[Pw_forum] Negative phonon frequencies in SiO_2

Гриша Гончаровский eiklm at mail.ru
Mon Jan 2 09:40:17 CET 2012


Dear QE users,

I attempt to calculate phonon dispersion in alpha-quartz and keep getting the wrong result. Optical branches look rather plausible, but acoustic branches include some negative frequencies, moreover, one branch seems to be negative in the whole Brillouin zone. I tried to use different pseudopotentials, to select different q-grids, ecutwfc and ecutrho values, but result remains the same.
My scf input:

&control
    calculation='scf',
    restart_mode='from_scratch',
    prefix='sio2',
    pseudo_dir = '/home/grysha/espresso/pseudo/',
    outdir='/home/grysha/espresso/tmp/'
 /
 &system
    ibrav=4,
    celldm(1)=9.289897331,
    celldm(3)=1.099552482,
    nat=  9, ntyp= 2,
    ecutwfc = 20.0
    ecutrho = 150.0
 /
 &electrons
    electron_maxstep=1000
    mixing_mode = 'plain'
    mixing_beta = 0.7
    conv_thr =  1.0d-8
 /
ATOMIC_SPECIES
 Si  28.086  Si.pz-vbc.UPF
 O   15.999  O.pz-rrkjus.UPF
ATOMIC_POSITIONS crystal
 Si   0.46990000   0.00000000   0.66666667
 Si   0.00000000   0.46990000   0.33333333
 Si  -0.46990000  -0.46990000   0.00000000
 O    0.41410000   0.26810000   0.78540000
 O   -0.26810000   0.14600000   0.45206667
 O   -0.14600000  -0.41410000   0.11873333
 O    0.26810000   0.41410000  -0.78540000
 O   -0.41410000  -0.14600000  -0.11873333
 O    0.14600000  -0.26810000  -0.45206667
K_POINTS automatic
4 4 4 0 0 0

Phonon input:

phonons of SiO_2
 &inputph
  tr2_ph=1.0d-12,
  prefix='sio2',
  ldisp=.true.,
  nq1=4, nq2=4, nq3=4
  amass(1)=28.086,
  amass(2)=15.999,
  outdir='/home/grysha/espresso/tmp',
  fildyn='sio2.dynFull',
 /

q2r input:

&input
   fildyn='sio2.dynFull', zasr='simple', flfrc='sio2444.fc'
 /


Trying to calculate dispersion e.g. along the Gamma --- A line with such input

&input
    asr='simple',  amass(1)=28.086, amass(2)=15.999,
    flfrc='sio2444.fc', flfrq='sio2.freq'
 /
 11
  0.0 0.0 0.000000  0.0
  0.0 0.0 0.0866025  0.0
  0.0 0.0 0.173205  0.0
  0.0 0.0 0.259808  0.0
  0.0 0.0 0.346410  0.0
  0.0 0.0 0.433013  0.0
  0.0 0.0 0.519615  0.0
  0.0 0.0 0.606218  0.0
  0.0 0.0 0.692820  0.0
  0.0 0.0 0.779423  0.0
  0.0 0.0 0.866025  0.0

I obtain in the sio2.freq file

&plot nbnd=  27, nks=  11 /
            0.000000  0.000000  0.000000
   -3.9029   -1.6909   -0.9482   55.4410   58.8549  184.0277
  232.6261  232.8523  328.4843  362.4152  372.6477  373.7721
  424.6636  425.2693  446.6598  518.9579  681.4918  682.2090
  775.2395  775.6317  778.7409 1081.8532 1082.0740 1096.3594
 1172.6020 1172.7318 1231.8798
            0.000000  0.000000  0.086602
  -27.8791  -24.0601   34.3374   57.7231   79.0905  185.5367
  222.9925  242.2063  317.0109  365.2996  371.2517  382.6987
  418.6292  430.0279  445.5397  521.3980  655.6902  706.8253
  771.8412  778.2792  779.1211 1079.7256 1084.3283 1096.2590
 1161.7487 1183.7450 1233.8392
            0.000000  0.000000  0.173205
  -42.4501   -9.1280   61.2822   73.4236  100.4741  191.4075
  214.2298  250.4044  293.9354  359.5278  385.7910  392.7004
  412.1738  433.0018  442.8980  527.3652  629.0604  729.1692
  768.1371  779.8632  780.1202 1077.7929 1086.6829 1095.9627
 1152.2769 1193.5893 1237.5794
            0.000000  0.000000  0.259808
  -52.8832   51.6657   77.3462   97.0335  116.7849  202.2120
  207.1785  255.6953  266.4743  356.9882  398.4107  401.5138
  405.7037  434.3569  440.0324  534.6626  602.8471  747.0288
  764.9138  780.6564  780.8533 1076.3172 1088.9413 1095.4574
 1144.7947 1202.0362 1238.6139
            0.000000  0.000000  0.346410
  -58.6717   74.3589   85.0705  128.2999  130.9468  201.2035
  208.4521  243.0812  255.8998  358.7639  398.8673  406.8712
  407.6452  435.0435  437.8321  542.7959  579.1644  758.8697
  762.7038  780.9131  781.1744 1075.3784 1091.0091 1094.6777
 1139.5788 1210.4478 1234.1293
            0.000000  0.000000  0.433013
  -60.8276   83.9302   87.2635  134.9317  167.8644  185.4667
  197.4218  245.0537  249.7366  362.0234  393.8809  410.1527
  410.7431  435.4336  436.7230  553.4463  560.5617  761.4133
  764.3324  781.0491  781.1413 1074.9821 1092.7449 1093.5520
 1137.1544 1220.0745 1225.2006
            0.000000  0.000000  0.519615
  -60.1379   80.5518   86.7839  132.6697  149.3933  198.7177
  200.3888  241.4125  253.5269  360.6371  395.7978  409.1896
  409.5822  435.2824  437.1175  547.6607  569.0894  761.4928
  762.8666  780.9771  781.1909 1075.1131 1091.9347 1094.1553
 1137.9753 1215.1177 1230.0095
            0.000000  0.000000  0.606218
  -56.3113   65.0439   82.0152  112.8013  123.3145  204.0657
  207.4619  252.7666  256.6986  357.3738  402.2872  403.2411
  404.9860  434.7431  438.8123  538.5747  590.4774  753.7985
  763.6404  780.8314  781.0603 1075.7777 1090.0095 1095.1065
 1141.8689 1206.1496 1237.0771
            0.000000  0.000000  0.692820
  -48.3077   33.4208   70.6393   84.0546  109.2574  196.4080
  210.5025  253.5698  280.4439  357.7446  392.5521  397.3795
  408.9672  433.8191  441.4193  530.9458  615.7606  738.7961
  766.4268  780.3237  780.5636 1076.9885 1087.8329 1095.7382
 1148.2613 1197.9403 1238.6959
            0.000000  0.000000  0.779423
  -35.4182  -25.8479   49.1690   64.6900   90.4259  187.7989
  218.3935  246.5521  306.2412  362.0995  378.5373  387.7484
  415.3698  431.7777  444.3207  524.1062  642.3589  718.4768
  769.9587  779.2468  779.5653 1078.7078 1085.5161 1096.1347
 1156.7679 1188.8888 1235.7641
            0.000000  0.000000  0.866025
  -17.2456  -13.7809   17.5251   53.9027   67.3275  184.3786
  227.8145  237.5840  325.1353  365.1032  368.9991  377.8392
  421.8521  427.7529  446.3640  519.5984  668.8173  694.5999
  773.6796  776.9999  778.8399 1080.8109 1083.1472 1096.3340
 1167.0735 1178.3041 1232.4106

Single-phonon calculation at the A point also gives some negative frequencies:

     omega(  1 -  1) =        -96.6  [cm-1]   --> A                  
     omega(  2 -  2) =        -69.9  [cm-1]   --> A                  
     omega(  3 -  3) =        -52.9  [cm-1]   --> A                  
     omega(  4 -  4) =        -25.6  [cm-1]   --> A                  
     omega(  5 -  5) =        -13.8  [cm-1]   --> A                  
     omega(  6 -  6) =        169.3  [cm-1]   --> A                  
     omega(  7 -  7) =        218.6  [cm-1]   --> A                  
     omega(  8 -  8) =        227.9  [cm-1]   --> A                  
     omega(  9 -  9) =        318.1  [cm-1]   --> A                  
     omega( 10 - 10) =        360.1  [cm-1]   --> A                  
     omega( 11 - 11) =        362.7  [cm-1]   --> A                  
     omega( 12 - 12) =        371.0  [cm-1]   --> A                  
     omega( 13 - 13) =        412.5  [cm-1]   --> A                  
     omega( 14 - 14) =        419.1  [cm-1]   --> A                  
     omega( 15 - 15) =        441.1  [cm-1]   --> A                  
     omega( 16 - 16) =        524.0  [cm-1]   --> A                  
     omega( 17 - 17) =        668.1  [cm-1]   --> A                  
     omega( 18 - 18) =        694.6  [cm-1]   --> A                  
     omega( 19 - 19) =        773.0  [cm-1]   --> A                  
     omega( 20 - 20) =        776.0  [cm-1]   --> A                  
     omega( 21 - 21) =        779.8  [cm-1]   --> A                  
     omega( 22 - 22) =       1078.2  [cm-1]   --> A                  
     omega( 23 - 23) =       1080.8  [cm-1]   --> A                  
     omega( 24 - 24) =       1094.3  [cm-1]   --> A                  
     omega( 25 - 25) =       1165.6  [cm-1]   --> A                  
     omega( 26 - 26) =       1175.8  [cm-1]   --> A                  
     omega( 27 - 27) =       1244.2  [cm-1]   --> A       

but I wonder that positive frequencies in these two cases are not significantly different.
Where should I dig to avoid instability? I would be grateful for any suggestion!

Regards

Mikhail Goncharovski


More information about the Pw_forum mailing list