Microplastic Sensing: A Major Global Problem

In the global effort to regulate microplastics and nanoplastics (MNPs), large-scale detection of these pollutants remains a significant challenge. The lack of standardized methods (standardized methods em negrito) and detailed characterization of MNP concentrations in different environments compromises both the effectiveness of legislation and public awareness (legislation and public awareness em negrito) of this issue¹.

The vast diversity in size, composition, and shape of MNP particles, combined with the variability of the environments where they are found, makes their detection and quantification a complex task (complex task em negrito). Existing methods often struggle to accommodate these variations and may fail to provide consistent and accurate results².

Recently, advances have been made in the search for more effective alternatives to detect and quantify MNPs. Established technologies such as Fourier-transform infrared spectroscopy (FTIR), scanning electron microscopy (SEM), and fluorescence-based techniques (Fourier-transform infrared spectroscopy (FTIR), scanning electron microscopy (SEM), and fluorescence-based techniques em negrito) have been refined to handle the diversity and complexity of these particles³. On other hand, emergent electrochemical methods (electrochemical methods em negrito) represent a potential alternative for rapid, low cost, user-friendly, sensitive and reproducible (sensitive and reproducible em negrito) quantification of microplastics4.

In particular, electrochemical biosensors are a powerful system to boost specificity since their responses are mediated by target recognition through an electrode-tethered bioreceptor. Beyond their primary role in quantifying the analyte with enhanced analytical performance, such biosensing devices can also be used to study the bioreceptor/target binding kinetics and thermodynamics. For instance, they can be used to determine the affinity constant involved in these interactions, along with the number of binding sites5. These insights are essential for experimentally reaching a more comprehensive understanding of the protein proposed by our project (“our project” link to https://2024.igem.wiki/cnpem-brazil/model).

Therefore, the fabrication and validation of hardware suitable for coupling with genetically engineered proteins (hardware suitable for coupling with genetically engineered proteins em negrito) aimed at contaminant quantification could serve dual purposes. It offers a fast and cost-effective solution for experimentally determining the dissociation constant, which is crucial for advances in synthetic biology (synthetic biology em negrito).

At the same time, it provides a highly accurate method for contaminant sensing, addressing a significant global challenge (“significant global challenge” em negrito).

The Journey of Microplastic Detection Techniques throughout the B.A.R.B.I.E. Project

In this section, we present an overview of recent high-impact works on the detection of micro- and nano plastics. (“In this section, we present an overview of recent high-impact works on the detection of micro- and nano plastics” em negrito e italico.)

In general, electrochemical alternatives for the detection of MNPs aim to quantify both the concentration and the average size of particles. However, a significant challenge is dealing with sensitivity and specificity when interrogating the devices with real and complex samples, which can introduce various types of interferences, e.g., nonspecific adsorptions. These interferences can arise from macromolecules and particles similar in size and electrochemical properties of MNPs or from the interaction of MNPs with different molecules on their surfaces6.

Currently, the main established methods for detecting micro and nanoplastics include image-based methods such as optical microscopy7,8, fluorescence microscopy9, electron microscopy10,11, and spectroscopy-based methods such as Fourier-transform infrared (FTIR)12,13,14 and Raman spectroscopy15,16,17. The primary limitation of these techniques is the time required for analysis and the restrictions in point-of-care determination of microplastics. Moreover, none of these characterizations are sufficiently precise on their own, necessitating a combination of techniques to obtain reliable results. For instance, image-based techniques might be used to identify particles and their sizes, while spectroscopy is employed to determine the composition and degradation state of the MNP. There are also approaches that allow coupling microscopy analysis with spectroscopy, such as μ-FT-IR or μ-Raman17,18, though these systems are typically sensitive only to particles larger than ~10 μm19.

Recent approaches explore the use of electrochemical arrays for the detection and quantification of MNPs. In 2021, Colson and Michel proposed using impedance spectroscopy to detect individual plastic particles within a microfluidic channel coupled with an electrochemical working cell21. In this setup, the passage of microplastics was individually detected by a peak in the impedance scan. Moreover, using the real and imaginary parts of the impedance measured at six different current frequencies, a machine learning model was trained to distinguish signals from polymeric particles, biological residues, or other sediments of similar size, enhancing the sensor's prediction capability This method advanced the detection of microplastics due to its precision, speed, and scalability. However, the results showed a considerable drop (≈50%) in sensitivity for microplastics smaller than 200 μm.

In 2022, Ching et al proposed an impedance-based sensor to predict the concentration, composition, and size of microplastics in water22. This system relied on static sample measurements and used the k-nearest neighbors algorithm for predictions. Although sensitive for microplastics larger than 20 μm, the proposed sensor had a limitation in selectivity affecting its use in real samples. In 2023, H Du et al proposed a graphene-based impedance sensor capable of enhancing selectivity for micro- and nanoplastics23. Although similar to other electrochemical sensors, this system employed machine learning techniques such as principal component analysis to integrate signals from impedance measurements across multiple frequencies. Despite its innovative approach, the method faced reproducibility challenges in microplastic detection. Furthermore, the sensor lacked a selective recognition element for MNPs beyond machine learning analyses limiting its specificity in detecting these pollutants.

More recently, in April 2024 Z Xiao et al proposed a photoelectrochemical biosensor based on the aggregation effect of MNPs driven by protein corona formation24. In this sensor an electrode with $Cu_3SnS_4$ nanostructures was fabricated to which bovine serum albumin (BSA) was anchored. The aggregation effect induced by interaction of BSA with MNPs was used to determine concentration of MNPs through changes in measured photocurrent. However, specificity of this system is expected to be low because of nonspecific character of this intermolecular interaction.

In conclusion while electrochemical methods show promise for detecting MNPs by quantifying particle concentration and size they encounter significant challenges when applied to complex real-world samples. Interferences from macromolecules or similarly-sized particles can compromise accuracy current approaches often require a combination of techniques to achieve reliable results. Recent advancements including integration machine learning development innovative sensor designs like impedance spectroscopy photoelectrochemical biosensors have enhanced sensitivity scalability prediction capability. However several challenges persist in detecting smaller MNPs ensuring selectivity reproducibility underscoring need further innovation in this field.

Our Proposal

To address urgent need rapid accessible accurate detection MNPs in B.A.R.B.I.E project we developed an electrochemical biosensor capable quantifying micro- nanoplastics (MNPs) by plastic binding proteins as recognition element coupled label-free electrochemical biosensor.

The sensing device consists gold microfabricated device glass substrate. Additionally device includes three electrodes: working counter reference electrodes. This set electrodes allows faradaic charge-transfer reaction-mediated electrochemical measurements. Moreover active area each electrode photolithographically defined by SU-8 polymer which has been reported essential ensuring reproducibility measurements25.

The sensor fabricated cleaning glass substrate applying HMDS adhesion layer patterning design via photolithography A Cr/Au metal layer deposited followed by liftoff surface cleaning The process concludes SU-8 deposition final cleaning As final result manufacturing process 12 plates obtained each containing 8 different sensors The figure below shows resulting plate (Fig 1).

To construct biosensor idea using plastic binding proteins (PBPs) includes Barbie1 engineered (“engineered” link https://2024.igem.wiki/cnpem-brazil/model) our team specifically binds MNPs The concentration these particles then quantified using square wave voltammetry (SWV) arrangement label-free biosensor redox probe solution To facilitate quantification point-of-need handheld low-cost one-channel potentiostat capable performing EIS SWV other electrochemical measurements employed.

Furthermore three distinct methods employed evaluate sensor's performance:

• Blank (BA): Utilizes electrochemical chip without surface modification PBPs.
• Bind (BI): Utilizes electrochemical chip modified PBPs working electrode surface faradaic SWV measurements.
• Bulk (BU): Involves prior process bulky corona-protein formation PBPs analyzing entire MNP-PBP complex adsorbed working electrode.

Detection methods evaluation

We present different methods identifying quantifying micro nano plastics proposed project discuss their advantages disadvantages.(We present different methods identifying quantifying micro nano plastics proposed project discuss their advantages disadvantages em negrito e italico.)

Blank (BA)

For this detection methodology MNPs placed aqueous solution containing electrolyte known electrochemical response SWV technique explore quickly efficiently work electrode surface occupation Since MNPs small particles composed dielectric material can trigger various electrochemical processes solution such hindering ion diffusion accumulating charges surfaces filling surface region blank working electrode These effects impact impedance SWV measurements. While all mechanisms contribute quantifying MNPs water other small materials present water can also affect results extent.

Deposited - $\text{Na}_2\text{SO}_4$:

As initial approach exploring electrochemical sensing MNPs solution $\text{Na}_2\text{SO}_4$ tested reported effective detecting microplastics21 First EIS measurements conducted blank sensor using solution containing 0.5 M $\text{Na}_2\text{SO}_4$ Typical $\text{Na}_2\text{SO}_4$ curves Bode plot shown Figure 2.

At higher frequencies impedance phase below 45° magnitude remains constant relative frequency characteristics more resistive behavior In mid-to-low frequency range phase exceeds 45° magnitude starts varying linearly frequency indicating more capacitive behavior solution.

Based analysis electrolyte’s behavior circuit shown Figure 3 chosen fit EIS data In circuit $\text{Na}_2\text{SO}_4$ solution modeled resistor R_s interface between work electrode solution represented parallel combination capacitor C_surf resistor R_surf mathematically capturing capacitive resistive electrochemical processes occurring surface It expected presence microplastics interface between working electrode electrochemical solution will result changes measured values R_surf C_surf.

The verify relationship between concentration MNPs electrochemical EIS response each sensor rinsed with 1 mL deionized water PSNBs (1000 nm) progressively added concentrations 0.01 0.1 1 mg/L At each stage samples incubated 10 minutes measured $\text{Na}_2\text{SO}_4$ then rinsed with 2 mL deionized water Obtained values R_surf C_surf normalized against blank ($\text{Na}_2\text{SO}_4$) evaluated relation PS (polystyrene) concentration shown Figure 4.

The results equivalent circuit fitting using impedance spectroscopy NaSO4 solution yielded unsatisfactory R² values 0.22 0.25 indicating poor correlation between modeled experimental data Additionally reproducibility compromised causing excessively high deviations measurements Essentially conclude approach too simple detect microplastics effectively.

Deposited - $\text{K}_3[\text{Fe}(\text{CN})_6]$/$\text{K}_4[\text{Fe}(\text{CN})_6]$:

The $\text{K}_3[\text{Fe}(\text{CN})_6]$/$\text{K}_4[\text{Fe}(\text{CN})_6]$ probe widely used electrochemical detection experiments due well-established oxidation reduction processes literature. Additionally concentration microplastics evaluated using square wave voltammetry (SWV) simple fast method demonstrated high reproducibility This makes results conclusions obtained reproducible valuable future iGEM teams other projects aiming detect microplastics.

Due redox processes occurring $\text{K}_3[\text{Fe}(\text{CN})_6]$/$\text{K}_4[\text{Fe}(\text{CN})_6]$ probe SWV measurements show current peak specific applied potential where maximum current (I_Max) related electroactive area working electrode The PS concentration thus connected surface coverage working electrode consequently measured I_Max To explore relationship samples PSNBs (polystyrene nanobeads) (1000 nm) progressively analyzed concentrations 0.01 0.1 1 10 100 mg/L At each stage samples incubated 10 minutes washed with 2 mL deionized water Results analytical curve shown Figure 5.

While approach yielded slightly better outcomes compared impedance spectroscopy method results remained inadequate R² values (~0.76) still falling short acceptable levels (>0.9) By depositing MNPs surface using SWV there poor linearity sensitivity Thus not expected enable determination MNPs real-world samples.

Bind (BI)

In the Bind (BI) methodology, the sensor is pre-modified by binding the target protein on the work electrode. After this process, the sample of MNPs dispersed in the electrolyte is brought into contact with the electrodes. The specific interaction between the plastic particles and the protein can produce electrochemical results that determine the concentration of the MNPs with minimal interference from other potential materials in the medium.

