Two approaches for addressing electrochemical electrode arrays with reduced external connections

Abstract

Although patterning hundreds or thousands of electrochemical electrodes on lab-on-a-chip devices is straightforward and cost-effective using photolithography, easily making connections between hundreds of electrodes and external amplifiers remains a bottleneck. Here we describe two electrode addressing approaches using multiple fluid compartments that can potentially reduce the number of external connections by ~100-fold. The first approach enables all compartments on the device to be filled with solution at the same time, and then each fluid compartment is sequentially electrically activated to make the measurements. The second approach achieves lower measurement noise by sequentially filling recording chambers with solution. We propose an equivalent circuit to explain measurement noise in these recording configurations and demonstrate application of the approaches to measure quantal exocytosis from individual cells. A principle advantage of using these approaches is that they reduce the fraction of the microchip area that needs to be dedicated to making external connections and therefore reduces the cost per working electrode.

Fig. 1

Prototype multiplexed device with 6 fluid compartments and 16 conductive paths. (a) Photo of prototype device. (b) Schematic illustration of Method 1 where the active fluid compartment is selected by switching the potential of the bath solution. (c) Schematic illustration of Method 2 where the active compartment is selected by filling it with electrolyte solution and connecting it to a reference electrodes. Compartments are inactivated by removing the solution. (d) Sample Cyclic Voltammetry traces recorded using Method 1. Electrodes in electrolyte filled compartments (Comp 2 and Comp 3) have no faradaic current whereas electrodes in the solutions containing FCA (Comp 1 and Comp 4) exhibit faradaic currents due to oxidation of FCA.

Fig. 2

Chronoamperometry calibration curves demonstrate a lack of crosstalk between compartments with Method 1. (a) Sample chronoamperometry traces for electrodes in adjacent fluid-filled compartments with or without FCA. (b) Plot of steady-state current versus concentration of FCA in either the FCA-containing compartment (triangles: Method 1, circles: Method 2) or the adjacent compartment without FCA (squares: Method 1). The best-fit line has a slope of either 1.81 nA/mM (Method 1) or 1.77 nA/mM (Method 2). The asterisks represent the expected diffusion-limited current obtained from COMSOL simulations.

Fig. 3

Adding fluid to additional compartments leads to proportional increases in capacitance and noise whether or not the additional compartments are grounded. (a) Schematic of approach with additional compartments grounded or (b) Floating (c) The capacitance and noise power spectral density (S_I, measured at 1000 Hz) increase linearly with the number of compartments containing fluid. The bottom chart plots S_I normalized to capacitance to demonstrate that they increase in parallel upon fluid addition. Error bars are standard errors from the measurements on n=7 floating electrodes and n=5 grounded electrodes.

Fig. 4

“Floating” fluid-filled compartments provide a pathway to ground through other electrochemical electrodes in the array. (a) Schematic illustration of multiple pathways to ground. (b) Equivalent circuit representation.

Fig. 5

The noise increase that results upon addition of fluid to multiple compartments can be reversed by removing the fluid. Following removal of electrolyte solution, each well was washed with distilled water and dried with compressed air. (a) Variance of the current increases with solution addition in more compartments and decreases with solution removal. (b) Current noise level from the working electrode after solution addition and removal remains similar to original noise level.

Fig. 6

Both multiplexing approaches can resolve amperometric spikes due to quantal exocytosis from individual chromaffin cells. The four traces in each column result from sequential addressing of an electrode in each of four fluid compartments. Method 1 recordings were acquired by loading all compartments with cells and solution simultaneously. Method 2 recordings were acquired by addressing the compartments by sequentially filling with cells and solution followed by removing fluid from the previously used compartment.