Detection of Mercury Ions Using Graphene Nanoribbon-DNA Sensors Fabricated via Template Methods

Abstract

To enhance the sensitivity of graphene-DNA sensors for Hg²⁺ detection, a novel graphene nanoribbon-DNA sensor was fabricated using a template-assisted approach. Silicon nanowires served as templates to decorate the graphene device, followed by plasma etching to delineate graphene nanoribbons. After template removal, the resulting sensors based on silicon nanowire templates were successfully constructed. DNA sequences containing four guanine bases were conjugated with graphene sensors prepared using the templates. The carboxyl groups on the edges of the graphene nanoribbons were activated with EDC/NHS chemistry to facilitate covalent bonding with amino-modified DNA. The kinetic response and Hg²⁺ detection capability of the fabricated sensors were characterized using a semiconductor parameter analyzer. Results indicated that the silicon nanowire-templated graphene nanoribbon sensor exhibited high sensitivity, with a detection limit of 3.62 pM. This innovative approach further improved the sensitivity of graphene-DNA sensors for Hg²⁺ detection.

1. Introduction

A highly toxic heavy metal ion, mercury is widely present in the atmosphere, soil, and aquatic environments. Its toxic nature affects complex biological processes such as cellular metabolism, signal transduction, and gene expression, posing serious threats to human health and ecosystems. When exposed, mercury ions can cross the blood–brain barrier, damaging the central nervous system [1,2], and may also impair kidney function, leading to renal failure [3]. Long-term exposure can cause neurotoxicity, immune disorders, and multi-organ failure [2]. In ecosystems, mercury ions readily convert into methylmercury, accumulating in aquatic food chains, especially in fish [3,4], and disrupting aquatic life by inhibiting growth and disturbing ecological balance. Additionally, mercury contamination can impact terrestrial plants and soil microbes [4]. Due to these hazards, the World Health Organization (WHO, 5 nM) and the U.S. Environmental Protection Agency (EPA, 2 ppb) have established exceedingly low safety thresholds [1,5]. Industrial processes risk abrupt release of high concentrations of mercury ions, resulting in environmental contamination and toxicity levels sufficient to cause significant health hazards, necessitating the detection sensitivity at the sub-micromolar or even nanomolar scale. Consequently, developing real-time, on-site visual detection methods for mercury ions is crucial [6,7]. Many countries have also established strict emission standards, requiring regular monitoring by industries [5]. Advances in detection technology have increased the potential for on-site, real-time measurement [1,8,9]. Consequently, creating highly sensitive, rapid, and portable visual detection systems for mercury ions is essential for early warning, risk assessment, and regulatory compliance, ultimately protecting human health and ecological stability.

The template method is a technique that precisely controls the morphology and structure of materials by introducing a tailored structure (template). Its main steps include template preparation, material synthesis, and template removal [10,11,12]. Depending on the type of template used, it can be classified into hard-template, soft-template, and composite template methods. Hard- and soft-template approaches are commonly used in sensor fabrication, with the hard template method especially employed for producing humidity sensors and composite piezoelectric materials. This involves using rigid templates, such as mesoporous silica or mesoporous carbon, which shape the material structure through filling or coating processes. Hard templates offer high stability and confinement, enabling precise control over the morphology and dimensions of the final materials [10,13,14,15,16]. The soft template method demonstrates excellent performance in the fabrication of high-frequency ultrasonic transducers and mesoporous materials. It employs non-rigid nanostructures as surfactants and polymers as templates, which facilitate the formation of materials through self-assembly or chemical reactions. The ease of template removal, mild processing conditions, and scalability make this approach highly suitable for large-scale production [13,16,17]. By integrating the advantages of both hard and soft templates, a synergistic effect is achieved that facilitates the formation of the material’s nanostructures [18]. Template methods are extensively employed in sensor fabrication, encompassing the synthesis of nanostructures [19,20], deposition of functional materials, preparation of molecularly imprinted polymers(MIPs) [21,22], and the development of specialized sensor types.

Silicon nanowires ingeniously integrate the material properties of silicon with the unique advantages of nanoscale structures, rendering them an ideal platform for the development of diverse biosensors [23]. In the field of biosensing, they are highly regarded for their high sensitivity, rapid response times, and exceptional electrical characteristics. Bai et al. successfully fabricated graphene nanoribbons (GNRs) with widths less than 10 nm by employing chemically synthesized silicon nanowires as physical protective masks during oxygen plasma etching. Atomic force microscopy analysis demonstrated that the resulting GNR patterns were a precise replication of the template’s nanowire morphology, enabling the formation of strip-like, branched, or intersecting graphene nanostructures [24]. Currently, GNRs, with a width controllable down to 6 nm, are successfully created. Additionally, GNR field-effect transistors (FETs) have been constructed, wherein the nanoribbons are directly connected to bulk graphene electrodes. This approach paved a novel pathway for the preparation of graphene nanostructures within the deep nanoscale regime, circumventing the need for complex lithographic techniques [25]. Consequently, it opened exciting new opportunities for the advancement of graphene nano-electronic device engineering.

This study reported the synthesis of graphene nanoribbons via chemical vapor deposition (CVD) [26,27,28]. Silicon nanowires served as templates, enabling plasma etching to pattern the graphene into ribbon structures. Subsequently, a brief ultrasonic treatment was employed to remove the masking layer, leaving behind the corresponding graphene nanoribbons. Gold electrodes were deposited using a resistance evaporation coating system, successfully fabricating a graphene nanoribbon-based sensor with silicon nanowires as templates. Four DNA sequences, comprising G bases, were subsequently functionalized onto the fabricated graphene nanoribbon sensor for downstream hybridization reactions. The sensor’s response to Hg²⁺ was characterized by monitoring its kinetic behavior and sensitivity using a semiconductor parameter analyzer (SPA).

2. Materials and Methods

Reagents and Experiments

DNA: 5′-CCACCACTTTTTTTTTGGGGTTTTTTTTT-3′, The DNA sequences used in the experiment were synthesized by Sangon Biotech (Shanghai, China) Co., Ltd., and their sequence accuracy was verified via HPLC purification and mass spectrometry. Ultrapure water (UP) used for HPLC was prepared using a laboratory ultrapure water system (UPF-10 L). All other solutions employed in the experiments were also prepared with UP water. The current profile was measured using a semiconductor parameter analyzer (SPA), and the data were processed with Origin software (Origin 2022, Northampton, MA, USA). For further details on other reagents and instruments, refer to Tables S1 and S2.

The prepared graphene nanoribbon sensors had edges functionalized with carboxyl groups (-COOH). After activation with EDC/NHS chemistry, a semi-stable Sulfo-NHS ester was generated, which could subsequently react with the amino groups (-NH₂) to form amide bonds [29]. This enabled conjugation with NH₂-terminated DNA strands, facilitating the fabrication of graphene nanoribbon-DNA biosensors (R-COOH + H₂N-R′ → -C(O)-NH-R′ + H₂O). These sensors were capable of detecting target analytes (e.g., Hg²⁺), with the underlying reaction mechanism illustrated in Figure 1.

Figure 1. Schematic diagram of the graphene nanoribbon-DNA biosensor for Hg²⁺ detection. Graphene nanoribbons possessed intrinsic carboxyl groups (-COOH) at their edges. Upon activation with EDC/NHS, these groups formed semi-stable Sulfo-NHS esters at the reaction sites. Amino-modified DNA then covalently attaches via nucleophilic attack, forming stable amide bonds. Subsequently, Hg²⁺ was selectively captured through the formation of T-Hg²⁺-T structures for detection purposes.

