What made Arkema’s peroxides unstable in Harvey’s aftermath? | September 11, 2017 Issue - Vol. 95 Issue 36 | Chemical & Engineering News
Volume 95 Issue 36 | p. 5 | News of The Week
Issue Date: September 11, 2017 | Web Date: September 7, 2017

What made Arkema’s peroxides unstable in Harvey’s aftermath?

Chemists explain peroxide bond stability and thermal runaway
Department: Science & Technology
Keywords: Chemical bonding, Safety, Peroxides, Industrial safety, Process chemistry, Synthesis, Polymers, Hurricane, Tropical storm, Harvey

Organic peroxides, the chemicals that led to evacuations and explosions at the flooded Arkema chemical plant in Crosby, Texas, are well-known for their instability—the characteristic that provides both their chemical utility and their hazardous properties.

Two of the peroxides stored at Arkema’s Crosby site were tert-butyl 2-ethylhexaneperoxoate, which has an SADT of 35 °C, and bis(2-ethylhexyl) peroxydicarbonate, which has an SADT of 5 °C for a 75% solution.
Structure of bis(2-ethylhexyl) peroxydicarbonate and tert-butyl 2-ethylhexaneperoxoate.
Two of the peroxides stored at Arkema’s Crosby site were tert-butyl 2-ethylhexaneperoxoate, which has an SADT of 35 °C, and bis(2-ethylhexyl) peroxydicarbonate, which has an SADT of 5 °C for a 75% solution.

Organic peroxides contain the peroxide functional group ROORʹ.The O–O bond is inherently weak; the bond dissociation energy of CH3O–OCH3 is 157.3 kJ/mol compared with 335 kJ/mol for CH3–OCH3.

Exactly why O–O bonds are so weak is an “excellent question,” says Weston T. Borden, a computational chemist at the University of North Texas, “because I do not think that the correct answer is known with any degree of certainty.”

Likely, the weak peroxide bond arises from repulsion between the lone pairs of electrons on the oxygen atoms, combined with the atoms’ high electronegativity, or tendency to attract electrons. “Covalent bond formation involves transfer of electron density to the region between the two atoms that are covalently bonded, and this transfer should be more energetically costly as the electronegativity of the atoms increases,” Borden says.

The instability of the peroxide bond means that it takes little energy, such as in the form of heat, to get peroxides to break apart. That easy breakup is where organic peroxides’ chemical utility originates. Organic peroxides readily decompose into free organic radicals, RO, that may then serve as initiators for polymerization.

But peroxides’ effortless split into highly reactive radicals also means that they must be stored well below their “self-accelerating decomposition temperature” (SADT). Above that temperature, a dangerous cycle will develop that leads to a thermal runaway: Peroxide decomposition releases heat that causes more peroxide decomposition that releases more heat. The temperature at which emergency procedures should be implemented is 10 °C below the SADT, according to Arkema’s storage guidelines.

Media reports suggest that the exploding peroxide containers at the Arkema plant were a “textbook example” of a thermal runaway, says explosives expert Jimmie Oxley of the University of Rhode Island. Arkema’s storage of the materials in thousands of containers from 3.8 to 18.9 L, which prevented easy neutralization, may have been part of a safe storage strategy, Oxley adds. “If you have a larger mass, it’s easier for the material to self-heat” because the heat won’t dissipate unless the material is stirred, Oxley says.

Oxley thinks it’s unlikely that the Arkema organic peroxides detonated. Instead, the explosions were more likely pressure ruptures from buildup of gaseous peroxide decomposition products. If the decomposition products were flammable, they could also have autoignited from the heat.

Chemical & Engineering News
ISSN 0009-2347
Copyright © American Chemical Society
KM (September 8, 2017 8:39 AM)
It would be better if this article included the specific definition of SADT (self-accelerating decomposition temperature). That temperature is specific to 1) mass of material and 2) how long that mass is held at that temperature (usually defined as within 7 days). Because it is mass specific, a 5 gram sample of a peroxide does not have the same SADT as a 5 kg sample. SADT is inversely proportional to mass: larger the sample mass, lower the SADT. The SADTs quoted are usually (always?) given for the largest package used to ship the material in commerce, likely kilograms to tens of kilograms for these peroxides.
Melody Bomgardner (September 8, 2017 3:47 PM)
Hi KM - In an earlier Harvey-related story, we covered the information that Arkema has disclosed about the amounts of organic peroxides stored at the Crosby site. [http://cen.acs.org/articles/95/web/2017/08/Arkema-discloses-chemicals-quantities-risk.html ] Here's the relevant quote "containers in the range of 1-gal [3.8-L] jugs up to 4- to 5-gal [15.1- to 18.9-L]" So the SADTs in question would be those that relate to {roughly) a mass of 3.8 kg and up. Hope that helps. EPA is investigating the details about how these peroxides were stored.
Dave (September 12, 2017 5:58 PM)
A classic qualitative test to distinguish aliphatic from aromatic compounds is to burn a small amount in a flame. Aliphatic compounds burn cleanly with little to no soot. Aromatics produce a black, sooty residue. The black and sooty smoke seen at Arkema is more likely due to Luperox® A75 benzoyl peroxide powder or some other grade of aromatic peroxide.
Patrick Reynolds (ReynoldsIDC) (September 24, 2017 10:39 AM)
Having viewed many aerial photographs of the aftermath of this incident, I have yet to see anything that would resemble and "explosion", and certainly not a "blast" as many in the media have reported. I believe these trailers simply ignited, and burned. In fact, the last six trailers had to be artificially ignited just to get on with it, and get it over with.

While some of the containers may have "popped", I believe as likely as not that many of the pops may well have been the semi-trailer tires rupturing due to heat from the fires. When one of these trailer tires pops, it can often times be heard from a mile away or more.

It will be interesting to me to see if that theory proves true, or not.

Please also, see my post on LinkedIn regarding this incident, which I posted as it was happening. https://www.linkedin.com/feed/update/urn:li:activity:6309471236653146112/

Leave A Comment

*Required to comment