Effective determination of food contaminations such as pathogens (Escherichia coli, E. coli) and heavy metal ions (Pb²⁺, Zn²⁺, etc.) is important in the field of food safety and public health. Recently, sensing probe-modified graphene field-effect transistors (G-FETs) have been widely used for highly sensitive and selective detection of food contaminations. Understanding the interaction between the sensing probe and the target, interfacial carrier generation and transport in the target/sensing probe-modified-graphene system are the keys to improving sensing performance of G-FETs biosensors. This thesis aims to design and fabricate antibody modified-G-FETs (ABG-FETs) and aptamer modified G-FETs (APG-FETs) for E. coli detection, and to design new sensing probes for Pb²⁺ and Zn²⁺ detection. For G-FETs biosensors, a theoretical model is used to correlate charged target-induced carriers on nanodevices to explain electrical responses. Based on the model, the performance of biosensors can be optimized by controlling the Debye length, the sensing probes, and the carrier mobility of G-FETs. The main contributions of this thesis are summarized as follow:

First, functionalization processes and their influence on carrier mobility of G-FETs were investigated. Aromatic molecule (1-pyrenebutanoic acid succinimidyl ester, PBASE) acting as a linker was used to anchor sensing probes onto graphene surface in this study. Dimethyl formamide (DMF) and methanol (CH₃OH) were used as two solvents to dissolve the PBASE. The PBASE was stably immobilized on the graphene surface, which was confirmed by Raman microscopy. Electrical measurements and Fermi level shift analysis further revealed that the PBASE imposed a p-doping effect while DMF and CH₃OH imposed an n-doping effect. More importantly, CH₃OH caused a smaller reduction in the carrier mobility of G-FETs (from 1095.6 cm²/V·s to 802.4 cm²/V·s) than DMF (from 1640.4 cm²/V·s to 5.0 cm²/V·s). Therefore, CH₃OH was used as the solvent for linker functionalization in the following experiments.

Second, two types of sensing probes, namely antibody and aptamer, were employed to achieve specific detection of E. coli. Quantitative analysis of the carrier density induced by the bacteria was explored. For antibody-modified G-FETs (ABG-FETs), the average estimated number of holes induced by per E. coli K12 was 1488. While for aptamer-modified G-FETs (APG-FETs), the average estimated number of holes induced by per E. coli 8739 was 2042. Besides, Debye length was identified as another important parameter to affect the number of E. coli-induced carriers. Phosphate buffered saline (PBS) was used as the electrolyte to provide a liquid environment for the detection of E. coli. The concentration of the PBS affected the ionic strength which was related to the Debye length. High concentration of PBS (10 mM PBS) yielded a Debye length of 0.7 nm. When the Debye length was smaller than the height of sensing probes, no carriers were induced in G-FETs due to the electric-field screening effect. Systematic studies were conducted and the results demonstrated that the optimal concentration of PBS was 0.1 mM PBS. Enhanced efficiency was found in the shorter height of the aptamer, which facilitated the carrier generation process in G-FETs. Moreover, the aptamer possessed more advantages such as simple modification process, stable under harsh conditions, and high affinity toward targets.

Third, the influence of gate voltage-dependent carrier mobility of G-FETs on the bio-detection signal was investigated to provide better electrical response and lower detection limit. Higher carrier mobility suggests that the E. coli induced-carriers in graphene can be collected efficiently. The carrier mobility of G-FETs can be tuned by the gate voltage. Therefore, by optimizing the gate voltage, the carrier mobility can be maximized accordingly. For both ABG-FETs and APG-FETs biosensors, we experimentally demonstrated that the lowest detection limit was obtained at the optimized gate voltage that resulted in highest carrier mobility. The detection limit of 10² CFU/mL was achieved at the gate voltage of 75 mV by using the ABG-FETs and 40 mV by using the APG-FETs. Furthermore, the selectivity of the ABG-FET and the APG-FET was experimentally verified. The experimental results indicated that either the ABG-FETs or the APG-FETs exhibited excellent selectivity for E. coli over other bacteria strains.

Fourth, the operation of the sensing mechanism of G-FET biosensors were modeled and simulated with COMSOL Multiphysics. The motion of the negatively charged E. coli particles in solution at different time, and the surface charge of graphene induced by charged E. coli were systematically studied. Base on the simulation results, the relationship between source-drain current, graphene-bacteria distance, and bacterial concentration are established, in which the graphene-bacteria distance may play a key role in improving the sensing performance of the G-FET biosensors. We believe that the simulation of the G-FET biosensors could serve as a guideline for the design and optimization of G-FETs biosensors in the near future.

Finally, two aromatic molecules (T1 and T2), acting as the sensing probes, were developed for Pb²⁺ and Zn²⁺ detection with high sensitivity and selectivity. Coumarin was used as the fluorophore and 8-hydroxyquinoline (8-HQ) was used as the metal-ions receptor. Schiff base (C=N) and vinyl group (C=C) were employed as a spacer to connect coumarin and quinoline in T1 and T2, respectively. The sensing probe T1 exhibited a fluorescent response to Pb²⁺ with a visually detectable color change from colorless to yellow, and 30-fold fluorescence enhancement. The binding of Pb²⁺ with T1 inhibited photo-induced electron transfer (PET) process, resulting in an increase of fluorescence intensity. T1 also displayed a high affinity towards Pb2+ ions with a dissociation constant (K_d) of 0.1 μM and possessed a high selectivity for Pb²⁺ ions. T2 showed a selective colorimetric and ratiometric response towards Zn2+ with a visually detectable color change from yellow to pale red in absorption, and from green to yellow in emission. The detection limit for Zn²⁺ was estimated at 48.1 nM, which was below the threshold limit value of Zn²⁺ in drinking water (46 µM). Moreover, T2 displayed high potential for Zn²⁺ monitoring in living cells and in real water samples. This study provides a simple approach to monitoring the metal ions in the environment base on the change of the fluorescent signal.

In summary, we developed several sensors for food contaminations (bacteria and heavy metal ions) detection with high sensitivity and selectivity. The antibody and aptamer modified G-FETs biosensors were fabricated for E. coli detection. The sensing probes based on aromatic molecules served as the fluorescent probes for heavy metal ions detection. The effects of the gate voltage and the electrolyte with different ionic strengths on the sensing performance of G-FETs-based biosensors were systematically investigated. The operation and the sensing mechanism of the G-FET biosensors for bacterial detection were modeled and simulated to gain a better understanding of the design and the optimization of the graphene biosensors. We believe that our study provides a comprehensive understanding and consideration to develop new biosensors for food safety monitoring with high performance.