1. INTRODUCTION

Guided-mode resonance (GMR) devices consist of grating and waveguide structures, where the grating and incident wave vectors are combined to match the propagation mode of the waveguide layer under the resonance condition. This resonance principle can be used to implement a narrowband transmission bandstop or reflection bandpass filters [1]. GMR devices are commonly used as biochemical sensors based on the detection of grating surface refractive index (RI) changes caused by biochemical reactions, because this resonance condition is very sensitive to the RI of the medium on the grating surface [2,3]. In addition, the resonance condition of a GMR device is very sensitive to its grating period. Therefore, when a GMR device is fabricated on a flexible substrate, it can also be used as a physical sensor based on grating period changes caused by pressure, strain, or torque [4–8]. In pressure sensing applications, the most commonly used material for the flexible substrate is poly-dimethylsiloxane (PDMS), because the sensors can be integrated into a PDMS-based microfluidic channel. In Ref. [4], a high-RI ${{\rm TiO}_2}$ film as the waveguide layer was directly sputter-coated on the replica grating surface of a PDMS substrate. They found that the difference in stiffness and the coefficient of thermal expansion between PDMS and ${{\rm TiO}_2}$ caused cracking damage on the ${{\rm TiO}_2}$ film; however, this PDMS-GMR device was demonstrated as a strain sensor based on the wavelength interrogation. Another study proposed a PDMS-GMR pressure sensor to embed a ${{\rm TiO}_2}$ grating inside the PDMS substrate to solve the cracking issue, and they achieved a limit of detection (LOD) of 0.26 mbar [5]. However, this requires a complicated fabrication process and a high-resolution spectrometer. In Ref. [6], instead of a coating film, ${{\rm TiO}_2}$ nanoparticles were deposited on the PDMS grating surface to act as the waveguide layer. The pressure changes caused a reflected light color change, and an LOD of 1.6 mbar was achieved. This requires expensive nanoparticle materials. In another approach, a photoresist was used as the waveguide layer on a PDMS substrate, and a chemical process was used to hydrophilize the substrate surface because PDMS is hydrophobic [7]. A color-tunable pressure sensor based on a two-dimensional inverse colloidal photonic crystal with PDMS was also reported to vary the pneumatic pressure in the microfluidic channel and detect its reflective color, depending on the bending of PDMS membranes [8]. In this nanostructure, no waveguide layer was deposited on the PDMS surface, and its LOD was 1.6 mbar. In addition, a non-optical micro-pressure sensor was proposed for microfluidic applications, and an LOD of 1.0 mbar was obtained [9]. This sensor is based on a special Ag/PDMS conducting composite, and the pressure changes the resistance of the composite. Another physical sensor, a gradient grating period GMR filter was fabricated on a flexible polyethylene terephthalate (PET) substrate to fabricate a lightweight torque sensor [10]. This PET-GMR device also had a sputtered-${{\rm TiO}_2}$ waveguide layer. Recently, high-RI core and low-RI clad polymers were combined to form a GMR structure on a PET foil to form an all-polymer strain sensor [11].

In this study, we investigated the fabrication and measurement of a non-PDMS-based GMR pressure sensor. It uses a highly flexible low-density polyethylene (LDPE) film as its substrate. The LDPE film used here is much more flexible than that used in previous studies, and relatively better pressure detection sensitivity can be obtained using this film. The LDPE is a low-cost material that is not hydrophobic like PDMS; and it is easy to attach to a microfluidic channel and is very suitable for disposable device applications. The LDPE film is commonly used for packing purposes and can be fabricated via an extrusion blown film process. An LDPE film has good strength and clarity when its thickness is approximately 80 µm [12]. Our experimental results show that the ${{\rm TiO}_2}$ layer can be directly coated on the LDPE film substrate and is suitable for the fabrication of GMR pressure sensors. Furthermore, instead of wavelength interrogation by using a spectrometer, the transmitted power change caused by the resonance angle shift of the GMR sensor was used to measure the pressure. This intensity detection method can achieve a faster measurement than wavelength interrogation, and a pulse pressure measurement is demonstrated.

