Home

1 Runwien user`s and reference guide Contents

1. label replaces the atom i integer the RMT order affects all the atoms in the crystal Default AUTO let WIEN2k determine it if no atom is indicated 1 is assumed RO atom i AUTO range r First radial mesh point in bohr for the indicated non equivalent atom Wien2k manual recommends a value between 0 0005 and 0 00005 Similarly to most orders in this section AUTO selects the standard WIEN2k value A list of values a range or even a list of ranges is also acceptable 47 A RO order affects to a single atom type by default to the first atom in the input indicated by the integer atom i The number atom i is a non equivalent atom index pointing to the corresponding EQUIV LIST The non equivalent atom index are assigned in order of appearance of the EQUIV LIST environment 1 for the first atom 2 for the second atom etc If your order omits the index of the atom which amounts to applying the RO variaton to the first atom be careful to avoid white characters within the declared range If a label replaces the atom i integer the RO order affects all the atoms in the crystal Default 0 0001 If no atom index is indicated 1 is assumed RKMAX AUTO range r Give a value or a list of values for the RMT KMAX product where KMAX is the maximum k for plane waves and RMT is the smallest of the muffin tin radii of the different types of atoms This parameter does not appear on the struct input file bu
2. 38 p2lab p2xa 2vba P pc21b p2yb 2zcb P pbc21 p2zc 2ybc P p2lca p2xa 2zca P pb2la p2yb 2xab P pnc2 p2z 2xbc P p2na p2x 2yca P pb2n p2y 2zab P pcn2 p2z 2yac P p2an p2x 2zba P pn2b p2y 2xcb P pmn21 p2zac 2x P p21mn p2xba 2y P pn21m p2ycb 2z P pnm21 p2zbc 2y P p21nm p2xca 2z P pm21n p2yab 2x P pba2 p2z 2xab P p2cb p2x 2ybc P pc2a p2y 2zca P pna21 p2zc 2xn P p21nb p2xa 2yn P pc21n p2yb 2zn P pbn21 p2zc 2yn P p21cn p2xa 2zn P pn2la p2yb 2xn P pnn2 p2z 2xn P p2nn p2x 2yn P pn2n p2y 2zn P cmm2 c2Z 2x CXY a2mm a2x 2y CYZ bm2m b2y 2z CXZ cmc21 c2zc 2X CXY a21ma a2xa 2y CYZ bb21m b2yb 2z CXZ ccm21 C2ZC 2V CXY a2lam a2xa 2z CYZ bm21b b2yb 2x CXZ ccc2 c2z 2x CXY a2aa a2x 2ya CYZ bb2b b2y 2zb CXZ amm2 a2z 2x CYZ b2mm b2x 2y CXZ 39 cm2m c2y 2z CXY abm2 a2z 2xb CYZ b2cm b2x 2yc CXZ cm2a c2y 2za CXY bma2 b2z 2ya CXZ c2mb c2x 2zb CXY ac2m a2y 2xc CYZ ama2 a2z 2xa CYZ b2mb b2x 2yb CXZ cc2m c2y3 2z CXY bbm2 b2z 2yb CXZ c2cm 2x3 2z CXY am2a a2y 2xa CYZ aba2 a2z 2xab CYZ b2cb b2x 2ybc CXZ cc2a c2y 2zca CXY bba2 b2z 2yba CXZ c2cb c2x 2zcb CXY ac2a a2y 2xac CYZ fmm2 f2z 2x F f2mm
3. Default 0001 XEND x r y r z r den r X end of the plane Same syntax as above Crystallographic units Default 1001 YEND x r y r z r den r Y end of the plane Same syntax as above Crystallographic units Default 0101 NSHELLS nx i ny i nz i Number of unit cell shells in each direction which will be used to compute the electron density Default 3 3 3 NPT nx i ny i Number of points in each axis Default 100 100 ZMIN zmin r Lowest bound for electron density graph in atomic units e bohr73 This parameter will be passed down to a gnuplot script so a value like is allowed meaning automatic fit to the lower bound of the function Default ZMAX zmax r Upper bound for electron density graph in atomic units e bohr 3 This parameter will be passed down to a gnuplot script so value is allowed meaning automatic fit to the upper bound of the function Default DC dc i Number of contour lines used Default 20 TYPE 3D C Choose between a 3d representation of the electron density or a 2d contour plot Default 3d SCALE NORMAL LOG Select a normal or logarithmic representation of the plotted scalar func tion Default normal NOLABELS Do not include labels on the plot Default false DO range i ALL NONE NEW OLD Used together with the EXCEPT order below EXCEPT range i As a default printrho will be run only on the new points skipping those points already examined The DO EXCEPT
4. Default numerical 0 if lt value gt is not specified DEBYE STATIC DEBYESTATIC DEBYEITER DEBYESTATICBV value r Select the method used to compute the debye temperature of the crystal Valid elections are e static static EOS no thermal effects e debyestatic get the Debye temperature from computed static bulk moduli e debyeiter get the Debye temperature from computed adiabatic bulk moduli self consistent approach e debyestaticbv Debye temperature from static bulk moduli at static equilibrium volume lt value gt is poisson s ratio only used if debyestatic or debyeiter Default debyestatic PRESSURE AUTO range r Specify pressures for gibbs Tf the list of pressures specified ex ceeds the possibilities of your data it will be cut down to an appropriate maximum pressure on gibbs error Default a single pressure p 0 gibbs04 M A Blanco E Francisco and V Lua a GIBBS isothermal isobaric thermodynamics of solids from energy curves using a quasi harmonic Debye model Comput Phys Commun 158 2004 57 72 89 TEMPERATURE AUTO range r Specify temperatures for gibbs Default one temperature T 0 FREE ENERGY energ r Value of free in vacuo energy for the molecular formula This value will only be used to estimate the formation energy of the crystal It is possible to run gibbs in absence of this datum Runwien will try to use the results of a previous FREE section if it is av
5. y 0 70937250 z 0 20566670 y 0 91769900 z 0 20566670 96 64 32 C 2 4 1 6 CI 18 2 28 2 2P 2 C 4 11000e 04 Fe 12 14 26 Fe 18 2 28 2 2P 6 38 2 Fe 3 47700e 03 H ili H 18 1 H 0 00000e 00 2545 1312760 75 6207230 0 9665040 3311 0035460 000 18 general rmt atom 2 1 25000 general rmt atom 3 1 25000 general rmt atom 4 1 25000 general rmt atom 5 1 25000 general rmt atom 6 1 25000 general rmt atom 7 0 82000 general rmt atom 8 0 82000 general rmt atom 9 0 82000 general rmt atom 10 0 82000 general rmt atom 11 0 82000 general ro 0 00050000 general rkmax 3 00000 general lmax 12 general lnsmax 5 general gmax 23 000 general mix 0 40000 general kpts 1 initialization potential ggapbe96 initialization ecoreval ry 6 000 initialization emin ry 9 00000000 initialization emax ry 2 00000000 scf commands used i 100 ec 0 00001 it 6 Only one structure no variable parameters Energy ry 3318 8725220000 E fermi ry 0 13297 No critic data available Output index plot and gnuplot files out files scfsummary out synopsis out index files temp index ps files gnuplot files As can be seen a very complicated structure can be calculated easily by using runwien adequately Next steps towards ferrocene electronic structure could include examining kpts and rkmax effect on energy and properties
6. 57 NOSUMMARY Don t run the summarizing script This script will generate a scfsummary out file in root directory containing information about the results of the scf process Default false DO range i ALL NONE NEW OLD Used together with the EXCEPT order below EXCEPT range i The DO EXCEPT pair is used to select a subset of calculations to run when more that one calculation is implied from the input data The behavior of this orders is described in the GLOBAL section Default Use the local DO EXCEPT pair if present otherwise apply the global DO EXCEPT pair if present if neither is defined calculate only new structures The DO and EXCEPT orders are frequently used the SCF section for instance to repeat the costly SCF process on a particular geometry that has failed to converge while keeping intact the geometries already converged As a matter of fact WIEN2k internal orders like run lapw can be run manually if necessary Default if no do or except was specified in scf and global calculate only new structures If there is a global do or except use it 3 6 3 Requirements of the scf section General initialization and prescf sections are required 3 7 Spinorbit section 3 7 1 Syntax runwien SPINORBIT DEFAULT DIRECTION h i k i l i ADDLO at i el r de r EXCLUDE ati i at2 i NEWEMAX emax r NEWKPTS kpts r MAX ITERATIONS maxit i CHARGE CONV cconv r ENERGY CONV econv r FORCE CONV fconv r NICE nice value i
7. ITDIAG Levc il NEW INI cyc i DO range i ALL NONE NEW OLD EXCEPT range i END SPINORBIT Re runs the scf cycle including the spin orbit interaction through a second variational method see WIEN2k user s guide for details This section is run right after the scf cycle using the converged orbitals as the starting point for the new scf The command runwien spinorbit default 58 will invoke spinorbit assuming default values on all variables In spinpolarized cases a reduction in the symmetry of the system may occur Note that the new struct file generated by symmetso is adopted but it is not parsed so runwien will still use the old structural information Also in cases where LDA U is being used the inorb input for the orb program and indm input for the lapwdm program are not adapted This shortcomings may be fixed in later versions of runwien 3 7 2 Keywords and environments DIRECTION h i k i Li Chooses the direction of magnetization Default 001 ADDLO at i el r de r Add an LO with a p1 2 radial function to the non equivalent atom at i with energy el r If de r is not zero the energy is automatically adjusted with steps de r Default el r 4 97 de r 0 005 EXCLUDE atl i at2 i Exclude the spin orbit correction for the non equivalent atoms at1 i at2 i Default NEWEMAX emax r This keyword allows the user to change the upper limit of the eigenvalue window calculated in lapw1 by writin
8. JOIN A UP DN BIUPIDNIPI C UP DN In case you are reading the ASCII documen tation stands for the asterisk character Join several DOS in one plot a b c stand for DOS plot numbers corresponding to line numbers in the dos list environment up and dn flags are only applicable in spinpolarized calcu lations and indicate the DOS spin that is to be used if none is indicated then total DOS flag instructs runwien to multiply the DOS by the multiplicity of the non equivalent atom Note is not permitted for interstitial nneq 1 or total 0 DOS plots up and dn can exchange positions as long as they appear after the dos line number and they form one field Default PLOTXMIN plotxmin r EV Lower bound for energy in dos graph in Ry If eV is indicated runwien will assume lt plotxmin gt is in eV This value will be passed to the gnuplot script so is acceptable input meaning that gnuplot automatically fits the graph limit Default PLOTXMAX plotxmax r EV Upper bound for energy in dos graph in Ry If eV is indicated runwien will assume lt plotxmax gt is in eV This value will be passed to the gnuplot script so is acceptable input meaning that gnuplot automatically fits the graph limit Default INIMAXENERGY in1maxenergy r EV Upper bound for extending the eigenvalue calculation using lapw1 in Ry If eV is indicated runwien will assume inlmaxenergy is in eV Defaul
9. in Ry If eV is indicated runwien will assume lt emin gt is in eV Default lower bound for lowest energy band 2 0 Ry if nothing appropriate is found ENERGYMAX emax r EV Upper energy bound for density of states mesh calculation in Ry If eV is indicated runwien will assume lt emax gt is in eV Default upper bound for highest energy band 1 0 Ry if nothing appropriate is found DE de r EV Energy increment in the mesh in Ry If eV is indicated runwien will assume lt de gt is in eV Default 0 0025 Ry BROAD broad r Gauss broadening factor Default 0 003 DOS LIST environment The syntax of this environment corresponds to 72 runwien DOS LIST atom i descr r label s END DOS LIST Sequence of lines each of them indicating a DOS calculation Atom represents the atom for which the dos is calculated 0 for total dos nneq 1 is the interstitial region where nneq is the number of non equivalent atoms descr refers to the column of the qtl file 1 total 2 s 3 p label is an identification text Default 0 1 tot SPIN MERGE NEW NO Specifies the treatment of spin up and spin down density of states Merge means dos up dos down and total dos will share the same plot New dos up dos down and total dos each of them will occupy an individual plot No dos up and dos down will be in separate plots and total dos will not be plotted The spin keyword can only be used in spinpolarized calculations Default new
10. optimization of rmt etc All this steps can be synthesized in just a few lines of input and all the output generated by them is extracted sorted and presented in a comfortable way 2 8 Full WIEN2k calculation in sodium chloride Next a calculation of the sodium chloride crystal is presented in full detail The geometry of the unit cell of NaCl can be retrieved from a public crystallography database such as the COD crystallography open database http www crystallography net or alternatively inserted by hand In our example we have chosen to download the cif file from the COD and store it as nacl cif in the nacl directory which is to hold our calculations tasciit mkdir nacl 19 cd nacl In a simple crystal such as NaCl it is reasonable to start with an all default calculation up to the SCF cycle This requires a nacl wien file in nacl containing runwien general loadcif nacl cif end general initialization default prescf default scf default synopsis default The job is run using tasciit runwien awk nacl wien In approximatelv one minute the calculation ends Vou can check that the geometrv of the unit cell is the correct one bv running xervsden on the struct file ascii xcrysden wien struct nacli nacli struct Also you can readily examine the struct file and check that everything is in order The synopsis out file contains some useful information Some important things to check are that the act
11. up to 10 in recent versions of WIEN2k so redimensioning is not necessary Tf there are atoms of very low local symmetry 1 m etc and the LM list is written up to LMAX 10 it may happen that an additional parameter has to be redimensioned depending on your choice of Insmax This parameter is NGAU in the param inc file of the lapwl program that controls the maximum number of gaunt coefficients calculated The default value for NGAU is 2350 We have found a value of 2650 enough for an atom in local symmetry 1 with a LM list up to ten and Insmax 5 Default set by wien LM LIST neq_atom i Imax i Same as above but the LM pairs are automatically generated by run wien awk This involves parsing the output of the sgroup program outputsgroup reading the local symmetry of each center and generating the appropriate list of LMs up to a maximum L of lmax i for the non equivalent atom neg atom i This form of LM LIST can not be used in cases where two equivalent atoms in the unit cell are forced into non equivalence see for instance antiferromagnetic Cr test023 In such cases use the LM LIST environment above The specific list of LM pairs for each local symmetry is shown below LM pairs up to LMAX 10 Loc sym LM list 52 001011 11222120 21 2233323130 31 32 3344 43424140 41 42 43 44555453525150 51 52 5 3 54 5566656463626160 61 62 63 64 65 6677 167574737271 70 71 72 73 74 75 76 7788878 6858483828180 81 82
12. with data from the initialization section If a previous runwien calculation must be loaded use LOADCHECK Otherwise runwien will issue an error when it finds the old directories and exit The loaded structures become old and are not messed with unless required by an do or except keyword Also you can not change non variable parameters from the old calculation such as cell constants atomic positions etc Should you want to do this start another runwien job in a different directory The creation of structures using variables parameters is fairly easy These variable parameters are npt rmt r0 rkmax Imax Insmax gmax mix and kpts see below for a definition The user can vary one of this parameters with a line like runwien rkmax 7 5 8 0 8 5 Thus generating three structures each with a different rkmax value but all with the same values for the rest of the options If the user adds another variable parameter trunwient rkmax 7 5 8 0 8 5 kpts 1000 2000 There will be 6 structures whose values will be those of the direct product of both sets There is a mechanism to avoid the coupling of variables the also kevword When an also kevword appears further variable parameter specifications will be independent of the previous For example runwien rkmax 7 5 8 0 8 5 also kpts 1000 2000 will only create five structures 3 with rkmax 7 5 8 0 and 8 5 and the default value for kpts and 2 with kpts 1000 2000 and t
13. 20 0 end general Use default values for the initialization and prescf sections initialization default prescf default scf Increase the default number of scf iterations and ask for a convergence of 0 00001 Ry in the energy max iterations 50 energy conv 0 00001 end scf synopsis exhaustive end synopsis This input is divided into sections each one corresponding to a main part in a WIEN2K calculation There are at least four sections required for all meaningful jobs that must be entered into a precise order e The general section is used to enter a reference geometry for the crystal lattice lengths and angles space group symmetry atoms in the main cell etc and give values to the main parameters for the FP LAPW method muffin tin spheres number of sample points in the reciprocal cell etc The user can ask here for an exploration of one or more of the model parameters but the crystal geometry is held fixed The initialization section is used to decide low level but important details of the FP LAPW model The exchange and correlation functional core valence separation energy local orbital special treatments etc are defined in this section The default values are correct in many cases The prescf section is in charge of preparing the sampling of k waves and the initial electron density The default values are usually enough for the corresponding parameters The scf section controls the self consistent iteration process
14. 83 84 85 86 87 88999 89796959493929190 91 92 93 94 95 96 97 9 8 991010109108107106105104103102101100 101 10 2 10 3 10 4 10 5 10 6 10 7 10 8 10 9 10 10 00 22 21202122 44 43 42 414041424344 66 6 5 64 63 6 2 6160616263646566 8 8 87 8 6 8 5 8 4 83 82 81808182838485868788 1010109 108 10 7106105104103 102101 100 101 10 2 10 3 10 4 10 5 10 6 10 7 10 8 10 9 10 10 2 00102022 223032 324042 4244 445052 5254 546062 6264 6466 668082 8284 8486 8688 8 8 10 0 10 2 10 2 10 4 10 4 10 6 10 6 10 8 10 8 10 10 10 10 m 0011 112022 2231 3133 334042 4244 4451 5 153 5355 556062 6264 6466 6671 7173 7375 7577 778082 8284 8486 8688 8891 9193 939 5 9597 9799 99100102 102104 10 410 6 10 6 10 8 10 8 10 10 10 10 2 m 00 222022 44 42404244 66 64 6260626466 8 8 86 84 828082848688 10 10 10 8 10 6 10 4 10 2 10 0102104106108 1010 222 00202230 3240424450 52 546062646670 72 74 76808284868890 92 94 96 98 100102104 10 6 10 8 10 10 mm2 001020223032404244505254606264667072 74768082848688909294969810010210410610 8 10 10 mmm 002022404244606264668082848688100 102 104106 10 8 10 10 4 0010203040 444450 545460 646470 747480 8484 888890 9494 989810010 410410 8 108 4 002032 3240 444452 5260 646472 7276 7680 8484 8888972 9296 9610 0 10 4 10 4 10 8 10 8 4 m 002040 444460 646480 8484 8888100 104104 10 8 10 8 422 00102030404450 5454606470 74748084889 0 94 98100104108 4mm 0010203040445054606470748084889094
15. Add a spinorbit section in wien input file to all the structures in the elastic calculation The lines included in the body of the environment are passed verbatim to the corresponding wien input files The information in this environment and the following is not parsed runwien will not be able to tell you if it is right or wrong until it is calculating the elastic structures CLEAN WIEN FULL Append a clean wien full at the end of each deformation point wien This will make each point in elastic section clean itself after being calculated For details on clean keyword see above global 69 NOSEND Create the wien input files but do not run the calculation of the elastic constants Useful if you plan to do it manually and then reading it back with a LOADCHECK REREAD Default 3 9 3 Requirements of the elastic section General initialization and prescf sections are required 3 10 Printrho section 3 10 1 Syntax runwien PRINTRHO RHO TOTALRHO VALRHO SPIN ATOMIC DEFORM VTOTAL VCOUL VXC ENERGYMIN emin r ORIGIN x r y r z r den r XEND x r y r z r den r YEND x r y r z r den r NSHELLS nx i ny i nz i NPT nx i ny i ZMIN zmin r ZMAX zmax r DC dc i TYPE 3D C SCALE NORMAL LOG NOLABELS DO range i ALL NONE NEW OLD EXCEPT range i END PRINTRHO Using a general structure prints its electron density valence electron density deformation density spin density atomic promolecular den
16. a large number of Fortran and C codes that exchange information through a set of ASCII and binary data files A collection of shell scripts is used to drive the main tasks The user is expected to work through a web interface that greatly simplifies the trouble of creating and formatting the several input files needed for any calculation However we have felt the need of a tool capable of processing large series of calculations such as a potential energy surface exploration Therefore we have devised a new approach better suited for large scale production runs runwien Runwien is a new script driven interface for the WIEN2K package The user is expected to write a single file describing the system to be calculated and the tasks to perform and runwien is in charge of the ordered execution of the many programs integrated in the WIEN2K bundle The input file can be quite short and simple because runwien knows default values for many of the internal parameters governing the electronic structure calculation On the other hand direct control of every relevant parameter is also available Exploring the convergence of the results with respect to the internal parameters is facilitated because a grid instead of a fixed value can be used for one or more of the parameters Similarly a potential energy surface can be explored by using a grid on one or more of the independent variables that control the system geometry The execution concludes with the production o
17. b eee 77 3 15 3 Requirements of the aim section 0000 ee 77 3 16 CritiGseChiOn sec eee ew eaaa en bade pp bags cadet ae CHEER RS Swe EE 78 IN 2 2424 RENE ak bbe dade ea ewe elise ea oe ee ok Eko 4 78 3 162 Keywords and environments 2 Ls aos em vie ee obs BARRE DS 79 3 16 3 Requirements of critic section s o oeoo eee 80 SC SWEEP SECHON 42 aa 2 ho no Ba RAWSKA KA RRR RE eR W RO AS Ba 80 Salil ylides A IA 80 3 17 2 Keywords and environments s s s sisa ate a a 000002 p ay a ao ee eee 82 3 17 3 Requirements of the sweep section LL o e 88 a DIN SCHIOUN 4 rira n at vi aa ee d E eee r z eee A A 89 JAGA TOYA osad bh ee ARA 89 3 18 2 Keywords and environments ss s a sg ggg gek a eee eee eee 89 3 18 3 Requirements of the gibbs section e e eu 90 3 19 Synopsis section lt o s sise wan a W 822 SY Yee ee eed ja 90 3191 SMA ew gw eke Bon DARE EEE ER RRA KW dat a BR 90 3 19 2 Keywords and environments ao sosoo eee 90 3 19 3 Requirements of the synopsis section 000000 90 Directory tree for runwien File locations 91 Test files 91 Programming techniques used in runwien 93 In case something goes wrong 94 Copyright notice 94 2 User s Guide 2 1 Introduction Developed by K Schwartz P Blaha and their coworkers the WIEN2K WIEN2k electronic structure package performs Full Potential Linear Augmented Plane Wave FP LAPW calculations on solids The package is made of
