The octet rules from chemistry is the idea that chemicals, when they react, will always try to gain eight electrons in their outer shell, as this is similar the noble gases and is the most stable configuration (in most cases). You can get free radicals floating around which are atoms or molecules that have unpaired electron but these radicals are very reactive and will quickly react to stabilise.
Another example of the octet rule breaking down would be phosphorous pentachloride (PCl5). Phosphorous has 5 electrons in its outer shell (one pair and three not in a pair). To get up to the stable 8 electrons it should logically share electrons with three chlorines and although PCl3 is a stable chemical often it is passed over. It is actually more favourable for that pair of electrons in the phosphorous outer shell to separate and then all 5 electrons to pair up with an electron from chlorine to form five covalent bonds leading to 10 electrons around the phosphorous. When a species is able to surpass the common 8 electron limit it is called hypervalent with being a PCl5 hypervalent molecule.
There have been some predictions about the where we should look for hypervalent molecules but those formed form atoms on the second row of the periodic table, lithium to neon, have been quite elusive. This investigation put a small amount of methyl fluoride (CH3F) in a large quantity of solid argon and bombarded the mixture with electrons. This produced hydrogen fluoride (HF) and hopefully the respective hydrogen fluoride anion (HF–). It was confirmed by infrared spectroscopy that the H – F was indeed present Now HF– demonstrates hypervalency as the fluorine (a group two element) has now got 9 electrons in its outer shell.
It was found that the bombardment electron energy and electron flux were not related linearly to the rate of production of HF–, it can therefore be concluded that its production is an indirect result of the electrons. Through photolysis (attempting to break the molecule with electromagnetic radiation) it was found that the bond dissociation energy is very similar to a dihalides. Further analysis of the charge densities showed that the fluorine was the more negative of the two but the characteristic of the bond is still more covalent than ionic. It is also believed that the effects on the argon matrix could reveal more detail about the nature of hypervalent bonding but the current theory is that the H atoms and F– ions are broken from CH3F independently but then associate again later to from the HF–.