PolyMEMS INAOE, a Surface Micromachining Fabrication Module and the Developm...
SWANG_CSNE final poster
1. Corrosion Analysis
• MATLAB Image Processing Toolbox
• k-means image segmentation
• One-way ANOVA
Results
Product
Testing
Flexible Glassy Carbon Microelectrode Arrays for
Chronic µECoG & Intracortical Recording & Stimulation
Sarah Wang1, Maria Vomero2, Noah Goshi2, Calogero Gueli3, and Sam Kassenge2
1Harvey Mudd College 2Mechanical Engineering, San Diego State University 3Microsystems Engineering, University of Freiburg
Chelsea Sabrina
Dimensions 4.2 mm x 7.4 mm 4.2 mm x 8.4 mm
Thickness 50 µm 15 – 50 µm
Channels 6 12
Diameter 60 µm 60 µm
Figure 6. Devices at the end of production. (a) Devices ready
to be released from the wafer. (b) Devices placed in 1% agar to
demonstrate intended configuration in the brain.
Dry Aging Stimulation
• Electrochemical Impedance
Spectroscopy (EIS)
• Charge Storage Capacity (CSC)
• 1 week in 30mM H2O2 @ 37°C
to mimic immune response
• 1,500 cycles of cyclic
voltammetry (CV) to “stress”
electrodes after aging
Motivation
A new class of neural microelectrode arrays (MEA) fabricated using photo-
patternable polyimide enable:
• High-volume fabrication;
• Small, thin, and flexible features.
These advantages offer increased biocompatibility and creative MEA designs, e.g.
combined electrocorticography (ECoG) and intracortical recording capabilities.
Glassy carbon (GC) is biocompatible, electrochemically inert material with
• tunable mechanical and electrical properties
• high charge storage capacity (CSC), and
• high corrosion resistance,
These motivate the use of GC to sense and stimulate brain activity. It is also photo-
patternable, complementing the process of fabricating flexible MEA.
Two novel designs of flexible GC neural MEA were developed:
(1) penetrating design for intracortical signal recording;
(2) surface + penetrating design for combined µECoG and intracortical recording.
Design & Fabrication
(a) (b) (a) (b)
References
(1) S. Kassegne, M. Vomero, R. Gavuglio, M. Hirabayashi, E. Özyilmaz, S. Nguyen, J. Rodriguez, E.
Özyilmaz, P. van Niekerk, A. Khosla, Electrical impedance, electrochemistry, mechanical stiffness, and
hardness tunability in glassy carbon MEMS µECoG electrodes, Microelectronic Engineering (133) 36-44.
(2) M. Vomero, P. van Niekerk, V. Nguyen, N. Gong, M. Hirabayashi, A. Cinopri, K. Logan, A. Moghadasi, P.
Varm, S. Kassegne, A novel pattern transfer technique for mounting glassy carbon microelectrodes on
polymeric flexible substrates, Journal of Micromechanics and Microengineering (26): 025018.
(3) N. Goshi, M. Vomero, I. Dryg, S. Seidman, S. Kassegne, Modeling and Characterization of Tissue/
Electrode Interface in Capacitive µECoG Glassy Carbon Electrodes, ECS Trans. 2016 72 (1): 83-90.
Figure 5. SEM images of the penetrating
portion of Chelsea. (a) Pt electrodes. (b)
GC electrodes.
(a) (b)
Figure 4. Chelsea and Sabrina specifications.
Figure 2. GC fabrication. (a) GC is made from
pyrolyzed SU-8 10 negative resist. (b) A comparison of
platinum electrodes (top) and GC electrodes (bottom).
Figure 9. Porosity calculation via k-means image
segmentation. The image is first converted to greyscale and
filtered to remove noise. Then it is segmented into three
colors, with the darkest two representing corrosion.
Figure 9. Corrosion in Pt. (a) Surface of dry
electrode. (b) Surface after aging and stim-
ulation, where a large portion has been eroded
away.
(a) (b)
Table 2. Results of image analysis. GC does
not exhibit much corrosion for reasonable sig.
levels, e.g. α = 0.05, whereas Pt experiences
more corrosion.
Conclusions
Two novel neural MEAs were successfully designed and fabricated within the given
guidelines:
• Chelsea: penetrating MEA for intracortical signal recording and stimulation
• Sabrina: combined surface and penetrating MEA for µECoG and intracortical recording.
These flexible, implantable, minimally-invasive designs will enable doctors and
researchers to chronically acquire neural information.
Further testing demonstrates the advantages of glassy carbon over platinum.
• Higher charge storage capacity, larger water window, and lower impedance enables
more reliable electrical measurements
• Resistance to H2O2 exposure and electrical stimulation ensures safety and robustness in
long-term applications
Acknowledgements
This project is supported by NSF Grant No. EEC-1028725 under ERC program. The author
would like to thank her mentor Maria Vomero, her NeuroMEMS colleagues, Dr. Sam
Kassegne, and CSNE for enabling her research experience. The NeuroMEMS team would
also like to acknowledge its collaborators at the University of Washington and the
University of Ferrara.
Figure 7. EIS of dry GC and Pt electrodes.
n = 18 for both groups. GC demonstrates low-
er impedance within the applicable frequency
range
GC Pt
CSC (mC/cm2) 26.6 ± 8.4 12.8 ± 1.6
Water window -0.9 to 1.1 V -0.8 to 0.8 V
Figure 1. Mask designs.
(a) Penetrating only. (b) Surface
and penetrating.
Figure 3. Lithography and pyrolysis process for fabricating GC on flexible polyimide substrate.
Table 1. Comparison of electrochemical properties of the
different electrodes.
Figure 8. CV of Pt and GC demonstrating the wider
electrochemical or water window of GC.
GC Pt
% n % n
Dry 30.7 ± 4.1 16 29.4 ± 2.3 5
Aged 30.5 ± 5.3 12 39.8 ± 1.7 9
Aged + Stim 33.9 ± 2.1 12 40.0 ± 9.3 10
ANOVA p = .075 p = .011
(a)
µECoG
intra-
cortical
1% agar
(b)
Electrochemical Characterization
GC window: -0.9 to 1.1 V
Pt window: -0.8 to 0.8 V