1: %
2: \documentclass[prl,aps,twocolumn,showpacs]{revtex4}
3: % \documentclass[prl,preprint,aps]{revtex4}
4:
5: \usepackage[dvips]{graphicx}
6: \usepackage{dcolumn}
7:
8: \newcommand{\be}{\begin{equation}}
9: \newcommand{\ee}{\end{equation}}
10: \newcommand{\bea}{\begin{eqnarray}}
11: \newcommand{\eea}{\end{eqnarray}}
12: \newcommand{\nn}{ \nonumber}
13: \newcommand{\ds}{\displaystyle}
14: \thispagestyle{empty}
15: \mathsurround=2pt
16: \topmargin=-20mm
17: \begin{document}
18: %\Large
19:
20:
21:
22:
23:
24: \title{ Fermi-liquid effects in the transresistivity in quantum Hall double layers near $\nu= 1/2 $}
25:
26:
27:
28: \author{Natalya A. Zimbovskaya }
29: %$^{a,b}$ and Steven H. Simon$^b$}
30:
31:
32: \affiliation{Department of Physics and Astronomy, St. Cloud State
33: University, 720 Fourth Avenue South, St. Cloud, MN 56301, USA;\\
34: Urals State Mining University, Kuibysheva Str. 30, Yekaterinburg, Russia, 620000 }
35: %\\$^b$Lucent Technologies Bell Labs, Murray Hill, New Jersey 07974}
36:
37: \date{\today}
38:
39:
40: \begin{abstract}
41: Here, we present theoretical studies of the temperature and magnetic field dependences of the Coulomb drag transresistivity between two parallel layers of two dimensional electron gases in quantum Hall regime near half filling of the lowest Landau level. It is shown that Fermi-liquid interactions between the relevant quasiparticles could give a significant effect on the transresistivity, providing its independence of the interlayer spacing for spacings taking on values reported in the experiments. Obtained results agree with the experimental evidence.
42: \end{abstract}
43: \vspace{1mm}
44:
45: \pacs {71.27.+a 73.43.-f}
46:
47:
48:
49: \maketitle
50:
51:
52:
53: %\begin{multicols}{2}
54: %\narrowtext
55:
56:
57:
58: During the last decade double-layer two-dimensional electron gas (2DEG)
59: systems were of significant interest due to many remarkable phenomena they
60: exhibit, including so called Coulomb drag. In Coulomb drag experiments two
61: 2DEGs are arranged close to each other, so that they can interact via
62: Coulomb forces. A current $ I $ is applied to one layer of the system, and
63: the voltage $ V_D $ in the other nearby layer is measured, with no current
64: allowed to flow in that layer. The ratio $- V_D/I $ gives a
65: transresistivity $ \rho_D $ which characterizes the strength of the
66: effect. The physical interpretation of the Coulomb drag is that momentum
67: is tranferred from the current carrying layer to the nearby one due to
68: interlayer interactions \cite{one,two,three}.
69:
70:
71: It was shown theoretically \cite{four,five} and confirmed with experiments
72: \cite{five} that the transresistivity between two 2DEGs in quantum Hall
73: regime at one half filling of the lowest Landau level for both layers is
74: proportional to $ T^{4/3} \ (T $ is the temperature of the system) which
75: is quite different from the temperature dependence of $ \rho_D $ in the
76: absence of the external magnetic field applied to 2DEGs. This
77: temperature dependence of the drag at $ \nu
78: = 1/2 $ originates from the ballistic contribution to the
79: transresistivity. The latter reflects the response of the two-layer system
80: to the driving disturbance of finite wave vector $ \bf q $ and finite
81: frequency $\omega $ when considering scales are smaller than the mean free
82: path $ l $ of electrons $ (ql \gg 1) $, and times are shorter than their
83: scattering time $ \tau \ (\omega \tau \gg 1)$ \cite{six}.
84:
85: In further experiments \cite{seven} the Coulomb drag was measured between
86: 2DEGs where the layer filling factor was varied around $ \nu = 1/2. $ The
87: transresistivity was reported to be enhanced quadratically with $ \Delta\nu = \nu - 1/2. $ It was also reported that the curvature of the enhancement depended on temperature but it was insensitive to both sign of $ \Delta \nu $ and distance $ d $ between the layers. The present work is motivated with these experiments
88: of \cite{seven}. We calculate the transresistivity between two layers of 2DEGs subject to a strong magnetic field which provides $ \nu $ close to $ 1/2 $ for both layers.
89:
90: We start from the well-known expression \cite{one,three} which relates the Coulomb drag transresistivity to density-density components of the polarization in the layers $ \Pi_{(1)} (\bf q,\omega) $ and $ \Pi_{(2)} (\bf q,\omega) :$
91: \bea %f1
92: \rho_D &=& \frac{1}{2(2\pi)^2} \frac{h}{e^2} \frac{1}{Tn^2} \int
93: \frac{q^2 d \bf q}{(2\pi)^2} \int \frac{\hbar d \omega}{\sinh^2 (\hbar\omega/2T)}
94: \nn \\ \nn\\ &&
95: \times \big|U ({\bf q}, \omega)\big|^2 {\mbox {Im}} \Pi_{(1)}({\bf
96: q,}\omega) {\mbox {Im}} \Pi_{(2)}({\bf q,}\omega).
97: \eea
98: Here, $ U ({\bf q}, \omega)$ is the screened interlayer Coulomb
99: interaction, and electron densities in the layers are supposed to be equal$ (n_1 = n_2 = n) .$
100:
101: %+1
102: Within the usual Composite Fermion (CF) approach \cite{eight} a single layer polarizability describes that part of the density-current electromagnetic response which is irreducible with respect to the Coulomb interaction. Adopting for simplicity the RPA, we obtain the following expression for the $ 2 \times 2 $ polarizability matrix:
103: \be % f2
104: \Pi^{-1} = (K^0)^{-1} + C^{-1} .
105: \ee
106: Here, the matrix $ K^0 $ gives the response of noninteracting CFs and C is the Chern-Simons interaction matrix. Assuming for certainty the wave vector $ \bf q $ to lie in the $ "x"$ direction we have:
107: \be %f3
108: %\begin{eqnarray}
109: \mbox{ C\/} \mbox{= }%&
110: \left(
111: \begin{array}{cc}
112: \mbox{\/0} & \mbox{\/\ $\ds\frac{iq}{4\pi\hbar}$}\nonumber\\
113: \mbox{\/ $-\ds\frac{iq}{4\pi\hbar}$} & \mbox{\/0}%\nonumber
114: \end{array} \right)\mbox{.}
115: %\end{eqnarray}
116: \ee
117: %\be
118: %C_{00} = C_{11} = 0; \qquad
119: %C_{01} = - C_{10} = \frac{iq}{4 \pi \hbar} ,
120: % \ee
121: %\noindent
122: Starting from the expression (2) we arrive at the following results for the density-density response function $ \Pi_{00(i)} ({\bf q}, \omega ):$
123: \bea %f4
124: && \Pi_{00(i)} ({\bf q}, \omega ) =
125: \Pi_{(i)} ({\bf q}, \omega)
126: \nn \\ \nn\\&
127: =& \frac{K_{00 (i)}^0 ({\bf
128: q,}\omega)}{\displaystyle 1 -
129: \frac{8i\pi\hbar}{q}
130: K_{01(i)}^0 ({\bf q,}\omega) -
131: \bigg(\frac{4\pi\hbar}{q} \bigg)^2
132: \Delta_{(i)} ({\bf q,}\omega)}
133: . \eea
134: Here
135: \be %f5
136: \Delta_{(i)}({\bf q,}\omega)
137: = K_{00(i)}^0 ({\bf q,}\omega) K_{11(i)}^0
138: ({\bf q,}\omega) + \big(K_{01(i)}^0 ({\bf q,}\omega)\big)^2.
139: \ee
140: Within the RPA response functions included in Eqs. (4), (5) are simply
141: related to components of the CF conductivity tensor $ \tilde \sigma $ \cite{eight}:
142: \bea %f6
143: \frac{1}{\tilde \sigma_{xx}^{(i)} ({\bf q,}\omega)} &=& \frac{iq^2}{\omega
144: e^2} \left [\frac{1}{K_{00(i)}^0 ({\bf q,}\omega)} -\frac{1}{K_{00(i)}^0
145: ({\bf q,}0)} \right];
146: \nn \\ \nn\\
147: {\tilde \sigma_{yy}^{(i)}({\bf q,}\omega)} &=& -\frac{i e^2}{\omega}
148: \bigg [{K_{11(i)}^0 ({\bf q,}\omega)} - {K_{11(i)}^0 ({\bf
149: q,}0)} \bigg ];
150: \nn \\\nn\\
151: {\tilde \sigma_{xy}^{(i)} } &=& - {\tilde \sigma_{yx}^{(i)} } =
152: \frac{ie^2}{q} {K_{01(i)}^0 ({\bf q,}\omega)}.
