Integration of Metal–Organic Polyhedra onto a Nanophotonic Sensor for Real-Time Detection of Nitrogenous Organic Pollutants in Water

The grave health and environmental consequences of water pollution demand new tools, including new sensing technologies, for the immediate detection of contaminants in situ. Herein, we report the integration of metal–organic cages or polyhedra (MOCs/MOPs) within a nanophotonic sensor for the rapid, direct, and real-time detection of small (<500 Da) pollutant molecules in water. The sensor, a bimodal waveguide silicon interferometer incorporating Rh(II)-based MOPs as specific chemical receptors, does not require sample pretreatment and enables minimal expenditure of time and reagents. We validated our sensor for the detection of two common pollutants: the industrial corrosion inhibitor 1,2,3-benzotriazole (BTA) and the systemic insecticide imidacloprid (IMD). The sensor offers a fast time-to-result response (15 min), high sensitivity, and high accuracy. The limit of detection (LOD) in tap water for BTA is 0.068 μg/mL and for IMD, 0.107 μg/mL, both of which are below the corresponding toxicity thresholds defined by the European Chemicals Agency (ECHA). By combining innovative chemical molecular receptors such as MOPs with state-of-the-art photonic sensing technologies, our research opens the path to implement competitive sensor devices for in situ environmental monitoring.

X-ray photoelectron spectroscopy (XPS) measurements were performed at room temperature with a SPECS PHOIBOS 150 hemispherical analyzer (SPECS GmbH, Berlin, Germany)) in a base pressure of 5x10 -10 mbar using monochromatic Al-Kα radiation (1486.74 eV) as excitation source operated at 300 W. The energy resolution was measured by the FWHM of the Ag 3d5/2 peak for a sputtered silver foil was 0.62 eV. The spectra were calibrated with respect to the C1s at 284.8 eV.

Matrix-assisted laser desorption/ionization-time of flight (MALDI-TOF) mass spectrometry
Nuclear magnetic resonance (NMR). All 1 H NMR spectra were recorded using a Bruker Avance NEO 300 NMR spectrometer at 25 °C. Chemical shifts (δ) are reported in ppm.

S1.3 Fabrication and functionalization of the BiMW sensor
BiMW sensor chip (1 x 3 cm; Figure S1) was fabricated at a wafer-scale using silicon photonic technologies at the ICTS/cleanroom from the IMB-CNM (CSIC). 3 Each chip integrates an array of 20 independent bimodal waveguides and 12 reference waveguides. Its working principle relies on the behavior of two transverse electric polarized light modes propagated through a waveguide.
Briefly, light from a polarized diode laser (λ = 660 nm, HL6545MG, Thorlabs) is first confined through the waveguide core (rib thickness = 150 nm, height = 2 nm) in a single mode (fundamental). After a specific distance, the fundamental mode is coupled into a bimodal section through a step junction that allows the first propagating mode to emerge. These two modes travel          In the calibration curves, each signal corresponds to the mean ± SD of triplicate measurements on different batches of BiMW sensor chips.

Data analysis
Data were analyzed using Origin 8.0 (OriginLab) and GraphPad Prism (Graphpad Software, US).
Calibration curves were plotted as mean ± SD (standard deviation) of the accumulated biosensor response signals in triplicate (after signal stabilization at t = 1000 s) versus the accumulated BTA/IMD concentrations. The data were fitted to one-site specific binding model regression according to the following formula (eq.1): y= (A •X)/((B+X)) (Eq. 1) where y is the sensor response, X is the concentration of BTA/IMD, A is the extrapolated maximum number of receptor sides on the surface and B is the equilibrium binding constant which corresponds to the analyte concentration needed to achieve half-maximum receptor sides occupied at equilibrium.
The limit of detection (LOD) is defined as the minimum variation of analyte concentration that produces a signal at least three times higher than the detectable noise of the system which corresponds to the standard deviation (SD) of the baseline in between measurements. Therefore, LoD was calculated as the concentration corresponding to the blank signal plus three times its SD.