mopdf.com

2Project Description2.1Status of Current ProjectsThe research effort in medium energy nuclear physics at the University of Richmond is part of theprogram at the Thomas Jefferson National Accelerator Facility (JLab) in Newport News, VA. Wenow discuss our work to analyze recently collected data using the CLAS12 detector and continuedpreparations for future experiments. CLAS12 consists of a Forward Detector with a toroidal fieldgenerated by six sectors of superconducting coils. An array of drift chambers, time-of-flight counters, Cerenkov counters, calorimeters, and other devices measure and identify the reaction products[9]. The Central Detector covers large angles and is built around a solenoid magnet and anothersuite of detectors. In 2011, we made a commitment to the ‘design, prototyping, development, andtesting of software for event simulation and reconstruction in CLAS12’ as part of a Memorandum ofUnderstanding with JLab. The importance of software development in the 12 GeV era has grown.We remain committed to these goals.[10, 11, 12, 13]2.1.1Magnetic Form Factor of the NeutronThe elastic electromagnetic form factors are basic observables that describe the distribution ofcharge and magnetization inside the proton and neutron. Their measurement is a goal of thecurrent NSAC Long-Range Plan (see Section 2.1) [14], and forms a central part of the physicsprograms at JLab [15, 16, 17, 18]. We are part of a broad campaign to measure the four elastic,electromagnetic, nucleon form factors (electric and magnetic ones each for the proton and neutron)at JLab that includes seven experiments approved for running after the 12 GeV Upgrade at JLab[16, 17]. Gilfoyle is the spokesperson and contact person for JLab experiment E12-07-104 to measureGnM , the neutron magnetic form factor in Hall B using CLAS12 [19]. The experiment has an A rating from the Program Advisory Committee (PAC) and was awarded 30 days of beamtime in thefirst five years running of CLAS12 [20]. A large portion of our work now is analyzing the recentlycompleted deuterium runs to extract GnM and preparations for further CLAS12 operations. We arealso members of the collaboration to measure GnM in Hall A (E12-09-019).We first outline the steps of the the measurement of the neutron magnetic form factor followedby a summary of recent experimental activity. To extract GnM we form the ratio R of e n to e pevents in quasielastic scattering (QE) from deuterium. This quantity R depends on all four elastic,electromagnetic form factors (EEFFs): GnM , GnE , GpM , and GpE . The proton EEFFs are known tohigher precision than the neutron ones so their contribution to the GnM systematic uncertainty willbe limited. The neutron electric form factor GnE is not well known at high Q2 , but it’s contributionto R is small and parameterizations are available that account for its high-Q2 behavior. Moredetails on the method are in Section 2.2.1.We now discuss the status of the GnM analysis of the deuterium data. The data were collectedas part of the CLAS Collaboration’s Run Group B (RGB). The run statistics for the three separateexperimental runs are listed in Table 2. Nearly forty PAC days of running generated more than43 billion triggers constituting about 43% of total PAC days approved for RGB. The PI (Gilfoyle)served as Run Coordinator for RGB during the spring, 2019 run (seven days) and during the winter,2020 run (11 days) which came immediately after the JLab winter shutdown.With the arrival of the first CLAS12 deuterium data the PI formed a collaboration with Dr. B.Raue, a CLAS Collaboration colleague at Florida International University (FIU) and his doctoralstudent Ms. L. Baashen. The GnM analysis is Ms. Baashen’s PhD thesis. The group typicallymeets 1-2 times per week and pre-COVID Gilfoyle spent one day each week at JLab. Since March,2020 the group has maintained these regular meetings remotely.3

