Joseph Talghader obtained his B.S. in electrical engineering from Rice University. He was awarded an NSF Graduate Fellowship and attended the University of California at Berkeley where he received his M.S. in 1993 and Ph.D. in 1995. He worked at Texas Instruments as a Process Development Engineer, where he investigated EEPROM memory design and reliability issues. After graduating from Berkeley in 1995, he joined Waferscale Integration where he developed microfabrication processes for high-density nonvolatile memory devices. In 1997 Dr. Talghader joined the faculty at the University of Minnesota as an Assistant Professor and was later promoted to Associate and then Full Professor. He has been extensively involved in thermal infrared and radiation heat transfer devices, optical coatings materials science, and the miniaturization of micro-opto-mechanical systems. His group has demonstrated the highest sensitivity uncooled thermal detectors and the first tunable multispectral thermal detectors. He has several patents and patent applications in these areas. His group works in a number of different fields including recent research on the measurement of the optical and materials properties of glacial ice. Dr. Talghader has received 3M Faculty Awards on three occasions. His technology has been a Finalist for the Minnesota Cup for entrepreneurs. He has served on various program committees and reviews, including service on the triennial strategic planning panel for the Army Research Office Electronics Division, twice as Program chair of the IEEE/LEOS Optical Microelectromechanical Systems Conference, and as Guest Editor of the IEEE Journal of Selected Topics in Quantum Electronics Special Issue on Optical Microsystems. He was Conference chair of the IEEE/LEOS Optical MEMS and Nanophotonics Conference and served as Lead Editor of the JSTQE Special Issue on Optical Micro and Nanosystems. He currently an Editor of the NPG journal Light: Science and Applications.
Abstract The optical absorption of contaminants on high reflectivity mirrors was measured using photo thermal common-path interferometry before and after exposure to high power continuous-wave laser light. The contaminants were micron-sized graphite flakes on hafnia-silica distributed Bragg reflectors illuminated by a ytterbium-doped fiber laser. After one-second periods of exposure, the mirrors demonstrated reduced absorption for irradiances as low as 11 kW cm−2 and had an obvious threshold near 20 kW cm−2. Final absorption values were reduced by up to 90% of their initial value for irradiances of 92 kW cm−2. For shorter pulses at 34 kW cm−2, a minimum exposure time required to begin absorption reduction was found between 100 μs and 200 μs, with particles reaching their final minimum absorption value within 300 ms. Microscope images of the surface showed agglomerated particles fragmenting with some being removed completely, probably by evaporation for exposures between 200 μs to 10 ms. Exposures of 100 ms and longer left behind a thin semi-transparent residue, covering much of the conditioned area. An order of magnitude estimate of the time necessary to begin altering the surface contaminants (also known as ”conditioning”) indicates about 200 μs seconds at 34 kW cm−2, based on heating an average carbon particle to its sublimation temperature including energy loss to thermal contact and radiation. This estimation is close to the observed exposure time required to begin absorption reduction.
Solar selective coatings that absorb solar incident light but that inhibit emission of thermal light can have a dramatic impact on the performance of Direct Steam Generation (DSG) systems. In prior art, perfect coatings with instantaneous transitions between high absorption and low absorption have been modeled. In this paper non-ideal spectral, environmental, and material properties of solar selective coatings are shown to have significant effects on the efficiencies and optimum transition wavelengths of real systems. By introducing more real world parameters into the coating optimization it is desired that a more cost effective and efficient system can be built. It is shown that using ideal conditions the optimum transition wavelength for a DSG system is 1.4μm with an efficiency of 55.7%. Whereas for a realistic non-ideal DSG system the optimum transition is at 3.4μm resulting in an efficiency around 30% depending on the concentration factor used. It is then shown that an optimized selective coating will be outperformed by a simple non-selective black absorber with 95% absorption at concentration factors above 80 and above 130 when the AM1.5 spectrum is used.
Many current quantum optical systems, such as microcavities, interact with thermal light through a small number of widely separated modes. Previous theories for photon number fluctuations of thermal light have been primarily limited to special cases that are appropriate for large volumes or distances, such as single modes, many modes, or modes of uniform spectral distribution. Herein, a theory for the general case of spectrally dependent photon number fluctuations is developed for thermal light. The error in variance of prior art is quantitatively derived for an example cavity in the case where photon counting noise dominates. A method to reduce the spectral impact of this variance is described.
A subsampling technique for real-time phase retrieval of high-speed thermal signals is demonstrated with heated metal lines such as those found in microelectronic interconnects. The thermal signals were produced by applying a current through aluminum resistors deposited on soda–lime–silica glass, and the resulting refractive index changes were measured using a Mach–Zehnder interferometer with a microscope objective and high-speed camera. The temperatures of the resistors were measured both by the phase-retrieval method and by monitoring the resistance of the aluminum lines. The method used to analyze the phase is at least 60× faster than the state of the art but it maintains a small spatial phase noise of 16 nm, remaining comparable to the state of the art. For slowly varying signals, the system is able to perform absolute phase measurements over time, distinguishing temperature changes as small as 2 K. With angular scanning or structured illumination improvements, the system could also perform fast thermal tomography.
Spatial modulation spectroscopy (SMS) is a powerful method for interrogating single nanoparticles. In these experiments optical extinction is measured by moving the particle in and out of a tightly focused laser beam. SMS is typically used for particles that are much smaller than the laser spot size. In this paper, we extend the analysis of the SMS signal to particles with sizes comparable to or larger than the laser spot, where the shape of the particle matters. These results are important for the analysis of polydisperse samples that have a wide range of sizes. The presented example images and analysis of a carbon microparticle sample show the utility of the derived expressions. In particular, we show that SMS can be used to generate extinction cross-section information about micrometer-sized particles with complex shapes.
High temperature microheaters have been designed and constructed to reduce the background thermal emission radiation produced by the heater. Such heaters allow one to probe luminescence with very low numbers of photons where the background emission would overwhelm the desired signal. Two methods to reduce background emission are described: one with low emission materials and the other with interference coating design. The first uses platforms composed of material that is transparent to mid-infrared light and therefore of low emissivity. Heating elements are embedded in the periphery of the heater. The transparent platform is composed of aluminum oxide, which is largely transparent for wavelengths less than about 8 μ m. In the luminescent microscopy used to test the heater, an optical aperture blocks emission from the heating coils while passing light from the heated objects on the transparent center of the microheater. The amount of infrared light transmitted through the aperture was reduced by 90% as the aperture was moved from the highly emissive heater coils at 450 \,^∘C to the largely transparent center at the same temperature. The second method uses microheaters with integrated multilayer interference structures designed to limit background emission in the spectral range of the low-light luminescence object being measured. These heaters were composed of aluminum oxide, titanium dioxide, and platinum and were operated over a large range of temperatures, from 50 \,^∘C to 600 \,^∘C. At 600 \,^∘C, they showed a background photon emission only 1/800 that of a comparison heater without the multilayer interference structure. In this structure, the radiation background was sufficiently reduced to easily monitor weak thermoluminescent emission from CaSO 4 :Ce,Tb microparticles.
While there are innumerable devices that measure temperature, the nonvolatile measurement of thermal history is far more difficult, particularly for sensors embedded in extreme environments such as fires and explosions. In this review, an extensive analysis is given of one such technology: thermoluminescent microparticles. These are transparent dielectrics with a large distribution of trap states that can store charge carriers over very long periods of time. In their simplest form, the population of these traps is dictated by an Arrhenius expression, which is highly dependent on temperature. A particle with filled traps that is exposed to high temperatures over a short period of time will preferentially lose carriers in shallow traps. This depopulation leaves a signature on the particle luminescence, which can be used to determine the temperature and time of the thermal event. Particles are prepared—many months in advance of a test, if desired—by exposure to deep ultraviolet, X-ray, beta, or gamma radiation, which fills the traps with charge carriers. Luminescence can be extracted from one or more particles regardless of whether or not they are embedded in debris or other inert materials. Testing and analysis of the method is demonstrated using laboratory experiments with microheaters and high energy explosives in the field. It is shown that the thermoluminescent materials LiF:Mg,Ti, MgB4O7:Dy,Li, and CaSO4:Ce,Tb, among others, provide accurate measurements of temperature in the 200 to 500 \,^∘C range in a variety of high-explosive environments.
A new instrument for high-resolution optical logging has been built and tested in Antarctica. Its purpose is to obtain records of volcanic products and other scattering features, such as bubbles and impurities, preserved in polar ice sheets, and it achieves this by using long wavelength near-infrared light that is absorbed by the ice before many scattering events occur. Longer wavelengths ensure that the return signal is composed primarily of a single or few backscattering event(s) that limit its spatial spread. The compact optical logger features no components on its body that draw power, which minimizes its size and weight. A prototype of the logger was built and tested at Siple Dome A borehole, and the results were correlated with prior optical logging profiles and records of volcanic products from collected ice core samples.
