Silicon Photonics for Biosensing Applications

Introduction

Biosensing devices that can rapidly and accurately detect biomolecules have become increasingly important in fields ranging from medical diagnostics to environmental monitoring. While biological organisms have evolved highly specialized biomolecular recognition mechanisms, such as antibodies, enzymes, and nucleic acids, integrating these sensitive bioreceptors into robust and practical sensing devices requires careful engineering.

One of the key components in biosensing devices is the signal transducer, which converts the biorecognition event into a measurable signal. Optical transduction is one of the most widely used methods, as it can take advantage of the diverse spectroscopic properties of materials. Silicon photonics, in particular, has emerged as a powerful platform for optical biosensing due to the unique optical properties and fabrication capabilities of silicon-based microstructures.

This tutorial provides an overview of the different silicon photonic structures that have been explored for biosensing applications, including porous silicon thin films, Bragg mirrors, microcavities, waveguides, and photonic crystals. The capabilities and limitations of each architecture are discussed, along with the various bioreceptor molecules and surface chemistry approaches that have been integrated. Key experimental results and performance metrics, such as detection limits, are highlighted to showcase the state-of-the-art in silicon photonic biosensors.

Bioreceptor Molecules

The vast majority of biosensing devices leverage bioreceptor molecules that are either derived from living organisms or engineered to mimic their function. Common bioreceptors include:

• Antibodies/Antigens: Antibodies are complex proteins that can selectively bind to specific target antigens based on size, shape, and chemical functionality (Figure 1c). The "lock-and-key" interaction between antibodies and antigens is a widely used biorecognition mechanism.

• Enzymes: Enzymes are catalytic proteins that can modify target molecules, often with high specificity. The enzymatic activity can be coupled to the biosensing transducer to achieve very high sensitivity (Figure 1e).

• Nucleic Acids: DNA and RNA oligonucleotides can hybridize to complementary strands, providing a versatile biorecognition event that can be detected optically (Figure 1d).

• Cellular Structures: Whole cells or cellular components, such as transport proteins and lipids, can also serve as bioreceptors for detecting toxins, viruses, and other biomolecules.

Figure 1. depicts the process of selective biomolecule capture: (a) a sensor surface with attached antibodies (bioreceptors) captures target antigens from a solution, while mismatched antigens are not captured; (b) a diagram showing common bioreceptor-target pairs, including antibody-antigen, DNA with complementary DNA oligo, and an enzyme with its target protein.

Surface Chemistry and Passivation

Integrating bioreceptor molecules with silicon photonic structures requires careful surface preparation and functionalization. For porous silicon, the as-fabricated hydride-terminated surface is highly reactive and susceptible to degradation in aqueous solutions.

Two common passivation approaches are (1) growing a stable oxide layer on the porous silicon, and (2) forming a robust silicon-carbon bond through hydrosilylation with alkenes or alkynes (Figure 2). The oxide layer can then be further functionalized with silane chemistry to attach bioreceptors. Hydrosilylation directly produces a more stable alkyl monolayer on the porous silicon surface.

Regardless of the passivation method, the goal is to create a surface that is inert to the surrounding environment while providing appropriate attachment points for the desired bioreceptor molecules.

Figure 2 Sample reactions for porous silicon surface passivation. (a) Hydrosilylation reactions, involving either alkene or alkyne reagents, form silicon–carbon bonds. (b) Silylation reactions attach silane molecule to oxidized porous silicon via linkage to surface oxygen.

Optical Reflectance Transducers in Porous Silicon

One of the most widely explored silicon photonic architectures for biosensing is porous silicon, which can be fabricated by electrochemical etching to produce a nanoscale porous matrix. The high surface area and tunable optical properties of porous silicon make it a versatile transducer material.

Single-Layer Thin Films: The simplest porous silicon biosensor utilizes the change in effective optical thickness of a single thin film upon binding of target molecules. As biomolecules infiltrate the porous matrix, the effective refractive index increases, causing a shift in the Fabry-Pérot interference fringes in the reflectance spectrum (Figure 1a,b). This can provide detection limits in the sub-picomolar range but is susceptible to unwanted surface etching in some cases.

Bragg Mirrors: By periodically varying the porosity during etching, porous silicon Bragg mirrors can be fabricated. The stop-band wavelength in the reflectance spectrum shifts as biomolecules infiltrate the multilayer structure, offering improved sensitivity over single-layer films.

