This PDF file contains the front matter associated with SPIE Proceedings Volume 10629, including the Title Page, Copyright information, Table of Contents, and Conference Committee listing.

Within the framework of the first European Defence Agency (EDA) call for protection against chemical, biological, radiological and nuclear threats (CBRN Protection) we established a project on active multispectral reflection fingerprinting of persistent chemical agents (AMURFOCAL). A first paper on the project AMURFOCAL has been issued last year on the SPIE conference in Warsaw, Poland. This follow up paper will be accompanied by an additional paper that deals specifically with the aspect of the 100 W-level peak power laser system tunable in the LWIR. In order to close a capability gap and to achieve detection at stand-off distances our consortium built a high peak power pulsed laser system with fast tunability from 8 to 10 μm using an external-cavity quantum cascade laser and optical parametric amplification. This system had to be tested against different substances on various surfaces with different angles of inclination to evaluate the ability for an active stand-off technology with an eye-safe laser system to detect small amounts of hazardous substances and residues. The scattered light from the background surface interferes with the signal originating from the persistent chemicals. To account for this additional difficulty new software based on neutral networks was developed for evaluation. The paper describes the basic setup of the instrument and the experiments as well as some first results for this technology.

The ability to detect and ultimately successfully identify threat agents, including chemical warfare agents (CWAs), remains a priority in both military and domestic environments globally. A number of detection technologies exist, however, active infrared (IR) hyperspectral imaging offers a number of operational advantages, making it a desirable deployment option. Most importantly, active hyperspectral imaging offers a stand-off detection approach, removing the operator and equipment from close proximity to the potentially lethal threat agents.

M Squared’s Negative Contract Imager (NCI) is a turn-key chemical sensor, built on a flexible technology platform, making it suitable for a range of chemical sensing applications. A key benefit of the NCI platform is that it can be operated remotely, further removing the operator from the potentially hazardous environment. The NCI used for the here presented results operates in the mid wave IR (nominally 2.7 μm to 3.7 μm), allowing the fundamental absorption bands of agents to be addressed and analyzed. The presented work demonstrates how the NCI was able to successfully detect and identify a number of CWAs deposited on a range of surfaces. A key challenge to positive identification of a threat is analyzing the recorded absorption signature from the sensor and comparing it to reference signatures. Depending on the nature of the deposition of the agent the resulting absorption features can be distorted compared to the reference signature. The discussed results demonstrate how spectral fitting algorithms can be used to assist in agent identification.

Raman hyperspectral imaging (HSI) can be an advantageous technique for the detection and identification of threat materials (i.e. homemade and military grade explosives), especially if those materials are located in a complex scene. Raman spectroscopy has the capability to provide a distinct molecular fingerprint of a threat material, which gives it the ability to provide near unambiguous threat identification. Unfortunately, the current generation of Raman sensors have numerous limitations that hinder their performance and limit their ability to be applied in real world scenarios. These systems offer low optical throughput, have larger size/weight requirements, and can only interrogate an area of interest the size of a focused laser spot. These limitations are typically due to a system’s spectrometer, which traditionally utilizes a dispersive grating and requires a narrow entrance slit width and long focal length optics to accurately accept and pass the collected scattered light onto the detector. In addition, the use of focused laser excitation creates eye-safety concerns that restrict the usage of Raman sensors for most real-world applications. With these issues in mind, ChemImage Sensor Systems (CISS) is developing a next generation Raman sensor capable of providing a wide-area of coverage and improved eye-safety using defocused laser excitation. This is made possible by utilizing a spatial heterodyne spectrometer (SHS), a slitless grating-based Michelson interferometer with no moving parts. The entrance aperture to the SHS can be centimeters in diameter, which provides the SHS an etendue orders of magnitude greater than a traditional spectrometer. This feature also allows the excitation laser to be defocused to centimeters in diameter. In addition, the sensor utilizes a fiber-array spectral translator (FAST) bundle, a 2-D hyperspectral imaging fiber composed of hundreds of smaller fibers, which gives the sensor the ability to spatially distinguish the area of interrogation. The combination of these two technologies is termed FAST-SHS. This paper will discuss the background of spatial heterodyne spectroscopy and Raman hyperspectral imaging, the initial setup and design of the sensor, and provide initial detection results.

A significant remaining challenge in chemical detection is the ability to rapidly detect with high fidelity a full suite of CWAs and TICs in a single point-detection system. Gas chromatography (GC) is a proven laboratory technique that can achieve the stated detection goal, but not at the required speed and not in a wearable (or even portable) form factor. Efforts in miniaturizing GCs yielded small devices, but they remain slow as they retain the end-of-column detection paradigm which results in long elution times of CWAs and TICs. We describe a novel concept of in-column detection by probing the sorbent coating (stationary phase) of a micro-GC column through optical evanescent field interactions in the long-wave infrared (“chemical fingerprint”) spectral region (U.S. Patent US9599567B2). Detection closer to the injection port ensures a rapid response for slow-eluting analytes. Although this results in poor separation (i.e. poor ability to identify chemicals), this is more than compensated by having full IR absorbance spectra at each location. This orthogonal spectral signature (along with GC retention times) is used in a powerful algorithm to quickly identify components in a complex mixture under conditions of incomplete separation. We present results with an ATR-based system that uses a focused tunable quantum cascade laser beam directed by galvo mirrors at points along a molded micro-GC column whose bottom wall is the sorbent coated ATR prism. Efforts are under way to further miniaturize this device by employing novel long-wave-IR photonic waveguides for a truly portable integrated photonic chromatographic detector of CBRNE threats.

UV Raman hyperspectral datacubes were acquired by filtered imaging of a laser illuminated scene. The laser excitation wavelength was scanned over the Stokes Raman band relative to fixed narrow bandpass optical filters. Two different 0.3 nm wide bandpass filters with center wavelengths 248.4 and 264.1 nm were used and the excitation wavelength was scanned in steps of 0.2 nm. Results are presented for persistent chemical warfare agents and simulant chemicals on different surfaces.

A standoff chemical detection system is being developed to detect and identify a wide range of trace chemicals on a variety of natural and artificial surfaces. The system is based on active mid-infrared (MIR) hyperspectral imaging in which the target surface is illuminated using miniature, rapidly tunable, external-cavity quantum cascade lasers (ECQCLs). These lasers are tuned across the wavelength range of 7.7 – 11.8 μm while a HgCdTe camera captures images of the reflected light. Hypercubes with 128x128 pixels and more than 130 wavelengths are captured within 0.1 s. By operating the camera in sub-window mode, hypercubes with 16x96 pixels and 138 frames are captured in only 14 ms. To the best of our knowledge, these represent the world’s fastest acquisition of active MIR hypercubes. Raster-scanning of the laser beam is used to scan large regions. In this talk, we will present results for detecting traces of solid chemicals (with loadings on the order of 100 μg) on natural outdoor surfaces such as roofing shingles, concrete, sand, and asphalt at a standoff distance of 5 m. The measured spectra are found to correlate very well with those of reference measurements made of pure chemicals after accounting for the substrate reflectance.

The global defense community requires new approaches for standoff detection of chemical, biological, radiological, nuclear and explosive (CBRNE) threats. Such standoff detection methods must be capable of discriminating the target hazardous materials from the environmental background. Therefore these sensors must exhibit high selectivity. High selectivity detection of CBRNE threats can be accomplished using infrared (IR) spectroscopy, which produces a unique spectral “fingerprint” of the target chemical, enabling discrimination of the target chemical from other chemicals in the background. Standoff detection using IR spectroscopy however requires that enough of the incident source light may be collected at the detector; therefore a high-power source is needed. Commercially available quantum cascade laser (QCL) sources are capable of projecting high power, coherent laser light at targets down range from the source. In order to collect complete IR spectra throughout the entire fingerprint region, the output of multiple QCL modules are combined into a single exit aperture. This is typically achieved using mirrors and other optics which are susceptible to vibrational and temperature misalignments in field systems. In order to provide a more ruggedized solution to combining the beam output of multiple QCL modules, we developed a unique chalcogenide optical fiber beam combiner which combines the output of four commercial QCL modules. This allows for scanning across a spectral range from 6.01 – 11.20 μm encompassing parts of both the IR functional groups and fingerprint regions. We demonstrate the ability of this QCL system to generate high quality IR spectra of hazardous materials.

One of the greatest dangers for first-responders when approaching a disaster site is their lack of incident-related information. This problem can be overcome by deploying and operating a range of high-performance sensors mounted on an unmanned aerial vehicle (UAV). For an initial survey and subsequently detailed reconnaissance of disaster areas where radiation sources represent a threat for rescue operations, the following specialized equipment can be used: the RIEGL VUX-1UAV laser scanner, a highly sensitive gamma radiation probe provided by CBRN Protection and the RiCOPTER-M UAV. An embedded processing and radio communication system integrated onboard the UAV enables real-time access to georeferenced 3D LiDAR pointclouds and gamma radiation levels. Precise localization of the gamma radiation sources and simulation of corresponding radiation activity patterns is achieved by automatic processing on the ground, taking the up-to-date topography into account. The results are displayed to the person in charge of response forces in an intuitive and user-friendly way on, e.g., a tablet computer in real-time and while the UAV is still in the air. Resolution and precision are continuously increased by a semi-autonomous flight path generation approach. This takes into account real-time radiation measurements in order to fly additional lines over automatically detected locations of specific interest. We present results of extended field tests with live radiation sources to demonstrate real-time data acquisition, processing, and refinement. The system shown represents a highly flexible and possibly life-saving asset for first-responders in time-critical scenarios.

The transition metal oxide embodied organic composites have great promise for high energy radiation detection. The interaction of high energy radiation such as γ-rays with the organic composite can generate photoelectric responses, Compton scattering and electron hole pairs, which can provide favorable properties to enhance the radiation detectivity of the composite. These effects along with changes of oxidation state of metal oxides, provide significant change in the electrical characteristics of composites due to radiation exposure. We have developed nickel oxide (NiO2) nanoparticles embodied urea composite (urea-NiO₂), and determined effect of γ-radiation on the current – voltage characteristics in the frequency range of 100 Hz to 100,000Hz. In this paper, we describe the results of effect of additional oxidizing agent MnO₂ (urea-NiO₂-MnO₂) on the morphology, processing and current voltage characteristics due to exposure of Cs-137 γ-radiation. It was observed that addition of MnO₂ in urea-NiO₂ composite decreases the sensitivity of detection. However, urea-NiO₂-MnO₂ composite recovers to original properties after irradiation much faster than urea-NiO₂ composite.

Studies of amorphous silica fibers bombarded with neutral particles indicate that both photons and neutrons have the same mechanism for creating defects, however rate of creation and affected wavelengths pertaining to each type remains unclear. It is difficult to positively attribute defects to one or the other when both are introduced in the sample concurrently. We sought to mitigate this issue in the current experiment by placing lead shielding of various thicknesses in the line of radiation from a nuclear reactor to the Yb-doped fiber, which is then exposed to neutrons and photons from a nuclear reactor source. Reducing photon fluence via various thicknesses of high-Z shielding materials, while maintaining the same neutron fluence, provides for a comparison of defect formation rates in YDF. In comparing neutron dominant and gamma dominant radiation sources for a similar total dose, the absorption spectrum of the 20/400 YDF deconvolves to several Gaussian peaks, with a 0.1 eV shift for the 1-1.5 eV peak and a 0.15 eV shift for the 2-2.5 eV peak (lower energies). Noticeably, the gamma radiation dominant source absorption spectrum shows a peak around 1.6 eV significantly more pronounced than in the neutron radiation dominant source dataset.

Law enforcement officers and public safety personnel are a critical component of the Global Nuclear Detection Architecture, and would benefit from additional opportunities to train for this mission in realistic threat scenarios. Physical Sciences Inc. (PSI) is developing a Virtual Source Training Toolkit (VSTT) system capable of reproducing the response of handheld radiation detectors to a virtual source in a complex occlusion and shielding environment. The toolkit will allow additional low-cost training opportunities for these officers inside operationally relevant public areas in order to reduce the time required to detect and localize a realistic radiological threat. The main components of the VSTT are a user position estimation system and a radiation propagation algorithm. Both algorithms operate at 10 Hz update rate on a handheld Android smart device that simulates the user interface of a radiation detector. The user position and orientation are determined through a Bayesian fusion process between the smart phone IMU measurements and range estimates to Bluetooth beacons. The radiation propagation algorithm simulates both attenuation and scattering of radiation between the programmed virtual source position and the user’s estimated position. The VSTT has been demonstrated to provide an average localization error < 1.2 m while traversing a complex interior space including walls and magnetic perturbations. The simulated radiation spectra achieve Spectral Angle Mapping values < 0.93 between simulated and measured source configurations through multiple shielding materials and thicknesses. In a series of experiments, an operator is able to rapidly localize a virtual source using a prototype VSTT.

Uranium Ore Concentrate (UOC, often called yellowcake) is a generic term that describes the initial product resulting from the mining and subsequent milling of uranium ores en route to production of the U-compounds used in the fuel cycle. Depending on the mine, the ore, the chemical process, and the treatment parameters, UOC composition can vary greatly. With the recent advent of handheld spectrometers, we have chosen to investigate whether either commercial off-the-shelf (COTS) handheld devices or laboratory-grade Raman instruments might be able to i) identify UOC materials, and ii) differentiate UOC samples based on chemical composition and thus suggest the mining or milling process. Twenty-eight UOC samples were analyzed via FT-Raman spectroscopy using both 1064 nm and 785 nm excitation wavelengths. These data were also compared with results from a newly developed handheld COTS Raman spectrometer using a technique that lowers the background fluorescence signal. Initial chemometric analysis was able to differentiate UOC samples based on mine location. Additional compositional information was obtained from the samples by performing XRD analysis on a subset of samples. The compositional information was integrated with chemometric analysis of the spectroscopic dataset allowing confirmation that class identification is possible based on compositional differences between the UOC samples, typically involving species such as U₃O₈, α-UO₂(OH)₂, UO₄•2H₂O (metastudtite), K(UO₂)₂O₃, etc. While there are clearly excitation λ sensitivities, especially for dark samples, Raman analysis coupled with chemometric data treatment can nicely differentiate UOC samples based on composition and even mine origin.

Matrix assisted laser desorption ionization (MALDI) is a powerful technique that improved the mass spectrometry (MS) characterization of biological molecules. However this technique requires the mixing of matrix compound with the analyte of interest. The matrix compound used in MALDI process is not universal and usually depends heavily on the nature of analyte of interest being analyzed. As such there are many matrices that are used and without knowing the nature of your analyte it will be hard to predict which matrix is optimal for the most effective MALDI-MS analysis. Moreover, a high energy laser exposure is needed to initiate the ionization process through a charge transfer process between the matrix and analyte molecules. Recent advancement in the metalorganic framework (MOF) field introduced desirable surfaces that can be modified for various applications. Such MOFs can be synthesized with porous solid, and could have regular or predicted geometry. This project is introducing a novel idea of utilizing a modified MALDI substrate with MOF that can provide charge transfer between immobilized functionalized groups and analyte molecules that mimic the solvation process when a solution matrix is used. Begin the abstract two lines below author names and addresses.

