Activity for Lars Nordström
-
Lars Nordström posted a comment on discussion Elk Users
Dear Ronald! Nice to "meet" you again.,at least if you are the one I think. I remember with pleasure my visit to Carnegie Lab around 96' when Igor was still there and I was at NRL. I am not 100% sure about your problem, but ... If you put the quantization axis along x there would probably be no splitting of "up" and "down" spins as the band structure part assumes a quantization axis along z. But in order to fully understand your problem I have some questions for you: Without spin orbit orbit coupling...
-
Or, if you instead want to include some 5f-character in the valence bands, you can add a local orbital in the species file ... 1 : apword 0.1500 0 F : apwe0, apwdm, apwve 0 : nlx 7 : nlorb. --- 6->7 , ie one extra local orbital 0 2 : lorbl, lorbord 0.1500 0 F : lorbe0, lorbdm, lorbve 0.1500 1 F 1 2 : lorbl, lorbord 0.1500 0 F : lorbe0, lorbdm, lorbve 0.1500 1 F 2 2 : lorbl, lorbord 0.1500 0 F : lorbe0, lorbdm, lorbve 0.1500 1 F 3 2 : lorbl, lorbord. --- the extra lo taking care of f-character (l=3)...
-
Dear Ssingh, It seems to be a mix up of conventional cell which is face centered cubic and primitive cell that is cubic. I have no idea why ... Maybe you can send me the GEOMETRY.OUT you try to plot? In general, at Bilbao web-page you can see the Wyckoff positions of Space group #225 at the top you can read beneath Coordinates the four face centering vectors which bring you from primitive to conventional site. Non-trivial "Coordinates" indicate that they use multiplicity for a conventional site and...
-
Dear topgun, this is not related to the earlier problem you had with accuracy, In fact this result it is not even in error. Unfortunately the international crystallograhic association long ago decided that the Wyckoff multiplicity is referred to the conventional cell not the primitive. Elk will use the primitive and since the concentional cell is four times larger than the primitive, the site 4a has only one site per (primitive) cell ... Sincerely, Lars
-
Dear Top Gun, For the H at wyckoff site 4d you need to give 1/3 and 2/3 in higher precision, otherwise the rounding of error will give 12 sites that are extremely close by since Elk uses a smaller criterion for when they are equivalent . Good luck! Lars Nb: the 0.33333 for the Y site might seemingly be 1/3, but it is in fact a free parameter of the 6f wyckoff position ...
-
Dear Casey, I forgot to point out two points. 1) The python-script assumes spin-orbit coupling. That is that rotations regards both orbital and spin space. But that is also assumed in tabulations of crystallographic space groups as those in the Bilbao crystallographic server. When spin polarized set up but without spin-orbit coupling the SYMCRYS.OUT file is too general for Bilbao as they take the form as described by Kay. 2) Please be aware that if the atomic positions are obtained by lattice optimisation,...
-
Dear Casey, I actually have created a python script which I think does what you want. . It reads your SYMCRYS.OUT and produce a format that is readable by Bilbao and their web-program "Identify Group" at https://www.cryst.ehu.es/cgi-bin/cryst/programs/checkgr.pl?tipog=gesp You just run it with python3 SymCrys2xyz.py in a directory where you have a SYMCRYS.OUT file. for BaTiO3 with lattice input in elk.in: scale 7.46 avec 1.000000000 0.000000000 0.000000000 0.000000000 1.000000000 0.000000000 0.000000000...
-
Dear Casey, I actually have created a python script which I think does wha you want . It reads your SYMCRYS.OUT and produce a format that is readable by Bilbao and their web-program "Identify Group" at https://www.cryst.ehu.es/cgi-bin/cryst/programs/checkgr.pl?tipog=gesp You just run it with python3 SymCrys2xyz.py in a directory where you have a SYMCRYS.OUT file. for BaTiO3 with lattice input in elk.in: scale 7.46 avec 1.000000000 0.000000000 0.000000000 0.000000000 1.000000000 0.000000000 0.000000000...