The present work describes the procedures and results from the first temperature measurements in closed chamber detonations obtained using the thermoluminescence (TL) of particles specifically developed for temperature sensing. Li2B4O7:Ag,Cu (LBO), MgB4O7:Dy,Li (MBO) and CaSO4:Ce,Tb (CSO) were tested separately in a total of 12 independent detonations using a closed detonation chamber at the Naval Surface Warfare Center, Indian Head Explosive Ordnance Disposal Technology Division (NSWC IHEODTD). Detonations were carried out using two different explosives: a high temperature plastic bonded explosive (HPBX) and a low temperature plastic bonded explosive (LPBX). The LPBX and HPBX charges produced temperatures experienced by the TL particles to be between ~550–670 K and ~700–780 K, respectively, depending on the shot. The measured temperatures were reproducible and typically higher than the thermocouple temperatures. These tests demonstrate the survivability of the TL materials and the ability to obtain temperature estimates in realistic conditions, indicating that TL may represent a reliable way of estimating the temperature experienced by free-flowing particles inside an opaque post-detonation fireball.
Thermal effects in optical substrates are vitally important in determining laser damage resistance in long-pulse and continuous-wave laser systems. Thermal deformation waves in a soda-lime-silica glass substrate have been measured using high-speed interferometry during a series of laser pulses incident on the surface. Two-dimensional images of the thermal waves were captured at a rate of up to six frames per thermal event using a quantitative phase measurement method. The system comprised a Mach–Zehnder interferometer, along with a high-speed camera capable of up to 20,000 frames-per-second. The sample was placed in the interferometer and irradiated with 100 ns, 2 kHz 𝑄-switched pulses from a high-power Nd:YAG laser operating at 1064 nm. Phase measurements were converted to temperature using known values of thermal expansion and temperature-dependent refractive index for glass. The thermal decay at the center of the thermal wave was fit to a function derived from first principles with excellent agreement. Additionally, the spread of the thermal distribution over time was fit to the same function. Both the temporal decay fit and the spatial fit produced a thermal diffusivity of 5×10E-7 m^2/s.
Incorporation of bismuth telluride/antimony telluride co-sputtered thermoelectric junctions into MEMS devices requires process developments for patterning and encapsulation as well as characterization of properties such as film stress and contact resistance. Test structures are presented for measuring important thermoelectric properties, resistivity, thermal conductivity, carrier concentration, and Seebeck coefficient. A fabrication process is presented that allows the junctions to be deposited, patterned, encapsulated, and etch released. Measurement of the thermoelectric junctions reveals a room temperature figure of merit, Z_T, of 0.43 with a total Seebeck coefficient difference of 150 μV/K, resistivities of 17.4 and 7.6 μΩ-m, and thermal conductivity of 0.34 and 0.30 W/mK for antimony telluride and bismuth telluride, respectively. The junctions have been incorporated into state of the art uncooled thermopile infrared detectors with a peak detectivity of 3 × 10^9 cm*Hz^(1/2)/W.
A model is presented and confirmed experimentally that explains the anomalous behavior observed in continuous wave (CW) excitation of thermally isolated optics. Distributed Bragg Reflector (DBR) high reflective optical thin film coatings of HfO2 and SiO2 were prepared with a very low absorption, about 7 ppm, measured by photothermal common-path interferometry. When illuminated with a 17 kW CW laser for 30 s, the coatings survived peak irradiances of 13 MW/cm2, on 500 μm diameter spot cross sections. The temperature profile of the optical surfaces was measured using a calibrated thermal imaging camera for illuminated spot sizes ranging from 500 μm to 5 mm; about the same peak temperatures were recorded regardless of spot size. This phenomenon is explained by solving the heat equation for an optic of finite dimensions and taking into account the non-idealities of the experiment. An analytical result is also derived showing the relationship between millisecond pulse to CW laser operation where (1) the heating is proportional to the laser irradiance (W/m2) for millisecond pulses, (2) the heating is proportional to the beam radius (W/m) for CW, and (3) the heating is proportional to W/m⋅tan−1(t√/m) in the transition region between the two.
Most thermoluminescent materials are created using crystal growth techniques; however, it would be of great utility to identify those few thermoluminescent materials that can be deposited using simpler methods, for example to be compatible with the early portions of a silicon integrated circuit or microelectromechanical fabrication process. In this work, thin films of yttrium oxide with a terbium impurity (Y2O3:Tb) were deposited on silicon wafers by electron beam evaporation. The source for the Y2O3:Tb was made by combining Y2O3 and Tb4O7 powders. The approximate thicknesses of the deposited films were 350 nm. After deposition, the films were annealed at 1100 \,^∘C for 30 s to improve crystallinity. There is a strong correlation between the x-ray diffraction (XRD) peak intensity and the thermoluminescent glow curve intensity. The glow curve displays at least two peaks at 140 \,^∘C and 230 \,^∘C. The emission spectra was measured using successive runs with a monochromator set to a different wavelength for each run. There are two main emission peaks at 490 nm and 540 nm. The terbium impurity concentration of approximately 1 mol% was measured using Rutherford backscattering spectrometry (RBS). The Y2O3:Tb is sensitive to UV, x-ray, and gamma radiation. The luminescent intensity per unit mass of UV irradiated Y2O3:Tb was about 2 times that of x-ray irradiated TLD-100.
On extremely small scales, traditional microcooler performance estimates must be corrected to include losses due to radiation. We present a method for analysis of microcoolers having a significant radiative contribution to their thermal conductance. We have fabricated ultrasmall microcoolers from sputtered Bi2Te3/Sb2Te3 thermoelectric junctions with cooling volumes of 200 μm × 200 μm × 65 nm, which we believe to be the smallest microcoolers ever made. The devices are highly thermally isolated with total thermal conductance under 5 × 10−7 W/K in vacuum at room temperature. By fitting the temperature response to input power of the devices in vacuum, we have quantified the nonlinearity of the response to calculate the radiative and film contributions to the total thermal conductance of the device. Three device geometries are presented, with radiative contributions to thermal conductance of 15%, 26%, and 100% depending on their emissive area and support structure. The cooling capabilities of these devices are also measured with maximum cooling of 3.1 K for the 15% radiation-limited device and 2.6 K for the 26% radiation-limited device, with power consumptions below 5 μW.
The laser-damage thresholds of single material and nanolaminate thin films were compared under continuous-wave (CW) illumination conditions. Nanolaminate films consist of uniform material interrupted by the periodic insertion of one or more atomic layers of an alternative material. Hafnia and titania were used as the base materials, and the films were deposited using atomic-layer deposition. The nanolaminates were less polycrystalline than the uniform films, as quantified using x-ray diffraction. It was found that the nanolaminate films had reduced laser-damage thresholds on smooth and patterned substrates as compared to uniform single-material films. This behavior is unusual as prior art indicates that amorphous (less polycrystalline) materials have higher laser-damage thresholds under short-pulse excitation. It is speculated that this may indicate that local thermal conduction affects breakdown more strongly under CW excitation than the dielectric properties that are important for short-pulse excitation.
A model is presented and confirmed experimentally that explains the anomalous behavior observed in continuous wave (CW) excitation of thermally isolated optics. Distributed Bragg Reflector (DBR) high reflective optical thin film coatings of HfO2 and SiO2 were prepared with a very low absorption, about 7 ppm, measured by photothermal common-path interferometry. When illuminated with a 17 kW CW laser for 30 s, the coatings survived peak irradiances of 13 MW/cm2, on 500 μm diameter spot cross sections. The temperature profile of the optical surfaces was measured using a calibrated thermal imaging camera for illuminated spot sizes ranging from 500 μm to 5 mm; about the same peak temperatures were recorded regardless of spot size. This phenomenon is explained by solving the heat equation for an optic of finite dimensions and taking into account the non-idealities of the experiment. An analytical result is also derived showing the relationship between millisecond pulse to CW laser operation where (1) the heating is proportional to the laser irradiance (W/m2) for millisecond pulses, (2) the heating is proportional to the beam radius (W/m) for CW, and (3) the heating is proportional to W/m⋅tan−1(t√/m) in the transition region between the two.
Strong correlations have been found in the polarization of light transmitted through a polycrystalline material and the grain sizes and orientations of that material. Experiments and supporting simulations with irregularly shaped single quartz crystals show that linear polarization is lost more rapidly as grain sizes decrease and the angular spread of the crystal orientations increase. A quantitative method using Stokes matrices to predict such changes is described and experimentally verified using an apparatus to vary the orientation of irregular quartz crystals. Grain sizes are varied between 1mm and 4mm, and the angular spreads in the crystal orientation are varied between 9\,^∘ and 27\,^∘. This technique has applications to identify changes in crystal structure of transparent uniaxial polycrystalline materials, especially in the nondestructive characterization of glacial ice.
Micromachined structures that repeatedly contact their substrate must be actuated in such a way to minimize stiction. A study of electrostatic actuation for adaptive thermal detectors has been performed that shows only a few tenths of a volt can separate normal actuation, snap-down without stiction, and snap-down with stiction. The applied voltage needed to achieve snap-down without stiction is largely constant if the pulse width exceeds the mechanical response time, but increases dramatically if the pulse is applied for a shorter time.
A continuously tunable vertical actuator with subnanometer resolution is presented. It consists of a heterostructure cantilever which has collapsed over a 125 nm thick nanogap. Its operating principle relies on the temperature dependence of the adhesion energy between two InGaAs surface quantum wells. Deflections from −17 to 5 nm with a precision better than three atomic layers have been measured.
The adhesion energy is measured between InGaAsquantum wells that have collapsed across a 125 nm air gap in an InP/InGaAs heterostructure. The method relies on measuring the unadhered length and shape of collapsed microcantilevers with optical interferometry. The adhesion energy is found to be 72 \pm 16 mJ m^-2 . Since the air gap is much smaller than has been measured previously, the influence of van der Waals forces across the gap was included in theoretical modeling. It was found that the forces should not cause significant deviation from the standard adhesion models unless the adhesion energy drops below 25 mJ m^-2 .
The thermal contact conductance (TCC) of microactuated mechanical interfaces has been characterized using an electronic technique, where micromachined test structures were heated with a current and the TCC was inferred from the change in resistance. For every device tested, the TCC was higher in vacuum than in air. This is in stark contrast to the behavior of bulk interfaces, and several experiments suggest that it may be the result of a decreased solid–solid contact area in air caused by the pressure of the interstitial gas. The average effective TCC of a polysilicon/nitride interface brought together by electrostatic actuation varies about values of 6.0×10 4 W /( K m 2 ) in air and 9.5×10 4 W /( K m 2 ) under vacuum for applied pressures of 1 MPa. These values are significantly higher than commonly reported for nonmetallic materials and probably reflect the very smooth surfaces of depositedthin films.
With the recent interest in adaptive IR imaging,focal plane arrays are desired that can operate linearly over an enormous dynamic range. Unfortunately, large signals can cause thermal detectors to operate at temperatures significantly above their ambient resulting in intensity dependent performance or even device damage. In this letter, the responsivity of microbolometer devices is controlled using the detector and substrate as a simple electrostatic actuator. Microbolometers are demonstrated to switch between states that are over a factor of 50 apart in responsivity. The limits of the switching are theoretically separated by four to five orders of magnitude. In addition, intermediate values of responsivity can be obtained by designing devices in which the support beams snap down at lower voltage than the detector plate. Combining this idea with the pressure dependence of the thermal contact conductance, continuous thermal conductance tuning over a factor of 3 is demonstrated.
Free carrier generation due to the presence of absorbing contaminants is considered as a primary mechanism for initiating continuous-wave laser damage in low absorption, high bandgap optical coatings. Thermal, optical, and ion-generation are examined as means of exciting electrons above the bandgap of the material. Once electrons exist above the bandgap, they absorb incident light, causing a runway thermal-absorption process leading to material breakdown. Testing this theory, high reflectivity distributed Bragg reflectors and half-wave coatings were tested with a 17kW 1070nm continuous wave laser in the presence of carbon contamination. Damage thresholds of titania, tantala, hafnia, alumina, and silica were compared to theory and found to follow similar trends. In an effort to prevent laser damage, samples were conditioned using lower irradiance levels, prior to higher exposure. This was found to reduce absorption by up to 90% as well as increase damage thresholds by over an order of magnitude for some samples.
The minimum and maximum laser irradiances that are effective in cleaning etch-released membranes and microstructure was measured for carbon microparticles on LPCVD silicon nitride thin films and suspended platforms. A 1064nm Nd:Yag laser was scanned across the samples to ablate the contaminants. Microscope images of the membranes showed mass removal for an irradiance as low as 600W cm-2. More thorough cleaning was achieved by increasing the irradiance. For 2.5×2.5mm, 205nm thin silicon nitride membranes, 79% of the contaminated area was removed with an exposure of 3.7kW cm-2. Catastrophic damage was seen at a power level of 8.4kW cm-2. Ablation effects were also measured as a change in optical absorption using a photo thermal common-path interferometer. Peak absorption values were decreased from over 100,000ppm to less than 20,000ppm. Silicon nitride platforms were also tested. Despite significant substrate heating, the platforms survived intact up to power levels of 8.4kW cm-2 with near perfect cleaning of carbon particles from their surfaces.
Infrared-transparent microheaters have been constructed to reduce the background blackbody radiation produced by the heater. Among other applications, such heaters allow one to probe the high temperature peaks of thermoluminescent(TL) materials. The microheater consists of peripheral platinum heating elements on a mid-infrared transparent alumina platform. Alumina has a relatively low blackbody signal at high temperature for wavelengths less than 8μm. To test the reduced blackbody emission, an aperture was placed over the heating coils and then the transparent center of the microheater. The amount of infrared light transmitted through the aperture was reduced by 90% as the aperture moved from the highly emissive heater coils at 450\,^∘C to the largely transparent center at the same temperature.
The blackbody radiation limit has traditionally been set forth as the ultimate performance limit of thermal detectors. However, this fundamental limit assumes that the detector absorbs uniformly throughout the thermal spectrum. In much the same way as photon detectors can achieve very high D* because they do not absorb photon energies below their bandgap, so too can thermal detectors except that thermal detectors are not limited to cryogenic operation. In both cases, the enhanced theoretical D* is achieved because the radiation noise is reduced in a device that does not absorb at a uniform high level throughout the thermal emission band. There are multiple ways to achieve high D* in thermal detectors. One is to use materials that absorb only in a certain spectral range, just as in photon detectors. For example a detector made from PbSe, with proper optical coupling, absorbs only photons with wavelengths shorter than 4.9μm. The radiation limited detectivity of such a device can theoretically exceed 9 x 1010cmHz1/2/W in the MWIR. Even with Johnson and 1/f noise estimates included, it can still exceed 2.5x1010cmHz1/2/W in the MWIR. Another technique, applicable for narrowband thermal detectors, is probably even more powerful. Consider a thermal detector that is almost completely transparent. Here, the radiation noise has been reduced but the signal has been reduced even more. However, if the device is now placed inside an optical cavity, then at one wavelength and in one direction, the nearly transparent detector couples to the cavity resonance to absorb at 100%. Radiation from all other wavelengths and directions are rejected by the cavity or are absorbed only weakly by the detector. It is shown that theoretically, the D* of these devices are roughly proportional to the inverse square root of the spectral resonant width under certain conditions. It is also shown that even including Johnson noise and 1/f noise, the practically achievable D* approaches or exceeds 1011 cmHz^1/2/W.
Microbolometers and other thermal detectors have traditionally been limited to seeing objects in a broad wavelength band at a single sensitivity. Recent advances in interface heat transfer and optical cavity design promise to change that. In this paper, we present recent work on thermal infrared detectors with tunable responsivity and wavelength. First, we demonstrate that extended dynamic range in thermal detectors can be achieved by electrostatically bringing a portion of the detector support structure in contact with the substrate. The exact amount of heat transfer can be controlled by adjusting the contact area and pressure. The thermal conductance and responsivity can be switched more than an order of magnitude using this technique. Next, we demonstrate that a wavelength tunable device in the LWIR can be achieved by modifying the structure of a microbolometer to incorporate a modified Gires Tournois optical cavity. The cavity couples light at a single wavelength into the microbolometer while other wavelengths are rejected. We demonstrate that resonance can be tuned from 8.7 to 11.1 μm with applied voltages from 0 to 42 V. The FWHM of the resonance can be switched between around 1.5 μm in a narrow-band mode and 2.83 μm in a broad-band mode.
An electrostatic actuation scheme for creating a soft contact is described and applied to cantilevers and doubly clamped beams. The actuation voltage form is composed of a high voltage pulse with a low DC voltage. The voltage pulse height and duration is mapped for achieving minimum impact force, and the results are well predicted by the simulation. Furthermore, a new phenomenon is observed for clamped beams that pulse longer than the soft-landing requirement bounces the beam back rather than stick.
This paper covers the design, fabrication, and testing of a step-wise tunable long-wavelength infrared filter. The design uses a modified type of microbolometer to filter infrared light in the 7-10 micrometer wavelength range. Fabrication of the device is accomplished using standard microelectronic and microelectromechanical systems (MEMS) fabrication techniques. Germanium and zinc sulfide are used as optical layers for good reflectivity in the desired infrared range. Testing of the device shows that it filters light in three absorption bands in the specified wavelength range. The tuning of the device is found to be step-wise in response to electrostatic actuation using voltages up to 100 volts.
