This journal is©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015 New J. Chem.
Cite this:DOI: 10.1039/c5nj00254k
Absolute asymmetric synthesis of five-coordinate complexes†
Anders Lennartson*a and Mikael Håkansson*b
In 2009, we reported the first example of the optical resolution of a five-coordinate complex. In this study, we discuss the chirality of five-coordinate enantiomers displaying mono- and bi-dentate ligands by optical resolution of three complexes. Through total spontaneous resolution, it was possible to obtain optically active bulk products starting from achiral precursors only. It was discovered that crystallisation of two of the complexes was influenced by undetectable amounts of an unidentified source of optical activity. 1. Introduction
Pasteur performed the first optical resolution of a four-coordinate tetrahedral compound in 1848 (although the true nature of the tetrahedral carbon atom was not realised until 1874). In 1911,
Werner reported the first optical resolution of a six-coordinate octahedral coordination compound,1 which was an achievement constituting the final proof of the coordination theory. Werner’s success was due to the fact that he worked with configurationally inert cobalt(III) complexes with very low racemisation rates.
Isolating enantiomers of complexes displaying coordination numbers other than four and six provedmore intriguing, since such complexes typically are configurationally labile. The solution to this problem turned out to be spontaneous resolution occasionally, racemic mixtures crystallise as conglomerates, i.e. as a 1 :1 mixture of the two enantiomorphic crystal forms; such compounds are said to undergo spontaneous resolution since the two enantiomers spontaneously separate into different crystals. It may occur when the substance crystallises in one of the so-called Sohncke space groups.2 If the substance is stereochemically labile, total spontaneous resolution or crystallisation-induced asymmetric transformation3 is possible. If the solution crystallises slowly, all crystals may grow from the first crystal nucleus formed, and all crystals will be identical and have the same absolute structure or handedness. This is possible since the solution of a labile solute will remain racemic throughout the crystallisation, and the crystallising enantiomer will therefore never be depleted.
Which enantiomer that actually crystallises will be random.
Using solid-state CD (circular dichroism) spectroscopy,4 it is possible to measure the optical purity in the solid state.5 It should be noted that preparing bulk samples of an optically active compound in a yield exceeding 50% from achiral or racemic precursors only constitute examples of absolute asymmetric synthesis.6,7 We have previously reported a number of examples of absolute asymmetric synthesis based on total spontaneous resolution.5,8–17
Of special interest are the five-coordinate enantiomers, the subject of this study, since these were for many years the missing link between the work of Pasteur and Werner. Five-coordinate enantiomers were first believed to have been isolated over a century ago when Pope and Peachey resolved allyl-benzyl-methyl-phenylammonium hydroxide.18 In those days, quaternary ammonium salts were believed to be five-coordinate. van ’t Hoff suggested a trigonalbipyramidal (TB-5) coordination geometry, while Bischoff suggested a distorted square pyramidal (SPY-5) geometry.19 The chirality of five-coordinate complexes, including the use of stereodescriptors, was discussed in 1996 by von Zelewsky, but at that time it was not expected that five-coordinate coordination compounds would be optically resolved.20 Since the situation now has changed, we believe that the stereochemistry of five-coordinate compounds needs a more thorough investigation.
In 2009, after reporting spontaneous resolution of eight-,21 seven-5 and nine-16 coordinate enantiomers, we achieved the first optical resolution of a five-coordinate complex, [Zn(S2CNEt2)2(2-vinim)] (vinim = 1-vinylimidazole).15 This complex was trigonalbipyramidal and chirality in a SPY-5 (the other common coordination geometry for coordination number five) complex was observed by Biswas et al. in 2012.22 They were able to obtain solid-state CD-spectra from individual single crystals of a copper(II) coordination polymer. In 2015, we reported the optical resolution of a palladium complex being intermediate between four- and five-coordination.23 a Department of Chemistry and Chemical Engineering, Chalmers University of
Technology, 412 96, Gothenburg, Sweden. E-mail: email@example.com b Department Chemistry and Molecular Biology, University of Gothenburg, 412 96,
Gothenburg, Sweden. E-mail: firstname.lastname@example.org † CCDC 1045814–1045816. For crystallographic data in CIF or other electronic format see DOI: 10.1039/c5nj00254k
Received (in Victoria, Australia) 29th January 2015,
Accepted 20th April 2015
DOI: 10.1039/c5nj00254k www.rsc.org/njc
Pu bl ish ed o n 20
A pr il 20 15 . D ow nl oa de d by
N or th
D ak ot a S ta te
U ni ve rs ity o n 24 /0 5/ 20 15 0 2: 55 :3 2.
View Article Online
New J. Chem. This journal is©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2015
We have limited our discussion to complexes displaying only achiral monodentate ligands (a–e) or bidentate ligands, which may be symmetrical (a^a) or non-symmetrical (a^b). The complexes can, to simplify the discussion, be divided into three groups having only monodentate ligands (group 1), one bidentate ligand and three monodentate ligands (group 2) and two bidentate ligands and one monodentate ligand (group 3).
For complexes displaying exclusively monodentate ligands, seven different compositions are possible: [Ma5], [Ma4b], [Ma3b2], [Ma3bc], [Ma2b2c], [Ma2bcd] and [Mabcde], in each case either TB-5 or SPY-5 coordination geometry is possible (Fig. 1). The complexes are chiral if aa ba c and da e in the TB-5 case. In the SPY-5 case, the complexes are chiral if a a c and b a d and at least one of the following two conditions are fulfilled: ba c (if a = d) or aa b (if c = d). The apical ligand e has no influence.