Ultra-high purity delafossite metals

October 24, 2018

MPI-CPfS co-workers: Pallavi Kushwaha, Veronika Sunko, Nabhanila Nandi, Helge Rosner, Markus Schmidt, Horst Bormann, Clifford Hicks, Frank Arnold, Elena Hassinger. Philip Moll and Andy Mackenzie

Fig. 1: a) As-grown PtCoO2 single crystals; b) x-ray characterization; peaks marked * correspond to unreacted PtCl2 adhering to as-grown crystals; c) focused ion beam sculpted sample for precise transport measurements. — Fig. 1: a) As-grown PtCoO₂ single crystals; b) x-ray characterization; peaks marked * correspond to unreacted PtCl₂ adhering to as-grown crystals; c) focused ion beam sculpted sample for precise transport measurements.

Fig. 1: a) As-grown PtCoO₂ single crystals; b) x-ray characterization; peaks marked * correspond to unreacted PtCl₂ adhering to as-grown crystals; c) focused ion beam sculpted sample for precise transport measurements.

The delafossite structural series of general formula ABO₂ features a wide range of physical properties including transparent conductors, candidate magneto- and thermoelectrics and magnetic insulators. Arguably less well-known are an astonishing series of metals such as PdCoO₂, PtCoO₂ and PdCrO₂. Quasi-two-dimensional and with conduction taking place in planes of triangular lattice noble metals, they have both the highest known conductivity. Our work has established that, at low temperatures, resisitivities of only a few nWcm, corresponding to mean free paths as high as 50 μm, can be obtained in PdCoO₂. The microscopic origins of these record-breaking electrical properties are not well understood.

Although first synthesized over forty years ago, full appreciation of the properties and potential of the delafossite metals has been reached only relatively recently. To put their purity levels in context, a 50 μm mean free path in the impurity scattering limit would correspond to there being only one defect approximately every 200000 lattice spacings. For this to be observed in melt-grown single crystals implies either unprecedented chemical perfection of the noble metal conducting layers or some suppression of electron scattering.

After becoming interested in this problem five years ago following discussions with our collaborators in the group of Yoshi Maeno at Kyoto University, we performed a series of de Haas-van Alphen effect and transport studies on PdCoO₂ and PdCrO₂, confirming the extremely high purity of PdCoO₂ [1] and the existence of weak coupling between Cr³⁺ spins and Pd conduction electrons in PdCrO₂ [2]. We then embarked on a programme of in-house crystal growth, succeeding in growing the first ‘large’ single crystals of PtCoO₂ (sub-mm plates a few μm thick). Using modern focused ion beam technology of the kind now set up in our department at the Institute, we cut samples such as the one shown in Fig. 1c for precise transport measurements. These revealed that the room temperature resistivity of PtCoO₂ is only 2.1 μΩcm, corresponding to a room temperature conductivity per carrier nearly double that of elemental copper.

In collaboration with the group of Dr. Phil King at the University of St Andrews, we performed angle-resolved photoemission spectroscopy of the new PtCoO₂ crystals [3], and in-house torque magnetization measurements of their quantum oscillation spectrum. These measurements revealed the nearly hexagonal Fermi surface shown in Fig. 2, and the rather astonishing dispersions shown in Fig. 3. Although the Fermi surface is rather faceted, strongly suggesting that it originates from Pt 5d orbitals, the measured Fermi velocities are close to the free electron value, and the single conduction band shows no broadening even to 0.5 eV below the Fermi level. This is qualitatively in agreement with the previous transport work on the other delafossite metals, in that it also suggests anomalously weak scattering.

Fig. 2: Raw ARPES data showing the single faceted Fermi surface of PtCoO2 — Fig. 2: Raw ARPES data showing the single faceted Fermi surface of PtCoO₂

Fig. 2: Raw ARPES data showing the single faceted Fermi surface of PtCoO₂

Fig. 3: a) The resolution-limited quasiparticle dispersion of PtCoO2 extending at least 500 meV below the Fermi level; b) Expanded data along two cuts in the Brillouin zone showing the extremely large Fermi velocity, which is consistent with independent quantum oscillation measurements. — Fig. 3: a) The resolution-limited quasiparticle dispersion of PtCoO₂ extending at least 500 meV below the Fermi level; b) Expanded data along two cuts in the Brillouin zone showing the extremely large Fermi velocity, which is consistent with independent quantum oscillation measurements.

Fig. 3: a) The resolution-limited quasiparticle dispersion of PtCoO₂ extending at least 500 meV below the Fermi level; b) Expanded data along two cuts in the Brillouin zone showing the extremely large Fermi velocity, which is consistent with independent quantum oscillation measurements.

Our work on these intriguing metals continues on several fronts. Firstly, we will extend our photoemission studies, checking for orbital character and possible spin texture of the states near the Fermi level, and performing a careful comparison of bulk states with various kinds of surface state that we believe will exist. At all stages the experiments will interface closely with electronic structure calculations. Secondly, interesting avenues of research open if one takes the very low observed scattering rates at face value, without microscopic understanding, and investigates their consequences. In that spirit we decided to investigate taking PdCoO₂ into the mesoscopic regime in which we constrained conduction channel widths to much less than the mean free path, and obtained, for the first time, evidence for hydrodynamic-like electronic transport in a bulk material [4]. Thirdly, we are actively attempting crystal growth of other delafossites that are likely to be metallic, but have never previously been synthesized in single crystal form. So little is understood about this fascinating family of materials that we anticipate a major on-going research programme investigating them. Our primary goals in doing so are to obtain fundamental understanding, but we also note that they have the potential for epitaxial growth both in combination with each other and with the large number of related non-metallic compounds; multilayer thin films have the promise to yield fascinating properties.