Maria V. Schiza

Multilayer sol/gel membranes for optical sensing applications: single layer pH and dual layer CO2 and NH3 sensors

Organo-silica sol-gel membranes have been prepared and demonstrated in a single layer format for pH measurement and multiple-layer format for both carbon dioxide and ammonia. The sensors are simple and versatile since the same chemistry and membranes are used for each sensor. The sensors use hydroxypyrenetrisulfonic acid (HPTS) as the indicator immobilized in a base-catalyzed sol-gel containing poly(dimethyl)siloxane, aminopropyltriethoxysilane (APTES) and tetraethylorthosilicate (TEOS). This indicator gel is over coated with a hydrophobic sol-gel to reduce cross reactivity to pH when either CO(2) or NH(3) are examined. The gels are very stable and the sensors retain response up to a 12-month period. Sensors can be stored in buffer or dry without loss of function and have response times to that are comparable to literature values.

Growth and Characterization of a Porous Aluminum Oxide Film Formed on an Electrically Insulating Support

Thin films of porous anodic aluminum oxide have been prepared on an electrically insulating support by the anodization of aluminum films sputtered onto glass slides. The resulting transparent aluminum oxide films were characterized by scanning electron microscopy and variable angle ellipsometry. Subsequently, the film was modeled from the ellipsometric data taken. An underlying conductive medium is not necessarily needed to bring about nearly complete anodization of the aluminum layer.

Application of multivariate optical computing to simple near-infrared point measurements

Quantitative multivariate spectroscopic methods seek spectral patterns that correspond to analyte concentrations even in the presence of interferents.By embedding a spectral pattern that corresponds to a target analyte in an interference filter in a beamsplitter arrangement;bulky and complex instrumentation can be eliminated with the goal of producing a field-portable instrument.A candidate filter design for an rganic analyte,of military interest,and an interferent is evaluated.

On-line reoptimization of filter designs for multivariate optical elements

An automated method for producing multivariate optical element (MOE) interference filters that are robust to errors in the reactive magnetron sputtering process is described. Reactive magnetron sputtering produces films of excellent thickness and uniformity. However, small changes in the thickness of individual layers can have severe adverse effects on the predictive ability of the MOE. Adaptive reoptimization of the filter design during the deposition process can maintain the predictive ability of the final filter by changing the thickness of the undeposited layers to compensate for the errors in deposition. The merit function used, the standard error of calibration, is fundamentally different from the standard spectrum matching. This new merit function allows large changes in the transmission spectrum of the filter to maintain performance.

Application of multivariate optical computing to near-infrared imaging

Rapid quantitative imaging of chemical species is an important tool for investigating heterogenous mixtures, such as laminated plastics, biological samples and vapor plumes. Using traditional spectroscopic methods requires difficult computations on very large data sets. By embedding a spectral pattern that corresponds to a target analyte in an interference filter in a beamsplitter arrangement; the chemical information in an image can be obtained rapidly and with a minimal amount of computation. A candidate filter design that is tolerant to the angles present in an imaging arrangement is evaluated in near-infrared spectral region for an organic analyte and an interferent.

Use of a 2D to 1D Dimension Reduction Fiber-Optic Array for Multiwavelength Imaging Sensors

A dimension reduction fiber-optic array is used to measure the response of a stacked-layer, image-guide CO2/O2 sensor, simultaneously at several different wavelengths. Two different image-guide CO2/O2 sensor configurations are described: a stacked-layer sensor, where luminescence indicators for CO2 and O2 are uniformly coated on the tip of the sensor; and a side-by-side coated sensor where the two indicators are coated on different halves of the fiber tip. It is shown that a single image-guide measurement, made by using the dimension reduction array, can be used to generate response plots, intensity profiles, and reconstructed images at different luminescence wavelengths. The spatial resolution of an image guide sensor is limited by the number of fibers used to construct the dimension reduction array.

Miniature Stereo Spectral Imaging System for Multivariate Optical Computing

Chemical or hyperspectral imaging is a rapidly developing field that has applications ranging from materials characterization to remote environmental sensing.1 Thanks to developments in processing and instrumentation over the past two decades,2–4 it is now possible to use hyperspectral imaging routinely in the laboratory.5–9 However, the technique still suffers from long data collection times and the need for post-collection computer processing.10 Other than satellite remote earth sensing, uses of hyperspectral imaging outside the laboratory have been limited because of the restricted portability of most instruments. Speed, size, maintenance, sensitivity, complexity, and cost remain significant challenges to widespread adoption of hyperspectral measurement platforms outside a laboratory or satellite setting. Interference filter based multivariate optical computing (MOC) attempts to combine the data collection and processing steps of a traditional multivariate chemical analysis in a single step, offering an all-optical computing technology with no moving parts.11 MOC instruments have characteristics that lend themselves well to compact, portable, rapid, and sensitive imaging of multivariate information content in optical spectra. In filter-based MOC, a specialized interference filter called a multivariate optical element (MOE) is used as an optical beamsplitter. The intensity difference between the light rays that are transmitted and reflected by the MOE is designed to equate to the magnitude of a specific multivariate pattern in the light spectrum. Details of the theory and design of these MOEs have been previously reported.12–14 Tradi-

Improved Dispersion of Bacterial Endospores for Quantitative Infrared Sampling on Gold Coated Porous Alumina Membranes

An improved method for qualitative and quantitative sampling of bacterial endospores using Fourier transform infrared (FT-IR) microscopy on gold-coated porous alumina membranes is presented. Bacillus subtilis endospores were filtered onto gold-coated alumina membranes serving as substrates. Studies in the mid-infrared (MIR) region revealed the characteristic bacterial absorption spectrum at low surface concentration, while scanning electron microscopy (SEM) images of the same samples provided precise calculation of the surface concentration of the bacterial endospores. Under the conditions of study, the average concentration of endospores was determined to be 1356 ± 35 spores in a 100 × 100 μm2 area, with a relative standard deviation of 0.0260. Examination of ten random spots on multiple substrates with FT-IR microscopy apertured to the same area gave an average relative standard deviation of 0.0482 in the signal strength of the amide A band at 3278 cm−1. An extinction cross-section in reflection of σext = (7.8 ± 0.6) × 10−9 cm2/endospore was calculated for the amide A band at the frequency of its peak absorbance, 3278 cm−1. The absorption cross-section of the amide A band in reflection is estimated to be σabs ≈ (2.10 ± 0.12) × 10−9 cm2/endospore.

Hyperspectral imaging sensors using a novel 2D to 1D fiber array

A new technique is described for simultaneously obtaining chemical images in an imaging sensor with high spectral resolution over a wide spectral range. The technique is demonstrated using a combination CO2/O2 sensor, in three different sensor configurations. It is shown that it is potentially feasible to measure as many as 300 indicators on the tip of an image guide simultaneously at many different wavelengths.

Multilayer fiber optic chemical sensors employing organically modified SiO2 and mixed TiO2/SiO2 sol gel membranes

Fiber-optic sensors have been developed that incorporate multi-layer organically modified silica sol-gel membranes. pH sensors use a single layer hydrophilic organo-silica sol-gel membrane with a covalently attached pH sensitive dye, hydroxypyrene trisulfonic acid. The hydrophilic coating is made by copolymerizing silanol-terminated polydimethylsiloxane and tetraethylorthosilicate with 3-aminopropyltriethoxysilane. Unlike previous methods, which use acid as a catalyst, we have found that a base catalyst produces optically transparent gels. The sol-gel coated sensors are simple to make and require drying and aging times of as little as one day. Sensors made using these gels exhibit very good long-term stability, fast response times and no dye leaching. pCO2 sensors were fabricated using the same pH sensitive sol-gel layer overcoated with a hydrophobic high organic content sol- gel membrane. The response of the pH and pCO2 sensors is very fast due to the high porosity of the sol-gel membranes. Although in-situ sensors have been described for a number of organic and inorganic species, many volatile organochloride compounds (VOCs), such as perchloroethylene (PCE) and trichloroethylene (TCE), have been difficult to measure using current fiber-optic sensor transduction schemes. One of the optical sensors described here is a multilayer (3-4) sol-gel system that incorporates a TiO2/SiO2 membrane to degrade VOCs into smaller, detectable products. Upon exposure to UV light, TiO2, a semiconductor with a bandgap of 3.2 eV, produces highly reactive electron-hole pairs that are capable of photodegrading most organic compounds. The VOCs mentioned above are sensitive to degradative oxidation on TiO2 surfaces. During photodegradation of VOCs a number of products are formed including H+, HCl, CO2 and a number of smaller hydrocarbons. These products are produced in the TiO2 membrane and on TiO2 surfaces and the products diffuse into the nearby indicator membrane where they are detected. Carbon dioxide and protons produced are detected by the pH sensitive indicator layer described above. pH and CO2 sensors, and preliminary data for the measurement of VOCs will be presented.