Shining Light on the Deep—High Quality Factor Metasurfaces for Real-Time Ocean Observation

Phytoplankton are microscopic photosynthetic organisms responsible for half of the global photosynthetic carbon fixation and are an essential part of the earth’s carbon cycle. But, under certain conditions, phytoplankton can undergo explosive growth forming dense blooms that can cover hundreds of square kilometers and can release powerful biotoxins (phycotoxins) that contaminate drinking water sources, harming humans and wildlife. Understanding how environmental drivers impact phycotoxin production is key to advancing climate resilience but remains an outstanding challenge due to the difficulty of measuring toxin prevalence and persistence in situ. Predominant methods of studying phycotoxins are based on mass spectroscopy, polymerase chain reaction, and DNA/RNA sequencing. These lab-based methods are costly, require sophisticated infrastructure, and lack remote in situ detection capabilities. Current in situ approaches, such as surface plasmon resonance (SPR), are limited by low sensitivity, poor quantification range, and are challenging to scale.<br/><br/>Here, I will describe the development of a nanophotonic sensor for in situ aquatic toxin detection using surface-functionalized high quality factor (high-Q) silicon metasurfaces. Our metasurfaces are composed of sub-wavelength silicon nanobars with dimensions 500 nm x 600 nm x 160 nm on sapphire substrates. First, we show that by introducing small biperiodic perturbations to the lateral block dimension, free space radiation can be coupled into guided mode resonances to produce high-Q resonances with Q factors exceeding 1,000 in aquatic environments. We demonstrate in simulation and experiment that by varying the biperiodic perturbation from 10 nm to 50 nm, the quality factor can be modulated from a few hundred to nearly 10,000 and that the local electric field can be driven to the silicon surface, increasing the overlap between the probe field and the binding surface by nearly four-fold. We show that long resonance lifetimes together with the increased field penetration at the binding surface results in strong spectral shifts of the resonance mode due to small perturbations to the local dielectric environment. We fabricate our metasurfaces by e-beam lithography and demonstrate that the optical responses of individual resonators can be spatially resolved as individually addressable ‘pixels’ and simultaneously read out as a change in resonance wavelength or as a change to the scattering intensity on a 2D InGaAs CCD array using a home-built infrared optical reflection microscope in a cross-polarized reflection configuration.<br/><br/>We then apply this platform to the detection of two aquatic phytotoxins: domoic acid, a neurotoxin responsible for human and wildlife mortalities, and microcystin, a liver toxin produced by cyanobacteria that poses a threat to drinking and agricultural water supplies. We selectively target domoic acid and microcystin using tailored surface functionalization of self-assembled monolayers (SAM) and a competitive antibody binding assay. We show that domoic acid and microcystin can be directly crosslinked to the metasurface through a step-wise covalent silanization. We observe in both calculations and experiment that consecutive molecular layers generate 0.5 nm - 4 nm shifts to the resonant wavelengths as sequential layers bind to the resonator surface. We then describe selective and quantitative toxin detection using a competitive antibody binding assay, where the metasurface optical signal is inversely proportional to the toxin concentration. Finally, I will discuss the integration of our high-Q metasurfaces with the Environmental Sample Processor (ESP), an autonomous robotic water sampler developed at the Monterey Bay Aquarium Research Institute (MBARI), that offers a pathway for in situ phycotoxin detection, processing, and analysis.