Development and characterization of in vivo flexible electrodes compatible with large tissue displacements

Abstract

Electrical activity is the ultimate functional measure of neuronal tissue and recording that activity remains a key technical challenge in neuroscience. The mechanical mismatch between rigid electrodes and compliant brain tissue is a critical limitation in applications where movement is an inherent component. An electrode that permits recording of neural activity, while minimizing tissue disruption, is beneficial for applications that encompass both normal physiological movements and those which require consistent recording during large tissue displacements. In order to test the extreme of this range of movement, flexible electrodes were developed to record activity during and immediately following cortical impact in the rat. Photolithography techniques were used to fabricate flexible electrodes that were readily insertable into the brain using a parylene C base and gold conduction lines and contact pads, permitting custom geometry. We found that this electrode configuration retained mechanical and electrical integrity following both durability studies and large movements within the cortex. This novel flexible electrode configuration provides a novel platform for experimentally examining neuronal activity during a range of brain movements.

Figure 1.

(A) Schematic of the flexible electrode prototype design showing the orientation of the gold leads (not-to-scale). Insertion depth control is accomplished using the fabricated bead and barb. (B) Montage of photomicrographs of a prototype flexible electrode with an insertion section of 0.5 cm past barb. (C) Dark-field images of various features of finalized flexible electrode, including exposed 1 mm1 × mm Au contact pad for pin attachment, bead and barb structural features for controlled insertion depth, and exposed Au contact sites for 100 μm and 200 μm wide electrodes. Contact sites are 40 μm in diameter; all scale bars are 100 μm.

Figure 2.

(A) Fabrication process: 15 μm parylene deposited on glass substrate. AZ 4620 positive photoresist is patterned for electrode lines. (B) Gold is vertically deposited using the CVC E-Beam Evaporator. (C) Acetone wash is used for lift-off processing to remove photoresist from the electrode surface. A top layer of parylene is deposited, embedding the gold lines. (D) Photoresist is patterned, and a protective aluminum layer is deposited. Lift-off processing with an acetone wash defines the electrode contact sites and electrode shape. (E) Unprotected parylene is etched with O₂ plasma using the Plasma Therm Reactive Ion Etcher. (F) Protective aluminum layer is wet etched, and mechanical separation releases the electrode from the substrate.

Figure 3.

Schematic outlining the whisker-barrel cortex recording protocol. Whiskers were mechanically stimulated, which elicited a consistent, repeatable neuronal response from layer IV of the barrel field cortex.

Figure 4.

Buckling force tests were performed on five prototype electrodes fixed 2.5 mm from the tip of the electrode. Force loads on electrode tips were measured using an MTS NanoUTM until both visual inspection and measured data confirmed electrode shank buckling. Mean buckling force thresholds of 1970 μN for the 25 μm thick electrodes with a column length of 2.5 mm were measured.

Figure 5.

Average, min and max magnitude of the electrode impedance as a function of frequency prior to, and then following induced bend (n = 7). Featured electrodes are 100 μm wide at the shank with a 40 μm diameter contact site. Mechanical deflection is induced on the electrode shank by an Instron Tensile Testor and consists of a permanent 90° bend 5 mm from the tip of the shank across the smallest cross-sectional dimension. Prior to the bend, the magnitude of impedance at 1 kHz ranges from 44 kΩ to 81 kΩ, with an average of 59 kΩ. Following the bend, the magnitude of impedance at 1 kHz ranges from 86 kΩ to 126 kΩ with an average of 101 kΩ.

Figure 6.

Flexible electrode recording (A) from layer IV pyramidal cells prior to, during and following cortical impact in a representative animal. Injury occurs when time = 0 s. Whisker stimulations prior to injury are indicated by arrows. Red (dark gray) mechanical artifact trace obtained from recording in the barrel cortex in a dead animal. Typical whisker responses are also shown before (B) and following cortical impact (C). Note the different voltage scale. Gains for all recordings shown are the same.