Magnetic structures are usually determined by neutron powder diffraction. In difficult cases one may have to use single crystals or even polarized neutrons. Although X-ray magnetic scattering is very weak, orders of magnitude gain in intensity and brilliance of synchrotron X-rays in modern synchrotron sources should allow magnetic structures to be determined routinely by X-ray diffraction (if not by powder diffraction), thus making the use of hazardous and expensive neutron diffraction experiments unnecessary. At least, that was what confidently expressed by the founders of X-ray magnetic scattering community around 1990, almost a quarter of a century before. I wonder how much of that has become a reality in recent years. Can anyone please comment on this?
I would like to add my support to the concept of complementarity between X-rays and neutrons for determining magnetic structures. Naturally the neutron scattering cross section possesses several large advantages for the study of magnetic structures, but there also the downsides, for example associated with the need for large quantities of sample and highly absorbing isotopes. For these reasons, recently x-ray resonant scattering has proved most advantageous for the recent study of the iridates, which can only be synthesized in the form of very small crystals, and which typically have small moments. Also non-resonant X-ray magnetic scattering should not be disregarded entirely, despite the very small cross-section, because it can also have great power, particularly when looking at materials with poor resonant enhancements, such as the 3d transition metals at the K edges. The non-resonant signal is typically only of order 2 times weaker than the resonant signal in the manganites, and is on a comparable scale to charge scattering from magnetoelastic distortions, opening up interesting possibilities to look at interference effects.
Currently I agree that it is not feasible to investigate magnetic structures using powder X-ray diffraction, but potentially with the development and routine use of 2d detectors, hopefully combined with polarisation analysers, this might become a more realistic possibility.
I think that much more easy to obtaine infirmation conserning of magnetic structure with the help of photoemission spectroscopy with angle resolution (ARPES). Magnetic structure modify Brilloune zone and this effect may be observed by ARPES.
Yes. Please refer the article: Journal of Physics: Condensed Matter Volume 16 Number 29;Maxim V Lobanov et al 2004 J. Phys.: Condens. Matter 16 5339 doi:10.1088/0953-8984/16/29/023;Crystal and magnetic structure of the Ca3Mn2O7 Ruddlesden?Popper phase: neutron and synchrotron x-ray diffraction study
Maxim V Lobanov et al. 2004
The crystallographic and magnetic structures of Ca3Mn2O7 Ruddlesden?Popper phase have been determined by a combination of neutron and synchrotron x-ray diffraction. Two-phase behaviour observed at room temperature is attributed to an incomplete structural phase transition. The magnetic structure was solved in the Cm'c2_1' Shubnikov group with dominant G-type antiferromagnetic order in the perovskite bilayers. The temperature evolution of the structural and magnetic parameters is presented.
@Klaus: It is good to see that you spend some time to coney your experience and knowledge in a lucid way.
Klaus here is exactly correct. Neutron data is far more useful than X-Ray data for magnetic structure. X-ray diffraction is useful for structural analysis but the interaction with magnetic fields is weak compared to its interaction with the actual structure. Neutrons also penetrate deeper to give a better profile.
What is interesting with magnetic materials is that if you can do both experiments with similar equivalent 'wavelengths', you can directly compare just the structural pattern from X-ray, with the enhanced structure and magnetic structure generated by neutron diffraction - this is a great tool.
Also for magnetic structure, consider Mossbauer spectrometry, as this deals directly with the magnetic interactions in the material, see one of my papers for a Mossbauer and Neutron study on FeGa.
It is very difficult..as you know. Here they have tried...
http://iopscience.iop.org/0953-8984/17/46/L02
I would like to add my support to the concept of complementarity between X-rays and neutrons for determining magnetic structures. Naturally the neutron scattering cross section possesses several large advantages for the study of magnetic structures, but there also the downsides, for example associated with the need for large quantities of sample and highly absorbing isotopes. For these reasons, recently x-ray resonant scattering has proved most advantageous for the recent study of the iridates, which can only be synthesized in the form of very small crystals, and which typically have small moments. Also non-resonant X-ray magnetic scattering should not be disregarded entirely, despite the very small cross-section, because it can also have great power, particularly when looking at materials with poor resonant enhancements, such as the 3d transition metals at the K edges. The non-resonant signal is typically only of order 2 times weaker than the resonant signal in the manganites, and is on a comparable scale to charge scattering from magnetoelastic distortions, opening up interesting possibilities to look at interference effects.