Drop casting - BaCBM2

To explore the effectiveness of using plastic-binding proteins for MNP sensing, the drop casting method was employed. This protein immobilization method simply involves the protein adsorption by depositing onto the substrate. First, SWV measurements were taken with the $\text{K}_3[\text{Fe}(\text{CN})_6]$/$\text{K}_4[\text{Fe}(\text{CN})_6]$ probe for each clean sensor. After washing each device with 1 mL of PBS (10 mM), 5 µL of BaCBM2 protein (500 µg/mL) was added to the working electrode, followed by a 20-minute incubation.

After incubation, the devices were washed with 2 mL of PBS, and SWV measurements were taken again to define the blank, as shown in Figure 5, where a decrease in current after protein deposition indicates the proteins were likely fixed to the surface.

Each sensor was then washed again with 1 mL of PBS, and PSNBs (1000 nm) were progressively added at concentrations of 0.01, 0.1, 1, 10, and 100 mg/mL. At each stage, the samples were incubated for 10 minutes and washed with 2 mL of PBS. The current from each SWV measurement was normalized relative to the blank to create an analytical curve based on the microplastic concentration, shown in Figure 6.

By using the BaCBM2 proteins interaction with the MNPS the results became significantly better, with the R² going from 0.76 to 0.94 just by adding the protein. This already allows the biosensing of microplastics. However there is room for improvement, especially because the standard deviation between measurements is quite large, making the concentration values uncertain.

Self-Assembled Monolayers (SAMs) - BaCBM2

To ensure BaCBM2 immobilization and amplify the measured signal, the modification of the working electrode with three different self-assembled monolayers (SAMs) was tested, L-cysteine, cysteamine and graphene oxide26, as shown in Figure 7.

First, SWV measurements were taken using the $\text{K}_3[\text{Fe}(\text{CN})_6]$/$\text{K}_4[\text{Fe}(\text{CN})_6]$ (5 mM) probe on the clean sensors. Afterward, the sensors were rinsed with 1 mL of PBS to allow for SAM deposition. Each modification (L-cysteine, cysteamine, and graphene oxide) was performed on 4 different sensors. For this, 5 µL of each solution was added to the sensors at concentrations of 50 mM, 50 mM, and 2 mg/mL, respectively, for L-cysteine, cysteamine, and graphene oxide. The samples were then incubated at 4°C for 15 hours. After incubation, the sensors were washed with 2 mL of PBS per device, and SWV analysis was conducted with $\text{K}_3[\text{Fe}(\text{CN})_6]$/$\text{K}_4[\text{Fe}(\text{CN})_6]$ (5 mM).

Following this, the devices were rinsed with 1 mL of PBS, and 5 µL of BaCBM2 protein solution (150 µg/mL) was added. The sensors were incubated with EDC:NHS (16.6:33.33 mM) for 45 minutes at 37°C. After incubation, each device was washed with 2 mL of PBS, and SWV measurements were taken again.

Next, the samples were rinsed with 1 mL of PBS before adding 5 µL of SuperBlock solution, which was incubated for 90 minutes. After this stage, each device was washed with 2 mL of PBS, and new SWV measurements were taken as a blank.

Finally, each sensor was washed again with 1 mL of PBS, and PSNBs (1000 nm) were progressively added at concentrations of 5, 10, 50 and 100 mg/L. At each step, the samples were incubated for 20 min and washed with 2 mL of PBS. The current obtained from each SWV measurement was normalized against the blank to create an analytical curve for the microplastic concentration, shown in Figure 8.

#

Based on this initi 0.85) was the one using Cysteamine to bind the protein to the gold surface. Therefore we repeated the experiment increasing the incubation time (20 min to 30 min), the EDC:NHS concentration (16.6:33.33 mM to 25:50 mM) and testing an improved set of concentrations: 0.01, 0.1, 1, 10 and 100 mg/L, to obtain the analytical curve shown in the Figure 9.

The result obtained was an extraordinary R² value of 0.99 and an extremely reproducible curve with minimal deviation between the repetitions of the experiment. This means that by strongly binding the protein on the surface of the electrode, the sensor becomes highly precise in detecting even low concentrations of microplastic. For reference, the lowest concentration tested of 0.01 mg/L, that is the lowest concentration quantified by Du, H. et al21 and is the equivalent of adding 1 gram of plastic in 100.000 liters of water.

Bulk (BU)

In the Bulk (BU) methodology, MNPs are initially placed in a solution containing a high concentration of target proteins. After a certain period, the surface of the particles is coated with these proteins, forming protein coronas (Model), thus, the interaction process is in the solution bulk. After this process, part of this solution is deposited specifically onto the working electrode. Following the immobilization of these MNP-protein complexes, an electrolyte is added to establish contact between the three electrodes.

For the experimental tests of the bulk methodology, the BaCBM2 protein (150 µg/mL) and PSNBs (0.01, 0.1 and 1 mg/mL) were first mixed in PBS (10 mM) using a Microtube Rotating Mixer for 10 minutes to allow the formation of protein coronas.

Following this, SWV measurements were performed using clean sensors. Next, a sample containing only 150 µg/mL of BaCBM2 in PBS, which was also shaken for 10 minutes, was incubated for 10 minutes to measure the blank by SWV, where a decrease of 26% in the peak current was observed.

The four concentrations of PSNBs were then progressively incubated on the sensor. After each incubation, the sensor was washed with 2 mL of PBS (10 mM), followed by SWV measurement. The resulting analytical curve is presented in figure 10.

It is noticeable that the bulk detection method showed a good response for the detection of PS particles (R² = 0.96). However, due to the limited number of results, this method still needs to be tested in various ways. Its potential advantages justify continued investigation, with future efforts focusing on optimizing certain parameters such as the interaction time between the protein and MNPs, as well as the incubation time on the device surface.

Statistical Comparison of Detection Methods

Based on the results obtained for each detection method tested, it is possible to evaluate them in terms of several figures of merit. For our analysis, we evaluated, in addition to the R² metric, the mean standard deviation (σ) and the sensitivity (S), shown in equations 1, 2, and 3, where the space of the N_i analyzed concentrations is represented by i and the N_j replicates at each concentration are given by j.

Therefore, based on the calculated values for the metrics, the linearity (R² normalized relative to the methods), the precision ((-1) * σ) normalized relative to the methods), and the sensitivity (S normalized relative to the methods) were analyzed in a more didactic manner. Thus, Figure 11 provides a comparative analysis of the detection methods.

In this analysis, we can observe how the first method evaluated (blank) is limited in terms of linearity (R² = 0.76) and precision (σ = 12.4 %), due to the absence of a recognition element. In contrast, the bind method using protein applied by drop casting showed a significant improvement in linearity (R² = 0.94), thanks to the presence of the BaCBM2 protein as a biological recognition element. However, it still exhibited poor precision (σ = 10.88 %), likely due to variability in the electrode modifications through this method.

The bind method with BaCBM2 linked to a self-assembled monolayer (SAM) of cysteamine achieved high linearity (R² = 0.99) and excellent precision (σ = 0.88 %), owing to both the biological recognition element and a more precise fabrication method. Finally, the bulk method demonstrated good linearity (R² = 0.96) and precision (σ = 4.39 %), indicating the potential of this approach. However, more tests are needed to confirm its reliability. In terms of sensitivity, the methods presented similar results.

Dissociation Constant

Based on the results of the electrochemical assays, we demonstrated that it is possible to estimate the dissociation constant involved in the protein-plastic interaction.

Since 1960, the analysis of the dissociation constant has been a studied method for evaluating the affinity between molecules through a binding assay 27,28. In electrochemical assays based on Square Wave Voltammetry (SWV), the measured signal often depends on the occupancy of target molecules or particles (T) on the working electrode surface. This surface coverage (Θ) is influenced by the concentration of particles in the medium ([T]), the maximum surface coverage (Θ_Max), and the interaction between the target and the surface, which is related to the dissociation constant (K_d). The general expression for this process can be represented by the Langmuir isotherm29, as shown in Equation 4.

Thus, the general graphs observed for this process can be analyzed in Figure 12. It illustrates a progressive surface filling with increasing target concentration. Additionally, when plotted against the logarithm of the target concentration, the graph shows an initial exponential region, followed by a linear region, and finally an asymptotic region at the end.

