Side-Gate-FET Test Strips for Invasive and Non-Invasive Saliva-Based Glucose Monitoring Using GO-Catalyzed CuO-ZnO Hollow Spheres

Saliva contains diverse diagnostic biomarkers, including glucose, steroid hormones, and HIV antibodies, and provides accuracy comparable to blood tests. Unlike invasive blood tests, saliva sampling is painless, simple, and stress-free process. Yet, no user-friendly saliva-based glucose test is currently available for home use.

Traditional methods, such as Liquid Chromatography-Mass Spectrometry (LC-MS) and Ultraviolet-Visible (UV-Vis) spectroscopy, require laboratory settings. The newer wearable or optical sensors often face challenges, including environmental sensitivity and signal drift.

Recently, field-effect transistor (FET) biosensors have emerged as a promising solution. They detect glucose through electrical changes in a semiconductor channel, offering fast response, high sensitivity, and compatibility with flexible, low-power devices suitable for everyday monitoring.

This paper presents the development of advanced side-gate FET biosensor test strips for both invasive and non-invasive saliva-based glucose monitoring. The core sensing material consists of graphene oxide (GO)–catalyzed copper oxide–zinc oxide (CuO-ZnO) hollow spheres.

The researchers developed glucose-sensing FET biosensors through a simple fabrication process. CuO/ZnO hollow spheres with varying GO concentrations (25–200 mg) were synthesized via a hydrothermal method. Precursors of copper and zinc nitrate were mixed with GO nanosheets and reacted with potassium carbonate. The mixture was heated in an autoclave at 140°C for 12 hours, washed, and dried to produce the final composite. The resulting composites, along with multi-walled carbon nanotubes (MWCNTs), were drop-cast onto flexible substrates to form the sensor’s source and drain regions, and were then finished with a protective Nafion coating. A side-gate electrode was incorporated for signal amplification.

Six sensors were created in total: one using only CNT, one using CuO/ZnO, and four using CuO/ZnO combined with varying GO concentrations. The solution volume, drop-casting speed, and drying temperature were precisely controlled throughout the process to ensure uniform and consistent sensor coatings.

The sensor was integrated into a flexible FET test strip connected to a mobile-enabled IoT device to create a user-friendly system. When saliva contacts the strip, glucose alters the channel current, which is then processed by a microcontroller and transmitted via Wi-Fi or Bluetooth.

The sensing mechanism involves the non-enzymatic oxidation of glucose at CuO sites, which is enhanced by GO and ZnO, thereby altering the conductivity of the FET channel. This change produces a drain current directly proportional to glucose concentration.

The CNT/CuO-ZnO-GO configuration with 100 mg of GO delivered optimal performance, achieving a sensitivity of 1600 µA·mM⁻¹ and a remarkably low limit of detection (LOD) of 0.001 µM. The sensor also exhibited strong linearity (R² > 0.99) across a wide detection range (0–15 mM) in both artificial and real saliva.

Validation with saliva samples from ten fasting individuals showed strong agreement between invasive-mode readings and those from a commercial glucometer. Non-invasive saliva glucose measurements also exhibited a strong correlation with blood glucose levels, confirming saliva as a reliable diagnostic medium.

Overall, this sensor outperforms existing saliva-based glucose sensors, offering the widest detection range, the highest sensitivity, and the lowest detection limit. It opens the door to affordable, accurate, and painless glucose monitoring, especially for diabetic patients.