1: \documentstyle[prl,aps,epsfig,floats]{revtex}
2: \newcommand\beq{\begin{equation}}
3: \newcommand\eeq{\end{equation}}
4: \newcommand\bea{\begin{eqnarray}}
5: \newcommand\eea{\end{eqnarray}}
6:
7: \begin{document}
8: \draft
9:
10: \textheight=23.8cm
11: \twocolumn[\hsize\textwidth\columnwidth\hsize\csname@twocolumnfalse\endcsname
12:
13: \title{An alternate model for magnetization plateaus in the molecular magnet
14: $\rm V_{15}$}
15:
16: \author{Indranil Rudra$^1$, S. Ramasesha$^1$ and Diptiman Sen$^2$}
17: \address{\it $^1$ Solid State and Structural Chemistry Unit,
18: Indian Institute of Science, Bangalore 560012, India \\
19: $^2$ Centre for Theoretical Studies,
20: Indian Institute of Science, Bangalore 560012, India}
21:
22: \date{\today}
23: \maketitle
24:
25: \begin{abstract}
26: Starting from an antiferromagnetic Heisenberg Hamiltonian for the fifteen
27: spin-$1/2$ ions in $\rm V_{15}$, we construct an effective spin Hamiltonian
28: involving eight low-lying states (spin-$1/2$ and spin-$3/2$) coupled to a
29: phonon bath. We numerically solve the time-dependent Schr$\ddot o$dinger
30: equation of this system, and obtain the magnetization as a function of
31: temperature in a time-dependent magnetic field. The magnetization exhibits
32: unusual patterns of hysteresis and plateaus as the field sweep rate and
33: temperature are varied. The observed plateaus are not due to quantum tunneling
34: but are a result of thermal averaging. Our results are in good agreement with
35: recent experimental observations.
36: \end{abstract}
37: \pacs{PACS numbers: ~75.45.+j, ~75.60.Ej, ~75.50.Xx}
38: \vskip.5pc
39: ]
40:
41: The synthesis of high-nuclearity transition metal clusters such as
42: ${\rm Mn_{12},~Fe_8~and~V_{15}}$ \cite{synth} has provided an impetus to the
43: study of magnetism on the nanoscale. These transition metal clusters
44: are basically isolated transition metal complexes involving multi-dentate
45: ligands; the chemical pathway between the metal ions in the transition metal
46: complex dictates the nature of exchange interactions. The complex interplay of
47: the topology of exchange interactions, magnetic dipolar interactions and
48: spin-lattice coupling has yielded a rich physics on the nanoscale which
49: includes quantum tunneling, quantum phase interference and quantum coherence
50: \cite{qtm,v15}. Quantum resonance tunneling is characterized by the
51: observation of discrete steps or plateaus in the magnetic hysteresis loops
52: at low temperatures. The signature of quantum interference is seen in the
53: variation of the tunnel splitting as a function of the azimuthal angle of
54: the transverse field for tunneling between $M_s~ =~ -10$ and $10-n$ states in
55: molecular magnets with ground state spin-$10$ \cite{wernsdorferfe8}.
56: Quantum coherence-decoherence studies are important from the stand point
57: of application of these systems in quantum computations \cite{qcompu}.
58:
59: There have been several models proposed to understand these phenomena
60: \cite{model}. Quantum hysteresis and interference have largely been studied
61: by using an effective spin Hamiltonian with dipolar interactions with a time
62: varying external magnetic field. The time evolution of the states of the
63: system are carried out within a master-equation approach \cite{mastereq}. The
64: decoherence phenomena has been studied by using a simple two state model in a
65: transverse magnetic field \cite{decoherence}. Even though most of these
66: clusters contain a fairly small number of metal ions, the spin on the metal
67: ion, at least in the case of $\rm Fe_8$ and $\rm Mn_{12}$, is fairly large; a
68: full quantum mechanical study of these systems is difficult because of the
69: large Fock space dimensionalities of $1.69$ million and $100$ million
70: respectively. However, the $\rm V_{15}$ cluster is far more amenable
71: to a rigorous quantum mechanical analysis because of the much smaller
72: Fock space ($\approx 33,000$ dimensional) spanned by the unpaired spins of
73: the system. A quantitative study of these systems requires at least the
74: low-lying states of the full spin-Hamiltonian to be evolved in time quantum
75: mechanically as the external magnetic field is ramped with time (as is done in
76: experiments). In this letter, we study the magnetization of $\rm V_{15}$ by
77: following its evolution as a function of a time-dependent magnetic field at
78: different temperatures. The low-lying states are obtained by solving the
79: exchange Hamiltonian corresponding to all the spins of the system.
