We tried to determine the small atomic shifts in multiferroic YMnO3, HoMnO3 and LuMnO3 at the magnetic phase transition temperature by neutron powder diffraction and so far have not succeeded. We first measured on the neutron diffractometer D20 that has a limited Q range (Q_max = 7 Å^-1) and then on POWGEN at SNS (spallation source) up to a very high Q. All these data lead to no definite conclusions about the atomic shifts at T_N. We then tried polarized neutron diffraction on single crystals of the multiferroic YMn2O5. By measuring flipping ratios we could estimate the atomic displacements at T_N. The structural modulation has the same propagation vector as the magnetic modulation so we get very small nuclear (structural) contribution on top of strong magnetic contribution of diffracted intensity. By measuring flipping ratios we could estimate the atomic (Mn) displacements. In neutron antiferromagnetic reflections contain a very small structural contribution. In resonant or even nonresonant X-ray scattering the structural (charge) contribution should dominate the magnetic contribution of diffracted intensity. So one should in principle be able to determine the atomic displacements more accurately. I know that TbMnO3 has been studied by resonant X-ray diffraction but the atomic displacement (O) was found to be rather small. We are convinced that atomic displacements of Mn ions (Mn^3+ and Mn^4+) are relatively large in YMn2O5. Will it be possible to measure it at the ESRF on some beam line?
I guess that should be possible. Beamlines that come to mind are BM28 XMaS at the ESRF and P06 at Petra III/Desy.. I need to know more about the expected movements to give better advice about how to best set up the experiment. Maybe we should have a coffee together one of these days.
yes XMas can get down in the soft range and you might get also the YL lines. In principle also ID31 high resolution (and of course higher flux being on ID) powder diffraction beamline can also be suitable (it gets below the ca. 6.5 keV of the MnK). Another possibility for sure is I16 at Diamond and Steve Collins for sure has also great experience with magnetism. In both cases you can work in resonant, non resonant and over a large range of energies (to reduce possible surface effects).
Those are for pure powder diffraction: on the other hand, XMas is also suited for single crystal diffraction (should you have one)
Sure ID 31 is good for the present purpose. The problem of the neutron diffraction data for a magnetic sample which orders with propagation vector k = 0 (hexagonal manganites) is that the magnetic and structural reflections are at the sample positions. So the refinement of simultaneous crystal and magnetic structures leads to correlations between structural and magnetic parameters especially close to T_N where diffuse magnetic scattering appears. Since the X-ray powder diffraction on ID31 will contain almost no signal (I assume) then the data will give a good structural refinement. My only worry is that the O positions will probably not determined very accurately due to the presence of heavy scatters like Y, Ho, Lu etc.
For RMn2O5 the propagation vector is 1/2, 0, 1/4 and the magnetic and structural reflections are separated except for the small structural contribution on magnetic reflections. But single cystal X-ray diffraction with polarization analysis can separate them I guess. We could do that with polarized neutrons as well.
Tapan, you are perfectly right: oxygens are on the "transparent" side for X-rays in your case... that's why people always do a combined neutron + X-ray refinement! Right now both fullprof and topas can do that
Another possibility, if you are interested in the Mn displacement, could be the single crystal diffraction at the Swiss Norwegian Beamline at the ESRF, though I anticipate it will be difficult. Have a look at M. Pregelj, O et al, Phys. Rev. Lett. 103, 147202 (2009).
Jane Brown and I tried single crystal neutron diffraction on hexagonal YMnO3 and HoMnO3 on the diffractometer D9 of ILL. We wished to check whether Mn moves out of the position (1/3,0,0) at the antiferromagnetic transition and found no shift. Some authors have reported shift by powder neutron diffraction but I suspect the validity of those results. This is because of the magnetic structure with the propagation k = 0 that causes overlap of the magnetic and structural reflections. This leads to strong correlations between structural and magnetic parameters in neutron diffraction. X-ray diffraction is better because there are practically no magnetic intensities.
- Badges
- Science method
- Similar topics
- Methodology
- Crystallography
- X-ray Diffraction