Biological nanoparticles such as viruses and exosomes are important biomarkers for a range of medical conditions, from infectious diseases to cancer. Biological sensors that detect whole viruses and exosomes with high specificity, yet without additional labeling, are promising because they reduce the complexity of sample preparation and may improve measurement quality by retaining information about nanoscale physical structure of the bio‐nanoparticle (BNP). Towards this end, a variety of BNP biosensor technologies have been developed, several of which are capable of enumerating the precise number of detected viruses or exosomes and analyzing physical properties of each individual particle. Optical imaging techniques are promising candidates among broad range of label‐free nanoparticle detectors. These imaging BNP sensors detect the binding of single nanoparticles on a flat surface functionalized with a specific capture molecule or an array of multiplexed capture probes. The functionalization step confers all molecular specificity for the sensor’s target but can introduce an unforeseen problem; a rough and inhomogeneous surface coating can be a source of noise, as these sensors detect small local changes in optical refractive index. In this paper, we review several optical technologies for label‐free BNP detectors with a focus on imaging systems. We compare the surface‐imaging methods including dark‐field, surface plasmon resonance imaging and interference reflectance imaging. We discuss the importance of ensuring consistently uniform and smooth surface coatings of capture molecules for these types of biosensors and finally summarize several methods that have been developed towards addressing this challenge.

Single-Particle Interferometric Reflectance Imaging Sensor (SP-IRIS) enables optical visualization and characterization of individual nanoparticles without any labels. Using this technique, we have shown end point and real-time detection of viral particles using laminate based active and passive cartridge configurations. Here, we present a new concept for low-cost microfluidic integration of the sensor chips into compact cartridges through utilization of readily available Si fabrication technologies. This new cartridge configuration allows simultaneous detection of individual virus binding events on a 9-spot microarray, and will provide the needed simplicity and robustness for routine real-time operation for discrete detection of viral particles in a multiplex format.

Protein microarrays have gained popularity as an attractive tool for various fields, including drug and biomarker development, and diagnostics. Thus, multiplexed binding affinity measurements in microarray format has become crucial. The preparation of microarray-based protein assays relies on precise dispensing of probe solutions to achieve efficient immobilization onto an active surface. The prohibitively high cost of equipment and the need for trained personnel to operate high complexity robotic spotters for microarray fabrication are significant detriments for researchers, especially for small laboratories with limited resources. Here, we present a low-cost, instrument-free dispensing technique by which users who are familiar with micropipetting can manually create multiplexed protein assays that show improved capture efficiency and noise level in comparison to that of the robotically spotted assays. In this study, we compare the efficiency of manually and robotically dispensed α-lactalbumin probe spots by analyzing the binding kinetics obtained from the interaction with anti-α-lactalbumin antibodies, using the interferometric reflectance imaging sensor platform. We show that the protein arrays prepared by micropipette manual spotting meet and exceed the performance of those prepared by state-of-the-art robotic spotters. These instrument-free protein assays have a higher binding signal (~4-fold improvement) and a ~3-fold better signal-to-noise ratio (SNR) in binding curves, when compared to the data acquired by averaging 75 robotic spots corresponding to the same effective sensor surface area. We demonstrate the potential of determining antigen-antibody binding coefficients in a 24-multiplexed chip format with less than 5% measurement error.

Investigating the binding kinetics of small molecule analytes to larger ligands, such as proteins and antibodies, is a compelling task for the field of drug and biomarker development, as well as the food industry and agro-biotechnology. Here, we improve the limit of detection of the Interferometric Reflectance Imaging Sensor (IRIS), a label-free, highly multiplexed biosensor, to perform real-time affinity measurement of small molecules binding to immobilized antibodies in a microarray format. As the analytes bind to the surface probes, the biomass accumulation on the surface is quantified by measuring the optical reflectance from the layered Si/SiO2 chip through the solution, in a common-path interferometer configuration. As a proof of concept, label-free detection of biotin molecules binding to immobilized streptavidin probes is performed, achieving 1 pg/mm2 sensitivity through signal averaging in a shot noise limited operation. Furthermore, we apply the optimized sensor to the screening of a 20-multiplexed antibody chip (MW~150 kDa ligands) against Fumonisin B1 (MW = 721.8 Da), one of the most prevalent mycotoxins found in many cereal grains such as corn and wheat. The simultaneously recorded binding curves of the toxin to the multiplexed sensor yield a signal-to-noise ratio of ≈8 when noise reduction methods of spatial and temporal averaging are utilized.