80: A spin-phonon interaction is then introduced in the Hamiltonian. We thermally
81: average the magnetization over the low-lying states after each of these states
82: is independently evolved. We find that this model reproduces quantitatively
83: all the experimental features found in the magnetization studies of $\rm
84: V_{15}$ \cite{wernsdorferv15}.
85:
86: The schematic structure of $\rm V_{15}$ is shown in Fig. 1. Structural and
87: related studies on the cluster indicate that within each hexagon,
88: there are three alternating exchanges $J \approx 800~K$ which are
89: the strongest in the system, and they define the energy scale of the problem.
90: Besides, there are weaker exchange interactions between the spins involved
91: in the strong exchange and also with the triangle spins which lie between the
92: hexagons. All the exchange interactions are antiferromagnetic in nature.
93: The exchange pathways and their strengths \cite{wernsdorferv15} are also
94: shown in Fig. 1. What is significant in the cluster is the fact that the
95: spins in the triangle do not experience direct exchange interactions of any
96: significance. The exchange Hamiltonian of the cluster
97: is solved using a valence bond basis in each of the total spin subspaces,
98: for all the eigenvalues. It is found that two spin-$1/2$ states and a
99: spin-$3/2$ state are split-off from the rest of the spectrum by a gap of
100: $0.6 J$ \cite{synth}. These eight states almost exclusively correspond
101: to the triangle spins and they are the only states which will make significant
102: contributions to sub-kelvin properties. We therefore set up an effective
103: Hamiltonian in the Fock space of the three spins. We find that the form
104: \bea
105: H_{sp-sp}~ =~ && \epsilon ~I ~+ ~\alpha ~(S_1 \cdot S_2 + S_2 \cdot S_3 + S_3
106: \cdot S_1) \nonumber \\
107: && +~ \delta ~S_1 \cdot S_2 \times S_3 ~,
108: \eea
109: where $\epsilon = -4.781808$, $\alpha =-0.001491$, and $\delta = 0.015144$ in
110: units of the exchange $J$ (see Fig. 1), reproduces the low-lying eigenstates
111: to numerical accuracy. Note the unusually large value of the three-spin
112: interaction; this term has not been considered in the earlier literature
113: on this subject (such as Ref. [9]), and it is essential for the spin-$3/2$
114: state to lie between the two spin-$1/2$ states as is found to be the case in
115: this system.
116:
117: The direct spin-spin interaction terms permitted by the $C_3$ symmetry are
118: given by
119: \beq
120: H_{dip} ~=~ \gamma ~[~(S_{+}^3 + S_{-}^3) ~+~ i ~(S_{+}^3 - S_{-}^3)~] ~.
121: \eeq
122: We have also introduced a coupling between the spin states of the cluster and
123: the phonons. The spin-phonon interaction Hamiltonian which preserves the $C_3$
124: symmetry is phenomenologically given by \cite{spph}
125: \bea
126: H_{sp-ph} ~=~ q ~ (b + b^\dagger) ~[~ && (S_+^2 + S_-^2) ~+~ i
127: (S_+^2 - S_-^2) \nonumber \\
128: && +~ (S_z^2 - {\frac{1}{3}} S^2) ~] ~,
129: \label{hspph}
130: \eea
131: where {\it q} is the spin-phonon coupling constant, $b$ ($b^\dagger$) is the
132: phonon annihilation (creation) operator, and $\hbar \omega$ is the phonon
133: frequency. For simplicity, we have assumed a single phonon mode although the
134: molecule has various possible vibrational modes. The form of the interaction
135: in Eq. (\ref{hspph}) means that the phonons couple only to states with
136: spin-$3/2$. We have restricted the dimensionality of the Fock space of the
137: phonons to $15$ considering the low temperatures of interest.