18. calculation v 1 0 5 root temp pwd home alberto temp machine XXXX date mie oct 24 00 07 22 CEST 2007 Sections run and time information general section run time s 0 initialization section run time s 1 prescf section run time s 0 scf section run time s 0 sweep section run time s 1507 total run time s 1508 Structural information Mol formula Bel Mol mass 9 012182 a bohr 4 32104800 b bohr 4 32104800 c bohr 6 77145600 alpha deg 90 00000000 beta deg 90 00000000 gamma deg 120 00000000 c a 1 56708650 c b 1 56708650 b a 1 00000000 volume bohr73 109 49413924 Mol volume 54 74706962 space group P63 mmc lattice H system hexagonal nonequivalent atom list in struct and position of a representative atom atom 1 is Be mult 2 x 0 66666667 y 0 33333333 z 0 75000000 Additional information Number of electrons in unit cell 8 12 Valence electrons in unit cell 8 Core electrons in unit cell Core val Core val el Core leaking e for atom Atomic energy lstart Be Total in vacuo energy lstart Calculation fixed parameters spinpolarized relativistic first density general npt general rmt atom 1 general ro general rkmax general general general general mix general kpts initialization initialization lmax lnsmax gmax potential ecoreval ry emin
19. can be used to modify this behavior At difference from other sections the global DO EXCEPT pair do not affect sweep Default Use the local DO EXCEPT specification if available Do the sweep task only on new points otherwise 3 17 3 Requirements of the sweep section Requires general initialization prescf and scf sections 88 3 18 Gibbs section 3 18 1 Syntax runwien GIBBS DEFAULT EOS NUMERICAL VINET BIRCH BCNT VINET BIRCH BCNT value r DEBYE STATIC DEBYESTATIC DEBYEITER DEBYESTATICBV value r PRESSURE AUTO range r TEMPERATURE AUTO range r FREE ENERGY energy r END GIBBS The gibbs code by Blanco et al gibbs04 uses the energy versus volume curve and a quasi harmonic Debye model to estimate the thermal and pressure effects on the crystal geometry The equation of state EOS bulk modulus thermal expansion etc are calculated in the process Runwien s GIBBS task uses the energies from a sweep run to prepare the input for the external gibbs code and extract a selection of the most important results The minimal code that activates this section is runwienti gibbs default It will run gibbs section assuming default values for all variables 3 18 2 Keywords and environments EOS NUMERICAL VINET BIRCH BCNT VINET BIRCH BCNT value r Type of equa tion of state employed to adjust data If bent or numerical bent a value is required See gibbs manual for more info
20. changing the cell parameters is heavily used in the sweep section see below Keep in mind that crystallographic coordinates like 1 3 1 6 and similar should be given with a large precision to prevent ghost duplicacies of atoms when applying the symmetry operators Default DO range i ALL NONE NEW OLD Used together with the EXCEPT order below EXCEPT range i The DO EXCEPT pair is used to select a subset of calculations to run when more that one calculation is implied from the input data The behavior of this orders is described in the GLOBAL section LDA U environment Formed as LDA U SIC AMF HMF iat1 i 11 i l1 s defu r defj r iat2 i 12 i 12 s defu r defj r END LDA U 49 Includes an orbital dependent correction to the exchange correlation potential This method allows LDA or GGA calculations to successfully reproduce the properties of the so called strongly correlated systems Common systems of this kind are transition metal oxides and rare earth elements and compounds For more information see the WIEN2k user s guide and references therein The first keyword in the environment selects the flavor of LDA U e SIC Self Interaction Correction Anisimov et al 1993 e AMF Around the Mean Field Czyzyk et al 1994 e HMF Hubbard model in Mean Field Anisimov et al 1991 Each line in the LDA U environment activates the U correction for a particular atom and atomic symmetry 1 The first
21. cm c 2x CXY am a 2y CYZ bm b 2z CXZ bm b 2x CXZ cm c 2y CXY am a 2z CYZ cc c 2xc CXY aa a 2ya CYZ bb b 2zb CXZ bb b 2xb CXZ ce c 2yc CXY aa a 2za CYZ p2 m p2x P p2 m p2y P p2 m p2z P p21 m p2xa P p21 m p2yb P p21 m p2zc P c2 m C2x CXY a2 m a2y CYZ b2 m b2z CXZ b2 m b2x CXZ c2 m c2y CXY a2 m a2z CYZ p2 c p2xc P p2 a p2ya P p2 b p2zb P p2 b p2xb P p2 c p2yc P p2 a p2za P p2 n p2xbc P p2 n p2yac P p2 n p2zab P p21 c p2xca P p2l a p2yab P p21 b p2zbc P p21 b p2xba P p21 c p2ycb P p21 a p2zac P p21 n p2xabc P p21 n p2yabc P 37 p21 n p2zabc P c2 c C2xc CXY a2 a a2ya CYZ b2 b b2zb CXZ b2 b b2xb CXZ c2 c c2yc CXY a2 a a2za CYZ p222 p2z 2x P p2221 p2zc 2x P p2122 p2xa 2y P p2212 p2yb 2z P p21212 p2z 2xab P p22121 p2x 2ybc P p21221 p2y3 2zca P p212121 p2zac 2xab P c2221 c2zc 2x CXY a2122 a2xa 2y CYZ b2212 b2yb 2z CXZ c222 c2z 2x CXY a222 a2x 2y CYZ b222 b2y 2z CXZ f222 f2z 2x F 1222 122 2x B 1212121 i2zac 2xab B pmm2 p2z 2x P p2mm p2x 2y P pm2m p2y 2z P pmc21 p2zc 2x P p21ma p2xa 2y P pb21m p2yb 2z P pem21 p2zc 2y P p2lam p2xa 2z P pm21b p2yb 2x P pcc2 p2z 2xc P p2aa p2x 2ya P pb2b p2y 2zb P pma2 p2z 2xa P p2mb p2x 2yb P pc2m p2y 2zc P pbm2 p2z 2yb P p2cm p2x 2zc P pm2a p2y 2xa P pca21 p2zc 2xac P
22. file keyword is used with the path and name of the file with or without k list K list files are easily created using xcrysden available at http www xcrysden org Default according to the lattice type a template file will be selected bcc klist for B fcc klist for F simple cubic klist for P and hcp klist for H If the lattice type is not one of those an error will be risen and no band plot will be done INIMAXENERGY inlmaxenergy r EV Upper bound for the eigenvalue calculation using lapwi in Ry If eV is indicated runwien will assume inlmaxenergy r is in eV Default maxenergy in initialization section DO range i ALL NONE NEW OLD Used together with the EXCEPT order below EXCEPT range i As a default bandplot will be run only on the new points skipping those points already examined The DO EXCEPT orders can be used to modify this behavior Default Use the local DO EXCEPT specification if available Use the global DO EXCEPT data otherwise In absence of user orders do the bandplot task only on new points 3 12 3 Requirements of bandplot section General initialization prescf and scf sections are required 3 13 Kdos section 3 13 1 Syntax trunwient 74 KDOS PLOTXMIN plotxmin r RY PLOTXMAX plotxmax r RY PLOTYMAX plotymax r KDOS ndos i UP DN MERGE UP DN DO range i ALL NONE NEW OLD EXCEPT range i END KDOS Print a diagram combining the band structure spaghetti plots K with the d
23. files for the calculation Runs in sequence a host of programs related to symmetry initial density generation see WIEN2k manual for more information about the process Also completes the struct file with the options required by the user Specifically the programs run are setrmt_lapw sets the rmt if not given by the user instgen lapw nn sgroup symmetry and lstart Also prepares inl in2 and inm files runwienti initialization default Runs initialization assuming all default parameters It is equivalent to runwien initialization end initialization 3 4 2 Keywords and environments XCPOTENTIAL LSDA GGAPBE96 GGAPW91 GGAWCO06 Chooses one of the available exchange correlation potentials Default ggapbe96 ECOREVAL core energy r Initial core valence separation energy in Ry This energy value is compared to the results of the atomic calculation and core and valence states are thus assigned This assignment does not vary during the scf cycle unlike the band energies Default 6 0 ENERGYMIN min_energy r Lower energy cutoff for eigenvalue search inl energy min Default auto wien2k determines which is best ENERGYMAX max energy r Higher energy cutoff for eigenvalue search inl energy max Default auto NNFACTOR nn_dist r Search for nearest neighbors nn in structure up to a distance of nnfactor nn distance Default 2 51 ORBITALS environment The syntax of this environment corresponds to
24. flatness is relatively insensitive to reasonable variations in the muffin tin radius of atoms The cause for this is that the flatness is defined as a ratio of densities both of them corresponding to points that are almost always located in the interstitial region Both energy and flatness plots against the rmt of both atoms do not show any clear trend As we are interested in PES exploration we will go on with rmt 2 for both atoms to avoid muffin tin collisions The NaCl crystal is highly ionic so saturation of k points is achieved easily In this calculation it makes no major difference whether the system is calculated using 2000 or 10000 k points as both energy and flatness remain unchanged Concerning RKmax the flatness seems quite insensitive to a variation in this parameter The energy decreases on increasing RKmax and a minimum is not observed Contrary to kpts rkmax is not a variational parameter and an increase in energy usually indicates problems related to linear dependencies in the basis set We have selected kpts 2000 and rkmax 7 5 in order to keep the computational cost low but a better rkmax is needed if the results are to be meaningful The selected parameters correspond to the first structure studied in the parameter exploration Just for checking we can calculate the density of states band structure and kdos K points and DOS for it To do this rewrite the wien input file 24 runwien loadcheck do 1 dos
25. format lt neq atom gt is the number of the non equivalent atoms for which the basis is specified global energy is the default linearization energy and global apw 1 if apw method basis function is to be used 0 if 62 lapw First lt orb 1 gt appearing in the list refers to local orbitals plane waves basis functions of that l Subsequent orb l will refer to local orbitals orb energy is the linearization energy orb var gt 0 commands wien to search for a good linearization energy and if stop is set stop in case it is not found Finally use lt apw gt 1 in case you want an apw basis function type 0 in case of lapw See WIEN2k s manual for more information about basis Default reference structure LM list The syntax of this environment corresponds to runwien LM LIST neq_atom i 11 mi 12 m2 13 m3 lk mk 1k 1 mk 1 END LM LIST Specify LM lattice harmonics expansion of charge density See wien s manual for L M pairs adequate for each structure Default reference structure FERMI ROOT TEMP TEMPS GAUSS TETRA ALL value r Method to find fermi energy See WIEN2k s manual for details on each of the method Value is a method specific characteristic number Default reference structure MAX ITERATIONS maxit i Maximum number of cycles Default reference structure CHARGE CONV cconv r Charge convergence criterion Default reference structure ENERGY CONV econv r Energy convergence criterion Default r
26. local symmetry in the unit cell does not correspond to the points where the inversion center is located In such cases a second origin is defined at these points In WIEN2k it is computationally cheaper to place the origin at a point where inversion symmetry is present so that the real versions of the code are used Therefore in cases where ambiguity could arise spacegroup assumes the second origin has been selected The space group label follows the format accepted by spacegroup which is also used in the w2web interface The exhaustive list of space group labels follows The lattice of the struct file generated using the corresponding space group is also indicated Note that there are more Spacegroup internal label s than Short label s This allows the user to specify the orientation of a given privileged axis in the crystal Space group labels Short label Spacegroup internal label Associated lattice pl pl P p 1 pl P p2 p2x P p2 p2y P p2 p2z P p21 p2xa P p21 p2yb P p21 p2zc P c2 c2x CXY a2 a2y CYZ b2 b2z CXZ b2 b2x CXZ c2 c2y CXY a2 a2z CYZ pm p 2x P pm p 2y P pm p 2z P pe p 2xc P pa p 2ya P pb p 2zb P pb p 2xb P pe p 2yc P pa p 2za P pn p 2xbe P 36 pn p 2yac P pn p 2zab P
27. p 2yca 2zc P pbcn p 2zn 2xab P pnca p 2xn 2ybc P pbna p 2yn 2zca P pcnb p 2yn 2xac P pcan p 2zn 2yba P pnab p 2xn 2zcb P pbca p 2zac 2xab P pcab p 2yab 2xac P pnma p 2zac 2xn P pbnm p 2xba 2yn P pmcn p 2ycb 2zn P pnam p 2yab 2xn P pmnb p 2zbc 2yn P pemn p 2xca 2zn P cmcm C 2zZC 2x CXY amma a 2xa 2y CYZ bbmm b 2yb 2z CXZ bmmb b 2yb 2x CXZ ccmm C 2zC 2y CXY amam a 2xa 2z CYZ cmca C 2DzaC 2x CXY abma a 2xba 2y CYZ bbem b 2ycb 2z CXZ bmab b 2yab 2x CXZ ccmb C 2zbc 2y CXY acam a 2xca 2z CYZ cmmm C 22 2xX CXY ammm a 2x 2y CYZ bmmm b 2y 2z CXZ cccm C 22 2xc CXY amaa a 2x 2ya CYZ bbmb b 2y 2zb CXZ cmma C 2za 2x CXY abmm a 2xb 2y CYZ bmcm b 2yc 2z CXZ bmam b 2ya 2x CXZ cmmb c 2zb 2y CXY acmm a 2xC 2z CYZ ccca C 2za 2xac CXY abaa a 2xb 2yba CYZ bbcb b 2yc 2zcb CXZ bbab b 2ya 2xab CXZ 42 cccb c 2zb 2ybc CXY acaa a 2XC 2ZCa CYZ fmmm f 22 2x F fddd f 2zuv 2xvw F immm 1 22 2X B ibam i 22 2xab B imcb i 2x 2ybc B icma i 2y 2zca B ibca i 2zac 2xab B icab i 2yab 2xac B imma il 2zac 2x B ibmm i 2xba 2y B imcm i 2ycb 2z B imam i 2yab 2x B immb i 2zbc 2y B icmm 1 2xca 2z B p4 p4 P p41 p41 P p42 p4c P p43 p43 P i4 i4 B 141 141b B p 4 p 4 P 1 4 1 4 B p4 m p4 P p42 m p4c
28. right or wrong until it is calculating the sweep structures Sweep section combined with elastic section can build up an impressive amount of data Use this with care PRINTRHO environment The syntax of this environment corresponds to runwien PRINTRHO item_list i END PRINTRHO Add a printrho section in the wien input file on some of the sweep structures The same consid erations described for the elastic environment apply here DOSPLOT environment The syntax of this environment corresponds to runwien DOSPLOT item_list i END DOSPLOT Add a dosplot section in the wien input file on some of the sweep structures The same consider ations described for the elastic environment apply here RXPLOT environment The syntax of this environment corresponds to runwien RXPLOT item list i END RXPLOT Add a rxplot section in the wien input file on some of the sweep structures The same considera tions described for the elastic environment apply here BANDPLOT environment The syntax of this environment corresponds to 87 runwien BANDPLOT item list i END BANDPLOT Add a bandplot section in the wien input file on some of the sweep structures The same consid erations described for the elastic environment apply here KDOS environment The syntax of this environment corresponds to runwien KDOS item list i END KDOS Add a kdos section in the wien input file on some of the sweep structures The same consideration
29. runwien ORBITALS neq_atom i global energy r global apw i orb l i orb energy r orb var i CONT STOP apw i END ORBITALS Orbital list in the same format as accepted in wien but of course free format lt neq atom gt is the number of the non equivalent atoms for which the basis is specified global energy is the default linearization energy and global apw 1 if apw method basis function is to be used 0 if lapw First lt orb 1 gt appearing in the list refers to local orbitals plane waves basis functions for that l Subsequent orb l will refer to local orbitals orb energy is the linearization energy orb var gt 0 commands wien to search for a good linearization energy and if stop is set stop in case it is not found Finally use lt apw gt 1 in case you want an apw basis function type 0 in case of lapw See wien s manual for more information about basis Default set by wien LM LIST environment The syntax of this environment corresponds to runwien LM LIST neq_atom i 11 mi 12 m2 13 m3 1k mk 1k 1 mk 1 END LM LIST Specify LM lattice harmonics expansion of charge density See wien s manual for L M pairs adequate for each structure Be careful with dimensioning parameters in WIEN2k code By default L gt 6 require redimen sioning LMAX2 in aim dstart lapw0 and lapw2 This redimensioning is required for some of the tests included in this distribution The dimensioning parameter for the maximum L value is higher
30. s and the list of non equivalent atoms given as the atomic symbol at s and the crystallographic coordinates of a representative x r y r and z r The atoms given in the body of spglist are automatically replicated and written to the struct file Runwien calls the WIEN2k s spacegroup utility to find the rest of equivalent atoms in the unit cell Note the difference in behaviour to that of EQUIV LIST only one atom for each non equivalent atom type is required to be present in the environment In the case of crystals that may be described using either hexagonal or rhombohedral cells two options are available The first and default is writing the cell parameters and atomic positions 35 in the hexagonal setup The other is to write both cell parameters and atomic coordinates using the rhombohedral cell In the latter case the keyword RHOMB must be passed to SPGLIST The appropriate conversion for working with WIEN2k is done by Runwien The struct file may end up having a H lattice hexagonal cell parameters and atomic coordinates or a R lattice hexagonal cell parameters but rhombohedral atomic coordinates with a conversion matrix defined in WIEN2k user s guide The choice depends on the symbol in the list below In some cases the convention as consigned in the International Tables for Crystallography indicates two different cell origins for the same crystal This usually happens in the space groups where the position of maximum
31. the runwien execution immediately CLEAN WIEN FULL Remove some files from the root directory and the cases subdirectories Runwien must know the number of directories to clean so this keyword must come after a LOADCHECK a GENERAL or a SWEEP section In general only those files whose name start with lt root gt or lt case gt are involved in the cleaning So if you want to keep your own files safe from the cleaning procedures just name them differently The default cleaning removes the lt case gt vector and lt case gt help files The WIEN option uses clean_lapw WIEN2K utility script to remove a large collection of files usually not required after the SCF convergence has been achieved The FULL option removes a large list of files from all the directories involved In fact the only lt root gt and case files kept will be clmsum clmup clmdn struct int in2 inm klist outputst scf dayfile incritic outputcritic ingibbs output gibbs 22 66 ps eps and pdf In other words the FULL cleaning keeps the input and output files and the intermediate files needed to run the final property calculations UNDO section Undo a section removing its check and all its information Currently only SWEEP CRITIC ELASTIC FREE and GIBBS sections can be undone LOADCHECK REREAD Loads a s
32. the user requires the calculation of some particular properties or techniques Most of the runwien orders are particular to a given section so the structure of the present guide reflects the structure of the input file Before listing sections and keywords there are some general caveats and advices you have to take into account when using runwien e Any line whose first non blank character is is interpreted as a comment and subsequently ignored Blank lines are ignored too Tab characters are equivalent to blanks e Input is case insensitive except for titles file names and similar We will use CAPITALS in this reference to indicate runwien keywords Other conventions used in describing runwien orders include optional keywords and data are enclosed in square brackets something alternative settings are separated by vertical bars op1 op2 most keywords expect data in the form of an integer or real number or a character string The type of data will be indicated by appending a suffix to the data name i for an integer r for a real and s for a string some keywords can receive as input a range of values The range can be formed as any of value min max min max step where the value min max and step can be integer range i will be used to stress this character or real values range r depending on the particular keyword on which they appear A range like 1 10 corresp
33. 1 lattice 5 IP system cubic structure loaded from nonequivalent atom list in struct atom atom 1 is Fe mult 1 x 0 00000000 atom 2 is C mult 2 x 0 08686660 atom 3 is C mult 2 x 0 90739070 atom 4 is C mult 2 x 0 85589760 atom 5 is C mult 2 x 0 00354910 atom 6 is C mult 2 x 0 14629600 atom 7 is H mult 2 x 0 16444000 atom 8 is H mult 2 x 0 82278690 atom 9 is H mult 2 x 0 72603630 atom 10 is H mult 2 x 0 00789420 atom 11 is H mult 2 x 0 27884260 Additional information Number of electrons in unit cell Valence electrons in unit cell electrons in unit cell val total e for atom val el conf for atom leaking e for atom Core Core Core Core Core val total e for atom Core val el conf for atom 3P 6 3D 6 5 48 1 5 Core leaking e for atom Core l val l total e for atom Core l val el conf for atom Core leaking e for atom Atomic energy 1start Fe Atomic energy Istart C Atomic energy lstart H Total in vacuo energy lstart Calculation fixed parameters spinpolarized no relativistic first density yes general npt 381 general rmt atom 1 2 10 struct file ferrocen struct and position of a representative y 0 00000000 z 0 00000000 y 0 12560000 z 0 20900000 y 0 12142770 z 0 20900000 y 0 94944650 z 0 20900000 y 0 84732850 z 0 20900000 y 0 95619730 z 0 20900000 y 0 23976270 z 0 20566670 y 0 23048250 z 0 20566670 y 0 90268330 z 0 20566670
34. 1 48 11 8 10 91393 10 91393 10 91393 325 0000 9 1248 1412300 0 084610000 177 0 1508880 0 2 4 4 0 2 24 24 11 48 11 8 28 11 18689 11 18689 11 18689 350 0000 9 1248 1391160 0 111400000 189 0 1479630 0 2 4 4 0 2 24 24 1 48 1 8 11 44714 11 44714 11 44714 375 0000 9 1248 1352460 0 133760000 201 0 1450710 0 2 4 4 0 2 24 24 11 48 11 8 9 83048 9 83048 9 83048 237 5000 10 1248 1166690 0 065370000 195 0 0000000 n a n a 10 16396 10 16396 10 16396 262 5000 9 1248 1324490 0 010100000 196 0 0000000 n a n a 10 47690 10 47690 10 47690 287 5000 10 1248 1390750 0 033630000 248 0 0000000 n a n a 10 77217 10 77217 10 77217 312 5000 10 1248 1413790 0 069280000 257 0 0000000 n a n a 11 05209 11 05209 11 05209 337 5000 9 1248 1404140 0 098650000 258 0 0000000 n a n a 11 31851 11 31851 11 31851 362 5000 9 1248 1373680 0 123110000 270 0 0000000 n a n a The file nacl sweep print1 eps shows that the energy curve is smooth With this extended set of data we can run the gibbs program using the runwien awk input runwien loadcheck gibbs temperature 0 278 15 end gibbs synopsis default The gibbs program is run in the subdirectory nacl gibbs It writes the output file nacl outputgibbs The energy against volume data is fitted using a numerical procedure involving polynomials of increasing order The Debye temperature is calculated from the static bulk modulus as obt
35. 1 Runwien user s and reference guide Authors Alberto Otero de la Roza and Victor Luana Contact albertoQcarbono quimica uniovi es Address Departamento de Quimica F sica y Anal tica Universidad de Oviedo Juli n Claveria 8 33007 Oviedo Spain Version 1 0 11 Date 2008 10 09 Contents 1 Runwien user s and reference guide 1 2 User s Guide 3 24 dmtroducti miu 1448429 dezi t k de i ba di 402404 ad ir a 3 22 Obtaining and installing runwie n 2 2 s s e erso aca a a a 4484 pee e RARA e Da 3 2 3 First steps Calculatioris n hep Becomes 3 2 4 Potential energy surface PES calculations e rene 8 2 5 Continuing and completing a calculation s s s soso es oaa k ee ee 10 2 6 Calculation of properties 2 02444 0884449 da 05444 ie e bii hi 14 20 ASGOMPIEX cases ferrocene saa aie g a idea a me Boba dede e deze Bez ki hE wi 16 2 8 Full WIEN2k calculation in sodium chloride 19 3 Reference guide 29 3 1 General Conventions as 2 saa sled Behe a eae BREESE Bi v WE 29 3 2 Global keywords and environments A 30 3 8 General section 2 e eze w won aa a Pele eed we hee ee kura 33 A ISP a a o de EE ee SR ae a ee ee eke PO 33 3 32 Keywords and environments ao sosoo eee 35 3 3 3 Requirements of the general section oe e 50 BA Initialization section wou o es ca waw a f n been e KOK 50 A ray Ge 50 3 4 2 Keywords and
36. 2000 5000 10000 rkmax 7 0 9 0 0 5 would ask for 75 different calculations in which five different values of rmt are combined with three values for kpts and with five values for rkmax This multidimensional search can easily give rise to a large number of calculations Runwien provides two different mechanisms to reduce this dimensional escalade The also keyword can be used to perform two or more independent explorations of parameters For instance runwien rmt 1 1 8 2 0 0 5 kpts 5000 rkmax 8 O also rmt 1 1 9 kpts 2000 5000 10000 rkmax 8 0 also rmt 11 9 kpts 5000 rkmax 7 0 9 0 0 5 will produce three independent one dimensional scans On the other hand two or more parameters can be linked to vary in a correlated way runwien link rmt rkmax rmt 1 1 8 2 0 0 5 rkmax 7 0 9 0 0 5 In this example there will be just five different calculations where both variables will change at the same time Of course the linked variables should have the same number of values 2 4 Potential energy surface PES calculations A very important feature of runwien is its ability to calculate a large amount of points in a potential energy surface with a minimal effort The section in charge of this task is called sweep and requires that the general initialization prescf and scf sections be executed first The sweep section works by creating a subdirectory for each of the required geometries and preparing a wien file on each direc
37. 2442 101 0 158959 0 2 4 4 0 2 24 24 1 48 11 8 11 2 50 2 50 7 50 4000 150 9 1248 1388410 0 12441 139 0 159010 0 2 4 4 0 2 24 24 1 48 11 8 12 2 50 2 50 7 50 6000 150 9 1248 1388490 0 12438 229 0 158995 0 2 4 4 0 2 24 24 1 48 11 8 13 2 50 2 50 7 50 8000 150 9 1248 1388460 0 12439 268 0 159012 0 2 4 4 0 2 24 24 1 48 11 8 14 2 50 2 50 7 50 10000 150 9 1248 1388450 0 12440 310 0 158970 0 2 4 4 0 2 24 24 1 48 11 8 15 2 50 2 50 7 50 2000 150 9 1248 1388390 0 12442 98 0 158959 0 2 4 4 0 2 24 24 1 481 11 8 16 2 50 2 50 8 00 2000 182 9 1248 1397580 0 11914 109 0 157354 0 2 4 4 0 2 24 24 1 48 11 8 17 2 50 2 50 8 50 2000 194 9 1248 1402350 0 11442 119 0 157093 0 2 4 4 0 2 24 24 1 48 1 8 18 2 50 2 50 9 00 2000 272 9 1248 1404990 0 11021 134 0 157530 0 2 4 4 0 2 24 24 1 481 11 8 19 2 50 2 50 9 50 2000 296 8 1248 1406700 0 10617 133 0 158164 0 2 4 4 0 2 24 24 1 48 11 8 20 2 50 2 50 10 00 2000 344 7 1248 1408050 0 10160 132 0 158840 0 2 4 4 0 2 24 24 1 48 11 8 This information can be plotted in order to select the adequate parameters The first 9 structures correspond to the 2D rmt grid mentioned above Varying the rmt of any of the two atoms affects the energy in the mRy scale The effect is similar on the flatness of the electron density with difference near 0 001 The electron density
38. 35 8 178735 8 178735 90 90 90 There are two non equivalent atoms Ti and C occupying the positions of a rock salt structure equiv list Ti 000 end equiv list equiv list C 0 5 0 5 0 5 end equiv list Explore a small change from the recommended rmt values 14 The rmt of both atoms will be linked giving rise to a 1D scan link rmti rmt2 rmt 1 3 2 1 rmt 2 3 2 1 gmax 15 0 r0 0 00005 kpts 1000 end general initialization xcpotential ggapbe96 energymax 2 0 end initialization prescf default scf Use charge convergence criterion charge conv 0 0001 end scf do 1 Run the following sections except sweep only for structure number 1 i e for the one with 3 rmt The do and except keywords provide a rich mechanism for this kind of selection printrho Plot the electron valence density in a plane rho valrho Other options total deformation spin and atomic densities coulomb exchange correlation and total potential Energy cutoff for the bands adding to the valence density energymin 1 0 Plane specification corners and number of grid points origin 1 1 0 4 xend 1 3 0 4 yend 3 1 0 4 npt 100 100 Number of unit cells added up on each direction nshells 3 23 Cap the electron density fixing a min and max value in e bohr73 zmin 0 075 zmax 0 3 end printrho dosplot Density of States plot total and projected Use eV Rydberg is the default and fix plot limits plotunit
39. 80 10 32280 10 32280 275 0000 9 1248 1370460 0 013030000 154 0 1575500 0 24 4 0 12 24 24 11 48 11 8 10 62659 10 62659 10 62659 300 0000 10 1248 1407190 X 0 052370000 0 0 0000000 n a n a 10 91393 10 91393 10 91393 325 0000 not done 11 18689 11 18689 11 18689 350 0000 not done 11 44714 11 44714 11 44714 375 0000 not done 26 that indicates that the fourth sweep point is in the middle of the calculation and that the last three points were not calculated at all To continue the interrupted job use an input file runwien loadcheck sweep reference struct 1 do 4 7 dosplot end dosplot bandplot end bandplot kdos kdos 1 end kdos critic iws 2 newton le 15 end critic end sweep synopsis default and run it Note that we have removed the with line and included the do 4 7 to continue the calculation Runwien awk will then rewrite the wien files in the sweep directory tree with exactly the same data and calculate the missing points After running the input file the result is tasciit Variable parameters number of structures 7 a bohr b bohr c bohr v bohr73 it warning energy ry efermi ry time s planarity morsesum topology 9 65489 9 65489 9 65489 225 0000 9 1248 1042110 X 0 098260000 128 0 4215360 24 2 4 4 0 11124111 48 11 8 10 00000 10 00000 10 00000 250 0000 9 1248 1258250 0 036130000 141 0 1632730 0 2 4 4 0 12 24 24 11 48 1 8 10 32280 10 322
40. 80 10 32280 275 0000 9 1248 1370460 0 013030000 171 0 1575500 0 2 4 4 0 2 24 24 1 48 11 8 10 62659 10 62659 10 62659 300 0000 10 1248 1407190 X 0 052370000 191 0 1542220 0 2 4 4 0 2 24 24 1 48 11 8 10 91393 10 91393 10 91393 325 0000 9 1248 1412300 X 0 084610000 178 0 1508880 0 2 4 4 0 2 24 24 1 48 11 8 11 18689 11 18689 11 18689 350 0000 9 1248 1391160 X 0 111400000 190 0 1479630 0 2 4 4 0 2 24 24 1 48 11 8 11 44714 11 44714 11 44714 375 0000 9 1248 1352460 X 0 133760000 201 0 1450710 0 2 4 4 0 2 24 24 1 48 11 8 The DOS spaghetti and KDOS plots are generated in ascii nacl sweep nacl sweep nacl sweep 1 ps eps The data of energy can be readily graphed manually and also using the runwien awk input file runwien loadcheck sweep 27 print energy vs v end sweep synopsis default that generates the nacl sweep print1 eps file in the root directory of the calculation Finally we want to obtain a estimation of the thermodynamic properties of the NaCl system by using the quasiharmonic Debye model implemented in gibbs However we feel that 7 points is not enough data for this task so we would like to refine the potential energy surface by adding new points intertwined with the previous grid Additionally it turns out that we have at our disposal a new batch calculation system so we do not want runwien awk to calculate the new points serially To this end we write a ne
41. 98 100104108 42m 0020 324044 526064 72 76808488 92 96 100 104108 Am2 0020 324044 526064 72 76808488 92 96 100 104108 4 mmm 002040446064808488100104108 53 3 0010203033 334043 435053 536063 6 366 66 7073 7376 768083 8386 869093 9396 9699 9 9100103 103106 106109 109 3 002040 434360 6363 666680 8383 8686100 10 3103 106106 109109 312 00102030 33404350 5360636670 73 768083 8690 93 96 99100103106 109 321 00102030 33404350 5360636670 73 768083 8690 93 96 99100103 106 109 31m 001020303340435053606366707376808386 90939699100103106109 3m1 001020303340435053606366707376808386 90939699100103106109 31m 00204043606366808386100103106109 3m1 00204043606366808386100103106109 6 00102030405060 666670 767680 868690 96 96100 106106 6 002033 334053 53606673 7380869093 9399 99100106 6 m 00204060 666680 8686 100 106 106 622 001020304050606670 76808690 96 100 106 6mm 0010203040506066707680869096100 106 6m2 002033405360667380869399 100 106 62m 002033405360667380869399 100 106 6 mmm 00204060668088100108 23 0040446064 326266 72 7680848 8 9 2 96 94 9 8100104108 10 2 106 10 10 m 3 00404460646266808488 100104 108 10 2106 10 10 432 0040446064808488 94 98100104108 43m 0040446064 32 72 76808488 92 96100104108 m 3m 0040446064808488100104108 FERMI ROOT TEMP TEMPS GAUSS TETRA ALL value r Method to find fermi energy See WIEN2k manual for detai
42. AL LSDA GGAPBE96 GGAPW91 GGAWC06 Chooses one of the available exchange correlation potentials Default reference structure ECOREVAL ecore r Initial core valence separation energy in Ry Default reference structure ENERGYMIN emin r Lower energy cutoff for eigenvalue search Default reference structure ENERGYMAX emax r Higher energy cutoff for eigenvalue search Default reference structure ORBITALS environment The syntax of this environment corresponds to runwien ORBITALS neq_atom i global_energy r global_apw i orb l i orb energy r orb var i CONT STOP apw i END ORBITALS Orbital list in the same format as accepted in wien but of course free format neq atomp is the number of the non equivalent atoms for which the basis is specified global energy is the default linearization energy and global apw 1 if apw method basis function is to be used 0 if lapw First lt orb_1 gt appearing in the list refers to local orbitals plane waves basis functions of that I Subsequent lt orb_1 gt will refer to local orbitals orb energy is the linearization energy orb var gt gt 0 commands wien to search for a good linearization energy and if stop is set stop in case it is not found Finally use lt apw gt 1 in case you want an APW basis function type 0 for a LAPW function See the WIEN2k manual for more information about bases Default reference structure LM LIST environment The syntax of this envir
43. In our example we have decided to increase the maximum iteration cycles to 50 instead of the default 30 The convergence in the energy up to 10 microRydberg is actually the default The remaining sections in the input are all optional Printrho Tasks related to the electron density and several interaction potentials Dosplot Calculation and plot of the Density of States DOS and similar tasks Rxplot Print the calculated x ray emission or absorption spectra Bandplot Plot the band structure Kdos Combined plot of the band structure and the density of states Aim Perform an Atoms in Molecules i e Bader analysis of the crystal electron density using Sofo et al code included in WIEN2k Critic Perform an Atoms in Molecules i e Bader analysis of the crystal electron density using Mart n Pend s et al critic code POL2008 Sweep Produces a Potential Energy Surface by means of the variation of one or more geometrical parameter cell lengths and angles internal atomic positions or some combinations of these Gibbs Use the energy versus volume data to determine the Equation of State EOS and produce an estimation of thermal effects using the quasiharmonic Debye model imple mented in Blanco et al gibbs code BFL2004 Elastic Drive the calculation of the elastic constants of the crystal Free Used to determine the energy of free atoms under the same calculation conditions used for a crystal Synopsis Flexible tool for
44. N LOG ATAN BADER nptsu i nptsv i niso i END GRDVEC DOPLOT DENSITY RHO GRADMOD LAPLACTAN LAP GRADIENT FORMAT DEFAULT GNUPLOT D2D D3D CARTESIAN 3DCUBE CUBE FILE rootfile s 1D x0 r yO r zO r xi r yi r zi r nu i 2D xO r yO r zO r xi r yi r zi r x2 r y2 r z2 r nu i nv i 8DC xO r yO r zO r xi r yi r zi r nu i nv i nw i USE RHO VALRHO ATOMIC NOSUMMARY DO range i ALL NONE NEW OLD EXCEPT range i END CRITIC Locate critical points in electron density by using the critic program Optionally calculate basin files adequate as tessel input for plotting atomic basins Creates a final summary in criticsummary out in root directory The critic output file contains much more information than the topology and planarity extracted by runwien to the general summary This output file with extension outputcritic is saved to the general structure subdirectory and can be read processed printed by the user The input file incritic is also saved and can be edited to continue the topological analysis beyond the routine tasks runwien critic default Same as above but assuming defaults on all variables 78 3 16 2 Keywords and environments LINE ATOM NNEQ mult i x r y r z r ATOM NNEQ mult ii x r y r z r npts i Print the elec tron density and other properties on a regular grid of lt npts gt points in the indicated segment The end points of the segment can be atoms or general pos
45. OM a r b r c r alpha r beta r gamma r Unit cell parameters in bohr and sexagesimal degrees If the ANGSTROM keyword appears after CELL PARAMETERS a r b r and c r are to be in angstrom Compulsory parameter TITLE runtitle s Sets a title for the calculation This will be used on several places including titles within gnuplot scripts Including single or double quotes will interfere severely To be safe you should stick to the characters 0 9a zA Z_ Default runwien lapw calculation v lt runwien version gt SPINPOLARIZED YES NO Spin polarized calculation Express the system density as a sum of spin up and spin down contributions and calculate each of them separately Default no RELATIVISTIC YES NO Relativistic calculation of the core electron density Default yes NPT atom i range i AUTO Number of radial mesh points It must be an odd number The WIEN2k manual recommends 381 for LDA calculations 781 for GGA The use of a first last or first last step range will start a scan of NPT values The AUTO keyword selects the default value for NPT A NPT order affects to a single atom type by default to the first atom in the input indicated by the integer atom i The number atom i is a non equivalent atom index pointing to the cor responding EQUIV LIST The non equivalent atom index are assigned in order of appearance of the EQUIV LIST environment 1 for the first atom 2 for the second atom etc If
46. P p4 n p4a P p42 n p4bc P i4 m 14 B i41 a idad B p422 p4 2 P p4212 p4ab 2ab P p4122 p41 2c P p41212 p43n 2nw P p4222 p4c 2 P p42212 p4n 2n P p4322 p43 2c P p43212 p41n 2abw P 1422 14 2 B 14122 i41b 2bw B p4mm p4 2 P p4bm p4 2ab P p42cm p4c 2c P 43 p42nm p4n 2n P p4cc p4 2c P p4nc p4 2n P p42mc p4c 2 P p42bc p4c 2ab P 14mm 14 2 B i4cm 14 2ab B i41md i41b 2 B i4lcd 141b 2c B p 42m p 4 2 P p 42c p 4 2c P p 421m p 4 2ab P p 421c p 4 2n P p 4m2 p 4 2 P p 4c2 p 4 2c P p 4b2 p 4 2ab P p 4n2 p 4 2n P i 4m2 1 4 2 B i 4c2 1 4 2c B i 42m 1 4 2 B i 42d i 4 2bw B p4 mmm p4 2 P p4 mcc p4 2c P p4 nbm p4a 2b P p4 nnc p4a 2bc P p4 mbm p4 2ab P p4 mnc p4 2n P p4 nmm p4a 2a P p4 ncc p4a 2ac P p42 mmc p4c 2 P p42 mcm p4c 2c P p42 nbc p4ac 2b P p42 nnm p4ac 2bc P p42 mbc p4c 2ab P p42 mnm p4n 2n P p42 nmc p4ac 2a P p42 ncm pdac 2ac P i4 mmm i4 2 B i4 mcm 14 2c B 141 amd i4bd 2 B i41 acd i4bd 2c B p3 p3 H p31 p31 H 44 p32 p32 H r3 r3 R p 3 p3 H r 3 13 R p312 p3 2 H p321 p3 2 H p3112 p31 2 0 0 1 3 H p3121 p31 2 H p3212 p32 2 0 0 1 6 H p3221 p32 2 H r32 r3 2 R p3
47. PLOTXMIN emin r Minimum energy for x ray spectra graph in eV Default 2 0 PLOTXMAX emax r Maximum energy for x ray spectra graph in eV Default 15 0 DE de r Energy increment for x ray spectra graph in eV Default 0 02 TYPE EMIS ABS Absorption or emission x ray spectra Default abs INIMAXENERGY emax r Maximum energy for the eigenvalue window calculation in nl in Ry Default 10 0 Ry DO range i ALL NONE NEW OLD Used together with the EXCEPT order below EXCEPT range i As a default rxplot will be run only on the new points skipping those points already examined The DO EXCEPT orders can be used to modify this behavior Default Use the local DO EXCEPT specification if available Use the global DO EXCEPT data otherwise In absence of user orders do the rxplot task only on new points 76 3 14 3 Requirements of the rxplot section General initialization prescf and scf sections are required 3 15 Aim section 3 15 1 Syntax runwien AIM DEFAULT ATOM nneq i USES TWO THREE ALL FOUR NSHELIS nx i ny i nz i DO range i ALL NONE NEW OLD EXCEPT range i END AIM Locate critical points in electron density by using the WIEN2k program aim For now calculating the basin properties is not supported nor does the analysis of the output file runwien aim default Same as above but assuming defaults on all variables 3 15 2 Keywords and environments ATOM nneq i Atom from where the search is
48. ST ANY PATH path s Use the converged density of a previously calculated structure as starting point replacing the superposition density generated by dstart There are 5 modes of REUSE In CHAIN each general section structure tries to reuse the con verged density of the preceding structure In FIXED one given structure number n i is selected and consistently used as starting point FIRST is equivalent to FIXED 1 i e use always the density of the first structure If ANY is used runwien tries to find a directory among the cal culated structures that contains a valid clmsum and struct files Finally a path may be given directly with the keyword PATH pointing to the location of a calculation With independence of the method employed in determining the path containing the initial density runwien has three levels of modification of the clmsum file If the struct file of the source calcu lation differs in any structural parameter cell parameters or atomic positions from the actual calculation a extrapolation is carried out using the clmaddsub utility only available in recent versions of WIEN2k This process involves subtracting the old superposition density stored in the clmsum atomic file and adding the new one Second if the structural parameters are the same but the struct files of both calculations differ in npt rmt or r0 for any atom a interpolation of the radial grid is performed using the WIEN2k s clminter program Finall
49. The sweep calculation inherits as a default the cell parameters entered in the general section but these can be overridden manually The results may be printed to a xy or xyz file and plotted out using a gnuplot script The sweep section thus is runwien s power tool for potential energy surface exploration This explo ration may be done for any arbitrary crystal system Notice however that the reference crystal system and symmetry entered in the general section can not be modified in the sweep section For example if the reference structure is cubic the cell can t be deformed into anything else than a cube This can be tricked however by using a tiny deformation on the reference structure For instance use something like 90 0001 on the three cell angles if you want to explore a rhombohedral deformation All the properties seen in previous sections printrho dosplot can be calculated for whatever sweep structures you like These properties are calculated in a natural way for the internal working of the sweep structure involves one wien input file for each structure calculated Therefore all the property section lines are passed directly to the indicated structures wien file and runwien executed on them All the sweep calculations are contained in a directory that is called lt root gt sweep where lt root gt is the root name for the calculation This directory contains one subdirectory for each structure and an i
50. X Imax i Maximum value for spherical harmonics expansion inside muffin tins Required in initialization same as above Default reference structure LNSMAX Insmax i Maximum value for spherical harmonics expansion used to computate matrix elements with non muffin tin functions Required in initialization same as above Default reference structure GMAX gmax r Maximum reciprocal vector for fourier expansion of electron density Wien2k manual recommends using a large gmax 25 for systems with short H bonds Also a larger gmax for gga calculations 14 than for lsda potentials Required in initialization same as above Default reference structure MIX lambda r Mixing factor used in the creation of the new charge density Default reference structure XCPOTENTIAL LSDA GGAPBE96 GGAPW91 GGAWC06 Chooses one of the available exchange correlation potentials Default reference structure ECOREVAL ecore e Initial core valence separation energy in Ry Default reference structure ENERGYMIN emin r Lower energy cutoff for eigenvalue search Default reference structure ENERGYMAX emax r Higher energy cutoff for eigenvalue search Default reference structure ORBITALS environment The syntax of this environment corresponds to runwien ORBITALS neq_atom i global_energy r global_apw i orb l i orb energy r orb var i CONT STOP apw i END ORBITALS Orbital list in the same format as accepted in wien but of course free
51. a label replaces the atom i integer the NPT order affects all the atoms in the crystal An order like NPT 300 500 100 will confound the parser into assuming that 300 is the atom number rather than the first value of a range If your order omits the number of the atom be careful to avoid white characters within the declared range Default 781 for every atom If no atom index is indicated 1 is assumed RMT atom i range r AUTO Muffin tin radius in bohr for the indicated non equivalent atom The AUTO keyword lets WIEN2k choose the RMT according to the geometry of the system Other possibilities are e a range first last or first last step asks runwien to perform a set of calculations with a variable RMT values like x correspond to multiply the internal RMT value by a factor according to the formula rmt auto value 1 x 100 If the symbol occurs anywhere within a range each RMT value will be interpreted as a multiplying factor This is the case of for instance 3 4 5 which is equivalent to 3 4 5 e a list of values or ranges separated by commas is also admisible A RMT order affects to a single atom type by default to the first atom in the input An order like RMT 3 5 0 5 will confound the parser into assuming that 3 is the atom number rather than the first value of a range If your order omits the number of the atom be careful to avoid white characters within the declared range If a
52. ailable or the atomic energies estimated from the INITIALIZATION section Default free in vacuo energy If free was not run use the atomic energies estimated on the initialization section 3 18 3 Requirements of the gibbs section General initialization prescf scf and sweep sections are required for gibbs to run 3 19 Synopsis section 3 19 1 Syntax runwien SYNOPSIS OUTPUT file s EXHAUSTIVE list_item i END SYNOPSIS Synopsis creates a whole runwien run summary and writes it to a specified file This summary includes general information calculation times important input parameters and key output values all pretty printed As it may use information from every section synopsis is best run at the end but this is by no means necessary Most significant quantities are already included in the output However the possibilities for this task are huge We may add optional information using switches to trigger it in future versions As it is common this section can be run using default values for all variables runwien synopsis default 3 19 2 Keywords and environments OUTPUT file s Specify output filename Default synopsis out EXHAUSTIVE list item i Print exhaustive information for the general section structures indicated in the list item This list follows the same syntax as with in sweep structure and rmt in general section Exhaustive information includes not so important values like convergence
53. ained from the fit The thermodynamic properties of the crystal at zero and room temperature are plotted with the above input rendering excerpt ascii RESULTS AT P 0 FOR ALL TEMPERATURES 0 00 322 39 602 20 4 83 0 00000 0 00000 278 15 335 62 612 96 14 36 76 39760 48 19218 Each point calculated in sweep is stored under the corresponding subdirectory in nacl sweep The index file nacl sweep nacl sweep index contains the information relating the unit cell geometry with the numbering These subdirectory contain also runwien awk jobs Thus a detailed analysis of one of these points is always possible using runwien awk 3 Reference guide 3 1 General conventions WIEN2k calculations take place in a directory that is referred to as root Most of the many files involved in the calculation are named as lt root gt lt something gt Following this convention the runwien input file must be called lt root gt wien 29 A runwien run can give rise to many WIEN2k calculations either because the user wants to explore the effect of modifying some model parameter or because a scan of the geometry has been called for Every one of those calculations is done in a particular subdirectory that is automatically created by the runwien script We will refer to the calculations as cases and to their subdirectories as lt case gt The runwien input is divided into several sections some of them mandatory and some used only when
