The formation of poly(alkylene sebacate-crown ether pseudorotaxane)s by condensation of linear alkylene diols and sebacoyl chloride in the presence of crown ethers in the neat state has been studied. It was found that the average number of crown ether molecules per repeat unit in the polypseudorotaxane was a function of (a) the ring size and (b) the stoichiometric ratio of macrocycle to diol but independent of (i) the equilibration time of the diol and crown ether prior to addition of the diacid chloride, (ii) the length of the diol, and (iii) the temperature of equilibration and polycondensation. All of these observations are consistent with the involvement of hydrogen bonding between the diol and the crown ether as a driving force for threading, except the lack of temperature dependence. Dethreading of the isolated polypseudorotaxanes was shown to be extremely slow. Therefore, it was reasoned that the lack of temperature dependence was due to dethreading during the polymerization, inasmuch as once the ester bond has formed there is no strongly attractive force between the linear and cyclic species and the low molecular weights of the growing oligomeric esters would permit relatively facile dethreading. Based on this idea, a bulky tetraphenylmethane-based bisphenol was employed to make a copolymer (1:4) with 1,10-decanediol; indeed, the purified polyrotaxane contained more than twice as much crown ether as the polypseudorotaxane from the linear diol, confirming that dethreading does occur during the polymerization process. The polyrotaxanes all were capable of extracting metal ions from aqueous solutions. In the cases with high loadings of crown ether two distinct crystalline phases were detected by DSC: one due to the polyester backbone and one due to the crown ether; glass transitions were also observed for the crown ether component of the polyrotaxanes. Polyrotaxanes possess higher intrinsic viscosities than the backbone polymers of the same molecular weight due to increased hydrodynamic volume brought about by the macrocyclic components. However, differential solvation of the backbone and cyclic components of the polyrotaxanes was demonstrated; the intrinsic viscosity of the polyrotaxane decreased in a good solvent for the crown ether. The temperature dependence of the melt viscosity of a polyrotaxane was essentially the same as that of the polyester model backbone, but the absolute melt viscosity was much lower due to reduced chain entanglement.

Physically branched and cross-linked polymeric structures were produced for the first time by rotaxane formation during reaction of a pendant group of a preformed macromolecule. The rotaxane structure is believed to form from a hydrogen-bonded bimolecular complex of 5-(hydroxymethyl)-1,3-phenylene-1‘,3‘-phenylene-32-crown-10 (16) by esterification of the hydroxy group of one macrocycle through the cavity of the second in its reaction with poly(methacryloyl chloride) (12). For esters formed in model reactions of 12 with methanol and with 5-(hydroxymethyl)-1,3-phenylene-16-crown-5 (14), which is too small to be threaded, the degrees of polymerization were identical; however, the polymer from reaction of 12 and 16 under the same conditions had a significantly higher degree of polymerization and polydispersity, i.e., was highly branched via rotaxane formation. Increasing the concentration in the reaction of 12 with 16 led to the formation of a gel fraction along with a high molecular weight sol fraction; the gel represents a novel network structure based on mechanical interlocking via rotaxane structures. 2D NOESY NMR experiments clearly demonstrated the rotaxane structure as manifest in the through-space correlation of the benzylic protons of the “thread” with the intra-annular protons of the “bead”.

No abstract

A diol with a sterically large core, namely bis(p-tert-butylphenyl)bis[p-(2-(2-hydroxyethoxy)ethoxy)phenyl]methane (diol BG 7), was synthesized by a multiple step method and successfully used for the preparation of a sebacate-based polyester 30-crown-10 (30C10) rotaxane with blocking groups or stoppers along the backbone. It was found that the resulting polyrotaxane had a threading efficiency (m/n value, the average number of cyclic molecules threaded per repeat unit) 5 times as high as that without BG under the same conditions, which proved that the BG can effectively prevent threaded 30C10 from slipping off the polymeric backbone during the preparation of the polyrotaxane. Additionally, new evidence for the formation of the polyrotaxane was demonstrated, including a chemical shift of threaded 30C10 protons different from that of the unthreaded species, a through-space interaction between 30C10 and the backbone proved by 2D NOESY measurements, and hydrolytic recovery of 30C10.

g, = g, = 1.99, g , = 1.96, and D = ~ 4.57 cm-'. In this case the agreement between observed and calculated zero field splitting can be considered as very satisfactory. Interestingly, non-zero fourth-order parameters of the order of cm-are calculated. This can have a large significance for the interpretation of the magnetic properties of clusters that show quantum tunneling of the magnetization at low temperat~re.~"] Improvements to the fit can be achieved by taking into consideration the difference in the ligand field parameters of the equatorial ligands. The average M n -0 distance on the x axis is 1.913 A, and that for the ligands on they axis is 1.933 A; thus some anisotropy may be anticipated. If we assume that the calculated difference in the Dq values for the axial and equatorial ligands is due to an exponential decrease of the parameter with the metal-ligand distance, the Dq parameters for the ligands on the x and y axes can be calculated as 1850 and 1750 cm-', respectively. By setting e: = 9 595 cm-', e: = 2570 cm-', ei = 9080 cm-', e: = 2430 cm-', and leaving the parameters of the axial ligand at the previously determined value, we calculate g, = 1.99, g, = 1.99, g, = 1.96, D = -4.55 cm-', and E = 0.28 cm-', in good agreement with the experimental data.[Mn(dbm),] is the first molecular manganese(u1) complex for which a detailed analysis of the EPR parameters could be performed. The results given here show that HF-EPR can easily provide the g factors and the zero field splitting of manganese@ 11) complexes. The success of the ligand field analysis confirms that structural information can be readily obtained from the spectra. Furthermore, we have shown how reasonable estimates of theg and D tensors in pseudo-octahedral manganese(II1) complexes can be made simply by recording the electronic spectra in solution. We have not observed hyperfine splittings in our sample, but, given the success in the interpretation of the other tensors it is reasonable to assume that the ligand field analysis is applicable also for the A tensor. This is of great significance for the current efforts in the understanding of the structure of photosystem 11. Experimental Section[Mn(dbm),] was obtained by an alternative procedure to those already reported for the synthesis of manganese(II1) tris(P-diketonate) complexes [20]. A solution of Hdbm (4.4 mmol) and triethylamine (12 mmol) in anhydrous, oxygen-free CH,CN (40mL) was prepared under dinitrogen. Solid MnBr,.4H,O (2 mmol) was then added, and the solution was stirred for 1.5 h under dinitrogen. The yellow precipitate was collected by filtration and suspended in CH,CN with stirring. Upon contact with air, the precipitate progressively dissolved and the mixture turned brown; black lustrous crystals of [Mn(dbm),J formed quickly. The purity of the samples was checked by microanalysis: Calcd. for MnC,,H,,O, (%): C 74.58, H 4.59; found: C 74.84, H 4.57