Based on the reported template method for silicon nanowire fabrication, and with reference to graphene nanoribbon preparation protocols, the process began by dispensing 200 μL of diluted silicon nanowire solution onto a pre-fabricated graphene device. To ensure intimate contact between the silicon nanowires and the graphene substrate, the substrate was briefly immersed in isopropanol (IPA) for 4–5 s, followed by nitrogen blow-dry to induce capillary forces. Subsequently, the devices were allowed to air dry naturally at room temperature. At this stage, the silicon nanowires were precisely positioned. The graphene device, now decorated with silicon nanowires, was then subjected to oxygen plasma etching, which introduced a high density of -COOH functional groups along the graphene edges. Final cleaning was performed via an ultrasonic bath to remove the silicon nanowire template, resulting in exposed graphene nanoribbons on the substrate. Electrical contacts were established through thermal evaporation of chromium and gold electrodes, ensuring a stable electrical connectivity for sensor operation. This process yielded graphene nanoribbon sensors, as illustrated in Figure 2. Utilizing this approach, graphene nanoribbon sensors templated with silicon nanowires were successfully fabricated.

Figure 2. Schematic diagram of the preparation of graphene nanoribbon sensors. Diluted silicon nanowires can physically be adsorbed onto graphene films, enabling oxygen ion plasma etching to remove graphene regions not protected by the template. Subsequently, brief ultrasonic treatment was employed to remove the silicon nanowire masking layer, resulting in graphene nanoribbons that precisely replicated the morphology of the original nanowires.

For sensors templated with silicon nanowires, 20 μL of activation solution containing 5 nM N-hydroxysuccinimide (NHS) and 500 nM 1-ethyl-3-(3-dimethylaminopropyl) (EDC) (at a molar ratio of 1:100) was drop-cast onto the surface for pretreatment, followed by an overnight incubation at 4 °C. The devices were then rinsed with phosphate-buffered saline (PBS) buffer and allowed to air-dry naturally. Subsequently, 20 μL of 40 nM DNA solution was drop-cast and incubated at 4 °C for 8 h to facilitate conjugation. After incubation, the surface was rinsed with PBS buffer to remove unreacted DNA strands, completing the fabrication of the graphene nanoribbon-DNA biosensor.

Finally, 20 μL of Hg²⁺ solutions with varying concentrations was added to the sensor surface, and the system was incubated at room temperature for 1 h. The resultant changes in electrical current were measured using a semiconductor parameter analyzer.

3. Results

In the initial phase, Raman spectroscopic characterization was conducted on graphene nanoribbon sensors fabricated using silicon nanowire templates. Analysis of the intensity ratio between the G and 2D peaks, along with peak position shifts, enabled the determination of the graphene layer number [30]. Monolayer graphene exhibited higher G and 2D peak intensities with a pronounced 2D peak near 2700 cm⁻¹. Thus, confirming it was a monolayer graphene structure [31]. Relevant results were presented in Figure S1, confirming that the fabricated graphene nanoribbons were of monolayer nature. Figure S2 displayed the characterization of nanoribbons prepared using silicon nanowires as templates.

3.1. Optimization of DNA Concentration

The concentration of DNA conjugated to graphene nanoribbon sensors fabricated with silicon nanowire templates was optimized. Initially, the baseline current I₀ of the prepared graphene nanoribbon devices was measured. Following overnight activation of the sensor surface with EDC/NHS solution, varying concentrations of DNA solution (10 nM, 20 nM, 30 nM, 40 nM, and 50 nM) were drop-cast onto the surface and reacted at 4 °C for 8 h. Subsequently, the sensors were rinsed with PBS and air-dried naturally. The current I was measured using a semiconductor parameter analyzer, and the change in current (I₀-I) was recorded.

As shown in Figure 3, the current change (I₀-I) exhibited a gradual increase and approached a plateau with increasing DNA concentration for the silicon nanowire-templated sensors. When the DNA concentration reached 30 nM, the rate of current change became more gradual, stabilizing at this level. Therefore, a DNA concentration of 30 nM was selected for further studies involving silicon nanowire-templated graphene nanoribbon-DNA sensors.