54. alculation so as to check everything is correctly setup 91 Together with the input wien files our distribution includes the most significant output files syn opsis out runwien out the log data that runwien prints to stdout and some postscript images Keep our original outputs and compare with the results obtained in your machine to check that your installation works correctly Some tests like dosplot tests portray some of the new capabilities of runwien in the determination of electronic properties You can use them as templates for your calculations if you desire as they are fairly complete The complete list of tests follows The estimated times refer to a desktop PC dual core 3 40 GHz 1 GB RAM running debian etch kernel version 2 6 23 14 test001 TiC this is the example in the WIEN2k user s guide It is a single point calculation of TiC plot of valence density DOS x ray spectrum spaghetti plot QTAIM analysis and 5 points in the potential energy surface Estimated time 7m 23s test002 Ni another example from the WIEN2K user s guide Fcc nickel is calculated mostly with default options Estimated time 1m 50s test003 Ti02 WIEN2K user s guide example Minimal input mostly with default options Estimated time 3m 36s test004 Be bcc an example of the usage of the elastic section in a cubic crystal The tetragonal deformation is very flat so there is a good amount of noise in the corresponding energy p
55. an alternatively use the input runwien loadcheck sweep reference struct 1 25 with v 900 1500 100 dosplot end dosplot bandplot end bandplot kdos kdos 1 end kdos critic iws 2 newton ie 15 end critic end sweep synopsis default which does exactly the same thing as the former The wien file is processed with the command tasciit runwien awk nacl wien Let us assume that for whatever reason the calculation is killed oh nol in the middle of the sweep run Specificallv the fourth point is affected ascii info sweep Calculating sweep structure 3 info sweep Extracting information from structure 3 info sweep Marking as done info sweep Writing checkpoints info sweep Calculating sweep structure 4 Killed To check the actual point at which the calculation was killed you can examine the stdout as above the sweep script temporary file created by runwien awk or simply execute runwien awk on the input runwien loadcheck reread synopsis default This produces an output tasciit Variable parameters number of structures 7 a bohr b bohr c bohr v bohr73 it warning energy ry efermi ry time s planarity morsesum topology 9 65489 9 65489 9 65489 225 0000 9 1248 1042110 X 0 098260000 129 0 4215360 24 2 4 4 0 1 24 1 48 1 8 10 00000 10 00000 10 00000 250 0000 9 1248 1258250 X 0 036130000 138 0 1632730 0 2 4 4 0 12 24 24 1 48 11 8 10 322
56. analysis of the scf process real rkmax used space filling by muffin tin spheres etc 3 19 3 Requirements of the synopsis section No other section is required for synopsis section In fact synopsis will detect if other sections have been run and print a summary according to it 90 4 Directory tree for runwien File locations All work runwien performs is constrained to root directory which is where wien file lives The root directory has the same name as the wien file except for the wien extension This is a requirement for runwien to run The directory tree for a full exhaustive calculation is as follows ascii lt root gt lt root gt wien main input file lt root gt index general section structure indexfile lt root gt sweep print sweep section print files lt root gt sweep print eps sweep section plots out summaries critic scf sweep synopsis lt root gt lt n gt general section s structure lt root gt lt n gt WIEN2k s calculation files gnuplot scripts ps files output files lt root gt free lt at gt free section directory for atom lt at gt lt root gt free wien free section main input file l lt root gt freel free section calculation files lt root gt elastic elastic section directory __ lt root gt xxxxxx deformation xxxxxx subdirectory lt root gt lt n gt point number n in def
57. aved calculation from the check files If the REREAD keyword is present runwien does not rely on the check information and tries to determine as much as possible from all the existing files This way externally calculated sweep or general structures can be loaded REREAD does rely on a set of minimal variables general and sweep iterations number of non equivalent atoms and the like which should never vary Some unimportant variables like the printing level in kgen or the ones related to printrho dosplot etc are plainly ignored Be however careful when rereading a structure which has only gone through the general section its struct file is usually badly set the struct file is heavily modified in the initialization section When loading from a previous calculation you must be aware that non variable parameters can not be changed That is changing from energy conv 0 0001 to energy conv 0 00001 is not allowed and in fact will stop the calculation if tried Also specifying the same value as the one it is stored in check is redundant but valid in general There is however one exception to this environments such as equiv list etc can not be specified when loading checkpoints SYSTEM order s Execute an external OS order This is prone to produce any malware action in a wien file inherited from an untrusted source Due to this potential security risk this command is disabled by default Should you want to use this
58. bset of calculations to run when more that one calculation is implied from the input data The behavior of this orders is described in the GLOBAL section Default Use the local DO EXCEPT pair if present otherwise apply the global DO EXCEPT pair if present if neither is defined calculate only new structures 3 7 3 Requirements of the spinorbit section General initialization prescf and scf sections are required 3 8 Free section 3 8 1 Syntax runwien FREE ATOM atname s GLOBAL REFERENCE STRUCT n i CELL a r SPINPOLARIZED YES NO NPT r r RMT r r RO r r RKMAX rkmax r LMAX lmax i LNSMAX Insmax i GMAX gmax r MIX lambda r XCPOTENTIAL LSDA GGAPBE96 GGAPW91 GGAWCO6 ECOREVAL ecore e ENERGYMIN emin r ENERGYMAX emax r ORBITALS neq_atom i global_energy r global_apw r orb l i orb energy r orb_var i CONT STOP apw i END ORBITALS LM LIST neq_atom i 11 mi 12 m2 13 m3 1k mk 1k 1 mk 1 END LM LIST FERMI ROOT TEMP TEMPS GAUSS TETRA ALL value r MAX ITERATIONS maxit i CHARGE CONV cconv r ENERGY CONV econv r FORCE CONV fconv r ITDIAG cyc il NEW INI cyc i CLEAN NOSEND 60 NICE nice i DO ALL NONE NEW OLD atnamei s atname2 s EXCEPT atname1 s atname2 s END FREE The calculation of the free in vacuo energy of an atom using WIEN2k is a delicate matter Following the recommendations of the WIEN2k mailing list this section tries to obtain a value for the in vacuo energy
59. ccept runwien s default value Npt value will not change from one structure to the other but generally speaking auto keyword should be avoided in sweep section as all the points in the potential energy surface must have the same calculation level All this comments are valid for general section s varying parameters in sweep section rmt r0 82 Default reference structure RMT atom i rad r Muffin tin radius in bohr for non equivalent atom index atom i Default reference structure if no atom is indicated 1 is assumed RO atom i rad r First radial mesh point in bohr for non equivalent atom index atom i WIEN2k manual recommends a value between 0 0005 and 0 00005 Default reference structure RKMAX rkmax r rmt Kmax where Kmax is the maximum k for plane waves and rmt is the smallest muffin tin radius Default reference structure LMAX Imax i Maximum value for spherical harmonics expansion inside muffin tins Required in initialization same as above Default reference structure LNSMAX Insmax i Maximum value for spherical harmonics expansion used to computate matrix elements with non muffin tin functions Required in initialization same as above Default reference structure GMAX gmax r Maximum reciprocal vector for fourier expansion of electron density Wien2k manual recommends using a large gmax 25 for systems with short H bonds Also a larger gmax for gga calculations 14 than for Isda pote
60. cf sections 3 17 Sweep section 3 17 1 Syntax runwien SWEEP REFERENCE STRUCT n i NPT atom i pt i RMT atom i rad r RO atom i rad r RKMAX rkmax r LMAX lmax i LNSMAX Insmax i GMAX gmax r MIX mixfactor r KPTS kpts i LDA U SIC AMF HMF 80 iati i li i li s u r j r iat2 i 12 i 12 s u r j r END LDA U WITH AIB C ALPHA BETA GAMMA V C AIC BIB A range r ALSO PRINT zvar s VS yvar s VS xvar s XCPOTENTIAL LSDA GGAPBE96 GGAPW91 GGAWCO6 ECOREVAL ecore r ENERGYMIN emin r ENERGYMAX emax r ORBITALS neq_atom i global_energy r global_apw i orb_1 i orb energy r orb var r CONT STOP apw i END ORBITALS LM LIST neq atom i 11 mi 12 m2 13 m3 1k mk 1k 1 mk 1 END LM LIST FERMI ROOT TEMP TEMPS GAUSS TETRA ALL value r MAX ITERATIONS miter i CHARGE CONV cconv r ENERGY CONV econv r FORCE CONV fconv r ITDIAG cyc il NEW INI cyc i NICE nice i REUSE CHAIN REFERENCE PATH DETECT NOREUSE NOSUMMARY SPINORBIT item_list i END SPINORBIT ELASTIC item_list i END ELASTIC PRINTRHO item_list i END PRINTRHO DOSPLOT item list i END DOSPLOT RXPLOT item list i END RXPLOT BANDPLOT item list i END BANDPLOT AIM item list i END AIM CRITIC item list i 81 END CRITIC MINI commandline s CLEAN WIEN FULL DO range i ALL NONE NEW OLD EXCEPT range i END SWEEP Examine the properties of the crystal as one or more cell parameters change their value independently or not
61. critic data available The basis set information is contained in the nacl1 nacll in1 file and reads tasciit WFFIL WFPRI SUPWF 7 5 10 4 R MTxK MAX MAX L IN WF V NMT 43099 4 0 GLOBAL E PARAMETER WITH n OTHER CHOICES global APW LAPW 0 0 23100 0 00000CONT 1 0 3 71800 0 00200CONT 1 1 0 19700 0 00000GO0NT 1 1 1 65900 0 00000GO0NT 1 43099 3 0 GLOBAL E PARAMETER WITH n OTHER CHOICES global APW LAPW 0 0 21700 0 00000GO0NT 1 0 1 03900 0 00000GO0NT 1 1 0 18700 0 00000GO0NT 1 K VECTORS FROM UNIT 4 9 0 2 0 emin emax window which translates almost without modification to a new wien input file nacl wien runwienti general loadcif nacl cif rmt 1 2 0 2 25 rmt 2 2 0 2 25 kpts 2000 rkmax 7 5 2 5 2 5 also kpts 2000 10000 2000 rkmax 7 5 also kpts 2000 rkmax 7 5 10 0 0 5 end general initialization lm list 1 0040446064808488100104 108 21 end lm list lm list 2 0040446064808 4 8 8 10 0 10 4 10 8 end lm list orbitals 1 43099 0 0 0 231 0 000 CONT 0 3 718 0 002 STOP 1 O 197 0 000 CONT 1 1 659 0 000 CONT end orbitals orbitals 2 43099 0 0 O 217 0 000 CONT 1 O 1 039 0 000 CONT 1 1 0 187 0 000 CONT 1 end orbitals m n pa pa end initialization prescf default scf default critic iws 2 newton 1d 15 end critic synopsis default Before proceeding with the calculation we will try to determine the best values for some other WIEN2k parameters Concretely rmt the number of k poin
62. d densities from one deformation point to the other in a elastic calculation The CHAIN keyword makes each deformation use the converged density of the previ ous one The first structure of each deformation type is calculated starting from the superposition of atomic densities With REFERENCE each structure extrapolates the charge density of the selected reference structure of the general section In a way similar to SCF section PATH allows the user to use consistently a given calculation as starting point from every strained structure Finally if DETECT is read runwien searches for existing clmsum and struct files in the directories of the elastic section and uses the density files if found Default DETECT NOREUSE Reusing the converged densities can lead to large time savings However problems can arise specially in the deformations involving lowering of symmetry The NOREUSE keyword allows the user to avoid the reutilization of densities of previous points completely Default MINI commandline s Perform an internal parameter optimization using the mini program through the min_lapw script The mini line is passed down to all the wien input files in the elastic calculation This is a temporary solution to the problem of including the inner strain in the elastic constant calculation Default no internal optimization SPINORBIT environment The syntax of this environment corresponds to runwien SPINORBIT END SPINORBIT
63. d of runwien awk This option however is a temporary solution and a further development of this feature in runwien awk is planned Should this keyword fail to calculate the inner strain you can fall back to running mini manually and rereading the results in runwien awk 3 9 2 Keywords and environments REFERENCE STRUCT n i General section s reference structure for calculations This structure pro vides default values in case they are not user provided Default the first which was run through general section DEFORMATION POINTS n i MAXLENGTH mlenght r MAXANGLE mangle r Selects the num ber of points calculated in each deformation the maximum length spanned in a b and c cell parameter s deformations and the max angle spanned in alpha beta gamma cell parameter s deformations There is nothing wrong about an even number of points but it seems more sensible to choose and odd number for you are interested in the second derivative at zero strain The mlength parameter is given as a fraction of the original cell length For example if mlength 0 05 the deformation in the a parameter will comprise the interval a 1 mlength a 1 mlength The mangle value is the maximum distance the deformation reaches from the original angle cell parameter in sexagesimal degrees For example if mangle 5 the deformed alpha interval for a cubic system will be alpha 5 alpha 5 If an odd number of points is used the central point will be sli
64. dd your own gnuplot scripts into runwien code To do this you must add the name of your gnuplot script to the array const gnuplot where the dimensions of the array are indicated by a string containing the type of quantity the script is intended for For example ascii const_gnuplot v c a energy lapw plot e ca v gnuplot The gnuplot script must be available through your PATH and readily executable it will be called gnuplot script not gnuplot gnuplot script Default no prints XCPOTENTIAL LSDA GGAPBE96 GGAPW91 GGAWC06 Chooses one of the available exchange correlation potentials Default reference structure ECOREVAL ecore r Initial core valence separation energy in Ry Default reference structure ENERGYMIN emin r Lower energy cutoff for eigenvalue search Default reference structure ENERGYMAX emax r Higher energy cutoff for eigenvalue search Default reference structure ORBITALS environment The syntax of this environment corresponds to runwien ORBITALS neq_atom i global_energy r global_apw i orb l i orb energy r orb var i CONT STOP apw i END ORBITALS Orbital list in the same format as accepted in wien but of course free format lt neq atom gt is the number of the non equivalent atoms for which the basis is specified global energy is the default linearization energy and global apw 1 if apw method basis function is to be used 0 if lapw First lt orb 1 gt appearing in the list refers to
65. e WIEN2k package Cif files should be used with caution particularly those files coming from a external database In particular e Atomic coordinates are usually stored with low precision This has bad consequences for symmetry determined values like 1 3 and 2 3 A coordinate of 0 3333 instead of 1 3 can produce a spurious atom when the nonequivalent position is replicated by symmetry The best action in this case is hand editing this file to change these coordinates to something like 0 333333333 Note that the RMT NPT and RO values generated by cif2struct are not meaningful so they are completely ignored These parameters need to be set in the general section specification Default LOADSTRUCT filestruct s Load the information contained in the indicated struct file The behavior of this order is analogous to that of LOADCIF that is values explicitly set are maintained and the structural information is completed using the struct file In this case the number of parameters to be read is larger lattice type list of non equivalent atoms positions and mutiplicities cell parameters NPT RO and RMT for each atom Contrarily to LOADCIF the RMT NPT and RO contained in the struct file are read These values act as default if the corresponding keyword does not appear in the general section If you want to recover the WIEN2k value for any of this use the AUTO value The strategy of loading all the information from a struct file and then
66. e create a directory for the calculation ferrocene and copy the original struct file into ferrocene orig ferrocen struct The ferrocene wien file contains runwien general loadstruct orig ferrocen struct Load structural info from existing struct kpts 1 rkmax 3 0 lmax 12 lnsmax 5 gmax 23 0 end general initialization xcpotential ggapbe96 end initialization prescf default scf max iterations 100 energy conv 0 00001 itdiag 6 Require iterative diagonalization with a full diagonalization each 6 scf cycles This value has no effect in newer WIEN2k versions end scf synopsis default For the sake of computational simplicity only one k point will be used The synopsis out generated by runwien is tasciit General information title Ferrocen Molecule root temp pwd home alberto temp machine xxxx date mie oct 24 02 22 37 CEST 2007 Sections run and time information general section run time s 0 initialization section run time s 4 prescf section run time s 124 scf section run time s 3043 total run time s 3171 Structural information Mol formula Fe1C2C2C2C2C2H2H2H2H2H2 Mol mass 186 031400 a bohr 15 00000000 b bohr 15 00000000 17 c bohr 15 00000000 alpha deg 90 00000000 beta deg 90 00000000 gamma deg 90 00000000 c a 1 00000000 c b 1 00000000 b a 1 00000000 volume bohr73 3375 00000000 Mol volume 3375 00000000 space group P
67. e this keyword to separate two independent sets of structures For instance runwien rkmax 3 0 kpts 1000 5000 2000 also rkmax 3 2 kpts 1000 5000 2000 will explore the effect of the KPTS parameter for two different values of RMT Every one of the explorations separated by ALSO will be done independently Default LINK varl s var2 s var3 s Link the indicated variables so that they vary simultaneously The linking must be done between variables that receive the same number of values or a mismatch error will be found The variable names are the same as the keywords described before RKMAX GMAX etc In the case of RMT RMTI corresponds to the first atom RMT2 to the second etc The same syntax is used in NPT NPTI NPT2 and RO R01 RO2 Note that 48 runwien awk input is case insensitive The linking is global and affects every ALSO block Several LINK keywords may be issued in order to set more than one ligature among variables However the sets of variables linked must be disjoint Default LOADCIF filecif s Use the structural data from the indicated cif file to complete the system geometry Explicitly set data will not be overwriten For instance runwien loadcif graphite cif cell 4 8 4 8 6 5 90 90 120 will keep the cell parameters from the explicit CELL order but read the rest of the structural data from the graphite cif file The LOADCIF task depends on the cif2struct code distributed with th
68. ee Eh L 70 Ok GWADA Kan san oh aha Ste te GO Ae tee AR ane ee ee ee r e 70 3 102 Keywords and environments s s sxs p 424424 24 69 eee KR 70 3 10 3 Requirements of the printrho section 00 0 a eee 71 3 11 Dosplat SE tiG i i 4045 42 2 ee 4a A ea L a 72 Sl Syntaks za ze hea ee ini kata eee eee PEELE ARE eee ee mat 72 IL Keywords and environments s osos sc co 44 den b OWAK 22 Ree HS 72 3 11 3 Requirements of the dosplot section LL Luomo 000020000 73 3 12 Bandplot Section as a dve ee ik a da e W EO A i eee niu aron l 74 Sol Qe ONA AE thee Bo a ee ee ee ee 74 3 12 2 Keywords and environments eee ee 74 3 12 3 Requirements of bandplot section o o o 74 SA o 4 4 66 a ana lat Bow di eat je fie bi Bee tak s ie Phe Mets ged woo we FL 74 3 15 11 SMA eee ye a be AE eee 48 ee HE ate Si 74 3 13 2 Keywords and environments ee 75 3 13 3 Requirements of the kdos section o oo 75 3 14 Rxplot section coopera a ii eee ed eee eee eee dd 75 314 1 Sythe ee Ee eee ee eee de e ee Kd was 76 3 14 2 Keywords and environments 2 76 3 14 3 Requirements of the rxplot section o o o e TI 310 Aim SOCHOM e 3 23 9 64 44444 2446 eed Ze ente wl Pepe eB t ep de E 77 ZAIN ISO MW sooo E gy era yee dae oe aa a PS we eee s hae BGG By ie WAN al he OS 77 3 15 2 Keywords and environments e
69. eference structure FORCE CONV fconv r Force convergence criterion Note that reference values are only used if all three criteria are not given Default reference structure ITDIAG cyc i Use iterative diagonalization and do a complete diagonalization each cyc i cycles In versions of WIEN2k gt 13 08 2007 cyc i has no effect Default cyc i 3 NEW IN1 cyc i Generate a new basis set in cycle lt cyc gt Default CLEAN Append this clean line to the wien input file for the selected atom Default NOSEND Create the wien input file but do not run the atom for which this keyword applies atom global mechanism Default NICE nice i Process priority for every section of local wien Note that runwien child processes can not have a lower nice than runwien itself Default DO ALL NONE NEW OLD atnamel s atname2 s Used together with the EXCEPT or der below 63 EXCEPT atnamel s atname2 s As a default the FREE section will obtain the in vacuo en ergy for the atoms in the crystal molecular formula that have not been calculated before The DO EXCEPT orders can be used to modify this behavior Notice that the syntax of the order in this section differs slightly from their counterpart in other sections Here a list of atom name substitutes the range of integer indices On the other hand the FREE section is not affected by the general DO EXCEPT only by the local options Default calcu
70. eference structure cyc i 3 NEW INI cyc i Generate a new basis set in cycle lt cyc gt Default no even if it is set in reference structure NICE nice i Process priority for every section of local wien Runwien child processes can not have a lower nice than runwien itself Default reference structure s scf_ nice REUSE CHAIN REFERENCE PATH path s DETECT The REUSE keyword allows the user to con trol the reutilization of converged densities from one deformation point to the other in a sweep calculation The CHAIN keyword makes each point use the converged density of the previous one The first structure is calculated starting from the superposition of atomic densities With REFERENCE each structure extrapolates the charge density of the selected reference structure of the general section In a way similar to SCF section PATH allows the user to use consistently a given calculation as starting point from every geometry Finally if DETECT is read runwien searches for existing clmsum and struct files in the directories of the sweep section and uses the density files if found Default CHAIN NOREUSE Reusing the converged densities can lead to large time savings The NOREUSE keyword allows the user to avoid the reutilization of densities of previous points completely Default NOSUMMARY Don t run the summarizing script This script will generate a sweepsummary out file in root directory which contains useful information ab
71. ensity of states DOS Use gnuplot version 4 2 or superior for best appearance of the plot runwienti kdos default Same as above but assuming defaults on all variables 3 13 2 Keywords and environments PLOTXMIN plotxmin r RY Lower bound for energy in dos graph in eV If Ry is indicated runwien will assume lt plotxmin gt is in Ry The character may be used but the behavior is the same as if no plotxmin were issued Default if a dosplot section was executed in the same run use its plotxmin Else 12 0 PLOTXMAX plotxmax r RY Upper bound for energy in dos graph in eV If Ry is indicated runwien will assume lt plotxmax gt is in Ry The character may be used but the behavior is the same as if no plotxmax were issued Default if a dosplot section was executed in the same run use its plotxmax Else 8 0 PLOTYMAX plotymax r Upper limit for DOS plot Default maximum value for DOS in the selected x range KDOS ndos i UP DN MERGE UP DN The kdos keyword creates a kdos diagram The DOS used is the lt ndos gt line in the dos list see dosplot section If an up resp dn is appended to the lt dos gt identifier use the spin up resp spin down density of states Also the band structure diagram may be generated using the spin up Cup spin down dn or both merge spin contributions to the energy bands by writing the appropriate second field Default no extra kdos plots DO range
72. environments eee eee 51 3 43 Requirements of the initialization section o oeo 55 3 0 Presch SeGhlaul xi 4 4 4 8 8 gd wy See EE ed bh A Eh ee i 55 35 ONLO teat ae ah A de A afe AE ete re ee ee AAA 55 302 Keywords and environments 2 00444 aa bee Ee Re PR b 55 3 5 3 Requirements of the prescf section o 56 360 SerSection cit eek BA dan vie oo Se eee ee ee ee ee 56 36 1 2202 6eeee sete ee ee eee EEL EASE eee ee kok 56 3 6 2 Keywords and environments e ssas 446068444 kw oe ee eee eer Es 56 3 6 3 Requirements of the scf section 00000 58 at SpIMOrbit SECO 40 e ow ack dw se EL fa A S G de ee oo BLL 58 Jl IVAN AE ee ne ten YA Wa owi Moog DP ETE b ja AW A WA EE a POW PE 58 3 7 2 Keywords and environments s s s as ec aa ac aa erraia eaaa 59 3 7 3 Requirements of the spinorbit section LL 2 e 000022000 60 BB FFOSSECGION 3 3 5 4 gece a eR wa Bag eS SE ae Gk eae red w EB 60 38l IMA oe ee ee a ae ee UW ee 48 ee a dete a 60 3 8 2 Keywords and environments LL 0000 eee 61 3 8 3 Requirements of the free section o oo 64 3 9 Elastic section w kesa dda bb i di 44404 eee iia dag eee eee hn 64 SAID i ft Aran din vi f e eee sik ki ug de ka S SE Te A a ear r r 64 3 9 2 Keywords and environments a sosoo oe e e e e e e 66 3 9 3 Requirements of the elastic section LLL eee 70 DAQ Printrho Secon a ez sek a etat wy e ea E A A A
73. f a flexible summary of the most useful results This user manual is divided into two main parts The first part introduces runwien through the dis cussion of a selected set of examples The second part consists of a formal description of the orders implemented in runwien and some of the main mechanisms used in programming the interpreter 2 2 Obtaining and installing runwien See README 2 3 First steps Calculations on hcp Be Our first example corresponds to a calculation of hcp Be at the experimental NPT normal pressure and temperature geometry We use a large GMAX 20 0 value as recommended for light elements and fix an appropriate muffin tin radius RMT Let us take a look to the be wien input runwienti Comment lines must start with a as the first non blank character Blank lines spaces tabs and upper lower caps are ignored general Crystal description H corresponds to a hexagonal lattice Check the lattice types in the WIEN2K user guide lattice H equiv list Be WIEN2k P Blaha K Schwarz G K H Madsen D Kvasnicka and J Luitz WIEN2k An Augmented Plane Wave Local Orbitals Program for Calculating Crystal Properties Karlheinz Schwarz Techn Universitat Wien Austria 2001 ISBN 3 9501031 1 2 http www wien2k at 0 66666667 0 33333333 0 75000000 0 33333333 0 66666667 0 25000000 end equiv list relativistic yes cell parameters 4 3210478 4 3210478 6 7714557 90 90 120 rmt 1 1 8 gmax
74. f2x 2y F fm2m f2y 2z F fdd2 f2z 2xd F f2dd f2x 2yd F fd2d f2y 2zd F imm2 i2z 2x B i2mm 12x 2y B im2m 12y 2z B iba2 i2z 2xab B i2cb i2x 2ybc B icza i2y 2zca B ima2 122 2xa B i2mb 12x 2yb B ic2m 12y 2zc B ibm2 122 2yb B 12cm 12x 2z B im2a 12y 2xa B pmmm p 22 2x P pnnn p 2zab 2xbc P pecm p 22 2xc P pmaa p 2x 2ya P pbmb p 2y 2zb P pban p 2zab 2xb P 40 pncb mn p 2xbc 2yc a p 2yca 2za an p 2za 2xa ie p 2xb 2yb a p 2yc 2zc p 2ya 2xa en p 2zb 2yb p 2xc 2zc M p 2za 2xbc MO p 2xb 2yca p 2yc 2zab a p 2ya 2xbc a p 2zb 2yca a p 2xc 2zab n p 2zac 2x ken p 2xba 2y gt p 2ycb 2z b p 2yab 2x p 2zbc 2y p 2xca 2z A p 2za 2xac se p 2xb 2yba m at p 2yc 2zcb a p 2ya 2xab e p 2zb 2ybc e p 2XC 2zca On p 2z 2xab o p 2x 2ybc p 2y 2zca a p 2zab 2xac se p 2xbc 2yba SE p 2yca 2zcb e a p 2zc 2xb p 2xa 2yc sare p 2yb 2za m p 2yb 2xc n p 2zc 2ya ae p 2xa 2zb pen p 22 2xn l p 2x 2yn a p 2y 2zn p 2zab 2xa p 2xbc 2yb P 41 pmnm
75. field corresponds to the non equivalent atom index in the unit cell The second field selects the angular momentum quantum number l either by its value 0 1 2 or by the corresponding letter s p d The optional values defu r and defj r set the default values for U and J for the atom and 1 specified in the line The U Coulomb on site parameter and J exchange parameter characterize the LDA U correc tion These parameters are treated in runwien through the keywords U and J that can generate series of calculations in the same way as RMT NPT RO See below for more information U line i urange r Coulomb parameter for LDA U environment line line i Runwien accepts a range of U parameters in urange r Each of the U in the range generates a new structure much in the same way as the other calculation parameters listed above providing for an easy way of evaluating the effect of U on the system If no line i is indicated the first line is assumed This syntax however forces the range urange r to be a single field i e without spaces or tabs If line i the U values apply to all the lines in the LDA U environment Default defu r if it was set for line line i in the LDA U environment Otherwise 0 J line i jrange r Exchange parameter for LDA U environment line line i Runwien accepts a range of J parameters in jrange r Each of the J in the range generates a new structure much in the same way as the other calculati
76. g a new emax r in the in1 c file Default NEWKPTS kpts r kpts r In spinpolarized cases where symmetry may be reduced regenerate the k list with a number of k points equal to kpts r in the whole 1BZ If the value provided contains the character the kpts r value will be interpreted as a percentage of the original kpts for each of the structures run in the spinorbit section Default 100 same number of kpts as in the original MAX ITERATIONS maxit i Maximum number of cycles Default the value provided by the user in the SCF section otherwise 30 CHARGE CONV cconv r Charge convergence criterion Default none ENERGY CONV econv r Energy convergence criterion Default none FORCE CONV fconv r Force convergence criterion The default convergence criterion is the same as in the SCF section or energy conv 0 00001 if it was not set Default none NICE nice_value i Process priority for run_lapw Child processes can not have a lower nice than runwien itself Default 0 increase to decrease priority ITDIAG cyc i Use iterative diagonalization and do a complete diagonalization each cyc i cycles In versions of WIEN2k gt 13 08 2007 cyc i has no effect Default no it diag cyc i 3 if applicable NEW INI cyc i Generate a new basis set in cycle lt cyc gt Default DO range i ALL NONE NEW OLD Used together with the EXCEPT order below 59 EXCEPT range i The DO EXCEPT pair is used to select a su
77. generated in a previous run with LOADCHECK or LOADCHECK REREAD 3 3 2 Keywords and environments LATTICE lattice letter s Crystal Bravais lattice in WIEN2k naming convention That is P for primitive F for face centered B for body centered H for hexagonal etc See WIEN2k user s guide for details R lattices must be handled with special care In a R lattice the cell parameters are expected to be those of an hexagonal cell while the atomic positions are given in a rhombohedral setup The conversion between these two is uniquely defined and documented in the WIEN2k user s guide Compulsory parameter EQUIV LIST environment Formed as EQUIV LIST symbol s X Y y r Z r END EQUIV LIST List of equivalent positions for a non equivalent atom in the unit cell in crystallographic units Symbol is the symbol for the atom case insensitive The symbol representing an atom uses the same conventions as WIEN2k the identity of the atom is determined from the first two one letters in the name Further additions to the atom symbol are allowed and they are passed down to the struct file Several equiv list environments are allowed each of them corresponding to a new non equivalent atom Compulsory parameter SPGLIST environment Formed as SPGLIST spg s RHOMB at s X r y r Z r END SPGLIST The spglist environment determines the crystal geometry completely along with the cell param eters It accepts the space group label spg
78. ghtly displaced from the zero strain position to ensure that all points are calculated with the same low symmetry This displacement is 0 001 bohr in distances and 0 005 degrees in angles The POINTS MAXLENGHT and MAXANGLE keywords may exchange positions Default 7 points 0 05 in maxlength and 5 0 in maxangle POLYORDER ord i Order of the fitting polynomial The only valid orders are 2 4 6 and 8 Default order 4 FIXMIN YES NO If the user activates the fixmin fixmin yes then the point around which the polynomial is expanded is fixed to the zero strain value 0 0 in general 0 5 for the hexagonal monoclinic deformation Otherwise the equilibrium point is optimized in the fitting process Default no TERMI YES NO The term keyword instructs runwien to include in the polynomial the 1st grade term al x x0 Including this term is equivalent to the variation of x0 allowed by FIXMIN NO If term1 is selected then FIXMIN is forcibly set to YES Default no NPT atom i n i Number of radial mesh points for non equivalent atom with index atom i Must be an odd number Wien2k manual recommends 381 for lda calculations 781 for gga Please note that the argument for npt is not a list of values i e npt is fixed throughout all the sweep run auto can be used and it will be passed into local wien files Auto keyword will make runwien accept runwien s default value Npt value will not change from one structure to
79. he default value for rkmax It is important to keep in mind that when an ALSO order is found all the variables are reset to their defaults That is if you set the mixing parameter mix to 0 05 and then write an ALSO structures created after that point will have the default value of mix 0 40 and not 0 05 There is another way to couple variables the LINK keyword This keyword accepts two or more variables When the structures are created these variables run simultaneously over their respective values For example runwien link lmax 1nsmax lmax 10 11 12 lnsmax 4 5 6 will create three structures with values of Imax lnsmax 10 4 11 5 12 6 The variables keep linked even when an ALSO keyword is found All the variables linked must have the same number of values and runwien will issue an error if this requirement is not met For a calculation to start and to be correctly run all necessary data must be provided through input There is a relatively small set of indispensable parameters lattice type equivalent atom list and positions in the unit cell and cell parameters This information can be provided by e Indicating LATTICE EQUIV list and CELL PARAMETERS in the general section 34 e Indicating SPGLIST and CELL PARAMETERS in the general section e Loading this information from a cif file with the LOADCIF keyword e Loading it from a struct file using the LOADSTRUCT keyword e Loading it from a check file
80. hexagonal structure nor to indicate the geometry of an orthorhombic system using v c a and a Notice that the volumes entered in the WITH order are conventional cell volumes but the ones in the synopsis and in the summary files are volumes per unit formula Be very careful when copying and pasting from an output to an input Default one structure equal to the reference structure ALSO Used to separate two or more different and independent grid scans For example runwien with c a 1 82 with v 300 500 50 also with c a 2 5 2 6 with v 310 corresponds to two different scans of the cell volume the first one having fixed the ratio c a and the second one a fixed volume Default none 84 PRINT zvar s VS yvar s VS xvar s Variables to be used in a 2D or 3D property versus geometry plot lt zvar gt or if z is missing lt yvar gt may be one of energy efermi planarity topology a b c alpha beta gamma v c a c b b a Each specification of these keywords generates a file lt root gt sweep print lt n gt where n is the print line number and root the calculation root name If lt z var gt exists a xyz file is created first column x second column y third column z adequate for gnuplot If only lt x var gt and lt y var gt are present an xy file is created Finally if adequate gnuplot scripts are present and indicated in constant part of runwien s code they can be run to create the plot You can a
81. i ALL NONE NEW OLD Used together with the EXCEPT order below EXCEPT range i As a default bandplot will be run only on the new points skipping those points already examined The DO EXCEPT orders can be used to modify this behavior Default Use the local DO EXCEPT specification if available Use the global DO EXCEPT data otherwise In absence of user orders do the bandplot task only on new points 3 13 3 Requirements of the kdos section General initialization prescf scf dosplot and bandplot sections are required 3 14 Rxplot section 75 3 14 1 Syntax runwien RXPLOT DEFAULT ATOM nneq i N n i L 1 i PLOTXMIN emin r PLOTXMAX emax r DE de r TYPE EMIS ABS IN1MAXENERGY emax r DO range i ALL NONE NEW OLD EXCEPT range i END RXPLOT Prints calculated x ray emission or absorption spectra for general structures The rxplot section uses WIEN2k xspec facility to obtain the spectra data but then creates a gnuplot file to print it As in previous sections this gnuplot file is saved in the general structure subdirectory runwien rxplot default Same as above but assuming defaults on all variables 3 14 2 Keywords and environments ATOM nneq i Calculate the spectra for the indicated non equivalent atom Default 1 N n t Quantum number n for orbital involved in electronic absorption or emission Default 1 L Li Quantum number 1 for orbital involved in electronic absorption or emission Default 0
82. i CP id i up i down i CPALL BCPALL up i down i 79 RBCPALL bup i bdown i rup i rdown i CHECK x r y r z r ENDCHECK RHO DENSITY LAP LAPLACIAN LOG ATAN BADER nptsu i nptsv i niso i END GRDVEC Include a grdvec environment in the critic input file Grdvec plots a plane containing the gradient paths originating from a critical point For a detailed explanation of the syntax see the critic program documentation Default DOPLOT DENSITY RHO GRADMOD LAPLACIAN LAP GRADIENT FORMAT DEFAULT GNUPLOT D2D D Generalized plot function in critic Allows plotting of 1d 2d and 3d representations of density gradient and laplacian For an exhaustive description of the input consult critic s manual Default USE RHO VALRHO ATOMIC The USE keyword switches the clmsum used in critic input to that of valence density valrho and promolecular density atomic Default rho NOSUMMARY Don t generate the critic summary Default DO range i ALL NONE NEW OLD Used together with the EXCEPT order below EXCEPT range i As a default critic will be run only on the new points skipping those points already examined The DO EXCEPT orders can be used to modify this behavior Default Use the local DO EXCEPT specification if available Use the global DO EXCEPT data otherwise In absence of user orders do the critic task only on new points 3 16 3 Requirements of critic section Requires general initialization prescf and s
83. in the topological analysis of the electron density of NaCl To that end the LM list for the expansion of the density in the muffin tins is extended up to LMAX 10 runwien lm list 1 0040446064808 48 8 10 0 10 4 10 8 end lm list lm list 2 0040446064808 48 8 10 0 10 4 10 8 end lm list Note that this LM list most likely will require you to recompile WIEN2k with an augmented LMAX2 parameter However for the sake of clarity you can omit the lm list specification in the input file defaults to LMAX 6 The basis set specification follows almost directly copied from the in1 file of our previous calculation The prescf and scf sections are run with the default values and finally the localization of the critical points for all the calculated systems is requested the critic program is needed for this step Along with the energy we are going to use the electron density flatness as a criterion to assess the adequacy of the WIEN2k parameters The electron density flatness flatness is a measure of the planarity of the electron density in the interstitial sometimes the term planarity is used instead of flatness but they are equivalent It is calculated as the ratio between the global minimum of the electron density and the maximum density at the bond critical points Run this calculation after removing the old output files tasciit rm rf nacl index out runwien awk nacl wien In approximately one hour the calculat
84. ion finishes The part of the synopsis out file relevant to us is ascii Calculation variable parameters number of structures 20 num rmt 1 rmt 2 rkmax kpts basis it warning energy ry N efermi ry time s planarity morsesum topology 1 2 00 2 00 7 50 2000 296 9 1248 1409550 0 05629 164 0 154138 0 2 4 4 0 12 24 24 1 48 11 8 2 2 25 2 00 7 50 2000 296 9 1248 1415330 0 06395 146 0 156605 0 2 4 4 0 12 24 24 11 48 11 8 3 2 50 2 00 7 50 2000 296 9 1248 1409760 0 07045 146 0 157923 0 2 4 4 0 12 24 24 1 48 11 8 4 2 00 2 25 7 50 2000 296 9 1248 1412770 0 07409 145 0 153908 0 2 4 4 0 12 24 24 11 48 11 8 5 2 25 2 25 7 50 2000 182 9 1248 1398230 flatness P Mori S nchez A Mart n Pend s and Victor Lua a A classification of covalent ionic and metallic solids based on the electron density J Am Chem Soc 124 2002 14721 14723 23 0 08988 114 0 154612 0 2 4 4 0 2 24 24 1 48 11 8 6 2 50 2 25 7 50 2000 182 9 1248 1400320 0 09598 115 0 155934 0 2 4 4 0 2 24 24 1 48 11 8 7 2 00 2 50 7 50 2000 296 7 1248 1407630 0 09291 114 0 155285 0 2 4 4 0 2 24 24 1 48 1 8 8 2 25 2 50 7 50 2000 182 9 1248 1399670 0 10967 113 0 156159 0 2 4 4 0 2 24 24 1 48 11 8 9 2 50 2 50 7 50 2000 150 9 1248 1388390 0 12442 101 0 158959 0 2 4 4 0 2 24 24 1 48 11 8 10 2 50 2 50 7 50 2000 150 9 1248 1388390 0 1
85. itions entered as crystallographic coordinates Default don t calculate any line NICE nice i Set the critic process priority Some critic tasks determining the atomic basins and inte grating properties in them require a large cpu time so it is useful to lower the process priority in this case Notice that runwien child processes can not have a lower nice than runwien itself Default 0 INTEGRALS nr i ntheta i nphi i Calculate the atomic basin integrals that currently provide the atomic charge and volume for each basin Use lt nr gt radial points and ntheta and lt nphi gt angular points The algorithm used in critic is a bisection like method where the criterion for bisection is the nuclear cp where the gradient line ends starting from an arbitrary point Default do not integrate BASINPLOT level i delta r theta r phi r Plot attraction basin for each of the ncp Basinplot starts with an octahedron and subdivides it level times if level gt 0 or starts with a cube and subdivides it lt level gt times if level 0 Then it finds the zero flux surfaces bisection method on the rays determined by the resulting polyhedron Delta is the precision cryst coords with which the surface is determined in each ray Occasionally a ray is traced along a high symmetry line and the bisection method fails to determine the true surface A small rotation in theta and or phi le 3 or 1e 4 typically can then be applied t
86. late only the new atoms The global DO EXCEPT pair has no effect 3 8 3 Requirements of the free section General initialization and prescf sections are required 3 9 Elastic section 3 9 1 Syntax runwien ELASTIC DEFAULT REFERENCE STRUCT n i DEFORMATION POINTS n i MAXLENGTH mlenght r MAXANGLE mangle r POLYORDER ord i FIXMIN YES NO TERM1 YES NO NPT atom i n i RMT atom i radius r RO atom i r r RKMAX r r LMAX 1 i LNSMAX 1 i GMAX g r MIX lambda r KPTS k r LDA U SIC AMF HMF iati i li i li s u r j r iat2 i 12 i 12 s u r j r END LDA U XCPOTENTIAL LSDA GGAPBE96 GGAPW91 GGAWCO6 ECOREVAL ecore r ENERGYMIN emin r ENERGYMAX emax r ORBITALS neq_atom i global energy r global_apw r orb l i orb energy r orb var i CONT STOP apw i END ORBITALS LM LIST neq_atom i 11 mi 12 m2 13 m3 lk mk 1k 1 mk i END LM LIST FERMI ROOT TEMP TEMPS GAUSS TETRA ALL value r 64 MAX ITERATIONS miter i CHARGE CONV cconv r ENERGY CONV econv r FORCE CONV fconv r ITDIAG cyc il NEW INI cyc i NICE nice i REUSE CHAIN REFERENCE PATH DETECT NOREUSE MINI commandline s SPINORBIT END SPINORBIT CLEAN WIEN FULL NOSEND END ELASTIC The elastic sections allows the user to calculate the elastic constants of a general structure This section expands the capabilities of WIEN2k by providing a means for calculating the elastic constants of cubic and hexagonal systems The calcula
87. lculation into runwien using the LOADCHECK system As a final remark remember this program is provided with no warranty at all You should always double check your data for errors 8 Copyright notice Copyright c 2007 8 Alberto Otero lt alberto carbono quimica uniovi es gt and Victor Lua a lt victor carbono quin Universidad de Oviedo 94 runwien awk is free software you can redistribute it and or modify it under the terms of the GNU General Public License as published by the Free Software Foundation either version 3 of the License or at your option any later version runwien awk is distributed in the hope that it will be useful but WITHOUT ANY WARRANTY without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE See the GNU General Public License for more details You should have received a copy of the GNU General Public License along with this program If not see http www gnu org licenses Generated on 2008 12 31 00 44 12 CET Generated by Docutils from reStructuredText source 95
88. lization can not be run unless general has been run or loaded 3 5 Prescf section 3 5 1 Syntax runwien PRESCF KGENOUTPUT LONG SHORT KGENSHIFT YES NO NICE nice_level i DO range i ALL NONE NEW OLD EXCEPT range i END PRESCF Runs programs which are the last step before scf calculation These include kgen generates plane waves and dstart initial density In kgen inversion is always added to the k point list if asked runwien prescf default Runs prescf section assuming default values for all variables 3 5 2 Keywords and environments KGENOUTPUT LONG SHORT Select the printing level of kgen output This is asked for only in older versions of WIEN2k lt 20080421 Default short KGENSHIFT YES NO If it is allowed shift k point mesh away from high symmetry directions Default yes NICE nice level i Process priority for dstart can take quite long Child processes can not have a lower nice than runwien itself Default 0 increase to decrease priority 55 DO range i ALL NONE NEW OLD Used together with the EXCEPT order below EXCEPT range i The DO EXCEPT pair is used to select a subset of calculations to run when more that one calculation is implied from the input data The behavior of this orders is described in the GLOBAL section Default Use the local DO EXCEPT pair if present otherwise apply the global DO EXCEPT pair if present if neither is defined calculate only ne
