Jefferson Lab Frozen Spin Target

maintained by: Chris Keith,

The Jefferson Lab Frozen Spin Target (FroST) is a nuclear-spin polarized target utilized for scattering experiments with a tagged photon beam inside the CEBAF Large Acceptance Spectrometer (CLAS).

Protons in frozen, 1.5 mm beads of butanol are highly polarized via a technique called Dynamic Nuclear Polarization (DNP): the off-center saturation of the Electron Spin Resonance (ESR) frequency of paramagnetic radicals dissolved within the butanol. The DNP process is performed at a "moderate" temperature of approximately 0.3 Kelvin while inside an external, 5.0 Tesla "polarizing" magnet. Under these conditions the electronic spins of the paramagnetic radicals are completely polarized, and saturating the sample with microwaves near the ESR frequency effectively transfers the electrons' polarization to the nuclear spins.

A bespoke, horizontal 3He/4He dilution refrigerator is used to cool the target material during and after the DNP process, while a microwave generator is used to saturate the ESR frequency of the paramagnetic radicals (approximately 140 GHz at 5 Tesla). The nuclear spins can be polarized either parallel or anti-parallel to the direction of the magnetic field, depending on whether the microwave frequency is slightly below or above the ESR frequency. During tests the dilution refrigerator achieved a base temperature of 26 milliKelvin (mK). In the experimental hall, where vibration is more problematic, a base temperature of 28 mK was observed. Cooling power at 50 mK is about 800 microwatts, and between 50 and 100 milliwatts at 300 mK.

When the ultimate target polarization is reached, the microwave generator is switched off, and the refrigerator cools the target beads to a temperature of about 30 mK. At such low temperatures, the polarization of the protons decays very slowly. In other words, the spins are "frozen" (hence the name). This allows a weaker magnetic field to be used for "holding" the polarization. The holding field is generated by a 0.56 Tesla superconducting solenoid mounted inside the Frozen Spin cryostat. Unlike the polarizing magnet, this solenoid is thin enough for scattered particles to pass through and be detected by the CLAS spectrometer. However its magnetic field is strong enough to keep the polarization decay at an acceptibly low rate (less than 1% per day). Both the polarizing and solenoidal holding magnets produce a field along the direction of the photon beam (longitudinal field direction). However, we have recently (Feb. 2008) tested a four-layer, superconducting dipole magnet that produces a 0.51 Tesla field perpendicular to the beam (transverse field direction). This magnet will be used for the second round of FROST experiments, scheduled for 2010.

The polarization process begins by inserting the target cryostat into the warm, horizontal bore of the 5 Tesla polarizing magnet and energizing the microwave generator. Although the target polarization increases rapidly at first, two to three hours are required to reach a polarization of 80%. During this time the microwaves warm the target to about 0.3 K. After the microwave generator is switched off, 30-45 minutes are required for the target to cool into the so-called "Frozen Spin Mode", denoted by a target temperature less than 50 mK. At that time the polarizing magnet is de-energized while the holding magnet is simultaneously energized to 0.56 Tesla (another 45 minutes). Next the Frozen Spin cryostat is removed from the polarizing magnet and moved 4 meters into the center of CLAS (about 2 minutes). Finally the tagged photon beam is activated and scattering data is acquired for a period of 5-10 days after which the polarization process is repeated (usually to reverse the target polarization). The photon beam deposits about 10 microwatts to the refrigerator, warming it 2 mK or so. Even with the beam on target, the polarization loss remains at 1-1.5% per day.

Two NMR coils and Q-meters are used to measure the target polarization. The first is tuned to 212 MHz and is used during the DNP process at 5 Tesla. The second, at 24 MHz, is used at 0.56 Tesla when the target is operating in the Frozen Spin mode.

In 2010 the Frozen Spin Target was utilized for the second set of experiments inside CLAS. In this case the longitudinal holding coil was replaced with a superconducting, racetrack-style dipole magnet that produced a 0.50 T field perpendicular to the incident beam direction. Minor modifications were made to the heat exchangers inside the dilution unit that resulted in a lower target temperature of about 24--25 mK. As a result, 1/e depolarization times up to 4000 hours were obtained without beam and 3200 hours with beam, or less than 0.5% polarization loss per day. Also, the maximum substainable He-3 circulation rate was increased from about 16 mmol/s to above 30 mmol/s.

- Polarization: > 80%
- Base temperature: 25 mK
- Holding field: 0.56 Tesla (longitudinal), 0.50 Tesla (transverse)
- Relaxation rate: 0.5% per day (positive pol.), 1.0% per day (negative pol.)

Documents (Includes Hazard Analysis)
Document 1: 3He Pump and Purge Procedure (pdf)
Document 2: Cooling the 4He (pdf) (includes 4He pump/purge)
Document 3: Condensing Pure 4He in the Dilution Unit (pdf)
Document 4: How to Remove Pure 4He from Circulation (pdf)
Document 5: Condensing the 3He/4He in the Dilution Unit (pdf)
Document 6: How to Remove the 3He/4He Mash (pdf)
Document 7: Swapping the LN2 Traps (pdf)
Document 8: How to Warm the Target to Room Temperature (pdf)
Document 9: How to Insert the Frozen Spin Target Stick (pdf)
Document 10: Removing the Frozen Spin Target Insert (pdf)

2004 JLab Seminar: Freezing the Spin of the Proton (pdf)
2005 Invited Talk: European Workshop on Solid Polarized Targets (pdf)
2006 Hall B Seminar: Frozen Spin Targets in a Nutshell, Part 1 (pdf)
2006 Hall B Seminar: Frozen Spin Targets in a Nutshell, Part 2 (pdf)
2007 MCC Presentation: FROST: The JLab Frozen Spin Target(pdf)
2008 Core Managers Meeting Seminar: The JLab Frozen Spin Target(pdf)
2008 Hall B Seminar: Frozen Spin Targets in a Nutshell, Part 3: RESULTS(pdf)
2008 JLab Science and Technology Review (pdf)
2008 SPIN2008: The JLab Frozen Spin Target (pdf)

Pictures and Drawings
Frozen Spin Poster (pdf)
Flow Diagram, 2007 (pdf)
Cryostat Wiring Diagram, 2007 (pdf)
3-D Construction Drawings

Additional Links
JLab Webpage for Frozen Spin Project (out of date)
JLab Wiki for the Frozen Spin Project
GWU Webpage for Frozen Spin Experiments (secure)