We report on the frozen-spin polarized hydrogen–deuteride (HD) targets for photoproduction experiments at SPring-8/LEPS. Pure HD gas with a small amount of ortho-H2 (∼0.1%) and a very small amount of para-D2 (∼0.001%) was liquefied and solidified by liquid helium. The temperature of the produced solid HD was reduced to about 30 mK with a dilution refrigerator. A magnetic field (17 T) was applied to the HD to grow the polarization with the static method. After the aging of the HD at low temperatures in the presence of a high-magnetic field strength for three months, the polarization froze. Almost all ortho-H2 molecules were converted to para-H2 molecules. Most remaining para-D2 molecules were converted to ortho-D2 molecules. The para-H2 and ortho-D2 molecules exhibited weak spin interactions with the HD. If the concentrations of the ortho-H2 and para-D2 were reduced appropriately at the beginning of the aging process, the aging time can be shortened. We have developed a new nuclear magnetic resonance (NMR) system to measure the relaxation times (T1) of the 1H and 2H nuclei with two frequency sweeps at the respective frequencies of 726 MHz and 111 MHz and succeeded in the monitoring of the polarization build-up at decreasing temperatures from 600 mK to 30 mK at 17 T. Automatic NMR measurements with the frequency sweeps enabled us to omit the use of a manual tuning circuit and to remove magnetic field sweeps with eddy current heat. This technique enables us to optimize the concentration of the ortho-H2 and to efficiently polarize the HD target within a shortened aging time.

We have been carrying out photoproduction experiments at the Laser Electron Photon beamline at SPring-8/LEPS since 2000.1 Linearly or circularly polarized photon beams in the energy range of 1.5 GeV–3.0 GeV are produced by the backward Compton scattering of an ultraviolet laser from 8 GeV electrons.2 Photoproduction of various mesons and baryons, such as ϕ,3,4π,5K*,6 and hyperons,7–11 was studied with unpolarized liquid hydrogen or deuterium targets. If a polarized nucleon target is introduced for the LEPS experiments, new types of experiments can be realized to measure double-spin asymmetries that provide precious knowledge to the understanding of the hadron structure, its production mechanism, and the existence of exotic particles.

Honig suggested the use of the hydrogen–deuteride (HD) molecule as a frozen-spin polarized target after the important work on the relaxation mechanism.12 The HD target had been developed at Syracuse,13 at the Brookhaven National Laboratory (BNL),14–17 and at ORSAY,18,19 and was used for physics experiments at the BNL20 and the Jefferson Lab (JLAB)21,22 by Sandorfi et al. We started the development of the HD target at Osaka University in 2005.23–25 

Given that the purity of commercially available HD gas is ∼96%, the HD gas is purified up to 99.99% using a distiller26 and is analyzed using a gas chromatograph with a quadrupole mass spectrometer.27 Pure H2 gas with an amount of ∼0.1% is used as the catalyst and is added to the purified HD gas to achieve long relaxation times. The HD gas is liquefied and solidified by liquid helium and the solid HD is cooled down to ∼20 mK–30 mK with a 3He–4He dilution refrigerator. A 17 T magnetic field is generated by a superconducting solenoid and is applied to build-up the HD polarization. The use of the static nuclear polarization at low temperatures and at a high-magnetic field strength for a three-month period leads to the generation and freezing of the HD spin polarization. The H2 gas is composed of ortho-H2 (o-H2) with a spin of S = 1 and para-H2 (p-H2) with a spin of S = 0. In addition, the population ratio of o-H2 to p-H2 is 3:1 at room temperature. The o-H2 molecules have a decay time of ∼1 week at low temperatures and generate ∼2 μW heat at the beginning of the aging process. The p-D2 molecules have a decay time of about 18 days. At the end of the aging period, almost all the o-H2 and p-D2 molecules are converted to the p-H2 and o-D2 molecules, respectively. The latter molecules engage in weak spin interactions with the HD. We obtained a relaxation time of ∼8 ± 2 months for the 1H nucleus, which was adequately long for the conduct of the planned experiments at SPring-8.28 Once the spin polarization is frozen at Osaka University, the temperature can be increased and the magnetic field can be decreased. The HD target is transported to SPring-8 at a temperature of 1.5 K at a magnetic field of 1 T for photoproduction experiments.

In the past, the calibration of the 1H polarization was carried out based on Nuclear Magnetic Resonance (NMR) measurements with magnetic field sweeps at ∼1 T with a frequency of 40 MHz at 4.2 K.29 After the aging of the HD target at 17 T for three months, the magnetic field was decreased to 1 T and the 1H polarization was obtained based on the estimation of the ratio of the area of the final NMR signal to that of the calibration signal. A three-month aging period is very long and the consumption of liquid helium at a rate of ∼24 l/day for the operation of the dilution refrigerator and the superconducting solenoid is costly. Most of the liquid helium can be supplied by the Low Temperature Center of Osaka University. To ensure smooth operations, we purchased commercial liquid helium for use during long holiday periods. Evaporated helium gas was returned to the Low Temperature Center for recycling.