138:
139: The evolution of the magnetization as a function of the magnetic field has
140: been studied by using the total Hamiltonian $H_{total}$, given by
141: \bea
142: H_{total} ~=~ && H_{sp-sp} ~+~ H_{dip} ~+~ H_{sp-ph} ~+~ \hbar \omega
143: (b^\dagger b + 1/2) \nonumber \\
144: && +~ h_z (t) S_z ~+~ h_x (t) S_x ~,
145: \eea
146: where we have assumed that besides an axial field $h_z (t)$, a small
147: transverse field $h_x (t)$
148: could also be present to account for any mismatch between the crystalline
149: $z$-axis and the molecular $z$-axis. The numerical method involves setting up
150: the Hamiltonian matrix in the product basis of the spin and phonon states
151: $|i,j>$, where $|i>$ corresponds to one of the eight spin configurations of
152: the three spins, and $j$ varies from $0$ to $14$, corresponding to the fifteen
153: phonon states retained in the problem. The values we have assigned to the
154: different parameters are $\gamma=10^{-3}$, $q=10^{-4}$ and $\hbar \omega =
155: 1.25 \times 10^{-4}$, all in units of the exchange $J$ (see Fig. 1).
156:
157: To study the magnetization phenomena, we start with the direct product
158: eigenstates of $H_{sp-sp}$ and $\hbar \omega (b^\dagger b + 1/2)$, and
159: independently evolve each of the $120$ states $\psi_{ij}$ by using the time
160: evolution operator
161: \beq
162: \psi (t + \Delta t) ~=~ e^{-i H_{total} \Delta t/\hbar} ~\psi(t) ~.
163: \eeq
164: The evolution is carried out in small time steps by applying the evolution
165: operator to the state arrived at in the previous step. The magnetic field
166: is changed step-wise in units of $0.015 T$. At each value of the magnetic
167: field, the system is allowed to evolve for $300$ time steps of size $\Delta
168: t$, before the field is changed to the next value. At every time step, the
169: average magnetization $<M(t)>$ is calculated as
170: \bea
171: < M(t) > ~=~ \sum_{i=1}^8 ~\sum_{j=0}^{14} && ~
172: \frac{e^{-\beta [w_i+ h_{z}(t)m_i]}}{Z_{spin}} ~\frac{e^{-\beta \hbar \omega
173: (j + 1/2)}}{Z_{ph}}~ \nonumber \\
174: && ~~~ <\psi_{ij} (t)| \hat{S_z} |\psi_{ij} (t)> ~,
175: \eea
176: where $w_i$ and $m_i$ are the eigenvalues and magnetizations of eigenstates of
177: $H_{sp-sp}$, $\beta = 1/k_B T$, $Z_{ph}$ is the phonon partition function of
178: $\hbar \omega (b^\dagger b +1/2)$, and $Z_{spin}$
179: is the partition function of $H_{sp-sp}$ in the presence of the axial
180: magnetic field. In Fig. 2, we show the energy level ordering of the effective
181: spin Hamiltonian and the effect of the magnetic field on the eigenvalue
182: spectrum. We also show the couplings between various states brought about by
183: the magnetic dipolar terms and the spin-phonon terms; note that the spin-$1/2$
184: and spin-$3/2$ states are not connected to each other by these terms.
185:
186: In Fig. 3, we show the hysteresis plots of the system for different
187: temperatures. We see that at low temperatures, the plateaus in the
188: hysteresis plots are very pronounced. The plateau width at $<S_z> = - 0.5$
189: corresponds to $2.64$ T which is in excellent agreement with the experimental
190: value of $2.8$ T \cite{wernsdorferv15} assuming that $J=800~K$.
191: We also find that the plateau vanishes above a temperature of $0.8$ K
192: which is also in excellent agreement with the experimental value of $0.9$ K
193: \cite{wernsdorferv15}. The inset in figure shows the temperature variation
194: of the plateau width. We note that the plateau width falls off rapidly with
195: temperature, and an exponential fit to $W~=~A~\exp (-T/\Omega)$
196: (see Fig. 3) gives the characteristic temperature $\Omega$
197: to be $0.2$ K. This small value of $\Omega$ is because there are no large
198: barriers between the different magnetization states in this system, unlike
199: the high spin molecular magnets such as $\rm Mn_{12}$ \cite{model}.
200:
201: We also observe that when the field is swept more rapidly, there are
202: additional plateaus at intermediate values of magnetization. For example, in
203: Fig. 4 the field sweep rate is increased by a factor of five compared to
204: Fig. 3, and we find a small plateau of width $0.03 ~T$ near $H=0.15 ~T$
205: at a value of $< S_z > = -0.375$. This is because near that field, some of
206: the spin-$3/2$ states become degenerate in energy; subsequently, as the
207: magnetic field is increased, the system stays locked
208: in some of those states if the sweep rate is too high. This plateau vanishes
209: on warming the system slightly. We have also studied the effect of cycling
210: the field on the plateaus. In Fig. 5, we show the magnetization as a
211: function of field at different temperatures and sweep rates. It is
212: interesting to see that the system does not have either remnance or
213: coercivity; all the hysteresis plots pass through the origin. The effect of
214: varying the rate of scanning the field is also shown in Fig. 5.
