1: %\documentclass[twocolumn,shortnote]{jpsj2}
2: \documentclass[letter]{jpsj2}
3: %\documentclass[shortnote]{jpsj2}
4: \usepackage{graphicx}
5:
6: \def\runtitle{Dielectric Response of the Interacting 1D Spinless Fermions with Disorder}
7: \def\runauthor{Masato \textsc{Kishi} and Yasuhiro \textsc{Hatsugai}}
8:
9: \title{%
10: Dielectric Response of the Interacting 1D Spinless Fermions with Disorder
11: }
12: \author{%
13: Masato \textsc{Kishi}\thanks{E-mail: kishi@pothos.t.u-tokyo.ac.jp} and
14: Yasuhiro \textsc{Hatsugai}
15: }
16:
17: \inst{%
18: Department of Applied Physics, University of Tokyo, 7-3-1 Hongo\\
19: Bunkyo-ku, Tokyo 113-8656, Japan \\
20: }
21:
22: \recdate{\today}
23:
24: \abst
25: {
26: Dielectric responces of the one-dimentional electron system is
27: investigated numerically.
28: We treat an interacting one-dimentional spinless fermion model with
29: disorder by using the Density Matrix Renormalization Group(DMRG)
30: method which is extended for nonuniform systems.
31: We apply an electric field $E$ to the system
32: and calculate dielectric responces.
33: Dielectric responce of the Mott insulator and the Anderson insulator
34: are calculated respectively.
35: Steplike behaviors in $P-E$ curve are obtained which corresponds to
36: breakdown of the insulating behavior.
37: For the Mott insulator, the steps originate from generation of kink-pairs.
38: For the Anderson insulator on the other hand,
39: the origin of the steps is a crossing of the localized one particle
40: energy levels.
41: We also treat random systems with interaction.
42: From one parameter scaling analysis of the susceptibility $\chi$,
43: the metal-insulator transition in attractively interacting region
44: is confirmed and a phase
45: diagram of the random spinless fermion model is obtained.
46: }
47: %%% keywords %%%%%
48: \kword{%
49: Anderson localization, polarization, dielectric response, density
50: matrix renormalization group
51: }
52:
53: \begin{document}
54: \sloppy
55: \maketitle
56:
57:
58: \section{Introduction}
59: Effects of randomness and interaction in electronic systems are two
60: major problems in condensed matter physics.
61: These two problems have
62: been studied intensively for several decades and a lot of fundamental
63: results are accumulated.
64: The presence of disorder entails a localization of electronic states
65: due to a quantum mechanical interference of the Bloch states and this
66: phenomenon has been well studied as the Anderson localization.
67: According to the scaling theory, all states localize in one-
68: or two-dimensional systems without interaction,
69: no matter how weak the randomness is.
70: Interaction between electrons also changes states of electrons drastically.
71: Strong Coulomb interaction can lead a metallic system to an
72: insulator when the filling factor is rational and
73: attractive interaction may cause superconductivity.
74: Although we have well established understanding
75: of the localization of non-interacting electrons,
76: study on the correlated electrons with randomness is
77: still unsatisfactory\cite{rf:lee}. It is mainly due to lack of
78: reliable numerical techniques for the correlated electrons with randomness.
79:
80: Today, for the one-dimensional correlated systems without randomness,
81: a lot of numerical results are collected by the
82: Density Matrix Renormalization Group (DMRG)\cite{rf:1} method
83: and consistent understanding with analytical predictions
84: has been achieved\cite{rf:d}.
85: Recently, the DMRG was extended to random systems
86: and it has been applied to random spin chains\cite{rf:3}.
87: It brings us new possibilities to investigate correlation effects on the
88: random system.
89: In this paper, we concentrate on dielectric responses of
90: one-dimentional electronic systems with both randomness and interaction.
91: Applying a finite electric field $E$, we calculate a polarization $P$
92: as a derivative of ground state energy
93: by the field strength $E$. Linear susceptibility $\chi$ is also
94: obtained numerically as a derivative of the polarization by the electric field.
