Vinegar syndrome

Why does a decaying reel of film destroy itself faster the longer it has already been decaying?

Two cans of acetate film from the same production run sit on the same shelf for fifty years. One opens clean. The other announces itself before the lid is off — a sharp smell of vinegar — and inside, the roll has shrunk and buckled into ridges.

The odd part is not that one decayed but the shape of the decay. Most deterioration behaves like weathering: a roughly constant nibble, so an object half-ruined is losing material at about the rate it always did. Film does not do that. Archivists describe an item as fine for decades and then gone in a very few years. Why would damage already done make the next damage come faster?

Start with what the base actually is. Cellulose is a long chain of sugar rings carrying hydroxyl groups; to make a film base, those hydroxyls are converted to acetate esters — acetic acid units hung off the backbone. That is what makes the material clear, flexible and, unlike the nitrate stock it replaced in the mid-twentieth century, not violently flammable.

Now ask what an ester does in the presence of water. It hydrolyses: water splits the bond, restoring a hydroxyl on the chain and releasing the acid that was attached. So a triacetate base sitting in ordinary humid air is quietly coming apart, and what it releases is acetic acid.

That already explains the smell. Notice what it does not yet explain: the smell is not a symptom produced on the side. It is the decay's product, and it is a reagent. Does that matter?

It matters because ester hydrolysis is acid-catalysed. Acid is not used up; it lowers the barrier and moves on to the next bond. So the rate at which the film releases acetic acid depends on how much is already present, and every unit released stays available to help release the next. A quantity whose rate of increase is proportional to its own size does not grow in a straight line. It grows exponentially, which is why the curve looks flat for a long time and then turns a corner.

Two things then decide whether a particular reel takes that corner. The first is whether the acid can leave. A reel is a tightly wound roll, often inside a sealed can, and acetic acid vapour diffuses out of that geometry very slowly, so the acid is held against the material that made it, at its highest concentration at the core — which is where archivists typically find the damage worst. The second is temperature and moisture. Water is a reagent, so drier air starves the reaction, and the rate rises steeply with temperature, which is why cold storage is the intervention that actually works rather than a counsel of tidiness.

The visible ruin follows from the chemistry rather than adding to it. As acetate groups come off, the base shrinks; the gelatin emulsion carrying the image does not, so it is put into compression and buckles into the ridges called channelling. The picture is often still there and simply cannot be got through a scanner.

The feedback does not create the reaction; it shapes its timing. Hydrolysis begins on day one, slowly, in every acetate reel in the world. What autocatalysis adds is a long, deceptively quiet induction period followed by an acceleration — so the absence of a vinegar smell means the loop has not yet closed, not that this reel is exempt.

It also explains the archivist's habit of separating a smelly reel from its neighbours, which otherwise looks superstitious. Acetic acid vapour catalyses any acetate ester, not just the one that emitted it, so a decaying reel is a source of the reagent that starts its shelf-mates. Cold, dry, ventilated storage and acid scavengers in the can address exactly this, and none of them reverse anything already lost.

One distinction is worth pressing, because this domain trades in both kinds of fact. The vinegar smell is a genuinely perceptual instrument: the nose detects acetic acid far below the concentrations that produce visible damage, which is why a chemical warning reaches us as a sensory one. The thresholds drawn on top of it are conventions. The commonly cited "autocatalytic point" — a free-acidity level beyond which decay is treated as self-sustaining — is a line drawn across a continuous curve for triage, not a phase change the material undergoes. I recall the figure quoted as around 0.5 in free-acidity units, and the rule of thumb that each 5 °C of cooling roughly doubles remaining life; both are recalled approximations that vary by stock and source, and I would not build an argument on either.

How would we know the loop is real rather than a story fitted to the observation? Take reels matched for stock, age and measured acidity; store half sealed and half vented with an acid absorber; measure free acidity over years. Autocatalysis predicts the vented set diverges downward, and that each set's own curve bends upward rather than climbing straight. The refuting observation: acidity rising linearly with time and venting making no difference — which would mean the acid is a marker of damage rather than a cause of it.

• This is the climate counterpart to the light-dose budget in Museum lighting, which noted that humidity and temperature can destroy some objects faster than light ever will. The fixed difference: light damage accumulates in proportion to the dose received, this in proportion to the damage already done.
• Why polyester bases are immune to this particular failure, and what they fail at instead.

Every term the collection defines is gathered in the glossary.