Spontaneous Resolution by Crystallization of an Octanuclear Iron(III) Complex Using Only Racemic Reagents (original) (raw)

The P and M enantiomers of the octanuclear [Fe 8-(m 4-O) 4 (m-4-Cl-pz) 12 Cl 4 ] complex, having T symmetry, were resolved by temporary substitution of chloride ligands by racemic 4-s Bu-phenolates and subsequent crystallization, where the (S)-and (R)-phenolates coordinate selectively to the M and P complexes, respectively. The complexes were characterized by circular dichroism analysis and X-ray structure determination. This work constitutes a rare example of enantiomeric recognition resulting in spontaneous resolution upon crystallization. Interest in the resolution of racemic mixtures stems from the importance of enantiopure substances to biological processes and the pharmaceutical industry. [1] Enantiopure materials have also applications in magnetics and nonlinear optics. [2, 3] While synthetic asymmetric molecules abound, in the absence of a chiral initiator only their racemic mixtures are synthesized , with the notable exemption of asymmetric autocatal-ysis. [4] The latter may in some cases resolve spontaneously by crystallization, yielding enantiopure single crystals. Chiral stationary phase chromatographic separation and attrition-enhanced deracemization are currently the principal resolution methods. [5, 6] Enantioselective synthetic techniques employing chiral catalysts have been developed to circumvent the difficult problem of racemate resolution. [7] The 2001 Nobel prize was awarded for the development of chiral catalysts, [8] a field that remains active and vibrant to date. [9] Most homogeneous racemic catalysts are transition-metal complexes, which can undergo facile isomerization, because of the lability of metal-ligand bonds and flexibility of their coordination sphere (e.g., Berry pseudo-rotation, Bailar twist). In contrast, successful chiral transition-metal catalysts, such as the family of Chiralphos, [10] contain ligands whose steric bulk renders them inflexible, preventing racemization. However, ligand lability can still be a problem, and is further exacerbated for metal ions with low or no crystal field stabilization (e.g., high-spin d 5 centers). Unfortunately, this list includes Fe 3+ complexes whose extensive role in homogeneous catalysis has been recognized for a long time. [9d, 11] Clearly, the field of specialty chiral chemicals stands to benefit from the availability of new, nonracemizable Fe 3+ enantiopure compounds. Examples of mononuclear transition-metal complex race-mates that have been successfully resolved are rare. For instance, monomeric ruthenium complexes containing the 1-Me-3PhCp ligand were resolved by temporary substitution of a terminal chloride ligand by (S)-a-methylbenzenemethane-thiol and subsequent fractional crystallization to isolate the two enantiomers of the complex. [12] A similar approach was used for the resolution of rhodium complexes with tripodal tetradentate ligands, employing (S)-phenylglycinato as the resolving agent. [9] Chromatographic methods have been also utilized for the resolution of metal complexes. [13] In contrast to mononuclear species, the racemization of polynuclear complexes, requiring the simultaneous rearrangement of the coordination spheres of every component metal center, adding up to a higher activation energy, is a more unlikely process. The resolution of tetrahedral M 4 L 6 metal-ligand hosts made of labile components (M = Ga III , Fe III , Al III , In III , Ti IV , Ge IV), including a water-soluble derivative, has been demonstrated. [14] It follows then that, conceptually, racemic poly-nuclear complexes may be an easier target for resolution, even if they involve somewhat labile metal cations. The area of chiral metal cluster and nanoparticle applications in catalysis has been recently reviewed and the need for further development of the field has been stressed. [15] We have characterized a family of octanuclear Fe 3+ complexes of the general formula [Fe 8 (m 4-O) 4 (m-4-R-pz) 12 X 4 ], [Fe 8 ] (Figure 1 a), where pz = pyrazolato anion