Practical: Refinement of powder diffractogram and extraction of single crystal intesities for ZnCr2Se4. Paweł Zajdel

Transcription

1 Practical: Refinement of powder diffractogram and extraction of single crystal intesities for ZnCr2Se4 Paweł Zajdel NCNR and University of Silesia, Katowice, Poland /43

2 Goals: Obtain possible magnetic structure(s), which are in agreement with the symmetry imposed by the propagation vector group and fit the powder pattern. Predict intensities for a single crystal experiment. Methods: Representational analysis for list of possible solutions (SARAh, MODY, pencil). Your favourite refinement package (Fullprof Suite). It is good to know in advance: The nuclear structure? XRD Any constrains on the moments? Magnetization, Mössbauer,... Order of the transition? Latent heat? Some crystallography... we are solving a structure after all 2/43

3 Normal spinel ATBO2X4 Cubic Fd-3m, (a ~.5 Ǻ ), ZnCr2Se Å nonstoichiometric? B-site trigonal antiprism "octahedral" (u=.25). A-site always ideal tetrahedron. 2 settings: setting Fd-3mS and setting 2 (standard) Fd-3mZ 3/43

4 What we know about ZnCr2Se4 μeff = 5.5μB, μsat = 5.7μB, ΘCW~K Cr3+ (S=3/2) μeff = 5.46μB, μsat = 6.μB, R. Plumier, C. R. Acad. Sc. Paris, t. 26 (965) 3348; J. Phys. (Paris) 27, 23 (966) R. Kleinberger, R. de Kouchkovsky Comp. Rend. Acad. Sci 262, 628 (966) K. Dwight, N. Menyuk, Phys Rev. B, 63(2),435 (967) K. Siratori, J. Phys. Soc. Jap. 3(3), 79 (97) F. Yokaichiya, A. Krimmel, V. Tsurkan, I. Margiolaki, P. Thompson, H. N. Bordallo, A. Buchsteiner, N. Stüßer, D. N. Argyriou, A. Loidl Phys. Rev. B 79, (29) TN= 2K, NTE< K I4/amd, c/a=.9995 Ordered moment.7μb! 4/43

5 Distortion and NTE - 966! R. Kleinberger, R. de Kouchkovsky Comp. Rend. Acad. Sci 262, 628 (966) 5/43

6 Fullprof Suite survival kit: Output files (main) Data Single or many patterns Fullprof wfp2k, Fp2k *.prf - Winplotr *.out - Text PCR - instructions *.hkl - Text *.fou - Text EdPCR (cif, ins, self) *.fst - FPStudio *.cif - Text 6/43

7 Getting started preparing nuclear structure (from scratch) New PCR Import CIF Import SHELXL *.ins Use 7/43

8 Getting started nuclear structure from scratch Important things to type-in (General, Pattern, Phases):. One experimental pattern (Znc.dat). 2. Wavelength 2.782Å. 3. Background type start with 6 order poly (we shall convert it later later) 3. CW-neutron powder diffraction 4. One phase: a) Nuclear only, b) Rietveld, c) Gaussian profile. 8/43

9 Pattern 9/43

10 GSAS, XY or whatever you have Pattern/ Data file This is needed only If you click "Calculation" /43

11 Pattern: Geometry Works better for me Absorption - may help if your B< Always asymmetry. /43

12 Phase/s Title for your phase. If not active add one first. Rietveld nuclear Add phase 2/43

13 Phase/s: contribution This section connects Patterns with Phases. We connect current Phase () with previously entered Pattern(). Pattern () was CW-neutron. Make sure the correct one is selected! BT- and our sample. Ask your instrument scientist! Check after cycle if FP accepted it. (May go wrong) Warning! This will automatically change to "HKL file" during LeBail fits. Reset to "Automatic" if you have problems. 3/43

14 Phase: Nuclear Symmetry Should work for most standard symbols Type-in your SG And hit Symm Op. This should fill operators. Check with ITC, when in doubt. Need this number for SOF, 4/43

