Germanium tin oxide thin films for uncooled infrared detectors

10337927 ยท 2019-07-02

    Inventors

    Cpc classification

    International classification

    Abstract

    Microbolometer is a class of infrared detector whose resistance changes when the temperature changes. In this work, we deposited and characterized Germanium Oxide thin films mixed with Sn (GeSnO) for uncooled infrared detection. GeSnO were deposited by co-sputtering of Sn and Ge targets in the Ar+O environment using a radio frequency sputtering system. Optical characterization shows that the absorption in GeSnO was most sensitive in the wavelength ranges between 1.0-3.0 m. The transmission data was further used to determine the optical energy band gap (0.678 eV) of the thin-film using Tauc's equation. We also found the variations of absorption coefficient (1.480210.sup.5 m-.sup.1-1.009710.sup.7 m.sup.1), refractive index (1.242-1.350), and the extinction coefficient (0.3255-8.010) for the wavelength ranges between 1.0-3.0 m. The thin film's resistivity measured by the four point probe was found to be 4.55 -cm and TCR was in the range of 2.56-2.25 (%/K) in the temperature range 292-312K. In light of these results it can be shown that this thin film is in keeping with the current standards while also being more cost and time effective.

    Claims

    1. A method of forming a thin film comprising: Ge.sub.xSn.sub.yO.sub.z where, 39x54, 09y20, 19z46.

    2. The method of claim 1, wherein the thin film is deposited by one of radio frequency sputtering, direct current sputtering, chemical vapor deposition, metal organic chemical vapor deposition techniques.

    3. The method of claim 1, wherein the thin film of Ge.sub.xSn.sub.yO.sub.z where x, y, z values, 39x54, 09y20, 19z46 is deposited in Ar:O environment by radio frequency and direct current sputtering method.

    4. The method of claim 1, wherein the transmittance, reflectance and absorptance of the thin film varies between 0% to 90%, 0% to 75% and 0% to 100% respectively for the wavelength ranges 1.0 m to 3.0 m.

    5. A thin film comprising: germanium tin oxide with a chemical formula Ge.sub.xSn.sub.yO.sub.z with x, y, z values 39x54, 09y20, 19z46.

    6. The thin film of claim 5, whose refraction index varies between 1.2 to 14 and extinction coefficient varies between 0.32 to 8.01 for the wavelength ranges from 1.0 m to 3.0 m.

    7. The thin film of claim 5, whose transmittance, reflectance and absorptance varies between 0% to 90%, 0% to 75% and 0% to 100% respectively for the wavelength ranges 1.0 m to 3.0 m.

    8. The thin film of claim 5, wherein the optical bandgap of the thin film varies between 0.51 eV to 0.93 eV.

    9. The thin film of claim 5, wherein the activation energy of the thin film has a value between 0.007 eV to 0.196 eV depending on the atomic composition of the thin film.

    10. The thin film of claim 5, wherein the resistivity of the thin film varies between 3.24 Ohm-cm to 1.88 Ohm-cm between the temperature ranges 292K-312K.

    11. The thin film of claim 5, wherein coefficient of Resistance (TCR) varies in the range2.56%/K-2.25%/K between the temperatures ranges 292K-312K.

    12. A method of forming a microbolometer comprising: a thin film comprising Ge.sub.xSn.sub.yO.sub.z where x, y, z values, 39x54, 09y20, 19z46.

    13. The method of claim 12, the thin film is deposited by one of radio frequency sputtering, direct current sputtering, chemical vapor deposition, metal organic chemical vapor deposition techniques.

    14. The method of claim 12, wherein the thin film of Ge.sub.xSn.sub.yO.sub.z deposited in Ar:O environment by radio frequency sputtering method.

    15. The method of claim 12, wherein the thin film has a refraction index varies between 1.2 to 1.4 and extinction coefficient varies between 0.32 to 8.01 for the wavelength ranges 1.0 m to 3.0 m.

    16. The method of claim 12, wherein the optical bandgap of the thin film varies between 0.51 eV to 0.93 eV.

    17. The method of claim 12, wherein the activation energy of the thin film has a value between 0.007 eV to 0.196 eV depending on the atomic composition of the thin films.

    18. The method of claim 12, wherein the resistivity of the thin film varies between 3.24 Ohm-cm to 1.88 Ohm-cm between the temperature ranges 292K-312K.

