Functional Nanofibers for Sensors

Abstract

Electrospun nanomaterials and their applications have increasingly gained interest over the last decade. Nanofibers are known for their exceptional surface area and wide opportunities for their functionalization. These properties have been attractive for various sensing applications; however, mostly electric sensing principles have been reported. An overview of most frequently studied concepts will be presented. A novel approach based on optical detection will be described. Various functionalized nanofiber materials have been used to demonstrate feasibility of realization of miniature sensors of biomedical and chemical values (enzymes reactions, metal ions content, concentration, etc.). Compactness and sensitivity of the sensors are significantly enhanced through original hybrid fiber-optic/nanofiber design. The potential of the new detection principle for various applications (bio-medical, chemical, forensic, automotive, etc.) will be discussed.

1. Introduction

Electrospinning is a highly versatile technique to produce continuous fibers with diameters ranging from several micrometers down to few nanometers by applying a high voltage on a solution or melts, mainly from polymers. At nanoscale, several superior characteristics occur such as large surface to volume ratio that can reach values as large as 10³ times of that of micrometer, easy adaptability to surface functionalization, and extraordinary supreme mechanical properties such as stiffness and tensile strength. These outstanding characteristics make electrospun nanofibers an optimal candidate for many important applications [1].

Beside electrospinning, a number of processing methods have been used in recent years to produce polymer nanofibers, such as drawing, self-assembly, template synthesis, and phase separation [2]. Each of these techniques has its limitation, whereas drawing is only limited to viscoelastic materials that can handle the stresses developed during pulling to produce nanofibers, while self-assembly is time consuming in producing continuous polymer nanofibers. Template synthesis uses nanoporous membrane as a template to produce nanofibers of solid (a fibril) or hollow (a tubule) shape. Phase separation takes relatively long time to transfer solid polymer into nanoporous foam. Electrospinning process due to its ease of fabrication appears to be the only technique, which could be further developed for mass production.

The term electrospinning has been used relatively recently; however, its fundamentals dated back more than 60 years earlier. Formhals published a succession of patents [1, 3, 4, 5, 6] from 1934 to 1944. Through this series, he specified the experimental setup for producing polymer filaments using electrostatic force, whereas the polymer solution was exposed to electric field through two electrodes with different polarity. One is placed into the solution, and the other onto the collector. Once the jet solution ejected out from a metal spinneret, it evaporated to become fibers and these fibers were collected on the collector. The potential difference depended on the properties of the solution such as polymer molecular weight and viscosity. The problem occurred that was the fibers favored to stick to each other as well as to the collectors. This problem was due to the insufficient distance between the spinneret aperture and the collectors, which led to inadequate time for jet solution to evaporate. In 1936, C.L. Norton approach was patented due to his contribution to electrospinning from a melt rather than solution using air blast to boost fiber formation [7], Rozenblum and Petryanov-Sokolov [8] in 1938 produced electrospun fiber that was developed into filter materials. These filter materials were then mass manufactured for gas masks. Sir Geoffrey Ingram Taylor stablished the underpinning of a theory for electrospinning between 1964 and 1969. He explained the mathematical model of the cone shape of the fluid droplet under the electric field [9, 10, 11]. In the report of the National Institutes of Health (NIH), The Small Business Innovation Research 1988, Simon produces a submicron- and nanoscale fibrous mats from electrospinning. These mats were especially created for use as substrates in vitro cell [12]. In the beginning of 1990, many organic polymers have been successively elctrospun into nanofibers. Credit for that goes to many research groups, remarkably Reneker and Rutledge, who familiarize the name electrospinning for the process. This process can be simply explained when a sufficiently high voltage is applied to a liquid droplet, and it will charge the body of the droplet. The electrostatic repulsion will generate to counteract the surface tension. Hence, the droplet will stretch. At critical point named Taylor zone, a stream of liquid will be erupted from the surface. If the molecular cohesion of the liquid is sufficiently high, a charged jet will be formed; otherwise, droplets are electrosprays. As the jet flies in air, it will eventually dry and will deposit on the grounded collector.

The standard laboratory setup for electrospinning apparatus consists of spinneret connected to high-voltage (5–50 kV) direct current power supply as illustrated in Figure 1.