89. local orbitals plane waves basis functions of that l Subsequent orb l will refer to local orbitals orb energy is the linearization energy orb var gt 0 commands wien to search for a good linearization energy and if stop is set stop in case it is not found Finally use lt apw gt 1 in case you want an apw basis function type 0 in case of lapw See WIEN2k s manual for more information about basis Default reference structure LM LIST environment The syntax of this environment corresponds to runwien LM LIST neq_atom i 11 mi 12 m2 13 m3 lk mk 1k 1 mk 1 END LM LIST 85 Specify LM lattice harmonics expansion of charge density See wien s manual for L M pairs adequate for each structure Default reference structure FERMI ROOT TEMP TEMPS GAUSS TETRAJALL value r Method to find fermi energy See WIEN2k s manual for details on each of the method Value is a method specific characteristic number Default reference structure MAX ITERATIONS miter i Maximum number of cycles Default reference structure CHARGE CONV cconv r Use a charge convergence criterion Default reference structure ENERGY CONV econv r Energy convergence criterion Default reference structure FORCE CONV fconv r Force convergence criterion Default reference structure ITDIAG cyc i Use iterative diagonalization and do a complete diagonalization each cyc i cycles In versions of WIEN2k gt 13 08 2007 cyc i has no effect Default r
90. lot This affects the accuracy of the elastic constants Estimated time 18m 29s test005 Be hcp example of usage of elastic in an hexagonal crystal Estimated time 2m 29s test006 CaF 2 the cohesive energy of CaF2 is calculated using the free section Estimated time 9m 2s test007 Cr hcp several examples of the usage of printrho in ferromagnetic hcp Cr Total valence spin atomic and deformation densities are plotted Also total Coulomb and exchange correlation potentials Estimated time 28m 20s test008 Sr fcc example of dosplot The total atomic and interstitial and angular contributions to the DOS are plotted Several joint plots are also calculated including total atomic interstitial atomic s p d f and d dz2 dx2y2 dxy dxz dyz Estimated time 7m 53s test009 Cr hcp another DOS example with spinpolarized ferromagnetic hcp Cr The plots are the analogous to test008 but taken spin contributions into account Estimated time 20m 33s test010 Ca fcc example of the rxplot section Estimated time 1m 54s test011 MgO example of bandplot section for a typically ionic system Estimated time 3m 18s test012 Mg hcp bandplot section for a metallic system Estimated time 8m 17s test013 Li topological analysis of Li using the AIM program The position of the critical points is determined Estimated time 11m 34s test 014 Li same as test013 but using CRITIC our own AIM code E
91. ls on each of the method Value is a method specific characteristic number Default set by WIEN2k tetra 0 000 In case no value is provided use 0 000 SGROUP Use the struct file proposed by sgroup This struct file contains the optimum geometrical parameters cell lengths origin etc for the calculation However sometimes the user may be interested in overriding this behavior For instance if he wants to calculate a system with non equivalent atoms of the same type at symmetry equivalent positions supercells etc An example of this is bcc Cr where both atoms in the unit cell are non equivalent allows antiferromagnetic ordering runwien equiv list Cr 1 0 0 0 0 0 0 end equiv list 54 equiv list Cr 2 0 5 0 5 0 5 end equiv list If sgroup is run on the struct file generated by this input both atoms are made equivalent Due to these reasons copying the sgroup struct file is disabled by default Default do not use it DO range i ALL NONE NEW OLD Used together with the EXCEPT order below EXCEPT range i The DO EXCEPT pair is used to select a subset of calculations to run when more that one calculation is implied from the input data The behavior of this orders is described in the GLOBAL section Default Use the local DO EXCEPT pair if present otherwise use the global DO EXCEPT pair if present if neither is defined calculate only new structures 3 4 3 Requirements of the initialization section Initia
92. m1 p3 2 H palm p3 2 H p3cl p3 2 c H p3lc p3 2c H r3m r3 2 R r3c r3 2 c R p 31m p3 2 H p 31c p3 2c H p 3m1 p3 2 H p 3c1 p3 2 c H r 3m 13 2 R r 3c r3 2 c R p6 p6 H p61 p61 H p65 p65 H p62 p62 H p64 p64 H p63 p6c H p 6 p 6 H p6 m p6 H p63 m p6c H p622 p6 2 H p6122 p61 2 0 0 1 12 H p6522 p65 2 0 0 1 12 H p6222 p62 2 0 0 1 3 H p6422 p64 2 0 0 1 6 H p6322 p6c 2 40 0 1 4 H p6mm p6 2 H p6cc p6 2c H p63cm p6c 2 H p63mc p6c 2c H p 6m2 p 6 2 H 45 p 6c2 p 6c 2 H p 62m p 6 2 H p 62c p 6c 2c H p6 mmm p6 2 H p6 mcc p6 2c H p63 mcm p6c 2 H p63 mmc p6c 2c H p23 p2 2 3 P f23 f2 2 3 F 123 12 2 3 B p213 p2ac 2ab 3 P 1213 i2ac 2ab 3 B pm 3 p2 2 3 P pn 3 p2ab 2bc 3 B fm 3 f2 2 3 F fd 3 f2uv 2vw 3 F im 3 12 2 3 B pa 3 p2ac 2ab 3 P ia 3 i2ac 2ab 3 B p432 p4 2 3 P p4232 p4n 2 3 P f432 f4 2 3 F f4132 f4d 2 3 F i432 i4 2 3 B p4332 p4bdn 2ab 3 P p4132 p4bd 2ab 3 P 14132 i4bd 2ab 3 B p 43m p 4 2 3 P f 43m f 4 2 3 F i 43m i 4 2 3 B p 43n p 4n 2 3 P f 43c f 4c 2 3 F i 43d i 4bd 2ab 3 B pm 3m p4 2 3 P pn 3n p4a 2bc 3 P pm 3n p4n 2 3 P pn 3m p4bc 2bc 3 P fm 3m 14 2 3 F fm 3c f4n 2 3 F fd 3m f4vw 2vw 3 F fd 3c f4ud 2vw 3 F im 3m 14 2 3 B ia 3d i4bd 2ab 3 B 46 Compulsory parameter CELL PARAMETERS ANGSTR
93. me 6m 34s e test024 As elemental As The spglist environment is used in this example As has a rhombo hedral unit cell which is fed into runwien using the spglist rhomb keyword Estimated time 2m 39s 6 Programming techniques used in runwien Runwien is released under a GPL license This means you can modify the code to fit your needs provided the resulting code is also under the GPL and you don t remove our names from it However in order to do this you must take into account how runwien could reach the size it has now without ending up in a programming chaos This was achieved by following some simple programming conventions e AWK is a scripting language It does not have fixed variables types nor local variables so a big deal of care must be taken when creating or removing variables In runwien every variable has a prefix which delimits its scope For example elastic section variables have the prefix elastic elastic_ points elastic_ maxlength A variable with a section prefix should never be modified outside of its section Some variables are global and receive the global prefix and can only be written in general and initialization sections Temporal variables and scratch variables have the temp prefix you should try to avoid carrying information over sections with them Constants have the const prefix and must not be modified e Variable naming is a problem in functions It is extremely important to avoid variable c
94. mechanism is that the crystallographic coordinates of the atoms within the cell cannot be changed The sweeping mechanism is thus limited to the cell parameters 2 5 Continuing and completing a calculation There are many occasions specially when exploring large PES when you will not do all the calculations in a row For instance your initial set of geometries does not bracket the energy minimum appropriately and you want to add points to the grid or some kind of problem has stopped your run in the middle of a calculation or simply you want to distribute your calculations on a collection of machines and afterwards put back together for the final summary Whatever the circumstance runwien provides a method for stopping and resuming calculations by using checkpoint files These files are automatically compiled during the run and form a kind of black box of the calculation The LOADCHECK order allows the user to load all this information back into runwien This CHECK LOADCHECK mechanism can also be useful if runwien run is interrupted If that happens you can manually clean the point which was being calculated and restart from that point Let us examine an example of this recovery mechanism using again the hcp Be PES as a test case The calculation will be divided into three main parts first create the sweep tree structure and input files then run the runwien jobs under the sweep subdirectory manually and then load back the results in
95. modynamics of solids from energy curves using a quasi harmonic Debye model Comput Phys Commun 158 2004 57 72 Mol formula Bel Mol mass 9 012182 a bohr 4 32104800 b bohr 4 32104800 c bohr 6 77145600 alpha deg 90 00000000 beta deg 90 00000000 gamma deg 120 00000000 c a 1 56708650 c b 1 56708650 b a 1 00000000 volume bohr73 109 49413924 Mol volume 54 74706962 space group P63 mmc lattice H system hexagonal nonequivalent atom list in struct and position of a representative atom atom 1 is Be mult 2 x 0 66666667 y 0 33333333 z 0 75000000 Additional information Number of electrons in unit cell 8 Valence electrons in unit cell 4 Core electrons in unit cell 4 Core val total e for atom Be 21 2 4 Core val el conf for atom Be 18 2 28 2 Core leaking e for atom Be 1 08300e 03 Atomic energy lstart Be 29 2657470 Total in vacuo energy lstart 29 2657470 Calculation fixed parameters spinpolarized no relativistic first density yes general npt 781 general rmt atom 1 1 80000 general rO 0 00010000 general rkmax 7 00000 general lmax 10 general lnsmax 4 general gmax 20 000 general mix 0 40000 general kpts 1000 initialization potential ggapbe96 initialization ecoreval ry 6 000 initialization emin ry 9 00000000 initialization emax ry 2 00000000 scf commands used i 50 ec 0 00001 Only one str
96. ndex file Each subdirectory starts with one wien file When every wien file is written runwien batch calculate all of them unless stated otherwise by the user The same considerations for the variation of parameter as in the general section apply in this section Thus if you request runwien to calculate a 4 3 4 4 and c 6 7 6 8 four structures will be calculated with all the possible combinations One difference is that in sweep section there is no link option The also keyword is available too At the end of the sweep section a summarizing script will be run and a summary file sweepsum mary out will be created in the root directory containing the relevant information of the sweep run The sweep section has no default orders Enter explicitly whatever you want it to perform 3 17 2 Keywords and environments REFERENCE STRUCT n i Select which structure from the general section will be used as reference Any parameter or condition not explicitly set will be inherited from this reference Default the first structure that appears in the general section NPT atom i pt i Number of radial mesh points for non equivalent atom index atom i Must be an odd number WIEN2k manual recommends 381 for lda calculations 781 for gga Please note that the argument for npt is not a list of values i e npt is fixed throughout all the sweep run auto can be used and it will be passed into local wien files Auto keyword will make runwien a
97. ntials Required in initialization same as above Default reference structure MIX mixfactor r Mixing factor used in the creation of the new charge density Default reference structure KPTS kpts i Number of k points in the whole first brillouin zone Default reference structure LDA U environment Formed as LDA U SIC AMF HMF iat1 i 11 i l1 s wer jer iat2 i 12 i 12 s u r jer END LDA U Includes an orbital dependent correction to the exchange correlation potential This method allows LDA or GGA calculations to successfully reproduce the properties of the so called strongly correlated systems Common systems of this kind are transition metal oxides and rare earth elements and compounds For more information see the WIEN2k user s guide and references therein The first keyword in the environment selects the flavor of LDA U e SIC Self Interaction Correction Anisimov et al 1993 e AMF Around the Mean Field Czyzyk et al 1994 e HMF Hubbard model in Mean Field Anisimov et al 1991 Each line in the LDA U environment activates the U correction for a particular atom and atomic symmetry 1 The first field corresponds to the non equivalent atom index in the unit cell The second field selects the angular momentum quantum number 1 either by its value 0 1 2 or by the corresponding letter s p d 83 The U Coulomb on site parameter and J exchange parameter characterize the LDA U cor recti
98. nts 4 5 6 7 8 and 10 prescf default scf default critic do 3 Run critic task only on point 3 end critic the GENERAL and INITIALIZATION sections will affect to 20 points the PRESCF and SCF steps to 6 cases and the CRITIC task to a single case Default calculate only the NEW structures PARALLEL machines s Information about the details of a run using parallelization is done through a machines file in WIEN2k see WIEN2k user s guide for details The PARALLEL keyword indicates the location of the machines file and commands runwien to run the calculations in parallel using the p flag The machines file is copied to the directory where the calculations are actually carried out The file is also passed down to the elastic free and sweep sections The method of generating machines files in a batch system job using a script some examples are given in the WIEN2k FAQ can also be applied as long as there is a pointer to the newly generated machines file 32 If no machines s is indicated the parallelization is explicitly negated canceling the effect of previ ous PARALLEL orders Note that PARALLEL related variables are not stored in the check files so this keyword must be indicated in a runwien execution using LOADCHECK if it is intended to run parallelized Default serial calculations DUPLICATE range i Duplicate the calculations indicated in the range i The numbers parsed from range i are treated sequentially For each s
99. o prevent this problem Default do not plot basin files IWS NOIWS i i Defines the criteria to be used in the search for critical points The search will start on a partition of the whole unit cell option NOIWS or of the irreducible part of the Wigner Seitz cell option IWS lt i gt degrees of subdivision will be used in any case Default noiws and i 3 or whatever is the current critic default NEWTON r r Do an automatic search for the critical points of the electron density using Newton s algorithm If critic fails to find all the critical points try first to increase the subdivision level in the IWS or NOIWS order There are other techniques for this same purpose but Newton s is the best method on most cases and it is the only one that can be used through runwien Edit the critic input file and run critic interactively if you want to explore other possibilities Notice that by default critic section does not calculate anything Newton must be indicated in order to find the critical points of the electron density Also be careful with exponential numbers with d or q as the exponential symbol as AWK does not understand this notation Default r le 15 GRDVEC environment The syntax of this environment corresponds to runwien GRDVEC FILES rootname s PLANE x0 r yO r zO r xi r yi r zi r x2 r y2 r z2 r SCALE sx r sy r PARAM radi r rad2 r rad3 r step r endpt r proj i ORIG x r y r z r atr i up i down
100. of all the atoms composing a given reference system one of the general structures by calculating a fcc structure with only one atom type only 1 k point and very large cell parameter The atomic in vacuo free energies are then used to calculate the total in vacuo energy of the system This free energy may be used in the gibbs section as reference structure and to calculate the dissociation energy of the crystal The calculations required for this task are stored in lt root gt free lt at gt where lt root gt is the root for the calculation and lt at gt the atom name runwien free default Runs free section assuming default values for all variables 3 8 2 Keywords and environments ATOM atname s Atom to which the following keywords apply Free section calculates the in vacuo energy of every atom type in the molecular formula However this task can be worked out with different calculation levels for different atoms With atom lt atname gt you indicate that the following lines apply only to the atom whose symbol is lt atname gt until runwien finds another atom keyword or a global keyword see below Orders appearing before any ATOM specification i e right after the free keyword will apply to all the atom types Default GLOBAL After this keyword is read all the input refers to all the atom types For example if you are calculating CaTiO3 and wish to set up a cell parameter of 20 0 for Ca 21 0 for Ti and 22 0 fo
101. ollision when functions call other functions For this each function is assigned an integer that corresponds 93 to its level in the calling tree Thus a function that is only called in the main program is level 1 a function that is called in the main program and by level 1 functions is level 2 etc If n is the level of the function its variables must have the local lt n gt prefix e Runwien is a modular program You could for example remove a non fundamental section by eliminating the input reading section the execution section and all the references and related functions This removal is easier than it sounds Adding a new section is equally easy If you plan to do this use any of the other sections as templates e There are some mechanisms your new sections must honor LOADCHECK LOADCHECK REREAD SAVECHECK Design a checkfile and put into it the variables you want to save if the user decides to work with a two step process This usually means result variables for properties sections and almost all variables for general sweep and the like e The interesting results of your section should end up in variables accessible for the rest of the sections This is specially important if you want to pass information to synopsis For example the energy obtained from the SCF calculation goes into scf_energy which is later read by synopsis Modify synopsis if you think the quantities obtained are important enough e do except time con
102. on These parameters are set using the u r and j r values respectively Default reference structure WITH A B C ALPHAJBETAJGAMMA V C A C B B A range r Grid specification A with line specifies a varying parameter for current crystal system range follows the same lt min gt lt max gt lt step gt syntax as the general section For example one can vary the parameter of a cubic structure with a line runwien with a 9 0 12 5 0 5 12 7 12 8 You can also use in the specification if you like In that case quantities are calculated against the reference structure s parameters A structure of lesser symmetry must have more with keywords added For example an hexagonal lattice can vary runwien with c a 1 5 1 82 2 with v 300 500 50 600 Please note all parameters refer to conventional unit cell dimensions and atomic units Angles are measured in sexagesimal degrees For reference structure s crystal system only certain parameters and parameter rations are allowed by runwien to vary These are e cubic a or v e hexagonal a and c or c a and v e rhombohedric a and alpha e tetragonal a and c e orthorhombic a b and c e monoclinic a b c and alpha e triclinic a b c alpha beta gamma These and only these parameters are permitted for a given crystal system All other parameter specification different from the above will be ignored For example it is not allowed to vary v and b a for an
103. on parameters listed above providing for an easy way of evaluating the effect of J on the system If no line i is indicated the first line is assumed This syntax however forces the range jrange r to be a single field i e without spaces or tabs If line i the J values apply to all the lines in the LDA U environment Default defj r if it was set for line line i in the LDA U environment Otherwise 0 3 3 3 Requirements of the general section The GENERAL section does not require previous sections to be run In fact this is almost surely the first section you will be running It needs however some essential variables lattice cell parameters and non equivalent list of atoms This variables can be set either manually via the correspondent keywords reading a struct reading a cif or loading a previous calculation 3 4 Initialization section 3 4 1 Syntax runwienti INITIALIZATION DEFAULT 50 XCPOTENTIAL LSDA GGAPBE96 GGAPW91 GGAWCO6 ECOREVAL core_energy r ENERGYMIN min_energy r ENERGYMAX max energy r NNFACTOR nn dist r ORBITALS neg atom i global energy r global apw i orb 1 i orb energy r orb var i CONT STOP apw i END ORBITALS LM LIST neq atom i 11 mi 12 m2 13 m3 1k mk 1k 1 mk 1 END LM LIST LM LIST neq_atom i lmax i FERMI ROOT TEMP TEMPS GAUSS TETRA ALL value r DO range i ALL NONE NEW OLD EXCEPT range i END INITIALIZATION Generates all the necessary input
104. onds to 1 2 3 10 whereas 0 5 0 5 is equivalent to 0 0 5 1 5 e The crystal geometry can be read in from several sources As a general rule but in particular when loading from a cif file care must be taken with precision in the atom coordinates a 1 3 carelessly written as 0 3333 will give rise to an incorrect Wyckoff position and to the wrong multiplicity Use 0 333333333 and similar to make sure this does not happen In addition cif files downloaded from databases use to have ms dos newline format and need to be converted before being used with runwien One more warning be careful with atom names crystallographers tend to assign names to atoms such as O h1 Ca2 and the like These names will confuse runwien and WIEN2k e Stdout of almost every program is stored as lt root gt lt nameofprogram gt err or similar Error files coming from the scf process are assigned the name lt root gt lt nameofprogram gt error e Try to keep the structure and keyword list without tabs and extra blanks This will make your life easier later on see the description of RMT in the GENERAL section 3 2 Global keywords and environments runwien EXIT CLEAN WIEN FULL UNDO lt section gt LOADCHECK REREAD SYSTEM order s DO range i ALL NONE NEW OLD EXCEPT range i PARALLEL machines s DUPLICATE range i 30 These keywords and environments must appear outside any section EXIT Terminates
105. onment corresponds to runwien LM LIST neq_atom i 11 mi 12 m2 13 m3 lk mk 1k 1 mk 1 END LM LIST Specify LM lattice harmonics expansion of charge density See wien s manual for L M pairs adequate for each structure Default reference structure FERMI ROOT TEMP TEMPS GAUSS TETRA ALL value r Method to find fermi energy See WIEN2k s manual for details on each of the method lt Value gt is a method specific charac teristic number Default reference structure MAX ITERATIONS miter i Maximum number of cycles Default reference structure CHARGE CONV cconv r Charge convergence criterion Default reference structure 68 ENERGY CONV econv r Energy convergence criterion Default reference structure FORCE CONV fconv r Force convergence criterion Note that the default values only will be used if all three criteria are undefined Default reference structure ITDIAG cyc i Use iterative diagonalization and do a complete diagonalization each cyc i cycles In versions of WIEN2k gt 13 08 2007 cyc i has no effect Default cyc i 3 NEW INI cyc i Generate a new basis set in cycle lt cyc gt Default NICE nice i Process priority for every section of local wien Runwien child processes can not have a lower nice than runwien itself Default reference structure s scf_ nice REUSE CHAIN REFERENCE PATH path s DETECT The REUSE keyword allows the user to con trol the reutilization of converge
