CMOS-Based Ion Image Sensor Enabling pH Measurement Under Light Irradiation

Plant growth is heavily influenced by weather factors like rainfall, temperature, and sunlight, which fluctuate across seasons and years. Traditionally, farmers rely on experience to manage their crops. However, smart agriculture—using information technology in farming—has gained interest in making production more efficient and reliable.

This study developed a minimally invasive pH image sensor for real-time plant monitoring. Traditional pH sensors often produce errors under light exposure, a common challenge in agricultural settings. The proposed sensor overcomes light interference, enhancing measurement accuracy and expanding its potential for agricultural applications.

A single pixel in the charge-transfer pH image sensor, which is the foundation of the sensor developed in this study, has an electrolyte–insulator semiconductor structure with an ion-sensitive membrane. The sensor detects pH using charge-transfer technology by sensing changes in the semiconductor surface potential beneath the sensing area (SA). Variations in the ion-sensitive membrane potential alter the SA’s potential well depth, producing electrical signals that shift with pH.

However, light exposure generates electron-hole pairs in the Si-based sensor, contaminating the signal and causing inaccurate pH readings, making distinguishing between pH and light-induced changes difficult. The study developed a novel approach that leverages the distinct gain characteristics of two sensing pixels to mitigate light contamination.

While both pixels responded similarly to light, their pH sensitivity varied based on the slope of the characteristic curve, which was adjusted by modifying the sensing area. By combining the sensing voltages of the two pixels and using simultaneous equations, changes in pH and light can be measured separately.

The sensor design involved circuit simulations to optimize pixel parameters, creating high-gain Pixel A and low-gain Pixel B with overlapping characteristic curves for method feasibility. The pH image sensor was designed and fabricated using a custom CMOS (Complementary Metal-Oxide-Semiconductor) process. The sensor features a 32x32 array of staggered pixels, integrated vertical and horizontal scanners, readout circuitry, and a silicon nitride (SiN) proton-selective membrane. The fabricated sensors were packaged with epoxy resin for waterproofing.

The sensor was evaluated using solutions of pH 4.01 and 6.86. High-gain Pixel A and low-gain Pixel B showed gains of 0.75 and 0.41 respectively. The sensor was then evaluated under light illumination. The light sensitivities for both pixels were 489 mV/(mW/cm2). Simultaneous equations were formulated using these previously measured pH sensitivities and light sensitivities. A pH measurements under light irradiation were performed using the proposed method. The results showed that the errors were within 10% for both pH and light intensity, thus verifying the effectiveness of the proposed method. Conventional methods under light irradiation resulted in significant errors, highlighting the advantages of the proposed approach.

The study proposed and demonstrated a pH image sensor capable of separately measuring pH and light intensity under light-irradiated conditions for smart agriculture. The sensor can detect various ions (K⁺, Ca₂⁺, Cl^-, NO₃^-) by altering the sensitive membrane, enabling multi-ion sensing for plant growth monitoring in smart agriculture.