Advancements in optical biosensing techniques: From fundamentals to future prospects

Biosensors have become ubiquitous in our day-to-day life due to advancements in different technologies that simplify human life. They consist of a light source, a multi-layered structure that converts the signal into a measurable form, and a detector system.¹ Biosensors can be classified on the basis of biorecognition elements and sensing mechanisms. Sensing mechanisms include electrochemical, optical, and mechanical, depending upon the transduction methods.^2,3 In case of electrochemical sensors, the chemical reaction is converted into an electrical signal using electrodes that changes the flow of electrons.⁴ An optical biosensor involves the interaction of light with a biorecognition element that modifies the input signal and generates a measurable signal.^5,6 Mechanical biosensors alter the physical quantity, such as pressure or motion, into electrical or detectable signals.⁷ Biorecognition elements include antibodies, aptamers, nucleic acids, and enzymes.⁸  Figure 1 shows the key components of a biosensor: biological recognition elements, transducers, different samples, and photodetectors to amplify the signal that can be measured using different platforms.

These biosensors have been explored in different areas, which encompass chemical, neurotransmitter, defense, bio-sampling, environmental, food, medical, and industrial based applications.^10,11 Progress in the biosensing field has been made year-wise, for example, enzyme-based biosensors in 1962, followed by new techniques for material synthesis, detection methods, microfluidic chips, and machine learning.¹² 

Biosensors play a significant role due to their ability to detect a variety of analytes with high specificity, sensitivity, multiplexing capabilities, and portability in real time, enabling them to detect diseases at early stages. Fueled by various applications of biosensors in different areas, global market reaches 24.9 × 10⁹ USD in 2021.¹³ Due to advancements in technologies driven by applications that include medical, environmental, and industrial areas, projected evaluation could reach 50 × 10⁹ USD by 2030. The key areas of biosensing include health care,¹⁴ environmental monitoring,¹⁵ pollutants,¹⁵ pathogens,¹⁵ food safety,¹⁶ and toxins.¹⁶ Various biosensing platforms have been reported in the food industry to sense contaminants such as micro-organisms, pesticides, and heavy metals.^17–19 Adulterants include melamine, fillers, and industrial chemicals. Common allergens include peanuts, tree nuts, milk, eggs, wheat, soy, fish, shellfish, sesame seeds, and sulfites.²⁰ Biosensors have also been used in finding the pathway of progression of diseases and effective treatment, not limited by diagnosis. They have also been explored in implantable biosensors that can continuously monitor the vibrant biomolecule levels and symptoms that could lead to personalized medicine using specific biomarkers as they maximize the chances of successful treatment.^21–23 

In case of optical biosensors, label-free plasmonic [surface plasmon resonance (SPR) and localized SPR], interferometric, colorimetric, surface-enhanced Raman scattering-based (SERS), and fluorescence-based sensors have been reported for different sensing applications. SPR excites the surface plasmon in thin films and measures refractive index (RI) change near the metal surface caused due to binding of the analyte, while LSPR excites plasmons in nanostructures or nanoparticles. Interferometric biosensors are label-free that measure the path length change or interferometric pattern due to attachment of the sensing analyte on the surface of a biorecognition element. In a colorimetric biosensor, as a target binds with the bioreceptor, color change is observed. In SERS, metallic nanostructures have been used to enhance the Raman scattering signal of the molecules. SERS is limited as it is associated with specific molecules/analytes due to Raman active modes. A fluorescent biosensor measures the fluorescent signal emitted by a labeled sensing analyte.

Prism-based SPR sensors are used in Otto and Kretschmann configurations, which have been explored for a wide range of applications. In the Otto configuration, light wave couples with SPs at an angle through few nm gaps, while the Kretschmann configuration couples light through the metal directly to SPs, as shown in Fig. 2(a). The Kretschmann configuration is commonly used as compared to the Otto configuration due to the easy fabrication process and provides high sensitivity values as coupling between light and SPs is strong. However, it is important to mention that the Otto configuration provides tunable resonances as airgap can be adjusted while it is limited in the Kretschmann configuration. Recently, a fiber sensor based on Kretschmann’s configuration has been reported with high sensitivity, portability, miniaturized size, remote sensing ability, low cost for bulk production, strong electromagnetic coupling, and immunity toward electromagnetic interference. In fiber SPR configuration, the prism is replaced with a core–clad fiber that excites the SPs; as the target binds to the bioreceptor, a sharp dip in the output intensity of light was observed at a wavelength called resonance wavelength. Figure 2(b) shows the resonance and angular wavelength shifts. In the figure, θ_SP1 and θ_SP2 represent the resonance angle in case of reference and target analytes or that can shift with a change in the refractive index of the analyte in angular interrogation and λ is the resonance wavelength for a specific analyte refractive index that shift to another value with a change in the analyte refractive index in wavelength interrogation. LSPR sensors involve the use of metallic nanoparticles of few nm as compared to thin metal films. As the target binds to the biorecognition layer, it shows a dip in output spectra at a resonance angle that alters the resonance angle or wavelength shift that can be measured using a detector. LSPR sensors can detect a small analyte volume and have high sensitivity due to more localization of SPs, higher multiplexing ability, and low cost due to NPs as compared to bulk SPR sensors. Figure 2(b) shows the prism and fiber based SPR and LSPR sensor configurations, in which light is incident via a prism and passes through the metal to excite plasmons. EW and SP wave represent the evanescent wave and the surface plasmon wave, respectively. In interferometric biosensors, a reference beam and a sample beam are used; the sample beam interacts with the bioreceptor and causes a binding that produces phase changes or intensity changes. An interference pattern with the reference beam is obtained, which is analyzed using the detector. There are numerous types of interferometric sensors, which include Michelson interferometer, Mach–Zehnder Interferometer (MZI), Fabry–Pérot interferometer-based biosensor (FPIB), and Sagnac interferometer. A Michelson interferometric sensor is the simplest structure among four types of interferometric structures that utilize a single beam splitter to divide a beam of light into two paths. It measures the path difference caused by one beam that reflects off from a fixed mirror and another from a movable mirror. MZI also utilizes two mirrors and two beam splitters that split the light beam into two paths using interference by division of amplitude. One beam splitter is used to split the beam, and the other is used to combine the beam. In MZI, fiber biosensor light is launched using a single mode fiber that split into arms and it passes through the analyte placed in the sensing arm. FPIB involves coupling of light from the source using two mirrors in the cavity that trap the beam that suffers multiple reflections. A Sagnac interferometer uses a beam splitter and two mirrors and involves the propagation of light in a closed loop in a rotating frame. In a Sagnac interferometric sensor, split light propagates in opposite directions inside the ring and rotation causes the difference in distance traveled. Traveled light combines, having phase difference proportional to the rate of rotation. In case of RI sensors, the fiber section sensitive to changes can measure any changes in the surrounding RI. Among all four configurations, MZI and FPI are commonly used and widely explored for various biochemical applications. Figure 3 shows the Michelson, MZI, FPIB, and Sagnac interferometers’ structure-based sensor configuration. Colorimetric sensors rely on a color change upon binding events. This color change can be due to several mechanisms, including aggregation of nanoparticles, enzymatic activity involving colored substrates, or changes in the properties of indicator dyes, as shown in Fig. 4(a). SERS sensors utilize metallic nanostructures to enhance the Raman scattering signal of molecules adsorbed on their surface. Binding events on the biorecognition layer bring target molecules close to the nanostructures, significantly amplifying their Raman signal, which provides a fingerprint for identification, as shown in Fig. 4(b). Fluorescence sensors use biorecognition elements labeled with fluorescent molecules. Figure 4(c) shows that binding events can alter the fluorescence intensity, lifetime, or polarization due to changes in the fluorophore’s environment.

Optical biosensors have been explored on the basis of different phenomena, such as light reflection, polarization, bending, interference, diffraction, absorption, light emission properties (fluorescence and luminescence), and scattering. Advances in technology that involve the use of prisms, fibers, waveguides, and gratings as a substrate; advancements in synthesis methods; and improvements in bioreceptor functionalization have made biosensors a promising tool for various applications. This section discusses the details of the reported work, advancements, and improvements in the different types of optical biosensors.