106. order uncomment the corresponding line in the beginning of runwien variable const allow system DO range i ALL NONE NEW OLD Used together with the EXCEPT order below EXCEPT range i The DO EXCEPT pair can be used to select a subset of calculations The de fault in absence of DO EXCEPT orders is to perform only the calculation of new structures avoiding the cases already declared as done within the check status files The apparition of a Before cleaning runwien needs to determine the value of two important internal variables general iterations and sweep iterations They are calculated in general section and sweep section respectively Alternatively they can be read from the checkfiles in the course of the LOADCHECK task Both variables are used to determine the number and the names of the lt case gt directories involved in the run Runwien prints to the standard output the cleaning operations done so you can check if everything has been correctly done 2The remaining sections are GENERAL INITIALIZATION PRESCF SCF PRINTRHO DOSPLOT RXPLOT BANDPLOT KDOS AIM and SYNOPSIS You can run PRINTRHO DOSPLOT RXPLOT BANDPLOT AIM and SYNOPSIS several times each new run overwriting the previous files You can not undo GENERAL INITIAL IZATION PRESCF and SCF but you can rerun the old structures with LOADCHECK DO OLD or similar see below However it is not possible to change their parameters Of course you can al
107. orders can be used to modify this behavior Default Use the local DO EXCEPT specification if available Use the global DO EXCEPT data otherwise In absence of user orders do the printrho task only on new points 3 10 3 Requirements of the printrho section This section requires general initialization prescf and scf sections 71 3 11 Dosplot section 3 11 1 Syntax runwien DOSPLOT PLOTUNITS EV RY ENERGYMIN emin r EV ENERGYMAX emax r EV DE de r EV BROAD broad r DOS LIST atom i descr r label s END DOS LIST SPIN MERGE NEW NO JOIN A UP DN BCUP DN C UPIDNI N PLOTXMIN plotxmin r EV PLOTXMAX plotxmax r EV IN1MAXENERGY inimaxenergy r EV DO range i ALL NONE NEW OLD EXCEPT range i END DOSPLOT Print the density of states plot for general structures The DOS is generated using tetra for the selected atom or interstitial and angular momentum values The dosplot section expands the utility set of WIEN2k the user can join several DOS plots in one optionally scaled with multiplicity merge or split spin DOS etc Also as in printrho section the gnuplot is saved so the user can customize it runwien dosplot default Same as above but assuming defaults on all variables 3 11 2 Keywords and environments PLOTUNITS EV RY Units for the DOS representation electron volts or rydbergs Default ry ENERGYMIN emin r EV Lower energy bound for density of states mesh calculation
108. ormation xxxxxx lt root gt lt n gt wien deformation point input file lt root gt lt n gt 1 deformation point calculation files lt root gt sweep sweep section directory lt root gt sweep index sweep section indexfile lt root gt sweep lt n gt sweep point local dir lt root gt sweep lt n gt wien local wien lt root gt sweep lt n gt index local index file lt root gt lt n gt struct reference struct file lt root gt sweep lt n gt 1 local structure lt root gt sweep lt n gt 1 local wien s files lt root gt gibbs gibbs section directory lt root gt ingibbs gibbs input file root outputgibbs gibbs output file lt root gt check checkfile directory check checkfiles user files cif struct etc Postscript files containing the results of the properties sections are stored in their respective structure directories This also happens for gnuplot scripts which generate them and general output files for WIEN2k programs 5 Test files Some twenty full WIEN2k cases are included along with runwien Mostly they consist in low cost simple systems adequate for testing purposes and for proving the utility of runwien as an interface The index file test index contains a summary of the type of systems and a hint of what the calculation is about I lively recommend you to run these tests before trying any real c
109. out the sweep calculation Default no SPINORBIT environment The syntax of this environment corresponds to runwien SPINORBIT item_list i END SPINORBIT 86 Add a spinorbit section in the wien input file to some structures item list is a list of structures where this section will be included It follows the same syntax as with keyword or rmt etc in general section using n m l etc If no item list is included all the structures will receive a copy of this code The lines included in the body of the environment are passed verbatim to the corresponding wien input files The information in this environment and the following is not parsed runwien will not be able to tell if it is right or wrong until it is calculating the sweep structures ELASTIC environment The syntax of this environment corresponds to runwien ELASTIC item_list i END ELASTIC Add a elastic section in wien input file to some structures lt item_list gt is a list of structures where this section will be included It follows the same syntax as with keyword or rmt etc in general section using n m l etc If no lt item_list gt is included all the structures will receive a copy of this code The lines included in the body of the environment are passed verbatim to the corresponding wien input files The information in this environment and the following is not parsed runwien will not be able to tell if it is
110. p sweep temp sweep3 runwien out temp sweep temp sweep4 runwien out temp sweep temp sweep5 runwien out temp sweep temp sweep6 runwien out temp sweep temp sweep7 runwien out sweepsummary out temp sweep temp sweep1 synopsis out temp sweep temp sweep2 synopsis out temp sweep temp sweep3 synopsis out temp sweep temp sweep4 synopsis out temp sweep temp sweep5 synopsis out temp sweep temp sweep6 synopsis out temp sweep temp sweep7 synopsis out synopsis out index files temp index temp sweep temp sweep index ps files gnuplot files 2 6 Calculation of properties The runwien programming paradigm is based on clearness and modularity Once the SCF process has converged the calculation of many crystal properties can be achieved by running a collection of independent property sections dosplot for the Density of States bandplot for the Band of Structure analysis critic for the electron density topology and so on The modularity of runwien makes it easy to add or remove sections from the wien input file We will show some of the post scf possibilities that runwien offers using the popular WIEN2k user s guide example TiC A commented lt TiC wien gt input file can be as follows runwien general title keyword sets the global title Try to avoid exotic characters here like or title TiC calculation chapter 3 in wien guide lattice is fcc lattice F cell parameters 8 1787
111. plot default bandplot default kdos kdos 1 end kdos synopsis default The previous calculation is loaded with loadcheck and only the first structure is calculated with the do 1 keyword Next the density of states spaghetti and kdos plots are requested all with the default options The calculation produces the neat plots ascii nacl01 nacl01 dosband 1 eps nacl01 nacl01 dosplot 1 ps nacl01 nacl01 spaghetti ps Once the best calculation conditions have been selected we advance to the determination of the energy against the cell volume We select a reasonable set of volumes around the equilibrium volume and make runwien awk calculate the chosen points by rewriting the wien input file runwien loadcheck sweep rkmax 7 5 kpts 2000 rmt 1 2 0 rmt 2 2 0 with v 900 1500 100 dosplot end dosplot bandplot end bandplot kdos kdos 1 end kdos critic iws 2 newton le 15 end critic end sweep synopsis default This input file commands runwien awk to vary the cell volume from 900 to 1500 in steps of 100 atomic units Note that the information about the system is loaded in with loadcheck so it is not necessary to run general etc again In addition to the topological analysis critic the density of states dosplot band structure bandplot and merged DOS and bandstructure kdos plots are issued The WIEN2k parameters have values that correspond exactly to the first structure calculated in the parameter ex ploration Therefore you c
112. procal vector for the fourier expansion of the electron density kpts Number of k points in the whole first brillouin zone Imax Maximum angular moment for the spherical harmonics expansion in the basis func tions Insmax Maximum angular moment for the spherical harmonics used to calculate matrix elements with non muffin tin functions mix Mixing factor of the previous densities used on the self consistent calculation npt Number of radial mesh points It must be an odd number 381 is the recommended value for LDA and 781 for GGA calculations r0 First radial mesh point a logarithmic scale is used that starts in 0 0001 bohr by default rkmax Product rmt times kmax where kmax is the maximum k used for the plane waves set rmt Muffin tin radius in bohr Each non equivalent atom in the crystal can have a different value although difficult convergence problems can arise if the crystal includes atoms with very different rmt s The above list represent some of the most influential parameters on the final LAPW results The exploration of the convergence of the total energy with respect to rkmax for instance is a quite common task when we start working on a new crystal Similarly the rmt radius needs to be adapted when high pressure phases are involved or we are studying the topological properties of the electron density The exploration can involve two or more parameters at once For instance runwien rmt 1 1 8 2 0 0 5 kpts
113. r O you can write runwien cell 20 0 atom Ti cell 21 0 atom O cell 22 0 Default REFERENCE STRUCT n i Use the lt n gt th general structure to fill up the variables which were not set in the input file Default the first general structure which was run through the general section CELL a r Cell parameter for the fcc conventional cell used in the free energy calculation Keep this value high but not too much Values between 20 0 and 25 0 are just right See WIEN2k mailing list for reference Default 25 0 SPINPOLARIZED YES NO Perform a spinpolarized or non spinpolarized calculation for the se lected atom Default reference structure 61 NPT r r Number of radial mesh points Must be an odd number Wien2k manual recommends 381 for lda calculations 781 for gga Please note that the argument for npt is not a list of values i e npt is fixed throughout all the sweep run auto can be used and it will be passed into local wien files Default reference structure RMT r r Muffin tin radius in bohr Note that this rmt does not require the non equivalent atom the system only has one atom Default reference structure RO r r First radial mesh point in bohr Wien2k manual recommends a value between 0 0005 and 0 00005 Default reference structure RKMAX rkmax r rmt Kmax where Kmax is the maximum k for plane waves and rmt is the smallest muffin tin radius Default reference structure LMA
114. r used in the creation of the new charge density Default reference structure KPTS k r Number of k points in the whole first brillouin zone Default reference structure LDA U environment Formed as LDA U SIC AMF HMF iat1 i 11 i l1 s wer jer iat2 i 12 i 12 s u r jer END LDA U Includes an orbital dependent correction to the exchange correlation potential This method allows LDA or GGA calculations to successfully reproduce the properties of the so called strongly correlated systems Common systems of this kind are transition metal oxides and rare earth elements and compounds For more information see the WIEN2k user s guide and references therein The first keyword in the environment selects the flavor of LDA U e SIC Self Interaction Correction Anisimov et al 1993 e AMF Around the Mean Field Czyzyk et al 1994 e HMF Hubbard model in Mean Field Anisimov et al 1991 Each line in the LDA U environment activates the U correction for a particular atom and atomic symmetry 1 The first field corresponds to the non equivalent atom index in the unit cell The second field selects the angular momentum quantum number 1 either by its value 0 1 2 or by the corresponding letter s p d 67 The U Coulomb on site parameter and J exchange parameter characterize the LDA U cor rection These parameters are set using the u r and j r values respectively Default reference structure XCPOTENTI
115. re run The keyword do none is indicated both in scf and sweep so that the directories and input files are generated but no actual SCF calculation is done The user is expected to run the sweep calculations manual In a second pass the information is read into runwien using LOADCHECK REREAD and printed with synopsis Estimated time 11s 1h 26m 20s e test021 TiC the input file generating the example in the article The TiC rocksalt structure is calculated Using the converged density DOS and spaghetti plots are generated Using these two the kdos section is used to create a joint DOS band plot A topological analysis is performed including a representation of the laplacian of the electron density in the 001 plane The elastic constants are calculated using the elastic section A sweep gibbs run is done and the isolated atoms energy is calculated using free Estimated time 9h 32m e test022 ZnO an example of a non centrosymmetric structure Plots of the total electron density DOS band structure KDOS RX are generated Also topological analysis is carried out with critic The Morse sum condition on the critical points is not fulfilled pointing to the need of increased computational parameters larger LM list for example Estimated time 50m 39s e test023 Cr bcc antiferromagnetic Cr This is done using a spinpolarized calculation in bcc where atoms in 0 0 0 and in 1 2 1 2 1 2 are forcibly made non equivalent Estimated ti
116. runwien feature you can select which structures are going to be calculated The do none order forbids the calculation of the structures defined in the general section a single one in this case The structures defined in the next sweep part will be the do ones being calculated max iterations 50 energy conv 0 0001 end scf sweep with v 80 85 90 95 100 105 110 keep c a sqrt 8 3 as in an ideal hcp structure with c a 1 632993 print energy vs v end sweep synopsis default The with instruction in the sweep section indicates the geometrical variables that are going to be explored The cell lengths and angles can be explored in this way but as the example shows the user can ask for a variation of the cell volume the ratio between two cell lengths the internal position of an atom etc On the other hand runwien avoids those changes that are not compatible with the crystal symmetry For instance an instruction like runwien with b 4 0 5 0 0 2 will be simply ignored for an hexagonal crystal The sweep exploration can involve changes in two or more independent variables to produce a multidi mensional energy surface runwienti with v 80 100 5 with c a 1 60 1 80 0 05 Two or more independent scans can be produced in sequence runwien with a 4 20 4 40 0 05 with c 6 70 6 80 0 05 also with a 4 10 4 50 0 05 with c 6 75 also with a 4 30 with c 6 60 6 90 0 05 A significant limitation of the current sweeping
117. ry emax ry commands used initialization initialization scf total e for atom conf for atom 4 20 000 0 40000 7500 ggapbe96 0 Be 1414 Be 18 2 28 2 Be 0 00000e 00 29 2657470 29 2657470 no yes 781 1 88100 0 00010000 7 50000 10 10 000 13 00000000 2 00000000 i 50 ec 0 0001 Only one structure no variable parameters No sc data available No critic data available Structural grid sweep results Fixed parameters alpha deg 90 beta deg 90 gamma deg 120 c a 1 63299 c b 1 63299 b a aod Variable parameters number of structures 7 a bohr b bohr c bohr 3 83877 3 83877 6 26868 3 91713 3 91713 6 39665 3 99248 3 99248 6 51969 4 06508 4 06508 6 63825 4 13519 4 13519 6 75273 4 20299 4 20299 6 86345 4 26867 4 26867 6 97071 Output index plot and gnuplot out files scfsummary out v bohr73 it warning energy ry 40 0000 5 29 5183990 42 5000 5 29 5266380 45 0000 5 29 5323950 47 5000 5 29 5361680 50 0000 5 29 5383420 52 5000 5 29 5392095 55 0000 5 29 5390180 files 13 ooooooo efermi ry 724170000 680700000 641240000 605240000 572460000 542130000 514280000 time s pl 153 217 225 230 227 226 228 O O O O OO OS SO O temp sweep temp sweep1 runwien out temp sweep temp sweep2 runwien out tem
118. s described for the elastic environment apply here AIM environment The syntax of this environment corresponds to runwien AIM item list i END AIM Add an aim section in the wien input file on some of the sweep structures The same considerations described for the elastic environment apply here CRITIC environment The syntax of this environment corresponds to runwien CRITIC item list i END CRITIC Add a critic section in the wien input file on some of the sweep structures The same considerations described for the elastic environment apply here MINI commandline s Perform an internal parameter optimization using the mini program through the min lapw script The mini line is passed down to all the wien input files in the sweep calculation This is a temporary solution and will be modified in later versions Default no internal optimization CLEAN WIEN FULL Append a CLEAN WIEN FULL order at the end of each sweep point wien This will make each point in sweep section clean itself after being calculated For details on clean keyword see description in global section Sweep calculations can get very big Apply cleaning if you plan to calculate many structures of a system with a lot of valence electrons DO range i ALL NONE NEW OLD Used together with the EXCEPT order below EXCEPT range i As a default sweep will be run only on the new points skipping those points already examined The DO EXCEPT orders
119. s ev plotxmin 12 5 ev plotxmax 3 0 ev DOS list as set in the qtl file see WIEN2k manual dos list O 1 tot 1 4 Tid 15 23Cp end dos list end dosplot rxplot X Ray emission or absorption spectra Extend the eigenvalue window to 2 5 Ry use lapwi As a general rule each time this value is set in lapwi to calculate a property it is reset at the end of the section to its original value inimaxenergy 2 5 Calculating absorption spectra for Ti n 2 1 1 atom 1 n 2 11 x limits and dx for the plot plotxmin 2 0 de 0 02 plotxmax 15 0 end rxplot bandplot default Plot the band structure with all the default values critic Topological analysis of the electron density Get the critical points using Newton s search and 3 level recursive division of the irreducible wedge of the wigner seitz cell IWS newton ie 15 iws 3 end critic sweep Calculate 5 points from 10 to 10 of the volume of the reference structure which in this case is the first one indicated in the general section with v 10 10 5 print energy vs v end sweep synopsis exhaustive end synopsis 2 7 A complex case ferrocene Let us examine now a complex case ferrocene The structure can be found in one of the WIEN2K distribution example struct files Runwien offers the possibility of loading a structure from a struct 16 file as well as from a cif file through the use of the cif2struct utility To do that w
120. s information is used to calculate the elastic constants and evaluate the Born stability conditions The energy information and exact procedure and formulas used are written to the file lt root gt elastic elasticsummary out The LOADCHECK REREAD option is useful in this context Using this keyword the elastic calcu lations can be manually and the results read in into runwien awk The elastic summary fit scripts graphs and elastic constants are regenerated There is however one exception to this every file with the name of a target fit script fit xxxxxx gnuplot where xxxxxx is the deformation identifier will be recognized and used for the fitting instead of being overwritten The results of the elastic section are printed in a synopsis run The scripts and data employed to generate the elastic constants and some guidance about the process are stored in lt root gt elastic The DO EXCEPT mechanism is not implemented in elastic section For the elastic section to proceed to the calculation of the elastic constants all the energies for all the deformations must be known The easiest way of calling this section is accepting all default conventions and parameters 65 runwienti elastic default Note that an accurate determination of the elastic constants requires the inner deformation of atomic positions to be taken into account This inner strain can be calculated by using the mini optimization program through the mini keywor
121. s sweep be pes sweep4 be pes sweep be pes sweep5 be pes sweep be pes sweep6 be pes sweep be pes sweep7 be pes sweep be pes sweep8 The lt be pes check gt directory will contain the check status reports essential to the stop and continue mechanism The lt be pes1 gt directory contains the input files to perform the single WIEN2k calculation required by the GENERAL section Finally the eight subdirectories in lt be pes sweep gt correspond to the eight cell volumes asked for in the SWEEP section Each subdirectory contains a wien file that 11 can be run by hand to produce the calculation of the corresponding point On a Beowulf cluster for instance mpiexec or a similar job launcher can be used to run each sweep point on a different core thus achieving a parallel execution When all the 8 points have been calculated the information is taken back into runwien with a new lt be pes wien gt runwien loadcheck reread synopsis default LOADCHECK reads all the information stored in the checkfiles structure etc and reread instructs runwien to check if actual information is different from checkfile s information This is the case for we want the energies and properties of sweep points loaded Having read all this information runwien pretty prints a synopsis containing all the data you have manually calculated The lt synopsis out gt of the run is tasciit General information title runwien awk lapw
122. sity coulomb potential exchange correlation potential and total potential in an arbitrary plane This section extends WIEN2k s default utilities by providing a way to print planes defined by non orthogonal axis Additionally the scripts employed are saved as gnuplot scripts so the user can modify them to fit his her particular needs The gnuplot script is directly applicable to the data contained in the structure subdirectory The resulting ps files are placed in the structure s subdirectory runwien printrho default Same as above but assuming defaults on all variables 3 10 2 Keywords and environments RHO TOTALRHO VALRHO SPIN ATOMIC DEFORM VTOTAL VCOUL VXC Type of potential density plot Totalrho means total valence core density Valrho valence density Spin spin density only allowed in spinpolarized calculations Atomic density as superposition of atomic densities calculated in initialization Deform deformation density Vtotal total potential Vcoul coulomb potential Vxc exchange correlation potential 70 Default total density plot ENERGYMIN emin r Lower energy cutoff for the processing of electron density in Ry Choose ecoreval see above in order to calculate only valence densities Default total density energymin value in in2 file ORIGIN x r y r z r den r Origin of the plane x y and z must be integers den is the denominator The final point will be x den y den z den Crystallographic units
123. started counting multiplicity Default 1 USES TWO THREE ALL FOUR Start the critical point search between pairs of atoms all three atoms all two and three atoms or all four atoms respectively Default four NSHELLS nx i ny i nz i Number of unit cell shells which will be used to compute electron density in each direction Default 3 3 3 DO range i ALL NONE NEW OLD Used together with the EXCEPT order below EXCEPT range i As a default aim will be run only on the new points skipping those points already examined The DO EXCEPT orders can be used to modify this behavior Default Use the local DO EXCEPT specification if available Use the global DO EXCEPT data otherwise In absence of user orders do the aim task only on new points 3 15 3 Requirements of the aim section General initialization prescf and scf sections are required 77 3 16 Critic section 3 16 1 Syntax runwien CRITIC DEFAULT LINE ATOM NNEQ mult i x r y r z r ATOM NNEQ mult i X r y r z r npts i NICE nice i INTEGRALS nr i ntheta i nphi i BASINPLOT level i delta r theta r phi r IWS NOIWS i i NEWTON r r GRDVEC FILES rootname s PLANE x0 r yO r zO r xi r yi r zi r x2 r y2 r z2 r SCALE sx r Sy r PARAM radi r rad2 r rad3 r step r endpt r proj i ORIG x r y r z r atr i up i down i CP id i up i down i CPALL BCPALL up i down i RBCPALL bup i bdown i rup i rdown i CHECK X T y r z r ENDCHECK RHO DENSITY LAP LAPLACIA
