Extended Data Fig. 9: Soft and stretchable high-density mesh electrode array for tracking neural activities.

a, Schematics showing the electron-beam (e-beam) lithographic fabrication of PFPE-DMA-encapsulated stretchable mesh electronics with a 32-channel high-density electrode array. First, a Ni layer is deposited on a blank silicon oxide wafer as a sacrificial layer. A SU-8 layer is patterned as a spacer, Pt electrodes are photolithographically patterned (step 1). Then, the bottom PFPE-DMA (step 2), Au interconnects (step 3), and top PFPE-DMA layer (step 4) are lithographically patterned. Au layer is patterned by e-beam lithography. PFPE-DMA layers are patterned by photolithography. Zoomed-in images show the details of the electrode arrays (highlighted in blue dashed boxes) and individual electrodes (highlighted in red dashed boxes). b, Photographic images showing the free-floating 32-channel high-density electrode array during stretching, bending, and twisting. c, BF microscopic image showing a representative 128-channel electrode array. d, Zoomed-in view of the red dashed box-highlighted region in (c) showing the high-density electrodes and interconnects. e, Zoomed-in view of the blue dashed box-highlighted region in (d) showing the stretchable design. f, Zoomed-in view of the green dashed box-highlighted region in (e) showing the individual electrodes and interconnects. g, Single-unit waveform centroids (n = 20 neurons) from the continuous 5-day recording of the cyborg axolotl embryo in (Fig. 5). Centroid computed using spatial average across electrode positions weighted by the square of the mean waveform amplitude at each electrode. Grey patterns indicate the positions and sizes of the mesh electrodes. h, Single-unit waveform centroid displacement throughout the 5-day recording in (Fig. 5). Centroids from the same day are labeled with the same color. Centroids for the same units across different days are connected by lines. Grey patterns indicate the positions and sizes of the mesh electrodes. i, Average displacement of single-unit centroids across different days of the 5-day recording in (Fig. 5). Grey contours indicate quintile boundaries of the distribution of centroid position displacement.