Open Access
Add to BibSonomy
Abstract
The metal-type microbolometers in CMOS technology normally suffer low resistivity and high thermal conductivity, limiting their performance and application areas. In this paper, we demonstrate a polysilicon microbolometer fabricated in 0.18 µm CMOS and post-CMOS processes. The detector is composed of a SiO2 absorber coupled with a salicided poly-Si thermistor that has a high resistivity of 1.37×10−4 Ω·cm and low thermal conductivity of 18 W/m·K. It is experimentally shown that the microbolometer with a 40 µm × 40 µm pixel size has a maximum responsibility and detectivity of 2.13×104 V/W and 2.33×109 cmHz1/2/W, respectively. The results are superior to the reported metal-type and diode-type microbolometers in the CMOS process and provide good potential for a low-cost, high-performance, uncooled microbolometer array for infrared imaging applications.
© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Microbolometer infrared detectors have shown impressive developments in night vision, remote sensing, space exploration etc. [1–3]. In the past few years, CMOS-based microbolometer is developed via custom CMOS technology. Remarkable achievements have been acquired in infrared-sensitive materials, device fabrication, readout circuits, and device package [4–7]. However, they have typically shown low responsivity and detectivity values. For metal-type CMOS-based microbolometers [8], the maximum dc responsivity and the detectivity is about 1751 V/W and 8.37×108 cmHz1/2/W, respectively. The poor performance is normally due to the low-temperature coefficient of resistance (TCR) and resistivity of the metal films. An alternative semiconductor material such as n-well/p-well in the CMOS process has provided a relatively large TCR of 0.5–0.7%/K [9]. Diode-type microbolometers exhibit a higher responsivity and detectivity than metal-type ones [10]. A maximum dc responsivity of 4970 V/W and detectivity of 9.7×108 Hz1/2/W have been reported. To further improve the detectivity, poly-SiGe has been proposed to implement high-format bolometer focal plane arrays [11]. However, SiGe materials need to be deposited at high temperatures, requiring the special deposition process in the standard CMOS line. On the other hand, the poly-silicon gate material used in the CMOS process has been widely applied for the thermoelectric infrared sensor [12,13], biosensor [14], humidity Sensor [15–17], and so on. For the salicided poly-Si material in the CMOS process, it can provide a high resistivity and a low thermal conductivity. However, the microbolometers using salicided poly-Si material in the CMOS process have not been reported, and there remain issues to be resolved in the bolometer fabrication. Such as the material and thickness of the infrared absorption layer cannot be arbitrarily selected and controlled due to the limitation of standard CMOS process design rules. In addition, the metal pad on the top layer cannot be effectively protected and therefore etched partially when the metal mask is etched.
In this work, a novel polysilicon microbolometer is demonstrated in CMOS technology. The salicided poly-Si is selected as thermistor material. The thermally isolated cavity and the suspended microbridge structure of the bolometer are achieved by post-CMOS process steps. The performance of the microbolometer including the temperature coefficient of resistance, effective thermal conductance, voltage responsivity, detectivity and the thermal time constant is further investigated by simulation and experiments.
2. Structural design and fabrication
The novel microbolometer is designed in 0.18 µm standard CMOS technology with six metal layers. Figure 1(a) shows the 3D view of the designed microbolometer. The microbolometer consists of two major parts. One is the temperature sensor with the size of 40 × 40 µm, including SiO2 absorber and poly-Si thermistor. This part is normally fabricated in a CMOS process. Figure 1(b) shows the cross-section view of this part and the location of the readout circuit. In this part, the salicided poly-Si thermistor with thickness and width of 0.2 µm and 0.25 µm is made from gate material in the CMOS process, which is designed to be a serpentine and narrow strip shape in order to increase temperature change quickly, wrapped with SiO2 dielectric material. The second part of the microbolometer is the suspended bridge with clearance a and width b respectively for the thermally isolated cavity. They are mainly consists of SiO2 dielectric material, in which a thermistor used to output electrical signals is wrapped. The absorber suspended by two bridges is located over a thermally isolated cavity with a depth of c, which is formed by etching the Si substrate using simple post-CMOS steps. For obtaining the air bridge, the CMOS metal layer above the poly-Si in Fig. 1(b) is taken as a etch mask to achieve narrow etch openings between the support arms.
