Hi Lin Xiao, You are right, it means hartree atomic units. Below is an explanation of how to get from relaxation time in seconds to Ha.
The Drude dielectric function looks like: epsilon = 1 - wp^2 / [ w ( w + i/tau ) ] Where tau is the relaxation time (average time between collisions of an electron with impurities/defects/ions) Usually the relaxation time is specified in seconds. Siesta uses the damping rate or scattering rate (inverse relaxation time): gamma = 1 / tau where gamma is in Hartree units. To convert from inverse seconds to hartrees: electron_volts = planck/ ( elec_charge * time ) = 6.626e-34 / ( 1.606e-19 * t ) = J s / ( J/eV * s) = eV then divide by 27.2114 to get Hartrees. So a relaxation time of 3e-14 seconds (which is the value for gold at 300K) works out to be 0.137856 eV or 0.005066 Ha Cheers, Marty On 14 January 2011 08:35, lin xiao <[email protected]> wrote: > Dear Marty, > There is a small problem remaining. > When i do as you said, > [yangxu@WimaxServer 20110113-10]$ /opt/siesta-3.0-rc2/bin/optical > Do you want to include a Drude term? > This is typically needed for metals > if yes: enter 1, if no: enter 0 > 1 > We need an empirical relaxation time > Enter the invers in Ha units > here,what is the meaning of "Ha units"?Does 'Ha units' means HATREE,ie > atomic units,or something else? > BEST WISHES! > xiaolin, from ZJU > 2011/1/14 lin xiao <[email protected]> >> >> Dear Marty, >> I appreciate your explaining.It really helps a lot. >> Thx again. >> xiaolin, from zju >> >> 2011/1/14 Marty Blaber <[email protected]> >>> >>> Hi Lin Xiao, >>> >>> At the command line, you must type only: >>> >>> optical >>> >>> without the <e2.dat >>> the < operator has the effect of pasting the file into the standard >>> input stream of the program. This is where you were supposed to type >>> in the "1". >>> Now, optical will ask the question: >>> > Do you want to include a Drude term? >>> > This is typically needed for metals >>> > if yes: enter 1, if no: enter 0 >>> and you may enter 1. >>> >>> Cheers, >>> Marty >>> >>> >>> >>> On 12 January 2011 21:52, lin xiao <[email protected]> wrote: >>> > Dears, >>> > I'm using the optical model of siesta to calculate graphene.Here is a >>> > question: >>> > [guozhendong@yangxu-433 20110112-2]$ /opt/siesta-3.0-rc2/bin/optical >>> > <e2.dat >>> > Do you want to include a Drude term? >>> > This is typically needed for metals >>> > if yes: enter 1, if no: enter 0 >>> > [guozhendong@yangxu-433 20110112-2]$ >>> > Then whatever i enter,for instance 1 ,the system answer is: >>> > 1: Command not found. >>> > What should i do if i want to include the Drude term? >>> > Thx! >>> > >>> > >> > >
