Multi-functionalization Strategies Using Nanomaterials: A Review and Case Study in Sensing Applications

Remarkable advances in nanomaterials and nanotechnology have led researchers in various fields. The scale effects imparted by nanomaterials are associated with unexpected macroscale phenomena and properties that find many applications. However, multi-functionalization may be accompanied by physical and commercial limitations. Therefore, research must proceed in several different directions. Here, we define multi-functionalization and the electrical applications thereof in terms of increasing performance, addition of new and valuable properties, and multi-physics in play. We deal with sensors, actuators, energy harvesters, and solar cells and explore research that seeks to increase sensitivity, append “stretchability”, and facilitate untethered communication. Furthermore, we analyze research trends in materials use and manufacturing, and highlight useful fabrication methods. With the aim of predicting future research trends, our review presents a roadmap that will aid research on sensing and multi-functional applications.


Introduction
Applications of nanotechnology have received a great deal of attention since the late 1980s. In the United States, for example, federal government funding in nanotechnology increased by 15% annually from 2001 to 2010, and stands at over $1.6 billion today [1]. Large companies such as GE and Intel devote significant resources to nanotechnology [2,3]. Carbon nanotubes (CNTs) and graphene lattices created from nanoconfined forms of graphite have been used to create high-performance transistors and ultra-strong composite materials [4,5]. These materials have been employed to manufacture new displays, super-efficient batteries, and solar cells [6][7][8]. Pharmaceutical companies seek to use nanoparticles to target drugs for diseases such as cancer [9,10]. In the fields of electronics and semiconductor manufacturing, graphene-based electronics, spintronics, and photonics may replace silicon [11][12][13]. Nano-structuring increases the battery storage capacity, yields cheap and efficient solar cells, and delivers ultra-strong composites [13][14][15]. Nanomaterials are used to produce smart materials that are self-healing or self-cleaning, memory metals that revert to their original shapes, and piezoelectric ceramics and crystals that turn pressure into energy [16][17][18][19].
Here, we review nanomaterials ranging from several tens of nanometers to several micrometers in size (Fig. 1); thus, nanoplates, nanoparticles, nanowires, and nanotubes find most applications [20]. Their behavior is governed by both quantum and classical physics [21]. For example, nanomaterials have much greater surface areas than macro-scale materials of similar weights. Furthermore, nanomaterials interact very extensively with their surroundings [22].
Ordinary substances have different properties on the nanoscale; thus, greater reactivity, unusual electrical behavior, and remarkable strength per weight allows the creation of new medicines, super-slick coatings, and stronger composites [23][24][25]. Figure 2 shows that nanomaterials exhibit 1 3 (1) support and confinement effects, (2) shape and size effects, and (3) composition and oxidation state effects. The first effect occurs when the nanomaterial scale is comparable to the wavelength of an electron [26]. The motion of randomly moving electrons is restricted to specific energy levels, and the quantum realm reigns [27]. Such effects depend on the material properties, Bohr radius, available energy levels, potential well, and valence and conduction bands [28].
Nanomaterial shape and size (the "morphology effect") influence the melting point, fluorescence, electrical conductivity, magnetic permeability, and chemical reactivity of the materials [29][30][31][32][33]. Vacancies ("missing" atoms in a lattice) play major roles in electrical and thermal transport and mechanical behavior [34,35]. Furthermore, nanomaterials may exhibit a variety of oxidation states and are easily oxidized by ambient air (which is sometimes undesirable) [36]. Usually, the composition and oxidation status of magnetic nanomaterials are determined by the chosen synthetic phase, which may include both surfactants and precursor ligands [37,38].
The explosive interest in nanomaterials is reflected in the number of papers listed in the Science Citation Index (Extended) (SCI(E)). We analyzed research trends of 2011-2013, 2014-2016, and 2017-2019 based on the number of published articles and citations, impact factors, and the H-index [39,40]. This index allows the analysis of journal quality and principal research trends. The H-index evaluates the performance of individuals and the impact of their work. Figure 3 shows that research interest and quality increased greatly from 2011 to 2019 [41]. All nanotube, nanoparticle, quantum dot, nanowire, and nanoplate studies have increased greatly in number over the past five years. Carbon nanotubes (CNTs) are the most active topic in terms of the number of journal articles and the H-index, followed by quantum dots and nanoparticles. Nanoplate research exhibited the most rapid increase in H-index, and nanoparticles showed the slowest. Interest in nanomaterials shows no signs of abating.
Nanomaterials are used in a variety of products including paints, filters, insulators, and lubricants; their very small scales impart specific properties, and they have become invaluable tools in materials science [42][43][44][45]. Recently, nanomaterials have been applied in biosensing, bioimaging, tumor diagnosis, and anti-biofouling [46,47]. Pharmaceutical companies use nano-enabled biotechnologies to produce rapid and sensitive diagnostic agents and more effective therapeutics [48]. The food manufacturing industry employs nanolaminates of edible lipids or polysaccharides to spray food products, protecting them from air and moisture [49]. Figure 4 shows the scope of nanomaterial applications, including sensors, actuators, energy harvesters and storage devices, solar cells, catalysts, healthcare materials, compounds used in food and agriculture, cosmetics and paint, and industries. In this study, we focus on their electrical applications. We will deal with sensors, actuators, energy harvesters, storage devices, and solar cells. Engineers seek to enhance performance (increase sensitivity), add new properties (for example, stretchability), ensure untethered communication, and develop a multi-physics that qualitatively  Fig. 1 The scales of nanomaterials with respect to nano objects and nanostructures Fig. 2 The physical attributes of various nanomaterials with respect to composition and oxidation effect, shape and size effect, and support and confinement effect guides trends in functionalization. Such research must consider the physical and commercial limitations of nanomaterials used to create real-world physical devices. To this end, we presented a case study.

Towards Multi-functionalization
Nanomaterials enable their application to increase and maximize the value proposition. Figure 5 shows that nanomaterials add value by delivering unprecedented performance and versatility to end users. It is possible to tailor material structures at extremely small scales to achieve specific properties, thus greatly extending the scope of material science. A multi-functionalization, which is defined as allowing to provide more than one attribute at a time, has been generated in the process of maximizing value. As new attributes have been added that have not been possible in a single product before, end-users could have benefited from using the product. It can be explained by a product bundling strategy that offers several products or services for sale as one combined product or service package in marketing [50,51]. A good bundle is worth more than the sum of its parts to the end user, replaces confusing or tedious choices with order and simplicity (Fig. 5). The extra properties afforded by nanomaterials generally reduce costs; even the convenience of not hand-picking and adding-on items is attractive. Since bundling of two different properties by nanomaterials can be effective in maximizing application value, researchers have been working on appending the properties of an existing product in various ways using nanomaterials.

Physical Limitations of Nanomaterials
The physical limitations of nanomaterials are usually evident during synthesis, which may be either top-down or bottom-up. During top-down synthesis, nanostructures are created by etching away existing crystal planes or by miniaturizing bulk fabrication [52]. Top-down processes dominate microscale fabrication but are limited by the wavelength of light and the available tools [53]. During bottom-up synthesis, nanostructures are synthesized on a substrate by atom stacking, which creates crystal planes [54]. The structures are smaller than those of top-down syntheses, but the surfaces and stacked molecules must be compatible [55]. Neither process yields completely homogeneous nanomaterials of the desired sizes and shapes (Fig. 6). Free or surface energy imposes physical limitations during manufacturing. The surface (or Gibb's) energy is the difference in energy between a packed material and the same number of atoms in an infinitely extended solid [56,57]. As nanomaterials exhibit large surfaceto-volume ratios (the "size effect"), their reactivities are very high [58]; thus, they exhibit large surface energies associated with a tendency to self-assemble and melt at low temperatures.

Commercialization Limitations of Nanomaterials
Nanomaterial commercialization is compromised by high processing costs, a lack of investment capital, poor process scalability, and a shortage of qualified engineers [59]. The investment timeline commences with a breakthrough invention, followed by attempted commercialization, and then (in successful cases) product acceptance and profit ( Fig. 7; [60]). Researchers must bear in mind that expenditure is the greatest in the developmental phase and may be difficult to maintain.

Multi-functionalize Strategy
The term "multi-functionalization" is commonly used to explain new research trends, but confusion reigns: Does the term differ from "multi-physics" and "self-adjustability"? We define multi-functionalization as increasing performance, adding properties, and deriving multi-physics (Fig. 8). We sought trends in multi-functionalization; we do not focus on the historical phases of technological innovation, but are concerned with real-world applications. We seek to help researchers place their work in the context and suggest how they might best proceed. A performance increase reflects the value of a researcher's work and is appreciated by the end user. For example, enhanced sensor sensitivity and utility, a reduction in the actuating force, and an increase in the actuating frequency band are valued by end users. The novel properties of nanomaterials (Fig. 2) facilitate the addition of further properties, allowing (for example) construction of lightweight body armor, or clothes that resist wrinkling, staining, and bacterial growth. Clear nanoscale films on eyeglasses, computer and camera displays, windows, and other surfaces render them water-and residue-repellent, antireflective, self-cleaning, resistant to ultraviolet or infrared light, nonfogging, antimicrobially active, scratch-resistant, and electrically conductive. Nanomaterials are beginning to enable washable, durable "smart fabrics" equipped with flexible nanoscale sensors and electronics that monitor health and capture solar and motion energy. Furthermore, lightweighting of cars, trucks, airplanes, boats, and spacecraft saves significant fuel. Polymer

