1. Introduction

Up to date, multifunctional electronic sensors have been investigated by using a variety of resistive or capacitive strategies due to their excellent mechanical flexibility and electrical sensing properties [1–3]. For example, stretchable conductors made of silver nanowires are demonstrated for strain and pressure sensing based on capacitance [4]. Polyurethane fibers assembling ZnO nanowires are used to detect strain, temperature, and ultraviolet light by resistance changes [5]. These electronic devices exhibit the high performance and multiple sensing capabilities, but they fail to distinguish different sensing signals simultaneously. Despite Integrating multiple sensing elements into independent electrical signals can obtain multi parametric sensors with simultaneous sensitivity readout and low cross sensitivity, it increases the complexity and the cost of manufacturing process. In addition, electronic sensors are susceptible to electromagnetic interference (EMI) and are affected by electrical safety issues [2].

Compared with electronic sensors, functionalized fiber sensors offer attractive characteristics [6–8], such as light weight, compact size, EMI immunity, and safety. However, conventional optical fiber sensors made of silicon or glass are highly stiff and rigid, which make them be fragile and not prone to large curvature bending [9–11]. To overcome these limitations, flexible polymer optical fiber sensors are used to achieve a wide range of curvature or strain sensing through the deformation of sensors [12–14]. For example, Guo’s group demonstrated a dye-doped PDMS optical fibers sensor for highly wearable strain sensing which can detect strains over a large dynamic range of 100% with a precision of +/-0.91% [15]. Despite polymer optical fiber sensor have made great progress recently, the optical fiber sensor based on polymer doped by UCL materials for simultaneously temperature and curvature sensing have not been achieved yet.

In this work, we demonstrated a soft and transparent optical fiber sensor based UCL which can synchronously detect the temperature and the curvature. The sensor is made by thermally curing polymer materials PDMS doped with UCL materials NaYF₄: Yb³⁺/Er³⁺. Under the excitation of a 980 nm laser, the sensor generated the dual green UCL. The FIR technology was used to temperature measurements. In the lower temperature region, the fluorescence intensity corresponding to 525 nm green peak which is used to measure the curvature based on the principle of the emission intensity demodulation almost did not change with temperature. Furthermore, the curvature changes of the sensor only change the emission intensity value without affecting the FIR. Hence, deformation produced by the sensor through bending results in detectable and reversible changes in its reflected light, allowing the curvature to be simultaneously measured. It has great application prospect in biological health monitoring and flow rate monitoring because the sensor can realize temperature and strain sensing simultaneously and is not sensitive to different pH values.

2. Fabrication and principle

Figure 1(a) shows the fabrication processes of the hydrogel optical fiber sensor, which was fabricated based on soft and transparent PDMS materials doped with NaYF₄: Yb³⁺/Er³⁺ nanomaterials. Firstly, to fabricate these sensors, molds with appropriate sizes are manufactured. Secondly, two bundled multimode optical fibers (core/cladding, 105/125 µm) were fixed the mold and aligned to the center of the mold. Thirdly, the hydrogel precursor containing PDMS materials mixed at base to curing agent ratio of 10:1 and NaYF₄: Yb³⁺/Er³⁺ materials is injected into the mold, which is heated at 70 °C for 60 min subsequently. Finally, the sensor after thermal curing was peeled off from the mold.

 

Fig. 1. (a) The fabrication process of the hydrogel optical fiber sensor. (b) The real picture of the sensor probe. (c) The state of the sensor probe at a torsion angle greater than 90°. (d) The state of the sensor probe when the bending angle is greater than 90°. (e) The transmission electron microscopy (TEM) image of NaYF₄: Yb³⁺/Er³⁺ nanomaterials.

