MEASURING FUSION CROSS-SECTIONS FOR THE 20 O + 12 C SYSTEM AT NEAR BARRIER ENERGIES

Transcription

1 MEASURING FUSION CROSS-SECTIONS FOR THE 20 O + 12 C SYSTEM AT NEAR BARRIER ENERGIES Michael Rudolph Submitted to the faculty of the University Graduate School in partial fulfillment of the requirements for the degree Master of Sciences in the Department of Chemistry, Indiana University December 2011

2 Accepted by the Graduate Faculty, Indiana University, in partial fulfillment of the requirements for the degree Master of Sciences. Master s Thesis Committee: Romualdo T. de Souza, Ph. D., Chair Caroline C. Jarrold, Ph.D. W. Michael Snow, Ph. D. ii

3 Copyright 2012 Michael Rudolph iii

4 Dedication I wish to dedicate this thesis first and foremost to my savior Jesus Christ, in whom I can do all things. I also dedicate this thesis to my wife Laura and to my parents, Phillip and Marian. Their love, support, encouragement, and advice has given me the strength to persevere. iv

5 Acknowledgements I wish to thank my graduate advisor, Romualdo de Souza, for the example he set with his passion, hard work, and dedication to chemistry. I thank him for the time he spent working with me individually and for the lessons and experiences he has imparted to me all of which have proven invaluable to me. I am grateful for the service of the faculty serving on my Master s committee: Professors Caroline Jarrold and W. Michael Snow. I would like to thank Sylvie Hudan for her hard work and particularly for the time she invested over the past year to help me with the data analysis. Thank you for not only providing the tools to analyze the data but also for teaching me how to use them. I thank Zack Gosser and Kyle Brown for their literal blood, sweat, and tears they poured into the experiment. Thank you for your hard work, honesty, selflessness, and friendship. I am also grateful to have worked with Alan McIntosh, Brice Floyd, and Damien Mercier as part of the Nuclear Chemistry group at Indiana and for their contributions to the experiment. I thank the staff at Oak Ridge National Lab for accommodating us with beam time on more than one occasion to prepare for the experiment. I owe a great deal of gratitude to Dan Shapira and Felix Liang for the time, knowledge, and expertise they shared and for the role each of them has played in my development. I appreciate the hard work and service of my collaborators on the experiment: Kyle Brown, Zack Gosser, Sylvie Hudan, Damien Mercier, and Romualdo de Souza of Indiana University; Felix Liang and Dan Shapira of ORNL; Abdelouhad Chbihi of GANIL; and v

6 Michael Famiano of Western Michigan University. Thank you to the staff at GANIL for delivering the high purity oxygen beam for the experiment, for your knowledge and insight, and for your support and hospitality. I also wish to acknowledge the efforts of both the EIS and MIS groups in the chemistry department at Indiana. Without their skill and expertise the equipment and instrumentation for the experiment could not have been realized. We are all indebted to the United States Department of Energy, who graciously funded our efforts. Finally, I wish to thank Michael Famiano, Asghar Kayani, and Allan Kern of Western Michigan University for providing us with beam time and support following the GANIL experiment. vi

7 Michael Rudolph MEASURING FUSION CROSS-SECTIONS FOR THE 20 O + 12 C SYSTEM AT NEAR BARRIER ENERGIES The fusion of neutron-rich 20 O on 12 C at energies in the range of 20 MeV E lab 41 MeV was measured at the GANIL accelerator facility located in Caen, France. The design, principles of operation, and calibration of the detectors used to perform the measurement are described. Analysis of coincident evaporation residues and light charged particles produced in the decay of the compound nucleus, 32 Si*, has been performed. The fusion cross-section associated with the de-excitation of the compound nucleus leading to the emission of charged particles has been determined and compared with predictions of a simple fusion model. For reference, the fusion of 16 O + 12 C has been measured in the energy range of 20 MeV E lab 42 MeV and the fusion cross-section for the charged particle decay channels extracted. The 16 O fusion cross-section has also been compared with statistical model calculations in order to demonstrate the ability of the model to predict the formation and de-excitation of the compound nucleus for β-stable nuclei. Romualdo T. de Souza, Ph.D., Chair Caroline C. Jarrold, Ph.D. W. Michael Snow, Ph. D. vii

8 Table of Contents 1 Introduction 1 Bibliography Experimental Setup Overview Mechanical Setup The Detectors Active Collimator Active Degrader Ionization Chamber (CID) mm Microchannel Plate Det. / Active Target T3, T2 Annular Silicon Dets mm Microchannel Plate Det Zero Degree Ionization Chamber (ZDIC) Electronic Signal Processing Bibliography Inspection of Data Online CID viii

9 Table of Contents 3.2 SBD mm MCP, 40mm MCP T3, T ZDIC Bibliography Calibration Energy Calibration of the SBD Energy Calibration of T3, T Bibliography Evaporation Residues and Light Charged Particles Resulting from the 129 Fusion of 20 O + 12 C 5.1 Purity of the Beam Experimental Results Fusion of 16 O + 12 C Conclusions Bibliography Testing of the T2 silicon and 18mm MCP Overview Test Config. 1: Comparison of MCP Orientations Test Config. 2: Comparison of T2 Orientations Test Config. 3: Comparison of Different MCP Dets ix

10 Table of Contents 6.5 Test Config. 4: Coincidence of Two MCP Dets Bibliography Conclusions 171 Bibliography A Summary of Programs and Scripts Used in the Data Analysis 176 x

11 List of Figures 1.1 Isolated neutron star Accreting neutron star Structure of a neutron star Simple 1D model of the nuclear interaction potential Classification of nucleus nucleus collisions Fusion excitation functions of 12 C + 28, 29, 30 Si and 12 C + 16 O from Ref. [22] The Gamow peak Fusion excitation function of 9 Li + 70 Zn from Ref. [1] Fusion excitation functions of 16, 17, 18 O + 12 C from Ref. [4] Predicted energy distributions of evaporation residues for 20 O C from PACE Predicted angular distributions of evaporation residues for 20 O C from PACE Isometric view of the fusion chamber Isometric view of the full mechanical setup xi

12 List of Figures 2.7 Configuration of the detectors in the chamber Isometric view of the active collimator detector Operating regions for gas detectors Electron drift velocity for CF 4 gas mixtures Electron drift velocities for common ion chamber gases Basic Frisch grid ionization chamber Full assembly of the degrader ionization chamber (CID) Interior components of CID Segmented CID anode CID window assembly Tack welding apparatus for the CID and ZDIC windows CID detector with Plexiglas flange CID pre-amplifier box a Gain comparison between a single MCP and a Chevron MCP stack b Chevron MCP stack configuration Schematic of a triangular MCP detector Design of the 18mm MCP detector Circuit diagram of the 18mm MCP voltage divider board S1 silicon detector S2 silicon detector xii

13 List of Figures 2.26 Silicon detector cable conduits Block diagram for processing of silicon detector signals from the pie segments Circuit diagram of the silicon frequency splitter Circuit diagram of the silicon fast timing amplifier Silicon fast timing amplifier unit T2 pre-amplifier box T2 fast timing amplifier pulse for 226 Ra α source mm MCP detector design mm MCP anode Full ZDIC detector assembly ZDIC interior components ZDIC entrance window assembly ZDIC segmented anode ZDIC pre-amplifier box ZDIC Frisch grid assembly ZDIC electric field shaping lines Block diagram of the signal processing electronics Block diagram of the trigger electronics E loss spectra of the six CID anodes for 20 O E loss spectra of CID anode 3 for all pressures used in 20 O xiii

14 List of Figures experiment E loss spectra of the six CID anodes for 16 O One-dimensional SBD energy spectra for 20 O Two-dimensional energy spectra of E loss in CID vs. E SBD for 20 O and 16 O One-dimensional TOF between 18mm MCP and 40mm MCP for 20 O One-dimensional energy spectra of T3 pies for 20 O One-dimensional energy spectra of T2 pies for 20 O One-dimensional energy spectra of T3 rings for 20 O One-dimensional energy spectra of T2 rings for 20 O Two-dimensional Energy-TOF spectra of T2 pie 1 and T3 pie 9 for 20 O Two-dimensional Energy-TOF spectra of T2 pie 1 and T3 pie 9 for 16 O One-dimensional E loss spectra of ZDIC anodes for 20 O One-dimensional E loss spectra of select ZDIC anodes for all pressures used in 20 O experiment One-dimensional SBD energy spectrum for 226 Ra α source Energy calibration of SBD using a 226 Ra α source and 16, 20 O xiv

15 List of Figures beams Deviation of SBD calibration points from fit Calibrated one-dimensional SBD energy spectrum One-dimensional energy spectra of T2 pie 1 and T3 pie 9 using a 226 Ra α source Energy calibration of T2 pie 1 and T3 pie 9 using a 226 Ra α source and 16, 20 O beams Deviation of T2 pie 1and T3 pie 9 from calibration fit Calibrated two-dimensional Energy-TOF spectra of T2 pie 1 and T3 pie 9 for 20 O Two-dimensional spectra of E T2 Pies vs. E T2 Rings for calibration of T2 rings TProfile histrograms for E T2 Pies vs. E T2 Rings for calibration of T2 rings Calibrated two-dimensional Energy-TOF spectra summed over all T2 pies using E Pies compared to E Max of the pies and rings One-dimensional CID E loss spectra using 20 O and 20 F Two-dimensional Energy-TOF spectrum of T2 pies with the residue gate drawn for 20 O Two-dimensional Energy-TOF spectrum of T3 pies using 20 O Two-dimensional Energy-TOF spectra of particles detected in T2 xv

16 List of Figures for coincident particles in T3 for 20 O + 12 C Efficiency of detecting coincident evaporation residues and light charged particles for 20 O + 12 C One-dimensional energy distributions of evaporation residues and light charged particles for 20 O + 12 C Measured fusion cross-sections for charged particle emission channels for 20 O + 12 C compared to evapor predictions Efficiency of detecting coincident evaporation residues and light charged particles for 16 O + 12 C Two-dimensional Energy-TOF spectra of particles detected in T2 for coincident particles in T3 for 16 O + 12 C One-dimensional energy distributions of evaporation residues and light charged particles for 16 O + 12 C Measured fusion cross-sections for charged particle emission channels for 16 O + 12 C compared to evapor predictions Two-dimensional Energy-TOF spectrum of one T2 pie with 20 O Mechanical configuration for first series of α source tests using the 18mm MCP and T Two-dimensional Energy-TOF spectra of T2 using a 226 Ra α source with alternate orientations of the 18mm MCP xvi

17 List of Figures 6.4 Electronics configuration for first series of α source tests using the 18mm MCP and T Two-dimensional Energy-TOF spectra of T2 using a 226 Ra α source with alternate orientations of T Mechanical configuration for the third series of α source tests using the 18mm MCP and T Two-dimensional Energy-TOF spectra of T2 using a 226 Ra α source with two different 18mm MCP detectors Electronics configuration for the fourth series of α source tests using the 18mm MCP and T Two-dimensional Energy-TOF spectra of T2 using a 226 Ra α source with and without a coincidence of the two MCP detectors Two-dimensional Energy-TOF spectrum of T2 using a 226 Ra α source with an MCP coincidence and a cut on the T2 rings xvii

18 List of Tables 2.1 Primary electrons liberated by an incident ion in common ion chamber fill gases. From Ref. [13] Energy loss calculations for 20 O ions at 3 MeV/A in the CID anodes at gas pressures in the range of 20 to 180 mbar CID anode and cathode bias voltages at each pressure used in the experiment ZDIC gas pressures and anode and cathode bias voltages for each beam energy in the experiment Degraded beam energy as determined by the SBD of 20 O for each gas pressure utilized in CID Predictions from evapor of the percentages of evaporation residue distribution for various nuclides for 20 O + 12 C Predictions from evapor of the percentages of evaporation residue distribution for various nuclides for 16 O + 12 C Measurements and predictions of the average energy for evaporation residues and light charged particles for the 16 O + 12 C system xviii

19 Chapter 1 Introduction Neutron stars are interesting astrophysical objects that have been widely studied in recent years, yet in many respects remain poorly understood. A neutron star contains matter under extreme conditions of pressure, temperature, and density. An interesting category of neutron stars is that of accreting neutron stars. An accreting neutron star exists in a binary system and accretes matter from its companion star. An accreting neutron star provides an exotic environment for the formation of neutron-rich isotopes far from beta stability in its outer crust. In order to understand the dynamics and the structure of these highly neutron-rich nuclei, one must determine the nuclear equation-of-state (EOS). The nuclear EOS describes the fundamental properties of nuclear matter as a function of temperature, pressure, mass density, and composition. The accuracy of theoretical models for predicting the properties of rare-isotopes is limited, and recent experiments that have investigated exotic nuclei have led to some unexpected results [1, 2, 3]. Further advances toward understanding the structure and dynamics of these exotic nuclei and ultimately the properties of neutron stars are largely driven by new experiments. 1

20 Chapter 1: Introduction Fig. 1.1: An artist s depiction of an isolated neutron star is shown [8]. A neutron star, depicted in Fig. 1.1, is a stellar remnant composed largely of neutrons that has a typical radius of ~10 km and a total mass on the order of 1 2 M ʘ [4]. Neutron stars are among the densest objects in the known universe as a result of having average mass densities of ~7 x g/cm 3, which is 2-3 times the normal nuclear density of 2.8 x g/cm 3 [5]. The origin of neutron stars was first theorized by Baade and Zwicky in 1933 who predicted the supernova explosion of a massive or super-massive star could lead to the formation of a neutron star [6]. This explanation remains the accepted theory describing neutron star formation. Supernova explosions are presently understood as a shockwave resulting from the gravitational collapse of a star following the depletion of its nuclear fuel. This shockwave, originating in the star core, permeates through the layers 2

21 Chapter 1: Introduction of the star structure and upon reaching the surface explodes away the outer layers of the crust, leaving only the stellar core. A supernova explosion leads to the profuse release of neutrinos and gravitational energy roughly ergs in total [7]. The characteristics of neutron stars are incredibly diverse. Nevertheless, characteristics such as the type and frequency of emitted electromagnetic radiation provide a basis for classification. A first distinction among neutron stars is based upon whether they exist in isolation or as part of a binary system. Isolated neutron stars can be further classified as belonging to one of two groups: radio quiet or radio-loud, according to the amount of observable radiation emitted. Radio-quiet neutron stars are more difficult to discover in part because they emit very little detectable radiation and have been found solely by observation of pure thermal radiation and soft x-rays. Discovery of Radio-loud neutron stars, however, is much more prevalent due to their emission of a considerable amount of electromagnetic radiation which is detectable from Earth. One class of Radio-loud neutron star is the pulsar - a rotating body that releases observable bursts of electromagnetic radiation. Pulsars account for roughly 1700 of the 2000 known neutron stars and have been discovered both in isolation and in binaries [5]. Pulsars are further identified as a radio pulsar, soft gamma repeater, or an anomalous x- ray pulsar. Radio pulsars are rotationally powered neutron stars that emit electromagnetic radiation in the form of radio waves. Soft gamma repeaters (SGRs) and anomalous x-ray pulsars (AXPs) are magnetars, or neutron stars whose electromagnetic radiation is attributed to very large magnetic fields on the order of Gauss [9]. 3

22 Chapter 1: Introduction A neutron star may be found in a binary system with a white dwarf, a second neutron star, or a luminous star, however, it is still considered to be in isolation if material is not accreted from the neighboring object. An accreting neutron star in a binary system is depicted in Fig Neutron stars in binaries are characterized by the emission of electromagnetic radiation in the x-ray region, which is fueled by the accretion process. X- ray binaries are classified according to the mass of the neighboring star. A system consisting of an x-ray binary in which the neighboring mass is greater than 2M ʘ is considered a high mass x-ray binary (HMXB). When the mass of the neighboring star is less than 2M ʘ the system represents a low mass x-ray binary (LMXB). Further classification of binary neutron stars is based upon the rate of recurrence of the x-rays. One category is the x-ray transient, which is marked by x-ray emissions lasting several hours followed by longer periods of quiescence. A second category of x-ray binary is the x-ray pulsar a rotating neutron star that emits periodic x-rays and is typically found in Fig. 1.2: An artist s depiction of an accreting neutron star [10]. 4

23 Chapter 1: Introduction HMXBs. The strong flow of accreted material is responsible for the periodic x-ray emission. Yet another category of x-ray binary is the x-ray burster, which produces distinct bursts of x-rays with considerably less regularity by comparison with pulsars. An x-ray burst is typically in the soft energy range, can last between several seconds and tens of seconds, and is identified as either a Type-I or Type-II burst. Type-I bursts are the most common, have been linked to the thermonuclear ignition of accreted material on the surface, and release approximately ergs of energy [5,11]. An especially interesting case of Type I burst is the x-ray superburst. In this case, a burst lasts from 2 to 12 hours and releases a total energy of ergs [12]. The energy source of superburst production is presently unknown, however the ignition of carbon in the crust has been proposed [13]. Type-II bursts are relatively rare and considerably less is known regarding the source of these bursts. Type-II burst activity can be highly irregular as in the case of the rapid bursters attributed to sudden fluctuations in the accretion rate [14]. The structure of a neutron star remains a somewhat open question, particularly with regard to the composition and equation of state of matter in the inner core. Nonetheless, most theoretical models identify a neutron star as having an atmosphere and four interior regions: the outer and inner crust (or envelopes) and the outer and inner core. A diagram of the proposed structure is shown in Fig The atmosphere is composed of a plasma ranging in depth from a few millimeters for cold neutron stars with T ~ 3 x 10 5 K to ~ 10 cm for hot neutron stars with T ~ 3 x 10 6 K [5]. Extremely cold or ultramagnetized 5

24 Chapter 1: Introduction neutron stars may have a liquid or solid atmosphere rather than a plasma. The behavior of the atmosphere is not fully understood, and current theoretical work has focused on determining the equation of state (EOS) and equilibrium conditions of the plasma. Fig. 1.3: The proposed structure of a neutron star is depicted with a description of the matter contained in each region [15]. Directly beneath the atmosphere is an outer crust with an approximate thickness of several hundred meters and a density that ranges between 10 4 g/cm 3 and g/cm 3. The composition of the outer crust is considered to be an electron gas that becomes a solid lattice of nuclei in the inner crust [4]. The mass density increases at greater depths in the crust and raises the electron Fermi energy, leading to electron capture reactions. 6

25 Chapter 1: Introduction Successive electron captures by the atomic nuclei produce extremely neutron-rich nuclei that eventually reach the neutron drip line at the outer/inner crust boundary. Below the outer crust lies an inner crust having a thickness of ~1 km and a density that ranges between 4 x g/cm 3 and the normal nuclear density, ρ o = 2.8 x g/cm 3 [5]. The inner crust is composed of neutron-rich nuclei immersed in a superfluid of electrons and neutrons [16]. Nuclei near the outer/inner core boundary approach the neutron drip line. Additional neutrons produced by electron capture reactions occupy neutron continuum states, giving rise to a sea of free neutrons. At a density of ~ρ o /3 the spherical nuclei become unstable with respect to fission and the competition between the Coulomb and surface terms of the liquid drop model produces exotic nuclear shapes. Once a density of ~ρ o /2 is reached in the neutron star crust, all nuclei are unstable and cease to exist [4]. At greater depths in the neutron star one finds an outer core composed primarily of neutrons with small quantities of electrons, protons, and possibly muons [5]. The outer core ranges in density from ~0.5 ρ o to ~2ρ o and has an approximate thickness of a few kilometers [5]. The outer core matter is electrically neutral and in beta equilibrium. Located beneath the outer neutron star core lies an inner core with a radius of a few km and a density of roughly 2ρ o. Experiments have led to considerable progress in understanding the properties of nuclear matter at near nuclear densities; however, at densities beyond ρ o the state of the nuclear matter is largely unknown. Theories involving hyper-ionization of matter, pion or kaon condensation, and quark and strange quark 7

26 Chapter 1: Introduction matter have been proposed to describe the composition of the inner core [16]. The composition of the crust of an accreting neutron star in a binary system is determined in part by the accreted material. Material accreted from a luminous companion star onto the surface of a neutron star is a mixture of hydrogen and helium [12]. This material becomes compressed after reaching the surface of the neutron star as accreted material continues to accumulate above it. The compression increases the pressure and temperature of the nuclei on the neutron star surface, resulting in thermonuclear fusion of hydrogen and helium through the 3α reaction, the αp reaction, and the rapid proton capture (rp) process [17]. Continued accretion further raises the temperature of the crust, providing energy and increasing the probability for subsequent thermonuclear reactions. Heavy nuclei resulting from thermonuclear induced fusion reactions in the neutron star crust are buried deeper in the crust as accretion layers continue to form on the neutron star surface. Electron capture reactions modify the N/Z of these heavy nuclei, producing highly neutron-rich nuclei. After several hundred years of accretion the density and pressure of the neutron star crust are raised sufficiently to allow for gravity driven fusion reactions to occur deeper in the crust. These gravity driven fusion reactions termed pycnonuclear reactions result from the overlap of nuclear wave functions at extremely high densities [18]. The processes of pycnonuclear and thermonuclear fusion can be qualitatively understood by examining a simple one-dimensional barrier model. 8

