Archive | magnetism RSS for this section

StrucFact or whatever it is called

Wrote a crude program to calculate X-ray, neutron and neutron magnetic diffraction scattering factors.

Introduction

StrucFact or StrucFact.exe is a crude program to calculate expected intensities for neutron diffraction (magnetic and nuclear) and X-ray diffraction (in the version from 2013). It is distributed as source code because any serious user will likely need to modify the source to make sensible use of it. It’s mathematics is essentially taken from Neutron Diffraction by George Bacon and H. M. Rietveld J. Appl. Cryst. (1969) 2 65-71 (see also here), and its mandate is limited; its job is to calculate F2 for neutron scattering for arbitrary cells, magnetic or nuclear, and more recently for X-ray diffraction.

It is `developed’ solely on an `as needed’ basis, which means I add `features’ when I need them to solve some problem I am working on. The inverted commas may seem gratuitous, but they are not!

I am sure there are better tools out there for everything that this program does, and I advise against using it. There should be a README.TXT and the code itself distributed with this file.

Please Note

1. It does not work for incommensurates (unless you want to define an enormous cell).

2. It treats every cell as P1 (i.e. you have to give it all atoms explicitly).

3. The nuclear and magnetic cells must be the same size, which means that if one is bigger than the other (usually magnetic bigger than nuclear) you have to put the atoms from the smaller cell into the bigger, including inserting all redundant copies of atoms.

4. No corrections for intensities (e.g. not even Lorentz), no B-factors beyond the isotropic.

In other words, it is remarkably limited. Why anyone would want to use it I do not know when FullProf and GSAS and the like are around. Having said that, it is quite simple (in the sense that everything has to be done explicitly, so it is laborious but not as conceptually demanding) compared to such programs, and the (minimally tested) code is here available.

2 Compiling

This is Fortran90 code that does not need any extra libraries. gfortran is to be preferred.

Non-static binary:

gfortran -o StrucFact.exe strucfact.f90

Please report errors in the code to /dev/null, although if desperate you can email me.

Download from http://djg.altervista.org/downloads/StrucFact_Files.tar.gz or http://djg.altervista.org/downloads/StrucFact_Files.zip. There’s a PDF inside the archive that gives more info.

Meh.

Advertisements

AANSS 2016 — it’s approximately that time of year again, again.

Get that neutron feeling.

Get that neutron feeling.

 

www.anbug.org

The AANSS is a great mix of formality and informality, quality science in a relaxed atmosphere. Anyone who has or might or ought to use neutron scattering in their work (and isn’t that all of us, really?) is invited. And here’s a trick: Registration is $50 cheaper for ANBUG members but ANBUG membership is free! So join up!

AINSE

 

Oldly.

A Nice Piece of Work: Evolution with applied field of the magnetic structure of TbNiAl4

More years ago than I like to recall, I helped out with a study of the magnetic structure of TbNiAl4. The first paper we did on that goes back to about 2006. I did a refinement of the neutron diffraction data, and I got some things right, but what was not apparent — partly due to the limitations of powder diffraction and partly due to the limitations of the user (me) — was the subtlety and complexity of the real magnetic structure.

reyner2

Fortunately, PhD Scholar Reyner White has now done it properly, with better data.  He found that the propagation vector I determined was pretty good, but has been able to show that the magnetic structure shows an “`elliptical helix’ type structure in which the moments rotate in the ab-plane as one moves along the c-axis”, which is far more concrete — I showed that it was incommensurate, and I found the propagation vector, but exactly what it was that was incommensurate I was wily enough not to say.

It sure is nice to work with good people who take the time to do things well.

 

Chemical and magnetic ordering in Fe0.5Ni0.5PS3, again.

An article referred to before has been officially published. It is called ‘Chemical and magnetic ordering in Fe0.5Ni0.5PS3‘ and is available (subscription needed, sorry) at http://link.springer.com/article/10.1007%2Fs10751-014-1108-6.

D. J. Goossens, G. A. Stewart, W. T. Lee, A. J. Studer, ‘Chemical and magnetic ordering in Fe0.5Ni0.5PS3‘, Hyperfine Interactions, April 2015, Volume 231, Issue 1-3, pp 37-44.

What’s it about? Well…

A version, possibly final but I'm not sure, of the abstract I put in to the conference.

A version, possibly final but I’m not sure, of the abstract I put in to the conference.

Many thanks to Glen and Hal Lee.

Quote of the Day: “there is a lot of dull, hard work to be done.”

Cover of <i>Magnets</i> by Francis Bitter.

Cover of Magnets by Francis Bitter.

Before you can read or write stories you must learn spelling and grammar; before you can play a sonata on the piano you must learn scales, harmony, and musical notation; and before you can go into a laboratory and make an intelligent stab at discovering something new, there is a lot of dull, hard work to be done.  No one escapes that.

Francis Bitter, Magnets: The Education of a Physicist

Link.

The Wrong Way to Write a Paper

Maybe eighteen months or even two years ago I notice that the conference on Hyperfine Interactions and Nuclear Quadrupole Interactions (HFINQI) was coming to Canberra. I always take a close look at conferences that come to Canberra, because I work there.   At the time I was working on a project studying the properties of some magnetically unusual materials, the MPS3 family of compounds.  In particular, we were looking at ones that contained 50:50 ratios of two different magnetic transition metals on the M site, and we had found some interesting behaviour in Fe0.5Ni0.5PS3.  This compound seems to show time-dependent magnetic properties suggesting glassiness and disorder.

Now, iron-containing materials can be studied using Mössbauer spectroscopy, which is sensitive to the crystal and magnetic environments of the Fe atoms, in this case Fe2+ ions.

My students and I had studied the materials using magnetometry and neutron diffraction, but if I wanted to put the work in to HFINQI I needed a hyperfine technique, which Mössbauer is.  I contacted Glen Stewart at PEMS at UNSW Canberra, who is an expert in Mössbauer and we talked over an experiment.  In the end, Glen collected some lovely data and was able to fit it with a really nice model, and here’s a picture:

Mossbauer figure

A sample of the 57Fe-Mössbauer spectra recorded for Fe0.5Ni0.5PS3 as a function of temperature. In each case, the fitted theory (black line) is a sum of three magnetically-split sextets (coloured lines) which are associated with three distinct magnetic Fe-site environments.

We find that the time-dependence is not apparent in the Mössbauer, which is not surprising as it is a slow technique.  We also find that the Fe environments are not random, which is perhaps the most significant result, as it will relate to the interactions between magnetic species and so to the magnetic properties.

Now, while putting a paper into HFINQI prompted the Mössbauer work, which is a backwards way of choosing what science to do, it was by no means a silly experiment to do, as the site-symmetry information available from the technique complements the crystal symmetry information from neutron diffraction, and has indeed enhanced our understanding of the compound.

The paper is available (paywalled, I’m sorry to say) at http://dx.doi.org/10.1007/s10751-014-1108-6.

Chemical and magnetic ordering in Fe0.5Ni0.5PS3 by D.J.Goossens · G.A.Stewart · W.T.Lee · A.J.Studer, Hyperfine Interactions (full reference not yet available as of Dec 27, 2014).

 

Approximately never

part1part2

This is an old drawing from a Physics tutorial.  Nothing very flash.

Rapid Microwave Synthesis and Structural Phase Diagram of Doped Yttrium Manganate

Last year Fred Marlton did his Honours in Chemistry at ANU, and this year the little paper we wrote came out.  It was more done to give Fred a sense of the publishing process, and communicating science in that way, but it does have a few handy results in it as well, ones that could save time and effort for people synthesising compounds from the family LnxY1−xMnO3.  The paper is published in Zeitschrift für Naturforschung B, which is a venerable journal with, based on this experience, an excellent editorial process.  The abstract can be viewed here.

In essence the title says it all; we used microwave sintering to make the samples, which saved an order of magnitude in time and electricity, and allowed Fred to quickly map out the whole phase diagram describing where these compounds form.

 

The phase purty versus composition diagram for lanthanide-doped yttrium manganate.

The phase purity versus composition diagram for lanthanide-doped yttrium manganate.

 

What we found was that solid-state microwave synthesis allows manufacture of high-quality samples in hours rather than days. The resulting phase diagram accords well with results from the literature, and from calculations based on the Goldschmidt tolerance factor for the stability of perovskite structures, suggesting that the transformation from hexagonal to perovskite with doping is driven essentially by ion sizes. Some results concerning the microwave synthesis of BaLnInO4 compounds, where Ln is a lanthanide, are noted. Microwave sintering of BaNdInO4 yields single-phase samples where conventional sintering does not.

A Minor Tool

Magneticians sometimes want to plot curves of spontaneous magnetization as a function of temperature for magnetically ordered materials, usually to compare to experimental results from magnetic neutron scattering or magnetometry.  These curves are based on the Brillouin function. Their shape depends on the total angular momentum, J or sometimes S.  They look something like this:

Brillouin curve for J=1.

Brillouin curve with J=1, plotted along with some experimental data of some sort.

All this program, found here at http://djg.altervista.org/downloads/Brillouin_curve_calc.tar.gz does is take the J from the user and write out the numbers for the plot.  The actually machinery is taken from M I Darby 1967 Br. J. Appl. Phys. 18 1415 doi:10.1088/0508-3443/18/10/307.  And it is this paper that should be referenced, or in fact used directly. Comments on g95 for Windows can be found here.

And, yes, I did draw the graph in Excel.  Ugly but good enough…