cond-mat0302467/nqw.tex
1: % \documentclass[aps,twocolumn,amsmath,amssymb]{revtex4}
2: % \documentclass[twocolumn,showpacs,preprintnumbers,amsmath,amssymb]{revtex4}
3: \documentclass[preprint,aps,amsmath,amssymb,showpacs]{revtex4}
4: 
5: % Some other (several out of many) possibilities
6: %\documentclass[preprint,aps]{revtex4}
7: %\documentclass[preprint,aps,draft]{revtex4}
8: %\documentclass[prb]{revtex4}% Physical Review B
9: 
10: \bibliographystyle{apsrev}
11: 
12: \usepackage{graphicx}% Include figure files
13: \usepackage{dcolumn}% Align table columns on decimal point
14: \usepackage{bm}% bold math
15: 
16: \begin{document}
17: 
18: \title{Vertical Confinement and Evolution of Reentrant Insulating 
19: Transition in the Fractional Quantum Hall Regime}
20: % \title{Reentrant Insulating Phase near $\nu = 1/3$ Fractional Quantum 
21: % Hall Effect in a Vertically Confined Two-Dimensional Electron System}
22: 
23: \author{I. Yang$^{1}$, W. Kang$^{1}$, S.T. Hannahs$^{2}$, L.N. Pfeiffer$^{3}$, 
24: and K.W. West$^{3}$}
25: 
26: \affiliation{
27: $^{1}$James Franck Institute and Department of Physics,
28:  University of Chicago, Chicago, Illinois 60637\\
29:  $^{2}$National High Magnetic Field Laboratory, 1800 East Paul 
30: Dirac Drive, Tallahassee, Florida 32310\\
31: $^{3}$Bell Laboratories, Lucent Technologies, 600 Mountain
32: Avenue, Murray Hill, New Jersey 07974}
33: 
34: \date{\today}
35: 
36: \begin{abstract}
37: We have observed an anomalous shift of the high field reentrant insulating 
38: phases in a two-dimensional electron system (2DES) tightly confined 
39: within a narrow GaAs/AlGaAs quantum well. Instead of the well-known
40: transitions into the high field insulating states centered around 
41: $\nu = 1/5$, the 2DES confined within an 80\AA-wide quantum well exhibits 
42: the transition at $\nu = 1/3$. Comparably large quantum lifetime of the 2DES 
43: in narrow well discounts the effect of disorder and points to 
44: confinement as the primary driving force behind the evolution of the 
45: reentrant transition.
46: \end{abstract}
47: 
48: \pacs{73.43.-f}
49: 
50: \maketitle
51: 
52: The prospect for a quantum Wigner crystal has driven the study of
53: two-dimensional electron system (2DES) under the conditions of 
54: low temperature and high magnetic field\cite{WignerCrystal}.  
55: Experiments on high mobility GaAs/AlGaAs heterostructures have 
56: shown that in the limit of zero temperature the $\nu = p/(2p \pm 1), 
57: p = 1, 2, 3\ldots$ series of incompressible quantum liquid states of 
58: fractional quantum Hall effect (FQHE) terminates with a transition 
59: into a field-induced insulator at low fillings. A dramatic sequence 
60: of transitions involving a reentrant insulating phase, a $\nu = 1/m$ 
61: primary FQHE state, and the final insulating phase can be realized 
62: in the highest quality samples.  
63: Initially observed in 2DES around the $\nu = 1/5$ FQHE\cite{Jiang90}, the 
64: reentrant insulating phase has been also detected adjacent to the 
65: $\nu = 1/3$ FQHE state in two-dimensional hole system (2DHS)\cite{Santos92}. 
66: Reentrant insulating behavior is also seen in high quality
67: silicon MOSFET and $p$-SiGe heterostructures 
68: in the integer quantum Hall regime\cite{D'Iorio92,Kravchenko95,Sakr01}.
69: Because of the correlated nature of the FQHE, the insulating states 
70: above and below the FQHE states at $\nu = 1/3$ and 1/5 are expected 
71: to be driven by electron-electron correlation rather than disorder, 
72: enhancing the likelihood of formation of a Wigner crystal. This has 
73: given impetus for various electrical, acoustic, microwave, and optical
74: investigations of the high field insulating 
75: phases\cite{Goldman90,Li91,Williams91,Buhmann91,Paalanen92,Goldys92,Manoharan94,Li97,Pan02}. 
76: 
77: For an ideal, disorder-free 2DES, the ground state in the limit of zero 
78: temperature is an ordered electron crystal at small Landau level fillings. 
79: However, a positive confirmation of the Wigner 
80: crystalline order has remained controversial and somewhat elusive. This 
81: is partly due to absence of scattering experiments that can directly probe 
82: the crystalline order of the insulating phases. In addition, 
83: interpretation of various experiments in the Wigner crystalline regime 
84: is complicated by general lack of understanding of the effects of 
85: disorder in presence of strong interaction. 
86: Even the highest quality samples presently available possess 
87: non-negligible disorder at low fillings, and the competition of 
88: disorder and interaction is thought to modify the ground state 
89: of 2D systems in some fundamental way. At short distances, the 
90: ground state is conjectured to evolve into a partially ordered Wigner 
91: crystal consisting of finite-size domains that are pinned by the disorder 
92: potential\cite{WignerCrystal}. In presence of strong disorder, the ground 
93: state evolves into a disorder-driven correlated insulator called 
94: Hall insulator\cite{Kivelson92,Shahar95,Wong96}.
95: 
96: In this paper we present an unexpected observation of reentrant 
97: insulating phase around the $\nu = 1/3$ FQHE state in a two-dimensional 
98: {\it electron} system. The 2DES in question is found in 
99: a GaAs/AlGaAs quantum well whose narrow width produces a tight 
100: vertical confinement of the electronic wave function. 
101: In spite of its relatively low mobility, we observe a clear sequence 
102: transitions to an insulator, a FQHE, and back to an insulating phase 
103: in the vicinity of $\nu = 1/3$ filling in a fashion reminiscent of the 
104: Wigner crystalline regime in high mobility $n-$ and $p$-type 
105: GaAs/AlGaAs heterostructure. 
106: Tilted magnetic field study shows that the insulating phases are found 
107: to be insensitive to the presence of parallel magnetic field. We also
108: compare the properties of the 2DES in the NQW with a wider quantum well 
109: and the conventional heterostructure. We find that the single particle 
110: lifetime of the 2DES in NQW is comparable to other higher mobility specimen 
111: and conclude that confinement within the NQW may be responsible for the 
112: shift in the reentrant transition.
113: 
114: The experiment was performed using a modulation doped AlGaAs/GaAs 
115: quantum well of 80-\AA\ in width. The density of the sample was $n = 
116: 1.1 \times 10^{11} \rm{cm}^{-2}$ with a low temperature mobility of $2.56 
117: \times 10^{5} \rm{cm}^{2}/\rm{V~sec}.$  The transport in narrow 
118: quantum well (NQW) is dominated by interfacial fluctuations\cite{Motohisa92}, 
119: yielding 
120: a substantially lower mobility than comparable heterostructure samples.
121: Samples in Hall bar and van der Pauw configurations were studied 
122: inside a dilution refrigerator with a 14 tesla superconducting magnet.
123: Tilted-field study was performed up to 40 tesla using the 
124: hybrid magnet at the National High Magnetic Field Laboratory. 
125: A light emitting diode
126: was used to illuminate the sample at low temperatures. 
127: Depending on the illumination conditions 
128: and the thermal cycling history, small variation in the data was detected.
129: This, however, does not alter our conclusions.
130: 
131: 
132: Fig. \ref{fig:narrow1} illustrates the magnetoresistance $R_{xx}$ of 
133: a NQW sample at a
134: temperature of 35 mK. The most striking feature of the data is the 
135: sharp increase of longitudinal resistance between $\frac{1}{3} < \nu < 0.45$. 
136: Slightly after $\nu = 1/2$ $R_{xx}$ increases dramatically by more 
137: than 2 orders of magnitude, completely overwhelming other transport 
138: features. The peak resistance at $\nu = 0.38$ exceeds $900 \rm{k}\Omega$ 
139: before dropping precipitously as it enters the $\nu = 1/3$ FQHE 
140: state. $R_{xx}$ subsequently diverges upon entering a high field
141: insulating phase. The inset of Fig. \ref{fig:narrow1} illustrates
142: longitudinal and Hall resistances of the same sample under a 
143: different illumination condition. The $R_{xy}$ is quantized at 
144: $3h/e^{2}$ at $\nu = 1/3$, demonstrating the formation of the 
145: $\nu = 1/3$ FQHE state. The $\nu = 3/7$ 
146: FQHE state is found as a weak $R_{xx}$ minimum prior to the 
147: reentrant insulating phase.  
