Hello, i'd like to study Ehrenfest force (both vector field and |F|). Multiwfn has a stress tensor, but is it possible to calculate the Ehrenfest force as its negative divergence? Does Multiwfn have such an option? I couldn't find it in the manual. (I understand that the Ehrenfest force derived from the stress tensor has problems due to the uncertainty of the stress tensor https://doi.org/10.1063/5.0174905 , but I'd still like to see it).

Dear Alexey,

Mortera-Carbonell et al. propose an alternative to the stress-tensor approach, using the exchange–correlation Ehrenfest force field, Fₓc(r), to define and analyze the Ehrenfest force. https://doi.org/10.1021/acs.jctc.4c00890
---
About vector field representation, perhaps adapting the scripts for gnuplot, to read the .txt (as Professor Tian Lu mentioned), could be useful. See: http://gnuplotting.org/vector-field-fro … index.html

call orbderv(5,1,nmo,x,y,z,wfnval,grad,hess,tens3)

F=0D0
do imo=1,nmo
    occ = MOocc(imo)
    if (occ==0D0) cycle
    phi = wfnval(imo)
    g(:) = grad(:,imo)
    h(:,:) = hess(:,:,imo)
    t3(:,:,:) = tens3(:,:,:,imo)

    do i=1,3
        sumj = 0D0
        do j=1,3
           
            sumj = sumj + 2D0*(h(j,i)*g(j) + g(i)*h(j,j) - t3(j,i,j)*phi - h(i,j)*g(j))
        end do
       
        F(i) = F(i) - (-0.25D0*occ*sumj)
    end do
end do

ehr = sqrt(F(1)**2 + F(2)**2 + F(3)**2)
end function

could you please check is it correct?

Ehrenfest force has not been explicitly implemented in Multiwfn. But you can easily extend the source code Multiwfn to realize this aim. Stress tensor is calculated by "subroutine stress_tensor" in function.f90. Basic knowledge about modifying source code of Multiwfn can be found in http://sobereva.com/multiwfn/res/Modify … ltiwfn.pdf