ABSTRACT:A flexible polyurethane foam fire-retarded with 7.8% tris(1,3-dichloro-2-propyl)phosphate (TDCPP) was found to lose by volatilization 80% of this material at 200ЊC, a temperature at which there is as yet little decomposition of the foam. It is therefore to be expected that when this foam burns, most of its TDCPP will vaporize into the flame. The latter would therefore be expected to be the main site of activity of this retardant. However, when TDCPP was injected directly into the flame of the burning unretarded foam, no inhibiting effect was apparent. This, in contrast to the above conclusion, points to the condensed phase as the main site of the retarding activity. Although there is some additional evidence that supports the latter interpretation, it is difficult to reconcile it with the small amount of residual TDCPP available for reaction in the condensed phase. Flames of unretarded foams were also unaffected by the injection of other halogen containing materials, such as trichloropropane, HCl, and HBr. It is suggested that the temperatures of these polyurethane flames may not be sufficiently high to initiate the radical-trapping-based flame-poisoning mechanism classically attributed to the halogenated fire retardants. This could be the reason why the latter are less effective as fire retardants for polyurethanes than they are for many other substrates.

Tris(1,3-dichloro-2-propyl) phosphate (TDCPP) is used commercially as a fire retardant for flexible polyurethane foams. It was found to act both in the condensed (pyrolysis zone) and the vapor (flame) phases. The extent of its activity in the individual phases depends on the way in which the specimen is ignited. Under conditions of candle-like, top-down burning, retardation seems to occurs mainly in the condensed phase by a mechanism apparently based largely on the barrier properties of a phosphorus-containing carbonaceous layer that builds up on top of the liquid pyrolyzing layer beneath the flame. As the formation of this barrier requires time, extinguishment in this mode is relatively slow. Extinguishment is much faster in bottom-up burning, where the flame appears to be the main site of the retardation. In this mode, because of the orientation of the specimen relative to the flame, a disproportionately large amount of TDCPP enters the flame. It is proposed that TDCPP decomposition products, HCl and other low-fuel-value materials dilute the fuel vapors sufficiently to reduce their flame propagation velocity to below that at which they stream out of the pyrolysis zone. This pushes the flame away from the pyrolysis zone, uncoupling the thermal feedback mechanism that produces the fuel.

No abstract