The Langmuir isotherm approximation can be particularly useful in describing the measurements obtained from the self-assembled monolayers (SAM) in the bind (BI) methodology, as seen in the experiments conducted with the electrochemical sensor. During the incubation of PS particles on the working electrode, the interaction of particles with the affinity proteins (BaCBM2) follows the relationship described by Equation 4. As a result, surface coverage—and thus the measured signal—depends non-linearly on the concentration of PS in the medium, explaining part of the behavior observed in the obtained results.

In the second analytical curve, obtained with the cysteamine modification (Figure 8), a highly linear relationship (R² = 0.99) between the measured signal and the logarithm of PS concentration was observed. This suggests that, for this configuration, the chosen concentration range falls within the linear region of Figure 12. On one hand, this may indicate greater measurement precision30, but on the other hand, it precludes the estimation of maximum surface coverage and thus the dissociation constant.

Conversely, the first analytical curve generated with the cysteamine modification shows some saturation of the signal as log([PS]) increases (Figure 8.B), suggesting that the selected concentration range for this configuration is closer to the asymptotic region. While this may affect measurement precision, it allows for the estimation of the dissociation constant involved in the interaction between PSNBs and the surface in this system. Thus, Equation 4 was applied to fit the data obtained from the first cysteamine modification, as shown in Figure 13.

Through this analysis, the dissociation constant (K_d) for PS particles binding to the surface of the working electrode was determined to be 15.13 mg/L. This value reflects the interaction between the anchored BaCBM2 proteins on the electrode and the plastic particles.

It is important to note that, although this value relates to the PS-protein interaction, it is influenced by various experimental factors such as temperature, stirring, incubation time, PS-gold interaction, protein-solution interaction, the reduced degrees of freedom of the protein after immobilization on the surface, among other factors. These variables complicate direct comparison with computationally predicted K_d values, for instance.

Nevertheless, despite the limitations when comparing computational data, this method of calculating the dissociation constant can be particularly useful for comparing the affinity of different proteins for PS particles, or more broadly, for any target molecule of interest.

Dissociation Constant

We presented the results obtained from scanning electron microscopy and X-ray photoelectron spectroscopy for the characterization of the device and the modifications with the BaCBM2 protein.

Scanning Electron Microscopy

In order to characterize the MNPs interaction with the surface of the electrode, as well as the reusability of the sensor, Scanning Electron Microscopy (SEM) imaging was used. The SU-8 boundary, responsible for delimiting the active area of the working electrode, was characterized in Figure 14 and shown to have a thickness of 4.7 μm.

Figure 15 demonstrates the effectiveness of using Ar plasma for cleaning the gold surface. This guarantees the quality of experiments that were carried subsequently on the same sensor.

Finally, the MNPs adsorbed on the sensor can be observed in Figure 16. The MNPs display an affinity with the SU-8 polymer, causing them to accumulate on the boundary of the working electrode (Figure 16A). In some instances, interactions between particles were observed, where they almost seem to merge and cluster, as in Figure 16B. Another interesting phenomena was the apparent marks left by particles on the gold surface. It is unclear if these patterns in Figure 16C are caused by the physical “pressing” of MNPs against the gold or potentially some other electrochemical mechanism.

X-ray Photoelectron Spectroscopy

To determine the composition of material deposited on the surface of the sensor, the technique of x-ray Photoelectron Spectroscopy (XPS) was utilized. This characterization allows us to determine whether the SAMs and the protein adhere properly to the gold surface after washing. Figure 17 shows the full survey spectrum of four samples: Clean electrode (Blank), electrode with cysteamine monolayer (Cysteamine SAM), electrode with cysteamine monolayer and BaCBM2 (Cysteamine SAM + Protein) and only the BaCBM2 adsorbed on the surface via drop casting (Protein Drop Casting). First we look at the peaks of the 4f orbital of gold atoms for reference.

Next, we analyzed the peaks provided by the 1s orbital of Nitrogen atoms. Nitrogen is present on both the protein and the cysteamine, but it is more abundant on the protein because of the number of amino acids. Figure 18 is consistent with this expected result, and shows both cysteamine and BaCBM2 remained fixed on the surface.

Finally, Figure 19 confirms this result by demonstrating the presence of the sulfur from the cysteamine monolayer on the surface. The drop casting also generates a sulfur peak due to the cysteine group of the protein itself, although less abundant than the SAM.

Experimental Information

We shared some experimental specifications regarding the manufacturing and cleaning processes of the sensors, electrochemical assays, and measurements used, with the aim of making the hardware more accessible for new projects.

Sensor Fabrication Process

chosen architecture for the electrochemical sensor, as well as the fabrication methodology, were based on the device proposed by Flavio M. Shimizu and Anielli M. Pasqualeti et al 25.

1. Glass Substrate Preparation: The sensor fabrication begins with a glass substrate (25.4 × 76.2 mm) (Fig. 20). First, the glass substrate is cleaned using N2 plasma.

https://static.igem.wiki/teams/5396/hardware-figures/glass.webp Figure 20. Glass substrates

2. HMDS Adhesion Layer: After cleaning, a 2.3 μm thick layer of HMDS adhesion promoter is deposited on the substrate by spin coating (4000 RPM for 30 seconds) (Fig. 21). The plate is then heated at 100 °C for 3 minutes.

https://static.igem.wiki/teams/5396/hardware-figures/hmds1.webp Figure 21. HMDS spin-coating

3. Photoresist Deposition: A layer of positive photoresist AZ4220 (Fig. 22) is applied by spin coating (4000 RPM for 30 seconds), followed by heating at 100 °C for 5 minutes.

https://static.igem.wiki/teams/5396/hardware-figures/resin.webp Figure 22. Positive photoresist AZ4220 deposition.

4. Photolithography and development: The sensor’s architecture is patterned using the mask aligner (MJB3, Karl Suss) with 15 seconds of exposure to light (Fig. 23). After exposure, the architecture is developed using AZ4000 developer for 3 minutes, followed by rinsing with water, nitrogen jet cleaning, and O2 plasma treatment (100W and 70 mT for 2 minutes).

https://static.igem.wiki/teams/5396/hardware-figures/align.webp Figure 23. Mask aligner (MJB3, Karl Suss) during light exposition process.

5. Metal Layer Deposition: After developing the positive resist, a 20 nm layer of chromium (Cr) and a 100 nm layer of gold (Au) are deposited onto the substrate (Fig. 24).

https://static.igem.wiki/teams/5396/hardware-figures/gold-plate.webp Figure 24. Substrates prepared for metal deposition.

6. Liftoff Process: The next step is the liftoff process, where the photoresist is dissolved in acetone, removing the Cr/Au layers that were on top of the resist.

7. Surface Cleaning and Hydrophilization: The surface is cleaned and hydrophilized using O2 plasma treatment (100W and 100 mT for 2 minutes).

- **[VIDEO OF PLASMA]**

8. SU-8 Deposition: A 5.4 μm thick layer of SU-8 is deposited by spin coating (1300 RPM for 25 seconds), followed by a pre-bake at 65 °C for 2 minutes and 95 °C for 5 minutes.

**[VIDEO OF SPIN COATING]**

9. Photolithography for SU-8: The defined pattern for the SU-8 layer is photolithographed using the mask aligner (MJB3, Karl Suss) with 15 seconds of exposure (Fig. 25), followed by post-baking at 75 °C for 1 minute and 95 °C for 3 minutes.

https://static.igem.wiki/teams/5396/hardware-figures/align2.webp Figure 25. Mask aligner (MJB3, Karl Suss) being operated.

10. Development of SU-8: The SU-8 pattern is developed using Metoxy-2-acetate developer (30 seconds for development, 30 seconds for rinsing).

11. Final Cleaning: The final step involves cleaning the surface with O2 plasma treatment (100W and 100 mT for 2 minutes) (Fig. 26).

https://static.igem.wiki/teams/5396/hardware-figures/plasma.jpeg Figure 26. Ar plasma cleaning of the chips.