We may be perceiving things differently, but used once is one thing, as opposed to used once one way AND used once another way, as well as it has to be both at the same time. That kind of representation is representative of procedural thinking as opposed to set-based logic, and it largely eliminates the benefits of using an RDBMS. It also doesn't take into account the more complex set of rules usually associated with "creating an isotope", which are usually bound by the inherent valence properties of the various atoms involved. Of course, what the poster actually means by the word "isotope" isn't addressed, and could certainly have a sizable impact on what's actually needed. Just for grins and giggles, I created an ATOMS table on my local machine and inserted some of the basic elements and their valence values from a copy of the periodic table I found online, that is fairly detailed. I then tried to think a little bit more procedurally, with the Valence values being a comma-separated string. Of course, I couldn't help but use DelimitedSplit8K to table-ize that information, and then a cartesian product to map a single positive valence element to a single negative valence element, and see how many unique combinations I could come up with where the sum of the valence values was equal to zero. It was an interesting exercise, but nothing I did there even remotely resembled what this poster seems to want. Without a good "why" ... I just can't get excited about trying to help the poster any further.