153: \eea
154:
155: To proceed we calculate the components of the CF conductivity at $ \nu $
156: slightly away from $ 1/2 .$ In this case CFs experience a nonzero
157: effective magnetic field $ B_{eff} = B - B_{1/2}.$ We concentrate on the
158: ballistic contribution to the transresistivity, so we need asymptotics for
159: the relevant conductivity components applicable in a nonlocal $ (ql \gg 1)
160: $ and high frequency $ (\omega \tau \gg 1) $ regime. Corresponding
161: expressions for $ \tilde \sigma_{ij} $ were obtained in earlier works
162: \cite{eight}. However, these results are not appropriate for our analysis
163: for they do not provide a smooth passage to the $ B_{eff} \to 0 $ limit at
164: finite $ \bf q.$ Due to this reason we do not use them
165: in further calculations. To get a suitable
166: approximation for the CF conductivity we start from the standard solution
167: of the Boltzmann transport equation for the CF distribution function. This
168: gives us the following results for the CF conductivity components for a
169: single layer \cite{nine}:
170: \bea %f7
171: \tilde \sigma_{\alpha \beta} &=& \frac{m^* e^2}{(2\pi \hbar)^2}
172: \frac{1}{\Omega} \int_0^{2\pi} d \psi v_\alpha (\psi) \exp \left [\! -\frac{iq}{\Omega} \int_0^\psi v_x(\psi'')d \psi'' \right]
173: \nn\\\nn\\ &&
174: \times \int_{-\infty}^\psi v_\beta (\psi') \exp \left
175: [\frac{iq}{\Omega} \int_0^{\psi'} v_x (\psi'') d \psi'' \right.
176: \nn\\\nn\\ &&
177: \left. +\frac{1}{\Omega \tau} (\psi' - \psi)(1 - i \omega \tau)\right ] d\psi'.
178: \eea
179: Here, $ m^*, \ \Omega $ are the CF effective mass and the cyclotron
180: frequency at the effective magnetic field $ B_{eff}; \ \psi $ is the
181: angular coordinate of the CF cyclotron orbit. Now we carry out some formal
182: transformations of this expression (7) following the way proposed
183: before \cite{nine,ten}. First, we expand the CF velocity components $ v_\beta (\psi') $ in Fourier series:
184: \be %f8
185: v_\beta (\psi') = \sum_k v_{k\beta} \exp (i k \psi').
186: \ee
187: Substituting this expansion (8) into (7) we obtain:
188: \bea %f9
189: \tilde \sigma &=& \frac{m^* e^2}{(2 \pi \hbar)^2} \sum_k v_{k\beta}\int_0^{2\pi} d \psi v_\alpha (\psi) \exp (ik\psi)
190: \nn\\\nn\\ &&
191: \times \int_{-\infty}^0 \exp \bigg [\big(ik\Omega - i\omega +
192: \frac{1}{\tau} + iqv_x (\psi)\big) \theta
193: \nn\\\nn\\ &&
194: + iq \int_0^\theta \big(v_x (\psi + \Omega \theta') - v_x
195: (\psi)\big) d\theta' \bigg ] d \theta
196: \eea
197: where $ \theta = (\psi' - \psi)/\Omega. $
198:
199: Then we introduce a new variable $ \eta $ which is related to the
200: variable $ \theta $ as follows:
201: \bea %f10
202: \eta & =& \big(ik \Omega - i\omega + \frac{1}{\tau} + iqv_x
203: (\psi) \big) \theta
204: \nn\\&&
205: + iq \int_0^\theta \big [v_x (\psi + \Omega \theta') - v_x (\psi) \big ] d \theta' ,
206: \eea
207: and we arrive at the result:
208: \bea %f11
209: \tilde \sigma_{\alpha\beta} &=& \frac{i m^* e^2}{(2\pi\hbar)^2} \sum_k v_{k\beta} \int_{-\infty}^0 e^\eta d \eta
210: \nn\\\nn\\ \!\! &&\!
211: \times \!\int_0^{2\pi} \!\!\frac{v_\alpha (\psi) \exp (ik\psi)}{\omega + i/\tau - k \Omega - qv_x (\psi + \Omega \theta)} d \psi
212: . \eea
213: Under the conditions of interest $\omega \tau \gg 1, \ ql \gg 1,$ and also assuming that the filling factor is close to $ \nu = 1/2,$ so that $ qv_F \gg \Omega \ (v_F $ is the CFs Fermi velocity), the variable $ \theta $ is
214: approximately equal to $ \eta \tau (1 + iql \cos \psi + i k\Omega \tau - i \omega \tau)^{-1}.$ Taking this into account and expanding the last term in the denominator of (11) in powers of $ \Omega \theta $ we obtain:
215: \bea %f12
216: && q v_x (\psi + \Omega \theta)
217: \nn\\
218: &\approx& \!\!q v_x (\psi) + \eta \Omega q \tau (1 + iql
219: \cos \psi + i k \Omega \tau - i \omega \tau)^{-1} \frac{dv_x}{d \psi}
220: \nn\\\nn\\ &
221: +& \!\! q\frac{\eta^2}{2}(\Omega \tau)^2 (1 + i q l \cos \psi + i k \Omega \tau
222: - i \omega \tau)^{-2} \ \frac{d^2 v_x}{d \psi^2}. \,
223: \eea
224: Substituting this asymptotic expression into (9) we can calculate first terms of the expansions of relevant components of the CF conductivity in powers of the small parameter $ (qR)^{-1} $ where $ R = v_F/\Omega $ is the CF cyclotron radius. Within the "collisionless" limit $ 1/\tau \to 0 $ we have:
225: \bea %f13-14
226: \tilde \sigma_{xx} & =& - N \frac{i\omega}{q^2} e^2 \left \{1 + \frac{i \delta}{\sqrt{1 - \delta^2}}
227: + \frac{ i \delta}{\sqrt{(1 - \delta^2)^5}} \right.
