Comment:
ADESH has been extremely useful. I'm starting to get the hang of the package and am really impressed with its capabilities.
Comment:
ADESH has been extremely useful. I'm starting to get the hang of the
package and am
really impressed with its capabilities.
Well, thank you very much.
Question
How to make ZrO2 (Zirconia) cell with ADESH?
Is there a way to calculate energy of this cell once it is made?
Answer
You can use following steps to make ZrO2 cell on ADESH.
{U.Cell [ User defined
[ 0, 0, 0 ]
[ 0.5, 0.5, 0 ]
[ 0, 0.5, 0.5 ]
[ 0.5, 0, 0.5 ]
[ 0.25, 0.25, 0.25]
[ 0.75, 0.75, 0.75]
[ 0.25, 0.25, 0.75]
[ 0.25, 0.75, 0.25]
[ 0.75, 0.25, 0.25]
[ 0.75, 0.75, 0.25]
[ 0.25, 0.75, 0.75]
[ 0.75, 0.25, 0.75] ] }
{AlloY [ Systematic
[ Lattice Constant X: 5.152]
[ Lattice Constant Y: 5.152]
[ Lattice Constant Z: 5.152]
[ 0, 0, 0 ] Zr
[ 1/2, 0, 1/2] Zr
[ 1/2, 1/2, 0] Zr
[ 0, 1/2, 1/2] Zr
[ 1/4, 1/4, 1/4] O
[ 1/4, 1/4, 3/4] O
[ 1/4, 3/4, 1/4] O
[ 3/4, 1/4, 1/4] O
[ 3/4, 1/4, 3/4] O
[ 1/4, 3/4, 3/4] O
[ 3/4, 3/4, 1/4] O
[ 3/4, 3/4, 3/4] O ] }
{ Compute [ Make computational cell ] }
The parameters for ZrO2 potential:
Input these in the file Element.dat
{Compute [ Adjust IAP parameters
[ Modify current settings
[ O]
[ IONIC]
[ Cutoff: R**2: 16.0]
[ Sigma: 5435]
[ n = 10]
[ Charge: -2]
[ Zr ]
[ IONIC]
[ Cutoff: R**2: 16.0]
[ Sigma: 5435]
[ n = 10]
[ Charge: 4] ] }
[ Save to a file ....]
(Give the name 'element.dat' if you want to alter it.
Save the old one in some other name. Element.dat is the
default data file read by ADESH.)
Question
I am curious how you obtained the Zr and O potential values you sent to me
are these from a journal article? Also, what is the minimum cell size that
is appropriate for the zirconia energy calculations in ADESH? Is there
a critical number of atoms that must be in the cell?
Answer
ZrO2 potential parameters were obtained by fitting them to the lattice constant. If you use a different lattice constant, they would be different. They are not from a journal article. We design them. The parameters and potential functions of ionic crystals are still being tested. Usually we also fit the parameters to cohesive energy and bulk modulus of the material. If you know any reliable values for ZrO2, please let us know. Other experimental data such as vacancy formation or migration energy are also used to corroborate the parameters.
Another way to corroborate is to compare the calculated configurations of Gbs, dislocations, interfaces etc. with the HRTEM pictures. Sometimes it is not known whether the configurations seen in HRTEM micrographs are equilibrium structures. However, it is one way. The migrations of vacancies etc. can be tracked and time frames can be estimated with comparative energy values from ADESH. These estimates can also be compared with experimental values.
The number of atoms in the cell depend on the boundary conditions you use. For a rigid boundary condition: the size of the cell should be bigger by at least the cutoff radius than the size of the cell you would relax. (For rigid boundaries, you keep outer part of the cell fixed and relax the interior). Periodic boundary conditions: Cell size of the basic cell should be at least as large as the cutoff radius. Free boundary conditions: Use the subset option to keep the relaxing cell size larger than the actual cell size. Example: If a +Y surface is free surface, and the cell extends to 10 Angstroms along +Y, make the relaxing size to be +15 along +Y. In any case, the minimum size of the cell should be at least as large as the cutoff radius. The maximum number of atoms is 3000 in ADESH.