148: 
149: Fig. \ref{fig:narrow2} shows the temperature dependence of the 
150: insulating state at a 
151: slightly higher density ($n = 1.2 \times 10^{11} \rm{cm}^{-2}$). As 
152: temperature is raised, $R_{xx}$ decreases sharply with the insulating 
153: features largely disappearing above T = 300 mK.
154: The inset of Fig. \ref{fig:narrow2} 
155: shows an Arrhenius plot of peak resistance at $\nu = 
156: 0.38$. The resistance at the peak is activated with an activation 
157: energy of $E_{g} \sim 0.26 $K. 
158: At lower temperatures, there is a
159: saturation of the resistivity. Measurement of I-V characteristics 
160: shows that the transport in the insulating regime is highly 
161: nonlinear, similar to previously observed
162: reentrant insulating 
163: phases\cite{Jiang90,Santos92,Goldman90,Li91,Williams91,D'Iorio92,Kravchenko95,Sakr01}. 
164: 
165: Fig. \ref{fig:narrow3} shows the effect of parallel magnetic field 
166: in the insulating regime above and below the $\nu = 1/3$ FQHE state.
167: Comparison of magnetoresistance for the tilt angle of $\theta = 
168: 0^{\circ}$ and $\theta = 58^{\circ}$ at T = 50 mK shows that there 
169: is no appreciable change in both $\nu = 1/3$ FQHE state and the 
170: insulating states above and below $\nu = 1/3$ even though the total 
171: magnetic field was nearly double of the perpendicular magnetic field. 
172: This shows that neither the increase in the Zeeman energy nor the 
173: deformation of the wave function due to strong parallel magnetic field
174: appears to play a large role in the insulating phase.
175: 
176: 
177: 
178: 
179: In Fig. \ref{fig:narrow4} we explore the role of vertical confinement 
180: by comparing the transport between quantum well samples with different 
181: widths. Fig. \ref{fig:narrow4}a illustrates magnetoresistance of our
182: 80\AA\ wide NQW. Fig. \ref{fig:narrow4}b illustrates the 
183: magnetoresistance of a quantum well sample that is 300\AA\ wide and 
184: possessing a mobility of $\mu = 7.8 \times 10^{6} cm^{2}/Vs$ and 
185: a density of $n = 6.2 \times 10^{10} cm^{-2}.$ In the wider quantum 
186: well specimen, magnetoresistance at 30 mK shows 
187: a well-developed sequence of FQHE states centered around  $\nu = 1/2$ 
188: and $\nu = 1/4$ followed by the reentrant insulating phase prior to the
189: $\nu = 1/5$ FQHE state. This contrasts sharply with the 80 \AA\ -wide 
190: NQW which only exhibits a weakly developed FQHE state at $\nu = 2/3$ 
191: prior to the reentrant insulating state above $\nu = 1/3$. These 
192: results suggest strongly that narrow confinement is likely to be
193: important in altering the properties of 2DES in the insulating regime.
194: 
195: The prevailing view on the high field reentrant transitions in 
196: various 2DES involves 
197: either an entry into the Wigner crystalline regime\cite{WignerCrystal} or 
198: an approach based on the global phase diagram of quantum Hall 
199: effect\cite{Kivelson92}. Consideration of the former scenario follows 
200: from the importance of the electron-electron interaction in the limit 
201: of low fillings. Since the Laughlin states at $\nu = 1/m, m = 3, 5...$ 
202: occur from strong 
203: electronic correlation\cite{Laughlin83}, it follows that the interaction 
204: should also play an important role in the adjacent insulating states. 
205: The reentrance is explained in terms of competition between the FQHE 
206: liquid and Wigner solid. In this point of view, the insulating phases 
207: seen in the NQW is likely to 
208: be some kind of strongly correlated ground state driven by interaction.
209: Alternatively, the insulating phases in NQW may occur due to some electron
210: localization effect associated with disorder as suggested by its modest 
211: mobility. In this context, the global phase diagram picture of quantum 
212: Hall transitions\cite{Kivelson92} is relevant as a transition
213: from an insulator into a FQHE state at $\nu = 1/3$ is permitted.
214: 
215: However, in either pictures, there is no obvious explanation for 
216: the shift in the reentrant insulating transition to $\nu = 1/3$ for 
217: the 2DES in the NQW. This feature is particularly puzzling since 
218: the reentrant behavior is always found near $\nu = 1/5$ in 2DES based 
219: on GaAs/AlGaAs structures\cite{WignerCrystal,Jiang90}. In fact, the observed sequence 
220: of transitions in NQW resembles the reentrant insulating transitions near 
221: $\nu = 1/3$ in 2DHS in GaAs/AlGaAs heterostructure\cite{Santos92}. 
222: In the reentrant insulating transitions seen in Si MOSFET and p-SiGe 
223: heterostructure, the reentrant insulating phases are found in the integer 
224: quantum Hall regime and spin is thought to play an important 
225: role\cite{D'Iorio92,Kravchenko95,Sakr01}. In our NQW, tilted field experiment
226: appears to rule out the role of spin in the reentrant insulating phase. 
227: 
228: Theoretically the transition into a Wigner crystal in 2DES 
229: is predicted for $\nu \leq 1/6.5$\cite{Lam84,Levesque84,Zhu95}. 
230: In the case of 2DHS in GaAs/AlGaAs heterostructures, the stability of the
231: insulating phases near $\nu = 1/3$ is explained in terms of increased Landau 
232: level mixing associated with its heavier effective mass 
233: ($m^{*}_{h} = 0.3m_{\circ}$) compared to that of electrons 
234: ($m^{*}_{e} = 0.067m_{\circ}$)\cite{Santos92,Zhu93,Price93}. 
235: Since the effective mass of electrons in narrow quantum wells has
236: been shown to be comparable to heterostructures\cite{Huant92}, 
237: Landau level mixing does not appear to play a significant role in NQW. 
238: As there is no universally accepted explanation for the reentrant transitions 
239: observed in various two-dimensional semiconductor systems, understanding 
240: of the shift in the reentrant behavior in the NQW is likely to be important 
241: in clarifying the nature of the associated insulating phases.
242: 
243: The role of disorder remains an important 
244: question as the mobility of 2DES in NQW is modest compared to other 
245: 2DES based on GaAs/AlGaAs quantum structures. 
246: However, previous experiments on GaAs/AlGaAs heterostructures with mobilities 
247: comparable to that of our NQW have not found reentrant behavior next to 
248: the  $\nu = 1/3$ FQHE state\cite{Shahar95,Wong96}. While this 
249: appears to discount the importance of disorder associated with the 
250: reentrant behavior in the NQW, interfacial roughness serves to 
251: restrict the transport in narrow quantum wells\cite{Motohisa92} and 
252: a more quantitative measure of disorder is necessary. This is 
253: particularly important since electronic transport under magnetic 
254: field is largely determined by large angle scattering instead of the small 
255: angle scattering which dominates the zero field transport. The single 
256: particle relaxation time, $\tau_{s}$, in semiconductors is consequently
257: substantially smaller than the transport scattering time, 
258: $\tau_{t}$\cite{DasSarma85,Coleridge91}. 
259: 
260: In Table \ref{tab:table1}  we summarize the 
261: properties of 2DES derived from 3 different GaAs/AlGaAs structures that 
262: exhibit reentrant insulating phase in the lowest Landau level. In addition 
263: to the 80\AA\ quantum well, 2DES from a heterostructure and 300\AA\ quantum 
264: well with mobilities that are respectively 10 and 30 times larger were 
265: compared. $\tau_{t}$ was deduced from zero field mobility, $\mu = 
266: e\tau_{t}/m^{*}$, and $\tau_{s}$ was determined from the Dingle analysis 
267: of the Shubnikov-de Haas (SDH) oscillations as suggested by 
268: Coleridge\cite{Coleridge91}. In contrast to the large differences 
269: in $\tau_{t}$, we found that $\tau_{s}$'s determined from different structures
270: were surprisingly close. The $\tau_{s}$ in  the high mobility 
271: 300 \AA\ wide quantum well was 8.5 $ps$ with the 80 \AA\ NQW yielding
272: a $\tau_{s}$ of 3.7 $ps$. The heterostructure sample was
273:  found to possess a $\tau_{s}$ comparable to other 2DES samples.
274:  
275: The proximity of $\tau_{s}$'s is also reflected in the onset of SDH oscillations, $B_{onset}$. In 
276: the insets of Fig. \ref{fig:narrow4} we show the SDH oscillations for 
277: 80 \AA\ NQW and 300 \AA\ quantum well. A $B_{onset}$ of $\sim$ 60 mT 
278: for the NQW and $\sim 30$mT for other samples were obtained. The
279: comparable $\tau_{s}$ and $B_{onset}$ suggest that 
280: in spite of its low mobility, the electronic lifetime in NQW is not 
281: adversely affected by the tight confinement, increasing the likelihood
282: that the reentrant behavior around $\nu = 1/3$ is driven by 
283: interaction rather than disorder. On the other hand, the larger 
284: resistivity at $\nu = 1/2$ and the absence of high order FQHE states 
285: is consistent with presence of stronger disorder in the NQW than the 
286: wider well. However, coexistence of the $\nu = 1/3$ FQHE state next 
287: to the insulating phases indicates that disorder is not enough to 
288: suppress the electron correlation in the limit of low fillings.
289: 
290: Since the most distinguishing characteristic of the 2DES in a NQW involves its 
291: vertical confinement, the physics of reentrance may potentially occur from 
292: a confinement-induced evolution of the Coulomb interaction. The ground state 
293: energy and the interaction parameters of 2DES depends on the finite vertical 
294: extent of the electronic wave function\cite{Zhang86,Morf02}. A thicker 2DES 
295: consequently experiences a softer 
296: Coulomb potential and thereby possesses a reduced FQHE energy gap compared to a 
297: thinner 2DES. Our estimate of the thickness of electrons based on the interaction
298: parameters in quantum wells\cite{DasSarma85a}
299:  points to a substantial reduction in its thickness
300: compared to the 2DES found in heterostructures\cite{Morf02}. For the 2DES in
301: Table \ref{tab:table1}, we obtain a thickness of 128\AA\ in the heterostructure specimen
302: which contrasts sharply against the estimated thickness of 19\AA\ and 67\AA\ for 
303: the 80\AA\ and 300\AA\ quantum wells. The role of thickness in the insulating 
304: phases in 2DES remains unknown and further theoretical investigation is 
305: necessary to clarify the effect of confinement in relation to the enhancement 
306: of $r_{s}$ in the insulating regime.
307: 
308: In summary, we have observed a puzzling shift in the reentrant insulating 
309: phases in a 2DES confined within a NQW. The apparent shift in the reentrant 
310: transition to $\nu = 1/3$ cannot be reconciled in terms of disorder and points
311: to importance of confinement. Understanding the shift in the reentrant transition 
312: may be important in uncovering the physics behind the reentrant insulating
313: phases in the NQW as well as other two-dimensional electron and hole systems.
314: While the confinement within a NQW is expected to produce a thinner 2DES, its 
315: effect in the high field insulating states remains to be clarified.
316: 
317: We would like to thank H. Fertig, H.W. Jiang, A.H. MacDonald, H.L. 
318: Stormer, and D.C. Tsui for useful 
319: discussions. A portion of this work was performed at the National 
320: High Magnetic Field Laboratory which 
321: is supported by NSF Cooperative Agreement No. DMR-9527035.
322: The work at the University of Chicago is supported by NSF 
323: DMR-9808595 and DMR-0203679.
324: 
325: \pagebreak
326: 
327: \begin{thebibliography}{99}
328: \bibitem{WignerCrystal} See for example, review articles by H.A. Fertig and 
329:  M. Shayegan, in {\it Perspectives in Quantum Hall Effects,} edited by S. 
330:  Das Sarma and A. Pinczuk (Wiley, New York, 1996), and references therein. 
331: \bibitem{Jiang90} H.W. Jiang {\it et al.}, Phys. Rev. Lett. {\bf 65}, 633 (1990). 
332: \bibitem{Santos92} M.B. Santos {\it et al.}, Phys. Rev. Lett. {\bf  68}, 1188 (1992). 
333: \bibitem{D'Iorio92} M.D. D'Iorio {\it et al.}, Phys. Rev. B. {\bf 46}, 15992 (1992).
334: \bibitem{Kravchenko95} S. Kravchenko {\it et al.}, Phys. Rev. Lett. {\bf 75}, 910 (1995).
335: \bibitem{Sakr01} M.R. Sakr {\it et al.}, Phys. Rev. B {\bf 64}, 161308R (2001).
336: \bibitem{Goldman90} V.J. Goldman {\it et al.}, Phys. Rev. Lett. {\bf 65} 2189 (1990). 
337: \bibitem{Li91} Y.P. Li {\it et al.}, Phys. Rev. Lett. {\bf 67}, 1630 (1991).  
338: \bibitem{Williams91} F.I.B. Williams {\it et al.}, Phys. Rev. Lett. {\bf 66}, 3285 (1991). 
339: \bibitem{Buhmann91} H. Buhmann {\it et al.}, Phys. Rev. Lett. {\bf 66}, 926 (1991). 
340: \bibitem{Paalanen92} M.A. Paalanen {\it et al.}, Phys. Rev. B {\bf 45}, 11342 (1992). 
341: \bibitem{Goldys92} E.M. Goldys {\it et al.}, Phys. Rev.  B {\bf 46}, 7957 (1992). 
342: \bibitem{Manoharan94} H.C. Manoharan {\it et al}, Phys. Rev. B {\bf 50}, 17662 (1994).
343: \bibitem{Li97} C.C. Li {\it et al.}, Phys. Rev. Lett. {\bf 79}, 1353 (1997).  
344: \bibitem{Pan02} W. Pan {\it et al.}, Phys. Rev. Lett. {\bf 88}, 176802 (2002). 
345: \bibitem{Kivelson92} S. Kivelson {\it et al}, Phys. Rev. B {\bf 46}, 2223 (1992). 
346: \bibitem{Shahar95} D. Shahar {\it et al.}, Phys. Rev. Lett. {\bf 74}, 4511 (1995).  
347: \bibitem{Wong96} L.W. Wong {\it et al}, Phys. Rev. B {\bf 54}, 17323 (1996). 
348: \bibitem{Motohisa92} J. Motohisa {\it et al}, Appl. Phys. Lett. {\bf 60}, 1315 (1992).
349: \bibitem{Lam84} P.K. Lam and S.M. Girvin, Phys. Rev. B {\bf 30}, 473 (1984). 
350: \bibitem{Levesque84} D. Levesque {\it et al}, Phys. Rev. 
351:  B {\bf 30}, 1056 (1984). 
352: \bibitem{Zhu95} X. Zhu and S.G. Louie, Phys. Rev. B {\bf 52}, 5863 (1995). 
353: \bibitem{Zhu93} X. Zhu and S.G. Louie, Phys. Rev. Lett. {\bf 70}, 335 (1993). 
354: \bibitem{Price93} R. Price {\it et al}, Phys. Rev. Lett. 
355:  {\bf 70}, 339 (1993). 
356: \bibitem{Huant92} S. Huant {\it et al}, Phys. Rev. B {\bf 46}, 2613 (1989). 
357: \bibitem{Laughlin83} R.B. Laughlin, Phys. Rev. Lett {\bf 50}, 
358: 1395 (1983).
359: \bibitem{DasSarma85} S. Das Sarma and F. Stern, Phys. Rev. B {\bf 32}, 8442 (1985).
360: \bibitem{Coleridge91} P.T. Coleridge, Phys. Rev. B {\bf 44}, 3793 (1991).
361: \bibitem{Zhang86} F.C. Zhang  {\it et al}, Phys. Rev. B {\bf 33}, 2903 (1986).
362: \bibitem{Morf02} R.H. Morf {\it et al}, Phys. Rev. B {\bf 66}, 075408 (2002).
363: \bibitem{Ortalano97} M.W. Ortalano {\it et al}, Phys. Rev. B {\bf 55}, 7702 (1997).
364: \bibitem{DasSarma85a} S. Das Sarma {\it et al}, Ann. Phys. (N.Y.) {\bf 163}, 78 (1985).
365: \end{thebibliography}
366: 
367: \pagebreak
368: 
369: \begin{table*}
370: \caption{\label{tab:table1} Comparison of the 
371: properties of two-dimensional electron systems in 80\AA\ GaAs/AlGaAs 
372: quantum well, 300\AA\ GaAs/AlGaAs quantum well, and a GaAs/AlGaAs
373: heterostructures. All exhibit reentrant behavior in 
374: the lowest Landau level.}
375: \begin{ruledtabular}
376: \begin{tabular}{cccccccccc}
377: % \begin{tabular}{c.cc.c} \hline\hline
378: sample & density & mobility, $\mu$ & scattering 
379: & onset of & single particle & $\tau_{t}/\tau_{s}$ & $\rho_{xx}$ 
380: at $\nu = 1/2$  & reentrance \\
381:   & ($10^{11} {cm}^{-2}$) & ($\rm{cm}^{2}/\rm{Vs}$) & time, $\tau_{t}$$(ps)$ & 
382:   SDH (mT) & lifetime, $\tau_{s}$$(ps)$ & & ($k\Omega/\Box$) & \\
383: \hline
384: % \vspace*{1mm}
385: %     &&&&& 
386: 80\AA\ QW & 1.1 & $2.5 \times 10^{5}$ & 9.7 & 61 & 3.6 & 2.7 &  15.3 & $\nu = 1/3$ \\  
387: % heterostructure & 1.1 & $6.5 \times 10^{6}$ & 248 & 46 & 5.5 & 45 & $\nu = 1/5$ \\
388: 300\AA\ QW & 0.62 & $7.8 \times 10^{6}$ & 297 & 29 & 8.5 & 35 & 0.241 & $\nu = 1/5$ \\
389: heterostructure & 0.53 & $2.5 \times 10^{6}$ & 95 & 30 & 6.7 & 
390: 14 & 0.741 & $\nu = 1/5$ \\
391: % \hline\hline
392: 
393: \end{tabular}
394: \end{ruledtabular}
395: \end{table*}
396: 
397: \vspace*{4in}
398: \pagebreak
399: 
400: \begin{figure}
401: % \vspace*{1.5in}
402: %\includegraphics[width=3.25in]{nqwfig1.eps}
403: \caption{\label{fig:narrow1} 
404: Longitudinal magnetoresistance of a narrow quantum well at 35 mK. 
405: Inset: longitudinal (blue) and transverse (red) magnetoresistivities of a sample 
406: with a slightly higher density. Integers and fractions indicate the Landau level 
407: filling of the two-dimensional electron system.}
408: \end{figure}
409: 
410: \begin{figure}
411: % \vspace*{1.25in}
412: %\includegraphics[width=3.25in]{nqwfig2.eps}
413: \caption{\label{fig:narrow2} Longitudinal magnetoresistance of a narrow 
414: quantum well at 32, 46, 60, 80, 100, 140, 210, 300 and 420 mK. Inset: 
415: Arrhenius plot of the 
416: resistivity at the peak of the reentrant insulating phase.}
417: \end{figure}
418: 
419: \begin{figure}
420: % \vspace*{1.25in}
421: %\includegraphics[width=3.25in]{nqwfig3.eps}
422: \caption{\label{fig:narrow3} Magnetoresistance vs perpendicular 
423: magnetic field for tilt angles at $\theta = 0^{\circ}$ (solid) and 
424: $\theta = 58^{\circ}$ (dashed). The measurement was performed at 
425: 50 mK.}
426: \end{figure}
427: 
428: \begin{figure}
429: % \vspace*{1.25in}
430: %\includegraphics[width=3.25in]{nqwfig4.eps}
431: \caption{\label{fig:narrow4} Comparison of reentrant insulating 
432: transitions in a  80 \AA\ (top) and 300 \AA\ (bottom) wide quantum well with 
433: mobility of $2.5\times 10^5 cm^{2}/Vs$ and $7.8\times 10^{6}cm^{2}/Vs$, 
434: respectively, at T $\approx 30mK$.  
435: Insets: expanded view of the low field Shubnikov-de Haas (SDH)
436: oscillations. While the mobilities differ by a factor of $\sim$30, 
437: the onsets of SDH oscillations differ by a factor of $\sim$2.}
438: \end{figure}
439: 
440: \end{document}
441: