The physical and NMR characterizations of allyl‐ and crotylcelluloses

Tri-O-allylcellulose (degree of polymerization, DP ∼112) was prepared in ∼91% yield, and tri-O-crotylcellulose (DP ∼138) was prepared in ∼56% yield from microcrystalline cellulose (DP ∼172, and polydispersity index, PDI ∼1.95) using modified literature methods. Number-average molecular weight (Mn = 31,600), weight-average molecular weight (Mw = 191,800), and PDI = 6.07 data suggested that tri-O-allylcellulose may be crosslinking in air to generate branched chains. The polymer was stabilized with 100 ppm butylated hydroxy toluene (BHT). The material without BHT experienced glass transition (Tg, differential-scanning calorimetry, DSC) between −2 and +3 °C, crosslinked beyond 100 °C, and degraded at 298.6 °C (by thermogravimetric analysis, TGA). Mn (45,100), Mw (118,200), PDI (2.62), and thermal data (Tg − 5 to +3 °C, melting point 185.8 °C, recrystallization 168.9 °C, and degradation 343.6 °C) on tri-O-crotylcellulose suggested that the polymer was formed with about the same polydispersity as the starting material and is heat stable. While allylcellulose generated continuous flexible yellow films by solution casting, crotylcellulose precipitated from solution as brittle white flakes. Dynamic mechanical analysis (DMA) data on allylcellulose films (Tg − 29.1 °C, Young's modulus 5.81 × 108 Pa) suggest that the material is tough and flexible at room temperature. All 1H and 13C resonances in the NMR spectra were identified and assigned using the following methods: Double-quantum filter correlation spectroscopy (DQF COSY) was used to assign the network of seven protons in the anhydroglucose portion of the repeat unit. The proton assignments were verified and confirmed by total correlation spectroscopy (TOCSY). A combination of heteronuclear single-quantum coherence (HSQC) and 13C spectroscopies were used to identify all bonded carbon–hydrogen pairs in the anhydroglucose portion of the repeat unit, and assign the carbon nuclei chemical shift values. Heteronuclear multiple bond correlation (HMBC) spectroscopy was used to connect the resonances of methines and methylenes at positions 2, 3, and 6 to the methylene resonances of the allyl ethers. TOCSY was used again to identify the fifteen 1H resonances in the three pendant allyl groups. Finally, a combination of HSQC, HMBC, and 13C spectroscopies were used to identify each carbon in the allyl pendants at 2, 3, and 6. Because of line broadening and signal overlap, we were unable to identify the conformational arrangement about the C5 and C6 bond in tri-O-allyl- and tri-O-crotylcelluloses. © 2000 John Wiley & Sons, Inc. J Polym Sci A: Polym Chem 38: 1889–1902, 2000