Complexes of the Triolide from (R)‐3‐Hydroxybutanoic Acid with Sodium, Potassium, and Barium Salts: Crystal Structures, Ester Chelates and Ester Crowns, Crystal Packing, Bonding, and Electron‐Localization Functions

Publisher Website
Google Scholar
Abstract
The triolide of (R)‐3‐hydroxybutanoic acid ((R,R,R,))‐3,7,11‐trimethyl‐2,6,10‐trioxadodecane‐1,5,9‐trione; (1), readily available from the corresponding biopolymer P(3‐HB) in one step, forms crystalline complexes with alkali and alkaline earth salts. The X‐ray crystal structures of three such complexes, (3 NaSCN)·4 1 (2), (2 KSCN)·2 1 · H2O (3), and (2) Ba(SCN)2 · 2 1 · 2 H2O · THF (4), have been determined and are compared. The triolide is found in these structures (i) as a free molecule, making no contacts with a cation (clathrate‐type inclusion), (ii) as a monodentate ligand coordinated to a single ion with one carbonyl O‐atom only, (iii) as a chelator, forming an eight‐membered ring, with two carbonyl O‐atoms attached to the same ion, (iv) as a linker, using two carbonyl O‐atoms to bind to the two metals of an ion‐X‐ion unit (ten‐membered ring), and (v), in a crown‐ester complex, in which an ion is sitting on the three unidirectional CO groups of a triolide molecule (Figs. 1–3). The crystal packing is such that there are columns along certain axes in the centers of which the cations are surrounded by counterions and triolide molecules, with the non‐polar parts of 1 on the outside (Fig. 4). In the complexes 2–4, the triolide assumes conformations which are slightly distorted, with the carbonyl O‐atoms moved closer together, as compared to the ‘free’ triolide 1 (Fig. 5). These observed features are compatible with the view that oligo (3‐HB) may be involved in the formation of Ca polyphosphate ion channels through cell membranes. A comparison is also made between the triolide structure in 1–4 and in enterobactin, a super Fe chelator (Fig. 5). To better understand the binding between the Na ion and the triolide carbonyl O‐atoms in the crown‐ester complex, we have applied electron‐localization function (ELF) calculations with the data set of structure 2, and we have produced ELF representations of ethane, ethene, and methyl acetate (Figs. 6–9). It turns out that this theoretical method leads to electron‐localization patterns which are in astounding agreement with qualitative bonding models of organic chemists, such as the ‘double bond character of the COOR single bond’ or the ‘hyperconjugative n → σ* interactions between lone pairs on the O‐atoms and neighbouring σ‐bonds’ in ester groups (Fig. 8). The noncovalent, dipole/pole‐type character of bonding between Na+ and the triolide carbonyl O‐atoms in the crown‐ester complex (the NaOC plane is roughly perpendicular to the OCO plane) is confirmed by the ELF calculation; other bonding features such as the CN bond in the NaSCN complex 2 are also included in the discussion (Fig. 9).