Neutron-neutron scattering can be described using the phenomenological Lagrangian $$\mathcal L_I=g \pi^0(x)\bar\psi_n(x)\gamma^5\psi_n(x). $$ Since the Lagrangian contains three fields and the process $n+n\to n+n$ involves four particles (let $p_1,k_1$ and $p_2,k_2$ be the momenta of the 'in' and 'out' pairs respectively), the terms in $g^2$ are the lowest that contribute to the scattering amplitude: the corresponding Feynman diagrams are
However, I have a small doubt. Why is there no diagram with $p_1$ and $k_1$ entering the same internal vertex?