Value increase
Value maximise Fig. 6 The physical limitations of nanomaterials, from synthesis to inclusion in a device composite nanomaterials are used to make golf clubs, tennis rackets, bicycles, motorcycle helmets, automobile parts, luggage, and power tool housings, all of which are lightweight, stiff, durable, and resilient. In addition, CNT sheets will soon be used in next-generation air vehicles. Innovations may be incremental or radical. Performance increases and property additions are usually incremental, and new physics is radical. Such physics reveals unsuspected applications and opens new markets. The research and development costs are high, as are the expected returns. The process is risky. The physics includes connections between mechanics, electromagnetics, and optics. Of the three elements of multi-functionalization, the first is the improvement of properties, namely, mechanical, electrical, magnetic, optical, thermal, or chemical properties (Fig. 9). For example, the stiffness and strength are representative mechanical properties. The electrical properties include conductivity and resistivity. The second element is the addition of new properties, achieved using various nanomaterials and novel design strategies. The third element is new physics. Below, we will deal with sensors, actuators, energy harvesters, and solar cells. Thus, the original properties are electrical; as will become clear, most appended properties are mechanical. Some yield new magnetic, optical, thermal, or chemical physics.

Sensors
Nanotechnology has enabled electronic devices to become smaller and have a higher performance than before. Nanometer-scale materials and manufacturing processes quickly and effectively save and transmit data. Flexible, bendable, foldable, rollable, and stretchable electronics are being integrated into wearable medical applications, aerospace instruments, and the Internet of Things devices. Flexible electronics include semiconductor nanomembranes for smartphones and e-readers. Graphene and cellulosic nanomaterials are being used to create wearable tattoos, photovoltaics sewn into clothing, and rollable electronic paper. Flat, flexible, lightweight, non-brittle, highly efficient electronics find applications in a host of smart products. Other computing and electronic products include flash memory chips for smartphones and thumb drives, ultra-responsive hearing aids, antimicrobial/antibacterial keyboard coatings and cell phone casings, conductive inks for printed electronics [controlled via radio-frequency identification (RFID)], smart cards, smart packaging, and flexible e-book readers. Table 1 and Fig. 10 show the research trends and H-index trajectories of the sensing applications. Sensor sensitivity is commonly used to evaluate sensor quality. The commonly associated keywords are flexibility, stretchability, waterproofness, transparency, and self-healing. Various nanomaterials and designs have been used to increase sensor sensitivity. Sensors are commonly fabricated from silver and gold nanoparticles, graphene, and CBTs. Liu et al. placed a woven fabric in a graphene sensor and achieved a gauge factor (GF) as high as 223 [94]. Yan et al. electrospun a nanofiber yarn made of carbon black and graphene nanocomposites [95], achieving a GF of 1700. Chen et al. reported a GF of 4000 using a silver nanowire and graphene [96]; the sensor exploited mechanical cracking. Sensors are now additionally    Table 1). The most popular appended properties are flexibility and stretchability, and the sensors can be placed on free-formed or deformable surfaces. Flexibility and stretchability also increase the sensing range. Flexibility is achieved by the use of nanoparticles [61,62,64,65] or graphene [63]. Su et al. printed silver nanoparticles and fabricated a liquid channel (to introduce flexibility), achieving a GF of 3000 [64]. Stretchability was achieved by printing nanoparticles [74] or graphene [70] onto stretchable substrates. In many cases, nanomaterials are mixed or specifically designed to maximize GF and extend stretchability. Zhang et al. used nanocomposites of silver nanoparticles and CNTs to attain a stretchability of 95% and a GF of 39.8 [66]. Jeong et al. developed fragmented, structured strain sensors made from graphene foam with a stretchability of 70% [73]. Guo et al. created a sandwich strain sensor fabricated from CNTs and a carbon black nanocomposite, achieving a stretchability of 120% [72]. Park et al. used a graphene yarn to develop a sensor with a stretchability of 150% and a sensitivity of 10 [67]. Zhou et al. increased the sensitivity of a 50% stretchable sensor fabricated using CNTs to 100,000, exploiting mechanical cracking [68]. Zhao et al. extended the stretchability to 120%, with a GF of 10,000,000, by combining graphene and silver nanoparticles in a fragmented structure [75].
The sensor performance typically varies with pressure, temperature, and humidity. Temperature, in particular, exerts major effects; nanomaterials are temperature-sensitive because of their size (Fig. 2). The sensor performance must be independent of the temperature unless the sensor is to measure the temperature. Yi et al. fabricated a temperaturetunable sensor using a thermally evaporated gold nanofilm [82]. Rahman et al. created a three-dimensional (3D), printed, silver, nanoparticle sensor stable at 500 °C [81].  [92]. Furthermore, wireless communication is also possible; electromagnetic methods use radiofrequency (RF) signals [83,84] or structural color [85][86][87][88], as will be discussed in Sect. 4.3.
The self-healing property is the highest increasing property in terms of sensor research trends using nanomaterials. Self-healing materials take inspiration from biological systems that can self-organize and self-repair. Self-healing materials would reduce the need for costly maintenance by healing themselves when damage occurs. One example includes ingredients that are automatically released or expand to fill cracks when they appear. Because of the increasing cost of the sensors and their properties, which are difficult to maintain in case of failure due to miniaturization, it is expected that research on self-healing properties will continue.

