
N1—H2N···Br2ⁱ	0.88 (1)	2.79 (6)	3.323 (6)	120 (5)
N2—H3N···Br1ⁱⁱ	0.88 (1)	2.69 (5)	3.422 (6)	142 (6)
N2—H4N···Br1ⁱⁱⁱ	0.88 (1)	2.55 (3)	3.355 (6)	153 (6)

Symmetry codes: (i) x−1/2, −y+3/2, −z+1; (ii) −x+1, y−1/2, −z+1/2; (iii) −x+1/2, −y+1, z+1/2.

Acknowledgements

We thank Dr Volker Huch for the measurement of the data set. Funding for this research was provided by Universität des Saarlandes.

References

Aureggi, V., Ehmke, V., Wieland, J., Schweizer, W. B., Bernet, B., Bur, D., Meyer, S., Rottmann, M., Freymond, C., Brun, R., Breit, B. & Diederich, F. (2013). Chem. Eur. J. 19, 155–164. Web of Science CSD CrossRef CAS PubMed Google Scholar
Banothu, J., Khanapur, M., Basavoju, S., Bavantula, R., Narra, M. & Abbagani, S. (2014). RSC Adv. 4, 22866–22874. Web of Science CSD CrossRef CAS Google Scholar
Bernstein, J., Davis, R. E., Shimoni, L. & Chang, N.-L. (1995). Angew. Chem. Int. Ed. Engl. 34, 1555–1573. CrossRef CAS Web of Science Google Scholar
Boessenkool, I. K. & Boeyens, J. C. A. (1980). J. Cryst. Mol. Struct. 10, 11–18. CrossRef CAS Web of Science Google Scholar
Bondi, A. (1964). J. Phys. Chem. 68, 441–451. CrossRef CAS Web of Science Google Scholar
Brandenburg, K. (2007). DIAMOND. Crystal Impact GbR, Bonn, Germany. Google Scholar
Bruker (2010). APEX and SAINT. Bruker AXS Inc., Madison, Wisconsin, USA. Google Scholar
Cocco, M. T., Congiu, C., Onnis, V. & Ianelli, S. (2004). Heterocycles, 63, 259–270. CAS Google Scholar
Edshammar, L.-E. (1969). PhD thesis, University of Stockholm, Sweden. Google Scholar
Eliel, E. L., Allinger, N. L., Angyal, S. J. & Morrison, G. A. (1965). In Conformational Analysis. New York: Interscience Publishers–John Wiley. Google Scholar
Entrena, A., Campos, J., Gomez, J. A., Gallo, M. A. & Espinosa, A. (1997). J. Org. Chem. 62, 337–349. CrossRef PubMed CAS Web of Science Google Scholar
Glusker, J. P., Lewis, M. & Rossi, M. (1994). In Crystal Structure Analysis for Chemists and Biologists. Weinheim: VCH Publishers. Google Scholar
Gowri, S., Uma Devi, T., Priya, S., Surendra Dilip, C., Selvanayagam, S. & Lawrence, N. (2015). Spectrochim. Acta Part A, 143, 192–199. Web of Science CSD CrossRef CAS Google Scholar
Groom, C. R., Bruno, I. J., Lightfoot, M. P. & Ward, S. C. (2016). Acta Cryst. B72, 171–179. Web of Science CrossRef IUCr Journals Google Scholar
International Tables for Crystallography (2002). Vol. E, Subperiodic Groups, edited by V. Kopsky & D. B. Litvin. Chester: International Union of Crystallography, and Dordrecht: Kluwer Academic Publishers. Google Scholar
Ivanov, I. L., Mazurin, M. O., Tsvetkov, D. S., Malyshkin, D. A., Sereda, V. V. & Zuev, A. Y. (2019). Thermochim. Acta, 674, 58–62. Web of Science CrossRef CAS Google Scholar
Krause, L., Herbst-Irmer, R., Sheldrick, G. M. & Stalke, D. (2015). J. Appl. Cryst. 48, 3–10. Web of Science CSD CrossRef ICSD CAS IUCr Journals Google Scholar
Liang, F., Li, Y.-Z., Wu, C.-T. & Liu, S.-M. (2002). Acta Cryst. E58, m950–m952. Web of Science CSD CrossRef IUCr Journals Google Scholar