If the concentrations of the o-H2 and p-D2 are decreased at the beginning of the aging, the aging time can be shortened. Shortening the aging time is very important for decreasing the costly liquid helium consumption and for avoiding power, mechanical, and cryogenic failures during the aging. The relaxation time of the solid HD depends on the concentrations of the o-H2 and p-D2, temperature, magnetic field, and so on. The relaxation time of the solid HD was measured in the temperature range of 1.2 K–4.2 K at various concentrations of the o-H2 and p-D2.13,32–35 However, no measurements were conducted at temperatures lower than 1 K and at magnetic fields higher than 10 T. Although it was necessary to measure the relaxation time of the HD polarization within the temperature range below 1 K at 17 T for the optimization of the concentrations of the o-H2 and p-D2, there were technical difficulties in monitoring the build-up of the polarization by NMR measurements at high-frequencies of ∼700 MHz. To overcome these difficulties, we developed a new NMR system that could be operated within a broad frequency range up to 726 MHz. Some brief explanations have been reported elsewhere.36 

We measured the polarization of the 1H and 2H nuclei at Osaka University and at SPring-8. In order to make reliable calibration for the polarization, it was necessary to use the same NMR system at both sites. However, the weight of the conventional NMR system was 80 kg, and frequent transportation of the system (which was mounted on a rack with a height of 2 m) between Osaka University and SPring-8 was not easy. We constructed a portable NMR system with an operating software system with PCI eXtensions for Instrumentation (PXI), as shown in Fig. 1.29 The weight of the portable NMR system was only 7 kg and the cost was reduced to 25%.

FIG. 1.

Schematic of the portable Nuclear Magnetic Resonance (NMR) system. The single-coil method was applied in conjunction with the use of magnetic field sweeps.

FIG. 1.

Schematic of the portable Nuclear Magnetic Resonance (NMR) system. The single-coil method was applied in conjunction with the use of magnetic field sweeps.

Close modal

The portable NMR system consisted of PXI-1036 (chassis), PXI-8360 (connection between laptop PC and PXI), PXI-5404 (signal generator), and PXI-5142 (ADC), which were developed by the National Instruments Company. This system was controlled by a LabVIEW program on the laptop. The frequency range of the PXI-5404 ranged from 0 MHz to 100 MHz and was suitable for NMR measurements of the 1H nucleus with magnetic field sweeps at a field strength of ∼1 T. The signal-to-noise (S/N) ratio of the portable NMR system depended on the performance of the laptop that was used to operate it.

Two coaxial cables were used between the tuning circuit and the coil. One was a semi-flexible coaxial cable (MULTIFLEX 141 produced by Suhner) with a length of 2 m and a diameter of 4.2 mm used at room temperature. The capacitance of the cable is 95.0 pF/m. The other is a semi-rigid coaxial cable (SC-119/50-SCN-CN produced by Coax Co., Ltd.) with a length of 2.5 m and a diameter of 1.19 mm. The capacitance of the cable is 95.2 pF/m. The latter cable was used at low temperatures on the dilution refrigerator and has thermal anchors at 4.2 K (liquid helium), 1.5 K (1 K-pot), 0.6 K (still), and the mixing chamber. The height of the thermal anchor at 4.2 K changed by 0.12 m/day depending on the liquid helium level. The self-inductance of the coil was 0.893 μH.

We developed a new NMR system that operated in a wide frequency range up to 726 MHz. Given that the polarization measurement was performed during the aging of the HD target at 17 T, the superconducting solenoid was operated with a persistent current mode, and a frequency sweep method was applied for the polarization measurements.

The signal generator PXI-5404 was replaced with PXIe-5650 that generated radiofrequency (RF) signals at frequencies up to 1.3 GHz. The analog-to-digital converter (ADC) PXI-5142 was also replaced with PXIe-5162 that sampled the data at different rates up to 5 G sample/s. Instead of the single-coil method used in the previous NMR system,29 the crossed-coil method, which had been used for tens of years by Syracuse University,13 BNL,14–17 ORSAY,18,19 and JLAB,21,22 was applied, as shown in Fig. 2. Two sets of the coaxial cables written above were used. The self-inductances of the coils were 0.876 μH (input) and 0.893 μH (output), and the mutual inductance between the crossed-coils was 4 nH. The capacitive coupling of the crossed-coils was 4.2 pF. Given that the tuning circuit in the previous NMR system specified the frequency and required manual operations for each nucleus, the tuning circuit was not used. Although the measurements without the tuning circuit resulted in poor S/N ratios, automatic NMR measurements within wide frequency ranges were performed.

FIG. 2.

Schematic of the new NMR system. The crossed-coil method and frequency sweeps were applied.

FIG. 2.

Schematic of the new NMR system. The crossed-coil method and frequency sweeps were applied.

Close modal

Figure 3 shows the HD target cell and NMR coils attached to the bottom of the DRS2500 dilution refrigerator. The HD target cell and the support frame of the NMR coils were made of Kel-F [Poly-Chloro-Tri-Fluoro-Ethylene (PCTFE)] that did not contain any hydrogen. A Teflon-coated silver wire with a diameter of 0.3 mm was used for input signals.

FIG. 3.

(a) Mixing chamber of the DRS2500 dilution refrigerator with the cold finger surrounded by the superconducting solenoid made of NbTi and Nb3Sn. (b) Cross-section of the target cell and support frame of the NMR coils. (c) Structure of the support frame of the NMR coils with the directions of the magnetic field H0 of the superconducting solenoid and the applied radiofrequency (RF) field H1.

FIG. 3.

(a) Mixing chamber of the DRS2500 dilution refrigerator with the cold finger surrounded by the superconducting solenoid made of NbTi and Nb3Sn. (b) Cross-section of the target cell and support frame of the NMR coils. (c) Structure of the support frame of the NMR coils with the directions of the magnetic field H0 of the superconducting solenoid and the applied radiofrequency (RF) field H1.

Close modal

It was wound to form a single-turn saddle coil on the support frame. The other wire was also wound on the support frame in the perpendicular direction and served as the crossed-coil for picking up output signals. The RF field of H1 was produced by the coil for input signals, and the direction of H1 was perpendicular to that of the magnetic field H0 produced in the superconducting solenoid. The diameter of the support frame was 38 mm and that of the external thermal shield made of copper was 56 mm. The effect of the external metal was estimated to be about 40%. Thin aluminum wires (20% in weight of the HD target) with a purity higher than 99.999% were soldered on the target cell to insure the cooling of the solid HD.

We used a 3He–4He dilution refrigerator (DRS2500) produced by Leiden Cryogenics B.V.30 in The Netherlands to cool the HD target. The DRS2500 refrigerator has a lowest temperature of 6 mK and a cooling power of 2500 μW at 120 mK. A strong magnetic field was produced by the superconducting solenoid (NbTi and Nb3Sn) produced by Japan Superconductor Technology, Inc.31 in Japan. The target cell was attached to a cold finger made of pure copper (99.99%) with a length of 500 mm. In turn, the cold finger was attached to the mixing chamber with a lowest temperature, as shown in Fig. 3(a). A carbon resistance thermo sensor was used to measure the temperature of the mixing chamber where the magnetic field was reduced to be about 300 gauss.

The DRS2500 refrigerator and superconducting solenoid were precooled to 77 K by liquid nitrogen. After the liquid nitrogen was blown out, liquid helium cooled them to 4.2 K. The HD gas (1 mol), that had an o-H2 impurity of 0.3% and a p-D2 impurity smaller than 0.001%, was liquefied at ∼20 K and solidified at 4.2 K in the target cell. The superconducting solenoid was excited with a current of 270 A to produce a magnetic field strength of 17 T, and the operation was changed to the persistent current mode. The stability of the magnetic field is better than 1 ppm/h with the persistent current mode. NMR measurements with frequency sweeps for 1H, 2H, and 19F nuclei were initiated. The 1H and 2H nuclei were the main components of the HD target, and the 19F nucleus was contained in the target cell and support frame of the NMR coils. The NMR frequencies for the 1H, 2H, and 19F nuclei were 726 MHz, 111 MHz, and 683 MHz, respectively, at 17 T. Since slow frequency sweeps caused non-negligible background shifts, the speed of the frequency sweeps was set to 0.544 MHz/s. We accumulated 100 k data points and estimated the average values at each frequency point.

The temperature of the target decreased to 600 mK when the 1 K pot of the DRS2500 was pumped and 3He gas was liquefied into the mixing chamber. When 4He gas was also liquefied into the mixing chamber, the temperature of the HD target became lower than 100 mK and gradually dropped down to 30 mK. A heater (power of 0.09 W) was applied to increase the flow of the circulating 3He gas.

HD molecules do not have symmetry restrictions on their total wave function since the two particles are distinguishable. As a result, all HD molecules are in the lowest rotational energy state with L = 0. The L = 0 state leads to infinite spin–lattice relaxation time; therefore, there is no way to polarize pure HD molecules with the static method.

On the other hand, H2 and D2 molecules are indistinguishable under interchange of nuclei. This requires their total wave function to be anti-symmetric for H2 (Fermi particles) and symmetric for D2 (Bose particles). Table I summarizes the symmetries for the molecules of o-H2, p-H2, p-D2, and o-D2.

TABLE I.

The spin S, rotator L, and symmetries for the molecules of o-H2, p-H2, p-D2, and o-D2.

MoleculeSS symmetryLL symmetryTotal symmetry
o-H2 Symmetric Anti-symmetric Anti-symmetric 
p-H2 Anti-symmetric Symmetric Anti-symmetric 
p-D2 Anti-symmetric Anti-symmetric Symmetric 
o-D2 Symmetric Symmetric Symmetric 
MoleculeSS symmetryLL symmetryTotal symmetry
o-H2 Symmetric Anti-symmetric Anti-symmetric 
p-H2 Anti-symmetric Symmetric Anti-symmetric 
p-D2 Anti-symmetric Anti-symmetric Symmetric 
o-D2 Symmetric Symmetric Symmetric 

Because of their L = 1 molecular rotation states, o-H2 and p-D2 impurities relax to the lattice host efficiently, switching on the relaxation of HD spins to the lattice through spin–spin interaction with o-H2 and p-D2 spins. Then 1H and 2H polarizations build up and distribute throughout the HD sample by spin diffusion. The process of polarization build-up is related not only to spin–lattice relaxation but also to complex spin–spin relaxation. By aging the target at the polarizing field and temperature conditions, the decay to p-H2 and o-D2 states is used as a switch off mechanism to freeze the HD polarization.

The o-H2 state decays to the p-H2 state with a life time of 6.4 days and heat release of 2.6 mW/mol, while the p-D2 state decays to the o-D2 state with a life time of 18.2 days and heat release of 0.45 mW/mol. After the aging process of 3 months, the heat release is negligibly small and the HD polarization becomes frozen.

