A Highly Selective Real-Time Electroanalytical Detection of Sulfide Via Laser-Induced Graphene Sensor

SEM was leveraged to understand the surface morphology and the material structure's pore-size distribution. This implies that surface carbonization has been performed, which is required for the sensing. In Figure 1, the SEM images of different resolutions are provided. In addition, energy-dispersive X-ray analysis was performed to understand the material composition on the surface. Herein, different compositions of the material with their respective percentages are provided. The results indicate that the carbon has the very high conductivity and here in this work the laser induced graphene has been used which also has good conductivity as well as high surface area to volume ratio which is the required for the sensing. Furthermore, the surface modification carried out with MB can be seen. MB plausibly reacts with H₂S for electrooxidation to take place [11].

Figure 1:

The XRD of the sensing material has been carried out to understand the lattice effect of modifying the sensing material. Figure 2(a) is the XRD peak of the LIG/MB. The peak at 32° can be referred to as MB, the peak around 26° can be associated with the LIG, and the peak around 22° can be referred to as the polyimide sheet peak.

Figure 2:

For understanding bond formation, XPS is useful. Figure 2(b) is the XPS analysis of the LIG/MB working electrode. The respective element has a peak at the relevant binding energy, such as for carbon at 285.27 eV, nitrogen at 400.31 eV, oxygen at 533.05 eV, chlorine at 198.76 eV, and sulfur at 165.3 eV. The atomic weight percentages are as follows: carbon 81.1, nitrogen 3, oxygen 5.76, chlorine 0.18, and sulfur 0.76.

The electrochemical experiment was performed using the MB-modified working electrode via CV, wherein the potential window was selected between −1 V to +1 V and the scan rate was 10 mV/s. Figure 4(a) shows a comparative CV of bare LIG in 0.1 M phosphate buffer solution (PBS) and 1 mM sodium sulfide (Na₂S) solution. While the characteristic peak was observed, the MB-modified LIG shows a discrete oxidation peak at a potential of 0.17 V, as shown in Figure 4(b) with 1 mM Na₂S at 10 mV/s for n = 4. MB was often used as the redox mediator for H₂S detection [9].

Figure 4:

The scan rate analysis verified that the modification done using MB was successful, as the peak pertaining to MB was observed in the PBS due to electrochemical uptake. For this analysis, a buffer solution of 0.1 M PBS (pH = 7.4) was taken, and the scan rate was changed between 10 mV/s and 100 mV/s. As shown in Figure 5(a), a clear peak of MB has been observed. Further, the effect of varying scan rates over the electrocatalytic oxidation of LIG/MB over 1 mM Na₂S was tested, as depicted in Figure 5(b) [34, 35].

Figure 5:

As revealed in Figure 5(b), the redox peak was seen to increase gradually while enhancing the current. It can be seen that by using the Randle–Sevick equation, the current displays a linear relationship with the square root of the scan rate [19, 27, 28]. The linear plot shown in Figure 5(c) regarding the diffusion coefficient was 9 × 10⁻⁶ cm² ⋅ s⁻¹ [29]. To calculate the number of electrons that take part in the reaction, the Laviron plot was used, and the Laviron plot's slope gives the value. Ip=2.69×105×n×(α×nα)0.5×A×D0.5ν0.5CA {I_{\rm{p}}} = 2.69 \times {10^5} \times n \times {(\alpha \times {n_\alpha})^{0.5}} \times A \times {D^{0.5\nu 0.5}}{C_{\rm{A}}} where I_p indicates the peak current, n the number of electrons, α the charge transfer coefficient, n_α the number of electrons involved in the reaction, D the diffusion coefficient of the analyte, A the area of the electrochemical sensor, and C the concentration of the analyte in bulk. The calculated electrochemical active surface area of the fabricated electrochemical sensor was 0.1495 mm². The empirical Wilke–Chang formula can help calculate the diffusion coefficient [36].

Where T indicates the temperature, M the molecular weight of the solvent, χ the association parameter (2.6 in the present case), ν the viscosity at 25°C, and V the molar volume of the solute at the boiling point.

The Laviron equation: Epa=E°+RTαnFlnRTKsαnF+RTαnFlnv {E_{{\rm{pa}}}} = E^\circ + {{RT} \over {\alpha nF}}ln{{RTKs} \over {\alpha nF}} + {{RT} \over {\alpha nF}}lnv where α represents the transfer coefficient, K_s the standard rate constant, n the electron's number, E° the redox potential, T the absolute temperature, and F the Faraday constant. Thus, E_pa is proportional to lnν. The slope value was 1.32, α_n = slope of the E_pa vs. lnν = 1.32, and α = 0.5. This leads to the value of n as 2.64, concluding that the number of electrons in the reaction was 3 [33].

Further, to understand the device performance, it was necessary to perform the concentration effect analysis in which the sensor detection limit was determined. The device performance was analyzed using the CA technique in 0.1 M PBS (pH = 7.4). Each concentration was tested for 4 min in the potential value of 0.17 V obtained by the CV. While performing the CA, there was an oxidation of S²⁻ to S⁰ on the surface of the LIG/MB electrochemical sensor.

