Dear Andrew, First, thank you very much for taking the time for this really detailed and helpful answer. It is very much appreciated and I hope that I can give some useful feedback besides consuming your time.
Just to let you know, I run everything in vmware fusion on a mac (but it should be exactly the same as native Windows). When I installed SARAh a desktop icon was created that launched SARAh-refine but I didn't 'know anything' of SARAh-representational analysis.That's fine because a beginner would click on the video tutorial button immediately. This however doesn't do anything in my version (7.1.3). I guess it should open a browser and go to the some webpage. Anyway, that's just a minor detail and I figured out that going to ftp://ftp.bcc.ac.uk/pub/users/uccaawi/ has all the tutorials. They are really excellent and I could easily understand how all individual parts of the software are used (of course the underlying representational analysis much less so)! Since we suspect from earlier neutron powder diffraction that the magn structure is incomm I decided to work with Fullprof. So off I went and created the MAT file with Pnma as the underlying crystal structure, one Mn atom in a Wyckoff 4c position (for simplicity I neglected Co moments) and a 'guess' (or rather a value from literature) for k (0 0 tau). The creation of the files was smooth and I was able to load the MAT file in SARAh-refine. For a simple start I selected all the basis functions with the common IR gama(1) and created the pcr file. As the next step I copied the relevant information from sarah62.pcr into my.pcr . Please note that the 'phase header' is slightly different in my pcr because ultimately I want to refine multi-bank patterns from GEM. Therefore some parameters (such as Irf Npr Jtyp .... Pr1 Pr2 Pr3 ) are defined individually for each bank on a separate line. I don't know if that's crucial for describing my problem but I thought I'll mention it. The rest of the magnetic phase information was, however, copied from sarah62.pcr as a second phase into my.pcr. For a start I fixed all parameters except the C coefficients (set the value to 1.00 and the code to 1.00) as explained in the tutorial. Running Fullprof results in some pattern calculation BUT unfortunately the scale factor is off so much that the magnetic phase is a factor 100000 more intense than the nuclear phase. Refining the scale factor (either for the magn phase or the nuc phase) doesn't help. Neither does setting it to some fixed large value to compensate the scaling difference. I was wondering whether this is a known problem or whether you'd have any guesses as to what I might be doing wrong. If you want to see any files please let me know. I'm more than happy to send them. The possibility to solve a magnetic structure from basic principles such as group theory and RA seems very appealing and unbiased to me so I hope that there is a solution to problems arising on the way. Once again I'm very grateful for your help and time. Alex 2009/7/23 Andrew Wills <a.s.wi...@ucl.ac.uk> > Dear Alex > The MAT file is generated from the symmetry analysis calculations using > SARAh-Representational Analysis (there are video help files to show you how > to do this). > > If you are setting up a refinement in GSAS, you can use SARAh Refine to > set up the magnetic phase for you. There are 2 options: > > 1) you set up the phase, make it magnetic and put in an atom with a moment > on it to make sure it works. Select <GSAS Controls>/Insert magnetic atoms , > to overwrite this atom with all the atoms for your P1 magnetic structure. > (this is required so that SARAh can match the atom to its moment > orientation > > 2) Use SARAh Refine to set up the magnetic phase for you. > -To do this, delete all the atoms in the P1 phase, then delete the phase > itself. Run Powpref to make sure it works. > -Then Select <GSAS Controls>/Make magnetic phase (Fully automatic) > -Define the nuclear phase that will be used as a template for the magnetic > phase's lattice parameters, profile, etc > -Define the form factors of the different moments > > -The phase will now be set up for you. (If there are problems with this, > let me know and I'll try to fix them) > > Refining the structure is then based on selecting the basis vectors and > firing off reverse-Monte Carlo cycles. There is no need to define red/black > symmetries in GSAS as the moment orientations are all looked after by SARAh. > Start with selecting the basis vectors of a single irreducible > representation, and then combining them if that fails to fit your data. It > is important to learn to read basis vectors and what they correspond to- > SARAh has some help files to help you get to grips with this. > > A word on GSAS vs Fullprof: > GSAS can only handle a magnetic structure that can be defined by a unit > cell. So, your incommensurate structure would need to be close to a lock-in > value to work, i.e. (0 0 0.32) is close to (0 0 > 1/3). That said, the definition of the magnetic structure can be very simple, > a set of moments in P1, so setting understanding > what it all means is quite unambiguous. > > FullProf has a k vector and is far more flexible in the ways in which > magnetic structures can be defined. This flexibility can make it hard to set > up a magnetic phase. SARAh -Refine tries to deal with most of these problems > for you (see the video help for the entire process), but it is sometimes > less obvious to see what is happening in the pcr file. > > If you have any questions, feel free to get in contact. Good luck! > > Andrew > > ps/ to see the SARAh documentation- install SARAh (which contains > SARAh-Representational Analysis and SARAh-Refine) on a windows computer > (including macs with parallels or vmfusion), and look in the help menu > > > On 23 Jul 2009, at 12:27, Alexander Barcza wrote: > > Dear Ross, > > Thanks for your reply. Sorry for being so exclusive in my initail mail, of > course I should have included everyone! The title of PhD student also > applies to me. > > I've seen a talk about SARAh by Andrew Wills at ISIS and was impressed by > it. Therefore I've already downloaded it but didn't succeed in using it yet. > > The problem is actually already at the beginning: > In my GSAS file I created a second phase with P1 and just copied the atoms > of the main phase over. > > I loaded that file by clicking <GSAS Controls>, <Load GSAS EXP file> > > How do I then define the MAT file? > > Sorry, it must be obvious but I can't find it. I am happy to read some > documentation but couldn't find any on the website ( > http://www.chem.ucl.ac.uk/people/wills/index.html). > > Thanks again for your help > Alex > > > > 2009/7/23 Ross H Colman <ucca...@ucl.ac.uk> > >> Dear Alex, >> >> You addressed your email to "experts on magnetic refinement" so I >> apologise if I am the only person to reply for I am a lowly PhD student. >> >> I am afraid I am not going to answer your questions directly but instead >> point you to the program that I use for magnetic refinements. Its free >> software available on CCP14 called SARAh. It allows you to input your >> magnetic atom sites, space group and k-vector then has the ability to >> modify your fullprof pcr file depending on how you wish to do the >> refinement. >> >> The process is a little involved, as it uses representational theory to >> reduce your problem to symmetry allowed structures, rather than freely >> refining all variables, but it has step by step instructions and I think >> there are even some online help videos for using it now. >> >> Available from: >> www.CCP14.ac.uk or >> www.chem.ucl.ac.uk/people/wills/index.html >> >> Best regards and good luck >> Ross Colman >> >> >> > Dear experts on magnetic refinement, >> > >> > We recently got some nice data on GEM, ISIS and I was able to get a >> decent >> > structural fit using GSAS and FullProf. >> > >> > Now I would like to refine the magnetic structure but this is my first >> > time >> > so I struggle to give FullProf what it wants. >> > >> > Since I'm a novice treating the magnetic data I'd have some general >> > questions about the input (which I attached below). The general >> structure >> > of >> > the pcr file seems to be OK, because it is read by FP but unfortunately >> in >> > the current form leads to a singular matrix. I'm sure I've done >> something >> > wrong with the input parameters. >> > >> > Here is a quick outline of what I want to achieve: >> > There are two magnetic ions in the structure. Mn and Co, but Co has a >> very >> > small moment (we know that from neutron diffraction from the 80's). The >> > moments seem to order in two incommensurate spiral structures (one for >> Mn >> > and one for Co) with the k-vector along c-axis. I thought it is best to >> > include a separate purely magnetic phase in FP. Would you agree and if >> yes >> > can I still do a multi-bank refinenment? >> > >> > Here is the snippet from the pcr file which was inspired by Ho2Cu2O5, an >> > example from the FP webpage (questions below): >> > >> ---------------------------------------------------------------------------------------------------------------------------------------------- >> > CoMnSi(magnetic >> > reflection) >> > >> > ! >> > !Nat Dis Mom Jbt Isy Str Furth ATZ Nvk More >> > 2 0 0 -1 -1 0 0 0.0000 2 1 >> > !Jvi Jdi Hel Sol Mom Ter >> > 3 0 0 0 0 0 >> > !Contributions (0/1) of this phase to the 6 patterns (FOR simplicity I >> > deleted all but the first pattern to get started) >> > 1 >> > !Irf Npr Jtyp Nsp_Ref Ph_Shift for Pattern# 1 >> > -1 13 -1 0 0 >> > ! Pr1 Pr2 Pr3 Brind. Rmua Rmub Rmuc for Pattern# 1 >> > 0.000 0.000 1.000 1.000 0.000 0.000 0.000 >> > P -1 <--Space group symbol >> > !Nsym Cen Laue MagMat >> > 4 1 3 1 >> > ! >> > SYMM x,y,z >> > MSYM u,v,w,0.0 >> > SYMM -x,-y,z+1/2 >> > MSYM u, v,-w,0.0 >> > SYMM x+1/2,-y+1/2,z >> > MSYM -u,v,-w,0.0 >> > SYMM -x+1/2,y+1/2,z+1/2 >> > MSYM u,-v,-w,0.0 >> > ! >> > !Atom Typ Mag Vek X Y Z Biso Occ Rm >> > Rphi Rtheta >> > ! Im Iphi Itheta beta11 beta22 beta33 MagPh >> > Co1 MCO2 1 1 0.15609 0.25000 0.05324 0.00000 1.00000 >> > 0.3 23.0 >> > 45.0 >> > >> > 220.00 0.00 230.00 280.00 0.00 >> 0.0 >> > 0.0 0.0 >> > 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 >> > 0.00 0.00 0.00 0.00 0.00 0.00 >> > Mn1 MMN2 1 2 0.02093 0.25000 0.68070 0.00000 1.00000 >> > 3.1 178.0 >> > 13.0 >> > >> > 0.00 0.00 0.00 0.00 0.00 >> > 0.0 0.0 0.0 >> > 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 >> > 0.00 0.00 0.00 0.00 0.00 0.00 >> > !-------> Profile Parameters for Pattern # 1 >> > ! Scale Extinc Bov Str1 Str2 Str3 >> Strain-Model >> > 4.8798 0.0000 0.0000 0.0000 0.0000 0.0000 0 >> > 1.00000 0.00 0.00 0.00 0.00 0.00 >> > ! Sig-2 Sig-1 Sig-0 Xt Yt Z1 >> > Z0 Size-Model >> > ! 0.000 186.666 0.000 <- Instr. par. + shifts. Only shifts >> are >> > refined >> > 0.000 33.363 0.000 0.000 0.000 0.000 >> > 0.000 0 >> > 350.00 31.00 0.00 0.00 0.00 0.00 >> > 0.00 >> > ! Gam-2 Gam-1 Gam-0 LStr LSiz >> > ! 0.000 0.000 0.000 <- Instr. par. + shifts. Only shifts are >> > refined >> > 0.000 0.000 0.000 0.000 0.000 >> > 0.00 0.00 0.00 0.00 0.00 >> > ! a b c alpha beta gamma >> #Cell >> > Info >> > 5.866537 3.692333 6.856417 90.000000 90.000000 >> > 90.000000 >> > >> > 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 >> > ! Pref1 Pref2 alph0 beta0 alph1 Kappa >> > ! Instr. par. + shifts: 1.597107 24.016417 1.496805 46.025921 >> > <- >> > Only shifts are refined >> > 1.000000 0.000000 0.000000 -3.937651 0.000000 0.000000 >> > 0.00 0.00 0.00 40.00 0.00 0.00 >> > !Absorption correction parameters >> > 0.00000 0.00 0.00000 0.00 ABS: ABSCOR1 ABSCOR2 >> > !Propagation vector 1 >> > 0.00000 0.00000 0.2300000 >> > 0.00 0.000 1.000 >> > !Propagation vector 2 >> > 0.00000 0.00000 0.2000000 >> > 0.0000 0.0000 1.000 >> > >> ------------------------------------------------------------------------------------------------------------------------------------------------ >> > >> > Questions (forgive me if they are obvious but I'm really just starting) >> > >> > 1) How do I decide on the space group? With incommensurate structures is >> > it >> > always just P -1 >> > >> > 2) Cen: IF I choose P -1 as the space group don't I have to set Cen to 2 >> > (i.e. centro-symmetric) >> > >> > 3)Laue: Is that the Laue symm of the the underlying crystallographic >> > phase? >> > >> > 4)SYM and MSYM: Is there a way to create them or do I have to deduce >> them >> > by >> > the space group I choose. Because for P -1 I wouldn't expect a +1/2 >> > component in them (do they come from the Laue sym?). There is really not >> > much about this in the FP manual >> > >> > 5) Atom types: Where in the manual do I find the correct types I have to >> > use? Is it always just M in front of the atom? >> > >> > 6)Do I enter the propagation vectors just as I did it, or do I have to >> do >> > something different? >> > >> > >> > To whoever takes the time in answering even just a small number of these >> > questions or giving any advice I am really thankful. I'm very intrigued >> by >> > this magnetic neutron scattering business but it's sometimes hard to >> keep >> > on >> > top of all the arising issues. >> > >> > With best regards >> > Alex >> > >> >> >> > >
