Is science sometimes in danger of getting tunnel vision? Recently published ebook author, Ian Miller, looks at other possible theories arising from data that we think we understand. Can looking problems in a different light give scientists a different perspective?

A 1960s PhD (2) Project! What project?

Mid December, and time to select a supervisor and project. That was easy; I selected my supervisor on the grounds he was enthusiastic over a project. I did my background reading, got some glassware to my bench, and was all ready to start, but with two days before time to go away for Christmas, I headed for the library. Then, for a Christmas present, I gave my supervisor the results of the project: it was neatly written up in the latest JACS. Oops! Worse, there was no safety net; apart from these measurements there was nowhere to go with the compounds.
 
Back from Christmas, my supervisor gave me two new projects. One was to examine solvolyses of a number of substituted acetylacetonates of various metals, to see the degree of electronic effects that could be related to various metals. That seemed a great project, until the literature indicated that the rate constants of at least some were zero. Not promising! The alternative project involved fusing a cyclopropanecarboxylic acid to the 9-10 position of phenanthrene, and examine the possibility of conjugative effects of substitution on the phenanthrene on acidity. At the time, there was a debate as to whether cyclopropane transmitted conjugative effects.
 
I saw difficulties. To get a starting material, one took 3 kg of refluxing phenanthrene and carefully added about 700 g of ethyl diazoacetate while avoiding using scratched glassware, etc. If things went wrong, bang! The report indicated that it was possible to extract about 25 g of product from the 3.6 kg of resultant tar. As if that were not bad enough, how did one get substitution? Trying to make 3 kg of a substituted phenanthrene did not seem to be particularly attractive, especially since the most reactive carbons of phenanthrene itself are 9 and 10. Substituting the basic product was possible, but there was a risk the 2-position (para to the other phenyl) would be the most reactive. Accordingly, I was fairly dejected when the Head of Department met me walking towards the department. When I explained, he smiled and said, “Why don’t you select your own project?” That was probably just to get rid of me, but I took him up on this offer. (Note to self at time: supervisor not functioning on all cylinders, and most of time so far, he was somewhere else! Little did I know I was getting into the groove!)
 
At that time, two results had come in from the dissociation of a sequence of 2-phenycyclopropane carboxylic acids, employing the Hammett equation. Recall my baggage? I knew about that. Now the interesting thing about this was the two determinations, one in water and one in aqueous ethanol, came to opposite conclusions! Further, the conclusions were based on the Hammett rho values, which were known to attenuate substitution effects by about a third for every saturated carbon in the chain, but full conjugation, e.g. a double bond, showed little attenuation. The determination in water had substitution effects attenuated by the same as two methylene units, i.e. no conjugation, while the aqueous ethanol results had the rho value about 30%  higher, which led to the conclusion that cyclopropane conjugated about 30-50% as well as a vinyl, as in cinnamic acids. What do you think about that?
 
I regarded this as hopelessly naïve. Suppose whatever induction measured acted like a fluid. An example might be pushing electron charge, or, dare I think it at the time, enhancing probability? Cyclopropane has two routes to go through, therefore, when you add the second route, it should be 30% higher anyway. More interesting were the sigma values; these represent the ability of the substituent to "push or pull" charge/potential, and each has an inductive value, which is enhanced if the substituent can participate in a mesomeric effect. A quick graphing of the carboxylic acid results showed the rho values were too small to be sure, but if we discarded the para nitro substitution, the water results were also consistent, but with worse scatter, with the rho value being 30% higher. Discarding inconvenient results, I here you say. Well, yes, but there was a reason. The paranitro compound is almost totally insoluble in water, the others are not a lot better, so I considered that it was doubtful if true equilibrium was reached. The ethanol results would be better. The basic problem with considering sigma values was that rho was too small that experimental uncertainty made a conclusion unreliable, but what you could see indicated, based on para methoxy, that there was very little or no conjugation.
 
So, I could see a project: measure the dissociation constants of the 2-phenylcyclopropylamines. Amines, with one less atom in the road, have higher rho values, which I knew from my previous summer work. Baggage at work for me! So, all I had to do was wait for supervisor to reappear. More to come.
Posted by Ian Miller on Oct 19, 2012 11:07 PM Europe/London

Share this |

Share to Facebook Share to Twitter Share to Linked More...

Leave a comment?

You must be signed in to leave a comment on MyRSC blogs.

Register free for an account at http://my.rsc.org/registration.