The concentration effect was performed thrice (n = 3) to ensure the accuracy and repeatability of the developed sensor. Then, the mean value and error bar were taken as the response values for the calculation. Figure 6 shows the concentration effect analysis in the 0.5 μM–1 mM Na₂S range. It was observed that as the concentration was decreasing, the current response was also decreasing. This was due to the buildup of sulfur on the surface of the working electrode and the leaching of MB from the surface of the working electrode.

Figure 6:

There were two linear ranges observed from the calibration curve of the current vs. concentration, 0.5–50 μM and 100 μM–1 mM, and their respective linear equations are as follow: Ip=0.0041×concentration+0.181,R2=0.969(0.5−50μM)Ip=0.0007×concentration+0.4964,R2=0.957 (100 μM−1 mM) \matrix{{{I_{\rm{p}}} = 0.0041 \times {\rm{concentration}} + 0.181,} \hfill \cr {{R^2} = 0.969(0.5 - 50\mu {\rm{M}})} \hfill \cr {{I_{\rm{p}}} = 0.0007 \times {\rm{concentration}} + 0.4964,} \hfill \cr {{R^2} = 0.957\,(100\,\mu {\rm{M}} - 1\,{\rm{mM}})} \hfill \cr}

The LoD was calculated to be 0.435 μM, and the sensitivity of the two linear ranges (low and high) for the electrochemical sensor was calculated as Sensitivity=0.295μA/(μM mm2) for 0.5−50 μMSensitivity=0.0047μA/(μM mm2) for 100−1 mM \matrix{{{\rm{Sensitivity}} = 0.295\mu {\rm{A}}/(\mu {\rm{M}}\,{\rm{m}}{{\rm{m}}^2})\,{\rm{for}}\,0.5 - 50\,\mu {\rm{M}}} \hfill \cr {{\rm{Sensitivity}} = 0.0047\mu {\rm{A}}/(\mu {\rm{M}}\,{\rm{m}}{{\rm{m}}^2})\,{\rm{for}}\,100 - 1\,{\rm{mM}}} \hfill \cr}

The developed sensor's signal-to-noise (S/N) ratio was found to be 2.76. The limit of quantification (LoQ) was found to be 2.45 μM.

The pH dependency of the solvent was a significant parameter for the oxidation reaction, as it can affect the reaction between LIG/MB and sulfide. Here the pH range was taken between 3 and 11 to study the behavior of the electrocatalytic oxidation in various acidic and basic pH ranges. The value of the E_pa was plotted against the respective pH value. It shows that the electrochemical sensor was pH dependent. The number of protons involved in the reaction can be measured using the Nernst equation;

Through the pH analysis, it can be observed in Figure 7 that as the pH of the buffer solution varied from low to high, there was a shift in the peak potential and peak current. The respective E_pa vs. pH and I_pa vs. pH, is shown in Figure 7(b, c). Thus, it can be observed that for each pH, a different number of protons was involved in the reaction and thus the pH of the PBS plays and important role in the sensing. As the charge transfer rate changes, the detection also gets affected [25].

Figure 7

Interference tests have been performed to determine the electrochemical sensor's selectivity. For this, various interfering analytes such as NaNO₂, Na₂SO₃, Na₂CO₃, DA, and AA with 100 μM concentration were taken and mixed in 5 μM concentration of Na₂S predissolved in 0.1 M PBS (pH = 7.4). In addition, O₂ and CO₂ were purged in the Na₂S solution for 10 min to vary the effect of gases. It can be seen that there was no significant interference of the various analytes mentioned above, and the sensor shows good selectivity. The current change mentioned in percentage at the top of the bar graph and the result of the interference study are displayed in Figure 8

Figure 8:

The repeatability of the device has been analyzed and the respective bar graph is shown in Figure 9(a). The relative standard deviation (RSD) was found to be 2.62% for the repeatability of the sensor as can be seen in Figure 9(a). For the reproducibility of the sensor, three different sensors have been fabricated and tested for sulfide detection. The bar graph has been shown in Figure 9(b) along with the error bar and the respective RSD is 2.26%. This analysis indicates that the fabricated sensor shows good repeatability and reproducibility for sensing applications. For stability analysis, the device was used for 15 days The fabricated electrochemical sensor shows good stability as the deviations were around 3.67% for the peak potential value and 4.24% for the peak current value, respectively.

Figure 9:

Real sample testing was performed to verify the fabricated electrochemical sensor's practical ability. The lake water sample was collected from different lakes, such as Kapra Lake, Hussain Sagar Lake, and Shameerpet Lake in Hyderabad, India. For the real sample analysis, the sample preparation involved mixing the sample with 10-fold of 0.1 M PBS (pH = 7.4) solution. Further, different concentrations of the Na₂S, such as 10 μM, 50 μM, and 100 μM, were used in the sample prepared. The electrochemical technique used was CA at a potential of 0.17 V and was run for 5 min. Following this, recovery analysis was performed in which the electrochemical sensor showed good recovery. The low RSD indicates that the device performance is good. The result of the real sample analysis is given in Table 2. Also, the respective CA was provided for each lake with different concentrations spiked in Figure 10(a–c).

Figure 10: