R. D. S. Yadava

Multifrequency characterization of viscoelastic polymers and vapor sensing based on SAW oscillators.

Simplified relations for the changes in SAW velocity and attenuation due to thin polymer coatings and vapor sorption are presented by making analytic approximations to the complex theoretical model developed earlier by Martin et al. [Anal. Chem. 66 (14) (1994) 2201-2219]. The approximate velocity relation is accurate within 4% for the film thicknesses up to 20% of the acoustic wavelength in the polymer film, and is useful for analyzing the mass loading, swelling and viscoelastic effects in SAW vapor sensors. The approximate attenuation relation is accurate within 20% for very thin films, (less than 2% of the acoustic wavelength in the film). Based on these relations, a new procedure for determination of polymer viscoelastic properties is described that exploits the frequency dependence of the velocity and attenuation perturbations, and employs multifrequency measurement on the same SAW platform. Expressions for individual contributions from the mass loading, film swelling and viscoelastic effects in SAW vapor sensors are derived, and their implications for the sensor design and operation are discussed. Also, a new SAW comb filter design is proposed that offers possibility for multimode SAW oscillator operation over a decade of frequency variation, and illustrates feasibility for experimental realization of wide bandwidth multifrequency SAW platforms.

Mass sensitivity analysis and designing of surface acoustic wave resonators for chemical sensors

The sensitivity of surface acoustic wave (SAW) chemical sensors depends on several factors such as the frequency and phase point of SAW device operation, sensitivity of the SAW velocity to surface mass loading, sensitivity of the SAW oscillator resonance to the loop phase shift, film thickness and oscillator electronics. This paper analyzes the influence of the phase point of operation in SAW oscillator sensors based on two-port resonator devices. It is found that the mass sensitivity will be enhanced if the SAW device has a nonlinear dependence on the frequency (delay ~ frequency−1). This requires the device to generate and operate in a ωτg(ω) = const region in the device passband, where ω denotes the angular frequency of oscillation and τg(ω) denotes the phase slope of the SAW resonator device. A SAW coupled resonator filter (CRF) that take advantage of mode coupling is considered in realizing such a device to help in shaping the phase transfer characteristics of a high mass sensitivity sensor. The device design and simulation results are presented within the coupling-of-modes formalism.

Effect of film thickness and viscoelasticity on separability of vapour classes by wavelet and principal component analyses of polymer-coated surface acoustic wave sensor transients

The transient response of a polymer-coated surface acoustic wave (SAW) vapour sensor depends on partitioning and diffusion of vapour species into the polymer in conjunction with its thickness and viscoelastic properties. The shapes of transient signals carry information about vapour identities due to specificity of the partition coefficient and the diffusion coefficient. The analysis of transient signals therefore offers a simpler approach for vapour identification in comparison to conventional electronic nose systems that employ a broadly selective sensor array. The transient response-based methods are however not developed to a similar level of maturity as their sensor array counterparts. The main reason for this is associated with complex signal generation kinetics and polymer viscoelasticity. The latter is independent of vapour identities (assuming low concentrations) but influences sensor response through nonlinear dependences on polymer thickness and viscoelastic coefficients. In this paper, we endeavour to find out whether viscoelasticity and its manifestation through thickness dependences could be turned into an advantage for transient-based vapour identification. Using an established SAW sensor model with additive noise we analyse sensor transients by wavelet decomposition and principal component analysis (PCA) for various combinations of polymer thickness, viscoelastic storage and loss moduli and noise level. We calculate vapour class separability measures defined on the basis of scatter matrices of principal components of wavelet coefficients to determine the discrimination ability of sensor transients for various combinations of film thickness and viscoelastic parameters. The simulation experiments are performed by considering a polyisobutylene-coated SAW oscillator sensor under exposure to seven volatile organic compounds (chloroform, chlorobenzene, o-dichlorobenzene, n-heptane, toluene, n-hexane and n-octane). The film thicknesses are varied from thin film (where mass loading dominates) to thick film through fundamental film resonance (where viscoelastic effects dominate) to very thick film regions spanning over a few higher order resonances. The storage and loss shear moduli are varied to simulate conditions of glassy, glassy-rubbery and rubbery phases of polymer. The transient response generation incorporates an additive noise source with uniform distribution over a specified range. Effect of noise variation on class separability is also studied. A comparison of the wavelet transform method is made with the phase space-based partial-least-squares regression method for feature extraction. In conclusion, it is found that (i) vapour class separability increases with polymer thickness for all viscoelastic conditions from glassy to rubbery, except near film resonances, (ii) near resonance class separability dramatically declines for films with no or low loss, (iii) by imparting finite viscoelastic losses to polymer coatings, not only are the detrimental effects of film resonance eliminated but also the sensor performance improves, and (iv) viscoelastic effects produce better noise immunity in thick film sensors. This analysis provides a new perspective for designing high-performance transient sensors through optimization of film thickness for specific polymer selections.

Fuzzy C-means clustering based uncertainty measure for sample weighting boosts pattern classification efficiency

The paper presents a fuzzy c-means clustering based fuzzy measure for weighting samples in a dataset for pattern classification. The method improves classification efficiency. The fuzzy c-means generated membership grades of a sample for belonging to different clusters are interpreted as measures of uncertainty for assigning specific crisp class label to this sample. The fuzzy measure of total uncertainty for a sample is defined as U = -Σk=1c Mk log2 Mk where Mk denotes the membership grade in k-th cluster, and the summation extends is over all the c clusters. The data samples in feature space are then transformed according to X → (1 + U)X. By using a radial basis function neural network classifier the classification efficiency is compared based on the original and the transformed feature vectors. Several data sets collected from open sources were used for validation.

Enhancing chemical identification efficiency by SAW sensor transients through a data enrichment and information fusion strategy—a simulation study

The paper proposes a new approach for improving the odor recognition efficiency of a surface acoustic wave (SAW) transient sensor system based on a single polymer coating. The vapor identity information is hidden in transient response shapes through dependences on specific vapor solvation and diffusion parameters in the polymer coating. The variations in the vapor exposure and purge durations and the sensor operating frequency have been used to create diversity in transient shapes via termination of the vapor–polymer equilibration process up to different stages. The transient signals were analyzed by the discrete wavelet transform using Daubechies-4 mother wavelet basis. The wavelet approximation coefficients were then processed by principal component analysis for creating feature space. The set of principal components define the vapor identity information. In an attempt to enhance vapor class separability we analyze two types of information fusion methods. In one, the sensor operation frequency is fixed and the sensing and purge durations are varied, and in the second, the sensing and purge durations are fixed and the sensor operating frequency is varied. The fusion is achieved by concatenation of discrete wavelet coefficients corresponding to various transients prior to the principal component analysis. The simulation experiments with polyisobutylene SAW sensor coating for operation frequencies over [55–160] MHz and sensing durations over [5–60] s were analyzed. The target vapors are seven volatile organics: chloroform, chlorobenzene, o-dichlorobenzene, n-heptane, toluene, n-hexane and n-octane whose concentrations were varied over [10–100] ppm. The simulation data were generated using a SAW sensor transient response model that incorporates the viscoelastic effects due to polymer coating and an additive noise source in the output. The analysis reveals that: (i) in single transient analysis the class separability increases with sensing duration for a given frequency of operation, and also with frequency for a given sensing duration, and (ii) the information fusion based on both the multiple sensing cycles and the multiple sensing frequencies enhances the class separability by nearly an order of magnitude.

Application of Fuzzy c-Means Clustering for Polymer Data Mining for Making SAW Electronic Nose

Polymers provide chemical interfaces to electronic nose sensors for detection of volatile organic compounds. An electronic nose combines a properly selected set of sensors (array) with pattern recognition methods for generating information rich odor response patterns and extracting chemical identities. Each sensor in the array is functionalized by a different polymer for diversely selective sorption of target chemical analytes. Selection of an optimal set of polymers from a long list of potential polymers is crucial for cost-effective high performance development of electronic noses. In this paper we present an application of fuzzy c-means clustering on partition coefficient data available for target vapors and most prospective polymers for selection of minimum number of polymers that can create maximum discrimination between target chemical compounds. The selection method has been validated by simulating a polymer coated surface acoustic wave (SAW) sensor array for monitoring fish freshness.

Fuzzy Subtractive Clustering for Polymer Data Mining for SAW Sensor Array Based Electronic Nose

Fuzzy subtractive clustering (FSC) has been applied as data mining tool for making selection of a small set of polymers from a large set of prospective polymers having potential for being chemical interfaces for electronic nose sensor array. The basic idea behind applying FSC selection is to cluster the prospective polymers according to some measure of similarity among them in relation to their interaction with the chemicals targeted for sensing. The polymers defining the cluster centers are taken to make the selection set. The basis for defining similarity among different polymers is the partition coefficients associated with sorption of chemical analytes from vapor phase to polymer phase in thermodynamic equilibrium. The goal for selection is to identify a minimal set of polymers that provide the most diversely interaction possibilities with the target vapors. The proposed selection method has been validated by simulating responses of a polymer-coated surface acoustic wave (SAW) sensor array for detection of freshness and spoilage of milk and fish food products. The end use of the proposed selection method is suggested for developing low-cost high-performance sensor array based electronic noses for commercial and consumer applications.

Transient feature extraction based on phase space fusion by partial-least-square regression analysis of sensor array signals

Pattern classification based on transient signal analysis provides an effective method for identification of dynamical systems. The partial-least-square regression (PLSR) is most commonly used to generate parametric representation of phase space defined by measured signals and their time derivatives. The PLS component scores are interpreted as object features for pattern identification. In this paper, we consider sensor array transients, and propose PLSR based fusion of phase spaces of individual sensors into a single virtual phase space. Motivation for this approach comes from realizing that (i) multiplicity of array sensors encodes information about object diversity, and (ii) PLSR models object diversities in terms of small number of latent variables. The approach is validated through a case study of vapor identification by electronic nose based on surface-acoustic-wave (SAW) chemical sensor array. A comparison of results with and without fusion shows substantial improvement in vapor class separability after phase space fusion.

Evidence generation for Dempster-Shafer fusion using feature extraction multiplicity and radial basis network

Feature extraction methods in pattern recognition tasks seek to transform data variables to abstract mathematical variables such that their scores (called features) reveal hidden data structure of high cognitive value. Various feature extraction methods process raw data from different perspectives. Some depend on statistical correlation or independence such as principal component analysis (PCA), independent component analysis (ICA), singular value decomposition (SVD) and linear discriminant analysis (LDA), and some others aim to model parametric representations such as partial-least-square regression (PLSR). These methods can be viewed as independent observers who generate different feature sets for describing the same data. In supervised pattern recognition problems, this viewpoint can be combined with a classifier function to generate independent sets of class likelihood. The latter can be interpreted as evidences for class identities assigned by independent expert systems consisted of feature extraction method and classifier function combinations. Having created such set of experts, one can employ an information fusion system that could predict class identities. Following this paradigm, we used above mentioned feature extraction methods paired with a radial basis network to generate evidences, and applied Dempster-Shafer (D-S) fusion for pattern classification in a number of benchmark data sets. It is found that DS fusion results in enhanced classification rates compared to results from individual expert systems.

Modeling electrical response of polymer-coated SAW resonators by equivalent circuit representation

The paper presents an equivalent circuit model of the polymer coated surface acoustic wave (SAW) resonators by combining coupling-of-mode (COM) description of SAW resonators and perturbation calculation of SAW propagation under polymer loading. An expression for the motional load produced by polymer coating is deduced in terms of COM parameters and polymer characteristics. In addition, expressions for the shifts in resonance frequency and attenuation due to polymer loading are obtained. Simulation results are presented for one-port and two-port resonator devices coated with viscoelastic thin polymer film. The influence of polymer film on resonator response is studied with regard to variations in film thickness and shear modulus. The model simplifies understanding of polymer-coated SAW sensors.