Automated real-time spectral characterization of phase-change tunable optical filters using a linear variable filter and infrared camera
AAutomated real-time spectral characterization of phase-changetunable optical filters using a linear variable filter and infraredcamera
David Bombara a , Calum Williams b , Stephen Borg c , Hyun Jung Kim c,d,* a University of Nevada, Reno, Department of Mechanical Engineering, 1664 N. Virginia Street, Reno, NV, USA,89557 b Department of Physics, Cavendish Laboratory, University of Cambridge, JJ Thomson Avenue, Cambridge, CB30HE, UK c NASA Langley Research Center, 1 Nasa Drive, Hampton, VA, USA, 23666 d National Institute of Aerospace, 100 Exploration Way, Hampton, VA, USA, 23666
Abstract.
Actively tunable optical filters based on chalcogenide phase-change materials (PCMs) are an emergingtechnology with applications across chemical spectroscopy and thermal imaging. The refractive index of an embeddedPCM thin film is modulated through an amorphous-to-crystalline phase transition induced through thermal stimulus.Performance metrics include transmittance, passband center wavelength (CWL), and bandwidth; ideally monitoredduring operation (in situ) or after a set number of tuning cycles to validate real-time operation. Measuring theseaforementioned metrics in real-time is challenging. Fourier-transform infrared spectroscopy (FTIR) provides the gold-standard for performance characterization, yet is expensive and inflexible—incorporating the PCM tuning mechanismis not straightforward, hence in situ electro-optical measurements are challenging. In this work, we implement anopen-source M
ATLAB ® -controlled real-time performance characterization system consisting of an inexpensive linearvariable filter (LVF) and mid-wave infrared camera, capable of switching the PCM-based filters while simultaneouslyrecording in situ filter performance metrics and spectral filtering profile. These metrics are calculated through pixelintensity measurements and displayed on a custom-developed graphical user interface in real-time. The CWL isdetermined through spatial position of intensity maxima along the LVF’s longitudinal axis. Furthermore, plans aredetailed for a future experimental system that further reduces cost, is compact, and utilizes a near-infrared camera. Keywords: spectroscopy, phase-change materials, tunable filter, spectral imaging, linear variable filter, GUI (graphicaluser interface). *Corresponding author:
Hyun Jung Kim, [email protected]
Optical bandpass filters are critical components utilized in a plethora of systems and applications,from fluorescence microscopy to remote sensing.
These filters are designed to transmit only acertain band of wavelengths (passband) and block all others. Conventional optical bandpass filtersare passive—offering discrete static passbands, arising from the interference of alternating indexdielectric thin-films.
4, 5
There are a growing number of imaging applications requiring precisespectral filtering across a range of wavelengths (tunability), with motorized filter wheels typi-cally utilized.
6, 7
However, filter wheels are bulky, have limited spectral coverage, and offer slowswitching speeds. In recent years, there have been significant research efforts toward cost-effectivetunable optical bandpass filters that can operate across multiple wavebands and exhibit fast switch-ing capability, based on nonstandard material platforms. Chalcogenide phase-change materials(PCMs), and their integration into micro-and nano-structured filter designs, are emerging as anattractive candidate for such a tunable filter. These exotic materials exhibit a large refractive indexshift across multiple wavebands through a reversible phase transition (amorphous-to-crystalline)1 a r X i v : . [ phy s i c s . i n s - d e t ] F e b igure 1: Active wavelength tuning of the phase-change material (PCM) cavity narrowband filter.(a) The PCM-cavity narrowband filter takes an optical input (illumination) and returns an opticaloutput that is filtered at a corresponding center wavelength (CWL). The CWL is a function of thestate of the PCM. (b) The CWL of the optical output is shortest when the PCM is in its amorphousphase, hence the lowest refractive index value. Partially crystalline states increase the refractiveindex, hence the CWL, until the PCM reaches the fully crystalline state (highest refractive index).(c) To “reset” the PCM to the amorphous phase, a pulse of high power but short duration must beapplied (e.g., laser pulse). To set the material to its crystalline phase, a lower-power pulse of longerduration must be applied. (d) The material resets to its amorphous phase if it exceeds its meltingtemperature T m , then is rapidly quenched below its glass transition temperature T glass in time lessthan t . The crystalline transition requires that the PCM exceed T glass for a duration of at least t .initiated through external energy stimuli. When integrated into established or novel optical fil-ter design schemes—such as Fabry-Perot interference-based or metasurface-based filters—PCMscan provide an optically active medium to tune the passband center wavelength (CWL) throughrefractive index switching. PCMs undergo a phase transition through rapid localized melting and recrystallization, whichcan be induced through optical or electrical stimuli. The former is typically used in rewritable op-tical storage media. Compounds consisting of germanium, antimony, and tellurium (GeSbTe)are commonly used for optical devices, from optical storage media, integrated photonic elements,to tunable filters. To switch the material to its amorphous phase, an ultrafast (nanosecond), high-intensity pulse must be applied. In contrast, switching to the crystalline phase requires a longer,lower-intensity pulse. For amorphization, the material must be heated above its melting tempera-2ure, then rapidly quenched. For the material to crystallize, the material must be heated above itsglass transition temperature for a duration long enough to reach crystallization. PCM-based tunableoptical filters are typically switched using laser-based irradiation. An overview of the PCM filterconcept, laser input sequence, and corresponding temperatures of a PCM filter is shown in Fig. 1.Partial crystallization can also be realized, and intermediate phase switching has been previouslyshown to increase the number of accessible CWLs for a single multistate tunable filter. PCM-based tunable filters have been shown to provide all-solid-state ultrafast (nanosecond)switching from the near-infrared (NIR) to mid-wave infrared (MWIR).
