Development of 2D quantum materials pipelines (2D-QMaPs)

The MonArk Quantum Foundry is developing two prototypes of a 2D quantum material pipeline (2D-QMaP).  One 2D-QMaP will be at Montana State University and one will be at the University of Arkansas. The 2D-QMaPs will be developed to provide automated capabilities that unite and accelerate the three primary stages of sample preparation and device fabrication for 2D quantum materials research:

1.   Robotic exfoliation, flake hunting, and stacking: The 2D-QMaPs will automate these tasks using a robotic exfoliating roller, a computer-controlled microscope, layered material optical contrast libraries, and a robotic flake pick-up and stacking apparatus. These processes will build upon prior work and open source resources [1].

2.   Device fabrication and packaging: Samples will then be transferred for device fabrication, without removal to air, to a high-resolution (<20 nm) thermal scanning probe lithography system with a laser writer for high throughput that is also integrated with an electron beam evaporator. Within the same inert environment, samples will then be packaged for (opto)electronic measurement.

3.   Characterization: Finally, quantum device characterization tools will be integrated with Stages 1 and 2 to generate detailed reports of the material and device properties of virtually any 2D heterostructure.  Capabilities will include µPL and µRaman, magneto-optical probes, nano-optical spectroscopy, bulk magnetometry, nanoscale scanning magnetometry, and magneto-transport and qubit characterization with supporting instruments to extract mobilities, transition temperatures, photon autocorrelations, coherence/decay times, cavity coupling/decay rates, gate fidelities, and more.

Two key principles guide the design and development of the 2D-QMaPs. First, as many functionalities and capabilities as possible will be integrated into an interconnected glovebox system to preserve 2D quantum materials by avoiding parasitic exposure to water and oxygen. When systems cannot be integrated into the glovebox system, vacuum suitcases will be used to facilitate sample transfer. Second, particularly at the exfoliation and fabrications stages, varying degrees of robotic and machine vision automation will be strategically employed to perform the most time-consuming and unrepeatable tasks. Automation will be crucial for the acquisition of datasets across a large parameter space with which machine learning algorithms will be trained.

References

[1] Masubuchi, S. et al. Autonomous robotic searching and assembly of two-dimensional crystals to build van der Waals superlattices. Nature Communications 9, 1413, doi:10.1038/s41467-018-03723-w (2018).