228: \nn\\\nn\\
229: && \times \left.
230: \frac{1}{2(qR)^2} \bigg (1 - \frac{5}{4} \, \frac{1}{1 - \delta^2}
231: \bigg) \right \};
232: \\\nn\\
233: \tilde \sigma_{yy} &=& N \frac{v_F e^2}{q} \left \{\sqrt{1 - \delta^2} + i \delta + \frac{1}{2 (qR)^2} \right.
234: \nn\\\nn\\ && \left.
235: \times \left [\frac{7}{4}
236: \frac{1}{\sqrt{(1 -\delta^2)^5}} -
237: \frac{1}{\sqrt{(1 -\delta^2)^3}} \right ] \right \};
238: \\ \nn\\
239: \tilde \sigma_{xy} & =& i N \frac{v_F e^2}{q} \frac{\delta}{2 qR} \bigg [\frac{1}{\sqrt{1 - \delta^2}} + \frac{\delta^2}{\sqrt{(1 - \delta^2)^3}}
240: \bigg ].
241: \eea
242: Here, $ N = m^*/2\pi \hbar^2 $ is the density of states at the CF Fermi
243: surface, and $ \delta = \omega /q v_F.$ Using these results we can easily
244: get approximations for the functions $ K_{\alpha \beta (i)}^0 ({\bf q},
245: \omega) \ (\alpha, \beta = 0.1)$ and, subsequently, the desired
246: density-density response function given by (4). It was shown \cite{three}
247: that the integral over $ \omega $ in the expression for $ \rho_D $ (1) is
248: dominated by $ \omega \sim T,$ and the major contribution to the integral
249: over $q$ in this expression comes from $ q \sim k_F (T/T_0 )^{1/3} ,$
250: where $k_F$ is the Fermi wave vector and the scaling temperature $ T_0 $
251: is defined below. Therefore we get an estimate for $ \delta, $ namely $
252: \delta \sim (T/\mu ) (T_0 /T)^{1/3}, $ where $ \mu $ is the chemical
253: potential of a single 2DEG included in the bilayer. For the parameter $ T_0
254: $ taking on values of the order of room temperature, $\delta $ is small compared to unity at low temperatures $
255: (T \sim 1 K).$
256:
257:
258: Here, we limit ourselves to the case of two identical
259: layers $ (\Pi_{(1)} = \Pi_{(2)} \equiv \Pi) $. %+2
260: For $ \delta \ll 1$ we obtain the approximation:
261: \bea % f16
262: && \Pi_{00} ({\bf q}, \omega )
263: \nn\\ \nn \\&
264: = &\frac{q^3}{\displaystyle q^3 \Big (\frac{d n}{d \mu}\Big)^{-1} - 8 \pi i \hbar \omega k_{F} \Big(1 + 2(k_{F} R)^{-1} +
265: \frac{3}{8} (q R)^{-2} \Big)}.\nn\\
266: \eea
267: Here, $ dn/d\mu $ is the compressibility of the $ \nu = 1/2 $ state which
268: is defined as \cite{three}:
269: % f17
270: \be
271: \frac{dn}{d \mu} \equiv \Pi_{00} ({\bf q} \to 0; \ \omega \to 0 ) =\frac{3 m^*}{8 \pi \hbar^2 }.
272: \ee
273: This differs from the compressibility of the noninteracting 2DEG in the
274: absence of an external magnetic field (the latter equals N). The
275: difference in the compressibility values is a manifestation of the
276: Chern-Simons interaction in strong magnetic fields.
277:
278: In following calculations we adopt the expression used in the work
279: \cite{three} for the screened interlayer potential $ U ({\bf q,}\omega),$
280: namely:
281: \bea %f18
282: U ({\bf q,}\omega) &=& \frac{1}{2}\ \frac{V_b + U_b}{1 + \Pi ({\bf q,}\omega) (V_b + U_b)}
283: \nn\\ \nn\\ &&
284: -\frac{1}{2}\ \frac{V_b - U_b}{1 + \Pi ({\bf
285: q,}\omega) (V_b - U_b)}
286: \eea
287: where $\ V_b {\bf (q)} = 2 \pi e^2/\epsilon q $
288: and $\ U_b {\bf (q)} = (2 \pi e^2/\epsilon q) e^{-q
289: d} $ are Fourier components of the bare Coulomb potentials for
290: intralayer and interlayer interactions, respectively, and $ \epsilon $ is
291: the dielectric constant. Substitung (18) into (1) and using our result
292: (16) for $ \Pi \bf (q, \omega) $ we can present the transresistvity in the
293: "ballistic" regime as:
294: %f19
295: \be
296: \rho_D = \rho_{D0} + \delta \rho_D.
297: \ee
298: Here, the first term $ \rho_{D0} $ is the transresistivity at $ \nu =
299: 1/2 $ when the effective magnetic field is zero, and the second term gives
300: a correction arising in a nonzero effective magnetic field (away from $
301: \nu = 1/2 $). As it was to be expected, our expression for $ \rho_{D0} $
302: coincides with the already known result \cite{three}:
303: \be %f20
304: \rho_{D0} = \frac{h}{e^2} \frac{\Gamma (7/3) \zeta (4/3)}{3 \sqrt 3}
305: \bigg(\frac{T}{T_0} \bigg)^{4/3}
306: \ee
307: with $ T_0 = \big(\pi e^2 n d/\epsilon\big) (1 + \alpha),$
308: and
309: %f21
310: \be
311: \displaystyle{\frac{1}{\alpha} = \frac{2 \pi e^2 d}{\epsilon}
312: \frac{dn}{d \mu}} .
313: \ee
314: The leading term of the correction $ \delta \rho_D $ at low temperatures $ (T/T_0)\ll 1 $ can be writen as follows:
315: \bea %f22
316: \delta \rho_D &\!
317: = \!&\! \frac{2}{3} \rho_{D0} \frac{1}{k_F R} \left(1 +
318: \frac{3}{8} \frac{1}{k_FR}\right)
319: + a^2 \frac{h}{e^2} \left(\frac{2T}{T_0} \right)^{2/3}\!\!
320: \frac{1}{(k_F R)^2}
321: \nn\\\nn\\
322: &&\!\!\approx \frac{4}{3} \rho_{D0} \Delta \nu \left(1 + \frac{3}{4} \Delta
323: \nu \right) + 4a^2 \frac{h}{e^2} \left( \frac{2T}{T_0}
324: \right)^{2/3}\!\! (\Delta \nu)^2 . \nn\\
325: \eea
326: Here, the dimensionless positive constant $ a^2 $ can be approximated as:
327: \be %f23
328: a^2 = \frac{7}{24\sqrt 3} \int_0^\infty \bigg(\frac{y^{2/3}}{\sinh^2 y} - \frac{1}{y^{4/3} \cosh^2 y} \bigg) dy.
329: \ee
330:
331: We have to remark that our result (23) cannot be used in the limit $ T \to
332: 0. $ Actually, this expression provides a good asymptotic form for the
333: coefficient $ a^2 $ when $ (T k_F l / \mu )^{1/3} \geq 1.5 .$ Assuming
334: that the mean free path is of the order of $ 1.0 \mu m $ as in the
335: experiments \cite{eleven} on dc magnetotransport in a single modulated
336: 2DEG at $ \nu $ close to $1/2,$ and using the estimate of \cite{seven} for
337: the electron density $ n = 1.4 \times 10^{15} m^{-2} ,$ we obtain that the
338: expression (23) gives good approximation for $ a^2 $ when $ T/\mu $ is no less than $10^{-2}$.
339:
340: It follows from our results (19), (22) that transresistivity $
341: \rho_D $ enhances nearly quadratically with $ \Delta \nu $ when the
342: filling factor deviates from $ \nu = 1/2 .$ The linear in $\Delta \nu $
343: term is also present in the expression for $ \delta \rho_D. $ This causes
344: an asymmetric shape of the plot of Eq. (22) relative to $ \Delta \nu = 0.
345: $ However, this asymmetry is not very significant for the linear term is
346: smaller than the last term on the right hand side of (22). This difference
347: in magnitudes is due to different temperature dependences of the
348: considered terms. The first term including the linear in $ (k_F
349: R)^{-1}$ correction is proportional to $ (T/T_0)^{4/3},$ whereas the
350: second one is proportional to $ (T/T_0)^{2/3}$ and predominates at low
351: temperatures. %+3
352: So, the magnetic field dependence of the transresistivity near $\nu = 1/2
353: $ matchs that observed in the experiments (See Fig. 1).
354:
355:
356: \begin{figure}[t]
357: \begin{center}
358: \includegraphics[width=7.8cm,height=8.7cm]{jetp04_1.eps}
359: \caption{
360: Scaled drag resistivity versus $ \Delta \nu $ at T = 0.6; lowest dashed curve is the plot of Eq. (22) at $ m^* = 4 m_b; \ A_0 = 15, $ and remaining curves present experimental data of [7];
361: }
362: \label{rateI}
363: \end{center}
364: \end{figure}
365:
366:
367:
368:
369: Keeping only the greatest term in (22), the ratio $ \rho_D /\rho_{D0}$ can
370: be presented in the form:
371: %f24
372: \be
373: \frac{\rho_D}{\rho_{D0}} = 4 \beta (\Delta \nu)^2 + 1.
