Is there any general rule about predicting whether a compound would be thermally labile from it's chemical structure? If so, could you please pass along any references you have found that provide a good description?