Letters to Nature

Nature 414, 887-889 (20 December 2001) | doi:10.1038/414887a; Received 9 August 2001; Accepted 19 October 2001

A limit on spin–charge separation in high-Tc superconductors from the absence of a vortex-memory effect

D. A. Bonn1,2, Janice C. Wynn2, Brian W. Gardner2, Yu-Ju Lin2, Ruixing Liang1, W. N. Hardy1, J. R. Kirtley3 & K. A. Moler2

  1. Department of Physics and Astronomy, University of British Columbia, Vancouver, British Columbia, V6T 1Z1 Canada
  2. Departments of Applied Physics and Physics, Stanford University, Stanford, California 94305, USA
  3. IBM T.J. Watson Research Center, PO Box 218, Yorktown Heights, New York 10598, USA

Correspondence to: D. A. Bonn1,2 Correspondence and requests for materials should be addressed to D.A.B. (e-mail: Email: bonn@physics.ubc.ca).

There is a long-standing debate about whether spin–charge separation is the root cause of the peculiar normal-state properties and high superconducting transition temperatures of the high-T c materials. In the proposed1 state of matter, the elementary excitations are not electron-like, as in conventional metals, but rather the electron 'fractionalizes' to give excitations that are chargeless spin-1/2 fermions (spinons) and charge +e bosons (chargons). Although spin–charge separation has been well established in one dimension, the theoretical situation for two dimensions is controversial and experimental evidence for it in the high-T c materials is indirect. A model2 with sharp experimental tests for a particular type of separation in two dimensions has recently been proposed. Here we report the results of those experimental tests, placing a conservative upper limit of 190 K on the energy of the proposed topological defects known as visons. There is still debate3 about the extent to which this experiment can settle the issue of spin–charge separation in the high-T c copper oxides, because some forms of the separation are able to avoid the need for visons. But at least one class4, 5, 6 of theories that all predict a vortex-memory effect now are unlikely models for the copper oxides.