    19. The method of claim 12, wherein the temperature coefficient of Resistance (TCR) of the thin film varies in the range 2.56%/K-2.25%/K between the temperatures ranges 292K-312K.

    Description

    4. BRIEF DESCRIPTION OF FIGURES

    (1) FIG. 1: Flowchart illustrating the method for preparing the Ge.sub.xSn.sub.yO.sub.z thin films where 0.39x0.54, 0.09y0.20, 0.19z0.46.

    (2) FIG. 2: Atomic Composition of GeSnO thin films determined by energy dispersive spectroscopy at 5 keV.

    (3) FIG. 3: Refractive index and extinction coefficient of GeSnO thin film.

    (4) FIG. 4: Transmittance, reflectance and absorptance of GeSnO thin film.

    (5) FIG. 5: Optical band gap of GeSiSnO determined from transmittance, reflectance and absorptance data.

    (6) FIG. 6: Arrhenius plot at various temperature to determine the activation energy of GeSnO thin films.

    (7) FIG. 7: Variations of resistivity and TCR with temperature.

    (8) FIG. 8: Cross section of a GeSnO Microbolometer.

    5. DETAILED DESCRIPTION OF EXEMPLARY EMBODIMENTS

    (9) We present here the microbolometer elements and methods for forming the same. The sensing layer of microbolometer is fabricated using various atomic compositions of Germanium, Tin, and Oxygen to form an alloy of germanium-tin-oxide (GeSnO). The variation in atomic composition in the alloy of GeSnO will allow the materials' properties to be varied. The variation of the atomic composition in GeSnO alloy will vary the fundamental material properties such as activation energy and carrier mobility. This way the key microbolometer device figures of merits such as resistivity, TCR, responsivity, absorption in IR region of interest, noise, detectivity, noise equivalent temperature difference are varied and optimized for better device performance.

    (10) Variation and optimization of atomic composition in the alloy of GeSnO allow us to increase the room temperature TCR to a value which is significantly higher than the current reported values (from 2%/k to 5%/k). TCR is one of the important properties of microbolometer and it defines the sensitivity of the microbolometer device which is defined as change in electrical resistance of the device with change in temperature. The greater the value of TCR, the better will be the sensitivity of the microbolometer device. Herein, we used oxygen to bond with other semiconductors (Ge and Sn) to increase the TCR among other properties. In the preferred embodiment we found the TCR varied between 2.25%/K to 2.56%/K for the temperature ranges 292K-312K. The resistivity for this temperature range for the thin films of GeSnO is varied between 0.5426 Ohm-cm to 45.67 Ohm-(cm). The higher resistivity of the Ge.sub.xSn.sub.yO.sub.z where 0.39x0.54, 0.09y0.20, 0.19z0.46, alloy had been resulted from lower electrical conductivity which arises because of the low carrier mobility associated with this.

    (11) The atomic composition of the microbolometer sensing layer made of Ge.sub.xSn.sub.yO.sub.z thin film layer consists of two semiconducting materials (Ge, Sn) forming an alloy and bonded with oxygen where 0.39x0.54, 0.09y0.20, 0.19z0.46. The element Ge, may or may not be doped with impurities such as n-type or p-type impurities while they are used as one of the constituent elements. In one embodiment, we disclose the atomic composition of Ge.sub.xSn.sub.yO.sub.z where the value of x is the atomic percentage of Ge in that alloy relative to the atomic percentage of Sn. The value of x varies between 0.39 to 0.54 while the value of y varies between 0.09 to 0.20. The value of z varies between 0.19 to 0.46.

    (12) In another embodiment, the thin film of Ge.sub.xSn.sub.yO.sub.z will be deposited using the sputtering process, where 0.39x0.54, 0.09y0.20, 0.19z0.46. The sputtering process consists of radio frequency as well as direct current method.

    (13) In the present invention, the exemplary embodiments are described in details in the related art work hereinafter. The drawings presented here, will explain the current invention better although it should be noted that the present invention is not limited to the drawings.

    (14) FIG. 1 is a flow chart illustrating a method for fabricating the Ge.sub.xSn.sub.yO.sub.z thin film to be used as the sensing layer of the microbolometer in accordance with an embodiment of the present invention where 0.39x0.54, 0.09y0.20, 0.19z0.46. FIG. 1 shows the method for manufacturing a Ge.sub.xSn.sub.yO.sub.z thin film for a bolometer's sensing layer where 0.40x0.90, 0.39x0.54, 0.09y0.20, 0.19z0.46. The preferred embodiment of the present invention includes is completed in a single step: this includes the deposition of Ge.sub.xSn.sub.yO.sub.z thin film on a substrate; In accordance with the invention, the thin films of Ge.sub.xSn.sub.yO.sub.z are deposited by co-sputtering of Sn and Ge targets in the Ar+O environment using a radio frequency sputtering system. The films are deposited on silicon or cover glass or other conformal substrates. The thin films are also deposited by DC sputtering or metal-organic chemical vapor deposition (MOCVD) or chemical vapor deposition process using Ge, Sn, and O based gases and precursors.

    (15) In this case, the DC sputter and the RF sputter use argon (Ar) or nitrogen (N.sub.2) plasma. The thickness of the Ge.sub.xSn.sub.yO.sub.z thin film may ranges from 1400 nm to 1450 nm. According to the invention, the thin films of Ge.sub.xSn.sub.yO.sub.z comprise of Ge, Sn, and O elements with values of where 0.39x0.54, 0.09y0.20, 0.19z0.46. When the z of Ge.sub.xSn.sub.yO.sub.z is >0.46, the electric resistance of the thin film is too large. Accordingly, it is not suitable for the bolometer applications. When the oxygen component, z, is close to zero, the electric resistance is too small and the TCR value is small as well, thereby making it not suitable for the bolometer.

    (16) A turbo pump evacuated the chamber to a base pressure of 310.sup.7 Torr or less before sputtering. This compound target of Ge and Tin (Sn) target both were used simultaneously to deposit the Ge.sub.xSn.sub.yO.sub.z thin films on silicon or glass or other substrates which may be flexible or rigid. The deposition will be done by using DC and/or RF sputter. The DC sputter and RF sputter use Argon (Ar) or nitrogen (N.sub.2) plasma. The deposition process takes place in an oxygen atmosphere. The sputtering will take as long as 2 hours but in this case, the Ge and Sn targets were sputtered for 40 minutes. Ge was sputtered at 250 W while Sn was sputtered at 75 W. The preferred embodiment includes a design of material for each of the layers of the microbolometer whose composition and thickness leads to high figures of merits. Germanium Tin Oxide (GeSnO) was created by co-sputtering Ge with Sn targets in the Ar+O environment using a radio frequency sputtering system.

    (17) Radio frequency sputtering system is defined as a technique involved in alternating the electrical potential of the current in the vacuum environment at radio frequencies to avoid a charge building up on certain types of sputtering target materials, which over time will result in arching into the plasma that spews droplets creating quality control issues on the thin films and will even lead to the complete cessation of the sputtering of atoms terminating the process. In this case, Ge at 250 W and Sn at 75 W were sputtered for 40 minutes with a deposition rate of 5.997 per second on a silicon wafer.

    (18) After sputtering process, energy dispersive spectroscopy function from a scanning electron microscope was used to determine the atomic compositions of the thin film.

    (19) FIG. 2: shows the atomic Composition of Ge.sub.xSn.sub.yO.sub.z thin films determined by energy dispersive spectroscopy done at 5 keV where 0.39x0.54, 0.09y0.20, 0.19z0.46. The unique peaks in FIG. 2 indicate the estimated value of abundance of the different elements in the thin film. In the preferred embodiment, Ge.sub.xSn.sub.yO.sub.z thin films consist of 30.2% Oxygen, 14.3% Tin, 48.3% Germanium, and 7.1% Carbon where 0.39x0.54, 0.09y0.20, 0.19z0.46. As mentioned earlier, the thin films of Ge.sub.xSn.sub.yO.sub.z comprise of Ge, Sn, and O elements. In the preferred embodiment, the values ranges 0.39x0.54, 0.09y0.20, 0.19z0.46 Determination of atomic composition of the constituent elements in the thin films of GeSnO is crucial for not only the correct elemental composition in GeSnO but also for unique behavior of Ge.sub.xSn.sub.yO.sub.z thin film (where 0.39x0.54, 0.09y0.20, 0.19z0.46) for sensing layer of microbolometer.

    (20) FIG. 3 indicates the refractive index and extinction coefficient of the GeSnO thin film. Refractive index is a number that describes how light propagates through that medium, while the extinction coefficient indicates several different measures of the absorption of light in a medium. Extinction coefficient refers to a measure of the rate of decrease in the intensity of electromagnetic radiation (as light) as it passes through a given substance;

    (21) Extinction coefficient k() is given by:

    (22) k ( ) = a 4 ( 6 )

    (23) Where a is the absorption coefficient, is the wavelength.

    (24) Reflective index n() is the measure of the propagation of a ray of light as it passes from one medium to another and it was also derived by:

    (25) n ( ) = n s ( ) ( 1 + R 1 - R ) 0.5 ( 7 )
    Where, n.sub.s is the refractive index of the substrate (Silicon wafer in our case) and R is the Reflectance. The refractive index for GeSnO varied between 1.2 to slightly above 1.3 for the wavelength ranges between 1.0 m to 3.0 m. These values are different from pure Ge, Sn, or their oxides. The extinction coefficient varied between 0.32 to 8.01 for the wavelength ranges between 1.0 m to 3.0 m. The values of the extinction coefficient are also different from pure Ge, Sn, or their oxides. The values of extinction coefficient and refractive index at various wavelengths determines the absorption, reflection and transmission through the thin films. The absorption in the film at a particular wavelength is crucial as because of the absorption the thin films' temperature changes and we get a TCR. The values of absorption coefficient along with the reflection coefficient and transmittance at various wavelength is mentioned in FIG. 4.

    (26) FIG. 4 indicates the transmittance, reflectance and absorptance of Ge.sub.xSn.sub.yO.sub.z thin film, where 0.39x0.54, 0.09y0.20, 0.19z0.46. These measurements were carried out using monochromator, infrared (IR) light source, pyroelectric detector, mechanical chopper, dynamic signal analyzer and computer. Infrared (IR) light signal was generated using a IR source which came out of the monochromator and was chopped at 100 Hz. Results of peak signal at 100 Hz and noises were recorded from dynamic signal analyzer in order to find the corresponding transmittance and reflectance in the thin films between the wavelength ranges 2.5-5.5 m. Kirchhoff's law was then used to calculate the absorptance.

    (27) Transmittance () through a thin film is expressed as the ratio of transmitted flux (.sub.t) to the incident flux (.sub.i)

    (28) ( ) = t i ( 8 )
    where, is the wavelength.

    (29) Kirchhoffs law relating to absorptance (), transmittance () and reflectance () is expressed as:
    ++=1(9)

    (30) Absorption coefficient is a measure of the rate of decrease in the intensity of electromagnetic radiation (as light) as it passes through a given substance. For application as the sensing layer of microbolometer, we need the absorption value to be as high as possible. Higher absorption in the sensing layer will heat up the material and then the resistance of the material will reduce. So the higher absorption is associated with greater sensitivity. The addition of Sn in Ge.sub.xSn.sub.yO.sub.z thin films, increase the absorption in the wavelength ranges between 2 to 4.5 where 0.40x0.90, 0.08y0.60, 0.01z0.20.

    (31) FIG. 5 is Optical band gap of Ge.sub.xSn.sub.yO.sub.z determined from transmittance, reflectance and absorptance data which shown earlier. Extrapolation of (h).sup.1/2=0 of linear portion in the plot of (h) versus the photon energy h, gave the value of optical bandgap as mentioned in FIG. 5 where 0.39x0.54, 0.09y0.20, 0.19z0.46. To determine the optical bandgap, we used Tauc's equation for direct bandgap semiconductor which is expressed below:
    h=B(hE.sub.9).sup.1/2(10)
    Where B is a constant, h is the photon energy and E.sub.g is the energy band gap. (h).sup.2 was graphed as a function of h and the energy band gap was determined at the h value where =0. Optical bandgap is an important parameter for optical detectors such as microbolometers. The optical bandgap of GeSnO in this case is 0.678 eV which is unique for the Ge.sub.0.483Sn.sub.0.143O.sub.0.302 thin film. The optical bandgap plays an important role in optoelectronic properties of the material. When the photon falls on top of the material, if its energy is higher than the optical bandgap then it creates and electron hole pair. This is the principle of operation for a photon detector. For thermal detectors like microbolometers, the detector material's temperature rises and the resistivity fo the material changes. Hence, electrical bandgap is an inherent property of the materialGeSnO.

    (32) FIG. 6 illustrates Arrhenius plot at various temperature to determine the activation energy of Ge.sub.xSn.sub.yO.sub.z thin films where 0.39x0.54, 0.09y0.20, 0.19z0.46. Activation energy is the minimum energy required to cause a process (such as a chemical reaction) to occur. It is expressed using Arrhenius formula as follows;

    (33) k = - E a R ( 1 T ) + ln A ( 11 )
    where, k represents the rate constant, E.sub.a is the activation energy, R is the gas constant (8.3145 J/K mol), and T is the temperature expressed in Kelvin. A is known as the frequency factor, having units of L mol.sup.1 s.sup.1, and takes into account the frequency of reactions and likelihood of correct molecular orientation.

    (34) Four point probe instrument was used to measure average resistance of the film by passing current through the outside two points of the probe and measuring the voltage across the inside two points. This resistance is called sheet resistance (R.sub.s).

    (35) If the spacing between the probe points is constant, and the conducting film thickness is less than 40% of the spacing, and the edges of the film are more than 4 times the spacing distance from the measurement point, the average resistance of the film or the sheet resistance is given by;
    Rs=4.53V/I(11)
    Total resistance R is given by;
    R=R.sub.s*t(12)

    (36) Where, R.sub.s is the measured sheet resistance and t is the films thickness. From FIG. 6 it is seen that the plot of natural logarithm of resistivity () versus 1/kT is a straight line and the slope of the straight line provides the value of the activation energy associated with this process. The higher the activation energy, higher will be the TCR associated with it. Adding oxygen with the GeSn compound increases the activation energy (Ea). The activation energy for Ge.sub.0.483Sn.sub.0.143O.sub.0.302 thin film is 0.1859 eV.

    (37) In FIG. 7, we show the variations of temperature coefficient of resistance, TCR and resistivity for a resistor is determined by measuring the resistances values over an appropriate temperature range in Kelvin. The TCR is calculated as the average slope of the resistance value over this interval. TCR exhibits how rapidly the resistance of the sensing material responds to a change in temperature and is expressed as;

    (38) = 1 R .Math. d R d T = 1 R R T = - E a kT 2 ( 13 )
    Here, E.sub.a is the activation energy and k is the Boltzmann constant.

    (39) TCR is a material property, so the higher the value, the better it is for IR uncooled detection. The room temperature (301K) TCR value is 2.53%/K which is at on par with the Vanadium Oxide that is used as one of the most popular materials for microbolometers. The resistivity of the for Ge.sub.0.483Sn.sub.0.143O.sub.0.302 thin film varied between 3.24 Ohm-cm to 1.88 Ohm-cm for the temperature range between 290 K to 325 K

    (40) FIG. 8 shows the cross sectional view of the microbolometer using GeSnO sensing layer. The substrate 101 in this case is Silicon. The depositions of different layers other than the Al and polyimide were done by a RF magnetron sputtering system equipped with a turbo pump and a three-inch target holder. Prior to sputtering, the process chamber was evacuated to 310.sup.7 Torr by the turbo pump. Sputtering was done at 3 mTorr pressure. An Ar flow of 54 SCCM was used in every sputtering process other than deposition. In fabricating the Ge.sub.xSn.sub.yO.sub.z bolometer, lift off technique was used for patterning all the films because of its simplicity where 39x54, 09y20, 19z46.

    (41) The fabrication of the bolometer starts by depositing 400 nm of silicon nitride 111 on a cleaned lightly doped p-type, three inch diameter silicon wafer 101. This layer of silicon nitride 111 serves as the electrical insulation for the substrate and would withstand all the solvents used in next fabrication steps. Then a 400-nm-thick layer of Al layer 141 was deposited by thermal evaporation and patterned. The cryo pump of the evaporator was cooled down to 20 K after which the sample was mounted. The chamber was evacuated to 110.sup.6 Torr prior to evaporation. A deposition rate of 5 angstrom/sec was achieved at 100 Ampere of applied current. To perform the lift off, negative resist NR7-1500P from Futurrex Corporation was spin coated on the wafer at 3000 rpm for 30 seconds. Then the wafer was pre baked at 150 C. for 60 seconds on a hot plate and exposed under the ultraviolet light. A post exposure bake for 60 seconds was done at 120 C. on a hot plate. Then the resist was developed in RD6, a negative resist developer from Futurrex Corporation, for 50 seconds. The resist thickness was found to be about 1.8 m after developing. This was found to be thick enough to lift off 0.4-m-thick Al layer 141. After depositing Al film, the wafer was kept in 1165 photo resist striper for about two hours to complete the lift off process. The wafer was then rinsed with acetone, methanol, and DI water followed by a blow dry in nitrogen to make it clean. This Al layer 141 would serve as a mirror for reflecting the infrared rays and form the basis of an optically resonant cavity.

    (42) A sacrificial layer 151 of photo definable polyimide PI-2737 from HD Microsystems was spin coated, patterned by conventional photolithography and wet etching process. The polyimide was cured in the convection oven. After curing the polyimide thickness was found 2.2 For patterning, the polyimide was spin coated at a speed of 1650 rpm for 60 seconds. Then it was baked on two step hot plate. In first step, it was baked at 70 C., while in second step it was baked at 100 C., both for three minutes. Then it was exposed in ultraviolet light. For developing polyimide, developer DE 9040 along with rinse solution RI 9180, both from HD Microsystem, were used. To store these solutions and complete developing process, four tanks and one squeeze bottle were used. First two of the four tanks were filled with 100% DE 9040 solution, third tank was filled with 50% DE 9040 and 50% RI 9180 solutions, fourth tank was filled with 25% DE 9040 and 75% RI 9180 solutions, while the squeeze bottle was filled with 100% RI 9180 solution. The wafer was kept inside the first two tanks for 10 seconds and agitated ultrasonically. Then it was transferred to the second and third tank respectively where it was kept for 15 seconds in each of them. Then the wafer was rinsed with RI 9180 solution for 20 seconds, by holding it vertically to remove all the polyimide flakes. By using four tanks instead of one tank, the problem of flakes generation was solved. The polyimide thickness at that point was about 5.3 m. This was then cured in an oven at 250 C. for four hours in nitrogen ambient. The temperature was ramped slowly (from room temperature to 250 C. in one and half hours) to avoid possible thermal stress in the film.

    (43) To achieve low thermal mass, the sensing layer of Ge.sub.xSn.sub.yO.sub.z 181 was made 0.6579 m-thick. In order to make the sensing layer mechanically strong as well as free of warping, a sandwich structure of Ge.sub.xSi.sub.ySn.sub.1(x+y)O.sub.z was made employing silicon nitride where, 39x54, 09y20, 19z46. The sandwich layers increase the thermal mass of the detector. However, they were found to be necessary, since the first type of the bolometers were fabricated without silicon nitride sandwich layers and they were found to be warped after removing the sacrificial polyimide. To achieve the sandwich structure, first the bottom silicon nitride layer 161 was deposited and patterned. The thickness of this layer was set to 100 nm. Silicon nitride was chosen for this case, because silicon nitride is known to passivate silicon dioxide, although in current work it was not observed any significant effect of passivating the Ge.sub.xSi.sub.ySn.sub.1(x+y)O.sub.z sensing layer 181 with silicon nitride where 39x54, 09y20, 19z46. Next, a 200-nm-thick NiCr (20% Ni, 80% Cr) electrode arm 121 was deposited and patterned. NiCr has very low thermal conductivity, and thus provides good thermal isolation between bolometer thermometer and substrate.

    (44) To form an Ohmic contact with p-type Ge.sub.xSn.sub.yO.sub.z, a 50-nm-thick Ni film 171 was deposited on top of NiCr arm and patterned where 39x54, 09y20, 19z46. Then, the sensing layer of Ge.sub.xSn.sub.yO.sub.z was deposited in an Ar:O.sub.2 environment from a compound target of Ge and Sn where 39x54, 09y20, 19z46. Next, a 14-nm-thick NiCr absorber 191 and 100-nm-thick silicon nitride 201 layers were deposited. This silicon nitride layer 201 works as the top layer of the sandwich structure. These three layers were lifted of together. Finally, a 300-nm-thick Ni bond-pad-layer 131 was deposited on top of NiCr for the simplicity of bonding the device ultrasonically. At this point, the bolometer fabrication was completed. The sacrificial layer of polyimide under the bolometer was not removed for the sake of simplicity in device fabrication.