Dear all Quantum Espresso users and developers,<br><br>I am not able to reproduce the zone cente phonons in ZrO2 calculated<br>in PRL 78, 21, 1997.<br><br>My results, without LO-TO splitting(can be calculated, once accoustical mode=0 cm-1 ) are:<br>

<br>T1u: 14.27 cm-1<span style="color: rgb(255, 102, 0);"> <---------not zero</span><br>T1u: 219.6 cm-1<br>T2g: 550.35cm-1 <br><br> I am not able to make the accoustical mode=0 cm-1.<br> Although  the born charges and dielectric constants are <br>

 ok with relative error~1%.<br><br><br><span style="color: rgb(255, 0, 0);">Input files:-----------------------------------------------</span><br><br><span style="color: rgb(51, 51, 255);">scf file:</span><br>&control<br>

    calculation='scf',             ! relax,self consistent calc.<br>    verbosity='high',<br>    restart_mode='from_scratch',   ! 'from_scratch',<br>    prefix='ZrO2',<br>    wf_collect=.true.,                    ! collect wave func in ./ZrO2.save<br>

    tstress=.true.,<br>    tprnfor=.true.,<br>    nstep=100,                             ! default=50<br>    dt=20,                                      !10,      ! default=20<br>    etot_conv_thr=1.0D-5,          ! default=1.0D-4<br>

    forc_conv_thr=1.0D-5,          !1.0D-4,  default=1.0D-3<br>    pseudo_dir = './',<br>    outdir='./',<br> /<br><br> &system<br>    ibrav = 2,<br>    celldm(1)=9.67                 <br>    nat=3, ntyp=2,             <br>

    ! nbnd=                              ! can calculate itself<br>    ecutwfc=30,                      !50,    !in Ry, no default, initially=60<br>    ecutrho=350,                    !350 !300, ! 8*50=400<br>    !    nr1=14,<br>

    !    nr2=14,<br>    !    nr3=14,<br>    !    nr1s=14,<br>    !    nr2s=14,<br>    !    nr3s=14,<br>    !    nosym=.true.,<br>    nspin=1,                           ! 1,2,4, "noncolin=.true."<br>    !    input_dft=                    ! read from PP files.<br>

 /<br><br> &electrons<br>    mixing_beta = 0.2,              ! 0.7<br>    conv_thr = 1.0d-12,            ! default=1.0d-06, ! 1.0d-12,1.0d-8<br>    diagonalization=cg              ! 'cg',!'cg', 'david', ! <br>

<br> /<br><br>ATOMIC_SPECIES <br> Zr   0.0000 Zr.pbe-nsp-van.UPF<br> O    0.0000 O.pbe-van_ak.UPF<br><br>ATOMIC_POSITIONS (crystal)<br>Zr 0.0  0.0  0.0<br>O  0.25 0.25 0.25<br>O  0.75 0.75 0.75       <br><br>K_POINTS (automatic)<br>

3 3 3 0 0 0                             !8 8 8 0 0 0                             <br><br><span style="color: rgb(51, 51, 255);">ph file</span>:----------------------------<br> &inputph<br>  tr2_ph=1.0d-14,             !1.0d-14,     ! threshold for self consistency=1.0d-12(default)<br>

  alpha_mix(1)=0.2,         ! default=0.7<br>  prefix='ZrO2',<br>  iverbosity=1,<br>  epsil=.true.,       <br>  trans=.true.,                   ! default=true ,<br>  asr=.true.,<br>  zue=.true.,                     ! eff charges<br>

  amass(1)=91.22400,<br>  amass(2)=15.99940,<br>  outdir='./',<br>  fildyn='ZrO2.dynG',       ! file where the dyn matrix is written<br><br> /<br>0.0 0.0 0.0<br><br>Effect of  convergences (plane waves, k point sampling, augmentation functions)<br>

is checked. Forces are zero  and press~1 kbar.<br><br>Bulk mod~233 GPa, expt~267Gpa => error~13 %.<br><br>Also,in above paper LDA pp was used, but i used USPP. <br><br>So in nutshell:---<br>        i think USPP has to be modified or norm conserving pp should be used.<br>

 <br>I need your comments over this.<br><br>Thanks!!<br><br>With Regards,<br>Sonu Kumar<br><br>Phd Student<br>Physics Department<br>Indian Institute of Technology <br>Delhi-110016, India<br>web:-<a href="http://www.iitd.ac.in/">http://www.iitd.ac.in/</a><br>