-
Dear Zhiwei, yes the reason for slow convergence is definitely moment rotations. As you can see by plotting the different y and z components you are not close to convergence. So either be patient, as the torques build up the speed of rotation will increase and you will reach convergence sooner or later ... Or, from the plots you extrapolate to a probable solution. Set up a calculation where the different site moments are constrained, either in direction and value or only in direction, see elk.pdf...
-
Dear Zhiwei, as everything is stable in absence of SOC I wonder if it is not so that you have encountered a case where the collinear anti-ferromagnetic order is unstable, due to Dzyaloshinskii-Moriya like interactions which vaish in absence of SOC. At least that is my experience. Then the local moments will rotate away from their initial direction. That is, they acquire components along y and/or z too. This process can be slow and you should be patient or you could do an "intelligent" guess and constrain...
-
Dear Attique! OK, a little what I assumed. Calculations in VASP and other popular pseudo-potential methods ignore the core conribution to total energy. As of Cu 29 electrons you with pseudo potentials only treat at most 11 . Ignoring the deeply bound core states. This definitely affects the total energy ... Elk on the other hand is an all-elecron approach. Look at eg DOI 10.1126/science.aad3000 for comparisons in accuracy and value. Again, welcome to elk community and using the all-electron DFT and...
-
Dear Attique! OK, a little what I assumed. Calculations in VASP and other popular pseudo-potential methods ignore the core conribution to total energy. As of Cu 29 electrons you with pseudo potentials only treat at most 11 . Ignoring the deeply bound core states. This definitely affects the total energy ... Elk on the other hand is an all-elecron approach. Look at eg DOI 10.1126/science.aad3000 for comparisons in accuracy and value. Again, welcome to elk community and using the all-electron DFT and...
-
Dear Attique! OK, a little what I assumed. Calculations in VASP and other popular pseudo-potential methos ignore the core conribution to total energy. As of Cu 29 electrons you with pseudo potentials only treat at most 11 . Ignoring the deeply bound core states. This definitely affects the total energy ... Elk on the other hand is an all-elecron approach. Look at eg DOI 10.1126/science.aad3000 for comparisons in accuracy and value. Again, welcome to elk community and using the all-electron DFT and...
-
Dear Attique! OK, a little what I assumed. Calculations in VASP and other popular pseudo-potential methos ignote the core conribution to total energy. As of Cu 29 electrons you with pseudo potentials only treat at most 11 . Ignoring the deeply bound core states. This definitely affects the total energy ... Elk on the other hand is an all-elecron approach. Look at eg DOI 10.1126/science.aad3000 for comparisons in accuracy and value. Again, welcome to elk community and using the all-electron DFT and...
-
Dear Attique! Welcome to elk community. I do not know what your earlier experience is so my answer might not be spot on what you are worrying about ... Then let us know, for instance what energy do you compare with. You run for a supercell consisting of four atoms per unit cell, what is referred to as the conventional cell for fcc. With the given setup you have you can simply add two lines in elk.in primcell .true. Then elk will run with corresponding primitive cell, the true fcc cell which would...
-
Dear Karima, Yes, bfcmt is important since it breaks the time reversal symmetry in the set up of crystal symmeries. Without it, the result will usually be a time reversal symmetric solution. The exception is that magnetic solution can still be allowed in low symmery cases, without breaking the symmetry. Bfcmt also has a second role. From an essentially non-magnetic state you can induce a magnetic state by starting with a fairly large field and reduce it with reducebf parameter. Depending on how steep...
-
Dear Karima! as you might have noticed I was a couthor on the paper you mentioned, although the code was pre-elk in that work. I can safely assure you elk can do everything that code could, but better, since I was involved in the programming of both. As regards the actual calculations I was mainly contributing as external supervisor to Jussi Enkovaara, ss he came as visitor to Uppsala, but I do know the details behind his calculations. If you have any specific questions I will help you if I can....
-
Hi again Jack, Just comments to your statements, to be clear: 1) In fact, from previous experience, use of symmetry in geometry optimisations can prevent the optimiser from distorting into a lower symmetry if it wants to, and such distortions can sometimes lower the energy. That is not just from "previous experience" -- it is an obvious fact. If you enforce a symmetry you wil not allow any symmetry breaking. But structures far from equilbrium might have instabilities although the stable symmetric...
