Birth of a Polaron

Introduction

Magnetoresistance is the phenomenon of a material changing its electrical resistivity when it is placed in a magnetic field. This phenomenon has aroused great interest recently because of the possibility of making use of it in devices such as read/write heads in computer disc-drives. The magnetoresistance of traditional materials is quite small: the passage of current is not greatly affected by external magnetic fields. Recently, however, various materials have been made with extremely large magnetoresistances. "Giant" magnetoresistive were discovered first. These are multilayers of ferroelectric and non-magnetic metals. When the magnetoresistance of the special kind of perovskite manganese oxide, La1-xCaxMnO3+d, was discovered to be much larger, the name "Colossal" magnetoresistance was coined.

Our Interest

There is thought to be an important role played by the lattice in the phenomenon in this material. In particular, it is known that the material undergoes a metal-insulator transition accompanied by a ferromagnetic to paramagnetic transition. It has been suggested that charges tend to localize in this material as polarons, localized charges surrounded by a local structural distortion cloud. Using the Pair Distribution Function analysis of neutron powder diffraction data we were able to observe the formation of these polarons as the material went through the metal-insulator transition.

This work has appeared in Phys. Rev. Lett. 77 715 (1996) in a paper which looked like this: (gziped postscript, 93kB)

This plot shows the height of a measured PDF peak in 3 samples as a function of temperature. The height of the peak gives information about the degree of order of the manganese to oxygen and oxygen to oxygen bonds in the material.

The arrows show the temperature of the metal-insulator transition. Sample (c) never becomes metallic, (a) does at 235K and (b) at 180K. What we can see is that the PDF peak height comes down sharply in the vicinity of the metal-insulator transition.

At the metal-insulator transition the Mn-O and O-O bonds are suddenly becoming more disordered and changing their lengths. This is exactly what is expected if polarons are forming: in the polaronic (above TMI) state there will be a range of Mn-O bond lengths on different sites depending on whether or not a charge resides there. In a non-polaronic state, all of the Mn sites will be identical, the Mn-O bond length distribution better ordered, and the PDF peaks sharper.

We can actually see what the polaron looks like by comparing what the PDFs look like above and below the metal-insulator transition and taking the difference between them. This is shown in this figure. The experimental PDFs taken above and below TMI are shown as curves (i). The difference between them is shown, magnified, as the open circles in the lower panel.

We have calculated the PDFs from 2 models and these are shown as curves (ii). The difference between the models is shown as the solid line in the second panel. It is clear that the difference between the two models matches well the difference between the two experimental PDFs. This difference is a breathing mode distortion of the oxygen ions around 1-in-4 of the manganese ions. There is approximately one charge for every 4 manganese ions in this material and they are causing the oxygens to collapse around themselves.

[Postcard] Return to Simon Billinge:Home Page