1: \documentclass[twocolumn,showpacs,preprintnumbers,amsmath,amssymb]{revtex4}
2:
3: %\documentclass[preprint,preprintnumbers,amsmath,amssymb]{revtex4}
4: %\documentstyle[aps]{revtex}
5:
6: \begin{document}
7: \draft
8:
9: \title{FRACTIONAL PERIODICITY OF PERSISTENT CURRENTS:
10: A SIGNATURE OF BROKEN INTERNAL SYMMETRY}
11:
12: \author{P. Singha Deo$^1$, P. Koskinen$^2$, M. Koskinen$^2$,
13: and M. Manninen$^2$}
14: \address{$^1$ S. N. Bose National Centre for Basic Sciences, JD Block,
15: Sector III, Salt Lake City, Kolkata 98, India.}
16: \address{$^2$ Department of Physics, University of Jyv\"askyl\"a,
17: FIN-40351 Jyv\"askyl\"a, Finland}
18: \date{\today}
19:
20:
21:
22: \begin{abstract}
23: We show from the symmetries of the many body Hamiltonian, cast
24: into the form of the Heisenberg (spin) Hamiltonian, that the fractional
25: periodicities of persistent currents are due to the breakdown
26: of internal symmetry and the spin Hamiltonian
27: holds the explanation to this transition.
28: Numerical diagonalizations are performed
29: to show this explicitely.
30: Persistent currents therefore, provide an easy way to experimentally
31: verify broken internal symmetry in electronic systems.
32: \end{abstract}
33:
34: \pacs{PACS: 67.40.Db, 73.21.La}
35:
36: \maketitle
37:
38:
39:
40: Remarkable advances in fabrication techniques, now make it possible
41: to confine a few electrons in a conducting wire where electron motion
42: is governed by quantum mechanics, rather than classical mechanics.
43: The system becomes an electron waveguide within the confinement
44: potential. The phase coherence of the electrons is maintained over the
45: sample, making it possible to observe several intrinsic quantum
46: mechanical phenomena including
47: Aharonov-Bohm oscillations, universal conductance fluctuations,
48: quantized conductance in point contact, quenching of Hall resistance
49: in narrow cross \cite{boo},
50: current magnification effect \cite{deo1}, etc. A remarkable consequence
51: of such coherence is the existence of equilibrium persistent
52: currents \cite{hun}
53: in a ring threaded by an Aharonov-Bohm flux,
54: that was first predicted theoretically \cite{but}
55: and subsequently detected experimentally \cite{exp} in mesoscopic
56: systems. Aharonov-Bohm flux here refers to a situation
57: when the magnetic field is restricted to a small region in the
58: center of the ring, and the electrons in the ring do not feel
59: the magnetic field. Although these are
60: equilibrium currents, they can yield information about transport
61: \cite{deo2}
62: and may help us to understand the effects of electron-electron
63: interaction on transport, using Hamiltonian diagonalization techniques.
64:
65: Interacting fermions exhibit very novel properties that fascinate
66: scientists for a long time. Interacting nucleons for example has
67: shown many novel features of interactions \cite{boh}.
68: Of special relevance to this work, is the discovery that
69: certain heavy nuclei can exhibit rotational excitations,
70: that could not be explained by the shell structure of a spherical
71: neuclous. Initial understanding was provided by Bohr and Mottelson,
72: in terms of collective modes of a deformed nuclei, in the simplest
73: model as rigid rotation of the deformed nucleus.
74: Similar ideas of {\it internal} symmetry breaking,
75: explained the details of the mass spectra of alkali metal
76: clusters\cite{clemenger85}, and suggest an existence of static
77: spin-density waves in quantum dots\cite{kos1}.
78: Indeed, mesoscopic systems give us an unique opportunity
79: to access regimes that do not occur naturally and study
80: a few electrons in man made quantum dots, both
81: experimentally and
82: simultaneously with almost
83: exact theoretical methods. Hence these systems can give us a rare
84: opportunity to study how few electron properties evolve into
85: macroscopic collective properties as we increase the number of electrons.
86:
87: Wigner crystalization of electrons is one such bulk phenomenon
88: and is still a very debatable
89: issue, although it was first proposed a very long time ago.
90: We hereby exclude the situation when the quantum mechanical
91: kinetic energy or uncertainty of an electron can be quenched
92: by a strong magnetic field.
93: Theoretically one can find signatures of a crystal structure when one looks
94: at the conditional probability (the probability of
95: finding other electrons when the coordinates of one electron
96: is fixed by hand) of the interacting electrons, while the
97: probability itself (density) does not show any sign of a crystal structure.