Question
We have deposited a zirconia film with (100) normal orientation
on a (111) zirconia single crystal. How would I make the cell for the (111)
substrate to model this interface? There seems to be a problem in assigning a
cell because there are obviously no 3 orthogonal directions to assign to define
the (111) substrate.
Answer
Use a two step procedure here.
(1) Make a zirconia cell with X (-1 1 0), Y (1 1 1), Z (1 1 -2) orientations. Use the M.Index option. Use {suBset [ Size of the cell < Cubic > ] } and make the size along +Y to be -0.1 . Then make the cell by { Compute [ Make cell ... ] }. Save it as "ZrO2SUB.bef ". ( You can do this by the registered copy).
(2) Then make the ZrO2 cell with X ( 1 0 0 ), Y (0 1 0), Z (0 0 1) (default).
Size along -Y would be -0.1 and +Y to be 10.0.
Save this one as ZrO2film.bef.
Then read them back in. Say yes when ADESH asks
" Do you want to append ..." for the second file.
(Or you can make a second cell and append it to first one when
you make it.
When you make the second cell, say "Yes" to 'Do you want to append?'.
This will work with the demo version).
Hence appending the dis-similar materials for substrate and film
can be done in two ways. Both times you make them
separately and then append them when you make the second one or
save them first and them read them in and append that time.
Example 10 in the manual uses similar procedure to put
a diamond film on top of a Silicon substrate. You need to
'fit' the film on top of the substrate by giving appropriate
coherent strain and adjusting the Y distances between
the film and substrate. Also you may need to use the
translatory movement to move the film against substrate
in horizontal directions to adjust the configuration
across the interface.
Question
I am concerned about the atomic radii I should be using to model zirconia. The
default radii for Zr and O are 1.615 and 1.4 angstroms, respectively. If
one consults the Shannon-Prewitt radii for Zr4+ and O2- for the coordinations
present in zirconia, the radii are 0.98 and 1.28 angstroms, respectively.
One would observe different results along the film/substrate lattice using the
differing sets of radii, especially when trying to model coincident site lattices.
My question is: are the IAP parameters you gave me for Zr and O only appropriate
for the default radii, or are they general
for any desired atomic size (within reason)?
Answer
ADESH uses the radii values from "Element.dat" for plotting (graphics)
puposes only.
IAP parameters are fitted to the lattice constant. Not to the radii
of atoms (ofcourse usually the radii depend on the lattice constant).
If you wish to use a different lattice constant, let us know.
We will have to re-calculate the parameters. Otherwise, simply change
the radii in "Element.dat" and when you plot atoms, you will get
appropriate sizes. It won't affect the parameters or energy calculations.
Question
Can I control which atoms terminate the free surface (say of a
substrate). I tried creating different cell sizes and wound up with
the same terminating atom species. Does ADESH allow one to terminate the
substrate with either Zr or O or
is one automatically set based on energetic considerations?
Answer
You can terminate the substrate with Zr or O atoms. We created two files (ZrO2a.bef & ZrO2b.bef). They terminate +Y plane with different atom planes. You can get this by changing the cell size in 'subset' before making the cell.
Try: (1) -5 to +6 for X, Y, Z (ZrO2b.bef)
(2) -5 to +6 for X, Z and -5 to +5 for Y (ZrO2a.bef)
These files are small (<100 atoms).
Send a note if
you have problems making them. We will send them to you.
May be you can download them through FTP.
Question
I am running ADESH in Windows 95 and am marking and pasting the
ADESH graphical
output directly into Microsoft Word. When I print on a standard inkjet
printer, it is
difficult to distiguish the different atoms because the grayscales are so
similar. I can make
the ADESH output a bitmap in Corel PhotoPaint and manually change the colors
of the atoms, but
that takes a lot of time. Have you come up with any procedures to
facilitate printing the
graphical output instead of just text output of atomic positions?
Answer
Some colors give a better gray scale contrast.
We change the colours with 'Word Perfect presentations'
just like you are using 'Corel photopoint'.
Color inkjet printer sometimes makes a difference.
Our registered version allows a screen dump to an HP
compatible printer as a black and white copy.
You can run ADESH in Windows 95 and use its copy and paste
facility to transfer the image into another application running
under windows (e. g. Word, Word Perfect, Paint etc.).
Steps:
Answer
If the films are non epitaxial, the coherent strain need not
be incorporated.
After making the substrate and film (separately with different
Miller Indices), put them together (Example 10 will help).
Rigid displacement of film against the substrate
can be tried after this. Use suBset option to move only the
film. Translational displacement along X, Y, Z as well as
rotational displacement of the film is important.
Monitor the energy at each step. The energy we are
calculating is internal energy. Please be aware that when you choose a
part of the whole cell to move (film only), after the movement,
some atoms from film may go out of the subset. They won't
move next time. So Monitor the subset. Also to compare
the energies, calculate energies of atoms from some central
part of the cell (away from free surfaces: distance
equal to cutoff distance of the potential function). Also
try to get equal number of atoms from film and substrate in
the Energy Subset when you calculate energy. Normalise the energy
to say 100 atoms so you can compare properly.
If you are simulating the film that is 5000 A thick, hold a top
few layers of the film rigid. If you are simulating layer by layer growth,
you may relax full film. The best strategy is to corroborate the
calculated configurations with experimental ones whenever you can.
Also keep in mind that the experimental configurations are not
always at equilibrium. Especially with thin films, some non-equilibrium
configurations tend to be relatively stable (local energy minima).
ADESH simulates all range of these configurations as relaxation
takes place. ADESH works as a good investigative tool when
used along with the experimental findings.
Question
I'm trying to model another material we had been sputtering:
La(1-x)CaxMnO3.
The compound
is composed of La(+3)Mn(+3)O3 with 33% Ca(+2)Mn(+4)O3. The material has the
perovskite
crystal structure with a lattice parameter of 3.903 Angstroms, according to
theta-2theta
X-ray diffraction. I've been trying to model the parent LaMnO3 structure
before trying
to get to the 33% Ca substitution case. I've been using the Shannon-Prewitt
radii
for La3+ with CN=12 (1.50 angstroms), Ca2+ with CN=12 (1.48 angstroms), O2-
with CN=6 1.26 angstroms)
and Mn4+ with CN=6 (0.67 angstroms). The positions are:
Mn in (1/2 1/2 1/2)
O in (0 1/2 1/2)
(1/2 0 1/2)
(1/2 1/2 0)
La and
Ca in (0 0 0)
I am first trying to model LaMnO3 correctly. I entered the atom positions
correctly (user defined
cell, then systematic alloy). I modified the lookup table in order to get
the correct ionic radii.
I've been having trouble graphically displaying this structure. Usually I
wind up with a blob
of color on the screen. Do you know how to rectify this problem?
Answer
We modified Element.dat for radii of La (1.5 A), Ca (1.48 A),
O (1.26 A) and Mn (0.67 A).
Made a LaMnO3 cell with the above atom positions as you described.
The atoms are small enough to plot. We did not see any blob.
If you modified element.dat while ADESH was running, it may
still use the old values and hence the blob if the old radii
were large. Incidentally , ADESH uses the radii values in
Element.dat only for plotting.
Answer
Make the cell with just La atoms with 3.903 A lattice constant.
(U-Cell-SC; Atom-La, Alloy-Systematic-Lattice const.: 3.903 etc.)
Answer
About ZrO2 and the energies:
(1) It does not matter which direction you take as
open surface. Any one will do as long as
the three (X, Y, Z) are mutually perpendicular.
(2) Ionic potentials work best with periodic boundaries.
With other situations (heavy deformation, free surfaces
GBs etc), the accuracy reduces.
(3) We made cells with following orientations:
(4) Make the boundaries periodic
Use ({Compute [ Energy < Boundaries < Periodic >>]}) (For X, Y , Z)
(5) Relax this configurations (Use 0.05 or smaller for DeltaR)
(6) Then open X surface (for 110 Zirconia)
by choosing non-periodic boundaries for X.
[Open Y surface (for 111 Zirconia)
by choosing non-periodic boundaries for Y]
(7) Relax again.
Use smaller DeltaR (less than 0.05 Angstrom) when you relax.
Use periodic boundaries wherever you can.
The energy occasionally goes up slightly since the program processes
one atom at a time and when the energy for the atom being processed
goes down, it may sometimes increase the energy for the neighbors.
I have noticed this to happen sometimes with Ionic crystals.
The increase should be small when it is there.
Overall energy should be reduced with occasional small increase.
Sorry for the inconvenience.
We are experimenting with a new parameter
to take care of this problem and make it more robust. It works very well
with covalent crystals. Ionic will take little more work.
Metallic and covalent crystals do not show this increase.
Thanks for your patience.
We will take care of it soon and mail you a new copy.
MC usually increases the energy if the temperature is greater than 0K.
Static should decrease it since it is at 0K.
(8) When you want to use free surface relaxation, the relaxed
size should be higher than the cell size for the free surface.
The relaxed size for other directions should be smaller
by cutoff radius (for rigid boundaries),
OR it should be periodic.
The relaxing size should be
(11) You can try monte-carlo instead of Static as well.
See how readily the surface distorts.
Answer
You can compare the XYZ coordinates of respective atoms
before and after the relaxation.
Answer
We relaxed (110) Zirconia (450 atoms)
with MC for 2500 iterations with Delta R = 0.05, temp. = 455 K
(about 125000 moves).
Mostly oxygen atoms moved (vibrated at their sites). No major distortion.
For the cell (450 atoms) the energy went up by about 70 eVs.
The energy of Zr atom in ideal cell is: -104.471 eV and
energy of O atom in ideal cell: -23.438 eV
Answer
Use following steps to make GaN cell.
Assuming the length of Ga-N bond to be 1.94 A.
The ratio of 1.94A to length of 'c' axis (5.185 A) is 0.3761.
You will need the new version of ADESH (ADESH 2.0) to make GaN
cell and calculate its energy.
It is available on our web page. Download adesh.zip.
You can download it and try the following GaN parameters.
You will need pkunzip.exe to unzip this demo version.
Run ADESH (demo version).
Answer
As a first step, you can build the geometric
models of your KNbO3 and LaAlO3 lattices (and interfaces)
and include appropriate defects in them.
Use these models to compare with your HREM images
to verify the types of defects that exist in your samples.
Atomistic configurations are thus established. For example,
core configuration of a partial dislocation can be calculated
with ADESH. This atom arrangement is corroborated with HREM image.
A partial dislocation will show atom arrangement unique for itself.
You can use following steps to create your lattices with ADESH:
(See Example 7 in the manual of ADESH)
Answer
ADESH includes three types of potential functions.
The functional forms (equations) are hard-coded.
The parameters can be input by users.
Anjali Nandedkar, Ph. D.
You may want to use {Subset [ Plot < Slice
< Z-Min : -1 >
If you start to get some surprises (blobs etc.), save
your data file, exit ADESH and run it again and read the
file back in. Start fresh. It helps. we have noticed
that if we are using ADESH for a while, changing options for too long,
changing materials, it restores things to start fresh.
Problems go away.
Question
Do you have any suggestions for finding a way to substitute the Ca randomly
for 33% of the La sites only? (In LaMnO3: see above).