124. stimated time 11m 57s test015 Cr hcp topological analysis of spinpolarized hcp Cr using CRITIC Only critical point position is calculated Estimated time 39m 5s test016 Be hcp example of usage of the sweep section in an hexagonal crystal A 3x3 grid in V c a is calculated then an independent grid in a with fixed c 3 points This amounts to a total of 12 points Estimated time 5m 48s test017 Be bec calculation of the thermodynamic quantities in bcc Be using the GIBBS program A total of 13 volumes are calculated and passed into GIBBS Estimated time 6m 34s 92 e test018 Be hcp an example of the usage of grids in the general section The evolution of the topology of the electron density is evaluated against the variation of rmt rkmax kpts and mix Estimated time 31m 41s e test 019 C diam a two step runwien execution In a first run the general initialization prescf and scf sections are run Then in a second pass the information is loaded into runwien awk exploration of rkmax kpts and rmt is done Copying the struct file generated by sgroup in the first pass is necessary to shift the origin of the unit cell to a point with inversion symmetry The sgroup struct file is not necessary in the second pass for the information is updated right after the copy Estimated time 2m 30s 30m 46s e test020 Be hcp another two step runwien job In the first pass general initialization prescf scf and sweep sections a
125. t maxenergy in initialization section DO range i ALL NONE NEW OLD Used together with the EXCEPT order below EXCEPT range i As a default dosplot will be run only on the new points skipping those points already examined The DO EXCEPT orders can be used to modify this behavior Default Use the local DO EXCEPT specification if available Use the global DO EXCEPT data otherwise In absence of user orders do the dosplot task only on new points 3 11 3 Requirements of the dosplot section General initialization prescf scf sections are required 73 3 12 Bandplot section 3 12 1 Syntax runwien BANDPLOT DEFAULT KLIST TEMPLATE tplfile s FILE lt file gt IN1MAXENERGY inimaxenergy r EV DO range i ALL NONE NEW OLD EXCEPT range i END BANDPLOT Print the band structure spaghetti plots Bandplot section uses WIEN2k utility spaghetti to create the plots runwien bandplot default Same as above but assuming defaults on all variables 3 12 2 Keywords and environments KLIST TEMPLATE tplfile s FILE lt file gt Enters the k list file to be used in the spaghetti plot If the keyword template is found then file must be one of the template files in WIEN2k root directory SWIENROOT SRC templates with or without the klist extension At the time we are writing this the available files are bcc klist fcc klist hcp klist and simple cubic klist If the user wishes to provide a different k list file
126. t we have included it within the GENERAL section to let the user explore different values different RKMAX need different directories and sets of initialization files even when their struct are the same The same reason applies to other parameters included here like LMAX LNSMAX and GMAX Default AUTO assume WIEN2k proposed value LMAX AUTO range i Maximum value of to be used on the spherical harmonics expansion inside the muffin tins A large value is needed to satisfy continuity of the electron density at the muffin tin surface but there are limitations hard coded in the WIEN2k source Default AUTO don t change WIEN2k proposed value LNSMAX AUTO range i Maximum value for spherical harmonics expansion used to com putate matrix elements with non muffin tin functions Default AUTO don t change WIEN2k proposed value GMAX AUTO range i Maximum reciprocal vector used on the fourier expansion of electron density WIEN2k manual recommends using a large GMAX 25 for systems with short H bonds GGA calculations also require larger GMAX than the L S DA calculations a value of 14 is typical in this case Default AUTO don t change WIEN2k proposed value MIX AUTO range r Mixing factor used in the creation of the new charge density Default AUTO don t change WIEN2k proposed value KPTS AUTO range i Number of k points in the first brillouin zone Default 1000 ALSO Us
127. the creation of reports on the calculation results Following WIEN2k usual practices the runwien input file should reside in its own directory and the directory name provides the root name for the large abundance of files created during the calculation In other words let us assume the calculation will reside in the directory tasciit alberto calc wien tests be Then the runwien input should be called be wien The calculation starts by invoking the runwien interpreter tasciit cd alberto calc wien tests be runwien awk be wien Runwien will start writing status information to stdout vou can redirect it to a file After a while the calculation will end and the file svnopsis out will contain a summarv of the run In the case of our example tasciit General information title runwien awk lapw calculation v 1 0 5 root be pwd home alberto calc wien tests be machine XXXX date jue nov 29 17 34 35 CET 2007 Sections run and time information general section run time s 0 initialization section run time s 1 prescf section run time s 0 scf section run time s 23 total run time s 24 Structural information POL2008 A Otero de la Roza M A Blanco A Mart n Pend s and Victor Lua a Critic a new program for the topological analysis of solid states electron densities to be published BFL2004 M A Blanco E Francisco and V Lua a GIBBS isothermal isobaric ther
128. the other but generally speaking auto keyword should be avoided in elastic section as all the points must have the same calculation level All this comments are valid for general section s varying parameters in sweep section rmt r0 Default reference structure 66 RMT atom i radius r Muffin tin radius in bohr for non equivalent atom with index atom i Default reference structure if no atom is indicated 1 is assumed RO atom i r r First radial mesh point in bohr for non equivalent atom with index atom i WIEN2k manual recommends a value between 0 0005 and 0 00005 Default reference structure RKMAX r r rmt Kmax where Kmax is the maximum k for plane waves and rmt is the smallest muffin tin radius Default reference structure LMAX l i Maximum l value for spherical harmonics expansion inside muffin tins Required in initial ization same as above Default reference structure LNSMAX Li Maximum value for spherical harmonics expansion used to computate matrix elements with non muffin tin functions Required in initialization same as above Default reference structure GMAX g r Maximum reciprocal vector for fourier expansion of electron density Wien2k manual rec ommends using a large gmax 25 for systems with short H bonds Also a larger gmax for gga calculations 14 than for lsda potentials Required in initialization same as above Default reference structure MIX lambda r Mixing facto
129. tions run automatically but it is also possible to do the deformation points manually and load the results afterwards All the elastic section work is done in the lt root gt elastic directory where root is the root of the calculation In this directory there is a subdirectory for each deformation 3 for cubic systems volume tetragonal and rhombohedral deformations and 5 for hexagonal systems volume a and b c alpha and beta and gamma deformations These deformations are annotated by a six digit code where a 0 represents the original cell parameter and a 1 a varying cell parameter Thus for example 111000 represents the volume deformation while 000001 means that gamma is the only cell parameter that is varied Each deformation directory contains several subdirectories each corresponding to one point in the deformation Runwien avoids having a different symmetry for the unstrained point by adding it up a tiny deformation so as to give it the same symmetry as the rest Note that the struct file generated by sgroup is used in the deformation points This is planned to change in future versions Once all the energies of all the deformed geometries are known runwien proceeds to calculate the elastic constants First the finite lagrangian strain tensor is calculated from the cell parameter deformations Then gnuplot scripts with name fit xxxxxx gnuplot are used to fit the data and extract the second derivative of the energy Finally thi
130. to a general runwien The wien file for the first step would be runwien general lattice H equiv list Be 0 66666667 0 33333333 0 75000000 0 33333333 0 66666667 0 25000000 end equiv list relativistic yes cell parameters 4 3210478 4 3210478 6 7714557 90 90 120 rmt 1 10 This time rmt is set to 10 less than the recommended value Otherwise muffin tin spheres would collide on compression 10 gmax 20 0 rkmax 7 5 kpts 7500 end general initialization ecoreval 10 0 orbitals 1 24243 0 0 0 054 0 002 CONT O 0 6 813 0 002 CONT O 1 0 261 0 002 CONT O end orbitals lm list 1 00 20 33 40 53 60 66 73 80 86 93 99 100106 end lm list end initialization prescf default scf do none max iterations 50 energy conv 0 0001 end scf sweep do none with v 50 60 70 80 90 100 110 120 with c a 1 632993 end sweep Let us assume that this file is be pes wien and it is located in a directory called be pes The calculation is performed by issuing the order ascii runwien awk be pes wien The calculation ends in a few seconds According to our input runwien creates a collection of subdi rectories and input files but the DO NONE orders in the SCF and SWEEP sections prevents any lengthy calculation to actually take place The subdirectories created are ascii be pes check be pesi be pes sweep be pes sweep be pes sweep1 be pes sweep be pes sweep2 be pes sweep be pes sweep3 be pe
131. tory The runwien script will be applied afterwards to all those new input files unless the user asks otherwise All the structures created by the sweep section share the same WIEN2k parameters which can include a list or range of values for some of them but vary in their cell constant values We will refer to this mechanism as sweep structures from now on Let us see an example From the calculations in the previous section we have determined the best WIEN2k parameters to use on the hcp Be crystal Some more tricks that our next input uses are a precise set of linearization energies and the best set of basis functions for the whole PES runwien general lattice H equiv list Be 0 66666667 0 33333333 0 75000000 0 33333333 0 66666667 0 25000000 end equiv list relativistic yes cell parameters 4 3210478 4 3210478 6 7714557 90 90 120 Reduce rmt to avoid muffin collision on compression rmt 1 10 gmax 20 0 rkmax 7 5 kpts 7500 end general initialization ecoreval 10 0 orbitals 1 24243 0 0 0 054 0 002 CONT O 0 6 813 0 002 CONT O 1 0 261 0 002 CONT O end orbitals LM list for spherical harmonics expansion of charge density inside the muffin tin sphere You probably will have to recompile WIEN2K with some augmented dimension parameter If this is your case you can simply remove this line lm list 1 00 20 33 40 53 60 66 73 80 86 93 99 100106 end lm list end initialization prescf default scf do none Another
132. trol for the section do done global atom etc are mechanisms designed for easy programming and friendly usage Try to maintain coherence by copying them if you see it fit In general think that a script this size is difficult to maintain unless a good amount of self discipline is put in it If you make some deep interesting modifications to our code we would be very grateful to hear from it 7 In case something goes wrong Runwien is provided as a mere tool to automatize tedious tasks If something goes wrong runwien usually saves the standard output and standard error of every program it runs in files with err extension For example the output of the nn program s name is lt root gt nn err Scf programs lapw0 lapw1 lapw2 and mixer also produce error files which are saved as lt root gt lt progname gt error The dayfile and the scf lt n gt files together with error files are the main sources you will have to check out if there is a problem in your calculations Runwien incorporates a warning system for the SCF cycle for example it detects the presence of ghost bands and preforms some checks on output files to see everything is running smoothly These checks are the coded result of some of the knowledge accumulated in the WIEN2k mailing list http www wien2k at reg_user inde but it is by no means complete If something goes awry but you feel you can repair it easily you can manually do it and reload the ca
133. tructure number increment the structure counter and copies the whole directory renaming its contents The checkfiles are automatically updated 3 3 General section 3 3 1 Syntax runwien GENERAL LOADCIF filecif s LOADSTRUCT filestruct s LATTICE lattice letter s EQUIV LIST environment L Y ZE END EQUIV LIST SPGLIST spg s RHOMB at s x r y r Z r END SPGLIST CELL PARAMETERS a r b r c r alpha r beta r gamma r TITLE runtitle s SPINPOLARIZED YES NO RELATIVISTIC YES NO NPT atom il range i AUTO RMT atom il range r AUTO RO atom il frange r AUTO RKMAX AUTO range r LMAX AUTO range i LNSMAX AUTO range i GMAX AUTO range i MIX AUTO range r KPTS AUTO range i ALSO LINK vari s var2 s var3 s DO range i ALL NONE NEW OLD EXCEPT range i LDA U SIC AMF HMF iati i 11 ill1 s defu r defj r iat2 i 12 i 12 s defu r defj r END LDA U U line i urange r J line i urange r END GENERAL 33 The general section is perhaps the most important one in runwien It reads the characteristic calcu lation variables structural and WIEN2k verifies needed parameters are set parses expressions in the wien input loads info from external files creates header in stdout and whole index file creates the iteration counter and the directory tree structure and starts the struct files that will be completed later
134. ts and RKmax The output of the nn program nacl1 nacl1 outputnn contains tasciit RMT 1 2 50000 AND RMT 2 2 50000 SUMS TO 5 00000 LT NN DIST 5 32956 So rmt must be significantlv smaller than 2 5 for both atoms as we want to explore the potential energv surface In our example a grid of 3x3 values of rmt for Na and Cl atoms was selected The block runwien rmt 1 2 0 2 25 2 5 rmt 2 2 0 2 25 2 5 kpts 2000 rkmax 7 5 calculates all the 3x3 9 possibilities found by combining the three values for the rmt of Na 2 0 2 25 and 2 5 with the corresponding three possibilities for the Cl atom also 2 0 2 25 and 2 5 The kpts and rkmax parameters are fixed at 2000 and 7 5 respectively for all the 9 structures The keyword runwienti also starts a new grid independent of the previous one The keywords runwien kpts 2000 10000 2000 rkmax 7 5 in this new block vary kpts from 2000 up to 10000 in steps of 2000 amounting to 5 points while rkmax is fixed to 7 5 Note that the also keyword causes this 1D parameter grid to be independent from the 2D rmt grid described above Here rmt for both atoms assumes its default value 2 5 Finally some values of rkmax are also studied using the input 22 runwien also kpts 2000 rkmax 7 5 10 0 0 5 Again the also keyword uncouples the rkmax variation from the rest of the parameter grids The initialization section comes next in the input file Our interest in this example is
135. ual cell geometry corresponds to your expectations possible core leakings and other warnings from the SCF process and the correct core valence assignment tasciit Core val el conf for atom C1 18 2 28 2 2P 6 3S 2 3P 5 Core val el conf for atom Na 18 2 28 2 2P 6 38S 1 In our case an energv was obtained but runwien awk indicates the presence of ghost bands tasciit Energy ry 1248 1317470000 E fermi ry 0 13756 Warnings 5 No critic data available 5 Ghost bands The scf output file contains more details on the problem ascii WARN QTL B value eq 5 08 in Band of energy 0 27651 ATOM 2 L 1 WARN You should change the E parameter in case ini or use ininew switch These data indicates that a further adjustment of the basis set is needed To this end remove the all default calculation ascii rm rf nacl index out Change the wien file in order to let WIEN2k calculate an appropriate basis set in the third cycle runwien general loadcif nacl cif end general 20 initialization default prescf default scf new ini 3 end scf synopsis default Execute it tasciit runwien awk nacl wien Note that letting WIEN2k generate a good basis set does not alwavs work and sometimes vou will be forced to do it by hand However it works fine in NaCl rendering a correct energy and no ghost band warnings tasciit Energy ry 1248 1369720000 E fermi ry 0 13097 No
136. ucture no variabl Energy ry 29 5397615000 E fermi ry 0 50969 e parameters No critic data available Rest of summary skipped As this output clearly reveals there is a large number of internal parameters that runwien knows about and that the user can keep as their defaults of can control at wish Occasionally the user may want to explore the influence of some of these parameters on the results or just optimize the best calculation conditions This is facilitated because runwien provides the possibility of entering a list and a range of values for some of the parameters For instance runwienti rmt 1 1 70 1 75 1 80 will perform three different calculations using the enlisted three different values for the muffin tin radius of Be non equivalent atom number 1 The same purpose could be achieved using a range trunwient rmt 1 1 70 1 80 0 05 where the three values of the range are the starting and ending values and the step 1 bv default respectivelv A list of ranges and values is also possible runwienti rmt 1 1 7 1 800 1 900 0 025 The actual value of the parameter can even be given as a increase or decrease of the value given by WIEN2k runwien rmt 1 10 15 where 10 means ten percent less than the default or recommended value and 15 fifteen percent more The internal parameters that can receive this treatment of using a list or range percent values only work for rmt include gmax Maximum reci
137. w input file runwien loadcheck sweep do none with v 950 1450 100 end sweep synopsis default and execute runwien awk on it Due to the do none keyword no actual calculation is performed but the directory and the wien input file for the new points 8 to 13 are generated The user is in charge of running these points manually either using a queue system or by any other means For example for point 8 ascii cd nacl sweep nacl sweep08 runwien awk nacl sweep08 wien When the points from 8 to 13 are calculated the data can be read back into runwien awk and the energy plotted against the atomic volume using the input file runwien loadcheck reread sweep print energy vs v end sweep synopsis default which compiles the information in all the structures either manually or automatically run tasciit Variable parameters number of structures 13 a bohr b bohr c bohr v bohr73 it warning energy ry efermi ry time s planarity morsesum topology 9 65489 9 65489 9 65489 225 0000 9 1248 1042110 0 098260000 128 0 4215360 24 2 4 4 0 111241 11 48 11 8 10 00000 10 00000 10 00000 250 0000 9 1248 1258250 0 036130000 141 0 1632730 0 2 4 4 0 12 24 241 11 48 11 8 10 32280 10 32280 10 32280 275 0000 9 1248 1370460 0 013030000 170 0 1575500 0 2 4 4 0 12 24 241 11 48 11 8 10 62659 10 62659 10 62659 300 0000 10 1248 1407190 0 052370000 191 0 1542220 0 2 4 4 0 12 24 24
138. w structures 3 5 3 Requirements of the prescf section General and initialization sections are required for prescf to run correctly 3 6 Scf section 3 6 1 Syntax runwien SCF DEFAULT MAX ITERATIONS maxit i CHARGE CONV cconv r ENERGY CONV econv r FORCE CONV fconv r NICE nice value i REUSE CHAIN FIXED n i FIRST ANY PATH path s ITDIAG cyc i NEW INI cyc i MINI commandline s NOSUMMARY DO range i ALL NONE NEW OLD EXCEPT range i END SCF Runs the scf cycle By default asummary will be generated in the root directory called scfsummary out containing some relevant information about the scf process Notice that the simple runwien scf default will invoke scf assuming default values on all variables 3 6 2 Keywords and environments MAX ITERATIONS maxit i Maximum number of cycles Default 30 CHARGE CONV cconv r Charge convergence criterion Default none ENERGY CONV econv r Energy convergence criterion Default 0 00001 unless one of cc ec or fe is indicated 56 FORCE CONV fconv r Force convergence criterion Note that if any of energy conv force conv or charge conv are set they constitute the scf stop criterion If none of them are set then the default value is energy conv 0 00001 Ry Default none NICE nice value i Process priority for run lapw Child processes can not have a lower nice than runwien itself Default 0 increase to decrease priority REUSE CHAIN FIXED n i FIR
139. ways resort to manual execution and rereading the results 31 DO EXCEPT set outside any section changes globally the default from that point on Contrarily a DO EXCEPT declared within a section range will only affect locally to that particular section As an exception the ELASTIC FREE SWEEP GIBBS and SYNOPSIS sections use particular DO EXCEPT mechanisms e ELASTIC ignores the DO EXCEPT options but the deformations required to calculate some elastic constants can be skipped with the NOSEND order e The FREE section is only affected by the local DO EXCEPT declaration with a particular syntax allowing element symbols See the FREE section keywords e The SWEEP section uses only the local DO EXCEPT declarations and ignores the global values e The DO EXCEPT list has no meaning within the GIBBS and SYNOPSIS tasks The DO EXCEPT orders can be qualified with range i A range of integers indicating the cases to include or exclude The range can be any combination of a list of integers e g 1 3 9 a first last pair e g 2 10 a first last step triplet e g 2 10 2 ALL Include all new and old cases NONE Exclude all cases NEW Include only the new cases This is the global default OLD Include just the old cases In the next example runwien general Imax 1 20 This involves 20 points or structs end general initialization default Change default behavior for next sections do 4 10 except 9 Run only poi
140. y if the struct files contain the same information the clmsum file is simply copied and effectively replaces the superposition density as starting point for the SCF cycle In older versions of runwien lt 14 12 2007 the extrapolation of the charge density is substituted by a mere renormalization It is interesting to note that REUSE may lead to some drastic shortening of the computational cost of your calculations specially if the old densities correspond correspond to a good set of parameters Therefore routine usage of this keyword should be considered Default no reuse ITDIAG cyc i Use iterative diagonalization and do a complete diagonalization each cyc i cycles In versions of WIEN2k gt 13 08 2007 cyc i has no effect Analogously to REUSE ITDIAG may impose dramatic cuts in the computational cost of the calculation Default no it diag cyc i 3 if applicable NEW INI cyc i Generate a new basis set in cycle lt cyc gt Default no MINI commandline s Perform an internal parameter optimization using the mini program through the min_lapw script The commandline s optional field is appended to min lapw in the calling command min lapw lt commandline s gt If no commandline s is found j run _lapw scfemd line s is used where scfemdline s are the arguments for the run _lapw script This keyword is provided as a temporary solution It will be improved in later versions Default no internal optimization

1 Runwien user`s and reference guide Contents

Contents

Download Pdf Manuals

Related Search

Related Contents