Organic Peroxides & Self-Reactive Substances: Examples & Storage

What is an organic peroxide?

OSHA’s classification criteria define an organic peroxide as a liquid or solid organic chemical which contains the bivalent -O-O- structure, and which is “as such considered a derivative of hydrogen peroxide, where one or both of the hydrogen atoms have been replaced by organic radicals” (29 CFR 1910.1200 Appendix B, B.15.1.1).

Read that definition slowly, because it does two jobs at once. It tells you what the class is — a peroxide bond with organic groups attached — and it tells you where the class came from. Every organic peroxide is hydrogen peroxide with something bolted on in place of a hydrogen.

That oxygen-oxygen bond is weak, and it is the whole story. It holds two oxygen atoms together with far less enthusiasm than almost any other bond in the molecule, so it is the first thing to break when the substance is warmed, struck, or contaminated. When it breaks it releases two reactive fragments and heat — into a molecule that is also made of carbon and hydrogen. Fuel and oxidizer, in the same structure, waiting on temperature.

The same appendix spells out the consequences. Organic peroxides are thermally unstable and may undergo exothermic self-accelerating decomposition; in addition they may be liable to explosive decomposition, burn rapidly, be sensitive to impact or friction, or react dangerously with other substances. Not “one of these” — potentially all four at once.

Organic peroxide vs hydrogen peroxide

This is the question Google itself asks on this topic, and the answer falls straight out of the definition above.

Hydrogen peroxide is not an organic peroxide. It fails on both halves of the test: it is inorganic, and both of its hydrogens are still attached. It is the parent compound of the entire class — the thing you modify to make an organic peroxide — not a member of it.

The regulation is remarkably explicit about the family relationship. Appendix B sets the boundary of the organic-peroxide class using the hydrogen peroxide content of the mixture: a formulation is excluded when it holds no more than 1.0% available oxygen from organic peroxides while containing no more than 1.0% hydrogen peroxide, or no more than 0.5% available oxygen while containing between 1.0% and 7.0% hydrogen peroxide (B.15.2.1). The criteria for the children are written in terms of the parent.

Hazard-wise the two behave in opposite ways. Hydrogen peroxide is classified as an oxidizer — it donates oxygen to someone else’s fuel. An organic peroxide already has the fuel. That is why hydrogen peroxide belongs on our oxidizers page and not this one.

A note on our own data. Hydrogen peroxide currently appears in this class’s substance list below. It should not. The entry arrives through a reactivity-group link that treats it as a member of “Peroxides, Organic” rather than as their parent, and we have not yet corrected it. We would rather show you the flaw than quietly filter it and let you find it in an audit. Everything else on this page is unaffected.

Why organic peroxides and self-reactive substances share H240 to H242

Here is the thing almost no reference will tell you, and it changes how you read a safety data sheet.

Organic peroxides and self-reactive substances are two different hazard classes that use identical hazard codes. OSHA’s definition of a self-reactive chemical — thermally unstable, liable to strongly exothermic decomposition without oxygen — ends by excluding anything classified as an explosive, an organic peroxide, an oxidizing liquid or an oxidizing solid (Appendix B, B.8.1). The classification procedure repeats it operationally: a chemical is not considered for the self-reactive class if it is already an organic peroxide under B.15 (B.8.2.1). The two classes cannot overlap by design.

And yet:

The type criteria in B.8.2.3 and B.15.2.2 are word-for-word identical for types A through F. The pictogram allocation in Appendix C lists both classes under the flame and both under the exploding bomb (Figure C.1). The UN’s own hazard-statement codification puts self-reactive chapter 2.8 and organic-peroxide chapter 2.15 in the same cell for H241 and H242.

There is exactly one point of divergence, and it is useless to you. Type G differs: a self-reactive type G has an SADT between 60 °C and 75 °C, while an organic peroxide type G is thermally stable at 60 °C or above with no upper bound — because a substance with an SADT above 75 °C is not considered for the self-reactive class at all (B.8.2.1). But type G carries no pictogram, no signal word and no hazard statement. The one place the two classes differ is the one place neither of them is labelled.

