Attempts to unveil the intertwined influences which cause the complex properties of strongly correlated electron systems call, in principle, for a simultaneous investigation of all degrees of freedom in response to a change of external parameters, a task which is hardly possible. As an example, neutron scattering experiments on the manganites can certainly reveal structural and magnetic (spin) properties but not directly those related to the charge degree of freedom.

We focus on electronic transport measurements: global magnetotransport and local tunneling. In both cases, the conduction electrons close to the Fermi energy E_F are probed, i.e., those electrons which are as closely involved in the correlations as possible. As an example, our Hall effect measurements on CeCoIn₅ suggested the existence of a certain group of electrons that might be involved in magnetism and quantum criticality but not in superconductivity. This, of course, does not necessarily rule out the possibility of the antiferromagnetic fluctuations being involved in the formation of unconventional superconductivity, even if they do not become critical right at the critical field of superconductivity.

Moreover, the fact of probing the conduction electrons directly at the Fermi energy is crucial for the determination of spin polarization which refers to a difference in the density of states for the spin-up and spin-down channel right at E_F. Spin polarization is measured, e.g., in the skutterudite samples, by point-contact spectroscopy, a close relative of tunneling spectroscopy.

Clearly, such measurement techniques are most suitable for the investigation of strongly correlated electron systems which not only can be carried out within the relevant parameter space (e.g., low temperatures) but which also allow for the above-mentioned control parameters to be varied within an appropriate range. In case of magnetotransport, this is within experimental feasibility even though it typically calls for complex setups. We conduct magnetoresistance and Hall effect measurements down to about 50 mK, in magnetic fields up to 15 T and under pressure of up to 1.2 GPa. We have optimized our setup for sensitivity which is now better than 0.01 nV in two channels such that ρ_xx and ρ_xy can be measured simultaneously. We use lock-in technique (very important to keep track of the phase of the Hall signal) in connection with low-temperature transformers and low-noise voltage amplifiers. Typically, we conduct isothermal field sweeps, with the field ramped in steps to enable averaging.

Figure 1. Cryostat used for magnetotransport investigations.

It should also be noted that those measurements are most attractive the results of which can be compared to theoretical predictions or which could serve as input for theoretical considerations. In this context, the Hall measurements on YbRh₂Si₂ constitute a good example. However, as appealing as Hall measurements might appear an interpretation in case of strongly correlated electron systems is often exceedingly demanding because of their complex band structure.