15 Getting started nuclear structure from scratch Lattice parameter/s (.48) and atomic positions. (Roughly refine later). Atoms Zn.25,.25,.25 (8a) chemical SOF =, we type.467 Cr (6d) chemical SOF=, we type Se.26,.26,.26 (32e) chemical SOF=, we type Chemical SOF = like in GSAS, FOX, RIETAN FP, Shelxl crystallographic SOF CrystSOF = Chem SOF * Site_Multi / General_Multi Zn_SOF = * 8 / 92 = FP will round it Make sure it's correct otherwise your moment will not scale properly. 5/ 43

16 Getting started nuclear structure from scratch Atoms Lattice and profile 6/43

17 Atoms Warning! Make sure you choose an atom with chemical SOF= as Atom #. Otherwise FP will "correct" you and you will get wrong scaling factor. DIY 7/43

18 Profile Your job Default but you can write your own if known. Save yours or use ZnCr2Se4_start.pcr 8/43

19 Refinement... your job Try to choose initial scale by hand so you can see peaks! Than refine a constant background. Than is your call. ZnCr2Se4_part.pcr refined offset, profile, lattice and zero shift, scale. 9/43

20 Indexing was yesterday Try your luck thats an example, which doesn't look good. Kx Ky Kz R-factor In fact k is close to (,,.46) Or if you use only first you may get k2 = (.24,.24, ) Check everything! 2/43

21 Insert second phase and do pattern matching but first lets look into PCR file COMM Title! Files => DAT-file:, PCR-file: X:\Dokumenty\RAMS_2\start!Job Npr Nph Nba Nex Nsc Nor Dum Iwg Ilo Ias Res Ste Nre Cry Uni Cor Opt Aut 7!!Ipr Ppl Ioc Mat Pcr Ls Ls2 Ls3 NLI Prf Ins Rpa Sym Hkl Fou Sho Ana 4 2!! lambda Lambda2 Ratio Bkpos Wdt Cthm mur AsyLim Rpolarz ->Patt# !!NCY Eps R_at R_an R_pr R_gl Thmin Step Thmax PSD Sent !! 7!Number of refined parameters!! Zero Code SyCos Code SySin Code Lambda Code MORE ->Patt# ! Background coefficients/codes for Pattern# ! ! Data for PHASE number: ==> Current R_Bragg for Pattern# :.! spinel!!nat Dis Ang Pr Pr2 Pr3 Jbt Irf Isy Str Furth ATZ Nvk Npr More ! F d -3 m <--Space group symbol!atom Typ X Y Z Biso Occ In Fin N_t Spc /Codes Zn Zn Cr Cr Se Se ! > Profile Parameters for Pattern #! Scale Shape Bov Str Str2 Str3 Strain-Model ! U V W X Y GauSiz LorSiz Size-Model ! a b c alpha beta gamma #Cell Info ! Pref Pref2 Asy Asy2 Asy3 Asy4 S_L D_L ! 2Th/TOF 2Th2/TOF2 Pattern # There are several blocks that you need to know. General info number of phases, type of background, number of exluded regions, input and output options. Not a good idea to change them by hand except Nph number of phases, Nex excluded regions, Nba type of background. Pattern properties: Wavelength, zero shift, background but also number of parameters and cycles. Phase properties Plotting range leave is as last line I am going to change Nph to 2 and copy the lines in light purple. 2/43

22 Second phase and do pattern matching If everything went fine you should have pcr file with 2 identical phases. Also variables like lattice and shape are constrained to be the same. Now we are going to: ) change the name of the phase, Phases select Phase 2, 2) change dropdown list "Calculation" type to "Pattern matching with constant scale factor," 3) remove constrains by unlclicking and clicking again. 4) add k-vector and -k since they are not equivalent. 5) and try to refine. It will do Rietveld on Phase and LeBail on satellites. Pcr file after this stage ZnCr2Se4_Phase2.pcr. 22/43

