M. Paulsson, T. Frederiksen, and M. Brandbyge

**Modeling inelastic phonon scattering in atomic- and molecular-wire junctions**

Phys. Rev. B **72**, 201101(R) (2005); **75**, 129901(E) (2007) [cond-mat/0505473]

Computationally inexpensive approximations describing electron-phonon scattering in molecular-scale conductors are derived from the nonequilibrium Green's function method. The accuracy is demonstrated with a first-principles calculation on an atomic gold wire. Quantitative agreement between the full nonequilibrium Green's function calculation and the newly derived expressions is obtained while simplifying the computational burden by several orders of magnitude. In addition, analytical models provide intuitive understanding of the conductance including nonequilibrium heating and provide a convenient way of parameterizing the physics. This is exemplified by fitting the expressions to the experimentally observed conductances through both an atomic gold wire and a hydrogen molecule.

M. C. Sullivan, D. R. Strachan, T. Frederiksen, R. A. Ott, and C. J. Lobb

**Effects of self-field and low magnetic fields on the normal-superconducting phase transition**

Phys. Rev. B **72**, 092507 (2005) [cond-mat/0505334].

Researchers have studied the normal-superconducting phase transition in the high-Tc cuprates in a magnetic field (the vortex-glass or Bose-glass transition) and in zero field. Often, transport measurements in "zero field" are taken in the Earth's ambient field or in the remnant field of a magnet. We show that fields as small as the Earth's field will alter the shape of the current vs voltage curves and will result in inaccurate values for the critical temperature T_{c} and the critical exponents ν and z, and can even destroy the phase transition. This indicates that without proper screening of the magnetic field it is impossible to determine the true zero-field critical parameters, making correct scaling and other data analysis impossible. We also show, theoretically and experimentally, that the self field generated by the current flowing in the sample has no effect on the current vs voltage isotherms.