Plasmonic biosensors can primarily be classified in two types: SPR and LSPR. SPR involves propagating SPs at metal–dielectric surfaces, and LSPR sensors are based on localized SPs on the surface of metal nanoparticles. Figure 5 shows the output spectra of four interrogation methods depending on the parameter that is varied in each technique: (1) angular interrogation, (2) spectral interrogation, (3) intensity interrogation, and (4) phase interrogation.²⁴ The angular interrogation method involves variation of the angle of incidence; a reflected/transmitted light intensity is observed to obtain the resonance angle θ_SPR. At θ_SPR, the intensity of light encounters a sharp decrease indicated by a dip in output spectra, as shown in Fig. 5.

Wavelength interrogation involves measuring variation of the wavelength of light at a fixed angle, and a dip is observed in output spectra at resonance wavelength λ_SPR. The wavelength interrogation method does not require moving parts and eliminates the need of scanning over a wide range of angles. This method requires a spectrometer that increases its cost.²⁵ The intensity interrogation method measures the intensity of output light at a fixed incidence angle. The phase interrogation method involves the measurement of phase difference of output light under the effect of p- and s-polarized incident light. Each method has different advantages and disadvantages, and its choice depends upon the requirement of applications.³ 

A variety of plasmonic (SPR AND LSPR) biosensors have been reported in recent years. A prism-based SPR sensor fabricated using Au, DNA tetrahedrons, and fluorescent dye has been reported to detect the human immune deficiency virus with a LOD of fM range.²⁶ Various prism and fiber based SPR/LSPR sensors have been reported using graphene,²⁷ MoS₂,^28,29 black phosphorus,³⁰ WS₂,³¹ and graphene oxide/g-C₃N₄³² nanocomposites using Ag and Au metals to detect DNA hybridization,³³ miR let-7a,³⁴ cancer and its types,³⁵ cadmium, lead ions,³⁶ cryptococcus neoformans,³⁷ amyotrophic lateral sclerosis, E. coli³⁷ and other bacteria, and tuberculosis.³⁸ DOL, in case of different samples, varies from femtomolar (fM) to picomolar (pM) concentrations with high sensitivity values. SPR and LSPR sensors differ in excitation, electric field, sensitivity, and different types of instruments. In SPR, excitation of SPs occurs at M–D thin film interfaces with electric field confinement of 200–1000 nm, while in LSPR, it occurs at metal nanoparticles with field confinement in the range of 20–40 nm. An SPR curve is affected by the thickness of the metal film, angle, wavelength, and width of curve, while in Refs. 39 and 40, LSPR shape, size, and type of metal nanoparticles are shown to affect the absorption spectra. Due to wide field confinement, SPR detects the target molecules near, on the surface, and in bulk of the sample solution that provide incorrect positives. LSPR sign is confined to few nm to provide better selectivity sensitivity compared to other methods.^41–45 SPR sensor design involves the requirement of more precise control of the incident angle of light and light polarization to produce the SPR effect and better design sensing.²⁵ Plasmonic sensors detect an analyte by measuring changes in the analyte refractive index. Section II B discusses the colorimetric sensors that detect the presence of target analytes by the change in the color of solution.

In colorimetric sensors, metal nanoparticles (Ag or Au) are functionalized using bioreceptors to detect the specific target that aggregates nanoparticles' corresponding color change which can be detected visually or with photometer. Various colorimetric biosensor designs using different bioreceptors have been reported so far.^46–52 They include thiolated Au nanoparticles with hexadecyl trimethyl ammonium bromide and cysteamine capped for hepatitis C virus detection,⁵³ Au nanoparticles conjugated with DNA hairpin and polyethylene glycol to detect miRNA-203,⁵⁴ and Au nanoparticles functionalized using a snowball assembly with an enzyme that detects thrombin.^55, Figure 6 shows a colorimetric sensor constructed using a snowball assembly of palindromic DNA–AuNPs used to detect target DNA.⁵⁵ Various other biosensors that include a paper-based structure to detect miRNA, Pb2+, influenza A virus, COVID-19, estradiol, prostate cancer, and mycobacterium have been also investigated with a LOD of nM range.⁵⁶ 

Advantages of colorimetric sensors include easy detection through smart phones, tablets, and flatbed optical scanners; naked eye visibility; and low cost.⁵⁷ The color change in colorimetric sensors is visible to the human eye, while luminescence biosensors involve the use of light emitted by specific molecules or materials for detection purpose. In Sec. II C, luminescence-based biosensors and their applications will be explored in different areas.