Figure 1.

Electrospinning apparatus schematic.

There is a variety of solutions that can be loaded into the syringe, for example, polymer solution, sol–gel, particulate suspension, or melt [13].

By controlling the processing parameters, different nanofiber morphologies can be obtained (Figure 2a–m [14]), beaded, smooth [15], helical [16], ribbon [17], necklace-like [18], porous [18], core-shell [19], hollow [20], multichannel-tubular [21], nanowire-in microtube [22], muli-core cable-like [23], tube-in-tube structured nanofibers [24].

Figure 2.

Different nanofiber morphologies: (a) beaded (b) smooth, (c) helical, (d) ribbon, (e) necklace-like, (f,g) porous, (h) core-shell, (i) hollow, (j) multichannel-tubular, (k) nanowire-in microtube, (l) muli-core cable-like (m) tube-in-tube structured nanofibers.

2. Electrospinning and sensors

The main properties that should be provided in any material to be used as a sensor are: firstly, to be responsive to the external stimuli; secondly, this response can be accessible to electronic interface; thirdly, this has a large specific area since sensing preferentially occurs at interface. The material that possesses the first property is called smart material. The stimuli can be pressure, temperature, PH, moisture, chemical substances, electric, magnetic field, or light. In order for these smart materials to a sensor, it should act as a transducer. That means to response to the external stimuli in a way which can be measurable. In other words, it converts the external stimuli into a quantity that can be measurable.

Electrospinning manifests the capabilities of smart materials at the nanoscale dimension, especially as sensing materials. At nanoscale dimensions, many features are accessible, for example, excellent mechanical properties, especially flexibility, high porosity, large surface area, ability to surface functionality, and the ability to produce not only one-dimensional (1D), but also three-dimensional (3D) materials. Due to dramatic decrease in the diameter of the fibers, this has a great impact on the surface area, which is significantly increased, consequently the number of sites increase to interact with the external environment more effectively. High porosity provides utmost channels for transporting among nanofibers in electrospun mats, hence speeding up the transportation mechanism and increasing sensitivity. Another aspect of electrospinning is in its ability to form continuous nanofibers. This feature is so important in sensors, because sensors are usually assembled into a certain measuring instrument or analog-to-digital conversion circuit; therefore, it should provide a stable continuous circuit to supply a path for the current. Hence, electrospinning is irreplaceable to provide a stable circuit. Additionally, electrospinning makes use of various materials from inorganic to organic matters.

Smart materials, also named stimuli-responsive materials, are capable of undergoing reversible physical/chemical change upon exposure to external stimulus, such as temperature, PH, electrical, magnetic, light, chemicals, ions. Integration of these stimuli responsive materials with nanotechnology, such as electrospinning, has enormously accelerated the development of sensors.

3. Transduction sensing mechanisms

3.1 Capacitive sensors based on electrospun nanofibers

Capacitive sensors mainly depend on changing the relative permittivity of the dielectric material between two conducting electrodes. Capacitive devices are often used as displacement and pressure sensors.

Yang et al. [25] developed a flexible capacitive pressure sensor based on electrospun polyvinylidene fluoride (PVDF) nanofiber membrane with carbon nanotubes (CNTs). The fabrication process and schematic diagram of CNT-PVDF composite nanofiber are shown in Figure 3a,b. Two pieces of indium tin oxide polyethylene terephthalate films connected with cooper wires were fixed on the top and bottom surface of the composite nanofiber membrane as electrodes to record the capacitance variation under external pressure. The schematic diagram and the actual diagram of the single sensor are shown in Figure 3c. The SEM images of the composite nanofibers are shown in Figure 3d,g with different CNT weight ratios of 0.03, 0.05, 0.1, and 0.2 wt % of PVDF. At the beginning, the increase of CNT led to decrease the diameter of nanofibers, and then, the diameter increases again. This is because the when CNT increases the conductivity, the electrostatic force between the collector and the syringe spinneret will increase. Therefore, the nanofibers will be pulled thinner. Any further increase of CNT will lead to increase of nanofibers diameter as the CNTs tend to agglomerate due to strong Van der Waals force.

Figure 3.

Flexible capacitive pressure sensor: (a) schematic of the fabrication of the composite nanofiber. (b) schematic diagram of the composite nanofiber. (c) Diagram of the sensor. (d)–(h) SEM images of the composite nanofiber membrane with 0.03, 0.05, 0.1, 0.2 wt % carbon nanotube additions, respectively [25].

By increasing the permittivity and decreasing the young’s modulus of the CNT-PVDF dielectric layer, the capacitive sensor exhibited high sensitivity (∼0.99/kPa) with a composition of 0.05 wt% CNTs (Figure 4a), fast response (∼29 ms), and excellent cyclic loading/unloading stability (>1000 cycles) (Figure 4b).

Figure 4.

Characterization of the pressure sensing performance of the flexible capacitive sensor. (a) The relative change in capacitance of the sensor with different weight ratio CNTs addition under low pressure applied. (b) Experimental systems for dynamic pressure applying and measuring capacitance, and enlarged view of a portion of the figure.

3.2 Resistive sensors based on electrospun nanofibers

Resistive sensors rely on the measuring the change of electrical resistivity as a variable of the amount of the reactive analytical samples through surface reaction. Up to now, numerous attempts have been carried on to develop ultrasensitive sensors to detect NH₃, CO₂, CO, O₂, H₂S, moisture, volatile organic compounds (VOCs) [26, 27, 28, 29, 30]. Resistive sensors based on electrospun nanofibers provide high and quick gas response via rapid and effective diffusion of analytic gas on the sensing surface. The porosity of the sensitive layer supply high-diffusivity paths for target molecules, promoting gas diffusion and mass transportation across sensing surface. This leads to faster response/recovery times. Moreover, porous surface boosts the formation of more chemisorbed oxygen species [31, 32].

3.2.1 Inorganic nanofibers

Resistive sensors based on metal oxide semiconductors (MOS) are the most simple and versatile gas sensors [33].

The resistance of the metal oxide is changing in accordance with the adsorption of the gases. One-dimensional (1D) nanostructured MOS have attracted much attention as chemical sensor materials as a result of their large surface-to-volume ratio, high porosity, excellent surface activities, and high surface charge modulation depth. ZnO, SnO₂, TiO₂, NiO, and LaFeO₃ are the most 1D MOS, which have been widely developed in the creation of highly sensitive gas sensors [34, 35].

The sensing mechanism of MOS gas sensors can be illustrated as follows:

In pure air, donor electrons in metal oxide attract to the oxygen, which is adsorbed into the surface of sensing material, preventing current flow (Figure 5a), while in the presence of the target gas (Figure 5b), oxygen reacts with the reducing gases. Hence, surface density of adsorbed oxygen decreases, and those electrons are then released into MOS, allowing current to flow freely through the sensor.

Figure 5.

Sensing mechanism of MOS gas sensors: (a) in clean air and (b) in the presence of the target gas.

Abundance of n-type semiconductors such as ZnO, SnO₂, TiO₂, In₂O₃, WO₃, and ZnO/SnO₂ have turned out to be excellent gas materials for detecting both reducing and oxidizing gases, including H₂, NH₃, ethanol, acetone, and toluene. TiO₂ is the most well-known MOS used in ultrasensitive resistive sensors. Kim et al. have indicated the use of TiO₂ nanofibers as a detector for NO₂ [30]. Wang et al. reported that ZnO nanofibers with an average diameter of 150 nm display excellent sensing properties against ethanol at an operating temperature of 300°C, with a rapid response of about 3 s, including short recovery time of about 8 s and high sensitivity [36]. Lately, SnO₂ has attracted much attention because of its high transparency, semi-conductivity, wide-band gap, and huge magneto-optic and chemical sensing effects [37, 38]. A highly porous SnO₂ nanofibers were prepared by combining electrospinning with oxygen plasma etching. They displayed fast response (7 s), wide linear response range, and low detection limit (< 1 ppb) [39].

In addition, doping is an efficient method to improve the sensing properties of the sensors. Li et al. have demonstrated that LiCl-doped TiO₂ nanofibers have an enhanced sensitivity toward humidity better than pure TiO₂ nanofibers [40]. Moreover, the composite nanofibers have ultra-fast response and recovery time. Zhang et al. developed double-layer ZnO/In₂O₃ composite nanfibers for sensing ethanol. ZnO/In₂O₃/ZnO displayed improved and excellent sensing properties compared with ZnO nanofibers (detection limit of 1 ppm, shorter response, and recovery time of 2 and 1 s, respectively).

In addition to n-type semiconductors, p-type semiconductors have also been used to prepare vapor sensors, including NiO, Cr₂O₃, LaFeO₃, CuO, LaOCl/NiO, etc. Fan et al. [41] produced LaFeO₃ nanofibers-based ethanol sensor with good reversibility and selectivity and fast response and recovery time. Electrospun LaOCl/NiO composite nanofibers have significant performance in ethanol sensing against CO, NO₂, H₂, NH₃, due to incorporation of NiO that catalyzes gas sensing reaction [42].

Many methods have been carried on to improve the sensitivity, response, and recovery time, for example, combining p-type with n-type metal oxide semiconductors to form p-n junction remarkably improving the sensing characteristic [43, 44], functionalizing the surface of nanofibers with catalytic nanoparticle (such as Ag, Pd, Pt) [45], and doping salts (KCl, LiCl, NaCl, and MgCl₂) into nanofibers, especially in humidity sensors [46].

3.2.2 Organic nanofibers

Organic polymers, especially conducting polymers (CPs) as an alternative to inorganic semiconductors, provide attractive features such as mechanical flexibility, easy processing, and adaptable electrical conductivity. Many research efforts have been dedicated to the development of nano-sensors based on CPs such as PANI, polythiophene, and their derivatives. However, CPs have poor solubility in common solvents, which restrict its application. Many routes have been developed to overcome this drawback, for example, incorporating CPs into other polymeric systems (such as PS, PEO, CA) or synthesized in other conducting forms (oxidized, reduced) [47]. The charge transport for CPs is primarily due to hopping mechanism. This hopping occurred because of changing polymer resistance in the presence of a sensing gas. This change can be due to chemical change (doping/de-doping), conformational change, or polymer swelling.

Electrospinning provides abundance of activated sites for CPs immobilization due to its unique features, such as large surface area, high porosity, and large stacking density. Pinto NJ et al. [48] demonstrated that the electrospun-isolated nanofibers of poly(3,4-ethylenedioxythiophene) doped with poly(styrene sulfonic acid) can be used to sense vapors (NH₃, HCL, NO₂, aliphatic alcohols).

3.2.3 Hybrid nanofibers

Hybrid nanofiber-based sensors have been developed to overcome the drawbacks of inorganic nanofiber-based sensors (require high-operating temperature) and organic nanofiber-based sensors (low sensitivity). Few researchers have investigated hybrid nanofibers-based sensing devices, and they got promising results in terms of sensitivity, response time, and reversibility [49, 50]. Researchers should devote their work to improve stability, selectivity, and reusability of the sensors.

3.3 Fiber optic sensors

All the aforesaid nanofibers-based sensors depend mainly on electrical sensing principle; however, in some cases, electricity is not suitable for the target sensing analytes. Hence, the importance of using optical sensors is necessary. Among the optical properties that have been utilized at sensors is reflectivity, refractive index, color, and absorption coefficient. Refractive index has been investigated by our work team, and it was effective.

Fiber optic sensor technology has been rapidly developed in the past 30 years due to the innovations in telecommunication, semiconductor, and electronics sectors that have significantly reduced the prices of optical components and stimulated the development of optical fiber sensor [51]. Optical fiber sensors are capable of measuring a wide variety of physical properties, such as chemical changes, strain, electric and magnetic fields, pressure, temperature, displacement (position), radiation, flow, liquid level, vibrations, and light intensity. Optical fiber sensors exhibit a number of advantages over the conventional electrical and electronic sensors:

• Are non-electrical devices

• Require small cable sizes and weight that enable small sensor sizes

• Allow access into inaccessible areas

• Permit remote sensing

• Immune to radio frequency and electromagnetic interference

• Do not contaminate their surroundings and are not subject to corrosion

• Provide high sensitivity, resolution, and dynamic range

• Offer sensitivity to multiple environmental parameters

It is believed that optical fiber sensors will replace the conventional devices for the measurement of various physical, chemical, and biological parameters. Optical fiber sensors are dielectric devices that are chemically inert. They do not require electric cables and are technically ideal for working in hostile media, and corrosive environment for remote sensing applications [52, 53, 54, 55].

In the following section, we will display some of our team efforts of hybrid fiber-optic/nanofiber sensors developments.

3.3.1 Hybrid fiber-optic/nanofiber sensors

3.3.1.1 Bio-medical sensors

Petrík et al. [52] have produced SiO₂ nanofibers. The surface of SiO₂ nanofibers was functionalized with enzymes. Figure 6 shows the SEM images of nanofibers with and without enzyme immobilization. The functionalized nanofibers were attached at the tip of Y junction plastic optical fiber. The setup of the experiment is displayed in Figure 7. One end of Y junction is attached to an LED as light source, and the other is connected to the photodetector. The light emitted by the LED is carried to the sensing point and partially reflected back. The reflected power was observed to be a function of enzyme concentration as shown in Figure 8. The detection limit was nearly 10% of the full initial level.

Figure 6.

SEM pictures of nanofibers without (a) and with (b) immobilized enzyme.

Figure 7.

Setup of the tested optical fiber sensor with an enlargement of the detection part.

Figure 8.

Reflected intensity vs. concentration of a model enzyme-substrate.

The results of these experiments are very optimistic to the effectiveness of the prescribed optic fiber system with nanofibers. This system can be used as a basis of a wide family of optic fiber sensors sensitive to various chemical and biological substances. The proposed optical fiber sensor can be integrated into security systems for fast and cost effective.

Main advantages of the approach are as follows:

• Chemically inert materials—possibility to disinfect/sterilize

• Miniature dimensions

• Sufficiently high sensitivity

Another work effort from our group is a trial to estimate the water content in brake fluid using hybrid optical fiber/nanofiber as it will be explained later.

3.3.1.2 Waste-water cleaning process monitoring with nanofiber/fiber-optic sensors

We have used similar approach in a proof-of-concept study of using nanofiber/fiber-optic sensors for monitoring of waste-water bio-cleaning process.

Activity of bacteria in sludge water was monitored using online and offline optical fiber sensing system that utilizes the nanofibrous membranes. The used optics showed reasonable sensitivity levels to the slight changes in water compositions due to the presence of slurry matters and the formation of biofilms on the surface of the nanofibrous membranes. In general, the online setup showed better performance compared with the offline system that has inhomogeneous formation of bacterial films. As a future continuation of this work, other fibrous systems with higher compatibility and growing conditions for bacteria will be used. Also, functionalization of the fibers with elements that attract bacteria will be implemented in the upcoming work. Moreover, the experimental setup will adopt an online measurement system for mobilized and flowing bioreactors. The experimental setup is shown in Figure 9. An example of the sensor response to the bacteria activity is illustrated in Figure 10.

Figure 9.

Pictures for the used fiber optics and the online measurement setup.

Figure 10.

Reflected intensity as a function of time detected by the hybrid fiber optic/nanofiber sensor.

3.3.1.3 Silica nanofibers-based sensor concept for water content measurement

Focus of this research is to build and investigate an optical fiber sensor based on silica nanofibers prepared by a reliable and low-cost electrospinning technique to detect water content in DOT-4 brake fluid. To the best of our knowledge, this is a novelty study of optical fiber sensor to detect water content in an aqueous substance using electrospinning nanofibers.

The nanofiber processed by electrospinning has a larger specific surface area compared with conventional coating film, which can absorb a large number of water molecules. In some recently published articles [56, 57], dielectric properties of silica-based hybrid nanostructures and thin films have been investigated in which capacitance and dielectric constant act as a function of frequency. Batool et al. [58] studied the effect of RH on dielectric response of SiO₂ nanofibers; however, it is rarely investigated the effect of RH on refractive index of SiO₂ nanofibers.

The method used in this work involves utilization of silica nanofibers. The full description of preparation of the nanofibers can be found in patent WO 2017/186201 [59]. The composite PVP/SiO₂ nanofibers were left in the air for 24 h for hydrolysis of TEOS. Subsequently, PVP/SiO₂ nanofibers were annealed at 800°C for 6 h in furnace to obtain pure SiO₂ nanofibers. Figure 11a shows the scanning electron microscope (SEM) image of pure SiO₂ nanofibers after the removal of PVP, annealed at 800°C for 6 hours. The nanofibers have diameters ≈ 150 ̶ 200 nm. Figure 11b shows the energy-dispersive spectrum of SiO₂ nanofibers. The presence of atomic % of Si and O in the sample indicates formation of SiO₂ and complete removal of PVP. Si-O-Si bonds (siloxane groups) at 1087 and 797 Cm⁻¹ become more intense after heat treatment as shown in Figure 11c.

Figure 11.

(a) SEM images of SiO₂ nanofibers heat treated at 800°C for 6 h, (b) EDS spectrum, (c) Fourier-transformed infrared spectroscopy of TEOS/PVP electrospun fibers before (B) & after (A) heat treatment from both sides interior (i) & exterior (o).

Measurements were made with 0–7% water added to the brake fluid. The amount of water that was added to the brake fluid was determined according to dry basis moisture content (designated M_d in the text) is described by the percentage equivalent of the ratio of the weight of water (W_W) to the weight of the dry matter (W_d), herein is DOT-4.

Dry Basis Moisture Content is defined by Eq. (1):

$$\mathrm{Md}=100\times \left(\mathrm{Wet}\text{Weight}–\mathrm{Dry}\text{Weight}\right)/\mathrm{Dry}\text{Weight}.$$E1

Commercial brake fluid tester was utilized for checking the percent of water content presented in brake fluid.

Silica nanofibers were glued on the tip of 2x Multimode optical fiber 50/125 μm, optical power meter as a source of input light, and a detector of the reflected light. OFS has been immersed in a brake fluid while changing its water content.

Figure 12 shows the change in the power intensity as a function of water content in brake fluid which is almost linearly. That is probably related to water molecules that will be absorbed and concentrated in the pores of the silica nanofibers. This effect will alter the refractive index (RI) of the silica nanofibers, hence changing the optical power intensity. As a result, the accumulation of the water molecules will cause the increase of effective refractive index of the surrounding medium. This will lead to the leakage of the light through evanescent field [60, 61]. This proposed sensor based on reflected light intensity modulation. Based on Fresnel reflection, a proportion of lights are leaked when the sensor is in the liquid. This amount of light depends on the refractive index of the liquid. For normal incidence, the reflectance simplifies to the following equation Eq. (2).

Figure 12.

OFS response to the water change in brake fluid in terms of changing light intensity.

$$\mathrm{R}={\left|\frac{n_1-{\mathrm{n}}_2}{n_1+n_2}\right|}^2$$E2

• $n_1$, $n_2$ are the RIs of optical fiber and exterior; $n_1$ are known by the fabrication of the optical fiber,

• $n_2$ is variable according to the exterior medium.

The experiments are still going on while controlling humidity and temperature to assign the parameters, which could influence the accuracy and repeatability of the potential sensor.

4. Conclusion

Hundreds of papers are being published per year on “sensing” nanofibers.

Electrospinning looks like the most versatile method for their fabrication. Many unique sensor designs require just mm² of the nanofiber mat per unit (single use or multiple/continuous measurement). Definitely, they will not generate market opportunities for 1–2 m width production lines available on the market (Elmarco, Innovenso). The producers should probably consider the development of small volume special machines. Some of the “lab tools” the offer will probably fulfill the market needs.

Smart membranes/textiles are much more compatible with current production lines offered to the market. But the technological processes will be probably challenging and will need further development (chemistry, depositions of special substances, inter-operations, after-treatments, etc.).

Acknowledgments

This work was supported by the Ministry of Education, Youth and Sports in the Czech Republic under the “Inter Excellence – Action programme” within the framework of project “Micro-struCtural imaging as a Tool for modelinG fibrOus materiALS (μ-CT GOALS)” (registration number LTAUSA18135). Also, this work was supported by the Ministry of Education, Youth and Sports of the Czech Republic and the European Union - European Structural and Investment Funds in the frames of Operational Programme Research, Development and Education - project Hybrid Materials for Hierarchical Structures (HyHi, Reg. No. CZ.02.1.01/0.0/0.0/16_019/0000843).