So: the H-code cannot tell you which class you have. Neither can the type, the signal word, or the pictogram. The only reliable source is the hazard class name written out in section 2 of the SDS. Failing that, the chemistry: an oxygen-oxygen bond means peroxide, while azo compounds, diazonium salts, azides and sulfonyl hydrazides are self-reactives.

This has a direct consequence for the list on this page. Because our classification is anchored on hazard codes, this class holds both — and we would rather name them than pretend otherwise. These members are self-reactive substances, not organic peroxides:

They stay in the class deliberately. Their storage requirements are identical to the peroxides’, and dropping them would mean dropping their segregation guidance — a worse error than a broad label.

Types A to G and the SADT

Both classes grade severity with letters instead of numbers, and the letters are about what the substance does in its package, not what it does in a beaker (B.15.2.2):

Two consequences that are easy to miss. Type A is not a labelling problem — it is a transport prohibition: organic peroxides of type A are not accepted for carriage under Class 5.2 (ADR 2.2.52.1). And type G gets no pictogram, no signal word and no hazard statement, which is why it can never appear in any list built from hazard codes, including ours.

Underneath the letters sits one number. The self-accelerating decomposition temperature — write it out; the abbreviation SADT mostly returns a software design methodology — is the lowest temperature at which the substance in its package decomposes faster than it can lose heat. Above it, the reaction heats itself and accelerates. It is determined by a single method shared by everyone: the UN Manual of Tests and Criteria, Part II, section 28 (B.15.3.2 for peroxides, B.8.3.2 for self-reactives).

SADT is a property of the package, not the molecule. Same peroxide, bigger drum, lower SADT — a larger mass sheds heat more slowly. This is why appendix B states its type G threshold for a 50 kg package rather than in the abstract, and why mixing two thermally stable peroxides obliges you to redetermine the SADT of the mixture: two stable ingredients can form a less stable blend (B.15.3.3).

What our data shows

Hazard code plus signal word together recover the type band, because types C and D take Danger while types E and F take Warning. Applied to the members of this class that carry a harmonised classification:

Zero type A, in both. That is not a gap in our data — it is the transport prohibition showing up in a classification inventory. Type A barely exists commercially because it cannot legally move. At the other end, type G cannot appear by construction, since it has no hazard code to be found by.

The middle is where this class lives. And note the single type E/F entry: one substance in this entire list is labelled Warning rather than Danger.

Organic peroxide examples

The recognisable ones, all drawn from this class’s members:

Benzoyl peroxide deserves a note, because it is the organic peroxide most people have met — and they have met the wrong one. The acne product is a few per cent of benzoyl peroxide dispersed in a carrier. The industrial article is type B: H241, heating may cause a fire or explosion, and it carries the exploding bomb. Same molecule, entirely different artefact. This page is about the second one.

Peroxy acids like peracetic acid are worth flagging too, since they read like acids and get filed with them. They are not: the -O-O- bond makes them organic peroxides, and the acid group is incidental to the physical hazard.

For the pictograms themselves — the exploding bomb and the flame — see GHS01 and GHS02.

Why organic peroxides are dangerous

Line this class up against its two neighbours and the answer is obvious.

A flammable liquid is fuel. It needs oxygen from the air and an ignition source. Remove either and nothing happens — which is why a flammables cabinet works.

An oxidizer is the oxygen. It needs someone else’s fuel and an ignition source. Keep it away from fuel and it sits there inert.

An organic peroxide is both, in one molecule — and the oxygen-oxygen bond is its own ignition source, triggered by nothing more exotic than warmth. There is no third party to separate it from. The fire triangle is fully assembled inside the container, and the only variable left is temperature.

That is the difference between a hazard you control by segregation and one you control by refrigeration.

OSHA’s flammable liquids standard makes the point in the driest possible way. It names organic peroxides exactly once, to exclude them: because they undergo autoaccelerating thermal decomposition, organic peroxides are excluded from every flash point determination method in the standard (29 CFR 1910.106(a)(14)). You cannot measure their flash point, because heating them does not liberate vapour to ignite — it starts the decomposition. The test destroys the sample.

A class of substances that burns fiercely but cannot be assigned a flash point does not belong in a cabinet whose whole design logic is flash point. That single exclusion explains more about organic peroxide storage than any storage rule does.

Storing organic peroxides and self-reactive substances

This is the most restrictive class on this site. Of the twelve other storage classes we track, nine are prohibited, three require separation, and none is compatible. It is the only class with no green at all — see the compatibility panel at the top of this page.

That result is not an artefact of merging two classes together. It falls out of the chemistry both of them share: acids and bases catalyse the decomposition, metals catalyse it, oxidizers and reducers both attack the peroxide bond, and every flammable in the room is fuel for a fire that does not need air. There is no neighbour that is merely neutral.

What US regulation actually requires

Almost nothing — and saying so is more useful than implying otherwise.

There is no OSHA standard for organic peroxide storage. Subpart H of 29 CFR 1910 covers flammable liquids (1910.106), explosives and blasting agents (1910.109), liquefied petroleum gases (1910.110) and anhydrous ammonia (1910.111). Organic peroxides have no section of their own. This is the same shape as oxidizers, where OSHA regulates exactly one substance and leaves the rest to consensus standards.

What does exist:

• 1910.106(a)(14) — the flash point exclusion above. Not a storage rule, but the most informative sentence OSHA has written about this class.
• 1910.119 Appendix A — the Process Safety Management list of highly hazardous chemicals names diacetyl peroxide above 70% concentration, with a threshold quantity. Hydrogen peroxide at 52% by weight or greater is listed too. Above the threshold, PSM applies: process hazard analysis, mechanical integrity, management of change. This is the only substance-specific federal hook for a peroxide, and it triggers on quantity, not on the fact that you have one.
• 1910.1200 — classification and labelling, via appendices B and C. Communication, not storage.

Fire codes carry the practical quantity limits, through NFPA 400 and the model fire codes adopted by your authority having jurisdiction. Be aware these use a different classification system — organic peroxides are graded Class I to V plus “unclassified detonable”, not types A to G. Two parallel schemes for the same substances; do not translate between them by intuition.

Temperature control comes from transport

The refrigeration requirement people associate with this class is a transport provision, not a workplace one. Dangerous-goods rules assign each formulation two numbers derived from its SADT: a control temperature, the maximum at which it can be safely carried, and an emergency temperature, at which emergency procedures begin (ADR 2.2.52.1; in the US, 49 CFR 173.21 and 173.225). Only thermally sensitive formulations get them — each type splits into temperature-controlled and non-temperature-controlled UN entries.

That split is why the transport panel in our storage tool shows nothing for these substances, and the reason is worth knowing. Organic peroxides and self-reactive substances are the two classes whose UN numbers are generic: UN 3101 to 3120 encode type, physical state and whether temperature control applies — “Organic peroxide type C, liquid” is UN 3103 — rather than naming a chemical. The named entries live in a separate formulation table (ADR 2.2.52.4; 49 CFR 173.225(c)), and its rows are keyed on concentration and diluent, not on a CAS number. Our transport data is imported from the main dangerous-goods table, so for this class there is nothing to match against. The classification depends on how the peroxide is formulated, which is exactly the point.

What this means in practice

Reading the above together: keep the class isolated rather than merely segregated, treat temperature as the primary control rather than a comfort factor, know the control temperature of what you actually hold, and check your fire code for the quantity limits — because your federal workplace regulator has not set any.

Always verify against the substance’s own safety data sheet, sections 7 and 10.