Figure 3. Optimization of DNA concentration in graphene nanoribbon-DNA sensors based on silicon nanowire templates.

3.2. Optimization of Hg²⁺ Reaction Time

When Hg²⁺ interacted with a graphene nanoribbon-DNA sensor, it formed a T-Hg²⁺-T mismatch structure through coordinate bonding specifically with thymine (T) bases within the DNA molecule. This triggered a dual structural modulation: firstly, it induced intra- and inter-molecular folding of the DNA into dense hairpin-like or double-stranded conformations, significantly reducing steric hindrance and electrostatic repulsion among GNRs; secondly, the T-Hg²⁺-T mismatch acted as a molecular bridge, facilitating cross-linking between adjacent DNA-modified GNRs. This process ultimately drove the assembly of GNRs into microscale aggregates. The mechanism was directly mediated by conformational rearrangements of the DNA, representing the most fundamental structural response of GNRs to Hg²⁺ binding.

As shown in Figure 4, within the first 5 min after Hg²⁺ application, the graphene nanoribbon-DNA sensor prepared on a silicon nanowire template exhibited a sharp increase in current, which gradually declined with increasing reaction time. The emergence of this phenomenon was attributed to the enhancement of DNA’s affinity for Hg²⁺ by T-Hg²⁺-T base pair modifications, which directly influenced the electrochemical interface by altering the charge distribution at the electrode–solution boundary. Consequently, the stabilization of the DNA double helix promoted more effective electron transfer from the electrode through the DNA to the detection medium, which led to a marked increase in the peak current signal in electrochemical detection [32]. By 40 min, the current change plateaued and showed the saturation of the binding process. Due to the initial extensive chelation of Hg²⁺ with T-T, the available binding sites became saturated over time, leaving residual Hg²⁺ predominantly in weakly associated states. Concurrent surface passivation, thickening of the diffusion layer, and Hg²⁺ desorption/reoxidation collectively contribute to a decrease in electronically active transfer centers, resulting in an attenuation of the signal rather than further enhancement [33]. Therefore, 40 min was selected as the optimal reaction time for Hg²⁺-DNA interaction on this sensor. Leveraging this mechanism, the detection system could achieve rapid and sufficient accumulation of Hg²⁺ ions on the electrode surface within a short timeframe, effectively addressing the low target ion enrichment efficiency and lengthy analysis duration characteristic of conventional methods.

Figure 4. The reaction time of graphene nanoribbon DNA sensors based on silicon nanowire template with Hg²⁺.

3.3. Sensitivity Testing

Graphene nanoribbon-DNA sensors fabricated with silicon nanowire templates were subjected to Hg²⁺ sensitivity analysis using a semiconductor parameter analyzer. The gate voltage was varied from −8 V to 8 V, while the source-drain voltage was maintained at −50 mV, and the initial current (I₀) was recorded. Subsequently, solutions of Hg²⁺ at various concentrations (1 pM, 5 pM, 10 pM, 50 pM, 100 pM, 500 pM, 1 nM, 5 nM, and 10 nM) were drop-cast onto the sensor surface. The reaction was carried out at room temperature, with a reaction duration of 40 min. The semiconductor parameter analyzer was used to measure the current response under identical voltage conditions, recording the current values corresponding to each Hg²⁺ concentration.

