Quantum microscopy in flatland

A new two-dimensional host crystal for quantum microscopy could spur advances for the technique, alongside new challenges. We demonstrate the first widefield quantum microscopy with a van der Waals material, exploiting one apparent limitation to highlight a key benefit over existing platforms.
Quantum microscopy in flatland

Our group is interested in applying the tools, techniques and systems of quantum physics to construct new types of microscopes. Where a conventional microscope captures the light reflected from a sample, a 'quantum microscope' uses an array of quantum systems embedded in a crystalline host to map quantities such as a sample's temperature distribution. In our recent work we explored a fundamentally new type of host - one that is intrinsically two-dimensional, which allows for a new range of possible applications.

At the heart of a quantum microscope is a crystal containing atomic 'defects' (such as a missing atom) that behave as tiny collections of electrons with strictly defined (quantised) energy levels. We illuminate these defects with green laser light, causing them to re-emit red light that can be used to infer their electronic energy levels. In turn, these energy levels can be related to the local environment, whether that's an intrinsic property, such as strain in the crystal lattice, or extrinsic, such as the magnetic field produced by a nearby magnet or electric current. These relations between energy levels and environmental perturbations are known and precise, allowing these quantum sensors to operate as calibration-free sensors, making them fundamentally different from their traditional ('classical') counterparts.

Comparison of diamond and hBN as quantum microscopy host crystals
Diamond (blue, left) is the predominant crystal used in quantum microscopy, however its bulky three-dimensional character increases sample-sensor standoff, reducing spatial resolution and magnetic signal (purple lines and blue/red gradient) from the sample (orange). We explored hexagonal boron nitride (hBN; right, green) as an alternative crystal that is intrinsically two-dimensional.

Until recently these defects have been hosted in rigid three-dimensional crystals such as diamond, which do a good job of protecting the system from undesired disturbances, but are inflexible and thus difficult to interface with samples we want to measure. Promisingly, our coauthors at the University of Technology Sydney have identified a quantum defect in a two-dimensional or 'van der Waals' material, hexagonal boron nitride (hBN). Van der Waals materials are comprised of layers, where each sheet is covalantly bonded internally, but with only the weak van der Waals force holding adjacent layers together - the most famous example is graphene, made out of graphite. This layered property allows the crystals to be made very thin or even isolated as 'monolayers', as well as conform to arbitrarily bumpy surfaces. These properties led us to the idea of using `quantum-active' hBN foils to perform quantum microscopy, with the hope of extending to new areas of application where the smaller sample-sensor standoff would allow for higher spatial resolution, as well as simpler interfacing.

To test the capabilities of our van der Waals quantum microscope, we set about creating a test structure. Our chosen sample was a flake of magnetic chromium ditelluride (CrTe2), itself a van der Waals material and a ferromagnet with a critical temperature just above room temperature; our quantum-active hBN was layered on top to form a van der Waals 'heterostructure'. We were initially disappointed to find no magnetic signal, although indications were that our microscope was working as intended. However we soon realised that the high laser power we were using could be a significant source of heating, enough to heat the CrTe2 above its critical temperature of about 50°C. Upon changing to conditions of less heating, we happily found that our microscope still worked well and we were now able to measure a magnetic signal of the expected magnitude. Furthermore, the shifting of the energy levels could be used to obtain a temperature map simultaneous with our magnetic maps, meaning our microscope can be used to perform correlative imaging between the two quantities - for instance, to work out the temperature dependence of magnetisation. This capability turned out to be a consequence of the van der Waals nature of our host: because it is so thin, not much heat is able to escape through the sensor and the temperature profile remains undisturbed. If we repeated the experiment with a thicker three-dimensional host such as diamond, a lot of heat would dissipate through it and we would not be able to capture a meaningful temperature map. What began as an experimental annoyance ended up being a hint towards a new capability for quantum microscopy, unique to this form of host.

Simultaneous magnetic and temperature mapping of chrome ditelluride flakes
(a) Photograph of imaging structure. Chromium ditelluride (CrTe2) magnetic sample is dark, with the green hexagonal boron nitride (hBN) placed on top. (b,c) A temperature (b) and magnetic field (c) map of the CrTe2 flake under minimal heating conditions. (d) Schematic of the flake magnetization under minimal heating conditions. (e,f,g) As (b,c,d) but under moderate heating conditions. The magnetic signal is killed as the sample is heated by roughly 50°C; above its critical temperature.

Alongside imaging magnetic materials, mapping current flow in electrical devices is a major application of quantum microscopy. In this modality, the imaged magnetic field is generated by moving charge carriers (electrons), and thus a reconstructed current density map can be formed. To test this technique we imaged current flowing in a graphene device; hBN is routinely used in such van der Waals heterostructures as a protective insulating layer and thus the construction of the device was trivial. This simple sample interfacing contrasts with microscopes that use a three-dimensional sensor where fabrication of the device would be far more challenging. We were additionally able to detect an increase in temperature over the graphene device caused by Joule heating from the electrical current.

The sum of our results suggests that our van der Waals quantum microscope is an attractive new measurement platform. The hBN host can interface with arbitrarily rough topologies, is already commonly integrated in van der Waals heterostructures, and is capable of unique multi-modal measurements. We are optimistic that many of the lessons learned in the wider quantum sensing community over past few decades can extend to this new system and lead to an even broader range of capabilities, facilitating studies in a number of scientific fields in the future.

Please sign in or register for FREE

If you are a registered user on Research Communities by Springer Nature, please sign in