Dear Peter,
The mbj calculations were well converged for our carbon-based compound with week van der Waals interactions among each clusters in the unit cell. We doped impurity endohedrally into the carbon-based compound. Therefore the symmetry of the system decreased, and the number of non-equivalent atoms increased. However, the interstitial region has not changed substantially. It almost remains the same as before manipulating impurities. For the new structure file the mbj calculation stops with the same error as before even employing the new BRJ subroutine. Could reducing the TOL from 1D-6 to 1D-7 fix the problem? Or should we change the IF condition on iint?