| Version | Summary | Created by | Modification | Content Size | Created at | Operation |
|---|---|---|---|---|---|---|
| 1 | Bruno Miranda | + 2700 word(s) | 2700 | 2021-05-12 08:39:36 | | | |
| 2 | Bruno Miranda | + 1 word(s) | 2701 | 2021-05-18 13:22:03 | | | | |
| 3 | Peter Tang | Meta information modification | 2701 | 2021-05-19 04:51:04 | | |
Optical biosensors based on nanostructured materials have obtained increasing interest since they allow the screening of a wide variety of biomolecules with high specificity, low limits of detection, and great sensitivity. Among them, flexible optical platforms have the advantage of adapting to non-planar surfaces, suitable for in vivo and real-time monitoring of diseases and assessment of food safety.
Optical biosensors have emerged as analytical devices for the rapid [1], cost-effective [2], selective [3], and specific detection of biological compounds (antibodies, nucleic acids, peptides, toxins, etc.), as well as bacteria [4], viruses [5][6], and cells [7]. The specificity of biosensors is an intrinsic property arising from the biorecognition probe immobilized on the surface of the transducing element. To this aim, noble metals nanomaterials represent very efficient transducers, due to their capability of supporting localized surface plasmons (LSPs) [8] and of significantly enhancing Raman scattering of molecules adsorbed onto their surface (SERS) [9].
Localized surface plasmon resonance (LSPR) is the size and shape-dependent coherent oscillation of the conduction electrons of a noble metal, arising when the size of the object is much smaller than the excitation wavelength [10][11][12][13][14]. The excitation of LSPs gives rise to a strong enhancement of the electromagnetic field in the surroundings of the nanoparticles, which makes their resonance locally sensitive to refractive index variations [15]. In particular, silver (Ag) and gold (Au) nanoparticles (NPs) have been studied deeply due to their capability of exhibiting LSPs in the visible region of the spectrum, thus allowing the design of refractive index [16][17] and colorimetric [18][19][20] optical biosensors. When a target analyte is recognized by the nanoparticles, a resonance shift, proportional to the concentration of the analyte, can be measured through UV-vis spectroscopy.
Noble metal nanoparticles immobilized onto a substrate can be used also for surface-enhanced Raman spectroscopy (SERS). SERS is a sensitive and powerful optical technique providing resolutions up to single-molecule detection [21][22]. It has been extensively used for label-free biochemical assays and cell studies [23][24][25]. Two main mechanisms are involved in SERS: the charge transfer between the molecules and the substrate (chemical effect), and the LSPR modes of noble-metal nanoparticles (electromagnetic effect) [26][27]. SERS spectroscopy is performed to collect information about molecular vibrational states, guaranteeing high sensitivity to conformational changes [9]. Metallic nanoparticles provide the selected substrates with strong enhancement factors (EF) of the molecular Raman signals [28]. For SERS spectroscopy, strong efforts have been made to design and fabricate efficient substrates, with enhancement factors of the Raman signals up to 1014, to reach ultra-low limits of detection [29].
All the advantages shown by optical devices based on plasmonic nanoparticles have stimulated the continuous improvement of their fabrication techniques. The nanotechnological fabrication processes are based on two main approaches: top-down and bottom-up, which are sometimes combined to obtain a “hybrid approach” [30][31]. While the top-down approach usually requires nanolithographic techniques, which permits the mechanical or chemical etching of the bulk material, the bottom-up approach is based on the chemical synthesis of nanoparticles [32][33][34], starting from “molecular bricks.” In the case of the bottom-up approach some other methods are required to graft the nanomaterials onto the substrates, usually made of rigid materials (glass, silicon, quartz, etc.) [35][36][37].
The concept of flexible optical biosensors has been introduced more recently, due to the necessity of creating some optical platforms capable of adapting to non-planar surfaces, boosted by the advent of flexible electronics [38] and photonics [39]. This property finds its natural application in wearable sensors, conforming to the skin [40][41][42], food-packaging (sensors for food monitoring) [1][43], real-time monitoring of healing processes [44], and 3D cell cultures in scaffolds and organoids (cellular growth rate monitoring) [45]. Other advantages of flexible plasmonic substrates rely on the rapid, real-time, and cost-effective monitoring of a target analyte.
Flexibility allows rapid and high processability, thus extending plasmonic platforms to daily life applications [46]. For these reasons, many researchers have introduced very promising hybrid/nanocomposite transducers, based on the combination of synthetic or natural polymers with metallic nanoparticles. The combination of polymers with optically active nanomaterials generates platforms with extreme ease of integration within microfluidics and microelectronics devices, showing promising developments toward smart and efficient technologies.
