In my point of view, the basic issue in this question is the light emission. A recorder is, in fact, a light detector. In the 10^-3 to 10^-6 m scale (wich involves large atomic structures), we can detect the light reflected from the structure (this is exactly what our eyes do). Think about itl: if you throw a lot of small balls against a wire fence (with holes not much bigger than the ball diameter), there is a great chance that a lot of balls hit the fence and "reflect" backwards. If you were blind, you could estimate the structure of the fence based on the pattern of the reflected balls. Well, a recorder does the same thing, but with photons instead of balls and atomic structures instead of the fence.

When you go to smaller scales (10^-9 for instance), it is like trying to hit the ball in a single wire of the fence. The chance here is much smaller, as the reflection of photons by molecular structures. To detect reactions, it is even worse: you will need to hit the interchanging electrons of the reaction in a so tunned way that you could track its "path". Even in a non-quantum regime, it would be almost impossible; it's like trying to "see" the trajectory of a flying bird throwing stones on it (please don't do that).

Some reactions emit light, wich we can detect. But still in those cases, we're not able to record a "movie" of the reaction, but only estimate the energy involved and, maybe, some other useful information. We can say that there is a bird flying around, but we can't see it or know where it is going.

Thank you Marcos! This is a fantastic explanation.

Would it be theoretically possible to record a movie of the reaction despite its difficult implementation? What I mean is that since light is nature's speed limit, are chemical reactions significantly slower than the speed of light? If so, I can imagine a path able to be tracked.

Or, better yet, imagine covering a bird with paint that flew against a wall? As the bird flies, you could see its path traced along the wall thanks to the paint. I wonder if a similar tagging can occur between molecules...

Insights would be appreciated :)

There is no real need for such a recorder.

We now have sufficient knowledge that we can accurately predict chemical reactions by computer modelling.

The pharma industry, in particular, are using complex chemical/biological models to investigate processes prior to any physical molecular generation.

Search on "Computer Simulation of Chemical Reactions"