2. DEVICE FABRICATION AND SIMULATION

The fabrication of the proposed LDPE-GMR pressure sensor is based on a simple thermal- nanoimprinting technology [13]. In our process, a metal grating mold (nickel), which was a replica of a holographic grating mold, was obtained using the electroforming method. A schematic of the fabrication process is shown in Fig. 1. The metal mold was placed on the heating plate of a nanoimprinting machine (NIL-AH1, AHEAD Optoelectronics), and the LDPE film substrate was placed on the metal mold. Thereafter, the quartz plate of the nanoimprinting machine was pressed onto the LDPE substrate under a pressure weight of 7 kg. The temperature of the metal plate was gradually increased from 25°C (room temperature) to 90°C within 30 min and maintained for 5 min to form the grating on the LDPE film. The heating power was then turned off, and the heating plate was naturally cooled to 25°C. The quartz plate was then elevated, and the LDPE film was easily peeled off the metal mold. Next, the LDPE film with the grating was attached to a glass plate using polyvinyl alcohol (PVA) as the adhesive and sacrificial layer. Photographs and atomic force microscopy (AFM) images of the LDPE film with grating before coating the ${{\rm TiO}_2}$ layer are shown in Figs. 2(a) and 2(b), respectively. AFM measurements showed that the average values of the grating period and the modulation depth were 416 and 90 nm, respectively.

 

Fig. 1. Fabrication process of the LDPE GMR sensor.

Download Full Size | PDF

 

Fig. 2. LDPE-GMR device. (a) and (b)/(c) and (d): photograph and AFM image before/after ${{\rm TiO}_2}$ coating.

Download Full Size | PDF

Subsequently, the LDPE film with a grating attached to the glass plate was placed in a sputtering machine to coat the ${{\rm TiO}_2}$ waveguide layer. The sputter coating process was conducted at a low power (${\lt}{100}\;{\rm W}$) to produce a low deposition rate (${\lt}{4}\;{\rm nm/min}$) and to avoid increasing the temperature of the LDPE film very much. As the thermal expansion coefficients of LDPE and ${{\rm TiO}_2}$ are different, a large temperature difference during the coating process may cause serious crack damage in the ${{\rm TiO}_2}$ layer. Finally, a ${{\rm TiO}_2}$ waveguide layer with a thickness of approximately 250 nm was formed on the grating surface. Photographs and AFM images of the LDPE-GMR device after ${{\rm TiO}_2}$ coating are shown in Figs. 2(c) and 2(d), respectively. The AFM results were obtained before removing the PVA and showed that, compared with the previous results, the average values of the grating period and the modulation depth were increased to 422 and reduced to 55 nm, respectively, and these average values were used in our simulation. Although the sine grating profile after coating was inferior to that before coating, the structure was acceptable and exhibited the resonance phenomenon in our measurement. Subsequently, the ${{\rm TiO}_2}$-coated LDPE-GMR film sensor on the glass plate was immersed in the water for 2–3 min to dissolve PVA and release the film sensor, as shown in Fig. 2(c). Curvatures around the edges were observed on the film sensor, because the edge of the membrane that adhered to the glass partially peeled during the coating process, resulting in internal residual stress. However, most of the area near central was good and could be used for pressure sensing.

The schematic and simulation results of the fabricated LDPE-GMR device are shown in Figs. 3(a) and 3(b), respectively. The simulation was based on the rigorous coupled-wave analysis method. The parameters of the device were as follows: grating period $\Lambda = {422\! -\! 428}\;{\rm nm},$ modulation depth $d = {55}\;{\rm nm}$; ${{\rm TiO}_2}$ layer thickness $h = {250}$; RI of the LDPE film ${n_S} = {1.55}\;{\rm RIU}$; RI of the waveguide (${{\rm TiO}_2}$) ${n_W} = {2.29}\;{\rm RIU}$; RI of air ${n_A} = {1.0}\;{\rm RIU}$, and incident wavelength $\lambda = {632.8}\;{\rm nm}$. The simulation results show that increasing the grating period Λ from 422 to 428 nm causes the resonance angles to increase linearly from 18.82° to 19.65°. Therefore, based on these results, the average rate of increase in the resonant angle caused by the increasing grating period is approximately 0.12°/nm, as shown in the inset of Fig. 3(b).

 

Fig. 3. (a) Schematic of the LDPE GMR pressure sensor. (b) Simulation results of the resonant angle for the increasing grating period from 422 to 428 nm.

Download Full Size | PDF

3. MEASUREMENT SETUP AND RESULTS

Figure 4(a) shows a schematic of the LDPE-GMR pressure sensor. The LDPE-GMR film was sandwiched between two rubber sealing rings with radii of 1.85 cm. These two rubber rings were squeezed by two glass substrates using binder clips; their photographs are shown in Fig. 4(b). Two chambers separated using the LDPE-GMR sensing film were formed. In the upper chamber, a few punching holes were created on the rubber sealing ring to maintain a pressure of 1 atm. In the lower chamber, two syringe needles were punctured through the rubber ring to control the pneumatic pressure inside the chamber by injecting (the inlet) and releasing (the outlet) air. When the pressure inside the lower chamber was higher than 1 atm, the LDPE-GMR film bent upward as a parabolic profile. Consequently, the grating period increased as schematically shown in Fig. 4(a), and the corresponding resonance angle increased, as shown in the simulation results in Fig. 3(b). During measurement, the pressure sensing module was mounted on a motorized rotation stage to scan the incident angle θ, as shown in Fig. 5. A HeNe laser beam was incident on the center of the LDPE-GMR film to avoid a change in the angle of incidence due to the curvature of the membrane and to ensure that the resonance angle shift was caused by the increase in the grating period. A polarizer was used to generate a TM-polarized incident wave, and a homemade micropump was used to inject air into the chamber. The micropump was realized using an actuator to press a soft rubber ball, and the volume of the injection air could be controlled by the displacement of the actuator. The chamber pressure was monitored using a manometer (GM505, BENETECH) with a measurement range of $\pm {24.90}\;{\rm mbar}$ and a resolution of $\pm {0.01}\;{\rm mbar}$. Silicone tubes were used to connect the micropump to the chamber and manometer. The power of the transmitted light was detected using a power meter, and all the data were recorded using a computer.

 

Fig. 4. (a) Schematic and (b) photograph of the LDPE-GMR pressure sensor testing module.

Download Full Size | PDF

 

Fig. 5. Schematic of the measurement setup.

Download Full Size | PDF

In the first experiment, the transmitted power of the LDPE-GMR sensor was investigated by scanning the incident angle for different pressures. In this experiment, the outlet (releasing) needle was stuck; therefore, the injected air was used to maintain the pressure inside the chamber. Consequently, the LDPE-GMR film remained laterally strained. The transmitted powers versus the incident angles for the three pressures are shown in Fig. 6. The black line indicates the absence of air injection, and the red and blue lines represent two pneumatic pressures, 10 and 24 mbar, respectively, inside the chamber. These measured cures have broader resonance widths than the simulation results. This may be induced by the nonuniformity in the thickness of the waveguide and the grating period after the ${{\rm TiO}_2}$ coating. The resonance angles (transmission dip) shifted from 18.70° to 18.83° and 19.08°, as shown in the figure. The pressure increases the grating period, producing the resonance angle shift to the right, which agrees with the simulation results. We can estimate that the grating period may have been strained from 422.0 to 423.1 and 425.2 nm, respectively.

 

Fig. 6. Resonance angle curves of three pressures of 0, 10, and 24 mbar.

Download Full Size | PDF

Next, the pulse pressure measurement was demonstrated. The incident angle was fixed at 17°, at which the slope of the transmitted power curve, as shown in Fig. 6, was the most abrupt, and we could obtain a large transmitted optical power change caused by different pressures. In this experiment, the outlet (releasing) needle was not stuck, which was different from the first measurement setup; therefore, the chamber pressure increased to a maximum during the injection of air and then decreased back to 1 atm because the outlet released the air. Five different volumes of air were sequentially injected into the air chamber within the same time interval of approximately 1 s. Five peak pressure values were recorded by the manometer. The time-evaluation results for the transmitted power changes are shown in Fig. 7(a). The five power peaks corresponded to five pulse pressures: 3.6, 5.3, 12.6, 15.0, and 18.1 mbar. The relationship between the transmission power change and pulse pressure is shown in Fig. 7(b), and the linear fitting line shows a slope of 8.27 µW/mbar. Therefore, if the power meter noise can be controlled to be less than 0.1 µW, this LDPE-GMR pressure sensor can achieve an LOD of 0.012 mbar. The performance of the proposed sensor was compared with other relevant sensors in the references, as listed in Table 1. In practical use, a UV light may degrade the LDPE film; therefore, a transparent cover with a UV cutoff filter can be used to extend good performance stability.

 

Fig. 7. (a) Time evaluation of transmission power change for five pulse pressures, 3.6, 5.3, 12.6, 15.0, and 18.1 mbar. (b) Relationship between the transmission power change and pulse pressure.

Download Full Size | PDF

Table 1. Comparison between Our LDPE-GMR Pressure Sensor and Other Relevant Sensors

View Table

4. CONCLUSION

In this study, we successfully fabricated a GMR pressure sensor based on a flexible LDPE film. The proposed fabrication method and material of the sensor are low in cost, while the sensitivity and LOD are acceptable. The transmitted power changes of the LDPE-GMR sensor were proportional to the pressure values, and fast pulse pressure measurement was achieved. The proposed LDPE-GMR sensor has significant potential for pressure-measurement applications in many instruments and microfluidic devices.