23 Second phase and do pattern matching At this point it should be evident that the background is not correctly modelled and the first peak ()+q has nasty asymmetry. Phases Background type It inputs default lines in the pcr file. Check them under Refinements>Background. 23/43

24 Second phase and do pattern matching It inputs default lines in the pcr file. Check them under Refinements>Background. Less is better but use as much as is necessary ZnCr2Se4_Phase3.pcr 24/43

25 Representational analysis At this point we have propagation vector and good atomic positions. What may go wrong? Use Basireps first. Output: => Number of Space group: 227 => Hermann-Mauguin Symbol: F d -3 m => Hall Symbol: -F 4vw 2vw 3 => Table Setting Choice: 2 Setting <cut> => Number of elements of G_k: 8 => Number of irreducible representations of G_k: 5 => Dimensions: <cut> => The input propagation vector is: K=(...46 ) => K.. IS NOT.. equivalent to -K <cut> -> GAMMA(Magnetic): 3 Irep_k() + Irep_k( 2) + 2 Irep_k(3) + 3 Irep_k(4) + 3 Irep_k(5) Wrong! OK 25/43

26 Sarah/KAREP Sarah227_summary.txt ORDERS OF THE REPRESENTATIONS: : 2 2: 3: 4: 2 5: 2 DECOMPOSITION OF THE MAGNETIC REPRESENTATION INTO IRs OF Gk: 3 TIMES THE REP OF ORDER 2 TIMES THE REP 2 OF ORDER 2 TIMES THE REP 3 OF ORDER 3 TIMES THE REP 4 OF ORDER 2 3 TIMES THE REP 5 OF ORDER 2 Identical to Basireps and Wrong! 26/43

27 Sarah/Kovalev Used to be correct ORDERS OF THE REPRESENTATIONS: : 2: 3: 4: 5: 2 DECOMPOSITION OF THE MAGNETIC REPRESENTATION INTO IRs OF Gk: TIMES THE REP 2 TIMES THE REP 2 TIMES THE REP TIMES THE REP 3 TIMES THE REP OF ORDER 2 OF ORDER 3 OF ORDER 4 OF ORDER 5 OF ORDER 2 but since yesterday got overflow? Correct! 27/43

28 MODY/Kovalev But use setting! Chromium position.625,.625,.625 Number of representations: 5 REPRESENTATION tau, dim, REPRESENTATION tau 2, dim, REPRESENTATION tau 3, dim, REPRESENTATION tau 4, dim, REPRESENTATION tau 5, dim 2, occurring time **** occurring 2 times **** occurring 2 times **** occurring time **** occurring 3 times **** Correct! = 2 = 4*3 OK 28/43