95: The polarization $P$ reflects deviation of mass center of electrons
96: by the field.
97: The dielectric response is a fundamental quantity
98: to characterize whether the system is a metal or an insulator.
99: It is suitably defined in open boundary conditions,
100: while Resta and Sorella defined polarization in periodic system recently.\cite{rf:c1}
101: Hence the DMRG is naturally applied to investigating the dielectric response.
102: Using the conventional DMRG, Aebischer {\it et al.} investigated the dielectric responce of Hubbard model with
103: next nearest neighbor hopping by
104: and confirmed the metal-insulator transition of the system.\cite{rf:4}
105: We apply the DMRG for nonuniform system by Hida\cite{rf:3} to random fermionic systems.
106:
107: We treat an interacting one-dimensional spinless fermion model in random
108: potential.
109: The Hamiltonian of the $L$ site system is given as
110: \begin{equation}
111: H=-t\sum_{i=1}^{L-1} (c_i^\dagger c_{i+1} + h.c.) + V
112: \sum_{i=1}^{L-1} n_in_{i+1} + \sum_{i=1}^L \epsilon_i n_i,
113: \end{equation}
114: where $n_i = c_i^\dagger c_i$ and $\epsilon_i$ is a random potential,
115: which distributes over an interval $[-W/2,W/2]$ uniformly.
116: We set $t = 1$ and consider the half-filling case
117: and impose an open boundary condition.
118: In the absence of disorder, the system is metallic for $-2<V<2$.
119: For
120: half-filling case, at $V=2$ the system undergoes a metal-insulator
121: transition and the system has a finite charge gap for $V > 2$. In an
122: attractive interaction region at $V=-2$, the system becomes unstable
123: due to phase separation.
124: When the randomness is present, the system is always insulating due to
125: the Anderson localization without interaction. Then the interplay between the
126: randomness and interaction can be interesting\cite{rf:ad1,rf:ad2} and possible metallic
127: phase is expected for some range of negative $V$.
128: Chui and Bray, and Apel treated the effect of interplay between randomness and
129: interaction on Tomonaga model.\cite{rf:8,rf:7} The former authors determined critical
130: value of the interaction constant of localize-delocalize transition by
131: analysis of the density autocorrelation function. The
132: latter considered the dynamic conductivity.
133: The spin-dependent backward scattering was also treated by Chui and
134: Bray\cite{rf:9a}, Apel and Rice\cite{rf:9b}, and Suzumura and Fukuyama.\cite{rf:9}
135: Giamarchi and Schulz took into account the renormalization of
136: interaction by the disorder and obtained a phase diagram.\cite{rf:6}
137: Runge and Zymanyi, and Bouzerar and Poilblanc estimated the size of
138: delocalized region in random interacting spinless fermion model
139: by the exact diagonalization.\cite{rf:a,rf:2}
140: Schmitteckert {\it et al.} used the DMRG and estimated the size of the
141: delocalized region from analysis of the phase sensitivity.\cite{rf:5}
142:
143: In sec. II, we discuss on the dielectric responses of systems.
144: Sec. III is for results and discussion.
145: Sec. IV is a summary.
146:
147: \section{Dielectric response}
148:
149: We focus on dielectric responses of the one dimensional system.
150: In order to observe the dielectric response directly, we apply the electric field $E$
151: to
152: the system. As a second quantized form of the potential, $-Ex$,
153: the coupling term
154: \begin{equation}
155: H_E = - E \sum_{i = 1}^L\left(i - \frac{L+1}{2}\right) en_i,
156: \end{equation}
157: is added to the Hamiltonian, where $e$ is the charge of the electron and we set $e = 1$ in this paper.
158: Then the full
159: Hamiltonian of the system is given by $H_T = H + H_E$.
160: As a function of $E$, the polarization $P$ of
161: the system is defined as
162: \begin{eqnarray}
163: P =- \frac{1}{L} \frac{ \partial E_{0}}{\partial E }
164: &=&- \frac{1}{L} \left\langle \frac{\partial H_E}{\partial E}
165: \right\rangle _E \\ \nonumber
166: &=& \frac{1}{L} \sum_{i = 1}^L \left(i-\frac{L+1}{2}\right) \langle n_i \rangle_E,
167: \end{eqnarray}
168: where $E_0$ is the ground state energy and $\langle n_i \rangle _E$
169: represents the ground state expectation value of $n_i$\cite{rf:4}.
170: Here we used the Feynman's theorem to derive the expression.
171:
172: In the Mott insulator, for a finite value of the electric field
173: whose potential energy is
174: comparable with the Mott gap, we expect a collapse of the charge
175: gap due to the interaction.
176: Then one can expect a gap closing which is a collapse of the
177: Mott gap at $E_{\rm g} \sim EL$ where $E_g$ is the Mott gap.
178: In the Anderson insulator, reconstruction of the charge occurs by
179: transferring electrons above the tunneling barrier.
180: The critical field strength $E_{\rm c}$ is estimated as $E_{\rm c}
181: \sim E_{\rm g}^{\rm A} / L$ where $E_{\rm g}^{\rm A}$ is an energy difference between
182: the highest occupied and the lowest empty one particle states.
183: $E_{\rm g}^{\rm A}$ is of the order of $1/L$ on the average.
184: In each case, we can obtain information on the charge degree of
185: freedom above its ground state.
186:
187: As for a linear response regime, $E \to 0$, we calculate
188: zero-field dielectric susceptibility as
189: \begin{equation}
190: \chi = \left. \frac{\partial P}{ \partial E} \right|_{E = 0}
191: = - \left. \frac{1}{L} \frac{\partial^2 E_0}{\partial E^2} \right|_{E=0}.
192: \end{equation}
193:
194: From the susceptibility $\chi$, we directly obtain information whether
195: the ground state is metallic or an insulator.
196: In the thermodynamic limit, $\chi$ is diverging if the system is metallic, but
197: converging to a finite value if it is an insulator.
198: Indeed $\chi \sim L^2$ is expected by the perturbation calculation for the pure
199: non-interacting system($W=0,V=0$). It is also confirmed numerically for
200: the pure interacting systems($W=0,-2<V<0$) by the DMRG(See later).
201:
202: In order to calculate the charge distribution and the ground state
203: energy, we use the exact diagonalization for $V=0$, and the DMRG
204: for the finite $V$.
205: In the application of the DMRG, we use the extended infinite-size
206: algorithm by Hida\cite{rf:3}, which enables us to treat
207: non-uniform lattice models also.
208:
209: %\section{DMRG for random systems}
210:
211: %In this study, we use the random DMRG method.\cite{rf:3}
212: %In the conventional DMRG, at each infinite-system steps,
213: %disorder configuration near the center of the lattice is incorrect.
214: %Therefore ground state may be far from the exact one and even if
215: %many finite system iterations are performed,
216: %it may converge to an incorrect state.
217: %By using the random DMRG, this problem shall be solved.
218: %In random DMRG, the disorder configuration is always kept when the size of system is increased.
219: %The random DMRG method for $N$ site systems is as follows.
220:
221: %1. For every odd $i (i=1..N-3)$, consider the configuration
222: %$\bullet_{i}\bullet_{i+1}\bullet_{i+2}\bullet_{i+3}$ (Fig\ \ref{fig:dmrg}(a)).
223: %Here $\bullet $ represents the single site on which the potential
224: %corresponds to $i$th site is added.
225: %Apply the iteration procedure of the standard DMRG to the four-sites
226: %system above
227: %and obtain the important $m$ states of $\bullet _{i}\bullet_{i+1}$ and $\bullet
228: %_{i+2}\bullet_{i+3}$.
229: %Then the $m$ states of the blocks $[2i-1:2i|$ and $|2i+1:2i+2]$ are obtained.
230: %Here $[i:j|$ represents the block made up of $i,i+1,..,j$-th sites and
231: %open boundary condition is imposed at the end of left-hand side
232: %and the end of right-hand side have been connected to neighbor blocks in
233: %previous iteration procedure.
234:
235: %2. Construct superblocks from two blocks which are separate three
236: %sites each other.
237: %Then apply the iteration procedure of
238: %the DMRG. i.e. use the configuration $[2i-1:2i|\bullet
239: %_{2i+1} \bullet _{2i+2} |2i+3:2i+4]$ for $i = 1..N/2-4$(Fig\ \ref{fig:dmrg}(b)).
240:
241: %3. Similar to Step. 2, combine two blocks separating three sites each
242: %other.
243: %The length of each block increases by one site at each iteration step.
244: %Repeat the iteration step till $[1:N/2-1|$ and $|N/2+1:N]$ is obtained.
245:
246: %This procedure corresponds to infinite system method of the conventional DMRG.
247: %We can apply the finite system method of conventional DMRG to random
248: %systems as the same way.
249: %The number of necessary iteration steps for the infinite system method of the random DMRG
250: %is of $O(N^2)$, though that of the conventional DMRG is $O(N)$.
251:
252:
253: \section{Results}
254: \subsection{Response of the two different insulators}
255: \subsubsection{Mott insulator}
256:
257: We calculate the polarization $P$ in the presence of interaction for
258: the pure systems by the DMRG.
259: For $V=1.0$, where the ground state of the system is the
260: Luttinger liquid.
261: Then the $P-E$ curve is smooth as shown in Fig.\ \ref{fig2}\cite{rf:loc}.
262: However for $V=6.0$, where the system is in a Mott insulator phase ,
263: the $P-E$ curve exhibits a stepwise behavior.
264: Since the polarization corresponds to the mass
265: center of the electrons, these steps represent a discontinuous change
266: of the charge configuration.
267: These steps actually come from the generation of kink-pairs.
268: In the inset of Fig.\ \ref{fig2}, the charge distribution at $E=E_{\rm s} \pm \Delta
269: E$ are shown, where $E_{\rm s}$ is the smallest value of the electric field
270: at which the kinks are generated.
271: At the step, the kink-pair is generated
272: and the electrons between the kinks are shifted by one site to the right which
273: compensates for the collapse of the Mott gap.
274: When the kinks are separated by length $l$,
275: the number of electrons between kinks is $l/2$. Then the energy gain
276: of the length $l$ kinks is
277: estimated as $\sim El/2$.
278: Then the first step is due to the kinks with longest length $L$.
279:
280: \subsubsection{Anderson insulator}
281:
282: We also calculated the polarization $P$ for the Anderson insulator.
283: Fig. \ref{fig2b} is the $P-E$ curve for the randomness strength $W=5$ without interaction.
284: In the presence of randomness, the $P-E$ curve
285: exhibits a stepwise behavior also
286: ,which is caused by crossing of one particle localized states.
287: Matrix elements of the Hamiltonian between arbitrary localized states
288: are exponentially small as a function of the distance.
289: Therefore energy repulsion between the highest occupied and the
290: lowest unoccupied states at both ends of the system is practically
291: negligible.
292: Then charge reconstruction of the ground state occurs
293: which corresponds to the electron tunneling between the localized states.
294: The inset of Fig.\ \ref{fig2b} is the charge distributions
295: at $E = E_{\rm s} \pm \Delta E$ where $E_{\rm s}$
296: is the field strength of the step.
297: The charge distribution near the center of the lattice is unchanged
298: when $E$ crosses the step, but that near the edges is modified.
299:
300: \subsection{Linear Response Regime}
301: \subsubsection{Susceptibility of non-interacting systems}
302: In this section we focus on the linear response region($E \ll t/L$).
303: We calculate the susceptibility $\chi$ as a function of $L$
304: by numerically differentiating the ground state energy.
305: At first, we consider the non-interacting case($V=0$).
306: In order to calculate $\chi$, we apply small electric field
307: $\pm \Delta E$. $L \Delta E \sim
308: 10^{-3}$.
309: We need to avoid the occurence of the step in $\pm \Delta E$.
310: When the step occurs near $E=0$ accidentally, we do not use the data.
