Consider using accurate data from IUPAC

[Ellipse0934]

I was trying to compute the mass of BerylIodideTechnetogen (BITcH) but the accuracy was outside the expected bounds. Migrating to a floating point system and using accurate data from IUPAC will be helpful to all.

https://www.ciaaw.org/atomic-weights.htm is a good project from where we can pull data.

Also since the isotope distribution is an ever evolving question maybe we should should also change the format of this library to use csv rather than defining a dict in elements.py. I suggest each row be of the format:
atomic number, symbol, name, wt/[wt_1,frac_1, wt_2, frac_2]

Here we can choose wt to describe the final weight OR enter a isotope weight and fraction array. E.g [6.420, 0.69, 6.399, 0.31]