Strain tuning and precision instrumentation
We perform technically challenging measurements on correlated electron materials. Our most important tool is controlled lattice distortion, using home-built piezoelectric-based uniaxial pressure cells, versions of which have been commercialised. Through careful sample preparation and mounting, we have been able to compress hard oxide materials including Sr2RuO4 and YBa2Cu3O6+x by over 1%. In the process, we induce strong changes in electronic properties. For example, at a uniaxial compression of ~0.6% one of the Fermi surfaces of Sr2RuO4 passes through a topological transition, and in connection with this change the superconducting transition temperature more than doubles.
Uniaxial pressure was until recently an underused experimental technique, due to the technical difficulties. The largest difficulty was to achieve both large and highly homogeneous uniaxial stress in samples; we have solved this by preparing samples as relatively narrow bars, mounting them carefully, and applying force along the long axis. The quantitative effects of laboratory-achievable uniaxial pressures often far exceeds those of achievable magnetic field: Depending on the material, uniaxial compression by 1% can be equivalent to an applied field of around a thousand tesla. By lifting lattice symmetries, for example by applying an orthorhombic lattice distortion to a tetragonal crystal, the effects of uniaxial pressure are also often qualitatively different from hydrostatic pressure.
Correlated electron materials are in general complicated beasts, with many processes occuring simultaneously. Lattice distortion can strengthen certain process and suppress others. By continuously tuning the lattice parameters, a given material can be evolved into another with different properties, and by studying each step along the way can learn more than from looking at a single fixed point. We show below two examples, Sr2RuO4 and Sr3Ru2O7.
- Clifford Hicks, Group Leader
- Alexander Steppke, Postdoctoral Research Assistant (joint with Mackenzie group)
- Joonbum Park, Postdoctoral Research Assistant
- Mark Barber, Doctoral Student (joint with Mackenzie group)
- You-Sheng Li, Doctoral Student (joint with Mackenzie and Nicklas groups)
- Jack Bartlett (joint with Mackenzie group)
- Veronika Sunko, Doctoral Student (joint with Mackenzie group and King group, U. of St Andrews)
- Fabian Jerzembek, Doctoral Student (IMPRS, joint with Mackenzie group and Prof. H.-H. Klauss TUD)
- Po-Ya Yang, Doctoral Student (IMPRS, joint with Mackenzie group and Prof. M. Vojta TUD)
Uniaxial pressure apparatus
We apply pressure using piezoelectric actuators. They are arranged in a way that their thermal contraction cancels.
The applied pressure can be varied continuously. For example, to compress the sample, the central actuator is lengthened, and/or the outer actuators shortened.
Electronic structure calculations indicate that if Sr2RuO4 is compressed uniaxially by about 0.75%, its largest Fermi surface should go through a Lifshitz transition, where it connects with its copies in the adjacent Brillouin zones.
The effects of this transition are enhanced by the quasi-two-dimensional electronic structure of Sr2RuO4. In our experiments, we find that the superconducting transition temperature Tc of Sr2RuO4 passes through a pronounced peak when the lattice is compressed by ~0.6%, at which it is more than double its value in the unstrained material. The electrical resistivity in the normal state also peaks at about this strain. These changes are the result of tuning the material to the Lifshitz transition.
Sr3Ru2O7 has a field-induced magnetically ordered phase, occuring when a field of about 8 T is applied along the c axis. This phase is associated with a metamagnetic quantum critical point. Metamagnetism is a field-induced jump in magnetisation – one can imagine it as field-induced ferromagnetism. Metamagnetic transitions are usually first-order transitions, however in some materials the critical endpoint of a metamagnetic transition can be at unusually low temperatures. When this is the case, one gets soft fluctuations all over the Fermi surfaces and a high susceptibility towards ordered phases.
This ordered phase turns out to be exceptionally sensitive to applied orthorhombic lattice distortion. When Sr3Ru2O7 is compressed, the resistivity along the compression axis rapidly increases, and along the transverse axis rapidly falls. The regions of enhanced longitudinal and transverse resistivity overlap. This shows that we have two distinct order parameters, a density wave along x and a density wave along y, that can coexist.
Piezoelectric-based strain tuning is a recently-developed technique, and we are expanding its use to other materials. We are also pursuing a number of technical developments, including:
- Heat capacity and thermal conductivity measurements of samples under uniaxial pressure.
- Our present apparatus has very high stiffness, and so controls the strain in samples more than the stress. For samples with structural transitions, of which there are many, it would be better to controll the stress, by building apparatus with either much lower stiffness, or with feedback.
- Reducing the sample size: using a focused ion beam to prepare small samples into precisely-defined shapes with smooth surfaces, we expect to achieve higher strains.
- Scanning SQUID: by combining scanning SQUID susceptometry and uniaxial pressure, we can perform precision measurements of, for example, how the magnetic susceptibility of a material varies with uniaxial pressure.
Below is an SEM image of a sample that has been shaped with a plasma focused ion beam. The sample is placed across a ~100 µm-wide gap, and we apply stress by varying the width of this gap.