-
Dear Jack, you do want you want. However, I would first optimize the structure within the space group and then perhaps switch off the symmery and increase the accuracy to check if it stays there. Alternatives require high accuracy and large amount of computer time, to be certain that one stay on an accurate Born-Oppenheimer surface and do not stray off ... /Lars
-
Dear Jack, I do not know what you want to do, but if you want to optimize the structure within the P3_1 21 symmetry you need to give a structure that do belong to the space group. Your data is slightly deviating. But avec a -asqrt(3) 0.0 a +asqrt(3) 0.0 0.0 0.0 c would fullfill the criteria together with atoms 1 : nspecies 'Te.in' : spfname 3 : natoms; atposl below x 0.0 0.333333333 0.0 x 0.6666666666 -x -x 0.0 where in you case you can start with a=4.2629 c=11.2477 x=0.2683 This set up will stay...
-
Dear Myron, please let us know why you want that and we might be able to help. Best wishes, Lars
-
Dear Malte, similar issues have been dicussed before in this forum. The answer is found in a fairly old paper: Spin-Orbit Coupling of the Actinide Elements: A Critical Evaluation of Theoretical Equilibrium Volumes by Lars Nordström, John Wills, Per Andersson, Per Söderlind, and Olle Eriksson in PRB, 63, 035103 (2001). Best regards, LARS
-
Dear Cenna, It looks like you have set up the structure in units of Ångström instead of atomic (Bohr) units. You can fix this by adding two lines, which will multiply the lattice vectors with the conversion factor: scale 1.8897 Good luck, Lars
-
Dear Thilo, You use the wrong k-path. Remember that the k-points under plot1d is given in reciprocal lattice vectors. So you go from Gamma (0,0,0) to another Gamma (0,1,0) centered in next BZ. The reciprocal lattice vectors are given in INFO.OUT and you have to express X as a linear combination of them ... Good luck, Lars
-
Dear Malte, it is not implemented, but the worst effect comes when you e.g. optimise volume with varying muffin-tin radius. So by keeping this radius constant the effect is almost gone. However, if you are interested in e.g. the spin-orbit splitting of the 6p state you are in trouble, if you cannot keep it in the core ... Best wishes, Lars
-
Dear Malte, it is not implemented, but the worst effect comes when you e.g. volume with varying muffin-tin radius so by keeping this radius constant the effect is almost gone. However, if you are interested in e.g. the spin-orbit splitting of the 6p state you are in trouble, if you cannot keep it in the core ... Best wishes, Lars
-
No, varying the linarization energy is not truely variational in that sense. Best regards, Lars
-
Yes, one can look at the linearization energies and compare with the occupied part of the partial DOS. For wide bands the exact linearization energy makes little effect but on the other hand narrow bands they do. One should make certain that the linearization energy corresponds to the center of the partial DOS. The default value 0.15 is not aware at all of the bands as eg the position of the Fermi energy. Autolinengy takes care that the linearization energy is situated below the Fermi energy but...
-
OK, the shift is by default -0.1, which means the linearization energy is 0.1 below Fermi energy. This usually put the linearization eneregies at a relevant energy for band states. However, you can change this default value in elk.in through the variable dlefe. In your case with narrow f-bands, they are sensitive on the linearization energy if is too low (maybe even outside the range of f-bands) as it seems to be in your case with default value 0.15. So yes it is better to use autlinengy. You can...
