This blog post was jointly prepared by Ahmet Avsar, Michele Pizzochero and Alberto Ciarrocchi.
Fifteen years after the discovery of graphene, the family of 2D crystals has tremendously expanded and currently comprises a broad range of materials, including metals, semi-metals, semiconductors, trivial and topological insulators. Furthermore, many electron-correlation effects like superconductivity, Mott transitions and magnetic instabilities have emerged in some of these low-dimensional systems. The recently-discovered intrinsically-magnetic 2D crystals (e.g. CrI3 and Fe3GeTe2) have attracted substantial attention for their unique properties including layer-dependent magnetism and its electric field modulation. However, it is not very surprising that such systems present a magnetic response down to the atomically-thin limit, since their bulk parent materials were known to be magnetic as well. Unfortunately, the experimental isolation of 2D magnets has been held back by their extreme sensitivity to ambient conditions. In fact, even after a tedious encapsulation process in the inert environment of a glove-box, only short-term stability can be achieved. This restricts the flexibility of device architectures, slows down experimental progress, and casts doubts over their immediate integration into spin-based electronic devices.
Defect engineering has been considered a promising approach to introduce magnetism in otherwise non-magnetic materials, and several theoretical studies identified many types of magnetic defects in transition metal dichalcogenides. Specifically, earlier first-principles investigations predicted that PtSe2 is more prone to hosting metal atom vacancies as compared to conventional group VI dichalcogenides (e.g., MoS2), and that these defects act as magnetic centers. On the experimental side, STM measurements performed on as-grown PtSe2 under selenium-rich environments revealed the existence of a high-concentration of Pt vacancies.
This observation has driven our curiosity towards the magneto-transport response of this air-stable material. Since in the laboratory of Prof. Andras Kis we were already familiar with the fabrication of ultra-thin PtSe2-based devices, it did not take long before the first set of devices was characterized, which unexpectedly revealed the presence of antiferromagnetic ground state orderings. These findings were surprising, as pristine PtSe2 is a non-magnetic material. We immediately focused on measuring additional devices in a wider thickness range and experimentally demonstrated the presence of ferromagnetic ground state orderings as well. Thanks to the collaboration with the group of Prof. Oleg Yazyev, we have provided a theoretical picture able to capture the observed effects. Specifically, we suggest that such a competition between ferro- and antiferro-magnetic ground states stems from the RKKY exchange couplings between magnetic Pt vacancies located at the surface of metallic films of PtSe2.
While reducing defect concentration is vital for enhancing device performance, our work shows that the presence of impurities is not always detrimental: indeed, they can be exploited to tailor the magnetic properties of air-stable materials. Since such growth-related vacancies are ubiquitous in many 2D materials, our findings expand the range of 2D magnets into stable materials that would normally be overlooked. We suggest that the currently observed small critical temperatures may be improved by either controlling the defect concentration in PtSe2 or investigating the effect in other 2D material systems.
The detailed study has been published and can be found in the following Nature Nanotechnology letter: https://www.nature.com/articles/s41565-019-0467-1