29 Let us use SARAh/Kovalev Look into the end of *.lst2 IR #, BASIS VECTOR: # (ABSOLUTE NUMBER:# ) : ( 2-2 ) + i( ) 2: ( -2 2 ) + i( ) 3: ( ) + i( ) 4: ( ) + i( ) ****** IR # 2, BASIS VECTOR: # (ABSOLUTE NUMBER:# 2) : ( 2 2 ) + i( ) 2: ( -2-2 ) + i( ) 3: ( ) + i( ) 4: ( ) + i( ) IR # 2, BASIS VECTOR: # 2 (ABSOLUTE NUMBER:# 3) : ( 4) + i( ) 2: ( 4) + i( ) 3: ( 2.984) + i( ) 4: ( 2.984) + i( ) ****** IR # 3, BASIS VECTOR: # (ABSOLUTE NUMBER:# 4) : ( 2 2 ) + i( ) 2: ( -2-2 ) + i( ) 3: ( ) + i( ) 4: ( ) + i( ) IR # 3, BASIS VECTOR: # 2 (ABSOLUTE NUMBER:# 5) : ( 4) + i( ) 2: ( 4) + i( ) 3: ( ) + i( ) 4: ( ) + i( ) ****** IR # 4, BASIS VECTOR: # (ABSOLUTE NUMBER:# 6) : ( 2-2 ) + i( ) 2: ( -2 2 ) + i( ) 3: ( ) + i( ) 4: ( ) + i( ) IR # 5, BASIS VECTOR: # (ABSOLUTE NUMBER:# 7) : ( 2 ) + i( ) 2: ( 2 ) + i( ) 3: (.332 ) + i( ) 4: (.332 ) + i( ) IR # 5, BASIS VECTOR: # 2 (ABSOLUTE NUMBER:# 8) : ( 2 ) + i( ) 2: ( 2 ) + i( ) 3: (-.332 ) + i( ) 4: (-.332 ) + i( ) IR # 5, BASIS VECTOR: # 3 (ABSOLUTE NUMBER:# 9) : ( 2) + i( ) 2: ( -2) + i( ) 3: (.332) + i( ) 4: ( -.332) + i( ) IR # 5, BASIS VECTOR: # 4 (ABSOLUTE NUMBER:# ) : ( ) + i( ) 2: ( ) + i( ) 3: (.492 ) + i( ) 4: (.492 ) + i( ) IR # 5, BASIS VECTOR: # 5 (ABSOLUTE NUMBER:# ) : ( ) + i( ) 2: ( ) + i( ) 3: ( ) + i( ) 4: ( ) + i( ) IR # 5, BASIS VECTOR: # 6 (ABSOLUTE NUMBER:# 2) : ( ) + i( ) 2: ( ) + i( ) 3: ( -.492) + i( ) 4: (.492) + i( ) Can we make spiral with moments only in ab plane with one IRREP? YES 29/43

30 Let us use SARAh/Kovalev Select "promissing" basis functions From IRREP5 Press "Generate model PCR" for powder. It will create file Sarah.pcr in the folder with your 227.MAT file 3/43

31 Let us use SARAh/Kovalev! ! Data for PHASE number: ==> Current R_Bragg for Pattern# :.! Magnetic Phase!!Nat Dis Mom Pr Pr2 Pr3 Jbt Irf Isy Str Furth ATZ Nvk Npr More ! F - <--Space group symbol! Nsym Cen Laue Ireps N_Bas 4-4! Real()-Imaginary() indicator for Ci! SYMM X, Y, Z BASR 2 2 BASI SYMM -X+3/4, -Y+3/4, Z BASR 2 2 BASI SYMM Y+/4, -X+, Z+/4 BASR BASI SYMM -Y+, X+/4, Z+/4 BASR BASI!!Atom Typ Mag Vek X Y Z Biso Occ C C2 C3! C4 C5 C6 C7 C8 C9 MagPh CR MCR ! > Profile Parameters for Pattern # We need only part of it between the begining and profile parameters. I am going to replace this fragment in the second phase of our pcr file. 3/43

32 We now have pcr file with 2 phases When you check - the second phase is now Rietveld (magnetic) with symmetry given by basis functions. SARAh can do it for you but you have to know what is going on. There are 3 corrections that need to be made.. Magnetic phase uses now CHEMICAL OCCUPANCIES!!! 2. There is unique chromium atom in magnetic cell and SARAh writes its SOF as 4. But in fact it is taken into account by the centering in SG symbol F -. Since we use F-centering change the occupancy to. 3. In the Refinement window find Convergence Criterion and change it back to None. 4. Constrain scale. 32/43

33 It will not refine since it is not yet a spiral.! Real()-Imaginary() indicator for Ci! SYMM X, Y, Z BASR 2 2 BASI SYMM -X+3/4, -Y+3/4, Z BASR 2 2 BASI SYMM Y+/4, -X+, Z+/4 BASR BASI SYMM -Y+, X+/4, Z+/4 BASR BASI BASI!!Atom Typ Mag Vek X Y Z Biso Occ C C2 C3! C4 C5 C6 C7 C8 C9 MagPh CR MCR /43

34 Let us normalize the functions (/2) to see what they are! Real()-Imaginary() indicator for Ci! SYMM X, Y, Z =cos BASR BASI = sin SYMM -X+3/4, -Y+3/4, Z α=4.759 BASR Sin α =.666 = sa BASI Cos α =.746 = ca SYMM Y+/4, -X+, Z+/4 BASR BASI SYMM -Y+, X+/4, Z+/4 BASR BASI BASI!!Atom Typ Mag Vek X Y Z Biso Occ C C2 C3! C4 C5 C6 C7 C8 C9 MagPh CR MCR /43

35 Let us normalize the functions (/2) to see what they are! Real()-Imaginary() indicator for Ci! SYMM X, Y, Z BASR BASI SYMM -X+3/4, -Y+3/4, Z BASR BASI SYMM Y+/4, -X+, Z+/4 BASR sa -sa ca BASI SYMM -Y+, X+/4, Z+/4 BASR sa -sa ca BASI BASI -ca -ca This is an example where: "Keeping it real goes wrong!" One needs complex coefficients to obtain a screw/helix. =cos = sin Sin (α+) =.666 = sa Cos (α+) =.745 = ca α=4.759deg This is the exact angle bewteen Cr planes or Or α=2pi *.46/4 If your precision is not good enough i.e. sin2α+cos2α <> Extra peaks!!!! 35/43

36 Now we just need coefficients to refine!atom Typ Mag Vek X Y Z Biso Occ C C2 C3! C4 C5 C6 C7 C8 C9 MagPh CR MCR We want a perfect helix with equal components, so C = C2 = C3 = C4, C = C2 and C3 = C4, C = -C3 Which gives C, C, -C, -C and file ZnCr2Se4_withIRandC.pcr Warning! If you refine propagation vector change basis functions. Make sure that sum of squares your values of sin and cos add to. 36/43

37 After several cycles! Real()-Imaginary() indicator for Ci! SYMM X, Y, Z BASR BASI SYMM -X+3/4, -Y+3/4, Z BASR BASI SYMM Y+/4, -X+, Z+/4 BASR BASI SYMM -Y+, X+/4, Z+/4 BASR BASI !!Atom Typ Mag Vek X Y Z Biso Occ C C2 C3! C4 C5 C6 C7 C8 C9 MagPh CR MCR /43

38 Now, how to get single crystal intensities? Tell FP to write them in different formats available from those 2 tabs. Write to *.out 38/43

39 Now, how to get single crystal intensities? Then select more specific information. We want reflection file *.hkl FP will create Pcrname.hkl Pcrnamex.hkl 39/43

40 Output file Pcrname.out depending on the choice either integrated or profile integrated. No. Code H K L IVK H3 d-hkl CORR Mult Hw 2theta/TOF Icalc Iobs Sigma HwG HwL ETA H H Pcrname.hkl does not list explicitly which k wector and gives multi * F 2.! SPGR: F -; h k l CELL: Mult sint/lamb 2T Fwhm F sf < -k < +k < -k < +k Pcrnamex.hkl does explicitly list k wector and gives multi * F2.! Phase No: 2 magph2 2. RF2 = Sum{ F2obs-F2cal }/Sum{F2obs} Number of propagation vectors: 2 Kx Ky Kz h k l n Mult F2(obs) F2(calc) Qx = Qy Qz d-spacing /43

41 Problems Try to find satellite intensities of 55 and 7. Due to perfect degeneracy od d-spacings (similar thing happens in tetragonal groups for 43 and 5) FP makes mistake. Pcrname.hkl Seems to be fine but the same F Pcrname2.hkl Only 7 satallites are listed with inttensity and multiplicity including 5 but separate 5 entry. 4/43

42 Remarks: Check SOF and ATZ! - Scaling of moments. Before removing a phase fix all variables. If a variable does not change its value during refinement and you are sure it should, check the "code" if it is not damped. "Sum Iobs=, do you really have reflections" check scale and lattice parameters, then profile, B and SOFs. - Bilbao Crystallographic server Use FPstudio to get moments! 42/ 43

43 Final remarks: Trust no one (software). Have fun! 43/ 43