98: The conditional probability shows oscillations\cite{borrmann01}
99: which suggest broken internal symmetry. However, in finite systems
100: the situation is not so clear due to the fact that the
101: correlation coming from the Pauli exclusion principle alone
102: will cause oscillating conditional probability at short distances,
103: even in noninteracting systems.
104: On the other hand a recent work of Koskinen et al \cite{kos1}
105: raises the
106: issue that a few electron system in a quantum dot
107: and quantum ring can exhibit broken internal
108: symmetry. Their mean field studies
109: of the electron probability (density) showed a perfect crystal structure.
110: It was subsequently shown \cite{vei} that the quantum
111: mechanical superposition of states
112: is destroyed by the non-linearity of the mean field and
113: as a result the internal symmetry
114: is mapped out (as a consequence the
115: angular momentum quantum number does not take
116: integral values anymore).
117: Effects of non-linearity
118: cannot be ignored (the exact cause of non-linearity
119: not being important) and can well lead to
120: a quenching of the
121: quantum mechanical uncertainty and result in Wigner crystals.
122: Hence at this state it would be useful
123: to find some experimental ways of determining if the internal symmetry
124: is broken or not and what are the signatures that one
125: should look for when treating larger systems using
126: approximate methods like mean field theories and
127: effective Hamiltonians. For example, in nuclei, the experimental evidence
128: of such broken symmetry states can be obtained from the
129: rotational spectrum.
130: The purpose of this work is to show that the measurement of the
131: persistent current can give signature of broken internal symmetry
132: of electronic states in small quantum rings.
133:
134: The many-body Hamiltonian for electrons in a quasi-1D-ring
135: can be written as
136: \begin{equation}
137: H=\sum_{i}\left(-{{\hbar}^2\over{2m^*}}\nabla_i^2 + V(r_i)\right)
138: +\sum_{i<j}{{e^2}\over{4\pi\epsilon_0\epsilon\vert {\vec r}_i
139: -{\vec r}_j\vert}}
140: \label{hamilton}
141: \end{equation}
142: where $V$ is the potential confining the electron in the ring.
143: $V$ is assumed to have a circular symmetry.
144: Koskinen {\it et al} \cite{kos2,koskinen2002}
145: have performed exact numerical computations
146: for a few electrons confined in such a ring and shown that
147: for narrow rings the many-body spectrum can be described
148: essentially exactly with a simple model Hamiltonian
149: \begin{equation}
150: H_{eff}= J\sum_i {\vec S}_i\cdot {\vec S}_j +
151: {{1}\over{2I}}M^2 + \sum_\alpha \hbar\omega_\alpha n_\alpha
152: \label{modelh}
153: \end{equation}
154: where the first term is an antiferromagnetic Heisenberg
155: Hamiltonian, the second term describes rigid rotations of
156: the electron system ($M$ is the centre of mass angular momentum
157: and $I$ is the moment of inertia), and
158: the last term describes the vibrational states of localized
159: electrons. Koskinen {\it et al} \cite{kos2,koskinen2002}
160: compared the energy spectra
161: of exact diagonalization of Hamiltonian (\ref{hamilton}) to those
162: of the model Hamiltonian (\ref{modelh}) and found
163: an excellent agreement for hundreds of many-body states in rings with
164: from 2 to 7 electrons.
165:
166: The model Hamiltonian can be understood as a result of localization
167: of electrons and forming a Wigner molecule which is freely
168: rotating in the external potential\cite{kos2}.
169: Related ideas of electron localization in noncircular dots
170: had been suggested earlier\cite{jef,cre}.
171: Assuming localization, the Coulomb energy of the
172: exact Hamiltonian (\ref{hamilton}) can be expanded around the
173: classical equilibrium positions of electrons.
174: This leads to potential wells at each classical site.
175: The tight binding model of the system can be described by a
176: half filled Hubbard model, which in the limit of the strong
177: Coulomb energy (Hubbard $U$) leads to the antiferromagnetic
178: Heisenberg model \cite{vol}.
179: It is important to note, however, that for our continuous system
180: where the localizing potential
181: is not an external potential, we do not get an insulating
182: phase for the half filled case \cite{kus}.
183: The rigid rotation and the the vibrations of the localized
184: electrons can be assumed to separate out from the spin Hamiltonian
185: leading to the simple effective Hamiltonian of Eq. (\ref{modelh}).
