Enhancement - temperature dependent K_w value based on updated IAPWS formulation

[rkingsbury]

pyEQL uses a third-party IAPWS library to compute bulk water properties. There appears to be a more modern implementation here. Evaluate whether this library would offer faster performance or reduce dependency count.

EDIT: the focus of this issue has shifted - see next comment.

[rkingsbury]

Update - the more modern IAPWS library does not implement all the properties we need in pyEQL, so is not worth pursuing further.

Specifically, Solution only needs the following from the water model:

• density
• dynamic viscosity
• dielectric constant (in one place)

The IAPWS also has a formulation for the water dissociation constant K_w, which Solution currently treats as a constant.

A nice future enhancement would be to use the IAPWS formulations to make K_W temperature (and pressure) dependent using the latest (2024) formulation

Considering that we require so few properties, it may be worth considering reverting to in-house implementations of the relevant equations (which were part of pyEQL prior to version 0.6.0.

[rkingsbury]

Considering that the new equation for K_W requires density as input, this creates further incentive to re-implement the density model alongside the K_W model.

We should also cache the water density on __init__ to improve performance.