10, 11
Despite the promisingresults, no studies, to the knowledge of the authors, have focused on device reliability metrics,which is crucial to understand their behavior over many tuning cycles, device longevity, failurerates, and suitability for practical and real-world applications. For example, spectral propertiesthat remain constant for each phase cycle would indicate that the filter functions reliably. How-ever, when its properties begin to deviate from typical values, the filter may no longer be fit forits purpose. To determine these reliability metrics and characterize the tunability of the PCM-based filter, the spectral properties must be measured after each phase change. These include thepassband CWL, transmittance intensity, and bandwidth (in terms of full width at half maximum,FWHM). Further, good homogeneity across the spatial extent of the filter is highly desirable, there-fore the monitoring of this information is needed across the entire filter (i.e., 1D spatiospectralinformation as opposed to single point spectroscopic measurements). Fourier-transform infrared(FTIR) spectroscopy is arguably the most common method for obtaining this information and isthe most accurate.
10, 11
For example, the Thermo Scientific ™ Nicolet ™ iS ™
10 FTIR spectrometerhas a 10,000:1 signal-to-noise ratio and spectral resolution of 0.4 cm − . However, it is expensiveand bulky instrumentation that requires long data collection times to calculate the spectral proper-ties from the interferogram. Critically, it is challenging to integrate the PCM-based filter switchingsystems within the FTIR apparatus and FTIR systems are generally limited to providing spectralproperties from only a single point on the sample.In this study, we present an approach to real-time PCM-based tunable filter characterizationthat utilizes a linear variable filter (LVF) and IR camera to measure the filter’s CWL. This methodis simpler to construct for the research community and enables in situ measurement across theentire filter area (1D spatiospectral measurement). A graphical user interface (GUI) is developedin M
ATLAB ® to control separate instruments and plot the tunable filter’s passband CWL in real-time. These properties are measured based on acquired images from an IR camera. Furthermore,we detail progress on the PCM’s reliability evaluation and a system for tuning and characterizingthe PCM using commonly available, off-the-shelf hardware. FTIR based analysis, albeit accurate, is expensive, time-consuming (especially if needed to per-form spectral 2D measurements) and challenging to modify for our real-time performance char-acterization requirements. Instead, our alternate approach is to utilize a compact and inexpensive(relative to an FTIR spectrometer) LVF—a small rectangular optical filter that has its CWL vary-ing linearly along the length of the filter. Therefore, in combination with the PCM-based filter,simply by imaging (recording) the intensity response through an LVF, it is possible to characterizethe spectral performance. The accuracy of this method is high enough such that the phases of thePCM (amorphous and crystalline) are unambiguously distinguishable and the error margin of the3igure 2: Experimental setup. (a) A schematic of the experimental setup. The infrared (IR) cameraimages the LVF with a fixed-wavelength bandpass filter behind it. They are both illuminated bythe contact hot plate (blackbody). (b) A photograph of the experimental setup. The hot plate actsas a blackbody thermal radiation source. (c) The LVF is mounted in the IR camera’s field of view.(d) The bandpass filter is between the blackbody source and LVF.CWL measurement is 0.085 µm. Error analysis is provided in Section 2.2, “Experimental Results.”Our previous work realized a PCM-based filter with a CWL of approximately 2.9 µm and 3.3 µm inits amorphous and crystalline states, respectively. With this LVF-based system, the CWL shift of0.4 µm is detectable. Further, this approach has the advantageous feature that it is simpler to mod-ify in order to evaluate (test) other switching mechanisms or imaging approaches, such as trialingdifferent electrode designs, different laser technologies, off-axis illumination, etc.
