<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<style type="text/css" style="display:none;"><!-- P {margin-top:0;margin-bottom:0;} --></style>
</head>
<body dir="ltr">
<div id="divtagdefaultwrapper" style="font-size:12pt;color:#000000;font-family:Calibri,Helvetica,sans-serif;" dir="ltr">
<p>Hi,</p>
<p><br>
</p>
<p>It might be because you have not geometry optimized your system completely. </p>
<p>I recommend you to run a relax calculation with following parameters changed, they might seem overkill but they work:</p>
<p><br>
</p>
<p>in CONTROL </p>
<p></p>
<div> etot_conv_thr = 1.0D-6,</div>
<div> forc_conv_thr = 1.0D-5,</div>
<div>in SYSTEM</div>
<div>
<div> nosym = .false.,</div>
<div>in ELECTRON</div>
</div>
<div>
<div> diagonalization = 'cg',</div>
<div> conv_thr = 1D-8,</div>
</div>
<div><br>
</div>
<div>With a dense k-point grid, however as the symmetry is broken the k-point grid might be dense - you need to check.</div>
<div>Then a vc-relax run with same parameters (and update atomic positions), but set cell_dofree as per your requirement.</div>
<div>You also need to check your structure after the geometry optimization, there might be a change in angle from say 90 degrees to 89.9 degrees, this will also give you negative or imaginary frequences.</div>
<div><br>
</div>
<div>I also recommend you to get the crystal structure data form: <span>materialsproject.org </span></div>
<div><span>They are generally optimized or with very fewer forces. </span></div>
<div><span><br>
</span></div>
<div><span>Regards,</span></div>
<div><span>Ghadiyali Mohammed Kader</span></div>
<div><span>Dept. of Physics,</span></div>
<div><span>University of Mumbai.</span></div>
<p></p>
</div>
</body>
</html>