215: We find that as the scanning rate increases, the hysteresis in the plot of
216: magnetization {\it vs} field decreases, and the plateau feature is
217: almost identical in both the scanning directions. This could be
218: because of the slow relaxation of the magnetization which is indeed the
219: reason why the plateaus occur in the first place. We also find that the
220: transverse field term does not affect any of our results significantly.
221:
222: To summarize, we have derived an effective Hamiltonian from the
223: exchange Hamiltonian of the full $\rm V_{15}$ system. In the presence of a
224: time varying magnetic field, the states of the effective Hamiltonian are
225: allowed to evolve under the influence of magnetic dipolar interactions and
226: a spin-phonon coupling. During the time evolution, the magnetization is
227: followed as a function of the applied magnetic field. The calculated
228: $M$ {\it vs} $H$ plots show magnetization plateaus at low temperatures. The
229: width of the plateau at low temperature as well as the temperature at which
230: the plateau vanishes are in excellent agreement with experimental values.
231: It is also shown that the number of plateaus observed depends upon the
232: scanning speed of the magnetic field. When the magnetic field is cycled, the
233: hysteresis plots pass through the origin indicating the absence of remnance
234: and coercion. The hysteresis is pronounced for slow scanning speeds. From our
235: results, it appears that the magnetization plateaus in $\rm V_{15}$ is not a
236: consequence of quantum resonant tunneling but is a result of thermal
237: averaging. We also find that the magnetization does not show any oscillation
238: with time during evolution indicating the absence of quantum tunneling.
239:
240: We thank the Council of Scientific and Industrial Research, India for their
241: grant No. 01(1595)/99/EMR-II.
242:
243: \begin{thebibliography}{99}
244:
245: \bibitem{synth} T. Lis, Acta Crystallogr. B {\bf 36}, 2042 (1980); C. Delfas,
246: D. Gatteschi, L. Pardi, R. Sessoli, K. Wieghardt and D. Hanke, Inorg. Chem.
247: {\bf 32}, 3099 (1993); A. L. Barra, D. Gatteschi, L. Pardi, A. M${\rm \ddot
248: u}$ller and J. D${\rm \ddot o}$ring, J. Am. Chem. Soc. {\bf 114}, 8509 (1992).
249:
250: \bibitem{qtm} {\it Quantum Tunneling of Magnetization - QTM'94}, edited by L.
251: Gunther and B. Barbara, NATO ASI Ser. E, Vol. 301 (Kluwer, Dordrecht, 1995);
252: J. R. Friedman, M. P. Sarachik, J. Tejada and R. Ziolo, Phys. Rev. Lett. {\bf
253: 76}, 3830 (1996); L. Thomas, F. Lionti, R. Ballou, D. Gatteschi, R. Sessoli
254: and B. Barbara, Nature {\bf 383}, 145(1996) and references therein.
255:
256: \bibitem{v15} D. Gatteschi, L. Pardi, A. L. Pardi, A. M${\rm \ddot u}$ller and
257: J. D${\rm \ddot o}$ring, Nature {\bf 354}, 465 (1991).
258:
259: \bibitem{wernsdorferfe8} W. Wernsdorfer and R. Sessoli, Science {\bf 284}, 133
260: (1999).
261:
262: \bibitem{qcompu} E. M. Chudnovsky, Science {\bf 274}, 938(1996); P. C. E.
263: Stamp, Nature {\bf 383}, 125 (1996); L. Thomas, F. Lionti, R. Ballou, D.
264: Gatteschi, R. Sessoli and B. Barbara, Nature {\bf 383}, 145 (1996).
265:
266: \bibitem{model} H. De Raedt, S. Miyashita, K. Saito, D. Garcia-Pablos and N.
267: Garcia, Phys. Rev. B {\bf 56}, 11761 (1997); D. A. Garanin and E. M.
268: Chudnovsky, Phys. Rev. B {\bf 56}, 11102 (1997); A. Garg, Phys. Rev. Lett.
269: {\bf 81}, 1513 (1998); I. S. Tupitsyn, JETP Lett. {\bf 67}, 32 (1998),
270: cond-mat/9712302; N. V. Prokofev and P. C. E. Stamp, J. Low. Temp. Phys. {\bf
271: 104}, 143 (1996); M. I. Katsnelson, V. V. Dobrovitski and B. N. Harmon, Phys.
272: Rev. B {\bf 509}, 6919 (1999); I. Rudra, S. Ramasesha and D. Sen,
273: cond-mat/0008144.
274:
275: \bibitem{mastereq} P. Politi, A. Rettori, F. Hartmann-Boutron and J. Villain,
276: Phys. Rev. Lett. {\bf 75}, 537 (1995); A. Fort, A. Rettori, J. Villain, D.
277: Gatteschi and R. Sessoli, Phys. Rev. Lett. {\bf 80}, 612(1998).
278:
279: \bibitem{decoherence} V. V. Dobrovitski, M. I. Katsnelson and B. N. Harmon,
280: Phys. Rev. Lett. {\bf 84}, 3458 (2000).
281:
282: \bibitem{wernsdorferv15} I. Chiorescu, W. Wernsdorfer, A. M${\rm \ddot
283: u}$ller, H. B${\rm \ddot o}$gge and B. Barbara, Phys. Rev. Lett. {\bf 84},
284: 3454 (2000) and J. Magn. Magn. Mat. {\bf 221}, 103 (2000).
285:
286: \bibitem{spph} M. N. Leuenberger and D. Loss, Europhys. Lett. {\bf 46}, 692
287: (1999) and Phys. Rev. B {\bf 61}, 1286 (2000); E. Callen and H. B. Callen,
288: Phys. Rev. {\bf 139}, A455 (1965).
289:
290: \end{thebibliography}
291:
292: \begin{figure}
293: \begin{center}
294: %\epsfig{figure=fig1.ps,height=11cm}
295: \epsfig{figure=fig1.ps,height=9cm}
296: \end{center}
297: \caption{Schematic exchange interactions in a $\rm V_{15}$ cluster. There is
298: no direct exchange interaction amongst the triangle spins. Interactions not
299: shown explicitly can be generated from the $C_3$ symmetry of the system.}
300: \end{figure}
301:
302: \begin{figure}
303: \begin{center}
304: %\epsfig{figure=fig2.ps,bbllx=50,bblly=0,bburx=500,bbury=800,height=11cm}
305: \epsfig{figure=fig2.ps,bbllx=50,bblly=0,bburx=500,bbury=800,height=9cm}
306: \end{center}
307: \caption{(a) Eigenstates of the effective spin Hamiltonian
308: $H_{sp-sp}$, (b) Eigenstates in the presence of a moderate axial field.
309: Arrows show the states connected by the dipolar terms and the transverse
310: field. (c) is the same as (b) but in a stronger field, (d) describes the
311: effect of spin-phonon terms (shown by arrows with broken lines) on (c).}
312: \end{figure}
313:
314: \begin{figure}
315: \begin{center}
316: %\epsfig{figure=fig3.ps,bbllx=50,bblly=0,bburx=500,bbury=800,height=11cm}
317: \epsfig{figure=fig3.ps,bbllx=50,bblly=0,bburx=500,bbury=800,height=9cm}
318: \end{center}
319: \caption{Plot of magnetization {\it vs} axial field at different temperatures.
320: Inset shows plateau width as a function of temperature (full circles).
321: Triangles in the inset correspond to the values from the fit to an exponential
322: function.}
323: \end{figure}
324:
325: \begin{figure}
326: \begin{center}
327: %\epsfig{figure=fig4.ps,bbllx=50,bblly=0,bburx=500,bbury=800,height=11cm}
328: \epsfig{figure=fig4.ps,bbllx=50,bblly=0,bburx=500,bbury=800,height=9cm}
329: \end{center}
330: \caption{Magnetization as a function of axial field for a faster sweep rate at
331: three different temperatures.}
332: \end{figure}
333:
334: \begin{figure}
335: \begin{center}
336: %\epsfig{figure=fig5.ps,bbllx=50,bblly=0,bburx=500,bbury=800,height=11cm}
337: \epsfig{figure=fig5.ps,bbllx=50,bblly=0,bburx=500,bbury=800,height=9cm}
338: \end{center}
339: \caption{Magnetization {\it vs} axial field for a full cycling of the field at
340: different temperatures and sweep rates.}
341:
342: \end{figure}
343:
344: \end{document}
345:
346: