In this keynote presentation to the 20th meeting of SPIE’s CBRNE Sensing Conference, the author will review key papers, trends, and impacts over the past two decades. The author will also present his view on the status of CBRNE sensing within the Department of Defense, potential trends towards development in the future, and gaps in the overall landscape.

This article presents new hyperspectral imaging (HSI) results from a standoff chemical detection system that utilizes monolithic arrays of Distributed Feedback (DFB) Quantum Cascade Lasers (QCLs) as a source, with each array element at a slightly different wavelength than its neighbor. In this rastering approach to HSI, analysis of analyte/substrate pairs benefits from a laser source with characteristics offered uniquely by a QCL Array. In addition to describing the HSI system developed, a description of experimental standoff detection results using the man-portable system from 1.4 meters are presented. We present HSI results on two very different chemical substrate pairs; trace solid PETN on aluminum and the liquid VX on polycarbonate.

The results from infrared backscatter imaging spectroscopy on a mobile platform for stand-off detection of trace amounts of explosive materials on relevant substrates are presented. This technique utilizes an array of tunable infrared quantum cascade lasers to illuminate targets. The spectral range of the QCL system spans from 6 - 11 μm, which enables excitation of a wide variety of absorption bands present in analytes of interest. Targets are prepared by sieving particles through a 20 μm mesh onto substrates to simulate relevant qualities (particle size, fill factor, and mass loading) expected of real world targets. The backscatter signal from targets is collected with an IR focal plane array. This information is stored in a hyperspectral image cube to allow for post processing in a detection algorithm. We demonstrate the selectivity and sensitivity of the discussed technique down to the nanogram level for RDX and PETN on glass. Spectra are generated by extracting the signal from small regions of interest to simulate targets with miniscule coverage areas. Preliminary comparison of backscatter data with simulated data from a model that incorporates particle size, mass loading, and substrate response show good agreement. Confusant agents, such as sand, are introduced to the targets loaded with analyte to illustrate the selectivity of this technique. The results of these studies are presented, along with future improvements to the technique.

As the nature of explosive, chemical and narcotic threats become more pervasive and lethal to innocent bystanders and unsuspecting military and law enforcement authorities, there is a growing demand for rapid and effective detection of materials in real-time with a high degree of autonomy at safe distances. In an effort to address this need, ChemImage has been developing novel, adaptable, short-wave infrared (SWIR) molecular chemical imaging systems for real-time analysis of complex environments, including for detection of hazardous materials (e.g., explosives, chemical warfare agents, drugs of abuse). At the heart of these systems is the Conformal Filter (CF), which is a liquid crystal (LC)-based tunable filter that transmits multi-band waveforms. Building on concepts of multivariate optical computing, the CF is tuned electro-optically and dynamically to mimic the functionality of a discriminant vector for classification. The resulting integrated detector response approximates the detection response of conventional hyperspectral imaging with only two discrete measurements instead of hundreds to thousands. Real-time detection is achieved by operating two CFs in tandem within a dual polarization (DP) system, which exploits the polarization sensitivity of the LC filters and allows for simultaneous acquisition of the compressed hyperspectral imagery. This paper will discuss the development, characterization, and test results of a prototype, handheld CF sensor, with a focus on its application to explosives, chemical and narcotic threat detection.

Infrared spectroscopy is a powerful laboratory technique for detection of and discrimination between hazardous chemical threats, such as explosive materials, and non-hazardous background chemicals. Transitioning this method to the field however has significant challenges, especially when being deployed as a standoff, or proximate standoff technique. The main issue for proximate standoff detection lies in the ability of a system to collect enough reflected light from a source illuminating a surface to provide chemical information for the target material. This issue is further exacerbated when trying to detect highly scattering materials on surfaces, such as particulates or powders, which are typical forms for explosive materials. While diffuse reflectance from such materials provides good chemical vibrational absorption band information, when being collected at some distance away from the sample, this scattering provides a significant detection challenge. We present detection results for highly scattering inkjet printed explosive standards, collected at proximate standoff distances (~ 0.5 m) in a laboratory environment. Discrimination between these standards using both conventional infrared spectroscopy, as well as NRL’s unique optical filter based biomimetic sensing approach are discussed. This work is meant to inform the rational development of portable infrared optical sensors for detection of explosives at proximate distances in operational environments.

With the purpose of validating dispersion models, ammonia (NH₃) releases were performed in September 2018 and a network consisting of NH₃ detectors and temperature sensors were positioned in a grid in front of the source. In addition, the test grid was also monitored by a focal plane array imaging system based on a LWIR detector, which was positioned at a safe standoff distance of 1 km. With this setup, it was possible to monitor the release and the development of the generated cloud during the dissemination, as well as monitoring surrounding areas for risk assessment purposes during and after each challenge. As the observation was performed in near real time (approximately 0.5 Hz frame rate for the measurement, data transfer, Fourier transform and analysis), it was possible to give immediate feedback to the release team and test control personnel. Of special interest are background concentrations below the detection limit, as once these are achieved this indicates whether an area is safe and/or when additional challenges/disseminations can occur.

The detection of chemical hazards on surfaces continues to be a challenge for the sensing community. In order to minimize risk to users, a desirable configuration is a non-contact (standoff) system, which can operate a safe distance from the hazard. A conceptual solution to this challenge is the Wide-Area Mapping and Identification (WAMId) system. The WAMId prototype breadboard combines two distinct technologies, hyperspectral imaging and standoff Raman spectroscopy, operating in tandem to locate and identify anomalous areas of interest and then presumptively identify surface contaminates. In the developed configuration, a single short to mid wave infrared (SWIR/MWIR) hyperspectral camera images a scene of interest, data is processed to locate anomalous materials and the resulting coordinates from the scene are uploaded to a gimbal control which then slews an 830 nm Raman system to perform presumptive identification measurements. In this work, we present the results of the program, to include system development, and sample testing data for three chemicals.

Alakai Defense Systems will present improvements and data from its DUV Raman system. Specifically, this is an Ultraviolet (UV) Raman microscope for the rapid scanning of fingerprints for the detection of trace explosives. Since this sensor operates in the UV spectrum, it can rapidly scan an area mapping out the results in minutes versus tens of minutes to hours for Near-infrared (NIR) systems. A short description of the instrument and performance is presented.

We present a fully integrated photonic chip spectrometer for near-infrared tunable diode laser absorption spectroscopy of methane (CH₄). The integrated photonic sensor incorporates a heterogeneously integrated III-V laser/detector chip coupled to a silicon external cavity for broadband tuning, and a long waveguide element (>20 cm) for ambient methane sensing. An on-chip sealed CH4 reference cell is utilized for in-situ wavelength calibration of the external cavity, and a real-time wavelength compensation method for laser calibration is described and demonstrated. The resulting signal is guided back to the III-V photodiodes for spectral signal readout using a custom-designed acquisition board, remotely controlled and operated by a Raspberry Pi unit. Component-level testing of the waveguide sensitivity, external cavity laser, and reference cell is demonstrated. Full-stack testing of the integrated sensor chip yields sub-100 ppmv∙Hz-^1/2 sensitivity, and spectral density analysis demonstrates our integrated chip sensor to have a fundamental performance within an order of magnitude of commercially available fiber-pigtailed DFB laser units. We envision our integrated photonic chip sensors to provide disruptive capability in SWaP-C (size, weight, power, and cost) limited applications, and we describe an achievable short-term pathway towards sensitivity enhancement to near-10 ppmv levels.

Waveguide-enhanced Raman spectroscopy (WERS) enables the detection and identification of trace concentrations of vapor-phase analytes using a chip-scale photonic circuit coated with a sorbent material. Previous demonstrations of WERS utilized a hydrogen-bond acidic hyperbranched carbosilane fluoroalcohol-based sorbent polymer and focused on detection limits for different nerve agent simulants. In this work, we examine the Raman spectra of a number of new sorbent materials obtained using WERS. By comparing the spectra pre-exposure to the modified spectra measured during analyte exposure, the effects of hydrogen-bonding on the sorbent and analyte molecules are observed. Changes to the Raman transition strength or frequency of individual lines due to analyte binding shed light on the partitioning of vapor-phase molecular agents into the sorbent, and can be used to design sorbent materials with even higher sensitivity. We examine two new types of sorbents: Fluorinated bisphenol-based materials that increase the steric bulk of the substituents ortho- to the hydroxyl group, designed to reduce self-binding; and carbosilane fluoroalcohol polymers synthesized with a novel hydrosilylation reaction. The WERS detection limits for these new sorbents are measured for nerve-agent simulants and compared to previous generation materials.

Enhanced Raman relies heavily on finding ideal hot-spot regions which enable significant enhancement factors. In addition, the termed “chemical enhancement” aspect of SERS is often neglected due to its relatively low enhancement factors, in comparison to those of electromagnetic (EM) nature. Using a metal-semiconductor hybrid system, with the addition of induced surface oxygen vacancy defects, both EM and chemical enhancement pathways can be utilized on cheap reusable surfaces. Two metal-oxide semiconductor thin films, WO3 and TiO2, were used as a platform for investigating size dependent effects of Au nanoparticles (NPs) for SERS (surface enhanced Raman spectroscopy) and PIERS (photo-induced enhanced Raman spectroscopy – UV pre-irradiation for additional chemical enhancement) detection applications. A set concentration of spherical Au NPs (5, 50, 100 and 150 nm in diameter) was drop-cast on preirradiated metal-oxide substrates. Using 4-mercaptobenzoic acid (MBA) as a Raman reporter molecule, a significant dependence on the size of nanoparticle was found. The greatest surface coverage and ideal distribution of AuNPs was found for the 50 nm particles during SERS tests, resulting in a high probability of finding an ideal hot-spot region. However, more significantly a strong dependence on nanoparticle size was also found for PIERS measurements – completely independent of AuNP distribution and orientation affects – where 50 nm particles were also found to generate the largest PIERS enhancement. The position of the analyte molecule with respect to the metal-semiconductor interface and position of generated oxygen vacancies within the hot-spot regions was presented as an explanation for this result.

Exposure to chemicals in everyday life is now more prevalent than ever. Air and water pollution can be delivery mechanisms for toxins, carcinogens, and other chemicals of interest (COI). A compact, multiplexed, chemical sensor with high responsivity and selectivity is desperately needed. We demonstrate the integration of unique Zr-based metal organic frameworks (MOFs) with a plasmonic transducer to demonstrate a nanoscale optical sensor that is both highly sensitive and selective to the presence of COI. MOFs are a product of coordination chemistry where a central ion is surrounded by a group of ligands resulting in a thin-film with nano- to micro-porosity, ultra-high surface area, and precise structural tunability. These properties make MOFs an ideal candidate for gaseous chemical sensing, however, transduction of a signal which probes changes in MOF films has been difficult. Plasmonic sensors have performed well in many sensing environments, but have had limited success detecting gaseous chemical analytes at low levels. This is due, in part, to the volume of molecules required to interact with the functionalized surface and produce a detectable shift in plasmonic resonance frequency. The fusion of a highly porous thin-film layer with an efficient plasmonic transduction platform is investigated and summarized. We will discuss the integration and characterization of the MOF/plasmonic sensor and summarize our results which show, upon exposure to COI, small changes in optical characteristics of the MOF layer are effectively transduced by observing shifts in plasmonic resonance.

Sweat provides direct information of the real-time emotional and cognitive state of the subject, with applications ranging from situational awareness and mission effectiveness of armed forces to disease diagnosis for clinicians. Development of a broad class of human performance monitoring devices to quantify sweat biomarkers necessitates non-invasive, real-time monitoring of ultra-low concentrations (μM to fM) of hormones, proteins, and neurotransmitters. Field effect transistors are the predominant sensor approach whereby the gate electrode is modified with a selective bio-recognition element (BRE). However, FETs have diminished sensitivity in high ionic strength environments associated with sweat. Alternatively, BRE-modified photonic integrated circuits (PICs) have high sensitivity in high ionic strength fluids, low cost at the manufacturing scale, and enable a number of novel device concepts to achieve ultra-low levels of detection. One major technological challenge is to predict the limit of detection (LoD), or sensor response function, for a particular PIC geometry in a microfluidic chamber. LoD is highly dependent on analytic capture efficiency, fluid dynamics and affinity, analyte/light interaction, and analyte concentration. This work presents finite element simulations to emulate microfluidic BRE sweat sensors and provide a predictive limit of detection for different sensing structures or elements. Specifically, the optimum mass transfer and kinetics for sensing approaching single molecule detection is discussed, including flow characteristics, biomarker size, adsorption and desorption kinetics, and sensor geometry. Key metrics include capture efficiency (molecules being captured over molecules entering channel ), time to reach steady state, and temporal adsorption site occupancy to predict PIC system LoD. It is found that these systems are kinetically controlled, with capture efficiencies remaining below 1% even for k_ads/k_des ratios of 10¹⁰ . The need for adsorption kinetics measured for flow systems instead of stationary fluid systems is stressed, as these parameters are what need to be optimized to greatly increase analyte capture.

Trace quantities of explosives left behind by those handling explosives materials present an opportunity to identify both the handlers and the objects they have contaminated. Understanding the evolution of these particles is critical for tailoring detection strategies of optical techniques as well as non-optical contact harvesting methods. We are working towards a complete particle persistence model that captures the contribution of environmental factors such as temperature, airflow, and humidity as well as physical factors such as vapor pressure, particle size and inter-particle spacing to predict particle lifetimes for explosives and other chemicals. Our approach involves depositing particles onto glass substrates using particle sizes and loadings known to be deposited by fingerprint deposition, and then studying their behavior in a custom flow cell with controlled airflow, humidity and temperature. Optical microscope images of the sample taken at fixed time intervals are analyzed to monitor particle sublimation, and those images used to determine the mass loss as a function of time. The data are then fit to a model and from the fitting constants the sublimation rate is calculated. We find that the measured sublimation rate exhibits the expected dependence on vapor pressure for a given material, and that the dependence on vapor pressure is largely material independent. We focus on the behavior of a model material, 2,4-dinitrotoluene and select explosive materials under controlled conditions. We are able to use the data from 2,4-dinitrotoluene to predict the behavior of 2,4,6-trinitrotoluene using the physical properties (e.g., vapor pressure) of the respective materials and compare it to experimental results.

Raman hyperspectral imaging (HSI) is a valuable technique for the detection of threat materials (i.e. explosives and/or narcotics), especially if those materials are located in a complex area with varied background constituents. Raman spectroscopy can provide a unique molecular fingerprint of a threat material, which allows it 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 limitations include low optical throughput, larger size/weight requirements, and area of interrogation size limited to the size of a focused laser spot. These limits are typically due to a system’s spectrometer, commonly a dispersive grating based approach that requires a narrow entrance slit width and long focal length optics to accurately resolve and pass the collected scattered light onto the detector. In addition, using focused laser excitation creates eye-safety concerns that can restrict the usage of Raman sensors for most real-world applications. To address these issues, ChemImage Corporation 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 slit-less grating-based Michelson interferometer with no moving parts. The entrance aperture to the SHS can be orders of magnitude larger than a traditional spectrometer’s entrance slit, which provides an etendue gain of equal magnitude. This feature also allows the laser to be utilized in a defocused configuration, providing an area of coverage up to centimeters in diameter. The sensor also comprises a fiber-array spectral translator (FAST) bundle, a 2-D hyperspectral imaging fiber composed of dozens of smaller fibers, which gives the sensor the ability to spatially discriminate the area of interrogation. The combination of these two technologies is termed FAST-SHS. This paper will provide the background of spatial heterodyne spectroscopy and Raman hyperspectral imaging, the setup and design of a breadboard FAST-SHS, and provide initial results.

Trace particle quantities of explosives left behind by those handling explosives materials present an opportunity to identify both the handlers, secondary handlers and the objects they have contacted. Understanding the nature of these particles is critical for tailoring optical detection strategies as well as non-optical contact harvesting methods. We are working towards developing a model to understand and quantify the nature of particles transferred from the hands to different substrate surfaces. In this preliminary paper we report on a newly developed finger test-bed to produce a robotically controlled series of fingerprints, with an artificial finger designed to mimic the physicochemical properties of the human finger. In an initial set of experiments, we examine the effect of a range of applied forces, the effect of a range of initial particle sizes, and the serial print number on the deposited mass and deposited particle sizes, for a surrogate explosive loaded as particles on gloved fingers which are subsequently pressed against a set of clean glass slides.

Hyperspectral imaging (HSI) has become increasingly popular for sensing in defense, commercial, and academic research for its ability to acquire vast amounts of information, relatively quickly, at stand-off distances. As such, the need for rapid or near-real time data reduction is becoming more evident especially when immediate knowledge of the area under investigation is required such as in contested areas, the scene of natural disasters, and other similar scenarios. While analysis of the underlying spectral information may provide specific information about materials present, in HSI determining an anomaly can be just as informative in scenarios such as CB detection for avoidance. Therefore, a rapid, real-time HSI anomaly detection algorithm is merited. In this paper, we present work towards an algorithm for near-real time anomaly detection utilizing higher-order statistics and, in particular, implications due to changes in skewness and kurtosis, the 3rd and 4th central moments. We demonstrate using a visible-SWIR hyperspectral line scanner that anomalies (thiodiglycol and acetaminophen) can be detected in data that is updated to simulate real-time analysis. Changing spectral features result in changes in the probability density function, and can be specifically realized with comparisons of higher order statistics (i.e. skewness and kurtosis), thereby reducing a full spectral analysis at each voxel to a comparison of two values at each pixel. This paper explores utilizing this concept as a means for anomaly detection, evaluating different surfaces that an analyte may be present on, and lastly presents work towards automated background updates for anomaly detection on dynamic surfaces.

Machine learning based perception algorithms are increasingly being used for the development of autonomous navigation systems of self-driving vehicles. These vehicles are mainly designed to operate on structured roads or lanes and the ML algorithms are primarily used for functionalities such as object tracking, lane detection and semantic understanding. On the other hand, Autonomous/ Unmanned Ground Vehicles (UGV) being developed for military applications need to operate in unstructured, combat environment including diverse off-road terrain, inclement weather conditions, water hazards, GPS denied environment, smoke etc. Therefore, the perception algorithm requirements are different and have to be robust enough to account for several diverse terrain conditions and degradations in visual environment. In this paper, we present military-relevant requirements and challenges for scene perception that are not met by current state-of-the-art algorithms, and discuss potential strategies to address these capability gaps. We also present a survey of ML algorithms and datasets that could be employed to support maneuver of autonomous systems in complex terrains, focusing on techniques for (1) distributed scene perception using heterogeneous platforms, (2) computation in resource constrained environment (3) object detection in degraded visual imagery.

Real-time assessment of suspicious substances on surfaces is an important capability needed by warfighters/first responders. It is also important to perform these assessments without contact or spreading of the suspicious substance or use of reagents. We present work conducted under DTRA and Army funding to develop a hand-held, 1 m to 5 m standoff, optical sensor which detects and classifies trace and bulk concentrations of a wide range of chemical, biological, and explosives (CBE) materials in real-time and full daylight with a fully integrated analyzer weighing less than 12 pounds, including batteries. The sensor method described here combines the complementary chemical information of molecular bonds using Raman and the electronic configuration information using fluorescence, with excitation below 250 nm. There are six main advantages of excitation below 250 nm compared to near-UV, visible or near-IR counterparts: 1) Solar blind detection enabling standoff operation in full daylight; 2) Fluorescence-free Raman and Raman-free fluorescence enabling enhanced detection and identification of target materials without mutual interference; 3) Resonance Raman signal enhancement for improved Raman sensitivity; 4) Simplification of Raman spectra due to resonance enhancement, 5) Short penetration depth, providing physical separation of surface contaminant materials from substrate; and 6) Eye retina safe. These detection capabilities are not possible with near UV, visible, or near IR sensors. A special feature of our sensor is the ability to detect trace biological materials at standoff distances in real time. Photon Systems and JPL have developed these methods over many years, enabling instruments deployed to extreme environments on Earth and an upcoming lander mission to Mars in 2020.

Reflectance (emittance) spectroscopy, especially at infrared wavelengths, continues to grow in utility as an analytical technique for contact, standoff, and remote sensing. The reflectance spectra of solids, however, are complex, depending on many parameters, even for the same material. Granule or powder particle size, crystal morphology, layer thickness, and substrate material all affect the spectral distribution of reflected light. However, such phenomena can all be modeled if the optical constants n(ν) and k(ν) are available. If the quantitative absorption coefficient K(ν) is known, the k value can be obtained via the relation k(ν) = 2.303K(ν)/4πν. The absorption coefficient can in turn be derived from a simple KBrpellet infrared absorption measurement, provided the pellet mass ratio is prepared quantitatively. The method requires the pellet’s mass and diameter, along with the analyte mass fraction and density. In this paper we demonstrate the requisite experimental details in preparing the pellets, as well as methods to reduce light scattering in order to obtain more quantitative values. Theoretical methods to derive the related optical constants will also be detailed, in particular the assumptions used to obtain the scalar refractive index n. Ideally, this value is known or measured separately, but in some cases we have found that it can be approximated (first approximation) for most organic chemicals by n~1.5 at the shortest wavelength. The results are presented for a couple of species.

Other than open bodies of water, bulk liquids are rarely encountered in the environment. Rather, liquids are typically found as aerosols, liquid droplets, or liquid layers on sundry substrates: glass, concrete, metals, etc. The layers can be of varying thicknesses, from micron-level to millimeter thick deposits. The infrared (IR) reflectance spectra of such deposits vary greatly, approximating the bulk reflectance for thicker deposits and for thinner layers on reflective surfaces, producing “transflectance” spectra that more closely replicate simple transmission of the IR light twice traversing the absorptive medium. Rather than recording large numbers of such spectra to serve as endmembers of a spectral reflectance library, we have recognized that the spectra can be modeled so long as the complex optical constants n(ν) and k(ν) are known as a function of frequency, ν. Here n is the real (dispersion) part and k is the imaginary (absorption) component of the complex index of refraction. However, in many cases the bands in the longwave IR (7 to 13 μm) can become saturated, and better signal-to-noise and specificity can be realized at shorter wavelengths. In earlier studies, we obtained the n/k values from 1.28 to 25 μm for a series of liquids, but are now expanding those measurements to include additional liquid species and extending the spectral range to lower wavelengths. In this paper we describe the methodologies for compiling and fusing the two data sets collected to provide better and more complete spectral coverage from 1 to 25 μm (10,000 to 400 cm^-1 ). The broad spectral range means that one needs to account for both strong and weak spectral features, all of which can be useful for detection, depending on the scenario. To account for the large dynamic range, both long and short path length transmission cells are required for accurate measurements.

Optical remote sensing has become a valuable tool in many application spaces because it can be unobtrusive, search large areas efficiently, and is increasingly accessible through commercially available products and systems. In the application space of chemical, biological, radiological, nuclear, and explosives (CBRNE) sensing, optical remote sensing can be an especially valuable tool because it enables data to be collected from a safe standoff distance. Data products and results from remote sensing collections can be combined with results from other methods to offer an integrated understanding of the nature of activities in an area of interest and may be used to inform in-situ verification techniques. This work will overview several independent research efforts focused on developing and leveraging spectral and polarimetric sensing techniques for CBRNE applications, including system development efforts, field deployment campaigns, and data exploitation and analysis results. While this body of work has primarily focused on the application spaces of chemical and underground nuclear explosion detection and characterization, the developed tools and techniques may have applicability to the broader CBRNE domain.

Synthesis and crystal growth of scintillators and semiconductor materials for radiation detectors have been proven to be time consuming and very costly. Several alternative crystals such as Tl3ASSe3, TlGaSe2, Tl4HgI6, PbSe(1-x)Ix have developed in our laboratory. These heavy metal and high Z based compounds have shown great promise. We have been working on some innovative approaches based on Cerenkov radiation and nanocomposites of ionizing organics for faster and efficient sensors. By combining some metallic oxides with an organic material, it should be possible to both extend the energy range of particles capable of being detected while also providing more discrimination for high energy gamma-rays, based on local secondary effects in the surrounding organic matrix. We have been working with a highly ionizing organic compound p-chloranil (2,3,5,6-Tetrachloro-1,4-benzoquinone) matrix. In addition, we have determined effect of oxidizing compounds MnO2 on urea-based composites. We use metal oxide as active ingredient in this matrix. We will present effect of morphology and processing on the performance of nanocomposite for sensing gamma-rays.

The detection, location, and identification of suspected underground nuclear explosions (UNEs) are global security priorities that rely on integrated analysis of multiple data modalities for uncertainty reduction in event analysis. Vegetation disturbances may provide complementary signatures that can confirm or build on the observables produced by prompt sensing techniques such as seismic or radionuclide monitoring networks. For instance, the emergence of non-native species in an area may be indicative of anthropogenic activity or changes in vegetation health may reflect changes in the site conditions resulting from an underground explosion. Previously, we collected high spatial resolution (10 cm) hyperspectral data from an unmanned aerial system at a legacy underground nuclear explosion test site and its surrounds. These data consist of visible and near-infrared wavebands over 4.3 km² of high desert terrain along with high spatial resolution (2.5 cm) RGB context imagery. In this work, we employ various spectral detection and classification algorithms to identify and map vegetation species in an area of interest containing the legacy test site. We employed a frequentist framework for fusing multiple spectral detections across various reference spectra captured at different times and sampled from multiple locations. The spatial distribution of vegetation species is compared to the location of the underground nuclear explosion. We find a difference in species abundance within a 130 m radius of the center of the test site.

Fixed-angle reflectance spectroscopy using a commercial Fourier transform infrared (FTIR) spectrometer was employed to derive the optical constants n and k of several uranium compounds. This technique relies upon measurement of the quantitative reflectance R(ν) spectra from a polished surface across a broad spectral range (in this case, the mid- and far-IR covering ca. 7500 to 50 cm^-1 ) followed by application of the Kramers-Kronig transformation (KKT). Near-normal fixed-angle measurements as used in this technique require continuous reflectance spectra to as low a wavenumber value as possible. Here, we discuss some of the many challenges in measuring the far-IR and very far-IR (terahertz) spectra using an interferometric instrument, particularly those stemming from small sample sizes, typically just millimeters on a face for crystalline samples, as well as limitations due to optical components and diffraction. We apply this method to single-crystal UO₂ and its mineralogical form uraninite, as well as other Ubearing minerals such as autunite [Ca(UO₂)₂(PO₄)₂·8-12H₂O] and the dehydrated form of autunite, meta-autunite. In addition to the specular reflectance spectra, x-ray diffractometry was used as a confirmatory technique to analyze the surface composition of the species. Deriving the infrared optical constants for such U-bearing species (as well as other solids) will enable nondestructive detection under a variety of environmental and compositional conditions.

Chemical detection is a priority for the Intelligence Community (IC) with applications such as forensic analysis, border/facility protection, and stockpile/production monitoring. In particular, the IC has an interest in long term monitoring of a chemical environment without human oversight. The technology necessary for monitoring of this type must provide high sensitivity and accuracy, be robust to false alarms in the presence of complex chemical mixtures, and be contained in a small, ruggedized package with autonomous operation. The Intelligence Advanced Research Project Activity’s (IARPA’s) Molecular Analyzer for Efficient Gas-phase Low-power INterrogation (MAEGLIN) program is developing an ultra-low-power chemical analysis capability for chemical detection and identification of explosives, chemical weapons, industrial toxins and pollutants, narcotics, and nuclear materials with a low false alarm rate in the presence of complex interferents. In Phase 1 the MAEGLIN program separately developed component technology for chemical collection, separation, and identification tasks. This paper will describe the MAEGLIN program’s results from the nine separate Phase 1 performers. Highlights include development of an array of micro ion trap mass spectrometers with dual photoelectric and electron impact ionization, a sector mass spectrometer with a charge coupled device (CCD) detector that uses permanent magnets for ion bending, a miniature tandem ion mobility spectrometer system with a wire grid fragmenter, adaptive multi-channel three dimensional gas chromatography, various active and passive microfabricated preconcentrators, including one with a dedicated “hold and fire” stage for <0.25 second injection, and miniature electrostatic and Knudsen pumps.

Adaptive Infrared Imaging Spectroradiometer (AIRIS) is a longwave infrared (LWIR) sensor for remote detection of chemical agents such as nerve gas. AIRIS can be considered as a hyperspectral imager with 20 bands. In this paper, we present a systematic and practical approach to detecting and classifying chemical vapor from a distance. Our approach involves the construction of a spectral signature library of different vapors, certain practical preprocessing procedures, and the use of effective detection and classification algorithms. In particular, our preprocessing involves effective vapor signature extraction with adaptive background subtraction and normalization, and vapor detection and classification using Spectral Angle Mapper (SAM) technique, which is a signature-based target detection method for vapor detection. We have conducted extensive vapor detection analyses on AIRIS data that include TEP and DMMP vapors with different concentrations collected at different distances and times of the day. We have observed promising detection results both in low and high-concentrated vapor releases.

The initial objective of this work was to demonstrate detection of explosives and precursor materials using existing, fielded CWA (Chemical Warfare Agent) point detection systems, specifically the Joint Chemical Agent Detector (JCAD), and to integrate explosives detection into the family of CBRN sensors. A key premise of the objective is to demonstrate this utility without modifying the detector itself. Recent RDECOM efforts expanded to demonstrate the ability of the JCAD to detect and analyze explosives, low volatility compounds (LVC), narcotics, and pharmaceutical based compounds (PBC) in this manner. JCAD is a currently fielded, man-portable, ion mobility spectrometry-based detector for CWA that operates at ambient atmospheric pressure conditions. Tens of thousands JCAD systems have been produced and many remain in warehouse storage. Smiths Detection and RDECOM have produced a new chemical detector system using an unmodified JCAD. A JCAD from the warehouse can be plugged into a new solid-liquid adapter (SLA) peripheral hardware, or “cradle,” and analysis software residing in the cradle turns the JCAD into a detector for explosives, LVC, narcotics, and PBC. The system has been demonstrated for trace detection of these compounds. To maintain high versatility and low interferences, its spectrometric resolving power is improved through rigorous instrument calibration and real-time signal processing techniques. Separate unique instrumentation to measure ion mobility values an order of magnitude more accurately (± 0.2%) than literature values is in operation at ECBC and provides the reference values for JCAD calibration. Measured ion mobility values of target analytes are adjusted for water vapor effects and background contaminants during calibration.

We have shown recently that unique optical signatures can be observed with the measurement of ultrashort middle infrared laser pulses that have been transmitted through molecular vapors. Here, we report on an increased signal-to-noise ratio of the pulse measurements by using a cross-correlation technique with a lockin amplifier. Carbon tetrafluoride and dimethyl methylphosphonate (DMMP) cross-correlation signatures are highly discriminated using principal component analysis. A squared exponential Gaussian process regression model is used to quantitatively predict the concentration of DMMP.

A Bakman Technologies PB7220-2000-T portable, frequency-domain, THz spectrometer connected to a custom fabricated, light-weight 10 meter White cell was employed to measure the 1.492 THz, 1.507 THz and 1.523 THz molecular transitions in DHO. Sensor sensitivity levels are then compared to what is required to detect naturally occurring DHO in atmospheric water vapor.

The integration of thermal infrared (TIR) hyperspectral systems into Unmanned Aerial Vehicles (UAVs) platforms is expected to open doors toward a wide variety of demanding thermal imaging applications ranging from academics and research to industry. Currently, the UAV remote sensing technology in TIR region is still in its infancy and the main expectations are the reduction of both, sensor sizes and cost while maintaining their performances at a high level. In this communication, we report on Telops newly designed compact, light and robust TIR hyperspectral module of less than 10 kg with about 50W of power consumption. The new module can be integrated into a complete stand-alone imager with applications such as 360˚ Hyperspectral Surveillance. Integration in complete, highly flexible UAV based, infrared hyperspectral imaging solutions, such as airborne real-time gas detection, identification and quantification is also possible. The need for a reliable and cost-efficient gas detection system is of prime importance especially when security threatening situations like gas leaks and emissions occur. The knowledge of the precise localization of the leaks, identification of the chemical nature of the gases involved and quantification of the gas flux emanating from the leaks are the crucial inputs needed for the incident response team to take actions based on relevant information. In this regard, UAVs based TIR remote sensing technology offers many benefits over traditional gas detection systems as it allows safely monitoring and imaging of large areas. The sensor can fly several hundreds of meters above the scene, avoiding the need to access restricted and potentially dangerous zones in the installations. Beside the newly designed compact and light TIR hyperspectral module, Telops have also developed solutions for gas detection and identification along with some tools for the quantification of gas flow rates emanating for leak source. These solutions were recently demonstrated during a flight campaign up to 4600 feet above the ground for detection and identification of ethylene, methanol and acetone gas release experiment. The Fourier transform technology used in our hyperspectral imaging systems on an airborne platform allows recording of airborne hyperspectral data using mapping and targeting modes. These two acquisition modes were used for gas detection and real time quantitative airborne chemical images of the three gas clouds were obtained paving the path toward a viable solution for gas leak surveys and environmental monitoring.

Gas Chromatography (GC) is routinely used in the laboratory to temporally separate chemical mixtures into their constituent components for improved chemical identification. This paper will provide a overview of more than twenty years of development of one-dimensional field-portable micro GC systems, highlighting key experimental results that illustrate how a reduction in false alarm rate (FAR) is achieved in real-world environments. Significantly, we will also present recent results on a micro two-dimensional GC (micro GCxGC) technology. This ultra-small system consists of microfabricated columns, NanoElectroMechanical System (NEMS) cantilever resonators for detection, and a valve-based stop-flow modulator. The separation of a 29-component polar mixture in less than 7 seconds is demonstrated along with peak widths in the second dimension ranging from 10-60 ms. For this system, a peak capacity of just over 300 was calculated for separation in about 6 s. This work has important implications for field detection, to drastically reduce FAR and significantly improve chemical selectivity and identification. This separation performance was demonstrated with the NEMS resonator and bench scale FID. But other detectors, suitably fast and sensitive can work as well. Recent research has shown that the identification power of GCxGC-FID can match that of GC-MS. This result indicates a path to improved size, weight, power, and performance in micro GCxGC systems outfitted with relatively non-specific, lightweight detectors. We will briefly discuss the performance of possible options, such as the pulsed discharge helium ionization detector (PDHID) and miniature correlation ion mobility spectrometer (mini-CIMS).

A passive LWIR camera based on a focal plane array detector was used to capture hyperspectral images of different scenes where low volatile chemical warfare agent (CWA) and simulant droplets were deposited on a variety of substrates. Four different CWA; mustard gas (HD), cyclosarin (GF), VX, VR, and two simulant liquids; dimethyl methylphosphonate (DMMP) and triethyl phosphate (TEP) were used as surface contaminants and applied to the substrates as 5 μl droplets. The trials were performed outdoors with the scene close to the ground and the camera imaging at an angle of approximately 35° and a stand-off distance of about 2 m, i.e., mainly the reflected radiation from the cold sky in combination with the thermal emission from the scene was observed. Brightness temperature spectra were extracted from the hyperspectral data cubes and compared to results from a thin film model as well as reference LWIR spectra of the liquids.

We present a spectroscopic technique based upon optical phase-fluctuation spectroscopy for very high levels of sensitivity and specificity with application for detecting the presence of concealed explosives by detection in the vapor phase. The approach enables recent advances in deep-infrared QCL spectroscopic sources to be utilised without the need for cooled detectors and gives multi-pass Herriott-type cell performance from a highly compact form factor. The system has been evaluated in the mid-infrared using a continuous-wave optical parametric oscillator as a spectroscopic excitation source, and Ethane as a sample molecule for detection. With this setup we have demonstrated the specificity of the device by being able to resolve characteristic spectral lines of the molecule of interest against other contaminants in the sample with similar spectral response, and a noise-equivalent sensitivity of 15ppb. Sensitivity is currently limited by ambient mechanical noise and routes to minimize this are considered.

We report on a fiber-format laser system for Fourier-transform coherent anti-Stokes Raman spectroscopy (FT-CARS) of toxic chemical hazards, such as chemical warfare agents (CWAs). The system is based on ytterbium-fiber technology featuring ultra-broad spectral coverage and high-sensitivity. High-quality CARS spectra with maximum Raman shifts of 3000 cm^-1 and signal-to-noise ratio >200 for observation times of 160 μs are measured; a detection limit of one part per thousand is demonstrated with a cyanide/water solution. The system was developed as the basis for a highly accurate, sensitive, reliable and portable device for the real time detection of water contaminants and deposited CWAs at trace levels.

Recent developments have shown that an optical filter-based sensing approach, inspired by human color vision, is capable of high-confidence discrimination between chemicals with similar infrared vibrational absorption bands. A key design point for this technique lies in the selection of the optical filters, which provide good discrimination between chemicals. Filter selection is also intrinsically tied to the classification method employed for the discrimination itself. Thus, it is imperative that the classification method or methods to be used are well understood and that mathematical means exist to compare the discrimination results provided by independent sets of optical filters. To meet this challenge, we are examining means to assign cost values to each set of optical filters for a given associated classification method. In this effort, the cost value used is the volume formed by three unique discrimination vectors. This method is developed from machine learning approaches, which define cost functions for stochastic optimization routines. We discuss multiple computational methods to discriminate between chemicals with similar infrared vibrational absorption bands using unique infrared (IR) tristimulus values for each chemical. These IR-tri-stimulus values are determined by the interaction with the chemical absorption bands and three individual optical IR band-pass filters. Methods to determine the associated cost for various selections of these IR band-pass filters and associated mathematical operations are described and compared for the computational methods. We discuss the methods employed to select the IR optical filters and discuss how the flexibility of this approach demonstrates the power of this biomimetic sensing method.

Detection of explosives on surfaces could potentially be achieved using handheld standoff optical sensors, providing rapid intelligence and safety to the warfighter or first responder. Recent work has shown the capability to discriminate various chemical vapors using a bioinspired or biomimetic detection system modeled on human color vision. This biomimetic system utilizes three overlapping broadband infrared optical filters to discriminate between various chemicals. Preliminary reflectance data of chemicals on surfaces indicate a capability for discrimination of target chemicals and interferents by analysis of biomimetic sensor output using novel analytical methods such as “Comparative Discrimination Spectral Detection” (CDSD). Transitioning this detection method to threats on surfaces at proximate standoff distances (~1 m) requires additional considerations including surface characteristics and angle of detection. This work explores sensor detection parameters for ammonium nitrate residue on aluminum surfaces of various roughnesses. Samples of the explosives component ammonium nitrate, NH₄NO₃, diluted at 10%, 5%, and 1% in DI water, were prepared by dropcasting onto aluminum coupons with four surface finishes: polished, brushed, extruded, and sandblasted. Surface roughnesses were measured. Single beam reflectance spectra (2 – 20 μm) were collected using a FTIR spectrometer over multiple independent angles of incidence and collection (15° - 80°). Multiple factors were analyzed including albedo, and potential sensor configurations. Characteristics for future MWIR and LWIR sensors, such as illumination power and detector sensitivity, are evaluated which enable chemical spectral identification across range of aluminum surface roughness for proximate standoff distances and different angles of incidence.

Whispering gallery mode devices have emerged as a powerful class of optical devices in which light-matter interactions are significantly enhanced within micron-scale structures, making them an ideal platform for both fundamental science and applications such as advanced sensors, low-threshold lasers and nonlinear optics, just to name a few. Typically, fibers tapers or prisms are used as optical couplers for the resonators. We demonstrate angle-polished fibers as an alternative option to efficiently couple light to a high-quality whispering-gallery resonator. Angle-polished fibers offer the advantage of rapid fabrication, increased mechanical stability versus fiber tapers, and the ability to tune the excitation angle of light. We demonstrate the use of angle-polished fibers for coupling light out of a whispering-gallery device as well as a rapid, low-cost method for fabricating the couplers.

Optical Whispering Gallery Mode (WGM) microresonators have become a powerful tool in fundamental physics as well as significant applications. Optical sensors based on WGM resonators have shown ultra-high sensitivity levels for various analytes. Among various kinds of WGM resonators, microbubble resonators (MBRs) are especially appealing as sensors, since the optical and fluidic elements are combined into a single component. We have developed a simple, rapid, and reliable packaging technique for silica microbubble resonators using 3-D printed packaged chips. The packaged MBR offer Q-factors at high as 106 with stability to environmental fluxes such as temperature. As an initial application, we demonstrate internal pressure sensing with our packaged MBR devices. The integration of both optical and fluidic components shows potential in a wide variety of field-based and point-of-care applications.

We are currently developing a novel fabric spectrometer-based colorimetric chemical sensor that is lightweight, sensitive, and person-borne. This research will enable a new class of chemical sensors with a much flexible form factor to open up a variety of other person-borne and distributed sensing applications.

Several recently proposed methods for detecting radioactivity at range involve driving laser induced avalanche breakdown seeded by electrons or negative ions whose density are elevated in the vicinity of a radioactive source. Using a chirped, mid-IR laser, we drive breakdowns at 1 meter standoff distances and monitor the breakdown timing using the backscattered spectrum. In addition to the on/off radiation detection based on the increased probability of finding a seed electron in the focal volume, we also can determine the spatial distribution of these seed electrons in the focal volume through temporal information encoded in this backscatter spectrum. We demonstrate that the backscatter spectrum is a superior detection method relative to visible plasma fluorescence, total pump backscatter, or absolute backscatter timing in its ability to determine the relative radiation level. We discuss scaling to longer focal geometries inherent in remote sensing and possible limitations to the technique, supported by modeling

A concept for all-optical remote detection of radioactive materials is presented and analyzed. The presence of excess radioactivity increases the level of negative ions in the surrounding air region. We model irradiated air to estimate the density of negative ions and use a set of coupled rate equations to simulate a subsequent laser-induced avalanche ionization. This can act as a source of seed electrons for a laser-induced avalanche ionization breakdown process. We examine avalanche ionization behavior in several laser parameter regimes, and determine the time required for saturation of the breakdown for both a single seed ion as well as for a population of ions present in the focused volume of a highintensity laser pulse, corresponding to two methods of remotely measuring the ion density, which is a signature of radioactive materials.