Beam [GeV]PAC daysTriggers [B]Charge [mC]DAQ [kHz]Spring, 201910.2, 10.621.721.479.614Fall, 201910.46.7921.724Winter, 202010.410.512.935.219Sum38.9 (43%)43.3136.5Table 2: Run Group B data collection properties and statistics.Analysis (“cooking”) of the RGB data is underway. As of October, 2020 cooking is complete(Pass 1) for the spring, 2019 run and Pass 1 for the fall, 2019 run will occur in the next month.Calibrations for the winter 2020 are in process (Pass 0).Some preliminary results from our group’s analysis are shown in Figure 1. The left-hand panelshows W2 plotted versus θpq for the 2 H(e, e0 p) reaction. The angle θpq is between the exchanged,virtual photon γ and the ejected proton. We expect the quasi-elastic production to be at small θpqwhile the inelastic background will occur at higher W2 and larger θpq for both protons and neutrons.The proton results in the left-hand panel of Figure 1 display that property - the quasielastic groupFigure 1: Preliminary results for RGB GnM analysis of the 2 H(e, e0 p) reaction showing W2 versusθpq , the angle between the virtual photon and the ejected proton (left panel). The right panel showsthe missing mass squared for the same reaction before selecting the quasi-elastic events.is clearly separated from the inelastic background. The right hand panel shows the square of themissing mass for neutral particles in our data sample. There is a clear separation between thephotons and neutrons. Optimizing the event selection is ongoing.A key step in measuring GnM with the ratio method is to also measure the neutron detectionefficiency (NDE) to determine precisely the number of e n events in R. We use the 1 H(e, e0 π n)reaction on a hydrogen target as a source of tagged neutrons to measure NDE in the CLAS12calorimeters [21]. We select 1 H(e, e0 π )n events with a missing mass cut and other kinematicconstraints and predict the location of the neutron in the the CLAS12 calorimeters. These eventsform the denominator in the NDE. We reject events if the predicted neutron path misses theCLAS12 fiducial volume. We then search for a neutron hit close to the expected location. If aneutron is detected it goes into the numerator of the efficiency.We are extracting the NDE from hydrogen data collected during Run Group A (RGA) running.Run statistics for the three separate runs are listed in Table 3. Over sixty PAC days of runningand 285 mC of charge collected constitutes over 40% of total beam-time approved for RGA. The4

Spring, 20186.4, 10.621.7126Beam [GeV]PAC daysCharge [mC]Fall, 201810.63099Spring, 201910.21060Sum61.7285 (46%)Table 3: Run Group A data collection properties and statistics.1.2NDENDEanalysis of the RGA data is far along and we have exploited that to begin extracting the neutrondetection efficiency. Preliminary results are shown in Figure 2. In the left-hand panel we show the1 Blue - Inbending electrons1.21 Black - both polaritiesBlue - CLAS6 4.2 GeV ResultsRed - Outbending electrons0.80.80.60.60.40.4 NDE in 0.740 0.006, pn 3.5 GeV NDE out 0.744 0.006, pn 3.5 GeV0.200123456 NDE 0.741 0.004, pmm 3.5 GeV0.2007pn [GeV]1234567pn [GeV]Figure 2: Preliminary results for the neutron detection efficiency from the 1 H(e, e0 π n) reaction asa function of neutron momentum pn . Data are from Run Group A.NDE extracted for opposite polarities of the CLAS12 torus magnet. The NDE rises rapidly in therange 0.5 GeV/c pn 3 GeV and levels off at higher Q2 . There are differences between the twomeasurements at low missing momenta where the NDE is changing rapidly. At pn & 3.2 GeV/cthe NDE for each polarity reaches at plateau at NDE 0.74. In the region of the plateau theaverage values for the plateau for the two torus polarities agree within less than 1%. In the righthand panel we show the NDE for the combined torus magnet polarities (black, filled, squares). Asexpected we see improved statistical precision at high pn . For comparison we show results fromthe previous CLAS6 detector in Hall B. The current CLAS12 calorimeters include the refurbishedCLAS6 calorimeter used to measure the blue points (open circles) in the right-hand panel. The twomeasurements are consistent with each other in the low-pn region where the NDE changes rapidly.With the construction of CLAS12 a new calorimeter (Pre-shower calorimeter or PCAL) was added.This addition explains the higher plateau for the CLAS12 detector because there is more materialfor the neutrons to pass through and a greater chance for them to interact and produce a signal.Finally, it is worth noting these results are from just the fall, 2018 RGA run so the amount of datain the final NDE sample will more than double.Simulation of the CLAS12 detector and the physics processes is an essential tool to understandthe CLAS12 response, to validate our own codes and algorithms, and to explore the physics moredeeply. We have developed a full, end-to-end simulation for the GnM project [5, 22, 23]. Theseprojects include development of an event generator for quasielastic events [24] and in the last 18months new tools for simulating the inelastic background. We have begun to use a variation ofthe Pythia event generator [25] that includes nuclear effects. The goal is to study the inelastic5

background under the neutron peak to optimize the extraction of the yield from the e n and e pevents. Our initial results are encouraging.In the fall of 2020, the JLab Program Advisory Committee (PAC) performed a jeopardy reviewof Run Group B. In a jeopardy review the PAC evaluates the continued relevance of the experimentsin the run group and can change the amount of remaining beam time, the scientific rating, etc.PAC48 (fall, 2020) recommended that Run Group B remain active, keep the full amount of theoriginal, approved beam time, and maintained the scientific rating at A. A related report by theTechnical Advisory Committee to the PAC assessed the theoretical goals of the run group. Itwas supportive of the Run Group B physics program and noted in the section discussing the GnMexperiment that “Theoretical support for these measurements remains very strong.” [18].2.1.2CLAS12 Software DevelopmentThe University of Richmond group continues to develop software to support continuing CLAS12operations. Since 2017 Richmond undergraduates and Surrey students have written codes to extractGnM , measure NDE, monitor analysis run-by-run, and a variety of other tasks.[2, 5, 7, 8, 22, 26].Here we describe two projects from the current grant period from the Surrey students that aresupported by this DOE grant.Code testing and validation are essential for developing robust, accurate, high-performancesoftware. The goal is to catch mistakes/bugs introduced as the programs evolve and to monitorchanges in performance. This goal is especially important in large collaborations where many peopleare working on the software so it changes at a rapid pace. During the nightly software builds, unittests are run on the subsystems to test their performance. The unit tests are part of an array ofautomatic tests the “have proven invaluable in overseeing software development” [27].Mr. M.Armstrong, a Surrey masters studentworking with the PI and JLab staff scientistDr.V.Ziegler updated and expanded a unit test onthe CLAS12 drift chambers to test changes to theCLAS12 reconstruction code. The unit test here ispart of the Drift Chamber (DC) subsystem. A previous unit test consistently produced false positivesso its warning had become unusable. The source ofthis failure was identified and the test modified foruse with the latest version of the CLAS12 CommonTools. An example is shown in Figure 3. Trackswere simulated with the CLAS12 standard simulation gemc and reconstructed with the CLAS Col- Figure 3: Distribution of the z-componentlaboration’s Common Tools package. In Figure 3 of the track vertex for simulated electrons insimulated data have been reconstructed to extract CLAS12. Previous test point shown in green.vz , the z-component of the track vertex. The red Current test point and range shown in blue.curve is a fit to the region around the peak and thecentroid and range of the unit test are shown by the blue dot and vertical lines. The previous testevent is shown by the green dot at vz 3.84 cm, more than 8σ from the centroid of the peak.The reconstruction code had evolved so the old unit test returned a vz far from the expected valueof zero. This code is now part of the regular CLAS12 software distribution [6].Mr. A.Saina, another Surrey masters student is now completing a study of the reconstructionresolution of the CLAS12 software also working with Dr. Ziegler. The resolution here is measuredby swimming two simulated tracks through the CLAS12 Forward Detector. One track uses the6

known/generated vertex and initial 3-momentum from the event generator that was the input tothe simulation. The second track uses the reconstructed vertex and momentum produced by thereconstruction code itself. Differences between these two tracks at different detector subsystems areused to fill histograms and the width of the histogram is a measure of the reconstruction resolution.The effects of particle energy, particle type, torus field polarity, sector dependence, and geometrywere studied using two recent versions of the CLAS12, physics-based simulation code gemc.In Figure 4 we show the change of the resolution between gemc version 4.4.0 (yellow points) and4.3.2 (blue points) for θ and φ at fixed z. In upgrading gemc the generation of the subsystemsignals was made more realistic. The z-component here is in a special sector coordinate systemwhere x and y lie in the plane of the CLAS12 drift chambers, TOF panels, and calorimeters andthe z-axis is perpendicular to those planes. There is about a 50% increase in the widths of thedistributions at fixed z in the more recent version of gemc and the results are closer to what weobserve. This work will be the topic of a CLAS12-NOTE now in preparation.0.08Δθ [deg]Δϕ [deg]0.200.150.10gemc 4.3.20.05gemc d Z [cm]gemc 4.3.2gemc 4.4.0300400500600700Fixed Z [cm]Figure 4: Fit results for the widths of the φ (left) and θ (right) distributions for gemc version4.4.0 (yellow points) and 4.3.2 (blue points).2.1.3SVT AlignmentDuring the first half of the current grant period, the PI continued working on the track-basedalignment of the Silicon Vertex Tracker (SVT) in the Central Detector. This project built on thework of a Surrey Masters student, Peter Davies, who developed and validated the SVT geometrysoftware and made it consistent with the design drawing of the SVT [28]. These tools are still inuse. With the start of Run Group B and the arrival of deuterium data, the demands of the GnMproject prevented the PI from continuing. A description of the SVT work is in Ref. [29].2.1.4SummaryWith the arrival of the first deuterium date in Run Group B, the Richmond group has begunthe analysis of the GnM experiment with collaborators from Florida International University. Thepreliminary results on the neutron detection efficiency are encouraging. We continue to optimize theselection of e n and e p events that go into the ratio. Our University of Richmond undergraduatesand University of Surrey masters students are contributing to the software enterprise in the CLASCollaboration in general and specifically for the GnM analysis. Their work is described in twoCLAS12 NOTES and presentations at the annual DNP meeting [2, 5, 7, 8, 22, 26]. The geometryand alignment of the SVT project is far along.7

2.2Plan of WorkThe research effort here in nuclear physics is part of the program at the Thomas Jefferson NationalAccelerator Facility (JLab) in Newport News, VA. The primary goal of JLab is to unravel the quarkand gluon structure of protons, neutrons, and atomic nuclei and to understand how they emergefrom Quantum Chromodynamics (QCD). In this section we describe the experimental environmentand the proposed physics program. Gilfoyle will spend his sabbatical during 2022-2023 at JLab.JLab is a unique tool for basic research in nuclear physics. The central instrument is the Continuous Electron Beam Accelerator Facility (CEBAF); a superconducting electron accelerator with amaximum energy of 12 GeV, a 100% duty cycle, and a maximum current of 85 µA. There are fourexperimental halls (Halls A-D) that can collect data simultaneously and provide complementarycapabilities (see left-hand panel in Figure 5). CLAS12 in Hall B is a large particle detector withForward DetectorCentralDetectorBeamFigure 5: CEBAF layout (left) and CLAS12 design drawing (right).a toroidal, multi-gap magnetic spectrometer with large solid angle coverage at forward angles (theForward Detector) and a solenoid centered on the target for large angles (the Central Detector).See the right-hand panel in Figure 5. There are over 100,000 readout channels. The toroidal magnetic field in the Forward Detector is generated by six sectors of iron-free superconducting coils.The particle detection system in each sector consists of drift chambers [30] to measure chargedparticle trajectories, Cerenkov detectors [31] to identify electrons, pions, and kaons, scintillators[32] for time-of-flight measurements, and electromagnetic calorimeters [33]. The six segments areinstrumented individually to form six independent spectrometers. The Central Detector is builtaround a solenoid magnet with silicon and micromegas trackers, a time-of-flight system, and a central neutron detector. The CLAS Collaboration operates and manages CLAS12 with support fromthe US Department of Energy. The Richmond group has been part of the CLAS Collaborationsince its inception.Our focus for the next three years is (1) on the analysis of data already collected in RunGroups A and B in Hall B to measure GnM , (2) continued development of software to supportCLAS12 reconstruction and analysis, (3) participate in the Hall A GnM measurement as resourcesand time permit, (4) participate in the JLab experiment E12-06-117 Quark Propagation and HadronFormation, and (5) other activities.2.2.1Magnetic Form Factor of the NeutronWe now describe our program to measure the neutron magnetic form factor GnM at high Q2 .One of the central goals of nuclear physics now is to push our understanding of QCD into the8

non-perturbative region (see Sect 2.1 of the NSAC Long-Range Plan) [14]. Here, the nonlinearnature of QCD dominates and defies traditional mathematical solutions; forcing us to resort tophenomenological models, effective field theories, and the daunting numerical calculations of latticeQCD. Our understanding of the structure of the proton and neutron is still clouded. The neutronmagnetic form factor GnM is one of the fundamental quantities of nuclear physics and its evolutionwith Q2 characterizes the distribution of magnetization within the neutron. It is central to ourunderstanding of nucleon structure [14, 15, 16, 17, 34]. We are part of a broad campaign to measurethe four elastic nucleon form factors (electric and magnetic ones each for the proton and neutron)at JLab that includes seven experiments approved for running [17, 35].Gilfoyle is the spokesperson and contact person for JLab Experiment E12-07-104 which willmeasure GnM with the CLAS12 detector in Hall B and was approved by JLab PAC32 [20]. He isalso a co-spokesperson on a Hall A measurement of GnM E12-09-019. For the next budget periodour focus will be on the CLAS12 experiment. We propose to continue the analysis of data fromRun Groups A and B in Hall B to extract a precision measurement of the neutron magnetic formfactor. We are part of a collaboration with Dr. Brian Raue and his doctoral student Ms. LamyaBaashen at Florida International University (FIU). Our progress so far is described in Section 2.1.1and more details of the method are below.Here we present some background to the study of elastic electromagnetic form factors (EEFFs)and motivate their measurement. The most general form of the hadronic current for a nucleon is iσ µν qν2µ0µ2kj F2 (Q ) ν(p) check ν qσ(1)J ieν(p ) γ F1 (Q ) 2Mwhere M is the nucleon mass, kj with j p, n is the anomalous magnetic moment in units ofthe nuclear magneton, the ν and ν are the Dirac spinors, qν is the momentum transfer, andµN e /(2Mp )[36]. The Dirac and Pauli form factors are F1 (Q2 ) and F2 (Q2 ) respectively. Theseare routinely written in terms of Sachs form factors(p,n)GE(p,n) F1(p,n) τ F2(p,n)and GM(p,n) F1(p,n) F2(2)where τ Q2 /4M 2 .Measuring GnM and other elastic electromagnetic form factors (EEFFs) will decisively impactour understanding of the nucleon in the 12-GeV era. By measuring all four nucleon EEFF’s andinvoking charge symmetry the quark Dirac and Pauli form factors can be extracted in the followingway [37, 38].ppundnF1(2) 2F1(2) F1(2)andF1(2) 2F1(2) F1(2)(3)The result of this flavor decomposition is shown in the left-hand panel of Figure 6 [37]. There arelarge diffe

Here, we have collaborated with Dr. B.Raue from Florida International University (FIU) and his doctoral student Ms. L. Baashen. The Gn M analysis will be her thesis. An essential quantity in our analysis is the neutron detection e ciency (NDE) to provide an accurate measure of the number o