To characterize a filter with tunable CWL, a system that validates the spectral performance ofstatic bandpass filters is first needed. The experimental setup consists of (1) an IR camera (Indiumantimonide detector, FLIR SC8000 Series), (2) a radiation shield tube, (3) a contact hot plate(blackbody radiation source), and (3) the filter stack (Fig. 2). The IR camera is capable of imagingthe MWIR waveband (2.5–5.0 µm) with a resolution of 1344 ×
784 pixels. The camera utilized alens with a focal length of 50 mm. The distance between the lens and filter (target) was 122 cm.The IR images are recorded using FLIR ResearchIR software. The blackbody radiation source wasan iron plate with emissivity ε ≈ and a diameter of 17.8 cm. A tube between the camera andfilter stack shielded against unwanted background radiation.A top-down schematic of the experimental setup is shown in Fig. 2(a), whereas the photo-graph of the entire setup is shown in Fig. 2(b). The LVF (Vortex Optical Coatings, Ltd.) andbandpass filter are shown in Fig. 2(c) and 2(d), respectively. The bandwidth of the LVF at 50%peak transmittance is specified by the manufacturer to be 2% of the peak wavelength with peak4ransmittance typically greater than 60% across the CWL band. The LVF also has out-of-bandblocking within its spectral range of an optical density of 3.0. The LVF has nominal dimensionsof 15 mm × × ± ±
30 nm and 140 nm ±
30 nm.
Using the MWIR camera, the LVF was imaged with seven different bandpass filters behind it.Figure 3 shows the experimental results. The superimposed view of the IR intensity image andRGB image is shown in Fig. 3(a) for illustration purposes. As shown in Fig. 3, the CWL of the LVFvaries along the length of the filter beginning at 2.5 µm and gradually increasing to 5.0 µm at theopposite end. However, due to variations in the manufacturing of our particular LVF, experimentaldata confirmed that it reached its 5 µm peak in just 10 mm of filter length. The “scan line pixelnumber” in Fig. 3 refers to the numbering of the camera’s pixels along the horizontal axis ofthe LVF. The CWL here is defined to be the wavelength at which the transmittance is maximum.“Detector counts” from the IR camera are used as a proxy for transmittance; the location of peakdetector counts is the location of the peak filter transmission. To find this location on the LVF, asingle horizontal line of pixels is sampled in the center of the filter and parallel to the long axis. InFig. 3(b), the locations of peak transmission can be identified as regions of higher relative intensity,where the brightness of the pseudocolor image corresponds to the number of counts recordedduring a given detector integration time. For each bandpass filter, the particular integration timewas determined experimentally to obtain the high-contrast image and depended on the temperatureof the blackbody radiation source.In this experiment, we demonstrate that we can detect the CWL and accurately measure itslocation along the LVF. The plot of normalized detector counts versus scan line pixel numberis shown in Fig. 3(c) for each bandpass filter. The experimental results confirm that the LVFaccurately indicates the CWL of the filter that is placed behind. The correlation between CWL andthe scan line pixel number is shown in Fig. 3(d). The pixel number of the scan line used as a proxyfor the linear distance along the filter.For our experimental setup and detector/optics combination, the minimum horizontal spatialresolution for each pixel is approximately 0.341 mm. The instantaneous field of view (IFOV) ofour imaging system represents the spot size on the LVF for an individual pixel and also correspondsto our estimated spectral measurement uncertainty. Using a 50-mm lens in combination with a1344 × ± × y scan , is shown in Fig. 4. To find the location of the CWL, first, aregion of interest (ROI) is drawn over the filter image. y scan was chosen to be in the center of theLVF. It was initially assumed that the horizontal position of the peak transmittance may depend onthe chosen location of y scan , due to fabrication tolerances. However, Fig. 5 shows otherwise, whichshows the variation of the detector counts in the 2D pixel grid. Figure 5 shows that the locationof peak transmission along x does not change depending on the chosen scan line location, y scan .Another method to determine the location of peak transmittance would be to scan every pixel inthe 2D ROI for every CWL measurement. Compared to scanning a single line, this method is lesscomputationally efficient, making it less suited for the real-time characterization application. Thenumber of detector counts is evaluated within two bounds, where x lb and x ub are the lower andupper bounds, respectively. In Fig. 4(a), x and y refer to the numbering of the image’s horizontaland vertical pixels, respectively. The detector counts must be evaluated within bounds becausethe detector counts increase dramatically outside the range of pixel numbers that do not cover theLVF. This increase can be seen in Fig. 4(b), where beyond x lb and x ub the colors that correspondto the detector counts appear brighter. Figure 4(b) shows results for the 3.60-µm filter with labeledspectral properties. The locations of the CWL, FWHM, and peak transmittance are shown. TheCWL is a function of the scan line pixel number; the two variables vary according to the equation, CW L = a ( x − x lb ) + λ , (1)where a is the proportionality constant, λ is the passband CWL when x = x lb , and x is thenumbering of the pixels horizontally. Least-squares linear regression identified a = 0 . µ m and λ = 2 . µ m . The PCM tunable filters are based on thin-film Ge Sb Te and fabricated in a laboratory at NASALangley Research Center; the fabrication process is described in detail in other studies.
10, 11, 17
Theevaluation of tunable PCM-based filters may be automated in two aspects: phase switching and7igure 4: The method for determining the CWL of the filter from raw pixel data. (a) The IRimage used to obtain CWL. The scan line is located at y = y scan . The upper and lower bounds ofthe horizontal pixel number are x lb and x ub , respectively. (b) The spectral results of the 3.60-µmbandpass filter. Spectral properties shown include the locations of the CWL, peak transmittance,intensity half-maximums, and bounds of the filter.Figure 5: A three-dimensional plot showing the variation of the detector counts with the horizontaland vertical pixel numbers that correspond to the area of the IR image that the LVF occupies. Aslong as the scan line passes through the entire length of the LVF, the location of peak transmittancedoes not depend on the exact location of the scan line.8igure 6: Operation flowchart: The general commands and processes executed by the M ATLAB ® graphical user interface (GUI). The program runs in a loop until the cycle limit is reached but canbe terminated by the user at any time.performance characterization. A M ATLAB ® -based application is developed to control the hard-ware systems and analyze incoming data. Figure 6 details the main loop that runs in order toswitch and characterize the PCM-based filter. The source code for the M ATLAB ® application isavailable at https://dbombara.github.io/automated-pcm-characterization, as well as a descriptionof the experiment and setup.Once the user begins the experiment, the filter automatically switches between the amorphousand crystalline states while its CWL is automatically calculated and saved for each cycle. Thesequence of functions that automates the system is shown in Fig. 6. The GUI allows the userto conveniently control and monitor the experiment’s progress. At the start of the experiment,M ATLAB ® establishes connections to each instrument. The initial IR image is then loaded, withthe ROIs displayed on the screen. At this point, the user may redraw the ROIs, based on the exactposition of the LVF in the view of the camera. The selected ROIs are displayed on the screen andremain in view while the experiment runs. Before firing a laser pulse, first, the desired state isevaluated. Typically, the desired state alternates on each cycle between amorphous and crystalline.The desired state of the PCM determines the intensity and duration of the fired laser pulse. Aclosed relay switch instructs the IR camera to save its image. The laser waveform is then plotted9or that cycle. Then, the file of the IR image is imported and displayed in the GUI. An image maskis applied that allows the CWL to be calculated and plotted within two bounds on the horizontalaxis. All data from that cycle are then saved to a file. The experiment runs until the cycle limit isreached, which can be preset by the user or determined by the GUI based on the data from eachcycle. More details on these processes may be found below. The switching of the PCM between amorphous and crystalline states is triggered by the heat ofan incoming laser pulse. A Quantel Evergreen HP pulsed laser is chosen in this study to deliverup to 340 mJ of energy at 532 nm center wavelength. The pulses are delivered at a rate of up to20 Hz. The wavelength and energy level were chosen based on our previous work on GeSbTephase-change metasurface spectral filters. The pulsed laser output is controlled with a functiongenerator and is viewed with an oscilloscope; this conveniently enables the user to monitor thewaveform. The laser consists of two components: the power supply and the “head”, which emitsthe pulsed laser beam. In this study, the oscilloscope (Tektronix MDO4024C) and function gen-erator (Tektronix MDO4AFG) functionalities are combined into a single instrument (MDO standsfor mixed-domain oscilloscope). This choice of oscilloscope and function generator was due to thehigh-level programming commands that are available in M
ATLAB ® .The laser, function generator, and oscilloscope are controlled within the M ATLAB ® user inter-face. From M ATLAB ® ’s Instrument Control Toolbox ™ , Quick Control Oscilloscope commands areemployed to issue high-level commands to the instruments. This aspect of the GUI builds upon theOscilloscope App from MathWorks ® and incorporates similar functionality. Serial commandscan also be issued directly to the laser to adjust the settings of the instruments, such as the powerlevel. The performance characterization is conducted in a way similar to the characterization in Section2. To automate the image acquisition process, a USB-controllable relay (8-Channel 5-Amp ProXRLite, Relay Pros) triggers the camera to record the IR images. In this study, the switch is “dry”;the closure of the relay switch draws no electrical current. The FLIR ResearchIR software savesthe image in a proprietary file format known as ATS. Then, the FLIR File Reader Software Devel-opment Kit (SDK) is used to import the ATS file into M
ATLAB ® . Functions from the SDK thenextract the image’s metadata and raw pixel information so that the image can be displayed in theGUI. Using the imported image, the CWL of the tunable filter is calculated in the manner explainedin Section 2, using Eq. 1. The GUI is designed with the user’s inputs on the left quadrants of the window and the system’soutputs on the right quadrants. A screenshot of the GUI is shown in Fig. 7. The various aspectsof the GUI are organized with tabs. The names of the tabs include “Instruments”, “MeasurementMode”, “Stop Condition”, “Pulse Settings”, and “Testing”. There are additional groups of tabsfor monitoring the performance characterization of the PCM by displaying raw data and plottingspectral properties. 10igure 7: Real-time filter characterization: A screenshot of the custom M
ATLAB ® GUI.The source code and packaged application are available at https://dbombara.github.io/automated-pcm-characterization. The window contains four quadrants (a)-(e), which each containtabs. The left half of the GUI is generally for inputs by the user, whereas the right half displaysthe output from the instruments and computed spectral properties. (a) The top-left quadrant of theGUI contains tabs for hardware-based settings and conditions. (b) The top-right quadrant of theGUI displays “raw” data from the IR camera and laser. (c) The Instruments tab, which is a part ofthe top-left quadrant, enables the user to connect to the different instruments. (d) The bottom-leftquadrant is for software-based settings. (e) The bottom-right quadrant displays the spectral prop-erties of the tunable filter. With the current LVF-based approach, only the CWL may be plotted. Infuture work, other properties, such as transmittance and full width at half maximum (FWHM) willbe calculated. The (f) Stop Condition tab, (g) Testing tab, and (h) Files tab allow the user to havemore control over the experiment.
It is crucial to ensure all instruments are properly connected before beginning the experiment. Forthat reason, the Instruments tab, shown in Fig. 7(c), includes information to connect and operatethe oscilloscope and function generator. Options such as the resource (the instrument identifier),driver, channel, voltage range, and acquisition time are selected by the user. The Instrument tabalso contains buttons for connecting to the relay and laser power supply via serial connections.11 .4.2 Measurement Mode and Stop Condition Tabs
The Measurement Mode tab (Fig. 7(d)) and Stop Condition tab (Fig. 7(f)) both allow the userto select options for terminating the experiment when a particular condition, or combination ofconditions, is met. These tabs are especially useful for hours-long reliability tests. In the StopCondition tab, the user can choose to stop the experiment when either the transmittance, CWL,or both properties fall outside a predetermined limit. Similarly, the experiment may stop after aparticular number of occurrences outside that limit, or when the number of occurrences reaches aset percentage of tuning cycles completed thus far. In contrast, the conditions in the MeasurementMode tab—total cycles and total time—are calculated in the GUI itself without relying on theconnected hardware.
The required energy and pulse duration may change depending on the particular PCM that is un-der test. In the Pulse Settings tab (Fig. 7(a)), the user can set the desired laser voltages (V),pulse widths (ns), and delays (µs) between pulses. The GUI then calculates the duty cycles (%),frequencies (MHz), and periods (µs) of the pulses for amorphization and crystallization. The spe-cific values of the laser voltage, pulse width, and delay are determined experimentally and varydepending on the particular materials utilized in the tunable filter.
The top-right quadrant (7(b)) of the GUI window is for displaying “raw” data: in general, datathat come directly from instruments, not values that are computed as a result of the instruments’data. In this quadrant, there are two tabs: the “Laser (Osc/FGen)” tab and the IR Image tab. As thenames imply, the former plots the input to and output from the laser, as determined by the functiongenerator and oscilloscope. The input signal to the laser differs from the output signal; the laserrequires two pulses of 5-V amplitude and 10-µs duration, with 170 µs between pulses. However,the output pulse from the laser head is only nanoseconds-long. The IR Image tab displays the IRimage and ROI that are selected before the experiment. As the CWL of the PCM-based filter istuned, the IR image will change based on the indicator on the LVF.