Actuators
An actuator converts stimuli such as light, heat, and electrical and magnetic fields into a mechanical output. Electrostatic, piezoelectric, shape memory alloy (SMA), thermal, electrochemical [107,108], and photo actuators have been described. We focus on electrical actuators, where an electric signal triggers a mechanical output. Such actuators have various applications (for example, artificial muscles). The outputs differ in terms of displacement, force, and frequency. Most actuators are flexible, allowing deformation or deflection; however, we do not consider flexibility to be an appended property. Nanomaterials including nanoparticles [109][110][111], carbon black [112], CNTs [113,114], and graphene [115][116][117][118][119] have been used to fabricate actuators. Lu et al. used CNTs to develop a micro-optomechanical actuator via photolithography, achieving a maximum deflection of 22 mm [114]. Sul et al. fabricated a bimorphic actuator using multi-walled CNTs (MWCNTs) and aluminum [120]. Liang et al. prepared an electromechanical actuator by stacking graphene sheets fabricated via filtration. Fe 3 O 4 nanoparticles were placed between the sheets to enhance their performance [117]. Park et al. developed an actuator triggered by humidity that featured a bilayer of graphene oxide and MWCNTs [118]. Li et al. created an actuator exhibiting 360º deflection using a reduced graphene oxide (rGO)/polydimethylsiloxane (PDMS) bilayer and direct ink printing [119]. Kong et al. developed an ionic polymer metal composite (IPMC) actuator employing CNTs and graphene [121].
The properties of the actuators are listed in Table 2, and Fig. 11 shows the research trends revealed by the H-index of journal keywords. The most popular keyword in terms of an appended property is "soft". Polymers are mixed with nanomaterials. The actuation mechanism differs based on the material used. Photoactuators (activated by light) were created by mixing photoresponsive nanoparticles with poly(N-isopropylacrylamide) (PNIPAAm) (a thermally sensitive hydrogel) [97,98,103,122]. Sun   N, N-dimethylacrylamide, employing various geometries (swollen, twisted, spiral, or helicoidal) to study the actuation characteristics [99]. The other property attributed to actuators is transparency (Table 2). Zhu et al. used graphene and epoxy to create a bimorphic structure via chemical vapor deposition, photolithography, and reactive ion etching. The transparent actuator mimics the shape and motion of the fly wing. The maximum deflection range was 4.5 μm [104].
Multi-functionalization has also been realized by rendering actuators multi-responsive. Han et al. developed an actuator that reacts to humidity, light, and electrical currents. MWCNTs, carbon nanofibers, and poly(3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOTPSS) were printed via electrospinning [105]. Shepherd et al. placed molecular spin-crossover materials in a polymer matrix when preparing a multi-responsive actuator in response to both temperature and light [106].
Among the important properties of actuators, force and frequency are greatly affected by the weight of the actuator itself, and therefore, efforts to reduce the weight itself are expected to continue. Manufacturing composite materials composed of nanomaterials is a widely known technique for achieving lightweight materials. In addition, considering the industry to which actuators are applied, advanced composites can help build strong, lighter components for vehicles, including aircrafts. In addition to next-generation nanocomposites, ongoing advances in composites made from carbon fiber and other materials could make it possible to substitute composites for materials such as aluminum in increasingly more applications. These advances include new methods for producing and binding carbon fibers, allowing for less expensive fabrication.
In addition, memory metals that revert to a prior shape when heated to a specific temperature could be a direct method for self-adaptation to the surrounding environment. These materials are being considered as a way of producing movement in an actuator using a charge to expand or contract the materials, imitating muscle movement. Some versions of memory metals can be programmed to take on multiple shapes at different temperatures.

