sensors Article
Piezoelectric Active Humidity Sensors Based on Lead-Free NaNbO3 Piezoelectric Nanofibers Li Gu 1 , Di Zhou 2 and Jun Cheng Cao 1, * 1 2
*
The Key Laboratory of Terahertz Solid-State Technology, Shanghai Institute of Microsystem and Information Technology, Chinese Academy of Science, Shanghai 200050, China; [email protected] School of Physics and Information Engineering, Jianghan University, Wuhan 430056, China; [email protected] Correspondence: [email protected]; Tel.: +86-21-6251-1070 (ext. 8805)
Academic Editor: Vittorio M. N. Passaro Received: 8 April 2016; Accepted: 30 May 2016; Published: 7 June 2016
Abstract: The development of micro-/nano-scaled energy harvesters and the self-powered sensor system has attracted great attention due to the miniaturization and integration of the micro-device. In this work, lead-free NaNbO3 piezoelectric nanofibers with a monoclinic perovskite structure were synthesized by the far-field electrospinning method. The flexible active humidity sensors were fabricated by transferring the nanofibers from silicon to a soft polymer substrate. The sensors exhibited outstanding piezoelectric energy-harvesting performance with output voltage up to 2 V during the vibration process. The output voltage generated by the NaNbO3 sensors exhibited a negative correlation with the environmental humidity varying from 5% to 80%, where the peak-to-peak value of the output voltage generated by the sensors decreased from 0.40 to 0.07 V. The sensor also exhibited a short response time, good selectively against ethanol steam, and great temperature stability. The piezoelectric active humidity sensing property could be attributed to the increased leakage current in the NaNbO3 nanofibers, which was generated due to proton hopping among the H3 O+ groups in the absorbed H2 O layers under the driving force of the piezoelectric potential. Keywords: sodium niobate; humidity sensors
piezoelectric energy harvesting;
nanofibers;
self-powered;
1. Introduction Nowadays, the development of high-performance humidity sensors has attracted great attention because monitoring and controlling environmental humidity is important for industry, agriculture, and people’s daily lives [1]. Thanks to their high sensitivity, simple fabrication processes and device structures, as well as the high integrity of IC techniques, low-dimensional semiconductor nanomaterials are widely investigated as resistance-type humidity sensing materials [2–6]. Their working mechanism can be attributed to the increase in conductivity after the water molecules are absorbed on the surface of the sensing layers. Moreover, there are also capacitive sensors with a high dielectric response to the variation in relative humidity, which exhibit fast response speed and high temperature compensation [7–9]. However, both the resistive and capacitive humidity sensors require an external power unit such as lithium ion batteries. The relatively large size of the batteries and their need for frequent recharging or replacement has limited the miniaturization and intelligentization of modern smart-sensor systems. Since the power consumption of the micro-scaled sensors has decreased to a much lower level due to the miniaturization of sensing layers, the electrical power generated by energy harvesters such as solar cells, thermoelectric devices, and piezoelectric nanogenerators is sufficient to maintain the Sensors 2016, 16, 833; doi:10.3390/s16060833
www.mdpi.com/journal/sensors
Sensors 2016, 16, 833
2 of 9
sensors [10]. Among them, the piezoelectric nanogenerators that can convert mechanical vibration energy into electrical power has exhibited unique advantages compared to the former versions because mechanical energy is ubiquitous in our environment [11–14]. As a result, mechanical energy harvesting will be less affected by working conditions than optical and thermal energies. For instance, Xu et al. have demonstrated self-powered pH and UV sensors with a ZnO nanowire-based piezoelectric nanogenerator as their powering sources [15]. Partial voltage applied to the sensors by the nanogenerators changed with the sensor resistance when the pH value and UV irradiation around the sensor changed. Moreover, Lee and co-authors have reported a self-powered environment sensor, which consists of a piezoelectric nanogenerator and single-wall carbon nanotubes (SWCNT) [13]. The sensor could be used for detecting mercury ions in a water solution by lighting up a light-emission diode (LED) with the nanogenerators once the concentration of mercury ions is beyond the alarm level. However, the integration of nanogenerator and sensor units in the self-powered system is complicated because the interference of mechanical vibration from the nanogenerators on the sensor units should be eliminated to guarantee system stability. Recently, a novel type of self-powered sensor system was demonstrated by Zhu and his co-authors, showing that the ZnO nanowires could act as both the piezoelectric energy harvester and the sensor unit [10,16,17]. The working mechanism of such “active sensors” can be attributed to the polarization screening effect, which was induced by the redistribution of free charge carriers in the nanowires under piezoelectric potential. As the free charge carrier density of the ZnO nanowires was sensitive to the surface-absorption of chemicals such as hydrogen, ethanol, water, and glucose, those devices could be used for detecting the target subject by monitoring the variation in generated output voltage [18–21]. Therefore, both the piezoelectric and chemical sensing properties of the nanowires are crucial for building high-performance active sensors. Although the ZnO nanowires can play both roles in the devices, the limited piezoelectric constant and poor gas sensing selectivity may limit the practical application of such active sensors. As reported, the perovskite NaNbO3 nanofibers (NFs) have exhibited ultra-high sensitivity and selectivity to water molecules at room temperature. The surface-absorption of water molecules leads to a great enhancement of a NF’s conductivity [22]. Simultaneously, the NaNbO3 nanowires have also been reported to have an outstanding piezoelectric energy-harvesting performance by Jung et al. [23]. Those results inspired us to consider the potential of the NaNbO3 nanowires for building high-performance active humidity sensors. In this work, the perovskite NaNbO3 NFs were synthesized using a far-field electrospinning method. A flexible active humidity sensor based on the NaNbO3 NFs was assembled on the polydimethylsiloxane (PDMS) substrate. The energy-harvesting behavior of the devices under various humidity conditions was investigated. The device exhibited highly sensitive active humidity sensing performance at room temperature. Moreover, the working mechanism of the active humidity sensor was also discussed in detail. 2. Materials and Methods The NaNbO3 NFs were firstly synthesized on the clean silicon substrates through a far-field electrospinning method. The detailed experimental procedures are as follows: First, the NaNbO3 sol-gel of 0.4 mol/L was prepared by the reflux reaction of the niobium ethoxide and sodium acetate anhydrous in the mixture of acetic acid and 2-methoxyethanol with a volume ratio of 1:3. The reaction lasted for 1 h at 95 ˝ C and was protected in a dry N2 atmosphere to prevent hydrolysis of the niobium ethoxide. All chemical reagents were analytically pure and purchased from Alfa Aesar in China. After the reaction, a transparent yellow NaNbO3 solution was obtained. Then, the electrospinning precursor was prepared by mixing the NaNbO3 solution with the ethanol solution of poly(vinylpyrrolidone) (PVP, MW-1300000, 0.08 g/mL). After stirring for 12 h at room temperature, the precursor was transferred into a 5-mL plastic syringe, which was then positioned onto the syringe pump to electrospin the NFs. The clean silicon substrate was positioned on the metal sample collector, which was placed 10 cm away from the syringe tip. The electrospinning process was then started at a voltage of 1.4 kV/cm, and the precursors ejected at a constant rate of 0.36 mL/h. Finally, the
Sensors 2016, 16, 833
3 of 9
substrates were covered by white composite NFs of the NaNbO3 solution, and the PVP was dried at 80 ˝ C for 5 h. The dried composite NFs were firstly pre-sintered at 450 ˝ C for 1 h to combust the carbons and then annealed at 700 ˝ C to form the perovskite NaNbO3 polycrystalline NFs. All heat treatments were done in an oxygen atmosphere. The structure and morphology of the samples was characterized by the X-ray diffractometer (XRD, Bruker D8 Advanced, CuKα, λ = 0.15406 nm) and field-emission scanning electron microscopy (FESEM, JSM7100F). In order to2016, fabricate the flexible nanogenerators, the prepared NaNbO3 NFs were Sensors 16, 833 3 oftransferred 9 from the silicon substrate onto the flexible PDMS substrate. Firstly, a certain amount of PDMS carbons and then annealed at 700 °C to form the perovskite NaNbO3 polycrystalline NFs. All heat silicone was spin-coated on the samples and then cured at 60 ˝ C to make the NFs stiff on the PDMS treatments were done in an oxygen atmosphere. The structure and morphology of the samples was polymers. characterized Then, the silicone was diffractometer carefully lifted offBruker from the substrateCuKα, to transfer the NFs onto the soft by the X-ray (XRD, D8 Advanced, λ = 0.15406 nm) and polymer substrate. After that, electron the Pt interdigital electrodes (IDEs) were deposited on the NFs/PDMS field-emission scanning microscopy (FESEM, JSM7100F). In order to fabricate theduring flexible the nanogenerators, preparedFinally, NaNbO3the NFsdevices were transferred surface by using a shadow mask sputteringthe process. were completed from the silicon substrate onto the flexible PDMS substrate. Firstly, a certain amount of PDMS silicone after wire leading. Then, the NaNbO3 NFs were poled with a DC voltage of 20 kV/cm at room was spin-coated on the samples and then cured at 60 °C to make the NFs stiff on the PDMS polymers. temperature. The the devices then shorted for 30 the min toonto remove the stored charges. Then, theelectrodes silicone was of carefully lifted offwere from the substrate to transfer NFs the soft polymer The piezoelectric harvesting and active humidity sensing performance of the devices were substrate.energy After that, the Pt interdigital electrodes (IDEs) were deposited on the NFs/PDMS surface by using a shadow mask during the sputtering process. Finally, the devices were completed after measured with a self-designed measuring system as well as a data acquisition (DAQ) card (National wire leading. Then, the NaNbO3 NFs were poled with a DC voltage of 20 kV/cm at room temperature. Instruments USB-6210), a digital multimeter (Keithley 2000), a charge amplifier (Shenzhen Cheng Tec, The electrodes of the devices were then shorted for 30 min to remove the stored charges. The Shenzhen, piezoelectric China, CT5002), the WS-60A gas humidity sensor testing Among them, the were Keithley 2000 energyand harvesting and active sensingsystem. performance of the devices multimetermeasured was used for testing the output voltage by the sensors when the experiment was with a self-designed measuring system generated as well as a data acquisition (DAQ) card (National USB-6210), a digital multimeter (Keithley 2000), a charge (Shenzhenwas Cheng conductedInstruments inside an air-dry oven to test the temperature stability. Theamplifier DAQ USB-6210 used to test Tec, Shenzhen, China, CT5002), and the WS-60A gas sensor testing system. Among them, the Keithley the output voltage at room temperature. 3.
