Overview

Our group uses positrons and light for investigations in several highly relevant areas.

    1. Analytic, nondispersive Raman spectroscopy for measuring dissolved gases in seawater, gaseous emissions associated with earthquake pre-tremors, and breath analyses associated with complex medical conditions
    2. Further, we are working to develop intense positron sources

Our research extends from H₂ molecules to liver diseases with a significant emphasis on using our know-how to minimize human catastrophes and maximize medical treatment efficacies.

Faculty Collaborators

    Philip Varghese, Professor, Department of Aerospace Engineering and Engineering Mechanics, University of Texas at Austin
    Jacek Borysow, Professor, Department of Physics, Michigan Tech University

Graduate and Undergraduate Students

    Jeremy Johnson and Katie Schalk
    Sean Lantz and Santiago Jose Benavides

Research Underway

1 Spontaneous Raman Spectroscopy

A. Introduction

Raman spectroscopy is becoming a more significant spectroscopic tool in science, technology and medicine for the following reasons:

(1) the spectra frequently consist of tightly defined sets of very narrow lines, thus minimizing interferences,

(2) laser frequencies can be optimized for each analysis; allowing single mode, near-infrared, tunable laser diodes, which are small, efficient, and cost effective,

(3) photo-avalanche diode detectors exist with high quantum efficiencies to be used to study very dilute gas mixtures,

(4) Raman spectra can be recorded from gases, liquids and surfaces.

B The Analytic Non-Dispersive Raman Spectrometer (ANDRaS^TM)

The group's research efforts are based on measurements of the presence and density distributions of small molecules and their isotopologues as tracers at the location of interest. The Raman spectra of all small molecules are well-known, and the detailed study of one Raman line is sufficient to measure concentrations or chemical reaction rates; this increases the sensitivity and resolution of the final results. Often, Raman lines are broadened very little in solution; thus experiments can be carried out for suspensions in gases, for solvates in liquids, and for deposits on surfaces. The spectrometer can be replaced by a narrow band pass filter (±3 to 10 nm); this relaxes the optical constraints and provides access to extended emitting volumes and a large acceptance cone. The CCD detector is exchanged for a photo avalanche diode (APD) with high quantum efficiency.

The team's final, modified non-dispersive Raman analyzer consists of few components, for carrying out the tracer determinations. A temperature-controlled laser diode forms a tunable, single-mode, incident beam. It enters a confocal multi-pass cell, where the laser intensity in the central cross-over increases about 50-fold. The scattering volume at the center of the cell is about 1mm long, 0.5mm deep and 0.1mm thick. The volume imaged onto the detector is somewhat smaller. Molecules present in this volume scatter photons which are collected by an aspherical lens with a large acceptance cone at a right angle to the cell axis. The focal point of the lens is moved to the scattering center, so that the collected photons form a parallel beam. A narrow band pass filter is inserted and about ±3 nm of the spectrum will pass. The frequency of the laser beam is set so that the spectral line of interest will be able to pass the filter. The spectrally selected beam is split into two components with a pellicle. The intensity in either beam is proportional to the concentration of the species in the center of the multi-pass cell.

Often two Raman lines are particularly closely spaced, as is common when mixtures of isotopologues have to be analyzed. A particularly effective solution to detect their line intensities without loss is shown in Figure 1. For ¹²CO₂ and ¹³CO₂ starting with a single mode diode laser, the Raman dyads lines will be at 1388 cm^-1 and 1370 cm^-1. With a laser wavelength of 795 nm, the lines will pass a narrow band-pass filter whose center of the transition function is at resonance with D1 line of Cs at 894 ± 2nm. This requirement determines the frequencies of the incident laser and the narrow band pass filter. One beam of target photons will pass through the Cs-vapor cell (at about 250°C and 10 cm long). The frequency of the diode laser beam is tuned so that the resonance D1 line coincides with the Raman component which originates from the most abundant isotopologue (¹²C). After the line has been absorbed by the Cs-vapor, the remaining components, from other isotopologues (such as ¹³C), reach the detector. The widths of these Raman lines are determined by the Doppler effect of the Brownian motion of the target gas and the monochromaticity of the incident laser beam. Both are less than 0.02cm^-2. Line removal with alkali vapor cells can be achieved with high resolution as the Doppler line width of the Cs-vapor is about 500 MHz [1]. The same procedure can be used to remove the Rayleigh line if spectral features with small Raman shifts, such as rotational transitions [2] are to be investigated. An alternate approach avoids the beam set-up all together; the laser frequency can be tuned over the 2 spectral components of CO₂.

The alkali-vapor-filtered beam and the original beam are focused on the APD after passing a double-slotted optical chopper. The opening time of the shutters controls the time during which each beam is recorded. Sometimes the photon intensity is too low to gain a useful signal-to-noise ratio with the APD, and then the signal is recorded with a lock-in amplifier. Figure 2 shows the totally assembled unit.

Figure 1: Raman analyzer for closely-spaced Raman lines for any intensity ratio

Figure 2: The assembled ANDRaS^TM unit

Results for chapter B

In summary, our apparatus, described above, is an Analytic Non Dispersive Raman Spectrum (ANDRaS) analyzer, which detects a variety of trace impurities, isotopes resolved, with high sensitivity. The examples below describe several applications in more details. In general, our unit replaces a gas chromatography, mass spectrometer system at a much lower cost, it provides absolute intensities and its compact size makes it portable.

C Medical applications

There are several well-understood and widely applied techniques available to physicians to diagnose diseases with breath analysis. The breath is investigated for changes of its composition before and after the administration of a ¹³C-labeled compound. The molecules are metabolized at a specific site, such as the liver or the small intestine, and ¹²CO₂ with ¹³CO₂ are generated as a byproduct. This is a noninvasive technique, and hopefully it will see many new applications in the future. The analytic instrument of choice today is gas chromatography with mass spectrometry. However these analyzers are expensive and sufficiently complex that for absolute results repeated calibrations are required. The size of these instruments makes portability difficult and they are available only at facilities with strong research interests.

The Raman system, we have developed, are particularly suitable to determine the isotope ratios of the constituent atoms in molecules such as CO₂ offering the possibility to become a major alternative to mass spectroscopy. The ¹³C to ¹²C ratio is about 1% for all natural products. However, when one ingests a suitable compound, which is enriched with ¹³C, then the ¹³C2 content in the breath will increase, and the higher carbon isotope ratio does reflect the biological functionality of the active site in the body [3]. The first and most common diagnostic technique today, based on ¹³C in the breath, is applied to detect Helicobacter Pylori viruses in the stomach. They initiate ulcers, and eventually cancer. Currently, ten different diseases can be analyzed by this technique. It is an inexpensive, noninvasive diagnostic tool to be used by a nurse in a physician's office, Note our Raman unit will simultaneously measure congenital Lactose sensitivity in newborn babies as well as 4 distinct liver diseases [4]. After a disease has been identified and the treatment begins, the recovering processes can be studied with follow-up measurements. This opens the possibility to adjust the treatments for optimal recovery. We hope that in the near future new molecules will be identified which are decomposed at a characteristic site, and ¹³CO₂ provides access to a new analysis for this organ.

We expect that the first applications of ANDRaS^TM will consist of measurements on patients which have symptoms that are common for liver malfunctions. It is known for many years that the major activities of the liver can be subdivided into 4 segments. The functionality of each group can be investigated with ¹³C-labeled enzymes or amino acids. The analysis is currently done with gas chromatographs and mass spectrometers. It sets the golden standard. The same breath samples can be analyzed by ANDRaS^TM. We plan to work with several physicians to evaluate 2400 patients half of them healthy; the other half suffers of one of the four liver malfunctions.

Comparisons of the results will be the foundation for approval from the Health and Food Administration (FDA). We expect few problems since ANDRaS is a non-invasive procedure but replaces the expensive GC-mass spectrometer units only.

Another breath analysis with ANDRaS^TM involves lactose intolerance. Lactose is a disaccharide ingested by humans with milk products and their derivatives. In the small intestine, the enzyme lactase is produced and it hydrolyzes the D-lactose to form D-galactose and D-glucose. They transfer into the bloodstream through the intestinal walls. These mono-saccharine are then metabolized to H₂O and CO₂ in the muscles. For two reasons the decomposition might not take place; either the small intestine is infected (temporary) or a key codon in a critical position of the DNA is missing (congenital) and no lactase is produced. In these situations the lactose moves to the large intestine and bowl where it is fermented and generates hydrogen. Therefore lactose sensitivity is accompanied by the presence of hydrogen in the breath and the patients have often diarrhea. The situation is particularly serious when a newly-born baby has congenital lactose intolerance. A quick hydrogen test is a first indicator but it is only 60% reliable. When no H2 is found in the breath, then the patients' problems are most likely an allergy toward milk born proteins. A definite answer requires a ¹³C test with either mass spectrometry analysis or ANDRaS^TM. The baby will be fed ¹³C enriched lactose (5 mg per lb body weight) and the breath will be analyzed for the ratio of ¹³CO₂ to ¹²CO₂. Assuming that 0.5 % of all newborns have congenital lactose intolerance, about 20,000 babies are at risk per annum at the current birth rate of about 4 millions.

Results for Chapter C

Many more unique combinations of ¹³C-labeled compounds and a specific organ in the human body will be found, and the analysis of the breath for ¹³CO₂ will reveal the status of this organ's functionality. The non-invasive detection with Raman spectroscopy will provide the optimal analyzer.

D Impurities in Natural Gas

The Raman analyzer described above is not limited to detect CO₂ in air. Any mixture of gases can be monitored for a wide range of concentrations. In order to select the suitable narrow band filters, the Raman lines of the compounds to be recorded have to be known with uncertainties of less than ±1 cm^-1. The narrow band pass filters frequencies will identify either the carrier gas or the impurities. The pressures in the Raman cell should be kept under two atmospheres because contributions from collision induced emissions will either require constant monitoring or contribute significantly to the uncertainties.

Occasionally accidental resonances occur, they increase the pressure broadening and the spectral lines can no longer be insulated with a narrow band pass filter.

Figure 3: The totally depolarized Raman spectrum of ¹²CH₄

Figure 4: Spectrum of ¹²CH₄, observed (top) and simulated (bottom)

Two special tasks are presents in more detail: the determination of presence of CO₂ and/or H2O in natural gas with a precision of less than 1 ppm. Natural gas contains on average 64% methane, 10% ethane, 7% propane, 6% butane, 2% isobutene and 4% pentane.(Impurities: 7% nitrogen, 1% carbon dioxide) The Raman spectra of these hydrocarbons are known [5] The measurements have to be carried out at 20° above room temperature to avoid distortions of the densities of the organic compounds due to condensation at their dew points. Figs. 3 and 4 show the Raman spectrum of pure, gaseous ¹²CH₄.[6] The most intense line at about 3000 cm^-1 will be removed with a narrow notch filter to avoid that the detector becomes saturated. The data recorded from the mixture will be normalized to the CH₄ line at 1530 cm^-1. The rotational side bands can be avoided with a narrow band pass filter of ±3 nm width. To detect the concentration of the CO₂ a filter at 1285 cm^-1 will allow the higher energy line of the CO₂ Fermi dyad to pass without spurious contributions of the methane Raman scattering [2]. These settings will be certified by measurements with test gas samples. The water detection is rather straight forward since its main line of water occurs at 3650 cm^-1 and Fig. 3 shows that there are no interferences in this spectral range. [6]

E Detection of Precursors of Earth Quakes

Earthquakes have caused and will continue causing localized destruction around the world without substantial notice. In the last 12 years, the area around Istanbul and Izmit (Turkey) and Athens (Greece) have suffered substantial damage and casualties repeatedly from 1509 to 1999AD. The latest seismic events were not as catastrophic as previous ones, but future quakes are predicted to occur during the next 30 years in the Marmara region with a probability of greater than 65%. This area is particularly unstable due to the existing seismic gap and the post-1999 earthquake stress transfer at the west portion of the 1000 km long North Anatolian Fault Zone (NAFZ) which passes through the Marmara Sea about 15 km from Istanbul. The city is fully aware of this impending problem and the authorities are taking all conceivable physical and social steps for preparedness and risk mitigation.

Parallel efforts by local scientists have initiated the built-up of extensive on- and offshore seismographic, accelerometric and GPS networks with about 400 instruments, including dedicated early earthquake warnings and rapid response systems.[8]

The expectation to sense foreshocks and to design an efficient early warning system is based on the analysis of the 1999 data recorded around Izmit with seismographs. Figure 5 shows the seismic traces at the station UCG 1.3 km north of the epicenter; the main-shock occurring at time 00h 01min 39sec UTC on August 16 1999 with a strength ML = 7.4. The seismograms above the main shock trace show precursors with various strengths, from 0.1 ML to 2.5 ML at the UCG station. The earliest recorded event occurred 43 minutes before the main event. The seismic traces for all foreshocks start with a high frequency S-wave which is followed by a P-wave, with lower frequency. The speed of sound of these waves is different by approximately a factor 2. This speed differential shows the location of the source is 14 km underground, the same for all traces of the quakes. The event occurred due to the seismotectronic slip rate of 2.7 to 3.0 cm/year of the right lateral E-W slip across the Maramara region of the NAFZ.

This series of recordings raised the expectation that anticipating an earthquake event via foreshocks using a multitude of detectors may be possible. Seismographs are the current preferred technology. Non dispersive Raman analyzers are presently being discussed to be included for testing in the field. Changes of the composition and concentration of dissolved gases in springs around an epicenter have been found at several sites. The gases have been released by rock displacements at the epicenter of the quake.

Figure 5: Seismograms of Izmit Mainshock and the Foreshocks

Water samples collected after quakes had shown excess of H₂, CO₂, CH₄ and H2S. The ANDRaS^TM system is very sensitive; therefore there is a chance sufficient amount of gases can be detected due to the foreshocks. In case both technologies find correlated signals, an underground movement between the rocks most certainly had occurred and support prediction of a significant earth quake.

The Marmara region has been chosen (as a "supersite") for a large scale test program to improve our understanding of the occurrence of geological disasters. An international project is currently underway to carry out long-term hazard-monitoring; and development of novel detectors and their instrumentation that will serve the next generation of geo-hazard/observation systems. The results will strengthen the predictability of geo-hazards through the development of sophisticated models focusing on extreme events at the regional level.[5] The improvements of understanding of hazard and risk management will lead to a better protection of citizens and socio-economic life and the built environment. Our Raman unit can be considered a novel instrument to monitor dissolved gases as listed above; however some changes have to be made to adapt them for this environment.

Results from Chapter E

As described in section B and C the Raman analyzer used so far was dedicated to gaseous targets with traces of impurities. For the Marmara project the Raman head has to be submerged in water to the sources of interest. At full operation, all locations marked in Fig. 6 by red and blue circles will be equipped with our instrument for continuous recording of the dissolved volcanic gases for an extended period of time. The northern branch of the NAFZ is constantly in motion and this large data set of long duration aims at the determination of a base line to identify the variations of the seismic and chemical parameters.

Figure 6: The location of the on and offshore observation systems in MARSITE project region. Red: Broad-band seismology stations. Black: broad-band OBS stations. Yellow: GPS stations. Green: soil radon stations. Blue: Spring water stations. Purple Till meter stations.

F Determination of Impurities, Deposits and Defects on Surfaces with isotopic Resolution

A review on laser based techniques capable of detecting minute traces of inorganic compounds have been published recently [9]. However, these instruments are built on single mode lasers, high resolution spectrometers, charge coupled detectors (CCD) and special environments like vibration free and temperature controlled setting. In general, samples are taken at the site of interest and they are transferred and analyzed in a suitable equipped laboratory. During this transfer one assumes that the samples do not alter their compositions Therefore it is preferred that the analyzers are capable to be transported to the sampling site and operate there without loss of functionality. Raman and IR absorption spectroscopy are often used for identification and quantization of mixtures of chemical species with high selectivity. IR spectra are usually very complex and extremely high resolution is required to separate individual lines. In addition most of the molecular transitions are deep in the infrared and few useful intense lasers are available to use the ground state to excite from. Nevertheless IR absorption spectroscopy is ubiquitous and found its most sensitive application in ring-down spectroscopy (RDS) [10]

There are many Raman systems on the market today; however, they all suffer from the same drawback. Raman cross sections are extremely small; therefore only dense materials (solids or liquids) in sufficiently large quantities are routinely analyzed. Raman spectrometers capable of detecting low concentrations of gaseous substances have been reported in elaborate intra-cavity laser setups as demonstrated by Taylor at all [11].

Recently we have built a Raman analyzer which can differentiate the isotopes of hydrogen at concentrations as low as 5×10¹³ cm^-3 in air at STP [2]. The high sensitivity was achieved using a multi-pass cell in conjunction with an atomic vapor Rb absorption filter which eliminates the Rayleigh scattered light [1]. However, the vapor pressures of many organic solids such as polycyclic ether, natural products or nitrates at room temperature are significantly lower than this detection limit. Therefore this spectrometer cannot be used in its current form for the detection of these compounds as tracers.

The design of the improved analyzer presented here [12] takes advantage of the ability of the oxidized silicon wafer surface to attract via electrostatic forces large variety of organic and inorganic molecules. The presence, the rate of adsorption and desorption of organic molecules, especially hydrocarbons have been studied because they cause serious problems in the advanced electronics fabrication processes [13]. Still, relatively little is known about the interactions of the vapors of many inorganic compounds with silicon dioxide surfaces. When the sticking coefficients for these species on silicon wafers are as high as for most organic hydrocarbons, SiO₂ will be an extremely efficient sample holder. The Raman spectrometer design, described here will reach ultimate detection limits of nano-moles as will be shown in this study for ammonium nitrate (NH₄NO₃ or AN).

Our approach will be the first step on the way to the development of systems, which can measure quickly over a very large dynamic range a rich variety of molecules, calibrated to an absolute scale. Traces of AN have been deposited on a silicon wafer surface, to demonstrate the performance of our novel Raman set-up. We expect that measurements of the concentrations of AN in a river, such as the Mississippi, will be very

Figure 7: Optical arrangement of a Raman spectrometer designed to measure deposits and defects on surfaces.

useful for the evaluation of data collected in research projects which focus on the environment. The nitrate compounds are routinely used as fertilizers. An appreciable amount is transferred by the weather to the local tributaries. For instance the Iowa land area accounts for 5% of the Mississippi River basin, Iowa's streams almost 25% of the nitrate-nitrogen the River delivers to the Gulf of Mexico. The average nitrate-N was 5.2 mg/L [14]. Our Raman spectrometer is rugged enough and affordable that one can equip a high school in every district on the river with a Raman instrument and ask the seniors at a local high school to record the daily changes of AN for extended time periods to establish seasonal and temporal variations.

All experiments presented here were performed using a modified setup similar as shown in Fig. 6 described by Fink [15] and used for the first time by Kiefer [16]. The major components of the Raman apparatus are:

    (1) A laser diode (LD) tuned to the D2 line of rubidium (Rb) near 780 nm. The diode current was constant to .01 micro-ampere
    (2) A pair of 50.2-mm diameter concave mirrors (SM1, SM2) made out of BK7 glass with a 100.0-mm radius of curvature, separated by a distance of about 200 mm. The nominal reflectivity of the mirror at normal incidence is better than 99.99%
    (3) A polished, N-doped silicon wafer (S), with a native layer of oxide with a thickness of about 50nm;
    (4) collection optics (CL, L2);
    (5) and a 0.275-m spectrograph (SP) coupled to a cooled CCD camera.
    (6) The laser beam formed at the beam splitter is fed by 2 mirrors into an Rb cell (Rb) and a Fabry-Perot (IN) to show the operator that the laser beam is monochromatic and at the right wavelength.

The holographic notch filter (NF) with OD >5 was used to removed the Rayleigh scattered light. A Dove prism (DP) was used to rotate the line image of the focused laser beam at the sample and to bring it in parallel to the entrance slit of the spectrograph.

A tunable laser diode system was set up composed of a 6 cm laser cavity with a laser diode (SHARP GH0781JA2C, power of 120 mW), a collimating lens, and a diffraction grating (G) provide the optical feedback. The 1200 grooves/mm grating (G) blazed at 750 nm was mounted in the Littrov configuration and the 0-order was used as the laser output. A mirror (M1) mounted at right angle, next to the diffraction grating compensated for horizontal beam displacement caused by grating rotation during tuning. The temperature of the chip was kept constant by a thermoelectric cooler. Once tuned to the frequency of the rubidium D2 line the laser drift was less than 1.0 GHz per hour without an active locking frequency mechanism. The fluorescence monitored by the Infrared viewer from the rubidium reference cell RC was used to tune the laser to Rb D2 line. The mode structure was monitored by the scanning confocal Fabry-Perot interferometer with 7.5 GHz free spectral range and finesse of 200 denoted as (IN) in Fig. 6. The laser diode was powered by the commercial current source. The estimated power of the monochromatic light was about 40 mW. The laser light was polarized vertically to the table plane.

A 120-mm focal length convex lens L1 placed at such a distance from the spherical concave mirror SM1 that the laser at nearly grazing angle is focused at the silicon surface. This image is subsequently projected at almost the same place after reflection from mirror SM2. After each subsequent reflection the angle of incident at the silicon surface increased together with the fraction of laser light entering the silicon (refracted light). The refracted laser light was completely absorbed in the silicon. The estimated increase of the light intensity at the target in the multi-pass configuration and the consequent increase in Raman signal is nearly factor of 5 compared to a single pass. We assumed silicon's index of refraction at 780 nm to be equal to 3.70 [17] and used Fresnel equations [18] to compute the intensity of the reflected light as a function of the angle of incidence.

The Raman scattered light was collected by a multi-element uncoated condenser CL made of Borosilicate Crown Glass, with f/0.7 with a back focal length of about 25 mm. A plano-convex lens L2, 50-mm diameter and 200-mm focal length, imaged the scattering volume to the entrance slit of a 0.275-m Turner-Czerny spectrograph. The light image was rotated by 90o by a Dove prism (DP) before entering the spectrograph. This arrangement imaged approximately 20 micron×0.8 mm of a focused laser beams area on the Si disk onto the 160 micron×6 mm entrance slit of the spectrograph. The imaging optics matched closely the f/4 number of the spectrograph. The overall magnification of the collection optics was about 8. The grating used in all measurements has 600 grooves/mm and it is blazed at 1.0 micron. The resulting resolution is 0.2 nm (or 3.2 cm^-1). A back illuminated, cooled (243 K) Hamamatsu CCD array with 1024×256 24 micron pixels with a well capacity of 300,000 electrons per pixel, was used as a light detector. The dark electron count at 243 K was about 10 electrons per pixel per second. Most spectra were taken with an exposure time between 0.01 s and 1.0 minute.

Solid AN was dissolved in water with a concentration of 43 g/liter. Then, a pre-measured drop of the solution, approximately 1/37 ml, was applied to the silicon to cover the surface area of 1.7x1.7 cm³. The resulting surface density of AN was 2.9×10²² molecules/m² or about 1000 monolayer of NH₄NO₃, (specific density of 1.725 gr/cm³). The area, where the laser beam interacted with the NH₄NO3 crystals was estimated to be 2.0×10^-8 m². The laser was operating in the TM00 mode, with a Gaussian beam profile. Assuming a uniform distribution of AN on the surface of silicon, we concluded that the excitation laser light interacted with 4.5×10¹⁴ molecules or 0.5nmol of AN.

Fig. 6 shows a representative Raman spectrum of deposited AN on a silicon dioxide surface, obtained with our apparatus, is. The spectrum was taken in open air at room temperature of about 295 K and relative humidity near 70 %. The spectrum shown in Fig. 6 is not corrected for response function of the spectrograph and CCD camera. The identification of the ammonium spectral lines is according to reference [19, 20] for phase III of AN. The unfiltered remnants of Rayleigh line at 780 nm were outside the active area of the CCD detector. A background was subtracted for Figure 6 to bring the baseline to zero counts. Based on our estimates there were approximately 1000 layers of AN on the surface of silicon dioxide.

Figure 6: Raman spectrum of NH₄NO₃ deposited on SiO₂

This translates to a signal-to-noise ratio of 1 for 20 layers of NH₄NO₃, a value usually defined as the detection limit. Under our experimental conditions (very short integration times) the shot noise determines the noise level. Thus increasing observation time from 0.01 s to 1 minute will lower the noise level 80 fold and brings the detection limit to the 1/4 of a monolayer of deposited AN on the silicon.

Results for chapter F

We have studied the capability of a simple, relatively inexpensive Raman spectrometer as a monitor for traces of molecular species attached to the surface of silicon. We used as an example of AN, which is a commonly used as fertilizer, to demonstrate that with this spectrometer sub-monolayer concentrations of molecules can be detected. It should be straight forward to extent this research to dissolved NO₃-. The substrate will be replaced by a gold plated disk with plasma discharged deposited SiO₂ (50nm). This will increase the Laser power in the multi-pass cell by a factor 8. A known amount of diluted AN solution will be deposited on the SiO₂ substrate, the water will evaporate and the remaining NO₃ concentration can be recorded.

2 Development of an intense positron source

Positron annihilation spectroscopy is a well established research tool to characterize the surface and bulk electron distributions of any chosen material. However, the impressive results obtained with this technique have not yielded positron investigations and monitoring in an industrial environment. The greatest hindrance is the lack of sufficiently intense positron sources to record a simple Doppler distribution function with good statistics within minutes at a fabrication line. In general, most positron spectrometers are equipped with radioactive sources which produce at best only modest intensities of 10⁶ e+/s. An improvement by at least a factor of at least 100 is needed to become viable for on-line positron metrology. High-energy accelerators or nuclear reactors are not the answer due to their size and costs.

We propose to use standard technologies to generate an e⁺ beam with good electron optical properties, such as a small diverging angle, a small diameter and with a flux of 10⁸ e⁺/s. Beta positrons from a 10Ci source will be moderated with a 12-times folded W-mesh network. The low energy positrons will be accelerated into an electron optic focusing device which forces the positron trajectories to cross at a small (1-2 mm²0) area above the electrodes. A second moderator (solid rare gas) will form the source for a gun to generate a beam of mono-energetic positrons of any desired energy.

The positron will be used to research the new dielectrics in MOSFETs with HfO² doping and to studies of micro-voids in plastic membranes, which are used in industry to separate gases mixtures

References

[1] J. Borysow and M. Fink, "NIR Raman spectrometer for monitoring protonation reactions in gaseous hydrogen", J. Nucl. Materials, <341>, 224-230, (2005)

[2] J. Borysow and M. Fink, "Ultra high resolution NIR Raman spectrometer" Appl. Spectrosc. <60>, 54-56, (2006)

[3] P. Klein, "Clinical Application of ¹³CO₂ Measurements". Fed Proc. <42>, 2698 – 2701, (1982)

[4] Nista, L. Miele, P. Cermonini, I. A. Cazzato, A. Grieco, and G. Gasparrini Review article: "Breath testing for human liver function assessment Aliment" Pharmacol Ther. <16>, 1977-1997, (2002)

[5] N. A. Atamas, A. M. Yaremko, T. Seeger,A. Leipertz, A. Bienco, Z. Latajka, H. Ratajczak and A. J. Barnes, "A Study of the Raman spectra of Alkanes in the Fermi-resonance region", J. of Molecular Spectroscopy, <708>, 189-195, (2004)

[6] E. Joiurdanneau, F. Chaussard, R. Saint-Loup, T. Gabard and H. Berger, "The Methane Raman Spectrum from 1200 to 5500 cm^-1". J. of Molecular Spectroscopy, <233>, 219-230, (2005)

[7] G. Avila, J. M. Fernandez, G. Tejeda and S. Montero, "The Raman spectra and cross-section of H2O, D2O in the OH/OD stretching region", J. of Molecular Spectroscopy, <223>, 166-180 (2004)

[8] S. Oezalaybey, M. Ergin, M.Aktar, C. Tapirdamaz, F. Bicmen and A. Yoeruek,"The Izmit Earthquake Sequence in Turkey; Seismological and Tectonic Aspects", Bull. of the Seismological Society of America <92>, 376-386 (2002).

[9] A. Weber,"High resolution Raman spectroscopy of gases", Handbook of High-resolution Spectroscopy; Edi. M. Quack and F. Merkt, Wiley, <2>, 1153-1236 (2011).

[10] M. D. Wheeler, S. M. Newman, A. J. Orr-Ewing, and M. N. R. Ashford, "Cavity ring-down spectroscopy," J. Chem. Soc. Faraday Trans. <94>, 337–351 (1998).

[11] D. J. Taylor, M, Glugla, and R-D. Penzhorn, Rev. Sci. Instrum., <72>, 1970-1977 (2001)

[12] Nearly the same design of multi-pass Raman spectrometer was first used in investigation of silicon wafer contamination by FeCl3, A. Ross, University of Texas at Austin, unpublished, (1993)

[13] H. Habuka, M. Shimada and K. Okuyama J. Electrochem., Soc., <147>, 2319, (2000).

[14] R. D. Libra, "Nitrate Nitrogen: Iowa's unintended export", Iowa Geology, Iowa Dep. of Natural Resources, 1998.

[15] J. C. Robinson, M. Fink and A. Mihill, Rev. Sci. Instr. <63>, 3280 (1992).

[16] W. Kiefer, H. J. Bernstein, H. Wieser and M. Danyluk, J. Mol. Spectrosc. <43>, 125, (1972).

[17] CRC Handbook of Chemistry and Physics, <75> ed. CRC Press (1999)

[18] E. Hecht "Optics" pg. 111, Addison Wesley Longman Inc (1998)

[19] H.C. Tang and B.H. Torrie, J. Phys. Chem. Solids, <38>, 125 (1976)

[20] K. Awazu and H. Kawazoe, J. Appl. Phys.; Appl. Rev., <94>, 6243 (2003).

[21] R.N. Brown and A.C. McLaren, Proc. R. Soc. London, 226A, 329 (1962)