Currently I agree that it is not feasible to investigate magnetic structures using powder X-ray diffraction, but potentially with the development and routine use of 2d detectors, hopefully combined with polarisation analysers, this might become a more realistic possibility.
If you want to probe magnetic structures upon x-rays you may also consider Resonant X-ray Scattering. This technique combines absorption and diffraction. In these experiments the energy of the photons is tuned close to the absorption edge of the element you want to probe.
Then you have to consider the imaginary part of the scattering tensor f. This component, when the energy is tuned close to an absorption edge, depends on the energy, the light polarization and the wave vectors orientation.
You can have a look here for a summary of the physics behind this technique, which is widely used: L. Paolasini, F. de Bergevin, "Magnetic and resonant X-ray scattering investigations of
strongly correlated electron systems." C. R. Phys. 9, 550 (2008).
few examples I'm aware of powder diffraction with X-rays, definitly not hte first step in your investigations:
Orbital and magnetic ordering in Pr1−xCaxMnO3 and Nd1−xSrxMnO3 manganites near half doping studied by resonant soft x-ray powder diffraction
U. Staub, M. García-Fernández, Y. Bodenthin, V. Scagnoli, R. A. De Souza, M. Garganourakis, E. Pomjakushina, and K. Conder
Phys. Rev. B 79, 224419 – Published 17 June 2009
PRB 1 citation
Magnetic and electronic orderings in orthorhombic RMnO3 (R=Tm, Lu) studied by resonant soft x-ray powder diffraction
M. Garganourakis, Y. Bodenthin, R. A. de Souza, V. Scagnoli, A. Dönni, M. Tachibana, H. Kitazawa, E. Takayama-Muromachi, and U. Staub
Phys. Rev. B 86, 054425 – Published 17 August 2012
Magnetic and electronic properties of RNiO3 (R = Pr, Nd, Eu, Ho and Y) perovskites studied by resonant soft x-ray magnetic powder diffraction
By: Bodenthin, Y.; Staub, U.; Piamonteze, C.; et al.
JOURNAL OF PHYSICS-CONDENSED MATTER Volume: 23 Issue: 3 Article Number: 036002 Published: JAN 26 2011
I know of just one study in which (resonant) x-ray scattering was applied to a sample in power form: S.P. Collins et al., J. Phys. Cond. Matt., vol. 7 , pages : L223-L229 (1995). There, they didn't "determine" a magnetic structure but took a known magnetic structure (that of uranium dioxide) and showed that one of its magnetic powder diffraction peaks could be detected below TN working with linear polarized x-rays tuned to the U M4 edge.
Thus, to answer your question, synchrotron x-rays CAN in principle be used to solve magnetic structure in a powder sample (well, at least to get a handle on the magnetic periodicities), HOWEVER in practice -- for reasons expressed by others who've addressed your query -- it is technically (very) difficult.
For the case of a very small quantity of powder sample, prohibitively small to allow neutron diffraction, resonant x-ray scattering in the soft x-ray regime (e.g., at L2,3 edges for 3d element o M4,5 edges for a 4f element) could be considered a route to determine magnetic periodicity/ies. For example, working in transmission geometry, as people do in the magnetic SAXS community -- see, e.g., Kortright et al. PRB vol. 71, 012402 (2005) -- one could envisage being able to detect the lowest Q ring of magnetic diffraction in a transition-metal or rare-earth based material. I don't know if anyone has ever tried such a study, with powder mounted on, e.g., an Si3N4 membrane to allow scattering in transmission geometry.
When I talk about determination of magnetic structure then I of course think of determining (1) magnetic periodicity or the propagation vector, (2) spin directions of magnetic atoms and also (3) the magnetic moments associated with each magnetic atoms. The first goal is of course to determine the magnetic periodicity and it looks like that we can achieve this by X-ray diffraction in some cases at least. But for the other two steps you have to put the magnetic intensities in absolute scale. That is not at all a easy job in X-ray diffraction. In resonant X-ray scattering this is hardly possible as far as I know but I would liked to be informed by the specialists. Even for non-resonant X-ray diffraction this is not an easy job either. It is not trivial to put the intensities of charge and magnetic intensities in absolute scale or at least in the same scale. Again I would like to hear about it from the specialists. Things may have changed meanwhile!