27 Chapter 1: Introduction In a simple one dimensional picture, the interaction between two nuclei is determined by the interplay of the nuclear strong force and the Coulomb force. Collectively the nuclear and Coulomb potentials form an interaction potential that depends on the distance of separation between the two nuclear centers, r, as depicted in Fig In reality when two nuclei collide, the type of reaction that results is determined by the angular momentum as well as the distance of nearest approach (impact parameter) [20]. Fig. 1.4: A one-dimensional, radially dependent interaction potential composed of the nuclear and Coulomb potentials between two nuclei is depicted [19]. A projectile nucleus approaching from the right can enter the potential well if it possesses sufficient kinetic energy to pass over the barrier of height V. Classically if the projectile nucleus approaches the target nucleus from below the barrier at point A, it will be unable to enter the potential well since the region between B and C is the classically forbidden region. However, in a quantum mechanical treatment the projectile has some small, finite probability of penetrating the barrier. Fig. 1.5 shows the classification of nuclear reactions as a function of the impact parameter. At the largest impact parameters, r > R int, nuclear reactions do not occur and the interaction leads to elastic scattering or Coulomb excitation. Slightly smaller impact parameters, where r ~R int, produce inelastic scattering or nucleon exchange reactions, 9

28 Chapter 1: Introduction however, these exchange reactions are typically limited to the transfer of one or two nucleons. For impact parameters where r < R int, strongly damped reactions can result, leading to the transfer of considerable mass and energy. Lastly, deeply penetrating collisions occur at the smallest impact parameters, r <<R int, and lead to the formation of a compound nucleus [20]. Fig. 1.5: The classification of nucleus-nucleus collisions based on the impact parameter is depicted [20]. The impact parameter decreases from top to bottom, and r indicates the distance of closest approach relative to the interaction radius, R Int, that determines the reaction characteristics. Fusion consists of the amalgamation of two nuclei to form a single compound nucleus (CN). From Fig. 1.4, the two ways through which colliding nuclei can fuse are illustrated: 10

29 Chapter 1: Introduction either by passing over the potential barrier or by quantum mechanically tunneling through it if sufficient kinetic energy is not available. The probability for a fusion reaction to occur is expressed through the fusion cross-section, which is written classically as [20]: ( ) (Eq. 1.1) where l is angular momentum, T l are the transmission coefficients, is the reduced de Broglie wavelength, and is the cross-section. Often one assumes a sharp cut-off for T l : { which causes Eq. 1.1 to take the form [20]: ( ) ( ) (Eq. 1.2) If one then assumes l o is associated with the highest partial wave to fuse for a given energy E just above the height of the Coulomb barrier, then Eq. 1.2 becomes the classical expression for the fusion cross-section [20]: ( ) (Eq. 1.3) where R B is the interaction barrier radius, V(R B ) is the barrier potential, and E CM is the projectile energy in the center-of-mass frame. A useful insight into the fusion process is provided by the fusion excitation function - the dependence of the fusion cross-section on incident kinetic energy. As an example, the fusion excitation functions measured by W. J. Jordan et al are shown in Fig. 1.6 for the reactions of 12 C and 16 O on 28, 29, 30 Si [22]. The 12 C cross-sections were measured in the energy range 20 MeV E Lab 49 and the 16 O cross-sections were measured in the energy 11

30 Chapter 1: Introduction range 21 MeV E Lab 61 MeV. In each panel of Fig. 1.6, the measured fusion crosssection (solid points) is provided in addition to the predicted cross-section using both the Bass potential (dashed line) and the Glas Mosel model (solid line). The overall trend for the 12 C and for the 16 O systems is an increase in the fusion cross-section from 28 Si (peaked at ~1000 mb) to 29,30 Si (peaked at ~1200 mb) [22]. For the systems studied, the Glas Mosel model provides a better fit to the data in comparison with the Bass potential. Fig. 1.6: The fusion excitation function for the reactions 12 C + 28, 29, 30 Si are shown in the lefthand column and 16 O + 28,29,30 Si in the right-hand column. The points constitute experimental data, the solid curves are fits according to the Glas-Mosel model, and the dashed curves are calculations using the Bass potential [22]. 12

31 13 Chapter 1: Introduction As shown in Fig. 1.6, excitation functions exhibit a sharp decline in fusion crosssections at some point with decreasing center-of-mass energy. This result is consistent with a barrier driven process as one would expect from the previous discussion of the interaction potential of Fig Excitation functions are generally smooth functions, although light systems such as 12 C + 28 Si shown in Fig. 1.6 can contain oscillations in the cross-section with increasing energies [22]. This oscillatory structure has also been observed for the 12 C + 12 C and 16 O + 12 C systems and could be the result of shell structure effects of the colliding nuclei [22, 23, 24]. Nuclei in an astrophysical environment, such as that of an accreting neutron star crust, will fuse through thermonuclear reactions if both nuclei possess sufficient kinetic energy. As the colliding nuclei exhibit a distribution of energies due to the finite temperate of the system, the probability for thermonuclear fusion depends on both the reaction crosssection and on the Maxwell Boltzmann velocity distribution of the nuclei. The average thermonuclear rate for interacting nuclei can be expressed mathematically as [25]: ( ) ( ) ( ) (Eq. 1.4) where S(E) is the astrophysical S-factor, μ is the reduced mass, k is the Boltzmann constant, and T is the gas temperature. The Maxwell Boltzmann distribution is given by the term e -E/kT, and the term e -2πη represents the probability of tunneling through the Coulomb barrier - the Gamow factor. The astrophysical S-factor is used to describe the properties of the interacting nuclear matter. The tails of the Gamow factor and the Maxwell-Boltzmann velocity distribution

32 Chapter 1: Introduction overlap, producing a maximum reaction probability known as the Gamow peak. The Gamow peak occurs at an energy E o as described by the equation [25]: ( ) (Eq. 1.5) where Z p and Z T are the atomic numbers of the projectile and target nuclei, M p and M T are the masses of the projectile and target nuclei, and T is the temperature of the system in kelvin. Displayed in Fig. 1.7 is the relative probability for fusion between colliding nuclei as a function of center-of-mass energy described by both the Maxwell-Boltzmann velocity distribution and the Gamow factor. The region of overlap between the two functions constitutes the Gamow peak, which occurs at an energy of E o and has been scaled in the illustration for clarity. Fig. 1.7: The fusion probability as a function of the energy of interacting nuclei using a Maxwell-Boltzmann velocity distribution and the Gamow factor. The overlap of the two functions produces the Gamow peak, located at E o, which has been scaled for convenience [26]. 14

33 Chapter 1: Introduction The maximum of the velocity distribution occurs at an energy of E = kt, which clearly falls below E o. The Gamow peak is very important for thermonuclear reactions in astrophysical environments as it is used to determine the energy range in which most thermonuclear reactions will occur. It has been suggested the thermonuclear burning of carbon in the crust of an accreting neutron star is responsible for the production of x-ray superbursts [13]. However, simulations indicate that the temperature is insufficient for the thermonuclear fusion of carbon to occur [27]. Thus, an additional source of heat which serves to trigger the carbon burning process has been sought. It has been postulated that fusion of higher Z nuclei found at greater depths in the crust release energy that could represent a significant source of heat. Naturally, the higher Z nuclei have higher Coulomb barriers which suppress the fusion process. Theory indicates thermonuclear fusion of nuclei with Z 8 will not occur until sufficient densities in the neutron star crust where electron capture reactions have produced very neutron-rich nuclei. For example, the fusion of 24 O + 24 O will occur at densities approaching g/cm 3. Horowtiz et al found the astrophysical S- factor for the fusion of 24 O to be a factor of eight larger than the S factor for the fusion of 16 O. This enhancement could be larger still if the eight valence neutrons are weakly bound to the inner core [27]. Weakly bound valence neutrons in 24 O could form a polarizable neutron-rich skin that would increase the fusion cross-section. Determining whether neutron-rich nuclei manifest an enhancement in the fusion cross-section near and below the Coulomb barrier, and thus constitute a viable heat source in the crust of an 15

34 Chapter 1: Introduction accreting neutron-star, can only be determined by experiment. 16

35 Chapter 1: Introduction Bibliography: Chapter 1 [1] J.F. Liang et al, Nucl. Phys. A 746, 103c (2004). [2] W. Loveland et al, Phys. Rev. C 74, (2006). [3] C. Signorini et al, Nucl. Phys. A 735, 329 (2004). [4] G. Baym. Nucl. Phys. A 590, 233c (1995). [5] P. Haensel, A. Y. Potekhin, D.G. Yakovlev. Astrophysics and Space Science Library. (Springer Science, New York, 2007), Vol [6] W. Baade and F. Zwicky, Phys. Rev. 43, (1934). [7] Supernova and Gamma Ray Bursts, edited by M. Livio, N. Panagia, and K. Sahu. (Cambridge University Press, Cambridge and New York, 2001). [8] Science Daily. Possible Closest Neutron Star to Earth Found. (2007). [9] K. Hurley. Physica E (2011). [10] NASA. [11] W.H.G. Lewin. X-Ray Binaries. (Cambridge University Press, Cambridge and New York, 1995). [12] H. Schatz. Nucl. Phys. A 718, 247c (2009). [13] A. Cumming and J. Bildsten. Astrophys. J. Lett. 559, L127 (2001). [14] Y. Tawara. Astrophysics and Space Science (1986). Vol [15] Neutron star theory group, UNAM. Picture/NStar/NStar_l.gif [16] H. Heiselberg and M. Hjorth-Jensen. Phys. Rep. 328, (2000). [17] H. Schatz et al. Nucl. Phys. A 688, 150c (2001). [18] S. Son. Phys. Lett. A 337, 397 (2005). 17

36 Chapter 1: Introduction [19] J.M. Reid, The Atomic Nucleus. (Penguin Books, Dover and Manchester, 1984). [20] Treatise on Heavy-Ion Science, edited by D.A. Bromley (Plenum Press, New York, 1984). Vol. 2. [21] G. Friedlander. Nuclear and Radiochemistry. (John Wiley and Sons, New York, 1981). [22] W.J. Jordan et al. Phys. Lett. B 87, 1 (1979). [23] P. Sperr et al. Phys. Lett. 36 (1976). [24] P. Sperr et al. Phys. Lett. 37 (1976). [25] J.R. Newton et al. Phys. Rev. C 75, (2007). [26] C. Rolfs and W. Rodney. Cauldrons in the Cosmos: Nuclear Astrophysics. (The University of Chicago Press, Chicago 1988). [27] C.J. Horowitz et al. Phys. Rev. C 77, (2008). 18

37 Chapter 2 Experimental Setup 2.1 Overview of the Experiment The discovery of an enhancement in the fusion cross-section beyond that predicted by standard fusion models would have important implications with respect to the structure and fusion dynamics of neutron-rich nuclei. Furthermore, an enhancement in the fusion cross-section could help to answer fundamental questions concerning the properties of neutron star crusts and the production of x-ray bursts and superbursts as discussed in Chapter 1. In order to determine if neutron-rich, light nuclei exhibit an enhancement in fusion with respect to their beta-stable counterparts an experiment was performed to investigate the fusion of 20 O on 12 C at sub-barrier energies. For a given Z, an increase in neutron number will allow interacting nuclei to fuse at greater distances between the nuclear centers simply due to the increased size of the nuclei. As a result, the astrophysical S factor should systematically increase as a function of increasing neutron number. The presence of additional neutrons could also open additional excitation modes and lead to a further increase in the total fusion cross-section. 19

38 Chapter 2: Experimental Setup A recent experiment by Loveland et al measured the fusion excitation function of a light neutron-rich projectile, 9 Li + 70 Zn, at sub-barrier energies in the range of 9.7 MeV E C.M MeV [1]. A large enhancement in the fusion cross-section was found beyond predictions made by coupled channel calculations [2] as shown in Fig The determination of a large fusion radius was considered evidence for a neutron-rich skin due to the valence neutrons. The observation of such enhancements in the fusion crosssection suggests that systematic measurements of the fusion cross-section for neutronrich nuclei at sub-barrier energies are necessary [3]. Fig. 2.1: The measured fusion excitation function of 9 Li + 70 Zn (points) is shown and compared to coupled-channel calculations [1]. Previous measurements by Eyal et al of the fusion excitation functions of 16 O + 12 C, 17 O + 12 C, and 18 O + 12 C provide a reference for the measurement of 20 O + 12 C [4]. Displayed in Fig. 2.2 are the measured excitation functions of these systems for projectile energies in the range of 7 MeV E C.M. 14 MeV. Also shown are the calculated fusion cross-sections obtained using the Wong formalism [5]. Clearly, a systematic increase is present in the fusion cross-section from 16 O to 18 O. This increase is attributable to an 20

39 Chapter 2: Experimental Setup increase in the astrophysical S factor. Consequently, the fusion cross-section for fusion of 20 O with 12 C is expected to increase beyond that of 18 O. Additional fusion dynamics such as the formation of a neutron-rich neck could lead to an enhancement beyond the systematic trend. Fig. 2.2: Measured fusion excitation functions for the reactions 16, 17, 18 O + 12 C (points) and calculated excitation functions using the Wong formalism (lines) [4]. The experiment was performed at the GANIL-SPIRAL facility located in Caen, France using a radioactive 20 O beam to bombard a thin natural carbon target. A primary beam of 22 Ne accelerated to an energy of 79 MeV/A impinged upon a production target of carbon. The 20 O beam was extracted and re-accelerated by the CIME cyclotron to an 21

40 Chapter 2: Experimental Setup energy of 3 MeV/A before being transported to the experimental area. This energy represents the minimum energy that could be delivered by the beam transport system. The beam intensity at the target was typically 1-2 x 10 4 p/s. In the near Coulomb barrier energy region investigated, compound nuclei produced by fusion reactions de-excite by the emission of neutrons, protons, and alpha particles deflecting the resulting residue from zero degrees. A large fraction of the evaporation residues are expected to occur in the angular region of 3 θ lab 20 based on statistical model calculations performed using the PACE code [6]. The predicted angular and energy distributions of the principal evaporation residues are provided in Figs. 2.3 and 2.4. The average residue energy is expected to increase from ~11 MeV to ~ 22 MeV to ~ 34 MeV as the incident beam energy is increased from E lab = 20 MeV to 40 MeV and finally to 60 MeV. Two annular silicon detectors were used to measure evaporation residues in the angular regions of 11.3 θ lab 22.8 and 3.53 θ lab 10.8, while a zero-degree ionization chamber covered the region of θ lab 3.5. The angular coverage of the three detectors is shown over the total angular distribution of the evaporation residues in Fig. 2.4 at incident beam energies of 20, 40, and 60 MeV. Described in the following sections of this chapter is the vacuum chamber used to house the charged particle detectors; the function, design, and principles of operation of each detector; and the electronics required to process and record the signals produced by these detectors. 22

41 Chapter 2: Experimental Setup Fig. 2.3: The energy distribution of the evaporation residues predicted by PACE for 20 O + 12 C at three incident beam energies. Fig. 2.4: Predicted angular distributions of evaporation residues for the fusion of 20 O + 12 C at beam energies of E Lab = 20, 40, and 60 MeV. Calculations were performed using PACE. 23

42 Chapter 2: Experimental Setup 2.2 Mechanical Setup The detectors used to measure the fusion of 20 O on 12 C were mounted inside a large rectangular stainless steel vacuum chamber measuring 41.2 along the beam line, 14.0 wide, and 11.7 high. The chamber alone is shown in Fig. 2.5 with its side flanges removed while it is shown with the detector assembly inside in Fig As operation of Fig. 2.5: Isometric view of the chamber used in the experiment. The chamber dimensions are given in inches. the microchannel plate detectors used in the experiment required high vacuum, the chamber was initially evacuated using an oil displacement mechanical pump followed by a turbomolecular pump (500 L/s). A Cryoplex-10 cryopump (3000 L/s) was then used to evacuate the chamber to a pressure of 2 x 10-6 torr and maintain the vacuum environment for the duration of the experiment. 24

43 Chapter 2: Experimental Setup Fig 2.6: Above is a depiction of the full mechanical setup for the experiment. The beam enters from the left and passes through the chamber until stopping in ZDIC. The vacuum chamber consists of a rectangular frame with four rectangular flanges. Two of these flanges, constructed from 0.5 thick stainless steel, provide for beam entry and exit from the chamber. The two larger rectangular flanges, located transverse to the beam, can easily be removed. They are constructed from aluminum and allow easy access from both sides to the detectors mounted inside the chamber. An ISO-160 half nipple was welded to the entrance flange for connection to the GANIL beamline while the exit 25

44 Chapter 2: Experimental Setup flange allowed direct mounting of the zero degree ionization chamber. Located on the top of the chamber were three 8 conflat flanges. On the bottom of the chamber three 10 and one 4 conflat flanges were provided. The detection system used in the experiment consisted of an active collimator, two microchannel plate detectors (MCPs), two segmented annular silicon detectors, and two gas ionization chambers. The position of each detector relative to the target is indicated in Fig The first element of the setup inside the chamber was an active collimator with a center hole measuring in diameter. This detector was placed directly at the entrance of the chamber and served to ensure the beam was centered and had no appreciable halo. The active collimator was particularly important during initial tuning of the beam into the chamber. It also provided feedback during the experiment that the location of the beam did not wander as a function of time. The active collimator was constructed from a 1/8 thick plastic scintillator and measured 3 in diameter. It was coupled to a fast photomultiplier tube via a Plexiglas light guide. This entire assembly was mounted on a 3/8 thick aluminum plate, which contained a diameter hole to allow the beam to pass through. This aluminum plate was able to be centered with respect to the beam axis inside the chamber. The second element along the beam path was the active degrader ionization chamber (CID). This detector was a pivotal component of the experiment as it allowed the beam to be degraded for multiple energy measurements making efficient use of the allocated beam time. An equally important function it served was determination of the beam 26

45 Chapter 2: Experimental Setup Fig. 2.7: The detectors are shown mounted inside the fusion chamber at the proper distances for the experiment. The position of each detector is measured relative to the target foil located on the target ladder and is given in inches. identity and any beam contaminants on a particle-by-particle basis. This latter point is crucial in radioactive beam measurements. By adjusting the gas pressure in CID the incident beam energy of E Lab = 60 MeV could be decreased to energies in the range of 20 to 40 MeV. As multiple incident beam energies are necessary for the measurement of the fusion excitation function, use of CID as a gas degrader eliminates the need for the time consuming process of re-tuning the cyclotron associated with each change in beam energy. The choice of a gas degrader is important as solid degraders do not provide the necessary thickness uniformity. The gas utilized in CID was carbon-tetrafluoride, chosen in part because of its high electron drift velocity, large number of primary electrons generated by an incident ion, and stability. CID possesses a window diameter of 0.91 and an active path length of By mounting the detector to an adjustable 6 27

46 Chapter 2: Experimental Setup Huntington linear positioner [7], the detector could be inserted or removed from the beam path. CID was located as close to the target as possible in order to minimize the size of the beam spot on the target due to the beam s divergence. The next detector in the path of the beam was the 18mm microchannel plate detector (MCP). This detector provided the start signal for the time-of-flight measurement of evaporation residues and beam ions. The 18mm MCP detector is a conventional design that utilizes a carbon foil for emission of secondary electrons which are then reflected by an electrostatic mirror onto the microchannel plates [8]. The microchannel plates used are a two-plate stack mounted in a chevron configuration. The microchannel plates, manufactured by Photonis, have an active area of 18 mm [9]. Initially, a 20 µg/cm 2 carbon foil was used in the experiment together with a 100 µg/cm 2 carbon foil located on a target ladder immediately following the MCP. Later in the experiment the 20 µg/cm 2 foil was replaced by a 100 µg/cm 2 foil and the MCP detector was used as an active target. The 18mm MCP was mounted on a 4 Huntington positioner, and thus could be easily inserted or retracted from the beam path. This linear positioner was fixed to the second 8 conflat flange located on the top of the vacuum chamber. Immediately following the 18mm MCP was the target ladder used in the experiment. This ladder had four target positions and was manipulated using a computer controlled stepper motor located outside the vacuum, mounted on the 8 conflat flange. The target ladder contained the following: a passivated ion implanted silicon detector (PIPS Si), a 100 µg/cm 2 natural carbon foil, a 600μg/cm 2 Au target, and an empty target frame to 28

47 Chapter 2: Experimental Setup allow target out measurements. For each new pressure in CID the beam intensity was reduced and the PIPS Si used to determine the absolute energy of the beam. During the second portion of the experiment when the 18mm MCP was used as an active target, the target ladder was fully retracted. Downstream of the target were two annular silicon detectors used to detect evaporation residues following fusion and elastically scattered beam particles. When a target on the target ladder was used the T3 annular silicon detector covered the angular region of 13.0 θ lab 24.8 and the T2 annular silicon detector covered the angular region of 3.87 θ lab Later in the experiment when the 18mm MCP detector was used as an active target, T3 covered the angular region of 11.3 θ lab 22.8 and T2 covered the angular region of 3.53 θ lab These silicon detectors were used to measure both the deposited energy and time of arrival of incident charged particles. Time-of-flight information could then be obtained between the silicon detector and the 18mm MCP detector. Segmentation of the silicon detectors provided position information in two dimensions for ions entering the detector. The thickness of both T3 (305 μm) and T2 (249 μm) was sufficient to stop incident beam particles as well as evaporation residues produced in the experiment. T3 and T2 were rigidly mounted on a 3/8 thick aluminum table located within the vacuum chamber (see Figs. 2.6, 2.7). Four leveling legs on the bottom of the aluminum table allowed vertical alignment with the beam axis. Two additional leveling legs on each side allowed horizontal positioning of the table. Positioned behind the T2 silicon detector were two detectors used to detect fusion 29

48 Chapter 2: Experimental Setup residues emitted in the angular range θ Lab The first of these detectors was a 40mm MCP detector which provided a stop signal allowing measurement of the beam time-of-flight from the 18mm MCP detector to the 40mm MCP. In addition, the coincidence of the two microchannel plate detectors provided a measure of the beam intensity. The 40mm MCP detector is of the same triangular design utilized for the 18mm detector [8]. For reasons unrelated to this experiment, the 40mm microchannel plate stack, manufactured by Photonis, consists of three channel plates in a Z stack orientation with an active area of 40mm. A 0.9 µm thick mylar foil with a 12 nm gold layer deposited onto it was utilized as the emission foil. The detector was rigidly mounted on the aluminum table inside the vacuum chamber. The final detector, a zero degree ionization chamber (ZDIC), measured the energy loss of beam and fusion residues in the angular region of θ lab ZDIC features a segmented anode divided into 3 rows parallel to the beam path and 5 columns perpendicular to the beam path. Segmentation of the anode was included in the design of ZDIC to aid in separating evaporation residues from beam particles. ZDIC has an entrance window with diameter of 4.83 and an active path length of Carbon tetrafluoride was used as the fill gas for ZDIC. The gas pressures used to operate ZDIC were not sufficient to stop all incident beam particles, however, any such particles were stopped in the 3/8 thick aluminum flange of the detector. ZDIC mounts directly to and seals with the chamber exit flange. 30

49 Chapter 2: Experimental Setup Charge sensitive amplifiers (CSAs) were used for amplification of the signals produced in the charged particle detectors and were located directly on the exterior of the chamber. Placement of the pre-amplifiers on the chamber itself was done in order to minimize cable capacitance between the detectors and the CSA. Custom pre-amplifier boxes were designed in order to house the CSAs associated with T3, T2, CID, and ZDIC. 2.3 The Detectors Active Collimator The active collimator served to protect the silicon detectors and the microchannel plates during tuning of the 20 O beam as well as provide feedback during the beam tuning process. For the duration of the experiment the active collimator ensured a significant change did not occur in the beam optics. The active collimator consists of a plastic scintillator mounted to a 3/8 thick aluminum plate and read out by a photomultiplier tube (See Fig. 2.8). Located at the entrance of the fusion chamber (20.4 relative to the target ladder) the detector collimated the beam as it entered the setup. The 1/8 thick fast plastic scintillator was machined into an annulus with an active outer diameter of 3 and given a center hole with a diameter of The diameter of the inner hole was based on the calculations of the GANIL beam physicist who estimated this to be larger than the maximum anticipated beam diameter. Additional considerations governing the choice of the active collimator hole diameter included the foil diameter of the 18mm MCP (0.70 ) and its position downstream. 31

50 Chapter 2: Experimental Setup Fig. 2.8: Isometric view of the active collimator detector. A single layer of aluminized mylar was used to wrap the scintillator, leaving a region of 2mm around the perimeter of the hole opening exposed. The mylar was then wrapped in a black plastic in order to shield the PMT from external light. A small region of the outer diameter was machined flat to allow attachment of a Plexiglas light guide. The light guide was optically coupled to a fast photomultiplier tube. A V792 charge-to-digital converter was used to integrate the charge in the PMT signal. A diameter hole was also located in the center of the aluminum plate to allow a well tuned beam to pass through without striking the plate. Two leveling legs were located on each side of the aluminum plate to align the detector and secure its position once aligned. 32

51 2.3.2 Active Degrader Ionization Chamber (CID) Introduction to Ionization Chambers Chapter 2: Experimental Setup A simple transverse field ionization chamber design consists of a pair of parallel electrodes located within an enclosed gas volume. Ionizing radiation entering the ionization chamber, typically through a thin window material, produces ion electron pairs by collisions with gas molecules. A DC electric potential applied to the electrodes generates an electric field that causes the positive ions and electrons to separate. Movement of the electrons toward the anode followed by their collection generates a measureable signal. If the electric field is sufficiently large, the primary electrons, n o, produced by the ionization are accelerated and produce secondary electrons following Fig. 2.9: The operating regions for gas detectors are shown. The pulse amplitude is given as a function of applied potential across the electrodes [11]. 33

52 Chapter 2: Experimental Setup subsequent collisions with other gas molecules. This effect is the amplification of the signal in the gas. In order for a gas detector to operate in the ionization chamber region, the applied potential must be sufficiently large to prevent re-combination of ion electron pairs yet small enough to avoid gas amplification of the signal. Shown in Fig. 2.9 is gas amplification for different applied voltages at a given pressure of the fill gas. The different operating regions are clearly visible. As one raises the bias potential from 0V the collected signal will increase and approach an asymptotic limit as seen in Fig After determining the voltage at which this plateau occurs, the applied potential V o should be set to the middle of the saturation region [10]. Once ionization occurs, the time required for an electron to reach the anode is determined by its drift velocity in the gas. Drift velocity is dependent upon the reduced electric field (X/p) and can be calculated approximately using the formula: ν = (eλx) / (µmp) where ν is the drift velocity, λ is the mean free electron path, X is the electric field strength, µ is the rms agitation electron velocity, m is the electron mass, and p is the gas pressure [10]. X is typically given in units of V/cm and p in units of torr or atm. In addition to these dependencies, the drift velocity is a function of the chosen gas. Similarly, the drift velocity can be calculated for a positive ion. However, a positive ion generates a signal approximately 1000 times slower than an electron given the typical electron drift velocity approaches 5 x 10 6 cm/s compared with only 1000 cm/s for the ion [10]. This effect is a consequence of the difference in mass between a positive ion and an 34

53 Chapter 2: Experimental Setup electron. The degrader ionization chamber (CID) and the zero degree ionization chamber (ZDIC) were operated using carbon tetrafluoride. The gas, CF 4, was chosen primarily for its high electron drift velocity which continues to increase with applied potential (see Fig. 2.10). Another advantage of CF 4 is the large number of primary electrons produced by an incident ion relative to other common ion chamber gases, which is shown in Table 2.1. Additionally, CF 4 presents minimal working hazards since it is non-toxic, stable, and non-flammable [12]. The electron drift velocities for several common fill gases are depicted in Fig Only pure CH 4 offers a drift velocity comparable to pure CF 4, however, drift velocity in CH 4 saturates at a lower electric field. Fig. 2.10: The electron drift velocity is shown for CF 4 gas mixtures as a function of the reduced electric field [14]. 35

54 Chapter 2: Experimental Setup Table 2.1: The number of primary electrons produced for an incident ion in a number of common ion chamber gases [13]. Fig. 2.11: The electron drift velocities of common ion chamber fill gases are shown. Of the four gases methane offers the largest increase in electron drift velocity as the reduced electric field is increased [10]. An important design feature of the degrader ionization chamber, and the zero degree ionization chamber as well, is the Frisch grid. Once ionization occurs, the electron-ion 36

55 Chapter 2: Experimental Setup pair separate due to the electric field and generate a current due to the motion of the charges. In an ionization chamber without a Frisch grid, the anode signal commences as soon as the electron cation pair start to separate and reaches a maximum when all electrons have reached the anode. Thus, the amplitude of the signal depends on the position of the ionization relative to the anode. This position dependence, which is undesirable, can be removed by locating a Frisch grid at an intermediate position between the two electrodes. The active volume is then bounded by the Frisch grid and the cathode, and ionizing particles are prevented from entering the region between the anode and the Frisch grid. An example of a simple gridded ionization chamber is depicted in Fig Signal Anode Grid V 1 Fig. 2.12: A basic Frisch grid ionization chamber design is shown. Inclusion of a Frisch grid essentially removes pulse height dependence on ionization track within the active gas volume [11]. Cathode V 2 The Frisch grid is held at a potential intermediate between that of the cathode and anode. When ionization occurs, the ion moves toward the cathode and the electron toward the grid. The presence of the grid screens the anode from the motion of the electron in the active volume. When the electron reaches the grid it is further accelerated by the higher electric field between the grid and anode. A signal is only observed once 37

56 Chapter 2: Experimental Setup the electrons have passed the grid. A fine wire mesh with high transparency (> 90%) is used to construct the grid. To prevent the collection of electrons at the grid the electric field in the grid-anode region is chosen to be higher than in the active volume. Description and Design of the Degrader Ionization Chamber (CID) The degrader ionization chamber (CID) was designed and constructed in order to degrade the 20 O radioactive beam and to identify contaminants within the beam. The CID detector has a window diameter of and an active path length of 3.45 parallel to the beam path. This path length is sufficient to degrade the 20 O beam with incident energy of E Lab = 60 MeV to an energy of 20 MeV without exceeding a pressure of 135 torr. The entrance window of the detector and the target located on the target ladder are separated by just 7.1 in order to limit the size of the beam spot on the target due to the beam divergence. The body of CID was machined from a single block of stainless steel using a wire electric discharging machine (see Fig. 2.13a,b). This was done to create a leak tight chamber with a smooth uniform interior surface to prevent sparking during operation. Three flanges attach to the CID body to close the detector volume. These flanges are: an entrance flange, an exit flange, and a flange with the signal feedthroughs. The entrance and exit flanges each contain a window frame with support wires and a gas feedthrough - one for each of the six anodes, the cathode, and a spare. The ionization chamber has exterior dimensions of 4.68 x 3.03 x 3.95 which allow operation of the detector in an ISO 160 standard cross. 38

57 Chapter 2: Experimental Setup Fig a: (Left) The full assembly of CID is depicted with the entrance flange partially removed. The interior components are mounted to the entrance flange for unimpaired access. Fig. 2.13b: (Right) CID is shown fully assembled. The LEMO connectors used for the anode signals and bias voltages and for the cathode bias voltage are located on the top of the detector. Fig 2.14: The CID interior components are shown mounted to the entrance flange. The anodes (top) as well as the cathode are inset in plastic for isolation and to prevent sparking. 39

58 Chapter 2: Experimental Setup A depiction of the CID interior components is presented in Fig A distance of 0.34 separates the anodes and Frisch grid and 1.18 separates the cathode and Frisch. The Frisch grid consists of a nickel mesh grid tack welded taut onto a 3/32 thick stainless steel frame. The grid material contains 70 lines per inch, and is ~ 90% transparent to incident electrons [15]. Each anode was machined from a 3/32 thick copper plate and measures 0.57 along the beam direction and 1.56 in the perpendicular direction. A durable plastic called PEEK (polyetheretherketone) was used to isolate the grid and electrodes and provide mechanical support. Rails machined from PEEK provide alignment of the six anodes and mount them to the rest of the assembly. An important feature of these rails is the precision pockets machined to maintain a separation of 0.01 between each anode (see Fig. 2.15). The CID window assembly consists of three stainless steel frames (see Fig. 2.16a,b). The window was produced by attaching a 1.5 µm doubly aluminized mylar foil to the stainless steel window frame. The two support wire frames each contains a set of 0.15 Fig 2.15: Five of the six CID anodes are shown within the plastic rails. The rails separate the anodes by a distance of to ensure electrical isolation and allow for six independent signals. 40

59 Chapter 2: Experimental Setup Fig 2.16a : (Left) The CID window assembly is shown mounted to the entrance flange. Gold plated tungsten wires were used to support the mylar window foil. Fig. 2.16b: (Right) An isometric view of the CID window assembly is shown. An o-ring groove on the front of the flange provides sealing between the window frame and the flange. spaced diameter gold-plated tungsten wires for support of the mylar foil. One support wire frame is designed to slide inside the inner diameter of the window frame. The second wire frame slides over the outer diameter of the window frame, locating it within the active volume of the detector. Thus, the extent to which the window can bow in either direction is minimized. The interior support frames were not used during the experiment, however. The gold-tungsten wires were tack-welded into place on both sets of support frames. Wires attached using this technique proved to be exceptionally durable as they have survived numerous tests and the experiment without the need for replacements. The apparatus used for tack welding is shown in Fig When assembled, the window frame extends through the flange up to the active volume (a distance of 0.3 ). 41

60 Chapter 2: Experimental Setup Fig. 2.17: The tack welding apparatus used for attachment of the Au-W wires to the CID windows. The copper piece to which the wire frame is mounted serves as one electrode. The second electrode (at left) is only touched to the wire and frame when a weld is desired. The two electrodes were connected to a spark welding machine. Edge effects in the electric field of CID were minimized by including field shaping lines which gradually step down the voltage between the electrodes and the grid. To accomplish this, a single sided copper G-10 printed circuit board panel was machined to remove sections of copper resulting in isolated conducting strips. Two panels of this design were produced and located on the beam left and beam right sides of the CID active volume. Initial tests of CID resulted in the detector sparking at a relatively low voltage. To determine the location of the sparking, the steel entrance flange was replaced with a 1/2 thick Plexiglas flange (see Fig. 2.18) and the chamber operated in a darkened environment. Based upon our observations, two principal changes were made. The first 42

61 Chapter 2: Experimental Setup Fig. 2.18: CID is shown with the steel exit flange replaced by a ½ thick Plexiglas flange for observation of sparking between the electrodes and the chamber walls. change was the addition of a PEEK plastic plate containing a machined pocket to encompass the cathode, leaving only the side of the cathode facing the Frisch exposed. Secondly, thin plastic sheets (Kaptan) were appropriately cut and placed inside the ion chamber walls for isolation. These changes prevented electrical discharges from occurring for the reduced electric field strengths necessary in the experiment. CID has been safely operated at a pressure of 180 mbar and with bias voltages of -900V on the cathode and +300V on the anode without sparking. The signals produced in each of the CID anodes were amplified using charge sensitive amplifiers (CSAs) located outside the vacuum on the 8 Conflat flange to which the 43

62 Chapter 2: Experimental Setup detector is mounted. The energy deposited by an ion as it passes over an anode results in a certain number of primary electrons due to the ionization. This charge is integrated by the CSA and a proportional voltage pulse is produced. Each CSA output signal was processed by a shaping amplifier to produce a Gaussian-shaped voltage pulse. The shaping amplifiers utilized were high gain, four channel NIM modules referred to as quad shapers. The voltage pulse from the quad shaper was digitized by a CAEN V785 peaksensing analog-to-digital converter and recorded by the data acquisition system [17]. Fig. 2.19a: The CID pre-amplifier box is shown mounted to the ISO-160 standard flange. Fig. 2.19b: The CID pre-amplifier box is shown with the side panels removed. Two of the charge sensing pre-amps can be observed. The design of the CID pre-amplifier box was required to satisfy several constraints. It was necessary for the box to be able to mount to an ISO 160 standard flange to allowing testing of the detector within an ISO 160 cross. The box needed to be sufficiently large to 44

63 Chapter 2: Experimental Setup house 7 pre-existing CSAs (one board for each anode and one for the cathode), while leaving sufficient space for the detector gas feedthroughs and the Huntington linear positioner. The final realization of the CID pre-amplifier box is depicted in Fig a,b. Calculations of the Energy Loss in CID A series of calculations were performed using SRIM (The Stopping Range of Ions in Matter) 2008 [18] to determine the energy loss of 20 O ions at an incident energy of 3 MeV/A. Provided in Table 2.2 is the calculated energy loss in each of the six anodes as well as the total energy loss in the detector for each pressure used in the experiment. Included in the total energy loss are the energy losses attributed to the entrance and exit windows. From table 2.2, both the average total energy loss and the energy loss in each anode segment increases with increasing pressure. The calculations also indicate the Pressure (mbar) ΔE A1 (MeV) ΔE A2 (MeV) ΔE A3 (MeV) ΔE A4 (MeV) ΔE A5 (MeV) ΔE A6 (MeV) ΔE Total (MeV) Table 2.2: Calculations of the energy loss in each CID anode and the total energy loss in the detector is given at each pressure of CF 4 used in the experiment. Energy loss in the 1.5 µm mylar windows was taken into account for the total energy loss. 45

64 Pressure (mbar) Cathode Bias (V) Chapter 2: Experimental Setup Anode Bias (V) Table 2.3: The CID anode and cathode bias voltages are given for each pressure used in the detector during the experiment. energy deposited in the detector increases from anode 1 to anode 6 at each given pressure. Listed in Table 2.3 are the anode and cathode bias voltages applied for the pressures used in the experiment mm Microchannel Plate / Active Target Introduction to Microchannel Plate Detectors Microchannel plate timing detectors (MCPs) are commonly used in nuclear physics to provide excellent spatial and timing resolution. Such detectors can generate an overall gain of 10 4 to 10 7 and timing resolution on the order of 100 ps [19]. A typical MCP is constructed from a lead glass disc that contains between 10 4 and 10 7 channels with diameter in the range of 2 to 100µm [19]. Most of the important characteristics of a microchannel plate are determined by the ratio of channel length to diameter. When an electron of sufficient kinetic energy is incident on a microchannel plate, electrons in a channel are ejected. These electrons are accelerated in the channel and collide with the 46

65 Chapter 2: Experimental Setup channel walls, resulting in the ejection of additional electrons. Thus, each channel of the MCP behaves as a continuous dynode. The result is an avalanche of electrons at the exit of the microchannel plate. Microchannel plates can be used singly or as a stack of several plates to achieve a higher overall gain. The gain generated by a two channel plate stack in a chevron configuration is compared with that of a single channel plate in Fig. 2.20a for the same applied voltages. Use of a chevron stack is common and consists of two parallel plates containing channels angled between 5 and 10 relative to the plate surface [20]. The chevron configuration suppresses the phenomenon of ion feedback. Ion feedback is the result of residual gas molecules within the channel of a microchannel plate that can become ionized by electron collisions. Following ionization, the positive ion will be Fig.2.20a: The signal gain is shown as a function of applied voltage per microchannel plate using a Chevron and a straight channel MCP [19]. Fig. 2.20b: The effect of incident radiation on the surface of a Chevron MCP stack is depicted. The avalanche of electrons produced in the channel plates are collected at the metal anode [19]. 47

66 Chapter 2: Experimental Setup accelerated towards the input of the MCP and generate a signal. This effect is almost entirely eliminated by bending the channels as is done by a chevron stack. A common configuration for use of microchannel plates in time-of-flight measurements is the transmission detector depicted in Fig This triangular design uses a negatively biased carbon foil perpendicular to the path of incident particles to generate primary electrons. Parallel wire grids are located on either side of the carbon foil and are maintained at ground potential to provide a uniform electric field. The inner of the two grids thus accelerates ejected electrons in the direction of an electrostatic mirror. The outer grid provides a symmetric arrangement that removes electrons ejected in the opposite direction. The path of an electron ejected from the foil and accelerated towards the mirror is then bent 90 onto the surface of the first MCP by the electrostatic mirror Fig 2.21: The basic triangular microchannel plate detector design is shown. Each of the components are labeled as follows: the carbon foil: s, the wire grids: g, the microchannel plate stack: MCP, the electrode between microchannel plates: e, and the anode : a. The curved arrow represents the path of an electron ejected from the foil by an incident particle [8]. that is angled 45 relative to the carbon foil. The electrostatic mirror consists of two wire grids. The inner grid is grounded while the outer grid is operated at a potential ~10% higher than the voltage of the carbon foil to ensure reflection of all electrons. A chevron 48

67 Chapter 2: Experimental Setup microchannel plate stack is typically used with an applied potential of approximately V per plate to amplify the electrons. The avalanche of electrons exiting the second channel plate is then collected by a flat metal anode plate. This configuration has been demonstrated to produce pulses on the anode up to 1V in amplitude into 50Ω for 5.48 MeV alphas emitted by an 241 Am source [8]. The efficiency of a timing detector of this design was found to be 72.0% +/- 0.7% [8]. Description and Design of the 18mm Microchannel Plate Detector The 18mm microchannel plate detector (MCP) provided a start signal for time-offlight measurements with the T3 and T2 silicon detectors as well as the 40mm MCP. The detector design is the standard triangular MCP timing detector described previously and depicted in Fig The 18mm MCP was oriented such that incident beam ions entered Fig. 2.22a: An isometric view of the 18mm microchannel plate detector is shown. Beam enters the detector through the reflecting grids before passing through the carbon emission foil. Fig. 2.22b: Schematic of the 18mm MCP detector illustrating the typical voltages during the experiment. 49

68 Chapter 2: Experimental Setup through the reflecting wire planes prior to passing through the carbon emission foil. The distance between the carbon foil of the 18mm MCP and the carbon target foil located on the target ladder was A 20 µg/cm 2 carbon foil was used during the first portion of the experiment but was later replaced by a 100 µg/cm 2 foil for use of the detector as an active target. The 18mm microchannel plate detector used in the experiment utilizes a chevron plate stack from Photonis (Model 3018 MA 60:1) with an active area of 0.71 (18mm). The detector was mounted on a Huntington linear positioner with 4 travel located on the second 8 conflat flange on the chamber top. This positioner allowed the MCP detector to be inserted into or retracted from the beam path. A stainless steel frame with an 18mm (0.71 ) diameter hole was used to mount the carbon foil. Parallel grids are located on either side of the carbon foil at the entrance of the detector and contain diameter wires with 0.04 spacing. Each set of entrance grid wires is mounted on 1/8 thick singlesided copper G10 printed circuit board, which contains a 1 diameter opening. Parallel reflecting grids are located 45 relative to the foil and contain diameter wires with 0.04 spacing. The reflecting grid wire planes are each mounted to a 1/16 thick singlesided copper G10 printed circuit board with a rectangular opening measuring 1.1 x 1.6. The entrance wire planes and carbon foil, the reflecting wire planes, and the microchannel plate stack are all mounted to a triangular frame constructed from stainless steel. The MCP anode is a 1/16 thick stainless steel disc located 1/16 behind the 50

69 Chapter 2: Experimental Setup microchannel plate stack. The detector was biased to high voltage using the voltage divider circuit depicted in Fig The printed circuit board with this voltage divider circuit was placed roughly 1 behind the MCP stack inside the vacuum chamber (See Fig. 2.22a). Signals from the Fig. 2.23: The circuit diagram of the voltage divider board used in the 18mm microchannel plate detector is shown. MCP anode were propagated on a short RG-58 BNC coaxial cable to ensure signal integrity. Directly outside the chamber the MCP anode signal was amplified by a factor of 200 using an ORTEC VT120 fast timing amplifier [21]. Following amplification the analog signal was discriminated by a Tennelec TC454 constant fraction discriminator to 51

70 Chapter 2: Experimental Setup generate a logical signal. The logic signal was used both in the generation of the trigger electronics and in the time-to-digital converter as a start signal The T3, T2 Silicon Detectors Introduction to Semiconductor Diode Detectors Silicon semiconductor detectors were chosen for the experiment to detect evaporation residues and separate them from scattered beam particles. This requires measurement of both incident particle energy and time-of-flight with excellent resolution. When an ion with a kinetic energy of a few MeV is incident on a semiconductor detector a large number of electron-hole pairs are produced. The large number of produced electrons is due to the small band gap of semiconductors (Si is 3.6 ev) in comparison to an ionization energy of ~ ev for a gas ionization chamber [11]. This larger number of electronhole pairs produced in a semiconductor detector minimizes the impact of small statistical fluctuations, ultimately leading to an improved energy resolution over other types of detectors. The T2 and T3 silicon detectors are constructed from passivated-ion implanted silicon wafers, Si(IP). Current Si(IP) wafer technology allows production of detectors with active areas of 6 or even 8 [22], which allows a large solid angle to be covered using only a few detectors. Improvements in detector fabrication techniques also allow the construction of the highly segmented Si(IP) detectors used in this work. 52

71 Chapter 2: Experimental Setup Description of the T3 and T2 Silicon Detectors The T3 silicon detector used in the experiment is of the S1 design (see Fig. 2.24), while the T2 silicon detector is of the S2 design (see Fig. 2.25), both manufactured by Micron Semiconductor [23]. The S1 and S2 designs are each a highly segmented passivated ion implanted silicon detector able to provide good angular resolution. The S1 and S2 detectors used in the experiment have nominal thicknesses of 305 µm and 290 µm, sufficient to stop all incident heavy charged particles produced in the experiment. Fig. 2.24a: The ohmic side of the T3 silicon detector (S1 design) which is divided into 22.5 segments to produce 16 pies. Fig. 2.24b: The junction side of the T3 silicon detector is shown. The junction side is segmented into four quadrants of 16 concentric rings. While the S1 and S2 designs are similar, they possess some key differences namely the center hole diameter, the active area, and the segmentation pattern of the junction side of the detector. The S1 design has a center hole diameter of 1.8 (46mm) and an overall active area of 8.2 in 2 (53 cm 2 ), while the S2 design has a center hole diameter of

72 Chapter 2: Experimental Setup (20mm) and an active area of 5.42 in 2 (35 cm 2 ). The junction side of the S1 detector is segmented into 16 concentric rings (Fig. 2.24b) that are partitioned into four quadrants. The junction side of the S2 silicon detector (Fig. 2.25b) has segmentation consisting of 48 concentric rings. The ohmic side of both detectors is segmented into 16 azimuthal pie sectors (Figs. 3.24a, 3.25a), where the capacitance of a segment is approximately 70 pf [22]. Fig. 2.25a: The ohmic side of the T2 silicon detector of the S2 design manufactured by Micron Semiconductor is shown. The ohmic side has 16 pie divisions. Fig. 2.25b: The junction side of the T2 silicon detector is segmented into 48 concentric rings. The T3 and T2 annular silicon detectors were located at distances of 4.1 (10.4 cm) and 6.4 (16.3 cm) downstream relative a target positioned at the target ladder and covered the angular regions of 13.0 θ lab 24.8 and 3.87 θ lab For the case of 54

73 Chapter 2: Experimental Setup the carbon target positioned at the 18mm MCP detector the T3 silicon covered the angular region of 11.3 θ lab 22.8 and the T2 silicon covered the angular region of 3.53 θ lab Each silicon detector provided both a slow energy signal and a fasttiming signal for a particle incident on a given pie sector. The fast timing signals of the silicon detectors were used with the timing signals of the 18mm MCP to determine particle time-of-flight. Determination of both particle energy and time-of-flight is critical for separation of fusion residues from elastically scattered beam in the silicon detectors. The T3 and T2 detectors were mounted on a 3/8 aluminum table inside the vacuum chamber and were located downstream relative the target ladder and upstream relative the 40mm MCP (see Figs. 2.6, 2.7). The T3 signals were propagated from the detector to the 10 conflat feedthrough flange and the T2 signals from the detector to the 8 conflat feedthrough using 50Ω, RG-316 Electroweave cables [24]. It was determined that in order to additionally shield the cables from fast noise pickup, it was necessary to house Fig. 2.26a: Side cutaway view of the T3 silicon detector with the cable conduit and pre amplifier box. Fig. 2.26b: The T2 cable conduit is shown shielding the cables used to pass the signal from the detector to the feedthrough flange. 55

74 Chapter 2: Experimental Setup the cables within a metal conduit. The T3 and T2 detectors are shown with the cable conduit systems in Figs. 2.26a and 2.26b. Description of the T3, T2 Electronics The T3 electronics contains 80 independent channels: one channel for each of the 16 rings in each of the four quadrants (i.e. 64 signals) and one channel for each of the 16 pies. Similarly, the T2 electronics contains 64 independent channels: one channel for each of the 48 rings and one channel for each of the 16 pies. The treatment of the ring signals utilizes a standard approach with each ring channel sent through a charge sensing pre-amplifier (CSA) followed by a conventional shaping amplifier (PICO-Systems) [25]. The amplitude of the resulting Gaussian-shaped pulse is then digitized by a peak-sensing analog-to-digital converter (Caen V785 ADC) and recorded. The treatment of the silicon pie channels, depicted in Fig. 2.27, follows a new approach. The pie signal is split by a passive frequency dependent splitter with a transition frequency of 7.5 MHz. The impedance of the passive splitter is 50Ω to match the RG-316 cables used to transport the signals from the detector to the feed-through flange. The circuit diagram of the frequency splitter is depicted in Fig The low frequency portion of the pie signal is sent through an LC low-pass filter and is used to provide the particle energy information. The low frequency or slow signal was first amplified by a charge-sensing amplifier (CSA) and then by a PICO-Systems shaping amplifier. The charge sensing amplifiers used for both the T3 and T2 rings and slow pie signals provide good energy resolution of 26 kev [26]. The resulting peak of the 56

75 Chapter 2: Experimental Setup Gaussian-shaped pulse was digitized by a V785 ADC. Thus, for each ionizing particle redundant energy information is provided by the ring and the pie channels. Fig. 2.27: A diagram is shown depicting the treatment of the signals from the pie side (ohmic side) of the T2 silicon detector [22]. Fig. 2.28: The circuit diagram of the silicon frequency splitter is shown [22]. The high frequency portion of the pie signal was sent through an LC high-pass filter and amplified by the fast timing amplifier. The circuit diagram for the fast-timing amplifier is depicted in Fig The fast timing amplifier utilizes a pair of monolithic amplifiers in two stages: the MAR 8ASM+ and the GALI 74+ manufactured by Mini 57

76 Chapter 2: Experimental Setup Circuits [27]. Together the two stages of amplification produce a total gain of 170 for the signal input. The amplifier unit pictured is constructed on a two-layer printed circuit board and has overall dimensions of 1.5 x 4.8. An important aspect of the final amplifier unit is the copper shield visible in Fig This grounded shield mounted on the second stage of the amplifier circuit prevents cross-talk between the second stage of the amplifier and the adjacent channel. The fast timing signal from the amplifier was discriminated by a Caen V895 Leading Edge discriminator and sent to a Caen V1290 time-to-digital converter (TDC) as a stop signal. Fig. 2.29: Circuit diagram of the fast timing amplifier [22]. 58

77 Chapter 2: Experimental Setup Fig. 2.30: The fast timing amplifier used to amplify the fast silicon signals in the experiment [22]. Fig. 2.31: The T2 pre-amplifier box is shown with the side panel removed. The box was located outside the vacuum on the third 8 conflat of the vacuum chamber. The CSAs and fast timing amplifiers for T2 were mounted on the 8 conflat flange immediately outside the vacuum and housed within an aluminum box that provided both 59

78 Chapter 2: Experimental Setup mechanical support and electrical shielding (Fig. 2.31). A similar box housing the CSAs and fast amplifiers for T3 was mounted on the second 10 conflat flange on the bottom of the vacuum chamber. Performance of the Silicon Fast Timing Electronics The performance of the fast timing electronics for the T2 annular silicon detector was determined by performing time-of-flight measurements using alpha particles emitted from a radioactive source. The setup consisted of a 226 Ra alpha source, an 18mm MCP, and the T2 silicon used in the experiment arranged in a collinear geometry from the T2 detector. The alpha particles were incident on the 18mm MCP detector (foil first) before striking the T2 detector located 34.5 downstream. Fast signals were discriminated by a Caen V895 leading edge discriminator LED before being sent to a V1290 time-to-digital converter (TDC) as stop signals. The start signal for the TDC was provided by the 18mm MCP signal discriminated by a Tennelec TC 454 constant fraction discriminator. Each of the 16 slow pie channels and each of the 48 ring channels were processed by a PICO Systems shaping amplifier unit and a Caen V785 analog to digital converter (ADC). The TDC and the two ADCs were read out by a high speed VME based data acquisition system. A typical pulse observed from the output of a fast timing amplifier channel is shown in Fig The rise-time of the fast signals is generally between 5 and 6 ns. The overshoot of the fast timing signal effectively limits the count rate that can be observed in the silicon detector. This limitation is not an issue for the type of experiment for which 60

79 Chapter 2: Experimental Setup this system was developed where the radioactive beam intensity does not exceed 1 x 10 6 particles/sec [22]. Evaluation of the data taken from the test setup determined the time resolution of the T2 fast timing electronics (accounting for the previously measured 210 ps of the MCP) to be 365 ps (FWHM) for the MeV α particles [22]. The lower energy alpha peaks of 4.784, 5.489, and MeV show that the time resolution increases by just 70 ps for a decrease of 1 MeV in alpha energy. Fig. 2.32: A representative pulse for a T2 fast timing amplifier channel is shown for the 226 Ra α source [22] The 40mm Microchannel Plate Detector Description and Design of the 40mm MCP A 40mm microchannel plate timing detector was located at a distance of 9.8 downstream from the target ladder to provide time-of-flight measurements to the 18mm MCP detector. The 40mm detector covered the angular region of θ lab 3.52 for a target 61

80 Chapter 2: Experimental Setup positioned at the target ladder. Relocation of the 100 µg/cm 2 carbon target to the 18mm MCP detector made a negligible change to the angular coverage of the 40mm detector. In addition to the time-of-flight, the intensity of the 20 O beam was determined with a coincidence measurement of the two MCP detectors. The 40mm MCP detector design is essentially a large scale version of the 18mm MCP detector that has been previously described. The 40mm microchannel plate detector utilizes a chevron Z-stack of microchannel plates from Photonis that has an active diameter of 1.6 (40mm). The emission foil is 0.9 µm thick mylar with a nm evaporated gold layer deposited onto one side of the foil. The foil was mounted to a stainless steel frame containing a 35 mm diameter hole, and the frame oriented with the gold side of the foil facing the 45 reflecting wire planes. The diameter of the foil frame was made to be 35 mm in order to provide a large angular acceptance for particles transmitted through the detector. Mylar was used rather a traditional carbon foil given the need for a more rigid material to cover the large hole opening in the frame. The field shaping grids located on either side of the foil consist of diameter gold coated tungsten wires with 0.04 (1mm) spacing mounted to singlesided copper, ¼ thick printed circuit board. The reflecting wire planes positioned 45 relative to the foil also contain diameter gold tungsten wires with 0.04 spacing mounted to single-sided copper, 1/16 thick printed circuit board. The full 40mm MCP detector assembly is depicted in Fig

81 Chapter 2: Experimental Setup Fig. 2.33: MasterCAM depiction of the 40mm microchannel plate detector. Incident beam particles pass through the gold-mylar foil and exit through the reflecting grids. The MCP stack was precisely mounted using a Techtron plastic piece. This piece contained a recessed pocket on one side to center the channel plate stack and raised regions on the opposite side to provide stand-offs. A nearly identical Techtron piece is located opposite the first piece and is used to position and mount the stainless steel anode (see Fig. 2.34). The anode Techtron piece contains a recessed pocket for the anode and raised stand-offs on the opposite side. The stand-offs provide the desired spacing between the MCP stack and the anode. In an attempt to build a position sensitive detector, a position sensitive XY-wire grid was located between the channel plate stack and the anode during the experiment. This capability of the detector was not utilized during the experiment. 63

82 Chapter 2: Experimental Setup Fig. 2.34: The 40mm metal anode is shown mounted in the recessed pocket of the Techtron piece. Two machined grooves extend from the anode for the electrical contact tabs used to apply a bias potential. The 40mm anode signal was first amplified by an ORTEC VT120 fast-timing amplifier before being discriminated by a Tennelec TC454 constant fraction discriminator. The resulting logic signal was used as a stop in one channel of the V1290 time-to-digital converter. In addition to producing a stop signal in the TDC, the 40mm logic signal generated by the constant fraction was used in the trigger electronics, which is depicted in Fig The Zero Degree Ionization Chamber Description and Design of the Zero Degree Ionization Chamber A zero degree ionization chamber (ZDIC) was constructed to measure the energy loss of beam particles and fusion residues subtending the angular region of θ lab 3.52 in the fusion experiment of 20 O on 12 C. ZDIC has an entrance window diameter of 4.8 and an active path length of 7.4. The start of the ZDIC active volume was located 20.8 downstream relative a target foil positioned at the target ladder. The ZDIC exit flange 64

83 Chapter 2: Experimental Setup was constructed from 3/8 thick aluminum which is more than sufficient to stop beam particles in the experiment. Carbon tetrafluoride was used as the fill gas of ZDIC, and the pressure of the detector was varied in order to stop evaporation residues within the active volume. Provided in Table 2.4 is the gas pressure, anode bias voltage, and cathode bias voltage of ZDIC at each given pressure in CID. A key feature of ZDIC is the segmented anode designed to separate beam particles from evaporation residues. Other important ZDIC features include the window assembly, a Frisch-grid, electric field-shaping lines, and insulation panels to prevent sparking. CID Pressure (mbar) Beam Energy (MeV) ZDIC Pressure (mbar) ZDIC Cathode Voltage (V) ZDIC Anode Voltage (V) Table 2.4: The corresponding pressures and voltages of the ZDIC detector are provided for each pressure of CID in the experiment. The average beam energy following CID is also provided and was calculated using the Stopping Range of Ions in Matter (SRIM) 2008 [18]. ZDIC was designed to be mounted directly onto the fusion chamber exit flange. This design maximizes the window diameter and overall size of the detector. A CAD depiction of the assembled detector with the entrance flange removed is shown in Fig The body of ZDIC was machined from a single piece of aluminum using electrical discharging machining in order to minimize welds in the detector body. Thus, only an entrance, exit, and connector flange are required to seal the detector. The assembled ZDIC has overall exterior dimensions of 8.75 x 8.8 x All interior components 65

84 Chapter 2: Experimental Setup Fig. 2.35: The full ZDIC detector assembly is shown with the entrance flange removed. are mounted to the ZDIC exit flange for easy removal (see Fig. 2.36). The window assembly (see Fig. 2.37) consists of two stainless steel pieces: one on which the mylar window foil is mounted and another to support a wire support grid. The mylar window mount frame defines the overall active diameter of the detector as 4.8 and has a 3/32 wide lip for increased contact area between the frame and the mylar. To produce a ZDIC window, 2.5 µm doubly aluminized mylar was stretched over a Plexiglas surface before taping the edges of the mylar. A thin layer of five minute epoxy (Hardman Double/Bubble) was then applied to the stainless steel window frame which was subsequently set onto the foil. Pressure was applied to the window as it dried by placing a ~2.5 kg mass on top of the steel frame. The window was given 20 to 30 minutes to dry 66

85 Chapter 2: Experimental Setup Fig. 2.36: The ZDIC interior components are shown mounted to the exit flange of the detector. Fig. 2.37: The ZDIC entrance window assembly. 67

86 Chapter 2: Experimental Setup to dry prior to cutting along the outer diameter of the frame with a soldering iron. Windows produced using this technique resulted in the vacuum chamber pressure increasing from ~1 x 10-6 torr to 2 x 10-6 torr when ZDIC was filled with 100 torr of CF 4. The support wire grid was produced from a thick stainless steel panel which was cut using a wire EDM (Electrical Discharging Machine) to create 0.20 diameter wires spaced by a distance of 1.0 (see Fig. 2.37). The resulting wire grid was then tack-welded into position on the steel frame. ZDIC contains 15 independent anode segments, divided such that 5 anode columns lie along the beam path and 3 rows lie perpendicular to it. We chose this segmentation for this detector in an attempt to separate evaporation residues from beam particles incident on the detector during the experiment. 20 O ions will primarily enter the detector at zero degrees and generate a signal in the center anode column, while the residues will be deflected by secondary emission to either the beam left or beam right column. Additionally, due to the pressure chosen for ZDIC the fusion residues should stop in the first three anode rows. These anodes were constructed from a single 3/32 thick copper plate that was cut using a wire EDM. The three columns were produced by taking two cuts from a single copper sheet. One cut was oriented 1.5 to the left of the beam axis, and the other cut was made 1.5 to the right of the beam axis. The anodes are electrically isolated from each other by their insertion into a plastic plate constructed from PEEK (polyetheretherketone) that maintains a separation of between individual anodes. The anodes mounted to the plastic plate are displayed in Fig

87 Chapter 2: Experimental Setup Fig. 2.38: The 15 ZDIC anode segments are pictured. A separation of is kept between the anodes by a system of dowel pins and screws attaching each anode to the plastic panel. Each of the independent anodes has an associated LEMO [16] connector located on the feedthrough flange (on top of the detector) connecting the detector to a charge sensing amplifier (CSA). A spring-loaded contactor was soldered to each LEMO connector to provide electrical contact between the LEMO connector and the anode. Although somewhat fragile, these contactors greatly simplify the process of disassembling the detector. On the air-side of the feedthrough flange, the fifteen charge sensing amplifiers (CSAs) were plugged directly into the LEMO connectors. These CSAs were enclosed by an aluminum box pictured in Fig The box consists of a top and a bottom plate that possesses machined grooves to locate the CSAs to the correct position and provide mechanical support. These two plates are spaced by aluminum stand-offs and the box is enclosed by a series of thick aluminum panels. A Frisch grid was included in the design of ZDIC to eliminate position dependence of the signal. The Frisch grid mesh was clamped between two stainless steel frames to provide a taut screen (Fig. 2.40). The grid 69

