One of the biggest shocks for our students when they start projects is the difference between the highly scripted experiments in the undergraduate teaching laboratories and the research project lab.
Like many other chemistry programmes, our students spend two years doing lab experiments following a lab script. Doing experiments that are pretty much proven to work and really, will stand a fair amount of carelessness in the prep is a great way to learn techniques, produce products for analysis and to get a bit of confidence in the lab. It is, however, poor preparation for following literature preps (of indeterminate detail and reproducibility), modifying or adapting preps to make derivatives, and generally planning complex synthetic schemes. As noted in the previous post, there seems to be some reluctance to interpret spectral data fully and close to the time the experiment was conducted. I don’t quite understand why it all gets left until write-up time but I think many students assume the reaction has worked and so it’s all fine.
You have to work out what you’ve made, how much and what the impurities are before you move on. No matter what. If chemistry research was simply successful based on the intention of the chemist, we’d all be Nobel Laureates. Not doing this strikes me as a fundamental lack of understanding of…well…chemistry really and how chemical reactions work. Or perhaps it is failing to ‘join up the dots’. For example, if you’re following a literature prep with a 60% yield, you have to consider what the other 40% of the moles of the limiting reactant are up to. Why then assume it’s worked perfectly. Or if you’ve added one reactant in a significant excess (which we do a fair bit, aiming for best possible yields, dendrimers and all that), you’ll need a plan for getting rid of the rest of that excess.
Following or adapting preps is one thing, leaping off into the unknown is very different. This can be as straightforward as trying to adapt a prep for a significantly different desired product. That’s when critical appraisal of every scrap of evidence that you can obtain is required. Observations during the reaction, crude spectral data, purified spectral data. And you can’t make assumptions that the desired product has formed. Checking over a reaction when it has gone wrong always starts with checking what you did and whether you did exactly as intended.
Ultimately it doesn’t matter whether you’ve made your intended product or an unexpected product: interpreting all data fully right there and then is the correct way to do science. If you have to repeat the reaction, figure out a way to check the data more quickly by all means, but don’t do that the first time through. The question is how to design and run undergraduate labs to develop that level of critical thinking.