I assume the graphical depiction is not a function of the valence of the
ions. It's a little tricky
here because the Mn has mixed oxidation state.
Then put 33% Ca atoms with Alloy-Random (If it is 33% of La atoms).
Remember to choose Ca under Alloy option. Then do Compute- Alloy.
After this make the cell with Mn and O atoms only using
U.Cell-User defined
(same lattice constant). Append the two files.
We made the files and they looked fine.
We will send them to you if you wish.
Send us a note.
Question
I'm working with zirconia. What I'm trying to do is model the
surface relaxation
(if any) that would occur for films with (110) and (111) orientation normal
to the substrate
plane. I have created cubic cells of X +/- 10 Ang. Y +/- 10 Ang. and Z+/-
Ang. I have
chosen the +z surface to correspond to the film growth direction (maybe this
is wrong, I don't know).
This is the way I tried to model the relaxation. It may be way off from the
proper procedure.
Do you have any suggestions to better model the surface relaxation? Do I
relax the entire cell
or only part of it? Should I reconfigure the geometry so the Y direction is
[110] and [111]?
Should I be using a Monte Carlo simulation instead of static minimization?
I want to see if
the atoms more readily relax on a (110) plane versus a (111) plane.
{U.Cell [ User defined
[ 0, 0, 0]
[ 0.5, 0.5, 0]
[ 0, 0.5, 0.5]
[ 0.5, 0, 0.5]
[ 0.25, 0.25, 0.25]
[ 0.75, 0.75, 0.75]
[ 0.25, 0.25, 0.75]
[ 0.25, 0.75, 0.25]
[ 0.75, 0.25, 0.25]
[ 0.75, 0.75, 0.25]
[ 0.25, 0.75, 0.75]
[ 0.75, 0.25, 0.75] ] }
Below are the cell geometries:
(110) Zirconia: Size: 450 atoms | (111) Zirconia: Size: 540 atoms
X [1 1 0] -8 to +10 | X [1 1 -2] -9 to +10
Y [ 0 0 1] -8 to +7 | Y [1 1 1] -8 to +10
Z [-1 1 0] -8 to +10 | Z [-1 1 0] -8 to +10
{ Alloy [ Systematic
[ Lattice Constant X: 5.152]
[ Lattice Constant Y: 5.152]
[ Lattice Constant Z: 5.152]
[ 0, 0, 0] Zr
[ 1/2, 0, 1/2] Zr
[ 1/2, 1/2, 0] Zr
[ 0, 1/2, 1/2] Zr
[ 1/4, 1/4, 1/4] O
[ 1/4, 1/4, 3/4] O
[ 1/4, 3/4, 1/4] O
[ 3/4, 1/4, 1/4] O
[ 3/4, 1/4, 3/4] O
[ 1/4, 3/4, 3/4] O
[ 3/4, 3/4, 1/4] O
[ 3/4, 3/4, 3/4] O ] }
{ Compute [ Make computational cell ] }
(1) smaller by cutoff size (for rigid boundaries)
(2) Periodic (for periodic boundaries)
(3) Larger (Free surface)
It should not be equal to cell size any time.
(9) The energies of the cells were:
(a) (110) Zirconia: 450 atoms:
Perioic Boundaries: -22701.887 eV
X surfaces open: (Y and Z periodic):
-22026.752 ev (Before Relaxation)
-22055.072 ev (After Relaxation with 0.05, 0.02, 0.01 DeltaRs)
Area of X surface: 15.46 X 18.22 A**2 = 281.68 A**2.
The surface Energy for (110) surface =
= -22055.752 + 22701.887 eV / 2*281.68
= 646.815 eV /2*281.68 eV/A**2 =
= 1.14815 eV/A**2 = 18370.4 ergs/cm**2.
(b) (111) Zirconia: 540 atoms:
Perioic Boundaries: -27242.26 eV
Y surfaces open: (X and Z periodic):
-26775.754 ev (Before Relaxation)
-26783.307 ev (After Relaxation with 0.05, 0.02, 0.01 DeltaRs)
Area of Y surface: 18.93 X 18.22 A**2 = 344.9046 A**2.
The surface Energy for (111) surface =
= -26783.307 + 27242.26 eV / 2*344.9046
= 458.953 eV /2*344.9046 eV/A**2 =
= 0.6653 eV/A**2 = 10645.332 ergs/cm**2.
The change is not much. With motion you will see how and how far
the atoms moved (see 10 next).
(10) With above relaxations I did, I also made the LOG file
by saying YES to 'Create Log File' (just press enter,
it changes to Yes from No).
Then look at the log file (Relax.log) with 'motion'.
Exit ADESh and run 'Motion'. Read in Relax.log.
I copy relax.log into other name (zro2110.log) to save it.
Then read the new log file in 'motion'.
This is because relax.log gets purged every time
you run ADESH.
Run this animation to see the results of relaxation.
The atom movements are more visible there.
Question
How do I investigate the movements of atoms at different
surfaces in materials?
You can also see the atom movements directly with 'Motion' program.
Go through different slices and orientations with 'Motion'
to get the whole picture. You can go through the changes
by looking at resulting XYZ coordinates as well.
A strong suggestion is to corroborate the results of
ADESH with some known experimental results.
Since you are depositing ZrO2 and have samples, it
may be feasible for you. Please send me a note
if you need suggestions. For example, if
you are doing any characterizations (X-Ray analysis,
HRTEM etc.) we can compare those with ADESH results.
Question
In Zirconia, which atom is more mobile, Zr or O?
Question
How can I calculate the wurtzite structures such as GaN? I can't find the
wurtzite crystal structures in ADESH program.
How to calculate the energy of GaN cell?
(1) To make GaN cell:
{U.Cell [ User Defined
[ 0, 0, 0,]
[ 0, 0, 0.3761]
[ 0.3333, 0.6667, 0.5]
[ 0.3333, 0.6667, 0.8761] ] }
{Alloy [ Systematic
[ Lattice Constant X: 3.189]
[ Lattice Constant Y: 3.189]
[ Lattice Constant Z: 5.185]
[ Angle Gamma(X / Y): 120.0]
[ At 0, 0, 0 : Ga ]
[ At 0, 0, 3/8 : N ]
[ At 1/3, 2/3, 1/2 : Ga ]
[ At 1/3, 2/3, 7/8 : N ] ] }
{ Compute [ Make the Computational Cell
683 atoms etc.] }
(2) Input Potential Function Parameters:
{ Compute [ Adjust IAP Parameters
< Modify Current Settings
< Nitrogen
< Ionic (CASA)
< R_CUTFF**2 = 36.00>
< Sigma = 2468.0 >
< n = 10.0 >
< Charge = -3.0 > > > > ] }
Repeat the procedure for Gallium. Input same parameters as
Nitrogen except for 'charge'. Charge is +3.0 for Gallium.
Save these parameters into a file (say elementi.dat).
By default ADESH reads parameters from element.dat.
You may save new parameters in element.dat itself as well.
Choose:
{ Compute [ Adjust IAP Parameters
< Save IAP Parameters to a File
( File Name: Element.dat) > ] }
We have tested the parameters for strains.
The energy went up for
compressive and tensile strains. Hence the above
configuration had the minimum energy for ideal cell.
We tested it upto 5% strain.
The ideal energy of atoms is rather high, but bulk modulus
was calculated to be close to 209 GPa.
We calculated the vacancy formation energy to be
about 63.5 eV.
Please play with these parameters. They are rather crude
at this time. Send me a note if you have any questions.
Question
We are looking at perovskite thin films on perovskite substrates with
mismatch between 0 and 7 percent. We have HREM images of the interface
between a KNbO3 film and a LaAlO3 substrate which have a mismatch of 5 %.
These images show the insertion of misfit dislocations to accommodate the 5
% mismatch (19 film lattice fringes for every 20 substrate lattice
fringes.) However, the dislocation appear to be spread out over many
lattice fringes which suggest the dislocation are disassociated into
partial dislocations. Between these partial dislocation may exist a defect
similar to a stacking fault.
Both materials have the perovskite crystal structure.
The atomic positions are:
A B O3
K Nb O3
La Al O3
A (0, 0, 0)
B (0.5, 0.5, 0.5)
O (0.5, 0.5, 0), (0.5, 0, 0.5), (0, 0.5, 0.5)
Both KNbO3 and LaAlO3 are cubic at the growth temperature.
KNbO3 4.016 =C5
LaAlO3 3.818 =C5
Since the dislocation are expected to reappear with a regular spacing, I
figured the model could just be repeated. This would mean a model which is
20 x 20 LaAlO3 (19 x 19 KNbO3) unit cells. The bottom (LaAlO3 substrate)
would be fixed. The top (KNbO3 film) would be free. We would probabley
need a thickness of 2 or 3 unit cells of LaAlO3 for the substrate and at
least 3 or 4 unit cells of KNbO3 for the film. This would result in
approximately 2000 atoms in the simulation.
Please let me know what addition data I need to obtain for the model.
{U.Cell [User Defined < 0, 0, 0>
< 0.5, 0.5, 0.5>
< 0.5, 0.5, 0>
< 0.5, 0, 0.5>
< 0, 0.5, 0.5>]}
{Atom [Potassium] }
{alloY [ Systematic ]}
(Input appropriate lattice constants,
Keep the angles 90 degrees (default for cubic)
Also choose appropriate atoms for each site by going to the site
and pressing
Default orientation of the lattice is [100] [010] [001].
You can change it by choosing {M. Index} option in the top
manu bar in ADESH.
Input the dislocations by choosing the option {Dislocation}
Input Dislocation type, Burger's vector etc.
A mixed dislocation (neither edge nor Screw) can be calculated
by putting edge component of the Burger's vector in dislocation 1
and by putting screw component in dislocation 2.
Dislocation line is along Z.
(You choose your Z direction in {M. Index})
The next step is to calculate energy and analyze the
energetics of the configurations etc.
I will contact you soon about the
energy calculations.
Question
ADESH calculates energy for what type of materials?
Is it possible to use the user-defined potential in ADESH program? If
possible, how?
How did you decide the parameters for the potential
in Si? Do you have a potential for GaAs?
(1) Two body: For materials with metallic bonds (Standard Lennard-Jones)
(2) Three body: For materials with covalent bonds (Tersoff: Three body 2
and Keating: Three body 1 )
(3) Ionic : For materials with ionic bonds: (coulomb interaction
+ two body term).
Hence sometimes we are able to design parameters for given material.
(Mostly we have done it for Two body and Ionic potentials so far).
Other times the parameters come from literature.
Parameters for Si are from a paper by (1) Keating, (2) Tersoff. The manual
has the references to these papers.
We have used Keating function with parameters fo GaAs before. I will send you
those soon if you are interested. They came from the books on potential
functions edited by Stoneham.
I used Ionic potential for GaN.
Please download a demo version from our web page. Ionic potential
is available there. Please test it with the parameters for GaN.
(I had sent them to you earlier).
The parameters for the potential functions that are
included in ADESH can be changed by the user.
Go to {Compute [ Adjust IAP paramters < Modify current settings > ] }
Choose any element. For Si, you will see that the three body (Tersoff)
potential is a default one. All parameters you see listed, can be changed by you.
Use arrow keys to go to the option and push
Questions? Comments? Suggestions?
Send a note to Anjali Nandedkar at CASA Engineering
CASA Engineering
15 Dartantra Drive
Hopewell Junction
NY 12533
U.S.A
Tel.: (845) 226 - 1925
anjali@casaengineering.com
Return to the Main Page.