FAQ

0 Like 0 Dislike

Anonymous

BCC or FCC lengths are not equal

Hello,

While generating FCC or BCC SQS structure the lattice parameter are not exactly equal why is that? For Cubic systems, it should be equal?
I went through your code and saw supercell.py fixes the cell (natoms/2: BCC) and invoking the mcsqs -rc command.

What is the reason? Does mcsqs automatically fixes the cell size according to the number of atoms? I would guess if I fix the cell and number of atoms then it should do according to these constraints.

Thank you for your answer?
Regards,

Report abuse

2 Responses

  1. 0 Like 0 Dislike

    Shivam Tripathi

    If your total number of atoms is such that it can not form cubic supercell then the cell parameters will be different in different directions (lattice parameters will always be the same in all directions). In order to form a cubic supercell in BCC total no. of atoms should be 2*n^3 and in case of FCC it should be 4*n^3 (n is an integer). Can you please let me know how many atoms you are using?

    Reply Report abuse

    Please login to answer the question.

    1. 0 Like 0 Dislike

      Anonymous

      Thanks for your response. In my case, I have a system composed of Ti0.2Hf0.2Nb0.2Zr0.2V0.2. I was generating the smallest structure of 50 atoms using the lattice parameter of 3 angstroms.  However, with mcsqs code from Atat sqscell lengths are different it doesn't generate a cubic cell but different lengths below is some of the cell parameters:

      89

      5.000000 -1.000000 0.000000
      -5.000000 -1.000000 0.000000
      0.000000 0.000000 -2.000000

      5.000000 -1.000000 -1.000000
      5.000000 1.000000 1.000000
      -5.000000 -1.000000 1.000000

      0.000000 0.000000 -2.000000
      0.000000 2.000000 0.000000
      5.000000 0.000000 0.000000

      .....

      However, I tried your approach.  For n is 4, natoms=128 then atomic composition of Ti becomes 0.2*128=25.6 and the code gives error -> it is not integer. I was thinking atoms of different radius might also be the issue in not generating cubic structure however, mcsqs i think do not take into account radii. 

      Let me know where I am wrong. I have read the manual by Atat I could not find these rules. I have read your supercell code which takes in natoms and produce 3 factors and then you run code for these cell lengths (orthogonal). (if (strtype == 'BCC'):     if(natoms%2!=0):) you are here also enforcing the for BCC atoms should be a multiple of 2.

       

      Any help would be great in resolving this problem. let me know if you require more info on my system. With above system i want to generate BCC structure. 

      Reply Report abuse

      Please login to answer the question.

  2. 0 Dislike

    Shivam Tripathi

    We are taking conventional cells of FCC (4 atoms) and BCC  (2 atoms) as our unit cells. So, the total number of atoms should be integer multiple of 4 (FCC) and 2 (BCC). You can only make cubic cells if your total number of replication (repeating unit cell) is same in all three direction so total number of atoms should be (n^3*4 (FCC) and n^3*2 (BCC)). In case of your choice of n, it is showing error because total no. of atoms of any of the 1 specie (Ti) is not an integer and your structure can not have fractional atoms. In your case, I will suggest total number of atoms to be 250 to get cubic supercell.

    Reply Report abuse

    Please login to answer the question.