In Figure 5, the correlation between Hg²⁺ ion concentration (ranging from 1 pM to 10 nM) and the sensor current was illustrated: as the Hg²⁺ concentration increased from 1 pM to 10 nM, the current exhibited a monotonic decline. Based on the specificity of the T-Hg²⁺-T coordination mechanism, this trend reflected the modulation of interfacial charge that suppresses charge carrier transport in the GNR. The sensor exhibited saturation beyond 1 nM during the response process; therefore, in order to ensure the reliability and linearity of the calibration curve, a fit was performed using the current response data at the first five characteristic points with concentrations ≤1 nM (0.01 to 1 nM), the ionic current response demonstrated a strong linear correlation with Hg²⁺ concentration (R² = 0.997), described by the linear equation y = 0.07104x + 0.23555. This indicated that in the present concentration interval, the T-Hg²⁺-T coordination system remained unsaturated, with binding proportions increasing uniformly with concentration, and the signal response maintaining a stable concentration dependence. The maximum relative standard deviation (RSD) observed was 2.96%, which signified high reproducibility in mercury ion detection. Based on these findings, the detection limit was established at 3.62 pM, underscoring the sensor’s high precision in detecting trace levels of Hg²⁺. The LOD of this work had an advantage compared with that of other works, such as 1 nmol/L [1].

Figure 5. Sensitivity analysis of Hg²⁺ detection using graphene nanoribbon-DNA sensors based on silicon nanowire template.

The selection of analytical methods required balancing sensitivity, selectivity, cost, portability, and analysis throughput. Table 1 provides synthesized data for a comprehensive comparison of the principal techniques. In comparison with the existing mainstream biosensor detection technologies, the graphene nanoribbon-DNA sensor developed in this study demonstrated a superior detection sensitivity. Conventional methods, such as fluorescence sensing, electrochemical sensing, and colorimetric detection, typically achieve detection limits for Hg²⁺ in the nanomolar (nM) to micromolar (μM) range. Although electrochemical techniques based on differential pulse anodic stripping voltammetry (DPASV) can analyze trace amounts of Hg²⁺, the graphene nanoribbon-DNA sensing system employed in this research further overcame sensitivity bottlenecks through its unique interfacial charge regulation mechanisms, including the inductive effects of T-Hg²⁺-T coordination and the modulation of carrier transport at GNR edge states. This approach enabled the sensor to exhibit enhanced trace detection performance while simultaneously surpassing the capabilities of traditional methods.

Table 1. Comparison of mercury ion detection using various sensors.

Sensing Method	Method	Linear Range	Detection Limit (LOD)	References
Colorimetric	Ligand-induced aggregation of gold nanoparticles (AuNPs)	0.08–1.25 μM	37 nM	[34]
	Cysteine-modified Au-Ag core–shell nanorods	1–60 μM	1.273 μM	[35]
	GO-PtNPs composite materials	5–250 nM	4.5 nM	[36]
Electrochemical	Photoelectrochemical/photovoltaic amperometry (PEC); Au nanoparticle-decorated g-C3N4 photoelectrodes	0.01–0.5 nM	0.13 nM	[37]
	Differential pulse anodic stripping voltammetry (DPASV)	0.5–1.8 nM	0.16 pM	[38]
	Nitrogen and boron co-doped carbon dots, graphene oxide/carbon nanotube composites	0.05–25 μM	9.7 nM	[39]
Fluorescence	Carbon quantum dots (CQDs)	0–1 μM	0.5 μM	[40]
	Static Quenching	-	5.3 nM	[41]
	Coupled fluorescence-thiol-alkene (excimer) probe	-	1.98 nM	[42]

To evaluate the reproducibility and storage stability of the graphene nanoribbon-DNA sensor for trace Hg²⁺ detection, targeted testing was carried out. Under identical experimental conditions at room temperature, the same sensor was used to perform no less than ten repeated measurements of the target analyte. The calculated RSD averaged approximately 2.8%, satisfying the reproducibility criteria for biosensors used in trace analysis. Additionally, the sensors used were stored at 4 °C within a vacuum chamber for 30 days, after which their current response was remeasured. The results demonstrated a signal retention rate of 93%, effectively ensuring the accuracy of detection. Collectively, the graphene nanoribbon-DNA sensor designed in this study exhibited outstanding performance characteristics.