This paper presents a theoretical and experimental investigation of photon diffusion in highly absorbing microscale graphite. A Nd:YAG continuous wave laser is used to heat the graphite samples with thicknesses of 40 μm and 100 μm. Optical intensities of 10 kW cm-2 and 20 kW cm-2 are used in the laser heating. The graphite samples are heated to temperatures of thousands of kelvins within milliseconds, which are recorded by a 2-color, high speed pyrometer. To compare the observed temperatures, differential equation of heat conduction is solved across the samples with proper initial and boundary conditions. In addition to lattice vibrations, photon diffusion is incorporated in the analytical model of thermal conductivity for solving the heat equation. The numerical simulations showed close matching between experiment and theory only when including the photon diffusion equations and existing material properties data found in the previously published works with no fitting constants. The results indicate that the commonly-overlooked mechanism of photon diffusion dominates the heat transfer of many microscale structures near their evaporation temperatures. In addition, the treatment explains the discrepancies between thermal conductivity measurements and theory that were previously described in the scientific literature.
It is shown that an aberrated wavefront incident upon a Fabry-Perot optical cavity excites higher order spatial modes in the cavity and that the spectral width and distribution of these modes is indicative of the type and magnitude of the aberration. The cavities are purely passive, and therefore frequency content is limited to that provided by the original light source. To illustrate this concept, spatial mode decomposition and transmission spectrum calculation are simulated on an example cavity; the effects of various phase delays, in the form of two basic Seidel aberrations and a composite of Zernike polynomial terms, are shown using both Laguerre-Gaussian and plane wave incident beams. The aggregate spectral width of the cavity modes excited by the aberrations is seen to widen as the magnitude of the aberrations/ phase delay increases.
Laser-Induced optical breakdown often occurs unexpectedly at optical intensities far lower than those predicted by ultra-short pulse laser experiments, and is usually attributed to contamination. To determine the physical mechanism, optical coatings were contaminated with carbon and steel microparticles and stressed using a 17 kW continuous-wave laser. Breakdown occurred at intensity levels many orders of magnitude lower than expected in clean, pristine materials. Damage thresholds were found to strongly follow the bandgap energy of the film. A thermal model incorporating the particle absorption, interface heat transfer, and free carrier absorption was developed, and it explains the observed data, indicating that surface contamination heated by the laser thermally generates free carriers in the films. The observed bandgap dependence is in direct contrast to the behavior observed for clean samples under continuous wave and long-pulse illumination, and, unexpectedly, has similarities to ultra-short pulse breakdown for clean samples, albeit with a substantially different physical mechanism.
Typically when microsensor particles are used to monitor temperatures in harsh environments, a statistical number of particles are collected and measured in aggregate. While this method is undoubtedly the most practical method for the rapid acquisition of overall temperature data, the limitations of collective measurements of numbers of particles versus individual ones has never been explored. In this paper, the thermoluminescent (TL) magnesium borate microparticles are used to measure temperatures inside the periphery of explosions with collective and individual behavior contrasted. It is found that individual measurement indicate that some particles undergo extreme temperature while others seem to have had no exposure to high temperature at all. The microparticles were irradiated with 200Gy of gamma radiation to fill the traps in the band gap. Several grams of the irradiated microparticles were placed at various distances from the source of the detonation. After the microparticles were collected the TL curve was measured for microparticles that were in the detonation and a control group of microparticles not in the detonation. The TL curve of an individual microparticle was measured by placing the microparticle ranging in size from 25 μm to 75 μm on a microheater with an area of 300 μm \texttimes 300 μm. The microheater was then used to heat the microparticle at a linear rate while the thermoluminescence of the microparticle was measured. A summation of first-order kinetics curves were used to do a fit to the thermoluminescence curves of the microparticles that were in the detonation and the control group. By comparing the ratio of first-order kinetic curve peaks of the particles that were in the detonation to the control group the temperature that the particles in the explosion were calculated. This process was carried out for many different particles that were in the same detonation and collected from the same location in the explosive test chamber. Individual extracted temperatures from the microparticles show a large distribution ranging from room temperature to 516\textdegreeC, but in aggregate, the microparticles show a clustering of temperatures around 290\textdegreeC.
Abstract The optical absorption of contaminants on high reflectivity mirrors was measured using photo thermal common-path interferometry before and after exposure to high power continuous-wave laser light. The contaminants were micron-sized graphite flakes on hafnia-silica distributed Bragg reflectors illuminated by a ytterbium-doped fiber laser. After one-second periods of exposure, the mirrors demonstrated reduced absorption for irradiances as low as 11 kW cm−2 and had an obvious threshold near 20 kW cm−2. Final absorption values were reduced by up to 90% of their initial value for irradiances of 92 kW cm−2. For shorter pulses at 34 kW cm−2, a minimum exposure time required to begin absorption reduction was found between 100 μs and 200 μs, with particles reaching their final minimum absorption value within 300 ms. Microscope images of the surface showed agglomerated particles fragmenting with some being removed completely, probably by evaporation for exposures between 200 μs to 10 ms. Exposures of 100 ms and longer left behind a thin semi-transparent residue, covering much of the conditioned area. An order of magnitude estimate of the time necessary to begin altering the surface contaminants (also known as ”conditioning”) indicates about 200 μs seconds at 34 kW cm−2, based on heating an average carbon particle to its sublimation temperature including energy loss to thermal contact and radiation. This estimation is close to the observed exposure time required to begin absorption reduction.
Solar selective coatings that absorb solar incident light but that inhibit emission of thermal light can have a dramatic impact on the performance of Direct Steam Generation (DSG) systems. In prior art, perfect coatings with instantaneous transitions between high absorption and low absorption have been modeled. In this paper non-ideal spectral, environmental, and material properties of solar selective coatings are shown to have significant effects on the efficiencies and optimum transition wavelengths of real systems. By introducing more real world parameters into the coating optimization it is desired that a more cost effective and efficient system can be built. It is shown that using ideal conditions the optimum transition wavelength for a DSG system is 1.4μm with an efficiency of 55.7%. Whereas for a realistic non-ideal DSG system the optimum transition is at 3.4μm resulting in an efficiency around 30% depending on the concentration factor used. It is then shown that an optimized selective coating will be outperformed by a simple non-selective black absorber with 95% absorption at concentration factors above 80 and above 130 when the AM1.5 spectrum is used.
Many current quantum optical systems, such as microcavities, interact with thermal light through a small number of widely separated modes. Previous theories for photon number fluctuations of thermal light have been primarily limited to special cases that are appropriate for large volumes or distances, such as single modes, many modes, or modes of uniform spectral distribution. Herein, a theory for the general case of spectrally dependent photon number fluctuations is developed for thermal light. The error in variance of prior art is quantitatively derived for an example cavity in the case where photon counting noise dominates. A method to reduce the spectral impact of this variance is described.
A subsampling technique for real-time phase retrieval of high-speed thermal signals is demonstrated with heated metal lines such as those found in microelectronic interconnects. The thermal signals were produced by applying a current through aluminum resistors deposited on soda–lime–silica glass, and the resulting refractive index changes were measured using a Mach–Zehnder interferometer with a microscope objective and high-speed camera. The temperatures of the resistors were measured both by the phase-retrieval method and by monitoring the resistance of the aluminum lines. The method used to analyze the phase is at least 60× faster than the state of the art but it maintains a small spatial phase noise of 16 nm, remaining comparable to the state of the art. For slowly varying signals, the system is able to perform absolute phase measurements over time, distinguishing temperature changes as small as 2 K. With angular scanning or structured illumination improvements, the system could also perform fast thermal tomography.
Spatial modulation spectroscopy (SMS) is a powerful method for interrogating single nanoparticles. In these experiments optical extinction is measured by moving the particle in and out of a tightly focused laser beam. SMS is typically used for particles that are much smaller than the laser spot size. In this paper, we extend the analysis of the SMS signal to particles with sizes comparable to or larger than the laser spot, where the shape of the particle matters. These results are important for the analysis of polydisperse samples that have a wide range of sizes. The presented example images and analysis of a carbon microparticle sample show the utility of the derived expressions. In particular, we show that SMS can be used to generate extinction cross-section information about micrometer-sized particles with complex shapes.
High temperature microheaters have been designed and constructed to reduce the background thermal emission radiation produced by the heater. Such heaters allow one to probe luminescence with very low numbers of photons where the background emission would overwhelm the desired signal. Two methods to reduce background emission are described: one with low emission materials and the other with interference coating design. The first uses platforms composed of material that is transparent to mid-infrared light and therefore of low emissivity. Heating elements are embedded in the periphery of the heater. The transparent platform is composed of aluminum oxide, which is largely transparent for wavelengths less than about 8 μ m. In the luminescent microscopy used to test the heater, an optical aperture blocks emission from the heating coils while passing light from the heated objects on the transparent center of the microheater. The amount of infrared light transmitted through the aperture was reduced by 90% as the aperture was moved from the highly emissive heater coils at 450 \,^∘C to the largely transparent center at the same temperature. The second method uses microheaters with integrated multilayer interference structures designed to limit background emission in the spectral range of the low-light luminescence object being measured. These heaters were composed of aluminum oxide, titanium dioxide, and platinum and were operated over a large range of temperatures, from 50 \,^∘C to 600 \,^∘C. At 600 \,^∘C, they showed a background photon emission only 1/800 that of a comparison heater without the multilayer interference structure. In this structure, the radiation background was sufficiently reduced to easily monitor weak thermoluminescent emission from CaSO 4 :Ce,Tb microparticles.
A new instrument for high-resolution optical logging has been built and tested in Antarctica. Its purpose is to obtain records of volcanic products and other scattering features, such as bubbles and impurities, preserved in polar ice sheets, and it achieves this by using long wavelength near-infrared light that is absorbed by the ice before many scattering events occur. Longer wavelengths ensure that the return signal is composed primarily of a single or few backscattering event(s) that limit its spatial spread. The compact optical logger features no components on its body that draw power, which minimizes its size and weight. A prototype of the logger was built and tested at Siple Dome A borehole, and the results were correlated with prior optical logging profiles and records of volcanic products from collected ice core samples.
While there are innumerable devices that measure temperature, the nonvolatile measurement of thermal history is far more difficult, particularly for sensors embedded in extreme environments such as fires and explosions. In this review, an extensive analysis is given of one such technology: thermoluminescent microparticles. These are transparent dielectrics with a large distribution of trap states that can store charge carriers over very long periods of time. In their simplest form, the population of these traps is dictated by an Arrhenius expression, which is highly dependent on temperature. A particle with filled traps that is exposed to high temperatures over a short period of time will preferentially lose carriers in shallow traps. This depopulation leaves a signature on the particle luminescence, which can be used to determine the temperature and time of the thermal event. Particles are prepared—many months in advance of a test, if desired—by exposure to deep ultraviolet, X-ray, beta, or gamma radiation, which fills the traps with charge carriers. Luminescence can be extracted from one or more particles regardless of whether or not they are embedded in debris or other inert materials. Testing and analysis of the method is demonstrated using laboratory experiments with microheaters and high energy explosives in the field. It is shown that the thermoluminescent materials LiF:Mg,Ti, MgB4O7:Dy,Li, and CaSO4:Ce,Tb, among others, provide accurate measurements of temperature in the 200 to 500 \,^∘C range in a variety of high-explosive environments.
The present work describes the procedures and results from the first temperature measurements in closed chamber detonations obtained using the thermoluminescence (TL) of particles specifically developed for temperature sensing. Li2B4O7:Ag,Cu (LBO), MgB4O7:Dy,Li (MBO) and CaSO4:Ce,Tb (CSO) were tested separately in a total of 12 independent detonations using a closed detonation chamber at the Naval Surface Warfare Center, Indian Head Explosive Ordnance Disposal Technology Division (NSWC IHEODTD). Detonations were carried out using two different explosives: a high temperature plastic bonded explosive (HPBX) and a low temperature plastic bonded explosive (LPBX). The LPBX and HPBX charges produced temperatures experienced by the TL particles to be between ~550–670 K and ~700–780 K, respectively, depending on the shot. The measured temperatures were reproducible and typically higher than the thermocouple temperatures. These tests demonstrate the survivability of the TL materials and the ability to obtain temperature estimates in realistic conditions, indicating that TL may represent a reliable way of estimating the temperature experienced by free-flowing particles inside an opaque post-detonation fireball.
Thermal effects in optical substrates are vitally important in determining laser damage resistance in long-pulse and continuous-wave laser systems. Thermal deformation waves in a soda-lime-silica glass substrate have been measured using high-speed interferometry during a series of laser pulses incident on the surface. Two-dimensional images of the thermal waves were captured at a rate of up to six frames per thermal event using a quantitative phase measurement method. The system comprised a Mach–Zehnder interferometer, along with a high-speed camera capable of up to 20,000 frames-per-second. The sample was placed in the interferometer and irradiated with 100 ns, 2 kHz 𝑄-switched pulses from a high-power Nd:YAG laser operating at 1064 nm. Phase measurements were converted to temperature using known values of thermal expansion and temperature-dependent refractive index for glass. The thermal decay at the center of the thermal wave was fit to a function derived from first principles with excellent agreement. Additionally, the spread of the thermal distribution over time was fit to the same function. Both the temporal decay fit and the spatial fit produced a thermal diffusivity of 5×10E-7 m^2/s.
We have derived an expression for the effective absorbing area for thermal infrared detectors having non-zero absorption in the support legs, which is different from the geometric areas of the constituent detector elements. This technique is particularly applicable to devices where sensitivity is more important than fill-factor, as opposed to standard imaging arrays. The effective area can simply be substituted in standard equations to obtain a good estimate of the detector performance under uniform flood illumination conditions. The formalism can also be used for estimating the contributions of the individual signal generating elements to the total measured signal. This approximation has been tested for MEMS infrared detectors with thermoelectric readout operating under vacuum. The responsivity of the same device calculated using the effective area approximation and measured using a tightly constrained absorbing area are found to match very closely, within 5% over the most wavelengths and within 15% at the shortest thermal infrared wavelengths.
A model is presented and confirmed experimentally that explains the anomalous behavior observed in continuous wave (CW) excitation of thermally isolated optics. Distributed Bragg Reflector (DBR) high reflective optical thin film coatings of HfO2 and SiO2 were prepared with a very low absorption, about 7 ppm, measured by photothermal common-path interferometry. When illuminated with a 17 kW CW laser for 30 s, the coatings survived peak irradiances of 13 MW/cm2, on 500 μm diameter spot cross sections. The temperature profile of the optical surfaces was measured using a calibrated thermal imaging camera for illuminated spot sizes ranging from 500 μm to 5 mm; about the same peak temperatures were recorded regardless of spot size. This phenomenon is explained by solving the heat equation for an optic of finite dimensions and taking into account the non-idealities of the experiment. An analytical result is also derived showing the relationship between millisecond pulse to CW laser operation where (1) the heating is proportional to the laser irradiance (W/m2) for millisecond pulses, (2) the heating is proportional to the beam radius (W/m) for CW, and (3) the heating is proportional to W/m⋅tan−1(t√/m) in the transition region between the two.
Most thermoluminescent materials are created using crystal growth techniques; however, it would be of great utility to identify those few thermoluminescent materials that can be deposited using simpler methods, for example to be compatible with the early portions of a silicon integrated circuit or microelectromechanical fabrication process. In this work, thin films of yttrium oxide with a terbium impurity (Y2O3:Tb) were deposited on silicon wafers by electron beam evaporation. The source for the Y2O3:Tb was made by combining Y2O3 and Tb4O7 powders. The approximate thicknesses of the deposited films were 350 nm. After deposition, the films were annealed at 1100 \,^∘C for 30 s to improve crystallinity. There is a strong correlation between the x-ray diffraction (XRD) peak intensity and the thermoluminescent glow curve intensity. The glow curve displays at least two peaks at 140 \,^∘C and 230 \,^∘C. The emission spectra was measured using successive runs with a monochromator set to a different wavelength for each run. There are two main emission peaks at 490 nm and 540 nm. The terbium impurity concentration of approximately 1 mol% was measured using Rutherford backscattering spectrometry (RBS). The Y2O3:Tb is sensitive to UV, x-ray, and gamma radiation. The luminescent intensity per unit mass of UV irradiated Y2O3:Tb was about 2 times that of x-ray irradiated TLD-100.
Incorporation of bismuth telluride/antimony telluride co-sputtered thermoelectric junctions into MEMS devices requires process developments for patterning and encapsulation as well as characterization of properties such as film stress and contact resistance. Test structures are presented for measuring important thermoelectric properties, resistivity, thermal conductivity, carrier concentration, and Seebeck coefficient. A fabrication process is presented that allows the junctions to be deposited, patterned, encapsulated, and etch released. Measurement of the thermoelectric junctions reveals a room temperature figure of merit, Z_T, of 0.43 with a total Seebeck coefficient difference of 150 μV/K, resistivities of 17.4 and 7.6 μΩ-m, and thermal conductivity of 0.34 and 0.30 W/mK for antimony telluride and bismuth telluride, respectively. The junctions have been incorporated into state of the art uncooled thermopile infrared detectors with a peak detectivity of 3 × 10^9 cm*Hz^(1/2)/W.
We describe uncooled thermal detectors with a peak detectivity of at least 3 \texttimes109 cm \surdHz/W with spectrally selective absorption in the long-wave infrared. The spectral selectivity in absorption is achieved through resonant cavity coupling of a thin metal film with a low-order air-gap optical cavity. The electrical readout uses thermoelectric thin films with a Johnson noise limited performance. The detectors are of multiple sizes but those with 100- \mum2 area have time constants of 58 ms and thermal conductances of 2.3 \texttimes10-7 W/K.
Vertical electrostatic wedge actuators are described that control nanometer-scale gaps between surfaces. Standard parallel-plate electrostatic actuators become difficult to stabilize across extremely small gaps because the nature of the forces and the force laws that describe them often deviate from a Coulomb’s law dependence. In this work, a nanometer-scale air gap between a collapsed cantilever structure formed by two facing In0.53Ga0.47As surfaces, with areas of tens of microns, was controlled by a wedge electrostatic actuator. Upon actuation, the gap spacing between the surfaces was tuned over a maximum range of 55 nm with an applied voltage of 60 V.
On extremely small scales, traditional microcooler performance estimates must be corrected to include losses due to radiation. We present a method for analysis of microcoolers having a significant radiative contribution to their thermal conductance. We have fabricated ultrasmall microcoolers from sputtered Bi2Te3/Sb2Te3 thermoelectric junctions with cooling volumes of 200 μm × 200 μm × 65 nm, which we believe to be the smallest microcoolers ever made. The devices are highly thermally isolated with total thermal conductance under 5 × 10−7 W/K in vacuum at room temperature. By fitting the temperature response to input power of the devices in vacuum, we have quantified the nonlinearity of the response to calculate the radiative and film contributions to the total thermal conductance of the device. Three device geometries are presented, with radiative contributions to thermal conductance of 15%, 26%, and 100% depending on their emissive area and support structure. The cooling capabilities of these devices are also measured with maximum cooling of 3.1 K for the 15% radiation-limited device and 2.6 K for the 26% radiation-limited device, with power consumptions below 5 μW.
The laser-damage thresholds of single material and nanolaminate thin films were compared under continuous-wave (CW) illumination conditions. Nanolaminate films consist of uniform material interrupted by the periodic insertion of one or more atomic layers of an alternative material. Hafnia and titania were used as the base materials, and the films were deposited using atomic-layer deposition. The nanolaminates were less polycrystalline than the uniform films, as quantified using x-ray diffraction. It was found that the nanolaminate films had reduced laser-damage thresholds on smooth and patterned substrates as compared to uniform single-material films. This behavior is unusual as prior art indicates that amorphous (less polycrystalline) materials have higher laser-damage thresholds under short-pulse excitation. It is speculated that this may indicate that local thermal conduction affects breakdown more strongly under CW excitation than the dielectric properties that are important for short-pulse excitation.
A review is made of the physics and technology of spectrally selective thermal detectors, especially those operating at non-cryogenic temperatures. The background radiation noise fluctuations are rederived for arbitrary spectral characteristics. Infrared absorption due to phonons and free carriers is discussed followed by a review of published works on artificial infrared absorption materials such as patterned grids, nanoparticles, plasmonic structures, metamaterials and others. Subsequently, the literature of the spectral characteristics of broadband thermal detectors and spectrally selective thermal detectors is reviewed. Finally, the authors speculate on the directions that future research and development in the area will take regarding architectures, sensitivity and spectral characteristics.
The far field radiation efficiency achievable in narrowband thermal emitters is investigated, taking into account the full spatial and spectral variation of the emissivity. A coupled Fabry-Perot cavity model is used to develop an insight into the efficiency variation with cavity coherence and device temperature. It is found that the spatial variation of emissivity has to be explicitly included in the radiation power calculations to accurately estimate the achievable power efficiencies. The calculated radiation efficiencies of an ideal coherent cavity coupled emitter were found to vary from 0.1% to 9%, with a corresponding increase in the emission linewidth from 6.3 nm to 930 nm, and were much lower than that estimated without accounting for effects of spatial coherence. The analysis presented here can be used to determine the optimal operating temperature of a coherent thermal emitter once its emission characteristics and conduction losses are known and it is demonstrated that this optimum temperature is different from the temperature of peak blackbody emission at the resonant absorption wavelength.
A model is presented and confirmed experimentally that explains the anomalous behavior observed in continuous wave (CW) excitation of thermally isolated optics. Distributed Bragg Reflector (DBR) high reflective optical thin film coatings of HfO2 and SiO2 were prepared with a very low absorption, about 7 ppm, measured by photothermal common-path interferometry. When illuminated with a 17 kW CW laser for 30 s, the coatings survived peak irradiances of 13 MW/cm2, on 500 μm diameter spot cross sections. The temperature profile of the optical surfaces was measured using a calibrated thermal imaging camera for illuminated spot sizes ranging from 500 μm to 5 mm; about the same peak temperatures were recorded regardless of spot size. This phenomenon is explained by solving the heat equation for an optic of finite dimensions and taking into account the non-idealities of the experiment. An analytical result is also derived showing the relationship between millisecond pulse to CW laser operation where (1) the heating is proportional to the laser irradiance (W/m2) for millisecond pulses, (2) the heating is proportional to the beam radius (W/m) for CW, and (3) the heating is proportional to W/m⋅tan−1(t√/m) in the transition region between the two.
Hafnia-alumina nanolaminates show improved smoothness and reduced crystallinity relative to pure hafnia in films formed by atomic layer deposition (ALD). However, typical nanolaminates also show reduced cross-plane thermal conductivity due to the much larger interface density relative to continuous films. We find that the interface thermal resistance in hafnia-alumina nanolaminates is very low and does not dominate the film thermal conductivity, which is 1.0 to 1.2 W/(m K) at room temperature in 100 nm thin films regardless of the interface density. Measured films had a number of interfaces ranging from 2 to 40, equivalent to interface spacing varying from about 40 to 2 nm. The degree of crystallinity of these films appears to have a much larger effect on thermal conductivity than that of interface density. Cryogenic measurements show good agreement with both the minimum thermal conductivity model for disordered solids and the diffuse mismatch model of interface resistance down to about 80 K before diverging. We find that the films are quite smooth through a 400:5 ratio of hafnia to alumina in terms of ALD cycles, and the refractive index scales as expected with increasing alumina concentration.
A measurement technique, pulsed thermoluminescence, is described which uses short thermal pulses to excite trapped carriers leading to radiative recombination. The pulses are obtained using microstructures with \sim500 \mathrmμs thermal time constants. The technique has many of the advantages of pulsed optically stimulated luminescence without the need for optical sources and filters to isolate the luminescent signal. Charge carrier traps in α-Al2O3:C particles on microheaters were filled using 205 nm light. Temperature pulses of 10 and 50 ms were applied to the heaters and compared with a standard thermoluminescence curve taken at a ramp rate of 5 K s-1. This produced curves of intensity verses temperature similar to standard thermoluminescence except shifted to higher temperatures. The luminescence of single particles was read multiple times with negligible loss of population. The lower limit of the duration of useful pulses appears to be limited by particle size and thermal contact between the particle and heater.
Mechanical position is used to control the wavelength of light emission of semiconductor heterostructures. The heterostructures are coupled across a gap that varies with position to tune electron states in much the same manner that optical cavities can be coupled across a tunable reflectivity mirror to control photon states. In the experiments, a SixN/InP cantilever containing an InGaAs surface well collapses over another InGaAs quantum well. The spacing between the wells varies along the cantilever, such that the heterostructure band gap is determined by the mechanical bending of the cantilever. Photoluminescence measurements of the coupled 200 Å surface wells show a wavelength shift of up to 22 nm. Associated theory shows that mechanical quantum coupling enables interband or intersubband devices with unprecedented spectral tuning ranges for gain or absorption.
Strong correlations have been found in the polarization of light transmitted through a polycrystalline material and the grain sizes and orientations of that material. Experiments and supporting simulations with irregularly shaped single quartz crystals show that linear polarization is lost more rapidly as grain sizes decrease and the angular spread of the crystal orientations increase. A quantitative method using Stokes matrices to predict such changes is described and experimentally verified using an apparatus to vary the orientation of irregular quartz crystals. Grain sizes are varied between 1mm and 4mm, and the angular spreads in the crystal orientation are varied between 9\,^∘ and 27\,^∘. This technique has applications to identify changes in crystal structure of transparent uniaxial polycrystalline materials, especially in the nondestructive characterization of glacial ice.