Flexible biosensors find unprecedented applications in the design of wearable, point-of-care testing, and food monitoring devices. First, rigid substrates commonly employed for the accommodation of the plasmonic nanoparticles are difficult to employ as wearable sensors since they cannot easily adapt to skin. Also, rigid platforms on the skin could be uncomfortable and they could not find patient’s compliance [47]. Secondly, concerning POCT devices, researchers are moving toward the use of microfluidics to reduce sample volumes and enhance the capability of an analyte to interact with the bioprobe on the sensing surface. In this context, polydimethylsiloxane (PDMS) and poly(methyl methacrylate) (PMMA) represent the gold standards to fabricate microfluidic channels [48][49]. The typical approach to combine microfluidics and rigid plasmonic substrates is the bonding of the two components. The final result is a microfluidic channel having only one wall covered with the transducing element. However, at the microscale, it may be worth having a channel completely covered with plasmonic nanoparticles to increase the detection efficiency and the contact area. While this is not possible with rigid substrates, it can be done with polymeric nanocomposites [50][51]. Finally, in food safety monitoring polymeric optical devices show appealing features to be easily integrated into food packaging, which is mainly involving polymeric materials [52][53].
The elasticity, bending capability, and stretchability of polymers over/in which plasmonic nanoparticles can be impregnated has been opening novel fundamental studies on the coupling mechanisms between plasmonic nanoparticles. This is something that was not feasible with rigid platforms. As an example, the optical response of plasmonic NPs dispersed in a polymeric film can be coupled by compression, due to the reduction of the distance among NPs, or can be decoupled by stretching the polymer [54]. For this reason, flexible nanoplasmonic is rapidly evolving in optomechanics, which combines theoretical physics with optics and material sciences [55][56]. Moreover, these platforms find applications in many other research fields, such as homeland security (i.e., drugs [57] and explosives [58][59] detection), seismology [60], and plant biology [61].
Figure 1 briefly schematizes the most used approaches to obtain functional biohybrid nanocomposites, together with the setups usually employed for their optical characterization and the main advantages. The most used nanomaterials are spherical gold and silver nanoparticles, but more complex shapes, such as nanorods and nanostars [29][62][63], are also employed for the fabrication of these optical devices. The shape and the size of the NPs are important design parameters to tune and optimize the optical responses. The biorecognition elements may include antibodies, enzymes, single-stranded DNAs, and/or aptamers, which provide the platform with high selectivity and specificity for the target analytes (antigens, substrates, RNAs, and cells). The LSPR optical setup usually consists of a white light source directly connected to an optical fiber probe. The resonant spectra can be collected in transmission mode if the device is optically transparent, or in reflectance mode, for devices with high reflectivity. A spectrometer is used to collect the transmitted/reflected light. Vice versa, a typical SERS setup consists of a laser source at different wavelengths, whose light is directly conveyed to the devices and collected with a CCD camera to register the Raman signal. In this review, we mainly focus on the description of LSPR-based flexible biosensors and SERS-based flexible biosensors, reporting the most innovative technologies and protocols for the fabrication of bio-responsive materials combining synthetic or natural polymers with gold or silver nanoparticles having diverse shapes and sizes.

Figure 1. A schematization of flexible optical biosensing platforms reporting the combination of a polymeric substrate with differently shaped gold/silver nanoparticles, the most used biorecognition elements, and target analytes. Furthermore, a schematization of the detection setups for LSPR and SERS signals is reported. Finally, the main advantages are summarized.
The demand for optical biosensors based on LSPR rather than SPR has increased conspicuously in the last two decades. This is mainly due to the different spatial decay of the two sensing platforms. While surface plasmon polaritons (SPPs) exploited for SPR are generated on a thin metallic surface (thickness ~10–250 nm) and have a large spatial decay, localized surface plasmons (LSPs), also known as non-propagating plasmons, are generated on noble-metal nanoparticles, which have characteristic dimensions always well below the excitation wavelength. In the second case, the spatial decay of the electromagnetic field is much smaller and limited to the surrounding of the NPs [64][65][66][67]. This significant difference allows the design of platforms, whose sensitivity is strictly associated with the surface of the NPs and independent from what happens far away from the surface (bulk) [32][33]. In this context, LSPR biosensors show appealing properties such as miniaturization, minimal interferences, and scalable production. However, while both periodic [34][35] and non-periodic [31] arrays of noble-metal nanoparticles on hard substrates have been already proposed as sensing platforms, there is still a lot of active research to propose novel approaches toward the fabrication of flexible, polymer-based LSPR biosensors. The main issue associated with the fabrication of such optical platforms is the limited number of polymers that can be used as substrates. A good LSPR biosensor must be highly transparent or reflective to allow the detection of the optical signal from noble metal nanoparticles. For this reason, opaque polymers, such as nanofibers (commonly employed as substrates for SERS-based biosensors) are not suitable for LSPR sensing.
SERS optical biosensors leverage on the design and fabrication of periodic, quasi-periodic, or random metal nanostructured arrays (nanohole arrays [68][69][70], nanocanals [71], porous structures [72][73][74][75][76][77], etc.) on rigid substrates (alumina, silicon, glass, etc.), showing very efficient performances, in terms of sensitivity and limits of detection. However, for the SERS measurements, the adsorption of the analyte of interest onto the surface is a necessary step, which is not always straightforward. It requires the extraction and the collection of the biomolecule and the selection of suitable surface chemistry for the successful binding of the analyte onto the substrate [78]. To obtain efficient and fast in situ detection, a rigid and opaque substrate may limit the applications to planar surfaces.
