Towards Integrated Surface-Enhanced Raman Spectroscopy of Anionic Pollutants in Water
Pablo Muñoz Galindo is a PhD student in the research group Optical Sciences. Supervisor is prof.dr. S.M. García Blanco from the Faculty of Science & Technology.
Ensuring the quality of water reserves is essential to guarantee not only the safety of the citizens of a country but also the prosperity of a nation. At the same time, human activity produces often contamination of water reserves as a side-effect. Contamination events generate important economical and health consequences such as restricted access to water for bath or consumption, intoxication, destruction of local fauna and flora, and reduction of the trust of the users in the water companies. Traditional drinking water quality monitoring by the water providers requires periodic sample collection, preparation and analysis in specialized laboratories. Consequently, the first alarm after a contamination event is often given by the consumers and not by the water companies. Sensor technology and “smart water-supply networks” can help to reduce the time response to contamination events. The ideal sensor to be integrated in a smart water supply network should have low-size, low-cost, low-maintenance and long lifetime. Furthermore, the need of sample preparation should be avoided, and the sensor should provide early detection and identification of the pollutant agents.
The experiments described in this thesis were carried out as part of the WaterPrint project, which pursued the design and fabrication of an optical sensor for the real-time detection and fingerprinting of pollutants in the water distribution network. The thesis is organized in 7 chapters.
Chapter 1 (introduction), presents the necessary theory to understand the different experiments contained in the subsequent chapters as well as the state of the art in waveguide enhanced Raman spectroscopy and waveguide enhanced SERS. It also introduces the general structure of the thesis.
Chapter 2 (surface modification of SERS substrates) demonstrates surface-enhanced Raman spectroscopy (SERS) of anionic pollutants in water using modified commercial gold-based SERS substrates. The limit of detection (LOD) of the analytes of interest (i.e., perchlorate, nitrite and nitrate anions) with conventional Raman spectroscopy turned out to be insufficient to detect these pollutants below legal maximum concentrations. The usage of SERS substrates alone did not improve the signal intensity due to the hydrophobicity of the substrates and low affinity of the gold material for the analytes of interest. Modification of the SERS substrates using Cys, DMAET and MMP self-assembled monolayers improved the achievable LOD by 2 orders of magnitude while using 1400 times lower power. The samples coated with cysteamine (Cys) showed the highest performance for the three analytes of interest. Experiments with drinking water samples revealed some of the challenges that must be tackled before a SERS-based sensor can be implemented in the field: (1) the existence of competing anions that may reduce the measured intensity for the analytes of interest and (2) the degradation of the substrate material and coatings when in continuous contact with water samples.
Chapter 3 (preconcentration devices based on isotachophoresis) explores the fabrication of different depletion-zone Isotachophoresis (dzITP) devices based on polydimethylsiloxane (PDMS) and Nafion®. The devices included multiple microchannels intersected by a perm-selective membrane formed with Nafion®. Different techniques to deposit the Nafion® resin were evaluated including deposition using a PDMS stamp and deposition in a cut made with a razor-blade. The fabricated devices were used to separate and concentrate the ions contained in a mixture of fluorescein and phosphate-buffered saline (PBS) buffer solution as a first effort to understand the technique and evaluate its application to other type of analytes such as perchlorate, nitrite and nitrate anions. The samples were characterized with an inverted microscope (Olympus IXS1) and an UV mercury lamp. The results showed a 20-fold best-case preconcentration rate. Such a device could help to push the LOD for perchlorate in the SERS experiments introduced in chapter 2 down to concentrations below the legal limit. Further optimization is still required to improve the LOD of other analytes with lower Raman cross- section such as nitrite and remains an open task.
Chapter 4 (nano-antennas on non-conductive substrates) demonstrates the fabrication of gold plasmonic antenna arrays on fused-silica substrates using electron-beam lithography (EBL). The proposed fabrication route used a double metal layer lift-off process to tackle both the problems of charging during the EBL process and adhesion of the gold material for the antennas. Two metals that can take these roles in an interchangeable way were identified: chromium and titanium. Both metals have associated selective wet etching processes (i.e., 0.1% HF for titanium and Cr etcher for chromium) that allow to etch the anti-charging layer while preserving the adhesion layer. The anti-charging layer also contributes to protect any pre-existing devices on the substrate (e.g., integrated waveguides) and it allows for imaging of the processed samples using conventional scanning electron microscopy (SEM) after the lift-off process to confirm the topography of the resulting nano-antennas. Multiple bow-tie nano antenna arrays were fabricated on fused-silica substrates following the proposed method. The scattering spectra of these samples were explored using a custom dark-field microscopy setup. The obtained results were in agreement with finite-difference time-domain (FDTD) simulations.
Chapter 5 (waveguides for integrated Raman spectroscopy) demonstrated waveguide Raman spectroscopy using aluminum oxide (Al2O3), silicon nitride (Si3N4) and titanium oxide (TiO2) channel waveguides in both transverse electric (TE) and transverse magnetic (TM) polarizations. Different waveguide designs were created based on mode solver calculations (Lumerical MODE solutions). The waveguides were designed to operate both in TE and TM polarization at pump (785 nm) and Stokes (<880 nm) wavelengths. Chips with waveguides of varying width were produced using UV lithography. Specific fabrication process flows for each of the tested materials were introduced in the chapter. A custom Raman setup, developed at the Optical Sciences group, was used to characterize the samples. Measurements of the intrinsic background and with toluene as analyte were performed for both TE and TM polarization. The measurements were performed in back-scattering configuration. The results with Si3N4 waveguides excited with TM polarization reported the highest signal to noise ratio (SNR) in the study.
Chapter 6 (plasmonic antennas on integrated waveguides for SERS) explored the fabrication of plasmonic antennas on integrated waveguides for SERS. FDTD simulations were performed to understand the behavior of dipole and bow-tie antennas integrated onto TiO2 waveguides. The designed antennas showed SERS enhancement factors in the order of 5⋅103-5⋅105 and simulated insertion losses in the order of 1-1.8 dB/antenna. Different antenna arrays were fabricated on TiO2 and Si3N4 channel waveguides following the fabrication process flow proposed in chapter 4. Antennas with both chromium and titanium as anti-charging layer were successfully produced, confirming the interchangeable role of these two metals and their associated selective wet-etching processes. Scanning electron microscopy (SEM) images of samples used to perform a dose-test and optical microscopy images of the final samples confirm the existence of the nano-antennas. The resulting chips were functionalized with (3-Aminopropyl)trimethoxysilane (APTMS) and poly-L-lysine (PLL) to improve the sensing surface affinity for perchlorate. Preliminary results showed no detection of PLL, APTMS or perchlorate. The existence of the SiO2 over-cladding on the waveguides, the position of the sensing window in the center of the sample, the strong intensity of the intrinsic background of TiO2 and Si3N4 or the very weak input coupling achieved by the experimental setup possible causes that can explain the lack of signal from the analytes of interest or the deposited coatings.
Chapter 7 (conclusions and outlook) presents the main conclusions of this work and introduces multiple research directions for the continuation of the experiments presented in this thesis.
