PHD polyols are prepared for the same reason that copolymer polyols are: to provide a stable filler phase for flexible foams with greater load bearing at low density . PHD polyols are created by rapid reaction of polyisocyanates, usually toluene diisocyanate (TDI), with low-molecular-weight polyamines or hydrazine in a polyol phase. While reaction of isocyanates with amines is much faster than that of isocyanates with polyol hydroxyls (see Chapter 3), there is sufficient urethane formation with the polyol to create a stabilization mechanism for the formed particles.
A continuous process for making PHD polyols generally involves impingement mixing of the components. Batch descriptions have been described and economic analyses presented . The process involves rapid addition of TDI to hot (80 °C) aqueous hydrazine dispersed in a polyol used for molded foam applications under vigorous mixing conditions. The stoichiometry is such that isocyanate functionality exceeds amine functionality to insure reaction with the polyol phase and militate against the need for system neutralization. The heat of reaction is a major source of control and manufacturing attention. Subsequently, the system is dehydrated under heat and vacuum.
The reaction of isocyanates with amines is 1000 times faster than their reaction with primary hydroxyls (see Chapter 3). Thus, the PHD polyol reaction does not promote the level of compatibilization and size limitation available with the copolymer polyol process that employs macromer stabilizer technology. Additionally, the rapid amine-isocyanate reaction results in poor control of particle nucleation and formation. The high concentration of reactants in the initial stages results in many initial nuclei but also diffusion-controlled reaction and growth. The result is a relatively broad distribution of particle sizes. The low concentration toward the end of the reaction results in the formation of fewer and smaller particles with an increased incidence of polyol reaction (abetted by the high temperatures of the reaction at the end). This usually results in PHD polyols exhibiting bimodal distribution or a broad monomodal distribution of particle sizes from hundreds of nanometers to several micrometers. Additionally, viscosities of PHD polyols tend to be somewhat higher than the SAN copolymer polyol. The causes of this effect are not completely understood but probably result from a significant amount of particle-polyol colloidal interaction resulting in less free polyol in the system and very small particles acting as chain extenders increasing the effective molecular weight of the fluid phase. The effect is not huge, and process improvements have been able to produce 30% solid PHD polyols in the same viscosity realm (~3000^1000cP) as those from 40% solid SAN copolymers .
Like SAN copolymers, most PHD products are spherical morphology. A less desirable, though still interesting, result can occur to form rods of polyurea-phase dispersion. The less desirable aspect is that the high aspect ratio dispersion particles can significantly increase the viscosity of the solution due to increased opportunity for particle-particle interactions. On the other hand, higher aspect ratio particles can significantly increase the load-bearing efficiency of PHD polyols that would then be incorporated into the subsequent foam. The different morphology can result by controlling the disposition of isocyanate and amine phases within the polyol-reacting medium. In the case that the solid phase forms either interfacially or due to two-phase (amine-isocyanate) interdiffusion, the typical and desired (from a low-viscosity standpoint) spherical particle is obtained. In the case that the solvent polyol acts as an intermediating and diffusion rate controlling medium, it becomes possible to obtain rodlike structures.