Analogous to Mott insulators we also have to consider superexchange in charge-transfer insulators. One contribution is similar to the Mott case: the hopping of a d electron from one transition metal site to another and then back the same way. This process can be written as
d i n p 6 d j n → d i n p 5 d j n + 1 → d i n − 1 p 6 d j n + 1 → d i n p 5 d j n + 1 → d i n p 6 d j n {\displaystyle d_{i}^{n}p^{6}d_{j}^{n}\rightarrow d_{i}^{n}p^{5}d_{j}^{n+1}\rightarrow d_{i}^{n-1}p^{6}d_{j}^{n+1}\rightarrow d_{i}^{n}p^{5}d_{j}^{n+1}\rightarrow d_{i}^{n}p^{6}d_{j}^{n}} .
This will result in an antiferromagnetic exchange (for nondegenerate d levels) with an exchange constant J = J d d {\displaystyle J=J_{dd}} .
J d d = 2 t d d 2 U d d = 2 t p d 4 Δ C T 2 U d d {\displaystyle J_{dd}={\frac {2t_{dd}^{2}}{U_{dd}}}={\cfrac {2t_{pd}^{4}}{\Delta _{CT}^{2}U_{dd}}}}
In the charge-transfer insulator case
d i n p 6 d j n → d i n p 5 d j n + 1 → d i n + 1 p 4 d j n + 1 → d i n + 1 p 5 d j n → d i n p 6 d j n {\displaystyle d_{i}^{n}p^{6}d_{j}^{n}\rightarrow d_{i}^{n}p^{5}d_{j}^{n+1}\rightarrow d_{i}^{n+1}p^{4}d_{j}^{n+1}\rightarrow d_{i}^{n+1}p^{5}d_{j}^{n}\rightarrow d_{i}^{n}p^{6}d_{j}^{n}} .
This process also yields an antiferromagnetic exchange J p d {\displaystyle J_{pd}} :
J p d = 4 t p d 4 Δ C T 2 ⋅ ( 2 Δ C T + U p p ) {\displaystyle J_{pd}={\cfrac {4t_{pd}^{4}}{\Delta _{CT}^{2}\cdot \left(2\Delta _{CT}+U_{pp}\right)}}}
The difference between these two possibilities is the intermediate state, which has one ligand hole for the first exchange ( p 6 → p 5 {\displaystyle p^{6}\rightarrow p^{5}} ) and two for the second ( p 6 → p 4 {\displaystyle p^{6}\rightarrow p^{4}} ).
The total exchange energy is the sum of both contributions:
J t o t a l = 2 t p d 4 Δ C T 2 ⋅ ( 1 U d d + 1 Δ C T + 1 2 U p p ) {\displaystyle J_{total}={\cfrac {2t_{pd}^{4}}{\Delta _{CT}^{2}}}\cdot \left({\cfrac {1}{U_{dd}}}+{\cfrac {1}{\Delta _{CT}+{\tfrac {1}{2}}U_{pp}}}\right)} .
Depending on the ratio of U d d and ( Δ C T + 1 2 U p p ) {\displaystyle U_{dd}{\text{ and }}\left(\Delta _{CT}+{\tfrac {1}{2}}U_{pp}\right)} , the process is dominated by one of the terms and thus the resulting state is either Mott-Hubbard or charge-transfer insulating.3
Khomskii, Daniel I. (2014). Transition Metal Compounds. Cambridge: Cambridge University Press. doi:10.1017/cbo9781139096782. ISBN 978-1-107-02017-7. 978-1-107-02017-7 ↩
Zaanen, J.; Sawatzky, G. A.; Allen, J. W. (1985-07-22). "Band gaps and electronic structure of transition-metal compounds". Physical Review Letters. 55 (4): 418–421. Bibcode:1985PhRvL..55..418Z. doi:10.1103/PhysRevLett.55.418. hdl:1887/5216. PMID 10032345. https://link.aps.org/doi/10.1103/PhysRevLett.55.418 ↩