July 3, 2003

Measuring Neutrino-Nucleus Cross Sections at the Spallation Neutron Source

Submitted to the ORNL Physics Division by:
Fred Bertrand: feb@ornl.gov
Vince Cianciolo: cianciolotv@ornl.gov
Yuri Efremenko: efremenk@utkux.utcc.utk.edu
on behalf of the future SNS2 Collaboration

Table of Contents

I. Executive Summary

II. Scientific Motivation

III. Proposed enclosure and location

IV. Possible detectors

    A. Segmented detector

    B. Homogeneous detector

V. Schedule

VI. Collaboration

I. Executive Summary

The high beam intensity and short duty-factor of the Spallation Neutron Source (SNS) at ORNL would allow an exciting program to measure charged-current neutrino-nucleus cross sections, ZA(ne,e-)X. These cross sections are of great interest to the fields of nuclear structure physics and astrophysics; in particular, they would be of immediate use in the study of core collapse supernova physics. This is especially true due to the fortuitous similarity between the SNS neutrino spectrum and the neutrino spectrum produced in a supernova explosion as shown in Figure 1

With the termination of the neutrino programs at Los Alamos and at ISIS, there are no stopped-pion neutrino facilities, which are necessary to do such neutrino experiments. We propose to build at the SNS a shielded, instrumented enclosure which will host an initial set of two detectors in order to carry out a long term program of cross section measurements on a range of appropriate nuclear targets. The tentative name for the proposed facility is Supernova Neutrino Studies at the Spallation Neutron Source or SNS2

At present, only neutrino cross sections on 2H, 12C have been measured well (~4-10% errors). The only other reported results are for 2H, 56Fe and 127I (~40% errors). We show below that at SNS2 it is feasible to measure the charged-current neutrino-nucleus cross section for any selected nuclear target species to a statistical accuracy of 10% in one year. With clever detector designs that allow reuse of the detector with different target materials it becomes feasible to develop a true program to measure the cross sections for many nuclear species. Theoretical interest in these cross sections spans the entire range of nuclear species and must be tempered by the realization that neutrino cross sections are exceedingly small (~10-41 cm2) so that even with the prodigious neutrino flux at the SNSthe required target size is of order 10 tons. Therefore viable target materials must be naturally monoisotopic and affordable in large quantity. Based on these considerations an initial three-year program might consist of 10% measurements for the neutrino cross section on 56Fe, 209Bi, 12C, 16O. Future directions would be guided by the results of this initial program and by theoretical interest. For instance, total cross sections on additional targets; additional data on a smaller number of targets to allow for differential measurements as a function of energy and/or angle; or neutral-current measurements. The neutral current measurements, which are of significant theoretical interest due to the complete lack of experimental data to test theories, are perhaps possible at SNS2 given appropriate detector development. 

II. Scientific Motivation

A star more massive than 8-10 solar masses ends its life in a catastrophic explosion known as a core collapse supernova. These explosions are extremely important for galactic dynamical and chemical evolution because they disperse almost the entire mass of the star into interstellar space. Neutrino-nucleus weak interactions play a central role in both supernova dynamics and supernova nucleosynthesis. These interactions are also central to existing or proposed terrestrial facilities to detect neutrinos from the next Galactic or near-extra-Galactic supernova, which in turn will provide detailed neutrino ?light curves? from which supernova models and supernova nucleosynthesis models can be diagnosed and improved. Although many nuclei participate in these weak interactions, and it is impossible to measure all of the relevant cross sections, measurement of several key cross sections would provide invaluable checks on the elaborate theoretical models used in current computations.A new aspect of the supernova mechanism, the importance of electron-capture on nuclei heavier than Fe, has recently been recognized. No information is currently available for calculations, or even estimates, of these electron-capture rates. The proposed neutrino facility would enable measurements of the inverse reaction ZA(ne,e-)Z+1A, on e.g. 75As, 93Nb and 127I, in the relevant nuclear mass region. From these measurements, the electron-capture rates can be estimated. Fortunately the neutrino energy spectra available at the SNS are exactly what is needed for these measurements.

III. Proposed enclosure and location

The neutrino detector and shielding enclosure would be located in the SNS target hall.In discussions with the SNS Target Systems Division we have identified a mutually acceptable location on the north side of the beam line, at a mean distance of 21m from the spallation target, and at an angle of 160 degrees relative to the incoming proton beam direction. The available floor space is ~ 4.5m X 4.5m with a clear height of 6.5m. This is shown in Figure 2.

At full power (1.4 MW) the SNS will bombard its mercury target with a 1.1 mA, 1.3 GeV proton beam, producing~0.2 neutrinos per proton. The resulting neutrino flux through this volume will be ~2.5 X 107n/s/cm2, giving several tens of neutrino interactions per day for a ten ton detector. This must be compared with the cosmic ray muon (neutron) flux through this volume of ~2.5 X 108 (1.4 X 106) events per day. Such events must be suppressed through a combination of the SNS time structure, an active veto counter and shielding. 

The SNS time structure (~700 ns proton pulses at 60 Hz) allows us to eliminate a large amount of both types of cosmic backgrounds by turning off the detector except for the small fraction of time during which neutrinos can come from the target. Target neutrinos, which result from the p®m®n decay chain, will all arrive within several muon decay lifetimes (tm = 2.2 ms). This results in an active time of only 4 X 10-4 seconds for every second of machine operation, thus reducing the effective cosmic ray muon (neutron) flux through this volume to 1 X 105 (5.6 X 102) events per day. 

An active veto system can only provide rejection for the charged component of the cosmic background. We assume an achievable efficiency of 99% for this veto system which would further reduce the cosmic ray muon flux to 1 X 103 events per day. Most of the remaining muons can be rejected by their signature in the detector, but those which produce a neutron in the shielding can be confused with the desired signal. To a reasonable approximation we are only concerned with those muons which produce a neutron in the last interaction length of shielding. The observed m®n+X yield is 4X10-5 n/m/(g/cm2), giving a background rate of ~5.2 events/day coming from neutrons generated in the shielding enclosure by cosmic ray muons which failed to fire the active veto system. 

Without shielding the primary cosmic neutron flux is (as discussed above) ~560/day, two orders of magnitude above the irreducible background from cosmic muons. This sets the scale for the thickness of shielding: a steel enclosure, with a 1 m thick roof and 0.5 m thick walls will reduce the primary cosmic neutron flux by two orders of magnitude. Preliminary calculations, incorporating the SNS target and shielding assemblies and materials from nearby neutron scattering instruments show that this shielding is also sufficient to shield against SNS-generated neutrons.

Preliminary calculations show that the weight of the resulting shield package, plus the detectors is within the load limit of the SNS floor at our desired location. The resulting 3.5m X 3.5m X 5.5m interior space would be sufficient to house the active veto-counter system and one or two neutrino target/detector systems.


IV. Possible detectors

Several different types of detectors are being investigated; the two that appear most promising are:

IV-A. Segmented detector

Individual elements of the segmented detector would be composed of a position-sensitive gas proportional counter surrounded by a thin-walled cylindrical tube made of the target element. Signals would be read out from both sides of each individual channel to provide three-dimensional position information. Particle energy is reconstructed from the range of the particle track or by the total number of fired cells. Direction information can be extracted from the reconstruction of the track. In principle, this detector can be constructed in such a way that the detector elements are reusable; when a measurement with one target material is complete the target/detector combinations would be unstacked,the detector elements would be removed and loaded into a new set of target tubes, and the new target/tube combinations re-stacked. As a result, neutrino interactions with different nuclei can be studied with the same detector. This has great benefits for systematic error reduction as well as for long-term ease of operation and cost. From the list of nuclei of interest given above, cylindrical tubes of 9Be, 27Al, and 56Fe are easily obtainable; suitable tubes of many other target other elements can possibly be manufactured as powder contained in a plastic matrix.

Detectors with gas tubes as the sensitive elements have a number of advantages compared, for example, to those with scintillator rods. In general, gas tubes are less expensive and do not require an expensive light-readout system. In addition the low detector mass eliminates the necessity to subtract interactions in the target from the interactions in the detector itself. The energy resolution obtained by measurement of the track length is as good as that obtained by measurement of the visible energy deposition for electrons in the energy range of a few tens of MeV.

Monte Carlo studies show that a reasonable design for the target tubes would be 1.0 cm diameter tubes with a wall thickness of 0.5 mm. An Fe target with a 10 ton fiducial mass would be 1.85 m X 1.85 m and would have 2.5 m long tubes. The total target/detector volume should include an additional 20 cm beyond this fiducial volume such that accepted events are totally contained.. This results in a size of 2.9 m X 2.25 m X 2.25 m - or a total of 50,600 2.9 m target/detector tubes. For an Fe target with 10 ton fiducial mass, at a mean distance of 21 m from the SNS spallation target, and with an expected cross section of ~2.6X10-40 cm2, the expected reaction rate is 61 per day.We estimate a detector efficiency of 30%, leading to a signal rate of 18 neutrinos per day or about 3600 per year. 

IV-B. Homogeneous detector

Some nuclei are difficult or impossible to obtain as solid compounds. It is more efficient to measure such targets (e.g., 2H, 12C, 16O, 127I) in the form of a liquid or an aqueous solution. Therefore we propose to have a second, homogeneous detector that can be filled with various liquids. The detector would be built as a steel, spherical or rectangular light tight vessel, with a size of ~27 m3Scintillator or Cherenkov light would be detected by ~300 8? PMTs (or other type of photodetectors) mounted on the inner wall. 

The first target for this detector would likely be 12C (liquid scintillator).The 12C cross section has been previously measured by both KARMEN and LSND, but better data are desired. For SNS2 the expected rate is 10 events per day assuming a detection efficiency of 50%, or 2000 events per year. In one year this is four times the total number ofevents measured by KARMEN. An additional benefit of an early 12C measurement is that it will provide a calibration of the SNS neutrino flux. This is critical to allow a systematic accuracy comparable to our statistical precision.

V. Schedule

We expect to submit a letter of intent discussing the physics motivation of the proposed neutrino facility and requesting the identified floor space to SNS management by August 2003. Preparation of this document is well along and prior to submission we will be hosting a workshop/collaboration meeting. By January 2004, we expect to have a proposal submitted to the Nuclear Physics Division of the DOE. 

Our experimental schedule is guided by the expectation of full-power beam at the SNS by the end of 2008 - at that point we could efficiently begin commissioning our detector. In order to achieve this we would need to perform detailed design studies and some R&D in FY04-06; begin detector construction in FY06; erect the shielding enclosure in FY07; and install the detector in FY08.


VI. Collaboration

It is our intention that SNS2 will operate as a user facility for the neutrino community and that experimental priority will be set by an independent Program Advisory Committee.
The following institutions are currently a part of the SNS2 collaboration:

Oak Ridge National Laboratory

University of Tennessee

University of Alabama

University of South Carolina

and discussions are actively under way with additional collaborators. We anticipate that this collaboration will be open to all users, and we welcome interested parties to join.


SNS Neutrino Spectrum
Supernova Neutrino Spectrum

Figure 1. Energy spectrum of neutrinos produced by the SNS (top) and by a supernova (bottom).

Detector Location

Figure 2. Proposed SNS2 facility location at an angle of 160 degrees relative to the incoming proton direction and a distance of 21 m from the neutrino source.