-
Dear Harry! Autolinengy=T set the linearization energies to a certain energy below Fermi energy, so that one does not have to manually reset it if the default value 0.15 Ha is too low or high. Usually you want the linearization energy to be in the middle of the occupied bands. What is Fermi energy in your case? Best / Lars
-
Sorry I was probably too brief ... When you have very narrow bands close to Fermi energy, as your f-bands, details in bonding and hybridization will change occupations drastically leading to oscillations and unstable numerics. The cause of this problem is that these bands are unphysically (unchemically) high in energy due to LDA/GGA errors. Have a nice weekend, Lars
-
OK, to be clear. That the 4f states are at positive energies I think is a LDA error, they are lying to high ... What happens when you put the f in core? Remember, you do not expect any f-bonding in rare earths as LDA/GGA always predict. Therefore it is common to treat them as core or to use LDA+U to localize them. Best, Lars
-
Hi!! I would guess you have semi-core s and p states below 0 energy at the corresponding linesrisation energies too. Isuggest you try f in core to compare Best eishes, Lars
-
Hi! You say you use default species file for Lu.in, which assumes looks like 'Lu' : spsymb 'lutetium' : spname -71.0000 : spzn 318945.3295 : spmass 0.237356E-06 2.8000 57.8887 700 : rminsp, rmt, rmaxsp, nrmt 21 : nstsp 1 0 1 2.00000 T : nsp, lsp, ksp, occsp, spcore 2 0 1 2.00000 T 2 1 1 2.00000 T 2 1 2 4.00000 T 3 0 1 2.00000 T 3 1 1 2.00000 T 3 1 2 4.00000 T 3 2 2 4.00000 T 3 2 3 6.00000 T 4 0 1 2.00000 T 4 1 1 2.00000 T 4 1 2 4.00000 T 4 2 2 4.00000 T 4 2 3 6.00000 T 4 3 3 6.00000 F 4 3 4 8.00000...
-
Dear Yo, Is your main concern that you get E(H)>E(0)? Well, it should be. Probably their Eq 3 lacks a minus sign ... It is clearly written that the energy minimum corresponds to the maximum in their J(q) curve. and hence E(H)-E(0)>0 also for them. The factor two I cannot explain, but I suggest you do the full q-range with cone angle 90 degrees, which is more stable numerically and compare the curves ... Good luck, Lars
-
Hi Alaska! Are you the Alaska I know as David's post-doc? Anyhow, the release is ready for the next elk version . If you ate willing to act as early bird, we can send you you a link for a repository, if you then let us know any short comes Best regards, Lars
-
yes
-
Dear Chris, Yes the atomic moment for Ho is 10 Bohrs per atom, decomposed into a spin moment of 4 and orbital moment of 6. Hund's rules tell you that you have 7 spin up electrons and 3 spin down, hence spin moment 4. The spin up electrons do not give rise to any orbital moment, but the 3 spin down states add up with 3+2+1=6. As the 4f shell is more than half filled these two moments add upp to 10 bohr magnetons per atom. Now, you have to realise that the numbers you have given says that the two Ho...
-
Dear Chris! Yes the atomic moment for Ho is 10 Bohrs per atom, decomposed into a spin moment of 4 and orbital moment of 6. Hund's rules tell you that you have 7 spin up electrons and 3 spin down, hence spin moment 4. The spin up electrons do not give rise to any orbital moment, but the 3 spin down states add up with 3+2+1=6. As the 4f shell is more than half filled the moments add upp to 10 per atom. Now, you have to realise that the numbers you have given says that the two Ho atoms have spin moments...
-
Dear Chris, First I suggest you read the manual "elk.pdf" enclosed in the release. There is always some good information. For instance if you search for "band" you get 15 hits including the relevant text: Produces a band structure along the path in reciprocal space which connects the vertices in the array vvlp1d. The band structure is obtained from the second-variational eigenvalues and is written to the file BAND.OUT with the Fermi energy set to zero. If required, band structures are plotted to...
-
It seems I was not logged in, but the last post was by me ... /Lars
-
Elk wants everything in atomic units. Neither Ånström nor Electronvolts are atomic units. 1 Å = 0.52917 a.u. For length a.u. is also referred as Bohr radius. Hence in elk.in your choice should be specified as avec -3,2445 3.2445 3,2445 3,2445 -3,2445 3,2445 3,2445 3,2445 -3,2445 Does it help? /Lars
-
Hi! The lattice constants of spacegroup.in and elk.in seem to be inconsistent. What cubic lattice constant did you intend to use and in what units? /Lars
-
Dear Cenna, although off the topic, I just want to comment on your AF setup. Andrii is correct in that your AF is not what is observed experimentally. For the conventional cell, it is the two in-plane sites that order AF with respect to each other, the ordering in between plane are less important. It is known that you need a U on the Cu site in order to get a stable AF and insulatig solution, as I recall above 6 eV, which means that your 7.5 eV should be OK. The U on La might improve the low lying...
-
Dear Ana, As I read from your plots the position of the band of interest is due to the SO splitting of the Bi-p bands at Gamma point. It is a little hard to read off from the plots, but I would get that there is a trend of increasing splittings from 1.4, 1.6 to 1.75 when going from ELK -> W2k -> FPLO. These differences are due to a known problem of scalar relativistic treatments of the SO splittings of p-states in heavy materials, see eg PRB 63, 035103 (2001). This problem was at least partially...
-
yes & yes
-
Hi Harry! Yes you missed the fact that the density matrix is in general hermitean. Although you might expect a real matrix in the case of time reversal symmetry and without spin-orbit coupling this will not happen due to the fact that it is written in a complex basis, spherical harmonics. Best/ Lars
-
Dear Lizard! Sorry I remembered wrong: For task 102 the k-mesh is set up by np3d not ngridk. You set it up through plot3d (see manual and the Fe-FS example). If you do not give a mesh, the deafult np3d mesh is 20,20,20 which corresponds to a 21,21,21 mesh for the output to xcrysdens (well, an old bug in xcrysdens which is now a feature ..). Good luck /Lars
-
OK, it is still unclear when you say "I change ... and I add ...". What is the final elk.in file and are you sure that a new FERMISURFACE.bxfs was created? Are you aware that if you have a STATE.OUT and EFERMI.OUT from a scf run, you can calculate just task 102 with whatever k-mesh you like? Best /Lars
-
Dear Lizard, I do not understand, task 102 should use your given k-point mesh. How does your elk.in file look like? There is no limit of number of k-points what I am aware of, only those dictated by available time and memory ... Best wishes /Lars
-
Dear Joel, Although I have no personal experience with ELF plots I can at least guess the origin of your problem. As the ELF involves gradients of the density, a non-converged APW basis set will give you jumps since these basis functions do not have a continuous gradient. Hence there are two ways to improve your ELF plot: 1) converge the basis set more by increasing RGKMAX parameter - the jump will decrease but probably not disappear unless you go to infinite basis ... 2) switch to LAPW basis set,...
-
Dear Jianpeng, Please read the ref I gave, to see how to do that. Best wishes / Lars
-
Dear Jianpeng, There are some simple implementations with an operator added to the hamiltonian that couple B-field with angular momentum operator, in the line you seem to ask for, see eg Hjelm PRB 50, p4332 (1994). However, to do it properly you should get the corresponding vector potential A, given by B=grad A, and substitute the gradient in the kinetic energy with the corresponding covariant gradient that does inclue A. In a crystal with translational symmetry, this problem is non-trivial to solve...
-
Dear Jianpeng, There are some simple implementations with an operator added to the hamiltonian that couple B-field with angular momentum operator, in the line you seem to ask for, see eg Hjelm PRB 50, p4332 (1994). However, to do it properly you should get the corresponding vector potential A, given by B=grad A, and substitute the gradient in the kinetic energy with the corresponding covariant gradient that does inclue A. In a cryustal with translational symmetry, the problem to solve is non-trivial...
-
Dear Hiro, Your file has 815 lines, corresponding to 16 blocks with 500 lines with energy running from -.5 Ha to .5 Ha in each block. These blocks correspond to the s,p,d,f states resolved in their magnetic quantum number, in order (l,m)=(0,0),(1,-1),(1,0),...,(3,3). You just have to find a plotting program that can plot these partial DOS. You probably want to change the default energy window too. Best /Lars
-
Dear T Budi, The example Fe-spiral is supposed to treat this case. By varying vqlss= x,x,0 from x=0 to x=0.5, you get a spiral with q from Gamma to X. Have you tested it? You may want to incresase number of k-point to get well converged energies. In your setup you make an effort to constrain the moment along x-direction but then put a global magnetic field along z, which does not make sense. Without this global field you do not have to constrain the moments to get a planar spiral since that is stable...
-
The plan is to have it going in next version. For the moment we have an interface that is working but it needs some more tweeking to make it more efficient and updated to the latest version. Best /Lars
-
OK, I would follow Michael's suggestions. As for references, I suggest the ones I already gave. Good luck, Lars
-
Dear Harry, you do not explain how it fails. Does it converge to zero moment or does it not at all converge? Without knowing I have a problem to help. If it is non-conergence you might have to lower the beta0 parameter. However I can say that it usually better to start from an already converged solution if you increase a parameter as U, that is use task 1 instead of 0. In addition I think it is odd to use J=0 as in your calculation. Are you sure that is what you want? There is also other parameters...
-
You need higher accuracy in the Te site coordinates. Now spacegroup differ between 0.66666700 0.33333300 0.25000000 0.00000000 0.00000000 0.00000000 0.66666600 0.33333300 0.25000000 0.00000000 0.00000000 0.00000000 So I suggest you try 'Te' : spsymb, spfname 1 : nwpos 0.333333333 0.666666666 0.75 : wpos as input. /Lars
-
You have in both cases ! reduction to primitive cell : T you have to switch off this flag. /Lars
-
It might be units, Hartree vs Rydbergs ...
-
It might be units, Hatree vs Rydbergs ...
-
you might get good/decent results for values of RGKMAX as low as 4 or lower. At least...
-
Dear Bradford, your choice of basis set is unnecessary large, or rather your lack...
-
Dear Bradford, your choice of basis set is unnecessary large, or rather your lack...
-
I have a similar setup of slurm and for me the following setup works: In elk.in scrpath...
-
Dear SS (or whatever you call yourself), you have not specified why you ask this...
-
Dear SS (or whatever you call yourself), you have not specified why you ask this...
-
Dear SS (or whatever you call yourself), you have not specified why you ask this...
-
Yes. There has been a sleeping project for a while, but recently we have started...
-
Dear Harry, your information is very sparse and it is not clear what results you...
-
There is nothing wrong with either elk or spacegroup. Your calculations are probably...
-
Are you sure you are using a.u. and not Ångström as length units? Best wishes, L...
-
Dear Dae kwang Jun, You are supposed to give the k-points in fractions of reciprocal...
-
Hi Pietro! I sincerely hope you are not Italian, then you do not deserve any answer;-(...
-
Dear Fei! That is the difference between an all electron method and a pseudo-potential...
-
If you read this post, you will find that Markus has answered your question. PDOS...
-
Hi Michael, is it not that the code does not know what quantization axis it ought...
-
Dear Pietro, My argument above was wrong. The rotation direction is independent of...
-
Dear Pietro, My argument above was wrong. The rotation direction is independent of...
-
Hi Pietro, OK now I understand what you mean. First I want to remind that the spiral...
-
Dear Pietro, I would benefit the understanding if you would present the elk.in files...
-
Since you are interested in small energy differences it could be wise to choose the...
-
Dear all, Yes. I was involved in calculating elastic constants long ago and they...
-
Dear Praveen, I agree with Markus two points. I just want to add that your curve...
-
Hi Pietro! As Marc stated DFT+U is a variational approach, if the double counting...
-
Hi! You should use more digits in atomic positions in spacegroup.in, instead of 0.3333...
-
In Elk time inversion symmetry is utilised in case of spin-independent calculations...
-
Dear Lukas, Since spin-orbit coupling cannot be included in a spin spiral calculation,...
-
Please try the latest version which includes a patch for electron-phonon coupling....
-
OK! I might have misunderstand. I thought you recalculated the E-V data ... /Lar...
-
Hi! OK, the make.inc files seems to be very different in between individuals, so...
-
Hi! Do you have primcell=.true.? Best, Lars
-
Hi Paul, I have now actually read the article by de Boer and de Groot, and noticed...
-
Dear Paul, I am sorry I missed that there were two 4s lo's. You have to get rid of...
-
Have you tried the (new) flag ncgga to be true, it fixes some problems with gga combined...
1 >