Infection with the spirochete Borrelia burgdorferi leads to Lyme disease, the most common tick-borne disease in North America, Europe, and Asia. Currently, Lyme disease is diagnosed using a two-tiered approach of ELISA/immunofluorescence, followed by Western blot analysis. These assays measure serological immune response to the infection, namely levels of IgG or IgM antibodies that bind to B. burgdorferi antigens. However, the existing approach is non-quantitative, lacks sensitivity, and may contribute to delayed diagnosis. In this study, grating-coupled fluorescence plasmonics (GC-FP) was used for rapid, highly-multiplexed detection of antibodies that bind B. burgdorferi proteins in human and mouse blood serum. GC-FP is an optical plasmonic method that enables quantitative detection of molecular interactions and can be incorporated into microfluidic format for highly multiplexed testing. We have demonstrated that this technique allows us to use only three microliters of blood serum to quantitatively detect multiple target antibodies within 30 minutes. We have also shown that GC-FP is faster and more sensitive than the traditional two-tiered Lyme disease testing scheme, making it attractive for diagnostic purposes. This proof-of-concept study provides foundations to develop GC-FP as a highly sensitive diagnostic tool to enhance the efficiency of assessment for Lyme disease patients, which will ultimately improve treatment outcomes.

This paper describes a rapid, reagentless, standoff method of detection and classification of bulk and trace suspicious substances on natural surfaces using solar-blind deep UV excitation and detection. Detection is typically accomplished in less one second. The detection method is solar blind and can be employed at standoff distances up to 5 m or more without interference from natural or man-made light sources. By this method, unknown suspicious powders, that potentially contain biological hazards, are automatically triaged using a four-step sequential iteration of Principal Component Analysis methods using pre-determined eigenvector sets to: 1) detect and differentiate whether a sample is bio or non-bio; 2) whether the detected bio is microbial, protein, or plant; 3) if microbial, whether the sample is a bacterial cell or spore, yeast, fungi, or fungal spore; and 4) to provide some higher level of cellular differentiability. The same method is also applicable to a wide range of chemical agents and explosives materials. The method and related instruments employ sample excitation at 248.6 nm and detection over a spectral range from 250 nm to below 350 nm, a spectral region blind to solar and most man-made light sources. Detection and classification is accomplished in less a few seconds. Sample detection and classification rates can be over 20 per second. Fully integrated and self-contained hand-held instruments are presently under development with an overall weight less than about 8 lbs, including a battery for over 8 hours of typical use. The standoff detection range is nominally 5 cm to 5 m.

Alakai Defense Systems has upgraded its short range UV Raman standoff explosive detection sensors called the Portable Raman Improvised Explosive Detection (PRIED) sensor by adding near trace detection capability. The PRIED sensor is a standoff detection sensor that works at ranges of 1-10m for a wide variety of explosives (50m for some selected chemicals). Recent improvements have focused on expanding PRIED’s capabilities to include near-trace detection of explosives at 1-2m range on fluorescent substrates. Data will be presented showing this new capability along with a brief description of the design upgrades

Raman spectroscopy is a powerful tool capable of identifying unknown materials. In the past 20 years, laser sources and detectors have been getting smaller which has led to the development of handheld Raman sensors for use by the military and first responders. One of the advantages of Raman sensors is that it requires no sample preparation, however the incident laser must be able to interrogate the sample which means that the measurement must be taken in the open air or a transparent container. If an unknown material is found in an opaque container, it is typically transferred into a transparent sample jar which represents a handling hazard for the operator. More recently, a technique known as Spatially Offset Raman Spectroscopy (SORS) has shown the ability to measure Raman signals for materials stored in opaque (non-metallic) containers which would eliminate this hazard. Alternatively, advanced algorithm techniques can be used with traditional epi-illumination laser excitation to extract weak Raman signatures from noisy backgrounds or complex mixtures caused when looking at chemicals stored in Raman-active containers. For this study we compare spectral results obtained from two commercially available Raman instruments which use an epi-illumination configuration and SORS against a common set of transparent and opaque containers.

We are developing a stand-off technique for the detection of trace amounts of explosive materials. The motivation behind this work is to prevent loss of life and injury to military and civilian personal by detecting threats at distance. The matured technique will allow for the facile identification of possible threats with minimum user effort and enough time to take appropriate action. This manuscript illustrates the results from our infrared backscatter imaging spectroscopy mobile stand-off method to detect trace amounts of explosive materials under laboratory conditions. The described technique uses tunable quantum cascade lasers, with full spectral coverage from 6-11 μm, to illuminate a target and an infrared focal plane array to collect the backscattered signal into hyperspectral images cubes. The quantum cascade lasers are operated under eye safe levels which allows for safe and stealthy probing of objects, vehicles, and even people. Experiments are performed on tilted substrates to simulate real world conditions where it is unlikely to collect the specular reflections. The collected hyperspectral image cubes contains spectral, spatial, and temporal information that can be fed to a detection algorithm.

Field detection of chemical, biological, radiological, nuclear and explosive (CBRNE) threats requires the development of highly selective sensors with low size, weight, power and cost (SWaP-c). Recent developments have demonstrated that an optical biomimetic sensing approach, based on human-eye color detection can provide high-confidence discrimination of target chemicals while rejecting potential interferents with similar chemical structures. This biomimetic sensing method operates by identifying differences in the overlap between target and interferent chemical infrared absorption bands utilizing three, overlapping, optical bandpass filters. This method is non-spectroscopic and requires only the use of commercially available, off-the-shelf optical components. This approach has been demonstrated for volatile chemical vapors in the mid-wave-infrared (3 – 5 μm). Based on this success, experimental studies of this biomimetic sensing approach have been expanded further into the long wave infrared spectral region (6 – 12 μm) and for detection of explosives on surfaces, including aluminum and plastics. We present discrimination results using this biomimetic sensing approach for explosive samples on surfaces in both the mid- and long- wave infrared. Numerical data, along with experimentally collected data, are discussed. We demonstrate that this method is capable of discriminating between similar explosives on surfaces as well as between these explosives and potential environmental interferents. We present the results of these experiments and discuss potential transition of this approach to future field-ready stand-off devices and applications.