Fig. 1. (a) 3D schematic structure of the proposed CMOS microbolometer and the readout circuit is located next to the detector as shown in the red dotted box. (b) The cross-section schematic view of the microbolometer in 0.18 µm standard CMOS technology, where the material of metal 1–6 is Al of the metal interconnection layer, which can be used as a mask for the post-CMOS process.
When the infrared (IR) light irradiates the microbolometer, the IR light is absorbed by SiO2 absorber, resulting in the temperature elevation of the structure. Then, the heat transfer quickly from SiO2 to the salicided poly-Si, which in turn results in resistance change of the polysilicon resistors placed on thermally isolated regions. The absorber and thermistor are separated from the Si substrate by the thermally isolated cavity, while the detector unit connects to the substrate only through two suspended bridges. Therefore, the substrate is treated as a heat sink and its temperature is consistent with the environment. The electrical results are further read out by the circuit. For the microbolometer, the responsivity is normally determined by the temperature rise of the bolometer, which is mainly influenced by IR absorptivity and effective thermal conductance. The temperature rise of the bolometer ΔT at time t can be given by:
The absorber structure plays an important role in receiving and transmitting IR signals. To obtain the optimal one, the infrared absorptivity and thermal conductance of the structure are studied using lumerical finite-different time-domain (FDTD) and ANSYS simulation respectively based on standard 0.18 µm CMOS process conditions. In the FDTD simulation, the size of the absorber is set to 40 µm × 40 µm, as shown in Fig. 1(a). The SiO2 thickness d can be selected to 0.75 µm/2.13 µm/3.51 µm/4.89 µm, respectively, which corresponds to the thickness of the back-end SiO2 dielectric layers in CMOS process. The clearance between the adjacent suspended bridge a, the width of suspended bridge b, and the depth of thermally isolated cavity c are set to 10 µm, 10 µm, and 70 µm respectively. The plane wave source and the perfect match layer (PML) boundary conditions are chosen in the simulation. Figure 2(a) shows the absorptivity as a function of wavelength in the range of 7 µm −14 µm for different SiO2 thicknesses d. It is found that there are three main absorptivity peaks at about 8 µm, 10 µm, and 12.5 µm in the spectrum, which are substantially consistent with the IR absorption peaks of SiO2 film reported in [18,19]. The average IR absorptivity η for various d is listed in Table 1. It is seen that η increases with d. The thicker the SiO2 layer, the larger the IR absorptivity. The maximum average value of η is up to 74.61% at d = 4.89 µm. Figure 2(b) shows the time dependence of the temperature rise ΔT by ANSYS simulation. Here, the color dots denote the ANSYS simulated data, while the solid lines indicate the calculated results from Eq. (1). In the ANSYS simulation, the density, thermal conductance and specific heat capacity of SiO2 are set as 2200 kg/m3, 1.4 W/m·K and 730 J/kg·K respectively. The initial temperature of the absorber is set to 300 K. The uniformly distributed heat power of 20 nW is applied to the surface of the SiO2 absorber with the constant temperature boundary condition.
Fig. 2. (a) The simulated absorptivity under the SiO2 thickness of 0.75, 2.13, 3.51, and 4.89 µm. (b) The simulated temperature rise with time under different SiO2 thickness. The curves represent the fitting results of simulated data by Eq. (1).
Table 1. The parameters obtained from the simulation and the calculation
It can be seen that the temperature rise in the absorber region decreases with the increase of d due to the increase of the thermal conductance of the supporting leg. In addition, a thermal time constant of 18.3 ms is obtained, which hardly changes with d. When d is 0.75µm, the highest steady-state temperature rise of 59 mK could be expected. The higher temperature rise ΔTs enables the larger responsivity of the detector. It is therefore considered that the optimal thickness d is 0.75 µm. This optimal thickness can be obtained by choosing the first layer metal in the CMOS technology as the mask for etching the materials above in the post-CMOS steps.
The main fabrication process flow of the microbolometer is illustrated in Fig. 3(a), including standard 0.18 µm CMOS process steps marked with hollow circles and post-CMOS process steps with solid circles. Firstly, the implantation of N-well and thermal annealing are performed in turn. Then, the gate oxide is grown and the poly-silicon film is deposited and salicided. After the SiO2 deposition, the first Al interconnect layer is grown. Finally, the subsequent Si3N4 passivation layer is deposited and the basic structure fabrication is accomplished, shown in Fig. 1(b). After that, the post-CMOS process steps are performed to fabricate the thermally isolated cavity. We use the first metal layer as a mask to etch the materials without the mask cover using inductively coupled plasma (ICP) until the silicon substrate is exposed. After this step, the cross-section structure of the microbolometer is shown in Fig. 3(b-I); then, the chip is immersed in 50 mL TMAH solution with Si powder and (NH4)2S2O8 for 2.5 h to etch Si substrate under SiO2 absorber and suspended bridge; finally, the first metal layer functioned as the mask is dissolved in 80 mL H3PO4 solution (≥ 85.0%) with HNO3 (65%−68%) and acetic acid at 50 °C for 2 h while the pad is protected by photoresist in this step. After etching is completed, the photoresist is dissolved in acetone, and the thermally isolated cavity is formed. The cross-section structure after post-CMOS steps is shown in Fig. 3(b-II).
Fig. 3. (a) Process flow for the fabrication. The hollow and color circles represent standard CMOS and post-CMOS processes, respectively; (b-I) The schematic view of the microbolometer structure using the ICP etching until the silicon substrate is exposed. The red arrows indicate the etching direction; (b-II) Suspended microbolometer with the thermally isolated cavity is formed after the TMAH wet etching and mask is removed.
In terms of the process flow mentioned above, the proposed microbolometer was fabricated. Figure 4 shows the SEM image of the microbridge structure of the microbolometer with different rotation angles. The top-side view in Fig. 4(a) reveals that the edges of the cavity beneath are clear and there is no obvious deformation, moreover, the backside of the microbolometer is flat without redundant etching residue, as shown in Fig. 4(b). Those results inform that the microbolometer structure with the thermally isolated cavity has been successfully prepared with excellent shape and toughness. The microbolometer is separated from the thermally isolated cavity and moved to a flat surface and the SiO2 absorber thickness of microbolometer is measured by the KLA Tencor P6 stylus profiler and shown in Fig. 4(c). It can be found that the thickness of the SiO2 absorber is about 0.85 µm, which is basically similar to the value of the experimental design. The difference is mainly because the microbolometer is not closely attached to the surface of the chip during the measurement process.
Fig. 4. SEM images of suspended microbolometer: (a) top-side and (b) back-side view of the microbolometer. (c) The SiO2 absorber thickness of microbolometer. The inset shows the microbolometer separated from the thermally isolated cavity and the yellow dashed line represents the measure position.
3. Experimental results and discussions
The temperature coefficient of resistance (TCR), effective thermal conductance (Geff), voltage responsivity (R (f)), detectivity (D*), noise equivalent power (NEP) and thermal time constant (τ) were experimentally characterized. Figure 5(a) demonstrates the extracted resistance R as a function of temperature rise ΔT for the microbolometer, which is linearly fitted with the following equation [20]:
where $\alpha $ is the TCR of the salicided poly-Si material, T is the test temperature, T0 is the room temperature of 300 K, and R0 is the resistance of the thermistor at T0. It is observed that all experimental data appear to fall on the fitting curve, informing the R linearly increases with temperature rise. Through the linear fitting, the TCR of about 0.35%/K is obtained, which is higher than those values for non-salicided polysilicon [21], Ti [22] and Wu [23].Fig. 5. (a) The measured resistance of the microbolometer in the temperature rise range of 0 - 40 K. The red line represents the linear fitting by Eq. (2); (b) Inverse of the resistance vs. square of the bias current measured in vacuum condition. The red line represents the linear fitting by Eq. (3).
The effective thermal conductance Geff of the microbolometer is further evaluated by measuring the resistance change caused by the bias current self-heating effect [21] following the Eq. (3):
where Ib is the bias current. Figure 5(b) shows the inverse of the measured resistance versus the square of bias current in the vacuum pressure. It is seen that the measured data can be fitted well by Eq. (3). The fitted Geff is estimated as 3.08 × 10−7 W/K, which is consistent with the simulation value of 1.75×10−7 W/K listed in Table 1 and can be comparable to the reported a-Si bolometers of 6.2 × 10−7W/K, 1.8 × 10−7 W/K, and 6.72 × 10−7 W/K, respectively [24].The measurements of R(f), D*, and NEP are particularly concerned for the microbolometer, and the schematic of the experimental set-up is shown in Fig. 6(a). Here, RL and R are the load and microbolometer resistance, respectively. The RL is selected as 520 kΩ to obtain a constant bias current. First of all, the test chip is packaged in dual In-Line ceramic package (DIP) CSB04839 and placed in a customized vacuum chamber, in which the pressure is kept less than 10 Pa by using VRD-8 vacuum pump. IR lights with a wavelength of 8.26 µm were generated by a mid-IR quantum cascade laser (QCL) source and modulated by MC200B chopper. Then, the signal voltage ΔV will be generated and collected by the SR830 Lock-In amplifier. The noise spectral density was measured by Lakeshore CRX-4K cryogenic probe station combined with the FS-Pro noise testing system. The voltage responsivity is then calculated under different bias currents by the following formula:
Here R (f) is voltage responsivity at the chopper frequency of f, ΔV is the signal voltage, and Pin is the input power of the IR illumination.
Fig. 6. (a) The schematic of the experimental set-up of the R(f) measurements. (b) R as a function of bias current and the red line represents the linear fitting result. (c) The bias current dependence of noise voltage at the bandwidth of 10 kHz. The inset shows the noise power spectral density of the microbolometer under the vacuum condition in the current range of 3 µA to 29 µA. (d) Evaluated D* and NEP of the microbolometer under various bias currents.
Figure 6(b) shows the bias current dependence of dc voltage responsivity R for the microbolometer at the chopper frequency f = 0 Hz. All data fall on the linear fitting curve, indicating that the R is linearly dependent on the bias current. Figure 6(c) shows the total noise voltage Vn as a function of bias current at the bandwidth of 10 kHz. The value of Vn is extracted from the noise spectral density in the bias current range from 3 µA to 29 µA, shown in the inset of Fig. 6(c). It can be found that Vn gradually increases with the bias current increases from 3.4 µV to 5.4 µV. This is because an increase of bias current gives rise to the increase of the low-frequency 1/f noise and thermal noise of the microbolometer. The 1/f noise at low frequencies of salicided poly-Si mainly comes from electron-mobility fluctuations occurring in the depletion regions near grain boundaries [25,26] and the thermal noise of polysilicon mainly comes from the irregular thermal motion of carriers. Figure 6(d) shows the bias current dependence of D* and NEP at f = 0 Hz by using the following formulas [13]:
Here, Ad is the area of the absorber, Δf is electrical bandwidth. As can be seen from Fig. 6(d), the D* increases at first and then saturates and finally decreases with the bias current. The main reason for the decrease is the noise increasing with the bias current. The maximum value of D* of 1.91 × 109 cmHz1/2/W is achieved at about 25 µA. In contrast, the NEP decreases significantly and then tends to stabilize gradually with the increase of bias current, and the value reaches a minimum of 2.11 × 10−10 W at 25 µA. Thus, the optimal operating current of the microbolometer is about 25 µA, and the R is about 2.37 × 104 V/W at this point.