Lidin, S., Popp, T., Somer, M. & Schnering, H. G. (1992). Angew. Chem. Int. Ed. Engl. 31, 924–927. CrossRef Web of Science Google Scholar
Lin, L.-Z., Cordell, G. A., Ni, C.-Z. & Clardy, J. (1990). Phytochemistry, 29, 965–968. CSD CrossRef CAS Web of Science Google Scholar
Liu, S.-H., Xue, G.-P., Zhang, Z.-H., Fu, E.-Q., Wu, C.-T. & Yang, Q.-C. (1997a). Chin. J. Struct. Chem. 16, 355–357. Google Scholar
Liu, Y., Xue, G.-P., Wu, C.-T., Luo, B.-S. & Chen, L.-R. (1996). Chin. J. Struct. Chem. 15, 371–373. CAS Google Scholar
Liu, Y., Xue, G.-P., Wu, C.-T., Luo, B.-S. & Chen, L.-R. (1997b). Chin. J. Struct. Chem. 16, 338–341. CAS Google Scholar
Longuet-Higgins, H. C. (2002). Mol. Phys. 100, 11–20. Google Scholar
Martincigh, B. S. & Marsicano, F. (1995). J. Coord. Chem. 34, 365–381. CrossRef CAS Web of Science Google Scholar
Monte, A. P., Marona-Lewicka, D., Kanthasamy, A., Sanders-Bush, E. & Nichols, D. E. (1995). J. Med. Chem. 38, 958–966. CSD CrossRef CAS PubMed Web of Science Google Scholar
Müller, P., Herbst-Irmer, R., Spek, A. L., Schneider, T. R. & Sawaya, M. R. (2006). Crystal Structure Refinement – A Crystallographer's Guide to SHELXL. Oxford University Press. Google Scholar
Neis, C., Petry, D., Demangeon, A., Morgenstern, B., Kuppert, D., Huppert, J., Stucky, S. & Hegetschweiler, K. (2010). Inorg. Chem. 49, 10092–10107. Web of Science CSD CrossRef CAS PubMed Google Scholar
NIST (2019). Standard Reference Database, Number 69, edited by P. J. Linstrom & W. G. Mallard. National Institute of Standards and Technology, Gaithersburg MD, 20899, https://doi.org/10.18434/T4D303 (retrieved February 27, 2019). Google Scholar
Parsons, S., Flack, H. D. & Wagner, T. (2013). Acta Cryst. B69, 249–259. Web of Science CrossRef CAS IUCr Journals Google Scholar
Romba, J., Kuppert, D., Morgenstern, B., Neis, C., Steinhauser, S., Weyhermüller, T. & Hegetschweiler, K. (2006). Eur. J. Inorg. Chem. pp. 314–328. Web of Science CSD CrossRef Google Scholar
Růžička, A., Padělková, Z., Švec, P., Pejchal, V., Česlová, L. & Holeček, J. (2013). J. Organomet. Chem. 732, 47–57. Google Scholar
Saari, W. S., Raab, A. W. & King, S. W. (1971). J. Org. Chem. 36, 1711–1714. CrossRef CAS Google Scholar
Sheldrick, G. M. (2015a). Acta Cryst. A71, 3–8. Web of Science CrossRef IUCr Journals Google Scholar
Sheldrick, G. M. (2015b). Acta Cryst. C71, 3–8. Web of Science CrossRef IUCr Journals Google Scholar
Shubnikov, A. V. & Koptsik, V. A. (1974). In Symmetry in Science and Art. New York: Plenum Press. Google Scholar
Spek, A. L. (2009). Acta Cryst. D65, 148–155. Web of Science CrossRef CAS IUCr Journals Google Scholar
Surendra Dilip, C. & Gowri, S. (2014). CSD Communication (refcode MOMVIV). CCDC, Cambridge, England. Google Scholar
Wang, H. M. J., Chen, C. Y. L. & Lin, I. J. B. (1999). Organometallics, 18, 1216–1223. Web of Science CSD CrossRef CAS Google Scholar
Zhang, F. L. & Shen, L. (2011). CSD Communication (deposition number 737932). CCDC, Cambridge, England. Google Scholar
Zhao, W., Wang, Y. & Wang, A. (2017). J. Mater. Sci. Chem. Eng. 5, 1–11. CAS 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.