Pyridinium Dichromate Fire

Anne Marie van den Braken
Akzo Nobel, Deventer, the Netherlands

Chemical & Engineering News (21 Apr 2003) Vol. 81, No. 16, pp. 4.

Recently, we investigated a runaway reaction of 15-kg pyridinium dichromate (PDC) during repackaging. The incident exposed personnel to PDC, and a large area had to be decontaminated.

The incident: PDC was delivered in small drums, containing 15 kg of PDC each. For production, the required amount was weighed and repackaged in three stainless steel 50-L corrugated drums, each containing 15 19 kg of the product. The lids on the drums were closed with clamping rings and stack piled. Repackaging took about 45 minutes at ambient temperature (about 20 °C). About one hour later, the lowest drum started to hiss. Within a few seconds, the lid flew open and blue flame and green smoke emerged. The force threw off the drums on top. The resulting fire could be extinguished with water. A few hours later, the middle drum started to leak gas and green solids as well. Immersing the drum in water stopped this. The top drum remained undamaged.

The investigation: The operation had been carried out since 1994. The stainless steel drums were used for weighing out products and were thoroughly cleaned and dried afterward. This incident happened with the first delivery by a new PDC supplier. Interviewing the operator made clear that the decomposed material had a unusual appearance: Instead of free-flowing orange crystals, the product was orange-brown and compacted to a big lump. Using a Tripod analysis, two possible causes were defined: a low quality and high instability of the batch as supplied, or a different quality of PDC, possibly in combination with a contaminant.

A low PDC stability as such was not likely, because the product had been in storage for some time without causing problems. Repackaging the product was considered to be an important parameter in the cause. The presence of water was highly unlikely but could not be excluded completely. Several contamination experiments with water and PDC (from the same lot but from a different container) were performed. Neither gas evolution nor a runaway reaction could be initiated. Also, DSC (differential scanning calorimeter) experiments were performed on the PDC from the incident lot and compared with PDC from another supplier. The only difference that could be established was a small endothermic peak in the area around 100 120 °C that was present in the incident lot only. The decomposition of the second drum was attributed to the heat development in the drum underneath it.

Remaining questions: Although we have investigated this incident thoroughly, we were not able to find a cause. Although we had some indications that the PDC involved was of a different quality, this could not be proven, as all material from the original drum had decomposed. PDC is generally considered to be a stable product at room temperature. Because this incident had considerable health and safety consequences, we would like to hear from anyone who has experienced a "spontaneous" decomposition of commercial PDC at room temperature or knows conditions that initiate such a decomposition.

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

This page last revised December 22, 2006
© Copyright 2006 by the American Chemical Society