3.4. Selectivity Analysis

Various metal ions for selective analysis were selected. Under the optimal conditions described above, identical concentrations (1 nM) of different metal ions were introduced into graphene nanoribbon-DNA sensors fabricated with silicon nanowire templates. The sensors were incubated at room temperature for 40 min. The resulting current responses were measured using a semiconductor parameter analyzer, with the results depicted in Figure 6.

Figure 6. Selective analysis of Hg²⁺ using graphene nanoribbon DNA sensors based on silicon nanowire template.

Figure 6 illustrates the influence of different metal ions at an identical concentration (1 nM) on the current response of the graphene nanoribbon-DNA sensor: it was evident that only Hg²⁺ elicited a significant current response, far exceeding that of other ions, while responses from interfering ions such as Cu²⁺, K⁺, and Fe²⁺ were minimal. This disparity arose from the specificity of T-Hg²⁺-T coordination bonds and the charge modulation of GNR. Specifically, other metal ions (e.g., Cu²⁺, Fe²⁺, Pb²⁺) lacked the capacity for selective coordination with thymine bases; for example, Cu²⁺ interacted with DNA predominantly through non-specific electrostatic forces, preventing stable complex formation. Even weak interactions induced minimal interfacial charge changes, such as positive charge accumulation, which were substantially lower than those caused by Hg²⁺. As a result, Cu²⁺ maintained the sensor’s low-current response. Conversely, Hg²⁺ induced substantial positive charge accumulation at the GNR edges, effectively modulating the edge state carrier concentration and amplifying the current response. The weak electrostatic influence of interfering ions failed to induce notable changes in the edge states, resulting in a filtered response signal and conferring high selectivity of the sensor for Hg²⁺.

The observed decrease in current intensity upon addition of Hg²⁺ was remarkably greater than that caused by other metal ions. This showed that graphene nanoribbon-DNA sensors fabricated with silicon nanowire templates exhibited high selectivity and a significant response specifically to Hg²⁺. In summary, the results in Figure 6 confirmed that the selective recognition of Hg²⁺ by the graphene nanoribbon-DNA sensor was fundamentally mediated by the specific T-Hg²⁺-T coordination chemistry, with the charge-sensitive properties of GNR edge states further enhancing this selectivity, enabling precise detection of Hg²⁺ within complex ion matrices.

The introduction of Hg²⁺ ions promoted the formation of a stable “T-Hg²⁺-T” hairpin structure within the DNA, thereby impeding electron conduction and resulting in a significant reduction in current. Dithiothreitol (DTT) adsorbs Hg²⁺ effectively through its thiol (-SH) groups via strong coordination, restoring the electron transfer pathway and causing the current to increase. Upon subsequent addition of Hg²⁺, it bound to the thiol groups of DTT, forming a stable complex that reinhibited electron conduction and led to a decrease in current. This sensing system exhibited reversibility due to these mechanisms, as illustrated in Figure S3.

3.5. Detection in Actual Water Samples

Based on the sensitivity analysis, graphene nanoribbon-DNA sensors constructed using the templates demonstrated that those prepared with silicon nanowire as templates exhibited significantly higher sensitivity to Hg²⁺. The developed graphene nanoribbon-DNA sensor was systematically evaluated for its detection performance of target analytes, with Hg²⁺ serving as the model pollutant, in real-world samples. Standard solutions of Hg²⁺ at concentrations of 0.01 nM, 0.05 nM, 0.1 nM, 0.5 nM, and 1 nM were prepared, while simultaneously collecting Yangtze River water samples to simulate environmental water testing scenarios. To ensure analytical accuracy, the river samples underwent pre-treatment via filtration through 0.22 μm membrane filters to effectively remove particulate impurities, thus minimizing interference in detection outcomes. The sensor demonstrated excellent performance in detecting known Hg²⁺ concentrations; recovery rates ranged stably from 97.38% to 114.69% (Table 2), indicating its capability to accurately quantify the true Hg²⁺ levels in solutions. Relative standard deviations (RSD) were maintained between 2.87% and 3.25%, further confirming the stability and reliability of the measurements. After applying the sensor to Yangtze River water samples, the Hg²⁺ concentration was measured at 142.35 pM. Consequently, this validated the sensor’s reliability, accuracy, and potential utilization for in situ environmental surveillance.

Table 2. The detection results of Hg²⁺ solution with known concentrations.

Samples	Added (nM)	Obtained (nM)	Recovery (%)	RSD (%)
1	0.01	0.011	114.69	3.25
2	0.05	0.049	98.67	3.06
3	0.1	0.103	102.93	2.99
4	0.5	0.503	100.68	3.00
5	1	0.974	97.38	2.87

To address the insufficiency of real-sample validation, we supplemented gold-standard inductively coupled plasma-mass spectrometry (ICP-MS) cross-validation, which systematically verified the accuracy and reliability of the quantitative results of the graphene nanoribbon-DNA sensor. The results showed that the ICP-MS detection value was 138.67 pM, and the relative deviation from the sensor’s measured value (142.35 pM) was only 2.65%, falling within the acceptable error range (<5%) for trace heavy metal analysis in environmental samples. This high consistency confirmed that the sensor’s quantitative results were highly comparable to the gold-standard ICP-MS method, validating both the accuracy of the measured Hg²⁺ concentration in the Yangtze River sample and the sensor’s reliability for detecting trace Hg²⁺ in complex environmental matrices.

This sensor offered high sensitivity combined with field-deployable capabilities, with analysis completed within 40 min using a label-free, single-step incubation mode, and required minimal sample preparation. The custom-designed portable device maintained low manufacturing costs, enabling deployment in remote regions. The detection relied on DNA-specific recognition via T-Hg²⁺-T base pairing, providing robust anti-interference in complex environmental matrices. However, its linear detection range was narrow (0.01–1 nM), limiting its applicability for high-concentration Hg²⁺ detection. Future improvements could involve optimizing DNA grafting density or graphene nanoribbon aspect ratios or integrating standardized portable readout devices. The method achieved a favorable balance among sensitivity, detection speed, cost, and on-site usability, establishing itself as a highly competitive practical alternative for environmental mercury ion monitoring, particularly suited to rapid, low-cost, and in-field testing scenarios.

4. Conclusions

In this study, a template-assisted fabrication approach was used to develop a type of graphene nanoribbon-based DNA sensor, utilizing silicon nanowires as templates. The templates were physically adsorbed onto the graphene device surface and subsequently patterned through oxygen plasma etching, precisely producing GNRs with shapes conforming to that of the templates. The edges of etched GNRs were functionalized with abundant carboxyl groups. Chromium and gold electrodes were deposited via high-vacuum resistance evaporation, with the two gold electrodes serving as the source and drain, respectively. Thereby, GNR-based sensors were constructed. The fabricated sensors were activated using EDC/NHS chemistry, followed by conjugation with amino-modified DNA through amide bond formation with Sulfo-NHS ester-activated carboxyl groups. The response kinetics of these sensors to Hg²⁺ were characterized using a semiconductor parameter analyzer, revealing a detection limit of 3.62 pM for sensors fabricated with silicon nanowire. The graphene nanoribbon-DNA sensor designed in this study exhibited several key advantages for Hg²⁺ detection: it offered straightforward operation, rapid response times, visual readouts, and ease of integration into existing systems. More importantly, its sensitivity surpassed that of current comparable detection methodologies.

Mercury ion detection has evolved from traditional large-scale analytical instruments to a modern sensing era centered on innovative materials and new detection principles. The application of fluorescence, electrochemical, and other techniques demonstrated significant potential for real-time, on-site monitoring across various fields. Future research should address challenges related to stability, interference resistance, and cost efficiency. Through interdisciplinary integration, the goal was to advance technological industrialization and system integration, thereby supporting mercury pollution control and safeguarding public health.