<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,Arial,Helvetica,sans-serif;" dir="ltr">
<pre><i>Hello,
if your system is isolated, i.e. it is a finite piece of graphene, and it is
small enough to simulate it entirely in DFT, than you have to use a single k-
point, and you do not need to build any supercell. I.e. you only need a cell
bug enough to keep the ribbon in it, and ensure that there is a bit of vacuum
on every side.
Than for both phonopy and phono3py calculation, you specify 1x1x1 as a
supercell size. This will still require a huge amount of DFT calculations, as
your cell includes many atoms.
Finally, I don't know how much sense it makes to compute thermal transport for
a nanometer sized finite system. Heat clearly transfer ballistically inside the
ribbon, as it is much smaller than the mean-free-path of any phonon. More in
general: heat is a macroscopic quantity, and what does it mean to have a
temperature gradient in a single molecule of carbon, formed by just 84 atoms?
--
Dr. Lorenzo Paulatto</i> <br><br><br>Dear Lorenzo,<br><br>As you say in the first paragraph, that is what i in first place tried. I calculated<br>a 1x1x1 supercell dimension, with a single gamma point (k_points {gamma}), and the band results<br>were absurd (it did not look at all alike the reference i am using). I deem it could be related to the fact<br>that phonon bands are supposed to be calculated in an infinite system, so with a 1x1x1 supercell that is not posible...<br><br>The idea i have is firstly perform a band calculation which results i know in advanced,<br>so i can try after a thermal calculation of a bigger isolated nanorribon, which is my real task. The point is that<br>the finite size of the nanoribbon, will play a role in thermal contuctivity as you suggest.<br><br>So, since is not a problem of k points, since i am using a single gamma point, what could be? <br>And, if i try to solve the phonon band problem with a 4x4x1 supercell, which k_point mesh would be convenient?<br><br>Once again, thank you for your time.<br><br>Pablo.<br></pre>
<br>
</div>
</body>
</html>