Add setter for ionization potential

[wtbarnes]

When making an Ion, it is possible to set the abundance through a keyword argument. This keyword argument can be either the dataset name or an actual value. The ionization potential should work the same and should have similar setter/getter logic to how the abundance property works.

[wtbarnes]

This should be done in tandem with #323

[jwreep]

Are there significantly different values or precisions from different atomic databases? I guess I'm wondering why a setter is needed for ionization potential, as it should just be a fundamental property of an ion.