After more than 20 years of intense research, the nature of the phase which appears below 18K in the compound URu₂Si₂ remains a mystery. Until now, no measurement using a microscopic probe has been able to directly link a physical parameter to this phase. We have just shown that this phase is characterized by low-energy magnetic excitations. This finding was made possible thanks to thermal expansion and inelastic neutron-scattering experiments simultaneously performed under pressure.

The intermetallic URu₂Si₂ is a so-called heavy fermion system (see Figure 1) whose ground state superconducts at ambient pressure and is anti-ferromagnetic for applied pressures larger than 0.5GPa, i.e. 5 kbar (Figure 1). Close to 18 K, at small pressure, a thermodynamical phase transition is observed in specific heat measurement, as well as in numerous other macroscopic measurements including electrical resistivity and thermal expansion. The nature of the order adopted by the system at this transition is not known as yet. Many types of order have been proposed - for instance orbital order, multipolar order, spin order associated with orbital currents, charge density waves - but none of these has been confirmed. In the absence of identification, this order is coined hidden order (HO).

Using the IN12 and IN22 triple axis spectrometers at the Institute Laue Langevin, we measured the neutron scattering intensity as a function of temperature for a pressure of 0.67GPa. In order to in situ determine the transitions towards the hidden and antiferromagnetic orders, which is a requirement for associating observed signals to the different phases, a strain gauge was glued onto the URu₂Si₂ crystal (Figure 2). By probing the variation of the sample length it gives access to its thermal expansion, whose anomalies signal the phase transitions (Figure 3) at T₀ and T_x. Below T_x, a strong neutron elastic scattering intensity is observed at the scattering vector Q0 = (100), in agreement with the presence of the antiferromagnetic order. The negligible intensity measured between T_x and T₀ comes from parasitic contributions, due to the survival of AF droplets generated near defects, and does not allow differentiation between the HO and paramagnetic phases. The hidden order phase therefore has no specific signature in these measurements: this is also the case for other experimental microscopic probe techniques like NMR.

Only inelastic neutron scattering intensity evidences it. As shown in Figure 4, a resonance is observed which is centered at 1.25meV. This inelastic magnetic signal is a signature of the hidden order, since it disappears in the paramagnetic and antiferromagnetic phases. What is the nature of the order parameter associated with these excitations? As of today, it remains unknown. Nevertheless, it can be noticed that this signal is observed at Q₀, where the elastic scattering is seen in the antiferromagnetic phase. Interestingly, it reminds us of the ubiquitous resonance in high temperature superconductors based on copper.

Heavy fermions The highly unusual characteristics of strongly correlated electron-metallic systems appear in the properties of their conduction electrons. Strong correlations between these electrons enhance their apparent mass to values much higher than in ordinary metals, like copper. Electrons being fermions in the statistical physics language, these metals are quoted as heavy fermion systems. They are known for their intriguing properties: an example is the competition of magnetic and superconducting states at low temperature.