# Why doesn't every pure enantiomer racemise?

After having been acquainted with enantiomerism for nearly six months, I have one standing doubt.

Consider the structure of alanine. Suppose I have a pure mixture of exclusively 'S-alanine'(left in the picture) in a flask.

There are hundreds, if not, thousands of molecular collisions in the flask per second. Why can not a bump to one of the S-alanine change it to R-alanine? This way, the mixture should racemise itself if left standing for some time. But this does not happen. Why? Can someone please explain?

• If I somehow had a tank full of left hands, and I agitated the tank, you might get lots of different conformations such as a tight fist, an open palm, etc... but there's no way I'm going to turn a left hand into a right hand (a different configuration) without chopping off the fingers and gluing them back the wrong way. Similarly, racemisation (at a chiral centre) requires the breakage of a single bond, which is not very likely to happen. When it does occur, there is often some mechanism to explain it, e.g. keto-enol tautomerism for alpha carbons. Sorry for the morbid analogy. – orthocresol Aug 31 '15 at 15:51
• @orthocresol You should write this up into an answer. – bon Aug 31 '15 at 17:06