<div dir="ltr">Hello qe users,<div>I am performing the calculation of the relaxed structure of a supercell of Fe(CO)5 (Z=4, C2/c). </div><div>After the first scf cycle (CG style diagonalization) the new energy of the system is calculated.</div><div>In the next cycle the scf does not converge after 1000 steps.</div><div>I believe this comes from incorrect input.<br></div><div>Were could the error in the input be?</div><div>The coordinates of the supercell I have extracted from the supplementary material of the article <a href="https://doi.org/10.1038/srep15139" target="_blank">https://doi.org/10.1038/srep15139</a>.</div><div>Here are some parts of the output:</div><div>Self-consistent Calculation<br><br>   iteration #  1   ecut=   80.00 Ry   beta= 0.70<br>   CG style diagonalization<br>   ethr =  1.00E-02,  avg # of iterations =  3.0<br><br>   total cpu time spent up to now is   1803.9 secs<br></div><div><br></div><div>  ......</div><div><br></div><div>  iteration # 32   ecut=   80.00 Ry   beta= 0.70</div>   CG style diagonalization<br>   ethr =  2.04E-09,  avg # of iterations =  3.1<br><br>   total cpu time spent up to now is   43842.1 secs<br><br>   total energy        =  -1850.88412269 Ry<br>   Harris-Foulkes estimate  =  -1850.88412271 Ry<br>   estimated scf accuracy   <    0.00000137 Ry<br><br>   iteration # 33   ecut=   80.00 Ry   beta= 0.70<br>   CG style diagonalization<br>   ethr =  5.19E-10,  avg # of iterations =  3.1<br><br>   total cpu time spent up to now is   45086.7 secs<br><br>   End of self-consistent calculation<div><br></div><div>   ...</div><div><br></div><div>Total force =   27.930798   Total SCF correction =   0.002498<br><br><br>   Computing stress (Cartesian axis) and pressure<br><br>   Message from routine stres:<br>   noncollinear stress + GGA not implemented<br><br>   BFGS Geometry Optimization<br><br>   number of scf cycles   =  1<br>   number of bfgs steps   =  0<br><br>   energy  new       =  -1850.8841224519 Ry<br><br>   new trust radius     =    0.5000000000 bohr<br>   new conv_thr       =    0.0000010000 Ry<br></div><div><br></div><div>  ...</div><div>  Writing output data file pwscf.save/</div>   Message from routine punch:<br>   wavefunctions written to file<br>   NEW-OLD atomic charge density approx. for the potential<br><br>   total cpu time spent up to now is   45155.5 secs<br><br>   Self-consistent Calculation<br><br>   iteration #  1   ecut=   80.00 Ry   beta= 0.70<br>   CG style diagonalization<br>   c_bands:  4 eigenvalues not converged<br>   ethr =  1.00E-06,  avg # of iterations = 27.7<br><div>   ...</div><div>   iteration #999   ecut=   80.00 Ry   beta= 0.70</div>   CG style diagonalization<br>   c_bands:  4 eigenvalues not converged<br>   ethr =  1.04E-08,  avg # of iterations =  8.9<br><br>   total cpu time spent up to now is  2033318.7 secs<br><br>   total energy        =  -1859.84580191 Ry<br>   Harris-Foulkes estimate  =  -1859.84718501 Ry<br>   estimated scf accuracy   <    0.25486994 Ry<br><br>   iteration #***   ecut=   80.00 Ry   beta= 0.70<br>   CG style diagonalization<br>   c_bands:  2 eigenvalues not converged<br>   ethr =  1.04E-08,  avg # of iterations =  9.4<br><br>   total cpu time spent up to now is  2036333.6 secs<br><br>   total energy        =  -1859.85310299 Ry<br>   Harris-Foulkes estimate  =  -1859.84865768 Ry<br>   estimated scf accuracy   <    0.25711820 Ry<br><br>   End of self-consistent calculation<br><br>   convergence NOT achieved after *** iterations: stopping<div><br></div><div>   I also attach the input. </div><div>  Thanks </div></div>