Optical whispering gallery mode (WGM) biochemical sensors operate by tracking changes in resonant frequency as materials enter the evanescent near-field of the resonator. To achieve the smallest limit of detection, it is desirable for WGM sensors to exhibit as large a frequency shift as possible for a material of a given size and refractive index, as well as the ability to track as small a frequency shift as possible. Previously, plasmonic nanoantennas have been coupled to WGM resonators to increase the magnitude of resonance shifts via plasmonic enhancement of the electric field, however this approach also results in increased scattering from the WGM, which degrades its quality factor, making it less sensitive to extremely small frequency shifts. This degradation is caused by the ohmic and scattering dissipation caused by metallic nanoantennas. Using simulations, we show here that the precise positioning of nanoantennas coupled to a microtoroid WGM resonator can be used to overcome this drawback and achieve ultrahigh-Q plasmonic cavity modes simultaneously with electric field enhancement. It is shown that a nanoantenna composed of two similarly coupled nanorods supports higher Q modes than a single nanorod antenna. Our results have immediate application in the context of optical sensing.

Ultra-high-quality, whispering-gallery-mode (WGM) optical micro-resonators provide an unprecedented capability to trap light in a highly confined volume, smaller than a strand of human hair. A beam of light can travel the boundary of a WGM resonator over 10⁶ times, significantly enhancing light-matter interactions for a variety of sensing applications. Our group has leveraged WGM sensors for a wide array of applications, including the detection of infrared light, gas sensing, photo-acoustic imaging, and nanoparticle detection and sizing. We have also developed a new hand-held microresonator platform that can be combined with commercial devices such as drones, as well as a fiber-based, WGM microprobe sensor.

U.S. Army Dugway Proving Ground (DPG) is a major defense test range located in the remote west desert of Utah, USA. DPG is made up of various testing facilities, extensive test grids, and impact areas. DPG’s mission is testing for chemical and biological defense. Recently, a series of large-scale chlorine releases were held at DPG, known as the Jack Rabbit II test program. The purpose of the testing was to better define public safety parameters in the event of a large-scale chlorine release. DPG deployed 100s of point sensors to quantify the test events. Three single-wavelength UV lidar systems were also developed and deployed with the goal of providing a more overall picture of these events. This was an experimental effort using principles similar to Differential Absorption Lidar (DIAL) to estimate chlorine concentration and track clouds downrange. Lidar systems are typically configured with two wavelengths for DIAL measurements. As our effort was experimental and had very limited funds, we used on hand ND:YAG lasers at the 355 nm wavelength only. The second wavelength was later simulated from portions of the data in which no chlorine was present. The main assumption made in using only a single wavelength was that very limited aerosols and other types of chemicals would be mixed with the chlorine cloud. This single-wavelength approach was found to be an effective method for tracking absorbing chemical vapors. We obtained an overall picture of the test event and were able to estimate concentrations in post processing.

The system and mechanical design of a four-wavelength lidar system is described. The system is designed to be maximally adaptive to deployment scenario in terms of both size/weight/power and detection application. The wavelengths included in the system are 266 nm, 355 nm, 1064 nm, and 1574 nm – all generated from Nd:YAG based pump laser sources. The system is designed to have a useful range from 400 meters to 5,000 meters, depending on the wavelength and atmospheric conditions.

Dugway Proving Grounds (DPG) plays a key role in the open-air field-testing of systems used in defense against chemical and biological threats. The performance of systems under test are benchmarked against a suite of wellcharacterized point and standoff instrumentation. Elastic-backscatter lidar systems with large power-apertures operating at 1.06 μm provide standoff detection, quantification, and location of aerosol plumes. The accuracy and sensitivity these systems provide comes at the cost of a large NOHD (>5 km) which limits their utility. To this end, Space Dynamics Lab (SDL) developed an eye-safe system following system requirements from DPG. The system provides a standoff capability for field tests where a NOHZ and required PPE would be an undue burden. CELiS (Compact Eye-Safe Lidar System) is an elastic-backscatter lidar that operates at 1.57 μm, using a commercial 30 Hz Nd:YAG laser and OPO combination. The short pulse length and low repetition rate give the system an advantage in range resolution and daytime operation over a similarly sized system based on a fiber laser. CELiS uses LidarView, an SDL-developed lidar display package, for data acquisition and hardware control. The Joint Ambient Breeze Tunnel (JABT) is used to perform calibration and sensitivity measurements of the various lidar systems at DPG. The JABT provides confinement of an aerosol plume and allows for comparison of TSI APS (Aerodynamic Particle Sizer) concentrations to the lidar backscatter values over an extended period. CELiS was used to support a recent JABT test and the data analysis and performance results from the test are described.

Laser absorption spectroscopy utilizes a tunable infrared source, providing the necessary selectivity, to detect the characteristic fingerprint spectral absorption of an abundant gas. In a simple embodiment such as single-pass absorption, sensitivity is limited as attenuation becomes minuscule for trace level concentrations; a problem exacerbated in the midinfrared region due to significant detector noise. Sensitivity can be improved by increasing interaction between the optical field and molecular ensemble with methods such as a multiple-pass Herriot cell or resonant cavity ring-down spectroscopy but these techniques have a substantial overhead in instrumentation. An alternative approach to this problem is Phase Fluctuation Optical Heterodyne (PFLOH) spectroscopy. Here, interferometric effects are used to detect the minute heating of the sample gas when incident laser light of the appropriate wavelength is absorbed. More specifically, by placing the absorption chamber within one arm of a Mach-Zehnder interferometer, heat-induced changes in the optical path length can be detected with great sensitivity through the resulting fringe modulation. A secondary benefit is that although excitation occurs in the infrared, its effects can be detected using visible lasers and silicon detectors, thereby obviating the need for cooled, infrared detectors. We will present our results used to detect ethane using absorption in the 3.33-3.37 μm region. The Mach-Zehnder interferometer used a Helium Neon laser for the probe laser, and a broadly tunable Optical Parametric Oscillator (OPO) for spectroscopic excitation. We have demonstrated detection levels at parts per billion with further sensitivity possible by implementing several identified improvements.

The output of a laser frequency comb is composed of 100,000+ perfectly spaced, discrete wavelength elements or comb teeth, that act as a massively parallel set of single frequency (CW) lasers with highly stable, well-known frequencies. In dual-comb spectroscopy, two such frequency combs are interfered on a single detector yielding absorption information for each individual comb tooth. This approach combines the strengths of both cw laser spectroscopy and broadband spectroscopy providing high spectral resolution and broad optical bandwidths, all with a single-mode, high-brightness laser beam and a simple, single photodetector, detection scheme. Here we show that this novel spectroscopy source can be employed for regional (~kilometer squared) monitoring using an array of stationed retros or in conjunction with an unmanned aerial systems (UAS). Both fixed and UAS systems combine the high-precision, multi-species detection capabilities of open-path DCS with the spatial scanning capabilities to enable spatial mapping of atmospheric gas concentrations. The DCS systems measure the atmospheric absorption over long, 100m to 1 km, open air paths with 0.007cm-1 resolution over 1.57 to 1.66 um, covering absorption bands of CO2, CH4, H2O and isotopologues.

Past chemical warfare agent attacks often mention the use of mixtures of chemicals or impure-incomplete formulation. Terrorist groups may also generate new chemical toxic agents. Those situations involve unknown compounds and thus may be undetectable by traditional methods. Indeed, standoff gas detection with infrared devices traditionally relies on the comparison between measured signal with a library of signals included in a database. Observing the gas absorption in infrared band III (LWIR 8-14 μm), our multispectral infrared camera is used to detect gas clouds up to a range of several kilometers, to provide identification of gas type and to follow the motion of the cloud in real time. The approach described in this paper develops an algorithm that enables the device to detect gas even if the measured signature is not in the database – pattern-matching-free algorithm. This detection process has been evaluated in the laboratory and subjected to significant experimental feedbacks. The results are a capability to detect unknown gases and gas mixtures.

We present a portable optical spectrometer for fugitive emissions monitoring of methane (CH4). The sensor operation is based on tunable diode laser absorption spectroscopy (TDLAS), using a 5 cm open path design, and targets the 2ν₃ R(4) CH₄ transition at 6057.1 cm^-1 (1651 nm) to avoid cross-talk with common interfering atmospheric constituents. Sensitivity analysis indicates a normalized precision of 2.0 ppmv·Hz-^1/2, corresponding to a noise-equivalent absorbance (NEA) of 4.4×10^-6 Hz^-1/2 and minimum detectible absorption (MDA) coefficient of αmin = 8.8×10^-7 cm^-1·Hz^-1/2. Our TDLAS sensor is deployed at the Methane Emissions Technology Evaluation Center (METEC) at Colorado State University (CSU) for initial demonstration of single-sensor based source localization and quantification of CH₄ fugitive emissions. The TDLAS sensor is concurrently deployed with a customized chemi-resistive metal-oxide (MOX) sensor for accuracy benchmarking, demonstrating good visual correlation of the concentration time-series. Initial angle-ofarrival (AOA) results will be shown, and development towards source magnitude estimation will be described.

An effective response to human biowarfare agent exposure events requires the availability of simple, sensitive, reliable, and manufacturable sensing and diagnostic tools. While ring resonators fabricated on a silicon-on-insulator platform have found wide application as enabling components for biosensors, and have even been commercialized successfully, silicon nitride-based ring resonators have received less attention. We hypothesized that silicon nitride would provide both manufacturing and performance advantages over silicon in a biosensing context. To test that hypothesis, we designed a series of silicon nitride ring resonators. Designs were fabricated at the American Institute for Manufacturing Integrated Photonics (AIM Photonics) foundry. We will discuss the design process, optical performance of the manufactured devices, and their use in the label-free detection of biomedically relevant protein targets.

Recent advances in subwavelength grating (SWG) structures have shown promise in sensing applications. Though prior sensors have obtained high index sensitivities, their designs have not focused on providing strong analyte/light interaction efficiency for low analyte concentration flows. We have explored high-contrast grating “fish-bone” and segmented SWG structures in the hopes of improving this aspect. We present below the progress of our design exploration of these structures and compare their performance to a typical slot waveguide to understand their impact on spectral transmission and analyte interaction. The best performing structures will be experimentally functionalized with label-free analyte capture materials to sense specific threat molecules via index change. We envision building manyanalyte multiplexed sensor arrays from these devices to enable simultaneous monitoring of multiple chemical and biological threats or human biomarkers with high sensitivity, specificity, and low probabilities of false alarm.

In-situ gas analysis was demonstrated using a mid-infrared (mid-IR) microcavity. Optical apertures were made of ultrathin silicate membranes using the complementary metal-oxide-semiconductor (CMOS) process. Fourier transform infrared spectroscopy (FTIR) shows that the silicate membrane is transparent in the range 2.5 - 6.0 μm, overlapping with gas absorption lines and therefore enables gas detection applications. CH4, CO2, and N2O were selected as analytes due to their strong absorption bands corresponding to functional group stretching: C-H, C-O, and O-N, respectively. A short response time of subsecond and high accuracy of gas identification were achieved. The chip-scale mid-IR sensor is a new platform for an in-situ, remote, and embedded gas monitoring system.

We present accurate measurements for the optical constants for a series of organic liquids, including organophosphorous compounds. Bulk liquids are rarely encountered in the environment, but more commonly are present as droplets of liquids or thin layers on various substrates. Providing reference spectra to account for the plethora of morphological conditions that may be encountered under different scenarios is a challenge. An alternative approach is to provide the complex optical constants, n and k, which can be used to model the optical phenomena in media and at interfaces, minimizing the need for a vast number of laboratory measurements. In this work, we present improved protocols for measuring the optical constants for a series of liquids that span from 7800 to 400 cm-1. The broad spectral range means that one needs to account for both strong and weak spectral features that are encountered, all of which can be useful for detection, depending on the scenario. To span this dynamic range, both long and short cells are required for accurate measurements. These protocols are presented along with experimental and modeling results for thin layers of silicone oil on aluminum.

Prototype simulations of attenuated total re ection (ATR) applied for infrared molecular binding spectroscopy, which is for nerve-agent detection, are presented. The simulations use: calculated estimates of permitivity functions (for the custom sorbent SiFA4H, nerve agent simulant DMMP and molecular structure SiFA4H-DMMP); and a model of re ection from multicomponent-multilayer systems, which is based on the scattering-matrix representation of electromagnetic-wave propagation. The physical assumptions and approximations underlying these simulations, and model-parameter sensitivity are examined with respect to quantitative prediction of ATR spectra associated with nerve-agent detection. Experimentally measured ATR spectra are utilized for qualitative comparison and quantitative adjustment of model parameters.

This study examines using simulated spectra for analysis of diffuse IR reflectance from explosive materials that are sparsely distributed upon a surface. The simulated spectra are calculated using ensembles of reflectance spectra for non-interacting explosive particles on surfaces, which have specified dielectric response properties and particle-size distributions. Reflectance spectra for individual particles upon a surface are calculated numerically using a model based on Mie scattering theory, which assumes a spherical particle on a surface. This validation study considers a prototype system comprising a sparse distribution of RDX particles upon a soda-lime glass surface compared with experimental results.

Results are presented for the detection of trace explosive residues on real-world surfaces using active mid-infrared (MIR) hyperspectral imaging. The target surface is illuminated using miniature, rapidly tunable, external-cavity quantum cascade lasers (EC-QCLs) and the reflected light is imaged using a HgCdTe camera with a spatial resolution of 70 μm. Hypercubes with 128x128 pixels are captured with more than 256 wavelengths that span 7.7 – 11.8 μm. The samples consisted of PETN residues which were applied to keyboard keys at various levels of chemical loading. We estimate a limit of detection of less than 6 ng per pixel for the as-deposited chemical. The explosive residue remains detectable by HSI even after wiping the surface several times using isopropyl alcohol. Simple signature models for solid particles (i.e., Mie scattering) and thin-films account for the many of the spectral features observed in the chemical signatures.

We present the rationale, methods, and results of modeling of thin film organic liquids on various substrates. These liquids may coat surfaces (substrates) either as a result of their production, dispersal via aerosols or spills. Identification of unknown coated surfaces using either reflectance or emittance spectroscopy cannot be accomplished simply through reference to reflectance signature libraries since neither the thickness of the liquid layer nor the substrate type is known beforehand and both contribute to the signature. Liquid spectral libraries offer the complex index of refraction (n,k) as a function of wavelength which by itself is useful only for thick (bulk) liquid layers via computation of reflectance and transmittance coefficients using the Fresnel equations. Thin liquid layers both reflect and refract incident light in combination with reflectance from the substrate. We show modeling of various organic liquids on substrates using commercial thin film design and modeling software, as well as Monte Carlo ray tracing software to demonstrate the variety of potential signatures encountered that depend on the thickness of the liquid layer as well as the characteristics of the substrate (metal or dielectric). These substrates give rise to transflectance behavior, while many dielectric substrates have rich absorption features that provide complex signatures that combine attributes of both the liquid and the substrate. Knowledge of the complex index of refraction of both target liquids and substrates is essential in order to synthesize spectra necessary in the application of target identification algorithms.

Infrasound propagation through various atmospheric conditions and interaction with environmental factors in- duce highly non-linear and non-stationary effects that make it difficult to extract reliable attributes for classi- fication. We present featureless classification results on the Library of Typical Infrasonic Signals using several deep learning techniques, including long short-term memory, self-normalizing, and fully convolutional neural net- works with statistical analysis to establish significantly superior models. In general, the deep classifiers achieve near-perfect classification accuracies on the four classes of infrasonic events including mountain associated waves, microbaroms, auroral infrasonic waves, and volcanic eruptions. Our results provide evidence that deep neural network architectures be considered the leading candidate for classifying infrasound waveforms which can directly benefit applications that seek to identify infrasonic events such as severe weather forecasting, natural disaster early warning systems, and nuclear weapons monitoring.

Infrasonic waves continue to be a staple of threat identification due to their presence in a variety of natural and man-made events, along with their low-frequency characteristics supporting detection over great distances. Considering the large set of phenomena that produce infrasound, it is critical to develop methodologies that exploit the unique signatures generated by such events to aid in threat identification. In this work, we propose a new infrasonic time-series classification technique based on the recently introduced Wavelet Scattering Transform (WST). Leveraging concepts from wavelet theory and signal processing, the WST induces a deep feature mapping on time series that is locally time invariant and stable to time-warping deformations through cascades of signal filtering and modulus operators. We demonstrate that the WST features can be utilized with a variety of classification methods to gain better discrimination. Experimental validation on the Library of Typical Infrasonic Signals (LOTIS)—containing infrasound events from mountain associated waves, microbaroms, internal atmospheric gravity waves and volcanic eruptions—illustrates the effectiveness of our approach and demonstrate it to be competitive with other state-of-the-art classification techniques.

An active, standoff, all-phase chemical detection capability has been developed under IARPA’s SILMARILS program. The detection platform utilizes reflectance spectroscopy in the longwave infrared coupled with an automated detection algorithm that implements physics-based reflectance models for planar chemical films, particulate in the solid and liquid phase, and vapors. Target chemicals include chemical warfare agents, toxic industrial chemicals, and explosives. The platform employs broadband Fabry-Perot quantum cascade lasers with a spectrally selective detector to interrogate target surfaces at tens of meter standoff. A statistical F-test in a noise whitened space is used for detection and discrimination over a large target spectral library in high clutter environments.

The capability is described with an emphasis on the physical reflectance models used to predict spectral reflectivity signatures as a function of surface contaminant presentation and loading. Developmental test results from a breadboard version of the detector platform are presented. Specifically, solid and liquid surface contaminants were detected and identified from a library of 325 compounds down to 30 μg/cm² surface loading at a 5 m standoff. Vapor detection was demonstrated via topographic backscatter.

We report an optical molecular gas sensor exhibiting high levels of selectivity and sensitivity. The outstanding sensitivity demonstrated by our technology is rooted in a novel combination of photoacoustic spectroscopy (PAS) operated within the cavity of a continuous-wave, intra-cavity Optical Parametric Oscillator (OPO). We exploit the very high circulating field present within the resonant down-converted cavity as the excitation source of the photoacoustic effect, conferring orders-of-magnitude improvement in optical excitation power. Additionally, the wide selectivity of the system arises from the inherent broad tunability and narrow optical linewidth of an OPO. Here we report the use of this technology for the detection of ammonia (NH₃) as a simulant target molecule. A 3-D printed miniature PAS cell with microelectromechanical systems based (MEMS) microphone is used for the gas detection. The resonance frequency of the cell was measured at 17.9 kHz with a Q-factor of 9. The down-converted signal wave resonating within its optical cavity was tuned to 6605.6cm^-1 (corresponding to a strong local NH₃ absorption line) through a combination of phase matching and intra-cavity etalon control. The laser was amplitude modulated at the resonance frequency of the PAS cell, producing an average optical excitation power of ~10W in the signal arm of the OPO, to induce the photoacoustic effect for only 4W of primary diode pump power. In this work we show detection limit at the level of single parts-per-billion (ppb). Additionally, we will discuss how this technology could be readily refined to potentially demonstrate a sensitivity of tens parts-per-quadrillion.

Chemical warfare agents (CWAs) such as nerve and blister agents are expected to pose continuing and growing dangers for the Warfighter in the future. We investigate a novel chemical detection modality, based on a new platform for colorimetric detection of chemical threats incorporated in hollow fibers, which are miniature in two dimensions and extendable (“extrudable”) in the third dimension (along the fiber length). By exploring fibers, and films that can be scaled to a fiber geometry, we will enable a new fiber-based chemical threat detector that can serve in textiles worn by the Warfighter (e.g., uniform), as well as in non-worn textiles and an outlying fence or perimeter for early detection of a threat cloud near an expeditionary shelter, outpost, encampment, or base.

Discrimination between different rocket types is an important application for utilizing infrasound in event monitoring within a range of 0-100 km. This is in contrast to traditional nuclear weapons monitoring which leverages infrasound propagation over thousands of kilometers. The motivation of this research is to demonstrate the utilization of deep neural network architectures to discriminate infrasonic signals produced by rocket launches and collected by an near-field infrasound sensor array. The data collection contains three space bound rocket classes: Delta IV, Atlas V, and Falcon 9. In particular, we investigate the classification accuracy of a multi-class convolutional neural network (CNN) and a deep neural network (DNN) on various feature representations, such as neural network derived features, spectrograms, and wavelet scattering transform coefficients. Our experiments validate the viability of a CNN and DNN framework for near-field infrasonic applications, with our proposed method achieving favorable results.