311: The number of occurence of such step is typically $1 \sim 2$ in 100 samples
312: when $W$ is large($W \sim 5$).
313: The calculations are always carried out in the localized region except
314: $W=0$, then we expect $\chi \sim e^{-\xi /L}$ when the system size is
315: sufficiently large.
316: Therefore we take an average of $\log \chi $ as is shown in Fig.\ \ref{fig3}.
317: We averaged over 500 realizations of the disorder potentials.
318: For $W=0$, $\chi$ increases as $\chi \sim L^2$ and
319: $\chi$ seems to diverge. This implies the system is metallic. On the
320: other hand, for $W=3$, $\chi$ is convergent to a finite value
321: which is consistent with the insulating ground state.
322: For intermediate $W$,
323: saturation of $\chi$ is not clearly observed up to $L = 500$ and
324: only deviation from $\chi \sim L^2$ is observed.
325: This is because the localization
326: length $\xi$ is larger than the system length we used.
327:
328: Therefore we need a finite-size scaling analysis to determine localization
329: length $\xi$.
330: We perform the finite-size scaling analysis by assuming
331: \begin{equation}
332: \frac{\chi (L)}{L^2} = g\left( \frac{\xi(W)}{L} \right),
333: \end{equation}
334: where $\xi(W)$ is a localization length and $g(x)$ is a scaling
335: function.
336: If $L$ is much larger than $\xi$, $\chi$ becomes constant.
337: Therefore the scaling function $g$ behaves as $g(x)\sim x^2$ for $x \to 0$.
338: On the other hand, when $L$ is much smaller than $\xi$ $(L \ll \xi )$, the
339: wavefunction spread over entire system and
340: the system looks like metallic. Then,
341: $\chi \sim L^2$ is expected and
342: the scaling function $g$ behaves as $g(x) \sim const.$ for $x \to \infty$.
343: The scaling function $g$ we obtained for $V=0$ is shown in Fig. \ref{fig3}.
344: It shows the above-mentioned one parameter scaling hypothesis works quite well
345: in the model.
346:
347: \subsubsection{Metal-insulator transition in attractive $V$}
348: Using the random DMRG, we study the systems with attractive electron-electron
349: interaction.
350: In our implementation of the DMRG, three or four finite lattice sweeps are
351: performed to get the convergence of the ground state energy and
352: the number of retained
353: states for each block is 60-100 to keep the truncation error to be less
354: than $10^{-9}$.
355: The susceptibility $\chi$ is calculated for various values of $V$ and $W$.
356: The calculations are carried out up to the system size $L=100$.
357: We take an average over 128 realizations of the disorder potentials
358: for $V=-1.4$ and over 64 realizations of the disorder potentials for
359: $V=-0.5,-0.8,-1.0,-1.1,-1.2,-1.3,-1.5,-1.6$ and $-1.8$.
360: The steps in $P-E$ curve become smoother when interaction is introduced.
361: When $\log \chi$ is larger than $5$, we consider this as an influence of
362: the step and we do not use the data.
363: Then we perform the finite-size scaling analysis similarly to the
364: noninteracting case assuming the same one parameter scaling hypothesis.
365: In order to avoid finite size effect, we use the
366: data for $L \ge 20$ to determine the localization length $\xi$.
367: The obtained scaling function $g$ for $V=-1.4$ is shown in Fig.\ \ref{fig4}.
368: The scaling hypothesis for the interacting case also seems to work
369: well in the present model.
370: For $V = -1.8$ the $L$ dependence of $\chi$ deviates from $\chi \sim
371: L^2$ in $L <
372: 50$ even for $W = 0$.
373: We could only treat systems $V \ge -1.6$ due to the finite size effects.
374:
375: In Fig.\ \ref{fig5}, $W$ dependence of the localization length $\xi$ is
376: shown for various value of $V$.
377: Note that the localization length $\xi$ is normalized by $\xi (W=3)$.
378: For $V=0$, it seems that $\xi$ is divergent at $W=0$, but
379: for $V=-1.4$, $\xi$ increases rapidly with decreasing $W$ and
380: diverge around $W \sim 1$.
381:
382:
383: In order to determine the critical disorder strength $W_{\rm c}$,
384: we fit the localization length $\xi$ as
385: \begin{equation}
386: \xi(W) = (A+B(W-W_{\rm c}))(W-W_{\rm c})^\beta , \label{fit}
387: \end{equation}
388: where $A,B$ and $\beta$ are fitting parameters.
389: For $V=0$, we obtained $W_{\rm c}=0.02 \pm 0.09$ and $\beta = -2.1 \pm 0.3$,
390: which is consistent with expected value $W_{\rm c}=0$ and $\beta =-2$.
391:
392: In Fig. \ref{fig6}, the obtained phase diagram is shown.
393: The vertical lines represent errorbars.
394: This phase diagram is consistent with the one obtained
395: from the analysis of the phase sensitibity.\cite{rf:5}
396:
397:
398: \section{Summary}
399: In the present paper we have studied the dielectric response of the
400: one-dimensional spinless fermion model with interaction and disorder by
401: using the random DMRG.
402: At first, we have calculated the polarization of the Mott insulator
403: and the Anderson insulator.
404: In the $P-E$ curve we observed stepwise behaviors
405: both for the Mott insulators and the Anderson insulators.
406: From the change of the charge distribution,
407: we could understand the stepwise behaviors.
408: For the Mott insulator the steps come from
409: the generation of kink-pair and occur at $E \sim 1/L$.
410: On the other hand,
411: the steps for the Anderson insulator represent the crossing of the
412: energy levels
413: and the steps occur at $E\sim 1/L^2$.
414: From the zero field susceptibility,
415: we performed the finite-size scaling and determined the localization
416: length $\xi$.
417: Also we confirmed the existence of the metallic region
418: in attractive interacting regime.
419:
420: The computation in this work has been done in part using the
421: facilities of the Supercomputer Center, ISSP, University of Tokyo.
422: Y. H. was supported in part by a Grant-in-Aid from the Ministry of
423: Education, Science and Culture of Japan and also by the Kawasaki steel
424: 21st Century Foundation.
425:
426: \begin{thebibliography}{99}
427: \bibitem{rf:lee} P. A. Lee and T. V. Ramakrishnan:
428: Rev. Mod. Phys. {\bf 57} (1985) 287.
429: \bibitem{rf:1} S. R. White:
430: Phys. Rev. Lett. {\bf 69} (1992) 2863;
431: Phys. Rev. {\bf B 48} (1993) 10345.
432: \bibitem{rf:d} Density-Matrix Renormalization: A New Numerical Method
433: in Physics, edited by I. Peschel {\it et al}. (Springer-Verlag,
434: Berlin, 1999).
435: \bibitem{rf:3} K. Hida:
436: J. Phys. Soc. Jpn. {\bf 65} (1996) 895.
437: \bibitem{rf:c1} R. Resta and S. Sorella,
438: Phys. Rev. Lett. {\bf 82} (1992) 370.
439: \bibitem{rf:4} C. Aebischer, D. Baeriswyl and R. M. Noack:
440: Phys. Rev. Lett. {\bf 86} (2001) 468.
441: \bibitem{rf:ad1} S. Fujimoto, N. Kawakami:
442: Phys. Rev. {\bf B 54} (1996) R11018.
443: \bibitem{rf:ad2} Y. Otsuka and Y. Hatsugai:
444: J. Phys, Condens. Matter {\bf 12} (2000) 9317.
445: \bibitem{rf:8} S. T. Chui and J. W. Bray:
446: Phys. Rev. {\bf B 16} (1977) 1329.
447: \bibitem{rf:7} W. Apel:
448: J. Phys. C {\bf 15} (1982) 1973.
449: \bibitem{rf:9a} S. T. Chui and J. W. Bray:
450: Phys. Rev. {\bf B 19} (1979) 4020.
451: \bibitem{rf:9b} W. Apel and T. M. Rice:
452: Phys. Rev. {\bf B 26} (1982) 7063.
453: \bibitem{rf:9} Y. Suzumura and H. Fukuyama:
454: J. Phys. Soc. Jpn. {\bf 52} (1983) 2870.
455: \bibitem{rf:6} T. Giamarchi and H. J. Schulz:
456: Phys. Rev. {\bf B 37} (1988) 325.
457: \bibitem{rf:a} K. J. Runge and G. T. Zimanyi:
458: Phys. Rev. {\bf B 49} (1994) 15212.
459: \bibitem{rf:2} G. Bouzerar and D. Poilblanc:
460: J. Phys. I (France) {\bf 4} (1994) 1699.
461: \bibitem{rf:5} P. Schmitteckert {\it et al.}:
462: Phys. Rev. Lett. {\bf 80} (1998) 560.
463: \bibitem{rf:loc} M. Kishi and Y. Hatsugai:
464: J. Phys. Soc. Jpn. {\bf 72} (2003) Suppl. A pp.147.
465: \end{thebibliography}
466:
467: \begin{center}
468: \begin{figure}[hbp]
469: \hspace*{0cm}
470: \includegraphics[scale=0.6]{figure1.eps}
471: \caption{Polarization $P$ as a function of the applied electric field
472: $E$. (a) For the pure system($W = 0$).
473: The Luttinger liquid regime ($V = 1.0$) and
474: the Mott insulator regime ($V = 6.0$). The system size is $L = 30$. (b)
475: The charge distribution at each site,
476: solid line :$E=E_{\rm s} - \Delta E$, broken line :$E=E_{\rm s} + \Delta E$ where
477: $E_{\rm s}$ is the value of the electric field where the stepwise behavior is
478: observed.}
479: \label{fig2}
480: \end{figure}
481: \end{center}
482: \begin{center}
483: \begin{figure}
484: \includegraphics[scale=0.6]{figure2.eps}
485: \caption{The polarization $P$ as a function of the applied electric field
486: $E$, non-interacting system($V = 0$,$W=5$,$L=20$). Inset: the charge
487: distribution at $E=E_{\rm s} + \Delta E$(broken line) and $E=E_{\rm
488: s} -
489: \Delta E$(solid line).
490: }
491: \label{fig2b}
492: \end{figure}
493: \end{center}
494: \begin{center}
495: \begin{figure}
496: \includegraphics[width=0.45\linewidth]{figure3a.eps}
497: \includegraphics[width=0.45\linewidth]{figure3b.eps}
498: \caption{Left: Average $\log\chi$ as a function of $\log1/L$.
499: The system sizes are between 10 and 500. $V=0$. $W =$ 0, 0.5,
500: 1.5, 3.0 and 5.0. We averaged over 500 realizations of the disorder potentials.
501: Right: The scaling function $g$ for $V=0$ is shown. }
502: \label{fig3}
503: \end{figure}
504: \end{center}
505: \begin{center}
506: \begin{figure}
507: \includegraphics[scale=0.6]{figure4.eps}
508: \caption{Scaling plots of $\chi / L^2$ vs $ \xi / L$ in a log-log
509: scale.
510: $V = -1.4$. }
511: \label{fig4}
512: \end{figure}
513: \end{center}
514: \begin{center}
515: \begin{figure}
516: \includegraphics[scale=0.6]{figure5.eps}
517: \caption{$\xi$ vs $W$. $V=0$ and $-1.4$. $\xi$ is scaled by the value at $W=3$.
518: The dashed lines show fitting curves by eq (\ref{fit}).
519: }
520: \label{fig5}
521: \end{figure}
522: \end{center}
523: \begin{center}
524: \begin{figure}
525: \includegraphics[scale=0.6]{figure6.eps}
526: \caption{Phase diagram of the half-filled spinless fermion
527: model. The vertical lines are errorbars.}
528: \label{fig6}
529: \end{figure}
530: \end{center}
531: \end{document}
532:
533:
534:
535:
536:
537:
538:
539:
540:
541:
542:
543:
544:
545:
546:
547:
548:
549:
550:
551:
552:
553:
554:
555:
556:
557: