Hyperpolarized cesium ions doped in a glass material.

Journal of Magnetic Resonance 249 (2014) 94–99

Contents lists available at ScienceDirect

Journal of Magnetic Resonance journal homepage: www.elsevier.com/locate/jmr

Hyperpolarized cesium ions doped in a glass material Kiyoshi Ishikawa ⇑ Graduate School of Material Science, University of Hyogo, Ako-gun, Hyogo 678-1297, Japan

a r t i c l e

i n f o

Article history: Received 30 April 2014 Revised 3 October 2014 Available online 17 October 2014 Keywords: Hyperpolarized ion Spin injection Optical pumping

a b s t r a c t Hyperpolarized (HP) 133 Cs nuclear magnetic resonance signals were measured from borosilicate glass cell walls during optical pumping of cesium vapor at high magnetic ﬁeld (9.4 T). Signiﬁcant signal enhancements were observed when additional heating of the cell wall was provided by intense but non-resonant laser irradiation, with integrated HP 133 Cs NMR signals and line widths varying as a function of heating laser power (and hence glass temperature). Given that virtually no Cs ions would originally be present in the glass, absorbed HP Cs atoms rarely met thermally-polarized Cs ions already at the surface; thus, spinexchange via nuclear dipole interaction cannot be the primary mechanism for injecting spin polarization into the glass. Instead, it is concluded that the absorption and transport of HP atoms into the glass material itself is the dominant mechanism of nuclear spin injection at high temperatures—the ﬁrst reported experimental demonstration of such a mechanism. Ó 2014 Elsevier Inc. All rights reserved.

1. Introduction Spin polarization of optically-pumped atoms can be transferred to nuclei in a solid [1]. The dilute vapor signiﬁcantly polarizes the dense nuclear system of the solid. In the sense that spin polarization transfer is caused by spin interaction when the polarized atoms come close to unpolarized nuclei at the surface, it is considered to be similar to spin-exchange optical pumping, where optically polarized atoms exchange spin angular momentum during atomic collisions with unpolarized atoms [2–5]. However, there are still many unknowns on the physics of spin polarization. In the past few years, optically-polarized cesium (Cs) atoms have greatly enhanced nuclear magnetic resonance (NMR) signal of Cs ions in the hydrides, CsH and CsD, and the halides, CsCl and CsI [6,7]. The polarized atoms release valence electrons close to the salt surface, resulting in spin polarized ions. The electronic process is so quick that the atoms never degrade nor enhance their own and salt’s nuclear polarization. Since nuclear–nuclear dipole interaction is always on resonance between the ions of the same nuclear species [8], nuclear polarization is efﬁciently transferred at any magnitude of applied magnetic ﬁeld [9]. At the temperature of 100 C, nuclear polarization has been transported in the bulk salt by spin diffusion via dipolar interaction [10]. Despite the difference in each combination of nuclear and electron spins, the abovedescribed spin transfer and transport are induced by the spin interaction, hence the name spin-exchange optical pumping. ⇑ Fax: +81 791 58 0131. E-mail address: [emailprotected] http://dx.doi.org/10.1016/j.jmr.2014.10.002 1090-7807/Ó 2014 Elsevier Inc. All rights reserved.

In the salt at high temperature, ion migration is important for ionic and electrical conductivity [11,12] and the same holds true for spin transport. Owing to diffusion migration of spin-polarized ions, nuclear spin polarization is transported from the surface to the bulk salt faster than the spin diffusion via dipolar interaction. Therefore, 133 Cs NMR signal has been several orders of magnitude enhanced in the polarized salts [7,13]. In the same way as ion migration in the bulk, atom-exchange are also possible at the surface, that is, the polarized atoms can substitute for the surface ions. The mechanism has not been studied previously. Here, we classify potential mechanisms for spin polarization injection into the salt, as schematically shown in Fig. 1(a) and (b). Although the atom-exchange rate should be signiﬁcantly different between the hydride that decomposes approximately at 200 C [14] and the halides that are stable beyond the melting point, similar signal enhancement was observed in both ionic solids [10]. Therefore, nuclear dipolar spin-exchange plays an important role in spin injection into the salts at 100 C. On the other hand, atom-exchange is expected to dominate at the higher temperature. The large signal of the bulk salt hides the physics about adsorbed ions at the surface [15,16] even by measurements in a wide temperature range. As is the case in spin-exchange via resonant dipolar interaction, atom-exchange is expected to be nearly independent of magnetic ﬁeld. As the result, separate measurements of the mechanisms have not been successful, and the salt is not the best material for study of atom-exchange at the surface. The studies on the other materials such as solid xenon [17–19] and optical pumping of atoms in solid helium [20] also provide little information on the surface. Alkali-metal atoms and ions at the

K. Ishikawa / Journal of Magnetic Resonance 249 (2014) 94–99

surface of glass and other substrates has been studied by X-ray induced photo-electron spectroscopy [21,22], light induced atom desorption [23,24], and sublevel spectroscopy by evanescent wave [25], posing the challenge for investigation of spin dynamics at the surface. The spin relaxation is accelerated by electric and magnetic ﬁelds and their gradient at the surface. Therefore, the spin injection mechanism faster than the relaxation is needed to hyperpolarize the bulk. This paper presents an NMR study of hyperpolarized (HP) Cs ions in sodium borosilicate glass. We will demonstrate that spinexchange mechanism shown in Fig. 1(c) is intrinsically inhibited in the Cs-free glass. Instead, the polarized ion injection shown in Fig. 1(d) is measured with no background from the bulk. The measurements can be understood by taking into account doping (absorption) of HP ions and migration (diffusion) in the glass as well as the motional narrowing of NMR signal at high temperature. It means that hyperpolarization survives the entire process of ionization, penetration into the glass, and diffusion into the bulk. The two of them, ionization and diffusion, have been tested by using the salts [6,7,9,10,13]. Therefore, the present work provides the ﬁrst evidence for polarized ion injection to hot softened materials. Experimental background is described brieﬂy. Polarized atomic beam is a good probe to detect the surface state which the reﬂected atoms carry information on [27,28]. It might be capable of doping the atoms into a glass substrate and we could independently control experimental parameters such as atom ﬂux density, injection velocity, and substrate temperature, in return for complicated

Fig. 1. Schematic diagram of potential mechanisms for spin polarization injection to the salts and the glasses. (a) Spin exchange via resonant nuclear dipolar interaction at the salt surface. Spin polarization is transferred between physisorbed HP ions (formerly vapor atoms) and thermally-polarized ions in the salt when vapor atoms release valence electrons. Since ½Cs ¼ ½X in the halide CsX, spin-exchange works effectively. (b) Atom exchange (also called as chemical exchange and cation replacement) at the salt surface. HP ions (formerly vapor atoms) replace the thermal ions or penetrate into vacancy in the salt. (c) Spin exchange via resonant nuclear dipolar interaction between HP ions (formerly vapor atoms) and the thermal ions already at the glass surface. The mechanism is less effective if the thermal ions are dilute at the surface. (d) Ion injection at the glass surface. HP ions (formerly vapor atoms) are injected into the structure of glass. The mechanism that Cs ions replace sodium ions may also be signiﬁcant because sodium ions are present in sodium borosilicate glass with the proportion of ½Na ½Si=10 [26]. Electron spin polarization of the atoms contributes little to nuclear polarization in these materials. The released electrons will be trapped in the salt and the glass, emitted as photoelectrons, or reduce end group at the glass surface. Nuclear polarization is transported from the surface into the bulk via dipolar interaction and ion migration, depending on density and mobility of the ions. Because of small electric quadrupole moment, spin polarization of Cs nuclei decays slowly in a local ﬁeld of cubic symmetry. For example, the decay time is 600 s in CsCl at 100 C.

setup. Instead, we adopt a simple way that Cs atoms diffusing in a buffer gas are optically pumped in a sealed glass cell. On contact with the glass walls, the polarized atoms are doped as polarized ions into the glass. The amount of doped ions is increased when the glass is heated up to the softening temperature. Localized laser heating enables the experiment in a wide range of temperatures in a standard NMR coil, as described below. 2. Experiment The fabrication details of cylindrical borosilicate glass cells have been described [6]. The glass cell used in this work contains simply Cs metal and N2 buffer gas, and any Cs compounds have not been intentionally added. Virtually no Cs ions are originally present in the sodium borosilicate glass. Due to magnetic ﬁeld produced by conduction electrons, the NMR frequency in the Cs metal is higher by the Knight shift of 770 kHz than in the compounds at the applied ﬁeld of 9.4 T [29,30]. Therefore, no 133 Cs NMR signal was observed in a frequency region of interest prior to establishment of optical pumping conditions. Optical alignment near the NMR coil is shown in Fig. 2. An NMR probe was maintained at 90 C by an electric heater. In the probe, the glass cell was heated from above by unfocused laser beam at the maximum power of 50 W and at the wavelength of 1060 nm, and also heated from below by focused laser beam of a few watts at 790 nm. Heating from both sides made the Cs metal stay near the bottom. The lasers heated the glass like a torch would. Because the OH-stretch band at 3 lm absorbs the heating lasers at most on the order of 1% [31], impurity bands should have helped the laser heating. We were careful about macroscopic impurities, which would absorb the intense light and make a hole through the glass wall [32]. The heating lasers are far from resonance and unlikely to have a signiﬁcant impact on atomic transitions. A single-mode laser of 1.2 W for optical pumping was tuned to the D2 line (852 nm) of Cs atoms, linearly polarized, and focused from below. The pumping light heated the glass and also the buffer gas because of quenching of atomic ﬂuorescence. Focusing of the lasers on the metal led to bright spot and sudden increase of vapor density.

Fig. 2. Optical and NMR setup. A cylindrical glass cell (outer diameter 10 mm) contains Cs metal and N2 gas at 2 kPa. When the metal stay in the detection region, quality factor of the coil is not so high but sensitive enough for the NMR measurement. Tuning of NMR circuit has been readjusted at each measurement condition. Air temperature in the NMR probe was measured by non-magnetic thermocouple placed about 10 mm from the glass cell, and maintained at 90 C by an electric heater. The glass walls were hotter than the air due to laser heating. One of heating lasers (6 50 W, 1060 nm) lighted the glass cell from above introduced with optical ﬁber. The heating light was scattered by the sealed glass stem and signiﬁcant part of them warmed the surrounding parts such as the coil, the cell holder, and the inner walls of NMR probe. Another heating laser (a few watts, 790 nm) and optical pumping laser (1.2 W, 852 nm) were focused from below through free space.

K. Ishikawa / Journal of Magnetic Resonance 249 (2014) 94–99

3. Result and discussion

Optical pumping

NMR amplitude (arb. units)

The pumping light was scattered and unpolarized by the glass at the focusing position, but we were able to polarize atomic vapor nonetheless. The details have been described [5,6,33,34]. In close relation to this work, we brieﬂy explain high-ﬁeld optical pumping of atoms in a buffer gas. The electron Zeeman splitting of the Cs atoms is large enough at 9.4 T to resolve eight transitions 2 S1=2 ðmS ¼ 1=2; mI Þ $ 2 P3=2 ðmJ ¼ 1=2; 3=2; m0I Þ from (a) to (h), where mS and mJ are magnetic quantum numbers of electron and mI of nucleus. Each transition has partially-resolved components for jmI j 6 7=2, where the differences, DmI ¼ m0I mI and Dme ¼ mJ mS , satisfy DmI þ Dme ¼ mp (photon’s spin angular momentum). According to DmI and Dme for six transitions from (b) to (g), r and p-polarized lights induce several optical-pumping processes which are competitive with each other. The direction and the degree of nuclear spin polarization of pumped atoms depend on mixing ratio of the light polarizations, buffer gas pressure, vapor density, and so on. In some cases, due to cancellation between these processes, net nuclear polarization is much lower than what would be induced in each process. On the other two resonance transitions (a) and (h), simple optical pumping scheme is expected; atoms are excited only by rþ polarized component (mp ¼ 1) of scattered pumping light for D2(a) ðmS ¼ 1=2; mJ ¼ 3=2; DmI ¼ 1Þ, and only by r polarized light (mp ¼ 1) for D2(h) ðmS ¼ 1=2; mJ ¼ 3=2; DmI ¼ 1Þ. At the two transitions in high magnetic ﬁeld, therefore, the laser polarization and the optical axis with respect to the external ﬁeld are not responsible for the direction of atomic spin polarization. Though they are not the strongest transitions in the D2 line, we expect sufﬁcient and stable spin polarization in atomic vapor. At D2(h), optical pumping induces the nuclear polarization being aligned parallel to the applied magnetic ﬁeld, we call this direction as positive, and the negative (reversed) electron spin polarization. Optical pumping at D2(a) induces the negative nuclear polarization and the positive electron polarization.

D2(h) 351 329 GHz 351 349 GHz D2(a) 352 122 GHz

1 0 -1 -2

Heating laser power 45 W

-5

Fig. 3. 133 Cs NMR signal of Cs ions in the sodium borosilicate glass. Free induction decay signals were averaged twenty times with repetition time of 4 s, and Fourier transformed. The signal was positively enhanced by optical pumping at the transition D2(h), and reversely at D2(a). No signal was measured by detuning the pumping laser. Optical pumping and heating lasers were focused from below, and the other heating laser was 45 W with unfocused beam at the glass cell. The frequency difference between D2(h) and D2(a) at 9.4 T is so large that it takes a few minutes to change the laser tuning. Signal averaging time of 80 s is much larger than the spin buildup time of a few seconds. Within the lifetime of the glass cell, therefore, it was difﬁcult to measure the dynamics of spin polarization. Just for information, it took 440 s to measure the three signals, including the time for signal averaging and laser tuning.

CsCl / H2O

fused

CsI / glass

CsCl / glass

2 CsI

CsCl

0 -5

During continuous laser heating and optical pumping, we applied a single radio-frequency pulse to detect free-induction decay signal with the NMR coil. Fourier transform of the averaged signal is shown in Fig. 3. Unexpectedly, the signals showed up although it was unclear at that time where they came from in the virtually empty glass cell. The signal was reversed by different optical-pumping transitions not by maladjustment of NMR circuit. The NMR frequency was close to reference frequencies of the water solution, the salts, and the fused glass in the different glass cells, as shown in Fig. 4. Therefore, we infer that the signals were from Cs ions with no nearby paramagnetic electrons. The atoms are rarely ionized by our lasers, eliminating the possibility of detecting the polarized Cs ions in the gas phase. Since no compounds have ever been added in the cell, we come up with the only solution that the signal comes from Cs ions in the glass walls. Because of small electric quadrupole moment of Cs nuclei, line splitting is negligible and thus the NMR signal is inhom*ogeneously broadened in the local ﬁeld of glass material. The line shape is neither Gaussian nor Lorentzian, suggesting the effect of several relaxation processes. The repetition time for signal averaging of four seconds was found large enough for the saturated maximum amplitude. Therefore, the spin polarization was built up in a few seconds. In the case that spin diffusion length is smaller than the wall thickness, the buildup time is equal to the spin relaxation time T 1 [10]. However, we were unable to measure it in an accurate way such as saturation recovery because the signal stably lasted only a few hours. When the laser was tuned to the transition D2(h), nuclear polarization of atomic vapor was positive and the NMR signal of Cs ions was posi-

NMR frequency (rel. to 52 495) (kHz)

NMR amplitude (arb. units)

NMR frequency (rel. to 52 495) (kHz) Fig. 4. Reference 133 Cs NMR signals measured for water solution of CsCl (40 C) and solid-state CsCl (100 C) and CsI (110 C). With no optical pumping, nuclear spin state is in thermodynamic equilibrium. The signals are also measured in the middle of fusing the salt and the borosilicate glass by laser heating, where the frequency shift is 2.5 kHz (CsCl/glass) and 7.5 kHz (CsI/glass), respectively, from the signals of pure salt. The sample should have been heated above the melting point of the salts 626 C (CsI), 645 C (CsCl) [35], and the soften point of the glass (821 C). For reference, the working point (glass blowing temperature) is approximately 1250 C [26].

tive from the glass walls. Optical pumping at D2(a) enhanced the signal reversely, meaning that the observed nuclear polarization of ions was negative. Namely, the direction of nuclear polarization of Cs ions in the glass was the same as nuclear polarization of atomic vapor. In the case of detuning the pumping laser from the atomic transitions, the atomic nuclei were thermally polarized and Cs ions in the glass also remained in the thermal equilibrium spin state. Since the thermal signal is below the noise level, we have yet to calculate the signal enhancement compared to the thermal equilibrium in the same experimental condition. Comparing to the signal for known amount of Cs salt in the different glass cells, we ﬁnd that the signal amplitude is signiﬁcantly enhanced by optical pumping. At the salt surface, dipolar interaction between Cs nuclei (one is from formerly vapor atoms and the other from salt ions as shown in Fig. 1(a)) was dominant at about 100 C in nuclear spin injection [6]. On the other hand, Cs ions in the glass are originally present

K. Ishikawa / Journal of Magnetic Resonance 249 (2014) 94–99

(a) 2:30

(b) 2:33

(d) 2:48

(e) 2:53

(f) 2:58

(g) 3:03

(h) 3:27

(i) 3:31

NMR amplitude (arb. units)

0.5

FWHM (kHz, )

NMR amplitude (arb. units)

6 4 4 3

2 35

Area (arb. units, )

much less than in the salt. Since the polarized atoms rarely meet thermally-polarized ions already at the surface, spin transfer between nuclei shown in Fig. 1(c) is initially inefﬁcient. When the borosilicate glass absorbs alkali metal, the glass cells take on brown discoloration as it has often been seen by overheating. The same happened for the glass cell used in this work, as quantitatively described below. Therefore, we carefully examine if spinexchange between the polarized physisorbed ions and the absorbed ions is signiﬁcant. Fig. 5 shows the signals repeatedly measured at the ﬁxed heating laser power of 35 W. For an hour, the enhanced signal is not increased though the thermally polarized ions are gradually accumulated in the glass walls. Since the signal amplitude is unaffected by the accumulated density of dilute ions, spin-exchange shown in Fig. 1(c) is less likely to be the primary mechanism of spin injection into the borosilicate glass. Part of metal ions might diffuse into the glass at the metal-glass interface. However, because nuclear polarization decays fast in the metal (T 1 1 ms) [36], the signal would not have been enhanced if the vapor atoms penetrated through the metal ﬁlm into the glass. Therefore, the enhanced signal arises from direct doping of polarized Cs ions into the borosilicate glass, as shown in Fig. 1(d). The amount of polarized ions depends on temperature because the polarized vapor density is increased by heating and the ions are more easily doped into the softened glass. As shown in Fig. 6, the signal is greatly enhanced with an increase in the heating laser power. In contrast, the resonance frequency ﬂuctuates over time, because the local temperature is unstable in our heating method and the chemical shift is dependent on temperature. The signal is narrowed by motional averaging at high temperature. The doped ions can stay temporarily at unstable sites in the glass network [37,38] and such ions are stabilized by ion migration. Therefore, we also assume line narrowing as ensemble average by annealing at high temperature. Since the linewidth of 3 kHz is an order of magnitude broader than the signal of fused samples shown in Fig. 4, the glass wall is not completely melted by heating at the maximum power of 50 W. Although the glass viscosity is more than ten orders of magnitude changed by heating up to the melting point [26], the enhancement and the linewidth have been changed only by a factor of four and two, respectively, throughout the measurement. The signal enhance-

Optical pumping D2(h) 351 329 GHz Heating laser power 35 W 40 W 45 W 50 W

Heating power (W)

0 -5

NMR frequency (rel. to 52 495) (kHz) Fig. 6. 133 Cs NMR signal of Cs ions doped in the borosilicate glass measured by changing the heating power of unfocused laser. The optical pumping laser is tuned to the transition D2(h). The power of focused lasers is maintained constant. Although the NMR frequency ﬂuctuates over time, the full width at half maximum (FWHM, ) and the integral of Fourier transform (Area, M) changes dependently on the heating power, as shown in the inset.

ment is theoretically expected proportional to J s T 1 in Eq. (A4) of Ref. [10], where J s is the injected nuclear spin current and T 1 in a thick hom*ogeneous solid. The measurements are, however, inadequate for quantitative analysis using this relation. The body of glass cell was heated by the unfocused laser as well as an electric heater. By the above-described experiment of approximately six hours, the borosilicate glass walls has absorbed the metal atoms and taken on a hom*ogeneous brown discoloration, as shown in Fig. 7. Based on our experience of heating the glass alkali-metal cells in a conventional oven, the temperature for discoloration is higher than 200 C. We conﬁrmed no metal left in the cell both by eye and by the NMR Knight-shift measurement. Therefore, on the order of 1020 ions should have been absorbed in the brown color glass. For lack of metal, the enhanced signal lasted only a few hours in the optical pumping experiment. Because of broad line width (> 10 kHz) in the limit of weak heating, as shown in Fig. 6, the NMR signal of thermally-polarized Cs ions was below the detectable level at room temperature even by averaging many times. Fig. 7 shows the dip made by focusing the lasers on the bottom window. It means that the glass was locally heated up to the softening point. From motional narrowing, we assume that the enhanced signals have originated from the doped ions at the dip. The integral of a Fourier transform is proportional to the number of polarized ions

0.0

0.5 0.0

0.5 0.0 0

NMR frequency (rel. to 52 495) (kHz) Fig. 5. Time variation of 133 Cs NMR signal by optical pumping at D2(h). The number on each ﬁgure indicates the measurement time. Each signal has been averaged ﬁfty times with the repetition period of 4 s. The brand-new glass cell was warmed up and maintained at 90 C by the electric heater for two and a half hours, and then the series of measurement was started with heating unfocused laser at 35 W. The glass cell was heated totally for six hours until all the metal was absorbed in the glass walls. The absorption of Cs metal into the glass walls is negligible at 90 C during the warm-up time.

Fig. 7. Photograph of the glass cell and a scale with minimum scale value of 1 mm, taken after the optical and NMR experiment. The cell has taken on a hom*ogeneous brown discoloration by absorption of the Cs ions. The focused laser beams have produced a dip at the bottom (left-hand side) of the cell because the softened glass is distorted in atmospheric pressure. The glass near the dip metamorphoses into crystal-like material.

K. Ishikawa / Journal of Magnetic Resonance 249 (2014) 94–99

(the number of ions spin polarization). Comparing the enhanced signal from the glass (Fig. 3) with the thermal equilibrium signal from the salt (Fig. 4), roughly 1016 polarized ions were absorbed into the glass. Assuming hom*ogeneous absorption of Cs ions, the volume at the dip ( 2 mm3 ), and the volume of the walls ( 600 mm3 ), we estimate nuclear spin polarization 3% ¼ ð1016 =2Þ=ð1020 =600Þ attained at the dip of the bottom glass window. 4. Summary HP Cs ions were doped in the sodium borosilicate glass during optical pumping of Cs atoms in the sealed glass cell. When the glass walls were heated up to the softening point, 133 Cs NMR signal was greatly enhanced and motionally narrowed. The signal of thermally polarized doped ions in the same glass cell was below the detectable level at high temperature and at room temperature. Therefore, using the reference signal in the different cells, we estimated that the percent level of nuclear polarization was attained at the softened glass dip. The ability to generate HP signal lasted a few hours—it was better to have enhanced the signal and lost the vapor soon at high temperature than never to have enhanced at all. It was also a challenge to measure HP ions in the glass walls without breaking the glass cell. Such a fragile situation will be overcome in near future. Although the conditions such as high temperature may not be suitable for every material, the results present the ﬁrst experimental evidence that atom-exchange and ion-injection can work for nuclear spin polarization injection. In future applications, it might become possible to laser write HP ions on the softened materials. Acknowledgments This work was supported in part by JSPS KAKENHI Grant No. 25610115. References [1] K. Ishikawa, B. Patton, Y.-Y. Jau, W. Happer, Spin transfer from an optically pumped alkali vapor to a solid, Phys. Rev. Lett. 98 (18) (2007) 183004, http:// dx.doi.org/10.1103/PhysRevLett.98.183004. [2] P. Franken, R. Sands, J. Hobart, Polarization of free potassium atoms by exchange collisions with sodium atoms and free electrons, Phys. Rev. Lett. 1 (1958) 52–54, http://dx.doi.org/10.1103/PhysRevLett.1.52. . [3] L.C. Balling, R.H. Lambert, J.J. Wright, R.E. Weiss, Optical pumping at high temperatures, Phys. Rev. Lett. 22 (1969) 161–163, http://dx.doi.org/10.1103/ PhysRevLett.22.161. . [4] T.G. Walker, W. Happer, Spin-exchange optical pumping of noble-gas nuclei, Rev. Mod. Phys. 69 (2) (1997) 629–642, http://dx.doi.org/10.1103/ RevModPhys.69.629. [5] W. Happer, Y.-Y. Jau, T.G. Walker, Optically Pumped Atoms, Wiley-VCH, Weinheim, 2010. [6] K. Ishikawa, B. Patton, B.A. Olsen, Y.-Y. Jau, W. Happer, Transfer of spin angular momentum from cs vapor to nearby cs salts through laser-induced spin currents, Phys. Rev. A 83 (6) (2011) 063410, http://dx.doi.org/10.1103/ PhysRevA.83.063410. [7] K. Ishikawa, Hyperpolarisation of cs salts by optical pumping of cs atoms in a random scattering medium at high magnetic ﬁeld, Micropor. Mesopor. Mater. 178 (2013) 123–125, http://dx.doi.org/10.1016/j.micromeso.2013.02.043. . [8] A. Abragam, Principles of Nuclear Magnetism, Oxford University Press, 1961. [9] K. Ishikawa, Glass-wool study of laser-induced spin currents en route to hyperpolarized cs salt, Phys. Rev. A 84 (1) (2011) 013403, http://dx.doi.org/ 10.1103/PhysRevA.84.013403. [10] K. Ishikawa, Spin accumulation in thin cs salts on contact with optically polarized cs vapor, Phys. Rev. A 84 (2011) 033404, http://dx.doi.org/10.1103/ PhysRevA.84.033404. . [11] J. Arends, H. Nijboer, Ionic conductivity of cscl, Solid State Commun. 5 (3) (1967) 163–166, http://dx.doi.org/10.1016/0038-1098(67)90510-8. . [12] I.M. Hoodless, R.G. Turner, Self-diffusion in single crystals of cscl, J. Phys. Chem. Solids 33 (10) (1972) 1915–1919, http://dx.doi.org/10.1016/S00223697(72)80490-6. .

[13] K. Ishikawa, Hyperpolarization of 133 cs nuclei enhanced by ion movement in a cesium salt, Phys. Rev. A 84 (2011) 061405(R), http://dx.doi.org/10.1103/ PhysRevA.84.061405. . [14] N.D. Bhaskar, J. Camparo, M. Ligare, W. Happer, Photoproduction of spinpolarized hydrogen atoms and electrons in mixtures of cesium vapor and hydrogen gas, Phys. Rev. Lett. 46 (21) (1981) 1387–1390, http://dx.doi.org/ 10.1103/PhysRevLett.46.1387. [15] H.W. Long, H.C. Gaede, J. Shore, L. Reven, C.R. Bowers, J. Kritzenberger, T. Pietrass, A. Pines, P. Tang, J.A. Reimer, High-ﬁeld cross polarization nmr from laser-polarized xenon to a polymer surface, J. Am. Chem. Soc. 115 (18) (1993) 8491–8492, http://dx.doi.org/10.1021/ja00071a086. . [16] B. Driehuys, G.D. Cates, W. Happer, Surface relaxation mechanisms of laserpolarized 129 xe, Phys. Rev. Lett. 74 (1995) 4943–4946, http://dx.doi.org/ 10.1103/PhysRevLett.74.4943. . [17] W.M. Yen, R.E. Norberg, Nuclear magnetic resonance of xe129 in solid and liquid xenon, Phys. Rev. 131 (1) (1963) 269–275, http://dx.doi.org/10.1103/ PhysRev.131.269. [18] R.J. Fitzgerald, M. Gatzke, D.C. Fox, G.D. Cates, W. Happer, 129 xe spin relaxation in frozen xenon, Phys. Rev. B 59 (1999) 8795–8811, http://dx.doi.org/10.1103/ PhysRevB.59.8795. . [19] N.N. Kuzma, B. Patton, K. Raman, W. Happer, Fast nuclear spin relaxation in hyperpolarized solid 129 xe, Phys. Rev. Lett. 88 (14) (2002) 147602, http:// dx.doi.org/10.1103/PhysRevLett.88.147602. [20] S. Lang, S. Kanorsky, T. Eichler, R. Müller-Siebert, T.W. Hänsch, A. Weis, Optical pumping of cs atoms in solid 4 he, Phys. Rev. A 60 (1999) 3867–3877, http:// dx.doi.org/10.1103/PhysRevA.60.3867. . [21] J. Zemek, P. Jiricek, O. Gedeon, B. Lesiak, A. Jozwik, Electron irradiated potassiumsilicate glass surfaces investigated by XPS, J. Non-Cryst. Solids 351 (19–20) (2005) 1665–1674, http://dx.doi.org/10.1016/j.jnoncrysol.2005.04.059. . [22] O. Gedeon, K. Jurek, Microanalysis of glass surfaces after thermal exposure, Microchimica Acta 155 (1-2) (2006) 147–150, http://dx.doi.org/10.1007/ s00604-006-0533-x. . [23] A. Gozzini, F. Mango, J. Xu, G. Alzetta, F. Maccarrone, R. Bernheim, Lightinduced ejection of alkali atoms in polysiloxane coated cells, Il Nuovo Cimento D 15 (5) (1993) 709–722, http://dx.doi.org/10.1007/BF02482437. . [24] S. Villalba, H. Failache, A. Lezama, Light-induced atomic desorption and diffusion of rb from porous alumina, Phys. Rev. A 81 (2010) 032901, http:// dx.doi.org/10.1103/PhysRevA.81.032901. . [25] E. Ulanski, Z. Wu, Measurement of dwell times of spin polarized rubidium atoms on octadecyltrichlorosilane- and parafﬁn-coated surfaces, Appl. Phys. Lett. 98 (20) (2011), http://dx.doi.org/10.1063/1.3591172. . [26] for example, Corning Wafer Sheet [link]. http://www.corning.com/ lightingmaterials/images/wafersheet.pdf. [27] A. Nass, C. Baumgarten, B. Braun, G. Ciullo, G. Court, P. Dalpiaz, A. Golendukhin, G. Graw, W. Haeberli, M. Henoch, R. Hertenberger, N. Koch, H. Kolster, P. Lenisa, H. Marukyan, M. Raithel, D. Reggiani, K. Rith, M. Simani, E. Steffens, J. Stewart, P. Tait, T. Wise, The HERMES polarized atomic beam source, Nucl. Instr. Meth. A 505 (3) (2003) 633–644, http://dx.doi.org/10.1016/S01689002(03)00986-0. . [28] H. Jänsch, K. Becker, K. Blatt, H. Leucker, D. Fick, R. Butsch, B. Heck, D. Krämer, K.-H. Möbius, W. Ott, P. Paul, R. Suntz, G. Tungate, I.M. Turkiewicz, A. Weller, E. Steffens, Nuclear spin polarized alkali beams (li and na): production and acceleration, Nucl. Instr. Meth. A 254 (1) (1987) 7–12, http://dx.doi.org/ 10.1016/0168-9002(87)90475-X. . [29] C.H. Townes, C. Herring, W.D. Knight, The effect of electronic paramagnetism on nuclear magnetic resonance frequencies in metals, Phys. Rev. 77 (1950) 852–853, http://dx.doi.org/10.1103/PhysRev.77.852. . [30] B. Patton, K. Ishikawa, Y.-Y. Jau, W. Happer, Intrinsic impurities in glass alkalivapor cells, Phys. Rev. Lett. 99 (2) (2007) 027601, http://dx.doi.org/10.1103/ PhysRevLett.99.027601. [31] W.H. Kohl, Handbook of Materials and Techniques for Vacuum Devices, Springer, New York, 1995. [32] D.K. Walter, W.M. Grifﬁth, W. Happer, Energy transport in high-density spinexchange optical pumping cells, Phys. Rev. Lett. 86 (2001) 3264–3267, http:// dx.doi.org/10.1103/PhysRevLett.86.3264. . [33] B.A. Olsen, B. Patton, Y.-Y. Jau, W. Happer, Optical pumping and spectroscopy of cs vapor at high magnetic ﬁeld, Phys. Rev. A 84 (2011) 063410, http:// dx.doi.org/10.1103/PhysRevA.84.063410. . [34] X. Zeng, C. Wu, M. Zhao, S. Li, L. Li, X. Zhang, Z. Liu, W. Liu, Laser-enhanced lowpressure gas nmr signal from 129 xe, Chem. Phys. Lett. 182 (6) (1991) 538–540, http://dx.doi.org/10.1016/0009-2614(91)90120-X. .

K. Ishikawa / Journal of Magnetic Resonance 249 (2014) 94–99 [35] M.A. Bredig, Molten Salt Chemistry, Interscience, New York, 1964. [36] J. Korringa, Nuclear magnetic relaxation and resonnance line shift in metals, Physica 16 (7-8) (1950) 601–610, http://dx.doi.org/10.1016/0031-8914(50)90105-4. . [37] P.M. Aguiar, V.K. Michaelis, C.M. McKinley, S. Kroeker, Network connectivity in cesium borosilicate glasses: 17 o multiple-quantum MAS and double-resonance NMR, J. Non-Cryst. Solids 363 (0) (2013) 50–56, http://dx.doi.org/10.1016/

j.jnoncrysol.2012.12.010. . [38] A. Osipov, L. Osipova, V. Eremyashev, Structure of alkali borosilicate glasses and melts according to raman spectroscopy data, Glass Phys. Chem. 39 (2) (2013) 105–112, http://dx.doi.org/10.1134/S1087659613020119. .

Hyperpolarized cesium ions doped in a glass material. - PDF Download Free (2024)