The nuclear spins of 1H, 2H, and 19F nuclei are 1/2, 1, and 1/2, respectively. If the population distribution of the spin system obeys the Boltzmann statistics, the polarizations of the 1H, 2H, and 19F nuclei at the thermal equilibrium state at 17 T are calculated, as shown in Fig. 4. Given that the magnetic moment of 19F is close to and slightly smaller than that of 1H, these two nuclei have similar polarizations. The 2H nucleus has a smaller magnetic moment than those of the 1H and 19F nuclei and yields a smaller 2H polarization.

FIG. 4.

Polarization of 1H (solid curve), 2H (dashed curve), and 19F (dotted curve) nuclei calculated for the magnetic field of 17 T in the thermal equilibrium state.

FIG. 4.

Polarization of 1H (solid curve), 2H (dashed curve), and 19F (dotted curve) nuclei calculated for the magnetic field of 17 T in the thermal equilibrium state.

Close modal

Figure 5 shows NMR signals of the 1H nucleus measured before and after the aging process based on the use of the portable NMR system (Fig. 1). The NMR signals before the aging were measured at 4.2 K and 0.87 T. Since the temperature of 4.2 K was easily produced by liquid helium at the atmospheric pressure, the calibration at 4.2 K was reliable. The NMR signals of the 1H nucleus were measured again at 0.3 K and 0.87 T for comparison after the aging process at temperatures of about 20 mK at 17 T over a 3-month period. The speed of the magnetic field sweeps can be changed from 0.0085 T/min to 0.2 T/min, and the NMR signals were measured with a speed of 0.034 T/min not to generate much eddy current heat. The NMR signals measured after the aging were ∼2000 times larger than those measured before the aging process. The 1H polarization was obtained as 44 ± 1% where the uncertainty of 1% originated from the uncertainty in the background subtraction at 4.2 K and 0.87 T. Taking calibration data at a temperature lower than 4.2 K may give more precise polarization because the background subtraction is easy and the NMR system has better linearity. Figure 5(e) shows the relaxation of the 1H polarization at 0.3 K and 1 T after the aging process. The polarization data were fitted with an exponential function, and the relaxation time of 8 ± 2 months was obtained. The relaxation time is much longer than that obtained by ORSAY32 and enables us to take more physics data at SPring-8.

FIG. 5.

NMR signals of the 1H nucleus measured with the portable NMR system (Fig. 1) inside the DRS2500 dilution refrigerator before [(a) and (b)] and after [(c) and (d)] the aging process. The peak shift of about 5 gauss was caused by magnetic hysteresis. (e) The relaxation of the polarization after the aging process.

FIG. 5.

NMR signals of the 1H nucleus measured with the portable NMR system (Fig. 1) inside the DRS2500 dilution refrigerator before [(a) and (b)] and after [(c) and (d)] the aging process. The peak shift of about 5 gauss was caused by magnetic hysteresis. (e) The relaxation of the polarization after the aging process.

Close modal

Figure 6 shows the circuit gain as a function of the frequency measured by the new NMR system. The frequency was changed from 1 MHz to 1000 MHz for the input coil and the response was measured by the output coil. During the aging of the HD target at 17 T, the 1H NMR signals were measured with frequency sweeps around 726 MHz, where large gains were obtained. However, gains were not large in the frequencies around 111 MHz for 2H and 683 MHz for 19F. Since NMR measurements for the 1H, 2H, and 19F nuclei were carried out with automatic frequency sweeps, tuning the NMR circuit for a particular frequency was impossible in the present experiment.

FIG. 6.

The circuit gain (Q-curve) as a function of the frequency for the new NMR system at 17 T.

FIG. 6.

The circuit gain (Q-curve) as a function of the frequency for the new NMR system at 17 T.

Close modal

NMR signals measured with frequency sweeps at 4.2 K and 30 mK at 17 T based on the use of the new NMR system (Fig. 2) are shown in Fig. 7. Phase shifts for the separation of the real and imaginary signals were determined to minimize χ2 in the fitting to the imaginary signals with a Gaussian function. The widths of the NMR signals of 1H and 2H nuclei are determined by the field homogeneity of about 10−4 in the target.

FIG. 7.

NMR signals of the 1H, 2H, and 19F nuclei measured inside the DRS2500 dilution refrigerator at 4.2 K (left) and 30 mK (right) at 17 T with the new NMR system (Fig. 2).

FIG. 7.

NMR signals of the 1H, 2H, and 19F nuclei measured inside the DRS2500 dilution refrigerator at 4.2 K (left) and 30 mK (right) at 17 T with the new NMR system (Fig. 2).

Close modal

The signals of the 1H and 2H nuclei were small and the S/N ratios were poor at 4.2 K. Given that the number of 19F nuclei was much larger than those of the 1H and 2H nuclei, and given that the 19F nuclei were located near the NMR coils where the coil sensitivity was high, the 19F signals were clearly observed even at 4.2 K. The intensities of the 1H and 2H signals became larger by ∼100 and 30 times, respectively, when the temperature decreased from 4.2 K to 30 mK. The intensities of the 19F signals increased by a factor of ∼3 at 30 mK. It was considered that the deterioration of the 1H and 19F signals at 30 mK was caused by the high-frequency signal detection difficulties. The 1H and 19F signals were clearly observed at magnetic fields below 7 T; however, these deteriorated at field strengths above 7 T.

Only the central regions of the NMR imaginary signals in Fig. 7 were fitted with a Gaussian function, and the signal heights were obtained. The temperature of the mixing chamber and the build-up of the polarization of the 1H, 2H, and 19F nuclei are shown in Fig. 8. We succeeded in monitoring the build-up of the polarization during the aging process of the HD target at 17 T.

FIG. 8.

(a) Temperature of the mixing chamber measured by a carbon resistance thermometer. The NMR signal heights of (b) 1H, (c) 2H, and (d) 19F nuclei at the beginning of the aging of the HD target at 17 T. The solid curves are the results of the fits obtained with the use of function (3).

FIG. 8.

(a) Temperature of the mixing chamber measured by a carbon resistance thermometer. The NMR signal heights of (b) 1H, (c) 2H, and (d) 19F nuclei at the beginning of the aging of the HD target at 17 T. The solid curves are the results of the fits obtained with the use of function (3).

Close modal

The 3He condensation was completed at 0 h, while the 4He condensation was initiated at 2.5 h and was completed at 9 h. The temperature of the mixing chamber decreased from 600 mK to 30 mK during the 4He condensation. The 1H polarization started to grow at 9 h. The polarization grew up to its maximum value within 1 day. The 2H polarization also started to grow at 9 h. The speed of the growth of the 2H polarization was slower than that of the 1H polarization. Both the 1H and 2H polarizations became ∼10 times larger at 30 mK than those at 600 mK. The 19F polarization became larger by only 10% when the temperature decreased from 600 mK to 30 mK. Given that the NMR signals of 19F at 30 mK are larger than those at 4.2 K by ∼3 times, the actual temperature was estimated to be approximately equal to 1.5 K. The thermal conductivity of Kel-F at 500 mK was ∼2 × 10−5 W/cm K and was adequately large for cooling the NMR support frame.37 Insufficient cooling of the NMR support frame was inferred owing to the poor thermal conductivity between the support frame and the cold finger. Grease should be added to the screws for better thermal conductivity in the next cooling attempt.

The NMR signal heights within the range of 9 h–25 h were fitted by the function

P=P0(1exp(tt0T1)),
(1)

where P0, t0, and T1 are free parameters. The T1 values of the 1H and 2H nuclei at 30 mK–600 mK at 17 T were 2.96 ± 0.03 h and 7.72 ± 0.72 h, respectively. By fitting the data with narrow regions, the T1 value of the 1H nucleus increased as time elapsed. Given that the concentration of o-H2 did not decrease so fast, it was considered that the prolonged value of T1 was inferred to be caused by the low temperature.

We also carried out data analyses based on the areas of the NMR peak regions that were estimated by integration. As a result of this analysis, the relaxation times of the 1H and 2H nuclei were 3.69 ± 0.03 h and 7.90 ± 0.12 h, respectively. The uncertainties due to the selection of the integration range were about 0.22 h and 0.03 h for the 1H and 2H nuclei, respectively.

The non-linearity of the preamplifier and ADC was checked and was found to be smaller than 2.8%, which gave the systematic uncertainties of the relaxation times of 0.03 h and 0.06 h at most for the 1H and 2H nuclei, respectively. Since the 1H signals are about 10 times larger than the 2H signals, as shown in Fig. 8, the non-linearity of all the NMR circuits, if it exists, affects the 1H signals strongly. However, temperatures estimated from the 1H signal heights and those from the 2H signal heights are not largely different. The systematic uncertainties caused by the non-linearity of the NMR circuit are estimated from these temperature differences and are found to be 0.21 h and 0.17 h for the 1H and 2H nuclei, respectively.

It is questionable if the cold finger made of copper with a weight of 3.7 kg and the HD target have the same temperature as the mixing chamber whose temperature was monitored during the aging. The 63Cu nucleus has a spin of 3/2 and polarizing all Cu nuclei of the cold finger and target cell needed part of cooling power of the DRS2500 dilution refrigerator and delayed the cooling of the HD target.38 However, we found that the 2H relaxation time was longer than the 1H relaxation time, as shown in Fig. 8. Since the 1H nuclei have the same temperature as the 2H nuclei, different relaxation times between 1H and 2H prove that the cooling of the cold finger and the HD target cell was faster than the build-up of the target polarization. In previous studies, the same results that the 2H nuclei had longer relaxation time than the 1H nuclei were obtained at 1.8 K and 0.85 T.34 At the beginning of the aging process, we used H2 as a catalyst but did not use D2 in the HD gas, which may have led to a longer 2H relaxation time.

The NMR data were measured 12 days after the liquefaction and solidification of the HD. The o-H2 concentration was estimated to decrease from 0.3% to 0.05%, and the p-D2 concentration was estimated to decrease from 0.001% to 0.0005% during the measurements. Given that the relaxation times of the 1H and 2H nuclei were found to be short enough, the o-H2 concentration of 0.05% should be reduced to shorten the aging time of the HD target. After the aging process of 3 months, the o-H2 concentration will be expected to decrease to 3.6 × 10−8%, and the p-D2 concentration will be 3.7 × 10−6%. If remaining p-D2 concentration has a strong effect on the relaxation times after the aging process, we would decrease it by some HD gas distillations26 before the aging process.

In order to optimize the amounts of o-H2 and p-D2 in the HD target and the aging time, we developed a new NMR system and succeeded in the monitoring of the build-up of the polarizations of the 1H and 2H nuclei at 17 T. The polarizations were found to grow within one day when the temperature decreased from 600 mK to 30 mK. The relaxation times of the 1H and 2H nuclei at 30 mK–600 mK and 17 T were obtained as 2.96 ± 0.03 (stat.) ± 0.73 (syst.) and 7.72 ± 0.72 (stat)±0.18 (syst.) h, respectively, where the differences between the results of two different analyses are considered as the systematic uncertainties. The o-H2 concentration of 0.05% was excessively large for the build-up of the polarizations. Accordingly, in future work, we will optimize the concentration of o-H2 and reduce that of p-D2 and shorten the traditionally used three-month aging period. We can decrease the probability of various failures and the helium consumption during the aging.

The crossed-coil NMR system presented in the present manuscript has the potential to help studying complex phenomena in the brute force polarization build-up of HD targets. In addition, the present frequency sweep method will be useful for the monitoring of the polarization of the HD target during the photoproduction experiments at SPring-8.

In these experiments, we also observed the NMR signals of the 27Al and 35Cl nuclei. Recently, we started developing a polarized 139La target for a T-violation experiment with polarized neutron beams at J-PARC. The present NMR system can be used for a broad range of frequencies, including the frequency of 102 MHz for the 139La nucleus at 17 T. Although the maximum frequency in this experiment was 726 MHz, the present NMR system can generate and observe signals at higher frequencies up to 1.3 GHz. The present technique would play an important role not only for the target development of HD but also for various other polarized nuclear target developments.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

The presented study involved the conduct of experiments at the BL33LEP of SPring-8 with the approval of the Japanese Synchrotron Radiation Research Institute (JASRI) as the contract beam line (Proposal No. BL33LEP/6001). We are grateful to the staff of the Low Temperature Center of Osaka University for supplying us with the required liquid helium. We thank Dr. J.-P. Dideletz, Dr. S. Bouchigny, Dr. G. Rouille, Professor G. Frossati, Dr. N. R. Hoovinakatte, Dr. A. M. Sandorfi, Dr. X. Wei, Dr. M. M. Lowry, and Dr. T. Kageya for their important advice. We also thank Professor K. Fukuda, Dr. T. Kunimatsu, and Professor M. Tanaka for the construction of the primary NMR system and for the provision of some additional modules. The present work was supported in part by the Ministry of Education, Science, Sports, and Culture of Japan and by the National Science Council of the Republic of China.

