A multilayer-based fiber optic sensor enabling simultaneous measurement of humidity and temperature is proposed and demonstrated. The sensitive elements were multilayer coatings consisting of nano-porous TiO2 and SiO2 films, which were deposited on fiber end-face to form a Fabry-Perot (F-P) filter structure. Relative-humidity (RH) sensing is correlated with the shift of interference fringe due to the change of effective refractive index of porous coatings when exposed to different RH environments. The sensor is sealed in a glass tube in case of temperature measurement. Experimental results show that the average sensitivity are 0.43nm/%RH and 0.63nm/°C respectively when environmental RH changes from 1.8%RH to 74.7%RH and temperature changes from 21.4°C to 38.8°C. The proposed sensors present high repeatability, and especially highly sensitive to lower moisture measure.
© 2014 Optical Society of America
Humidity measurement and control is of great importance in many application fields such as meteorological services, biomedical device, food and electronic processing, chemical gas purification. In the past decades, there were extensive investigations on humidity sensing techniques including chilled mirror hygrometer, wet and dry psychrometer, electronic humidity sensor and optical sensor. Conventional humidity sensors have many limitations such as low sensitivity, relatively bulky, while a powered active device is required in case of electronic sensor. In many cases, humidity parameter is usually associated with others, such as temperature and pressure. When environmental temperature is changed, the measured humidity will also change. Therefore simultaneous measurement on humidity and temperature is necessary. Conventional humidity or temperature sensor can only measure one parameter, which is not acceptable for some applications.
Fiber-optic sensors have many distinctive characteristics such as small size, flexibility and immunity to electromagnetic interference [1–3]. There already exists fiber-optic temperature sensor, for example fiber Bragg grating (FBG), which enables temperature measurement with sensitivity of 0.1°C. Also, several optical fiber RH sensors based on different configurations for measuring humidity have been reported, such as long period gratings (LPGs) , tilted fiber Bragg’s grating (TFBG) , U-bend , hetero-core optical fiber [7, 8]. The principle of these humidity sensors rely on the use of the moisture sensitive material to generate secondary effects such as refractive index change or strain on the sensing fiber that result in the shift of output spectra or change in intensity. However, these sensors can only measure temperature or humidity separately.
In this paper, a fiber-optic sensor enabling simultaneous measurement on humidity and temperature is proposed and demonstrated. The sensing probes are porous oxide multilayer deposited on fiber tip, realized by e-beam evaporation without ion-source assistance, forms thin film Fabry–Perot interferometer (FPI). FPI sensors are extremely sensitive to perturbations that affect the optical path difference between two reflective mirrors and the sensing region can be very compact to ensure miniature size of the sensor. There are two sensing probes in the system; one is sensitive to environmental humidity change, while the other is sealed in a glass tube for temperature measurement.
2. Sensing mechanism and theoretical simulation
The relative humidity of an air-water mixture is defined as the ratio of the partial pressure of water vapor (H2O) in the mixture to the saturated vapor pressure of water at a given temperature. Here the saturated vapor pressure of water is correlated with temperature, the correlation presented by Buck  is commonly encountered in the literature and provides a reasonable balance between complexity and accuracy.
The miniature optical fiber humidity sensing probe is consisted of three-layer optical thin films deposited on a multimode fiber (MMF: 62.5nm/125nm) tip as shown in Fig. 1. The first and third layers are TiO2 coatings, which are employed as mirror layers. A very thick SiO2 coating is used as cavity layer in the Fabry-Perot structure. The coatings are deposited by evaporation technology, while porous micro-structure in the coating is realized. The sensing mechanism is based on the change of reflected interference spectrum when water molecules are absorbed in the porous oxide coating, which means the change of effective refractive index, as result this will shift the interference fringe. Therefore the drift of FPI fringe is correlated with the variation of humidity level.
The dielectric thin films manufactured by e-beam evaporation without ion-source assistance have columnar and porous structures . When the TiO2 and SiO2 porous coatings absorb water molecules from environment as the influence of capillary condensation [11–13], their effective refractive index will change. In this way the correlation of humidity level with effective refractive index of the films are maintained. Variations in the refractive index of dielectric films would affect the propagation of light in the element, which in turn leads to a phase change. Figure 2(a) shows the theoretical simulation of interference fringe shift of the proposed three-layer F-P structure when the effective refractive index have 1% and 2% increase. The third interference fringe dips are located at 526.6nm, 530.9nm and 535.4nm respectively in case of original fringe, 1% and 2% of effective refractive index increase. The means 4.3nm and 8.8nm of characteristic wavelength drift exist when effective refractive index increase 1% and 2% when compared to the original fringe.
As mentioned above, relative humidity in a sealed system is varying with the change of environmental temperature. This can be employed for the principle of temperature sensing. Figure 2(b) plots the change of relative humidity in a sealed system with changing temperature under different original levels of relative humidity. It can be concluded that relative humidity inside the sealed Fabry-Perot sensor decreases with the increase of surrounding temperature. As for the temperature measurement in the work, the sensing probe with similar porous thin film coating is sealed in a glass tube, and therefore is isolated with environmental humidity change. However, when environmental temperature changes, as result this will change the relative humidity in the sealed tube. In this way the interference fringe drift is correlated with relative humidity change in the sealed metal tube, and also correlated to the environmental temperature change.
A three-layer FPI structure was realized on a multimode fiber tip by e-beam evaporation, it is possible to control the structural and morphological properties of the deposited thin film by optimization of deposition process parameters [14, 15]. In the experiments, the basic vacuum pressure of coating chamber is set at 0.01Pa, oxygen (O2) with a velocity of 100sccm is supplied as procedure gas, and the fiber sample baking temperature is set at 100°C. The first and third layer are 168.55 nm TiO2 film with deposition rate of 0.2nm/s, while the second layer is 1621.34 nm SiO2 film with deposition rate of 0.5nm/s. The dielectric coatings are realized without ion-source assistance, which enables the porous structure.
The multilayer dielectric thin films were deposited on fibers using physical vapor deposition, adhesion between dielectric thin films are fine due to the continuously physical deposition, in some case ion-source assistance is employed for a better adhesion between thin films. Also the adhesion between thin films and fiber can be optimized by changing deposition parameter such as temperature and so on. The proposed sensor can still work well after 1 year, which shows good performance of multilayer thin film attachment to fiber.
The experimental set-up shown in Fig. 3 consists of a broadband light source (HL-2000 Tungsten Halogen Light Sources from Ocean Optics, wavelength range: 360-2500 nm), miniature fiber spectrometer (S3000-VIS Micro Spectrometer made from Seeman Technology, wavelength range: 320-1050 nm, wavelength resolution: 0.3 nm), multimode optical fiber coupler and the proposed F-P sensor probe working in reflection mode. During sensing characterization experiments, the light emitted from the Tungsten Halogen broadband light source goes to the optical coupler, one output port is connected to the OSA for measuring the reflected optical spectrum, while the second output port is fusion spliced with the fabricated RH sensor. Two sensors are connected to the system with an optical switch; one sensor is used for relative humidity measurement, while another is sealed in a glass tube for temperature measurement. Considering its flow ability and rapid solidification, the ultraviolet curing adhesive was used to seal the temperature sensing probe into the glass tube (the packaging was finished in the room environment, about 50%RH, 25°C).
The fiber-optic humidity sensor was enclosed in an accurate humidity generator (Model SRH-1 made by SHINYEI, Japan) with a high performance dew point hygrometer for calibration. The relative-humidity (RH) can be varied from 1.6RH% to 90%RH with control accuracy at 0.1%RH.
4. Experimental results and discussion
Figure 4(a) shows the packed sensor in a sealed glass tube for temperature measurement, the both sensors have the same multilayer coating as sensitive elements. A key issue to manufacture the fiber-optic humidity and temperature sensor is to realize and control the porous structure in the oxide films. Micro-structure of the deposited multilayer samples were investigated by scanned TEM as shown in Fig. 4(b). It can be clearly demonstrated that the coating are porous in micro-structure, and the pore size is between 20 and 50 microns dependent on the deposition process.
Figure 5(a) shows the actual spectrum in the wavelength region of 450 nm-750 nm measured by the miniature fiber spectrometer (time of exposure 510 ms, time sampling interval 5 ms, spectrum smooth degree 10) under 10%RH, 30%RH and 70%RH respectively. There exist several spectral dips of minimum reflectivity in the spectral region, since the wavelength position of minimum reflectance will be drifted by the effective refractive index change of coating material due to the absorption and desorption of water molecular, each spectral peak can be regarded as a characteristic wavelength of humidity sensing measurement. It can be observed that the interference fringe shows red-shift when the relative–humidity increases, which corresponds to an increase of effective refractive index of the sensing films. This can be explained that when RH increases, the dielectric coating absorbs more water molecules from atmosphere, and the increase of water molecules (refractive index = 1.33) filling air pores (refractive index = 1) in the coating will lead to an increase of effective refractive index in the sensing films.
The measurement results were analyzed, the shift of characteristic wavelength to different RH levels are fit and plotted in Fig. 5(b), and it can be found that the average humidity sensitivity of proposed sensor is 0.43nm/%RH approximately. Moreover, it should be mentioned that the spectrum shift reaches 8.2 nm in RH level ranging from 1.8%RH to 14.3%RH (humidity sensing test lower than 1.8%RH is not available because of the limit of experimental device), which means RH sensitivity of 0.66nm/%RH. It can be concluded that the fiber-optic humidity sensor is more sensitive at lower RH level, whereas its wavelength shift at higher RH level is not apparent. This can be understood with the fact the micro-pore is easy to absorb water molecules due to capillary condensation effect, while at higher RH level, the micro-pore is easy to be saturated.
Temperature cross-sensitivity is a main concern for the unsealed relative-humidity sensor, cross-sensitivity to temperature was investigated as for the proposed F-P fiber-optic humidity sensor. Characteristic wavelength shifts of the fiber sample at different RH levels were tested under 15°C, 25°C and 35°C respectively. Figure 6(a) shows the temperature effect to the humidity measurement. It can be concluded that the temperature cross-sensitivity is rather low, the characteristic wavelength shift is less than 2nm within 20°C of temperature change, while the change is typically more than 20nm when RH level changes from 20%RH to 80%RH. As for the proposed F-P temperature sensor, the sensor head is packed in a sealed system, cross-sensitivity to environmental RH change should be avoided to ensure the accurate measurement of temperature. Figure 6(b) shows the stability of characteristic dip wavelength of the sealed F-P temperature sensor under different environmental RH levels, while the temperature is kept at 20°C. It can be concluded that the sealed F-P sensor head is insensitive to RH, which means the sealing of the sensor by ultraviolet curing adhesive is good and the cross-sensitivity to RH of the sealed sensor can be ignored.
Temperature response of the sealed F-P sensor is conducted and the spectral shift under different temperature from 21.4°C to 38.8°C is plotted and shown in Fig. 7(a), it can be found that the characteristic dip wavelength presents blue shift with the increase of temperature. The correlation of wavelength shift to temperature is generated as shown in Fig. 7(b). Compared to the unsealed RH sensor which shows red shift with the increase of RH level, the sealed temperature presents blue shift with the increase of temperature. The blue shift caused by temperature can be explained by the Kelvin’s relation .
Experiments for sensing repeatability, response time and stability are also studied; the sensing experiments have been repeated for 4 times in order to investigate the measurement repeatability. Wavelength shifts of the proposed sensor under different humidity and temperature levels in both ascending and descending phase are shown in Fig. 8. The measurement results are quite stable at each humidity and temperature reading, taking into account that the reading error of 0.3nm. The sensor can still function even if it is over saturated, the sensor was immersed in water or alcohol several times, and it can still function normally after drying in the air.
The sensor’s response time was also evaluated. The sensor was placed in 14.3% and 60.5% RH moisture condition supplied by saturated salt solution of LiCl and NaBr in bottles. A few cycles were performed to determine the repeatability, rise time, decay time and hysteresis of the sensor. The spectrum movement was recorded with a rate of 1 frame per second, and the peak shift is plotted as shown in Fig. 9. Experimental results show that the sensor has a response/recover time of 5 seconds between 14.3% and 60.5% RH. As to the sealed temperature sensor, the response time is more than 20s from 20°C to 40°C because of the delay of heat transfer. Furthermore the stability of the proposed sensor was also examined, Fig. 9 shows the characteristic wavelength shift when the unsealed relative-humidity sensor was held at 41%RH and the sealed temperature was held at 26°C for 12 hours. The data was recorded at every 3 min interval. Only small fluctuations are observed, which demonstrate good stability of the proposed fiber-optic relative-humidity and temperature sensor. The small random errors displayed in Fig. 9 can be attributed to slight instability of micro spectrometer and temperature change during recordings.
A fiber optic sensor enabling simultaneous measurement of humidity and temperature is proposed and demonstrated. The sensing elements are multilayer porous films deposited on fiber tip by e-beam evaporation. Relative-humidity sensing is correlated with the shift of interference fringe due to the change of effective refractive index of porous coatings when exposed to different RH environments. In case of temperature measurement, the sensor is sealed in a glass tube. Experimental results show that the average sensitivity are 0.43nm/%RH and 0.63nm/°C respectively when environmental RH changes from 1.8%RH to 74.7%RH and temperature changes from 21.4°C to 38.8°C. The cross-sensitivity to temperature change for relative-humidity sensor and to humidity change for temperature sensor is low. Meanwhile the sensor presents good repeatability and stability. The proposed fiber-optic humidity sensor is very promising for application, especially for the lower humidity level.
This work is finically supported by the Project of National Science Foundation of China, NSFC (Number: 62190311, 51175393).
References and links
1. M. Yang, Y. Sun, D. Zhang, and D. Jiang, “Using Pd/WO3 composite thin films as sensing materials for optical fiber hydrogen sensors,” Sens. Actuators B Chem. 143, 750–753 (2010).
2. J. Dai, M. Yang, Y. Chen, K. Cao, H. Liao, and P. Zhang, “Side-polished fiber Bragg grating hydrogen sensor with WO3-Pd composite film as sensing materials,” Opt. Express 19(7), 6141–6148 (2011). [CrossRef] [PubMed]
3. M. Yang, H. Liu, D. Zhang, and X. Tong, “Hydrogen sensing performance comparison of Pd layer and Pd/WO3 composite thin film coated on side-polished single-and multimode fibers,” Sens. Actuators B Chem. 149, 161–164 (2010).
4. L. Alwis, T. Sun, and K. T. V. Grattan, “Design and performance evaluation of polyvinyl alcohol/polyimide coated optical fibre grating-based humidity sensors,” Rev. Sci. Instrum. 84(2), 025002 (2013). [CrossRef] [PubMed]
6. A. Vijayan, M. Fuke, R. Hawaldar, M. Kulkarni, D. Amalnerkar, and R. C. Aiyer, “Optical fibre based humidity sensor using Co-polyaniline clad,” Sens. Actuators B Chem. 129(1), 106–112 (2008). [CrossRef]
7. L. Xia, L. Li, W. Li, T. Kou, and D. Liu, “Novel optical fiber humidity sensor based on a no-core fiber structure,” Sens. Actuators A Phys. 190, 1–5 (2013). [CrossRef]
8. S. Akita, H. Sasaki, K. Watanabe, and A. Seki, “A humidity sensor based on a hetero-core optical fiber,” Sens. Actuators B Chem. 147(2), 385–391 (2010). [CrossRef]
10. Z. Li and R. Zhang, “Study on a gas sensor based on the optical character of porous silicon microcavities,” Chinese Journal of Sensors and Actuators 20, 54–57 (2007).
11. S. J. Regg, K. S. W. Sing, and M. Adsorption, Surface Area and Porosity, 2nd ed. (Academic Press, 1982).
12. E. V. Astrova and V. A. Tolmachev, “Effective refractive index and composition of oxidized porous silicon films,” J Materials Science and Engineering 69-70(60), 142–148 (2000). [CrossRef]
13. J. T. W. Yeow and J. P. M. She, “Capacitive humidity sensing using carbon nanotube enabled capillary condensation,” 5th IEEE Conference on Sensors 439–443 (IEEE, 2006).
14. K. Robbie and M. J. Brett, “Sculptured thin films and glancing angle deposition: Growth mechanics and applications,” J. Vac. Sci. Technol. A 15(3), 1460–1465 (1997). [CrossRef]
15. D. Wolfe and J. Singh, “Titanium carbide coatings deposited by reactive ion beam-assisted, electron beam–physical vapor deposition,” Surf. Coat. Tech. 124(2-3), 142–153 (2000). [CrossRef]