Microcavities: Introducing a defect layer within a Bragg mirror creates a porous silicon microcavity, which exhibits a sharp resonance peak in the reflectance spectrum. Binding of target molecules in the defect layer results in a resonance shift, providing high sensitivity due to the narrow linewidth. Techniques to enlarge the microcavity pores have enabled detection of larger biomolecules.

• Rugate Filters: Instead of the stepwise porosity changes in Bragg mirrors, porous silicon rugate filters are fabricated by sinusoidally varying the porosity. This produces a narrow reflectance stop-band that is sensitive to biomolecule binding, including enzymatic digestion of immobilized peptides.

• Waveguides: Porous silicon waveguides, with a low-porosity guiding layer and high-porosity cladding, allow light to be coupled into and guided along the sensing region. Changes in the effective refractive index upon biomolecule binding can be detected by monitoring the coupling angle or resonance wavelength (Figure 3).

Figure 3. (a) Porous silicon waveguide consisting of a low-porosity waveguiding layer and a high-porosity cladding layer. The prism couples light at a specific angle α across the air gap and into the waveguide through an evanescent wave. When biomolecules infiltrate into the waveguide, the effective refractive index, and thus the angle at which light is coupled into the waveguide, change. (b) Cross-sectional SEM of a porous silicon waveguide with a 310 nm low-porosity layer and a 1330 nm high-porosity layer

Overall, porous silicon optical reflectance transducers have demonstrated detection limits ranging from sub-picomolar to micromolar, depending on the specific architecture and bioreceptor-target system. The ability to precisely control the porous structure provides a great deal of flexibility in designing sensitive and selective biosensors.

Optical Reflectance in Other Silicon Nanostructures

Beyond porous silicon, other silicon photonic structures have also been explored for biosensing applications.

• Ring Resonators: Silicon ring and disk resonators support high-Q whispering gallery modes that are sensitive to changes in the surrounding environment, enabling detection of small molecule binding, protein interactions, and cellular processes.

• Slot Waveguides: By confining light in a narrow low-index slot between high-index silicon regions, slot waveguides can achieve extreme field enhancement and sensitivity to surface binding events. Slot waveguide-based biosensors have detected antibody-antigen binding down to the ng/mm^2 level.

• Photonic Crystals: Two-dimensional silicon photonic crystals, with periodic arrays of air holes, can be designed with defect cavities to concentrate the optical field. The transmission or reflection spectrum of these photonic crystal sensors shifts in response to biomolecule binding, with predicted detection limits down to the fg scale (Figure 4).

Figure 4 Schematics of photonic crystal microcavities with (a) single hole defect, (b) multihole defect, (c) L3 defect, and (d) multihole L3 defect.

Figure 5 (a) Cross-sectional SEM view of a porous silicon microcavity. (b) Top view SEM of a porous silicon microcavity, including an inset of the surface pores with average diameter 88 nm. Note the erythrocytes on the surface, which are filtered out of the porous matrix and cross-linked to the surface via gluteraldehyde fixation and demonstrating the size exclusion properties of the sensor even within complex media.

Intensity-Based Transducers: Mach-Zehnder Interferometers

In contrast to the reflectance-based porous silicon and photonic crystal sensors, Mach-Zehnder interferometers (MZIs) detect biomolecular binding through changes in the interference intensity at the output. An MZI consists of a input waveguide that splits into a sensing arm and a reference arm, which are then recombined.

The sensing arm is exposed to the target biomolecules, causing a change in the optical path length and resulting in a phase shift between the two arms. This phase shift modulates the output intensity, providing a highly sensitive readout of the binding event. The built-in reference arm helps compensate for environmental fluctuations.

Fully integrated MZI biosensors have demonstrated detection of DNA oligonucleotides down to 300 pM, as well as antibody-antigen interactions with surface coverages less than 0.3 pg/mm^2.

Figure 6 (a) Optical image and (b) intensity spectra for the detection of an important gelatinase (MMP-2), often associated with metastatic potential in tumors. Using a porous silicon transducer resulted in a decrease in the detection limit by two orders of magnitude when compared with the standard clinical detection methods toward this enzyme.

Photoluminescence Transducers

In addition to optical reflectance and interference-based detection, the intrinsic photoluminescence of porous silicon has also been explored as a transduction mechanism for biosensing. Changes in the photoluminescence intensity or spectrum can be monitored upon binding of target biomolecules.

While less commonly used than reflectance methods due to greater measurement uncertainties, photoluminescence-based porous silicon biosensors have demonstrated detection of immunocomplexes, DNA hybridization, bacteria, and other analytes through photoluminescence quenching or spectral shifts.

Summary and Outlook

Silicon photonics has emerged as a powerful platform for label-free optical biosensing, offering a diverse range of transducer architectures that can be tailored to the specific target analyte and bioreceptor. Key advantages include the ability to precisely control the optical properties through nanoscale engineering, the potential for integration with microfluidics and electronic readout, and the leveraging of mature silicon fabrication processes.

As summarized in Table 1, silicon photonic biosensors have demonstrated detection limits spanning many orders of magnitude, from the attomolar to millimolar range, depending on the transducer design and biomolecular interaction. Ongoing research continues to push the sensitivity, selectivity, and integration capabilities of these devices to enable practical applications in areas such as medical diagnostics, environmental monitoring, and drug discovery.

Table 1 Summary Data for Silicon Photonic Biosensors

Bioreceptor	Transducer	Signal	Detection range	Detection limit	Reference
Antibody	MZI	Intensity change (phase shift)	10 µg/mL		Bronsinger 1997
Antibody	MZI	Intensity change (phase shift)		0.25 pg/mm²	Densmore et al. 2009
Antibody	PSi microcavity	Reflectance shift	2-10 mg/mL		Bonanno 2007
Antibody	PSi microcavity	Reflectance shift	0.5-2.5 mg/ml		Ouyang 2005
Antibody	PSi microcavity	Photoluminescence quenching	10-1000 µg/mL	10 µg/mL	Starodub et al. 1996
Antibody	PSi single layer	Reflectance shift	2.5 mg/ml		Dancil 1999
Antibody	PSi single layer	Reflectance shift	10 M	1 ng/mm²	Janshoff 1998
Antibody	Ring resonator	Transmission shift	0.1 ₐM to 1 µM	5 ₐM	Armani et al. 2007
Antibody	Ring resonator	Transmission shift	1-199 ng/ml	25 ng/ml	Washburn et al. 2009
Antibody	Slot waveguide	Transmission shift	2-75 pg/mL	16 pg/mm²	Barrios et al. 2008
Biomimetic	PSi rugate filter	Reflectance shift	0.01-1 µM		Kilian 2007
Cell structure	PSi microcavity	Photoluminescence shift		1.7 µg	Chan et al. 2001
Cell structure	Ring resonator	Transmission shift	0.005-0.5 mg/L		Wang 2009
Coenzyme	2D photonic crystal	Transmission shift	10mM		Buswell et al. 2008
Coenzyme	PSi microcavity	Reflectance shift	0-2 mg/mL	0.3 ng/mm²	Ouyang 2005
Coenzyme	Ring resonator	Transmission shift	10 ng/mL to10 ug/mL	10 ng/mL	De Vos et al. 2007
DNA	PSi double layer	Reflectance shift	1 pM to 10 µM	55 fg/mm²	Steinem 2004
DNA	PSi microcavity	Photoluminescence shift		1 µM	Chan et al. 2000
DNA	PSi microcavity	Photoluminescence qunching	10 µM		DiFrancia 2005
DNA	PSi single layer	Reflectance shift	10 nM to 1 fM	9fg/mm²	Lin 1997
DNA	PSi single layer	Reflectance shift	1 µM		Voelker et al. 2008
DNA	PSi waveguide	Reflectance shift	50 µM		Rong et al. 2008a
DNA	PSi waveguide	Reflectance shift	1-10 µM	42 nM	Rong et al. 2008b
DNA/virus	PSi microcavity	Photoluminescence shift		194.2 fM	Chan et al. 2000
Enzyme	PSi microcavity	Reflectance shift	1-40 µM	50 pg/mm²	DeLouise et al. 2005
Enzyme	PSi microcavity	Reflectance shift	4-15 µM		Ouyang 2007
Protein	2D photonic crystal	Transmission shift	10 pM to 0.1 ₘM	500 pg/mm²	Dorfner et al. 2009
Protein	PSi rugate filter	Reflectance shift	7-14 µM		Orosco 200+O8:T306

Looking forward, further innovations in silicon photonic structures, surface functionalization techniques, and integrated system design will likely lead to even more sophisticated and capability biosensing platforms. The unique optical properties and scalable fabrication of silicon offer tremendous potential for translating cutting-edge biosensing research into robust, reliable, and cost-effective devices.

Reference

[2] J. L. Lawrie and S. M. Weiss, "Silicon Photonics for Biosensing Applications,"