Quantitative Determination of Threading in Rotaxanated Polymers by Diffusion-Ordered NMR Spectroscopy Tiejun Zhao and Haskell W. Beckham* Polymer Education and Research Center, School of Textile and Fiber Engineering, Georgia Institute of Technology, Atlanta, Georgia 30332-0295 Harry W. Gibson Department of Chemistry, Virginia Polytechnic Institute & State University, Blacksburg, Virginia 24061 Received December 12, 2002; Revised Manuscript Received April 10, 2003 ABSTRACT: Two-dimensional diffusion-ordered NMR spectroscopy (DOSY) was used to examine a rotaxanated polymer and its blends with unthreaded macrocycles. The threaded macrocycles exhibit self- diffusion coefficients similar to those of the polymer, while the diffusion coefficients for the unthreaded macrocycles are much larger. For poly[(styrene)-rotaxa-(30-crown-10)], the method provided definitive proof that the 30-crown-10 is threaded; for blends, the method allowed quantitative determination of the threaded-macrocycle fraction. The DOSY technique is particularly useful for those rotaxanated polymers in which weakly interacting macrocycle and polymer offer no other detectable spectroscopic signature. Threading can be proven, and the fraction of threaded macrocycles can be determined. Introduction Molecular architectures that consist of cyclic mol- ecules threaded onto linear polymer segments are called polyrotaxanes when bulky end groups block dethreading of the cyclic component and polypseudorotaxanes when blocking groups are not present. 1,2 We generally refer to both types as rotaxanated polymers. Syntheses of rotaxanated polymers are typically followed by experi- ments to determine whether the macrocycles are threaded or not. In some cases, threaded macrocycles cause or experience changes in local electronic structure that are detectable by various spectroscopic techniques. For example, changes in NMR chemical shifts of mac- rocycles are observed when there exists hydrogen bond- ing, charge transfer, or dipolar coupling interactions with the backbone. 3-9 However, many rotaxanated structures have been prepared for which threading is not accompanied by changes in the local electronic structure. This is often the case for rotaxanated materials with no specific interactions between macrocycle and linear polymer. For example, some polyrotaxanes based on aliphatic crown ethers do not exhibit NMR chemical shift changes from those of their respective unthreaded components. 10-13 Many of these threaded architectures are created by polymerization in macrocycle solvents or cosolvents. Following polymerization, the product contains both threaded and unthreaded macrocycles. It is dissolved in a good solvent for both components and then precipi- tated in a selective solvent for the unthreaded macro- cycle. The macrocycle content of the product can be followed spectroscopically. The dissolution/precipitation procedure is repeated until the macrocycle content becomes constant. At this point it is assumed that the remaining macrocycles are threaded. However, a more direct method to determine whether the macrocycles are threaded is desired for (1) definitive proof of threading, (2) quantification of threaded and unthreaded fraction, and (3) examination of threading/dethreading kinetics in polypseudorotaxanes. Self-diffusion depends on molecular size. Since mac- rocycles are typically much smaller than the polymers on which they are threaded, the two components of a rotaxanated polymer should exhibit vastly different intrinsic self-diffusion behavior. When the two compo- nents are threaded, the macrocycle self-diffusion should be governed by the self-diffusion of the polymer. Diffu- sion-ordered NMR spectroscopy (DOSY) 14-18 can be used to correlate diffusion coefficients with molecular struc- ture via chemical shifts. It has been successfully applied to study molecular association, which causes changes in self-diffusion coefficients; the association constants of several guest-host complexes, such as cyclodextrin and various small molecules, have been determined. 19-21 In fact, DOSY has been used to study pseudorotaxane formation between R-cyclodextrin and a series of ho- mologous R,ω-diaminoalkanes. 22 The technique should be useful also for quantitative determination of thread- ing in rotaxanated polymers. Experimental Section The materials chosen for this study include polystyrene (PS), 30-crown-10 (30c10), and the rotaxanated polymer (see Figure 1) consisting of these components, poly[(styrene)-rotaxa-(30- crown-10)] (PS-rotaxa-30c10). All materials were dried at 50 °C under vacuum for 2 days before use and then dissolved into CDCl 3 (Aldrich, 99.9%) at concentrations around 0.7% (w/w) for NMR analysis. Physical blends of 30c10/PS and 30c10/PS- rotaxa-30c10 were prepared by dissolving each component directly in CDCl 3. The PS, a commercial sample chosen because of its molecular-weight similarity to the rotaxanated polymer, was characterized by gel permeation chromatogra- phy: M n ) 7 × 10 4 g/mol and Mw ) 3 × 10 5 g/mol. The 30c10 and PS-rotaxa-30c10 were prepared according to published methods. 23,24 The PS-rotaxa-30c10 (molecular weight ∼ 10 4 g/mol) contains bulky tris(p-tert-butylphenyl)methyl groups at the PS chain ends that prevent dethreading of 30c10 (see * To whom correspondence should be addressed. E-mail: haskell.beckham@tfe.gatech.edu. 4833 Macromolecules 2003, 36, 4833-4837 10.1021/ma025959g CCC: $25.00 © 2003 American Chemical Society Published on Web 05/28/2003