Explosion hazard with pipe from chloroprene tank

Howard E. III Simmons; James E. Johnston
DuPont Science & Engineering Laboratory, Wilmington, Del.
DuPont Explosion Hazards Laboratory, Carneys Point, N.J.

Chemical & Engineering News (7 Aug 1995) Vol. 73, No. 32, pp. 4.

We would like to alert C&EN readers to a safety hazard of a class of compounds that we have not previously seen reported as being dangerous. An explosion at one of our Neoprene manufacturing sites severely injured a worker using a portable band saw to remove a flange from an open-ended 1-inch-diameter, 12-foot-long Schedule 10 stainless-steel pipe. The pipe had been decommissioned from a refined chloroprene (2-chloro-1,3-butadiene) storage tank. The usual decontamination procedures using water and steam cleaning had been employed.

The explosion was traced to the thermally induced deflagration or possibly even detonation of a dense deposit lining the walls of the piping connected to the transfer tank. The material consisted of almost pure 2-chloro-1-nitro-4-nitroso-2-butene dimer (1). It was determined that this compound, of limited solubility in chloroprene, had formed by the reaction of nitrogen oxides (NOx) with chloroprene.

The reaction is formally the 1,4-addition product of N2O3 with chloroprene. NOx had formed in the chloroprene refining column by thermal decomposition of inhibitors used to stabilize the monomer and prevent "popcorn" polymer in vapor spaces. The inhibitor is a complex mixture of thermally and hydrolytically unstable compounds containing nitro, nitrito, nitroso, and nitrato functional groups, formed by bubbling nitrogen tetroxide (N2O4) gas through a mixture of two low molecular weight olefins. Under the refining conditions, this mixture decomposes, liberating NOx, which along with the more volatile inhibitor components, act as vapor space inhibitors.

The nitroso-dimer (1) was independently synthesized in moderate yield (43% crude yield) by the reaction of aqueous nitrous acid with a benzene solution of chloroprene to give a fine powdered product. Differential scanning calorimetry measurements show an onset of decomposition at about 80 C with a very sharp strong increase in the exotherm at 100 C. The total energy release is 2,500 joules per g, putting it in the same category as some well-known explosives. Compound 1 was only mildly sensitive in drop-weight impact testing. A shock-wave propagation measurement using powdered material destroyed the carbon steel pipe and indicated a strong deflagration. Thermodynamic calculations, however, suggest energies should be higher, in the detonation range. A shock test using dense material, as found in the plant piping, would be likely to result in a true detonation.

The 1,4-addition of N2O3 across monoand dienes appears to be a general, fairly high yield reaction. We wanted to alert your readers to the potentially hazardous products formed by the reaction of NOx and olefins.

Return to List of Safety Letters
Return to Chemical & Engineering News Home Page

This page last revised December 7, 1998
© Copyright 1998 by the American Chemical Society