Levodopa Sensing with a Nanosensor Array via a Low-Cost Near Infrared Readout

Abstract

Near infrared (NIR) signals are beneficial for biomedical applications due to reduced light absorption, scattering, and autofluorescence in this range, which promises higher signal-to-noise ratios (SNR). Single-walled carbon nanotubes (SWCNTs) fluoresce in the NIR (800–1700 nm) and serve as building blocks for biosensors. To quantify the benefits of NIR fluorescence biosensing, we simulate the SNR considering wavelength-dependent scattering/absorption, autofluorescence, dark currents, and excitation background. We also compare Si and InGaAs PIN phototdiodes (pn diode with an additional intrinsic layer) as detectors for the NIR region. The simulation shows that the SNR of fluorophores in the NIR is higher, but InGaAs detectors are outperformed by Si detectors in the short NIR (<1050 nm). This was also validated in experiments with (6,5)-SWCNTs (emission 990 nm), showing a 1.2-fold higher SNR for Si PIN photodiodes. Next, SWCNTs were chemically modified to create sensor arrays/barcodes that detect levodopa. Monitoring levodopa blood levels is a crucial step for personalized Parkinson’s disease treatment. We then combine nanosensors and detectors to engineer a portable low-cost fluorescence reader that scans (6,5)-SWCNT sensor barcodes. It detects levodopa at relevant concentrations (10 μM) in human blood serum. Thus, we combine NIR fluorescent sensors with high SNR and low-cost Si detectors to make use of beneficial NIR signals, which opens opportunities for point-of-care applications.