1: \documentclass[]{elsart}

2: \usepackage{amssymb}

3: \usepackage{epsfig}

4: 

5: \begin{document}

6: \runauthor{Huffman \textit{et al.}}

7: \begin{frontmatter}

8: 

9: \title{Magnetically Stabilized Luminescent Excitations in Hexagonal 

10: 	Boron Nitride}

11: 

12: \author[NIST]{P.\,R. Huffman\thanksref{Corr}}, 

13: \author[HARVARD]{C.\,R. Brome}, 

14: \author[HARVARD]{J.\,S. Butterworth}, 

15: \author[HARVARD]{S.\,N. Dzhosyuk},

16: \author[HMI]{R.~Golub}, 

17: \author[LANL]{S.\,K. Lamoreaux}, 

18: \author[HARVARD]{C.\,E.\,H. Mattoni}, 

19: \author[HARVARD]{D.\,N. McKinsey}, and

20: \author[HARVARD]{J.\,M. Doyle}

21: 

22: \address[NIST]{National Institute of Standards and Technology, 

23: 	    Gaithersburg, MD 20899, USA}

24: \address[HARVARD]{Harvard University, Cambridge, MA 02138, USA}

25: \address[HMI]{Hahn-Meitner Institut, Berlin, Germany}

26: \address[LANL]{Los Alamos National Laboratory, Los Alamos, NM 87545, 

27: 	    USA}

28: \thanks[Corr]{Correspondence and requests 

29: 	for materials should be sent to PRH (paul.huffman@nist.gov).}

30: 	    

31: \begin{abstract} 

32: Magnetically stabilized luminescence is observed in hexagonal boron

33: nitride.  The luminescence is induced by absorption of cold neutrons

34: and is in the visible region.  In the absence of a magnetic field, the

35: photon emission level is observed to decay over several hundred

36: seconds.  A fraction of this luminescence can be suppressed if the

37: temperature is $T \lesssim 0.6$~K and the magnetic field is $B \gtrsim

38: 1.0$~T\@.  Subsequent to irradiation and suppression, luminescence can

39: be induced by an increase in $T$ or lowering of $B$.  Possible

40: explanations include stabilization of triplet states or the

41: localization and stabilization of excitons.

42: \end{abstract}

43: 

44: \begin{keyword}

45: Boron Nitride, Magnetic Stabilization, Luminescence

46: \end{keyword}

47: \end{frontmatter}

48: 

49: \section{Introduction}

50: The study of luminescence in hexagonal boron nitride (h-BN) informs us

51: about properties intrinsic to the material as well as its defects and

52: impurities.  Techniques used to study h-BN include electron

53: paramagnetic resonance\cite{gei64,kvk69}, thermally stimulated

54: luminescence\cite{lop92}, thermally stimulated conduction\cite{lop92},

55: and luminescence induced by light, cathodic, electric, ionic, and

56: radioactive sources\cite{lar56,lsg92,ksz75,Kob94}.  Identified defects

57: include F-centers, nitrogen vacancies and elemental impurities

58: \cite{ksz75,khu73,gkl89,gkl90,dzh69,emk81}.  Information about the

59: intrinsic nature of h-BN has been obtained through studies of the

60: near-band-gap luminescence\cite{tbs94}.

61: 

62: A continuum photon emission spectrum in h-BN has been observed as a

63: result of thermal neutron exposure which produces energetic He$^{+}$

64: (1.47~MeV) and $^{7}$Li$^{+}$ (0.84~MeV) ions in the material via the

65: $^{10}$B(n,$\alpha$)$^{7}$Li reaction.  Other irradiation studies used

66: 1.0~MeV H$^{+}$ and 1.8~MeV He$^{+}$ ions and resulted in the

67: characterization of a continuum photon emission spectrum of

68: 280~nm to 600~nm\cite{Kob94}.  Here we present observations of magnetic

69: suppression of cold neutron induced luminescence in h-BN at low

70: temperature.  The luminescence is stimulated by the by-products of

71: neutron capture and can be suppressed by application of a magnetic

72: field.  The studies we present here were prompted by the use of h-BN

73: as a neutron shielding material in an experiment to magnetically trap

74: ultracold neutrons (UCN)\cite{huf00}.