The Testing tab (Fig. 7(g)) contains buttons for the user to test the functionality of the GUI beforerunning the experiment. The relays can be turned on and off with the click of a button. Samplefiles can also be imported, after which the IR image can be plotted, and the spectral properties canbe calculated. The Files tab (Fig. 7(f)) allows the user to specify file and folder names from whichdata will be imported and to which the data will be saved.The bottom-right quadrant (Fig. 7(e)) of the GUI window contains plots of values computedfrom the IR image above. The cycle number is plotted on each abscissa whereas the CWL, FWHM,and transmittance intensity are each plotted in separate tabs on the ordinate axes.The M
ATLAB ® GUI has much utility in the current experimental system, but in future work, theGUI will be applied to (1) an experiment to investigate the PCM-based filter’s reliability and (2) acompact, and custom-designed circuit to optically switch the PCM-based filter and characterize itsphase changes in the NIR waveband. 12
Applications
Despite the current capabilities of the automated experimental system, future developments arenecessary to make PCM-based filters ready for commercial and government applications. Bybuilding upon the M
ATLAB ® GUI and IR camera-based characterization in this work, two addi-tional developments are proposed. First, experiments will be conducted to evaluate the reliabilityand longevity of the PCM-based filter. Second, a system to characterize and tune the PCM-basedfilter in the NIR waveband will be developed with commonly-available and cost-effective compo-nents. In addition, the proposed system will have reduced size and weight due to a custom-designedcircuit for optically switching the phase of the PCM.
The reliability of the PCM-based filters will be evaluated using the M
ATLAB ® GUI in future ex-periments. Figure 8 shows the schematic of the reliability test that will utilize the M
ATLAB ® GUI.The setup is similar to that in Fig. 2, but with added components to facilitate the laser-based tuningof the PCM-based filter. Because the laser will heat the PCM-based filter during the phase change,the IR camera must wait for that heat to dissipate before taking an image. This is because the IRcamera cannot differentiate between surface heating of the filter and blackbody radiation. Dur-ing the reliability experiment, the laser firing and the IR image acquisition will be separated for aduration that allows the heat of the PCM to sufficiently dissipate. However, the experiment willstill be automated, since the filter may be potentially be switched for – cycles until it fails;the moment of failure can be identified with our approach. Previous PCM reliability studies havedemonstrated long operational longevity, with one example demonstrating . × switching cy-cles before failure. However, PCM-based optical filters face challenges over their resistive-basedPCM counterparts. In optical PCM-based devices, the entire area of the tunable film must undergoa complete and reversible phase transition. However, resistive-based PCM-based devices do notrequire complete crystallization of the tunable film to switch from “high” to “low”. To be used in NASA’s Space and Science missions, the filter must be reliable, with any drop-offin performance as a function of the tuning cycle known in advance. For example, one potential ap-plication of the tunable filter would be the Space Launch System (SLS) for the Artemis-1 Mission.The filter could enable accurate temperature measurement of the rocket’s core booster stage viaradiometric techniques because of its nanosecond-scale filter switching speeds and narrow band-width. If the error margin of the temperature measurement is reduced, the maximum temperaturesfrom the rocket may be smaller than originally designed for, therefore, the heat shield may notneed to be as massive. The tunable filter could also be utilized for chemical and gas sensing inNASA’s Science missions. If the mission time is one year, for example, the tunable filter mustwork reliably for that duration.
Despite the advantages of PCM-based filters, there is a need for decreased size, weight, power,and cost; this would enlarge the range of possible applications for the tunable filter. Figure 9shows the design of a cost-effective and compact tuning system for PCM-based filters that havedetectable phase transitions in the NIR waveband, albeit with typically an increased absorption.The cost of components for this system was less than $400. The schematic is shown in Fig. 9(a),the photographs of the initial build are shown in Fig. 9(b)–(c), and the comparison of input and13igure 8: The reliability test setup: The reliability test will rely on the M
ATLAB ® GUI for auto-mated laser control and automated measurement of the filter’s CWL. The CWL will indicate thecurrent state of the filter.output signals are shown in Fig. 9(d)–(e). The system consists of two main subsystems: thephase-switching subsystem and the imaging subsystem.
In Fig. 9(a), the DC power supply (LRS-350-24, MEAN WEll Enterprises Co., Ltd.) providespower to the overall system and allows the laser diode to emit high-current pulses. A programmablevoltage regulator determines the input to the step-up DC-DC converter (FS02-12, XP Power). Theoutput of the voltage regulator (LM317, Texas Instruments) is adjusted using a digital potentiome-ter (X9C104). The amplitude of the DC-DC converter output determines the amplitude of the laserpulse.The microcontroller (Due, Arduino) functions as both a waveform generator and controller forthe digital potentiometer. A common-emitter amplifier circuit is utilized to amplify the pulse wavefrom the Arduino. In Fig. 9, the values of the resistors and capacitors are chosen based on circuitsimulations. It is noted that other amplification strategies may be used, such as an operationalamplifier with a high slew rate. A previous project created an arbitrary waveform generator usingan Arduino Due and obtained pulse widths as low as 12 ns. Similarly, the Arduino in Fig. 9 wasprogrammed to output a wide range of pulse widths and frequencies. The output from the signalamplifier determines the pulse width and frequency of the laser diode’s output. Although the pro-totype was constructed on a breadboard, the system may be constructed on a printed circuit board(PCB) to for more compactness. This prototype was tested using one-Watt LEDs to approximate14igure 9: The cost-effective and compact tuning system. (a) The schematic of the proposed system.(b) A prototype and (c) the wired connections on the breadboard. (d) The system takes an inputpulse wave from the Arduino microcontroller. (e) The output pulse waveform is amplified nearly30 times compared to the input pulse, with minimal distortion.15he behavior of a high-powered laser diode.The components for the prototype are available frompopular retailers, making them accessible for both research and education purposes.Preliminary results from the prototype are shown in Fig. 9(d)–(e). Figure 9(d) shows the inputpulse wave from the microcontroller whereas Fig. 9(e) shows the output of the DC-DC converter.The input signal has a frequency of 4.98 kHz and a peak-to-peak voltage of 3.64 V. The outputsignal retained the same frequency as the input signal, but its peak-to-peak voltage was 97.6 V. Asseen in Fig. 9(d)–(e), the signal is minimally distorted after amplification.
For the imaging subsystem, a single-board computer (Model 3 B+, Raspberry Pi) is proposedto interface with an NIR camera (NoIR Camera, Raspberry Pi). This camera is similar to theRaspberry Pi V2 Camera Module, except the NoIR camera possesses no IR-cutoff filter. Thespectral response of the Raspberry Pi V2 Camera Module has previously been studied, and itsCMOS image sensor (Si-detector) has a spectral response from approximately 400–1000 nm. Theutility in the Raspberry Pi NoIR camera is in imaging PCMs that exhibit index modulation inthe NIR as well as the short-wave infrared (SWIR) and MWIR. Thus, phase transitions can beinferred based on the NIR intensity changes. Using the NIR camera, Fresnel reflections of the PCMat normal incidence would indicate the phase transitions, due to the PCM’s changing refractiveindex upon switching. The Raspberry Pi computer and camera are directly controllable using theM ATLAB ® Raspberry Pi Support Package, making their integration into the current M
ATLAB ® GUIstraightforward.
PCM tunable filters are emerging as an alternative technology platform for ultrafast spectral filter-ing across a wide range of operating wavelengths. However, the ability to accurately and easilycharacterize their performance remains challenging. In this study, a system for automated tunablefilter characterization was designed and implemented as an alternative to FTIR spectroscopy. TheLVF was shown to accurately indicate the CWL of the bandpass filter behind it. A M
ATLAB ® GUIwas designed to automatically control the subsystem instrumentation, display the filters’ spectralproperties, and record the output data. The overall system enables 1D spatiospectral (bright field)characterization and allows for easier integration of other tuning mechanisms in comparison toFTIR systems. Plans were detailed for a future reliability experimental system that is compact,low-cost, and utilizes a NIR camera.Future work will build upon the M
ATLAB ® application and experimental results to advance thestate of the art in PCM-based optical filters. Reliability studies in the future will reveal the filter’slong-term performance. The compact system for imaging and switching of the filter will decreasethe cost of optical PCM-based devices. This work has wider applications in systems that utilizespatially-varying optical filters, such as circular variable filters. The software that was developedin this work can also be modified to incorporate more-sophisticated image processing algorithms,such as an algorithm to automatically detect the location of the LVF and determine the ROI. Fordevices and materials with more complicated spatially-varying spectral profiles, machine learningtechniques may be used for more robust spectral characterization. The M ATLAB ® application isavailable at https://dbombara.github.io/automated-pcm-characterization.16 .1 Paper author contributions D.B. developed the M
ATLAB ® application. H.J.K and C.W. conceived the idea. H.J.K. and S.B.carried out the experiments. C.W. advised the device characterization. All authors wrote themanuscript with D.B. as the lead. D.B. would like to thank the NASA Internship, Fellowship, and Scholarship (NIFS) program co-ordinators and/or project manager for the internship opportunity. The authors would also like tothank Scott Bartram, Equipment Specialist at the NASA Langley Research Center, for his expertiseand assistance with laboratory instruments. Beginning in June 2020, this work has been completedunder the NIFS program with the project entitled “LabVIEW controlled evaluation system of ac-tively tunable filter for NASA Science & Space Mission”.
Langley Research Center (CIF, IRAD); Engineering and Physical Sciences Research Council(EP/R003599/1); Wellcome Trust (Interdisciplinary Fellowship).
The authors declare that they have no competing interests.