88 Chapter 2: Experimental Setup was assembled by first stretching the mesh to make it taut before placing the first frame above the mesh and the second frame below the mesh and fastening the two halves together. This technique simplified the process of assembling a new Frisch grid. Fig. 2.39: The ZDIC pre-amplifier box containing the charge sensitive amplifiers. The box is located directly outside the vacuum on the ZDIC connector flange. Fig. 2.40: The ZDIC Frisch grid assembly is pictured. A pair of stainless steel frames is used to clamp the mesh and hold its position through a series of screws. 70

89 Chapter 2: Experimental Setup To minimize deformation of the electric field at the edges of the detector, electric field shaping lines have been included in the ZDIC design. Single sided copper-clad G-10 printed circuit board was machined on the copper side to remove strips of copper and leave isolated conducting strips (see Fig. 2.41). The remaining copper strips connected by 10 MΩ resistors form a set of equipotential lines when attached to the anode, cathode, and Frisch grid. Fig. 2.41: A G-10 printed circuit board panel containing the electric field shaping lines of the ZDIC detector. Two such panels are mounted on opposite sides of the ZDIC active volume parallel to the beam path. Initial testing of ZDIC demonstrated that sparking to the chamber walls proved to be an issue at the gas pressures and voltages necessary for the experiment. The precise location of the sparking was determined using a Plexiglas flange for the entrance flange in the same manner as the testing of CID. To prevent sparking a series of changes were made to the detector design. One change required the rounding of the cathode edges to eliminate sharp corners. A second modification was the insertion of a G-10 panel along each side of the ionization chamber wall (except the anode connector side) to provide insulation. In order to insulate the exit flange of the detector, a recess was made in the exit flange and a G-10 panel mounted within the recess using epoxy. It was also 71

90 Chapter 2: Experimental Setup discovered that sparking regularly occurred between the cathode and the KF 40 tube located on the ZDIC exit flange. This issue was resolved by including a G-10 ring that was placed within the inner diameter of the tube. 2.4 Electronic Processing of Analog Detector Signals The analog signals produced in each charged particle detector were amplified, processed by a chain of electronics, and digitized to be recorded by the data acquisition system onto the hard disk of a computer. The electronic chain used to process the analog signals is depicted in Fig along with the trigger electronics in Fig The data acquisition system, a PC/VME based multi-parameter system developed by Michigan State University [28], was responsible for recording the experimental data. The signal processing electronics for the detectors used in the experiment is subsequently described. Each microchannel plate detector produces a fast timing signal for an incident charged particle in the detector. The signal can be used to provide particle time-of-flight information or used in the trigger electronics. Each MCP detector utilized an ORTEC VT120 fast timing pre-amplifier unit to amplify the anode signal by a factor of 200. The anode signal was sent from the detector to the pre-amplifier using a 50Ω RG-316 co-axial cable. The pre-amplifiers were located immediately on the outside of the vacuum chamber to minimize noise pick-up prior to amplification of the signal. The output of the amplifier was sent to a cable patch panel using a 20 ft. long RG-58 co-axial cable. From the patch panel the signals were routed via high quality RG-8 cables to a radiation safe 72

91 Chapter 2: Experimental Setup Fig. 2.42: The chain of electronics for the processing of analog detector signals in the experiment is shown. 73

92 Chapter 2: Experimental Setup area (casemate) resulting in a delay of ~120 ns. The signals were sent to the casemate in order to provide access to the fast timing signals of the MCP detectors. This allowed the zero-crossing of the constant fraction discriminators (CFDs) to be adjusted in addition to the trigger electronics. In the casemate, the analog signal was split using a passive 50Ω splitter. The first branch of the signal was returned to the experimental vault using the patch panel and sent to a Caen V792 charge-to-digital converter (QDC). The second branch was discriminated by a Tennelec TC454 constant fraction discriminator (CFD). For each MCP, the first output of the CFD was returned to the vault and delivered to one channel of both V1290 time-to-digital converters (TDCs). Additional outputs of the CFD were used in generating the trigger electronics that will be discussed later in this section. Each analog signal generated in a silicon detector provided two essential pieces of information: the total charge deposited in the detector and the arrival time of the signal. From this information one can determine the energy of the incident particle and the particle time-of-flight to the 18mm MCP. The T3 silicon detector is segmented into four quadrants of 16 concentric rings on the junction side, while the T2 silicon is segmented into 48 concentric rings on the junction side. The four ring quadrants of T3 were cabled together during the experiment because the number of available electronic channels was limited. A standard approach was used for the processing of each ring segment: the ring signal was amplified by a charge sensitive amplifier (CSA) and sent to a PICO - systems shaping amplifier using a 20 ft., 50Ω co-axial ribbon cable [29]. The output of the shaping amplifier is a smooth Gaussian-like voltage pulse with an amplitude that is 74

93 Chapter 2: Experimental Setup proportional to the amount of charge in the input signal. The output of the shaping amplifier was digitized by a peak-sensing ADC (Caen V785) and recorded by the data acquisition. The ohmic side of the T2 and T3 detectors is segmented into 16 azimuthal pie sectors. Each of the pie segments utilizes a passive frequency-dependent splitter to produce a high frequency (fast) signal used for timing and a low frequency (slow) signal used to extract particle energy information. The slow energy signals generated in each pie were sent through an LC low pass filter and amplified by a CSA. The energy signals were sent to a PICO-systems shaping amplifier using a 20, 50Ω co-axial ribbon cable. The resulting Gaussian-shaped pulses were then digitized by a V785 ADC and recorded by the data acquisition. Pie and ring signals of both detectors used CSAs with gains of 45 mv/ MeV. One T3 ring quadrant utilized CSAs with lower gains (15 mv/ MeV) due to a shortage of high gain CSAs. The shaping amplifier gains were adjusted in order for the full-scale range of the ADC to be ~60 MeV. Following an LC high pass filter the fast-timing signals from each pie segment were amplified by a fast-timing amplifier with a voltage gain of 170 and subsequently discriminated by a Caen V895 leading-edge discriminator (LED). The individual outputs from the LED module provided the stop signals for a V1290 TDC. The OR signal from the LED module was sent to the casemate to be used in the trigger electronics. A metal box was used to house the charge-sensitive amplifiers and fast-timing amplifiers of each silicon detector. The box was located on the chamber flange nearest 75

94 Chapter 2: Experimental Setup the associated detector and served as both mechanical support for the amplifiers and as a shield against noise pick-up. Electroweave 50Ω cables were used to send the signals from the silicon detector to the pre-amplifier box. A system of metal conduit located inside the vacuum chamber was used to electrically shield these Electroweave cables (Fig. 2.26). Signals from the degrader ionization chamber (CID) and the zero degree ionization chamber (ZDIC) were amplified using CSAs located directly outside of the vacuum chamber. The CSAs of CID were contained in a pre-amplifier box mounted on the chamber flange nearest the detector. The CSAs of ZDIC, on the other hand, were contained in a pre-amplifier box mounted directly to the case of the detector. Once amplified by the CSA, each anode signal was processed by a NIM quad shaping amplifier unit. This shaping amplifier has both a slow and a fast output. The slow output is a smooth Gaussian-shaped pulse, which was sent to a V785 ADC. The fast output of the quad shaper for the second left ZDIC anode and the second right ZDIC anode was discriminated by a NIM leading edge discriminator. An OR of these two ZDIC anodes was used in the trigger electronics. To monitor the position of the beam during the initial tuning and throughout the experiment, the active collimator provided a count of the incident beam particles, which could then be compared with the number of incident particles in the MCP detectors. The signal produced in the plastic scintillator was amplified by a photomultiplier tube mounted to the light guide of the scintillator. The PMT signal was sent to the chamber exit flange using a 5, 50Ω RG-58 co-axial cable and from the exit flange to the cable 76

95 Chapter 2: Experimental Setup patch panel using a 20 RG-58 co-axial cable. Once in the casemate, the analog PMT signal was passively split, with one branch returning to the vault to be sent to a V792 QDC. The second output of the splitter was discriminated by a LeCroy 821 LED and utilized in the trigger electronics. The surface barrier silicon detector (SBD) was positioned on the target ladder for calibration of CID. The signal of the SBD was sent from the detector to a CSA located outside the vacuum chamber using an RG-316 co-axial 50Ω cable. Once amplified by the CSA, the signal was sent through a 20, 50Ω RG-58 co-axial cable and passively split into 50Ω. The first branch of the signal was amplified by an ORTEC 572 amplifier and sent to one channel of a V785 ADC. The second branch was amplified by an ORTEC 579 fast filter amplifier (FFA) and discriminated by a LeCroy 821 LED. The resulting logic signal was used in the trigger electronics. The main purpose of the trigger electronics is to produce a master trigger signal. This signal determined which events will be recorded by the computer. The master trigger was also used to create the ADC gates, TDC triggers, and QDC gate. The first portion of the trigger electronics resulting in the generation of the master trigger is depicted in Fig Following generation of the master trigger the electronics used to generate the ADC gates and triggers is depicted in Fig Prior to the master trigger, a series of individual triggers were formed. Trigger 1 consisted of an OR between the T2 fast silicon OR and the T3 fast silicon OR, which had been generated in the V895 LEDs. Trigger 2 formed an AND between the 40mm MCP anode signal and the delayed 18mm 77

96 Chapter 2: Experimental Setup Fig. 2.43: The chain of trigger electronics following generation of the master trigger is depicted. MCP anode signal. The 18mm MCP delay was chosen so that a coincidence resulted only for incident charged particles with time-of-flights longer than that of the beam. Trigger 3 is an AND between the 40mm MCP and a delayed 18mm MCP signal, however, the delay of the 18mm MCP is sufficient to include only only particles with time-of-flights consistent with the direct beam. After forming the coincidence, Trigger 3 was downscaled by a factor of 300 to minimize the amount of data recorded by the computer. Trigger 4 was the logic signal generated from the surface barrier silicon detector (SBD). An OR of Triggers 1 through 4 was taken, and the resulting signal is considered the Submaster trigger. Trigger 4 was only included in the Sub-Master for calibration runs during which the SBD was inserted in the beam path. The final individual trigger, Trigger 5, was formed by taking an OR of the 2 nd left ZDIC anode and the 2 nd right ZDIC anode. In 78

97 Chapter 2: Experimental Setup creating the master trigger, an OR was taken of the Sub-master trigger and Trigger 5. By design, evaporation residues triggered the computer through Trigger 1, Trigger 2, or Trigger 5. The master trigger was used to generate not only the gate/trigger of each VME module but also the computer trigger. Triggering each module at the proper time is critical in order for the module to recognize a new event, and to digitize the information of each channel. Triggering of the computer must also occur at the correct time so the contents of each VME module can be read and recorded by the computer. The ADC gate was formed using one output of the master trigger that was sent to a Philips 794 gate and delay generator (GDG) and given a width of 6 μs. The ADC gate was sent to the vault and used to gate each of the four V785 ADCs. A second output of the master trigger produced the TDC trigger and the computer/ hardware busy. To accomplish this, the master trigger was provided to a LeCroy 222 GDG, given a width of 30 μs, and sent to a logic fan-in/ fanout (FIFA). The logic fan was used to form an OR between the GDG output and the computer busy generated by the VM USB (VME master controller). The first output of the logic fan constitutes the computer/ hardware busy and was used to veto the master trigger. The second output of the logic fan was provided to a LeCroy 821 LED to decrease the width of the signal to 315 ns. The signal was then sent to a LeCroy 222 GDG as the stop, and another output of the master trigger was sent to the GDG as the start. The resulting signal represents the TDC trigger, which was sent to the vault and used to trigger the two V1290 TDCs. The width of the TDC trigger was ~ 10 ns. The 79

98 Chapter 2: Experimental Setup final output of the master trigger was provided to a logic FIFA to create the QDC gate and the computer trigger. An OR of the 18mm MCP, the 40mm MCP, and the Active Collimator was formed. The result of the OR was taken as an AND with the master trigger in order to generate the QDC gate. The QDC gate was given a width of 100 ns, sent to the vault, and used to gate the QDC. The computer trigger was generated by taking the master trigger at the output of the logic fan and the width of the gate was adjusted using a LeCroy 222 GDG. The computer trigger was sent to the VM USB, which read each module located in the VME crate. Information was sent from the VM USB to the computer using a USB cable in order for the information to be recorded onto the computer. 80

99 Chapter 2: Experimental Setup Bibliography: Chapter 2 [1] W. Loveland et al. Phys. Rev. C 74, (2006). [2] K. Hagino, N. Rowley, and A. T. Kruppa, Comput. Phys. Commun. 123, 143 (1999). [3] de Souza, R.T. Sub-barrier fusion cross-sections of neutron-rich oxygen and carbon nuclei. Proposal to the GANIL program advisory committee. May [4] Y. Eyal et al. Phys. Rev. C 13, 4 (1976). [5] C. Y. Wong. Phys. Rev. Lett. 31, 766 (1973). [6] A. Gavron, Phys. Rev. C 21, 1 (1980). [7] Huntington, [8] G. D Erasmo and V. Paticchio, Nucl. Inst. and Meth. A 234, 91 (1984). [9] Photonis, [10] H. W. Fulbright, Nucl. Inst. and Meth. 162, 21 (1979). [11] G. Knoll, Radiation Detection and Measurement (John Wiley and Sons, New York, 1989). [12] Airgas, [13] Fischer, et al. Nucl. Inst. and Meth. A 238, 249 (1985). [14] J. Va Vra et al., Nucl. Inst. and Meth. A 324 (1993). [15] Industrial Netting, [16] LEMO, [17] CAEN, [18] James F. Ziegler et al., Nucl. Inst. and Meth. B 268 (2010). [19] J. Wiza, Nucl. Inst. and Meth (1979). 81

100 Chapter 2: Experimental Setup [20] Experimental Methods in the Physical Sciences. edited by F.B. Dunning, R. Hulet (Academic Press, San Diego and London, 1995), Vol. 29A. [21] ORTEC, [22] R.T. desouza et al., Nucl. Inst. and Meth. A 632 (2011). [23] MicronSemiconductor, [24] Electroweave, [25] PicoSystems, [26] C. Metelko et al., Nucl. Inst. and Meth. A 569 (2006). [27] Mini-Circuits, [28] R. Fox, Ron s DAQ Software, 1918 Pinecrest Dr., East Lansing, MI [29] TycoElectronics, 82

101 Chapter 3 Inspection of Data Online As the aim of the experiment was to measure the fusion excitation function for the 20 O + 12 C system, clearly extraction of the fusion cross-sections is of primary importance. Determination of the fusion cross-section allows a comparison with the known fusion cross-sections of stable oxygen or carbon isotopes as well as with theoretical predictions from a fusion model. These comparisons place the measurement in a broader context and allow one to determine if an enhancement is present in the fusion cross-section for 20 O. Extraction of the fusion cross-section is performed after the experiment in an offline analysis. During the experiment, however, it is also important to examine the raw data from each detector to ensure that each detector is performing adequately and there are no problems with the experiment. In this chapter, online data is presented from all the detectors in the experiment. 3.1 CID Data The fusion cross-section was measured with 20 O incident on the target at eight incident beam energies, 20 MeV E lab 41 MeV. To obtain multiple beam energies the 83

102 Chapter 3: Inspection of Data Online accelerated beam of 20 O at 3 MeV/A was degraded in the degrader ionization chamber (CID), which was operated at pressures between 90 and 180 mbar. An additional series of measurements were performed in the same manner using 16 O, where an incident beam at 3 MeV/A was degraded in CID to produce three incident beam energies in the same energy range as for 20 O. The CID detector not only degraded the energy of the incident beam but also provided a means to separate 20 O ions from beam contaminants by measuring the energy loss of incident particles. Each of the six anodes in CID provided a measure of energy loss as the beam traversed the detector. Following a rough calibration with an alpha source and the 16 O beam, the energy loss measured in the first 5 anodes for the 20 O beam were consistent with the expected energy loss based upon energy loss calculations. The 6 th anode, on the other hand, did not produce a reasonable energy spectrum due to a poor contact to the anode and is consequently omitted in the subsequent analysis. Provided in Fig. 3.1 are the one-dimensional energy loss spectra of the CID anodes for an 20 O beam at 3 MeV/A and a pressure of 89.5 mbar in the detector. The energy spectrum shown represents the energy deposited by an incident ion in a particular anode segment. In each spectrum, the peak of highest intensity appears between channels 400 and 700 and corresponds to 20 O ions. A second, much smaller peak is located immediately to the right of the 20 O. This peak is attributed to the primary contaminant in the radioactive beam, 20 F. The relationship between these two peaks gives a measure of the beam purity. In anode 3 for example, the ratio of the 20 O peak to the 20 F peak is 39:1, constituting a contamination of ~ 2.5% due to 20 F. The 20 O and 20 F peaks are 84

103 Chapter 3: Inspection of Data Online Fig. 3.1: One-dimensional energy loss spectra of the 6 degrader ionization chamber anodes at a pressure of 89.5 mbar and for an incident 20 O beam at E lab = 60MeV. Anode 1 is the furthest upstream of the CID anodes while Anode 6 is the furthest downstream. 85

104 Chapter 3: Inspection of Data Online well separated in anodes 1 through 4, but are less discernible in anode 5 due to a broadening of the two peaks. This broadening is due to the improper biasing of anode 6 due to the poor contact which consequently affects the electric field in the vicinity of the anode 5. Present in each spectrum are two additional large peaks to the right of 20 O and 20 F which are the result of pile-up events in the detector. Pile-up occurs when the time necessary to collect the total charge due to an incident particle exceeds the arrival time between particles. This phenomenon leads to the collection of multiple particle energies in one event, shifting the location of the peak to an energy channel roughly twice that for a single particle event. Displayed in Fig. 3.2 are energy loss spectra of the third CID anode at each pressure used in the detector for the 20 O beam. Clearly the 20 O peak remains the dominant peak in each spectrum and becomes broader as the pressure in the detector increases. This trend can be described quantitatively by the full-width at half maximum (FWHM) of the peak at several pressures: 43.3 channels at 90 mbar, 55.6 at 131 mbar, and 65.5 at 180 mbar. A broadening of the 20 O peak with increasing pressure is anticipated because of the increasing multiple scattering in the gas as a function of pressure. In contrast to the visibility of the 20 O peak, the 20 F peak is only somewhat noticeable at gas pressures of 100, 109, 120, and 130 mbar but is clearly evident at 90, 140, 160, and 180 mbar. In support of this observation is the ratio between the 20 O and 20 F peaks which is considerably larger in the first set of pressures (i.e., 113:1 at 130 mbar) compared to the second set of pressures (i.e., 71:1 at 140 mbar). A change in the beam tune over the 86

105 Chapter 3: Inspection of Data Online Fig. 3.2: One dimensional energy spectra of the 3 rd CID anode at each pressure used to operate the detector in the experiment. 87

106 Chapter 3: Inspection of Data Online Fig. 3.3: One dimensional energy loss spectra of the degrader ion chamber anodes for 16 O at 3 MeV/A and 90 mbar in the detector. 88

107 Chapter 3: Inspection of Data Online course of the experiment is the most likely source of this difference, as the first set of pressures were utilized earlier in the experiment than the second. The pile-up peaks also remain prominent in each spectrum and increase in width as a function of increasing pressure (FWHM 1 = 84.3, FWHM 2 = 66.2 at 90 mbar; FWHM 1 = 185, FWHM 2 = 104 at 180 mbar). For comparison to the 20 O, the energy loss spectra of all CID anodes are shown in Fig. 3.3 for an 16 O beam at 3 MeV/A and 90 mbar of CF 4 in the detector. The spectra of anodes 1 5 contain a strong 16 O peak similar to the 20 O peak exhibited in the spectra of Fig Pile-up for 16 O appears as a single, broad peak in anodes 1 and 5 but becomes two separate peaks in anodes 2 4. The existence of two pile-up peaks is in agreement with the 20 O spectra of anodes 1 4 and indicates that neither peak is due to 20 F ions. The 20 F peak located to the immediate right of the 20 O peak in Fig. 3.1 is clearly absent from the 16 O spectra. This result is significant because it confirms the origin of the peak to be a contaminant in the radioactive beam and not an effect produced in the detector. 3.2 SBD Data Following the reduction of the beam energy with CID, the surface barrier silicon detector (SBD) was used to measure the incident beam energy. After each change of the pressure in CID, the beam intensity was reduced and the SBD, which was mounted on the target ladder, was inserted into the beam path. Presented in Fig. 3.4 are one-dimensional energy spectra of the SBD for two different pressures of the CID detector. The top SBD spectrum corresponds to the case in which CID was removed from the beam path and 20 O 89

108 Chapter 3: Inspection of Data Online Fig. 3.4: One dimensional energy spectra of the surface barrier silicon detector for 20 O with the degrader ion chamber removed, inserted at 89.5 mbar, and inserted at mbar 90

109 Chapter 3: Inspection of Data Online Fig. 3.5: Two dimensional spectra of the energy loss in the 4 th anode of the degrader ion chamber (CID) vs. the energy deposited in the surface barrier silicon detector. The top panel is the result of 20 O at 3 MeV/A and 90 mbar in the CID, and the bottom panel is the result of 16 O at 3 MeV/A and 90 mbar in the CID. 91

110 Chapter 3: Inspection of Data Online was directly measured in the SBD. The lower two panels depict the effect of inserting CID into the beam path with pressures of 89.5 mbar and mbar respectively. Each spectrum exhibits a sharp peak that represents the full energy deposited in the SBD by the incident beam particles. The width of the peak increases from the un-degraded beam (FWHM = kev) to the beam degraded by 180 mbar of CF 4 (FWHM = 1027 kev). This broadening of the beam energy is expected due to increased multiple scattering as the gas pressure increases. In addition to the dominant peak, a small secondary peak corresponding to 20 F is present in each case of degraded beam but absent from the undegraded beam spectrum. This can be understood by the difference in energy loss of the 20 O and 20 F nuclei in the CID. In the case of direct beam, the two nuclei possess the same energy and consequently become located in the same peak of the SBD spectrum. Once degraded in the CID, however, the two nuclei are transmitted with different energies, and appear as separate peaks. To examine the correlation between the energy deposited in the CID and the SBD, the energy loss of the fourth CID anode (ΔE) is plotted against the energy of the SBD detector (E) in Fig A ΔE vs. E plot is provided both for an 20 O beam and for an 16 O beam each with a pressure of 90 mbar in the CID detector. Two peaks are clearly exhibited in the 20 O spectrum, the most dominant of which corresponds to 20 O ions. The second peak in the spectrum is located higher in ΔE CID but lower in E SBD relative to the 20 O, which is consistent with its previous assignment as 20 F. The relationship between the 20 O and 20 F peaks in the spectrum demonstrates the anti-correlation between the energy 92

111 Chapter 3: Inspection of Data Online of the two nuclei deposited in each detector. In contrast, only one peak is present in the delta E vs. E spectrum of 16 O corresponding to the incident 16 O ions. This result is expected as there are no contaminants in the 16 O beam. In addition to these features, both spectra exhibit a tail leading from the oxygen peak. While particles located in this tail exhibit less deposited energy in the 4 th CID anode in comparison to particles located in the oxygen peak, they exhibit the same deposited energy in the SBD detector mm and 40mm MCP Data The 18mm microchannel plate detector (MCP) was utilized as an active target to provide a start signal for time-of-flight measurements, while the 40mm MCP detector provided a stop signal for incident particles in the angular region of θ lab 3.5. Thus, a measure of the beam intensity and beam time-of-flight were provided by the two detectors. A one dimensional time-of-flight spectrum between the 18mm and 40mm MCPs is shown in Fig. 3.6 for 20 O at E lab = 41.0 MeV. For convenience the time-of-flight is already provided in nanoseconds since each channel of the time-to-digital converter corresponds to 25 ps. The sharp peak located at 0 ns in the spectrum is the result of direct beam particles, and has a width of 447 ps (FWHM). To the right of the beam peak and starting at a time-of-flight of ~ 2ns is a continuous distribution of counts that decreases logarithmically extending up to a time-of-flight of ~ 100 ns. The counts located in this tail can be evaporation residues, elastically scattered beam particles, or beam particles that have been slit scattered from one of the detectors in the beam path. A broad peak is also present to the left of the dominant peak. This broad peak corresponds to random 93

112 MCP coincidences due to thermal electrons present in each MCP. Chapter 3: Inspection of Data Online Fig. 3.6: MCP18 to MCP40 time-of-flight spectrum at an incident beam energy of E lab = 41 MeV. 3.4 T3 and T2 Data The T3 and T2 annular silicon detectors measured incident charged particles in the angular regions of 11.3 θ lab 22.8 and 3.53 θ lab Each detector is doublesided and capable of producing an energy signal from the segments of the ohmic side (pies) as well as the junction side (rings). However, only the pie segments were used to provide fast-timing signals for incident particles. The performances of all pie and ring segments in both detectors during the experiment were consistent with alpha source and stable beam tests with the exception of T2 rings 31 33, which exhibited large non- 94

113 Chapter 3: Inspection of Data Online Fig. 3.7: One-dimensional energy spectra for charged particles incident on each of the T3 pies at a beam energy of E lab = 25 MeV. 95

114 Chapter 3: Inspection of Data Online Fig. 3.8: One-dimensional energy spectra of the T2 pie segments at an incident beam energy of E lab = 25 MeV. 96

115 Chapter 3: Inspection of Data Online linearities in gain as well as noise pick-up from adjacent rings. As an example of the energy spectra measured by the pie segments, we display in Figs. 3.7 and 3.8 the onedimensional energy spectra for an incident 20 O beam at E lab = 25 MeV. A peak corresponding to beam particles elastically scattered in the carbon target is located on the far right of each spectrum. The elastic peak is considerably more pronounced in T3 pies 8 13 than in the remaining T3 pies. This behavior is mirrored by the T2 detector as the elastic peak is clearly less intense in pies 6 11 than in pies 0 5 and The most likely reason for this asymmetry is an off-centeredness of the beam. In each spectrum, counts to the left of the elastic peak represent particles with a lower energy than the beam. Such lower energy particles could either be lower energy reaction products such as fusion residues or beam particles that have been slit scattered prior to entering the T2 or T3 detector. The most likely source of slit scattering is the acceleration and reflector wire grids of the 18mm MCP detector. In addition to these features held in common by the T3 and T2 pies, a number of important differences are also observed. One difference is the width of the elastic peak, which is appears broader in the T3 pies (FWHM 75 channels) than the T2 pies (FWHM 59 channels). A second difference is the sharp peak(s) located between channels 250 and 350 in each T2 pie spectrum that is not present in the T3 pies. Subsequent to the experiment it was determined that this peak is the result of incomplete charge collection in the T2 pie segments, and clearly occurs with less regularity in the T3 pies. Still another observable difference between the detectors are the two additional peaks in T2 pies 0 and 15 located at channels ~300 and ~350 that are not observed in any 97

116 Chapter 3: Inspection of Data Online of the remaining T2 pies or the T3 pies. The origin of these two peaks is unclear, but based upon prior experience with similar detectors could result from charge sharing between the two adjacent pie segments due to insufficient isolation by the SiO 2 on the wafer [1]. The one dimensional energy spectra of the T3 rings and the inner 16 rings of T2 are shown in Figs. 3.9 and 3.10 for an 20 O beam at E lab = 25 MeV. The T2 rings shown are representative of the remaining 32 ring segments aside from rings The T3 and T2 rings spectra exhibit a number of features in common with the pies spectra, namely a prominent elastic peak and a distribution at lower energies. A characteristic that is unique to the T3 rings, however, is the presence of additional sharp peaks at low energy channels. These peaks, which are particularly evident in rings 0-5, were determined to be due to multiple pedestals. The occurrence of more than one pedestal resulted from each of the four quadrants of 16 rings in T3 being shorted together by an external cable in order to reduce the number of electronic channels required. As this short was performed after the charge sensitive amplifier, slightly different gains for each charge-sensitive amplifier resulted in the multiple pedestals. Located between channels 800 and 1100 in the T2 rings spectra are several peaks that are not visible in the T3 rings. Energy vs. time-of-flight (ETOF) spectra for the T3 and T2 detectors are extremely important as both time and energy are utilized to separate evaporation residues from scattered beam particles. Displayed in Fig are the representative ETOF spectra of T3 pie 9 and T2 pie 1 for 20 O at an energy of E lab = 41.0 MeV. The energy shown is the 98

117 Chapter 3: Inspection of Data Online Fig. 3.9: One-dimensional energy spectra for particles incident on each of the 16 T3 ring segments at a beam energy of E lab = 25 MeV. 99

118 Chapter 3: Inspection of Data Online Fig. 3.10: One-dimensional energy spectra of the T2 ring segments at an incident beam energy of E lab = 25 MeV. 100

119 Chapter 3: Inspection of Data Online Fig. 3.11: Two dimensional energy versus time-of-flight spectra of T2 pie 1 and T3 pie 9 before the energy calibration. The spectra were generated for an incident 20 O beam at E lab = 41.0 MeV. 101

120 Chapter 3: Inspection of Data Online energy deposited in the given pie segment, and the time-of-flight is the difference between the start signal of the 18mm MCP and the stop signal produced by the pie displayed. The first distinct feature in both spectra is the peak located at channel ~2500 in T2 pie 1 and ~2700 in T3 pie 9 which is due to elastically scattered beam particles. Also prominent is a dark line extending from the elastic peak that increases in TOF as the energy decreases. This line extends to channel ~200 in both detectors. Counts that lie along this locus correspond to particles degraded in energy prior to impinging on the silicon detector. The most likely cause of this slit scattering as previously mentioned is the 18mm MCP wire planes. A significant difference between the two detectors is the presence in T2 of a dark near vertical band starting at the elastic peak and extending to an energy of roughly 400. Particles located on this vertical band possess a time-of-flight comparable to elastically scattered beam particles but represent cases in which the pie segment failed to collect the full energy deposited in the detector. The vertical band does not appear in the T3 ETOF, indicating that incomplete charge collection occurred far less frequently in the T3 detector. Located below the slit scatter line of both detectors is a haze resulting from slit scattered beam particles where only a fraction of the deposited energy is collected by the pie segment. Incomplete charge collection and charge splitting between adjacent pies can both lead to an incomplete measurement of the particle s energy. A locus of counts is also found on the longer time side of the slit-scatter line in the region where the evaporation residues are expected to occur. During the experiment we determined that the cross-section associated with this locus corresponds to 102

121 Chapter 3: Inspection of Data Online Fig. 3.12: Two dimensional energy vs. time-of-flight spectra of T3 pie 9 and T2 pie 1 for 16 O at E lab = 42.1 MeV. 103

122 Chapter 3: Inspection of Data Online approximately b. This large cross-section well in excess of the total nuclear cross-section of ~1 barn at this energy points to an atomic process. This locus therefore represents a significant problem for the experiment, as it prevents the direct determination of the fusion cross-section. A series of alpha tests demonstrated the origin of these counts to be slit scattered particles from the 18mm MCP wire planes that eject an electron and produce a false early time signal in the MCP detector (see Chapter 6). In order to make a comparison between the performance of the silicon detectors with 20 O and 16 O, the ETOF spectra of T3 pie 9 and T2 pie 1 are provided in Fig for an 16 O beam at E lab = 42.1 MeV. The 16 O ETOF spectra exhibit the same principle features as the 20 O spectra: a strong elastic peak, slit scatter line, and a haze of counts both below as well as to the right of the slit scatter line. Also consistent between the 16 O and 20 O spectra is the intense vertical band present in T2 pie 1. An observable difference between the spectra of the two beams, however, is the width of the elastic peak which is clearly broader in the 20 O ETOF. This difference is the result of a higher pressure in the CID detector for 20 O (90 mbar) compared with 16 O (20 mbar) to obtain a similar incident beam energy. That the T3 and T2 ETOF spectra are largely the same for 20 O and 16 O indicates the origin of the unexpected spectral features is not the result of the radioactive beam itself but is a detector artifact. 3.5 ZDIC Data The zero-degree ionization chamber (ZDIC) was used to detect beam particles and evaporation residues in the angular region θ lab 3.5. The pressures used to operate ZDIC 104

123 Chapter 3: Inspection of Data Online were sufficient to stop evaporation residues in the active volume but insufficient to stop direct beam particles. As described in detail in Chapter 2, ZDIC contains 15 independent anode segments divided into 5 segments along the beam path with 3 columns transverse to the beam path. The one-dimensional energy spectra of the ZDIC anodes are shown in Fig for an incident 20 O beam at E lab = 41 MeV and a pressure of 40 mbar in the detector. The energy shown in each spectrum is a measure of the energy deposited by incident particles in the anode segment displayed. From Fig. 3.13, the spectra of the center anode column are markedly different from those of the left and right anode columns. The spectra of these center anodes are similar to those observed with CID. Clearly evident are peaks attributable to 20 O, 20 F, and pile-up events. Incident 20 O ions produce the peak of greatest intensity in each of the center anodes (C1 C5) spectra. The 20 F peak is not visible in the C1 spectrum but begins to appear as a shoulder to the right of the 20 O peak in C2. The 20 O and 20 F peaks become increasingly separated from anodes C2 - C4 until becoming fully resolved in anode C5. Located to the right of the 20 O and 20 F peaks is an intense peak occurring from pile-up events in the detector. In contrast to the CID spectra, only one pile-up peak is observed in each of the center ZDIC anodes. A ratio of 3:1 between the 20 O and pile-up peak in the center anodes indicates a greater occurrence of pile-up in the ZDIC by comparison to the CID, which exhibits a ratio of 5:1. This larger pile-up in ZDIC is not unexpected due to the larger size of the detector. The left and right ZDIC anodes do not exhibit any of the strong peaks characterized in the center anodes. These spectra exhibit intensity at small energies which decreases with 105

124 Chapter 3: Inspection of Data Online Fig. 3.13: One-dimensional energy spectra for each of the zero degree ion chamber anodes at a beam energy of E lab = 41 MeV. 106

125 Chapter 3: Inspection of Data Online Fig. 3.14: One-dimensional energy loss spectra of the 2 nd left, center, and right zero-degree ion chamber anodes at each pressure used to operate the detector during the experiment. 107

126 Chapter 3: Inspection of Data Online increasing energy followed by a peak at larger energies. The performances of each left and right anode pair appear consistent with one another; however, the average energy deposited in the left anodes is greater than in the right anodes suggesting a beam asymmetry. The energy spectra of the second left, center, and right ZDIC anodes are displayed in Fig for each pressure used in the ZDIC during the experiment. The second anode in each column exhibits an increase in deposited energy as a function of increasing gas pressure. This result is expected given most of the incident beam particles are not stopped within the active volume hence the measured spectrum for the beam is an energy loss spectrum and the energy los increases with increasing pressure. Distinct 20 O and pile-up peaks are observed in the C2 anode spectrum at each pressure. However, the 20 F peak is not visible in anode C2 at the first 3 pressures, but begins to appear to the right of the 20 O peak at 40 mbar. The L2 and R2 anodes spectra do not exhibit any narrow, dominant peaks only a single broad peak at large values of the energy. The behavior of the L2 and R2 anodes is similar at each pressure between 27 and 40 mbar, however, at 20 mbar the overall shape of the R2 spectrum differs from that of the L2 anode. 108

127 Chapter 3: Inspection of Data Online Bibliography: Chapter 3 [1] B. Davin, Ph. D. thesis, Indiana University,

128 Chapter 4 Calibration Calibration of the detectors used in the experiment is necessary in order to extract the fusion cross-section of 20 O + 12 C in the energy region of 19.8 MeV E lab 41.0 MeV. The detectors calibrated were the SBD, the microchannel plates, and the annular silicon detectors. Despite its important role in the experiment, it was not necessary to calibrate the degrader ionization chamber (CID). Both of the tasks CID was used for: degrade the energy of the incident beam and identify contaminants in the beam based on energy loss could be accomplished without an absolute energy calibration being performed. The energy calibration of the SBD detector, on the other hand, was critical because the SBD was used to determine the degraded beam energy at each different pressure utilized in the CID. Performing energy calibrations of the annular silicon detectors was also important because the two detectors T2 and T3 combined to cover the angular region of 3.5 θ lab 22.8, in which approximately 70 percent of the fusion cross-section is located. Furthermore, the annular silicon detectors provide the largest amount of information for incident particles: the arrival time, deposited energy, and angle of incidence. Of the 110

129 Chapter 4: Calibration remaining fraction of the fusion cross-section, 20 percent was located in the angular region covered by the zero-degree ionization chamber (ZDIC), θ lab 3.5. Although this cross-section is not negligible, due to time constraints calibration of the ZDIC was not undertaken as part of this work. It is also necessary to calibrate the microchannel plate detectors and the fast-timing of the annular silicon detectors in order to measure the timeof-flight of reaction products. This chapter focuses on the methods used in the energy calibration of the SBD and the T2 and T3 silicon detectors. The thicknesses of the T3 silicon (305 μm), the T2 silicon (249 μm), and the SBD (80 μm) were sufficient to stop all incident charged particles in the energy range associated with the experiment. Therefore, the energy deposited in these detectors largely measures the total energy of the incident particle excluding energy loss in the target or entrance window of the detector. The SBD was calibrated using a 226 Ra alpha source and incident 16 O and 20 O beams at E lab = 3 MeV/A. The alpha energies of 226 Ra are well known, and the incident 16 O and 20 O beam energies were measured and provided by the GANIL accelerator staff. 4.1 SBD Calibration The raw one dimensional energy spectrum of the SBD is shown in Fig. 4.1 for the alpha source and in Fig. 3.3 for the 20 O beam. Four prominent peaks are observed in the alpha spectrum, corresponding to alpha energies of 4.784, 5.489, 6.002, and MeV [1]. Two minor peaks are also present and are the result of alpha particles with energies of and MeV. However, these two peaks were not used as calibration points 111

130 Chapter 4: Calibration Fig. 4.1: One dimensional energy spectrum of the surface barrier silicon detector for a 226 Ra alpha source. The four dominant peaks correspond to energies of 4.784, 5.489, 6.002, and MeV, and the two smaller peaks correspond to energies of and MeV. given their low relative intensities and close proximity to the MeV and MeV peaks. To determine the position in channel number of each peak in the alpha spectrum and the elastic peak in the 16 O and 20 O spectra a Gaussian fit was applied and the centroid position extracted. The known energy of each peak in MeV was then plotted as a function of its position in channel number, and a 1 st order polynomial fit was applied. The polynomial fit and the six calibration points are shown in Fig. 4.2 as well as the chisquare and the 0 th and 1 st order terms describing the function. The polynomial fit constitutes the SBD calibration and describes all calibration points to within 0.5% as seen 112

131 Chapter 4: Calibration in Fig As an example, a calibrated one dimensional energy spectrum of the SBD at a CID pressure of 89.5 mbar is provided in Fig Fig. 4.2: The 1 st order polynomial fit to the 6 calibration points of the surface barrier silicon detector is shown. The dominant peak in the spectrum is associated with 20 O and is centered at 41.0 MeV, while the smaller peak is attributed to 20 F and is centered at 37.1 MeV. Following the calibration of the SBD, the beam energy was determined for each pressure utilized in the CID during the experiment. The resulting incident energies are provided in Table 4.1. The full-width at half maximum of the 20 O peak is also listed in Table 4.1, and clearly increases with increasing pressure in the CID. This result is expected, due to the increasing importance of multiple scattering as the CID pressure is increased. More 113

132 Chapter 4: Calibration multiple scattering leads to a broader distribution of trajectories through CID resulting in a broader energy distribution. Fig. 4.3: The deviation of the measured energy points from the calibration fit for the SBD. P CID (mbar) Beam Energy (MeV) FWHM (kev) Table 4.1: The degraded beam energy as determined by the surface barrier silicon at each pressure utilized in the degrader ion chamber. The full-width at half maximum (FWHM) of the 20 O peak is also provided. 114

133 Chapter 4: Calibration Fig. 4.4: Calibrated one dimensional energy spectrum of the surface barrier silicon detector for an 20 O beam degraded by a pressure of 89.5 mbar in the degrader ion chamber. 4.2 T2 and T3 Calibration The T3 and T2 silicon detectors are both double-sided detectors. Migration of the electrons produced to the pie segments and holes to the ring segments results in two dimensional position information. The pies were chosen for the energy measurement in part because the fast timing signals were also generated by the pies. In addition, use of the ring energy is less ideal than the pie energy given the higher segmentation of the rings, which can lead to a greater occurrence of charge splitting between segments. Thus, the decision was made to calibrate the pie segments first and to use the pie energy whenever possible. Nonetheless, calibration of the ring segments is also important in this experiment as a means to verify the energy collected by the pies. This cross-check of the 115

134 Chapter 4: Calibration energy as determined by both the ring and pie is particularly important for the T2 detector given the large number of events exhibiting incomplete charge collection by the pies. The T3 and T2 pies were calibrated using a 226 Ra alpha source and the elastic peak of 16 O and 20 O beams at several incident energies. To calibrate the T3 pies, an 16 O beam at energies of E lab = 23.8 and 28.4 MeV and an 20 O beam with an energy of 25.0 MeV were used. An additional 20 O beam energy of 41.0 MeV was utilized for T3 pies 8 13, but could not be included in the calibration of the remaining pies due to limited statistics. This discrepancy in the statistics between the two groups of pies is most likely the result of an off-centeredness of the beam as previously noted. Calibration of the T2 pies made use of the following beam energies: 16 O at E lab = 23.8, 28.4, 42.6 MeV and 20 O at 19.8, 25.0, and 41.0 MeV. In determining each of the beam energies, the energy loss code SRIM [2] was used to account for the energy loss of incident beam particles in the CID windows, the given gas pressure in the CID, and the carbon target. The elastic peaks are clearly evident in the raw one-dimensional energy spectra of the T3 and T2 pies, examples of which are provided in Figs. 3.5 and 3.6 for 20 O at 25.0 MeV. Raw alpha spectra for T3 pie 9 and T2 pie 1 are shown in Fig. 4.5 and are representative of the remaining T3 and T2 pies. As described previously for the SBD, the alpha spectra contain 4 principal peaks used in the calibration and 2 minor peaks that were excluded. To determine the location of each alpha and elastic peak, a Gaussian function was fitted to the peak and the centroid position was extracted. Subsequently, the calculated energy in MeV of each point was plotted against the peak position in channel number, and a 2 nd 116

135 Chapter 4: Calibration Fig. 4.5: One dimensional energy spectra of T2 pie 1 and T3 pie 9 using a 226 Ra alpha source. 117

136 Chapter 4: Calibration Fig. 4.6: The 2 nd order polynomial fits used in the calibration of T2 Pie 1 and T3 Pie 9 are shown. The chisquare and polynomial terms used to describe each fit are also provided. 118

137 Chapter 4: Calibration Fig. 4.7: Relative deviation of the measured points from the calibration fit for T2 Pie 1 and T3 Pie

138 Chapter 4: Calibration order polynomial fit was applied. The polynomial fits and calibration points for T2 pie 1 and T3 pie 9 are plotted in Fig. 4.6, with the chi-square and the polynomial terms describing the functions displayed in the legends. As show in Fig. 4.7, all calibration points are within 4.5% of the fit for all T2 pies, and within 6% for T3 pies The fits for T3 pies 0 7 and describe the points to within 5%. However, the highest elastic energy used in calibrating the second group of T3 pies was 25.0 MeV. As a result, the energy of the elastic peak for T3 pies 0-7, following the calibration differs by as much as 19% from the calculated energy for 20 O at E lab = 41.0 MeV. This does not present a significant problem for the experiment as fusion residues in the angular range subtended by T3 cannot have such high kinetic energies. Following energy calibration of each pie segment, one-dimensional energy spectra and two-dimensional energy versus time-of-flight (ETOF) spectra were generated. The ETOF spectra of T3 and T2 are extremely important as they are used to distinguish evaporation residues from beam particles. Shown in Fig. 4.8 are the calibrated ETOF spectra of T2 pie 1 and T3 pie 9, which are representative of the remaining T2 and T3 pies. For reference the corresponding uncalibrated spectra were shown in Fig As described previously in Chapter 3, the ETOF spectra of both detectors exhibit a distinct elastic peak, a slit scatter line, and a distribution of counts forming a haze both below and above the scatter line. A significant difference between the T3 and T2 detectors, however, is the dark vertical line in the T2 pies spectra that ranges from the elastic peak at about 40 MeV to an energy of roughly 6 MeV. This line is the result of incomplete charge 120

139 Chapter 4: Calibration Fig. 4.8: Two dimensional energy versus time-of-flight spectra of T2 pie 1 and T3 pie 9 calibrated in energy for 20 O at E lab = 41.0 MeV. 121

140 Chapter 4: Calibration collection in the pie segments. The vertical band is nearly absent from the T3 ETOF in Fig. 4.8, indicating the occurrence of incomplete charge collection to be far less frequent for the T3 pies than the T2 pies. The method to calibrate the ring segments utilized the pie calibration previously described. In calibrating the T2 rings, the sum of the pies energy in MeV was plotted as a function of the ring energy in raw channel number for each ring segment. Examples of these two dimensional energy spectra using rings 2 and 11 are given in Fig. 4.9 for an 20 O beam at E lab = 41.0 MeV. A positive correlation between the pie and ring energy is observed in the form of a near linear, positively-sloped line that extends from approximately 1 MeV to the elastic energy of ~ 40 MeV. Counts located on this line represent cases in which the pie and ring collected approximately the same energy. The slight curvature of the line is the result of non-linearities in the charge-sensitive amplifiers and electronics used to process the signals. A second positively-sloped line is present in the spectra, located below the good pie-ring correlation and resulting from a higher energy collected in the rings than in the pies. At the base of this line, a horizontal band appears and ranges in ring energy from approximately channel 150 to 1300 but remains at an energy of about 6 MeV in the pies. This band can also be observed in the T2 pie 1 ETOF of Fig. 4.8 as the horizontal line located at 6 MeV, which spans in time from -368 to -362 ns. Another prominent feature of the pies vs. ring energy spectra is the vertical band that extends from the elastic peak, ranging in pie energy from ~41 MeV to ~1 MeV while remaining constant in ring energy. Counts located on this line possess a 122

141 Chapter 4: Calibration Fig. 4.9: Two dimensional spectra of the T2 pies energy vs. the energy in T2 rings 2 and 11, for an 20 O beam at E lab = 41.0 MeV 123

142 Chapter 4: Calibration ring energy consistent with elastically scattered beam particles, but due to incomplete charge collection in the pies exhibit a lower pie energy. To proceed with the T2 ring calibration a gate was drawn to encompass the positive correlation in each pies vs. ring energy spectrum. Displayed in Fig. 4.9 are the gates for ring 2 and ring 11. A profile histogram of the pies energy in MeV vs. the ring energy in channel number was generated for each ring but only included counts located within the gate. A profile histogram displays the average y-value for each bin in the x-axis, which is ideal for minimizing the effect of statistical fluctuations in order to obtain an accurate fit to the data. A 2 nd order polynomial function was fitted to the counts located in the profile histogram and used for the ring calibration. The gated profile histograms and polynomial fits for rings 2 and 11 are shown in Fig. 4.10, where the black points depict the data located within the gate and the dashed red line represents the fit. As previously noted for the T2 pie segments, incomplete charge collection was a significant problem during the experiment. Apparent in Fig. 4.9 is the fact that the T2 pie segments consistently collected a lower energy than the ring segments. Consequently, in the subsequent analysis the highest energy observed in either the rings or pies was used for the energy of the particles. This approach was justified since the probability of two particles entering the same detector is low. Moreover the cases of multiple non-adjacent pies and rings could be easily distinguished and excluded. To demonstrate the difference between the pies energy and the maximum energy, ETOF spectra representing both cases are provided in Fig for an 20 O beam at E lab = 41.0 MeV. The energy shown in the 124

143 Chapter 4: Calibration Fig. 4.10: Profile histograms of T2 rings 2 and 11 following application of the gate for good pie-ring energy correlation. The data is displayed in black and the 2 nd order polynomial fits in red. 125

144 Chapter 4: Calibration Fig. 4.11: T2 energy vs. time-of-flight spectra using the pies energy (top) and using the maximum energy between the rings and the pies (bottom) for 20 O at E lab = 41.0 MeV. 126

145 Chapter 4: Calibration top panel of Fig corresponds to the energy deposited in any of the 16 T2 pies, and the time-of-flight is the difference in time between the 18mm MCP and the corresponding pie segment. The spectrum in the bottom panel again utilizes the timing signals produced in the T2 pies, but the energy shown is the maximum energy observed for a particle in either the pie or ring segments. The principal features of the ETOF spectrum using the maximum energy remain largely unchanged from those exhibited when strictly using the pie energy. The major difference between the two spectra is the vertical incomplete charge collection band that extends from the elastic peak. The vertical line, which ranges from ~ 6 to 40 MeV in the pies ETOF spectrum, ranges instead from ~ 25 to 40 MeV in the maximum energy ETOF. Though reduced, the persistence of the vertical band indicates that incomplete charge collection is present in the rings as well as the pies. Nonetheless, use of the maximum energy observed in all T2 segments is clearly an improvement over use of the pie energy alone. 127

146 Chapter 4: Calibration Bibliography: Chapter 4 [1] National Nuclear Data Center, [2] James F. Ziegler et al., Nucl. Inst. and Meth. B 268 (2010). 128

147 Chapter 5 Evaporation Residues and Light Charged Particles Resulting from the Fusion of 20 O + 12 C 5.1 Purity of the 20 O Beam An important concern in radioactive beam experiments is the purity of the incident beam. In order to identify contaminants in the 20 O beam throughout the experiment, the degrader ionization chamber was used to measure the energy loss of incident beam particles. As an example, a one dimensional energy spectrum measured by the third CID anode during the experiment is provided in Fig The most prominent peak in the spectrum corresponds to 20 O ions. A much smaller peak located to the immediate right of 129

148 Chapter 5: ERs and LCPs Resulting from 20 O + 12 C Fig. 5.1: One dimensional energy loss spectra of the third CID anode for an 20 O beam (Left) and for a 20 F beam (Right). the 20 O is attributed to 20 F, which is the primary contaminant expected in the radioactive beam. This assignment of the peaks is based upon energy loss calculations for the two ions. To verify these two peak assignments, a test was performed at the conclusion of the experiment by tuning the CIME cyclotron to accelerate 20 F rather than 20 O. The resulting CID anode spectrum for this test is displayed in the right panel of Fig Clearly the 20 F peak becomes dominant while the intensity of the 20 O peak decreases, confirming that the 130

149 Chapter 5: ERs and LCPs Resulting from 20 O + 12 C peaks have been correctly assigned. In the subsequent analysis, only events corresponding to incident 20 O have been selected based upon the 20 O peak in the CID spectra. The 20 O gate used is shown for reference in the left panel of Fig Experimental Results Displayed in Fig. 5.2 is a two-dimensional spectrum of the energy versus time-offlight for particles entering the T2 detector. The energy displayed is the energy deposited in the silicon detector while the time-of-flight corresponds to the time-of-flight between the target MCP and the T2 detector. Clearly evident is the prominent elastic peak with an energy of 40 MeV. Originating from the elastic peak and extending down to approximately 25 MeV is a near vertical band. Particles along this locus have a time-offlight consistent with elastically scattered particles. For these particles, a lower energy is measured due to the incomplete collection of charge by the silicon detector. Walk of the leading edge discriminators used for the silicon timing signals is responsible for the slight positive slope of this band. This incomplete charge collection occurs despite the detector being biased to -80V, well above the manufacturer s nominal full depletion value. Biasing at a voltage of -90V did not noticeably reduce the incomplete charge collection problem. A quoted breakdown voltage of -100V prevented biasing the detector to a substantially higher value during the experiment. Also evident in the spectrum is a locus of points originating from the elastic peak and increasing in time as the deposited energy decreases. Points along this locus correspond to particles degraded in energy prior to 131

150 Chapter 5: ERs and LCPs Resulting from 20 O + 12 C Fig. 5.2: Two dimensional energy-tof spectrum for particles entering the T2 detector. The polygonal gate indicates the region used to search for coincidences between evaporation residues in T2 and light charged particles in T3. See text for details. Nuclide 22 MeV 27 MeV 42 MeV 30 Si Si Al Al Mg Mg Table 5.1: Predictions of the evapor model of the percentage of the evaporation residue distribution populated by various nuclides. 132

151 Chapter 5: ERs and LCPs Resulting from 20 O + 12 C entering the silicon detector. Consequently, their measured time-of-flight reflects their degraded energy. Slit scattering of particles from the acceleration and reflecting grids of the target MCP detector is the most likely source of the energy loss. Located below the slit scatter line is a haze corresponding to slit scattered particles for which incomplete charge collection occurs. Located at higher energies and longer times with respect to the slit scatter line is yet another locus of points. This locus exhibits a similar energy-time relationship as the slit scatter line. The large cross-section of barns associated with these points clearly indicates that they have an atomic and not nuclear origin. However, as they occupy the same region of the energy-tof spectrum as that expected by the evaporation residues they represent a daunting background for measuring the fusion cross-section. Subsequent to the experiment bench tests with alpha sources conclusively demonstrated that scattering from the reflecting grids of the MCP provided a false early start signal resulting in this background (see Chapter 6). The energy-tof spectrum for particles entering the T3 detector is displayed in Fig While the principal spectral features of the elastic peak, incomplete charge collection, and slit scatter line observed for T2 also exist for T3, it is interesting to note that the incomplete charge collection in T3 appears to be significantly less severe than for T2. The background due to false early target MCP signals from the reflecting grid is also evident in this detector supporting the result that the problem arises from the MCP detector and not the T2 silicon detector. The evaporation residues present in the experiment result from de-excitation following the fusion reaction. The nuclide 133

152 Chapter 5: ERs and LCPs Resulting from 20 O + 12 C Fig. 5.3: Two dimensional energy-tof spectrum for particles entering the T3 detector. composition of the evaporation residue distribution can be calculated by utilizing a multiparticle Monte Carlo evaporation code such as evapor [1]. This model uses a simple fusion model (Bass) [2] to predict the fusion cross-section, followed by a Hauser Feschbach approach to model its subsequent decay. Calculations were performed at incident energies varying from E lab = 20 MeV to E lab = 45 MeV assuming a triangular angular momentum distribution. In Table I, the percentage of the evaporation distribution attributable to various nuclides is shown for three incident energies. Even at the lowest 134

153 Chapter 5: ERs and LCPs Resulting from 20 O + 12 C energy, E lab = 22 MeV, while 84% of the residues are formed through purely neutron decay channels, a significant fraction of the evaporation residue yield 16% is formed through some charged particle emission. With increasing incident energy, the percentage of yield involving some charged particle emission increases to 27% at E = 42 MeV. In order to extract fusion events from the large atomic background evident in Fig. 5.2 and Fig. 5.3, we therefore elected to require a coincidence between detection of an evaporation residue in T2 and a charged particle in T3. The potential region of interest for evaporation residues is well-defined by the slit-scatter line in Fig. 5.2 and extends to longer times. It is indicated by the polygonal gate displayed in Fig Coincident with detection of a particle in this gate a particle in T3 was required. All coincident particles in T3 are clustered in time occurring with a time spread of 2 ns. This time spread of the charged particle in T3, short compared to the RF beam burst period of 100ns, indicates that random coincidences play no significant role. This result is hardly surprising due to the low beam intensity. The E-TOF spectra of coincidences in T2 is shown in Fig. 5.4 for three incident energies. For reference the position of the slit scatter line is indicated as a solid line. It should be noted that the energy depicted corresponds to the energy measured in the silicon detector. At the highest incident energy, E lab = 41.0 MeV, all of the coincident particles are clustered in the range 8 MeV E 22 MeV. At the intermediate energy, E lab = 25.0 MeV, although particles for the most part are clustered in the same energy range, a couple of particles are observed with E 5 MeV. At the lowest incident energy, E lab = 19.8 MeV, the detected energies are significantly lower, E 13 MeV. 135

154 Chapter 5: ERs and LCPs Resulting from 20 O + 12 C Fig. 5.4: Energy-TOF spectra of particles in the T2 detector observed in coincidence with a charged particle in T3. The solid line indicates the position of the slit-scatter line observed in Fig

155 Chapter 5: ERs and LCPs Resulting from 20 O + 12 C Fig. 5.5: (Color online) Efficiency for detecting the residue and light charged particle. To ascertain whether the measured energies for evaporation residues and light charged particles match the expected energies for these reaction products, we compared the experimental data with the predictions of the statistical model code evapor [1]. Evaporation residues and light charged particles from the model were subsequently filtered by the geometrical acceptance of the experimental setup and detection thresholds were accounted for. The efficiency for detecting a LCP in coincidence with a residue in T2 is principally determined by the small solid angle of the T3 detector and the isotropic emission of the LCPs. As evident in Fig. 5.5 this efficiency is relatively constant with incident energy and approximately 3.0% to 3.5%. The efficiency calculations include the 137

156 Chapter 5: ERs and LCPs Resulting from 20 O + 12 C influence of the finite beam spot size as well as multiple scattering in CID. The beam spot size was assumed to be a gaussian with a width of 7mm at 4 sigma, based upon beam optics calculations. The multiple scattering in CID as a function of pressure was determined by an experimental measurement at OakRidge National Laboratory. In this experiment, a low-intensity beam of 3 MeV/A 18 O was passed through a gas cell containing CF 4 and the resulting beam spot was imaged on a multiwire proportional counter located downstream. It is evident from Fig. 5.5 that neither the multiple scattering in CID nor the finite beam spot has a significant influence on the calculated coincidence efficiency. Presented in Fig. 5.6 are the energy distributions for evaporation residues (left column) and LCPs (right column) measured in the experiment (solid line), together with the predictions of the evapor model (dashed red line). As we do not experimentally determine the identity of the light charged particle, the predicted energy distributions shown are summed over all charged particles. The predicted distributions for both evaporation residues, as well as light charged particles, have been corrected for energy loss in half the target as well as a nominal dead layer on the surface of the silicon detectors. The thickness of this dead layer was assumed to have an effective thickness of 0.7 μm Si equivalent, consistent with similar detectors [3]. The model predictions have been arbitrarily scaled for clarity, as indicated in the figure. It is evident that at all three incident energies shown, the energy distributions for the evaporation residues are in reasonable agreement with the model calculations though at the highest incident energy 138

157 Chapter 5: ERs and LCPs Resulting from 20 O + 12 C Fig. 5.6: (Color online) Energy distribution of light charged particles and residues compared to predictions of the statistical model evapor. 139

158 Chapter 5: ERs and LCPs Resulting from 20 O + 12 C the experimental energies are lower by approximately 5.5 MeV. In the case of the light charged particle spectra, somewhat larger differences between the measured and predicted distributions are observed. Some of the difference between the measured and predicted distributions may arise because of differences between the emitted particles produced and those predicted. Having corrected for the geometric efficiency, it is possible to extract the cross-section associated with fusion followed by charged particle emission, σ cp. The total number of incident beam particles was determined from the target MCP detector and cross-checked against the CID detector at the running intensity and the SBD detector at low intensity. In both cases, the integrated counts of the target MCP and the other two detectors were in reasonable agreement though the integrated counts in CID were typically 10% higher. This slightly larger number of counts in CID can be understood due to the divergence of the beam on degrading. The resulting cross-sections are presented in Fig These measured cross-sections decrease with decreasing incident energy from 492 ± 105 mb at E lab = 40.6 ± MeV to 82.3 mb ± 26 at E lab = 19.6 ± MeV. Vertical error bars reflect the statistical errors associated with the measurement while horizontal error bars indicate the dispersion (sigma) in incident energy due to degrading the incident beam. The observed change in the cross-section with energy is influenced by both the overall decrease in the fusion cross-section with decreasing energy, as well as changes in the population of charged particle decay channels with decreasing energy. Shown for comparison in Fig. 5.7 are the predictions for both the total fusion cross-section (solid 140

159 Chapter 5: ERs and LCPs Resulting from 20 O + 12 C Fig. 5.7: Measured fusion cross-sections associated with charged particle emission channels (solid points) compared to predictions of the evapor model. The total predicted fusion cross-section is shown as a solid line while the cross-section predicted for the charged particle channels is represented by the dashed line. line) and the cross-section associcated with charged particle channels (dashed line) predicted by the evapor model. Comparison of the experimental data with the model predictions yields two significant results. The first noteworthy point is that the measured cross-sections exceed those predicted by the model for the charged particle channels by a factor of 2. This is true at all incident energies including the highest incident energy which is well above the Coulomb barrier. The second interesting result is that the dependence of the experimental cross-section with decreasing incident energy is weaker 141

160 Chapter 5: ERs and LCPs Resulting from 20 O + 12 C than that predicted by the fusion-evaporation model. Each of these observations is independently interesting. For example, if one might question the uncertainty involved with degrading the incident beam to the lowest incident energy, the impact of this uncertainty in energy is significantly less, and less important for the highest energy point where the excitation function is relatively flat. Thus, the larger cross-section at the highest energy point is particularly noteworthy. 5.3 Fusion of 16 O + 12 C To provide a reference for the fusion of neutron-rich oxygen and carbon nuclei as well as ascertain whether the simple fusion model correctly predicts the fusion cross-section for β-stable nuclei, we measured the reaction 16 O + 12 C at the conclusion of the GANIL experiment. A 16 O beam was accelerated to an energy of 3 MeV/A by the CIME cyclotron and delivered to the experimental area where it was degraded by CID to obtain multiple beam energies between E Lab = 23 and 42 MeV. The beam became incident on the 100 μg/cm 2 carbon foil located on the active target MCP, and the identical setup as the 20 O measurement was used. The percentages of the different nuclides produced from the fusion of 16 O + 12 C have been predicted using evapor and are provided in Table 5.2 at the two incident beam energies measured at GANIL. In contrast to the decay of 32 Si*, the fusion product of this reaction, 28 Si*, decays principally by charged particle emission. The efficiency for detecting light charged particles in coincidence with fusion residues is shown in Fig In contrast to the efficiency calculations for the 20 O reaction shown 142

161 Chapter 5: ERs and LCPs Resulting from 20 O + 12 C Nuclide 23.5 MeV 28 MeV 42 MeV 27 Si Al Al Mg Mg Na Ne Table 5.2: Predictions of the evapor model code of the percentage of the evaporation residue distribution for different nuclides in the fusion of 16 O + 12 C. Fig. 5.8: (Color online) Efficiency for detecting the residue and light charged particle in the reaction 16 O + 12 C measured at GANIL. 143

162 Chapter 5: ERs and LCPs Resulting from 20 O + 12 C in Fig. 5.5, a dependence on the multiple scattering in CID is observed as well as an overall general increase with incident energy. With increasing incident energy, namely a decreasing pressure in CID, the influence of the multiple scattering in CID diminishes as expected. The analysis of the 16 O data was performed in the same manner as 20 O: polygonal gates were used to select counts in the T2 energy-tof spectra in order to identify coincident light charged particles in T3. Shown in Fig. 5.9 are the resulting energy-tof spectra of the T2 silicon detector for coincident particles at three incident beam energies. The position of the slit-scatter line is again provided for reference. At a beam energy of 42.1 MeV, all evaporation residues are located within the energy region 3 MeV E lab 35 MeV, while at a beam energy of 28.1 MeV all residues fall within the range 3 MeV E lab 24 MeV. The residues at the third beam energy, 23.6 MeV, are located between 3 MeV and 20 MeV. Presented in Fig are the one-dimensional energy spectra of the evaporation residues (left column) and the light charged particles (right column) for three incident beam energies. The measured energies are indicated using a black line, while the predicted energy distributions from the evapor model are indicated by a red line. As with 20 O, the predicted energies of the evaporation residues and light charged particles for 16 O have been corrected for energy loss in half the target thickness and the dead layer of the Si detector, estimated to be 0.7 μm. Provided in Table 5.3 are the measured and predicted average energies of the evaporation residues and light charged particles. The 144

163 Chapter 5: ERs and LCPs Resulting from 20 O + 12 C Fig. 5.9: Two dimensional energy-tof spectra of evaporation residues in T2 at 3 incident beam energies for 16 O. 145

164 Chapter 5: ERs and LCPs Resulting from 20 O + 12 C Fig. 5.10: Energy distributions of the evaporation residues and light charged particles are shown for the fusion of 16 O + 12 C at three incident beam energies. The measured energies (black) are provided in addition to the predicted energies from the evapor model (red). 146

165 Chapter 5: ERs and LCPs Resulting from 20 O + 12 C Measured Predicted E lab (MeV) < E res > (MeV) < E LCP > (MeV) < E res > (MeV) < E LCP > (MeV) Table 5.3: Measured average energies of evaporation residues and light charged particles produced in the reaction 16 O + 12 C. The predicted average energies from the evapor model code are also provided for comparison. information provided in the table demonstrates the measured energies of both the evaporation residues and the light charged particles are in good agreement with the model predictions. A second measurement of 16 O + 12 C was undertaken at Western Michigan University (WMU) following the experiment at GANIL to provide additional points on the fusion excitation function. A beam of 16 O was accelerated by the 6 MV tandem to energies between 20 and 35 MeV and impinged on a 100 μg/cm 2 carbon foil. The target foil also served as the electron emission foil of the target MCP detector. As the beam energy could be easily varied and a low energy beam could be transported, there was no need to degrade the beam energy using the CID detector. Consequently, the uncertainty in the energy is determined by the tandem accelerator and is typically on the order of 20 kev. Evaporation residues and light charged particles were detected with the silicon detectors as in the GANIL experiment. Evaporation residues in T2 coincident with light charged particles in T3 were identified by their energy and time-of-flight. The subsequent analysis 147

166 Chapter 5: ERs and LCPs Resulting from 20 O + 12 C of the experimental data closely followed the analysis performed for the GANIL experiment. Shown in Fig as the solid points is the extracted fusion cross-section that decays by charged particle emission. The measured cross-section includes the following energies from the GANIL experiment: E CM = 10.1 and 12.0 MeV, and the WMU experiment: E CM = 8.6, 10.3, 12.4, and 15.0 MeV. For comparison to the experimental data, both the predicted total fusion cross-section and the fusion cross-section associated with charged particle decay are shown as the solid and dashed lines respectively. The predictions represent the Bass fusion cross-sections calculated using the evapor code. It is apparent that the measured cross-section agrees well with the model prediction. This agreement indicates that the model does a good job of predicting the fusion cross-section of the reaction 16 O + 12 C and the subsequent charged particle de-excitation of the 28 Si *. The fusion cross-section for the highest energy point in the GANIL experiment, 14.8 MeV, has been omitted from the cross-section plot. An error occurred during the measurement of this point which lead to a measured fusion cross-section ~ 30% below the predicted fusion cross-section for the charged particle channels. As this was the first 16 O point measured during the experiment, one possible explanation for the error is the beam tune, which was adjusted before measurement of the next point at 12.0 MeV. Another possible explanation is the timing of the trigger electronics, which was adjusted for each beam energy. To verify that an error was responsible for this discrepancy and not an intrinsic flaw in the method or the detectors, a new measurement was made at WMU 148

167 Chapter 5: ERs and LCPs Resulting from 20 O + 12 C using a similar incident beam energy, E CM = 15.0 MeV. The fusion cross-section measured at WMU was found to be in excellent agreement with model predictions as shown in Fig Fig. 5.11: Measured fusion cross-sections from the reaction 16 O + 12 C associated with charged particle emission channels compared to the predictions of the evapor model. The points measured in the GANIL experiment (solid points) as well as the WMU experiment (open triangles) are provided. The total predicted fusion cross-section is shown as a solid line while the cross-section predicted for the charged particle channels is represented by the dashed line. 149

168 Chapter 5: ERs and LCPs Resulting from 20 O + 12 C 5.4 Conclusions This first attempt to measure the total fusion cross-section in the system 20 O and 12 C demonstrated that while the overall approach utilized appears feasible, there are still some technical obstacles to overcome. Most notably it is necessary to eliminate slit scattering from the target MCP reflecting grid if one is to measure the fusion excitation function. Despite this setback with the present dataset, we have successfully extracted the cross-section for fusion of 20 O and 12 C nuclei into a compound nucleus which subsequently undergoes charged particle decay. The measured cross-section exceeds that of a simple fusion model for the same channels. This discrepancy could either point to an enhancement of the fusion cross-section relative to this simple model, or may result from under-prediction of the charged particle channels as compared to neutron decay in the deexcitation cascade. It may be instructive to compare the present experimental results with more sophisticated fusion models which account for transfer channels prior to fusion. The present data suggests that future experiments to measure the total fusion cross-section should also attempt to measure the different de-excitation channels. Such data might provide new insights on the de-excitation of neutron-rich light nuclei. 150

169 Bibliography: Chapter 5 Chapter 5: ERs and LCPs Resulting from 20 O + 12 C [1] N.G. Nicolis, J.R. Beene, EvapOR, a multi-particle Monte Carlo evaporation code, 1993, unpublished. [2] R. Bass, Phys. Rev. Lett. 39 (1977) 265. [3] D. Fox et al., Nucl. Instr. and Meth. A 368, 709 (1996). 151

170 Chapter 6 Testing of the T2 silicon and 18mm MCP 6.1 Overview Precise measurement of both incident particle energy and time-of-flight by the annular silicon detectors was critical in the experiment to successfully separate evaporation residues from scattered beam particles. However, the two dimensional energy versus time-of-flight (ETOF) spectra exhibited a number of unexpected features as described in Chapter 3. For reference, an ETOF spectrum of T2 pie 4 is provided in Fig. 6.1 and is representative of the remaining T2 pies. The first set of features that was not well understood at the time of the experiment is the vertical and horizontal bands extending from the elastic peak. These bands are particularly prominent in the T2 detector. Of even greater concern are the large number of counts to the right of the slit scatter line, in the 152

171 Chapter 6: Testing of the T2 silicon and 18mm MCP region where the evaporation residues are expected to occur. Attempts to distinguish true data points from anomalous ones through analysis proved unsuccessful. There was Fig. 6.1: Two dimensional energy vs. time-of-flight spectrum of T2 pie 4 for 20 O at E lab = 41.0 MeV. clearly a need to test in order to understand the performance of the 18mm MCP and silicon detectors in the experiment and to investigate the source of the unexpected features in the energy versus time-of-flight (ETOF) spectra. To accomplish this, we performed a series of tests at IU with a partial setup consisting of the 18mm MCP and the T2 silicon detector. 153

172 Chapter 6: Testing of the T2 silicon and 18mm MCP 6.2 Test Configuration 1 The first test setup, depicted in Fig. 6.2, consisted of a 226 Ra alpha source, an 18mm MCP detector, and a T2 silicon detector. The same MCP and T2 silicon used in the fusion experiment were used in testing. We investigated changing the relative orientation of the 18mm MCP with respect to the alpha source. The distance between the source and the 18mm MCP carbon foil was 2.0 when the reflecting wire planes faced the alpha source compared to 0.75 when the MCP carbon foil faced the source. In both cases the T2 silicon detector was located at a distance approximately 34.5 downstream from the alpha source. Fig. 6.2: The configuration for the first series of tests at I.U. is shown. Only 1 MCP detector is utilized here, and is the same detector used in the fusion experiment. 154

173 Chapter 6: Testing of the T2 silicon and 18mm MCP The first test was performed to compare the two orientations of the 18mm MCP detector. In the first case, incident alpha particles entered the 18mm MCP through the reflecting wire planes, while in the second case incident alphas entered through the carbon emission foil. It had been hypothesized during the GANIL experiment that scattering of incident beam ions (or alpha particles) from the reflecting grid planes of the 18mm detector could result in the ejection of electrons from the grids which could become incident directly on the microchannel plates. If the MCP reflecting wire planes are located upstream relative to the emission foil, electrons ejected from the grid could effectively produce a signal that would arrive earlier in time compared with electrons ejected from the emission foil. Generation of an earlier time signal in the MCP would result in an apparent longer time-of-flight to the silicon detector, therefore shifting the data point to the longer time side of the slit-scatter line. The complement to this scenario is also true with the carbon foil located upstream from the reflecting wire planes and an apparent shorter time-of-flight resulting from false signals generated by the reflecting grids. However, this is less of a concern because shorter flight times would lie on the opposite side of the slit-scatter line and would not interfere with detection of evaporation residues. For the MCP orientations test the T2 silicon detector was biased at -80V and the junction (ring) side of the detector faced the alpha source: the same conditions as in the fusion experiment. The spectra presented in Fig. 6.3 are the sums of all T2 pies with the exception of 0 and 15. These two pies have been omitted due to poorer energy resolution 155

174 Chapter 6: Testing of the T2 silicon and 18mm MCP in comparison with the remaining pies. Scaling of the individual pies was performed using the known alpha energies of the 226 Ra source and using pie 2 as a reference. The MCP detector used in this test is referred to as MCP2 in order to avoid confusion with the MCP detector used in subsequent tests. The configuration of the electronics used in this test is presented in Fig Fig. 6.3a (Left): The 18mm MCP was oriented with the reflecting wire grid planes located upstream relative to the carbon emission foil. The T2 silicon detector was biased at -80V and the ring side of the detector faced the alpha source, the same conditions as during the fusion experiment. Fig. 6.3b (Right): The 18mm MCP was oriented with the carbon emission foil located upstream relative to the reflecting wire grids. The T2 silicon detector was biased at -80V, and the ring side of the detector faced the source. 156

175 Chapter 6: Testing of the T2 silicon and 18mm MCP The configuration of the electronics in the first series of tests is straight-forward. Signals produced in each pie segment of the T2 silicon detector were passively split using a frequency dependent splitter with a transition frequency of 7.5 MHz, which resulted in a low frequency (slow) component and a high frequency (fast) component. The low frequency component provides particle energy information and was amplified using a charge sensitive pre-amplifier (CSA) and a PICO-systems shaping amplifier [1]. The signal from the shaping amplifier was sent to a V785 analog-to-digital converter (ADC) [2] to be digitized and recorded by the computer. The high frequency component of the pie signal was amplified by a fast timing amplifier and discriminated using a V895 leading-edge discriminator (LED). The individual outputs of the LED were sent to a V1290 time-to-digital converter (TDC). In addition to the individual outputs, the LED produces an OR of the 16 pie signals, which was used to create the electronic trigger. Signals produced in each ring segment of the T2 silicon were amplified using a CSA and sent to a shaping amplifier. The output of the shaping amplifier was then sent to a V785 ADC. The signal from the 18mm MCP anode was amplified using an ORTEC VT120 fast timing amplifier [3], and the amplified output was discriminated using a Tennelec TC 454 constant fraction discriminator. The resulting logic signal was used as a stop signal in a V1290 TDC. The fast T2 silicon OR from the V895 was used to generate the computer and TDC triggers, the ADC gates, and to provide a reference to the TDC. The T2 silicon OR was sent from the LED to a four-fold logic unit to be vetoed by the computer busy. The 157

176 Chapter 6: Testing of the T2 silicon and 18mm MCP computer busy was provided by the VM-USB (VME controller). The first output of the vetoed silicon OR (logic unit) was delayed 1.1 μs, widened appropriately, and then used to gate both ADCs. A second output of the vetoed silicon OR was sent to a logic fan-in/ fan-out (FIFO). One output of the logic fan triggered the TDC, while a second output of the logic fan was delayed 90 ns and triggered the computer. A third output of the logic fan was provided to the TDC as a reference time. The silicon detector signal was chosen Fig. 6.4: The first configuration of the electronics is shown. FTA = fast timing amplifier, LED = Leading Edge Discriminator, TDC = Time to Digital Converter, GDG = Gate and Delay Generator, FIFA = Fan In/ Fan Out, CSA = Charge Sensing Pre-Amplifier, ADC = Analog to Digital Converter, CFD = Constant Fraction Discriminator. 158

177 Chapter 6: Testing of the T2 silicon and 18mm MCP to trigger the computer because only events in which an alpha particle is incident on the silicon detector are of interest. In comparing the two cases for the differing MCP orientations one observes a lesser number of counts in the ghost region for the case of the carbon foil facing the alpha source (Fig. 6.3b) as opposed to the case of the reflecting grids facing the alpha source (Fig. 6.3a). The ghost region, which refers to the haze of counts located directly to the right of the slit-scatter line, is the region in which the evaporation residues were predicted to occur in the fusion experiment. Ideally this region should not contain any counts in the case of the alpha source. This result demonstrates that the MCP should be oriented such that incident ions enter the detector through the carbon emission foil and not the reflecting wire planes. The orientation of the MCP does not appear to impact the other prominent features of the ETOF spectra, namely the horizontal or vertical bands. 6.3 Test Configuration 2 A second test was performed to compare particle entry on the ohmic (pie) side with particle entry on the junction (ring) side of the T2 silicon detector. This test made use of the mechanical setup presented in Fig. 6.2 and the electronics configuration presented in Fig The T2 silicon bias was increased from -80V to -150V for this test because at a bias voltage of -80V alphas incident on the pie side of the detector produced poorly shaped fast timing signals with long rise-times. This result can be understood as the detector depletes from the ring side so that at -80V, the field at the pie surface is 159

178 Chapter 6: Testing of the T2 silicon and 18mm MCP apparently insufficiently strong to produce good timing signals. The rise-time for pie side entry at -80V was on the order of ns, however, the rise-time for ring side entry with the same bias voltage was typically 5 6 ns. With increasing bias voltage the fast signal shape improved and the rise time approached 5 ns at a bias voltage of -150V. MCP2, the detector used in the fusion experiment, was used for this test. The resulting Fig. 6.5a (Left): The ETOF spectrum between the 18mm MCP and the T2 silicon is shown with alphas entering the ring side of the silicon at -150V. The MCP is oriented with the reflecting wire planes facing the alpha source. Fig. 6.5b (Right): The ETOF spectrum between the 18mm MCP and T2 silicon with alpha particles entering the pie side of the T2 silicon at -150V is shown. The MCP is oriented with the carbon foil facing the alpha source. 160

179 Chapter 6: Testing of the T2 silicon and 18mm MCP ETOF spectra for the two orientations of the T2 detector are shown in Fig. 6.5, and are the sums of all T2 pies with the exception of 0 and 15. It should be noted that both the orientation of the 18mm MCP and the orientation of the T2 silicon are different between the two spectra in Fig The principal difference between the two silicon orientations is the disappearance of the vertical bands from the ring side entry case (Fig. 6.5a) to the pie side entry case (Fig. 6.5b). This suggests the vertical bands are a result of incomplete charge collection when ions are incident on the junction side rather than the ohmic side of the detector. The improvement in the number of counts located in the ghost region to the right of the slitscatter line from Fig. 6.5a to Fig. 6.5b can largely be attributed to the change in MCP orientation from reflector to foil as previously described in the discussion of Figs. 6.3a and 6.3b. 6.4 Test Configuration 3 The next series of tests utilized a new mechanical geometry, pictured in Fig The geometry was changed in order to allow for an MCP coincidence and to decrease the distance between the source and the T2 silicon. The first test to use the new geometry was a comparison between the 18mm MCP used in the experiment (MCP2) and a second 18mm MCP of essentially the same mechanical design (MCP1). However, there are a few differences between the two MCPs that should be mentioned. First, MCP1 contains a chevron channel plate stack where as MCP2 contains a chevron stack followed by a single channel plate. Second, MCP1 possesses a pair of position-sensitive wire planes 161

180 Chapter 6: Testing of the T2 silicon and 18mm MCP Fig. 6.6: The 2 nd mechanical setup used for MCP + Si testing at I.U. is shown. This setup differs from the 1 st test setup in that two MCPs can be tested. The distance between the source and the T2 silicon has been decreased. located between the channel plates and the metal anode. These sense wire planes are not present in MCP2. The position sensitive capability of MCP1 was not used during the tests, however. Finally, the distance separating the channel plate stack and the metal anode differs between the two MCPs due to the presence of the sense wire planes in MCP1. The distance between the channel plate stack and the anode in MCP2 is ~ 1/32 compared with a distance of ~ 1 in MCP1. To compare the two different MCP detectors alpha data was taken with the MCP of interest inserted into the ISO cross and with the other MCP fully retracted. The resulting ETOF spectra are summaries of T2 pie channels 1 5 and 7 14 and are shown in Fig Pie 6 was omitted because its recorded alpha energies oscillated in time between two 162

181 Chapter 6: Testing of the T2 silicon and 18mm MCP Fig. 6.7a (Left): The resulting ETOF spectrum is shown for MCP1 + T2 with the pie side of T2 facing the alpha source and a T2 bias of -150V. MCP1 is oriented with the carbon foiling facing the alpha source. Fig. 6.7b (Right): The resulting ETOF spectrum for MCP2 + T2 with the pie side of the T2 facing the alpha source and a T2 bias of -150V. MCP2 was oriented with the reflecting grid facing the source, and was the detector used in the fusion experiment. values. It should be noted that the two MCPs used the same channel of the electronics for the data presented in Fig. 6.7 in order to demonstrate the electronics are not the source of the different results. The state of the T2 detector (pie side entry and a bias voltage of -150V) is the same in Figs. 6.7a and 6.7b. Fig. 6.7a is the result of using MCP1 with incident alphas entering 163

182 Chapter 6: Testing of the T2 silicon and 18mm MCP the carbon foil side of the detector and Fig. 6.7b is the result of using MCP2 with incident alphas entering the reflecting grid side of the detector. The two spectra are clearly different with regard to the horizontal bands that extend from short time-of-flights to the 4 He scatter line. The bands are considerably less pronounced in Fig. 6.7a (MCP1) than in Fig. 6.7b (MCP2). The horizontal bands are also more dominant in Fig. 6.5b where MCP2 was used with foil side entry as compared with MCP1 in Fig. 6.7a. This difference is most likely a result of the difference in the spacing between the channel plate stack and the anode in the two detectors. The greater prevalence of ghost points in Fig. 6.7b compared with Fig. 6.7a can again be attributed to the orientation of the MCP. 6.5 Test Configuration 4 A final test was performed using both microchannel plate detectors to form a coincidence. The T2 silicon was oriented with the pie side facing the alpha source and was biased to a voltage of -150V. The mechanical setup depicted in Fig. 6.6 was used for this test; however, the electronic configuration was altered from the previous tests to the configuration shown in Fig The electronics involved with processing the silicon detector signals and generating the trigger electronics is unchanged from the previous configuration. However, the processing of the MCP2 anode signal has been slightly altered, and the electronics to process the MCP1 signal has been added. Each MCP anode signal was amplified using a VT120 fast timing amplifier, and the resulting signal sent to a linear FIFO. The first output of the linear fan was amplified by a variable gain amplifier and delayed prior to being sent to a V792 charge-to-digital converter (QDC). The second 164

183 Chapter 6: Testing of the T2 silicon and 18mm MCP Fig. 6.8: The electronics configuration for the MCP coincidence test is depicted. The MCP coincidence is generated using a Four Fold Logic Unit. FTA = Fast Timing Amplifier, LED = Leading Edge Discriminator, TDC = Time to Digital Converter, GDG = Gate and Delay Generator, FIFA = Fan In/ Fan Out, CSA = Charge Sensing Pre-Amplifier, VGA = Variable Gain Amplifier, QDC = Charge to Digital Converter, CFD = Constant Fraction Discriminator. output of the linear fan was discriminated by a constant fraction discriminator, with one discriminator output sent to the TDC. A second channel of the MCP2 CFD was delayed by 10 ns, and taken as an AND with an output of the MCP1 CFD. The coincident MCP signal was used in two ways: one output was sent to the TDC, and a second output was 165