1.
T.
Nakano
,
J. K.
Ahn
,
M.
Fujiwara
,
H.
Kohri
,
N.
Matsuoka
,
T.
Mibe
,
N.
Muramatsu
,
M.
Nomachi
,
H.
Shimizu
,
K.
Yonehara
 et al, “
Multi-GeV laser-electron photon project at SPring-8
,”
Nucl. Phys. A
684
,
71
(
2001
).
2.
N.
Muramatsu
,
Y.
Kon
,
S.
Daté
,
Y.
Ohashi
,
H.
Akimune
,
J. Y.
Chen
,
M.
Fujiwara
,
S.
Hasegawa
,
T.
Hotta
,
T.
Ishikawa
 et al, “
Development of high intensity laser-electron photon beams up to 2.9 GeV at the SPring-8 LEPS beam line
,”
Nucl. Instrum. Methods Phys. Res., Sect. A
737
,
184
(
2014
).
3.
T.
Mibe
,
W. C.
Chang
,
T.
Nakano
,
D. S.
Ahn
,
J. K.
Ahn
,
H.
Akimune
,
Y.
Asano
,
S.
Date
,
H.
Ejiri
,
H.
Fujimura
 et al, “
Near-threshold diffractive ϕ-meson photoproduction from the proton
,”
Phys. Rev. Lett.
95
,
182001
(
2005
).
4.
S. Y.
Ryu
,
J. K.
Ahn
,
T.
Nakano
,
D. S.
Ahn
,
S.
Ajimura
,
H.
Akimune
,
Y.
Asano
,
W. C.
Chang
,
J. Y.
Chen
,
S.
Date
 et al, “
Interference effect between ϕ and Λ(1520) production channels in the γp → K+Kp reaction near threshold
,”
Phys. Rev. Lett.
116
,
232001
(
2016
).
5.
H.
Kohri
,
S. H.
Shiu
,
W. C.
Chang
,
Y.
Yanai
,
D. S.
Ahn
,
J. K.
Ahn
,
J. Y.
Chen
,
S.
Date
,
H.
Ejiri
,
H.
Fujimura
 et al, “
Differential cross section and photon-beam asymmetry for the γp → πΔ++(1232) reaction at forward π angles for Eγ = 1.5–2.95 GeV
,”
Phys. Rev. Lett.
120
,
202004
(
2018
).
6.
S. H.
Hwang
,
K.
Hicks
,
J. K.
Ahn
,
T.
Nakano
,
D. S.
Ahn
,
W. C.
Chang
,
J. Y.
Chen
,
S.
Date
,
H.
Ejiri
,
H.
Fujimura
 et al, “
Spin-density matrix elements for γp → K*0Σ+ at Eγ = 1.85–3.0 GeV with evidence for the κ(800) meson exchange
,”
Phys. Rev. Lett.
108
,
092001
(
2012
).
7.
R. G. T.
Zegers
,
M.
Sumihama
,
D. S.
Ahn
,
J. K.
Ahn
,
H.
Akimune
,
Y.
Asano
,
W. C.
Chang
,
S.
Date
,
H.
Ejiri
,
H.
Fujimura
 et al, “
Beam-polarization asymmetries for the p(γ, K+)Λ and p(γ, K+0 reactions for Eγ = 1.5–2.4 GeV
,”
Phys. Rev. Lett.
91
,
092001
(
2003
).
8.
H.
Kohri
,
D. S.
Ahn
,
J. K.
Ahn
,
H.
Akimune
,
Y.
Asano
,
W. C.
Chang
,
S.
Date
,
H.
Ejiri
,
S.
Fukui
,
H.
Fujimura
 et al, “
Differential cross section and photon-beam asymmetry for the γn → K+Σ reaction at Eγ = 1.5–2.4 GeV
,”
Phys. Rev. Lett.
97
,
082003
(
2006
).
9.
K.
Hicks
,
D.
Keller
,
H.
Kohri
,
D. S.
Ahn
,
J. K.
Ahn
,
H.
Akimune
,
Y.
Asano
,
W. C.
Chang
,
S.
Date
,
H.
Ejiri
 et al, “
Cross sections and beam asymmetry for K+Σ*− photoproduction from the deuteron at Eγ = 1.5–2.4 GeV
,”
Phys. Rev. Lett.
102
,
012501
(
2009
).
10.
N.
Muramatsu
,
J. Y.
Chen
,
W. C.
Chang
,
D. S.
Ahn
,
J. K.
Ahn
,
H.
Akimune
,
Y.
Asano
,
S.
Date
,
H.
Ejiri
,
H.
Fujimura
 et al, “
Near-threshold photoproduction of Λ(1520) from protons and deuterons
,”
Phys. Rev. Lett.
103
,
012001
(
2009
).
11.
H.
Kohri
,
D. S.
Ahn
,
J. K.
Ahn
,
H.
Akimune
,
Y.
Asano
,
W. C.
Chang
,
S.
Date
,
H.
Ejiri
,
S.
Fukui
,
H.
Fujimura
 et al, “
Near-threshold Λ(1520) production by the γp → K+Λ(1520) reaction at forward K+ angles
,”
Phys. Rev. Lett.
104
,
172001
(
2010
).
12.
A.
Honig
, “
Highly spin-polarized proton samples-large, accessible, and simply produced
,”
Phys. Rev. Lett.
19
,
1009
(
1967
).
13.
A.
Honig
,
Q.
Fan
,
X.
Wei
,
A. M.
Sandorfi
, and
C. S.
Whisnant
, “
New investigations of polarized solid HD targets
,”
Nucl. Instrum. Methods Phys. Res., Sect. A
356
,
39
(
1995
).
14.
M.
Rigney
,
M.
Breuer
,
S.
Buhler
,
C.
Commeaux
,
J.-P.
Didelez
,
P.
Hoffman-Roth
,
A.
Honig
,
A.
Sandorfi
,
C.
Schaerf
,
R.
Skowron
 et al, “
Solid HD polarized target: Conceptual design of transport and in-beam cryostat
,”
Nucl. Instrum. Methods Phys. Res., Sect. A
356
,
116
(
1995
).
15.
X.
Wei
,
C. M.
Bade
,
A.
Caracappa
,
T.
Kageya
,
F. C.
Lincoln
,
M. M.
Lowry
,
J. C.
Mahon
,
A. M.
Sandorfi
,
C. E.
Thorn
, and
C. S.
Whisnant
, “
New improvements leading to higher polarization frozen spin HD targets at the LEGS facility
,”
Nucl. Instrum. Methods Phys. Res., Sect. A
526
,
157
(
2004
).
16.
C.
Thorn
and
A.
Caracappa
, “
Precision HD polarimetry with the LEGS crossed-coil NMR polarimeter
,”
AIP Conf. Proc.
980
,
397
(
2008
).
17.
C.
Thorn
, “
Precision polarimetry of solid HD photonuclear targets at LEGS
,”
AIP Conf. Proc.
1149
,
880
(
2009
).
18.
G.
Rouille
,
M.
Bassan
,
C.
Commeaux
,
J. P.
Didelez
,
C.
Schaerf
,
V.
Bellini
,
J. P.
Bocquet
,
M.
Capogni
,
M.
Castoldi
,
A. D.
Angelo
,
G.
Gervino
,
F.
Ghio
,
B.
Girolami
,
M.
Guidal
,
E.
Hourany
,
R.
Kunne
,
P. L.
Sandri
,
A.
Lleres
,
D.
Moricciani
,
D.
Rebreyend
, and
A.
Zucchiatti
, “
A polarized HD target factory in Europe
,”
Nucl. Instrum. Methods Phys. Res., Sect. A
464
,
428
(
2001
).
19.
M.
Bassan
,
S.
Bouchigny
,
C.
Commeaux
,
J.-P.
Didelez
,
G.
Rouillé
,
C.
Schaerf
,
V.
Bellini
,
J.-P.
Bocquet
,
M.
Castoldi
 et al, “
Static and dynamic polarization of HD
,”
Nucl. Instrum. Methods Phys. Res., Sect. A
526
,
163
(
2004
).
20.
S.
Hoblit
,
A. M.
Sandorfi
,
K.
Ardashev
,
C.
Bade
,
O.
Bartalini
,
M.
Blecher
,
A.
Caracappa
,
A.
D’Angelo
,
A.
d’Angelo
,
R.
Di Salvo
 et al, “
Measurements of HD(γ, π) and implications for the convergence of the Gerasimov-Drell-Hern integral
,”
Phys. Rev. Lett.
102
,
172002
(
2009
).
21.
D.
Ho
,
P.
Peng
,
C.
Bass
,
P.
Collins
,
A.
D’Angelo
,
A.
Deur
,
J.
Fleming
,
C.
Hanretty
,
T.
Kageya
,
M.
Khandaker
 et al, “
Beam-target helicity asymmetry for γn → πp in the N* resonance region
,”
Phys. Rev. Lett.
118
,
242002
(
2017
).
22.
M. M.
Lowry
,
C. D.
Bass
,
A.
D’Angelo
,
A.
Deur
,
G.
Dezern
,
C.
Hanretty
,
D.
Ho
,
T.
Kageya
,
D.
Kashy
,
M.
Khandaker
 et al, “
A cryostat to hold frozen-spin polarized HD targets in CLAS: HDice-II
,”
Nucl. Instrum. Methods Phys. Res., Sect. A
815
,
31
(
2016
).
23.
H.
Kohri
,
M.
Fujiwara
,
K.
Fukuda
,
T.
Hotta
,
T.
Kunimatsu
,
C.
Morisaki
,
T.
Ohta
,
S.
Ono
,
K.
Ueda
,
M.
Uraki
 et al, “
Polarized HD target for future LEPS experiments at SPring-8 in Japan
,”
Int. J. Mod. Phys. E
19
,
903
(
2010
).
24.
M.
Utsuro
,
T.
Kuwai
,
T.
Ohta
,
H.
Kohri
, and
M.
Fujiwara
, “
Anisotropic thermal conductivity of hydrogen deuteride crystal for polarized proton target development
,”
Physica B
418
,
36
(
2013
).
25.
Y.
Yanai
,
H.
Kohri
,
M.
Yosoi
,
T.
Hotta
,
H.
Katsuragawa
,
T.
Ohta
,
S.
Tanaka
, “
Development of polarized HD target for SPring-8/LEPS experiment
,” in
Proceedings of XVII International Workshop on Polarized Sources, Targets and Polarimetry—PoS(PSTP2017)
,
2017
.
26.
T.
Ohta
,
S.
Bouchigny
,
J.-P.
Didelez
,
M.
Fujiwara
,
K.
Fukuda
,
H.
Kohri
,
T.
Kunimatsu
,
C.
Morisaki
,
S.
One
,
G.
Rouille
 et al, “
Distillation of hydrogen isotopes for polarized HD targets
,”
Nucl. Instrum. Methods Phys. Res., Sect. A
664
,
347
(
2012
).
27.
T.
Ohta
,
S.
Bouchigny
,
J.-P.
Didelez
,
M.
Fujiwara
,
K.
Fukuda
,
H.
Kohri
,
T.
Kunimatsu
,
C.
Morisaki
,
S.
Ono
,
G.
Rouille
 et al, “
HD gas analysis with gas chromatography and quadrupole mass spectrometer
,”
Nucl. Instrum. Methods Phys. Res., Sect. A
640
,
241
(
2011
).
28.
H.
Kohri
, “
Photoproduction for N* and related topics at LEPS/LEPS2
,” in
Proceedings of the 10th International Workshop on the Physics of Excited Nucleons (NSTAR2015)
(
The Physical Society of Japan
,
2016
), Vol. 10, p.
010008
.
29.
T.
Ohta
,
M.
Fujiwara
,
K.
Fukuda
,
H.
Kohri
,
T.
Kunimatsu
,
C.
Morisaki
,
S.
Ono
,
M.
Tanaka
,
K.
Ueda
,
M.
Uraki
 et al, “
Development of portable NMR polarimeter system for polarized HD target
,”
Nucl. Instrum. Methods Phys. Res., Sect. A
633
,
46
(
2011
).
30.
See https://leiden-cryogenics.com/ for Leiden Cryogenics B.V.
31.
See https://www.jastec-inc.com/e_top/ for Japan Superconductor Technology, Inc.
32.
S.
Bouchigny
,
C.
Commeaux
,
J.-P.
Didelez
, and
G.
Rouillé
, “
Progress in the polarization of HD
,”
Nucl. Instrum. Methods Phys. Res., Sect. A
544
,
417
(
2005
).
33.
S.
Bouchigny
and
J.-P.
Didelez
, “
Towards the polarization of DT molecules
,”
Nucl. Instrum. Methods Phys. Res., Sect. A
577
,
320
(
2007
).
34.
S.
Bouchigny
,
J.-P.
Didelez
,
F.
Dubois
, and
G.
Rouille
, “
Distillation of HD gas and measurement of spin-lattice relaxation times
,”
Nucl. Instrum. Methods Phys. Res., Sect. A
607
,
271
(
2009
).
35.
W. N.
Hardy
and
J. R.
Gaines
, “
Nuclear spin relaxation in solid HD with H2 impurity
,”
Phys. Rev. Lett.
17
,
1278
(
1966
).
36.
T.
Ohta
,
H.
Kohri
,
M.
Yosoi
,
M.
Fujiwara
, and
M.
Tanaka
, “
Development of 17T-NMR system for measurement of polarized HD and 3He targets
,” in
Proceedings of XVIth International Workshop in Polarized Sources, Targets, and Polarimetry—PoS(PSTP2015)
,
2015
.
37.
A. C.
Anderson
,
W.
Reese
, and
J. C.
Wheatley
, “
Thermal conductivity of some amorphous dielectric solids below 1 K
,”
Rev. Sci. Instrum.
34
,
1386
(
1963
).
38.
F.
Pobell
, in
Matter and Methods at Low Temperatures
(
Springer
,
2007
), Chap. 10.