I disagree with the statement that the magnetic intensities have to be put on an absolute scale in order to complete step (2), the determination of the spin directions. For example, the measurement of azimuthal dependence of resonant x-ray scattering is done (quite routinely, I think) to determine spin directions.
I agree, however, that to complete step (3) -- moment determination in muB -- RXS intensities may be difficult to model, although I don't how far code such as the FDMNS package has got to doing this succesfully.
Oooops, I'd forgotten that we we're still talking about a sample in powder form. Thus, step (2) couldn't be achieved by the azimuthal method.
@Peter: I agree that the step (2) can be achieved if you can at least put all magnetic intensities in the same scale. Can one do that easily? May be possible in powder diffraction (you know better) but for single crystals with odd shape and in reflecting geometry (I am not thinking of very high energy X-ray diffraction where the absorption is considerably reduced) it is hardly possible. I guess if you are talking about resonant X-ray scattering (may be for powder diffraction the resonance enhancement is necessary) then achieving step (2) is impossible. What do you think?
If you are considering soft X-ray powder diffraction to measure a single magnetic reflection (or may be two) then forget about solving the magnetic structure complete with all the three steps.
@Klaus: You may explain the difficulties we encountered to put the magnetic intensities in absolute scale (or at least in the same scale of charge scattering) for our EuAs3 sample even with the hard X-rays on ID15.
Ideally no; Synchrotron xrd is not the suitable for magnetic structure at all because x-ray see only coherent volume of electrons no matter material is magnetized or not. But one can correlate nicely their magnetism with that of the lattice modifications. You can follow the link
http://arxiv.org/ftp/arxiv/papers/1309/1309.6417.pdf
Dear Tapan, Klaus, Helen, Valerio, Peter, Vera, etc. Just to add more meat into the discussion, I agree with all what have been said before as most/all of the participants in this blog are RXS experts. To me X-ray powder diffraction experiments, either resonant or not, in view of extracting the magnetic structure is not the correct technique. The signal, if any, is way too small for any sensible magnetic structure refinement. Indeed there are some papers showing some reflections here and there but nothing beyond a curiosity. In addition you have to consider that "powder" is just a collection of small single crystals. And presently it is possible to select one of these small single crystals (1 mm^3 is enough, even less should be feasible) and carry out diffraction experiments with it, either with a 2D detector or by applying polarization techniques that Helen, Peter or Valerio are expert on. Remember that by chosen adequately the reflection and the experimental conditions a very few reflections are required to get a magnetic structure.... if antiferromagnetic. This approach is to me more meaningful than "powder" diffraction. In any case, and if samples are not pathological, neutron diffraction is simply.... the best and most straightforward technique for magnetic structures. But we all know that, n'est pas Tapan ?
I agree with all of you. One paradigmatic case is Holmium metal, where magnetic X-ray scattering re-vealed that the wave vector did not change smoothly
below ~ 30 K; instead a series of commensurate lock-ins were observed. These were explained by a spin-slip model for Ho!.
Well can we then conclude that the early X-ray magnetic scattering enthusiasts claimed more than they should have perhaps? We still need neutron scattering for studying magnetic structure and a complete understanding of magnetism in general.
G. Srajer at al., Advances in nanomagnetism via X-ray techniques, Journal of Magnetism and Magnetic Materials 307, 1-31 (2006).
24 years later the statement posted above is still valid. Up to now a complex magnetic structure cannot be determined with resonant magnetic x-ray scattering from scratch. Once the magnetic structure is known, many more details can be revealed by XRMS, such as the spin and orbital ordering not available from neutron scattering.
Dear Colleagues, one, more general, point to the discussion of the problem.
X-ray definition of the magnetic structure needs f- or d-electrons as a source of magnetic moment. Due to this, no organic or polymer magnetism for X-ray study, that's only for neutrons forever.
- Badges
- Science topic
- Similar topics
- Physics
- Condensed Matter Physics