Theses

Frequently, in nuclear physics research, nuclear events are studied using the electrical or optical signals they produce in various materials. Research relating to neutrons has an especially strong need for robust and efficient signal classification algorithms that can be adapted quickly to varying materials and experiment types. Previous methods are considered, and a new graphically-driven software-based pulse shape discrimination approach is proposed. Our implementation of the method is explained in detail. Use of the method at BYU has significantly reduced the time required to begin new neutron-related research projects and adapt to changes in experimental setup. Furthermore, it has dramatically increased user understanding of the pulse classification process, reduced the number of errors made, and helped identify mistakes made in processing previous datasets.

With homeland security applications in mind, I constructed and tested a gamma ray and neutron detector designed to detect spontaneous fission events from highly shielded sources. The detector consisted primarily of a 25.4 cm $\times$ 25.4 cm $\times$ 15.2 cm block of Eljen EJ-200 plastic scintillator coupled to four Adit 5-inch photomultiplier tubes. The signals from all the tubes were added following gain matching with a $^{60}$Co gamma source. Using a $^{252}$Cf source, I measured the efficiency of the detector for various shielding types and thicknesses. Data were acquired both with a Cd foil placed on the front of the detector and with the Cd foil removed. The difference of the pulse-height histograms for these two configurations was shown to be a good measure of the neutron source strength. Neutron detection efficiency per fission neutron in a 4$\pi$ detector peaked at 13.5\%, corresponding to a shielding diameter of 20.4 cm. Through single and double pulse analysis, we were able to confidently determine if a fission source was present and also if boron was being used in the shielding.

We investigated the response magnitude across the full spatial extent of two commercially produced 5-inch photomultiplier tubes, the Hamamatsu R1250 and the Adit B133D01S. Each tube was translated by motorized stages across an incident light source. The response of the photomultiplier tube was recorded, as well as the response of a fast photodiode that simultaneously measured a portion of the incident light. Peak height and area responses were analyzed, with constant fraction discrimination implemented to determine the start and stop times of the peaks for the area calculations. The Hamamatsu response varied linearly across the tube, but had responses that varied by up to a factor of 10 when normalized to the center. The Adit was much more uniform, varying only by a factor of 0.3 across the majority of its spatial extent. These results indicate that the Adit is superior for magnitude of response experimental applications.

In recent years, there has been a helium-3 shortage, resulting in a need for alternative neutron detectors that can efficiently distinguish between neutrons and gammas. A hybrid liquid organic scintillator and lithium-6 glass neutron detector has been developed and tested to determine its neutron-gamma discrimination capabilities using pulse shape discrimination (PSD). Two liquid scintillators, Eljen EJ-325a and EJ-325UV, were compared. The EJ-325UV lacks a wavelength-shifting phosphor so as to minimize the significance of the absorption of lithium glass light in the scintillator. Principal component analysis (PCA) is explored as a method to improve neutron and gamma region separation. Previous work is extended from single- to correlated-pulse analysis, which declares an event a neutron if a liquid scintillator recoil pulse precedes a lithium capture pulse. It was found that PCA fails to improve region separation and efficiencies, but it does provide quantifiable measures of the significance of PSD parameters in distinguishing between events. The correlated-pulse analysis method reduces the ratio of misidentified gammas by an order of magnitude, but reduces the detector efficiency to less than one percent. The detector utilizing EJ-325UV identified 20-40% more neutrons than the EJ-325a detector. Further analysis is needed to determine the importance of self-absorption in the hybrid detector.

Thediscriminationofneutronsandgammarayshasbecomeanimportantfieldofresearchboth for scientific purposes as well as security purposes. Until recently helium-3 was used to discriminate neutron radiation from gamma radiation, however, with the depletion of helium-3 reserves, newmethodsmustbefound. TheBYUNuclearResearchGrouphasdevelopedahybridlithium-6 glass and liquid scintillator that combines the discrimination characteristics of both materials. Parameters such as pulse area, area distribution, and after peaking count are used to sort the gammas and neutrons into different regions. Particularly, the single pulse analysis methods as well as data analysis improvements are reported. These improvements provide tools for double pulse analysis. Future detector innovations are also discussed.

It is often difficult to accurately determine the number of neutrons incident on a detector because of room return, or the scattering of neutrons from nearby material into the detector. We developed a neutron detection platform mounted on a scissor lift so that room return can be minimized in our counting experiments. We measured the neutron counting rate as a function of the height of the platform above a concrete slab located below the lift. This measurement was taken with a Li-6 glass neutron detector used in conjunction with a Li-7 glass detector to subtract gamma background. We used neutrons from a Cf-252 source located at a fixed position on the platform. We found that room return became minimal when the lift was raised to a height of five meters above the concrete. This result is important for helping reduce room return in neutron detection research. With neutron detectors that are sensitive to low energy or thermal neutrons, as is our Li-6 detector, an effective way to account for room return is to have the detector and source beyond five meters from the wall and floor.

Lanthanide series elements are used in many different materials at small concentrations. One application is as the scintillating component of detector materials. X-Ray florescence (XRF) spectroscopy can be used to identify trace elements within materials. The goal of this research is to be able to build an XRF library of lanthanide series elements that can be used as a basis for other measurements. Also, I was able to make an estimate for the limits of detection in our XRF system. I found that K peaks are visible up to 0.1% concentrations by weight; and L series lines are visible up to 1% concentrations by weights. At smaller ratios than these benchmarks the spectrum become too uncertain to make valid measurements.

Prosthetic legs are very expensive, and persons with amputations in developing countries often don’t have access to prosthetic devices. 2ft Prosthetics researches low cost feet to give amputees an affordable option. One of the current designs is made out of layers of PVC. Amputees often complain that the foot is excessively loud during gait. To reduce this, we tested two different designs. One clinic in a developing country reported using an inner tube from a bike tire helped mute the sound. The second design tested used a screw drilled through the toe of the prosthesis to hold the layers of PVC together. To test the feet, we made multiple models of each design and tested them against the original PVC design. These feet were put on our testing fixture that compresses the toe and heel using two pneumatic pistons to simulate walking on the foot. While on the fixture, the sounds from these feet were recorded. We found the original, screw in toe, and tire design produced -27.2 dB, -31.0 dB, and -32.2dB of sound respectively during gait. The inner tube design was also tested for durability against the original design. Durability testing showed no significant difference between the durability of the two designs.

We have constructed a number of prototype cadmium capture-gated neutron detectors that are simple and inexpensive in design. For many applications, a large volume detector is desirable, so we built and tested a detector with a 10 in. x 10 in. x 6 in. head coupled to four Hamamatsu R1250 photomultiplier tubes. This detector has a maximum efficiency of about 12 percent for 2.00 MeV neutrons. These tubes are excellent for timing which was important for measuring the detector efficiency as a function of neutron energy, but not optimum for light collection. To improve efficiency, we are building a similar detector using Adit B133D01S photomultiplier tubes. We will describe the construction and operation of these detectors.

Abstract 1 Using the combination of a neutron sensitive 6Li glass scintillator detector with a neutron insensitive 7Li glass scintillator detector we are able to accurately measure the neutron capture rate of our 6Li detector. We used this detector with a Cf neutron source to measure the efects that both polyethylene and 5% borated polyethylene shielding has on our capture rates. Both of these measurements were compared with MCNP calculations to determine how well the calculations simulated the measurements, particularly when the source is highly shielded. MCNP is shown to have a general tendency to underestimate detector eficiency with polyethylene shielding. For pure polyethylene it underestimates the measured value at an average of 10%. This increases to an average of 18%for borated polyethylene. Abstract 2 The method for arriving at the statistical errors for the mass, multi-plication and alpha-n rate for a Pu source from a multiplicity counting measurement is explained. Next, three methods for calculating the errors from a simulation using MCNP-Polimi are presented and compared with the measured results. The three approaches considered here are: a multinomial distribution, a Poisson distribution, and a semi-empircal approach.

We investigated the uniformity of electron transit times across the full spatial extent of two 5” photomultiplier tubes, the Hamamatsu R1250 and the Adit B133D01. The photomultiplier tubes were translated across a localized incident light source while a portion of the incident light was simultaneously measured and recorded by a fast photodiode. Constant fraction discrimination was utilized to calculate electron transit times as the difference between the start times of the photodiode and photomultiplier traces. The Hamamatsu tube provided a uniform timing response that varied by no more than 1.7 ns. The Adit tube was extremely non-uniform with transit times that varied by as much as 57 ns, yet the symmetry of variation in transit times differed significantly when analyzed with two different algorithms to determine pulse timing. These results indicate that the Hamamatsu tube is superior to the Adit for experimental timing applications, and that the method of analysis significantly affects final timing results of the Adit tube.

An analysis of the U-235 fission spectrum measured using a cadmium capture-gated neutron detector is presented. A better knowledge of the fission spectrum of U-235 is beneficial for nuclear reactor and bomb design. The detection of low-energy neutrons from such fission is difficult when using pulse shape discrimination. A cadmium capture-gated neutron detector was used to analyze the fission spectrum of the U-235 fission chamber at Los Alamos Neutron Science Center. Capture-gated neutron detectors are more efficient at low energies because the cross section of neutron capture goes up as the inverse of energy. Also, they can discriminate well between gamma rays and neutrons at low energies because of the presence of a neutron capture pulse. The data show probable neutrons detected at energies lower than 0:5 MeV. However, due to discrepancies when compared with other data from the detector, we cannot draw any quantitative conclusions.

The cadmium capture-gated neutron spectrometer utilizes a dual-pulse signal from incoming neutrons to differentiate between neutrons and gamma rays. We have built such a detector and performed a time-of-flight experiment at Ohio University to measure incident neutron energy. We determined the detector efficiency as a function of neutron energy for neutrons with energies 0.5 MeV - 9 MeV. The detector has a peak efficiency of 12% for 2 MeV neutrons. The cadmium capture of the neutrons provides a low energy neutron detection boost that keeps the efficiency above 9% for neutrons with energy less than 2 MeV. A properly calibrated cadmium-capture gated neutron detector can be used to measure low energy neutrons from fission sources.

Neutron detection is an important component to Homeland Security. Portal monitors are put at points of entry into the country to detect illegal nuclear material entering the United States. In the past and currently, 3He-based detectors have and are being used in these portals. However, because of the current shortage of 3He, the BYU Nuclear Group is looking for alternative methods for neutron detection. In particular, I have been doing preliminary work on a two-photomultiplier tube hybrid neutron detector that utilizes lithium and cadmium components. This work focuses on how each component of the hybrid (lithium and cadmium) performs on its own. It then outlines what results are seen when both sides are simultaneously "watching" the same radiation source. Both the cadmium and lithium components work as expected when operating alone. When combined– i.e. both components are on and looking at the same radiation source–we see that the cadmium component is the dominating detection component in the hybrid detector. Relatively few events are seen from the lithium side. Further work should be done to confirm the results herein, to consider other setups that may yield more balanced results as well as to confirm the hope that a hybrid detector, such as the one here, can detect neutrons over a broader energy range than either component by itself. If this can be accomplished, the hybrid detector will become a more viable candidate for potential use in homeland security.

With the global supply of He-3 in short supply and an increased need for neutron detectors, Cd-based detectors provide an attractive alternative. I have done research on various possible designs using Monte Carlo methods to optimize detectors for efficient neutron capture.By examining Cd-based detectors with the highest neutron capture rate will be better able to distinguish neutrons and gamma rays in order decrease the possibility of false neutron detection.

The double pulse response of the Gadolinium Lithium Borate Cerium detector suggested its effectiveness for applications in Laboratory Nuclear Astrophysics, specifically in sparse neutron spectroscopy amidst the competition of background radiation. Using Cf-252 as a neutron source, a time of flight facility was built to determine the efficiency of the LGB detector. The detector has a $10\%$ efficiency in neutron detection, however, the energy detection efficiency is $2\%$. Thus, the total efficiency for neutron spectroscopy is $0.2\%$. Therefore, the LGB detector is not a good choice for neutron spectroscopy.

The fission spectrum of neutrons plays a signi cant role in the construction of nuclear reactors. However, the fi ssion spectrum of neutrons with kinetic energies less than 1 MeV, emitted from the spontaneous fission of 252Cf, is not as precisely known as that for higher energy neutrons. A method for making improved measurements of low-energy cross sections is described and results of preliminary tests are reported.

Due to a worldwide shortage of 3He, a key component of many neutron detectors, alternative technologies are needed to fully secure the world against the threat of nuclear terrorism. Lithium- 6-loaded glass scintillator is a material that has been widely used in neutron detectors for decades, but its comparatively high gamma sensitivity has made it unattractive for use in homeland security applications. We have recently identified and tested a new method for reducing the gamma sensitivity of a neutron detector constructed using 6Li-loaded glass scintillator. Our prototype neutron detector consists of small ( 1 mm3) shards of broken 6Li-loaded glass scintillator embedded in a matrix of Eljen Technology EJ-500 optical epoxy. Several inches of mineral oil can be placed behind the matrix for moderation. Monte Carlo simulations (conducted using the particle transport codes MCNP and PENELOPE) and experimental tests both show that this detector can achieve good neutron detection efficiency while obtaining dramatically reduced gamma sensitivity when it is compared to similar detectors that employ sheets of 6Li glass.

We have improved on our original design for a cadmium capture-gated neutron detector by designing a "wedge" detector. This new design allowed for more consistent pulses for the entire face of the attached photomultiplier tube. Testing at the Ohio University accelerator facility showed efficiency of this new detector lies around 12%. Testing was also performed with bismuth as a gamma shield to reduce the number of accidental events recorded. Three inches of bismuth proved to reduce the amount of accidental pulses recorded by the detector by a factor of 10.

The characterization of a newly developed capture-gated neutron spectrometer is presented. Such spectrometers produce a dual signal from incoming neutrons, allowing for differentiation between other particles, particularly gamma rays. The neutron provides a primary light pulse in either plastic or liquid scintillator through neutron-proton collisions. A capture material then captures the moderated neutron exciting the material. The capture material promptly de-excites with the emission of gamma rays, which then produces a second pulse. The present spectrometer alternates one-centimeter thick plastic scintillators with sheets of cadmium inserted in between for neutron capture. Each neutron captured in cadmium releases about 9 MeV of gamma energy. Many factors influence the operation of the spectrometer, including: materials, geometry, optics, and methods of analysis. Experiments are presented that weigh the effect that these factors have on the capabilities of the detector. Theoretical, computational and experimental data presented here also illustrates the detector's ability to be utilized in applications of neutron detection for nuclear non-proliferation. The intrinsic efficiency of the detector for the detection of spontaneous fission neutrons from Cf-252 is measured to be 5.23%.

We have modeled, built, and tested a neutron detector composed of three neutron detecting materials (cadmium, lithium, and scintillating plastic) into a single detector in an attempt to broaden the neutron energy detection range and to decrease the gamma-ray sensitivity compared to lithium plastic or cadmium-plastic. Computational modeling optimizes the geometry of the detector. Initial testing of the efficiency of the detector has shown ability to discriminate a neutron signal from a gamma-ray signal in the 6Li-glass. This work provides the foundation for further improvement of the detector concept.

We have designed a neutron detector using 6Li glass scintillator. 6Li has a large neutron capture cross section, which gives high neutron detection rates. 6Li captures the neutrons and the resulting 7Li rapidly decays into charged particles which are subsequently detected. We have measured neutron detection efficiency of 30 percent with gamma contamination of only 1 part in 10000.

After 9/11/2001, the increase of security measures has made an important role for neutron detectors. The installation of detectors to scan for radioactive material has caused a shortage of helium-3. We have been working on making a cadmium capture-gated detector as a replacement for helium-3 detectors. Through using MCNP, a Monte Carlo simulation software, I have been working at optimizing the capture efficiency of the cadmium detector. By varying plastic scintillator and cadmium sheet thicknesses I have shown that doubling the thickness of the cadmium sheets has almost no effect on the efficiency. Variances in the plastic thickness have a much greater effect.

Pulse height distribution of He-3 detectors used in international safeguards monitoring vary from that of a typical He-3 neutron counter. Data taken shows that between eighteen and twenty-five percent of the neutron counting efficiency is lost when using low-level discrimination of 100mV, per manufacturer’s recommendation. Also, the charge-coupled amplifiers of the detectors may be responsible for changing the pulse height distributions of the detectors.

My research group and I have built a cadmium capture-gated neutron detector, as envisioned by Dr. Bart Czirr. This novel design is intended to address the growing need for a helium-3-free neutron detector for national security and scientific research. Our detector takes advantage of neutron interactions with a scintillating plastic moderator and neutron capture in cadmium. A neutron entering the plastic loses energy to nuclear scattering and knocks protons free from nuclei. The protons lose energy to Coulomb scattering, producing a pulse of light in the scintillating plastic. The slowed neutron is captured in cadmium foil and releases 9 MeV of gamma rays, which Compton scatter to produce a second light pulse. This double pulse signals a neutron capture. This neutron detector was constructed using methods developed by my group. Testing has shown that our detector is able to detect ionizing radiation and discriminate between neutrons and gamma rays.

Data acquisition systems used for spectroscopy have undergone many changes over the past several years. Rapid changes in both hardware and software cause these systems to become obsolete very quickly. A new technology known as waveform digitization has recently become available. In principle, waveform digitizers could replace several components of older data acquisition systems. In this study, waveform digitizers were applied to x-ray and gamma-ray spectroscopy. These results are compared to the older data acquisition systems. Waveform digitizers have been found to be useful; however, some problems need to be overcome before they are widely used. Waveform digitizers work well when the signal does not have to be processed through an amplifier, as the amplifier makes the signal too wide for the digitizer to process. The different preamplifiers used with solid state detectors present other problems for waveform digitizers. Once the initial challenge of obtaining a spectrum was overcome, we set out to answer two questions: (a) Can waveform digitizers handle count rates typically encountered in spectroscopic applications? and (b) Can we obtain energy resolutions comparable with other data acquisition systems? We found that the waveform digitizer we used worked well at a rate of 20,000 counts per second, but that the resolution was not as good as expected. This problem with the preamplifier needs to be solved to make the digitizer suitable for spectroscopic applications.

In the last two decades, many studies have been published on metal enhanced fusion. Various types of metals have been used in these studies, but never pure lithium due to the difficulty of ensuring a clean lithium target. I demonstrate that exposing and maintaining a clean surface of a small quantity of lithium metal can be accomplished by forming a liquid lithium pendant droplet. This report includes a brief overview of past research of the deuteron-deuteron fusion reaction as well as a history of metal catalyzed fusion. I show that molten lithium does form a pendant droplet and discuss the variables that determine droplet formation. This report includes a description of the equipment development and the experimentation process, as well as the results.

Low ux neutron detection is challenging because of the low signal to back- ground ratio. Lithium gadolinium borate crystal embedded in a plastic scin- tillator enhances low ux neutron detection through paired-pulse scintillation. The signal from this detector contains an energy pulse and a capture-gated pulse. To calibrate this new scintillator, a method was developed to recognize the neutron signal and extract the energy information. Progress was made toward nding the proper gains to calibrate the detector against an accepted standard.

My research has been to try to establish and perfect non-destructive methods to discern between counterfeit coins and genuine ones. I was able to use XRF spectroscopy to develop methods of identifying incorrect materials in coins, incorrect proportions of materials in coins, and even invalid claims to ages of coins. I used an electron-microscope to identify minute alterations such as tampering with mint marks and/or dates. These methods greatly increase the ability to determine if a coin is counterfeit without damaging the coins.

The Open Physics Abstraction Layer (OPAL) is an open-source soft ware project that provides a library of physics routines to use in other programs. In this project, the efficacy of using OPAL to perform physics simulations was examined and found lacking. Among its major deficiencies are the inability to export numerical data, and the lack of commonly-used physics elements such as springs, dampers, or linear oscillators. To address these issues, a data-logging framework was created and directions are provided to guide a future developer.

This thesis paper examines the feasibility of using correction factors on human head hair with PIXE spectroscopy. Using computer models, we compared x-ray production from thin samples made by a lengthy acid-digestion process and thick samples which require minimal preparation. By comparing the two numbers, we determined experimental correction factors for hair in thick targets. This paper first describes the properties of hair and how PIXE is used in trace element analysis. It then details the separate computer programs used to collect data. The resulting x-ray counts are shown with an analysis on how the data are used to determine the corrections. We found that the computer programs are an accurate representation of the hair for proton energy loss. We determined correction factors, but were not able to experimentally test them due to time and equipment constraints.

In this thesis we present a proof of principle for the construction of pigment databases to facilitate the analysis of paintings using Particle-Induced X-ray Emission (PIXE) spectroscopic techniques. A small data set is constructed using internal-beam PIXE on 10 modern pigments. We compare this data set to the data obtained from thick targets made from combinations of these same pigments analyzed using external beam mode in a helium atmosphere. We show that the data set accurately predicts the composition of six of the eight targets. Limitations such as charging-induced background, the inability to resolve layering, and thickness-caused inaccuracies are discussed.

The purpose of the project was to develop a method for studying trace elements in eighteenth and nineteenth century printed materials. The elemental content in the ink will be used to distinguish ink of one source from ink of another. External PIXE (Particle Induced X-ray Emission) was the method used in this experiment to determine the elements in the inks. This paper discusses the steps of setting up an experiment of this nature as well as possible ways to make the results more effective. The only element that proved to be distinguishable in the ink was lead; however, lead was not present in most of the samples that were analyzed.

Thirty-six sediment samples were extracted from the bottom of Utah Lake in a grid-like pattern in the area of the former site of the Geneva Steel plant. A few of the samples were initially analyzed using an energy dispersive X-ray fluorescence spectrometer (EDXRF) to determine a rough estimate of the sediment matrix, which consisted mainly of calcium carbonate and silicon dioxide. All thirty-six samples were then analyzed with wavelength-dispersive XRF (WDXRF). The majority of the results were consistent with findings of previous investigations. However, higher counts of lead, arsenic, copper, and zinc were found in one of the samples. This was likely due to Geneva’s pumping of materials into the lake. This local sample was then analyzed with Particle-induced X-ray emission (PIXE) spectroscopy, and similar results were found. PIXE was also used to show that the preparation of samples for the XRF was introducing some trace amounts of tungsten into the samples.

The neutron-neutron (n-n) scattering length is a fundamental parameter in nuclear physics, but experiments give one of two different numbers and there is still no adequate explanation for this discrepancy. However, measurements are plagued with large uncertainties caused by neutron detector cross talk. Many experimentalists also rely upon computer code to calibrate their neutron detectors. We have developed a new neutron detector expressly for the purpose of improving the n-n scattering length measurement. It offers two important advantages: 1) minimal cross talk and 2) high counting efficiency. We calibrated the detector from 1 MeV to 6 MeV at 1 MeV increments. We have shown that the computer code, MCNP, does not always give the correct detector efficiency, and that reliance upon this code for calibration could be a large factor for error in previous experiments. Preliminary tests show no cross talk between two like detectors and suggest that these detectors are appropriate for a n-n scattering length measurement.

Before low-cost, hybrid neutron detectors can be tested, they must be able to produce strong optical signals. In order to study actual detector conditions, the optics of several detector designs with varying numbers of 2 mm acrylic disks in mineral oil were measured, and the total signal attenuation calculated was calculated. This was accomplished by first determining the amount of signal transmission using laser optics and comparisons with theoretical models of Fresnel Coefficients. The optical efficiency of multiple detector geometries was evaluated. Simulations for an actual detector were then created using Monte Carlo for Neutral Particles (MCNP) which gave information on the photon energies resulting from each neutron interaction. These simulations yielded varying light energies on the order of 0.5 MeVee (electron volt electron equivalent). With an approximate total signal attenuation of 61.2\% in both the mineral oil and acrylic disks and an actual attenuation of approximately 62.6\%, the resulting optical signal would have energies on the order of 300 keVee. As a result, we were able to conclude that a detector with this geometry would be optically viable.