Energy Harvesters
Nanomaterials may reduce our reliance on traditional energy sources and help meet increasing energy demands by allowing us to use clean, affordable, and renewable energy sources and reduce energy consumption. Thin-film solar panels attached to computer cases and flexible piezoelectric nanowires woven into clothing generate usable energy from light, friction, and body heat, thus powering mobile electronic devices. Similarly, various nanoscience-based approaches are aimed at converting the waste heat of computers, automobiles, homes, and power plants into usable electric power. Energy-efficient and energy-saving products are increasing in both number and type. Nanotechnology may yield affordable, clean drinking water via rapid low-cost detection and treatment of impurities. The trend shows that efforts are underway to append new properties to energy-harvesting devices. Table 3 and Fig. 12 list and show the trends in properties attributed to piezoelectric and triboelectric energy-harvesting systems. Flexibility is often imparted for piezoelectric applications. A piezoelectric device converts mechanical energy into electric energy, and vice versa. Flexibility thus expands the device utility. Many researchers have developed flexible piezoelectric energy harvesters using nanomaterials such as ZnO, lead titanate (PbTiO 3 ), lead zirconate titanate (PZT), and polyvinylidene fluoride (PVDF) as nanoparticles [128,130,132,138], nanofibers [125,129,131], nanotubes [123], or nanosheets [124,126]. Liu et al. developed a flexible piezoelectric energy harvester by electrospinning PVDF nanofibers and lead oxide nanoparticles [125]. Shin et al. screen-printed barium zirconium titanate (BZT) nanoparticles onto polyvinylidene fluoride (PVDF) nanofilms to ensure flexibility. The energy density was 2.79 mJ/cm 3 [126]. Shi et al. created a flexible piezoelectric energy harvester with an open-circuit voltage of 11 V from graphene nanosheets and barium titanate (BaTiO 3 ) nanoparticles [124]. Flexible harvesters have also been electrospun from nanocomposites of titanate nanoparticles/PVDF nanofibers [129] and nickel oxide (NiO) nanoparticles/silicon oxide (SiO 2 ) nanoparticles [130]. Jeong et al. reported a stretchability of 200% using lead magnesium niobate-lead titanate (PMN-PT) nanoparticles, silver nanowires, and CNTs [123]. Nanomaterials have also been used to generate triboelectric energy generators [138,140]. Chen et al. employed strontium titanate (SrTiO 3 ) nanoparticles with nanopores to fabricate a flexible triboelectric energy device. The opencircuit voltage and power density were 338 V and 6.47 W/ m 2 , respectively [138].
One of the main objectives of energy harvesters is to harness energy from its surrounding environment and store this harnessed energy for later use. We expected that research on energy harvesters would be more than contributing to the environment by simply saving and generating energy. Because piezoelectric materials that turn pressure into electricity are not new and have reached technical limitations in terms of efficiency, efforts to change the materials and structures themselves into eco-friendly materials are likely to be attempted. Green materials that attempt to solve environmental issues in a more direct manner are estimated to account for low carbon-dioxide emissions. Furthermore, the lead-free property could be an existing example of researchers' efforts toward going green.

Solar Cell
Energy storage technology includes batteries and other systems that store energy for later use. Lithium-ion (Li-ion) batteries and fuel cells are already powering electric and hybrid vehicles and billions of portable consumer electronic devices. Li-ion batteries, in particular, have improved greatly in terms of performance and fall in price; the cost per unit of storage capacity has declined considerably over the past decade. Over the next 10 years, advances in energy storage technology may render electric vehicles (hybrids, plug-in hybrids, and all-electronic vehicles) cost-competitive with vehicles using internal combustion energy. In power grids, advanced battery storage systems integrate solar and wind power, improve power quality by controlling frequency variations and handling peak loads, and reduce costs by enabling utilities to postpone infrastructure expansion. In developing economies, battery/solar systems may deliver reliable power to places that have never had power. Here, we focus on solar cells as an example of multi-functionalization. The original property is energy density. The materials used for construction include silver [144], titanate [145][146][147], gold [148], and ZnO nanoparticles [149], among others [150][151][152][153]. A photovoltaic device with a magnesium oxide nanolayer afforded efficient (18.23%) power conversion [154].
Solar cells have become flexible [141] and transparent [142] (Table 4). Figure 13 shows the related research trends. Zi et al. developed a flexible photovoltaic device with a photocurrent density of 3.0 mA/cm 2 using silver nanoparticles in a hydrogenated, amorphous silicon/germanium base [141]. Bush et al. fabricated semi-transparent photovoltaic devices using aluminum and zinc oxide nanoparticles [142]. Solar cells exhibiting multi-physics have been studied from the viewpoint of multi-functionalization. Dou et al. used bismuth tellurium nanotube/ZnO nanoparticle composites to convert solar energy into both thermal and photo energy, thus enhancing the solar cell efficiency [143].
Recently, metamaterials that revert their original properties, including permittivity, permeability, emissivity, and thermal/ electrical conductivity by light, have been introduced. They have the advantage that they can be applied to applications that include a solar cell in a compact manner because it may not require additional equipment or physical sources that are required to change various intrinsic properties.

Case Study: Sensing Application
Sensing applications, usually termed sensors, are actively researched by those interested in advanced nanomaterials. Since 1950, efforts have been made to explain the physical behaviors of nanomaterials used in sensing applications. Usually, simplified models based on quantum mechanics, or molecular dynamic or simulation methods, have been employed to model crowded nanomaterial environments. Recently, such approaches have provided valuable guidance to those interested in the synthesis and exploitation of novel nanomaterials. State-of-the-art sensors use nanoparticles, nanotubes, and graphene to exploit the physical phenomena associated with the scale effects of nanomaterials (absent from bulk materials), to increase sensor performance, and to expand their potential applications. The scale of the nanomaterials can be found in Fig. 1 in the introduction.
In the present era of big data, sensors are finding applications that were unimaginable earlier. Wearable sensors measure human body motion, control human-machine interfaces during manufacturing, and monitor manufacturing robots. Sensors must be increasingly flexible, stretchable, and communicative (Fig. 14), and it is expected that future applications may be even more demanding. Here, we deal with research trends through the lens of our multi-functionalizing strategy. Sensor performance indicators include sensitivity, stretchability, measurement range, mechanical durability, signal linearity, and hysteresis; however, here we focus on the first three, as these are the most relevant. Furthermore, we explore only one type of sensor that finds applications in biology, medicine, mechanical engineering, and manufacturing. A strain sensor detects the strain caused by stretching, bending, or twisting and emits an electrical signal (Fig. 15). Such sensors may be resistive, capacitive, or piezoelectric ( Fig. 16) [155]. The piezoresistive sensor detects a change in resistance and is fabricated from conductive materials such as graphene, CNTs, metal nanoparticles, or a conductive polymer. The resistance changes on mechanical deformation and the piezocapacitive sensor detect the associated capacitance change. The basic sensor features conductors and a dielectric. When strain develops, the distance between the conductors (and hence the capacitance) changes. The piezoelectric sensor detects the electrical charges produced. Piezoelectric materials generate charges on mechanical deformation, and either the voltage or current can be measured. Sensitivity is usually defined by reference to how external stimuli change sensor properties. The sensitivity of a piezoresistive sensor is the extent of the external force required to induce a measurable resistance change. GF is often used to compare such sensors, and is where ΔR is the increase in the sensor resistance compared to the initial resistance R and of the induced strain. As it is difficult to compare the performances of sensors that differ in terms of sensing mechanisms, we focus only on the piezoresistive sensors below. Figure 17 and Table 5 show the multi-functionalization research trends among studies that studied piezoresistive strain sensors. Below, the methods used to increase the sensitivity and add stretchability are discussed in detail. Wireless communication by sensors is discussed in Sect. 4.3.

Increasing Performance: Toward High Sensitivity
Sensors exploit geometrical changes, such as cracking, percolation, and tunneling. Figure 18 shows the sensing mechanisms of the 0D, 1D, and 2D nanomaterials. The latter two materials exhibit more stable electrical connections under strain. As both graphene and CNTs exhibit excellent electrical conductivity and stability, they are frequently used in sensor applications requiring high sensitivity [174,180,187,188]. The scale of these materials is in tens of nanometers in GF = ΔR∕R ,  diameter and a few to tens of micrometers in length. To further increase the sensitivity, graphene and CNTs are mixed with other nanomaterials [172,181,182,184] or combined to form auxetic [177] or yarn [67] structures. Rahim et al. mixed graphene and CNTs and achieved a GF of 20,000 [182]. Xiao et al. reported a higher GF (1,000,000) using a nanocomposite of graphene and copper mesh [181]. Zero-dimensional materials exhibit fewer and shorter electrical connections when stretched compared with 1D and 2D materials. However, metal nanoparticles that have a diameter of a few to hundreds of nanometers have recently been used to increase sensor GFs. For example, Zhao et al. fabricated a sensor from silver nanoparticles and graphene; the GF achieved was 100,000 [75]. Metal nanoparticles are added not because of their intrinsic electrical properties, but rather because of changes in the nanogaps between the nanoparticles. In contrast to a typical commercial strain gauge featuring a bulk metal thin film, the resistance changes with an increase in length (because of the associated thinning of the crosssectional area) governed by Ohm's law, the contact resistance between nanoparticles increases dramatically as the nanogap increases. An electron tunneling model is used to guide the development of nanoparticle synthesis (in terms of diameter and homogeneity); due to this, sensitivity increases further. However, because the connection is unstable, stretchability is limited when only 0D materials are employed. There is a trade-off between sensitivity and stretchability, which is discussed in Sect. 4.2. Sensors exploiting mechanical cracking are similar to those of nanoparticle sensors. When strain is applied, a mechanical crack creates a nanogap, greatly changing the resistance and thus imparting high sensitivity. Zhou et al. exploited mechanical cracking when fabricating sensors from CNTs [68].
In electromechanical sensing applications, which is the scope of this section, the main advantage of using nanomaterials is their higher electrical conductivity and sensitivity. In nanomaterials, the electron mean free path is in a similar range as the diameter; therefore, the quantum confinement effect occurs. The electrical properties differ from those of bulk materials. Therefore, nanomaterials enable faster and higher performance. Both graphene and CNTs have excellent electrical conductivity. As discussed, sensors fabricated with 0D materials can have high sensitivity with nanogaps and electron tunneling effects. Graphene and CNTs also consist of structures with high porosities, which can increase the sensitivity. The materials themselves have a high piezoresistive gauge factor.
In addition, the performance of sensors can be easily designed through appropriate material selection. For nanoparticles, as the diameter decreases, the electromagnetic properties change accordingly. The bandgap of nanowires varies with their wire diameter, and their electrical properties can be tuned by doping. The properties of CNTs depend  on their diameter, length, crystallinity, surface, number of walls, and shape. Furthermore, nanomaterials have stronger mechanical properties than bulk materials. Therefore, sensors fabricated with nanomaterials can exhibit higher mechanical properties with light weight. The good adhesion of nanomaterials can be another advantage of using nanomaterials. When printing sensors using the bottom-up method, the patterns should attach firmly to the substrate to increase stability. As nanomaterials have a high surface energy and maintain good adhesion with the substrate, long-term stability can be achieved.
Another advantage of nanomaterials is that they have a high reaction rate because of their high surface-to-volume ratio. A large number of broken bonds increase the surface reactivity. They normally have a lower activation energy than the bulk materials. Chemical and bio-sensing applications utilize these characteristics to increase the number and speed of reactions in sensors. The number of effective reaction areas increases, resulting in higher sensitivity.

Appending Properties: OVERCOME the Trade-Off Between Sensitivity and Stretchability
Given the increasing demand for wearable, skin-mountable devices that measure human body motion, respiration, and voice, current strain sensors must be flexible and stretchable. Stretchability increases the measuring range and is important if the deformations are large. In addition, soft actuators and soft robotics require stretchable sensors and electronics. Three strategies have been used to enhance stretchability. The first employs 1D or 2D nanomaterials that are very stretchable because their longitudinal components are (obviously) greater than those of 0D materials; they remain conductive under strain (Fig. 18). Zhang et al. used a carbonized, plain weave cotton fabric on an Ecoflex substrate to increase the stretchability to 140% [164]. Liao et al. increased the stretchability to 150% using CNTs and polyurethane (PU), employing a soak coating/hydrothermal method to create textiles [92]. Liu et al. described a sensor with GF and stretchability of 100% and 150%, respectively. Single-walled CNTs (SWCNTs) were allowed to self-assemble on a PDMS substrate [175]. Yamada et al., using a similar approach, achieved a stretchability of 280% [166]. Two-dimensional materials are even more stretchable. Cheng et al. used graphene to develop a strain sensor with a stretchability of 700% [178]. Furthermore, liquids can be used to fabricate stretchable strain sensors. Chossat et al. fabricated a 100% stretchable strain sensor employing an ionic solution and eutectic gallium-indium (EGaIn) Ecoflex substrate [190].
The second method employs nanocomposites. Metal nanoparticles may crack when used to measure large deformations. Strain measurement is very sensitive, but only within a limited range [156,157]. Hence, CNTs and graphene have been used to create sensors with high aspect ratios that are mechanically stiff and electrically conductive (Fig. 19). CNTs and graphene exhibit features that render them especially suitable for stretchable sensing, but are not very sensitive. Hence, nanocomposites of both metal nanoparticles and CNTs or graphene have been formulated to ensure both high sensitivity and stretchability. Zhang et al. increased the stretchability by up to 95% by mixing silver nanoparticles with CNTs [66]; stretchability was greater than that imparted by nanoparticles alone. Zhao et al. combined AgNPs with   [197] 1 3 graphene to ensure high sensitivity and stretchability. A fragment structure was employed to maximize sensitivity. The GF was 100,000, and the stretchability was 120% [75]. Guo et al. mixed CNTs with carbon black, and the sensor stretchability was 120% [72]. The nanoparticles imparted sensitivity and the CNTs and graphene stretchability; the nanocomposite exploits both features via a mechanical or chemical reaction. However, a trade-off is in play; sensitivity or stretchability predominates depending on the composition ratio of the two nanomaterials. Advanced manufacturing processes were used to reduce this trade-off. Finally, designs that endure many applications of strain increase stretchability. The most popular structure is wrinkles [70,159,167,168,170]. Figure 20a shows the fabrication process. A stretchable substrate is pre-stretched during material deposition and, on release, forms wrinkles that can be initially stretched flat and then further stretched. Wei et al. fabricated a wrinkled structure from a silver nanowire and polyimide fiber with a stretchability of 400% [159]. Park et al. reported a higher stretchability (750%) using CNTs on an Ecoflex substrate to create wrinkles [168]. Ryu et al. used a similar approach and achieved a stretchability of 900% [170]. Yarns and fiber networks (especially the latter) have been widely studied [67,176]. Figure 20b shows the shape of the sample yarn. This is durable under strain because the fibers are strongly interwoven. Foroughi et al. twisted SWC-NTs and PU into yarn with a stretchability of 300% [176]. Wang et al. electrospun a fibrous network using rGO and thermoplastic polyurethane. The stretchability was 250%, and the GF was 79 [185]. Wang et al. woven MWCNTs and rubber fibers into textiles; the stretchability was 800% [89].
Likewise, three strategies exist for achieving stretchability; here we analyze each strategy. The use of nanomaterials with length components is the most popular and effective method. Graphene and CNTs are the most frequently used materials based on a literature review. This is because they not only have high electrical conductivity and strong mechanical properties, but also have good stretchability. Nanocomposites can be more effective by utilizing the advantages of nanomaterials with different shapes and sizes. However, the cost of the materials increases, and it is difficult to handle and fabricate compared to a single material. The third method can append stretchability by using a structural design that has high stretchability, rather than using the advantages of nanomaterials. By introducing the structural design, this method has a greater possibility of increasing stretchability while having more flexibility in material selection. Wrinkle structures as well as sandwich, multilayer, auxetic, yarn, and porous structures are widely used. Normally, these structures are fine structures with nanometer to micrometer scale. Therefore, it is necessary to develop effective fabrication methods. The fabrication processes are discussed in detail in Sect. 4.4. In addition, studies on substrate  [199] materials that can stretch more than 1000% without cracks are required for a highly stretchable strain sensor.

New Physics: Beyond the Limits of Wired Sensing
There is a need for strain sensors that operate in environments that are hostile to humans. However, the requirement for additional data acquisition and the wire used to transmit sensor signals increase costs and may reduce accuracy. Thus, untethered (wireless) transmission capacity has been added to certain sensors. Such sensors can be used to monitor structural health when an electrical connection is expensive or when a human cannot enter the sensor environment (perhaps because of lethal gas). Wireless communication exploits various methods that are not mechanical or electrical (Table 5). Wireless communication is often achieved by integrating a commercial wireless communication module into a strain sensor [158,200,201]. However, only sensors with standalone wireless communication systems have been studied. RFID electromagnetic circuits are simple, cheap, and untethered. In addition, as the size of the material decreases, the magnetic characteristics become dominant. Therefore, a smaller magnetic field is required for magnetization. Cheng et al. developed a stretchable, largearea, wireless strain sensor using microfluidic, stretchable radiofrequency electronics. The sensor is fabricated as a mechanically reconfigurable antenna to measure the strain and communicate data via RF signals [84]. Kim et al. used silver nano-ink to create an antenna for LC resonator-based, chipless passive RFID tags [83]. The performance of wireless communications depends on the design of the patterns, normally in the range of hundreds of nanometers to tens of micrometers. In addition, to prevent any loss of space and make the product compact, it is important to design and print an antenna in a limited space. Therefore, direct printing methods are normally used to print nanomaterials and secure device density. Recently, strain sensors exploiting structural colors have been prepared via advanced optical nanofabrication technology. These sensors exploit Bragg's diffraction law. The distance between the nanostructures or fabricated nanopatterns varies within the wavelength range of visible light (380-700 nm), and the strain changes the color. It should be noted that the GF unit for wireless   [192]. Fudouzi et al. embedded polystyrene colloidal particles in a PDMS film to change the structural color [193]. Shang et al. used an external magnetic field to align superparamagnetic colloidal nanocrystal clusters into a 1D, chain-like, photonic crystal structure embedded in a polymer matrix [195]. Haque et al. employed unidirectionally aligned lamellar bilayers to develop a structural color in a hydrogel [197].  Figures 21,22,23,24 show the manufacturing processes combining nanoparticles, nanowires, nanotubes, and graphene, respectively. This illustrates a variety of real-world fabrication methods. Nanoparticle self-assembly was earlier a popular form of manufacturing (Fig. 21), but it has decreased over the past 10 years. Electrospinning is rapidly becoming popular. In addition, the demand for direct printing has increased. 3D printing is widely used for design autonomy and flexibility to mass-produce customized items and reflect each customer's needs. Chemical vapor deposition and physical vapor deposition processes have developed considerably and have reached their optimizations. They are frequently used in various nanomaterials. In particular, the chemical vapor deposition of nanowires and nanotubes has been widely studied. However, this method is currently less used to fabricate nanowires, but it remains commonly employed to create nanotubes. In addition, to mass-produce a variety of products and to meet a quick timeline of the products, molding has become popular. Research on the molding process is conducted to enable precise and rapid production, replacing conventional complex and expensive processes. Graphene structures are usually prepared via chemical vapor deposition and self-assembly, but 3D printing and molding are becoming more widely used. This analysis helps in understanding the trends and combination of the fabrication process used for handling various nanomaterials. This can be a guideline for researchers attempting to use nanomaterials for the first time.

Combinations of Nanomaterials and Manufacturing Process
The sensor market is expanding exponentially, and the importance of sensors is constantly increasing. Based on the literature, sensors made with nanomaterials exhibit high

Conclusions
Nanomaterials have opened up new horizons in natural science and engineering. Scientists have focused on nanomaterial synthesis and physical phenomena that are lacking on a macro scale; furthermore, engineers are constantly seeking valuable applications. Multi-functionalization increases the value of nanomaterials by enhancing and appending properties and deriving new physics. We explored the use of nanomaterials in sensors, actuators, energy harvesters, and solar cells. We then applied our strategy to a sensor case study. We expect that not only sensors but also other devices will be multi-functionalized using novel nanomaterials. Our work will aid researchers in drawing research roadmaps that will enable them to achieve commercial success.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.