75: 

76: \section{Experimental Apparatus}

77: Hexagonal boron nitride is used in our neutron trapping apparatus to

78: collimate an incident neutron beam and to absorb those neutrons

79: scattered out of the collimated beam.  Boron compounds are commonly

80: used as shielding materials for absorbing low energy neutrons. 

81: Natural boron has a 19.9~\% abundance of $^{10}$B, which has a large

82: thermal neutron absorption cross section, $\sigma_{\rm thermal} =

83: 3840$~b.  Hexagonal boron nitride is particularly suitable because of

84: its high purity and mechanical properties.  Hexagonal BN can be

85: obtained with $<0.4~\%$ oxygen and $<0.1~\%$ all other total elemental

86: impurities\cite{combatbn}.  It is easily machinable and has a low

87: thermal expansion coefficient.  The h-BN used in our experiment is

88: obtained commercially as a solid cylinder and machined into the form

89: of interlocking tubular pieces with inner diameters of 3.86~cm and

90: outer diameters of 4.16~cm.  The assembled h-BN tube fits inside a

91: cupronickel (CuNi) cell body and extends its entire length, 110~cm

92: (see figure~\ref{fig:fibercell}).  Two cylindrical rings are used as

93: neutron collimators and are positioned at distances of 36~cm and 74~cm

94: relative to the front (beam entrance) of the h-BN tube.  These rings

95: each have an inner diameter of 2.0~cm, length of 1~cm and fit snugly

96: within the h-BN tube.  The cell is filled with liquid $^4$He and is

97: cooled by a dilution refrigerator to 0.1~K to 1.0~K\@.

98:  

99: A cold neutron beam (moderated to 40~K) traverses the cell parallel to

100: the long axis with a fluence of approximately $5 \times

101: 10^{8}~{\mathrm s}^{-1}$.  About 60~\% of the neutrons entering the

102: cell are absorbed by the h-BN\@.  A set of superconducting magnets

103: resides just outside the cell body and when at maximum current, the

104: h-BN is in a magnetic field ranging from 1.0~T to 1.8~T\@.

105: 

106: Under normal operating conditions, a light collection system is

107: situated just inside the tube of h-BN between the two collimators (see

108: figure~\ref{fig:fibercell}).  This system consists of a diffuse

109: reflector made of expanded polytetrafluoroethylene (PTFE) onto which

110: is evaporated an ultraviolet-to-blue downconverter (tetraphenyl

111: butadiene, TPB)\cite{films} and two helically coiled blue-to-green

112: wavelength shifting fibers\cite{Kuraray}.  The fibers exit the

113: cryogenic apparatus and are each coupled to a photomultiplier tube. 

114: This system is sensitive to photons over the entire spectrum from the

115: extreme ultraviolet (EUV; $<100$~nm) to the visible blue ($\sim

116: 500$~nm) and is not capable of further wavelength

117: discrimination\footnote{The fiber light collection system was not

118: actually used in the measurement made in Ref.~\cite{huf00}.}.

119: 

120: \section{Results}

121: 

122: In a typical run, the detection region is exposed to neutrons for

123: 900~s.  During this irradiation period, the photomultiplier tubes

124: (PMTs) are turned off.  After 900~s, the neutron beam is blocked, the

125: PMTs are brought up to voltage and luminescence is detected.

126: 

127: The time dependence of the luminescence signal is shown in

128: figure~\ref{fig:nomagnet} (solid line), starting at the point when the

129: beam is turned off ($t=0$~s).  This decay does not fit to a single

130: exponential and varies roughly as inverse time.  When the magnet is

131: energized during the irradiation and observation, a significant

132: fraction of this luminescence signal is suppressed (see

133: figure~\ref{fig:nomagnet}, dashed line).  When the field is lowered

134: (starting at 1275~s), luminescence from the ``stored'' component is

135: emitted.

136: 

137: Numerous tests were performed to identify the source of the

138: luminescence signal and to characterize its behavior.  By the process

139: of elimination using multiple experimental setups, we determined that

140: the luminescence signal was originating from the boron nitride and

141: simplified the detection system to verify this conclusion.  In the

142: experimental setup shown in figure~\ref{fig:cleanapp}, only h-BN and

143: an acrylic lightguide were present (the TPB downconverter and

144: wavelength-shifting fibers were removed), and the geometry was

145: arranged such that neutrons could not directly impact the acrylic. 

146: The luminescence signal was observed.  Both components of this decay

147: signal can be seen in figure~\ref{fig:cleandata}.  In separate tests

148: performed at 4~K, we did not observe luminescence in an acrylic sample

149: under similar neutron irradiation conditions.

150: 

151: Studies to characterize the behavior of the magnetically stabilized

152: component of the luminescence signal were performed using the fiber

153: cell apparatus (figure~\ref{fig:fibercell}).  Data resulting from

154: lowering the magnetic field in various ways is shown in

155: figure~\ref{fig:storage}.  The initial irradiation takes place with

156: the magnets at maximum current.  In figure~\ref{fig:storage}A the

157: magnet current is held constant for 800~s after the neutron beam is

158: turned off.  Then the magnetic field is reduced in four steps: from

159: $100~\% \rightarrow 75~\%$, $75~\% \rightarrow 50~\%$, $50~\% \rightarrow

160: 25~\%$ and $25~\% \rightarrow 0~\%$ of its maximum value.  The largest

161: fraction of the luminescence is emitted in the last two reduction

162: steps.  Figure \ref{fig:storage}B further demonstrates the suppression

163: of luminescent emission.  In this measurement, the magnetic field is

164: slowly ramped down from 100~\% to zero over 2700~s.  Immediately after

165: this, the field is rapidly ($\sim 100$~s) turned back to 100~\% for

166: 570~s and then back to zero.  As can be seen, the luminescence is

167: suppressed while the field is on.  We have verified, in a separate

168: experiment, that the magnetic field has no significant effect on the

169: efficiency of the detection system.

170: 

171: To characterize the time dependence of the decay of the stabilized

172: component, the superconducting magnet was quenched resulting in a

173: reduction in field from its maximum value to zero in less than 1~s. 

174: The time dependence of this decay signal is roughly the same as the

175: initial decay (see figure~\ref{fig:quench}).  In addition, the total

176: integrated number of counts using the magnet quench signal is roughly

177: equal to that obtained using a slow field ramp.  Presumably the

178: luminescence is the result of relaxation of some type of solid state

179: excitation in the h-BN\@.

180: 

181: The persistence of these excitations under conditions of high magnetic

182: field is most dramatically demonstrated by leaving the magnetic field

183: on for 28,800~s after irradiation.  After this time the magnetic field

184: is lowered to zero and luminescence is observed.  The number of

185: photons emitted after 28,800~s was within 50~\% of the number emitted

186: after only 1000~s of storage (under the same irradiation conditions).

187: 

188: We also observed that raising the temperature of the cell has a

189: similar effect to lowering the magnetic field.  Figure

190: \ref{fig:variation} shows data for different field and temperature

191: conditions.  In both sets of data the cell was irradiated by the

192: neutron beam for 900~s.  In figure \ref{fig:variation}A the magnetic

193: field is lowered from its initial value to zero in approximately 100~s

194: beginning 1275~s after the beam is turned off.  Traces are shown for

195: initial magnet currents of 100~\%, 75~\%, 50~\% and 25~\% of its maximum

196: current.  The data shown in figure \ref{fig:variation}B is taken

197: maintaining the magnet at its maximum current while raising the

198: temperature of the cell from its initial value to 1~K in approximately

199: 100~s beginning 1275~s after the beam is turned off.  The traces are

200: for initial temperatures of 170~mK, 400~mK, 525~mK and 650~mK\@.  It is

201: important to note that temperature changes brought about by eddy

202: current heating while ramping the magnetic field ($< 20$~mK) are much

203: smaller than those required to initiate the release of luminescence

204: ($> 200$~mK).

205: 

206: \section{Discussion}

207: Given the myriad of mechanisms for producing luminescence, we can only

208: offer general hypotheses about the magnetically stabilized component

209: of the luminescence signal and offer some general points.  The energy

210: of interaction of the magnetic moment of an electron or hole in h-BN

211: with a magnetic field of 1~T in temperature (Kelvin) units is 1.3~K\@. 

212: This is the same energy scale as that observed in

213: figure~\ref{fig:variation}.  Our highest applied field of 1.8~T

214: results in an energy of interaction $\sim 2.3$~K\@.  The applied

215: magnetic field should lead to polarization of paramagnetic species as

216: the thermal fluctuations present in the solid are not energetic enough

217: for depolarization.  In addition to electrons and holes, other

218: possibilities for paramagnetic species present in the h-BN include

219: triplet state excitons.  These typically have a magnetic moment of one

220: Bohr magneton ($1 \mu_{B}$) or greater.  Spin polarization of trapped

221: electrons, holes or excitons may take place and then affect the

222: possible relaxation paths to luminescence.

223: 

224: One possibility is that the magnetic field stabilizes a triplet state

225: of a defect center such as an F-center.  Triplet states can have very

226: long lifetimes\cite{tripst} and the stabilization would prevent

227: conversion to the much shorter lived singlet state.  Observations of

228: magnetic suppression of singlet state formation, resulting in enhanced

229: triplet state formation has been observed in alkali

230: halides\cite{Por71,Bal83}.  Unlike our case, these triplet states

231: exhibit pure exponential decay which is uneffected by magnetic field

232: changes.  Another possibility is that the interaction of an electron

233: or exciton with a paramagnetic center causes trapping, similar to that

234: seen in indium antimonide\cite{kst94}.  The applied magnetic field

235: could keep the two species aligned.  As the temperature is raised (or

236: the field lowered) thermal fluctuations would release the excitations

237: which could then undergo radiative recombination.  The apparent

238: distribution of trapping energies may indicate a combination of

239: thermoluminescent activity with a magnetic interaction.  That is, the

240: usual thermoluminescent trapping is modified by some type of magnetic

241: interaction, perhaps one of those discussed above.  In this case, the

242: spatially varying magnetic field would provide a continuum of trapping

243: energies.

244: 

245: \section{Conclusion}

246: 

247: We have observed a magnetically stabilized component in the

248: luminescence of hexagonal boron nitride under neutron irradiation. 

249: The excitations that give rise to the luminescence can be stabilized

250: for long periods by high magnetic fields ($\approx 1$~T) at low

251: temperatures ($\lesssim 600$~mK).  Further experiments are necessary

252: to elucidate the nature of the excitations, including spectroscopic

253: analysis of the luminescence, quantitative analysis of the magnetic

254: field dependence and optically detected electron paramagnetic

255: resonance.

256: 

257: 

258: \section{Acknowledgments}

259: 

260: This work was supported in part by the National Science Foundation

261: under grant No.  PHY-9424278.

262: 

263: \begin{thebibliography}{99}

264: 

265: \bibitem{gei64}

266: D.~Geist and G.~R{\"o}melt, Sol.\ St.\ Comm.\ {\bf2}, 149 (1964).

267: 

268: \bibitem{kvk69}

269: V.\,A. Krasnoperov, N.\,V. Vekshina, M.\,B. Khusidman and V.\,S.

270: Neshpor, Zh.\ Prik.\ Spek.\ {\bf11}, 299 (1969).

271: 

272: \bibitem{lop92}

273: V.\,V. Lopatin and F.\,V. Konusov, J.\ Phys.\ Chem.\ Solids {\bf53},

274: 847 (1992).

275: 

276: \bibitem{lar56}

277: S.~Larach and R.\,E. Shrader, Phys.\ Rev.\ {\bf104}, 68 (1956).

278: 

279: \bibitem{lsg92}

280: A.\,I. Lukomskii, V.\,B. Shipilo and L.\,M. Gameza, Zh.\ Prik.\ Spek.\

281: {\bf57}, 100 (1992).

282: 

283: \bibitem{ksz75}

284: A.~Katzir, J.\,T. Suss, A.~Zunger and A.~Halperin, Phys.\ Rev.\ B

285: {\bf11}, 2370 (1975).

286: 

287: \bibitem{Kob94}

288: H.~Kobayashi, H.~Shibata and S.~Tagawa, Nucl.\ Inst.\ Meth.\ B

289: {\bf90}, 556 (1994).

290: 

291: \bibitem{khu73}

292: M.\,B. Khusidman, Sov.\ Phys.\ Sol.\ St.\ {\bf14}, 2791 (1973).

293: 

294: \bibitem{gkl89}

295: Yu.\,I. Galanov, F.\,V. Konusov and V.\,V. Lopatin, Izv.\ Vys.\ Uch.\

296: Zav.\ Fizika {\bf11}, 72 (1989).

297: 

298: \bibitem{gkl90}

299: Yu.\,I. Galanov, F.\,V. Konusov and V.\,V. Lopatin, Crys.\ Res.\

300: Tech.\ {\bf25}, 1343 (1990).

301: 

302: \bibitem{dzh69}

303: U.\,D. Dzhuzeev and P.\,E. Ramazanov, Izv.\ Vyss.\ Uch.\ Zav., Fizika

304: {\bf12}, 81 (1969).

305: 

306: \bibitem{emk81}

307: K.~Era, F.~Minami and T. Kuzuba, J.\ of Lum.\ {\bf24/25}, 71 (1981).

308: 

309: \bibitem{tbs94}

310: C.\,A. Taylor, S.\,W. Brown, V.~Subramaniam, S.~Kidner, S.\,C. Rand

311: and R.~Clarke, Appl.\ Phys.\ Lett.\ {\bf65}, 1251 (1994).

312: 

313: \bibitem{huf00}

314: P.\,R. Huffman \textit{et al.} Nature {\bf403}, 62 (2000).

315:     

316: \bibitem{combatbn}

317: AX05 grade h-BN, Carborundum Inc., Amherst NY. Certain trade names and

318: company products are mentioned in the text in order to adequately

319: specify the experimental procedure and equipment used.  In no case

320: does such identification imply recommendation or endorsement by NIST.

321: 

322: \bibitem{films}

323: D.\,M. McKinsey \textit{et al.} Nucl.\ Inst.\ Meth.\ B {\bf132}, 351

324: (1997).

325: 

326: \bibitem{Kuraray}

327: Y11 scintillating fiber purchased from Kuraray Co., Ltd.  8F, Mauzen

328: Building, 3-10, 2-Chome, Nihonbashi, Chuo-ku, Tokyo, 103-0027, Japan.

329: 

330: \bibitem{tripst}

331: ed.  A.\,B. Zahlan, The Triplet State, Cambridge University Press,

332: Cambridge, 1967.

333: 

334: \bibitem{Por71}

335: F. Porret and F. L{\"u}ty, Phys.\ Rev.\ Lett.\ {\bf26}, 843 (1971).

336: 

337: \bibitem{Bal83}

338: G. Baldacchini, G.\,P. Gallerano, U.\,M. Grassano, and F. L{\"u}ty, 

339: Rad.\ Eff.\ {\bf72}, 153 (1983).

340: 

341: \bibitem{kst94}

342: I.\,V. Kavetskaya, N.\,N. Sibel'din and V.\,A. Tsvetkov, JETP {\bf78},

343: 926 (1994).

344: 

345: \end{thebibliography}

346: \vspace{1.0in}

347: 

348: \begin{figure}[h]

349: \begin{center}

350: \epsfig{file=figure1.eps,width=3.0in}

351: \end{center}

352: \caption{Neutron shielding and light collection system.  Neutrons 

353: enter from left, are collimated by a h-BN collimator, pass through the 

354: detection region and exit through a second collimator.  The detection 

355: system consists of an expanded PTFE diffuse reflector with a thin 

356: layer of TPB evaporated onto the inner surface.  The light is 

357: transported out of the apparatus using two helically coiled 

358: wavelength shifting fibers which are coupled into PMTs at room 

359: temperature.}

360: \label{fig:fibercell}

361: \end{figure}

362: 

363: \vspace{0.5in}

364: \begin{figure}[h]

365: \begin{center}

366: \epsfig{file=figure2.eps,width=3.0in}

367: \end{center}

368: \caption{The time dependence of the luminescence signal with no

369: magnetic field (solid line) and when the magnetic field is energized

370: during the irradiation and deenergized 1275~s after the irradiation

371: ends (dashed line).}

372: \label{fig:nomagnet}

373: \end{figure}

374: 

375: \begin{figure}

376: \begin{center}

377: \epsfig{file=figure3.eps,width=3.0in}

378: \end{center}

379: \caption{Schematic diagram of the simplified apparatus used to verify the 

380: luminescence was originating from the boron nitride.  The beam stop 

381: has three 4~mm slits machined at a radial distance of 14~mm from the 

382: center of the beam stop to allow the light to enter the acrylic 

383: lightguide.  No unscattered neutrons can reach the acrylic lightguide.}

384: \label{fig:cleanapp}

385: \end{figure}

386: 

387: \begin{figure}

388: \begin{center}

389: \epsfig{file=figure4.eps,width=3.0in}

390: \end{center}

391: \caption{Luminescence signal using the setup shown in

392: figure~\ref{fig:cleanapp}.  The magnetic field is energized during the

393: irradiation and initial observation times and deenergized 1275~s after

394: the irradiation ends.}

395: \label{fig:cleandata}

396: \end{figure}

397: 

398: \begin{figure}

399: \begin{center}

400: \epsfig{file=figure5.eps,width=3.0in}

401: \end{center}

402: \caption{{\bf A}: The effect of lowering the magnetic field ($B$) in four 

403: steps as follows: a. $t=1275$~s, $B=100~\% \rightarrow 75~\%$;

404: b. $t=1700$~s, $B= 75~\% \rightarrow 50~\%$;

405: c. $t=2125$~s, $B= 50~\% \rightarrow 25~\%$;

406: d. $t=2550$~s, $B= 25~\% \rightarrow  0~\%$.  

407: {\bf B}: The magnetic field is slowly lowered

408: from its initial maximum value to zero over 2700~s. Just after this ramp 

409: ends, the field is rapidly raised to its maximum value beginning 

410: at point ``a'' and 570~s later, the field is lowered back to zero (point 

411: ``b'').}

412: \label{fig:storage}

413: \end{figure}

414: 

415: \begin{figure}

416: \begin{center}

417: \epsfig{file=figure6.eps,width=3.0in}

418: \end{center}

419: \caption{The time dependence

420: of the luminescence signal when the magnet is deenergized in less 

421: than one second (solid line) and the time dependence when the magnet 

422: is deenergized using a slower ramping speed of $\approx 200$~s (dashed 

423: line).}

424: \label{fig:quench}

425: \end{figure}

426: 

427: \begin{figure}

428: \begin{center}

429: \epsfig{file=figure7.eps,width=3.0in}

430: \end{center}

431: \caption{{\bf A}: The change in intensity of the luminescence signal

432: due to varying the magnitude of the magnetic field.  Each trace is for

433: a different initial magnetic field intensity.  In each case the magnet

434: is brought from its initial value to zero in 100~s.  {\bf B}: The

435: change in intensity of the luminescence signal due to varying the

436: temperature of the cell.  Each trace is for a different initial

437: temperature.  In each case the temperature is raised from its initial

438: value to 1~K while the magnet is kept at 100~\% of its full value.}

439: \label{fig:variation}

440: \end{figure}

441: 

442: \end{document}

443: 

444: 

445: 

446: