1: %\documentclass[aps,prb,preprint,groupedaddress]{revtex4}

2: %\documentclass[aps,prb,preprint,superscriptaddress]{revtex4}

3: \documentclass[aps,prb,twocolumn,groupedaddress,showkeys,showpacs]{revtex4}

4: 

5: \usepackage{multirow,amssymb,amsbsy,amsmath}

6: \usepackage{epsfig}

7: 

8: %\usepackage[active]{srcltx}

9: 

10: 

11: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

12: \newcommand {\nifour} {Ni$_{4}$}

13: \newcommand {\nimo} {\{Ni$_{4}$Mo$_{12}$\}}

14: \newcommand {\nitwo} {Ni$^{\text{II}}$}

15: \newcommand {\nifourfull} {[Mo$^{\text{V}}_{12}$O$_{30}$($\mu_2$-OH)$_{10}$H$_2$\{Ni$^{\text{II}}$(H$_2$O)$_3$\}$_4$]}

16: \newcommand{\op}[1]{%

17:     \fontdimen12\textfont3=2pt\fontdimen12\scriptfont3=1.4pt%

18:     \!\null\mathop{\vphantom{#1}\smash{#1}}\limits_{\sim}\null\!}

19: \newcommand{\xref}[1]{\protect\ref{#1}}

20: \newcommand{\figref}[1]{Fig.~\protect\ref{#1}}

21: \newcommand{\fmref}[1]{(\protect\ref{#1})}

22: \def\bra#1{\langle \, {#1} \, | \,}

23: \def\ket#1{\, | \, {#1} \, \rangle}

24: \newcommand{\braket}[2]{\langle \, {#1} \, | \, {#2} \, \rangle}

25: \newcommand{\vek}[1]{{\!\vec{\,#1}}}

26: \newcommand{\dint}{\text{d}}

27: %\newcommand{\new}[1]{\textbf{#1}}

28: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

29: 

30: 

31: 

32: \begin{document}

33: 

34: %\title{Field-dependent magnetic parameters in \nimo: Magnetostriction at the molecular level?}

35: \title{Observation of field-dependent magnetic parameters

36:   in the magnetic molecule \nimo}

37: 

38: 

39: \author{J\"urgen Schnack}

40: \email{jschnack@uos.de}

41: \author{Mirko Br\"uger}

42: %\email{mbrueger@uos.de}

43: \affiliation{Universit\"at Osnabr\"uck, Fachbereich Physik,

44: D-49069 Osnabr\"uck, Germany}

45: 

46: \author{Marshall Luban}

47: %\email{luban@ameslab.gov}

48: \author{Paul K\"ogerler}

49: %\email{kogerler@ameslab.gov}

50: \author{Emilia Morosan}

51: \author{Ronald Fuchs}

52: %\email{fuchs@ameslab.gov}

53: \affiliation{Ames Laboratory \& Department of Physics and Astronomy,

54: Iowa State University, Ames, Iowa 50011, USA}

55: 

56: 

57: \author{Robert Modler}

58: \affiliation{Johann Modler GmbH, Postfach 100462, D-63741

59:   Aschaffenburg, Germany}

60: 

61: 

62: \author{Hiroyuki Nojiri}

63: %\email{nojiri@imr.tohoku.ac.jp}

64: \affiliation{Institute for Materials Research, Tohoku

65:   University, Katahira 2-1-1, Sendai 980-8577, Japan}

66: 

67: \author{Ram C. Rai}

68: \author{Jinbo Cao}

69: \author{Janice L. Musfeldt}

70: \affiliation{Department of Chemistry, University of Tennessee,

71:   Knoxville, TN 37996, USA}

72: 

73: \author{Xing Wei}

74: \affiliation{National High Magnetic Field Laboratory, Florida State University,

75: Tallahassee, FL, 32310}

76: 

77: 

78: 

79: \date{\today}

80: 

81: \begin{abstract}

82:   We investigate the bulk magnetic, electron paramagnetic

83:   resonance, and magneto-optical properties of \nimo, a magnetic

84:   molecule with antiferromagnetically coupled tetrahedral

85:   \nitwo\ in a diamagnetic molybdenum matrix. The

86:   low-temperature magnetization exhibits steps at irregular

87:   field intervals, a result that cannot be explained using a

88:   Heisenberg model even if it is augmented by magnetic

89:   anisotropy and biquadratic terms.  Allowing the exchange and

90:   anisotropy parameters to depend on the magnetic field provides

91:   the best fit to our data, suggesting that the molecular

92:   structure (and thus the interactions between spins) may be

93:   changing with applied magnetic field.

94: \end{abstract}

95: 

96: \pacs{75.50.Xx,75.10.Jm,75.40.Cx}

97: \keywords{Magnetic Molecules, Anisotropy, Susceptibility, EPR, Magneto-optics}

98: \maketitle

99: 

100: 

101: 

102: 

103: 

104: 

105: 

106: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

107: \section{Introduction}

108: \label{sec-1}

109: 

110: Despite the enormous progress that has been made in

111: understanding magnetic molecules over the past

112: decade,\cite{SGC:Nat93,GCR:S94,Gat:AM94,Cor:NATO96,CGS:JMMM99,WSK:IO99,MKD:CCR01,MLS:CPC01,WAH:Nature02}

113: it is still a challenge to deduce the underlying microscopic

114: spin Hamiltonian. Mn$_{12}$-acetate is a good example of this

115: problem.\cite{Lis:ACB80,SGC:Nat93,TLB:Nature96,Cor:NATO96,JLB:PRL00,FWK:PRB01}

116: Although known for almost twenty years, only extensive

117: investigation elucidated the model

118: parameters.\cite{RJS:PRB02,CSO:PRB04} Small magnetic molecules,

119: with their simpler chemical and magnetic structure, can

120: therefore provide an important opportunity to understand the

121: dependence of magnetic observables on model

122: parameters.\cite{WHM:PRB98,KWM:PRB03,LBB:PRB03}

123: 

124: In this work we report our joint experimental and theoretical

125: efforts to understand the behavior of \nifourfull, henceforth

126: abbreviated as \nimo, a magnetic molecule which is comprised of

127: \nitwo\ centers positioned at the nucleophilic sites of an

128: $\epsilon$-Keggin cluster forming an almost ideal

129: tetrahedron.\cite{MBK:IC00} In contrast to several other nickel

130: compounds which exhibit

131: ferromagnetic\cite{ABB:CEJ02,KWM:PRB03,YWH:P03,MHR:EJIC04,KCW:DT04}

132: or mixed coupling,\cite{DLM:IC05} the Ni centers of this molecule

133: are antiferromagnetically coupled, as is also the case for certain

134: Ni-$2\times 2$-grid molecules.\cite{WHM:PRB98} Because the structure

135: of \nimo\ is almost perfectly tetrahedral one might anticipate that

136: the magnetic energy levels are reasonably well described by an

137: isotropic Heisenberg model with a single antiferromagnetic exchange

138: parameter.\cite{MBK:IC00} Such a Hamiltonian can be expressed in

139: terms of the total spin:

140: %--------------------------------------------------------

141: \begin{eqnarray}

142: \label{E-1-1}

143: \op{H}

144: &=&

145: -2 J

146: \sum_{u<v}\;

147: \op{\vec{s}}(u) \cdot \op{\vec{s}}(v)

148: +

149: g \mu_B \vec{B}\cdot \sum_{u}\;\op{\vec{s}}(u)

150: \\

151: &=&

152: -J

153: \left[

154: \op{\vec{S}}^2-4s(s+1)

155: \right]

156: +

157: g \mu_B B \;\op{S}_z

158: \nonumber

159: \ ,

160: \end{eqnarray}

161: %--------------------------------------------------------

162: where $\op{\vec{s}}(u)$ is a single-spin operator at site $u$

163: and $\op{\vec{S}}$ is the total spin operator. The spin quantum

164: number of each \nitwo\ ion in an octahedral ligand field is

165: $s=1$. For antiferromagnetic coupling ($J<0$) the resulting

166: low-temperature magnetization curve $\mathcal{M}(B)$ that

167: follows from \fmref{E-1-1} displays four steps before reaching

168: saturation. These steps occur at the magnetic fields

169: %--------------------------------------------------------

170: \begin{eqnarray}

171: \label{E-1-2}

172: B_{S\rightarrow(S+1)}

173: &=&

174: -

175: \frac{2 J}{g \mu_B} (S+1)

176: \ ,

177: \end{eqnarray}

178: %--------------------------------------------------------

179: for $S=0, 1, 2, 3$, where the lowest Zeeman-split levels of

180: adjacent multiplets cross. In particular, from \fmref{E-1-2} it

181: follows that the level crossing fields are uniformly

182: spaced.\cite{ACC:ICA00,ScL:PRB01} In stark contrast to the

183: expectation of uniformly spaced crossing fields, the

184: experimental magnetization ${\mathcal M}(B)$ curve of \nimo\

185: features non-equidistant steps at 4.5, 8.9, 20.1, and 32~T. Even

186: assuming an anisotropic Hamiltonian\cite{Brueger03} with two

187: exchange couplings and biquadratic terms as done for other

188: \nitwo-compounds,\cite{WHM:PRB98,KWM:PRB03} we were unable to

189: account for the specific sequence of steps in the

190: low-temperature magnetization of \nimo.  In order to provide a

191: comprehensive picture of the unusual high-field magnetic

192: behavior of \nimo, we have been led to invoke field-dependent

193: exchange and anisotropy parameters.  We argue that this

194: dependence emanates from changes in molecular structure (and

195: thus the overlap of those atomic orbitals that determine the

196: exchange interactions as well as the coordination geometries

197: which affect the electronic single-ion properties) with applied

198: magnetic field. The magneto-optical response of \nimo\ supports

199: a small change in the \nitwo\ coordination environment and the

200: associated electronic single-ion properties.

201: 

202: 

203: 

204: 

205: 

206: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

207: \section{Chemical and Crystal Structure}

208: \label{sec-6}

209: 

210: %===================    figure   =================================

211: \begin{figure}[!ht]

212: \begin{center}

213: %\epsfig{file=Paul-ni4mo12-neu2.eps,width=55mm}

214: \epsfig{file=fig-1.ps,width=55mm}

215: \vspace*{1mm}

216: \caption[]{(color online) Ball-and-stick representation of the

217:   \nimo\ molecule (Ni: numbered big spheres; Mo: medium size

218:   spheres; O: small spheres; H: not shown) emphasizing a

219:   slightly stretched Ni$^{\text{II}}_4$ pyramid with a

220:   near-equilateral triangle base (Ni ions labeled 2, 3, and 4)

221:   and an elevated apex (Ni ion labeled 1).  Ni-Ni distances:

222:   $d_{12}=6.700(5)$~\AA, $d_{13}=d_{14}=6.689(1)$~\AA,

223:   $d_{23}=d_{24}=6.616(1)$~\AA,

224:   $d_{34}=6.604(1)$~\AA.\cite{MBK:IC00}}

225: \label{F-1}

226: \end{center}

227: \end{figure}

228: %===================    figure ===============================

229: 

230: The neutral \nimo\ cluster is isolated in the form of crystals

231: of

232: [Mo$^{\text{V}}_{12}$O$_{30}$($\mu_2$-OH)$_{10}$H$_2$\{Ni$^{\text{II}}$(H$_2$O)$_3$\}$_4$]$\cdot$14~H$_2$O

233: and is based on the diamagnetic, highly-charged

234: $\epsilon$-Keggin anion

235: [Mo$^{\text{V}}_{12}$O$_{38}$($\mu_3$-OH)$_{2}$]$^{18-}$, built

236: up from four edge-sharing \{Mo$_3$\} groups (each consisting of

237: three edge-sharing MoO$_6$ octahedra).  Within the

238: $\epsilon$-Keggin framework, the Mo positions form six

239: Mo$^{\text{V}}_2$ groups with short Mo-Mo single bonds.  The

240: $\epsilon$-Keggin structure is formally derived from the common

241: $\alpha$-Keggin isomer by rotating all four {Mo$_3$} groups by

242: 60$^\circ$, preserving the T$_d$ symmetry. In \nimo, four

243: [\nitwo(H$_2$O)$_3$]$^{2+}$ groups are coordinated each to three

244: (unprotonated) $\mu_2$-oxo centers that interlink the Mo

245: positions of the Mo$^{\text{V}}_2$ groups (\figref{F-1}). This

246: results in a octahedral O$_3$\nitwo(H$_2$O)$_3$ coordination

247: environment with all-trans-positioned oxo and water ligands and

248: nearly identical Ni-O distances (Ni-($\mu_2$-O):~2.05~\AA,

249: Ni-OH$_2$:~2.06~\AA).  This capping of the {Mo$_3$} groups of

250: the $\epsilon$-Keggin fragment by four \nitwo\ positions

251: produces a near-regular \nifour\ tetrahedron. Contrary to many

252: other tetrahedral \nifour\ structures in which Ni pairs are

253: connected by mononuclear bridging centers, in \nimo\ each Ni

254: pair is interconnected via one -O(-Mo-)$_2$O- bridging motif

255: serving as a superexchange pathway (see \figref{F-1} where one

256: Ni-O-Mo-O-Ni pathway is highlighted by dark bonds).  The

257: geometry of each of these pathways is therefore characterized by

258: four bond lengths, three bond angles, and two dihedral angles,

259: as opposed to \nifour-type structures comprising mononuclear

260: linker groups (two bond lengths, one bond angle).  Importantly

261: however, the molecular geometry of \nimo\ slightly deviates from

262: the $T_d$-symmetric ideal, resulting in a slightly stretched

263: \nifour\ pyramid with elongated Ni-Ni distances between the Ni

264: positions of a basal Ni$_3$ plane and one apex.  While the

265: crystallographic symmetry operations result in the molecular

266: point group $C_s$, the actual geometry is virtually of $C_{3v}$

267: symmetry, and the geometric parameters for all

268: Ni-O(-Mo-)$_2$O-Ni pathways correspondingly fall into two sets.

269: Within these two sets, the individual bond lengths and angles

270: display minimal deviations (typically $< 0.8$~\%) from the

271: respective averages: Intra-basal Ni-Ni contacts are

272: characterized by $\langle$Ni-O$\rangle$ = 2.05\AA,

273: $\langle$Mo-O$\rangle$ = 1.95~\AA, $\langle$Ni-O-Mo$\rangle$ =

274: 135.4$^\circ$, and $\langle$Mo-O-Mo$\rangle$ = 89.6$^\circ$. For

275: Ni-Ni contacts between the apex to the base positions average

276: values of $\langle$Ni-O$\rangle$ = 2.03~\AA,

277: $\langle$Mo-O$\rangle$ = 1.95~\AA, $\langle$Ni-O-Mo$\rangle$ =

278: 137.9$^\circ$, and$\langle$<Mo-O-Mo$\rangle$ = 94.6$^\circ$ are

279: found.  As the geometric parameters do not vary significantly

280: within the intra-basal and within the apex-basal Ni-Ni contacts, 

281: we do not take into account slight deviations from the idealized

282: $C_{3v}$ symmetry, and use only exchange constants $J$ and $J'$

283: in this paper (see \figref{F-2}).

284: 

285: %===================    figure   =================================

286: \begin{figure}[!ht]

287: \begin{center}

288: %\epsfig{file=Ni4-structure.eps,width=45mm}

289: \epsfig{file=fig-2.eps,width=45mm}

290: \vspace*{1mm}

291: \caption[]{Simplified structure of \nifour: the superexchange

292:   interactions $J'$ and $J$ are represented by dashed and solid

293:   lines, respectively.}

294: \label{F-2}

295: \end{center}

296: \end{figure}

297: %===================    figure ===============================

298: 

299: It should be added that magnetic

300: exchange through polyoxomolybdates frameworks, especially if

301: more than one possible pathway exists for each contact, has been

302: found to be fairly insensitive to the separation distance of the

303: pair of spin centers. Also, for similar systems based on

304: mononuclear linker groups the bond angles have a stronger effect

305: on the exchange energies than the contact

306: distance.\cite{HSH:IC95,WHM:PRB98,EFK:P99,CMP:CC01,ABB:CEJ02,KWM:PRB03,YWH:P03,MHR:EJIC04}

307: Due to the presence of crystal water molecules in the

308: solid-state structure of \nimo$\cdot$14~H$_2$O which space the

309: cluster entities apart, the closest intermolecular Ni$\cdots$Ni

310: distance in the solid state exceeds 7.15~\AA, rendering

311: inter-molecular (dipole-dipole) magnetic exchange insignificant.

312: 

313: 

314: 

315: 

316: 

317: 

318: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

319: \section{Experimental results}

320: \label{sec-2}

321: 

322: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

323: \subsection{Magnetic properties of \nimo}

324: \label{sec-2-1}

325: 

326: %===================    figure   =================================

327: \begin{figure}[!ht]

328: \begin{center}

329: %\epsfig{file=ni4-T-MdB-combined.eps,width=70mm}

330: \epsfig{file=fig-3.eps,width=70mm}

331: \vspace*{1mm}

332: \caption[]{(color online) Low-field susceptibility: Experimental

333:   data are given by black squares whereas the solid curve shows

334:   the result assuming a simple Heisenberg Hamiltonian

335:   \fmref{E-1-1} with $J/k_B=-3.4$~K and an isotropic

336:   spectroscopic splitting factor $g=2.25$. The inset shows a

337:   close-up view of the low-temperature region for different

338:   magnetic fields.}

339: \label{F-3}

340: \end{center}

341: \end{figure}

342: %===================    figure ===============================

343: 

344: Figure~\xref{F-3} displays the magnetic susceptibility

345: $\mathcal{M}(B)$ of \nimo\ as a function of temperature.  These

346: data were collected on a SQUID magnetometer (Quantum Design

347: MPMS-5) at various magnetic fields in a temperature range of $2

348: - 290$~K. Using a simple Heisenberg Hamiltonian \fmref{E-1-1}

349: one obtains $J/k_B=-3.4$~K and an isotropic $g$-factor of

350: $g=2.25$, see also Ref.~\onlinecite{MBK:IC00}.  The model fit

351: given by the solid curve in \figref{F-3} is acceptable, although

352: the low-temperature behavior is not well reproduced, as can be

353: seen in the inset of \figref{F-3}, where the magnetic

354: susceptibility for various field values and low temperatures is

355: displayed. We suggest that the deviation is not only due to the

356: presence of impurities but also due to anisotropic as well as

357: biquadratic terms in the Hamiltonian which are known to be

358: needed to model the low-temperature behavior of

359: \nitwo-compounds.\cite{WHM:PRB98,KWM:PRB03}

360: 

361: %===================    figure   =================================

362: \begin{figure}[!ht]

363: \begin{center}

364: %\epsfig{file=ni4-B-M-combined.eps,width=70mm}

365: \epsfig{file=fig-4.eps,width=70mm}

366: \vspace*{1mm}

367: \caption[]{(color online) Magnetization: Experimental data are

368:   given by squares (dark symbols -- NHMFL, $T=0.44$~K; light

369:   symbols -- OHMFL, $T=0.40$~K). The theoretical magnetization

370:   assuming a simple Heisenberg Hamiltonian \fmref{E-1-1} with

371:   $J/k_B=-3.4$~K and an isotropic spectroscopic splitting factor

372:   $g=2.25$ is given by a solid curve for $T=0.44$~K. The inset

373:   shows the experimental differential magnetization $\dint

374:   {\mathcal M}/\dint B$ as data points as well as the

375:   theoretical $\dint {\mathcal M}/\dint B$ (solid curve) using

376:   the same parameters as above.  }

377: \label{F-4}

378: \end{center}

379: \end{figure}

380: %===================    figure ===============================

381: 

382: High-field magnetization measurements are a valuable tool to

383: extract information on the spin Hamiltonian which is not

384: accessible from magnetization measurements on commercial SQUID

385: magnetometers.  The high-field magnetization for a powder sample

386: of \nimo\ has been independently measured in pulsed magnetic

387: fields at the facility of the National High Magnetic Field

388: Laboratory (NHMFL) at Los Alamos as well as at the Okayama High

389: Magnetic Field Laboratory (OHMFL) by using a standard inductive

390: method (maximum at NHMFL $B=60$~T, whereas the maximum at OHMFL

391: is $B=40$~T, $dB/dt=10000\dots 15000$~T/s).  The results of

392: these two measurements are in very close agreement. No

393: hysteresis is found between up and down sweep runs, indicating

394: thermal equilibrium behavior of $\mathcal{M}$ vs. $B$.

395: 

396: 

397: Figure \xref{F-4} shows the magnetization as a function of

398: applied external magnetic field at $T=0.44$~K.  Strikingly, four

399: non-equidistant steps are observed in the magnetization. These

400: steps are found near 4.5, 8.9, 20.1, and 32~T.  Saturation of

401: the magnetization is not observed until 60~T. For comparison, we

402: show the expected response of a simple Heisenberg Hamiltonian

403: (Eq.~\fmref{E-1-1}, ideal tetrahedron) for the model parameters

404: extracted above. Note that this model predicts equidistant steps

405: in the magnetization at 4.5, 9.0, 13.5, and 18.0~T with

406: saturation above the fourth step. The drastic deviation between

407: the predictions of Eqn.~\fmref{E-1-1} and the experimental

408: results cannot be the result of heating via the magnetocaloric

409: effect. Although observed in other compounds, such an effect

410: would only smear out the steps but not shift the step positions.

411: In addition sample heating or cooling in a varying field is

412: often accompanied by hysteresis, which is absent in this

413: measurement.\footnote{The interested reader is referred to the

414:   extensive literature on ``butterfly hysteresis loops'' in

415:   molecular magnets, e.g.

416:   Refs.~\onlinecite{CWM:PRL00,WKS:PRL02,WKS:PRL03}.}  The

417: observed step positions thus constitute a challenge not only to

418: the simple Heisenberg model given by \eqref{E-1-1}, but also to

419: more elaborate models \eqref{E-3-1-1} incorporating anisotropy

420: terms and biquadratic exchange. We suggest that these models

421: should be extended to include field-dependent parameters in

422: order to account for our experimental results.  This will be

423: discussed in Sec.~\xref{sec-4}.

424: 

425: 

426: 

427: 

428: 

429: 

430: 

431: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

432: \subsection{EPR response of \nimo}

433: \label{sec-2-3}

434: 

435: 

436: %===================    figure   =================================

437: \begin{figure}[!ht]

438: \begin{center}

439: %\epsfig{file=ni4-ESR.eps,width=70mm}

440: \epsfig{file=fig-5.eps,width=70mm}

441: \vspace*{1mm}

442: \caption[]{(color online) The figure shows the dependence of the

443:   observed EPR resonance frequencies P1 (circles), P2

444:   (triangles), and P3 (squares) on the magnetic field. The lines

445:   provide linear fits to the data. The data were taken at

446:   $T=4.2$~K.}

447: \label{F-5}

448: \end{center}

449: \end{figure}

450: %===================    figure ===============================

451: 

452: Figure \xref{F-5} displays the results of our Electron

453: Paramagnetic Resonance (EPR) measurements. The transmission of a

454: powder sample was determined as a function of applied magnetic

455: field in a frequency range from 95~GHz to 381.5~GHz. The figure

456: shows the dependence of the observed EPR resonance frequencies

457: P1 (circles), P2 (triangles), and P3 (squares) on magnetic

458: field. One immediately notices that two different slopes can be

459: assigned to the data, one corresponding to $\Delta M = 1$ and

460: another one corresponding to forbidden transitions with $\Delta

461: M = 2$. These dependencies can be qualitatively -- and to some

462: extent quantitatively -- explained by looking at the Zeeman

463: level scheme of the simple Heisenberg model \fmref{E-1-1} as it

464: is schematically depicted in \figref{F-6}.

465: 

466: %===================    figure   =================================

467: \begin{figure}[!ht]

468: \begin{center}

469: %\epsfig{file=ni4-levels.eps,width=55mm}

470: \epsfig{file=fig-6.eps,width=55mm}

471: \vspace*{1mm}

472: \caption[]{(color online) The figure shows schematically the

473:   Zeeman level splittings in a pure Heisenberg model together

474:   with the assignments of allowed and forbidden transitions.  }

475: \label{F-6}

476: \end{center}

477: \end{figure}

478: %===================    figure ===============================

479: 

480: The strongest transition which we observe in the spectra is the

481: allowed transition P3 with $\Delta M = 1$. Since the applied

482: temperature is of the order of the coupling, this transition is

483: actually a sum of several transitions. At low-field values it is

484: dominated by the transition between ($S=1, M=-1$) and ($S=1,

485: M=0$), whereas at higher fields the dominant contribution stems

486: from the transition between ($S=2, M=-2$) and ($S=2, M=-1$).

487: The dependence of the resonance frequency of P3 on the applied

488: field suggests that the zero-field splitting in the triplet is

489: small.

490: 

491: P1 is a low-field transition which connects ($S=1, M=-1$) and

492: ($S=1, M=+1$) and thus should be forbidden. In the spectra its

493: strength is much weaker than that of P3. We believe that this

494: transition appears due to mixing of $\op{S}_z$ eigenstates that would

495: arise from anisotropic contributions to the Hamiltonian. The

496: line which is plotted through the data points suggests that the

497: zero-field splitting between ($S=1, M=-1$) and ($S=1, M=+1$)

498: is probably small, although, the experimental data points --

499: which extend only down to 1.3~T -- would also allow a somewhat

500: bigger splitting, especially since the lowest-lying P1 data

501: points appear to deviate from a straight line.

502: 

503: P2 is another rather weak forbidden transition which shares the

504: slope with P1. We believe that this transition connects ($S=2,

505: M=-2$) and ($S=1, M=0$). This transition is not observable below

506: about 7~T due to the fact that this transition occurs only when

507: the mixing of $\op{S}_z$ eigenstates is sufficiently strong

508: which is the case around the level crossing at 9~T.  The

509: dependence of this transition on temperature, i.e. on the

510: thermal occupation of the level with ($S=2, M=-2$) is small. The

511: extrapolation of the field dependence allows one to deduce an

512: approximate isotropic Heisenberg coupling from the zero-field

513: energy separation of the triplet and the pentuplet. From

514: Eq.~\fmref{E-1-1} one deduces that $E(S=2)-E(S=1)=-4J$, thus one

515: obtains $J\approx -3.4$~K, which is in very good agreement with

516: the exchange constants deduced from our susceptibility

517: measurements. The spectroscopic splitting factor can be

518: determined from the slopes of P3 and P1 in \figref{F-5} to be

519: $g=2.23\pm 0.03$, which is also in good agreement with the value

520: deduced from the high-temperature behavior of the

521: susceptibility. It is noteworthy that the high field data of

522: transition P3 have a smaller slope which, if fitted alone

523: (dashed-dotted line in \figref{F-5}), suggests $g\approx 2.11\pm

524: 0.03$. However, we want to point out that these considerations

525: are done on the basis of the simple Heisenberg model

526: \fmref{E-1-1}, that does not take into account that a realistic

527: Hamiltonian has to contain anisotropic terms.  A more detailed

528: explanation is given at the end of sec.~\xref{sec-4}.

529: 

530: 

531: Summarizing this part, we find that the zero-field splitting is

532: small. We explain this observation by the fact that the local

533: principal axes of the Ni-ions point in different directions

534: (radially outwards), and thus the average global anisotropy is

535: small. Our use of a powder sample led to additional averaging.

536: 

537: 

538: 

539: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

540: \subsection{Optical properties of \nimo}

541: \label{sec-2-4}

542: 

543: 

544: 

545: %===================    figure   =================================

546: \begin{figure}[!ht]

547: %\includegraphics[width=70mm]{Ni4conductivity3.EPS}

548: \includegraphics[width=80mm]{fig-7.eps}

549: \vskip -0.2cm \caption{(color online) Optical conductivity of a

550:   \nimo\ pellet at 4.2~K (solid curve) and 300~K (dashed curve),

551:   calculated from reflectance measurements (inset) by

552:   Kramers-Kronig analysis. The energy range of our

553:   magneto-optical work is indicated by the arrow.}

554: \label{F-7}

555: \end{figure}

556: %===================    figure ===============================

557: 

558: Figure \ref{F-7} displays the optical conductivity of \nimo.

559: These experiments were carried out on a pressed powder sample

560: using a Lambda-900 grating spectrometer equipped with

561: reflectance stage and cryostat.\cite{CWM:PRB04} \nimo\ is a

562: semiconductor with an optical gap of $\sim$0.6 eV. Based upon

563: comparisons with chemically similar nicklates as well as

564: existing electronic calculations,\cite{PYH:JAP05,PBS:PT05} we

565: assign the excitations centered at 1.5 eV as on-site nickel $d$

566: to $d$ transitions, likely activated by modest hybridization

567: with the coordinating oxygens.  These excitations take place in

568: both majority and minority channels according to recent DFT

569: calculations.\cite{PBS:PT05} The features above 3 eV are

570: assigned as O $p$ to Ni $d$ charge transfer excitations. The

571: energy range of our magneto-optical investigation is also shown

572: in Fig. \ref{F-7}, providing a preview of the physical origin

573: of the field-induced spectroscopic changes, discussed below.

574: 

575: %===================    figure   =================================

576: \begin{figure}[!ht]

577: %\includegraphics[width=70mm]{Ni4MagnetoOptical4.EPS}

578: \includegraphics[width=80mm]{fig-8.ps}

579: \vskip -0.2cm

580: \caption{(color online) (a) The normalized magneto-optical

581:   response, R(B)/R(B = 0 T), of \nimo\ pressed powder in an

582:   applied magnetic field from 0 to 32 T at 4.2 K. 1 T steps are

583:   shown. No hysteresis is observed.  (b) Absolute value of

584:   the integrated area of the magneto-optical reflectance ratio

585:   feature as a function of applied magnetic field (solid

586:   symbol).  Results from the integrated optical conductivity

587:   data are similar.  The inset shows the change in the optical

588:   conductivity of \nimo\ with magnetic field, $\sigma_1$($B$ =

589:   30 T)/$\sigma_1$($B$ = 0 T).  The $B$ = 0 T curve ratios data

590:   taken before and after the field sweep, giving an indication

591:   of overall spectral reproducibility.}

592: \label{F-8}

593: \end{figure}

594: %===================    figure ===============================

595: 

596: Figure \ref{F-8}(a) displays the magneto-optical response of

597: \nimo\ as a function of magnetic field from 0 to 32~T at 4.2~K.

598: These experiments were carried out on the same pressed powder

599: sample using a grating spectrometer (0.8 - 3.5 eV) equipped with

600: InGaAs and CCD detectors and a 33~T steady field resistive

601: magnet at the NHMFL in Tallahassee, FL.\cite{CWM:PRB04} The

602: reflectance ratio, $R(B)/R(B = 0)$, is a normalized response and

603: highlights changes in the optical properties with applied

604: magnetic field.  With increasing field, the reflectance of

605: \nimo\ decreases by $\sim$2\%. It is notable that this effect

606: occurs in the visible spectral range, hence the name

607: ``magnetochromism". Based upon the aforementioned peak

608: assignments in the optical conductivity spectrum in \figref{F-7}

609: and the $\sigma_1$($B$)/$\sigma_1$($B$ = 0 T) ratio data in the

610: inset of \figref{F-8}(b), we attribute the observed

611: magnetochromic effect to a field-induced modification of the Ni

612: $d$ to $d$ on-site excitation. Distortion of the

613: pseudo-octahedral \nitwo\ crystal field environment is a

614: plausible driver. Note that this is a local, molecular-level

615: distortion rather than a bulk effect.  No field-induced changes

616: were observed on the leading edge of the O $p$ to Ni $d$ charge

617: transfer bands above 3~eV.

618: 

619: 

620: We quantify the magneto-optical effect in \nimo\ by plotting the

621: absolute value of the integrated intensity of the reflectance

622: ratio as a function of applied magnetic field

623: (\figref{F-8}(b)).\footnote{Similar results are obtained using

624:   the optical conductivity ratio, $\sigma_1(B)/\sigma_1(0)$.}

625: The magneto-optical effect is small at low fields, becomes

626: appreciable above 10~T, and continues to grow for $B\ge 30$~T.

627: There is no evidence of saturation to 33 T. The overall rising

628: magnetochromic response can be fit with several different

629: functions including a cubic polynomial in $|B|$ and a simple

630: exponential, suggesting a likely functional form for the

631: field-dependence of the magnetic parameters in \nimo, described

632: below.  We propose that the applied magnetic field interacts

633: with the spin centers, deforming the local structure around the

634: \nitwo\ sites.  This process modifies the crystal field

635: environment, the result of which is a field-induced change in

636: electronic structure (\figref{F-8}~(a)).  Thus, these

637: spectroscopic results support the picture of field-dependent

638: exchange and anisotropy terms in the spin Hamiltonian of \nimo\

639: that derive from magnetoelastic (and consequent spin-orbit)

640: interactions.  Magneto-optical effects due to field-induced

641: changes in local structure have also been observed in other

642: materials including (CPA)$_2$CuBr$_2$ and

643: K$_2$V$_3$O$_8$.\cite{WCM:PRB,rai05}

644: 

645: 

646: 

647: 

648: 

649: 

650: 

651: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

652: \section{Theoretical Models}

653: \label{sec-3}

654: 

655: In order to understand the magnetic properties of \nimo\ we

656: adopt the following general Hamiltonian that has been used for

657: other \nitwo-compounds,\cite{WHM:PRB98,KWM:PRB03}

658: %--------------------------------------------------------

659: \begin{eqnarray}

660: \label{E-3-1-1}

661: \op{H}

662: &=&

663: \op{H}_{\text{H}}

664: +

665: \op{H}_{\text{ani}}

666: +

667: \op{H}_{\text{biq}}

668: +

669: \op{H}_{\text{Z}}

670: \ ,\ \text{where}

671: \\

672: \label{E-3-1-1-A}

673: \op{H}_{\text{H}}

674: &=&

675: -

676: 2

677: \sum_{u<v}\;

678: J_{uv}

679: \op{\vec{s}}(u) \cdot \op{\vec{s}}(v)

680: \\

681: \label{E-3-1-1-B}

682: \op{H}_{\text{ani}}

683: &=&

684: D

685: \left[

686: \sum_{u}\;

687: (\vec{e}_r(u) \cdot \op{\vec{s}}(u))^2

688: -\frac{8}{3}

689: \right]

690: \\

691: \label{E-3-1-1-C}

692: \op{H}_{\text{biq}}

693: &=&

694: -

695: 2

696: \sum_{u<v}\;

697: j_{uv}

698: \left(\op{\vec{s}}(u) \cdot \op{\vec{s}}(v)\right)^2

699: \\

700: \label{E-3-1-1-D}

701: \op{H}_{\text{Z}}

702: &=&

703: g\, \mu_B\, \vec{B} \cdot

704: \sum_{u}\;

705: \op{\vec{s}}(u)

706: \ .

707: \end{eqnarray}

708: %--------------------------------------------------------

709: Here $\op{H}_{\text{H}}$ denotes the Heisenberg Hamiltonian and

710: $\op{H}_{\text{ani}}$ describes the single-site anisotropic

711: (ligand field) contribution, which is compatible with the

712: approximate tetrahedral symmetry of \nimo.  The unit vector

713: $\vec{e}_r(u)$, which points radially outwards, serves as a

714: local anisotropy axis for site $u$. The term $8/3$ is convenient

715: in order to render the Hamiltonian traceless.

716: $\op{H}_{\text{biq}}$ represents biquadratic terms, which are

717: the next higher order compared to the Heisenberg

718: Hamiltonian,\cite{BeG90} and $\op{H}_{\text{Z}}$ is the Zeeman

719: term. We employ a single spectroscopic splitting factor $g$

720: since the $g$-tensor anisotropy was found to be very small for

721: the present system; a similar result has been found for \nitwo\

722: squares.\cite{WHM:PRB98} We also assume that a possible

723: anisotropic exchange between the \nitwo\ centers is small

724: because the orbital contribution to the ground state is

725: small.\cite{BBC:CPC03}

726: 

727: In the following sections, we simplify \mbox{$J_{uv} = J$} and

728: \mbox{$j_{uv} = j$}, or, if we use two different constants,

729: \mbox{$J' = J_{12} =J_{13} = J_{14}$}, \mbox{$J= J_{23} =J_{24}

730:   = J_{34} $} and \mbox{$j' = j_{12} =j_{13} = j_{14}$} ,

731: \mbox{$j= j_{23} =j_{24} = j_{34} $}, as illustrated in

732: \figref{F-2}. We also account for impurity effects (additional

733: paramagnetic \nitwo\ ions) and their batch-to-batch variation by

734: adding a paramagnetic term to the Hamiltonian \fmref{E-3-1-1}.

735: 

736: We have made numerous attempts to model the experimental

737: magnetization curve of \nimo\ (Fig. \ref{F-4}) with

738: field-independent values of $J$ and $D$, as described

739: previously.  Despite these efforts, an explanation for the

740: non-uniform spacing of the crossing fields has not been

741: forthcoming.  Assuming a Hamiltonian with two exchange couplings

742: and anisotropic as well as biquadratic terms as given in

743: \eqref{E-3-1-1} did not result in a satisfactory description of

744: all magnetic observables on a common footing.\cite{Brueger03}

745: Therefore, we extended our model to allow $J$ and $D$ to vary

746: with applied magnetic field. The possible field-induced

747: distortion of the crystal field environment around the Ni

748: centers motivates this ansatz.

749: 

750: 

751: 

752: 

753: 

754: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

755: \section{Field-dependent model parameters}

756: \label{sec-4}

757: \subsection{Low-field properties of \nimo}

758: 

759: It is not \emph{a priori} clear how the model parameters should

760: depend on the magnetic field since this dependence is indirect.

761: The exchange couplings, the biquadratic contributions as well as

762: the anisotropy result from the electronic structure, and are

763: (complicated) functions of orbital overlaps, lattice stiffness,

764: and spin-orbit coupling. If the electronic structure is

765: noticeably altered by an applied field it is probable that the

766: parameters entering Hamiltonian \fmref{E-3-1-1} are also

767: changed, but very likely in a highly non-linear manner. To the

768: best of our knowledge, so far such dependencies were only

769: observed in certain Mn grid molecules,\cite{WZT:PRL02} where the

770: anisotropy changes sign at a level crossing.  It appears

771: plausible to us that besides continuous variations, the

772: parameters can also change abruptly since the molecular

773: structure may relax into new ground states at certain field

774: values. Therefore, one can anticipate that the field-dependence

775: of the model parameters might only be piecewise analytic between

776: the abrupt changes. In this section we therefore investigate

777: simple piecewise parameterizations of the Hamiltonian. We also

778: neglect a possible anisotropic field dependence of the

779: parameters, i.e. the exchange parameters of different bonds

780: would be modified differently, and therefore the effect should

781: depend on the relative orientation of the molecule with respect

782: to the field.

783: 

784: %===================    figure   =================================

785: \begin{figure}[!ht]

786: \begin{center}

787: %\epsfig{file=ni4-T-MdB-B-best.eps,width=65mm}

788: \epsfig{file=fig-9.eps,width=65mm}

789: \vspace*{1mm}

790: \caption[]{(color online) Low-field susceptibility: The measured

791:   data are given by symbols.  The fits for $B=0.5$~T are

792:   obtained using $J/k_B=-3.2$~K, $J'/k_B=-3.1$~K, $j/k_B=1.6$~K,

793:   $j'/k_B=0$, $D/k_B=-1.0$~K. The data for field values of

794:   $B=3.5$~T up to $5$~T are approximated using $J/K=-3.2$~K,

795:   $J'/k_B=-3.2$~K, $j/k_B=1.5$~K, $j'/k_B=0$, and

796:   $D/k_B=-3.2$~K. $g=2.195$ in all cases. The estimated impurity

797:   concentrations are 0.216 and 0.146 individual \nitwo\ ions per

798:   \nimo\ molecule for sample 1 and 2, respectively.  The

799:   simulated data are averaged over 100 orientations.}

800: \label{F-9}

801: \end{center}

802: \end{figure}

803: %===================    figure ===============================

804: 

805: It turns out that the low-field susceptibility versus

806: temperature data, shown in \figref{F-9}, can be modeled by two

807: sets of parameters. The magnetization of two different samples

808: of \nimo\ at $B=0.5$~T can be understood assuming

809: $J/k_B=-3.2$~K, $J'/k_B=-3.2$~K, $j/k_B=1.6$~K, $j'=0$,

810: $D/k_B=-1.0$~K together with the separate term to account for

811: the paramagnetic impurities (free \nitwo\ ions) contained in the

812: samples (parameterization \textbf{1}). The parameters of the

813: Hamiltonian compare nicely to those of $2\times 2$--\nitwo\ grid

814: molecules.\cite{WHM:PRB98} The sign of the anisotropy as well as

815: that of the biquadratic term signal the same behavior, only the

816: absolute value of the biquadratic term is somewhat larger.  The

817: resulting zero-field splitting of the triplet is only 0.15~K, in

818: very good agreement with our EPR results. The fact that the

819: biquadratic term is one order of magnitude larger than usually

820: observed ($j\propto J/100$)\cite{HuO:PRL64,BeG90,WHM:PRB98}

821: signals that the bonds in the basal triangle of the molecule

822: might be rather soft. The other four susceptibility

823: measurements, \figref{F-9}, which are obtained from $B=3.5$~T to

824: $5$~T, can be better approximated by a second set of parameters:

825: $J/k_B=-3.2$~K, $J'/k_B=-3.1$~K, $j/k_B=1.5$~K, $j'/k_B=0$, and

826: $D/k_B=-3.2$~K (parameterization \textbf{2}).  These parameters

827: differ from those for $B=0.5$~T in two ways: (1) the biquadratic

828: contribution is 7~\% smaller, (2) the anisotropy coefficient is

829: three times larger.

830: 

831: \subsection{High-field properties of \nimo}

832: 

833: Although the above parameterization describes the low-field data

834: up to $B\approx 5$~T reasonably well, it fails to reproduce the

835: magnetization data at higher fields. The dramatic increase of

836: the field spacings between adjacent magnetization steps at 4.5,

837: 8.9, 20.1, and 32~T cannot be explained by small changes of the

838: anisotropy $D$ or the coupling $J$. The third field spacing is

839: about twice the first one, which using relation \fmref{E-1-2}

840: suggests an exchange coupling that is about 1.25 of the original

841: one. In the following we therefore investigate a model where

842: only the exchange parameters are allowed to depend on field.

843: This is of course a simplification since all parameters of the

844: Hamiltonian ($J$, $j$, $D$, and $g$) should be modified in a

845: varying field. For our purpose we assume a hypothetical

846: exponential dependence of $J$ on the absolute value of the

847: external magnetic field

848: %--------------------------------------------------------

849: \begin{eqnarray}

850: \label{E-3-3-1}

851: J(B)

852: &=&

853: J_0

854: \exp\left(

855: \frac{|B|}{\gamma}

856: \right)

857: \ ,

858: \end{eqnarray}

859: %--------------------------------------------------------

860: which we motivate by a possible change of the overlap of those

861: orbitals that are involved in the superexchange. Such a change

862: could be caused by variations of the bond distances and angles;

863: the latter are known to have dramatic effects on the exchange

864: parameters. Based on the magneto-structural correlations for

865: Ni-O cubanes in Ref.~\onlinecite{HSH:IC95}, a change of the

866: exchange parameter in \nimo\ by the necessary amount would

867: correspond to a change of the bond angles by approximately one

868: quarter of a degree if the effect was attributed solely to bond

869: angle modifications. A literature survey of chemically-similar

870: Ni(II) cluster

871: compounds\cite{HSH:IC95,WHM:PRB98,EFK:P99,CMP:CC01,ABB:CEJ02,KWM:PRB03,YWH:P03,MHR:EJIC04}

872: reveals that $J$ and $D$ vary from at least -10 to

873: 17.5~cm$^{-1}$ (-7 to 12.2~K) and -5.5 to 0.5 cm$^{-1}$ (-3.8 to

874: 8.5~K), respectively. In these systems, bond angle is the

875: governing structural parameter due to the intimate relationship

876: between angle and magnetic orbital overlap.\footnote{In

877:   contrast, the Ni..Ni distances have only a small effect on the

878:   observed values of $J$.} There is a linear correlation between

879: the value of $J$ and Ni-O-Ni and Ni-O-O-Ni bond angles in

880: several nickel clusters. Although the transition metal centers

881: in these nickel clusters have different local environments, the

882: aforementioned range of $J$ and $D$ represents the variation in

883: physical parameters that can be accessed by chemical tuning in

884: these compounds. Magnetic field-induced effects are anticipated

885: to be in this range as well.

886: 

887: 

888: %===================    figure   =================================

889: \begin{figure}[!ht]

890: \begin{center}

891: %\epsfig{file=ni4-B-M-combined-best.eps,width=65mm} \vspace*{1mm}

892: \epsfig{file=fig-10.eps,width=65mm} \vspace*{1mm}

893:   \caption[]{(color online) Magnetization of sample 2:

894:     Experimental data (NHMFL) are given by squares. The solid

895:     curve provides the best fit using a Hamiltonian with an

896:     exponentially field-dependent coupling. The simulated data

897:     are averaged over 100 orientations. The inset shows the

898:     experimental differential magnetization $\dint {\mathcal

899:       M}/\dint B$ as data points as well as the theoretical

900:     $\dint {\mathcal M}/\dint B$ (solid curve) using the same

901:     parameters as above. }

902: \label{F-10}

903: \end{center}

904: \end{figure}

905: %===================    figure ===============================

906: 

907: The best fit using two exchange parameters was obtained for

908: $J/k_B=-4.2~\text{K}*\exp(B/96~\text{T})$,

909: $J'/k_B=-3.2~\text{K}*\exp(B/52~\text{T})$, $j/k_B=0.16$~K,

910: $j'/k_B=0.39$~K, $D/k_B=-8.9$~K, and $g=2.195$ together with the

911: paramagnetic impurity contribution already determined for sample

912: 2 (parameterization \textbf{3}). This fit is shown in

913: \figref{F-10}. Except for small fields, where the two

914: aforementioned parameter sets are appropriate, this model

915: provides a rather good description of the magnetization over a

916: large field range. The assumption of an exponential dependence

917: is not essential; a polynomial dependence yields similar

918: results.

919: 

920: 

921: 

922: \subsection{EPR revisited}

923: 

924: The level scheme and possible EPR transitions can be shown more

925: closely using the parameters of the general Hamiltonian

926: \fmref{E-3-1-1}.

927: 

928: %===================    figure   =================================

929: \begin{figure}[!ht]

930: \begin{center}

931: \epsfig{file=fig-11.ps,width=55mm,angle=0}

932: \vspace*{1mm}

933: \caption[]{(color online) The figure shows the Zeeman level

934:   splittings for the realistic Heisenberg model \fmref{E-3-1-1}

935:   (parametrization \textbf{2}) for several orientations of the

936:   field relative to the molecules. The various curves, which

937:   differ especially at avoided level crossings, represent

938:   relative angles of $10^\circ$.}

939: \label{F-11}

940: \end{center}

941: \end{figure}

942: %===================    figure ===============================

943: 

944: In \figref{F-11} the Zeeman split levels are presented for

945: fields up to 8~T using parametrization \textbf{2}. The thick

946: curves show levels for an orientation of the field axis pointing

947: from the mid point of the basal triangle through the top Ni

948: center (Ni ion 1 in \figref{F-2}). The thinner curves show the

949: levels for orientations with relative angle steps of $10^\circ$

950: along a great circle through one of the basal Ni centers. The

951: variation of the level positions with orientation for a fixed

952: absolute value of the external field is especially large at

953: avoided level crossings.  Although the final EPR line is given

954: by averaging over all orientations one can already deduce from

955: \figref{F-11} that the EPR measurement is in full agreement with

956: the microscopic Hamiltonian. The attached symbols for the three

957: observed transitions match the calculated level spacings very

958: well.

959: 

960: 

961: %===================    figure   =================================

962: \begin{figure}[!ht]

963: \begin{center}

964: \epsfig{file=fig-12.eps,width=55mm,angle=0}

965: \vspace*{1mm}

966: \caption[]{Relative EPR transmission observed at $\nu=190$~GHz,

967:   which corresponds to an energy level difference of 9.12~K.:

968:   The dominant (allowed) transition P3 is rather broad, whereas

969:   the the weaker (forbidden) transition P1 results in a much

970:   sharper peak. The other weak (forbidden) transition P2, which

971:   is masked by P3, is broader than P1.}

972: \label{F-12}

973: \end{center}

974: \end{figure}

975: %===================    figure ===============================

976: 

977: To some extent one can even explain the widths of the three

978: transitions, which are shown in \figref{F-12} for a frequency of

979: $\nu=190$~GHz and a temperature of $T=4.2$~K.  The transition

980: P1, which occurs between energy levels marked by filled circles

981: in \figref{F-11}, is rather narrow; this is connected to the

982: fact that variations of the field direction do not lead to

983: strong variations of the respective levels.  This also implies

984: that the mixing of $\op{S}_z$ eigenstates is weak and thus the

985: amplitude of this forbidden transition is small. P3 on the

986: contrary is an allowed transition which dominates the spectrum.

987: In the investigated field range it is given mainly by the

988: transition between levels marked by open squares in

989: \figref{F-11}, but transitions between other (unmarked) levels

990: also contribute. he rather broad line can in part be explained

991: by the rather strong variation of the upper marked level with

992: variations of the field direction. The transition P2 between

993: levels marked by open triangles in \figref{F-11} is stronger

994: than P1 but also broader, which can again be understood by

995: looking at the variation of levels.

996: 

997: 

998: 

999: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1000: \section{Conclusions}

1001: \label{sec-5}

1002: 

1003: We have presented a comprehensive experimental and theoretical

1004: investigation of the magnetic molecule \nimo. We find that the

1005: main model parameters of the Hamiltonian (i.e. exchange and

1006: anisotropy parameters) have a strong dependence on magnetic

1007: field, an effect that may be accompanied by molecular

1008: magnetostriction. All of our efforts to avoid this conclusion

1009: lead to a blatant contradiction between theory and experiment.

1010: The discovery of a dependence of Hamiltonian parameters on field

1011: in \nimo\ is quite novel. Nevertheless, it may be a general

1012: characteristic of magnetic cluster-based materials with strong

1013: lattice coupling. It is well known that most materials

1014: (including magnetic molecules) are not

1015: rigid.\cite{SMW:PRB02,SDG:CEJ02,SDM:CEJ04,WCM:PRB,rai05} In

1016: addition, only relatively few high-field magnetization studies

1017: have been performed to date. Thus, it is of interest to discover

1018: whether other magnetic molecules might display a similarly

1019: strong dependence of the model parameters on magnetic field in

1020: the regime above 20~T.

1021: 

1022: 

1023: 

1024: 

1025: 

1026: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1027: \begin{acknowledgments}

1028:   This work was supported by the Deutsche Forschungsgemeinschaft

1029:   (Grant No.  SCHN~615/8-1).  Ames Laboratory is operated for

1030:   the U.S. Department of Energy by Iowa State University under

1031:   Contract No.  W-7405-Eng-82. H.~N.  acknowledges the support

1032:   by Grant in Aid for Scientific Research on Priority Areas (No.

1033:   13130204) from MEXT, Japan and by Shimazu Science Foundation.

1034:   Work at the University of Tennessee was supported by the

1035:   Materials Science Division, Basic Energy Sciences, U.S.

1036:   Department of Energy (DE-FG02-01ER45885) and the Petroleum

1037:   Research Fund, administered by the American Chemical Society

1038:   (38164-AC5).  The National High Magnetic Field Laboratory is

1039:   supported by NSF Cooperative Agreement DMR-0084173 and by the

1040:   State of Florida.  We thank K.~B\"arwinkel, S.~Hill,

1041:   V.~Kataev, G.~Khaliullin, M.~Pederson, A.~Postnikov,

1042:   R.~Saalfrank, J.~van Slageren, O.~Waldmann, and R.~Winpenny

1043:   for helpful discussions.

1044: \end{acknowledgments}

1045: 

1046: 

1047: %\bibliography{/home/schnack/tex/bibtex/js-own,/home/schnack/tex/bibtex/js-mag,/home/schnack/tex/bibtex/js-mis}

1048: 

1049: \begin{thebibliography}{45}

1050: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi

1051: \expandafter\ifx\csname bibnamefont\endcsname\relax

1052:   \def\bibnamefont#1{#1}\fi

1053: \expandafter\ifx\csname bibfnamefont\endcsname\relax

1054:   \def\bibfnamefont#1{#1}\fi

1055: \expandafter\ifx\csname citenamefont\endcsname\relax

1056:   \def\citenamefont#1{#1}\fi

1057: \expandafter\ifx\csname url\endcsname\relax

1058:   \def\url#1{\texttt{#1}}\fi

1059: \expandafter\ifx\csname urlprefix\endcsname\relax\def\urlprefix{URL }\fi

1060: \providecommand{\bibinfo}[2]{#2}

1061: \providecommand{\eprint}[2][]{\url{#2}}

1062: 

1063: \bibitem[{\citenamefont{Sessoli et~al.}(1993)\citenamefont{Sessoli, Gatteschi,

1064:   Caneschi, and Novak}}]{SGC:Nat93}

1065: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Sessoli}},

1066:   \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Gatteschi}},

1067:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Caneschi}}, \bibnamefont{and}

1068:   \bibinfo{author}{\bibfnamefont{M.~A.} \bibnamefont{Novak}},

1069:   \bibinfo{journal}{Nature} \textbf{\bibinfo{volume}{365}},

1070:   \bibinfo{pages}{141} (\bibinfo{year}{1993}).

1071: 

1072: \bibitem[{\citenamefont{Gatteschi et~al.}(1994)\citenamefont{Gatteschi,

1073:   Caneschi, Pardi, and Sessoli}}]{GCR:S94}

1074: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Gatteschi}},

1075:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Caneschi}},

1076:   \bibinfo{author}{\bibfnamefont{L.}~\bibnamefont{Pardi}}, \bibnamefont{and}

1077:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Sessoli}},

1078:   \bibinfo{journal}{Science} \textbf{\bibinfo{volume}{265}},

1079:   \bibinfo{pages}{1054} (\bibinfo{year}{1994}).

1080: 

1081: \bibitem[{\citenamefont{Gatteschi}(1994)}]{Gat:AM94}

1082: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Gatteschi}},

1083:   \bibinfo{journal}{Adv. Mater.} \textbf{\bibinfo{volume}{6}},

1084:   \bibinfo{pages}{635} (\bibinfo{year}{1994}).

1085: 

1086: \bibitem[{\citenamefont{Coronado et~al.}(1996)\citenamefont{Coronado, Delhaes,

1087:   Gatteschi, and Miller}}]{Cor:NATO96}

1088: \bibinfo{editor}{\bibfnamefont{E.}~\bibnamefont{Coronado}},

1089:   \bibinfo{editor}{\bibfnamefont{P.}~\bibnamefont{Delhaes}},

1090:   \bibinfo{editor}{\bibfnamefont{D.}~\bibnamefont{Gatteschi}},

1091:   \bibnamefont{and} \bibinfo{editor}{\bibfnamefont{J.}~\bibnamefont{Miller}},

1092:   eds., \emph{\bibinfo{title}{Localized and Itinerant Molecular Magnetism: From

1093:   Molecular Assemblies to the Devices}}, vol. \bibinfo{volume}{321} of

1094:   \emph{\bibinfo{series}{NATO Advanced Studies Institute, Series E: Applied

1095:   Sciences}} (\bibinfo{publisher}{Kluwer Academic},

1096:   \bibinfo{address}{Dordrecht}, \bibinfo{year}{1996}).

1097: 

1098: \bibitem[{\citenamefont{Caneschi et~al.}(1999)\citenamefont{Caneschi,

1099:   Gatteschi, Sangregorio, Sessoli, Sorace, Cornia, Novak, Paulsen, and

1100:   Wernsdorfer}}]{CGS:JMMM99}

1101: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Caneschi}},

1102:   \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Gatteschi}},

1103:   \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Sangregorio}},

1104:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Sessoli}},

1105:   \bibinfo{author}{\bibfnamefont{L.}~\bibnamefont{Sorace}},

1106:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Cornia}},

1107:   \bibinfo{author}{\bibfnamefont{M.~A.} \bibnamefont{Novak}},

1108:   \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Paulsen}}, \bibnamefont{and}

1109:   \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Wernsdorfer}},

1110:   \bibinfo{journal}{J. Magn. Magn. Mater.} \textbf{\bibinfo{volume}{200}},

1111:   \bibinfo{pages}{182} (\bibinfo{year}{1999}).

1112: 

1113: \bibitem[{\citenamefont{Waldmann et~al.}(1999)\citenamefont{Waldmann,

1114:   Sch\"ulein, Koch, M\"uller, Bernt, Saalfrank, Andres, G\"udel, and

1115:   Allenspach}}]{WSK:IO99}

1116: \bibinfo{author}{\bibfnamefont{O.}~\bibnamefont{Waldmann}},

1117:   \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Sch\"ulein}},

1118:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Koch}},

1119:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{M\"uller}},

1120:   \bibinfo{author}{\bibfnamefont{I.}~\bibnamefont{Bernt}},

1121:   \bibinfo{author}{\bibfnamefont{R.~W.} \bibnamefont{Saalfrank}},

1122:   \bibinfo{author}{\bibfnamefont{H.~P.} \bibnamefont{Andres}},

1123:   \bibinfo{author}{\bibfnamefont{H.~U.} \bibnamefont{G\"udel}},

1124:   \bibnamefont{and}

1125:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Allenspach}},

1126:   \bibinfo{journal}{Inorg. Chem.} \textbf{\bibinfo{volume}{38}},

1127:   \bibinfo{pages}{5879} (\bibinfo{year}{1999}).

1128: 

1129: \bibitem[{\citenamefont{M\"uller

1130:   et~al.}(2001{\natexlab{a}})\citenamefont{M\"uller, Luban, Schr\"oder, Modler,

1131:   K\"ogerler, Axenovich, Schnack, Canfield, Bud'ko, and Harrison}}]{MLS:CPC01}

1132: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{M\"uller}},

1133:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Luban}},

1134:   \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Schr\"oder}},

1135:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Modler}},

1136:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{K\"ogerler}},

1137:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Axenovich}},

1138:   \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Schnack}},

1139:   \bibinfo{author}{\bibfnamefont{P.~C.} \bibnamefont{Canfield}},

1140:   \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Bud'ko}}, \bibnamefont{and}

1141:   \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Harrison}},

1142:   \bibinfo{journal}{Chem. Phys. Chem.} \textbf{\bibinfo{volume}{2}},

1143:   \bibinfo{pages}{517} (\bibinfo{year}{2001}{\natexlab{a}}).

1144: 

1145: \bibitem[{\citenamefont{Wernsdorfer et~al.}(2002)\citenamefont{Wernsdorfer,

1146:   Aliaga-Alcalde, Hendrickson, and Christou}}]{WAH:Nature02}

1147: \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Wernsdorfer}},

1148:   \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Aliaga-Alcalde}},

1149:   \bibinfo{author}{\bibfnamefont{D.~N.} \bibnamefont{Hendrickson}},

1150:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Christou}},

1151:   \bibinfo{journal}{Nature} \textbf{\bibinfo{volume}{416}},

1152:   \bibinfo{pages}{406} (\bibinfo{year}{2002}).

1153: 

1154: \bibitem[{\citenamefont{M\"uller

1155:   et~al.}(2001{\natexlab{b}})\citenamefont{M\"uller, K\"ogerler, and

1156:   Dress}}]{MKD:CCR01}

1157: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{M\"uller}},

1158:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{K\"ogerler}},

1159:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{A.~W.~M.}

1160:   \bibnamefont{Dress}}, \bibinfo{journal}{Coord. Chem. Rev.}

1161:   \textbf{\bibinfo{volume}{222}}, \bibinfo{pages}{193}

1162:   (\bibinfo{year}{2001}{\natexlab{b}}).

1163: 

1164: \bibitem[{\citenamefont{Lis}(1980)}]{Lis:ACB80}

1165: \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Lis}}, \bibinfo{journal}{Acta

1166:   Chrytallogr. B} \textbf{\bibinfo{volume}{36}}, \bibinfo{pages}{2042}

1167:   (\bibinfo{year}{1980}).

1168: 

1169: \bibitem[{\citenamefont{Thomas et~al.}(1996)\citenamefont{Thomas, Lionti,

1170:   Ballou, Gatteschi, Sessoli, and Barbara}}]{TLB:Nature96}

1171: \bibinfo{author}{\bibfnamefont{L.}~\bibnamefont{Thomas}},

1172:   \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Lionti}},

1173:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Ballou}},

1174:   \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Gatteschi}},

1175:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Sessoli}}, \bibnamefont{and}

1176:   \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Barbara}},

1177:   \bibinfo{journal}{Nature} \textbf{\bibinfo{volume}{383}},

1178:   \bibinfo{pages}{145} (\bibinfo{year}{1996}).

1179: 

1180: \bibitem[{\citenamefont{Jang et~al.}(2000)\citenamefont{Jang, Lascialfari,

1181:   Borsa, and Gatteschi}}]{JLB:PRL00}

1182: \bibinfo{author}{\bibfnamefont{Z.~H.} \bibnamefont{Jang}},

1183:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Lascialfari}},

1184:   \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Borsa}}, \bibnamefont{and}

1185:   \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Gatteschi}},

1186:   \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{84}},

1187:   \bibinfo{pages}{2977} (\bibinfo{year}{2000}).

1188: 

1189: \bibitem[{\citenamefont{Furukawa et~al.}(2001)\citenamefont{Furukawa, Watanabe,

1190:   Kumagai, Borsa, and Gatteschi}}]{FWK:PRB01}

1191: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Furukawa}},

1192:   \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Watanabe}},

1193:   \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Kumagai}},

1194:   \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Borsa}}, \bibnamefont{and}

1195:   \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Gatteschi}},

1196:   \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{64}},

1197:   \bibinfo{pages}{104401} (\bibinfo{year}{2001}).

1198: 

1199: \bibitem[{\citenamefont{Regnault et~al.}(2002)\citenamefont{Regnault,

1200:   Jolic{\oe}ur, Sessoli, Gatteschi, and Verdaguer}}]{RJS:PRB02}

1201: \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Regnault}},

1202:   \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Jolic{\oe}ur}},

1203:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Sessoli}},

1204:   \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Gatteschi}},

1205:   \bibnamefont{and}

1206:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Verdaguer}},

1207:   \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{66}},

1208:   \bibinfo{pages}{054409} (\bibinfo{year}{2002}).

1209: 

1210: \bibitem[{\citenamefont{Chaboussant et~al.}(2004)\citenamefont{Chaboussant,

1211:   Sieber, Ochsenbein, G\"udel, Murrie, Honecker, Fukushima, and

1212:   Normand}}]{CSO:PRB04}

1213: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Chaboussant}},

1214:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Sieber}},

1215:   \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Ochsenbein}},

1216:   \bibinfo{author}{\bibfnamefont{H.-U.} \bibnamefont{G\"udel}},

1217:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Murrie}},

1218:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Honecker}},

1219:   \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Fukushima}},

1220:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Normand}},

1221:   \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{70}},

1222:   \bibinfo{pages}{104422} (\bibinfo{year}{2004}).

1223: 

1224: \bibitem[{\citenamefont{Koch et~al.}(2003)\citenamefont{Koch, Waldmann,

1225:   M\"uller, Reimann, and Saalfrank}}]{KWM:PRB03}

1226: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Koch}},

1227:   \bibinfo{author}{\bibfnamefont{O.}~\bibnamefont{Waldmann}},

1228:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{M\"uller}},

1229:   \bibinfo{author}{\bibfnamefont{U.}~\bibnamefont{Reimann}}, \bibnamefont{and}

1230:   \bibinfo{author}{\bibfnamefont{R.~W.} \bibnamefont{Saalfrank}},

1231:   \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{67}},

1232:   \bibinfo{pages}{094407} (\bibinfo{year}{2003}).

1233: 

1234: \bibitem[{\citenamefont{Luban et~al.}(2003)\citenamefont{Luban, Borsa, Bud'ko,

1235:   Canfield, Jun, Jung, K\"ogerler, Mentrup, M\"uller, Modler

1236:   et~al.}}]{LBB:PRB03}

1237: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Luban}},

1238:   \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Borsa}},

1239:   \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Bud'ko}},

1240:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Canfield}},

1241:   \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Jun}},

1242:   \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Jung}},

1243:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{K\"ogerler}},

1244:   \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Mentrup}},

1245:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{M\"uller}},

1246:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Modler}},

1247:   \bibnamefont{et~al.}, \bibinfo{journal}{Phys. Rev. B}

1248:   \textbf{\bibinfo{volume}{66}}, \bibinfo{pages}{054407}

1249:   (\bibinfo{year}{2003}).

1250: 

1251: \bibitem[{\citenamefont{Waldmann et~al.}(1998)\citenamefont{Waldmann, Hassmann,

1252:   M{\"u}ller, Volkmer, Schubert, and Lehn}}]{WHM:PRB98}

1253: \bibinfo{author}{\bibfnamefont{O.}~\bibnamefont{Waldmann}},

1254:   \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Hassmann}},

1255:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{M{\"u}ller}},

1256:   \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Volkmer}},

1257:   \bibinfo{author}{\bibfnamefont{U.~S.} \bibnamefont{Schubert}},

1258:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{J.~M.} \bibnamefont{Lehn}},

1259:   \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{58}},

1260:   \bibinfo{pages}{3277} (\bibinfo{year}{1998}).

1261: 

1262: \bibitem[{\citenamefont{M\"uller et~al.}(2000)\citenamefont{M\"uller, Beugholt,

1263:   K\"ogerler, Bogge, Bud'ko, and Luban}}]{MBK:IC00}

1264: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{M\"uller}},

1265:   \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Beugholt}},

1266:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{K\"ogerler}},

1267:   \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Bogge}},

1268:   \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Bud'ko}}, \bibnamefont{and}

1269:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Luban}},

1270:   \bibinfo{journal}{Inorg. Chem.} \textbf{\bibinfo{volume}{39}},

1271:   \bibinfo{pages}{5176} (\bibinfo{year}{2000}).

1272: 

1273: \bibitem[{\citenamefont{Andres et~al.}(2002)\citenamefont{Andres, Basler,

1274:   Blake, Cadiou, Chaboussant, Grant, Gudel, Murrie, Parsons, Paulsen

1275:   et~al.}}]{ABB:CEJ02}

1276: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Andres}},

1277:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Basler}},

1278:   \bibinfo{author}{\bibfnamefont{A.~J.} \bibnamefont{Blake}},

1279:   \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Cadiou}},

1280:   \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Chaboussant}},

1281:   \bibinfo{author}{\bibfnamefont{C.~M.} \bibnamefont{Grant}},

1282:   \bibinfo{author}{\bibfnamefont{H.~U.} \bibnamefont{Gudel}},

1283:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Murrie}},

1284:   \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Parsons}},

1285:   \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Paulsen}},

1286:   \bibnamefont{et~al.}, \bibinfo{journal}{Chem. Eur. J.}

1287:   \textbf{\bibinfo{volume}{8}}, \bibinfo{pages}{4867} (\bibinfo{year}{2002}).

1288: 

1289: \bibitem[{\citenamefont{Yang et~al.}(2003)\citenamefont{Yang, Wernsdorfer,

1290:   Hill, Edwards, Nakano, Maccagnano, Zakharov, Rheingold, Christou, and

1291:   Hendrickson}}]{YWH:P03}

1292: \bibinfo{author}{\bibfnamefont{E.~C.} \bibnamefont{Yang}},

1293:   \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Wernsdorfer}},

1294:   \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Hill}},

1295:   \bibinfo{author}{\bibfnamefont{R.~S.} \bibnamefont{Edwards}},

1296:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Nakano}},

1297:   \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Maccagnano}},

1298:   \bibinfo{author}{\bibfnamefont{L.}~\bibnamefont{Zakharov}},

1299:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Rheingold}},

1300:   \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Christou}}, \bibnamefont{and}

1301:   \bibinfo{author}{\bibfnamefont{D.~N.} \bibnamefont{Hendrickson}},

1302:   \bibinfo{journal}{Polyhedron} \textbf{\bibinfo{volume}{22}},

1303:   \bibinfo{pages}{1727} (\bibinfo{year}{2003}).

1304: 

1305: \bibitem[{\citenamefont{Moragues-Canovas

1306:   et~al.}(2004)\citenamefont{Moragues-Canovas, Helliwell, Ricard, Riviere,

1307:   Wernsdorfer, Brechin, and Mallah}}]{MHR:EJIC04}

1308: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Moragues-Canovas}},

1309:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Helliwell}},

1310:   \bibinfo{author}{\bibfnamefont{L.}~\bibnamefont{Ricard}},

1311:   \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Riviere}},

1312:   \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Wernsdorfer}},

1313:   \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Brechin}}, \bibnamefont{and}

1314:   \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Mallah}},

1315:   \bibinfo{journal}{Eur. J. Inorg. Chem.} pp. \bibinfo{pages}{2219--2222}

1316:   (\bibinfo{year}{2004}).

1317: 

1318: \bibitem[{\citenamefont{King et~al.}(2004)\citenamefont{King, Clerac,

1319:   Wernsdorfer, Anson, and Powell}}]{KCW:DT04}

1320: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{King}},

1321:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Clerac}},

1322:   \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Wernsdorfer}},

1323:   \bibinfo{author}{\bibfnamefont{C.~E.} \bibnamefont{Anson}}, \bibnamefont{and}

1324:   \bibinfo{author}{\bibfnamefont{A.~K.} \bibnamefont{Powell}},

1325:   \bibinfo{journal}{Dalton Trans.} pp. \bibinfo{pages}{2670--2676}

1326:   (\bibinfo{year}{2004}).

1327: 

1328: \bibitem[{\citenamefont{Demeshko et~al.}(2005)\citenamefont{Demeshko,

1329:   Leibeling, Maringgele, Meyer, Mennerich, Klauss, and Pritzkow}}]{DLM:IC05}

1330: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Demeshko}},

1331:   \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Leibeling}},

1332:   \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Maringgele}},

1333:   \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Meyer}},

1334:   \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Mennerich}},

1335:   \bibinfo{author}{\bibfnamefont{H.~H.} \bibnamefont{Klauss}},

1336:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Pritzkow}},

1337:   \bibinfo{journal}{Inorg. Chem.} \textbf{\bibinfo{volume}{44}},

1338:   \bibinfo{pages}{519} (\bibinfo{year}{2005}).

1339: 

1340: \bibitem[{\citenamefont{Abbati et~al.}(2000)\citenamefont{Abbati, Caneschi,

1341:   Cornia, Fabretti, and Gatteschi}}]{ACC:ICA00}

1342: \bibinfo{author}{\bibfnamefont{G.~L.} \bibnamefont{Abbati}},

1343:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Caneschi}},

1344:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Cornia}},

1345:   \bibinfo{author}{\bibfnamefont{A.~C.} \bibnamefont{Fabretti}},

1346:   \bibnamefont{and}

1347:   \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Gatteschi}},

1348:   \bibinfo{journal}{Inorg. Chim. Acta} \textbf{\bibinfo{volume}{297}},

1349:   \bibinfo{pages}{291} (\bibinfo{year}{2000}).

1350: 

1351: \bibitem[{\citenamefont{Schnack and Luban}(2001)}]{ScL:PRB01}

1352: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Schnack}} \bibnamefont{and}

1353:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Luban}},

1354:   \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{63}},

1355:   \bibinfo{pages}{014418} (\bibinfo{year}{2001}).

1356: 

1357: \bibitem[{\citenamefont{Br\"uger}(2003)}]{Brueger03}

1358: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Br\"uger}},

1359:   \bibinfo{type}{diploma thesis}, \bibinfo{school}{Universit\"at Osnabr\"uck},

1360:   \bibinfo{address}{Universit\"at Osnabr\"uck} (\bibinfo{year}{2003}).

1361: 

1362: \bibitem[{\citenamefont{Halcrow et~al.}(1995)\citenamefont{Halcrow, Sun,

1363:   Huffman, and Christou}}]{HSH:IC95}

1364: \bibinfo{author}{\bibfnamefont{M.~A.} \bibnamefont{Halcrow}},

1365:   \bibinfo{author}{\bibfnamefont{J.-S.} \bibnamefont{Sun}},

1366:   \bibinfo{author}{\bibfnamefont{J.~C.} \bibnamefont{Huffman}},

1367:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Christou}},

1368:   \bibinfo{journal}{Inorg. Chem.} \textbf{\bibinfo{volume}{34}},

1369:   \bibinfo{pages}{4167} (\bibinfo{year}{1995}).

1370: 

1371: \bibitem[{\citenamefont{Escuer et~al.}(1999)\citenamefont{Escuer,

1372:   Font-Bard\`{i}a, Kumar, Solans, and Vicente}}]{EFK:P99}

1373: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Escuer}},

1374:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Font-Bard\`{i}a}},

1375:   \bibinfo{author}{\bibfnamefont{S.~B.} \bibnamefont{Kumar}},

1376:   \bibinfo{author}{\bibfnamefont{X.}~\bibnamefont{Solans}}, \bibnamefont{and}

1377:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Vicente}},

1378:   \bibinfo{journal}{Polyhedron} \textbf{\bibinfo{volume}{18}},

1379:   \bibinfo{pages}{909} (\bibinfo{year}{1999}).

1380: 

1381: \bibitem[{\citenamefont{Cadiou et~al.}(2001)\citenamefont{Cadiou, Murrie,

1382:   Paulsen, Villar, Wernsdorfer, and Winpenny}}]{CMP:CC01}

1383: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Cadiou}},

1384:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Murrie}},

1385:   \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Paulsen}},

1386:   \bibinfo{author}{\bibfnamefont{V.}~\bibnamefont{Villar}},

1387:   \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Wernsdorfer}},

1388:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Winpenny}},

1389:   \bibinfo{journal}{Chem. Commun.} pp. \bibinfo{pages}{2666--2667}

1390:   (\bibinfo{year}{2001}).

1391: 

1392: \bibitem[{\citenamefont{Choi et~al.}(2004)\citenamefont{Choi, Woodward,

1393:   Musfeldt, Wei, Whangbo, He, Jin, and Mandrus}}]{CWM:PRB04}

1394: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Choi}},

1395:   \bibinfo{author}{\bibfnamefont{J.~D.} \bibnamefont{Woodward}},

1396:   \bibinfo{author}{\bibfnamefont{J.~L.} \bibnamefont{Musfeldt}},

1397:   \bibinfo{author}{\bibfnamefont{X.}~\bibnamefont{Wei}},

1398:   \bibinfo{author}{\bibfnamefont{M.~H.} \bibnamefont{Whangbo}},

1399:   \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{He}},

1400:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Jin}}, \bibnamefont{and}

1401:   \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Mandrus}},

1402:   \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{70}},

1403:   \bibinfo{pages}{085107} (\bibinfo{year}{2004}).

1404: 

1405: \bibitem[{\citenamefont{Park et~al.}(2005)\citenamefont{Park, Yang, and

1406:   Hendrickson}}]{PYH:JAP05}

1407: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Park}},

1408:   \bibinfo{author}{\bibfnamefont{E.-C.} \bibnamefont{Yang}}, \bibnamefont{and}

1409:   \bibinfo{author}{\bibfnamefont{D.~N.} \bibnamefont{Hendrickson}},

1410:   \bibinfo{journal}{J. Appl. Phys.} \textbf{\bibinfo{volume}{97}},

1411:   \bibinfo{pages}{10M522} (\bibinfo{year}{2005}).

1412: 

1413: \bibitem[{\citenamefont{Postnikov et~al.}(2005)\citenamefont{Postnikov,

1414:   Br{\"u}ger, and Schnack}}]{PBS:PT05}

1415: \bibinfo{author}{\bibfnamefont{A.~V.} \bibnamefont{Postnikov}},

1416:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Br{\"u}ger}},

1417:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Schnack}},

1418:   \bibinfo{journal}{Phase Transitions} \textbf{\bibinfo{volume}{78}},

1419:   \bibinfo{pages}{47} (\bibinfo{year}{2005}).

1420: 

1421: \bibitem[{\citenamefont{Woodward et~al.}(2005)\citenamefont{Woodward, Choi,

1422:   Musfeldt, Haraldsen, Wei, Koo, Dai, Whangbo, Landee, and Turnbull}}]{WCM:PRB}

1423: \bibinfo{author}{\bibfnamefont{J.~D.} \bibnamefont{Woodward}},

1424:   \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Choi}},

1425:   \bibinfo{author}{\bibfnamefont{J.~L.} \bibnamefont{Musfeldt}},

1426:   \bibinfo{author}{\bibfnamefont{J.~T.} \bibnamefont{Haraldsen}},

1427:   \bibinfo{author}{\bibfnamefont{X.}~\bibnamefont{Wei}},

1428:   \bibinfo{author}{\bibfnamefont{H.-J.} \bibnamefont{Koo}},

1429:   \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Dai}},

1430:   \bibinfo{author}{\bibfnamefont{M.-H.} \bibnamefont{Whangbo}},

1431:   \bibinfo{author}{\bibfnamefont{C.~P.} \bibnamefont{Landee}},

1432:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{M.~M.}

1433:   \bibnamefont{Turnbull}}, \bibinfo{journal}{Phys. Rev. B}

1434:   \textbf{\bibinfo{volume}{71}}, \bibinfo{pages}{174416}

1435:   (\bibinfo{year}{2005}).

1436: 

1437: \bibitem[{\citenamefont{Rai et~al.}()\citenamefont{Rai, Cao, Musfeldt, Wei,

1438:   Singh, Sales, and Mandrus}}]{rai05}

1439: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Rai}},

1440:   \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Cao}},

1441:   \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Musfeldt}},

1442:   \bibinfo{author}{\bibfnamefont{X.}~\bibnamefont{Wei}},

1443:   \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Singh}},

1444:   \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Sales}}, \bibnamefont{and}

1445:   \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Mandrus}},

1446:   \bibinfo{note}{submitted}.

1447: 

1448: \bibitem[{\citenamefont{Bencini and Gatteschi}(1990)}]{BeG90}

1449: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Bencini}} \bibnamefont{and}

1450:   \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Gatteschi}},

1451:   \emph{\bibinfo{title}{Electron parametric resonance of exchange coupled

1452:   systems}} (\bibinfo{publisher}{Springer, Berlin, Heidelberg},

1453:   \bibinfo{year}{1990}).

1454: 

1455: \bibitem[{\citenamefont{Basler et~al.}(2003)\citenamefont{Basler, Boskovic,

1456:   Chaboussant, Gudel, Murrie, Ochsenbein, and Sieber}}]{BBC:CPC03}

1457: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Basler}},

1458:   \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Boskovic}},

1459:   \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Chaboussant}},

1460:   \bibinfo{author}{\bibfnamefont{H.~U.} \bibnamefont{Gudel}},

1461:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Murrie}},

1462:   \bibinfo{author}{\bibfnamefont{S.~T.} \bibnamefont{Ochsenbein}},

1463:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Sieber}},

1464:   \bibinfo{journal}{ChemPhysChem} \textbf{\bibinfo{volume}{4}},

1465:   \bibinfo{pages}{910} (\bibinfo{year}{2003}).

1466: 

1467: \bibitem[{\citenamefont{Waldmann

1468:   et~al.}(2002{\natexlab{a}})\citenamefont{Waldmann, Zhao, and

1469:   Thompson}}]{WZT:PRL02}

1470: \bibinfo{author}{\bibfnamefont{O.}~\bibnamefont{Waldmann}},

1471:   \bibinfo{author}{\bibfnamefont{L.}~\bibnamefont{Zhao}}, \bibnamefont{and}

1472:   \bibinfo{author}{\bibfnamefont{L.}~\bibnamefont{Thompson}},

1473:   \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{88}},

1474:   \bibinfo{pages}{066401} (\bibinfo{year}{2002}{\natexlab{a}}).

1475: 

1476: \bibitem[{\citenamefont{Huang and Orbach}(1964)}]{HuO:PRL64}

1477: \bibinfo{author}{\bibfnamefont{N.~L.} \bibnamefont{Huang}} \bibnamefont{and}

1478:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Orbach}},

1479:   \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{12}},

1480:   \bibinfo{pages}{275} (\bibinfo{year}{1964}).

1481: 

1482: \bibitem[{\citenamefont{Sushkov et~al.}(2002)\citenamefont{Sushkov, Musfeldt,

1483:   Wang, Achey, and Dalal}}]{SMW:PRB02}

1484: \bibinfo{author}{\bibfnamefont{A.~B.} \bibnamefont{Sushkov}},

1485:   \bibinfo{author}{\bibfnamefont{J.~L.} \bibnamefont{Musfeldt}},

1486:   \bibinfo{author}{\bibfnamefont{Y.~J.} \bibnamefont{Wang}},

1487:   \bibinfo{author}{\bibfnamefont{R.~M.} \bibnamefont{Achey}}, \bibnamefont{and}

1488:   \bibinfo{author}{\bibfnamefont{N.~S.} \bibnamefont{Dalal}},

1489:   \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{66}},

1490:   \bibinfo{pages}{144430} (\bibinfo{year}{2002}).

1491: 

1492: \bibitem[{\citenamefont{Saalfrank et~al.}(2002)\citenamefont{Saalfrank,

1493:   Demleitner, Glaser, Maid, Bathelt, Hampel, Bauer, and Teichert}}]{SDG:CEJ02}

1494: \bibinfo{author}{\bibfnamefont{R.~W.} \bibnamefont{Saalfrank}},

1495:   \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Demleitner}},

1496:   \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Glaser}},

1497:   \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Maid}},

1498:   \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Bathelt}},

1499:   \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Hampel}},

1500:   \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Bauer}}, \bibnamefont{and}

1501:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Teichert}},

1502:   \bibinfo{journal}{Chem.-Eur. J.} \textbf{\bibinfo{volume}{8}},

1503:   \bibinfo{pages}{2679} (\bibinfo{year}{2002}).

1504: 

1505: \bibitem[{\citenamefont{Saalfrank et~al.}(2004)\citenamefont{Saalfrank,

1506:   Deutscher, Maid, Ako, Sperner, Nakajima, Bauer, Hampel, Hess, Hommes

1507:   et~al.}}]{SDM:CEJ04}

1508: \bibinfo{author}{\bibfnamefont{R.~W.} \bibnamefont{Saalfrank}},

1509:   \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Deutscher}},

1510:   \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Maid}},

1511:   \bibinfo{author}{\bibfnamefont{A.~M.} \bibnamefont{Ako}},

1512:   \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Sperner}},

1513:   \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Nakajima}},

1514:   \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Bauer}},

1515:   \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Hampel}},

1516:   \bibinfo{author}{\bibfnamefont{B.~A.} \bibnamefont{Hess}},

1517:   \bibinfo{author}{\bibfnamefont{N.~J. R.~V.} \bibnamefont{Hommes}},

1518:   \bibnamefont{et~al.}, \bibinfo{journal}{Chem.-Eur. J.}

1519:   \textbf{\bibinfo{volume}{10}}, \bibinfo{pages}{1899} (\bibinfo{year}{2004}).

1520: 

1521: \bibitem[{\citenamefont{Chiorescu et~al.}(2000)\citenamefont{Chiorescu,

1522:   Wernsdorfer, M\"uller, B\"ogge, and Barbara}}]{CWM:PRL00}

1523: \bibinfo{author}{\bibfnamefont{I.}~\bibnamefont{Chiorescu}},

1524:   \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Wernsdorfer}},

1525:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{M\"uller}},

1526:   \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{B\"ogge}}, \bibnamefont{and}

1527:   \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Barbara}},

1528:   \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{84}},

1529:   \bibinfo{pages}{3454} (\bibinfo{year}{2000}).

1530: 

1531: \bibitem[{\citenamefont{Waldmann

1532:   et~al.}(2002{\natexlab{b}})\citenamefont{Waldmann, Koch, Schromm, M\"uller,

1533:   Bernt, and Saalfrank}}]{WKS:PRL02}

1534: \bibinfo{author}{\bibfnamefont{O.}~\bibnamefont{Waldmann}},

1535:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Koch}},

1536:   \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Schromm}},

1537:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{M\"uller}},

1538:   \bibinfo{author}{\bibfnamefont{I.}~\bibnamefont{Bernt}}, \bibnamefont{and}

1539:   \bibinfo{author}{\bibfnamefont{R.~W.} \bibnamefont{Saalfrank}},

1540:   \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{89}},

1541:   \bibinfo{pages}{246401} (\bibinfo{year}{2002}{\natexlab{b}}).

1542: 

1543: \bibitem[{\citenamefont{Waldmann et~al.}(2003)\citenamefont{Waldmann, Koch,

1544:   Schromm, M\"uller, Bernt, and Saalfrank}}]{WKS:PRL03}

1545: \bibinfo{author}{\bibfnamefont{O.}~\bibnamefont{Waldmann}},

1546:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Koch}},

1547:   \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Schromm}},

1548:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{M\"uller}},

1549:   \bibinfo{author}{\bibfnamefont{I.}~\bibnamefont{Bernt}}, \bibnamefont{and}

1550:   \bibinfo{author}{\bibfnamefont{R.~W.} \bibnamefont{Saalfrank}},

1551:   \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{90}},

1552:   \bibinfo{pages}{229904} (\bibinfo{year}{2003}).

1553: 

1554: \end{thebibliography}

1555: 

1556: 

1557: 

1558: 

1559: \end{document}

1560: 

1561: