Photonic-plasmonic microcavity for ultrasensitive protein detection
Measuring proteins in real-time down to fM solution concentration levels
WGM sensors derive their unprecedented sensitivity from the use of high quality-factor (Q-factor) optical resonances to monitor wavelength shift signals upon binding of biomolecules or nanobeads to the resonator surface. Even a single virus could be detected. Yet, if e.g. a single protein molecule shall be detected, the sensitivity has to be boosted. There have been several approaches, such as the generation of hot spots using a hybrid photonic-plasmonic sensing concept with a gold nanoparticle (NP) layer coupled to a WGM biosensor. However, there are some drawbacks: First, measurements cannot be done directly in solution. Second, real-time analysis is not possible since the proteins have to be pre-adsorbed on the NPs. Third, proteins are adsorbed randomly within the NP layer – outside of plasmonic field enhancements sites – which lowers the detection sensitivity.
A German-American team led by Frank Vollmer and Melik C. Demirel now proposes an alternative concept overcoming these problems: optical trapping of protein molecules at the sites of plasmonic field enhancements in a random gold NP layer. The stable integration of the microsphere WGM biosensor with a wetted gold NP layer is critical for achieving ultra-sensitive detection. Therefore, the silica microsphere cavity remains fixed on the Au NP layer. The Q-factor of the microsphere drops slightly but is still in the 105 range. After adding bovine serum albumin (BSA) solution at microliter of sample volumes, which enters the NP layer by capillary suction, the researchers observed an unexpectedly large significant wavelength shift.
The achieved sensitivity in the order of femtomole concentration levels was very surprising, and cannot be explained from random binding of the BSA molecules to the NP surface. Instead, the scientists hypothesized that the protein molecules prefer to bind to hotspot locations (i.e. closely spaced random NPs) of plasmon resonances excited in the NP layer due to optical trapping. To validate this hypothesis, they calculated the electromagnetic field distribution in a model NP layer using generalized Mie theory and simulated the expected wavelength shift due to the binding of proteins. Their calculations showed that, indeed, optical trapping of the proteins at highly sensitive plasmonic hotspot locations is essential for achieving high sensitivity in microcavity biosensing.
The achieved sensitivity in the order of femtomole concentration levels was very surprising, and cannot be explained from random binding of the BSA molecules to the NP surface. Instead, the scientists hypothesized that the protein molecules prefer to bind to hotspot locations (i.e. closely spaced random NPs) of plasmon resonances excited in the NP layer due to optical trapping. To validate this hypothesis, they calculated the electromagnetic field distribution in a model NP layer using generalized Mie theory and simulated the expected wavelength shift due to the binding of proteins. Their calculations showed that, indeed, optical trapping of the proteins at highly sensitive plasmonic hotspot locations is essential for achieving high sensitivity in microcavity biosensing.
Original publication
Santiago-Cordoba, M. A., et al.; J. Biophotonics 5(8-9), 629-638 (2012)