Chemists at Merck & Co. were performing experiments using nitric oxide at high pressure (10–20 bar) when two instances of an explosion occurred during rapid depressurization of the NO headspace from a 500-mL closed reactor system. No injuries occurred, and damage was contained to the barricaded cell area. Both before and in between these events, NO had been used successfully about 100 times. Each explosion occurred after completion of the reaction, while venting through three-eighths- or one-quarter-inch i.d. Teflon-lined steel-braided tubing to atmospheric pressure.
Static electricity was suspected as the ignition source that, in conjunction with the presence of an oxidant (NO) and fuel (CH3OH), would lead to combustion. To confirm this hypothesis, an investigation was conducted. Preliminary results are communicated here.
The reaction system consisted of NO in conjunction with methanol under basic conditions. A literature search didn’t point to any existing cautionary notes about this reaction. Experimental ignition testing of NO systems was conducted by Fauske & Associates, which showed no combustion unless ignition energy greater than 3 J was used. This exceeds the energy typical of a static discharge, so it does not fully explain the observed combustion.
Further analysis of the reaction headspace using gas chromatography/mass spectrometry revealed that N2O was formed over time from a simple model system of NO + sodium methoxide + methanol. The conversion of NO to N2O and concomitant oxidation of methanol to formic acid proceeds to 50% in about six hours. Testing showed the energy needed to ignite the headspace of methanol under 1 bar of 50/50 NO/N2O is less than 3 mJ, several orders of magnitude lower than for similar systems without N2O.
On the basis of these results, the likely cause of the explosions is the combination of (a) formation of N2O gas and (b) generation of static potential caused by the rapid flow of gas and condensing methanol through the Teflon-lined tubing during rapid depressurization (while venting), which leads to sparking of sufficient energy to cause the combustible vapor to ignite.
We wanted to alert the chemical process industry to risks associated with this particular procedure. Anyone contemplating use of this chemistry should thoroughly evaluate its safety.
Daniel Muzzio, Ephraim Bassan, Erik Dienemann, Mark Weisel, Cameron Cowden, Scott Hoerrner, William Olsen, Michael Man-Chu Lo, Amjad Ali