Fiber-Optic Sensor Tracks Marine Sulfide with High Stability

The device integrates a nanotransferred plasmonic metasurface onto the end facet of a multimode fiber, creating a specialized interface for detecting sulfur species in water. This approach addresses a significant limitation in conventional sulfide analysis, which typically relies on sample collection and laboratory testing, thereby hindering rapid on-site analysis and dynamic monitoring of marine ecosystems.

The study, published in Discover Nano, introduces a strategy that combines thermal nanoimprint lithography, electrodeposition, and a novel nanotransfer process. By using these techniques, the team fabricated ordered gold plasmonic nanoarrays and successfully transferred them onto the confined surface of an optical fiber. This method overcomes the challenges associated with direct nanofabrication on small fiber facets, which are often limited by low throughput and high equipment costs in traditional approaches.

To enable specific detection of sulfide, the researchers integrated a chitosan-copper functional layer onto the gold metasurface. This coating, referred to as CS-Cu, acts as a sulfide-responsive interface. When exposed to dissolved sulfide, the copper coordination centers within the chitosan network react to form sparingly soluble copper-sulfur species. This chemical interaction reduces the effective coordination sites and induces partial restructuring of the functional layer, which in turn alters the local optical environment surrounding the gold nanostructures.

The optical response of the sensor was characterized using a fiber-optic spectrometer with a broadband halogen light source. The team observed that exposure to sulfide produced a blue shift in the localized surface plasmon resonance wavelength. This shift is consistent with the chemical and structural changes occurring in the CS-Cu layer, including the restructuring of the coordination network. The sensor demonstrated a refractometric sensitivity of 271.7 nm per refractive index unit, indicating its high capability to detect changes in the surrounding medium.

In terms of quantitative performance, the sensor exhibited a logarithmic response to sodium sulfide-derived dissolved sulfide concentrations ranging from 0.05 to 10 mmol L⁻¹. The calibration curve, constructed using wavelength shifts measured after five minutes of reaction, showed a strong linear relationship with the logarithm of sulfide concentration. The coefficient of determination for this relationship was 0.9745, suggesting a reliable detection mechanism across the tested range. A preliminary detection limit of 0.08 mmol L⁻¹ was estimated using a three-sigma criterion based on the standard deviation of the blank response.

Characterization of planar model films using Fourier-transform infrared spectroscopy and X-ray photoelectron spectroscopy provided further insight into the sensing mechanism. FTIR spectra revealed changes in characteristic bands associated with amino, hydroxyl, and C-O groups after sulfide exposure. XPS analysis showed the emergence of new low-binding-energy features in the S 2p spectra, consistent with sulfide-like sulfur in copper-sulfur species. These results confirm that the functional layer undergoes significant chemical changes that drive the optical response of the sensor.

Structural analysis via contact profilometry indicated that the average film thickness decreased by approximately 40% after five minutes and 70% after ten minutes of exposure to 1 mM sodium sulfide. Although these measurements were performed on dried planar films, they collectively demonstrate that the CS-Cu layer undergoes substantial physical and chemical modifications upon sulfide exposure. The stability of the underlying plasmonic metasurface was also evaluated, showing that the optical response remained intact after multiple cycles of functional layer deposition and removal.

The environmental stability of the sensor was tested in high-ionic-strength solutions representative of marine conditions. The device maintained stable optical baselines during seven-day immersion in deionized water, 600 mmol L⁻¹ sodium chloride, 10 mmol L⁻¹ sodium bicarbonate, and 80 mmol L⁻¹ magnesium chloride solutions. No sustained unidirectional baseline drift or obvious signal degradation was observed in these media. After removal from the solutions, the resonance wavelength returned close to the initial air-state baseline, indicating that the immersion did not cause irreversible optical changes.

Selectivity experiments were conducted to evaluate the sensor’s tolerance to major seawater ions and potential interferents. The team tested various substances, including chloride, bromide, sulfate, sulfite, thiosulfate, bicarbonate, phosphate, nitrate, metal ions, and humic acid, at concentrations substantially higher than their typical levels in natural seawater. After 24 hours of exposure, most interferents produced no obvious resonance wavelength shift. However, noticeable responses were observed for thiosulfate and humic acid, likely due to their sulfur-containing nature or nonspecific surface adsorption.

Recovery experiments in artificial seawater further validated the sensor’s accuracy in complex matrices. Recoveries were 94.7%, 99.3%, and 76.0% at sulfide concentrations of 5, 1, and 0.5 mmol L⁻¹, respectively. The lower recovery at 0.5 mmol L⁻¹ indicates a negative bias at this concentration, highlighting the need for improved low-concentration precision. Device-to-device reproducibility was also assessed, with a coefficient of variation of 4.6% for refractometric sensitivity across three independently fabricated sensors. Response reproducibility was concentration-dependent, deteriorating at the low end of the tested range.

The development of this nanotransferred fiber-optic plasmonic platform offers a promising solution for in situ marine sulfide monitoring. By combining a compact probe format with remote optical interrogation, the sensor avoids the need for reagent addition during measurement. This feature is particularly advantageous for deep-sea or remote environmental monitoring where sample handling and instrument maintenance are challenging. The ordered gold nanoarray provides a controlled plasmonic interface, enhancing device-to-device consistency compared to sensors relying on nanoparticle assembly.

Despite these advancements, the study acknowledges several limitations that require further investigation. The current detection limit and response speed could be improved through optimization of the nanostructure and CS-Cu formulation. Additionally, while the sensor showed stability in representative ionic solutions, further validation in real seawater samples and long-term field conditions is necessary. The authors emphasize that improved low-concentration precision, matrix recovery, and independent detection-limit validation are needed before broader quantitative environmental use. This work lays a foundational step toward practical, real-time monitoring of marine sulfur cycling and environmental quality.

The fabrication process relies on a multi-step sequence to overcome the geometric constraints of the fiber end facet. The team began by preparing a 3.5 wt% solution of poly(butyl methacrylate) in anisole, which served as the thermal nanoimprint resist. This resist was spin-coated onto indium tin oxide-coated glass substrates and patterned using a silicon mold fabricated via interference lithography and reactive ion etching. To facilitate the subsequent transfer, the silicon molds were treated with a silane agent to reduce adhesion during demolding. After thermal nanoimprinting at 140 °C, the imprinted resist pattern was separated from the mold. Gold nanostructures were then selectively grown within the imprinted template through electrodeposition at a current density of 2.5 mA cm⁻². This step produced ordered gold nanoarrays with the designed periodic geometry on the conductive glass substrate.

To move these nanoarrays from the glass to the fiber, the researchers employed a poly(vinyl alcohol) film as a temporary transfer carrier. The gold nanoarrays were coated with a 10 wt% PVA solution, which was then peeled off the glass substrate. The PVA-supported metasurface was placed onto the fiber end facet, which had been pre-coated with a UV-curable adhesive. After curing the adhesive under ultraviolet illumination, the PVA layer was dissolved in water, leaving the gold metasurface attached to the fiber. This nanotransfer strategy achieved a high fidelity of 99.7 ± 0.23% across six independently fabricated sensors, demonstrating that the process is reproducible and scalable for constructing fiber-optic sensors.

The optical performance of the bare sensor was validated through finite-difference time-domain simulations and experimental measurements. Simulations based on structural parameters extracted from scanning electron microscopy images showed pronounced electric field enhancement around the gold nanostructures under resonant excitation. This localized plasmonic contribution provides the physical basis for detecting refractive-index variations near the sensing interface. Experimentally, the sensor exhibited a linear dependence of the resonance wavelength on the surrounding refractive index, with a coefficient of determination of 0.9863. The refractometric sensitivity was calculated to be 271.7 nm RIU⁻¹, with a coefficient of variation of 4.6% across three independently fabricated sensors, confirming the reliability of the optical transduction platform.

The preparation of the sulfide-responsive layer involved dissolving chitosan in an acetic acid solution and adjusting the pH to 5.5–6.0. A copper sulfate solution was then added to create a precursor with a final Cu²⁺ concentration of 110 µM. This precursor was drop-cast onto the fiber sensor tip and dried at 25 °C and 60% relative humidity for 12 hours. The resulting CS-Cu layer has a refractive index of approximately 1.49 to 1.50 in the visible range, which is higher than that of water. This difference in refractive index contributes to the local dielectric environment of the gold nanostructures, ensuring that the functional layer remains optically coupled with the plasmonic interface after modification.

For the sulfide sensing experiments, sodium sulfide solutions were prepared and standardized using potassium permanganate titration to ensure accurate concentration values. The sensor was immersed in these solutions, and reflection spectra were recorded continuously with a sampling interval of one minute. The dynamic spectral evolution was monitored to evaluate the response performance. The calibration curve was established from the resonance wavelength shifts measured after five minutes of sulfide exposure, a fixed readout time that provided differentiation among the tested concentrations. This approach allows for quantitative detection of sulfide in aqueous media, with the optical response directly linked to the chemical restructuring of the CS-Cu layer.

Subject of Research: Development of a nanotransferred plasmonic metasurface fiber sensor for in situ sulfide monitoring in marine environments

Article Title: Nanotransferred plasmonic metasurface fiber sensor for in situ sulfide monitoring in marine environments

Article References: Liao, C., Zhu, H., Guan, B., Xu, Y., Chen, S., Ye, X., Zhou, J., & Cai, J. (2026). Nanotransferred plasmonic metasurface fiber sensor for in situ sulfide monitoring in marine environments. Discover Nano, 21(1), Article 610. https://doi.org/10.1186/s11671-026-05020-8

Image Credits: AI Generated

DOI: 10.1186/s11671-026-05020-8

Keywords: plasmonic sensing, fiber-optic sensors, sulfide monitoring, marine environments, nanotransfer lithography, chitosan-copper, environmental sensing, LSPR, optical detection, water quality, Nanotransferred, plasmonic