My application: I am attempting to create a device to map redox potential in soils at ~1 mm resolution, using a grid of platinum electrodes.
Background: Redox potential is a measure of the "electron pressure" created by the reduced species of chemicals in the soil, and is a general measure of how anaerobic a soil is. It is typically done in the following setup, illustrated here: A platinum electrode is inserted in the soil, as is a reference electrode that uses the Ag/Ag2+ redox couple and which completes a salt bridge with the soil. Both are connected to a high-impedance (~10 GOhm) voltmeter. The high impedance of the voltmeter is very important to the whole setup, because the charge density in the soil is very small and the kinetics of oxidation are slow, so any electron flow from the soil to the reference electrode will cause the voltage to drift. Typical values of redox potential range from +700 mV to -400 mV, and they are essentially DC, changing only over the course of hours.
This setup is all well and good for taking single point measurements in the soil, but I want to be able to map redox potential over a 5 x 5 cm grid (at a resolution of ~1 mm, so 2500 electrodes total, lower resolution acceptable , say 2mm, if 1mm^2 can't be achieved) to explore how heterogeneous redox conditions are over short spatial scales. Because there are so many electrodes, I want to take reasonably fast measurements, say 1 per second.
My approach: I've landed on the idea of creating a PCB that has copper pads on one side, and then electroplating the pads with 100 um of platinum, in order to create the grid of electrodes that I can insert into soil. The electrodes would each be connected to the reverse side of the PCB with a via, where I want to implement switching circuitry to route signal to the high-impedance voltmeter, one pixel at a time. (I really want to avoid running 2500 wires to another circuit board for the switching, as I want to use potting compound to insulate the circuit from soil water.)
My questions: I'm stuck on how to implement the switching circuitry in a compact fashion that will allow sufficient electrode density. For true microelectrode arrays (e.g. used in neurophysiology), hundreds of electrodes are able to be placed into a square millimeter because they put a 3T CMOS switch array on an IC die. I think that would be the best sort of architecture for this project, where an entire row of electrodes is switched on at one time, with all the electrodes in each column connected to their own bus that goes to ADC circuitry (see Figure 2 on this page to see what I'm talking about) . But the comparatively large size of the electrode array in my case makes this impractical, and anyway I completely lack the money or the knowledge to be able to create my own IC die. (I'm a soil scientist.)
- Is there a way that I can implement this sort of switching system using off-the-shelf SMD devices, allowing me to achieve high electrode density? (I would pay to have the boards pick-and-place fabricated.)
- What sort of switch design would be appropriate for my purposes? To my mind, the key design considerations are ultra-low leakage current, both on and off, low stray capacitance (so that the voltage doesn't take a long time to stabilize), and a small footprint. Are there other important design parameters I should consider?
- Does anyone have particular components in mind they could suggest?
- Am I approaching this project completely wrong and there's a much better way to do what I want?
Thanks for your help!