References
Applied Spectroscopy , 1091–1120 (2012).2 N. Gat, “Imaging spectroscopy using tunable filters: a review,” in Wavelet Applications VII , , 50–64, International Society for Optics and Photonics (2000).3 J. W. Lichtman and J.-A. Conchello, “Fluorescence microscopy,” Nature Methods (12), 910–919 (2005).4 H. Macleod, Thin-Film Optical Filters , Institute of Physics Publishing, third ed. (1986).5 M. Lequime, “Tunable thin film filters: review and perspectives,” in
Advances in Optical ThinFilms , , 302–311, International Society for Optics and Photonics (2004).6 N. A. Hagen and M. W. Kudenov, “Review of snapshot spectral imaging technologies,” Op-tical Engineering (9), 1 – 23 (2013).7 G. Lu and B. Fei, “Medical hyperspectral imaging: a review,” Journal of Biomedical Optics (1), 1 – 24 (2014).8 M. Wuttig, H. Bhaskaran, and T. Taubner, “Phase-change materials for non-volatile photonicapplications,” Nature Photonics , 465–476 (2017).9 Q. Wang, E. T. Rogers, B. Gholipour, et al. , “Optically reconfigurable metasurfaces andphotonic devices based on phase change materials,” Nature Photonics (1), 60–65 (2016).10 C. Williams, N. Hong, M. Julian, et al. , “Tunable mid-wave infrared Fabry-Perot bandpassfilters using phase-change GeSbTe,” Opt. Express , 10583–10594 (2020).11 M. N. Julian, C. Williams, S. Borg, et al. , “Reversible optical tuning of GeSbTe phase-changemetasurface spectral filters for mid-wave infrared imaging,” Optica , 746–754 (2020).172 M. Jafari, L. J. Guo, and M. Rais-Zadeh, “Application of phase change material in tunableoptical filters and shutters,” in Micro- and Nanotechnology Sensors, Systems, and Applica-tions XI , T. George and M. S. Islam, Eds., , 110 – 116, International Society for Opticsand Photonics, SPIE (2019).13 M. Wuttig and N. Yamada, “Phase-change materials for rewriteable data storage,”
NatureMaterials , 824–832 (2007).14 “Advantages of a Fourier Transform Infrared Spectrometer,” Technical Note 50674, ThermoScientific.15 R. Sczupak, M. Fredell, T. Upton, et al. , “Measurement and alignment of linear variablefilters,” in Photonic Instrumentation Engineering V , Y. G. Soskind, Ed., , 207 – 212,International Society for Optics and Photonics, SPIE (2018).16 “Infra Red Linear Variable Filters | vortexopticalcoatings.co.uk.”17 H. J. Kim, M. Julian, C. Williams, et al. , “Active tunable filters based on GeSbTe phase-change materials and surface plasmon resonance (Conference Presentation),” in
Optical Com-ponents and Materials XVII , S. Jiang and M. J. F. Digonnet, Eds., , International Soci-ety for Optics and Photonics, SPIE (2020).18 A. Ursache, “Oscilloscope App (mathworks.com/matlabcentral/fileexchange/69847-oscilloscope-app), MATLAB File Exchange. Retrieved January 21, 2021,” (2021).19 “Quantel EverGreen HP LRR User Manual,” (2018).20 J.-S. Moon, H.-C. Seo, K. K. Son, et al. , “Reconfigurable infrared spectral imaging withphase change materials,” in
Micro- and Nanotechnology Sensors, Systems, and ApplicationsXI , T. George and M. S. Islam, Eds., , 117 – 121, International Society for Optics andPhotonics, SPIE (2019).21 B. Evans, “Arduino Due Arbitrary Waveform Generator,” (2017). https://github.com/Bruce-Evans/ArduinoDueArbitraryWaveformGeneratorAndController.22 M. A. Pagnutti, R. E. Ryan, G. J. C. V, et al. , “Laying the foundation to use Raspberry Pi 3V2 camera module imagery for scientific and engineering purposes,”
Journal of ElectronicImaging (1), 1 – 13 (2017). David Bombara is a master’s student in mechanical engineering at the University of Nevada,Reno. He is also an intern since June 2020 for the NASA Langley Research Center through theNIFS (NASA Internship, Fellowship, and Scholarship) program. He obtained his bachelor’s degreein mechanical engineering from the University of Nevada, Reno in 2020. His broad researchinterests include optical instrumentation, measurement, and control systems.
Dr. Calum Williams is a Research Fellow in the Department of Physics at the University of Cam-bridge. His research focuses on miniaturized optical imaging technologies for biomedical diag-nostics. He completed his Ph.D. in nanophotonics and holography at the University of Cambridgein 2017.
Stephen E. Borg is an aerospace technologist at the NASA Langley Research Center. He receivedbachelor’s degrees in mechanical engineering (1988) and physics (1988) from Old Dominion Uni-versity and has been working primarily in the areas of radiometry, infrared imaging, and tempera-ture measurement in support of NASA’s subsonic & hypersonic aeronautics research programs.18 r. Hyun Jung Kimr. Hyun Jung Kim