The mammalian olfactory system is able to detect many more odorants than the number of receptors it has by utilizing cross-reactive odorant receptors that generate unique response patterns for each odorant. Mimicking the mammalian system, artificial noses combine cross-reactive sensor arrays with pattern recognition algorithms to create robust odor-discrimination systems. The first artificial nose reported in 1982 utilized a tin-oxide sensor array. Since then, however, a wide range of sensor technologies have been developed and commercialized. This review highlights the most commonly employed sensor types in artificial noses: electrical, gravimetric, and optical sensors. The applications of nose systems are also reviewed, covering areas such as food and beverage quality control, chemical warfare agent detection, and medical diagnostics. A brief discussion of future trends for the technology is also provided.
A rapid method for vapor pressure measurement was developed and used to derive the vapor pressure curve of the thermally labile peroxide-based explosive hexamethylene triperoxide diamine (HMTD) over the temperature range from 28 to 80 °C. This method uses a controlled flow of vapor from a solid-phase HMTD source that is presented to an ambient-pressure-ionization mass spectrometer equipped with a secondary-electrospray-ionization (SESI) source. The subpart-per-trillion sensitivity of this system enables direct detection of HMTD vapor through an intact [M + H](+) ion in real time at temperatures near 20 °C. By calibrating this method using vapor sources of cocaine and heroin, which have known pressure-temperature (P-T) curves, the temperature dependence of HMTD vapor was determined, and a Clausius-Clapeyron plot of ln[P (Pa)] vs 1/[T (K)] yielded a straight line with the expression ln[P (Pa)] = {(-11091 ± 356) × 1/[T (K)]} + 25 ± 1 (error limits are the standard error of the regression analysis). From this equation, the sublimation enthalpy of HMTD was estimated to be 92 ± 3 kJ/mol, which compares well with the theoretical estimate of 95 kJ/mol, and the vapor pressure at 20 °C was estimated to be ∼60 parts per trillion by volume, which is within a factor of 2 of previous theoretical estimates. Thus, this method provides not only the first direct experimental determination of HMTD vapor pressure but also a rapid, near-real-time capability to quantitatively measure low-vapor-pressure compounds, which will be useful for aiding in the development of training aids for bomb-sniffing canines.
Abstract. This paper presents Active Class Selection (ACS), a new class of problems for multi-class supervised learning. If one can control the classes from which training data is generated, utilizing feedback during learning to guide the generation of new training data will yield better performance than learning from any a priori fixed class distribution. ACS is the process of iteratively selecting class proportions for data generation. In this paper we present several methods for ACS. In an empirical evaluation, we show that for a fixed number of training instances, methods based on increasing class stability outperform methods that seek to maximize class accuracy or that use random sampling. Finally we present results of a deployed system for our motivating application: training an artificial nose to discriminate vapors.
The chemical and physical fates of trace amounts (<50 μg) of explosives containing 2,4,6-trinitrotoluene (TNT), hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX), and pentaerythritol tetranitrate (PETN) were determined for the purpose of informing the capabilities of tactical trace explosive detection systems. From these measurements, it was found that the mass decreases and the chemical composition changes on a time scale of hours, with the loss mechanism due to a combination of sublimation and photodegradation. The rates for these processes were dependent on the explosive composition, as well as on both the ambient temperature and the size distribution of the explosive particulates. From these results, a persistence model was developed and applied to model the time dependence of both the mass and areal coverage of the fingerprints, resulting in a predictive capability for determining fingerprint fate. Chemical analysis confirmed that sublimation rates for TNT were depressed by UV (330-400 nm) exposure due to photochemically driven increases in the molecular weight, whereas the opposite was observed for RDX. No changes were observed for PETN upon exposure to UV radiation, and this was attributed to its low UV absorbance.
The design and implementation of a portable fluorescence-based vapor sensing system are described. The system incorporates previously developed microsensor array technology into a compact, low-power device capable of collecting and delivering ambient vapor samples to the array while monitoring and recording the fluorescent responses of the sensors. The sensors respond differentially when exposed to a sample vapor and, when processed using a support vector machine (SVM) pattern recognition algorithm, are shown to discriminate between three classes of petroleum distillates. The system was characterized using sample vapors prepared under several different conditions in three sensing scenarios. The first scenario demonstrates the basic operational capability of the device in the field by presenting high concentration vapors to the array. The second scenario introduces the potential for a greater degree of variability in both sample vapor concentration and composition in an effort to emulate real-world sensing conditions. The third scenario uses an on-board trained pattern recognition algorithm to identify unknown vapors as their responses are collected. The device demonstrated high classification accuracy throughout the field tests and is capable of improving its classification accuracy when challenged with samples presented under variable ambient conditions by enhancing the signal-to-noise ratio of the array response.
High-sensitivity (ng/cm²) optical detection of the explosive 2,4,6-trinitrotoluene (TNT) is demonstrated using photodissociation followed by laser-induced fluorescence (PD-LIF). Detection occurs rapidly, within 6 laser pulses (~7 ns each) at a range of 15 cm. Dropcasting is used to create calibrated samples covering a wide range of TNT concentrations; and a correspondence between fractional area covered by TNT and PD-LIF signal strength is observed. Dropcast data are compared to that of an actual fingerprint. These results demonstrate that PD-LIF could be a viable means of rapidly and remotely scanning surfaces for trace explosive residues.
Nanostencil lithography has a number of distinct benefits that make it an attractive nanofabrication processes, but the inability to fabricate features with nanometer precision has significantly limited its utility. In this paper, we describe a nanostencil lithography process that provides sub-15 nm resolution even for 40-nm thick structures by using a sacrificial layer to control the proximity between the stencil and substrate, thereby enhancing the correspondence between nanostencil patterns and fabricated nanostructures. We anticipate that controlled proximity nanostencil lithography will provide an environmentally stable, clean, and positive-tone candidate for fabrication of nanostructures with high-resolution.Significant advances in device physics across optical, electrical, and magnetic modalities have been enabled by the ability to fabricate structures with nanoscale precision. There is currently a large variety of top-down nanostructure fabrication methods available, such as electron beam lithography 1 ,
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.