For this reason, the development of flexible, transparent substrates is very promising to overcome these issues, allowing the non-destructive detection of the target analytes. Among the currently used materials, we can distinguish between synthetic and natural polymers as SERS flexible substrates, whose performance is comparable to the previously mentioned rigid platforms [79][80].
There is an increasing demand for portable biosensors, where the clinical diagnostics is directly transferred from equipped laboratories to the patient on site-diagnosis. This need asks for renovated fabrication strategies of point-of-care testing (POCT) devices, which show ease-of-use, compact size, and limited costs [81][82]. Many examples of already commercialized POCT have been reviewed recently and include pregnancy tests, glucose testing, and HIV testing [83]. LSPR- and SERS-based flexible biosensors are promising transducers for the design of a POCT due to the ease of integration with microelectronics and microfluidics [84] (Figure 2a).

A first example of the integration of an LSPR platform with microfluidics has been reported by Huang et al. [87]. They introduced an approach to continuously monitoring the light transmission from an array of AuNPs arranged in a microfluidic channel. A green LED was used in substitution the typical halogen light source. The authors reported a sensitivity of 10−4 in RIU. The sensing capabilities of the proposed biosensor were shown by measuring the absorbance variation arisen from biotin/anti-biotin interaction. A LOD of 270 ng/mL was successfully achieved. This first example highlights the importance of the design process of both microfluidics and miniaturized optical components. More precisely, microfluidic channels must be highly transparent, to be compatible with light pathways, they should ensure an efficient sample delivery and minimize reagents and sample consumption [82]. On the other side, spectrometers and light sources (optical components) must be miniaturized to obtain a compact device and, although this is often not very easy, some methods to integrate LEDs for the transducer illumination and miniaturized spectrometers for the collection of the signal have been already proposed to overcome this issue [83]. POCTs for the diagnosis of disease especially in developing countries, where expensive laboratory equipment and specialized operators are not easily available are crucial for the rapid screening of a population. In this scenario, the low-cost polymers-based plasmonic devices offer the possibility to extend the modern lab technologies all over the world and give the less well-off the possibility to access to fast diagnosis and appropriate health care [88].
The design of wearable biosensors for the early diagnosis of diseases has seen many efforts in sensors research. Unfortunately, these novel platforms generally suffer from low reproducibility in sensing capabilities as well as a lack of accuracy in the robust quantification of biomarkers from the skin due to the very tiny concentrations and species of targetable analytes in sweat. Moreover, some crucial issues are still topics of active research: data acquisition, processing, power supply, adaptability to non-planar surfaces (e.g., skin) [42]. Of course, sensitivity, selectivity, and low limits of detection are crucial in any sensing platform, but, in the case of wearable sensors, the collection of skin fluids from the body in a non-invasive way is still an open challenge. Some attempts involving textiles and hydrogels for their absorbing capability have been proposed. However, these materials are not suitable for the precise control of the collected volume.
The combination of micro-and nano-technology for flexible plasmonic biosensors has given rise to platforms with integrated functions all focusing on a single device having a few millimeters size. In these cases, microfluidics and microelectronics can be combined with flexible plasmonic platforms to produce wearable optical biosensors, whose readout can be performed to the naked eye or via integration with smartphones [41]. Wearable optical biosensors find their potential applications in the fast screening of the population for the detection of a target pathogen, which, in the era of SARS-CoV-2 (severe acute respiratory syndrome coronavirus 2), have revealed as crucial to avoid the pandemic spreading of disease. LSPR and SERS-based optical biosensors have already shown their potential in the detection of viral pathogens as SARS-Cov-2 [89][90][91]; for this reason, combining this to wearable biosensors by embedding the transducing elements on a flexible substrate could be a winning strategy to pursue, as reported also by Choe et al. [85] (Figure 2b).
Due to the overall increase of the world population in the past decades, avoiding food waste is becoming a fundamental necessity; for this reason, the growing food industry is working on the improvement of the long-storage and preservation of food with novel packaging and delivery systems [53]. In this scenario, the biosensing of freshness markers, pathogens, allergens, and toxic agents in food is evolving toward the so-called smart active packaging [52]. Many sensors have been already proposed for food monitoring, but, again, some of the commonly encountered issues hide in the robustness, selectivity, and sensitivity of the proposed devices. Smart colorimetric labels could provide a “quality index” of the food by simply exhibiting a color variation visible to the naked eye [92]. Even though the implementation of some devices in smart and active packaging has already been proposed, for instance in refs. [53][93], one of the main challenges remains the achievement of a multiplexed sensing of the many different factors affecting the quality of certain food. Flexible optical biosensors have appealing multifunctional capabilities enabling both contaminants detection and longer shelf-life of food due to the sensing mechanisms, herein reported, and to the antimicrobial activity of noble-metal NPs [44][94]. For this reason, the use of polymers combined with optically active nanomaterials exhibit promising potential also in food quality monitoring. A smart application for the SERS-based detection of pesticides in fruits and vegetables has been reported in ref. [86] (Figure 2c), but many other flexible platforms are currently ready for these applications.