mp-776632_Li in the MP battery explorer predicts a voltage of 9.739 V for the redox couple Mn4(P2O7)3 / LiMn4(P2O7)3. So there must be an error somewhere.
Can you clarify a bit what you think is wrong here? There are plenty of poor redox couples with either unphysically high voltage or even negative voltage (!) in the battery explorer. This really just indicates that particular material isn’t physically realizable as a cathode:
from mp_api.client import MPRester
with MPRester() as mpr:
li_mn_p_o = mpr.materials.insertion_electrodes.search(working_ion='Li',elements=['Mn','P','O'])
voltages = sorted(doc.average_voltage for doc in li_mn_p_o)
print(f"Min voltage = {voltages[0]:.2f} V; Max voltage = {voltages[-1]:.2f} V")
>>> Min voltage = -2.24 V; Max voltage = 9.74 V
You are right. I underestimated the effect of the energy above hull of the charged state that looked borderline acceptable passing an eV/atom filter, but due to the large formula unit contributes a large part of the calculated voltage.