XueMaXue commented 2 months ago

Hi everyone,

I have installed Matjes-1.1. I would like to know if I want to simulate an AFM structure, I can provide J1, J2 and D1, D2. Which parameters are necessary in my input file? Or in the examples in the package, is there any explanation about the meaning of different parameters? If I want to run a magnetic field-temperature phase diagram after relaxed the structure, how should I do it? Are there any examples?

Looking forward to your reply!

Best regards, Xue Ma

bertdupe commented 2 months ago

Hello, I indeed did not put examples of simple MC - sorry. This is the input file with the parameters you need. Look in the wiki for the key words.

Lattice

Periodic_log .T. .T. .F. Nsize 10 10 1 alat 1.0 1.0 1.0 lattice 1.0 0.0 0.0 # a_11 a_12 a_1 0.0 1.0 0.0 # a_21 a_22 a_23 0.0 0.0 1.0

H_ext 0.0 0.0 00.0

Atoms

atomtypes 2 Fe 0.5 0.0 0.0 F 0 # atom type: (name, mag. moment, mass, charge, displacement, number TB-orb.) Ir 0.0 0.0 0.0 F 0

atoms 2 Fe 0.0 0.0 0.0 Ir 0.5 0.5 0.79

Magnetic interactions

magnetic_J 1 1 1 1.0d-3 # between atoms 1 and 1, shell 1, amplitude in eV 1 1 2 0.05d-3 # between atoms 1 and 1, shell 2, amplitude in eV

magnetic_anisotropy 1 0.0 0.0 1.0 -0.05d-3

magnetic_D 1 1 2 1 -0.5d-3 #between atoms 1 and 1, mediated by 2, shell 1, amplitude in eV 1 1 2 2 -0.05d-3 #between atoms 1 and 1, mediated by 2, shell 1, amplitude in eV

for the initialization as an AFM (thinking that it is collinear). Use the file init.config

file that tells what the initial configuration should be

#############################

Define magnetic structure

############################# configuration_magnetic spiral # initialize as a spiral

#########################################

Configuration of magnetic spin spiral

#########################################

vector of the SS and 2 vectors for the polarization

qvec_magnetic 0.5 0.0000 0.0 # edge spiral at q=0.5 is an AFM - check the wiki for more information Rq_magnetic 0.0 0.0 1.0 Iq_magnetic 1.0 0.0 0.0

XueMaXue commented 2 months ago

Thanks for your reply. According to your input, my input is as follows (structure: BiFeO3; R3c).

begin input

Lattice

Periodic_log .T. .T. .T. Nsize 10 10 10 alat 1.0 1.0 1.0 lattice -0.0661926633566317 3.7468483555038055 3.7468483555038055 3.7468483555038055 -0.0661926633566317 3.7468483555038055 3.7468483555038055 3.7468483555038055 -0.0661926633566317

H_ext 0.0 0.0 00.0

Atoms

atomtypes 2 Fe1 4 0.0 0.0 F 0 # atom type: (name, mag. moment, mass, charge, displacement, number TB-orb.) Fe2 4 0.0 0.0 F 0

atoms 2 Fe1 0.4998724586952532 0.4998724586952532 0.4998724586952532 Fe2 -0.0001275413047467 -0.0001275413047467 -0.0001275413047467

Magnetic interactions

magnetic_J 1 2 1 -26.998d-3 # between atoms 1 and 2, shell 1, amplitude in eV 1 1 2 -2.005d-3 # between atoms 1 and 1, shell 2, amplitude in eV 2 2 2 -2.005d-3 # between atoms 2 and 2, shell 2, amplitude in eV

magnetic_anisotropy 1 1.0 1.0 1.0 -0.054d-3

magnetic_D 1 2 2 1 0.369d-3 #between atoms 1 and 1, mediated by 2, shell 1, amplitude in eV 1 1 2 2 -0.0036d-3 #between atoms 1 and 1, mediated by 2, shell 1, amplitude in eV 2 2 2 2 -0.0036d-3

end input

Can you help me find any errors? Can I directly use Matjes to execute this file?

In addition, I want to confirm whether the units of J and D are eV/atom or eV/pair.

If a magnetic field is applied, do I only need to change H_ext?

What if temperature is applied? Does the same Tini and Tfin apply to determine the temperature?

Can I use MD?

I understand that the number of MC steps is controlled by Total_MC_Steps? Is MD Total_MD_Steps?

How do I start from a random structure?

In addition, if I want to use a structure under a certain condition, how do I load it?

In addition to input, are other input files required?

Looking forward to your reply!

Best regards, Xue Ma

Mexxregout commented 2 months ago

Hello Xue,

I can answer a few of your questions.

Q: I want to confirm whether the units of J and D are eV/atom or eV/pair A: The units are atomic (not per bond).
Q: Can I directly use Matjes to execute this file? A: If you have an input file and an init.config file in the same directory then you can call Matjes directly to run it.
Q: If a magnetic field is applied, do I only need to change H_ext? A: yes, this is enough.
Q: What if temperature is applied? Does the same Tini and Tfin apply to determine the temperature? A: In the spin dynamics the temperature is enabled by the Tini 3.0d0 (should give 3 Kelvin). The flag rnd_num_type 1 or 2 gives the random number generator (internal or mkl) and name_rnd normal gives a normal/Gaussian distribution. Be aware that the fluctuations for the temperature are governed by the gilbert damping, timestep (dt) and k_bT. This means that the you should check if the Tini you give matches with the desired simulation temperature you want. Matjes can print the spin temperature to check if it matches Tini (flag gra_log .T.).
Q: Can I use MD? A: I think Matjes has some implementation of phonons but I never used it so Bertrand can help you more with the capabilities of Matjes in this regard.
Q: How do I start from a random structure? A: in the init.config file you can define configuration_magnetic random to get a random spin configuration.
Q: In addition to input, are other input files required? A: You need an input file to specify all the simulation details (Monte Carlo, Spin dynamics, etc and the needed parameters). The other input you need is the initial magnetic texture which is defined in the init.config file (or magnetic_init.dat if you want to fully specify the texture yourself). The init.config file has more general initial conditions like a spin spiral, skyrmions or random spins.

Greetings Mexx

bertdupe commented 2 months ago

yes, that will now work because of the DMI So BFO is a difficult case. You do not start with the easy part.

the anisotropy will be difficult because the different 111 axes are not orthogonal. You have an easy axis along that 111 - if it is the polarization axis, it should ba hard axis (according to our calculations but not according to the experimentalists - your call) the exchange - smart move but that will not do what you intend I think. I would do this: magnetic_J 1 2 1 -26.998d-3 # between atoms 1 and 2, shell 1, amplitude in eV 1 1 1 -2.005d-3 # between atoms 1 and 1, shell 2, amplitude in eV 2 2 1 -2.005d-3 # between atoms 2 and 2, shell 2, amplitude in eV

without the DMI it works fine. The rest of the input looks like this - I just took arbitrary numbers. You will have to optimize these.

Tini 1.0d0 !loop on T: start with Tini, finish with Tfin and do n_Tsteps Tfin 100.0d0 n_Tsteps 51 !number of MC simulations at different T, uses Tini and Tfin Total_MC_Steps 2000 T_relax 10 !number of thermalization steps (will be T_relax*N) T_auto 5 !measure steps occur every T_auto autocorrelation steps

cone 3.1415926535897931
sphere_samp .T. print_relax .F. Cor_log .F.

n_sizerelax 1 n_relaxation 10

calc_topo .F. gra_log_position .F. gra_freq 1

bertdupe commented 2 months ago

now for the DMI, it is a bit more difficult. You have to introduce a third atom OR an antisymmetric exchange tensor. The way I did works fine but you will have to check carefully the direction of the DMI if you would to explore spin spiral symmetries and so on - Good luck :-)

Atoms

atomtypes 3 Fe1 4.0 0.0 0.0 F 0 # atom type: (name, mag. moment, mass, charge, displacement, number TB-orb.) Fe2 4.0 0.0 0.0 F 0 O 0.0 0.0 0.0 F 0

atoms 3 Fe1 0.4998724586952532 0.4998724586952532 0.4998724586952532 Fe2 -0.0001275413047467 -0.0001275413047467 -0.0001275413047467 O 0.25 0.25 0.25

Magnetic interactions

magnetic_J 1 2 1 -26.998d-3 # between atoms 1 and 2, shell 1, amplitude in eV 1 1 1 -2.005d-3 # between atoms 1 and 1, shell 2, amplitude in eV 2 2 1 -2.005d-3 # between atoms 2 and 2, shell 2, amplitude in eV

magnetic_anisotropy 1 1.0 1.0 1.0 -0.054d-3

magnetic_D 1 2 3 1 0.369d-3 #between atoms 1 and 1, mediated by 2, shell 1, amplitude in eV 1 1 3 1 -0.0036d-3 #between atoms 1 and 1, mediated by 2, shell 1, amplitude in eV 2 2 3 1 -0.0036d-3

end input

XueMaXue commented 2 months ago

I will try the above suggestions. Thank you so much for your prompt and insightful response.

Thanks and Regards, Xue Ma

bertdupe / Matjes

How to simulate AFM structures #147

Lattice

Atoms

Magnetic interactions

file that tells what the initial configuration should be

Define magnetic structure

Configuration of magnetic spin spiral

vector of the SS and 2 vectors for the polarization

begin input

Lattice

Atoms

Magnetic interactions

end input

Atoms

Magnetic interactions

end input