Broken windshield with the words Accidental Scientific Discoveries

Accidental Scientific Discoveries: Mistakes are KEY

There was a great discussion this week on Twitter about science communication inspirations. One thing that came up was how important it was that the Magic School Bus encouraged kids to go out there and make mistakes. During the discussion, it became clear that some of the public doesn’t realize just how important mistakes really are in the advancement of science. So I figured this would be a good time to talk about some amazing accidental scientific discoveries.

Broken windshield with the words Accidental Scientific Discoveries

The important thing is that you wrote it down.



I feel like it’s important to emphasize that good lab hygiene is very important. But the discovery of penicillin really undermines my point on that one. Alexander Fleming famously left a dish of staphylococcus bacteria partly open. When he returned to the lab after some time away, he found that mold had started growing in the dish. More importantly, the area around the mold was completely free from the bacteria. Other scientists had speculated that this particular mold had some sort of useful effect on bacteria, but it was this accidental experiment that got to the bottom of the situation. Fleming proceeded to isolate the compound that was killing the bacteria and medicine was revolutionized.

Pencillin in several commercial forms

I’m allergic but I’m still really glad this stuff exists. Image from Wellcome Images.

Still, try to keep your bacterial cultures properly covered.


Continuing our theme of “actually follow safety procedures even though not doing so worked out”, we have the discovery of the first artificial sweetener. Constantin Fahlberg was doing experiments using coal tar and he neglected to properly wash his hands. Which is rightly horrifying to any chemist and just about any non-chemist. Chemistry is one of those fields where you wash your hands before you go to the washroom as well as after. The fact that his unwashed hands wound up anywhere near his mouth after lab work well… they certainly should not have.

But they did, and that’s how he noticed that benzoic sulfimide was actually quite sweet. He produced it commercially as saccharin, which was quite useful during sugar rationing. Later, its value for making diet sweets also became apparent.

But please still wash your hands after working in the lab. You’re far more likely to get sick than rich.

Safety Glass

Glass is an incredibly useful substance. You can see through it, which is pretty key really. If you put glass in the front of a vehicle, you can see the road ahead of you and not get a face full of bugs for your trouble. Only problem is glass has a bad habit of breaking when struck by an object.

This was actually a two part accident. The first part of the accident was when a flask became coated with the nitrocellulose film Edouard Benedictus was studying. He probably didn’t really want to deal with getting it off so he could continue experimenting. And he certainly didn’t intend to drop the flask.

But then he noticed that although the flask was clearly cracked, it had remained intact. Usually when you drop a flask the pieces end up halfway across the lab! It took him another six years to actually file a patent on the incredibly useful new material. The specific plastics involved in safety glass have changed since then, but the basic principle of laminating the glass with plastic remains the same. So you don’t get a face full of glass shards if a pebble happens to strike your windshield.

Shattered but unbroken windshield

Expensive, but less so than putting your eye out

All I ever had happen when I dropped a flask was having to repeat a reaction.


I think every chemist has at some point had a reaction result in something that can only be described as “horrible goo”. Or at least I hope so because otherwise it was just me and I feel bad. Harry Coover was trying to develop a new plastic for making high precision gun sights during WWII. Instead, cyanolacrylate proved to be a horrible goo that stuck to everything.

There was a war going on and Coover had a very specific task, so he didn’t initially do anything with the material. It wasn’t until he was working at Eastman Kodak and trying to develop a heat resistant jet canopy that it came back into his life. Another worker noticed that cyanoacrylate caused two prism to stick firmly together. This time, Coover did further testing and realized that he had an incredibly strong adhesive that didn’t require heat to activate it. And while the stickiness was incredibly annoying for the gun sight project, it was actually incredibly useful. And so the horrible goo went into production as Superglue. Not quite as versatile a solution as duct tape, but still an incredibly useful material.

An Inordinate Number of Polymers

I’ve talked about how silicone was discovered while trying to make a completely different compound. It turns out this is a very common way to discover exciting new polymers.


Hans von Pechmann was investigating diazomethane when he happened to obtain a white waxy substance. He determined that it was a long chain of CH2 units and termed it polymethylene. Diazomethane is a rather awful starting material though, so later synthesese used ethylene gas (CH2=CH2) as the monomer, hence polyethylene. Any plastic item with a 2 or 4 in the recycling symbol is made of polyethylene. Which is actually a bit of a problem at the time of writing as half of all American ethylene production is currently offline due to Hurricane Harvey.


Leo Baekland was trying to develop a synthetic alternative to shellac, since the insect source did not make it the easiest material to come by. What he wound up with instead of was the first synethetic thermoplastic. This provided all sorts of new molding opportunities. Bakelite has since been replaced by other materials, but it was still the ground breaker and has a distinctive look.


While it’s not nearly as ubiquitous in the outside world, I would be remiss if I were to omit the first polymerization of a silicon-carbon double bond. While adding an initiator to polymerize a double bond is a common way of making pure carbon polymers, it wasn’t reported in silicon-carbon double bonds until 2008. Though it’s quite likely that the polymerization of this particular silene had actually been observed many times. Since the easiest way to achieve it is to add a slight excess of one of the reagents while trying to make the initial double bond. My labmates Laura Pavelka and Kaarina Milnes took the time to find out what exactly the material they got when they overshot the stoichiometry was.

New polymers are probably the best example of making the initial mistake being the easy part. The science comes in when you figure out what actually happened so you can do it again on purpose later.

So with a new season of The Magic School Bus upon us, let’s all go out there and take chances, get messy and most important of all make mistakes.