374: \ee
375: Here, the coefficient $ \beta $ equals:
376: %f25
377: \be
378: \beta = \frac{3 \sqrt 3 a^2} {\Gamma(7/3) \zeta(4/3)}
379: \left(\frac{2T_0}{T} \right)^{2/3}.
380: \ee
381: This coefficient is proportional to the curvature of the plot of Eq. (22)
382: assuming that the first term is neglected. The curvature reveales a
383: strong dependence on temperature whose character also agrees with
384: experiments of \cite{seven} as it is shown in Fig. 2.
385:
386:
387: \begin{figure}[t]
388: \begin{center}
389: \includegraphics[width=8cm,height=8.4cm]{jetp04_2.eps}
390: \caption{Temperature dependence of the coefficient $ \beta^{-1} $ for
391: interlayer distances $ d = 10 nm $ (upper curve) and $ d = 22.5 nm $
392: (lower curve) compared to the summary of experimental curvature at both
393: spacings [7]}
394: \label{rateI}
395: \end{center}
396: \end{figure}
397:
398:
399:
400: A striking feature in the experimental results is that they appear to be
401: nonsensitive to the distance between the 2DEGs. Sets of data corresponding
402: to samples with different interlayer spacings $ d_A = 10 nm$ and $ d_B =
403: 22.5 nm$ fall on the same curve. This concerns both magnetic field
404: dependence of the transresistivity and temperature dependence of the
405: parameter $\beta$. Results of the present analysis provide a possible
406: explanation for this feature. It follows from (20)--(25) that the
407: dependence of $ \rho_D$ of the interlayer spacing is completely included
408: in the characteristic temperature $ T_0 $ which is defined with Eq. (21).
409: The above quantity is nearly independent of the interlayer separation $ d
410: $ when the parameter $ \alpha $ takes on values larger that unity.
411: Estimating the parameter $ \alpha $ as it is
412: given by Eq. (21), we obtain that the condition $ \alpha > 1 $ could be
413: satisfied for small values of the compressibility of the $ \nu = 1/2 $
414: state. %+4
415: However, within the RPA the effective mass of CFs coincides with the
416: single
417: electron band mass $ m_b $ which takes on the value $ m_b \approx 0,07 m_e
418: $ for GaAs wells ($ m_e $ is the mass of a free electron). Using this
419: value to estimate the compressibility as it is introduced by Eq. (17)
420: we get $ \alpha \approx 0.44. $ This is too small to provide insensitivity
421: of the coefficient $ \beta $ determined by Eq. (25) to the interlayer
422: distance for interlayer spacings reported in the experiments
423: \cite{three}. The above discrepancy
424: could be removed taking into account Fermi liquid interactions among
425: quasiparticles (CFs). To include Fermi liquid effects into consideration
426: we write the renormalized polarizability $ \Pi^* $ in the form
427: \cite{eight}:
428: % f26
429: \be
430: \Pi^{* -1} = \Pi^{-1} + F_{(0)} + F_{(1)}.
431: \ee
432: Here, $\Pi $ is the polarizability of noninteracting CFs defined with Eq.
433: (2), and the remaining terms present contributions arising due to Fermi
434: liquid interaction in the CF system. Only contributions from the first and
435: greatest two terms in the expansion of the Fermi liquid interaction
436: function in Legendre polynomials ($f_0$ and $ f_1$, respectively) are kept
437: in Eq. (26) to avoid too lengthy calculations. Matrix elements of the $
438: 2\times 2 $ matrices $ F_{(0)}$ and $ F_{(1)}$ equal:
439: \bea % f27
440: F_{(0)} &= &
441: \left( \begin{array}{cc}
442: f_0 & 0 \nn \\
443: 0 & 0
444: \end{array}
445: \right)
446: \\ \\
447: F_{(1)} & = &
448: \left( \begin{array}{cc}
449: {\ds \frac{m^* - m_b}{ne^2} \frac{\omega^2}{q^2} } &
450: 0 \nn\\
451: 0 & {-\ds \frac{m^* - m_b}{ne^2}}\\
452: \end{array} \right) .
453: \eea
454:
455: Within the Fermi liquid theory the effective mass $ m^* $ is related to
456: the "bare" mass $ m_b $ as follows:
457: % f28
458: \be
459: \frac{1}{m_b} = \frac{1}{m^*} + \frac{f_1}{2 \pi \hbar^2} \equiv
460: \frac{1 + A_1 }{m^*}.
461: \ee
462: Using these expressions (26)--(28) and carrying out calculations within
463: the relevant limit $ \delta \ll 1 ,$ we obtain that the expression for the
464: density-density response function for a single layer keeps the form given
465: by Eq. (16) where the compressibility $ dn/d\mu $ is replaced with the
466: quantity $ dn^*/d\mu $ renormalized due to the Fermi liquid interaction:
467: % f30
468: \be
469: \frac{dn^*}{d \mu}
470: = \frac{3m^*}{8 \pi \hbar^2} \bigg (1 + \frac{3m^*}{8 \pi \hbar^2}
471: f_0 \bigg )^{-1} \!\!\equiv \frac{dn}{d\mu} \bigg( 1+ \frac{dn}{d\mu} f_0
472: \bigg )^{-1} \!.
473: \ee
474:
475: For strongly correlated quasiparticles this renormalization may
476: significantly reduce the compressibility of the CF liquid, and,
477: consequently, increase the value of the parameter $ \alpha. $ It is
478: usually assumed \cite{three,eight} that the Fermi liquid renormalization
479: of the effective mass significantly changes its value: $ m^* \sim 5-10 \
480: m_b. $ This gives for the Fermi liguid coefficient $A_1$ values of the
481: order of 10. Using this estimate, and substituting our renormalized
482: compressibility (29) into the expression (21) we arrive at the conclusion
483: that $dn^*/d\mu $ is low enough for the condition $ \alpha > 1, $ to be
484: satisfied when the Fermi liquid parameter $ A_0 \equiv f_0 / 2\pi \hbar^2
485: $ takes on values of the order of $10-100.$ This conclusion does not seem
486: an unrealistic one for it is reasonable to expect $ A_0 $ to be of the
487: order or greater than the next Fermi liquid parameter $ A_1.$
488: We obtain
489: a reasonably good agreement between the plot of our Eq. (22) and the
490: experimental results, using %+6
491: $A_0 = 15$ and $A_1 = 3 \ (m^* = 4 m_b).$ (Fig. 1).
492:
493: Our results for temperature dependence of $ \beta^{-1}$ also agree with
494: the results of experiments \cite{seven}. The upper curve in Fig.2
495: corresponds to the double-layer system with with smaller interlayer
496: spacing $ d_A = 10 nm $ which gives $ T_0 = 487 K,$ and the lower curve
497: exhibits the temperature dependence of $ \beta^{-1} $ for greater spacing
498: $ d_B = 22.5 nm \ (T_0 = 587 K).$ The curves do not coincide but they are
499: arranged rather close to each other.
500:
501:
502:
503: Finally, the results of the present analysis enable us to qualitatively
504: describe all important features observed in experiments of \cite{seven} on
505: the Coulomb drag slightly away from one half filling of lowest Landau
506: levels of both interacting 2DEG. %+5
507: They also give us grounds to treat these experimental results as one
508: more evidence of strong Fermi liquid interaction in the CF system near one
509: half filling of the lowest Landau level. The above interaction provides a
510: significant reduction of the compressibility of the CF liquid and a
511: consequent enhancement in the screening length in single layers.
512: Essentially, the parameter $ \alpha $ characterizes the ratio of the
513: Thomas--Fermi screening length in a single 2DEG at $ \nu = 1/2 $ and the
514: separation between the layers \cite{three}. When $ \alpha > 1, $
515: intralayer interactions predominate those between the layers which could
516: be the reason for low sensitivity of the bilayer to changes in the
517: interlayer spacing. It is likely that here is an explanation for the
518: reported nearly independence of the drag on the interlayer separation
519: \cite{seven}. We believe that at larger distances between the layers the
520: dependence of the transresistivity of $ d $ could be revealed in the
521: experiments. At the same time the results of \cite{seven} give us a
522: valuable opportunity to estimate a strength of Fermi liquid interactions
523: between quasiparticles at $ \nu = 1/2 $ state which is important for
524: further studies of such systems.
525:
526:
527:
528:
529: \vspace{0mm}
530:
531: {\it Acknowledgments:}
532: The author thank K.L. Haglin and G.M. Zimbovsky for help with the manuscript.
533:
534:
535: \begin{thebibliography}{99}
536:
537: \bibitem{one} L. Zheng and A.H. MacDonald,
538: Phys. Rev. B {\bf 48}, 8203
539: (1993).
540:
541: \bibitem{two} A. Kamenev and Y. Oreg,
542: Phys. Rev. B {\bf 52}, 7516 (1995).
543:
544: \bibitem{three} I. Ussishkin and A. Stern,
545: Phys. Rev. B {\bf 56}, 4013 (1997).
546:
547:
548: \bibitem{four} S. Sakhi,
549: Phys. Rev. B {\bf 56}, 4098 (1997);
550:
551:
552: \bibitem{five} M.P. Lilly, J.P. Eisenstein, L.N. Pfeiffer and K.W. West,
553: Phys. Rev. Lett. {\bf 80}, 1714 (1998).
554:
555:
556: \bibitem{six} When the external driving disturbance applied to one of the
557: layers is of small $ {\bf q,}\omega \ (ql \ll 1, \omega \tau \ll 1)$ the
558: transresisitivity is dominated with the diffusion contribution, and new
559: effects could emerge (See e.g. F. von Oppen, S.H. Simon and A. Stern,
560: Phys. Rev. Lett. {\bf 87}, 106803 (2001) and references therein).
561:
562:
563: \bibitem{seven} M.P. Lilly, J.P. Eisenstein, L.N. Pfeiffer and K.W. West,
564: cond-mat/9909231.
565:
566: \bibitem{eight} B.I. Halperin, P.A. Lee and N. Read,
567: Phys. Rev. B {\bf 47}, 7312 (1993);
568: S.H. Simon and B.I. Halperin,
569: Phys. Rev. B {\bf 48}, 17368 (1993).
570:
571:
572: \bibitem{nine} N.A. Zimbovskaya and J.L. Birman, Phys. Rev. B, {\bf 60}
573: 16762 (1999).
574:
575: \bibitem{ten} N.A. Zimbovskaya, "Local Geometry of the Fermi Surface and
576: High-Frequency Phenomena in Metals", Springer-Verlag, New York, 2001.
577:
578: \bibitem{eleven} R.L. Willett, K.W. West, and L.N. Pfeiffer Phys. Rev.
579: Lett. {\bf 83}, 2624 (1999).
580:
581: \end{thebibliography}
582: %\end{references}
583:
584:
585:
586:
587:
588:
589:
590:
591:
592:
593:
594:
595:
596:
597:
598:
599:
600: %\end{multicols}
601:
602: \end{document}
603:
604:
605:
606:
607:
608: The composite fermion approach is used for theoretical studies of the
609: Coulomb drag between two parallel layers of two-dimensional electron gases
610: in the quantum Hall regime near Landau level filling factor $\nu = 1/2 $.
611: It is shown that as the filling factor deviates from $\nu = 1/2 $ the
612: magnitude of the transresistivity gets an additional term which enhances
613: with the effective magnetic field. This agrees with
614: experimental results. Temperature dependence of the transresistivity near
615: one half filling is also analyzed, and obtained results exhibit a
616: reasonable fit with the experiments.
617:
618:
619:
620:
621: \begin{eqnarray}
622: \mbox{ M\/} &\mbox{= }&
623: \left(
624: \begin{array}{ll}
625: \mbox{\/0} & \mbox{\/\ 1} \nonumber\\
626: \mbox{\/1} & \mbox{\/0}\nonumber
627: \end{array}
628: \right)\mbox{.}
629: \end{eqnarray}
630:
631: \vfill
632:
633:
634:
635:
636:
637: \begin{eqnarray}
638: \mbox{ $F_{(1)}$\/} &\mbox{= }&
639: \left(
640: \begin{array}{cc}
641: \mbox{\/$\ds \frac{(m^* - m_b)}{ne^2} \frac{\omega^2}{q^2} $} &
642: \mbox{\/\ 0} \nonumber\\
643: \mbox{\/0} & \mbox{\/ $ -\ds \frac{(m^* - m_b)}{ne^2}$ }\nonumber
644: \end{array}
645: \right)\mbox{;}
646: \end{eqnarray}
647:
648:
649:
650:
651:
652: \begin{eqnarray}
653: \mbox{ $F_{(1)}$\/} &\mbox{= }&
654: \left(
655: \begin{array}{cc}
656: \mbox{\/$\ds \frac{(m^* - m_b)}{ne^2} \frac{\omega^2}{q^2} $} &
657: \mbox{\/\ 0} \nonumber\\
658: \mbox{\/0} & \mbox{\/ $ -\ds \frac{(m^* - m_b)}{ne^2}$ }\nonumber
659: \end{array}
660: \right)\mbox{;}
661: \end{eqnarray}
662:
663:
664:
665:
666: $$
667: F_{(0)00} = f_0; \quad
668: F_{(0)01} = F_{(0)10} = F_{(0)11} = 0;
669: $$ $$
670: F_{(1)00} = \frac{(m^* - m_b)}{ne^2} \frac{\omega^2}{q^2}; \quad
671: F_{(1)11} = - \frac{(m^* - m_b)}{ne^2} ;
672: $$\be
673: F_{(1)01} = F_{(1)10} = 0.
674: \ee
675:
676:
677:
678:
679:
680:
681:
682:
683:
684:
685:
686:
687:
688:
689:
690:
691:
692:
693:
694:
695:
696:
697:
698:
699:
700:
701:
702:
703:
704:
705:
706:
707:
708:
709:
710:
711:
712:
713:
714:
715:
716:
717:
718:
719:
720:
721:
722:
723:
724:
725:
726:
727:
728:
729:
730:
731:
732:
733:
734:
735:
736:
737:
738:
739:
740:
741:
742:
743: \bibitem{seven} cond-mat/9710041
744:
745: Title: Coulomb Drag in the Extreme Quantum Limit
746: Authors: M.P. Lilly, J.P. Eisenstein, L.N. Pfeiffer, K.W. West
747:
748: