Chemical Clusters in Polyurethane Networks. SAXS, Photoelastic and Dynamic Mechanical Behavior of Networks from Poly(Oxypropylene)Diol, Diisocyanate, and Trimethylolpropane Prepared One‐Stage and Two‐Stage Process

Abstract

The effect of the initial ratio of components and of the one‐ or two‐stage process of formation of polyurethane networks prepared from poly(oxypropylene)diol (PD, M _n=2100), 4,4′‐diisocyanatodiphenylmethane (MDI) and trimethylolpropane (TMP) on their extraction, swelling, small‐angle X‐ray spectroscopy (SAXS), photoelastic, and dynamic mechanical behavior was investigated. The networks were prepared at various ratios of hydroxy (OH) to isocyanate (NCO) groups, r _HT=[OH]_TMP/(0.5[NCO])=1–1.5, by a one‐ or two‐stage process (in the first step [OH]_PD/[NCO]=1/2) up to the full conversion of NCO groups. Due to decreasing reactivity of OH groups on TMP a nonuniform distribution of hard (TMP‐MDI‐TMP) segments (chemical clusters) was developed in both network series. The SAXS experiments proved that one‐stage networks exhibit a more heterogeneous structure than two‐stage ones. As expected, the weight fraction of the gel, w _g, and equilibrium modulus, G _e, decrease with increasing deviation from stoichiometry; the G _e and w _g values of one‐stage networks at constant r _HT are always higher than those of two‐stage networks. The frequency‐temperature (f‐T) superposition of both components of the complex compliance J* (=J′–iJ″) was performed and the horizontal shift factor, log a _T, satisfied the WLF equation for all networks. From the dependence of superimposed storage, J′_p, and loss, J″_p compliances on reduced frequency f.a _T, three contributions could be determined from the Cole‐Cole distribution; each process was characterized by the relaxation strength, ΔJ _i, relaxation time, τ_i and distribution broadness, b _i. The broadness b and strength ΔJ of the high and medium frequency process (i=1,2), which correspond to the glass‐rubber transition, are roughly independent of r _HT and the preparation method. The τ₃ and ΔJ ₃ values of the low‐frequency process in networks prepared by the two‐stage process are always greater than those in networks prepared by the one‐stage process.

Dedicated to Professor John L. Stanford on the occasion of his 60th birthday.

Keywords:

• polyurethanes
• sol fraction
• photoelasticity
• dynamic compliance
• frequency‐temperature superposition

Acknowledgments

Financial support of the Ministry of Education of the Czech Republic (project MSM0021620835) is gratefully acknowledged.

Notes

Dedicated to Professor John L. Stanford on the occasion of his 60th birthday.