Update Ecker2015 solid-phase diffusivity functions to more closely match the paper

[DrSOKane]

The solid-phase diffusivity functions in Ecker2015 and Ecker2015_graphite_halfcell are updated to better match the functions in the original paper. A new citation Yuan2023, for which these parameters were used, is also added.

Advantages:

• The new functions reproduce phase separation behaviour in a similar way to the Cahn-Hilliard equations
• With them, the results in Yuan2023 can be reproduced.

Disadvantages:

• This level of physical accuracy comes at a noticeable computational cost, comparable with that of ORegan2022
• The results in Richardson2019 are no longer reproduced. However, that paper still made a major contribution towards our understanding of the Ecker2015 parameters and is therefore still cited.

[rtimms]

My preference would be to add a new Yuan2023 parameter set so that the Ecker results don't change from version to version.

[valentinsulzer]

What's the formula for D_s_n and D_s_p actually used by Ecker? I can't find it in Part I or Part II.

[kratman]

This should probably be under only "breaking changes" or "features" instead of both

[kratman]

@DrSOKane Is this one fixed by this change as well?

#2152

[DrSOKane]

What's the formula for D_s_n and D_s_p actually used by Ecker? I can't find it in Part I or Part II.

They don't give a formula, only data. It's possible that they used an interpolation between the data points. LiionDB has this data (I put it there!) so an alternative approach would be to use that.

[kratman]

@DrSOKane This PR was marked as merged because the commits were also present in #4168