Luminescence is emission of light by molecules or different materials through different processes, such as electrochemiluminescence, chemiluminescence, fluorescence, and bioluminescence. Luminescence biosensors do not require excitation devices, have low background noise, high sensitivity, and selectivity; and are widely used in biosensing. In these types of biosensors, as the target and bioreceptor molecules bind to each other, there is a change in the output signal, which is used as a working principle in biosensing applications.^58,59 In fluorescence, the absorbed light is emitted at a particular wavelength by molecules, and it has been explored in biomedical imaging, diagnostics of diseases, and environmental monitoring. In chemiluminescence, light is emitted due to a chemical reaction by exciting the molecule to a higher state. It is used to detect targets in immunoassays and DNA samples.³⁴ Bioluminescence and electrochemiluminescence emit light due to biochemical reactions inside the living organism and electrochemical reaction between a luminescent molecule and an electrode.^60–63 Various luminescent biosensors have been reported to detect different targets, for example, Salmonella enterica bacteria,^64, M. aeruginosa,⁶⁵ estradiol, PSA,⁵² CaMV35S promoter, lectin gene,⁶⁶ COVID-19,⁶⁷ SARS-Cov-2,^68, S. aureus,⁶⁹ hypochlorous acid,⁶⁶ Salmonella Typhimurium,⁷⁰ and miRNAs.⁷¹ In Sec. II D, interferometric biosensors that utilize interference patterns for detecting an analyte are explored.

Interferometric biosensors can detect the refractive change of 10⁻⁷ order that can detect pg/ml concentration of targets, such as toxins and thousands of viruses or whole cells.⁷² Interferometric biosensor designs have been explored for versatile biochemical detections.^73–79 These are limited by sensitivity due to a small path length change in case of biosamples, non-specific adsorption signal, variability in bioconjugate systems, temperature fluctuations, and material constraints as changes in the material do not cause changes in the limit of detection and figure of merit.⁸⁰ These biosensors have the potential to improve the current performance, such as by employing multiple detectors, as at present most biosensors involve a single detector in order to reduce cost, reduce thermal noise, and integrate components on a single chip and evanescent field using a 3D sensor matrix.⁸¹ 

Surface-enhanced Raman spectroscopy biosensors are used to enhance the Raman scattering signal with ultra-sensitivity and fingerprint of molecules.

A SERS biosensor is made up of 4-mercaptobenzoic acid and 5′-NH₂-ssDNA. Ag/TiO₂@3′-NH₂-ssDNA with a core of Au and a shell of Ag is used to detect miRNA-21. It can detect miRNA-21 with the detection limit of 0.75 fM and a linear range of 1.0 fM–1.0 nM.⁸² Au nano-flowers deposited on the substrate with streptavidin, cyanine-5, and a hairpin assembly with biotin-based SERS biosensors are used to detect genetically modified organism DNA in maize samples. These sensors can detect the DNA sample in the concentration range of 0.01–10000 fM and a detection limit of 6.31 aM.⁸³ A ¹⁸O stable isotope labeling based SERS biosensor that revealed changes in physiological information of cells to detect E. coli bacteria, is explored.⁸⁴ An Ag nano-cube-based SERS sensor, with the Raman reporter of 4-mercaptobenzoic acid, is used to detect the oral cancer DNA. A cancer sample can be detected with a LOD of 3.1 fM in the concentration range of 10 fM–0.1 nM and a RSD of 4.6%–7.6%.^85–87 Other SERS biosensors have been reported for various analytes.^85,88–98 Existing optical biosensing techniques are powerful but limited by several challenges. Therefore, researchers are exploring various other techniques to overcome the challenges with traditional biosensor designs.

Various biosensing techniques are emerging in recent years that provide better capabilities and high performance to detect the target. They include waveguides, surface-enhanced infrared absorption, integration with machine learning, and artificial intelligence. Waveguide-assisted biosensors make use of guided electromagnetic waves through narrow channels to measure the reflected/transmitted intensity with high sensitivity and can be integrated into miniaturized devices to develop point-of-care, label-free applications. Waveguide techniques include fibers, photonic crystals, and planar waveguides to guide the light along a specific path. It can confine the light inside the waveguide and a small volume, which enhances the light and target interaction yield that can detect the target with high performance parameters.^99,100 Surface-enhanced infrared absorption nanostructured metal surfaces are used to amplify the absorption of molecules in infrared region after interaction with nanostructures. In this technique, enhancement of light occurs due to the field created by interaction of metal-based nanostructures and infrared light that enhances specificity and sensitivity.¹⁰¹ It can detect the functional groups inside biomolecules providing excellent sensitivity and selectivity with characterization details. Machine learning provides a more detailed analysis of data belonging to the artificial intelligence branch that develops algorithms and models to provide more predictions about data.¹⁰² Artificial intelligence is a simulation technique using human intelligence with the help of algorithms, data, and computational power. These techniques have the capability to analyze the complex and large data produced by different biosensors and can establish correlations that are difficult to achieve with conventional techniques. It can also optimize the performance parameter in a better way and predict biological responses at an early stage in case of disease diagnosis.^103–105 

Bio-recognition elements are structures or biomolecules that specifically bind a target with high sensitivity and specificity. They play an important role due to the requirement of specificity in medical, industry, and other applications. Antibodies, aptamers, nucleic acids, enzymes, whole cells, phages, peptides, lectins, and molecularly imprinted polymers have been reported to be detected in various applications. Biorecognition elements are deposited on the surface of transducers in case of biosensor design, and it should effectively interact with a specific target molecule in the presence of other chemicals substances.

Antibodies or immunoglobulins (Ig) are Y-shaped protein structures made up chains of polypeptides that consist of heavy and light chains used to detect a specific antigen, as shown in Fig. 7(a).¹⁰⁶ An antibody is made up of the Fab region present at the top of a structure that binds the antigen and stem Fc region, which activates immune cells and their binding.¹⁰⁷ It can be polyclonal, monoclonal, or recombinant, depending upon the number of epitopes that it can bind. Monoclonal and polyclonal antibodies detect a single epitope and multi-epitopes.¹⁰⁸ Recombinant antibodies detect a single epitope due to their monoclonal nature produced in the lab synthetically using synthetic genes.¹⁰⁹ Polyclonal and monoclonal antibodies have advantages and disadvantages associated with them. A monoclonal antibody can be synthesized at a high cost, while polyclonal antibodies have high cross-reactivity issues. Monoclonal or recombinant antibodies bind only one epitope and, therefore, are used to design highly specific biosensors. The orientation of antibodies and immobilization plays a significant role in the detection of a target with high specificity and high performance.¹¹⁰ 

Enzymes are molecules that catalyze various reactions and mainly exist in the form of proteins.¹¹¹  Figure 7(b) shows the simple diagram of an enzyme–substrate active cite, in which a substrate is a molecule where non-absorbed enzymes interact to speed up the chemical reaction. Enzymes can be conjugated with antibodies, cells, and aptamers and, therefore, are preferred for labeling purpose to detect an analyte. Lactate dehydrogenase, horseradish peroxidase, glucose oxidase, and acetylcholinesterase enzymes are commonly used for various biosensing applications. Various enzymatic optical biosensors have been reported for different applications.^112–119 

Aptamers or chemical antibodies are high affinity sequences of single stranded RNA or DNA molecules that can fold into three-dimensional structures and can detect an analyte with high specificity.^120,121 Aptamers are produced using systematic evolution of ligands by an exponential enrichment (SELEX) procedure, an iterative procedure that consists of oligonucleotide sequences around 10¹⁵–10¹⁸ to generate excellent affinity to detect a target.¹²² Aptamers have tertiary structures with basis sequences bound through hydrogen bonding.^123–126 Aptamers can bind to a wide variety of targets [Fig. 7(c)], which include from small molecules to toxins and micro-organisms, and are more stable than antibodies as they can survive at high temperatures, in harsh environments, and across a wide pH range.³⁵  Figure 7(c) shows an aptamer sequence with a three-dimensional structure, which binds to the target cell through conformational recognition. Figure 8 shows the five most commonly used high-affinity DNA aptamer structures for various applications in food, medical, and environment-related industries. Aptamers have been explored in various applications and used in industries due to their high specificity and other chemical–physical properties.¹²⁷  Table I presents the various biosensor designs with aptamers for different applications.

A nucleic acid is a complementary sequence of DNA or RNA and transforms the genetic information as it involves information to produce an organism inside the body [Fig. 7(d)].¹³⁹ They are made of polymer chains of nucleotides and named as cytosine, guanine, adenine, thymine, and uracil. A DNA is a double-stranded helical molecule that uses thymine as a base, and a single-stranded RNA molecule makes use of uracil to form sequences.¹³⁹ A nucleic acid-based biosensor can be easily fabricated and operated, and it provides fast results with low-price and high performance to detect analytes. Numerous biosensor designs have been described to detect viruses, cancer cells, micro-organisms, and various biochemicals.^{89,114,140–144}

Bacteriophage is a virus family member that makes a copy of other bacteria used to detect the different micro-organisms associated with different diseases.¹⁴⁵ They consist of DNA or RNA as a genetic material, with a protein coating that can detect strains of bacteria specifically. Phages are very specific in nature and can affect a single bacterial species. T4 bacteriophage is an example that affects Escherichia coli bacterium.¹⁴⁶ A polyhedral head is made up of proteins and genetic materials that can have different shapes, and a hollow tube tail that helps in attaching to the host material. Based on reproductive cycles, phages are classified into two cycles: (1) lytic cycle and (2) lysogenic cycle. In the lytic cycle, phages first replicate the genetic material inside a host cell; in the second stage, they burst out and produce new phages. The lysogenic cycle involves a replicated genetic material inside the chromosome that is transferred to the daughter cell to produce new phages.^147–150 Phages are used to design the biosensor used to detect a variety of viruses, bacteria, and other organisms.^{37,81,151–155}

Molecular imprinted polymers are synthetic materials generated in the lab to detect the target with strong affinity, sensitivity, and selectivity.^156,157 Molecular imprinted polymer synthesis involves template, monomer, polymerization, and removal of the template processes. Small molecules called monomers that can interact with a template through chemical reactions are polymerized so that functional groups inside the monomer rearrange themselves to produce cavities with a polymer network system.^158–161 Molecular imprinted polymers employ a biosensor design that faces difficulty in commercialization due to the requirement of large-scale production, low selectiveness, and complete, though easy, template removal.¹⁶² 

Whole cells employ living microbial cells that include viruses, bacteria, and fungi that have not been fractionated and maintain their cell structure and organelles with functionality.^163,164 Various biosensor designs based on whole cells that have been reported so far exhibit good sensitivity and specificity values to detect different analytes.^155,165,166

Choosing the right biorecognition element among antibodies, enzymes, aptamers, and nucleic acids plays an important role for high specificity and sensitivity.¹⁶⁷ Bioreceptors should have the capability to bind the specific target molecule only with minimum interaction in the presence of multiple components in the target sample.⁸ Nucleic acids that consist of linear polymers of different nucleotides or nucleic acids are used to design bioreceptors that match the complementary DNA of the concerned individual. The matching strand can be detected with different transducing mechanisms. A snowball assembly-based biosensor was constructed using AuNPs functionalized with a DNA sequence and determined by the Nt.BbvCI enzyme. It is used to detect the thrombin and DNA samples by incubation of the hairpin DNA–AuNPs–target DNA with the BbvCI enzyme to cleave the target DNA that detaches the hairpin DNA from nanoparticles. The released hairpin DNA sequences bind themselves to another palindromic DNA sequence to form a snowball structure assembly. A sensor can detect a target with a LOD and concentration in the range of 8.3 nM and 10–100 nM.⁵⁵ 

The selection of materials for the sensor platform, including substrates, electrodes, metallic nanomaterials, and transducers, is critical. The materials should be biocompatible, stable, and capable of producing the desired signal.¹⁶⁸ Glass, silicon, and polymers are most commonly used as a substrate. Among metals and metallic nanoparticles in different sensing platforms, Au is chosen due to its stability toward oxidation, inertness toward chemical reactions, larger surface area, stability, and long-lasting nature in commercialization.

Various fabrication techniques, which include thin-film deposition (sputtering, evaporation, and chemical vapor deposition), photolithography, surface functionalization, and microfluidics, can be used to create biosensors. Challenges include maintaining the integrity of biorecognition elements and ensuring reproducibility.^169,170 Functionalization is done using self-assembled monolayers and biotin–streptavidin binding printing. Reproducibility, point of care, wearability, scalability, and cost-effectiveness are the main challenges in the fabrication and commercialization of biosensors.

Proper functionalization of the sensor surface with biorecognition elements involves controlling surface density and orientation to maximize binding kinetics.¹⁷¹ 

Immobilization techniques, such as physical adsorption, covalent binding, and entrapment, should be carefully chosen to ensure stability and functionality of the biorecognition element.^{4,25,80,172,173}

The selection of the correct biorecognition element and its immobilization, materials, and fabrication techniques to design the sensor play an important role in effective and reliable detection for various applications. For example, self-assembled monolayers can be used to design a sensor that has specific control over the surface of the sensor and orientation of molecules.

This section discusses combining various elements or bringing a laboratory on a chip, their miniaturization, and challenges in fabrication and commercialization of biosensors.

Lab-on-a-chip or biosensing on a chip is a miniature laboratory that joins the different functions of a laboratory on a single chip in a size of few square centimeters. This technology reduces the amount of chemical waste, requires minimum sample volumes (picoliters) for all the complex investigations, and has short response times. It also acts as an interface between traditional optical devices, such as microscopes, diodes, and microfluidic instruments.^174–178 The key feature of a lab-on-a-chip biosensor is the execution of multiple tasks simultaneously, which include sample injection, mixing, chemical reaction, isolation of substances, enrichment, and detection process.¹⁷⁹ It should also include pathways for fluids (microchannels), micropumps to push solutions through the microchannels, microscopic valves to regulate flow, and micromixers to mix the samples in the microchannels.¹⁸⁰ There are also certain disadvantages, which include capillary forces, roughness of surface, and interference effect from synthesis materials that dominates at a small scale. Detection methods at microscale level may not scale down as a convention detection system due to background noise increment and thedecrease in signal strength due to less detection area and low sensitivity.¹⁸¹ 

Wearable biosensors are in direct contact with the human body; therefore, it should be biocompatible.

Wearable biosensors can be integrated into smartwatches, fitness trackers, and patches to monitor various vital levels that encompass heart, sweat, pulse rates; oxygen blood levels; and urea and glucose levels. Implantable biosensors are implanted inside the body to continuously monitor the biomarkers or physiological parameters.^182–184 It provides a platform for mobile health devices and direct conversion of biological data into measurable signals.

Multiplexing detection indicates the capability of a biosensor to detect multiple bio-analytes, such as proteins and antibodies, simultaneously on a single chip or any other platform, as shown in Fig. 9.¹⁸⁵ It saves time and resources, requires a less sample volume and fewer materials, and reduces exposure to toxic drugs.¹⁸⁶ Various sensor designs using different platforms and multiplexing strategies have been described, such as Simoa and the proximity extension assay.^186,187 Difficulties in multiplexing detection involves signal interference from multiple analytes, expensive complex engineering, cross-reactivity due to non-specific binding, processing and analysis of large data, and calibration of individual analytes.¹⁸⁸ For effective multiplexing, future optical biosensor designs should address these challenges.

The other significant hurdle in next-generation optical biosensors is miniaturization that can be achieved using microfluidics with miniaturized channels, photonic integrated circuits, and nanomaterials. The use of fabrication techniques, such as inkjet printing, low-cost materials, and reusable bioreceptors, can address the cost-effectiveness of biosensor production. Currently available biosensors can detect a target on time scales of milliseconds to minutes. It is also a challenge to develop biosensor designs that can detect analytes on fast time scales and is, therefore, a central challenge to increase the applicability of single-molecule optical biosensors. Single-molecule detection is another important area to explore as it provides information about molecular properties distributions compared to existing biosensing techniques that focus on averages. The signal-to-noise ratio should be improved for better and reliable biosensing by reducing its potential sources that include electronic noise, thermal fluctuations, and environmental interference. Stability and reproducibility of sensors with passage of time and electronic noise require development of a new material that works under varying conditions, such as temperature, humidity, and chemical exposure. Commercial biomedical sensors and studies to detect the targets are mostly in buffer solution of filtering bio-solutions as compared to complex media such as blood, saliva, sweat, and urine. These complex fluids suffer from interference due to autofluorescence or scattering process and detection of non-target species that produce weak signals. Currently, researchers are concentrating on accuracy and signal enhancement combining plasmonic with fluorescence and nanoparticle labeling, as sensitivity and selectivity depend upon size, shape, orientation of the nanomaterial, and functionalization of chemical substances. Non-target species detection can be resolved using blocking agents or molecules with antifouling properties. Recently, new antifouling molecules, such as zwitterionic polymers and peptoids, have been reported to reduce non-target species detection. Self-powered biosensors coupled with mobile phones have huge potential to transform healthcare systems efficiently with a broad range of analytes due to seamless collection, analysis, and sharing of data.

In this Tutorial, the operating principle, design, optimization, classification of design, and applications of biosensors in various fields have been discussed in detail. Several detection mechanisms, including plasmonics, color change, fluorescence, SERS, and interferometry, have been demonstrated. Different parameters that indicate the performance of a sensor have been summarized. In recent years, researchers have been focusing on developing a continuous monitoring of vitals, glucose, and diabetes for early disease detection. Efforts are being made to implement biosensors as a wearable device for easy monitoring at a lower cost. Current research also includes the development of nano-biosensors, artificial intelligence, and improved synthesis techniques for nanomaterials and 2D-materials. Research to detect cancer and heart-related diseases, leading causes of death worldwide, involves the investigation of new biomarkers, cost-effectiveness, and commercialization of optical biosensors at clinical levels. It has been used to detect pathogenic bacteria at different sensing platforms also. The key features of lab-on-a-chip biosensors and wearable and implantable biosensors and future aspects and progress made in multiplex biosensors have been described.

The research was co-funded by financial support from the European Union under the REFRESH project—Research Excellence For REgion Sustainability and High-tech Industries Project No. CZ.10.03.01/00/22_003/0000048 via the Operational Programme Just Transition. This work was also supported by the Ministry of Education, Youth, and Sports of the Czech Republic conducted by the VSB-Technical University of Ostrava under Grant Nos. SP2024/081 and SP2024/059.

The authors have no conflicts to disclose.

Baljinder Kaur: Conceptualization (equal); Investigation (equal); Methodology (equal); Visualization (equal); Writing – original draft (lead); Writing – review & editing (equal). Santosh Kumar: Data curation (equal); Formal analysis (equal); Investigation (equal); Supervision (equal); Writing – original draft (equal); Writing – review & editing (equal). Jan Nedoma: Data curation (equal); Formal analysis (equal); Project administration (equal); Resources (equal); Writing – original draft (equal); Writing – review & editing (equal). Radek Martinek: Data curation (equal); Project administration (equal); Resources (equal); Validation (equal); Writing – review & editing (equal). Carlos Marques: Conceptualization (equal); Formal analysis (equal); Methodology (equal); Project administration (equal); Supervision (equal); Validation (equal); Writing – original draft (equal); Writing – review & editing (equal).

Data sharing is not applicable to this article as no new data were created or analyzed in this study.