Download Full Size | PDF

The size of the sensor was 1 mm (thickness) × 5 mm (width) × 20 mm (length). By varying the size of rectangular molds, sensors with various sizes can be fabricated. For this sensor, an excessively long insertion depth results in a reduction in the area available for curvature sensing. Besides, if the insertion length is too short, the fiber will not be firmly connected with the probe, and the structure will be easily damaged. Therefore, we choose 5 mm insertion length. The bright field picture of the sensor probe is shown in Fig. 1(b). It can be seen that the sensor probe is transparent and colorless, which is attributed to the ratio of mixed PDMS materials and NaYF₄: Yb³⁺/Er³⁺ nanomaterials with the nanoscale particle size. As shown in Fig. 1(e), the TEM image indicates that the average particle size of NaYF₄: Yb³⁺/Er³⁺ nanomaterials is approximately 20 nm. The sensor can be freely twisted up to 90° as shown in Fig. 1(c). Moreover, no change occurs to the sensor when the sensor is repeatedly bent over 90° as shown in Fig. 1(d). These results show that the sensor has excellent deformation properties. Therefore, it can be used to detect reversibly the curvature or strains at a large dynamic range.

The experimental setup of the fabricated sensor is shown in Fig. 2(a). A 980 nm fiber coupled laser which is used as the excitation source was coupled into the MMF. Then excitation light transmits to the sensor fabricated by PDMS and NaYF₄: Yb³⁺/Er³⁺ nanomaterials through the MMF, and the excitation light induces NaYF₄: Yb³⁺/Er³⁺ nanomaterials to generate fluorescence. The divergent fluorescence couples into the another MMF and transmits to the spectrometer (USB2000+, Ocean Optics). Under the 980 nm excitation with the out power of 209.9 mW (The output power density is 2.425×10⁴ mW/mm²), the upconversion emission spectrum of NaYF₄: Yb³⁺/Er³⁺ nanomaterials at room temperature (20.0 °C) is shown in Fig. 2(b). As shown in Fig. 2(b), the UCNPs exhibited two distinct emission bands centered at 525 nm and 545 nm in the wavelength from 510 nm to 580 nm, corresponding to the ²H_11/2 → ⁴I_15/2, ⁴S_3/2 → ⁴I_15/2 transitions, respectively. Since two energy bands corresponding to green emission peaks are thermally coupled, the spectral features of emission peaks centered at 525 nm and 545 nm are used to detect the temperature based on the FIR technique.

 

Fig. 2. (a) The experimental setup of the fabricated sensor. (b) Green emission spectrum of the achieved NaYF₄: Yb³⁺/Er³⁺ nanomaterials under 980 nm excitation. (c) The upconversion process of the synthesized NaYF₄: Yb³⁺/Er³⁺ nanomaterials.

Download Full Size | PDF

In order to further explain the upconversion of the synthesized NaYF₄: Yb³⁺/Er³⁺ nanomaterials, the energy transition mechanism of Yb³⁺/Er³⁺ pairs is built as shown in Fig. 2(c). Under the excitation of a 980 nm laser, the populations in ²F_7/2 state of Yb³⁺ ions are excited to their ²F_5/2 excited state. Then, the energy transfer (ET) process ²F_5/2 (Yb³⁺) + ⁴I_15/2 (Er³⁺) →⁴I_11/2 (Er³⁺) + ²F_7/2 (Yb³⁺) (ET1) easily takes place, populating the ⁴I_11/2 excited state of the Er³⁺ ions. And then by energy transfer process ²F_5/2 (Yb³⁺) + ⁴I_15/2 (Er³⁺) → ⁴F_7/2 (Er³⁺) + ²F_7/2 (Yb³⁺) (ET2) from the Yb³⁺ ions, the ²F_7/2 state of Er³⁺ is populated. Next, the populations in the ²F_7/2 state relax non-radiatively to the ²H_11/2 and ⁴S_3/2 state. Finally, the green emissions at 525 nm and 545 nm are achieved through the ²H_11/2 → ⁴I_15/2 and ⁴S_3/2 → ⁴I_15/2 transitions, respectively.

3. Experiment and results

In order to research the temperature sensing performance of the fabricated sensor, the sensor was placed in an oven that was heated from 303 K (30 °C) to 423 K (150 °C) with the temperature interval of 10 K (10 °C). The target temperature was set through the oven, and the actual temperature was shown on the display panel of the oven, as shown in Fig. 3(a). The fluorescence spectra were monitored in real time by using the spectrometer (USB2000+, Ocean Optics). At each temperature, the luminescence spectra were recorded after the temperature had been steady for 5 minutes for keeping the temperature inside the oven stable and balanced. Under the 980 nm excitation with the out power of 209.9 mW (The output power density is 2.425×10⁴ mW/mm²), the measured luminescence spectra from 303 K to 423 K were shown in Fig. 3(b). As can be seen from Fig. 3(b), I₅₄₅ (the fluorescence intensity of 545 nm emission peak) decreases greatly with the increasing of the temperature at the temperature range of 303 K to 423 K, while I₅₂₅ (the fluorescence intensity of 525 nm emission peak) changes little. Especially at the temperature range of 303 K to 363 K, I₅₂₅ is almost unchanged. In addition, the wavelength of luminescence peaks does not change with temperatures changing. Figure 3(c) shows that I₅₂₅/I₅₄₅ (the fluorescence intensity ratio) increases monotonously with the increasing of temperatures, which can be described by Boltzmann distribution [16,17]. Thus, FIR can be express as [16]

 

Fig. 3. (a) Optical setup for temperature sensing of the sensor. (b) Fluorescence spectra of the sensor at various temperatures. (c) Dependence of I₅₂₅/I₅₄₅ on temperatures. (d) Linear plot of ln(I₅₂₅/I₅₄₅) versus the inverse temperatures. (e) I₅₂₅/I₅₄₅ response to five cycles of alternating temperatures between 293 K (20 °C) and 333 K (60 °C).

Download Full Size | PDF

It can be seen from Fig. 3(d), there is a linear behavior between ln(I₅₂₅/I₅₄₅) and 1/T in accordance with Eq. (2), showing the sensor has the temperature sensitivity of 714.82 K^-1 with excellent linearity (R²= 0.997) at a temperature range of 303 to 423 K. To investigate the repeatability and the stability of the sensor for temperature sensing, the temperature cycle experiment was conducted. In this temperature cycle experiment, alternating temperatures of the oven were set as 293 K (20 °C) and 333 K (60 °C), respectively. Figure 3(e) showed I₅₂₅/I₅₄₅ response to five cycles of alternating temperatures. As can be seen from Fig. 3(e), maximum deviations which are defined as (R_max-R_min)/2R (R_max is the maximum of I₅₂₅/I₅₄₅, R_min is the minimum of I₅₂₅/I₅₄₅ and R is the average of I₅₂₅/I₅₄₅) of I₅₂₅/I₅₄₅ at 293 K and 333 K are 0.84% and 0.24%, respectively. Results validated that the sensor has the capability of reversible thermal response and an excellent repeatability even though it experiences such temperature mutation.

In order to further study the relationship between the fluorescence intensity corresponding to the sensor’s 525 nm green peak and temperatures, spectra were obtained at the temperature range from 303 to 363 K with the temperature interval of 10 K, which is shown in Fig. 4(c). As shown in the inset of Fig. 4(c), the maximum deviation which is defined by (I_max-I_min)/2I (I_max, I_min and I are the maximum, the minimum and the average of the 525 nm green emission, respectively) of the emission intensity at 303 to 363 K is 2.89%. While that of the emission intensity under 303 to 333 K is 1.29%. These results indicate that the emission intensity of the 525 nm luminescence peak can be ignored in the temperature range of 303 to 363 K, especially in the temperature range of 303-333 K, which can be used for curvature sensing based on emission intensity demodulation principle due to the fact that the temperature only changes the FIR and does not affect the luminescence intensity. In addition, in this temperature range, the sensor is quite suitable for the monitoring of human physiological activities such as respiration and the body temperature.

 

Fig. 4. (a) Optical setup for curvature sensing of the sensor. (b) Schematic diagram of the sensor for curvature measurements. (c) The 525 nm luminescence peak of the sensor at 303-363 K, the inset shows the variation of emission intensity with the increasing of temperatures. (d) Fluorescence spectra of the sensor at various curvatures. (e) Emission intensities at 525 nm and 545 nm as a function of curvatures. (f) Dependence of I₅₂₅/I₅₄₅ on curvatures.

Download Full Size | PDF

To test the curvature sensing performance of the sensor, we set up the experimental device in Fig. 4(a). The equipment of Fig. 4(a) can be used for the detection of bending curvature through bending the sensor. The bending of the sensor is realized by changing the length in the horizontal direction with an interval of 200 µm. Since optical fiber is inserted into the PDMS material about 5 mm, and the fluorescence band is assumed to be uniform green band, as shown in Fig. 4(b). In order to simplify the calculation process, we build the model in Fig. 4(b) to transform the horizontal length change of the sensor into the curvature change of the sensor. Upon bending, the changes in bending curvature of the sensor can be obtained from the geometric structures in Fig. 4(b), given by

Figure 4(f) shows dependence of I₅₂₅/I₅₄₅ on curvatures. In the process of bending curvature sensing experiment, I₅₂₅/I₅₄₅ is between 0.45573 and 0.47992. And the maximum deviation is 2.58%, indicating that the curvature response is almost independent of I₅₂₅/I₅₄₅.

In order to study the repeatability of the sensor for bending curvature sensing, the bending-recovering cycles experiment was conducted. Figure 5(a) shows emission intensities at 525 nm as a function of curvatures in four cycles of alternating curvatures between 0.192 and 0.213 mm^-1, where the sensor exhibited reversible emission intensity changes with a negligible deviation during the bending-recovering cycles. Furthermore, I₅₂₅/I₅₄₅ is almost the same in two cycles of alternating curvatures, which is shown in Fig. 5(b). The above experimental results show that the sensor has an excellent repeatability for curvature sensing, which can be used in a variety of applications. In addition, for curvature measurement, the laser power directly changes the fluorescence intensity, affecting the accuracy of measurement results. Therefore, in order to enable the laser to output a specific power value, we choose the laser with selectable power value and the power range can be stabilized by automatic circuit feedback.

 

Fig. 5. (a) Emission intensities at 525 nm response to four cycles of alternating curvatures between 0.192 and 0.213 mm^-1. (b) Dependence of I₅₂₅/I₅₄₅ on curvatures for two cycles.

Download Full Size | PDF

In practical application, pH in the environment will affect the accuracy of the parameter sensing, so it is of great significance to study the spectral stability of the sensor in different pH solutions. First, the sensor was placed in different pH solutions, such as deionized water, NaOH solutions and H₂SO₄ solutions, and the temperature of these solutions was kept at 303 K. Testing the luminescence spectra for three times in each solution, and wiping the sensor with an ethanol solution to avoid the effects of the previous testing solution before each testing. As shown in Fig. 6, the luminescence wavelengths and peak intensities of the luminescence spectra in solutions with different pH levels are almost the same, indicating that the sensor can maintain a particularly stable performance in different pH environments. Therefore, the sensor can show the excellent performance for temperature and strain sensing simultaneously in different pH environments. It has great application prospect in the monitoring of biological health and flow rate, providing an optical alternative for multi-parameters monitoring in future.

 

Fig. 6. Emissions spectra in solutions with different pH levels, where the sensor was kept at a constant temperature of 303 K.

Download Full Size | PDF

4. Conclusions

In summary, we demonstrated an optical fiber sensor based UCL for synchronously temperature and curvature sensing. Temperature measurements were only achieved through the FIR of green fluorescence intensities of 525 nm and 545 nm emission peaks, providing the temperature sensitivity of 714.82 K^-1 with excellent linearity of 0.997 at a temperature range of 303 to 423 K. Especially, in the low temperature range (303-363 K), the temperature measurement is almost irrelevant to the green UCL emission centered at 525 nm. Thus, the sensor produces detectable and reversible changes in the reflected fluorescence through the deformation produced by bending, allowing measurement of the bending curvature. Experimental results show that the sensor has excellent linearity (R²= 0.998) and repeatability for curvature sensing. In addition, the influence of the bending curvature on the FIR can be ignored. It has great application prospect in biological health monitoring and flow rate monitoring because the sensor can realize temperature and strain sensing simultaneously and is not sensitive to different pH values, providing an optical alternative for multi-parameters monitoring in the future.