I shouldn’t have to tell any scientist
this, but I have to remind myself of this, so presumably some of you
have forgotten as well. That’s okay. It’s natural. But here’s the
thing – never stop investigation at the point of reaction failure.
I’ve got an anecdote.
I work in a polymer lab which means my
goal is usually to produce a piece of plastic. Typically this will
be a film – think like saran wrap but thicker. A sheet of plastic
that can be cut, shaped, punched, and all other manners of
manipulations to test it for novel physical or chemical properties.
I’ve spent a few months trying to
replicate a reaction from literature. It should have been easy.
Should have been. And all together it has been fairly successful –
I can certainly make that polymer but my molecular weight has been
too low. When your molecular weight is too low, you can’t make a
film. You can make a spread of some plastic that crackles like paint
from the 1500’s but you can’t make a useful film. Such has been the
First we tried a procedure that was
scribbled by hand into a Moleskin from a colleague. Didn’t work. But
he hadn’t done that reaction in years so I got to thinking about what
might be the issue.
So I broke it up into two steps after
pitching an idea to my PI in a whirlwind of whiteboard drawings.
Didn’t work. It worked….better…but still not good enough.
We tried exactly the literature too, of course. Somehow, that didn’t work that well either. Polymer, yes, but crackling, brittle, and useless.
We tried inventing our own procedure
using what we knew about the synthesis. Wouldn’t it be neat if you
could do a one pot synthesis of a polymer in a sonicator? …doesn’t
work. Oh well.
So, one day, I’m making an attempt at
producing this polymer. It’s a 24 hour reaction. (That was the appeal
of trying the sonication route – could we cut time by agitation?
Turns out we can’t vent enough of the by products off for to be
viable, at least in a bath sonicator. A tip sonicator could work but
they’re expensive and we don’t have our own to mess around with.)
again, 24 hour reaction.
My worst nightmare.
My biggest fear.
nightmares about what could happen when I run these – the stir bar
shifts slightly off center and smashes the thermometer or the water
flow to the condensing tubes slips and I end up having to clean up 12
hours of water just pouring into the hood and all over my reaction.
Or the gas tank runs out. (That one has actually happened.) Or the
grad student doesn’t realize you’re running an oxygen sensitive
reaction and takes your argon tank for his project. (That one has
also actually happened.) Or the variac is old and the components
inside are not to be trusted and it takes your 60C reaction up to
170C while you’re having nightmares about this kind of thing at
home…and you come into lab to a burnt up round bottom flask with no
solvent and a stir bar trapped in a polymer mess.
guess. That one actually happened.
Luckily, this polymer was
not prone to crosslinking (crosslinked polymers can’t be dissolved)
so I just got more solvent and let it soak in a round bottom flask of
solvent until I could free the stir bar, filter out some weird solid
parts (a few crosslinks? Burnt…stuff? I don’t know.) and get the
solution. The solution that had been through more than I intended for
it to suffer through.
At this point the procedure was to
continue to part two of the reaction, crash out the polymer, filter
it under vacuum, obtain a solid powder (the prospect of a solid
powder is attractive to industries that use polymers, as opposed to
purchasing a wet resin), redissolve said powder and attempt to cast a
Did I mention the powder should have
been a yellowish/peachy/off-white color? Did I mention it came
out…well, lavender? Purple!
It shouldn’t have worked. I tabled it,
marked it off as a failure. Sure, I dissolved some of it but I just
left it on my shelf and decided it was probably worthless. I
mean…it was purple. And the polymer I’m seeking is known to be
yellowish. So, this had to be another failure in a long string of
And then I looked up at my shelf one day and it
wasn’t a purple solution anymore. It had turned orange…over the
course of some days. Weird.
So I popped the damn thing in the
oven on a really un-engineered curing cycle, which is to say, I just
threw the damn thing in a vented oven and said, “screw it.” (I
would have used the real one I know works for the commercial material
but I accidentally locked the programmable oven into a separate program earlier in
the day. Whoops!)
We were about to go home. I took it out of
the oven I had put it in.
And then I peeled it off the glass.
Polymer film. Thin, flexible, strong.
And I was so freaking
shocked that I just handed it to my PI silently while staring at the
sink. Perfectly still, staring at the sink at the end of the day
before a long weekend.
that reaction that got supposedly ruined when the variac
“The burnt one?”
So, moral of the story – KEEP YOUR
FAILURES. TREASURE THEM. SEE WHAT THE HECK HAPPENS WHEN YOUR REACTION
GOES VERY WRONG.
Monday morning, we start a whole new round of
testing because we just got some clues. And maybe it’s the polymer I
was trying to make but maybe it’s a little be different. It’s gonna
be chemical Nancy Drew and I am pumped.