An official website of the United States government
Here's how you know
Official websites use .gov
A
.gov website belongs to an official
government organization in the United States.
Secure .gov websites use HTTPS
A lock (
) or https:// means you've safely
connected to the .gov website. Share sensitive
information only on official, secure websites.
As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with,
the contents by NLM or the National Institutes of Health.
Learn more:
PMC Disclaimer
|
PMC Copyright Notice
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.
The structures of seven complexes of general formula LAuX3 (L = methylpyridines or dimethylpyridines, X = Cl or Br) are presented. In the crystal packing, a frequent feature is the offset-stacked and approximately rectangular dimeric moiety (Au—X)2, linked by Au⋯X contacts.
The structures of seven gold(III) halide derivatives of general formula LAuX3 (L = methylpyridines or dimethylpyridines, X = Cl or Br) are presented: trichlorido(2-methylpyridine)gold(III), [AuCl3(C6H7N)], 1 (as two polymorphs 1a and 1b); tribromido(2-methylpyridine)gold(III), [AuBr3(C6H7N)], 2; tribromido(3-methylpyridine)gold(III), [AuBr3(C6H7N)], 3; tribromido(2,4-dimethylpyridine)gold(III), [AuBr3(C7H9N)], 4; trichlorido(3,5-dimethylpyridine)gold(III), [AuCl3(C7H9N)], 5; tribromido(3,5-dimethylpyridine)gold(III), [AuBr3(C7H9N)], 6, and trichlorido(2,6-dimethylpyridine)gold(III), [AuCl3(C7H9N)], 7. Additionally, the structure of 8, the 1:1 adduct of 2 and 6, [AuBr3(C6H7N)]·[AuBr3(C7H9N)], is included. All the structures crystallize solvent-free, and all have Z′ = 1 except for 5 and 7, which display crystallographic twofold rotation symmetry, and 4, which has Z′ = 2. 1a and 2 are isotypic. The coordination geometry at the gold(III) atoms is, as expected, square-planar. Four of the crystals (1a, 1b, 2 and 8) were non-merohedral twins, and these structures were refined using the ‘HKLF 5’ method. The largest interplanar angles between the pyridine ring and the coordination plane are observed for those structures with a 2-methyl substituent of the pyridine ring. The Au—N bonds are consistently longer trans to Br (average 2.059 Å) than trans to Cl (average 2.036 Å). In the crystal packing, a frequent feature is the offset-stacked and approximately rectangular dimeric moiety (Au—X)2, with antiparallel Au—X bonds linked by Au⋯X contacts at the vacant positions axial to the coordination plane. The dimers are connected by further secondary interactions (Au⋯X or X⋯X contacts, ‘weak’ C—H⋯X hydrogen bonds) to form chain, double chain (‘ladder’) or layer structures, and in several cases linked again in the third dimension. Only 1b and 7 contain no offset dimers; these structures instead involve C—H⋯Cl hydrogen bonds combined with Cl⋯Cl contacts (1b) or Cl⋯π contacts (7). The packing patterns of seven further complexes LAuX3 involving simple pyridines (taken from the Cambridge Structural Database) are compared with those of 1–8.
1. Chemical context
In the previous part (Döring & Jones, 2024a ▸) of our series of publications ‘Gold complexes with amine ligands’, we reported the structures of four gold(I) halide complexes involving methylpyridine (picoline) and dimethylpyridine (lutidine) ligands. That publication presents much introductory material that we do not repeat here. For convenience, we have interpreted the term ‘amine’ liberally to include aza-aromatics.
In this publication we describe the structures of seven gold(III) halide derivatives of general formula LAuX3 (L = methylpyridines or dimethylpyridines, X = Cl or Br). These are: trichlorido(2-methylpyridine)gold(III) 1 (as two polymorphs 1a and 1b); tribromido(2-methylpyridine)gold(III) 2; tribromido(3-methylpyridine)gold(III) 3; tribromido(2,4-dimethylpyridine)gold(III) 4; trichlorido(3,5-dimethylpyridine)gold(III) 5; tribromido(3,5-dimethylpyridine)gold(III) 6 and trichlorido(2,6-dimethylpyridine)gold(III) 7. Additionally, we present the structure of 8, the 1:1 adduct of 2 and 6.
Note added during revision: A referee commented that 8 might be referred to as a co-crystal rather than an adduct. This is certainly a reasonable suggestion in view of the IUCr definition of a co-crystal (https://dictionary.iucr.org/Co-crystal): ‘Solid consisting of a crystalline single-phase material composed of two or more different molecular and/or ionic compounds, generally in a stoichiometric ratio, which are neither solvates nor simple salts.’ The problem in our view is that a solid is not necessarily the same as a crystal. We would therefore prefer to say that we studied a co-crystal of the adduct 8. The IUCr dictionary is an extremely useful document, but it is often difficult to provide watertight definitions of any given concept. For example, Bombicz (2024 ▸) recently offered reasoned criticism of the IUCr definition of ‘isostructural/isotypic’, and we supported her views in our previous paper (Döring & Jones, 2024a ▸).
The structure of the parent compound (py)AuCl3, Cambridge Structural Database (CSD; Groom et al., 2016 ▸) refcode PYAUCL10, was presented by Adams & Strähle (1982 ▸) (‘py’ = ‘pyridine’ throughout this paper). Two other compounds with the composition (py)AuX3 were in fact adducts of the type {[(py)2AuX2]+[AuX4]−·[(py)AuX3]} [X = Cl, KILFIV; Bourosh et al. (2007 ▸); X = Br, WOQMEU; Peters et al., 2000 ▸)]. The only related alkylpyridine structure is that of (4-Et-py)AuCl3 (ESITIM; Hobbollahi et al., 2019 ▸). Other derivatives involving ‘simple’ substituted pyridines as ligands are the isotypic pair (4-CN-py)AuX3 (X = Cl, WIRGAH or Br, WIRFUA; Mohammad-Natij et al., 2013 ▸) and several complexes (3-X1-py)AuX23 (X1 = halogen, X2 = Cl or Br) (Pizzi et al., 2022 ▸). The structures of this latter series and of (4-Et-py)AuCl3 are discussed in the section Database survey.
2. Structural commentary
All the structures crystallize solvent-free; Z′ values are 0.5 for 5 and 7, which display crystallographic twofold symmetry (with atoms N11, C14, Au1 and Cl1 on the twofold rotation axes 0.5, y, 0.75 and 0.5, y, 0.25, respectively), 2 for 4 and 1 for all other structures. Structures 1a and 2 are isotypic, but 5 and 6, which also differ only in the halogen, are not. Figs. 1 ▸–9 ▸ ▸ ▸ ▸ ▸ ▸ ▸ ▸ show the molecules of these compounds in the crystal, with ellipsoids drawn at the 50% probability level. Selected bond lengths and angles are given in Tables 1 ▸–9 ▸ ▸ ▸ ▸ ▸ ▸ ▸ ▸. The molecules are numbered such that atoms X1 (and X4, where two independent molecules are present) are trans to the pyridinic nitrogen atoms. The numbering of X2/X3, cis to the pyridinic nitrogen, is chosen to make X2—Au1—N11—C12 the smallest absolute torsion angle (with appropriately altered numbering for structures with two residues). This does not apply to 5 and 7, for which the cis sites are symmetry-related. The ring numbering of 6 (C12 to C16), otherwise ambiguous, is assigned by the same criterion.
Table 1. Selected geometric parameters (Å, °) for 1a.
The pyridine rings are as expected planar, with r.m.s. deviations of the six ring atoms between 0.002 and 0.01 Å. The coordination geometry at the central gold(III) atoms is, also as expected, square-planar; the r.m.s. deviations from the plane of Au, N and the three X atoms range from zero for 5 and 7 (by symmetry) to 0.058 Å for 3, whereby the donor atoms alternate above and below the plane by ca 0.06 Å; a similar alternation is observed for the dimethylpyridine component of the adduct 8, whereas the same molecule alone (structure 6) has a much lower r.m.s. deviation of 0.012 Å. The angles between these two planes are 78.4 (1)° for 1a, 84.7 (2)° for 1b, 78.7 (2)° for 2, 57.2 (1)° for 3, 84.5 (1)° and 74.8 (1)° for the two molecules of 4, 51.0 (1)° for 5, 56.0 (1)° for 6, 83.4 (1)° for 7 and 58.2 (2) and 84.3 (2)° for the two components of the adduct 8, corresponding to compounds 6 and 2. The largest angles are thus observed for those structures with a 2-methyl substituent of the pyridine ring, and presumably serve to reduce steric stress between these substituents and the X atoms cis to the nitrogen donor atom at Au. The gold atoms lie up to 0.15 (1) Å (for 1b) out of the pyridine plane, but lie exactly in this plane (by symmetry) for 5 and 7.
Bond lengths and angles may be regarded as normal. The Au—N bonds are consistently longer trans to Br [average (av.) of seven bonds: 2.059 Å] than trans to Cl (av. of four bonds: 2.036 Å), reflecting a greater trans influence of the bromido ligand compared to chlorido. There is no clear difference between Au—Cl bond lengths trans to N compared with those cis to N, whereas Au—Br bonds trans to N (av. of seven bonds: 2.395 Å) are significantly shorter than those cis to N (av. of fourteen bonds: 2.421 Å). The bond angles at Au are close to the ideal 90°/180°; the angles with the largest deviations for the former are 88.25 (10)° for N1—Au1—Br2 of 3 and 91.17 (5)° for Cl3—Au1—Cl1 of 1b, and for the latter 176.590 (17)° for Br6—Au2—Br5 of 4. The C—N—C angles of the py ligands are all close to 120° (av. of eleven angles: 120.8°).
A least-squares fit of the polymorphs 1a and 1b gave an r.m.s. deviation of 0.08 Å; a similar fit of the two independent molecules of 4 (one inverted) gave a deviation of 0.16 Å. Fits of molecules 2 and 6 (the latter inverted) to the same molecules of the adduct 8 gave r.m.s. deviations of 0.091 and 0.061 Å, respectively. More informative figures are however obtained by fitting only the pyridine ligands, which are closely similar; the differences associated with the AuX3 moieties are then shown more clearly. For 1a/1b, the atoms Cl2 and Cl3 differ in position by 0.26 and 0.20 Å respectively (Fig. 10 ▸). For 4, the gold atoms lie on opposite sides of the pyridine plane, and this, coupled with the 10° difference in the interplanar angle, leads to significant differences in the positions of the bromine atoms (0.39, 0.50, 0.51 Å, respectively for Br1–3; Fig. 11 ▸). A similar effect, although the interplanar angles are almost equal, is seen for the fit of 1b with its counterpart in the adduct 8 (deviations 0.50, 0.43, 0.40 Å; Fig. 12 ▸), whereas the largest difference for the fit of 6 with its counterpart in 8 is for Br1 (0.22 Å; Fig. 13 ▸).
3. Supramolecular features
Hydrogen bonds of the type C—H⋯X for all structures are given in Tables 10 ▸–18 ▸ ▸ ▸ ▸ ▸ ▸ ▸ ▸. These include several borderline cases that are not discussed explicitly. For all packing diagrams, the labelling indicates the asymmetric unit, and hydrogen atoms not involved in secondary contacts are omitted for clarity. The choice of ‘important’ interactions and their hierarchy is necessarily subjective, at least to some extent; diagrams with a small number of heavy-atom contacts are easier to interpret than those involving a larger number of hydrogen bonds, and this is especially true for H⋯Br contacts, which are probably weaker than H⋯Cl. Primes (′,′′) indicate symmetry-equivalent atoms; operators are not given in full each time. A summary of the packing features is given in Table 19 ▸.
Notes: (a) Refcodes refer to structures whose packing is discussed in the section Database survey; (b) these columns do not necessarily present an exhaustive list; see text for further details.
Before discussing the packing of 1–8 in detail, it is useful to look back on the packing of (py)AuCl3 (Adams & Strähle, 1982 ▸; space group C2/c, Z = 8), to see what types of secondary interaction can arise. Short non-bonded contacts were observed between the gold atom and two chlorine atoms, positioned axially to the main coordination plane in such a way as to complete a highly stretched octahedron at the gold atom (Au⋯Cl 3.636 and 3.648 Å, Cl⋯Au⋯Cl 173.0°; operators − x, − + y, 1 − z and − x, + y, 1 − z ). This leads to ladder-like double chains of residues (Fig. 14 ▸), parallel to the b axis, in which the molecules display offset stacking of the AuX3 groups; one Au—Cl bond of each molecule (the rungs of the ladder) shares two Au⋯Cl contacts with antiparallel Cl—Au bonds of each neighbouring molecule (the side rails of the ladder). The (Au—X)2 quadrilaterals, approximately rectangular and with side lengths corresponding to the Au—Cl bond length and the Au⋯Cl contact distance, are a recurring feature in the structures discussed here. Offset stacking of this type is a common feature in AuX3 complexes, and we have observed it e.g. in four modifications of (tetrahydrothiophene)AuCl3 (Upmann et al., 2017 ▸). In general, any suitable donor atoms can occupy these two contact sites. It might be argued that such contacts are merely connected with the steric ease of approach to the two sides of the coordination plane; this has also been argued for short contacts to the linearly coordinated gold atom of gold(I) complexes, although H⋯Au hydrogen bonding in such systems is reasonably well established (Schmidbaur, 2019 ▸; Schmidbaur et al., 2014 ▸). However, recent studies and calculations (Daolio et al., 2021 ▸; Pizzi et al., 2022 ▸) have indicated that there is a π-hole at the gold atom, and that there is thus a definite attractive interaction, a ‘coinage bond’, between the gold atom and the additional donor(s) (see below).
At the time of publication of the (py)AuCl3 structure, more than 40 years ago (the data were probably recorded in the late 1970s), the main interest in crystal structure determinations generally centred on the molecule being studied, whereas intermolecular contacts were often neglected. The analysis of the Au⋯Cl contacts in (py)AuCl3 constituted a welcome exception. However, the structure contains other secondary contacts that were not mentioned, probably because at the time such contacts were not regarded as significant. First, there is a short Cl⋯Cl contact of 3.462 Å connecting the ‘ladders’. Such formally non-bonding contacts between halogen atoms have been the subject of considerable interest for some time and are usually termed ‘halogen bonds’. For C—X⋯X—C systems, they are considered to involve a small region of positive charge in the extension of the C—X bond vectors beyond the atom X, often leading to one C—X⋯X angle of ca 90° and one of ca 180° (see e.g. Metrangolo et al., 2008 ▸, or Cavallo et al., 2016 ▸, for review articles); they are quite common for Au—X systems, but we are not aware of any systematic and/or theoretical study of X⋯X contacts in these systems. We have drawn attention to X⋯X contacts in various tetrahalogenidoaurate(III) salts (e.g. Döring & Jones, 2016 ▸) and in LAuX3 complexes (e.g. Döring & Jones, 2024b ▸), and recently presented a short database survey of the latter (Döring & Jones, 2023 ▸). Secondly, there are short contacts of the type C—H⋯Cl that are now regarded as hydrogen bonds and are, somewhat misleadingly, often termed ‘weak’ hydrogen bonds. These were not mentioned in the 1982 publication, and indeed no hydrogen atoms were included in the refinement, which was not unusual at the time for heavy-atom structures. We used the program XP (Bruker, 1998 ▸) to calculate the hydrogen-atom positions, and established that there are three short H⋯Cl contacts, one as short as 2.79 Å; this connects neighbouring molecules in the ladders. Two further such contacts (2.85 and 2.86 Å) connect the ladders; these are omitted from Fig. 14 ▸ for clarity.
The packing diagram of compound 1a is shown in Fig. 15 ▸. The molecules are linked to form inversion-symmetric dimers (operator 1 − x, 1 − y, 2 − z) in an offset packing pattern, with Au1⋯Cl2′ = 3.441 (2) Å; reinforcement is provided by the shortest hydrogen bond H16⋯Cl2′. The dimers are in turn linked by a short contact Cl2⋯Cl3(1 + x, y, z) = 3.239 (2) Å to form double chains parallel to the a axis. The angles Au1—Cl2⋯Cl3′ and Au1—Cl3⋯Cl2′ are 161.09 (6) and 162.30 (7)°, respectively. There are no other Cl⋯Cl contacts < 3.8 Å. In the packing of compound 2 (isotypic to 1a) the corresponding dimensions are Au1⋯Br2′ = 3.5654 (8), Br2⋯Br3 = 3.3840 (9) Å, Au1—Br2⋯Br3′ = 164.38 (3), Au1—Br3⋯Br2′ = 159.34 (3)°. The second polymorph 1b has no Au⋯Cl contact shorter than 3.833 (2) Å for Au1⋯Cl2(−x, 1 − y, −z), but has an even shorter Cl⋯Cl contact: Cl2⋯Cl3(x, y, −1 + z) = 3.164 (2) Å. This combines with the shortest H⋯Cl hydrogen bond, again H16⋯Cl2, to form double chains of molecules parallel to the a axis (Fig. 16 ▸), with Au1—Cl2⋯Cl3′ = 172.65 (7)° and Au1—Cl3⋯Cl2′ = 174.55 (8)°. The double chains are linked in the b direction by the hydrogen bond H13⋯Cl1. It is notable throughout this series of structures that the cis (to the pyridine ligands) halogen atoms X2 and X3 (or X5 and X6) tend to be involved in the main packing features, whereas the trans halogen atom X1 (or X4) often provides the additional linkages.
The packing of compound 3 consists of layers parallel to the bc plane (Fig. 17 ▸) at x ≃ 0.25 and 0.75, in which the two contacts Au1⋯Br2( − x, − + y, − z) = 3.4688 (5) Å and Au1⋯Br3( − x, − y, 1 − z) = 3.6535 (5) Å complete a stretched octahedron at the gold atom [angles Br2′⋯Au1⋯Br3′ = 173.91 (1)°, Au1—Br2⋯Au1′ = 148.42 (2)°]. The Au1⋯Br3 contacts generate an offset-stacked dimer, but this stacking is not further extended. The contact Br3⋯Br3( − x, − − y, 1 − z) = 3.5256 (9) Å completes the layer, with Au1—Br3⋯Br3′ = 170.31 (3)°. The two shortest H⋯Br contacts also lie within this layer, but are omitted from Fig. 17 ▸ for clarity. The layers are linked by the Br1⋯Br1(−x, y, + z) contact of 3.4531 (9) Å, with Au1—Br1⋯Br1′ = 174.62 (3) Å, and by stacking of the pyridine rings [intercentroid distances 3.657 (2) and 3.619 (2) Å, slippage 1.28 and 1.38 Å, operators 1 − x, −y, 1 − z and 1 − x, 1 − y, 1 − z respectively] (Fig. 18 ▸). No other structure presented here has an intercentroid distance between the rings < 3.70 Å.
The packing of compound 4 is closely related to that of 1a. Each independent molecule forms double chains parallel to the a axis that are topologically analogous to those of 1a, with dimers arising from antiparallel (Au—Br)2 contacts and further linked by Br⋯Br contacts; dimensions (Å and °) are Au1⋯Br2(1 − x, 1 − y, −z) = 3.6606 (5), Br2⋯Br3(1 + x, y, z) = 3.3117 (6), Au—Br2⋯Br3′ = 164.67 (2), Au1—Br3⋯Br2′ = 165.45 (2) for the first molecule (Fig. 19 ▸) and Au2⋯Br5(−x, 1 − y, 1 − z) = 3.8328 (5), Br5⋯Br6(−1 + x, y, z) = 3.5191 (6), Au2—Br5⋯Br6′ = 147.93 (2), Au2—Br6⋯Br5′ = 151.00 (2) for the second molecule (Fig. 20 ▸). It is notable that the contact distances are shorter and the angles more linear for molecule 1 than for molecule 2, for reasons that are not apparent. The main difference from 1a is that the two chains are further linked in 4 by the contacts Br1⋯Br6(x, − y, − + z) = 3.5977 (6) and Br3⋯Br4 = 3.5591 (6) Å (Fig. 21 ▸).
The packing of compound 5 resembles that of (py)AuCl in that the molecules are assembled into chains linked by offset stacking, with Au1⋯Cl2(1 − x, 1 − y, 1 − z and x, 1 − y, + z) = 3.6401 (7) Å, Cl2′⋯Au1⋯Cl2′′ = 163.15 (2)°; the chains run parallel to the a axis. However, the Au2Cl2 quadrilaterals are not edge-linked as in (py)AuCl, but apex-linked. Similar chains were observed in the isotypic pair (4-CN-py)AuX3 (X = Cl or Br; Mohammad-Natij et al., 2013 ▸). Adjacent chains are linked by the contact Cl2⋯Cl2(2 − x, y, − z) = 3.5501 (13) Å to form layers parallel to the ac plane (Fig. 22 ▸). The hydrogen bond H2⋯Cl1, 2.67 Å, is not included in Fig. 22 ▸ because of the view direction, in which the rings, lying in or close to the planes at x ≃ 0, 0.5, 1 etc., are seen edge-on. Layers are connected in the b direction by the three-centre hydrogen bond from H14 to two Cl2 atoms (Fig. 23 ▸).
The packing of compound 6 involves double chains, parallel to the a axis (Fig. 24 ▸) that are topologically the same as those of 1a and 4. The usual dimers are formed, although the contact distance is rather long: Au1⋯Br3(−x, 1 − y, 1 − z) = 3.7738 (6) Å. The dimers are connected by the contacts Br2⋯Br3(1 + x, y, z) = 3.4644 (6) Å, with angles Au1—Br2⋯Br3′ = 165.93 (2)° and Au1—Br3⋯Br2′ = 166.15 (2)°. The double chains are connected by the contacts Br1⋯Br1(−x, −y, −z) = 3.4284 (9) Å, with Au1—Br1⋯Br1′ = 156.36 (3)°, to form a double layer parallel to (01) (Fig. 25 ▸).
The packing of compound 7 is unexpected; it involves neither Au⋯Cl nor Cl⋯Cl interactions. Instead, the two important contacts are the short hydrogen bond H14⋯Cl1′(x, 1 + y, z) and a Cl⋯π contact from Cl2 to the centroid (Cg) of the pyridine ring at ( − x, − y, 1 − z); the contact distance Cl2⋯Cg′ is 3.5458 (5) Å, with angles Au1—Cl2⋯Cg′ = 162.6° and Cg′⋯Cl2⋯Cg" = 171.4°. The Cl⋯π interactions propagate parallel to [101], so that the result is a layer structure parallel to (10) (Fig. 26 ▸). This type of interaction can be regarded as a halogen bond from the chlorine atom to the π electron cloud of the pyridine ligand.
The main feature of the packing of adduct 8 (composed of 2 and 6) is a layer structure (Fig. 27 ▸) parallel to (110). Chains of alternating molecules of 2 and 6, horizontal in Fig. 27 ▸, run parallel to [1]; they are propagated by the contacts Br2⋯Br5(1 − x, −y, −z) = 3.2915 (11) Å and Br3⋯Br6(−x, 1 − y, 1 − z) = 3.5493 (13) Å, with Au1—Br2⋯Br5′ = 161.86 (5), Au2—Br5⋯ Br2′ = 169.76 (5), Au1—Br3⋯Br6′ = 162.49 (5) and Au2—Br6⋯Br3′ = 151.56 (5)°. As in the structure of 2 alone, there are no axial contacts to the gold atom Au2 [discounting Au2⋯Br2 3.9093 (11) Å as too long]. The gold atom of molecule 6 has two axial contacts, Au1⋯Br2(1 − x, −y, 1 − z) = 3.5279 (10) and Au1⋯Br6 = 3.5169 (10) Å, with Br2′⋯Au1⋯Br6 = 169.74 (2)°, in contrast to its single axial contact in the structure of 6 alone. The former contact is part of an offset-stacked dimer (see the small quadrilaterals in Fig. 27 ▸), but these quadrilaterals do not associate directly to form more extensive elements of the packing. The linkages between layers are provided by the contacts Br1⋯Br1(−x, −y, 1 − z) = 3.362 (2) and Br4⋯Br4(1 − x, 1 − y, −z) = 3.343 (2) Å, with Au1—Br1⋯Br1′ = 175.87 (7) and Au2—Br4⋯Br4′ = 153.75 (6)° (Fig. 28 ▸), in a manner reminiscent of the interlayer links in 3 and those within the double layers of 6.
4. Database survey
The searches employed the routine ConQuest (Bruno et al., 2002 ▸), part of Version 2024.1.0 of the CSD (Groom et al., 2016 ▸). A search for ‘simple’ compounds of the form LAuCl3 (L = pyridine ligand with no substituents involved in further rings, X = halogen) gave 21 hits. The Au—N bond lengths were 2.015–2.073, av. 2.043 (13) Å, the Au—Cl bond lengths trans to N were 2.255–2.273, av. 2.263 (3) Å, and the Au—Cl bond lengths cis to Au—N were 2.221–2.29, av. 2.275 (11) Å. No clear trans influences can be recognised in these values. The three hits for X = Br were the (py)AuBr3 component of {[(py)2AuBr2]+[AuBr4]−·[(py)AuBr3]} (WOQMEU, Peters et al., 2000 ▸); (4-CN-py)AuBr3 (WIRFUA, Mohammad-Natij et al., 2013 ▸); and (3-F-py)AuBr3 (WEFRAE, Pizzi et al., 2022 ▸). All showed Au—Brtrans bonds significantly shorter than Au—Brcis, by ca 0.02–0.03 Å, but the Au—N bond lengths were variable at 2.040–2.098 Å. The sample is probably too small to draw reliable conclusions.
It is instructive to take six of the simplest compounds thus found and briefly compare their packing features with those of 1–8. The compounds chosen are: L = 4-ethylpyridine, X = Cl (ESITIM, Hobbollahi et al., 2019 ▸); L = 3-bromopyridine, X = Cl (WEFQAD); L = 3-fluoropyridine, X = Cl (WEFQEH); L = 3-chloropyridine, X = Cl (WEFQIL); L = 3-iodopyridine, X = Cl (WEFQOR); and L = 3-fluoropyridine, X = Br (WEFRAE; all from Pizzi et al., 2022 ▸). In all cases, the authors drew attention to the short Au⋯X contacts. These compounds are included in Table 19 ▸. C—H⋯X hydrogen bonding is neglected.
ESITIM crystallizes in Pcab with Z = 8. In the original publication, the Au⋯Cl contacts (3.244, 3.409 Å) were described as linking the molecules to form infinite chains. In fact, they combine to form a layer structure, involving Au4Cl4 rings, parallel to the ab plane at z ≃ 0.25, 0.75 (Fig. 29 ▸). In the series of 3-halopyridine complexes, the halogen substituents of the pyridine rings are ‘non-innocent’ atoms as regards to intermolecular interactions. In WEFQAD (P, Z = 2), the Au⋯Cl contacts (3.492, 3.579 Å) combine to form a ‘ladder’ structure parallel to the a axis. Two short Br⋯Cl contacts to the trans chlorine atom (3.490, 3.690 Å) are observed, which link the layers (Fig. 30 ▸). In the corresponding 3-fluoro derivative WEFQEH (P21/n, Z = 4), the Au⋯Cl contacts (3.373, 3.426 Å) combine to form a layer structure, involving Au4Cl4 rings, parallel to the ab plane at z ≃ 0.25, 0.75 (Fig. 31 ▸; an equivalent diagram was presented by Pizzi et al. (2022 ▸) but we include this Figure for completeness and for consistency of format). A short F⋯F contact of 2.684 Å links the layers. In the 3-chloro derivative WEFQIL (P21/n, Z = 4), the Au⋯Cl contacts (3.402, 3.412 Å) combine to form a chain of apex-linked quadrilaterals (analogous to the chains in 5) parallel to the a axis; these are liked by Cl⋯Clpy contacts of 3.536 Å to form layers parallel to the ac plane at y ≃ 0.25, 0.75 (Fig. 32 ▸), and the layers are connected in the third dimension by another Cl⋯Clpy contact of 3.495 Å. In the 3-iodo derivative WEFQOR (C2/c, Z = 8), the Au⋯Cl contacts (3.368, 3.483 Å) combine to form a ‘ladder’ structure parallel to the b axis; ladders are linked by I⋯Cl contacts of 3.500 Å to form layers parallel to the ab plane at z ≃ 0.25, 0.75 (Fig. 33 ▸). The layers are linked in the third dimension by a Cltrans⋯Cltrans contact of 3.433 Å. In WEFRAE (P212121, Z = 4), the tribromido analogue of WEFQEH, the Au⋯Br contacts (3.542, 3.588 Å) combine to form a chain of apex-linked quadrilaterals parallel to the a axis; these are linked directly by quite long Br⋯Br contacts of 3.710 Å to form a layer structure parallel to the ab plane at z ≃ 0.25, 0.75 (Fig. 34 ▸). Layers are linked in the third dimension by Brtrans⋯Brtrans contacts of 3.712 Å. The fluorine atom is not involved in short contacts.
A search for AuIII structures related to 7, containing Au—Cl and Au—Npyridinic bonds together with a short Cl⋯π contact (defined by the distance from Cl to the pyridine ring centroid Cg) gave thirteen hits with Cl⋯Cg < 3.7 Å. The shortest distance is 3.344 Å in trichlorido-(1,7,15,15-tetramethyl-3,10-diazatetracyclo[10.2.1.02,11.04,9]pentadeca-2,4,6,8,10-pentaene)gold(III), a camphorquinoxaline complex (SUYXAN; Glišić et al., 2018 ▸).
5. Synthesis and crystallization
Trichlorido(2-methylpyridine)gold(III) (1): 114.2 mg (0.351 mmol) of the gold(I) precursor chlorido(2-methylpyridine)gold(I) was prepared by the method of Ahrens (1999 ▸). This was dissolved in 5 ml of dichloromethane, and the solution was added to a solution of 100 mg (0.363 mmol) of PhICl2 in 5 ml of dichloromethane. Equal (0.4 ml) portions of the solution were transferred to five ignition tubes and overlayered with the five precipitants n-pentane, n-heptane, diethyl ether, diisopropyl ether and petroleum ether (b.p. 313–333 K). The tubes were stoppered and transferred to the refrigerator overnight. Crystals of compound 1, polymorph a, were obtained as yellow prisms and tablets from the tube with diisopropyl ether. In general for these syntheses, crystals also formed in at least some of the other tubes, but the best, judged by inspection under a microscope, were selected for X-ray measurements. Elemental analysis [%]: calc.: C 18.18, H 1.78, N 3.53; found C 17.78, H 1.79, N 3.59. Because of the problem of incomplete oxidation that we have sometimes encountered using PhICl2, the procedure was repeated in parallel using two equivalents of PhICl2, although this precaution later proved to have been unnecessary for the reactions presented here. The same crystallization experiments were carried out. Crystals of compound 1, polymorph b, were obtained as yellow plates from the tube with n-pentane. Elemental analysis [%]: calc.: C 18.18, H 1.78, N 3.53; found: C 17.63, H 1.78, N 3.58.
Tribromido(2-methylpyridine)gold(III) (2): 90 mg (0.247 mmol) of (tht)AuBr (tht = tetrahydrothiophene) were converted to bis(2-methylpyridine)gold(I) dibromidoaurate(I) (Döring & Jones, 2024a ▸), which was immediately (without drying) dissolved in 2 ml of dichloromethane, and two drops of elemental bromine were added. The usual crystallization experiments were carried out. Crystals of compound 2 were obtained in the form of red blocks and tablets from the tube with n-pentane. Elemental analysis [%]: calc: C 13.60, H 1.33, N 2.64; found: C 12.60, H 1.37, N 2.62.
Tribromido(3-methylpyridine)gold(III) (3): 90 mg (0.247 mmol) of (tht)AuBr were converted to bis(3-methylpyridine)gold(I) dibromidoaurate(I) (Döring & Jones, 2024a ▸), which was immediately (without drying) dissolved in 2 ml of dichloromethane. Two drops of elemental bromine were added. The usual crystallization experiments were carried out. Crystals of compound 3 were obtained in the form of red plates from the tube with n-pentane. Elemental analysis [%]: calc.: C 13.60, H 1.33, N 2.64; found: C 13.47, H 1.35, N 2.78.
Tribromido(2,4-dimethylpyridine)gold(III) (4): 45,2 mg (0.124 mmol) of (tht)AuBr were dissolved in 2 ml of 2,4-dimethylpyridine. The solution was transferred to a 5 ml glass vial and overlayered with diisopropyl ether. The vial was closed and stored in the refrigerator. The supernatant was then pipetted off and the remaining colourless crystals, assumed to be bis(2,4-dimethylpyridine)gold(I) dibromidoaurate(I), dried in vacuo, yielded 32.5 mg (48%). The crystals proved to be unsuitable for structure determination because of streaking of the diffraction peaks. They were dissolved in 2 ml of dichloromethane and 3 drops of elemental bromine were added, leading to a red solution. This was overlayered with n-pentane and stored in the refrigerator for a week. Crystals of 4 were obtained in the form of red plates and needles. The elemental analysis gave an unsatisfactory value for C: [%] calc.: C 15.46, H 1.67, N 2.58; found: C 13.81, H 1.53, N 2.43.
Trichlorido(3,5-dimethylpyridine)gold(III) (5): 166 mg (0.518 mmol) of (tht)AuCl were converted to bis(3,5-dimethylpyridine)gold(I) dichloridoaurate(I) (Döring & Jones, 2024a ▸). The sample was divided in half; each half was dissolved in 5 ml of dichloromethane, and then treated with one or two equivalents of PhICl2. The solutions were subjected to the usual crystallization experiments. Crystals of 5 were obtained in the form of yellow blocks from all tubes; those chosen were from the 1:2 experiment using diethyl ether. Elemental analysis [%]: calc.: C 20.48, H 2.21, N 3.41; found: C 20.23, H 2.121, N 3.58.
Tribromido(3,5-dimethylpyridine)gold(III) (6): see (8) below.
Trichlorido(2,6-dimethylpyridine)gold(III) (7): 122.5 mg (0.382 mmol) of (tht)AuCl were converted to 119 mg (0.175 mmol) of (2,6-dimethylpyridine)gold(I) dichloroaurate(I) (Hashmi et al., 2010 ▸). This was divided into two portions, and 2 ml of dichloromethane were added to each. A solution of 24.1 mg (0.088 mmol) of PhICl2 in 2 ml of dichloromethane was added to one aliquot and a solution of 48.2 mg (0.175 mmol) of PhICl2 in 2 ml of dichloromethane to the other. These solutions were subjected to the usual crystallization experiments. Crystals of 7 in the form of yellow blocks were obtained from the 1:2 experiment using n-heptane. Elemental analysis [%]: calc.: C 20.48, H 2.21, N 3.41; found: C 20.71, H 2.20, N 3.53.
Tribromido(2-methylpyridine)gold(III)/tribromido(3,5-dimethylpyridine)gold(III) (1/1) (8): Crystals of compounds 8 and 6 arose serendipitously, partly as a result of human error, as follows. 137.3 mg (0.376 mmol) of (tht)AuBr were converted to 84.0 mg (0.114 mmol) of bis(2-methylpyridine)gold(I) dibromoaurate(I) as above, of which 75.1 mg (0.102 mmol) were dissolved in 5 ml of dichloromethane. Five drops of elemental bromine were added. Half of the resulting red solution was overlayered with n-pentane. At some stage, which can no longer be identified (but the 2-picoline was checked by NMR and was pure), the system became contaminated with 3,5-dimethylpyridine. One of the red crystals that formed was investigated and proved to be the 1/1 adduct 8. The 1H NMR spectrum of the sample showed the expected two methyl singlets, but in the ratio 4:1 rather than the expected 2:1 for a 1/1 mixture of 2 and 6; this would suggest that the sample of red crystals from which 8 was taken consisted of both 6 and 8. Consistent with this, the solution of the red crystals in CDCl3, left to stand for some time, deposited a few red crystals that proved on X-ray examination to be compound 6.
6. Refinement
Details of the measurements and refinements are given in Table 20 ▸. For all structures, multi-scan absorption corrections were applied using spherical harmonics, as implemented in the SCALE3 ABSPACK scaling algorithm (Rigaku OD, 2020 ▸). For compound 6, analytical numeric absorption corrections using a face-indexed crystal model, based on expressions derived by Clark & Reid (1995 ▸), were applied first.
Computer programs: CrysAlis PRO (Rigaku OD, 2020 ▸), SHELXS97 (Sheldrick, 2008 ▸), SHELXL2019/3 (Sheldrick, 2015 ▸) and XP (Bruker, 1998 ▸).
Aromatic hydrogen atoms were included at calculated positions and refined using a riding model with C—H = 0.95 Å. Methyl groups were included as idealized rigid groups with C—H = 0.98 Å and H—C—H = 109.5°, and were allowed to rotate but not tip (command ‘AFIX 137’). U values of the hydrogen atoms were fixed at 1.5 × Ueq of the parent carbon atoms for methyl groups and 1.2 × Ueq of the parent carbon atoms for other hydrogens. A small number of badly fitting reflections were omitted (1a, two reflections with deviations > 8σ; 1b, seven reflections > 7σ; 2, one reflection > 15σ; 5, one reflection > 6σ; 8, one reflection > 29σ).
Four of the crystals (1a, 1b, 2 and 8) were non-merohedral twins, with twinning by 180° rotation about the a axis for 1a, 1b and 2 and about [11] for 8. These structures were refined using the ‘HKLF 5’ method (Sheldrick, 2015 ▸). The relative volumes of the smaller twinning components refined to 0.4710 (6), 0.4583 (6), 0.4641 (5) and 0.4440 (5), respectively. The twin data reduction merges equivalent reflections before writing the intensity file, so that Rint is meaningless (and is not given in Table 20 ▸). The intensity datasets comprise all non-overlapped reflections from both components and all overlapped reflections, so that the number of reflections should be interpreted with caution. More stringent checks during the data reduction of twins (e.g. the command ‘remove outliers’) mean that the completeness of some datasets is less than ideal, typically around 95%.
Special features and exceptions: For 1b, the large difference peak of 4.5 e Å−3 has coordinates that are arithmetically related to those of the gold atom and thus may represent residual twinning errors. For 3, the x and y coordinates of the gold atom are ca 0.25, which leads to systematically weak reflection classes; checkCIF comments on (pseudo-) B-centring. The second weighting parameter b (Sheldrick, 2015 ▸) does not converge, but oscillates over a small range. For 6, the methyl hydrogen atoms at C18 were unclear, and were therefore refined as an ideal hexagon of half-occupied sites (command ‘AFIX 127’). However, the disorder may be more extensive than this simple model.
Tribromido(3,5-dimethylpyridine-κN)gold(III) (6) . Special details
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.
Bourosh, P., Bologa, O., Simonov, Y., Gerbeleu, N., Lipkowski, J. & Gdaniec, M. (2007). Inorg. Chim. Acta, 360, 3250–3254.
Bruker (1998). XP. Bruker Analytical X-Ray Instruments, Madison, Wisconsin, USA.
Bruno, I. J., Cole, J. C., Edgington, P. R., Kessler, M., Macrae, C. F., McCabe, P., Pearson, J. & Taylor, R. (2002). Acta Cryst. B58, 389–397. [DOI] [PubMed]
Cavallo, G., Metrangolo, P., Milani, R., Pilati, T., Priimagi, A., Resnati, G. & Terraneo, G. (2016). Chem. Rev.116, 2478–2601. [DOI] [PMC free article] [PubMed]
Clark, R. C. & Reid, J. S. (1995). Acta Cryst. A51, 887–897.
Daolio, D., Pizzi, A., Terraneo, G., Ursini, M., Frontera, A. & Resnati, G. (2021). Angew. Chem. Int. Ed.60, 14385–14389. [DOI] [PMC free article] [PubMed]
Döring, C. & Jones, P. G. (2016). Z. Anorg. Allg. Chem.642, 930–936.
Glišić, B. Đ., Hoffmann, M., Warżaitis, B., Genčić, M. S., Blagojević, P. D., Radulović, N. S., Rychiewska, U. & Duran, M. I. (2018). Polyhedron, 105, 137–149.
Groom, C. R., Bruno, I. J., Lightfoot, M. P. & Ward, S. C. (2016). Acta Cryst. B72, 171–179. [DOI] [PMC free article] [PubMed]
Hashmi, A. S. K., Lothschütz, C., Ackermann, M., Doepp, R., Anantharaman, S., Marchetti, B., Bertagnolli, H. & Rominger, F. (2010). Chem. Eur. J.16, 8012–8019. [DOI] [PubMed]
Hobbollahi, E., List, M. & Monkowius, U. (2019). Monatsh. Chem.150, 877–883.
Metrangolo, P., Meyer, F., Pilati, T., Resnati, G. & Terraneo, G. (2008). Angew. Chem. Int. Ed.47, 6114–6127. [DOI] [PubMed]
Mohammad-Nataj, R., Abedi, A. & Amani, V. (2013). Synth. React. Inorg. Met.-Org. Nano-Met. Chem.43, 1375–1380.
Peters, K., Peters, E.-M., von Schnering, H. G., Hönle, W., Schmidt, R. & Binder, H. (2000). Z. Kristallogr. New Cryst. Struct.215, 413–414.
Pizzi, A., Calabrese, M., Daolio, A., Ursini, M., Frontera, A. & Resnati, G. (2022). CrystEngComm, 24, 3846–3851.
Rigaku OD (2020). CrysAlis PRO. Rigaku Oxford Diffraction, Yarnton, England.
Schmidbaur, H. (2019). Angew. Chem. Int. Ed.58, 5806–5809.
Schmidbaur, H., Raubenheimer, H. G. & Dobrzańska, L. (2014). Chem. Soc. Rev.43, 345–380. [DOI] [PubMed]
Sheldrick, G. M. (2008). Acta Cryst. A64, 112–122. [DOI] [PubMed]
Sheldrick, G. M. (2015). Acta Cryst. C71, 3–8.
Upmann, D., Näther, C., Jess, J. & Jones, P. G. (2017). Z. Anorg. Allg. Chem.643, 311–316.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.