fac-{2-[Bis(2-aminoethyl)amino]ethanaminium}trichloridorhodium(III) chloride hemihydrate

The crystal structure of the title compound, [Rh(C6H19N4)Cl3]Cl·0.5H2O, is isotypic with the previously reported Ru analogue. The structure contains two crystallographically independent [Rh(Htren)Cl3]+ cations with a facial tridentate coordination of the monoprotonated tren ligand [tren = tris(2-aminoethyl)amine], leading to an overall distorted octahedral coordination environment around the Rh(III) atom. In one of the two cations, the ethylene groups of the two chelate rings as well as the non-coordinating ethylammonium group are disordered over two sets of sites [0.579 (3):0.421 (3) occupancy ratio]. A series of N—H⋯Cl and O—H⋯Cl hydrogen bonds stabilizes the structure.

The crystal structure of the title compound, [Rh(C 6 H 19 N 4 )Cl 3 ]ClÁ0.5H 2 O, is isotypic with the previously reported Ru analogue. The structure contains two crystallographically independent [Rh(Htren)Cl 3 ] + cations with a facial tridentate coordination of the monoprotonated tren ligand [tren = tris(2-aminoethyl)amine], leading to an overall distorted octahedral coordination environment around the Rh(III) atom. In one of the two cations, the ethylene groups of the two chelate rings as well as the non-coordinating ethylammonium group are disordered over two sets of sites

Related literature
The preparation of the title compound has been described by Hyvä rinen et al. (2009) and the crystal structure of the isotypic Ru III complex has been reported by Sakai et al. (1996). Disorder phenomena, caused by a superposition of differently folded chelate rings of the tren ligand have been observed by Dü pre et al. (1999). Hypodentate coordination of polyamine ligands has been discussed by Blackman (2005) and Neis et al. (2010). For disorder phenomena, see: Hirshfeld (1976).
We thank Dr Volker Huch (Universitä t des Saarlandes) for the data collection.

Comment
The title compound has recently been obtained as a byproduct in the synthesis of [RhCl 2 (tren)] + . Based on the slightly longer wave lengths of the d-d transitions, a partial coordination of the tren ligand was assigned to [RhCl 3 (Htren)] + . Considering a step by step binding of the nitrogen donors to a mononuclear aqua-chlorido-Rh III precursor, such an intermediate with three coordinated amino groups could either have a meridional or a facial geometry. The two forms cannot be distinguished by NMR spectroscopy, because both diastereomers exhibit C s symmetry for the averaged solution structure.
The crystal structure analysis confirmed tridentate binding for the Htren + ligand and exhibited a facial geometry for the coordinated diethylenetriamine unit (Fig. 1). Partial metal binding ("hypodentate coordination") of polyamine ligands is well known. Hypodentate coordination can either be caused by the specific steric requirements of the ligand or by slow ligand substitution. The observation, that vigorous conditions in the synthetic procedure resulted in an exclusive formation of [RhCl 2 (tren)] + indicates that the incompletely coordinated ligand of the title compound is due to kinetic rather than thermodynamic reasons. [RhCl 3 (Htren)] + should thus be regarded as an intermediate, trapped on its way to [RhCl 2 (tren)] + .
The structure of the title compound is isotypic with the previously reported Ru analogue. The asymmetric unit contains two crystallographically independent [RhCl 3 (Htren)] + cations, two crystallographically independent chloride anions and one water molecule. The entire structure is stabilized by a three dimensional network of hydrogen bonds (Table 2). Notably, one of the hydrogen atoms of the water molecule (H1WB) has no acceptor: its nearest neighbors are the hydrogen atoms of an ethylenediamine group. Two coordinated chloride ions (Cl2 and Cl3) already exhibit O···Cl separations of 3.675 and 3.784 Å. The molecular structure of one of the cations closely approaches C s symmetry with the two chelate rings having a λ and δ conformation. The second cation exhibited some disorder for the five membered Rh-N-C-C-N rings and the non coordinating ethylammonium group. This disorder could be resolved and has been refined as a superposition of two distinct conformers. Within one particular form, the same type of conformation (i.e. λ/λ or δ/δ) was observed for the two chelate rings.
In comparison to the isotypic Ru complex, the M-N and M-Cl bonds of the title compound are, as expected, slightly shorter. In both congeners, the M-N bonds of the tertiary nitrogen atoms were slightly longer. In addition, a trans influence (push-pull mechanism) has found to be operative (Table 1). However, for the Ru congener, these effects were -if at all -barely significant. This is now different for the title compound, where the accuracy of the structure has been increased by almost one order of magnitude.

Refinement
In the second cation (Rh2), the ethylene groups of both chelate rings as well as the non-coordinating ethylammonium group are disordered, and were considered as a major and minor component with an occupancy of 57.9 (3) % and 42.1 (3) %, respectively. The partially occupied positions of all non hydrogen atoms (major component: C7, C9, C11, C13, C15, C17, N81; minor compounent: C8, C10, C12, C14, C16, C18, N82) could be refined anisotropically. C7 and C8, C11 and C12, N81 and N82 were each refined with equal displacement parameters. However, the disorder obviously generated some inequality of the displacement parameters for neighboring atoms such as N7 and C9 or C14 (Hirshfeld, 1976). H atoms bonded to the water O atom were located in an electron density map and refined with distance restraints of O-H = 0.84 (2) Å, and with U iso (H) = 1.2U eq (O). Other H atoms bonded to C-and N-atoms were positioned geometrically and refined using a riding model with free rotation about the C-NH 3 bond, with C-H = 0.99 Å, N-H = 0.91 (NH 3 groups) or 0.92 Å (NH 2 groups), and U iso (H) = 1.2 or 1.5 (NH 3 groups) of U eq of the pivot atom.  Refinement. Refinement of F 2 against ALL reflections. The weighted R-factor wR and goodness of fit S are based on F 2 , conventional R-factors R are based on F, with F set to zero for negative F 2 . The threshold expression of F 2 > σ(F 2 ) is used only for calculating Rfactors(gt) etc. and is not relevant to the choice of reflections for refinement. R-factors based on F 2 are statistically about twice as large as those based on F, and R-factors based on ALL data will be even larger.