186: The antisymmetry of the total Hamiltonian have to be taken into
187: account by matching the symmetries of the different parts of the
188: wave function (spin, rotations, vibrations)\cite{koskinen2002}.
189: In the present case when we are studying only the ground state
190: properties (persistent current being a ground state
191: property) the vibrational states do not play any role.
192:
193: An external magnetic field will bring two additional terms
194: in the model Hamiltonian. The gauge field will change the
195: angular momentum part and the direct interaction with the
196: electron spins will add a Zeeman term.
197: Since we are interested in the equilibrium persistent currents,
198: we will assume that the magnetic field is confined only
199: inside the ring so that the Zeeman energy is absent.
200: By ignoring the vibrations
201: (which have higher energy than rotations)
202: the model Hamiltonian in the presence of a magnetic field
203: flux $\phi$ penetrating the ring is
204: \begin{equation}
205: H_B=J\sum_{i,j}{\vec S}_i\cdot {\vec S}_j
206: +{{(M-N\phi)^2}\over{2I}}
207: \label{modelb}
208: \end{equation}
209: where $M$ is the angular momentum, $N$ the number of electrons in the ring and
210: $\phi$ is the flux through the ring in units of $\phi_0=hc/e$
211: i.e., $\phi$=${e \over hc} \int {\vec A}\cdot {\vec dr}$
212: $\vec A$ being the vector potential.
213: The strongly interacting case here correspond to $J \rightarrow 0$,
214: $J$=0 being the classical case when the electrons do not overlap.
215: In this case, there is no
216: uncertainty in the internal frame of the system and it is a perfect
217: crystal in its internal frame.
218: It is important to notice that keeping $I$ fixed, the small $J$
219: limit correspond to a narrow ring with strong correlation and the
220: large $J$ limit approaches the non-interacting case.
221:
222: The correspondence between (1) and (3) shown
223: in Ref. \cite{kos2} in the absence of flux,
224: can be easily extended to the case
225: when there is Aharonov-Bohm flux. First of all
226: the Hamiltonian in (3) cannot have any extra
227: contribution from the 2nd term in the Hamiltonian in (1).
228: This is because when we write the Hamiltonian in (1)
229: in terms of a centre of mass coordinate
230: and relative coordinates,
231: the second term contains only
232: relative coordinates,
233: and will only affect the vibrational states,
234: which in turn does not affect
235: the persistent currents. This means that $J$
236: is independent of flux because the second term in
237: (1) only depends on relative coordinates and is not
238: affected by the flux.
239: Another way to prove this is to show that the Coulomb
240: matrix elements are independent of Aharonov Bohm flux,
241: which can be shown analytically.
242: This is also evident in the numerical calculations \cite{pie,tap}.
243: Secondly, the flux also cannot affect the relative motions (kinetic
244: energy part) of the electrons and this was proved in Ref. \cite{wei}.
245: The proof essentially puts all the flux dependence on the wavefunction
246: by gauging away the flux dependence of the Hamiltonian. Writing the many
247: body wavefunction in presence of flux as a linear combination of Slater
248: determinants, constructed from the flux dependent, non-interacting,
249: single particle wave functions, it is easy to show that all the flux
250: dependence of the relative coordinates cancel each other. The second term
251: in (3) also directly follow from there, once we put the flux dependence
252: back into the Hamiltonian.
253:
254: The persistent current can be determined as a derivative of the
255: ground state energy with respect of the flux\cite{hun}.
256: Consequently, it is sufficient to study the periodicity of the
257: ground state energy as a function of the flux.
258: First of all it can be shown that the Heisenberg Hamiltonian remains
259: unchanged under the transformation $M \rightarrow M+N$ and so we need to
260: consider only the first N eigen-energies of the system.
261: In Fig. 1 we show a contour plot of the eigen-energies for N=
262: 4, 5, 6 and 8 in the $J-\phi$-plane.
263: When $J \rightarrow $0
264: then the periodicity with flux is $\phi_0/N$ and correspond to
265: the case when the flux quantum is $\phi_0/N$ corresponding to the
266: rigid rotor of charge $Ne$. As $J$ is increased, signifying that
267: the electrons are getting delocalized and overlapping with each other,
268: the periodicity changes smoothly to $\phi_0$. For even number of
269: electrons, $\phi_0/N$
270: periodicity first changes over to $\phi_0/2$ periodicity before
271: the full $\phi_0$ periodicity is recovered. For odd $N$ also
272: $\phi_0$ periodicity changes to $\phi_0/2$
273: periodicity, which unlike the case of even $N$,
274: remain all the way up to $J=\infty$. This can also be see in
275: Fig. 2 where we plot the $M$ values of the ground state for
276: different flux $\phi / \phi_0$
277: and $Jr_0^2$ ($r_0$ being the radius of the ring, $I=Nmr_0^2$).
278: The ground state switches
279: its $M$ values as shown in the figure 2
280: in the different parameter regimes.
281: While for even $N$ one can see converging
282: phase regions that cannot be extended to infinity,
283: for $N=5$ the line separating $M=1$ and $M=4$
284: is a vertical line that can be extended to infinity. This is also
285: consistent with the fact that when $J$ is large we should recover
286: the free electron results and odd number of spin-full free
287: electrons in a ring always give $\phi_0/2$ periodicity
288: \cite{los}. But,
289: for even $N$ the $\phi_0/N$ periodicity first changes to
290: $\phi_0/2$ periodicity as $J$ is increased,
291: and then the free electron result
292: of $\phi_0$ periodicity is recovered for very large $J$.
293:
294: The value of flux which gives the minimum total energy
295: for large $J$ depends
296: on the number of electrons in the ring. For even particles with
297: $N=4$, 8, 12, etc. ($N/2$ is even)
298: the minimum is at flux $\phi=\phi_0/2$ (see fig. 3) while
299: for $N=6$, 10 etc. ($N/2$ is odd)
300: the minimum is at $\phi=0$. The reason is the
301: symmetry of the solution of the Heisenberg Hamiltonian
302: as can be proved for any $N$ from group theoretical analysis.
303: For odd $N/2$ the minimum energy corresponds
304: to $M=0$ and the second lowest state
305: has $M=N/2$ while
306: for even $N/2$ the minimum energy has $M=N/2$ and the second lowest state
307: has $M=0$ (see fig. 3 as an example).
308: In both cases it happens that at a certain region of $J$, when the
309: angular momentum is increased, the ground state jumps between these
310: two lowest states leading to $\phi_0/2$ periodicity in the
311: total energy and persistent current. When $J$ becomes large enough
312: then the splitting between these two lowest states also become
313: very large and
314: only the ground state of the Heisenberg Hamiltonian matters
315: and $\phi_0$ periodicity is obtained.
316: For odd number of electrons (odd $N$)
317: the situation is different
318: since there are two angular momentum values corresponding
319: to the minimum energy of the Heisenberg Hamiltonian.
320: For noninteracting electrons these two values appear
321: at angular momenta (in the limit of $N\rightarrow \infty$)
322: $N/4$ and $3N/4$.
323: Consequently, there are two equal energy minima leading to
324: $\phi_0/2$ periodicity even at infinitely large values of $J$.
325: In the case of small number of electrons the energy minimum
326: can not occur exactly at $N/4$ and
327: the $\phi_0$ and $\phi_0/2$ periodities are
328: superimposed as in Fig. 1(b). This
329: signifies that for intermediate $J$ values, when there is
330: no longer any correlation between all the $N$ particles as
331: in a rigid rotor,
332: there still seems to be correlation between electron pairs
333: leading to quasi-particles with charge $2e$.
334: %there is still a correlation between
335: %nearest neighbors, and there are quasi-particles of charge $2e$.
336: For odd $N$ since every electron cannot find a pair (similar
337: to the parity effect in superconducting grains), such
338: correlation is not possible. Spin
339: values may or may not change as flux is changed from 0 to
340: $\phi_0$. The sequence of change is depicted in Fig. 2.
341: The $\phi_0/N$ periodicity occurs when $J$ is small and
342: the splitting between the states is so small that the flux
343: can create transition through all the states.
344:
345: It should be noted that the flux dependence of the eigenenergies,
346: in a clean ring, should be the same in presence and absence of interactions
347: \cite{wei}. A many body eigenenergy, in absence of interactions, should
348: change parabolicaly with $N\phi$.
349: Our model Hamiltonian in (3) is consistent with this
350: and the flux dependence of a given state is always parabolic in the
351: calculated eigenenergies. However, as the periodicity is reduced by N,
352: the amplitude of the persistent currents is also reduced by N as
353: the Brillouin zone becomes 1/N of the non-interacting case. For large
354: $J$, when we recover the $\phi_0$ periodicity, once again the flux
355: dependence of a particular many body state changes parabolically with
356: $N\phi$, all the way up to the non-interacting zone boundary.
357:
358: The relation of the parameters of the model Hamiltonian (2) to those of the
359: original Hamiltonian (1) require exact diagonalization of the latter.
360: The results of Koskinen et al\cite{kos2,koskinen2002} indicate that the onset
361: of the $\phi_0/N$ periodicity, which happens at $Jr_0^2\approx 0.1$,
362: can be obtained with a ring with $V(r)={1\over 2}m^*\omega_0^2(r-r_0)^2$,
363: by choosing (in effective atomic units) $\omega_0\approx 1/m^*r_0^2$
364: (for $N=6$).
365: For a material with $m^*=0.1$ and $\epsilon=10$ this condition could
366: be achieved,
367: for example, with $r_0=80$ nm and $\omega_0=1.7$ meV.
368:
369: Fractional periodicities of persistent currents in a 1D
370: Hubbard ring has been discussed before \cite {kus, wu, kot}.
371: These studies correspond to a situation when there is an
372: externally applied periodic potential or a lattice.
373: In presence of such an externally applied potential
374: the concept of Wigner crystal is not meaningful.
375: The observed fractional periodicities, Kotlyar et al\cite{kot} associate
376: with magnon excitations.
377: How this interpretation relates to our findings
378: is an interesting subject of future studies.
379:
380: Breakdown of internal symmetry of a many body system \cite{kos1},
381: crystallization of electrons in the bulk (Wigner crystals),
382: fractional periodicities of persistent currents \cite{tap} are three
383: different intersting research topics that are brought together
384: in this work. It is shown that fractional periodicity of
385: persistent currents is due to the breakdown of internal
386: symmetry.
387: The Hamiltonian,
388: diagonalized upto 8 electrons in a ring threaded by a flux
389: show this explicitely and symmetry considerations establish
390: this for any $N$. Broken internal symmetry in electronic
391: systems is of
392: special importance as the interaction between electrons
393: is well known as compared to that between neucleons.
394: It is difficult to observe such broken symmetry states
395: because most of the physical quantities that can be measured,
396: do not depend on whether the internal symmetry is
397: broken or not. In the nuclei, the only evidence of such
398: broken internal symmetry comes from the rotational and vibrational
399: spectrum of a nuclei. For electrons embedded in a solid
400: the equilibrium persistent currents provide a way
401: to find this evidence.
402: At present a few electron ring can be realized \cite{lor} and possibly
403: reveals the much sought experimental proof
404: of broken internal symmetry in an electronic system.
405:
406:
407: This work has been supported by the Academy of Finland under
408: the Finnish Centre of Excellence Programme 2000-2005 (Project No.
409: 44875, Nuclear and Condensed Matter Programme at JYFL).
410:
411: \begin{thebibliography}{99}
412: \bibitem{boo} Quantum coherence in mesoscopic systems, Vol. 254 of NATO
413: Advanced Study Institute, Series B: Physics, edited by
414: B. Kramer (Plenum, New York, 1991); Mesoscopic Phenomenon in solids,
415: edited by B.L. Altshuler, P.A. Lee, and R.A. Webb (North-Holland,
416: Amsteradam, 1991).
417:
418: \bibitem{deo1} A. M. Jayannavar and P. Singha Deo, Phys. Rev. B {\bf 51},
419: 10175 (1995)
420:
421: \bibitem{hun} F. Hund, Ann. Phys. (Leipzig) {\bf 32}, 102 (1938);
422: N. Byers and C.N. Yang, Phys. Rev. Lett. {\bf 46}, 7 (1961);
423: F. Bloch, Phys. Rev. {\bf 137}, A787 (1965).
424:
425: \bibitem{but} M. B{\"u}ttiker, Y. Imry and R. Landauer, Phys. Lett. A
426: {\bf 96}, 365 (1983).
427:
428: \bibitem{exp} L.P. Levy, et al. Phys. Rev. Lett. {\bf 64}, 2074 (1990).
429: V. Chandrasekhar, et al. Phys. Rev. Lett. {\bf 67}, 3578 (1991);
430: D. Maily, et al. Phys. Rev. Lett. {\bf 70}, 2020 (1993).
431:
432: \bibitem{deo2} P. Singha Deo, Phys. Rev. B {\bf 55}, 13795 (1997).
433:
434: \bibitem{boh} A. Bohr and B. Mottelson, Nuclear structure (Benjamin,
435: New York, 1975).
436:
437: \bibitem{clemenger85} K. Clemenger, Phys. Rev. B {\bf 32},1359 (1985).
438:
439: \bibitem{borrmann01} P. Borrmann and J. Harting, Phys. Rev. Lett.
440: {\bf 86}, 3120 (2001).
441:
442: \bibitem{kos1} M. Koskinen, M. Manninen and S. M. Reimann,
443: Phys. Rev. Lett. {\bf 79}, 1389 (1997);
444: C. Yannouleas and U. Landman, Phys. Rev. Lett. {\bf 82}, 5325 (1999);
445: A. Puente and L. Serra, Phys. Rev. Lett. {\bf 83}, 3266 (1999).
446:
447: \bibitem{vei} S. Viefers, P. Singha Deo, M. Koskinen, S.M. Reimann,
448: and M. Manninen, Phys. Rev. B {\bf 62}, 10668 (2000)
449: \bibitem{kos2} M. Koskinen, M. Manninen, B. Mottelson and S. M. Reimann,
450: Phys. Rev. B. {\bf 63},205323 (2001).
451:
452: \bibitem{koskinen2002} P. Koskinen, M. Koskinen, and M. Manninen,
453: Eur. Phys. J. B {\bf 28}, 483 (2002).
454:
455: \bibitem{jef} J.H. Jefferson and W. Hausler, Phys. Rev. B {\bf 54},
456: 4936 (1996).
457:
458: \bibitem{cre} C.E. Creffield, W. Hausler, J.H. Jefferson and S. Sarkar,
459: Phys. Rev. B {\bf 59} 10719 (1999).
460:
461: \bibitem{vol} D. Vollhardt in ``Proceedings of the international
462: school of physics, Enrico Fermi - Perspectives in many particle
463: physics'', North Holland, 1994.
464: \bibitem{kus} F.V. Kusmartsev, J. Phys.: Cond. Mat. {\bf 3}, 3199 (1991).
465: \bibitem{pie} T. Chakraborty and P. Pietilainen, Phys. Rev. B {\bf 52},
466: 1932 (1995).
467: \bibitem{los} D. Loss and P. Goldbart, Phys. Rev. B {\bf 43}, 13762 (1991).
468: \bibitem{tap} K. Niemela et al, Europhys. Lett. {\bf 36}, 533 (1996).
469: \bibitem{wei} A. Muller-Groeling, H. A. Weidenmuller and C. H. Lewenkopf,
470: Europhys. Lett. {\bf 22}, 193 (1993).
471: \bibitem{lor} A. Lorke et al, Phys. Rev. Lett. {\bf 84}, 2223 (2000).
472: %\bibitem{tcp} Proccedings of the
473: %``International workshop in artificial atoms'' in Trento, Italy (19.4.99
474: %to 30.4.99). Organized by ``European Center for Theoretical Studies in
475: %Nuclear Physics and Related Areas''.
476: \bibitem{wu} N. Wu and M. Fowler, Phys. Rev. B {\bf 45}, 11795 (1992).
477: \bibitem{kot} R. Kotlyar, C.A. Stafford, and S. Das Sarma, Phys.
478: Rev. B {\bf 58} 3989 (1998).
479: \end{thebibliography}
480:
481: \centerline{\bf Figure Captions}
482:
483: \noindent Fig. 1. 2D plot of the energy of the Heisenberg Hamiltonian
484: for electrons in a 1D ring versus $Jr^2$ and $\phi/\phi_0$
485: for N=4 (Fig. a), 5 (Fig. b), 6 (Fig. c) and 8 (Fig. d).
486: The dark areas are maxima and bright areas are minima. The persistent
487: is the derivative of the energy with respect to the flux and so
488: for the persistent currents dark areas are minima and bright areas
489: are maxima.
490:
491: \noindent Fig. 2. Phase diagram of the ground state angular momentum
492: of the Heisenberg Hamiltonian
493: for electrons in a 1D ring versus $Jr^2$ and $\phi/\phi_0$
494: for N=4 (Fig. a), 5 (Fig. b), 6 (Fig. c) and 8 (Fig. d).
495: The region with a particular shade denotes
496: the $Jr^2$ and $\phi/\phi_0$, corresponding to which the ground state
497: angular momentum is $M$, where $M$ is designated in the shaded regions
498: as M(S), where S is the total spin.
499:
500: \noindent Fig. 3. States of an 8 electron system in a 1D ring as obtained
501: by diagonalizing the Heisenberg Hamiltonian. $\phi/\phi_0$=0,
502: $Jr^2=1.0$, the M values
503: are given at the base and the S values are labeled by the side of each state.
504: The difference between the lowest state and the highest state in the Fig.
505: is 5.651$Jr^2$.
506:
507: \end{document}
508: