Hi, the point is that your input file is wrong, so there is no way to avoid the error but trying to understand which structure you want to simulate.
1) you say you're trying to match a lattice with ibrav = 2 (cubic F (fcc)) with one with ibrav = 12 (Monoclinic P, unique axis c ), is that right? If so, you should carefully try to understand which kind of lattice could describe the matching of the two. At the moment, you're trying to build a crystal composed of ALL carbon atoms within a simple cubic cell, is that what you want to do? This is not the source of your error, but it is something you should care about to arrive to a reasonable input file 2) given the unit cell you set up in your input, that is simple cubic, most of the provided atoms will produce the same error (so as you should remove those atoms if you want the code not to provide that error). The reason is that, in the alat unit you use (the same would be for crystal units), for example the two atoms C 0.00 0.00 0.00 C 1.00 1.00 0.00 are equivalent by a lattice translation. Indeed, the coordinates of the second can be obtained by translating the first atom by a1 + a2 with a1 = a ex, a2 = a ey, ex and ey the versors of the x and y axes respectively, and a1, a2 two primitive vectors out of the three defining the primitive cell. The input file should specify only inequivalent atoms in the unit cell, that is no atom coordinates should be obtained from those of another by simply applying a lattice translation. Your error could be associated with the fact that you are building a supercell, so a "larger" unit cell containing two or more unit cells of an initial lattice. In this case it can be correct to select the atoms using XCrysDen, but it will provide you such coordinates in units of alat of the initial cell. So, if the super cell has a side which is for example twice that of the original one, then the coordinates in alat units should be all divided by 2! Example: - start with a simple cubic cell, only one atom in the unit cell in position (0 0 0) - imagine you want to to double the super cell along all directions, so building a 2x2x2 supercell. Then the atoms in the unit cell become (in the OLD alat units): 0 0 0 1 0 0 0 1 0 0 0 1 1 1 0 1 0 1 0 1 1 1 1 1 which are 8 = 2x2x2 - because new_alat = old_alat x 2, the coordinates expressed in terms of new alat become: 0 0 0 1/2 0 0 0 1/2 0 0 0 1/2 1/2 1/2 0 1/2 0 1/2 0 1/2 1/2 1/2 1/2 1/2 The first set of coordinates would produce the error you find, the second one will not. Not related with your error: you use PAW pseudopotentials, you might need to specify ecutrho as well. The default value, = 4 * ecutwfc, might not provide converged results in many or most cases. Usually values from 6*ecutwfc to even 12*ecutwfc (depending on the system and on the property you want to compute) are needed. Giovanni -- Prof. Giovanni Cantele Dipartimento di Fisica "Ettore Pancini" Universita' degli Studi di Napoli "Federico II" Complesso Universitario M. S. Angelo - Ed. 6 Via Cintia, I-80126, Napoli, Italy e-mail: [email protected] Phone: +39 081 676910 Web page: https://sites.google.com/view/giovanni-cantele/home Il giorno mer 7 gen 2026 alle ore 20:43 Мищенко В.Н. <[email protected]> ha scritto: > Dear QE forum! > > I would like to ask a question regarding the modeling of a hybrid > semiconductor structure that combines two materials with crystal lattices > having different values for the parameter ibrav=2 and ibrav=12. To describe > this structure, the parameters ibrav=0 and nosym=.true. were chosen, and > the values of the remaining input parameters are presented in the appendix, > in the file diam.scf.in. The coordinates of all atoms were selected using > the XCrySden program. > > When running the diam.scf.in file in QE 7.4.1, I receive the error > message: > > Error in routine check_atoms (1): atoms # 1 and # 2 differ by lattice > vector (-1, -1, 0) in crystal axis > > When attempting to continue the modeling after removing atom number 2 from > the list, the program again returns a similar error, but this time between > the first and third atoms specified in the ATOMIC_POSITIONS alat. > > I'd like to ask: is there any way to avoid this error while maintaining > the original number of atoms at 24? > > Best regards, > > Valery Mishchanka > > PhD, Assoc. Prof., BSUIR > > _______________________________________________________________________________ > The Quantum ESPRESSO Foundation stands in solidarity with all civilians > worldwide who are victims of terrorism, military aggression, and > indiscriminate warfare. > > -------------------------------------------------------------------------------- > Quantum ESPRESSO is supported by MaX (www.max-centre.eu) > users mailing list [email protected] > https://lists.quantum-espresso.org/mailman/listinfo/users
_______________________________________________________________________________ The Quantum ESPRESSO Foundation stands in solidarity with all civilians worldwide who are victims of terrorism, military aggression, and indiscriminate warfare. -------------------------------------------------------------------------------- Quantum ESPRESSO is supported by MaX (www.max-centre.eu) users mailing list [email protected] https://lists.quantum-espresso.org/mailman/listinfo/users