2000 multimeter was used for testing the output voltage generated by the sensors when the experiment was conducted inside an air-dry oven to test the temperature stability. The DAQ USBResults6210 andwas Discussions used to test the output voltage at room temperature.
Figure3. Results 1a shows the SEM image of the as-synthesized NaNbO3 NFs on the silicon substrate and Discussions after the annealing treatment. The NFs were randomly arranged on the substrate with a good size Figure 1a shows the SEM image of the as-synthesized NaNbO3 NFs on the silicon substrate after distribution and smooth surface morphology. The average diameter of the NFs was ~78 nm according the annealing treatment. The NFs were randomly arranged on the substrate with a good size to the statistical result shown Figure 1b. The The XRD resultdiameter shownofin 2 ~78 confirms the perovskite distribution and smoothin surface morphology. average theFigure NFs was nm according to the statistical result shown in Figure 1b. The XRD result shown in Figure 2 confirms the perovskite structure of the NFs, with all diffraction peaks indexed into the monoclinic phase of NaNbO3 (JCPDS structure of the NFs, with all diffraction peaks indexed into the monoclinic phase of NaNbO3 (JCPDS card No. 74-2437). card No. 74-2437).
Figure 1. (a) The field-emission scanning electron microscopy (FESEM) images of the as-synthesized
Figure 1. (a) The 3field-emission electronatmicroscopy (FESEM) images of the image, as-synthesized NaNbO nanofibers (NFs) scanning after being annealed 700 °C. The inset picture is the magnified ˝ and (b) is the statistical result of the NFs’ diameters. NaNbO3 nanofibers (NFs) after being annealed at 700 C. The inset picture is the magnified image, and (b) is the statistical result of the NFs’ diameters.
Figure 3 shows the schematic fabrication procedure of the flexible NaNbO3 active sensor. Because the heat treatment is necessary for the formation of perovskite NaNbO3 NFs, silicon werethe usedschematic for the synthesis of the NFs. procedure Due to the porous structure of the NFs on the3silicon Figuresubstrates 3 shows fabrication of the flexible NaNbO active sensor. the top layer of the NFs could be packaged by the PDMS soft polymers through the spinBecause thesubstrate, heat treatment is necessary for the formation of perovskite NaNbO3 NFs, silicon substrates coating and the curing process. Thereafter, the NFs could be easily transferred onto the flexible PDMS were used substrate for the synthesis ofthe thePDMS NFs.layer. Due After to thetheporous structure the NFs on based the silicon by lifting off IDE deposition, theofflexible device on the substrate, the top layer of the NFs could becould packaged by the polymers through the3,spin-coating and electrospun NaNbO 3 NFs be obtained. AsPDMS shown soft by the photo image in Figure the as3 in dimension. fabricated device was approximately 2.0 × 1.0 × 0.2 cm the curing process. Thereafter, the NFs could be easily transferred onto the flexible PDMS substrate
by lifting off the PDMS layer. After the IDE deposition, the flexible device based on the electrospun NaNbO3 NFs could be obtained. As shown by the photo image in Figure 3, the as-fabricated device was approximately 2.0 ˆ 1.0 ˆ 0.2 cm3 in dimension.
Sensors 2016, 16, 833
4 of 9
Sensors 2016, 16, 833
4 of 9
Sensors 2016, 16, 833
4 of 9
Sensors 2016, 16, 833
4 of 9
Figure 2. The X-ray diffractometer (XRD) pattern of the NaNbO3 NFs after being annealed at 700 °C
Figure 2. The X-ray diffractometer (XRD) pattern of the NaNbO3 NFs after being annealed at 700 ˝ C in oxygen. Figure in oxygen. 2. The X-ray diffractometer (XRD) pattern of the NaNbO3 NFs after being annealed at 700 °C in oxygen.
Figure 2. The X-ray diffractometer (XRD) pattern of the NaNbO3 NFs after being annealed at 700 °C in oxygen.
Figure 3. NaNbO3 NF-based active sensors fabrication procedure. Figure 3. NaNbO3 NF-based active sensors fabrication procedure.
The inset picture 4a illustrates the piezoelectric energy-harvesting performance-testing Figurein3.Figure NaNbO 3 NF-based active sensors fabrication procedure. method, in which the poled device is fixed at one end and forced by a rotating stick to bend. As shown The inset picture in Figure 4a illustrates the piezoelectric energy-harvesting performance-testing in Figure 4a, the output voltage generated by the device without NaNbO3 NFs was lower than 0.2 V, method, in which the poled device is fixed at one end and forced by a rotating stick to bend. As shown Thewhich insetwas picture in Figure illustrates theactive piezoelectric energy-harvesting performance-testing Figure 3.4a NaNbO 3 NF-based sensors fabrication procedure. too weak to be recognized from the strong background noise, and only 10% of the output in Figure 4a, the output voltage generated by the device without NaNbO3 NFs was lower than 0.2 V, voltage was the generated by the sensors with NaNbO 3 NFs (Figure 4b). The much lower voltage signal method,which in which poled device is fixed at one end and forced by a rotating stick to bend. As shown wasinset too picture weak toinbeFigure recognized from the background noise, and only 10% of the output The 4a illustrates thestrong piezoelectric energy-harvesting performance-testing generated from the blank sample may be by due to releasing stored remnant charges between the in Figure 4a, the output voltage generated the device without NaNbO NFs was lower than voltage generated by the sensors with at NaNbO 3 NFs (Figure 4b). The much lower voltage signal method,was in which thethe poled device is fixed one end forced by a rotating stick 3to bend. As shown electrodes, because thickness of the PDMS layerand changed during the bending process. These 0.2 V, which was4a, too weak tovoltage be recognized from the strong background noise, only 10% generated from the blank sample may beby due to releasing stored remnant chargesand between the of the in Figure thethe output generated without NaNbO 3 NFs was lower than 0.2 V, results suggest energy-harvesting behaviorthe of device the device should be attributed to the piezoelectric because the thickness of the PDMS layer changed during the bending process. These voltage output electrodes, voltage was generated by the sensors with NaNbO NFs (Figure 4b). The much lower 3 which was too weak to be recognized from the strong background noise, and only 10% of the output effect of the NaNbO3 NFs. results thethe energy-harvesting the device should beThe attributed to the piezoelectric voltagesuggest wasfrom generated by thesample sensorsbehavior with NaNbO 3 to NFs (Figure 4b). much lower voltage signal signal generated blank may be of due releasing stored remnant charges between the effect of thefrom NaNbO generated the3 NFs. blank sample may be due to releasing stored remnant charges between the
electrodes, because the thickness of the PDMS layer changed during the bending process. These results because the thickness of of thethe PDMS layer changed the bending process. Theseeffect of suggest electrodes, the energy-harvesting behavior device should be during attributed to the piezoelectric results suggest the energy-harvesting behavior of the device should be attributed to the piezoelectric the NaNbO3 NFs. effect of the NaNbO3 NFs.
Figure 4. The output voltage generated by the active sensor (a) without and (b) with the NaNbO3 NFs when the sensors were periodically bent by a rotating stick. Figure 4. The output voltage generated by the active sensor (a) without and (b) with the NaNbO3 NFs when the sensors were periodically bent by a rotating stick. Figure 4. The output voltage generated by the active sensor (a) without and (b) with the NaNbO3 NFs
Figure 4.when Thethe output voltage generatedbent by the sensor sensors were periodically by aactive rotating stick. (a) without and (b) with the NaNbO3 NFs when the sensors were periodically bent by a rotating stick.
Sensors 2016, 16, 833
5 of 9
Sensors52016, 16, 833 5 ofreversely 9 Figure shows the inversion of voltage polarity when the electrodes of the device were connected to the DAQ card. When it was connected to the DAQ card with forward direction, a positive Figure 5 shows the inversion of voltage polarity when the electrodes of the device were reversely voltage connected peak wastofirstly recorded with the device bending motion (Figure 5a). The positive polarity the DAQ card. When it was connected to the DAQ card with forward direction, a of the voltage peak can be attributed to the axial stretching of NaNbO under the positive bending state, 3 NFs positive voltage peak was firstly recorded with the device bending motion (Figure 5a). The Sensors 2016, 16, 833 5 of 9 which may generate the piezoelectric potential along the same direction of the poling electrical polarity of the voltage peak can be attributed to the axial stretching of NaNbO3 NFs under the field. bending state, which may generate piezoelectric potential along ofnegative the polingvoltage Once the connection wasthe reversed, the device bending resulted in the thesame generation of Figure 5 shows inversion ofthe voltage polarity when the electrodes of thedirection device were reversely electrical field. Once thecard. connection was reversed, device bending resulted in the of connected toThose the DAQ When it was connected topiezoelectric the DAQ card with forward direction, a peaks (Figure 5b). results also confirmed thatthe the effect of the generation NaNbO was 3 NFs negative voltage peaks (Figure 5b). Those results also confirmed that the piezoelectric effect of the positive voltage peak was firstly recorded with the device bending motion (Figure 5a). The positive the predominant mechanism of the energy-harvesting behavior [24].
NaNbO3 of NFs the predominant behavior polarity thewas voltage peak can bemechanism attributed oftothe theenergy-harvesting axial stretching of NaNbO[24]. 3 NFs under the bending state, which may generate the piezoelectric potential along the same direction of the poling electrical field. Once the connection was reversed, the device bending resulted in the generation of negative voltage peaks (Figure 5b). Those results also confirmed that the piezoelectric effect of the NaNbO3 NFs was the predominant mechanism of the energy-harvesting behavior [24].
Figure 5. The output voltage generated by the NaNbO3 sensor with the electrodes connected (a)
Figure 5.straight The output generated bysystem. the NaNbO3 sensor with the electrodes connected (a) straight and (b)voltage in reverse to the DAQ and (b) in reverse to the DAQ system. It is known that the sensitivity of sensor systems mainly depends on the signal-to-noise ratio (SNR) of the Because the high-frequency sampling and much higher impedance of the Figure 5. signals. Thethe output voltageofgenerated by the NaNbO 3 sensor with the electrodes (a) It is known that sensitivity of sensor systems mainly depends on theconnected signal-to-noise ratio NaNbO 3 NFs (up to TΩ) compared with the input impedance of the DAQ cards (~MΩ), the noise in straight and (b) in reverse to the DAQ system. (SNR) of the signals. Because of the high-frequency sampling and much higher impedance of the the recorded voltage signals was unstable and too high to recognize the relatively lower voltages NaNbOduring NFs (up to TΩ) compared with the input impedance of the DAQ cards (~MΩ), the noise in the sampling process [25]. of This limited the sensitivity of theon energy harvesters asratio the 3 It isthe known that the sensitivity sensor systems mainly depends the signal-to-noise piezoelectric active sensors, in which the generated was the sensitivity signal for thevoltages detection recorded voltage signals was unstable too highvoltage tosampling recognize relatively lower (SNR) of the signals. Because of theand high-frequency andthe much higher impedance of the during of humidity As a result, a charge wasof used for theDAQ signal conditioning, including NaNbO 3 NFschanges. (up[25]. to TΩ) compared with input impedance the cards (~MΩ), noise in the sampling process This limited thetheamplifier sensitivity theof energy harvesters asthe the piezoelectric the impedance matching and filtering. Figure 6 shows the to comparison of recorded the recorded voltage was unstable and toothe high recognize the relatively lowersignals voltages active sensors, in which thesignals generated voltage was sensitivity signal for thevoltage detection of by humidity the DAQthe cards betweenprocess the unconditioned and the conditioned signals. shown,harvesters the ratio between sampling [25]. This limited of theAsenergy as the changes.during As a result, a charge amplifier was used forthe thesensitivity signal conditioning, including the impedance the output voltage and noise amplitude was ~10 for the unconditioned signals and could be increased piezoelectric active sensors, in which the generated voltage was the sensitivity signal for the detection matching and filtering. Figure 6 shows the comparison of recorded voltage signals by the DAQ cards to humidity ~20 after signal conditioning. the voltage decreased due to the impedance of changes. As a result, Although a charge amplifier wasamplitude used for the signal conditioning, including betweenthe the unconditioned and the conditioned signals. As the ratio between matching, the increased could enhance sensitivity of theshown, sensors. Moreover, the stability of output impedance matchingSNR and filtering. Figurethe 6 shows the comparison of recorded voltage signalsthe by the voltage signal also improved duefor to filtering theconditioned high-frequency noise, which resulted inbetween a stable to ~20 voltage the and noise amplitude was ~10 the unconditioned signals and could increased DAQ cards between the unconditioned and the signals. As shown, thebe ratio baseoutput line signal andand improved thethe accuracy the voltage noise amplitude was of ~10the forsensors. the unconditioned and could be increased after signal conditioning. Although voltage amplitude decreasedsignals due to the impedance matching, to ~20 after Although the voltage amplitudeMoreover, decreased due the impedance the increased SNRsignal couldconditioning. enhance the sensitivity of the sensors. the to stability of the voltage matching, the increased SNR could enhance the sensitivity of the sensors. Moreover, the stability of signal also improved due to filtering the high-frequency noise, which resulted in a stable base line the voltage signal also improved due to filtering the high-frequency noise, which resulted in a stable signal and the improved accuracythe of the sensors. baseimproved line signal and accuracy of the sensors.
Figure 6. Comparison of recorded voltage signals generated by the NaNbO3 sensors with and without signal conditioning.
Figure 6. Comparison of recorded voltage signals generated by the NaNbO3 sensors with and without
Figure 6. Comparison of recorded voltage signals generated by the NaNbO3 sensors with and without signal conditioning. signal conditioning.
Sensors 2016, 16, 833
6 of 9
Figure 7 shows the piezoelectric active humidity response of the NaNbO3 sensors with the voltage signal conditioned by the charge amplifier. The sensing performance was tested by changing the humidity of the environment when the NaNbO3 sensors were continuously vibrating under the knocks of the rotating stick. As shown in Figure 7a–h, the sensor could generate distinguishable impulsive voltage with the humidity changing from 5% to 95% RH, where the peak-to-peak value of the voltage decreased from 0.4 to 0.06 V. Figure 7i shows the relationship between the voltage and the humidity of the environment. The voltage amplitude exhibited a negative linear correlation with the humidity varying from 5% to 80% RH and saturated when the humidity rose higher than 80% RH. As reported, the piezoelectric energy-harvesting behavior of the NaNbO3 NFs could be attributed to the electron flow in the external circuit under the driving force of the piezoelectric potential [26–28]. The electrons would accumulate at the positive side of the electrode due to the Schottky potential barrier between the Pt electrode and NaNbO3 NFs and compensate the piezoelectric potential. Once the piezoelectric potential was removed due to the release of strain, the accumulated electrons would flow back to the positive side along the external circuit and generate a negative voltage peak. As reported, the electrical field applied along the axial direction of the NaNbO3 NFs could lead to protons hopping among the surface-absorbed H3 O+ groups [22,29]. Therefore, an additional charge flow could be generated on the surface of the NaNbO3 NF when the piezoelectric potential was generated due to bending motion. This additional charge flow would lead to leakage current inside the piezoelectric materials and thus decrease the output voltage of the devices. When the environmental humidity changed, the concentration of protons on the surface of the NFs also changed. As a result, the output voltage changed due to the variation in leakage current. Moreover, the leakage current density in the NFs was in a linear correlation with the concentration of protons and absorbed water molecules [17]. Therefore, the output voltage exhibited a negative linear relationship with the environmental humidity at 5% to 80% RH. However, the absorption of water molecules was saturated when the humidity increased further (>80% RH), which resulted in the slight decrease in output voltage shown in Figure 7i. Finally, the sensitivity of the active sensor could be obtained as ~2 mV/%RH according to the linear result fitting within the humidity range of 5% to 80% RH. The relatively lower sensitivity of this NaNbO3 sensor compared to the reported CeO2 /ZnO and SnO2 /ZnO sensors should be attributed to the decreased output voltage amplitude after the signal conditioning. However, it is worth noting that the structure of the NaNbO3 nanofiber-based sensors was much more stable than the sensors based on the ZnO nanorod arrays, which was assembled through the stacking of the top electrode and the nanorod arrays. Moreover, the sensor also exhibited fast response speed to the variation in humidity. As shown in Figure 8a, the response time of the sensors for the humidity changing from 65% to 95% RH is ~12 s. Considering the time taken for the change in humidity, the real value of the response time should be shorter. Figure 8b shows the output voltage generated by the sensors when ethanol steam and H2 gas was introduced to and then removed from the testing chamber, respectively. The output voltage did not exhibit any change during the introducing and removing process of ethanol and H2 , which confirmed the excellent selectivity of the sensor. In order to evaluate the influence of temperature on the sensing result, the energy conversion property of the sensors under 38% RH was tested when the temperature increased from 24 to 80 ˝ C. As shown in Figure 8c, no obvious change in the voltage amplitude could be found, which suggested that the sensor possessed great temperature stability during the testing process.
did not exhibit any change during the introducing and removing process of ethanol and H2, which confirmed the excellent selectivity of the sensor. In order to evaluate the influence of temperature on the sensing result, the energy conversion property of the sensors under 38% RH was tested when the temperature increased from 24 to 80 °C. As shown in Figure 8c, no obvious change in the voltage amplitude could Sensors 2016, 16, 833 be found, which suggested that the sensor possessed great temperature stability 7 of 9 during the testing process.
Sensors 2016, 16, 833
7 of 9
Sensors 2016, 16, 833
7 of 9
Figure 7. (a–h) The output voltage generated by the active sensor at different humidity conditions; (i) Figure 7. (a–h) The output voltage generated by the active sensor at different humidity conditions; the relationship between the peak value of output voltage and the humidity. (i) the relationship the peak value ofby output voltage and at thedifferent humidity. Figure 7. (a–h) The between output voltage generated the active sensor humidity conditions; (i) the relationship between the peak value of output voltage and the humidity.
Figure 8. (a) The transient response of the sensor when the humidity increased from 65% to 95% RH. (b) The transient response of the sensor when ethanol steam and H2 was introduced and removed, Figure 8. (a) The transient response of the sensor when the humidity increased from 65% to 95% RH. respectively. (c) The output voltage of of the sensor when at 38%the RHhumidity under different temperature. Figure 8. (a) The transient response increased from 65% to 95% RH. (b) The transient response of the sensor when ethanol steam and H2 was introduced and removed, (b) The transient response of the sensor when ethanol steam and H2 was introduced and removed, respectively. (c) The output voltage of the sensor at 38% RH under different temperature. 4. Conclusions respectively. (c) The output voltage of the sensor at 38% RH under different temperature.
In this work, monoclinic NaNbO3 piezoelectric NFs with a uniform size distribution were 4. Conclusions 4. Conclusions synthesized using a far-field electrospinning method. By lifting off the PDMS layer coated on the NFs In this work, monoclinic NaNbO3 piezoelectric NFs with a uniform size distribution were and In thethis electrode flexible active humidity obtained. sensors could work, deposition, monoclinic aNaNbO NFssensor with awas uniform size The distribution were 3 piezoelectric synthesized using a far-field electrospinning method. By lifting off the PDMS layer coated on the NFs generate an impulsive output voltage with an amplitude up to 2 V during the vibration process. synthesized using a far-field electrospinning method. By lifting off the PDMS layer coated on The the and the electrode deposition, a flexible active humidity sensor was obtained. The sensors could energy-harvesting performance was confirmed to have originated the piezoelectric effect could of the NFs and the electrode deposition, a flexible active humidity sensorfrom was obtained. The sensors generate an impulsive output voltage with an amplitude up to 2 V during the vibration process. The NaNbO3 an NFs. By conditioning the voltage with a charge thethe noise in the recorded generate impulsive output voltage withsignal an amplitude up toamplifier, 2 V during vibration process. energy-harvesting performance was confirmed to have originated from the piezoelectric effect of the voltage signals was suppressed bywas half, which could be originated attributedfrom to matching impedance The energy-harvesting performance confirmed to have the piezoelectric effectand of NaNbO3 NFs. By conditioning the voltage signal with a charge amplifier, the noise in the recorded filtering high frequency signals. The output voltage of the NaNbO 3 sensors after the signal the NaNbO NFs. By conditioning the voltage signal with a charge amplifier, the noise in the recorded 3 voltage signals was suppressed by half, which could be attributed to matching impedance and conditioning a negative linear correlation with the environmental humidity the voltage signalsexhibited was suppressed by half, which could be attributed to matching impedance andwhen filtering filtering high frequency signals. The output voltage of the NaNbO3 sensors after the signal humidity changed fromThe 5% output to 80%voltage RH. The humidity-dependent output provides the high frequency signals. of the NaNbO3 sensors after thevoltage signal conditioning conditioning exhibited a negative linear correlation with the environmental humidity when the potential of the NaNbO3 device to be applied as an active humidity sensor with the sensitivity of 2 humidity changed from 5% to 80% RH. The humidity-dependent output voltage provides the mV/%RH. The variation in output voltages with humidity can be attributed to the humiditypotential of the NaNbO3 device to be applied as an active humidity sensor with the sensitivity of 2 dependent leakage current inside the NaNbO3 NFs, which was induced by protons hopping among mV/%RH. The variation in output voltages with humidity can be attributed to the humiditythe surface-absorbed H3O+ under the driving force of piezoelectric potential. The active sensor also dependent leakage current inside the NaNbO3 NFs, which was induced by protons hopping among exhibited a fast response speed, as well as great selectivity and temperature stability. The active +
Sensors 2016, 16, 833
8 of 9
exhibited a negative linear correlation with the environmental humidity when the humidity changed from 5% to 80% RH. The humidity-dependent output voltage provides the potential of the NaNbO3 device to be applied as an active humidity sensor with the sensitivity of 2 mV/%RH. The variation in output voltages with humidity can be attributed to the humidity-dependent leakage current inside the NaNbO3 NFs, which was induced by protons hopping among the surface-absorbed H3 O+ under the driving force of piezoelectric potential. The active sensor also exhibited a fast response speed, as well as great selectivity and temperature stability. The active humidity sensor based on NaNbO3 NFs provides an effective solution for self-powered sensor systems with high sensitivity, simple structure and fabrication processes, and low production costs. Acknowledgments: This work was supported by the National “973” Program of China (Grant No. 2014CB339803), the “863” Program of China (Project No. 2011AA010205), and the National Science Foundation of China (Grant No. 51502114). Author Contributions: Li Gu performed the materials synthesis, device fabrication, and sensor testing experiments. Di Zhou contributed reagents and materials. Juncheng Cao conceived and designed the experiments. Conflicts of Interest: The authors declare no conflict of interest.
Abbreviations The following abbreviations are used in this manuscript: SWCNTs NFs LED SNR XRD DAQ IDE
single-wall carbon nanotubes Nanofibers light emission diode signal-to-noise ratio X-ray diffraction data acquisition interdigital electrode
References 1. 2. 3. 4. 5. 6. 7.
8.
9. 10.
Blank, T.A.; Eksperiandova, L.P.; Belikov, K.N. Recent trends of ceramic humidity sensors development: A review. Sens. Actuators B Chem. 2016, 228, 416–442. [CrossRef] Fu, X.Q.; Wang, C.; Yu, H.C.; Wang, Y.G.; Wang, T.H. Fast humidity sensors based on CeO2 nanowires. Nanotechnology 2007, 18, 776–789. [CrossRef] Kuang, Q.; Lao, C.; Wang, Z.L.; Xie, Z.; Zheng, L. High-Sensitivity Humidity Sensor Based on a Single SnO2 Nanowire. J. Am. Chem. Soc. 2007, 129, 6070–6071. [CrossRef] [PubMed] Matko, V.; Donlagic, D. Sensor for high-air-humidity measurement. IEEE Trans. Instrum. Meas. 1996, 45, 561–563. [CrossRef] Matko, V. Next generation AT-cut quartz crystal sensing devices. Sensors 2011, 11, 4474–4482. [CrossRef] [PubMed] Matko, V.; Milanovi´c, M. Temperature-compensated capacitance-frequency converter with high resolution. Sens. Actuators A Phys. 2014, 220, 262–269. [CrossRef] Dai, C.-L.; Liu, M.-C.; Chen, F.-S.; Wu, C.-C.; Chang, M.-W. A nanowire WO3 humidity sensor integrated with micro-heater and inverting amplifier circuit on chip manufactured using CMOS-MEMS technique. Sens. Actuators B Chem. 2007, 123, 896–901. [CrossRef] Zhang, Y.; Fu, W.; Yang, H.; Li, M.; Li, Y.; Zhao, W.; Sun, P.; Yuan, M.; Ma, D.; Liu, B.; et al. A novel humidity sensor based on Na2 Ti3 O7 nanowires with rapid response-recovery. Sens. Actuators B Chem. 2008, 135, 317–321. [CrossRef] Zhang, Y.; Yu, K.; Jiang, D.; Zhu, Z.; Geng, H.; Luo, L. Zinc oxide nanorod and nanowire for humidity sensor. Appl. Surf. Sci. 2005, 242, 212–217. [CrossRef] Guo, H.; Chen, J.; Tian, L.; Leng, Q.; Xi, Y.; Hu, C. Airflow-induced triboelectric nanogenerator as a self-powered sensor for detecting humidity and airflow rate. ACS Appl. Mater. Interface 2014, 6, 17184–17189. [CrossRef] [PubMed]
Sensors 2016, 16, 833
11.
12. 13. 14.
15. 16.
17.
18.
19.
20.
21.
22. 23. 24. 25. 26. 27. 28. 29.
9 of 9
Hwang, G.T.; Park, H.; Lee, J.H.; Oh, S.; Park, K.I.; Byun, M.; Park, H.; Ahn, G.; Jeong, C.K.; No, K.; et al. Self-powered cardiac pacemaker enabled by flexible single crystalline PMN-PT piezoelectric energy harvester. Adv. Mater. 2014, 26, 4880–4887. [CrossRef] [PubMed] Wang, Z.L. Self-powered nanosensors and nanosystems. Adv. Mater. 2012, 24, 280–285. [CrossRef] [PubMed] Lee, M.; Bae, J.; Lee, J.; Lee, C.-S.; Hong, S.; Wang, Z.L. Self-powered environmental sensor system driven by nanogenerators. Energy Environ. Sci. 2011, 4. [CrossRef] Wang, Z.; Zhang, Y.; Yang, S.; Hu, Y.; Wang, S.; Gu, H.; Wang, Y.; Chan, H.L.W.; Wang, J. (K,Na)NbO3 Nanofiber-based Self-Powered Sensors for Accurate Detection of Dynamic Strain. ACS Appl. Mater. Interface 2015, 7, 4921–4927. [CrossRef] [PubMed] Xu, S.; Qin, Y.; Xu, C.; Wei, Y.; Yang, R.; Wang, Z.L. Self-powered nanowire devices. Nat. Nanotechnol. 2010, 5, 366–373. [CrossRef] [PubMed] Hsueh, H.-T.; Chen, Y.-H.; Lin, Y.-D.; Lai, K.-C.; Chen, J.-W.; Wu, C.-L. Integration of flower-like ZnO nanostructures with Crystalline-Si interdigitated back contact photovoltaic cell as a self-powered humidity sensor. Appl. Phys. Lett. 2013, 103. [CrossRef] Zhu, D.; Fu, Y.; Zang, W.; Zhao, Y.; Xing, L.; Xue, X. Piezo/active humidity sensing of CeO2 /ZnO and SnO2 /ZnO nanoarray nanogenerators with high response and large detecting range. Sens. Actuators B Chem. 2014, 205, 12–19. [CrossRef] Zhao, Y.; Lai, X.; Deng, P.; Nie, Y.; Zhang, Y.; Xing, L.; Xue, X. Pt/ZnO nanoarray nanogenerator as self-powered active gas sensor with linear ethanol sensing at room temperature. Nanotechnology 2014, 25, 115502. [CrossRef] [PubMed] Zang, W.; Nie, Y.; Zhu, D.; Deng, P.; Xing, L.; Xue, X. Core-Shell In2 O3 /ZnO Nanoarray Nanogenerator as a Self-Powered Active Gas Sensor with High H2 S Sensitivity and Selectivity at Room Temperature. J. Phys. Chem. C 2014, 118, 9209–9216. [CrossRef] Xing, L.; Hu, Y.; Wang, P.; Zhao, Y.; Nie, Y.; Deng, P.; Xue, X. Realizing room-temperature self-powered ethanol sensing of Au/ZnO nanowire arrays by coupling the piezotronics effect of ZnO and the catalysis of noble metal. Appl. Phys. Lett. 2014, 104. [CrossRef] Fu, Y.; Zang, W.; Wang, P.; Xing, L.; Xue, X.; Zhang, Y. Portable room-temperature self-powered/active H2 sensor driven by human motion through piezoelectric screening effect. Nano Energy 2014, 8, 34–43. [CrossRef] Zhang, Y.; Pan, X.; Wang, Z.; Hu, Y.; Zhou, X.; Hu, Z.; Gu, H. Fast and highly sensitive humidity sensors based on NaNbO3 nanofibers. RSC Adv. 2015, 5, 20453–20458. [CrossRef] Jung, J.H.; Lee, M.; Hong, J.I.; Ding, Y.; Chen, C.Y.; Chou, L.J.; Wang, Z.L. Lead-Free NaNbO3 Nanowires for a High Output Piezoelectric Nanogenerator. ACS Nano 2011, 5, 10041–10046. [CrossRef] [PubMed] Mantini, G.; Gao, Y.; D’Amico, A.; Falconi, C.; Wang, Z.L. Equilibrium piezoelectric potential distribution in a deformed ZnO nanowire. Nano Res. 2009, 2, 624–629. [CrossRef] Koszewnik, A.; Grzes, P.; Walendziuk, W. Mechanical and electrical impedance matching in a piezoelectric beam for Energy Harvesting. Eur. Phys. J. 2015, 224, 2719–2731. [CrossRef] Wang, Z.; Pan, X.; He, Y.; Hu, Y.; Gu, H.; Wang, Y. Piezoelectric nanowires in energy harvesting applications. Adv. Mater. Sci. Eng. 2015, 2015. [CrossRef] Falconi, C.; Mantini, G.; D’Amico, A.; Wang, Z.L. Studying piezoelectric nanowires and nanowalls for energy harvesting. Sens. Actuators B Chem. 2009, 139, 511–519. [CrossRef] Wang, Z.L.; Song, J. Piezoelectric nanogenerators based on zinc oxide nanowire arrays. Science 2006, 312, 242–246. [CrossRef] [PubMed] Liang, S.; He, X.; Wang, F.; Geng, W.; Fu, X.; Ren, J.; Jiang, X. Highly sensitive humidity sensors based on LiCl–Pebax 2533 composite nanofibers via electrospinning. Sens. Actuators B Chem. 2015, 208, 363–368. [CrossRef] © 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).
Piezoelectric Active Humidity Sensors Based on Lead-Free NaNbO3 Piezoelectric Nanofibers Li Gu 1 , Di Zhou 2 and Jun Cheng Cao 1, * 1 2
*
The Key Laboratory of Terahertz Solid-State Technology, Shanghai Institute of Microsystem and Information Technology, Chinese Academy of Science, Shanghai 200050, China; [email protected] School of Physics and Information Engineering, Jianghan University, Wuhan 430056, China; [email protected] Correspondence: [email protected]; Tel.: +86-21-6251-1070 (ext. 8805)
Academic Editor: Vittorio M. N. Passaro Received: 8 April 2016; Accepted: 30 May 2016; Published: 7 June 2016
Abstract: The development of micro-/nano-scaled energy harvesters and the self-powered sensor system has attracted great attention due to the miniaturization and integration of the micro-device. In this work, lead-free NaNbO3 piezoelectric nanofibers with a monoclinic perovskite structure were synthesized by the far-field electrospinning method. The flexible active humidity sensors were fabricated by transferring the nanofibers from silicon to a soft polymer substrate. The sensors exhibited outstanding piezoelectric energy-harvesting performance with output voltage up to 2 V during the vibration process. The output voltage generated by the NaNbO3 sensors exhibited a negative correlation with the environmental humidity varying from 5% to 80%, where the peak-to-peak value of the output voltage generated by the sensors decreased from 0.40 to 0.07 V. The sensor also exhibited a short response time, good selectively against ethanol steam, and great temperature stability. The piezoelectric active humidity sensing property could be attributed to the increased leakage current in the NaNbO3 nanofibers, which was generated due to proton hopping among the H3 O+ groups in the absorbed H2 O layers under the driving force of the piezoelectric potential. Keywords: sodium niobate; humidity sensors
piezoelectric energy harvesting;
nanofibers;
self-powered;
1. Introduction Nowadays, the development of high-performance humidity sensors has attracted great attention because monitoring and controlling environmental humidity is important for industry, agriculture, and people’s daily lives [1]. Thanks to their high sensitivity, simple fabrication processes and device structures, as well as the high integrity of IC techniques, low-dimensional semiconductor nanomaterials are widely investigated as resistance-type humidity sensing materials [2–6]. Their working mechanism can be attributed to the increase in conductivity after the water molecules are absorbed on the surface of the sensing layers. Moreover, there are also capacitive sensors with a high dielectric response to the variation in relative humidity, which exhibit fast response speed and high temperature compensation [7–9]. However, both the resistive and capacitive humidity sensors require an external power unit such as lithium ion batteries. The relatively large size of the batteries and their need for frequent recharging or replacement has limited the miniaturization and intelligentization of modern smart-sensor systems. Since the power consumption of the micro-scaled sensors has decreased to a much lower level due to the miniaturization of sensing layers, the electrical power generated by energy harvesters such as solar cells, thermoelectric devices, and piezoelectric nanogenerators is sufficient to maintain the Sensors 2016, 16, 833; doi:10.3390/s16060833
www.mdpi.com/journal/sensors
Sensors 2016, 16, 833
2 of 9
sensors [10]. Among them, the piezoelectric nanogenerators that can convert mechanical vibration energy into electrical power has exhibited unique advantages compared to the former versions because mechanical energy is ubiquitous in our environment [11–14]. As a result, mechanical energy harvesting will be less affected by working conditions than optical and thermal energies. For instance, Xu et al. have demonstrated self-powered pH and UV sensors with a ZnO nanowire-based piezoelectric nanogenerator as their powering sources [15]. Partial voltage applied to the sensors by the nanogenerators changed with the sensor resistance when the pH value and UV irradiation around the sensor changed. Moreover, Lee and co-authors have reported a self-powered environment sensor, which consists of a piezoelectric nanogenerator and single-wall carbon nanotubes (SWCNT) [13]. The sensor could be used for detecting mercury ions in a water solution by lighting up a light-emission diode (LED) with the nanogenerators once the concentration of mercury ions is beyond the alarm level. However, the integration of nanogenerator and sensor units in the self-powered system is complicated because the interference of mechanical vibration from the nanogenerators on the sensor units should be eliminated to guarantee system stability. Recently, a novel type of self-powered sensor system was demonstrated by Zhu and his co-authors, showing that the ZnO nanowires could act as both the piezoelectric energy harvester and the sensor unit [10,16,17]. The working mechanism of such “active sensors” can be attributed to the polarization screening effect, which was induced by the redistribution of free charge carriers in the nanowires under piezoelectric potential. As the free charge carrier density of the ZnO nanowires was sensitive to the surface-absorption of chemicals such as hydrogen, ethanol, water, and glucose, those devices could be used for detecting the target subject by monitoring the variation in generated output voltage [18–21]. Therefore, both the piezoelectric and chemical sensing properties of the nanowires are crucial for building high-performance active sensors. Although the ZnO nanowires can play both roles in the devices, the limited piezoelectric constant and poor gas sensing selectivity may limit the practical application of such active sensors. As reported, the perovskite NaNbO3 nanofibers (NFs) have exhibited ultra-high sensitivity and selectivity to water molecules at room temperature. The surface-absorption of water molecules leads to a great enhancement of a NF’s conductivity [22]. Simultaneously, the NaNbO3 nanowires have also been reported to have an outstanding piezoelectric energy-harvesting performance by Jung et al. [23]. Those results inspired us to consider the potential of the NaNbO3 nanowires for building high-performance active humidity sensors. In this work, the perovskite NaNbO3 NFs were synthesized using a far-field electrospinning method. A flexible active humidity sensor based on the NaNbO3 NFs was assembled on the polydimethylsiloxane (PDMS) substrate. The energy-harvesting behavior of the devices under various humidity conditions was investigated. The device exhibited highly sensitive active humidity sensing performance at room temperature. Moreover, the working mechanism of the active humidity sensor was also discussed in detail. 2. Materials and Methods The NaNbO3 NFs were firstly synthesized on the clean silicon substrates through a far-field electrospinning method. The detailed experimental procedures are as follows: First, the NaNbO3 sol-gel of 0.4 mol/L was prepared by the reflux reaction of the niobium ethoxide and sodium acetate anhydrous in the mixture of acetic acid and 2-methoxyethanol with a volume ratio of 1:3. The reaction lasted for 1 h at 95 ˝ C and was protected in a dry N2 atmosphere to prevent hydrolysis of the niobium ethoxide. All chemical reagents were analytically pure and purchased from Alfa Aesar in China. After the reaction, a transparent yellow NaNbO3 solution was obtained. Then, the electrospinning precursor was prepared by mixing the NaNbO3 solution with the ethanol solution of poly(vinylpyrrolidone) (PVP, MW-1300000, 0.08 g/mL). After stirring for 12 h at room temperature, the precursor was transferred into a 5-mL plastic syringe, which was then positioned onto the syringe pump to electrospin the NFs. The clean silicon substrate was positioned on the metal sample collector, which was placed 10 cm away from the syringe tip. The electrospinning process was then started at a voltage of 1.4 kV/cm, and the precursors ejected at a constant rate of 0.36 mL/h. Finally, the
Sensors 2016, 16, 833
3 of 9
substrates were covered by white composite NFs of the NaNbO3 solution, and the PVP was dried at 80 ˝ C for 5 h. The dried composite NFs were firstly pre-sintered at 450 ˝ C for 1 h to combust the carbons and then annealed at 700 ˝ C to form the perovskite NaNbO3 polycrystalline NFs. All heat treatments were done in an oxygen atmosphere. The structure and morphology of the samples was characterized by the X-ray diffractometer (XRD, Bruker D8 Advanced, CuKα, λ = 0.15406 nm) and field-emission scanning electron microscopy (FESEM, JSM7100F). In order to2016, fabricate the flexible nanogenerators, the prepared NaNbO3 NFs were Sensors 16, 833 3 oftransferred 9 from the silicon substrate onto the flexible PDMS substrate. Firstly, a certain amount of PDMS carbons and then annealed at 700 °C to form the perovskite NaNbO3 polycrystalline NFs. All heat silicone was spin-coated on the samples and then cured at 60 ˝ C to make the NFs stiff on the PDMS treatments were done in an oxygen atmosphere. The structure and morphology of the samples was polymers. characterized Then, the silicone was diffractometer carefully lifted offBruker from the substrateCuKα, to transfer the NFs onto the soft by the X-ray (XRD, D8 Advanced, λ = 0.15406 nm) and polymer substrate. After that, electron the Pt interdigital electrodes (IDEs) were deposited on the NFs/PDMS field-emission scanning microscopy (FESEM, JSM7100F). In order to fabricate theduring flexible the nanogenerators, preparedFinally, NaNbO3the NFsdevices were transferred surface by using a shadow mask sputteringthe process. were completed from the silicon substrate onto the flexible PDMS substrate. Firstly, a certain amount of PDMS silicone after wire leading. Then, the NaNbO3 NFs were poled with a DC voltage of 20 kV/cm at room was spin-coated on the samples and then cured at 60 °C to make the NFs stiff on the PDMS polymers. temperature. The the devices then shorted for 30 the min toonto remove the stored charges. Then, theelectrodes silicone was of carefully lifted offwere from the substrate to transfer NFs the soft polymer The piezoelectric harvesting and active humidity sensing performance of the devices were substrate.energy After that, the Pt interdigital electrodes (IDEs) were deposited on the NFs/PDMS surface by using a shadow mask during the sputtering process. Finally, the devices were completed after measured with a self-designed measuring system as well as a data acquisition (DAQ) card (National wire leading. Then, the NaNbO3 NFs were poled with a DC voltage of 20 kV/cm at room temperature. Instruments USB-6210), a digital multimeter (Keithley 2000), a charge amplifier (Shenzhen Cheng Tec, The electrodes of the devices were then shorted for 30 min to remove the stored charges. The Shenzhen, piezoelectric China, CT5002), the WS-60A gas humidity sensor testing Among them, the were Keithley 2000 energyand harvesting and active sensingsystem. performance of the devices multimetermeasured was used for testing the output voltage by the sensors when the experiment was with a self-designed measuring system generated as well as a data acquisition (DAQ) card (National USB-6210), a digital multimeter (Keithley 2000), a charge (Shenzhenwas Cheng conductedInstruments inside an air-dry oven to test the temperature stability. Theamplifier DAQ USB-6210 used to test Tec, Shenzhen, China, CT5002), and the WS-60A gas sensor testing system. Among them, the Keithley the output voltage at room temperature. 3.
2000 multimeter was used for testing the output voltage generated by the sensors when the experiment was conducted inside an air-dry oven to test the temperature stability. The DAQ USBResults6210 andwas Discussions used to test the output voltage at room temperature.
Figure3. Results 1a shows the SEM image of the as-synthesized NaNbO3 NFs on the silicon substrate and Discussions after the annealing treatment. The NFs were randomly arranged on the substrate with a good size Figure 1a shows the SEM image of the as-synthesized NaNbO3 NFs on the silicon substrate after distribution and smooth surface morphology. The average diameter of the NFs was ~78 nm according the annealing treatment. The NFs were randomly arranged on the substrate with a good size to the statistical result shown Figure 1b. The The XRD resultdiameter shownofin 2 ~78 confirms the perovskite distribution and smoothin surface morphology. average theFigure NFs was nm according to the statistical result shown in Figure 1b. The XRD result shown in Figure 2 confirms the perovskite structure of the NFs, with all diffraction peaks indexed into the monoclinic phase of NaNbO3 (JCPDS structure of the NFs, with all diffraction peaks indexed into the monoclinic phase of NaNbO3 (JCPDS card No. 74-2437). card No. 74-2437).
Figure 1. (a) The field-emission scanning electron microscopy (FESEM) images of the as-synthesized
Figure 1. (a) The 3field-emission electronatmicroscopy (FESEM) images of the image, as-synthesized NaNbO nanofibers (NFs) scanning after being annealed 700 °C. The inset picture is the magnified ˝ and (b) is the statistical result of the NFs’ diameters. NaNbO3 nanofibers (NFs) after being annealed at 700 C. The inset picture is the magnified image, and (b) is the statistical result of the NFs’ diameters.
Figure 3 shows the schematic fabrication procedure of the flexible NaNbO3 active sensor. Because the heat treatment is necessary for the formation of perovskite NaNbO3 NFs, silicon werethe usedschematic for the synthesis of the NFs. procedure Due to the porous structure of the NFs on the3silicon Figuresubstrates 3 shows fabrication of the flexible NaNbO active sensor. the top layer of the NFs could be packaged by the PDMS soft polymers through the spinBecause thesubstrate, heat treatment is necessary for the formation of perovskite NaNbO3 NFs, silicon substrates coating and the curing process. Thereafter, the NFs could be easily transferred onto the flexible PDMS were used substrate for the synthesis ofthe thePDMS NFs.layer. Due After to thetheporous structure the NFs on based the silicon by lifting off IDE deposition, theofflexible device on the substrate, the top layer of the NFs could becould packaged by the polymers through the3,spin-coating and electrospun NaNbO 3 NFs be obtained. AsPDMS shown soft by the photo image in Figure the as3 in dimension. fabricated device was approximately 2.0 × 1.0 × 0.2 cm the curing process. Thereafter, the NFs could be easily transferred onto the flexible PDMS substrate
by lifting off the PDMS layer. After the IDE deposition, the flexible device based on the electrospun NaNbO3 NFs could be obtained. As shown by the photo image in Figure 3, the as-fabricated device was approximately 2.0 ˆ 1.0 ˆ 0.2 cm3 in dimension.
Sensors 2016, 16, 833
4 of 9
Sensors 2016, 16, 833
4 of 9
Sensors 2016, 16, 833
4 of 9
Sensors 2016, 16, 833
4 of 9
Figure 2. The X-ray diffractometer (XRD) pattern of the NaNbO3 NFs after being annealed at 700 °C
Figure 2. The X-ray diffractometer (XRD) pattern of the NaNbO3 NFs after being annealed at 700 ˝ C in oxygen. Figure in oxygen. 2. The X-ray diffractometer (XRD) pattern of the NaNbO3 NFs after being annealed at 700 °C in oxygen.
Figure 2. The X-ray diffractometer (XRD) pattern of the NaNbO3 NFs after being annealed at 700 °C in oxygen.
Figure 3. NaNbO3 NF-based active sensors fabrication procedure. Figure 3. NaNbO3 NF-based active sensors fabrication procedure.
The inset picture 4a illustrates the piezoelectric energy-harvesting performance-testing Figurein3.Figure NaNbO 3 NF-based active sensors fabrication procedure. method, in which the poled device is fixed at one end and forced by a rotating stick to bend. As shown The inset picture in Figure 4a illustrates the piezoelectric energy-harvesting performance-testing in Figure 4a, the output voltage generated by the device without NaNbO3 NFs was lower than 0.2 V, method, in which the poled device is fixed at one end and forced by a rotating stick to bend. As shown Thewhich insetwas picture in Figure illustrates theactive piezoelectric energy-harvesting performance-testing Figure 3.4a NaNbO 3 NF-based sensors fabrication procedure. too weak to be recognized from the strong background noise, and only 10% of the output in Figure 4a, the output voltage generated by the device without NaNbO3 NFs was lower than 0.2 V, voltage was the generated by the sensors with NaNbO 3 NFs (Figure 4b). The much lower voltage signal method,which in which poled device is fixed at one end and forced by a rotating stick to bend. As shown wasinset too picture weak toinbeFigure recognized from the background noise, and only 10% of the output The 4a illustrates thestrong piezoelectric energy-harvesting performance-testing generated from the blank sample may be by due to releasing stored remnant charges between the in Figure 4a, the output voltage generated the device without NaNbO NFs was lower than voltage generated by the sensors with at NaNbO 3 NFs (Figure 4b). The much lower voltage signal method,was in which thethe poled device is fixed one end forced by a rotating stick 3to bend. As shown electrodes, because thickness of the PDMS layerand changed during the bending process. These 0.2 V, which was4a, too weak tovoltage be recognized from the strong background noise, only 10% generated from the blank sample may beby due to releasing stored remnant chargesand between the of the in Figure thethe output generated without NaNbO 3 NFs was lower than 0.2 V, results suggest energy-harvesting behaviorthe of device the device should be attributed to the piezoelectric because the thickness of the PDMS layer changed during the bending process. These voltage output electrodes, voltage was generated by the sensors with NaNbO NFs (Figure 4b). The much lower 3 which was too weak to be recognized from the strong background noise, and only 10% of the output effect of the NaNbO3 NFs. results thethe energy-harvesting the device should beThe attributed to the piezoelectric voltagesuggest wasfrom generated by thesample sensorsbehavior with NaNbO 3 to NFs (Figure 4b). much lower voltage signal signal generated blank may be of due releasing stored remnant charges between the effect of thefrom NaNbO generated the3 NFs. blank sample may be due to releasing stored remnant charges between the
electrodes, because the thickness of the PDMS layer changed during the bending process. These results because the thickness of of thethe PDMS layer changed the bending process. Theseeffect of suggest electrodes, the energy-harvesting behavior device should be during attributed to the piezoelectric results suggest the energy-harvesting behavior of the device should be attributed to the piezoelectric the NaNbO3 NFs. effect of the NaNbO3 NFs.
Figure 4. The output voltage generated by the active sensor (a) without and (b) with the NaNbO3 NFs when the sensors were periodically bent by a rotating stick. Figure 4. The output voltage generated by the active sensor (a) without and (b) with the NaNbO3 NFs when the sensors were periodically bent by a rotating stick. Figure 4. The output voltage generated by the active sensor (a) without and (b) with the NaNbO3 NFs
Figure 4.when Thethe output voltage generatedbent by the sensor sensors were periodically by aactive rotating stick. (a) without and (b) with the NaNbO3 NFs when the sensors were periodically bent by a rotating stick.
Sensors 2016, 16, 833
5 of 9
Sensors52016, 16, 833 5 ofreversely 9 Figure shows the inversion of voltage polarity when the electrodes of the device were connected to the DAQ card. When it was connected to the DAQ card with forward direction, a positive Figure 5 shows the inversion of voltage polarity when the electrodes of the device were reversely voltage connected peak wastofirstly recorded with the device bending motion (Figure 5a). The positive polarity the DAQ card. When it was connected to the DAQ card with forward direction, a of the voltage peak can be attributed to the axial stretching of NaNbO under the positive bending state, 3 NFs positive voltage peak was firstly recorded with the device bending motion (Figure 5a). The Sensors 2016, 16, 833 5 of 9 which may generate the piezoelectric potential along the same direction of the poling electrical polarity of the voltage peak can be attributed to the axial stretching of NaNbO3 NFs under the field. bending state, which may generate piezoelectric potential along ofnegative the polingvoltage Once the connection wasthe reversed, the device bending resulted in the thesame generation of Figure 5 shows inversion ofthe voltage polarity when the electrodes of thedirection device were reversely electrical field. Once thecard. connection was reversed, device bending resulted in the of connected toThose the DAQ When it was connected topiezoelectric the DAQ card with forward direction, a peaks (Figure 5b). results also confirmed thatthe the effect of the generation NaNbO was 3 NFs negative voltage peaks (Figure 5b). Those results also confirmed that the piezoelectric effect of the positive voltage peak was firstly recorded with the device bending motion (Figure 5a). The positive the predominant mechanism of the energy-harvesting behavior [24].
NaNbO3 of NFs the predominant behavior polarity thewas voltage peak can bemechanism attributed oftothe theenergy-harvesting axial stretching of NaNbO[24]. 3 NFs under the bending state, which may generate the piezoelectric potential along the same direction of the poling electrical field. Once the connection was reversed, the device bending resulted in the generation of negative voltage peaks (Figure 5b). Those results also confirmed that the piezoelectric effect of the NaNbO3 NFs was the predominant mechanism of the energy-harvesting behavior [24].
Figure 5. The output voltage generated by the NaNbO3 sensor with the electrodes connected (a)
Figure 5.straight The output generated bysystem. the NaNbO3 sensor with the electrodes connected (a) straight and (b)voltage in reverse to the DAQ and (b) in reverse to the DAQ system. It is known that the sensitivity of sensor systems mainly depends on the signal-to-noise ratio (SNR) of the Because the high-frequency sampling and much higher impedance of the Figure 5. signals. Thethe output voltageofgenerated by the NaNbO 3 sensor with the electrodes (a) It is known that sensitivity of sensor systems mainly depends on theconnected signal-to-noise ratio NaNbO 3 NFs (up to TΩ) compared with the input impedance of the DAQ cards (~MΩ), the noise in straight and (b) in reverse to the DAQ system. (SNR) of the signals. Because of the high-frequency sampling and much higher impedance of the the recorded voltage signals was unstable and too high to recognize the relatively lower voltages NaNbOduring NFs (up to TΩ) compared with the input impedance of the DAQ cards (~MΩ), the noise in the sampling process [25]. of This limited the sensitivity of theon energy harvesters asratio the 3 It isthe known that the sensitivity sensor systems mainly depends the signal-to-noise piezoelectric active sensors, in which the generated was the sensitivity signal for thevoltages detection recorded voltage signals was unstable too highvoltage tosampling recognize relatively lower (SNR) of the signals. Because of theand high-frequency andthe much higher impedance of the during of humidity As a result, a charge wasof used for theDAQ signal conditioning, including NaNbO 3 NFschanges. (up[25]. to TΩ) compared with input impedance the cards (~MΩ), noise in the sampling process This limited thetheamplifier sensitivity theof energy harvesters asthe the piezoelectric the impedance matching and filtering. Figure 6 shows the to comparison of recorded the recorded voltage was unstable and toothe high recognize the relatively lowersignals voltages active sensors, in which thesignals generated voltage was sensitivity signal for thevoltage detection of by humidity the DAQthe cards betweenprocess the unconditioned and the conditioned signals. shown,harvesters the ratio between sampling [25]. This limited of theAsenergy as the changes.during As a result, a charge amplifier was used forthe thesensitivity signal conditioning, including the impedance the output voltage and noise amplitude was ~10 for the unconditioned signals and could be increased piezoelectric active sensors, in which the generated voltage was the sensitivity signal for the detection matching and filtering. Figure 6 shows the comparison of recorded voltage signals by the DAQ cards to humidity ~20 after signal conditioning. the voltage decreased due to the impedance of changes. As a result, Although a charge amplifier wasamplitude used for the signal conditioning, including betweenthe the unconditioned and the conditioned signals. As the ratio between matching, the increased could enhance sensitivity of theshown, sensors. Moreover, the stability of output impedance matchingSNR and filtering. Figurethe 6 shows the comparison of recorded voltage signalsthe by the voltage signal also improved duefor to filtering theconditioned high-frequency noise, which resulted inbetween a stable to ~20 voltage the and noise amplitude was ~10 the unconditioned signals and could increased DAQ cards between the unconditioned and the signals. As shown, thebe ratio baseoutput line signal andand improved thethe accuracy the voltage noise amplitude was of ~10the forsensors. the unconditioned and could be increased after signal conditioning. Although voltage amplitude decreasedsignals due to the impedance matching, to ~20 after Although the voltage amplitudeMoreover, decreased due the impedance the increased SNRsignal couldconditioning. enhance the sensitivity of the sensors. the to stability of the voltage matching, the increased SNR could enhance the sensitivity of the sensors. Moreover, the stability of signal also improved due to filtering the high-frequency noise, which resulted in a stable base line the voltage signal also improved due to filtering the high-frequency noise, which resulted in a stable signal and the improved accuracythe of the sensors. baseimproved line signal and accuracy of the sensors.
Figure 6. Comparison of recorded voltage signals generated by the NaNbO3 sensors with and without signal conditioning.
Figure 6. Comparison of recorded voltage signals generated by the NaNbO3 sensors with and without
Figure 6. Comparison of recorded voltage signals generated by the NaNbO3 sensors with and without signal conditioning. signal conditioning.
Sensors 2016, 16, 833
6 of 9
Figure 7 shows the piezoelectric active humidity response of the NaNbO3 sensors with the voltage signal conditioned by the charge amplifier. The sensing performance was tested by changing the humidity of the environment when the NaNbO3 sensors were continuously vibrating under the knocks of the rotating stick. As shown in Figure 7a–h, the sensor could generate distinguishable impulsive voltage with the humidity changing from 5% to 95% RH, where the peak-to-peak value of the voltage decreased from 0.4 to 0.06 V. Figure 7i shows the relationship between the voltage and the humidity of the environment. The voltage amplitude exhibited a negative linear correlation with the humidity varying from 5% to 80% RH and saturated when the humidity rose higher than 80% RH. As reported, the piezoelectric energy-harvesting behavior of the NaNbO3 NFs could be attributed to the electron flow in the external circuit under the driving force of the piezoelectric potential [26–28]. The electrons would accumulate at the positive side of the electrode due to the Schottky potential barrier between the Pt electrode and NaNbO3 NFs and compensate the piezoelectric potential. Once the piezoelectric potential was removed due to the release of strain, the accumulated electrons would flow back to the positive side along the external circuit and generate a negative voltage peak. As reported, the electrical field applied along the axial direction of the NaNbO3 NFs could lead to protons hopping among the surface-absorbed H3 O+ groups [22,29]. Therefore, an additional charge flow could be generated on the surface of the NaNbO3 NF when the piezoelectric potential was generated due to bending motion. This additional charge flow would lead to leakage current inside the piezoelectric materials and thus decrease the output voltage of the devices. When the environmental humidity changed, the concentration of protons on the surface of the NFs also changed. As a result, the output voltage changed due to the variation in leakage current. Moreover, the leakage current density in the NFs was in a linear correlation with the concentration of protons and absorbed water molecules [17]. Therefore, the output voltage exhibited a negative linear relationship with the environmental humidity at 5% to 80% RH. However, the absorption of water molecules was saturated when the humidity increased further (>80% RH), which resulted in the slight decrease in output voltage shown in Figure 7i. Finally, the sensitivity of the active sensor could be obtained as ~2 mV/%RH according to the linear result fitting within the humidity range of 5% to 80% RH. The relatively lower sensitivity of this NaNbO3 sensor compared to the reported CeO2 /ZnO and SnO2 /ZnO sensors should be attributed to the decreased output voltage amplitude after the signal conditioning. However, it is worth noting that the structure of the NaNbO3 nanofiber-based sensors was much more stable than the sensors based on the ZnO nanorod arrays, which was assembled through the stacking of the top electrode and the nanorod arrays. Moreover, the sensor also exhibited fast response speed to the variation in humidity. As shown in Figure 8a, the response time of the sensors for the humidity changing from 65% to 95% RH is ~12 s. Considering the time taken for the change in humidity, the real value of the response time should be shorter. Figure 8b shows the output voltage generated by the sensors when ethanol steam and H2 gas was introduced to and then removed from the testing chamber, respectively. The output voltage did not exhibit any change during the introducing and removing process of ethanol and H2 , which confirmed the excellent selectivity of the sensor. In order to evaluate the influence of temperature on the sensing result, the energy conversion property of the sensors under 38% RH was tested when the temperature increased from 24 to 80 ˝ C. As shown in Figure 8c, no obvious change in the voltage amplitude could be found, which suggested that the sensor possessed great temperature stability during the testing process.
did not exhibit any change during the introducing and removing process of ethanol and H2, which confirmed the excellent selectivity of the sensor. In order to evaluate the influence of temperature on the sensing result, the energy conversion property of the sensors under 38% RH was tested when the temperature increased from 24 to 80 °C. As shown in Figure 8c, no obvious change in the voltage amplitude could Sensors 2016, 16, 833 be found, which suggested that the sensor possessed great temperature stability 7 of 9 during the testing process.
Sensors 2016, 16, 833
7 of 9
Sensors 2016, 16, 833
7 of 9
Figure 7. (a–h) The output voltage generated by the active sensor at different humidity conditions; (i) Figure 7. (a–h) The output voltage generated by the active sensor at different humidity conditions; the relationship between the peak value of output voltage and the humidity. (i) the relationship the peak value ofby output voltage and at thedifferent humidity. Figure 7. (a–h) The between output voltage generated the active sensor humidity conditions; (i) the relationship between the peak value of output voltage and the humidity.
Figure 8. (a) The transient response of the sensor when the humidity increased from 65% to 95% RH. (b) The transient response of the sensor when ethanol steam and H2 was introduced and removed, Figure 8. (a) The transient response of the sensor when the humidity increased from 65% to 95% RH. respectively. (c) The output voltage of of the sensor when at 38%the RHhumidity under different temperature. Figure 8. (a) The transient response increased from 65% to 95% RH. (b) The transient response of the sensor when ethanol steam and H2 was introduced and removed, (b) The transient response of the sensor when ethanol steam and H2 was introduced and removed, respectively. (c) The output voltage of the sensor at 38% RH under different temperature. 4. Conclusions respectively. (c) The output voltage of the sensor at 38% RH under different temperature.
In this work, monoclinic NaNbO3 piezoelectric NFs with a uniform size distribution were 4. Conclusions 4. Conclusions synthesized using a far-field electrospinning method. By lifting off the PDMS layer coated on the NFs In this work, monoclinic NaNbO3 piezoelectric NFs with a uniform size distribution were and In thethis electrode flexible active humidity obtained. sensors could work, deposition, monoclinic aNaNbO NFssensor with awas uniform size The distribution were 3 piezoelectric synthesized using a far-field electrospinning method. By lifting off the PDMS layer coated on the NFs generate an impulsive output voltage with an amplitude up to 2 V during the vibration process. synthesized using a far-field electrospinning method. By lifting off the PDMS layer coated on The the and the electrode deposition, a flexible active humidity sensor was obtained. The sensors could energy-harvesting performance was confirmed to have originated the piezoelectric effect could of the NFs and the electrode deposition, a flexible active humidity sensorfrom was obtained. The sensors generate an impulsive output voltage with an amplitude up to 2 V during the vibration process. The NaNbO3 an NFs. By conditioning the voltage with a charge thethe noise in the recorded generate impulsive output voltage withsignal an amplitude up toamplifier, 2 V during vibration process. energy-harvesting performance was confirmed to have originated from the piezoelectric effect of the voltage signals was suppressed bywas half, which could be originated attributedfrom to matching impedance The energy-harvesting performance confirmed to have the piezoelectric effectand of NaNbO3 NFs. By conditioning the voltage signal with a charge amplifier, the noise in the recorded filtering high frequency signals. The output voltage of the NaNbO 3 sensors after the signal the NaNbO NFs. By conditioning the voltage signal with a charge amplifier, the noise in the recorded 3 voltage signals was suppressed by half, which could be attributed to matching impedance and conditioning a negative linear correlation with the environmental humidity the voltage signalsexhibited was suppressed by half, which could be attributed to matching impedance andwhen filtering filtering high frequency signals. The output voltage of the NaNbO3 sensors after the signal humidity changed fromThe 5% output to 80%voltage RH. The humidity-dependent output provides the high frequency signals. of the NaNbO3 sensors after thevoltage signal conditioning conditioning exhibited a negative linear correlation with the environmental humidity when the potential of the NaNbO3 device to be applied as an active humidity sensor with the sensitivity of 2 humidity changed from 5% to 80% RH. The humidity-dependent output voltage provides the mV/%RH. The variation in output voltages with humidity can be attributed to the humiditypotential of the NaNbO3 device to be applied as an active humidity sensor with the sensitivity of 2 dependent leakage current inside the NaNbO3 NFs, which was induced by protons hopping among mV/%RH. The variation in output voltages with humidity can be attributed to the humiditythe surface-absorbed H3O+ under the driving force of piezoelectric potential. The active sensor also dependent leakage current inside the NaNbO3 NFs, which was induced by protons hopping among exhibited a fast response speed, as well as great selectivity and temperature stability. The active +
Sensors 2016, 16, 833
8 of 9
exhibited a negative linear correlation with the environmental humidity when the humidity changed from 5% to 80% RH. The humidity-dependent output voltage provides the potential of the NaNbO3 device to be applied as an active humidity sensor with the sensitivity of 2 mV/%RH. The variation in output voltages with humidity can be attributed to the humidity-dependent leakage current inside the NaNbO3 NFs, which was induced by protons hopping among the surface-absorbed H3 O+ under the driving force of piezoelectric potential. The active sensor also exhibited a fast response speed, as well as great selectivity and temperature stability. The active humidity sensor based on NaNbO3 NFs provides an effective solution for self-powered sensor systems with high sensitivity, simple structure and fabrication processes, and low production costs. Acknowledgments: This work was supported by the National “973” Program of China (Grant No. 2014CB339803), the “863” Program of China (Project No. 2011AA010205), and the National Science Foundation of China (Grant No. 51502114). Author Contributions: Li Gu performed the materials synthesis, device fabrication, and sensor testing experiments. Di Zhou contributed reagents and materials. Juncheng Cao conceived and designed the experiments. Conflicts of Interest: The authors declare no conflict of interest.
Abbreviations The following abbreviations are used in this manuscript: SWCNTs NFs LED SNR XRD DAQ IDE
single-wall carbon nanotubes Nanofibers light emission diode signal-to-noise ratio X-ray diffraction data acquisition interdigital electrode
References 1. 2. 3. 4. 5. 6. 7.
8.
9. 10.
Blank, T.A.; Eksperiandova, L.P.; Belikov, K.N. Recent trends of ceramic humidity sensors development: A review. Sens. Actuators B Chem. 2016, 228, 416–442. [CrossRef] Fu, X.Q.; Wang, C.; Yu, H.C.; Wang, Y.G.; Wang, T.H. Fast humidity sensors based on CeO2 nanowires. Nanotechnology 2007, 18, 776–789. [CrossRef] Kuang, Q.; Lao, C.; Wang, Z.L.; Xie, Z.; Zheng, L. High-Sensitivity Humidity Sensor Based on a Single SnO2 Nanowire. J. Am. Chem. Soc. 2007, 129, 6070–6071. [CrossRef] [PubMed] Matko, V.; Donlagic, D. Sensor for high-air-humidity measurement. IEEE Trans. Instrum. Meas. 1996, 45, 561–563. [CrossRef] Matko, V. Next generation AT-cut quartz crystal sensing devices. Sensors 2011, 11, 4474–4482. [CrossRef] [PubMed] Matko, V.; Milanovi´c, M. Temperature-compensated capacitance-frequency converter with high resolution. Sens. Actuators A Phys. 2014, 220, 262–269. [CrossRef] Dai, C.-L.; Liu, M.-C.; Chen, F.-S.; Wu, C.-C.; Chang, M.-W. A nanowire WO3 humidity sensor integrated with micro-heater and inverting amplifier circuit on chip manufactured using CMOS-MEMS technique. Sens. Actuators B Chem. 2007, 123, 896–901. [CrossRef] Zhang, Y.; Fu, W.; Yang, H.; Li, M.; Li, Y.; Zhao, W.; Sun, P.; Yuan, M.; Ma, D.; Liu, B.; et al. A novel humidity sensor based on Na2 Ti3 O7 nanowires with rapid response-recovery. Sens. Actuators B Chem. 2008, 135, 317–321. [CrossRef] Zhang, Y.; Yu, K.; Jiang, D.; Zhu, Z.; Geng, H.; Luo, L. Zinc oxide nanorod and nanowire for humidity sensor. Appl. Surf. Sci. 2005, 242, 212–217. [CrossRef] Guo, H.; Chen, J.; Tian, L.; Leng, Q.; Xi, Y.; Hu, C. Airflow-induced triboelectric nanogenerator as a self-powered sensor for detecting humidity and airflow rate. ACS Appl. Mater. Interface 2014, 6, 17184–17189. [CrossRef] [PubMed]
Sensors 2016, 16, 833
11.
12. 13. 14.
15. 16.
17.
18.
19.
20.
21.
22. 23. 24. 25. 26. 27. 28. 29.
9 of 9
Hwang, G.T.; Park, H.; Lee, J.H.; Oh, S.; Park, K.I.; Byun, M.; Park, H.; Ahn, G.; Jeong, C.K.; No, K.; et al. Self-powered cardiac pacemaker enabled by flexible single crystalline PMN-PT piezoelectric energy harvester. Adv. Mater. 2014, 26, 4880–4887. [CrossRef] [PubMed] Wang, Z.L. Self-powered nanosensors and nanosystems. Adv. Mater. 2012, 24, 280–285. [CrossRef] [PubMed] Lee, M.; Bae, J.; Lee, J.; Lee, C.-S.; Hong, S.; Wang, Z.L. Self-powered environmental sensor system driven by nanogenerators. Energy Environ. Sci. 2011, 4. [CrossRef] Wang, Z.; Zhang, Y.; Yang, S.; Hu, Y.; Wang, S.; Gu, H.; Wang, Y.; Chan, H.L.W.; Wang, J. (K,Na)NbO3 Nanofiber-based Self-Powered Sensors for Accurate Detection of Dynamic Strain. ACS Appl. Mater. Interface 2015, 7, 4921–4927. [CrossRef] [PubMed] Xu, S.; Qin, Y.; Xu, C.; Wei, Y.; Yang, R.; Wang, Z.L. Self-powered nanowire devices. Nat. Nanotechnol. 2010, 5, 366–373. [CrossRef] [PubMed] Hsueh, H.-T.; Chen, Y.-H.; Lin, Y.-D.; Lai, K.-C.; Chen, J.-W.; Wu, C.-L. Integration of flower-like ZnO nanostructures with Crystalline-Si interdigitated back contact photovoltaic cell as a self-powered humidity sensor. Appl. Phys. Lett. 2013, 103. [CrossRef] Zhu, D.; Fu, Y.; Zang, W.; Zhao, Y.; Xing, L.; Xue, X. Piezo/active humidity sensing of CeO2 /ZnO and SnO2 /ZnO nanoarray nanogenerators with high response and large detecting range. Sens. Actuators B Chem. 2014, 205, 12–19. [CrossRef] Zhao, Y.; Lai, X.; Deng, P.; Nie, Y.; Zhang, Y.; Xing, L.; Xue, X. Pt/ZnO nanoarray nanogenerator as self-powered active gas sensor with linear ethanol sensing at room temperature. Nanotechnology 2014, 25, 115502. [CrossRef] [PubMed] Zang, W.; Nie, Y.; Zhu, D.; Deng, P.; Xing, L.; Xue, X. Core-Shell In2 O3 /ZnO Nanoarray Nanogenerator as a Self-Powered Active Gas Sensor with High H2 S Sensitivity and Selectivity at Room Temperature. J. Phys. Chem. C 2014, 118, 9209–9216. [CrossRef] Xing, L.; Hu, Y.; Wang, P.; Zhao, Y.; Nie, Y.; Deng, P.; Xue, X. Realizing room-temperature self-powered ethanol sensing of Au/ZnO nanowire arrays by coupling the piezotronics effect of ZnO and the catalysis of noble metal. Appl. Phys. Lett. 2014, 104. [CrossRef] Fu, Y.; Zang, W.; Wang, P.; Xing, L.; Xue, X.; Zhang, Y. Portable room-temperature self-powered/active H2 sensor driven by human motion through piezoelectric screening effect. Nano Energy 2014, 8, 34–43. [CrossRef] Zhang, Y.; Pan, X.; Wang, Z.; Hu, Y.; Zhou, X.; Hu, Z.; Gu, H. Fast and highly sensitive humidity sensors based on NaNbO3 nanofibers. RSC Adv. 2015, 5, 20453–20458. [CrossRef] Jung, J.H.; Lee, M.; Hong, J.I.; Ding, Y.; Chen, C.Y.; Chou, L.J.; Wang, Z.L. Lead-Free NaNbO3 Nanowires for a High Output Piezoelectric Nanogenerator. ACS Nano 2011, 5, 10041–10046. [CrossRef] [PubMed] Mantini, G.; Gao, Y.; D’Amico, A.; Falconi, C.; Wang, Z.L. Equilibrium piezoelectric potential distribution in a deformed ZnO nanowire. Nano Res. 2009, 2, 624–629. [CrossRef] Koszewnik, A.; Grzes, P.; Walendziuk, W. Mechanical and electrical impedance matching in a piezoelectric beam for Energy Harvesting. Eur. Phys. J. 2015, 224, 2719–2731. [CrossRef] Wang, Z.; Pan, X.; He, Y.; Hu, Y.; Gu, H.; Wang, Y. Piezoelectric nanowires in energy harvesting applications. Adv. Mater. Sci. Eng. 2015, 2015. [CrossRef] Falconi, C.; Mantini, G.; D’Amico, A.; Wang, Z.L. Studying piezoelectric nanowires and nanowalls for energy harvesting. Sens. Actuators B Chem. 2009, 139, 511–519. [CrossRef] Wang, Z.L.; Song, J. Piezoelectric nanogenerators based on zinc oxide nanowire arrays. Science 2006, 312, 242–246. [CrossRef] [PubMed] Liang, S.; He, X.; Wang, F.; Geng, W.; Fu, X.; Ren, J.; Jiang, X. Highly sensitive humidity sensors based on LiCl–Pebax 2533 composite nanofibers via electrospinning. Sens. Actuators B Chem. 2015, 208, 363–368. [CrossRef] © 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).