To obtain the thermal time constant τ of the microbolometer, the detectivity of the microbolometer as a function of chopping frequency is measured at the bias current of 25 µA, and the result is displayed in Fig. 7. It is seen that as the chopping frequency is increased from 4 Hz to 150 Hz, the D* decreases from 1.67 × 109 cmHz1/2/W to 8.61 × 107 cmHz1/2/W. Further, the frequency dependence of the test detectivity is fitted by the following equation:
where D*(f) is the detectivity at the chopper frequency of f. τ and f are thermal time constant and frequency of the chopper respectively. D* is obtained from Fig. 6(d) with the value of 1.91 × 109 cmHz1/2/W at 25 µA. It is found all data fall on the fitting curve. The thermal time constant of 22 ms is then obtained by the fitting. The value exhibits a similar value to the simulated one of 18.3 ms from Fig. 2(b). The little difference between them is due to the non-ideal vacuum environment of the fabricated one.Fig. 7. The detectivity of the microbolometer concerning modulation frequency was measured at 25 µA under vacuum condition. The red curve represents the fitting result by Eq. (7). All data appears to fall on the fitting line.
The effects of the resistance of salicided poly-Si on the properties of microbolometers are further investigated. Table 2 compares the main performance parameters of the microbolometers using different resistances. All the samples are designed with the same parameter except the resistance. It is found that with the increase of resistance, the R value of the microbolometer is improved. However, D* increases first and then decreases, and a maximum value of 2.33 × 109 cmHz1/2/W is obtained when the resistance is 43.45 kΩ. The decrease of D* after the resistance is larger than 43.45 kΩ can be attributed to the dramatic increase of the noise voltage Vn. Additionally, the NEP shows a change opposite to the D* with the increase of resistance and it reaches a minimum value of 1.69 × 10−10 W at the same point of 43.45 kΩ.
Table 2. The properties of the microbolometers with different resistance. All measurements are performed at a current of 25 µA
Table 3 compares the voltage responsivity and detectivity between the proposed microbolometers and the currently reported ones in CMOS or non-CMOS technologies. It can be found that the proposed new microbolometer exhibits a relatively high voltage responsivity of 2.13 × 104 V/W; moreover, its detectivity is higher than the normal Al and N-well microbolometers by CMOS technology. This is due to the salicided poly-Si having a high resistivity and a lower thermal conductivity. For example, the resistivity and thermal conductivity of the metal Al are 2.83 × 10−6 Ω·cm and 238 W/m·K, respectively, while those of the salicided poly-Si are 1.37 × 10−4 Ω·cm and 18 W/m·K, respectively. Additionally, the detectivity of the salicided poly-Si microbolometer is comparable even to that of VOx, a-Si, and SiGe fabricated by the non-CMOS process. More importantly, the performance of the microbolometer with the salicided poly-Si thermistor can be further improved by the structure design of surface plasmonic resonance, the resonant cavity, etc. Further study is now going on.
Table 3. The comparison of the performance between ours and the previously reported works.
4. Conclusion
The microbolometer using salicided poly-Si as the thermistor has been successfully prepared by the standard CMOS technology and the simple post-CMOS process steps. It is experimentally shown that the salicided poly-Si in the detector has a positive TCR of 0.35%/K, and the thermal conductance of the microbolometer is up to 3.08 × 10−7 W/K. A promising performance, such as high voltage responsivity of 2.13 × 104 V/W, high detectivity of 2.33 × 109 cmHz1/2/W, low noise equivalent power of 1.69 × 10−10 W are achieved for the proposed microbolometer when the resistance is about 43.45 kΩ. These distinguishing features reveal that the salicided poly-Si microbolometer detector has the potential to provide fascinating performance in an infrared imaging system.
Funding
National Natural Science Foundation of China (61627804); National Key Research and Development Program of China (2016YFB0402403, 2016YFB0400402); Key Laboratory of Infrared Imaging Materials and Detectors (IIMDKFJJ-19-07).
Disclosures
The authors declare no conflicts of interest.
Data availability
The data underlying results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. B. Du, Y. He, Y. He, and C. L. Zhang, “Progress and trends in fault diagnosis for renewable and sustainable energy system based on infrared thermography: A review,” Infrared Phys. Technol. 109, 103383 (2020). [CrossRef]
2. L. Yu, Y. Z. Guo, H. Y. Zhu, M. C. Luo, P. H. and X, and L. Ji, “Low-Cost microbolometer type infrared detectors,” Micromachines 11(9), 800 (2020). [CrossRef]
3. C. H. Shin, D. P. Pham, J. Park, Y. J. Lee, S. Kim, and J. Yi, “Investigation of boron-doped hydrogenated silicon films as a thermo-sensing layer for uncooled microbolometer,” Thin Solid Films. 690, 137515 (2019). [CrossRef]
4. A. Voshella, N. Dharb, and M. M. Rana, “Materials for microbolometers: vanadium oxide or silicon derivatives,” Proc. SPIE 10209, 102090M (2017). [CrossRef]
5. C. Fang, X. Q. Chen, and X. J. Yi, “High-speed CMOS readout integrated circuit for large-scale and high-resolution microbolometer array,” Optik 125(13), 3315–3318 (2014). [CrossRef]
6. T. Zhou, H. Liu, Y. Xu, B. Jiang, and Y. Su, “A high-performance operational amplifier for infrared readout circuit application,” J. Phys.: Conf. Ser. 1176, 062014 (2019). [CrossRef]
7. B. Fisette, M. Tremblay, H. Oulachgar, F. Généreux, D. Béland, P. Beaupré, C. Julien, D. Gay, S. Deshaies, M. Terroux, B. Tremblay, D. Dufour, and C. Alain, “Novel vacuum packaged 384 ( 288 broadband bolometer FPA with enhanced absorption in the 3-14 µm wavelength range,” Proc. SPIE 10177, 101771R (2017). [CrossRef]
8. N. Shen, J. Yu, and Z. A. Tang, “An uncooled infrared microbolometer array for low-cost application,” IEEE Photonics Technol. Lett. 27(12), 1247–1249 (2015). [CrossRef]
9. D. S. Tezcan, S. Eminoglu, O. S. Akar, and T. Akin, “An uncooled microbolometer infrared focal plane array in standard CMOS,” Proc. SPIE 4288, 112–121 (2001). [CrossRef]
10. S. Eminoglu, M. Y. Tanrikulu, and T. Akin, “A Low-Cost 128 × 128 Uncooled Infrared Detector Array in CMOS Process,” J. Microelectromech. Syst. 17(1), 20–30 (2008). [CrossRef]
11. R. Jimenez, M. Moreno, A. Torres, A. Morales, A. Ponce, D. Ferrusca, J. R. Magdaleno, J. C. Ramos, J. H. Perez, and E. Cano, “Fabrication of microbolometer arrays based on polymorphous silicon-germanium,” Sensors 20(9), 2716 (2020). [CrossRef]
12. P. S. Lin, T. W. Shen, K. C. Chan, and W. Fang, “CMOS MEMS thermoelectric infrared sensor with plasmonic metamaterial absorber for selective wavelength absorption and responsivity enhancement,” IEEE Sens. J. 20(19), 11105–11114 (2020). [CrossRef]
13. T. W. Shen, K. C. Chang, C. M. Sun, and W. Fang, “Performance enhance of CMOS-MEMS thermoelectric infrared sensor by using sensing material and structure design,” J. Micromech. Microeng. 29(2), 025007 (2019). [CrossRef]
14. M. C. Chen, H. Y. Chen, C. Y. Lin, C. H. Chien, T. F. Hsieh, J. T. Horng, J. T. Qiu, C. C. Huang, C. H. Ho, and F. L. Yang, “A CMOS-Compatible Poly-Si nanowire device with hybrid sensor/memory characteristics for System-on-Chip applications,” Sensors 12(4), 3952–3963 (2012). [CrossRef]
15. C. L. Zhao, M. Qin, and Q. A. Huang, “A fully packaged CMOS interdigital capacitive humidity sensor with polysilicon heaters,” IEEE Sens. J. 11(11), 2986–2992 (2011). [CrossRef]
16. S. K. Park, Y. H. Yang, C. Y. Lee, Y. J. Kwon, T. S. Shi, J. H. Park, C. Hong, I. W. Cho, and K. D. Yoo, “A study of vertical thin Poly-Si channel transfer gate structured CMOS Image sensors,” IEEE Electron Device Lett. 38(2), 232–235 (2017). [CrossRef]
17. S. K. Yeh, M. L. Hsieh, and W. Fang, “CMOS-Based tactile force sensor: a review,” IEEE Sensors J. 21(11), 12563–12577 (2021). [CrossRef]
18. J. Schäfer, J. Hnilic, J. Šperk, A. Quade, V. Kudrle, R. Foest, J. Vodák, and L. Zajíčková, “Tetrakis (trimethylsilyloxy) silane for nanostructured SiO2-like films deposited by PECVD at atmospheric pressure,” Surf. Coat. Technol. 295, 112–118 (2016). [CrossRef]
19. Y. Song, T. Sakurai, K. Maruta, A. Matusita, S. Matsumoto, S. Saisho, and K. Kikuchi, “Optical and structural properties of dense SiO2, Ta2O5 and Nb2O5 thin-films deposited by indirectly reactive sputtering technique,” Vacuum 59(2-3), 755–763 (2000). [CrossRef]
20. R. K. Bhan, R. S. Saxena, C. R. Jalwania, and S. K. Lomash, “Uncooled infrared microbolometer arrays and their characterisation techniques,” Def. Sci. J. 59(6), 580–589 (2009). [CrossRef]
21. N. Shen, Z. X. Huang, and Z. A. Tang, “A Low-Cost uncooled infrared microbolometer by CMOS process,” ECS Solid State Lett. 4(7), Q29–Q31 (2015). [CrossRef]
22. X. Gu, G. Karunasiri, and G. Chen, “Determination of thermal parameters of microbolometers using a single electrical measurement,” Appl. Phys. Lett. 72(15), 1881–1883 (1998). [CrossRef]
23. N. Shen, J. Yu, and Z. A. Tang, “The design and fabrication of a tungsten microbolometer array integrated chip in a standard CMOS process,” Chinese Journal of Sensors and Actuators 27(6), 726–729 (2014).
24. S. Gupta, A. Katiyar, R. K. Bhan, and R. Muralidharan, “Design optimization of pixel structure for α-Si based uncooled infrared detector,” Def. Sci. J. 63(6), 581–588 (2013). [CrossRef]
25. H. C. Degraaff and M. T. M. Huybers, “1/f noise in polycrystalline silicon resistors,” J. Appl. Phys. 54(5), 2504–2507 (1983). [CrossRef]
26. S. L. Jang, “A model of 1/f noise in polysilicon resistors,” Solid-State Electron 33(9), 1155–1162 (1990). [CrossRef]
27. S. Eminoglu, D. S. Tezcan, M. Y. Tanrikulu, and T. Akin, “Low-cost uncooled infrared detectors in CMOS process,” Sens. Actuator A-Phys. 109(1-2), 102–113 (2003). [CrossRef]
28. B. Wang, J. J. Lai, H. Li, H. M. Hu, and S. H. Chen, “Nanostructured vanadium oxide thin film with high TCR at room temperature for microbolometer,” Infrared Phys. Technol. 57, 8–13 (2013). [CrossRef]
29. H. J. Lee, D. Wang, T. H. Kim, D. H. Jung, T. H. Kil, K. S. Lee, H. J. Choi, S. H. Baek, E. Yoon, W. J. Choi, and J. M. Baik, “Wide-temperature (up to 100 ◦C) operation of thermostable vanadium oxide based microbolometers with Ti/MgF2 infrared absorbing layer for long wavelength infrared (LWIR) detection,” Appl. Surf. Sci. 547, 149142 (2021). [CrossRef]
30. M. H. Unewisse, B. I. Craig, R. J. Watson, O. Reinhold, and K. C. Liddiard, “The growth and properties of semiconductor bolometer for infrared detection,” Proc. SPIE 2554, 43–54 (1995). [CrossRef]
31. M. Almasri, “Amorphous silicon two-color microbolometer for uncooled IR detection,” IEEE Sens. J. 6(2), 293–300 (2006). [CrossRef]
32. S. Sedky, P. Fiorini, K. Baert, L. Hermans, and R. Mertens, “Characterization and optimization of infrared poly SiGe bolometers,” IEEE Trans. Electron Devices 46(4), 675–682 (1999). [CrossRef]
