Optimized Fabrication Procedure for High-Quality Graphene-based Moiré Superlattice Devices

Our research is focused on the quantum electronic properties of two-dimensional materials and their heterostructures, aiming to discover and investigate new quantum phenomena and develop novel electronic devices. We pioneered the field of twist turnings, where the properties of two-dimensional materials' heterostructures change as a function of the angle between the layers. The most impactful discovery was superconnectivity in magic angiography.

Achieving a precise twist angle remains challenging due to heater strain disorder and lattice relaxation introduced during the nanofabrication process, which often leads to devising homogeneity and limit reproducibility across samples. This protocol reflects our experience optimizing each fabrication step to improve uniformity and yield, enabling more researchers to reliably build high quality graphene moire devices, and accelerate progress in this field.

[Narrator] To begin, place a graphene wafer on the sample stage of the transfer setup, integrated with a super continuum laser. Use peek through to make the Inkscape window semi-transparent while floating above the camera window. Use the draw BZA curves and straight lines function in Inkscape to outline the graphene boundary and the planned laser cutting lines. Now power on the laser and increase the intensity until the beam spot is visible. Adjust the sample stage to align the beam spot with the start of a planned laser cutting line. Next, move the sample stage to guide the beam spot along the planned laser cutting line. After zeroing the laser intensity, inspect the laser cutting line. Repeat the steps until all planned lines are laser cut. To pick up the top hexagonal boron nitrite flake, adjust the orientation of the flake until the straight edge is vertical and positioned on the right side. Gently clamp the glass slide with the PC and PDMS stamp inside the socket at a downward tilt angle of two to three degrees. Slide the socket to the left using the sliding tray until the glass slide is positioned above the sample wafer, then manually lock it in place. After setting the sample stage temperature to 50 degrees Celsius, engage the glass slide downward using the Z direction actuator. Focus the microscope on the polycarbonate film and inspect the surface cleanness. Move the hexagonal boron nitride flake to the selected clean region and engage the slide further downward. Now engage the stamp onto the wafer with a speed of five micrometers per second, until it completely covers the hexagonal boron nitride flake, then set the sample stage temperature to 80 degrees Celsius. After the sample stage temperature reaches 80 degrees Celsius, disengage the stamp with five micrometers per second until the wavefront is close to the hexagonal boron nitrite straight edge, then pick up the flake with a reduced speed of two micrometers per second. Once picked up, increase the speed to five micrometers per second to detach the polycarbonate film from the wafer. Switch off the stage heater and open the water cooling system. When the temperature drops below 45 degrees Celsius, close the water cooling system. Disengage the glass slide all the way up at one millimeter per second. Place the wafer with the freshly laser cut graphene on the sample stage, and locate the graphene using the 10X objective. Adjust the flake orientation so the laser cutting lines are parallel to the hexagonal boron nitride straight edge. Outline the graphene boundary in Inkscape, including the laser cutting lines. Now move the glass slide with the top hexagonal boron nitride flake to the pre-engaged position with the PC and PDMS stamp two millimeters above the graphene wafer. Align the flake straight edge with the center of the laser cut line to match the hexagonal boron nitride with the graphene drawing. Tune the microscope to focus on the graphene and align it to its Inkscape drawing. Then focus on the top flake and move the glass slide in XY direction to align it to its drawing. After refocusing on a focal plane slightly higher than the precut graphene, engage the glass slide downward at 0.1 millimeter per second until the top flake comes into focus. Refocus on the graphene, and set the speed to five micrometers per second. Engage the stamp slowly until the flake becomes roughly visible while ensuring no contact with the wafer. Continue to engage the stamp downward until it contacts the wafer. Engage the stamp at 0.5 micrometers per second until the wavefront is close to the left edge of the flake. Reduce the speed 0.02 micrometers per second, and continue engagement to contact the first graphene piece. Once the flake fully covers the first graphene piece and is pinned at the straight edge, stop the engagement. Clear hysteresis by moving the stage upward at five micrometers per second until the wavefront starts retracting. Set a disengaging speed of 0.02 micrometers per second to pick up the graphene gently. When the wavefront moved away from the flake, increase the speed to five micrometers per second to fully detach the stamp from the wafer. Now copy the entire graphene drawing in Inkscape, and rotate it by the desired twist angle. Align the second graphene piece in the copy drawing with the first piece in the original drawing to maximize overlap. Rotate the sample by the desired twist angle. Align the remaining graphene on the wafer to the copied drawing. Focus on the top hexagonal boron nitride flake on the glass slide and align it to the drawing. After aligning the flakes to the drawings, repeat the demonstrated steps to pick up the remaining graphene. Once picked up, examine the quality of the top stack under the microscope. Place the wafer with the pre-cleaned, bubble-free bottom gate on the sample stage. Outline the boundary of the bottom gate using Inkscape. Align the drawing of the top stack to that of the bottom gate to place the twisted bilayer graphene on the graphite finger. Set the sample stage temperature to 160 degrees Celsius to tear off the polycarbonate film after encapsulation, then load the glass slide with the top stack, and move it to the pre-engaged position. Roughly align both the bottom gate and the top stack to the drawings. Focus on the bottom gate, then slightly raise the focal plane. Identify another middle focal plane between the bottom gate and top stack, then set the speed to five micrometers per second, and engage the stamp until the top stack is visible. Align both to the drawings accurately. When the sample stage is above 150 degrees Celsius, engage the stamp at five micrometers per second until the wavefront is close to the bottom gate, then reduce the speed to 0.5 micrometers per second. Slowly engage the top stack onto the bottom gate. Once the wavefront passes the top stack, increase the speed back to five micrometers per second to fully engage the stamp onto the wafer, then disengage the stamp at five micrometers per second. After fully separating the stamp from the wafer, lift the stamp upward for three seconds at five micrometers per second. Begin moving it in the XY directions to tear off the polycarbonate film. Fully disengage the glass slide at one millimeter per second and remove it from the socket. Switch off the stage heater and activate the water cooling system. Once the temperature drops to room temperature, remove the wafer and inspect under an optical microscope. Fourfold degenerate land out fans emerged from both the charge neutrality point and super lattice gaps in the high quality magic angle twisted bilayer graphene device, breaking down into twofold or onefold degeneracies at higher magnetic fields. Symmetry broken twofold degenerate land out fans were identified at half filling states, indicating correlated insulating behavior. The superconducting dome was located underneath the land out fan around the half filling state. Temperature dependent resistance measurements revealed two superconducting domes at both sides of the half filling state, with the maximum critical temperature reaching approximately 1.7 Kelvin. Among 14 measured devices, the optimal critical temperature peaked at a twist angle near 1.08 degrees, consistent with theoretical predictions. Disordered devices near the magic angle exhibited significantly reduced critical temperatures compared to high quality devices, as seen in their resistance versus temperature profiles. These observations underscore the importance of optimizing the fabrication procedure to achieve high quality graphene-based Moire super lattice devices.