The sprawling GAF plant was upgrading some equipment in their dye-manufacturing department. One of the oldest and most common designs for production reactors is the glass-lined metal vat, or glass kettle.

These old vessels were made of cast iron, but often coated with glass or ceramic similar to a porcelain cooking pot. A new glass-lined reactor had been installed because the old one was too small. The new reactor was made of a newer metal alloy instead of cast iron, but it was glass-lined just like the old cast iron reactor.

The new reactor was ready for testing, but the old reactor was left in place until testing was completed. The first batch was run in the new reactor, but something was wrong. The expected color reaction hadn't taken place.

A deep blue color was expected, yet there wasn't a trace. Analysis by infrared spectroscopy showed that the dye was not forming. The ingredients were sitting there without reacting. One theory was that the large batch they tested needed more heat, but that didn't make sense.

Possibly there was a weighing error or some other coincidence. The batch of materials was drawn off from the bottom spigot of the new reactor. Half of it was put in the old reactor. The reaction began, the color changed, and all went smoothly.

This was mystifying. Why didn't it react in the new unit? Maybe there was something about the amount of material, but there wasn't an obvious principle of science or engineering to apply. The remaining half-batch of reactant was placed in the new reactor, and the heat was raised a little above normal. Nothing happened.

This batch also reacted when placed in the old reactor. So what was going on here? Only the old reactor worked. Could it be that the new glass lining was somehow different and was poisoning the reaction? Was the new reactor inadvertently adding something? If that were true, why did the batch still work when moved into the old reactor?

There was a mystery here and there had to be an invisible agent or some subtle physical difference between the two reactors that was causing the problem.

Chemical analysis of water that was run through the new reactor didn't show anything unusual. The reactor was not poisoning the batch. The new reactor was adding absolutely nothing. Evidence now pointed to the old reactor as having something special, but something unknown.

The Rest Of The Story

Was the old reactor adding something? Water was run through the old reactor and analyzed for common chemical elements by atomic absorption and Iron showed up in the test. Where would iron come from? There had to be a crack in the old glass lining.

This was the magic key, and it would open up production in the new reactor. Here's what was going on. The old reactor, with its exposed iron, was catalyzing the reaction. It only took a trace of iron and the hairline crack was enough.

Since such a small amount of iron was being dissolved from the reactor, the reactor was likely to last for a century or more before being depleted. Armed with this “forensic” detail, the problem was easily fixed. A trace of iron salt was added to each batch. The old reactor had been adding a pinch of iron for many decades and now the secret was out.

There was an added benefit. The knowledge that iron was a catalyst let the chemical engineers improve the reaction by adjusting the level of iron. When all the evaluations were completed, the new reactor worked quite well, and the new formula ran three times faster than the old one. In this case, there was no substitute for lab work.