Micromachined structures that repeatedly contact their substrate must be actuated in such a way to minimize stiction. A study of electrostatic actuation for adaptive thermal detectors has been performed that shows only a few tenths of a volt can separate normal actuation, snap-down without stiction, and snap-down with stiction. The applied voltage needed to achieve snap-down without stiction is largely constant if the pulse width exceeds the mechanical response time, but increases dramatically if the pulse is applied for a shorter time.
Optical coating degradation under laser irradiation can take several forms. Perhaps the most common that is not due to particulates is thermal breakdown, caused by heating of the coating to a catastrophic failure induced by local melting, delamination, evaporation, or some other change. We demonstrate that micromachined dielectric membranes show strong differences in their hydroxyl signatures as measured by Fourier-transform IR spectroscopy. The changes correspond to regions of high fluence (3200 J/cm2) from a Nd:YAG laser. It is found that the absorption peaks associated with OH decrease after laser treatment, indicating a reduction in the number of film hydroxyl groups.
The thermomechanical response of electron beam deposited nanoporous silicon dioxide is examined using substrate curvature measurements and nanoindentation. Analysis of the thin film bond angle strain distributions versus temperature indicates that low temperature ( T < 100 ^\circC) stress hysteresis and tensioning are primarily attributed to hydrogen bonded water desorption. However, at higher temperatures, the absence of water desorption suggests that the thermomechanical behaviour is related to thermally induced bond angle strain redistributions towards the local bonding environment of quartz and thermally grown silicon dioxide. This is supported by the co-efficient of thermal expansion data that trend lower with higher annealing temperatures. The re-absorption of water into the thin film accounts for the reproducibility of the open-loop stress hysteresis and tensioning observations.
A continuously tunable vertical actuator with subnanometer resolution is presented. It consists of a heterostructure cantilever which has collapsed over a 125 nm thick nanogap. Its operating principle relies on the temperature dependence of the adhesion energy between two InGaAs surface quantum wells. Deflections from −17 to 5 nm with a precision better than three atomic layers have been measured.
Morphological changes due to adsorbed gases in nanoporous silicon dioxide thin films are demonstrated using in situ x-ray photoelectron spectroscopy at temperatures in the range 20\leqslantT\leqslant300^\circC. Adsorbed hydrogen bonded water vapor is observed to relax the surface bond strain of low-temperature electron-beam deposited silicon dioxide up to 100^\circC. This was determined by measuring the width of the Si 2p and O 1s photoemission peak full widths at half maximum, which are distinctly smaller for films with adsorbed water vapor than for the same films after vapor has been outgassed by heating above 100^\circC. In situ heating in the range 100
Continuous deformable membrane mirrors are becoming more attractive for use in adaptive optics because they cause no diffraction in the reflected beam and ensure smooth and continuous phase variations across the mirrors. However, when such mirrors are used to correct a high-power incident wave front, the absorption in the coatings causes the temperature of the membrane to increase, thereby creating in-plane thermal stress due to the rigidly clamped boundaries. We present a technique to measure thermal stress in such nondeforming membrane structures. The directional stress and temperature effects are simultaneously measured and decoupled in micromachined membrane mirrors by using a group of three ion-implanted silicon resistors with different orientations. In stress measurements made with incident power, the sensors measure changes in compressive thermal stress to within 80-90 kPa.
A remote microfluidic metal-ion sensor is developed using an electrochemical system integrated with a compact photovoltaic cell power supply. The sensor is designed to detect the sum of metal ions in a remote environment. The sensor uses electrodeposition to remove ions from the fluid around the sensor and deposit them on an electrode at the tip of a cantilever. The electrodeposited mass changes the resonant frequency of the cantilever, which can be determined upon read-out. The sensor is designed to be dropped in liquids or flow through microfluidic systems and can be used in parallel with many other similar sensors. The photovoltaic cells are directly integrated on the device and are capable of producing tens of microwatts of power at about 15% efficiency with laser excitation. However, the sensor operates at power levels of 50 nW with small voltages and currents using only scavenged daylight or room light. The complete device is integrated into a total volume below 0.046 mm3, which is more than two orders of magnitude smaller than other remote electrically-powered sensors reported to date. Although it is expected that multiple devices will be used in parallel to gain statistical data, individual particles detect metal-ion concentration within 24% of the actual concentration, making them suitable for safety testing and endpoint monitoring among other applications.
We report the first practical results on design and characterization of adaptive microbolometers with a thermally tuned responsivity. Such devices are needed to simultaneously image scenes that contain objects at ambient and extremely hot temperatures. In the high detectivity state, the microbolometers are operated similar to standard commercial devices. In the low detectivity state, portions of the support beams are brought in contact with the substrate, which thermally shorts the devices on a pixel-by-pixel basis. This is the first time this concept has been practically implemented in a microbolometer device architecture as opposed to simple cantilever beams and plates. The maximum actuation voltage is set to 17 V, and the thermal conductances, responsivities and detectivities of a typical device can be switched more than an order of magnitude between 1.65 × 10−5 W K−1 and 2.99 × 10−4 W K−1, between 1.5 V W−1 and 0.2 V W−1 and between 1.8 × 106 cm Hz1/2 W−1 and 1.5 × 105 cm Hz1/2 W−1, respectively. This extends the dynamic range of the device more than an order of magnitude. The device pixel size is 100 µm ×100 µm.
The adhesion energy is measured between InGaAsquantum wells that have collapsed across a 125 nm air gap in an InP/InGaAs heterostructure. The method relies on measuring the unadhered length and shape of collapsed microcantilevers with optical interferometry. The adhesion energy is found to be 72 \pm 16 mJ m^-2 . Since the air gap is much smaller than has been measured previously, the influence of van der Waals forces across the gap was included in theoretical modeling. It was found that the forces should not cause significant deviation from the standard adhesion models unless the adhesion energy drops below 25 mJ m^-2 .
The performance of optical microcavities is limited by spectral degradation resulting from thermal deformation and fabrication imperfections. In this paper, we study the spatial-mode properties of micromirror optical cavities with respect to commonly seen aberrations. Electrothermal actuation is used to slightly adjust the shape and position of micromirrors and study the effects this has on the spatial-mode structure of the cavity spectrum. The shapes of the micromirrors are changed using Joule heating with thermal expansion deformation. Significant differences in mirror tilt, curvature, and astigmatism are measured, but the tilt has by far the biggest impact on cavity finesse and resolution. We demonstrate that unwanted higher order spatial modes can be suppressed electrically and an amplitude reduction for the higher order modes of over 60% has been obtained with a tuning current of 5.5 mA. A fundamental mode finesse of approximately 60 is maintained throughout tuning. These tunable cavities have great potential in applications using cavity arrays or requiring dynamic mode control.
Aberrated optics degrade the cavity finesse and point spread function of any optical system. It is frequently overlooked that the phase errors of these aberrations cause characteristic spectral features that can be observed and measured by an optical cavity. A general theory of the spectral decomposition of aberrations is developed and applied to an electrically deformable membrane feeding a high-finesse resonator. The micromirror, which consists of a 1 cm metallized SiNxelectrostatically-actuated membrane with phase offsets of about 1.5μm, leads to optical power being distributed to over twenty higher-order Laguerre-Gaussian cavity modes.
The minimum and maximum laser irradiances that are effective in cleaning etch-released membranes and microstructure was measured for carbon microparticles on LPCVD silicon nitride thin films and suspended platforms. A 1064nm Nd:Yag laser was scanned across the samples to ablate the contaminants. Microscope images of the membranes showed mass removal for an irradiance as low as 600W cm-2. More thorough cleaning was achieved by increasing the irradiance. For 2.5×2.5mm, 205nm thin silicon nitride membranes, 79% of the contaminated area was removed with an exposure of 3.7kW cm-2. Catastrophic damage was seen at a power level of 8.4kW cm-2. Ablation effects were also measured as a change in optical absorption using a photo thermal common-path interferometer. Peak absorption values were decreased from over 100,000ppm to less than 20,000ppm. Silicon nitride platforms were also tested. Despite significant substrate heating, the platforms survived intact up to power levels of 8.4kW cm-2 with near perfect cleaning of carbon particles from their surfaces.
Free carrier generation due to the presence of absorbing contaminants is considered as a primary mechanism for initiating continuous-wave laser damage in low absorption, high bandgap optical coatings. Thermal, optical, and ion-generation are examined as means of exciting electrons above the bandgap of the material. Once electrons exist above the bandgap, they absorb incident light, causing a runway thermal-absorption process leading to material breakdown. Testing this theory, high reflectivity distributed Bragg reflectors and half-wave coatings were tested with a 17kW 1070nm continuous wave laser in the presence of carbon contamination. Damage thresholds of titania, tantala, hafnia, alumina, and silica were compared to theory and found to follow similar trends. In an effort to prevent laser damage, samples were conditioned using lower irradiance levels, prior to higher exposure. This was found to reduce absorption by up to 90% as well as increase damage thresholds by over an order of magnitude for some samples.
Under low light conditions, high temperature measurements of luminescence are limited by the overlap of the thermal emission spectra and the luminescent emission spectra being measured. A solution to this is to have a heat source that can be designed not to emit in a certain wavelength range(s) by coating it with an interference multilayer. The multilayer effectively changes the emissivity of the heat source. Microheaters made from aluminum oxide platforms with platinum heating elements were coated with aluminum oxide and titanium oxide multilayers. This multilayer structure was used to measure the thermoluminescence of CaSO4:Ce,Tb up to 420\,^∘C. They also showed a thermal emission background 800 times lower at 600\,^∘C than the same microheater with no multilayer structure.
Infrared-transparent microheaters have been constructed to reduce the background blackbody radiation produced by the heater. Among other applications, such heaters allow one to probe the high temperature peaks of thermoluminescent(TL) materials. The microheater consists of peripheral platinum heating elements on a mid-infrared transparent alumina platform. Alumina has a relatively low blackbody signal at high temperature for wavelengths less than 8μm. To test the reduced blackbody emission, an aperture was placed over the heating coils and then the transparent center of the microheater. The amount of infrared light transmitted through the aperture was reduced by 90% as the aperture moved from the highly emissive heater coils at 450\,^∘C to the largely transparent center at the same temperature.
The thermoluminescence characteristics of a thin film of terbium-doped yttrium oxide change upon repeated stress application through electrostatic actuation. A maximum 42% decrease in the intensity of two thermoluminescent peaks is seen when voltage is applied in 5V increments to 25V, translating to 0.15 μm of center deflection. While the overall intensity decreases, the higher temperature peak - corresponding to deeper traps - is affected more than the lower temperature one. Two possible physical explanations for the behavior are mechanical stress and dielectric charging.
Thermoluminescent (TL) particles show promise as robust direct-contact thermal history sensors for explosive events. Research with microheaters has shown that TL microparticles can measure temperature excursions of hundreds of degrees; however, microheaters do not generate the severe pressure and shock stimuli present in post- detonation environments. To address this, TL particles were tested under conditions produced by the detonation of an aluminized explosive formulation. TLD-100 (LiF:Mg,Ti) powder was irradiated with 220 Gy of gamma radiation from a ^167Cs source before being exposed to the free field detonation of a 20 gram charge. Particles were recovered post-detonation from two separate tests and their TL glow curves measured. At least two TL emission peaks 50 ^oC apart are clearly distinguishable in both samples, with peak intensity ratios decreasing 33.7% and 60.0% from an original 8.88:1, indicative of distinct carrier traps emptying at rates depending on the trap energy. These ratios agree well with thermocouple measurements from within the post-detonation fireball.
The blackbody radiation limit has traditionally been set forth as the ultimate performance limit of thermal detectors. However, this fundamental limit assumes that the detector absorbs uniformly throughout the thermal spectrum. In much the same way as photon detectors can achieve very high D* because they do not absorb photon energies below their bandgap, so too can thermal detectors except that thermal detectors are not limited to cryogenic operation. In both cases, the enhanced theoretical D* is achieved because the radiation noise is reduced in a device that does not absorb at a uniform high level throughout the thermal emission band. There are multiple ways to achieve high D* in thermal detectors. One is to use materials that absorb only in a certain spectral range, just as in photon detectors. For example a detector made from PbSe, with proper optical coupling, absorbs only photons with wavelengths shorter than 4.9μm. The radiation limited detectivity of such a device can theoretically exceed 9 x 1010cmHz1/2/W in the MWIR. Even with Johnson and 1/f noise estimates included, it can still exceed 2.5x1010cmHz1/2/W in the MWIR. Another technique, applicable for narrowband thermal detectors, is probably even more powerful. Consider a thermal detector that is almost completely transparent. Here, the radiation noise has been reduced but the signal has been reduced even more. However, if the device is now placed inside an optical cavity, then at one wavelength and in one direction, the nearly transparent detector couples to the cavity resonance to absorb at 100%. Radiation from all other wavelengths and directions are rejected by the cavity or are absorbed only weakly by the detector. It is shown that theoretically, the D* of these devices are roughly proportional to the inverse square root of the spectral resonant width under certain conditions. It is also shown that even including Johnson noise and 1/f noise, the practically achievable D* approaches or exceeds 1011 cmHz^1/2/W.
Thermal expansion mismatch between the layers of an optical coating and its substrate alters the shape of an optical element. We demonstrate predictable coating behavior using atomic layer deposition and apply to high-reflectivity mirror design.
Microbolometers and other thermal detectors have traditionally been limited to seeing objects in a broad wavelength band at a single sensitivity. Recent advances in interface heat transfer and optical cavity design promise to change that. In this paper, we present recent work on thermal infrared detectors with tunable responsivity and wavelength. First, we demonstrate that extended dynamic range in thermal detectors can be achieved by electrostatically bringing a portion of the detector support structure in contact with the substrate. The exact amount of heat transfer can be controlled by adjusting the contact area and pressure. The thermal conductance and responsivity can be switched more than an order of magnitude using this technique. Next, we demonstrate that a wavelength tunable device in the LWIR can be achieved by modifying the structure of a microbolometer to incorporate a modified Gires Tournois optical cavity. The cavity couples light at a single wavelength into the microbolometer while other wavelengths are rejected. We demonstrate that resonance can be tuned from 8.7 to 11.1 μm with applied voltages from 0 to 42 V. The FWHM of the resonance can be switched between around 1.5 μm in a narrow-band mode and 2.83 μm in a broad-band mode.
An electrostatic actuation scheme for creating a soft contact is described and applied to cantilevers and doubly clamped beams. The actuation voltage form is composed of a high voltage pulse with a low DC voltage. The voltage pulse height and duration is mapped for achieving minimum impact force, and the results are well predicted by the simulation. Furthermore, a new phenomenon is observed for clamped beams that pulse longer than the soft-landing requirement bounces the beam back rather than stick.
The adhesion energy is measured between InGaAs quantum wells that have collapsed across a 120 nm airgap in a InP/InGaAs heterostructure. The method relies on measuring the unadhered length and shape of collapsed microcantilevers with optical interferometry. The adhesion energy is found to be 72 plusmn 16 mJ m^-2. Since the airgap is much smaller than has been measured previously, the influence of van-der-Waals forces across the gap was included in theoretical modeling. It was found that the forces should not cause significant deviation from the standard adhesion models unless the adhesion energy drops below 25 mJ m^-2.
This paper covers the design, fabrication, and testing of a step-wise tunable long-wavelength infrared filter. The design uses a modified type of microbolometer to filter infrared light in the 7-10 micrometer wavelength range. Fabrication of the device is accomplished using standard microelectronic and microelectromechanical systems (MEMS) fabrication techniques. Germanium and zinc sulfide are used as optical layers for good reflectivity in the desired infrared range. Testing of the device shows that it filters light in three absorption bands in the specified wavelength range. The tuning of the device is found to be step-wise in response to electrostatic actuation using voltages up to 100 volts.
In this paper, the fabrication of microbolometers with electronically controllable responsivity is presented. The first generation devices are built in a standard polysilicon-based micromachining process with HF etch-release and demonstrated with a responsivity that can be tuned over a factor of 50. The responsivity is controlled by applying a voltage between the microbolometer and the substrate. The resulting electrostatic force causes a small portion of the support beam to contact the substrate, which thermally shorts the device at that point. The thermal contact points are defined using curved support beams with residual-stress from a Cr/Au metallization. The lowest portion of the beams contacts the substrate, and the curvature protects the device from full "snap-down," which might induce stiction. The fabrication of the second generation microbolometers based on VOx and silicon nitride materials with a polyimide etch-release is also described. The thermal contact points for these devices are defined by beam mechanics rather than by beam curvature induced by stress, and they actuate at 17 volts. The test array has a fill-factor of 91% for a pixel period of 140μm limited by our photolithography equipment.
When thermal infrared detectors are exposed to large signals, they are susceptible to a host of unwanted effects including intensity dependent noise and detectivity, nonlinearities in materials characteristics, and even temporary blindness or device damage. To combat these problems and to extend dynamic range, the responsivity and time constant of microbolometers have been controlled using electrostatic actuation. The responsivity has been demonstrated to switch over a factor of 60, with theoretical limits encompassing 4 to 5 orders of magnitude. High responsivity states correspond to free-standing bolometers, while discrete lower responsivity states are created by partially or completely actuating the device supports into contact with the substrate. Continuous tuning over a part of the range is demonstrated by utilizing electrostatic pressure to increase the thermal contact between discrete switching states.